The functions of WRINKLED1 (WRI1) transcription factor in regulating fatty acid (FA) biosynthesis are highly conserved in crop plants, including maize (Zea mays ; Pouvreau et al. , 2011), rapeseed (Brassica napus ; Li et al. , 2015), oil palm (Elasis guineensis ; Ma et al. , 2013), camelina (Camelina sativa ; An et al. , 2017) and soybean (Glycine max ; Chen et al. , 2018; Chen et al. , 2020; Zhang et al. , 2017). Recently, Kong et al. (2017) found that the WRI1 plays a role in root auxin homeostasis and affects root development in Arabidopsis , suggesting its involvement in root architecture and potentially shoot architecture as well. In soybean, there are two WRI1 homologs (GmWRI1a and GmWRI1b ), and the expression levels of GmWRI1a showed a positive correlation with the seed oil content (SOC; Zhang et al. , 2017), whereas such correlation between GmWRI1b and SOC remains unknown. The ‘RY(CATGCA)’ cis‐element, to which the GmABI3a (an ortholog of Arabidopsis ABSCISIC ACID INSENSITIVE3) protein directly binds, is absent in the GmWRI1b promoter, resulting in low GmWRI1b expression levels and consequently only basal function of the GmWRI1b in soybean (Zhang et al. , 2017). Chen et al. (2018) reported that overexpression of GmWRI1a increased the SOC and FA content, and altered the FA composition in soybean. However, the detailed functions of GmWRI1b in soybean remain elusive. Recently, individual overexpression of GmWRI1a or GmWRI1b gene in soybean hairy roots altered phospholipid and galactolipid syntheses, soluble sugar and starch contents in nodules derived from transgenic hairy roots, and exhibited an increase in nodule number (Chen et al. , 2020). On the other hand, knockdown of both GmWRI1a and GmWRI1b genes in soybean hairy roots interfered with glycolysis and lipid biosynthesis in developed nodules, and resulted in a decrease in nodule number (Chen et al. , 2020). Although Chen et al. (2020) explored the overexpression of the GmWRI1 genes in hairy roots for functional studies, their results suggest that the GmWRI1 genes may have other pleiotropic functions in addition to regulation of genes related to FA synthesis in soybean. Furthermore, the previous work only looked at the hairy root system; and therefore, it could not monitor the phenotypic outcomes at whole‐plant level (Chen et al. , 2020).

Here, we obtained three stable transgenic soybean lines overexpressing GmWRI1b (GmWRI1b‐OX ) (Figure 1a), and examined the agronomic traits of the T4‐generation homozygous GmWRI1b‐OX plants grown under field conditions at three types of interval distances (10, 30 and 40 cm) between two plants within a row in the Hanchuan Transgenic Biosafety Station in 2019. All three GmWRI1b‐OX lines showed improved agronomic performance (Figure 1b), including decreases in plant height (at 10‐ and 30‐cm plant distances) and internode length (Figure 1c–d), but increases in node number, branch number, stem diameter, shoot dry weight, pod number per plant, seed number per plant, yield per plant and yield per ha at three plant density levels, when compared with wild‐type (WT) plants (Figure 1e–l). In addition, the SOCs were found to be higher in GmWRI1b‐OX than in WT plants at all three types of plant distances (Figure 1m). As a result, the total seed oil production per plant increased by 41.3% to 54.8% at 40‐cm, 39.3% to 60.2% at 30‐cm, and 63.8% to 93.2% at 10‐cm distances (Figure 1n). The seed protein contents and 100‐seed weights of three GmWRI1b‐OX lines and WT were found to be comparable (Figure 1o, p). These data collectively indicate that GmWRI1b is a promising gene, which can be used to alter plant architecture, thereby improving yield, and increase SOC in soybean and perhaps in other crops.

Previously, Chen et al. (2018) reported that overexpression of GmWRI1a also increased yield per plant, which was resulted from the larger seed size, not from the changes in node number, pod number per plant and seed number per plant. In the present study, overexpression of GmWRI1b in soybean resulted in the increase in yield per plant in GmWRI1b‐OX lines (Figure 1k), which was the result of improvement of plant architecture, including the increase in the pod number per plant (Figure 1i) and seed number per plant (Figure 1 j), but not seed size (Figure 1p). This finding was not reported by any previous study on any GmWRI1 genes. It would be then interesting to investigate whether the GmWRI1a also plays a role in the regulation of soybean architecture. It should also be mentioned that in this study, overexpression of GmWRI1b elevated SOC but did not impact total protein content (Figure 1m and o). Recently, Manan et al. (2017) reported that ectopic expression of the G. max LEAFY COTYlEDON2a (GmLEC2a ), which has a role in up‐regulating the expression of GmWRI1 in soybean, increased not only the SOC but also the protein content in transgenic Arabidopsis seeds. Thus, the relationship between protein and oil contents in soybean deserves detailed investigations on a case‐by‐case basis.

