Impact of “Green Revolution” gene Rht-B1b on coleoptile length of wheat

Wheat coleoptile is a sheath-like structure that helps to deliver the first leaf from embryo to the soil surface. Here, a RIL population consisting of 245 lines derived from Zhou 8425B × Chinese Spring cross was genotyped by the high-density Illumina iSelect 90K assay for coleoptile length (CL) QTL mapping. Three QTL for CL were mapped on chromosomes 2BL, 4BS and 4DS. Of them, two major QTL QCL.qau-4BS and QCL.qau-4DS were detected, which could explain 9.1%–22.2% of the phenotypic variances across environments on Rht-B1 and Rht-D1 loci, respectively. Several studies have reported that Rht-B1b may reduce the length of wheat CL but no study has been carried out at molecular level. In order to verify that the Rht-B1 gene is the functional gene for the 4B QTL, an overexpression line Rht-B1b-OE and a CRISPR/SpCas9 line Rht-B1b-KO were studied. The results showed that Rht-B1b overexpression could reduce the CL, while loss-of-function of Rht-B1b would increase the CL relative to that of the null transgenic plants (TNL). To dissect the underlying regulatory mechanism of Rht-B1b on CL, comparative RNA-Seq was conducted between Rht-B1b-OE and TNL. Transcriptome profiles revealed a few key pathways involving the function of Rht-B1b in coleoptile development, including phytohormones, circadian rhythm and starch and sucrose metabolism. Our findings may facilitate wheat breeding for longer coleoptiles to improve seedling early vigor for better penetration through the soil crust in arid regions.


Introduction
Wheat (Triticum aestivum L.) is one of the most important food crops in the world, providing large amounts of starch, rich protein and dietary fiber for humans (Asseng et al., 2020). Maintenance of high and stable wheat yields is crucial for global food security (Boyer, 2004). Drought is an important abiotic stress seriously limiting wheat production (Gupta et al., 2020). Arid and semiarid regions account for about 60% of global crop production, and drought stress caused by frequent extreme weather events often leads to severe reduction of wheat production (Puttamadanayaka et al., 2020). To ensure the emergence rate under drought stress, deeper sowing is often adopted for better utilization of the water in soil . However, a sowing depth beyond the coleoptile length (CL) will result in poor stand establishment, late emergence, and slow early leaf development (BR, 1976;Schillinger et al., 1998). Wheat coleoptiles facilitate the stem and the first leaf to break the ground, and directly determine the maximum sowing depth (Rebetzke et al., 2007a;Rebetzke et al., 2014). However, the short coleoptiles of modern semi-dwarf wheat varieties reduce emergence when sown deep . Understanding the genetic basis for CL will help developing high-yield semi-dwarf varieties with longer coleoptiles and suitable for deep sowing. Previous studies have demonstrated that CL has high heritability and additive effects and is controlled by multiple genes Rebetzke et al., 2007b). Hence, it is feasible to increase the CL through genetic manipulation.
In the 1960s and 1970s, the wide application of dwarf genes Rht1 (Rht-B1b) and Rht2 (Rht-D1b) combined with the increased application of chemical fertilizer greatly promoted the increase of wheat yield, which was called the "Green Revolution" of wheat (Peng et al., 1999;Hedden, 2003). However, compared with the wild type Rht-B1a, the dwarf gene, while improving the resistance to colonization and harvesting index, led to increased nitrogen fertilizer requirement, decreased 1000-grain weight, lower grain protein content, drought tolerance, lower anthers exposure rate and susceptibility to scab (Tang et al., 2009;Lanning et al., 2012;Zhang et al., 2013;He et al., 2016). Genetic analysis has also predicted that Rht-B1b and Rht-D1b loci have certain shortening effects on the CL of wheat, but there has been no further evidence for this speculation (Ellis et al., 2004;Rebetzke et al., 2007b;Yu and Bai, 2010;Li et al., 2011). In contrast, other two widely used Rht genes, Rht8 and Rht24, have been proved to have no negative effect on CL, providing an opportunity to breed semi-dwarfing wheat cultivars with long coleoptiles (Würschum et al., 2017;Chai et al., 2022;Tian et al., 2022;Xiong et al., 2022).
Here, we demonstrated that Rht-B1 is the functional gene underlying a CL QTL on chromosome 4B and its dwarfing allele (Rht-B1b) reduces the CL through multiple pathways such as phytohormones, circadian rhythm and starch and sucrose metabolism. These results provide valuable information for wheat breeding of longer coleoptiles to improve the seedling early vigor and penetration through soil crust in arid regions.

