Deciphering the genetic architecture of fruit color in strawberry

Abstract Fruits of Fragaria species usually have an appealing bright red color due to the accumulation of anthocyanins, water-soluble flavonoid pigments. Octoploid cultivated strawberry (Fragaria × ananassa) is a major horticultural crop for which fruit color and associated nutritional value are main breeding targets. Great diversity in fruit color intensity and pattern is observed not only in cultivated strawberry but also in wild relatives such as its octoploid progenitor F. chiloensis or the diploid woodland strawberry F. vesca, a model for fruit species in the Rosaceae. This review examines our understanding of fruit color formation in strawberry and how ongoing developments will advance it. Natural variations of fruit color as well as color changes during fruit development or in response to several cues have been used to explore the anthocyanin biosynthetic pathway and its regulation. So far, the successful identification of causal genetic variants has been largely driven by the availability of high-throughput genotyping tools and high-quality reference genomes of F. vesca and F. × ananassa. The current completion of haplotype-resolved genomes of F. × ananassa combined with QTL mapping will accelerate the exploitation of the untapped genetic diversity of fruit color and help translate the findings into strawberry improvement.


Introduction
Octoploid cultivated strawberry (Fragaria × ananassa) is a major horticultural crop that issued from a fortuitous cross in the 18th century in France between F. virginiana and F. chiloensis (Edger et al., 2019).Other species of the Fragaria genus cultivated for human consumption include the wild diploid woodland strawberry (F.vesca), which serves as a model for fruit crops of the Rosaceae family (Gaston et al., 2020).The cultivated and approximately 25 wild strawberry species currently described, which include species ranging from diploids to decaploids, exhibit a high genetic and phenotypic diversity that can be used to decipher the mechanisms responsible for strawberry fruit color (Castillejo et al., 2020;Qiao et al., 2021).
Considerable effort has been devoted in recent years to breeding cultivated strawberry varieties displaying not only increased resistance to pathogens, high yield, and extended storage period, but also improved fruit sensorial and nutritional quality (Senger et al., 2022).The main sensorial traits include fruit flavor and taste (aroma, sugars, organic acids) and appearance (size and shape, brightness, color).Specialized metabolites responsible for fruit color, mainly anthocyanins, also participate in fruit nutritional value (Mezzetti et al., 2018;Miller et al., 2019).As a result, fruit color, which plays a major role in the attractiveness and nutritional quality of strawberry, has been intensively investigated over the past decade.
Substantial advances have been made in our capacity to identify genetic variations responsible for differences in fruit color, thanks to progress in sequencing and genotyping technologies.Whole genome sequencing (WGS) of F. × ananassa led to the release of a reference genome (Edger et al., 2019) and, more recently, of several haplotype-resolved genomes (Hardigan et al., 2021a, Preprint;Lee et al., 2021;Fan et al., 2022;Mao et al., 2023).The implementation of high-throughput genotyping technologies (Bassil et al., 2015;Hardigan et al., 2020) now enables the highresolution mapping of fruit color quantitative trait loci (QTLs) and the identification of candidate genes (CGs) (Castillejo et al., 2020;Davik et al., 2020;Labadie et al., 2020Labadie et al., , 2022;;Pott et al., 2020;Manivannan et al., 2021).Excellent reviews (e.g.Whitaker et al., 2020) published recently describe in detail the present state of genetic and genomic studies in strawberry and their application to breeding.In this review, we focus on how our understanding of the genetic architecture of fruit color in strawberry can benefit from recent developments in fruit color phenotyping and in strawberry genomics.We further provide a current and forward-looking perspective on how these findings can be translated to improvement of strawberry fruit color.

The genus Fragaria is genetically diverse
Strawberry belongs to genus Fragaria, which contains about 25 species (Liston et al., 2014;Qiao et al., 2021) ranging from diploid (2n=2×=14) to decaploid (2n=10×=70).All species are restricted to single continents or specific areas, except the diploid woodland species F. vesca, which is found in both Eurasia and America (Staudt, 1962(Staudt, , 1989)).The majority of diploid species and all five tetraploid species are present in China (Liston et al., 2014).The single hexaploid species, F. moschata, is present in Europe, including Russia.The two related wild octoploid species, F. chiloensis and F. virginiana, are present in America with different distributions in South and North America (Staudt, 1962(Staudt, , 1989)).
