Plants | Free Full-Text | Transcriptome Analysis of Resistant Cotton Germplasm Responding to Reniform Nematodes

Reniform nematode (Rotylenchulus reniformis Linford and Oliveira) is an obligate parasite that has a wide host range, which hinders management of this pest in several important crop species [1]. The nematode is a sedentary, semi-endoparasite, and has become one of the most important nematode species present in soils of the cotton-producing regions of the United States [1,2]. In the southeast states, such as Mississippi, Alabama, and Louisiana, reniform nematode has replaced Meloidogyne incognita (root-knot nematode) as the major nematode species infecting cotton [1,3]. Female vermiform nematodes can infect cotton roots throughout the growing season [4] and during this process, the nematode will modify host cells at the feeding site to form a syncytium, which provides the sole nutrient source for nematode development and reproduction [5]. Feeding on the root system interferes with the uptake of water and nutrients by the plant, resulting in stunting, delayed maturity, and yield reductions [4,6]. Moreover, root damage from nematode feeding makes cotton plants more vulnerable to soilborne diseases such as Fusarium wilt, which can compound yield losses [7]. In 2022, it was estimated that yield losses caused by reniform nematodes were 36.67 thousand metric tons, which was 1.2% of the U.S. total cotton production, and represents a loss of USD 74 M [8]. Yield losses in specific regions of the country were, however, significantly higher than the nationwide average with losses as high as 50% reported [9,10].

Limited approaches are available to manage reniform nematodes in Upland cotton (Gossypium hirsutum L.) production. Nematicide soil fumigation, seed treatment, in-furrow and foliar applications have been used to mitigate yield losses [1,10]. However, only a few nematicides are currently available and their efficacy can vary based on growing conditions [11,12,13,14]. Nematicides are effective in providing early season protection, but the nematode population can quickly and dramatically increase during the growing season due to the nematode’s short life cycle and high reproduction rate [1]. Additionally, nematicides are expensive and not all nematicide applications are profitable [14,15]. Human health and environmental concerns also need to be considered with the use of nematicides. Nematode management using crop rotation with non-host crops, such as corn and peanuts, or with resistant host cultivars, such as those available for soybean, has been recommended to reduce nematode populations [11,16]. This approach will require long rotational cycles to significantly lower the nematode population and therefore crop rotation is not always feasible due to the economic and resource constraints associated with cotton production [1,11]. Growing resistant Upland cotton cultivars or combining the application of nematicides with the use of resistant cultivars has been reported as an effective management strategy [1,11,17,18,19]. McCulloch et al. (2021) reported that resistant cultivars significantly suppress the reniform nematode population resulting in a 26% seedcotton yield increase compared to susceptible controls [20], whereas Koebernick et al. (2021) reported 8–20% yield increase using a resistant cultivar with a nematicide application during planting [17].

The development of resistant Upland cotton cultivars for reniform nematode management will require the use of multiple sources of resistance to limit the ability of the nematode to overcome a single source of resistance. Upland cotton accounts for 97% of cotton production in the U.S. (https://www.ers.usda.gov) (accessed on 1 October 2023); however, extensive screening of the Upland cotton germplasm collection only identified a few sources with moderate resistance [21,22,23,24]. These sources of resistance showed an inconsistent response [21,25] and have not been employed in commercial cultivar development. More desirable sources of resistance have been identified from tetraploid species G. barbadense L. [21,22] and several diploid Gossypium species, including G. aridum (Rose & Standl.) Skovst, G. arboreum L., G. herbaceum L., and G. longicalyx Hutch. & Lee [21,26,27,28].

Introgression of reniform nematode resistance from Gossypium germplasm into Upland cotton has been the focus for multiple breeding programs. Screening of G. barbadense germplasm collections identified a few accessions showing high levels of resistance [21,22] and multiple Upland cotton germplasm lines with improved resistance to reniform nematode have been released using two G. barbadense resistance sources. The G. barbadense accession TX 110 was used to develop two Upland cotton lines, TAM RKRNR-9 (Reg. No. GP-941, PI 662039) and TAM RKRNR-12 (Reg. No. GP-942, PI 662040), which showed suppression of reniform nematode reproduction by 40–70% [29]. The most widely used source of resistance has been the G. barbadense accession GB 713 and breeding lines derived from this source of resistance have been used in commercial cultivar development. Three lines, namely M713Ren1 (Reg. No. GP-958, PI 665928), M713Ren2 (Reg. No. GP-959, PI 665929), and M713Ren5 (Reg. No. GP-960, PI 665930), were developed in Mississippi and showed approximately 90% suppression in reniform nematode egg production [30]. The lines BARBREN-713 (Reg. No. GP-987, PI 671965) and BARBREN-713-32 (Reg. no. GP-1134, PI 701076) [31,32] were developed in Texas and suppressed nematode egg production by 74–92% [32]. Three QTLs controlling reniform nematode resistance were mapped for GB 713 with QTLs Ren^barb1 and Ren^barb2 located on chromosome AD_ch21_Dt.11 and QTL Ren^barb3 on chromosome AD_ch18_Dt.13 [33]. A recent study has shown that two QTLs controlled resistance with QTL Ren^barb2 mapped to a 17.7 MB interval (36.5–54.2 MB) on AD_ch21_Dt.11 and QTL Ren^barb3 mapped to a 1.8 MB interval (5.9–7.7 MB) on AD_ch18_Dt.13 in the genome assembly of BARBREN-713-32 [34]. Resistance was associated with QTL Ren^barb2, whereas QTL Ren^barb3 did not provide resistance, but was required to recover the high level of resistance characteristic of GB 713 [35,36].

