Multiple transcriptome comparisons reveal the essential roles of FLOWERING LOCUS T in floral initiation and SOC1 and SVP in floral activation in blueberry

Abstract

The flowering mechanisms, especially chilling requirement-regulated flowering, in deciduous woody crops remain to be elucidated. Flower buds of northern highbush blueberry cultivar Aurora require approximately 1,000 chilling hours to bloom. Overexpression of a blueberry FLOWERING LOCUS T (VcFT) enabled precocious flowering of transgenic "Aurora" mainly in non-terminated apical buds during flower bud formation, meanwhile, most of the mature flower buds could not break until they received enough chilling hours. In this study, we highlighted two groups of differentially expressed genes (DEGs) in flower buds caused by VcFT overexpression (VcFT-OX) and full chilling. We compared the two groups of DEGs with a focus on flowering pathway genes. We found: 1) In non-chilled flower buds, VcFT-OX drove a high VcFT expression and repressed expression of a major MADS-box gene, blueberry SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (VcSOC1) resulting an increased VcFT/VcSOC1 expression ratio; 2) In fully chilled flower buds that are ready to break, the chilling upregulated VcSOC1 expression in non-transgenic "Aurora" and repressed VcFT expression in VcFT-OX "Aurora", and each resulted in a decreased ratio of VcFT to VcSOC1; additionally, expression of a blueberry SHORT VEGETATIVE PHASE (VcSVP) was upregulated in chilled flower buds of both transgenic and non-transgenic' "Aurora". Together with additional analysis of VcFT and VcSOC1 in the transcriptome data of other genotypes and tissues, we provide evidence to support that VcFT expression plays a significant role in promoting floral initiation and that VcSOC1 expression is a key floral activator. We thus propose a new hypothesis on blueberry flowering mechanism, of which the ratios of VcFT-to-VcSOC1 at transcript levels in the flowering pathways determine flower bud formation and bud breaking. Generally, an increased VcFT/VcSOC1 ratio or increased VcSOC1 in leaf promotes precocious flowering and flower bud formation, and a decreased VcFT/VcSOC1 ratio with increased VcSOC1 in fully chilled flower buds contributes to flower bud breaking.

Keywords: FLOWERING LOCUS T; SHORT VEGETATIVE PHASE; SUPPRESSOR OF OVEREXPRESSION OF CONSTAN 1; Vaccinium corymbosum; chilling requirement; dormancy release; flowering mechanism; transcriptome analysis.

FIGURE 1

Phenotypic changes in when VcFT-OX transgenic “Aurora” and differentially expressed transcripts detected in four pairs of comparisons. (A) VcFT-OX transgenic “Aurora”. (B) Non-transgenic “Aurora”. (C) Flower bud formation in VcFT-OX transgenic “Aurora” and non-transgenic “Aurora”. Each data point represent an average of data from six plants. (D) Flowering chilled and non-chilled flower buds in transgenic “Aurora” VcFT-OX and non-transgenic “Aurora”. Each data point represents an average of data from three plants. Six transgenic lines (VcFT-OX1 to VcFT-OX6) and three groups of wild type (non-transgenic WT1-WT3) plants were investigated after they reached 2–3 year old. Bars show standard deviation.

FIGURE 2

Differentially expressed flowering pathway genes. (A) VcFT-OX transgenic “Aurora” and non-transgenic “Aurora” were compared in leaves and non-chilled buds, respectively. Non-chilled and chilled flower buds from 3 to 4 year old bushes were collected in late November and February, respectively. Annotated flowering pathway genes (flowering), hormone-related genes (hormone), sugar-related (sugar) genes, and MADS-box (MADS) genes were presented. *The numbers of the shared DEGs were counted according to the shared DETs. (B) Transgenic vs. non-transgenic leaves. (C) Transgenic vs. non-transgenic buds. (D–E) Responses of flowering pathway genes to VcFT-OX in mature leaf (D) and bud (E); the positive or negative regulation sign is based on the information from Arabidopsis, but it may not match the results obtained in this study. White, green, and red boxes indicate no differential expression, upregulated expression, and downregulated expression, respectively. The boxes with “?” indicate inconsistent differences among homologues.

FIGURE 3

Differentially expressed flowering pathway genes. (A) Chilled and non-chilled flower buds in transgenic “Aurora” VcFT-OX and non-transgenic “Aurora”, respectively. Non-chilled and chilled flower buds from 3 to 4 year old bushes were collected in late November and February, respectively. Annotated flowering pathway genes (flowering), hormone-related genes (hormone), sugar-related (sugar) genes, and MADS-box (MADS) genes were presented. *The numbers of the shared DEGs were counted according to the shared DETs. (B) Chilled vs. non-chilled buds of transgenic VcFT-OX “Aurora”. (C) Chilled vs. non-chilled buds of non-transgenic “Aurora”. White, green, and red boxes indicate no differential expression, upregulated expression, and downregulated expression, respectively. The boxes with “?” indicate inconsistent differences among homologues. (D,E) Responses of flowering pathway genes to fully chilled vs. non-chilled buds of non-transgenic plants (D) and VcFT-OX transgenic plants (E). The positive or negative regulation sign is based on the information from Arabidopsis, but it may not match the results obtained in this study.

FIGURE 4

Comparison of the RT-qPCR analysis result and the RNA-seq data of the selected DETs (A,B) and relative expression of three flowering pathway genes in blueberry floral buds during chilling hour accumulation (C). (A) Non-chilled flower buds (VcFT-OX “Aurora” vs. non-transgenic “Aurora” bud). (B) Non-transgenic bud (chilled vs. non-chilled). −∆∆Ct is an average of three biological and three technical replicates for each DET. EUKARYOTIC TRANSLATION INITIATION FACTOR 3 SUBUNIT H was used to normalize the RT-qPCR results. (C) The expression was normalized to VcACTIN. The first normalized data point (October) was arbitrarily set as “1” and then used to normalize the other data points. Each data point is an average of three biological and three technical replicates. The error bars indicate standard deviation. Gene IDs: VcSOC1-1 (SOC1_ARATH), VcSOC1-2 (SOC1_ARATH), VcSVP (SVP_ARATH), VcFT (HD3A_ORYSJ), VcLFY (FLO_ANTMA), and VcFUL (AGL8_SOLTU).

FIGURE 5

Interactions of FT-mediated floral initiation and SOC1-regulated floral activation in blueberry. Floral initiation signals are mainly produced in leaves. The diagram was drawn based on the data sets in Table 2. The genes listed are the major differentially expressed genes induced by either VcFT overexpression or chilling in blueberry. The positive or negative regulation sign is based on the information from Arabidopsis, but it may not match the results obtained in this study.