The ZmSOC1 and GmSOC1 protein sequences were used as queries to conduct BLAST searches against the tomato proteome in Phytozome v13. The identified tomato SOC1 and SOC1-like proteins were subsequently selected for phylogenetic analysis ( Figure 1 ).

A 696-bp ZmSOC1 (also known as ZmMADS1), identical to the sequence derived from HQ858775.1 in GenBank, was cloned using polymerase chain reaction (PCR) from the cDNA of the maize inbred line B104 (Han et al., 2021). The corresponding protein sequence matches NP_001105152.1 in GenBank ( Figure 1 ). The fragment was inserted into a modified pCAMBIA1300 vector between a CaMV 35S promoter and a CaMV poly(A) signal at the KpnI and XbaI restriction sites ( Figure 1 ).

Similarly, a 630-bp GmSOC1, identical to the sequence derived from NM_0011249448.2 in GenBank, was cloned from the cDNA of the soybean cultivar Thorne. The fragment was inserted into the same modified pCAMBIA1300 vector between a CaMV 35S promoter and a CaMV poly(A) signal at the KpnI and XbaI sites ( Figure 1 ). The corresponding protein sequence is identical to NP_001236377.1 in GenBank ( Figure 1 ).

The VcSOC1K gene was previously cloned into the T-DNA region of the binary vector pBI121, positioned between the CaMV 35S promoter and the Nos terminator for constitutive expression (Song and Chen, 2018). In this study, the VcSOC1K expression cassette was excised from pBI121 using HindIII and EcoRI digestion and subsequently ligated into the T-DNA region of a HindIII- and EcoRI-digested pCAMBIA1300 (PC1300) vector. This ensures that all three constructs share the same binary vector backbone ( Figure 1 ).

All three constructs were confirmed by sequencing the target genes and subsequently transformed into Agrobacterium tumefaciens strain EHA105. For tomato transformation, indeterminate tomato (Lycopersicon esculentum) ‘Ailsa Craig’ was used. Seeds were sterilized in a 2.5% (v/v) sodium hypochlorite solution and germinated to produce cotyledons. Non-inoculated cotyledons were cultured on antibiotic-free regeneration medium to produce non-transgenic (NT) regenerants, which were used as controls. The regeneration medium consisted of Murashige and Skoog (MS) basal salts (Murashige and Skoog, 1962), Gamborg B5 vitamins (Gamborg et al., 1968), 2.85 μM zeatin riboside, 2.86 μM indole-3-acetic acid (IAA), 30 g/L sucrose, and 6 g/L agar. Transformation was carried out following established protocols (Danial et al., 2021). Regeenrated shoots approximately 2–3 cm in length were excised and transferred to 30 mL MS medium for rooting. For NT plants, the rooting medium lacked antibiotics, whereas for transgenic plants, it was supplemented with 10 mg/L hygromycin, 250 mg/L timentin, and 250 mg/L cefotaxime.

After rooting, T₀ plantlets were transplanted into 4-inch plastic pots containing water-soaked Suremix Perlite planting medium (Michigan Grower Products Inc., Galesburg, MI). The plants were covered with plastic bags and placed in a growth room maintained at 25 °C with a 16/8-hour light/dark photoperiod. Over a two-week period, the plastic bags were gradually removed to acclimate the plantlets. Once acclimated, the plants were repotted into one-gallon pots and transferred to a greenhouse.

For each construct, approximately 10–20 T₀ transformants were cultivated in the greenhouse. Each transformant, derived from a separate explant, was considered an independent transgenic line. At the time of repotting into one-gallon pots, transformed and non-transformed plants of similar size, 3–5 lines per construct and non-transformants, were selected. These comparable plants were used as representatives for phenotyping of T₀ plants. First-generation (T₁) seeds were harvested separately from each T₀ plant.

For phenotyping T₁ plants, 20–30 seeds from each of three selected T₀ transgenic lines per construct were germinated in soil. Ten plants from each transgenic line were transferred to one-gallon pots and grown in a greenhouse.

Phenotypic assessments for each plant included measuring the time and height of the first flower appearance, recording the time of the first mature fruit appearance, counting the total number of branches and fruits harvested, and weighing all harvested fruits. Photographs were taken to document phenotypic variations.

For genotyping of the T₀ and T₁ plants, genomic DNA was extracted from young leaves using the cetyltrimethylammonium bromide (CTAB) method (Doyle and Doyle, 1987). PCR was performed to detect the hygromycin phosphotransferase (hpt) transgene and the full-length sequences of GmSOC1, ZmSOC1, and VcSOC1K using the primers listed in Supplementary Table S1 .

