Modified Primers for the Identification of Nonpathogenic Fusarium oxysporum Isolates That Have Biological Control Potential against Fusarium Wilt of Cucumber in Taiwan

Previous investigations demonstrated that Fusarium oxysporum (Fo), which is not pathogenic to cucumbers, could serve as a biological control agent for managing Fusarium wilt of cucumber caused by Fo f. sp. cucumerinum (Foc) in Taiwan. However, thus far it has not been possible to separate the populations of pathogenic Fo from the nonpathogenic isolates that have biological control potential through their morphological characteristics. Although these two populations can be distinguished from one another using a bioassay, the work is laborious and time-consuming. In this study, a fragment of the intergenic spacer (IGS) region of ribosomal DNA from an Fo biological control agent, Fo366, was PCR-amplified with published general primers, FIGS11/FIGS12 and sequenced. A new primer, NPIGS-R, which was designed based on the IGS sequence, was paired with the FIGS11 primer. These primers were then evaluated for their specificity to amplify DNA from nonpathogenic Fo isolates that have biological control potential. The results showed that the modified primer pair, FIGS11/NPIGS-R, amplified a 500-bp DNA fragment from five of seven nonpathogenic Fo isolates. These five Fo isolates delayed symptom development of cucumber Fusarium wilt in greenhouse bioassay tests. Seventy-seven Fo isolates were obtained from the soil and plant tissues and then subjected to amplification using the modified primer pair; six samples showed positive amplification. These six isolates did not cause symptoms on cucumber seedlings when grown in peat moss infested with the isolates and delayed disease development when the same plants were subsequently inoculated with a virulent isolate of Foc. Therefore, the modified primer pair may prove useful for the identification of Fo isolates that are nonpathogenic to cucumber which can potentially act as biocontrol agents for Fusarium wilt of cucumber.


Introduction
Numerous formae speciales of Fusarium oxysporum (Fo) Schlechtend:Fr. are important pathogens that cause wilt diseases in many different host plants. This species is a ubiquitous soil-inhabiting fungus that is also a normal constituent of fungal communities in the rhizospheres of plants [1], [2], [3]. Historically, pathogenic Fo isolates showing high host specificity have been classified into more than 150 formae speciales based on plant species and cultivars they infect [4]. Among these formae speciales, F. oxysporum f. sp. cucumerinum (Foc) is one of the important pathogens that causes cucumber Fusarium wilt in most production areas, including North America, Europe and Asia [5]. Currently, several methods are being employed to control this disease, including breeding for resistance, fungicide application, crop rotation, soil amendments, and biological control. However, breeding for resistance and crop rotation practices are time-consuming, and the use of fungicides can be environmentally hazardous [6]. Soil amendments are less hazardous to the environment, but efficacy often depends on soil structure and pH [7]. Among biocontrol agents evaluated for the control of Fusarium wilt caused by Fo formae speciales, the use of nonpathogenic Fo isolates appears to hold much promise.
Nonpathogenic Fo isolates have been used for the control of Fusarium wilt caused by various Fo formae speciales [8]. A nonpathogenic Fo strain, Fo47, has been shown to be an effective biocontrol agent for managing Fusarium wilt in several vegetable and flower crops [9]. The introduction of nonpathogenic Fo into the stems of sweet potatoes and carnations [10], [11] resulted in the control of Fusarium wilt diseases in each respective host. In Taiwan, there are several reports of nonpathogenic Fo isolates found to be useful for the control of Fusarium wilt [12], [13]. Furthermore, Chen [12] reported that the Fo isolate Fo366 reduced the severity of cucumber Fusarium wilt caused by Foc. However, Fravel et al. [8] and Alabouvette et al. [14] showed that not all nonpathogenic Fo isolates are effective biocontrol agents. Screening nonpathogenic Fo isolates to assess the potential to serve as biocontrol agents has been difficult and time consuming. Thus, establishing a new method for rapid and reliable identification of nonpathogenic Fo isolates that have potential for use as biocontrol agents could be very beneficial.
