Genetic Dissection of Anthocyanin Accumulation in Tomato Using GWAS and Hybridization Probe Melting (HPM) for Marker-Assisted Breeding

Abstract

Tomato fruit color is primarily influenced by the accumulation of pigments such as carotenoids and anthocyanins, which are regulated by a complex network of genes and environmental factors. The presence of anthocyanins, in particular, contributes to the purple phenotype in tomatoes, which has been associated with improved nutritional quality and disease resistance. Previous studies have identified key regulatory genes, including SlMYBATV and SlANT1, that control anthocyanin biosynthesis in tomatoes. A total of 48 tomato accessions were selected, including both anthocyanin-producing and non-producing cultivars, to assess genetic variation in relation to fruit color. GWAS analysis identified significant associations between single-nucleotide polymorphisms (SNPs) on chromosomes 7 and 10 and the purple fruit phenotype. These genomic regions contained key anthocyanin regulatory genes, SlMYBATV on chromosome 7 and SlANT1 on chromosome 10, confirming their roles in anthocyanin biosynthesis. Linkage disequilibrium (LD) analysis further revealed strong correlations between SNPs within these regions, facilitating the selection of representative SNPs for genotyping. The genotyping of SNPs Ch07_60981501 and Ch10_64354129 using HPM demonstrated clear differentiation between purple and non-purple cultivars based on melting temperature differences, validating the functional significance of the identified loci. The results confirmed that the nonfunctional SlMYBATV allele at the atv locus allows for enhanced anthocyanin accumulation by relieving repression of anthocyanin activators like SlANT1. This interaction between the Aft and atv loci drives light-dependent anthocyanin biosynthesis in purple tomatoes. HPM genotyping offers a cost-effective tool for allele identification, supporting breeding programs for anthocyanin-rich tomatoes. SNP markers enable marker-assisted selection (MAS) for improved nutritional and aesthetic traits. This study highlights SlMYBATV and SlANT1 in anthocyanin biosynthesis, advancing efforts to develop enriched tomato varieties and supporting agricultural productivity and health.

1. Introduction

The tomato (Solanum lycopersicum L.), originated from South America, which was then domesticated in Mexico and later brought to Europe. Its introduction to Africa dates back to the 17th century. Several wild species of tomato exist, but only “current tomato” (Solanum pimpinellifolium) and the “cherry tomato” (Solanum lycopersicum var. cerasiforme) are suitable for consumption by humans [1]. Tomato is the world’s second most widely consumed vegetable after potato (Solanum tuberosum), with global production reaching 161.7 million metric tons. While the majority of tomatoes are consumed fresh, there is an increasing demand for processed tomato products, especially in Europe and North America [2,3].

Tomatoes provide a wide range of health benefits, including potential protection against cancers, cardiovascular diseases, and osteoporosis. Their high lycopene content, along with nutrients like vitamin C, folate, and polyphenols, contributes to their antioxidant and anti-cancer properties. Tomatoes also support liver health, assist in detoxification, and may help lower high blood pressure [4,5]. Regular consumption of tomatoes and tomato-based products has been associated with improved skin health, fertility, immune function, and recovery after exercise. Factors such as light intensity, growing media, and temperature can affect their nutritional value. However, excessive intake of tomato products or lycopene supplements may lead to adverse effects [6]. Studies have shown that purple tomatoes offer anti-inflammatory, antimicrobial, anticancer, cardioprotective, hepatoprotective, and neuroprotective properties. Notably, anthocyanin-rich tomatoes have shown promise in extending the lifespan of cancer-prone mice, indicating further health-promoting effects. These tomatoes have been developed through both conventional breeding and transgenic techniques, using specific transcription factors from snapdragon flowers. In addition to their health benefits, purple tomatoes exhibit improved shelf-life and storability, making them advantageous for farmers, the processing industry, and consumers alike [7,8]. Significant importance is attributed to the presence of certain secondary plant metabolites, such as anthocyanins and other flavonoids, due to their health-promoting effects [9,10,11].

