Natural Variation of OsHd8 Regulates Heading Date in Rice

: Heading date, as one of the most important agronomic traits, is a fundamental factor determining crop yield. Although diverse genes related to heading date have already been reported in rice, the key gene that regulates heading date is still poorly understood. Here, we identiﬁed a heading date regulator, heading date 8 ( OsHd8 ), which promoted the heading date under long-day conditions and encoded a putative HAP3 subunit of the CCAAT-box-binding transcription factor. It is localized in the nucleus and expressed in various tissues. Sequence analysis revealed that there were four SNPs and one InDel in the promoter region of OsHd8 , which was involved in the regulation of some ﬂoral regulators including GHD7.1 , SDG718 , OsGI and HDT1 . Further evolutionary analysis showed that OsHd8 presents divergence between indica and japonica , showing natural selection during the domestication of cultivated rice. These results indicate that OsHd8 plays an important role in the regulation of heading date, and may be an important target for rice breeding programs.


Introduction
Flowering is an important process in plant transition from vegetative to reproductive growth [1,2]. Rice flowering is often affected by the external environment, including light, temperature, and nutritional conditions [3]. Appropriate flowering time is not only beneficial to the reproductive development of rice but also affects the yield of rice [4].
To date, many flowering-related genes have been identified in plants [5][6][7][8][9][10]. Ghd7 encodes a protein with a CCT (CO, CO-LIKE and TIMING OF CAB1) domain that delays the heading date, increases the plant height and promotes panicle size development [5]. In Arabidopsis thaliana, CO (CONSTANS), as a key factor in the photoperiod pathway, promotes flowering under long-day (LD) conditions [11]. FT (FLOWERING LOCUS T) is a member of the PEBP gene family that shares homology with RAF kinase inhibitor proteins (RKIPs; these are activated by CO to regulate flowering [12][13][14]. Hd1 and Hd3a, a rice ortholog of the Arabidopsis CO and FT gene, respectively, have been identified to promote heading under short day (SD) conditions [6,9]. However, the regulation of Hd1 and Hd3a in rice is different from that in Arabidopsis thaliana. PhyB (phytochromes B) is involved in the post-translation regulation of Hd1 to regulate Hd3a expression, while overexpression of Hd1 inhibits Hd3a expression and delays flowering depending on phyB under SD conditions [15]. RFT1 (RICE FLOWERING LOCUS T 1), encoding the mobile flowering signal, is the closest homolog of Hd3a. In Hd3a-RNAi transgenic plants, the RFT1 gene is activated to promote flowering under SD conditions [10,16]. Ehd1 (Early heading date 1) encodes a B-type response regulator, which can promote the flowering of rice under SD conditions by regulating expression of the FT-like gene [8]. Ehd1, as a unique

Heading Date Investigation Map-Based Cloning
The 1880, JH2B two parents and F 2 population were investigated for the heading date. Each plant recorded the heading date when the rice plant began to head with a single spike. A total of 606 F 2 plants with extreme phenotypes, which defined as the first 15% of heading in the whole F 2 population, derived from the cross of 1880/JH2B were used for mapping of OsHd8. Firstly, it was primarily located in the 3.3~6.53 Mb region of the chromosome 8 by BSA-Seq (bulked segregant sequencing), and then we used the F 2 population of the target Agronomy 2022, 12, 2260 3 of 16 QTL and mapped OsHd8 between the marker SNP7730 and SSR-1 (Supplementary Table  S2) using a chromosome fragment substitution analysis.

BSA-Seq and Analysis of the Seq-BSA Data
For the BSA-seq, two DNA pools were developed by selecting the extreme early heading date plants and extreme late heading date plants from the F 2 population. The extreme phenotype of the early growth period was defined as the first 15% of heading in the whole F 2 population, and the extreme phenotype of the late growth period was defined as the last 15% of the heading in the whole F 2 population. The early heading date pool (Z-pool) was made by mixing equal amounts of DNA from 50 extreme early heading date plats, and the late heading date pool (W-pool) was made by mixing equal amounts of DNA from 50 extreme late heading date plats. The DNA isolated from the two DNA pools were prepared for BSA sequencing.
Libraries for all the DNA pools were prepared according to the Illumina TruSeq Library Construction Kit. The DNA libraries were sequenced on Illumina Hiseq Xten PE15 (Illumina Inc., San Diego, CA, USA). The short reads from the two DNA pools were aligned to a Nipponbare reference genome (MSU Rice Genome Annotation Project Release 7) using the BWA software [36]. Reads of the Z-pool and W-pool were separately aligned to a Nipponbare reference genome and consensus sequence reads to call SNPs with the SAM tools software [36]. The SNP loci between the test samples and reference genome were obtained using the GATK software [37]. The Euclidean distance (ED) and SNP-index were calculated to identify the candidate regions of the genome associated with the heading date [38].