Plant materials and phenotyping
A total of 245 F 2:10 RILs derived from the cross of Zhou 8425B × Chinese Spring were used in this study. Zhou 8425B (Pedigree: Zhou 78A/Annong 7959) and Chinese Spring are an elite facultative wheat line and Chinese landrace, respectively. Zhou 8425B contains two dwarfing alleles Rht-B1b and Rht-D1b and has a short coleoptile length (CL) of about 3.3 cm. Chinese Spring contains two wildtype alleles Rht-B1a and Rht-D1a, which contribute to a long CL of about 4.8 cm. Seeds were sampled from plants grown and harvested at Shijiazhuang of Hebei Province and Qingdao of Shandong Province during the 2020-2021 and 2021-2022 cropping seasons, respectively. Good-quality seeds without any visible damage were selected for all lines. Seeds of all parental and progeny lines were sown in cylindrical pots (100 mm high and 80 mm in diameter) at a sowing depth of 2 cm below the soil surface. The CL was determined from the scutellum to the tip of the coleoptile.

SNP genotyping and QTL analysis
For the Zhou 8425B × Chinese Spring population, the 245 RILs and their parents were genotyped with the 90K iSelect SNP array . Twenty-one linkage groups corresponding to the 21 chromosomes were constructed from 14,955 polymorphic markers. All linkage maps covered 2290.06 cM with marker densities of 7.04 (A), 8.60 (B) and 2.19 (D) markers per cM (Wen et al., 2017). Broad-sense heritability was estimated using IciMapping 4.1 software (https://isbreeding.caas.cn/index.htm). Quantitative trait loci (QTLs) mapping was conducted using IciMapping 4.1 software with inclusive composite interval mapping (ICIM) algorithm (Li et al., 2007). The CL of all lines and the average phenotypic values from the two environments were used for QTL detection. The mapping parameters were chosen as step=1.0 cM and PIN = 0.01. A LOD threshold of 2.5 was chosen for declaration of putative QTLs.

Plant materials for Rht-B1 functional study and RNA-Seq analysis
To study the association between Rht-B1 and the 4B QTL in the current study, an overexpression line and loss of function line of Rht-B1 were created. The complete coding sequence (CDS) of Rht-B1b (GenBank: MG681100.1) was overexpressed in a hexaploid wheat cultivar Fielder (Rht-B1b and Rht-D1a) under driving by maize ubiquitin promoter (All primers were listed in Table S1). CRISPR/SpCas9 was used to create knockout line o f R h t -B 1 b . The sgRNA (PAM-guide sequence 5' -GGAGCCGTTCATGCTGCAG-3') was designed to target conserved regions of Rht-B1b. The resultant construct was transformed into immature embryos by the Agrobacterium tumefaciens (Ishida et al., 2015). Sixty good-quality seeds of each transgenic null lines (TNL), Rht-B1b overexpression lines (Rht-B1b-OE) and Rht-B1b CRISPR/SpCas9 edited lines (Rht-B1b-KO) were evenly sown in ten 10 cm (top diameter) × 8.9 cm (height) plant pots with a sowing depth of 2 cm below the soil surface. Before the coleoptile broke the ground, coleoptile tips and whole coleoptiles of ten TNL and Rht-B1b-OE plants were collected and immediately put into liquid nitrogen for RNA-Seq. Each group included three biological replicates.

