Background

Upland cotton with higher yield properties attributed to a significant proportion of cotton production worldwide to fulfill the ever-increasing demands of the textile industry. Although upland cotton yields higher than other cotton species, the bottleneck in further improving fiber quality remains the fundamental concern [1]. Therefore, improving the fiber quality of cotton, specifically Upland cotton, is one of the current crucial research dimensions in cotton breeding programs. Fiber length is an important indicator in assessing fiber quality [2]. Therefore, the exploitation of fiber developmental mechanisms in different cotton species is fundamental to improve cotton cultivars. Fiber development mechanisms comprise four distinct yet overlapping phases viz., fiber initiation, fiber elongation, synthesis of the secondary cell wall, and fiber maturation [3, 4]. Cotton fibers, generally described as single-celled trichomes, are an excellent source material for studying single-cell elongation [5, 6]. Many studies showed that phytohormones, i.e., IAA, GA3, BR, etc., promoted cell and fiber elongation in cotton fiber development [7,8,9]. Previous research extensively emphasized that the Gibberelic Acid Stimulated Transcript/Arabidopsis (GAST/GASA) family genes were regulated by phytohormones to promote cell elongation and cell division in many plants [10].

GAST-like genes have been involved in complicated biological processes modifying further plant growth and development [11]. GASAs domains are remarkably conserved in most plants, resulting in conserved proteins [10, 12]. These proteins are generally divided into three well-conserved subgroups; however, OsGASTs were divided into four subgroups in rice [13]. The encoded protein of GAST family members usually contains three sections: the N-terminal signal peptide sequence, the hydrophilic, and C-terminal regions as GASA domain. GASA domains generally comprise 60 amino acids, including 12 cysteine, key residues for the functional domain [14]. The exact subcellular localization of GASTs proteins is the sticking point to determine the protein function. Many in vivo studies demonstrated that GASTs proteins were discovered in the cell wall and/or apoplast. However, a few GASTs proteins were localized in the plasma membrane, cytoplasm, and nucleus. The above-mentioned reports imply the divergent functional trends of GASTs, which are likely corresponding to functional association with cell elongation and division [15]. The presence of conserved motifs is the main reason for GASTs to be conserved in vascular plants. Further, the phylogenetic studies of GASTs family genes from the plant kingdoms showed that GASTs might be initially evolved in S. moellendorffii [11]. GAST1, which is one of GAST family genes, initially detected in tomato [16] with subsequent discoveries in different species, such maize (Zea mays L.) [10], Arabidopsis [11], Glycine max [17], Grapevine (Vitis vinifera L.) [12].

Previously published reports emphasized the regulatory role of GAST family genes in cell elongation and cell division and involved in responses towards abiotic and biotic stresses [18,19,20]. AtGASA family members have been reported with their regulatory role in hormone syntheses such as ABA, GA, BR, IAA, JA, and SA in Arabidopsis [15], while the OsGASR9 regulated grain length, width, and thickness in rice [21]. Interestingly, two ortholog from Arabidopsis (AtGASA10) and rice (OsGAST9) showed conserved (Functional and structural) positive regulatory role in germination between dicot and monocot plants. Phytohormones may regulate both AtGASA10 and OsGASR9 through the signal element in their promoter regions and correspond to the signal element in likely feedback response cycles in GA/ABA-mediated regulation [11]. GAs are among the universal plant phytohormones that play a crucial role in the growth and development processes, including germination, stem elongation, photosynthesis, flowering, and seed development, while GASA-like genes possess a key role in GAs signaling pathway [12, 22, 23] in different plants viz. gerbera [24], and maize [10]. Some GASA genes are regulated by other hormones, as RSI1 was regulated by IAA [25], and OsGSR1/2 was regulated by GA3 [26].

Reported statistics suggested that different GASA-like genes may exhibit differential expression patterns, mostly built on the spatiotemporal pattern of gene expression for regulating their speculated functions [27]. In A. thaliana, AtGASA5 is highly expressed in the shoot tip and the inflorescence meristems of the reproductive stage [27]. Interestingly, AtGASA6 is the key node in GA, ABA, and glucose signal interaction network, which regulates GA, ABA, and glucose in seed germination mesocotyl elongation [28]. The expression analysis in Moso bamboo revealed that most PhGASTs might be related to M. bamboo flower development and shoot growth. Importantly, PhGASR1 was presumed to play a key role in the rapid shoot growth involved in the ABA pathway [29]. However, the expression of AtGASA10 was not significantly influenced by exogenous GA treatment of suspension cells in Arabidopsis [30].

