Genome-Wide Identification, Molecular Evolution, and Abiotic Stress-Responsive Regulation of Cupin Superfamily Genes in Rice (Oryza sativa L.)

Abstract

The Cupin superfamily, characterized by a conserved β-barrel structure, plays crucial roles in plant growth, development, and stress responses. However, comprehensive analyses of this gene family in rice remains limited. Here, we performed a genome-wide identification, molecular evolution, and expression analysis of Cupin genes in rice under abiotic stress. Utilizing the telomere-to-telomere (T2T) genome of rice, 54 Cupin genes (OsCupins) were identified and classified into four subfamilies (GLP, PIRIN, TRR14, and ARD) based on phylogenetic relationships with Arabidopsis homologs. These genes were unevenly distributed across ten chromosomes, with tandem and segmental duplications driving their expansion. Structural and synteny analyses revealed conserved motifs and orthologous relationships with sorghum and Arabidopsis. The promoter regions of OsCupins were enriched in stress-responsive cis-elements, including ABRE, MYB, and MYC motifs. qRT-PCR data demonstrated the significant upregulation of multiple OsCupins (e.g., OsGLP15, OsGLP38, and OsGLP43) under NaCl and PEG 6000 treatments. Functional validation in yeast showed that the overexpression of OsGLP15, OsGLP38, or OsGLP43 enhanced salt and drought tolerance in yeast, with OsGLP43 exhibiting the strongest stress resilience. Our findings provide insights into the evolutionary dynamics and stress-responsive regulatory mechanisms of the Cupin superfamily in rice, offering potential targets for enhancing abiotic stress tolerance in this critical crop.

1. Introduction

The Cupin superfamily is a protein family widely found in plants, animals, and microorganisms, characterized by its unique β-barrel structure, which enables it to perform various biological functions [1]. Besides this conserved β-barrel structure, Cupin superfamily proteins also have two conserved motifs: [G(x)5Hx H(x)3,4E(x)6G] and [G(x)5Px G(x) 2H(x)3N]. These motifs are separated by 15 to 50 amino acids and together form a metal–ion-binding region shielded within the β-barrel structure [2,3]. The Cupin superfamily exhibits remarkable functional diversity, encompassing metal-binding enzymes, seed storage proteins, sugar-binding proteins, sugar isomerases, and other metal-dependent enzymes with catalytic activity. To date, several families of proteins have been recognized within the Cupin superfamily, such as germin-like proteins (GLPs) [4], PIRIN proteins [5], acireductone dioxygenase (ARD) proteins [6], and Trehalose Resistance 14 proteins (TRR14) [7]. The Cupin superfamily gene, initially discovered in wheat and termed germin, encodes a protein featuring the Cupin_1 domain, which is a key marker during wheat germination [3].

The GLP family consists of soluble glycoproteins that exhibit high sequence homology to wheat germin [8]. These GLPs are expressed in a variety of plant organs, such as roots, stems, leaves, flowers, and seeds, and demonstrate diverse functions throughout different stages of plant growth [9,10]. They primarily function as receptors, enzymes, and structural proteins in multiple physiological processes [11]. Research has indicated that PsGLP1-2 in plum is involved throughout fruit development and ripening [12], ClGLP1 in lemon is abundantly expressed in ripe fruits [13], and the silencing of OsGLP1 in rice leads to dwarfism, decreased SOD activity, and reduced disease resistance [8]. Furthermore, GLPs belong to a widespread class of pathogenesis-related proteins in plants, playing a vital role in the plant’s response to pathogen attacks and external stresses [14]. The Pirin family proteins were first identified in humans through yeast two-hybrid screening, interacting with the transcription factor NFI/CTF1 to carry out their related functions [5]. Pirin proteins are not limited to humans; they are also found in other mammals, plants, fungi, and even prokaryotes, with their N-terminal protein structure being highly conserved across species [5,15]. In Arabidopsis, AtPRN1 is found in both the cytoplasm and nucleus, and its expression is stimulated by abscisic acid (ABA) and red light. It interacts with the α subunit of the G protein GPA1 and acts downstream of it in the signaling pathway [15,16]. Research has demonstrated that AtPRN1 is a shuttling protein that moves into the nucleus in response to signals and binds to the transcription factor NF-Y. Together with the G protein-coupled receptor GCR1, GPA1, and NF-Y, AtPRN1 forms a signaling pathway that mediates responses to blue light and ABA, influencing seed germination [17]. Meanwhile, AtPRN2 is also localized to the cytoplasm and nucleus and is primarily expressed in cells near the vascular bundles. It regulates the expression of lignin biosynthesis genes, thereby inhibiting the accumulation of S-type lignin near the xylem vessels [18]. ARD proteins are enzymes that are ubiquitous in organisms. In the methionine cycle, they catalyze the conversion of cis-ketones to 2-keto-4-methylthiobutyrate (KMTB), facilitating methionine synthesis. As methionine is a crucial component in the ethylene synthesis pathway, ARD genes play a pivotal role in regulating ethylene synthesis and metabolism in plants. The Arabidopsis AtARD gene family comprises four members: AtARD1, AtARD2, AtARD3, and AtARD4 [19]. Notably, AtARD1 is an effector of the β subunit of the heterotrimeric G protein and regulates cell division to control hypocotyl length in Arabidopsis seedlings [20]. In rice, OsARD1 is involved in ethylene synthesis via the methionine cycle, enhancing the plant’s tolerance to drought and salinity [21,22]. TRR14 is a novel protein that may play a part in plant responses to environmental stresses, particularly in response to trehalose treatment [23]. Researchers used a reverse genetics approach to isolate two trr14 T-DNA insertion mutants from the SALK collection, which exhibited heightened sensitivity to salt and drought stress compared to wild-type plants [23].