RNA Isolation and Quantitative Reverse Transcription-PCR (qRT-PCR) Analysis
Total RNAs were extracted from young rice leaves of 3-week-old seedlings using the TRIzol Reagent according to the manufacturer's protocol (Invitrogen, Carlsbad, CA, USA). Total RNA was used for synthesizing the cDNA with a reverse transcription kit (Vazyme, Nanjing, China). Quantitative reverse transcriptase (qRT-PCR) was performed on a Roche LightCycler 480II instrument using the Hieff qPCR SYBR Green Master Mix (No Rox) following the manufacturer's instructions (Yeasen, Shanghai, China). The actin gene was used as the internal control. All assays were performed with three biological replicates and the relative expression level was analyzed with the 2 −∆∆CT method [39].

Vector Constructions and Plant Transformation
In order to construct a complementary vector, the upstream 2 kb promoter region of the ATG, gene coding region, and downstream 1 kb region of the OsHd8 gene, were amplified from the genomic DNA of 1880 and then constructed into the complementary vector pCAMBIA1301. To prepare the construction of the Hd8 overexpression vector, the CDS of OsHd8 was amplified from 1880 and was introduced into the vector pCAMBIA1301-Ubi. All the constructed vector plasmids were transformed into an Agrobacterium tumefactions strain EHA105 and transferred by Agrobacterium-mediated transformation into JH2B.

Dual Luciferase (LUC) Analysis
The promoter regions were amplified from 1880 and JH2B, then cloned into the pGreenII 0800-LUC vector. Subsequently, the vectors co-transformed with GV3101 (pSoup-p19) chemically competent cells. Overnight, A. tumefaciens were cultured at 28 • C and collected by centrifugation and re-suspended in the MS medium with OD 600 = 1.5, and incubated at 28 • C and 200 rpm for 3 h. The strains were infiltrated into tobacco (Nicotiana benthamiana) leaves and tested after 3 days (long day/white light). Leaves were infiltrated with 1 mM D-luciferin solution and images were captured using a Tanon 5200 Multi imaging system. Quantification was performed using the Dual-Luciferase Reporter Assay System (Promega, Madison, WI, USA). All the assays were performed on three biological replicates.

Evolutionary Analysis
The genome sequences of 3024 cultivars and 32 wild rice accessions were obtained from the Rice Functional Genomics and Breeding Database (RFGB, http://www.rmbreeding. cn/Snp3k, accessed on 20 April 2022) and OryzaGenome (http://viewer.shigen.info/ oryzagenome/, accessed on 22 April 2022) [40,41]. The geographic information of cultivated rice populations was obtained from the MKBASE (http://www.mbkbase.org/rice/ germplasm, accessed on 28 April 2022) and marked on the map using R software. The PopGenome package in the R software was used to calculate the parameters of genetic divergence for OsHd8 and its flanking regions between indica and japonica subspecies, including haplotype and nucleotide F ST , Nei's G ST , Hudson's G ST and H ST [42]. A phylogenetic tree of OsHd8 was constructed using the UPGMA method with MEGA7.0 [43], and a haplotype network was constructed using the pegas package in the R software [44].

Statistical Analysis
All the assays were performed on three biological replicates. The data were analyzed using the GraphPad Prism 9 software (https://www.graphpad.com/, accessed on 28 January 2022) and the means were compared by Student's t-test, the *, ** and *** mean p < 0.05, 0.01 and 0.001, respectively. The primers used for genetic mapping, vector construction, PCR and qRT-PCR analysis were all listed in Supplementary Table S4.

