Molecular Characterization of an Endopolygalacturonase from Fusarium oxysporum Expressed during Early Stages of Infection

Abstract

The tomato vascular wilt pathogen Fusarium oxysporum f. sp. lycopersici produces an array of pectinolytic enzymes that may contribute to penetration and colonization of the host plant. Here we report the isolation of pg5, encoding a novel extracellular endopolygalacturonase (endoPG) that is highly conserved among different formae speciales of F. oxysporum. The putative mature pg5 product has a calculated molecular mass of 35 kDa and a pI of 8.3 and is more closely related to endoPGs from other fungal plant pathogens than to PG1, the major endoPG of F. oxysporum. Overexpression of pg5 in a bacterial heterologous system produced a 35-kDa protein with endoPG activity. Accumulation of pg5 transcript is induced by citrus pectin and d-galacturonic acid and repressed by glucose. As shown by reverse transcription-PCR, pg5 is expressed by F. oxysporum in tomato roots during the initial stages of infection. Targeted inactivation of pg5 has no detectable effect on virulence toward tomato plants.

Among the wide array of cell wall-degrading enzymes (CWDEs) are polygalacturonases (PGs), which are involved in the degradation of pectin, a complex polysaccharide that is primarily found in the middle lamella and primary cell wall of higher plants (33). Endopolygalacturonases [endoPGs; poly(1,4-α-d-galacturonide) galacturonohydrolase; EC 3.2.1.15], which are produced by many organisms and have long been studied with regard to their role in many aspects of pathogenicity, are responsible for depolymerization of homogalacturonan, a major component of plant cell walls (10). The genes encoding a number of fungal endoPGs, including those of Aspergillus flavus, Aspergillus niger, Botrytis cinerea, Cochliobolus carbonum, Colletotrichum lindemuthianum, Cryphonectria parasitica, Fusarium moniliforme, Fusarium oxysporum, and Sclerotinia sclerotiorum, have been cloned and characterized (3–6, 14, 17, 23, 27, 28, 34). However, with a few notable exceptions (28, 30), conclusive evidence for the role of fungal CWDEs in pathogenesis by targeted gene disruption is lacking (17, 26, 27). Besides acting as virulence factors, endoPGs may also function as avirulence determinants through release of oligogalacturonide inducers of plant defense (11) and interactions with plant proteins that modulate activities of PG-inhibiting proteins (PGIPs) (7). Regulation of endoPG gene expression is generally subject to the carbon source available, with the exception of a constitutively expressed gene in B. cinerea (31). These genes are mainly induced by pectin and subject to glucose repression (14, 27, 35), although complete regulation of endoPG gene expression is not completely understood.

EndoPGs are among the first CWDEs secreted by F. oxysporum f. sp. lycopersici upon contact with the host tissue (21). Like other pectinolytic enzymes, endoPGs have been suggested to be of prime importance in a number of key steps during infection, such as root penetration, colonization of the vascular tissue, and perforation of xylem vessel plates (2). To investigate this hypothesis, we have previously isolated an endoPG gene and an exopolygalacturonase (exoPG) gene from F. oxysporum. Both genes are expressed by the fungus during infection, but inactivation of either gene had no effect on virulence (14, 19).

The soilborne plant pathogen F. oxysporum Schlecht causes vascular wilt disease on a wide variety of crops (2). This fungus produces a wide variety of extracellular CWDEs, including xylanases, cellulases, proteases, pectate lyases, and exo- and endoPGs, (9, 12, 13, 16, 18, 20, 24), that may contribute to the degradation of the structural barriers constituted by plant cell walls. The gene encoding the major endoPG secreted by F. oxysporum, PG1, has been cloned, and mutants lacking an active copy of pg1 were shown to retain full virulence (14). These mutants still exhibited extracellular PG activity, although it was strongly reduced in comparison with the wild type, suggesting the presence of additional PG genes in F. oxysporum. The present report describes the isolation of pg5, encoding a novel endoPG, from F. oxysporum. We show that pg5 is expressed during saprophytic growth on citrus pectin and in planta during the initial stages of infection. Targeted inactivation of pg5 suggests that the enzyme is not essential for pathogenesis on tomato plants.

