Development of a transformation system for the

nematophagous fungus Pochonia chlamydosporia

Simon D. ATKINS*, Tim H. MAUCHLINE, Brian R. KERRY and Penny R. HIRSCH

Rothamsted Research, Harpenden, Hertfordshire, AL5 2JQ, UK.E-mail : simon.atkins@bbsrc.ac.uk

Received 7 August 2003; accepted 2 March 2004.

The nematophagous fungus Pochonia chlamydosporia is a potential biocontrol agent against root knot and cystnematodes. Genetic transformation of the fungus to introduce visual marker genes, novel traits, or changes in expression

levels of endogenous genes, would greatly enhance understanding of its behaviour on nematode-infested roots and of itsinteractions with other soil and rhizosphere microorganisms. A transformation system for the introduction of novelgenes into P. chlamydosporia has been developed. Methods to generate protoplasts, introduce DNA and regenerate

transformed viable fungal mycelium have been optimised, using plasmids carrying the green fluorescent protein markergene gfp and the hygromycin resistance gene hph. Cultures of P. chlamydosporia were resistant to high levels of a rangeof fungal inhibitors, including hygromycin, that are commonly used with dominant selectable marker genes in the

transformation of other fungi. However, regenerating protoplasts transformed with hph could be selected by their abilityto grow through an agar overlay containing 1 mg mlx1 hygromycin. Green fluorescence was observed in protoplasts andregenerating mycelium after transformation with gfp, but the GFP phenotype was lost on subculture. Maintenance ofintroduced genes was not stable, and during subculture, PCR assays indicated that the transformants lost both hph and

gfp. When these genes were introduced on the same plasmid, segregation of hph and gfp was observed prior to their loss.It was unclear whether the introduced plasmids were able to replicate autonomously in P. chlamydosporia, or if theyintegrated transiently into the fungal genome. Possible reasons for the instability of the transformants are discussed.

INTRODUCTION

Pochonia chlamydosporia (Zare, Gams & Evans 2001;syn. Verticillium chlamydosporium) is a widespread andnaturally occurring facultative fungal parasite of cystand root-knot nematode eggs and has shown potentialas a biological control agent of these nematodes inpot and microplot (De Leij & Kerry 1991, De Leij,Dennehy & Kerry 1993) and field trials (Atkins et al.2003). The fungus occurs naturally in soil and can col-onise the rhizosphere of a range of plants to a varyingdegree (Bourne, Kerry & De Leij 1994). This is animportant criterion in the selection of isolates as bio-logical control agents.

The potential of P. chlamydosporia as a nematodebiological control agent will become increasinglyimportant when methyl bromide, a major nematicidefor agriculture and horticulture, is phased out in 2005(Thomason 1987), and other methods of control areinvestigated and implemented. It is likely to form part

of an integrated pest management strategy (Atkinset al. 2003).

Methods for the isolation, enumeration and identi-fication of the fungus from root and soil sampleshave been developed (Hirsch et al. 2001). A recentlydesigned competitive PCR assay (Mauchline, Kerry &Hirsch 2002) has allowed more accurate quantificationfrom soil samples, but these methods cannot informon the physiological or developmental status of thefungus, which are of primary importance in under-standing the ecology of different isolates. To fullyunderstand the tri-trophic interactions between thefungus, plant and nematode hosts, methods for in situvisualisation of the fungus are needed. Monoclonalantibodies have been developed that allow visualisationof the fungus on root samples (Hirsch et al. 2001), buthave showed some cross-reactivity with other soilfungi, and further work is necessary to improve theirspecificity.

The development of a transformation system togenerate targeted deletions of candidate genes, or trans-genic fungi stably expressing a visible marker gene* Corresponding author.

Mycol. Res. 108 (6): 654–661 (June 2004). f The British Mycological Society 654

DOI: 10.1017/S0953756204009906 Printed in the United Kingdom.

such as green fluorescent protein (GFP), would enablea more detailed assessment of the interactions withplant and nematode hosts. The reason for choosing gfpas a visual marker gene over other reporter genes, suchas GUS (b-glucuronidase), is that it does not requireeither exogenous substrates, co-factors or antibodiesfor detection. Problems in substrate uptake, the necess-ity to fix, and hence kill the treated tissue, and cellpermealization, especially in multicellular organisms,sometimes limits the use of certain reporter genes(Maor et al. 1998). All of these techniques are un-necessary for the detection of GFP and makes it anideal visual marker gene for in situ studies.

With the range of fluorescent proteins now available(Chalfie & Kain 1998) a transformation system wouldenable a number of isolates to be transformed withmarker genes for different fluorescent proteins, whichcould subsequently be co-inoculated into soil andused to monitor fungal interactions in the rhizosphere.The gene encoding GFP, cloned from the jellyfishAequorea victoria (Prasser et al. 1992) has been suc-cessfully expressed in a wide range of organismsincluding filamentous fungi (Lorang et al. 2001).Expression of gfp in filamentous fungi requires anefficient transformation system, a fungal promoterthat satisfies the requirements of a given experimentalobjective, and a gfp variant that is efficiently trans-lated in fungi. A range of gfp mutants are reportedin the literature offering enhanced expression or site-directed expression and the types and uses are reviewedby Kendall & Badminton (1998) and Lorang et al.(2001).