RNA-Seq and data analysis
Total RNA of three biological replicates was extracted using the TRIzol ® reagent, and mRNA was purified from total RNA using poly-T oligo-attached magnetic beads. The first strand cDNA was synthesized using random hexamer primer and RNase H. Subsequently, the second strand cDNA synthesis was obtained using DNA Polymerase I and RNase H. Library preparation for RNA-Seq was conducted by Novogene and sequenced on an Illumina Novaseq platform with 1 ug of total RNA (http:// www.novogene.com/). IWGSC RefSeq v2.1 and annotation v2.1 were used for the reference genome and gene model annotation (Zhu et al., 2021). Raw data were processed to obtain clean reads by removal of adapter, ploy-N and low-quality reads. Paired-end clean reads were aligned to the reference genome using Hisat2 (Kim et al., 2019). FeatureCounts was used to count the read numbers mapped to each gene (Liao et al., 2014). Differential expression analysis was performed using the DESeq2 R package (Love et al., 2014). Genes with an adjusted P-value < 0.05 found by DESeq2 were assigned as differentially expressed genes (DEG). GO and KEGG enrichment analysis of DEGs were implemented by the TBtools (Chen et al., 2020). GO terms and KEGG pathways with corrected P-value lower than 0.05 were considered as significantly enriched by the DEGs.

QTL analysis of coleoptile length
The parental lines, Zhou 8425B and Chinese Spring, differed significantly (P < 0.05) for coleoptile length (CL). Based on data averaged across all environments, CL ranged from 2.5 to 6.1 cm with an average of 4.0 cm. CL showed continuous variation in RIL population and had a high heritability of 0.86 ( Figure 1A). Three QTLs for CL were identified on chromosomes 2BL, 4BS and 4DS in the Zhou 8425B × Chinese Spring population (Table 1). Two major QTLs, QCL.qau-4BS and QCL.qau-4DS, were stably detected in all environments, which explained 9.1%-22.2% of the phenotypic variance across environments (Table 1). Based on the genomic position of the flanking markers, we found that QCL.qau-4BS and QCL.qau-4DS spanned the Rht-B1 and Rht-D1 loci, respectively. QCL.qau-2BL explained about 3.0%-3.1% of phenotypic variance, and thus was a minor QTL for CL.

Validation of the effect of Rht-B1b on coleoptile development
Many studies have reported CL QTL on the Rht-B1 locus, indicating that Rht-B1b may reduce wheat CL (Botwright et al., 2001;Li et al., 2017). However, there has been no direct evidence for this speculation. Here, Rht-B1b overexpression and CRISPR/ SpCas9 gene-editing were performed and homozygous plants were generated by self-crossing for CL evaluation. The results demonstrated that Rht-B1b overexpression could reduce the CL about 8.6%, while loss-of-function of Rht-B1b would increase the CL about 17.9% relative to that of null transgenic plants (TNL) (Figures 1B, C). Thus, Rht-B1 can be a target gene of QCL.qau-4BS.

Transcriptome analysis of Rht-B1 on coleoptile development
Although Rht-B1b is known to reduce the CL, the underlying regulatory mechanism remains unclear. To dissect the regulatory mechanism, whole coleoptiles and coleoptile tips of TNL and Rht-B1b-OE were collected for RNA-Seq analysis before the coleoptile breaks the ground. Compared with those of TNL, 142/523 and 191/ 1993 differentially expressed genes (DEG) were upregulated/ downregulated by Rht-B1b in the transcriptome of whole coleoptile and coleoptile tips (Table S2 and S3). There were more down-regulated DEGs than up-regulated DEGs, indicating that Rht-B1b mainly represses the gene expression in coleoptiles. GO enrichment analysis of coleoptile tips revealed that Rht-B1b mainly reduces the CL via the process of "photosynthesis", "oxidationreduction process", "nitrate assimilation carbohydrate metabolic process", and "pigment biosynthetic process". In the whole coleoptile, DEGs were enriched in the GO processes of "oxidation-reduction", "glucan metabolism" and "cellular carbohydrate metabolism" (Table 2). In the coleoptile tips, DEGs were mainly enriched in the GO processes of "photosynthesis", "oxidation-reduction", "nitrate assimilation", "carbohydrate metabolism" and "pigment biosynthetic process" (Table 2).

