Integrated Metabolome and Transcriptome Analysis Reveals the Mechanism of Anthocyanin Biosynthesis in Pisum sativum L. with Different Pod Colors

Abstract

Pea (Pisum sativum L.) is a significant source of dietary protein, starch, fiber, and minerals, offering health benefits and serving as both a green vegetable and dry grain. The pigment contents in pea pods with different colors and related genes are still unclear. We conducted an integrated transcriptome and metabolome analysis on three cultivars, including QiZhen (QZ) with green immature pods, FengMi (FM) with yellow immature pods, and ZiYu (ZY) with purple immature pods, to identify the key genes and metabolites involved in anthocyanin accumulation. ZY showed the highest total anthocyanin content compared with FM and QZ. Subsequent quantification revealed that four metabolites, including Delphinidin-3-O-galactoside, Delphinidin-3-O-(6″-O-xylosyl)glucoside, Cyanidin-3-O-galactoside, and Pelargonidin-3-O-(xylosyl)glucoside, were the most highly accumulated in the ZY cultivar, suggesting their role in the purple pigmentation of ZY pea pods. There were 49 differentially accumulated anthocyanidins in ZY vs. FM, 43 differentially accumulated anthocyanidins in ZY vs. QZ, and 21 differentially accumulated anthocyanidins in FM vs. QZ. These findings highlight the importance of the type and concentration of anthocyanin compounds, especially those based on delphinidin, cyanidin, and pelargonidin, in the development of purple pea pods. The transcriptomic analysis revealed that certain anthocyanin biosynthetic genes were expressed at higher levels in ZY than in FM and QZ. In ZY, the higher expression levels of five key genes (PAL, 4CL, CHS, F3H, and UFGT) resulted in elevated anthocyanin content compared to FM and QZ. Furthermore, the BSA-seq analysis identified a candidate region associated with purple color in pea pods, which is located on chromosome 6 and contains 21 DEGs. Sequence variation in KIW84_061698, which encodes a bHLH transcription factor, was identified as the key candidate gene controlling anthocyanin content. This study clarifies the molecular mechanisms behind pea pod coloration and identifies potential genetic engineering targets for breeding anthocyanin-rich sugar snap peas.

1. Introduction

Pea (Pisum sativum L.) is an essential nutritional crop for sustainable agriculture and food security, and it serves as a prominent genetic model in Mendelian genetics [1]. Pea, a key cool-season legume, is cultivated extensively across the globe. In 2023, over 13.7 Mt of dry peas and 21.4 Mt of green peas were produced globally [2]. Dried peas are mainly used in the feed and food industries, whereas fresh and crispy peas are employed in the canned and frozen food sectors [3]. Pea pods contain a high amount of dietary fiber, protein, calcium, iron, and phenolic antioxidants [4,5]. Research indicates that legume seed development is genetically programmed and associated with alterations in metabolite levels [6,7]. Although it is recognized, molecular genetic research on pea has been limited due to its large genome size (4.45 Gb) and high proportion of repeat sequences [8]. The availability of pea genome sequences offers a crucial molecular foundation for investigating key agronomic traits and supporting crop improvement [9]. Mendel’s study of seven pea agronomic traits led to the formulation of the laws of segregation and independent assortment. To date, seven loci including A (seed coat/flower color), LE (stem length), I (cotyledon color), R (seed shape), pod color (GP), pod form (V/P), and the position of flowers (FA) have been cloned and characterized at the molecular level in detail [10,11,12]. In previous studies, the chlorophyll content of pod mesocarp was decreased in yellow-podded (gp) compared with wild-type green pods [13]. GP was mapped to chromosome 3, and the 3′-exoribonuclease genes were identified as potential candidate genes controlling the yellow pod color [14]. Recently, a study focused on uncovering the gene identities and genomic context of alleles underlying the seven Mendel’s agronomic characteristics was completed [12]. A 100 kb deletion caused an intergenic transcriptional fusion that disrupted ChlG (Chlorophyll synthase) function and chlorophyll synthesis, leading to the yellow pod phenotype [12]. In fact, immature pea pods have a third color, purple, but few studies have focused on elucidating the biochemical content change and molecular regulator mechanism that result in this significant trait.

Anthocyanins are prevalent water-soluble pigments and contribute to colors such as orange, red, purple, and blue for distinctive tissues and organs in plants [15,16,17]. Anthocyanins are secondary metabolites originating from the phenylpropanoid and flavonoid pathway [18]. These compounds are produced in the cytosol and moved to vacuoles for storage [19]. Anthocyanins, known for their strong antioxidant properties, aid plants in combating biotic and abiotic stresses, including UV-B radiation [20], salinity [21], cold, and drought [22]. Moreover, anthocyanins have demonstrated potential benefits for humans by mitigating damage from viruses and certain chronic diseases [23,24,25]. Due to their biochemical properties, anthocyanins are crucial in crop breeding and plant evolution [26,27]. The current research has extensively explored the synthesis pathways of anthocyanins in plants like Arabidopsis [28], grape [29], apple [30], pear [30], blueberry, and bilberry [31]. Anthocyanin synthesis involves phenylalanine and malonyl-coenzyme A, catalyzed by structural genes, including CHS (chalcone synthase), CHI (chalcone isomerase), F3H (flavanone 3-hydroxylase), F3′H (flavonoid 3′-hydroxylase), F3′5′H (flavonoid 3′,5′-hydroxylase), DFR (dihydroflavonol 4-reductase), ANS (anthocyanidin synthase), and UFGT (uridine diphosphate-glucose: flavonoid 3-O-glucosyltransferase) [32]. Anthocyanin biosynthesis is regulated by both structural and numerous regulatory genes. MBW (MYB-bHLH-WD40) ternary complexes, which include MYB and bHLH transcription factors associated with a WD40 repeat protein, play a key role in initiating various cellular differentiation pathways in numerous plants and are essential for regulating the anthocyanin biosynthesis pathway [33]. Additionally, anthocyanin biosynthesis is also influenced by environmental stimuli and plant hormones. Recently, transcriptomics and metabolomics have been primarily utilized to investigate plant growth processes [34,35,36,37,38]. This method has been applied in many species, such as transcriptome and metabolic analysis of anthocyanin and proanthocyanidin accumulation in red mutant pear [39].

