Transfer of Meloidogyne incognita Resistance Using Marker-assisted Selection in Sorghum

Genotyping of each BC₁F₆ line with microsatellite markers in the QTL-Sb.RKN.3.1 region confirmed that each line contained DNA from Honey Drip in this region (Table 3, Fig. 1). GT-IR7-BC1F6 and Top 76-6-BC1F6 were both heterozygous for QTL-Sb.RKN.3.1 marker RKNP529, which is at 54,678,311 bp on the sorghum chromosome 3 genomic sequence (Table 3).

Figure 1:

The fresh weights of the sorghum root systems differed among genotypes in both trials (Table 4). However, there was a significant statistical Trial × Genotype interaction (P = 0.0190), so each trial was analyzed separately; the interaction appears to have been due to greater differences in root weight in Trial 2 rather than to inconsistent relative performance of the genotypes. In both trials, Entry 22 had the lowest numerical root weight and Top 76-6 had the greatest numerical root weight (Table 4). In Trial 1, the root weight of each BC₁F₆ genotype was statistically similar to its recurrent parent (Table 4). However, in Trial 2, Top 76-6-BC1F6 had lower root weight than Top 76-6, whereas the other BC₁F₆ genotypes did not differ from their recurrent parent (Table 4).

There was no Trial × Genotype interaction for the total number of nematode eggs produced, therefore, the trials were combined for a pooled analysis. The number of nematode eggs produced differed among genotypes, with the greatest number of eggs (numerically) produced on Dale and the fewest on GT-IR7-BC1F6 (Table 5). The recurrent parental genotypes (Dale, Top 76-6, Collier, GT-IR7, and Entry 22) were all statistically similar to each other, and the BC₁F₆ genotypes (Dale-BC1F6, Top 76-6-BC1F6, Collier-BC1F6, GT-IR7-BC1F6, and Entry 22-BC1F6) were similar to each other and to the resistant standard, Honey Drip (Table 5).

Because root weights differed among genotypes, nematode reproduction per gram of root was also calculated. Although the mean number of eggs per gram of root changed the numerical ranking of the genotypes, the results were similar to those for the total number of eggs produced. The recurrent parental genotypes were all statistically similar to each other, and the BC₁F₆ genotypes were similar to each other and to the resistant standard, Honey Drip (Table 5).

Host-plant resistance to a nematode species is a relative term that is based on comparing the level of nematode reproduction on a plant genotype to the level of reproduction on a designated susceptible genotype of the same species. A genotype that reduces reproduction by 90% is typically acknowledged as highly resistant, whereas smaller reductions are often called partially or moderately resistant (Davis and Stetina, 2016; Hussey and Janssen, 2002). Levels of Meloidogyne spp. reproduction can be documented in two different ways: total number of eggs produced and eggs per gram of root produced. Standardizing the amount of reproduction on a per gram of root basis is often used when the plants being evaluated have large differences in root mass. However, the two measurements address different questions. The total reproduction is useful for evaluating whether a genotype is likely to suppress nematode levels in a field and therefore be beneficial to a subsequent susceptible crop. Reproduction per gram of root is useful for evaluating the parasitic load on that plant (Davis and Stetina, 2016), which should be correlated with the amount of damage caused to that crop. In the study reported herein, all BC₁F₆ genotypes would be considered highly resistant to M. incognita based on either total reproduction or reproduction per gram of root. Therefore, we conclude that all of the BC₁F₆ genotypes should suffer little or no damage from M. incognita and also be effective at suppressing M. incognita in the field thereby serving as an effective rotation crop.

Movement of favorable alleles at QTL regions into different plant backgrounds is an important next step after the identification of QTL for a trait. Although a QTL may account for a large amount of the phenotypic variance of a trait, the movement of favorable alleles in this QTL region into a different genetic background may not always confer the desired trait to the progeny. For example, a single recessive gene conferring resistance to the Zucchini yellow mosaic virus Florida strain (ZYMV-FL) was identified in a watermelon (Citrullus lanatus) F₂ mapping population of PI 595203 (resistant) x ‘New Hampshire Midget’ (NHM, susceptible) (Ling et al., 2009). When the QTL region was moved using marker-assisted selection from PI 595203 into the ‘Charleston Gray’ background, the resulting BC₂F₂ plants did not exhibit the same level of resistance found in PI 595203 as they exhibited virus symptoms and virus replication was detected on even the most resistant plants (Harris et al., 2009). The authors speculated that a modifier gene controlling ZYMV-FL replication may exist in ‘Charleston Gray’.

Marker-assisted selection has been used to move disease resistance QTL from a donor parent to a recurrent parent resulting in resistant phenotypes. Resistance to rice (Oryza sativa) blast, caused by Magnaporthe oryzae, and sheath blight, caused by Rhizoctonia solani, was moved from the donor line Tetep to the susceptible rice hybrid Pusa 6B (Singh et al., 2015). Blast resistance is controlled by the blast resistance gene Pi54 and sheath blight resistance is controlled by three QTL. BC₁F₂ plants that were homozygous for Pi54 were selfed to generate BC₁F₃ families that were then subjected to a stepwise reductive screening utilizing markers for the three sheath blight resistance QTL. The BC₁F₅ plants were phenotyped and the lines containing the blast and sheath blight resistance QTL in the Pusa 6B background were resistant to both rice blast and sheath blight.

Following the original cross between the susceptible sorghum line Collier and the resistant line Honey Drip, F₁ plants were found to be resistant to M. incognita, and the F₂ generation segregated in approximately a 3:1 ratio of resistant to susceptible (Harris-Shultz et al., 2015a). Thus, the resistance was inherited as a single dominant locus. In the study reported herein, incorporating QTL-Sb.RKN.3.1 into a diverse set of genotypes resulted in a high level of resistance in all genotypes, which suggests that the resistance QTL is likely to confer resistance to M. incognita when incorporated into other genetic backgrounds. If any of the BC₁F₆ genotypes evaluated for this study were crossed to a susceptible line, it is expected that the resulting F₁ hybrid would be highly resistant to M. incognita.

The effectiveness of using crop rotation to minimize damage from Meloidogyne spp. to a crop of primary economic importance has been known for more than a century (Bessey, 1911). To effectively manage nematode populations in a cropping system, each crop should leave a lower nematode population density than the economic damage threshold of the following crop (Nusbaum and Ferris, 1973). Cotton is one of the primary field crops in the southern United States, and although several species of nematodes can cause severe damage to cotton, M. incognita causes the greatest total loss because of its widespread distribution (Davis and Stetina, 2016; Davis et al., 2018). Additional options for rotation crops that are profitable and do not require expensive, specialized equipment would be beneficial, and M. incognita-resistant crops such as sorghum would increase the options available for nematode management in cotton. Additionally, Meloidogyne-resistant crops can provide benefits beyond suppression of the target nematode species. For example, in contrast to the effect of some nematicides (Timper et al., 2012), host resistance to a plant-parasitic nematode should not affect population levels of beneficial nematodes or most other soil organisms, which may allow fields to maintain a level of natural suppression of nematodes or other pathogens.

In this study we moved the resistance QTL from ‘Honey Drip’ into five different sorghum backgrounds that included forage, sweet, and grain sorghum. The resistance to M. incognita in all of the backcross lines was equivalent to ‘Honey Drip’ and suggests this gene could be introgressed using marker-assisted selection into many sorghum genotypes and confer resistance. Thus, this QTL and its associated markers will be useful for sorghum breeding programs that want to incorporate M. incognita resistance into their sorghum lines.