Novel Alleles of the Potato Leaf Gene Identified in Italian Traditional Varieties Conferring Potato-like Leaf Shape in Tomato

Abstract

The genetic diversity of tomato in Italy and the growing interest in high-quality food products highlight the importance of establishing varietal distinctiveness through molecular strategies to ensure agrifood product quality and traceability. In this study, four Italian potato-like leaf (PL) landraces were analyzed: “Spagnoletta di Formia e di Gaeta” (SPA) from southern Lazio, “Giagiù” (GIA) and “Patanara” (PTN) from Campania, and “Pomodoro di Mola” (MOL) from Apulia. These landraces were genotyped for the potato leaf gene (C), with two PL American genotypes and a non-allelic PL mutant line included as outgroups. Nagcarlang served as control. An allelism test confirmed C as determinant gene. The SCAR marker for C revealed that the Italian landraces presented determinants other than the most representative one responsible for PL. Whole-genome sequencing of SPA identified a private novel nonsense SNP variant allele, confirmed through dCAPS marker analysis. Additionally, two novel PL alleles responsible for missense variations were identified in GIA/PTN and MOL. In silico protein analysis suggested that novel C alleles could be functional determinants for the protein activity. Overall, PL mutations identified for the first time could serve as molecular tools for agrifood chain traceability, enabling early differentiation and recognition of genotypically similar varieties.

1. Introduction

Tomato (Solanum lycopersicum L.) is among the most widely cultivated vegetable crops worldwide [1] and serves as a model species in genetic studies due to its relative ease of cultivation and reproduction [2]. High phenotypic variability has been reported and studied in tomato species concerning fruit, reproductive, and vegetative traits [3]. Italy, as well as Spain, is a European center for tomato diversification [3,4,5]; numerous phenotypically differentiated landraces have been developed across all Italian regions and are protected by national and regional laws. In the Lazio region (central Italy), there is a considerable number of well-characterized traditional varieties [6], some of which are officially recognized [7]. Among these, “Spagnoletta di Formia e Gaeta” (SPA) is a landrace characterized by flat, ribbed, medium-sized fruits and potato-like leaves (PLs). This latter trait, marked by reduced leaf complexity compared to the wild type, features smooth leaf blade margins lacking lobes and serrations, making it a morphological marker useful for distinguishing SPA from other tomato landraces in Lazio [6]. The PL trait has been reported to positively affect both yield performance and fruit quality by increasing the soluble solid content of fruits [8]. The canonical gene determining the PL phenotype is encoded by Solyc06g074910, commonly known as the potato leaf (C) gene. It measures 2161 bp and consists of three exons on chromosome 6, encoding an R2R3 MYB transcription factor family protein [9]. A total of 11 C mutations have been reported in the literature, including eight single nucleotide polymorphisms (SNPs), two deletions, and one transposon insertion [8,9]. Among these, the latter represents the most common spontaneous mutation, consisting of a Rider retrotransposon element insertion of 4867 bp in the third exon of the gene, identified in heirloom tomatoes from the USA and Northern and Eastern Europe [8,9]. The Rider family is considered a major source of phenotypic variation in tomatoes, and the Rider insertion sequence for the PL trait has also been reported to be responsible for other known tomato mutations [10]. Alterations in leaf shape in tomatoes could be attributed to different genes [2]. Indeed, it has been reported that lines carrying a mutation known as the entire allele of the Aux/IAA family of transcription factors (TFs) in tomato (iaa9 mutants), obtained through TILLING or CRISPR-Cas9 technologies, exhibit the PL trait [11,12] similar to the potato leaf (C) phenotype (PL), which could lead to confusion between them. Currently, sequencing data for thousands of tomato accessions, including several Italian traditional varieties [13], obtained via next-generation sequencing technologies are publicly available [14] and can be used to study genetic variability to explore distinctiveness among genotypes. Additionally, whole-genome resequencing projects consisting of high-quality SNP datasets are highly useful for developing molecular markers. Based on these datasets, simple methods to target InDels and point mutations—such as designing sequence-characterized amplified regions (SCARs) and cleaved amplified polymorphic sequence (CAPS) markers—can be adopted as low-cost and powerful PCR-based analyses that reveal genetic polymorphisms. Specifically, SCAR markers allow the identification of polymorphic regions differing in size due to insertion or deletion via simple PCR amplification and agarose gel electrophoresis [15], whereas CAPS markers enable analysis of SNP mutations through PCR amplification followed by enzyme-mediated restriction, specifically by cutting the amplicon at the mutation site [16]. The derived CAPS (dCAPS) is an upgrade based on amplified PCR fragments from mismatched primers to generate restriction sites digested by an appropriate restriction enzyme [17,18].

