Switching genetic effects of the flowering time gene Hd1 in LD conditions by Ghd7 and OsPRR37 in rice.

Flowering time control in plants is a major limiting factor on the range of species. Day length, perceived via the photoperiodic pathway, is a critical factor for the induction of flowering. The module of GIGANTEA (GI)-CONSTANS (CO)-FLOWERING LOCUS T in the long day (LD) plant Arabidopsis is conserved in diverse plant species including the short day (SD) plant rice, where this module comprises OsGI-Heading date 1 (Hd1)-Heading date 3a. Hd1, the rice ortholog of Arabidopsis CO, has dual functions in the regulation of flowering time, promoting flowering in SD conditions and delaying it in LD conditions. We herein show genetic interactions among three LD repressor genes: Hd1, Grain number, plant height and heading date 7 (Ghd7), and Oryza sativa Pseudo-Response Regulator37 (OsPRR37). Genetic analyses, including segregation analyses, evaluations of near isogenic lines, and transformation for flowering time demonstrated that Hd1 promoted flowering time in inductive SD and non-inductive LD conditions in genetic condition of loss-of-function Ghd7 and OsPRR37 (ghd7osprr37) in rice. Functional Ghd7 or OsPRR37 may switch the genetic effects of Hd1 from the promotion to the delay of flowering times in LD conditions.

Introduction

Flowering time in crop species is a major factor limiting the range of species and a source of local adaptation. Responses to seasonal cues from changing environmental factors, including day length, temperature, and specific light quality, influence flowering time. Extensive studies on the regulation of flowering time have been conducted using the long day (LD) plant Arabidopsis, which comprises four flowering pathways: photoperiod, vernalization, autonomous, and GA-dependent (Andrés and Coupland 2012, Boss , Mouradov , Simpson and Dean 2002, Song ). A similar photoperiod flowering pathway has been reported in the short day (SD) plant rice (Tsuji ).

A series of genes are associated with photoreceptors and circadian rhythms for flowering time control in plants. CONSTANS (CO) and FLOWERING LOCUS T (FT) are the central integrators of the photoperiod pathway for flowering in Arabidopsis (Samach and Coupland 2000). CO induces FT expression in inductive LD conditions (Corbesier , Kardailsky , Kobayashi , Putterill , Tiwari Turck ) and mediates the circadian clock and photoperiod for flowering time (Suarez-Lopes ). GIGANTEA (GI) is a clock-associated gene that regulates CO expression (Huq ). The Arabidopsis GI-CO-FT module is the central mechanism and is conserved in a diverse range of plant species, including as the OsGI-Heading date 1 (Hd1)-Heading date 3a (Hd3a) module in rice (Tsuji , 2013, Valverde 2011). Hd1, the rice ortholog of the Arabidopsis floral activator CO, promotes flowering by activating Hd3a in SD conditions, whereas it delays flowering by suppressing the expression of Hd3a in LD conditions (Hayama , Izawa , Kojima , Yano ). Although the module is conserved between LD and SD plants, functional differences between CO in Arabidopsis and Hd1 in rice remain unclear.

Furthermore, a unique pathway, Grain number, plant height and heading date 7 (Ghd7)-Early heading date 1 (Ehd1)-Hd3a/Rice FT-like 1 (RFT1), is involved in rice flowering time control. Ehd1 encodes a B-type response regulator and promotes flowering time through the induction of RFT1 in SD and LD conditions (Doi , Komiya , Tamaki ). Ghd7 and Oryza sativa Pseudo-Response Regulator37 (OsPRR37) suppress Ehd1 expression, delaying flowering time only in LD conditions (Gao , Shrestha , Xue ). Ghd7 encodes a CCT (CO, CO-LIKE, and TIMING OF CAB1) domain protein and exerts major effects on the number of grains per panicle, plant height, and heading date (Xue ). Ghd7 is induced in LD conditions and represses Hd3a expression, thereby delaying flowering time (Weng , Xue ). There are no orthologs of Ehd1 or Ghd7 in the Arabidopsis genome.

