Agronomic Trait Analysis and Genetic Mapping of a New Wheat Semidwarf Gene Rht-SN33d

Plant height is a key agronomic trait that is closely to the plant morphology and lodging resistance in wheat. However, at present, the few dwarf genes widely used in wheat breeding have narrowed wheat genetic diversity. In this study, we selected a semi-dwarf wheat mutant dwarf33 that exhibits decreased plant height with little serious negative impact on other agronomic traits. Genetic analysis and mutant gene mapping indicated that dwarf33 contains a new recessive semi-dwarf gene Rht-SN33d, which was mapped into ~1.3 Mb interval on the 3DL chromosome. The gibberellin metabolism-related gene TraesCS3D02G542800, which encodes gibberellin 2-beta-dioxygenase, is considered a potential candidate gene of Rht-SN33d. Rht-SN33d reduced plant height by approximately 22.4% in mutant dwarf33. Further study revealed that shorter stem cell length may be the main factor causing plant height decrease. In addition, the coleoptile length of dwarf33 was just 9.3% shorter than that of wild-type Shaannong33. These results will help to expand our understanding of new mechanisms of wheat height regulation, and obtain new germplasm for wheat improvement.


Introduction
Wheat (Triticum aestivum L.) is a gramineae crop, and the third most widely cultivated crop in the world, after rice and maize. As such, enhancing wheat yield remains a major target of wheat breeding. Wheat cultivation conditions have been continuously improved, with more fertilizer and irrigation facilities used to achieve higher yields. However, excessive application of fertilizers and irrigation cause excessive elongation of the internodes, resulting in lodging [1]. Beginning in the 1960s, the widespread application of semi-dwarfing genes in commercial wheat cultivars greatly improved wheat lodging resistance and yield [2]. Therefore, wheat height modification and the molecular mechanism underlying wheat height regulation are important for wheat improvement.
To date, more than 20 quantitative trait loci (QTLs), and a few cloned genes related to wheat plant height regulation, have been reported, such as Rht1, Rht2 [3], Rht3 [4], Rht4, Rht5, Rht12, Rht13 [5], Rht24 [6], and Rht8 [7,8]. Rht1 and Rht2 were widely introduced into modern commercial wheat varieties during the first 'Green Revolution'. Approximately 24.5% of wheat varieties from autumn-sown regions of China have the gene Rht1, and the frequency of Rht2 is as high as 45.5% [9]. Rht1 and Rht2 genes were originally identified from Norin 10 [10]. Rht1 has been mapped to the short arms of wheat chromosome 4B, and Rht2 is a homoeologous gene of Rht1 that was mapped to the short arms of chromosome 4D [3]. According to the present study, the functions underlying Rht1 and Rht2 in plant height regulation have been well studied. RhtB1a and RhtD1a, wild-type alleles of Rht1 Int. J. Mol. Sci. 2023, 24, 583 2 of 12 and Rht2, encode a protein that contains a DELLA domain at the N-terminus. RhtB1b and RhtD1b, mutant-type alleles of Rht1 and Rht2, contain a mutation which results in a stop codon within the open reading frame, and produces truncated Rht proteins. The wild-type Rht protein represses plant growth, but this repression can be removed by gibberellin (GA)-induced Rht degradation. However, the truncated Rht protein can avoid degradation and exert a continuous inhibitory effect on plant growth [11]. Recent research has revealed new insights into Rht1 and Rht2 [12]. Another widely used dwarf gene, Rht8, encodes an RNase H-like protein carrying a zinc finger BED-type motif, has a high distribution in the Yellow and Huai River Valleys, and has been mapped to chromosome 2D [6,7,13]. Rht3 (the Rht-B1c, allele of the gene Rht1) is a dominant dwarfing gene and results in a reduction of plant height of up to 46% [14,15]. Rht10 is an allele of Rht2 [5]. Other dwarf genes, Rht4, Rht5, Rht9, Rht12, and Rht13, have also been identified and result in varying degrees of dwarfing [5]. Although the above~20 dwarfing genes have been identified, only Rht1, Rht2, Rht8 and Rht24 have been widely introduced into commercial varieties [6,9]; this has resulted in a narrow genetic diversity of wheat cultivars. On the other hand, the negative effects of these genes have been widely introduced into commercial varieties. For example, RhtB1b and RhtD1b result in shortened coleoptile length and reduced seedling vigor [16]. Some results have also revealed that excellent water and fertilizer conditions are necessary for higher production from varieties with the RhtB1b or RhtD1b allele [17]. Therefore, identifying new dwarf wheat germplasm resources is important and necessary for wheat improvement.
In this study, a dwarf wheat mutant, dwarf33, was identified in Shannonong33 (SN33), a commercial wheat variety, using a mutant library induced by the chemical mutagen NaN 3 . dwarf33 exhibited some good agronomic traits. Interestingly, it had no significant differences in grain length and width compared to SN33, and there was no serious negative impact on coleoptile length. Genetic analysis revealed that dwarf33 contains a recessive dwarf gene, named Rht-SN33d, which is limited to a~1.3 Mb region at the end of the wheat 3DL chromosome. Gene prediction indicated that the GA metabolic gene TraesCS3D02G542800, encoding gibberellin 2-beta-dioxygenase, is the key candidate gene. Overall, the identification of new wheat dwarf germplasm resources and genes not only expand genetic diversity but also provide new germplasm for wheat breeding. This study also provides a potential target gene for improving wheat varieties and production using molecular biological techniques.