Genetic Analysis and Mapping of OsHd8 for Heading Date
In the breeding practice, we found that the heading date of the O. longistaminata introgression line 1880 and an early flowering variety JiaHong2B (JH2B), was about 92 days and 65 days ( Figure 1A,B), respectively, showing a great difference. To explore the genetic basis for the heading date in 1880, genetic linkage analysis of 1147 F 2 individuals derived from the cross of 1880/JH2B displayed a continuous distribution with an apparent valley bottom between 62 and 92 days ( Figure 1C). These plants were then used for the short and long heading date pools. The two pools were then subjected to whole-genome sequencing up to >121× coverage, and 835,204 high-quality single-nucleotide polymorphisms (SNPs) were identified. SNP-index analysis showed that there was only one obvious single peak on the short arm 3.3~6.53 Mb of chromosome 8 (Supplementary Figure S1), meaning that the candidate gene controlling heading date is possibly located in this region, and named as Hd8.
A total of 660 plants with extreme phenotypes were then selected from a 4500 F 2 population of 1880/JH2B cross and were used for fine mapping; the OsHd8 successfully narrowed the locus to a 31.8 kb region between the marker SNP7730 and SSR-1 ( Figure 2A). According to the information from the RGAP (Rice Genome Annotation Project), four predicted genes were present in this region, namely, LOC_Os08g07730, LOC_Os08g07740, LOC_Os08g07760, and LOC_Os08g07774 (Figure 2A). qRT-PCR showed that the LOC_Os08g07740, encoding a histone-like transcription factor and archaeal histone, had a large expressional difference between 1880 and JH2B ( Figure 2B-E). And it was reported a flowering suppressor named EF8/LHD1 [45,46], we deduced that the LOC_Os08g07740 was responsible for OsHd8. A total of 660 plants with extreme phenotypes were then selected from a 4500 F2 population of 1880/JH2B cross and were used for fine mapping; the OsHd8 successfully narrowed the locus to a 31.8 kb region between the marker SNP7730 and SSR-1 ( Figure 2A). According to the information from the RGAP (Rice Genome Annotation Project), four predicted genes were present in this region, namely, LOC_Os08g07730, LOC_Os08g07740, LOC_Os08g07760, and LOC_Os08g07774 ( Figure 2A). qRT-PCR showed that the LOC_Os08g07740, encoding a histone-like transcription factor and archaeal histone, had a large expressional difference between 1880 and JH2B ( Figure 2B-E). And it was reported a flowering suppressor named EF8/LHD1 [45,46], we deduced that the LOC_Os08g07740 was responsible for OsHd8.

OsHd8 Encodes a Transcriptional Repressor
To investigate whether LOC_Os08g07740 was responsible for the phenotypic changes, a 4.7 kb genomic fragment, including a 2 kb upstream regulatory sequence, a 1.7 kb gene coding region, and a 1 kb downstream fragment of OsHd8 from 1880, was cloned into the vector pCAMBIA1301, and introduced into JH2B, through agrobacterium-mediated transformation. Compared to JH2B, the heading date of transgenic complementary plants was delayed by about one week and showed an increased expression level ( Figure 3A-C).  To further validate the function of LOC_Os08g07740, an overexpression construct was transformed into JH2B. The OsHd8 transcript level increased by about fivefold, and the heading date was delayed by 15 to 35 days compared with control plants ( Figure  3D,E). These results demonstrated that LOC_Os08g07740 is OsHd8, which is essential for regulating the heading date in rice.
To further investigate the function of OsHd8, protein subcellular localization was performed and found that it was located in the nucleus ( Figure 4A); spatiotemporal analysis To further validate the function of LOC_Os08g07740, an overexpression construct was transformed into JH2B. The OsHd8 transcript level increased by about fivefold, and the heading date was delayed by 15 to 35 days compared with control plants ( Figure 3D,E). These results demonstrated that LOC_Os08g07740 is OsHd8, which is essential for regulating the heading date in rice.
To further investigate the function of OsHd8, protein subcellular localization was performed and found that it was located in the nucleus ( Figure 4A); spatiotemporal analysis showed that OsHd8 was expressed in all tissues, including the roots, young leaves, culms, and panicle ( Figure 4B), corresponding to the A protein-BLAST(BLASTp) at NCBI online revealed that OsHd8 encodes a CBFD_NFYB_HMF domain nuclear transcription factor belonging to the HAP3 subunit [46]. The transcriptional activity assays were then performed in rice protoplasts. The luciferase reporter gene contained five copies of binding sites for GAL4, and the Renilla luciferase gene was used as the internal reference. Compared with the transactivator control constructs GALBD-VP16, GALBD-VP16 fused with OsHd8 induced significantly less LUC activity in rice protoplasts ( Figure 4C,D), indicating that OsHd8 functions as a transcriptional repressor.

