Association of GhGeBP genes with fiber quality and early maturity related traits in upland cotton

Abstract

Transcription Factors (TFs) are key regulators of how plants grow and develop. Among the diverse TF families, the Glabrous-enhancer binding protein (GeBP) family plays a key role in trichome initiation and leaf development. The specific roles of GeBP TFs in plants remain largely unexplored, although GeBP transcription factors play important roles in plants. This study identified 16 GhGeBP genes in Gossypium hirsutum, ranging from 534 bp (GhGeBP14) to 1560 bp (GhGeBP2). Phylogenetic analysis grouped 16 GhGeBP genes clustered into three subgroups, unevenly distributed across 14 chromosomes. Analysis of the cis-acting elements revealed 408 motifs in the 2 kb upstream regions of the promoters, including stress-responsive, phytohormone-responsive, and light-responsive elements. Tissue-specific expression analysis showed 8 GhGeBP genes were highly expressed across all tissues, while GhGeBP4 and GhGeBP12 were down-regulated under conditions of drought, salt, cold, and heat stress. A genome-wide association study (GWAS) identified GhGeBP4 was associated with fiber micronaire (FM) and fiber strength (FS), while GhGeBP9 was linked to the node of the first fruiting branch (NFFB) and flowering time (FT). Haplotype analysis revealed that GhGeBP4-HAP2 exhibited higher fiber quality traits, while GhGeBP9-HAP2 was associated with early maturity. The results of this study offer significant insights that are worthy of further investigation into the role of the GhGeBP gene family in G. hirsutum and promising targets for marker-assisted selection strategies in cotton breeding programs, particularly for improving fiber quality and early maturity traits.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-024-10983-y.

Background

Transcription factors (TFs) play a pivotal role in regulating the biological processes of plant growth and development. The complexity of higher organisms necessitates the involvement of a diverse array of TFs to orchestrate a broad spectrum of biological functions [1]. TFs typically bind to cis-acting elements, modulating gene expression in answer to various environmental stresses [2–4]. Among the various TF families, the GeBP family is unique to plants and plays a crucial role in initiating trichome formation that regulats the generation of trichomes initiation by specially recognizes and binds with the Glabrous1(GL1) and have a decisive function in leaf growth and development [5]. Despite its importance, research on GeBP TFs specific roles in plants remains limited. Model organisms are the main focus of current studies, like Arabidopsis thaliana [6–9] and Glycine max [10, 11], which possess 22 and 9 GeBP TF members, respectively. In Arabidopsis, the GeBP TFs exhibit a conserved central DNA-binding domain and a C-terminal region characterized by a putative leucine-zipper pattern. These structural features suggest a shared mechanism of action among family members, though functional diversity is evident from the differential expression patterns observed across various tissues and developmental stages.

In the regulatory networks governing plant development, transcription factors (TFs) play a central role, particularly through their participation in hormonal pathways and interactions. A notable example is the Glabrous-enhancer binding protein binding to the cis-regulatory element of the GL1 gene and regulating its transcription. The GL1 gene, belongs to the MYB family, is essential for determining epidermal cell fate and is modulated by phytohormones such as gibberellins and cytokinins, reflecting the intricate link between genetic regulation and hormonal control in plant development [12]. Further regulation of GeBP expression is mediated by the KNOX family TF, BREVIPEDICELLUS (BP) [6], which not only upregulates GeBP but also enhances the cytokinin signaling pathway in the apical meristem of the stem [13]. Others studies also speculated that GeBP may control the appearance of epidermal hair through regulating the gibberellin and cytokinin pathways [9]. In addition, previous studies have suggested that four TF families including, MBF1, jumonji, ULT, and GeBP are mainly involved in plant development and hormone responses, and usually do not participate in stress responsiveness under stress conditions [6, 9, 14–16]. However, a study by Ray et al. (2011) indicated an upregulation of these TFs under water deficit conditions, suggesting a broader role for GeBP in drought stress responsiveness [17]. Likewise, a research by Wang et al. (2023) demonstrated that, in Brassica rapa, BrGeBP5 may regulate low-temperature stress as well as BrGeBP3 and BrGeBP14 were probably in response to drought stress [18]. Additionally, a research by Liu et al. (2023) showed that overexpression of MdGeBP3 from apples in Arabidopsis led to a decrease in drought tolerance [19]. These researches revealed that GeBP may be crucial for developing stress-tolerant crop varieties.

AtCPR5 is an important regulatory factor that controls the growth and development of plants [20, 21]. In the classic disease resistance pathway, AtCPR5 is downstream in identifying pathogen genes and simultaneously participates in both the Non Expressor of PR Genesl (NPR) and NPR dependent disease resistance signaling pathways. AtGeBP/GPL regulates several genes in the AtCPR5 signaling pathway, which participate in stress resistance and cell wall metabolism processes in the AtCPR5 signaling pathway. Research involving multiple mutants (Atgebp, gpl1, gpl2, gpl3, and cpr5) has further elucidated the role of AtGeBP/GPL, showing that it specifically controls cell elongation processes dependent on AtCPR5 signaling, without affecting cell proliferation [22].

