Tomato Mi-gene Resistance-Breaking Populations of Meloidogyne Show Variable Reproduction on Susceptible and Resistant Crop Cultivars

Origin, isolation, and propagation of Meloidogyne populations: Over the course of two growing seasons, roots of plants from 16 processing tomato fields in California with Mi-gene resistant cultivars exhibiting stunted growth were dug up. Their roots were examined for root galling symptoms, and they were sent to the Department of Nematology at UC Riverside when symptomatic (Table 1). The roots were then washed free of soil, cut into approximately 1-cm-long pieces, and placed in a misting chamber (Niblack and Hussey, 1985) for 5 days for nematode extraction. The suspensions were then examined at ×40 magnification using a dissecting microscope to check for the presence of Meloidogyne second-stage juveniles (J2). If they were present, these suspensions were added to a 6-week-old tomato cv. Celebrity plant with the Mi-gene (Bayer/Seminis, St. Louis, MO) grown in steam-sterilized sand (93% sand, 4% silt, 3% clay, pH 7.1) in a 3.8-liter pot in a greenhouse. Plants were grown under natural light at soil temperatures between 21–26 °C, watered daily through an automated drip system, and fertilized with 10g of a slow-release fertilizer (Osmocote 17-6-10, Scotts, Marysville, OH) one week after transplanting. Eight weeks later, the roots of these plants were washed and examined, and if galling was present, a single egg mass was removed from the roots under a dissecting microscope at ×10 magnification and added to a 4-week-old cv. Celebrity tomato seedling growing in a 55-ml plastic cone (Stuewe and Sons, Corvallis, OR) with the same steam-sterilized sand in a greenhouse. Plants were carefully removed from the cone four weeks later and transferred to a 3.8-liter pot. They were grown for another 6 weeks, and then nematodes were extracted from the roots in a misting chamber as described above. Nematodes from these roots were used as inoculum for further multiplication on tomato cv. Celebrity plants, and roots from these cv. Celebrity plants were used as a source of inoculum for experiments and for nematode identification.

Nematode identification: RKN females were excised from galled tomato cv. Celebrity roots, and perineal patterns from five females per root system were cut for morphological identification (Riggs, 1990). Patterns were examined under ×500 magnification using a Leitz DMR compound light microscope and based on the patterns, nematodes were identified to species level (Eisenback, 1985). In addition to morphological identification, five populations that were used for further experiments were also identified through PCR. Of those, four that were identified as M. incognita based on perineal patterns were used to amplify and sequence the nad5 mtDNA gene fragment. Three μl of extracted DNA were transferred to a 0.2-ml Eppendorf tube containing: 10 μl DreamTaq Green PCR Master Mix (2×) (Thermo Fisher Scientific, Waltham, MA), 10 μl water and 0.15 μl of each primer (1.0 μg/μl). Forward NAD5F2 (5′-TAT TTT TTG TTT GAG ATA TAT TAG-3′) and reverse NAD5R1 (5′-CGT GAA TCT TGA TTT TCC ATT TTT-3′) primers as described by Janssen et al. (2016) were used in PCR. The PCR amplification profile consisted of 4 min at 94°C, followed by 40 cycles of 1 min at 94°C, 1 min at 45°C, and 1 min 30 sec at 72°C, with a final extension at 72°C for 10 min. Two μl of the PCR product were run on a 1% TAE-buffered agarose gel (100 V, 40 min). PCR products were purified using QIAquick PCR Purification Kit (Qiagen, Hilden, Germany) and sequenced directly. Sequencing was performed by Genewiz (CA, USA). New sequences were compared with reference sequences using Blastn. Sequence was submitted in the Genbank under accession number OM418740. To confirm the identity of one population that was identified based on perineal patterns as M. javanica, J2 were used in a species-specific assay for this species, as described by Dong et al. (2001), using primers Fjav 5′-CCT TAA TGT CAA CAC TAG AGC C-3′ and Rjav 5′-GGC CTT AAC CGA CAA TTA GA-3′.