Expression Level of OsHd8 Affects Heading Date
To illustrate how OsHd8 regulates the heading date, we compared the genomic sequence of OsHd8 between 1880 and JH2B; no nucleotide difference was detected in the coding regions of OsHd8 (Supplementary Figure S2). Instead, four polymorphisms and

Expression Level of OsHd8 Affects Heading Date
To illustrate how OsHd8 regulates the heading date, we compared the genomic sequence of OsHd8 between 1880 and JH2B; no nucleotide difference was detected in the coding regions of OsHd8 (Supplementary Figure S2). Instead, four polymorphisms and one 8-bp InDel were found in the promoters of the two parent lines ( Figure 5A).
To verify whether the 8-bp InDel or 4 SNPs in the OsHd8 promoter affect the OsHd8 expression, we generated constructs by installing two type promoter fragments into the pGreenII0800-LUC vector and then introduced the constructs into rice protoplasts for transient expression assays. Results showed that 1880 promoter activity was much stronger than that of JH2 ( Figure 5B). Furthermore, to validate the promoter activation capacity between 1880 and JH2B, a Dual-LUC assay was carried out in tobacco leaves and found that the 1880 promoter had a greater LUC/REN value than JH2B ( Figure 5C), meaning that the sequence variations in the promoter of 1880 led to a high expression of OsHd8, and hence the delayed heading date. To verify whether the 8-bp InDel or 4 SNPs in the OsHd8 promoter affect the OsHd8 expression, we generated constructs by installing two type promoter fragments into the pGreenII0800-LUC vector and then introduced the constructs into rice protoplasts for transient expression assays. Results showed that 1880 promoter activity was much stronger than that of JH2 ( Figure 5B). Furthermore, to validate the promoter activation capacity between 1880 and JH2B, a Dual-LUC assay was carried out in tobacco leaves and found that the 1880 promoter had a greater LUC/REN value than JH2B ( Figure 5C), meaning that the sequence variations in the promoter of 1880 led to a high expression of OsHd8, and hence the delayed heading date.

Comparative Transcriptome Analysis of JH2B and OE-OsHd8
To further explore the regulatory network underlying the OsHd8 function, we performed RNA-sequencing analysis with young leaves from JH2B and OE-Hd8. A total of 1198 differentially expressed genes (DEGs) were identified, of which 825 genes were upregulated and 373 genes were downregulated in the OE-OsHd8 plants. Gene Ontology (GO) assay showed that these DEGs were significantly enriched in terms of the metabolic

Comparative Transcriptome Analysis of JH2B and OE-OsHd8
To further explore the regulatory network underlying the OsHd8 function, we performed RNA-sequencing analysis with young leaves from JH2B and OE-Hd8. A total of 1198 differentially expressed genes (DEGs) were identified, of which 825 genes were upregulated and 373 genes were downregulated in the OE-OsHd8 plants. Gene Ontology (GO) assay showed that these DEGs were significantly enriched in terms of the metabolic process, binding and transcription regulator activity ( Figure 6A). Further analysis of transcription factors of DEGs revealed that transcription factors associated with flowering were enriched in the OsHd8 pathway, such as the MADS family, HAP2 family, bZIP family and WRKY family (Supplementary Figure S3) [25,[47][48][49]. In particular, several genes controlling the rice heading date, such as GHD7.1, SDG718, OsGI and HDT1 [2,[50][51][52], were differen-Agronomy 2022, 12, 2260 9 of 16 tially expressed in OE-OsHd8 plants just as confirmed with qRT-PCR analysis ( Figure 6B). These results demonstrated that OsHd8 could be involved in complicated transcriptional regulation processes governing the rice heading date.

Variations in the OsHd8 Promoter Affect its Expression and Heading Date
In order to understand the effects of the promoter variation on the expression of OsHd8, we then selected 101 rice accessions, including 59 indica (IND), 9 temperate japonica (TEJ), 9 tropical japonica (TRJ), 12 aus (AUS), 2 aromatic (ARO) and 10 admix (ADM) lines from different countries for clustering analysis ( Figure 7A). Results showed that OsHd8 diverges into six haplotypes ( Figure 7A), of which the haplotype 2 (Hap2), belonging to JH2B, showed the shortest heading date. 1880 belongs to Hap1, which showed the longest heading date, while Hap3, Hap4 and Hap5 were derived from Hap1 by single-base mutations and had a heading date between the Hap1 and Hap2 ( Figure 7B). Further transcriptional analysis of OsHd8 in rice leaves showed that the Hap2 expression was significantly lower than the other five haplotypes ( Figure 7C), consistent with the phenotype of the haplotypes. These results indicate that the promoter sequence variation seems significantly correlated with the gene expression and heading date of rice.

