Introduction

Promoters are collections of cis-regulatory motifs and sequences located upstream of the transcription start site of genes that are recognized by transcription factors involved in transcription initiation (Dynan and Tjian 1985). The isolation and characterization of varied promoters suitable for plant genetic engineering based on their constitutive, tissue-specific, temporal, or inducible expression patterns is an essential step to introduce desirable traits into plants. In particular, promoters with high and constitutive expression are widely used in basic and applied research to control the expression of genes of interest in transgenic plants. However, the same promoter should not be used to drive the expression of separate transgenes in the same plant, as this may lead to transgene inactivation due to silencing (Que and Jorgensen 1998). Promoters from plant viruses, such as the cauliflower mosaic virus (CaMV) 35S promoter and figwort mosaic virus (FMV) promoter, are routinely employed as constitutive promoters in plant research (Govindarajulu et al. 2008). However, they may drive the expression of transgene at levels in the wrong tissue or at the wrong time at a much higher level than the native plant promoters. Exploring efficient constitutive native promoters will be a critical step to develop transgenic plants that express the desired traits.

Soybean (Glycine max (L.) Merr.), which is a good source of protein and oil for human dietary needs, is a valuable agronomic crop worldwide. Soybean also can form a symbiotic relationship with Rhizobium species, producing nodules that can fix atmospheric nitrogen by converting it into ammonia for plant consumption in exchange for sugars (Yang et al. 2021). Despite its agronomic importance, the yield of soybean is relatively low, which has motivated extensive work on molecular breeding to overcome this limitation and produce new and improved transgenic soybean. Therefore, various strong native soybean promoters are required to drive transgene expression. Several promoters have been identified to direct expression in constitutive, tissue-specific, temporal, or inducible manners. For instance, 20 highly expressed soybean genes were identified using transcriptome deep sequencing (RNA-seq) data sets based on their high expression levels in one or more tissues or following a given treatment (Zhang et al. 2015). Their promoter activities were confirmed in lima bean (Phaseolus lunatus) cotyledons and soybean hairy roots. The activity of the promoters from two ELONGATION FACTOR 1A (eEF1A) genes was also detected in stable primary soybean transformants. Of the strong constitutive promoters employed to direct transgene expression in soybean, the polyubiquitin (GmUbi) promoter is the most extensively used (Grant et al. 2017). Several specific promoters have also been reported, such as the HYPERSENSITIVE-INDUCED RESPONSE1 (GmHIR1) promoter expressed specifically in flowers and developing seeds (Koellhoffer et al. 2015), the pathogen-inducible POLYPHENOL OXIDASE12 (GmPPO12) promoter (Chai et al. 2013), the heat- and 1-aminocyclopropane-1-carboxylic acid (ACC)-induced TONOPLAST INTRINSIC PROTEIN2;6 (GmTIP2;6) promoter (Feng et al. 2019), the root-preferential PROLINE-RICH PROTEIN2 (GmPRP2) promoter (Chen et al. 2014), and the abscisic acid (ABA)- and polyethylene glycol (PEG)-induced RESPONSIVE TO DESICCATION26 (GmRD26) promoter (Freitas et al. 2019).

