Real-Time PCR Assays for Races of the Spinach Fusarium Wilt Pathogen, Fusarium oxysporum f. sp. spinaciae

Production of spinach (Spinacia oleracea L.) in the United States has nearly quintupled since the mid-1980s to meet growing consumer demand (USDA-ERS 2007; USDA-NASS 2022). The rapid increase in demand has been due, in part, to greater awareness of the health benefits of spinach and the convenience of prepackaged, fresh-market spinach (Boriss and Kreith 2006). This, in turn, has necessitated a commensurate supply of spinach seed that can only be produced in regions with the necessary climatic conditions for this species to produce seed of high quality (Foss and Jones 2005). Only regions of the world with a long summer day length, mild summer temperatures, and low humidity and rainfall during seed maturation are suitable for production of spinach seed (Metzger and Zeevaart 1985; Rackham 2002). Consequently, the maritime Pacific Northwest, including western Oregon and western Washington, is the only region in the United States where spinach seed can be produced, and accounts for up to one fifth of the global spinach seed supply (Foss and Jones 2005). Thus, any constraint to acreage for spinach seed crops represents a threat to production of spinach nationally and internationally.

Despite the optimal climatic conditions for seed production in the maritime Pacific Northwest region of the United States, the low pH typical of this region’s soils is highly conducive to Fusarium wilt of spinach, caused by Fusarium oxysporum f. sp. spinaciae (Gatch and du Toit 2017; Gyawali et al. 2021). Characteristic symptoms of Fusarium wilt of spinach plants include chlorosis, wilting, and dark brown to black vascular necrosis (Correll et al. 1994). In young seedlings, F. oxysporum f. sp. spinaciae has been reported to cause damping-off prior to the emergence of true leaves, especially when soil temperatures are warm (Fiely et al. 1995). Fusarium wilt is the greatest biotic limitation for seed producers in the maritime Pacific Northwest and has led to losses of entire spinach seed crops (Gatch 2013). Moreover, the pathogen can survive in these conducive soils for >10 years, necessitating seed growers to utilize extremely long crop rotations of 10 to 15 years or more (Gatch and du Toit 2015; Gyawali et al. 2021). Growers use a variety of management strategies to mitigate losses to Fusarium wilt, the most effective being amending soil with agricultural limestone (calcium carbonate) and/or compost (du Toit et al. 2014; Gatch and du Toit 2017). However, each of these strategies provides only transient and partial suppression of spinach Fusarium wilt. Thus, growers have relied on estimating risk of the disease by assessing soil sampled from prospective field(s) for spinach seed crops with a direct soil bioassay that has been offered annually since 2010, whereby spinach seed is planted into the soil samples and evaluated for severity of Fusarium wilt over 6 weeks (Gatch and du Toit 2015; Gyawali et al. 2021).

Although Fusarium wilt remains a major constraint for spinach seed production in the maritime Pacific Northwest, accurate diagnosis of this disease and identification of isolates of F. oxysporum f. sp. spinaciae are limited to time- and labor-intensive pathogenicity tests or bioassays (Gatch and du Toit 2015). To address the need for rapid detection, Okubara et al. (2013) developed a TaqMan real-time PCR assay for F. oxysporum f. sp. spinaciae that targeted a single nucleotide polymorphism (SNP) in the intergenic spacer (IGS) region of rDNA. Initially, the assay appeared promising as it proved specific to a broad collection of isolates of the spinach Fusarium wilt pathogen (Okubara et al. 2013). However, when the assay was evaluated against a more diverse sample of F. oxysporum isolates, including many isolates not pathogenic on spinach but which originated from fields in which spinach had been grown or from spinach seed, the assay cross-reacted with DNA extracted from many of these nonpathogenic, spinach-associated isolates of F. oxysporum as well as a few isolates of four other formae speciales, rendering the assay ineffective for identifying isolates of the target pathogen exclusively (Okubara et al. 2013).

For other formae speciales of F. oxysporum, diagnostic assays based on regions of DNA used for general identification or associated with some, but not all, lineages of a forma specialis have either cross-reacted with nontarget isolates or organisms or failed to detect all individuals of the same target forma specialis (Magdama et al. 2019; Suga et al. 2013). Since many formae speciales of F. oxysporum are polyphyletic (Baayen et al. 2000; Laurence et al. 2014; O’Donnell et al. 2009; van Dam et al. 2016), diagnostic assays based on regions of the genome used for genus- or species-level identification have often lacked specificity. Thus, the most robust diagnostic assays are likely those targeting loci or genes associated functionally with pathogenicity or host specificity.

Many of the genes that determine pathogenicity among formae speciales, i.e., effector genes, are encoded on accessory chromosomes that are transferable among isolates (Li et al. 2020a, b; Ma et al. 2010; Schmidt et al. 2013; van Dam et al. 2017). This explains, in part, the complicated evolutionary relationships among isolates of many formae speciales (Henry et al. 2021; O’Donnell et al. 2009; van Dam et al. 2016). The profile of effector genes, or effector gene candidates, has proven a reliable set of markers for distinguishing isolates of one special form of F. oxysporum from others. Recently, isolates of F. oxysporum f. sp. spinaciae were shown to possess a distinct profile of candidate effector genes that is distinguishable from those of isolates of other special forms and avirulent, spinach-associated F. oxysporum isolates (Batson et al. 2021). Additionally, isolates of the spinach Fusarium wilt pathogen were placed into two pathogenicity groups based on the presence or absence of specific candidate effector genes. These pathogenicity groups are now recognized as races (Batson et al. 2022a) as some cultivars of spinach differ in susceptibility to the two races. Therefore, knowing which race is present or predominant in a field or region has practical significance. Of the 52 candidate effector genes identified by Batson et al. (2021), five were unique to the genome assemblies of isolates of F. oxysporum f. sp. spinaciae and absent from the genomes of 222 other isolates of F. oxysporum, making them promising markers for diagnostic assays for the spinach Fusarium wilt pathogen.

