Soybean (*Glycine max*) is a crucial crop globally, providing edible oil, vegetable protein, and animal feed. However, its yield significantly trails behind major cereals like rice, wheat, and maize. Cereals have benefited from the Green Revolution, where semi-dwarf varieties with shorter, more compact plants allow for high-density planting, leading to increased yields. In contrast, soybean plant height is determined by both node number and internode length. Reducing plant height by decreasing node number also decreases pod number (since pods attach to nodes), ultimately reducing yield. Therefore, shortening internodes while maintaining node number is a promising strategy to increase soybean yield, especially under high-density planting conditions. Reduced plant height and lodging tolerance are critical for dense planting. Gibberellins (GAs) are plant hormones that significantly influence plant height. In rice, the *sdl* allele (encoding GA20-oxidase 2) and in wheat, the *Rht* alleles (encoding N-terminally truncated DELLA proteins) are key ‘Green Revolution genes’ that reduce plant height to enhance high-density planting and nitrogen use efficiency, significantly increasing grain yield. While several genes regulating soybean plant height via the gibberellin pathway have been identified, research on internode length has been limited. The CRY1-STF (HY5)-gibberellin regulatory module plays a role, with STF1 and STF2 (homologs of Arabidopsis ELONGATED HYPOCOTYL 5 (HY5)) activating *GA2ox* expression under normal light conditions, thus decreasing GA1 levels and suppressing internode elongation. In Arabidopsis, the E3 ubiquitin ligase COP1, with SPA family members, interacts with HY5, leading to HY5 ubiquitination and degradation. Four *SPA* genes exist in Arabidopsis, with all SPA proteins containing WD-repeat, coiled-coil, and kinase-like domains. The WD-repeat domain is essential for COP1-SPA complex function, and SPA1 also interacts independently with HY5, reducing HY5 protein levels in light. The role of SPA proteins in soybean plant height and internode length, however, remains unexplored. This study aims to identify and characterize genes controlling internode length in soybean to enhance yield under high-density planting conditions.

Extensive research on plant architecture and yield improvement in major cereal crops has revealed the importance of semi-dwarfing genes and their impact on high-density planting. The Green Revolution highlighted the success of reducing plant height while maintaining or improving yield through manipulation of gibberellin biosynthesis and signaling pathways. In rice, the discovery of the *sd1* gene, encoding a defective gibberellin 20-oxidase, revolutionized rice production. Similarly, in wheat, the identification of the *Rht* genes, which encode truncated DELLA proteins, has had a significant impact on wheat yield. These studies serve as a basis for understanding the potential of manipulating plant height in other crops, particularly soybean. Previous work on soybean has identified several genes involved in height regulation through the gibberellin pathway, but relatively less is known about specific mechanisms regulating internode length. The role of the CRY1-STF-gibberellin module in regulating soybean height has been explored, demonstrating the importance of light signaling in controlling internode elongation. The function of SPA proteins in regulating plant height and internode length in soybean remained largely unknown before this study. Understanding the role of these genes and pathways in soybean is crucial for developing high-yielding cultivars suitable for high-density planting. The study builds upon existing knowledge of plant architecture and gibberellin signaling to investigate a novel gene in soybean controlling internode length.

This study employed a multi-faceted approach combining genetic mapping, positional cloning, molecular biology techniques, and field trials. A soybean mutant library was generated using γ-ray irradiation of the elite cultivar Heinong 35 (HN35). A dwarf mutant with extremely short internodes, named *reduced internode 1* (*rin1*), was identified. To identify the underlying gene, the *rin1* mutant was crossed with another elite cultivar, Heihe 43 (HH43), creating an F2 population. Genotyping-by-Sequencing (GBS) data, along with phenotypic data (internode length and plant height), were used to map the *rin1* locus. Fine-mapping narrowed down the candidate region, identifying a homolog of Arabidopsis *SPA3* (named *SPA3a* in soybean) with a mutation in the coding region as the most likely candidate gene. CRISPR/Cas9 technology was used to generate knockout mutants of *SPA3a* to confirm its role. Near-isogenic lines (NILs) were developed to further confirm the phenotype. Natural variation in *SPA3a* was analyzed using resequencing data from 1295 soybean accessions to investigate the prevalence of the identified mutation. Gene expression patterns were investigated using RT-qPCR and in situ hybridization. Yeast two-hybrid assays, pull-down assays, and bimolecular fluorescence complementation (BiFC) were used to investigate protein-protein interactions. Cell-free and *in vivo* degradation assays were performed to examine the effect of RIN1 on STF protein abundance. The effect of *rin1* on the expression of GA2 oxidase genes was investigated using RT-qPCR and dual-luciferase transient expression assays. Gibberellin levels were quantified by LC-MS/MS. Finally, field trials were conducted to evaluate the effect of *rin1* on grain yield under different planting densities.