The vibrant colors of flowers and fruits are a result of different pigments from the phenylpropanoid and terpenoid classes, primarily chlorophylls, carotenoids, and anthocyanins. Natural pigments produced in plants, such as chlorophylls, carotenoids, and anthocyanins, are typically synthesized through the phenylpropanoid and terpenoid pathways [8]. In some crops, domestication may have favored traits related to cultivation over those affecting color. In others, the natural pigments in the edible parts have been intensified or fine-tuned. This is especially true for tomatoes, which are rich in carotenoid pigments like lycopene and phytoene [12]. However, tomatoes contain only small amounts of certain flavonoids, and somewhat surprisingly, the fruits typically do not produce anthocyanins, unlike other members of the Solanaceae family such as eggplant (Solanum melongena) or pepper (Capsicum spp.) [8].

Tomato plants possess a variety of flavonoids in their vegetative tissues, including anthocyanins. However, in the fruit, only small amounts of naringenin chalcone (an intermediate in the flavonoid biosynthetic pathway) and certain flavonols, such as quercetin and kaempferol glycosides, accumulate. These flavonols are concentrated in the peel, but tomatoes do not produce anthocyanins. The ‘black’ or ‘purple’ fruits found in some heirloom varieties are a result of mutations that affect chlorophyll breakdown and carotenoid content, rather than anthocyanin production [13]. Some wild cultivars of tomatoes retained the ability of producing anthocyanins.

Anthocyanins, the most common pigments responsible for purple, red, and blue colors in plants, include more than 300 identified compounds. Structurally, they are planar molecules with a C6-C3-C6 carbon backbone, typical of flavonoids. These compounds are categorized into distinct classes based on the presence of hydroxyl groups at positions 3′, 4′, or 5′ of the phenyl “B” ring, the type of glycosidic subunits attached, and whether the sugar hydroxyl groups are acylated [14,15]. The flavonoid biosynthetic pathway has been studied in various lines of S. lycopersicum, including S. lycopersicum V. cerasiforme, which is believed to be the most likely wild ancestor of domesticated tomatoes. A significant limitation in the flavonoid biosynthetic pathway is the lack of expression of the Chalcone isomerase (CHI) gene in the fruit peel. This deficiency is likely due to a mutation in a fruit-specific regulatory element of the promoter region. As a result, the high levels of naringenin chalcone found in the tomato peel can be attributed to this enzyme’s absence, as naringenin chalcone is the primary substrate for the CHI enzyme. Conversely, the biosynthetic pathway appears to be consistently inactive in the fruit’s flesh, where none of the structural genes analyzed are expressed. Interspecific crosses with wild tomato species have introduced the ability to produce small quantities of anthocyanins in the peels of cultivated tomatoes. For instance, the dominant gene Anthocyanin fruit (Aft) induces limited pigmentation when stimulated by high light intensity and was transferred to domesticated tomatoes through a cross with S. chilense. Similarly, the gene Aubergine (Abg), introgressed from Solanum lycopersicoides, can create strong and variegated pigmentation in the peels of tomatoes. The dominant Aft gene and the recessive atroviolacea (atv) gene, when introduced into domesticated tomatoes from different wild Solanum species, lead to only modest anthocyanin pigmentation. Interestingly, when these genes are combined in the double mutant Aft/Aft atv/atv, the result is tomatoes with deep, vibrant purple pigmentation, far more intense than with either gene on its own [16,17].

The transcription factors SlAN2 and ANTHOCYANIN1 (SlANT1) both play roles in anthocyanin synthesis in tomatoes, but their functions differ. SlAN2 is a key regulator of anthocyanin accumulation in response to high light and low temperatures. When expressed, it induces the expression of genes like SlDFR, SlAN1, and SlJAF13, which are involved in the anthocyanin biosynthetic pathway. SlAN2 is transcriptionally activated under specific environmental conditions, leading to significant pigment accumulation [18]. In contrast, SlANT1, while capable of inducing anthocyanin synthesis when overexpressed, does not naturally respond to high light or low temperatures and has lower expression levels under normal conditions. Silencing SlAN2 reduces anthocyanin biosynthesis, but silencing SlANT1 has no significant effect. This indicates that SlAN2 plays a more central role in regulating anthocyanin accumulation, while SlANT1 appears to have an additive or synergistic role [18]. The transcription factor SlAN2-like, part of the Aft locus, plays a key role in regulating fruit pigmentation in tomatoes. In wild-type tomatoes, splicing mutations in SlAN2-like can impact its functionality, leading to variations in the intensity of pigmentation [19].