dwarf33 Shows Decreased Plant Height
The wheat mutant dwarf33, identified from an NaN 3 -mutagenized library in an SN33 background, exhibited shorter plant height compared to wild-type SN33 from the jointing stage ( Figure 1A-C). At the grain filling stage, dwarf33 was 63.3 cm tall, which was 22.4% shorter than SN33, while there was not significant difference in the number of internodes and spike length between them ( Figure 1D-F). Compared to SN33, internode lengths of dwarf33 were shorter by varying degrees, especially the second and fifth internodes, which decreased by 6.8 cm (30.3%) and 2.2 cm (40.9%), respectively ( Figure 1F). As shown in Figure 2, there was no significant difference in spike length between dwarf33 and SN33 ( Figures 1F and 2A). The grain length and width of dwarf33 was close to that of SN33 (Table S3). The grain number per spike and TKW were lower than in SN33, but the spike number per plant was greater than in SN33 ( Figure 2D-F). In addition, the results indicate that dwarf33 was more sensitive to irrigation than SN33. The plant height of dwarf33 decreased by 17.2 cm (27.2%) in a drought environment compared to that in an irrigated environment. However, the SN33 decreased by only 14.9 cm (18.3%) ( Table 1).    irrigated environment. However, the SN33 decreased by only 14.9 cm (18.3%) ( Table  1).

Cytological Observation of Internode Cells
As shown in Figure 1D,F, the second internode under the spike exhibited the largest difference in length between dwarf33 and SN33. Therefore, the cell length of the second internode was measured to reveal the differences in cell characteristics between dwarf33 and SN33. The results indicate that the cell length of the mutant dwarf33 was significantly less than that of SN33. The cell length decrease in dwarf33 reached approximately 15% ( Figure 3C).

Cytological Observation of Internode Cells
As shown in Figure 1D,F, the second internode under the spike exhibited the largest difference in length between dwarf33 and SN33. Therefore, the cell length of the second internode was measured to reveal the differences in cell characteristics between dwarf33 and SN33. The results indicate that the cell length of the mutant dwarf33 was significantly less than that of SN33. The cell length decrease in dwarf33 reached approximately 15% ( Figure 3C).

Coleoptile Length Measurement
Coleoptile length is an important trait because it is closely associated with seedling vigor. Therefore, the coleoptile length of dwarf33 and SN33 was measured under dark conditions. As shown in Figure 4A,C, the average length of coleoptiles in dwarf33 was 4.5 cm, whereas the coleoptile length of SN33 was 4.9 cm. The coleoptile length of dwarf33 decreased by 9.3% compared to SN33.