Integrating differentially accumulated metabolites (DAMs) with differentially expressed genes (DEGs) provides a robust correlation between phenotypes and genes [40,41]. Although anthocyanin biosynthesis pathways have been investigated in several crops [42,43], the mechanism of coloration in peas is yet to be fully elucidated. A previous study performed a comparative analysis of the transcriptome and metabolome to investigate alterations in metabolite accumulation and gene expression at two pod developmental stages in two pea accessions with green pods (GP) and yellow pods (YP) [3]. They also performed a systematic analysis of the difference between two pea accessions with green pods (GPs) and purple pods (PPs). Their findings revealed that eight genes encoding enzymes (C4H, CHI, F3H, F3′H, F3′5′H, DFR, ANS, and FLS) that participated in the flavonoid synthesis pathway were significantly up-regulated in PPs; led to a substantial accumulation of flavonoid and anthocyanin metabolites [44]. However, anthocyanin-targeted metabolome analysis of different pea pods needs to be more fully explored. Moreover, the anthocyanins synthesis in pea pods is controlled by multiple genes. Using different materials could lead to the discovery of more related candidate genes and potential molecular mechanisms that regulate the anthocyanin levels of pea pods with different colors. This study systematically investigates the mechanism of anthocyanin-based color development in pea pods of QiZhen (QZ), FengMi (FM), and ZiYu (ZY) through integrated transcriptomic and metabolomic analyses. Metabolomic analyses found purple pea pods were mainly determined by four metabolites with high accumulation levels, including Delphinidin-3-O-galactoside, Delphinidin-3-O-(6″-O-xylosyl)glucoside, Cyanidin-3-O-galactoside, and Pelargonidin-3-O-(xylosyl)glucoside. Consequently, we identified six DEGs (PAL, 4CL, CHS, F3H, FLS, and UFGT) involved in the anthocyanin biosynthesis pathway in pea pods, which may regulate anthocyanin synthesis. By comprehensively analyzing the results of BSA-seq and RNA-seq, we found that KIW84_061698, which encodes a bHLH transcription factor, was the key candidate gene related to anthocyanin content. These results offer understanding for identifying metabolites and candidate genes related to anthocyanin formation and color variations in green, yellow, and purple peas, establishing a molecular basis for anthocyanin content enhancement and breeding programs in the future.

2. Materials and Methods

2.1. Plant Materials and Treatments

Our study utilized three types of pea pods (QZ, FM, and ZY) sourced from experimental fields in Hefei (31°53′24″ N, 117°14′26″ E), China. Samples were chosen from healthy plants without visible pests or diseases in April 2024. QiZhen (QZ) is an elite sugar snap pea cultivar, and its young pods are green. FengMi (FM) is also a sugar snap pea cultivar, and its young pods are yellow. The young pods of ZiYu (ZY) are dark purple and inedible. Samples of immature pods without seeds were collected from each cultivar 20 days post-flowering. The samples were preserved in liquid nitrogen for subsequent analyses. Three biological replicates were conducted, each derived from 10 plants. In order to analyze the segregation model of pod color, two crosses (FM × QZ and QZ × ZY) were constructed.

2.2. Quantification of Anthocyanin and Proanthocyanidin Content

According to the method described by Zhang et al. [45], the total anthocyanin content was assessed using hydrochloric acid and methanol. A 2 g sample was combined with 5 mL of HCl/methanol (1/99, v/v) solution, thoroughly mixed, and stored in the dark at 4 °C for over 20 h, followed by ultrasonic extraction at 4 °C for 30 min. Then, the mixture was centrifuged at 8000 rpm and 4 °C for 10 min. Absorbance was measured at 530, 620, and 650 nm. Anthocyanin content was calculated using the formula:

$$\mathrm{A}\mathrm{n}\mathrm{t}\mathrm{h}\mathrm{o}\mathrm{c}\mathrm{y}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{n}\mathrm{e}\mathrm{t}={\displaystyle \frac{\mathrm{A}}{4.62\times {10}^4}}\ast {\displaystyle \frac{\mathrm{V}}{\mathrm{W}}}\ast {10}^6$$

V means volume, and W means wight; A = (A530 − A620) − 0.1(A650 − A620). The expressed outcome as the nmol of anthocyanins per gram of sample. Following the instructions, the proanthocyanidin content was determined using the Micro Plant Proanthocyanidins Assay Kit from Solarbio Life Sciences, Beijing, China. A 0.1 g sample was extracted with 2 mL of extraction solution (60% ethanol) at 60 °C for 2 h. The supernatant was moved to a clean tube for testing after being centrifuged at 10,000 g for 10 min. A mixture of 0.2 mL supernatant and 0.8 mL of either buffer A [Solution 1 (8% HCl): Solution 2 = 1:1] or buffer B [Solution 1 (8% HCl): methanol = 1:1] was prepared, and the absorbance at 500 nm (A500) was measured for both buffers. The Micro Plant Proanthocyanidins Assay Kit from Solarbio Life Sciences in Beijing, China, was used to obtain Solution 1 and Solution 2. The standard curve is represented by y = 0.0194x + 0.0006, with an R² value of 0.999. Proanthocyanidin content (mg g fresh weight⁻¹) = 0.515 × (ΔA − 0.0006)/W. W means sample weight; ΔA = A500(test) − A500(control).

2.3. Anthocyanin-Targeted Metabolome Analysis

The sample was freeze-dried and pulverized into a powder at 30 Hz for 1.5 min, then stored at −80 °C. A 50 mg sample of this powder was extracted with 0.5 mL of a methanol, water, and hydrochloric acid solution at a 500:500:1 volume ratio. The sample was mixed using a vortex for 10 min and then centrifuged at 12,000 g at 4 °C for 3 min. The remaining material was extracted again following the same method and conditions. Before LC-MS/MS analysis, the supernatants were collected and filtered through a 0.22 μm membrane filter (Anpel, Shanghai, China).

An analysis of the sample extracts was conducted using a UPLC-ESI-MS/MS system (Sciex, Framingham, MA, USA), which includes the ExionLC™ AD UPLC and the Applied Biosystems 6500 Triple Quadrupole MS. The analysis was conducted using UPLC with a Waters ACQUITY BEH C18 (MS/MS, QTRAP®, 6500+, 1.7 µm, 2.1 mm × 100 mm). The solvent system used was water with 0.1% formic acid and methanol with 0.1% formic acid. The gradient program started at 95:5 v/v at 0 min, shifted to 50:50 v/v at 6 min, and reached 5:95 v/v at 12 min, where it was held for 2 min. Then it returned to 95:5 v/v at 14 min, where it was held for 2 min. The flow rate was 0.35 mL/min, the temperature was 40 °C, and the injection volume was 2 μL.