OsPRR37 functions as a floral repressor in LD conditions (Koo ). OsPRR37 is an ortholog of the circadian clock genes PRR3/7 in Arabidopsis and is expressed in LD and SD conditions. However, OsPRR37 represses Hd3a expression in LD conditions only (Gao , Koo , Murakami , Nakamichi ). The genetic interactions of OsPRR37 may play an important role in flowering time control in LD conditions (Fujino and Sekiguchi 2005a, Gao , Koo , Shibaya ).

Asian cultivated rice, Oryza sativa L., originated from the tropics. Extensive efforts by rice breeding programs to improve flowering time may have contributed to make rice production possible in various climatic conditions at latitudes ranging between 53°N and 40°S (Lu and Chang 1980). Early flowering time, through a decrease in photoperiod sensitivity, may have played an important role in expanding the range of rice (Izawa 2007, Shrestha , Zheng ). Based on a series of genetic analyses on early flowering time, we identified QTLs for extremely early flowering behavior using unique natural variations originating from Hokkaido (41–45°N latitude), Japan, a region with a long natural day length of more than 15 hours (Fujino and Sekiguchi 2005a, 2005b, 2008, Nonoue ).

In the present study, we investigated genetic interactions expressing flowering time in non-inductive LD conditions by focusing on three LD repressor genes: Hd1, Ghd7, and OsPRR37. We found that functional Hd1 promoted flowering time in LD and SD conditions in genetic conditions with ghd7osprr37. The present results provide new rice phenotypes with extremely early flowering behaviors, which are useful for not only commercial plant breeding programs, but also further investigations on flowering time control in SD cereals.

Materials and Methods

Plant materials

Progenies in the F2 and F3 generations of a cross between Hoshinoyume (HS) and Akitakomachi (AKT) were used in genetic analyses of flowering time. The F2 population (n = 144 plants) derived from the self-pollination of F1 plants was used in a segregation analysis of flowering time. Four types of F3 populations (96 plants/population), populations I–IV, were developed from the self-pollination of F2 plants. Based on the genotype of the three genes, Hd1, Ghd7, and OsPRR37, four F2 plants with heterozygous Hd1 were selected: Ghd7OsPRR37, Ghd7osprr37, ghd7OsPRR37, and ghd7osprr37. In addition, three types of near isogenic lines (NILs) of HS, Hd1ghd7osprr37 described in Nemoto were used: HSGhd7 for functional Ghd7, HShd1 for loss-of-function hd1, and HShd1Ghd7 for hd1 and Ghd7. The genotypes of the flowering time genes in the plant materials used in the present study are summarized in Supplemental Table 1.

Cultivation, trait evaluation, transformation, growth conditions in the growth chamber, and DNA analyses (Supplemental Table 2) are described in Supplemental Text 1.

Results

Segregation analyses of flowering time in natural field conditions in Hokkaido

To clarify the roles of the three LD repressor genes, Hd1, Ghd7, and OsPRR37, in flowering time control, segregation analyses on flowering time were performed in natural field conditions in Hokkaido. Based on the sequence alignments of the three LD repressors in the parents, HS and AKT, we found functional nucleotide polymorphisms (FNPs) that are known to cause the loss-of-function of these genes (Supplemental Fig. 1) (Gao , Koo , Xue , Yano ). HS carried a functional Hd1 and loss-of-function ghd7 and osprr37, Hd1ghd7osprr37, whereas AKT carried a loss-of-function hd1 and functional Ghd7 and OsPRR37, hd1Ghd7OsPRR37 (Supplemental Fig. 1).

