Quantification in Soil and the Rhizosphere of the Nematophagous Fungus Verticillium chlamydosporium by Competitive PCR and Comparison with Selective Plating

Abstract

A competitive PCR (cPCR) assay was developed to quantify the nematophagous fungus Verticillium chlamydosporium in soil. A γ-irradiated soil was seeded with different numbers of chlamydospores from V. chlamydosporium isolate 10, and samples were obtained at time intervals of up to 8 weeks. Samples were analyzed by cPCR and by plating onto a semiselective medium. The results suggested that saprophytic V. chlamydosporium growth did occur in soil and that the two methods detected different phases of growth. The first stage of growth, DNA replication, was demonstrated by the rapid increase in cPCR estimates, and the presumed carrying capacity (PCC) of the soil was reached after only 1 week of incubation. The second stage, an increase in fungal propagules presumably due to cell division, sporulation, and hyphal fragmentation, was indicated by a less rapid increase in CFU, and 3 weeks was required to reach the PCC. Experiments with field soil revealed that saprophytic fungal growth was limited, presumably due to competition from the indigenous soil microflora, and that the PCR results were less variable than the equivalent plate count results. In addition, the limit of detection of V. chlamydosporium in field soil was lower than that in γ-irradiated soil, suggesting that there was a background population of the fungus in the field, although the level was below the limit of detection. Tomatoes were infected with the root knot nematode (RKN) or the potato cyst nematode (PCN) along with a PCN-derived isolate of the fungus (V. chlamydosporium isolate Jersey). Increases in fungal growth were observed in the rhizosphere of PCN-infested plants but not in the rhizosphere of RKN-infested plants after 14 weeks using cPCR. In this paper we describe for the first time PCR-based quantification of a fungal biological control agent for nematodes in soil and the rhizosphere, and we provide evidence for nematode host specificity that is highly relevant to the biological control efficacy of this fungus.

The soil fungus Verticillium chlamydosporium has the potential to be used as a biological control agent for some plant-pathogenic nematodes belonging to the genera Meloidogyne, Globodera, and Heterodera (2, 4, 16), although much about the diversity in the genus Verticillium and the extent of nematode host preference by different fungal isolates is not known. To understand this potential, it is necessary to monitor the growth, spread, and persistence of the fungus in soil and on plant roots with and without nematodes. Quantification of filamentous nematophagous fungi, such as V. chlamydosporium, is notoriously difficult (15). Unlike many microorganisms, these fungi are not composed of units that are approximately the same size and have the same genetic content. They have several different life stages, which may include hyphae with various numbers of nuclei as well as unicellular conidiospores and, in the case of V. chlamydosporium, multicellular chlamydospores. Several methods have been used to quantify such fungi in soil; these methods include estimating the number of nematodes parasitized (15), physically extracting spores (5), most-probable-number techniques (8), using selective media (17), and enzyme-linked immunosorbent assays (3). However, all of the techniques examined have met with only limited success. Sensitive techniques for identifying groups of organisms that rely on recognition of specific DNA sequences offer advantages if they can be made quantitative. Methods to extract DNA from the general microbial community based on bead beating have been developed (6). Such methods are suitable for extracting DNA from fungal spores and outperform methods such as sonication, freeze-thaw lysis, and grinding in liquid nitrogen (11).

The PCR is a powerful tool for amplifying and detecting specific nucleic acid molecules present at low levels (9). However, quantification of PCR products by conventional PCR is limited, because during exponential amplification of the template, small variations in amplification efficiency can drastically change the yield of the PCR product, resulting in a nonquantitative assay. For this reason, the use of PCR in combination with most-probable-number techniques is also limited. One method that accounts for variations in amplification efficiency is competitive PCR (cPCR). Coamplification of the molecule of interest with a known amount of competitor molecules bearing primer sites identical to those of the target allows reproducible quantification of templates (19). Competitor molecules can be designed so that they are a different size than the target fragment but retain the original primer binding sites. Such molecules can be engineered from completely foreign DNA or can be obtained by PCR amplification of nontarget DNA with target primers at low annealing temperatures (10, 21, 23). Another approach is to alter the size of the original PCR product by either adding or deleting a fragment between the primer binding sites.

