No abstract

BackgroundDigitalis purpurea is an important ornamental and medicinal plant. There is considerable interest in exploring its transcriptome.ResultsThrough high-throughput 454 sequencing and subsequent assembly, we obtained 23532 genes, of which 15626 encode conserved proteins. We determined 140 unigenes to be candidates involved in cardiac glycoside biosynthesis. It could be grouped into 30 families, of which 29 were identified for the first time in D. purpurea. We identified 2660 mRNA-like npcRNA (mlncRNA) candidates, an emerging class of regulators, using a computational mlncRNA identification pipeline and 13 microRNA-producing unigenes based on sequence conservation and hairpin structure-forming capability. Twenty five protein-coding unigenes were predicted to be targets of these microRNAs. Among the mlncRNA candidates, only 320 could be grouped into 140 families with at least two members in a family. The majority of D. purpurea mlncRNAs were species-specific and many of them showed tissue-specific expression and responded to cold and dehydration stresses. We identified 417 protein-coding genes with regions significantly homologous or complementary to 375 mlncRNAs. It includes five genes involved in secondary metabolism. A positive correlation was found in gene expression between protein-coding genes and the homologous mlncRNAs in response to cold and dehydration stresses, while the correlation was negative when protein-coding genes and mlncRNAs were complementary to each other.ConclusionsThrough comprehensive transcriptome analysis, we not only identified 29 novel gene families potentially involved in the biosynthesis of cardiac glycosides but also characterized a large number of mlncRNAs. Our results suggest the importance of mlncRNAs in secondary metabolism and stress response in D. purpurea.

No abstract

Salvia miltiorrhiza is one of the most economically important medicinal plants. Cytochrome P450 (CYP450) genes have been implicated in the biosynthesis of its active components. However, only a dozen full-length CYP450 genes have been described, and there is no systematic classification of CYP450 genes in S. miltiorrhiza. We obtained 77,549 unigenes from three tissue types of S. miltiorrhiza using RNA-Seq technology. Combining our data with previously identified CYP450 sequences and scanning with the CYP450 model from Pfam resulted in the identification of 116 full-length and 135 partial-length CYP450 genes. The 116 genes were classified into 9 clans and 38 families using standard criteria. The RNA-Seq results showed that 35 CYP450 genes were co-expressed with CYP76AH1, a marker gene for tanshinone biosynthesis, using r≥0.9 as a cutoff. The expression profiles for 16 of 19 randomly selected CYP450 obtained from RNA-Seq were validated by qRT-PCR. Comparing against the KEGG database, 10 CYP450 genes were found to be associated with diterpenoid biosynthesis. Considering all the evidence, 3 CYP450 genes were identified to be potentially involved in terpenoid biosynthesis. Moreover, we found that 15 CYP450 genes were possibly regulated by antisense transcripts (r≥0.9 or r≤–0.9). Lastly, a web resource (SMCYP450, http://www.herbalgenomics.org/samicyp450) was set up, which allows users to browse, search, retrieve and compare CYP450 genes and can serve as a centralized resource.

microRNAs (miRNAs) play vital regulatory roles in many organisms through direct cleavage of transcripts, translational repression, or chromatin modification. Identification of miRNAs has been carried out in various plant species. However, no information is available for miRNAs from Panax ginseng, an economically significant medicinal plant species. Using the next generation high-throughput sequencing technology, we obtained 13,326,328 small RNA reads from the roots, stems, leaves and flowers of P. ginseng. Analysis of these small RNAs revealed the existence of a large, diverse and highly complicated small RNA population in P. ginseng. We identified 73 conserved miRNAs, which could be grouped into 33 families, and 28 non-conserved ones belonging to 9 families. Characterization of P. ginseng miRNA precursors revealed many features, such as production of two miRNAs from distinct regions of a precursor, clusters of two precursors in a transcript, and generation of miRNAs from both sense and antisense transcripts. It suggests the complexity of miRNA production in P. gingseng. Using a computational approach, we predicted for the conserved and non-conserved miRNA families 99 and 31 target genes, respectively, of which eight were experimentally validated. Among all predicted targets, only about 20% are conserved among various plant species, whereas the others appear to be non-conserved, indicating the diversity of miRNA functions. Consistently, many miRNAs exhibited tissue-specific expression patterns. Moreover, we identified five dehydration- and ten heat-responsive miRNAs and found the existence of a crosstalk among some of the stress-responsive miRNAs. Our results provide the first clue to the elucidation of miRNA functions in P. ginseng.

