HvVRN2 responds to daylength, whereas HvVRN1 is regulated by vernalization and developmental status

Abstract

Two genetic loci control the vernalization response in winter cereals; VRN1, which encodes an AP1-like MADS-box transcription factor, and VRN2, which has been mapped to a chromosome region containing ZCCT zinc finger transcription factor genes. We examined whether daylength regulates expression of HvVRN1 and HvVRN2. In a vernalization-responsive winter barley (Hordeum vulgare), expression of HvVRN1 is regulated by vernalization and by development, but not by daylength. Daylength affected HvVRN1 expression in only one of six vernalization-insensitive spring barleys examined and so cannot be a general feature of regulation of this gene. In contrast, daylength is the major determinant of expression levels of two ZCCT genes found at the barley VRN2 locus, HvZCCTa and HvZCCTb. In winter barley, high levels of HvZCCTa and HvZCCTb expression were detected only when plants were grown in long days. During vernalization in long-day conditions, HvVRN1 is induced and expression of HvZCCTb is repressed. During vernalization under short days, induction of HvVRN1 occurs without changes in HvZCCTa and HvZCCTb expression. Analysis of HvZCCTa and HvZCCTb expression levels in a doubled haploid population segregating for different vernalization and daylength requirements showed that HvVRN1 genotype determines HvZCCTa and HvZCCTb expression levels. We conclude that the vernalization response is mediated through HvVRN1, whereas HvZCCTa and HvZCCTb respond to daylength cues to repress flowering under long days in nonvernalized plants.

Figure 1.

Induction of HvVRN1 during development and by vernalization treatment under LD or SD conditions in a winter barley cultivar. A, RNA gel-blot analysis of HvVRN1 levels in plants in SD or LD conditions. RNA was extracted from plants (cv Sonja; Hvvrn1/HvVRN2) harvested after 4, 6, 8, 10, or 12 weeks. HvVRN1 expression was assayed by high-stringency hybridization of RNA gel blots with a HvVRN1-specific riboprobe. Ethidium bromide staining of ribosomal RNA is shown as a loading comparison. B, Average days until head emergence of the winter barley (cv Sonja) grown under SD or LD conditions with or without vernalization. Error bars show sd. The asterisk indicates that in SD, four of eight plants grown died without producing heads. C, Real-time RT-PCR quantification of HvVRN1 expression, relative to ACTIN, in SD or LD conditions. SD-grown plants (21-d-old) were shifted to LD for 14 d, whereas control plants were maintained under SD. The indicated error is the sd for three technical repeats. D, RNA gel-blot analysis of HvVRN1 levels during vernalization in either SD or LD conditions. RNA was extracted from plants grown for 3 weeks in SD conditions, then vernalized in either SD or LD conditions. Plants were harvested after 1, 3, 5, 7, or 9 weeks of vernalization. E, Real-time RT-PCR quantification of HvVRN1 expression, relative to ACTIN, in LD or SD conditions after 7 and 9 weeks of vernalization (+V), and 2 weeks after the end of the vernalization treatment. The indicated error is the sd for three technical repeats. LD, Long day; SD, short day; SDV, short day with vernalization; LDV, long day with vernalization.

Figure 2.

Comparison of HvVRN1 expression in different spring barley genotypes under LD or SD conditions. Real-time RT-PCR quantification of HvVRN1 expression, relative to ACTIN, in 14-d-old plants grown in glasshouse conditions under either LD or SD. The indicated error is the sd for three technical repeats. The cultivars used are Golden Promise (HvVRN1/Hvvrn2), Morex (HvVRN1/Hvvrn2), Icheon Naked (HvVRN1/Hvvrn2), Chame 11 (HvVRN1/HvVRN2), Sikangense (HvVRN1/HvVRN2), Olli (HvVRN1/Hvvrn2/HvVRN3), and Himalayense type 15 (HvVRN1/HvVRN2). LD, Long day; SD, short day.

Figure 3.

Expression of HvZCCTa and HvZCCTb is higher in LD-grown plants. A, RNA gel-blot analysis of HvZCCTa and HvZCCTb expression in LD- or SD-grown plants (21-d-old cv Sonja; Hvvrn1/HvVRN2). Blots were hybridized at high stringency with a riboprobe that is predicted to hybridize to the HvZCCTa, HvZCCTb, and HvZCCTc genes. Ethidium bromide staining of ribosomal RNA is shown as a loading comparison. B, Semiquantitative RT-PCR assay of HvZCCTa or HvZCCTb expression in LD- or SD-grown plants using gene-specific primers. Expression of ACTIN is shown as a loading comparison. LD, Long day; SD, short day.