MATERIALS AND METHODS

Fungal isolates and culture conditions.

F. oxysporum f. sp. lycopersici strain 4287 (race 2) and F. oxysporum f. sp. melonis strain 18 M (race 1) were obtained from J. Tello (Universidad de Almería, Almería, Spain) and stored as microconidial suspensions in 30% glycerol at −80°C. The pathotypes of the isolates were periodically confirmed by plant infection assays in a growth chamber. The detailed origins of the other F. oxysporum isolates used for Southern analysis are described elsewhere (15).

For extraction of genomic DNA, mycelia were obtained from cultures grown in potato dextrose broth (Difco, Detroit, Mich.) on a rotary shaker at 150 rpm and 28°C. For analysis of gene expression, microconidia were germinated in potato dextrose broth, washed in sterile water, and transferred to synthetic medium (SM) as previously described (14). Prior to being autoclaved, SM was supplemented with one of the following substrates: 1% (wt/vol) citrus pectin, 0.5% (wt/vol) polygalacturonic acid sodium salt (PGA), 1% (wt/vol) d-galacturonic acid, 1% (wt/vol) rhamnose (all from Sigma), 1% (wt/vol) glucose, and 2.5% (wt/vol) tomato vascular tissue (TVT) obtained as described previously (14). Plant seeds were kindly provided by Novartis, Almería, Spain.

Nucleic acid manipulations.

pg5 was isolated from a lambda EMBL3 genomic library of F. oxysporum f. sp. lycopersici isolate 42-87 by using the Cryphonectria parasitica enpg1 clone as a probe (17). Library screening, subcloning, and other routine procedures were performed as described in standard protocols (25). Sequencing of both DNA strands was performed at the Servicio de Secuenciación Automática de DNA, CIB, Madrid, Spain, using a Dyedeoxy Terminator Cycle Sequencing Kit (Perkin-Elmer, Foster City, Calif.) on an ABI Prism 377 Genetic Analyzer apparatus (Applied Biosystems, Foster City, Calif.). Analyses of sequencing data were carried out with the Lasergene programs (DNAStar Inc., Madison, Wis.). DNA and protein sequence databases were searched by using the BLAST algorithm (1) at the National Center for Biotechnology Information (Bethesda, Md.).

Genomic DNA was extracted from F. oxysporum mycelium as described previously (22), digested with appropriate restriction enzymes, and subjected to Southern hybridization analysis, as described in standard protocols (25), by using a nonisotopic digoxigenin labeling kit (Roche Molecular Biochemicals, Mannheim, Germany) according to the instructions of the manufacturer. For Northern analysis, 5 μg of total RNA extracted as described previously (8) was separated on a formaldehyde–1% agarose gel and transferred to a positively charged nylon membrane (Roche Molecular Biochemicals) by capillary transfer. For quantification, transferred RNA on the membrane was stained for 5 min in 0.02% methylene blue–0.3 M sodium acetate, pH 5.2. After destaining in 20% ethanol, filters were subjected to hybridization by the use of a nonisotopic digoxigenin labeling kit. A nonisotopically labeled single-stranded antisense DNA probe was generated by asymmetric PCR, as described previously (14), using as a template a 1.6-kb HindIII fragment, containing the pg5 coding region, cloned into the Bluescript KS(+) vector.

Assays of pathogenicity on tomato plants and fruits.

Infection of tomato plants was performed as reported elsewhere (14). Briefly, tomato seedlings of cultivar Vemar were inoculated with F. oxysporum f. sp. lycopersici strains by dipping the roots in a microconidial suspension, planting seedlings in minipots with vermiculite, and maintaining them in a growth chamber at 25°C with 14 h of light and 10 h of darkness per day. Plants immersed in sterile water were used as controls. For pathogenicity assays, the severity of disease symptoms was recorded at different times after inoculation, using a scale ranging from 1 (healthy plant) to 5 (dead plant) (14). Twenty plants were used for each treatment group.