Polymerase chain reaction (PCR) is a useful tool for the molecular characterization of fungi [15]. Many reports indicate that fungal species with similar morphology can be further classified based on PCR results [16], [17], [18], [19]. The intraspecies diversity of several fungi, including formae speciales and races, has been further distinguished using the PCR approach [20], [21]. Recent studies showed that the intergenic spacer (IGS) region of ribosomal DNA (rDNA) is a source of phylogenetic markers in Fo [22], [23], [24], [25], [26] and that the region amplified with the general primers FIGS11/FIGS12 is suitable for the study of populations, including relationships among the Fo isolates [22], [27]. The objectives of this study were to identify polymorphisms in the IGS region of rDNA that differentiate nonpathogenic from pathogenic Fo isolates and to develop a method to assess the efficacy of nonpathogenic Fo isolates for use as potential biocontrol agents to manage Fusarium wilt of cucumber.
For long-term storage of the cultures used in this study, single spore isolates were grown on PDA plates at 28uC for 5 days. Agar disks (0.5 cm in diameter) were cut from the colony margins and transferred into test tubes (12 cm in length, 1.5 cm in diameter) containing a soil-agar medium (1% WA plus 10% loamy sand soil, autoclaved for 30 min at 121uC, 15 lbs). The tube cultures were incubated at room temperature, with caps kept loosely, until the soil-agar medium was dry. Then, the caps were tightened, and the cultures were stored at room temperature.
Fungal DNA Extraction, PCR and Analysis of the IGS Region DNA extraction was conducted by the method of Saitoh et al. [29] with some modification. Ten-to fourteen-day-old mycelia from single spore isolates grown on potato dextrose agar (PDA) slants were transferred with an inoculation needle into a microtube containing 500 ml lysis buffer (200 mM Tris-HCl, 50 mM ethylenediaminetetraacetic acid, 200 mM NaCl, 1% n-lauroylsarcosine sodium salt, pH 8.0). The mycelia were dispersed in the buffer and incubated for 20 to 30 min at room temperature. The mixture was centrifuged at 18,000 g (Rotor: Nr. 12154, Sigma 3k20) for 10 min at 4uC, and then the supernatant (300 ml) was transferred into a new microtube. After the supernatant was mixed with 750 ml of ethanol by inverting the tube, the DNA was precipitated by centrifugation at 18,000 g for 4 min at 4uC. After one wash with 70% ethanol, the DNA pellet was air dried and dissolved in 50 ml TE buffer (pH 8.0). The PCR reactions were performed with genomic DNA from five selected Fo isolates using the primer set FIGS11 (59 -GTAAGCCGTCCTTCGCCTCG -39)/FIGS12 (59 -GCAAAATTCAATAGTATGGC -39) to amplify a part of the IGS region [30]. The five selected isolates include the known nonpathogenic biocontrol isolate Fo366 and four formae speciales isolates identified as F. oxysporum f. sp. cucumerinum (Foc100), F. oxysporum f. sp. luffae (Fol114), F. oxysporum f. sp. phaseoli (Fop04) and F. oxysporum f. sp. tracheiphilum (Fot60). Amplified DNA fragments were sequenced (Mission Biotech Co., Taiwan) and searched using the BLAST algorithm in GenBank from the National Center for Biotechnology Information (NCBI, Bethesda, MD).

Specificity, Sensitivity and Application of the Primer Pair
To assess the specificity of the primer pair FIGS11/NPIGS-R to detect nonpathogenic Fo isolates, genomic DNA from the following 137 Fo isolates and 8 Fusarium spp. isolates were used as template DNA for the PCR assay: 15 formae speciales, which included one isolate each of chrysanthemi; 3 isolates each of basilici and phaseoli; 4 isolates each of luffae, lycopersici, momordicae and radicis-cucumerinum; 5 isolates of melonis; 6 isolates of gladioli; 8 isolates of lilii; 10 isolates of lactucae; 12 isolates of tracheiphilum; 13 isolates each of cucumerinum and niveum; 32 isolates of cubense, 15 nonpathogenic isolates of Fo and 8 isolates of seven different Fusarium spp. (Table 1). Each 25-ml PCR mixture contained 1 ml Fo DNA, PCR Master Mix (Genemark Technology Co., Ltd., Taiwan) and 0.2 mM of each specific primer. PCR amplification was performed under the following temperature cycles: 95uC for    2 min, followed by 30 cycles of denaturing at 95uC for 30 sec, annealing at 58uC fo r 1 min, and polymerizing at 72uC for 45 sec, and then a final extension at 72uC for 10 min. PCR products were subjected to electrophoresis in 1.5% agarose gels.