The promoters of reference genes typically used for normalization during quantitative PCR may also be used as constitutive promoters, as they are by definition evenly expressed across multiple tissues, developmental stages, or conditions. For instance, CYCLOPHILIN2 (GmCYP2) and UBIQUITIN-CONJUGATING ENZYME4 (GmUBC4) are stable reference genes expressed in different organs (roots, stems, leaves, flowers, and pods) (Miranda Vde et al. 2013). Similarly, eEF1A and eEF1B are the most stable reference genes across various developmental stages, including the third fully developed trifoliate leaf, full flowering, and developed pods. For spatial and temporal gene expression studies, GmUBC4, GmUBC2, GmCYP2, and ACTIN11 (GmACT11) are considered as reference genes. eEF1A and TUBULIN A5 (GmTUA5) are the most stable reference genes when studying gene expression of soybean roots infected with the roundworm Meloidogyne incognita. Likewise, GmCYP2 and eEF1A are the most stable reference genes in soybean leaves under attack by velvetbean caterpillar (Anticarsia gemmatalis) (Miranda Vde et al. 2013). ELONGATION FACTOR 1B (GmEF1B) and HYPOTHETICAL PROTEIN 2 (UKN2) are the best-suited reference genes in plants during virus infection (Ma et al. 2013), while GmEF1A and GmACT11 are the best reference genes under high-salinity stress. Other genes have also been shown to exhibit stable expression under drought stress (TUBULIN4 [GmTUB4], GmTUA5, and GmEF1A), in the dark (GmACT11 and GmUKN2), or upon high salinity or drought stress and in the dark (GmEF1B and GmUKN2) (Ma et al. 2013). GmEF1B and GmACT11 may be employed as references in O2-depleted environments, such as in flooded plants (Nakayama et al. 2014). However, the activities of these respective promoters have not been tested across the entire vegetative growth period, over multiple development stages, especially during nodule development. At the same time, the coefficient of variation of expression over the multiple tissues has not been evaluated.

In this study, we used bioinformatic analysis methods to search for native soybean promoters exhibiting strong and constitutive expression throughout vegetative growth and development as well as during nodule development. We selected ten candidate promoters and further tested two from the 40S RIBOSOMAL PROTEIN SMALL SUBUNIT S28 (GmRPS28, referred to as RPS28 herein) and EUKARYOTIC TRANSLATION INITIATION FACTOR 1 (GmEIF1, referred to as EIF1 herein) genes. We generated transcriptional fusions between these two promoters and the β-GLUCURONIDASE (uidA) reporter gene and introduced them into soybean, Arabidopsis (Arabidopsis thaliana), and Nicotiana benthamiana for evaluation.

Results

Identifying highly expressed soybean genes from microarray and RNA-seq data

To identify genes with broad and high expression during plant growth and development, and nodule development, we first analyzed the expression levels of soybean genes from published 1072 microarray samples which generated with Affymetrix gene chips (Wu et al. 2019). We retained only those genes with an average value in the top 100 and coefficients of variation (CoV) less than 0.035, indicative of relatively constant expression across all samples (Table S1). To further remove the effect of environmental factors, we then ordered the retained 25 genes from published soybean RNA-sequencing samples (Machado et al. 2020; Table S2) on the basis of their minimum transcript per million (TPM) (mini TPM > 50) and CoV (CoV < 0.4) and kept the top 10 genes (Table 1). Among the top 10 genes, the ones which stably expressed under stress and have not been reported were selected as candidates. Gene 1 (Glyma.19G052400) and gene 6 (Glyma.15G050200) have been published previously (Yim et al. 2015; Bansal et al. 2015). Gene 2 (Glyma.19G172200) was upregulated by GmbZIPE2 overexpression. Gene 3 (Glyma.17G103100) and gene 7 (Glyma.05G024200) were reported to be induced by the soybean mosaic virus G7 strain (Chen et al. 2017). Gene 5 (Glyma.07G115900) encodes a nucleoside diphosphate kinase, which was reported to be regulated by NaCl-CaCl2 and NaCl-LaCl3 treatment (Yin et al. 2015). Gene 9 (Glyma.08G126200) was identified as a target gene of gma-miR5037c which is involved in salt stress and phosphate starvation (Ning et al. 2019; Tripathy et al. 2021). Gene 10 (Glyma.13G290900) was reported to be inhibited by drought stress (Wang et al. 2019). Finally, Glyma.19G203300 (hereafter as RPS28) and Glyma.06G269600 (hereafter EIF1) were selected for further characterization (Fig. S1A and B).