In this paper, we describe development of a cPCR assay for V. chlamydosporium in soil and the rhizosphere and compare the results obtained with this assay to the results obtained by plate counting on selective agar. We also used the method to investigate the host nematode preference of V. chlamydosporium isolated from potato cyst nematode (PCN) with tomato plants infested with PCN or root knot nematode (RKN).

MATERIALS AND METHODS

Soil properties and fungal inoculum.

Soil samples were taken from Sawyers Field at Rothamsted Experimental Station, Hertfordshire, United Kingdom. The soil at this site is a flinty silty clay loam of the Batcombe-Carstens Series, a Typic Paleudalf (USDA 1992). V. chlamydosporium isolates 10 and Jersey were grown on 1.7% corn meal agar (Oxoid, Basingstoke, United Kingdom) at 25°C prior to use. A chlamydospore inoculum was prepared in a sand-milled barley mixture as previously described (7). Soil samples were inoculated with known numbers of chlamydospores and mixed thoroughly.

Germination of tomato seeds.

Tomato seeds were planted in soil and allowed to germinate for 2 weeks; seedlings were then transplanted into fungus-inoculated soil. The experiments were conducted at 20°C in glasshouses, and the pots were watered daily.

Culturing and application of nematodes.

Meloidogyne incognita (RKN) was cultured on aubergine at 25°C, and infective juveniles were harvested by standard methods (14). Globodera pallida (PCN) was cultured on potato cultivar Maris Piper; cysts were harvested by using a fluidizing column (22), and infective juveniles were extracted from the cysts by standard procedures (14). A total of 1,000 second-stage juveniles of one of the nematodes were added to each pot in holes made in four places around the root system of a plant that had been grown for 2 weeks after transplantation.

Experimental design: inoculation of V. chlamydosporium into γ-irradiated soil and field soil at various concentrations.

For each treatment, three replicate pots containing 50 g of γ-irradiated or unsterilized field soil were prepared with inoculum concentrations ranging from 10¹ to 10⁵ V. chlamydosporium 10 chlamydospores g of soil⁻¹. Partially sterilized soil samples were obtained by γ-irradiating 5-kg batches in sealed plastic bags by using a dose of 1 megarad (Irradiated Products Ltd., Swindon, United Kingdom). These samples were stored at ambient temperature for several years; during this time the concentration of surviving residual bacteria and fungi stabilized at 10² CFU g of soil⁻¹. Samples were taken at different times after inoculation with V. chlamydosporium, and experiments were scheduled for up to 8 weeks. In these experiments, the soil moisture level was maintained at 50% of the water-holding capacity by covering pots with aluminum foil and placing the pots in trays of moist sterilized sand with a standardized amount of water added. After DNA extraction or plating, the moisture content was calculated for each soil sample and used to adjust the final results.

Inoculation of V. chlamydosporium into soil containing nematode-infested and healthy tomato plants.

For each treatment, three replicate pots of γ-irradiated soil were prepared by adding 5,000 V. chlamydosporium Jersey chlamydospores g of soil⁻¹. Experiments were scheduled for up to 14 weeks, and tomato fruits were removed from the plants after 10 weeks in order to prevent senescence. The following treatments were used: V. chlamydosporium Jersey in soil alone; V. chlamydosporium Jersey in soil with tomato plants; V. chlamydosporium Jersey with PCN and tomato plants; and V. chlamydosporium Jersey with RKN and tomato plants. Control treatments were identical except that the fungus was not added.

For both experiments at each sampling time, soil samples were removed and used for DNA extraction and/or dilution plating onto a V. chlamydosporium semiselective medium as described below.

