Effect of qGN4.1 QTL for Grain Number per Panicle in Genetic Backgrounds of Twelve Different Mega Varieties of Rice

Abstract Background Rice is a major source of food, particularly for the growing Asian population; hence, the utilization of genes for enhancing its yield potential is important for ensuring food security. Earlier, we have mapped a major quantitative trait loci (QTL) for the grain number per panicle, qGN4.1, using biparental recombinant inbred line (RIL) populations involving a new plant type Indica rice genotype Pusa 1266. Later, three independent studies have confirmed the presence of a major QTL for spikelet number by three different names (SPIKE, GPS and LSCHL4) in the same chromosomal region, and have implicated the overexpression of Nal1 gene as the causal factor for high spikelet number. However, the effect of qGN4.1 in different rice genetic backgrounds and expression levels of the underlying candidate genes is not known. Results Here, we report the effect of qGN4.1 QTL in the genetic backgrounds of 12 different high-yielding mega varieties of rice, introgressed by marker assisted-backcross breeding (MABB) using two QTL positive markers for foreground selection and two QTL negative flanking markers for recombinant selection together with phenotypic selection for the recovery of recipient parent genetic background. Analysis of the performance of BC2F3 plants showed a significant increase in the average number of well-filled grains per panicle in all the backgrounds, ranging from 21.6 in CSR 30-GN4.1 to 147.6 in Samba Mahsuri-GN4.1. Furthermore, qGN4.1 caused a significant increase in flag leaf width and panicle branching in most backgrounds. We identified BC3F3 qGN4.1 near-isogenic lines (NILs) with 92.0–98.0% similarity to the respective recipient parent by background analysis using a 50 K rice SNP genotyping chip. Three of the NILs, namely Pusa Basmati 1121-GN4.1, Samba Mahsuri-GN4.1 and Swarna-GN4.1, showed a significant yield superiority to their recipient parents. Analysis of differential gene expression revealed that high grain number in these QTL-NILs was unlikely due to the overexpression of Nal1 gene (LOC_Os04g52479). Instead, another tightly linked gene (LOC_Os04g52590) coding for a protein kinase domain-containing protein was consistently overexpressed in the high grain number NILs. Conclusion We have successfully introgressed the qGN4.1 QTL for high grain number per panicle into 12 different mega varieties of rice using marker-assisted backcross breeding. The advanced near-isogenic lines are promising for the development of even higher yielding versions of these high-yielding mega varieties of rice