This work adopted SCAR and dCAPS marker analysis to explore the molecular determinants of locus C mutations in SPA and other Italian landraces characterized by the PL phenotype. The aim was to identify unknown mutations responsible for the PL phenotype involved in functional stability changes in mutant proteins and potentially suitable for early differentiation and recognition of phenotypically similar elite landraces, ensuring their protection and traceability along the agri-food chain.

2. Materials and Methods

2.1. Plant Material

Four Italian tomato landraces characterized by the potato-like leaf (PL) phenotype were analyzed: “Spagnoletta di Formia e di Gaeta” (SPA) from the Lazio region (central Italy), “Patanara” (PTN) and “GiaGiù” (GIA) from Campania, and “Pomodoro di Mola” (MOL) from the Apulia region (southern Italy) (Figure 1). Additionally, two PL American landraces, LA2374 (LA) and PC711571 (PC), obtained from the C.M. Rick Tomato Genetics Resource Center (TGRC), along with the “Red Setter” iaa9 mutant isogenic line carrying the entire mutation (IAA) [11], were included as outgroups. The commercial variety “Nagcarlang” (NAG) served as a control for the “wild-type” leaf shape (WT). All seed materials were sourced from the DAFNE Seed Bank of the University of Tuscia (Table S1).

2.2. Growth Conditions and Allelism Test

In 2023, the plant material was cultivated in a greenhouse at the experimental farm of the University of Tuscia “Nello Lupori” in Viterbo (42°42′66″ N, 12°08′03″ E). During the flowering stage, immature flowers from four SPA plants were emasculated and subsequently hand-crossed with mature pollen collected from six PL genotypes (GIA, IAA, LA, MOL, PC, PTN) and the wild-type (WT) line (NAG) to establish seven F₁ hybrid progenies. At the ripening stage, seeds were harvested, and eight seeds of each F₁ hybrid and their parental lines were sown in Petri dishes with paper filters soaked in double-distilled water (ddH₂O), incubated at 23 °C in the dark until the emergence of 1-cm radicles. After emergence, seedlings were transplanted into Jiffy pots. Phenotypic evaluation of leaf shape was performed 20 days post-germination on the parent and F₁ progeny plants by acquiring images using Perfection 2480 photo scanner (EPSON Corp., Suwa, Japan) of the second and third leaves.

2.3. DNA Extraction

Fresh leaves were collected from each F₁ hybrid and parental line and ground in liquid nitrogen. Approximately 150–200 mg of frozen powder was used to extract genomic DNA (gDNA) using a modified protocol of the cetyltriethylammonium bromide (CTAB) extraction method [19]. The gDNA quantity was evaluated using a Qubit fluorometer (Life Technologies, Carlsbad, CA, USA) and a Qubit dsDNA BR Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA) following the manufacturer’s instructions.