The diploid cotton species are also a potential source of novel resistance genes; however, transferring resistance to Upland cotton is technically demanding and laborious. G. longicalyx was the only cotton species reported to show an immune response to reniform nematode infection [21]. This source of resistance was successfully transferred to Upland cotton using the bridging lines (G. hirsutum × G. longicalyx) and (G. hirsutum × G. herbaceum) with the breeding lines LONREN-1 (Reg. No. GP-977, PI 669509) and LONREN-2 (Reg. No. GP978, PI 669510) being developed [37,38,39]. The G. longicalyx resistance gene Ren^lon was mapped on chromosome AD_ch21_Dt.11 [40]. Resistance derived from G. longicalyx was associated with a hypersensitive response resulting in root necrosis and severe plant stunting, therefore this resistance source has not been used for commercial cultivar development [41,42]. Bhandari et al. (2015) [43] reported hypersensitive reactions in G. hirsutum germplasm lines LONREN-1 and LONREN-2 with G. longicalyx resistance, and G. arboreum accession A2-190, to two isolates from Louisiana based on plant height reductions; however, reductions in root growth for A2-190 was not reported. Severe root necrosis associated with resistance has not been reported for other diploid cotton species showing high levels of resistance, although, limited research has been conducted on these resistant sources. Resistance from G. aridum and G. arboreum were transferred to Upland cotton using the (G. hirsutum × G. aridum) bridging line [27,44]. The resistance QTL Ren^ari derived from G. aridum was mapped to AD_ch21_Dt.11 in Upland cotton [27] and at least two resistance genes were reported for the G. arboreum source of resistance [44]. Limited breeding research has been conducted on these resistance sources and no Upland cotton breeding lines have been released.

The G. arboreum germplasm collection includes more than 1600 accessions and resistant genotypes have been frequently observed in this collection [28]. A few G. arboreum accessions have been genetically characterized for nematode resistance with resistance being conferred by one or a few genes [45,46,47]. A genome-wide association study of 246 accessions revealed 15 SNPs significantly associated with reniform nematode resistance [48]. The lack of genomic tools has hindered the transfer of resistance from G. arboreum to Upland cotton. To enhance the utilization of resistance sources, in the present study, the transcriptomes of G. arboreum and G. barbadense resistant genotypes responding to reniform nematode were evaluated to identify differentially expressed genes and genomic regions potentially associated with nematode resistance.

Gene expression studies are one approach to identify genes associated with resistance. Transcriptome analysis of G. hirsutum cultivars that are susceptible (DP90 and SG747), resistant (BARBREN-713) and hypersensitive (LONREN-1) to reniform nematodes revealed many DEGs associated with resistance, such as cell wall architecture, hormone metabolism and signaling, ROS levels, cell death pathways, and pathogenesis [49]. In the present study, DEGs with a range of functions were identified from the four cotton genotypes in response to reniform nematode infection. Resistant gene analogs and non-specific resistance genes, such as chitinase, glucan endo-1,3-beta-glucosidase, and pathogenesis-related proteins, were differentially expressed and may have an important role in response to reniform nematode infection. The DEGs were more often upregulated than downregulated, higher expression of certain genes may help to fight the parasites and mitigate the damage of nematodes. Li et al. (2015) [49] reported that more DEGs were downregulated for GB 713, whereas the other two evaluated genotypes show slightly more DEGs being upregulated.

Meloidogyne induced cotton genes MIC-3 and MIC-4 were differentially expressed in A2-100 and A2-190, and MIC-3 gene was differentially expressed for GB 713, and they were upregulated in these genotypes. MIC-3 was found to be specifically expressed in cotton resistant genotypes when inoculated with M. incognita [50]. Overexpression of MIC-3 in G. hirsutum cv. Coker 312 reduced egg production of M. incognita by 60–75%, but no effect on reniform nematode reproduction was reported [51]. These genes showed expression in A2-100 and A2-190 on D5, which may indicate that expression of these genes occurs early in the infection process and may influence the establishment of the feeding site. In contrast, MIC-3 expression was observed in GB 713 on D9, suggesting a possible role in hindering nematode development. These DEGs were associated with chromosome A05 in both species. Several cotton resistance genes/QTLs to biotic and abiotic stresses have been mapped on chromosome AD-ch05_At.05 [52,53,54], which may result from natural selections in cotton evolution. The physical locations on cotton genomes for these genes/QTLs were unknown, thus their distance to the resistance genes identified in this study are undetermined.