V. chlamydosporium semiselective medium.

A corn meal agar-based medium containing rose bengal, fungicides, and antibiotics (7) was used to reisolate and enumerate V. chlamydosporium from soil and root samples. Soil samples (1 g) were suspended in a 0.05% sterile agar solution, and a dilution series was prepared. Root samples (1 g) were homogenized in 9 ml of a 0.05% sterile agar solution by using a pestle and mortar, and a dilution series was prepared. Aliquots (0.2 ml) of the 10⁻² and 10⁻³ dilutions were plated and incubated at 24°C for 2 weeks. V. chlamydosporium colonies were counted, and the numbers of CFU per gram of soil were calculated.

PCR conditions.

DNA samples extracted from soil were tested with a primer set designed to amplify a fragment of the V. chlamydosporium β-tubulin gene (12). These primers are specific for V. chlamydosporium and do not amplify DNA from other species of Verticillium or from members of other fungal genera (13). They were utilized to detect V. chlamydosporium in soil and root samples. The primer sequences were as follows: tub1f, 5′-TTT GCA GTA TCT CAG TGT TC-3; and tub1r, 5′-ATG CAA GAA AGC CTT GCG AC-3′. All PCR products were electrophoresed on 3% agarose gels (Helena Bioscience, Sunderland, United Kingdom) and stained with ethidium bromide (0.5 μg ml⁻¹). Product sizes were estimated by comparison with a 123-bp DNA ladder (Gibco BRL, Paisley, United Kingdom). The PCR mixtures (20 μl) contained 1 μl of sample, each primer at a concentration of 0.1 μM, 1× PCR buffer (1.5 mM Mg²⁺; Roche Diagnostics, Lewes, United Kingdom), 1 mM MgCl₂, each deoxynucleoside triphosphate (Roche) at a concentration of 0.2 mM, and 1 U of Taq polymerase (Roche). The thermocycling conditions were as follows: 95°C for 1 min, followed by 45 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 2 min and a final extension step consisting of 72°C for 5 min.

Competitor construction.

To construct a competitor for the tub1r-tub1f primer set, the 270-bp PCR product of V. chlamydosporium 10 obtained by using these primers was ligated into pGEM-T (Promega, Southampton, United Kingdom) and cloned in SURE competent cells (Stratagene, Amsterdam, The Netherlands). Plasmids were extracted from transformed colonies with a Qiagen Midi kit (Qiagen, Crawley, United Kingdom) and digested with BamHI (Roche). There are no BamHI restriction sites in pGEM-T, but one is present in the V. chlamydosporium 10 tub1f-tub1r PCR product. The digest was treated with phenol-chloroform and then ligated to a heat-treated Sau3AI digest of pBR322. Colonies arising from this ligation were screened by PCR using primers tub1f and tub1r, and one colony with a 75-bp insert was selected. This provided a PCR product that could be distinguished from the 270-bp V. chlamydosporium 10 PCR product and had similar amplification kinetics. The exact size of the insert was determined by sequencing the competitor using ABI PRISM dye terminator cycle sequencing Ready Reaction kits (Perkin-Elmer, Warrington, United Kingdom).

Soil DNA extraction.

An UltraClean soil DNA isolation kit (Mo Bio Laboratories, Solana Beach, Calif.) was used as described by the manufacturer, with minor modifications. The protocol used involved 0.25-g soil samples which were bead beaten with a Mikro-Dismembrator II (B. Braun, Meslungen, Germany) for 15 min at an amplitude of 12 mm. The DNA was resuspended in 200 μl of 10 mM Tris (pH 8), and 1 μl of each extract was used as a PCR template.

cPCR.

Each cPCR assay consisted of a series of reactions in which the reaction mixtures contained different amounts of competitor DNA and a constant amount of sample DNA extracted from soil or roots. Control reaction mixtures contained sample DNA template, competitor DNA, V. chlamydosporium 10 genomic DNA, or no DNA. Products were separated on agarose gels (Fig. 1), which were digitized with the EagleEye II system (Stratagene), and band intensities were quantified using the GelDoc 2000 system (Bio-Rad, Hemel Hempstead, United Kingdom) after background fluorescence had been subtracted. Differences in competitor and target PCR product lengths were accounted for by multiplying the competitor band intensity by the ratio of the product sizes (270:345).