2.4. Design of SCAR and dCAPS Molecular Markers

To evaluate the putative presence of the Rider transposon insertion in Solyc06g074910 (C gene) in PL genotypes, a primer pair was designed upstream and downstream of the putative Rider transposon insertion site (Table S2) located at 44,079,695 bp (Tomato genome version SL4.0), measuring 4867 bp [8]. The putative Rider genomic region (accession number: EU195798) sequences were retrieved and manually inserted at the breakpoint position in the third exon of the C gene as previously identified [9,14]. The PCR were conducted in a final volume of 10 µL, containing 1 µL of gDNA (20 ng/µL), 1 µL of each primer (10 pmol/μL), 5 µL of 2× Green GoTaq buffer (Promega Corporation, Madison, WI, USA), and 2 µL of ddH₂O. The PCR conditions included an initial denaturation step at 95 °C for 7 min, followed by 30 cycles at 95 °C for 30 s, 60 °C for 30 s, and 72 °C for 30 s, with a final elongation step at 72 °C for 7 min. To investigate other mutations in the C gene associated with PL, data from the “360 tomato variants” project were analyzed via raw reads from SPA, MOL, and NAG. The raw data were sourced from the NCBI Sequence Read Archive under accession SRP045767 [13] and mapped against the tomato genome (Tomato genome version SL4.0, annotation ITAG4.0) via a modified pipeline reported by [20]. A marker located within the coding region of C was retained for analysis. SNPeff software version 4.3T [21] was used for gene annotation and prediction of potential effects of identified SNPs on protein function. A dCAPS marker based on the SNP was designed using primer pairs with a mismatch at the 24th nucleotide of the forward primer targeting a region of length. 161 bp (Table S2). The PCR mixture was prepared as above, except for increasing the number of cycles to 35. The amplicons were subsequently digested with TaqI (New England Biolabs, Ipswich, MA, USA) at 37 °C for 8 h. The resulting digestion pattern was visualized using a UV light-transilluminator system with ethidium bromide used for DNA staining. dCAPS analysis was extended to assess the presence of the SPA allele in different SPA accessions and to confirm breeding events in F₁ progeny during the allelism test.

2.5. Sanger DNA Sequencing

For Sanger sequencing of seven accessions, primer pairs were designed to generate four different overlapping amplicons covering the whole C gene (Table S2). PCR was performed in a final volume of 20 μL containing 1 µL of gDNA (20 ng/µL), 2 µL of each primer (10 pmol/μL), 10 µL of 2× Green GoTaq buffer (Promega Corporation, Madison, WI, USA), and 7 µL of ddH₂O. The PCR conditions included an initial denaturation step at 95 °C for 7 min followed by 35 cycles at 95 °C for 30 s, an annealing temperature for each primer pair for 30 s, and elongation at 72 °C for 30 s with a final elongation step at 72 °C for another 7 min. The annealing temperature depended on each primer pair (Table S2). The PCR amplicons were purified using the NucleoSpin™ Gel and PCR Clean-up Kit (Macherey-Nagel, Düren, Germany) and sent for Sanger sequencing (Eurofins Genomics, Luxemburg City, Luxemburg).

2.6. In Silico Gene and Protein Structure Analysis

To analyze sequencing data obtained from Sanger sequencing, CDSs were identified via multiple alignment of the seven landraces against the Heinz1706 reference gene (Tomato version SL4.0) [14] using ClustalX 2.1 software [22]. To assess the identification of novel allele variants in the C gene, identified CDS of each landrace were realigned against those carrying known SNP mutations [8,9] To evaluate whether SNP mutations detected by sequencing could influence protein function, Ensembl Variant Effect Predictor online tool [23] was queried. ExPASy [24] was used to translate the predicted CDSs carrying PL phenotype to obtain the WT and mutant protein sequences. Multiple alignment on the resulting protein sequences were undergone via ClustalX 2.1. To assess amino acid (AA) substitution impacts on protein structure and function, predicted WT three-dimensional (3D) protein structure via AlphaFold [25], resulting.pdb file was downloaded. A BlastP analysis on Arabidopsis Information Resource (TAIR) database, release 10.0 [26] and ClustalX 2.1 alignment were performed to identify Arabidopsis thaliana homolog proteins followed by querying SWISS-MODEL (https://swissmodel.expasy.org/interactive, accessed on 17 November 2024). The transcription factor WER was used as best predicted model (QMEANDisCo: 0.77 ± 0.08) [27]. Main structural domains were identified; Arabidopsis protein model served as WT tomato protein comparison model. Single AA substitutions were introduced into tomato protein model via PyMol mutagenesis tool [28]. Structural changes analyzed by calculating distances between mutant AA and the surrounding functional AA chain sides through PyMol distance measurement tool. Furthermore, mutation effects on protein stability and dynamics were evaluated via Dynamut2 webserver tool [29]. All Arabidopsis and tomato WT and mutant protein structures were visualized via PyMOL v. 3.0.5.

