The DMPs (DUF679 membrane proteins) are membrane proteins found exclusively in green plants (Viridiplantae) and especially in flowering plants (Cyprys et al., 2019). The members of this uncharacterized plant-specific gene family are predicted to be involved in various physiological processes, particularly senescence and reproduction (Kasaras and Kunze, 2010; Cyprys et al., 2019; Zhu et al., 2021). The DMPs are integral membrane proteins with no sequence similarity to functionally assigned protein domains, channels, transporters, or any other membrane proteins in any kingdom (Kasaras and Kunze, 2010). A total of ten members, located across four chromosomes, were identified in the DMP gene family in Arabidopsis (Kasaras and Kunze, 2010). The Arabidopsis DMP family proteins have four transmembrane spans and amino- and carboxy-termini in the cytosol (Kasaras and Kunze, 2010). Among 10 AtDMPs, AtDMP1 was reported as a senescence-associated gene, as it was upregulated during natural or developmental senescence of siliques, rosettes, and cauline leaves, as well as in dark-induced senescence of attached and detached leaves (Van der Graaff et al., 2006). Additionally, AtDMP1 is highly expressed in dehiscence and abscission zones of siliques (Kasaras and Kunze, 2010). Furthermore, AtDMP1, with dual endoplasmic reticulum (ER) and tonoplast localization, is engaged in membrane fission during the breakdown of these organelles during leaf senescence as well as membrane fusion during root vacuole biogenesis (Kasaras et al., 2012). Like AtDMP1, AtDMP3 and AtDMP4 were upregulated in senescing rosette leaves, cauline leaves, and siliques, indicating overlapping functions during senescence. Additionally, AtDMP3 and AtDMP4 transcripts were detected in roots and flowers, respectively. Three AtDMPs, AtDMP2, AtDMP6, and AtDMP7, were expressed in all organs (Kasaras and Kunze, 2010). AtDMP4 was coregulated with the core of dPCD (developmental-programmed cell death) marker genes, including BIFUNCTIONAL NUCLEASE1 (BFN1), PUTATIVE ASPARTIC PROTEASE A3 (PASPA3), RIBONUCLEASE3 (RNS3), CYSTEINE ENDOPEPDITASE 1 (CEP1), and EXITUS1 (EXI1). They were commonly upregulated in differentiation-induced and age-induced dPCD types (Olvera-Carrillo et al., 2015). Additionally, during Arabidopsis stigma senescence, AtDMP4, along with BFN1, RNS3, EXI1, CEP1, DMP4, and PASPA3 were reported to copromote senescence and dPCD (Gao et al., 2018). Furthermore, a comprehensive genome-wide analysis of four cotton species (Gossypium arboreum, G. raimondii, G. hirsutum, and G. barbadense) identified a total of 58 DMPs. Analysis of the expression patterns of these DMPs unveiled their possible involvement in key biological processes, such as plant senescence, flower reproductive development, and stress response (Zhu et al., 2021).

The sperm-specific AtDMP8 and AtDMP9 are reported to involve gamete fusion with a more significant effect on sperm-egg fusion (Takahashi et al., 2018; Cyprys et al., 2019). Moreover, the detection of a mutation in the haploid inducer-associated locus qhir8 (specifically the ZmDMP gene) in maize and loss-of-function mutations in the Arabidopsis AtDMP8 and AtDMP9 led to the development of an effective haploid induction system in dicots (Zhong et al., 2019; Zhong et al., 2020). Although AtDMP8 and AtDMP9 orthologues of the DMP gene family are utilized to induce maternal haploids in dicots (Zhong et al., 2020; Wang et al., 2022; Zhong et al., 2022a; Zhong et al., 2022b), information about other members in the family and their functional relevance needs to be investigated in crops. Soybean (Glycine max), the most widely grown commercial legume and oilseed crop, is commonly used for human consumption, livestock feed, oil production (Hartman et al., 2011). Soybeans contributed 70.39% of plant-based protein meals and 28.55% of plant-based oils in 2020/2021 globally (Market View Data Base, United Soybean Board 2021, accessed on 26 July 2022) and have been referred to as a ‘Wonder crop’ and the ‘Golden bean’ for their high nutritional content, oil content (18%), high-quality proteins (40%), high productivity, versatile uses, and profitability (Rajendran et al., 2022). Being a major oilseed crop, soybean has gained increasing attention in the genomics area, and the availability of genome sequence information accelerated the characterization of important gene families in soybean (Zhang et al., 2022a). However, studies on soybean DMPs investigating their genome-wide occurrence, phylogenetics, and functions are still lacking.