Rice, a staple food crop worldwide, plays a pivotal role in global food security due to its consistent yield and quality. However, during its growth period, rice often faces various abiotic stresses, such as drought, salinity, and extreme temperatures. Recent studies indicate modern high-yielding rice varieties experience approximately 12% yield reduction at salinity levels above 3 dS/m, increasing to roughly 50% yield loss at 6 dS/m [24]. Therefore, gaining a thorough understanding of the molecular mechanisms underlying rice’s response to these stresses is crucial for enhancing rice’s stress resistance and ensuring stable food production. The rice genome comprises 12 chromosomes, totaling approximately 430 Mb of nucleotides [25]. In 2005, the International Rice Genome Sequencing Project (IRGSP) released the initial draft of the rice genome (Nipponbare), known as IRGSP-0.5 [26], and has since refined it to produce a more accurate reference genome for Nipponbare rice, termed IRGSP-1.0 [27]. Nevertheless, challenges in assembling the rice genome persist, including the abundance of complex repetitive sequences, the unique structures of telomeres and centromeres, and limitations in sequencing technology. The completion of the telomere-to-telomere (T2T) genome of rice (AGIS1.0) resolves ambiguities in repetitive regions and centromeres [28], enabling a comprehensive identification of previously overlooked Cupin genes, particularly tandem duplicates in complex loci, and the precise analysis of their molecular evolution. Despite reports in other plants, the molecular evolution, functional diversification, and stress-responsive roles of Cupins in rice remain underexplored, warranting genome-wide analysis.

The objective of this study is to conduct a systematic analysis of the molecular evolution of Cupin superfamily genes in rice by identifying them genome-wide and investigating their expression patterns under abiotic stress conditions. This research aims to offer new perspectives and theoretical groundwork for a better understanding of rice’s stress resistance mechanisms.

2. Materials and Methods

2.1. Identification of Cupin Genes in Rice

The AGIS1.0 version of the rice T2T genome and annotation files were obtained from the RiceSuperPIRdb website (http://www.ricesuperpir.com/web/nip, accessed on 15 February 2025). Amino acid sequences of Cupin genes from Arabidopsis were retrieved from the Arabidopsis Information Resource (TAIR) (https://www.arabidopsis.org/, accessed on 15 February 2025) and aligned using BLASTp. Hidden Markov Model (HMM) files for Cupin domains, namely, Cupin_1 (PF00190), Cupin_3 (PF05899), ARD (PF03079), Pirin (PF02678), and Pirin_C (PF05726), were downloaded from the Pfam protein family database (http://pfam.xfam.org, accessed on 15 February 2025). Using HMMER 3.0 (http://hmmer.org/, accessed on 15 February 2025), we searched for Cupin genes in the rice genome database, targeting all genes that contained Cupin domains. To validate the identification of these genes, we utilized Pfam, SMART, and InterPro programs. Lastly, the longest coding sequence (CDS) transcript was chosen as the representative sequence for each gene. Domain architecture was characterized using Pfam, SMART, and InterPro to classify OsCupins based on the number of Cupin domains per protein. Proteins containing a single Cupin domain (e.g., Cupin_1 for GLPs, Cupin_3 for TRR14, ARD for ARD subfamily) were categorized as single-domain type, while those harboring two distinct Cupin domains (specifically PIRIN + PIRIN_C for OsPIRIN1-3) were classified as double-domain type.

2.2. Physicochemical Property Analysis

The ProtParam tool (https://web.expasy.org/protparam/, accessed on 16 February 2025) was used to calculate the physicochemical properties of the predicted rice Cupin proteins. These properties encompassed the length of the amino acid sequence, the theoretical isoelectric point (pI), molecular weight (MW), instability index, aliphatic index, and grand average of hydropathicity (GRAVY) index. For predicting the subcellular localization of rice Cupin genes, we utilized the Wolf Psort online website (https://wolfpsort.hgc.jp/, accessed on 16 February 2025) and selected the top prediction as the final result.