FIG. 1.

Example of a cPCR gel. A titration series of competitor molecules was prepared with a standard amount of template DNA from soil samples.

To relate the amounts of template DNA in PCR mixtures to a known number of chlamydospores, a standard curve was produced for the tub1r-tub1f primer set by adding known amounts of chlamydospores to γ-irradiated Sawyers Field soil before DNA was extracted from the samples. The log₁₀ number of competitors added to each of a series of reaction mixtures was plotted against the log₁₀ ratio of the target product band intensity to competitor product band intensity. The number of target molecules in each sample was then estimated by determining the value at which a straight line through the points crossed the x axis [log₁₀ (target band intensity/competitor band intensity) = 0] (Fig. 2, upper panel). It should be noted that log₁₀ (target band intensity/competitor band intensity) = 0 is equivalent to a target/competitor ratio of 1. The values were plotted against the appropriate log₁₀ number of chlamydospores per gram of soil to form the standard curve (Fig. 2, lower panel). The assays were performed in triplicate, and the results were used to calibrate experimental assays and demonstrate that they were quantitative. In the analysis of experimental assays the log₁₀ number of competitors added to a reaction mixture was adjusted to represent the corresponding number of chlamydospores by using the standard curve equation. Each value was further altered to take into account the wet weight of the sample. The resulting values were plotted against the log₁₀ ratio of target band intensity to competitor band intensity in order to estimate the number of chlamydospores (or equivalent gene copies) in each sample (Fig. 3).

FIG. 2.

Data set and standard curve. (Upper panel) Example of a data set used to make a single point on the standard curve. Ratios of the log₁₀ cPCR target intensity to the log₁₀ cPCR product intensity obtained with primers tub1f and tub1r are plotted against the log₁₀ numbers of competitor molecules in assay mixtures. (Lower panel) Standard curve for primers tub1f and tub1r. The log₁₀ number of chlamydospores in each extract is plotted against the log₁₀ number of competitor copies required for equivalence. t, target; c, competitor.

FIG. 3.

Example of an experimental data set. Ratios of the log₁₀ cPCR target intensity to the log₁₀ cPCR product intensity obtained with primers tub1f and tub1r are plotted against log₁₀ adjusted numbers of competitor molecules (using the standard curve) in sets of assays.

RESULTS

cPCR assays.

We found that increases in the number of competitor molecules were directly proportional to increases in the amount of cPCR product and inversely proportional to the amounts of V. chlamydosporium PCR product (Fig. 1).

Standard curve for cPCR.

The standard curve for estimating V. chlamydosporium populations showed that there was a positive linear relationship between the log₁₀ amount of V. chlamydosporium 10 chlamydospores added to soil and the concentration of competitor molecules required for equivalence. The coefficient of determination (r²) was 0.99 for the regression curve (Fig. 2, lower panel). The standard curve was used to calculate the number of chlamydospores corresponding to the log₁₀ number of competitors added to a reaction mixture. Seeding soil with identical amounts of V. chlamydosporium 10 chlamydospores and V. chlamydosporium Jersey chlamydospores resulted in equivalent estimates of the number of β-tubulin gene copies by cPCR. For this reason it was not necessary to create a separate standard curve for experiments in which V. chlamydosporium Jersey was used rather than V. chlamydosporium 10.

Determining growth dynamics of V. chlamydosporium 10 after inoculation into γ-irradiated soil.

V. chlamydosporium 10 was undetectable in γ-irradiated control samples, suggesting that any indigenous population of this fungus was killed in the irradiation process and that genetic material from the V. chlamydosporium cells had been degraded by the surviving residual microflora. Both plate count and cPCR results showed that V. chlamydosporium 10 was able to proliferate in the γ-irradiated soil, but the two methods gave different results (Fig. 4). The cPCR estimates increased for all inoculation rates until a plateau of approximately 10⁵ gene copies g of soil⁻¹ was reached after 1 week of incubation (Fig. 4, lower panel). The plate count data showed that the fungal populations increased but at a lower rate, reaching a plateau of about 10⁵ CFU g of soil⁻¹ after 3 weeks (Fig. 4, upper panel).

