Gene silencing and gene expression inphytopathogenic fungi using a plant virus vectorTiziana Masciaa,b,1, Franco Nigroa, Alì Abdallaha, Massimo Ferraraa, Angelo De Stradisb, Roberto Faeddac,Peter Palukaitisd,1, and Donato Gallitellia,b

aDipartimento di Scienze del Suolo, della Pianta e degli Alimenti, Università degli Studi di Bari Aldo Moro, 70126 Bari, Italy; bIstituto di Virologia Vegetale delConsiglio Nazionale delle Ricerche, Unità Operativa di Supporto di Bari, 70126 Bari, Italy; cDipartimento di Gestione dei Sistemi Agroalimentari e Ambientali,Università degli Studi di Catania, 95123 Catania, Italy; and dDepartment of Horticultural Sciences, Seoul Women’s University, Seoul 139-774, South Korea

Edited by David C. Baulcombe, University of Cambridge, Cambridge, United Kingdom, and approved February 7, 2014 (received for review August 29, 2013)

RNA interference (RNAi) is a powerful approach for elucidatinggene functions in a variety of organisms, including phytopatho-genic fungi. In such fungi, RNAi has been induced by expressinghairpin RNAs delivered through plasmids, sequences integrated infungal or plant genomes, or by RNAi generated in planta by aplant virus infection. All these approaches have some drawbacksranging from instability of hairpin constructs in fungal cells todifficulties in preparing and handling transgenic plants to silencehomologous sequences in fungi grown on these plants. Here weshow that RNAi can be expressed in the phytopathogenic fungusColletotrichum acutatum (strain C71) by virus-induced gene silenc-ing (VIGS) without a plant intermediate, but by using the directinfection of a recombinant virus vector based on the plant virus,tobacco mosaic virus (TMV). We provide evidence that a wild-typeisolate of TMV is able to enter C71 cells grown in liquid medium,replicate, and persist therein. With a similar approach, a recombi-nant TMV vector carrying a gene for the ectopic expression of thegreen fluorescent protein (GFP) induced the stable silencing of theGFP in the C. acutatum transformant line 10 expressing GFP de-rived from C71. The TMV-based vector also enabled C. acutatum totransiently express exogenous GFP up to six subcultures and for atleast 2 mo after infection, without the need to develop transforma-tion technology. With these characteristics, we anticipate this ap-proachwill findwider application as a tool in functional genomics offilamentous fungi.

transfection | plant pathogen adaptation | host species jump

RNA interference (RNAi), RNA silencing, and gene quellingare conserved processes in animals (1), plants (2), and fila-mentous fungi (3), where they operate through diverse pathwaysbased on a set of core reactions (4, 5). Briefly, these include thesynthesis of a double-stranded RNA (dsRNA), which is recog-nized and diced into 21- to 25-long dsRNA fragments by con-served ribonucleases of the Dicer-like protein family. The smallfragments denoted small interfering RNAs (siRNA) are thenloaded onto members of the Argonaute protein family to forman RNA-induced silencing complex that uses one of the twostrands of the siRNAs to direct RNA degradation, translationalrepression, or DNA methylation of homologous target genes(4, 5). These processes, commonly referred to as RNAi, playa natural role in the regulation of endogenous gene expression,development, and maintenance of genome integrity and stabilityin reproductive cells, and in the protection against transposableelements and viral infections. Besides its biological role, RNAiwas recognized rapidly as a powerful approach for elucidatinggene functions in a variety of organisms, including phytopatho-genic fungi (5, 6). In these organisms, RNAi has been inducedmainly with siRNAs derived from constructs expressing dsRNAsor hairpin RNAs (hpRNAs) delivered through plasmids orsequences integrated in fungal or plant genomes (7). In the lattercase, the technique is referred to as host-induced gene silencing(5) because it implies that siRNA molecules generated in trans-genic plants expressing dsRNAs of fungal sequences traffic

between the host and the infecting fungus. Conversely, virus-induced gene silencing (VIGS) exploits viruses to deliver sequencehomologous to a target gene fragment and trigger RNAi (8). Itrequires the engineering of a viral genome to include fragmentsof the target host gene or transgene to be silenced, the ability ofthe recombinant virus to establish a systemic infection in the hostof interest, and the silencing of the target gene by a sequencehomology-dependent mechanism. VIGS is initiated by DNA andRNA recombinant viral vectors as soon as their replication pro-duces a dsRNA. This highly structured substrate is cleaved intosiRNA by the enzymes of the RNAi pathway and the signalamplified and spread throughout by an RNA-dependent RNApolymerase coded by the host to down-regulate the expression ofthe target gene (8). Thus, VIGS is a powerful reverse geneticstool originally developed for plants, but new protocols and viralvectors have expanded its utility also for functional genomics infungi. For example, the barley stripe mosaic virus (BSMV)-VIGSsystem has been used successfully for RNAi of specific patho-genicity genes in Puccinia triticina by siRNA generated in plantathrough infection of the BSMV vector expressing dsRNAs fromP. triticina genes (9). Besides the complexity of some models thatinclude transformation and handling of transgenic plants, thestability of these dsRNA transcripts also has been questioned infungi (10) and in viral vectors (11). When present on the samestrand of viral nucleic acid, the engineered, self-complementarysequences can enhance the effectiveness of silencing but such in-verted repeat structures are often unstable inside the viral genomes(11). Hence, we examined whether the plant virus tobacco mo-saic virus (TMV) could be used for both gene expression and

