Skip to main content
NCBI home page
Genetics logoLink to Genetics
. 2016 Aug 24;204(2):799–806. doi: 10.1534/genetics.116.193086

Insights into Interspecific Hybridization Events in Allotetraploid Cotton Formation from Characterization of a Gene-Regulating Leaf Shape

Lijing Chang 1, Lei Fang 1, Yajuan Zhu 1, Huaitong Wu 1, Zhiyuan Zhang 1, Chunxiao Liu 1, Xinghe Li 1, Tianzhen Zhang 1,1
PMCID: PMC5068863  PMID: 27558139

Abstract

The morphology of cotton leaves varies considerably. Phenotypes, including okra, sea-island, super-okra, and broad leaf, are controlled by a multiple allele locus, L2. Okra leaf (L2°) is an incomplete mutation that alters leaf shape by increasing the length of lobes with deeper sinuses. Using a map-based cloning strategy, we cloned the L2 locus gene, which encodes a LATE MERISTEM IDENTITY 1 (LMI1)-like transcription factor (GhOKRA). Silencing GhOKRA leads to a change in phenotype from okra to broad leaf. Overexpression of GhOKRA in Arabidopsis thaliana greatly increases the degree of the leaf lobes and changes the leaf shape. Premature termination of translation in GhOKRA results in the production of broad leaves. The sequences of OKRA from diploid progenitor D-genome species, and wild races and domesticated allotetraploid cottons in Gossypium hirsutum show that a premature termination mutation occurred before and after the formation of tetraploid cotton, respectively. This study provides genomic insights into the two interspecific hybridization events: one produced the present broad leaf and another formed okra leaf phenotype with complete OKRA, that occurred during allotetraploid cotton formation.

Keywords: cotton, leaf shape, HD-ZIP I transcription factor, interspecific hybridization, evolution

THE cotton genus Gossypium includes ∼45 diploid (2n = 2x = 26) and five allotetraploid (2n = 4x = 52) species (Fryxell 1992; Wendel et al. 2009). Diploid Gossypium species fall into eight different genome groups (A–G, plus K) based on the observations of meiotic pairing behavior (Fryxell 1979; Endrizzi et al. 1985; Percival et al. 1999). Allotetraploids originated from interspecific hybridization between diploid species in the New World (Wendel 1989; Wendel et al. 1995). Gossypium herbaceum or G. arboreum and G. raimondii are generally regarded as the best extants of the A- and D-subgenome progenitors, respectively (Endrizzi et al. 1985; Wendel et al. 1995; Zhao et al. 1998).

The length of the leaf lobes is highly variable in the Gossypium genus, but the predominant form in cultivated cottons is the broad leaf type. The lobe of the leaf varies in shape from broadly triangular to lanceolate or even willow leaf. Okra (L2°), sea-island (L2e), super-okra (L2s), sub okra (L2u), and broad leaf (l2) are members of a multiple allelic series that exist at the L2 locus in chromosome (Chr) D1 (or Chr 15) (Stephens 1945). Okra leaf shape solely found in the D subgenome in tetraploid cotton is incomplete dominant inheritance and alters leaf shape by increasing the length of lobes with deeper sinuses. Broad leaf cotton has wider leaf lobes and less indentation between the major lobes than the okra types. Broad leaf (l2) is recessive to okra leaf (L2) (Stephens 1945). Cotton with the okra leaf phenotype has a potential resistance to insect pests and drought. The okra leaf type also has a greater CO2 exchange rate, water use efficiency, and a lower stomatal conductance compared to normal (broad) leaf type of cotton (Pettigrew et al. 1993). It also provides less favorable microenvironmental conditions for the habitat of Bemisia tabaci, leading to fewer adults and immatures (eggs and nymphs) because of the more open canopy, as evidenced by the higher leaf perimeter-to-leaf area ratio (Chu et al. 2002). Okra leaf traits can be used as a source of biotic and abiotic stress resistance. Many extra-long staple cotton (G. barbadense) cultivars have the L2e sea-island genotype, which is similar to L2u and heterozygous okra (L2°l2) in phenotype (Meredith 1984). Okra leaf Upland cotton was grown in Australia on a major scale, accounting for between 40 and 60% of Australian seed sales between 1987 and 1993 (Stiller and Wilson 2014). The Commonwealth Scientific and Industrial Research Organization has released many okra leaf cultivars, including Siokra1-4 and Siokra S-324. Okra leaf cultivars such as Biaoza 1A have also been developed for a range of circumstances, such as short season production systems and dryland production systems in Xinjiang, China (Ma et al. 2001).