The flowering time of the parents were 95.0 ± 0.0 days in HS and 108.6 ± 2.2 days in AKT. The segregation of flowering time in the F2 population (n = 144) derived from the cross between HS and AKT was continuous (95–131 days) (Fig. 1, Supplemental Fig. 2). Significant differences in mean flowering time were detected at the three chromosomal regions around Hd1, Ghd7, and OsPRR37 (Table 1). Flowering time was significantly later in plants with Hd1 (120.2 days) than in those with hd1 (104.8 days). Additionally, flowering time was significantly later in plants with Ghd7 (115.7 days) than in those with ghd7 (105.1 days). Furthermore, flowering time was significantly later in plants with OsPRR37 (116.1 days) than in those with osprr37 (107.1 days). These results clearly indicated that the three chromosomal regions around the three LD repressors were responsible for flowering time control. However, ten plants with Hd1 expressed the early type flowering time (range 95–115 days) similar to that of plants with hd1 (Fig. 1, Table 1). These results clearly suggested putative interactions between the chromosomal regions around the three LD repressor genes, Hd1, Ghd7, and OsPRR37.

Fig. 1

Frequency distribution of flowering time (days) in the F2 population derived from the cross between Hoshinoyume and Akitakomachi. Black, white, and gray bars indicate the three genotypes of the Hd1 gene; Hd1/Hd1, hd1/hd1, and Hd1/hd1, respectively.

Table 1

Flowering time in the F2 population derived from the cross between Hoshinoyume (HS) and Akitakomachi (AKT)

Marker	Chromosome	Targeted gene	Homozygous for the HS type			Homozygous for the AKT type			Heterozygous
			n	FT	Range	n	FT	Range	n	FT	Range
Hd1H43	6	Hd1	34	120.2 ± 9.1***	95–131	32	104.8 ± 3.7	99–115	75	111.7 ± 7.0	95–127
RM1306	7	OsPRR37	39	107.1 ± 8.0	95–127	32	116.1 ± 8.2***	103–131	66	113.2 ± 8.2	103–131
GHD7CAPS2	7	Ghd7	32	105.1 ± 6.5	95–115	37	115.7 ± 7.4***	103–131	63	112.6 ± 8.4	103–131

n indicates the number of plants. Flowering time (FT) is expressed as mean ± SD. Range (days) shows the variation of FT in each genotype.

Asterisks indicate significant differences in FT between plants with the chromosomal region including functional allele and the loss-of-function allele in the genes (*** P < 0.001, Student’s t test).

Genetic interactions among LD repressors

To investigate genetic interactions among the three LD repressor genes, Hd1, Ghd7 and OsPRR37, the segregation of flowering time by the Hd1 genotype was assessed using all four possible combinations of Ghd7 and OsPRR37. We developed the four F3 populations between HS and AKT, populations I to IV, based on the genotypes of the three repressors.

In population I with functional Ghd7 and OsPRR37 (Ghd7OsPRR37), the flowering time of plants with Hd1 was 134.3 ± 3.3 days (range 120–135 days), whereas that with hd1 was 105.8 ± 2.1 days (range 104–112 days) (Fig. 2A, Supplemental Table 3). In population II with functional Ghd7 and loss-of-function osprr37 (Ghd7osprr37), flowering time was later in plants with Hd1 (118.6 ± 6.3 days; range 112–135 days) than in those with hd1 (101.0 ± 1.2 days; range 100–104 days) (Fig. 2B, Supplemental Table 3). In population III with loss-of-function ghd7 and functional OsPRR37 (ghd7OsPRR37), flowering time was later in plants with Hd1 (106.8 ± 2.3 days; range 102–110 days) than in those with hd1 (101.7 ± 1.5 days; range 100–104 days) (Fig. 2C, Supplemental Table 3). In these three populations, flowering time was significantly later in plants with Hd1 than in those with hd1. In contrast, in population IV with loss-of-function ghd7 and osprr37 (ghd7osprr37), flowering time was significantly earlier in plants with Hd1 (85.5 ± 2.2 days; range 82–90 days) than in those with hd1 (96.4 ± 1.3 days; range 94–98 days) (Fig. 2D, Supplemental Table 3).