Genes for receptor-like proteins showed expression for A2-100 and A2-190 on D5 and in GB 713 on D9. These DEGs were upregulated for these genotypes. Receptor-like proteins are cell surface receptors composed of several distinct domains, including signal peptide, extracellular leucine-rich repeat (LRR) region, and transmembrane domain, with these proteins playing an important role in plant development as well as disease resistance [55]. DEGs of receptor-like proteins were associated with chromosome A01 across the cotton genotypes in the present study, which may suggest a conservation response. Additionally, Li et al. (2018) [48] reported one SNP significantly associated with reniform nematode resistance for chromosome A01 in G. arboreum, which is about 9 Mb from the receptor-like gene XM_053030958.1 (Table 2).

Hevamine-A-like protein is a type of plant chitinase and lysozyme that are important for plant defense against pathogens [56]. Chitin is an essential component of the nematode eggshell and pharynx, and the disturbance of chitin synthesis or hydrolysis could lead to nematode embryonic lethality, defective egg laying, or molting failure [57]. Upregulated differential expression for hevamine-A-like protein was recorded for the D5 treatment across the four genotypes and for the D9 treatment in A2-190 and TX 110. This DEG was commonly associated with chromosome A12 across the four genotypes. A genome-wide association study of G. arboreum genotypes also identified candidate genes for reniform nematode resistance on chromosome A12 [48].

Several other genes associated with plant defense were also recorded for the genotypes evaluated in the current study. Multiple peroxidase genes were differentially expressed for A2-100, A2-190, and GB 713, which were upregulated. Peroxidase genes are critical for cell wall stiffening [58]. Li et al. (2015) [49] reported peroxidase genes were differentially expressed in response to reniform nematode infection across the four genotypes that showed different infection responses. Aspartyl protease family protein At5g10770-like showed differential expression and was upregulated for A2-100, A2-190, and TX 110. This protein could degrade pathogen effectors contributing to reducing virulence and degrade pathogenesis-related proteins that would induce the expression of genes involved in stress and defense responses, innate immunity, and systemic acquired resistance (SAR) [59]. A metacaspase-9-like gene was recorded for A2-100 and TX 110, and two ervatamin-B-like genes and one self-pruning gene were observed for A2-100. These genes have critical roles in programmed cell death during plant development and defense responses [60,61,62]. Two genes differentially expressed for A2-100 and one gene for TX 110 encode early-nodulin like proteins. These proteins were reported to have an important role for enhancing fitness of the pathogen during host colonization [63]. In addition, these proteins have functions associated with the transportation of nutrients and plant development [63]. One DEG (XM_017774501.2) recorded for A2-100 encodes the NIM1-interacting protein. This gene has an important role in systemic acquired resistance [64]. Two Kunitz trypsin inhibitor 5 genes were differentially expressed for TX 110. These genes were reported to have functions associated with the protection of plants from insect predators [65]. One nematocidal crystal cry1Ag gene observed for TX 110 may have resulted from Bacillus thuringiensis contamination [66].

Mapping reniform nematode resistance genes will be critical for the development of molecular markers to enhance the introgression of these genes from related Gossypium species for the development of resistant Upland cotton cultivars. Fifteen SNPs significantly associated with reniform nematode resistance were mapped to eight chromosomes for G. arboreum [48]. DEGs reported in the current study were widely distributed across the genome with DEGs observed on 13 chromosomes for A2-100 and 9 chromosomes for A2-190; however, the majority of the DEGs were associated with chromosome A05 for the 2 genotypes. Thus, DNA markers developed from chromosome A05 should be selected for screening segregating populations for nematode resistance derived from G. arboreum sources. DEGs associated with chromosome A12 may also have a role in nematode resistance for the G. arboreum genotypes. Some similarities were observed for the DEGs across the genotypes and these data may suggest some shared resistance genes; however, identifying G. arboreum genotypes with unique resistance genes will be necessary to increase genetic diversity.