Significance

RNAi is used to elucidate gene functions also in phytopatho-genic fungi, where it is expressed by hairpin RNAs deliveredthrough plasmids, sequences integrated in fungal or plantgenomes, or by RNAi generated in planta by a plant virus in-fection. These approaches have drawbacks ranging from in-stability of hairpin constructs to difficulties in preparing andhandling transgenic plants to silence homologous sequences infungi grown on these plants. Here we show that RNAi can beexpressed in phytopathogenic fungi by direct transfection witha plant virus-based vector and that the approach also can beused to obtain foreign protein expression in fungi. This tech-nique could find wider application for functional genomics infilamentous fungi of biomedical and phytopathological interest.

Author contributions: T.M., F.N., and D.G. designed research; T.M., F.N., A.A., M.F., A.D.S.,and R.F. performed research; F.N. and M.F. contributed new reagents/analytic tools; T.M.,F.N., A.D.S., P.P., and D.G. analyzed data; and P.P. and D.G. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.1To whom correspondence may be addressed. E-mail: tiziana.mascia@uniba.it orscripath1@yahoo.co.uk.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1315668111 PNAS | March 18, 2014 | vol. 111 | no. 11 | 4291–4296

MICRO

BIOLO

GY

Dow

nloa

ded

by g

uest

on

June

1, 2

021

http://crossmark.crossref.org/dialog/?doi=10.1073/pnas.1315668111&domain=pdf&date_stamp=2014-03-06mailto:tiziana.mascia@uniba.itmailto:scripath1@yahoo.co.ukhttp://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplementalhttp://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplementalwww.pnas.org/cgi/doi/10.1073/pnas.1315668111

VIGS (8) by a direct infection in phytopathogenic fungi. It wasknown that TMV could enter and persist in the mycelia of thesoil-inhabiting phytopathogenic fungus Pythium sp. (now placedamong oomycetes) (12), although no direct evidence for viralreplication was provided. Similarly, Pythium arrehnomanes couldbe transfected with TMV and tobacco necrosis virus (13). Sincethen, the matter did not receive further attention, as the interestfor viruses of fungi focused on true mycoviruses (14, 15). In thisframework, a very recent study reported the infection of intactmycelium of Sclerotinia sclerotiorum by the geminivirus-like DNAmycovirus S. sclerotiorum hypovirulence-associated DNA virus 1,as a more efficient and easier to handle approach for fungaltransfection (16).Here we show that RNAi can be expressed in the phytopath-

ogenic fungus Colletotrichum acutatum by VIGS using a recombi-nant vector based on TMV. We first provided evidence that awild-type isolate of TMV was able to enter replicate and persist incells of C. acutatum, strain C71 (C71). Subsequently, we used arecombinant TMV vector, in which the ORF of the gene encodingthe green fluorescent protein (GFP) was transcribed in C71 cellsfrom a duplicate of the TMV coat protein (CP) subgenomic mRNApromoter and demonstrated that the approach could be used toobtain foreign protein expression in fungi. Finally we used theTMV-GFP construct to abolish transcription and endogenous ex-pression of GFP in the C. acutatum transformant line 10 expressingGFP (CATEF10) transformant line of C71.

ResultsTMV Enters and Replicates in Cells of Colletotrichum spp. A purifiedTMV suspension was added to a liquid culture of C71 and my-celia were collected and examined at 5, 10, and 20 d after in-cubation. Observations with a transmission electron microscope

(TEM) on dips of mycelia of C71 treated with sodium hypo-chlorite to eliminate virus particles adhering to hyphae externallyshowed that TMV particles were present in mycelia already after5 d of incubation and persisted therein at least for the sub-sequent 15 d (Fig. 1 A and B). Ultrathin sections of myceliacollected at 5, 10, and 20 d of incubation with TMV showedaggregates of TMV-like rods (Fig. 1A) and membrane vesiculationinside the C71 cells (Fig. 1B). These virus-like particles and vesi-cles were not seen in the hyphae not exposed to TMV (Fig. 1C),but were seen consistently in approximately 73% of an average ofthe 2,000 hyphae collected and observed at each sampling time.The presence of TMV particles in the C71 cells was confirmed byin situ immunogold labeling (IGL), as abundant IGL signals wereseen scattered in the cytoplasm at 5, 10, and 20 d of incubationwith TMV (Fig. 1D), but not in control cultures (Fig. 1E). IGL didnot indicate any preferential localization for TMV particles in thecytoplasm or cellular organelles, because of the poorly resolvedcellular ultrastructure due to the IGL treatment. Nevertheless,these results demonstrated that TMV particles entered the C71hyphae and/or germinating conidia, probably through woundscaused by either shaking or at structures or loci where conidiagerminate and initiate hyphal development, colonized the mycelia,and followed their growth up to 20 d of incubation. To estimate thepercent of germinating conidia entered by virus particles, C71 my-celia were sampled at 24 h after the addition of a purified virussuspension to the culture in liquid medium. Twenty mycelia sampleswere collected randomly, transferred individually onto a solid sub-strate, and allowed to grow further for 1 wk. Hybridization signalswith a TMV-specific probe were obtained consistently from thetotal RNA extracts prepared from each of the 20 cultures (Fig.S1D). Because it was expected that after a 24-h incubation inliquid medium no new conidia would be produced in the culture,