Variations in the OsHd8 Promoter Affect its Expression and Heading Date
In order to understand the effects of the promoter variation on the expression of OsHd8, we then selected 101 rice accessions, including 59 indica (IND), 9 temperate japonica (TEJ), 9 tropical japonica (TRJ), 12 aus (AUS), 2 aromatic (ARO) and 10 admix (ADM) lines from different countries for clustering analysis ( Figure 7A). Results showed that OsHd8 diverges into six haplotypes ( Figure 7A), of which the haplotype 2 (Hap2), belonging to JH2B, showed the shortest heading date. 1880 belongs to Hap1, which showed the longest heading date, while Hap3, Hap4 and Hap5 were derived from Hap1 by single-base mutations and had a heading date between the Hap1 and Hap2 ( Figure 7B). Further transcriptional analysis of OsHd8 in rice leaves showed that the Hap2 expression was significantly lower than the other five haplotypes (Figure 7C), consistent with the phenotype of the haplotypes. These results indicate that the promoter sequence variation seems significantly correlated with the gene expression and heading date of rice. The asterisks indicate significant differences (*, p < 0.05; ***, p < 0.001; Student's t-test).

OsHd8 is Subjected to Selection in Cultivated Rice
To investigate the genetic relationship of OsHd8 variations, a total of 3024 cultivated rice from the 3k database (http://www.rmbreeding.cn/Index/, accessed on 20 April 2022) and 32 O. rufipogon accessions (http://viewer.shigen.info/oryzagenome/, accessed on 20 April 2022) were selected to analyze the genetic diversity of this gene and its flanking region. The nucleotide diversity value (π) of OsHd8 is lower than its flanking regions in

OsHd8 Is Subjected to Selection in Cultivated Rice
To investigate the genetic relationship of OsHd8 variations, a total of 3024 cultivated rice from the 3k database (http://www.rmbreeding.cn/Index/, accessed on 20 April 2022) and 32 O. rufipogon accessions (http://viewer.shigen.info/oryzagenome/, accessed on 20 April 2022) were selected to analyze the genetic diversity of this gene and its flanking region. The nucleotide diversity value (π) of OsHd8 is lower than its flanking regions in both cultivated and O. rufipogon accessions ( Figure 8A, Supplementary Table S3), suggesting that OsHd8 might be subjected to natural selection.
both cultivated and O. rufipogon accessions ( Figure 8A, Supplementary Table S3), suggesting that OsHd8 might be subjected to natural selection. We further analyzed the parameters of genetic divergence in the OsHd8 locus between indica and japonica subspecies from 3024 cultivated accessions, including the estimates of haplotype and nucleotide FST, Nei's GST, and Hudson's GST and HST. Genetic analysis showed that the five parameters in the OsHd8 locus were all greater than 0.25 between indica and japonica subspecies, and the haplotype and nucleotide FST reached to 0.549 and 0.764, respectively, ( Figure 8B), suggesting strong genetic differentiation between indica and japonica subspecies at the OsHd8 locus [53].
Phylogenetic analysis with the coding sequence indicated that the OsHd8 could be categorized into 86 haplotypes, and that the indica rice was closer to the wild rice haplotypes ( Figure 8C). Meanwhile, 1880 was clustered together with the wild rice, meaning We further analyzed the parameters of genetic divergence in the OsHd8 locus between indica and japonica subspecies from 3024 cultivated accessions, including the estimates of haplotype and nucleotide F ST , Nei's G ST , and Hudson's G ST and H ST . Genetic analysis showed that the five parameters in the OsHd8 locus were all greater than 0.25 between indica and japonica subspecies, and the haplotype and nucleotide F ST reached to 0.549 and 0.764, respectively, ( Figure 8B), suggesting strong genetic differentiation between indica and japonica subspecies at the OsHd8 locus [53].
Phylogenetic analysis with the coding sequence indicated that the OsHd8 could be categorized into 86 haplotypes, and that the indica rice was closer to the wild rice haplotypes ( Figure 8C). Meanwhile, 1880 was clustered together with the wild rice, meaning Hd8 1880 evolved from wild rice O. rufipogon ( Figure 8D). These results suggest that Hd8 1880 originated from wild rice and that at least one mutational event was involved in the origin of Hd8 1880 in indica rice.

