Integrated analysis of transcriptomic and metabolomic data reveals critical metabolic pathways involved in rotenoid biosynthesis in the medicinal plant Mirabilis himalaica

Abstract

Mirabilis himalaica (Edgew.) Heimerl is among the most important genuine medicinal plants in Tibet. However, the biosynthesis mechanisms of the active compounds in this species are unclear, severely limiting its application. To clarify the molecular biosynthesis mechanism of the key representative active compounds, specifically rotenoid, which is of special medicinal value for M. himalaica, RNA sequencing and TOF-MS technologies were used to construct transcriptomic and metabolomic libraries from the roots, stems, and leaves of M. himalaica plants collected from their natural habitat. As a result, each of the transcriptomic libraries from the different tissues was sequenced, generating more than 10 Gb of clean data ultimately assembled into 147,142 unigenes. In the three tissues, metabolomic analysis identified 522 candidate compounds, of which 170 metabolites involved in 114 metabolic pathways were mapped to the KEGG. Of these genes, 61 encoding enzymes were identified to function at key steps of the pathways related to rotenoid biosynthesis, where 14 intermediate metabolites were also located. An integrated analysis of metabolic and transcriptomic data revealed that most of the intermediate metabolites and enzymes related to rotenoid biosynthesis were synthesized in the roots, stems and leaves of M. himalaica, which suggested that the use of non-medicinal tissues to extract compounds was feasible. In addition, the CHS and CHI genes were found to play important roles in rotenoid biosynthesis, especially, since CHS might be an important rate-limiting enzyme. This study provides a hypothetical basis for the screening of new active metabolites and the metabolic engineering of rotenoid in M. himalaica.

Keywords: Metabolome; Mirabilis himalaica (Edgew.) Heimerl; Rotenoid metabolic pathway; Tibetan Plateau; Transcriptome.

Fig. 1

The typical ecological distribution of M. himalaica in Tibet. a The distribution area of M. himalaica in Tibet and sampling location. b Tibetan medicinal plant resource survey areas. c–h Collection of M. himalaica tissues for transcriptomic and metabolomic analysis and experimental procedures. The red dots represent the county-level distribution of M. himalaica, and the coordinates represent the geographic location of the experimental region of this study. The blue lines in b represent the recorded route during the resource survey

Fig. 2

KEGG functional annotation of M. himalaica transcriptome assembly. a Functional annotation of M. himalaica transcriptome assembly in the KEGG pathways. b The top 30 metabolism class pathways are listed based on the number of assigned unigenes

Fig. 3

Statistics and analysis of DEGs among the different tissues in M. himalaica. a The numbers of DEGs among the different tissues. b Heatmap analysis of the DEGs among the different tissues. The analysis was based on the Nr, GO, and KEGG annotations for genes that had a |Fold change| > 2, an FDR ≤ 0.001, and at least one sample with an RPKM > 20

Fig. 4

Top 20 enrichment pathways of DEGs, as analysed via pairwise comparisons between the different tissues

Fig. 5

Different accumulation and expression patterns of metabolites and enzymes related to rotenoid biosynthesis in different tissues. a Differential accumulation of metabolites among the different tissues. The three red circles in a represent the different tissues, in which R represents the roots, S represents the stems and L represents the leaves. The colour scale represents the multiple of the relative content of each metabolite in the tissues compared to that in the roots. b Differential expression of the genes of key catalytic enzymes among the different tissues. The colour scale represents the transformed log₁₀ (RPKM + 1) value of the unigenes in different tissues

Fig. 6

qRT-PCR verification of the genes of key enzymes in the phenylalanine metabolic pathway of M. himalaica