Chimica delle Proteine e Medicina Molecolarepaduaresearch.cab.unipd.it/3378/1/PhD_Thesis.pdf ·...

UNIVERSITA’ DEGLI STUDI DI PADOVA

Dipartimento di Scienze Farmaceutiche

Scuola di Dottorato in Scienze Molecolari

Indirizzo Scienze Farmaceutiche

XXIII° Ciclo

Chimica delle Proteine e

Medicina Molecolare Direttore della Scuola: Ch.mo Prof. Maurizio CASARIN

Coordinatore di indirizzo: Ch.ma Prof.ssa Adriana CHILIN

Supervisore: Ch.mo Prof. Vincenzo DE FILIPPIS

Dottorando: Dr. Fabio MASET

UNIVERSITY OF PADUA

Department of Pharmaceutical Sciences

Ph.D. School in Molecular Sciences

Pharmaceutical Sciences Curriculum

XXIII° Cycle

Protein Chemistry and

Molecular Medicine School Director: Prof. Maurizio CASARIN

Curriculum Coordinator: Prof. Adriana CHILIN

Supervisor: Prof. Vincenzo DE FILIPPIS

Ph.D. Student: Fabio MASET

Non pregate per avere una

vita facile. Pregate piuttosto per

essere persone più forti.

John Fitzgerald Kennedy

Cover Illustration: Jackson Pollock - Enchanted Forest, 1947 Oil on canvas 221,3 x 114,6 cm Peggy Guggenheim Collection, Venice

CONTENTS

RIASSUNTO 1

ABSTRACT 3

1. INTRODUCTION

An introduction to mass spectrometry 7

Ionization methods 8

Electrospray ionization 8

Matrix-assisted laser desorption/ionization 11

Mass analysers 13

The quadrupole mass filter 14

The TOF mass analyzer 15

Protein preparation and purification 17

Special considerations for protein preparation methods 17

Analysis of intact proteins 18

Digestion and preparation of gel-separated proteins for MS analysis 19

Peptide sequencing by tandem mass spectrometry 20

Protein identification by database searching 22

Peptide mass fingerprint 22

Searching with tandem mass spectrometric data 23

Database searching and protein modifications 24

Liquid chromatography and tandem mass 24

Principles 24

Instrumental and practical considerations 26

References 29

2. PROTEASE NEXIN-1 (PN-1)

Conformational and biochemical characterization of a biological active rat

recombinant protease nexin-1 expressed in E.coli 37

Introduction 37

Materials and methods 38

Results 46

Discussion 61

References 63

3. PYTHON SEBAE SERUM

A novel protein from the serum of Python sebae, structurally homologous

to γ-phospholipase A2 inhibitor, displays antitumor activity 71

Introduction 71


Results 80

Discussion 93

References 96

Supplemental data 98

4. ALANINE:GLYOXYLATE AMINOTRANSFERASE (AGT)

Molecular defects of the glycine 41 variants of alanine:glyoxylate aminotransferase

associated with primary hyperoxaluria type I 107

Introduction 107


Results 111

Discussion 119

References 122

Supporting Information .124

5. VON WILLEBRAND FACTOR (VWF)

Serine proteases from primary granules of leukocytes efficiently cleave oxidized

von willebrand factor: divergence from ADAMTS-13 137

Introduction 137


Results 143

Discussion 152

References 155

APPENDICES

Appendix A. Abbreviations and symbols 163

Appendix B. Amino acids nomenclature 164

1

RIASSUNTO

La proteomica riguarda lo studio sistematico delle proteine al fine di fornire una visione

completa della funzione, della struttura e della regolazione dei sistemi biologici. I progressi

avvenuti negli ultimi decenni, sia per quanto riguarda la strumentazione sia le metodologie

utilizzate, hanno permesso di ampliare il campo di studi biologici passando dall’analisi di proteine

purificate all’analisi di miscele complesse. La proteomica sta rapidamente diventando una

componente essenziale della ricerca biologica ed associato ai progressi della bioinformatica, questo

approccio alla descrizione dei sistemi biologici avrà indubbiamente un impatto notevole sulla nostra

comprensione dei fenotipi sia delle cellule normali e malate.

Inizialmente la proteomica era focalizzata principalmente sulla generazione di mappe proteiche

bidimensionali utilizzando elettroforesi su gel di poliacrilammide. La verifica dell’espressione o la

misurazione quantitativa dei livelli globali di proteine può ancora essere fatta sulla base dei gel

bidimensionali, tuttavia oramai questi compiti sono affidati alla spettrometria di massa la quale può

contare su di un’elevata sensibilità e specificità. La spettrometria di massa applicata alle proteine

offre molti vantaggi: oltre a calcolare il peso molecolare con elevata precisione, questa tecnica

permette di analizzare e caratterizzare la sequenza aminoacidica. Può anche essere utilizzata nello

studio delle modificazioni post-traduzionali e per monitorare la formazione di complessi in

soluzione. Infine può essere applicata con differenti scopi, quali l'analisi conformazionale, l'analisi

della cinetica di ripiegamento e di studi sulle attività catalitiche delle proteine.

Durante il dottorato di ricerca la mia attenzione è stata focalizzata soprattutto sull’utilizzo di tale

tecnica abbinata a metodologie di chimica delle proteine quali ad esempio l’elettroforesi mono e

bidimensionali, differenti cromatografie in fase liquida, la sintesi peptidica in fase solida e l’utilizzo

di proteasi enzimatiche. In particolare in questa Tesi di Dottorato gli argomenti di studio sono stati

trattati singolarmente, distinguendo i principali progetti in cui sono stato coinvolto in capitoli

indipendenti. Brevemente, nel capitolo 2 è proposto lo studio di protease nexin-1 (PN-1), il

principale inibitore della trombina a livello cerebrale, volto a chiarire la funzione della porzione

glucidica sulla conformazione, stabilità e funzione della proteina mediante lo studio della proteina

ricombinante prodotta in E. coli. Nel capitolo 3 è riportato il lavoro concernente la purificazione e la

caratterizzazione chimica, in particolare dell’identificazione de novo della sequenza amminoacidica,

di un analogo dell’inibitore della fosfolipasi A2 estratto dal siero di Python sebae, il quale ha

dimostrato di possedere un effetto citotossico pro-apoptotico e che potrebbe essere sfruttato per lo

sviluppo di nuove strategie antitumorali. Nel capitolo 4 l’attenzione è stata concentrata a chiarire le

2

dinamiche molecolari che portano allo sviluppo di iperossaluria primaria di tipo I mediante lo studio

del mutante G41R dell’enzima alanina:gliossilato amminotransferasi (AGT) analizzando in

particolar modo i meccanismi che portano G41R ad essere maggiormente soggetto a degradazione e

aggregazione rispetto alla proteina WT. Infine, il capitolo 5 tratta dell’effetto dello stress ossidativo

sul metabolismo del fattore di von Willebrand (VWF). Il fattore di von Willebrand è una

glicoproteina plasmatica estremamente complessa le cui dimensioni contribuiscono a regolare

l’equilibrio emostatico. Nello specifico, è stato osservato come l’ossidazione di un residuo di

metionina situato nel dominio A2 della glicoproteina impedisca il taglio proteolitico da parte di

ADAMTS-13, mentre non vada ad influenzare o in alcuni casi addirittura favorisca la proteolisi di

VWF da parte di proteasi leucocitarie liberate dai polimorfonucleati in seguito a stati infiammatori.

3

ABSTRACT

Proteomics involves the systematic study of proteins in order to provide a comprehensive view

of the structure, function and regulation of biological systems. Advances in instrumentation and

methodologies have fueled an expansion of the scope of biological studies from simple biochemical

analysis of single proteins to measurements of complex protein mixtures. Proteomics is rapidly

becoming an essential component of biological research. Coupled with advances in bioinformatics,

this approach to comprehensively describing biological systems will undoubtedly have a major

impact on our understanding of the phenotypes of both normal and diseased cells.

Initially, proteomics was focused on the generation of protein maps using two-dimensional

polyacrylamide gel electrophoresis. Protein expression, or the quantitative measurement of the

global levels of proteins, may still be done with two-dimensional gels, however, mass spectrometry

has been incorporated to increase sensitivity, specificity and to provide results in a high-throughput

format.

Mass spectrometry applied to proteins offers many advantages: in addition to calculating the

molecular weight with high accuracy, this technique allows to analyze and characterize the amino

acid sequence. It can also be used in the study of post-translational modifications and to monitor the

formation of complexes in solution. Finally it can be applied with different purposes such as the

conformational analysis, analysis of the kinetics of refolding and studies on the catalytic activities

of proteins.

During my Ph.D. course my efforts have been mainly devoted on the use of this technique in

combination with methods of protein chemistry such as the one- and two-dimensional

electrophoresis, differents liquid chromatographies, solid phase peptide synthesis and use of

proteolytic enzymes. Herein reported in my Ph.D. Thesis, the different treated subjects are divided

into independent chapters each containing a single case study. Briefly in chapter 2 it has been

proposed the study on protease nexin-1 (PN-1), the main inhibitor of thrombin in the brain, aimed at

clarifying the role of the carbohydrate portion on the conformation, stability and function, through

studies on recombinant protein produced in E.coli. In chapter 3 it has been showed the work on

purification and chemical characterization, including de novo identification of amino acid sequence,

of a similar inhibitor of phospholipase A2 extracted from the Python sebae serum, which

demonstrated an effect cytotoxic pro-apoptotic and that could be exploited for the development of

new anticancer strategies. In chapter 4 it has been reported the molecular dynamics that lead to the

development of primary hyperoxaluria type I by studying the G41R mutant enzyme alanine:

4

glyoxylate aminotransferase (AGT). In particular, were investigated the mechanisms leading to

G41R be more susceptible to degradation and aggregation than wild type protein. Finally, chapter 5

deals with the effect of oxidative stress on the metabolism of von Willebrand Factor (VWF). The

von Willebrand Factor is a very complex plasma glycoprotein whose dimensions help to regulate

the hemostatic balance. Specifically, in the study was observed as the oxidation of a methionine

residue located in the A2 domain of glycoprotein prevents proteolytic cleavage by ADAMTS-13,

while not going to influence or, in some cases, promote proteolysis of VWF by proteases released

from polymorphonuclear leukocytes in inflammatory conditions.

1. INTRODUCTION

1.Introduction

7

CHAPTER 1

An Introduction to Mass Spectrometry

Mass spectrometry (MS) is an analytical technique whose beginnings date back to the early days

of the last century. Among the analytical techniques, MS holds a special place because it measures

an intrinsic property of a molecule, its mass, with very high sensitivity and therefore it is used in an

amazingly wide range of applications. Beginning in the 1980s and on a larger scale in the 1990s,

mass spectrometry has played an increasingly significant role in the biological sciences. Why has it

taken so long? Mainly because mass spectrometers require charged, gaseous molecules for analysis.

Biomolecules being large and polar, however, they are not easily transferred into the gas phase and

ionized. Electrospray (ES) (1) and matrix-assisted laser desorption ionization (MALDI) (2) are the

ionization techniques that should be credited most for the success of mass spectrometry in the life

sciences. These methods were developed in the late 1980s and were the basis for the increasingly

powerful instrumentation that became available a few years later. Major advances were also made

in sample preparation for MS, a crucial area for overall feasibility and sensitivity of analysis.

Starting in 1993, software algorithms were published that allowed the correlation of mass

spectrometric data obtained for a protein with the increasingly populated sequence databases. In

retrospect, this event marked the transformation of mass spectrometry into a largescale, functional

genomics technique. The last few years have seen development of even more powerful

instrumentation and algorithms for protein characterization, a trend that shows no signs of slowing

down.

At the same time MS was being developed to meet the demands of molecular biology for high

sensitivity, the concept of proteomics began to be popularized. Proteomics is now defined as the

large-scale analysis of the function of genes and is becoming a central field in functional genomics

(3). The major tool to study purified proteins in this field is mass spectrometry. The history of

proteomics dates back to the discovery of two-dimensional gels in the 1970s, which provided the

first feasible way of displaying hundreds or thousands of proteins on a single gel (4, 5).

Identification of the spots separated on these gels remained laborious and was limited to the most

abundant proteins until the 1990s, when biological mass spectrometry had developed into a

sufficiently sensitive and robust technique. Today, mass spectrometry is an essential element of the

proteomics field. Indeed, researchers are now successfully harnessing the power of MS to supersede

the two-dimensional gels that originally gave proteomics its impetus.

1.Introduction

8

Currently, the uses of MS in proteomics are in three major areas. MS is the preferred technique

for characterization and quality control of recombinant proteins and other macromolecules, an

important task in the field of biotechnology. It is also commonly used for protein identification,

either in classical biochemical projects or in large-scale proteomic ones. Finally, because MS

measures the molecular weight of a protein, it is the method of choice for the detection and

characterization of posttranslational modifications and potentially can identify any covalent

modification that alters the mass of a protein.

Ionization methods

Mass spectrometry measures the mass-to-charge ratio of ionised molecules in the gas phase.

Hence, the analytes need to be ionised and transferred to the gas phase prior to analysis. Earlier

techniques such as electron impact (EI) (6) and chemical ionisation (CI) (7) were efficient in

ionising small volatile thermally stable molecules. Fast-atom bombardment (FAB) was introduced

in the early 1980s by Barber and co-workers (8) and was the first ionisation technique that could be

used for routine analysis of biomolecules of mass up to a few thousand daltons (9). The soft

ionisation techniques, MALDI and electrospray, are the revolutionising methods that have made

mass spectrometry one of the most important tools to analyse large biomolecules.

Electrospray ionization

Electrospray is an atmospheric-pressure ionisation method that produces small charged droplets

from a liquid medium under the influence of an electric field. The process itself originates from the

beginning of the last century but it was Dole and co-workers (10) who first transferred large

molecules to the gas phase in the late 1960s. In 1984, electrospray was used for the first time to

create gas phase ions to be analysed by a mass spectrometer. The group of J. B. Fenn reported the

use of electrospray mass spectrometry (11,12) at approximately the same time as Alexandrov and

co-workers (13) but it was still a few more years before the first electrospray mass spectra of large

molecules were published, again from the group of Fenn (14,15).

As Alexandrov et al. (13) reported 1984, electrospray is a very suitable method for combining

chromatographic methods with mass spectrometry. This is because the electrospray process

transfers ions from the solution phase to the gas phase at atmospheric pressure. In conventional

electrospray, a flow of liquid, from a chromatographic system or a syringe pump, is passed through

1.Introduction

9

a thin conducting needle at high voltage (3–4 kV). The potential difference is applied between the

needle and the counter electrode (the inlet of the mass spectrometer). The analytes will to some

extent, depending on the pH of the solvent, exist in an ionised form in the liquid and the applied

potential will create an accumulation of like charges at the tip of the needle. Positive or negative

ions will migrate to the end of the capillary depending on the polarity of the applied field. Figure 1

illustrates the electrospray process in the positive-ion mode, where positively charged ions

accumulate at the liquid surface (meniscus). Ions of opposite polarity (negatively charged ions) will

migrate towards the positive capillary wall. The high density of positive charges at the tip leads to

the formation of a Taylor cone (16) due to the repulsive coulombic forces between the positive ions.

Under a high electric field, the repulsive forces become stronger than the surface tension at the tip

of the cone, and a liquid mist is ejected. The mist breaks up in small highly charged droplets when

moving towards the counter electrode (Fig. 1).

Figure 1. Schematic representation of the electrospray process. A positive potential is applied to the capillary, causing positiveions in solution to drift towards the meniscus. The meniscus destabilises, leading to the formation of a Taylor cone emitting droplets with excess positive charge. Gas phase ions are formed from charged droplets in a series of solvent evaporation cycles as they are accelerated towards the entrance of the mass spectrometer.

Both a potential and a pressure gradient will direct the droplets towards the inlet of the mass

spectrometer and a counter-current flow of gas facilitates solvent vaporisation and prevents ion

cluster formation. As the solvent evaporates, the charge remains constant in the droplets and at a

certain point, the surface coulombic forces exceed the surface tension forces, the Rayleigh limit

(17), and the droplets break up into smaller droplets (18). This process continues until droplets with

diameters in the nanometer range are generated. The formation of gas phase ions from the small

charged droplets is not yet fully understood. Two mechanisms have been proposed. The original

idea was that the solvent evaporates and droplets break up until those with only a single analyte ion

1.Introduction

10

are created (10). The evaporation continues until a gas phase ion is formed. This is usually referred

to as the charged-residue model. Iribarne and Thomson (19) suggested an alternative mechanism in

1976 (the ion evaporation model), in which they proposed that droplets with a radius less than 10

nm can allow field desorption, i.e. direct emission of a gaseous ion. The charge state of the ion will

depend on the number of charges that are transferred from the droplet surface to the ion during

desorption. The gas phase ion formation processes are still under debate, and while the ion

evaporation theory might be the most accepted, a mechanism related to the charged-residue process

may account for the formation of gaseous protonated macromolecules (20). A number of useful

papers and volumes have been published in which detailed descriptions of the electrospray process

are discussed (21-24).

The electrospray ion source has gone through major developments since its introduction.

Conventional electrospray instruments operate best at a flow rate of 3–10 µl/min, but the coupling

of liquid chromatography (LC) to mass spectrometry has sometimes demanded flow rates of up to 1

ml/min. The electrospray evaporation is then facilitated by a coaxial gas flow (a nebulising gas)

(25). This type of source is generally called pneumatically assisted electrospray. On the other hand,

the coupling of capillary electrophoresis (CE) to mass spectrometry gives very low flow rates and a

sheath flow of organic solvent might be needed. However, this process dilutes the sample solution

and sheathless interfaces have been developed (26). Like all other analytical techniques,

miniaturisation has been one of the key steps in the development of electrospray ionisation. The

first report on low-flow-rate electrospray ionisation came in 1993 (27). The following year, Emmett

and Caprioli (28) followed up with a continuous infusion source operating at a flow rate of

approximately 300–800 nl/min, resulting in a major sensitivity increase. This source is called the

microelectrospray source and neuropeptides have been analysed down to the zeptomole (10–21

mole) level (29). The first electrospray source without continuous infusion was developed by Wilm

and Mann (30,31) and called nanospray. The sample is sprayed from a metal-coated capillary

(needle) with an opening of 1–10 µm. This results in a flow rate of 20–50 nl/min and, hence, small

sample volumes are consumed. Less than 0.5 µl can be loaded into the needle and used for more

than 15 min in favourable situations. This time frame is usually enough for several MS/MS

experiments. The flow rate will depend on the orifice diameter of the needle, the applied voltage

(usually < 1 kV) as well as the viscosity and volatility of the solvent. The nanospray source has

higher ionisation efficiency than a conventional electrospray source due to the production of smaller

droplets, and it is also reportedly less sensitive to salts than conventional electrospray. Another

modification that has made electrospray sources overall more tolerant to inorganic salts is the Z-

spray configuration (32). In this set-up, the spraying device is mounted at a right angle to the

1.Introduction

11

inlet/cone of the mass spectrometer. The idea is that the major part of the inorganic salts will not be

desorbed from the droplets and will hence travel in a straight path and not be analysed, while the

organic ions will be more easily transferred to the gas phase and transported into the mass

spectrometer down the pressure and potential gradients.

Electrospray ionisation produces multiply charged ions, and proteins with molecular weights in

the 10–100 kDa mass range will in general produce an envelope of ions with m/z values below 2500

(Fig. 2). Ion transmission is generally good in this region and mass measurement statistics are

excellent due to many different charge states being observed. These features make electrospray the

most suitable ionization method for molecular-weight determination of large biomolecules. The

envelope of different charge states can be converted to a true mass scale via maximum-entropy

processing (Fig. 2), an iterative method that calculates the mass and the abundance from the

experimental m/z peaks (33) or via other deconvolution or transform processes.

Figure 2. Electrospray mass spectrum of a protein. The attachment of many protons per protein molecule (from less than 30 to more than 50 here) leads to a series of m/z peaks for this single protein. The inset shows a computer analysis of the data from this series of peaks that generates a single peak at the correct molecular mass of the protein. (Adapted from Figure 2 in Mann M, Wilm M, 1995. Trends in Biochemical Sciences 20:219–224.)

Matrix-assisted laser desorption/ionisation

MALDI was introduced in the late 1980s by the group of Hillenkamp (34). The MALDI

technique is, like electrospray ionisation, referred to as ‘soft’ and thereby compatible with analysis

1.Introduction

12

of large biomolecules like proteins (35,36). The analytes are mixed with a saturated solution of

ultraviolet-absorbing matrix. The most commonly used matrices in peptide/protein analysis are α-

cyano-4-hydroxy-cinnamic acid (for small peptides) and 3,5-dimethoxy-4-hydroxy-cinnamic acid, i.

e. sinapinic acid (for larger peptides and proteins). The matrix/analyte mixture is applied to a target

plate. The solvent evaporates and the matrix and the analytes co-crystallise on the target. A laser

beam (commonly a nitrogen laser at 337 nm) provides light that is absorbed by the aromatic matrix

molecules. Energy is subsequently transferred to the analyte that becomes desorbed into the gas

phase (Fig. 3). The ionisation mechanism is not fully understood and several suggestions are still

debated (37). Co-desorption of matrix and analyte succeeds proton transfer, which may take place

in the solid phase, and also in the expanding plume of matrix and analyte ions after the laser

irradiation.

Figure 3. Ionization of matrix and sample particles as a result of laser exposure.

The MALDI source has traditionally been coupled to TOF mass analysers because of its pulsed

nature. Recent developments have made possible atmospheric pressure MALDI mass spectrometry.

MALDI produces mainly singly charged ions, and this feature means it is excellently suited for

analysis of complex biological mixtures such as protein digests (38). MALDI TOF mass

spectrometry is thus the primary analytical technique in proteomics for identification of proteins

separated by two-dimensional gel electrophoresis (39). The basic amino acid residues in peptides

and proteins are easily protonated and, consequently, these biomolecules are preferably analysed in

the positive ion mode. The MALDI ionisation process is less sensitive to salts than electrospray

ionisation. Nevertheless, salts and other impurities will cause peak broadening with the formation of

adducts. This will lower the mass accuracy and limit sensitivity, and these problems highlight the

1.Introduction

13

importance of satisfactory sample preparation. A number of different clean-up approaches have

been reported (40-41) often including the use of microcolumns for desalting, which have recently

become available commercially (ZipTip; Millipore). Several biotechnology companies have

launched robotic systems that desalt the large sample sets produced, before the introduction of

sample to MALDI mass spectrometry.

For a long time, MALDI TOF was considered a low-resolution mass spectrometric method. The

introduction of delayed extraction (42-44) and the mass reflectron (45) has changed this view

dramatically. Descriptions of these two features can be found below in the ‘The time-offlight mass

analyser’ section.

MALDI TOF has lately become the instrument of choice for many laboratories that are

investing in mass spectrometry equipment. The major advantages are that it is an easy system to

operate, it requires minimal mass spectrometric expertise to obtain data and it usually gives a quick

result. The relatively low purchasing cost together with reasonable running costs are other

parameters that make the system attractive from a buyers point of view. However, as with all mass

spectrometric systems, the interpretation of data, whether mass spectra or a list of database search

‘hits’, requires a degree of expertise and provides the rate-determining step for protein analysis.

Mass analysers

When ions have been formed in the source, they are transported to the analyser region and

separated according to their mass-to-charge ratio. A number of mass analysers are available which

can be divided into two classes. The first class, electric-field mass analysers, consist of the

quadrupole mass filter, the quadrupole ion trap (Paul trap) and the TOF mass analyser. The second

class, magnetic field mass analysers, comprise magnetic sectors and ion cyclotron resonance mass

analysers. The different analysers vary in their mass accuracy, mass range, resolution, sensitivity,

speed, footprint, cost, and the choice will depend on the specific application. The quadrupole ion

trap, for example, has MSn capability (the possibility to perform dissociation analysis of created

product ions), and FTICR mass spectrometers have extraordinary resolution (up to 107) and mass

accuracy (46). Magnetic sector instruments are the most mature instruments on the market and have

been a mainstay for many decades. Today, they are the instrument of choice for environmental

pollutant analysis. However, quadrupoles, traps and TOF instruments have largely accommodated

the growth in biological applications while the market for sector instruments remains constant. The

characteristics of the quadrupole mass filter and the TOF mass analyzer are described below.

1.Introduction

14

The quadrupole mass filter

The quadrupole mass filter is the most common mass analyser in use today and can be regarded

as a real ‘workhorse’. It was introduced in the early 1950s (47) and the technique has only seen

modest developments since then. The mass filter is used extensively as both a stand-alone device

and in multistage mass spectrometers like triple quadrupoles (48) and quadrupole TOF instruments

(49,59). The quadrupole analyser is constructed of four electronically conducting cylindrical rods

and is operated by the application of a combination of direct current (DC) and radio frequency (RF)

voltages (Fig.4).

Figure 3. Schematic representation of the quadrupole mass filter and its connections.

The mass filter establishes a two-dimensional quadrupole field between the four cylindrical

electrodes with the two opposite rods connected electrically. One rod pair (+) is connected to a

positive DC voltage, upon which a sinusoidal RF voltage is superimposed. The other rod pair (–) is

connected to a negative DC voltage, upon which a sinusoidal RF voltage is also superimposed.

Successful selection of a specific ion requires the RF and DC values to be set such that only the ion

of interest has a stable trajectory through the quadrupole system. In one field direction, ions with a

low m/z (light ions) will follow the alternating component of the field, gain energy and oscillate

with increasingly large amplitudes until they hit one of the rods and are discharged. Only high-mass

ions will hence be transmitted to the other end of the quadrupole. However, in the other direction of

the field, ions with a high m/z (heavy ions) will be unstable because of the defocusing effect of the

DC component. Lighter ions, on the other hand, will be stabilised by the alternating component and

transferred to the other end of the quadrupole. Combining the two directions gives a mass filter that

is suitable for mass analysis. Application of a suitable RF/DC ratio to the quadrupole can make it

1.Introduction

15

discriminate against both high- and low-mass ions to a desired degree. The mass filter can be set to

transmit a single isotope or to scan over a wide m/z range.The mass filter is a continuous analyser

compared to the TOF analyser that has a pulsed nature. This feature makes the quadrupole highly

compatible with continuous infusion sources, e. g. electrospray and liquid separation techniques

such as high-performance liquid chromatography (HPLC) and CE. The sensitivity of the analyser

for mass spectral acquisition is limited by the necessity to scan the quadrupole. The mass range of

commercial instruments is now about m/z 4000. However, modern quadrupoles can transmit ions

with m/z values above 10,000 (51) and by reducing the operating frequency, the mass range can be

extended to m/z 45,000 (52). Even though calibration is a straightforward process in mass filters

(m/z depends linearly on RF and DC), mass accuracy has traditionally been poor. However, Green

and co-workers have shown that accuracies in the range of 5 ppm can be achieved with careful

operation (53).

The TOF mass analyzer

A major development in TOF mass spectrometry came in the mid 1950s when Wiley and

McLaren (54) described ‘time-lag focusing’which markedly improves resolution. The principle of

the TOF mass analyser is to measure the flight time of ions accelerated out of an ion source into a

field-free drift tube to a detector. The flight time is related to the m/z values of the ions according to

the following formula:

TOF = L (2Uacc e)–1/2 (m/z)1/2

where L is the drift length in the field-free region, Uacc is the potential difference in the accelerating

region, e is the charge of an electron, m is the mass of the ion and z is its charge state. The TOF is

usually measured from the time point at which the ions are accelerated out of the source to the time

point when they reach the detector. The ions will separate in the TOF mass analyser according to

their m/z ratios, light ions arriving at the detector earlier than heavy ions if they carry the same

number of charges.

The ions initial spatial spread and initial velocity of the ions limit the resolving power of a TOF

mass analyser. In a MALDI source, for example, the ionisation creates a burst of ions that will be at

different distances from the detector (spatial spread) and have different kinetic energies. Ions with

the same m/z value but with different distances to the detector will consequently be detected at

different time points, thus decreasing resolution. The same is true for ions with the same m/z value

1.Introduction

16

but with different initial kinetic energy. As mentioned earlier, powerful tools have been invented to

compensate for these two problems.

Mamyrin and co-workers (45) introduced the mass reflectron in 1973. The mass reflectron

compensates for the initial energy spread of the ions; a schematic of a TOF instrument with a

reflector flight tube is shown in figure 5. In this type of instrument, the ions are not detected after a

single pass through the field-free region. They are instead reflected back into the field-free drift tube

by the electric field of the reflectron and detected at the same end of the tube as the initial

acceleration region. Ions with a high energy will penetrate the reflectron deeper than ions with the

same m/z value but with lower energy. Ions with the same m/z value can then be focused at the

detector. In addition, the incorporation of a reflectron will extend the flight path with a resultant

improved resolving power.

Figure 5. Schematic diagram of a time-of-flight mass analyser with a two-stage accelerating region and a mass reflectron.

The second major instrumental development that has resulted in improved resolution in TOF is,

as mentioned above, time-lag focusing. Wiley and McLaren (54) described an instrument with a

two-stage accelerating region and, more recently, a number of related designs have been applied to

MALDI TOF (42-44). By introducing a delay between the end of the ionisation pulse and the

application of the extraction pulse, the ions are sorted in space according to their original kinetic

energy. An ion that starts further from the field-free region in the first electric field will then be

accelerated to a slightly higher energy than an ion with the same m/z value but with a starting

position closer to the field-free region. The second electric field accelerates the ions into the drift

tube such that those with the same m/z value will arrive at the detector position at the same time.

1.Introduction

17

TOF is considered to be a high-speed mass analyser. The basic cycle time for the TOF analyser

is limited by the flight time of the heaviest ions. Since this is frequently in the 100 µs range,

thousands of full mass spectra can be generated each second. TOF mass analysers can produce full

spectra at high sample utilisation efficiencies, because all the m/z values in the flight tube at any

time can be detected. These capabilities of fast spectral generation rate, high efficiency and a high

duty cycle have made TOF a target for continuous ionization sources. Dodonov and co-workers

(55) introduced the electrospray source to an orthogonal extraction TOF mass spectrometer in 1987.

The continuously infused ions are pulsed at a high frequency perpendicular to their initial direction

of movement and their flight times are measured. The group of Guilhaus (56,57) has made great

contributions to the area of orthogonal-acceleration TOF mass analysers. This instrumental set-up

with coupling to continuous ionisation sources has opened up many new important applications for

TOF. One new application is in the field of non-covalent interactions of proteins, where the

theoretically unlimited mass range of the TOF mass analyser is of great value (58). Nowadays, TOF

mass analysers provide a valuable second stage in hybrid tandem mass spectrometers (49,59).

Protein preparation and purification

The up-front isolation procedures can have the most significant impact on the outcome of an

MS-based investigation. For example, sensitivity of the overall procedure is usually determined

more by the purification strategy than by the sensitivity of the MS instrument per se. Typically,

protein purification starts with a whole-cell lysate and ends with a gel-separated protein band or

spot. MS analysis is usually carried out on peptides obtained after enzymatic degradation of these

gel-separated proteins. In special cases, the intact proteins are analyzed or the gel electrophoretic

step is omitted by digesting a collection of proteins in solution and analyzing the resulting complex

mixture of peptides.

Special considerations for protein preparation methods

In principle, any of the classical separation methods such as centrifugation, column

chromatography, and affinity-based procedures can precede the final gel electrophoresis. As long as

the proteins of interest can be adequately resolved, it is best to minimize the number of separations.

Generally, silver-stained amounts are necessary for successful MS identification of proteins (5–50

ng or 0.1–1 pmol for a 50 kDa protein], but even higher sensitivities have been achieved by

specialized groups. It is important to minimize contamination with keratins, which are introduced

1.Introduction

18

by dust, chemicals, handling without gloves, etc, as the keratin peptides can easily dominate the

spectrum. Most detergents and salts are incompatible with both 2-D gels and MS. Therefore,

dialysis of the sample may be required. If the protein can be eluted from reversed-phase media, the

best sample preparation is achieved on small, low-pressure traps that can be incorporated into MS

injection ports. In affinity-based protocols it is important that the bait is pure, as contaminating

proteins, for example bacterial proteins in a Glutathione S. transferase (GST) fusion preparation,

will hinder analysis.

Analysis of intact proteins

Several uses of mass spectrometry involve the characterization of recombinant proteins. For

example, the glycosylation and disulfide bonding pattern of therapeutic proteins, such as growth

factors, can be studied in detail. A mass spectrum of the intact protein provides the precise

molecular weight of the major and minor forms of the protein, data that cannot always be gained

from peptide mapping. However, larger proteins are typically heterogeneous, making molecular

weight determination difficult.

Electrospray is the method of choice for determining molecular weight of proteins, as MALDI

results in broad peaks and low sensitivity for proteins above about 30 kDa. As mentioned above,

proteins need to be free of detergents and salts, which is usually accomplished by reversed-phase

chromatography. Formic acid can be used to solubilize proteins for electrospray analysis.

Identification of intact proteins from cell lysates by molecular weight alone is difficult for

several reasons. Sensitivity of ESI-MS for large molecules is poorer than for peptide analysis

because the signal is distributed over many charge states and the heterogeneity of the protein

similarly splits the signal into many components. More fundamentally, the molecular weight of a

protein cannot be predicted precisely from its database entry, because of N- and C-terminal

processing, posttranslational modifications, and chemical modifications introduced during sample

purification (for example, oxidation of methionines). Therefore, even a precise molecular weight by

itself will not allow identification of the protein. Nevertheless, in a recent study several hundred

proteins were recorded in a single experiment in which isoelectric capillary electrophoresis of

lysates of an Escherichia coli cell was combined with FTMS (60). Identification of the proteins was

not achieved in this experiment for the above-mentioned reasons.

1.Introduction

19

Digestion and preparation of gel-separated proteins for MS analysis

For MS-based analysis, most of the detergents and salts are eliminated in the gel washing

procedure. Nevertheless, the protein should be as concentrated as possible in the gel, to avoid

excessive background in the MS analysis. Pooling of spots is not necessarily advantageous, as both

protein and background will increase. Coomassie staining, silver staining, or radioactive labeling,

which are all compatible with MS analysis, can visualize proteins. Cross-linkers and harsh

oxidizing agents should be avoided as they interfere with extraction of peptides from the gel or may

chemically modify the peptides (61). Silver staining provides adequate sensitivity, but it should be

recognized that it has a narrow linear range.

Protein bands are excised from the gel and subjected to reduction, alkylation, several washing

steps, and finally enzymatic digestion followed by peptide extraction. A small portion of the

resulting peptide mixture can be directly used for MALDI peptide mapping. For electrospray

analysis, and often for MALDI analysis, peptides are desalted and concentrated. This can be

performed by columns of reversed-phase material in nanospray needles or gel loader tips or by

injection into liquid chromatography columns. The peptides are then eluted onto the MALDI target

or into a nanoelectrospray spraying needle or are loaded onto a microcapillary column.

Figure 6. Scheme of enzymatic fingerprint of proteins in situ.

1.Introduction

20

Peptide sequencing by tandem mass spectrometry

The sequence of peptides can be determined by interpreting the data resulting from fragmenting

the peptides in tandem mass spectrometers (62,63). In this technique, one peptide species out of a

mixture is selected in the first mass spectrometer and is then dissociated by collision with an inert

gas, such as argon or nitrogen. The resulting fragments are separated in the second part of the

tandem mass spectrometer, producing the tandem mass spectrum, or MS/MS spectrum. In the

instruments in use today, multiple collisions impart energy onto the molecule until it fragments.

(This is low-energy fragmentation, in which any single hit is not sufficient to break the peptide

bond. In high-energy fragmentation, the molecules have higher velocity and a single hit can break

bonds).

As shown in figure 7, several bonds along the backbone can be broken by the collisions. The

most common ion types are the b and the y ions, which denote fragmentation at the amide bond

with charge retention on the N or C terminus, respectively. Most proteomics experiments are

performed with tryptic peptides, which have arginyl or lysyl residues as their C-terminal residues.

In this case, y ions are the predominant ion series observed.

Most peptide sequencing is performed on electrosprayed ions. These ions generally have a

charge state corresponding to the number of positively charged amino acids plus the charge

formally localized at the N terminus of the peptide. Thus tryptic peptides are often doubly charged,

or triply charged if they contain a histidyl residue. Larger peptides are multiply charged, and their

fragmentation spectra often contain multiply charged ion series as well.

Figure 7. Nomenclature for the product ions generated in the fragmentation of peptide molecules by tandem mass spectrometry. Collision-activated dissociation (CAD) causes a single cleavage to occur more or less randomly at the various amide bonds in the collection of peptide molecules. This process generates a series of fragments that differ by a single amino acid residue. Ions of type y contain the C terminus plus one or more additional residues. Ions of type b contain the N terminus plus one or more additional residues. Additional ion types correspond to cleavages at different positions in the backbone (dashed lines)

1.Introduction

21

Tandem mass spectra are usually interpreted with computer assistance, or matched against

databases directly. In very high quality spectra it is possible to interpret the fragmentation ladders

(the b and the y ion series) from the low mass end through to the highest mass ion. For example, the

y ion series of tryptic peptides will start with masses y1=147 (Lys) or 175 (Arg). The next

fragmentation peak in the y ion series, the y2 ion, differs by the mass of an amino acid residue and

thus “spells out” the next amino acid. Similarly, the b ion series starts with b1 for the N-terminal

amino acid and is traced upward in molecular weight. Note that the b and y ions are not

distinguishable a priori. Ideally, a complete set of b and y ions will (doubly) confirm the entire

peptide sequence. In practice, not all fragment ions are present at detectable levels and fragments

also arise by double fragmentation of the backbone (internal fragment ions). Therefore, it is often

possible to interpret part but not all of the sequence with confidence.

Other features of the MS/MS spectrum include immonium ions (i), which arise by double

cleavage of the peptide backbone, N-terminal and C-terminal to the amino acid residue. These

immonium ions can indicate or confirm the presence of individual amino acids.

Fragmentation is not equally likely along the entire length of the peptide backbone. For

example, fragmentation between the two N-terminal amino acids is energetically unfavorable and

therefore the b1 ion is often not observed. The b2 ion, however, and its companion, the a2 ion, at a

28 Da (CO) mass difference, is usually very prominent. Likewise, fragmentation at the C-terminal

bond of a praline residue is weak but cleavage at the N-terminal side is usually very prominent.

Another feature that can frequently be observed is the switchover from a C-terminal to a N-

terminal series because of charge retention of a positively charged amino acid residue, such as

histidine.

Bond breakage of the doubly charged tryptic peptides most often used in protein

characterization is believed to proceed via the localized charge on the arginines or lysine residue on

the C terminus and the delocalized proton formally located on the N terminus, which induces the

amide bond breakage along the backbone. The arginyl residue strongly localizes the proton, and

peptides containing internal arginines often result in MS/MS spectra that are very difficult to

interpret.

As mentioned above, fragmentation of large, multiply charged ions often leads to multiply

charged ion series. High resolution is advantageous in studying these ions because it allows direct

assignment of the charge state of fragments based on the spacing of the carbon isotope peaks (for

example, they are spaced 1/3 of a dalton apart if they are triply charged). Larger peptides often

fragment efficiently and provide long ion series, but because the precursor ion intensity is

distributed over several charge states, sensitivity may not be as high.

1.Introduction

22

Mass spectrometers in use today cannot distinguish between isoleucine and leucine, which have

the same mass (though the distinction can in principle be made by using the different retention time

of leucine- and isoleucine-containing peptides during chromatography or the side chain

fragmentation in high-energy collisions). The glutamyl and lysyl residues have the same nominal

mass but can be distinguished by their mass difference of 0.036 Da on modern TOF or FTMS

instruments.

Protein identification by database searching

A key advance in biological mass spectrometry was the development of algorithms for the

identification of proteins by mass spectrometric data matched to a database, originally using a set of

peptide masses and now increasingly using the fragmentation spectra of the individual peptides. For

the reasons described in previous sections, obtaining the complete sequence of a peptide from the

tandem mass spectrum was time consuming at best and often impossible. With the availability of

the complete sequence of an increasing number of model species, the peptide sequencing problem,

formerly a holy grail in biological mass spectrometry, is reduced to a database correlation, enabling

automation and the scaling up of proteomics experiments.

Peptide mass fingerprint

In this method, a “mass fingerprint ” is obtained of a protein enzymatically degraded with a

sequence-specific protease such as trypsin. This set of masses, typically obtained by MALDI-TOF,

is then compared to the theoretically expected tryptic peptide masses for each entry in the database.

The proteins can be ranked according to the number of peptide matches (Fig. 5). More sophisticated

scoring algorithms take the mass accuracy and the percentage of the protein sequence covered into

account and attempt to calculate a level of confidence for the match (64-66). Other factors can also

be included, such as the fact that larger peptides are less frequent in the database and should

therefore count more when matched. The accuracy obtained in the measurement of peptide mass

strongly influences the specificity of the search (67,68). When high mass accuracy (10 to 50 ppm) is

achieved, as a rule at least five peptide masses need to be matched to the protein and 15% of the

protein sequence needs to be covered for an unambiguous identification. After a match has been

found, a second-pass search is performed to correlate remaining peptides with the database

sequence of the match, taking into account possible modifications.

1.Introduction

23

Mass fingerprinting can also resolve simple protein mixtures, consisting of several proteins

within a roughly comparable amount. For example, databases can be searched iteratively by

removing the peptides associated with an unambiguous match (69).

Generally, peptide mass fingerprinting is used for the rapid identification of a single protein

component. Protein sequences need to be in the database in substantially full length. Isoforms can

be differentiated from each other, if peptides covering the sequence differences appear in the

peptide map.

Searching with tandem mass spectrometric data

Databases can also be searched by tandem mass spectrometric data obtained on peptides from

the proteins of interest. Because the tandem mass spectra contain structural information related to

the sequence of the peptide, rather than only its mass, these searches are generally more specific and

discriminating.

Several approaches exist. The peptide sequence tag method (70) makes use of the fact that

nearly every tandem mass spectrum contains at least a short run of fragment ions that

unambiguously specifies a short amino acid sequence. As few as two amino acids can be combined

with the start mass and the end mass of the series, which specify the exact location of the sequence

in the peptide and the known cleavage specificity of the enzyme. Such a peptide sequence tag will

then retrieve from the database one or a few sequences whose theoretical fragmentation pattern is

matched against the experimental one. The procedure can be automated and is highly specific,

especially when performed using instruments with a high accuracy for mass, such as the quadrupole

TOF instrument.

Other methods do not attempt to extract any sequence information at all from the MS/MS

spectrum (71). Instead, the experimental spectrum is matched against a calculated spectrum for all

peptides in the database. A score is given to determine howmuch the tandem mass spectrum agrees

with the calculated sequence. Another score indicates how differently the next most similar

sequence in the database fits the spectrum. Although this method can be highly automated, the

sequences need to be verified by manual inspection unless the score is very high. In a typical liquid

chromatography MS/MS experiment on an ion trap, about 10–20% of MS/MS spectra may require

no further interpretation, whereas a much larger percentage can be verified manually.

Tandem mass spectrometry allows direct analysis of protein mixtures. Bands from one-

dimensional gels, when analyzed with the high sensitivity of the mass spectrometer, often turn out

to contain several proteins. In such a case, the software reports a list of proteins, each matched by

1.Introduction

24

one or several peptides. In extreme cases, crude protein mixtures can be reduced to peptides, and

the peptides fragmented and searched in databases (72). In this way, large numbers of proteins, up

to hundreds, can be identified all at once (see LC-MS/MS below).

Database searching and protein modifications

Protein modifications do not present an obstacle to identification. Of a typical 50 peptides (for a

50 kDa protein) generated by tryptic digestion, only a few will be modified. As described above,

only a small number of peptides are required for unique matching to a database entry, especially in

the case of data from tandem mass spectrometry; therefore even extensive modification only

marginally increases the difficulty of protein identification.

Modifications may also be discovered while searching databases. For example, a

phosphopeptide can be correlated to a peptide sequence in the database with an additional mass

increment due to the phosphogroup (80 Da). In the algorithm for the peptide sequence tag, a mass

difference at either side of the tag sequence can be allowed. For example, the tag sequence and the

mass to the C terminus of the peptide could agree with the database entry. However, the mass to the

N terminus could be larger by 80 Da. In this case there would have to be a modification yielding a

mass difference of 80 Da between the N terminus of the peptide and the start of the tag sequence.

