The lysin motif-containing proteins, Lyp1, Lyk7 and LysMe3, play important roles in chitin perception and defense against Verticillium dahliae in cotton

Background Lysin motif (LysM)-containing proteins are important pattern recognition receptors (PRRs) in plants, which function in the perception of microbe-associated molecular patterns (MAMPs) and in the defense against pathogenic attack. To date, the LysM genes have not been systematically analyzed in cotton or effectively utilized for disease resistance. Results Here, we identified 29, 30, 60, and 56 LysM genes in the four sequenced cotton species, diploid Gossypium raimondii, diploid G. arboreum, tetraploid G. hirsutum acc. TM-1, and G. barbadense acc. 3–79, respectively. These LysM genes were classified into four groups with different structural characteristics and a variety of expression patterns in different organs and tissues when induced by chitin or Verticillium dahliae. We further characterized three genes, Lyp1, Lyk7 and LysMe3, which showed significant increase in expression in response to chitin signals, V. dahliae challenge and several stress-related signaling compounds. Lyp1, Lyk7 and LysMe3 proteins were localized to the plasma membrane, and silencing of their expression in cotton drastically impaired salicylic acid, jasmonic acid, and reactive oxygen species generation, impaired defense gene activation, and compromised resistance to V. dahliae. Conclusion Our results indicate that Lyp1, Lyk7, and LysMe3 are important PRRs that function in the recognition of chitin signals to activate the downstream defense processes and induce cotton defense mechanisms against V. dahliae. Electronic supplementary material The online version of this article (10.1186/s12870-017-1096-1) contains supplementary material, which is available to authorized users.


Background
Various immune responses are triggered in plants against invading pathogens upon the perception of socalled pathogen associated molecular patterns (PAMPs) at the plant cell surface. Most of the known PAMPs have been well-characterized as either polypeptide type or carbohydrate type, such as two ligands, Flg22 (a peptide representing the elicitor-active epitope of the bacterial flagellin) and Elf18 (an EF-Tu-derived peptide). And their interrelated receptor proteins FLS2 and EFR have been identified as the leucine-rich repeat receptor-like kinases (LRR-RLKs) in Arabidopsis thaliana [1,2]. Chitin, a major component of the fungal cell wall, is a well-known PAMP that can be recognized by pattern recognition receptors (PRRs) at the plant cell surface and activate PAMP-triggered immunity (PTI) [3]. PTI includes the activation of pathogenesis-related (PR) genes and the production of reactive oxygen species (ROS) [4]. In addition, chitin binding sites and plasma membrane receptor proteins have been detected in membrane preparations of different plants, and have been shown to activate downstream defense processes [5][6][7][8].
The Lysin motif (LysM) domain, a protein module that recognizes chitooligosaccharides, peptidoglycan and other related N-acetylglucosamine (GlcNAc)-containing oligosaccharides, which usually contains about 40 amino acids (AA) and is a ubiquitous modular cassette that exists in all living organisms except for Archaea [9,10]. In plants, a chitin elicitor binding protein, CEBiP, was first reported to be a LysM domain-containing protein and involved in the binding and perception of chitooligosaccharides in rice [3]. Subsequently, LysM-encoding genes were identified in a broad range of organisms at a genome-wide level using the increasing numbers of transcript and genomic sequences. According to the subcellular location and domain structure of LysM genes, members of the family can be divided into four subgroups; LysMcontaining receptor-like kinases (Lyks), LysM-type receptor-like proteins (Lyps), extracellular LysM proteins (LysMes) and nonsecretory intracellular LysM proteins (LysMns) [10,11]. The multiple domains and the complex structures of the LysM genes are indicative of the variety of functions they carry out.
As PAMP receptors, LysM domain-containing proteins can sense bacterial oligosaccharides, peptidoglycan (PGN) and fungal chitin, and respond by promoting the plant's defenses. In Arabidopsis, Lyks are defined as receptors for chitin, and hence, are implicated in plant defense mechanisms against fungal pathogens [12]. AtCERK1, which is a cell surface chitin elicitor receptor kinase 1, directly binds chitin through its LysMcontaining ectodomain (AtCERK1-ECD) to trigger immune responses [13,14]. At LYK4 also mainly assists chitin signal transduction and activates plant innate immunity through chitin recognition [15]. In addition, AtLYK5 not only shares functions with AtLYK4 in mediating the chitin response, but also binds to chitin with a higher affinity than AtCERK1. AtLYK5 is also necessary for chitin-induced AtCERK1 phosphorylation and homodimerization. AtCERK1, AtLYK4 and AtLYK5 are membrane-localized proteins, and are all involved in plant responses to chitin [16]. Two rice LysM receptor molecules, CEBiP and CERK1, have been identified as critical in the regulation of chitin elicitor signaling [17]. Moreover, two LysM receptor proteins, LYP4 and LYP6, have also been identified as peptidoglycan and chitin perception receptors in rice [18]. To date, however, the functions of the LysM genes in cotton remain largely unknown.
Verticillium dahliae, a destructive soil-borne fungal pathogen, causes cotton Verticillium wilt and leads to severe reductions in cotton yield across the world [19]. Although huge efforts have been made to generate wiltresistant cotton cultivars through traditional breeding, this remains a challenge [20]. Recently, progress has been made in excavating PR genes and exploring the molecular mechanisms of responses to V. dahliae invasion in cotton. Several V. dahliae-responsive genes, such as GhNDR1, GhNaD1, GbWRKY1, GhSSN, and GhMLP28, have been shown to be functionally related to defense responses to V. dahliae attack in cotton [21][22][23][24][25]. Nevertheless, there is an urgent requirement to mine more candidate genes in order to develop Verticillium wilt-resistant cotton cultivars.
