Functional characterization and allelic mining of OsGLR genes for potential uses in rice improvement

Abstract

Glutamate-like receptor (GLR) genes are a group of regulatory genes involved in many physiological processes of plants. With 26 members in the rice genome, the functionalities of most rice GLR genes remain unknown. To facilitate their potential uses in rice improvement, an integrated strategy involving CRISPR-Cas9 mediated knockouts, deep mining and analyses of transcriptomic responses to different abiotic stresses/hormone treatments and gene CDS haplotype (gcHap) diversity in 3,010 rice genomes was taken to understand the functionalities of the 26 rice GLR genes, which led us to two conclusions. First, the expansion of rice GLR genes into a large gene family during evolution had gone through repeated gene duplication events occurred primarily in two large GLR gene clusters on rice chromosomes 9 and 6, which was accompanied with considerable functional differentiation. Secondly, except for two extremely conserved ones (OsGLR6.2 and OsGLR6.3), rich gcHap diversity exists at the remaining GLR genes which played important roles in rice population differentiation and rice improvement, evidenced by their very strong sub-specific and population differentiation, by their differentiated responses to day-length and different abiotic stresses, by the large phenotypic effects of five GLR gene knockout mutants on rice yield traits, by the significant association of major gcHaps at most GLR loci with yield traits, and by the strong genetic bottleneck effects and artificial selection on the gcHap diversity in populations Xian (indica) and Geng (japonica) during modern breeding. Our results suggest the potential values of the natural variation at most rice GLR loci for improving the productivity and tolerances to abiotic stresses. Additional efforts are needed to determine the phenotypic effects of major gcHaps at these GLR loci in order to identify 'favorable' alleles at specific GLR loci specific target traits in specific environments to facilitate their application to rice improvement in future.

Keywords: functional alleles; gene-CDS-haplotype (gcHap) diversity; glutamate-like receptor (GLR) genes; knockout mutants; rice improvement.

Figure 1

The phylogenetic relations between plant GLRs and mammals iGluRs (ionotropic glutamate receptors). The phylogenetic relations of glutamate receptors from Oryza sativa Geng (Os) based on the Os-Nipponbare-Reference-IRGSP-1.0 reference genome and other flowering plants that are described to show a conserved phenotype, including Arabidopsis thaliana (At), Zea mays (Zm), Solanum lycopersicum (Sl), Gossypium hirsutum (Gh), and Raphanus sativus (Rs), as well as basal land plants such as the moss Physcomitrium patens (Pp), the liverwort Marchantia polymorpha (Mp), and the lycophyte Selaginella moellendorffii (Sm), compared to the invertebrate Caenorhabditis elegans (Ce) and AMPARs, NMDARs, KARs, and δ-receptors from mammals (without prefix). Also included is the bacterial GluR0 from Synechocystis PCC 6803 and AvGluR1 from the freshwater rotifer Adineta vaga (Av). OsGLRs are shown in colored text for clarity. Sequences were aligned using MUSCLE software, and the phylogenetic tree was constructed using the neighbor-joining method. Sequences are available in the Supplementary Data 1 . AMPAR, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor; KAR, kainate receptor; NMDAR, N-methyl-d-aspartate receptor; δ, δ-receptor.

Figure 2

The expression level of the OsGLR genes in different rice tissues under the normal conditions. The plant tissues mainly include aleurone, anther, callus, leaf, panicle, pistil, root, seed and shoot.

Figure 3

Integrative analysis of the rice OsGLR gene family expression profiles under ABA (A), Cd (B, C), JA (D), cold (E), and dry (F) treatment conditions.

Figure 4

Phenotypic comparisons for five agronomic traits of knockout mutants of five rice OsGLR genes evaluated in two environments. (A-E) The agronomic traits represented by (A-E) are 1000-grain weight (g), seed setting rate (%), panicle number, panicle length (cm) and plant height (cm), respectively. *, **, *** and **** meant statistically significant at a level of 0.05, 0.01, 0.001 and 0.0001, respectively.