Adaptation of semi-dwarf modern wheat to drought conditions
Numerous studies have demonstrated a positive association between wheat CL and plant number under deep sowing (Hadjichristodoulou et al., 1977;Matsui et al., 2002). However, the two gibberellin-insensitive dwarfing genes, Rht-B1b and Rht-D1b, tend to cause shorter CL and low seedling emergence rate (Schillinger et al., 1998;Rebetzke et al., 2007b). Here, we generated Rht-B1b over-expressing and CRISPR/SpCas9 editing plants to study its influence on coleoptile development. As a result, overexpression of Rht-B1b reduced the CL, while its loss of function increased the CL. Rht-B1b encodes an N-terminal truncated DELLA protein (lack of DELLA and TVHYNP motifs), which is gibberellin-insensitive protein in wheat (Van De Velde et al., 2021). DELLA proteins encoded by the Rht-B1a gene are the downstream repressors of GA signal transduction and, GA induces the degradation of DELLA proteins via the ubiquitin/proteasome pathway (Itoh et al., 2003). Thus, Rht-B1b led to a reduction of CL compared with tall allele Rht-B1a since the GA-induced seedling growth was repressed (Alabadiı́et al., 2004). So far, there has been no study to validate the effect of Rht-B1b on CL, not to mention the    Putative key downstream genes of Rht-B1b for coleoptile development. TNL, OE, Tip and CL represents transgenic null plant, Rht-B1b-OE plant, coleoptile tip and whole coleoptile, respectively. Orange and blue colors show genes with low and high expression level, respectively.  . SCL3 functions as a positive regulator of GA signaling, which induces the expression of GA biosynthesis genes and autoregulates its own expression via direct interaction with DELLA. In our transcriptome, 16 genes related to GA biosynthesis, three GA receptor GID1 genes and three SCL3 genes were identified as DEGs between Rht-B1b-OE and TNL, indicating that wheat SCL3 and DELLA antagonize each other in maintaining GA homeostasis and GA responses as in Arabidopsis (Figure 2, 3). In addition, DELLAs can physically interact with and block PIF3 and PIF4 activities by sequestering the transcription factors from binding to their targets, which ultimately results in the inhibition of hypocotyl elongation (De lucas et al., 2008;Feng et al., 2008). JAZ could interrupt DELLA-PIF3 interaction, allowing more PIF transcription factors to activate plant growth (Yang et al., 2012). In our transcriptome, several homologs of JAZs were down-regulated by Rht-B1b-OE (Figure 2, 3). HY5 is a key transcription factor for the regulation of seedling photomorphogenesis. COP1 negatively regulates HY5 by directly and specifically interacting with HY5 (Ang et al., 1998). BBX25 and BBX24 additively enhance COP1 and suppress HY5 functions to regulate of seedling deetiolation process in Arabidopsis (Gangappa et al., 2013). RUP1 is induced by CRYs in response to blue light, which is dependent on HY5 (Tissot and Ulm, 2020). These genes are key factors for cryptochrome blue-light signaling and their homologs in wheat were identified as DEGs, indicating that light and GAs might antagonistically regulate coleoptile in wheat (Figure 2, 3).

