QTL mapping for fruit quality in Citrus using DArTseq markers

Abstract

BackgroundCitrus breeding programs have many limitations associated with the species biology and physiology, requiring the incorporation of new biotechnological tools to provide new breeding possibilities. Diversity Arrays Technology (DArT) markers, combined with next-generation sequencing, have wide applicability in the construction of high-resolution genetic maps and in quantitative trait locus (QTL) mapping. This study aimed to construct an integrated genetic map using full-sib progeny derived from Murcott tangor and Pera sweet orange and DArTseq™ molecular markers and to perform QTL mapping of twelve fruit quality traits. A controlled Murcott x Pera crossing was conducted at the Citrus Germplasm Repository at the Sylvio Moreira Citrus Centre of the Agronomic Institute (IAC) located in Cordeirópolis, SP, in 1997. In 2012, 278 F1 individuals out of a family of 312 confirmed hybrid individuals were analyzed for fruit traits and genotyped using the DArTseq markers. Using OneMap software to obtain the integrated genetic map, we considered only the DArT loci that showed no segregation deviation. The likelihood ratio and the genomic information from the available Citrus sinensis L. Osbeck genome were used to determine the linkage groups (LGs).ResultsThe resulting integrated map contained 661 markers in 13 LGs, with a genomic coverage of 2,774 cM and a mean density of 0.23 markers/cM. The groups were assigned to the nine Citrus haploid chromosomes; however, some of the chromosomes were represented by two LGs due the lack of information for a single integration, as in cases where markers segregated in a 3:1 fashion. A total of 19 QTLs were identified through composite interval mapping (CIM) of the 12 analyzed fruit characteristics: fruit diameter (cm), height (cm), height/diameter ratio, weight (g), rind thickness (cm), segments per fruit, total soluble solids (TSS, %), total titratable acidity (TTA, %), juice content (%), number of seeds, TSS/TTA ratio and number of fruits per box. The genomic sequence (pseudochromosomes) of C. sinensis was compared to the genetic map, and synteny was clearly identified. Further analysis of the map regions with the highest LOD scores enabled the identification of putative genes that could be associated with the fruit quality characteristics.ConclusionAn integrated linkage map of Murcott tangor and Pera sweet orange using DArTseq™ molecular markers was established and it was useful to perform QTL mapping of twelve fruit quality traits. The next generation sequences data allowed the comparison between the linkage map and the genomic sequence (pseudochromosomes) of C. sinensis and the identification of genes that may be responsible for phenotypic traits in Citrus. The obtained linkage map was used to assign sequences that had not been previously assigned to a position in the reference genome.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-017-3629-2) contains supplementary material, which is available to authorized users.

Construction of high-density genetic maps and detection of QTLs associated with Huanglongbing infection in citrus

“…Although we initially developed numerous SNP markers for our linkage mapping, only a small portion of markers were successfully mapped as unique loci. This has been observed in all previous SNP-based genetic maps of citrus, where many markers were located in ‘zero recombination clusters’, indicating nearly no meiotic recombination occurs between markers within those regions (Curtolo et al , 2017; Guo et al , 2015; Imai et al , 2017; Lyon, 2008; Shimada et al , 2014; Yu et al , 2016). High redundancy of identified markers is commonly attributed to both small population size and closely adjacent physical marker location, and could be significantly improved by increasing the size of mapping populations and the evenness of marker distribution.…”

“…Guo et al successfully constructed an integrated genetic map of pummelo with 1543 SNP markers using an intraspecific full-sib F 1 population with only 124 individuals (Guo et al , 2015). By contrast, Curtolo et al attempted to construct a genetic map using an interspecific full-sib F 1 population with 278 individuals from crossing of tangor and sweet orange, but only 661 non-redundant SNP-based DArTseq markers were finally mapped on the integrated map (Curtolo et al , 2017). As shown in Fig.…”

“…Biological characteristics of citrus, such as long juvenility, seedlessness, ployembryony, apomixis, heterozygosis, gametophytic incompatibility, zygotic selection and gametic selection (Shimada et al , 2014), not only seriously hamper the development of numerous uniquely segregating markers and production of large full-sib populations, but also remarkably influence allelic segregation and recombination ratio (Bernet et al , 2010; Ollitrault et al , 2012a). In addition to parental genotype, differential fitness of gametal genotypes, crossing direction, and regulatory gene interactions likely also contribute to the high level of segregation distortion in citrus (Bernet et al , 2010; Curtolo et al , 2017). In this study, the mapping population are neither full-sib progenies nor interspecific or intraspecific hybrids, but intergeneric hybrids from mixed crosses of several parents.…”

“…This is the first report on identification of QTLs related to C Las infection and HLB tolerance. Most of the recently reported QTLs in citrus are related to morphological and physiological traits (Asins et al , 2015; Curtolo et al , 2017; Imai et al , 2017; Kepiro and Roose, 2010; Raga et al , 2014; Raga et al , 2016; Sahin-Cevik and Moore, 2012; Sugiyama et al , 2011; Yu et al , 2017; Yu et al , 2016). Only a few reports focused on disease-related QTLs in citrus, such as Citrus tristeza virus (CTV) (Asins et al , 2012; Ohta et al , 2015), Alternaria brown spot (ABS) (Cuenca et al , 2016; Cuenca et al , 2013), and Citrus leprosis virus (CiLV) (Bastianel et al , 2009).…”

“…While initially confined to herbaceous plants, high-density genetic mapping is increasingly extended to perennial woody plants. Using high-throughput genotyping, the saturation of genetic maps has been greatly improved for some citrus species, such as sweet orange, mandarin and pummelo (Curtolo et al , 2017; Guo et al , 2015; Imai et al , 2017; Lyon, 2008; Ollitrault et al , 2012a; Shimada et al , 2014; Yu et al , 2016). However, in comparison with well-studied model and agronomic plants, citrus lags in development of high-density, high-resolution genetic maps with fine accuracy and precision.…”