Combination of Genomics, Transcriptomics Identifies Candidate Loci Related to Cold Tolerance in Dongxiang Wild Rice

Abstract

Rice, a cold-sensitive crop, is a staple food for more than 50% of the world's population. Low temperature severely compromises the growth of rice and challenges China's food safety. Dongxiang wild rice (DXWR) is the most northerly common wild rice in China and has strong cold tolerance, but the genetic basis of its cold tolerance is still unclear. Here, we report quantitative trait loci (QTLs) analysis for seedling cold tolerance (SCT) using a high-density single nucleotide polymorphism linkage map in the backcross recombinant inbred lines that were derived from a cross of DXWR, and an indica cultivar, GZX49. A total of 10 putative QTLs were identified for SCT under 4 °C cold treatment, each explaining 2.0-6.8% of the phenotypic variation in this population. Furthermore, transcriptome sequencing of DXWR seedlings before and after cold treatment was performed, and 898 and 3413 differentially expressed genes (DEGs) relative to 0 h in cold-tolerant for 4 h and 12 h were identified, respectively. Gene ontology and Kyoto encyclopedia of genes and genomes (KEGG) analysis were performed on these DEGs. Using transcriptome data and genetic linkage analysis, combined with qRT-PCR, sequence comparison, and bioinformatics, LOC_Os08g04840 was putatively identified as a candidate gene for the major effect locus qSCT8. These findings provided insights into the genetic basis of SCT for the improvement of cold stress potential in rice breeding programs.

Keywords: Dongxiang wild rice; differentially expressed genes; quantitative trait locus; seedling cold tolerance; transcriptomics.

Figure 1

Development and construction of high-density bin map of BRILs. (A) Flowchart of developing backcross recombinant inbred line (BRIL) population. ⊗ represents self-cross. (B) Polymorphic SNPs between DXWR and GZX49 distributed on chromosomes. (C) Genetic linkage map constructed by population BRILs.

Figure 2

Phenotypic identification of seedling for cold tolerance. (A,B) represent the phenotypes of GZX49 and DXWR before and after cold treatment, respectively. (C) Phenotypes of some lines of the BRIL population after cold treatment. The scale bar of (A–C) means 5 cm. (D) Frequency distribution of seedling survival rate after cold treatment. Arrows indicate the means of parental lines GZX49 and DXWR.

Figure 3

QTL analysis of the BRIL population. (A) Manhattan plots of the loci for seedling cold tolerance (SCT). The x-axis represents single nucleotide polymorphism (SNP) along each numbered chromosome; the y-axis represents the negative logarithm of the p-value (-log10 p) for the SNP association. Horizontal dashed lines in the plots indicate the declaration thresholds. (B,D) Expression levels of the OsMYB3R-2 and OsTPP1 in GZX49 and DXWR after cold stress measured by qRT-PCR, respectively. The results were statistically analyzed using Student’s t-test (** p < 0.01, *** p < 0.005). Transcription levels relative to 0 h, which was set to 1, are presented as the mean and SE of triplicates. LOC_Os03g13170 (Ubiquitin) is the control gene. (C,E) Sequence comparison of OsMYB3R-2 and OsTPP1 among GZX49 and DXWR. The vertical bars and triangle represent SNPs and nucleotide deletion, respectively.

Figure 4

Transcriptome analysis of the genetic mechanism of DXWR in response to cold stress. (A) Differentially expressed gene statistics. (B,C) are Venn diagrams of up-regulated and down-regulated genes, respectively. (D,E) are gene ontology (GO) functional enrichment histogram and Kyoto encyclopedia of genes and genomes (KEGG) pathway enrichment histogram, respectively. 0_4, 0_12, and 4_12 represent the comparison of DXWR seedling treatment at 4 °C for 0 h and 4 h, 0 h and 12 h, and 4 h and 12 h, respectively. PGI, FAE, CSWB, SSM, ASNSM, GM, GPM, LAM, ALAM, NM, PB, FB, PST, PPER, and PPI in subfigure (E) represent pentose and glucuronate interconversions, fatty acid elongation, cutin suberine and wax biosynthesis, starch and sucrose metabolism, amino sugar and nucleotide sugar metabolism, glycerolipid metabolism, gycerophospholipid metabolism, linoleic acid metabolism, alpha-linolenic acid metabolism, nitrogen metabolism, phenylpropanoid biosynthesis, flavonoid biosynthesis, plant hormone signal transduction, protein processing in endoplasmic reticulum, and plant–pathogen interaction, respectively.

Figure 5

qSCT8 candidate gene analysis. (A) Clustering heat maps of the relative expression levels of genes within the qSCT8 localization interval determined using RNA-seq data. Standard scores (Z-scores) were used as the numerical signs to evaluate the standard deviations from the mean of the corresponding samples. (B) Expression levels of the LOC_Os08g04840 in GZX49 and DXWR after cold stress measured by qRT-PCR. The results were statistically analyzed using Student’s t-test (*** p  <  0.005). (C) Sequence comparison of LOC_Os08g04840 among GZX49 and DXWR. Vertical bars represent SNPs.

Figure 6

Cold tolerance phenotype of DXWR. (A) The phenotype of DXWR encountering heavy snow weather in its habitat. (B) Phenotypes of DXWR, Longjing 31, and GZX49 after 4 days of treatment at 0 °C at the seedling stage. LJ31 stands for LongJing31. The scale bar of (B) means 5 cm.