3. Results and Discussion

3.1. Allelism Test Assessing C Involvement in the PL Phenotype

Since the potato leaf trait associated with the C gene is considered recessive [9], an allelism test was performed to assess whether SPA and other Italian landraces included in this study (i.e., GIA, MOL, and PTN) showing a PL phenotype carried other mutations in this gene.

In the allelism test, SPA was crossed with three Italian landraces, two American accessions, and an IAA mutant line carrying the entire allele of Solyc04g076850, which is also associated with the PL phenotype [11]; WT NAG served as a control for the crosses. All F₁ plants obtained from SPA and the three Italian landraces, as well as the two American PL landraces, presented the PL phenotype, whereas plants derived from the crosses SPA × NAG and SPA × IAA presented the WT phenotype, suggesting a key role of the C gene in phenotype expression (Figure 2).

3.2. Molecular Integrated Approaches Revealed Novel SPA Private PL Alleles

As previously reported, the most representative mutation for the WT-to-PL transition in cultivated tomato consists of a retrotransposon Rider insertion in the third exon of the C gene [9]. The tomato landrace group showing the PL phenotype was screened for Rider insertion, to detect potential novel causal alleles for this gene. The dominant SCAR marker revealed that only LA and PC showed the Rider insertion in the C gene, as indicated by the absence of amplification in the target region (Figure 3a). The reverse primer was designed at the end of the third exon, beyond the Rider insertion. The expected amplicon size of 605 bp was predicted only for genotypes that lacked of insertions (i.e., WT genotypes) and that showed the WT leaf phenotype (i.e., NAG). On the other hand, no amplification was expected for PL genotypes carrying the Rider insertion, due to its considerable size (5472 bp) for standard PCR. The results indicated that in the PL genotype, other genetic determinants were involved in controlling leaf shape. Since the PL amplicon pattern of the Italian landraces was similar to that of the WT (NAG), the presence of mutations other than the Rider insertion was expected. Thus, publicly available data for SPA, MOL, and NAG data [13] were retrieved from [10], filtered and mapped against a newer tomato genome version (Tomato Genome version SL4.0) to identify variants falling within the C gene sequence. The choice to retrieve only two PL accessions (i.e., SPA and MOL) was due to the unavailability of data for other PL accessions (i.e., GIA and PTN). The analysis highlighted a SNP falling in the coding region of the gene of interest at position 44,079,418 (G to A substitution), which is private to SPA. A predicted HIGH impact effect on protein function corresponding to a premature stop codon was found as a result of this identified SNP. At the same position, MOL and NAG showed the Heinz reference allele, suggesting the presence of additional determinants different from both the Rider insertion and the SPA private allele that could cause the PL phenotype in MOL (Table S3).

To confirm the SPA private SNP identified by in silico analysis, a dCAPS marker was developed and analyzed on PL genotypes to validate the identification of this private mutation. According to in silico analysis, the expected restriction patterns corresponded to a single fragment of 161 bp for genotypes without the mutation and two fragments of 137 bp and 24 bp for those carrying the SNP. PCR analysis successfully amplified the expected fragment in all analyzed genotypes (Figure 3b), and subsequent restriction analysis revealed that only SPA presented the predicted restriction band at 137 bp, whereas no cut fragments were generated in other genotypes (Figure 3c). Notably, the fragment measuring 24 bp did not appear due to its low molecular weight. The TaqI cutting site (5′-TCGA-3′) introduced in the forward primer occurred only on SPA amplified sequence, specifically from positions 44,079,415 to 44,079,418, owing to a G to A substitution at the last position of the this cutting site. This sequence was present only in SPA amplicon, confirming identification of a SPA putative determinant for the PL phenotype and suggesting other C-related alleles for the remaining three Italian landraces.