Fig. 1. Modifications of cellular ultrastructure and accumulation of TMV-like particles in hyphae of C71 collected at 20 d of incubation with TMV. Onlysamples collected at 20 d are shown in this figure. Electron micrograph of a hypha of C71 exposed (A and B) and not exposed (C) to TMV inoculum. In situlocalization of TMV particles by immunogold labeling (IGL) in the same sample shown in A, exposed (D) or not exposed (E) to TMV inoculum. Arrows in Apoint to massive fibrous aggregates of TMV-like rods. (B) Proliferation of the endoplasmic reticulum and dictyosomal vesicles in hyphae of C71 coincubatedwith TMV. Insets show a closer view of the rod-like aggregates in A, of membrane proliferation in B and of IGL signals in D. Fibrillar materials outside the cellwall are not formed by TMV particles because they were not targeted by IGL. They probably represent fragments of the cell wall damaged by the sodiumhypochlorite treatment as they were not seen in hyphae not exposed to sodium hypochlorite treatment (Fig. S1C). (F) TMV particles purified frommycelium ofC71 at 8 dpi with TMV. (G) TMV particles purified from N. tabacum cv. Samsun nn. n, nucleus; cw, cell wall. (Scale bars in A–D, 250 nm; F and G, Insets, 100 nm.)

4292 | www.pnas.org/cgi/doi/10.1073/pnas.1315668111 Mascia et al.

Dow

nloa

ded

by g

uest

on

June

1, 2

021

http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF1www.pnas.org/cgi/doi/10.1073/pnas.1315668111

we estimate that soon after addition to the liquid medium TMVentered 100% of either germinating conidia or wounded hyphae.Quantitative estimates of total RNA extracted from groups of

15 samples of the C71 mycelia, each group being collected at 5,10, and 20 d postinoculation (dpi) and determined by dot-blothybridization with a TMV-specific probe, showed a sixfold in-crease of TMV RNA accumulation from 5 to 20 dpi, indicatingviral replication in C71 cells (Fig. S1E). A similar pattern wasobtained also from total RNA preparations after hybridizationwith a plus-sense DIG-labeled RNA probe, which detected thereplication-specific, negative-sense strand of the viral RNA (Fig.2A). A Northern blot analysis at 5 dpi with TMV provided fur-ther evidence for viral replication in C71 cells by the detection ofthe CP subgenomic RNA (Fig. 2B). However, the overall accu-mulation of TMV RNA in C71 was much lower than in plants ofNicotiana benthamiana at 5 dpi with TMV (Fig. 2B). Using thesame approach, the analysis was extended to Colletotrichumclavatum, isolate number 398854 from International MycologicalInstitute (IMI), and Colletotrichum theobromicola, isolate F27,and in these cases, TMV also entered, persisted, and replicatedin the mycelia of both species (Fig. S1F).

TMV Does Not Alter Pathogenicity, Morphology, or Growth Rate ofC71. Mycelia samples taken at 5, 10, and 20 dpi with TMV,treated with sodium hypochlorite, crushed in phosphate bufferand rub inoculated onto Samsun nn tobacco plants inducedchlorosis in inoculated leaves by 5 dpi and systemic infection withappearance of typical TMV-induced disease symptoms consist-ing of mosaic and leaf blade deformation within 2 wk after in-oculation (Fig. S2A). The presence of TMV RNA in both locallyand systemically infected leaves of tobacco was verified by dot-blothybridization (Fig. S2D). Plants rub inoculated with a mixtureobtained by crushing mycelia of C71 not exposed to coincubationwith TMV remained free of symptoms. The virus was purified frommycelia of C71 collected at 8 dpi with TMV and washed extensivelywith distilled water before purification obtaining about 25 μg ofpartially purified virus suspension from 1 g of mycelia; i.e., a 15-foldincrease compared with the inoculum applied to the medium. TMVparticles purified from C71 mycelia (Fig. 1F) were morphologi-cally undistinguishable from those routinely obtained from plantinfection (Fig. 1G) and were infectious to Nicotiana occidentalisand N. benthamiana plants inducing severe local and systemic

symptoms (Fig. S2 B and C). These results demonstrate thatTMV particles retained their original morphology and the abilityto infect tobacco plants after entering and replicating in the cellsof C71 mycelia. Moreover, TMV RNA was detected also in con-idia harvested from TMV-infected C71 subcultured on plates atweekly intervals for 2 mo (Fig. S1G). Thus, it appears that TMVpersisted in intact conidia of C71, through which it could betransmitted to other C71 generations during subculture.For pathogenicity tests, we inoculated flowers of N. benthamiana