To date, there is no rapid means of detection of F. oxysporum f. sp. spinaciae in culture, plants, seed, or soil. Accurate diagnosis of Fusarium wilt of spinach and differentiation of isolates of each race currently requires culturing and pathogenicity testing, which require weeks to months to yield results. Therefore, the objectives of this study were to: (i) develop rapid diagnostic assays for isolates of each race of F. oxysporum f. sp. spinaciae based on candidate effector genes, and (ii) evaluate the assays for detection and quantification of F. oxysporum f. sp. spinaciae in inoculated spinach plants.

Materials and Methods

Locus selection and real-time PCR assay design

Of the five candidate effector genes for F. oxysporum f. sp. spinaciae identified by Batson et al. (2021), Fos1_2 (GenBank accession no. ON411376.1) was common to all available genome assemblies of F. oxysporum f. sp. spinaciae and lacked sequence variation among isolates of this pathogen, and Fos2 (ON411377.1) was associated only with the race 2 isolates and also lacked variation in DNA sequence among isolates. Thus, Fos1_2 was selected to develop a real-time PCR assay to identify isolates of both races of F. oxysporum f. sp. spinaciae, while Fos2 was selected to develop a real-time PCR assay that distinguishes race 2 from race 1 isolates of the pathogen. Based on the DNA sequences of each locus, real-time PCR primers and TaqMan probes were designed with Primer Express 2.0 (Applied Biosystems, Waltham, MA; Table 1). The DNA fragment amplified for each assay was 54 bp. Both probes were designed with a 3′ minor groove binder (MGB) to enhance sequence discrimination of target from nontarget DNA sequences (Kutyavin et al. 2000).

Table 1. Primers and probes developed for real-time PCR assays based on two candidate effector genes specific to Fusarium oxysporum f. sp. spinaciae

Primer/probea	Sequence (5′–3′)b	Length (bp)	Locus	Source
UNUP18S42	CGTAACAAGGTTTCCGTAGGTGAAC	25	ITS	Bakkeren et al. (2000)
UNLO28S576B	CTCCTTGGTCCGTGTTTCAAGACG	24	ITS	Bakkeren et al. (2000)
Fos1_2F	CGCCGCCTGCTGAAG	15	Fos1_2	This study
Fos1_2R	CGCACAGACAGGTGTTTGG	19	Fos1_2	This study
Fos1_2P	FAM-CTTGCATGCGGTACCT-MGBNFQ	16	Fos1_2	This study
Fos2F	AATCATCATGGCGGCTTTC	19	Fos2	This study
Fos2R	TCCCAGGTGCCCAAGTG	17	Fos2	This study
Fos2P	VIC-TGGCCTCGCCTTC-MGBNFQ	13	Fos2	This study

^a Each assay amplifies a partial region of a candidate effector gene identified by Batson et al. (2021). The size of the PCR product for each assay is 54 bp.

^b Both probes have a 3′ minor groove binder (MGB) modification with a nonfluorescent quencher (NFQ), and the 5′ fluorescent dye for Fos2 and Fos1_2 is VIC and FAM, respectively.

View as image HTML

Isolates tested

In total, 118 fungal and oomycete isolates were used in this study, including 44 isolates of F. oxysporum f. sp. spinaciae (21 isolates of race 1 and 23 isolates of race 2), 10 isolates of F. oxysporum associated with spinach that were demonstrated to be avirulent on spinach, 26 isolates of a total of 12 other formae speciales, 31 isolates of a total of 27 other Fusarium spp., and 7 isolates of other fungal or oomycete genera associated with spinach as pathogens or saprophytes (Supplementary Table S1). All the isolates of F. oxysporum f. sp. spinaciae were evaluated for pathogenicity on spinach by Batson et al. (2021), Fiely et al. (1995), or as part of this study (see below), and were collected originally from wilting spinach plants, soil in which spinach was grown, or spinach seed (Supplementary Table S1). The species or forma specialis designation of each isolate received from others was provided by the source or contributing program.

Pathogenicity testing

Each isolate associated with spinach that either had not previously been tested for pathogenicity on spinach by Batson et al. (2021) or Fiely et al. (1995), or that had been validated as F. oxysporum f. sp. spinaciae but the race designation of the isolate was unknown, was tested for pathogenicity on spinach following the protocol described by Batson et al. (2022a). In summary, each of two proprietary spinach inbreds (A and C) that have been demonstrated to distinguish isolates of each race of F. oxysporum f. sp. spinaciae (Batson et al. 2021, 2022a), were inoculated with each isolate of F. oxysporum in a randomized complete block design that entailed a 2-by-18 inbred-by-inoculation factorial treatment design with three replicate blocks. In total, 14 of the isolates of F. oxysporum were tested for pathogenicity on each inbred during this study. The inbreds were inoculated with a known isolate of each of F. oxysporum f. sp. spinaciae race 1 (Fus254) and race 2 (Fus167) as two positive control treatments. In addition, spinach plants were treated with water or an isolate of F. oxysporum avirulent on spinach (Fus187) as two negative control treatments.