However, the factors that the mechanism of exogenous plant phytohormones IAA and GA3 promoted in cotton fiber cells elongation remain unknown. Given the important roles of GASA family proteins by exogenous plant phytohormones IAA and GA3 promoted in plant development and cell elongations, the analysis of GASA family is greatly valued. Our study systematically aimed to uncover the transcriptomic landscape of GASA genes in fiber development of G. arboretum, G. barbadense, G. hirsutum, specifically AtGASA10, and its ortholog genes, by utilizing RNA-Seq and qRT-PCR expression profiles. Furthermore, we exploited specific features of GASA gene family, including gene structures, conserved motifs, tissues-specific expression, subcellular localization, expression patterns, overexpression of Arabidopsis, and in vitro ovule culture. Using IAA, GA3, and their transport/biosynthesis inhibitors, we demonstrated that GhGASA10 plays a vital role promoted cell elongations in the overexpression of Arabidopsis and cotton fibers.

Results

Identification, phylogenetic analysis of the GASA gene family in the plant kingdoms

According to BLASTP and Query Sequence Search of TBtools, the GASA protein sequences from different plants were collected. To study the evolutionary relationships of GASA genes, the protein sequences, including Arabidopsis, G. darwinii, G. mustelinum, G. tomentosum, G. raimondii, G. arboretum, G. barbadense, and G. hirsutum, were exploited using phylogenetic analysis (Fig. 1A). GASA proteins depicted conserved phylogenetic relationships between Arabidopsis and Gossypium. Based on phylogeny, we further classified GASA proteins into three subfamilies viz. GASA1/2/3/9/11/14, GASA4/5/6/12/13, and GASA7/8/8 L/10. The ortholog genes corresponding to each AtGASA gene could be found in different Gossypium species. Most GASA genes did not have homologous in diploid cotton compared to Arabidopsis such as GASA1/2/3/4/5/6/9/14/12, while GASA7/8/8 L/10/11/13 were identified with homologies. Allotetraploid cotton should have twice the number of diploid cotton genes, but the numbers of GASA genes were less than twice. This result showed that some GASA genes were lost during the evolution process, which is in line with the previously published statistics demonstrating higher gene losses in allotetraploid cotton than diploid cotton [31]. Subsequently, GASA genes were explored in 20 species, extending from lower plants to higher plants, to make certain the origin and evolutionary relationship of these genes (Fig. 1B). Based on the genes number analysis, GASA genes were present in lower fern plants Selaginella moellendorffii, but not in the lower algae plants Micromonas pusilla, Ostreococcus Tauri, and Volvox carteri, and moss Physcomitrella patens, indicating that the GASA genes might have originated in fern. From the origin of ferns to angiosperms, the number of GASA genes has hardly changed in diploid plants. This result indicated that the number of GASA family genes is conserved in most plants.

Fig. 1
figure 1

Phylogenetic and evolutionary analysis of the GASA gene family in different plant species. (A) An unrooted phylogenetic tree of GASA protein sequences from Arabidopsis thaliana and seven Gossypium specie. The phylogenetic tree was constructed using GASA protein sequences and the Neighbor-Joining (NJ) method in MEGA 6.0 software. (B) Comparisons of GASA protein number across a wide range of plant species. Mp, Micromonas pusilla; Ot, Ostreococcus tauri; Vc, Volvox carteri; Pp, Physcomitrella patens; Sm, Selaginella moellemdorffii; Atr, Amborella trichopoda; Ps, Picea sitchensis; Pd, Phoenix dactylifera; Os, Oryza sativa Japonica; Vv, Vitis vinifera; Nc, Nymphaea colorata; Zm, Zostera marina; At, A. thaliana; Gr, G. raimondii; Ga, G. arboretum; Gd, G. darwinii; Gt G. tomentosum; Gm, G. mustelinum; Gb, G. barbadense; Gh, G. hirsutum

Structural characterizations and motifs analyses of GhGASA genes

A phylogenetic tree was constructed utilizing GhGASA protein sequences to exploit the evolutionary relationships of GhGASA family genes along with their structure and function. In general, GhGASA genes possessed one to three exons. Furthermore, gene structure and phylogenetic relationship displayed a high correlation. In total, 20 conserved motifs were identified in the GhGASA protein sequence (Fig. S1). The number of conserved motifs in each GhGASA varied from 3 to 11. Most GASA proteins contain the conserved Motif 1–3, showing that the three Motif of GASA proteins may have important role for the cell elongation and cell division.