L2o was mapped to a 5.4 cM interval with molecular markers, and two LATE MERISTEM IDENTITY 1 (LMI1)-like genes have been identified as candidates for okra leaf (Andres et al. 2014). Using a combination of targeted association analysis, F2 population-based fine mapping, and comparative sequencing of orthologs, Zhu et al. (2016) further identified one LMI1-like gene to be a candidate gene underlying the okra leaf trait in G. hirsutum and named as GhOKRA, which was a class I homeodomain leucine zipper (HD-ZIP I) transcription factor. The homologous gene of LMI1, REDUCED COMPLEXITY (RCO), was reported to sculpt developing leaflets by repressing growth at their flanks in Cardamine hirsute (Vlad et al. 2014). In the present research, we used a chromosomal segment introgression line (CSIL) IL-15-5-1 with sea-island leaves (Figure 1A) and map-based isolated the HD-ZIP I transcription factor, which is responsible for okra leaf development in Gossypium. Virus-induced gene silencing (VIGS) assay and transgenic Arabidopsis studies confirmed that GhOKRA regulates okra leaf development in Gossypium. Sequence analysis of OKRA in D-genome diploid species and G. hirsutum races further provides genomic insights into the two interspecific hybridization events that occurred during allotetraploid cotton formation.

Figure 1.

Figure 1

Open in a new tab

Map-based cloning of the leaf morphology gene L2e. (A) The leaf phenotype of parents used for mapping: TM-1 and IL-15-5-1. From left to right: Third leaf, second leaf, and first leaf in the top to represent mature to young leaves. Bar, 5 cm. (B) L2e was first mapped on the D1 chromosome between the markers JESPR152 and NAU3040 using an F2 population. The genetic distance between the markers was 6.91 cM. L2e was further fine mapped to a region between markers JESPR152 and H1776 using 112 dominant individuals. A 183-kb region containing 18 putative ORFs was obtained by mapping this region with sequences in G. raimondii Chr 02. The homologous gene of AtLMI1 is shown at the red arrow.

Materials and Methods

Plant materials and genotyping

TM-1 is a standard genetic line of Upland cotton and has broad leaves (Kohel et al. 1970). T586 is a G. hirsutum multiple gene marker line and has okra leaves (Fryxell 1984; Endrizzi et al. 1985). Hai7124 is a commercial Sea-island Verticillium-resistant cultivar with sea-island leaves. Materials used in our research are summarized in Supplemental Material, Table S1. Eighty-nine accessions from seven races of G. hirsutum were listed in Table S2, provided by the Institute of Cotton Research of the Chinese Academy of Agricultural Sciences (CAAS). IL-15-5-1 was derived from a CSIL with sea-island leaves that carried the Hai7124 segment from chromosome D1 in the background of the recurrent parent TM-1 (Wang et al. 2012). Two F2 mapping populations of 770 and 508 individuals were developed from a cross between TM-1 as the female parent and IL-15-5-1 as the male parent. Their F1 had an intermediate phenotype. In the first mapping population, (IL-15-5-1 × TM-1)F3 produced 176 homozygous sea-island, 392 segregation for sea-island and broad leaf, and 202 homozygous broad leaf family lines. And the second mapping population contained 112 homozygous sea-island individuals from (IL-15-5-1 × TM-1)F2. All the materials and mapping populations were grown at the Jiangpu Breeding Station with normal field practices and in the greenhouses of Nanjing Agriculture University. Fresh leaves were used as the source of genomic DNA, and leaf primordia of different leaf shape varieties were harvested for total RNA extraction. All samples were immediately frozen in liquid nitrogen after collection and stored at −70°.

Cloning of the L2 gene

The genome sequence of L2 was amplified from Super okra, T586, TM-1, and Hai7124 plants and from the other materials listed in Table S1. The gene-specific PCR primer pairs used are described in Table S3. The complementary DNA (cDNA) of GhOKRA was cloned from the leaf primordium of super-okra cotton by reverse transcription-polymerase chain reaction (RT-PCR). Amplification products produced using ExTaq DNA Polymerase (TaKaRa Bio, Shiga, Japan) were cloned into the pMD19-T vector (TaKaRa Bio) for sequencing by Jinsite Biotechnology (Nanjing, China). The A-subgenome and D-subgenome sequences were distinguished by alignment with the homologous genes in G. raimondii (Paterson et al. 2012) and G. arboreum (Li et al. 2014).

Phylogenetic analysis of OKRA sequences

Alignment of the sequences of OKRA from different cotton materials (Table S1) was performed with ClustalX (1.83) (Thompson et al. 1997) and CLC sequence view 6. A phylogenetic tree was constructed by the maximum likelihood (ML) method in MEGA 6.06 (www.megasoftware.net) and the bootstrap test of phylogeny was performed with 1000 replications (Tamura et al. 2013).