Each isolate of F. oxysporum was cultured on half-strength potato dextrose agar (PDA) supplemented with chloramphenicol (100 mg/liter) for 5 days at room temperature (22 ± 1°C) under ambient light from north-facing windows. Three agar blocks (each 1 mm³) were excised from the leading edge of hyphal growth and placed in 300 ml of sterile Kerr’s broth (Kerr 1963). The inoculated broth was incubated at room temperature on a gyratory shaker (150 rpm) with ambient light. After 7 days, the microconidial suspension was filtered through one layer of cheesecloth, the microconidia counted on a hemocytometer, and the suspension diluted to 1.75 × 10⁶ microconidia/ml using sterile, deionized water. A 150-ml aliquot of the diluted inoculum was sprayed into 7.0 liters of RediEarth Propagation Mix (SunGro Horticulture, Agawam, MA) using a handheld spray bottle to achieve a final concentration of 3.75 × 10⁴ microconidia/ml potting medium. The potting medium was mixed with the inoculum in a custom-made, Gustafson batch seed treater (Gustafson LLC, Shakopee, MN) for approximately 5 min. Additional water was added until the potting medium was moist but not saturated. The inoculated potting medium was dispensed into six-cell packs (T.O. Plastics, Clearwater, MN) with approximately 720 ml of potting medium/six-cell pack (120 ml/cell). Two seeds of the appropriate inbred were planted into the medium in each cell (12 seeds/pack), and the emerged seedlings thinned to 1 plant/cell.

For the first 11 days after planting (DAP) each trial, the temperature was set to 20 to 24°C during the day and 18 to 22°C at night. The temperature was then increased to range from 26 to 28°C during the day and 22 to 25°C at night to promote development of Fusarium wilt. The supplemental lighting was set for 9-h days and 15-h nights. Plants were irrigated daily with water supplemented with 20-20-20 fertilizer (Everris, Dublin, OH) at a final N concentration of 200 ppm. The severity of Fusarium wilt symptoms was assessed 14, 21, and 28 DAP using a 0-to-5 ordinal scale that measures the proportion of wilted foliage (0 = no wilt symptoms evident, 1 = up to 20% of foliage wilting, 2 = 21 to 40% of foliage wilting, 3 = 41 to 60% of foliage wilting, 4 = 61 to 80% of foliage wilting, and 5 = 81 to 100% wilting or a dead plant), and then converted to an index, as described by Gatch and du Toit (2015). The final rating (28 DAP) was used for statistical analyses since this provided the best resolution among treatments.

DNA extraction

Each fungal isolate was grown for up to 2 weeks on half-strength PDA supplemented with chloramphenicol. DNA was then isolated from ∼100 mg of mycelium of each isolate with the Plant Synergy 2.0 kit (OPS Diagnostics, Lebanon, NJ), following the manufacturer’s protocol. The DNA was then diluted 10-fold in sterile, molecular-grade water as a working stock. To validate the quality of each working stock of DNA, the internal transcribed spacer (ITS) region of rDNA was amplified, and the PCR products visualized on an agarose gel. The ITS rDNA was amplified with the primers UNUP18S42 and UNLO28S576B (Bakkeren et al. 2000; Table 1) in a 30-μl reaction volume with a final concentration of 1× PCR buffer II (Applied Biosystems), 1.5 mM of MgCl₂, 0.2 mM of dNTPs, 0.4 μM of each primer, and 0.05 units of AmpliTaq Polymerase/μl (Applied Biosystems). A 2-μl aliquot of the working stock of DNA was added to each reaction, and the mix incubated in a thermal cycler (Thermo Hybaid PCR Express, Thermo Fisher Scientific, Waltham, MA) for one cycle of 3 min at 94°C; followed by 31 cycles of 45 s at 92°C, 45 s at 60°C, and 60 s at 72°C; and then one cycle of 10 min at 72°C. The PCR products were run on a 1.5% agarose gel stained with GelRed (Biotium, Fremont, CA), and then visualized with UV light.

Real-time PCR assays

Real-time PCR assays were completed in a Corbett Rotor-Gene Q 5plex HRM thermal cycler (Qiagen, Hilden, Germany). Each real-time PCR reaction was completed in a 0.1-ml tube (Corning, Corning, NY) with a final volume of 20 μl, containing 10 μl of Luna universal probe qPCR master mix (New England Biolabs, Ipswich, MA) with primers and probes at a final concentration of 400 and 200 nM, respectively. The remaining volume included 2 μl of template DNA and molecular-grade water. The thermal cycling conditions for both real-time PCR assays were set for one cycle of 60 s at 95°C, followed by 40 cycles of 15 s at 95°C and 30 s at 60°C. Each DNA sample was tested in duplicate. A threshold of 0.1 was used to determine the cycle threshold (C_q) of each sample. All real-time PCR assays were analyzed with the Rotor-Gene Q Series Software version 2.3.1 (Qiagen). In each run, normalized fluorescence measurements were adjusted with the ‘Slope Correct’ function to compensate for a gradual increase in fluorescence as the reaction progressed.

Specificity and sensitivity of the real-time PCR assays

In addition to in silico specificity testing of the two candidate effector genes, Fos1_2 and Fos2, performed by Batson et al. (2021), the specificity of each real-time PCR assay was tested with DNA extracted from each isolate listed in Supplementary Table S1. All samples were tested in duplicate. C_q values >40 were considered negative. To identify the limit of detection of each assay, real-time PCR assays were completed with DNA extracted from a race 1 isolate (Fus254) of F. oxysporum f. sp. spinaciae and a race 2 isolate (Fus167). The amount of DNA ranged from 52.5 fg to 2.1 ng over the standard dilution series prepared for each isolate in sterile, molecular grade water. Real-time PCR assays were used as described above, with the exception that three technical replicates were run for each DNA concentration and two technical replicates were included with water instead of DNA as the no-template control samples. The limit of detection for each assay was determined as the lowest DNA concentration at which all three technical replicates amplified the target DNA. The dilution series of DNA evaluated as standards for each real-time PCR assay were quantified with a Qubit Fluorometer 2.0 (Invitrogen, Carlsbad, CA).