Cotton is one of the most important crops globally, serving as a crucial raw material for the textile industry. While the GeBP gene family has been well characterized in Arabidopsis and soybean [6, 10], a comprehensive genome-wide analysis of the GhGeBP gene family in upland cotton has yet to be conducted. Recent advancements have seen the release of high-quality, assembled genomes and re-sequencing data for cotton, significantly enhancing our capacity to study genomic variations [23–27]. Despite these developments, the identification of critical genomic variations that influence fiber quality, yield, and early maturity traits in upland cotton have not been sufficient to date. This void underscores a significant need for focused research on these economically vital traits. In response to this need, the aim of our study was to carry out a comprehensive analyses of the GhGeBP gene family in upland cotton. We utilized genome-wide data to analyze cis-elements and evaluated the expression patterns of GhGeBP genes in various tissues, and in response to diverse environmental treatments. Additionally, we identified elite haplotypic variations, which could be pivotal for breeding programs. Our research not only enhances the genetic understanding of the GhGeBP gene family in cotton but also provides valuable insights and resources for the development of marker-assisted selection strategies in cotton breeding programs.

Materials and methods

Identification and characterization of GhGeBP proteins in upland cotton

Twenty-two of 23 AtGeBP sequences of proteins from Arabidopsis thaliana were retrieved from PlantTFDB (https://planttfdb.gao-lab.org/family.php?%20sp=Ath&fam=GeBP) [9] and 1 (AT2G20805) from a study by Li et al.(2023) [28]. These sequences served as queries to screen for homologous GhGeBP genes. Corresponding sequences of proteins for upland cotton were obtained from the CottonGen (https://www.cottongen.org/) [23, 29]. Further, identification of putative GeBP proteins in upland cotton was conducted using a two-step bioinformatics approach. Initially, the BLAST was employed with a threshold of E-value less than 1 × 10⁻¹⁰ to ascertain preliminary candidates. Subsequently, The presence of the GeBP-specific conserved domain (DUF573) was tested in these candidates by using the HMMER software (version: 3.0) to identify GhGeBP proteins under default parameters (e-value less than 0.001) [30–32]. To confirm the structural integrity of the identified proteins, all candidates were further analyzed against the Conserved Domain Database (CDD) hosted by the NCBI (https://www.ncbi.nlm.nih.gov/cdd/). This step ensured the accuracy of our domain-specific identifications.

Physicochemical characterization and subcellular localization

GeBP protein family physicochemical properties, including amino acid number, molecular weight (MW), and isoelectric point (PI), were determined using ExPASy ProtParam (http://web.expasy.org/protparam/). For subcellular localization predictions, the cello BUSCA (http://www.busca.biocomp.unibot.it/) was utilized to infer the likely cellular compartments where the GhGeBP proteins might exert their biological functions, further informing our understanding of their role in cotton physiology.

Sequence alignment and phylogenetic tree construction

The full-length GeBP protein sequences of Arabidopsis thaliana, Oryza sativa, Glycine max, and upland cotton were aligned using the FFT-NS-2 algorithm implemented in the MAFFT software (version: 7.310) [33]. This alignment provided the basis for subsequent phylogenetic analyses. Iqtree (version: 2.0.3) was then used to construct a maximum likelihood (ML) phylogenetic tree, selecting ‘JTT + F + R4’ as the best fit of the substitution model, supported by 1,000 bootstrap replicates [34]. The resulting phylogenetic tree was visualized using the ‘ggtree’ package in R software(version 4.3.1) [35].

Chromosome distribution, collinearity analysis and selection pressure calculation

Chromosome localization of the GhGeBP genes was visualized using TBtools (version: 2.080) [36]. The MCScanX toolkit (https://github.com/wyp1125/MCScanX) was employed to set genome-wide synteny relationships and to identify collinear gene pairs within the GhGeBP gene family. Selection pressure on the GhGeBP family genes was assessed using the KaKs_Calculator (version: 2.0) [37], with the Yang-Nielsen (YN) model [38] specified for calculation of the non-synonymous (Ka) and the synonymous (Ks) substitution rate.

Gene structure, cis-acting elements, motifs analysis, and prediction of protein tertiary structure

Gene structure information, including promoter regions, 5’UTR, 3’UTR, and exon-intron organization for the GhGeBP genes, was extracted from the GFF3 file of the genome of upland cotton [23]. Visualization of these structures was conducted using the Gene Structure Display Server (version: 2.0) (http://gsds.gao-lab.org/). Additionally, these genes’ upstream sequences (2,000 bp) were retrieved using a custom Python script and analyzed for cis-acting elements using the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/). The MEME (http://meme-suite.org/index.html) was predicted motifs for GeBP proteins. Finally, the visualization were created by TBtools [36]. Protein tertiary structures were predicted through Phyre2 online websites (http://www.sbg.bio.ic.ac.uk/phyre2) [39], and the 3D structure prediction of GhGeBP proteins were in Fig. S1.

