Identification and Utilization of the Soybean Plant Height Locus PH19-31

Abstract

Plant height is a key agronomic trait in soybeans, directly influencing yield potential and agronomic performance. In this study, the tall-stem cultivar Heinong 63 and the dwarf-stem cultivar Longken 3077, along with their F₂ populations, were used as experimental materials. A preliminary mapping using the bulk segregant analysis sequencing (BSA-Seq) approach enabled us to locate the target locus within a 1.8 Mb interval, which was subsequently narrowed to a 48.3 kb region using simple sequence repeat (SSR) and insertion and deletion (InDel) markers. The interval was designated as PH19-31 and contained nine annotated genes. One InDel marker, PL7, was developed within this fine-mapped region. Based on an expression analysis of the nine candidate genes, Glyma.19G230000 and Glyma.19G230200 were identified as potential regulators of plant height in soybeans. This study provides valuable materials and technical support for the cloning of plant height-related genes and the molecular breeding of soybeans.

1. Introduction

Soybean (Glycine max L.) is a major global crop and an important source of plant-based protein and vegetable oil. As demand for plant-derived protein continues to grow, improving soybean yield and performance has become a key breeding objective. Among the traits affecting yield, plant height plays a particularly critical role, as it determines the plant architecture, lodging resistance, and suitability for mechanized harvesting. An optimal height enhances light capture and environmental adaptability [1,2]. Studies suggest that soybean yield increases with plant height up to an optimal range of 70–90 cm, while excessive elongation during late growth stages may reduce productivity [3]. Recent research has shown that optimizing plant height through the PH13 gene enhances yield under dense planting conditions, emphasizing the role of height regulation in improving soybean adaptability and productivity [4]. Therefore, plant height plays a crucial role in soybean cultivation and genetic improvement and is considered a fundamental target in modern breeding programs.

Significant progress has been made in dissecting the genetic basis of plant height in soybeans. Through the use of recombinant inbred line (RIL) populations and molecular markers, multiple quantitative trait loci (QTL) associated with plant height have been identified [5,6,7,8], reflecting the complex and polygenic nature of this trait. Recent advances have incorporated bulked segregant analysis sequencing (BSA-Seq) to improve mapping resolution and efficiency. BSA-Seq combined with SNP-index and Euclidean distance (ED) algorithms has been successfully applied to identify candidate genes in various species, including complex genomes such as hexaploid chrysanthemum, and has accelerated the identification of causal genes such as OsRR22, a salt-tolerant gene in rice, which enabled the rapid development of a new cultivar [9,10]. Studies using F₂ populations and BSA-Seq have successfully mapped key plant-height-related QTLs and explored their relationship with other traits such as lodging resistance [11,12]. These findings demonstrate the effectiveness of BSA-Seq as a powerful tool for identifying trait-associated genomic regions and facilitating molecular breeding.

In this study, we aimed to identify novel loci associated with soybean plant height using a BSA-Seq strategy. A significant QTL was detected on chromosome 19 in an F₂ population derived from a cross between a tall-stem and a dwarf-stem cultivar. This locus was fine-mapped to a 48.3 kb interval and was designated as PH19-31. Given its strong association with plant height variation and its potential applicability in marker-assisted breeding, PH19-31 was selected as the focal target for candidate gene discovery and functional analysis.

To further characterize this region, we used an integrated strategy combining BSA-Seq with the single nucleotide polymorphism (SNP)/insertion and deletion (InDel)-index, Euclidean distance (ED) algorithms, and SSR-based fine mapping to identify candidate loci in an F₂ population. Simple sequence repeat (SSR) markers, known for their high polymorphism and co-dominant inheritance, were further used for fine mapping and candidate gene analysis. They remain widely used in soybean QTL analysis and marker-assisted selection due to their reliability and cost-effectiveness [13,14]. The results provide theoretical and practical support for plant architecture improvement, offer valuable materials for breeding high-yield and lodging-resistant soybean varieties, and highlight the potential of BSA-assisted fine mapping in trait dissection and molecular breeding.