Fig. 2

Frequency distributions of flowering time by Hd1 in different genetic backgrounds with combinations of Ghd7 and OsPRR37. Populations heterozygous for the Hd1 gene with all four possible genetic combinations of Ghd7 and OsPRR37. (A) Population I, Ghd7OsPRR37, (B) population II, Ghd7osprr37, (C) population III, ghd7OsPRR37, (D) population IV, ghd7osprr37. Black, white, and gray bars indicate the three genotypes of the Hd1 gene; Hd1/Hd1, hd1/hd1, and Hd1/hd1, respectively. Functional alleles; Hd1, Ghd7, and OsPRR37, Loss-of-function alleles; hd1, ghd7, and osprr37.

The genetic effects of Hd1 with the Ghd7 and OsPRR37 background genotype are summarized in Fig. 3 as the difference in flowering time between the plants carrying Hd1 and hd1. Plants carrying Hd1 delayed flowering time by 28.5 days with Ghd7OsPRR37, 17.6 days with Ghd7osprr37, and 5.0 days with ghd7OsPRR37, whereas flowering time was promoted by 10.9 days with ghd7osprr37 from plants carrying hd1 (Fig. 3, Supplemental Table 3). These results clearly showed the genetic interactions between the three LD repressors. Based on the genotypes of these repressors, we hypothesized a role for LD repressors in flowering time control. Functional Hd1 may promote flowering time in ghd7osprr37. Ghd7 or OsPRR37 may switch the genetic effects of Hd1 from the promotion to the delay of flowering time.

Fig. 3

Genetic effects of Hd1 on flowering time with different genetic conditions of Ghd7 and OsPRR37 expressed as a difference in flowering time between plants carrying Hd1 and hd1. + and − indicate functional and loss-of-function alleles in each gene, respectively.

Response to day lengths in near isogenic lines for Hd1 and Ghd7

To clarify the genetic roles of the LD repressors, Hd1 and Ghd7, in day length responses, NILs were grown in growth chambers that included controlled temperature and light conditions: 14.5 hours (LD) or 10 hours (SD) (Table 2). Three types of NILs with the genetic background of HS were developed; HShd1 carrying hd1, HSGhd7 carrying Ghd7, and HShd1Ghd7 carrying hd1 and Ghd7, and compared with HS carrying Hd1 and ghd7 (Supplemental Table 1).

Table 2

Day length response of near isogenic lines

Line	Genotype of Hd1 and Ghd7	Day length
		Long day		Short day
		n	FT	n	FT
HShd1	hd1ghd7	5	59.8 ± 1.1	5	59.8 ± 1.5
HS	Hd1ghd7	4	56.5 ± 0.6**	3	49.7 ± 1.6***
HShd1Ghd7	hd1Ghd7	5	72.4 ± 2.1	5	77.6 ± 2.9
HSGhd7	Hd1Ghd7	3	87.3 ± 2.1***	5	54.8 ± 1.5***

n indicates the number of plants. Flowering time (FT) is expressed as mean ± SD.

Asterisks indicate significant differences from the loss-of-function hd1 in the presence/absence Ghd7 (** P < 0.01, *** P < 0.001, Student’s t test).

In LD conditions, the flowering time of HS (56.5 days) was significantly earlier than that of HShd1 (59.8 days). In SD conditions, Hd1 (49.7 days) promoted flowering time significantly more than HShd1 (59.8 days) (Table 2). Functional Hd1 promoted flowering time in LD and SD conditions in ghd7osprr37.

Functional Ghd7 inhibited this promotion of flowering time by Hd1 in LD conditions only. The flowering time of HSGhd7 (87.3 days) was significantly later than that of HShd1Ghd7 (72.4 days) in LD conditions (Table 2). Functional Hd1 in HSGhd7 (54.8 days) promoted flowering time significantly more than HShd1Ghd7 (77.6 days) in SD conditions (Table 2).