Association analysis of GhGeBP genes with early maturity and fiber quality related traits

To verify the potential functional of GhGeBPs, an association analysis was conducted between these genes and 12 key agronomic traits (5 fiber-related traits and 7 early maturity related traits) in 355 upland cotton accessions. Accessions used for fiber related traits were planted in Anyang, Henan, China (36°08’N,114°48’E), and Shihezi, Xinjiang, China (44°31’N,86°01’E) in 2014 and 2015. Every year normally open bolls are taken from each repeating fruit branch. The fiber-related quality traits (FL(mm), FS(cN/tex), FU(%), FM and FE(%)) were detected in 10–15 g fiber samples [40]. Besides, from 2015 to 2017, multi-environment experiments were conducted in Anyang, Henan, China (36°08’N,114°48’E), Hubei, Huanggang, China (30°57’N, 114°92’E) and Shihezi, Xinjiang, China (44°31’N,86°01’E). Seven early maturity related traits (FT (days), FBP (days), WGP (days), YPBF (%), NFFB, HNFFB (cm) and PH (cm)) were identified [27]. Total 67 single nucleotide polymorphisms (SNPs) within the 2,000 bp upstream and downstream regions of the GhGeBPs were taken from the resequencing data of the 355 upland cotton accessions [27]. A linear mixed model (LMM) was utilized to execute the associations by incorporating all phenotype and SNP data based on GEMMA software (version: 0.98.3) [41]. The significant threshold value for association was set at -log₁₀P ≥ 2. Manhattan plots were generated using the ‘ggplot2’ package in R software [42].

Haplotype and genetic analysis of GhGeBP genes

The phenotypic effects of elite haplotypes were visualized using the ‘ggplot2’ package in R software [42]. Nucleotide diversity values for each group were estimated using VCFtools software (version: 0.1.16) [43] based on the release and breeding years of each accession and geographical distribution.

Transcriptome expression analysis of GhGeBP genes

The expression profiling from GhGeBPs in different tissues were analyzed using four sets of public transcriptome data (PRJNA490626, PRJNA509318, PRJNA793063, and PRJNA529497), available from NCBI (https://www.ncbi.nlm.nih.gov/bioproject/). The dataset PRJNA490626 [24] comprised samples from multiple organs of the upland cotton variety ‘TM-1’, including roots, stems, leaves, bracts, anthers, fibers, ovules, and various reproductive organs. And beyond that, it also contains several stress treatments (NaCl (0.4 M), PEG (200 g/L), 4℃ and 37℃) that act on ‘TM-1’. PRJNA509318 [27] included flower bud samples collected at five developmental stages (square, 5 days post-square (DPS), 10 DPS, 15 DPS, and 20 DPS). PRJNA793063 [44] contained long and short fiber samples from an interspecific backcross inbred line (BIL) population that was derived from a cross between Gossypium hirsutum (CRI36) and Gossypium barbadense cotton (Hai7124) at 5, 10 and 15 DPA. PRJNA529497 [45] encompassed 10 and 20 DPA fiber samples from the semi-wild species Gossypium hirsutum ‘TX2094’ and the cultivated species ‘Acala Maxxa’. Quality control for each transcriptome dataset was conducted using Trimmomatic (version: 0.39), with modified parameters (PE threads: 50, HEADCROP: 7, SLIDINGWINDOW: 4:15, MINLEN: 80). GhGeBP genes expression levels were quantified as Transcripts Per Million (TPM) values using Salmon (version: 1.10.0) [46], based on the coding sequence file of ‘TM-1’ [23, 47]. The ‘pheatmap’ package [48] in R software was used to generate heatmaps, with expression values normalized to log₂ TPM.

qRT‑PCR analysis of GhGeBP9

For further validation, qRT-PCR analyses were conducted. Primers designed are detailed in Table S1. Additionally, accessions ‘Han9609’, ‘Xinluzhong 34’ were planted at the Sanya, Hainan, China (18°36’N, 109°17’E). Samples from axillary buds were collected, immediately frozen in liquid nitrogen, and stored at − 80℃. Total RNA was extracted using the FastPure Universal Plant Total RNA Isolation Kit (RC411–01, Vazyme). cDNA synthesis was conducted using HiScript^® II Q RT SuperMix for qPCR (R223). Quantitative real-time PCR was performed using the Light Cycler 480 II system, and gene expression levels were calculated using the 2^−ΔΔCt method [49] with three replicates per sample.

