A Soybean Pyrroline-5-Carboxylate Dehydrogenase GmP5CDH1 Modulates Plant Growth and Proline Sensitivity

Abstract

Soybean [Glycine max (L.) Merr.], as a globally commercialized crop, is an important source of protein and oil for both humans and livestock. With more frequent extreme weather disasters, abiotic stress has become one of the critical factors restricting soybean production. Proline (Pro) is a well-known substance in plants that responds to abiotic stress. To identify potential effector genes involved in soybean resistance to abiotic stress, we focused on the pyrroline-5-carboxylate dehydrogenase (P5CDH) which is a key enzyme in the degradation process of Pro. Through homologous sequence alignment, phylogenetic tree, and predicted expression, we chose GmP5CDH1 (Glyma.05G029200) for further research. Tissue-specific expression assay showed that GmP5CDH1 had higher expression levels in soybean seed and cotyledon development. Subcellular localization assay revealed that GmP5CDH1 was a nuclear-membrane-localized protein. As the result of the predicted cis-acting regulatory element indicates, the expression level of GmP5CDH1 was induced by low temperature, drought, salt stress, and ABA in soybean. Next, we constructed transgenic Arabidopsis overexpressing GmP5CDH1. The results showed that GmP5CDH1 also strongly responded to exogenous Pro, and overcame the toxicity of abiotic stress on plants by regulating the endogenous concentration of Pro. The interaction between GmP5CDH1 and GmSAM1 was validated through yeast two-hybrid, LUC fluorescence complementary, and BIFC. In conclusion, overexpression of a soybean pyrroline-5-carboxylate dehydrogenase GmP5CDH1 regulates the development of Arabidopsis thaliana by altering proline content dynamically under salt stress, especially improving the growth of plants under exogenous Pro.

1. Introduction

Oil and protein are essential nutrients for humans and livestock. Greater than 70% of edible oil and half of feed protein come from plants. Soybean provides nearly 60% of global oilseed production and over 25% of global food including animal feed protein consumption, making it a leading commercial crop for vegetable oil and protein worldwide [1]. By 2050, we need to double the grain production to meet the growing population; however, global soybean production is far below the required level [2]. With continuous rapid climate change and expanding soil pollution, it may be more difficult to obtain sufficient yields. At present, abiotic stress has become one of the major limiting factors in crop production [3,4]. It is necessary to consider the ongoing challenges of environmental pressures such as drought, floods, extreme temperatures, frost, and heavy metal pollution to increase soybean production [5]. However, the heritability of abiotic stress-related traits is relatively low and is greatly influenced by environmental conditions [6]. Exploring effective abiotic stress resistance genes in soybean has still been faced with serious challenges.

Pro is one of the substances with the greatest osmotic regulation effect in plants, first discovered in 1954 from withered detached leaves of ryegrass [7]. Pro can not only accumulate in plants to resist abiotic stress, but also plays a role in osmotic regulation in microorganisms and invertebrates in the ocean by affecting Pro content [8]. As an osmotic regulator in plants, it has good affinity with water and can maintain the balance of cell permeability and structural stability [9,10]. Under a hostile environment, Pro has the function of clearing reactive oxygen species by synergizing with the antioxidant system in plants, or chelating singlet oxygen and hydroxyl radicals to maintain the balance of reactive oxygen species [11]. The unique structure and properties of Pro (cyclic sub-amino acids, extremely hydrophilic) can interact with proteins to form hydrophobic skeletons, serving as molecular chaperones to protect protein integrity and enzyme activity [12,13]. In addition, due to the increase in Pro synthesis caused by environmental stress, a large amount of NADPH produced in plants is used to synthesize NADP⁺, maintain electron flow centered on photoreaction, stabilize redox states, and protect the thylakoid membrane from light damage caused by free radicals [14]. Additionally, other researchers have found that Pro metabolism is closely involved in plant growth, especially in seed development and flowering. The accumulation of Pro in plants is an important signal for regulating plant growth and development [15,16].

The biosynthesis of Pro in higher plants mainly occurs through two pathways: glutamate and ornithine [8]. The glutamate pathway mainly occurs in the cytoplasm and chloroplasts [16], catalyzed by two enzymes: Δ1-pyrroline-5-carboxylate synthase (P5CS) and pyrroline-5-carboxylate reductase (P5CR). The ornithine pathway uses ornithine as a precursor substance, and related reactions mainly occur in mitochondria. Ornithine and α- Ketoglutarate (KG) are reduced to L-glutamate γ-semi-aldehyde (GSA) by ornithine-δ-aminotransferase (δOAT). GSA spontaneously cyclizes into pyroline-5-carboxylate (P5C) in chloroplasts and enters the cytoplasm, generating Pro under the catalysis of P5CR [17]. The degradation process of Pro is essentially a reversal of its synthesis process. In eukaryotes, the degradation of Pro mainly occurs in mitochondria [18], so the synthesis and degradation of Pro in plants occur in different spaces. Pro is first oxidized to P5C by Pro dehydrogenase (ProDH) and then oxidized to glutamic acid under the action of pyrroline-5-carboxylate dehydrogenase (P5CDH) [19]. This process not only maintains the balance of amino acids within plant cells but also facilitates the provision of sufficient nitrogen sources and energy for plant metabolism, growth, and development. Therefore, ProDH and P5CDH are two key enzymes involved in Pro degradation in higher plants.