Nematode multiplication and effects on susceptible and resistant plant varieties Tomato: Five nematode populations that originated from single-egg-mass cultures were randomly selected for further experiments. In a greenhouse pot trial, the multiplication and symptom expression of these populations on a range of Mi-gene-carrying resistant processing tomato cultivars was determined. Tomato cultivars used in the trial were DRI319, H8504, H5608, HM3887, and N6366. These tomato cultivars were selected based on recommendations by the California Tomato Research Institute (CTRI) as processing tomato cultivars currently grown in California (Anonymous, 2021). Seed of these cultivars was also kindly provided by CTRI. As controls, the resistant Mi-gene-carrying fresh-market cultivar Celebrity (Bayer/Seminis, St. Louis, MO) and the susceptible cultivar Daniela (Osborne Quality Seeds, Mt. Vernon, WA) were included. Also, a “standard” avirulent M. incognita race 3 population (Mi-3) originally isolated from cotton in the San Joaquin Valley, California that had been maintained on susceptible tomato in our greenhouse for several years was included as an internal control. For the trial, tomatoes were seeded in seedling trays with potting mix (Sunshine Mix 5, Sungro, Vancouver, BC, Canada) and placed in a greenhouse. Three weeks after emergence, plants were carefully removed from the seedling trays and transplanted into 1-liter white plastic cups filled with steam-sterilized sand. Ten days after transplanting, each plant was inoculated with 1,000 RKN J2 in 9 ml water by adding 3 ml each of the suspension to three 3-cm-deep holes around the base of each plant. For inoculation, pots receiving the same RKN population were placed on a bench and separated from the other pots. This was repeated for each RKN population.

The trial was set up according to a completely randomized design with five replicates (a total of 6 RKN populations × 7 tomato cultivars × 5 replicates = 210 pots). The pots were placed on a greenhouse bench with a 40-cm spacing between the pots. Plants were watered with a slow water drip to avoid the splashing associated with automated drip systems, and fertilized with 17–6-10 controlled release fertilizer (Scotts-Sierra Horticultural Products Co, Marysville, OH). Six weeks after inoculation, the plants were removed from the pots, and the roots were washed free of soil, weighed, and rated for severity of root galling (0–10 scale: 0 = no galling, 10 = 100% of root system galled). Roots were then placed in a mist chamber for 7 days to extract RKN J2. The suspensions coming off the roots were collected, and the number of J2 from each root system was counted under a dissecting microscope at ×40 magnification. Nematode reproduction is expressed as the reproduction factor RF (Pf/Pi = number of J2 per root system at harvest/number of J2 inoculated). The complete experiment was repeated once (trial 1 and 2).

Differences between the populations, tomato cultivars, and their interactive effects were analyzed statistically using R software (R Core Team, 2022). The repeated trials were considered random effects. Prior to analysis, J2 counts were transformed by x¹ = log10 (x + 1) before analysis. Where treatment effects on J2 infestation levels were significant, they were separated with Fisher's protected Least Significant Difference test at the 95% confidence level. Differences in galling levels were analyzed using the non-parametric Kruskal-Wallis test, also at the 95% confidence level. In the experiments with tomato, the Mi-3 population was not included in the statistical analysis, as it was included only as an internal standard to confirm the resistant nature of the different tomato cultivars, and we were primarily interested in differences among the resistance-breaking nematode populations.

In a series of subsequent experiments, the same nematode populations were used as inoculum on susceptible and resistant crop cultivars. The experiments were done as with tomato, with two replicated trials. Statistical analysis was as with the tomato trial. In each of the experiments, the Mi-gene-carrying resistant tomato cv. Celebrity was included as an internal control to confirm the virulence of the populations on tomato, but not included in the statistical analysis.

Pepper: The M. incognita-susceptible pepper (Capsicum annuum) cv. Baron (Seminis, Oxnard, CA) and the M. incognita-resistant cv. Carolina Wonder (Reimer Seeds, Saint Leonard, MD) were used. Experiments were as with tomato, but time between seeding and transplanting was 4 weeks.

Sweet potato and green bean: Sweetpotato (Ipomea batatas) stem cuttings of M. incognita-susceptible cv. O’Henry and M. incognita-resistant cv. Bonita were kindly donated by the California Sweetpotato Council. The cuttings were planted in vermiculite and placed in a misting chamber in a greenhouse. Plants were carefully removed from the vermiculite 10 days later, and cuttings exhibiting root growth were planted in 1-liter plastic cups with steam-sterilized sand, as with the tomato experiments. One week later, these plants were inoculated with the nematode populations. At harvest, some plants had developed small storage roots, but these were not included for J2 extraction. In the same experiment, the M. incognita-susceptible green bean (Phaseolus vulgaris) cv. BlueLake (Harris Seeds, Rochester, NY) and the M. incognita-resistant cv. Nemasnap (USDA-GRIN) were seeded directly into 1-liter plastic cups with steam-sterilized sand. Bean plants were inoculated with the nematode populations 10 days after emergence. All plants in the trial were inoculated simultaneously.