Results

Identification and protein characteristics of GeBPs gene family in upland cotton

Genome-wide searches have identified total 16 GhGeBP gene members containing the complete DUF573 domain in the G. hirsutum genome and the genes were renamed according to their chromosomal positions (Table S2). The physical and chemical properties of these GhGeBP proteins were further analyzed. The coding sequences varied from 534 bp (GhGeBP14) to 1560 bp (GhGeBP2), with 1,046 bp average length. The GhGeBP genes encoded proteins with amino acid lengths ranging greatly from 177 aa to 519 aa, in which GhGeBP2 has the largest number of amino acids and GhGeBP14 has the smallest. The predicted molecular weights (MW) ranged from 19,939.60 to 58,833.48 Da. Isoelectric points (pI) varied from 4.63 to 10.23, with 11 genes encoding acidic proteins (pI ≤ 6.5, average 5.30) and 4 genes (GhGeBP3, GhGeBP6, GhGeBP11, and GhGeBP14) encoding alkaline proteins (pI > 8, average 9.81). The grand average of hydropathicity (GRAVY) values ranged from − 0.943 to -0.466, indicating that all GhGeBP proteins are predominantly hydrophilic. Predicted subcellular localization revealed that the majority of GhGeBP proteins are situated in the nucleus, except for GhGeBP14, which is predicted to be in the cytosol. This suggests that most of GhGeBP proteins are likely involved in nuclear functions (Table S2).

Phylogenetic analysis of the GeBP gene family

An unrooted phylogenetic tree was constructed by the full-length amino acid sequences from A. thaliana (23 genes), O. sativa (15 genes), G. max (9 genes), and G. hirsutum (16 genes) to compare the evolutionary relationships among GeBP genes (Table S3). In accordance with the phylogenetic tree, the 63 GeBP genes were divided into six distinct subgroups (Fig. 1). Among them, 17 AtGeBP genes formed an independent cluster in group I, 3 GmGeBP genes were grouped in group III. This indicates that GeBP genes are conserved across both monocotyledons and dicotyledons. In G. hirsutum, the 16 GhGeBP genes were divided into three subgroups, though unevenly distributed: 8 members in group II (the largest subgroup of GhGeBP genes), 2 in group V, and 6 in group VI. The homologous genes from the At and Dt subgenomes of G. hirsutum clustered together within the same groups due to their high similarity. For instance, GhGeBP3 and GhGeBP11 clustered together in group V, sharing an average similarity of 23.48% with AtGeBP7. In group VI, GhGeBP6 and GhGeBP14 were grouped together, sharing an average similarity of 32.59% with OsGeBP7. These close phylogenetic relationships suggest that these homologous genes may share similar genetic functions.

Fig. 1.

Phylogenetic tree of GeBP genes between upland cotton, Arabidopsis thaliana, Glycine max and Oryza sativa. Different symbols correspond to different species, The color-coded circular nodes at the branch points of the tree represent bootstrap values, providing statistical support for the phylogenetic inferences. Higher bootstrap values indicate greater reliability of the inferred evolutionary relationships

Chromosome distribution and collinearity analysis

The chromosome position analysis revealed: GhGeBPs are unevenly distributed on the upland cotton chromosomes. The GeBP genes in upland cotton were mainly paired distribution on 14 chromosomes (A01/D01, A05/D05, A07/D07, A10/D10, A11/D11, A12/D12, A13/D13). Chromosome A13 and Chromosome D13 contained two GeBP genes, and other chromosomes both had only one GeBP gene (Fig. 2a), which was due to genomic segmental duplication or rearrangement. In addition, we simultaneously identified 16 GhGeBP genes in upland cotton that were involved in 10 blocks of synteny, with the exception of GhGeBP2, GhGeBP10, and GhGeBP12 (Fig. 2b). Most of synteny blocks, 8 (80%) were located between the At and Dt sub-genomes, including 13 orthologous GhGeBP genes in whole upland cotton genome. Among these genes, GhGeBP8 and GhGeBP13 are located within three collinearity synteny blocks, respectively. GhGeBP5 and GhGeBP16 involve two collinearity synteny blocks, respectively. While the remaining genes were involved in one collinearity synteny blocks.

Fig. 2.

GhGeBP genes chromosomal localization and duplication relationships. a Localization of GhGeBP on the chromosomes of upland cotton; b Gene duplication relationship among GhGeBP genes. The outer circle is color-coded to denote the different subgenomes: pink means the At subgenome and blue means the Dt subgenome. The line graph within the circle represents gene density across the chromosomes. Background shading indicates various types of gene duplication relationships: the gray background signifies all gene duplications within the genome, the red background shows duplications between chromosome groups, and the purple and blue backgrounds highlight duplications within the At and Dt subgenomes, respectively

Selection pressure analysis (Ka/Ks ratios)

Across the entire genome, the GhGeBP gene family comprises five homologous gene pairs, all with Ka/Ks ratios below 1 (Table S4). Notably, four of these pairs have Ka/Ks ratios below 0.31, implying that the GhGeBP genes have undergone purifying selection during their evolution. This selective pressure has likely contributed to the contraction of this gene family.