Spinach infection time course trial

The real-time PCR assays were tested for detection of the pathogen in infected spinach plants. Based on preliminary greenhouse trials and tests of plants from spinach seed crops with Fusarium wilt, the pathogen could be detected with the real-time PCR assays using DNA extracted from the roots of infected plants (data not shown). Therefore, a trial was set up to evaluate: (i) if the pathogen could be detected in spinach plants with the real-time PCR assays as early as 7 days after planting seed in inoculated potting medium; (ii) whether the pathogen could be quantified in infected plants over time with the real-time PCR assays and correlated with severity of Fusarium wilt symptoms; and (iii) whether it was possible to use the assays to distinguish plants infected with race 1 and/or race 2 of F. oxysporum f. sp. spinaciae. Two proprietary spinach inbreds (A and C) were planted in potting medium inoculated with an isolate of race 1 (Fus254), race 2 (Fus167), a 1:1 mix of both isolates, or water. A microconidial suspension of each isolate of F. oxysporum f. sp. spinaciae was produced and applied to potting medium as described above, with the exception that the inoculum applied was at 1.25 × 10⁴ microconida/ml potting medium to monitor progression of infection in the plants and development of wilt symptoms more gradually than would occur at the greater inoculum density used for the pathogenicity trial. The trial was set up as a randomized complete block design with three replicate blocks of a 2-by-4 inbred-by-inoculation factorial treatment design. Plants were rated for severity of wilt symptoms 7, 10, 14, 21, and 28 DAP, as described for the pathogenicity trial. The trial was repeated.

One plant was harvested at random from each replicate six-cell pack of each treatment at each rating. The crown and roots were triple-rinsed in sterile, deionized water, and dried for 10 min. For the first three sample times (7, 10, and 14 DAP), all root tissue below the crown of each plant was used for DNA extractions. The plants were much larger by 21 and 28 DAP, so a 2- to 5-cm-long piece of the crown and root tissue was sampled from each plant. The tissue sampled was weighed and stored at −80°C. DNA was then extracted from the tissue with the Plant Synergy 2.0 DNA extraction kit, as described earlier, with one modification: 750 μl of plant homogenization buffer was used instead of the recommended 500 μl because the larger volume accommodated the plant tissue more effectively. Each sample was eluted in 100 μl of TE buffer (50 mM Tris and 1 mM EDTA, pH = 8.0). Real-time PCR assays were conducted as described earlier, and the amount of DNA in each sample was extrapolated from the standard curve estimated from the DNA dilution series of isolate Fus167, ranging from 11 ng to 1.1 pg (trial 1) or 21.0 ng to 0.21 pg (repeat trial). In the first trial, DNA was not included in the standard curve at the limit of detection (0.21 pg, Fig. 1), but all estimated DNA concentrations <0.21 pg were considered too low for quantification and, thus, negative for the target gene.

Download as PowerPoint

Descriptive statistics, graphics, and statistical analyses

Summary statistics and graphics for Figs. 1, 2, and 3 were generated with software from the R-package suite tidyverse (Wickham et al. 2019). Spearman’s correlation coefficient was estimated with the “cor.test()” function in R. Statistical analyses for the pathogenicity test results were conducted on wilt severity ratings using PROC MIXED in SAS (SAS Institute, Cary, NC). Inbred and isolate were considered fixed effects, while the effect of block was considered random. The assumption of homogeneity of variances was violated so the variances were modeled with an unstructured variance-covariance matrix with the following statement in PROC MIXED: “repeated parent*isolate/subject = block type = un(1);”. Posthoc comparisons were calculated for the significant interaction terms using Fisher’s protected least significant difference (LSD) at P < 0.05. Means separations were calculated with the macro pdmix800 (Saxton 1998).

Download as PowerPoint

Download as PowerPoint

For the time course trial, analyses of variance were conducted on the indices of Fusarium wilt severity, independently for each trial. Due to nonhomogenous variances among treatments, the Fusarium wilt severity index was rank-transformed. The model statement included the fixed factors of inbred, isolate, and time, while the effect of block was random. Control plants treated with water were excluded from the analyses because Fusarium wilt did not develop on those plants. Also, because Fusarium wilt did not develop consistently before 21 DAP, only the wilt severity ratings at 21 and 28 DAP were included in the analyses. Posthoc comparisons were calculated as described earlier.

For the DNA standard curve calculation, the linear relationship between the C_q and log₁₀-transformed amount of DNA was assessed with the function “lm()” in R, and the slope, intercept, and R² were calculated from the linear model (Fig. 1). The means of three replicates were used to calculate the regression relationships, and the C_q values were included for each DNA concentration if: (i) DNA was amplified for each technical replicate, and (ii) the SD among technical replicates of the same DNA concentration was <0.8. PCR efficiency was calculated with the formula: −1 + 10^(−1/slope), where “slope” represents the slope estimated from the regression model.

The number of copies of each locus/genome was inferred from an estimated genome size of 65 Mbp, derived from genome assemblies of isolate Fus254 (race 1) and Fus167 (race 2) (Batson 2022; Batson et al. 2021). In these assemblies, Fus254 has one copy of Fos1_2 and no copies of Fos2, while Fus167 has two copies of Fos1_2 and one copy of Fos2. For the race 2 isolates of F. oxysporum f. sp. spinaciae used in this study, the ratio in copy number was estimated by dividing the number of copies of Fos2 by that of Fos1_2. If the ratio in copy number is consistent across isolates of the same race, it should be possible to estimate the proportion of DNA of each race in a mixed culture or a mixed infection of a plant.