FIG. 4.

Estimation of V. chlamydosporium populations in γ-irradiated field soil. Soil was inoculated with 10 chlamydospores g of soil⁻¹ (⋄), 100 chlamydospores g of soil⁻¹ (□), 1,000 chlamydospores g of soil⁻¹ (▵), 10,000 chlamydospores g of soil⁻¹ (×), or 100,000 chlamydospores g of soil⁻¹ (○). Upper panel, plate counts; lower panel, cPCR estimates. The dotted line indicates the limit of detection by cPCR. The dashed lines are lines that are extrapolated back to the starting amounts of inoculum.

Determining the limits of detection in γ-irradiated soil using cPCR.

The theoretical limit of detection in a PCR is one molecule. The PCR assay had an observed limit of detection of 10² V. chlamydosporium 10 chlamydospores g of γ-irradiated soil⁻¹, although bands on gels were too faint to allow quantification. Chlamydospores are composed of 10 to 20 haploid cells. Thus, it can be calculated that each reaction mixture contained the equivalent of the DNA from 0.125 chlamydospore or 1.25 to 2.5 gene copies. The equation provided by the standard curve could also be used to estimate the limit of detection of V. chlamydosporium β-tubulin genes in soil by determining the value of x when y = 0. From the standard curve of y = 1.6398 x − 3.575, we calculated that x = 2.18, which could also be expressed as the inverse log₁₀ of 151 gene copies when y = 0. This is comparable to the observed experimental limit.

Determining the limits of detection in γ-irradiated soil using plate counting.

The theoretical limit of detection of a propagule by plate counting is 1 CFU per plate. The actual limit of detection was found to be 10³ chlamydospores g⁻¹ in γ-irradiated soil. Plates were each inoculated with 0.2 ml of a 100-fold dilution of 1 g of inoculated soil; thus, the actual number of chlamydospores added per plate when the inoculum was 10³ chlamydospores g of γ-irradiated soil⁻¹ was approximately two.

Investigating growth dynamics of V. chlamydosporium 10 after inoculation into field soil.

Both methods showed that the V. chlamydosporium populations in field soil did not reach the presumed carrying capacity (PCC) observed in γ-irradiated soil (Fig. 5). Plate counting indicated that at the lower inoculation rates (10² and 10³ chlamydospores g⁻¹) chlamydospores proliferated during the experiment, whereas when 10⁴ chlamydospores g of soil⁻¹ was used, the number of chlamydospores increased only 10-fold after inoculation (Fig. 5, upper panel). The highest inoculation rate only slightly exceeded the original application rate (Fig. 5, upper panel), and the initial inoculum densities, as determined by cPCR, remained constant throughout the experiment (Fig. 5, lower panel). Values for both methods fluctuated throughout the experiment, but the plate count data at a given sampling time were more variable, as indicated by the larger standard errors.

FIG. 5.

Estimation of V. chlamydosporium populations in field soil. Soil was inoculated with 100 chlamydospores g of soil⁻¹ (⋄), 1,000 chlamydospores g of soil⁻¹ (□), 10,000 chlamydospores g of soil⁻¹ (▵), or 100,000 chlamydospores g of soil⁻¹ (○). Upper panel, plate counts; lower panel, cPCR estimates.

Investigating growth dynamics of V. chlamydosporium Jersey after inoculation into soil containing healthy and nematode-infested tomato plants.

V. chlamydosporium Jersey was inoculated into γ-irradiated soil, as well as into soil containing tomatoes with or without PCN or RKN. The changes in the population of the fungus were observed over a 14-week period.

Changes in root morphology were detected after 10 weeks of incubation (8 weeks after nematodes were added), when slight galling was visible on roots of RKN-treated plant roots in both V. chlamydosporium Jersey-treated and control pots. After 14 weeks of incubation (12 weeks after nematodes were added), major galling was present on the roots of RKN-treated plants, regardless of the presence of V. chlamydosporium Jersey. However, with PCN no cysts were observed after 10 weeks of incubation. After 14 weeks, cysts were present on control plants, but there were no cysts in V. chlamydosporium Jersey-treated pots.