Three groups of T₁ plants, derived from three T₀ transgenic lines containing GmSOC1, were selected for RNA sequencing and transcriptome analysis. Each group of T₁ plants was segregated into transgenic and NT plants. For each group, two newly formed mature leaves from all transgenic plants before their flowering were pooled together to create one biological replicate for transgenic plants, while leaves from NT plants were similarly pooled to form one biological replicate for NT plants. In total, three transgenic samples and three NT samples were collected, immediately frozen in liquid nitrogen, and stored at -80°C for RNA isolation.

Total RNA from each sample was isolated using a modified CTAB method (Chang et al., 1993) and further purified with the RNeasy Mini Kit (Qiagen, Valencia, CA, United States). To eliminate any residual DNA, on-column DNase digestion was performed using the RNase-Free DNase Set (Qiagen, Valencia, CA, United States). RNA quality was assessed with the High Sensitivity RNA ScreenTape system (Agilent Technologies, Santa Clara, CA, United States). All RNA samples used for sequencing and RT-qPCR analysis had RNA integrity number equivalent scores exceeding 5.0.

The RNA samples were sequenced using the Illumina NovaSeq 6000 platform (150 bp paired-end reads) at the Research Technology Support Facility at Michigan State University (East Lansing, Michigan, United States). The quality of the sequencing reads was evaluated using the FastQC program, focusing on per-base quality scores. Each of the six biological samples yielded 19–21 million paired reads (MR), with average quality scores exceeding 35, ensuring suitability for transcriptome analysis.

A transcriptome reference was assembled from approximately 120 million paired reads (MR) obtained from all six NT and transgenic lines using Trinity v2.15.1 (Haas et al., 2013). This reference was used for differential expression analysis. Transcripts identified as differentially expressed (DETs) with a false discovery rate (FDR) below 0.05 were selected for further analysis of various pathway genes. The transcriptome reference was annotated using Trinotate/4.0.2.

Pathway genes for nine phytohormones in Arabidopsis, including auxin, cytokinin, abscisic acid (ABA), ethylene, gibberellin (GA), BRs, jasmonic acid, salicylic acid, and strigolactones, were retrieved from the RIKEN Plant Hormone Research Network. Similarly, sugar pathway genes in Arabidopsis were identified. These hormone, MADS-box, and sugar pathway genes from Arabidopsis were used as queries to perform BLAST searches against the transcriptome reference, and isoforms with e-values less than -20 were identified for transcriptome comparisons. Flowering pathway genes in Arabidopsis and cereals (Walworth et al., 2016) were used to analyze flowering-related differentially expressed transcripts (DETs) identified in this study. Gene networks of overrepresented gene ontology (GO) terms for the selected DETs were constructed using Cytoscape 3.10.3.

Six selected DEGs were further analyzed through RT-qPCR using using the SYBR Green system (LifeTechnologies, Carlsbad, CA). A tomato ACTIN gene served as the reference gene to normalize the RT-qPCR results. All primers used in the analysis are included in Supplementary Table S1 . The same RNA samples used for RNA sequencing, including three biological samples and three technical replicates, were used for the RT-qPCR analysis. Fold changes were calculated using 2^−ΔΔCt, where ΔΔCt = (Ct_TARGET – Ct_NOM)_transgenic – (Ct_TARGET – Ct_NOM)_{non-transgenic}.

Data were processed using Microsoft Office Excel. Graphing and statistical analysis were generated in R (version 4.4.0) using the ggplot, annova, and emmeans packages. Fisher’s Protected Least Significant Difference test was used to determine statistical significance at p < 0.01 and p < 0.05 levels.

The tomato proteome in Phytozome v13 includes one SOC1 protein and four SOC1-like proteins, which have been identified as SlTM3, SlSTM3, SlMBP18, and SlMBP23 by Zahn et al (Zahn et al., 2023). These five genes are distributed across four chromosomes ( Figure 1 ). Their protein sequences exhibit high similarity to Arabidopsis SOC1, GmSOC1, ZmSOC1, and VcSOC1K. Notably, GmSOC1 and ZmSOC1 cluster with SlMBP18 and the SlSOC1-like protein XP_004252597.2, while VcSOC1 clusters with SlTM3, SlSTM3, SlMBP23, and Arabidopsis SOC1.

