Enzyme Repertoires and Genomic Insights into Lycium barbarum Pectin Polysaccharide Biosynthesis

Abstract

Lycium barbarum, a member of the Solanaceae family, is an important eudicot with applications in both food and medicine. L. barbarum pectin polysaccharides (LBPPs) are key bioactive compounds of L. barbarum, notable for being among the few polysaccharides with both biocompatibility and biomedical activity. Although studies have analyzed the functional properties of LBPPs, the mechanisms underlying their biosynthesis and transport by key enzymes remain poorly understood. In this study, we assembled a 2.18-Gb reference genome of L. barbarum, reconstructed the first complete biosynthesis pathway of LBPPs, and elucidated the sugar transport system. We also characterized the important genes responsible for backbone extension, sidechain synthesis, and modification of LBPPs. Furthermore, we characterized the long non-coding RNAs (lncRNAs) associated with polysaccharide metabolism. We identified a specific rhamnogalacturonan I (RG-I) rhamnosyltransferase, RRT3020, which enhances RG-I biosynthesis within LBPPs. These newly identified enzymes and pivotal genes endow L. barbarum with unique pectin biosynthesis capabilities, distinguishing it from other Solanaceae species. Our findings thus provide a foundation for evolutionary studies and molecular breeding to expand the diverse applications of L. barbarum.

Keywords: Lycium barbarum; Lycium barbarum pectin polysaccharide; Phylogenetic expansion; Rhamnogalacturonan I rhamnosyltransferase; lncRNA.

Figure 1

Evolutionary and comparative genomic analyses A. Phylogenetic tree depicting the evolutionary relationships and divergence time of 12 species. The numbers in red and green on the branches indicate the numbers of expanded (+) and contracted (−) gene families, respectively, in each species during evolution. The numbers on the branch nodes indicate divergence time. The orange star represents the WGT event that occurred in Solanaceae species. B. Distribution of genes in species-specific gene families (orange) and shared gene families (blue) as well as unassigned genes (gray). C. Venn diagram illustrating unique and shared gene families among S. melongena, L. barbarum, N. sylvestris, S. lycopersicum, and S. tuberosum. D. Intergenomic syntenic analysis between L. barbarum and S. lycopersicum, S. melongena, and V. vinifera. Syntenic pairs of genomic blocks are linked by gray lines. Red lines mark a representative syntenic block with one copy in V. vinifera and three copies in L. barbarum. E. Density distribution of synonymous substitutions per synonymous site (Ks) for paralogous genes based on the screened collinear regions among L. barbarum, other Solanaceae species (N. tabacum, S. lycopersicum, S. tuberosum, and S. melongena), and non-Solanaceae model plants (V. vinifera, A. majus, and A. thaliana). Only blocks with > 10 genes were retained. The synonymous substitution rate per gene (Ks) between each pair of species is shown in the distribution curves. A. majus, Antirrhinum majus; P. axillaris, Petunia axillaris; N. sylvestris, Nicotiana sylvestris; S. melongena, Solanum melongena; S. tuberosum, Solanum tuberosum; S. lycopersicum, Solanum lycopersicum; L. barbarum, Lycium barbarum; P. mume, Prunus mume; C. papaya, Carica papaya; A. thaliana, Arabidopsis thaliana; V. vinifera, Vitis vinifera; O. sativa, Oryza sativa; N. tabacum, Nicotiana tabacum; WGT, whole-genome triplication.

Figure 2

Expansion and fruit-specific expression of sugar transporter genes in L. barbarum A. Maximum likelihood phylogenetic tree of SWEET genes from A. thaliana, S. lycopersicum, S. tuberosum, and L. barbarum. The bootstrap confidences are labeled on the branch nodes. Tomato SlSWEET15 and its orthologs are highlighted with red branches. B. Heatmap showing the higher expression of LySWEET and LyGH32 genes in mature (red) fruits. C. Dot plot showing the differential expression of LySWEET genes in red fruits compared to the average level in all samples. Dots in red and pink denote significantly up-regulated genes and other genes in the family, respectively. D. qRT-PCR validation of top differential expression of LySWEET genes. Data are represented by mean ± SE (n = 3). E. Dot plot showing the differential expression of LyGH32 genes in red fruits compared to the average level in all samples. Dots in red and pink denote significantly up-regulated genes and other genes in the family, respectively. F. qRT-PCR validation of top differential expression of LyGH32 genes. Data are represented by mean ± SE (n = 3). SWEET, sugar will eventually be exported transporter; RFt, red fruit; GFt, green fruit; Rt, root; Fr, flower; Sm, stem; Lf, leaf; SE, standard error; FC, fold change; FDR, false discovery rate.