2. Materials and Methods

2.1. Plant Material

The tall-stem cultivar Heinong 63 and the dwarf-stem cultivar Longken 3077 were crossed to generate the F₁ generation. The F₁ plants were grown in Harbin, Heilongjiang Province, China (approximately 45.75° N, 126.63° E) in 2022 and harvested in October of the same year. Subsequently, the F₁ plants were self-pollinated to generate the F₂ population. In 2023, the F₂ population was cultivated again in Harbin for phenotypic evaluation. The experimental population was cultivated in the field under uniform agronomic conditions using a single-row plot layout. Each row was 2 m in length, with a plant spacing of 10 cm. Standard field practices, including thinning, weeding, and irrigation, were applied consistently. Plant height was measured at the maturity stage and defined as the distance from the cotyledonary node to the top of the main stem. Both parental lines were kindly provided by the Beidahuang Kenfeng Seed Industry Co., Ltd. (Harbin, China).

2.2. Phenotypic Identification

Kurtosis and skewness were used to describe the distribution characteristics of the plant height phenotypes in the parental lines and the F₂ population. Kurtosis reflects the sharpness of the distribution peak or the heaviness of the tails. The specific calculation formula is as follows:

$$\mathrm{K}\mathrm{u}\mathrm{r}\mathrm{t}\mathrm{o}\mathrm{s}\mathrm{i}\mathrm{s}={\displaystyle \frac{\frac{1}{n}{\sum }_{i=1}^n{(x_i-\overline{x})}^4}{{\left(\frac{1}{n}{\sum }_{i=1}^n{\left(x_i-\overline{x}\right)}^2\right)}^2}}-3$$

where xᵢ is the individual data point, $\overline{x }$ is the mean of the dataset, and n is the number of data points.

Skewness reflects the asymmetry of the data distribution or the direction of deviation from the mean. The specific calculation formula is as follows:

$$\mathrm{S}\mathrm{k}\mathrm{e}\mathrm{w}\mathrm{n}\mathrm{e}\mathrm{s}\mathrm{s}={\displaystyle \frac{\frac{1}{n}{\sum }_{i=1}^n{(x_i-\overline{x})}^3}{{\left(\frac{1}{n}{\sum }_{i=1}^n{\left(x_i-\overline{x}\right)}^2\right)}^{3/2}}}$$

where xᵢ is the individual data point, $\overline{x }$ is the mean of the dataset, and n is the number of data points.

2.3. DNA Extraction

Genomic DNA was extracted using a modified CTAB method. The detailed protocol is provided in Supplementary Table S1. The DNA quality was assessed using two methods: (1) 1% agarose gel electrophoresis to check integrity; and (2) a NanoDrop spectrophotometer to evaluate purity. The DNA met the Illumina sequencing library preparation standards, with A260/A280 = 1.82 ± 0.03 and A260/A230 > 2.0.

2.4. Bulked Segregant Sequencing (BSA-Seq)

Based on the plant height phenotypes of the two parental lines (Heinong 63 and Longken 3077) and the F₂ population, two parental pools and two extreme phenotype pools were constructed. Each parental pool consisted of leaf samples from 10 individuals of Heinong 63 and 10 individuals of Longken 3077. In the F₂ population, 20 extremely tall and 20 extremely short individuals were selected to form two extreme phenotype pools.

To minimize environmental variation and eliminate border effects, phenotypic measurements and DNA sampling were conducted exclusively on plants located in the central rows of the field. All border rows were excluded from phenotypic evaluation. Individuals exhibiting extreme phenotypes were defined as those within the upper and lower 5% quantiles of plant height values in the F₂ population.

The sampling procedure was as follows: during the full flowering stage, young leaves from the apex of each plant were collected, and genomic DNA was extracted using the CTAB method. The DNA concentration was measured, and equal amounts of DNA from each individual were mixed to construct the four DNA pools. All pooled samples were submitted to Biomarker Technologies (Beijing, China) for high-throughput sequencing.

Base calling was performed using the Illumina CASAVA 1.8 platform, and sequencing was conducted using a 150 bp paired-end strategy to ensure high data quality. The parental pools were sequenced at a depth of 10×, while the F₂ phenotype pools were sequenced at a depth of 30× to ensure sufficient genome coverage and data reliability.

2.5. Sequencing Data Processing

During data processing, the raw sequencing reads were subjected to strict quality control and filtering to obtain high-quality, clean reads. Clean reads were aligned to the soybean reference genome (Wm82.a2.v1) using a BWA(v0.7.17) software package [15]. The alignment results were further analyzed using Picard(v2.27.1) [16] to evaluate the insert size distribution and assess the accuracy and completeness of the sequencing data.