The peptide fragmentation spectra can also be calculated for all possible combinations of

common modifications. For example, serines in suspected phosphopeptides can be substituted in

turn with phosphoserines when calculating fragmentation spectra. Because of the increase in

possible peptide fragmentation spectra to consider and the increase in computation time, this search

is usually not performed for the whole database but only for a small set of sequences, including the

protein sequence already identified in the database from unmodified peptides.

Liquid chromatography and tandem mass

Principles

Liquid chromatography (LC) coupled to tandem mass spectrometry, called LC-MS/MS, is a

powerful technique for the analysis of peptides and proteins. This methodology combines efficient

separations of biological materials and sensitive identification of the individual components by

mass spectrometry. Complicated mixtures containing hundreds of proteins can be analyzed directly

even when concentration levels of different proteins vary by orders of magnitude. LC-MS/MS can

1.Introduction

25

be used alone or in combination with 1-D or 2-D electrophoresis, immunoprecipitation, or other

protein purification techniques.

Although numerous methods for coupling liquid chromatography to mass spectrometry have

been explored, it is electrospray ionization that has transformed LC-MS/MS into a routine

laboratory procedure sensitive enough to analyze peptides and proteins at levels interesting in

biological research. As described above, ESI requires a continuous flow of liquid, and signal

strength is independent of flow rate. In order to obtain maximum sensitivity, research efforts have

focused on coupling nano-scale LC at submicroliter flow rates to the highly sensitive micro-scale

ESI interface (30,73, 74). Currently, detection limits of a few femtomoles of peptide material loaded

on the column make this technique compatible with silver-stained, fluorescently labeled, or faintly

stained Coomassie gel bands and capable of detecting proteins and peptides present at a low copy

number per cell.

In a typical LC-MS/MS experiment, the analyte is eluted from a reversed-phase column to

separate the peptides by hydrophobicity, and is ionized and transferred with high efficiency into the

mass spectrometer for analysis. A large amount of data regarding individual species in a

complicated mixture is generated. For example, the peptide ligands associated with the human

major histocompatibility complex (MHC) class I molecules, HLA-A2.1, are a mixture of

approximately 10,000 different peptide species (75). As shown in figure 8, an aliquot representing

the amount isolated from 1 x 108 cells was loaded onto a nanoLC column and eluted into an ion trap

mass spectrometer using a long gradient. Mass spectra were acquired over the mass range 300–

2000. The data acquired contain molecular weight information on the peptide species and their

amounts. The ion current for each scan can be summed and plotted as a function of time. This

display is a total ion current chromatogram (TIC), shown in figure 8A, and is similar to a UV

chromatogram. Postacquisition, the data can be interrogated to reveal the ion current recorded at a

particular m/z (molecular weight), or a selected ion current chromatogram (SIC) (Fig. 8B). Any

individual peptide can be sequenced without further purification by isolating the eluting peptide,

fragmenting it, and obtaining the MS/MS spectrum (Fig. 8D). In this manner a large number of

peptides can be sequenced in a single LC-MS/MS run, even those that have the same molecular

weight if they differ in hydrophobicity. In practice, the mass spectrometer is often programmed to

perform one scan to determine the peptide masses and then to sequence the three to eight most

abundant peptides (data-dependent acquisition). The separation principle is almost always reversed-

phase high-performance liquid chromatography (RP-HPLC) as applied elsewhere in protein

chemistry; the only difference is that the dimensions and flow rates are much smaller. Peptides elute

with a typical peak width of 30 seconds. Higher separation efficiencies could in principle be

1.Introduction

26

achieved with coated open tubular capillaries, 1–5 µm internal diameter, which have been used for

high-sensitivity analysis, but the small amount of wall surface area results in a lower capacity

factor.

Figure 8. LC-MS analysis of a complex mixture of peptides isolated from human MHC class I molecules. An aliquot of peptides was loaded onto a nanoLC column and eluted with a gradient of acetonitrile and 0.1 M acetic acid. Spectra over the mass range 300 to 2000 were acquired every 1.5 seconds. (A) The sum of all the ion current recorded at the detector plotted as a function of scan number, a total ion current chromatogram (TIC). (B) The data replotted to show the ion current for m/z D 591.5, called a selected ion current chromatogram (SIC). (C) Approximately 20 mass spectra at elution time 117 min were summed to generate a single mass spectrum, which shows all the peptides that co-elute with the peptide at m/z D 591.5 (MC2H)2C [doubly charged peptide indicated by (MC2H)2C]. (D) The MS/MS spectrum for the (MC2H)2C ion at m/z 591.5 and the interpreted amino acid sequence. Instrumental and practical considerations

Low-flow-rate LC, or nanoLC, once thought to be a specialized and difficult technique, has

become a routine procedure in many laboratories over the last few years. The low flow rates

required are achieved by modifying a conventional LC system, transfer lines, gradient mixture, and

use of a precolumn splitting device (76) or by the use of specially designed nanoLC systems. An

1.Introduction

27

online UV detector is incorporated into the system, if desired. Typical nanoLC columns are 50–100

µm in internal diameter and are packed with polymeric or silica-based, C18 coated, stationary

phases with typical particle sizes in the 3 to 10 µm range. The smaller the column diameter, the

lower the flow rate for the same chromatographic separation and hence the higher the sensitivity.

Column diameters of 75 µm are the current compromise between ultimate sensitivity and trouble-

free operation on a routine basis. Columns are available from commercial vendors or are packed in

individual laboratories. As in conventional LC, the sample can be introduced onto the column via

loop injections of the sample. Auto samplers are available that inject submicroliter volumes. To

minimize losses associated with handling of the sample, an alternative method is to displace the

sample directly from the sample tube onto the column by pneumatic displacement (bomb loading).

Ideally, the sample is loaded at a higher flow rate (microliters per minute) onto a trap column and is

then eluted onto the separation column. This arrangement has the further advantage of leading to

fewer problems of plugging in these very fine columns.

An electrical connection needs to be made between the liquid and a power supply in order to

supply the charges to the electrospray process. In one type of connection the electrospray needle, or

emitter, is coated with a conductive material and the voltage is applied directly. Alternatively, a

liquid junction, approximately 3 nl, is purposefully formed in a metal union (stainless steel, gold,

titanium) between the exit of the LC column and the electrospray needle (77) or by applying the

spraying voltage prior to the column (78). For high-sensitivity LC-MS/MS applications, careful

consideration must be given to solvent purity. For example, even though UV-detectable trace

contaminants may not be present, the solvents may contain ionizable impurities that reduce the final

signal to noise of the analysis. The nanoLC system couples to various mass spectrometers such as

triple quadrupole, quadrupole TOF, ion traps, and FTMS instruments.

For nanoLC systems, sensitivity at 1–10 fmol in LC-MS/MS mode is routine on a triple

quadrupole mass spectrometer (75) and has been reported in the low attomole range for selected ion

monitoring. The newer quadrupole TOF instrument promises to provide a large improvement in

sensitivity in MS/MS mode over the triple quadrupole instrument, in addition to increased mass

resolution. The ion trap mass spectrometer has sensitivity in the full-scan MS mode of 1–5 fmol and

is mainly limited by chemical noise introduced into the trap. However, in MS/MS mode, increased

duty cycle of the ion trap instrument results in an improvement of ultimate sensitivity reported in

the 10 to 50 attomole range (79). This feature of the ion trap can be used only when sequencing a

particular, known mass during an LC-MS/MS run (for example, when sequencing a putative

phosphopeptide or when the mass has been already been determined by a more sensitive full-scan

method of mass spectrometry such as FTMS). Recently, automated variable-flow LC, 5–200 nl/min

1.Introduction

28

(80), also known as peak parking (81), has been employed to improve sensitivity to the 10 to 50

attomole level in MS/MS experiments on an ion trap mass spectrometer. On the FTMS, LC-MS

sensitivities at the 10 attomole level with a dynamic range of 103 have been reported. Sample

carryover does not appear to be a problem in the analysis when high-sensitivity applications are

dedicated to a single instrumental setup. It should be noted that all these sensitivity numbers relate

to material applied to the column rather than protein material in a gel band.

Mass spectrometry is the core technique of proteomics. Progress in instrumentation continues to

be made at a fast pace. Automation will make it possible to obtain rates of data generation that will

exceed those of genomics. This will allow the study of all protein complexes and organelles that can

be purified. Many protein interactions can be studied by coimmunoprecipitation followed by mass

spectrometric identification. Quantitative proteomics will most likely be achieved by stable isotope

methods in combination with mass spectrometry. Apart from the pressing areas of automation of

data acquisition and interpretation, areas for future research will be the analysis of protein

modification on a large scale. Cross-linking studies will tell us not only about the composition but

also about the spatial organization of protein complexes. There is much room for creativity in

connecting cell and molecular biological strategies with the powerful mass spectrometric

capabilities to solve questions that could not previously be addressed.

1.Introduction

29

REFERENCES

1. Fenn JB, Mann M, Meng CK, Wong SF,Whitehouse CM (1989) Electrospray ionization for

mass spectrometry of large biomolecules Science 246:64–71

2. Karas M, Hillenkamp F (1988) Laser desorption ionization of proteins with molecular

masses exceeding 10,000 daltons. Anal. Chem. 60:2299–2301

3. Pandey A, Mann M (2000) Proteomics to study genes and genomes. Nature 405:837–846

4. Klose J. (1975) Protein mapping by combined isoelectric focusing and electrophoresis of

mouse tissues. A novel approach to testing for induced point mutations in mammals

Humangenetik 26:231–243

5. O’Farrell PH (1975) High resolution two-dimensional electrophoresis of proteins J. Biol.

Chem. 250:4007–4021

6. Dempster AJ (1918) A new method of positive ray analysis. Phys. Rev. 11: 316–324

7. Munson MSB, Field FH (1966) Chemical ionization mass spectrometry I. General

introduction. J. Am. Chem. Soc. 88: 2621–2630

8. Barber M, Bordoli RS, Sedgwick RD, Tyler AN (1981) Fast atom bombardment of solids

(FAB): a new ion source for mass spectrometry. J. Chem. Soc. Chem. Commun. 325–327

9. Barber M, Bordoli RS, Elliott GJ, Sedgwick RN, Tyler AN (1982) Fast atom bombardment

mass spectrometry. Anal. Chem. 54:645A–657A

10. Dole M, Mach LL, Hines RL, Mobley RC, Ferguson LD, Alice MB (1968) Molecular

beams of macroions. J. Chem. Phys. 49:2240–2247

11. Yamashita M, Fenn JB (1984) Electrospray ion source: another variation on the free-jet

theme. J. Phys. Chem. 88:4451–4459

12. Yamashita M, Fenn JB (1984) Negative ion production with the electrospray ion source. J.

Phys. Chem. 88:4671–4675

13. Alexandrov ML, Gall LN, Krasnov NV, Nikolaev VI, Pavlenko VA, Shkurov VA, Baram

GI, Grachev MA, Knorre VD, Kusner YS (1984) Direct coupling of a microcolumn liquid

chromatograph and a mass spectrometer. Bioorg. Khim. 10:710–712

14. Meng CK, Mann M, Fenn JB (1988) Of protons or proteins. Z. Phys. D. 10:361–368

15. Fenn JB, Mann M, Meng CK, Wong SF, Whitehouse CM (1989) Electrospray ionization for

mass spectrometry of large biomolecules. Science 246:64–71

16. Taylor G. (1964) Disintegration of water drops in an electric field. Proc. R. Soc. Lond

A280:383–397

1.Introduction

30

17. Rayleigh, Lord (1882) On the equilibrium of liquid conducting masses charged with

electricity. Philos. Mag. 14:184–186

18. Kebarle P, Ho Y (1997) On the mechanism of electrospray mass spectrometry. In:

Electrospray Ionization Mass Spectrometry: Fundamentals, Instrumentation and

Applications, pp. 3–63, Cole R. B. (ed), Wiley, New York

19. Iribarne JV, Thomson BA (1976) On the evaporation of small ions from charged droplets. J.

Chem. Phys. 64:2287–2294

20. Fenn JB (1993) Ion formation from charged droplets: roles of geometry, energy, and time. J.

Am. Soc. Mass Spectrom. 4:524–535

21. Kebarle P, Tang L (1993) From ions in solution to ions in the gas phase. Anal. Chem.

65:972A-986A

22. Cole RB.(ed) (1997) Electrospray Ionization Mass Spectrometry: Fundamentals,

Instrumentation and Applications, Wiley, New York

23. Kebarle J (2000) A brief overview of the present status of the mechanisms involved in

electrospray mass spectrometry. J. Mass Spectrom. 35:804–817

24. Cole RB (2000) Some tenets pertaining to electrospray ionization mass spectrometry. J.

Mass Spectrom. 35:763–772

25. Bruins AP (1991) Liquid chromatography-mass spectrometry with ionspray and

electrospray interfaces in pharmaceutical and biomedical research. J. Chromatogr. 554:39–

46

26. Barnidge DR, Nilsson S, Markides KE (1999) A design for low-flow sheathless electrospray

emitters. Anal. Chem. 71:4115–4118

27. Gale DC, Smith R. D. (1993) Small volume and low flow-rate electrospray ionization mass

spectrometry of aqueous samples. Rapid Commun. Mass Spectrom. 7:1017–1021

28. Emmett MR, Caprioli RM (1994) Micro-electrospray mass spectrometry: ultra-high-

sensitivity analysis of peptides and proteins. J. Am. Soc. Mass Spectrom. 5:605–613

29. Andren PE, Emmett MR, Caprioli RM (1994) Microelectrospray: zeptomole/attomole per

microliter sensitivity for peptides. J. Am. Soc. Mass Spectrom. 5:867–869

30. Wilm M, Mann M. (1994) Electrospray and Taylorcone theory: Dole’s beam of

macromolecules at last? Int. J. Mass Spectrom. Ion Proces. 136:167–180

31. Wilm M, Mann M. (1996) Analytical properties of the nano-electrospray ion source. Anal.

Chem. 68:1–8

1.Introduction

31

32. Richards D, Verrier H, Major H, Rontree J. (1998) The direct LC-MS analysis of drug

candidates in cell culture media using a dual orthogonal sampling API interface, p. 75, Proc.

16th IMMS, 4–6 May, Budapest

33. Ferrige AG, Seddon MJ, Green BN, Jarvis SA, Skilling J (1992) Disentangling electrospray

spectra with maximum entropy. Rapid Commun. Mass Spectrom. 6:707–711

34. Karas M, Bachmann D, Bahr U, Hillenkamp F (1987) Matrix-assisted ultraviolet laser

desorption of non-volatile compounds. Int. J. Mass Spectrom. Ion Proces. 78:53–68

35. Karas M, Hillenkamp F (1988) Ultraviolet laser desorption of ions above 10 kDa. In:

Abstracts, 11th Int. Mass Spectrom. Conf., Bordeaux, 29 August – 2 September

36. Karas M, Hillenkamp F (1988) Laser desorption ionization of proteins with molecular

masses exceeding 10 000 Daltons. Anal. Chem. 60:1299–2301

37. Cotter RJ (1997) Time-of-Flight Mass Spectrometry: Instrumentation and Applications in

Biological Research, pp. 131–134, Am. Chemical Society, Washington, D.C.

38. Roepstorff P (2000) MALDI-TOF mass spectrometry in proteinchemistry. In: Proteomics in

Functional Genomics: Protein Structure Analysis, pp. 81–97, Jollès P and Jörnvall H (eds),

Birkhäuser, Basel

39. Pandey A, Mann M (2000) Proteomics to study genes and genomes. Nature 405:837–846

40. Vorm O, Mann M. (1994) Improved mass accuracy in matrix-assisted laser

desorption/ionization time-of-flight mass spectrometry of peptides. J. Am. Soc. Mass

Spectrom. 5:955–958

41. Gobom J, Nordhoff E, Mirgorodskaya E, Ekman R, Roepstorff P. (1999) Sample

purification and preparation technique based on nano-scale reversed phase columns for the

sensitive analysis of complex peptide mixtures by matrix-assisted laser desorption/ionization

mass spectrometry. J. Mass Spectrom. 34:105–116

42. Colby SM, King TB, Reilly JP (1994) Improving the resolution of matrix-assisted laser

desorption/ionization timeof flight mass spectrometry by exploiting the correlation between

ion position and velocity. Rapid Commun. Mass Spectrom. 8:865–868

43. Vestal ML, Juhasz P, Martin SA (1995) Delayed extraction matrix-assisted laser

desorption/ionization time of flight mass spectrometry. Rapid Commun. Mass Spectrom.

9:1044–1050

44. Brown RS, Lennon JJ (1995) Mass resolution improvement by incorporation of pulsed ion

extraction/ionization linear time of flight mass spectrometry. Anal. Chem. 67:1998–2003

1.Introduction

32

45. Mamyrin BA, Karataev VI, Shmikk DV, Zagualin VA (1973) The mass-reflectron, a new

non-magnetic time of flight mass spectrometer with high resolution. Sov. Phys. JETP.

37:45–48

46. He F, Hendrickson CL, Marshall AG (2001) Baseline mass resolution of peptide isobars: a

record for molecular mass resolution. Anal. Chem. 73:647–650

47. Paul VW, Steinwedel H (1953) Ein neues Massenspektrometer ohne Magnetfeld. Z.

Naturforsch. 8a: 448–450

48. Yost RA, Enke CG.(1978) Selected ion fragmentation with a tandem quadrupole mass

spectrometer. J. Am. Chem. Soc. 100:2274–2275

49. Morris HR, Paxton T, Dell A, Langhorn B, Berg M, Bordoli RS, Hoyes J, Bateman RH

(1996) High sensitivity collisionally-activated decomposition tandem mass spectrometry on

a novel quadrupole/orthogonal-acceleration time-of-flight mass spectrometer. Rapid

Commun. Mass Spectrom. 10:889–896

50. Shevchenko A, Chernushevich I, Ens W, Standing KG, Thomson B, Wilm M, Mann M

(1997) Rapid ‘de novo’ peptide sequencing by a combination of nano electrospray, isotopic

labelling and a quadrupole/time of flight mass spectrometer. Rapid Commun. Mass

Spectrom. 11:1015–1024

51. Rostom AA, Robinson CV (1999) Detection of the intact GroEL chaperonin assembly by

mass spectrometry. J. Am. Chem. Soc. 121:4718-4719

52. Winger BE, Light-Wahl KJ, Ogorzalek Loo RR, Udseth HR, Smith RD (1993) Observation

and implications of high mass-to-charge ratio ions from electrospray ionization mass

spectrometry. J. Am. Soc. Mass Spectrom. 4:536–545

53. Tyler AN, Clayton E, Green BN (1996) Exact mass measurements of polar organic

molecules at low resolution using electrospray ionization and a quadrupole mass

spectrometer. Anal. Chem. 68:3561–3569

54. Wiley WC, McLaren I. H. (1955) Time of flight mass spectrometer with improved

resolution. Rev. Sci. Instrum. 26:1150–1157

55. Dodonov AF, Chernushevich IV, Dodonona TF, Raznikov VV, Talrose VL (1987) USSR

patent number 1681340A1

56. Guilhaus M, Mlynski V, Selby D (1997) Perfect timing: time of flight mass spectrometry.

Rapid Commun. Mass Spectrom. 11:951–962

57. Guilhaus M, Selby D, Mlynski V (2000) Orthogonal acceleration time of flight mass

spectrometry. Mass Spectrom. Rev. 19:65–107

1.Introduction

33

58. Tito MA, Tars K, Valegard K, Hajdu J, Robinson CV (2000) Electrospray time of flight

mass spectrometry of the intact MS2 virus capsid. J. Am. Chem. Soc. 122:3550–3551

59. Bateman RH, Green MR, Scott G, Clayton E (1995) A combined magnetic sector-time-of-

flight mass spectrometer for structural determination studies by tandem mass spectrometry.

Rapid Commun. Mass Spectrom. 9:1227–123

60. Jensen PK, Pasa-Tolic L, Anderson GA, Horner JA, Lipton MS, Bruce JE, Smith RD (1999)

Probing proteomes using capillary isoelectric focusing-electrospray ionization Fourier

transform ion cyclotron resonance mass spectrometry. Anal. Chem. 71:2076–208

61. Shevchenko A, Wilm M, Vorm O, Mann M. (1996) Mass spectrometric sequencing of

proteins silver-stained polyacrylamide gels. Anal. Chem. 68:850–858

62. Biemann K, Scoble HA (1987) Characterization by tandem mass spectrometry of structural

modifications in proteins. Science 237:992–998

63. Hunt DF, Shabanowitz J, Yates JR 3rd, Zhu NZ, Russell DH (1987) Tandem quadrupole

Fourier-transform mass spectrometry of oligopeptides and small proteins. Proc. Natl. Acad.

Sci. USA 84:620–623

64. Berndt P, Hobohm U, Langen H (1999) Reliable automatic protein identification from

matrix-assisted laser desorption/ionization mass spectrometric peptide

fingerprints.Electrophoresis 20:3521–3526

65. Perkins DN, Pappin DJ, Creasy DM, Cottrell JS (1999) Probability-based protein

identification by searching sequence databases using mass spectrometry data.

Electrophoresis 20:3551–3567

66. Eriksson J, Chait BT, Fenyo D (2000) A statistical basis for testing the significance of mass

spectrometric protein identification results. Anal. Chem. 72:999–1005

67. Jensen ON, Podtelejnikov A, Mann M (1996) Delayed extraction improves specificity in

database searches by matrix-assisted laser desorption/ionization peptide maps. Rapid

Commun. Mass Spectrom. 10:1371–1378

68. Clauser KR, Baker P, Burlingame AL (1999) Role of accurate mass measurement (+/- 10

ppm) in protein identification strategies employing MS or MS/MS and database searching.

Anal. Chem. 71:2871–2882

69. Jensen ON, Podtelejnikov AV, Mann M (1997) Identification of the components of simple

protein mixtures by high-accuracy peptide mass mapping and database searching Anal.

Chem. 69:4741–4750

70. Mann M, Wilm M (1994) Error-Tolerant Identification of Peptides in Sequence Databases

by Peptide Sequence Tags. Anal. Chem. 66:4390–4399

1.Introduction

34

71. Eng JK, McCormack AL, Yates JR 3rd (1994) An Approach to Correlate Tandem Mass

Spectral Data of Peptides with Amino Acid Sequences in a Protein Database. J. Am. Soc.

Mass Spectrom. 5:976– 989

72. Yates JR 3rd, McCormack AL, Schieltz D, Carmack E, Link A (1997) Direct analysis of

protein mixtures by tandem mass spectrometry. J. Protein Chem. 16:495–497

73. Gale DC, Smith RD (1993) Micro-electrospray mass spectrometry: ultra-high-sensitivity

analysis of peptides and proteins. Rapid Commun.Mass Spectrom. 7:1017–1021

74. Emmett MR, Caprioli RM (1994) Micro-electrospray mass spectrometry: ultra-high-

sensitivity analysis of peptides and proteins. J. Am. Soc. Mass Spectrom. 5:605–613

75. Cox AL, Skipper J, Chen Y, Henderson RA, Darrow TL, Shabanowitz J, Engelhard VH,

Hunt DF, Slingluff CL Jr. (1994) Identification of a peptide recognized by five melanoma-

specific human cytotoxic T cell lines. Science 264:716–719

76. Chervet JP, Ursem M, Salzmann JP (1996) Instrumental Requirements for Nanoscale Liquid

Chromatography. Anal. Chem. 68:1507–1512

77. Davis MT, Stahl DC, Hefta SA, Lee TD (1995) A Microscale Electrospray Interface for

Online, Capillary Liquid Chromatography/Tandem Mass Spectrometry of Complex Peptide

Mixtures. Anal. Chem. 67:4549–4556

78. Gatlin CL, Kleemann GR, Hays LG, Link AJ, Yates JR 3rd. (1998) Protein identification at

the low femtomole level from silver-stained gels using a new fritless electrospray interface

for liquid chromatography-microspray and nanospray mass spectrometry. Anal. Biochem.

263:93–101

79. Shabanowitz J, Settlage RE, Marto JA, Christian RE, White FM, (2000). In Mass

Spectrometry in Biology and Medicine, ed. AL Burlingame, SA Carr, MABaldwin, pp. 163–

177. Totowa, NJ: Humana Press

80. Martin SE, Shabanowitz J, Hunt DF, Marto JA (2000) Subfemtomole MS and MS/MS

Peptide Sequence Analysis Using Nano-HPLC Micro-ESI Fourier Transform Ion Cyclotron

Resonance Mass Spectrometry. Anal. Chem. 72:4266–4274

81. Davis MT, Lee TD (1997) Variable flow liquid chromatography-tandem mass spectrometry

and the comprehensive analysis of complex protein digest mixturesJ. Am. Soc. Mass

Spectrom. 8:1059-1069

2. PROTEASE NEXIN-1 (PN-1)

2.Protease Nexin-1

37

CHAPTER 2

Conformational and biochemical characterization of a biologically

active rat recombinant Protease Nexin-1 expressed in E. coli

Rosaria Arconea, Alberto Chinalib, Nicola Pozzic, Maddalena Parafatia, Fabio Masetc, Concetta

Pietropaolob and Vincenzo De Filippisc

aDipartimento di Scienze Farmacobiologiche, Università di Catanzaro “Magna Graecia”, Italy; bDipartimento di Biochimica e Biotecnologie Mediche, Università di Napoli “Federico II”, Italy; cDipartimento di Scienze Farmaceutiche, Università di Padova, Italy.

Published in Biochim. Biophys. Acta. (BBA) 2009; 1794(4):602-14. INTRODUCTION

Glia-derived nexin or Protease nexin-1 (PN-1) is a 44 kDa glycoprotein expressed and secreted

by a variety of cells, including fibroblasts (1), myoblasts (2), vascular smooth muscle cells (3),

astrocytes (4), and neuronal cells (5). PN-1 belongs to the serpin super-family (6) and inhibits

several serine proteases, including thrombin, urokinase, tissue plasiminogen activator and plasmin

(7, 8) with a mechanism of suicide substrate mediated by the formation of a covalent complex with

the target protease (9). Once formed, the serpin-protease complex binds back to the cells and is

internalized and degraded, thus providing a localized mechanism for inhibiting and clearing the

protease from the extracellular environment (8). Through the production of inhibitors such as PN-1,

glial cells would be able to modulate the extent of neuronal migration and cause modifications

compatible with the early, target independent outgrowth of neurite. A fine localized balance

between neuronal proteolytic activity and PN-1 inhibition would sustain neurite elongation (8).

Although the molecular nature of the PN-1 inhibitory complex in native conditions remains

controversial, in vitro it consists of a 1:1 covalent protease-PN-1 complex and it appears to be stable

both in in reducing SDS-electrophoresis (7) and physiological conditions (10). Studies on the

distribution of PN-1 in various tissues have shown that it is a major serpin found in physiologic

amounts in the brain (5, 11), primarily secreted by glial cells (12) and differentially expressed

during neuronal differentiation (13). PN-1 has been found to inhibit thrombin-mediated neurite

outgrowth retraction (14) and in the protects neuronal cells from proteolytic damage and thrombin-

2.Protease Nexin-1

38

induced apoptosis in brain injury (15, 16). Recent data (17) suggest that PN-1 exerts biological

functions not mediated by its inhibitory activity on proteases.

During our studies aimed to dissect the structure-function relationships of PN-1, we approached

the task of devising an efficient system to produce large amounts of recombinant protein, either in

native or mutated form. To this aim, a cDNA coding for rat PN-1 was isolated by RT-PCR and

cloned in vectors exploiting an inducible T7 RNA polymerase and able to express PN-1 either as a

His-tag fusion protein (pE15.b plasmid) or as a mature protein (pT7.7 plasmid). The recombinant

protein (rPN-1) was expressed in E. coli BL21(DE3)pLysE strain, purified to homogeneity,

characterized for its chemical identity, as well as for its conformational and biochemical properties

and tested for its biological activity in numerous functional assays. rPN-1 was also expressed in an

inducible eukaryotic expression system, using HeLa Tet-off cells.

Our results show that E. coli is a convenient expression system for obtaining fully active rPN-1

in sufficiently high yields for structural and functional studies.

MATERIALS AND METHODS

Construction of rat PN-1 vector. Total RNA was purified from rat cortical astrocytes according to

(18), dissolved in water (1 µg/µl) and stored at -20°C. Ten pmol of random hexamers were mixed

with 1 µg of total RNA in a final volume of 10 µl. The mixture was heated for 5 minutes at 70 °C,

then kept at room temperature. Subsequently, 4 µl of first-strand buffer (5x concentrated; Gibco

BRL), 2 µl of 100 mM dithiothreitol, 1 µl of dNTPs solution (10 mM each), 2 µl water and 1 µl

(2.5 U) of Mo-MuLV reverse transcriptase (Gibco, BRL) were added and incubated 10 minutes at

room temperature and 45 minutes at 37 °C. The reaction was terminated by adding 80 µl of water

and heating at 94 °C for 5 minutes. Two µl of this mixture were used for PCR amplification of

cDNA encoding PN-1. PCR primers were synthesized utilizing the published sequence of cDNA

encoding rat glia-derived nexin (19). The sense primer, 5' CTC GTC TGA ATT CAT GAA TTG

GCA TTT TCC C3' includes 11 nucleotides adjacent to the sequence encoding the first amino acid

of the signal peptide (-11 to +18, numbering starts from the adenine of the ATG triplet

corresponding to the translation initiation site) and containing EcoRI restriction site; the antisense

primer, 5' CTC ACT ATC TAG AGG CTT GTT CAC CTG CCC C3', complementary to the 3’-

untraslated region (UTR), position +1423 and containing a XbaI restriction site. The reaction was

carried out in a total volume of 50 µl containing 20 pmol of each PCR primer, 5 µl of Taq

polymerase buffer (10x, Perkin Elmer Cetus), 2 µl of 2.5 mM of each dNTP and 2 U of Taq

2.Protease Nexin-1

39

polymerase (Perkin Elmer Cetus). The PCR amplification was carried out through 5 cycles of

denaturing (94°C, 1 minute), annealing (57°C, 1.5 minute) and extension (72°C, 2 minutes),

followed by further 35 cycles using a different annealing temperature (60°C, 1.5 minute). The final

extension was at 72°C for 10 minutes. The 1451-bp PCR product containing the cDNA encoding

PN-1 was recovered by agarose gel electrophoresis purified and ligated into pGEM-T vector

(Promega). The resulting plasmid pGEM-T-PN-1 was controlled by restriction mapping and the

cloned cDNA was verified by nucleotide sequence analysis (20). A NdeI restriction site was

introduced at the serine-20 codon (numbering start at the methionine-1 of PN-1 leader peptide), in

frame with the remaining coding sequence, by site-specific mutagenesis directed by a synthetic

oligonucleotide carrying the appropriate nucleotide substitutions, used as a primer in a PCR on

pGEM-T-PN-1 DNA.

For production of polyclonal antibodies against PN-1, the PCR fragment was excised by NdeI-

XbaI restriction endonucleases and cloned in the pET-15b vector (Novagen, UK), which allows the

expression of a Hexa-His-tagged protein easily purified by affinity chromatography on a nickel-

agarose column. The resulting vector was named pET-15b-PN-1. Polyclonal antiserum was raised

in rabbit against affinity purified His-tagged PN-1 (21).

To produce a large amount of mature PN-1, a NdeI-SalI fragment (1170 bp) was excised from

pET-15b-PN-1 DNA and inserted in a pT7.7 plasmid (22). The resulting vector was named pT7.7-

PN1. For in vitro transcription and translation of PN-1 cDNA, a SalI restriction site was introduced

in the polylinker region of pGEM-T-PN-1 plasmid by PCR mutagenesis and the SalI-XbaI fragment

containing the PN-1 cDNA region was excised and cloned in the pGEM-4Z vector, resulting in a

pGEM-4Z-rPN-1 vector.

To obtain an inducible expression system of PN-1 in eukaryotic cells, a 1212 bp EcoRI-XbaI

fragment, including the sequence coding for the signal peptide, was cloned in the pTRE vector

(Clontech, Cambridge, UK). The resulting plasmid, pTRE-PN-1, was further modified by inserting

a Kozak consensus sequence (23) to optimize the translation in eukaryotic cells. The final vector

was indicated as pTRE-Kozak-rbs-PN-1. All the constructs were controlled by restriction mapping

and DNA sequence analysis.

Expression in E. coli, renaturation and purification of rPN-1. The pET-15b-PN-1 or pT7.7-PN-

1 expression vectors were introduced in E. coli BL21(DE3)pLysE strain which expresses the T7

RNA polymerase under the inducible lac UV5 promoter. Bacterial growth, isolation of inclusion

bodies and extraction of proteins were performed as previously reported (24). The in vitro refolding

of rPN-1 was obtained through a two-step dialysis; first step against a low concentration of the

2.Protease Nexin-1

40

denaturing agent (2 M GdnHCl, 50 mM Tris-HCl pH 8.0) containing 2 mM DTT. The second

dialysis step was performed against 50 mM Tris-HCl pH 8.0, 0.15 M NaCl, 1 mM DTT, 10%

glycerol. The refolded extract was centrifuged at 15000 g for 45 min to remove insoluble material

and loaded at 1 ml/min onto a HiTrap Heparin HP column (5 ml, GE Healthcare Life Science)

connected to a FPLC system (GE Healthcare Life Science). The column was equilibrated in the

dialysis buffer and after 10 volumes washing, the elution was performed at 4 ml/min by applying 10

volumes of a linear gradient to 1.5 M NaCl and collecting 0.5 ml fractions. Protein analysis was

performed by SDS-PAGE and Western blotting and fractions containing rPN-1 were pooled,

aliquoted and stored at -20°C.

The Hexa-His-PN-1 protein was purified from bacterial lysate using nickel agarose (Ni-NTA)

resin (Qiagen, Germany) according to the manufacturer’s instructions. The concentration of total

protein was determined by a Bio-Rad protein assay (25) using bovine serum albumin as a standard.

Twelve percent polyacrilamide SDS-PAGE was performed as described by Laemmli (26). After

electrophoresis, proteins were stained by Coomassie brilliant blue R-250 or electrophoretically

transferred to nitrocellulose filters (BA85; Schleicher & Schull). Blots were probed with rabbit

antisera against PN-1 followed by incubation with a peroxidase-conjugated anti-rabbit IgG in PBS,

containing 5% dry milk and developed according to the Enhanced Chemioluminescence (ECL)

technique.

Analytical techniques

RP-HPLC. The purity of rPN-1 preparations was checked by loading an aliquot (100 µl) of rPN-1

solution (0.12 mg/ml) onto a Vydac (The Separation Group, Hesperia, CA) C4 column (4.6 x 150

mm, 5µm particle size), eluted with a linear acetonitrile-0.078% TFA gradient at flow rate of 0.8

ml/min. The absorbance of the effluent was recorded at 226 nm.

Mass spectrometry. Accurate molecular weight determination of rPN-1 was obtained with a

Mariner ESI-TOF high-resolution mass spectrometer (Perseptive Biosystems, Stafford, TX).

Tipically, RP-HPLC purified rPN-1 (5 µg) was lyophilized, dissolved in 1:1 water:acetonitrile

mixture (20 µl), containing 1% (by vol.) formic acid, and then analyzed by mass spectrometry,

obtaining mass values in agreement with the expected amino acid composition within 0.01% mass

accuracy.

Two-Dimensional Electrophoresis. 2D electrophoresis was conducted on a BioRad Protean IEF-

Cell apparatus, using essentially the procedure provided by the manufacturer (27). For the first

dimension, an immobilized pH-gradient (IPG) strip (pH 3-10) (Biorad-1632009) was incubated

with 300 µl of rehydration buffer (i.e., 0.1% (w/v) CHAPS, 0.1% (w/v) Bio-Lyte 3-10, 8M urea, 0.1

2.Protease Nexin-1

41

M DTT, and 0.001% (w/v) Bromophenol Blue) containing the lyophilised rPN-1 sample (5 µg) and

5 µl of protein standard (BioRad, 161-0320). Rehydration was carried out at 20±0.5 °C for 15 hours

and applying a constant potential of 50 V. Prior to isoelectrofocusing (IEF), desalting of the strip

was achieved by applying a potential of 250 V for 15 min. IEF was conducted at 10 kV·h for 3

hours and then at 10 kV·h, up to 60 kV. Finally, the strip was frozen at –80°C. To reduce protein

disulfide bonds eventually present, the strip was then incubated under gentle stirring for 10 min

with 6 ml of 0.375 M Trs-HCl buffer, pH 8.8, 6 M urea, 2% SDS, 20% glycerol (buffer A),

containing 2% DTT. The reducing buffer was discarded and the strip incubated for 10 min with

buffer A containing 2.5% iodoacetamide, for blocking unreacted DTT. The second dimension was

carried out by loading the strip on a (20 x 20 cm) SDS-PAGE vertical slab (26) (stacking gel: 4%

acrylamide; running gel: 12% acrylamide) run overnight at a constant current of 10 mA. The gel

was stained using a modified silver staining protocol (28).

Spectroscopic Measurements

Determination of protein concentration. The concentration of PN-1 was determined by ultraviolet

(UV) absorption at 280 nm on a double beam model Lambda-2 spectrophotometer from Perkin-

Elmer. The extinction coefficient at 280 nm was calculated using a molar absorption coefficient of

1,280 M-1·cm-1 for tyrosine, 5,690 M-1·cm-1 for tryptophan, 120 M-1·cm-1 for disulfides (29), and

taken as 0.83 mg -1·cm 2. The concentration of thrombin was determined either by UV absorbance

at 280 nm, using a molar absorption coefficient of 65770 M-1·cm-1 (30) or, alternatively, by titration

of thrombin active-site with hirudin (31).

Circular dichroism. CD spectra were recorded on a Jasco model J-810 spectropolarimeter equipped

with a thermostated cell-holder connected to a NesLab (Newington, NH) model RTE-111 water-

circulating bath. Far- and near-UV CD spectra were recorded at 20±0.5 °C in 50 mM Tris-HCl

buffer, pH 8.8, containing 0.6 M NaCl and 10% (v/v) glycerol, using a 1- or 10-mm path length

quartz cells in the far- and near-UV region, respectively.

Fluorescence. Emission spectra were recorded at 20±0.5 °C on a Perkin-Elmer spectrofluorimeter

model LS-50B, equipped with a thermostated cell-holder connected to a Haake F3-C water-

circulating bath. Protein samples in 50 mM Tris-HCl buffer, pH 8.8, containing 0.6 M NaCl and

10% (v/v) glycerol were excited at 280 or 295 nm, using an excitation/emission slit of 5/10 nm.

Heparin binding to rPN-1. For measuring heparin binding to rNP-1, a Jasco model FP-6500

spectrofluorimeter, equipped with a Peltier model ETC-273T temperature control system, was used.

Excitation and emission wavelengths were at 280 and 341 nm, respectively, using an

2.Protease Nexin-1

42

excitation/emission slit of 10/10 nm. During titration experiments, the increase of fluorescence

intensity at 341 nm was recorded as a function of heparin concentration. For all measurements, the

Long-Time-Measurement software (Jasco) was used. Under these conditions, photobleaching of

Trp-residues was essentially absent. To a solution of rPN-1 (2 ml, 130 nM) in 5 mM Tris-HCl

buffer pH 7.5, 0.15 M NaCl, 0.1% PEG-8000, were added under gentle magnetic stirring aliquots

(2-8 µl) of a stock solution (146 U/ml) of unfractionated porcine heparin (Calbiochem). Of note, 1

U of heparin is an amount equivalent to 2 µg of pure heparin, having an average molecular weight

of 14.500 Da (32). Fluorescence intensities were corrected for dilution (2-3% at the end of the

titration) and subtracted for the signal of the ligand (i.e., heparin) at the indicated concentration.

Molecular Modelling and Computational Methods. The structure of rPN-1 in the active and

latent form was modeled on the corresponding crystallographic structures of PAI-1 (PDB codes:

1dvmA, for the active form, and 1dvnA, for the latent form) (33), that displays high sequence

similarity with PN-1 (19). The three-dimensional model of rPN1 was obtained using the Swiss-

Model automated comparative protein modeling server (http://swissmodel.expasy.org/SWISS-

MODEL.html) (34). Accessible surface area (ASA) calculations were carried out by using a

computer program available on-line at the site http://molbio.info.nih.gov/structbio/basic.html (35).

Alignment of the rat PN-1 sequence (SwissProt code: P07092) with that of the human PAI-1

(SwissProt code: P05121) was carried out using the computer program Clustal W (vs. 1.83)

available on-line at the site http://www.ebi.ac.uk/clustalw (36). The theroretical pI value of rPN-1

was calculated using the software ProtParam, available on-line at the site http://www.expasy.ch/cgi-

bin/protparam. N-Glycosylation sites were predicted using the program NetNGlyc, available on-line

at the site http://www.cbs.dtu.dk/services/NetNGlyc/.

Thrombin-rPN-1 complex formation monitored by SDS-PAGE. rPN-1 was mixed with natural

thrombin (Calbiochem) (2:1 molar ratio) at room temperature (22±1°C) in in 5 mM Tris-HCl buffer

pH 7.5, 0.15 M NaCl, 0.1% PEG-8000. Samples were taken at time intervals, mixed immediately

with reducing SDS-loading buffer, and heated for a further 3 min at 100°C. SDS-gel electrophoresis

was performed in a 12% acrylamide SDS-gel (26) and proteins were stained with Coomassie

Brilliant Blue R-250.

Thrombin inhibition assays. In the absence of heparin, the inhibition of natural thrombin by rPN-1

was determined at 25±0.5°C by a discontinuous assay procedure similar to that described for

thrombin inhibition by antithrombin-III (32). Briefly, the protease (1 nM) was incubated in the

2.Protease Nexin-1

43

presence of rPN-1 (40 nM) in 300 µl (final volume) of 5 mM Tris-HCl, 0.15 M NaCl, 0.1% PEG

8000, pH 7.5. The residual protease activity was determined at time intervals by diluting 50 µl of

the reaction mixture into 950 µl of the same buffer containing the chromogenic substrate (D)-Phe-

L-pipecolyl-L-Arg-p-nitroanilide (Sigma) (S-2238, 93 µM). The inhibition of p-nitroaniline release

from S-2238 was monitored by measuring the absorbance at 405 nm, using a molar absorption

coefficient for p-nitroaniline of 9920 M-1·cm-1. The concentration of S-2238 was determined by

measuring the absorbance at 342 nm, using a molar absorption coefficient of 8270 M-1·cm-1 (37).

The per cent residual thrombin concentration was determined as the ratio of vi/v0, where vi and v0

are the initial velocities of thrombin-catalyzed substrate hydrolysis in the presence or absence of

rPN-1. The % active thrombin was plotted as a function of time and the data fitted to equation 1,

from which pseudo a first-order association constant, ka, could be estimated (see Data Analysis).

In the presence of heparin, thrombin inhibition assays were carried out at 25±0.5°C by a

continuous assay procedure. rPN-1 (2.0 nM) was incubated for 15 min at the same temperature with

S-2238 (85.7 µM) and increasing concentrations (from 100 pM to 100 µM) of unfractionated

porcine heparin (Calbiochem) in 950 µl of 5 mM Tris-HCl, 0.15 M NaCl, 0.1% PEG 8000, pH 7.5.

The reaction was started by addition of 50 µl thrombin, up to a final concentration of 50 pM. After

rapid mixing (5 s), substrate hydrolysis was immediately recorded by measuring the increase of the

absorbance at 405 nm. Progress curves were fitted to equation 2 and analyzed as detailed below.