and wounded apple fruits of cv Golden Delicious with suspensionsof conidia of C71 wild type or C71 transfected with TMV. Both theoriginal wild-type and TMV-containing C71 cultures were patho-genic on N. benthamiana flowers and apple fruit producing typicalrot symptoms (Fig. S3 A and B). In these cases, the lesion diameterinduced by the two cultures at 12 dpi was very similar (Fig. S3C),suggesting that the presence of TMV in C71 neither enhanced nordecreased the fungal pathogenicity in apple fruit. Neither themorphology of C71 mycelia (Fig. S3D) nor the growth rate of C71cultures on four different media (Fig. S3E) was altered noticeablyby TMV infection. The involvement of endopolygalacturonase(PG) in pathogenicity has been demonstrated for several fungi,including C. acutatum (17, 18). However, TMV infection of C71wild-type mycelia had no significant effect on PG activity at 20 dpi(Fig. S3F).

Enhanced GFP Expresses Constitutively in Transformed C71. Wetransformed the isolate C71 with a new binary plasmid vector toexpress constitutively the enhanced GFP (eGFP) in Colleto-trichum acutatum (pCATefGFP). This plasmid contained a con-struct in which the expression of the eGFP was regulated by aconstitutive translation elongation factor (TEF) promoter fromAureobasidium pullulans, and the glucoamylase terminator fromAspergillus awamori. A hygromycin B (HygB) resistance gene wasused as a selectable marker. An average of 150 C71 transfor-mants per 105 conidia/mL were obtained for the pCATefGFPconstruct and one of the transformants, denoted CATEF10, wasselected for this study. The expression of eGFP was confirmed byepifluorescence microscopy analysis (Fig. 3 A and B), whereasthe results of sequencing and restriction endonuclease double-digestion analysis revealed the integration of the completeTEFefGFP cassette in the CATEF10 genome, as a single in-sertion (Fig. S4A). The mitotic stability of the integrated trans-gene was confirmed by more than 20 subcultures in liquid andsolid media without HygB, and storage at −80 °C in water con-taining 25% (vol/vol) glycerol. Pathogenicity tests conducted onapple fruits did not show significant differences between theCATEF10 line and the C71 wild-type isolate (Fig. S3C), al-though the morphology of the respective mycelia was clearlydifferent (Fig. S3D). The transgenic expression of eGFP in theCATEF10 isolate induced also a ∼1.5-fold reduction in both thegrowth rate and the PG activity, if compared with C71 wild-typeisolate (Fig. S3 E and F). When viewed under TEM, cells ex-pressing GFP either transgenically or ectopically showed elec-tron dense protein aggregates (Fig. S5 B and D), which were notseen in C71 cells with or without TMV infection (Fig. 1) or incells of CATEF10 with a silenced eGFP phenotype (Fig. S5E).

The TMV-GFP Vector Down-Regulates eGFP Expression in CATEF10.The inoculum for examining foreign gene expression and VIGSin the mycelia was a recombinant, gene expression/VIGS vector,denoted TMV-GFP, which had been established in N. occidentalisplants inoculated with RNA transcripts synthesized from the bi-ologically active plasmid pBSG1057-5. Sap from systemically in-fected leaves of N. occidentalis emitting green fluorescence (Fig.S4C) was added to the liquid medium during CATEF10 growth.The C71 wild-type isolate challenged with the same sap was usedas a control. Mycelia samples collected at 5, 10, and 20 dpi fromCATEF10 cultures challenged with the TMV-GFP vector andobserved under an epifluorescence microscope showed drasticallyreduced green fluorescence in hyphae and conidia, starting from5 dpi and up to 20 dpi (Fig. 3 C and D). This result occurred in

Fig. 2. TMV enters and replicates in cells of C. acutatum. (A) Determinationof the relative quantity (RQ) of TMV plus-sense RNA or minus-sense RNA inpreparations of total RNA extracted from mycelia samples of C71 collectedat 5, 10, and 20 d of incubation with TMV estimated by dot-blot hybrid-ization with DIG-labeled minus- and plus-strand RNA probes, respectively,against TMV RNA. Columns represent the values of intensity of each spotquantified on the basis of a standard curve obtained by fivefold dilutionseries of a plasmid preparation at the initial concentration of 50 ng, con-taining the same target sequence recognized by the probe. (B) Abundanceof TMV RNA in 2 μg total RNA preparations extracted from a N. benthamianaplant (T) and a C71 culture (C) at 5 dpi with TMV by Northern blot hybridizationwith a DIG-labeled DNA probe. Arrows indicate the positions of TMV genomic(gRNA) and coat protein subgenomic (sgRNA) RNAs. Coat protein subgenomicRNA is not present in RNA from purified virions. H, healthy N. benthamianaplant; T, N. benthamiana infected by TMV; C, C71 infected by TMV; P, 200 ngRNA extracted from a purified preparation of TMV.

