Repression of TERMINAL FLOWER1 primarily mediates floral induction in pear (Pyrus pyrifolia Nakai) concomitant with change in gene expression of plant hormone-related genes and transcription factors

Abstract

Floral induction is an important event in the annual growth cycle of perennial fruit trees. For pear, this event directly affects fruit production in the following year. The flower buds in many species are induced by FLOWERING LOCUS T (FT), whose effect is repressed by the meristem-expressed gene TERMINAL FLOWER1 (TFL1). In this study, we investigated the functions of pear FT and TFL1 genes during floral development. Expression of pear FTs (PpFT1a and PpFT2a) in reproductive meristems was not obviously induced prior to floral initiation, while expression of TFL1s (PpTFL1-1a and PpTFL1-2a) rapidly decreased. The induction of the productive meristem identity MADS-box gene AP1 after repression of PpTFL1s suggested a primary role for PpTFL1 in floral induction. RNA-seq analysis suggested that plant hormone-related genes and several transcription factors that were coexpressed with PpTFL1 were potentially involved in the PpTFL1-mediated floral induction. Our data indicate the essential function of TFL1 in pear floral induction and add another species in the family Rosaceae in addition to strawberry and rose that shows a role for TFL1 in floral induction.

Keywords: Floral induction; TERMINAL FLOWER1; pear; plant hormone-related genes; signal transduction; transcription factor.

Fig. 1.

Apical tips/buds of newly grown long shoots (A; white arrow) and relative expression levels of PpFT1a and PpFT2a genes in ‘Kosui’ from June to August in 2011, 2012, and 2014 (B). Error bars indicate standard error of three technical replicates. The stage that a dome-like structure (indicating occurrence of visible floral initiation) appeared is indicated by the grey zone. (This figure is available in color at JXB online.)

Fig. 2.

(A, B) Relative expression levels of PpTFL1-1a, PpTFL1-2a (A), and the meristem identity MADS-box genes PpAP1 (B) in the apical tips/buds of the newly grown long shoots in ‘Kosui’ from June to August in 2011, 2012, and 2014. (C, D) The relative expression levels of PpFUL (C) and PpSOC1 (D) in 2014. Error bars indicate standard error of three technical replicates. The stage that dome-like structure (indicating occurrence of visible floral initiation) appeared is indicated by the grey zone.

Fig. 3.

Vegetative and reproductive phase changes during floral development of apical buds from spurs of ‘Kosui’ in 2012 and 2014. Visible floral initiation of ‘Kosui’ in the apical buds of the newly grown long shoots occurred during mid to late July (arrows).

Fig. 4.

Analysis of global gene expression during floral development. (A) Cluster dendrogram showing global relationships between biological replicates and among different developmental stages. Samples were named in the format ‘year-developmental stage’; for example, ‘12–0.2’ indicates the flower bud sample of developmental stage 0.2 harvested in 2012. Note that the same stages from different years were clustered together. (B) Numbers of genes expressed in each library with RPKM>0.3. (C) Venn diagram showing the number of commonly and uniquely expressed genes among the developmental stages. (D) Developmental-stage-specifically expressed genes (+) and genes that were not expressed (−). Plant hormone-related genes and transcription factors are indicated with different shades/colors. The number of plant hormone-related genes is too low to be see in the graph. (This figure is available in color at JXB online.)

Fig. 5.

qRT-PCR confirmation of RNA-seq results for selected genes. (A) Fifty genes (Supplementary Table S3) were used for qRT-PCR confirmation using the same samples as those used for RNA-seq. The correlation coefficients (cor) between the normalized expression values of RNA-seq and qRT-PCR were calculated with the CORREL function in Excel. Note that the correlation coefficients of most genes were >0.8. (B) RPKM values (box) and relative expression identified by qRT-PCR (bar) for PpFT1a and PpTFL1-1a. (C) Relative expression of the floral differentiation-related genes PpLFY1 PpAP-1, PpFUL, and PpSOC1. Data are presented as averages of samples of the same developmental stage from 2012 and 2014. Error bars indicate standard error using the two years’ samples.

Fig. 6.

Weighted gene co-expression network analysis (WGCNA) of differentially expressed genes (DEGs) identified from ‘Kosui’ flower buds during three developmental stages. (A) Hierarchical cluster tree showing three modules of coexpressed genes. Each of the 2376 DEGs represents a leaf in the tree, and each of the three modules represents a major tree branch. The lower panel shows the modules in the designated colors. (B) Module–floral initiation stage correlations and corresponding P-values (in parentheses). The left panel shows the three modules and the number of genes in each module. The color scale on right shows module–trait correlation from –1 to 1. In the left panel, ‘floral differentiation’ is the floral developmental stages. In the right panel, ‘floral initiation’ is the binarized developmental stages with stage 0=0 and stages 0.2 and 2.0=1. (C) Heat maps showing the expression patterns of each module. (This figure is available in color at JXB online.)

Fig. 7.

Pathway analysis of the PpTFL1 coexpressed genes. The graph is the output of MapMan. (This figure is available in color at JXB online.)

Fig. 8.

Relative expression levels of selected genes in flower and leaf buds of the newly grown long shoots during floral development in 2013. Data are presented as averages of samples with the same date. Error bars indicate standard error of three technical replicates. (This figure is available in color at JXB online.)

Fig. 9.

Relative expression levels of selected genes during far-red irradiation. Far-red light treatment induced the flower bud formation as described in Ito et al. (2016). SD, short-day treatment; FR, far-red irradiation. Error bars indicate standard error of two biological replicates.