Data analysis. Either in the absence or presence of heparin, the inhibition of thrombin by PN-1 has

been previously shown to conform to the mechanism reported in Scheme 1 (32, 38):

k1 k2 E + I ↔ EI* → EI (Scheme 1)

k-1

where the protease, E, reversibly binds the inhibitor, I, to form the encounter complex, EI*, that

irreversibly converts into the stable protease-serpin complex, EI, with a rate constant k2. KEI is the

equilibrium constant for the noncovalent complex formation, KEI = k-1/k1, where k1 and k-1 are the

association and dissociation rate constants, respectively. If the equilibrium E + I ↔ EI* is rapid

compared with the rate of formation of EI (i.e., k2 << k -1) and if (I) << KEI (i.e., when (EI*) is

negligible), then the reaction follows simple second-order kinetics and appears to be a one-step

irreversible process:

ka E + I → EI (Scheme 2)

2.Protease Nexin-1

44

where ka is the second-order rate constant for the formation of nondissociating complex EI. From

Scheme 2 it follows that dE/dt = -ka (E)·(I). Integration of equation 1 under pseudo first-order

conditions (i.e., (I) > 10·(E)), yields equation 1:

(E)t = (E)0·exp(-kobs·t) (1)

where (E)0 and (E)t are the enzyme concentrations at time zero and t, respectively, and kobs is the

pseudo first-order rate constant given by kobs = ka·(I)0, where (I)0 is the serpin concentration. The

values of (E)t were determined by the discontinuous assay method (32) (see above), plotted as a

function of reaction time and fitted to equation 1, yielding kobs as a fitting parameter. In the

discontinuous method, thrombin was incubated with PN-1 and samples of the reaction mixture were

taken at various times. Measurements of residual enzyme activity were conducted by monitoring the

initial rate of chromogenic substrate hydrolysis, S (i.e., S-2238), that yields the product P (i.e., p-

nitroaniline). Finally, from the knowledge of serpin concetration, (I)0, the value of ka could be easily

calculated. Here we used this procedure to analyze thrombin inhibition data by rPN-1 in the absence

of heparin.

In the presence of heparin, the affinity of PN-1 for thrombin increases by about 1000-fold,

thereby preventing an accurate estimation of ka using the discontinuous method (32). To overcome

this limitation, we used a continuous assay method (see above) in which continuous monitoring of

thrombin-PN-1 reaction was conducted in the presence of S-2238. It has been previously shown

(38-40) that thrombin inhibition by PN-1 follows a reversible slow binding process according to the

mechanism depicted in Scheme 3:

k1 Km kcat EI* ↔ I + E + S ↔ ES → E + P (Scheme 3)

k-1

Under pseudo first-order conditions, and with negligible consumption of substrate (<10%), the

amount of product, P, as a function of time, t, is given by the exponential equation (39, 40):

(P)t = vs·t + ((v0 – vs)/kobs)·(1 – exp(-kobs·t)) (2)

where kobs is the pseudo first-order rate constant for product formation and v0 and vs are the initial

and steady-state velocities, respectively. Estimates of kobs, v0, and vs were obtained by fitting to

equation 2 the data of product generation obtained at different heparin concentrations, from 100 pM

2.Protease Nexin-1

45

to 100 µM. At each heparin concentration, the second-order rate constant, k1, can be related to kobs,

v0, vs, and Km by equation 3:

k1 = (kobs·(1 - vs/v0))/(I)·(1 + (S)/Km) (3)

where Km is the Michaelis constant for thrombin-catalyzed hydrolysis of S-2238 at 25°C,

previously estimated as 3.0±0.3 µM (41). When the value of k1 was plotted against the

concentration of heparin, a bell-shaped curve is obtained. Such a curve is empirically described by

equation 4:

k1 = k1°/(1 + (H)/K1 + K2/(H)) (4)

where k1° is the maximum value of k1 obtained at an optimal heparin concentration, and K1 and K2

are empirical constants that represent the concentrations of heparin at which the half-maximal value

of k1 is observed. Estimates of k1°, K1 and K2 were obtained by fitting the data of k1 vs. (H) to

equation 4.

Cell Cultures, DNA transfections and immunocytochemistry. NB2A mouse neuroblastoma cells

(kindly provided by Dr V. De Franciscis, CNR, Italy) and human HeLa Tet-Off cells (Clontech)

were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal

bovine serum (FBS), or 10% Tet system approved fetal bovine serum (Clontech) for the HeLa Tet-

off cells. All culture media were also supplemented with 1.5 mM L-glutamine, 50 IU/ml penicillin

and 50 µg/ml streptomycin. The cells were maintained at 37°C in a 5% CO2 atmosphere and were

subcultured twice a week by 3-5-fold dilution with culture medium.

For DNA transfection, cells were seeded subconfluent on cover slips and in 60-mm plates; 24 hr

after plating, transfection was carried out using 1-7 µg of pTRE-Kozak-rbs-PN-1 DNA using the

Lipofectamin 2000 reagent technique (Invitrogen, Life Technologies) according to the

manufacturer’s instructions; 18 hours after transfection, cells were washed with culture media and

maintained in the presence (10 ng/ml) or absence of doxycycline hydrochloride (Sigma-Aldrich,

Italy). At 24, 48 and 72 h following depletion of the antibiotic, cells were harvested and culture

media collected for western blotting analysis and the cover slip cultures processed for indirect

immunofluorescence staining.

Subconfluent cells on glass cover slips, were fixed with cold 3.7% formaldehyde (Fluka, France) in

PBS with Ca2+ (0.1g/lt) and Mg2+ (0.1g/lt) for 20 min, permeabilized with 0.1% Triton X-100 in

2.Protease Nexin-1

46

PBS for 5 min. The cover slips were incubated for 2 h at room temperature with polyclonal PN-1

antibodies (21) at a 1:20 dilution in a PBS-0.2% gelatin solution, washed three times for 5 min with

0.5 NaCl, 0.02 M NaPO4 pH 7.4 and finally, with PBS. An anti-rabbit IgG (Fc-specific)

fluorescein-isothiocyanate (FITC)-conjugated (Jackson Immunoresearch Laboratories) was used as

secondary antibody, at 1:200 dilutions in PBS-0.2% gelatin, for 1 h at room temperature.

Cover slips were washed and mounted with Mowiol (Hoechst, FRG) on microscope slides and

observed with a Zeiss Axiovert 10 photomicroscope using a 63xPlan-Apochromat lens and

fluorescein (450-490 nm, FT515-LP565) filters. After several washings, nuclear staining was

performed. Cover slips were incubated for 20 min with 385 nM 4',6-diaminide-2-phenylindole

(DAPI, Sigma) (42), then washed and mounted with Mowiol (Hoechst, FRG) on microscope slides.

Cells were observed by a Zeiss Axiovert 10 photomicroscope using a 100x oil immersion objective

Plan-Apochromat lens and DAPI (379-401 nm, 435-485 nm) filters.

For F-actin staining, cells were fixed, permeabilized, washed with PBS (with Ca2+ and Mg2+) and

then incubated in the dark for 20 min at room temperature with 0.768 µM rhodamine phalloidin

(Molecular Probes) in PBS. Cells were rapidly washed for three times with buffer, mounted with

Mowiol and observed with a Zeiss Axiovert 10M photomicroscope using a 63xPlan-Apochromat

lens and rhodamine (546 nm, 590 nm) filters.

Induction of neurite outgrowth in NB2A neuroblastoma cells. NB2A cells were grown

subconfluent on cover slips and cultured for 24 h in DMEM supplemented with 10% FBS. Cells

were then differentiated to a neuronal morphology by serum deprivation (0% FBS), or kept

undifferentiated in 0.8% FBS, for 4 hr. Cells were exposed to 50 nM rPN-1 alone or in combination

with 2 nM thrombin, and after 3hr incubation were fixed and processed for immunocytochemistry.

RESULTS Expression, purification, and chemical characterization. A cDNA fragment coding for rat PN-1

was isolated by RT-PCR using specific synthetic oligonucleotides on total RNA from rat cortical

astrocytes (18). The PCR DNA product was inserted in pGEM-T vector. The resulting vector,

pGEM-T-PN-1, was sequenced and found to contain the rat PN-1 cDNA, as reported in the

NCBI/Gene Bank (accession number M17784).

Site specific mutagenesis was performed to introduce a NdeI restriction site which would allow

us to excise a NdeI-XbaI DNA fragment coding for the mature PN-1. Thereafter, the mature PN-1

2.Protease Nexin-1

47

cDNA was cloned in pET-15b and in pT7.7 vectors, resulting in pET-15b-PN-1 and pT7.7-PN-.1

vectors, respectively. E. coli cells, strain BL21(DE3)pLysE, were transformed with pET-15b-PN-1

DNA and produced the His-tagged PN-1. Bacterial culture was processed as detailed in Materials

and Methods and the fused rPN-1 protein was purified by IMAC on a nickel-loaded column (see

below). The purified protein was used to produce polyclonal anti-PN-1 antibodies in rabbit, that

were able to recognize PN-1 expressed in rat oligodendrocytes (21).

The recombinant protein was efficiently expressed in E. coli following IPTG induction (Fig. 1A,

+ IPTG). As also observed for the His-tagged form, mature rPN-1 was delivered to the inclusion

bodies. After solubilization of inclusion bodies (24), rPN-1 was refolded in vitro by dialysing

against Tris buffer, pH 8.8, containing moderate concentrations of denaturants and reducing agents

(i.e., 2 M Gnd-HCl and 2 mM DTT) and then against the same buffer, without denaturant,

containing 0.15 M NaCl, 1 mM DTT, and 10% glycerol. Addition of DTT prevented oxidation of

the three free Cys-residues in the PN-1 structure during refolding (43), while glycerol was added to

increase protein solubility. The refolding mixture was purified by affinity chromatography on a

heparin-Sepharose column, whereby rPN-1 eluted as a single peak at 0.8 M NaCl. Expression,

refolding and purification of rPN-1 was monitored by SDS-PAGE (Fig. 1A) and Western blot (Fig.

1B) analyses, showing the presence of a single intense band, at the molecular weight expected for

rPN-1 (i.e., 42 kDa) and immunoreactive against PN-1 antibodies. Strikingly, the expression and

purification procedure herein reported allowed us to recover about 3 mg of homogeneous (see

below) rPN-1 per liter of cell culture.

The homogeneity of rPN-1 was also checked by reversed phase (RP) HPLC, demonstrating that

the protein elutes as a single peak on a C4 analytical column (Fig. 1C). Accurate molecular weight

determination of rPN-1, carried out by ESI-TOF mass spectrometry (Inset to Fig. 1C), yielded a

mass value of 41774.4 ± 5 a.m.u. in close agreement with that expected from the amino acid

composition of rPN-1 (average mass: 41777.6 u.m.a.) (19). The isoelectric point, pI, of rPN-1 was

estimated by two-dimensional electrophoresis (Fig. 1D), yielding a pI value of 9.7, consistent with

the theoretical value deduced form the amino acid composition, pI 9.72. The minor spot at pI 9.6

likely originates from the artefactual cyanylation of some Lys-residues that sometimes occurs

during long-time incubation (i.e., 15 h) of the IEF strip with protein samples in the presence 8 M

urea (see Methods). Within the limits of the analytical methods used, these results provide strong

evidence for the homogeneity and chemical identity of our recombinant rPN-1 preparation, that was

subsequently used for conformational and functional characterization.

2.Protease Nexin-1

48

68

97

68

97

45

30

20

kDa

Heparin-Sepharosepurified rPN-1

rPN-1

rPN-1

RE 32 34 36 38 40 42- +

45

30

20

45

30

20

kDa


68

97

68

97

rPN-1rPN-1

rPN-1rPN-

A

B

IPTG

45

30

20

45

30

20

kDa


rPN-1rPN-1

rPN-1rPN-1

RE 32 34 36 38 40 42- +

45

30

20

45

30

20

kDa


68

97

68

97

rPN-1rPN-1

rPN-1rPN-

A

B

IPTG

4 8 12 16 20 24

Abs

orba

nce

at 2

26 n

m (

)

Retention Time (min)

41200 41600 42000 42400

375

500

625

750

875

% In

tens

ity

Mass (a.m.u.)

41761.9 ± 5 C

kDa

976645

30

20

14

4.5

5.1

5.5

5.9

6.6

7.0

8.5

10.0

12

3

4

5

6

D

rPN-1rPN-1rPN-1

kDa

976645

30

20

14

4.5

5.1

5.5

5.9

6.6

7.0

8.5

10.0

12

3

4

5

6

D

rPN-1rPN-1rPN-1rPN-1rPN-1rPN-1rPN-1

Figure 1. Expression in E. coli and purification of rPN-1 monitored by SDS-PAGE and Western blotting. (A) SDS-gel elecrophoresis was conducted in a 12% acrylamide gel and proteins were visualized by Coomassie blue staining. From the left to the right: 20 µl of bacterial cell lysate from BL21(DE3) cells transformed with the pT7.7/PN1 vector DNA in the absence (-) and in the presence (+) of 0.4 mM IPTG; 100-ng aliquot of the renaturation mixture (RM) of rPN-1; 10-µl aliquots of the chromatographic fractions (32-42) eluted from the heparin-Sepharose column; molecular weight protein markers are indicated on the left-hand side. (B) Western blot analysis of the same samples run in the SDS-gel, as in panel A. rPN-1 was revealed by immunostaining and enhanced chemioluminescence. (C) RP-HPLC analysis of heparin-Sepharose purified rPN-1. An aliquot (10 g) of rPN-1 was loaded onto a C4 column (4.6 x 150 mm) eluted with an acetonitrile-0.1% TFA gradient (---) from 10 to 60 % in 15 min. (Inset) Deconvoluted ESI-TOF mass spectrum of RP-HPLC purified rPN-1. (D) 2D-gel electrophoresis of purified rPN-1. The first dimension was run on an immobilized pH-gradient strip, from 3 to 10, while the second dimension on a 12% acrylamide SDS-gel, as detailed in the Methods. Proteins were visualized by silver staining. The pI value of rPN-1 was estimated as high as 9.6-9.7, using the BioRad pI protein standard for 2D electrophoresis, that contains (1) hen egg white conalbumin (pI 6.0, 6.3, 6.6); (2) BSA (pI 5.4, 5.5, 5.6); (3) bovine muscle actin (pI 5.0, 5.1); (4) rabbit muscle GAPDH (pI 8.3, 8.5); (5) bovine carbonic anhydrase (pI 5.9, 6.0); (6) soybean trypsin inhibitor (pI 4.5). Model building. The molecular model of rPN-1 was built by homology modelling technique on the

crystallographic structures of PAI-1 in the active form (PDB code: 1dvmA) (33), with which PN-1

shares 42% sequence identity and 65% similarity (see Fig. 2A) (19). The resulting model structure

(Fig. 2B) resembles that typical of the serpin fold comprised of a bundle of 9 α-helices (A-I) and a

2.Protease Nexin-1

49

β-sandwich composed of three β-sheets (A-C). The reactive center loop (RCL), i.e., the loop which

is cleaved by the target protease, extends at the top of the molecule, spans from Asp330 at the C-

terminal end of strand 5A to Arg355, and contains the scissible bond Arg345-Ser346.

In our model, four lysines (i.e., Lys71, Lys74, Lys75 and Lys78) are gathered on the exposed

surface of helix D, which is followed by a surface loop containing three extra lysines (Lys83, Lys84

and Lys86), well positioned with helix D to form a large positively charged cluster, corresponding

to the proposed heparin-binding site on PN-1 structure (44). Notably, PN-1 contains seven positive

charges in the 71-86 segment; in the corresponding region, antithrombin III (AT-III) and heparin

cofactor II (HC-II) contain five positive charges and PAI-1 only three positive charges and one

negative charge. Likely, this stronger electrostatic potential may explain the higher affinity of PN-1

for heparin (Kd = 0.8±0.6 nM) (39, 40) compared to that of other serpins, like AT-III for instance

(Kd = 8.8±1.3 nM) (32).

Figure 2. Structural model of rPN-1. (A) Clustal-W alignment of the rat PN-1 (SwissProt code: P07092) and human PAI-1 (SwissProt code: P05121). (B) Schematic representation (solid ribbon, light gray) of the three-dimensional structure of rPN-1, as obtained form molecular modelling studies. The model was generated on the crystallographic coordinates of PAI-1 in the native/active form (1dvm_A.pdb), as detailed in the Methods. The side-chains of the Trp- and Cys-residues are indicated.

PN-1 sequence contains three cysteines (i.e., Cys117, Cys131, and Cys360), that are not

conserved in the template PAI-1 structure (33). In the model presented here, all cysteines are too far

in the three-dimensional structure to form any disulfide bridge. Moreover, ASA calculations suggest

P05121 VHHPPSYVAHLASDFGVRVFQQVAQASKDRNVVFSPYGVASVLAMLQLTTGGETQQQIQA 60

P07092 SQLNSLSLEELGSDTGIQVFNQIIKSQPHENVVISPHGIASILGMLQLGADGRTKKQLST 60

: . : .*.** *::**:*: ::. ..***:**:*:**:*.**** :.*.*::*:.:

P05121 AMGFKIDDKGMAPALRHLYKELMGPWNKDEISTTDAIFVQRDLKLVQGFMPHFFRLFRST 120

P07092 VMRYNVN--GVGKVLKKINKAIVSKKNKDIVTVANAVFVRNGFKVEVPFAARNKEVFQCE 118

.* :::: *:. .*::: * ::. *** ::.::*:**:..:*: * .: .:*:.

P05121 VKQVDFSEVERARFIINDWVKTHTKGMISNLLGKGAVD-QLTRLVLVNALYFNGQWKTPF 179

P07092 VQSVNFQDPASACDAINFWVKNETRGMIDNLLSPNLIDSALTKLVLVNAVYFKGLWKSRF 178

*:.*:*.: * ** ***..*:***.***. . :* **:******:**:* **: *

P05121 PDSSTHRRLFHKSDGSTVSVPMMAQTNKFNYTEFTTPDGHYYDILELPYHGDTLSMFIAA 239

P07092 QPENTKKRTFVAGDGKSYQVPMLAQLSVFRSGSTKTPNGLWYNFIELPYHGESISMLIAL 238

..*::* * .**.: .***:** . *. . .**:* :*:::******:::**:**

P05121 PYEKEVPLSALTNILSAQLISHWKGNMTRLPRLLVLPKFSLETEVDLRKPLENLGMTDMF 299

P07092 PTESSTPLSAIIPHISTKTINSWMNTMVPKRMQLVLPKFTALAQTDLKEPLKALGITEMF 298

* *...****: :*:: *. * ..*. ******: ::.**::**: **:*:**

P05121 RQFQADFTSLSDQEPLHVAQALQKVKIEVNESGTVASSSTAVIVSARMAPEEIIMDRPFL 359

P07092 EPSKANFAKITRSESLHVSHILQKAKIEVSEDGTKAAVVTTAILIARSSPPWFIVDRPFL 358

. :*:*:.:: .*.***:: ***.****.*.** *: *:.*: ** :* :*:*****

P05121 FVVRHNPTGTVLFMGQVMEP 379

P07092 FCIRHNPTGAILFLGQVNKP 378

* :******::**:*** :*

W350

W219

W137

W174

W261

C117

C131

C360 •

R345

2.Protease Nexin-1

50

that only one Cys is buried in the protein core (i.e., Cys360), whereas the remaining two are highly

(i.e., Cys117) or moderately (i.e., Cys 131) solvent exposed. These conclusions are fully consistent

with earlier chemical modification studies with 5,5'-dithiobis-nitrobenzoic acid (43), conducted on

the human recombinant PN-1 (SwissProt entry code: P07093) and showing that under native

conditions all Cys-residues are in the reduced, free thiol state and accessible to solvent. Human PN-

1 shows 84% sequence homology with the rat protein (SwissProt entry code: P02092) (19) and both

proteins contain three cysteines in their sequence. Of these, the solvent exposed Cys117 and Cys131

are conserved, whereas the buried Cys360 of rat PN-1 is replaced by Phe in the human protein. On

the other hand, Cys209 of human PN-1 is substituted by Ser at the corresponding position, which is

rather exposed on the rat PN-1 structure. The presence of at least one sulphydryl group freely

accessible on the protein surface is consistent with the observation that disulfide-mediated

dimerization may occur on standing (43).

Conformational characterization. The conformational properties of rPN-1 were investigated by

CD in the far- and near-UV region and by steady state fluorescence spectroscopy. The far-UV CD

spectrum of rPN-1 (Fig. 3A) is typical of a protein containing a mixed α/β secondary structure, with

two shallow minima centered at 220 and 210 nm (45), and is similar in both shape and intensity to

those of AT-III (46) and plasminogen activator inhibitor-1 (PAI-1) (47), both proteins displaying

remarkable sequence similarity to PN-1 (19). The CD spectrum of rPN-1 in the near-UV region

shows a broad positive band in the 255-300 nm range (Fig. 3B). The contribution of the 19 Phe-

residues appears as a typical 6-nm spaced band system between 260 and 272 nm, while the positive

band at 292 nm is diagnostic of Trp-residue(s) embedded in a rigid protein environment (48).

Moreover, the near-UV CD spectrum of rPN-1 shows similar spectral features when compared

with that of AT-III (46), reflecting again the sequence and structural similarities existing between

these two proteins.

The fluorescence spectra of rPN-1 obtained after exciting the sample at 280 and 295 nm show a

red-shifted λmax value at ∼345 nm (Fig. 3C), indicating that the majority of the five Trp-residues in

the PN-1 structure are located in polar and, likely, solvent exposed environment (49). Moreover, the

absence of the contribution of Tyr-residues in the 280-nm spectrum indicates that there is an

efficient Tyr-Trp energy transfer, reflecting the compact structure of rPN-1 in the folded state.

Under strong denaturing conditions (5 M Gdn-HCl), the value of λmax is further shifted to 351 nm,

typical of polypeptide chains in the fully unfolded state, while the contribution of Tyr appears as a

weak, shallow band at ∼303 nm. Comparison of The fluorescence spectra reveal that rPN-1 has a

λmax value red-shifted by ∼10 nm, compared to that of the homologous serpins AT-III (46) and PAI-

2.Protease Nexin-1

51

1 (47), suggesting that the five Trp-residues in rPN-1 are located, on average, in a more polar

environment.

This conclusion is supported by the model structure of rPN-1 and by accessible surface area

calculations, showing that the majority of Trp-residues are solvent exposed (Fig. 2B). In particular,

the side-chains of Trp137, Trp219 and Trp350 are much more accessible than those of Trp174

and Trp261, buried in the protein core. For comparison, mature PAI-1 contains four Trp-residues, at

positions 86, 139, 175, and 262 (the numbering follows the rPN-1 sequence). Of these, three

(namely, Trp139, Trp175, and Trp262) are conserved in rPN-1 sequence (namely, Trp137, Trp174,

and Trp261). Analysis of the crystallographic structure of PAI-1 (45) indicates that Trp175 and

Trp262 are shielded from the solvent, whereas Trp86 and Trp139 are only moderately accessible

to water, in keeping with the lower λmax value determined in the fluorescence spectrum of PAI-1

(39).

Figure 3. Conformational characterization of rPN-1. Far- (A) and near-UV(B) CD spectra were taken at a protein concentration of 78 µg/ml. (C) Fluorescence spectra were recorded after exciting the rPN-1 samples (7 µg/ml) at 280 and 295 nm, or in the presence of 5 M Gnd-HCl, as indicated. All spectra were recorded at 20±0.5 °C in 50 mM Tris-HCl buffer, pH 8.8, containing 0.6 M NaCl and 10% (v/v) glycerol.

200 210 220 230 240 250

-10

-8

-6

-4

-2

0

2

4

[ θ] x

10-3

(deg

·cm2 ·d

mol-1

)

Wavelength (nm)

A

250 260 270 280 290 300 310 320

-20

0

20

40

60

80

[ θ]

(deg

·cm2 ·d

mol-1

)

Wavelength (nm)

B

C

300 330 360 390 420 450

0

1

2

3

4

5

6

Rel

ativ

e F

luor

esce

nce

Wavelength (nm)

λex 280 nm

5 MGnd-HClλex 280 nm

λex 295 nm

200 210 220 230 240 250

-10

-8

-6

-4

-2

0

2

4

[ θ] x

10-3

(deg

·cm2 ·d

mol-1

)

Wavelength (nm)

A

250 260 270 280 290 300 310 320

-20

0

20

40

60

80

[ θ]

(deg

·cm2 ·d

mol-1

)

Wavelength (nm)

B

C

300 330 360 390 420 450

0

1

2

3

4

5

6

Rel

ativ

e F

luor

esce

nce

Wavelength (nm)

λex 280 nm

5 MGnd-HClλex 280 nm

λex 295 nm

2.Protease Nexin-1

52

Functional studies

Heparin binding. In the presence of 1.7 µM unfractionated porcine heparin, the fluorescence

intensity of rPN-1 is increased by 25% (Fig. 4A), as already observed with AT-III (32), without

changes in the λmax value (i.e., the wavelength at which the intensity of fluorescence is highest).

Fluorescence titration of rPN-1 with heparin (Fig. 4B) displays saturation behaviour both at 30 and

130 nM serpin. In these conditions, the Kd value for the heparin-rPN-1 complex could not be

determined, likely because the concentration of rPN-1 used in these measurements largely exceeds

the Kd of heparin-rPN-1 complex, previously determined (i.e., 0.8±0.6 nM) (40). On the other hand,

rPN-1 concentrations in the Kd range would result into a fluorescence change too small to be

accurately measured. From the data reported in Fig. 4B, however, a 1:1 stoichiometry for the

binding of heparin to rPN-1 could be easily determined.

Figure 4. Binging of heparin to rPN-1 monitored by fluorescence change. (A) Fluorescence spectra of rPN-1 in the absence (---) and presence (___) of unfractionated porcine heparin (1.7 µM). Spectra were taken after exciting rPN-1 samples (80 nM) at 280 nm. (B) Titration of rPN-1 with heparin. The binding of heparin (0-7 M) to rPN-1 at 30 nM (o) and 130 nM (●) was monitored after exciting the samples at 280 nm and recording the increase of fluorescence at λmax (i.e., 341 nm) as a function of (heparin)/(rPN-1) ratio. Fluorescence change is expressed as (F0-F), where F0 and F are the fluorescence values of rPN-1 in the absence and presence of heparin, respectively. The stoichiometry of heparin binding to rPN-1 was obtained as the (heparin)/(rPN-1) ratio where the regression line (---) of the initial linear increase in fluorescence intersects the regression line (---) of maximum fluorescence change corresponding to saturation of the inhibitor (Olson et al., 1993). All measurements were carried out at 25±0.1 °C in 5 mM Tris-HCl buffer pH 7.5, 0.15 M NaCl, 0.1% PEG.

300 350 400 450 500

0

100

200

300

400

Rel

ativ

e F

luor

esce

nce

Wavelength (nm)

A B0 1 2 3 4

0

30

60

90

120

150

180

[Heparin]/[rPN-1]

F-F

0

2.Protease Nexin-1

53

Formation of thrombin-rPN-1 complex monitored by SDS-PAGE. The time-course kinetics of

rPN-1 binding to thrombin (molar ratio of 2:1) without heparin was monitored by SDS-gel

electrophoresis in the time range 5 s - 1 h (Fig. 5). Even after 5-s reaction, thrombin is almost

quantitatively sequestered to form a stable thrombin-rPN-1 complex (i.e., the acyl-enzyme

intermediate) migrating in the gel with an apparent molecular weight of 65 kDa, roughly given by

the sum of the molecular weight of rPN-1 (~42 kDa) and that of thrombin heavy chain (~32 kDa).

As the reaction time increases, the intensity of the 65-kDa band (cpx) further increases and

concomitantly that corresponding to the protease (thb) is barely detectable after 1-h reaction.

Interestingly, the intensity of the rPN-1 band (native) remains constant over time and no cleaved

serpin form even after 1-h reaction could be detected, with Coumassie Blue staining at least. The

cleaved form is generated after formation of the acyl-enzyme intermediate and subsequent

conformational change. When loop insertion in the β-sheet is not rapid enough to compete with

deacylation, then the reaction proceeds directly to the cleaved product (10, 50). The absence of the

cleaved form is usually taken as a good indication for 1:1 stoichiometry in complex formation (6).

With respect to this, the stoichiometry of the reaction is defined as the number of moles of serpin

needed to inhibit one mole of proteinase as a kinetically trapped complex. As shown in Fig. 5, a

weak band appears at a molecular weight slightly lower than that of the thrombin-rPN-1 acyl-

intermediate. This band likely corresponds to a thrombin cleaved form of the serpin-protease

complex (cpx*). Serpins, indeed, have been demonstrated to remarkably increase the

conformational flexibility of the protease bound to the serpin (51), with a resulting increased

susceptibility of the inhibited protease towards proteolysis by the active protease molecules in

equilibrium with the serpin-bound protease (50, 52).

Recombinant PN-1 expressed in E. coli failed to form a stable complex with the functionally

inactive thrombin mutant (Ser195Ala) produced earlier (53), in which the catalytic Ser195 was

replaced by Ala (unpublished Results). This finding is in keeping with the notion that Ser195 is

involved in the inhibitory mechanism of serpins (6, 51, 54).

2.Protease Nexin-1

54

Figure 5. SDS-PAGE analysis of the reaction mixture of rPN-1 with thrombin. At each time-point, to 22 µll of the reaction mixture, containing 2.4 µg of rPN-1 and 0.9 µg of α-thrombin (2:1 molar ratio), incubated at room temperature (22±1°C) in 5 mM Tris-HCl buffer pH 7.5, 0.15 M NaCl, 0.1% PEG-8000, were added 6 µl of pre-heated reducing SDS-loading buffer. After rapid mixing, samples were heated for a further 3 min at 100°C. The gel was stained with Coumassie Blue. Lane 1 is rPN-1 alone; lanes 2-9 correspond to 5-, 15-, 30-, 60-, 120-, 600-, 1200-, and 3600-s time points of the reaction; lane 10 is thrombin alone. Molecular-weight markers were loaded in the left-handed well. Reactants and products are labelled as follows: native, intact rPN-1; thb, thrombin (heavy chain); cpx, stable thrombin-rPN-1 complex; cpx*, cleaved complex. Thrombin inhibition by rPN-1 in the absence and presence of heparin. Thrombin inactivation

by rPN-1 in the absence of heparin was determined by the discontinuous assay method, according to

which the protease and the serpin were incubated at 25±0.5°C under pseudo-first order conditions

(i.e., (rPN-1) >> (trombin)) in Tris-HCl buffer, pH 7.5. The residual protease activity was

determined at time intervals by measuring the initial velocity of S-2238 substrate hydrolysis at 405

nm in the absence (v0) and presence (vi) of rPN-1. The per cent active thrombin was plotted as a

function of time (Fig. 6A) and the data fitted to equation 1, from which a figure of the pseudo first-

order association rate constant, ka, could be estimated as 1.1±0.1 x 106 M-1·s-1 (see Methods for

details). This value is fully consistent with those reported for natural, ka = 1.4 x 106 M-1·s-1 (39), and

recombinant PN-1 expressed in chinese hamster ovary cells, ka = 0.8 x 106 M-1·s-1 (43), or in yeast,

ka =1.4 x 106 M-1·s-1 (40).

In the presence of heparin, the affinity of rPN-1 for thrombin increases by 103-fold (39, 40) and

therefore the discontinuous method would not yield reliable ka values. To overcome this limitation,

inhibition assays were carried out at 25±0.5°C by continuously measuring over time inactivation of

thrombin by rPN-1 in the presence of S-2238 and increasing concentrations of heparin (100 pM–

100 µM) (32). The reaction was started by addition of thrombin (50 pM) and the progress curves of

the release of p-nitroanilide as a function of time were fitted to equation 2 (Fig. 6B). From these

data, the second-order association rate constant, k1, could be estimated at each heparin

82 3 4 5 6 71 9

thb

native

cpx

97

kDa

66

45

30

10

2.Protease Nexin-1

55

concentration. Plotting k1 values as a function of heparin concentration (Fig. 6C) yielded a typical

bell-shaped curve, similar to previous data reported with natural (39) or recombinant (40) PN-1.

The data were fitted to equation 4, yielding an optimal heparin concentration, (H)opt, at which the k1

value is maximal, k1° = 0.45±0.02 x 109 M-1·s-1. Notably, this value is very similar to that reported

for natural PN-1, k1° = 0.46±0.04 x 109 M-1·s-1 (39). For further increase of heparin concentration,

the cofactor seems to have an inhibitory effect on serpin-thrombin interaction, consistent with the

template model of heparin-accelerated inhibition of thrombin by serpins (32. The affinity of heparin

for PN-1 (Kd = 0.8±0.6 nM) (40) is much higher than that for thrombin (Kd = 0.69±0.06 µM) (55).

This large difference in affinities implies that the assembly of the ternary serpin-heparin-protease

complex, I·H·P, occurs predominantly as a bimolecular association between the serpin-heparin

binary complex, I·H, and the free protease, P, according to the template model reported in Scheme

4:

KIHP kIH I·H + P ↔ I·H·P → I·P + H (Scheme 4)

where KIHP is the equilibrium dissociation constant of the encounter I·H·P complex and kIH is the

first-order rate constant for the conversion of the I·H·P complex to the stable I·P complex and

release of heparin (32). According to this model, at heparin concentrations higher than (H)opt, the

free heparin not bound to rPN-1 can compete with the inhibitor-heparin complex, I·H, for the

binding to thrombin thereby reducing the reaction rate for the formation of thrombin-rPN-1

complex.

2.Protease Nexin-1

56

Figure 6. Thrombin inactivation by rPN-1 in the absence (A) and presence (B, C) of heparin. In the absence of heparin (A), the inhibition of α-thrombin by rPN-1 was determined by the discontinuous assay method, according to which the protease (1 nM) was incubated with rPN-1 (40 nM) in 5 mM Tris-HCl, 0.15 M NaCl, 0.1% PEG 8000, pH 7.5. The residual protease activity was determined at time intervals by measuring at 405 nm the inhibition of p-nitroanilide release from S-2238 (93 µM). The % active thrombin was plotted as a function of time and the data fitted to equation 1, from which a kobs value of 4.3±0.4·10-2 s-1 was obtained. If kobs = ka·(rPN-1), then the pseudo first-order association rate constant, ka, could be estimated as 1.1±0.1 x 106 M-1·s-1 (see Data Analysis for details). (B) In the presence of heparin, thrombin inhibition assays were carried out at 25 ± 0.5°C by continuously measuring over time inactivation of thrombin (50 pM) by rPN-1 (2.0 nM) in the presence of S-2238 (85.7 µM) and 1 (●), 10 (o), and 100 nM (▲) heparin. The release of p-nitroanilide as a function of time was fitted to equation 2, from which the values of kobs, v0 and vs were obtained as fitting parameters. From these values, the second-order association rate constant, k1, can be estimated at each heparin concentration, using equation 3. (C) Plot of k1 as a function of heparin. The data were fitted to equation 4, yielding the optimal (heparin) at which k1 value is maximal, k1°.

0 20 40 60 80 100 120

0

20

40

60

80

100

Act

ive

Thr

ombi

n (%

)

Time (sec)

A

0 20 40 60 80 100 120

0

20

40

60

80

100

Act

ive

Thr

ombi

n (%

)

Time (sec)

A

0 600 1200 1800 2400 3000

0

1

2

3

4

5

6

7

vi

[p-N

itroa

nilin

e] (

µM)

Time (sec)

vsB

0.0

0.1

0.2

0.3

0.4

0.5

543210-1

k 1

(nM

-1s-1

)

log [Heparin] (nM)

-2

C

2.Protease Nexin-1

57

Induction of neurite outgrowth by rPN-1 in neuroblastoma cells. Previous studies have shown

the relevance of protease-inhibitor balance in brain plasticity (56), and, more specifically, the role

of PN-1 and thrombin (14, 57). In particular, it has been demonstrated that thrombin regulates

process outgrowth from neurons and astrocytes and that this effect is mediated via a thrombin

receptor which is activated by proteolytic cleavage (56). PN-1 can block or reverse this cellular

effect of thrombin by inhibiting its proteolytic activity and preventing receptor activation (14, 57).

Here, we explored the ability of rPN-1 to modulate neuronal differentiation and the effects of

thrombin-PN1 interaction on neuronal morphology. To this aim, we chose the mouse NB2A

neuroblastoma cell line, which has been extensively used as a model system for studying the effects

of PN-1 and thrombin on neurite (44). NB2A cells show an undifferentiated phenotype when grown

in presence of serum. However, accumulating evidences have demonstrated that mouse NB2A cells

elaborate axonal neurites in response to various agents such as dibutyryl cyclic AMP, or culture

conditions, e.g., serum deprivation (58). Hence, neurite outgrowth and cell rounding were

monitored as indicators of morphological differentiation following exposure of NB2A cells to rPN-

1. We also investigated whether rPN-1 effects could be reversed by thrombin. In these experiments

we used a wild-type, fully active recombinant thrombin, produced as described earlier (53).

Fig. 7 shows the results of experiments in which cells were exposed to different treatments and

visualized by using immunocitochemistry and photomicrography. When cultured in presence of low

amount of serum (0.8% FBS), NB2A cells showed an undifferentiated phenotype (Fig. 7A); the

addition of 50 nM rPN-1 induced after 3 h neuronal differentiation, with formation of neurite

processes (Fig. 7C). The differentiated phenotype was also observed following concomitant

treatment with 50 nM rPN-1 and 2 nM thrombin for 3 h (Fig. 7E), although the simultaneous

addition of rPN-1 and thrombin to NB2A cells resulted in a decreased neurite length and a lower

number of primary branches (Fig. 7E) when compared with cells treated with rPN-1 alone (Fig.

7C). Withdrawal of serum from culture media for 4 h caused neurite outgrowth (Fig. 7B) and this

effect was reversed by treatment with 2 nM thrombin for 3 h (Fig. 7D); notably, serum-deprived

cells exposed simultaneously to 2 nM thrombin and 50 nM rPN-1 showed a differentiated

phenotype (Fig. 7F).

Taken together, these results demonstrate that recombinant PN-1 has neurite promoting activity

in NB2A cells similar to that of endogenous PN-1 (4), and that likely this effect is mediated by

inhibition of thrombin amidolytic activity caused by irreversible binding PN-1 to thrombin active

site.

2.Protease Nexin-1

58

Figure 7. rPN-1 induces neurite outgrowth in NB2A cells by inhibition of thrombin. Cells were grown subconfluent on cover slips, kept in DMEM containing 10% serum for 24 hr, then differentiated to a neuronal morphology by serum deprivation, or kept undifferentiated in 0.8 % FBS for 4 hr. (A) Cells maintained in DMEM supplemented with 0.8% FCS; (B) cells in DMEM without FCS; (C) cells in DMEM supplemented with 0.8% FCS exposed to 50 nM rPN1 for 3 hr; (D) cells in DMEM without FCS exposed to 2 nM thrombin for 3 hr; (E) cells in DMEM supplemented with 0.8% FCS exposed to 50 nM rPN1 and 2 nM thrombin for 3 hr: (F) cells in DMEM without FCS exposed to 50 nM rPN1 and 2 nM thrombin for 3 hr. After the treatment, the cells were fixed, incubated with rhodamine phalloidin and processed for immunofluorescence microscopy. Inducible expression of PN-1 in HeLa Tet-Off system. Recombinant PN-1 has been also

expressed in the Tet-Off system (59) which allows repression of a single gene by tetracycline or

doxycycline. After transfection of HeLa Tet-off cells with pTRE-Kozak-rbs-PN-1 plasmid, we

analyzed PN-1 expression in the presence and absence of doxycycline at different time points.Fig. 8

2.Protease Nexin-1

59

shows the results of the immunofluorescence analysis of PN-1 expression. After removal of

doxycyline, PN-1 appeared in the cell cytoplasm of transfected HeLa cells (Fig. 8B; 24, 48 and 72

h), and no signal was observed in the presence of the antibiotic (Fig. 8D; 24, 48 and 72 h). At 24 h,

a strong perinuclear fluorescence was observed in the putatively Golgi region, as expected for a

secreted protein. At 48 and 72 h, cell cytoplasm was strongly stained, thus indicating a high level of

protein synthesis. As a control, the same fields were stained with DAPI (Fig. 8C, 24, 48 and 72 h).

Figure 8. Inducible expression of rPN-1 in pTRE-PN-1 transfected HeLa Tet-off cells. Immunofluorescence microscopy in the presence of doxycycline or following doxycycline removal from the culture media, at the indicated time. Nuclei were stained with DAPI. Cells were grown to 80% confluence on glass cover slips, transiently transfected with the pTRE-Kozak-rbs-PN-1 plasmid. At the indicated times after removal of doxycycline, cells were fixed with formaldehyde and processed for immunofluorescence staining, using anti-PN-1 polyclonal serum, followed by fluorescein isothiocyanate antirabbit IgG, and DAPI stained.

2.Protease Nexin-1

60

These results were confirmed by Western blot analysis of secreted PN-1: briefly, after

transfection, the culture media were collected and the secreted proteins concentrated by

ultrafiltration (Amicon, cut-off 30 kDa) and subjected to Western blotting with an anti-PN-1

polyclonal antiserum. To obtain a semiquantitative estimation of rPN-1 secretion following

doxycycline removal, we loaded on SDS-PAGE aliquots of culture media corresponding to 100 µg

of total proteins extracted from the cells. As shown in Fig. 9A, the removal of the antibiotic led to a

high-level secretion of PN-1 at about 43 kDa. From the time course analysis of the secretion yields,

we estimated a maximal secretion at 48 h, which remained stable at 72 h. PN-1 signal appeared as a

doublet, presumably because of some heterogeneity in the glycosylation of the protein (60). No

signal was detected in non-transfected HeLa Tet-Off cells (NT) as well as in transfected cells

cultured in the presence of doxycycline, thus indicating a tight control of the pTRE promoter.

Figure 9. Inducible secretion of PN-1 in HeLa Tet-off cells. (A) Time course of rPN-1 secretion, following removal of doxycycline from the culture media of HeLa Tet-off cells transfected with pTRE-Kozak-PN-1 plasmid. Cells were grown to 80% confluence in 100-mm dishes in the presence of doxycycline (10 ng/ml). Plates were prepared in duplicate and 18 h after transfection; cells were washed with DMEM supplemented with 10% FBS and incubated in the presence or absence of the antibiotic. At the indicated times, culture media were collected and the cells were harvested, resuspended in a lysis buffer and the total protein content measured by Bradford assay. The culture media were concentrated by microfiltration and volumes corresponding to equal amounts of cellular proteins were loaded on 12% SDS-PAGE followed by Western blotting; signal was revealed by enhanced chemioluminescence (ECL). NT, culture media from non transfected HeLa Tet-off cells; the arrow indicates rPN1 (Mr 42 kDa). Numbers on the left represent Mr x 10-

3 of molecular size standards. (B) Bar graphs represent the quantification of PN-1 secretion from Western-blotting analysis. The amount of rPN-1 is expressed in arbitrary units. (+) addition of doxycycline; (-) removal of doxycycline.

A B

2.Protease Nexin-1

61

DISCUSSION

Rat glia-derived PN-1 has an apparent molecular weight of 43 kDa (60), about 2 kDa higher

than that deduced form its cDNA sequence (i.e., 41.7 kDa) (19), assigned to glycosylation reactions.

Scanning of rat PN-1 amino acid sequence for possible N-glycosylation sites identifies Asn364 as a

likely candidate, in the loop region connecting strand 358-364 and strand 369-376 in the B8 sheet.

Notably, natural PN-1 was sensitive to periodic acid staining (60) and treatment of glioma cells with

tunicamycin, a well known inhibitor of glycosylation in eukaryots, resulted in a decrease of about 3

kDa in the apparent molecular weight of PN-1 isolated from these cells (44). Being a glycoprotein,

PN-1 has been expressed using several eukaryotic systems, including yeast (44), chinese hamster

ovary cells (43), and baculovirus (44). In all cases, the recombinant PN-1 proteins were found to be

functionally identical to the natural species (40). Nevertheless, the exceedingly low levels of

expression and the poor yields of purification of rPN-1 from cell cultures were sufficient only for

conducting functional studies whilst impairing structural investigation, for which larger amounts of

protein are required. More recently, during studies aimed at identifying in rat seminal vesicles the

protein responsible of the inhibition of prostasin, an invasion suppressor in prostate cancer cells, rat

PN-1 was also expressed in E. coli (61).

The aim of our study was to develop an expression system suitable for producing large amounts

of pure and biologically active rPN-1 to be used in crystallization studies aimed at elucidating the

threedimesional structure of this physiologically important serpin at the atomic level. Hence, we

explored the possibility of expressing the protein in an inducible manner either in an eukaryotic

expression system, using the HeLa Tet-Off cells, and in a simple prokaryotic system like E. coli.

Notably, in the former case we observed a tight doxycycline-regulated synthesis and secretion of

PN-1, as demonstrated by immunocychemical and protein analysis of HeLa Tet-off cells transfected

with pTRE-Kozak-PN-1 plasmid (Figg. 8 and 9). Notwithstanding, this system still suffered of poor

expression yields. At variance, high-level expression of either His-Tagged fusion protein and

mature form was obtained in E. coli under control of the T7 RNA polymerase promoter. The

purified His-tagged PN-1 was used to obtain an anti-PN-1 antiserum that specifically recognized the

endogenous protein expressed in rat oligodendrocytes, as previously shown (21).

