Differential expressions of anthocyanin synthesis genes underlie flower color divergence in a sympatric Rhododendron sanguineum complex

BACKGROUND: The Rhododendron sanguineum complex is endemic to alpine mountains of northwest Yunnan and southeast Tibet of China. Varieties in this complex exhibit distinct flower colors even at the bud stage. However, the underlying molecular regulations for the flower color variation have not been well characterized. Here, we investigated this via measuring flower reflectance profiles and comparative transcriptome analyses on three coexisting varieties of the R. sanguineum complex, with yellow flush pink, bright crimson, and deep blackish crimson flowers respectively. We compared the expression levels of differentially-expressed-genes (DEGs) of the anthocyanin / flavonoid biosynthesis pathway using RNA-seq and qRT-PCR data. We performed clustering analysis based on transcriptome-derived Single Nucleotide Polymorphisms (SNPs) data, and finally analyzed the promoter architecture of DEGs. RESULTS: Reflectance spectra of the three color morphs varied distinctively in the range between 400 and 700 nm, with distinct differences in saturation, brightness, hue, and saturation/hue ratio, an indirect measurement of anthocyanin content. We identified 15,164 orthogroups that were shared among the three varieties. The SNP clustering analysis indicated that the varieties were not monophyletic. A total of 40 paralogous genes encoding 12 enzymes contributed to the flower color polymorphism. These anthocyanin biosynthesis-related genes were associated with synthesis, modification and transportation properties (RsCHS, RsCHI, RsF3H, RsF3′H, RsFLS, RsANS, RsAT, RsOMT, RsGST), as well as genes involved in catabolism and degradation (RsBGLU, RsPER, RsCAD). Variations in sequence and cis-acting elements of these genes might correlate with the anthocyanin accumulation, thus may contribute to the divergence of flower color in the R. sanguineum complex. CONCLUSIONS: Our results suggested that the varieties are very closely related and flower color variations in the R. sanguineum complex correlate tightly with the differential expression levels of genes involved in the anabolic and catabolic synthesis network of anthocyanin. Our study provides a scenario involving intricate relationships between genetic mechanisms for floral coloration accompanied by gene flow among the varieties that may represent an early case of pollinator-mediated incipient sympatric speciation. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12870-021-02977-9.