2.3. Chromosomal Localization of Cupin Genes in Rice

The chromosomal positions of rice Cupin genes were depicted using the Gene Location Visualize module from GTF/GFF in TBtools v2.142.

2.4. Phylogenetic Analysis, Gene Structure, and Conserved Motif Analysis of CUPIN Genes in Rice

For phylogenetic relationship analysis, amino acid sequences of Cupin genes from both rice and Arabidopsis were chosen. These sequences were aligned using Clustal, and subsequently, a phylogenetic tree was constructed using the Neighbor-Joining (NJ) method in MEGA11 software (https://www.megasoftware.net/, accessed on 27 February 2024), with a Bootstrap value of 1000. The phylogenetic tree was then visualized and enhanced using iTOL (https://itol.embl.de/, accessed on 19 February 2025). The nomenclature of rice Cupin genes was based on the results of phylogenetic analysis and chromosomal localization. To analyze the motif structure of rice Cupin family members, we utilized the online MEME software (v5.5.7, https://meme-suite.org/meme/, accessed on 27 February 2024), setting the parameters to a maximum of 20 motifs, with a minimum width of 6 and a maximum width of 50. The gene structures of rice Cupin genes, comprising CDS and UTR, were obtained from rice genome annotation files. Lastly, TBtools v2.142 was employed for the visual examination of the data.

2.5. Collinearity Analysis and Gene Duplication of Cupin Genes in Rice

To identify homologous Cupin genes in rice, we utilized BLASTp with an e-value threshold of less than 1 × 10⁻⁵. For analyzing the collinear relationships among rice Cupin genes, MCScanX was applied using its default parameters. The resulting Cupin collinear gene pairs were then visualized using TBtools. For the collinearity analysis, we downloaded the genome sequences and annotation files of Arabidopsis (TAIR10.55) from the Phytozome v13 website (https://phytozome-next.jgi.doe.gov/, accessed on 22 February 2025). Similarly, the genome sequences and annotation files of sorghum (GCF_000003195) were obtained from the NCBI website (https://www.ncbi.nlm.nih.gov/, accessed on 22 February 2025). MCScanX was subsequently used to investigate the collinearity between rice and Arabidopsis, as well as between rice and sorghum.

2.6. Analysis of Cis-Acting Elements in Rice Cupin Gene Promoters

The promoter sequence for each rice Cupin gene was defined as the 2000 bp region upstream of the start codon (ATG). We used TBtools to extract these promoter sequences and then utilized PlantCARE to predict the cis-regulatory elements (https://bioinformatics.psb.ugent.be/webtools/plantcare/html/, accessed on 22 February 2025). For the visualization and summarization of the cis-elements, we employed the Basic Biosequence View modules in TBtools. We prioritized stress/growth/hormone/light elements due to their established roles in abiotic stress adaptation, aligning with our focus on stress-responsive regulation.

2.7. Plant Materials and Stress Treatment

To analyze the response of Cupin genes to abiotic stress, the rice plants of O. sativa japonica cv. Liaoxing 1 (LX1) were grown in a constant temperature incubator at 30 °C under16 h light/8 h darkness. After 2 weeks of growth, the rice seedlings were treated with 200 mM sodium chloride (NaCl), 20% polyethylene glycol (PEG 6000), and ddH2O, respectively. The chosen concentrations of 200 mM NaCl and 20% PEG 6000 represent severe osmotic stress conditions commonly employed in laboratory screening assays to rapidly and robustly elicit molecular stress responses and identify potentially resilient genotypes or genes [29]. The samples were selected at 0, 3, 6, 12, and 24 h after treatment, and RNA extraction and qRT-PCR analysis were performed.

2.8. RNA Extraction and qRT-PCR Analysis

Total RNA was isolated using the plant total RNA extraction kit (TAKARA, NO.9769) and subsequently reverse-transcribed into cDNA with the PrimeScript RT reagent Kit (TAKARA, RR037Q). The resulting cDNA served as the template for gene expression analysis. The SYBR Premix Ex TaqII kit (TAKARA, RR820A) was selected, and the reaction system was set up following the manufacturer’s instructions. Fluorescence detection was performed using the Applied Biosystems ABI7500 (Thermo Fisher Scientific, MA, USA). All primer sequences, including the internal control actin1, are provided in Table S1.

2.9. Expression Patterns of Cupin Family Genes in Rice

The transcriptome data, encompassing expression levels (FPKM values) in shoots, 20-day-old leaves, young seedlings at the four-leaf stage, pre- and post-heading inflorescences, anthers, pistils, 5- and 10-day post-fertilization seeds, 25-day post-fertilization young embryos, and 25-day post-fertilization endosperm, were retrieved from the RGAP rice database. The log2FPKM values were displayed in the form of heatmaps using the HemI software (version 1.0) [30]. RNA-seq data for 14-day-old rice seedlings (Oryza sativa japonica Group) under ABA/JA treatments were downloaded from NCBI (PRJDB2600). Raw reads were processed via TBtools SRAtoFastq plugin (v2.081) and aligned to the AGIS1.0 genome using HISAT2 (v2.2.1). Gene-level raw counts were quantified using featureCounts (v2.0.3) and normalized with DESeq2 (v1.34.0). Expression patterns of Cupin genes were visualized as a heatmap using DESeq2-normalized log2 counts in TBtools HeatMap plugin.