Plant architecture is one of the important factors for the development of high‐yield cultivars. Previously, ectopic expression of the Arabidopsis CYP714A2 in rice caused semi‐dwarfism with moderately decreased plant height, more yielding tillers and higher grain yield in comparison with WT plants, which was due to gibberellic acid (GA) deactivation (Zhang et al. , 2011). We found that the expression levels of a soybean CYP714A gene (GmCYP714A /Glyma.18G218500 ) significantly increased in the OX6 and OX8 plants (Figure 1q). Previous studies have reported that the GmWRI1s bind to the AW‐box ‘CnTnG(n)7CG’ in the promoters of genes in soybean (Chen et al. , 2020; Chen et al. , 2018). We next employed electrophoretic mobility shift assay (EMSA) and transactivation assay to validate whether GmWRI1b would directly regulate the expression of GmCYP714A . Results of these assays convincingly indicated that the GmWRI1b could directly bind to the AW‐box motif identified in the GmCYP714A promoter (Figure 1r), and activate its expression (Figure 1s). We also found that the endogenous GA₃ and GA₄ levels were significantly reduced in the GmWRI1b‐OX6 line, while the GA₁ and GA₇ levels were comparable in OX6 and WT plants (Figure 1t). The semi‐dwarf phenotype of GmWRI1b‐OX6 plants was rescued by GA₃ application (Figure 1u), suggesting the involvement of GmWRI1b in GA catabolism, probably through regulating the expression of its downstream gene GmCYP714A . These results demonstrated that overexpression of GmWRI1b improved plant architecture and yield per plant in soybean under field conditions by altering GA metabolism. Our findings also provide a strategy and opportunity for stably increasing yield and SOC in soybean using a single gene.

WRINKLED1 (WRI1) 转录因子在调节脂肪酸 (FA) 生物合成中的功能在作物中高度保守，包括玉米 ( Zea mays ; Pouvreau et al. , 2011 )、油菜 ( Brassica napus ; Li et al. , 2015 ) 、油棕（ Elasis guineensis ；Ma et al. ， 2013 ）、亚麻荠（ Camelina sativa ；An et al. ， 2017 ）和大豆（ Glycine max ；Chen et al. ， 2018 ；Chen et al. ， 2020 ；Zhang et al. ） ， 2017 ）。最近，孔等人。 ( 2017 ) 发现WRI1在拟南芥根部生长素稳态中发挥作用，并影响根部发育，表明它也参与根结构和潜在的芽结构。在大豆中，存在两个WRI1同源物（ GmWRI1a和GmWRI1b ）， GmWRI1a的表达水平与种子含油量呈正相关（SOC；Zhang et al. , 2017 ），而GmWRI1b和SOC之间的这种相关性仍然未知。 GmWRI1b启动子中不存在 GmABI3a（拟南芥脱色酸 INSENSITIVE3 的直系同源物）蛋白直接结合的“RY(CATGCA)”顺式元件，导致GmWRI1b表达水平低，因此仅具有 GmWRI1b 的基础功能。大豆（Zhang et al. ， 2017 ）。陈等人。 ( 2018 ) 报道， GmWRI1a的过度表达增加了大豆中的 SOC 和 FA 含量，并改变了 FA 组成。然而， GmWRI1b在大豆中的详细功能仍不清楚。最近，大豆毛状根中GmWRI1a或GmWRI1b基因的个体过表达改变了转基因毛状根根瘤中磷脂和半乳糖脂的合成、可溶性糖和淀粉含量，并表现出根瘤数量的增加(Chen等， 2020 )。另一方面，大豆毛状根中GmWRI1a和GmWRI1b基因的敲除会干扰发育根瘤中的糖酵解和脂质生物合成，导致根瘤数量减少(Chen et al. , 2020 )。尽管陈等人。 ( 2020 )探索了GmWRI1基因在毛状根中的过度表达进行功能研究，他们的结果表明GmWRI1基因除了调控大豆FA合成相关基因外，可能还具有其他多效性功能。此外，之前的工作只关注毛状根系统；因此，它无法监测全株水平的表型结果（Chen et al. , 2020 ）。