Analysis of the gene structure of GhGeBP genes and the cis-acting elements

Gene promoters contain specific sequences that bind to transcription factors to control the expression of genes that play a role in development and responses to stress throughout the plant’s life cycle. The promoter regions on 2,000 bp upstream of the start codon of each GhGeBP gene were analyzed using PlantCARE to further predict the cis-acting regulatory elements of GhGeBP genes. Total 408 cis-acting elements, which represents 19 types, were predicted and grouped into five categories (Fig. 3a, b and Table S5): light-responsive, stress-responsive, phytohormone-responsive, plant organogenesis-related, development-related elements and plant growth. Among these cis-acting elements, numerous light-responsive elements were found in the GhGeBP genes promoter regions, representing 55.29% of the predicted cis-elements. These included GT1-motif, G-Box, Box4, and others, which were distributed across each GhGeBP gene promoter region. We found that Auxin-responsive elements (AuxRR-core, AuxRE, and TGA-element) were in 11 GhGeBP genes promoter regions. We also identified five types of hormone-responsive regulatory elements. Thirty-eight ABRE motifs associated with abscisic acid responsiveness were particularly abundant, with GhGeBP12 containing the most motifs. Salicylic acid-responsive TCA-elements were present in GhGeBP1, GhGeBP2, GhGeBP7, GhGeBP9, GhGeBP15, and GhGeBP16. The MeJA-responsive CGTCA motif was uniquely found in 11 GhGeBP genes (GhGeBP2, GhGeBP3, GhGeBP4, GhGeBP5, GhGeBP7, GhGeBP8, GhGeBP10, GhGeBP11, GhGeBP12, GhGeBP13, and GhGeBP16). Including GARE-motif, P-box, and TATC-box, elements that react to gibberellin were identified in GhGeBP12, GhGeBP16, GhGeBP7, GhGeBP8, and GhGeBP9. These results demonstrate that most GhGeBP genes might engage in multiple phytohormone signaling pathways. In addition, regarding stress-related responsive, TC-rich repeats (responsing defense and stress), GC-motif (anoxic inducibility), ARE (anaerobic induction), and LTR (low-temperature responsiveness) were more prevalent. Endosperm-expressed elements, meristem-expressed elements, and palisade mesophyll cell differentiation elements, for plant organogenesis and development, were recognized in GhGeBP10 promoter regions. Circadian control regulatory elements were detected in several genes, including GhGeBP3, GhGeBP5, and GhGeBP6. These results suggest that GhGeBP genes may be of importance in hormone signal transduction, cotton development, and response to abiotic stress. Table S5 contains detailed information on the cis-elements.

Fig. 3.

Cis-acting elements and gene structures of the GhGeBP promoter were analyzed comparatively. a Clustering analysis of genes in line with the GhGeBP phylogenetic tree; b Cis-acting elements visualization and classification in the GhGeBP promoter region, where each color represents a specific functional category of cis-acting elements, enabling a visual differentiation of regulatory motifs associated with various biological processes; c Gene structure of GhGeBP, displaying the arrangement of exons, introns, and untranslated regions (UTRs); d The distribution of conserved domains on GhGeBP proteins

Ground on genomic sequence analysis and gff3 file, GhGeBP genes structures and domain distribution were conducted (Fig. 3c, d). Except for GhGeBP2, which contains an additional YhaN domain, all other proteins have only a single DUF573 domain. The exon number varied between 1 and 8, and GhGeBP genes mostly displayed either intron-less structures or contained only a few introns. Of the 16 GhGeBP genes, 56.25% (GhGeBP3, GhGeBP4, GhGeBP5, GhGeBP7, GhGeBP8, GhGeBP9, GhGeBP13, GhGeBP14, and GhGeBP16) contained only one exon and lacked introns, four were two-exon genes (GhGeBP1, GhGeBP6, GhGeBP11, and GhGeBP15), and three had three or more exons (GhGeBP2, GhGeBP10, and GhGeBP12). GhGeBP2 had the highest number of exons (8) and introns (7). Notably, GhGeBP genes within the same phylogenetic cluster displayed similar gene structures. Furthermore, the 3D structure prediction of GhGeBP proteins certified the result above (Fig. S1).