2.10. Yeast Transgene Validation

The cDNA of LX1 was utilized as the template for amplifying target genes. Gene-specific primers (listed in Table S1) were designed to clone the target sequences, which were subsequently integrated into the yeast expression vector pYES2 via homologous recombination. Recombinant plasmids were introduced into the yeast strain INVscl using the PEG/LiAc transformation method. Transformed yeast cells were cultured on SD/-Ura solid medium at 28 °C for 5 days to select positive clones. Positive colonies were pre-cultured in liquid medium for 24 h, followed by induction with galactose for 36 h. The optical density (OD₆₀₀) of the yeast cultures was adjusted to 1.8, and serial dilutions were spot-inoculated onto agar plates. To simulate stress conditions, SD/-Ura media were supplemented with 1 M NaCl or 30% PEG 6000 [31]. Yeast strains transformed with the empty pYES2 vector served as the control group.

2.11. Data Processing

The relative expression levels were calculated using the 2−ΔΔCt method based on the qRT-PCR data [32]. Statistical analysis of the expression levels was performed using one-way ANOVA in SPSS 18.0 software, and significance was determined using Duncan’s multiple comparison test (p < 0.05). The data were then organized and visualized using GraphPad Prism 8.0 software.

3. Results

3.1. Genome-Wide Identification of Cupin Genes in Rice

To identify Cupin members in rice and Arabidopsis, a BLASTp search was performed, resulting in the identification of 54 and 39 putative Cupin genes in the rice and Arabidopsis genomes, respectively. Chromosome mapping revealed that these Cupin genes were unevenly distributed across ten of the twelve rice chromosomes (Figure 1). Chromosome 8 had the highest number of Cupins, with fifteen, followed by chromosome 3, with eleven (Figure 1). However, no Cupins were found on chromosomes 6 and 7.

3.2. Phylogenetic Analysis of Cupin Genes in Rice

To investigate the phylogenetic relationships and evolutionary trends of Cupins in rice, a neighbor-joining tree was constructed based on the full-length protein sequence alignments of 54 OsCupins and 39 AtCupins. The Cupins were distinctly categorized into eight well-supported sub-clades, named after their similarity to Arabidopsis thaliana Cupins: GLP-1, GLP-2, GLP-3, GLP-4, GLP-5, PIRIN, TRR14, and ARD clades (Figure 2). Among these, GLP-1 formed the largest clade, comprising 39 Cupins (20 from rice and 19 from Arabidopsis). The GLP-2 to GLP-5 clades contained 10 (6 OsCupins and 4 AtCupins), 11 (8 OsCupins and 3 AtCupins), 12 (7 OsCupins and 5 AtCupins), and 3 (3 OsCupins) Cupins, respectively. These five clades (GLP-1 to GLP-5) were closely clustered and not clearly distinct from one another. Additionally, the PIRIN, TRR14, and ARD clades included 6 (3 OsCupins and 3 AtCupins), 5 (4 OsCupins and 1 AtCupin), and 7 (3 OsCupins and 4 AtCupins) Cupins, respectively. Based on their phylogenetic relationship with Arabidopsis and their physical location on the chromosomes, the 54 OsCupins were renamed OsGLP1 to OsGLP44, OsPIRIN1 to OsPIRIN3, OsARD1 to OsARD3, and OsTRR14-1 to OsTRR14-4 (Figure 2).

3.3. Physicochemical Property of Cupin Proteins in Rice

Additionally, an analysis was conducted on the sequence characteristics of Cupin proteins in rice. The open reading frame (ORF) lengths for all 54 Cupins varied between 330 bp and 1566 bp, encoding polypeptides ranging from 109 to 521 amino acids (aa) in length. These polypeptides had predicted molecular weights (MWs) ranging from 11.71 kDa to 56.44 kDa and theoretical isoelectric points (pIs) spanning from 4.68 to 9.62. Based on their instability index (II) values, 46 Cupins were classified as stable proteins (II < 40), while 8 were deemed unstable (II > 40). Furthermore, 53 out of the 54 Cupin proteins had an aliphatic index exceeding 65, with values ranging from 66.37 to 104.7, and only one Cupin protein had an aliphatic index below 65 (53.55), suggesting that the majority of Cupin proteins exhibit high thermostability. The GRAVY values, which are indicative of protein solubility, ranged from −0.86 to 0.54, with 72.2% (39/54) of the Cupin proteins having a GRAVY value of more than 0, indicating that most Cupin proteins are non-hydrophilic. Subcellular localization predictions showed that the majority of Cupin proteins are in the extracellular space, chloroplasts, plasma membrane, and cytoplasm, with only a few found in peroxisomes, vacuoles, the Golgi apparatus, and the mitochondrial matrix. The relevant information is listed in Table S2.