To assay invasive growth of F. oxysporum strains, tomato fruits (cultivar Daniela) were washed under running tap water and surface sterilized by immersion for 5 min in ethanol. After air drying, the epidermis was punctured with a sterile pipette tip and 10 μl of a microconidial suspension (5 × 10⁸ ml⁻¹) was injected into the fruit tissue. Fruits were incubated at 28°C under conditions of 100% humidity. Colonization of the fruit tissue and formation of a mycelial mat on the fruit surface were determined at different time points after inoculation. All pathogenicity assays were performed at least twice, with similar results.

RT-PCR.

Five plants from each treatment group, inoculated as described above, were sampled after different time periods; total RNA was isolated from roots and lower parts of stems, and reverse transcription (RT)-PCR was performed as previously reported (14). Total RNA was treated with RNase-free DNase (Roche Molecular Biochemicals) and reverse transcribed into cDNA with murine leukemia virus reverse transcriptase (Gibco BRL, Paisley, United Kingdom), using the specific antisense primer 5′-AAGTTGGTGACGCTGTTGATG-3′. A volume of the RT reaction product was used for PCR amplification with the sense primer 5′-CCGATGCTGCTACTGCTATT-3′ and the antisense primer described above, both flanking the intron of pg5. PCR conditions were as follows: 40 cycles with denaturation at 94°C for 35 s, annealing at 50°C for 35 s, and extension at 72°C for 90 s. An initial denaturation step of 2 min at 94°C and a final elongation step at 72°C for 6 min were performed. Total genomic DNA of F. oxysporum was used as a template for PCR to compare the sizes of the amplified fragments with and without the intron. Aliquots of the PCR products were separated on 2% agarose gels, transferred to positively charged nylon membranes, and subjected to Southern hybridization analysis with the labeled pg5 probe.

Expression, purification, and characterization of the pg5 product in a heterologous bacterial system.

Total RNA from F. oxysporum strain 4287 grown on citrus pectin was retrotranscribed to cDNA by RT-PCR using the oligonucleotides 5′-CCATCCCTCATATGCGTGCCGGCAGCTGC-3′ (sense) and 5′-CTACCGCTCGAGAATAGTATTAACTGATAG-3′ (antisense). The amplified cDNA lacked the putative signal peptide and had an NdeI site and a XhoI site introduced at its 5′ and 3′ ends, respectively. To add a 10-histidine tag at the amino-terminal end of the PG5 protein, the amplified fragment was inserted into the pET-16b vector by digestion with NdeI and XhoI restriction enzymes and ligation. Escherichia coli expression strain BL21(DE3) (29) was transformed with the ligation mixture, and a single transformant colony was isolated. Bacterial cultures were grown on a large scale at 28°C, and the recombinant protein, extracted from cell lysates of 500-ml cultures, was purified under denaturing conditions by using nickel-nitrolotriacetic acid (Ni²⁺-NTA) resin according to the manufacturer's instructions (Qiagen Ltd., Dorking, Surrey, United Kingdom). Aliquots of each eluted fraction (500 μl) were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) in 12% (wt/vol) acrylamide resolving gels followed by staining with Coomassie brilliant blue. Fractions containing a single band of the overexpressed protein were pooled and renatured by using a gradient of decreasing concentrations of urea (8 to 0 M). EndoPG activity was measured by determining the release of reducing groups from PGA and expressed in nanokatals as previously described (12). Analysis of degradation products by thin-layer chromatography was carried out as reported elsewhere (18).

Transformation-mediated gene replacement and analysis of transformants.