For the spinach infection time course trial, the proportion of each race of F. oxysporum f. sp. spinaciae detected in infected plants should be distinguishable based on the ratio between the copy number of Fos2 and Fos1_2. DNA extracted from pure cultures of isolate Fus167 should have a ratio in copy number of approximately 0.5 since there are two copies of Fos1_2 for each copy of Fos2 in the genome. In contrast, pure cultures of Fus254 should have a ratio of 0 since the isolate lacks Fos2. Thus, in mixed populations of both isolates, the ratio between Fos1_2 and Fos2 should be between 0 and 0.5. As the proportion of Fus167 (race 2) increases relative to Fus254 (race 1), the ratio in copy number between Fos1_2 and Fos2 will approach 0.5, and as the amount of Fus254 (race 1) increases, the ratio in copy number of Fos2 and Fos1_2 will approach 0. Therefore, the ratio of Fos2:Fos1_2 was calculated for each plant sampled in the infection time course trial by extrapolating the number of copies of both candidate effector genes from the standard curve.

Results

Pathogenicity testing

Symptoms of Fusarium wilt were observed on plants of inbreds A and C inoculated with isolate Fus254 (F. oxysporum f. sp. spinaciae race 1) and Fus167 (race 2) (Supplementary Table S2). Neither inbred developed symptoms of Fusarium wilt when treated with water or the nonpathogenic isolate, Fus187. Symptoms of Fusarium wilt were first observed 14 DAP, and the severity of wilt progressed over the duration of the trial. For the analysis of variance, there was a significant effect of inbred (P < 0.0001) and isolate (P = 0.0054), and the interaction between inbred and isolate was significant (P = 0.0013). The positive control isolates caused the expected differences in severity of symptoms of wilt on each inbred, i.e., the race 1 isolate (Fus254) caused more severe wilt on inbred A than inbred C, and the race 2 isolate (Fus167) caused comparable severity of wilt symptoms on the two inbreds (Supplementary Table S2).

Among the 14 test isolates evaluated for pathogenicity on spinach in this study, to supplement the 33 isolates of the spinach Fusarium wilt pathogen identified in prior studies, 11 induced symptoms of wilt on both inbred lines and three did not cause either inbred to wilt (Supplementary Table S2). Of the 11 isolates identified as F. oxysporum f. sp. spinaciae, six induced symptoms of wilt that were characteristic of race 1 (more severe wilt on inbred A than inbred C), and the remaining five isolates caused symptoms of wilt consistent with race 2 (similar severity on inbreds A and C). Neither inbred inoculated with isolates Fus1061 and Fus1063 from spinach or isolate Fus1069 of F. oxysporum f. sp. raphani from radish induced symptoms of wilt on either inbred.

Sensitivity of the real-time PCR assays

The limit of detection was 0.210 pg of DNA for consistent detection of isolate Fus254 (race 1) with the Fos1_2 real-time PCR assay, while the limit of detection was half that (0.105 pg of DNA) for isolate Fus167 (race 2) with this assay (Fig. 1A). At these limits of detection for Fus254 and Fus167, the SD in C_q was 0.64 and 1.16, respectively, and the PCR efficiencies were 101 and 97%, respectively (Fig. 1A). In comparison, the C_q SD ranged from 0.03 to 0.41 above the lower limit of detection. The variability in C_q was large at the limit of detection for isolate Fus167, so this measurement of C_q was excluded from the linear model (Fig. 1A). The intercepts calculated from the linear models developed for the serial DNA dilutions of Fus254 and Fus167 differed by 0.9 C_q, implying an approximately two-fold difference in the number of copies of Fos1_2/ng genomic DNA between the two isolates. For the Fos2 real-time PCR assay, serial dilutions of isolate Fus167 (race 2) only were evaluated because only race 2 isolates of F. oxysporum f. sp. spinaciae have the candidate effector Fos2. The limit of detection of this assay was 0.210 pg DNA (∼3 copies of the target locus; Fig. 1B), at which the SD of C_q was 0.73 and the assay had an efficiency of 95%.

Specificity of the real-time PCR assays

DNA was amplified from all 44 isolates of F. oxysporum f. sp. spinaciae with the assay targeting Fos1_2 (Table 2; Supplementary Table S1), whereas DNA was amplified only from the 23 isolates of race 2 of F. oxysporum f. sp. spinaciae when tested with the real-time PCR assay targeting the race 2-specific locus (Table 2; Supplementary Table S1). In addition to the >200 genome assemblies of F. oxysporum evaluated that lacked Fos1_2 and Fos2 (Batson et al. 2021), none of the 74 nontarget isolates evaluated in this study had either Fos1_2 or Fos2 (Table 2; Supplementary Table S1). The ratio in copy number between Fos2 and Fos1_2 ranged from 1:0.61 to 1:3.45 for the 23 isolates of race 2 of F. oxysporum f. sp. spinaciae (Supplementary Table S1). Most isolates had a Fos2:Fos1_2 ratio of approximately 1:1 (nine isolates) or 1:2 (12 isolates; Supplementary Table S1), indicating the copy number of each locus differed by isolate.