Here, the development of a transformation systemfor the nematophagous fungus P. chlamydosporia isdescribed and the transformation of protoplasts toexpress gfp and resistance to hygromycin.

MATERIAL AND METHODS

General methods

Pochonia chlamydosporia isolate 10 was revived fromthe Rothamsted Research culture collection. The iso-late was grown from freeze dried stocks on potatodextrose agar PDA (Oxoid, Basingstoke) and incu-bated at 28 xC. Fungal plates were stored at 4 x untilneeded, then sub-cultures were taken and used to in-oculate fresh PDA plates and Czapek Dox liquid broth(Oxoid). Liquid cultures were shaken at 130 rpm unlessstated.

All data were subjected to analysis using one wayanalysis of variance (ANOVA) using the Genstatprogramme (Genstat 5 Committee 1993). Significance(P) is below the 5% level of significance and the stan-dard error of deviation (sed) is recorded in the text.

Plasmids

The plasmid pDH33, described by Smith et al. (1990),contained the hygromycin resistance gene (hph) from

Escherichia coli fused to the Aspergillus niger glucamy-lase gene promoter, and theA. nidulans trpC terminatorfor selection in fungi. The construct was ligated intothe pBR322 cloning vector that contains an ampicillinresistance gene for propagation in E. coli.

The expression vector pTEFEGFP, described by vanden Wymelenberg et al. (1997), contained a red shiftedmutant GFP cDNA fused to the Aureobasidium pull-ulans translation elongation factor promoter, and aterminator region derived from the Aspergillus awamoriglucoamylase gene for expression in fungi. The con-struct was ligated into the pBluescript cloning vectorcontaining the ampicillin resistance gene and lacZ geneunder the control of the lac promoter for propagationin E. coli.

The expression plasmid pCT74, described by Loranget al. (2001), contained the GFP expression vectorSGFP-TYG under regulation by the toxA gene pro-moter, and the hygromycin resistance gene hph underthe regulation of the trpC promoter, ligated into thecloning vector pBluescript for propagation in E. coli.

Plasmids were isolated from 16 h cultures of E. coligrown under selection in LB media using a Qiagenplasmid midi prep (Qiagen, Crawley), quantified byspectrophometry, and quality was assessed by gel elec-trophoresis and digestion with restriction enzymes.

Sensitivity to fungal inhibitors

Fungal inhibitors were added to corn meal agar(Oxoid) after dissolving in an appropriate solvent(Table 1). Compounds were obtained from Sigma(Poole) with the exceptions of benomyl (Dupont,Stevenage) and hygromycin (Melford Laboratories,Ipswich). A range of inhibitor concentrations, up tothe maximum concentration shown in Table 1, weretested. These levels were in excess of ED50 levels thathave affected other fungi (Urban, Bhargava & Hamer1999). Spores were washed from a fresh culture ofP. chlamydosporia and added to each plate, and

Table 1. List of fungal inhibitors, solvents and maximum

concentrations tested for inhibition of Pochonia chlamydosporia.

Inhibitor Solvent

Maximum

concentration

tested mg mlx1

Hygromycin Water 1000

Puromycin Ethanol 100

Paromomycin Water 500

Geneticin (G418) Water 500

Blasticidin Water 300

Phosphinothricin (PPT) Water 1000

Bleomycin Water 100

Phleomycin Water 100

Cyclohexamide Water 500

Oligomycin Ethanol 200

Benomyl Ethanol 200

Thiabendazole Ethanol 50

Carbendazim Ethanol 50

N-phenylcarbamate (MDPC) Methanol 50

S. D. Atkins and others 655

sensitivity to the solvent was tested using plates con-taining the solvent alone. To demonstrate that theselective agent was not affected by the agar, cultures ofVerticillium dahliae, V. tricorpus and V. albo-atrum(supplied by Simon Foster, Rothamsted Research) wereadded to agar plates containing hygromycin at a con-centration range of 0–200 mg mlx1. V. dahliae had beenpreviously shown to be sensitive to hygromycin (Do-binson 1995), and all three fungal species are related toPochonia, which was previously classified in the samegenus, Verticillium.

Development of a system for protoplast generation

Each step in the process was tested to optimise thegeneration of protoplasts from fungal mycelium. Pro-toplasts were examined after each treatment usingEvans Blue staining (Coury et al. 1993).