Forlani et al. [20] found two P5CDH homologous expressions in tobacco-cultured cells after salt treatment. NAD⁺ or NADP⁺ can serve as substrates for P5CDH, and it has a high affinity for NAD⁺. Deuschle et al. [21] isolated and identified T-DNA insertion mutants in AtP5CDH to investigate the physiological function of P5CDH. The degradation of Pro was not detected in the p5cdh mutant, but no changes in growth phenotype were observed. This suggests that the AtP5CDH gene is not essential for plant nutritional growth, but is necessary for the degradation of Pro. Additionally, the gene is sensitive to Pro, intermediate products P5C, and other substances that produce P5C such as arginine and ornithine. The accumulation of P5C may be a trigger for inducing Pro degradation. Rizzi et al. [22] used p5cdh mutants as materials to investigate how P5CDH affects the activation of Pro synthesis in Arabidopsis tissue. They found that when exogenous Pro stress was stopped, the Pro content in the mutant did not decrease but increased, indicating that the p5cdh mutant activated the ornithine pathway. The most recent study found that P5CDH-mediated Pro catabolism in mobilizing carbon and nitrogen during seed development in Arabidopsis led to lower seed yield and lighter seeds of the p5cdh mutant [23]. In other plants, treating alfalfa with saline–alkali, the expression of MsP5CSs, MsP5CRs, MsOATs, and MsProTs was significantly upregulated, while the expression of MsPDH1.1, MsPDH1.3, and MsP5CDH was significantly downregulated. This indicated that genes related to proline metabolism play an important role in the salt–alkali stress tolerance of alfalfa [24]. Researchers also identified a new germplasm of alfalfa which has a greater growth performance under field drought conditions through overexpression of P5CDH from Cleistogenes songorica [25].

So far, there have been preliminary studies on proline and its related genes in metabolic pathways in many species. However, researchers have a limited understanding of their impact on soybean growth, development, and stress response, which could be related to the significant environmental influence on abiotic stress resistance. Here, we identified the P5CDH family genes in soybean using homologous sequence alignment and analyzed them through evolutionary relationships, prediction expression, and gene structure. For further research, we chose GmP5CDH1, which is more closely related to AtP5CDH in the evolutionary relationship and has a higher expression level and protein sequence conservation. As a protein located in the nucleus and membrane, GmP5CDH1 is expressed in various soybean tissues, especially in developing seeds and cotyledons. In soybean, the expression level of GmP5CDH1 is induced to varying degrees under abiotic stress and ABA treatment, especially salt stress. Therefore, we obtained transgenic plants with heterologous overexpression of GmP5CDH1 in Arabidopsis. The determination of phenotype and physiological indicators of transgenic Arabidopsis and wild-type (WT) clarified that GmP5CDH1 can respond to salt stress and Pro strongly, possibly affecting plant growth and development by changing the content of endogenous Pro in the plant. These findings provide a basis for further research on the stress resistance phenotype and functional mechanism of GmP5CDH1 in soybean. In addition, we also identified a protein GmSAM1 that interacts with GmP5CDH1, which has been repeatedly reported to be involved in stress resistance, plant growth, and development in other crops. Therefore, in the future we will continue studying the mechanisms of GmP5CDH1 and GmSAM1 in soybean stress resistance.

2. Materials and Methods

2.1. Plant Materials and Growth Conditions

The soybean cultivar Nannong 94-16 was sown at the Jiangpu Experimental Station of Nanjing Agricultural University under field conditions. After the experimental materials grew to the seedling stage, samples were taken from the leaves, cotyledons, roots, and stems of plants. Next, the flowers were obtained; pods 7 and 13 days after flowering; seeds at 25, 35, and 45 days after flowering. Samples were firstly placed in liquid nitrogen for freezing preservation and then transferred to a −80 °C refrigerator for storage.

This study used the WT Arabidopsis seed Columbia-0 for gene-related functional validation. After inducing vernalization at a low temperature (4 °C) for 48–72 h, the surface was disinfected and planted in MS medium. The medium was placed in a plant light incubator with a day–night time of 16 h/8 h, a day–night temperature of 22/23 °C, and a relative humidity of 70%.

WT tobacco (Nicotiana tabacum) was grown in a 24 °C light incubator and used for injection about 3 weeks after planting.

To verify whether the expression level of GmP5CDH1 is affected by different external environments, uniformly sized soybean seeds were first selected and disinfected. Then, they were sown in sterilized nutrient soil (mixed with vermiculite and nutrient soil in a 1:1 ratio). When the first pair of true leaves of soybeans were fully unfolded, seedlings of similar size were selected and pre-cultured for 3 days in 1/2 Hoagland nutrient solution. Next, it was subjected to abiotic stress (polyethylene glycol, salt, and cold stress) as well as ABA treatment. (1) Drought stress treatment: the seedlings were placed in a 15% polyethylene glycol (PEG) solution and leaves were sampled at 0.5 h and 2 h. (2) Salt stress and ABA treatment: the seedlings were submerged in 250 mM NaCl and 10 mg/L ABA solutions, and leaf samples were taken at 3 h and 6 h. (3) Cold treatment: the seedlings were transferred to a constant temperature incubator at 4 °C and leaves were sampled at 2 h and 4 h. (4) Control treatment: under drought stress, salt stress, and ABA treatment, soybean seedlings were submerged in pure water without treatment, and leaves were sampled at 0 h, 0.5 h, 2 h, 3 h, and 6 h. Under cold treatment, soybean seedlings were placed in a 24 °C light incubator, and leaves were sampled at 0 h, 2 h, and 4 h [26]. The collected samples were quickly placed in liquid nitrogen for freezing preservation, and then transferred to a −80 °C refrigerator for storage.

2.2. Extraction and Reverse Transcription of Total RNA from Soybean

The extraction of RNA was carried out according to the new total RNA extraction kit (TianGen, Beijing, China), and the entire extraction process was carried out in an environment of ultralow temperature and no RNA enzyme. The M-MLV reverse transcription kit (TaKaRa, Dalin, China) was used to reverse the extracted RNA into cDNA.

