Transcriptome profiling of early developing cotton fiber by deep-sequencing reveals significantly differential...

Cotton fiber as a single-celled trichome is a biological model system for studying cell differentiation and elongation. However, the complexity of its gene expression and regulatory mechanism allows only marginal progress. Here, we report the high-throughput tag-sequencing (Tag-seq) analysis using Solexa Genome Analyzer platform on transcriptome of -2 to 1 (fiber initiation, stage I) and 2-8 (fiber elongation, stage II) days post anthesis (DPA) cotton (Gossypium hirsutum) ovules (wild type: WT; Xuzhou 142 and its mutant: fuzzless/lintless or flM, in the same background). To this end, we sequenced 3.5-3.8 million tags representing 0.7-1.0 million unique transcripts for each library (WT1, WT2, M1, and M2). After removal of low quality tags, we obtained a total of 2,973,104, 3,139,306, 2,943,654, and 3,392,103 clean sequences that corresponded to 357,852, 280,787, 372,952, and 382,503 distinct tags for WT1, WT2, M1, and M2, respectively. All clean tags were aligned to the publicly available cotton transcript database (TIGR, http://www.tigr.org). About 15% of the distinct tags were uniquely mapped to the reference genes, and 31.4% of existing genes were matched by tags. The tag mapping to the database sequences generated 23,854, 24,442, 23,497, and 19,957 annotated genes for WT1, WT2, M1, and M2 libraries, respectively. Analyses of differentially expressed genes revealed the substantial changes in gene type and abundance between the wild type and mutant libraries. Among the 20 most differentially expressed genes in WT1/M1 and WT2/M2 libraries were cellulose synthase, phosphatase, and dehydrogenase, all of which are involved in the fiber cell development. Overall, the deep-sequencing analyses demonstrate the high degree of transcriptional complexity in early developing fibers and represent a major improvement over the microarrays for analyzing transcriptional changes on a large scale.