Quantitative RT-PCR analysis

Total RNA was extracted from cotton leaf primordia and A. thaliana rosette leaves using the Biospin Plant Total RNA Extraction Kit (BioFlux, cat. no. BSC65S1). Leaf primordia at squaring stage were used for detecting the expression difference between species. Leaf primordia at seedling stage were used in the VIGS experiment. First-strand cDNA was generated using TransScript One-Step Genomic DNA Removal and cDNA Synthesis SuperMix (TransGen Biotech, Beijing, China, cat. no. AT311) kits according to the manufacturer’s instructions. The cotton Histone 3 gene (His3, GenBank accession no. AF024716) and Actin2 (Act2, At3g18780) of A. thaliana were used as reference genes. The sequences of the quantitative RT-PCR (qRT-PCR) primers are listed in Table S3. qRT-PCR products were quantified by the ABI 7500 Real Time System (Applied Biosystems, Foster City, CA) and the Light Cycler Fast Start DNA Master SYBR Green I kit (Roche, Basel, Switzerland) according to the manufacturer’s instructions.

VIGS assay

The cDNA fragment of GhOKRA was cloned from the leaf primordium of super-okra cotton. Gene-specific primers (Table S3) and anchored oligo (dT)18 primers were used to obtain a cDNA fragment for 3′-rapid amplification of cDNA ends (3′-RACE). A 273-bp fragment from the 3′ end of GhOKRA cDNA was amplified from the super-okra cotton. The product was cloned into EcoRI–XbaI-digested pTRV2, generating a pTRV2–GhOKRA construct. The vectors, pTRV1 and pTRV2–GhOKRA, were introduced into Agrobacterium tumefaciens strain GV3101. More than 15 super-okra individuals were infiltrated with a 1:1 mixture of Agrobacterium carrying pTRV1 and pTRV2 as a negative control, or pTRV1 and pTRV2–GhOKRA, separately. The primers used for the construction of VIGS vectors are listed in Table S3. The VIGS assay was carried out independently three times, according to methods described previously (Liu et al. 2002).

Subcellular localization of GhOKRA

The coding region of the GhOKRA gene was amplified from the super-okra cotton and inserted into GFP vector pBinGFP4 (Liu et al. 2014) to generate the construct pBin-35S::GhOKRA-GFP4. The construct was transformed into A. tumefaciens strain GV3101 using methods described previously (Sparkes et al. 2006). The subcellular localization of the OKRA–GFP4 fusion protein was observed with a confocal laser scanning microscope (Zeiss LSM 780, Berlin, Germany).

A. thaliana transformation

The fragments of truncated GhOKRAM from TM-1, GrOKRAM from G. raimondii, and complete GhOKRA from super-okra cotton were obtained using primers (Table S3) and recombined into the pBI121 plasmid using a ClonExpress II One Step Cloning Kit (C112-01, Vazyme Biotech, Nanjing, China) to construct the expression vectors pBI121-35S::GhOKRA/GhOKRAM/GrOKRAM used for plant transformation. GhOKRAM and GrOKRAM represented the genome sequence of premature termination GhOKRA in TM-1 and GrOKRA in G. raimondii, respectively. The complete GhOKRA fragment was cloned from the cDNA of super-okra cotton. Vectors were introduced into A. tumefaciens strain GV3101 for transformation of A. thaliana using the floral dip method (Clough and Bent 1998). Transgenic plants were tested by kanamycin selection marker and PCR analysis of NPTII and 35S-GhOKRA.

Data availability

Table S1 contains detailed descriptions of materials used in this research. Table S2 contains accessions from seven races of G. hirsutum and genotypes for each accession. Table S3 contains all of the primers designed in this research.

Results

Mapping of the okra leaf gene

IL-15-5-1 is a CSIL derived from a cross between G. barbadense cv. Hai7124 with a sea-island leaf and the recurrent parent G. hirsutum accession TM-1 (Figure 1A). Although it is phenotypically a sea-island leaf plant (Wang et al. 2012), the CSIL line is a near-isogenic line to TM-1. IL-15-5-1 was crossed as a male parent with TM-1 to produce a mapping population. Their F1 had an intermediate phenotype and (IL-15-5-1 × TM-1)F3 produced 176 homozygous sea-island, 392 segregation for sea-island and broad leaf, and 202 homozygous broad leaf family lines, consistent with the segregation ratio of 1:2:1 (χ2 = 2.01 < χ20.05, 2 = 5.99, Table S4). Using microsatellites or simple sequence repeats (SSRs) distributed on Chr 1D, the L2e gene was anchored in the introgression segment and linked with SSRs, JESPR152 and NAU3040 (Table S3). To isolate the L2e gene, we reproduced (IL-15-5-1 × TM-1)F2, consisting of 508 individuals for the fine mapping of L2e (Table S4). We screened out 112 homozygous sea-island individuals from (IL-15-5-1 × TM-1)F2 to narrow down the L2e locus between markers JESPR152 and NAU3040, using other SSRs (Table S3). The L2e gene was delimited in a 183-kb region flanked by markers JESPR152 and H1776, based on the genome sequences of G. raimondii (Paterson et al. 2012) and G. hirsutum (Zhang et al. 2015) (Figure 1B).