The gene replacement vector pPg5::pANBlue3 was constructed as follows. A 0.6-kb HindIII-ApaI fragment containing 5′-flanking sequence and the first 312 bp of the pg5 coding region and a 0.5-kb BamHI-HindIII fragment containing the final 404 bp of the pg5 coding region and 3′-flanking sequence were cloned, in opposite directions, into the hygromycin resistance vector pAN7Blue3 (14). After linearization with HindIII, the final replacement vector contained 0.6 and 0.5 kb of colinear genomic DNA, with pAN7Blue3 replacing approximately 350 bp of the pg5 coding region. The linearized fragment was used to transform protoplasts of F. oxysporum strain 42-87 by a protocol described previously (14). To obtain protoplasts induced for pg5 expression, microconidia were germinated for 14 h in SM with 1% citrus pectin prior to protoplast formation (19). Transformants appeared after 5 days and were transferred to hygromycin plates and subjected to two consecutive rounds of single-spore isolation before being stored as microconidial suspensions at −80°C.

For analysis of PG activity, the wild-type strain as well as transformants lacking PG5 were grown in 400-ml volumes of SM with 1% citrus pectin as the sole carbon source for 36 h. Supernatants were harvested, fractionated by preparative isoelectric focusing, and assayed for PG activity as previously described (12). Active fractions were concentrated, dialyzed against water, and analyzed by SDS-PAGE and silver staining as previously described (12).

RESULTS AND DISCUSSION

Isolation and characterization of pg5

In the present study, a novel endoPG gene from F. oxysporum, pg5, was identified using a heterologous endoPG clone (enpg1) from Cryphonectria parasitica as the probe (17). The screening of an F. oxysporum f. sp. lycopersici isolate 4287 genomic library yielded three recombinant phage clones that were characterized by restriction and Southern hybridization analyses. Results indicated that the three clones encompassed different portions of the same genomic region. A 1.6-kb HindIII fragment that hybridized to the probe was subcloned in pBluescript KS(+) and sequenced on both strands. The pg5 coding region consists of an open reading frame of 1,083 bp, encoding a 361-amino-acid polypeptide with five potential N-glycosylation sites, interrupted by one intron of 49 bp whose position was confirmed by sequencing of the cDNA clone. The 5′-flanking region contains two TATA boxes, at the −50 and −64 positions. The N-terminal amino acid sequence of the predicted protein has the characteristic features of a signal peptide with a predicted cleavage site between residues 23 and 24 (32).

The putative mature protein has a calculated molecular mass of 35 kDa and a pI of 8.3. The amino acid sequence of the pg5 product shows significant homology to fungal endoPGs. Phylogenetically, PG5 is most closely related to Colletotrichum lindemuthianum PG1 and Cryphonectria parasitica ENPG1, with sequence identities being around 70% (Fig. 1). Remarkably, the only other endoPG characterized so far from F. oxysporum, PG1, has only 43% identity with PG5 and falls into a very distant class of endoPGs. The presence of multiple endoPGs belonging to distinct monophyletic groups has been reported for a number of fungal species, including the plant pathogen B. cinerea, which contains at least six endoPGs belonging to three different groups (34).

FIG. 1.

Multiple alignment of the deduced amino acid sequences of F. oxysporum f. sp. lycopersici (F.o.) strains PG5 (GenBank accession no. AF078156) and PG1 (U96456), Cryphonectria parasitica ENPG-1 (U49710), Colletotrichum lindemuthianum CLPG1 (X89370), Sclerotinia sclerotiorum (S.s.) PG1 (L12023), and Cochliobolus carbonum PGN1 (M37819). Identical amino acids are shaded.

To determine the copy number for pg5, Southern analysis with a gene-specific probe was performed on F. oxysporum f. sp. lycopersici genomic DNA digested with nine different restriction enzymes. The hybridization pattern observed was consistent with the presence of a single copy of pg5 in the genome (results not shown). Furthermore, the occurrence of pg5 in 15 F. oxysporum isolates belonging to the formae speciales ciceris, conglutinans, gladioli, lini, lycopersici, melonis, and niveum was studied by Southern blot analysis of genomic DNA digested with HindIII. A hybridizing band of 1.6 kb was present in all isolates except 24 ml and A15, both of which belong to forma specialis melonis, race 1,2.