Cultivated strawberry and its wild progenitors, F. virginiana and F. chiloensis, are allo-octoploids with 2n=8×=56 chromosomes (Box 1).Genetic diversity of F. × ananassa has been increased by recurrent interspecific hybridizations with various progenitor accessions to introduce new traits-of-interest (Hardigan et al., 2021b).This species results from a hybridization by chance in the early 1700s in Europe between two related octoploid species, F. virginiana and F. chiloensis (Darrow, 1966;Hancock, 1999), resulting in a new cultivated fruit species, F. × ananassa.Strawberry breeding began in England in the late 1700s, followed by France and Germany.The first selected European cultivars were used as genitors in early American breeding programs, together with American native cultivars (Darrow, 1966).Today, the genetic diversity in the cultivated strawberry seems to be preserved thanks to the large number of founders and introgression with wild octoploid strawberry germplasm (Hardigan et al., 2021b).Consequently, the genetic structure of this cultivated species displays a large amount of admixture (Bird et al., 2021;Zurn et al., 2022).
Fruit color diversity, ranging from white to dark red, is high in the genus Fragaria.For example, white fruits can be found in the diploid species F. vesca, F. nubicola, F. pentaphylla, and F. nilgerensis (Qiao et al. 2021) and in the octoploid F. chiloensis (Finn et al., 2013).In F. × ananassa, the diversity of skin color and flesh color pattern (e.g.white-fleshed genotypes versus red-fleshed genotypes; Fig. 1) is considerable (Aaby et al., 2012;Castillejo et al., 2020).As detailed in Box 1, recent developments in fruit color phenotyping and strawberry genomics now enable the high-resolution mapping and identification of causal genetic variations present amidst the strawberry genetic diversity.In addition, widespread gene editing technologies can simultaneously target the large numbers of homoeoalleles at a given locus found in cultivated strawberry, thus providing a technological means to modify fruit color.

Recent advances in strawberry genomics provide the key to exploring strawberry diversity
Reconstructing the evolutionary history of the Fragaria species was made possible thanks to the increased availability of strawberry genome sequences.Within the Fragaria genus, the first genome sequenced was that of the diploid F. vesca genotype 'Hawai-4' (Shulaev et al., 2011;Edger et al., 2018;Li et al., 2019), a white fruit-bearing genotype that has been widely used in studies involving stable genetic transformation (Gaston et al. 2020).Since then, high-quality chromosome-level reference genomes of a red-fruited F. vesca genotype (Alger et al., 2021) and of various diploid species (J.Zhang et al., 2020;Feng et al., 2021Feng et al., , 2023;;Qiao et al., 2021;Sun et al., 2022) have been made available.Recently, the reconstruction of the phylogeny of 10 wild diploid strawberries with whole genome resequencing data revealed complex and widespread introgression across the Fragaria genus (Feng et al., 2023).
The allo-octoploid F. × ananassa is complex, with four subgenomes derived from diploid species contributing to its genome (Rousseau-Gueutin et al., 2009).The first F. × ananassa genome sequence, that of the California variety 'Camarosa', confirmed that the diploid F. vesca and F. iinumae were ancestors of two of the four subgenomes (Edger et al., 2019).The origin of the remaining two subgenomes, possibly related to F. viridis and F. nipponica or to a now extinct species related to F. iinumae, is still under investigation (Edger et al., 2019(Edger et al., , 2020;;Liston et al., 2020).Until recently, this complexity has hampered the highresolution genetic mapping of traits, and the diploid F. vesca has been used in parallel as a strawberry model (Gaston et al., 2020).The suitability of F. vesca to explore the genetic control of fruit quality traits, including color, was further reinforced by expression analyses showing the dominance of the F. vesca subgenome for the control of fruit quality traits and, more specifically, anthocyanin biosynthesis (Edger et al., 2019).The recent availability of F. vesca and F. × ananassa genome sequences triggered the very quick development of genotyping technologies using single nucleotide polymorphism (SNP) arrays based on the F. vesca genome (Bassil et al., 2015).Subsequently, an 850K SNP array was developed from the F. × ananassa 'Camarosa' genome (Hardigan et al., 2020), from which a widely used (Senger et al., 2022) 50K array with selected chromosomespecific SNPs was designed (Hardigan et al., 2020).SNP array genotyping now allows the high-resolution localization of QTLs on specific chromosomes of F. × ananassa homeologous groups (HGs) (Box 1).