Variant calling was performed using GATK(v4.2.0.0) [17] to detect SNPs, and SnpEff(v5.0) [18] was used for the annotation of small InDels. Low-confidence and low-quality SNP and InDel sites were filtered out prior to the downstream analysis.

An association analysis was carried out using the Euclidean distance (ED) algorithm [19] and the SNP-index (or InDel-index) algorithm [20]. To reduce background noise, the raw ED values were squared [19], and a DISTANCE-based fitting method was applied. A significance threshold was determined based on the median plus three times the standard deviation (median + 3×SD) of the fitted ED values [17].

2.6. SSR Marker Screening and Genotyping

Based on the newly published molecular markers by Song et al. [21] and the candidate interval identified through BSA analysis, 24 SSR loci within the preliminarily mapped region on chromosome 19 were selected. The primer sequences were obtained from SoyBase (https://soybase.org/) (accessed on 11 September 2024) and synthesized by the TsingKe Biological Technology Co., Ltd. (Beijing, China). Polymorphism between the two parental lines—Heinong 63 (paternal, tall-stem) and Longken 3077 (maternal, dwarf-stem)—was assessed using an 8% non-denaturing polyacrylamide gel electrophoresis (PAGE) system. In total, 10 SSR markers with stable and clear polymorphic bands were identified (

Table 1. Polymorphic molecular markers for fine mapping.

Primer Name	Forward Sequence (5′-3′)	Reverse Sequence (5′-3′)
BARCSOYSSR_19_1452	TGACACATTCTGAAACGGATG	GTAGCATTTAAATTAAGGCAAAAGA
BARCSOYSSR_19_1456	TATGGCCCGAAAATAACGAA	CGCATATGACAAGGAAGCAA
BARCSOYSSR_19_1459	TCCAACCCTAATCTGTCCTGTT	GAGAAGGTTTTGCTACGCCA
BARCSOYSSR_19_1462	CATAACTTCATTACAATTTTTACACCA	TGGATAAACTAGGTTTTTGGCTT
BARCSOYSSR_19_1468	GGGAAAATCAAAGATAATGTCAAA	AAATTGCAGCGGGTGTGTAT
BARCSOYSSR_19_1469	TTAACGGTGTCTCCAGCGTA	ATGTTTGGATCCCCATTCCT
BARCSOYSSR_19_1472	AAGGTTGCATTGTCAGGGAG	CAAACATTTCTCTTAACATTTAGCCA
BARCSOYSSR_19_1478	GTTTGCTGGAAGGATGTGGT	TCTCTTTCCAACAAGAAGTCGTC
BARCSOYSSR_19_1483	TCAAAAGAAAAGAAATGAAAAAGAA	TTGTTGTTTTGCCTTCACGA
BARCSOYSSR_19_1491	CAGCTACATTGCATCCGTGT	CAATGGTGCTTTTTCCTTTCA
BARCSOYSSR_19_1496	ATATATAACCATATAACCCATCAGCA	TCAATGGTTTCTTTTGGAACAA
BARCSOYSSR_19_1501	CGAATCACCCTAAATCCTTGTG	ACAGGGTCATGCAACAATTT
BARCSOYSSR_19_1506	TGTTCCTTGTTGGGGTTTTC	GCCACTGTAATTTGTGGCAA
BARCSOYSSR_19_1513	CCCTCTCCCTCTTTGAATCC	TTGCCACCAAGGTTGATGTA
BARCSOYSSR_19_1516	TGTGCCCTACAACAGAACCA	TAGGTATACCATGGAGCGGC
BARCSOYSSR_19_1521	ATTTTCCTTAACGGGCACAA	ACAATCACATCAGAAAAGATGACTA
BARCSOYSSR_19_1527	TTTCCTCTAATAAACATAATGTCGAG	AAATTGTGAGATTAATGGGAATG
BARCSOYSSR_19_1532	CCTTTTCGAGACAATCCCAA	TGATCCTATTTTGTTTCCCACA
BARCSOYSSR_19_1536	AATCGAATTAACCCAAACCAAA	TCAATTGGATTTGATTTTTGAA
BARCSOYSSR_19_1541	GCGCATCACAAGTTTTATAGATGCTGA	GAGGTCTAGTGCTTTGGTAAGGTT
BARCSOYSSR_19_1545	TCAGATCAGGTTGGTGCTTCT	TCACTTTTTGGTCGTCACAAG
BARCSOYSSR_19_1550	TTCTCCACCCCATTTTAAGG	TTGGCATATGACTAAAAGGGAA
BARCSOYSSR_19_1554	TCCCCCTCTCAATCTTTTCC	AGGATCCATCGCTACCCTG
BARCSOYSSR_19_1555	ACTTGAGCTTGGCATTGAGC	AAGCTTGAGTTTGGTTTCTTTAGC

Note: the markers in bold black font are screened polymorphic SSR markers.