Expression of pg5 in E. coli and characterization of the gene product.

A polyhistidine (His₁₀) affinity tag was attached to the N terminus of the mature PG5 protein by subcloning the cDNA without the signal peptide coding sequence into the pET16b expression vector. High levels of a 35-kDa protein were produced in E. coli BL21(DE3) transformed with the recombinant vector and grown in the presence of isopropyl-β-d-thiogalactopyranoside (IPTG) (Fig. 2A). This protein band was absent from E. coli transformants grown in medium lacking IPTG and from a strain transformed with the pET16b vector alone. No PG activity was detected in the recombinant protein samples, indicating that most of the protein was insoluble and found in inclusion bodies (results not shown). Therefore, Ni²⁺-NTA affinity chromatography of the polyhistidine-tagged PG5 protein was carried out under denaturing conditions, in 8 M urea, after which the protein was renatured in vitro under a gradient of decreasing urea concentration. This protocol produced an apparently homogeneous 35-kDa protein band, as determined by SDS-PAGE (Fig. 2A), that had PG activity. This fraction showed a total activity of 0.75 nkat, corresponding to a specific activity of 0.5 nkat/μg of protein (Fig. 2B). To determine the mode of action of PG5, the end products of enzymatic hydrolysis of PGA were analyzed by thin-layer chromatography. The presence of oligogalacturonides with intermediate degrees of polymerization together with di- and monogalacturonic acids was consistent with a classical endo mode of cleaving activity of PG5 (results not shown).

FIG. 2.

Purification of the recombinant pg5 gene product. (A) pg5 was overexpressed in E. coli grown in the presence of IPTG, and proteins were separated on an SDS–12% polyacrylamide gel and stained with Coomassie blue. Lane 1, molecular size markers; lanes 2 and 3, lysates of BL21(DE3) cells containing pET16b vector without (lane 2) or with (lane 3) the insert; lane 4, purified PG5 protein eluted from the Ni²⁺-NTA column. The molecular mass of the recombinant PG5 protein (in kilodaltons) is indicated on the left. (B) PG activities (specific activities) in bacterial lysates and in pure protein extract, expressed in nanokatals per microgram of protein. Column numbers correspond to lane numbers in panel A. One nanokatal was defined as the amount of enzyme that produced 1 nmol of galacturonic acid.

Expression of pg5 during saprophytic growth and during infection of tomato plants.

Expression of pg5 was determined by Northern hybridization analysis of total RNA obtained from mycelia of F. oxysporum f. sp. lycopersici grown for different periods of time in SM supplemented with different carbon sources. A single 1-kb transcript was detected in mycelia grown on citrus pectin, with maximum expression at 12 h of growth, whereas no transcript was detected in mycelia grown on glucose, PGA, or TVT (Fig. 3). Moreover, 1% d-galacturonic acid also induced pg5 expression, whereas 1% rhamnose did not (data not shown). The highly reduced accumulation of transcript levels in mycelia grown on 1% pectin plus 1% glucose suggested that glucose acts as a partial repressor of pg5 expression (data not shown). The two endoPGs of F. oxysporum, PG1 and PG5, show similar patterns of regulation during growth in axenic culture. Both are strongly induced by pectin and repressed by glucose.

FIG. 3.

Northern hybridization analysis of pg5 transcript accumulation in F. oxysporum f. sp. lycopersici mycelium grown for the indicated time periods (in hours) in synthetic medium containing 1% citrus pectin, 1% glucose, 0.5% PGA, or 2.5% TVT as the carbon source. (Lower panel) Total RNA blotted onto a nylon membrane and stained with 0.02% methylene blue. (Upper panel) Same filter, destained and hybridized with the digoxigenin-dUTP-labeled pg5 probe. The size of the transcript is indicated in kilobases.