Once a QTL has been mapped at high-resolution, the next step in deciphering the genetic architecture of a trait is to identify the causal polymorphism.In F. × ananassa, up to eight candidate homoeoalleles carried by eight chromosomes can be found at a given locus (Gaston et al., 2020).A breakthrough advance to analyse the contribution of each homoeoallele to a color trait was the development of accurate and cost-effective longread technologies enabling the assembly and phasing of complex genomes (Dumschott et al., 2020;Hon et al., 2020) (Box 1).So far, haplotype-resolved genomes have been completed for four varieties of F. × ananassa ('Royal Royce', 'Wongyo', 'Florida Brillance', and 'Yanli') (Hardigan et al., 2021a, Preprint;Lee et al., 2021;Fan et al., 2022;Mao et al., 2023).Three of them are available at the Genome Database for Rosaceae (Jung et al., 2019).Haplotype-genome sequencing and transcriptome analysis of 'Yanli' has already demonstrated the potential of this approach to investigate mechanisms underlying anthocyanin biosynthesis and therefore to discover hitherto unknown natural genetic variations controlling fruit color.

Fruit color is a major visual and nutritional trait in strawberry
The red color of the strawberry fruit is due to the accumulation during ripening of red-colored anthocyanins.Anthocyanins are flavonoids derived from the phenylpropanoid pathway (Fig. 1).In addition to their role as water-soluble pigments conferring an attractive color to the fruit, anthocyanins are antioxidant molecules with major health benefits (Miller et al., 2019).As with other fruit traits, there is considerable variation between consumer preferences based on geographic region or country.In Asia, for example, white-skinned fruit varieties can be highly valued (Castillejo et al., 2020).Breeders should therefore adapt their objectives to the main market to be reached and consider the intensity, hue, and pattern of the fruit color.
When deciphering the architecture of fruit color, the first step is typically to phenotype large collections of genetic resources (Box 1).Most often, strawberry color is scored by (D) Individual phenylpropanoid, flavonoid and anthocyanin compounds can be exhaustively analysed by using MS-based technologies.(E) Automated phenotyping is key to reliably assessing the diversity of fruit color intensity and distribution in large collections (Zingaretti et al., 2022).
• Natural MYB10 variations (A) MYB10 is a master regulatory gene that targets structural genes of the anthocyanin biosynthetic pathway, resulting in the bright red color of strawberry fruit skin.(B) Various loss-of-function MYB10 mutations impair anthocyanin biosynthesis in F. × ananassa, resulting in white-skinned fruit (Castillejo et al., 2020;Wang et al., 2020;Yuan et al., 2022).(C) Transposon insertion in the MYB10 promoter induces ectopic expression of MYB10 in the fruit flesh, resulting in red-skinned and redfleshed fruit (Castillejo et al., 2020).
• Simultaneous editing of homoeoalleles with the CRISPR/Cas9 system (A) The CRISPR/Cas9 system targets specific gene sequences where it induces double strand breaks.DNA repair of the breaks can generate mutations.(B) In F. × ananassa, multiple homoeoalleles of an anthocyanin transport gene sharing common sequences could be targeted simultaneously (Gao et al., 2020).MYB 1-5-9-10-11-123  bHLH 3-9-25-29-33- visual comparison with a color chart (Labadie et al., 2022).In recent years, a key development was the automatization of the process whereby photographs of fruits taken in standardized conditions are analysed for intensity and pattern of fruit color (Zingaretti et al., 2022).Color hue is usually assessed by colorimeter with the CIELAB color space (Lerceteau-Köhler et al., 2012;Alarfaj et al., 2021).Destructive methods can measure the total anthocyanin content of a fruit extract and tens of different polyphenolic metabolites when using highly sensitive, but more laborious, mass spectrometry (MS)-based technologies (Ring et al., 2013;Urrutia et al., 2016;Crecelius et al. 2017;Pott et al., 2020;Wang et al., 2020).Detailed knowledge of the composition of anthocyanin-related chemicals in the fruit can further facilitate the identification of structural and regulatory CGs underlying color QTLs (Davik et al., 2020;Labadie et al., 2022).