To study the evolutionary relationships and functional divergence of the prominent GASA gene-family members, we extracted and examined the upstream 2.0 kb promoter regions. Many cis-acting regulatory elements, including 13 elements related to plant growth (including, Photo-responsive, cell cycle, and seed-specific regulation) and stress responses (including hormone-response, wound-response, and defense response towards stresses), were analyzed (Fig. 2B).

Fig. 2
figure 2

Promoter elements analysis of the GASA genes family in upland cotton. (A) The six phytohormones elements in the upstream 2.0 kb promoter regions; (B) The number of cis-acting regulatory elements of GASA genes family promoter sequences in upland cotton, and the sequences analysis were used by online software PlantCare

We mainly focused on cis-acting regulatory elements to verify gene functions concerning cotton fiber development. GhGASA genes promoter active elements and cotton transcriptomic data revealed that fiber development might not be linked to the cell cycle regulation and seed-specific regulation; it might be linked to hormone-responsive elements such as IAA, GAs (Fig. 2A). Interestingly, we found that the Gh.A04G144000 and Gh.GASA10–1(Gh.D04G182700) genes have gibberellin-responsive and abscisic acid responsiveness in similar sites of their promoters, while they have auxin-responsive in different sites of their promoters near to genes of 5′-UTR. However, other GhGASA genes promoters had different with the promoters of Gh.A04G144000 and Gh.GASA10–1. It might be suggested that IAA and GAs play essential roles in fiber development and cell elongation.

RNA-seq expression profile of GASA genes in three major cotton species

Expression profile and tissue specificity were explored using transcriptomic data of G. arboretum, G. barbadense, G. hirsutum. Firstly, we constructed a phylogenetic tree of GASA family genes of three major cotton species using the relative expression profiles with TBtools. The majority of GASA family genes from the same subfamily had similar expression patterns in six varieties in three major cotton species (Fig. 3, Fig. S2, and Fig. S3).

Fig. 3
figure 3

Expression profile of GhGASA family genes based on RNA-seq data in two uplant accessions viz., J02 and ZRI. The tissues include root, stem, leaf, and different stages (0, 3, 5, 10, 15, 20, 25 DPA)

In G. hirsutum (Fig. 3), only four genes are highly expressed at fiber development stages. Gh.A09G018000 and Gh.D04G053600 depicted relatively high expression at the early stages of fiber development. However, the Gh.A04G144000 and Gh.D04G182700 were highly expressed throughout whole fiber development stages, especially the critical period of fiber elongation for 5–15 DPA.

In G. arboretum (Fig. S2), the study found that the Ga14G0224.1 showed a significantly higher expression level in all the tissues. While Ga07G1350.1 was only highly expressed in leaves, and Ga04G0326.1 was only highly expressed in different fiber development stages, especially during the critical period of fiber elongation.

In G. barbadense (Fig. S3), only three genes were highly expressed at crucial fiber development stages, especially two genes viz. Gbar.D04G017490 and Gbar.A04G012790 showed significant expression levels at the critical period for fiber elongation (10-DPA).

Interestingly, the five genes Ga04G0326.1, Gbar.D04G017490, Gbar.A04G012790, Gh.A04G144000, and Gh.D04G182700 were all ortholog of AtGASA10 in three cotton species. Regardless of the evolutionary relationship, gene structure and expression changes were consistent among the six varieties of G. arboretum, G. barbadense, G. hirsutum. These results emphasized that higher expression and tissue specificity of these genes viz. Ga04G0326.1, Gbar.D04G017490, Gbar.A04G012790, Gh.A04G144000, and Gh.D04G182700 might play a direct critical role in fiber development and fiber cell elongation.

In this study, we identified five GASAs genes sited at A04/D04. They were explicitly expressed at critical fiber development stages in three major cotton species, which emphasized that the GASA10 might have a significant role in fiber development, specifically fiber cell elongation. We further performed functional verification to understand the genetic basis, characteristics, and functions of GhGASA10–1 (Gh.D04G182700).