Cotton and cowpea: M. incognita-susceptible cotton (Gossypium hirsutum) cv. SJ2 (originally obtained from California Planting Cotton Seed Distributers and propagated in our greenhouse), the M. incognita-resistant cotton cv. DP 1474NR B2XF (seed kindly provided by Dr. Wheeler, Texas A&M), the M. incognita-susceptible cowpea (Vigna unguiculata) cv. CB46 Null-1 and the M. incognita-resistant cv. CB46 (seed kindly provided by Dr. Huynh, Dept. Nematology, UC Riverside) were seeded directly into 1-liter plastic cups with steam-sterilized sand. All plants were inoculated with the nematode populations simultaneously (cowpea: 14 days after emergence, cotton: 21 days after emergence).

Nematode identification: Perineal patterns of females from all nematode populations were consistent with those described for M. incognita, except for population 4, which had perineal patterns with typical double lateral lines and was identified as M. javanica. PCR-based identification on the five populations used for further experiments (including population 4 identified as M. javanica) confirmed the results from the morphological identification.

Nematode multiplication and ef fects on susceptible and resistant plant varieties Tomato: Statistical analysis showed that both “tomato cultivar” and “RKN population” had a significant effect (P ≤ 0.01) on nematode reproduction and root-galling. The interaction between these two factors, however, was not significant (P > 0.05). The avirulent Mi-3 population, included as a control, behaved as expected, with a high reproduction on the susceptible cv. Daniela and very few nematodes recovered from the resistant tomato cultivars. There were significant differences among the five resistance-breaking populations, with populations 3, 10, and 11 reproducing at higher levels than populations 4 and 12 (Table 2). There were also significant differences in the susceptibility of the different tomato cultivars towards the resistance-breaking populations.

Although all cultivars allowed a strong multiplication (RF > 3) of the resistance-breaking populations, cultivars H5608 and DRI319 were even more susceptible than the susceptible Daniela and the Mi-gene-carrying resistant cultivars Celebrity, HM3887, and HH8504 (Table 2). The severity of root-galling on tomato largely reflected the nematode multiplication data. The avirulent Mi-3 control population caused severe galling on the susceptible cv. Daniela and very minor galling on the resistant tomato cultivars. Populations 3, 10, and 11 also caused the most severe galling, while galling resulting from populations 4 and 12 was less dramatic. All cultivars exhibited moderately high galling after inoculation with the resistance-breaking populations, although minor but significant differences existed. Galling on the Mi-gene-carrying resistant cv. Celebrity was less than on the Mi-gene-carrying resistant cultivars H5608, DRI319, and HM3887 (Table 2). Fresh root weights of the tomatoes were not significantly different between the different tomato cultivars (P > 0.05; data not shown).

Pepper: In the trials with peppers, the interactive effect of “pepper cultivar” × “RKN population” on nematode reproduction and root galling was significant (P ≤ 0.01). Differences between the nematode populations within each pepper cultivar are thus shown. The tomato cv. Celebrity, included as a control, confirmed the avirulent nature of the Mi-3 population and the virulent nature of the five resistance-breaking populations on tomato. There was a significant difference among the RKN populations (P ≤ 0.01) in nematode reproduction on pepper cv. Baron (Table 3). This cultivar was highly susceptible to all populations, except to population 4, which failed to reproduce or cause any galling on this cultivar. The M. incognita-resistant pepper cv. Carolina Wonder was highly resistant to all populations, and differences among the populations were not significant (P > 0.05). Only very minimal galling was observed on the roots of cv. Carolina Wonder plants, although population 12 caused slightly more galling on this cultivar than most other populations (Table 3).

Sweetpotato and green bean: In the trials with sweetpotato and green bean, results from the internal standard – the resistant tomato cv. Celebrity – were as before, showing strong resistance against the Mi-3 population, while being highly susceptible to the five virulent (resistance-breaking) populations. The interactive effect of crop cultivar × RKN population on nematode reproduction and root-galling was highly significant (P ≤ 0.01). The M. incognita-susceptible sweetpotato cv. O’Henry and the M. incognita-susceptible bean cv. BlueLake were highly susceptible to all RKN populations, and galling on the roots of these plants at harvest was severe, regardless of the RKN population that was used as inoculum (Table 4).

Significant differences among the RKN populations existed both in the M. incognita-resistant sweetpotato cv. Bonita and the green bean cv. NemaSnap. Sweetpotato cv. Bonita was resistant to the Mi-3 control population (RF = 0.7), but susceptible (RF = 2.8–5.2) to the five populations known to be able to break resistance in tomato. This difference between Mi-3 and the five resistance-breaking populations was not reflected in the severity of root-galling on sweetpotato cv. Bonita, as this was minor with all RKN populations. The green bean cv. NemaSnap was highly susceptible to population 4, but highly resistant to the other five populations. This corresponded with the severity of root-galling on cv. NemaSnap after inoculation with the different nematode populations (Table 4).