In the transcriptome reference assembled from all transcripts in the leaves of the ‘Ailsa Craig’ cultivar used in this study, transcripts corresponding to SlTM3, SlSTM3, SlMBP18, and the protein identical to XP_004252597.2 were detected, while SlMBP23 was absent.

The conserved nature of SOC1 and SOC1-like genes suggests that their orthologs SlTM3, SlSTM3, SlMBP18, and SlMBP23 in tomato may play similar roles to those in other plant species. Consequently, the constitutive expression of GmSOC1, ZmSOC1, and VcSOC1K may have the potential to influence fruit production.

A total of 69 hygromycin-resistant T₀ transgenic lines for GmSOC1_CX, 51 for ZmSOC1_CX, and 92 for VcSOC1K_CX were obtained across the three constructs, providing sufficient material for selecting comparable lines for phenotyping and seed harvesting ( Figures 1B ).

The constitutive expression of GmSOC1 (GmSOC1_CX) resulted in non-significant changes, such as earlier flowering and increases in plant height and the number of flower clusters ( Supplementary Figures S1A–C ). Significant changes included increases in branch and fruit numbers, enhanced fruit production, larger fruit size, and earlier fruit maturation ( Supplementary Figures S1D–H ).

The constitutive expression of ZmSOC1 (ZmSOC1_CX) led to significant changes, including earlier flowering, earlier fruit maturation, and increased fruit number and production ( Supplementary Figures S1A, E–G ). Non-significant changes included decreases in plant height, numbers of flower clusters and branches, and fruit size ( Supplementary Figures S1B–D, H ).

The constitutive expression of VcSOC1K (VcSOC1K_CX) induced significant changes, including earlier flowering, increased numbers of flower clusters and fruits, enhanced fruit production, and reduced fruit size ( Supplementary Figures S1A, C, F–H ). It also caused non-significant changes, such as earlier fruit maturation and increases in plant height and branch number ( Supplementary Figures S1B, D, E ).

Taken together, GmSOC1_CX, ZmSOC1_CX, and VcSOC1K_CX all contributed to increased fruit numbers and production ( Supplementary Figure S1 ).

T₁ transgenic plants were phenotypically compared with their corresponding plants ( Figures 2A ). Under greenhouse conditions, among the nine groups with constitutive expression, ZmSOC1_CX17 was the only group of T₁ transgenic plants to exhibit significantly earlier flowering, while the other T₁ groups showed no difference in flowering time ( Figure 2 ).

Consistent with observations in T₀ plants, GmSOC1_CX, ZmSOC1_CX, and VcSOC1K_CX all resulted in increased fruit numbers and overall production, while the average fruit size showed no significant decrease ( Figures 2E ).

The T₁ GmSOC1_CX plants were selected for transcriptome analysis to confirm the presence of transgenes and identify genes responsive to GmSOC1 expression in tomato. This selection builds on previous transcriptome analyses of ZmSOC1_CX in soybean and maize, as well as VcSOC1K_CX in blueberry and maize, which have been reported (Song and Chen, 2018; Han et al., 2021; Song and Han, 2021; Song et al., 2021).

The assembled transcriptome reference comprised 61,563 transcripts corresponding to 20,455 annotated genes. In GmSOC1_CX leaves, 565 differentially expressed transcripts (DETs) associated with 479 differentially expressed genes (DEGs) were identified. Among these DEGs, the two transgenes, GmSOC1 and hpt, showed fold change (FC) of 23,170 [Log₂ ^{FC(transgenic/non-transgenic)} = 14.5] and 7,131 [Log₂ ^{FC(transgenic/non-transgenic)} = 12.8], respectively, confirming the expression of GmSOC1 ( Table 1 ). Additionally, RT-qPCR analysis of six selected DETs yielded results consistent with their RNA-seq data ( Figure 3 ), supporting the reliability of the RNA-seq findings.

Analysis of the 479 DEGs using the GOSlim_Plants ontology file in BiNGO identified 27 overrepresented Gene Ontology (GO) terms. These included 11 terms under “Biological Process”, seven under “Molecular Function”, and nine under “Cellular Component” ( Figure 4 ). These overrepresented GO terms suggest a broad impact of GmSOC1_CX on plant development at the transcript level, contributing to the phenotypic changes observed in GmSOC1_CX plants. For example, six overrepresented GO terms within the “Biological Process” category, namely “reproduction”, “response to abiotic stimulus”, “response to stress”, “post-embryogenic development”, “regulation of gene expression, epigenetic”, and “response to endogenous stimulus”, affect fruit production. These processes are linked through the overrepresented GO terms in the “Molecular Function” category, which regulate “transcription regulator activity” and “catalytic activity” via “binding” ( Figure 4 ).