Table 2. Summary of isolates used to validate the specificity of each of two real-time PCR assays for detecting isolates of Fusarium oxysporum f. sp. spinaciae

Speciesa	No. of isolates (species)	Real-time PCR assaysb
		Fos1_2	Fos2
F. oxysporum f. sp. spinaciae (race 1)	21	+	–
F. oxysporum f. sp. spinaciae (race 2)	23	+	+
F. oxysporum (NPS)	10	–	–
F. oxysporum f. sp. basilici	1	–	–
F. oxysporum f. sp. apii	3	–	–
F. oxysporum f. sp. cepae	2	–	–
F. oxysporum f. sp. ciceris	1	–	–
F. oxysporum f. sp. cucumerinum	3	–	–
F. oxysporum f. sp. conglutinans	1	–	–
F. oxysporum f. sp. lycopersici	2	–	–
F. oxysporum f. sp. matthiolae	1	–	–
F. oxysporum f. sp. melonis	2	–	–
F. oxysporum f. sp. niveum	4	–	–
F. oxysporum f. sp. pisi	1	–	–
F. oxysporum f. sp. raphani	5	–	–
Fusarium spp.	31 (27)	–	–
Cercospora chenopodii	1	–	–
Cladosporium variabile	1	–	–
Colletotrichum dematium	1	–	–
Pythium ultimum	1	–	–
Stemphylium beticola	1	–	–
Stemphylium vesicarium	1	–	–
Verticillium dahliae	1	–	–

^a Species or F. oxysporum forma specialis. All isolates of F. oxysporum f. sp. spinaciae were validated for pathogenicity by Batson et al. (2021), Fiely et al. (1995), or in this study (see Supplementary Table S1). NPS = nonpathogen of spinach but recovered from spinach, spinach seed, or soil in which spinach was grown.

^b Mean cycle threshold (C_q) values from duplicate real-time PCR assays designed for two candidate effector genes, Fos1_2 and Fos2, identified by Batson et al. (2021). + = DNA was amplified; – = DNA was not amplified by 40 cycles.

View as image HTML

Spinach infection time course trial

In both infection time course trials, symptoms of Fusarium wilt were first observed on inoculated plants 14 DAP and, in general, progressed throughout the trials (Fig. 2A and B). The control plants treated with water did not develop symptoms of Fusarium wilt. In both trials, inbred A developed more severe symptoms of Fusarium wilt when inoculated with the race 1 isolate, while inbred C developed more severe symptoms of wilt when inoculated with the race 2 isolate (Fig. 2A and B). The two inbreds also responded differently to infection by the 1:1 mix of isolates of races 1 and 2.

In trial 1, there was a significant effect of inbred (P = 0.0035) on the Fusarium wilt severity index 21 and 28 DAP, but the main effect of isolate was not significant (P = 0.2658), and the main effect of time (P = 0.0587) reflected differences among sample times. There was a significant inbred-by-isolate interaction (P = 0.0035) and inbred-by-time interaction (P = 0.0054). The latter interaction was likely due to the slight decline in severity of wilt of inbred C plants inoculated with the race 1 isolate, Fus254, from 21 to 28 DAP (Fig. 2A). Both Fos1_2 and Fos2 were detected with the real-time PCR assays from plants grown in inoculated potting medium, starting at the first evaluation 7 DAP, but not from the control plants (Fig. 2A). For the 18 plants infected with F. oxysporum f. sp. spinaciae, Fos1_2 was detected with the PCR assay in 10 (56%), 14 (78%), 14 (78%), 18 (100%), and 18 (100%) of the plants sampled 7, 10, 14, 21, and 28 DAP, respectively. For the 12 plants grown in potting medium inoculated with either race 2 or the 1:1 mix of races 1 and 2, Fos2 was detected by real-time PCR assay in 6 (50%), 10 (83%), 12 (100%), 12 (100%), and 12 (100%) of the plants sampled 7, 10, 14, 21, and 28 DAP, respectively.

In trial 2, the Fusarium wilt severity index measured 21 and 28 DAP was affected significantly by the inbred-by-isolate interaction (P < 0.0001), and there was a significant main effect of time (P < 0.0001), but there was not a significant main effect of inbred (P = 0.2281) or isolate (P = 0.2121), and none of the other interaction terms was significant (P ≥ 0.0887). Regardless of sampling time after planting, inbred A developed similar severity of wilt when grown in potting medium inoculated with the race 1 isolate and the mix of isolates of races 1 and 2 (P = 0.9121). Inbred A developed less severe symptoms of wilt in potting medium inoculated with the race 2 isolate than when grown in potting medium inoculated with the race 1 isolate (P = 0.0004) or a mix of the two races (P = 0.0005; Fig. 2B). For inbred C, the most severe wilt was observed on plants infected with the race 2 isolate, followed by the mix of both isolates, and then the race 1 isolate (Fig. 2B). For the 18 plants inoculated with F. oxysporum f. sp. spinaciae, Fos1_2 was detected by real-time PCR assay in 16 (89%), 17 (94%), 18 (100%), 18 (100%), and 18 (100%) of the plants sampled 7, 10, 14, 21, and 28 DAP, respectively. Among the 12 plants inoculated with either race 2 or the 1:1 mix of races 1 and 2, Fos2 was detected by real-time PCR in 10 (83%) of the plants sampled 7 DAP and all 12 (100%) of the plants sampled thereafter.

In both trials, the amount of DNA detected per mg of tissue using the real-time PCR assays was positively correlated with severity of Fusarium wilt (Fig. 2C to F). The amount of DNA detected in plant tissue spanned ≤4 orders of magnitude, and neither locus was amplified from control plants treated with water. For plants inoculated with F. oxysporum f. sp. spinaciae, the amount of DNA of Fos1_2 detected was correlated positively with Fusarium wilt severity (Spearman’s rho (r) = 0.7285; P = 4.36 × 10⁻¹⁰), and for plants inoculated with the race 2 isolate, Fus167, or the 1:1 mix of races 1 and 2, the amount of Fos2 DNA detected also was correlated positively with Fusarium wilt severity (r = 0.7135; P = 3.42 × 10⁻⁷). For plant samples from which both candidate effectors were detected, the amount of DNA detected for the two loci was strongly correlated (r = 0.9732; P < 2.2 × 10⁻¹⁶).