SlHY5 is a key transcription factor that regulates anthocyanin production in tomatoes in response to blue light. It binds to specific elements in the promoters of anthocyanin biosynthesis genes, such as chalcone synthase and dihydroflavonol 4-reductase, leading to increased anthocyanin levels. Silencing SlHY5 reduces anthocyanin accumulation, highlighting its role in light-driven pigment production in tomatoes [20]. SlHY5 drives SlAN1, SlAN2-like, which activates anthocyanin synthesis-related genes. Several transgenic approaches have been employed to boost flavonoid levels in tomato fruit by overexpressing either structural or regulatory genes involved in the biosynthetic pathway. Many of these strategies have utilized heterologous genes to achieve this goal, aiming to modify and enhance the production of flavonoids in the fruit [21,22,23], whereas only homologous expression of ANT1 showed increased anthocyanin production and increased shade avoidance, thinner leaves, lower seed germination rate, suppressed side branching, increased chlorophyll concentration, and lower photosynthesis [24,25].

Recent research has focused on the development of purple tomatoes enriched with anthocyanins, potent antioxidants with a variety of health benefits. These genetically engineered tomatoes combine the nutritional benefits of anthocyanins with traditional tomato phytochemicals, leading to fruits with greater nutraceutical value and higher antioxidant capacity than conventional red tomatoes [23]. In addition to their health benefits, purple tomatoes exhibit improved shelf-life and storability, making them advantageous for farmers, the processing industry, and consumers alike [7].

Genome-wide association studies (GWASs) have emerged as a powerful method for identifying genetic variants linked to phenotypic traits in plants [26]. The process consists of several key steps, including assembling and phenotyping an association panel, genotyping, conducting association analysis, and identifying candidate genes [27]. GWAS has been effectively applied to a wide range of plant traits, such as biotic resistance, abiotic tolerance, yield, and metabolic composition. The introduction of mixed model frameworks has helped to minimize false positives in GWAS results. Current research aims to enhance statistical power, improve computational speed, and tackle challenges like rare-variant analysis and synthetic associations. The insights gained from GWAS can be utilized for gene cloning and to accelerate crop breeding through marker-assisted selection or genetic engineering [28]. Genome-wide association studies (GWASs) have become essential for identifying genetic loci linked to complex traits in tomatoes. These studies have established significant connections between genetic markers and traits such as fruit volatiles, carotenoid content, and climate adaptation. GWAS has also identified novel candidate genes for phosphate uptake efficiency and 44 candidate loci related to fruit metabolic traits, including amino acids, sugars, and ascorbate levels [29,30]. These insights improve our understanding of the genetic basis of important agronomic traits in tomatoes and provide targets for crop improvement. Utilizing diverse tomato accessions, including landraces and wild relatives, has effectively captured a broad spectrum of genetic diversity for trait mapping and gene discovery [29,31]. Harnessing this knowledge, this study aims to identify key genes related to variation in fruit anthocyanin content of tomato for breeding anthocyanin-rich tomatoes.

2. Materials and Methods

2.1. Plant Materials and Growth Conditions

The seeds of the 48 resources used in this study were self-pollinated breeding lines that had undergone more than 10 generations of advancement. These seeds were provided by the agricultural company BuNong Seed Co., Ltd. (Jeonnam, Republic of Korea). Among the 192 resources, only five exhibited purple traits, while the others did not. All seeds were sown on March 5, 2024, at the BuNong Seed Breeding Research Center (35.85° N, 126.89° E) using 40-cell plug trays (540 × 280 × 48 mm). The plug seedlings were transplanted into a greenhouse on 8 April 2024, at designated planting densities. The cultivation method used ridge planting with plastic mulching, and the plants were arranged in double rows. The ridge width was 90 cm, the plant spacing was 40 cm, and the row spacing was 30 cm. To minimize the influence of various factors, all resources were grown in the same greenhouse under identical conditions, following standard tomato cultivation practices.

