Skip main navigation
Close Drawer MenuOpen Drawer Menu
Previous
Next
MycologyOpen Access icon OPENOpen Access license

The Three Lineages of the Diploid Hybrid Verticillium longisporum Differ in Virulence and Pathogenicity

    Affiliations
    Authors and Affiliations
    • Fluturë Novakazi
    • Patrik Inderbitzin
    • German Sandoya
    • Ryan J. Hayes
    • Andreas von Tiedemann
    • Krishna V. Subbarao
    1. First and fifth authors: Plant Pathology and Crop Protection Division, Department of Crop Sciences, Faculty of Agriculture, Georg-August University of Göttingen, Grisebachstraße 6, D-37077 Göttingen, Germany; second and seventh authors: Department of Plant Pathology, and third author: Genome Center and Department of Plant Sciences, University of California, Davis; and fourth author: United States Department of Agriculture–Agricultural Research Service, 1636 E. Alisal Street, Salinas, CA 93905.
    Published Online:4 May 2015https://doi.org/10.1094/PHYTO-10-14-0265-R

    Abstract

    Verticillium longisporum is an economically important vascular pathogen of Brassicaceae crops in different parts of the world. V. longisporum is a diploid hybrid that consists of three different lineages, each of which originated from a separate hybridization event between two different sets of parental species. We used 20 isolates representing the three V. longisporum lineages and the relative V. dahliae, and performed pathogenicity tests on 11 different hosts, including artichoke, cabbage, cauliflower, cotton, eggplant, horseradish, lettuce, linseed, oilseed rape (canola), tomato, and watermelon. V. longisporum was overall more virulent on the Brassicaceae crops than V. dahliae, which was more virulent than V. longisporum across the non-Brassicaceae crops. There were differences in virulence between the three V. longisporum lineages. V. longisporum lineage A1/D1 was the most virulent lineage on oilseed rape, and V. longisporum lineage A1/D2 was the most virulent lineage on cabbage and horseradish. We also found that on the non-Brassicaceae hosts eggplant, tomato, lettuce, and watermelon, V. longisporum was more or equally virulent than V. dahliae. This suggests that V. longisporum may have a wider potential host range than currently appreciated.

    Verticillium longisporum is a fungal plant pathogen that infects the vasculature (9) of Brassicaceae crops (16). One of the most severely affected hosts is oilseed rape (canola) in Europe, specifically France, Germany, Russia, and Sweden (14), where disease incidence of up to 80% (7) and yield losses of up to 50% (8) were reported. Other hosts include broccoli in Italy; cabbage in Japan and the United States (California); cauliflower in Belgium, Germany, Netherlands, and the United States (California); field mustard in Italy, Japan, and Sweden; horseradish in Japan and the United States (Illinois); and radish and turnip in Japan (14). Sugar beet in Sweden is the only non-Brassicaceae host from which V. longisporum has been isolated (14). V. dahliae, a close relative (14,15), causes Verticillium wilt (11) on a variety of different crops (16). Similar symptoms can result from infection by V. longisporum (27) but, in oilseed rape, V. longisporum induces premature ripening and early senescence (25), while wilting symptoms are absent (10).

    Once established, V. dahliae and V. longisporum are difficult to eradicate (23). Verticillium spp. are soilborne and form microsclerotia, which are clusters of melanized cells that survive in the absence of a host for at least 14 years (30). In the proximity of a host root and in response to root exudates, microsclerotia germinate, and the hyphae penetrate and colonize the root vasculature, where they spread and proliferate into the plant vessels.

    V. longisporum has an unusual evolutionary history. It is a diploid hybrid that evolved at least three different times, by hybridization between four ancestral lineages representing three different species (15). The ancestral lineages include the unknown, hypothetical Species A1 and Species D1, the V. dahliae lineage D2, and the V. dahliae lineage D3. Species A1 hybridized with Species D1, V. dahliae lineage D2, and V. dahliae lineage D3, resulting in the V. longisporum lineages A1/D1, A1/D2, and A1/D3, respectively.