Hence, we decided to express mature rPN-1 on a larger scale in E. coli cell cultures. After

solubilization of inclusion bodies and in vitro refolding, the mature rPN-1 was purified to

homogeneity by heparin-sepharose affinity chromatography, allowing us to recover about 3 mg of

highly homogenous protein per liter of bacterial culture, as demonstrated by the appearance of a

single, immunoreactive band in reducing SDS-electrophoresis (Fig. 1 A, B) and of a single, sharp

2.Protease Nexin-1

62

peak eluting from a C4 RP-HPLC column (Fig. 1C). Notably, the molecular mass of rPN-1 agrees

well with the theoretical value of the mature rat PN-1 (Inset to Fig. 1C). Finally, a pI value of 9.7

was determined for rPN-1 by 2D-gel electrophoresis (Fig. 1D), consistent with the theoretical value

deduced form the amino acid composition of rat PN-1 (i.e., pI 9.72). The protein was stable when

kept at -20° C as eluted from the heparin-sepharose column and no significant loss of activity was

observed even for samples stored for six months.

The amount of purified rPN-1 obtained in this work was at least 10-50 fold higher than that

attainable with eukaryotic systems (44) and this allowed us to carry out for the first time a detailed

conformational characterization of rPN-1 in solution by means of steady state fluorescence

spectroscopy and circular dichroism in the far- and near-UV region (Fig. 3). These spectroscopic

data compared favourably with those obtained with other serpins, such as AT-III (46) and PAI-1

(47), and validated the theoretical model of PN-1 structure reported in Fig. 2. Next, the native-like

structure of rPN-1 and the resulting biochemical properties were probed with respect to the ability

of the recombinant protein to (1) bind heparin, (2) form a SDS-stable complex with thrombin, and

(3) inhibit thrombin amidolytic activity in the absence and presence of heparin.

The data shown in Fig. 4 demonstrate that rPN-1 binds heparin tightly and in a normal manner,

with a 1:1 serpin-cofactor stoichiometric ratio. As already observed with AT-III (32, 46), heparin

binding remarkably enhances the fluorescence intensity of rPN-1, without changing the λmax value.

These data are unprecedented and compatible with a rigidification of the chemical environment of

some Trp-residues (e.g., Trp174) nearby the heparin-binding site (49). However, further studies are

required to clarify whether this conformational change simply reflects a structural adaptation of PN-

1 in response to cofactor binding or, alternatively, if it plays an effective role in promoting PN-1

binding to the target proteases, according to the allosteric activation mechanism previously

highlighted for AT-III function (32).

Although PN-1 exhibits broad protease inhibitory specificity, α-thrombin has been recognized

as its primary physiological target (7, 8). With respect to this, the formation of a SDS-stable

complex with thrombin (see Fig. 5), occurring with a 1:1 stoichiometry, is a clear-cut proof of the

native-like structure of our recombinant PN-1 (9). In fact, upon complex formation both the serpin

and protease undergo massive conformational changes (51) involving the formation of a tetrahedral

covalent intermediate between the Ser195 Oγ of the enzyme and the carbonyl carbon of the amino

acid at P1 position of the scissible P1-P1' bond in the reactive centre loop (RCL) of the inhibitor. In

a second step, the Ser195-acyl-intermediate is formed while the peptide bond between P1 ad P1' has

been broken (6, 54). Once formed, the serpin-protease complex rapidly adopts its lowest energy

conformation through the incorporation of the N-terminal portion of the RCL into -sheet A and

2.Protease Nexin-1

63

translocation of the tethered protease for the top to the bottom of the serpin, with a resulting

distortion of the protease conformation (51). At this stage, deacylation of the acyl-enzyme trapped

intermediate, to form the active protease and the cleaved serpin, is prevented largely by destruction

of the oxyanion hole in the protease (51). For these complex and concerted reactions to occur, it is

necessary that strict stereochemical and dynamical requirements are fulfilled in both serpin and

protease structure. Therefore, the quantitative formation of a stable thrombin-rPN-1 complex (Fig.

5) can be taken as a signature of the native-like conformation of recombinant PN-1 expressed in E.

coli.

Similar conclusions can be drawn from thrombin inhibition experiments yielding a value of

1.1±0.1 x 106 M-1·s-1 for the association rate constant, ka, of thrombin-rPN-1 interaction (Fig. 6A).

Strikingly, this value agrees well with those reported earlier for natural (39), and recombinant PN-1

expressed in different systems (40, 43). In the presence of unfractionated porcine heparin (Fig. 6B,

C), the ka value increases by about three orders of magnitude (k1° = 0.45±0.02 x 109 M-1·cm-1) at

the optimal heparin concentration ((H)opt ~ 60 nM). These values are fully consistent with those

reported for natural PN-1 (39) and suggest that the rate of association of thrombin and PN-1 in the

presence of heparin is essentially at the diffusion-controlled limit.

Finally, the cellular effects of rPN-1 were investigated by measuring the ability of the

recombinant serpin to promote neurite outgrowth in neuroblastoma NB2A cells. The results shown

in Fig. 7 provide evidence that recombinant PN-1 has neurite promoting activity in NB2A cells

similar to that of natural PN-1 (4), and that this effect is mediated by inhibition of thrombin

amidolytic activity, in agreement with previous studies showing that PN-1 can block or reverse the

cellular effects of thrombin by inhibiting its proteolytic activity, thus preventing proteolytic

activation of thrombin receptor(s) on neuronal cells (14, 57).

Altogether, our results demonstrate that E. coli is a suitable expression system for obtaining

milligram quantities of pure and fully active recombinnat PN-1 to be used in future structural and

functional studies.

REFERENCES

1. Eaton DL, Baker JB (1983) Phorbol ester and mitogens stimulate human fibroblast secretions of

plasmin-activatable plasminogen activator and protease nexin, and antiactivator/antiplasmin. J.

Cell Biol. 97:323–328.

2.Protease Nexin-1

64

2. Guttridge DC, Lau A, Tran L, Cunningham DD (1997) Thrombin causes a marked delay in

skeletal myogenesis that correlates with the delayed expression of myogenin and

p21CIP1/WAF1. J. Biol. Chem. 272:24117–24120.

3. Bouton MC, Richard B, Rossignol P, Philippe M, Guillin MC, Michel JB, Jandrot-Perrus M

(2003) The serpin protease-nexin 1 is present in rat aortic smooth muscle cells and is

upregulated in L-NAME hypertensive rats. Arterioscler. Thromb. Vasc. Biol. 23:142–147.

4. Hoffmann MC, Nitsch C, Scotti AL, Reinhard E, Monard D (1992) The prolonged presence of

glia-derived nexin, an endogenous protease inhibitor, in the hippocampus after ischemia-

induced delayed neuronal death. Neuroscience. 49:397–408.

5. Reinhard E, Suidan HS, Pavlik A, Monard D (1994) Glia-derived nexin/protease nexin-1 is

expressed by a subset of neurons in the rat brain. J. Neurosci. Res. 37:256–270.

6. Silverman GA, Bird PI, Carrell RW, Church FC, Coughlin PB, Gettins PG, Irving JA, Lomas

DA, Luke CJ, Moyer RW, Pemberton PA, Remold-O'Donnell E, Salvesen GS, Travis J,

Whisstock JC (2001) The serpins are an expanding superfamily of structurally similar but

functionally diverse proteins. Evolution, mechanism of inhibition, novel functions, and a revised

nomenclature. J. Biol. Chem. 276:33293–33296.

7. Baker JB, Low DA, Simmer RL, Cunningham DD (1980) Protease-nexin: a cellular component

that links thrombin and plasminogen activator and mediates their binding to cells.Cell 21:37–45.

8. Crisp RJ, Knauer DJ, Knauer MF (2000) Roles of the heparin and low density lipid receptor-

related protein-binding sites of protease nexin 1 (PN1) in urokinase-PN1 complex catabolism.

The PN1 heparin-binding site mediates complex retention and degradation but not cell surface

binding or internalization. J. Biol. Chem. 275:19628–19637.

9. Le Bonniec BF, Guinto ER, Stone SR (1995) Identification of thrombin residues that modulate

its interactions with antithrombin III and alpha 1-antitrypsin. Biochemistry 34:12241–12248.

10. Lawrence DA, Olson ST, Muhammad S, Day DE, Kvassman JO, Ginsburg D, Shore JD (2000)

Partitioning of serpin–proteinase reactions between stable inhibition and substrate cleavage is

regulated by the rate of serpin reactive center loop insertion into beta-sheet A. J. Biol. Chem.

275:5839–5844.

11. Mansuy IM, van der Putten H, Schmid P, Meins M, Botteri FM, Monard D (1993) Variable and

multiple expression of Protease Nexin-1 during mouse organogenesis and nervous system

development. Development 119:1119–1134.

12. Giau R, Carrette J, Bockaert J, Homburger V (2005) Constitutive secretion of protease nexin-1

by glial cells and its regulation by G-protein-coupled receptors. J. Neurosci. 25:8995–9004.

2.Protease Nexin-1

65

13. Beilin O, Karussis DM, Korczyn AD, Gurwitz D, Aronovich R, Hantai D, Grigoriadis N,

Mizrachi-Kol R, Chapman J Increased thrombin inhibition in experimental autoimmune

encephalomyelitis. J. Neurosci. Res. 79:351–359.

14. Cavanaugh KP, Gurwitz D, Cunningham DD (1990) Bradshaw RA Reciprocal modulation of

astrocyte stellation by thrombin and protease nexin-1. J. Neurochem. 54:1735–1743.

15. Nitsch C, Scotti AL, Monard D, Heim C, Sontag KH (1993) The glia-derived protease nexin 1

persists for over 1 year in rat brain areas selectively lesioned by transient global ischaemia. Eur.

J. Neurosci. 5:292–297.

16. Cunningham DD, Pulliam L, Vaughan PJ (1993) Protease nexin-1 and thrombin: injuryrelated

processes in the brain. Thromb. Haemost. 70:168–171.

17. Onuma Y, Asashima M, Whitman M (2006) A Serpin family gene, protease nexin-1 has an

activity distinct from protease inhibition in early Xenopus embryos. Mech. Dev. 123:463–471.

18. Grimaldi M, Arcone R, Ciliberto G, Schettini G (1995) Synergistic stimulation of interleukin 6

release and gene expression by phorbol esters and interleukin 1 beta in rat cortical astrocytes:

role of protein kinase C activation and blockade. J. Neurochem. 64:1945–1953.

19. Sommer J, Gloor SM, Rovelli GF, Hofsteenge J, Nick H, Meier R, Monard D (1987) cDNA

sequence coding for a rat glia-derived nexin and its homology to members of the serpin

superfamily. Biochemistry 26:6407–6410.

20. Sanger F, Nicklen S, Coulson AR (1977) DNA sequencing with chain-terminating inhibitors.

Proc. Natl. Acad. Sci. U. S. A. 74:5463–5467.

21. Blasi F, Ciarrocchi A, Luddi A, Strazza M, Riccio M, Santi S, Arcone R, Pietropaolo C,

D'Angelo R, Costantino-Ceccarini E, Melli M (2002) Stage-specific gene expression in early

differentiating oligodendrocytes. Glia 39:114–123.

22. Studier FW, Rosenberg AH, Dunn JJ, Dubendorff JW (1990) Use of T7 RNA polymerase to

direct expression of cloned genes. Methods Enzymol. 185:60–89.

23. Kozak M (1987) At least six nucleotides preceding the AUG initiator codon enhance translation

in mammalian cells. J. Mol. Biol. 196:947–950.

24. Arcone R, Pucci P, Zappacosta F, Fontane V, Malori A, Marino G, Ciliberto G (1991) Single-

step purification and structural characterization of human interleukin-6 produced in Escherichia

coli from a T7 RNA polymerase expression vector. Eur. J. Biochem. 198:541–547.

25. Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities

of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248–254.

26. Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of

bacteriophage T4. Nature 227:680–685.

2.Protease Nexin-1

66

27. Jungblut PR, Seifert R (1990) Analysis by high-resolution two-dimensional electrophoresis of

differentiation-dependent alterations in cytosolic protein pattern of HL-60 leukemic cells. J.

Biochem. Biophys. Methods 21:47–58.

28. Heukeshoven J, Dernick R (1985) Simplified method for silver staining of proteins in

polyacrylamide gels and the mechanism of silver staining. Electrophoresis 6:103–112.

29. Gill SC, von Hippel PH (1989) Calculation of protein extinction coefficients from amino acid

sequence data. Anal. Biochem. 182:319–326.

30. Fenton JW 2nd, Fasco M, Stackrow AB (1977) Human thrombins: production, evaluation, and

properties of α-thrombin. J. Biol. Chem. 252:3587–3598.

31. Dang QD, Di Cera E (1994) A simple activity assay for thrombin and hirudin. J. Protein. Chem.

13:367–373.

32. Olson ST, Björk I, Shore JD (1993) Kinetic characterization of heparin-catalyzed and

uncatalyzed inhibition of blood coagulation proteinases by antithrombin. Methods Enzymol.

222:525–559.

33. Stout TJ, Graham H, Buckley DI, Matthews DJ (2000) Structures of active and latent PAI-1: a

possible stabilizing role for chloride ions. Biochemistry 39:8460–8469.

34. Kopp J, Schwede T. The SWISS-MODEL repository of annotated threedimensional protein

structure homology models. Nucleic Acids Res. 32:D230–234.

35. Gerstein M (1992) A resolution-sensitive procedure for comparing protein surface and its

application to the comparison of antigen-combining sites. Acta Cryst. A48:271–276.

36. Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTALW: improving the sensitivity of

progressive multiple sequence alignment through sequence weighting, positionspecific gap

penalties and weight matrix choice. Nucleic Acids Res. 22:4673–4680.

37. Lottenberg R, Jackson CM (1983) Solution composition dependent variation in extinction

coefficients for p-nitroaniline. Biochim. Biophys. Acta 742:558–564.

38. Stone SR, Nick H, Hofsteenge J, Monard D (1987) Glial-derived neurite-promoting factor is a

slow-binding inhibitor of trypsin, thrombin, and urokinase. Arch. Biochem. Biophys. 252:237–

244.

39. Wallace A, Rovelli G, Hofsteenge J, Stone SR (1989) Effect of heparin on the gliaderived-

nexin–thrombin interaction. Biochem. J. 257:191–196.

40. Rovelli G, Stone SR, Guidolin A, Sommer J, Monard D (1992) Characterization of the heparin-

binding site of glia-derived nexin/protease nexin-1. Biochemistry 31:3542–3549.

41. Wells CM, Di Cera E (1992) Thrombin is a Na(+)-activated enzyme. Biochemistry 31:11721–

11730.

2.Protease Nexin-1

67

42. Tarnowski BI, Spinale FG, Nicholson JH (1991) DAPI as a useful stain for nuclear quantitation.

Biotech. Histochem. 66:297–302.

43. Evans DL, McGrogan M, Scott RW, Carrell RW (1991) Protease specificity and heparin

binding and activation of recombinant protease nexin I. J. Biol. Chem. 266:22307–22312.

44. Stone SR, Brown-Luedi ML, Rovelli G, Guidolin A, McGlynn E, Monard D (1994)

Localization of the heparin-binding site of glia-derived nexin/protease nexin-1 by site-directed

mutagenesis. Biochemistry 33:7731–7735.

45. Brahms S, Brahms J (1980) Determination of protein secondary structure in solution by vacuum

ultraviolet circular dichroism. J. Mol. Biol. 138:149–178.

46. Bruch M, Weiss V, Engel J (1988) Plasma serine proteinase inhibitors (serpins) exhibit major

conformational changes and a large increase in conformational stability upon cleavage at their

reactive sites. J. Biol. Chem. 263:16626–16630.

47. Dwivedi AM, Woodeshick RW, Walton HL, Reilly TM (1991) A spectroscopic study of the

conformations of active and latent forms of recombinant plasminogen activator inhibitor-1.

Biochem. Biophys. Res. Commun. 175:437–443.

48. Strickland EH (1974) Aromatic contributions to circular dichroism spectra of proteins. CRC

Crit. Rev. Biochem. 2:113–175.

49. Lakowicz JR (1986) Fluorescence studies of structural fluctuations in macromolecules as

observed by fluorescence spectroscopy in the time, lifetime, and frequency domains. Methods

Enzymol. 131:518–567.

50. Zhou A, Carrell RW, Huntington JA (2001) The serpin inhibitory mechanism is critically

dependent on the length of the reactive center loop. J. Biol. Chem. 276:27541–27547.

51. Huntington JA, Read RJ, Carrell RW (2000) Structure of a serpin–protease complex shows

inhibition by deformation. Nature 407:923–926.

52. Stavridi ES, O'Malley K, Lukacs CM, Moore WT, Lambris JD, Christianson DW, Rubin H,

Cooperman BS (1996) Structural change in alpha-chymotrypsin induced by complexation with

alpha 1-antichymotrypsin as seen by enhanced sensitivity to proteolysis. Biochemistry

35:10608–10615.

53. Arcone R, Pagliuca MG, Chinali A, Grimaldi M, Schettini G, Gast A, Pietropaolo C (1999)

Thrombin mutants with altered enzymatic activity have an impaired mitogenic effect on mouse

fibroblasts and are inefficient modulators of stellation of rat cortical astrocytes. Biochim.

Biophys. Acta 1451:173–186.

2.Protease Nexin-1

68

54. Lawrence DA, Ginsburg D, Day DE, Berkenpas MB, Verhamme IM, Kvassman JO, Shore JD

(1995) Serpin–protease complexes are trapped as stable acylenzyme intermediates. J. Biol.

Chem. 270:25309–25312.

55. Olson ST (1988) Transient kinetics of heparin-catalyzed protease inactivation by antithrombin

III. Linkage of protease-inhibitor–heparin interactions in the reaction with thrombin. J. Biol.

Chem. 263:1698–1708.

56. Turgeon VL, Houenou LJ (1997) The role of thrombin-like (serine) proteases in the

development, plasticity and pathology of the nervous system. Brain Res. Brain Res. Rev. 25:85–

95.

57. Cunningham DD (1992) Regulation of neuronal cells and astrocytes by protease nexin-1 and

thrombin, Ann. N.Y. Acad. Sci. 674:228–236.

58. Klann E, Shelton KR (1990) A lead-associated nuclear protein which increases in maturing

brain and in differentiating neuroblastoma 2A cells exposed to cyclic AMP-elevating agents.

Brain Res. Dev. Brain Res. 57:71–75.

59. Gossen M, Bujard H (1992) Tight control of gene expression in mammalian cells by

tetracycline-responsive promoters. Proc.Natl.Acad. Sci.U. S. A. 89:5547–5551.

60. Guenther J, Nick H, Monard D (1985) A glia-derived neurite-promoting factor with protease

inhibitory activity. EMBO J. 4:1963–1966.

61. Chen LM, Zhang X, Chai KX (2004) Regulation of prostasin expression and function in the

prostate. Prostate 59:1–12.

3. PYTHON SEBAE SERUM

3.Python sebae serum

71

CHAPTER 3 A novel protein from the serum of Python sebae, structurally homologous to γ-phospholipase A2 inhibitor, displays antitumor activity

Sandra Donninia, Simona Franceseb, Gloriano Monetic, Guido Mastrobuonic, Fabio Masetd, Roberta Frassond, Michele Palmieria, Mario Pazzaglie, Enrico Garacif, Vincenzo De Filippisd and Marina Zichea aDepartment of Molecular Biology, Pharmacology Unit, University of Siena, Siena; bBiomedical Research Centre, Sheffield Hallam University, Sheffield, UK; cDepartment of Clinical and Preclinical Pharmacology, University of Florence, Florence; dDepartment of Pharmaceutical Sciences, University of Padua, Padua; eDepartment of Clinical Physiopathology, University of Florence, Florence; fDepartment of Experimental Medicine and Biochemical Sciences, University of Rome “TorVergata”, Rome.

INTRODUCTION

Snake venoms are complex mixtures of pharmacologically active proteins, including proteases

and secretory phospholipases (PLA2) (1). The latter enzyme family can be distinguished in two 12

groups (I-XII), according to their chain length and intramolecular disulfide-bond topology (2,3).

PLA2s catalyse the hydrolysis of the acyl-ester bond at the sn-2 position of 1,2-diacyl-sn-

phosphoglycerides, leading to the formation lysophospholipids and free fatty acids which are the

precursors of proinflammatory eicosanoids (i.e, prostaglandins, thromboxanes, leukotrienes, and

lipoxins) (4). Thus, PLA2s play a key role in the inflammatory pathway and has been implicated in

numerous severe inflammatory diseases (5). Besides their pro-inflammatory function, PLA2s also

exhibit important physiological functions such as lipid digestion, cell migration and proliferation,

and antibacterial defence (5,6). Although the snake venom PLA2 enzymes differ in the type of

pharmacological effects they induce, toxicity is not directly related to PLA2 enzymatic activity,

thus suggesting the presence of a “toxic” site in the PLA2 structure distinct from the “catalytic” site

(7).

To protect themselves from leakage of their own venom PLA2s into the circulatory system,

venomous snakes contain in their blood PLA2s inhibitors (PLIs), classified into three groups

according to their structure (i.e., PLIα, PLIβ and PLIγ) (5,8). PLIs are acidic oligomeric

glycoproteins with N-linked oligosacchardide chains, composed of three to six identical or different


72

non-covalently linked subunits having a molecular weight of 20-30 kDa (9,10). PLIα is a trimer

composed of identical 20-kDa subunits, having a C-type lectin domain, and specifically inhibits

group-II acidic PLA2s. PLIβ selectively inhibits group-II basic PLS2s and has nine tandem leucine-

rich repeats in its sequence. Contrary to PLIα and PLIβ, PLIγ is a rather nonspecific inhibitor and

its primary structure is characterized by two tandem patterns of Cys-residues, as are found in

urokinase-type plasminogen activator receptor and in Ly6-related proteins (8,11,12). All three types

of PLIs have been found both in the sera of venomous snakes like Viperidae, whereas only PLIγ has

been identified in the sera of venomous Elapidae snakes (10). However, PLIs have also been

identified in the sera of nonvenomous snakes like Elaphe quadrivirgata (i.e., PLIα and PLIβ, PLIγ)

(13-15) and Python reticulatus (16). Thus, the presence of PLIs in nonvenomous snakes suggest that

their physiological role might not be restricted to the protection against the venom PLA2s and that

PLIs may axhibit other different, still unknown, functions.

In the present study, we explored the possibility of identifying in the serum of the nonvenomous

African Rock Python (Python sebae) PLIs possessing novel biological activities not directly related

to the inhibition of PLA2s. With this aim, samples of Python sebae serum, hereafter denoted as

PSS, were tested on several assays for cell toxicity and viability and on in vivo tumor growth. As a

control, similar experiments were carried out with serum samples from Python regius species,

denoted as PRS. Here we report data showing that PSS from P. sebae, species specifically,

produces cytotoxic and antiproliferative effects on tumor cells growing in vitro and reduces tumor

growth in nude mice transplanted with A431 cell line without side effects. To possibly identify the

chemical component(s) responsible for the cytotoxic and antitumor activity, PSS was fractionated

by ion-exchange chromatography (IEC), yielding an active protein fraction (i.e., F5). Notably, F5

reduced tumor cell viability by inducing apoptosis, with a mechanism involving activation of

caspase-3. Further analysis by size-exclusion (SEC) and reversed-phase (RP-HPLC)

chromatography, indicated F5 mainly exists in solution as a tetramer of about 90 kDa, formed by

two dimers, each composed of two non-covalently linked subunits (i.e., P1 and P2) of about 23

kDa. Of interest, isolated P1 and P2 lack cytotoxic and antitumor effects. Enzymatic peptide mass

fingerprint analysis and N-terminal Edman sequencing allowed us to establish that P1 and P2 differ

only by a few amino acids and that both proteins share high sequence homology with type-γ PLI

from P. reticulatus, PIP (16). Despite high sequence similarity, PIP is a homo-hexamer (16),

whereas F5 is a hetero-tetramer. More importantly, PIP is able to effectively inhibit the hydrolytic

activity of PLAs2, whereas the anti-PLA2 function of F5 is negligible. In conclusion, here we have

identified and thoroughly characterized a novel PLI-γ molecule possessing remarkable cytotoxic

and antitumor effects, that can be exploited for potential pharmacological applications.


73


Python serum sampling. Serum from the African Rock Python specie (P. sebae, species sebae

sebae) was sampled from two females aged approximately three years. As a control, serum was also

sampled from two females of the P. regius specie, aged approximately three and four years. All

donor snakes were born in captivity. In animal starved for 24 h and sedated by gentle handling for 2

h, blood samples (average 10 ml) were withdrawn from the cardiac cavity by a 20 cc syringe with a

21G needle. After sedimentation and centrifugation at 3000 rpm at room temperature for 10 min,

the clear supernatant was removed and stored in aliquots at –20°C and used as such in the

experiments after proper dilution in culture media or PBS. In the experiments the serum was diluted

in culture medium deprived of FCS, and titration was performed by protein content quantified by

the Bradford assay. The P. sebae serum is identified all through the manuscript by the abbreviation

PSS, while P. regius serum is identified as PRS. Details about the complete procedure of sample

preparation are reported in the patent n. PCT/EP01/14727.

Functional assays

Cells and culture conditions. Human squamous epithelial carcinoma (A431), human glioblastoma

(U373 MG), human breast cancer (MCF-7) cell lines and human fibroblast (HF) cell line were from

the American Type Culture Collection (ATCC). The A431 and MCF-7 cell line were cultured in

DMEM with 4500 mg/L of glucose and 10% fetal calf serum (FCS). HF cell line was cultured in

DMEM with 1000 mg/L of glucose, 1% non essential aminoacids, 10% sodium pyruvate and 10%

FCS. U373 MG cell line was cultured in MEM with 1% non essential aminoacids, 10% sodium

pyruvate and 10% FCS. Mouse lung carcinoma cell line (LLC) was kindly provided by Dr. F. Pica

at the University of Tor Vergata, Rome, and were cultured in RPMI 1640 with 10% FCS. Capillary

venular endothelial cells (CVEC) were obtained and cultured as described (17).

Cell morphology. Cells (3 x103/well) were plated in 96 microtiter plate in the presence of 10%

FCS, let to adhere for 5 h, and then treated with different concentrations of PSS in 10% FCS for 1h.

Cells (3 x103/well) were plated in 96 microtiter plate in the presence of 10% FCS, let to adhere for 5

h, and then treated with different concentrations of PSS in 10% FCS for 1h. After supernatant

removal, cells were fixed with methanol at 4 °C for 18 h and stained with Diff Qick. Cell

morphology was evaluated by optical microscope Nikon Eclipse T400 at 200X magnification and

the images were recorded by Nikon CCD camera. Cell survival was reported as total cell number

counted/well.


74

Apoptosis and necrosis analysis. Bivariant flow cytometry was performed on adherent cells grown

in the presence or absence of PSS for 4 h in media containing 1% FCS. After incubation, cells were

washed in cold PBS and resuspended in 100 µl of binding buffer (HEPES containing 2.5 M CaCl2).

Fluorescein-labeled Annexin V and propidium iodide (PI) were added to the cell suspension.

Annexin V, a member of the calcium and phospholipids-binding proteins, binds strongly and

specifically to phosphatidylserine which is a marker of cell apoptosis, while PI binds cell DNA,

highlighting necrotic cells. Cells were then analysed by flow cytometry (Becton Dickinson, USA)

and the results are expressed as % of positive cells/total cells.

In vivo tumor growth. Animal studies were carried out according to the European Economic

Community for animal care and welfare (EEC Law No. 86/609), using female 5 week-old Balb/c

nude mice (Harlan Nossan) housed in a barrier care facility and caged in groups of six. Mice were

implanted subcutaneously with A431 (10x106) cells resuspended in 100 µl of sterile PBS. After 4

days, when the diameter of tumor mass ranged between 0.5 and 1 mm3, mice were treated every

other day with 100 µl of a solution containing 100 µg/ml PSS, injected subcutaneously in proximity

of the tumor mass for 15 days. Twelve animals were enrolled in this study, six control mice treated

with PBS and six with PSS. Mice were assigned randomly to either group. Tumor dimensions were

measured every two days with a calliper. Tumor volumes were calculated by taking width x length

x thickness. The survival time of mice was also recorded.

Fractionation of PSS

Ethanol fractionation. Ethanol fractionation has been performed using Ethanol at the final

concentration of 50%. For details see supplemental data.

Salt fractionation. Salt fractionation of 10 µg/ml PSS was performed using NaCl concentrations

between 0.2, 0.4, 0.6, 0.8 and 1 M. For details see supplemental data.

Anion exchange chromatography. PSS (100 µl) diluted 1:5 in 20 mM Tris-HCl, pH 7.5, 0.15 M

NaCl, was fractionated by ion-exchange chromatography on a FPLC system, by using a 6 x 60 mm

MonoQ (Amersham-Pharmacia Biotech) anion exchange column. The serum diluted in Tris-HCl

buffer was centrifuged for 2 min at 13000 rpm, filtered at 0.45 µm on a miniclarifying filter

(Millipore), and then applied to the column. The column was equilibrated in 20 mM Tris-HCl pH

7.5, containing 0.15 M NaCl at a flow rate of 0.5 ml/min, and eluted with a linear gradient of NaCl

from 0.15 to 1.0 M. The absorbance of the effluent was monitored at 280 nm. The protein material

eleuted in correspondence of the chromatographic peaks were collected and the protein content was

estimated by the method of Bradford and that of bicinchoninic acid (BCA). Aliquots of these

fractions were tested for cytotoxic and antitumor activities as described above.


75

RP-HPLC. The fractions eluted at 0.4 M NaCl concentration, denoted as F5 and deriving from

several different chromatographic runs, were pooled and lyophilised. An aliquot (about 50 µg) of

these fractions were successively solubilized in 0.5 ml of aqueous trifluoroacetic acid (0.1% TFA)

and purified by RP-HPLC, on a C4 analytical column from Vydac (The Separation Group,

Hesperia, CA) (4.6 x 150 mm, 5 µm particle size). The column was equilibrated with aqueous 0.1%

TFA and eluted with a linear acetonitrile-0.1% TFA gradient from 30 to 60 % in 35 minutes, at a

flow rate of 0.8 ml/min. The absorbance of the effluent was recorded at 226 nm. The presence of

two peaks was observed, eluted at about 19 and 25 minutes, denoted as P1 and P2. These fractions

were lyophilised and subjected to subsequent analyses.

Chemical characterization

Molecular weight determination by analytical size-exclusion chromatography. The apparent

molecular weight of the proteins in the active fraction F5 eluting at 0.4 M NaCl from the ion

exchange column, was determined by analytical gel filtration chromatography. Samples were

loaded onto a Superose-12 column (1 x 30 cm, Amersham-Pharmacia Biotech), and eluted with 20

mM Tris/HCl buffer, pH 7.5, at a flow rate of 0.3 ml/min. The column was calibrated using a

protein mixture of known molecular weight, in the range of 6.5 – 135 kDa: bovine serum albumin

(67 kDa), ovalbumin (43 kDa), charbonic anydrase (29 kDa), Ribonuclease A (13.7 kDa) and

aprotinin (6.5 kDa). The interstitial volume (Vi) and the void volume (V0) were determined by

loading the Gly-Tyr-Gly tripeptide and dextran blue (2000 kDa), respectively. The distribution

constant (KD) value was calculated by the equation KD = (Ve – V0)/(V i -Ve), where Ve was the

elution volume of the proteins loaded onto the column.

Determination of the cysteine content. Aliquots (100 µg) of RP-HPLC purified P1 and P2 fractions

were subjected to reduction of disulfide bonds and carboxaamidomethylation reaction of the

cysteine (Cys) residues, eventually present along the amino acid sequence, to yield the

corresponding S-carboxamidomethylated (S-CM) derivatives. The reduction reaction was

conducted at 37°C in 0.1 M Tris/HCl buffer, pH 7.8, 1 mM EDTA and 0.125 M dithiothreitol

(DTT). After 2-h reaction, iodoacetamide was added up to a final concentration of 0.25 M and the

reaction allowed to proceed for 90 minutes at 37°C. The reaction mixture was then fractionated by

RP-HPLC on a C4 analytical column (4.6 x 150 mm, 5 µm particle size), equilibrated with 0.1%

aqueous TFA and eluted with a linear acetonitrile-0.1% TFA gradient from 30 to 60% in 35 min, at

a flow rate of 0.8 ml/min. Alternatively, P1 and P2 (100 µg each), with Cys-residues in the reduced

state, were treated for 90 min at 37°C with 4-vinylpyridine (4VP) (0.25 M) to yield the

corresponding S-pyridylethylated (S-PE) derivatives. The reaction was fractionated by RP-HPLC


76

on a C3 (4.6 x 150 mm, 5 µm particle size) analytical column (Agilent Technologies), eluted with a

linear acetonitrile-0.1% TFA gradient from 10 to 30% in 5 minutes and from 30 to 60 % in 40

minutes, at a flow rate of 0.8 ml/min. The absorbance of the effluent was recorded at 226 nm.

Molecular weight determination of P1 and P2 by SDS-PAGE and mass spectrometry. Protein

fractions were analysed by gradient gel electrophoresis (SDS-PAGE) using 4-12% polyacrylamide

precast gels. Lyophilized fractions from anion exchange chromatography were diluted in 20 µl of

sample buffer. Samples were then loaded onto the gel, which was run in Tris glycine buffer, pH 8.3,

with 0.1% SDS. After SDS-PAGE, protein bands were stained by immersing the gel in 0.1%

Coomassie blue solution in water/methanol/acetic acid (5:4:1) for 30 min and destaining by several

changes in 40% methanol, 10% acetic acid until clear background was obtained. To possibly

identify carbohydrate chains, protein bands were stained with the GelCode Glycoprotein Staining

Kit (cat. N. 24562, Pierce chem Co.), according to the manufacture’s procedures.

Unprocessed fractions and their alkylated forms were lyophilized, solubilized in 20 µl of a H2O-

acetonitrile solution (1:1 v/v), containing 1 % formic acid, and analysed by mass spectrometry on a

Mariner Electro Spray Ionization-Time of Flight (ESI-TOF) instrument (Perseptive Biosystems,

Stafford, TX). Spray Tip Potential was set at 3.0 kV, the nozzle potential and temperature at 200

Volts and temperature were set at 140°C, respectively.

Deglycosylation reaction. 20 µg of purified P1 or P2 were dissolved in 200 µl of 20 mM sodium

phosphate buffer, pH 7.2, containing 100 mM EDTA, 1% (by vol.) β-mercaptothanol. This solution

was treated for 24 hours at 37°C with N-glycosidase F (Roche) (0.2 U of enzyme per µg of protein).

The reaction mixture was fractionated by RP-HPLC and analysed by SDS-PAGE. The

polyacrylamide gel was stained either with Coomassie or with the GelCode Glycoprotein Staining

Kit.

N-Terminal Sequence analysis. The first twelve amino acids of purified fractions (P1 and P2) were

determined by Edman degradation, carried out by PRIMM (Biotech Products and Services, Milan,

Italy) using an automatic protein sequencer mod. 477-A (Applied Biosystems, Foster City, CA).

Disulfide reduction and cysteine derivatization of P1 and P2 prior to trypsin digestion. Reduction

of protein (10 µg) disulfide bonds was carried out under denaturing conditions in 0.125 M Tris–HCl

buffer, pH 8.3, containing 1mM EDTA and 6 M guanidinium hydrochloride (Gdn-HCl) in the

presence of 50-fold molar excess of DTT over the total cysteine content and incubation was

conducted at 56 °C for 90 min. Iodoacetamide was then added in a 10-fold molar excess over the

total cysteine content, and the mixture was further incubated in the dark for 60 min at 37 °C.

Desalting was carried out on a PD-10 column (GE Healthcare Bio-Sciences, Uppsala, Sweden),

eluted in 40 mM ammonium bicarbonate buffer. Four 0.5-ml fractions were collected, lyophilised in


77

a SpeedVac (Thermo, San Josè, CA) concentrator and finally dissolved in 40 µl of 50 mM

ammonium hydrogen bicarbonate, pH 7.0. Alternatively, P1 and P2 (100 µg each), with Cys-

residues in the reduced state, were treated for 90 min at 37°C with 4-vinylpyridine (4VP) (0.25 M)

to yield the corresponding S-pyridylethylated (S-PE) derivatives. The reaction was stopped by

addition of 1 M citric acid solution down to pH 4, and fractionated by RP-HPLC on a C3 (4.6 x 150

mm, 5 µm particle size) analytical column from Agilent Technologies. The column was eluted with

a linear acetonitrile-0.1% TFA gradient from 10 to 30% in 5 minutes and from 30 to 60 % in 40

minutes, at a flow rate of 0.8 ml/min. The absorbance of the effluent was recorded at 226 nm.

Tryptic digestion of S-carboxamidomethylated P2 (S-CM-P2) and peptide mass fingerprinting.

Sequencing grade trypsin (Promega, Madison, WI) was added to a final enzyme:protein ratio of

1:50 w/w and the mixture was incubated at 37°C overnight. The tryptic digestion was stopped by

adding 5 µl of 5% TFA. One microliter of S-CM-P2 tryptic digest was deposited on an

AnchorChip target plate (Bruker Daltonics, Bremen, Germany) and allowed to dry; 0.35 µl of

matrix (α-cyano-4-hydroxycinnamic acid 5g/l in 50/50 acetonitrile/0.1% TFA) were then added

and, again, allowed to dry. Mass spectrometric (MS) analysis was performed on an Ultraflex

MALDI TOF TOF (Bruker Daltonics) by using Flex Control 2.4 data acquisition software. Mass

spectra were acquired in reflectron mode over the m/z range 800-3500. The instrumental parameters

were chosen by setting the ion source 1 at 25 kV, the reflector at 26.30 kV and the delay time at 20

ns. The instrument was externally calibrated prior to analysis by using the Bruker peptide calibrant

kit (1000-3000 Da) and the sample spectra internally recalibrated with trypsin autolysis signals. For

MS data acquisition, a total of 400 shots were collected. The peptide masses present in each mass

spectrum, through the integrated software Biotools 3.0, are used to search the NCBInr databank.

Subdigestion of protein fraction tryptic digest with leucine aminopeptidase (LAP) -S-CM. P2

tryptic digest was dried in a SpeedVac concentrator and then dissolved in 12 µl of 20 mM Tris-HCl

buffer (pH 7.5) containing 1 mM MgCl2. Five microliters of this solution were added to 1 µl of

LAP (30 units/µl) (Sigma-Aldrich) at room temperature. A time course reaction was carried out and

1-µl aliquots were withdrawn after 1, 2 and 5 min, mixed with 1 µl of α-cyano-4-hydroxycinnamic

acid (5 g/l) in 50/50 acetonitrile-0.1% TFA and deposited on the target plate for MALDI TOF

analysis.

Tryptic digestion of S-pyridylethylated P1 and P2 (S-PE-P1 and S-PE-P2) and peptide mass

fingerprinting. 100 µg of S-PE-P1 or S-PE-P2 were dissolved in 200 µl of 50 mM Tris-HCl, pH

8.0, 1 mM CaCl2. To this solution, sequencing grade trypsin (Promega) was added to a final

enzyme:protein ratio of 1:20 w/w. The reaction mixture was incubated at 37°C overnight, stopped

by adding 50 µl of 20% aqueous formic acid, and fractionated by RP-HPLC on a C18 analytical


78

column (Agilent Technologies) eluted at a flow rate of 0.8 ml/min with a linear acetonitrile-0.1%

TFA gradient from 5 to 50 % in 40 minutes. The absorbance of the effluent was recorded at 226

nm. Peptide material eluted in correspondence of the chromatographic peaks were collected,

lyophilised, and dissolved in 1%-aqueous formic acid for subsequent mass spectrometry analysis.

Cys-containing fragments were identified by spectrophotometric analysis of S-pyridylethylated

cysteine in the near-UV region (350-240 nm), using a model Lambda 2 Perkin-Elmer (Norwalk,

CA) UV-Vis spectrophotometer.

Subdigestion of S-pyridylethylated P1 and P2, after tryptic digestion, with Glu-C endoproteinase

V8. The peptide materials eluted in correspondence of selected peaks, that remained unidentified in

the tryptic RP-HPLC fingerprint analysis, were lyophilised, dissolved in 100 µl of sodium

phosphate buffer, pH 7.8, and treated at 37°C for 2 hours with Glu-C endoproteinase V8 from S.

aureus, using a peptide:protease ratio of 1:50 by weight. The reaction was stopped by acid addition

and analysed by RP-HPLC on a C18 (4.6 x 150 mm) column eluted with a linear acetonitrile-0.1%

TFA gradient from 2 to 60% in 10 minutes.

Digestion of S-pyridylethylated P1 and P2 (S-PE-P1 and S-PE-P2) with Glu-C endoproteinase

V8 and peptide mass fingerprinting. 100 µg of S-PE-P1 or S-PE-P2 were dissolved in 200 µl of

sodium phosphate buffer, pH 7.8. To this solution, Glu-C endoproteinase V8 (Boheringher) was

added to a final enzyme:protein ratio of 1:20 w/w. The reaction mixture was incubated at 37°C

overnight, stopped by adding 50 µl of 20% aqueous formic acid, and fractionated by RP-HPLC. The

chromatographic peaks were lyophilised, dissolved in 1% aqueous formic acid and analysed by

ESI-TOF mass spectrometry. Prior to mass analysis, all chromatographic fractions were scanned by

UV-Vis absorption spectroscopy for identifying S-pyridylethylated Cys-residues.

Data analysis and protein identification. We used Mascot as a search tool (Matrix Science Ltd.,

London, UK) available on line (http://www.matrixscience.com) to perform the peptide mass

fingerprinting data. Mascot probability score is assigned by computing the probability (P) that the

observed match between the experimental data and mass values calculated from a candidate peptide

sequence is a random event. The correct match, which is not a random event, has a very low

probability. Mascot score is calculated as -10Log10(P). By using NCBInr as protein databank to

search against, Mascot significance threshold is 64. Trypsin autolysis mass signals were excluded

from the protein searching. The taxonomy was set on Chordata; S-carbamidomethylation or S-

pyridylethylation was selected as complete modification; two missed cleavages were allowed for

trypsin chosen as cleaving agent. Searches were performed with a tolerance of 40 ppm. MALDI MS

MS spectra of the peptide signals at m/z 1564.76 and 1485.84 were acquired as well by using

LIFT technology (Bruker Daltonics).


79

Spectroscopic measurements. Protein concentration was determined by the method of

bicinconinic acid for both P1 and P2 recording the absorbance of the solution at 562 nm on a double

beam model Lambda-2 spectrophotometer from Perkin-Elmer (Norwalk, CT). Circular dichroism

(CD) spectra were recorded on a Jasco (Tokyo, japan) model J-810 spectropolarimeter equipped

with a thermostated cell-holder connected to a NesLab (Newington, NH) model RTE-111 water-

circulating bath. Far-UV CD spectra were recorded at 20±0.5°C in 10 mM sodium phosphate pH

7.0, using a 1-mm pathlength quartz cell.

Immunochemical assays

Western Blotting. Cells (3 x 105) were seeded in 60 mm diameter dishes. After adherence, cells

were serum starved (0.1% serum, 24 h) and then stimulated with F5 or F1 (5 ug/ml) in 1% FCS.

Cells were then lysed and centrifuged at 10000 g (20 min at 4°C). 30 µg of proteins were mixed

with 4X reducing SDS-PAGE sample buffer and denatured (10 min at 100°C). Anti-cleaved-

caspase-3 antibody was used (1:1000, Santa Cruz). Results were normalized with actin (1:10000).

Immunohistochemical analysis. CVEC, A431 or U373MG cells (2.5x104 cells/well on glass cover-

slips placed into 24 multiwell plates) were incubated in 0.1% FCS with F5 (50 µg/ml), or F1 (50

µg/mL) or Colchicin (25 µM). After 24 h cells were fixed and incubated overnight at 4°C with

monoclonal antibodies against phosphatydilserine (for A431) or α-tubulin (for CVEC, A431 and

U373MG) diluted 1:200 in PBS/0.5% BSA. Samples were then incubated with secondary

antibodies FITC or TRITC conjugated (Sigma, Italy). Markers were assessed by fluorescence

microscope (Eclipse TE300, Nikon) at 40X magnification and images taken by a digital camera.