To determine whether pg5 is expressed by F. oxysporum f. sp. lycopersici during infection of its host plant, RT-PCR with gene-specific primers was used to detect the presence of pg5 transcript in roots and lower stems of tomato plants at different time points after inoculation. As a control, total genomic DNA of F. oxysporum was used as a template for PCR to compare the sizes of the amplified fragments with and without the intron (447 and 398 bp, respectively). Southern analysis of the RT-PCR products with the pg5 probe showed that expression of the gene occurred only in roots of infected plants at the first time point sampled (3 days), coinciding with the initial stages of infection (Fig. 4). No amplified fragment hybridizing to pg5 was observed at any other time point or in the noninoculated control plants. Thus, the temporal expression pattern of pg5 during tomato plant infection differs considerably from that of pg1; whereas pg1 is expressed during the entire disease cycle, with expression levels increasing during the final disease stages (14), the pg5 transcript is detected only during the initial stages of infection. This suggests that PG5 may play a specific role during the early phase of interaction between the fungus and the plant host.

FIG. 4.

Southern hybridization analysis of RT-PCR products, showing the expression pattern of pg5 during infection of tomato plants by F. oxysporum f. sp. lycopersici. First-strand cDNAs were generated from total RNA isolated at the indicated time points (in hours) from roots and stems of uninfected or infected plants and used as templates for PCR with primers specific for pg5 (see Materials and Methods). Aliquots of the PCR products were electrophoresed on a 2% agarose gel, blotted onto a nylon membrane, and hybridized with the labeled pg5 probe. The position and size (in base pairs) of the pg5 fragment are indicated. The numbers represent days after inoculation. gDNA, genomic DNA.

Targeted replacement of pg5

Mutants carrying a disrupted copy of the pg5 gene were generated by a one-step gene replacement protocol. Vector pPg5::pANBlue3 was constructed, as shown in Fig. 5A and B, by replacing a 0.3-kb ApaI-BamHI fragment within the pg5 coding region with the hygromycin B resistance plasmid pAN7Blue3. A linear HindIII restriction fragment containing the disrupted pg5 gene was used to transform F. oxysporum f. sp. lycopersici strain 4287, which is highly virulent to tomato plants. Eight hygromycin-resistant transformants were selected, and their genomic DNAs were isolated, digested with HindIII, and subjected to Southern analysis with the pg5 probe. As shown in Fig. 5C, in two of the transformants, Dpg5-1 and Dpg5-2, the 1.6-kb HindIII fragment corresponding to the wild-type pg5 allele was replaced by a larger, 6.0-kb fragment. This indicates that both transformants contain a single copy of the replacement vector integrated by double homologous recombination, thereby generating a disrupted copy of pg5. Consistent with gene replacement, the same single fragment hybridized with a probe for the hygromycin resistance gene (data not shown). The rest of the transformants, exemplified by NDpg5-3, contained the wild-type pg5 allele together with an additional hybridizing fragment, indicative of ectopic insertion of the replacement vector. Northern analysis of pectin-grown mycelia of the mutant strains Dpg5-1 and Dpg5-2, as well as the wild-type strain and the ectopic transformant NDpg5-3, confirmed the absence of the pg5 transcript in the two mutants, while this transcript was readily detected in the other strains (Fig. 6). Transcript levels of the exo- and endoPG pgx4 and pg1 genes, respectively, were comparable in pg5-mutants and in the control strains.

FIG. 5.

Targeted replacement of the F. oxysporum pg5 gene. (A) Physical map of the pg5 genomic region. The pg5 coding region is shown as a black arrow orientated with the open reading frame. Restriction enzymes are abbreviated as follows: A, ApaI; B, BamHI; and H, HindIII. (B) Gene replacement vector pPg5::pANBlue3. (C) Analysis of transformants by Southern blotting. Genomic DNAs from the transformants Dpg5-1, Dpg5-2, and NDpg5-3 (lanes 1 to 3, respectively) and wild-type strain 4287 (lane 4) were digested with HindIII, separated in a 0.7% agarose gel, blotted onto a nylon membrane, and hybridized with a pg5 probe. Sizes of hybridizing bands are indicated in kilobases.