Cloning of the okra leaf gene

In the 183-kb region, 18 open reading frames (ORFs) were predicted through annotation of the genome sequence of G. raimondii Chr 02 (Table S5). qRT-PCR of ORFs revealed that the expression of ORF17 in leaf primordium at squaring stage was consistently different between the super-okra leaf (Super okra), okra leaf (T586), sea-island leaf (Hai7124), and broad leaf (TM-1) (Figure 2, A and B and Figure S1). This gene was not expressed detectably in the broad leaf (TM-1), but its expression was significantly higher in the super-okra leaf (Super okra) than in the okra leaf (T586) and sea-island leaf (H7124). This gene is annotated as a homologous gene of LATE MERISTEM IDENTITY 1 (LMI1) of Arabidopsis and is an HD-ZIP I transcription factor. This HD-ZIP I transcription factor likely controlling the deeply lobed leaf shape in Gossypium is the same as the gene GhOKRA identified by Zhu et al. (2016).

Figure 2.

Figure 2

Open in a new tab

Analysis of genome sequences and transcript levels of OKRA in leaves of different shapes. (A) Different leaf shapes of cotton. From left to right: broad (TM-1), sea-island (Hai7124), okra (T586), super-okra (Super okra). (B) qRT-PCR analysis of OKRA gene expression in leaves of different shapes. The expression of OKRA was consistent with the phenotype of the leaf lobes. (C) Genome sequence analysis of OKRA. One 8-bp deletion and one 1-bp deletion were located in the TM-1 genome sequence at positions 742 and 861 bp from the ATG start codon, respectively. In this super-okra line, the single-nucleotide “G” labeled by red frame was not deleted at the C terminus. Data represent the means ± SDs of three replicated experiments. Different letters indicate significant differences between columns at P < 0.05.

We cloned the genomic sequences of OKRA in Chr 1D (referred as OKRA-1D) from different varieties of leaf including Super okra, T586, Hai7124, and TM-1. Multiple sequence alignment revealed that two single-nucleotide transitions at positions 185 and 268 bp from the ATG start codon were detected, along with one 8-bp deletion and one 1-bp deletion at positions 742 and 861 bp from the ATG start codon, respectively, in TM-1 (Figure 2C and Figure S2 and Table S1). In our super-okra line, we did not observe a single-nucleotide deletion at the C terminus that caused the super-okra leaf phenotype reported by Zhu et al. (2016). It is supposed that the broad leaf is mutated from the wild-type (WT) okra leaf, due to a frameshift mutation. OKRA–GFP was located in the nucleus, which was consistent with its putative function as a transcription factor (Figure S3).

GhOKRA controls cotton okra leaf development

To explore whether GhOKRA participates in leaf morphogenesis or not, we cloned 3′-end fragments from GhOKRA of super-okra cotton into VIGS pTRV2 vectors (Liu et al. 2002) and constructed a pTRV2–GhOKRA construct (Figure 3A). At 45 days after the induction of TRV-mediated gene silencing, the leaf of all super-okra cotton infiltrated with pTRV2–GhOKRA changed from deeply lobed okra to palmate (Figure 3, B and C). qRT-PCR analysis showed that the GhOKRA transcript was down-regulated in pTRV2–GhOKRA VIGS lines as compared with controls (Figure 3D). These results demonstrate that GhOKRA regulates leaf shape development.

Figure 3.

Figure 3

Open in a new tab

VIGS phenotypes and suppression of the endogenous transcripts. (A) Schematic representation of the structure of pTRV2–GhOKRA containing the 3′ cDNA in a 273-bp length of GhOKRA. (B and C) Leaf shape phenotypes: Super okra, pTRV2, pTRV2–GhOKRA, and TM-1. The leaves of plants with pTRV2–GhOKRA transformed to a broad shape but the leaves of plants with pTRV2 did not change. pTRV2 was used as a negative control. (D) qRT-PCR analysis was used to examine the transcript levels of GhOKRA in leaf primordia of plants infected with pTRV2 and pTRV2–GhOKRA. The expression of GhOKRA was significantly silenced by VIGS in plants infected with pTRV2–GhOKRA. Data represent the means ± SDs of three replicated experiments. Asterisks indicate a significant difference at * P < 0.05 and ** P < 0.01.