Phospholipase A2 activity assay. The secretory PLA2 (sPLA2) activity was measured by using the

sPLA2 Assay Kit (Cayman). Briefly, 106 A431 cells/100 mm dishes were plated in 10% FCS for 24

h and then the conditioned medium was collected, concentrated by Amicon Concentrator columns

(cut off 3 kDa) and tested for secretory Phospholipase A2 (PLA2) activity in the presence or

absence of F5 or F1 following manufactures raccomandations.

The inhibitory activity of PSS or its fractions was assayed by the sPLA2 (type V) Inhibitor

Screening Assay kit (Cayman). Briefly, different concentrations of PSS or its cytotoxic fractions

were added to the plate containing the sPLA2 and the inhibitory activity evaluated as absorbance at

405 nm after 1 minute using a plate reader. Data are reported as % inhibition for each sample versus

sPLA2 activity.


80

Statistics. Results are expressed as means ± SD. Statistical analysis were performed using Student’s

t-test. Differences between groups were considered statistically significant at p<0.05.

RESULTS

PSS induces cytotoxicity in tumor and non tumor cell lines

Cytotoxicity was evaluated in cell suspension as well as on adherent cells. Exposure to PSS in

the 25-1000 µg/ml range, induced concentration-dependent cytotoxicity in human tumor cell lines

(A431 and U373MG, ED50 = 62.5 µg/ml; MCF-7, ED50 = 100 µg/ml ), in mouse tumor cell line

(LLC, ED50 = 75 µg/ml) and in the non-tumor cell lines bovine endothelial cells (CVEC, ED50 =

115 µg/ml) and human skin fibroblasts (HF, ED50 = 125 µg/ml). (Table 1A and B, Supplemental

data). To investigate the species specificity of PSS activity we assessed the effect of the serum

obtained from two specimens of P. regius, PRS1 and PRS2, and from another specimen of the

African Rock Python, PSS2. Similar results were obtained with PSS2, while PRS1 and PRS2 failed

to induce cytotoxic effects (Table 2S).

PSS reduces cell viability

Microscope observation of cells stained with Diff Quick after 1 h exposure to PSS (100 µg/ml)

showed that the cytotoxicity was mainly characterized by retraction of cytoplasmic expansion,

leading to round shaped cells and lysis of cell bodies when compared to basal condition (1% FCS

without PSS) (Fig S1). Then, the PSS effects on cell viability were investigated using the MTT

assay. Cells were exposed to PSS for 5 days at concentrations selected as the least toxic, based on

previous results. Exposure to repeated PSS treatment (every 2 days for 5 days) in the low

concentration range (10-50 µg/ml), evidenced that PSS at 10 µg/ml had no relevant effect on cell

viability for any cell line. However, at higher concentrations, PSS significantly and concentration-

dependently reduced cell viability (Fig. 1).


81

Figure 1. Effect of PSS on cell viability. A panel of tumor and non tumor cells was used: coronary venular endothelial cell (CVEC), human fibroblast (HF), human breast carcinoma cells (MCF-7), human glioblastoma cells (U373MG), human squamous cell carcinoma (A431). Cell viability was quantified by MTT assay after 5 days exposure to 10-25-50 µg/ml. Data are expressed as % cell viability and are the mean of three experiments run in triplicate. Basal = 1% FCS. *p<0.05 vs basal, **p<0.01 vs basal, ***p<0.001 vs basal.

PSS induces cell apoptosis and necrosis

As reported in Fig. 1S, microscope observations of cells stained with Diff Quick, following PSS

treatment documented that changes were compatible with the occurrence of cell apoptosis and

necrosis (retraction of cytoplasmic expansion, rounding of cell shape and lysis of cell bodies). To

further investigate the PSS toxic mechanism and to corroborate the hypothesis that cytotoxicity was

mainly associated with cell apoptosis and necrosis, we performed apoptosis/necrosis studies by

bivariant flow cytometry analysis using fluorescein-labelled annexin V and propidium iodide (PI)

stained cells. In Table 1 are reported the flow-cytometry results of the tumor cell lines A431 and

U373MG exposed to different PSS concentrations for 4h. At the maximal PSS concentration tested

(100 µg/ml), about 50 of cells were positive for PI, and 30 % were positive for Annexin V.


82

Table 1: Effect of PSS on cell death

Apoptosis (Annexin V positive cells) and necrosis (Propidium iodide positive cells) were measured by FACS analysis after 4 h exposure of A431 and U373MG cell lines to different concentration of PSS. Data ± SD are expressed as % of positive cells for annexin V or propidium iodide of three experiments run in triplicate. **p<0.01, ***p<0.001 vs basal (0 µg/ml PSS).

PSS reduces in-vivo tumor growth and increases survival

We investigated the effect of PSS treatment on solid tumor growth and overall survival in nude

mice. The short supply of test compound, limited the number of animals entering the study. A431

cell line was selected for in vivo experiments. Once A431 cells had produced palpable tumors (4

days after subcutaneous transplant of cells) with tumor mass ranging between 0.5 and 1 mm3,

treatment with subcutaneous PSS in proximity of the tumor mass was started with a dosing schedule

of 10 µg every other day for 15 days. A total of 7 treatments was perfomed. Six control mice,

bearing similar tumor masses, were treated with sterile PBS only. PSS treatment was well tolerated,

producing neither local nor systemic effect. An additional group of mice which received PSS by

intra-peritoneal injection survived the treatment with no apparent toxic effect (data not shown). PSS

treatment reduced A431 tumor growth, and after one week, the tumor volume compared to control

was reduced by 20% (days 11 and 13 of Figure 2A, P<0.05). PSS treatment also affected survival,

as 4 out of 6 mice survived at day 21 whereas none of the control survived. However, given the

small number of animals, this difference did not reach significance (data not shown).

48 ± 9**24 ± 7**45 ± 5**23 ± 6**100

44 ± 11**7 ± 515 ± 48 ± 350

27 ± 3**7 ± 38 ± 28 ± 125

8 ± 27 ± 210 ± 49 ± 30

U373MG

Necrosis

(% cell PI positive)

U373MG Apoptosis

(% cell AnnexinV positive

A431 Necrosis


A431 Apoptosis

(% cell AnnexinV positive)

PSS (µµµµg/ml)

48 ± 9**24 ± 7**45 ± 5**23 ± 6**100

44 ± 11**7 ± 515 ± 48 ± 350

27 ± 3**7 ± 38 ± 28 ± 125

8 ± 27 ± 210 ± 49 ± 30

U373MG

Necrosis


U373MG Apoptosis


A431 Necrosis


A431 Apoptosis


PSS (µµµµg/ml)

48 ± 9**24 ± 7**45 ± 5**23 ± 6**100

44 ± 11**7 ± 515 ± 48 ± 350

27 ± 3**7 ± 38 ± 28 ± 125

8 ± 27 ± 210 ± 49 ± 30

U373MG

Necrosis


U373MG Apoptosis

(% cell AnnexinV positive

A431 Necrosis


A431 Apoptosis


PSS (µµµµg/ml)

48 ± 9**24 ± 7**45 ± 5**23 ± 6**100

44 ± 11**7 ± 515 ± 48 ± 350

27 ± 3**7 ± 38 ± 28 ± 125

8 ± 27 ± 210 ± 49 ± 30

U373MG

Necrosis


U373MG Apoptosis


A431 Necrosis


A431 Apoptosis


PSS (µµµµg/ml)


83

Figure 2. A.Effect of PSS on A431 tumor growing subcutaneously. A431 cells (106/mouse) were injected subcutaneously (s.c.) in Balb/c nude mice (day 0). PSS (10 µg/100 µl) or vehicle (PBS, 100 µl ), was injected every other day s.c. close to the tumor mass. Tumor volumes are reported as mean ± SEM, n = 6,*P<0.05 vs vehicle. TI = Tumor Injection. B. Control mice at day 11. C. PSS treated mice at day 11.

Fractionation of PSS and functional characterization of the active component F5

To purify and further characterize the active component(s) triggering cell death, PSS was

fractionated by anion-exchange chromatography (Fig. 3A) on a Mono-Q column. Aliquots of the

protein fractions were analysed by reducing SDS-PAGE (see Inset to Fig. 3A). Of note, fraction F2

contains a minor component at 75 kDa, while F3 contains the major component, migrating at 67

kDa, likely corresponding to the snake albumin. Fraction 4 is contaminated by F3 and contains two

additional bands at 48 and 35 kDa. Of note, fraction F5 contains a highly homogeneous component

at 23 kDa. Of interest, fraction F1 does not contain protein materials, as inferred from the

corresponding UV-absorption spectra, lacking the characteristic features of the absorption bands of

aromatic amino acids (not shown). The eluted fractions, were tested at the final protein

concentration of 50 µg/ml on the viability of A431. Of all fractions, only F5, eluting at about 0.4 M

1 cm1 cm

B

1 cm1 cm

B

1 cm1 cm

C

1 cm1 cm

C

AA


84

NaCl, retained the cytotoxic and antiproliferative effect of PSS, as deduced from TB and MTT tests

and (Fig. 3B).

Figure 3. Fractionation of PSS and effect of F1-F6 fractions on A431 cell viability. A, Fractionation of PSS by anion-exchange chromatography. An aliquot (140 µl) of PSS was loaded onto a Mono-Q analytical column, equilibrated in 20 mM Tris-HCl pH 7.5, containing 0.15 M NaCl at a flow rate of 0.5 ml/min, and eluted with a linear gradient of NaCl from 0.15 to 1.0 M. All fractions were subjected to analysis of total protein content and then tested for cytotoxic antitumoral activity. Inset, SDS-PAGE (12% acrylamide) of protein fractions eluted from the ion-exchange column. S*, low molecular weight protein standard mixture; PSS, crude serum (0.5 µl) from P. sebae; fractions F2-F5 (20 µl each) as reported in Figure 3A. B. Eluted fractions (F1 to F6) were tested for their effects on A431. Cell viability was monitored by TB exclusion after 1 h treatment and MTT test after 24-h treatment. MTT data are expressed as % cell viability and are the mean of at least two experiments run in triplicate. TB data are expressed as % dead cells/total cell counted. Control = 10% FCS. *p<0.05 vs control.

To possibly gain some information on the mechanism(s) of the observed antiproliferative effect,

we investigated the ability of F5 to affect caspase-3 activation, phosphatidylserine expression, and

tubulin assembly. As shown in Figg. 4 and 5, the cellular effects of F5 appeared to be linked to

induction of cell apoptosis, as indicated by caspase-3 activation and phosphatidylserine expression

in A431 (Fig. 4A and B). Conversely, no effect was detected on tubulin assembly, as assessed by

immunofluorescence in CVEC, A431 and U373MG cells exposed for 24 h to F5 (50 µg/ml) (Fig.

5).

0 10 20 30 40

0

20

40

60

80

100

Per

Cen

t NaC

l (--

-)

Abs

orba

nce

at 2

80 n

m (

)


F1 F2

F3

F4

F5 F6

0 10 20 30 40

0

20

40

60

80

100

Per

Cen

t NaC

l (--

-)

Abs

orba

nce

at 2

80 n

m (

)


F1 F2

F3

F4

F5 F6

9766

45

30

kDa S* PSS F2 F3 F4 F5

20

9766

45

30


20

0 10 20 30 40

0

20

40

60

80

100

Per

Cen

t NaC

l (--

-)

Abs

orba

nce

at 2

80 n

m (

)


F1 F2

F3

F4

F5 F6

0 10 20 30 40

0

20

40

60

80

100

Per

Cen

t NaC

l (--

-)

Abs

orba

nce

at 2

80 n

m (

)


F1 F2

F3

F4

F5 F6

9766

45

30


20

9766

45

30


20

100 ± 32.3 ± 2.9F6

63 ± 4*14 ± 4*F5

84 ± 6*8 ± 1.7* F4

90 ± 103 ± 2.7F3

97 ± 204 ± 2.7F2

119 ± 92.2 ± 1.9F1

1001.4 ± 1.6Control

Survival(% cell viability)

% dead cells/total cellscounted

A431

100 ± 32.3 ± 2.9F6

63 ± 4*14 ± 4*F5

84 ± 6*8 ± 1.7* F4

90 ± 103 ± 2.7F3

97 ± 204 ± 2.7F2

119 ± 92.2 ± 1.9F1

1001.4 ± 1.6Control

Survival(% cell viability)

% dead cells/total cellscounted

A431


85

Figure 4. F5 induces cell apoptosis. A. Western blot analysis of cleaved caspase-3 in response to F1 and F5 (50 µg/ml, 24 h). Actin was used for normalization. The graph represents the quantification of gels (ADU = arbitrary densitometric analysis). Gels are representative of three experiments with similar results. B. Phosphatydilserine expression on A431 exposed for 24 h to F1 or F5 (50 ug/ml) in 1% FCS (Control). Magnification 40X.

AA

BB


86

Figure 5. Effect of F5 on α-tubulin assembly. α-Tubulin morphology in A431, CVEC, and U373MG exposed for 24 h to F1 or F5 (50 ug/ml) in 1% FCS (Control). Colchicin was used as positive control (25 uM). Magnification 40X.

Notably, F5 exhibited only a modest inhibitory activity on group II sPLA2 from bee venom and on

A431-released sPLA2 (Fig. 6A and B). Conversely, no effect was observed on group V sPLA2

(Fig. 6C).

A431

CVEC

U373MG

A431

CVEC

U373MG


87

Figure 6 . Effect of F5 on PLA2 activity. A. The effect of F5 was evaluated on group II PLA2 enzyme activity derived from bee venom (BV) or, B. from PLA2 enzyme activity released by A431 cultured for 24 h in 10% FCS. The effect of F5 was measured by using the sPLA2 Assay Kit (Cayman), and evaluated as absorbance at 405 nm. C. Effect of F1, F5 or PSS on group V sPLA2 enzyme activity. The inhibitory activity of PSS or its fractions was assayed by the sPLA2 (type V) Inhibitor Screening Assay kit (Cayman) and measured as absorbance at 405 nm after 1 minute using a plate reader. Data are reported as % inhibition for each sample versus sPLA2 activity reported as 100% activity.

Chemical characterization of fraction F5

The oligomerization state of active F5 component

The oligomerization state of P1 and P2 in the active F5 fraction, eluted from the ion-exchange

column (see Fig. 3A), was established under native conditions by analytical size-exclusion

chromatography (SEC) (Fig. 7A) by loading an aliquot (50 µl, 40 µg) of F5 on a Superose 12 (1 x

30 cm) column. The column was calibrated with a mixture of protein standards in the range 6.5 –

135 kDa, and F5 eluted essentially as a single peak between the monomeric and dimeric BSA, with

an apparent molecular weight of 90±10 kDa. A shoulder at about 120 kDa, accounting for about

20% of the major peak, is also observed Apparently, this result is non consistent with the

electrophoretic data shown in the Inset to Fig. 3A, where fraction F5 migrates as a single band at

about 23 kDa. Further fractionation of F5 by RP-HPLC (Fig. 6B), yielded essentially two

chromatographic peaks of comparable height, denoted as P1 and P2 and having molecular mass


88

values of 23135.2±0.9 u.m.a. and 23152.1±0.8 u.m.a., respectively, as obtained by ESI-TOF mass

spectrometry analysis. Notably, when P1 and P2 were separately dissolved in PBS, both proteins

eluted from the SEC column with a retention time compatible with that of a tetramer, as well.

However, when tested for biological activity, either P1 and P2 revealed complete loss of cytotoxic

activity (data not shown).

Altogether, these data suggest that F5 mainly exists in solution as a tetramer formed by two

hetero-dimers each containing P1 and P2 subunits in a stoichiometric 1:1 ratio.

Figure 7. A, Oligomerization state of P1 and P2 in the active protein fraction F5 by size-exclusion chromatography. The oligomerization state of the active protein material eluted from the ion-exchange column (see F5, Fig. 3A) was established under native conditions by analytical size-exclusion chromatography on a Superose 12 (1 x 30 cm) column. Calibration was performed with a mixture of protein standards in the range 10 – 130 kDa. The apparent molecular weight of the active component resulted to be 98 ± 10 kDa, indicating that the protein, presumably exists in the form of a non-covalent tetramer. B, RP-HPLC analysis of F5. Active fraction F5 was analysed by RP-HPLC for subsequent chemical characterization. An aliquot (100 µl) was loaded onto a Vydac (4.6 x 150 mm) C4 column eluted with a linear acetonitrile/0.1%TFA gradient ----) at a flow rate of 0.8 ml/min. Two peaks, named P1 and P2, were eluted at 19 and 25 min, respectively, were then collected and subjected to N-terminal sequence analysis by automated Edman degradation and peptide mass fingerprinting with trypsin or endoproteinase Glu-C from S. aureus. Determination of glycosylation state and cysteine/cystine content in P1 and P2

Purified P1 and P2 were subjected to SDS-PAGE analysis prior and after deglycosylation

reaction with N-glycosidase F, which hydrolyses N-linked glycosylation sites Asn-residues. The gel

was stained either with Coomassie R-250 or with the GelCode Glycoprotein Staining Kit. Both

subunits yield a pink colour when stained with the glycoprotein staining kit, whereas after reaction

with N-glycosidase F, the protein bands are be stained only with Coomassie, thus suggesting at least

one N-glycosytaion site exists in both P1 and P2.

0,2 0,3 0,4 0,5 0,6 0,7 0,8

3,8

4,0

4,2

4,4

4,6

4,8

5,0

5,2

5,4

Log

PM

Kd

F5

0,2 0,3 0,4 0,5 0,6 0,7 0,8

3,8

4,0

4,2

4,4

4,6

4,8

5,0

5,2

5,4

Log

PM

Kd

F5

0 10 20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

4A.

U.:0

.04

A.U

.:0.0

4 P1

P2

0 10 20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

4A.

U.:0

.04

A.U

.:0.0

4A.

U.:0

.04

A.U

.:0.0

4A.

U.:0

.04

A.U

.:0.0

4 P1

P2

A B

0,2 0,3 0,4 0,5 0,6 0,7 0,8

3,8

4,0

4,2

4,4

4,6

4,8

5,0

5,2

5,4

Log

PM

Kd

F5

0,2 0,3 0,4 0,5 0,6 0,7 0,8

3,8

4,0

4,2

4,4

4,6

4,8

5,0

5,2

5,4

Log

PM

Kd

F5

0 10 20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

4A.

U.:0

.04

A.U

.:0.0

4 P1

P2

0 10 20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

4A.

U.:0

.04

A.U

.:0.0

4A.

U.:0

.04

A.U

.:0.0

4A.

U.:0

.04

A.U

.:0.0

4 P1

P2

A B


89

Determination of Cys-content of P2 fraction was carried out by

reduction/carboxamidomethylation reaction with DTT/iodoacetamide. The Cys-reduced and

carboxamidomethylated P2 (S-CM-P2) eluted as a single peak in RP-HPLC on a C4 analytical

column (not shown), with an increase in the mass value of 928.9±0.5 a.m.u. Considering that

carboxamidomethylation of half-cystine adds 58.03 a.m.u, we conclude that P2 contains 16

cysteines (16 x 58.03 a.m.u.. = 928.48 a.m.u.), arranged to form eight disulfides. Of note, neither P1

nor P2 reacts with Ellman’s reagent (dithio-bis(2-nitrobenzoic acid), indicating that no free Cys-

residues is present.

Similar results were obtained either with P1 and P2, after reduction and derivatization with 4-

vinylpyridine, a more selective reagent for Cys-residues (18). In both cases, a mass increment of

1696±0.7 a.m.u. was obtained, thus confirming the presence of 16 cysteines in both proteins (106

a.m.u. x 16 = 1696 a.m.u.).

Sequencing of P1 and P2

Purified P1 and P2, (see Fig. 7B) were subjected to N-terminal sequencing by Edman

degradation on a liquid-phase protein sequencer. In both fractions, the sequence

HKXEIXHGFGDD was obtained in which X denotes the absence in the RP-HPLC chromatogram

of a well-defined phenylthiohydantoine (PTH) derivative, indicating the presence of a disulfide

bridge in the X position. The N-terminal amino acidic sequence (HKCEICHGFGDD), when

compared with a protein databank, using the BLAST software, matched the sequence 1-12,

DKCEICHGFGDD, of the sPLA2 inhibitory protein (PIP) isolated from P. reticulatus (SwissProt

accession number: Q9I8P7) (16), apart from an amino acid replacement in position 1 (His in P.

sebae instead of Asp in P. reticulatus).

Internal sequence information on P1 and P2 were obtained by peptide mass fingerprinting of the

reduced and S-alkylated species with trypsin and Glu-C endoproteinase (Fig. 8A, B and C; Table 2

and 3), spectrophotometric analysis of S-pyridylethylated P1 and P2 proteolytic fragments and by

MS-MS analysis of some selected peptides (Fig. 2, Supplemental data). In the enzymatic peptide

mass fingerprint analysis, Cys-containing peptides were readily identified during RP-HPLC

analysis by the absorbance spectrum of the S-β-(4-pyridylethyl)-moiety, showing a characteristic

shape with a maximum at 254 nm (18). During electrospray ionisation tandem mass spectrometry,

Cys-containing peptides were identified by the appearance of a pyridylethyl-fragment ion of 106

a.m.u. Some anomalous tryptic cleavages were found and confirmed by careful MS-MS sequence

determination. From these data, we could establish a high sequence similarity between P1 and P2,

constituting the active protein serum component (i.e., fraction F5), with that of PLA2 inhibitor from


90

Python reticulatus (PIP) (16). Based on PIP sequence, we could identify 125 and 120 amino acids

in P1 and P2, respectively. Of these, only seven mutations in P1 and four mutations in P2 were

identified with respect to PIP sequence (see Fig. 9).

20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

1

23

1

45

6

7

20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

1A.

U.:0

.01

23

1

45

6

7

4

6

5 7

8

9

10

11

12

20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

1

23

1

45

6

7

20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

1A.

U.:0

.01

23

1

45

6

7

4

20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

1

23

1

45

6

7

20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

1A.

U.:0

.01

23

1

45

6

7

4

6

5 7

8

9

10

11

12

A

20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

1

23

1

45

6

7

20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

1A.

U.:0

.01

23

1

45

6

7

4

6

5 7

8

9

10

11

12

20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

1

23

1

45

6

7

20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

1A.

U.:0

.01

23

1

45

6

7

4

20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

1

23

1

45

6

7

20 30 40

0

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


A.U

.:0.0

1A.

U.:0

.01

23

1

45

6

7

4

6

5 7

8

9

10

11

12

A

20 30 40

0

20

40

60

80P

er C

ent A

ceto

nitr

ile (

---)

Abs

orba

nce

at 2

26 n

m (

)


4

321

7

65

9

8

10 11

1213

14

15

A.U

.:0.0

7A.

U.:0

.07

A.U

.:0.0

7 B

20 30 40

0

20

40

60

80P

er C

ent A

ceto

nitr

ile (

---)

Abs

orba

nce

at 2

26 n

m (

)


4

321

7

65

9

8

10 11

1213

14

15

A.U

.:0.0

7A.

U.:0

.07

A.U

.:0.0

7 B


91

Figure 8. Proteolysis of S-pyridylethylated derivatives of P1 and P2 with trypsin and Glu-C endoproteinase. RP-HPLC analysis of P1 (A) and P2 (B) with trypsin. RP-HPLC analysis of P2 with Glu-C protease (C).

Table 2: Peptide mass fingerprint analysis of P1

Peak Number

MW (a.m.u.) (experimental)

MW (a.m.u.) (theoretical)

∆∆∆∆MW Position Sequence Deduced Mutation

1 382.50 382.48 -0.02 117-118 CR 2 616.34 616.35 -0.01 164-168 KNDLK 3 738.38 738.38 ~0 169-173 KVECR 4 1192.24 1192.32 -0.08 18-29 ECPSPEDKCGK R26K 5 1197.54 1197.53 0.01 47-55 NCFSSSICK 6 2106.51 2106.55 -0.04 3-18 CEICHGFGDDCDGYEE 7 1440.79 1440.75 0.04 30-42 ILIDIALAP*VSFR P38methyPro 8 1673.75 1673.83 -0.08 145-158 GCVSSCPLLTLSER V153L 9 1561.77 1561.73 0.04 119-131 GTETMCLDLVGYR 10 1488.76 1488.72 0.04 59-70 VDIHVWDGVYMR I69M 11 1033.57 1033.62 -0.05 85-94 PLPGLPLSLK Q94K 12 1490.61 1490.69 -0.08 104-116 GIFTEDSTEHEVK

10 20 30 400

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


1514

13

16

17

18 19

10 20 30 400

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


1514

13

16

17

18 19

A.U

.:0.0

1

10 20 30 400

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


1514

13

16

17

18 19

10 20 30 400

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


1514

13

16

17

18 19

A.U

.:0.0

1A.

U.:0

.01

17

16

18 19

21

20

22

C

10 20 30 400

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


1514

13

16

17

18 19

10 20 30 400

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


1514

13

16

17

18 19

A.U

.:0.0

1

10 20 30 400

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


1514

13

16

17

18 19

10 20 30 400

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


1514

13

16

17

18 19

A.U

.:0.0

1A.

U.:0

.01

17

16

18 19

21

20

22

10 20 30 400

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


1514

13

16

17

18 19

10 20 30 400

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


1514

13

16

17

18 19

A.U

.:0.0

1

10 20 30 400

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


1514

13

16

17

18 19

10 20 30 400

20

40

60

80

Per

Cen

t Ace

toni

trile

(--

-)

Abs

orba

nce

at 2

26 n

m (

)


1514

13

16

17

18 19

A.U

.:0.0

1A.

U.:0

.01

17

16

18 19

21

20

22

C


92

Table 3: Peptide mass fingerprint analysis of P2

1 10 20 30 40 50 60

P.reticulatus DKCEICHGFG DDCDGYQEEC PSPEDRCGKI LIDIALAPVS FRATHKNCFS SSICKLGRVD P.sebae-P1 HKCEICHGFG DDCDGYEEEC PSPEDKCGKI LIDIALAPVS FRATHKNCFS SSICKLGRVD P.sebae-P2 HKCEICHGFG DDCDGYQEEC PSPEDKCGKI LIDIALAPVS FRATHKNCFS SSICKLGRVD

70 80 90 100 110 120

P.reticulatus IHVWDGVYIR GRTNCCDNDQ CEDQPLPGLP LSLQNGLYCP GAFGIFTEDS TEHEVKCRGT P.sebae-P1 IHVWDGVYMR GRTNCCDNDQ CEDQPLPGLP LSLKNGLYCP GAFGIFTEDS TEHEVKCRGT P.sebae-P2 IHVWEGVYIR GRTNCCDNDQ CEDQPLPGLP LSLKNGLYCP GAFGIFTEYS TEHEVKCRGT

130 140 150 160 170 180

P.reticulatus ETMCLDLVGY RQESYAGNIT YNIKGCVSSC PLVTLSERGH EGRKNDLKKV ECREALKPAS SD P.sebae-P1 ETMCLDLVGY RQESYAGNIT YNIKGCVSSC PLLTLSERGH EGRKNDLKKV ECREALKPAS SD P.sebae-P2 ETMCLDLVGY RQESYAGNIT YNIKGCVSSC PLVTLSERGH EGRKNDLKKV ECREALKYES SD

Peak Number

MW (a.m.u.) (experimental)

MW (a.m.u.)

(theoretical)

∆∆∆∆MW Position Sequence Deduced Mutation

1 616.35 616.35 ~0 164-168 KNDLK 2 459.27 45927 ~0 174-177 EALK 3 610.29 610.29 ~0 170-173 VECR 4 738.38 738.38 ~0 169-173 KVECR

5 1040.43 1040.47 -0.04 174-182 EALKYESSD P178Y A179E

6 488.23 488.26 -0.03 165-168 NDLK 7 1197.53 1197.53 ~0 47-55 NCFSSSICK

8 3490.98 3490.90 0.08 3-29 CEICHGFGDDCDGY QEECPSPEDKCGK

R26K

9 1659.85 1659.81 0.04 145-158 GCVSSCPLVTLSER 10 1634.81 1634.77 0.04 43-55 ATHKNCFSSSICK 11 1539.62 1539.66 -0.04 104-116 GIFTEYSTEHEVK D109Y 12 1561.74 1561.73 0.01 119-131 GTETMCLDLVGYR

13 2719.15 2719.12 0.03 73-94 TNCCDNDQCED QPLPGLP LSLK Q94K

14 1484.76 1484.78 -0.02 59-70 VDIHVWEGVYIR D65E 15 1426.82 1426.85 -0.03 30-42 ILIDIALAPVSFR 16 511.23 511.22 0.01 172-174 CRE 17 620.28 620.27 0.01 1-4 HKCE D1H 18 896.44 896.45 -0.01 115-121 VKCRGTE 19 1185.69 1185.68 0.01 162-171 GRKNDLKKVE

20 622.32 622.33 -0.01 175-179 ALKYE P178Y A179E

21 1896.71 1896.70 0.01 5-19 ICHGFGDDCDGYQEE 22 1531.76 1531.72 0.04 122-133 TMCLDLVGYRQE


93

Figure 9. Comparison between PIP amino sequence and P1 and 2 from P. sebae. The peptide sequences of P1 and P2 fraction from P. sebae serum are aligned with the sequence of PIP from P. reticulatus. The N-terminal Edman analysis of the first 12 amino acids of P25 form P. sebae matches the PIP sequence from P. reticulatus, aside from substitution in position 1 (H instead of D). The mass spectrometry data allowed to cover about 71% and 89% of the P1 and P2 sequences, respectively. The segment of PIP that could not be identified in P1 and P2 are indicated in gray.

Conformational Characterization

The conformational properties of purified P1 and P2 were investigated by CD in the far-UV

region, a spectroscopic techniques that gives information on the secondary structure content of

proteins (19). The CD spectra of both P1 and P2 (Fig. 10) is typical of a protein containing a

predominant β-sheet secondary structure, with a minimum centered at 212-215 nm (19). In

addition, these spectra indicate that P1 and P2 share common structural properties.

Figure 10. Far-UV circular dichroism (CD) of P1 and P2. CD spectra were of P1 and P2 were recorded in 10 mM sodium phosphate, pH 7.0 at a protein concentration of 0.17 mg/ml for both proteins. Ellipticity data are expressed in millidegrees.

DISCUSSION

A variety of factors displaying cytotoxic activity against murine and human tumor cell lines in

vitro, and antitumor efficacy in vivo have been documented in the venom of snakes and several

200 210 220 230 240 250

-6

-5

-4

-3

-2

-1

0

θ (m

deg)

Wavelength (nm)

P1P2


94

proteins have been isolated and characterized (20-24). Conversely, little information is available on

the identification of antitumor products in the snake serum.

In this work for the first time, we purified from the serum of the non venomous snake P. sebae,

termed PSS, a protein having a selective anti tumor activity. PSS cytotoxicity was evaluated in cell

suspension and in cell adhesion experiments using both tumor and non tumor cell lines. In both

experimental conditions and for all the cell lines, PSS exhibited a remarkable and dose-dependent

cytotoxic effect. Exposure to low concentration of PSS for prolonged time also affected cell

viability as measured by the MTT assay. After exposure to PSS at concentration ranging from 400

to 2000 µg/ml, changes of morphology and cell lysis occurred within minutes. Microscope

observation of cells exposed for several h to PSS, evidenced that a consistent proportion of the cell

monolayer became detached. The time dependence of the changes in cell morphology (1 h), is

evocative of a rapid treatment-induced gradual loss of cell-matrix adhesion, similar to that

documented for other cytotoxic principle obtained from snakes (24-26) Consistent with the

morphological appearance of cytoplasmic retraction, rounding of cell shape and lysis of cell bodies,

PSS effects were mainly associated with cell apoptosis and necrosis. With respect to the decrease in

cell viability and the morphological changes observed following exposure to PSS, our data suggest

a direct cytotoxic/pro-apoptotic activity of the serum in vitro. An interesting finding of this work

was that the cytotoxic effect appears to be more pronounced on human tumor cells than in non

tumor cells. At concentration of 50 µg/ml PSS was devoid of any effect on non-tumor cell lines,

while inducing over 50% inhibition of tumor cell growth. These effects were absent at

concentrations lower than 50 µg/ml.

In vivo experiments demonstrated that the cytotoxic effect produced in vitro on tumor cell lines

translated in tumor growth inhibition in nude mice bearing epithelial tumors (A431). PSS

peritumoral treatment, which was well tolerated, reduced tumor growth (30%). Although biased by

the limited number of mice that could be studied given the limited availability of the source, we

found interesting that the treatment appeared to increase survival.

The active protein component, F5, was purified from the crude PSS by ion-exchange

chromatography and found to exist as a tetramer of about 90 kDa, formed by two dimers, each

composed of two non-covalently linked subunits (i.e., P1 and P2) of about 23 kDa. Of interest,

isolated P1 and P2 lack cytotoxic and antitumor effects. Notably, F5 reduced tumor cell viability

and induced apoptosis by activation of caspase-3. In an attempt to identify the molecular target of

F5, the β-tubulin distribution was measured in cells treated with the active protein, but no effect was

observed on microtubule dynamics. Limitations in the supply of the source of the active protein,

prevented a comprehensive assessment of the putative mechanism for the anti-proliferative activity.


95

The cloning and recombinant expression of the novel protein, may open toward the full

characterization of the mechanisms, and in particular, the molecular target(s) of this cytotoxic

factor.

The results of the detailed chemical characterization, conducted by enzymatic peptide mass

fingerprint analysis and N-terminal Edman sequencing, revealed that P1 and P2 are two isoforms of

the same protein and both share high sequence homology with the γ-type phospholipase inhibitor

(PLI) from P. reticulatus, PIP (16) (see Fig. 9). The blood of venomous as well as non-venomous

snakes contains specialized serum proteins as an endogenous mechanisms of natural resistance to

toxicity induced by toxins such as myotoxic and neurotoxic PLA2s, as well as hemorrhagic

metalloproteinases (27). In particular, the PLA2 enzyme family can be distinguished in 12 groups

(I-XII), according to their chain length and intramolecular disulfide-bond topology (2,3), and it play

a key role in numerous severe inflammatory diseases (5). Specific inhibitors (PLIs) of the toxic

activity of PLA2s, have been characterized, and classified into three groups according to their

structure (i.e., PLIα, PLIβ and PLIγ) (5,8) PLIs are acidic oligomeric glycoproteins with N-linked

oligosaccharide chains, composed of three to six identical or different non-covalently linked

subunits having a molecular weight of 20-30 kDa (10,12) PLIα specifically inhibits group-II acidic

PLA2s, PLIβ selectively inhibits group-II basic PLA2s, while PLIγ is a nonspecific inhibitor and its

primary structure is characterized by two tandem patterns of Cys-residues (8,11,12) The PLIγ from

P. reticulatus, has been reported to appear as a homo-hexamer able to inhibit different types of

PLAs2 (28). On the contrary, the isolated F5 protein exists as a hetero-tetramer, and does not

possess an appreciable inhibitory activity on both type II and V PLA2s tested. Earlier studies have

shown that the segment 85PLPGLPLSLQNGLY98 (29) and 59VDIHVWDGV 67 (30) are responsible

for the PLA2 inhibitory activity of PIP. As shown in Fig. 9, we have indentified in P1 the sequence 59VDIHVWDGV 67 and in P2 the same sequence containing the mutation D65E, 59VDIHVWEGV67.

In P1, we have also identified the sequence 85PLPGLPLSLK94 and in P2 the sequence 95PLPGLPLSLKNGLY98. Rather surprisingly, F5 does not show any appreciable PLA2 inhibition.

This can be due to a different arrangement of the PLA2 recognition sites in F5 and/or to the

mutations D65E and Q94K. Of note, substitutions of charged amino acids are involved in both

cases. Finally, the conformational characterization of P1 and P2 revealed the presence of high

content of β-sheet secondary structure. This finding is consistent with earlier studies showing

significant structural similarity of PLIγ from different species, including PIP, with the structure of

mammalian urokinase-type plasminogen activator receptor (uPAR), which displays a characteristic

β-sheet structure and unique disulfide bond pattern.


96

In conclusion, in this study we have shown that P. sebae serum contains a previously unknown

specie specific and selective cytotoxic protein component endowed with pro-apototic activity which

could be exploited for the development of novel antitumor strategies.

REFERENCES

1. Kim SI, Kim KS, Kim HS, Choi MM, Kim DS, Chung KH, Park YS (2004) Inhibition of

angiogenesis by salmosin expressed in vitro. Oncol. Res. 14:227-233

2. Heinrikson RL, Krueger ET, Keim PS (1977) Amino acid sequence of phospholipase A2-alpha

from the venom of Crotalus adamanteus. A new classification of phospholipases A2 based upon

structural determinants. J. Biol. Chem. 252:4913-4921

3. Dennis EA (1997) The growing phospholipase A2 superfamily of signal transduction enzymes.

Trends Biochem. Sci. 22:1-2

4. Kishino J, Ohara O, Nomura K, Kramer RM, Arita H (1994) Pancreatic-type phospholipase A2

induces group II phospholipase A2 expression and prostaglandin biosynthesis in rat mesangial

cells. J. Biol. Chem. 269:5092-5098

5. Lambeau G, Lazdunski M (1999) Receptors for a growing family of secreted phospholipases

A2. Trends Pharmacol. Sci. 20:162-170

6. Buckland AG, Wilton DC (2000) The antibacterial properties of secreted phospholipases A(2).

Biochim. Biophys. Acta 1488:71-82

7. Kini RM, Evans HJ (1989) A model to explain the pharmacological effects of snake venom

phospholipases A2. Toxicon 27:613-635

8. Ohkura N, Okuhara H, Inoue S, Ikeda K, Hayashi K (1997) Purification and characterization of

three distinct types of phospholipase A2 inhibitors from the blood plasma of the Chinese

mamushi, Agkistrodon blomhoffii siniticus. Biochem. J. 325: 527-531

9. Faure G (2000) Natural inhibitors of toxic phospholipases A(2). Biochimie 82:833-840

10. Dunn RD, Broady KW (2001) Snake inhibitors of phospholipase A(2) enzymes. Biochim.

Biophys. Acta 1533:29-37

11. Ploug M, Ellis V (1994) Structure-function relationships in the receptor for urokinase-type

plasminogen activator. Comparison to other members of the Ly-6 family and snake venom

alpha-neurotoxins. FEBS Lett. 349:163-168


97

12. Faure G, Villela C, Perales J, Bon C (2000) Interaction of the neurotoxic and nontoxic secretory

phospholipases A2 with the crotoxin inhibitor from Crotalus serum. Eur. J. Biochem. 267:4799-

4808

13. Okumura K, Inoue S, Ikeda K, Hayashi K (2003) Identification and characterization of a serum

protein homologous to alpha-type phospholipase A2 inhibitor (PLIalpha) from a nonvenomous

snake, Elaphe quadrivirgata. IUBMB Life. 55:539-545

14. Okumura K, Inoue S, Ohkura N, Ikeda K, Hayashi K. (1999) cDNA cloning of the two subunits

of phospholipase A2 inhibitor PLIgamma from blood plasma of the Chinese mamushi,

Agkistrodon blomhoffii siniticus. IUBMB Life. 48:99-104

15. Okumura K, Masui K, Inoue S, Ikeda K, Hayashi K (1999) Purification, characterization and

cDNA cloning of a phospholipase A2 inhibitor from the serum of the non-venomous snake

Elaphe quadrivirgata. Biochem J. 341:165-171

16. Thwin MM, Gopalakrishnakone P, Kini RM, Armugam A, Jeyaseelan K (2000) Recombinant

antitoxic and antiinflammatory factor from the nonvenomous snake Python reticulatus:

phospholipase A2 inhibition and venom neutralizing potential. Biochemistry 39: 9604-9611

17. Schelling ME, Meiningerm CJ, Hawker JR Jr, Granger HJ (1988) Venular endothelial cells

from bovine heart. Am. J. Physiol. 254:1211-1217

18. Moritz RL, Eddes JS, Reid GE, Simpson RJ (1996). S-pyridylethylation of intact

polyacrylamide gels and in situ digestion of electrophoretically separated proteins: a rapid mass

spectrometric method for identifying cysteine-containing peptides. Electrophoresis 17: 907-917

19. Brahms S, Brahms J (1980) Determination of protein secondary structure in solution by

vacuum ultraviolet circular dichroism. J. Mol. Biol. 138:149-178

20. Schmitmeier S, Markland FS, Chen TC (2000) Anti-invasive effect of contortrostatin, a snake

venom disintegrin, and TNF-alpha on malignant glioma cells. Anticancer Res. 20:4227-4233

21. Oliveira JC, de Oca HM, Duarte MM, Diniz CR., Fortes-Dias CL (2002) Toxicity of South

American snake venoms measured by an in vitro cell culture assay. Toxicon 40:321-325

22. Abu-Sinna G, Esmat AY, Al-Zahaby AA, Soliman NA, Ibrahim TM (2003) Fractionation and

characterization of Cerastes cerastes cerastes snake venom and the antitumor action of its lethal

and non-lethal fractions. Toxicon 42:207-215

23. Kim SI, Kim KS, Kim HS, Choi MM, Kim DS, Chung KH, Park YS (2004) Inhibition of

angiogenesis by salmosin expressed in vitro. Oncol. Res. 14:227-233

24. Baramova EN, Shannon JD, Bjarnason JB, Fox JW (1989) Degradation of extracellular matrix

proteins by hemorrhagic metalloproteinases. Arch. Biochem. Biophys. 275:63-71


98

25. Coelho AL, de Freitas MS, Oliveira-Carvalho AL, Moura-Neto V, Zingali RB, Barja-Fidalgo C

(1999) Effects of jarastatin, a novel snake venom disintegrin, on neutrophil migration and actin

cytoskeleton dynamics. Exp. Cell. Res. 251:379-387

26. Petretski JH, Kanashiro MM, Rodrigues FR, Alves EW, Machado OL, Kipnis TL (2000) Edema

induction by the disintegrin-like/cysteine-rich domains from a Bothrops atrox hemorrhagin.

Biochem. Biophys. Res. Commun. 276:29-34

27. Koh DCI., Armugam A, Jeyaseelan K (2006) Snake venom components and their applications

in biomedicine. Cell. Mol. Life Sci. 63:3030–304

28. Thwin MM, Gopalakrishnakone P (1998) Snake endovenomation and protective natural

endogenous proteins: a mini review of the recent development (1991-1997). Toxicon 36:1471-

1482

29. Thwin MM, Satish RL, Chan ST, Gopalakrishnakone P (2002) Functional site of endogenous

phospholipase A2 inhibitor from python serum. Eur. J. Biochem. 269:719-727

30. Thwin MM, Satyanarayanajois SD, Nagarajarao LM, Sato K, Arjunan P, Ramapatna SL, Kumar

PV, Gopalakrishnakone P (2007) Novel peptide inhibitors of human secretory phospholipase A2

with antiinflammatory activity: solution structure and molecular modeling. J. Med. Chem.

50:5938-5950

SUPPLEMENTAL DATA


Cells and culture conditions

The A431 and MCF-7 cell line were cultured in DMEM with 4500 mg/L of glucose and 10% fetal

calf serum (FCS). HF cell line was cultured in DMEM with 1000 mg/L of glucose, 1% non

essential aminoacids, 10% sodium pyruvate and 10% FCS. U373 MG cell line was cultured in

MEM with 1% non essential aminoacids, 10% sodium pyruvate and 10% FCS. Mouse lung

carcinoma cell line (LLC) was kindly provided by Dr. F. Pica at the University of Tor Vergata,

Rome, and were cultured in RPMI 1640 with 10% FCS. Capillary venular endothelial cells (CVEC)

were obtained and cultured as described.


99

Cell morphology

Cells (3 x103/well) were plated in 96 microtiter plate in the presence of 10% FCS, let to adhere for

5 h, and then treated with different concentrations of PSS in 10% FCS for 1h. After supernatant

removal, cells were fixed with methanol at 4 °C for 18 h and stained with Diff Qick. Cell

morphology was evaluated by optical microscope Nikon Eclipse T400 at 200X magnification and

the images were recorded by Nikon CCD camera. Cell survival was reported as total cell number

counted/well.