Scans of linear ion trap (LIT) and triple quadrupole (QQQ) were performed using a QTRAP mass spectrometer (Sciex, Shanghai, China), which combines the features of a triple quadrupole and a linear ion trap. Equipped with an ESI Turbo Ion-Spray interface (Thermo Fisher Scientific, Waltham, MA, USA), the QTRAP^® 6500+ LC-MS/MS System (Sciex, Framingham, MA, USA) operates in positive ion mode and is controlled by Sciex’s Analyst 1.6.3 software. The ESI source operated with the following parameters: the ion source was set to ESI+; the source temperature was 550 °C; the ion spray voltage (IS) was 5500 V; the curtain gas (CUR) was adjusted to 35 psi.

2.4. Metabolic Data Analysis

According to the MWDB database created by MetWare Biotechnology Co., Ltd. (Wuhan, China), secondary spectral information was used for the qualitative analysis of metabolic data. Anthocyanins were analyzed using scheduled multiple reaction monitoring (MRM). Data acquisitions were performed using Analyst software version 1.6.3.1569. All metabolites were quantified using Multiquant software version 3.0.3. Mass spectrometer parameters, including the declustering potentials (DP) and collision energies (CE) for individual MRM transitions, were carried out with further DP and CE optimization. A specific set of MRM transitions was monitored for each period according to the metabolites eluted within this period. The VIP (variable importance in projection) score was obtained using OPLS-DA (orthogonal partial least squares-discriminant analysis). Differential anthocyanin accumulation between the samples was identified using the criteria of a variable importance in projection (VIP) ≥ 1 and a fold change threshold of ≥2 or ≤0.5. The multiple test correction was performed using the false discovery rate (FDR) method.

2.5. Transcriptome Analysis

The standard RNA-seq protocol was carried out. Total RNA was extracted from 0.45 g frozen samples using the TransZol Plant Kit (Transgen, Beijing, China) according to the manufacturer’s instructions. The quality of RNA was evaluated using agarose gel electrophoresis and Qubit assays (Thermo Fisher Scientific, Waltham, MA, USA). MRNA was fragmented with divalent cations in NEB fragmentation buffer, and libraries were constructed according to the NEB standard method. The library’s concentration was measured with Qubit 2.0 and adjusted to 1.5 ng/µL. The Agilent 2100 Bioanalyzer was used to evaluate the insert size (about 450 bp). Once the expected insert size was confirmed, quantitative real-time PCR (qRT-PCR) was employed to measure the library concentration, confirming it was above 2 nM for quality assurance. Libraries were pooled according to their effective concentration and the necessary sequencing data volume for paired-end Illumina sequencing. The data undergoes strict quality control using fastp, with filtering criteria that include: (1) Remove reads containing adapters. (2) When the N content in any sequencing read exceeds 10% of the number of bases in that read, remove this paired read. (3) Remove the paired reads if over 50% of their bases have a low quality score (Q ≤ 20) in any sequencing read. All the high-quality reads were aligned to a reference sequence of Zhongwan 6 (https://peagdb.com/index/, accessed on 23 May 2024). The clean reads were sequentially aligned with the reference genome using HISAT2 (Version 2.2.1) to obtain the positional information on the reference genome, as well as the sequence characteristic information. The gene expression levels were estimated as transcripts per million (TPM). The calculation methods for the TPM value are as follows:

$$\mathrm{T}\mathrm{P}\mathrm{M}={\displaystyle \frac{\raisebox{1ex}{$\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{d}\mathrm{s}\mathrm{m}\mathrm{a}\mathrm{p}\mathrm{p}\mathrm{e}\mathrm{d}\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{c}\mathrm{r}\mathrm{i}\mathrm{p}\mathrm{t}$}\!\left/ \!\raisebox{-1ex}{$\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{c}\mathrm{r}\mathrm{i}\mathrm{p}\mathrm{t}\mathrm{l}\mathrm{e}\mathrm{n}\mathrm{g}\mathrm{t}\mathrm{h}$}\right.}{\mathrm{s}\mathrm{u}\mathrm{m}\mathrm{o}\mathrm{f}\mathrm{m}\mathrm{a}\mathrm{p}\mathrm{p}\mathrm{e}\mathrm{d}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{d}\mathrm{s}\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{c}\mathrm{r}\mathrm{i}\mathrm{p}\mathrm{t}\mathrm{n}\mathrm{o}\mathrm{r}\mathrm{m}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{z}\mathrm{e}\mathrm{d}\mathrm{b}\mathrm{y}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{c}\mathrm{r}\mathrm{i}\mathrm{p}\mathrm{t}\mathrm{l}\mathrm{e}\mathrm{n}\mathrm{g}\mathrm{t}\mathrm{h}}}\ast {10}^6$$

Unstandardized read count data of genes were input into DESeq2 version 1.22.1 to perform differential expression analysis on samples with biological replicates between the groups. In order to control the false discovery rate, the Benjamini–Hochberg method was applied to adjust p-values, identifying genes with adjusted p-values below 0.05 as differentially expressed. To identify differentially expressed genes (DEGs), a threshold of |Log₂ (fold change) | ≥ 1 was used, along with a false discovery rate and an adjusted p-value under 0.05. Identified metabolites were annotated using the KEGG compound database (http://www.kegg.jp/kegg/compound/, accessed on 2 June 2024), and annotated metabolites were then mapped to the KEGG Pathway database (http://www.kegg.jp/kegg/pathway.html, accessed on 2 June 2024). MSEA (metabolite set enrichment analysis) was used to analyze pathways with significantly regulated metabolites, with their significance evaluated via hypergeometric testing. The enrichment bar plot was generated using TBtools (Version 2.056). GOSeq (R package v.1) and KOBAS software (Version 3.0) were used for gene ontology (GO) and KEGG pathway enrichment analyses. Significance enrichment analysis of GO involves taking the GO Term in the GO database as the unit and applying the hypergeometric test to identify the GO terms that are significantly enriched in the differentially expressed genes compared with the entire genome background. Bonferroni correction was used for multiple test correction. To study the expression patterns of genes in different samples, the TPM of the union of all differentially expressed genes was first standardized using the scale function of the R language, and then K-means clustering analysis was conducted. After taking the union of the differential genes of all comparison groups, it was used as the differential gene set for hierarchical clustering analysis. The Z-score was used to standardize the data, and the clustering heat map of the differential genes in the union of all comparison groups and the clustering heat map of each differential group were drawn.