To evaluate the sensitivity of the test, the quality of Fo276 DNA was quantified in GeneQuant 100 classic spectrophotometer (GE Healtcare), and diluted into several concentrations from 200 to 10 -3 ng using 1 ml of each concentration in each treatment as template DNA in a 25-ml PCR reaction volume. The sensitivity experiments were replicated three times with independent dilutions. Isolate Fo276 was used instead of Fo366 as a nonpathogenic Fo type strain because it shared 100% similarity in nucleotide sequence identity with Fo366 and because isolate Fo366 started to show a decrease in its capacity to control Fusarium wilt over the course of this study (Wang, unpublished data).
To detect the fungal colonization of cucumber roots by nonpathogenic Fo isolates, three cucumber seeds (Showy Green, Known-You seed Co., Ltd, Taiwan) were sown into three 5 cm 6 5 cm plastic tray cells containing infested peat moss (2610 6 conidia/g) for each nonpathogenic isolate. Conidia were produced on 2-to 3-wk-old PDA plate cultures at 28uC. Conidia were washed from the plates with sterile water, filtered through Miracloth (Calbiochem, San Diego, CA, USA), and quantified by counting in an improved Neubauer bright-line counting chamber (Marienfeld, Germany). One week after sowing, root tissues were collected from the seedlings, and total genomic DNA was extracted using the Plant Genomic DNA Purification Kit (Genemark Technology Co., Ltd, Taiwan) and used as template DNA.
For the detection of Fo276 in soil, a 1.5610 4 conidia/ml suspension was prepared from PDA plate cultures, as described above, and diluted in 10-fold increments to obtain a series of conidial concentrations. One milliliter of each concentration was added to microtubes containing one gram of autoclaved soil to establish various concentrations from 150,000 to 150 conidia/g soil. Total genomic DNA from each treatment was extracted using the Soil Genomic DNA Purification Kit (Genemark Technology Co., Ltd, Taiwan). All the PCR reactions were conducted three times to confirm reproducibility.

Evaluation of Biocontrol Potential of Nonpathogenic Fo Isolates
Inoculum for infesting soil in pots with the pathogenic Foc isolate (Foc100) was produced on an oat/sand medium [31]. The propagation medium (200 g oats, 200 g sand, and 400 ml distilled water), contained in 1-L flasks, was autoclaved for 20 min on two consecutive days. Twenty ml of a 1610 6 spores/ml suspension collected from PDA plate cultures were aseptically pipetted into each culture flask, which was then incubated for 2-3 wk at room temperature to allow for colonization. Subsequently, the contents of the culture flasks were air dried for 1 wk and triturated into a fine powder using a blender (RT-04, Rong Tsong Precision Technology Co., Taiwan). Dry inoculum was mixed with nonsterilized (shown to be Foc-free) Taichung field soil to achieve an inoculum level of 10 3 propagules/g of dry soil, as determined by 10-fold serial dilution plating on PCNB medium.
To evaluate the potential of nonpathogenic Fo isolates in reducing the severity of Fusarium wilt of cucumber, 10 susceptible  Fo366) and pathogenic F. oxysporum of 24 formae speciales strains (strains 2-25 represented apii, asparagi, cepae, cubense, cucumerinum-T, gladioli, glycines, lactucae, lagenariae, lilii, loti, luffae-T, lycopersici, matthiolae, medicaginis, melonis, phaseoli-T, pisi, radicis-lycopersici, rapae, raphani, spinaciae, tulipae and tracheiphilum-T, respectively. The notation -T stand for the strains that were isolated from samples collected from Taiwan). The nucleotide bases in the Fo366 sequences different from the other 24 formae speciles are indicated below the sequences. Lowercase letters indicate the nucleotide bases that differ between the pathogenic and the (Fo366) nonpathogenic strains. The dashes indicate base gaps. The dashed line region represented the sequence of specific primer, NPIGS-R. doi:10.1371/journal.pone.0065093.g001 plants (Showy Green, Known-You seed Co., Ltd, Taiwan) per isolate were grown in 5 cm 6 5 cm cell plastic trays containing noninfested peat moss or peat moss infested with conidia from nonpathogenic isolates as described in the previous section of this paper. Trays were seeded and held in the greenhouse for 10-14 days, at which time the seedlings were transplanted into 12.5 cm diameter pots containing either noninfested or Foc100-infested field soil with 10 3 propagules/g of dry soil. These plants were incubated for 8 wk in a greenhouse at 25-35uC and observed for symptom development. Disease severity was assessed weekly on a 0-4 scale in which ''0'' = healthy plants, ''1'' = plants with yellowing of the cotyledons and the first leaf, ''2'' = stunted plants or yellowing of less than half of the leaves, ''3'' = plants with stem yellowing, vascular discoloration and wilting of more than half of the leaves, and ''4'' = plants completely wilted or dead. The disease severity for each replicate of each treatment was calculated by the following formula: (SS i 6 N i ) 6 100/(4 6 N t ), where Si is the severity ratings 0 to 4, N i is the number of plants in each rating, and N t is the total number of rated plants.