3.4. Gene Structure, Conserved Domain, and Motifs of Cupin Genes in Rice

To elucidate the characteristic domain features of Cupin subgroups, we analyzed the conservation of amino acid residues within their functional domains by aligning the Cupin sequences. We identified a total of 20 unique motifs, labeled as motif 1 through motif 20. Members belonging to the same subfamilies typically exhibited similar motif compositions. For instance, the GLP-1 clade contained 7 to 10 motifs, with motif 1–3, motif 5, motif 6, and motif 10 being highly conserved (Figure 3B). The GLP-2 clade had 9 motifs, including motifs 1–6, motif 10, motif 15, and motif 17, among which motifs 1–4 were highly conserved (Figure 3B). GLP-3 to GLP-5 clades possessed 4 to 8 motifs (motif 1–7, 9, 10, 15, and 17), 5 to 6 motifs (motif 1–6 and 15), and 3 to 4 motifs (motif 1, 2, 15, and 17), respectively (Figure 3B). In the other subfamilies, the ARD clade had only five motifs: motif 7, 13, 16, 18, and 20; the PIRIN clade had 5–6 motifs (motif 12–14, 16, 17, and 19); and the TRR14 clade had only 2–3 motifs (motif 7, 11, and 17). All motifs were highly conserved except motif 7 (Figure 3B).

To verify the authenticity and completeness of the identified rice Cupin genes, we analyzed their functional domains by querying the Pfam and SMART databases. This analysis identified 51 single-domain and 3 double-domain OsCupins (all OsPIRINs). All OsCupin proteins in the GLP-1 to GLP-5 clades possess a Cupin_1 domain and are classified as members of the OsGLP family (Figure 3C). Proteins in the ARD clade contain exclusively one ARD domain and belong to the ARD family (Figure 3C). All proteins in the PIRIN clade have two domains, namely, the PIRIN domain and the PIRIN_C domain, and are part of the PIRIN family (Figure 3C). The TRR14 clade protein is characterized by the presence of a single Cupin_3 domain (Figure 3C). Moreover, we examined the exon–intron organization of each Cupin protein. Out of the 54 OsCupin genes, 18 have a single exon, while 27 have two exons (Figure 3D). Additionally, both ARD and PIRIN genes exhibit multiple exons and introns (Figure 3D). The isoforms of all OsCupins within their respective subgroups share similar exon–intron structures. OsGLP7 (with eight exons) and OsGLP19 (with four exons) are exceptions, having a higher number of exons than the regular members of this subfamily.

3.5. Gene Duplication and Synteny Analysis of Cupin Genes in Rice

Gene duplication events are a critical part of gene family composition. Analysis of family gene duplication events with the MCScanX tool showed that there were only two segment duplication events (OsTRR14-1–OsTRR14-3 and OsARD3–OsARD1) within the rice genome, which belong to the TRR14 and ARD subfamilies, respectively (Figure 4 and Table S3). Subsequently, we analyzed tandem duplication events within the OsCupin gene family and detected seven tandem-duplicated gene clusters on chromosomes 2, 3, 8, 9, and 12, all of which are classified into the GLP subfamily (Figure 4 and Table S3).

We also examined the orthologous Cupin gene pairs between rice and Arabidopsis/sorghum. Our analysis showed three orthologous gene pairs between AtCupins and OsCupins, establishing three syntenic relationships between Arabidopsis and rice (Figure 5). These three Cupin genes in rice belonged to the GLP-2, ARD, and TRR14 clades, respectively. Additionally, we found twenty-three orthologous gene pairs between OsCupins and SbCupins, leading to twenty-nine syntenic relationships between rice and sorghum (Figure 5). Of the twenty-three Cupin homologous genes in rice, four, three, three, four, and one were from the GLP1 to GLP5 clades, respectively. The remaining eight belonged to the TRR14 clade, four were from the PIRIN clade, and two were from the ARD clade.

3.6. Promoter Region Cis-Acting Regulatory Elements Analysis

In total, we identified 40 functional cis-elements in 54 OsCupins, which are related to biological and abiotic stress, plant growth and development, hormone responsiveness, and light responsiveness (Figure 6A–C). The majority of these elements were associated with biological and abiotic stress (38.01%), followed by hormone-responsiveness-related elements (26.48%). Additionally, 11.76% were linked to plant growth and development, and 23.75% were related to light responsiveness (Figure 6C).