The availability of data on the whole-genome of different Gossypium cotton species, including G. raimondii (D5) [26], G. arboreum (A2) [27], G. hirsutum acc. TM-1 (AD1) [28], and G. barbadense acc. 3-79 (AD2) [29], has made it possible to systematically identify and analyze the targeted genes on a genome-wide level, and has thus enriched our understanding on the molecular mechanisms of cotton responses to V. dahliae invasion. In the present study, we systematically surveyed LysM members in four sequenced cotton species for the first time, and analyzed their phylogenetic relationships, gene structures, subcellular localizations and expression patterns in different tissues in control conditions, and in response to PAMPs, V. dahliae and different stressrelated signaling compounds. Furthermore, we found that silencing of Lyp1, Lyk7, and LysMe3 expression in cotton significantly impaired tolerance of V. dahliae, indicating that these genes played important roles in the defense response mediated by chitin recognition. Moreover, down-regulation of Lyp1, Lyk7, and LysMe3 expression hampered downstream resistance-related pathways and the activation of PR genes. The study not only enriches our knowledge of the networks involving in LysMs in cotton, but also provides effective gene resources for the development of Verticillium wilt-resistant cultivars through cotton-breeding programs.
Based on the conserved domains, motif components and exon-intron organization of the LysM genes, we classified them into four categories; Lyks, Lyps, LysMes and LysMns (Additional file 2: Figure S1). The LysM genes were relatively conserved within each group, but obvious differences existed between groups. To elucidate the chromosomal distribution of these LysM genes, we integrated 13 scaffolds of the G. raimondii genome (named Chr01 to Chr13) [26] into the previously reported high-density interspecific genetic map of allotetraploid cultivated cotton species [32], and then reordered the 13 G. raimondii scaffolds according to the corresponding D1 to D13 chromosomes of the tetraploid cotton species (Fig. 1). Following the new order of the G. raimondii chromosomes, the LysM genes in G. raimondii were named GrLyp1 to GrLyp4, GrLyk1 to GrLyk8, GrLysMe1 to GrLysMe9, and GrLysMn1 to GrLysMn8 (Fig. 1). The 29 LysM genes in G. raimondii were matched to 12 scaffolds, and no genes were matched to the D7 chromosome scaffold. The chromosomal distribution of the LysM members was uneven, with chromosomes D1, D6, D9, D10, D11, D12 and D13 containing more than three genes each, and the other chromosomes containing smaller numbers of LysM genes. Furthermore, the corresponding orthologs in three other sequenced cotton species, G. arboreum, G. hirsutum acc. TM-1, and G. barbadense acc. 3-79, were named respective to the orthologs in G. raimondii (Additional file 1: Table S1).
Two publicly available algorithms, SignalP and Tar-getP, were used to predict the subcellular locations of these LysM proteins. SignalP predicted that 15 of the 29 LysM proteins contained signal peptides, meanwhile TargetP analyses predicted that 18 of these proteins were involved in secretory pathways. In addition, SMART programs confirmed that 12 of the LysM proteins contained transmembrane domains (Additional file 3: Table S2). These findings indicate that LysM proteins are involved in MAMP recognition and therefore perform crucial roles in plant defense responses.

Expression analysis of LysM genes in response to two PAMPs and V. dahliae
To further investigate the roles of LysM genes in tetraploid cotton, we used transcriptome data from G. hirsutum acc. TM-1 vegetative tissues (root, stem, and leaf ), floral tissues (petal and anther), and ovule and fiber tissues at −3, 0, 3, 5, 10, 20, and 25 days post anthesis (DPA) to determine the expression patterns of these genes [28]. A total of 50 LysM genes were expressed in G. hirsutum acc. TM-1 with FPKM > 1.0, and the developmental and spatial regulation of these genes differed between the various tissues (Fig. 2). As previously reported, many LysM genes played important roles in chitin recognition, chitin signal transduction and the activation of plant innate immunity [13-15, 18, 33]. To better understand the function of LysM genes, we selected 20 genes with relatively high expression levels in the root tissue and analyzed their expression when induced by V. dahliae and by two PAMPs, insoluble crab shell chitin and soluble chitin fragment Nacetylchitohexaose. In detail, the transcripts of 16 LysM genes, 6 Lyks, 4 Lyps, 2 LysMes, and 4 LysMns, were significantly upregulated following treatment with two PAMPs (Fig. 3). These 16 LysM genes exhibited maximal expression levels 1 h, 2 h or 4 h after treatment, and both insoluble crab shell chitin and soluble chitin fragment N-acetylchitohexaose treatment caused a 6to 13-fold increase in their expression. Significantly, Lyk7 of the Lyk category had a higher expression level than other members; reaching maximal levels of a 13fold increase 4 h after N-acetylchitohexaose treatment, and a 10-fold increase 2 h after chitin treatment. The expression of Lyp1 in the Lyps category was significantly upregulated following treatment, with increases of 11-and 7-fold 2 h after N-acetylchitohexaose treatment and 1 h after chitin treatment, respectively. In the Fig. 1 Chromosomal distribution of LysM genes in G. raimondii. The chromosome numbers are shown at the top of each bar. The 29 LysM genes in G. raimondii were classified into four groups and marked on the linkage map. The names of the scaffolds from the genome are also indicated in brackets. The chromosome numbers from D1 to D6, and D8 to D13, were consistent with the newly-updated interspecific genetic map of allotetraploid cultivated cotton species [32]. The nomenclature of the LysM genes for each group was based on the order of the chromosomes in G. raimondii LysMes category, an 11-fold increase in LysMe3 transcript levels was induced 4 h after N-acetylchitohexaose treatment, and an 8-fold increase was seen 2 h after chitin treatment. In the LysMns category, the expression of the LysMn2 was increased by 9-fold at 4 h after N-acetylchitohexaose treatment, but the peak values of LysMn3, LysMn5, and LysMn6, of 