Use of FIGS11/NPIGS-R to Identify Fo Isolates with Biocontrol Potential
Seventy-seven Fo isolates, 63 from soil and 14 from plant tissues, were collected at various locations across Taiwan and evaluated by PCR using the FIGS11/NPIGS-R primer set for amplification of the 500-bp IGS fragment. Isolates that tested positive for amplification were tested for pathogenicity to cucumber seedlings in a greenhouse root dip-inoculation test (5) and for their potential to suppress Fusarium wilt development using the method described in the preceding section.

Phylogenetic Analysis of Nonpathogenic Fo Showing Efficacy on Control Cucumber Wilting Based on EF-1a Gene and IGS Region
Previous studies indicated that F. oxysporum might be an opportunistic pathogen of human [32], [33]. For realizing the relationship between nonpathogenic Fo with biocontrol activity and the pathogenic Fo of human and plant, two DNA regions of translation elongation factor (EF-1a) gene and intergenic spacer (IGS) were used to amplify and analyze following the method reported by O'Donell et al. [33]. The primers for amplifying EF-1a gene and IGS were EF-1H (59 -ATGGGTAAGGAAGACAA-GAC -39)/EF-2T (59 -GGAAGTACCAGTGATCATGTT-39) and CNS1 (59-GAGACAAGCATATGACTACTG -39)/CNL12 (59 -CTGAACGCCTCTAAGTCAG -39), respectively [34]. The amplified sequences were aligned by CLUSTAL X 1.8 [35], and further visual alignments were done in SEQUENCE ALIGN-MENT EDITOR (Se-Al) v.2.0 [36]. In this study, 11 and 18 isolates of the pathogenic Fo from human and plant respectively were used for analysis. Moreover, two isolates F. foetens were used as outgroups. The isolate number, species, host and accession no. in GenBank database of these added Fo isolates were showed in Table 2. Phylogenetic analysis of the aligned sequences was done by distanced methods. The distance matrix for the aligned   sequences was calculated with the neighbor-joining (NJ) method [37]. Reliability of the inferred trees was estimated by 1,000 bootstrap resampling using the same program. Bootstrap [38] values were generated with 1,000 replicate heuristic searches to estimate support for clade stability of the consensus tree using the same program.

Statistical Analyses
Data were analyzed with the software SPSS 10.0 for WindowsH (LEAD Technologies, Inc., Charlotte, NC, USA). ANOVA was performed, and the Duncan post-hoc test was conducted to assess the differences among the treatments within each week at p = 0.05.

Primer Design
Amplified fragments of nonpathogenic Fo isolate Fo366 (strain 1) and four pathogenic Fo isolates (strains 6, 13, 18 and 25) were purified, sequenced and then compared for nucleotide variation with 20 other Fo formae speciales using Clustal X 1.81 alignment software (Fig 1). The specific primer NPIGS-R (5' -ACCCTA-GAGTATACACTAAACT -3') was designed according to the polymorphisms found in the IGS DNA sequence of Fo366 when compared with the formae speciales isolates (Fig 1).

