Identification of quantitative trait loci (QTLs) associated with agronomic traits in tea (Camellia sinensis) (L.) O. Kuntze

Abstract

The identification of candidate genes associated with black tea quality and drought tolerance traits in tea (Camellia sinensis) plants remains complicated and time-consuming. This study aimed to identify molecular, physiological and biochemical characteristics associated with black tea quality and drought tolerance traits in two tea segregating populations for future marker-assisted selection breeding. The two tea segregating populations TRFK St. 504 and TRFK St. 524 were developed from a reciprocal cross of two heterozygous parental clones, TRFK 303/577 and GW Ejulu. Using Ultra-Performance Liquid Chromatography (UPLC), catechin fractions and theaflavin fractions was analysed in green and black tea, respectively, while caffeine content were analysed in both green and black tea. Black tea samples were further subjected to organoleptic evaluation by professional tea tasters to score liquor characters (colour, brightness, strength and briskness). The percent relative water content (%RWC) using Short-time Withering Assessment of Probability for Drought Tolerance (SWAPDT) method was used to distinguish between drought susceptible and drought-tolerant tea cultivars. A total of 16 phenotypic data from two segregating tea populations were used to identify the quantitative trait loci (QTLs) influencing tea biochemical, and drought stress traits based on a consensus genetic map constructed using the DArTseq platform. The map consisted of 15 linkage group and corresponded to chromosome haploid number of tea plant (2n = 2x = 30) and spanned 1 260.1 cM with a mean interval of 1.1 cM between markers. A total of 1 421 DArTseq markers derived from the linkage map identified 53 DArTseq markers to be linked to black tea quality and %RWC. The putative QTLs linked to black tea quality, and drought tolerance traits were submitted to BLAST and assigned functions using Gene Ontology (GO) terms and biosynthetic pathways in the tea genome, gene ontology database and Kyoto Encyclopedia of Genes and Genomes (KEGG). The approach of combining linkage mapping with association mapping allowed identification, and precise authentication of putative QTLs with an additional six more putative QTLs identified. The functional annotations of all the putative QTLs detected were involved in the metabolism of secondary metabolites associated with tea phenolic biomolecules and abiotic stress. The predictive ability of all machine learning models varied across the phenotypic traits. The putative QTLs + annotated proteins + KEGG pathways based prediction approach showed more robustness and usefulness in the prediction of phenotypes than the individual prediction based models. The Extreme Learning Machine (ELM) model had a better prediction ability for catechin, astringency, brightness, briskness and colour based on putative QTLs + annotated proteins + KEGG pathways approach. The percent variables of importance of putatively annotated proteins and KEGG pathways were associated with the phenotypic traits.