In F. × ananassa, the main anthocyanin is pelargonidin 3-glucoside while cyanidin 3-glucoside is a minor component; the second most abundant anthocyanin is pelargonidin-3-malonylglucoside, which may represent more than 30% of the anthocyanins, depending on the cultivar (Aaby et al., 2012).Anthocyanin composition can be different in other strawberry species; for example, F. vesca accumulates preferentially cyanidin-glucosides (Urrutia et al., 2016).Leucoanthocyanidins are colorless while anthocyanidins and their derivatives are colored.The relative proportion of the various anthocyanins found in strawberry may have a profound effect on fruit color, with pelargonidin derivatives giving the fruit a bright red appearance and cyanidin derivatives a dark-red appearance (Fischer et al., 2014).
Following their synthesis on the cytoplasmic face of the endoplasmic reticulum, anthocyanins are transported into the vacuole for storage (Fig. 1).The color hue of anthocyanins is sensitive to pH, a process well studied in anthocyanin-colored flowers (Verweij et al., 2008).Therefore, vacuolar localization provides the low pH conditions required to express the intense red coloration of the fruit.Anthocyanin transport may involve several proteins, including glutathione S-transferases that can conjugate glutathione (GSH) to anthocyanins, thus facilitating anthocyanin transfer to the vacuole (Luo et al., 2018;Gao et al., 2020).
The MYB, basic helix-loop-helix (bHLH), and WD40 proteins that form the MBW complex are the main transcriptional regulators of the flavonoid biosynthetic pathway genes (Xu et al., 2015) in strawberry as in other plant species including Rosaceae species such as apple (Espley et al., 2007) (Fig. 1).In particular, the MYB family play key roles in the positive or negative regulation of proanthocyanidin and anthocyanin genes in developing strawberry (Schaart et al., 2013) and in response to hormonal (abscisic acid, ethylene) treatments (Zhao et al., 2018;Wang et al., 2020;Martínez-Rivas et al., 2023) and abiotic signals such as light and cold, heat, and water stresses (Medina-Puche et al., 2014;Kadomura-Ishikawa et al., 2015;Mao et al., 2022).Various additional transcription factors belonging to different families have been shown to modulate anthocyanin biosynthesis in strawberry (Fig. 1; Table 1).

Roles of structural anthocyanin-related genes in fruit color formation
Alterations in the expression of enzymes of the phenylpropanoid pathway can lead to changes in anthocyanin content and composition, and thus in the intensity, hue, and/or pattern of color of fruit tissues (Fig. 1).The color diversity found in strawberry genetic resources has been abundantly exploited in diploid and cultivated strawberries to investigate the role of structural and regulatory genes in anthocyanin accumulation, using up-to-date transcriptome and metabolome technologies (Tables 1, 2).An additional source of color diversity that has been exploited is changes in fruit color and associated modifications in the expression of regulatory and structural genes along fruit development and ripening (Tables 1, 2).
As shown in Table 1, since the early identification of phenylpropanoid and flavonoid pathway genes by Almeida et al. (2007), a large number of studies have investigated structural genes involved in anthocyanin biosynthesis and storage and branched pathways in strawberry.Various omics technologies including metabolomics (e.g.LC-electrospray ionization-MS or ultraperformance LC-MS/MS) and transcriptomics (microarrays, RNAseq) used alone or in combination have been used to investigate the contribution of the anthocyanin pathway and, more broadly, of the phenypropanoid pathway to fruit color in F. vesca and F. × ananassa.Sources of color variability were fruit developmental stages (green versus ripening stages), response to various signals (abscisic acid, light, temperature) and natural or artificially induced variations in skin and/or flesh color.Most findings have been functionally validated by transient expression assay in the fruit (Hoffmann et al., 2006), a technique well-suited for the visual assessment of the function of anthocyanin-related CGs by overexpression or RNAi silencing.Stable genetic transformation, which is much more time-consuming and tedious, has also been used in F. vesca and F. × ananassa to evaluate the function of various GCs and their natural variants, including by CRISPR/Cas9 gene editing (Gao et al., 2020) (Box 1).These studies demonstrated that the modulation of the expression of enzymes involved in phenylpropanoid and anthocyanin synthesis or transport effectively affects the quantity and/or composition of fruit anthocyanins.They further pinpointed the important impact on fruit color of the trade-off between the formation of colored anthocyanins and that of colorless compounds synthesized via branched pathways, an example of which is the synthesis by anthocyanidin reductase (ANR) of flavan-3-ols at the expense of anthocyanins (Fischer et al., 2014).