In G. hirsutum, there are three homologous genes of AtLMI1 that include GhOKRA. The other two homologs were not related to leaf development by VIGS assay (data not shown), suggesting their function has diverged.

Overexpression of GhOKRA changes A. thaliana leaf shape

To determine whether premature termination or low expression of GhOKRA in TM-1 changes leaf development, we developed transgenic A. thaliana lines that overexpressed GhOKRAM (a genomic sequence of premature termination of GhOKRA in TM-1), GrOKRAM (a genomic sequence of GrOKRA in G. raimondii), and complete GhOKRA (encoding region of cDNA from super-okra cotton) under control of the 35S promoter (Pro35S::GhOKRA) in Col-0. The expression of GhOKRA and GhOKRAM increased significantly in these transgenic plants (Figure 4A). The rosette leaf blades in transgenic plants overexpressing GhOKRA became narrow, abaxially curled, and divided into lobes as compared to the WT in the T1 generation (Figure 4, B and C), and the petiole was longer than that in the WT (Figure 4, D and E). There was no phenotypic difference between transgenic plants overexpressing GhOKRAM, GrOKRAM , and the WT (Figure 4, B and F). These results show that premature termination of translation of GhOKRA derived from an 8-bp deletion made this gene nonfunctional in TM-1. This result further verifies that GhOKRA controls leaf shape in cotton in the same way as in Arabidopsis.

Figure 4.

Figure 4

Open in a new tab

Analysis of truncated GhOKRA by overexpression in A. thaliana. (A) Relative expression levels of GhOKRA in transgenic plants. The expression levels of GhOKRA and GhOKRAM in leaves from three transgenic individual plants were all significantly higher than that in the WT. (B) Rosette of A. thaliana WT (Col-0). (C) Rosette of A. thaliana expressing GhOKRA. (D and E) Leaf phenotype of WT and transgenic plants expressing GhOKRA. (F) Rosette of A. thaliana expressing GhOKRAM. Transgenic plants overexpressing GhOKRA showed narrow, abaxially curled and divided leaves compared with the WT, while transgenic plants overexpressing GhOKRAM showed no differences compared with the WT. Premature termination led to the nonfunction of GhOKRA in TM-1. Data represent the means ± SDs of three replicated experiments. Asterisks indicate a significant difference at * P < 0.05 and ** P < 0.01. Bar, 1 cm.

Evolution of leaf shape in D-genome diploid species

We found two GhOKRA homologs, Thecc1EG022391t1 and CGD0020808, in cacao (Theobroma cacao) (Argout et al. 2011). Of the sequenced species, cacao is most closely related to the cotton genus (Paterson et al. 2012). The multiple sequence alignments indicated that no such 1- and 8-bp deletions occur in cacao (Figure S4). There exist 13–14 D-genome diploid species (Wendel et al. 2009). G. raimondii (D5) is generally regarded as the best extant of the D-subgenome progenitor (Endrizzi et al. 1985; Wendel et al. 1995; Zhao et al. 1998). Leaf shapes in D-genome diploid species are variable from the broad or unlobed leaf in G. aridum, G. davidsonii, and G. raimondii, to the deeply lobed leaf in G. thurberi and G. trilobum. The leaf shape of G. gossypioides is more or less deeply three lobed. We isolated the sequences of OKRA from these six D-genome diploid species mentioned above and found that there were remarkable differences in sequences in D-genome diploid species, especially at the position of the termination codon (Figure S5). Sequence analysis showed that the 1-bp deletion present in TM-1 was also found in G. raimondii (Figure S6), but the 8-bp deletion in TM-1 was absent from all D-genome species. There was an 18-bp deletion in G. gossypioides and a transition mutation (T/C) delaying the termination in G. davidsonii (Figure S5). Phylogenetic analysis showed that the OKRA sequence in G. raimondii was closest to that in tetraploid cotton (Figure S6), confirming that G. raimondii (D5) is the best extant of the D-subgenome progenitor.

