Independent replication of mitochondrial genes supports the transcriptional program in developing fiber cells of cotton (Gossypium hirsutum L.)

Cell type-specific cytonuclear coevolution in three allopolyploid plant species

Cytonuclear disruption may accompany allopolyploid evolution as a consequence of the merger of different nuclear genomes in a cellular environment having only one set of progenitor organellar genomes. One path to reconcile potential cytonuclear mismatch is biased expression for maternal gene duplicates (homoeologs) encoding proteins that target to plastids and/or mitochondria. Assessment of this transcriptional form of cytonuclear coevolution at the level of individual cells or cell types remains unexplored. Using single-cell (sc-) and single-nucleus (sn-) RNAseq data from eight tissues in three allopolyploid species, we characterized cell type-specific variations of cytonuclear coevolutionary homoeologous expression and demonstrated the temporal dynamics of expression patterns across development stages during cotton fiber development. Our results provide unique insights into transcriptional cytonuclear coevolution in plant allopolyploids at the single-cell level.

The Immature Fiber Mutant Phenotype of Cotton (Gossypium hirsutum) Is Linked to a 22-bp Frame-Shift Deletion in a Mitochondria Targeted Pentatricopeptide Repeat Gene

Cotton seed trichomes are the most important source of natural fibers globally. The major fiber thickness properties influence the price of the raw material, and the quality of the finished product. The recessive immature fiber (im) gene reduces the degree of fiber cell wall thickening by a process that was previously shown to involve mitochondrial function in allotetraploid Gossypium hirsutum. Here, we present the fine genetic mapping of the im locus, gene expression analysis of annotated proteins near the locus, and association analysis of the linked markers. Mapping-by-sequencing identified a 22-bp deletion in a pentatricopeptide repeat (PPR) gene that is completely linked to the immature fiber phenotype in 2837 F₂ plants, and is absent from all 163 cultivated varieties tested, although other closely linked marker polymorphisms are prevalent in the diversity panel. This frame-shift mutation results in a transcript with two long open reading frames: one containing the N-terminal transit peptide that targets mitochondria, the other containing only the RNA-binding PPR domains, suggesting that a functional PPR protein cannot be targeted to mitochondria in the im mutant. Taken together, these results suggest that PPR gene Gh_A03G0489 is involved in the cotton fiber wall thickening process, and is a promising candidate gene at the im locus. Our findings expand our understanding of the molecular mechanisms that modulate cotton fiber fineness and maturity, and may facilitate the development of cotton varieties with superior fiber attributes.

AtRAV1 and AtRAV2 overexpression in cotton increases fiber length differentially under drought stress and delays flowering

There is a longstanding problem of an inverse relationship between cotton fiber qualities versus high yields. To better understand drought stress signaling and adaptation in cotton (Gossypium hirsutum) fiber development, we expressed the Arabidopsis transcription factors RELATED_TO_ABA-INSENSITIVE3/VIVIPAROUS1/(RAV1) and AtRAV2, which encode APETALA2-Basic3 domain proteins shown to repress transcription of FLOWERING_LOCUS_T (FT) and to promote stomatal opening cell-autonomously. In three years of field trials, we show that AtRAV1 and AtRAV2-overexpressing cotton had ∼5% significantly longer fibers with only marginal decreases in yields under well-watered or drought stress conditions that resulted in 40-60% yield penalties and 3-7% fiber length penalties in control plants. The longer transgenic fibers from drought-stressed transgenics could be spun into yarn which was measurably stronger and more uniform than that from well-watered control fibers. The transgenic AtRAV1 and AtRAV2 lines flowered later and retained bolls at higher nodes, which correlated with repression of endogenous GhFT-Like (FTL) transcript accumulation. Elevated expression early in development of ovules was observed for GhRAV2L, GhMYB25-Like (MYB25L) involved in fiber initiation, and GhMYB2 and GhMYB25 involved in fiber elongation. Altered expression of RAVs controlling critical nodes in developmental and environmental signaling hierarchies has the potential for phenotypic modification of crops.