Further analysis of the 565 DETs for key genes identified 51 key DETs associated with flowering, phytohormones, and sucrose. These were annotated to 39 DEGs and exhibited high similarity to 28 Arabidopsis genes ( Table 1 ). These DETs were annotated to 39 DEGs and showed high similarities to 28 Arabidopsis genes ( Table 1 ). Among the 39 DEGs, GmSOC1 was highly expressed (Log₂ ^{FC(transgenic/non-transgenic)} = 14.5) and suppressed the expression of two major endogenous tomato SOC1 orthologues, SlTM3 (TRINITY_DN7_c0_g1_i17) and SlSTM3 (TRINITY_DN7_c0_g1_i2). In tomato, SlTM3 and SlSTM3 promote the floral transition but serve opposing roles in inflorescence development (Zahn et al., 2023). High expression of SLSTM genes contributes to highly branched inflorescences (Wang et al., 2021).

The transcription factor BEL1-like homeodomain protein 1 (BLH1), which shares high similarity with the MADS-box gene BEL1, was downregulated. BLH family transcription factors are multifunctional, playing critical roles in plant morphogenesis, flower and fruit development, and responses to various environmental factors (Niu and Fu, 2022). Similarly, the polygalacturonase inhibitor precursor (PGIP), which exhibits a high similarity to the MADS-box gene FLOR1, known to promote flowering under long-day conditions (Torti et al., 2012), was also repressed.

The expression of two Histone-lysine N-methyltransferase genes CURLY LEAF (CLF) and ENHANCER OF ZESTE 1 POLYCOMB REPRESSIVE COMPLEX 2 SUBUNIT (EZA1), was enhanced. These genes encode catalytic subunits of the polycomb group (PcG) multiprotein complex. CLF is essential for regulating floral development by repressing the AGAMOUS homeotic gene. It achieves this by forming a nuclear complex with EZA1 and other components, targeting ABA- and glucose-responsive elements (Saleh et al., 2007; Liu et al., 2019; Shu et al., 2020; Liu et al., 2022). Notably, a loss-of-function mutant of Brassica rapa exhibits early flowering (Poza-Viejo et al., 2024), suggesting that the upregulation of CLF and EZA1 may contribute to either no significant impact or delaying flowering ( Figure 5 ).

Of the genes involved in brassinosteroids (BRs) metabolism (Bajguz et al., 2020), the DEGs with repressed expressions included BAS1 and four DWARF genes (DWF1, DWF5, DWF6, and DWF7) ( Table 1 ); these DEG can functionally lead to BRs-deficient changes, which are often associated with decreased BRs that can affect multiple agronomic traits (Bishop, 2003; Li et al., 2016; Zhan et al., 2022). The E5L3R8_SOLLC encodes tomato GLYCOALKALOID METABOLISM1 (GAME1) involved in the steroidal alkaloids (SAs) pathway (Itkin et al., 2011; Averello et al., 2025). The E5L3R8_SOLLC showed decreased expression ( Table 1 ), suggesting a potential lower production of α-tomatine that has impact on plant defense system and fruit quality (Cárdenas et al., 2016; Sonawane et al., 2018; Averello et al., 2025).

For the cytokinin pathway genes, the expression of CKX7 was enhanced, meaning a potential of a reduced cytokinin level in the GmSOC1-CX plants (Köllmer et al., 2014). Similarly, a decreased expression of the KAO2 was found ( Table 1 ), indicating a likely lower level of GA production (Regnault et al., 2014; Shani et al., 2024).

Among the DEGs related to sucrose, the tomato sucrose transporter SlSUT4 (SUC4_ARATH) was upregulated ( Table 1 ). While the downregulation of SlSUT4 has been shown to promote flowering by enhancing sucrose transport to the shoot apex, its overexpression does not significantly impact flowering time or the expression of key genes in the flowering pathway (Liang et al., 2023).