Table 1 Summary of the top 10 genes with high expression levels
Full size table

The RPS28 and EIF1 promoters contain multiple core and cis-acting elements

We next analyzed the gene structures of RPS28 and EIF1 (Fig. 1A). We found that the RPS28 locus consists of a single exon separated from the 5' UTR by one large intron, while the EIF1 locus comprises four introns, one of which was located in the 5' UTR. We then turned to the identification of the core promoter elements TATA-box and CAAT-box in the region upstream of the translational start site ATG using the databases PLACE (Higo et al. 1999) and PlantCARE (Lescot et al. 2002). We detected ten putative TATA-boxes and 17 putative CAAT-boxes in the RPS28 promoter (Fig. 1B), and 12 putative TATA-boxes and 13 putative CAAT-boxes in the EIF1 promoter (Fig. 1C). In addition, we observed additional putative cis-acting elements within the promoters, including the CGTCA motif, CAT-box, ABA-responsive element (ABRE), GT1 motif, G-box, and TCTCCCT motif in the RPS28 promoter (Fig. 1B) and the ABRE, W-box, wound-responsive cis-element (WRE3), ethylene responsive element (ERE), and dehydration-responsive element (DRE) core in the EIF1 promoter (Fig. 1C). These results suggested that the RPS28 and EIF1 promoters harbor both core and inducible promoter elements. To examine whether their expression levels were affected by these putative inducible cis-elements, we checked their expression levels under cold, drought, and salicylic acid (SA) treatment from a published soybean RNA-Seq database (Yu et al. 2022). The expression levels of these genes did not change significantly under different stress conditions (Fig. S2A–F; Table S3). Furthermore, RPS28 and EIF1 showed almost the same expression level compared with GmUbi (Fig. S2A–F). These results suggest that RPS28 and EIF1 are constitutively expressed genes under stress conditions, though there are some stress responsive cis-regulatory elements in their promoters.

Fig. 1
figure 1

Promoter analysis of RPS28 and EIF1. A Gene structures of RPS28 and EIF1. B and C Promoter analysis of RPS28 (B) and EIF1 (C). The cis-acting elements are indicated as red (TATA-box) or black (CAAT-box) boxes or underlined in red (other cis-acting elements)

Full size image

The RPS28 and EIF1 promoters drive high GUS gene expression at different growth and development stages in soybean

We next examined the tissue specificity of the RPS28 and EIF1 promoters in transgenic soybean plants. We first placed the b-GLUCURONIDASE (GUS) reporter gene under the control of the RPS28 or EIF1 promoters (1500 bp and 1472 bp, respectively) (Fig. 2A). For each promoter, we generated two reporter constructs: one consisting of the promoter and 5' UTR (RPS28pro and EIF1pro), and one also harboring the first intron (RPS28-Ipro and EIF1-Ipro), as studies have shown that the first intron also may influence gene expression (Le Hir et al. 2003). We also constructed another reporter containing the well-known constitutive promoter GmUbi with its intron in the 5' UTR as a positive control (Fig. 2A; Grant et al. 2017). We transformed Ws82 with the five constructs via Agrobacterium-mediated cotyledonary-node transformation and then analyzed their promoter activity in T2 stable transgenic lines.

Fig. 2
figure 2

The RPS28 and EIF1 promoters drive GUS expression at different stages of soybean transformation. A Schematic representation of the binary vectors used for transformation. RB right border, LB left border, Bar phosphinothricin N-acetyltransferase, TNOS, nopaline synthase terminator. RPS28pro, RPS28 promoter and 5′ UTR; RPS28-Ipro, RPS28 promoter with 5′ UTR and the first intron; EIF1pro EIF1 promoter and 5′ UTR; EIF1-Ipro EIF1 promoter with 5′ UTR and first intron; GmUbipro, Ubiquitin promoter with 5′ UTR and the first intron (positive control). B–E RPS28 and EIF1 promoters support GUS expression during co-culture (B), 2 (C) and 4 weeks (D) after bud induction, and during bud elongation (E). In B and C, Ws82 cotyledon without infection was used as negative control. Scale bars = 1 cm