Coleoptile Length Measurement
Coleoptile length is an important trait because it is closely associated with seedling vigor. Therefore, the coleoptile length of dwarf33 and SN33 was measured under dark conditions. As shown in Figure 4A,C, the average length of coleoptiles in dwarf33 was 4.5 cm, whereas the coleoptile length of SN33 was 4.9 cm. The coleoptile length of dwarf33 decreased by 9.3% compared to SN33.

Dwarf Phenotype of dwarf33 Is Controlled by a Recessive Gene
For the genetic assay, the F 2 population of a cross between dwarf33/SN33 and SN33/ dwarf33 were generated and analyzed. The phenotype of all the F 1 plants of the reciprocal cross between dwarf33 and SN33 resembled the wild-type SN33. This result revealed that the dwarf phenotype of dwarf33 was recessive compared to the tall-plant phenotype. In the F 2 population of dwarf33/SN33, the number of dwarf plants, similar to that of dwarf33, was 125 plants. The number of tall plants, similar to SN33, was 435 plants. The segregation ratio of the dwarf plants and tall plants in the dwarf33/SN33 population was 1:3. The dwarf plants and tall plants in population SN33/dwarf33 displayed a 3:1 segregation as well (Table 2). Overall, the genetic analysis suggested that the dwarfism trait in dwarf33 was controlled by a pair of recessive genes, which was named Rht-SN33d.

Dwarf Phenotype of dwarf33 Is Controlled by a Recessive Gene
For the genetic assay, the F2 population of a cross between dwarf33/SN33 and SN33/dwarf33 were generated and analyzed. The phenotype of all the F1 plants of the reciprocal cross between dwarf33 and SN33 resembled the wild-type SN33. This result revealed that the dwarf phenotype of dwarf33 was recessive compared to the tall-plant phenotype. In the F2 population of dwarf33/SN33, the number of dwarf plants, similar to that of dwarf33, was 125 plants. The number of tall plants, similar to SN33, was 435 plants. The segregation ratio of the dwarf plants and tall plants in the dwarf33/SN33 population was 1:3. The dwarf plants and tall plants in population SN33/dwarf33 displayed a 3:1 segregation as well (Table 2). Overall, the genetic analysis suggested that the dwarfism trait in dwarf33 was controlled by a pair of recessive genes, which was named Rht-SN33d.

Dwarf Gene Mapping Based on BSA and the Wheat SNP Chip
Rht-SN33d was mapped based on the F2 population derived from the dwarf33/CS cross. In the dwarf33/CS cross, the plant height of F1 was 89 cm on average, which was significantly higher than dwarf33 and lower than CS. Plants with a height lower than SN33 were defined as dwarf plants, and those with a height taller than SN33 were defined as wild-type plants ( Figure 5). A total of 19,416 homozygous SNPs were identified between the parents dwarf33 and CS. Only 287 of the 19416 SNPs were identified between dp and hp, and 207 of the 287 SNPs between dp and hp were located at 609.9-615.2 Mb of wheat chromosome 3D ( Figure 6A,B). KASP markers were designed to create a genetic map, as shown in Figure 6C. A QTL with a LOD value of 34 was detected. Because of the low

