Genomics

Anthocyanin biosynthesis is regulated through the interaction of the MYB-bHLH-WD40 complex, which has already been characterized in several crops from diverse plant families, including some Rosaceae species. Using genome wide phylogenetic analysis, we identified three possible bHLH candidates from F. vesca and two from R. idaeus based on sequence homology with reported bHLH proteins from Malus, Arabidopsis and Fragaria. To date it is not clear which of these bHLH proteins are regulators of anthocyanin biosynthesis, and how this regulation is occurring in strawberries and raspberries. Here we show the results of gene expression studies of these bHLH genes as well as of putative anthocyanin biosynthetic genes at several fruit developmental stages. In addition, we show the results of promoter activation studies of “key” genes of the anthocyanin pathway like CHS, DFR, UFGT. Gene activation was determined by co-transformation of the promoter regions of biosynthetic genes fused to the luciferase gene and bHLH genes under 35S promoter expression in Nicotiana benthamiana, followed by luciferase transient assay; we determined that some of the bHLH candidate proteins are essential components for activation of key genes inolved in the anthocyanin pathway.

Fruit development and maturation involve many physiological and biochemical changes. Some apple cultivars show undesirable ripening process leading to mealiness, which is characterized by texture deterioration resulting in soft, grainy and dry fruit.

To understand the molecular bases of mealiness development, we used the 120k AryANE microarray (1). Based on sensorial analysis over 6 years, six genotypes with contrasted fruit texture (mealy or not) were selected among a progeny. A global transcriptome analysis was performed over four time points during fruit development and cold storage. Interestingly, this analysis revealed one transcript coding a pectin methyl esterase (PME) which displayed an important modification in its expression level in accordance with the development of mealiness in fruits. Protein fusion experiment showed that this specific MdPME is secreted to the apoplast. Microscopic analysis revealed a progressive loss of cell to cell adhesion in mealy fruits in accordance with specific cell wall disassembly. Biochemical analysis pointed out specific alteration of pectin during fruit ripening. However, no global significant changes in pectin structural properties could be ascribed to mealiness. Consistent with the observations of Ng et al (2), we hypothesize that mealiness may be partially due to qualitative and spatial variations of pectin microarchitecture rather than quantitative pectin differences, and that these changes may occur early in fruit development. Overall, these data support the role of PME in cell wall remodeling during fruit maturation and suggest that PME act locally. This MdPME could be an early marker of texture disorder in apple.

(1) Celton et al., New Phytol, 2014. In Press.
(2) Ng et al., BMC Plant Biol. 2013.

A set of 45 sweet cherry (Prunus >avium) varieties which includes ancestors of modern varieties, modern breeding cultivars and local germplasm was genotyped and evaluated with the 6K SNP chip array of Illumina by an Infinium® assay. Genotyping filtering revealed that from 5696 SNPs available in the 6K array, 431 (7.5%) were not scored, 3368 (59%) were monomorphic and 1897 (33%) were polymorphic and informative (MAF>0,1). From these up to 1.5% failed Hardy-Weinberg equilibrium (p<0.001). The SNP dataset was used to study the genetic structure of the sample. Cluster analysis identified clear substructure within the dataset by differencing 3 major groups. Modern cultivars and breeding founders grouped together in a single cluster suggesting a common Western European Origin (WEO). Local varieties from Spain were separated in two clusters, one from Western Spain (WS) and another from Eastern Spain (ES). Preliminary studies pinpoint to the existence of a unique genomic pool in the Spanish sweet cherry germplasm, evidencing its potential for breeding programs.

Understanding the processes that occurred during domestication remains one of the central questions in genetics. In addition to fundamental importance, it can indeed provide knowledge on the genetic bases for further crop improvement. As an example, apple is one of the most important fruit crops in temperate regions, having both an economic and a cultural value. Dessert apples are used for consumption as fruits while cider apples are used to produce cider. Yet few traits are known to be different, despite overlapping between dessert and cider apples, namely fruit size and bitterness. A population genomics approach was used for detecting loci involved in the differentiation between cider and dessert apples, on two core collections of old cultivars, one for dessert apples and one for cider apples. A set of 96 gene fragments, localized in 6 areas of the apple genome bearing QTLs for traits of agronomic interest, was re-sequenced and the Illumina 8K SNP chip was used to genotype these two collections. The low estimates of per locus differentiation (ie Fst) between dessert and cider apples indicated that these two pools recently derive from common ancestors, except for 2 genomic regions exhibiting higher Fst values and potentially involved in selective processes. Departure from neutrality was also tested for each “population” using Tajima’s D and Fu and Li’s D and several areas of the genome were found to have significant decrease of both. Such knowledge could be helpful in new selection programs, bringing their efficiency to a higher level.