2.3. Cloning and Bioinformatics Analysis of GmP5CDH1

Based on the sequence information for GmP5CDH1 predicted by Phytozome13 (https://phytozome-next.jgi.doe.gov/, accessed on 5 January 2023), a pair of specific primers was designed using Primer5. Using the seeds 13 days after flowering as a template, PCR amplification was performed using Phanta-Max Super-Fidelity DNA Polymerase (Vazyme, Nanjing, China). Next, 5 uL 10 × loading buffer was added to the 50 uL PCR amplification product; electrophoretic detection was carried out in 1% level agarose gel, and the separated target bands were cut and recovered using a FastPure Gel DNA Extraction Mini Kit (Vazyme, Nanjing, China). The recycled product was connected to the pCE2 TA/Blunt-Zero Vector (Vazyme, Nanjing, China). Primers are shown in the Supplementary Materials.

Relevant information about GmP5CDH1 was obtained, including coding region sequence (CDS), whole genome sequence, and amino acid sequence from Phytozome13. Protein sequences of other species with high homology were searched for using BLASTp in NCBI (https://www.ncbi.nlm.nih.gov/, accessed on 5 January 2023), and the phylogenetic tree was constructed with MEGA 6.0 software, where the bootstrap value was set to 1000. DNAMAN9.0 software, MEME (https://meme-suite.org/meme/tools/meme, accessed on 7 January 2023), and ExPASy (http://expasy.org/, accessed on 7 January 2023) were utilized to draw amino acid sequence alignment maps, align gene domains, and predict protein molecular weight and isoelectric points.

2.4. Gene Expression Analysis

RNA from various soybean tissues was extracted during the growth and development period mentioned in the above method and then inverted into cDNA for quantitative real-time polymerase chain reaction (qRT-PCR) analysis. qRT-PCR was carried out with an ABI 7500 system (Applied Biosystems, Foster City, CA, USA) using ChamQ™ SYBR qPCR Master Mix (Vazyme, Nanjing, China). The expression level of GmP5CDH1 was analyzed using the soybean genomic expression gene Gm-Tubulin (GenBank accession no. AY907703) as an internal reference [27]. The relative expression levels of GmP5CDH1 were calculated utilizing the 2^−ΔΔCt method [28]. Primers are shown in the Supplementary Materials.

2.5. Subcellular Localization of GmP5CDH1

The coding sequence of GmP5CDH1 with specific primers was cloned into the pFGC5941 vector to produce a fusion construct that includes a green fluorescent protein (GFP). The 35S: GFP (a control vector) and 35S: GmP5CDH1: GFP were transferred into Agrobacterium (Agrobacterium tumefaciens) strain EHA105 and then infiltrated into Nicotiana benthamiana leaves for transient expression. After 48–72 h, infiltrated leaves were imaged using laser confocal microscopy (LSM780; Zeiss, Germany). Primers are shown in the Supplementary Materials.

2.6. Arabidopsis Transformation

The coding sequence of GmP5CDH1 was cloned into the pMDC83 vector to obtain an Arabidopsis overexpression vector that would be introduced into Agrobacterium strain EHA105. Staining method was applied to infect Arabidopsis; a 5% sucrose solution (containing 0.3% SilwetL-77) was used as the infection solution [29].

Arabidopsis seeds harvested from the T0 generation were sterilized with sodium hypochlorite for 4 h, which were planted in MS medium supplemented with hygromycin antibiotics on a super-clean workbench. After 15 days, Arabidopsis seedlings that could grow normally were transferred to sterilized 1:1 vermiculite and nutrient soil.

DNA and RNA were extracted from the leaf tissues of T1 plants and WT Arabidopsis. The DNA was used as a template for PCR amplification to confirm positive plants. The RNA was inverted into cDNA for qRT-PCR analysis. Constitutive gene AT-Tubulin (AT5G62690) in Arabidopsis was used as an internal reference [27]. Primers are shown in the Supplementary Materials.

2.7. Phenotypic Analysis of Transgenic Arabidopsis

(1)

Germination rate statistics

After qRT-PCR analysis, two transgenic lines with higher gene expression levels were selected, sterilized with sodium hypochlorite in a fume hood, and then planted in MS medium for seed germination rate detection. Each experiment was repeated three times.

Using the same sterilization method as described above, the transgenic plants and WT were cultured in MS medium with exogenous application of different concentrations of Pro (0 mM,2 mM, 70 mM, 100 mM) for seed germination rate detection and repeated three times [30,31,32].

(2)

Determination of root length and fresh weight of Arabidopsis thaliana

After 14 days of salt stress and exogenous Pro treatment, a vernier caliper and electronic balance were used to measure the root length and fresh weight of transgenic plants and WT.

(3)

Determination of amino acid content in Arabidopsis thaliana

The harvested T2 transgenic Arabidopsis seeds were sterilized and seeded in MS medium. After growing for two weeks, they were transferred to nutrient soil. After the plants matured, their seeds were collected seeds separately in 2 mL centrifuge tubes and placed in a 30 °C oven for drying over a week. Then, the transgenic seeds were ground in a mortar with liquid nitrogen and their amino acid content measured. After thorough grinding, approximately 0.100 g (accurate to three decimal places) of each sample was weighed and added to 5 mL of HCl for extraction and filtration, and brought to a constant volume of 10 mL. A pipette was used to add a 500 μL sample to a 500 μL 4% sulfosalicylic acid, after which the sample was thoroughly mixed and centrifuged at 15,000 rpm for 15 min. The supernatant was filtered with a filter membrane and added to the upper bottle of the L-8900 automatic amino acid analyzer (Hitachi, Japan) to measure the amino acid content [33].

