Because allotetraploid cotton has a large and complex genome, researchers had not previously elucidated any comprehensive sequence information to describe the transcriptome related to defense responses against Verticillium dahliae. Here, we investigated such responses through full-length cDNA library construction and EST sequencing. Although two tetraploid species, G. hirsutum and G. barbadense, exist, few examinations of the latter have been made because it is less commonly cultivated. However, because it is more resistant to this wilt pathogen (Zhang et al., d), we used G. barbadense 'Pima90-53’ in our tests. As an American type of Pima, this cultivar has proven to be a better germplasm resource than other available types (i.e., Egyptian or Mid-Asian types of G. barbadense) because of its desirable traits for immunity to Verticillium, fiber quality, and greater yield (Ma et al., , Wang et al., ). Its resistance and phenotypic characters have long been recognized in both field and greenhouse settings. Moreover, we have previously cloned several functional genes related to resistance and/or fiber quality from 'Pima90-53’, e.g., GbVe, GbWRKY1, and ADF1 (Chi et al., [53, 54]; Liu et al., ; Pan et al., ; Zhang et al., b; Zhang et al., a; Zhang et al., d). Therefore, we chose this specific Pima cotton for profiling the G. barbadense transcriptome. Our goal was to mine those genes for economically important traits so that we might potentially improve those traits in the Upland cotton G. hirsutum.
Infections with V. dahliae progress throughout the roots and into the rest of the plant, causing serious losses in both yield and quality. Because the root is the first barrier against such an attack, we selected this particular tissue for analysis. Fungal spores germinate and epidermal cells are often penetrated within the first 12 h (Fradin and Thomma, ). Complex perception, transduction, and exchange of signals usually occurs in the early stages of infection (Zhang and Klessig, ; Kunkel and Brooks, ; Jones and Dangl, ). Therefore, we sampled at 1, 2, 4, 6, 8, 12, 24, 36, 48, 72, 96, and 120 hpi to coincide with those crucial stages and to isolate early pathogen-responsive genes. We also utilized an experimental system that allowed for tight control of environmental conditions so that gene expression was not altered by any factors other than the pathogen. This enabled us to identify and monitor as many genes as possible from our library.
More than 46,192 high-quality ESTs were generated from our root cDNA library of Verticillium-infected G. barbadense seedlings. These ESTs were assembled into 23,126 unigenes. Annotation results showed that this library contained many previously reported key response genes, such as for PR protein, chitinase, members of the GST gene family, PAL, CYP, and NDR1 (McFadden et al., ; Li et al., ; Gao et al., ; Xu et al., ; Zhu et al., ; Ahmed et al., ). Our library also included several genes that function in the development of cotton fibers, e.g., SusA1, CesAs, UXS1, ADF1, tubulin, and aquaporin (Pan et al., ; Yuan et al., ; Jiang et al., ; Kim et al., ; Chi et al., ). Our findings suggested that defense-related genes were abundant in our library, and that genes contributing to fiber formation might also function in protecting G. barbadense against infection by V. dahliae. Therefore, this library is an important genomics resource for isolating genes with novel roles. We also demonstrated the importance of identify critical genes that code for different phenotypes, such as stress resistance or fiber quality. That is, researchers can now construct a subtracted library to find genes that are preferentially expressed in resistant lines. We previously completed an SSH library that used a resistant G. hirsutum cultivar and contained more than 200 genes that are differentially expressed between it and susceptible cultivars (Zhang et al. ). Future studies will incorporate microarrays to achieve this goal in related experiments.
When exposed to various environmental stimuli, plants utilize elaborate mechanisms to regulate cellular and molecular events so they can protect themselves with pre-formed defense barriers and induce appropriate responses. For example, pathogen-triggered immunity (PTI) constitutes the first layer of that response, restricting pathogen activity by blocking further colonization (Zhang et al., ). A second layer, ETI, specifically recognizes the effector by one of its NB-LRR proteins (Nürnberger et al., ). ETI is an accelerated response while PTI is amplified, resulting in disease resistance and, usually, a hypersensitive cell death response at the infection site. In our full-length cDNA library, groups such as LRR-RLKs, signalling-related genes, and TFs were expressed during the cotton defense response. All of them possibly contribute to these PTI- or ETI-related systems.
Most PTI-associated genes were obtained in our library, including the chitin elicitor receptor kinase (CERK1), which recognizes chitin oligosaccharides during plant–pathogen interactions. It acts as a representative general elicitor to induce defense responses in a wide range of plant cells (Kaku et al., ). We also identified chitinases, which, as PR proteins, play important roles in enhancing stress resistance in a variety of plants. These chitin-degrading enzymes hydrolyze b-(1, 4) linkages and are capable of degrading the cell walls of plant pathogenic fungi (Adams, ; Cheng et al., ; Lawrence and Novak, ) while releasing elicitors of defense reactions. Therefore, chitinases are thought to have crucial roles in plant defenses against biotic stresses. Based on this, we might conclude that chitinases and their elicitors confer pathogen resistance in cotton by hydrolyzing the cell walls of V. dahliae.