Effect of osmotic buffer on protoplast generation

Spores of Pochonia chlamydosporia were washed froman actively growing culture on an agar plate and usedto inoculate 100 ml Czapek Dox broth (Oxoid) in a250 ml conical flask. Previous work (Atkins 2000) hadshown that germinating spores yielded the highestlevel of protoplasts. The culture was incubated at 28 x

for 48 h and the mycelium collected by centrifugationat 8 K g for 15 min at 4 x in sterile centrifugationbottles. Aliquots (1 g wet weight mycelium) were addedto 10 ml sterile Osmotic Buffer (OM–10 mM Tris-HCl,pH 7.5) containing one of the following reagentsdissolved in water (0.8 M sorbitol ; 1 M sorbitol ; 1.2 M

sorbitol ; 0.6 M KCl; 1 M sucrose; 1.2 M MgSO4 ; 0.6 M

NaCl; 0.7 M NaCl). Filter sterilised Novozyme 234(Sigma) was added to each buffer at a concentration of5 mg mlx1. Each buffer was incubated at 28 x withgentle shaking, 80 rpm.

Effect of pre-incubation with b-mercaptoethanol

Pre-incubation of mycelium with b-mercaptoethanol,before addition of the lysing enzyme, had been shownto increase protoplast generation in some fungi (Davies1985). Fungal spores were collected and grown asabove. Mycelium was preincubated in 50 ml Buffer 1(10 mM sodium phosphate, pH 7.5, 10 mM EDTA)containing 1% v/v b-mercaptoethanol in a 250 mlconical flask and incubated at 28 x for 1 h, with shakingat 150 rpm. A control of Buffer 1 containing no b-mercaptoethanol was run in parallel. Mycelium wascollected after 2 h incubation by centrifugation for15 min at 8 K g at 4 x. Mycelium was re-suspended in50 ml OM containing 5 mg mlx1 Novozyme 234.

Effect of Novozyme 234 concentration onprotoplast generation

Mycelium was pre-incubated in Buffer 1 containing1% v/v b-mercaptoethanol, collected as above and

re-suspended in 50 ml OM. A range of Noyozyme 234concentrations (1, 5, 10, and 15 mg mlx1) was added toaliquots of mycelium and incubated as above.

Effect of length of incubation in Novozyme 234 onprotoplast regeneration

Protoplasts were prepared as above and incubated in10 mg mlx1 Novozyme. At hourly intervals, for 3 h,aliquots were removed and observed.

Transformation

Pochonia chlamydosporia was co-transformed withpDH33 and pTEFEGFP, or transformed with pCT74alone. The protoplast production and transformationsystems were optimised from the results of the exper-iments outlined above. Fungal spores were washed froma two week culture grown on corn meal agar plates,into 200 ml Czapek Dox liquid broth (Oxoid) in 1 litreflasks for 48 h. Mycelium was collected by filtrationthrough a sterile 30 mm nylon mesh (Lockertex,Cheshire), washed twice with sdH2O, once with Buffer1 before re-suspension in 50 ml Buffer 1 and incubatedat 28 x with gentle agitation (80 rpm) for 1 h. Myceliumwas collected by filtration as before, washed twice withsdH2O and once with OM, before re-suspension in25 ml OM and incubation, with gentle agitation(60 rpm) at 28 x for 2 h. Filter-sterilised Novozyme234 was added to the OM at a concentration of10 mg mlx1. Protoplasts were separated from mycelialdebris by filtering through a 30 mm mesh, collectingthe filtrate in a sterile centrifuge bottle, then pelletedby centrifugation in a swing-out rotor at 2.8 K g for7 min. The pellet was resuspended and washed twicewith STC (1 M sorbitol, 10 mM Tris-HCl, pH 7.5;50 mM CaCl2). The centrifugation step was repeatedafter each wash. Protoplasts were re-suspended inSTC and counted using a haemocytometer and theconcentration adjusted to 107 protoplasts mlx1.

Protoplasts were transformed by adding 50 ml PTC(25% PEGx8000 (Sigma) in STC) to 200 ml protoplastsuspension. DNA was added at a concentration of10 mg and gently mixed with the pipette tip and in-cubated on ice for 20 min. A further 2 ml of PTC wasadded and incubated for a further 20 min at roomtemperature. After incubation, 4 ml of RegenerationMedia (RMx1 M sorbitol ; 0.1% yeast extract ; 0.1%di-potassium orthophosphate ; 0.05% MgSO4) wasadded and the protoplasts incubated at 28 x in anorbital shaker with agitation (100 rpm) for 2 h. Theprotoplasts were plated on osmotically buffered plates(PDA+1 M sorbitol) and overlaid with osmoticallybuffered PDA containing 1 mg mlxl hygromycin(Melford Laboratories), and incubated at 28 x. Ascontrols, protoplasts were plated on unbuffered media,and untransformed protoplasts were plated onto selec-tive buffered agar. Stocks (200 ml) of transformed,or unused protoplasts were supplemented with 50 ml

Transformation of Pochonia chlamydosporia 656

PTC and 2.5 ml dimethyl sulfoxide (DMSO, Sigma)and stored at x80 x, according to Bardi, Perotto &Bonfante (1999) for further analysis if necessary. Allpreparations were examined using a Zeiss Axioskopmicroscope (Welwyn Garden City) fitted with epi-fluorescence illumination with a 455 nm excitation filterand a 520 nm barrier filter. Samples were observedusing r40 oil immersion lens and images collectedwith a Xillix Micro Image digital camera (Vancouver)attached to an OpenLab Image Analysis system(Improvision, Coventry).