7.8-, 5.2-, and 5.4fold increases, respectively, were observed 2 h after treatment. In addition, the expression levels of LysMn2 and LysMn3 increased 4.2-and 9.6-fold 4 h after chitin treatment; however, the level of LysMn5 and LysMn6 transcript expression increased more quickly, with respective peaks of 3.9-and 6.6-fold increases reached 2 h after treatment. These results suggest a role for LysM genes in chitin recognition; an important part of cotton innate immunity. Roots, stems, leaves, petals, stamens, ovules at −3, 0, and 3 DPA, and fibers at 5, 10, 20, and 25 DPA were used for comparative transcriptome analysis. The expression data were converted to Log 2 (FPKM) to calculate the expression levels of the LysM genes in TM-1. Differences in gene expression are shown in the colors indicated in the scale. The RNA-Seq data used here can be downloaded from http://www.ncbi.nlm.nih.gov/bioproject/PRJNA248163/ Plant receptor kinases and receptor proteins containing lysin motifs play crucial roles in the recognition of carbohydrate patterns commonly associated with microbial surfaces and in the defense against microbial infection [3,12,[14][15][16][17]33]. To better ascertain the functions of LysM proteins in plant defenses against V. dahliae, we analyzed the expression patterns of 16 LysM genes after V. dahliae inoculation in Hai7124 and Junmian 1, which exhibited resistance and susceptibility to V. dahliae, respectively. Overall, the expression levels of these LysM genes were higher in Hai7124 than in Junmian 1. Further, 12 LysM genes, 4 Lyks, 3 Lyps, 1 LysMe, and 4 LysMns, were significantly induced by V. dahliae and quickly reached peak expression levels at different time points in Hai7124 (Fig. 4). Among them, Lyk7, Lyp1, LysMe3, and LysMn6 showed the highest expression levels of their respective groups after V. dahliae treatment, suggesting that these genes played important roles in chitin recognition and defense against V. dahliae.
Silencing Lyp1, Lyk7, and LysMe3 impairs cotton tolerance to V. dahliae To obtain further evidence of the role of LysM genes in V. dahliae resistance, we selected Lyp1, Lyk7, LysMe3, and LysMn6, which had the highest expression levels of their groups and were significantly induced after chitin, chitin fragment N-acetylchitohexaose, and V. dahliae treatments, for functional identification via virus-induced gene silencing (VIGS) analysis.
We constructed TRV: GbLyp1, TRV: GbLyk7, TRV: GbLysMe3 and TRV: GbLysMn6 vectors to silence endogenous genes in Hai7124, and used TRV: 00 as a mock treatment and TRV: GhCLA1 as a positive control to validate the efficiency of the VIGS assay. As expected, the cotton leaves showed an obvious photobleaching phenotype 2 weeks after agroinfiltration with the GhCLA1 construct (Additional file 4: Figure S2), indicating that the VIGS system worked well in our experimental operations. The cotton seedlings that were confirmed to be infiltrated with one of the constructs and those with the mock treatment were selected for RNA isolation and quantitative real-time PCR (qRT-PCR) analysis. Hai7124 and Junmian 1 plants were used as controls resistant to and susceptible to V. dahliae, respectively, and were challenged with V. dahliae strain V991 by the dipinfection method at a final concentration of 1 × 10 7 spores per milliliter (ml) [34]. About 10 days after inoculation, the seedlings of the Junmian 1 plants showed Fig. 3 Induced expression of LysM genes by two PAMPs. Two-week-old seedlings were treated with either 200 mg/mL insoluble crab shell chitin or soluble chitin fragment N-acetylchitohexaose and sampled 0.5, 1, 2, 4, 6 h (h) after treatment, respectively. The induction of each gene was examined by qRT-PCR, and the 0 h expression levels were used as controls when calculating the level of induction. The data represent the mean ± SD of three samples from three independent tests at each time point. "*": significant difference at P < 0.05; "**": significant difference at P < 0.01 obvious cotyledon wilting, but only a small number of Hai7124 plants displayed the leaf wilting phenotype and these only appeared at least 15 days after inoculation with V. dahliae. These results further confirmed that Hai7124 and Junmian 1 plants were resistant and susceptible to V. dahliae, respectively, and could be used as test cultivars to investigate the virulence of V. dahliae in cotton. The phenotypes of the two control plants at 20 and 25 days after inoculation with V. dahliae are shown in Additional file 5: Figure S3.
Compared with the controls, the expression levels of Lyp1 and Lyk7 were significantly lower in Lyk7-and Lyp1-silenced plants (Fig. 5a); however, the transcript levels of the other genes in the same group were not changed in the VIGS plants (Additional file 6: Figure S4). Further, the Lyp1-and Lyk7-silenced plants were inoculated with V. dahliae. No significant phenotypic differences were observed between the VIGS plants and Hai7124 control plants before pathogen invasion. However, 15 days after infection, the VIGS plants began to show obvious leaf-yellowing phenotypes, and 20 or 25 days after V. dahliae invasion, the Lyp1-, and Lyk7-silenced plants displayed more severe wilting and yellowing symptoms and more etiolated leaves than the control plants, suggesting that the Lyp1-, and Lyk7-silenced plants had a higher susceptibility to V. dahliae (Fig. 5b). After a further 10 days, almost all the true leaves were defoliated in Lyp1and Lyk7-silenced plants, meanwhile, a similar phenotype was observed in Junmian 1 at 25 days after V. dahliae invasion (Additional file 5: Figure S3). To better investigate the susceptibility of Lyp1-and Lyk7-silenced plants to V. dahliae, more than 20 plants per treatment were used to calculate the ratio of diseased to non-diseased leaves. The Hai7124 seedlings without injection and TRV: 00 control plants had similar characteristics, including few wilted leaves and a ratio of diseased leaves to healthy leaves of approximately 50% 35 days after inoculation. However, about 85% of the Lyp1-and Lyk7-silenced plants displayed leaf wilting or defoliation, and nearly 100% disease was observed in the susceptible control Junmian 1 plants 35 days after inoculation (Fig. 5c, Additional file 7: Table S3).