2.2. GBS (Genotyping-by-Sequencing) Library Preparation

This study aimed to identify genes associated with purple traits in tomatoes and develop molecular markers for rapid assessment of breeding materials with these traits. A total of 48 tomato resources, including five with purple traits and more than 10 generations of advancement, were provided by the BuNong Seed Breeding Research Center for this experiment.

2.3. DNA Extraction

A total of 48 tomato leaf samples (100 mg each) containing five purple traits were placed into 2 mL tubes containing stainless steel beads (Benchmark Scientific, D1033-23, Sayreville, NJ, USA) and finely ground using a BeadBlaster 24 Microtube Homogenizer (Benchmark Scientific, Sayreville, NJ, USA). The ground plant samples were then used to extract DNA following the standard protocol of the DNeasy Plant Pro Kit (QIAGEN, Hilden, Germany). The extracted DNA was confirmed through electrophoresis on a 1.5% agarose gel for 25 min and quantified using a nano spectrophotometer (Microdigital Co., Ltd., Seongnam, Republic of Korea). All samples were diluted to a uniform concentration of 100 ng/µL before being used in subsequent experiments.

2.4. Preparation of Libraries for Next-Generation Sequencing

GBS libraries are constructed using restriction enzyme ApeKI (GCWGC) using a protocol modified from Elshire et al., 2011 [32]. Oligonucleotides comprising the top and bottom strands of each barcode adapter and a common adapter were diluted (separately) in TE (50 uM each) and annealed in a thermocycler (95 °C, 2 min; ramp down to 25 °C by 0.1 °C/s; 25 °C, 30 min; 4 °C hold). Barcode and common adapters were then diluted in 10× adapter buffer (500 mM NaCl, 100 mM Tris-Cl) to 10 uM, mixed together in a 1:1 ratio, and 2.4 μL of the mix was added into a 96-well PCR plate. DNA samples (100 ng/μL) were added to individual adapter-containing wells. Samples (DNA plus adapters) were digested overnight at 75 °C with ApeKI (New England Biolabs, Ipswitch, MA, USA) in 20 μL volumes containing 1× NEB Buffer 3 and 3.6 U ApeKI. Adapters were then ligated to sticky ends by adding 30 μL of a solution containing 10× ligase buffer and T4 DNA ligase (200 units) (MGMED, Rimavská Sobota, Slovakia) to each well. Samples were incubated at 22 °C for 2 h and heated to 65 °C for 20 min to inactivate the T4 DNA ligase. Sets of 96 digested DNA samples, each with a different barcode adapter, were combined (5 μL each) and purified using a commercial kit (QIAquick PCR Purification Kit; Qiagen, Hilden, Germany; Valencia, CA, USA) according to the manufacturer’s instructions. DNA samples were eluted in a final volume of 50 μL. Restriction fragments from each library were then amplified in 50 μL volumes containing 2 μL pooled DNA fragments, HerculaseII Fusion DNA Polymerase (Agilent, Santa Clara, CA, USA), and 25 pmol each of the following primers: (A) 5′-AAT GAT ACG GCG ACC ACC GAG ATC TAC ACT CTT TCC CTA CAC GAC GCT CTT CCG ATC T-3′ and (B) 5′-CAA GCA GAA GAC GGC ATA CGA GAT CGG TCT CGG CAT TCC TGC TGA ACC GCT CTT CCG ATC T-3′. Polymerase chain reaction (PCR) was performed using a Life ECO Thermal Cycler (Bioer Technology Co., Hangzhou, China) with the following conditions: one cycle at 95 °C for 2 min, followed by 16 cycles at 95 °C for 30 s, 62 °C for 30 s, and 68 °C for 30 s, and finishing at 68 °C for 5 min. These amplified sample pools constitute a sequencing “library.” Libraries were purified as above (except that the final elution volume is 30 μL). To evaluate the amplified fragment size, 2 μL was loaded onto an Agarose gel and library quality was checked using Agilent Tape station with high-sensitivity DNA chip.