Selection of transformants

Colonies that grew through the hygromycin-supple-mented overlay before untransformed colonies ap-peared on the control plates (a window of 12–24 h)were transfered onto fresh PDA agar supplementedwith 1 mg mlxl hygromycin. DNA was extractedfrom individual colonies by the method described byKlimyuk et al. (1993). DNA was used in PCR reactionsusing primers for the detection of hph or gfp (seebelow). Colonies that showed positive for transform-ation with either of the two genes were sub-culturedtwice more onto unselective PDA media at weeklyintervals and colonies were tested once more for thepresence of hph or gfp using PCR.

Detection of hph

Primers (HPHF: CGC ATA ACA GCG GTC ATTGAC TGG AGC; HPHR: GTC GGG GCG TCGGTT TCC ACT ATC) were taken from the GenBank/EMBL database (accession numbers E16482 andE16481 respectively). Detection of the hph gene wasperformed by PCR optimized in a reaction volumeof 20 ml containing 0.1 mM each primer, 2 mlr10 PCRreaction buffer (Roche, Lewes, 1.5 mM Mg2+), 0.1 mMeach dNTP, 1 U Taq polymerase (Roche), 1 ml DNA(20–60 ng). PCR conditions were optimised asfollows: hot start hold of 95 x followed by 40 cycles of94 x for 1 min, 60 x for 1 min, 72 x for 1 min, with afinal incubation at 72 x for 5 min. PCR products wereseparated on 2% Nu-Seive agarose gels (FMC Bio-Products, Rockland, ME) in TBE buffer. The 123 bpDNA ladder (Invitrogen, Paisley) was used as a sizemarker. Positive control DNA was amplified from pureplasmid DNA. The primers generated a single fragmentof 370 bp.

Detection of gfp

Primers (GFPF: GGG(C) GAA(G) TGG G(C)GATGC A(C)AC A(C)TA CGG; GFPR: ACG CTG(T)CCT TCG(C) TCA(G) ATG TTG T(G) were designedby comparison of GFP sequences from the GenBank/EMBL database. These were then tested for the detec-tion of a range of GFP variants (Atkins 2000). Theprimers generated a single fragment of 380 bp.

Conditions for PCR were the same as for thedetection of the hph gene. Isolates were confirmed asPochonia chlamydosporia by PCR diagnosis usingspecies specific primers described byHirsch et al. (2000).

RESULTS

Pochonia chlamydosporia showed an inherent resist-ance to a wide range of fungicides tested (Table 1,Fig. 1). This resistance was not shown by the otherfungal isolates tested for their ability to grow onmedia supplemented with hygromycin with LD50 levelsof 12.5 mg mlx1 for V. albo-atrum and V. dahliae, and50 mg mlx1 for V. tricorpus, demonstrating that themedia did not affect sensitivity. Growth of P. chlamy-dosporia was inhibited by methyl-N(3,5-dichloro-phenyl)carbonate (MDPC), a compound that bindsto the b-tubulin protein, but previous investigationsof the b-tubulin gene (Hirsch et al. 2001) had demon-strated that the P. chlamydosporia genotype did notmatch the phenotypic response to this compound.Therefore, transformation with a b-tubulin gene vari-ant resistant to MDPC (as demonstrated by Blakemore1990) was highly unlikely to change this phenotypeand further investigation into the phenomenon isnecessary.

Optimisation of protoplast generation

The optimum time for generation of mycelium was 48 hafter inoculation of liquid broth with spores. After thistime mycelium pellets formed that severely inhibitedprotoplast generation. The effect of the osmotic bufferon protoplast generation was drastic (Fig. 2). The useof 1 M sorbitol generated significantly greater numbersof protoplasts (P<0.001, SED¡3.8r104, n=8) thanthe other buffers. Protoplasts generated were sphericaland ranged in size, averaging 5 mm in diameter (hyphaldiameter averaged 2.5 mm). Significantly greater num-bers of protoplasts (P<0.001, SED¡6.2r105, n=3)were generated when mycelium was pre-washed in

0102030405060708090

100

Hyg

rom

ycin

Pur

omyc

inP

arom

omyc

inG

enet

icin

PP

TB

leom

ycin

Phl

eom

ycin

Cyc

lohe

xim

ide

Olig

omyc

inB

enom

ylT

hiab

enda

zole

Car

bend

azim

MD

PC

Inhibitor

% in

hib

itio

n

Fig. 1. Effect of fungicides (n=3) on colony growth of

Pochonia chlamydosporia.

