Mapping the QTL for yield traits in bitter gourd (Momordica charantia L.)

Abstract

Bitter gourd (Momordica charantia), being a rich source of phytonutrients such as carbohydrates, minerals, vitamins, and other medicinal compounds, has a great importance in healthy dietary habits. Breeders always seek to breed bitter gourd varieties for the traits such as early maturity and high yield. However, limited investigations have been made to identify the genetic loci governing yield related traits. Marker assisted selection (MAS) assures the presence of favourable alleles and fast recovery of recurrent parent genome in the cultivar under improvement. The success of MAS mainly depends on the availability of a marker-dense genetic linkage map locating quantitative trait loci (QTL) for the target traits. The present study “Mapping the QTL for yield traits in bitter gourd (Momordica charantia L.)” was carried out during October, 2018 to December, 2021 with the objective to map the quantitative trait loci and to develop chromosome-wise maps for the yield traits in bitter gourd. To develop the mapping population, high yielding bitter gourd cultivar Priyanka (Momordica charantia var. charantia) and a wild bitter gourd accession IC634896 (M. charantia var. muricata), were used as parents. A set of 450 microsatellites were screened for polymorphism using genomic DNA of parents and 47 were found polymorphic. Bitter gourd genome (GenBank acc. no. GCA_013281855.1) was scanned and new hypervariable microsatellites were identified using Genome wide Microsatellite Analysing Tool (GMATo) and named as KAUBG_n where n is a serial number. From the 75 microsatellites identified, 69 were validated through successful PCR amplification and 38 among them were polymorphic between the parents. This led to the development of a set of 85 markers polymorphic between the parents. Crosses were made between the parental lines and hybrids from the cross Priyanka × IC634896 yielded more number of fruits and total fruit produce compared to the reciprocal hybrid. An F2:3 population was developed through single seed descent method from the cross Priyanka × IC634896. A panel of 200 F2:3 plants were evaluated for twenty seven traits, including fruit-, flower-, seed-, vine-, and leaf-related traits, contributing directly or indirectly to the total yield. Wide variation was observed among the F2:3 plants for the traits studied. A group of ninety plants was selected from 200 F2:3 plants such that they represent the variation of the base population. Genomic DNA of these plants were genotyped using 85 polymorphic markers. Genotypic data from the screening of 85 markers in the mapping population were used to generate a linkage map spanning 1287.99 cM distance across eleven linkage groups (LGs) corresponding to eleven chromosomes, using IciMapping software. LG 7 (28 markers) consisted of maximum number of markers followed by LG 2 and LG 9, each having 11 markers. LG 1 had 10 markers whereas LG 3, 4 and 8 had seven markers each. LG 5, 6, 10 and 11 had only one marker each. LG 7 covered maximum map distance of 384.19 cM where LG 8 covered least map distance of 68.58 cM. The genetic map and phenotypic data were used to generate the QTL maps, using Inclusive Composite Interval Mapping (ICIM) method to locate twenty seven traits on Momordica genome. Sixty QTL, including 37 major QTL with LOD values ranging from 3.1 to 15.2, explaining 1.8 to 35.9 per cent of the phenotypic variation were identified for 24 traits, on seven chromosomes. Twenty three QTL were identified for fruit-traits with LOD values ranging from 3.1 to 7.6, explaining 5.5 to 35.9 per cent of phenotypic variation. Thirteen QTL were identified for flower-related traits with LOD value ranging from 3.1 to 15.2, explaining 7.0 to 26.0 per cent of phenotypic variation. Seven QTL each were identified for seed and leaf-related traits with LOD values ranging from 3.2 to 10.8 and 3.5 to 6.5, explaining 5.6 to 26.3 and 3.2 to 15.8 per cent of phenotypic variation, respectively. Ten QTL were identified for vine-related traits with 3.2 to 8.7 LOD values and explaining 1.8 to 17.6 per cent of phenotypic variation. Single marker analysis was performed to identify markers co-segregating with the yield contributing traits. There were 129 hits for the marker-trait association with LOD values more than 3.0, explaining 11.62 to 29.34 per cent of the phenotypic variation. Using the least and best performing F2:3 plants, markers S13, KAUBG_5 and KAUBG_11 were validated for co-segregation with fruit breadth, first pistillate flower node, and number of pistillate flowers and fruits per plant, respectively. This study gives insights into the relative locations of microsatellites and major effect QTL for yield traits in Momordica genome. QTL with shorter marker interval (qFrtL-8-1, qDPF-3-1, qDSF-3-1, qDSF-7-1, qFrtShp-8-1) can be directly used in MAS for improving yield characters. Linkage observed between microsatellites identified in this study with yield traits signifies their importance in further fine mapping as well as marker assisted selection. The linkage map constructed in this study, being the first with microsatellites from Momordica genome, paves the path for comparative and consensus map generation with other marker types. Further, fine mapping using markers within the identified QTL hotspots can lead to possible identification and cloning of genes underlying the yield traits.