Cotton and cowpea: Finally, in the trial with cotton and cowpea, results on the internal control tomato cv. Celebrity were again as expected, and the interactive effect of crop cultivar × RKN population on nematode reproduction and root-galling was again highly significant (P ≤ 0.01). The M. incognita-susceptible cotton cv. SJ2 was an excellent host for Mi-3 and population 12, a maintenance host for population 4, and highly resistant to populations 3, 10, and 11. Susceptible cowpea cv. CB46 Null-1 was an excellent host for all six RKN populations, although reproduction of population 3 (RF = 16.4) was significantly lower compared to the other five populations. Galling on the susceptible cotton cultivar corresponded with results regarding nematode reproduction, and galling on the susceptible cowpea cultivar was high and did not differ among the RKN populations. Resistant cotton cv. DP 1474NR B2XF prevented a nematode increase (RF < 1) of all populations except population 12, which increased twofold. Galling on the roots of the resistant cotton cultivar was very minor with all RKN populations. Two RKN populations, 4 and 12, were able to increase more than 10-fold on M. incognita-resistant cowpea cv. CB46, whereas this cultivar was resistant to the other four populations (RF < 1). Root-galling on cowpea cv. CB46 reflected these results with regards to nematode reproduction (Table 5).

Fifteen of the 16 populations isolated from fields planted with Mi-gene-carrying resistant processing tomato in California were identified as M. incognita, and one was identified as M. javanica. It is unknown if M. incognita is more common in areas where processing tomato is grown, or if resistance-breaking occurs more easily in M. incognita. Recent data on the occurrence of RKN in California are not available, but in surveys by Siddiqi et al. (1973), 19 detections of M. incognita and 14 of M. javanica were reported associated with tomato, which suggests M. incognita is only slightly more common than M. javanica on California tomato. Resistance-breaking by M. javanica was reported in California by Williamson (1998), but that population was obtained by repeated greenhouse culturing of an originally avirulent population on tomato with Mi-mediated resistance. In the only other report on resistance-breaking RKN populations from the USA outside of California, which is from Georgia, all populations were identified as M. incognita (Hajihassani et al., 2022). Thus, this is the first report of a field-isolated population of M. javanica on Mi-gene-carrying resistant tomato in the USA. Resistance-breaking field populations of M. javanica have been reported in several other countries, however (Tzortzakakis and Gowen, 1996; Eddaoudi et al., 1997).

Although all five resistance-breaking populations reproduced well on the tomato cultivars evaluated in this study, there were differences. Populations 4 and 12 had significantly lower RF values than the other three. Differences in reproduction among four resistance-breaking M. incognita populations on Mi-gene-carrying resistant tomato were also reported by Hajihassani et al. (2022). They speculated that this might be due to genetic diversity among the different nematode populations tested, and this may also explain our results. In addition to differences between the RKN populations, there were also differences among the Mi-gene-carrying resistant tomato cultivars, with two cultivars, H5608 and DRI319, being particularly susceptible. It can be assumed that the resistance of all cultivars used in this study is based on the same single dominant Mi-gene (Williamson, 1998).

Jacquet et al. (2005), evaluating a range of tomato Mi-gene-carrying cultivars, obtained similar cultivar effects, and concluded that differences in the genetic background of these cultivars influenced the level of resistance towards resistance-breaking RKN populations. Previous research (Castagnone-Sereno et al., 2007) showed that resistance-breaking in tomato by M. incognita is associated with a loss in reproductive fitness, and although we only included one avirulent M. incognita population in our study, our results are similar, as the RF of the avirulent population on the susceptible tomato cv. Daniela, was higher than of the four resistance-breaking populations.

The susceptible pepper cv. Baron was an excellent host for all M. incognita populations, but not for the M. javanica population. Pepper is generally considered a non-host for M. javanica, and is in fact used as a differential host to distinguish between M. incognita and M. javanica (Eisenback et al., 1981). The N-gene-mediated resistance in pepper cv. Carolina Wonder was highly effective in preventing nematode reproduction in all RKN populations. Thus, the populations breaking resistance in tomato were not able to overcome N-gene-mediated resistance in pepper. This is similar to results from Djian-Caporalino et al. (2011), who also reported that RKN resistance in pepper was maintained when challenged with populations virulent on Mi-gene-mediated resistant tomato.