Dwarf Gene Mapping Based on BSA and the Wheat SNP Chip
Rht-SN33d was mapped based on the F 2 population derived from the dwarf33/CS cross. In the dwarf33/CS cross, the plant height of F 1 was 89 cm on average, which was significantly higher than dwarf33 and lower than CS. Plants with a height lower than SN33 were defined as dwarf plants, and those with a height taller than SN33 were defined as wild-type plants ( Figure 5). A total of 19,416 homozygous SNPs were identified between the parents dwarf33 and CS. Only 287 of the 19,416 SNPs were identified between dp and hp, and 207 of the 287 SNPs between dp and hp were located at 609.9-615.2 Mb of wheat chromosome 3D (Figure 6A,B). KASP markers were designed to create a genetic map, as shown in Figure 6C. A QTL with a LOD value of 34 was detected. Because of the low density of SNPs from the wheat 55K SNP chip, few SNPs were identified and successfully converted to KASP markers, especially the near the end of the chromosome of 3D. Hence, wheat exome capture sequencing was used to identify SNPs between dwarf33 and CS to design more KASP markers. Only the plants lower than dwarf33 could be determined to contain an Rht-SN33d homozygous genotype in the F 2 population. For fine mapping of the gene Rht-SN33d, only the plants lower than dwarf33 were used to detect the genetic recombination. Finally, the mutant gene was limited to an~1.3 Mb interval (612.3-613.6 Mb, Tables S1 and S2) at the end of the wheat 3D chromosome.
converted to KASP markers, especially the near the end of the chromosome of 3D. Hence, wheat exome capture sequencing was used to identify SNPs between dwarf33 and CS to design more KASP markers. Only the plants lower than dwarf33 could be determined to contain an Rht-SN33d homozygous genotype in the F2 population. For fine mapping of the gene Rht-SN33d, only the plants lower than dwarf33 were used to detect the genetic recombination. Finally, the mutant gene was limited to an ~1.3 Mb interval (612.3-613.6 Mb, Table S1, 2) at the end of the wheat 3D chromosome.  density of SNPs from the wheat 55K SNP chip, few SNPs were identified and successfully converted to KASP markers, especially the near the end of the chromosome of 3D. Hence, wheat exome capture sequencing was used to identify SNPs between dwarf33 and CS to design more KASP markers. Only the plants lower than dwarf33 could be determined to contain an Rht-SN33d homozygous genotype in the F2 population. For fine mapping of the gene Rht-SN33d, only the plants lower than dwarf33 were used to detect the genetic recombination. Finally, the mutant gene was limited to an ~1.3 Mb interval (612.3-613.6 Mb, Table S1, 2) at the end of the wheat 3D chromosome.  In addition, the reported dwarf gene RhtD1b was detected in SN33, which is widely used in commercial varieties. To compare the dwarfing effect between Rht-SN33d and RhtD1b, the KASP marker "KASP-Rht2" was used to identify the genotype of the dwarf33/CS F 2 population [18]. As shown in Table 3

Candidate Gene Prediction
The region that contained the candidate gene was limited to a~1.3 Mb interval (612.3-613.6 Mb) on the 3D chromosome. Twenty-two high-confidence genes were predicted in the candidate region based on the Chinese Spring wheat genome (Table 4). Interestingly, a GA biosynthesis related gene TraesCS3D02G542800 was found that encoded gibberellin 2-beta-dioxygenase. GA and related metabolic intermediates are widely involved in plant height regulation. Therefore, TraesCS3D02G542800 was predicted as the candidate gene. Thereafter, we cloned the CDS of TraesCS3D02G542800 from dwarf33 and SN33. The result indicated that there was no difference in the CDS of TraesCS3D02G542800 between dwarf33 and SN33 (detail sequence was listed in Supplementary materials).

Expression Pattern of Rht-SN33d
There were another two homoeologous genes of Rht-SN33d on chromosomes 3A and 3B that exhibited highly conserved DNA sequences. A sequence-specific primer for qPCR could not be designed to distinguish their expression. Therefore, the expression result from the primer pair Rht-SN33-F/R represented the total expression level of Rht-SN33 on chromosomes 3A, 3B, and 3D. The results indicate that the expression pattern of Rht-SN33 was tissue-specific ( Figure 7 and Figure S1), and the stem and leaf exhibited higher expression levels than the root, spike, and grain. Furthermore, the expression level in dwarf33 was significantly higher than that in SN33 for the roots, stems and leaves. chr3D 613631318 613633402 TraesCS3D02G543200 SKP1-like protein

Expression Pattern of Rht-SN33d
There were another two homoeologous genes of Rht-SN33d on chromosomes 3A and 3B that exhibited highly conserved DNA sequences. A sequence-specific primer for qPCR could not be designed to distinguish their expression. Therefore, the expression result from the primer pair Rht-SN33-F/R represented the total expression level of Rht-SN33 on chromosomes 3A, 3B, and 3D. The results indicate that the expression pattern of Rht-SN33 was tissue-specific ( Figure 7 and Figure S1), and the stem and leaf exhibited higher expression levels than the root, spike, and grain. Furthermore, the expression level in dwarf33 was significantly higher than that in SN33 for the roots, stems and leaves. Figure 7. The expression pattern of the Rht-SN33d gene in SN33 and dwarf33. Stem 1: the stem was collected at the jointing stage; Stem 2: the stem was collected at heading stage. "ns" indicates no significant differences; Asterisks "**" indicate significant differences at p ≤ 0.01.