We identified ten cis-elements that are responsive to various stresses, including drought-responsive elements (40.10%, MYC/as-1), stress-responsive elements (21.29%, STRE/TC-rich repeats), and anoxic-responsive elements (12.75%, ARE). Additionally, we found elements related to damage (WRE3, 6.68%), an MYB binding site involved in drought inducibility (MBS, 6.19%), dehydration-responsive elements (DRE core, 4.83%), biotic stress elements responsive to wounding and pathogens (WUN-motif, 4.46%), and low-temperature-responsive elements (LTR, 3.71%). Furthermore, we discovered nine cis-elements that are responsive to five different plant hormones. These include an abscisic acid responsiveness element (ABRE, 29.48%) found in 45 OsCupins, MeJA-responsive elements (CGTCA-motif/TGACG-motif, 47.25%) in 46 OsCupins, gibberellin-responsive elements (GARE-motif/P-box, 5.86%) in 25 OsCupins, a salicylic-acid-responsive element (TCA-element, 3.73%) in 40 OsCupins, auxin-responsive elements (TGA-element/AuxRR-core, 6.93%) in 31 OsCupins, and estrogen-responsive elements (ERE, 6.75%) in 25 OsCupins (Figure 6C).

3.7. Spatial Expression Profiles of OsCupin Genes

The tissue-specific expression profiles of OsCupins were investigated using the RNA-seq data. A total of 16 different tissues encompassing almost all rice tissues and developmental stages were analyzed. The heatmap analysis of the OsCupin gene family expression profiles revealed that genes of the same branch showed similar expression patterns; for example, most members of the GLP-1 branch were highly expressed at the seeding stage and root site (Figure 7). Members of the GLP-2 clade, such as OsGLP1, OsGLP2, and OsGLP22, showed pronounced activation in floral organogenesis stages, particularly in the young panicle, anther, and glume stages, where their expression intensities surpassed those of other family members, suggesting specialized roles in floral organogenesis (Figure 7). In contrast, members of the ARD, TRR14, and PIRIN clades exhibited differential expression across tissues, with OsARD1 and OsTRR14-2 showing ubiquitously high expression levels in all of the examined tissues (Figure 7).

3.8. Expression Pattern Analysis of OsCupin Genes Under Different Hormones

Transcriptome analysis revealed distinct expression profiles of OsCupin genes under phytohormone treatments, with ABA and JA responses showing marked divergence. Under ABA treatment, OsGLP12, OsGLP13, OsGLP14, OsGLP15, OsGLP20, OsGLP28, OsGLP29, OsGLP41, OsGLP42, and OsARD3 exhibited significant upregulation, with OsGLP14 reaching peak expression at 24 h (Figure 8). Conversely, OsGLP1, OsGLP2, OsGLP16, OsGLP19, OsGLP21, OsGLP22, OsGLP26, OsGLP27, OsGLP34, OsARD2, OsTRR14-1, and OsTRR14-3 showed progressive downregulation, with OsGLP34 displaying minimal expression at 12 h (Figure 8). JA treatment induced the pronounced upregulation of OsGLP20 (peak at 12 h) and sustained high expression of OsGLP23 and OsARD3, while OsGLP1, OsGLP2, OsGLP16, OsGLP22, OsGLP26, OsGLP34, OsGLP40, and OsARD2 were suppressed, with OsGLP34 attaining the lowest expression at 24 h (Figure 8). These differential regulatory patterns underscore functional diversification within the OsCupin family, where specific members mediate hormone-specific stress responses while others may act as broad-spectrum modulators or negative regulators.

3.9. Expression Pattern Analysis of OsCupin Genes Under Abiotic Stress

To investigate the potential regulatory roles of OsCupin genes under abiotic stress conditions, we subjected Oryza sativa japonica cv. LX1 to drought stress (20% PEG 6000) and salt stress (200 mM NaCl) treatments. Twelve OsCupin genes spanning all eight phylogenetic clades were selected for qRT-PCR analysis, with prioritization given to candidates harboring enriched stress-responsive cis-elements in their promoter regions. The qRT-PCR analysis revealed contrasting expression patterns between treatments: under NaCl treatment, three OsGLP genes (OsGLP15, OsGLP38, and OsGLP43) showed significant induction, while OsGLP9, OsGLP16, and OsGLP22 were markedly suppressed (Figure 9).

This contrasted with PEG 6000 treatment, where six OsGLP genes (OsGLP9, OsGLP15, OsGLP22, OsGLP29, OsGLP38, and OsGLP43) exhibited strong upregulation, whereas OsGLP16 remained uniquely repressed (Figure 10). Notably, genes from the ARD, PIRIN, and TRR14 families (OsARD1, OsPIRIN1, OsPIRIN2, OsTRR14-2, and OsTRR14-3) displayed consistent downregulation patterns under both stresses, suggesting that their transcriptional suppression may constitute a conserved stress-response mechanism (Figure 9 and Figure 10).