The diploid Fragaria vesca, is recognized as model for more complex octoploid species such as F. × ananassa (the commercial strawberry) and F. chiloensis (the native Chilean strawberry). The genome of F. vesca has been recently sequenced nevertheless the annotation in its 1.1 version presented some inconsistencies. Therefore, in order to improve the annotation of its transcriptome the following steps were followed: 1) Prediction of open reading frames from scaffolds; 2) Assembling of ORFs in a contig or coding sequences (CDS); 3) Annotation of genes; 4) Functional classification of the genes through KOG, followed by homology with reference genomes using Blast (Arabidopsis, Populus and Prunus). From the 256 scaffolds available a total of 217,766 ORFs, longer than 150 bp, were predicted using Glimmer. A total of 17,818 ORFs were annotated: 6451 sequences were classified within the main KOG categories (737 in Information, storage and processing category; 4858 in Cellular processing and signalling; and 856 in Metabolism), 5409 sequences were classified as poorly characterized and 5958 as unknown. A total of 53,586 ORFs with a biological description were annotated by homology. The entire data has been deposited initially in a local database. The search of genes related to important quality traits of strawberry fruit was performed particularly on fruit softening. The search provided an entire list of genes including those previously described as main players in the softening of strawberry fruit. Research supported by Anillo ACT-1110 project.

The defense response of apple rootstock to necrotrophic soilborne pathogens inciting replant disease is poorly defined. In this study, apple rootstock seedlings were inoculated with Pythium ultimum, a primary member in replant disease pathogen complex. Root tissues including mock inoculated were sampled at 0, 1, 4, 8, 24, 48, 72 and 96 hour post inoculation (hpi). RNA-Seq based transcriptome profiling was performed with at least 20 million reads per sample. The large gap mapper plugin and the transcript discovery plugin in combination with the existing RNA-Seq tool in the CLC Genomics Workbench were used to map the reads and identify the differentially expressed genes (DEGs); gene ontology analysis and KEGG pathway analysis were applied to annotate DEGs. The peak response is at 48 hpi based on the number of the identified DEGs (1061). Members from gene families functioning in hormone signaling such as ET, JA, GA, CK and auxin, encoding NAC, WRKY, MYB and ERF transcription factors; encoding enzymes in several biosynthesis pathways of potential antimicrobial secondary metabolites and cell wall modification are among the identified DEGs. The results from this study represent a crucial step in dissecting the resistant mechanisms in perennial root system to soilborne pathogens. It seems that the molecular framework of defense responses in perennial root system is largely conserved with that characterized using foliar pathosystem. Further investigation on the defense responses specific to various pathogens and rootstock genotypes (resistant and susceptible) should reveal valuable information for future genomics-assisted breeding of resistant apple rootstock.

Genome and transcript sequencing are common tools in molecular biology and genetics research, but most researchers in these fields do not have the computer science knowledge or server access to use the common bioinformatics tools to analyze these data. Bioinformatics tools that are user-friendly and allow access to server resources via the web are becoming more common. The goal of this project was to design and implement a web-based DNA annotation pipeline that allows researchers to analyze their DNA sequences with command line based annotation tools though a graphical user interface. The Genome Sequence Annotation Server (GenSAS) allows users to upload DNA sequences, create a task of user selected annotation tools, and submit the task to the server for execution. Users return to the GenSAS web page to view and work with the results when the task is complete. Results from tasks are visualized in the GenSAS Browser and users can create custom curation tracks by selecting features from the results of the annotation tools. Users export their results in the GFF3 file format which is compatible with many downstream bioinformatics tools and genome browsers. We have tested GenSAS with DNA from a variety of organisms and of different lengths and GenSAS has been further tested in undergraduate and graduate classes. Based on feedback from users, GenSAS is an easy to use customizable online DNA annotation pipeline tool that allows users with little computer science knowledge to create a custom DNA annotation for their sequence(s) of interest.