Figure 3

Expansion of CAZme gene family involved in pectin biosynthesis in L. barbarum A. Expansion of major CAZyme gene families in L. barbarum compared with those in S. lycopersicum, S. melongena, and A. thaliana. B. WGCNA of expanded CAZyme genes in L. barbarum shows strong correlations among cell wall-related CAZyme genes in the families GT106 (glycosyltransferase), CE13 (carbohydrate esterase), GH32 (glycoside hydrolase), GT2, and GT8. Edge colors indicate correlation strength. CAZyme, carbohydrate-active enzyme; WGCNA, weighted gene coexpression network analysis.

Figure 4

Phylogenetic repertoires for pectin biosynthesis and remodeling A. Proposed pectin biosynthesis pathway in L. barbarum by integrating genomic and transcriptomic data with the monosaccharide metabolism pathways from the KEGG pathway annotations. Essential enzymes are labeled: α-1,4-galacturonosyltransferase (GAUT), RG-I:rhamnosyltransferase (RRT), arabinan arabinosyltransferase (ARAD), β-1,3-galactosyltransferase (GALT), glycerate 2-kinase (XK), phosphoglucomutase (PGMP), hexokinase (HK), xylose isomerase (XYLA), phosphomannomutase (PMM), mannose-1-phosphate guanylyltransferase 1 (CYT1), GDP-mannose 3,5-epimerase (GME), UTP-glucose-1-phosphate uridylyltransferase (UGP), UDP-glucose 6-dehydrogenase (UGD), UDP-glucuronate 4-epimerase (GAE), trifunctional UDP-glucose 4,6-dehydratase/UDP-4-keto-6-deoxy-D-glucose 3,5-epimerase/UDP-4-keto-L-rhamnose-reductase (RHM), UDP-glucuronic acid decarboxylase (UXS), UDP-glucose 4-epimerase (UGE). The expression level (FPKM) of each gene was log₁₀-transformed and normalized to Z score in six organs. B. qRT-PCR validation of gene expression for several key enzymes involved in pectin biosynthesis. Data are represented by mean ± SE (n = 3). RG-I, rhamnogalacturonan I; FPKM, fragments per kilobase of transcript per million mapped reads.

Figure 5

Pectin metabolism-related lncRNAs in L. barbarum A. Classification of lncRNAs identified in L. barbarum. B. Venn diagram showing the overlap of up-regulated lncRNAs among root, stem, flower, leaf, and red fruit tissues in L. barbarum. C. Number of L. barbarum lncRNAs aligned to genomes of related species via a BLAST search as well as unaligned lncRNAs. D. Ten most correlated lncRNAs for each of the selected RRT genes in the WGCNA network. Red indicates up-regulated in immature (green) fruit; blue indicates down-regulated in immature (green) fruit. E. Ks distribution of all lncRNAs and the top 5 lncRNAs correlated to each CAZyme gene. The dashed line indicates the position of the WGT event. lncRNA, long non-coding RNA.

Figure 6

Novel RRTs increase RG-I biosynthesis in L. barbarum A. Evolution and protein structures of RRT genes in Solanaceae. A maximum likelihood phylogenetic tree was constructed for clade RRT1–4 (GT106) genes from A. thaliana and three Solanaceae species. The bootstrap confidences are labeled on the branch nodes. Significantly conserved motifs were identified de novo from the list of protein sequences using MEME [47]. B. Immunofluorescence images in RRT3020-expressing BY-2 cells with eGFP (green) and DAPI (blue) showing RRT3020 localization and nucleus, respectively. This experiment was repeated three times with similar results. Scale bar, 200 µm. C. Whole-mount immunolabeling assays on BY-2 cells expressing RRT3020 as well as BY-2 cells transformed with empty vector (control). RG-I accumulation was detected by CCRC-M35 antibody and cellulose was stained by Calcofluor white. Two independent experiments were performed. Scale bar, 200 µm. eGFP, enhanced green fluorescent protein; DAPI, 4′,6-diamidino-2-phenylindole; BF, bright field.

Figure 7

Schematic illustration of pectin biosynthesis and sugar transport in L. barbarum VIN, vacuolar invertase; CIN, cytoplasmic invertase; CWIN, cell wall invertase; LBPP, Lycium barbarum pectin polysaccharide.