The private mutation of SPA is different from those previously reported [8,9], suggesting the identification of novel alleles associated to the SPA Italian landrace. To verify the preservation of this mutation within the SPA variety, dCAPS analysis was performed on 11 different SPA accessions, and all results confirmed the private SNP pattern. Moreover, the F₁ hybrids obtained for the allelism test were also screened for the SPA C allele to assess the heterozygous condition. The analysis confirmed the presence of both bands, which were separated by 24 bp, as expected. This finding demonstrated for the first time the existence of a new allele generated by spontaneous mutations in the C gene, characterized by a premature stop codon.

3.3. Sanger Sequencing Analysis Highlighted Two Other Novel PL Alleles in Italian Landraces

Until now, the only known spontaneous point mutation has consisted of a missense variant that alters the amino acid sequence [8], whereas other mutations were obtained solely from mutagenesis approaches. Although the SPA allele was found and represents an important source of variability associated with this trait, there were other unexplored genetic variations in GIA, MOL, and PTN. Thus, a paired-end Sanger sequencing experiment was performed on the seven genotypes under investigation to identify additional SNPs in the coding region functionally involved in the PL phenotype, aiming at developing molecular markers useful for early genotypic distinctiveness. After sequence alignment, a novel SNP at 44,078,499 bp (G to A) was found in GIA and PTN within the second exon. In the third exon, another novel allele was identified in MOL at position 44,079,418, consisting of a transition (C to T). Finally, sequencing confirmed the private G to A transition at 44,079,418 bp for SPA (Figure 4), as evidenced by the dCAPS results.

The sequencing experiment also confirmed that in LA and PC, the gene sequence was truncated at positions 44,079,610, which is near the Rider insertion site (Figure S3). Indeed, the primer pairs were designed upstream (overlapping with the third amplicon) and downstream of the putative Rider insertion region, confirming the genetic determinant for the PL phenotypes in these two American landraces. Moreover, for all the genotypes, no polymorphisms were detected in the introns of the genes (Figure S3). As previously reported [9], other SNPs were found to be responsible for the PL phenotypes, even though some of them were obtained via induced mutagenesis strategies, and only one was spontaneous [8].

In this work, the spontaneous origin of the identified mutations was proven by two main lines of evidence. The first is related to the nature of the material under investigation, namely landraces, which have been defined as “dynamic populations undergoing several naturally occurring mutations contributing to genetic variability” [30]. Additionally, alignment of the predicted CDS of each accession against sequences carrying known mutations in the C gene was carried out. The results confirmed the novelty and spontaneity of the identified SNPs in the CDS region found at positions G152A, C293T, and G327A for GIA/PTN, MOL, and SPA, respectively (Figure S4). These correspond to the second (GIA/PTN and MOL) or third (SPA) nucleotide of codons 51, 98, and 109, respectively. The mutations resulted in two missense variants (GIA/PTN and MOL) and a nonsense mutation (SPA) (Table S4). This suggests their putative role in the expression of the PL phenotype.

3.4. SNP Mutations Affect the R2R3 Domain of the Myb Factor

To evaluate the functional effects of the novel identified mutations on protein function, the Variant Effect Predictor (VEP) tool of the Ensembl plant database was queried for all novel C alleles. In SPA, the predicted introduction of an early stop codon prevents the synthesis of a complete protein, resulting in a loss-of-function protein (Figure S5), which induces ribosome dissociation from mRNA and transcript degradation to prevent the accumulation of toxic protein waste in the cell [31]. In contrast, in GIA/PTN and MOL, the missense variants produced alterations in the protein sequences, resulting in G51E and P98L, respectively. These changes are considered deleterious mutations according to the VEP SIFT score output, which is equal to 0 (Table S4). Indeed, a SIFT score ≤ 0.05 indicates that the mutation is harmful and has an adverse impact [32].