    Hybridization is known to generate new plant pathogens (3), including V. longisporum, which is more virulent on oilseed rape than V. dahliae (9). Due to the involvement of different ancestors and potential lineage-specific genetic and epigenetic changes following hybridization (22), differences in pathogenicity and virulence may also exist between V. longisporum lineages. This is supported by the fact that, among V. longisporum lineage A1/D1 and A1/D3 isolates included in pathogenicity tests on cauliflower and oilseed rape, the V. longisporum lineage A1/D3 isolates were the least virulent group (15,32,33). However, only two V. longisporum lineage A1/D3 isolates were used, and lineage A1/D2 isolates were missing in these studies.

    Therefore, the objectives of this study were to test whether the three V. longisporum lineages differ with respect to pathogenicity and virulence. We performed greenhouse pathogenicity assays with 20 V. longisporum and V. dahliae isolates on 11 different hosts, which included Brassicaceae hosts and other important crops.

    MATERIALS AND METHODS

    Pathogenicity and virulence of 20 V. dahliae and V. longisporum isolates were determined on 11 hosts. Artichoke, lettuce (Asteraceae), cotton (Malvaceae), eggplant, tomato (Solanaceae), and watermelon (Cucurbitaceae) experiments were performed at the University of California (UC) Davis and cabbage, cauliflower, horseradish, oilseed rape (Brassicaceae), and linseed (Linaceae) experiments at the University of Göttingen, Germany.

    Fungal strains and inoculum production.

    Information about the 20 fungal isolates used in the experiments including host and geographic origin is provided in Table 1. Isolates were previously identified based on DNA sequencing and phylogenetic analyses as V. longisporum lineages A1/D1, A1/D2, or A1/D3 and V. dahliae lineage D2 or V. dahliae main group by Inderbitzin et al. (15), and Tran et al. (28) for strain PD639 (=Vl43).

    TABLE 1. Verticillium strains used in this study

    To inoculate artichoke, cotton, eggplant, lettuce, tomato, and watermelon, all isolates were grown in 10-cm Petri plates on potato dextrose agar medium (PDA) (Sigma-Aldrich, St. Louis) at room temperature (24 ± 2°C) in the dark for approximately 6 weeks. Conidia stock suspensions were prepared by pouring sterile distilled water into plates and dislodging conidia with a sterile pipette tip. Stock volumes and concentrations were 50 ml and >20 × 106 conidia ml−1, respectively, as determined by using a hemocytometer. Stocks were prepared the day before inoculation and stored at 4°C overnight.

    To inoculate cabbage, cauliflower, horseradish, linseed, and oilseed rape, isolates were grown on PDA (Carl Roth GmbH + Co KG, Karlsruhe) in Petri dishes for 10 days at room temperature. Two to three agar discs (5 mm in diameter) were transferred into 500-ml Erlenmeyer flasks containing 250 ml of sterile potato dextrose broth medium (Carl Roth GmbH + Co KG) and incubated for 2 weeks at 22°C under constant shaking (120 rpm) in the dark. Conidia suspensions were prepared the day of inoculation. Conidia were harvested by filtering through sterile gauze into 100-ml Erlenmeyer flasks. Spore concentrations were determined by using a hemocytometer (Thoma chamber) and adjusted to 1 × 106 conidia ml−1 with sterile tap water.

    Seedling production and inoculation.