CRISPR (clustered regularly interspaced short palindromic repeats)/Cas9 technology relies on the precise and efficient introduction of double-stranded DNA breaks by the Cas9 nuclease, guided by a sgRNA. Predicting the cleavage efficacy of sgRNAs in vitro prior to their use in plant genome editing offers time, labor, and cost savings by enabling the selection of efficient sgRNAs, optimizing experimental design, and enhancing the success rate and accuracy of editing experiments (Mehravar et al., 2019; Bente et al., 2020). The integration of preassembled Cas9 enzyme with in vitro-transcribed sgRNA, known as the ribonucleoprotein complex (RNP), has been successfully employed in various plant species such as soybean (Kim et al., 2017; Subburaj et al., 2022), wheat (Liang et al., 2017), Brassica (Murovec et al., 2018; Jeong et al., 2019), maize (Sant’Ana et al., 2020), grapevine (Malnoy et al., 2016), apple (Malnoy et al., 2016), and pepper (Kim et al., 2020). This approach is used to assess the in vitro cleavage efficiency sgRNAs and their suitability for subsequent genome editing experiments.

In this study, we identified a total of 198 DMP genes from monocot and dicot crops and analyzed their phylogenetic relationships. Further, we comprehensively analyzed the physiochemical properties, subcellular location, gene structures, motifs, and promoters of the GmDMP family using bioinformatics tools. In order to understand the age- and tissue-specific expression levels of DMPs, as well as their responses to different stresses, publicly available transcriptome data were processed for analysis. Additionally, we conducted real-time quantitative PCR (qPCR) analysis specifically on floral and vegetative tissues to gain further insights into their expression profiles. Subsequently, based on floral expression pattern, we selected two genes with potential applications in haploid induction and conducted an in vitro cleavage assay to assess the cleavage efficiency of the selected sgRNAs. Our findings consolidate the information available in public databases on the DMP gene family and provide a comparative account of DMPs from soybean that would be useful for subsequent functional characterization. This study also identified potential candidate genes associated with haploid induction, growth, and stress response, which have the potential to accelerate soybean breeding.

With the availability of high-throughput technologies, genome datasets, and bioanalytical tools, soybean is receiving growing attention in genome-wide gene analyzes, identifying and characterizing multiple gene copies in each gene family. In this study, we mined the soybean genome for DUF679 family genes and characterized them using publicly available transcriptome datasets and various bioinformatics tools. A total of 14 DMP genes were identified in the soybean genome, which was in the range of other dicots and monocots (Kasaras and Kunze, 2010; Cyprys et al., 2019). The lengths of the identified DMP proteins ranged from 136 to 222 amino acids ( Table 1 ). The instability index showed that eight DMP proteins are stable, whereas six are unstable. The majority of the GmDMPs had positive GRAVY index scores, suggesting that they might be membrane proteins, mainly interacting with hydrophobic regions that span membranes, rather than globular proteins. Peptides with fewer charged groups are generally less soluble in water and are disposed to aggregation in and interaction with hydrophobic pockets of larger proteins (Lawrence et al., 2007; Niwa et al., 2009).

In the phylogenetic analysis, seven major clades were formed. The monocot-/dicot-specific sub-clustering patterns corroborate previous studies, suggesting a common ancestor before the divergence of monocot and dicot DMP genes (Cyprys et al., 2019; Zhu et al., 2021). The pollen-specific AtDMP8/9 orthologues (23 DMPs) included three soybean DMPs (Glyma.09G237500, Glyma.18G097400, and Glyma.18G098300). Orthologues of AtDMP8/9 were utilized in a DMP-based haploid inducer system to efficiently induce maternal haploids in maize (Zhong et al., 2019), Medicago truncatula (Wang et al., 2022), Brassica napus (Zhong et al., 2022b), tobacco (Zhang et al., 2022b), and tomato (Zhong et al., 2022a). Our qPCR analysis revealed that among the AtDMP8/9 orthologues, Glyma.18G097400 and Glyma.18G098300 exhibited pollen-specific and pistil-abundant expression patterns, respectively. While, Glyma.09G237500 was highly expressed in reproductive organs, pollen and pistil, at the reproductive stage ( Figure 5 ). This suggests their potential involvement in pollination and fertilization. A recent study demonstrated that the loss of function of the gynoecium-expressed phospholipase AII (pPLAIIγ) created haploid plants in dicotyledonous Arabidopsis (Jang et al., 2023), where the gynoecium-expressed pPLAIIγ induced female organ-mediated haploid induction. This highlights a potential of pistil-expressed Glyma.09G237500 and Glyma.18G098300 in female organ-mediated haploid induction, which could be further explored in future studies. So far, pollen-specific phospholipase A type of gene named MTL/NLD/ZmPLA1 was reported to the functional maternal haploid inducer in monocotyledonous plants such as maize, rice, wheat, and foxtail millet (Gilles et al., 2017; Kelliher et al., 2017; Liu et al., 2017; Yao et al., 2018; Liu et al., 2020; Cheng et al., 2021). Thus, we suggest that Glyma.18G097400 might be the potential target gene of maternal haploid inducer. Female organ-mediated haploid induction lines could be a promising tool as they allow the production of haploid plants without laborious emasculation procedures. Additionally, depending on germ lines, it may be more efficient than male-mediated haploid induction, as the female reproductive organ is the site of fertilization and zygote formation, providing a more direct route to haploid embryo production.