The evolution of leaf shape in allotetraploid cottons

Allotetraploid cotton originated from hybridization event(s) between an extant A-genome progenitor of G. herbaceum (A1) or G. arboreum (A2) and a D-genome progenitor, G. raimondii (D5), 1–1.5 million years ago (MYA) (Endrizzi et al. 1985; Wendel et al. 1995; Zhao et al. 1998; Zhang et al. 2015). However, the number of times hybridization events occurred to produce allotetraploid cottons remains to be explored. Based mainly on their morphologies and distinct geographical distributions, one wild and six domesticated races of G. hirsutum have been identified (Hutchinson 1951). We isolated GhOKRA genome sequences from 89 accessions of all seven races with variable leaf shapes to evaluate the evolution of leaf shape in tetraploid cotton (Figure 5A and Figure S7). There were three sequence types in the GhOKRA genome sequence among the 89 accessions (Table S2). Seventeen accessions from morrilli, palmeri, latifolium, and marie-galante races had sequences that were completely consistent with okra leaf cultivated species (type 1). Twelve accessions from punctatum, yucatanen, palmeri, marie-galante, and latifolium races had the same mutated sequence as three accessions from G. raimondii, containing 1-bp deletion (type 2), suggesting that the broad leaf may originate from diploid progenitor cotton G. raimondii, although we can not exclude the possibility that the same mutated sequence 1-bp deletion occurred after the formation of tetraploid cottons. The remaining accessions from richmondii, latifolium, morrilli, punctatum, palmeri, and marie-galante races had the same sequence as TM-1 and modern broad leaf cultivars, containing both the 8-bp and 1-bp deletions at positions 742 and 861 bp from the ATG start codon, respectively (type 3). These results revealed that another mutation (an 8-bp deletion) occurred to develop the broad leaf phenotype after the formation of tetraploid cotton.

Figure 5.

Figure 5

Open in a new tab

Multiple sequence alignment of GhOKRA in seven races of G. hirsutum. (A) Genome sequences of GhOKRA from some accessions randomly selected from 89 accessions from all seven races were isolated and classified into three sequence types based on the 1-bp deletion and 8-bp deletion at positions 742 and 861 bp from the ATG start codon, respectively, in TM-1. Five accessions from morrilli palmeri, and marie-galante races had sequences that were consistent with Super okra (type 1). Three accessions had the same sequence type as G. raimondii containing the 1-bp deletion (type 2). The remaining five accessions had the same sequence type as TM-1 (type 3). (B) The two interspecific hybridizations that occurred during allotetraploid cotton formation. One hybridization event occurred between extant progenitors of G. herbaceum or G. arboreum and G. raimondii. Another hybridization event occurred between G. herbaceum or G. arboreum and the WT or nascent G. raimondii, which had an okra leaf phenotype with complete OKRA.

Discussion

The tetraploid was spawned from the interspecific hybridization of an A-genome species resembling an African (G. herbaceum) or Asian (G. arboreum) species and a D-genome species resembling extant G. raimondii (Wendel 1989). Under long-term human selection of a wide range of morphological and physiological traits, the two tetraploid species, G. hirsutum and G. barbadense, have been domesticated in South and Central America, respectively. Twelve D-genome diploid species are endemic to western Mexico, indicating this area is the center of diversity of the D genome. However, G. raimondii is presently distributed only in Peru. The 1-bp deletion in OKRA present in wild race species and modern cotton cultivars was found only in G. raimondii, indicating that the mutation in tetraploid cottons originated from the D-diploid progenitor species G. raimondii. We identified three types of GhOKRA in 89 accessions from seven wild and domesticated races of tetraploid Upland cotton and found that another 8-bp deletion mutation occurred after the formation of tetraploid cotton. From the evolution of leaf shape in allotetraploid cottons, we can reasonably deduce that allotetraploid cottons may originate from at least two interspecific hybridizations (Figure 5B). One hybridization event occurred between extant progenitors of G. herbaceum or G. arboreum and G. raimondii in which a 1-bp “G” deletion had already occurred before the formation of tetraploid cotton and produced the present broad leaf tetraploid cottons. Although a premature termination mutation with 1-bp deletions at positions 861 bp could occur after the tetraploid cotton formation, the possibility of such a premature termination mutation occurred at the same position in the tetraploid cotton and G. raimondii would be very low. Another hybridization event occurred between G. herbaceum or G. arboreum and the WT or nascent G. raimondii, which had an okra leaf phenotype with complete OKRA, to form the okra leaf tetraploid cottons. The phenotype is validated with GhOKRA haplotypes in the wild and domesticated races of tetraploid Upland cotton. After the formation of tetraploid cotton, a further 8-bp deletion occurred in the tetraploid cottons, originating from already broad leaf tetraploid cottons, and resulting in the gene l2, as found in TM-1.

