Iso-Seq analysis of the Taxus cuspidata transcriptome reveals the complexity of Taxol biosynthesis

Abstract

Background Taxus cuspidata is well known worldwide for its ability to produce Taxol, one of the top-selling natural anticancer drugs. However, current Taxol production cannot match the increasing needs of the market, and novel strategies should be considered to increase the supply of Taxol. Since the biosynthetic mechanism of Taxol remains largely unknown, elucidating this pathway in detail will be very helpful in exploring alternative methods for Taxol production. Results Here, we sequenced Taxus cuspidata transcriptomes with next-generation sequencing (NGS) and third-generation sequencing (TGS) platforms. After correction with Illumina reads and removal of redundant reads, more than 180,000 nonredundant transcripts were generated from the raw Iso-Seq data . Using Cogent software and an alignment-based method, we identified a total of 139 cytochrome P450s (CYP450s), 31 BAHD acyltransferases (ACTs) and 1940 transcription factors (TFs). Based on phylogenetic and coexpression analysis, we identified 9 CYP450s and 7 BAHD ACTs as potential lead candidates for Taxol biosynthesis and 6 TFs that are possibly involved in the regulation of this process. Using coexpression analysis of genes known to be involved in Taxol biosynthesis, we elucidated the stem biosynthetic pathway. In addition, we analyzed the expression patterns of 12 characterized genes in the Taxol pathway and speculated that the isoprene precursors for Taxol biosynthesis were mainly synthesized via the MEP pathway. In addition, we found and confirmed that the alternative splicing patterns of some genes varied in different tissues, which may be an important tissue-specific method of posttranscriptional regulation. Conclusions A strategy was developed to generate corrected full-length or nearly full-length transcripts without assembly to ensure sequence accuracy, thus greatly improving the reliability of coexpression and phylogenetic analysis and greatly facilitating gene cloning and characterization. This strategy was successfully utilized to elucidate the Taxol biosynthetic pathway, which will greatly contribute to the goals of improving the Taxol content in Taxus spp. using molecular breeding or plant management strategies and synthesizing Taxol in microorganisms using synthetic biological technology. Electronic supplementary material The online version of this article (10.1186/s12870-019-1809-8) contains supplementary material, which is available to authorized users.

IsoSeq transcriptome assembly of C3panicoid grasses provides tools to study evolutionary change in the Panicoideae

“…The average length of consensus sequences ranged from 925 bp to 1,438 kb (Figure S1). The number of consensus transcripts significantly exceeded the expected number of expressed genes, however, this is consistent with other reference genome-free IsoSeq analyses (Li et al, 2017; Kuang et al, 2019; Yan et al, 2019). Inflated numbers of consensus transcripts can result from sequencing of multiple alternatively spliced isoforms of the same gene, sequencing of incompletely processed mRNA molecules (Martin et al, 2014), high sequence error rates preventing multiple sequences from the same transcript being collapsed into a consensus, divergent haplotypes of the same locus present in our clonally propagated, wild collected, or partially inbred starting material, or contamination of the original samples with mRNA from non-target organisms.…”

Characterization and Analysis of the Full-Length Transcriptomes of Multiple Organs in Pseudotaxus chienii (W.C.Cheng) W.C.Cheng

“…The GO annotation analyses showed that “metabolic process”, “cellular process”, “binding”, “catalytic activity”, “cell”, and “cell part” comprise the majority of subcategories, indicating that the organs of P. chienii experience active cell metabolism to accumulate sufficient nutrition for the future growth and development of the organs. These GO terms are consistent with the previously investigated Taxus cuspidata transcriptome [ 25 ] and the Taxus mairei transcriptome [ 40 ]. The KEGG annotation analysis showed that “carbohydrate metabolism” and “signal transduction” are the top two pathways with the most abundant unigenes.…”

“…TFs regulate gene expression via interactions with themselves and other proteins under plant stress response [ 43 ]. In our study, we identified 1678, 2432, 1706, and 2497 TFs for the root, stem, leaf, and strobilus, respectively, belonging to 66 different families, which were comparable to those of another conifer species T. cuspidata (1940 TFs), and the model organism A. thaliana (2357 TFs) [ 25 ]. C2H2, C3H, bHLH, and bZIP were the most represented TF families in all four organs of P. chienii .…”

“…Transcriptome analysis could identify the type and number of intracellular key genes and reveal the metabolic pathways at a molecular level [ 24 , 25 ]. At present, several technologies have been applied for transcriptome sequencing.…”

IsoSeq transcriptome assembly of C3 panicoid grasses provides tools to study evolutionary change in the Panicoideae

“…While H. amplexicaulis exhibited the shortest consensus transcript length, this was not reflected in a reduced number of complete ORFs-those containing both an in-frame ATG and stop codon and occupying at least 60% of the total transcript length. The number of consensus transcripts significantly exceeded the expected number of expressed genes; however, this is consistent with other reference genome-freeIsoSeq analyses(Kuang, Sun, Wei, Li, & Sun, 2019;Li et al, 2017;Yan et al, 2019). Inflated numbers of consensus transcripts can result from sequencing of multiple alternatively spliced isoforms of the same gene, sequencing of incompletely processed mRNA molecules(Martin et al, 2014), high sequence error rates preventing multiple sequences from the same transcript being collapsed into a consensus, divergent haplotypes of the same locus present in our clonally propagated, wild collected, or partially inbred starting material, or contamination of the original samples with mRNA from non-target organisms.…”