Subsequently, in an in vitro cleavage assay, we observed variations in the digestion efficiency of SgRNAs. SgRNA#1, targeting Glyma.18G097400, and Glyma.18G098300, exhibited higher cleavage efficiency compared to SgRNA#2, showing more cleaved bands from the originally amplified band size ( Figure 6 ). Comparative studies, have found that RNP complexes capable of efficiently cleaving their target sites in vitro consistently produce similar results in protoplast-based screening methods (Kim et al., 2017; Jeong et al., 2019; Kim et al., 2020; Sant’Ana et al., 2020; Subburaj et al., 2022). Notably, the use of this in vitro cleavage protocol offers several advantages over alternative methods. All required components are commercially available or can be easily produced in the laboratory at a low cost (Mehravar et al., 2019; Bente et al., 2020). Additionally, the entire procedure can be completed within a single day. These findings highlight the effectiveness of the rapid, simple, and cost-effective in vitro cleavage protocol in eliminating inefficient candidate SgRNAs and identifying those with optimal performance, thereby increasing the likelihood of successful in vivo functions.

To elucidate the potential regulatory roles of GmDMPs in the development and stress response of soybean, we studied the distribution and frequency of CAREs. The identification of CAREs is currently an intriguing area, allowing the study of complex gene expression by integrating computational, comparative, and functional genomics (Li et al., 2015; Ho and Geisler, 2019; Nawade et al., 2022). The light-responsive elements (LREs) were found to be prevalent in the GmDMP promoters. Most common LREs, Box 4,G-box, and GT1-motif have been demonstrated to be critical for the regulation of light-mediated transcriptional activity (Gangappa et al., 2013; Ezer et al., 2017). The highly conserved G-box motif (CACGTG) binds to the basic helix-loop-helix (bHLH) and basic leucine zipper (bZIP) families of protein motifs (Heim et al., 2003; Carretero-Paulet et al., 2010) and reported to involved in the regulation of chlorophyll biosynthesis in Arabidopsis (Menkens et al., 1995; Kobayashi et al., 2012). The GATA motif (detected in 4 GmDMPs) plays a role in light responsiveness and tissue specificity, and is involved in the light-dependent development of phloem tissues (Trishla et al., 2020). In photosynthetic-responsive gene promoters, both I-box and G-box elements have been shown to be essential for activation in response to phytochrome, cryptochrome, and plastid signals. (Martıínez-Hernández et al., 2002; López-Ochoa et al., 2007). Twelve GmDMP promoters have either I-box or G-box, suggesting their functional relevance as regulators of ribulose 1,5-bisphosphate carboxylase/oxygenase small subunit light-responsive units (Murata et al., 2002). The promoter of Glyma.18G098300 was found to have a site for the sp1 (GGGCGG) element ( Figure 3 , Supplementary Table S5 ), which is considered a mammalian promoter element, implicated in the regulation of a wide variety of housekeeping genes and tissue-specific genes (Hagen and Guilfoyle, 2002). Interestingly it was not found in the Arabidopsis or rice genomes (Yamamoto et al., 2007). The AAGAA-motif, activation sequence-1 (as-1; TGACG), O2-site, and GCN4_motif were the most frequently distributed CAREs related to plant growth and development ( Figure 3 , Supplementary Table S5 ). The as-1 element was found to be responsible for auxin- or salicylic acid-dependent enhanced expression in leaves (Niggeweg et al., 2000), whereas the AAGAA-motif was associated with secondary xylem development (Ain-Ali et al., 2021).