DEGs were also observed on chromosome A05 of the A-genome and the homoeologous chromosome 19 of the D-genome for TX 110 and GB 713 with more DEGs frequently associated with these chromosomes for the D9 treatment. DEGs were distributed across most of the chromosomes for these two genotypes, but the frequency of DEGs associated with individual chromosomes varied across genotypes and across inoculation treatments. The occurrence of DEGs across the genome may suggest expression of genes related to root tissue damage and not associated with nematode resistance. The resistance QTLs transferred from GB 713 were mapped to AD_ch21_Dt.11 and AD_ch18_Dt.13 in the genome assembly of BARBREN-713-32 [34]. Transcriptome analysis of GB 713 and TX 110 in the current research identified DEGs associated with these two chromosomes and the DEGs were in or adjacent to the intervals identified for BARBREN-713-32. Two DEGs for TX 110 and one for GB 713 that were mapped onto these chromosomes had functions associated with plant defense. A few DEGs were comparable across the two genotypes and may indicated some similarity in the mechanism of resistance.

Sampling time is very critical for the transcriptome analysis of cotton responding to reniform nematodes. While inspecting nematode infection on a small subset of plants two DAI, it was found that not all plants were infected, and the G. arboreum accessions had little root development. All sampled plants were infected on D5, and plant root mass was enough for RNA extraction. The parasitizing nematodes were in the gravid stage on D9. Therefore, root samples were collected on D5 and D9 in this study. A comparison across the two inoculation treatments showed different gene expression patterns among the genotypes evaluated in the present study. The number of DEGs recorded for the D9 treatment was dramatically reduced compared to the D5 treatment for the G. arboreum genotypes. These data could suggest resistance is associated with the early stages of nematode infection. However, fewer DEGs were observed for A2-190 for the two inoculation treatments compared to A2-100. This may suggest that resistance associated with A2-190 could hinder the establishment of the feeding site for the nematode, resulting in the low number of nematodes typically observed on the root system for this genotype. Collecting root samples prior to D5 may aid in the identification of additional DEGs associated with resistance. However, transcript evaluations prior to D5 may require a different sampling approach because of the few nematodes infecting the root system for highly resistant genotypes and low root weights characteristic of G. arboreum accessions. In contrast, resistance associated with A2-100 may occur later in the infection process, leading to a greater number of nematodes infecting the root system [67]; thus, contributing to the moderately resistant response recorded for this genotype and the higher number of DEGs observed. The difference in the number of DEGs across inoculation treatments was reduced for the G. barbadense genotype GB 713; although, fewer DEGs were recorded for the D9 treatment. These data could suggest a different mode of resistance compared to the G. arboreum genotypes. This trend was reversed for TX 110 with more DEGs recorded for the D9 treatment than the D5 treatment. A moderately resistant response to reniform nematode infection has been reported for TX 110 [22]. When infected with the MSRR04 reniform nematode population, TX 110 and GB 713 showed similar nematode development for two days after inoculation; then, more rapid progression of nematode development was observed for TX 110 compared to GB 713 [67]. The expression of many non-specific pathogenesis-related genes was also observed for TX 110 for the D9 treatment in the current study, which could be associated with the moderately resistant response. The availability of data on nematode development for the two genotypes could be used in future studies to select additional time points to evaluate changes in gene expression to assess the pathways contributing to the resistant response.

Reference genomes are critical for transcriptome analysis. ShiXiYa 1 (PI 615743) was reported to have a moderately resistance response to reniform nematode infection [28]. Using this reference genome, many receptor-like DEGs were identified from the two G. arboreum genotypes evaluated in the current study. While using the Pima 90 genome as a reference, relatively less receptor-like DEGs were identified from the two G. barbadense genotypes; this could result from the lack of resistance in Pima 90 to reniform nematode (its response to reniform nematode needs to be tested). Three reference genomes were used for gene identification and feature count when analyzing the transcriptomes for GB 713. The number of non-specific plant defense related DEGs recorded was reduced using the G. hirsutum genotypes BARBREN 713 and BARBREN 713-32 resistant to reniform nematodes. Since their resistance was derived from GB 713, it is possible that these DEGs of GB 713 confer resistance to reniform nematode. Three DEGs were found in each of the G. hirsutum reference genomes, suggesting some resistance genes are common in these lines. The discrepancy in the numbers of DEGs identified may result from the selections in breeding or the difference of genome assemblies and annotations.

DEGs associated with disease resistance genes identified from the current research will be further tested using other resistant and susceptible cotton genotypes to determine their role in suppressing reniform nematode infection and development. More than 100 resistant genotypes were identified from the G. arboreum germplasm collection [28] and would be useful for these evaluations. Accessions from this collection showed a wide range of variations for infection response, which could be useful for identifying additional DEGs associated with nematode resistance. Some accessions showed very low numbers of nematodes infecting the root system and resistance could be associated with the establishment of the feeding site, whereas hindering nematode egg production could be associated with host resistance for other accessions. Increasing root sampling times would also be useful to identify additional DEGs associated with this variation in infection response although constitutively expressed resistance genes cannot be identified using transcriptome analysis. Molecular markers for resistance genes can be developed using segregating populations, which would be beneficial for the introgression of resistance in Upland cotton breeding programs.