Subcellular localization of GhGASA10–1

According to the online tool analysis, TMHMM2.0 (http://www.cbs.dtu.dk/services/TMHMM/) predicted that GhGASA10–1 has the 26 N-term signal sequence with transmembrane and the other sequence outside the membrane. The CELLO version 2 [32] and Euk-mPLoc 2.0 [33] predicted that the subcellular localization of GhGASA10–1 is extracellular. YLoc [34] and BaCelLo [35] indicated that the localization of GhGASA10–1 is a secreted pathway.

To verify this prediction, the full-length CDS of GhGASA10–1 was ligated with 35S-1300-GFP vector. The constructed vector was infiltrated into N. benthamiana mature leaves and visualized by confocal microscopy. The fluorescence of 35S-GFP was detected in the nucleus and the cytomembrane (Fig. 4). In contrast, the GhGASA10–1::GFP fusion protein was localized in the cell membrane, appearing green, and the cell membrane presented red fluorescence stained by cell membrane marker Dil. Subsequently, the cell membrane was merged into yellow by GhGASA10–1::GFP fusion protein and cell membrane marker Dil. The above results demonstrated that GhGASA10–1 was localized in the cell membrane, which may synthesize secreted protein transport to the cell wall involved in cell wall synthesis and promote cotton fibers cell wall development through the secreted pathway.

Fig. 4
figure 4

Subcellular localization of GhGASA10–1 protein in the epidermal cells of tobacco leaf. GFP was a green fluorescent protein and as a blank control; GFP-GhGASA10–1 protein was only located in the cell membrane; Dil was the most commonly used cell membrane fluorescent probe; Merged was GFP-GhGASA10–1 and Dil collective effect. Bar = 20 μm

Overexpression and tissues specificity analysis of GhGASA10–1 in Arabidopsis

To further confirm the gene function, GhGASA10–1 was overexpressed in Arabidopsis. Among ten lines GhGASA10–1-overexpressing transgenic Arabidopsis of T3 generation, three lines were selected for further analysis. Tissue specificity expression analysis (Fig. 5A) showed that GhGASA10–1 is explicitly expressed in the roots at the seedling stage. However, GhGASA10–1 is significantly down-regulated in the roots and specifically expressed in the flower buds at the flowering stages.

Fig. 5
figure 5

Characterization of GhGASA10–1 overexpression in Arabidopsis. (A) Different expression characteristics of different organs. The data are shown in (A) as mean ± SD, n = 3, **P < 0.01, ***P < 0.001, Student’s t-test. ACTIN2/8 was used as an internal control. (B/C) Verification and statistics of the germination experiment of wild-type and transgenic Arabidopsis. (D/E) Comparing and statistics of root lengths wild-type and transgenic Arabidopsis. The data are shown in (E) as mean ± SD, n = 6, **P < 0.01, ***P < 0.001, Student’s t-test. (F/G) Comparing wild-type and transgenic root lengths at the cellular level. The data are shown in (G) as mean ± SD, n = 60, **P < 0.01, ***P < 0.001, Student’s t-test. Bars in (B), (E) =1 cm. Bars in (F) =100 μm

When grown on 1/2 MS medium, wild-type and three transgenic lines exhibited a noticeable phenotypic difference in Arabidopsis seedling germination stages. The seedling germinated after 14 days of standard cultivation (Fig. 5B), while GhGASA10–1-overexpressed seedlings germinated vigorously than wild-type seedlings. Statistics on seeds germination rate showed that transgenic seeds germination rate was significantly higher than the wild type, especially on the third day (Fig. 5C). The length of the taproot after 14 days of vertical cultivation (Fig. 5D), GhGASA10–1-overexpressed seedlings formed more main roots than wild-type seedlings. The seed from wild-type and transgenic plants, viz., OE1, OE2, and OE3, were germinated on MS-agar medium. The results comprehended the transgene, as shown in Fig. 5D; the root growth was observed with significantly higher expression in the transgenic lines than in the wild-type plants. Biological statistics (Fig. 5E) showed that the length of the root of transgenic lines was twice the wild-type of Arabidopsis seedling stages. These results indicated that GhGASA10–1 promotes seedling germination and root extension in Arabidopsis.