2.5. Preprocessing

After sequencing, raw reads were de-multiplexed according to the barcode sequences and trimmed using an in-house Python script. This script splits the raw Illumina fastq file into separate fastq files based on the barcode sequences associated with each sample while filtering out reads containing any ambiguous bases in the barcodes. Reads were also trimmed using the cutadapt if the sequence contained the common adapter.

The de-multiplexed reads was trimmed using the Solexa QA package v.1.13 [33]. It is common for the quality of bases from either end of Illumina reads to drop in quality. We therefore trimmed either end of reads when the Phred quality score dropped below Q = 20 (or 0.05 probability of error). In addition, we also trimmed all 5′ and 3′ stretches of ambiguous ‘N’ nucleotides. Poor-quality sequence reads along with reads shorter than 25 bases were discarded.

2.6. Alignment and Detection of SNPs and InDels

To align the clean reads to the reference genome of Solanum lycopersicum cv. Heinz, the Burrows–Wheeler Aligner (BWA, 0.6.1-r104) program was applied. The BWA default values for mapping were used, except for seed length (−l) = 30, maximum differences in the seed (-k) = 1, number of threads (-t) = 16, mismatch penalty (-M) = 6, gap open penalty (-O) = 15, gap extension penalty (-E) = 8. Mapped reads were extracted from the resulting BAM file using SAMtools v.0.1.16 [34] for further analyses. The high mapping quality ensures reliable (unique) mapping of the reads, which is important for variant calling. Using the varFilter command, SNPs were called only for variable positions with a minimal mapping quality (-Q) of 30. The minimum and maximum of read depths were set to 3 and 95, respectively. An in-house script considering biallelic loci was used to select a significant site in the called SNP positions. Depending on the ratio of SNP/InDel reads to mapped reads, the variant type is classified into three categories: homozygous SNP/InDel for more than 90%, heterozygous SNP/InDel for more than 40% and less than 60%, and the rest of them as Etc.

To control the quality of markers, missing proportion (MSP) < 0.3 and minor allele frequency (MAF) > 0.1 were selected.

2.7. GWAS Analysis

The GWAS analysis was conducted using the GAPIT (Genome Association and Prediction Integrated Tool) R (version 4.3.1) package. Genotypic data for this study were obtained from a HapMap file, while phenotypic data consisted of fruit color categorized as purple (coded as 0) and other colors (coded as 1). SNPs with a minor allele frequency (MAF) below 0.05 and missing data above 10% were excluded to ensure data quality. GWAS was performed by employing four models: the General Linear Model (GLM), which performs basic association analysis without accounting for population structure or kinship; the Mixed Linear Model (MLM), which incorporates a kinship matrix and population structure; FarmCPU (Fixed and Random Model Circulating Probability Unification), which balances detection power and false-positive control; and BLINK (Bayesian-information and Linkage-disequilibrium Iteratively Nested Keyway), a computationally efficient model that controls linkage disequilibrium. Population structure was assessed using principal component analysis (PCA), with the top three principal components included as covariates in MLM, FarmCPU, and BLINK models. A kinship matrix was computed to model relatedness among individuals in MLM. Genome-wide significance was evaluated using a Bonferroni-adjusted p-value threshold of 0.05, and results were visualized through Manhattan and QQ plots. Candidate genes near significant SNPs were identified based on their genomic positions, with functional annotations performed to hypothesize their roles in determining fruit color.