2.6. Quantitative RT-PCR Analysis

QRT-PCR was used to determine the expression levels of eight candidate genes associated with anthocyanin biosynthesis. Total RNA was extracted from the pea pods using the Plant Total RNA Isolation Kit (SK8631; Sangon Biotech, Shanghai, China). CDNA was then synthesized using the RNA with MonScript™ RTIII (MR05101, Monad, Shanghai, China). The QuantiNova SYBR PCR Mix Kit (QIAGEN, Shanghai, China) was used as the premix. The primers used in qRT-PCR were designed using a web-based tool (https://www.ncbi.nlm.nih.gov/tools/primer-blast/, accessed on 22 June 2024). Supplementary Table S1 provides details of the primers used in this study. The qRT-PCR protocol began with an initial denaturation at 95 °C for 2 min, followed by 40 cycles consisting of 95 °C for 5 s, and 60 °C for 30 s. The melting curve was drawn by continuously acquiring the fluorescence signal from 65 °C to 95 °C at a frequency of 5 reads/°C. The pea actin gene (LOC127126228) was selected as a reference gene in the study. All data were obtained from three biological repetitions. The gene expression levels were calculated using the 2^−ΔΔCt method [46].

2.7. Combined Transcriptome and Metabolome Analysis

In order to identify co-enriched pathway information, the KEGG pathway database was used to map DEGs and DAMs. GraphPad 8.0.2 was used to generate a histogram after identifying the most significantly enriched pathways for co-significant enrichment of DEMs and DEGs in each comparison group based on DEG enrichment pathways. The anthocyanin biosynthesis pathway map was created using the KEGG pathway database as a reference. Standardized TPM values of DEGs and relative quantitative values of DAMs were visualized as heat maps on the pathway map.

2.8. BSA-Seq Analysis for Candidate Genome Region

The F₂ segregation population from the QZ and ZY cross was utilized to identify candidate genes linked to pod anthocyanin content. The leaves of the F₂ population and the parental lines were sampled in April 2024 in Hefei (31°53′24″ N, 117°14′26″ E), China. Two DNA bulks (‘Green’ bulk and ‘Purple’ bulk) of progeny with extreme phenotypes were constructed using the bulked-segregant analysis (BSA) method. Genomic DNA was isolated from fresh pea leaves utilizing a DNeasy Plant Mini Kit (QIAGEN, Hilden, Germany). DNA quantification was performed using the Quant-iT^TM PicoGreen dsDNA reagent kits from Invitrogen. The DNA from 30 individuals with complete green pods was equally combined to create a ‘Green’ bulk DNA, while the DNA from individuals with dark purple pods was similarly pooled to form a ‘Purple’ bulk DNA. The bulk DNA and the two parents’ DNA were used to prepare libraries for whole-genome sequencing. The four libraries, each with a 400 bp insertion, were sequenced using the Illumina NovaSeq platform with a paired-end (2 × 150 bp) strategy. FastQC was used to assess the quality of the raw sequencing reads. All of the high-quality reads were aligned to a reference sequence of Zhongwan 6 (https://peagdb.com/index/, accessed on 12 June 2024) using BWA (Burrows–Wheeler Aligner, version 0.7.17-r1198,) software. The adapter sequences, short reads, and low-quality reads were deleted. GATK version 3.8 with default parameters [47] was used for variant calling. ΔSNP-index (sliding window analysis was applied with a 1 Mb window size and a 100 kb increment) [48] and the GradedPool-Seq with default parameters [49] method was used for BSA-seq analysis.

2.9. Statistical Analysis

The results were presented as the mean ± standard deviation (SD) of triplicate. ANOVA analysis was used to analyze the differences between the means, with a p-value of less than 0.05 indicating significance.

3. Results

3.1. Changes in Anthocyanin Content in Pea Pods of QZ, FM, and ZY Cultivars

The immature pods of QZ, FM, and ZY were dark green, yellow, and dark purple, respectively (Figure 1A). The QZ pods, lacking an internal parchment layer, were characterized by a fleshy, sweet, and crispy texture. The contents of anthocyanidin and proanthocyanidin in ZY were significantly higher than those in QZ and FM (Figure 1B,C). The anthocyanidin content in FM was significantly higher than that in QZ, but the proanthocyanidin content was significantly lower (Figure 1B,C).

3.2. Anthocyanin Metabolic Profiling in Pea Pods of Three Different Cultivars

To investigate variations in anthocyanin biosynthesis and accumulation, a targeted metabolome analysis was performed among the three cultivars. The LC-MRM-MS chromatogram is shown in Supplementary Figure S1. Differential anthocyanin accumulation between sample pairs (FM vs. QZ, ZY vs. FM, and ZY vs. QZ) was evaluated. Figure 2A illustrates a hierarchical heat map clustering analysis based on metabolite concentration data, revealing three distinct clusters with varying accumulation levels in dark green, yellow, and dark purple pea pods. As expected, the anthocyanin levels differed among the samples, with 61 anthocyanidin-related metabolites, including 16 delphinidin derivatives, 14 cyanidin derivatives, 10 pelargonidin derivatives, 7 peonidin derivatives, 4 malvidin derivatives, and 3 petunidin derivatives identified (Figure 2A). Among these, four metabolites, including Delphinidin-3-O-galactoside, Delphinidin-3-O-(6″-O-xylosyl)glucoside, Cyanidin-3-O-galactoside, and Pelargonidin-3-O-(xylosyl)glucoside, were the most highly accumulated in the ZY cultivar. ZY contained 48 kinds of anthocyanins, FM contained 27, and QZ contained 14. There were 21 differentially accumulated anthocyanins in FM vs. QZ, 49 differentially accumulated anthocyanins in ZY vs. FM, and 43 differentially accumulated anthocyanins in ZY vs. QZ (Figure 2B). A total of 18 anthocyanins showed a higher accumulation level in FM compared to QZ (Supplementary Table S2). Additionally, 37 more anthocyanins accumulated in the ZY pea pods compared to the FM pea pods, and 40 more anthocyanins showed high accumulation in the ZY pea pods compared to the QZ pea pods (Supplementary Table S2). All 37 up-regulated anthocyanins in the ZY vs. FM group also showed increased accumulation in the ZY vs. QZ group. Therefore, the variation in pod color is likely due to the high anthocyanin expression in the ZY group compared to the low levels in the FM and QZ groups.

The top 10 significantly accumulated anthocyanins among the three comparisons were identified (

Table 1. Top 10 identification of differentially abundant metabolites in the three comparisons.