The phenotypic differences in leaf shape are controlled by a single multiple allele series in tetraploid cotton (Stephens 1945). Leaf shape in the D-genome diploid is highly variable. In the present research, using a map-based cloning strategy, we cloned the L2 locus genes using a sea-island okra leaf (L2e) CSIL, IL-15-5-1 (Wang et al. 2012), and found that both 1- and 8-bp deletions cause premature termination of translation and the nonfunction of OKRA and lead to broad leaf development in Gossypium. Super-okra leaves had a higher expression of OKRA than other okra leaves at the transcriptional level. Although the remarkable morphological variation between okra (L2°), sea-island (L2e), super-okra (L2s), and sub-okra (L2u) types are attributed to the same L2 multiple allele locus (Stephens 1945), their ORFs in this OKRA did not show any differences in their sequences. It remains to be explored whether it is epigenetics, a cis-regulatory mutation at this locus, or a causative mutation on a distant gene exerting its influence on OKRA that has caused such varied expression in these different okra leaf lines with the same ORF sequences.

Acknowledgments

This work was financially supported in part by grants from the National Natural Science Foundation of China (31330058), National Key R & D Program for Crop Breeding (2016YFD0100300), and the Jiangsu Province Collaborative Innovation Center-Modern Crop Production (JCIC-MCP) project. We thank the Medium-Term Gene Bank of Cotton in China and the Institute of Cotton Research of CAAS for providing some of the cotton seeds used in the present study. The authors declare no competing financial interests.

Author contributions: T.Z. designed the research; L.C., L.F., and Y.Z. performed the research; T.Z., L.C., and L.F. analyzed all data and wrote the manuscript; Y.Z., H.W., and X.L. constructed mapping populations; H.W. designed SSR primers; L.C. and Z.Z. conducted subcellular localization; L.C., L.F., and C.L. analyzed the evolution; and all authors discussed results and commented on the manuscript.

Footnotes

Supplemental material is available online at www.genetics.org/lookup/suppl/doi:10.1534/genetics.116.193086/-/DC1.

