Integrated GWAS, linkage, and transcriptome analysis to identify genetic loci and candidate genes for photoperiod sensitivity in maize

Abstract

Introduction: Maize photosensitivity and the control of flowering not only are important for reproduction, but also play pivotal roles in the processes of domestication and environmental adaptation, especially involving the utilization strategy of tropical maize in high-latitude regions.

Methods: In this study, we used a linkage mapping population and an inbred association panel with the photoperiod sensitivity index (PSI) phenotyped under different environments and performed transcriptome analysis of T32 and QR273 between long-day and short-day conditions.

Results: The results showed that PSIs of days to tasseling (DTT), days to pollen shedding (DTP), and days to silking (DTS) indicated efficacious interactions with photoperiod sensitivity for maize latitude adaptation. A total of 48 quantitative trait loci (QTLs) and 252 quantitative trait nucleotides (QTNs) were detected using the linkage population and the inbred association panel. Thirteen candidate genes were identified by combining the genome-wide association study (GWAS) approach, linkage analysis, and transcriptome analysis, wherein five critical candidate genes, MYB163, bif1, burp8, CADR3, and Zm00001d050238, were significantly associated with photoperiod sensitivity.

Discussion: These results would provide much more abundant theoretical proofs to reveal the genetic basis of photoperiod sensitivity, which would be helpful to understand the genetic changes during domestication and improvement and contribute to reducing the barriers to use of tropical germplasm.

Keywords: GWAS; QTL; candidate gene; genetic loci; joint analysis; photoperiod sensitivity.

Figure 1

Linkage analysis for the photoperiod sensitivity index (PSI) of days to tasseling (DTT), days to pollen shedding (DTP), and days to silking (DTS), the leaf number (LN), plant height (PH), and ear height (EH) in the F_2:3 population. The plant (A) and ear (B) phenotypes of the parental lines, T32 and QR273, in the different latitude areas (Sanya and Zhangye). (C) The PSI distribution of flowering time traits, PH, EH, and LN in the F_2:3 population at the Sanya and Zhangye.

Figure 2

The histogram for PSIs of flowering time traits, (A) DTT, (B) DTP, (C) DTS, (D) PH, (E) EH, and (F) LN in the F2:3 population between pair-three latitude areas, including Guiyang vs. Zhangye (GY/ZY), Sanya vs. Guiyang (SY/GY), and Sanya vs. Zhangye (SY/ZY).

Figure 3

PSIs of DTT (A), DTP (B), DTS (C), PH (D), EH (E), and LN (F) in the association mapping population between pair-three latitude areas, including GY/ZY, SY/GY, and SY/ZY. The Student’s t-test was introduced for statistical analysis.

Figure 4

Manhattan and Q-Q plots from a mixed linear model for maximum quantum efficiency for PSIs of flowering time traits in the association mapping population. (A) PSIs for GY/ZY, (B) PSIs for SY/GY, and (C) PSIs for SY/ZY. (D–F) show the common significant QTNs of PSIs between pair-three latitude areas, respectively. (G) shows that the common significant QTNs were collectively related to PSIs in three comparisons.

Figure 5

The differentially expressed genes (DEGs) identified between the two foundation inbred lines of T32 and QR273 in long-day (Zhangye) and short-day (Sanya) environments. (A) The number of upregulated and downregulated DEGs by long-day condition. (B) The common upregulated and downregulated DEGs in T32 and QR273. (C, D) The most enriched GO terms for the upregulated (C) and downregulated (D) genes.

Figure 6

Information of candidate genes identified by genome-wide association study (GWAS) and transcriptome analysis. (A) Annotation information of seven selected candidate genes. Red text indicates DEGs in both T32 and QR273. (B–D) show the relative expression levels of candidate genes in T32 and/or QR273 by qRT-PCR. (B) The relative expression of Zm00001d001895 and Zm00001d034510 in both T32 and QR273, (C) the expression of Zm00001d034511 in T32, and (D) the expression of Zm00001d034961, Zm00001d034514, Zm00001d046900, and Zm00001d034515 in QR273. Asterisks (*) indicate statistically significant differences between long-day (LD) and short-day (SD) conditions (*p < 0.05, **p < 0.001).

Figure 7

The relative expression of key candidate genes selected by GWAS and transcriptome analysis and the number of DEGs in T32 and QR273. (A–E) The relative expression levels of key candidate genes including Zm00001d023664 (A), Zm00001d008573 (B), Zm00001d021303 (C), Zm00001d047664 (D), and Zm00001d049244 (E). Asterisks (*) indicate remarkably significant differences between long-day (LD) and short-day (SD) conditions (***p < 0.0001). (F) The number of upregulated and downregulated DEGs in T32 and QR273.

Figure 8

Information of candidate genes identified by QTL and GWAS analysis. (A) Annotation information of 13 selected candidate genes. (B–D) show the relative expression levels of main candidate genes by qRT-PCR including Zm00001d050238 (B), Zm00001d042136 (C), Zm00001d012255 (D), Zm00001d008753 (E), Zm00001d012252 (F), and Zm00001d008749 (G). Asterisks (*) indicate statistically significant differences between long-day (LL) and short-day (SL) conditions (*p < 0.05, **p < 0.001).