Full size image

We detected GUS enzyme activity by histochemical staining for all constructs during co-culture, bud induction, and bud elongation stages during transformation (Fig. 2B–E), indicating that these promoters may be useful to support transgene expression during stable transformation. To test if these constructs had any adverse effects on plant growth, we compared the growth phenotypes of stable transgenic lines with the wild-type Ws82 and found that all transgenic lines were comparable to the Ws82, indicating that GUS expression did not affect plant growth or development (Fig. S3A, B). In soybean transformants, we observed GUS activity derived from all constructs throughout the 5-day-old seedlings (Fig. 3A, B), and in the unifoliate leaf, trifoliate leaf, buds, petioles, internodes, roots, and nodules of 15-day-old plants (Fig. 3C, D, and Fig. S3C, D). We also detected GUS activity in reproductive tissues, such as young pods as well as pods and seeds at the seed-filling stage (Fig. 3E, and Fig. S3E, F). In addition, we quantified GUS activity with 4-methylumbelliferyl (4-MU) in the root, cotyledon, unifoliate leaf, and trifoliate leaf of 15-day-old transgenic soybean plants. Except RPS28-Ipro, RPS28pro, EIF1-Ipro, and EIF1pro showed almost similar GUS activity compared with GmUbipro in the root, cotyledon, unifoliate leaf, and trifoliate leaf (Fig. 3F). We concluded that the RPS28 and EIF1 promoters can confer constitutive highly and stably expression of transgenes in soybean and support transgene expression during transformation, growth and development, and in nodules with different expression levels.

Fig. 3
figure 3

The RPS28 and EIF1 promoters drive GUS gene expression in transgenic soybean plants. A The 5-day-old seedlings. Scale bars = 1 cm. B Cotyledon with buds excised from the same 5-day-old seedlings shown in (A). Scale bars = 1 mm. C 15-day old seedlings. Scale bars = 10 cm. D Roots and nodules collected from 15 DPI. Scale bars = 0.5 mm. E Seeds at the filling stage. Scale bars = 1 mm. F Quantitative fluorimetric GUS assay in root, cotyledon, unifoliate, and trifoliate leaf of 15-day-old seedlings. Data are means ± standard deviation (SD)

Full size image

The RPS28 and EIF1 promoters drive GUS gene expressed in Arabidopsis and Nicotiana benthamiana

We then tested whether the RPS28 and EIF1 promoters might also function in other species. To this end, we introduced the RPS28pro:GUS, RPS28-Ipro:GUS, EIF1pro:GUS, and EIF1-Ipro:GUS reporters into the model plants Arabidopsis and N. benthamiana. We observed GUS activity in almost all Arabidopsis tissues or organs, including seedlings, flowers, and pods (Fig. 4A–D). Likewise, we detected GUS activity in N. benthamiana leaves transiently transformed with the reporter constructs (Fig. 4E). Together, these results indicate that the RPS28 and EIF1 promoters can drive transgene expression in a variety of plant species.

Fig. 4
figure 4

Representative GUS histochemical staining assay of transgenic Arabidopsis plants and Nicotiana benthamiana leaves driven by the RPS28 and EIF1 promoters. A, B RPS28 promoter with the 5′ UTR and first intron (A) or 5′ UTR alone (B). C, D EIF1 promoter with the 5′ UTR and first intron (C) or 5′ UTR alone (D). E GUS staining of the RPS28 and EIF1 promoters in Nicotiana benthamiana leaves. In A–D, GUS activity was tested in 7-day-old seedlings, flowers, and pods. In A–D, Scale bars = 2 mm. In E, Scale bars = 1 cm

Full size image

Discussion

Here, we reported two constitutive endogenous promoters derived from RPS28 and EIF1. First, RPS28 and EIF1 were expressed in almost all soybean organs, as well as during nodule development, with a low coefficient of variation. Second, the RPS28 and EIF1 promoters exhibited high expression in stable soybean transformants. Third, the RPS28 and EIF1 promoters were both active in Arabidopsis and N. benthamiana.