).

PCR amplification was conducted using gene-specific primers, with a reaction volume of 20 μL for the parental lines and 10 μL for the F₂ population, following standard thermal cycling conditions. The selected 10 SSR primer pairs were used for genotyping 543 individual F₂ plants from the Heinong 63 × Longken 3077 population.

2.7. QTL Mapping and Validation for Plant Height

To refine the BSA-identified candidate region, QTL mapping was performed using genotypic data from 10 polymorphic SSR markers and phenotypic data from 543 F₂ individuals. These markers were selected from the set of 24 SSRs described above and were located within the 1.8 Mb interval on chromosome 19. Genotypic and phenotypic data were formatted according to the template requirements of QTL IciMapping software version 4.2. The inclusive composite interval mapping (ICIM) method [22] was applied to detect QTLs associated with plant height.

A threshold LOD value at the 95th percentile was determined by 1000 permutation tests, and this value was used as the significance cutoff. A QTL was considered significant when the observed LOD score exceeded the threshold in the actual analysis, indicating the presence of a putative QTL within the corresponding interval.

2.8. Development and Validation of InDel Markers

Based on the resequencing results of the parental lines within the mapped interval, an InDel variant was identified at position 48,052,853. The sequence containing the InDel site was downloaded from the Phytozome database (https://phytozome.jgi.doe.gov/pz/portal.html) (accessed on 15 September 2024), and primers flanking the InDel were designed using the NCBI Primer-BLAST tool (https://www.ncbi.nlm.nih.gov/tools/primer-blast/) (accessed on 15 September 2024) (

Table 2. Positions of the InDel markers, the primers, and the length of the amplified fragment.

Primer Name	Physical Position (bp)	Forward Sequence (5′-3′)	Reverse Sequence (5′-3′)	Product Size (bp)
				Heinong 63	Longken 3077
PL7	48,052,853–48,052,900	TTTGGAAAATCCACCAATTATGTT	ACTATTGTGCGACTTGATATACT	319	272

).

A PCR was conducted in a 20 μL reaction system optimized for genotyping analysis. The amplification products were resolved using 6% denaturing polyacrylamide gel electrophoresis (PAGE).

2.9. Quantitative Real-Time PCR (qRT-PCR)

Total RNA was extracted from the shoot apical meristems of the two parental lines, Heinong 63 and Longken 3077, using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA; Cat. No. 15596-026). For each genotype, three biological replicates were prepared, and each replicate consisted of pooled shoot apical meristems from three individual plants. To remove genomic DNA contamination, the extracted RNA was treated with DNase I (Vazyme, EN401, Nanjing, China) according to the manufacturer’s instructions. First-strand cDNA was synthesized from the extracted RNA using a reverse transcription kit (Vazyme, R312, Nanjing, China). The qRT-PCR was performed using SYBR qPCR Mix (Vazyme, Q711, Nanjing, China) on a LightCycler 480 II system (Roche Diagnostics, Basel, Switzerland).

The soybean Actin11 gene was used as the internal reference to normalize gene expression levels. Relative expression levels of the target genes were calculated using the 2^ΔCt method [23].

2.10. Functional Annotation and Candidate Gene Screening Within the Candidate Region

Based on the candidate region identified through the integration of the SNP and InDel association analyses, the functional annotation of genes within this region was performed. Gene sequences were aligned using BLAST against the NR [24], Swiss-Prot, GO [25], KEGG [26], and COG [27] databases to obtain comprehensive functional information. Candidate genes were subsequently screened by integrating the functional annotation results with gene expression level analysis.

2.11. Statistical Analysis

All statistical analyses were performed using Microsoft Excel 2016 and GraphPad Prism 8.0. Data are presented as the mean ± standard deviation (mean ± SD). The number of samples or biological replicates (n) is indicated in the corresponding figure legends. Statistical significance between two groups was determined using a two-tailed Student’s t-test.