S. D. Atkins and others 657

Buffer 1 containing 1% b-mercaptoethanol (mean4.0r106¡7.2r105) than when b-mercaptoethanolwas absent (mean 7.3r105¡8.3r104). Novozyme 234concentration also affected protoplast generation. Noprotoplasts were generated at the lowest concentrationof enzyme used (1 mg mlx1), whereas significantlygreater numbers of protoplasts were generated(P<0.001, SED¡3.8r104, n=16) at the higher con-centrations of 10 mg mlx1 (2.2r105¡3.9r104) and15 mg mlx1 Novozyme (1.6r105¡3.2r104). No sig-nificant differences were recorded between treatmentsthat had either 10 or 15 mg mlx1 Novozyme added,although fewer protoplasts were generated at thehighest concentration. The incubation time in thelysing solution also affected protoplast generation.At 1 h the protoplast yield was low (3.8r105¡3.5r104) but had increased at 2 h (7.2r105¡4r104) and3 h (5.8r105¡3.1r104). This increase was significant(P<0.001, SED¡1.3r104, n=16). The optimum con-ditions for protoplast generation from P. chlamydo-sporia mycelium are listed in Table 2, which gave ayield of protoplasts of 1.8r108 gx1 wet weightmycelium.

Fungal transformation

Expression of gfp was clearly seen in both protoplastsand regenerating mycelium transformed with pTE-FEGFP or pCT74, examples of which are shown inFig. 3. Expression of gfp was observed in regeneratingmycelium up to three weeks after transformation whenkept in Czapek Dox liquid medium, but was not

observed in colonies plated onto solid media, irrespec-tive of whether Czapek Dox agar or the richer cornmeal agar was used. This was not due to an inability toobserve fluorescence on agar media because myceliumexpressing GFP in liquid medium continued to fluor-esce when transferred onto agar.

Co-transformation with pDH33 and pTEFEGFP

Of the 50 colonies selected after co-transformation withpDH33 and pTEFEGFP, 11 contained the hph genefrom pDH33, but none contained the gfp gene frompTEFEGFP, although gfp expression was clearlyvisible in regenerating mycelium in liquid culture(Fig. 3) demonstrating that hph-transformed isolateswere preferentially selected. All colonies were con-firmed to be Pochonia chlamydosporia when testedusing specific primers for the identification of thespecies (Hirsch et al. 2000). The transformed colonieswere sub-cultured onto fresh PDA plates containing1 mg mlx1 hygromycin after 1 wk and screened againusing the hygromycin primers. PCR analysis indicatedthat six of the remaining 11 colonies contained thehph gene. After a further week these colonies weresubcultured onto fresh PDA plates with and withoutselection with hygromycin. PCR screening of thesesub-cultures with the hph-specific primers demon-strated that only one isolate under hygromycin selec-tion retained hph, but had lost the gene on non-selectiveplates. Conversely, two other isolates had retainedhph when cultured on non-selective plates, but hadnot retained it on the selective media. After one furthersub-culture hph could not be detected in any of thecolonies using PCR.

Transformation with pCT74

Of the 50 colonies taken from the selection platesinoculated with Pochonia chlamydosporia after trans-formation with pCT74, five colonies contained hphand gfp. These isolates were sub-cultured onto freshPDA plates and DNA was extracted and tested againusing the hph and gfp selective primers. Three isolatesstill retained hph, but only two contained gfp. The col-onies were sub-cultured again as described earlier.On the third sub-culture a spore suspension was madeof all isolates and spread onto plates with and withouthygromycin and DNA was extracted from the resultingcolonies. One isolate still retained hph and gfp in allthe resulting colonies sampled, whereas, the other iso-lates had retained only gfp. No gfp expression waspresent in the mycelia of any of the isolates.

DISCUSSION

The first report of a DNA-mediated transformationsystem of a fungus was in 1973 (Mishra, Szabo &Tatum 1973). Since then many fungal species have been

0

1

2

3

4

5

6

7

0.8 Msorbitol

1 Msorbitol

1.2 Msorbitol

0.6 MKCl

1 Msucrose

1.2 MMgSO4

0.6 MNaCl

0.7 MNaCl

Buffer

Pro

topl

asts

x 1

05 ml-1

Fig. 2. Effect of osmotic buffer (n=8) on protoplast gener-

ation.

Table 2. Optimised conditions for protoplast generation from

Pochonia chlamydosporia.