Mascia et al. PNAS | March 18, 2014 | vol. 111 | no. 11 | 4293

MICRO

BIOLO

GY

Dow

nloa

ded

by g

uest

on

June

1, 2

021

http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF2http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF2http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF2http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF4http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF5http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF5http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF4http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF4

all three samples collected at each sampling time, in contrast toCATEF10 cultures challenged with sap from healthy N. occi-dentalis, which remained green fluorescent (Fig. 3 E and F). Toavoid any interference by siRNAs and dsRNAs resulting from virusinfection in N. occidentalis and retained in plant sap used as in-oculum, we also challenged CATEF10 cultures with a preparationof particles of TMV-GFP (Fig. S5A) purified from N. occidentalisplants showing bright fluorescence (Fig. S4C). Results from in-oculation with a purified preparation of TMV-GFP were similar tothose obtained with infected sap (Fig. S6 C and E).To demonstrate a relationship between the reduction of green

fluorescence in CATEF10 cultures challenged with TMV-GFPand RNAi, we analyzed samples of CATEF10 mycelia collectedat 5 and 10 dpi with TMV-GFP. The abundance of eGFP tran-scripts in nonsilenced and silenced CATEF10 lines was esti-mated by quantitative PCR (qPCR) of reverse-transcribed RNApreparations extracted from samples of mycelia collected fromthe CATEF10 line at 5 and 10 dpi with TMV-GFP. The resultsrevealed an almost complete down-regulation in the eGFP geneexpression (Fig. 4A). As expected, the expression of the GFPfrom the TMV-GFP vector also was highly reduced if comparedwith the abundance of transcripts observed in the C71 isolate at5 and 10 dpi with TMV-GFP (Fig. 4A). Effective silencing of theGFP expressed from the TMV-GFP vector was shown usinga primer pair specific for this sequence. Results shown in Fig. 4Bdemonstrated that the expression of ectopic GFP in CATEF10was reduced almost completely, similar to transgenic eGFP (Fig.4A). Additional evidence for RNAi was given by total RNApreparations hybridized with hydrolyzed DIG-labeled probes,

specific for the transgenic eGFP and for the CP gene of the TMV-GFP vector, which showed the presence of siRNAs of approxi-mately 21 nt specific for the eGFP gene and CP gene of TMV-GFP (Fig. 4 C and D). These siRNAs were not seen in theCATEF10 line without TMV-GFP infection, providing indirectevidence that ingress, replication, and expression of TMV-GFPtook place in cells of CATEF10 and triggered a systemic RNAisignal that followed fungal growth. Interestingly, as in plants,a siRNA population was produced also against the CP gene ofthe TMV-GFP vector (Fig. 4D).

GFP Expression and Silencing Persists in Plated Cultures and in PlantTissues. To test the stability of the CATEF10 silenced phenotype13 monoconidial cultures of CATEF10 with a silenced eGFPwere collected at 20 dpi from liquid medium and transferred atweekly intervals on solid substrate for a further 45 d. Fluores-cence was monitored at each passage, showing no reversed si-lencing effects for the six subcultures (Fig. 3 G and H). Only atthe seventh passage did 2 of the 13 monoconidial CATEF10cultures again show the fluorescence signal. To assess whetherthe TMV-GFP vector was still present and viable in eGFP-silenced CATEF10 cells, total RNA was extracted from the sixthsubculture sample and subjected to hybridization with a DNAprobe specific for the CP gene of TMV-GFP. Fig. S4B showsa very weak signal of hybridization, which parallels data fromqPCR (Fig. 4B); however, when mycelia from this sixth subcul-ture of silenced CATEF10 were crushed in sodium phosphatebuffer and rubbed onto leaves of N. benthamiana, N. occidentalis,and tobacco, the plants remained free of symptoms and TMV

Fig. 3. VIGS of GFP in CATEF10 transformant line upon ectopic expression of GFP driven by the chimeric plant virus gene expression/VIGS vector TMV-GFP.(A) Confocal microscopy showing green fluorescence of the CATEF10 transformant line constitutively expressing eGFP. (C) Epifluorescence microscopyshowing a culture of CATEF10 with uniformly silenced eGFP at 10 dpi with TMV-GFP. (E) Epifluorescence microscopy showing green fluorescence emittedfrom a culture of CATEF10 at 10 d after incubation with sap of a healthy N. occidentalis plant. Picture shows one of the three cultures analyzed. (G) Epi-fluorescence microscopy showing a culture of CATEF10 with still uniformly silenced eGFP after six subcultures on solid medium (i.e., approximately 65 dpi withTMV-GFP). (I) Confocal microscopy showing green fluorescence emitted by conidia of a culture of C71 at 10 dpi with TMV-GFP. (K) Epifluorescence microscopyshowing a culture of C71 at 10 dpi with wild-type TMV. (M) Epifluorescence microscopy showing green fluorescence still emitted by conidia of C71 infectedwith TMV-GFP after six subcultures on solid medium (i.e., approximately 65 dpi with TMV-GFP). (O) Epifluorescence microscopy showing a culture of CATEF10with still uniformly silenced eGFP, recovered after pathogenicity test in apple. A, C, E, G, I, K, M, and O and B, D, F, H, J, L, N, and P show the same samplesviewed under UV and white light, respectively. (Scale bars, 10 μm.)

4294 | www.pnas.org/cgi/doi/10.1073/pnas.1315668111 Mascia et al.