(4)

Detection of Pro content in Arabidopsis thaliana

The Pro content of transgenic Arabidopsis and WT subjected to salt stress and control, treated with exogenous proline at 0 mM, 2 mM, and 150 mM, was measured by ninhydrin reaction [34]. Fresh leaves near the roots were selected, 0.1 g weighed, and added to 1 mL PRO Lysis buffer to homogenize. After boiling in water for 10 min, samples were filtered with filter paper or gauze, and the filtrate is the proline extract. Set the proline standard sample gradient, 10, 20, 30, 40, 50, 60 μg/mL proline standard sample tubes (0.1 mL), blank tubes (0.1 mL ddH₂O), and proline extraction solution determination tubes (0.1 mL proline extract). Next, 1.5 mL PRO Assay buffer and 1.5 mL ninhydrin color development solution were added in sequence, mixed well, and then soaked in boiling water for 30 min. The solution turns red, after which is was quickly cooled and 2 mL of toluene added, shaken for 30 s, then let stand for a moment; the supernatant was transferred to a new centrifuge tube or test tube, centrifuged at 3000 g for 10 min, the supernatant taken for later use, adjusted to zero with a blank tube, and the absorbance at 520 nm measured in a standard tube using a spectrophotometer, which was set at zero with toluene and the absorbance of the measuring tube recorded.

A standard curve was created with the content (ug) of series proline standards as the horizontal axis and the corresponding absorbance as the vertical axis, and the proline content calculated based on the absorbance of the measuring tube. The content of proline was calculated for the specific sample according to the following formula:

Plant tissue sample PRO (ug/g) = C × VT/(W × VS)

• where C = proline content obtained from the standard curve (ug);
• VT = total volume of proline extract (ml);
• W = fresh weight of sample (g);
• VS = volume of extraction solution added during measurement (ml).

(5)

Analysis of GmP5CDH1 expression under abiotic stress

The transgenic and WT were grown for one month under normal conditions, three seedlings with similar growth were selected from each transgenic line and WT for treatment (2 mM Pro, 150 mM Pro, 150 mM NaCl) [30,31,32,35]. The control group under Pro treatment and salt stress treatment was irrigated with fresh water. One week later, samples were taken, and 3–4 leaves near the roots were selected for RNA extraction and reversal into cDNA. Using the constitutive gene AT-Tubulin (AT5G62690) in Arabidopsis as an internal reference [27], the expression of GmP5CDH1 in transgenic Arabidopsis lines was analyzed using method described above. Primers are shown in the Supplementary Materials.

2.8. Yeast Two-Hybrid Assay

Using homologous recombination, the CDS sequence of GmP5CDH was ligated to the pGBKT7 vector. The bait vector pGBKT7-GmP5CDH was co-transformed with the cDNA library plasmid of soybean into prepared yeast-competent cells, coated on SD dropout medium without Trp- Leu- His- Ade, and incubated at 30 °C for 5–7 days. The universal primers were used to sequence and distinguish positive monoclonal antibodies. The sequencing results were compared on NCBI to preliminarily identify interacting proteins. Clone full-length coding sequences of candidate interacting proteins to construct an AD vector and co-transfer them with pGBKT7-GmP5CDH1 into yeast to observe growth. Primers are shown in the Supplementary Materials.

2.9. Luciferase (LUC) Fluorescence Complementary Assay

Full-length coding sequences of GmP5CDH1 and GmSAM1 were cloned into the LUC bimolecular fluorescence complementary vector, resulting in GmP5CDH1-LUC-C vector and GmSAM1-LUC-N vector, respectively. Similarly, these vectors were subsequently transferred into EHA105 for co-infection of young Nicotiana benthamiana leaves. After 48–72 h, 1 mM D-Luciferin Potassium Salt solution (Yeasen, Shanghai, China) was painted on the back of the leaf, left in the dark for 5 min, and observed using an In Vivo Plant Imaging System (Berthold LB 985). Primers are shown in the Supplementary Materials.

2.10. Bimolecular Fluorescence Complementation (BiFC) Assay

Agrobacterium-mediated transient infiltration of Nicotiana benthamiana leaves were taken in this study. The living environment for tobacco was entirely within the plant-growth culture room in our laboratory, with a light cycle of 16 h and 8 h of darkness. Vectors were constructed to clone the GmSAM1 amplification fragment into 35S-SPYCE (M) and clone GmP5CDH1 into the SPYNE173 vector. Afterward, these vectors were transferred into EHA105 for co-infection of tobacco. After 48–72 h, laser confocal microscopy was used to observe the infected tobacco leaves (Leica SP8). Primers are shown in the Supplementary Materials.