2.8. HPM Analysis

The SNPs were identified through Hybridization probe melting (HPM) analysis combined with a 3′-phosphorylated oligonucleotide probe. Total genomic DNA (gDNA) was isolated from leaf tissue as per manufacturer’s instructions and HPM analysis (Roche, Mannheim, Germany) was carried out using a LightCycler 96 instrument (Roche, Mannheim, Germany). After that, HPM was performed in 10 µL reaction mixtures containing 1 µL at 5 ng/µL DNA, 0.1 µL of forward primer (10 pmol), 0.5 µL of reverse primer (10 pmol), 0.5 µL of probe (10 pmol), 0.3 µL of SYTO9 fluorescent dye (Invitrogen, Carlsbad, CA, USA), 5 µL of HS prime LP premix (GENETBIO, Daejeon, Republic of Korea), and 2.6 µL of DDW. Three steps were needed in HPM conditions, where the first step was preincubation (300 s initial pre-incubation at 95 °C), second step was 40 cycles of 3 step amplification (95 °C for 10 s, 64 °C to 56 °C for 15 s under touchdown and 72 °C for 15 s), and the remaining last step was probe melting reading (last step was four readings per °C at the final step after 60 s at 95 °C, 60 s at 40 °C, and 1 at 97 °C). Then, LightCycler 96 software (version 1.1) was used for HPM data analysis.

3. Results

3.1. Phenotypic Analysis of Seedling Growth and Fruit Trait

Tomato fruit color changes during ripening, transitioning from green to red as chlorophyll degrades and carotenoids accumulate [35]. This process is influenced by genetic factors, with key genes like Psy1, SGR, and SlMYB12 playing crucial roles in pigment production [36]. Recent research has revealed complex post-transcriptional regulatory mechanisms, including RNA splicing, that fine-tune the expression of color-related genes [37]. Tomato fruit color assessment is crucial for determining ripeness and quality. A total of 48 tomato accessions from the core collection of tomato cultivars were selected for this study. Only healthy, disease-free plants showing no signs of stress were included to ensure consistent fruit development. Visual inspection of fruit color revealed that five accessions exhibited an extreme purple fruit phenotype, which was notably distinct compared to the other accessions (Figure 1). These five accessions were classified as anthocyanin-producing cultivars, while the remaining accessions, which lacked the purple phenotype, were grouped as anthocyanin-absent cultivars. All plants were grown in randomized hills to control for any environmental bias, and this grouping formed the basis for the GWAS analysis.

3.2. Identification of Makeres Related to Anthosyanin Content by GWAS Analysis

Genotyping of the 48 tomato accessions resulted in a total of 4,811,535 SNPs, with an average of 100,240 SNPs per accession. The homozygous to heterozygous ratio across all accessions was 1.2, with accessions ranging from 44,752 to 159,637 total SNPs and homozygous to heterozygous ratios varying between 0.23 and 1.78 (Figure 2a). GWAS analysis revealed significant associations of SNPs from chromosomes 7 and 10 with the variation in fruit anthocyanin accumulation (purple color). Specifically, on chromosome 7, SNPs between positions 60,368,016 and 60,981,501 were significantly associated with fruit color. Similarly, on chromosome 10, SNPs within the range of 63,572,363 to 64,640,219 were also strongly correlated with the purple phenotype (Figure 2b, Data S1).

3.3. Mapping of Significantly Associated Genes

To identify potential candidate genes involved in anthocyanin biosynthesis, all annotated genes within the identified significant SNP regions on chromosomes 7 and 10 were reviewed (Table S1). Two genes stood out as being particularly relevant: SlMYBATV (Solyc07g052490) on chromosome 7 and SlANT1 (Solyc10g086260) on chromosome 10. Both genes are known to play critical roles in anthocyanin production and were located within the significant SNP regions (Figure 3) [18,38,39,40].

3.4. Identification of Markers for SNP Genotyping

To identify specific SNP markers linked to SlMYBATV and SlANT1, we conducted a linkage disequilibrium (LD) analysis. This revealed that the genes of interest lie between SNPs Ch07_60758024 and Ch07_60981501 on chromosome 7, and Ch10_64228589 and Ch10_64354129 on chromosome 10. The LD analysis demonstrated that the SNPs within these regions exhibit high r² values, indicating a strong correlation and suggesting that these SNPs are inherited together (Figure 4 and Table S2). This high degree of linkage means that genotyping any one of these SNPs can reliably predict the allele associated with the gene. For genotyping, SNPs Ch07_60981501 on chromosome 7 and Ch10_64354129 on chromosome 10 were selected due to their proximity to SlMYBATV and SlANT1, respectively.