Passion fruit (Passiflora edulis Sims) is an economically valuable fruit that is cultivated in tropical and subtropical regions of the world. Here, we report an ~1341.7 Mb chromosome-scale genome assembly of passion fruit, with 98.91% (~1327.18 Mb) of the assembly assigned to nine pseudochromosomes. The genome includes 23,171 protein-coding genes, and most of the assembled sequences are repetitive sequences, with long-terminal repeats (LTRs) being the most abundant. Phylogenetic analysis revealed that passion fruit diverged after Brassicaceae and before Euphorbiaceae. Ks analysis showed that two whole-genome duplication events occurred in passion fruit at 65 MYA and 12 MYA, which may have contributed to its large genome size. An integrated analysis of genomic, transcriptomic, and metabolomic data showed that ‘alpha-linolenic acid metabolism’, ‘metabolic pathways’, and ‘secondary metabolic pathways’ were the main pathways involved in the synthesis of important volatile organic compounds (VOCs) in passion fruit, and this analysis identified some candidate genes, including GDP-fucose Transporter 1-like, Tetratricopeptide repeat protein 33, protein NETWORKED 4B isoform X1, and Golgin Subfamily A member 6-like protein 22. In addition, we identified 13 important gene families in fatty acid pathways and eight important gene families in terpene pathways. Gene family analysis showed that the ACX, ADH, ALDH, and HPL gene families, especially ACX13/14/15/20, ADH13/26/33, ALDH1/4/21, and HPL4/6, were the key genes for ester synthesis, while the TPS gene family, especially PeTPS2/3/4/24, was the key gene family for terpene synthesis. This work provides insights into genome evolution and flavor trait biology and offers valuable resources for the improved cultivation of passion fruit.

Background It is becoming increasingly clear that cancers can rarely be ascribed to just one or a few genomic variations. Genes generally do not function alone, but in groups that function as “networks”. This study aimed to develop a competing endogenous RNA (ceRNA) network to elucidate the role of long non-coding RNA H19 in colorectal cancer. Methods Large-scale RNA-seq data was obtained from The Cancer Genome Atlas database. Differentially expressed RNAs were identified by bioinformatics analysis, and a competing endogenous RNA network was constructed. Functional enrichment analysis and correlation analysis between competing endogenous RNAs and clinical features were performed to reveal their roles in the tumorigenesis of colorectal cancer. To verify the conclusions derived from bioinformatics analysis, we investigated the effect of lncRNA H19 knockdown in human colorectal cancer cell lines HT-29 and HCT116. Results The present study successfully identify various cancer-specific lncRNAs and pseudogenes in CRC. The lncRNA/pseudogene–miRNA–mRNA ceRNA network was constructed using 10 lncRNAs, 5 pseudogenes, 122 mRNAs and 39 miRNAs. In the ceRNA network of CRC, H19 up-regulates various cancer-related mRNA by competitively sponging various miRNA, and participates in PI3K–Akt signaling pathway in this manner. Cox regression and correlation analysis showed that H19 and some other competing endogenous RNAs in the network are associated with poor prognosis and clinical parameters such as tumor grade and metastasis. Knockdown of H19 reduces the protein level of MET, ZEB1, and COL1A1 in vitro. Conclusions H19 regulates PI3K–Akt signal pathway through a competing endogenous RNA network and predicts poor prognosis in colorectal cancer. The pseudogene RPLP0P2 may be an important oncogene like H19 and needs to be studied further. Electronic supplementary material The online version of this article (10.1186/s12935-019-0866-2) contains supplementary material, which is available to authorized users.

No abstract