MYB10 emerges as a master regulator of anthocyanin content and pattern in the fruit
Similar genetic resources and strategies have been used to identify regulatory genes involved in the control of fruit color (Table 2).Their possible roles in the regulation of fruit anthocyanins have been further validated in planta by transgenic or transient gene expression assays and for transcription factors by mechanistic studies (e.g.protein-protein interactions and promoter-transcription factor binding assays).Not surprisingly, given the well-established role of the MBW ternary complex in phenylpropanoid regulation (Xu et al., 2015), various members of the MYB family have been shown to play a prominent role in anthocyanin accumulation in strawberry.In most cases, their gene targets could be determined (Fig. 1; Table 2).MYB10 regulates the expression of early biosynthesis and late biosynthesis genes and therefore emerged as a master regulator of fruit color in strawberry (Medina-Puche et al., 2014).Explorations of diverse accessions of natural white/yellowfruited genotypes of F. vesca, F. nilgerrensis, and F. × ananassa led to the discovery of various loss-of-function MYB10 mutations that impaired anthocyanin accumulation in the fruit (Hawkins et al., 2016;Zhang et al., 2017;Chen et al., 2020;Wang et al., 2020;Yuan et al., 2022).In addition, genome sequencing of numerous accessions of five diploid Fragaria species uncovered multiple independent single base mutations of the MYB10 gene associated with white fruit in several diploid species (Qiao et al., 2021).A key development was the systematic analysis of MYB10 alleles in F. vesca, F. chiloensis, and F. × ananassa (Castillejo et al., 2020)    allele responsible for the red-color of fruit flesh, from which a genetic marker was derived (Box 1).The gain-of-function is due to the insertion of a transposon in the MYB10 promoter that triggers anthocyanin biosynthesis.Haplotype sequencing of 'Yanli' coupled with transcriptome analysis of the ripe fruit provided further insight into the structural diversity and expression complexity of anthocyanin-related homoeoalleles (Mao et al., 2023).It showed, for example, that homoeoalleles of several anthocyanin structural genes were absent from one or several chromosomes while a dominant MYB10 homoeoallele located in a single chromosome from the F. iinumae subgenome was responsible for ~75% of MYB10 expression (Mao et al., 2023).Upstream in the regulatory network, a MAP kinase (MAPK3) was shown to reduce the MYB10 transcriptional activity in response to cold (Mao et al., 2022).
More MYB genes regulating the flavonoid pathway in strawberry fruit have been characterized recently, including MYB5, MYB9, MYB11, MYB123, and the negative regulator MYB1 (Table 2).Other members of the MBW complex that were studied include bHLH genes (bHLH3, bHLH9, bHLH33) and WD40 genes (Zhao et al., 2018;Yue et al. 2023).Additional transcription factors of various families including TCP11 and SLC8 (Pillet et al., 2015), HY5 (Li et al., 2020) and an AP2/ ERF (RAV1) (Z.Zhang et al., 2020) were also shown to be involved in the regulation of anthocyanins in strawberry.These results further demonstrate that, in addition to MYB10, many actors may participate in the regulation of the natural variations in fruit color observed in genetic resources.