These results confirmed the role for Ghd7 in the effects of Hd1 in different day lengths. Functional Hd1 with ghd7osprr37 may promote flowering time in LD and SD conditions. Functional Ghd7 inhibits this promotion of flowering time by Hd1 in a manner that is dependent on day length in LD conditions only.

Promotion of flowering time by Hd1

To confirm whether functional Hd1 promotes flowering time in ghd7osprr37, the functional Hd1 gene was transformed into HShd1, hd1ghd7osprr37. All three independent transformants showed significantly promoted flowering time in LD and SD conditions (Table 3). These results confirmed that functional Hd1 promoted flowering time in LD and SD conditions in ghd7osprr37.

Table 3

Changes in flowering time in transformants of an NIL, Hoshinoyume (HS) carrying the loss-of-function hd1, HShd1, with the functional Hd1 gene

Line	Day length
	Long day		Short day
	n	FT	n	FT
Transgene (+)	16	55.8 ± 2.3**	18	51.0 ± 2.4***
Transgene (−)	6	61.8 ± 2.0 ns	6	59.3 ± 2.0 ns
Vector control	3	60.0 ± 1.0	4	61.8 ± 2.5
HShd1	3	62.7 ± 4.9 ns	3	64.0 ± 1.6 ns

n indicates the number of plants. Flowering time (FT) is expressed as mean ± SD.

Asterisks indicate significant differences from the vector control (** P < 0.01, *** P < 0.001, no significance ns, Student’s t test).

Transgene (+) and (−) include three and two independent T1 lines, respectively.

Discussion

We herein presented genetic evidence for flowering time control with a focus on the three LD repressors, Hd1, Ghd7, and OsPRR37. Depending on the genotype of Ghd7 and OsPRR37, the genetic effects of Hd1 on flowering time switched between promotion or delay (Fig. 2). Hd1 promotes flowering time in inductive SD and non-inductive LD conditions in ghd7osprr37 in rice. Additionally, we concluded that Ghd7 or OsPRR37 switched the genetic effects of Hd1. Ghd7 and OsPRR37 additively delayed flowering time in natural field conditions as LD repressors.

Ghd7 is involved in a unique phenotype for flowering time, the switching of the genetic effects of Hd1 (Zhang ). However, we clearly showed the complete genetic basis in that the genotypes of the two genes, Ghd7 and OsPRR37, switched the genetic effects of Hd1 on flowering time. CO in Arabidopsis is known as a hub for signal integration between the circadian clock and photoperiod sensitivity in flowering (Johansson and Staiger 2015, Shim , Suárez-López , Valverda , Yanovsky and Kay 2002). The redundancy of Ghd7 and OsPRR37 for Hd1 function suggested that Hd1 also plays a role as a hub between photoperiod and circadian rhythms in rice. Biological interactions have been reported between Ghd7 and Hd1 (Nemoto ). In addition, Hd1 and OsPRR37 form the transcriptional repressor complex (Goretti ). In Arabidopsis, PSEUDO RESPONSE REGULATORs stabilize the CO protein to promote flowering in response to daylength (Hayama ). The molecular characterization of genetic interactions among Hd1, Ghd7, and OsPRR37 identified in the present study may provide insights into flowering time control in rice (Hori , Tsuji ).

The optimization of flowering time is a key aspect of grain productivity in cereal crops, including rice. Ghd7 and OsPRR37 play a major role not only in the adaptability of cultivated rice around the world (Fujino and Sekiguchi 2005a, Nonoue , Shibaya ), but also in wheat (Turner ), barley (Hemming , Turner ,), and sorghum (Murphy , Yang ). Our results suggest that not only a deficiency in gene functions, but also altered genetic interactions may be involved in early flowering. Flowering time plays a major role in crop yield, and a clearer understanding of genotypes that influence flowering time control will facilitate the development of new varieties.