Anthocyanin identification from the peels of two different materials
Color mutants are widely used in horticultural and other crops, especially those that are commonly propagated vegetatively, such as most fruit trees [38, 39]. Purple color in the flower petals of alfalfa (M. sativa L., M. falcata L. and their hybrids) is due to the presence of sap-soluble anthocyanins [40]. The floral anthocyanins of alfalfa have been widely studied. Lesins [41] identified alfalfa flower with three pigments as glycosides of petunidin, malvidin and delphinidin. Furthermore, Cooper and Elliott [42] identified alfalfa flower with three anthocyanins as 3,5-diglucosides of petunidin, malvidin and delphinidin. Differently, using HPLC, we only found that malvidin 3-O-glucoside and petunidin 3-O-glucoside in the purple flower of C, while no color pigment were detectable in the cream flowers of M (Fig. 1). The results suggest that the drastic differences in anthocyanin accumulation are a result of cultivar and genetic specificity.
PacBio full-length sequencing extends the alfalfa annotation and increases the accuracy of transcript quantification
Due to technical limitations, the reference genome of alfalfa is not presently available. Our current knowledge on the alfalfa transcriptome is mainly based on RNA-Seq gene expression data. Thus, the alfalfa transcriptome has not been fully characterized due to the lack of full-length cDNA. In this work, we used PacBio third-generation technology to annotate the sequences of the C cultivar, and analyzed the DEGs in different flower development stages of C and M using Illumina sequencing platform. We obtained 140,995 isoforms, including 513 novel isoforms. After comparation in Swiss-Prot, 204 new isoforms specific to alfalfa, but with unknown functions, were identified and will be useful in future studies (Supplementary Table S2). In transcriptome studies of populus, maize, and sorghum by single-molecule long-read sequencing, 59,977 (69%), 62,547 (57%) and 11,342 (41%) new isoforms were identified, respectively [33]. Due to species divergence, we only identified 23% new isoforms. However, our data demonstrated that PacBio full-length sequencing could provide a more comprehensive set of isoforms than next-generation sequencing.
Through a genome-based reconstruction strategy, using the Medicago genome (M. truncatula Mt4.0v2) as a reference, the mapping ratio of the corrected isoforms by PacBio full-length sequencing was 98.52%. Unfortunately, the mapping ratio of the clean reads by RNA-Seq was less than 50% (data not shown). We also compared the match ratio of the isoforms and contigs, from which we found that 99% of the isoforms (33,518) could be matched to known unigenes, indicating that the results of the long-read RNA sequencing were more integrated and accurate.
Comparison of the genes related to the biosynthesis of flavonoids in different alfalfa materials
Flavonoids are among the most important pigments in the petals of many plants [22, 43]. Anthocyanins are end products of the flavonoid biosynthetic pathway, and produce the widest spectrum of colors, ranging from pale yellow to blue-purple [44]. Our results demonstrated that the colour difference between the purple and cream flowers of alfalfa is due to the loss of the flower anthocyanins malvidin and petunidin (Fig. 1). The shift from purple to cream requires a blockage of the anthocyanin biosynthetic pathway, which probably occurs in some reactions before malvidin and petunidin are formed. Therefore, the abundance of the candidate genes was compared in the C and M transcriptomes to identify the key genes of cream color metabolism. Most of the isoforms related to flavonoids synthesis, including PALs, 4CLs, CHIs, DFRs, ANSs and UFGTs, showed large-scale higher transcription expression in C with purple flowers than in M with cream flowers, particularly S1–S3 (Fig. 2), indicating that the mutation-induced change in expression by these genes might occur far earlier than the emergence of the phenotype. It is generally known that CHS catalyzes the first reaction for anthocyanin biosynthesis and helps to form the intermediate chalcone, the primary precursor for all classes of flavonoids [45]. So if CHS reactions are strongly constrained, not only anthocyanin production but also that of nearly all other flavonoids is effectively eliminated [46]. The mutation of a single CHS enzyme led to white flower lines in grape hyacinth [28], petunia [47], Silene littorea [30] and arctic mustard flower [48]. Conversely, in our study, we found that CHSs showed higher expression in M-S4 than C-S4 (Fig. 10). Interestingly, coumaroyl-CoA can be transformed into isoliquiritigenin (an important product for the isoflavone biosynthesis pathway) by the co-function of CHS and CHR [49, 50]. Upon further data analysis, we found that the expression patterns of CHRs were similar to CHSs (Fig. 10). We thus speculated that the higher abundance of CHSs participated in another branching point in flavonoid biosynthesis, being the intermediates in the production of isoflavone biosynthesis, and CHS and CHR in M-S4 might be crucial for the biosynthesis of isoliquiritigenin.