As mentioned above, GhGASA10–1 was screened as a putative candidate gene for fiber cell elongation. However, the tissue specificity of this gene in Arabidopsis suggested marked changes in root elongation. To validate our hypothesis that GhGASA10–1 play a crucial role in cell elongation, we further examined the roots of Arabidopsis and compared both the transgenic and wild type. Interestingly, we observed that overexpression of GhGASA10–1 extensively promotes root length with the elongation of root cells instead of an increase in the number of cells (Fig. 5F/G). These results strengthen our hypothesis that GhGASA10–1 may be important in fiber cell elongation.

GhGASA10–1 expression level associates with cellulose synthesis

As the over-expression of GhGASA10–1 in Arabidopsis promote cell elongation resulting in root elongations, we speculated that GhGASA10–1 might promote downstream transcription factors, leading to high expression of cellulose synthase genes, further promoting cell elongation. Comparing the expression of cellulose synthase genes (AtCesAs) (Fig.6) in wild-type and overexpressing Arabidopsis, five members of the AtCesAs family were found to be upregulated. AtCesA5b/9 were upregulated twice, while AtCesA4/7 were upregulated three times, and AtCesA10 was upregulated more than ten times. Taken together, this data provides strong evidence that over-expression of GhGASA10–1 strongly induced particular cellulose synthesis associated genes, and further to promote cell elongation.

Fig. 6
figure 6

Over-expression of GhGASA10–1 promotes cellulose synthase genes in Arabidopsis. Over-expression of GhGASA10–1 promotes cellulose synthase genes to regulate root elongations in Arabidopsis. The data are shown as mean ± SD, n = 3, **P < 0.01, ***P < 0.001, Student’s t-test. ACTIN2/8 was used as an internal control

GhGASA10–1 induced by IAA but not GA3

It has been shown that GASA family genes are involved in the regulation of phytohormones in different plants and act as binding promoter elements in upland cotton. IAA or GA3 might regulate GhGASA10–1 and GhA04G144000, so we take hormone-treated ovule in vitro culture to verify the expression level GhGASA10–1. Surprisingly, qRT-PCR (Fig.7A) results showed that GhGASA10–1 was significantly upregulated during crucial fiber elongation stages. Moreover, GhGASA10–1 was regulated by IAA, but not GA3 (Fig.S4) during cotton fiber development when treated with different concentrations of hormone IAA, GA3, and their inhibitors in the cotton ovule. In addition, we further verified its ortholog gene GhA04G144000, which was slightly up by IAA (Fig.7B). These results revealed that GhGASA10–1 may be induced by IAA to promote cell elongation.

Fig. 7
figure 7

IAA induces the expression of GhGASA10–1 to promote cotton fiber elongations. Different concentrations of (1 μM, 5 μM, 10 μM) IAA and their inhibitor 10 μmol NPA treatment ovules on the day of flowering, respectively. Relative expression of GhGASA10–1 in different periods (5 d, 10 d, 15 d, 20 d) of G. hirsutism. The data are shown as mean ± SD, n = 60, **P < 0.01, ***P < 0.001, Student’s t-test. UBQ was used as an internal control

Discussions

The whole-genome sequences (WGS) of different cotton species have been accomplished in recent years, which has boosted the research on genetic breeding and functional gene discovery of cotton [36,37,38]. Moreover, cotton fibers are single-celled trichomes and are excellent model materials for studying single-celled elongations [39]. Some studies showed that phytohormones IAA, GA3, BR etc., promoted cell elongation in cotton fiber development [40] and GASA family genes are regulated by different phytohormones to promote or inhibit cell elongation and cell division as well as other function in many plants [12]. Therefore, it is imperative to exploit and understand the functional mechanism of some crucial genes of the GASA family regulating fiber cell elongations for different fiber development stages in cotton.