Specificity of the Modified Primer Pair
The newly designed primer NPIGS-R combined with FIGS11 was used to assess the amplification of 122 formae speciales isolates, eight different Fusarium spp. isolates and 15 nonpathogenic Fo isolates. No fragment or expected size was amplified from the 122 pathogenic Fo and 8 different Fusarium spp. isolates (Table 1). A 500-bp fragment was amplified from the Fo isolates Fo95022, Fo95024, Fo95026, Fo276 and Fo366, but no PCR product was amplified from other Fo isolates (Table 1). The PCR-amplified fragments of the Fo isolates Fo95022, Fo95024, Fo95026, and Fo276 were sequenced, and the nucleotide sequences showed   The symbol ''+'' means that these isolates could PCR amplify a 500-bp product with the primers FIGS11 and NPIGS-R; all of these products were sequenced and confirmed to have 100% identity with Fo276. b F. oxysporum isolates were tested for their pathogenicity using the root dip assay on cucumber, and the symbol ''2'' means that there were no symptoms on the cucumber plant 3 weeks after inoculation. doi:10.1371/journal.pone.0065093.t004 Table 5. Biocontrol efficacy of the six Fusarium oxysporum isolates shown to be nonpathogenic to cucumber a . The experiment was conducted in a greenhouse (18-28uC) using the substrate infestation inoculation method. b Plants of each treatment were assayed on a scale of 0-4:0 = Healthy plants, 1 = cotyledon and first leaf with yellowing, 2 = stunting or ,1/2 leaves with yellowing, 3 = stem yellowing, vascular discoloration, and .1/2 leaves with wilt symptoms, and 4 = plant wilted or dead. The disease scale was converted to disease severity and rounded off, as described in the Materials and Methods.  100% identity with Fo366. In contrast, all isolates, both pathogenic, nonpathogenic and different Fusarium spp., had a 550 to 700-bp fragment amplified using the primer pair FIGS11/ FIGS12, thereby confirming the quality of the genomic DNA and the species of F. oxysporum. The electrophoresis picture of PCR reactions of partial F. oxysporum isolates was showed in Fig 2.

Sensitivity and Application of the Newly Designed Primer
The results of the PCR sensitivity test showed that the primers FIGS11 and NPIGS-R could amplify the 500-bp fragment from as little as 10 pg (10 -2 ng) template DNA in a 25-ml reaction mixture (Fig 3). The utility of these primers to detect nonpathogenic Fo in roots was shown by the detection of Fo276 in roots of artificially infected cucumber seedlings. A PCR product was obtained with extracted DNA down to a 200-fold dilution. In contrast, with the FIGS11/FIGS12 primer pair, PCR products were only obtained down to an 80-fold dilution. The primers FIGS11/NPIGS-R also detected DNA extracted from soil infested with conidia from the Fo isolate Fo276. Using a soil dilution series, the lowest detection limit was 150 conidia/g soil (Fig 4).

Evaluation of Biocontrol Potential of Nonpathogenic Fo Isolates
Seven nonpathogenic Fo isolates (Table 1) were evaluated by the pre-inoculation of cucumber seedlings for their potential to delay symptom expression of cucumber Fusarium wilt. Twoweek-old seedlings grown in peat moss infested with the nonpathogenic isolates were transplanted into soil infested with the Foc isolate Foc100 and observed over an 8-wk period for symptom development. Five of the nonpathogenic Fo isolates (Fo95022, Fo95024, Fo95026, Fo276 and Fo366) delayed symptom development. Using the FIGS11/NPIGS-R primer pair, the 500-bp fragment was amplified from the IGS DNA of these five isolates. The other two nonpathogenic Fo isolates (Fo95020 and Fo95021) failed to delay symptom development, and no PCR product was amplified by the FIGS11/NPIGS-R primer pair in these two isolates (Table 3). Disease severity was suppressed by the Fo isolates Fo95024, Fo95026, Fo276 and Fo366 throughout the 8-wk period following the transplantation of seedlings into Foc-infested soil, but isolate Fo95022 suppressed disease severity for only 7 wks. Two Fo isolates, Fo276 and Fo95024, were most effective in suppressing disease development; the disease severity ratings they provided after 8 wks were 6% and 14%, respectively, as compared with a 71% rating for the control, which was not pre-inoculated ( Use of FIGS11/NPIGS-R to identify Fo isolates with biocontrol potential A total of 77 Fo isolates isolated from plant tissues and soil from various locations across Taiwan were assayed by PCR using the FIGS11/NPIGS-R primer pair. The amplification of the 500-bp IGS fragment occurred in only six of the Fo isolates, including two from plants (isolates Fo7 and SPA7) and four from soil (isolates HS33, OSS11, OSS12, and OSS14) (Table  S1 and Fig S1 with partial gel results). All six isolates identified as PCR positive were found to be nonpathogenic by root dipinoculation of cucumber seedlings in a greenhouse (Table 4). All six isolates identified as PCR positive and shown to be nonpathogenic were found to suppress Fusarium wilt development ( Table 5). The disease severity ratings for the plants preinfected with nonpathogenic Fo isolates ranged from 9-25% 8 wks after being transplanted into Foc-infested soil, whereas plants that were not pre-infected received a 63% disease severity rating (Table 5).