Apoptosis and necrosis analysis

Bivariant flow cytometry was performed on adherent cells grown in the presence or absence of PSS

for 4 h in media containing 1% FCS. After incubation, cells were washed in cold PBS and

resuspended in 100 µl of binding buffer (HEPES containing 2.5 M CaCl2). Fluorescein-labeled

Annexin V and propidium iodide (PI) were added to the cell suspension. Annexin V, a member of

the calcium and phospholipids-binding proteins, binds strongly and specifically to

phosphatidylserine which is a marker of cell apoptosis, while PI binds cell DNA, highlighting

necrotic cells. Cells were then analysed by flow cytometry (Becton Dickinson, USA) and the results

are expressed as % of positive cells/total cells.

Ethanol fractionation

One ml of PSS was treated with 1 ml of cold Ethanol for 30 min at 4°C, centrifuged at 13.000 rpm

at 4°C for 30 min, and the supernatant was discarded. The pellet was diluted in 50% cold Ethanol.

This procedure was repeated for three times and the final pellet was reconstituted in 1ml of PBS and

an aliquote tested on cell viability using the Sulphorodamine B assay.

Salt fractionation

Salt fractionation of 10 µg/ml PSS was performed using NaCl concentrations between 0.2, 0.4, 0.6,

0.8 and 1 M. The serum was pre-treated with the reported NaCl concentrations for 30 min at room

temperature, centrifuged at 13.000 rpm for 5 min and the supernatants were tested on cell viability

using the Trypan Blue assay.

Tryptic digestion of S-carboxamidomethylated P2 (S-CM-P2) and peptide mass fingerprinting

Sequencing grade trypsin (Promega, Madison, WI) was added to a final enzyme:protein ratio of

1:50 w/w and the mixture was incubated at 37°C overnight. The tryptic digestion was stopped by


100

adding 5 µl of 5% TFA. One microliter of S-CM-P2 tryptic digest was deposited on an

AnchorChip target plate (Bruker Daltonics, Bremen, Germany) and allowed to dry; 0.35 µl of

matrix (α-cyano-4-hydroxycinnamic acid 5g/l in 50/50 acetonitrile/0.1% TFA) were then added

and, again, allowed to dry. Mass spectrometric (MS) analysis was performed on an Ultraflex

MALDI TOF TOF (Bruker Daltonics) by using Flex Control 2.4 data acquisition software. Mass

spectra were acquired in reflectron mode over the m/z range 800-3500. The instrumental parameters

were chosen by setting the ion source 1 at 25 kV, the reflector at 26.30 kV and the delay time at 20

ns. The instrument was externally calibrated prior to analysis by using the Bruker peptide calibrant

kit (1000-3000 Da) and the sample spectra internally recalibrated with trypsin autolysis signals. For

MS data acquisition, a total of 400 shots were collected. The peptide masses present in each mass

spectrum, through the integrated software Biotools 3.0, are used to search the NCBInr databank.

Table 1S. Cytotoxic effect of P. sebae serum (PSS)

A. Cell suspension: TB assay. (% dead cells/total cells counted)

PSS[µg/ml] CVEC HF A431 U373MG MCF-7 LLC

0 5±2 3±1 7±2 8±5 6±2.4 6±3

1000 99±5*** 96±6*** 100±0*** 100±0*** 95±6*** 98±1***

400 96±3*** 87±5*** 98±1*** 100±0*** 90±8.3*** 98±0***

200 86±7*** 70±8*** 94±2*** 99±1*** 91± 3*** 95±3***

100 21±5** 33±6** 88.6±5*** 80±5*** 51± 8** 77±6***

50 12±9 9±4 45±9* 42±11* 24±3* 28±11*

25 6±4 5±3 20±8* 17±5* 15±0.4* 4.2±3


101

Serum

[µg/ml]

CVEC HF A431 U373MG MCF-7

PSS

[0]

[200]

[50]

5 2

86 7***

12 9

3 1

70 8***

9 4

7 2

94.2 2***

45 9*

8.4 5

98.9 1***

42 11*

5.7 2.4

90 2***

ND

PSS2

[0]

[200]

[50]

6.1 3

79.2 7***

9.8 4.5

4.3 3

78.25 16***

15 7

8.3 1

96.2 4.5***

37 3.6*

6.2 1

87.2 2.9***

35 9.6*

5.7 2.4

91 3***

23.8 3*

PRS1

[0]

[200]

[50]

5 3

1.8 2.3

3.8 4

4 3

26.2 10.8

6.4 3.8

7.3 5

17.2 5.7

5 3.6

8.2 6

17.2 7

0.5 0.3

5 2

14.7 4.5

4.6 4

PRS2

[0]

[200]

[50]

3.8 2

2.2 2.2

2 3

5 2

6.9 5

2.7 2

5.5 3

14 4

0.8 1.6

3.2 3

3.8 2.5

2.3 3

9 1

10 8.3

3.9 0.6

B. Cell adhesion: Sulphorhodamine B assay (% dead cells/total cells counted)

PSS[µg/ml] CVEC HF A431 U373MG MCF-7

0 10±2 13±1 7±2 9±5 10±2.4

1000 55±3** 50±2** 95±5*** 79±5*** 82±6***

200 35±2* 30±3* 84±7*** 67±4*** 70± 5***

100 16±2 14±6 48±3** 53±5** 50± 4**

Cytotoxycity was measured by the Trypan Blue exclusion assay. Cell suspensions were exposed to serial dilutions of P. sebae serum (PSS) for 1 h at 37°C in medium containing 10% FCS. Data are expressed as the percentage of dead cells over the total number of cells counted and are the mean ± SD of four experiments run in duplicate.***p<0.001 vs basal; **p<0.01 vs basal; *p<0.05 vs basal.

Table 2S. Comparison of the effect of sera sampled from different Python species on cytotoxicity


102

Table 3S. Effect of temperature on PSS cytotoxic activity. TB assay. (% dead cells/total cells counted)

PSS 37°C 56°C

CVEC 98.25 ± 2.4 7.9 ± 1.5

A431 98.8 ± 1.6 9.75 ± 1.9

PSS (200 µg/ml) were kept at 37°C and 56°C for 15 min and then cytotoxicity was evaluated on CVEC and A431 viability by TB assay. Data are expressed as the percentage of dead cells over the total number of cells counted and are the mean ± SD of four experiments run in duplicate.

Figure 1S. Cytotoxic effect of PSS on tumor and non-tumor cells. Cell morphology of tumor and not tumor cell lines after 1 h exposure to PSS was evaluated following staining with Diff Qick. Cells treated with 10% FCS or with 10% FCS in the presence of 100 µg/ml PSS for 1 h (CVEC, panels I and II, HF, panel III and IV, A431, panel V and VI, MCF-7 panels VII and VIII and U373MG panels IX and X respectively). The cell morphology was evaluated by optical microscope Nikon Eclipse T400 at 200X magnification and the images were recorded by Nikon CCD camera at the end of incubation time.

(VIII) MCF-7 (IX)

(X) U373MG (XI)

(VI) A431

(I) CVEC (II)

(III) HF (IV)

(VII)

Basal (10% FCS) + PSS 10 µg/ml


103

Figure 2S. Peptide mass fingerprinting by MALDI TOF TOF. A. shows an enlargement of MALDI MS spectrum of P25 tryptic digest. Six peptides match with the sequence of PIP from P. reticulatus. The sequence 164-177 is observed both at m/z 1564.78, in which the two cysteine residues present are alkylated (C*), and at m/z 1448.69, in which the same two cysteine are linked through a disulphide bridge. The peptide signal at m/z 1485.80 corresponds to the sequence 59-70 with a glutamic acid residue (E) replacing an aspartic acid residue (D) in position 64. B. shows a MALDI MS MS spectrum of the peptide signal at m/z 1564.78.

A

B

4. ALANINE:GLYOXYLATE AMINOTRANSFERASE (AGT)

4.alanine:glyoxylate aminotransferase

107

CHAPTER 4

Molecular defects of the glycine 41 variants of alanine:glyoxylate

aminotransferase associated with primary hyperoxaluria type I.

Barbara Cellinia,1, Riccardo Montiolia,1, Alessandro Paiardinib, Antonio Lorenzettoa, Fabio Masetc, Tiziana Bellinid, Elisa Oppicia and Carla Borri Voltattornia

aDipartimento di Scienze Morfologico-Biomediche, Sezione di Chimica Biologica, Facoltà di Medicina e Chirurgia, Università degli Studi di Verona, Strada Le Grazie, 8, 37134 Verona, Italy; bDipartimento di Scienze Biochimiche "A. Rossi Fanelli" and Centro di Biologia Molecolare del Consiglio Nazionale delle Ricerche, Università "La Sapienza", 00185 Roma, Italy; cDipartimento di Scienze Farmaceutiche, Università di Padova, via F. Marzolo 5, 35131 Padova, Italy; dDipartimento di Biochimica e Biologia Molecolare, Università di Ferrara, via Borsari 46 44100 Ferrara, Italy. 1B.C. and R.M. contributed equally to this work Published in Proc Natl Acad Sci U S A. 2010;107(7):2896-901

INTRODUCTION

Alanine:glyoxylate aminotransferase (AGT) is a homodimeric pyridoxal 5’-phosphate (PLP)

dependent enzyme which catalyzes the interconversion of L-alanine and glyoxylate into pyruvate

and glycine. Human AGT has been cloned, expressed in E. coli and purified. The enzyme crystal

structure, complexed with the competitive inhibitor amino-oxyacetic acid, was determined at a

resolution of 2.5 Å. Each subunit includes a N-terminal extension (residues 1-21), a large N-

terminal domain (residues 22-282) containing the PLP-binding lysine (K209), and a smaller C-

terminal domain (residues 283-392) (1). Steady-state and pre-steady-state kinetic studies featuring

the AGT transamination revealed high specificity for glyoxylate to glycine processing, consistent

with a key role of AGT in glyoxylate detoxification (2). The human liver-specific AGT is localized

in the peroxisomal matrix (3). The enzyme has been the focus of extensive clinical research because

its functional deficiency causes primary hyperoxaluria type 1 (PH1). PH1 is a rare autosomal

recessive disorder characterized by excessive synthesis and excretion of oxalate and glycolate, and

progressive accumulation of insoluble oxalate in kidneys and urinary tract (4). The AGT gene

(AGXT) occurs normally as one of the two allelic forms: the major (AGT-Ma) or minor (AGT-Mi)

alleles. The latter, comprising two coding polymorphisms, P11L and I340M, and a non-coding


108

duplication in intron 1 (5), has no dramatic effect on the properties of AGT. To date, well over one

hundred pathogenic mutations associated to AGT-Ma and/or AGT-Mi are known (6). Three

categories of enzymatic phenotypes causing PH1 can be identified: deficiency of AGT catalytic

activity but not AGT immunoreactivity, catalytic activity and immunoreactivity deficiency, and

mistargeting to mitochondria (4, 7). Notably, clinical data for single PH1 AGT mutations are

generally limited to a small number of individuals, which may interfere with identifying clear

correlations between disease characteristics and properties of mutant proteins. Currently, the way of

treatment of this progressive and potentially fatal disease is poor, as the molecular bases of the

effects of the various disease-associated point mutations are unknown. In the last years, a

biochemical characterization of the pathogenic variants G82E-Ma and F152I-Mi allowed us to

correlate the clinical and enzymatic phenotypes with the structural and functional properties of the

corresponding variants (2,8).

The G41 series of pathogenic mutations, including the G41R encoded on the background of the

major (G41R-Ma) and the minor (G41R-Mi) alleles, and the G41V, which only co-segregates with

the major allele (G41V-Ma), is of special interest because (i) G41 is an interfacial residue making

van der Waals contacts with the same residue of the other subunit and belongs to an α-helix

connected with the active site loop 24-32 (Fig. S1), (ii) G41R mutation is more severe when occurs

on AGT-Mi than on AGT-Ma (4), and (iii) responsiveness to pyridoxine therapy for the patients

bearing these mutations is so far unknown (9). Previous clinical and cell biochemical studies

suggested that the weakening of the dimeric structure of AGT consequent to G41 replacements

could be responsible for depletion of immunoreactive AGT, its intraperoxisomal aggregation (4,

10), and sensitivity to proteasomal degradation (11,12). Since formal proves of the effect of these

mutations at the molecular level are so far absent, we thought to provide insights into the molecular

basis of the G41 mutation’s pathogenicity. Biochemical data indicate for the first time that G41

mutations, besides causing a weakening of the intersubunit interaction, alter the conformational

state of the AGT dimeric form, reduce its resistance to thermal inactivation and unfolding, and

induce susceptibility to proteolytic degradation and self-aggregation. Moreover, predictions of the

structural effects caused by G41 mutation by means of molecular dynamics (MD) provide a

possible interpretation and explanation of our in vitro results.


109


Construction, expression and purification of G41 variant. AGT-Ma and AGT-Mi were prepared

as reported (2, 8). Site-directed mutagenesis, expression and purification of G41 variants in the C-

terminal His-tagged form were performed using standard procedures as described under SI.

Enzyme activity assays. Pyruvate formation was measured by the spectrophotometric assay already

reported (2). Kinetic parameters for the pair alanine/glyoxylate of G41 variants were determined in

the presence of 150 µM PLP by varying the substrate concentrations at a fixed saturating co-

substrate concentration. Data were fitted to the Michaelis-Menten equation. Thermal inactivation

experiments were performed as follows: enzyme (10 µM) was incubated for 10 min in 100 mM

potassium phosphate buffer, pH 7.4, at different temperatures, and then chilled on ice.

Transaminase activity was measured as indicated above.

Binding affinity for PLP. The KD(PLP) of G41 variants were determined by measuring the PLP-

induced changes either on the intrinsic fluorescence or in the CD visible spectrum of the

apoenzymes. The experimental conditions and the relative data analysis are given in the SI.

Cross-linking and SEC experiments. Photo-induced cross-linking with TBPR was performed as

previously described (7). SEC experiments were done on an Akta FPLC system (GE Healthcare)

using a custom packed Sephacryl S-300 10/600 column. The data were analyzed according to the

method of Manning et al. (16). Details are given in SI.

Proteinase K digestion and mass spectrometry analysis. Holo and apo AGT-Ma, AGT-Mi and

G41 variants (15 µM) were treated with proteinase K in 100 mM potassium phosphate buffer, pH

7.4, at 25°C at various AGT/proteinase K (w/w) ratios. At various times, 15 µl-aliquots were

withdrawn from the reaction mixtures and subjected to enzymatic activity assay, SDS-PAGE and

immunoblotting. The reaction was stopped by adding PMSF or EGTA to a final concentration of 2

mM to each aliquot. After staining with Coomassie Blue the band intensities were visualized and

analyzed using ImageJ software (Wajne Rasband, Bethesda, NY). Immunoblotting was made as

described in SI. LC-MS analyses of G41R-Mi were carried out with a model Mariner Esi-Tof

spectrometer from PerSeptive Biosystems (Stafford, TX) connected to a C4 Grace-Vydac

microbore column (1 x 50 mm). Column elution was carried out with a CH3CN-1% HCOOH

gradient from 1 to 80% in 30 min. Proteolysis reaction of G41R-Mi with proteinase K was analyzed


110

with a model 4800 Plus MALDI TOF-TOF instrument from Applied Biosystems (Foster City, CA).

Samples were desalted on a P10 C4 Zip-Tip (Bedford, MA), eluted with a sinapinic acid (Sigma)

saturated solution (2 µl) in CH3CN:H2O (60:40 by vol). Details are given in supporting information.

Turbidimetry measurements. The aggregation experiments were carried out in potassium

phosphate buffer, pH 7.4, at different I values and/or different enzyme concentrations. The turbidity

was monitored by measuring the absorbance at 600 nm as already reported (17).

DLS measurements. DLS measurements were made on a Zetasizer Nano S device from Malvern

Instruments (Malvern, Worcestershire, UK). The temperature of sample cell was controlled by a

thermostating system within ± 0.1 °C and 12.5 x 45-mm disposable cells with stopper were used.

To study the aggregation kinetics, an aliquot of each enzymatic species was diluted to a final

concentration of 4 µM in potassium phosphate buffer pH 7.4 at the desired I and temperature. PLP

was added to the holoenzyme solutions to a final concentration of 60 µM. The buffer was filtered

immediately before use to eliminate any impurities. TMAO or betaine were added to the buffer

before the addition of G41R-Mi.

Spectroscopic measurements. Absorption, fluorescence and CD spectra were performed as

described under supporting information.

Difference Scanning Calorimetry. DSC experiments were conducted with a VP-DSC

microcalorimeter (Microcal) in the temperature interval from 20 to 90 °C, with a scan rate of 90

°C/hour in 100 mM potassium phosphate buffer, pH 7.4, and at 5.5 µM protein concentration.

Computational analyses. MD simulations of the molecular models of AGT-Mi and G41R-Mi,

derived from the crystal structure of human AGT (1), downloaded from Brookhaven Protein Data

Bank (PDB) (18), were performed. A detailed description of the generation of these structures and

MD simulations is given in supporting information. Electrostatic computations on AGT-Mi and its

N-terminus deleted form were carried out by solving the nonlinear Poisson-Boltzmann equation,

one of the most popular continuum models for describing electrostatic interactions between

molecular solutes in salty, aqueous media. Adaptive Poisson Boltzman Solver was used to this

purpose (19), with a protein and solvent dielectric of 2.0 and 80.0, respectively, and I = 150 mM .


111

RESULTS

G41 mutations affect the spectral features and the coenzyme binding affinity. Like AGT-Ma

and AGT-Mi, G41 variants bind 2 mol of PLP per dimer, and exhibit visible absorbance and CD

spectra similar to those of AGT-Ma or AGT-Mi, even if their absorbance and dichroic maxima are

about 10-12 nm blue shifted (Fig. S2 and inset). The KD(PLP) values for G41R-Ma, G41V-Ma and

G41R-Mi were found to be 1.5 ± 0.4 µM, 0.55 ± 0.1 µM and 6.0 ± 0.5 µM, respectively, that are ~

6-, 2- and 23-fold higher than that of AGT-Ma or AGT-Mi (2, 8). Both AGT-Ma and AGT-Mi bind

pyridoxamine 5’-phosphate (PMP) with a KD(PMP) value << 0.1 µM, while even a prolonged time of

incubation of the apo forms of G41 variants with PMP (up to 5 mM) does not result in a dichroic

signal at 340 nm, typical of the AGT-PMP complex (2). Thus, G41 mutations exert a decrease in

the PLP binding affinity and a dramatic reduction in the PMP binding affinity. The comparison of

near-UV CD spectra as well as of intrinsic and 1-anilinonaphthalene sulfonic acid (ANS)

fluorescence spectra of AGT-Ma, AGT-Mi and G41 variants provides evidence that a different

conformation exists between each apoenzymatic form and the corresponding holo form, and

between the holo and apo forms of G41 variants and the corresponding forms of AGT-Ma or AGT-

Mi (Fig. S3A and B). These data imply that (i) conformational changes seem to accompany the apo

to holo transition for each enzymatic form, and (ii) G41 mutations affect the overall conformation

of both holo and apo AGT.

The far-UV CD spectra of G41 variants have been compared with those of AGT-Ma and AGT-

Mi. All spectra exhibit minima at 210 and 222 nm, typical of proteins containing appreciable

amounts of α-helix. However, spectra deconvolution reveals that AGT-Ma, AGT-Mi and G41V-Ma

have an identical composition of the overall secondary structure, while both G41R-Ma and G41R-

Mi display about 5% less α-helical content.

G41 mutations slightly affect the kinetic parameters. The kinetic parameters of AGT-Ma, AGT-

Mi and G41 variants for the pair alanine-glyoxylate are reported in Table S1. The Km values of G41

variants for L-alanine and glyoxylate are not significantly altered while the kcat values decrease by

1.5-3.5-fold as compared to those of the corresponding AGT-Ma and AGT-Mi. Altogether, these

data indicate that the G41 mutations slightly affect the catalytic properties of AGT.

G41 mutations increase the dimer-monomer equilibrium dissociation constant (Kd). As a first

step to investigate the impact of G41 mutations on the AGT dimeric structure, photo-induced cross-

linking experiments with Tris(2,2’-bipyridyl) ruthenium(II) chloride (TBPR) have been carried out


112

at 1 µM enzyme concentration. TBPR cross-linking is incomplete, and also gives rise to the

formation of aggregates and intramolecular cross-linked monomers. Nevertheless, it is possible to

estimate the relative population of chemically cross-linked dimer to monomer which results to be

similar for all the holo species, but lower for apo G41 variants than for apoAGT-Ma and apoAGT-

Mi (Fig. S4). To validate these data we used the very accurate and sensitive size-exclusion

chromatography (SEC) method. Holo and apo forms of AGT-Ma and AGT-Mi as well as the holo

forms of the G41 variants from 5 to 0.1 µM concentration (the latter value being the detection

limit), eluted as a single peak with a retention volume corresponding to a dimer. Thus, the Kd values

of these species must be << 0.1 µM. On the other hand, the apo forms of G41 variants over the

range 50-0.1 µM enzyme concentration eluted as a single peak whose position varied between the

dimeric and the monomeric forms of the enzyme, indicating a rapid equilibrium process. Plots of

the percent dimer as a function of apoG41R-Ma and apoG41R-Mi concentrations give hyperbolic-

like curves, the linear transformation of which yields the Kd values of 0.32 ± 0.05 µM and 1.8 ± 0.5

µM, respectively. In the case of apoG41V-Ma, a decrease of the integrated peak area starts at 1.5

µM reaching at 0.1 µM a value about 30% that expected, possibly because of monomerization

followed by artefactual aggregation. A Kd value ranging from 1.5 to 0.1 µM can be estimated for

this mutant. The SEC results, consistent with those of cross-linking analyses, do not allow to

quantify the impact of G41 mutations on the Kd value of holoAGT, but clearly indicate that the

mutations increase the Kd value of apoAGT.

G41 mutations induce susceptibility to proteinase K digestion. Limited proteolysis was used to

further probe that replacements of G41 change the overall conformation of AGT. AGT-Ma, AGT-

Mi and the G41 variants in the holo and apo forms were incubated in 100 mM potassium phosphate

buffer, pH 7.4, at 25°C with proteinase K at different AGT/protease weight ratios (from 5000/1 to

100/1). When aliquots of these reaction mixtures were withdrawn at different times and subjected to

SDS-PAGE, it was observed that while the size of the band (~ 42.7 kDa) corresponding to the intact

holo or apo forms of AGT-Ma and AGT-Mi remains unaltered even after a prolonged time of

incubation (Fig. S5A), that of G41 variants gradually decreases and simultaneously a faster

migrating band (~ 40.2 kDa) appears. Both the 42.7 and 40.2 kDa bands stain with antibody raised

against the C-terminal hexahistidine tag, as revealed by Western blot analysis, indicating that the

cleavage occurs within the N-terminus. In fact, Maldi-mass spectrometry analysis yields a

molecular weight difference for the N-terminally truncated fragment of 5503±15 a.m.u. compared

with the full-length G41R mutant, compatible with a cleavage site located at the peptide bond

Met53-Tyr54. The effect of proteinase K on holo G41R-Mi is shown in Fig. 1, and that on holo


113

G41V-Ma in Fig. S5B. The initial velocity values of the cleavage (expressed as µg enzyme/min/µg

proteinase K) determined by measuring the decrease of the intensity of the band corresponding to

the intact enzyme, are 4 ± 1, 530 ± 40, and 6 ± 1 for the holo forms of G41R-Ma, G41R-Mi and

G41V-Ma, respectively, and 130 ± 30, 870 ± 90 and 30 ± 6 for the corresponding apo forms. The

proteolytic cleavage is accompanied by a time dependent loss of transaminase activity occurring for

holoG41R-Mi with an initial velocity of 420 ± 50 µg mutant/min/µg proteinase K, a value that

agrees with that of degradation. At a 10/1 AGT/proteinase K weight ratio, G41 variants are

degraded in less than half-hour to low-molecular weight peptides, whereas AGT-Ma and AGT-Mi

remain unaltered. Altogether these data indicate that, unlike AGT-Ma and AGT-Mi, G41 variants

undergo digestion and concomitant inactivation more pronounced for their apo than for the

corresponding holo forms.

Figure 1. Effect of proteinase K on holoG41R-Mi. HoloG41R-Mi (15 µM) was incubated at 25 °C in 100 mM potassium phosphate buffer, pH 7.4, at a 1000⁄1 (wt/wt) mutant/proteinase K ratio. At times indicated, aliquots were removed, treated (see Methods), and subjected to 12% SDS-PAGE. Plus and minus signs indicate presence or absence of proteinase K.

G41 mutations induce formation of high order inactive aggregates. We noticed that, unlike

AGT-Ma and AGT-Mi, holoG41 variants at an ionic strength (I) value lower than 260 mM and/or at

an enzyme concentration higher than 10 µM form visible insoluble aggregates. When these

solutions are centrifuged, a yellow pellet indicative of PLP-bound aggregates appears. The specific

activity of the precipitates resuspended in the assay buffer is <5% than that of the supernatant

containing the dimeric species. This observation, together with the documented presence of

intraperoxisomal aggregates in patients bearing G41R mutation (10), led us to investigate in some


114

detail this phenomenon by means of turbidimetry and DLS studies under physiological conditions,

i.e at 37°C, pH 7.4 at I = 150 mM. As shown in Fig. 2, significant changes in turbidity began after a

lag phase ranging from 50 to 120 min for both the holo and apo forms of G41R-Mi as well as for

the apo forms of G41R-Ma and G41V-Ma. No turbidity could be detected for these enzymatic

species at pH 7.4, and I = 260 mM at either 25°C or 37°C at least over 5 hours interval. Fig. 3A and

B shows plots of the total count rate as function of time for holo and apoenzymes of AGT-Ma,

AGT-Mi and G41 variants. The increase in count rate for the holo and apo forms of AGT-Ma and

AGT-Mi is very slow, while the more typical fast aggregation for G41 variants could be seen with

the count rate leveling off after ~ 20-40 min. The decrease in the light scattering intensity occurring

for holoG41R-Mi and for the apo forms of G41 variants is due to precipitation of the protein

aggregates.

Figure 2. Time dependence of turbidity of AGT-Ma, AGT-Mi, and G41 variants. Absorbance at 600 nm

as a function of time of 4 µM apo forms of AGT-Ma (▪), AGT-Mi (▾), G41R-Ma (▴), G41V-Ma (♦), and

G41R-Mi (•) in potassium phosphate buffer, pH 7.4, I = 150 mM at 37 °C. Corresponding holoenzymes, open symbols.


115

Figure 3. Time-dependence of total count rate (measured as kilo counts per second) of AGT-Ma, AGT-Mi, and G41 variants in the holo (A) and apo (B) forms. Measurements performed at 4 µM enzyme concentration, 37 °C, I = 150 mM, pH 7.4. Color code: black, AGT-Ma; red, AGT-Mi; blue, G41R-Ma; green, G41V-Ma; fuchsia, G41R-Mi.

Altogether, the turbidity and DLS data indicate that (i) replacements of G41 are responsible for

the propensity of mutants to self-association, the extent of which is more pronounced for the apo

forms than for the corresponding holo forms, and (ii) G41R-Mi results the variant most prone to

aggregation. The molecular size of the species present in the enzymic solutions has been also

evaluated. The particle size of the dimeric form of AGT-Ma, AGT-Mi and G41 variants, measured

at 25°C, pH 7.4, at I = 260 mM, i.e. under conditions of no detectable association, is about 10 nm.

This value is consistent either with that (9.32 nm) derived from the X-ray structure (1) or that (7 ± 2

nm) calculated by an appropriate empirical equation (13). When the aggregation process occurring

under the physiological conditions mentioned above was followed by DLS, the time dependence of

the apparent particle size indicates that the 10 nm species is always present for holo and apo AGT-

Ma and holoAGT-Mi, although small aggregates (100-800 nm) appear after about 30 min (Fig. 4A


116

and B). Considering that the scattering intensity is proportional to the sixth power of the particle

diameter, the dimer must be in these species very abundant in number. In contrast, the dimer

disappears over a 5-75 min time range, depending on the enzymatic species, for AGT-Mi in the apo

form and for G41 variants in both the holo and apo forms. Moreover, while in apoAGT-Mi only

small aggregates (100-800 nm) accumulate, in G41 variants a distinct population of higher order

aggregates (~ 5000 nm), along with small aggregates, can be identified (Fig. 4B-E). From these data

it can be also envisaged that (i) the turbidity is mainly associated to the presence of high order

aggregates, and (ii) the lack of detectable turbidity for G41R-Ma and G41V-Ma in the holo form

could be ascribed to the very low fractional contribution of the larger particles to the total scattering

intensity.

The effect of 200 mM trimethylamine-N-oxide (TMAO) or 100 mM betaine (two osmolytes

known to suppress protein aggregation (14)), on the aggregation process of holoG41R-Mi has been

followed by DLS. Either TMAO or betaine, although unable to decrease the amount of and size of

the aggregates formed at equilibrium, as detected by the final count rate leveling off, cause a slight

increase in the lag time duration (from 10 to 20 min) and in the persistence of the dimeric species

(from 35 to 70 min) in the aggregation kinetics.


117

Figure 4. Time dependence of the apparent diameters of AGT-Ma, AGT-Mi, and G41 variants. Experimental conditions (see legend to Fig. 3). (A) holo and apo AGT-Ma. (B) holo and apo AGT-Mi. (C) holo and apo G41R-Ma. (D) holo and apo G41V-Ma. (E) holo and apo G41R-Mi. Color code: black, holo dimer; green, apo dimer; red, holo small aggregates; blue, apo small aggregates; cyan, holo high aggregates; fuchsia, apo high aggregates.


118

G41 mutations decrease resistance to thermal denaturation and inactivation. A differential

calorimetric study (DSC) was carried out to reveal the impact of G41 mutations against thermal

denaturation. Thermal denaturation of holo and apo forms of AGT-Ma, AGT-Mi, G41R-Ma and

G41V-Ma as well as of holo form of G41R-Mi produces DSC profiles consisting of an endothermic

transition due to the protein denaturation followed by a large exothermic transition due to the

aggregation of the denaturated protein (Fig. S6), precluding the peaks deconvolution. Thus, the

maxima of the endothermic transitions of these species represent apparent melting temperature

(Tm) values. The Tm of apoG41R-Mi could not be determined because of the lack of an observable

transition, possibly due to overlapping of the denaturation and aggregation processes (Table 1). The

thermal effect on the catalytic function of AGT-Ma, AGT-Mi and G41 variants was also measured,

and the transition midpoints of thermal inactivation (Ti) are reported in Table 1. The results indicate

that (i) holo AGT-Ma, AGT-Mi and G41 variants display Tm and Ti values higher than those of the

corresponding apoenzymes, which supports different conformational states of holo and apo forms,

(ii) AGT-Ma is more resistant than AGT-Mi, and (iii) G41 mutations decrease the thermal stability,

being G41R-Mi the most thermally instable variant.

Table 1. Transition midpoints of thermal denaturation (Tm) and inactivation (Ti) of AGT-Ma, AGT-Mi, and G41 variants

ND, not detectable. *They represent apparent Tm values and are the mean of two independent experiments. The error is within ±0.3 °C error. †Data inactivation points were subjected to nonlinear regression analysis and Ti values were calculated. The data are the means of at least two independent experiments. The standard error of the mean was less than 5% of the mean value in every case.

MD studies. We examined in a comparative way the conformational space sampled by the putative

structures of AGT-Mi and G41R-Mi by high-temperature (500 K) MD simulations with explicit

water solvation. Both AGT-Mi and G41-Mi reach an equilibrated state after ~ 100 ps (Fig. S7A),

G41V-Ma

G41R-Mi

G41R-Ma

AGT-Mi

AGT-Ma

Enzyme

54.562.358.361.0

46.051.8ND53.7

53.057.757.660.3

52.272.655.673.2

59.177.462.477.3

Ti† (°C)Tm* (°C)

Apo formHolo formApo formHolo form

G41V-Ma

G41R-Mi

G41R-Ma

AGT-Mi

AGT-Ma

Enzyme

54.562.358.361.0

46.051.8ND53.7

53.057.757.660.3

52.272.655.673.2

59.177.462.477.3

Ti† (°C)Tm* (°C)

Apo formHolo formApo formHolo form


119

thereafter their global architecture remaining stable, as confirmed by the indicators commonly used

to analyse MD simulations (Fig. S7A-C). Unlike what observed for AGT-Mi, a marked fluctuation

can be observed for the region spanning residues 1-44 of both monomers of G41R-Mi. This region

comprises the active site loop (residues 24-32) and the α-helix 34-42 in which the G41R

substitution takes place. In particular, the MD simulation of G41R-Mi reveals that the N-terminal

α-helices (residues 34-42) undergo a progressive displacement (Fig. S8), and a partial unwinding of

the first and second turns of the helix (Fig. 5). It is likely that these events could be related to the

accommodation of the long side-chain of Arg41.

Figura 5. Comparison of the initial 3D model of G41R-Mi (dark gray) with the averaged structure obtained from MD simulation (light gray). Residues 1–46, roughly corresponding to the N-terminal arm of G41R-Mi, are highlighted in orange for the initial 3D model and in red for the averaged structure. (Inset) Detail of the α-helix in which the G41R substitution takes place (residues 34–42).

DISCUSSION

G41, located at the end of the α-helix 34-42 of AGT, is an interfacial residue which makes van

der Waals contacts with the same residue of the adjacent subunit (1). Thus, it is not surprising that


120

(i) the conversion of the small, uncharged Gly41 to the bulky, charged Arg causes a ~ 5% loss of

α-helix content, while the conversion to Val, a larger uncharged residue, has not a detectable effect

on the overall secondary structure, and (ii) the apo forms of G41 variants undergo reversible

dissociation of the subunits with Kd values from at least ~3-fold to ~20-fold higher than that of

apoAGT-Ma or apoAGT-Mi. Actually, we found that G41 variants in the dimeric form differ from

AGT-Ma or AGT-Mi under many respects. Their structural conformation and stability are altered,

as detected by difference in the near-UV CD and intrinsic emission fluorescence spectra, larger

exposure of hydrophobic surfaces, sensitivity to proteinase K cleavage, and decrease in Tm and Ti

with respect to AGT-Ma and AGT-Mi. Additionally, G41 variants show slightly altered visible

spectroscopic features, a reduced steady-state catalytic activity, a decreased PLP binding affinity

as well as a dramatic reduction in the PMP binding ability.

A comparative study of the putative structures of AGT-Mi and G41R-Mi by MD predicts that

the G41R mutation would cause the partial unwinding of the α-helix 34-42 as well as the

displacement of the N-terminal arm and its exposure to the environment. Although these predictions

are of course not an experimental evidence of the structural effects caused by the mutation, they are

consistent with the in vitro data of G41R-Mi, i.e., the 5% loss of the α-helix content and the

susceptibility of the Met53-Tyr54 peptide bond to proteinase K cleavage. It is worth noting that,

although G41 is far from the active site, it is located in the α-helix connected with the active site

loop (residues 24-32) belonging to the N-terminus (Fig. S1). Therefore, we might speculate that the

alterations of the visible spectroscopic and catalytic features of G41 variants could be due to the

rearrangements occurring around the mutated residue transmitted to the active site.

Another particularly interesting aspect is that, unlike AGT-Ma and AGT-Mi, G41 variants

spontaneously form insoluble inactive high order aggregates (~ 5000 nm) under physiological

conditions of temperature, I and pH. The finding that the aggregation extent greatly decreases as I

increases may be ascribed to the reduction in favorable attractive interactions due to screening

effects. This behavior indicates that aggregation is not due to hydrophobic interactions, in which

case the addition of salt would enhance aggregation and little effect would be seen at low I, but

rather to electrostatic intermolecular interactions arising from protein charge heterogeneity. It is

also worthy noting that the aggregation occurs at a pH value lower than pI, thus excluding an

isoelectric precipitation. Why are the G41 variants aggregation-prone ? In the absence of the

crystal structure of these variants, the protein electrostatic potential distribution for AGT-Mi and

the 1-44 N-terminus lacking form of AGT-Mi has been calculated by electrostatic computer

modeling. The truncated form was chosen for comparison because it would mimic the structure of

G41R-Mi upon the displacement of the N-terminal arm. As shown in Fig. 6A and B, AGT-Mi


121

displays a highly positive charge distribution around its surface, while the truncated form exhibits

a clear dipole segregation of charges. In fact, the depletion of the highly positively charged N-

terminal arm of AGT-Mi leads to the exposure of several negatively charged residues (Asp51,

Glu59, Glu62, Glu274, Glu281, Asp344, Glu346 of each monomer), i.e to a condition of

electrostatically-driven self-aggregation of AGT. Thus, one can reason out that the fluctuation of

the N-terminus in G41R-Mi could cause the exposure of negative charged residues similar to that

observed for the truncated AGT form. Although this could be a plausible explanation for the

propensity of G41 variants to aggregate spontaneously, a conclusive evidence for this

interpretation is lacking.

Figure 6. Electrostatic potential surface maps of AGT-Mi and its truncated form. Electrostatic gradient and map (kT⁄e) of AGT-Mi (A), and 1–44 truncated form of AGT-Mi (B).

The finding that G41R-Mi displays structural and functional alterations more pronounced than

G41R-Ma needs to be discussed. The thermal unfolding data indicate that the instability of G41R-

Mi is due to the additive contribution of the polymorphic and pathogenic mutations. The effect of

the polymorphic substitutions is possibly due to P11L in that Pro instead of Leu (i) contributes to

the increase of structural rigidity, (ii) exerts additional constraints to the backbone, since the ring

closure keeps its Ф angle value almost fixed, and (iii) makes hydrophobic interactions with a

surface cavity surrounded by residues Leu14, Glu62, Gly63, Tyr66 and Ala280.

Altogether, our data assist in assessing a picture of the G41 variants enzymatic phenotype

more exhaustive than that previously proposed. The pathogenicity of G41 variants has been until

now related to a disruption of the interface and impairment of dimerization resulting in formation

of monomers with reduced catalytic activity and prone to degradation and/or intraperoxisomal

aggregation (1, 4, 15). Our in vitro results would indicate not only the impact of G41 mutations on


122

the dimerization but also provide evidence that G41 variants in the dimeric form are prone to

degradation and aggregation. The presence of aggregates only within the peroxisomal matrix in

patients bearing G41R mutation (10) is consistent with this view in that the peroxisomal import

machinery acts on folded dimeric proteins. In any case, it will be important to establish if the

intraperoxisomal aggregates have ultrastructure similarities with the aggregates spontaneously

formed in vitro. If this were the case, the electrostatically-driven protein aggregation between

folded dimers could represent a novel pathogenic mechanism by which G41 inherited mutations

would cause protein aggregation. Indeed, many disease-related genetic mutations are known to

alter the folding or stability of proteins leading to intermediates in which hydrophobic patches

become exposed and prone to self-association.

Overall, our work improves the understanding of the correlation between the genotype and the

enzymatic phenotype, thus allowing us to foresee the response to pyridoxine in patients carrying the

G41 mutations. Administration of pyridoxine to these patients is not sufficient to counteract the

disease because the molecular defects of these variants appear to be related to structural

rearrangements yielding molecules that are both in the apo and the holo forms prone to degradation

and aggregation. A promising therapeutic strategy could be the administration of small molecules

able to stabilize the native state of the protein, thus preventing degradation and aggregation. In this

regard, our preliminary results on the effects of osmolytes on the aggregation behaviour of G41

variants could be an encouraging perspective.

REFERENCES

1. Zhang X, et al. (2003) Crystal structure of alanine:glyoxylate aminotransferase and the

relationship between genotype and enzymatic phenotype in primary hyperoxaluria type 1. J Mol

Biol 331:643–652.

2. Cellini B, Bertoldi M, Montioli R, Paiardini A, Borri Voltattorni C (2007) Human wild-type

alanine:glyoxylate aminotransferase and its naturally occurring G82E variant: Functional

properties and physiological implications. Biochem J 408:39–50.

3. Motley A, et al. (1995) Mammalian alanine/glyoxylate aminotransferase 1 is imported into

peroxisomes via the PTS1 translocation pathway. Increased degeneracy and context specificity

of the mammalian PTS1 motif and implications for the peroxisome- to-mitochondrion

mistargeting of AGT in primary hyperoxaluria type 1. J Cell Biol 131:95–109.


123

4. Danpure CJ (2005) Molecular etiology of primary hyperoxaluria type 1: New directions for

treatment. Am J Nephrol 25:303–310.

5. Purdue PE, Takada Y, Danpure CJ (1990) Identification of mutations associated with

peroxisome-to-mitochondrion mistargeting of alanine/glyoxylate aminotransferase in primary

hyperoxaluria type 1. J Cell Biol 111:2341–2351.

6. Coulter-Mackie MB, Rumsby G (2004) Genetic heterogeneity in primary hyperoxaluria type 1:

Impact on diagnosis. Mol Genet Metab 83:38–46.

7. Lumb MJ, Danpure CJ (2000) Functional synergism between the most common polymorphism in

human alanine:glyoxylate aminotransferase and four of the most common disease-causing

mutations. J Biol Chem 275:36415–36422.

8. Cellini B, Montioli R, Paiardini A, Lorenzetto A, Voltattorni CB (2009) Molecular insight into

the synergism between the minor allele of human liver peroxisomal alanine:glyoxylate

aminotransferase and the F152I mutation. J Biol Chem 284:8349–8358.

9. Coulter-Mackie MB, Lian Q, Wong SG (2005) Overexpression of human alanine:glyoxylate

aminotransferase in Escherichia coli: Renaturation from guanidine-HCl and affinity for

pyridoxal phosphate co-factor. Protein Expres Purif 41:18–26.

10.Danpure CJ, et al. (1993) Enzymological and mutational analysis of a complex primary

hyperoxaluria type 1 phenotype involving alanine:glyoxylate aminotransferase peroxisome-to-

mitochondrion mistargeting and intraperoxisomal aggregation. Am J Hum Genet 53:417–432.

11.Coulter-Mackie MB, Lian Q (2008) Partial trypsin digestion as an indicator of mis-folding of

mutant alanine:glyoxylate aminotransferase and chaperone effects of specific ligands. Study of a

spectrum of missense mutants. Mol Genet Metab 94:368–374.

12.Coulter-Mackie MB, Lian Q (2006) Consequences of missense mutations for dimerization and

turnover of alanine:glyoxylate aminotransferase: Study of a spectrum of mutations. Mol Genet

Metab 89:349–359.

13.Wilkins DK, et al. (1999) Hydrodynamic radii of native and denatured proteins measured by

pulse field gradient NMR techniques. Biochemistry 38:16424–16431.

14.Venkatesu P, Lee MJ, Lin HM (2007) Thermodynamic characterization of the osmolyte effect

on protein stability and the effect of GdnHCl on the protein denatured state. J Phys Chem B

111:9045–9056.

15.Danpure CJ (2006) Primary hyperoxaluria type 1: AGT mistargeting highlights the fundamental

differences between the peroxisomal and mitochondrial protein import pathways. Biochim

Biophys Acta 1763:1776–1784.


124

16.Manning LR, Dumoulin A, Jenkins WT, Winslow RM, Manning JM (1999) Determining

subunit dissociation constants in natural and recombinant proteins. Methods Enzymol 306:113–

129.

17.Yong YH, Foegeding EA (2008) Effects of caseins on thermal stability of bovine beta-

lactoglobulin. J Agric Food Chem 56:10352–10358.

18.Sussman JL, et al. (1998) Protein data bank (PDB): Database of three-dimensional structural

information of biological macromolecules. Acta Crystallogr D 54: 1078–1084.

19.Baker NA, Sept D, Joseph S, Holst MJ, McCammon JA (2001) Electrostatics of nanosystems:

Application to microtubules and the ribosome. Proc Natl Acad Sci USA 98:10037–10041.

SUPPORTING INFORMATION

Materials. Pyridoxal 50-phosphate (PLP), pyridoxamine 50-phosphate, L-alanine, glyoxylate,

rabbit muscle L-lactic dehydrogenase, isopropyl β-thiogalactopyranoside, Tris(2,20-bipyridyl)

ruthenium (II) chloride (TBPR), trimethylamine-N-oxide, and betaine were all purchased from

Sigma. ANS (1-anilinonaphthalene sulfonic acid) was purchased from Molecular Probes. Anti-his

(C-term) antibody and ECL™ peroxidase labeled anti-mouse antibody were from Invitrogen. All

other chemicals were of the highest purity available.