3. Results

3.1. Phenotypic Analysis of Plant Height in Heinong 63, Longken 3077, and the F₂ Segregating Population

The male parent Heinong 63 (HN63) exhibited a tall-stem phenotype, with an average plant height of 117.2 cm, while the female parent Longken 3077 (LK3077) displayed a dwarf-stem phenotype, with an average plant height of 81.0 cm. The difference between the two parents was highly significant (Figure 1A). HN63 had an average of 20.5 nodes on the main stem, whereas LK3077 had an average of 18.5 nodes, also showing a highly significant difference (Figure 1B). The average internode length of HN63 was 6.3 cm, compared to 5.5 cm in LK3077, indicating another highly significant difference (Figure 1C). These results suggest that the plant height difference between HN63 and LK3077 is caused by the combined effects of the node number and internode length.

A total of 543 F₂ individuals derived from the HN63 × LK3077 cross were phenotyped for plant height. Plant height ranged from 46.0 cm to 147.0 cm, with an average of 96.7 cm and a coefficient of variation (CV) of 16.7% (

Table 3. Statistical analysis of plant height phenotypes in parental and F₂ populations.

Population	Male Parent (Mean ± SD, cm)	Female Parent (Mean ± SD, cm)	F2 Segregation Population
			Range	Mean ± SD (cm)	CV (%)	Kurtosis	Skewness
HN63 × LK3077	136.0 ± 5.8	88.6 ± 4.2	46.0–147.0	96.7 ± 16.2	16.7	3.1	0.8

). The F₂ population exhibited a continuous and wide distribution of plant height, with clear evidence of transgressive segregation. The kurtosis and skewness values were 3.1 and 0.8, respectively, indicating that this population is suitable for QTL mapping.

3.2. Sequencing Data Analysis

3.2.1. Sequencing Data Quality Control

A total of 20 tall and 20 dwarf individuals from the F₂ population were selected to construct two extreme phenotype pools, which were sequenced along with the two parental samples using the Illumina HiSeq platform for high-throughput whole-genome resequencing. After filtering the raw data, a total of 75.7 Gbp of clean bases was obtained, with Q30 scores exceeding 92.9%. The GC content ranged from 33.9% to 34.6%. The insert size distribution exhibited a unimodal normal distribution, indicating uniform library construction.

The mapping rate of the samples to the soybean reference genome reached 99.1%, with an average sequencing depth of 17× and a genome coverage of 97.4%, meaning that at least one base was covered in the vast majority of the genome. These quality control metrics demonstrate that the sequencing data were of high quality and suitable for subsequent variant detection and trait association analysis.

3.2.2. Association Analysis

After filtering the BSA sequencing data of the F₂ population, a total of 636,540 high-quality SNPs and 123,698 InDel variants were obtained. To further investigate the association between these variants and the plant height trait, both the Euclidean distance (ED) algorithm and the index algorithm were applied. For each variant site, the ED values were calculated based on the allelic frequency differences and sequencing depth between the extreme phenotype pools, resulting in thresholds of 0.30 for SNPs and 0.25 for InDels. SNPs and InDels were analyzed for trait associations using the ED and index methods (Figure 2).

The SNP-based association analysis identified two candidate regions on chromosome 19, with a total length of 2.0 Mb. The InDel-based analysis identified a region of 2.6 Mb on the same chromosome. The intersection of the two methods revealed a single overlapping candidate region with a length of 1.8 Mb, containing 243 annotated genes (

Table 4. F₂-population-associated region on chromosome 19 based on SNP/InDel analysis conducted using two methods.

Association Analysis Type	Chr.	Start Position	End Position	Region Size (Mb)	Gene Number in the Regions
SNP correlates results	Chr.19	47,239,765	47,472,236	0.2	23
	Chr.19	47,558,547	49,368,580	1.8	243
InDel correlates results	Chr.19	47,526,896	50,124,372	2.6	342
The intersection of the two association results	Chr.19	47,558,547	49,368,580	1.8	243

). The sequencing and association analysis results were visualized using Circos software (v0.69-9; http://circos.ca/) (accessed on 2 September 2024) to display variant distributions and BSA-based associations across the genome (Figure 3).