    Eleven different hosts were used, and the cultivar of each host is listed in Table 2. Artichoke, cotton, eggplant, lettuce, tomato, and watermelon experiments were performed at the UC Davis Contained Research Facility, a BSL3-P facility. For each species, one to two 128-plug trays (item CN-PLG 128; International Greenhouse Company, Danville, IL) containing autoclaved Sunshine Basic Mix number 2 potting mix (Sun Gro Horticulture, Agawam, MA) were seeded. Trays were incubated on a greenhouse bench with natural daylight at 21 ± 5°C, covered with sterilized cheesecloth until germination, and watered daily. Artichoke, cotton, eggplant, lettuce, and watermelon seedlings were grown for 4 weeks and tomato seedlings for 3 weeks prior to inoculation. During the first run of the experiment, fertilizer was applied daily for the first 2 weeks. During the second run, fertilizer was applied upon seeding and 2 weeks thereafter. Just prior to inoculation, seedlings were gently uprooted, rinsed free of soil in tap water, and immersed in 200 ml of a conidial suspension (2 × 106 conidia ml−1) in 500-ml beakers for at least 15 min in groups of four representing the four replications of each of the 114 host–fungal strain combinations (6 hosts, 19 fungal strains). Inoculations were performed separately for each host species. Conidial suspensions were replaced approximately halfway through the inoculations to ensure that all hosts were inoculated with nearly the same conidial densities. Seedlings were planted into 0.5-liter (= 16 oz.) Styrofoam cups (item 16J16; Smart and Final, Inc., Commerce, CA) containing autoclaved Sunshine Basic Mix number 2. Three holes (7 mm in diameter) were punched into the bottom of each cup for drainage. An equal number of uninoculated controls were planted for each host. The experiment was conducted twice, with inoculations performed on 3 March 2012 and on 30 and 31 May 2012. Inoculated seedlings were returned to a greenhouse bench and arranged in a completely randomized block design, maintained for 2 months under daylight at 21 ± 5°C, and watered daily for the first week and every second day thereafter. Seedlings were watered individually by hand without splashing to minimize the risk of cross contamination. Fertilizer was applied daily for the first 3 weeks during the first run of the experiment, and upon planting and every second week thereafter during the repetition of the experiment. The exact amount of fertilizer applied is unknown but the fertilizer stock consisted of CaNO3, Grow More (Grow More Inc., Gardena, CA), KNO3, MgSO4, and MKP and was in line with recommended rates for these crops.

    TABLE 2. Hosts used in this study

    Cabbage, cauliflower, horseradish, linseed, and oilseed rape seedlings were produced in a University of Göttingen climate chamber and greenhouse. Seed of cabbage, linseed, and oilseed rape were surface sterilized with 70% ethanol for 5 s and sown in sterile silica sand in seed trays (11 by 11 by 6 cm). Cauliflower seedlings were treated the same way except that a 2:1 soil-sand mixture was used. Prior to inoculation, plants were incubated for 14 days in a climate chamber at 22°C, 70% relative humidity, and illuminated 14 h/day with a light source (Osram Lumilux T8 L58W/21-840) placed approximately 100 cm above the plants (light intensity: approximately 132 μmol m−2 s−1). Horseradish seedlings were generated vegetatively from fully developed plants approximately 30 to 40 cm in height (three seedlings per mature plant) and planted in pots (9 by 9 by 9 cm) in a greenhouse using an autoclaved 3:2:1 mixture of compost soil-sand-turf. The horseradish seedlings were grown for 4 weeks in a greenhouse at 22°C with a day-and-night cycle of 14 and 10 h, respectively, and a light intensity of approximately 82 μmol m−2 s−1 achieved by a light source (Osram Lumilux 54W/840) placed at approximately 150 cm from the plants.

    Cabbage, cauliflower, linseed, and oilseed rape seedlings were inoculated by a cut-root dip inoculation method (31,32). This involved careful uprooting of 14-day-old plants and removal of sand and soil particles under running tap water. Approximately one-third of the taproot was cut off and seedlings were dipped into a conidial suspension (1 × 106 conidia ml−1) for 30 min. Control plants were dipped into tap water for 30 min. After inoculation, seedlings were transplanted into autoclaved 3:2:1 compost soil-sand-turf mixture in pots (7 by 7 by 7 cm), with two seedlings in each pot. Inoculations with the 19 Verticillium isolates were performed separately for each host. Twenty seedlings per pathogen isolate were inoculated concurrently. Twenty seedlings per host served as control treatment. All plants were incubated for 6 weeks following inoculation.

    Horseradish inoculations differed between the first and second experimental repetitions. The plants of the first repetition were first inoculated using the root dip method (31,32) without clipping of the roots. Because of a lack of foliar symptoms, the same plants were reinoculated by injecting a maximum amount of conidia suspension into the upper part of the root with a 20-gauge hypodermic needle and a 2-ml syringe (6,19). Tap water was injected into control plants. Plants of the second repetition were inoculated using root dip inoculation only. Only 10 Verticillium isolates and 12 horseradish seedlings were used for this experiment (Table 1). Inoculated horseradish plants were incubated for 9 weeks. Horseradish plants were fertilized weekly with a 1% NPK solution (15:11:15%).