F3H, F3’H and F3’5’H play critical roles in the flavonoid biosynthetic pathway, catalyzing the hydroxylation of flavonoids, including dihydrokaempferol, dihydroquercetin, and dihydromyricetin, which are necessary for anthocyanin biosynthesis [28, 51]. Additionally, the dihydroflavonols represent a branching point in flavonoid biosynthesis, being the intermediates in the production of both the coloured anthocyanins through DFR, and the colourless flavonols through FLS [52]. As a result of the competition for substrate (dihydroflavonols), the up-regulation of FLS and flavonols might be closely accompanied by a decrease in DFR and anthocyanin accumulation [52]. In our study, much higher expression of FLS/F3Hs, F3’Hs, and F3’5’H was found in most stages of M than C. This was accompanied with the higher expression of DFR in C, but at a very low level from S2–S4 of M (Fig. 10). A similar observation was made in grape hyacinth by Lou et al. [28], who concluded that DFR might be the target gene for the loss of blue pigmentation (delphinidin) in white grape hyacinth. Thus, the higher expression of FLS/F3Hs, F3’Hs, and F3’5’H might increase the production of other flavonoid compounds, such as dihydroquercetin, dihydrokaempferol, dihydromyricetin, myricetin and kaempferol in M, and the down-regulated DFR might partially block the synthesis of anthocyanins, thereby eliminating the process of purple pigmentation.
The purple flower ripening of C suggests that the fundamental transcriptional regulation of the flavonoid and pigment biosynthetic pathways, from the upstream PAL to the end gene UFGT, could be a major factor in the mutation, coordinating gene expression, flower coloration, and the accumulation of flavonoid intermediates.
Hub genes related to flower formation were identified by WGCNA
The cream-colored Zhongtian No. 3 alfalfa represents a color mutation, as the purple Defu alfalfa is the wild-type. Understanding the changes in the cream flower phenotype as a mutant of the wild-type could elucidate the mechanisms of the alfalfa flower pigmentation. Any functional loss of key enzymes in the flavonoid biosynthetic pathway could lead to a cream color mutation, including via transcript abundance changes in genes, and branching changes in flavone products [53, 54]. A novel finding from this study is that, by performing WGNCA, we identified floral developmental stage-specific gene modules (Figs. 8 and 9). To this end, 9 PALs, 2 4CLs, 4 CHSs, 3 CHRs, F3’H4, 2 DFRs, and 2 UFGTs were highly associated in modules with close relationships to the M4 or M group. They all possessed evident differences in transcript abundance in C and M, indicating their important roles in floral formation variation. It is worth noting that, the above genes were not the genes with the highest expression levels, implying that the high expression of genes was not necessary for distinguishing different flower colors [29]. Thus, the WGCNA analyses in this study provided a useful approach for selecting important genes related to the specific phenotypes. Du et al. [55] identified hub genes operating in the seed coat network in the early seed maturation stage by WGCNA analysis. Similar WGCNA analysis was used in golden camellia to identify unigenes correlated with flower color, and CHS, F3H, ANS and FLS were found to play critical roles in regulating the formation of flavonols and anthocyanidins [29].
The 6 hub genes were upstream of the flavonoid biosynthesis pathway, implying that the cream flower pigmentation of M was mainly blocked upstream. The decreased expression of PAL6, PAL9, and 4CL8, whether in C or M, is in line with the results in fig [54]. Wang et al. [54] found that the decreased expression of PALs and 4CLs affected the cinnamic acid content in the “Purple Peel” mature fruit peel. We speculated that the decreased expression of PAL6, PAL9, and 4CL8 might also affect the cinnamic acid content in the petals both in C and M. The elevated expression of CHSs in M-S4 might play crucial roles in the biosynthesis of other flavones, such as isoflavone, which is also a crucial factor in the color formation of different flowers in alfalfa.
Based on the above results, different flavonoid biosynthesis pathways in purple- and cream-colored alfalfa were inferred (Fig. 11). Briefly, compared to C, the flavonoid biosynthesis of M is blocked upstream, by PAL and 4CL, following which a branch of isoflavone biosynthesis regulated by CHS and CHR is dominant, completing the anthocyanin synthesis pathway. Additionally, the up-regulation of F3H/FLS, F3’H, and F3’5’H causes an increase in other flavonoid compounds, such as myricetin and kaempferol, further reducing anthocyanin synthesis. Finally, the low expression level of DFR accompanied with the low abundance of UFGT might disrupt the anthocyanin synthesis, leading to the formation of the cream color.