3.10. Response of pYES2 Overexpression Yeast to Abiotic Stress

Based on their pronounced and significant upregulation under both NaCl and PEG 6000 stresses, OsGLP15, OsGLP38, and OsGLP43 were prioritized for functional validation in yeast to assess their potential direct contribution to abiotic stress tolerance. We constructed a yeast overexpression system using pYES2 vector, and yeast transgene validation was performed on three candidate genes (OsGLP15, OsGLP38, and OsGLP43) that were significantly upregulated under salt and drought stress. Under non-stress conditions, yeast strains expressing these genes exhibited growth rates comparable to the empty vector control, indicating no detrimental effects on basal cellular metabolism (Figure 11A). However, under salt stress (1 M NaCl), yeast expressing OsGLP15, OsGLP38, or OsGLP43 displayed markedly enhanced growth compared to the control, with OsGLP43 demonstrating the strongest salt tolerance (Figure 11B). Similarly, under drought stress (30% PEG 6000), all three OsGLP-expressing strains outperformed the control in growth capacity (Figure 11C). These results confirm that OsGLP15, OsGLP38, and OsGLP43 confer enhanced adaptability to both salt and drought stresses in yeast.

4. Discussion

The Cupin superfamily proteins constitute a functionally diverse protein superfamily, characterized by a conserved β-barrel domain and two conserved motifs, enabling them to bind metal ions and exert a variety of physiological functions [2,3]. These functions include serving as seed storage proteins to provide nutrition for plant growth and development, participating in multiple enzymatic reactions to regulate organismal metabolism, acting as transcription factor cofactors and redox sensors in cell signaling pathways, and playing crucial roles in plant defense responses [33]. The identification and functional validation of the Cupins in rice may hold the promise to breed improved crops with excellent agronomic traits and stress tolerances to combat the challenge of global climate change. Cupin protein was initially discovered through the shared amino acid sequences between a heat-resistant protein produced early in wheat embryo germination and a stress-related globulin produced by the slime mold Physarum polycephalum [34]. Previous studies have analyzed Cupin gene families in various plant species, such as soybean and oilseed rape, identifying 69 and 173 members, respectively. However, the comprehensive characterization of the Cupin gene family in rice remains largely unexplored. Leveraging the high-contiguity telomere-to-telomere (T2T) genome assembly of rice (AGIS-1.0), this study provides the first comprehensive genome-wide identification and characterization of the Cupin superfamily in this critical crop. We identified 54 Cupin genes (OsCupins) and classified them into four major subfamilies—GLP (44 members), PIRIN (3), ARD (3), and TRR14 (4)—based on phylogenetic relationships and conserved domain architectures. In the present study, utilizing the latest T2T reference genome, we conducted an identification of the Cupin superfamily, including the OsGLP family, and identified a total of 44 OsGLPs. This quantity is one more than that reported previously using IRGSP-1.0 (43) [35], and the newly discovered OsGLP19 is located on chromosome 4. Therefore, the application of the rice T2T genome (AGIS-1.0) has facilitated the discovery of genes that were previously missed due to incomplete genome assemblies, providing a solid foundation for rice genetic breeding and functional genomics research. The physicochemical property analysis of Cupin proteins revealed a wide range of molecular weights and isoelectric points, suggesting diverse functions across different species and conditions. The phylogenetic analysis of the 54 Cupin genes identified eight distinct groups, with GLP-1 having the largest number of members. In addition, the GLP-1 to GLP-5 clades together form the largest subfamily, the OsGLP family. Conserved motifs and gene structure analyses revealed significant differences among the different subfamilies within the Cupin superfamily, yet the conserved motifs within each subfamily were highly preserved. The gene structures of the GLP and TRR14 subfamilies were relatively simple, with most members containing only 1–2 exons, whereas the gene structures of the ARD and PIRIN subfamilies appeared more complex and disordered. Gene duplication serves as a primary driver for gene family expansion, enabling plants to diversify their genetic repertoire, thereby facilitating species differentiation and enhancing environmental adaptability [36]. Our genome-wide analysis identified 54 Cupin genes in rice, significantly exceeding the number identified in Arabidopsis (39 genes), suggesting lineage-specific expansion of this gene family during rice evolution. Tandem duplication events are widespread in plant genomes. Under strong selective pressure, these events frequently yield paralogs specialized in defense-related functions [37]. Notably, 28 OsCupin genes (51.85% of OsCupin family) were characterized as tandem duplicates, all belonging to the GLP subfamily. This pattern indicates these duplications occurred after the divergence of rice from other species, potentially driving functional diversification of Cupin genes in stress adaptation and pathogen resistance mechanisms. Although GLPs are known to function enzymatically in diverse physiological processes, the enzymatic activities of most GLPs remain poorly defined. Previous crystallographic studies revealed that GLPs typically possess a single metal-binding active site coordinated by three conserved histidine ligands and one glutamic acid ligand [38]. Notably, the four highly conserved active site residues exhibit substitution flexibility. For instance, the glutamic acid residue in Motif 1 of Arabidopsis ATH6 (Athaliana 6) can be replaced by aspartic acid or glutamine, and the third histidine residue in Motif 2 of Pharbitis nil PNI can be substituted with aspartic acid [2]. Our in subcellular localization predictions (Table S2) suggest diverse cellular compartments for OsCupins, aligning with their potential multifunctional roles. Experimental validation of these predictions will be crucial to pinpointing the precise subcellular context of OsCupin function during stress adaptation.