Communicating editor: S. Poethig

Literature Cited

  1. Andres R. J., Bowman D. T., Kaur B., Kuraparthy V., 2014.  Mapping and genomic targeting of the major leaf shape gene (L) in Upland cotton (Gossypium hirsutum L.). Theor. Appl. Genet. 127: 167–177. [DOI] [PubMed] [Google Scholar]
  2. Argout X., Salse J., Aury J. M., Guiltinan M. J., Droc G., et al. , 2011.  The genome of Theobroma cacao. Nat. Genet. 43: 101–108. [DOI] [PubMed] [Google Scholar]
  3. Chu C. C., Natwick E. T., Henneberry T. J., 2002.  Bemisia tabaci (Homoptera: Aleyrodidae) biotype B colonization on okra- and normal-leaf upland cotton strains and cultivars. J. Econ. Entomol. 95: 733–738. [DOI] [PubMed] [Google Scholar]
  4. Clough S. J., Bent A. F., 1998.  Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J. 16: 735–743. [DOI] [PubMed] [Google Scholar]
  5. Endrizzi J. E., Turcotte E. L., Kohel R. J., 1985.  Genetics, cytology, and evolution of Gossypium. Adv. Genet. 23: 271–375. [Google Scholar]
  6. Fryxell P. A., 1979.  The Natural History of the Cotton Tribe. Texas A&M University Press, College Station, TX. [Google Scholar]
  7. Fryxell P. A., 1984.  Taxonomy and germplasm resources. pp. 27–57 in Cotton, edited by Kohel R. J., Lewis C. F. American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America, Madison, WI. [Google Scholar]
  8. Fryxell P. A., 1992.  A revised taxonomic interpretation of Gossypium L. (Malvaceae). Rheedea 2: 108–165. [Google Scholar]
  9. Hutchinson J. B., 1951.  Intra-specifc differentiation in Gossypium hirsutum. Heredity 5: 161–193. [Google Scholar]
  10. Kohel R., Richmond T. R., Lewis C. F., 1970.  Texas Marker1. Description of a genetic standard for Gossypium hirsutum L. Crop Sci. 10: 670–673. [Google Scholar]
  11. Li F., Fan G., Wang K., Sun F., Yuan Y., et al. , 2014.  Genome sequence of the cultivated cotton Gossypium arboreum. Nat. Genet. 46: 567–572. [DOI] [PubMed] [Google Scholar]
  12. Liu T., Song T., Zhang X., Yuan H., Su L., et al. , 2014.  Unconventionally secreted effectors of two filamentous pathogens target plant salicylate biosynthesis. Nat. Commun. 5: 4686. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Liu Y., Schiff M., Dinesh-Kumar S. P., 2002.  Virus-induced gene silencing in tomato. Plant J. 31: 777–786. [DOI] [PubMed] [Google Scholar]
  14. Ma Q., Yang X., Liu J., Wang Z., Xing T., et al. , 2001.  Hybrid cotton Biaoza A1. China Cotton 28: 23–24. [Google Scholar]
  15. Meredith W. R., 1984.  Influence of leaf morphology on lint yield of cotton-enhancement by the sub okra trait. Crop Sci. 24: 855–857. [Google Scholar]
  16. Paterson A. H., Wendel J. F., Gundlach H., Guo H., Jenkins J., et al. , 2012.  Repeated polyploidization of Gossypium genomes and the evolution of spinnable cotton fibres. Nature 492: 423–427. [DOI] [PubMed] [Google Scholar]
  17. Percival A. E., Wendel J. F., Stewart J. M., 1999.  Taxonomy and germplasm resources, pp. 33–63 in Cotton: Origin, History, Technology, and Production, edited by Smith C. W., Cothren J. T. John Wiley and Sons, New York. [Google Scholar]
  18. Pettigrew W. T., Heitholt J. J., Vaughn K. C., 1993.  Gas-exchange differences and comparative anatomy among cotton leaf-type isolines. Crop Sci. 33: 1295–1299. [Google Scholar]
  19. Sparkes I. A., Runions J., Kearns A., Hawes C., 2006.  Rapid, transient expression of fluorescent fusion proteins in tobacco plants and generation of stably transformed plants. Nat. Protoc. 1: 2019–2025. [DOI] [PubMed] [Google Scholar]
  20. Stephens S. G., 1945.  A genetic survey of leaf shape in New World cottons: a problem in critical identification of alleles. J. Genet. 56: 313–330. [Google Scholar]
  21. Tamura K., Stecher G., Peterson D., Filipski A., Kumar S., 2013.  MEGA6: molecular evolutionary genetics analysis version 6.0. Mol. Biol. Evol. 30: 2725–2729. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Thompson J. D., Gibson T. J., Plewniak F., Jeanmougin F., Higgins D. G., 1997.  The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25: 4876–4882. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Vlad D., Kierzkowski D., Rast M. I., Vuolo F., Dello Ioio R., et al. , 2014.  Leaf shape evolution through duplication, regulatory diversification, and loss of a homeobox gene. Science 343: 780–783. [DOI] [PubMed] [Google Scholar]
  24. Wang P., Zhu Y. J., Song X. L., Cao Z. B., Ding Y. Z., et al. , 2012.  Inheritance of long staple fiber quality traits of Gossypium barbadense in G. hirsutum background using CSILs. Theor. Appl. Genet. 124: 1415–1428. [DOI] [PubMed] [Google Scholar]
  25. Wendel J. F., 1989.  New world tetraploid cottons contain old world cytoplasm. Proc. Natl. Acad. Sci. USA 86: 4132–4136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Wendel J. F., Schnabel A., Seelanan T., 1995.  An unusual ribosomal DNA sequence from Gossypium gossypioides reveals ancient, cryptic, intergenomic introgression. Mol. Phylogenet. Evol. 4: 298–313. [DOI] [PubMed] [Google Scholar]
  27. Wendel J. F., Brubaker C., Alvarez I., Cronn R., Stewart J., 2009.  Evolution and natural history of the cotton genus, pp. 3–22 in Genetics and Genomics of Cotton, edited by Paterson H. A. Springer-Verlag, New York. [Google Scholar]
  28. Stiller W. N., Wilson I. W., 2014.  Australian cotton germplasm resources, pp. 1–34 in World Cotton Germplasm Resources, edited by Abdurakhmonov I. Y. InTech, Rijeka, Croatia. [Google Scholar]
  29. Zhang T., Hu Y., Jiang W., Fang L., Guan X., et al. , 2015.  Sequencing of allotetraploid cotton (Gossypium hirsutum L. acc. TM-1) provides a resource for fiber improvement. Nat. Biotechnol. 33: 531–537. [DOI] [PubMed] [Google Scholar]
  30. Zhao X. P., Si Y., Hanson R. E., Crane C. F., Price H. J., et al. , 1998.  Dispersed repetitive DNA has spread to new genomes since polyploid formation in cotton. Genome Res. 8: 479–492. [DOI] [PubMed] [Google Scholar]
  31. Zhu Q. H., Zhang J., Liu D., Stiller W., Zhang Z., et al. , 2016.  Integrated mapping and characterization of the gene underlying the okra leaf trait in Gossypium hirsutum L. J. Exp. Bot. 67: 763–774. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Table S1 contains detailed descriptions of materials used in this research. Table S2 contains accessions from seven races of G. hirsutum and genotypes for each accession. Table S3 contains all of the primers designed in this research.

Articles from Genetics are provided here courtesy of Oxford University Press

RESOURCES