在这里，我们获得了三个过表达GmWRI1b （ GmWRI1b-OX ）的稳定转基因大豆品系（图1a），并检查了在三种间隔距离（10、30 2019 年汉川转基因生物安全站，一排内两株植物之间的距离为 40 厘米。所有三个GmWRI1b-OX品系均显示出农艺性能的改善（图 1b），包括株高降低（株距为 10 厘米和 30 厘米）和节间长度（图1c-d），但在三个植物密度水平下，节数、分枝数、茎直径、芽干重、每株荚数、每株种子数、每株产量和每公顷产量增加，当与野生型（WT）植物相比（图1e-l）。此外，在所有三种植物距离下，GmWRI1b-OX中的 SOC 均高于 WT 植物（图 1m）。结果，每株植物的种子油总产量在 40 厘米距离处增加了 41.3% 至 54.8%，在 30 厘米距离处增加了 39.3% 至 60.2%，在 10 厘米距离处增加了 63.8% 至 93.2%（图 1n）。三个GmWRI1b-OX品系和 WT 的种子蛋白质含量和 100 粒种子重量具有可比性（图 1o，p）。这些数据共同表明， GmWRI1b是一个有前途的基因，可用于改变植物结构，从而提高产量，并增加大豆和其他作物的 SOC。

此前，陈等人。 ( 2018 ) 报道， GmWRI1a的过度表达也增加了单株产量，这是由于种子尺寸较大，而不是节数、单株荚数和单株种子数的变化。在本研究中，大豆中GmWRI1b的过度表达导致GmWRI1b ‐ OX品系的单株产量增加（图 1k），这是植物结构改善的结果，包括单株荚数的增加（图 1i） ）和每株植物的种子数量（图 1 j），但不是种子大小（图 1p）。之前任何关于GmWRI1基因的研究都没有报道过这一发现。那么研究GmWRI1a是否也在大豆结构的调节中发挥作用将会很有趣。还应该提到的是，在本研究中， GmWRI1b的过度表达升高了 SOC，但不影响总蛋白含量（图 1m 和 o）。最近，马南等人。 ( 2017 ) 报道， G. max LEAFY COTYlEDON2a ( GmLEC2a ) 的异位表达在大豆中上调GmWRI1的表达，不仅增加了转基因拟南芥种子的 SOC，还增加了蛋白质含量。因此，大豆中蛋白质和油脂含量之间的关系值得根据具体情况进行详细研究。

株型是培育高产品种的重要因素之一。此前，拟南芥CYP714A2在水稻中的异位表达导致了半矮化，与野生型植物相比，株高适度降低，分蘖较多，籽粒产量较高，这是由于赤霉酸（GA）失活所致（Zhang et al. , 2011） ）。我们发现 OX6 和 OX8 植物中大豆CYP714A基因 ( GmCYP714A / Glyma.18G218500 ) 的表达水平显着增加（图 1q）。先前的研究报道，GmWRI1 与大豆基因启动子中的 AW 盒“CnTnG(n)7CG”结合（Chen等， 2020 ；Chen等， 2018 ）。接下来，我们采用电泳迁移率变动分析（EMSA）和反式激活分析来验证 GmWRI1b 是否会直接调节GmCYP714A的表达。这些测定的结果令人信服地表明，GmWRI1b 可以直接与GmCYP714A启动子中识别的 AW-box 基序结合（图 1r），并激活其表达（图 1s）。我们还发现， GmWRI1b - OX6系中的内源 GA ₃和 GA ₄水平显着降低，而 OX6 和 WT 植物中的 GA ₁和 GA ₇水平相当（图 1t）。 GmWRI1b - OX6植物的半矮化表型通过 GA ₃应用得以挽救（图 1u），表明 GmWRI1b 可能通过调节其下游基因GmCYP714A的表达参与 GA 分解代谢。这些结果表明， GmWRI1b的过度表达通过改变 GA 代谢，改善了田间条件下大豆的植物结构和单株产量。我们的研究结果还提供了使用单基因稳定提高大豆产量和 SOC 的策略和机会。