Analysis of the tissue-specific pattern of GhGeBPs expression

To fully understand how the GhGeBP genes function, we analyzed their spatial and temporal expression profiles utilizing RNA-seq data in kinds of tissues (root, leaf, stem, sepal, torus, anther, bract, filament, petal, pistil, ovule, and fiber) at every five days between 10 and 25 days post-anthesis (DPA) under normal growth conditions (Fig. 4a). The results showed that 8 GhGeBP genes (GhGeBP1, GhGeBP3, GhGeBP6, GhGeBP8, GhGeBP9, GhGeBP11, GhGeBP14, and GhGeBP16) exhibited high expression levels relatively across all tissues. Contrarily, GhGeBP12 and GhGeBP15 displayed weak or no expression in any of the tissues analyzed. Interestingly, GhGeBP5 and GhGeBP7 were specifically expressed during ovule development, while GhGeBP1 and GhGeBP9 were strongly expressed during fiber development. Notably, GhGeBP1 exhibited the highest expression level in fiber development (TPM ≥ 23), followed by GhGeBP9 (TPM ≥ 43). Overall, throughout the reproductive cycle of upland cotton, most members of the GhGeBP gene family were expressed., suggesting their involvement in regulating the cotton to grow and develop.

Fig. 4.

Heatmap of GhGeBP genes’ expression in diverse tissues and under various conditions of stress. a Heatmap of GhGeBP expression in diverse tissues, with color gradation from blue to red indicating an increase in expression levels. b Heatmap of the time-based expression pattern of GhGeBP genes at 1, 3, 6, 12, and 24 h after treatment under drought stress induced by 200 g/L PEG; c Heatmap of GhGeBP expression at 1, 3, 6, 12, and 24 h under salt (0.4 M NaCl) stress conditions; d Heatmap of GhGeBP expression at 1, 3, 6, 12, and 24 h under cold (4 °C) stress conditions; e Heatmap of GhGeBP genes expression at 1, 3, 6, 12, and 24 h under heat (37 °C) stress conditions

Analysis of the pattern of GhGeBP expression under the influence of abiotic stress

To investigate whether GhGeBP genes participate in abiotic stress responses, their expression patterns were analyzed under drought (200 g/L PEG), salt (0.4 M NaCl), cold (4 °C), and heat (37 °C) stress conditions (Fig. 4b, c, d, e). Collating the four stress conditions, more GhGeBP genes exhibited changed expression in answer to dryness and salt stresses than to cold and heat stresses. Most GhGeBP genes were up-regulated under salt stress, with the highest differential expression observed at 12 h, after which expression levels significantly decreased at 24 h. In contrast, GhGeBP4 and GhGeBP12 were consistently downregulated at 3, 6, 12, and 24 h, demonstrating that these two genes might make negative affect in answer to salt stress. Similarly, the pattern of GhGeBP expression under dryness stress (simulated with PEG) closely resembled those observed under salt stress. Notably, GhGeBP9 and GhGeBP15 were significantly downregulated at 3, 6, 12, and 24 h under PEG-induced drought conditions. Interestingly, GhGeBP15 exhibited relatively low expression in all tissues but showed increased expression and significant upregulation under all four stresses which suggests that GhGeBP15 may be important to enhance resistance to multiple nonliving stresses.

GWAS of characteristics of upland cotton that correlate with early maturity and fiber quality

In a GWAS of 355 cotton germplasm accessions, all 67 SNPs dispersed across the 16 GhGeBP genes with a minor allele frequency (MAF) > 0.05 and a resequencing data miss rate < 20% were used. 23 Of these SNPs were located in the At sub-genome and 43 in the Dt sub-genome. Most SNPs were found in intergenic upstream regions (52.23%). Using a linear mixed model (LMM), total 16 and 6 important SNPs were identified for seven early maturity characteristics (FT, FBP, WGP, YPBF, NFFB, HNFFB, and PH) and five fiber quality characteristics (FU, FM, FL, FE and FS), respectively, across five GhGeBP genes (GhGeBP4, GhGeBP9, GhGeBP11, and GhGeBP15) (Table S6). Only GhGeBP4 was found in the At subgenome, whereas the remaining genes (GhGeBP9, GhGeBP11, and GhGeBP15) were situated in the Dt subgenome. GhGeBP9, exhibited a robust SNP cluster (15 significant SNPs) associated with NFFB and FT, with -log₁₀(P) values ranging from 13.05 to 13.98 and a phenotypic variation explained (PVE) of over 14% (14.54–15.50). Additionally, a stable SNP (rsD07_14834977) in the promoter region of GhGeBP11 with identified through eight traits (FT, HNFFB, WGP, YPBF, NFFB, PH, FBP, and FE), with a PVE ranging from 2.45 to 4.89% and -log₁₀(P) values ranging from 2.52 to 4.59. Three SNPs (rsD13_1727983, rsD13_1727984, and rsD13_1728125) downstream of GhGeBP15 were associated with FM, explaining 1.98–2.09% of the observed PVE. Interestingly, two SNPs (rsA10_105653025 and rsA10_105655897) in the promoter and 3’UTR regions of GhGeBP4 were linked to FS and FM, explaining 1.99% and 2.05% of the PV, respectively. Thus, GhGeBP4 and GhGeBP9 could be considered major candidate genes to continue dissecting.