Construction, Expression, and Purification of G41 Variants. Site-directed mutagenesis was

performed with the QuikChange II sitedirectedmutagenesis kit (Stratagene). G41V-Ma and G41R-

Ma variants were generated using the pAGThis-Ma construct as template, whereas the G41R-Mi

construct was created on the pAGThis-Mi template. The oligonucleotides used for mutagenesis

were as follows: G41V forward primer, 5’GCAGCCGTAGGGCTGCAGATGATCGG and its

complement; G41R forward primer, 5’GCAGCCCGCGGGCTGCAGATGATCGG and its

complement. The underlined codons indicate mutated amino acids. All mutations were confirmed

by DNA sequencing. G41R-Ma, G41R-Mi, and G41V-Ma were expressed and purified by using the

same procedure as alanine:glyoxylate aminotransferase (AGT)-Ma and AGT-Mi (1, 2). In the case

of G41R-Mi, the expression and purification were also carried out in the presence of 60 µM

exogenous PLP. The purified variants were homogenous as indicated by a single band on SDS-

PAGE with a mobility identical to that of AGT-Ma and AGT-Mi. Yields of G41R-Ma and G41V-

Ma after the standard purification were slightly lower than that of AGT-Ma, whereas yield of

G41R-Mi was about 20% with respect to that of AGT-Ma but increases to 40% when exogenous


125

PLP was added to the bacterial culture and the buffer during purification. The protein concentration

in the AGT samples was determined by absorbance spectroscopy using an extinction coefficient of

9.4 × 104 M−1 cm−1 at 280 nm (1). The PLP content of G41R-Ma and G41V-Ma was determined

by releasing the coenzyme in 0.1 M NaOH and by using ε = 6600 M−1 cm−1 at 388 nm, whereas

that of G41R-Mi was determined by spectrophotometric titration of apoenzyme (11 µM) with

increasing PLP concentrations (2–50 µM). The transition point of the biphasic plot of A430 versus

[PLP] was taken as the maximum value of PLP bound to the mutant (i.e., 1.84 mol per mol of the

dimeric species). The apo forms of the enzymes were prepared as already described (1).

Binding Affinity for PLP. The KD(PLP) from G41R-Ma (0.5 µM), G41R-Mi (0.5 µM), and G41V-

Ma (0.1 µM) were determined by measuring the quenching of the intrinsic fluorescence of

apoenzyme in the presence of PLP at concentrations ranging from 0.2 to 60 µM, 0.5 to 60 µM, and

0.1 to 10 µM, respectively. The KD(PLP) from G41R-Mi (4.5 µM) and G41V-Ma (1.5 µM) was also

determined by CD spectrophotometric titration at 420 nm of their apo forms in the presence of PLP

at concentrations ranging from 1 to 100 µM and 0.1 to 50 µM, respectively. All these experiments

were carried out in 100 mM potassium phosphate buffer, pH 7.4. The KD(PLP) values of G41R-Ma,

G41R-Mi, and G41V-Ma mutant-coenzyme complexes were obtained using the following equation:

[ ] [ ] [ ] [ ]( ) [ ] [ ][ ]{ } [ ]ttt2

)PLP(Dtt)PLP(Dttmax E2/PLPE4KPLPEKPLPEYY −++−++= [S1]

where [E]t and [PLP]t represent the total concentrations of the mutant and PLP, respectively, Y

refers to either the intrinsic fluorescence quenching or the 420 nm dichroic signal changes at a PLP

concentration, [PLP]t and Ymax refers to the aforementioned changes when all enzyme molecules are

complexed with coenzyme.

Determination of Kd. Size-exclusion chromatography (SEC) was used to analyze the amount of

monomer or dimer present in the solutions of AGT-Ma, AGT-Mi, and G41 variants in the holo and

apo forms at different enzymes concentrations. The analysis was performed on an Akta FPLC

system (GE Healthcare) using a custom packed Sephacryl S-300 10⁄600 column at room

temperature. The flow rate was 0.4 mL⁄ min, and the injection volume was 500 µL with detection at

280 nm. The mobile phase was 100 mM potassium phosphate buffer, pH 7.4 for the apoenzyme

samples, and the same buffer containing 60 µMPLP in the case of holoenzymes. To determine the

Kd, a stock solution of AGT-Ma, AGT-Mi, or G41 variants in the apo or holo forms was dissolved

in 100 mM potassium phosphate buffer, pH 7.4, to different enzyme concentrations, and after 3 h of


126

incubation at 25 °C (a sufficient time to reach the equilibrium between monomer and dimer),

subjected to SEC. Each sample was injected in triplicate, and the elution volume and data handling

were done using the software Unicorn 5.01 (Amersham Biosciences). The rapid equilibrium process

observed for G41 variants in the apo form was analyzed according to the method of Manning et al.

(3). This is a treatment that mathematically relates the protein concentration (in terms of the

theoretical maximum concentration of dimer) to the expected amounts of dimer and monomer for

an associating–dissociating equilibrium. The elution volumes of each sample have been measured

in terms of average peak concentration, which is calculated as

(sample volume) × (sample concentration) / (peak volume),

where the total peak volume is estimated to be equivalent to the volume at half-peak height.

The Kd value has been estimated as follows (3). If the maximum amount of AGT dimer is [E]

and the concentrations of monomeric and dimeric species are [M] and [D], respectively, so that

%D=100[D] / [E], it follows that

Kd = [M]2 / [D] = 4([E] – [D])2 / [D] = [(100 - %D)2 [E] / (25%D)] = 0.04 (100-%D) [E] / %D

Hence,

[ ]( )

−−=

2dD%10004.0

)D(%logElog)Klog( [S2]

Thus, a plot of log[(%D ⁄ 0.04 (100 − %D)2] with respect to log[E] yields a straight line of slope 1.

When [%D ⁄ 0.04 (100 − %D)2] = 1, Kd = [E]. The quadratic equation [S2] can also be solved in

terms of (%D):

[ ]( ) [ ][ ] [ ]E08.0/)EK16K(KE8D% d2dd +−+= [S3]

Spectroscopic Measurements. Absorption spectra were made with a Jasco V-550

spectrophotometer. CD spectra were obtained using a Jasco J-710 spectropolarimeter with a

thermostatically controlled compartment at 25 °C. For near-UV and visible wavelengths, the protein

concentration varied from 0.5 to 1 mg⁄mL in a cuvette with a path length of 1 cm. Routinely, three

spectra were recorded at a scan speed of 50 nm min−1 with a bandwith of 2 nm and averaged


127

automatically. Secondary structure content was calculated from far-UV CD spectra using the

Dicroweb software. Fluorescence spectra were recorded with a Jasco FP-750 spectrofluorometer

using 5 nm excitation and emission bandwidths. Spectra of blanks, i.e., samples containing all

components except the enzyme, were taken immediately before the measurements of samples

containing protein. Tryptophan emission spectra were taken from 300 to 500 nm using an excitation

wavelength of 280 nm at a protein concentration of 1 µM. For ANS binding experiments, a stock

solution of ANS was added to a final concentration of 15 µM to solutions of AGT-Ma, AGT-Mi,

and G41 variants in the holo and apo forms. The excitation wavelength was 365 nm, and the

emission was recorded from 400 to 560 nm. The values were normalized by subtracting the baseline

recorded for the probe alone under identical conditions. To determine the emission maximum

position, fluorescence spectra were processed by Jasco software.

Immunoblotting. Samples of G41 variants (15 µM) before and after treatment with proteinase K,

were separated by SDS-PAGE. The proteins were then transferred to a nitrocellulose Immobilon P

membrane (Millipore). After blocking at 4 °C with 5% dry milk, 0.05% Tween 20 in Tris buffered

saline (TBS), the membrane was incubated with anti-His (C-term) antibody, washed three times

with TBS, and detected with horseradish peroxidase secondary antibody by chemiluminescence.

Mass Spectrometry. Mass spectrometry analyses of G41R-Mi were carried out with a model

Mariner ESI-TOF spectrometer from PerSeptive Biosystems. Specifically, 5 µL of a 35 µM solution

of G41R-Mi in 100 mM potassium phosphate buffer, pH 7.4, were loaded onto a C4 Grace-Vydac

microbore column (1 × 50 mm), equilibrated for 20 min at a flow-rate of 20 µL⁄ min with

H2O∶CH3CN (99∶1), containing 1% HCOOH and eluted with a linear CH3CN-1%HCOOH

gradient from 1% to 80% in30 min. Spray-tip potential was set at 3.0 kV, while the nozzle potential

and temperature were set at 200 V and 140.0 °C, respectively. Proteolysis reaction of G41R-Mi

with proteinase K was analyzed with a model 4800 Plus Maldi-Tof-Tof instrument from Applied

Biosystems. Samples were desalted and concentrated by loading 10 µL of proteolysis reaction of

G41R-Mi (35 µM) onto a P10 C4 Zip-Tip, eluted manually with a sinapinic acid (Sigma) saturated

solution (2 µL) in CH3CN∶H2O (60∶40 by vol). The relative intensity of the laser was set at 5000

and delay time at 1800 ns.

Molecular Dynamics (MD). For the MD simulation of AGT-Mi and G41R-Mi, the crystal

structure of human AGT, downloaded from Brookhaven Protein Data Bank (PDB code: 1H0C),

was used as starting structure. The structure was checked and missing residues (1–3; 121–122; 392)


128

were modeled by making use of Modeller 9v5 (4); the full-length protein sequence of human

AGTwas retrieved by the National Center for Biotechnology Information site (GI code: 4557289)

and manually aligned to 1H0C. Sequences were identical, except for missing residues. A dimeric

structure was modeled, with heteroatoms included and symmetry restraints applied to each

monomer. Ten different models were built by using the default built-in refinement procedure and

the one displaying the lowest objective function, which measures the extent of violation of

constraints from the template, was taken as the representative model. From this starting model, the

script “mutate_model.py” of the Modeller package (which takes a given PDB file and mutates a

single residue) was used to derive the AGT-Mi and G41R-Mi models. The residue’s position was

then optimized by conjugate gradients energy minimization. PROCHECK (5) was used to monitor

the stereochemical quality of the final models, whereas Prosa2003 (6) was used to measure the

overall protein quality in packing and solvent exposure. The energy profiles and stereochemical

quality of the models were, as expected, almost identical to the crystal structure used as template.

Because the dimeric crystal structure was obtained in the presence of two cofactor PLP moieties,

the Amber 10 force field and the auxiliary Antechamber tool (7) were adopted to recognize the

atom and bond type of the system, to generate residue topology file, and to find missing force field

parameters. To this end, the AnteChamber Python Parser Interface was implemented in an in-house

developed python script (dynamics.py; source code available upon request). The resulting structure

was energy minimized in vacuum, using the Amber 10 force field and Gromacs 3.3.3 package (8).

Briefly, after an initial minimization performed to allow added hydrogens to adjust to the

crystallographically defined environment, a 5000-step steepest descent minimization without

periodic boundary conditions was performed, until the maximum derivative was less than 0.01 kJ

Mol−1 Å−1. Then, the system was soaked by a water box of 97 × 97 × 97 Å3 dimensions, filled with

a total of 20,136 water molecules and counterions added to neutralize the net negative charge of the

protein and obtain a concentration of 0.01 M Na+⁄Cl−. Energy minimization was then performed for

5000-step steepest descent, but in this case periodic boundary conditions were introduced. With the

greatest strain dissipated from the system, the next step was to let the solvent adapt to the protein,

while keeping the non-hydrogen atoms of the proteins fixed to the reference positions. For this

purpose, a position restrainedMD was performed for 10 ps at 200 K. After pressure coupling, the

system was gradually heated at 500 K during 50 ps simulation at temperature increments of 50 K

every 10 ps. Finally, a production simulation was carried out at 500 K for 3 ns, with a time step for

integration of 0.2 fs. Bond lengths were constrained by using the LINCS algorithm. Standard

quality assurance tests were carried out by means of the Gromacs package for the convergence of

the following thermodynamic parameters: temperature, potential, and kinetic energy. Convergence


129

was also checked in terms of the structure, through the rmsd against the starting structure (Fig.

S7A). Next to that, it was checked that during MDsimulation there were not interactions between

adjacent periodic images, because such interactions could lead to unphysical effects. MD

parameters files and trajectory files are available upon request to the authors. MD simulations were

carried out on a eightcore MacPro running Mac OS X 10.5.

REFERENCES

1. Cellini B, Bertoldi M, Montioli R, Paiardini A, Borri Voltattorni C (2007) Human wild-type

alanine:glyoxylate aminotransferase and its naturally occurring G82E variant: Functional

properties and physiological implications. Biochem J 408:39–50.

2. Cellini B, Montioli R, Paiardini A, Lorenzetto A, Voltattorni, CB (2009) Molecular insight into

the synergism between the minor allele of human liver peroxisomal alanine:glyoxylate

aminotransferase and the F152I mutation. J Biol Chem 284:8349–8358.

3. Manning LR, Dumoulin A, Jenkins WT, Winslow RM, Manning JM (1999) Determining subunit

dissociation constants in natural and recombinant proteins. Methods Enzymol 306:113–129.

4. Sali A, Blundell TL (1993) Comparative protein modelling by satisfaction of spatial restraints. J

Mol Biol 234:779–815.

5. Laskowski D, Mac ArthurMW, Moss DS, Thornton JM (1993) PROCHECK: A program to

check the stereochemical quality of protein structures. J Appl Crystallogr 26:9.

6. Sippl MJ (1993) Recognition of errors in three-dimensional structures of proteins. Proteins

17:355–362.

7. Case DA, et al. (2008) AMBER 10 (University of California, San Francisco) http://ambermd.org.

8. Van Der Spoel D, et al. (2005) GROMACS: Fast, flexible, and free. J Comput Chem 26:1701–

1718.


130

Figure S1. AGT active site and G41 location. The figure shows the positioning of G41 and the active site loop 24–32 with respect to the active site. One subunit is colored orange, the other is light blue, G41 and residues 24–32 are violet, and PLP is green. Active site residues are also indicated.

Figure S2. Visible absorption and CD spectra of AGT-Ma and G41 variants. (A) Absorption of AGT-Ma (—), G41R-Ma (⋯), and G41V-Ma (-·-·-) in 100mMpotassium phosphate buffer, pH 7.4, at a concentration of 16 µM. (Inset) Differential absorption spectrum of 14 µMG41R-Mi in the presence of 60 µMexogenous PLP (—).(B) CD spectra registered in the presence of 60 µM exogenous PLP. Symbols for AGT-Ma and G41 variants are the same as for (A). Absorption and CD spectra of AGT-Mi are identical to the corresponding of AGT-Ma.

A

B

A

B


131

Figure S3. Near-UV CD and ANS emission spectra of AGT-Ma and G41 variants. (A) Near-UV CD spectra of AGT-Ma or AGT-Mi (—), G41R-Mi (- - -), G41R-Ma (⋯), and G41V-Ma (-·-·-) in 100 mM potassium phosphate buffer, pH 7.4, at a concentration of 16 µM. Symbols for apoenzymes are the same as the corresponding holoenzymes. (B) Holoenzymes in the presence of 60 µM exogenous PLP (—) and apoenzymes (—) at a concentration of 3 µM were incubated with 50 µM ANS for 1 h at 25 °C. Emission spectra were monitored upon excitation at 365 nm. ANS emission spectra of holo and apo AGT-Mi are identical to the corresponding of AGT-Ma.

Figure S4. Dimerization of AGT-Ma, AGT-Mi, and G41 variants as determined by photo-cross-linking and SDS-PAGE. AGT-Ma, AGT-Mi, and G41 variants (1 µM) were photo-cross-linked with TBPR and separated by SDS-PAGE. Gels were then immunoblotted and detected by anti-His C-term antibody. Molecular mass markers are shown on the right-hand side; positions of uncross-linked monomer, intramolecular cross-linked monomer, dimer and high-order cross-linked products are shown on the left-hand side. The dimer-monomer ratio, estimated from band intensities using ImageJ quantitation software, is indicated for each enzymatic species. (A) holoenzymes, (B) apoenzymes. Lanes: 1, uncross-linked AGT-Ma; 2, cross-linked AGT-Ma; 3, cross-linked AGT-Mi; 4, cross-linked G41R-Ma; 5, cross-linked G41R-Mi; 6, cross-linked G41V-Ma.


132

Figure S5. Effect of proteinase K on the subunit molecular weight of AGT-Ma, AGT-Mi, and G41V-Ma in the holo form. The enzymes (15 µM) were incubated at 25 °C in 100 mM potassium phosphate buffer, pH 7.4, at a 100⁄1 (wt⁄wt) mutant/proteinase K. At the indicated times, aliquots were removed, treated as indicated in the Methods section, and subjected to 12% SDS-PAGE. (A) HoloAGT-Ma and holoAGT-Mi, (B) G41V-Ma. In (A) and (B), lane C represents untreated AGT-Ma and G41V-Ma, respectively. Figure S6 Thermal denaturation of AGT-Ma, AGT-Mi, and G41 variants in the holo form. DSC profiles of holo AGT-Ma (―), holoAGT-Mi (-.-.-.-.-.), holoG41V-Ma (- - - - -), holoG41R-Ma (. . . . . .), and holoG41R-Mi (-.-.-.-).


133

Figure S7. MD simulation evolution. (A) rmsd from initial position for backbone atoms vs. time; (B) gyration radius vs. time; (C) fluctuations (rmsf) for backbone atoms around the averaged structure. Figure S8. Average putative structure of G41R-Mi. The superposition of 3,000 frames (every picosecond) obtained fromMD simulations was done. The backbone of G41R-Mi has been colored on the basis of the rmsf. The α-helix in which the Gly41 → Arg residue substitution takes place is boxed


134

Table S1. Steady-state kinetic parameters of AGT-Ma, AGT-Mi, and G41 variants for the pair alanineglyoxylate

*From ref. 1 †From ref. 2

0.41 ± 0.04

143 ± 220.13 ± 0.02

42 ± 417.1 ± 0.5

18.3 ± 0.7

Glyoxylate

L-alanineL-alanine

GlyoxylateG41V-Ma

0.35 ± 0.07

35 ± 20.32 ± 0.02

30 ± 510.6 ± 0.5

11.1 ± 0.3

Glyoxylate

L-alanineL-alanine

GlyoxylateG41R-Mi

0.90 ± 0.07

50 ± 50.41 ± 0.04

22 ± 219.8 ± 0.4

20.5 ± 0.6

Glyoxylate

L-alanineL-alanine

GlyoxylateG41R-Ma

1.2 ± 0.2†

168 ± 8†0.22 ± 0.01†

28 ± 2†33 ± 5†

37 ± 1†

Glyoxylate

L-alanineL-alanine

GlyoxylateAGT-Mi

1.4 ± 0.2*

196 ± 44*0.23 ± 0.05*31 ± 4*45 ± 2*

45 ± 3*

Glyoxylate

L-alanine

L-alanine

GlyoxylateAGT-Ma

Kcat/Km (S-1mM-1)Km glyoxylate (mM)Km L-alanine (mM)Kcat(S-1)CosubstrateSubstrateEnzyme

0.41 ± 0.04

143 ± 220.13 ± 0.02

42 ± 417.1 ± 0.5

18.3 ± 0.7

Glyoxylate

L-alanineL-alanine

GlyoxylateG41V-Ma

0.35 ± 0.07

35 ± 20.32 ± 0.02

30 ± 510.6 ± 0.5

11.1 ± 0.3

Glyoxylate

L-alanineL-alanine

GlyoxylateG41R-Mi

0.90 ± 0.07

50 ± 50.41 ± 0.04

22 ± 219.8 ± 0.4

20.5 ± 0.6

Glyoxylate

L-alanineL-alanine

GlyoxylateG41R-Ma

1.2 ± 0.2†

168 ± 8†0.22 ± 0.01†

28 ± 2†33 ± 5†

37 ± 1†

Glyoxylate

L-alanineL-alanine

GlyoxylateAGT-Mi

1.4 ± 0.2*

196 ± 44*0.23 ± 0.05*31 ± 4*45 ± 2*

45 ± 3*

Glyoxylate

L-alanine

L-alanine

GlyoxylateAGT-Ma

Kcat/Km (S-1mM-1)Km glyoxylate (mM)Km L-alanine (mM)Kcat(S-1)CosubstrateSubstrateEnzyme

5. VON WILLEBRAND FACTOR (VWF)

5.von Willebrand Factor

137

CHAPTER 5

Serine proteases from primary granules of leukocytes efficiently

cleave oxidized von willebrand factor: divergence from ADAMTS-13.

Stefano Lancellottia#, Vincenzo De Filippisb#, Nicola Pozzib, Laura Oggianua,c, Sergio Rotellad,e,Giovanni Luca Scaglionea, Fabio Masetb, Flora Peyvandif, Pier Mannuccio Mannuccif

and Raimondo De Cristofaroa

aInstitute of Internal Medicine and Geriatrics, and Haemostasis Research Centre, Catholic University School of Medicine, Rome, Italy; bDepartment of Pharmaceutical Sciences, University of Padua, Italy; cSection of Biology Applied to Human Health,“Rome Tre” University, Rome, Italy; dDepartment of Hematology, Catholic University School of Medicine, Rome, Italy; eIRCCS San Raffaele Pisana, Rome, Italy; fA. Bianchi Bonomi Hemophilia and Thrombosis Center, University of Milan and Department of Medicine and Medical Specialties, IRCCS Maggiore Hospital, Mangiagalli and Regina Elena Foundation, Luigi Villa Foundation, Milan, Italy. #S.L. and V.D.F. contributed equally to this work Journal of Thrombosis and Haemostasis. Submitted for pubblication. INTRODUCTION

Activated polymorphonuclear cells (PMNs) are actively involved in the defense of the human

body against infections and use several biochemical strategies to pursue this aim (1, 2). In response

to diverse stimuli, activated PMNs secrete granule proteases and a series of cytotoxins, as well as

superoxide anion (O2•-) and other ROS (3-5). Moreover, PMNs are not merely defensive cells but

participate also in the hemostatic functions by interacting with both platelets and endothelial cells in

primary hemostasis (6, 7). Recent studies have documented indeed a specific cleavage by the PMNs

serine proteases elastase (HLE), proteinase 3 (PR3) and cathepsin G (CG) of the A2 domain of von

Willebrand factor (VWF) (8). The cleavage sites are located near or at the same Tyr1605-Met1606

peptide bond hydrolyzed by ADAMTS-13 (A Disintegrin-like And Metalloprotease with

ThromboSpondin type I repeats), the zinc protease that specifically controls the proteolytic

processing of VWF, when the latter is in a stretched conformation induced by high shear stress (8).

On the other hand, recent studies showed that oxidation of Met1606 by hypochlorous acid and

peroxynitrite with the formation of methionine-sulfoxide blocks the VWF proteolysis by

ADAMTS-13 (9, 10).


138

VWF, beside its well known engagement in primary haemostasis, participates in other

biological phenomena, such as bacterial infections and leukocyte recruitment and extravasation.

The high molecular weight VWF multimers are involved for instance in bacterial adhesion,

mediated by surface adhesins molecules called MSCRAMMs (Microbial Surface Components

Recognizing Adhesive Matrix Molecules)(11-14). Staphylococcus aureus is known to express

numerous adhesins (15), such as Staphylococcal protein A (Spa), which binds to soluble and

immobilized VWF multimers (16). The process of bacterial adhesion causes migration and

activation of PMNs that secrete enzymes and ROS to eliminate the pathogens. Very recently, VWF

multimers have been demonstrated to mediate also PMN extravasation (17).

Thus, in this study we investigated a) whether PMN-induced oxidative stress influences VWF

hydrolysis by purified leukocyte serine proteases (LPSs) HLE, CG and PR3; b) whether cleavage of

VWF by LSPs secreted by PMNs in a highly oxidative milieu may alter the efficiency of VWF

proteolysis. The investigation of these mechanistic aspects may further unravel the role of

leukocytes-VWF interactions in both hemostatic and extra hemostatic functions of VWF.


Proteolysis of VWF74 and VWF74 Met-SO by human LSPs and identification of the cleavage

sites by mass spectrometry. The pseudo wild-type peptide VWF74, containing the amino acid

exchange Cys1669Ala and encompassing the VWF A2 domain sequence 1596-1669

(DREQAPNLVYM VTGNPASDEIKRLPGDIQVVPIGVGPNANVQELERIGWPNAPILIQDFET

LPREAPDLVLQRA), and its derivative containing MetSO at position 1606 were synthesized by

the solid-phase method and characterized as previously detailed (10). VWF74 or VWF74-MetSO

(120 µl, 20 µM) in 10 mM Hepes, 150 mM NaCl, 5 mM CaCl2, pH 7.4 (HBS) were separately

incubated for 90 min at 37°C with 20 nM PMN-derived HLE, CG and PR3 , all purchased from

Sigma Aldrich (Milano, Italy). For comparison, proteolysis reaction was carried for 60 min under

identical experimental conditions with 5 nM recombinant human ADAMTS-13 (R&D Systems,

Minneapolis, MN). The reaction was stopped by adding 4% aqueous TFA (80 µl) and 6 M Gdn-

HCl (200 µl) and fractionated on a Grace-Vydac (The Separation Group, Hesperia, CA) C18

analytical column (4.6 x 250 mm), using an acetonitrile-0.1%TFA gradient (10 to 30% in 10 min;

30 to 45% in 30 min, 0.8 ml/min). The peptides eluted in correspondence of the chromatographic

peaks were collected, lyophilized and then analyzed by mass spectrometry (MS) on a Mariner ESI

TOF instrument from PerSeptive Biosystems (Stafford, TX), as detailed elsewhere(10).


139

Determination of Michaelis-Menten parameters for proteolysis of VWF74 and VWF74-MetSO by

human LSPs. The Michaelis-Menten parameters for the hydrolysis of VWF74 and VWF74 MetSO

by LSPs and ADAMTS-13 were determined by quantifying the cleavage products by RP-HPLC, as

previously described (10). The VWF74- peptide substrates were used at concentrations ranging

from 1.2 to 20.8 µM, whereas HLE, PR3 and CG were all used at 5 nM. The steady state hydrolysis

of VWF peptides was performed in 10 mM Hepes, 0.15 M NaCl, 3 mM CaCl2, pH 7.5 at 37 °C.

The Michaelis-Menten parameters were calculated by quantifying at 210 nm the N-terminal

peptides (1596Asp-Tyr1605 or 1596Asp-Val1607) released from VWF74 and VWF74-MetSO,

respectively. RP-HPLC and MS analyses showed that proteolysis of VWF74 peptides with PR3,

CG, and ADAMTS-13 in the time-range of the analysis (30-90 min) occurred exclusively at a single

peptide bond, Tyr1605-Met1606 or Val1607-Thr1608 (see Results), allowing us to determine the

parameters kcat and Km pertaining to a single proteolytic process. At variance, HLE cleaved VWF74

peptides at multiple sites with comparable efficiency, even after short incubation time. Thus, in the

latter case only an observed pseudo-first order kinetic constant, kobs, of VWF74 disappearance was

derived at 25 °C and low substrate concentration (4 µM). A lower temperature was chosen in this

case to slow down the proteolytic reaction and monitor with higher precision the products’ release.

Oxidation of VWF multimers by hypochlorous acid and the myeloperoxidase/H2O2/Cl- system

(MOPSY). Plasma VWF multimers were purified and characterized as previously detailed (10,18)

and their hypochlorous acid (HClO)- or myeloperoxidase system (MOPSY)-mediated oxidation was

carried out using a procedure described earlier (19). Intact VWF samples (200 nM as a monomer) in

10 mM Hepes-buffered saline, pH 7.50, were treated for 15 min at 25°C with increasing

concentrations of HClO (i.e., 2.5, 5, 10, 20 and 40 µM), in the absence or presence of L-tyrosine,

added to promote the formation of tyrosyl radicals that contribute to oxidative protein damage and

dityrosine intermolecular cross-links (20). When MOPSY was used, human leukocyte

myeloperoxidase (MPO, 25 nM) was added to VWF samples (200 nM) in HBS in the presence of

H2O2 (200 µM) and L-tyrosine (100 µM). The concentration of hypochlorous acid and hypochlorite

ion in commercial NaOCl solution (Sigma) was determined spectrophotometrically at 230 nm,

using a molar absorptivity of 93.3 and 7.2 M-1·cm-1, respectively19. The concentration of H2O2

solution was determined by UV-absorption at 240 nm, using a molar absorptivity of 350 M-1·cm1

(21). Excess HOCl/NaOCl and tyrosine were eliminated by gel-filtration on a Bio-Rad (Richmond,

CA) DG10 column. Oxidation of the model compound Nα-acetyl-tryptophanyl-amide (NATA) to

its mono- (Oia) and di-oxyindolylalanine (Dia) derivatives was carried out by the DMSO-HCl


140

procedure (22). The reaction was conducted for 15 min and quenched by 1:20 (v/v) dilution with

water. The products were then analyzed by MS and UV-absorption, revealing the presence of both

Oia and Dia in a 3:2 ratio.

Proteolysis of the HClO- and MOPSY-treated VWF multimers by LPSs and ADAMTS-13. After

treatment with HClO or MOPSY, VWF samples (20 µg/ml) were separately incubated at 37 °C

with 10 nM ADAMTS-13, HLE, PR3, and CG in 5 mM Tris-HCl buffer, pH 8.0, containing 3 mM

CaCl2, in the absence or presence of 1.5 mg/ml sulphate-free ristocetin (Helena Laboratories,

Beaumont, TX). At time intervals (i.e, 0, 1, and 2 hr), aliquots (50 µl) of these solutions were

sampled and the reaction stopped by adding 0.3 M acetic acid (final concentration) or 10 mM

EDTA, when ADAMTS-13 was reacted. In all cases, proteolysis of oxidized VWF (VWF-Ox) was

assessed by SDS-agarose electrophoresis in 1.5% agarose gel, using rabbit anti-human VWF

polyclonal antibody and a HRP-conjugated secondary anti-rabbit antibody (Dako, Milano, Italy), as

previously detailed (18).

Hydrolysis of VWF multimers by activated PMNs. Purified VWF multimers (20 µg/ml) were

incubated with human PMNs, isolated from healthy volunteers and stimulated with 50 ng/ml 12-

phorbol 13-myristate acetate (PMA, Sigma) in HBS. PMNs were purified by following a procedure

published elsewhere, with only minor modifications (23): EDTA-treated blood (20 ml) was mixed

with PBS (100 ml) and layered over 30 ml of endotoxin tested Ficoll-Paque solution (GE

Healthcare, Milan, Italy), and centrifuged at 400 g for 30 min at 20°C. The supernatant and the

mononuclear cell/platelet layer were removed by suction, and the bottom fraction containing PMN

cells and red cells was diluted with 10 volumes of 0.83% 7(w/v) NH4Cl to promote red cells lysis.

PMNs were collected by centrifugation (800 g for 15 min), washed twice with 10 mM phosphate

buffered saline, and resuspended in HBS at a concentration of 1-5x l03 cells/µl. Purity of PMN cell

preparations was checked by a FACS Canto® flow cytometer (BD Biosciences, Mountain View,

CA) using FITC-conjugated anti-CD11b or isotype control antibody. Data were then collected by

recording FITC fluorescence at 525 nm. Time from blood withdrawal to testing was less than 2 hr.

Production of superoxide by isolated PMNs in response to PMA was assessed by monitoring the

reduction of ferricytochrome C (24), by recording at 25°C the absorbance at 550 nm, on a model

Benchmark Plus microplate spectrophotometer (Bio-Rad), expressed as µM per 106 PMNs, using

an extinction coefficient of 21 mM-1·cm-1 for (reduced-oxidized) cytochrome C (25). The

concentration of superoxide was determined in the absence and presence of 6 U/ml SOD, 1 mM L-

methionine, and 200 nM catalase, used as superoxide and H2O2 scavengers, respectively. VWF was


141

added to these PMNs suspensions and immediately afterwards PMA was added to stimulate PMNs.

An aliquot of the PMNs suspension was taken and centrifuged at 10,000 rpm for 30 sec, while the

supernatant containing added VWF was used to analyze the pattern of VWF multimers. To inhibit

LSP activity, 500 µM aprotinin and phenyl-methyl-sulphenyl-fluoride (PMSF) were added in

control samples under identical experimental conditions. The SDS-agarose electrophoresis of VWF

multimers treated with activated PMNs was carried out in 0.8 % (stacking gel)-1.5 % (running gel)

agarose gel. The Western blot analysis was performed as detailed above.

Flow cytometry measurement of oxidative metabolism and spectrophotometric measurements of

ROS/LSP secretion by PMN cells. PMNs (1x106/tube) were labeled with 2′-7′-dichlorofluorescein

diacetate (DCFH-DA; 20 µM final concentration) and hydroethidine (HE; 10 µM final

concentration) for 15 min at 37°C to measure H2O2 and superoxide anion production, respectively

(26). Cells loaded with the fluorescent probes were then stimulated with 50 ng/ml PMA. Samples

were immediately analyzed on a FACS Canto® flow cytometer to ensure a true zero value and then

each tube was run individually to measure DCFH-DA and HE fluorescence, every 5 min for up to

25 min. Control tubes consisted of PMN cells without addition of PMA. Fluorescence emission for

DCFH-DA and HE was collected at 525 nm and 620 nm, respectively. A minimum of 10,000

events was acquired in list mode. Samples were analyzed with the FACS Diva software package

(BD Biosciences). Superoxide generation in intact cells was measured at 25°C by

spectrophotometric assay of ferricytochrome c reduction in 1.0 ml cuvettes containing 2×106 cells

and 50 µM cytochrome c in phosphate-buffered saline (PBS), pH 7,4, supplemented with 1.2 mM

MgCl2, 2 mM CaCl2. The mixture was blanked at 550nm in a Bio-Rad Benchmark Plus microplate

reader before adding PMA, and the time-dependent release of superoxide from PMNs was recorded

with or without the addition of superoxide dismutase, catalase and methionine, as ROS scavengers.

Readings were taken every 20 s for a total run time of 30 min. Superoxide dismutase-inhibitable

absorbance was calculated using ε=2·11×104 M−1cm−1 for reduced cytochrome c.

The concentrations of human leukocyte elastase antigen (HLE) bound to the specific inhibitor α1-

anti-protease 1, present in the supernatants of PMA-stimulated PMNs were measured using “PMN-

Elastase ELISA” kit (Abnova, Milano, Italy) according to the manufacturer’s instructions.

Absorbance values were measured with a Bio-Rad Benchmark Spectrophotometer Microplate

Reader (Bio-Rad Laboratories, Hercules, CA) at 450 nm, subtracting the aspecific contribution at

620 nm. Aliquots of the same supernatant solutions, taken from 5 to 30 min after addition of PMA,

were used to measure the level of HLE, by monitoring at 405 nm the enzyme activity toward 200

µM of the synthetic substrate N-(Methoxysuccinyl)-Ala-Ala-Pro-Val 4-nitroanilide (Sigma-


142

Aldrich). The reference curve was constructed using purified HLE over a concentration ranging

from 10 to 100 nM.

Spectroscopic measurements. UV-absorption spectra of intact and oxidized VWF multimers at

different HClO concentrations were recorded on a Varian-Cary (Palo Alto, CA) model 2200

spectrophotomer or on a Jasco (Tokyo, Japan) V-630 instrument. For native VWF, an extinction

coefficient at 280 nm was calculated as 0.846 (mg/ml)-1 cm-1 determined by the method of Pace et

al. (27). Intrinsic fluorescence spectra were recorded on a model Eclipse spectrofluorimeter (Varian,

Santa Clara, CA), equipped with a Peltier temperature control system at 25°C. Both intact and

HClO-oxidized (HClO, 2.5-40 µM for 15 min) VWF multimers were used at a concentration of 200

nM. VWF solutions were excited at 280 or 295, using excitation and emission slits of 2.5 or 5 nm.

Dityrosine production was assessed by exciting the samples at 325 nm and recording the

fluorescence intensity at 410 nm (28). Denaturation experiments of VWF and VWF-Ox were

carried out by recording the fluorescence λmax as a function of guanidine hydrochloride (Gdn-HCl)

concentration. Unfolding data were analyzed according to a two-state model, as previously reported

(27,29). Far-UV CD spectra were taken on a Jasco (Tokyo, Japan) J-810 spectropolarimeter using a

1-mm pathlength cuvette. The final spectra resulted from four accumulations after base line

subtraction and CD signal, [θ], was expressed as the mean residue ellipticity. All spectroscopic

measurements were carried out at 25±0.1°C in HBS.

Dynamic light scattering measurements on native and oxidized VWF samples were carried

out at 25±0.1°C with a Zetasizer Nano S (Malvern Instruments, Malvern, U.K.) at a fixed angle

(i.e., 173°) from the incident light (He-Ne laser power: 4 mW; incident light: 633 nm). Samples

were manually injected into a 1-cm pathlength Suprasil quartz cuvette (45 µl) (Hellma Italia, Milan,

Italy). Data were collected for at least 15 minutes from injection. Each measurement consisted of a

subset of runs automatically determined, each being averaged for 10 s. Peak intensity analysis was

used to determine the average hydrodynamic diameters (Z-average diameter) of the scattering

particles. Polydispersity index was obtained by a cumulative analysis of the intensity

autocorrelation function using the Nano vs. 5.0 software (30).


143

RESULTS

Determination of the cleavage sites by LSPs of VWF74 and VWF74-MetSO.

Proteolysis reactions of wild-type synthetic VWF74 and its MetSO-derivative (VWF74-MetSO) (20

µM) with LSPs (20 nM) and ADAMTS-13 (5 nM) were carried out under identical experimental

conditions and fractionated by RP-HPLC (Fig.1 A-H). Mass spectrometry (MS) analysis of the

peptides eluted from the column allowed us to unequivocally establish the chemical identity of the

proteolytic fragments (Table 1) and to identify the cleavage site(s) in the VWF74 peptide sequence

(Fig. 1I). After 1h-reaction at 37°C, ADAMTS-13 cleaved VWF74 exclusively at the peptide bond

Tyr1605-Met1606 (Fig. 1A), generating the N-terminal peptide 1596Asp-Tyr1605 (p1 = 1203.58

a.m.u.) and the C-terminal peptide 1606Met-Ala1669 (p2 = 6969.94 a.m.u.), which was eluted from

the column at retention times slightly shorter than the parent, uncleaved VWF74 peptide (p3 =

8156.06 a.m.u.). Cleavage of VWF74-MetSO was markedly impaired (Fig. 1E), in agreement with

the results recently reported by us and others (9,10). However, the cleavage site for ADAMTS-13

was unchanged (p1* = 1203.63 a.m.u.). Notably, in the case of VWF74-MetSO, the C-terminal

peptide 1606MetSO-Ala1669 (p2* = 6985.63 a.m.u.) could not be resolved form the uncleaved

peptide (p3* = 8172.62 a.m.u.), likely because the presence of the hydrophilic MetSO-residue in

both p2* and p3* species reduced the difference in their hydrophilic/hydrophobic balance. A similar

behavior was observed for the cleavage of VWF74-MetSO with other LSPs (i.e., CG and PR3). As

found with ADAMTS-13, cathepsin-G (almost) exclusively cleaves either VWF74 and VWF74-

MetSO at the peptide bond Tyr1605-Met1606 (Fig. 1B,F). However, conversely to what we have

observed with ADAMTS-13, CG seems to hydrolyze VWF74-MetSO with significantly higher

efficiency compared to the unmodified VWF74. In Fig. 1C,G is shown that proteinase-3 (PR3)

predominantly hydrolyzes both VWF74 and VWF74-MetSO with comparable efficiency at the

peptide bond Va1607-Thr1608, even though a very minor cleavage at the Ala1612-Ser1613 bond

was also observed. Whereas ADAMTS-13, CG and PR3 proteolyzed wild-type and oxidized

VWF74 at a single peptide bond with nearly absolute specificity, HLE cleaved the substrate at

multiple sites (e.g., Val1607-Thr1608, Val1626-Pro1627, Ala1647-Pro1648, and Ile1649-Leu) with

comparable efficiencies (Fig. 1D,H), reflecting the wider substrate specificity of this protease (5).

As obtained with PR3, oxidation of Met1606 does not seem to influence proteolysis by HLE. This

is well documented by the RP-HPLC chromatograms of the proteolysis reactions of VWF74 (Fig.

1D) and VWF74-MetSO (Fig. 1H), almost superimposable to each other except for the fragment

1596Asp-Val1607 (p7 in Fig. 1D), which in the MetSO form is eluted at shorter retention times

(p2* in Fig. 1H).


144

Figure 1. RP-HPLC analysis of the proteolysis reactions of VWF74 (upper panels) and VWF74-MetSO

(lower panels) with ADAMTS-13 (A, E), and LSPs cathepsin-G (CG) (B, F), proteinase-3 (PR3) (C, G),

and human leukocyte elastase (HLE) (D, H). Proteolysis was conducted under identical experimental

conditions (20 µM substrate, 37°C in HBS, pH 7.4) with ADAMTS-13 (5 nM, 1 h) and LSPs (20 nM, 1.5 h).

Acid-quenched aliquots were fractionated by RP-HPLC and the proteolytic fragments analyzed by MS (see

Table 1). When two peptide species are co-eluted in the same peak, the less abundant fragment (as judged

from MS spectra) is indicated in grey. For each protease, the absorbace scale relative to the proteolysis with

VWF74 and VWF74-MetSO is the same. For clarity, the chromatograms of the reference peptides at reaction

time = 0 is shown only in panels A and E. (I ) Identification of the cleavage sites of VWF74 and VWF74-

MetSO after proteolysis with ADAMTS-13 and LSPs, as deduced from MS data. The major cleavage

sites are indicated by thick arrows and bold labels.