Dynamics of V. chlamydosporium Jersey growth in soil.

V. chlamydosporium Jersey was inoculated at a rate of 5,000 chlamydospores g of soil⁻¹ and was detected by both plate counting and cPCR at this level after 24 h of incubation. After 7 weeks of incubation, after nematodes had been present for only 5 weeks, plate counts indicated that the concentration of the fungus had increased to the expected carrying capacity of the soil, which was consistent with results obtained with V. chlamydosporium 10 in γ-irradiated soil.

There were no differences among the numbers of CFU obtained from different treatments after 7, 10, or 14 weeks of incubation, indicating that neither the nematodes nor the tomato plants influenced fungal growth in soil. No cPCR were performed with the soil samples from pots containing tomatoes or nematodes.

Dynamics of V. chlamydosporium Jersey growth on roots.

There were no differences among the numbers of CFU obtained from different treatments after 7, 10, or 14 weeks of incubation, indicating that nematodes did not influence fungal growth, as estimated by the numbers of propagules in the rhizosphere at these times (Fig. 6, upper panel).

FIG. 6.

Estimation of V. chlamydosporium populations on tomato roots. Soil was inoculated with 5,000 chlamydospores g of soil⁻¹, and tomatoes were inoculated with PCN (▵), RKN (□), or no nematodes (⋄). Upper panel, plate counts; lower panel, cPCR estimates.

Results for samples tested by cPCR showed that after 10 weeks of incubation the estimated fungal populations for all treatments were equivalent. However, after 14 weeks, the estimated population from healthy plants had increased 10-fold and the population from PCN-infested plants had increased 100-fold. Both of the values were clearly greater than those obtained with RKN-infested plants, which decreased after week 10 (Fig. 6, lower panel).

DISCUSSION

The use of a standard curve for cPCR is vital when working with a complex substrate such as soil, in order to account for any anomalies and biases in the assay. In the case of soil, phenomena such as binding of DNA to soil particles and the presence of humic substances in DNA extracts can alter the efficiency of amplification of templates during PCR (18, 20). However, the standard curve showed that there was a linear relationship between the amplification efficiencies for template and competitor molecules, indicating that there was consistency in soil DNA extraction and demonstrating that the assay is quantitative (Fig. 2). Although the standard curve was based on V. chlamydosporium 10 chlamydospores, a test extraction performed with known amounts of V. chlamydosporium Jersey chlamydospores in soil resulted in a population estimate that was the same as that obtained for the equivalent amount of V. chlamydosporium 10 chlamydospores, indicating that the standard curve was appropriate for use with V. chlamydosporium Jersey-treated samples.

Multicellular chlamydospores were used as the fungal unit for soil inoculation in experiments rather than hyphal fragments or single-celled conidiospores primarily because chlamydospores are the preferred form of inoculum for field application of the fungus. Hyphal fragments are inconsistent in terms of size and DNA content, while both conidia and hyphae have low survival rates in soil and are subject to fungistasis when they are added to soil without a supplementary energy source. Chlamydospores have sufficient food reserves to enable the fungus to become established without the addition of other energy sources (17). Chlamydospores contain more than one cell; thus, cPCR results are related directly to the number of gene copies per gram of soil rather than to the number of chlamydospores. This reflects the number of chlamydospores in the initial stage which subsequently develop into hyphal fragments and conidia.

Growth of V. chlamydosporium 10 and Jersey in soil was estimated by cPCR and semiselective medium plate counting. The two methods measure different aspects of the growth dynamics of the fungi in soil. Plate counting measures the abundance of viable fungal propagules, but due to the variable nature of fungal material (hyphal fragments, conidia, chlamydospores) it is impossible to know what type of fungal unit a given colony is related to, resulting in an estimate of fungal biomass that is difficult to interpret. In contrast, the cPCR method estimates the numbers of gene copies in soil DNA extracts, which gives a better indication of fungal biomass, since increases in biomass can be correlated with increases in DNA content. However, DNA from moribund material can also be amplified, and this could give misleading results.