3.3. Fine Mapping of the PH19-31 Locus

Based on the BSA-Seq data and the new SSR markers published by Song et al. [21], 24 primer pairs were selected from the vicinity of the initially mapped 1.8 Mb interval. Polymorphism screening using the two parental lines identified 10 SSR primer pairs that produced clear and stable polymorphic bands (Table 1). These 10 SSR markers were then used for genotyping 543 individual plants from the F₂ population for the linkage analysis (Figure 4).

The results show that the original 1.8 Mb region was gradually narrowed down to the interval between markers BARCSOYSSR_19_1478 and BARCSOYSSR_19_1483, located at positions 47,986,062 bp and 48,101,240 bp, respectively. The final candidate interval spanned 115.1 kb, which was designated as PH19-31, and it contained a total of 15 annotated genes.

3.4. Validation of the Candidate Region and Linkage Marker Analysis

Based on the fine-mapping results, QTL analysis for plant height was performed using the genotypic data of 10 SSR markers and phenotypic data from the F₂ population in QTL IciMapping 4.2. A single QTL associated with plant height was identified between markers SSR19_1478 and SSR19_1483, which overlapped with the previously defined fine-mapped interval. This QTL exhibited a LOD score of 3.83 and explained 4.01% of the phenotypic variation.

Further analysis of the genotypes and corresponding plant height phenotypes at the two flanking markers (Figure 5B,C) revealed the following: at the SSR19_1478 locus, 149 individuals carrying the Heinong 63 allele (A) had plant heights ranging from 104.0 to 143.0 cm, with an average of 121.7 cm. In contrast, 70 individuals with the Longken 3077 allele (B) showed plant heights ranging from 46.0 to 117.0 cm, with an average of 103.9 cm.

At the SSR19_1483 locus, 126 individuals with the Heinong 63 allele (A) exhibited plant heights between 69.0 and 144.0 cm, averaging 117.1 cm, whereas 105 individuals with the Longken 3077 allele (B) ranged from 46.0 to 138.0 cm, with an average of 115.4 cm. The differences in plant height between the different genotypes at both markers were highly significant, indicating that these markers are closely linked to the plant height trait.

3.5. Functional Validation and Genetic Association of the InDel Marker PL7

To improve the accuracy and practical utility of the mapping results, an InDel marker named PL7 was developed within the candidate region. PL7 is located within the gene Glyma.19G229400 and is positioned 66.7 kb and 48.3 kb away from the SSR markers BARCSOYSSR_19_1478 and BARCSOYSSR_19_1483, respectively.

Genotyping of the F₂ population using PL7 revealed that 125 individuals carrying the HN63 (A) allele had an average plant height of 117.6 cm, whereas 112 individuals carrying the LK3077 (B) allele had an average plant height of 114.6 cm (Figure 5D). These results indicate that PL7 is tightly linked to the plant height trait and has potential for use in marker-assisted selection (MAS) in soybean breeding.

3.6. Verification of the Mapping Interval in the F₂ Population

To further validate and refine the mapped interval, recombinant individuals in the F₂ population were screened using the flanking SSR markers of the PH19-31 locus along with the InDel marker PL7. Based on phenotypic data, the PH19-31 interval was ultimately narrowed from the original 115.1 kb region between BARCSOYSSR_19_1478 and BARCSOYSSR_19_1483 to a smaller region of approximately 48.3 kb between PL7 and BARCSOYSSR_19_1483, which contains nine annotated genes (Figure 6).

3.7. Functional Analysis of Genes Within the Mapped Interval

By conducting fine mapping, we identified a total of nine genes within the mapped interval, and their functional annotations are listed in

Table 5. Gene and function predictions in location range.

Gene ID	Start Position	End Position	Gene Function Annotation
Glyma.19G229400	48,048,375	48,052,530	Calmodulin-binding protein-like function
Glyma.19G229500	48,059,444	48,063,147	Calmodulin-binding protein-like function
Glyma.19G229600	48,064,703	48,067,937	Protein of unknown function
Glyma.19G229700	48,069,220	48,070,638	Chaperone DnaJ-domain superfamily protein
Glyma.19G229800	48,072,274	48,078,300	α Karyopherin (importin)α
Glyma.19G229900	48,081,061	48,081,390	Transcription factor GATA (GATA binding factor)
Glyma.19G230000	48,084,644	48,087,743	Polynucleotidyl transferase
Glyma.19G230100	48,093,731	48,099,790	Kinesin (KAR3 subfamily)
Glyma.19G230200	48,103,324	48,106,756	Serine/threonine protein phosphatase

. To further determine candidate genes, shoot apical meristem tissues from HN63 and LK3077 during the VC stage (when the first unifoliate leaves are fully expanded) under natural conditions were collected for an expression analysis of the genes within this region.