In parallel, we also performed qRT-PCR to analyze the expression of LysMes and LysMns in the LysMe3-and LysMn6-silenced plants, respectively. Only LysMe3 and LysMn6 were significantly down-regulated in the VIGS plants compared to the TRV: 00 control plants (Fig. 6a, Additional file 6: Figure S4). The LysMe3-and LysMn6silenced plants were inoculated with V. dahliae. Downregulation of LysMe3 resulted in impaired tolerance to V. dahliae, yet there were no significant changes in tolerance to V. dahliae in the LysMn6-silenced plants compared to controls (Fig. 6b). The percentage of diseased leaves in LysMe3-silenced plants was consistently higher than in controls and reached 80% 35 days after inoculation, but in LysMn6-silenced plants, only 56% of leaves displayed wilting or defoliation, which was similar to Verticillium dahlia were investigated in G. barbadense cv. Hai7124 and G. hirsutum cv. Junmian 1, which show resistance and susceptibility to V. dahliae, respectively. qRT-PCR analysis showed differences in the expression of LysM genes in Hai7124 and Junmian 1 after inoculation with V. dahliae strain V991. The statistical analysis compared expression levels at different time points following treatment with those at 0 h. Error bars show the standard deviation of three biological replicates. "*": significant difference at P < 0.05; "**": significant difference at P < 0.01 that in TRV: 00 and Hai 7124 plants (Fig. 6c, Additional file 8: Table S4).
Taken together, these results show that silencing Lyp1, Lyk7, and LysMe3 in cotton plants significantly increases their susceptibility to V. dahliae, suggesting that these three genes play important roles in the resistance to V. dahliae infection.

Lyp1, Lyk7 and LysMe3 are located in the plasma membrane
To gain direct evidence of the subcellular localization of Lyp1, Lyk7 and LysMe3, we transiently expressed these proteins fused to GFP in onion epidermal cells using biolistic bombardment. A pBIN-GFP4 vector was used as a control and fluorescence was detected under a confocal All experiments were repeated using at least 20 seedlings, and error bars show the standard deviation of three biological replicates. Asterisks indicate statistically significant differences in the percentage of diseased leaves between treated plants and TRV: 00 controls, as determined by Student's t-test (*P < 0.05, **P < 0.01) microscope. As shown in Fig. 7, fluorescence from the Lyp1, Lyk7 and LysMe3 GFP fusion proteins was detected in the cell membrane, and the shape of the membrane was changed during the plasmolytic processes. However, GFP control fluorescence (35S-GFP) was detected throughout the cell. Taken together, these findings suggest that Lyp1, Lyk7, and LysMe3 are located in the cell membrane, and these typically membrane-anchored proteins might be involved in recognizing PAMPs and activating a variety immune responses.
Lyp1, Lyk7, and LysMe3 are crucial for downstream defense reactions To better investigate the roles of Lyp1, Lyk7, and LysMe3 in the cotton-V. dahliae interaction, we further analyzed their expression patterns in cotton treated with the defense-related signaling molecules, salicylic acid (SA), jasmonic acid (JA), ethylene (ET) and hydrogen peroxide (H 2 O 2 ). As shown in Fig. 8a, Lyp1, Lyk7, and LysMe3 were significantly upregulated following SA, JA, or H 2 O 2 treatment at different time points. In detail, Lyp1 and Lyk7 were clearly induced by SA, JA, and H 2 O 2 treatment, whereas LysMe3 was only upregulated by treatment with SA. These findings hint at a potential involvement of Lyp1, Lyk7, and LysMe3 in the SA, JA, and H 2 O 2 signaling pathways.
As previously reported [40][41][42][43], the PR genes play important roles in signal recognition and plant immunity. To identify the PR proteins involved in cotton defense mechanisms against V. dahliae, we analyzed the expression patterns of PR1 [40], PR4 [41], PR5 [42], and PR10 [43] in Hai7124 and Junmian 1 after treatment with V. dahliae strain V991. As shown in Fig. 9a, these four genes were significantly induced in the two cotton cultivars after V. dahliae infection, however, the expression levels were higher in Hai7124 than in Junmian 1, suggesting that these genes play important roles in  protecting cotton plants against V. dahliae infection. In addition, the levels of transcripts of these four defenserelated genes were upregulated in cotton following exogenous application of SA, JA, and ROS (Fig. 9b).
To determine whether Lyp1, Lyk7, and LysMe3 silencing affected downstream defense-related genes, we further analyzed the expression of PR1, PR4, PR5, and PR10 in the VIGS plants. As shown in Fig. 9c, compared with that in TRV: 00 control plants, the expression of the four genes was significantly lower in the Lyp1-, Lyk7-, and LysMe3-silenced cotton plants. We suppose that Lyp1, Lyk7, and LysMe3 act as plasma membrane receptors in cotton plants to recognize chitin signals and activate a common downstream pathway, and subsequently induce the expression of related defense genes, therefore enhancing resistance to V. dahliae (Fig. 10).