The ratio of Fos2:Fos1_2 was 0 for plants of inbreds A and C inoculated with the race 1 isolate, Fus254 (Fig. 3). For inbred A and C plants inoculated with the race 2 isolate, Fus167, the mean ratio in copy number was 0.526 and 0.514, respectively (Fig. 3). In contrast, plants grown in potting medium inoculated with both isolates had a ratio ranging from 0.052 to 0.558 (Fig. 3). For plants inoculated with the 1:1 mix of isolates, inbred A had a mean ratio of Fos2:Fos1_2 of 0.342, and inbred C had a mean ratio of 0.519. Inbred A is susceptible to both isolates, and inbred C is highly resistant to race 1 isolates but susceptible to race 2 isolates (Fig. 2A and B), so the ratio in copy number reflected the relative susceptibility of each inbred to isolates Fus254 and Fus167.

Discussion

In this study, real-time PCR assays were developed for detecting isolates of Fusarium oxysporum f. sp. spinaciae and distinguishing isolates of each of the two known races of this spinach pathogen. Both assays amplified DNA from isolates of the intended race(s) of F. oxysporum f. sp. spinaciae, and neither assay cross-reacted with nontarget isolates, indicating robust specificity for the spinach Fusarium wilt pathogen. Moreover, the assays were sensitive, as they were used to amplify DNA from >100 fg of genomic DNA (<10 copies of the target locus). Accordingly, these assays enabled rapid identification of F. oxysporum f. sp. spinaciae from pure cultures and in infected plants, even before symptoms had developed. These tools will facilitate studies on the dissemination and survival of isolates of the two races of F. oxysporum f. sp. spinaciae and the efficacy of disease management practices.

The real-time PCR assays were designed based on loci identified through comparative genomics (Batson et al. 2021). This approach was inspired by diagnostic assays developed for other formae speciales of F. oxysporum (Burkhardt et al. 2019; Lievens et al. 2009; van Dam et al. 2018). In each of those studies, the loci used for diagnostic assays were either (candidate) effector genes (Lievens et al. 2009; van Dam et al. 2018) or loci shared by all known phylogenetic lineages of a forma specialis (Burkhardt et al. 2019). Although it is unknown whether the loci used in this study are functional effectors, they have, so far, proven to be robust targets for detecting the spinach Fusarium wilt pathogen and distinguishing isolates of each race. In contrast, a previous effort to develop a real-time PCR assay for F. oxysporum f. sp. spinaciae based on an SNP in the IGS region of rDNA cross-reacted with DNA of off-target isolates (avirulent isolates of F. oxysporum associated with spinach, seed, or soil in which spinach had been grown, and a few isolates of F. oxysporum ff. spp. lageneriae, lilii, melongenae, and raphani; Okubara et al. 2013). Four of these cross-reacting isolates were included in this study, none of which cross-reacted with the real-time PCR assays developed for Fos1_2 and Fos2, validating the specificity of each assay.

The assays used in this study proved promising for rapid diagnosis of Fusarium wilt of spinach in early stages of infection. In the spinach infection time course trial, DNA of F. oxysporum f. sp. spinaciae was detected in most spinach plants at the earliest sampling, 7 DAP, and the amount of DNA of the pathogen detected in plants increased over time in correlation with an increase in the severity of Fusarium wilt. It may be possible to detect the pathogen earlier than 7 DAP, but this was not evaluated in the study. These real-time PCR assays have been used for diagnostic purposes on DNA extracted from pure cultures of F. oxysporum isolated from crops other than spinach produced in the same regions as commercial spinach seed crops, and from wilting plants in spinach seed crops. In one instance, the assays were tested on DNA extracted from spinach plants with symptoms of wilt resulting from flooding of the field, not from Fusarium wilt. The real-time PCR assays did not amplify the target DNA from any of the plants, indicating the absence of F. oxysporum f. sp. spinaciae, which was determined <24 h after receiving the samples. Concurrently, F. oxysporum was isolated from the wilting spinach plants, and subsequent pathogenicity tests over four weeks revealed that none of these isolates was pathogenic on spinach (e.g., isolates Fus1061 and Fus1063 in Supplementary Tables S1 and S2). The real-time assays will be valuable for differentiating wilt symptoms caused by F. oxysporum f. sp. spinaciae from those induced by other pathogens (e.g., Verticillium wilt symptoms caused by V. dahliae or postemergence seedling blight caused by Pythium spp., Aphanomyces spp., or Rhizoctonia spp.) or abiotic causes (e.g., soil moisture stress and/or excessively high temperatures).

In addition to aiding rapid diagnoses of spinach Fusarium wilt, these assays will facilitate studies on managing the disease. Since 2010, spinach seed growers in the maritime Pacific Northwest have submitted samples from their fields for testing for Fusarium wilt risk using a soil bioassay that is offered annually at the Washington State University Mount Vernon Northwestern Washington Research and Extension Center (Gatch and du Toit 2015), with >600 fields having been tested in the past 14 years (Gyawali et al. 2021). Briefly, seed growers submit a ∼20-liter representative sample of soil collected from their field; the soil is processed (dried, crushed, and sieved), dispensed into pots, and seed of three spinach inbreds, ranging from highly susceptible to partially resistant to F. oxysporum f. sp. spinaciae, are planted and monitored for severity wilt. Although highly effective at quantifying the relative risk of Fusarium wilt, the soil bioassay is time-, labor-, and resource-intensive, and takes two months to complete. Space constraints limit the number of soils that can be evaluated each year. In years of high demand, soil samples from as many as 80 fields in western Washington have been submitted for testing in the soil bioassay. The real-time PCR assays designed in this study could potentially be used to quantify soilborne inoculum, and thus, estimate the risk of Fusarium wilt of spinach in 24 to 48 h. However, the risk of Fusarium wilt is influenced by many factors besides inoculum density, such as pH, soil microflora, soil amendments, and so on (Batson et al. 2021, 2022a; Gatch and du Toit 2017; Gyawali et al. 2021; Naiki and Morita 1983; Naito et al. 1996, 1997). Identification and quantification of the pathogen from soils using the real-time PCR assays would require optimization, but the assays could be used preliminarily to quantify the pathogen from DNA extracted from soils as an estimate of inoculum density. Soils with low inoculum concentrations could be validated with the soil bioassay, and soils with economically damaging concentrations of inoculum need not be tested with the resource-intensive soil bioassay. This also would necessitate developing a threshold for soilborne inoculum, which can be influenced by many factors such as the susceptibility of the spinach lines planted to Fusarium wilt (Batson et al. 2022a) or the use of soil amendments such as agricultural limestone to suppress the disease (Gatch and du Toit 2017).

