Collection of QTL and SNPs associated with fiber quality traits
A total of 12 QTL mapping studies for cotton fiber quality traits were used in this study, in which the mapping population size ranged from 107 to 503 lines (Table 1) (Ali et al. 2018; Diouf et al. 2018; Huang et al. 2017; Jia et al. 2018; Keerio et al. 2018; Li et al. 2016a; Liu et al. 2018b; Ma et al. 2018a; Tan et al. 2018; Wang et al. 2015; Zhang et al. 2015b; Zou et al. 2018), and the number of SNP markers ranged from 168 to 19191 (Table 1). As a result, a total of 884 initial QTL related to cotton fiber quality traits were collected, which were unevenly distributed on each chromosome, and ranged from 12 to 57 (Fig. 1, Table S1). Chromosomes A4 had the lowest number of QTL and chromosome A10 had the highest number of QTL. Among them, there were a large number of QTL related to FL, FS, MIC, FU, and FE, which were 204, 207, 179, 118, and 108, respectively. However, there were small number of QTL related to the SCI, SF, FR, and FY, which were 21, 19, 13 and 15, respectively (Table S1).
Table 1
Fiber quality traits QTL mapped by SNP markers from 12 papers
QTL | Traits | Population type | Population size | Number of markers | Reference |
8 | FS | RIL | 250 | 168 SNP | [Zou et al. 2018] |
37 | FL, FS, MIC | RIL | 196 | 106 SSR & 104 SNP | [Zhang et al. 2015] |
9 | FS | RIL | 161 | 304 SSR & 5571 SNP | [ Wang et al. 2015] |
104 | FL, FS, MIC, FU, FE | RIL | 180 | 12116 SNP | [Tan et al. 2018] |
21 | FL, FE | BIL | 176 | 15369 SNP | [Ma et al. 2018] |
134 | FL, FS, MIC | RIL | 231 | 122 SSR& 4729 SNP | [Liu et al. 2018] |
30 | FL, FS, MIC, FU, FE | ILs | 107 | 3157 SNP | [ Keerio et al. 2018] |
186 | FL, FS, MIC, FU, FE | RIL | 137 | 139 SSR & 6295 SNP | [Jia et al. 2018] |
50 | FL, FS, MIC, FU, FE, SF | Natural population | 503 | 19 191 SNP | [Huang et al. 2017] |
193 | FL, FU, MIC, FS, FE, FR, FY, SCI | F2:3 population | 277 | 5178 SNP | [ Diouf et al. 2018] |
48 | FL, FS, MIC, FU, FE | RIL | 188 | 2618 SNP | [Li et al. 2016] |
59 | FL, FS, MIC, FU, FE | RIL | 180 | 6254 SNP | [ Ali et al. 2018 ] |
FL, fiber length; FS, fiber strength; MIC, fiber micronaire; FU, fiber uniformity; FE, fiber elongation; SCI, spinning consistency index; SF, short fiber; FR, fiber reflectance; FY, fiber yellowness. |
Meta-QTL of fiber quality traits
In this study, a meta-analysis was performed with 884 QTL related to cotton fiber quality traits, and a total of 74 stable meta-QTL related to FL, FS, FE, MIC, and FU were obtained, including 19 for FL, 18 for FS, 11 for FU, 11 for FE, and 15 for MIC, which covered 26 upland cotton chromosomes. There were 33 meta-QTL in the At sub-genome and 41 in the Dt sub-genome. The confident intervals (CI) of all meta-QTL were smaller than their respective initial QTL, which were ranged from 2.4 to13.4 Mb, with an average of 8.5 Mb (Fig. 2, Table S2). Among the 74 meta-QTL, 19 were obtained from multiple QTL coincident regions of 4 or more studies (Table S2), indicating that these regions had high correlation with cotton fiber quality traits.
Meta-QTL for FL
A total of 19 meta-QTL related to FL were obtained, covering 17 chromosomes, including 7 meta-QTL on At sub-genome (A01, A05, A08, A09, A10, and A12), and 12 Meta-QTL on Dt sub-genome (D02, D03, D04, D05, D06, D08, D09, D10, D11, D12, and D13) (Fig. 2, Table S2). Three meta-QTL have been mapped in more than 5 studies, namely meta-QTL-3, meta-QTL-15, and meta-QTL-17, which were located in the 18.8–31.70 Mb region on A05 chromosome, the 54.17–61.86 Mb region on D10 chromosome, and the 15.70-24.32 Mb region on D11 chromosome respectively (Table S2).
Meta-QTL for FS
A total of 18 meta-QTL related to FS were obtained, covering 15 chromosomes, including 9 meta-QTL on At sub-genome (A01, A02, A03, A05, A07, A09, and A10), and 9 meta-QTL on Dt sub-genome (D04, D05, D06, D07, D08, D10, D11, and D12) (Fig. 2, Table S2). Among them, four meta-QTL have been identified in the four studies, namely meta-QTL-23, meta-QTL-25, meta-QTL-27, and meta-QTL-30, which were located in 5.81–14.90 Mb on A05, 63.33–73.57 Mb on A07, 93.51–100.20 Mb on A10, and 4.01–15.81 Mb region on D05 respectively (Table S2).