Sweetpotato cv. O’Henry was equally and highly susceptible to all RKN populations. On the RKN-resistant sweetpotato cv. Bonita there was a clear and significant separation in nematode reproduction between the avirulent Mi-3 control population and the five resistance-breaking populations. Whereas the RF of the former population was 0.7, i.e., a decrease in nematode levels, the five resistance-breaking populations all multiplied more than twofold (RF = 2.8–5.2) on sweetpotato cv. Bonita. Thus, sweetpotato cv. Bonita can be considered resistant to the avirulent Mi-3 population, but susceptible to the five resistance-breaking populations. Galling on cv. Bonita roots was minor, however, and did not differ among the RKN populations.

In previous field studies on sweetpotato by Roberts and Scheuerman (1984) and by us (Ploeg, unpublished) a large RKN population increase in soil was observed even after one cropping cycle with a RKN-resistant cultivar. However, Roberts and Scheuerman (1984) reported that nematode-resistant sweetpotato cultivars did not develop typical RKN symptoms on the storage roots (bumpiness and cracking), and we found very large differences between the number of RKN eggs extracted from field-grown sweetpotato storage roots between susceptible cv. O’Henry (117 eggs/g storage root) and those found in resistant cv. Bonita (0.1 eggs/g storage root) (Ploeg, unpublished). Thus, it appears that RKN resistance in sweetpotato is primarily related to preventing nematode reproduction and symptom development on the storage roots (tubers), rather than nematode reproduction in the true feeder roots. The genetic basis for RKN resistance in cv. Bonita is unknown, but La Bonte et al. (2011) reported cv. Bonita to be highly resistant to Mi-3 in greenhouse studies.

Green bean cv. Blue Lake was an excellent host for all RKN populations, while the resistant cv. NemaSnap was highly resistant to all M. incognita populations, yet susceptible to M. javanica (population 4). Resistance of cv. NemaSnap to M. incognita was reported previously (Wyatt et al., 1983; Mullin et al., 1991), but information on M. javanica is limited to cv. NemaSnap being moderately resistant to a mixed population of M. javanica and M. incognita.

The host status of the susceptible cotton cv. SJ2 differed depending on the RKN population. It was a host for Mi-3, M. incognita population 12 and M. javanica (population 4). Meloidogyne incognita populations can be categorized as belonging to four distinct races, with populations able to reproduce on cotton belonging to race 3 or 4 (Barker et al., 1985). The race 3 designation of the control population corresponds with this result. Population 12 should be designated as a race 3 or race 4 population.

We did not include tobacco in our test, a differential host for M. incognita race 3 versus race 4. Resistant cotton cv. DP 1474NR B2XF prevented a nematode increase in all populations except of M. incognita population 12, which on average doubled on this cultivar.

The susceptible cowpea cv. CB46 Null-1 was an excellent host for all RKN populations. The resistant cowpea cv. CB46 was a good host for M. javanica (population 4), and for M. incognita population 12. The resistance in cowpea cv. CB46 is based on the Rk gene, and although resistance to M. incognita is generally strong, variability in the effectiveness of its resistance towards different M. incognita populations has been documented (Huynh et al., 2016). Furthermore, Rk-gene-mediated resistance in cowpea is generally less effective against avirulent populations of M. javanica than of M. incognita (Huynh et al., 2016).

Overall, it can be concluded that RKN populations able to break Mi-gene resistance in tomato are common throughout tomato-growing areas in California. We observed minor variability in the reproductive potential among these populations on resistant tomato cultivars, as well as some variability in resistance towards them among different Mi-gene-carrying tomato cultivars. The statement by Djian-Caporalino et al. (2011) that “RKN isolates virulent on one resistant crop are definitely not virulent on a different crop” may not be always true, as in our study only those populations virulent on Mi-gene-carrying resistant tomato were able to reproduce on the resistant sweetpotato cv. Bonita. This clear separation between the avirulent Mi-3 population and the resistance-breaking populations was not observed on the other crops tested.

This study also demonstrates the variability in virulence among the resistance-breaking populations when inoculated onto other resistant crop cultivars, as significant differences in reproduction were observed among the RKN populations on resistant bean, cotton and cowpea. The pepper cv. Carolina Wonder was the only crop cultivar that was highly resistant to all the RKN populations used in this study, and this crop may have potential for use in rotation with tomato to reduce levels of resistance-breaking RKN populations. This study shows that relying solely on one strategy, such as the use of host plant resistance to manage RKN, may lose its effectiveness in the long run. It also stresses the importance of alternating or integrating different management strategies, such as chemical, biological, or cultural control, in order to maintain the required reductions in population levels over time.