Beyond MYB10: using untapped genetic diversity to discover new natural variations in fruit color
Thanks to the exploration of anthocyanin-related pathways and genes in Fragaria, the number of color-related CGs has greatly increased in recent years.These findings already provide a solid basis for understanding anthocyanin accumulation in the fruit.However, the demonstration by transient or stable genetic transformation that a CG affects fruit color is usually not sufficient to ascertain its value in the genetic control of color.Moreover, the octoploid status of F. × ananassa raises specific issues regarding loss-of-function mutations.As the inactivation of one or several homoeoalleles can be compensated by the expression of other homoeoalleles, loss-of-function mutations may have an impact on fruit color only if at most one or a few homoeoalleles are active in F. × ananassa.This is indeed the case for the various loss-of-function mutations found in the F. × ananassa MYB10 gene (Table 2).They likely produce white/yellow fruits in F. × ananassa (Table 2) only because at most two MY10 homoeoalleles of the same subgenome are expressed in the fruit (Castillejo et al., 2020;Mao et al., 2023).In contrast, genetic variations resulting in the overexpression or enhanced enzymatic activity of a single homoeoallele have likely a dominant effect on fruit color, whatever the number of active homoeoalleles in the genome.The few examples of CG-based genetic markers for color selection by markerassisted selection, e.g.MYB10 (Castillejo et al., 2020) or ANR (Labadie et al., 2022), were designed after natural genetic variations present in regulatory regions of the CGs.Information on the number and genomic localization of expressed CG homoeoalleles is therefore precious for the design of genetic markers, or the simultaneous edition of all the homoeoalleles (Gao et al., 2020).These findings also underscore the importance of using the large untapped diversity found in cultivated strawberry for color improvement.
A way to combine a CG catalog with the diversity of genetic resources is to analyse expression QTLs, with the aim to link natural variations of CG transcript level to color variations and thereby develop reliable markers for selection in F. × ananassa (Barbey et al., 2020) (Table 3).However, this approach limits the focus to known expressional CGs and does not allow the discovery of new genetic variants with original functions.To further explore the color diversity, biparental or multiparental (Wada et al., 2017) populations can be used to map chromosome regions responsible for quantitative variations of fruit color.Since the early studies (Zorrilla-Fontanesi et al., 2011;Lerceteau-Köhler et al., 2012) (Table 3), the mapping resolution of QTL analyses has been considerably increased and biochemical technologies now permit the identification and quantification of tens of secondary metabolites related to fruit color and nutritional value (e.g. in Pott et al., 2020) (Box 1).
The development of high-throughput genotyping arrays (Hardigan et al., 2020) now permits the high-resolution mapping of a color QTL to a narrow genomic region on a single subgenome (Davik et al., 2020;Pott et al., 2020;Manivannan et al., 2021;Labadie et al., 2022) (Table 3; Box 1).Once a CG has been identified in the QTL region, by mining literature or combining transcriptome and metabolome analyses with genetic data, WGS data obtained from the parents can be analysed to identify GC polymorphisms, design genetic markers and validate them for breeding.This strategy has been, for example, used to design an ANR-specific homeoallelic marker for a major color (pelargonidin-3-glucoside content) QTL (Labadie et al., 2022).High-resolution color QTLs can also be detected by screening genetic resources by genome-wide association studies (GWAS).A GWAS scans the genome for significant associations between genetic markers and the trait studied and is thus an alternative to the genetic mapping of QTLs.It can efficiently link color variations to the genetic information provided by genotyping arrays, as done for strawberry aroma (Fan et al., 2022), or by genotyping by sequencing (Vining et al., 2017).The recent completion of haplotyperesolved genomes for several F. × ananassa varieties (Hardigan et al., 2021a, Preprint;Lee et al., 2021;Fan et al., 2022;Mao et al., 2023) (Box 1) opens the way to accurately determining the chromosomal location and polymorphisms of the homeoallelic variants responsible for color variation.However, care must be taken the interpretation of GWAS data, as indirect association may produce statistically significant results at loci unrelated to the trait (Platt et al., 2010).

Towards the selection of strawberry varieties with improved fruit color
In addition to its potential value in understanding the mechanisms involved in fruit color, information on the genetic architecture of fruit color can be used to select strawberry varieties with desirable traits.Once identified and validated, genetic variations underlying a major color trait can be translated to strawberry color improvement through marker-assisted selection (Castillejo et al., 2020;Labadie et al., 2022).Because multiple homoeoalleles can be simultaneously edited by CRISPR/ Cas9 (Gao et al., 2020), gene editing can also be used to validate the findings or even engineer new strawberry varieties (Box 1).