In addition to providing a tool for aiding risk assessment, the real-time PCR assays will facilitate studies on the efficacy of management practices for Fusarium wilt and epidemiology of the pathogen. Very long durations of crop rotation for spinach seed crops are essential in the maritime Pacific Northwest, given the very high risk of Fusarium wilt after just one spinach seed crop (Gatch 2013). Factors that account for the spike in soilborne inoculum of F. oxysporum f. sp. spinaciae after a spinach seed crop could be evaluated using these real-time PCR assays. Spinach seed crops must be taken to full senescence for the seed to mature and dry, and the plants are swathed into windrows to finish drying the seed (Foss and Jones 2005). F. oxysporum f. sp. spinaciae colonizes the dying roots, crowns, and stems during crop senescence. After harvest of a spinach seed crop, the residues are incorporated into the soil, homogenizing inoculum throughout the field. Furthermore, some crops grown in rotation with spinach may favor persistence of F. oxysporum f. sp. spinaciae because of asymptomatic colonization of the roots and crowns of these plant species (Banihashemi and DeZeeuw 1973, 1975; Gordon and Okamoto 1990; Henry et al. 2019). The real-time PCR assays could facilitate addressing these epidemiological aspects of Fusarium wilt of spinach.

Production of high-quality spinach seed only takes place in about a half-dozen locations in the world that have the unique climatic conditions required for these seed crops (Foss and Jones 2005; Metzger and Zeevaart 1985; Rackham 2002), whereas vegetative spinach crops are grown in far more locations around the globe. Furthermore, with the popularity of baby leaf spinach products and the intensity of seeding rates for baby leaf production (up to 9 million seed/ha), large volumes of spinach seed are moved globally from the high latitude regions of seed production to regions of vegetative crop production. Since F. oxysporum f. sp. spinaciae can be seedborne in spinach, and other, nonpathogenic Fusaria can also be found on spinach seed, the real-time PCR assays may prove useful for differentiating inoculum of the spinach pathogen from other Fusarium isolates on spinach seed, and quantifying seedborne inoculum of F. oxysporum f. sp. spinaciae. Currently, seed health tests followed by fungal isolation and pathogenicity tests are the time-intensive standard for quantifying F. oxysporum f. sp. spinaciae on spinach seed (Bassi and Goode 1978; du Toit et al. 2005; Olesen et al. 2011). Further research is needed to evaluate and optimize these real-time PCR assays for differentiating isolates of F. oxysporum f. sp. spinaciae from other isolates of Fusarium found on spinach seed, and for quantifying DNA of the pathogen on seed. The current, resource-intensive method of quantifying the pathogen should be evaluated with a diversity of spinach seed lots and compared with the ability to detect and quantify DNA of the pathogen from samples of the same seed lots, following best practices for using PCR assays in seed health tests, as described by the International Seed Health Initiative of the International Seed Federation (ISF 2019).

The real-time PCR assays also potentially could be used to estimate the proportion of each race of F. oxysporum f. sp. spinaciae present in plant, soil, and/or seed samples. However, this can be achieved only if the number of the two candidate effector genes is consistent among isolates of the same race. Among the 23 race 2 isolates of F. oxysporum f. sp. spinaciae evaluated in this study, the estimated ratio of the copy number between Fos2 and Fos1_2 ranged from ∼1:1 to ∼1:4, i.e., for every copy of Fos2 there was a range of one to four copies of Fos1_2 in the genome, depending on the isolate. Thus, it may not be possible to estimate the proportion of each race of F. oxysporum f. sp. spinaciae reliably because of the variation in copy number of each gene of the pathogen. Nonetheless, the two real-time PCR assays effectively distinguished DNA extracted from plant samples infected with only race 1 isolates versus samples with mixed infections of both races or only race 2 isolates. In addition to the two effectors used to design these real-time PCR assays, identifying a locus (loci) unique to race 1 isolates would permit more accurate quantification of both races of the pathogen. However, the real-time PCR assays were very effective at identifying mixed infections of spinach by races of the spinach Fusarium wilt pathogen, as demonstrated with the spinach infection time course trial.

In summary, the real-time PCR assays developed in this study enabled rapid detection of isolates of F. oxysporum f. sp. spinaciae and distinguished isolates of the two races, saving time and resources needed for the traditional method of identification of isolates of this recalcitrant pathogen of spinach. Research is needed to assess the potential to use these assays for quantifying inoculum of the pathogen in soil and seed. Research on the epidemiology of F. oxysporum f. sp. spinaciae and the efficacy of management practices for spinach Fusarium wilt will benefit from access to these real-time PCR assays.

Acknowledgments

The authors acknowledge Eliza Mae Atterbury, Michael Derie, Babette Gundersen, Marilen Nampijja, Ryan Solemslie, Tomasita Villarroel, and Kayla Spawton for superb technical support. The authors thank Cynthia Gleason at Washington State University and Mathieu Pel at Enza Zaden for their thoughtful review of this manuscript prior to submission to the journal.