The aligned and truncated EF-1a+IGS sequences consisted of 2,752 characters, with 2374 characters constant, 148 parsimony uninformative and 230 parsimony informative. The NJ tree constructed from the EF-1a+IGS region showed that the five isolates of nonpathogenic Fo with biological control activity were fell into one group with 80% bootstrap values and distinct from other pathogenic Fo isolates (Fig 5).

Discussion
Based on the data presented, the new primer NPIGS-R combined with the published general primer FIGS11 [27,30] distinguishes Fo isolates that are both nonpathogenic and potential biocontrol agents for Fusarium wilt of cucumber from formae speciales of Fo and other nonpathogenic Fo isolates that lack biocontrol potential. However, this primer pair does not distinguish nonpathogenic Fo isolates lacking biocontrol potential from formae speciales of Fo. Nonpathogenic Fo isolates that recovered from Fusarium wilt-suppressive soils have been extensively studied for antagonistic activity against various formae speciales of Fo [8]. There are several reports of nonpathogenic Fo isolates being used as biological control agents to manage Fusarium wilt of various crops [9], [10], [11], [12], [13]. However, it has been shown that not all nonpathogenic Fo isolates possess biocontrol potential [8], [14]. Currently, bioassay is the only available and reliable method to identify Fo isolates with biological control potential, but this assay is very time-consuming and laborious. The highly specific FIGS11/NPIGS-R primer set appears to offer an opportunity to rapidly and efficiently screen large numbers of Fo isolates to identify those with biocontrol potential.
These new molecular tools were used to investigate the genetic relationships among pathogenicity, biological control and saprophytic Fo and to elucidate the genetic determinants of pathogenicity and biological control ability. Appel and Gordon [27] showed an interaction between pathogenic and nonpathogenic Fo and addressed the differences in pathogenic race, vegetative compatibility group (VCG), mitochondrial DNA (mDNA) haplotype and IGS haplotype, but could not directly separate nonpathogenic Fo from pathogenic Fo. Therefore, the genetic basis of pathogenic, nonpathogenic or biocontrol strains of Fo remains unclear [39], [40]. In this study, we developed a molecular marker to differentiate the Taiwanese nonpathogenic Fo isolates from the pathogenic isolates on cucumbers (Fig 1). The primers FIGS11/NPIGS-R were able to specifically amplify a DNA product from the Fo isolates that showed potential for controlling Fusarium wilt of cucumber (Table 1 and Fig 2). The variation in the intergenic spacer (IGS) region of ribosomal DNA is useful for resolving intra-specific relationships within Fo [23], [25], [26]. It has been suggested that the variation in the IGS of rDNA may have a considerable effect on development, evolution, and ecology through its effects on growth-rate regulation, resulting from the role of the IGS in the production of rRNA [41]. This study suggests that the variation in the IGS region could differentiate the Fo isolates with biological control abilities from the pathogenic and the nonpathogenic Fo isolates that did not have biocontrol potential.