Discussion
Dwarfism genes can reduce plant height, enhance lodging resistance, and increase tolerance to higher water and fertilizer which have been widely used to improve wheat agronomic traits and increase wheat yield. The first "Green Revolution" introduced dwarf genes into cultivars, which improved lodging resistance and yield [2].
At present, Rht1 and Rht2 [3] and Rht8 and Rht24 are widely applied in commercial varieties. It should be noted that the few dwarfism genes that are widely used could easily lead to the loss of genetic diversity and introduce adverse effects of dwarf genes into commercial varieties [19]. Better water and fertilizer conditions are indispensable condition for obtaining higher wheat yields [17]. Previous studies indicated that Rht-B1b and Rht-D1b could reduce plant height by ~23% [20]. Another dwarf gene, Rht8, effectively reduced plant height by approximately 20% [7]. In this study, the plant height of mutant dwarf33 decreased by ~22.4% compared to wild-type SN33 ( Figure 1E). Therefore, the dwarf gene in dwarf33 showed a similar dwarfing effect on plant height to that of Rht1 and Rht2, but slightly larger than that of Rht8. Interestingly, although Rht-SN33d has a similar dwarfing ability to Rht2, it only reduced coleoptile length by 9.3% ( Figure 4C). Rht1 and Figure 7. The expression pattern of the Rht-SN33d gene in SN33 and dwarf33. Stem 1: the stem was collected at the jointing stage; Stem 2: the stem was collected at heading stage. "ns" indicates no significant differences; Asterisks "**" indicate significant differences at p ≤ 0.01.