    Because 19 fungal isolates were tested on cabbage, cauliflower, linseed, and oilseed rape and 10 isolates on horseradish, there were 91 treatment combinations (isolates–plant species), including water controls. Treatments were arranged in a completely randomized block design with 20 replications for cabbage, cauliflower, linseed, and oilseed rape and 12 for horseradish. Plants were incubated under the same conditions as prior to inoculation. All experiments were performed twice.

    Disease assessment.

    Artichoke, cotton, eggplant, lettuce, tomato, and watermelon.

    Cups were removed and root systems were cleared of soil by shaking. External symptoms were noted. Roots and shoots were cut open with a scalpel parallel to the long axis to evaluate vascular discoloration, and a disease score of 0 to 5 was assigned to each plant. The disease severity rating system was based on Vallad et al. (29) but generalized to fit all hosts as follows: 0 = vascular discoloration absent; 1 = discoloration of taproot tip only; 2 = discoloration of the taproot not beyond crown; 3 = discoloration beyond crown into shoot; 4 = discoloration throughout taproot and shoot, up to several leaves wilted, plants possibly stunted; and 5 = discoloration throughout taproot and stem, all leaves wilted or plant dead. Koch’s postulates were performed on one inoculated and one control plant per host–isolate combination for the first repetition of the experiment. Side roots and remaining soil were removed and taproots were cut into five pieces that were sterilized in 10% bleach (sodium hypochlorite 0.6%) for 1 min, rinsed twice in sterilized water, dried on sterile paper towels, plated onto Verticillium semiselective NP-10 medium (18,26), and examined for the presence of microsclerotia in culture medium using a dissecting microscope 1 to 2 weeks after plating.

    Cauliflower, cabbage, horseradish, linseed, and oilseed rape.

    Vascular discoloration was assessed as a measure of disease severity. Cauliflower, cabbage, linseed, and oilseed rape plants were removed from the pots, soil was washed off under running tap water and roots were sliced vertically with a scalpel. Discoloration was rated as follows, according to Vallad et al. (29): 0 = no vascular discoloration, 1 = 25% of vascular tissues in tap root exhibiting discoloration, 2 = 26 to 50% of vascular tissues in tap root exhibiting discoloration, 3 = 51 to 75% of vascular tissues in tap root exhibiting discoloration, 4 = 76 to 100% of vascular tissues in tap root exhibiting discoloration, and 5 = 100% of vascular tissues in tap root exhibiting discoloration extending into crown.

    For horseradish, roots were cut horizontally with a sharp knife approximately 1.5 to 2 cm below the hypocotyl, and the vascular discoloration of root-cross sections was rated following the method by Khan et al. (20).

    Statistical analysis.

    The Verticillium disease severity scores represented ordinal data that were analyzed using a nonparametric approach where the effects of experimental factors, including fungal strain and host, on disease severity scores were quantified through estimated relative marginal effects (RME) (5,24). Null hypotheses of treatments having no effect were tested using the analysis of variance-type statistic (ATS). Linear contrasts were constructed to compare disease severities between groups of isolates representing the V. longisporum lineages and V. dahliae, and to assess susceptibility to Verticillium infection between different hosts.

    Calculations were performed using the PROC Mixed procedure in SAS (2010, version 9.3) to generate RME and the LD_CI macro to generate 95% confidence intervals (5,24). The RME values are generated by the equation RME = (R − 0.5)/N, where R is the mean treatment ranking and N is the total number of experimental units in the analysis. Replications nested within experiments were considered random effects in the analysis.

    The UC Davis and the University of Göttingen data were analyzed separately. Differences between isolates and hosts (fixed effects) were determined for the UC Davis crops artichoke, cotton, eggplant, lettuce, tomato, and watermelon and for the University of Göttingen crops cabbage, cauliflower, horseradish, linseed, and oilseed rape. For each crop, differences between individual isolates, between V. longisporum lineages, and between V. longisporum lineages and V. dahliae were determined. The same differences were also assessed for the collective Brassicaceae crops and the collective non-Brassicaceae crops, except linseed, which was excluded because linseed scores were obtained at the University of Göttingen with a different rating scale.