Phylogenetic reconstruction using organelles or small numbers of nuclear loci often fails to accurately reconstruct organismal relationships, especially in groups with high rates of speciation and/or reticulate evolution, such as Rosaceae. The apple, peach and strawberry genomes have been sequenced, and these genera represent 3 of the 4 lineages of the family. To simultaneously obtain sequence data from hundreds of nuclear loci, probes were designed for solution hybridization-based target enrichment followed by sequencing on the Illumina HiSeq platform (Hyb-Seq). A total of 1419 conserved orthologous exons (<15% sequence divergence in pairwise comparisons across the three genomes and >10% divergence among loci in a genome) were targeted from 257 single-copy nuclear loci. The analysis included 24 species representing the 12 recognized Rosaceae tribes and 4 unplaced genera. A mean of 1342 exons and 392,000 base pairs were obtained per species. In addition, nearly complete plastomes were assembled via genome skimming. Maximum likelihood phylogenetic analysis was conducted using nucleotide and amino acid sequences with concatenation and gene tree approaches. The robust phylogenetic resolution of all inter-tribal relationships provides a strong foundation for studies of gene family evolution, biogeography and character evolution in Rosaceae. The probes are available for phylogenetic analysis within genera of Rosaceae, and are currently being used in Fragaria, Rubus, Crataegus and others. These studies highlight their utility for the resolution of species delimitation, interspecific hybridization and polyploid origins.

Next generation sequencing has significantly reduced sequencing costs and has permitted de-novo assembly of complex genomes, and re-sequencing of multiple genotypes within a species. The availability of plant reference genomes has accelerated genetics and genomics research, providing tools to identify functional elements, leading to the efficient development of improved varieties. Understanding the links between phenotypic variation and DNA variation the major challenge for plant geneticists. High-density SNP arrays for genome-wide assessment of allelic variation, have made high resolution genetic characterization of crop germplasm feasible. Such arrays are now commercially available for plant species with complex polyploid genomes.
Here we present an Illumina Infinium 20K array in apple that will be implemented in Genome Wide Association Studies as well as Pedigree Based Analyses. In both cases it will be employed to dissect the genetic mechanism controlling important fruit quality traits, towards the identification of valuable markers suitable for the assisted selection activity in apple breeding programs.. SNPs contained on the array were predicted from re-sequencing data derived from 14 apple genotypes and two doubled-haploids of Golden Delicious. An customised pipeline for SNP selection was devised avoiding the pitfalls of paralogous sequence variants. The performance of the array was assessed using data from mapping 23 populations from controlled crosses and germplasm collections. Marker positions based on the 23 genetic maps provided new insights on the apple genome, which will be presented and which were used to improve the scaffolding and anchoring of the apple reference genome, now available in the 3.0 new assembly.

Discovering genetic variations that cause the observed variations for important plant traits, is one of the key challenges for plant researchers. For apple, the International RosBREED SNP Consortium developed an 8K SNP-chip on the Illumina Infinium platform that can be used for such association and genotyping studies. This SNP-chip was used to genotype our Malus x domestica germplasm collection consisting of both commercial and old cultivars as well as some genotypes from a Telamon x Braeburn progeny.
We used the SNP data obtained from this SNP-array to estimate the linkage disequilibrium (LD) between adjacent SNPs as well as the LD decay across larger regions because this determines the number of markers necessary for association studies. In addition, association analyses can be biased by the population structure present in the collection used for these studies. Therefore we determined the population structure of our germplasm collection using the software STRUCTURE.
Evanno’s criterium indicated that the most likely numbers of subpopulations present in our collection were two and three subpopulations with two being the most likely. The first subpopulation was a group with the Telamon x Braeburn progeny and other Braeburn-related cultivars. The other subpopulation contained all other commercial and old cultivars. Thislatter group split into a group of commercial cultivarsand a group of old cultivars when increasing the number of subpopulations to three while the first group remained unchanged.
The estimated LD between adjacent SNPs using the r2 parameter was 0.23 and decayed quickly below 0.18 after about35kb.