A 3D structure prediction of the WT and the two mutant forms of the protein was subsequently performed. For this purpose, Arabidopsis and tomato WT protein 3D models were built. BlastP analysis via TAIR and ClustalX2.1 alignment identified the SlMyb36 transcription factor as the orthologous protein in Arabidopsis and revealed that the most conserved region ranged from 13 to 117 AA (Figure 5), corresponding to the R2R3 functional domains that constitute the DNA binding domain (DBD) (Figure S6a,b) [27,33]. In both AtMyb36 and SlMyb36 proteins, these regions contained three alpha helices named H1, H2, and H3 for R2 (Figure S6c) and H4, H5, and H6 for R3 (Figure S6d), characterized by a hydrophobic core made up of three highly conserved and regularly spaced tryptophan (W) residues. In the H4 helix, W is replaced by a phenylalanine (F) in both tomato and Arabidopsis.

With respect to the structural conformation of WT SlMyb36, the G51 nonpolar side chain is located in the inner, hydrophobic part of the H3 domain, distancing 10.8 Å, 5.6 Å, and 11.3 Å from the tryptophan rings of the R2 repeat (at positions W17, W37, and W57, respectively) (Figure 6a).

In GIA and PTN, the G51E substitution in the R2 repeat increased the distance between W17, W37, and W57 and the residue at position 51 (Figure 6b) by 2 Å due to its negatively charged side chain. This alteration affects the normal binding activity because it modifies the third helix of R2, which is involved in recognizing the major groove of the DNA [27,33]. For the R3 repeat, the rigid ring of P98 is turned inward of the hydrophobic R3 core, distancing 11.0 Å, 15.0 Å, and 17.9 Å from residues F71, W90 and W109, respectively (Figure 6c), This positioning plays an important role in regulating the correct protein fold and conferring rigidity to the protein [34,35,36].

In MOL, the P98L substitution resulted in hydrophobic, aliphatic side chains that favor the formation of an alpha-helical structure, thus reducing protein backbone rigidity [37,38]. Additionally, an L size greater than P may cause bumps, further promoting a structural change in the R3 domain (Figure 6d).

Single point mutations within a protein may result in conformation impairments affecting the kinetics of protein folding or causing protein destabilization [39]. Therefore, the difference in Gibbs free energy (ΔΔG) between the wild-type and mutant proteins was assessed to predict the changes caused by the AA substitutions found in GIA/PTN and MOL. The main interatomic contacts and noncovalent interactions among the G51 residue and the surrounding amino acids in the 3D wild-type protein conformation included polar interactions, hydrogen bonds, and van der Waals forces (VDW) (Figure S7a and Video S1a). The G51E substitution would result in a greater number of different kinds of interactions (Figure S7b, Video S1b) that led to strong destabilization of the mutant protein in comparison to its WT counterpart, as indicated by the negative ΔΔG value (−1.55 kcal/mol) (Figure S7b) [40]. A negative ΔΔG value was also obtained when the P98L substitution was considered (Figure S7d). However, the lower absolute value of ΔΔG found in this case (−0.23 kcal/mol) indicated that the P98L mutation destabilized the MOL mutant protein to a minor extent (Figure S7c,d). Hence, protein destabilization may be primarily due to the loss of protein backbone rigidity. It negatively impact on protein DNA recognition and binding ability due to the P-cyclized structure in the WT protein and the bumps introduced by L. which. Considering the relatively lower protein destabilization resulting from the P98L substitution compared to the G51E substitution, it could be hypothesized that a partially functional SlMyb36 protein may be synthesized. However, experimental studies are needed to validate this assumption.