FIG. 6.

Northern hybridization analysis of F. oxysporum f. sp. lycopersici wild-type strain 4287 (lane 5), replacement mutants Dpg5-1 and Dpg5-2 (lanes 1 and 2, respectively), and ectopic transformants NDpg5-3 and NDpg5-4 (lanes 3 and 4, respectively). (Upper panels) Total RNA from mycelia grown in synthetic medium containing 0.5% PGA (sodium salt) was hybridized with probes of the endoPG pg5 and pg1 genes and of the exoPG pgx4 gene. (Lower panel) Total RNA was blotted onto a nylon membrane and stained with 0.02% methylene blue. ∗, sample was not included in the hybridization with the pg1 probe.

Inactivation of pg5 did not cause any measurable growth reduction in SM with citrus pectin as the sole carbon source. Both hyphal morphology and the extent of microconidial production were indistinguishable from those of the wild type. Total extracellular PG activities of the wild-type strain 4287, the pg5 replacement mutants, and the ectopic-insertion transformant did not differ significantly (data not shown). To check for differences in PG isozymes, supernatants of the wild-type strain and pg5 mutant Dpg5-1 grown in SM with citrus pectin were subjected to preparative isoelectric focusing, and the fractions obtained were analyzed for PG activity and by SDS-PAGE. Both in the wild type and in the mutant, two major activity peaks were detected, in the fractions with pH values of 6.8 and 8.5. When proteins from these fractions were analyzed by SDS-PAGE, no significant differences in banding patterns were observed. In both strains, two closely spaced (35- and 37.5-kDa) bands corresponding to PG1 isoforms (12, 14) were the dominant protein bands (data not shown). PG1 has been shown to be the major endoPG secreted by F. oxysporum, and mutants lacking a functional copy of the pg1 gene exhibit dramatically reduced PG activity and impairment of saprophytic growth on pectic substrates (12, 14). Conversely, PG5 appears to be a minor PG isozyme, at least under the growth conditions used in our study. Therefore, its activity may be masked by the more abundant PG1, since the pIs of the two enzymes are probably similar (7 to 8 for the PG1 isoforms [12] and 8.3 for PG5). From these results, we conclude that pg5 does not contribute measurably to total extracellular PG activity in F. oxysporum under the conditions used in this study. Our results are similar to those reported for the bean pathogen Colletotrichum lindemuthianum, whose two endoPG genes, CLPG1 and CLPG2, show different expression patterns. CLPG1 is the major extracellular endoPG both during culture and in planta, whereas CLPG2 is transiently produced only during early stages of growth (6).

Infection assays on tomato plants were performed to determine the effect of pg5 inactivation on pathogenicity. Both pg5 mutants were as virulent as the wild type and an ectopic-insertion strain. The patterns of colonization of the host plant, as determined by the presence of the fungus in roots and stems of inoculated plants, were identical in the mutants and the wild-type strain (results not shown). To test whether pg5 contributes to invasive growth of F. oxysporum on living host plant tissue, tomato fruits were inoculated by injecting a microconidial suspension into the fruit tissue. After 5 days of incubation, the pg5 mutants had colonized and rotted the fruit tissue surrounding the site of inoculation to the same extent as the wild-type strain and an ectopic-insertion transformant, forming a dense mycelial mat on the surface of the fruit (results not shown). We conclude that pg5 is not required for pathogenicity of F. oxysporum on tomato plants or for invasive growth on living plant tissue.

It is currently unknown whether different endoPG genes in a given fungal species are functionally redundant or whether each of them performs specific biological tasks. The fact that pg5 is expressed only during the first stages of infection suggested that it might play a defined role in the pathogenicity process. However, since no loss or reduction of virulence was detected in pg5 mutants, we concluded that PG5 does not significantly contribute to the pathogenicity of F. oxysporum.