Figure 5

Population diversity of OsGLR genes in 3010 Asia cultivated rice accessions. (A) OsGLR genes distribution of five diversity gene types (house-keeping (HK) genes without SNP variants, low-diversity genes (0 < E_H < 0.05, E_H = 0.020 ± 0.015, and gcHapN = 12 ± 10), medium- to low-diversity genes (0.05 ≤ E_H < 0.3, E_H = 0.171 ± 0.065, and gcHapN = 82 ± 67), medium- to high-diversity genes (0.3 ≤ E_H < 0.7, E_H = 0.444 ± 0.110, and gcHapN = 498 ± 305) and a high-diversity gene (0.7 ≤ E_H ≤ 1, E_H = 0.804 ± 0.075, and gcHapN = 1767 ± 458), Shannon’s equitability (E_H ) takes into account both the number of species present (species richness) and the relative abundance of each species within the community. It ranges from 0 to 1, with values closer to 1 indicating a more even distribution of species abundance, and values closer to 0 indicating a less even distribution. (B) The relationship between Shannon’s equitability (E_H ) and gcHapN of OsGLR genes. (C) E_H values of OsGLR genes in different population organization. (D) gcHap number (gcHapN) of OsGLR genes in different population organization. (E) The major gcHap (≥ 30 rice accessions) number of OsGLR genes.

Figure 6

The Nei’s genetic identity (I_Nei ) of rice OsGLR genes between all pairwise populations calculated from the gcHap data.

Figure 7

The lost, new and unchanged gcHap variation of rice 26 OsGLR genes during modern breeding. (A) Heatmap of the numbers of lost, newly emerged, and unchanged gcHaps. (B) The comparative analysis of the numbers of lost, newly emerged, and unchanged gcHaps. (C) The frequency comparison of lost, newly emerged, and unchanged gcHaps. The Tukey’s multiple comparison method was used. ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05, and n.s., not significant.

Figure 8

Haplotype networks of five GLR genes with CRSPR-9 mediated knocking-out and their associations with four agronomic traits in 3KRG. (A) Based on gcHap diversity and contribution to subspecific/population differentiation, the rice OsGLR2.2 is classified as the other gene. (B) The OsGLR4.1 is classified as a X-G divergent gene. (C) The OsGLR7.1 is classified as a X-G divergent gene. (D) The rice OsGLR6.8 is classified as the other gene. (E) The OsGLR9.8 is classified as a X-G divergent gene. Within each haplotype network, two adjacent gcHaps are separated by mutational changes with hatching line indicating differences between the two most related haplotypes. The right side of each gene haplotype network corresponds to the phenotypic variation among the haplotypes. Boxplots are shown for the following four traits: plant height (PH), panicle length (PL), panicle number (PN) and 1000-grain weight (TGW). The P values under trait names indicate differences between the haplotypes assessed by two-way ANOVA, with different letters on the boxplots indicating statistically significant differences at P < 0.05 based on Duncan’s multiple range test. The bar charts on the right show the differences in frequency of the predominant gcHaps between landraces (LANs) and modern varieties (MVs) in Xian and Geng. Chi-square tests were used to determine significant differences in the proportions of the same gcHap between the different populations with ****P < 0.0001, ***P < 0.0001, **P < 0.01, *P < 0.05, and n.s., not significant.

Figure 9

Frequencies of the favorable gcHaps of all rice GLR genes affecting important agronomic traits (PN and TGW) in Xian/indica (XI) and Geng/japonica (GJ) modern varieties and different rice subpopulations. The favorable gcHap of a gene is defined as the gcHap associated with the highest trait value. “#accession” indicates the number of accessions that possess the favorable gcHap. Five subpopulations of XI (XI-1A, XI-1B, XI-2, XI-3, and XI-adm) and four subpopulations of GJ (temperate GJ: GJ-tmp, subtropical GJ: GJ-sbtrp, tropical GJ: GJ-trp, and GJ-adm) were classified by Wang et al. (Wang et al., 2018).