Group	Index	Compounds	p-Value	Fold_Change	Type
	Anthocyanidin_400	Cyanidin-3-O-glucoside-5-O-galactoside	0.022	Inf	up
	Anthocyanidin_27	Delphinidin-3-O-glucoside	0.006	Inf	up
	Anthocyanidin_212	Delphinidin-3-O-(6″-O-xylosyl)glucoside	0.003	Inf	up
	Anthocyanidin_396	Pelargonidin-3,5-O-digalactoside	0.024	Inf	up
	Anthocyanidin_329	Peonidin-3-O-(6″-O-malonyl)diglucoside	0.003	Inf	up
FM vs. QZ	Anthocyanidin_109	Procyanidin B4	0.027	Inf	up
	Anthocyanidin_107	Procyanidin B3	0.036	Inf	up
	Anthocyanidin_06	Cyanidin-3-O-(6-O-p-coumaroyl)-glucoside	0.009	5.165	up
	Anthocyanidin_75	Peonidin	0.036	5.034	up
	Anthocyanidin_399	Delphinidin-3-O-glucoside-5-O-galactoside	0.046	3.187	up
	Anthocyanidin_10	Cyanidin-3-O-galactoside	0.006	Inf	up
	Anthocyanidin_131	Cyanidin-3-O-(6″-O-acetyl)glucoside-5-O-glucoside	0.001	Inf	up
	Anthocyanidin_223	Delphinidin-3-O-(2‴-O-p-coumaroyl)rutinoside	0.002	Inf	up
	Anthocyanidin_215	Delphinidin-3-O-(6″-O-galloy)glucoside	0.002	Inf	up
	Anthocyanidin_403	Pelargonidin-3-O-(xylosyl)glucoside	0.004	Inf	up
ZY vs. FM	Anthocyanidin_26	Delphinidin-3-O-galactoside	0.005	Inf	up
	Anthocyanidin_277	Pelargonidin-3-O-(6″-O-xylosyl)galactoside	0.006	Inf	up
	Anthocyanidin_67	Pelargonidin-3-O-galactoside	0.006	Inf	up
	Anthocyanidin_231	Delphinidin-3,5,7-triglucoside	0.007	Inf	up
	Anthocyanidin_50	Malvidin-3-O-galactoside	0.008	Inf	up
	Anthocyanidin_131	Cyanidin-3-O-(6″-O-acetyl)glucoside-5-O-glucoside	0.001	Inf	up
	Anthocyanidin_223	Delphinidin-3-O-(2‴-O-p-coumaroyl)rutinoside	0.001	Inf	up
	Anthocyanidin_212	Delphinidin-3-O-(6″-O-xylosyl)glucoside	0.002	Inf	up
	Anthocyanidin_215	Delphinidin-3-O-(6″-O-galloy)glucoside	0.002	Inf	up
ZY vs. QZ	Anthocyanidin_403	Pelargonidin-3-O-(xylosyl)glucoside	0.004	Inf	up
	Anthocyanidin_26	Delphinidin-3-O-galactoside	0.005	Inf	up
	Anthocyanidin_277	Pelargonidin-3-O-(6″-O-xylosyl)galactoside	0.006	Inf	up
	Anthocyanidin_10	Cyanidin-3-O-galactoside	0.006	Inf	up
	Anthocyanidin_67	Pelargonidin-3-O-galactoside	0.006	Inf	up
	Anthocyanidin_231	Delphinidin-3,5,7-triglucoside	0.007	Inf	up

Inf indicates infinity.

). We hypothesize that anthocyanin is the primary metabolite responsible for the color change in pea pods. Furthermore, the differentially accumulated anthocyanins in the three comparison groups were all enriched in the ‘anthocyanin biosynthesis’, ‘biosynthesis of secondary metabolites’, and ‘flavone and flavonol biosynthesis’ pathways (Figure 2C–E). In the ZY vs. FM and ZY vs. QZ comparisons, some DAMs were enriched in ‘flavonoid biosynthesis’ (Figure 2D,E). In FM vs. QZ, Cyanidin-3-O-(6-O-p-coumaroyl)-glucoside, Delphinidin-3-O-glucoside, and Malvidin-3-O-glucoside were significantly enriched in the ‘anthocyanin biosynthesis’ pathway. In ZY vs. FM, Cyanidin-3-O-(6-O-malonyl-beta-D-glucoside), Cyanidin-3-O-(6-O-p-coumaroyl)-glucoside, Delphinidin-3-O-glucoside, Malvidin-3-O-glucoside, Pelargonidin-3-O-glucoside, and Pelargonidin-3-O-rutinoside were significantly enriched in the ‘anthocyanin biosynthesis’ pathway. In ZY vs. QZ, Cyanidin-3-O-(6-O-malonyl-beta-D-glucoside), Cyanidin-3-O-(6-O-p-coumaroyl)-glucoside, Pelargonidin-3-O-glucoside, and Pelargonidin-3-O-rutinoside were significantly enriched in the ‘anthocyanin biosynthesis’ pathway.

3.3. Transcriptome Sequencing and Annotation

RNA-seq analysis was performed on RNA from the three pea cultivars to investigate variations in their gene expression. The sequencing produced 76.62 Gb of clean data, with each sample contributing 7 Gb and a Q30 base percentage of at least 96% (

Table 2. Sequencing and quality statistics.

Sample	Raw Reads	Clean Reads	Clean Base (G)	Error Rate (%)	Q20 (%)	Q30 (%)	GC Content (%)
FM-1	68,276,454	64,555,088	9.68	0.01	98.97	96.72	42.85
FM-2	57,937,582	55,187,626	8.28	0.01	99.02	96.91	42.86
FM-3	55,651,956	52,669,990	7.9	0.01	99.03	96.94	42.92
ZY-1	63,461,278	58,630,268	8.79	0.01	98.99	96.79	42.75
ZY-2	61,287,894	58,063,148	8.71	0.01	99.03	96.96	42.73
ZY-3	59,649,068	56,253,684	8.44	0.01	99.01	96.83	42.76
QZ-1	57,436,570	53,905,096	8.09	0.01	99.02	96.92	42.75
QZ-2	62,130,038	58,493,888	8.77	0.01	98.99	96.78	42.68
QZ-3	55,796,790	53,095,348	7.96	0.01	98.96	96.67	42.63