Dow

nloa

ded

by g

uest

on

June

1, 2

021

http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF5http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF4http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF6http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF4www.pnas.org/cgi/doi/10.1073/pnas.1315668111

RNA was not detected in either inoculated or newly emergingleaves up to 21 dpi. Collectively, these results demonstrated thatRNAi activated in cells of CATEF10 line targeted also TMV-GFP.Nonetheless, the eGFP knockdown in the CATEF10 transfor-mants still carried the eGFP gene insertion, because an ampliconof the expected size was obtained using the primer pair designed toamplify a 748-bp fragment from the eGFP sequence.Despite the eGFP silencing, the overall morphology and my-

celial growth of both CATEF10 and the silenced CATEF10 werevery similar, while retaining the original difference from the C71and C71 + TMV-GFP isolates (Fig. S3D). A pathogenicity teston wounded apple fruit (Fig. S3C) and olive seedlings (Fig. S7 Aand B) did not show significant differences between the eGFP-silenced and -nonsilenced CATEF10 lines. The green fluores-cence (or lack thereof) from the transgenic fungi could not bedetected in either infected apple fruit or olive seedlings due tofluorescence emitted from the necrotized infected tissues. However,isolation of the fungus from such fruit or leaves of olive seedlings 3wk postinoculation and propagation on agar plates showed that theeGFP in CATEF10 was still silenced (Fig. 3 O and P and Fig. S7E),indicating that silencing of the transgene was maintained also inliving tissues.Green fluorescent conidia and hyphae also were seen scat-

tered in preparations of the wild-type C71 isolate challengedwith the TMV-GFP vector (Fig. 3 I and J and Fig. S6 A and B),but not in C71 cultures challenged with wild-type TMV (Fig. 3 K andL). Expression of ectopic GFP in C71 cultures also was demon-strated by qPCR with a primer pair specific for this GFP sequence(Fig. 4B). The TMV-GFP–derived green fluorescence was main-tained in C71 in 10 monoconidial cultures collected at 20 dpi fromliquid medium, passaged on solid substrate six times at weekly in-tervals, and monitored at each passage (Fig. 3 M and N). At theseventh passage, 8 of the 10 isolates had lost the fluorescence signal.

DiscussionThis study both provides unique information on the ability ofa plant virus to infect and complete its replication cycle in a plantpathogenic fungus and exploits a unique VIGS protocol for fil-amentous fungi. Our results clearly show that TMV is able toenter and replicate in cells of C. acutatum, C. clavatum, andC. theobromicola, which may not be an exception, although it has

neither been found nor probably searched for in nature. C71hosting TMV replication retained overall morphology, growthrate on different solid substrates, and PG activity, which isconsidered one of the hallmarks for pathogenicity. These traitswere apparently unaffected after several subcultures and storageup to 2 mo after inoculation. Thus, TMV did not reduce thefitness of this fungus to any noticeable extent. The most prom-inent alteration was at the cell ultrastructure level, where accu-mulation of rod-like virus particles was seen in the cytoplasm.More pronounced proliferation of the endoplasmic reticulumand dictyosomal vesicles also were observed (Fig. 1B and Fig.S5F) and by analogy with the cytopathology of plants infected byTMV (19), they may represent the sites for virus replication.Despite the ultrastructural changes and accumulation of maturevirions, the overall load of TMV RNA in C71 tissues seemsrelatively poor if compared with that of infected plants (Fig. 2Band Fig. S1E). In plants TMV distributes uniformly during sys-temic infection, whereas we do not know whether and how TMVmoves through C71 mycelia. Because growth of the mycelia occursat the tips of the hyphae, we can speculate that TMV distributesfrom one cell to the other during cell division at the tips of hyphae.If this is the case, it would be expected that virus could be presentprevalently in cells developed after virus entry, which could resultin the relatively low estimates of viral RNA loads detected inthis study.The observations that we described here could be another

documented case of host species jumps for plant viruses (20)without any apparent adverse effect for the virus fitness afterexposure to a nonplant host. Replication of a few plant viruseshas been documented in yeast (21), but it is previously undoc-umented in filamentous fungi. We also showed that a chimericplant virus could serve as a vector to abolish expression ofa stably integrated and initially highly expressed transgene infungal cells by the ectopic expression of sequences with homol-ogy to the transgene. The obvious advantage of VIGS over otherRNAi approaches in plants is that from the site of infection theinoculum spreads systemically throughout the entire organism.The situation is clearly different for the model proposed in thisstudy in which viral inoculum is added to a culture of conidiaphysically dispersed in the medium. However, following sap or pu-rified virus inoculation of the liquid medium, TMV-GFP probably