3. Results

3.1. Phylogenetic Analysis of the Pyrroline-5-Carboxylate Dehydrogenase Gene Families

Using the amino acid sequence of the P5CDH gene family in Arabidopsis as a reference, three members of the P5CDH family with conserved domains in the soybean genome were identified through local BLAST analysis (https://www.ncbi.nlm.nih.gov/, accessed on 5 January 2023). A phylogenetic tree of the P5CDH gene family was built using MEGA software and the neighbor-joining method, which includes seven species: soybean, alfalfa, Arabidopsis, kidney bean, sorghum, corn, and wheat (Figure 1A). Phylogenetic tree assay displayed that the P5CDH protein in leguminous plants clustered into one branch, such as Phvul.003G192100 in kidney beans and Medtr4g107940 in alfalfa, followed by AT5G62530 in the dicotyledonous model plant Arabidopsis. The P5CDH proteins of sorghum, corn, and wheat are distributed on the outermost branches of the phylogenetic tree, all of which are monocotyledonous plants. Among them, pyrroline-5-carboxylate dehydrogenase in soybean can be divided into two subfamilies. Glyma.05G029200 has a closer genetic relationship with the homologous gene in Arabidopsis, and we named it GmP5CDH1. In addition, MEME (https://meme-suite.org/meme/tools/meme, accessed on 7 January 2023) was used to predict the motif of P5CDH proteins in seven species. The results showed that P5CDH proteins in different species exhibit highly consistent phenomena in both the number and type of motifs (Figure 1B). According to the expression data of Williams 82 provided by SoyBase (https://www.soybase.org/, accessed on 7 January 2023), Glyma.05G029100 is almost not expressed in various tissues of soybean, while Glyma.05G029200 is expressed in all tissues other than leaves. The expression level is higher in soybean seeds and shows a trend of first increasing and then decreasing as the seeds mature (Figure 1C).

3.2. Gene Structure and Conserved Motif Analysis of GmP5CDH1

GmP5CDH1 belongs to the pyrroline-5-carboxylate dehydrogenase gene which is one of the key enzymes in the Pro degradation pathway. We predicted the genomic sequence information for GmP5CDH1 on Phytozome13 (https://phytozome-next.jgi.doe.gov/, accessed on 5 January 2023). The genomic region of GmP5CDH1 contains 16 exons and 15 introns, with a total length of 1665 bp coding sequence (Figure 2A). The information on GmP5CDH1 protein was predicted through ExPASy (http://www.expasy.org, accessed on 7 January 2023). GmP5CDH1 encodes 554 amino acids with a molecular weight of 61.44 KDa and an isoelectric point of 7.23 (Figure 2B). Comparing the protein sequences of GmP5CDH1 with other homologous proteins, the conserved Aldedh domain of GmP5CDH1 is located at amino acids 60–524 and shows high similarity among different species (Figure 2C). The cis-acting regulatory element structure prediction analysis of the promoter region (2000 bp before 5’UTR) of GmP5CDH1 was performed using Plant CARE (https://bioinformatics.psb.ugent.be/webtools/plantcare/html/, accessed on 13 January 2023). The result showed that the promoter region of GmP5CDH1 contains a large number of light-responsive elements, including ACE, G-box, G-Box, box 4, and the TCCC motif; hormone-responsive elements include ABRE (abscisic acid), P-box (gibberellin), CGTCA and TGACG motifs (jasmonic acid), and the TCA element (salicylic acid); nonbiological stress-induced responsive elements include ARE (anaerobic responsive), MBS (drought responsive), TC-rich repeats (defense and stress-responsive), and other cis-acting elements including GCN4_ Motif (endosperm gene expression) (Figure 2D,

Table 1. Function for part of the predicted cis-acting regulatory elements.

Cis-Acting Regulatory Elements	Sequence	Function
ABRE	ACGTG/CACGTG	cis-acting element involved in the abscisic acid responsiveness
ACE	CTAACGTATT/GACACGTATG	cis-acting element involved in light responsiveness
ARE	AAACCA	cis-acting regulatory element essential for the anaerobic induction
AT-rich sequence	TAAAATACT	element for maximal elicitor-mediated activation (2 copies)
Box 4	ATTAAT	part of a conserved DNA module involved in light responsiveness
CAAT-box	CCAAT	common cis-acting element in promoter and enhancer regions
CGTCA-motif	CGTCA	cis-acting regulatory element involved in the MeJA-responsiveness
G-Box	CACGTT	cis-acting regulatory element involved in light responsiveness
G-box	CACGTG	cis-acting regulatory element involved in light responsiveness
GCN4_motif	TGAGTCA	cis-regulatory element involved in endosperm expression
GT1-motif	GGTTAAT/GTGTGTGAA	light responsive element
MBS	CAACTG	MYB binding site involved in drought-inducibility
P-box	CCTTTTG	gibberellin-responsive element
TATA-box	TATA/TATAA/TATTTAAA	core promoter element around −30 of transcription start
TCA-element	CCATCTTTTT	cis-acting element involved in salicylic acid responsiveness
TCCC-motif	TCTCCCT	part of a light-responsive element
TC-rich repeats	ATTCTCTAAC	cis-acting element involved in defense and stress responsiveness
TGACG-motif	TGACG	cis-acting regulatory element involved in the MeJA-responsiveness

).

3.3. Expression Pattern Analysis of GmP5CDH1 in Soybean

qRT-PCR analysis was performed to study the expression patterns of GmP5CDH1 in various tissues and stages of soybean. GmP5CDH1 was expressed in all soybean tissues, with relatively high expression levels in cotyledons, seeds 7 days after flowering, and seeds 13 days after flowering (Figure 3A).

Under various abiotic stresses (low temperature, drought, and salt stress) and hormone (ABA) induction conditions, we performed an expression analysis of GmP5CDH1 in soybean. The results showed that at 4 h after low-temperature treatment, the expression level of GmP5CDH1 was significantly increased compared to the control group (Figure 3(B1)). At 0.5 h of drought treatment, the expression level of GmP5CDH1 was higher; at 2 h, it was lower than that in the control (Figure 3(B2)). Notably, under salt stress, the expression level of GmP5CDH1 had a continuous and significant improvement (Figure 3(B3)). When treated with ABA, there was a roughly double-fold increase at 3 h compared to the control group (Figure 3(B4)).