    RESULTS

    Artichoke, cotton, eggplant, lettuce, tomato, and watermelon.

    Verticillium disease symptoms were observed on all hosts. Mean disease incidence, which is the percentage of plants with Verticillium disease symptoms, ranged from 99% in eggplant and watermelon to 53% in tomato. Mean disease scores across all isolates were highest for eggplant and watermelon, with 4.0 and 4.1, respectively, and lowest for tomato, with 1.0. The mean disease incidence across all hosts was 81% and the mean disease score was 2.6 (Table 3). RME of disease severity of hosts across all isolates of V. dahliae and V. longisporum showed significant differences (Fig. 1). The ATS of the RME showed that the effect of host was significant (P < 0.0001) (Table 4).

    TABLE 3. Mean disease incidence and disease score for the University of California (UC) Davis and University of Göttingen (UGoe) hosts across the two experimental replicates

    Fig. 1.

    Fig. 1. Estimated relative marginal effects of Verticillium longisporum and V. dahliae disease severity scores for individual crops in the University of California Davis (top) and University of Göttingen (bottom) experiments; 95% confidence intervals are indicated by vertical lines.

    Download as PowerPoint

    TABLE 4. Analysis of variance (ANOVA) using the ANOVA-type statistic on ranked means of disease severity from the University of California (UC) Davis hosts and the University of Göttingen hostsa

    There were differences in virulence between isolates combined across all hosts tested at UC Davis. V. dahliae strain PD322 was the most virulent isolate, with a mean disease score of 3.7, whereas V. longisporum lineage A1/D1 strain PD644 was the least virulent isolate, with a mean disease score of 2.2 (Table 5). Virulence score RME of individual isolates across all hosts is shown in Figure 2, and indicates significant differences between some isolates. The ATS of the RME shows that the effect of isolate was significant (P < 0.0001) (Table 4).

    TABLE 5. Mean disease scores for Verticillium dahliae and V. longisporum isolates across all University of California (UC) Davis and University of Göttingen (UGoe) hosts

    Fig. 2.

    Fig. 2. Estimated relative marginal effects of virulence across all individual Verticillium dahliae and V. longisporum isolates grouped according to affiliation with V. longisporum lineages A1/D1, A1/D2, and A1/D3 and V. dahliae, for the University of California (UC) Davis hosts (top) and the University of Göttingen hosts (bottom); 95% confidence intervals are provided. UC Davis hosts are artichoke, cotton, eggplant, lettuce, tomato, and watermelon; University of Göttingen hosts are cabbage, cauliflower, horseradish, linseed, and oilseed rape.

    Download as PowerPoint

    On individual hosts, the highest disease score means observed were 5 for V. dahliae strains PD322 on watermelon and PD502 on artichoke. V. longisporum lineages A1/D1 strains PD348 and PD643 on eggplant and PD638 and on watermelon also had disease score means of 5. The lowest score was 0, by V. longisporum lineage A1/D1 strains PD348 and PD644 on tomato (Supplementary Table S1). The isolate RME on individual crops are shown in Figure 3. The effect of isolate was significant for all hosts (P < 0.0001) (Supplementary Table S2).

    Fig. 3.

    Fig. 3. Estimated relative marginal effects of virulence of individual Verticillium dahliae and Verticillium longisporum isolates on University of California (UC) Davis hosts grouped according to affiliation with V. longisporum lineages A1/D1, A1/D2, and A1/D3 and V. dahlia; 95% confidence intervals are provided. UC Davis hosts are artichoke, cotton, eggplant, lettuce, tomato, and watermelon.

    Download as PowerPoint

    There were significant differences in virulence between Verticillium spp. and lineages across all hosts combined (P < 0.0001) (Table 6). V. dahliae was the most virulent group, with an overall disease score of 3.3, which was higher than the scores of the V. longisporum lineages, which varied between 2.4 and 2.5 (Table 7; Fig. 4).