(See figure on previous page.) Fig. 8 Lyp1, Lyk7, and LysMe3 were involved in SA, JA, and ROS production. a Comparison of the expression levels of Lyp1, Lyk7 and LysMe3 in cotton plants treated with SA, JA, and ROS signal compounds and H 2 O-treated control plants. b The transcription levels of genes related to SA, JA, and ROS signaling pathways were analyzed in Lyp1, Lyk7, and LysMe3-silenced plants and TRV: 00 control plants by qRT-PCR. Error bars represent the standard deviation of three independent experiments with three technical replicates for each experiment. Asterisks indicate statistically significant differences, as determined by Student's t-tests (*P < 0.05, **P < 0.01) Fig. 9 Lyp1, Lyk7, and LysMe3 activated downstream defense genes. a The expression of defense-related genes (PR1, PR4, PR5, and PR10) was analyzed in Hai7124 and Junmian 1 after inoculation with V. dahliae strain V991. The 0 h treatments of the Hai7124 and Junmian 1 acted as controls. Asterisks indicate statistically significant differences, as determined by Student's t-tests (*P < 0.05, **P < 0.01). b The transcript levels of the four PR genes in cotton treated with SA, JA, and H 2 O 2 were detected, with H 2 O treatment acting as a control. c The expression levels of the defense genes were compared in the TRV: 00 control and VIGS cotton plants. Error bars represent the standard deviation of three independent experiments with three technical replicates for each experiment. Asterisks indicate statistically significant differences, as determined by Student's t-tests (*P < 0.05, **P < 0.01)

Genome-wide distribution and characterization of LysM genes in various plant species
LysMs, carbohydrate-binding modules with a length of approximately 40 amino acids, can bind to Nacetylglucosamine (GlcNAc)-containing glycans, such as peptidoglycan, chitin, and chitin-like compounds [44]. These genes therefore usually act as PRRs to detect PAMPs of invading pathogens, and activate defense genes and plant innate immunity. Based on the genomes of several plants, systematic genomewide investigation of LysM genes has been carried out in these species.
LysM genes are distributed unevenly in the chromosomes in a variety of species [45]. For example, in Arabidopsis, 14 LysM genes, 5 Lyks, 3 Lyps, 3 LysMes, and 3 LysMns, are distributed throughout all five of the chromosomes. In Glycine max, 47 LysM genes, comprising 21 Lyks, 4 Lyps, 16 LysMes, and 6 LysMns, are also positioned on all chromosomes, expect chromosome 12 (https://phytozome.jgi.doe.gov/pz/portal.html). However, in rice, 20 LysM genes, comprising 6 Lyks, 6 Lyps, 4 LysMes, and 4 LysMns are only positioned on 8 of 12 chromosomes, and in P. trichocarpa, 35 LysM genes, comprising 11 Lyks, 7 Lyps, 10 LysMes, and 7 LysMns, are distributed in 12 of 19 chromosomes. The differences in the distribution of LysM genes between species might be related to the individual defense characteristics of each species that developed during the evolutionary process, and interestingly, many duplicated LysM genes also display different expression patterns [10,45]. Here, we first systematically identified 29, 30, 60, and 56 LysM genes in four sequenced cotton species; the diploid cottons G. raimondii and G. arboreum, and the tetraploid cottons, G. hirsutum acc. TM-1 and G. barbadense acc. 3-79, respectively. The 29 LysM genes in G. raimondii were anchored to all 12 chromosomes, except Chr. D7, implying that the LysM genes are widely distributed in the Gossypium genome. From an evolutionary point of view, we can consider that one member of the LysM gene family in the diploid species G. raimondii corresponds to one homologous gene in G. arboreum and two homologs from the A and D subgenomes in tetraploid G. hirsutum acc. TM-1 and G. barbadense acc. 3-79. We found that 14 members of the LysM gene family had such a correspondence in the four sequenced cotton species, indicating that the A-and Dsubgenomes evolved independently after polyploid formation (Additional file 1: Table S1). The other inconsistencies may result from chromosome segmental or tandem duplication events during the evolution of different cotton species, the sequence quality and type of sequencing methods used in different cotton species, or assembly error in partial chromosomal regions. This requires further investigation.
The secretory pathways, signal peptides and transmembrane domains of proteins are particularly crucial to cellular function during defense against both biotic and abiotic stresses [46]. In this study, characteristics of the LysM genes, including their signal peptides (SPs), subcellular localization, and transmembrane domains were investigated in G. raimondii (Additional file 3: Table S2). We found that SPs existed in 15 LysM proteins, suggesting that the proteins are synthesized as pre-proteins, are subsequently cleaved at the signal peptide site to form a mature protein, and function by targeting the general secretary pathway. In addition, 12 LysM proteins, comprising all the GrLyks, GrLyp1, GrLysMe3, GrLysMe9, and GrLysMn7, possessed transmembrane domains; indicating that they have multiple complex functions. In addition to containing transmembrane domains and being involved in secretory pathways, eight LysM proteins, Lyp1-5, Lyk7, Lyk8, and Lyp1, also possessed signal peptides, indicating that they play important roles in the defense against a variety of stresses in cotton.

LysM genes show multiple expression patterns
Previous reports have shown that the expression of plant LysM genes can be both constitutive and induced [10,11,47,48]. Most LysM kinase genes in Glycine max are predominantly expressed in the roots, and orthologous genes have similar tissue expression patterns [10]. In the present study, we systematically analyzed the expression patterns of LysM genes in cotton. Using transcriptome data from G. hirsutum acc. TM-1 vegetative and reproductive organs, a total of 50 LysM genes were found to have diverse developmental and spatial regulation patterns (Fig. 2), and most genes showed both diverse and overlapping expression patterns in various tissues and organs, suggesting that these genes have a range of functions but with the identical conserved domains.