When different concentrations of chlamydospores were added to γ-irradiated soil and incubated for up to 8 weeks, cPCR data indicated that the fungus grew rapidly at all inoculation levels until the PCC (10⁵ chlamydospores g of soil⁻¹ or equivalent number of β-tubulin gene copies) was reached after 1 week for all inoculation rates (Fig. 4, lower panel). The plate counts, however, did not reach a plateau until 3 weeks (Fig. 4, upper panel). These results illustrate the growth of a filamentous fungus. Initially, DNA replication occurs, as demonstrated by the rapid increase in the cPCR estimates during the first week of incubation. This is followed by nuclear division and hyphal growth and then by sporulation, a slower process that was indicated by the more gradual increases in colony counts for a given inoculation level.

Inoculation of various concentrations of V. chlamydosporium 10 into untreated field soil resulted in no increase in the number of gene copies detected by PCR, and the PCC in γ-irradiated soil was not reached with inoculation rates of less than 10⁵ chlamydospores g of soil⁻¹ (Fig. 5, lower panel). This suggests that competition between the added fungus and other soil microorganisms for nutrients in the soil limited the size of the V. chlamydosporium 10 population. Nevertheless, the fungus remained viable, as indicated by plate counting, suggesting that it is a competent saprophyte that is able to survive in soil in the absence of plants (Fig. 5, upper panel). This is an important consideration for a biological control fungus used against nematodes, which may need to survive the intercropping season, when plants as well as nematode hosts are absent. It was not clear whether V. chlamydosporium 10 was actively growing or if spore germination was repressed, but the observed increase in viable counts at the lowest inoculation rate indicates that some growth and probably some conidiospore production occurred. Since conidiospores are single celled, the observation that the number of gene copies remained constant whereas the number of propagules increased is consistent with the occurrence of chlamydospore germination, limited growth, and conidiospore production.

An unexpected observation was that quantification of V. chlamydosporium by cPCR was 10-fold more sensitive immediately after inoculation in untreated field soil than in γ-irradiated soil (Fig. 4, lower panel, and Fig. 5, lower panel). In field soil, 10² chlamydospores g⁻¹ were detected and quantified; however, in γ-irradiated soil 10² chlamydospores g⁻¹ were detected only in some DNA extracts and could not be quantified, suggesting that 10² chlamydospores g⁻¹ is the limit of detection, which agrees with the theoretical calculations. This suggests that there is a background population of V. chlamyd osporium or V. chlamydosporium DNA in the field soil, which is not present in the γ-irradiated stored soil and is too small to be detected alone since uninoculated control samples gave no PCR products.

The detection limits of the plate count method were similar with γ-irradiated and untreated field soil (Fig. 4, upper panel, and Fig. 5, upper panel). The fungus was not able to grow as rapidly or at the same level in untreated soil as in γ-irradiated soil. This may have been a consequence of chlamydospores dying or not germinating or a consequence of growth being inhibited due to competitive soil microflora. Detection by the PCR system was not impeded by the presence of other organisms due to the highly selective nature of the primers. The initial observations with untreated soil were made immediately after the soil was inoculated with chlamydospores that had been checked for viability. However, it is possible that at later sampling times some DNA was extracted from moribund fungal material, although the plate count data and cPCR estimates were not dramatically different.

The results of the experiments involving V. chlamydosporium Jersey indicated that there was no difference in the number of fungal CFU in bulk soil for any of the treatments, suggesting that the rhizosphere has little effect on the growth of V. chlamydosporium Jersey in soil. This is perhaps surprising as relatively small pots (diameter, 9 cm) were used, which by the end of the experiment had been completely colonized by plants. This suggests that the root system influences the soil only very close to the roots and implies that there is an intimate association between V. chlamydosporium and tomato roots. However, an alternative explanation is that nutrient-rich γ-irradiated soil was used, which could have resulted in root exudation which had relatively little additional effect on fungal growth.