    TABLE 6. Analysis of variance (ANOVA) using the ANOVA-type statistic on ranked means of disease severity grouped according to Verticillium dahliae and V. longisporum lineages across the University of California (UC) Davis and the University of Göttingen hosts

    TABLE 7. Mean disease scores for Verticillium dahliae and V. longisporum lineages across all University of California (UC) Davis and University of Göttingen hosts

    Fig. 4.

    Fig. 4. Estimated relative marginal effects of virulence across all University of California (UC) Davis (top) and University of Göttingen (bottom) hosts of Verticillium dahliae and V. longisporum lineages; 95% confidence intervals are provided. UC Davis hosts are artichoke, cotton, eggplant, lettuce, tomato, and watermelon; University of Göttingen hosts are cabbage, cauliflower, horseradish, linseed, and oilseed rape.

    Download as PowerPoint

    On individual hosts, V. dahliae was significantly more virulent than V. longisporum on artichoke and cotton (P < 0.0001) but not significantly different from V. longisporum lineage A1/D1 on watermelon (P = 0.57), V. longisporum lineages A1/D2 and A1/D3 on lettuce (P > 0.14), and V. longisporum lineage A1/D2 on tomato (P = 0.09). V. dahliae was significantly less virulent than V. longisporum lineage A1/D1 on eggplant (P < 0.0001) (Table 8; Fig. 5).

    TABLE 8. Contrasts between disease severity of Verticillium dahliae and V. longisporum lineages on individual University of California (UC) Davis hosts, University of Göttingen (UGoe) hosts, and the Brassicaceae hosts onlya

    Fig. 5.

    Fig. 5. Estimated relative marginal effects of virulence on individual University of California Davis hosts of Verticillium dahliae and V. longisporum lineages; 95% confidence intervals are provided.

    Download as PowerPoint

    Within species and lineages across all hosts combined, there were no significant virulence differences between isolates, except in V. dahliae, where strain PD322 was more virulent than strain PD637 (Fig. 2). On individual hosts within species and lineages, there were virulence differences between isolates but no single isolate was consistently most or least virulent across all hosts (Fig. 3). This agrees with the significant host–isolate interaction (Table 4), which indicates that isolate virulence is host dependent.

    Cabbage, cauliflower, horseradish, linseed, and oilseed rape.

    Verticillium disease symptoms were observed on all hosts. Mean disease incidence ranged from 87% in cauliflower to 32% in linseed. Mean disease scores across all isolates were highest for horseradish and oilseed rape, with 1.8, and lowest for linseed, with 0.8. The mean disease incidence across all hosts was 68% and the mean disease score was 1.4 (Table 3). RME suggested significant differences in the response of hosts to all strains of V. dahliae and V. longisporum (Fig. 1). The ATS of the RME showed that the effect of host was significant (P < 0.0001) (Table 4).

    There were differences in virulence between isolates across all hosts combined. V. longisporum lineage A1/D3 strain PD589 was overall the most virulent isolate, with a mean disease score of 2.7, whereas V. dahliae strains PD327 and PD502 were least virulent, with a mean disease score of 0.6 (Table 5). Virulence score RME of individual isolates (Fig. 2) across all hosts indicated significant differences between some isolates. The ATS of the RME shows that the effect of isolate was significant (P < 0.0001) (Table 4).

    On individual hosts, the highest disease score was 4.6 for V. longisporum lineage A1/D3 strain PD589 on oilseed rape. The lowest score was 0 by V. dahliae strains PD322 and PD327 and V. longisporum lineage A1/D1 strain PD588, all on linseed (Supplementary Table S3). The isolate RME on individual crops are shown in Figure 6. The effect of isolate was significant for all hosts (P < 0.0001) (Supplementary Table S4).

    Fig. 6.

    Fig. 6. Estimated relative marginal effects of virulence of individual Verticillium dahliae and V. longisporum isolates on University of Göttingen hosts grouped according to affiliation with V. longisporum lineages A1/D1, A1/D2, and A1/D3 and V. dahliae; 95% confidence intervals are provided.