There is an increasing volume of data to indicate that LysM-containing proteins can detect several PAMPs, such as the bacterial oligosaccharide, peptidoglycan, and fungal chitin. Upon detection of PAMPs, these proteins then activate a wide range of physiological responses, including the production of SA, JA, ET, and ROS, as well as mitogen activated protein kinase (MAPK) phosphorylation, calcium influx, and the expression of defenserelated genes [18,49]. In A. thaliana, AtCERK1 was shown to be a key chitin receptor, and mediates chitininduced signaling through homodimerization and phosphorylation [12,14], and in rice, CEBiP associates with chitin-elicitor receptor kinase 1 (OsCERK1) to mediate MAMP-triggered immunity (MTI) in response to chitin [17,50]. To better validate the roles of cotton LysM genes in chitin recognition and chitin signal transduction, we analyzed the expression patterns of 20 highly expressed LysM genes in different Hai7124 tissues following treatment with insoluble crab shell chitin and soluble chitin fragment N-acetylchitohexaose. Sixteen genes were significantly upregulated in response to these two PAMPs at different time points, suggesting that LysM proteins play important roles in chitin recognition (Fig. 3). Further, we found that 12 LysM genes, 4 Lyks, 3 Lyps, 1 LysMe, and 4 LysMns, had high expression levels in Hai7124 after inoculation with V. dahliae strain V991, reaching peak expression levels at different time points (Fig. 4). Lyp1, Lyk7, LysMe3, and LysMn6 were significantly induced following treatment with both chitin signals and V. dahliae. These findings imply that the LysM genes play critical roles in fungal perception and function in cotton defense mechanisms against V. dahliae.
The G. hirsutum and G. barbadense species probably originated from a single hybridization event between Aand D-diploid species, however, the two have very different agronomic and fiber quality characteristics. For example, most of modern G. barbadense cultivars are resistance to V. dahliae, however, G. hirsutum cultivars are not. As described above, we found that Lyp1, Lyk7 and LysMe3 were significantly upregulated in cotton roots at different time points after inoculation with V. dahliae, and had higher expression levels in G. barbadense cv. Hai7124 than G. hirsutum cv. Junmian 1, indicating that these genes act as positive regulators in plant resistance to V. dahliae.

LysM genes function in the defense against biotic stresses
LysM genes mainly function in recognizing chitin elicitor signals and activating plant immune responses. This has been investigated in various plant species, such as Arabidopsis, wheat, rice, and tomato [15,17,47,51]. In Arabidopsis, CERK1, a LysM receptor kinase, responds to chitin elicitors resulting in MAPK activation and ROS generation [14]. Like CERK1, LYK4 and LYK5 have been localized to the plasma membrane, and are involved in chitin recognition, chitin signal transduction and plant innate immunity [15,16]. LYM1 and LYM3, two plasma membrane proteins, physically interact with peptidoglycans and mediate immunity to bacterial infection [33]. In rice, a plasma membrane glycoprotein, OsCEBiP, and a receptor kinase, OsCERK1, act as critical components of chitin signaling recognition pathways [17]. Similarly, LYP4 and LYP6 are promiscuous PRRs for PGN and chitin recognition, and are involved in ROS generation, defense gene activation, and callose deposition in rice [18]. In Medicago, LYK3 is proposed to function as the entry receptor in rhizobial nodulation factor signaling and specifically controls responses to infection [48]. Taken together, these data show that LysMs are important in the recognition of PGN and chitin signals, and the activation of plant immune responses to biotic stresses.
In this study, cotton Lyp1, Lyk7, LysMe3, and LysMn6 genes were upregulated at different time points after V. dahliae treatment. We observed that Lyp1-, Lyk7, and LysMe3-silenced G. barbadense cv. Hai7124 plants had more-severe disease symptoms than the TRV: 00 control and LysMn6-silenced plants after V. dahliae infection (Figs. 5b and 6b). In addition, statistical analysis suggested that the percentage of diseased leaves in Lyp1-, Lyk7-, and LysMe3-silenced plants was significantly higher than in control plants (Fig. 5c, Additional file 7: Table S3; Fig. 6c, Additional file 8: Table S4). These findings imply that Lyp1, Lyk7, and LysMe3, which have transmembrane domains, function as important chitin receptors, allowing them to recognize V. dahliae and induce cotton immunity.
Lyp1, Lyk7, and LysMe3 contribute to V. dahliae resistance through the activation of plant innate immune responses Chitin, a major component of the fungal cell wall, is a typical PAMP and is recognized by PRRs, which activate PTI. PTI is characterized by a wide range of physiological responses, including the production of ROS, calcium influx, and the expression of defense-related genes [4]. In addition, chitin signaling pathways appear to be independent, and different PAMPs might activate a common downstream pathway to induce pathogen resistance [1,52,53]. An increasing volume of data further demonstrates that plasma membrane receptors recognize chitin signals and activate downstream SA, JA, ET, and ROS pathways [3,18,54].
PR genes play important roles in signal recognition and plant immunity. For example, in tobacco, PR1 is involved in the stress response and is associated with resistance to oomycete pathogens [40]. PR4 encodes chitinase, which is an endogenous plant defense enzyme that generates signaling molecules (elicitors) for the induction of further defenses [41]. In A. thaliana, PR5 is thought to regulate SA biosynthesis and lead to the accumulation of high levels of camalexin in order to protect the plant against pathogen infection [42]. In addition, PR10 is suggested to increase pepper's resistance to the oomycete pathogen, Hyaloperonospora arabidopsidis [43]. In the present study, we examined the expression of several defense-related marker genes, PR1, PR4, PR5, and PR10, in Lyp1-, Lyk7-, and LysMe3silenced VIGS plants. The four PR genes were significantly upregulated in control plants after V. dahliae infection (Fig. 9a); but were all significantly downregulation in the Lyp1-, Lyk7-, or LysMe3-silenced plants compared to the control, suggesting that defense-related processes were restrained in the VIGS plants, leading to moderate susceptibility to fungal pathogens such as V. dahliae. When challenged by pathogens, plant levels of the signaling compounds, SA, JA, ET, and ROS change, PR gene expression is induced and plant resistance to infection is enhanced [55,56]. In our study, the expression of PR1, PR4, PR5, and PR10 was significantly increased after SA, JA, or ROS treatment in cotton plants, but was significantly lower in the Lyp1-, Lyk7-, and LysMe3-silenced cotton plants (Fig. 9b, c). Taken together, these findings suggest that cotton LysM-containing proteins are involved in activating downstream defense processes to enhance resistance to V. dahliae.