Discussion
Dwarfism genes can reduce plant height, enhance lodging resistance, and increase tolerance to higher water and fertilizer which have been widely used to improve wheat agronomic traits and increase wheat yield. The first "Green Revolution" introduced dwarf genes into cultivars, which improved lodging resistance and yield [2].
At present, Rht1 and Rht2 [3] and Rht8 and Rht24 are widely applied in commercial varieties. It should be noted that the few dwarfism genes that are widely used could easily lead to the loss of genetic diversity and introduce adverse effects of dwarf genes into commercial varieties [19]. Better water and fertilizer conditions are indispensable condition for obtaining higher wheat yields [17]. Previous studies indicated that Rht-B1b and Rht-D1b could reduce plant height by~23% [20]. Another dwarf gene, Rht8, effectively reduced plant height by approximately 20% [7]. In this study, the plant height of mutant dwarf33 decreased by~22.4% compared to wild-type SN33 ( Figure 1E). Therefore, the dwarf gene in dwarf33 showed a similar dwarfing effect on plant height to that of Rht1 and Rht2, but slightly larger than that of Rht8. Interestingly, although Rht-SN33d has a similar dwarfing ability to Rht2, it only reduced coleoptile length by 9.3% ( Figure 4C). Rht1 and Rht2 not only reduced the plant height but also shortened the coleoptile by approximately 20-30% [21,22]. A previous study indicated that some dwarf genes not only generate short plant height but also decrease the grain number per spike [23]. Spike length showed no significant difference between dwarf33 and SN33 ( Figure 1F). Although, the grain number per spike of dwarf33 was less than SN33, the spike number per plant was more than SN33 ( Figure 2D). These results suggested that dwarf33 contained a dwarf gene that had a smaller negative effect on coleoptile length and spike development. dwarf33 is a potential resource for seedling vigor improvement in wheat dwarfing breeding.
The plant hormone GA is important in plant height regulation. In GA-mediated plant height regulation, it can form a complex with GID1 and DELLA, while SCF can degrade DELLA by this complex through ubiquitination protein degradation, thereby relieving growth inhibition by DELLA. However, mutant DELLA cannot be degraded, therefore inhibiting growth [11,12]. In this study, the candidate gene was in the~1.3 Mb region on the long arm of chromosome 3D. Gene prediction showed that this region contains a gene TraesCS3D02G542800 which encodes Gibberellin 2-beta-dioxygenase. Previous studies indicated that Gibberellin 2-beta-dioxygenases can catalyze the deactivation of bioactive GAs or their precursors [24]. Therefore, changing the GA content by modifying GA2ox genes is a potential tool for plant type improvement. For example, overexpression of GA2oxs could decrease liliaceous monocotyledon Tricyrtis sp. plant height [25]. The dwarfing phenotype was observed in the GA2ox genes PvGA2ox5 and PvGA2ox9 overexpression switchgrass lines, which were considered for plant architecture improvement [26]. In wheat, Rht24 encodes GA2ox-A9 which is widely used for plant modification [6]. These results suggest that TraesCS3D02G542800 may be the candidate gene of Rht-SN33d.
As mentioned above, Rht-SN33d was mapped at the end of the 3D chromosome. Previous studies revealed that dwarfing genes are distributed in multiple wheat chromosomes. Rht1 is located in 4B, Rht2 in 4B [3], Rht4 in 2BL, Rht5 in 3BS, Rht12 in 5AL, Rht13 in 2DS [5], Rht24 in 6A [6], and Rht8 in 2DS [7,8] and so on. In our study, Rht-SN33d was located in the end of 3DL and is indicated as a new dwarf gene. However, there was no difference in the CDS of TraesCS3D02G542800 between dwarf33 and SN33. Previous studies have shown that structural variation may alter local chromatin states and, in turn, affect gene expression. For example, the deletion of a large fragment of approximately 304 kb near the TaMLO-B1 site changed chromatin states and upregulated the expression level of the upstream gene TaTMT3 [27]. Interestingly, the expression level of Rht-SN33 in dwarf33 was significantly higher than that in SN33 in the stem at the jointing stage and heading stage ( Figure 7). As noted before, 2-beta-dioxygenases can inactivate GAs [24]. This suggests that sequence variation or methylation modification may have occurred in the promoter, and another possibility is that the expression pattern of TraesCS3D02G542800 was affected by the mutation of nearby sites. Further experiments should verify these predictions.
In conclusion, the dwarf wheat dwarf33 was identified in the SN33 mutant library. The Rht-SN33d gene was mapped in 3DL, the results indicating that this is a new locus involved in plant height regulation. Rht-SN33d has a similar dwarfing effect on plant height reduction to Rht1 and Rht2 but has a weaker negative effect on coleoptile elongation. Furthermore, TraesCS3D02G542800 was predicted as the possible candidate gene encoding a gibberellin 2-beta-dioxygenase. Although the molecular mechanism of Rht-SN33d in dwarf33 remains to be further studied, it is expected to provide new material for high yield and lodging resistance breeding.

Plant Material
The dwarf mutant dwarf33 was identified from the Shaannong33 (SN33) mutant library induced by 10 mM NaN 3 (80115560, sinopharm, Shanghai, China) following the method of [28]. Wild-type SN33 is a superior commercial wheat variety that is widely cultivated in Shaanxi province and partly in the Yellow and Huai River Valleys of China. An F 2 population with 1030 plants derived from dwarf33/CS (Chinese Spring wheat) was used to map the mutant gene, which was named Rht-SN33d. Parents and F 2 individuals were planted in an experimental field in Northwest A&F University, Yangling, Shaanxi province in the year 2019 to 2020. The rows were 2 m long and 0.25 m apart, with 20 seeds per row. Normal irrigation involved irrigating once before and once after winter. Drought was based on irrigating by natural rainfall.
Leaf tissue was collected and stored at −20 • C for genomic DNA extraction. At the jointing stage (marked Stem 1) and heading stage (marked Stem 2), the second internode below the spike was collected and stored at −80 • C for RNA extraction, and some internodes were stored in FAA (formalin acetic acid ethanol) solution at 4 • C for paraffin embedding.

Phenotypic Assessment
At the filling stage, plant height is stable. The plant height of parents and F 2 individuals was measured (2020 Yangling). Plant height was considered to be the distance from the ground to the top but did not include the length of the awn. Seedling height was defined as the distance between the seed to the top of the seedling. The data on plant height, internode length, and seed characteristics (such as seed length, width, and TKW), were collected from six independent plants (2019 Yangling).
For coleoptile length measurements, the seed was germinated in water, then germinating seeds were planted in 7 × 7 × 7.3 cm (length × width × height) flowerpots, and cultivated under dark conditions at 25 • C. After 14 days l, the length between the seed and the top of the coleoptile was measured.