Mapping-by-sequencing of Ligon-lintless-1 (Li 1 ) reveals a cluster of neighboring genes with correlated expression in developing fibers of Upland cotton (Gossypium hirsutum L.)

Mapping-by-sequencing and SNP marker analysis were used to fine map the Ligon-lintless-1 ( Li 1 ) short fiber mutation in tetraploid cotton to a 255-kb region that contains 16 annotated proteins. The Ligon-lintless-1 (Li 1 ) mutant of cotton (Gossypium hirsutum L.) has been studied as a model for cotton fiber development since its identification in 1929; however, the causative mutation has not been identified yet. Here we report the fine genetic mapping of the mutation to a 255-kb region that contains only 16 annotated genes in the reference Gossypium raimondii genome. We took advantage of the incompletely dominant dwarf vegetative phenotype to identify 100 mutants (Li 1 /Li 1 ) and 100 wild-type (li 1 /li 1 ) homozygotes from a mapping population of 2567 F2 plants, which we bulked and deep sequenced. Since only homozygotes were sequenced, we were able to use a high stringency in SNP calling to rapidly narrow down the region harboring the Li 1 locus, and designed subgenome-specific SNP markers to test the population. We characterized the expression of all sixteen genes in the region by RNA sequencing of elongating fibers and by RT-qPCR at seven time points spanning fiber development. One of the most highly expressed genes found in this interval in wild-type fiber cells is 40-fold under-expressed at the day of anthesis (DOA) in the mutant fiber cells.  This gene is a major facilitator superfamily protein, part of the large family of proteins that includes auxin and sugar transporters. Interestingly, nearly all genes in this region were most highly expressed at DOA and showed a high degree of co-expression. Further characterization is required to determine if transport of hormones or carbohydrates is involved in both the dwarf and lintless phenotypes of Li 1 plants.

Next generation genetic mapping of the Ligon-lintless-2 (Li₂) locus in upland cotton (Gossypium hirsutum L.)

Mapping-by-sequencing and novel subgenome-specific SNP markers were used to fine map the Ligon-lintless 2 ( Li 2 ) short-fiber gene in tetraploid cotton. These methodologies will accelerate gene identification in polyploid species. Next generation sequencing offers new ways to identify the genetic mechanisms that underlie mutant phenotypes. The release of a reference diploid Gossypium raimondii (D5) genome and bioinformatics tools to sort tetraploid reads into subgenomes has brought cotton genetic mapping into the genomics era. We used multiple high-throughput sequencing approaches to identify the relevant region of reference sequence and identify single nucleotide polymorphisms (SNPs) near the short-fiber mutant Ligon-lintless 2 (Li 2) gene locus. First, we performed RNAseq on 8-day post-anthesis (DPA) fiber cells from the Li 2 mutant and its wild type near isogenic line (NIL) Gossypium hirsutum cv. DP5690. We aligned sequence reads to the D5 genome, sorted the reads into A and D subgenomes with PolyCat and called SNPs with InterSNP. We then identified SNPs that would result in non-synonymous substitutions to amino acid sequences of annotated genes. This step allowed us to identify a 1-Mb region with 24 non-synonymous SNPs, representing the introgressed region that differentiates Li 2 from its NIL. Next, we sequenced total DNA from pools of F2 plants, using a super bulked segregant analysis sequencing (sBSAseq) approach. The sBSAseq predicted 82 non-synonymous SNPs among 3,494 SNPs in a 3-Mb region that includes the region identified by RNAseq. We designed subgenome-specific SNP markers and tested them in an F2 population of 1,733 individuals to construct a genetic map. Our resulting genetic interval contains only one gene, an aquaporin, which is highly expressed in wild-type fibers and is significantly under-expressed in elongating Li 2 fiber cells.