Recent studies have shown that common mechanisms for receptor activation involve receptor homodimerization or oligomerization and subsequent phosphorylation, and that the LysM receptors might function as protein complexes [16]. In Arabidopsis, AtFLS2 was associated with AtBAK1 upon flagellin treatment, which initiated cellular defense signaling [57,58]. In addition, the AtLYK5-AtCERK1 interaction is chitin dependent, yet AtLYK4 can interact with AtCERK1 independently of the presence of chitin [16]. Here, we confirmed that the three LysM-containing proteins, Lyp1, Lyk7, and LysMe3, act as plasma membrane receptors in cotton and recognize chitin signals, activate downstream SA, JA, and ROS pathways, induce PR gene expression, and enhance resistance to V. dahliae (Fig. 10). However, the functional characteristics of Lyp1, Lyk7, and LysMe3 in the resistance network remain to be clarified. Future investigations into the effect of the three LysMs on specific defense-related genes and their involvement in the defense network will be meaningful, not only to improve our understanding of the molecular mechanisms of LysM-chitin interactions, but also for developing fungal-resistant cultivars through breeding programs for cotton and other crops.

Conclusion
Cotton Verticillium wilt, a highly destructive vascular disease caused by the soil-borne pathogen V. dahliae, leads to devastating reductions in plant mass, lint yield, and fiber quality. Although efforts have been made to produce wilt-resistant cotton cultivars by traditional breeding, to date, Verticillium wilt is not effectively controlled in most cotton producing areas. Here, we focused on LysM-containing proteins, which can sense pathogen PAMPs and promote plant defenses. We mined the key LysM genes involved in the cotton defense response to V. dahliae infection. Lyp1, Lyk7, and LysMe3 were found to act as plasma membrane receptors that recognize chitin signals, activate downstream defense processes, and enhance resistance to V. dahliae. These findings not only clarify the important roles of LysM genes in the defense against V. dahliae infection in cotton, but also enrich our knowledge of the networks of plant resistance.

Chromosomal localization, phylogenetic tree construction, and characterization analysis
MapInspect software (http://www.softsea.com/review/ MapInspect.html) was used to analyze the distribution of LysM genes in Gossypium. ClustalX (version 2.0) was used to construct multiple sequence alignments of LysM proteins without gaps and poorly aligned sections. A phylogenetic tree was generated by the MEGA 5.1 software (http://www.megasoftware.net/) using the maximum likelihood method, and the reliability of interior branches was assessed with 1000 bootstrap resampling. The online Gene Structure Display Server (GSDS) program (http://gsds1.cbi.pku.edu.cn/) was used to analyze exon/intron structures through alignment of the genomic DNA sequences with their corresponding coding sequences. In addition, LysM protein motifs were investigated using MEME (http://meme-suite.org/), which set the maximum number of motifs at 8 to further verify the phylogenetic classification of each protein. Signal peptides were predicted using the SignalP 4.1 Server (http:// www.cbs.dtu.dk/services/SignalP/). The TargetP 1.1 Server (http://www.cbs.dtu.dk/services/TargetP/) was used to analyze the subcellular localization of LysM proteins and the SMART [30] program was used to analyze the transmembrane domains of LysM genes.

Plant materials and treatments
The expression of LysM1 genes was analyzed in G. barbadense cv. Hai7124, G. hirsutum cv. Junmian 1, and G. hirsutum cv. Jinmian 19 following different stress treatments. Seedlings were grown in the same controlled environment chamber under the same conditions: a 16 h light/8 h dark cycle at 28°C for 2 weeks. All necessary permits for collecting Hai7124, Junmian 1 and Jinmian 19 were obtained from Nanjing Agricultural University, Jiangsu Province, China.
Hai7124 and Junmian 1 seedlings, which showed resistance and susceptibility to V. dahliae, respectively, were inoculated with the fungal pathogen, V. dahliae, using the dip-inoculation method [34]. V991, a highly aggressive and defoliating strain of V. dahliae, was obtained from our lab (State Key Laboratory of Crop Genetics & Germplasm Enhancement, Nanjing Agriculture University, Nanjing, China), which was cultured on potato dextrose agar medium (PAD) at 24°C for 4-5 days, and then transferred to Czapek's medium for incubation at 25°C for 5 days [37]. Subsequently, we used deionized water to adjust the concentration to 10 7 conidia per milliliter for inoculation of the seedlings. Seedling roots were harvested at 0, 24, 48, 96, and 144 h after V991 treatment.
Hai7124 seedlings were also treated with 200 mg/mL of insoluble crab shell chitin and soluble chitin fragment N-acetylchitohexaose (Seebio, Shanghai, China). The insoluble crab shell chitin and soluble chitin fragment N-acetylchitohexaose were dissolved in 1% acetic acid and deionized water, respectively. Next, the leaves of Hai7124 seedlings were sprayed with chitin and Nacetylchitohexaose, and leaves were harvested at 0, 0.5, 1, 2, 4, and 6 h following treatment.
G. hirsutum cv. Jinmian 19 seedlings, which have a high tolerance to abiotic stresses, were treated with four solutions of defense-related signaling molecules, containing 100 μM JA, 1 mM ET, 100 mM SA and 10 mM H 2 O 2 , respectively. ddH 2 O was used as a solvent control, and leaves were collected at 0, 1, 4, 12, and 24 h after treatment.