Candidate gene for fiber quality on A10

In the association analysis of fiber-related traits, GhGeBP4 was significantly associated with the highest number of traits (FM and FS) (Fig. 5a). Four SNPs were identified within GhGeBP4, located in the 2 kb upstream region and the 3’UTR (Fig. 5b). According to publicly available transcriptome data, GhGeBP4 showed peak expression at 15 DPA as the fiber developed, followed by a gradual decline (Fig. 5c). Haplotype analysis classified GhGeBP4 into two haplotypes, with HAP2 showing higher FS, FL, FU, and FE (Fig. 5d). This suggests that HAP2 may represent a superior haplotype for fiber quality, providing a new foundation for breeding elite fiber accessions. Gain insight into the genetic basis of HAP2 and its geographical distribution, 50 domestic upland cotton varieties containing HAP2 were grouped into four regional clusters: Northwest Inland Cotton Region (NIR) (29 varieties), Northern Super-Early Maturing Region (NSER) (2 varieties), Yellow River Region (YRR) (9 varieties), and Yangtze River Region (YZRR) (10 varieties) (Fig. 5e). The NIR cluster accounted for over 50% of the germplasm, consistent with the superior fiber quality of upland cotton in Xinjiang. Based on the variety approval years, the 355 germplasm resources were categorized into four breeding periods: pre-1950 (10 varieties), 1950–1979 (32 varieties), 1980–1999 (65 varieties), and 2000 to the present (175 varieties). An additional 73 varieties had unknown approval years. Nucleotide diversity analysis of this region (A10: 105,652,245 − 105,658,012) across these four breeding periods showed the highest diversity in GhGeBP4 before 1950, followed by a gradual decrease, reflecting breeding goals focused on improving fiber quality traits (Fig. 5f). Further analysis of BILs derived from a cross between CRI36 and Hai7124 revealed that during the fiber elongation phase, although generally decreasing, the expression of GhGeBP4 was higher in LF Group (long fiber group) in comparison to SF Group cotton (short fiber group) (Fig. 5g). Additionally, expression analyses of GhGeBP4 at 10 DPA and 20 DPA between the cultivated variety (Maxxa) and a semi-wild type (TX_10) showed that GhGeBP4 expression was consistently higher in the cultivated upland cotton than in the semi-wild type (Fig. 5h). These findings suggest a central role of GhGeBP4 in developing fibers, potentially contributing to the enhanced fiber qualities observed in cultivated varieties.

Fig. 5.

Association analysis of characteristics of upland cotton that correlate with fiber quality. a Manhattan plot across five characteristics of upland cotton that correlate with fiber quality; b Gene structure and two haplotypes of GhGeBP4; c Heatmap of 15 GhGeBP family genes expression during the fiber development period; d Box plots comparing the phenotypic variation of the five different traits between the two haplotypes identified; e Pie charts of the distribution of the 50 HAP2-containing accessions in the four cotton ecological regions; f Nucleotide diversity during four distinct breeding periods, with a vertical line marking the location of GhGeBP4; g The expression of GhGeBP4 on LF Group (long fiber group) and SF Group (short fiber group) cotton accessions at 5–15 DPA; h Significance analysis of GhGeBP4 expression in cultivated (maxxa) and semi-wild (TX_10) species at 10–20 DPA. (** P < 0.01, * P < 0.05)

Candidate gene for early maturity on D01

In the association analysis of early maturity traits, SNPs within GhGeBP9 showed significantly higher associations than other members of the GhGeBP gene family, particularly for NFFB (Fig. 6a). Most SNPs were in downstream regions, with no variants identified within the gene itself (Fig. 6b). Further haplotype analysis identified two haplotypes of GhGeBP9, HAP1 and HAP2. Varieties carrying the HAP2 haplotype exhibited lower NFFB, lower HNFFB, a shorter FT, a shorter FBP, and a shorter WGP (Fig. 6c). This suggests that HAP2 may represent a superior haplotype for early maturity traits, providing valuable insights for future research. The geographical distribution of these haplotypes showed that the distribution rate of the superior HAP2 haplotype was higher in NIR and NSER compared to YRR and YZRR (Fig. 6d). As a representative, a nucleotide diversity analysis of NIR was conducted. Nucleotide diversity analysis of chromosome D01 across three breeding periods showed that the nucleotide diversity of GhGeBP9 was higher during the 1950–1970 s than in other periods, and its diversity has not changed significantly from the 1980s to the present, indicates that the compartment where GhGeBP9 is located has been domesticated (Fig. 6e). Public transcriptome data showed that the GhGeBP9 expression level in the standard upland cotton line TM-1 was slightly higher than in the early-maturing variety CRI50 at 0–5 DPS. After 5 DPS, the expression level of GhGeBP9 in TM-1 gradually decreased and became lower than that in CRI50 (Fig. 6f). Moreover, comparing the expression levels of GhGeBP9 in TM-1 and CRI50 at different developmental stages revealed considerably higher expression in the early-maturing variety (Fig. 6g). This indicates that GhGeBP9 is central to influence the maturity of upland cotton. The qRT-PCR further validated higher expression levels in HAP2 accessions compared with HAP1, aligning with observed phenotypic differences (Fig. 6h).