    Download as PowerPoint

    There were significant differences in virulence between Verticillium spp. and lineages across all hosts combined (Table 7). V. longisporum lineage A1/D1 was the most virulent group, with an overall disease score of 1.7, followed by V. longisporum lineages A1/D3 and A1/D2, with scores of 1.4 and 1.2, respectively, which was significantly higher than the score of V. dahliae (P < 0.0001), which was 0.8 (Table 7; Fig. 4).

    On individual hosts, V. longisporum lineage A1/D1 was significantly more virulent than other lineages and V. dahliae on cauliflower (P = 0.02) and oilseed rape (P < 0.0001), and equally virulent as V. longisporum lineage A1/D2 on linseed (P = 0.0885). V. longisporum lineage A1/D2 was significantly more virulent than other lineages and V. dahliae on cabbage (P < 0.0001) and horseradish (P < 0.0001) (Table 8; Fig. 7).

    Fig. 7.

    Fig. 7. Estimated relative marginal effects of virulence on individual University of Göttingen hosts of Verticillium dahliae and V. longisporum lineages; 95% confidence intervals are provided.

    Download as PowerPoint

    Within species and lineages across all hosts combined and on individual hosts, there were virulence differences between isolates within all V. longisporum lineages (Fig. 6). No single isolate was most or least virulent across all individual crops, as illustrated by the significant host–isolate interaction term in Table 4.

    Brassicaceae versus non-Brassicaceae hosts.

    To illustrate susceptibility of Brassicaceae and non-Brassicaceae hosts against V. dahliae and V. longisporum, we generated a Brassicaceae dataset that comprised the data from cabbage, cauliflower, horseradish, and oilseed rape. The effect of lineage and species was significant in the Brassicaceae dataset (Supplementary Table S5). Across all Brassicaceae hosts combined, V. longisporum lineage A1/D1 was the most virulent group and V. dahliae was least virulent (Table 8). This was contrasted with the non-Brassicaceae experiment involving artichoke, cotton, eggplant, lettuce, tomato, and watermelon, where V. dahliae overall was the most virulent group, and the three V. longisporum lineages were equally virulent (Fig. 8).

    Fig. 8.

    Fig. 8. Comparison of estimated relative marginal effects (RME) of virulence between Brassicaceae and non-Brassicaceae hosts for Verticillium dahliae and V. longisporum lineages; 95% confidence intervals are provided. The Brassicaceae and non-Brassicaceae data were generated in different experiments analyzed separately, so that the confidence intervals cannot be used to compare Brassicaceae and a non-Brassicaceae RME. Noncrucifer RME are from Figure 4 (top). Brassicaceae hosts are cabbage, cauliflower, horseradish and oilseed rape; non-Brassicaceae hosts are artichoke, cotton, eggplant, lettuce, tomato, and watermelon.

    Download as PowerPoint

    DISCUSSION

    We demonstrated that there are significant differences in virulence and pathogenicity among the V. longisporum lineages and between V. longisporum and V. dahliae. V. longisporum was more virulent than V. dahliae on all Brassicaceae hosts tested, including cabbage, cauliflower, horseradish, and oilseed rape, but virulence of V. longisporum depended on the host and lineage. V. longisporum lineage A1/D1 was the most virulent group across all Brassicaceae crops and, among individual crops, V. longisporum lineage A1/D1 was the most virulent group on oilseed rape and marginally more virulent on cauliflower than V. longisporum lineage A1/D3. V. longisporum lineage A1/D2 was the most virulent group on cabbage and horseradish, whereas V. longisporum lineage A1/D3 was never the most virulent lineage on any of the hosts tested. However, V. longisporum lineage A1/D3 strain PD589 was among the most virulent isolates on oilseed rape.

    Our finding that V. longisporum lineage A1/D1 overall is more virulent on cauliflower and oilseed rape than V. longisporum lineage A1/D3 agrees with Zeise and von Tiedemann (32,33), whose V. longisporum groups lsp and lsp* correspond to V. longisporum lineages A1/D1 and A1/D3, respectively (15). Babadoost et al. (1) showed that V. longisporum lineage A1/D2 isolates were virulent on horseradish. Our assays confirmed these results, and also showed that V. longisporum lineage A1/D2 was most virulent on horseradish among the V. longisporum lineages and V. dahliae. It is interesting that all V. longisporum lineage A1/D2 isolates used in our study were originally isolated from horseradish. However, the horseradish isolates were not host specific, because V. longisporum lineage A1/D2 was also the most virulent group on cabbage.