Breeding for longer coleoptiles with previously reported genetic loci
The response of plants to drought is dependent on multiple factors, including duration and severity of drought conditions, frequency of drought, and the growth stage when subjected to the drought stress (Jatayev et al., 2020). Although wheat can be grown in a variety of harsh environments, rising temperature and unpredictable drought exacerbate the impact of drought stress on wheat yield. If drought stress occurs when sowing, farmers tend to sow more seeds in a deeper depth to increase the seedling establishment rate. Short coleoptiles severely hider the application of deep sowing in wheat production since it influences the emergence rate of wheat seedlings, particularly in fields with thick stubble or crusted soil surface (Rebetzke et al., 2014). Most modern semi-dwarf wheat varieties harboring Rht-B1b or Rht-D1b have short coleoptiles and low yields under drought stress relative to tall plants (Li et al., 2017;Sidhu et al., 2020). Wheat CL is a typical quantitative trait controlled by multiple genes (Rebetzke et al., 2007a). Pyramiding of multiple QTLs in modern semi-dwarf wheat cultivars can efficiently increase the CL. Thus, a comprehensive screening was conducted on the genetic locus for CL by QTL mapping and genome wide association analysis (GWAS) from previous studies and assembled them on wheat chromosomes according to their physical locations (Figure 4). So far, a total of 114 QTLs for CL traits in wheat have been found from 20 studies of CL-related QTL mapping in wheat ( Figure 4 and Table  S4) (Rebetzke et al., 2001;Rebetzke et al., 2007b;Landjeva et al., 2008;Landjeva et al., 2010;Li et al., 2010;Yu and Bai, 2010;Li et al., 2011;Zhang et al., 2013;Nagel et al., 2014;Rebetzke et al., 2014;Zhang et al., 2014;Singh et al., 2015;Elbudony, 2017;Liu et al., 2017;Mo et al., 2018;Zhang et al., 2018;Bovill et al., 2019;Puttamadanayaka et al., 2020;Francki et al., 2021;Ren et al., 2021). About 33 GWAS loci were found to be associated with CL ( Figure 4 and Table S5) (Li et al., 2017;Ma et al., 2020;Sidhu et al., 2020). These genetic loci were used for QTL-rich cluster (QRC) detection, which was defined when markers from at least two independent studies were physically located in 10 Mb range (Cao et al., 2020). The genomic positions of flanking markers were obtained from the IWGSC RefSeq V2.1 (Zhu et al., 2021). A total of 18 QTL-rich clusters (QRC) for CL were found in this study (Table 3). Of them, Rht-B1 and Rht-D1 are strong candidate genes of QRC 4B-I and 4D-I, respectively. These QRC of CL provide valuable gene resources for marker-assisted selection breeding for longer coleoptiles. Breeding for longer coleoptiles with the wild allele Rht-B1a In the 1960s and 1970s, the wide application of semi-dwarf genes Rht-B1b and Rht-D1b combined with the increased application of chemical fertilizer greatly promoted the wheat yield improvement, which was referred to as the "Green Revolution" of wheat (Hedden, 2003). However, semi-dwarf wheat with Rht-B1b/Rht-D1b alleles produce lower grain yield than taller plants with Rht-B1a/Rht-D1a under drought environment (Zhang, J et al., 2013;Jatayev et al., 2020). Compared with wild-type Rht-B1a, semi-dwarf allele Rht-B1b resulted in shorter coleoptile. Seeds of semi-dwarf wheat cultivars are generally sown shallower than taller wheat varieties to ensure the emergence of semi-dwarf seedlings and early vigour (Rebetzke et al.,FIGURE 4 Distribution of genetic loci for wheat coleoptile length (CL) on chromosomes. QTLs and GWAS loci are indicated in black and pink colors, respectively. The black and cyan bars in chromosomes indicate the positions of centromeres and QTL-rich clusters (QRC), respectively.
2007a; Jatayev et al., 2020). Shallow seeding in dry fields reduces emergence for varieties with short coleoptile length (Rebetzke et al., 2001;Jatayev et al., 2020). It is likely that taller wheat cultivars with Rht-B1a/Rht-D1a have higher rate of emergence than semi-dwarf wheat genotypes under early drought environment. Thus, improving lodging resistance of tall wheat has become an important research direction. Besides of reducing the height of plants, an alternative way is to breed wheat varieties with solid-stemmed stems to enhance lodging resistance in wheat (Liang et al., 2022). Wheat with tall plant height Rht-B1a/Rht-D1a allele combined with solid-stemmed stems alleles of TdDof might have high lodging and drought tolerances, long coleoptile and produce higher yield in drought field (Jatayev et al., 2020;Nilsen et al., 2020).

Data availability statement
The data presented in the study are deposited in the Sequence Read Archive (SRA) of the National Center for Biotechnology Information in the BioProject with accession number PRJNA936995.

Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.