1600 1610 1620 1630 1640 1650 1660 1669

DREQA PNLV-Y-MV-TGN PA-SDEIKRLP GDIQVV-PI-GV GPNANVQELE RI-GWPNA-PI-L I-QDFETLPRE APDLVLQRA

HLEHLE

AD13

PR3

CG

PR3 HLE HLE HLE HLE HLE HLE

I

0

20

40

60

80

100

Per

Cen

t Ace

toni

trile

(

)

15 20 25 30 35 40

Abs

orba

nce

at 2

20 n

m


••

•

p1p3

p2

VWF74

A ADAMTS-13A

U: 0

.05

0

20

40

60

80

100

Per

Cen

t Ace

toni

trile

(

)

15 20 25 30 35 40

Abs

orba

nce

at 2

20 n

m


••

•

p1p3

p2

VWF74

A ADAMTS-13A

U: 0

.05

15 20 25 30 35 40

Abs

orba

nce

at 2

20 n

m


••

•

p1p3

p2

VWF74

A ADAMTS-13A

U: 0

.05

AU

: 0.0

5

15 20 25 30 35 40


B CG

p1

p2

p3

AU

: 0.0

3

15 20 25 30 35 40


B CG

p1

p2

p3

AU

: 0.0

3A

U: 0

.03

15 20 25 30 35 40


C PR3

p1 p2

p3p4

AU

: 0.0

4

15 20 25 30 35 40


C PR3

p1 p2

p3p4

AU

: 0.0

4A

U: 0

.04

15 20 25


D HLE

p5+p6

p4

p3p2p1 p11

p7p8

p9p10

AU

: 0.0

2

15 20 25


D HLE

p5+p6

p4

p3p2p1 p11

p7p8

p9p10

AU

: 0.0

2

0

20

40

60

80

100 P

er

Cen

t Ace

toni

trile

(

)

15 20 25 30 35 40

Abs

orba

nce

at 2

20 n

m


••

•

E

p1*

p2*+p3*

VWF74-MetSO

ADAMTS-13

AU

: 0.0

5

0

20

40

60

80

100 P

er

Cen

t Ace

toni

trile

(

)

15 20 25 30 35 40

Abs

orba

nce

at 2

20 n

m


••

•

E

p1*

p2*+p3*

VWF74-MetSO

ADAMTS-13

AU

: 0.0

5A

U: 0

.05

15 20 25 30 35 40


p1*

p2*+p3*F CG

AU

: 0.0

3

15 20 25 30 35 40


p1*

p2*+p3*F CG

AU

: 0.0

3

15 20 25 30 35 40


p3*+p4*G PR3

p1*+p2*

AU

: 0.0

4

15 20 25 30 35 40


p3*+p4*G PR3

p1*+p2*

AU

: 0.0

4A

U: 0

.04

15 20 25


H HLE

p1*

p4*

p2*

p3*p7*

p8*

p10*

p9*

AU

: 0.0

2

p5*+

p6*

15 20 25


H HLE

p1*

p4*

p2*

p3*p7*

p8*

p10*

p9*

AU

: 0.0

2

p5*+

p6*

1600 1610 1620 1630 1640 1650 1660 1669


HLEHLE

AD13

PR3

CG


I1600 1610 1620 1630 1640 1650 1660 1669


HLEHLE

AD13

PR3

CG


I

0

20

40

60

80

100

Per

Cen

t Ace

toni

trile

(

)

15 20 25 30 35 40

Abs

orba

nce

at 2

20 n

m


••

•

p1p3

p2

VWF74

A ADAMTS-13A

U: 0

.05

0

20

40

60

80

100

Per

Cen

t Ace

toni

trile

(

)

15 20 25 30 35 40

Abs

orba

nce

at 2

20 n

m


••

•

p1p3

p2

VWF74

A ADAMTS-13A

U: 0

.05

15 20 25 30 35 40

Abs

orba

nce

at 2

20 n

m


••

•

p1p3

p2

VWF74

A ADAMTS-13A

U: 0

.05

AU

: 0.0

5

15 20 25 30 35 40


B CG

p1

p2

p3

AU

: 0.0

3

15 20 25 30 35 40


B CG

p1

p2

p3

AU

: 0.0

3A

U: 0

.03

15 20 25 30 35 40


C PR3

p1 p2

p3p4

AU

: 0.0

4

15 20 25 30 35 40


C PR3

p1 p2

p3p4

AU

: 0.0

4A

U: 0

.04

15 20 25


D HLE

p5+p6

p4

p3p2p1 p11

p7p8

p9p10

AU

: 0.0

2

15 20 25


D HLE

p5+p6

p4

p3p2p1 p11

p7p8

p9p10

AU

: 0.0

2

0

20

40

60

80

100 P

er

Cen

t Ace

toni

trile

(

)

15 20 25 30 35 40

Abs

orba

nce

at 2

20 n

m


••

•

E

p1*

p2*+p3*

VWF74-MetSO

ADAMTS-13

AU

: 0.0

5

0

20

40

60

80

100 P

er

Cen

t Ace

toni

trile

(

)

15 20 25 30 35 40

Abs

orba

nce

at 2

20 n

m


••

•

E

p1*

p2*+p3*

VWF74-MetSO

ADAMTS-13

AU

: 0.0

5A

U: 0

.05

15 20 25 30 35 40


p1*

p2*+p3*F CG

AU

: 0.0

3

15 20 25 30 35 40


p1*

p2*+p3*F CG

AU

: 0.0

3

15 20 25 30 35 40


p3*+p4*G PR3

p1*+p2*

AU

: 0.0

4

15 20 25 30 35 40


p3*+p4*G PR3

p1*+p2*

AU

: 0.0

4A

U: 0

.04

15 20 25


H HLE

p1*

p4*

p2*

p3*p7*

p8*

p10*

p9*

AU

: 0.0

2

p5*+

p6*

15 20 25


H HLE

p1*

p4*

p2*

p3*p7*

p8*

p10*

p9*

AU

: 0.0

2

p5*+

p6*


145

Table 1. MS data of the proteolytic fragments obtained by reaction of VWF74 and VWF74-MetSO (*) with ADAMTS-13 and LSPs a

ADAMTS-13

VWF74 VWF74-MetSO

Peak ID

Mass (a.m.u.) Fragment Sequence

Peak ID Mass (a.m.u.) Fragment Sequence

p1 1203.58 (1203.59)b 1596Asp-Tyr1605 p1* 1203.63 (1203.59) 1596Asp-Tyr1605

p2 6969.94 (6969.98) 1606Met-Ala1669 p2*c 6985.63 (6985.97) 1606Met-Ala1669

p3 8156.06 (8156.26) 1596Asp-Ala1669 p3* 8172.62 (8172.25) 1596Asp-Ala1669

Cathepsin G (CG)

p1 1203.59 (1203.59) 1596Asp-Tyr1605 p1* 1203.61 (1203.59) 1596Asp-Tyr1605

p2 6969.92 (6969.98) 1606Met-Ala1669 p2* 6986.25 (6985.97) 1606Met-Ala1669

p3 8156.74 (8156.26) 1596Asp-Ala1669 p3* 8172.91 (8172.25) 1596Asp-Ala1669

Proteinase 3 (PR3)

p1 1433.42 (1433.70) 1596Asp-Val1607 p1* 1449.73 (1449.69) 1596Asp-Val1607

p2 1873.90 (1873.90) 1596Asp-Ala1612 p2* 1889.90 (1889.89) 1596Asp-Ala1612

p3 6739.56 (6739.65) 1608Thr-Ala1669 p3* 6986.25 (6985.97) 1608Thr-Ala1669

p4 8156.78 (8156.26) 1596Asp-Ala1669 p4* 8172.47 (8172.25) 1596Asp-Ala1669

Leukocyte Elastase (HLE)

p1 411.19 (411.18) 1605Tyr-Val1607 p1* 1040.52 (1040.52) 1596Val-Asp1604

p2 1040.57 (1040.52) 1596Val-Asp1604 p2* 1449.70 (1449.69) 1596Asp-Val1607

p3 1494.83 (1494.78) 1629Gly-Ile1642 p3* 1494.83 (1494.78) 1629Gly-Ile1642

p4 753.43 (753.38) 1643Gly-Ile1649 p4* 753.40 (753.38) 1643Gly-Ile1649

p5 1333.84 (1333.80) 1617Lys-Ile1628 p5* 1333.82 (1333.80) 1617Lys-Ile1628

p6 2097.16 (2097.08) 1652Gln-Ala1669 p6* 2097.07 (2097.08) 1652Gln-Ala1669

p7 1433.73 (1433.70) 1596Asp-Val1607 p7* 2218.20 (2218.19) 1608Thr-Ile1628

p8 2218.30 (2218.19) 1608Thr-Ile1628 p8* 2323.28 (2323.25) 1650Leu-Ala1669

p9 2323.37 (2323.25) 1650Leu-Ala1669 p9* 2230.17 (2230.15) 1627Pro-Ala1647

p10 2230.24 (2230.15) 1627Pro-Ala1647 p10* 979.61 (979.60) 1643Gly-Ile1651

p11 979.61 (979.60) 1643Gly-Ile1651 a The peptide materials eluted in correspondence of the chromatographic peaks in Fig. 1 were collected, liophilyzed, and analyzed by ESI-TOF MS. b The values in parenthesis refer to the theoretical mass values deduced from the amino acid sequence of VWF74 or VWF74-MetSO (*). c Grey labels indicate the minor of two components co-eluting in the RP-HPLC chromatograms reported in Fig. 1.


146

Cleavage of VWF multimers by LSPs and activated PMNs. Preliminary experiments showed that,

at variance with ADAMTS-13, LSPs do not need a stretched conformational state of VWF to

hydrolyze this substrate. Even under static conditions and in the absence of ristocetin, which

stabilizes the active conformation of VWF (30), the latter was indeed hydrolyzed by HLE, PR3, and

CG (data not shown). However, in the presence of ristocetin the cleavage velocity was enhanced

(see Fig. 2A), likely due to the stabilization of a VWF conformation more accessible to proteolytic

enzymes. This finding is in qualitative agreement with recent results showing that even though

under static conditions a slow cleavage was observed, high shear stress accelerates the hydrolysis of

native VWF by LSPs (8). As shown in Fig. 2A, the oxidative modifications induced on plasma

VWF by MOPSY, representative of a typical leukocyte oxidative function, resulted in a more rapid

and efficient cleavage of VWF by LSPs. A similar effect was observed when PMA-activated PMNs

were used as a source of both ROS and proteases that extensively proteolyzed plasma-derived VWF

multimers after 45-min incubation (Fig. 2B). Production of superoxide anion (O2·-) by PMA-

activated PMNs was monitored by a cytofluorimetric assay, as well as by using horse cytochrome c

as a superoxide tracer (Fig. 3). Our results indicate that after PMA addition, PMNs undergo a rapid

oxidative burst, as demonstrated by flow cytometry showing that in the PMN cells the generation of

both H2O2 and superoxide anions began rapidly after 5 min. In particular, ROS secretion by PMNs

reached a plateau concentration of O2·- of about 30 µM after approximately 20-min stimulation

(Fig. 3B). Notably, extracellular scavenging of ROS by a mixture of superoxide dismutase (SOD),

catalase, and L-methionine caused a net decrease of VWF cleavage by activated PMNs. Likewise,

the presence of either serine protease inhibitors (i.e., aprotinin and PMSF) alone or in combination

with ROS scavengers completely inhibited the cleavage of VWF by LSPs. Additional experiments

aimed at assessing the release and activity of HLE by activated PMNs showed that active enzyme

was rapidly secreted by 2x103/µl cells, reaching a high concentration of active enzyme of about 350

nM after 5 min. The high levels of active HLE are sufficient to cause a significant cleavage of

VWF, as experimentally demonstrated. After the plateau was reached at 5 min, a slight but

progressive decrease of HLE activity was measured up to 30 min, although the active enzyme was

always present at concentrations >100 nM (Fig. 3C). Notably, ELISA data showed that no HLE-α1-

Protease Inhibitor complex was found in the same supernatants of activated PMNs (data not

shown). Thus, no natural inhibitor secreted by PMNs was responsible for the decrease of HLE

activity. Instead, the progressive inhibition of HLE could be likely due to liberation of ROS, which

are known to react with aminoacid residues engaged in the catalytic machinery of serine proteases

(19,31).


147

Figure 2. A) Multimeric structure of intact and oxidized (ox) VWF treated with ADAMTS-13 (A13), elastase (HLE), proteinase 3 (PR3) and cathepsin G (CG). The in vitro oxidation of VWF was obtained by treating for 15 min with 40 µM HClO, as detailed under Methods. For each enzyme, the multimer VWF pattern obtained after 2 hr digestion in the presence of 1.5 mg/ml ristocetin is shown. The electrophoresis was performed using 1.5 % SDS-agarose, followed by immunoblotting using polyclonal anti-VWF antibodies. This gel is representative of three different experiments which yielded similar results. B) Multimeric structure of purified VWF (20 µg/ml) tre ated for 45 min with 1x106/ml PMNs in the absence and presence of PMA (50 ng/ml), 1.5 mg/ml ristocetin, ROS scavengers and serine protease inhibitors. Proteolysis reaction were conducted as detailed in the Methods. SOD (6 U/ml), catalase (100 nM) and free methionine (1 mM) were used as ROS scavangers, while PMSF and aprotinin (500 µµµµM) as serine protease inhibitors . Electrophoretic run was carried out using 0.8 % (stacking gel)-1.5 % (running gel) agarose gel. ADAMTS-13 (A13), elastase (HLE), Proteinase 3 (Pr3) and cathepsin G (CG). The results shown in this gel is representative of five different experiments which yielded similar results.

A

B

PMA - + + + +

SOD + Catalase + Met - - + - +

Ser. Prot. Inhibitors - - - + +

A

B

T0T0 ox A13

A13ox

HLEoxHLE PR3

PR3 ox CG

CG oxT0

T0 ox A13

A13ox

HLEoxHLE PR3

PR3 ox CG

CG ox


148

Figure 3. Evaluation of oxidative metabolism in PMNs. A) Flow cytometry analysis of PMN cells loaded with DCFH-DA and HE, to monitor H2O2 and superoxide anion production, respectively, prior to their activation with 50 ng/ml PMA. Samples were acquired on a FACS Canto® flow cytometer at time 0 and then every 5 minutes from the addition of PMA for up to 25 minutes (from left to right). The histograms are representative of 3 independent experiments performed with PMN preparations from different consented volunteers. B) Measurement of superoxide release from PMNs. The time-dependent release of superoxide from PMNs with (□) or without (●) the addition of 6 U/ml superoxide dismutase, 200 nM catalase and 1 mM methionine is shown. The arrow indicates the time point when VWF was separated from PMNs for the analysis of its multimer pattern. Typical results from 3 independent experiments are shown.C) HLE releaseD from PMA-stimulated PMNs. The concentration of active enzyme was measured by a spectrophotometric determination of the hydrolysis rate of 200 µM N-(Methoxysuccinyl)-Ala-Ala-Pro-Val 4-nitroanilide at 405 nm as detailed under the Methods section. Determination of the Michaelis-Menten parameters for the hydrolysis of VWF74 and VWF74-

MetSO by LSPs. In proteolysis experiments with ADAMTS-13, VWF74 peptide was hydrolyzed

with the best-fit Michaelis-Menten parameter values kcat=0.81±0.02 sec-1 and Km= 6.55±0.3 µM, in

agreement with our previous data (10). On the contrary, no kinetic parameter could be calculated for

the VWF74 peptide containing MetSO under the same experimental conditions over a concentration

range from 2.5 to 40 µM, as previously reported by us and others (9,10). LSPs cleaved VWF74

efficiently, with kcat/Km values even higher than that of ADAMTS-13 (Fig. 4 and Table 2). In the

case of HLE, due to proteolysis at multiple sites (Fig. 1D,H), only a pseudo-first order rate

A

A

A

A

16001400120010008006004002000

36

34

32

30

28

26

24

22

20

18

Time (sec)

Sup

erox

ide

rele

ase

from

PM

Ns

[µM

]

B

16001400120010008006004002000

36

34

32

30

28

26

24

22

20

18

Time (sec)

Sup

erox

ide

rele

ase

from

PM

Ns

[µM

]

B

16001400120010008006004002000

36

34

32

30

28

26

24

22

20

18

Time (sec)

Sup

erox

ide

rele

ase

from

PM

Ns

[µM

]

B

16001400120010008006004002000

36

34

32

30

28

26

24

22

20

18

Time (sec)

Sup

erox

ide

rele

ase

from

PM

Ns

[µM

]

B

0 5 10 15 20 25 30 350

100

200

300

400

Time (min)

HLE

rele

ase

dfr

om

P

MN

s (n

M)

C

0 5 10 15 20 25 30 350

100

200

300

400

Time (min)

HLE

rele

ase

dfr

om

P

MN

s (n

M)

C

0 5 10 15 20 25 30 350

100

200

300

400

Time (min)

HLE

rele

ase

dfr

om

P

MN

s (n

M)

C

A

A

A

A

16001400120010008006004002000

36

34

32

30

28

26

24

22

20

18

Time (sec)

Sup

erox

ide

rele

ase

from

PM

Ns

[µM

]

B

16001400120010008006004002000

36

34

32

30

28

26

24

22

20

18

Time (sec)

Sup

erox

ide

rele

ase

from

PM

Ns

[µM

]

B

16001400120010008006004002000

36

34

32

30

28

26

24

22

20

18

Time (sec)

Sup

erox

ide

rele

ase

from

PM

Ns

[µM

]

B

16001400120010008006004002000

36

34

32

30

28

26

24

22

20

18

Time (sec)

Sup

erox

ide

rele

ase

from

PM

Ns

[µM

]

B

0 5 10 15 20 25 30 350

100

200

300

400

Time (min)

HLE

rele

ase

dfr

om

P

MN

s (n

M)

C

0 5 10 15 20 25 30 350

100

200

300

400

Time (min)

HLE

rele

ase

dfr

om

P

MN

s (n

M)

C

0 5 10 15 20 25 30 350

100

200

300

400

Time (min)

HLE

rele

ase

dfr

om

P

MN

s (n

M)

C


149

constant, kobs, of VWF74 disappearance at 25 °C could be detrmined as 0.125±0.01 min-1. At

variance with what previously observed with ADAMTS-13 (9-10), oxidation of Met1606 did not

inhibit the hydrolysis of VWF74 by LSPs. Instead, the specific oxidation of Met1606 slightly

increased the kcat/Km value of CG, mainly due to a slight increase of the kcat value, whereas the

value of kcat/Km and kobs for PR3 and HLE, respectively, remained essentially unchanged within the

experimental error, if compared to that of VWF74 (Table 2). These findings were in qualitative

agreements with those obtained with oxidized VWF multimers, as described above

Figure 4. Determination of the Michaelis-Menten parameters of VWF74 (A) and VWF74-MetSO (B) hydrolysis by ADAMTS13 (▲), cathepsin G (○) and proteinase 3 (●). The steady state kinetic parameters were calculated by measuring with RP-HPLC the concentration of the N-terminal peptide released in the proteolysis reaction. The continuous lines were drawn according to the best-fit Michaelis parameters listed in Table 2.

0 5 10 15 200.0

0.2

0.4

0.6

0.8

1.0

VWF74-MetSO (µM)

Initi

al r

ate

(sec

-1)

0 5 10 15 200.0

0.2

0.4

0.6

0.8

VWF74 (µM)

Initi

al r

ate

(sec

-1)

A

B

0 5 10 15 200.0

0.2

0.4

0.6

0.8

1.0

VWF74-MetSO (µM)

Initi

al r

ate

(sec

-1)

0 5 10 15 200.0

0.2

0.4

0.6

0.8

VWF74 (µM)

Initi

al r

ate

(sec

-1)

A

B


150

Table 2. Michaelis-Menten parameters and pseudo-first order rate constant, kobs, of VWF74 and VWF74-MetSO hydrolysis by LSPs and ADAMTS-13

VWF74 Enzyme kcat (sec-1) Km (µM) k cat/Km(M -1 sec-1)

ADAMTS-13 0.81 ± 0.02 6.55± 0.32 1.24x105 HLE kobs (min-1)= 0.125 ±0.01 CG 0.93± 0.03 5.88±0.44 1.58x105 PR3 0.88±0.04 5.18±0.64 1.7 x105

VWF74-MetSO Enzyme kcat (sec-1) Km (µM) kcat/Km(M -1 sec-1)

ADAMTS-13 n.d. n.d. n.d. HLE kobs (min-1)= 0.128±0.02 CG 1.21± 0.06 5.76±0.87 2.10x105 PR3 1.22± 0.06 7.84±0.87 1.56x105

Spectroscopic properties of native and oxidized VWF multimers. The UV-absorption spectrum of

native VWF showed a λmax value centred at 278 nm and a shoulder at 290, typical of Trp-

contribution. Increasing the concentration of hypochlorous acid (HClO) from 0 to 41 µM, the λmax

value underwent a small (i.e., 2 nm) blue-shift, whereas the absorbance ratio 280/255 nm

progressively decreased by ~35% of the original value (Fig. 5A). These spectral changes are

compatible with the formation of mono- (Oia) and di-oxyindolyl-alanine (Dia), generated by

oxidation of Trp-residues and maximally absorbing at 250 nm (22). Likewise, the fluorescence

spectra of VWF, recorded after treatment with HClO (0-40 µM) showed a 5-nm blue-shifted λmax

value, from 345 to 340 nm, and a 45% decrease of fluorescence intensity (Fig. 5B). The formation

of poorly emitting oxyndole-derivatives (i.e., Oia and Dia) explains the observed fluorescence

spectral changes (32). The far UV-CD spectrum of native VWF is typical of a protein containing a

mixed α/β secondary structure (33), with two negative bands at 208 and 215 nm (Fig. 5C). After

treating VWF with 41 µM HClO, the shape of the spectrum was essentially unchanged, whereas

ellipticity decreased by ~33%. This finding would suggest a lower secondary structure content, with

a partial unfolding of VWF-Ox. However, oxidation of Trp-residues to Oia- and Dia-derivatives is

expected to decrease the CD signal of VWF-Ox even in the absence of large conformational

changes (33-35). This conclusion is also supported by DLS measurements, showing that increasing

HClO concentrations did not alter the hydrodynamic radius of VWF, which remained essentially

constant (r = 100±39 nm, Fig. 5D), thus excluding gross alterations in the conformation or

aggregation state of VWF. This conclusion is also consistent with the absence of intermolecular


151

dityrosine cross-links, undetected by fluorescence measurements (not shown), and with

fluorescence denaturation experiments showing that both native and oxidized VWF display a

similar stability to Gdn-HCl induced unfolding (not shown).

Figure 5. Conformational characterization of intact and HClO-oxidized VWF. (A) UV-absorption spectra of purified plasma VWF multimers (100 µg/ml) treated for 15 min at 25°C with increasing concentrations of HClO (from top to bottom: 0, 6, 12, 29, and 41 µM). (B) Fluorescence spectra (λex=295 nm) of VWF (200 nM) at increasing HClO concentrations (from top to bottom: 0, 2.5, 5, 10, 20 and 40 µM). (C) Far-UV CD spectra of native (○) and oxidized (●) VWF (0.14 mg/ml). Oxidation was carried out with 41 µM HClO for 15 min at 25 °C. (Inset) Far-UV CD spectra of the model compounds NATA (○) and Oia/Dia (●) recorded at 310 µM concentration. (D) DLS spectra of VWF multimers (200 nM) at increasing HClO concentrations (from top to bottom: 40, 2, 0, 4, 20 and 10 µM).

C

1 10 100 1000

0

2

4

6

8

10

12

Inte

nsity

Dis

trib

utio

n (%

)

Diameter (nm) (Log Scale)

C

1 10 100 1000

0

2

4

6

8

10

12

Inte

nsity

Dis

trib

utio

n (%

)

Diameter (nm) (Log Scale)

260 280 300 320 3400.00

0.02

0.04

0.06

0.08

0.10

0.12

Re

lativ

e A

bso

rban

ce

Wavelength (nm)

0 10 20 30 40

0.09

0.10

0.11

0.12

Inte

nsity

at 2

77 n

m

[HClO] (µµµµM)

A260 280 300 320 340

0.00

0.02

0.04

0.06

0.08

0.10

0.12

Re

lativ

e A

bso

rban

ce

Wavelength (nm)

0 10 20 30 40

0.09

0.10

0.11

0.12

Inte

nsity

at 2

77 n

m

[HClO] (µµµµM)

A

210 220 230 240 250 260

-10

-8

-6

-4

-2

0

[θ

] x 1

0-3 (

deg

· cm

2 · dm

ol-1

)

Wavelength (nm)

C

200 220 240 260

-4

0

4

8

12

CD

(m

deg)

Wavelength (nm)

210 220 230 240 250 260

-10

-8

-6

-4

-2

0

[θ

] x 1

0-3 (

deg

· cm

2 · dm

ol-1

)

Wavelength (nm)

C

200 220 240 260

-4

0

4

8

12

CD

(m

deg)

Wavelength (nm)

320 340 360 380 400

0

20

40

60

80

100

120

Flu

ores

cenc

e In

tens

ity

Wavelength (nm)

0 µµµµM

5 µµµµM

2.5 µµµµM

20 µµµµM10 µµµµM

40 µµµµM

320 340 360 380 400

0

20

40

60

80

100

120

Flu

ores

cenc

e In

tens

ity

Wavelength (nm)

0 µµµµM

5 µµµµM

2.5 µµµµM

20 µµµµM10 µµµµM

40 µµµµM


152

DISCUSSION

ADAMTS-13 has been considered in the recent past the only enzyme responsible for the

proteolytic processing of VWF (8,36). However, recent studies showed that LSP efficiently cleave

VWF multimers near or even at the same peptide bond hydrolyzed by ADAMTS-13 (8). The results

of mass spectrometry analysis agree with those recently reported (8), showing that ADAMTS-13

and CG cleave rFRETS-VWF73 at Met1605-Tyr1606, whereas both HLE and PR3 hydrolyze the

same substrate selectively at Val1607-Thr1608. Strikingly, our results also show for the first time

that, at variance with ADAMTS-13, the selective oxidation of Met1606 to MetSO does not affect or

even slightly enhances VWF proteolysis by LSPs and this was ascertained with both synthetic

VWF74-MetSO peptide (Fig. 1) and full-length VWF oxidized with HClO or leukocyte MPO,

either by using purified proteases or activated PMN cells (Fig. 3) as source of LSPs and ROS.

PMNs stimulation by endogenous or exogenous agents results in an increase of their oxygen

consumption (respiratory burst) with a production of ROS, starting with the generation of

superoxide anion through the activation of NADPH oxidase and its product hydrogen peroxide

(H2O2) (37,38). Neutrophils produce also peroxynitrite from •O2- by reaction with nitric oxide

(NO•) and hydroxyl radical (OH•) from H2O2 in the presence of transition metals (39,40).

Activated PMNs release from their cytoplasmic granules proteolytic enzymes as well as

myeloperoxidase (MPO), which in the presence of chloride is responsible for the generation of

hypochlorous acid (HOCl), one of the most powerful oxidant molecules in vivo. As shown in Fig. 3,

stimulation of PMNs with PMA (50 ng/ml), a receptor-independent exogenous leukocyte activator

(2), leads to the relatively rapid generation of intracellular ROS, reaching a plateau concentration

>30 µM after about 20 min under physiological conditions (Fig. 3A). ROS are able to pass the cell

membrane and oxidize external macromolecules, as demonstrated by the cytochrome c assay (Fig.

3B). The degranulation reaction occurs after the oxidative burst and ROS generation, leading to

release of serine- and zinc-proteases. The results shown in Fig. 2B (a me pare sia figura 2 e non 3)

unequivocally show that the proteolytic processing of VWF by LSPs not only is not inhibited by the

presence of ROS but is even significantly accelerated by the action of ROS released by activated

PMNs.

Overall, the results herein reported, while further confirming and extending previous studies (8-

10), are unprecedented and demonstrate that oxidative modifications of a single amino acid in VWF

produce opposite effects on the interaction with the different enzymes (i.e., ADAMTS-13 and

LSPs) involved in its proteolytic processing. Notably, at variance with what recently described for

the effect of oxidation on VWF cleavage by ADAMTS-13 (9,10), the formation of MetSO at

position 1606 of the A2 domain does even favor the proteolysis of the Tyr1605-Met1606 peptide


153

bond by leukocyte LSPs, especially CG, thus supporting the hypothesis that the proteolytic

processing of VWF in vivo does not rely on ADAMTS-13 proteolysis alone, but arises also from the

activity of various serine proteases secreted by activated leukocytes (8).

Based on these findings, the question arises as to whether the positive and negative effect of

ROS on VWF cleavage by PMN serine proteases and ADAMTS-13, respectively, was caused by a

change in the physical-chemical properties of the sensitive amino acids (e.g., side-chain volume,

polarizability, and hydrophobicity of Met, Tyr, or Trp residues) that might directly be engaged in

the protease binding sites, thus favoring/disfavoring proteolysis, or oxidation affected proteolytic

efficiency indirectly by inducing a global conformational alteration of VWF multimers.

The comparative analysis of UV-absorption, fluorescence and CD spectra of native and

oxidized VWF (Fig. 5 A,B,C) displayed significant differences, suggesting some perturbation in the

secondary/tertiary structure of VWF upon HClO-mediated oxidation, with a partial unfolding of

VWF-Ox, which therefore would become more susceptible to proteolysis (41). However, the results

of DLS (Fig. 5D) and guanidinium-induced denaturation (Fig. S1) measurements indicate that upon

oxidation VWF does not undergo any gross distortion of the tertiary conformation nor

polymerization state, and agree with the results obtained with VWF74 and VWF74-MetSO, which

are largely unfolded (10) thus excluding any conformational effect of Met oxidation. Hence, we

conclude that the observed differences in the proteolysis of VWF are predominantly due to the

different chemical/structural properties of MetSO compared with those of unmodified Met at the

cleavable bond. However, more discrete local conformational changes in the A2 domain, possibly

undetected by the spectroscopic techniques used in this study, cannot be ruled out.

In the case of ADAMTS-13, our recent model building and docking studies suggest that

replacing an apolar amino acid like Met with a highly polar residue like MetSO strongly hampers

productive binding of the cleavable Tyr-Met bond to the hydrophobic S1′ site of the protease (10,

42). Opposite, albeit conceptually similar, considerations hold for explaining the effect of Met

oxidation on substrate hydrolysis by LSPs. LSPs are extremely basic enzymes, with pI values

ranging from 9.5 of PR3 to 12 of CG. (5). Furthermore, the crystallographic structure of CG

(1kyn.pdb) is characterized by the presence of a strong positive electrostatic potential in the region

surrounding the S1′ site, thus favouring binding of negatively charged substrates at P1'. These

structural features are consistent with the enhanced catalytic efficency of CG for VWF74-MetSO

(Table 2). At variance, the S1′ site of PR3 (1fuj.pdb) (43, 44) and especially of HLE (2rg3.pdb) (45)

is less electropositive than that of CG and is principally formed by apolar/neutral amino acids.

Accordingly, Met oxidation does not appreciably affect proteolysis by PR3 or HLE (Table 2).


154

May these findings have patho-physiological implications? Any activation of PMNs in

inflammatory, infectious or onco-hematological settings may influence the properties and functions

of VWF. Under physiological conditions, a discrete and “tonic” activation of leukocytes may

contribute to proteolytic processing of circulating VWF, because HLE, PR3 and CG are able to

cleave VWF with a catalytic specificity similar or even higher than that of ADAMTS-13. Notably,

De Meyer et al have recently detected satellite bands in plasma from TTP-patients and

ADAMTS13-/- mice by using VWF multimer analysis (46). As these bands are commonly

interpreted as cleavage products, it is reasonable to assume that they are generated in the above

settings by LSPs. Moreover, massive and uncontrolled release of LSPs and ROS in promyelocytic

leukemia causes extensive proteolytic degradation of VWF resulting in a severe hemorrhagic setting

(47) . Finally, it is known that VWF multimers play also extra-haemostatic functions. In particular,

it is engaged by bacterial pathogens to adhere to vascular vessel and transmigrate into tissues.

Staphylococcal aureus protein A (Spa) binds with high affinity to soluble and immobilized VWF

and has been identified as a novel staphylococcal adhesin, especially for the high molecular weight

VWF multimers (16). The process of bacterial adhesion causes migration and activation of PMNs

that secrete enzymes and ROS to eliminate the pathogens. Thus, the LSPs-mediated VWF

proteolysis may be considered a mechanism aimed at limiting the VWF-mediated adhesion and

invasion of such common bacteria in a highly oxidant milieu, where ADAMTS-13 could not

efficiently proteolyze VWF. Our results are compatible with the existence of two different pathways

of coupled oxidative/proteolytic reactions: 1) chronic oxidative stress by peroxynitrite, mainly

arising from endothelial cells in chronic inflammatory settings, diabetes and in ageing (48), reduces

cleavage by ADAMTS-13, with a resulting pro-thrombotic effect (9,10); 2) a rapid and

simultaneous release of ROS and serine proteases from activated leukocytes attracted by bacterial

pathogens results in a positive effect on VWF proteolysis, under conditions where high

concentrations of ROS could impede the activity of ADAMTS-13. The latter effect might assist the

anti-bacterial activity of PMNs, impeding an efficient VWF-mediated adhesion of bacterial

pathogens to vascular wall and their tissue invasion (16). Finally, further studies are needed at

assessing whether or not perturbations of the proteolytic processing of VWF by LSPs could be also

involved in the pathogenesis of “atypical” thrombotic microangiopathies presenting with normal

ADAMTS-13 levels (49).


155

REFERENCES

1. Nathan C (2006) Neutrophils and immunity: challenges and opportunities. Nat Rev Immunol.

6:173-182.

2. Sheppard FR, Kelher MR, Moore EE, McLaughlin NJ, Banerjee A, Silliman CC (2005)

Structural organization of the neutrophil NADPH oxidase: phosphorylation and translocation

during priming and activation. J Leukoc Biol. 78:1025-1042.

3. Dahlgren C, Karlsson A (1999) Respiratory burst in human neutrophils. J Immunol Methods.

232:3-14.

4. Pham CT (2006) Neutrophil serine proteases: specific regulators of inflammation. Nat Rev

Immunol. 6:541-550.

5. Korkmaz B, Moreau T, Gauthier F (2008) Neutrophil elastase, proteinase 3 and cathepsin G:

physicochemical properties, activity and physiopathological functions. Biochimie. 90:227-242.

6. Piccardoni P, Sideri R, Manarini S, Piccoli A, Martelli N, De Gaetano G, Cerletti C, Evangelista

V (2001) Platelet/polymorphonuclear leukocyte adhesion: a new role for SRC kinases in Mac-1

adhesive function triggered by P-selectin. Blood. 98:108-116.

7. Zarbock A, Polanowska-Grabowska RK, Ley K (2007) Platelet-neutrophil-interactions: linking

hemostasis and inflammation. Blood Rev. 21:99-111.

8. Raife TJ, Cao W, Atkinson BS, Bedell B, Montgomery RR, Lentz SR, Johnson GF, Zheng XL

(2009) Leukocyte proteases cleave von Willebrand factor at or near the ADAMTS13 cleavage

site. Blood. 114:1666-1674.

9. Chen J, Fu X, Wang Y, Ling M, Mc Mullen B, Kulman J, Chung DW, Lopez JA (2010)

Oxidative modification of von Willebrand factor by neutrophil oxidants inhibits its cleavage by

ADAMTS13. Blood. 115:706-712.

10. Lancellotti S, De Filippis V, Pozzi N, Peyvandi F, Palla R, Rocca B, Rutella S, Pitocco D,

Mannuci PM, De Cristofro R (2010) Formation of methionine sulfoxide by peroxynitrite at

position 1606 of von Willebrand factor inhibits its cleavage by ADAMTS-13: A new

prothrombotic mechanism in diseases associated with oxidative stress. Free Radic Biol Med.

48:446-456.

11. Patti JM, Allen BL, McGavin MJ, Hook M (1994) MSCRAMM-mediated adherence of

microorganisms to host tissues. Annu Rev Microbiol. 48:585-617.

12. George NP, Wei Q, Shin PK, Konstantopoulos K, Ross JM (2006) Staphylococcus aureus

adhesion via Spa, ClfA, and SdrCDE to immobilized platelets demonstrates shear-dependent

behavior. Arterioscler Thromb Vasc Biol. 26:2394-2400.


156

13. Moreillon P, Entenza JM, Francioli P, McDevitt D, Foster TJ, François P, Vandaux P (1995)

Role of Staphylococcus aureus coagulase and clumping factor in pathogenesis of experimental

endocarditis. Infect Immun. 63:4738-4743.

14. Heilmann C, Herrmann M, Kehrel BE, Peters G (2002) Platelet-binding domains in 2

fibrinogen-binding proteins of Staphylococcus aureus identified by phage display. J Infect Dis.

186:32-39.

15. Hartford OM, Wann ER, Hook M, Foster TJ (2001) Identification of residues in the

Staphylococcus aureus fibrinogen-binding MSCRAMM clumping factor A (ClfA) that are

important for ligand binding. J Biol Chem. 276:2466-2473.

16. Hartleib J, Kohler N, Dickinson RB, Chhatwal GS, Sixma JJ, Hartford OM, Foster TJ, Peters G,

Kehrel BE, Herrmann M (2000) Protein A is the von Willebrand factor binding protein on

Staphylococcus aureus. Blood. 96:2149-2156.

17. Petri B, Broermann A, Li H, Khandoga AG, Zarbock A, Krombach F, Goerge T, Schneider SW,

Jones C, Nieswandt B, Wild MK, Vestweber D (2010) Von Willebrand factor promotes

leukocyte extravasation. Blood. 116:4712-471.

18. De Cristofaro R, Peyvandi F, Palla R, Lavoretano S, Lombardi R, Merati G, Romitelli F, Di

Stasio E, Mannucci PM (2005) Role of chloride ions in modulation of the interaction between

von Willebrand factor and ADAMTS-13. J Biol Chem. 280:23295-23302.

19. De Cristofaro R, Landolfi R (2000) Oxidation of human alpha-thrombin by the

myeloperoxidase-H2O2-chloride system: structural and functional effects. Thromb Haemost.

83:253-261.

20. Bhattacharjee S, Pennathur S, Byun J, Crowley J, Mueller D, Gischler J, Hotchkiss RS,

Heinecke JW. (2001) NADPH oxidase of neutrophils elevates o,o'-dityrosine cross-links in

proteins and urine during inflammation. Arch Biochem Biophys. 395:69-77.

21. Nelson DP, Kiesow LA (1972) Enthalpy of decomposition of hydrogen peroxide by catalase at

25 degrees C (with molar extinction coefficients of H2O2 solutions in the UV). Anal Biochem.

49:474-478.

22. Savige WE, Fontana A (1980) Oxidation of tryptophan to oxindolylalanine by dimethyl

sulfoxide-hydrochloric acid. Selective modification of tryptophan containing peptides. Int J

Pept Protein Res. 15:285-297.

23. Klock JC, Bainton DF (1976) Degranulation and abnormal bactericidal function of granulocytes

procured by reversible adhesion to nylon wool. Blood. 48:149-161.

24. Kessels GC, Krause KH, Verhoeven AJ (1993) Protein kinase C activity is not involved in N-

formylmethionyl-leucyl-phenylalanine-induced phospholipase D activation in human


157

neutrophils, but is essential for concomitant NADPH oxidase activation: studies with a

staurosporine analogue with improved selectivity for protein kinase C. Biochem J. 292(Pt

3):781-785.

25. Van GB, Slater EC (1962) The extinction coefficient of cytochrome c. Biochim Biophys Acta.

58:593-595.

26. Robinson J (1997) Oxidative metabolism. Vol. In Current Protocols in Flow Cytometry. New

York: John Wiley & Sons, Inc.

27. Pace CN, Vajdos F, Fee L, Grimsley G, Gray T (1995) How to measure and predict the molar

absorption coefficient of a protein. Protein Sci. 4:2411-2423.

28. Yan LJ (2009) Analysis of oxidative modification of proteins. Curr Protoc Protein Sci. Chapter

14:Unit 14

29. Di Stasio E, Bizzarri P, Misiti F, Pavoni E, Brancaccio A (2004) A fast and accurate procedure

to collect and analyze unfolding fluorescence signal: the case of dystroglycan domains. Biophys

Chem. 107:197-211.

30. Di Stasio E, Romitelli F, Lancellotti S, Arcovito A, Giardina B, De Cristofaro R. (2009) Kinetic

study of von Willebrand factor self-aggregation induced by ristocetin. Biophys Chem. 144:101-

107.

31. Shao B, Belaaouaj A, Verlinde CL, Fu X, Heinecke JW (2005) Methionine sulfoxide and

proteolytic cleavage contribute to the inactivation of cathepsin G by hypochlorous acid: an

oxidative mechanism for regulation of serine proteinases by myeloperoxidase. J Biol Chem.

280:29311-29321.

32. Hawkins CL, Davies MJ (1998) Hypochlorite-induced damage to proteins: formation of

nitrogen-centred radicals from lysine residues and their role in protein fragmentation. Biochem

J. 332 (Pt 3):617-625.

33. Brahms S, Brahms J (1980) Determination of protein secondary structure in solution by vacuum

ultraviolet circular dichroism. J Mol Biol. 138:149-178.

34. De Filippis V, De Dea E, Lucatello F, Frasson R (2005) Effect of Na+ binding on the

conformation, stability and molecular recognition properties of thrombin. Biochem J. 390:485-

492.

35. Pozzi N, Banzato A, Bettin S, Bison E, Pengo V, De Filippis V (2010) Chemical synthesis and

characterization of wild-type and biotinylated N-terminal domain 1-64 of beta2-glycoprotein I.

Protein Sci. 19:1065-1078.

36. Chung DW, Fujikawa K (2002) Processing of von Willebrand factor by ADAMTS-13.

Biochemistry. 41:11065-11070.


158

37. Klebanoff SJ (2005) Myeloperoxidase: friend and foe. J Leukoc Biol. 77:598-625.

38. Cheung EY, Uitte de Willige S, Vos HL, Leebeek FW, Dippel DW, Bertina RM, de Maat MP

(2008) Fibrinogen gamma' in ischemic stroke: a case-control study. Stroke. 39:1033-1035.

39. Carreras MC, Pargament GA, Catz SD, Poderoso JJ, Boveris A (1994) Kinetics of nitric oxide

and hydrogen peroxide production and formation of peroxynitrite during the respiratory burst of

human neutrophils. FEBS Lett. 341:65-68.

40. Winterbourn CC (1981) Hydroxyl radical production in body fluids. Roles of metal ions,

ascorbate and superoxide. Biochem J. 198:125-131.

41. Fontana A, Polverino de Laureto P, De Filippis V, Scaramella E, Zambonin M (1997) Probing

the partly folded states of proteins by limited proteolysis. Fold Des. 2:R17-26.

42. Pozzi N, De Cristofaro R, Lancellotti S, De Filippis V (2010) Resistance of peroxynitrite-

oxidized von Willebrand factor to proteolysis by ADAMTS-13: a novel pro-thrombotic

mechanism elucidated by targeted mass spectrometry and molecular modeling. Biochem J.

Submitted for pubblication.

43. Greco MN, Hawkins MJ, Powell ET, Almond HR Jr, Corcoran TW, de Garavilla L, Kauffman

JA, Recacha R, Chattopadhyay D, Andrade-Gordon P, Maryanoff BE (2002) Nonpeptide

inhibitors of cathepsin G: optimization of a novel beta-ketophosphonic acid lead by structure-

based drug design. J Am Chem Soc. 124:3810-3811.

44. Fujinaga M, Chernaia MM, Halenbeck R, Koths K, James MN (1996) The crystal structure of

PR3, a neutrophil serine proteinase antigen of Wegener's granulomatosis antibodies. J Mol Biol.

261:267-278.

45. Huang W, Yamamoto Y, Li Y, Dou D, Alliston KR, Hanzlik RP, Williams TD, Groutas WC

(2008) X-ray snapshot of the mechanism of inactivation of human neutrophil elastase by 1,2,5-

thiadiazolidin-3-one 1,1-dioxide derivatives. J Med Chem. 51:2003-2008.

46. De Meyer S.F., Budde U., Deckmyn HKV (2009) In vivo absence of ADAMTS13 does not

abrogate VWF size heterogeneity. Journal of Thrombosis and Haemostasis. 7:PP-TH-091, 964.

47. Federici AB, Berkowitz SD, Lattuada A, Mannucci PM (1993) Degradation of von Willebrand

factor in patients with acquired clinical conditions in which there is heightened proteolysis.

Blood. 81:720-725.

48. van der Loo B, Labugger R, Skepper JN, Bachschmid M, Kilo J, Powell JM, Palacios-Callender

M, Erusalimsky JD, Quaschning T, Malinski T, Gygi D, Ullrich V, Lüscher TF (2000)

Enhanced peroxynitrite formation is associated with vascular aging. J Exp Med. 192:1731-1744.


159

49. Vesely SK, George JN, Lammle B, Studt JD, Alberio L, El-Harake MA, Raskob GE (2003)

ADAMTS13 activity in thrombotic thrombocytopenic purpura-hemolytic uremic syndrome:

relation to presenting features and clinical outcomes in a prospective cohort of 142 patients.

Blood. 102:60-68.

APPENDICES

163

Appendix A Abbreviations and Symbols

Å Angstrom

aa amino acid

Da Dalton

DTT Dithiothreitol

EDTA Ethylene Diamino Tetracetic Acid

HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

IPTG IsoPropyl-β-D-ThioGalactopyranoside

LC Liquid Cromatography

LSPs Serine Proteases Leukocyte

MW Molecular Weight

PEG PolyEthylene Glycol

SDS Sodium Dodecyl Sulfate

SDS-PAGE SDS-PolyAcrylamide Gel Electrophoresis

ESI Electrospray Ionization

HPLC High-Pressure Liquid Chromatography

m/z Mass to charge ratio

MALDI Matrix-Assisted Laser Desorption Ionization

MetSO Methionine sulfoxide

MS Mass Spectrometry

ROS Reactive Oxygen Species

RP Reverse-Phase

TFA Trifluoroacetic acid

TOF Time-of-flight

w/v Weight/volume

UV ultraviolet

Tris Tris(hydroxymethyl)aminomethane

164

Appendix B

Amino Acids Nomenclature

Amino acid Three letter code Single letter code

Alanine Ala A

Arginine Arg R

Aspartic acid Asp D

Asparagine Asn N

Cysteine Cys C

Glycine Gly G

Glutamine Gln Q

Glutamic acid Glu E

Histidine His H

Isoleucine Ile I

Lysine Lys K

Leucine Leu L

Methionine Met M

Phenylalanine Phe F

Proline Pro P

Serine Ser S

Threonine Thr T

Tyrosine Tyr Y

Tryptophan Trp W

Valine Val V

Chimica delle Proteine e Medicina Molecolarepaduaresearch.cab.unipd.it/3378/1/PhD_Thesis.pdf ·...

Documents

Transcript of Chimica delle Proteine e Medicina Molecolarepaduaresearch.cab.unipd.it/3378/1/PhD_Thesis.pdf ·...