Fig. 4. Reduction of GFP transcripts in CATEF10 transformant line at 5 and 10 dpi with TMV-GFP. (A) Abundance of eGFP transcripts in samples of CATEF10transformant line at 10 dpi culture in liquid medium (CATEF10) challenged with sap from healthy N. occidentalis compared with those detected in the sameline at 5 and 10 dpi with sap of N. occidentalis infected by TMV-GFP showing nearly 100% reduction in the expression level of both transgenic eGFP andectopic GFP carried by the TMV-GFP vector, as determined by qPCR. C71, wild-type culture of C71 challenged with sap from healthy N. occidentalis. C71 +TMV-GFP, accumulation of both the GFP subgenomic RNA transcribed from TMV-GFP vector and the TMV-GFP vector itself in wild-type C71 at 5 and 10 dpiwith sap of N. occidentalis infected by TMV-GFP. (B) Amplification of GFP expressed ectopically in C71 and CATEF10 cells upon infection with TMV-GFP vectorby using selective primer pairs annealing to the coat protein (CP) subgenomic promoter (ProFor) placed upstream the GFP sequence and to a GFP fragment(GFPRev2) placed 137 nt downstream of the ProFor primer in the sequence of TMV-GFP. The reduction of TMV-GFP transcripts in CATEF10 +TMV-GFP at 5 and10 dpi provides additional evidence for RNAi of viral GFP. (C and D) Detection of siRNAs produced against GFP (C) and CP of TMV-GFP (D) in CATEF10 (CA10) at5 and 10 dpi with TMV-GFP (+TGFP) in preparations of total RNA extracted from mycelia were hybridized with hydrolyzed probes specific for GFP (C) and CPof TMV-GFP (D). M, size markers made by 19, 21, and 25 DNA oligonucleotides.

Mascia et al. PNAS | March 18, 2014 | vol. 111 | no. 11 | 4295

MICRO

BIOLO

GY

Dow

nloa

ded

by g

uest

on

June

1, 2

021

http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF7http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF7http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF7http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF6http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF5http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF5http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF1

entered nearly 100% of CATEF10 cells as indirectly shown from thecontrol C71 (Fig. S1D), so that the whole culture exhibited a fulleGFP silencing phenotype, which was maintained still after sixsubcultures. Interestingly, no infectivity of TMV-GFP was detectedin plants back inoculated with crushed mycelia of such subcultures,providing direct evidence that the silencing signal was amplified,probably in each recipient cell, during fungal growth and targetedboth the eGFP transcript and the TMV-GFP RNA. This, in turn,implies also that the suppressor of RNA silencing encoded byTMV would be ineffective in counteracting RNAi in fungal cells.The observation that 8 of 10 monoconidial isolates of C71 lost thefluorescence expressed by TMV-GFP between the sixth and theseventh subculture is consistent with this hypothesis. We alsoprovided preliminary evidence that electron-dense aggregatesaccumulated in cells of C71 upon infection with TMV-GFP (Fig.S5 B and C) and that similar accumulation was also detected inCATEF10 cells (Fig. S5D), but not in those with an eGFP silencedphenotype (Fig. S5E).Compared with conventional gene knockout strategies in

fungi, RNAi has several advantages and fewer drawbacks (22).One major advantage is that RNAi can be induced transiently,overcoming the need of permanent deletion of specific genes,which may be lethal for the organism. In filamentous fungi, ef-ficient gene silencing has been achieved by different approaches(7, 9, 10) compared with which the VIGS strategy used in thisstudy is more direct and easier to perform, as it obviates theplant-mediated transferral of the silencing construct and is in-dependent of the fungus asexual reproduction pathway.Like animal and plant viruses, mycoviruses are also inducers,

targets, and suppressors of RNAi involved in the antiviral de-fense response (23). In the case of Cryphonectria parasitica, theRNAi-based response against hypovirus infection contributed tohypovirus RNA recombination. In this regard, viral genome in-stability represents an important obstacle to the use of chimericmycoviruses as gene expression vectors and the discovery ofa role for RNAi in promoting hypovirus RNA recombinationpotentially hampers the use of chimeric mycoviruses as vectors toinduce silencing in fungi (24). Also in this respect, our approachto use VIGS vectors based on plant viruses could be a more robust

tool. However, here we can anticipate that its use will need somerefined insights to be more widely adopted. For example, somefungi lack basic components of the RNAi silencing mechanism(10) and the ability of viral vectors so far developed for VIGS inplants to enter and replicate in fungal cells has to be investigatedcase by case. For example, in an earlier study, an attempt to es-tablish infection by BSMV in Gaeumannomyces graminis, Aur-eobasidium bolleyi, and Pythium ultimum failed (25). Finally, theapproach described in this study would also enable fungi totransiently or chronically express other proteins without the needto develop transformation technology for those fungi. This hasbeen shown by the TMV-GFP infection in the C71 wild-typeisolate as the conidia emitted green fluorescence and still con-tained the TMV-GFP vector after six subcultures on agar plates.This paves the way for new opportunities in the study of bothfungal physiology and interactions with their hosts, applied tohuman, animal, and plant mycology.

Materials and MethodsDetailed descriptions are provided in SI Materials and Methods.

Fungi and Virus Isolates. The C71 strain of C. acutatum and the isolates IMI398854 of C. clavatum and F27 of C. theobromicola were grown on liquidmedium and inoculated with TMV or TMV-GFP. Mycelia were collected at 5,10, and 20 dpi with TMV or TMV-GFP and treated with sodium hypochlorite.

Transmission Electron Microscopy . Mycelia of C71 or CATEF10 collected at 5-,10-, and 20-d incubations with or without TMV or TMV-GFP were used forthin sectioning. Immunogold labeling was performed with a polyclonal an-tiserum raised against TMV.