An increase in the fungal population on uninfected roots was observed after 10 to 14 weeks in pots when cPCR was used but not when plate counting was used (Fig. 6). One explanation for this is that removal of tomato fruits from the plants after 10 weeks of incubation could have resulted in a major redirection of nutrients to other plant parts, including the root system. Recent work indicated that fruit removal resulted in reallocation of the resources that would have gone into fruit maturation (1). Consequently, increased root exudation could have increased fungal populations and stimulated multicellular chlamydospore production, which would explain why the estimate based on number of gene copies was larger than the estimate based on number of CFU.

Even higher levels of fungus were detected on PCN-treated tomato roots by cPCR after 14 weeks of incubation (Fig. 6, lower panel). This could have been because female nematodes emerge on the root system after about 14 weeks, and it is known that the fungus responds by increasing chlamydospore production (16). However, in this case the results were associated with nematode parasitism. Fruit removal in this system would probably facilitate nematode development as well as increase general root exudation. Both of these effects would ultimately increase the fungal population, as PCN is a favorable host for V. chlamydosporium Jersey. Increases in the V. chlamydosporium Jersey population on PCN-infested roots indicate that there is a parasite-host interaction between the fungus and the nematode, but attack is normally associated with parasitism of females on the root surface. However, as no cysts were observed in treated pots, it is possible that fungal attack of newly emerged females was successful enough to prevent cyst formation.

In pots containing V. chlamydosporium Jersey and RKN, the populations of the fungus from tomato roots determined by plate counting remained constant throughout the experiment (Fig. 6, upper panel). However, the V. chlamydosporium Jersey populations determined by cPCR analysis decreased from week 10 to week 14 (Fig. 6, lower panel). One reason for this may be that changes in root physiology caused by root galling resulted in fungal DNA extraction being suboptimal, revealing a limitation of this DNA extraction method for heavily galled root systems. This could be due to reduced bead-beating disruption of roots or to inhibition of the PCR by increased release of humic material from galled roots. However, the lack of growth indicates that V. chlamydosporium Jersey is a poor parasite of RKN.

It is clear that the presence of PCN in a rhizosphere previously colonized by V. chlamydosporium Jersey increases the fungal population as determined by the number of β-tubulin gene copies. However, it is not clear whether the increase is a saprophytic response to increased root exudation associated with nematode parasitism or is due to the action of a successful biological control agent against its host. Evidence that the increase is a parasitic response to PCN and not RKN is provided by the observation that cyst formation did not occur in PCN-infested pots but extensive root gall formation did occur in RKN-infested pots.

If nematode-induced root exudation is initiated by root-attacking juveniles, then an earlier increase in the sizes of the fungal populations for all root samples from pots infested with nematodes would have been expected, but this did not occur. The growth burst after 14 weeks observed for V. chlamydosporium Jersey on G. pallida-infested tomato roots may be explained by a combination of increases in saprophytic growth and parasitic growth.

In summary, the cPCR method is a more sensitive detection method than the semiselective medium method, and it seems that in the sterile system both methods worked up to their theoretical limits. However, in field soil, plate counting was a less reliable tool for quantification of V. chlamydosporium. This may have been due to a suboptimal germination rate of the fungal inoculum in the soil. It is not clear whether the apparently stable population density of the fungus at high dose rates in the soil was due to survival of the fungus in the form of dormant chlamydospores or whether there was population turnover. The former implies that spore germination was inhibited under these conditions. The latter could have been true if the fungus was parasitized by soil organisms. There have been no reports of natural antagonists of this fungus in soil, but no systematic survey has been made.

The results of these experiments suggest that a more accurate interpretation of fungal dynamics in soil can be made by using culture- and PCR-based techniques together rather than using either method alone. The fact that growth of V. chlamydosporium Jersey was observed in plants infested with PCN and not in plants infested with RKN when cPCR was used but not when plate counting was used illustrates the potential of using these methods to improve monitoring and selection of V. chlamydosporium isolates in tritrophic interactions with nematodes, as well as in host crops in biological control trials.