Of course, to select superior strawberry varieties, breeders must take into account additional traits linked to the quality of the fruit (sweetness, acidity, aroma, shape, etc.) and the plant (resistance to pathogens, flowering pattern, yield, etc.) (Labadie et al., 2020;Whitaker et al., 2020;Senger et al., 2022).Combining several markers linked to various traits is an effective way of accelerating marker-assisted selection or selection by means of the genomic prediction models (genomic selection, GS) recently developed.GS links the traits studied, including fruit color, to genome-wide molecular markers to select the most suitable individuals from the population to breed new strawberry varieties with the desired traits.GS has been explored for strawberry resistance to Botritis cinerea (Petrasch et al., 2022) and Phytophtora cactorum (Jiménez et al., 2022) and fruit quality and yield (Gezan et al., 2017;Osorio et al., 2021).GS can use SNP markers and machine learning approaches to determine the genetic value of unseen lines for fruit quality traits, e.g.fruit color and shape in Solanaceae (Tong et al., 2022).As previous knowledge of causal genetic is not required, GS can be well adapted to the selection of color traits in strawberry where, except for major QTLs such as MYB10, color intensity is mostly governed by numerous small QTLs (Davik et al., 2020;Pott et al., 2020;Labadie et al., 2022).A recent example is the application of GS to the selection of strawberry F 1 hybrids with high fruit firmness or high pericarp color (Yamamoto et al., 2021).

Conclusion and perspectives
The wide diversity of fruit color intensity and pattern found in cultivated strawberry and its wild relatives can help decipher the genetic architecture of fruit color.Gaining insight into the genetic factors affecting natural variations in external (skin) and internal (flesh) fruit color is crucial for the efficient modification of this trait in new cultivars.Considerable work has been already done to identify structural and regulatory genes involved in the accumulation of red anthocyanins in the fruit.Among these is the MYB10 gene whose natural variations play key roles in clear-cut color variations of Fragaria species.
A recent breakthrough advance for the allo-octoploid F. × ananassa is the development of accurate long-read sequencing technologies that, when coupled with high-throughput genotyping, enable tackling of the untapped color diversity of strawberry genetic resources.Given the plummeting cost of these technologies (Eisenstein, 2023), we may envisage in the near future that haplotype-resolved genome sequences can be obtained for several of the genotypes under investigation, or even all of them.Providing that robust phenotyping technologies can be implemented, this should considerably accelerate the pace at which we decipher the genetic architecture of fruit color in strawberry.High-resolution mapping of QTLs to a distinct chromosome allows mining candidate genes and polymorphisms in a narrow genomic region and combining this information with expression analysis of homoeoalleles.A recent strawberry flavor study that combined multi-omics (transcriptome, metabolome), haplotype sequencing, and GWAS could successfully uncover fruit flavor genes and their regulatory elements (Fan et al., 2022), thus providing a roadmap for exploring the color of strawberry.In the near future, we may expect that, in addition to known color-related genes, genes of previously unknown function are isolated.While this should deepen our understanding of color control in strawberry, it can also be very challenging.

Box 1 .
Key developments in deciphering the genetic architecture of fruit color in strawberry • Strawberry color phenotyping (A) Examples of the diversity of fruit color intensity, hue, and pattern found in cultivated strawberry.(B, C) Fruit color intensity and hue can be evaluated by using (B) scoring-based color chart or (C) physical means (CIELAB color space).

Table 2 .
Regulatory genes with roles in anthocyanin accumulation in strawberry fruit Jiang et al. (2023) genetic marker Biparental populations derived from white-fruited F. vesca and white or red fleshed or skinned F. × ananassa or F. chiloensis genotypes MYB10 loss-of-function or gain-of-function alleles in all three Fragaria spp., functional validation, MYB10 gene targets in PP/Flav pathways b , metabolome Positive regulation of Anth and PA biosynthesis MYB5 gene targets in Flav pathwayMYB5F.×ananassaJiang et al. (2023)
a Microarray transcriptome analysis.b RNAseq transcriptome analysis.c qRT-PCR transcriptome analysis.