The promoters of soybean DMP genes have been also found to contain CAREs that are known to modulate gene expression in response to various stresses. Among abiotic stress-responsive CAREs, MYC, and MYB binding sites were present in all promoters. They have been reported to play an important role in drought-inducible expression, indicating that GmDMP expression is associated with abiotic stress (Smita et al., 2015). Glyma.07G201500, Glyma.09G237500, and Glyma.07G253500, which showed higher expression during drought, harbored multiple binding sites for dehydration-responsive elements ( Figure 3 , Supplementary Table S5 ). Moreover, the presence of LTR (low temperature responsive) in the Glyma.16G157800 promoter suggests its involvement in cold stress response (Baker et al., 1994; Zhang et al., 2020). The ARE (anaerobic responsive elements) motifs, known as low oxygen and dehydration-induced elements (Dolferus et al., 2001), were found in 10 GmDMPs ( Supplementary Table S5 ). The abscisic acid responsive elements (ABREs) have higher binding sites in most of the GmDMP promoters ( Figure 3 , Supplementary Table S5 ). They are regulators of various processes, including stomatal closure and seed and bud dormancy, as well as mediators in plant responses to cold, drought, and salinity stress (Choi et al., 2000; Yoshida et al., 2015). CAREs involved in the methyl-jasmonate responsive elements (TGACG-motif and CGTCA-motif) were also present in 10 GmDMP promoters ( Figure 3 , Supplementary Table S5 ). These elements are crucial in TF-mediated gene regulation (Rouster et al., 1997). For example, TGA1, a bZIP TF, was reported to act as a positive regulator of disease resistance by binding at the TGACG-motif and CGTCA-motif of the pathogenesis-related (PR-1) gene promoter in the Arabidopsis (Shearer et al., 2012). Likewise, the TGACG motif from the rice 12-oxophytodienoic acid reductase-1 (OsOPR1) promoter has been shown to play essential roles in defense responses (Sobajima et al., 2007). Moreover, the existence of W-box, WRE3, and the WUN-motif in promoters suggested that GmDMPs might play a vital role in biotic stress responses. The WUN-motif was characterized as wound responsive in the WUN1 gene (Siebertz et al., 1989), while WRKY TFs are known to function in wound response by binding to W-boxes (Eulgem et al., 2000). Moreover, the presence of these CAREs and their elevated expression upon Fusarium infection suggest the possible role of Glyma.18G097400 and Glyma.18G098300 in fungal response ( Figure 4B , Supplementary Table S6 ). Between the two GmDMPs in sub-clade III-1, Glyma.16G157800 and Glyma.02G075800, Glyma.02G075800 expression was 3-fold more elevated in Macrophomina resistant genotype. Notably, these AtDMP10 orthologue harbors wounding and pathogen responsive CAREs ( Supplementary Table S5 ). The W-box (TTGACC) element, which interacts with WRKY transcription factors (TFs) and regulates the expression of defense-related (pathogenesis-related 10, PR-10) genes, has a role in biotic and abiotic stresses, seed dormancy, and senescence (Dhatterwal et al., 2019). Under different biotic and abiotic stresses, the DMP genes exhibited diverse expression patterns and presence of stress responsive CAREs, indicating their potential roles in improving stress resistance and survival in soybean ( Figure 7 ). However, further studies are required to validate the functional relevance of these genes during stress.

In summary, DMP genes retrieved from soybean were subjected to bioinformatics analyzes for characterization. In the phylogenetic analyzes, group V, composed of AtDMP3 and AtDMP5, orthologues were absent from the soybean genome. The AtDMP8/9 orthologues Glyma.18G097400 and Glyma.18G098300, showing pollen-specific and pistil-abundant expression, could be potential targets for haploid induction in soybean. The expression profiles for most GmDMPs showed varying expression patterns in response to biotic and abiotic stresses. The presence of binding sites for various regulatory elements in GmDMP promoter sequences is consistent with this divergent expression pattern and implicates their possible involvement in growth and stress responses.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/ Supplementary Material .

OL acquired the funds and resources, designed the experiment, and administered and supervised the project. OL, BN, and TB conceptualized the project. OL, BN, JL, and JJ designed the methodology, curated the data, and performed the formal analysis. TB performed the RNA-Seq analysis. BN and TB implemented the software and visualized and validated the data. OL, BN, TB, and JJ wrote the paper. All authors contributed to the article and approved the submitted version.