Detection of quantitative trait nucleotides (QTNs) and QTN-by-environment and QTN-by-QTN interactions for cotton early-maturity traits using the 3VmrMLM method

Abstract

Early maturity is among the most important and complex agronomic traits of upland cotton (Gossypium hirsutum). Using the recently established 3VmrMLM method, a genome-wide association study was conducted, focusing on three primary early-maturity traits: growth period (GP), the first fruiting branch node (FFBN), and height of the first fruiting branch node (HFFBN). As a result, 45, 37, and 1 quantitative trait nucleotides (QTNs) for GP, 51, 28, and 0 QTNs for FFBN, and 84, 39, and 3 QTNs for HFFBN were detected by the “Single_env”, “Multi_env”, and “Epistasis” modules of IIIVmrMLM software, respectively. Of these QTNs, two for GP, two for FFBN, and three for HFFBN were simultaneously detected in at least three datasets or by at least three modules. The three genes, GH_D08G2175, GH_D08G2201, and GH_D09G2513, associated with the stable QTNs were further verified by qRT-PCR and haplotype analysis. Fourteen QTN-by-environment interactions (QEIs), comprising two for GP, seven for FFBN, and five for HFFBN, were detected, of which nine QEIs exhibited significant phenotypic differences between different genotypes under various environments. Three of the nine QEIs were respectively located within the genes GH_D12G1405, GH_D05G2088, and GH_D11G1027, which are likely to interact with environments and thus affect early maturity. Altogether 21 QTN-by-QTN interactions (QQIs), comprising 5 for GP and 16 for HFFBN, representing interactions between various effects were detected. Ten protein–protein interactions involving 6 QQIs and 10 gene pairs were predicted. The findings offer fresh perspectives on the genes underlying early maturity, advancing the prospects for molecular design breeding in cotton.