Additionally, seven DEGs of HEAT SHOCK PROTEINs showed upregulation ( Table 1 ), potentially contributing to enhanced tolerance to various abiotic stresses (Ul Haq et al., 2019). Similarly, the increased expression of the GLUTATHIONE S-TRANSFERASE (GST) 23 is associated with both biotic and abiotic resistance (Hernández Estévez and Rodríguez Hernández, 2020). In contrast, the tomato ALCOHOL DEHYDROGENASE (ADH1_EUPLT) and its homolog YFE37, annotated as SDR3B_ARATH, were both repressed. This reduction in expression may negatively impact disease resistance (Xun et al., 2022).

As a key integrator in the plant flowering pathway (Fornara et al., 2010; Lee and Lee, 2010; Immink et al., 2013), enhanced expression of SOC1 or its orthologues, either though overexpression or ectopic expression, can promote flowering. This phenomenon has been reported for many plant species, including tomato (Zahn et al., 2023). Among the five SOC1 and SOC1-like genes identified in tomato ( Figure 1 ), SlTM3 and SlSTM3 play a more significant role in flowering initiation compared to SlMBP23 and SlMBP18, at least in the indeterminate cultivar Moneyberg (Zahn et al., 2023). Additionally, the high expression of SlSTM3 has been linked to a highly branched inflorescence phenotype (Wang et al., 2021).

In this study, phylogenetic analysis revealed that GmSOC1 and ZmSOC1 clustered closely with SlMBP18 and the fifth SlSOC1-like gene, while VcSOC1K showed closer similarity to SlTM3, SlSTM3, and SlMBP23 ( Figure 1 ). Phenotypic observations demonstrated that both ZmSOC1-CX and VcSOC1-CX transgenic plants exhibited earlier flowering compared to non-transgenic controls, with statistically significant differences (P < 0.01) in the T₀ generation and no significant differences (P < 0.01) in the T₁ generation ( Figure 2 ; Supplementary Figure S1A ). In contrast, GmSOC1-CX plants did not show significant changes in flowering time across both T0 and T₁ generations (P < 0.05). The early flowering phenotype observed in ZmSOC1-CX lines suggests that the expression of SlMBP18 and the fifth SOC1-like gene of tomato may play a role in promoting flowering. For the T₁ generation, it would have been valuable to investigate additional traits such as seed germination time and seedling growth, as previous studies have shown that repressed expression of DWARF genes can delay seed germination (Li et al., 2016). Notably, while GmSOC1-CX has been reported to induce early flowering in Arabidopsis (Zhong et al., 2012), its overexpression did not significantly promote flowering in soybean. However, GmSOC1 knock-out mutants exhibited delayed flowering (Kou et al., 2022), highlighting the complex and species-specific regulatory roles of GmSOC1 in flowering time control.

Interestingly, in this study, the T₀ generation of VcSOC1K-CX plants from four transgenic lines exhibited earlier flowering compared to the non-transgenic lines, whereas the T₁ generation of three transgenic lines did not show this early flowering phenotype. Previous studies have reported that overexpression of VcSOC1K in blueberry and VcSOC1K-CX in tobacco promoted flowering (Song et al., 2013; Song and Chen, 2018); however, VcSOC1K-CX did not significantly induce early flowering in maize (Song and Han, 2021). This variation may be attributed to many factors, including differences in expression levels, plant species, and genotype.

At the transcript levels, GmSOC1-CX did not promote flowering, at least not significantly, is likely due to the enhanced expression of two Histone-lysine N-methyltransferase genes, CLF and EZA1 ( Figure 5 ) (Saleh et al., 2007; Liu et al., 2019; Shu et al., 2020; Liu et al., 2022; Poza-Viejo et al., 2024).

Plant architecture, including shoot and inflorescence structure, is an omnigenic trait that can significantly influence tomato productivity (Alonge et al., 2020; Gaarslev et al., 2021). Tomato SOC1 genes, such as SlTM3 and SlSTM3, serve as core regulators of inflorescence structure by interacting with other MADS-box genes (Alonge et al., 2020; Gaarslev et al., 2021; Wang et al., 2021; Zahn et al., 2023). Overexpression of SlTM3 and SlSTM3 enhances inflorescence branching (Wang et al., 2021), while their suppression reduces branching (Alonge et al., 2020; Zahn et al., 2023). In this study, transgenic lines expressing GmSOC1, ZmSOC1, and VcSOC1K exhibited no noticeable changes in inflorescence structure. For GmSOC1-CX lines, this was further supported at the transcript level by the unchanged expression of tomato FRUITFULL1 (FUL1)( Table 1 ), a direct activation target of SlSTM3 (Wang et al., 2021).