Meta-QTL for FU
A total of 11 meta-QTL related to FU were obtained, covering 11 chromosomes, including 5 meta-QTL on At sub-genome (A02, A06, A09, A10, and A11), and 6 meta-QTL on Dt sub-genome (D01, D02, D03, D05, D11, and D12). Among the meta-QTL, meta-QTL-45 on D03 and meta-QTL-47 on D11 chromosomes are more reliable for identification in four studies (Fig. 2, Table S2).
Meta-QTL for FE
A total of 11 meta-QTL related to FE were obtained, covering 11 chromosomes, including 4 meta-QTL on At sub-genome (A05, A10, A11, and A13), and 7 meta-QTL on Dt sub-genome (D01, D04, D07, D11, and D12) (Fig. 2, Table S2). Among these meta-QTL, meta-QTL-53 on chromosome D01 was identified in four studies.
Meta-QTL for MIC
A total of 15 meta-QTL related to MIC were obtained, covering 14 chromosomes, including 8 meta-QTL on At sub-genome (A05, A06, A08, A09, A10, A11, and A13), and 7 meta-QTL on Dt sub-genome (D03, D05, D06, D08, D09, D11, and D12) (Fig. 2, Table S2). Two meta-QTL have been identified in the four studies, namely meta-QTL-61 and meta-QTL-71, which were located in 17.32–26.9 Mb on A05, and 52.98–62.38 Mb on D08 respectively (Table S2).
Candidate genes identification combined and meta-QTL intervals and significant SNPs
8589 significant SNP loci associated to cotton fiber quality traits were collected from 15 GWAS studies and mapped to TM-1 genome (Chandnani et al. 2018; Fang et al. 2017; Gapare et al. 2017; Handi et al. 2017; Huang et al. 2017; Islam et al. 2016; Li et al. 2018; Li et al. 2017b; Liu et al. 2018b; Ma et al. 2018a; Ma et al. 2018b; Su et al. 2016; Su et al. 2018; Sun et al. 2017; Wen et al. 2018) (Table S3, Table S4, Fig. 2), 4343 of which were mapped in the 74 meta-QTL regions (Table S5). 297 candidate genes were identified closely linked to the 4343 SNPs, including 126 genes for FL, 93 for FS, 40 for FU, 20 for FE, 18 for MIC (Table S6).
GO and KEGG enrichment analysis of candidate genes
To identify common characteristics of these genes in biological functions, gene ontology (GO) analysis were performed with the 297 candidate genes, and 200 of them had ontology annotation, which were classified into the three main GO categories (biological process, molecular function, and cellular component) and 15 GO terms (Fig. 3; Table S7). In the biological process category, protein modification process (30, 15%), cellular protein modification process (30, 15%), protein metabolic process (46, 23%), macromolecule modification (30, 15%), macromolecule metabolic process (69, 34.5%), cellular protein metabolic process (37, 18.5%), cellular macromolecule metabolic process (60, 30%), and proteolysis (11, 5.5%) were the major subcategories (Fig. 3, Table S7). In the cellular component category, 35 (17.5%) genes were enriched in the membrane subcategory (Fig. 3; Table S7). In the molecular function category, protein serine/threonine kinase activity (18, 9%), protein binding (55, 27.5%), peptidase activity (10, 5%), protein tyrosine kinase activity (20, 10%), DNA binding (22, 11%), and transporter activity (16, 8%) were the major subcategories (Fig. 3, Table S7).
To further comprehend the enriched pathways of the candidate genes, KEGG pathway analysis was performed, and 234 annotated genes were assigned to 4 KEGG pathways (P < 0.05), including pentose and glucuronate interconversions, acarbose and validamycin biosynthesis, vitamin digestion and absorption, and membrane trafficking (Table 2). Some of the pathways have been reported to be associated with fiber development, such as, pentose and glucuronate interconversions pathway is associated with fiber elongation.