For each treatment, three biological repeats were harvested, then quick-frozen in liquid nitrogen and stored at −80°C before RNA extraction.

RNA isolation and expression pattern analysis
Total RNA was isolated from cotton seedling leaves and roots using the CTAB-acidic phenolic method [61], and the RNA samples (2 μg per reaction) were reversely transcribed into cDNA using the HiScript Q RT Super-Mix for qPCR (+gDNA wiper) (Vazyme).
The gene-specific primers for qRT-PCR analysis were designed using Beacon Designer 7.0. Cotton histone3 (AF024716) was used as the reference gene. All primer information is available in Additional file 9: Table S5. Real-time PCR amplification reactions were performed on an ABI 7500 Real Time PCR System (Applied Biosystems, USA) using AceQ SYBR Green Master (Low Rox Premixed) (Vazyme) with three technical replicates for each biological sample. Expression data from three biologically independent experiments were analyzed and presented as means ± S.D.
For statistical analysis, all generated data in this study were repeated at least three times on three biological replicates. The data from three biologically independent experiments were analyzed and presented as means ± S.D. In addition, statistical significance was determined by Student's t-tests.

Cloning of the LysM genes in G. barbadense cv. Hai 7124
Based on the known sequences of four cotton species, gene-specific primers were designed using Primer 5.0 software to amplify the homologous genes of LysM with complete open reading frames (ORFs) in G. barbadense cv. Hai7124, a cultivar resistant to V. dahliae (Additional file 9: Table S5). High-fidelity ExTaq DNA Polymerase (TaKaRa Biotechnology [Dalian] Co. Ltd., China) was used in standard PCR reactions. All PCR products were cloned into pMD19-T cloning Vectors (TaKaRa, Dalian, China) and transformed into bacterial strains E.coli DH5α that was obtained from our lab (State Key Laboratory of Crop Genetics & Germplasm Enhancement, Nanjing Agriculture University, Nanjing, China). At least six clones per gene were randomly selected and sequenced.

Vector construction and functional characterization of candidate genes via virus-induced gene silencing (VIGS)
The pTRV1 and pTRV vectors used for VIGS analysis were generously provided by Dr. Libo Shan of Texas A & M University (College Station, TX, USA). GhCLA1 (Cloroplastos alterados 1), which encodes 1-deoxy-Dxylulose-5-phosphate synthase, was used to construct a pTRV: GhCLA1 vector that acted as a control to verify the efficiency of the VIGS procedure [62]. TRV vectors were constructed to silence target genes: TRV: Lyk7 contained a 400 base pair (bp) gene-specific fragment of Lyk7 cDNA, TRV: Lyp1 included a 447 bp fragment of Lyp1, TRV: LysMe3 contained the full-length 312 bp sequence of LysMe3, and TRV: LysMn6 included a 494 bp gene-specific fragment of LysMn6. The primers used for constructing the VIGS vectors are shown in Additional file 9: Table S5.
These vectors were transformed into Agrobacterium tumefaciens strain GV3101 that was obtained from our lab (State Key Laboratory of Crop Genetics & Germplasm Enhancement, Nanjing Agriculture University, Nanjing, China), and the Agrobacteria containing TRV1 were subsequently mixed with TRV: 00, TRV: GhCLA1, TRV: Lyp1, TRV: Lyk7, TRV: LysMe3, and TRV: LysMn6 at a 1:1 ratio, before being infiltrated into the fully expanded cotyledons of eight-day-old Hai7124 cotton seedlings [63]. All plants were grown in the same growth chamber at 23/21°C (day/night), with a 16 h light/8 h dark cycle and changes in plant phenotypes were observed. Two weeks after Agrobacterium infiltration, the TRV: GhCLA1 plants showed highly uniform bleaching in newly emerged leaves. We harvested the leaves from at least three seedlings per treatment and isolated their RNA. All VIGS cotton seedlings, as well as the susceptible control Junmian 1, and the resistant control Hai7124 without Agrobacterium infiltration, were removed from the soil and dip-inoculated with V991 conidia suspension (1 × 10 7 conidia/mL) as described previously [34]. Subsequently, all plants were grown in the same growth chamber at 25/23°C (day/night), with a 16 h light/ 8 h dark cycle for 7 weeks and the ratio of diseased to healthy leaves was analyzed. Each treatment was applied to more than 20 plants and the VIGS experiments were repeated at least three times to increase the reliability of the results. Statistical analyses were used to compare the percentage of diseased leaves in experimental plants with that in TRV: 00 controls. Subcellular localization of Lyp1, Lyk7, and LysMe3 in onion epidermal cells The ORFs of the Lyp1, Lyk7, and LysMe3 genes were fused to green fluorescent protein (GFP) under the control of the 35S promoter in the pBin-GFP4 expression vector [64]. Plasmids harboring GFP alone were used as controls. The onions were bought in the vegetable supermarket and the onion epidermal cells were cultured in MS culture medium at 24°C for 3 h in the lab. Plasmid DNA at a concentration of 1 μg/ μL or greater was used for biolistic bombardment. Helios Gene Gun (Bio-Rad) was used to bombard plasmid DNA into onion cells. After transformation, the onion cells were incubated at 25 ± 1°C for 16 h in dark conditions. Fluorescent signals were visualized with a confocal microscope (Zeiss, LSM710; Germany), and 20% sucrose solution was used for the plasmolysis of the onion cells.

Additional files
Additional file 1: Table S1. Identification of the LysM genes in G. raimondii and their phylogenetic relationships with those in G. arboreum, G. hirsutum and G. barbadense. (XLS 39 kb)