Age of culture Conidiospores germinated for 48 h in

Czapek Dox broth

Pre-incubation 1 h in Buffer 1 plus 1%

b-mercaptoethanol

Osmotic buffer OM plus 1 M sorbitol

Lysing enzyme and

concentration

Novozyme 234 (Sigma) at 10 mg mlx1

Incubation time 2 h with gentle shaking, 80 rpm

Transformation of Pochonia chlamydosporia 658

transformed and a range of protocols now exists(Fincham 1989, 1999, Hynes 1996) not only for CaCl2/PEG mediated transformation of protoplasts but alsoelectroporation, biolistics and Agrobacterium mediatedtransformation. What is apparent from the literatureis that a process that works for one fungus may notnecessarily work for another and optimisation of allstages of protoplast generation and transformation isessential. Various transformation protocols based onbiolistics, electroporation and Agrobacterium mediatedtransformation were described by Atkins (2000). Noneof these methods were successful in transformingPochonia chlamydosporia. A reliable method for thegeneration and regeneration of viable protoplasts isfundamental to any transformation strategy. For thefirst time, a successful protocol for protoplast gener-ation, transformation, and transient gene expressionin the nematophagous fungus P. chlamydosporia. Thisrequired optimisation of several factors that affectprotoplast generation and are listed in Table 2. A pre-wash of the mycelium with a buffer containing b-mercaptoethanol was essential for a high yield ofprotoplasts. The b-mercaptoethanol destabilises thefungal membrane sufficiently to enable the lysingenzyme to function more effectively, resulting in a sig-nificantly greater number of protoplasts.

The high level of resistance to a wide range offungicides in P. chlamydosporia was not seen with theother fungi tested. Consequently, background growthof untransformed propagules on selective media wouldswamp any transformants with increased resistance.Nevertheless, selection of colonies that grew quicklythrough the hygromycin selective overlay yielded ahigh percentage of hph-transformed isolates as demon-strated by the PCR detection, and this method hasbeen used previously to select for transformants offungal isolates that have a natural resistance to hygro-mycin (Punt & Hondel 1992). The ability of trans-formed protoplasts to regenerate more quickly in the

presence of a high concentration of hygromycin in-dicates that protoplasts may be more susceptible tothe compound than mycelium. Once the protoplastshad regenerated, this sensitivity was lost and main-taining positive selection of hph with the fungicide hadno effect, resulting in loss of the hph gene over time.

Pochonia chlamydosporia demonstrated inherentresistance to a wide range of fungal inhibitors, rep-resenting a variety of different modes of action. Pre-vious research (Hirsch et al. 2001) has demonstratedthat the phenotypic resistance of the fungus to beno-myl, and sensitivity to MDPC, was not the result of amutation in the b-tubulin gene, the site of action forthe fungicide. Resistance to any one inhibitor maynot be due to a specific change in the target, or a speci-fic detoxification mechanism, but rather may be theresult of a more general mechanism such as a multidrugresistant pump (Pereira, Fachin &Martinezrossi 1998).However, the exact mechanism(s) of resistance inP. chlamydosporia needs further investigation.

The protoplasts generated were able to take upand express exogenous DNA, a fundamental step in atransformation system. The results indicate that co-transformation of the fungus with two or more plas-mids may not be a feasible option as transformantsselected on hygromycin contained only the hph, withno trace of gfp, although regenerating protoplastswere seen to express GFP in liquid culture. Trans-formation of protoplasts with a single vector carryingboth the resistance gene and the marker gene yieldedtransformants containing both genes when screenedusing PCR. The gradual loss of the exogenous DNAover time and with sub-culturing points to the inabilityof the plasmid to stably integrate, or the ability ofthe fungus to excise unwanted DNA. PCR demon-strated that gfp was still present within the fungus, butthe lack of expression in the mycelium from coloniesgrown on plates demonstrated that the gene was eithernot being expressed, or was being expressed at an

(b)(a)

(c) (d)

Fig. 3. Detection of GFP in protoplasts and regenerating mycelium in liquid media transformed with pTEFEGFP (a, c)

and pCT74 (b, d). Bar=5 mm.

S. D. Atkins and others 659

undetectable level. It is unclear whether any vectorDNA had integrated into the fungal genomic DNA, orwhether the autonomous plasmid was replicating. Thehigh copy number of plasmids in the initial transfor-mants may be diluted during cell division followingprotoplast regeneration and gradually lost from thefungal mycelium during growth.

P. chlamydosporia may have a mechanism forsilencing or regulating gene expression from unwantedor exogenous DNA. This phenomenon, in which ex-pression of the transgene and of endogenous genescontaining sequences homologous to the transgenecan be blocked, is involved in virus resistance andgenome maintenance and has been investigated in anumber of fungal species (Cogoni & Macino 1999).Irelan & Selker (1996) investigated three methods ofgene silencing in filamentous fungi, methylation in-duced premeiotically (MIP), repeat-induced mutation,and quelling of the gene. Each of the gene silencingprocesses involved the silencing of repeated sequences.Thus each has the potential for preventing proliferationof selfish DNA elements and silencing repeated se-quences that may arise in a transformation systemwhere many exogenous gene copies may become in-corporated. Insertion of many copies of the exogenousvector into the genome may result in gene silencingmechanisms being triggered and further investigationof this phenomenon is necessary. PCR detection ofonly hph or gfp in the protoplasts transformed with thevector pCT74, indicated that the fungus was able todelete genes. Transformants contained either or both,of the genes, with no apparent preference for the generemoved.