In this study, we identified GASA genes in seven representative cotton species, including three wild allotetraploid cotton G. mustelinum, G. darwinii, G. tomentosum with 48, 45, 47 GASA genes respectively, two cultivated allotetraploid cotton G. hirsutum, G. barbadense with 45, 41 GASA genes respectively, and its two diploid ancestors, G. arboreum and G. raimondii with 24, 25 GASA genes respectively. This differential distribution of GASA family genes emphasized the loss of genes in different allotetraploid cotton species, consistent with the higher rate of gene loss in different allotetraploid cotton than in diploid species [15, 41]. The GASA proteins are quite conservative in higher plants, mainly divided into three subgroups in Arabidopsis and Gossypium, which is consistent with GASA family of Zea mays [10], Oryza sativa [11, 21], Vitis vinifera [12], and Glycine max [17]. GASA genes are present in lower fern plants S. moellendorffii [11]. In the life evolutionary tree, the number of GASA genes has hardly changed from ferns to angiosperms, firstly reported in our study. The above results showed that GASA family genes are relatively conservative in gene structure and quantity [10, 12].

The promoter region elements and expression profile of GhGASA family genes in cotton were analyzed, and fiber-specific expression of GASA10 members in different cotton species. The results emphasized the involvement of phytohormones viz. GA3 and IAA in promoting cell wall and fiber elongation. Most of the GASA genes are regulated by GA, ABA, SA [11], which are involved in the hormone signaling pathway in different plants, explicitly influencing many plant functions such as bud dormancy, bud germination, root length, stem elongation, seed size, and yield [12, 36, 42, 43]. Interestingly, GASA10–1 and Gh.A04G144000 have abscisic acid responsiveness, gibberellin-responsive, and auxin-responsive elements, while Gh.A04G144000 have two gibberellin-responsive elements, especially it are different to the sites of their promoter auxin-responsive elements.

According to the expression profile of GASA genes in three major cotton species, different GASA family genes showed tissue-specific expression such as GhGASA1/2 was highly expressed in leaves of both cultivars, whereas GhGASA10 showed high expression in the fiber and seed of both cultivars. GASA genes have tissue-specific expression, that CcGASA4, OsGASR1/9 genes have flower-tissue-specific expression [21, 23, 43]. Numerous studies have found that the function of GASA genes not only promotes cell elongation and other tissue development but also resists various abiotic stresses i.e., salinity, drought, cold, fungal, and paclobutrazol (PBZ) [29, 44, 45].

GhGASA10–1 was localized in the cell membrane, which may synthesize secreted protein transform into the cell wall involved in cell wall compound for fiber cell elongations. Citrus clementina CcGASA4 and Rice OsGASR9 localized to the plasma membrane and nucleus [23], while Pyrus pyrifolia PpyGAST1 localized to the cytoplasm, and AtGASA5 protein localized in the cell wall/extracellular matrix [46]. Potato Snakin-1(GASA-like), which follows the secretory pathway, regulated cell division, primary metabolism, and cell wall composition [47]. This study results showed that GhGASA10–1 might be consistent with the function of AtGASA10.

The overexpression of GhGASA10–1 in Arabidopsis was analyzed, which promoted seedling germination. Interestingly, the overexpression of GhGASA10–1 remarkably promoted the main root extension, and the cellular level of Arabidopsis roots length was found that GhGASA10–1 promotes Arabidopsis roots cell elongations, which further indirectly confirmed that GhGASA10–1 promote cell and fiber elongation of cotton. AtGASA10 is involved in changing the hydroxyl length facilitating cell wall growth by regulating cell elongation. The above results showed that the functions of GhGASA10–1 differ from AtGASA10 function and might have different regulation mechanisms, which is crucial for further characterization of the GhGASA10–1 gene for its involvement in fiber cell elongation [30].

Most previous studies have shown that changes in plant organs may be due to the regulation of CesAs genes. CesAs genes perform various functions of primary and secondary cell wall synthesis [48]. Our results suggested that GhGASA10–1 in Arabidopsis promotes root elongations, which leads to the hypothesis that GhGASA10–1 may promote downstream transcription factors, leading to high expression of cellulose synthase and further promoting cell elongation.

Our study found that the expression of GhGASA10–1 was not promoted by the exogenous phytohormones GA3, which is consistent with the expression of AtGASA10 being not regulated by GA3 [30]. OsGASR9 is involved in response to GA in rice [21]. Interestingly, GhGASA10–1 was upregulated by the exogenous phytohormones IAA; the result showed that IAA might play a crucial role in fiber cell elongation and development, which is further researched for cotton fiber quality main as fiber elongations.