3.4. GmP5CDH1 Is a Co-Localization Protein of the Nucleus and Cell Membrane

Through Softberry (http://www.softberry.com/, accessed on 26 February 2023) prediction, GmP5CDH1 was predicted as a mitochondrial localization protein. To investigate the expression position of GmP5CDH1 in plants, subcellular localization analysis of GmP5CDH1 was performed using an Agrobacterium-mediated transient expression system in tobacco plants. The specific distribution location of GmP5CDH1 in cells was determined based on the green fluorescence signal of the reporter gene. GmP5CDH1-GFP fusion protein was found by distributing it in the cell membrane and nucleus (Figure 3C).

3.5. Heterologous Overexpression GmP5CDH1 in Arabidopsis Has a Pleiotropic Effect on Seed Germination Rate and Amino Acid Content

By using the staining method, we transferred the GmP5CDH1 overexpression vector into WT Arabidopsis. Due to the significant impact of Pro content on seed maturation and germination, we selected two transgenic Arabidopsis lines and WT for the seed germination experiment. As a result, the germination rates for transgenic Arabidopsis: 29–11 and 29–21 were significantly lower than that for WT (Figure 4A).

To further investigate the effect of GmP5CDH1 on the development of Arabidopsis seeds, we detected the amino acid content of transgenic Arabidopsis and WT seeds. The results showed that the glutamate family including Pro, glutamic acid, and arginine, were significantly increased. In addition, the content of aspartic acid, threonine, glycine, alanine, valine, leucine, isoleucine, and tyrosine were also higher than that of WT, ultimately leading to a significant increase in the total amino acid content (Figure 4B).

3.6. Induction of GmP5CDH1 Expression in Response to Salt Stress in Arabidopsis

From the analysis of the expression patterns of GmP5CDH1 after the various stress treatments described above, the expression level was found to change significantly under salt stress. Therefore, we further observed the growth of transgenic Arabidopsis treated with salt stress.

We observed the growth status of transgenic plants after 14 days of salt stress treatment. The plant size of transgenic Arabidopsis was smaller than that of WT (Figure 5A). Using statistical methods to analyze the root length and fresh weight of plants, it was declared that both of them in transgenic lines were significantly reduced compared to WT (Figure 5B,E).

Then, we measured the Pro content of transgenic Arabidopsis and WT. Under control treatment, the Pro content of transgenic lines and WT were lower without obvious differences. After salt stress treatment, the Pro content in both transgenic lines and WT substantially increased (Figure 5C). However, the Pro content in transgenic lines was significantly lower than that in WT. Leaves of different plant lines were collected and qRT-PCR was used to detect the expression of GmP5CDH1. As expected, the expression level of the GmP5CDH1 gene in transgenic lines increased obviously after salt treatment (Figure 5D).

3.7. GmP5CDH1 Responded to Exogenous Proline Induction and Participated in Proline Synthesis

Different gradients of exogenous Pro concentrations were set to detect the effect of exogenous Pro treatment on the germination rate of transgenic Arabidopsis. The germination rate of transgenic plant seeds and WT showed an increasing trend when treated with 1 mM Pro, while the germination rate decreased remarkably when treated with 70 mM Pro and 100 mM Pro. Except for 100 mM Pro treatment, the overall germination rate of transgenic plants was lower than WT. As the concentration of exogenous Pro increased, the germination rates of both transgenic Arabidopsis lines decreased to a lower extent than WT (Figure 6A).

Higher concentrations of Pro (usually greater than 10 mM) are not conducive to plant growth and development. Therefore, we conducted phenotypic observations on transgenic Arabidopsis and WT treated with 150 mM Pro. The root of WT was seen to be shorter and weaker than that of both transgenic lines (Figure 6C). The transgenic lines root length and fresh weight data were larger than WT (Figure 6B,D).

Next, we measured the Pro content in Arabidopsis plants treated with different concentrations of Pro. The results showed that the Pro content in transgenic lines was completely higher than that in WT, whether at lower or higher concentrations of exogenous Pro treatment (Figure 6E). Nevertheless, the analysis of qRT-PCR showed an opposite trend between different concentrations. Under the treatment of 2 mM pro, the expression level of GmP5CDH1 significantly decreased compared to WT, while under the treatment of 150 mM Pro, the expression level of GmP5CDH1 was extremely higher than WT (Figure 6F).

3.8. Interaction between GmP5CDH1 and GmSAM1

Preliminary screening of interacting proteins with GmP5CDH1 was conducted through a yeast two-hybrid experiment in soybean cDNA library. A protein in yeast was repetitively found to interact with GmP5CDH1 (Figure 7A). Glyma.17g039100 encodes S-adenosylmethionine synthase, which catalyzes the synthesis of S-adenosyl-Met (SAM) in a single carbon metabolism cycle. We first constructed LUC fluorescence complementary vectors for GmP5CDH1 and GmSAM1. The result showed that when only GmP5CDH1-C-LUC and GmSAM1-N-LUC bacterial solution were co-injected, tobacco leaves emitted strong LUC fluorescence signals; other bacterial solution combinations were co-injected and did not show LUC fluorescence (Figure 7B). This phenomenon indicated that GmP5CDH1 and GmSAM1 exhibit protein–protein interactions in the LUC bimolecular fluorescence complementarity experiment. Furthermore, we constructed a bimolecular fluorescence complementary (BIFC) vector at the tobacco protein level to further verify their interaction and confirm interaction location. The complementary yellow fluorescence could fuse upon both nuclear and cell membranes, indicating its interaction location (Figure 7C).