Fig. 6.

Association analysis of characteristics of upland cotton that correlate with early maturity. a Manhattan plot of seven characteristics of upland cotton that correlate with early maturity; b Gene structure and two haplotypes of GhGeBP9; c Box plots of seven traits among different haplotypes; d Differences in haplotype distributions of GhGeBP9 among the four cotton planting regions; e Nucleotide diversity of GhGeBP9 on chromosome D01 over various breeding periods in NIR; f Expression pattern of GhGeBP9 in two cotton varieties, TM-1 and CRI50, from 0 to 20 days post-squaring (DPS); g Comparative expression analysis of GhGeBP9 in TM-1 and CRI50 across different leaf developmental stages; h Expression level analysis of GhGeBP9 between ‘Xinluzhong 34’ which carried HAP1, and ‘Han 9609’ which carried HAP2. (** P < 0.01, * P < 0.05)

Disscussion

Our extensive analysis of the GhGeBP gene family confirmed two key genes, GhGeBP4 and GhGeBP9, associated with significant traits such as fiber quality and early maturity. Here, we discuss the roles of these genes in cotton breeding. GhGeBP4 has been identified as a significant contributor to fiber quality traits and classified into two haplotypes. HAP2, one of which, exhibits superior fiber quality characteristics like higher FL, FS, and FE. This advantageous haplotype is prevalent in the Northwest Inland Cotton Region, aligning with the high-quality cotton produced in Xinjiang. The expression pattern of GhGeBP4 at 15 DPA aligns with the stages of fiber elongation and secondary wall thickening, which are crucial for cotton fiber quality [50, 51]. Comparisons of expression data between long- and short-fiber cotton materials, as well as between wild cotton and cultivated varieties, corroborated this finding. The presence of cis-acting regulatory elements like light-responsive (G-Box, Box4) and hormone-responsive motifs (P-box, GARE-motif, ABRE, CGTCA) in the GhGeBP4 promoter indicates its involvement in phytohormone signaling pathways. This is in line with the reported role of GeBP family genes in managing phytohormone pathways like gibberellins and cytokinins [9, 52]. These findings suggest that GhGeBP4 may regulate fiber development through light and hormone signaling.

GhGeBP9 was identified as a significant contributor to early maturity traits in upland cotton. SNPs within GhGeBP9 showed significant associations with the NFFB, FT. Haplotype analysis revealed two haplotypes, HAP1 and HAP2, with HAP2 exhibiting superior early maturity traits. The geographical distribution of HAP2 aligns with the breeding goals in these regions, where early maturity is desirable [53]. GhGeBP9 expression in CRI50 and TM-1 indicates that GhGeBP9 is crucial to determine maturity timing. Identification of GhGeBP4 and GhGeBP9 as candidate genes for fiber quality and early maturity traits provides valuable targets for applying marker-assisted selection (MAS) in cotton breeding programs. GhGeBP4 is a critical regulator of fiber development, while GhGeBP9 influences early maturity traits. Both genes exhibit specific haplotypes associated with desirable phenotypes, offering potential for developing high-quality and early maturity cotton varieties. Further functional validation of these genes through CRISPR/Cas9 or RNA interference (RNAi) will clarify their precise roles in cotton physiology.

Gene family evolution is a crucial driver of plant genome diversity and adaptation. Gene family expansion, involving gene duplication and subsequent divergence, has led to the proliferation of transcription factor families such as WRKY, ARF, and MYB [54–57]. However, in some cases, gene families may also undergo contraction. A reduction in the number of GhGeBP genes was observed in our study, supported by evolutionary analysis using the ratio of Ka to Ks substitutions (Table S4). A ratio of Ka to Ks < 1 suggests purifying selection, indicating that most mutations are deleterious and that the gene is under strong functional constraint [58]. Despite the contraction of the GhGeBP gene family, key members like GhGeBP4 and GhGeBP9 still regulate important traits in upland cotton. These key genes suggest that even with a reduced number of members, the GeBP family retains its importance in cotton biology.

Authors’ contributions

LL, JW and RL wrote the original draft manuscript text, ZF, LL, and SY revised  the main manuscript text, SZ, MY and YZ carried out data collation, JW, YX and FL completed data analysis, RL and NS performed data visualization. All authors reviewed the manuscript.

Funding

This research was sponsored by National Natural Science Foundation of China (32401297), Cotton Bio-breeding and Integrated Utilization Open Fund (CB2023A09) and Zhejiang A & F Student Research Training Program (2023KX007).

Data availability

The genomic resequencing data for GWAS analysis are available in the NCBI Sequence Read Archive under the accession PRJNA389777.Additional data are provided within the manuscript and supplementary information files.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.