Another important PTI-related gene, Brassinosteroid Insensitive 1-Associated Kinase 1, or BAK1, was isolated here. Known as Somatic Embryo Genesis Receptor-Like Kinase 3 (SERK3), it is required by PRRs (Heese et al., ; Dodds and Rathjen, ). SERK3/BAK1 not only interacts with BRI1, which is involved in brassinosteroid signal transduction, but it also rapidly forms a complex with FLS2, which functions in reactions for plant disease resistance. For example, the BAK1/SERK4 homolog in Nicotiana has a direct role in elicitor perception of bacterial cold shock protein, flagellin, and elicitin, but not chitin (Heese et al., ). However, we must still investigate whether similar elicitors exist in V. dahliae and how the FLS2–BAK1 complex acts to confer pathogen immunity in resistant cotton plants.
In addition to the genes already mentioned, we obtained another important clue regarding wilt resistance in cotton. The major pollen allergen (Bet v1) family protein was one of the most abundant in our library. Proteins in this family have previously been identified under various biotic and abiotic stresses in birch, potato, pea, soybean, and cotton (Breiteneder et al., ; Matton and Brisson, ; Barratt and Clark, ; Crowell et al., ; Cheng et al., ). Family members may also have essential roles in the defense response by Sea Island cotton to Verticillium wilt (Chen et al., ). The mRNA transcripts examined in our study demonstrated that Bet v1 family genes were more abundant than any others, suggesting that this family is a vital component of the cotton response to Verticillium infection.
With critical roles in complex signaling cascades, phytohormones have been integrated into current models for defense responses (Bari and Jones, ; Grant and Jones, ). For example, the pathways for salicylic acid (SA), jasmonic acid (JA), ethylene (ET), and brassinosteroids are important regulators of expression by defense-related genes (Bari and Jones, ). In general, SA induces systemically acquired resistance and is implicated in plant tolerance to biotrophic pathogens (Spoel and Dong, ; Leon-Reyes et al., ). Both ET and JA are typically associated with defense responses to necrotrophic pathogens (Spoel et al., ; Bari and Jones, ). Cross-talk can occur between those SA- and JA/ET-mediated defense reactions to abiotic- and biotic-stress stimuli (Grant and Jones, ). Results from our Q-PCR analysis showed that expression patterns were the same for EDS1 and PAD4. Both genes were up-regulated upon inoculation, and their transcript levels were several-fold higher than that of the mock. This implied that the EDS1–PAD4 complex interferes with activation of innate cotton immunity. In addition, while identifying the biochemical pathways that were active during the response to V. dahliae inoculation, we discovered a plant hormone signal transduction pathway, within which exists an important branch: "SA→NPR1→TGA→PR-1→Disease resistance". The presence of this branch might suggest its role in the resistance response. However, in contrast to earlier conclusions (Leon-Reyes et al., ), we determined that SA is not involved in resistance to necrotrophic pathogens, such as V. dahliae. Therefore, we deduced that the process by which cotton becomes infected by Verticillium involves both biotrophic and necrotrophic stages.
The plant cell wall serves not only as a physical barrier, but as a defense barrier against pathogen penetration. Secondary metabolites play a fundamental role in the plant’s ability to fight against invading pathogens (Dubery and Smit, ; Naoumkina et al., ). That capacity is derived through multiple pathways, including those for the biosynthesis of phenylpropanoids, terpenoids, and cellulose. Most of the genes associated with those pathways, e.g., genes for caffeic acid, 3-O-methyltransferase, glutathione S–transferase, 4-coumarate:CoA ligase, UDP-glucuronic acid decarboxylase, cellulose synthase, and sucrose synthase. In this study, we detected a large phenylpropanoid pathway that encompassed those for flavonoid and lignin biosynthesis. We also obtained most of the genes that encode enzymes involved in the lignin pathway. For example, PAL, located in the core and entry of the phenylpropanoid pathway, is responsive to both biotic and abiotic stresses, including pathogen attack, and wounding (Huang et al., ). Likewise, POD has a role in reinforcing cell walls against the effects of pathogens or wounding through the polymerization of monolignols into lignin (Marjamaa et al., ). In Arabidopsis, Laccase4 and Laccase17 contribute to the constitutive lignification of stems, with the latter being involved in the deposition of G lignin units in fibers (Berthet et al., ). Therefore, all of these findings demonstrate that the phenylpropanoid pathway has an essential role in preventing the invasion or expansion of pathogens by reinforcing the cell wall. Nonetheless, future research should focus on the exact function of related genes within the phenylpropanoid pathway. Further characterization of those genes may provide valuable candidates for efforts toward the genetic improvement of cotton.