4. Discussion

Pro plays a crucial role in both free amino acids and protein components in the cellular metabolism of plants. With the continuous deepening of research, the multiple roles of Pro in the normal growth and development of higher plants and their stress response are gradually being elucidated. Nowadays, there is a clear understanding of the physiological role of Pro in plants, which mainly includes the following aspects: (1) regulate the growth and development process of plants; (2) osmotic regulation function; (3) antioxidant effects; (4) participate in nitrogen metabolism and energy metabolism; (5) protective function of cellular structure; and (6) participate in photosynthesis. The functions of genes related to Pro metabolism have been reported in both the model plant Arabidopsis and the legume model crop alfalfa, and their important effects on abiotic stress have been confirmed. However, little is known about the effects of genes related to the Pro metabolism pathway in soybean on plant growth, development, and stress response. Due to the relatively low heritability of abiotic stress-related traits and the significant impact of environmental conditions, it may be difficult to identify abiotic stress-resistant genes through population construction and gene mapping methods [6]. Therefore, we first identified the homologous genes of the key enzyme gene AtP5CDH for Pro degradation in Arabidopsis through homologous sequence alignment in soybean. Phylogenetic tree analysis showed that the P5CDH gene family in soybean is homologous to that in many species such as Arabidopsis, alfalfa, and maize (Figure 1A). Moreover, the soybean P5CDH gene members are closely related to P5CDH in dicotyledonous plants such as Arabidopsis, and far from that in monocotyledonous plants such as maize, indicating that the development of P5CDH is consistent with the evolutionary history of species. The members of the P5CDH gene family in different species are roughly the same in terms of the number, length, and position of exons, also implying that this gene family is highly conserved in evolution and may have important implications for the growth and development of higher plants (Figure 1B). Through phylogenetic tree analysis and expression patterns in public databases, we selected Glyma.05G029200, which is closer to AtP5CDH and has a higher expression level for further study (Figure 1A,C) and named it GmP5CDH1 (Figure 2A). Protein multiple sequence alignment analysis revealed that it has a highly conserved Aldedh domain (Figure 2B,C). We predicted the cis-acting regulatory element structure of its promoter region with Plant CARE and identified many elements related to hormone induction and stress resistance. Through qRT-PCR analysis on its expression levels in soybean, the result was highly consistent with that in public databases. GmP5CDH1 is expressed in various tissues of soybean, and its expression level is higher in developing seeds, showing a trend of first increasing and then decreasing (Figure 3A). The expression of the AtP5CDH gene in Arabidopsis was higher in flowers and lower in seeds [19]. These results all suggest that GmP5CDH1 may be involved in resisting abiotic stress, play a role in seed development, and have species-specific expression patterns. In addition, the results of subcellular localization showed that GmP5CDH1 was located on the nucleus and membrane (Figure 3C). This result was different from the prediction of Softberry (http://www.softberry.com/, accessed on 26 February 2023) and the localization of homologous gene in Arabidopsis, also indicating that the function of the P5CDH genes family may have some difference among different species.

Due to the stress response of the Pro metabolism pathway and the results of cis-acting element analysis, we validated the expression of GmP5CDH1 under various abiotic stresses (low temperature, drought, and salt stress) and hormone (ABA) induction conditions in soybean. We found that GmP5CDH1 responded to abiotic stress and hormone induction (Figure 3(B1–B4)); especially under salt stress, GmP5CDH1 expression levels and expression patterns were significantly different compared with the control group, and showed a trend of continuous increase in expression level with salt treatment time. In rice, the expression of P5C dehydrogenase had no significant change at 6 h, and then showed a two-fold increase at 24 and 48 h [36]. In tobacco, NtP5CDH exhibits an increase in transcripts under water-deficient conditions [37]. In our study, the expression level of GmP5CDH1 was significantly increased at 0.5 h and then drastically decreased compared to the control at 2 h under drought treatment. This trend was not entirely consistent with that in rice and tobacco. It was speculated that the differences in species and drought treatment time led to the result. Extending the induction time of expression under drought treatment and increasing sampling points could further improve the expression pattern of GmP5CDH1 under drought stress. Next, we created transgenic Arabidopsis plants with heterologous overexpression of GmP5CDH1 to explore the growth and development of transgenic lines, as well as the relationship between GmP5CDH1 and abiotic stress. The previous study showed that the p5cdh mutant had lower seed yield and lighter seeds than WT because P5CDH mediated Pro catabolism in mobilizing carbon and nitrogen [23]. Without any treatment, we found that the transgenic lines showed a lower germination rate (Figure 4A). Furthermore, the contents of most amino acids and total amino acids in seeds were significantly increased compared to WT (Figure 4B). These two results proved that GmP5CDH1 may participate in the carbon and nitrogen metabolism pathway to affect seed components, resulting in a difference in germination rate. Under salt stress treatment, transgenic plant root length and fresh weight were significantly reduced compared to WT (Figure 5A,B,E). Furthermore, the expression level of GmP5CDH1 showed its positive response to salt stress induction in transgenic Arabidopsis (Figure 5D). Due to the important role of P5CDH in Pro synthesis, we speculate that the phenotype of transgenic plants may be caused by GmP5CDH1 affecting the endogenous Pro content of the plant. Consistent with previous observations [38], Pro content determination (Figure 5C) of WT and transgenic lines suggested that increased activity of P5CDH engaged in the degradation of Pro, and the decrease in Pro content may reduce the osmotic pressure in cells, thus affecting the normal growth of cells and the stress resistance under salt stress.