3.5. Genotyping of Purple and Non-Purple Cultivars by Hybridization Probe Melting (HPM)

Hybridization Probe Melting (HPM) was employed to genotype the selected SNPs in both purple and non-purple tomato cultivars. This high-resolution, closed-tube method provides rapid and cost-effective genotyping without the need for labeled probes. HPM was performed on SNPs Ch07_60981501 and Ch10_64354129, which showed G/Tand T/C polymorphisms, respectively, using primers indicated in Table S3. The melting temperature differences between the alleles were observed to be 9 °C for Ch07_60981501 and 10 °C for Ch10_64354129. For Ch07_60981501, the non-purple cultivars showed a melting peak at 56 °C, while purple cultivars exhibited a peak at 63 °C (Figure 5a). For Ch10_64354129, non-purple cultivars showed a peak at 55 °C, while purple cultivars showed a peak at 64 °C (Figure 5b).

4. Discussion

Anthocyanin biosynthesis in plants is governed by intricate regulatory mechanisms that integrate environmental signals, phytohormonal cues, and genetic factors. The MYB-bHLH-WD40 (MBW) complex serves as a key regulator of anthocyanin production, orchestrating gene expression in response to diverse stimuli [41]. Environmental factors such as light, salinity, drought, and cold stress significantly influence anthocyanin accumulation, alongside regulatory inputs from various phytohormones [41]. Additionally, epigenetic modifications and post-translational changes in transcription factors further refine the control of anthocyanin biosynthesis [42]. MYB transcription factors act as positive or negative regulators by modulating signaling pathways and interacting with other proteins [43]. Anthocyanins play critical roles in plant defense against both biotic and abiotic stresses, while also offering health benefits to humans, such as reducing the risk of chronic diseases [44]. Identification and understanding of anthocyanin biosynthesis regulatory networks have advanced the development of anthocyanin-rich crop cultivars through targeted breeding.

Tomatoes exhibit a diverse range of colors beyond the common red, resulting from the accumulation of various pigments from the epidermis to the pericarp (flesh) [37,45,46]. Tomatoes, especially heirlooms, show a wide range of colors, such as pink, orange, yellow, brown, black, green, purple, and even white [47,48,49]. Instead of anthocyanins, some heirloom varieties appear purplish-brown due to the interaction of various carotenoids in their pericarp. These mutations includes colorless fruit epidermis (y), uniform ripening (u), and green flesh (gf) [13,50]. In the case of tomato, some wild accessions such as Solanum chilenos cv. Dunal retained the genes related to anthocyanin-related genes and domestication favored red tomatoes for cultivation [40,51]. These genes are Anthocyanin fruit (Aft), Aubergine (Abg), and atroviolacium (atv), which were then used for breeding anthocyanin-containing tomatoes [13]. Other genes and mutations are well characterized in many studies [50]. Based on these evidences and markers, there are currently numerous commercial cultivars which truly produce anthocyanin in their fruits, produced by crossing with donor wild cultivars [50,52,53].

The Aft trait is controlled by a dominant gene and is closely linked to the insertion and deletion (InDel) marker HP1953 within the R2R3-MYB gene SlAN2 (Solyc10g086250) [38]. At the Aft locus, a cluster of four closely related R2R3 MYB transcription factor genes involved in anthocyanin regulation was identified other than SlANT2, which are Solyc10g086260 (SlANT1), Solyc10g086270 (SlANT1-like/SlMYB28), and Solyc10g086290 (SlAN2-like/SlMYB114) [18,54]. Their close physical proximity suggests that this cluster arose from a tandem gene duplication event [55]. The overexpression of SlAN1 and SlANT2 significantly enhances anthocyanin content in tomato fruit, suggesting that the Aft gene may encode SlANT1 [56] and/or SlAN2 [13,16]. Tomatoes with only the Aft locus, such as cv LA1996, develop purple spots with limited anthocyanin accumulation. However, combining Aft with the atv locus leads to a notable increase in anthocyanin content [17,38].