RPS28 and EIF1 were not only constitutively expressed in different tissues and developmental stages and in nodules, but also under different conditions. Indeed, both genes were highly and stably expressed across 1072 different microarray datasets that covered 45 different genotypes and 18 tissues, as well as different biotic and abiotic stresses, such as rust disease, drought, aluminum stress, iron treatment, ABA treatment, and SA treatment (Wu et al. 2019). Although the presence of important cis-regulatory elements in the RPS28 and EIF1 promoters, they are highly expressed under different conditions. For instance, the RPS28 promoter harbors ABRE element which is involved in ABA signaling (Yoshida et al. 2010); in the EIF1 promoter, besides ABRE, we identified W-box motifs, which are bound by WRKY transcription factors and are related to drought stress (Li et al. 2020), and DRE elements, which activate transcription in response to both low temperature and water deficit (Knight et al. 1999). We found that RPS28 and EIF1 constitutively expressed under cold, drought, and water deficit conditions.

The inclusion of introns in transgenes can dramatically affect the expression outcome (Le Hir et al. 2003). Although a GmUbi promoter construct without an intron still drives transgene expression to levels over two times greater than the CaMV 35S promoter, the intron in the 5' UTR of the GmUbi promoter contributes to the highest activity of this promoter (Chiera et al. 2007; Grant et al. 2017). Similarly, the first intron of soybean eEF1A also is essential for the high activity of this native promoter (Zhang et al. 2016). Based on our analysis of the RPS28 and EIF1 loci, the 5′ UTR of RPS28 is followed by a large intron, while the EIF1 5′ UTR is interrupted by a large intron. Notably, the omission of the intron from the RPS28 promoter construct resulted in enhanced expression in root, cotyledon, unifoliate leaf, and trifoliate leaf. RPS28 was independently screened as being highly expressed across multiple tissues, but it showed a low expression level in the latter evaluation (Zhang et al. 2015). In the paper of Zhang et al., the first intron of RPS28 was included in the promoter sequence, which is similar to RPS28-Ipro in our study. The promoter activity of RPS28 was evaluated in lima bean cotyledons and soybean hairy root (Zhang et al. 2015). In our study, the expression level of RPS28-Ipro is the lowest in cotyledon and root, which is consistent with Zhang et al. The first intron may be the reason for the low activity of RPS28 promoter in their paper. However, the presence of the RPS28 leader intron is pivotal under some conditions, as we observed much stronger activity from the RPS28-Ipro reporter construct in N. benthamiana leaves relative to the RPS28pro construct.

In summary, we described the two highly expressed constitutive RPS28 and EIF1 promoters in soybean. Both promoters drove strong and constant expression during plant growth and development and nodule development, and under several biotic and abiotic stress conditions, and should serve as efficient promoters to generate transgenic plants with stable quality and yield. The two promoters have at least two advantages compared with GmUbi. First, the expressions of RPS28 and EIF1 are more stable than GmUbi indicated by the Coefficients of Variation (CoV) from 1072 microarray samples (table S1). Second, different promoters are necessary to drive the expression of multiple genes of interest in the same plant, as using a single promoter multiple times in one plant may lead to transgene silencing. Thus, the RPS28 and EIF1 promoters can be used as candidate promoters when several genes were constructed in one construct.

Materials and methods

Plant materials and growth conditions

The soybean ecotype ‘Williams 82’ and transgenic plants were grown in a greenhouse at 28 °C, under a 12 h light/12 h dark photoperiod. For nodulation assays, seeds were surface-sterilized with chlorine gas for 16–20 h and sown in wet vermiculite. After ~ 8 days, seedlings were inoculated with 2 mL of the Rhizobium strain Bradyrhizobium diazoefficiens USDA110 at an OD600 of ~ 0.1. Arabidopsis seedlings were grown on half-strength Murashige and Skoog (MS) medium for 7–8 days and then transferred to soil at 23 °C under a 16 h light/8 h dark photoperiod. Nicotiana benthamiana plants were grown at 26 °C under a 16 h light/8 h dark photoperiod.