Previous studies mainly focused on endogenous Pro, but with the deepening of research, the influence of exogenous Pro on plant resistance has also been given more attention. The application of exogenous Pro in plant stress is mainly antifreeze [39], penetrant [40], and enzyme protectants [41,42]. The application of exogenous Pro can alleviate osmotic stress, salinization, and drought stress of plants [43]. Ben et al. [40] found that exogenous Pro could increase the content of Pro, photosynthetic rate (Pn), chlorophyll, carotenoid, and starch contents in olive trees and relieve the pressure of photosynthetic chain under salt stress. Although Pro is beneficial to biological growth under various stress conditions, it is interesting to note that externally supplied Pro is toxic to plant and animal cells under certain adverse conditions [19,44]. After the exogenous application of Pro to the Arabidopsis mutant rsr1-1, nearly all seedlings exhibited brown staining and necrosis. This observation suggests that the damage inflicted by Pro on plants may not be directly attributable to Pro itself, but rather to its degradation products P5C/GSA [45]. Exogenous Pro feedback was also found to inhibit Pro synthesis, and led to excessive reduction of light and electron acceptor, significantly increased reactive oxygen intermediates in chloroplasts and mitochondria, and destroyed chloroplast and mitochondrial ultrastructure in Arabidopsis leaves [46]. In our study, we observed that both WT seeds and transgenic seeds exhibited an increased germination rate at a beneficial concentration of Pro. Conversely, elevated concentrations of exogenous proline were found to inhibit their germination rate (Figure 6A). However, with the increase in exogenous Pro concentration, the germination rate of all transgenic Arabidopsis seeds had a lower trend than that of WT. Transgenic plants showed better growth conditions under low and high concentrations of exogenous Pro (Figure 6B–D). Even though the expression of GmP5CDH1 was opposite under the two concentrations (Figure 6F), the endogenous Pro content of transgenic materials was significantly higher than that of the control, making the plants more tolerant to the toxicity caused by exogenous Pro (Figure 6E). We hypothesized that low Pro concentration promoted plant growth and reduced the expression of GmP5CDH1, thus reducing the degradation of Pro and accumulating more Pro. A high concentration of exogenous Pro promoted the expression of GmP5CDH1, which may have weakened the Pro-P5C cycle, rapidly degraded the accumulated P5C in cells, prevented the production of ROS, and enabled the plants to maintain relatively favorable Pro content, thus protecting the plants from the toxicity of exogenous Pro. These results indicated that the transgenic lines could effectively resist the toxicity of Pro from outside, making them more tolerant to the toxicity of Pro.

The study on the resistance of P5CDH to abiotic stress in soybean has not been reported. We preliminarily demonstrated that there might be an interspecies similarity in the resistance effect of the P5CDH gene family on plants by heterologous overexpression of GmP5CDH1 in Arabidopsis. Next, we screened the interacting protein GmSAM1 in the soybean genome library by yeast two-hybrid and further verified it by LUC fluorescence complementary and BIFC. The results showed that GmP5CDH1 and GmSAM1 can bind and function on the nucleus and membrane. S-adenosylmethionine synthase is involved in the synthesis pathway of polyamines. Polyamines are biologically active aliphatic nitrogen-containing bases produced during biological metabolism, which play important regulatory roles in plant sex differentiation, morphogenesis, and response to external biotic and abiotic stress [47]. Dimethylsulfoniopropionate (DMSP) is a biomolecule synthesized from the methionine (Met) pathway. The accumulation of DMSP was proved to be utilized by the sea-ice diatom in response to salinity stress [48]. Therefore, GmSAM1 may also enhance the stress resistance of soybean by affecting the synthesis of polyamines or other metabolites on the pathway. In addition, SAM is crucial for the growth and development of plants. Chen et al. [49] discovered that the Arabidopsis SAMS3 gene deletion mutant, mat3, lacks polyamine synthesis, which inhibits pollen germination and pollen tube growth and reduces the seed setting rate. Lately, studies have shown that the knockout mutation of the Arabidopsis homologous gene AT3G17390 (MAT4) produced by CRISPR/Cas9 is lethal, making MAT4 an essential gene in Arabidopsis and indicating that MAT4 plays a dominant role in SAM production, plant growth, and development [50]. Additionally, S-adenosyl-Met synthetase 1 also plays an important role in the ethylene biosynthesis pathway, leading to increased S-adenosyl-Met accumulation and elevated ethylene production in tomato [51]. Studies on P5CDH and SAM in other species have shown that they significantly affect coping with abiotic stress, and plant growth and development. However, we still know very little about their function in soybean-related traits, the regulatory networks they participate in, and the interactions between them. Therefore, exploring the stress resistance effect of GmP5CDH1 on soybean and the synergistic regulatory networks between them will be our future research focus.

5. Conclusions

In our work, we identified a gene of P5CDH family named GmP5CDH1 in soybean. It was found located in the nucleus and cell membrane. High expression levels of GmP5CDH1 in seeds and cotyledons suggest its involvement in seed development. As expected, GmP5CDH1 positively responded to abiotic stress (low temperature, drought, and salt stress) and ABA treatment. Then, we mainly investigated the growth, development, and stress resistance of overexpression GmP5CDH1 transgenic Arabidopsis under salt stress and Pro induction. We emphasized that GmP5CDH1 participates in the response of the Pro pathway by degrading Pro while maintaining a certain level of endogenous Pro content to alleviate the toxic effects of high concentrations of exogenous Pro on plants. In addition, we found interaction protein GmSAM1, which was reported to be involved in plant stress resistance, growth, and development processes in other species. Although we have demonstrated the interaction between GmSAM1 and GmP5CDH1, their impact on the synthesis of Pro and another amino acid in soybean, and their functional mechanisms to maintain cell homeostasis under stress still need to be explored. To better understand the role of GmP5CDH1 in the resistance of soybean to abiotic stress and the regulatory network it relates to, we need to conduct more in-depth research in soybean.