SlAN2-like functions as the master regulator of anthocyanin biosynthesis, while SlMYBATV (Solyc07g052490), a gene identified at the atv locus, acts as a passive repressor. SlMYBATV encodes an R3-MYB protein that negatively impacts anthocyanin accumulation by competing with SlAN2-like for binding to SlAN1, the bHLH component of the MBW complex, thereby disrupting its formation [38,57]. In the purple tomato variety Indigo Rose, the atv locus contains a nonfunctional version of SlMYBATV, which permits enhanced anthocyanin accumulation when combined with a functional activator such as SlAN2-like. Normally, functional SlMYBATV represses anthocyanin biosynthesis by competing with activators like SlAN2-like for binding to the bHLH component of the MBW complex. However, its nonfunctional state at the atv locus eliminates this repression, facilitating light-dependent activation of anthocyanin biosynthesis in the fruit skin [58].

The genetic basis of anthocyanin accumulation in our collection of tomato cultivars was dissected through GWAS analysis, which identified significant associations between SNPs on chromosomes 7 and 10 and the purple fruit phenotype. The identified genomic regions contained critical anthocyanin regulatory genes, SlMYBATV on chromosome 7 and SlANT1 on chromosome 10, corroborating previous findings about their roles in anthocyanin biosynthesis [18,38,39,40]. The identification of these loci aligns with earlier studies, where SlANT1 was highlighted as a master regulator of anthocyanin biosynthesis, while SlMYBATV served as a repressor competing with activators for bHLH binding within the MBW complex [19,57,58]. The derived LD heatmaps confirmed strong linkage disequilibrium (LD) within the SNP blocks containing SlMYBATV and SlANT1, allowing the selection of representative SNPs for genotyping. The high r² values among these SNPs ensured that genotyping a single SNP could reliably predict the alleles for the entire LD block. This redundancy simplifies the genotyping process, providing a cost-effective and reliable method for allele verification in breeding programs targeting anthocyanin-rich cultivars.

To validate these findings, HPM genotyping was employed to genotype selected SNPs, Ch07_60981501 and Ch10_64354129, in purple and non-purple tomato accessions. The technique, which offers high precision and minimal development requirements, clearly distinguished between purple and non-purple accessions based on melting temperature differences [59]. The melting curve analysis confirmed the homozygosity of all accessions, ensuring consistency in allele identification. The successful application of HPM genotyping and the clear differentiation of alleles between purple and non-purple cultivars underscore the functional significance of SlMYBATV and SlANT1 in regulating anthocyanin biosynthesis. These results are consistent with the observation that Aft, encoding SlAN2-like, and atv, encoding SlMYBATV, interact synergistically to enhance anthocyanin accumulation in tomato fruit. Specifically, the nonfunctional SlMYBATV allele at the atv locus in purple cultivars permits unimpeded activity of the anthocyanin activator SlAN2-like, driving light-dependent anthocyanin accumulation. This identification was parallel to the previous studies, which identified aft and atv loci through classical methods [37,38,39]. Our results also highlight the potential for marker-assisted selection (MAS) using the identified SNPs to facilitate breeding programs targeting anthocyanin-rich tomato varieties. The ability to reliably differentiate alleles based on SNP genotyping provides a valuable tool for breeders aiming to enhance nutritional quality and aesthetic appeal in tomatoes. Furthermore, the robust and reproducible nature of the HPM approach supports its broader application for genotyping other traits of interest in tomatoes and beyond.

5. Conclusions

This study builds upon previous findings regarding the genetic regulation of anthocyanin biosynthesis in tomatoes, offering novel insights through the application of GWAS and HPM for genotyping. While the roles of SlMYBATV and SlANT1 in regulating purple fruit phenotypes were previously known, this study uniquely integrates GWAS to identify SNPs associated with these loci and utilizes HPM for efficient allele identification. The development and validation of SNP markers represent a significant advancement in marker-assisted breeding programs for enhancing anthocyanin content in tomatoes. By refining our understanding of anthocyanin biosynthesis, this study establishes a practical framework for efficient genotyping and breeding of anthocyanin-rich cultivars. Future research should explore additional anthocyanin-related loci and genotype–environment interactions to optimize anthocyanin accumulation under diverse conditions.