Conclusions

In this study, the evolutionary relationships of GASA gene family were identified in different Gossypium species, which were classified with three distinct subclasses and quite conservative. GASA genes might have originated in lower fern plants Selaginella moellendorffii. From the origin of ferns to angiosperms, the number of GASA genes has hardly changed. The GhGASA10–1 was localized in the cell membrane, which may synthesize proteins in the cell wall involved in cell wall compound for fiber cell elongations. The overexpression of GhGASA10–1 promotes seedling germination in advance and promotes Arabidopsis roots cell elongations in Arabidopsis, further indirectly confirmed that GhGASA10–1 might promote cell elongation. GhGASA10–1 promotes AtCesA4/5b/7/9/10 in Arabidopsis, especially AtCesA10 being more remarkable, which revealed that GhGASA10–1 not only promotes primary wall synthesis but also primarily promotes secondary wall synthesis. GhGASA10–1 was upregulated by IAA, emphasizing that IAA may play a crucial role in fiber cell elongation and development. These results reveal the structural characteristics and expression patterns of the GhGASA gene family and functional verification of GhGASA10–1 in cell elongation and provide crucial information for further regulation mechanism/ network of fiber elongation.

Materials and methods

Database search and sequence retrieval

Cotton Functional Genomics Database (CottonFGD) (https://cottonfgd.org/) platform was employed to obtain the genomic datasets and protein sequences of different cotton species, i.e., G. arboreum L, G. raimondii Ulbr, G. hirsutum L, G. barbadense L. The protein sequences of three wild cotton such G. darwinii, G. mustelinum, and G. tomentosum were downloaded from NCBI (https://www.ncbi.nlm.nih.gov/) [49]. The protein sequences of Arabidopsis thaliana were downloaded from the Arabidopsis Information Resource (TAIR) (https://www.arabidopsis.org/). The other species’ protein sequences come from NCBI (https://www.ncbi.nlm.nih.gov/). Based on the sequence similarity of the protein domains, GASA protein sequences in different plant species were extracted using TBtools [50].

TBtools, with default parameters [50], was employed to search for various GASA protein sequences where the GASA domain (PF02704), obtained from pfam database (http://pfam.xfam.org./) was used as a query sequence. Repeated proteins were redundant, thus deleted, and only GASA protein sequences with e-value> 30 were kept and double-checked through NCBI-CDD (NCBI conserved domain database, https://www.ncbi.nlm.nih.gov/cdd) for further analysis.

Phylogenetics, gene structure, and motif analysis

Multiple sequence alignments of obtained GASA protein sequences, including Arabidopsis and Gossypium (as mentioned above), were done using Muscle wrapper in TBtools. Subsequently, IQ-TREE in TBtools was utilized to generate a phylogenetic tree with 1000 bootstraps [50]. A bar graph was made by the number of GASA family protein sequences of different cotton species.

Furthermore, The gene structures of GASAs were analyzed using the Gene Structure Shower of TBtools [50]. We also exploited motifs with conserved domains of GASA proteins using MEME (http://meme-suite.org/tools/meme) with default parameters. The GASA family genes characteristic was visualized and integrated into graphics using TBtools [50].

Analysis of cis-elements related to plant hormone

2.0 kb upstream sequences of GASA family genes in G. hirsutum and GASA10 genes from other four cotton species G. arboreum L, G. raimondii Ulbr, G. hirsutum L, G. barbadense, were extracted by TBtools, and the cis-elements were determined utilizing the PlantCARE database (http://bioinformatics.psb.ugent. be/webtools/plantcare/html/).

Plant materials and growth conditions

The three major cotton species of G. hirsutum cv. J02–508 and ZRI-015; G. arboreum cv. 971 and 972; G. barbadense cv. XINHAI133 and MAROAD were grown in Anyang cotton farm. Three different tissues and 0 DPA (day post-anthesis) ovules, 3, 5 DPA ovules, and 10, 15, 20, & 25 DPA fibers in G. hirsutum, 0, 3, 5 DPA ovules as well as 8 DPA fibers in G. arboreum, 0 DPA ovules as well as 15, 25 DPA fibers in G. barbadense were sent to Biomarker Technologies company for completing transcriptome sequencing.

Col-0, ecotype of Arabidopsis thaliana, seeds were put in 4 °C for vernalization and later grown on agar-solidified Murashige and Skoog (MS) medium, which were placed in an incubator with a 16 h / 8 h (light/dark) cycle at 22 °C. The seedlings were transplanted in mixed soil (vermiculite: humus = 1:1). The Agrobacterium tumefaciens strain (GV3101) with constructed overexpression vector was transformed into Arabidopsis plants using the floral dip method [51].