DNA and RNA Isolation and RT-PCR and RT-qPCR
DNA was isolated from wheat leaves following the CTAB method [29]. Total RNA was extracted using RNAiso Plus reagent (Takara Biomedical Technology, Beijing, China) according to the manufacturer's instructions. Reverse transcription was performed using EasyScript ® One-Step gDNA Removal and cDNA Synthesis SuperMix (Transgen, Beijing, China). RT-qPCR was performed using TransStart ® Green qPCR SuperMix (Transgen, Beijing, China). RT-qPCR was performed using primers for Rht-SN33, where the TaGAPDH gene was used as an internal control (Table S1). Amplification was performed with one cycle of 2 • min at 94 • C and 40 cycles of 10 • s at 94 • C, 30 • s at 60 • C and 1 • min at 72 • C using a CFX96 Real Time PCR System (Bio-Rad, Hercules, CA, USA). Kompetitive allelespecific PCR (KASP) assays were performed following standard KASP guidelines (LGC Genomics, Middlesex, UK). All PCRs for cloning or sequencing were performed using a 2× Phanta ® Flash Master Mix (Vazyme, Nanjing, China). The data of gene expression were collected from three independent plants, the data from each plant representing three technical repetitions.

Microscopic Observation of Stem Tissue
Paraffin embedding of samples was performed using the method described by Weigel and Glazebrook [30]. Observation was performed on a Nikon Eclipse E100 microscope (Nikon Co., Ltd., Tokyo, Japan). Cell size was measured from three independent plants of dwarf33 and SN33.

Marker Design and Gene Mapping
First, a dwarf DNA pool (dp) was created from 50 dwarf plants of the F 2 population of dwarf33/CS in which the plant height was lower than that of dwarf33. A high-DNA pool (hp) was created from 50 dwarf plants from the same population in which the plant height was taller than dwarf33. Subsequently, single nucleotide polymorphisms (SNPs) were identified using a wheat 55 K SNP array from CapitalBio Technology (Beijing, China). The protocol for KASP marker design was based on the method developed in a previous report (Ramirez-Gonzalez et al. 2015). Then, QTLs and LOD values were detected by QTL IciMapping V4.1 [31] based on 1030 F 2 plants derived from the cross between dwarf33 and CS. Because of the low SNP density of the wheat 55K SNP chip, to identify more SNPs between dwarf33 and CS, Exome capture sequencing was used to call the SNPs based on the sequence of Chinese Spring 1.1 [32,33]. Lastly, only the plants which were shorter than dwarf33 were used to fine mapping the gene Rht-SN33d.

Gene Prediction and Cloning
Based on gene mapping, the candidate gene was limited to a certain chromosome region. The high-confidence genes were predicted based on the genome annotation of Chinese Spring v1.0 (WheatOmics 1.0, http://202.194.139.32/jbrowse.html, accessed on 5 May 2019) [34]. The primer pair Rht-SN33d-F1/R1 was used to isolate the CDS sequence of the candidate gene using the 2× Phanta Flash Master Mix (Vazyme, Beijing, China). The reaction system involved: step 1: 98 • C, 90 s; step 2: 98 • C, 10 s; step 3: 60 • C, 15 s; step 4: 72 • C, 15 s; 32 cycles from step 2 to step 4 and then maintenance at 12 • C. Thereafter, the PCR product was cloned into the vector using a Universal Ligation Mix kit (Vazyme, Beijing, China) following the manual. The DH5α spot selected on an LB media (50 mg·L −1 ampicillin) plate was used to identify the sequence of the candidate gene.

Statistical Analyses
The results are shown as mean ± standard deviation (SD) from at least three replicates (the details of replicates are listed in the relative section). A two-tailed Student's t-test was used to analyze the significance between samples by using SPSS. p-value < 0.05 (*) and p-value < 0.01 (**) are regarded as significant.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.