The results show that two genes, Glyma.19G230000 and Glyma.19G230200, were identified as candidate genes based on their differential expression patterns between the two parental lines (Figure 7).

3.8. Protein Function Analysis of Candidate Genes

A protein function analysis was performed for two genes within the candidate interval associated with plant height.

The gene Glyma.19G230000 encodes a 3′-repair exonuclease that plays a key role in DNA damage repair and the maintenance of genome integrity. This gene is believed to play an important role in cell membrane formation, DNA repair, and stress response in plants. Its function is closely related to cell-cycle regulation during plant growth and development, thereby potentially affecting plant height [28].

Glyma.19G230200 encodes the protein phosphatase 2C (PP2C), a widely recognized regulatory protein involved in hormone signal transduction, responses to abiotic stress, and developmental processes. PP2C proteins may interact with auxin signaling components to modulate cell division and elongation, thus influencing plant height [29] (

Table 6. Candidate gene protein function annotation.

Gene ID	Proteins with Homologous Sequences
Glyma.19G230000	Three prime repair exonuclease 1, 2
Glyma.19G230200	Protein phosphatase 2C 5-related

).

4. Discussion

4.1. Regulatory Genes and QTLs Associated with Plant Height in Soybeans

Plant height is one of the most important agronomic traits in soybeans, as it directly affects plant architecture, yield potential, lodging resistance, and harvesting efficiency. The genetic regulation of plant height is complex and involves multiple genes. Although numerous QTLs related to plant height have been identified, the underlying genetic mechanisms remain incompletely understood.

In previous studies, many QTLs associated with plant height have been reported, particularly on chromosome 19. For example, 11 QTLs were identified using a recombinant inbred line (RIL) population, 6 of which were located on chromosome 19, with the highest phenotypic variance explained (PVE) reaching 36% [30]. Further research revealed that the SNP marker Gm19_45000827 in this region was significantly associated with plant height, highlighting its core regulatory role [31]. In addition, whole-genome resequencing was used to narrow the major QTL qPH16 from 960 kb to 477.55 kb, and high-precision molecular markers were developed based on long-fragment InDels to support marker-assisted pyramiding breeding [32].

Two previously reported plant height QTLs (plant height 4–4 and plant height 5–10) were found to overlap with the interval identified in the present study. The lengths of these regions were 12.2 Mb and 2.0 Mb, respectively. Plant height 4–4 corresponds to the well-known Dt1 locus. Using RFLP markers, a major-effect QTL controlling plant height (Dt1) was identified in an F₂ population derived from PI 97100 × Coker 237, accounting for 67.7% of the variation in plant height. Plant height 5–10 was identified in the Young × PI 416937 population and is located near marker K385 on linkage group L, explaining 6.6% of the phenotypic variation [5].

In this study, we identified a QTL locus PH19-31 located within a known QTL hotspot for plant height. This region was mapped using BSA-Seq based on tall and dwarf phenotype bulks constructed from the F₂ population, in combination with the SNP/InDel-index and Euclidean distance (ED) methods. The final mapped interval, PH19-31, was located on chromosome 19 and spans only 48.3 kb, which is significantly narrower than previously reported QTL intervals. These results underscore the value of this region for dissecting the genetic basis of plant height and its potential application in molecular breeding.

4.2. Functions and Potential Biological Roles of Candidate Genes

In this study, two candidate genes for plant height, Glyma.19G230000 and Glyma.19G230200, were identified. Glyma.19G230000 is annotated as a polynucleotide adenylyltransferase, an enzyme involved in plant adaptation to biotic stimuli and environmental stress. It plays a critical role in RNA metabolism, particularly in RNA editing and degradation processes [28]. By regulating gene expression under adverse conditions, it may indirectly influence important agronomic traits such as plant height.