Table 2
KEGG analysis of candidate genes
Accession | Name | Input gene number | p-value |
ko00040 | Pentose and glucuronate interconversions | 2 (1.89%) | 0.01 |
ko00525 | Acarbose and validamycin biosynthesis | 1 (0.94%) | 0.02 |
ko04977 | Vitamin digestion and absorption | 1 (0.94%) | 0.02 |
ko04131 | Membrane trafficking | 10 (9.43%) | 0.04 |
Key candidate genes identified from expression patterns
To better understand the molecular function of the candidate genes, the expression in 10 tissues (root, stem, leaf, petal, anther, stigma, fibers at four developmental stages) obtained from the transcriptome datasets of the upland cotton genetic standard TM-1 were used for spatiotemporal expression analysis (Zhang et al. 2015a). Among the 126 candidate genes for FL, nine of which (Gh_D08G1950, Gh_D06G0479, Gh_D11G1626, Gh_D13G1900, Gh_D10G0833, Gh_D13G1965, Gh_A09G1231, Gh_D08G1970, and Gh_D04G1574) showed high expression specifically during the development of cotton fiber. Gh_D08G1950, Gh_D06G0479, Gh_D11G1626, Gh_D13G1900, Gh_D10G0833, and Gh_D13G1965 showed high expression specifically at fiber secondary wall biosynthesis stages (20 and 25 DPA); Gh_A09G1231, Gh_D08G1970, and Gh_D04G1574 showed high expression specifically at fiber elongation stages (5 and 10 DPA). Among the 93 candidate genes for FS, 5 of which (Gh_A01G1474, Gh_A05G2203, Gh_D08G2110, Gh_A10G2036, and Gh_A07G1801) showed high expression specifically during the development of cotton fiber. Gh_A01G1474, Gh_A05G2203 and Gh_D08G2110 showed specifically high expression at fiber secondary wall biosynthesis stages (20 and 25 DPA); Gh_A10G2036, and Gh_A07G1801 showed high expression specifically at fiber elongation stages (5 and 10 DPA). Among the 40 candidate genes for FU, 2 of which (Gh_A11G2663 and Gh_D11G2059) showed high expression specifically during the development of cotton fiber. Gh_A11G2663 showed high expression specifically at fiber secondary wall biosynthesis stages (20 and 25 DPA); Gh_D11G2059 showed high expression specifically at fiber elongation stages (5 and 10 DPA). Among the 20 candidate genes for FE, 3 of which (Gh_A13G0282, Gh_A13G0354 and Gh_A11G1313) showed high expression specifically at fiber elongation stages (5 and 10 DPA). Among the 20 candidate genes for FE, Gh_D11G1416 showed high expression specifically at fiber secondary wall biosynthesis stages (20 and 25 DPA) (Table S8).
To better understand the molecular function of the key candidate genes identified from expression patterns, they were annotated in CottonFGD (https://cottonfgd.org/). As a result, 4 genes (Gh_D10G0833, Gh_A05G2203, Gh_D11G2059, and Gh_A13G0354) have no protein annotation, and others encode a variety of proteins (Table 3). Gh_D08G1950 encodes a probable copper-transporting ATPase HMA5; Gh_D06G0479 encodes a basic endochitinase; Gh_D11G1626 encodes a COBRA-like protein 4; Gh_D13G1900 encodes a WAT1-related protein; Gh_D13G1965 encodes a protein WVD2-like 1; Gh_A09G1231 encodes a 40S ribosomal protein S28; Gh_D08G1970 encodes a probable aquaporin PIP1-2; Gh_D04G1574 encodes a Snakin-1; Gh_A01G1474 encodes WAT1-related protein; Gh_D08G2110 encodes a CASP-like protein 5A2; Gh_A10G2036 encodes a Rop guanine nucleotide exchange factor 5; Gh_A07G1801 encodes a peptidyl-prolyl cis-trans isomerase FKBP15-1; Gh_A11G2663 encodes a protein WVD2-like 1; Gh_A13G0282 encodes a xylulose kinase; Gh_A11G1313 encodes a EPIDERMAL PATTERNING FACTOR-like protein 9; Gh_D11G1416 encodes a transcriptional corepressor LEUNIG_HOMOLOG (Table 3).
Table 3
The annotation of key candidate genes
Traits | Gene ID | Annotation |
FL | Gh_D08G1950 | Probable copper-transporting ATPase HMA5 |
FL | Gh_D06G0479 | Basic endochitinase |
FL | Gh_D11G1626 | COBRA-like protein 4 |
FL | Gh_D13G1900 | WAT1-related protein At1g09380 |
FL | Gh_D10G0833 | Not available |
FL | Gh_D13G1965 | Protein WVD2-like 1 |
FL | Gh_A09G1231 | 40S ribosomal protein S28 |
FL | Gh_D08G1970 | Probable aquaporin PIP1-2 |
FL | Gh_D04G1574 | Snakin-1 |
FS | Gh_A01G1474 | WAT1-related protein At1g25270 |
FS | Gh_A05G2203 | Not available |
FS | Gh_D08G2110 | CASP-like protein 5A2 |
FS | Gh_A10G2036 | Rop guanine nucleotide exchange factor 5 |
FS | Gh_A07G1801 | Peptidyl-prolyl cis-trans isomerase FKBP15-1 |
FU | Gh_A11G2663 | Protein WVD2-like 1 |
FU | Gh_D11G2059 | Not available |
FE | Gh_A13G0282 | Xylulose kinase |
FE | Gh_A13G0354 | Not available |
FE | Gh_A11G1313 | EPIDERMAL PATTERNING FACTOR-like protein 9 |
MIC | Gh_D11G1416 | Transcriptional corepressor LEUNIG_HOMOLOG |