Expression profile of genes

The fragments per kilobase of exon per million fragments mapped (FPKM) values were obtained from the transcriptome data of G. hirsutum cv. J02–508 and ZRI-015; G. arboreum cv. 971 and 972; G. barbadense cv. XINHAI133 and MAROAD. The expression values of three different tissues and 0 DPA ovules, 3, 5 DPA ovules as well as 10, 15, 20, and 25 DPA fibers in G. hirsutum, 0, 3, 5 DPA ovules as well as 8 DPA fibers in G. arboreum, 0 DPA ovules as well as 15, 25 DPA fibers in G. barbadense, were analyzed. Genes with FPKM> 1 in at least one stage were selected for further analysis. log2FPKM, after normalization, was used to display the gene expression as a heatmap.

Cloning of GhGASA10–1, vector construction, and plant transformation

cDNA sequence of GhGASA10–1 was cloned and plugged into pBI121 vector using two restriction sites (Xba1 and Sac1). The primers were conceived by primer 5.0, and the sequences are presented in Table S1. Subsequently, we took the GV3101 strain containing constructed GhGASA10–1-pBI121 vector to transform it into Arabidopsis plants according to the floral dip method. Seedlings with transgene were selected carefully and transferred to mixed soil. The transgenic plants were grown in a greenhouse for further sample collection for PCR confirmation. Homozygous transgenic Arabidopsis lines were obtained, and the lines OE1, OE2, and OE3, which have high levels of GhGASA10–1 expression, were selected for further analysis. The wild-type Arabidopsis plants were used as the controls.

RNA extraction and quantitative PCR

The total RNA of cotton stem tips and leaves were isolated by the RNAprep Pure Plant kit (Tiangen, China). Approximately 1000 ng of RNA were reversely synthesized into cDNA by MonScript RTIII Super Mix with dsDNase (Two-Step) (Monad, China).

The real-time PCR detection system (RT-PCR) utilized ABI 7500 real-time PCR system and ChamQ Universal SYBR qPCR Master Mix (Vazyme, China). The UBQ gene which expresses stably in upland cotton, was used as an internal control. The relative expression levels of GASA genes were calculated by the 2-ΔΔCT method.

The cotton ubiquitin gene (Gh_A10G005800) and Arabidopsis β-actin genes (actin 2, actin 8) were used as internal references [52, 53]. To further explore OE-GhGASA10–1 in Arabidopsis, whether through regulation of cell wall synthesis cellulose synthase genes to promote main root elongation, and related 12 AtCesAs primer pairs [48]. The primers used in the quantitative PCR analysis are shown in Table S1. All qRT-PCR experiments were executed for at least three biological replicates.

Subcellular localization of GhGASA10–1

The amplified exonic region of ChGASA10–1, using specific primers (Table S1) corresponding Sma I and Kpn I restriction enzyme sites, was fused to the 5′ terminal of the GFP gene and consequently generated GhGASA10–1-GFP fusion construct comprising CaMV 35S promoter. The GhGASA10–1-GFP vector and positive control (empty vector) were transformed into the Agrobacterium tumefaciens strain (GV3101), then transformed into Nicotiana tabacum leaves [54]. Leaves of the seedings were stained with cell membrane CM-Dil (10 μM, Sigma-Aldrich) and visualized using a laser confocal microscope (Zeiss LSM710, Germany).

In vitro cotton ovule culture and hormone treatment

Randomly selected cotton bolls (TM-1) were collected and sterilized in 0.1% (w/v) HgCl2 and 75% (v/v) ethanol for 15 min and 5 min, respectively. Collected bolls were washed with sterilized distilled water after sterilization. Ovule samples were collected from air-dried bolls under sterile conditions. Collected ovules were then cultured in BT medium as a control treatment in a dark environment at 28–30 °C, as previously described by [55]. The ovules were also cultured for harmone treatment assay with different concentrations of GA3 (1 μM, 3 μM, 5 μM), GA biosynthesis inhibitor (PAC, 5 μM), IAA (1 μM, 5 μM, 10 μM), and IAA transport inhibitor (NPA, 5 μM), respectively.