The analysis of cis-acting regulatory elements in the promoter region has yielded invaluable insights into the transcriptional regulation and potential functions of OsCupin genes. The presence of elements related to plant hormone regulation and stress response further suggests that OsCupin genes may play a role in hormone signaling and stress adaptation. The expression profiles of OsCupin genes in response to hormone treatments underscore their potential involvement in hormone signaling pathways. Notably, the elevated expression of OsGLP14 and OsARD3 under ABA treatment hints at their potential roles in ABA-mediated stress responses. Similarly, the increased expression of OsGLP23, OsGLP33, and OsARD3 in response to JA treatment indicates their likely involvement in JA-mediated defense responses. Collectively, these findings imply that OsCupin genes may serve as integrators of various hormone signaling pathways, regulating both plant growth and stress responses. Numerous OsCupin genes were found to play crucial roles in abiotic stress responses in plants such as Arabidopsis and Tamarix hispida. For instance, the heterologous expression of Arabidopsis thaliana AtGER1, AtGER2, and AtGER3 in tobacco confers enhanced thermotolerance under high-temperature stress conditions [39]. Transcript levels of ThGLP from Tamarix hispida were significantly upregulated by multiple abiotic stresses, including drought, salinity, and low-temperature exposure [40]. This study utilized the japonica cultivar LX1 for expression analyses. While this provides a consistent genetic background, it limits the generalizability of our findings across diverse rice germplasm. To assess the breeding relevance and allelic variation in OsCupin genes, future investigations should compare their expression patterns, allelic sequences, and association with stress tolerance phenotypes across a panel of rice genotypes exhibiting contrasting sensitivities to drought and salinity. The heterologous expression of OsGLP15, OsGLP38, and OsGLP43 in yeast significantly improved stress tolerance under salt and drought conditions, suggesting their potential roles in modulating osmotic balance or ion homeostasis during stress responses in rice. Notably, OsGLP43 exhibited the most pronounced protective effects, highlighting its promise as a target for improving stress resilience in crops. Based on conserved OsGLP functions, these genes may mitigate oxidative damage via SOD-like activity to scavenge ROS. They could also maintain osmotic balance by facilitating osmolyte accumulation or reinforcing cell walls against water loss. Additionally, their promoter ABRE/MYB elements and stress-responsive expression imply integration with ABA signaling to regulate ion homeostasis or stomatal closure. Although our yeast heterologous expression system provides compelling preliminary evidence for the functional involvement of OsGLP15, OsGLP38, and OsGLP43 in abiotic stress tolerance, it is crucial to acknowledge the inherent limitations of this model. Yeast cells lack the complex multicellular architecture, tissue-specific signaling, and developmental programs inherent to plants. Therefore, the precise physiological roles and regulatory networks of these OsGLP genes in planta, particularly within the context of rice development and under field-relevant stress conditions, remain to be definitively established. Future work will focus on generating and characterizing transgenic rice lines (overexpression and/or CRISPR/Cas9 knockout mutants) to confirm the biological significance of these findings in the target organism, determining the transcriptional targets through ChIP technology, and elucidating the molecular mechanism of these targets within the complex regulatory network.

5. Conclusions

In this study, we conducted a systematic genome-wide analysis of the Cupin gene family in rice, identifying 54 OsCupins and classifying them into four subfamilies based on phylogenetic relationships and conserved domain architectures. Tandem and segmental duplication events, particularly on chromosomes 3 and 8, contributed to the expansion of this family. The structural conservation of motifs and exon–intron organization within subfamilies underscored functional specialization during evolution. The synteny analysis revealed orthologous relationships with sorghum and Arabidopsis, highlighting evolutionary conservation across monocots and dicots. The promoter analysis identified abundant stress-responsive cis-elements, suggesting roles in abiotic stress adaptation. Specifically, genes such as OsGLP9, OsGLP15, OsGLP22, OsGLP29, OsGLP38, and OsGLP43 were found to be differentially expressed under various stress conditions, indicating their potential roles in stress adaptation. Functional validation in yeast demonstrated that OsGLP15, OsGLP38, and OsGLP43 significantly enhance salt and drought tolerance in this heterologous system, highlighting their potential as candidate genes for improving stress resilience. However, confirming their essential roles and elucidating their precise mechanisms within rice require further genetic and physiological studies using transgenic approaches. This study advances our understanding of the molecular evolution and stress-responsive regulatory networks of Cupin genes in rice, providing a foundation for future functional studies and genetic engineering to improve rice resilience under adverse environmental conditions.