Figure 4.

Expression of HvZCCTa and HvZCCTb responds rapidly to changes in daylength. A, Semiquantitative RT-PCR analysis of HvZCCTa or HvZCCTb expression, using gene-specific primers, in samples from plants (cv Sonja; Hvvrn1/HvVRN2) grown in SD for 21 d and then shifted to LD. Samples were taken at 1, 2, and 7 d after the shift to LD treatment. Control plants that remained in SD conditions were harvested at identical time points. Expression of ACTIN is shown as a loading comparison for all samples. B, Real-time RT-PCR quantification of HvZCCTa and HvZCCTb expression, relative to ACTIN, at 1, 2, 3, 5, or 7 d after the beginning of LD treatment. The indicated error is the sd for three technical repeats. C, Real-time PCR quantification of HvZCCTa and HvZCCTb expression in plants grown under LD for 21 d and then shifted to SD in either glasshouse (−V) or vernalizing (+V) conditions. Samples were taken at 1, 2, 3, 5, and 7 d after the beginning of SD treatment. The indicated error is the sd for three technical repeats. LD, Long day; SD, short day.

Figure 5.

Expression of HvZCCTa and HvZCCTb shows diurnal rhythm in plants grown in LD. Real-time RT-PCR quantification of HvZCCTa and HvZCCTb expression levels, relative to ACTIN, in 21-d-old plants (cv Sonja; Hvvrn1/HvVRN2) growing under either LD (♦) or SD (▪) that were harvested at 2-h intervals throughout a 24-h period. The indicated error is the sd for three technical repeats. Light and dark periods are indicated below graph as white or black bars, respectively, for both LD and SD treatments. LD, Long day; SD, short day.

Figure 6.

Vernalization does not block daylength induction of HvZCCTa or HvZCCTb. Semiquantitative RT-PCR using gene-specific primers was used to compare HvZCCTa or HvZCCTb expression in plants that have been vernalized for 9 weeks under SD and then returned to glasshouse conditions for 14 d under either SD (V + SD) or LD (V + LD). Plants grown under glasshouse conditions for an equivalent length of time (98 SD) are shown as a developmental control. Expression was also assayed in nonvernalized plants grown under SD for 21 d and then shifted to LD for 2 weeks. Expression of HvZCCTa and HvZCCTb in plants maintained in SD (35 SD) are shown as a developmental control. Expression of ACTIN is shown as a loading comparison for all samples. LD, Long day; SD, short day.

Figure 7.

RT-PCR analysis of HvZCCTa and HvZCCTb expression during vernalization under long-day conditions. A, Real-time RT-PCR quantification of HvZCCTa and HvZCCTb expression, relative to ACTIN, in vernalized and nonvernalized plants. Plants (cv Sonja; Hvvrn1/HvVRN2) were germinated and grown in long days. After 21 d, plants were placed at vernalizing temperatures in long days for 9 weeks (+V) or remained at normal glasshouse temperatures (−V) for the same length of time. The indicated error is the sd for three technical repeats. B, HvZCCTa and HvZCCTb expression levels were assayed separately using gene-specific primers at a number of points during the vernalization treatment (1, 3, 5, 7, and 9 weeks). Expression levels are compared to those in nonvernalized control plants harvested at identical time points. Expression of ACTIN is shown as a loading comparison for all samples.

Figure 8.

Expression of HvZCCTb and HvZCCTb is repressed by HvVRN1. RT-PCR analysis of HvZCCTa and HvZCCTb expression, using gene-specific primers, in samples extracted from nonvernalized plants of the winter parent genotype (cv Halcyon; Hvvrn1/HvVRN2; H), four doubled haploid lines of the winter genotype (Hvvrn1/HvVRN2), and five doubled haploid lines that carry the spring allele at HvVRN1 (HvVRN1/HvVRN2). Expression of ACTIN is shown as a loading comparison for all samples.

Figure 9.

A new model for regulation of flowering time in winter varieties by VRN1 and VRN2. Plants progress developmentally from vegetative to reproductive growth. Flowering is promoted by vernalization, daylength, and developmental pathways. VRN1 has a role in determining flowering time in response to vernalization. VRN1 expression is also required for floral transition and occurs if flowering is triggered by developmental pathways. VRN2 acts in long-day (LD) conditions to slow developmental progression toward flowering in plants that have not been vernalized. Loss of VRN2 leads to more rapid floral transition, resulting in earlier expression of VRN1 under LD conditions. Induction of VRN1 by vernalization (or expression from spring VRN1 alleles) overrides repression of flowering by VRN2 and triggers floral transition, possibly by repressing VRN2 expression (dotted line).