Glyma.19G230200 encodes a serine/threonine protein phosphatase (PPP), which regulates signaling pathways mediated by plant hormones such as jasmonic acid (JA) and salicylic acid (SA). This gene is known to play a key role in plant responses to pathogens, wounding, drought, and low-temperature stress. PPPs have been shown to negatively regulate JA signaling, thereby potentially influencing plant growth, development, and defense mechanisms [33]. Given its role in hormonal signal transduction, it is speculated that Glyma.19G230200 may affect plant height by modulating hormone pathways related to growth, such as those involving jasmonic acid or gibberellins (GA).

The expression analysis revealed that both candidate genes were significantly differentially expressed in the shoot apical meristems of the two parental lines, with markedly higher expression observed in the dwarf parent LK3077 compared with the tall parent HN63. The shoot apical meristem serves as a key site of active cell division and hormonal regulation. Genes preferentially expressed in this region are frequently involved in developmental processes such as meristem maintenance, internode elongation, and hormone-mediated growth responses [34].

The elevated expression of Glyma.19G230000 and Glyma.19G230200 in LK3077 may suggest their participation in regulatory pathways that modulate cell proliferation or hormonal balance, potentially involving GA, auxin (IAA), or JA signaling. This suggests a potential role in negatively regulating stem elongation. Previous studies have shown that genes that are highly expressed in the shoot apex are often closely associated with the cell division rate, cell wall remodeling, or hormone signaling [35].

Among the genes located within the fine-mapped interval, Glyma.19G230200 exhibited the most pronounced expression difference and had the highest overall expression level, making it the most likely key regulator of plant height. In plants, many genes related to the cell cycle, microtubule assembly, or cell elongation are specifically expressed at high levels in the shoot apical region [36,37]. Therefore, it is plausible that Glyma.19G230200 influences the internode length and overall plant height by regulating cell division activity in the meristem. Furthermore, it may also be involved in hormone-responsive pathways, particularly those related to gibberellin (GA) or auxin (IAA) signaling, thereby affecting stem growth [38,39].

4.3. Utility of InDel Markers in Molecular Breeding

Due to their high polymorphism and low development cost, InDel markers have been widely applied in QTL mapping, gene cloning, and marker-assisted selection (MAS) in crop breeding. A recent study conducted the de novo genome assembly of seven wild soybean accessions to construct a soybean pan-genome, and identified between 500,000 and 770,000 InDel variations between G. soja and G. max. Some of these InDels were located within functional genes, resulting in frameshift mutations and contributing to variations in traits such as disease resistance, plant architecture, maturity, and seed composition [40]. Over 93% of 70 selected InDel loci were validated by Sanger sequencing, confirming their potential application in soybean genetic diversity analysis and agronomic trait improvement.

InDel markers have also been used in the fine mapping of plant-height-related QTLs. Using whole-genome resequencing data in combination with a BSA strategy, one study successfully narrowed the major-effect QTL qPH16 to a 400 kb interval, thereby improving the mapping resolution [32]. Another study developed the InDel marker Gm16P763998, based on an 18 bp insertion in the candidate gene Glyma.16g076600, which is associated with pod shattering. This marker showed high concordance with the pod-shattering phenotype in two RIL populations and achieved a predictive accuracy of 95.8% to 100% in 120 soybean accessions, indicating strong stability and applicability [41].

Given the demonstrated importance and broad application of InDel markers in soybeans, we developed an InDel marker, PL7, within the plant height candidate interval. Based on a genotype–phenotype association analysis in the F₂ population, the interval was further narrowed to 48.3 kb, and PL7 was found to be closely linked to the plant height trait. These results highlight its potential for use in marker-assisted selection to improve plant height in soybean breeding programs.

5. Conclusions

In this study, the plant height gene in soybeans was mapped to a 48.3 kb interval on chromosome 19 using BSA-Seq combined with map-based cloning. In total, nine genes were identified within this region. Based on an expression analysis that compared the two parental lines, Glyma.19G230000 and Glyma.19G230200 were identified as candidate genes associated with plant height. In addition, an InDel marker, PL7, was developed within this interval. These findings provide a foundation for further exploring the molecular regulatory mechanisms underlying plant height in soybean.

These findings demonstrate the effectiveness of integrating high-throughput sequencing with classical genetic mapping for narrowing QTL regions and identifying functional loci. The candidate genes and molecular markers identified in this study provide valuable resources for further investigation of the molecular mechanisms underlying plant height and for the development of high-yield, lodging-resistant soybean varieties. Future studies will focus on functional validation of the candidate genes and the exploration of their roles against different environmental and genetic backgrounds.