Gene structure and promoter analysis

The gene structures of RPS28 and EIF1 were deduced using the Gene Structure Display Server (GSDS2.0, http://gsds.cbi.pku.edu.cn/index.php). cis-Acting elements were predicted using the PLACE (http://www.dna.affrc.go.jp/PLACE/) and PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) online databases.

Vector construction

Upstream regulatory sequences of the PRS28 and EIF1 genes are downloaded from Phytozome (Goodstein et al. 2012). The respective promoter fragments from genomic DNA of the soybean ecotype ‘Ws82’ were then PCR-amplified with specific primers (Table S4). Specifically, the PCR amplicon for the RPS28 promoter (RPS28pro) included 1500 bp of promoter and 82 bp of 5′ untranslated region (5′ UTR); a promoter fragment with the first intron of RPS28, for a total amplicon size of 2357 bp (RPS28-Ipro), was also generated. The EIF1 promoter was amplified as a 1640-bp amplicon comprising 1472 bp of promoter and 168 bp of 5′ UTR (EIF1pro) or with an additional 730 bp corresponding to the first intron (EIF1-Ipro). All PCR amplicons were cloned upstream of uidA in the vector pCAMBIA-1391Z harboring the phosphinothricin acetyltransferase (bar) gene as selection marker. Similarly, the GmUbi promoter (Glyma.20G141600) was cloned as a 917-bp amplicon covering the promoter, 5′ UTR, and first intron upstream of uidA in pCAMBIA-1391Z vector as positive control. For transformation in Arabidopsis and Nicotiana benthamiana, the original selection marker in pCAMBIA-1391Z, Hygromycin B phosphotransferase (HygR), was used.

Soybean stable transformation

Stable transformation of soybean was performed following the previously reported cotyledonary-node transformation method (Luth et al. 2015).

Arabidopsis transformation assay and transient expression assays in Nicotiana benthamiana leaves

The corresponding constructs were introduced into Agrobacterium (Agrobacterium tumefaciens) strain GV3101 and cultured on solid medium containing 100 μg/mL kanamycin. Single colonies were grown in liquid medium with 100 μg/mL kanamycin at 28 °C for 2 days. After centrifugation, the cell pellets were resuspended to a final OD600 ≈1.0 in transformation buffer (10 mM MES pH 5.7, 10 mM MgCl2, and 200 μM acetosyringone). For Arabidopsis transformation, the prepared strain liquid was used to transform plants by the floral-dip method (Clough and Bent 1998). Transgenic plants were screened on 0.5xMS plates supplemented with 35 mg/L Hygromycin. For N. benthamiana transient expression assay, the prepared strain liquid was infiltrated into young N. benthamiana leaves. After an incubation for 24 h in the dark, N. benthamiana plants were allowed to grow for another 36–48 h under a 16 h light/8 h dark photoperiod, and then, the leaves were collected and stained for GUS activity.

Histochemical and fluorimetric GUS assay

For histochemical GUS assays, samples were soaked in 90% (v/v) acetone for 30 min for fixation and incubated in 1 mg/mL 5-bromo-4-chloro-3-indolyl-β-D-glucuronic acid, cyclohexylammonium salt (X-Gluc) solution overnight at 37 °C in the dark. After removal of the X-Gluc solution, samples were destained in 70% (v/v) ethanol. Photographs were taken using a digital camera or stereo light microscope (Zeiss V20). For fluorimetric GUS assays, GUS activity in root, cotyledon, unifoliate, and trifoliate leaf of 15-day-old seedlings was measured as previously described (Freitas et al. 2019).