The Fo isolates Fo95020 and Fo95021, which showed no PCRamplified 500-bp product with the primers FIGS11 and NPIGS-R, were unable to delay the disease progression of Fusarium wilt of cucumber in a greenhouse experiment. According to this result, we speculated that these two isolates might belong to one of the Fo formae speciales or a saprophytic one that lacks the biological control ability for Fusarium wilt of cucumber. To relieve the concern about the pathogenicity of the nonpathogenic Fo used in this study, the Fo isolates Fo276 and Fo366 were tested for their pathogenicity on fourteen species of the main cultivated crops or vegetables (such as asparagus bean, basil, bitter gourd, cucumber, loofah, melon, pea, radish, snap bean, spinach, sweet potato, tomato, watermelon and wax gourd) in Taiwan, and the results showed that no symptoms were induced on the inoculated plants by either isolate (Wang, unpublished data). This lack of symptoms may be because the nonpathogenic Fo isolates were defined as those that ''failed to induce disease on a limited number of plant species to which they had been inoculated'' [14]. Therefore, the pathogenicity test may lead to problems in differentiating the isolates. Future research will focus on the utilization of the primers FIGS11 and NPIGS-R to screen more Fo isolates with biological control ability and to reveal the difference in genomic or IGS sequence between the pathogenic and nonpathogenic isolates.
The sensitivity of this PCR assay with the primers FIGS11 and NPIGS-R was shown to detect as low as 1610 -2 ng of fungal DNA. Such minute amounts of fungal DNA can be obtained easily from several natural substrates or living plant tissues that harbor the target strains. Moreover, 77 isolates of Fo from soil and plants tissues were screened and examined with FIGS11 and NPIGS-R. The results obtained demonstrated that only six isolates could be detected by FIGS11 and NPIGS-R. To confirm pathogenicity, the isolates that were amplified by FIGS11 and NPIGS-R had been tested for pathogenicity to cucumber, and no symptoms occurred in plants after inoculation. Meanwhile, these newly selected isolates have shown efficacy in delaying the disease progression of Fusarium wilt of cucumber in a greenhouse experiment. Thus, the newly developed molecular detection method with the primers FIGS11 and NPIGS-R may have practical applications in the study of the epidemiology, fungal population genetics, and even the mechanism of nonpathogenic strains in combating Fusarium wilt diseases [21]. Moreover, as mentioned in the Materials and Methods, the reason why Fo366 lost its biocontrol potential is unknown, but mutation was suspected [42]. This finding reemphasized that additional Fo isolates with biological control ability are needed and reaffirms the significance of this new reliable and highly specific protocol in the identification of the biocontrol potential of Fo isolates. In addition, several studies on Fusarium disease show that mixtures of biocontrol agents have provided better control and that a range of biocontrol mechanisms may operate in mixed populations of biocontrol agents [8], [43], [44]. Moreover, different biocontrol mechanisms have been shown among different nonpathogenic Fo isolates [8] and, the combination of different nonpathogenic Fo isolates might use multiple mechanisms to control the Fusarium wilt disease and provide better control efficacy.
Previous studies revealed that the Fusarium oxysporum is a species complex (FOSC) containing several groups of isolates with differently pathogenic activity on human and plants (Ma et al., 2010 [45], O'Donnell et al., 2004 [33], O'Donnell et al., 2009 [46]). Our result demonstrated that some nonpathogenic Fo isolates of FOSC that showed positive reaction to PCR have activity on controlling Fusarium wilt of cucumber. According to the molecular topology, the F. oxysporum with pathogenicity on human or plants were scattered in different molecular groups. The scattered phenomenon of topology in human and plant Fo pathogens was also confirmed by O'Donnell et al. (2009 [46]). They also mentioned that the evolutionary relationships between plant pathogens and nonpathogenic Fo (untested with the biological control activity) were nested and still unclear (O'Donnell et al., 2009 [46]). However, the isolates of nonpathogenic Fo with biological control ability were grouped into a unique clade and differentiated from other reference isolates of human and plant pathogens in this study. Along with these results, the nonpathogenic Fo isolates with biological control ability amplified by FIGS11/NPIGS-R were monophylogeny. Thus, the newly designed primer is indeed with the ability of specificity and peculiarity in detecting the nonpathogenic Fo with biological control ability in Taiwan.
In the future, additional isolates of different formae speciales and Fo nonpathogenic to cucumber from various geographic origins will be used to further confirm the specificity of our PCR assay method for identifying nonpathogenic Fo isolates with biocontrol potential. Further studies will determine whether the markers can be used worldwide.