Construction of the Binary Vector and Transformation of C. acutatum C71. Theplasmid vector pCATefGFP construct containing the eGFP sequence betweenthe constitutive TEF promoter from A. pullulans and the glaA terminator fromA. awamori was electroporated to competent cells of Agrobacterium tume-faciens AGL1 strain. For the transformation, AGL1 cell lines were mixed withconidia of C71 and spread onto a sterile nitrocellulose membrane placed on topof induction medium agar plates supplemented with hygromycin B for selection.

1. Fire A, et al. (1998) Potent and specific genetic interference by double-stranded RNAin Caenorhabditis elegans. Nature 391(6669):806–811.

2. Ruiz MT, Voinnet O, Baulcombe DC (1998) Initiation and maintenance of virus-in-duced gene silencing. Plant Cell 10(6):937–946.

3. Romano N, Macino G (1992) Quelling: Transient inactivation of gene expression inNeurospora crassa by transformation with homologous sequences. Mol Microbiol6(22):3343–3353.

4. Voinnet O (2005) Non-cell autonomous RNA silencing. FEBS Lett 579(26):5858–5871.5. Nunes CC, Dean RA (2012) Host-induced gene silencing: A tool for understanding

fungal host interaction and for developing novel disease control strategies. Mol PlantPathol 13(5):519–529.

6. Salame TM, Ziv C, Hadar Y, Yarden O (2011) RNAi as a potential tool for biotechnologicalapplications in fungi. Appl Microbiol Biotechnol 89(3):501–512.

7. Tinoco MLP, Dias BBA, Dall’Astta RC, Pamphile JA, Aragão FJL (2010) In vivo trans-specific gene silencing in fungal cells by in planta expression of a double-strandedRNA. BMC Biol 8:27.

8. Senthil-Kumar M, Mysore KS (2011) New dimensions for VIGS in plant functionalgenomics. Trends Plant Sci 16(12):656–665.

9. Panwar V, McCallum B, Bakkeren G (2013) Host-induced gene silencing of wheat leafrust fungus Puccinia triticina pathogenicity genes mediated by the Barley stripe mosaicvirus. Plant Mol Biol 81(6):595–608.

10. Meyer V (2008) Genetic engineering of filamentous fungi—progress, obstacles andfuture trends. Biotechnol Adv 26(2):177–185.

11. Lacomme C, Hrubikova K, Hein I (2003) Enhancement of virus-induced gene silencingthrough viral-based production of inverted-repeats. Plant J 34(4):543–553.

12. Brants H (1969) Tobacco mosaic virus in Pythium spec. Neth J Plant Pathol 75:296–299.

13. Nienhaus F, Mack C (1974) Infection of Pythium arrehnomanes in vitro with tobaccomosaic virus and tobacco necrosis virus. Z Pflanzenkrankh 81:728–731.

14. Pearson MN, Beever RE, Boine B, Arthur K (2009) Mycoviruses of filamentous fungiand their relevance to plant pathology. Mol Plant Pathol 10(1):115–128.

15. Ghabrial SA, Suzuki N (2009) Viruses of plant pathogenic fungi. Annu Rev Phytopa-thol 47:353–384.

16. Yu X, et al. (2013) Extracellular transmission of a DNA mycovirus and its use as a naturalfungicide. Proc Natl Acad Sci USA 110(4):1452–1457.

17. D’Ovidio R, et al. (2004) Characterization of the complex locus of bean encodingpolygalacturonase-inhibiting proteins reveals subfunctionalization for defense againstfungi and insects. Plant Physiol 135(4):2424–2435.

18. Alkan N, Espeso EA, Prusky D (2013) Virulence regulation of phytopathogenic fungiby pH. Antioxid Redox Signal 19(9):1012–1025.

19. Esau K, Cronshaw J (1967) Relation of tobacco mosaic virus to the host cells. J Cell Biol33(3):665–678.

20. Lalíc J, Cuevas JM, Elena SF (2011) Effect of host species on the distribution of mu-tational fitness effects for an RNA virus. PLoS Genet 7:e1002378.

21. Nagy PD (2008) Yeast as a model host to explore plant virus-host interactions. AnnuRev Phytopathol 46:217–242.

22. Nakayashiki H, et al. (2005) RNA silencing as a tool for exploring gene function inascomycete fungi. Fungal Genet Biol 42(4):275–283.

23. Nuss DL (2011) Mycoviruses, RNA silencing, and viral RNA recombination. Adv VirusRes 80:25–48.

24. Suzuki N, Geletka LM, Nuss DL (2000) Essential and dispensable virus-encoded repli-cation elements revealed by efforts to develop hypoviruses as gene expression vec-tors. J Virol 74(16):7568–7577.

25. Lange L (1977) Experiments on establishing a BSMV infection in three phytopatho-genic fungi. Phytopathol Z 90(2):184–188.

4296 | www.pnas.org/cgi/doi/10.1073/pnas.1315668111 Mascia et al.

Dow

nloa

ded

by g

uest

on

June

1, 2

021

http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF5http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF5http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF5http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=SF5http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1315668111/-/DCSupplemental/pnas.201315668SI.pdf?targetid=nameddest=STXTwww.pnas.org/cgi/doi/10.1073/pnas.1315668111