This study identified *rin1*, a soybean mutant with significantly shorter internodes and improved grain yield under high-density planting conditions. The *rin1* phenotype is caused by a partial loss-of-function mutation in *SPA3a*, a gene homologous to Arabidopsis *SPA3*. The RIN1 protein physically interacts with STF1 and STF2, two homologs of Arabidopsis HY5, promoting their degradation. RIN1 negatively regulates the expression of *GA2ox7a* and *GA2ox7b*, affecting gibberellin metabolism and ultimately controlling internode development. Field trials demonstrated that *rin1* mutants exhibit shorter plant heights and internode lengths compared to the wild type, yet yield significantly higher grain yield per plant. Crucially, the plot yield of *rin1* mutants was consistently higher than the wild type at high planting densities, reaching the most dramatic yield advantage at the highest density (450,000 plants per hectare). In situ hybridization revealed *RIN1* expression is particularly high in the shoot apical meristem during early vegetative growth, consistent with a role in internode development. Analysis of natural variation in *RIN1* across various soybean accessions revealed three haplotypes, with non-synonymous SNPs at positions 223 and 2654 influencing flowering time. Genetic analysis suggested that RIN1 functions upstream of STF1 and STF2. Exogenous GA3 application partially restored internode length in *rin1* mutants, confirming the involvement of gibberellin in the *rin1* phenotype. CRISPR-generated *rin1* mutants showed earlier flowering under long-day photoperiods, suggesting pleiotropic effects of RIN1. However, NIL-*rin1* did not show significant differences in flowering time compared to NIL-*RIN1*, potentially due to genetic background differences or functional redundancy.

This research provides strong evidence for a novel pathway regulating internode length in soybean, ultimately impacting yield under high-density planting. The identification of *rin1* and its underlying genetic basis offers a promising target for breeding high-yielding soybean cultivars. The results suggest that manipulating the RIN1-STF-gibberellin module could be a valuable strategy for improving soybean yield. The finding that *rin1* improves yield even with slightly reduced node number is significant, highlighting the potential for enhancing soybean yield without the yield penalties typically associated with reduced node number. The contrasting effects of *rin1* on internode length and flowering time suggest that these processes are regulated independently by RIN1, implying potential for fine-tuning both traits in breeding programs. Further investigation into the interaction of *rin1* with nitrogen fixation and flowering time under different environmental conditions will refine our understanding and guide breeding efforts. The success of *rin1* in enhancing soybean yield under high-density conditions contrasts with the challenges in improving soybean yield compared to cereal crops. This success might be due to the unique characteristics of soybean, such as its nitrogen-fixing capabilities and adaptability to different environments. Future research should focus on exploring the interactions between *rin1* and other genes involved in plant architecture, nitrogen use efficiency, and flowering time to develop even more improved soybean varieties.

This study successfully identified and characterized the *rin1* mutant in soybean, revealing a crucial role for RIN1/SPA3a in controlling internode length and enhancing grain yield under high-density planting conditions. The findings illuminate a novel regulatory module involving RIN1, STF1/STF2, and GA metabolism. This work provides a valuable genetic resource and a mechanistic understanding for breeding high-yielding soybean cultivars adapted to high-density planting and intercropping systems. Future research should explore the pleiotropic effects of *rin1* on nitrogen fixation and flowering time, optimize its use in different environments, and investigate the potential for stacking *rin1* with other beneficial alleles for further yield enhancement.

The study primarily focused on the Heinong 35 background. Further validation in other genetic backgrounds is necessary to confirm the broad applicability of the findings. While field trials demonstrated yield improvement, further research is needed to evaluate the performance of *rin1* mutants under diverse environmental conditions and nutrient regimes. The analysis of natural variation was limited to a specific panel of soybean accessions; broader analysis might reveal additional functional variation in *RIN1*. Finally, while the role of RIN1 in flowering time is suggested, the exact mechanisms remain to be fully elucidated.