Identification of a Candidate Gene for Panicle Length in Rice (Oryza sativa L.) Via Association and Linkage Analysis

Abstract

Panicle length (PL) is an important trait for improving panicle architecture and grain yield in rice (Oryza sativa L.). Three populations were used to identify QTLs and candidate genes associated with PL. Four QTLs for PL were detected on chromosomes 4, 6, and 9 through linkage mapping in the recombinant inbred line population derived from a cross between the cultivars Xiushui79 (short panicle) and C-bao (long panicle). Ten SSR markers associated with PL were detected on chromosomes 2, 3, 5, 6, 8, 9, and 10 in the natural population consisting of 540 accessions collected from East and Southeast Asia. A major locus on chromosome 9 with the largest effect was identified via both linkage and association mapping. LONG PANICLE 1 (LP1) locus was delimited to a 90-kb region of the long arm of chromosome 9 through fine mapping using a single segment segregating F2 population. Two single nucleotide polymorphisms (SNPs) leading to amino acid changes were detected in the third and fifth exons of LP1. LP1 encodes a Remorin_C-containing protein of unknown function with homologs in a variety of species. Sequencing analysis of LP1 in two parents and 103 rice accessions indicated that SNP1 is associated with panicle length. The LP1 allele of Xiushui79 leads to reduced panicle length, whereas the allele of C-bao relieves the suppression of panicle length. LP1 and the elite alleles can be used to improve panicle length in rice.

Keywords: association analysis; gene identification; map-based cloning; panicle length; quantitative trait locus; rice.

Figure 1

Population structure analysis of 540 rice accessions. (A) Panicle length of 10 rice accessions selected from 540 accessions to represent the diversity of panicle length in the population. Bar = 5 cm. (B) Frequency distribution of the panicle length of the 540 accessions in 2011 and 2012. (C) Changes in mean LnP(K) and ΔK. (D) Neighbor-joining tree for the 540 accessions generated using Nei's genetic distance. (E) Posterior probability of the 540 accessions belonged to seven subpopulations. (F) Principal components analysis (PCoA) for 540 accessions and reference cultivars genotyped with 262 molecular markers.

Figure 2

Genotypic and phenotypic performance of the parents. (A) Workflow of the construction of one NIL plant (BC₄F₂), NIL-LP1. (B) Graphical genotype of NIL-LP1. The black bar indicates the fragment from C-bao, and the remainder was derived from Xiushui79. (C) Panicle morphology of Xiushui79, NIL-LP1 and C-bao. Bar = 5 cm. (D) Xiushui79, NIL-LP1, and C-bao plants. Bar = 10 cm.

Figure 3

Map-based cloning of LP1. (A) Primary mapping of LP1. LP1 was mapped between the SSR markers RM3600 and RM410. (B) High-resolution mapping of LP1. LP1 was delimited to a 90-kb region between the markers L04 and RM7289 in the BAC clones AP005655 and AP006057, using a total of 8650 plants from the NIL-LP1/Xiushui79 F₂ population. (C) Predicted ORFs based on the Nipponbare genome sequence in the Rice Genome Annotation Project database (http://rice.plantbiology.msu.edu/index.shtml). The horizontal arrows indicate the six predicted ORFs. (D) Gene structure of LOC_Os09g28300. The empty boxes refer to the 5′ and 3′ UTRs; the black boxes, to exons; and the lines between the boxes, to introns. The fourteen SNPs in LP1 are indicated by solid lines. (E) The Remorin_C domain predicted in the protein encoded by LOC_Os09g28300. The solid lines indicate the positions of two amino acid transitions.

Figure 4

Analysis of the expression of the three expressed candidate genes via real-time quantitative RT-PCR. (A) Relative expression of LOC_Os09g28300 in the roots, stems, flag leaves and young panicles. (B) Relative expression of LOC_Os09g28340 in the roots, stems, flag leaves and young panicles. (C) Relative expression of LOC_Os09g28370 in the roots, stems, flag leaves and young panicles. (D) Real-time quantitative RT-PCR results for the three candidate genes in the young panicles. The 18S rRNA gene was used as a control. ^** indicates significance at the α = 0.01 probability level.

Figure 5

Phylogenetic analysis of the LP1 protein. (A) Phylogenetic tree of the LP1-like proteins in higher plants. The amino acid sequences of 15 LP1 homologs were obtained from NCBI (http://www.ncbi.nlm.nih.gov/) and used to construct a bootstrap N-J phylogenetic tree. A total of 1000 replicates were performed to determine the statistical support for each node. LP1 is denoted in bold. (B) Diagram indicating the two amino acid substitutions in LP1 at positions conserved across LP1-like proteins.

Figure 6

Distribution of marker-PL and QTLs or genes for panicle length and panicle length related on rice chromosome.