Phenotypically, the effect of SOC1 expression on plant height and branching has been less consistent in the literature compared to its well-established role in flowering time. In soybean, soc1 mutants have more internodes than the wild type, but the impact of GmSOC1 overexpression on plant architecture remains unclear (Kou et al., 2022). In Medicago truncatula, MtSOC1 has been shown to influence both flowering and primary stem height in both mutant and overexpression lines (Jaudal et al., 2018). Additionally, ZmSOC1-CX expression has been associated with reduced plant height in transgenic maize and soybean plants (Han et al., 2021; Song et al., 2021). In tomato, SlTM3 and SlSTM3 expression have not displayed any significant impact on plant height and branching (Zahn et al., 2023). In this study, GmSOC1-CX, ZmSOC1-CX, and VcSOC1-CX did not significantly change plant height but enhanced branching for at least the GmSOC1-CX and VcSOC1-CX plants ( Figures 2A ).

At the transcript level, it is noteworthy that the repressed expression of four DWARF genes could theoretically lead to reducing plant size in tomato due to less BRs production, as suggested by previous studies ( Figure 5 ) (Bishop et al., 1999; Bishop, 2003; Montoya et al., 2005; Li et al., 2016; Zhan et al., 2022). The enhanced expression of CKX7 could result in reduced cytokinin levels, which are expected to shorten plant height, increase branching, and reduce flower number (Eckardt, 2003; Bartrina et al., 2011; Köllmer et al., 2014; Waldie and Leyser, 2018). Similarly, a decreased expression of KAO2 might reduce GA production (Regnault et al., 2014), potentially leading to delayed seed germination, stunted plant growth, and delayed flowering. In this study, although GmSOC1-CX did not exhibit phenotypic changes such as plant dwarfing, delayed flowering, or reduced flower number, it did show increased branching ( Figures 1C , Figures 2A , and Supplementary Figure S1D ). However, BR, cytokinin, and GA levels, as well as seed germination timing, were not investigated in this work.

Crop yield-defining traits vary across different crops (Bailey-Serres et al., 2019). For tomatoes, yield-defining traits include both direct factors, such as fruit number, fruit size, and fruit production efficiency per unit area, as well as related traits, including tolerance to abiotic and biotic stresses (Alonge et al., 2020; Bhandari et al., 2023). Accordingly, hormone and flowering pathway genes have become the targets for genetic improvement of yield (Krieger et al., 2010; Ariizumi et al., 2013; Wang et al., 2021; Cui et al., 2022; Kang et al., 2022). In this study, two full-length SOC1 genes from maize and soybean, along with the K-domain of the blueberry SOC1 gene, were constitutively expressed in tomato. The resulting transgenic lines exhibited an increased fruit count, leading to higher total fruit production per plant, suggesting enhanced yield potential. This rise in fruit number was associated with improved branching in the transgenic plants, likely due to decreased BRs resulting from the repression of DWARF genes. Notably, similar effects have not been reported in tomato through the overexpression of SlTM3 and SlSTM3. Interestingly, lower expression levels of GmSOC1 have been shown to enhance soybean yield (Kou et al., 2022).

In addition to its essential role in flowering, SOC1 also plays a role in other processes in Arabidopsis, such as root development and leaf senescence, both of which impact crop yield (Chen et al., 2017; Castañón-Suárez et al., 2024). However, the impact of SOC1 overexpression on root development and leaf senescence in crops remains largely unexplored. Previously, we found that constitutive expression of three SOC1 genes, either full-length or partial, has the potential to enhance yield in maize, soybean, and blueberry, primarily through the regulation of flowering pathway genes, as indicated by RNA-seq data (Song and Chen, 2018; Han et al., 2021; Song and Han, 2021; Song et al., 2021). In this study, unlike our previous findings, the observed increase in fruit production per transgenic plant is attributed to enhanced branching, likely due to the repressed expression of DWARF genes ( Figure 5 ). Additionally, the increased expression of CLF and EZA1 in the RNA-seq data of the GmSOC1-CX lines appears to provide evidence explaining the unchanged flowering time ( Figure 5 ).

In this study, the biotic and abiotic tolerance of the transgenic plants were not directly assessed. However, the identification of several DEGs associated with these traits suggests that the transgenes may have influenced plant resilience to biotic and abiotic stresses ( Figure 4 , Table 1 ). Additionally, fruit quality as well as the other traits may have been impacted by the DEGs.