). Principal component analysis (PCA) was applied to the transcription sample expression (Figure 3A). PC1 accounted for 35.58% of the total variance among the samples, while PC2 accounted for 26.75%. Three repeats of the same cultivar were clustered together and clearly separated from the other cultivars. The score map effectively distinguished each cultivar, emphasizing transcript differences among the three cultivars, with the threshold set at |log₂Fold Change| ≥ 1 and FDR < 0.05. The DEG results for the comparison among the different groups are shown in Figure 3B,C. In the FM vs. QZ group, 3827 DEGs were identified, with 2034 up-regulated and 1793 down-regulated. The ZY vs. FM group had 5710 DEGs, comprising 3197 up-regulated and 2513 down-regulated genes. In the ZY vs. QZ group, 5325 DEGs were found, with 3319 up-regulated and 2006 down-regulated. There were 3239 DEGs identified in both the ZY vs. FM and ZY vs. QZ groups, including 2182 up-regulated and 820 down-regulated DEGs in both groups. A hierarchical clustering heat map was generated using the TPM values of the DEGs (Figure 3D). Significant differences were observed among the three cultivars, aligning with the metabolic group findings. Notably, many genes in ZY showed significant up-regulation. The H-cluster analysis grouped all DEGs into nine expression patterns, indicating high repeatability across the samples (Figure 3E). Clusters 1, 4, and 7, comprising 420, 2033, and 886 DEGs, exhibited high expression in ZY. Clusters 3, 8, and 9 contained 894, 432, and 1121 DEGs, respectively, all exhibiting high expression in FM. Clusters 2, 5, and 6 contained 651, 616, and 1188 DEGs and were highly expressed in QZ. Gene Ontology (GO) enrichment analysis revealed significant enrichment in the molecular function and biological process categories, and primarily associated with the ‘flavonoid metabolic process’ (GO: 0009812), ‘flavonoid biosynthetic process’ (GO: 0009813), and ‘UDP−glycosyltransferase activity’(GO:0008194) (Figure 4).

3.4. QRT-PCR Verification

Subsequently, we conducted random qRT-PCR analysis on selected candidate genes to validate the transcriptome data, as shown in Figure 5. We confirmed that the expression patterns of KIW84_010411, KIW84_015661, KIW84_015663, KIW84_072663, KIW84_061556, KIW84_032743, KIW84_052804, and KIW84_052801 aligned with their TPM expression levels. Moreover, the correlation coefficient (y = 0.5275x + 0.8825, R2 = 0.8441, Supplementary Figure S2) validated that the RNA-seq data were duplicable and reliable.

3.5. KEGG Enrichment and Anthocyanin Biosynthesis Pathway Analysis

We plotted the top 10 KEGG pathways co-enriched with DEGs and DAMs in each control group, displaying 16 enriched pathways (Figure 6A–C). ‘Metabolic pathways’, ‘anthocyanin biosynthesis’, ‘biosynthesis of secondary metabolites’, and ‘flavone and flavonol biosynthesis’ were all enriched in the FM vs. QZ (Figure 6A), ZY vs. FM (Figure 6B), and ZY vs. QZ (Figure 6C) groups. In the ZY vs. FM and ZY vs. QZ comparisons, some DEGs and DAMs were specifically co-enriched in ‘flavonoid biosynthesis’ and ‘isoflavonoid biosynthesis’ (Figure 6B,C). In the FM vs. QZ group, DEGs and DAMs that were enriched in the ‘anthocyanin biosynthesis’ and ‘flavone and flavonol biosynthesis’ pathways showed high rich factors. In the ZY vs. FM group, DEGs enriched in ‘anthocyanin biosynthesis’ had the highest rich factor. In the ZY vs. QZ group, DEGs enriched in ‘isoflavonoid biosynthesis’, ‘flavone and flavonol biosynthesis’, and ‘anthocyanin biosynthesis’ showed higher rich factors. Based on these results, we speculated that ‘anthocyanin biosynthesis’ and ‘flavonoid biosynthesis’ play crucial roles in the mechanism of anthocyanin accumulation in pea pods. To elucidate the regulatory network of anthocyanin biosynthesis in pea pods, a network diagram was used to depict the correlation between metabolites and genes. DEGs and DAMs with an absolute Pearson correlation coefficient exceeding 0.8 and a p-value below 0.05 in each pathway were chosen for mapping. In FM vs. QZ, three significant correlations were identified between Quercetin-3-O-glucoside and certain genes (Figure 6D). There were 25 significant correlations found between Naringenin and certain genes in ZY vs. QZ and ZY vs. FM (Figure 6E,F). Naringenin was the key intermediate product in the anthocyanin biosynthesis pathway, and four CHS genes (KIW84_025518, KIW84_052803, KIW84_025517, and KIW84_025227) were significantly correlated with Naringenin content.

3.6. Structural Genes Involved in Anthocyanin Biosynthesis

This research mined hub genes involved in the anthocyanin biosynthesis pathway by utilizing transcriptomic and metabolic data to predict the molecular mechanisms controlling color variations in pea pods (Figure 7). Figure 7 illustrates that the ZY cultivar exhibited elevated expression levels of the PAL, 4CL, and CHS genes from the upstream synthesis pathway, while the FM cultivar showed higher expression of the FLS gene, supplying substrates for downstream anthocyanin synthesis. In the synthesis pathway, the F3H and UFGT genes, responsible for producing colored substances, exhibited the highest expression levels in the ZY cultivar. Enhanced F3H expression facilitates the accumulation of Dihydrokaempferol, a crucial precursor for synthesizing stable Cyanidin 3-O-galactoside, Pelargonidin-3-O-galactoside, and Delphinidin-3-O-galactoside, catalyzed by DFR, ANS, and UFGT in the purple ZY cultivar.

3.7. Genetic and BSA-Seq Analysis

The F₁ plants derived from the cross between FM and QZ were green. In the F₂ population, 47 plants produced green pods, while 17 plants produced yellow pods. The segregation ratio fit the Mendelian segregation ratio for a single gene (χ² = 0.08, less than χ²_3:1 = 3.84), indicating that yellow pod color is controlled by a single nuclear gene, with green being dominant over yellow. The immature pods of F₁ plants from the cross of QZ × ZY were all purple but with green dots, meaning that purple was mosaic dominant over green. The immature pods in the F₂ population displayed a quantitative characteristic since the pod color exhibited continuous distribution from dark purple to green (Figure 8A). These data indicate that the anthocyanin content in purple pods is controlled by multiple genes. Yellow and purple pea pod colors are controlled by different genes, and some genes controlling the anthocyanin content were found in the RNA-seq analysis results. Because purple pod color is controlled by multiple genes, pyramiding genes related to the anthocyanin content in the breeding process was an effective method for cultivating new sugar pea cultivars with high anthocyanin content.

