SUMMARY Tomato is a widely cultivated crop, which can grow in many environments. However, temperature above 30°C impairs its reproduction, subsequently impacting fruit yield. We assessed the impact of high‐temperature stress (HS) in two tomato experimental populations, a multi‐parental advanced generation intercross (MAGIC) population and a core‐collection (CC) of small‐fruited tomato accessions. Both populations were evaluated for 11 traits related to yield components, phenology and fruit quality in optimal and HS conditions. HS significantly impacted all traits in both populations, but a few genotypes with stable yield under HS were identified. A plasticity index was computed for each individual to measure the extent of the heat impact for each trait. Quantitative trait loci (QTL) were detected in control and HS conditions as well as for plasticity index. Linkage and genome‐wide association analyses in the MAGIC and CC populations identified a total of 98 and 166 QTLs, respectively. Taking the two populations together, 69 plasticity QTLs (pQTLs) were involved in tomato heat response for 11 traits. The transcriptome changes in the ovary of six genotypes with contrasted responses to HS were studied, and 837 genes differentially expressed according to the conditions were detected. Combined with previous transcriptome studies, these results were used to propose candidate genes for HS response QTLs.

SummaryIn this study, we looked for genetic factors in the pepper (Capsicum annuum) germplasm that control the number of potato virus Y (PVY) particles entering the plant (i.e. effective population size at inoculation) and the PVY accumulation at the systemic level (i.e. census population size). Using genotyping‐by‐sequencing (GBS) in a core collection of 256 pepper accessions, we obtained 10 307 single nucleotide polymorphisms (SNPs) covering the whole genome. Genome‐wide association studies (GWAS) detected seven SNPs significantly associated with the virus population size at inoculation and/or systemic level on chromosomes 4, 6, 9 and 12. Two SNPs on chromosome 4 associated with both PVY population sizes map closely to the major resistance gene pvr2 encoding the eukaryotic initiation factor 4E. No obvious candidates for resistance were identified in the confidence intervals for the other chromosomes. SNPs detected on chromosomes 6 and 12 colocalized with resistance quantitative trait loci (QTLs) previously identified with a biparental population. These results show the efficiency of GBS and GWAS in C. annuum, indicate highly consistent results between GWAS and classical QTL mapping, and suggest that resistance QTLs identified with a biparental population are representative of a much larger collection of pepper accessions. Moreover, the resistance alleles at these different loci were more frequently combined than expected by chance in the core collection, indicating widespread pyramiding of resistance QTLs and widespread combination of resistance QTLs and major effect genes. Such pyramiding may increase resistance efficiency and/or durability.

Characterizing the natural diversity of gene expression across environments is an important step in understanding how genotype-by-environment interactions shape phenotypes. Here, we analyzed the impact of water deficit onto gene expression levels in tomato at the genome-wide scale. We sequenced the transcriptome of growing leaves and fruit pericarps at cell expansion stage in a cherry and a large fruited accession and their F hybrid grown under two watering regimes. Gene expression levels were steadily affected by the genotype and the watering regime. Whereas phenotypes showed mostly additive inheritance, ~80% of the genes displayed non-additive inheritance. By comparing allele-specific expression (ASE) in the F hybrid to the allelic expression in both parental lines, respectively, 3005 genes in leaf and 2857 genes in fruit deviated from 1:1 ratio independently of the watering regime. Among these genes, ~55% were controlled by cis factors, ~25% by trans factors and ~20% by a combination of both types of factors. A total of 328 genes in leaf and 113 in fruit exhibited significant ASE-by-watering regime interaction, among which ~80% presented trans-by-watering regime interaction, suggesting a response to water deficit mediated through a majority of trans-acting loci in tomato. We cross-validated the expression levels of 274 transcripts in fruit and leaves of 124 recombinant inbred lines (RILs) and identified 163 expression quantitative trait loci (eQTLs) mostly confirming the divergences identified by ASE. Combining phenotypic and expression data, we observed a complex network of variation between genes encoding enzymes involved in the sugar metabolism.

Flavour and nutritional quality are important goals for tomato breeders. This study aimed to shed light upon transgressive behaviors for fruit metabolic content. We studied the metabolic contents of 44 volatile organic compounds (VOCs), 18 polyphenolics, together with transcriptome profiles in a factorial design comprising six parental lines and their 14 F1 hybrids (HF1) among which were five pairs of reciprocal HF1. After cluster analyses of the metabolome dataset and co-expression network construction of the transcriptome dataset, we characterized the mode of inheritance of each component. Both overall and per-cross mode of inheritance analyses revealed as many additive and non-additive modes of inheritance with few reciprocal effects. Up to 66% of metabolites displayed transgressions in a HF1 relative to parental values. Analysis of the modes of inheritance of metabolites revealed that: (i) transgressions were mostly of a single type whichever the cross and poorly correlated to the genetic distance between parental lines; (ii) modes of inheritance were scarcely consistent between the 14 crosses but metabolites belonging to the same cluster displayed similar modes of inheritance for a given cross. Integrating metabolome, transcriptome and modes of inheritance analyses suggested a few candidate genes that may drive important changes in fruit VOC contents.

Water deficit (WD) leads to significant phenotypic changes in crops resulting from complex stress regulation mechanisms involving responses at the physiological, biochemical and molecular levels. Tomato growth and fruit quality have been shown to be significantly affected by WD stress. Understanding the molecular mechanism underlying response to WD is crucial to develop tomato cultivars with relatively high performance under low watering conditions. Transcriptome response to WD was investigated through the RNA sequencing of fruit and leaves in eight accessions grown under two irrigation conditions, in order to get insight into the complex genetic regulation of WD response in tomato. Significant differences in genotype WD response were first observed at the phenotypic level for fruit composition and plant development traits. At the transcriptome level, a total of 14,065 differentially expressed genes (DEGs) in response to WD were detected, among which 7393 (53%) and 11,059 (79%) were genotype- and organ-specific, respectively. Water deficit induced transcriptome variations much stronger in leaves than in fruit. A significant effect of the genetic background on expression variation was observed compared to the WD effect, along with the presence of a set of genes showing a significant genotype x watering regime interaction. Integrating the DEGs with previously identified WD response quantitative trait loci (QTLs) mapped in a multi-parental population derived from the crossing of the eight genotypes narrowed the candidate gene lists to within the confidence intervals surrounding the QTLs. The results present valuable resources for further study to decipher the genetic determinants of tomato response to WD.

Tomato is an important crop worldwide and one of the major sources of vitamin C (ascorbate) in the human diet. Ascorbate contributes both to tomato nutritional quality and has roles in stress tolerance and adaptation to the environment. In this study we show the variability that exists in tomato germplasm in terms of ascorbate content (10 to 90 mg/100 g fwt) which could be starting point for evaluating correlations with physiological traits potentially linked to ascorbate. We have then manipulated genes involved in ascorbate metabolism, using RNA interference, to investigate their influence on fruit ascorbate levels, fruit physiology and yield under both normal and drought stress conditions. There is some evidence that one of the genes chosen affects fruit size and yield under different conditions, but clear effects on the fruit ascorbate pool are not seen. We conclude that ascorbate metabolism is complex and can also have wider effects on fruit physiology and growth.