3.5. Importance of Preserving and Protecting Elite Landraces

These findings highlight the importance of preserving and protecting landraces as essential sources of natural genetic variation, thus overcoming mutagenesis steps and maintaining the quality of traditional products or developing novel improved competitive varieties. The three novel C alleles originating from spontaneous mutations in Italian landraces have been maintained by both farmer selection and environmental adaptation but not through breeding activities, similar to what was previously reported in Prudence Purple Heirloom [8]. Phenotypically, all these landraces presented the green shoulder trait, an appreciable and ancient trait indicating their local and not modern or improved variety nature. GIA and PTN are traditional varieties known as “Piennolo del Vesuvio” cultivated in Campania; these varieties produce the same small, pear-shaped fruit type, differing only in fruit color. Indeed, due to a pointed mutation in the psy1 gene at the splice donor site of the exon-intron junction [41], GIA results in a yellow tomato [16], which differs from the red fruit (Figure 7a,b) of PTN. Interestingly, the similar fruit shape, comparable soluble solid content [16,41,42,43], region of origin (Figure 1), and the same SNP allele on C gene in GIA and PTN found in this work suggest a common origin. SPA is a known traditional variety that originated in southern Lazio, is listed in the Italian “Traditional Agro-Food Product” (PAT) registry, and is highly appreciated for its taste and sauce [6] and characterized by flat-ribbed fruit (Figure 7c). MOL is a traditional landrace originating in the Apulia region, known as “Pomodoro di Mola”, cultivated along the coast and regularly registered in the PAT registry [44]. It produces elongated fruits appreciated for its tomato sauce [45] (Figure 7d). Moreover, the PL phenotype was not related to the fruit shape of tomato fruits. Indeed, the Italian accessions showed three different fruit shapes but the same leaf phenotype. Furthermore, the fruit shape of Pruden’s Purple [8] is similar to that of SPA, while the SNP determinant in the C gene is different, indicating no correlation between fruit and leaf shape.

The four accessions, in addition to having a common PL phenotype, originated close to coastal zones. An improvement in the total soluble solid (TSS) content in accessions with a PL phenotype has been reported [8,46]. This could be due to the greater leaf area intercepting sunlight [47]. Thus, a positive correlation between the PL phenotype and the TSS value could be hypothesized. Moreover, the cultivation area can further enhance the organoleptic traits of tomato fruits because of the coastal environment [48,49]. For example, a TSS value greater than 7.0° Brix was found for GIA and PTN, while a value of 6.5° Brix was found for SPA [3,50]. The interest of this trait is due to its putative involvement in the yield improvement of tomato. Indeed, it is reported that PL phenotype has a positive correlation with yield, leading to increase yield as well as Brix [8]. The presence of PL private SNPs identified in Italian landraces could represent a useful molecular tool suitable not only for varietal distinctiveness but also for landrace traceability, protection, and product quality control within the entire agrifood chain.

4. Conclusions

In this work, three new spontaneous SNPs were identified for the first time as potentially responsible for the tomato PL phenotype. All these mutations were found in four Italian landraces residing in coastal areas across three different regions, representing important sources of genetic variability. Given that these mutations are hidden in local, traditional plant resources and were demonstrated to be primary alleles for the PL phenotype, the need arises for a program with the aim to protect and conserve these landraces. The mutations were demonstrated to be primary alleles for the PL phenotype in these landraces. The PL trait, together with the area of origin of the landraces under investigation, could influence fruit quality, especially TSS. Molecular markers specifically developed in this study represent useful tools to genetically recognize and differentiate products derived from elite traditional varieties within the agri-food chain. This ensures the authenticity and protection of both raw and transformed products. Additionally, the recessive nature of the C gene and its different allelic forms can be advantageous in breeding programs for marker-assisted selection, as they can be harnessed for design codominant markers to discriminate heterozygous progeny and assess crossing early in the development of potential improved F₁ hybrids.