    The fact that V. longisporum was more virulent on the Brassicaceae crops than V. dahliae was expected, because V. longisporum is a pathogen of Brassicaceae, whereas V. dahliae generally affects other hosts (9). The only non-Brassicaceae host from which V. longisporum was isolated was sugar beet (16). However, in our assays, at least one of the V. longisporum lineages was more or equally virulent than V. dahliae on eggplant, tomato, lettuce, and watermelon, suggesting that, under field conditions, V. longisporum may have a wider potential host range than is currently appreciated.

    Our studies led to the following practical considerations. V. longisporum lineage A1/D1 was previously introduced into the Salinas Valley of California, where it caused Verticillium wilt on cauliflower (21,27) but has since disappeared. Our results show that V. longisporum lineage A1/D1 is also among the most virulent groups on other crops grown in California, including eggplant and watermelon, which are potential hosts should V. longisporum lineage A1/D1 be reintroduced. Another issue is Verticillium infection of oilseed rape, which is a potentially yield-limiting disease in Europe but is absent from North America (12). The disease is caused by V. longisporum lineage A1/D1, which is known from Europe, California, and Japan but has not been documented from Canada or the oilseed-rape-growing regions of the United States (15). However, V. longisporum lineage A1/D2 is known from horseradish in Illinois, which is in relative proximity to some of the major oilseed-rape-growing areas in the United States, including Minnesota, North Dakota, and Oklahoma (http://www.uscanola.com/). We were unable to include North American varieties of oilseed rape in our assays but, on the European oilseed rape varieties, V. longisporum lineage A1/D2 was the least virulent lineage of V. longisporum and, thus, V. longisporum lineage A1/D2 may not pose a risk to the North American oilseed-rape-growing regions.

    Our results lay the framework for investigations into the genetic bases of the virulence differences between the V. longisporum lineages and between V. longisporum and V. dahliae. These differences may be due to genetic changes associated with the hybridization process, as has been documented in plants (13), and also to the unique genetic contributions made to the hybrid by the individual parental lineages. Brasier (3,4) speculated that previously geographically separated nonpathogens could hybridize to form new pathogens. This appears to be the case for V. longisporum, particularly for V. longisporum lineage A1/D1, which, overall, was the most virulent lineage of V. longisporum in Brassicaceae crops. None of the ancestral species of V. longisporum lineage A1/D1, Species A1 and Species D1, have ever been found (15), suggesting that Species A1 and Species D1 are not pathogenic on any of the agricultural crops and are more likely to be associated with nonagricultural hosts, or they are nonpathogenic saprotrophs, similarly to V. nubilum (14,17). Very little is known about the natural hosts of Verticillium spp. (16). Such knowledge, particularly pertaining to Species A1 and Species D1, would be crucial to understanding the genetics of the origin of V. longisporum lineage A1/D1, one of the most serious oilseed rape pathogens today.

    ACKNOWLEDGMENTS

    We thank the following individuals for providing strains used in this research: M. Babadoost (University of Illinois, United States), the late D. Barbara (University of Warwick, United Kingdom), H. Sakai (Department of Agriculture, Gunma Prefecture, Japan), G. Stanosz (University of Wisconsin, United States), M. Typas (University of Athens, Greece), T. Usami (Chiba University, Japan), and I. Happstadius (Lantmännen Lantbruk, Svalöv, Sweden); R. M. Davis (UC Davis) for providing cotton seed; R. M. Bostock (UC Davis) for making his lab space available for this work; K. Godfrey and T. Simpson (UC Davis) for help in the greenhouse; S. Herbert, R. Marchebout, K. Maruthachalam, K. McConnell, D. Short, E. Vorbeck, J. Schaper, D. Tacke, and M. Knobel for excellent assistance in the greenhouse and laboratory; and USDA-NIFA-SCRI grant number 2010-51181-21069 and the California Leafy Greens Research Board for the funding that made these studies possible.

    LITERATURE CITED