Discussion
Flowering is an important trait of plants, and the appropriate flowering time is responsible for the growth and successful sexual reproduction in flowering plants [45]. How to accurately control the flowering time in rice is of great practical significance for the improvement of rice yield. In the present study, we found an elite allele of EF8/LHD1/DTH8/Ghd8 from indica 1880, which delays the heading date by about 27 days, compared to the JH2B (Figure 1) [45,46,54,55]. Previously, the functional DTH8 allele from cv. Asominori delays the heading date by about 13 days compared to CSSL61 (1-bp deletion in the exon that carries the nonfunctional DTH8) [55]. In our study, the newly identified mutations in the promoter reduced the expression level of OsHd8 in JH2B, leading to a shorter heading date ( Figure 3). Furthermore, the DTH8 allele from japonica cv. Asominori could down-regulate the transcription of Ehd1 and Hd3a to regulate the heading date, and another EF8 functional allele delays the heading date by regulating the Hd3a and RFT1. However, our results indicated that EF8 could alter the expression patterns of the other genes, GHD7.1, SDG718, OsGI and HDT1 ( Figure 3). Therefore, we speculate that different alleles of a gene may have different regulatory roles and target genes, resulting in different regulatory mechanisms and phenotypes. Whether differences in gene sequences or genetic backgrounds lead to different allelic effects is an interesting question that deserves further investigation.
The promoter is located in the upstream of the gene coding region and contains many cis-acting elements (CRE), and transcriptional regulation is mainly determined by the promoter CREs [29]. In our study, the promoter activity of 1880 is much higher than that of JH2B ( Figure 4B,C) because of the SNP and InDel variations in the JH2B promoter ( Figure 4A), of which one SNP at -335 bp (G/A) is located in the ABRE cis-acting element in the JH2B promoter (Figure 7). It has been well characterized that ABRE, as an important response element of abscisic acid (ABA), plays an important role in the regulation of the ABA signal network [56]. When encountered with drought stress, plants may accelerate the initiation of the flowering transformation to shorten their growth cycle through the RCN1 mediated ABA signal process [57]. Interestingly, we found that the ABA response element in the OsHd8 promoter of the japonica variety Nipponbare with a short heading date was the same as that of JH2B. Haplotype analysis based on the promoter sequence showed that the JH2B haplotype had the shortest heading date and the lowest expression level (Figure 7). This reminds us that the mutation of the ABA-responsive element of OsHd8 in JH2B may lead to the change in the growth period of rice. However, how ABA response elements regulate the expression of OsHd8 needs to be further validated.
During the evolution of rice, the regional adaptability of cultivated rice is affected by the response to the length of daylight [58]. The heading date is determined by a variety of internal and external signals, including light time, temperature, and hormones [59]. The difference in sensitivity of rice to the photocycle and temperature makes the rice vary greatly in different areas. Many rice lines and wild rice in tropical and subtropical regions, such as O. rufipogon, have strong photoperiod sensitivity to flowering. This strong photoperiod sensitivity completely inhibits heading in long-day conditions, allowing it to induce heading only in short daylight conditions [60]. The Hd8 JH2B allele cloned in this study may be a potential genetic resource. This genotype promotes flowering in long-day conditions and is widely present in the 3K database ( Figure 8). With further analysis of the regional distribution map of Hd8 1880 and Hd8 JH2B haplotypes, we can see that the Hd8 JH2B haplotype distributes mainly in the higher latitude area (Figure 9), meaning that this haplotype rice may be less sensitive to photoperiod and have expanded more greatly than the other haplotype rice.

Conclusions
This study identified and mapped a heading date gene, OsHd8, from an early-flowering rice JiaHong2B (JH2B). Four SNPs and one InDel in the promoter region of OsHd8 led to the advance of the JH2B heading date. Comparative transcriptome analysis revealed OsHd8 to be involved in regulation of some floral regulators including GHD7.1, SDG718, OsGI and HDT1. OsHd8 presents strong genetic differentiation between indica and japonica subspecies and shows artificial selection during the domestication of cultivated rice. Our work will provide a valuable heading date gene for rice breeding programs

Conclusions
This study identified and mapped a heading date gene, OsHd8, from an early-flowering rice JiaHong2B (JH2B). Four SNPs and one InDel in the promoter region of OsHd8 led to the advance of the JH2B heading date. Comparative transcriptome analysis revealed OsHd8 to be involved in regulation of some floral regulators including GHD7.1, SDG718, OsGI and HDT1. OsHd8 presents strong genetic differentiation between indica and japonica subspecies and shows artificial selection during the domestication of cultivated rice. Our work will provide a valuable heading date gene for rice breeding programs    Data Availability Statement: The datasets generated during the current study are available from the corresponding author on reasonable request.