Although this paper demonstrates the productionand transformation of protoplasts of the fungus P.chlamydosporia, it also indicates problems with ex-pression of GFP. Expression of GFP may be deleteri-ous to the fungus. For example, bacterial cells becomeosmotically unstable when the protein is over expressed(Brendan Cormack, pers. comm.). The GFP was seento be clearly compartmentalised within regeneratingmycelium and protoplasts (Fig. 3). This may indicateadverse effects on P. chlamydosporia, resulting in astrong counter selection and consequent loss of trans-formants containing the expression vector.

The next stage in developing a transformationsystem for P. chlamydosporia is to make transformantsmore stable. A new range of fluorescent protein vectorsare now available (Reef Coral Fluorescent Proteins,CloneTech, Bedford) and have been used in thesuccessful transformation of fungi (Bourett et al.2002, Mikkelsen et al. 2003). These vectors are re-ported to offer a higher degree of stability in trans-formants (CloneTech).Fluorescentactivatedcell sorting(FACS) of transformed protoplasts could yield agreater number of transformed protoplasts than theselection method outlined in this paper, and therefore,increase the likelihood of stable transformants beingisolated. Vectors designed to facilitate site-directed

integration can result in a stable integration of thevector DNA (Goldschmidtclermont 1991) and thisoffers possibilities for future transformation of P.chlamydosporia.

The system described in this paper can be used toinvestigate the expression of transgenes in Pochoniaprotoplasts, although the genes were not maintainedin a stable manner after regeneration. A stable trans-formed isolate expressing a visible marker gene wouldgreatly enhance not only the knowledge of how P.chlamydosporia functions and interacts in the rhizo-sphere, increasing our understanding of how thisfungus can be developed as a biological control agent,it is also of fundamental importance in trying tounderstand how fungi interact in the rhizosphere ingeneral.

ACKNOWLEDGEMENTS

Rothamsted Research receives grant-aided support from the UK

Biotechnology and Biological Sciences Research Council. We also

acknowledge support provided by EU grant Fair-Pl97-3444 as part of

this work.

REFERENCES

Atkins, S. D. (2000) Development of a transformation system for the

nematophagous fungus Verticillium chlamydosporium. PhD thesis,

University of Nottingham, Nottingham.

Atkins, S. D., Hidalgo-Diaz, L., Kalisz, H., Mauchline, T. H.,

Hirsch, P. R. & Kerry, B. R. (2003) Development of a new

management strategy for the control of root-knot nematodes

(Meloidogyne spp.) in organic vegetable production. Pest

Management Science 59 : 183–189.

Bardi, L., Perotto, S. & Bonfante, P. (1999) Isolation and regener-

ation of protoplasts from two stains of the ericoid mycorrhizal

fungus Oidiodendron maius : sensitivity to chemicals and heavy

metals. Microbiological Research 154 : 105–111.

Blakemore, E. J. A. (1990) Molecular genetics of benzimidazole

resistance in Pseudocercosporella herpotrichoides. PhD thesis,

University of Nottingham, Nottingham.

Bourett, T. M., Sweigard, J. A., Czymmek, K. J., Carroll, A. &

Howard, R. J. (2002) Reef coral fluorescent proteins for visualizing

fungal pathogens. Fungal Genetics and Biology 37 : 211–220.

Bourne, J. M., Kerry, B. R. & De Leij, F. A. A. M. (1994) Methods

for the study of Verticillium chlamydosporium in the rhizosphere.

Journal of Nematology 25 : 587–591.

Chalfie, M. & Kain, M. (1998) Green Fluorescent Protein: properties,

applications and protocols. John Wiley, Chichester.

Cogoni, C. &Marcino, G. (1999) Gene silencing inNeurospora crassa

requires a protein homologous to RNA-dependent RNA poly-

merase. Nature 399 : 166–169.

Coury, D. A., Naganuma, T., Polnefuler, M. & Gibor, A. (1993)

Protoplasts of Gelidium rorustum (Rhodophyta). Hydrobiologia

261 : 421–427.

Davies, B. (1985) Factors influencing protoplast isolation. In Fungal

Protoplasts: application in biochemistry and genetics (J. F. Peberdy

& L. Ferenczy, eds) : 45–71. Marcel Dekker, New York.

De Leij, F. A. A. M., Dennehy, J. A. & Kerry, B. R. (1993) The effect

of watering on the distribution of Verticillium chlamydosporium

in soil and the colonisation of egg masses ofMeloidogyne incognita

by the fungus. Nematologica 39 : 250–265.

De Leij, F. A. A. M. & Kerry, B. R. (1991) The nematophagous

fungus, Verticillium chlamydosporium, as a potential biological

control agent for Meloidogyne arenaria. Revue de Nematologie 14 :

157–164.

Transformation of Pochonia chlamydosporia 660

Dobinson, K. F. (1995) Genetic transformation of the vascular wilt

fungus Verticillium dahliae. Canadian Journal of Botany 73 :

710–715.

Fincham, J. R. S. (1989) Transformation in fungi. Microbiological

Reviews 53 : 148–170.

Fincham, J. R. S. (1999) Fungal genetics-past and present. Journal of

Genetics 77 : 55–63.