BSA-seq analysis was conducted to identify the QTLs associated with the colors of pea pods. A total of 9,736,221 SNPs were detected, and the SNP-index/ΔSNP-index was calculated for each SNP. BSA-seq analysis showed highly contrasting patterns of SNP-index for ‘Green’ bulk and ‘Purple’ bulk in a slightly larger region between 37.5 and 170.5 Mb on chromosome 6 (CM044350.1) (Figure 8B). The peak of the regression line for the ΔSNP-index was near 80.0 Mb (Figure 8B). GPS analysis mapped the candidate region from 71.2 to 87.2 Mb on chromosome 6 (Figure 8C), and 21 DEGs were located in this region. Among these DEGs, five DEGs (

Table 3. The types of variants found in the candidate genes.

Gene ID	Gene Annotation	Variant Location	SNP			Amino Acid Change
			FM	QZ	ZY
KIW84_061556	Cytochrome P450 94A2-like protein	-				-
KIW84_061609	PDDEXK-like family of unknown function	-				-
KIW84_061692	CIPK1-like kinase	5th exon	A	G	A	nonsynonymous
KIW84_061701	Storage globulins with unclear roles in pigmentation	Exon	T	T	A	synonymous
KIW84_061702	Storage globulins with unclear roles in pigmentation	Exon	A	G	A	synonymous
		Exon	G	A	G	nonsynonymous
		Exon	G	A	G	synonymous
KIW84_061697	Transcription factor TT8-like protein with bHLH domain	Exon	C	T	T	synonymous
		Exon	G	A	G	synonymous
		Exon	A	G	G	synonymous
		Exon	C	G	C	nonsynonymous
		Exon	T	C	T	synonymous
KIW84_061698	BHLH transcription factor	Exon	G	G	A	nonsynonymous
		Exon	G	G	A	synonymous
		Exon	A	A	G	synonymous
		Exon	T	T	C	nonsynonymous
		Exon	G	G	A	synonymous
		Exon	A	A	T	synonymous
		Exon	T	T	C	synonymous
		Exon	G	G	A	synonymous

), including KIW84_061556, KIW84_061609, KIW84_061692, KIW84_061701, and KIW84_061702, were highly expressed in ZY but showed low expression levels in both FM and QZ. KIW84_061556, which encodes a cytochrome P450 94A2-like protein, is likely involved in fatty acid metabolism or signaling; KIW84_061609 belongs to the PDDEXK-like family with unknown function; KIW84_061692 encodes a CIPK1-like kinase, possibly linked to calcium signaling; KIW84_061701 and KIW84_061702 encode storage globulins with unclear roles in pigmentation. Coincidentally, gene A, which controls anthocyanin production in the seed coat/flower, was located in the candidate region. Gene A was divided into two genes (KIW84_061697 and KIW84_061698) in the PeaZW6 reference genome. KIW84_061697 and KIW84_061698 were highly expressed in ZY compared with QZ and FM, respectively. Furthermore, the base variants of these seven genes between the three cultivars were analyzed using RNA-seq data (Table 3). KIW84_061556 and KIW84_061609 showed no sequence variations in pea pods with different colors. Most of the base variations in other genes were synonymous mutations. The SNP in the fifth exon of KIW84_061692 resulted in a nonsynonymous mutation; however, FM (showing low anthocyanin content) and ZY (showing high anthocyanin content) shared the same base at this site. One SNP in KIW84_061701 caused a synonymous mutation. Two base variations in KIW84_061702 resulted in a synonymous mutation. Another one caused a nonsynonymous mutation, but FM and ZY had the same base at this site. Four SNPs in KIW84_061697 resulted in a synonymous mutation; one SNP caused a nonsynonymous mutation, but FM and ZY also had the same base. It is worth noting that two SNPs in KIW84_061698 resulted in nonsynonymous mutations. QZ and FM, which had lower anthocyanin content, shared the same base, differing from ZY. KIW84_061698 encodes a bHLH transcription factor. The mutation of the two SNPs may affect the binding to the DNA sequence and affect gene function, subsequently influencing anthocyanin content and pod color. Thus, KIW84_061698 is identified as the key candidate gene for anthocyanin content in pea pods.

4. Discussion

Pea is a widely consumed crop, yet the genetic basis for pod color variation remains unknown. The genome of a yellow-pod pea line was sequenced, and its genetic and transcriptome profiles were compared with those of green-pod lines. The role of this candidate gene in color variation is yet to be reported [14]. Additionally, the combined analysis of the transcriptome and metabolome revealed a direct link between the decreased expression levels of genes and the reduced accumulation of metabolites in the delphinidin biosynthesis pathway between two pea accessions with green pods and yellow pods [3]. Recently, an excellent study uncovered the molecular basis for yellow pods [12]. As secondary metabolites, anthocyanins are vital for determining leaf color and affect fruit and seed color by varying in content and type [50,51,52]. Cyanidin, delphinidin, pelargonidin, peonidin, malvidin, and petunidin are the main anthocyanidins that contribute to the diverse colors found in plants [53]. The reddening of jujube peel is associated with high abundances of 3-O-glucoside and delphinidin 3-O-glucoside [40]. The red and pink petals of Camellia oleifera contain peonidin-3-O-glucoside and cyanidin-3-O-(6″-O-p-Coumaroyl) glucoside, respectively [54]. In this study, ZY showed the highest total anthocyanin content, containing 99 times more than QZ and FM (Figure 1B), corresponding with the purple color phenotype.