Genstat 5 Committee (1993) Genstat 5 Release 3 Reference Manual.

Clarendon Press, Oxford.

Goldschmidtclermont, M. (1991) Transgenic expression of amino-

glycoside adenine transferase in the chloroplast-a selectable marker

for site directed transformation of Chlamydomonas. Nucleic Acid

Research 19 : 4083–4089.

Hirsch, P. R., Atkins, S. D., Mauchline, T. H., Morton, C. O.,

Davies, K. G. & Kerry, B. R. (2001) Methods for studying the

nematophagous fungus Verticillium chlamydosporium in the root

environment. Plant and Soil 232 : 21–30.

Hirsch, P. R., Mauchline, T. H., Mendum, T. A. & Kerry, B. R.

(2000) Detection of the nematophagous fungus Verticillium

chlamydosporium in nematode-infested plant roots using PCR.

Mycological Research 104 : 435–439.

Hynes, M. J. (1996) Genetic transformation of filamentous fungi.

Journal of Genetics 75 : 297–311.

Irelan, J. T. & Selker, E. U. (1996) Gene silencing in filamentous

fungi: RIP, MIP and quelling. Journal of Genetics 75 : 313–324.

Kendall, J. M. & Badminton, M. N. (1998) Aequorea victoria bio-

luminescence moves into an exciting new era. TibTech 16 : 216–224.

Klimyuk, V. I., Carroll, B. J., Thomas, C. M. & Jones, J. D. G.

(1993) Alkali treatment for rapid preparation of plant material for

reliable PCR analysis. Plant Journal 3 : 493–494.

Lorang, J. M., Tuori, R. P., Martinez, J. P., Sawer, T. L., Redman,

R. S., Rollins, J. A., Wolpert, T. J., Rodriquez, R. J., Dickmann,

M. B. & Cuiffetti, L. M. (2001) Green fluorescent protein is lighting

up fungal biology. Applied and Environmental Microbiology 67 :

1987–1994.

Maor, R., Puyesky, M., Horwetz, B. A. & Sharon, A. (1998) Use of

green fluorescent protein (GFP) for studying development and

fungal plant interactions in Cochliobolus heterostrophus. Myco-

logical Research 102 : 491–496.

Mauchline, T. H., Kerry, B. R. & Hirsch, P. R. (2002) Quantification

in soil and the rhizosphere of the nematophagous fungus,

Verticillium chlamydosporium by competitive PCR and comparison

with selective plating. Applied and Environmental Microbiology 68 :

1846–1853.

Mikkelsen, L., Sarrocco, S., Lubeck, M. & Jensen, D. F. (2003)

Expression of the red fluorescent protein DsRed-Express in fila-

mentous ascomycete fungi. FEMS Microbiology Letters 223 :

135–139.

Mishra, N. C., Szabo, G. & Tatum, E. L. (1973) Nucleic acid induced

genetic changes inNeurospora. In The Role of RNA in Reproduction

and Development (M. C. Niu & S. J. Segal, eds) : 259–268. Elsevier/

North Holland Publishing, Amsterdam.

Prasser, D. C., Eckenrode, G., Ward, W. W., Predergast, F. G. &

Cormier, M. J. (1992) Primary structure of the Aequorea victoria

green fluorescent protein. Gene 111 : 29–233.

Pereira, M., Fachin, A. L. & Martinezrossi, N. M. (1998) The gene

that determines resistance to tioconazole and to acridine deriva-

tives in Aspergillus nidulans may have a corresponding gene in

Trichophyton rubrum. Mycopathology 143 : 71–75.

Punt, P. J. & Hondel, C. A. M. J. J. V. D. (1992) Transformation

of filamentous fungi based on hygromycin B and phleomycin

resistance markers. Methods in Enzymology 216 : 447–457.

Smith, T., Gaskell, J., Berka, R., Henner, D. & Cullen, D. (1990)

Promoter for the Aspergillus niger glucoamylase gene functions

in Ustilago maydis. Gene 88 : 259–262.

Thomason, I. J. (1987) Challenges facing nematology: environmental

risks with nematicides and the need for new approaches. In Vistas

on Nematology (J. A. Veech & D. W. Dickson, eds) : 469–476.

Society of Nematologists, Hyattsville.

Urban,M., Bhargava, T. & Hamer, J. E. (1999) An ATP-driven efflux

pump is a novel pathogenicity factor in rice blast disease. The

EMBO Journal 18 : 512–521.

van den Wymelenberg, A. J., Cullen, D., Spear, R. N., Schoenike, B.

& Andrews, J. H. (1997) Expression of green fluorescent protein

in Aureobasidium pullulans and quantification of the fungus on

leaf surfaces. Biotechniques 23 : 686–690.

Zare, R., Gams, W. & Evans, H. C. (2001) A revision of Verticillium

section Prostrata. V. The genus Pochonia, with notes on Rotifero-

phthora. Nova Hedwigia 73 : 51–86.

Corresponding Editor: S. J. Assinder

S. D. Atkins and others 661