Anthocyanins, including delphinidin, cyanidin, and pelargonidin, usually endow plants with blue, purple, and red [55]. A targeted metabolomics approach was employed to extensively characterize the anthocyanin profile in pea pods, resulting in the identification of 61 distinct anthocyanins, including 16 delphinidin derivatives, 14 cyanidin derivatives, 10 pelargonidin derivatives, 7 peonidin derivatives, 4 malvidin derivatives, and 3 petunidin derivatives (Figure 2A). Afterward, we investigated the variation of anthocyanin content in these three pea cultivars. Our detailed quantitative analysis and comparison revealed that both the quantity and composition of anthocyanins are significantly responsible for the purple coloration of ZY. In ZY, Cyanidin-3-O-(6-O-malonyl-beta-D-glucoside), Pelargonidin-3-O-glucoside, and Pelargonidin-3-O-rutinoside were significantly enriched in the ‘anthocyanin biosynthesis’ pathway compared with FM. Additionally, Cyanidin-3-O-(6-O-malonyl-beta-D-glucoside), Cyanidin-3-O-(6-O-p-coumaroyl)-glucoside, Pelargonidin-3-O-glucoside, and Pelargonidin-3-O-rutinoside were significantly enriched in the ‘anthocyanin biosynthesis’ pathway in ZY compared with QZ. Specifically, we identified 37 more highly accumulated anthocyanins in the ZY pea pods compared to the FM pea pods. Furthermore, 40 anthocyanins were more highly accumulated in the ZY pea pods compared to the QZ pea pods. The primary anthocyanins responsible for the purple pigmentation in the ZY pea pods were four metabolites: Delphinidin-3-O-galactoside, Delphinidin-3-O-(6″-O-xylosyl)glucoside, Cyanidin-3-O-galactoside, and Pelargonidin-3-O-(xylosyl)glucoside. In the comparison, 21 differences were noted between FM and QZ, 49 between ZY and FM, and 43 between ZY and QZ (Figure 2B). These results suggest that the type and concentration of anthocyanin compounds, particularly those based on delphinidin, cyanidin, and pelargonidin, are crucial in the development of purple pea pods. Furthermore, the differentially accumulated anthocyanins were enriched in the ‘anthocyanin biosynthesis’, ‘biosynthesis of secondary metabolites’, and ‘flavone and flavonol biosynthesis’ pathways (Figure 2C–E). Some DAMs were enriched in ‘flavonoid biosynthesis’ in the ZY vs. FM and ZY vs. QZ comparisons (Figure 2D,E). Anthocyanins and procyanidins are considered beneficial for human health due to their antioxidant capacity to reduce excessive reactive oxygen species (ROS). Anthocyanins are predominantly found in fruits and vegetables. Therefore, ZY, with its high anthocyanin content, can be utilized in breeding anthocyanin-rich sugar snap pea cultivars to improve their nutritional value and health care benefits.

RNA-seq analysis found that some structural genes, including PAL, 4CL, CHS, FLS, F3H, and UFGT, showed different expression levels in these three cultivars. These results indicate that these genes are crucial for the determination of pigment contents in pea pods. KEGG pathway co-enrich analysis of DAMs and DEGs also revealed that anthocyanin and flavonoid biosynthesis are key in color-related metabolic processes (Figure 6A–C). It is proposed that the color differences among the cultivars QZ, FM, and ZY result from variations in anthocyanin and flavonoid metabolism. Genes and metabolites jointly regulate the production of anthocyanins [56]. Previous research showed that the overexpression of PbPIF3a and PbPIF4 significantly suppressed anthocyanin accumulation and altered skin color in pear [56]. Generally speaking, changes in DEGs and DAMs significantly contribute to increased anthocyanin content in ZY.

Research over the past decade on model plants and diverse fruits has discovered anthocyanin metabolism pathways and biosynthetic genes, highlighting a certain degree of conservation [18]. The production of anthocyanins is controlled by initial structural genes (CHS, CHI, and F3H) and genes specific to anthocyanin biosynthesis (F3′H, F3′5′H, DFR, ANS, and UFGT) [55]. The hub genes PAL, 4CL, CHS, F3H, and UFGT show elevated expression in ZY (Figure 7), leading to increased anthocyanin content in the ZY samples relative to FM and QZ. However, expression of the FLS gene was lower in QZ compared to FM and ZY. The flavonoid metabolism pathway is the most direct metabolic pathway for variation in pod coloration between green pods and yellow pods [3]. The high expression of FLS genes in FM may lead to high flavonol accumulation levels and the yellow pod phenotype. Chalcone synthase (CHS) is the key enzyme in anthocyanin biosynthesis, facilitating the conversion of 4-coumaroyl-CoA and malonyl-CoA into naringenin chalcone. The construction of the correlation network between DEGs and DAMs identified Naringenin as the key intermediate product in the anthocyanin biosynthesis pathway, and four CHS genes (KIW84_025518, KIW84_052803, KIW84_025517, and KIW84_025227) were significantly correlated with Naringenin content (Figure 6E,F). Furthermore, these four genes and the other two CHS genes, KIW84_025521 and KIW84_052986, exhibited higher expression levels in ZY compared to FM and QZ (Figure 7). UFGT is essential in the last stage of anthocyanin biosynthesis, aiding in the glycosylation and stabilization of unstable anthocyanins. UFGT also signals the transport of anthocyanins to vacuoles, where they act as pigments. This study identified two UFGT genes, KIW84_015663 and KIW84_057044, which exhibited higher expression levels in ZY compared to FM and QZ (Figure 7). The varying expression levels of these candidate genes among the three pea pods provide valuable insights into their role in the determination of pea pod color and anthocyanin accumulation.

The BSA-seq analysis successfully identified a candidate region associated with purple pea pod color. This region was located on chromosome 6 (Figure 8) and contained 21 DEGs. A total of 5 of the 21 DEGs were highly expressed in ZY compared with FM and QZ. Coincidentally, gene A (LOC127096702), encoding a bHLH transcription factor and responsible for seed coat/flower color, was located at the candidate region. A was divided into two genes (KIW84_061697 and KIW84_061698) in the PeaZW6 reference genome. A is necessary for anthocyanin production in the flowers. In fact, all the plants with white flowers and no anthocyanin in the leaf axils produced green pods in the F₂ population derived from QZ × ZY. As a transcription factor, A may interact with some special genes to affect anthocyanin content in pea pods. RNA sequence analysis identified two base mutations in KIW84_061698 that may affect the binding to the DNA sequence and influence gene function. Therefore, this gene may be regarded as the key candidate gene related to anthocyanin content.

This study utilized metabolomics and transcriptomics data to systematically analyze pod color variation among yellow, purple, and green pod peas. However, this study has potential limitations. The functions of hub genes for the mechanism of anthocyanin-based color development in pea pods need to be more fully elucidated.

5. Conclusions

In summary, analyses of the transcriptome and metabolome were conducted on peas (Pisum sativum L.) exhibiting various colors. Compared with other anthocyanins, four metabolites, including Delphinidin-3-O-galactoside, Delphinidin-3-O-(6″-O-xylosyl)glucoside, Cyanidin-3-O-galactoside, and Pelargonidin-3-O-(xylosyl)glucoside, were significantly more accumulated in the ZY cultivar with purple pods. The ZY cultivar shows elevated expression of hub genes (PAL, 4CL, CHS, F3H, and UFGT), leading to increased anthocyanin content compared to the FM and QZ samples. Therefore, ZY with high levels of anthocyanin content can be utilized in breeding anthocyanin-rich sugar snap pea cultivars to improve their nutritional value and health care benefits. BSA-seq and RNA-seq analyses found that KIW84_061698 is the key candidate gene controlling anthocyanin content in pea pods. Our findings offer in-depth insights into the molecular mechanisms underlying anthocyanin biosynthesis in pea pods. This understanding will aid in creating pea cultivars with enhanced anthocyanin content.