Phytozome database was used to obtain sequences of DNA2 proteins (Zhang et al., 2012). In the rice genome (Oryza sativa IRGSP-1.0), the Hidden Markov Model (HMM) profiles of DNA2 domain from the Pfam (protein family) database were used to scan the predicted proteins using HMMERv3 (Prince and Pickett, 2002). By using HMM model in HMMERv3, the protein sequences of rice DNA2 were aligned. For the confirmation of the presence of DNA2 conserved domain, the putative DNA2 gene core sequences were verified by searching against the SMART (http://smart.embl-heidelberg.de/) and Pfam database (https://pfam.xfam.org/). Protein sequences of Zea mays, Hordeum vulgare, Pennisetum glaucum, and Oryza sativa were obtained from previous studies (Liu and Widmer, 2014; Guo et al., 2017), and TAIR (https://www.arabidopsis.org/) source was used to download information of Arabidopsis DNA2 gene family protein sequences and annotation. ExPASy (https://www.expasy.org/) online server was used to obtain molecular weight, GRAVY, Iso-electric point information for OsDNA2 (Gasteiger et al., 2003).

Rice genomic database in phytozome was accessed to get the DNA2 genes genomic coordinated on rice chromosomes. All OsDNA2 were present on the eight chromosomes of rice genome. Protein sequences of DNA2 in Arabidopsis, rice, maize, barley, and millet were aligned using clustalW. Bootstrap 1000 replications were used to generate phylogenetic tree using maximum likelihood (ML) method in MEGA 10. Coding sequence was compared with the corresponding full-length sequence for the identification of intron insertion sites in the DNA2 genes. This identification was performed by using Gene Structure Display Server (Guo et al., 2007). The analysis of conserved DNA2 motifs was performed by using MEME Suit (Multiple EM for Motif Elicitation) Version 4.12.0 (http://meme-suite.org/tools/meme) (Rehman et al., 2022). Following parameters were used for this analysis: ten motifs to be found with motif width between 10 and 200; site distribution was one occurrence per sequence or set at zero (at most one occurrence for each motif was allowed for each sequence), whereas the maximum number of motifs was set to 10 (Bailey et al., 2009). TBtools (https://bio.tools/tbtools) program was used for further analysis of MEME results (Chen et al., 2020). Unipro UGENE software package (Okonechnikov et al., 2012), helped to examine the conserved domains of OsDNA2 proteins. This aligned the sequences by the ClustalW algorithm and conserved regions were displayed in the form of color patterns which differentiated each amino acid based on physiochemical properties. OsDNA2 protein sequences in SMART database containing Pfam domain search options, was used to perform protein domain analysis, and confirmation was carried out through the InterPro database (Letunic et al., 2012; Blum et al., 2021).

MicroRNAs (miRNAs) interacting with the DNA2 genes were predicted form the available rice miRNA reference sequences by submitting genome sequences of OsDNA2 to the psRNATARGET server (https://www.zhaolab.org/psRNATarget/). The visualization of miRNAs and OsDNA2 gene was done with Cytoscape software (https://cytoscape.org/). Online tool gProfiler (https://biit.cs.ut.ee/gprofiler/gost, accessed on 20 July 2022) was used to conduct gene ontology (GO) analysis of OsDNA2 protein sequences with default parameters. The rice genome database was downloaded from RGAP (https://cottonfgd.org/) to get the DNA2 promoter region sequence containing 2000 bp upstream of the inhibition codon (ATG). The prediction of regulatory elements in the DNA2 promoter regions was carried out by using online tool PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) and visualized by TBtools. For the sequence similarity patterns evaluation, synteny analysis and sequence identity visualization were performed using the TBtools (Rahim et al., 2022). For gene duplication, homologous gene pair were calculated with the help of Ka/Ks calculator 2.0. For Synteny analysis, genome sequence (FASTA) and annotation files (gff/gtf) were used in One-step MCScanX toolkit in the TBtools.

CELLO v.2.5: subCELlular LOcalization predictor was used to predict the sub-cellular location of DNA2 family (Yu et al., 2014). Protein sequences were used as input and output results were further analyzed/visualized by using TBtools software. Amino acid sequences of OsDNA2 proteins were used for the prediction of 3D structures by utilizing the SWISS-MODEL database (https://www.swissmodel.expasy.org) (Waterhouse et al., 2018), while visualization of such predicted structures was carried out with the help of Pymol software (https://pymol.org/2/). Ramachandran Plot—Zlab, (https://zlab.umassmed.edu/bu/rama/), was applied for the confirmation of predicted 3D models of OsDNA2 proteins (Anderson et al., 2005).

RNA seq data for drought and salt stresses were assessed from online data bases which are publically available with Bio-projects GSE145869 and GSE167342 (Tarun et al., 2020; Bundó et al., 2022). In these studies, Nil-95 and Swarna genotypes were used as drought tolerant and sensitive, while IL22 and PL12 were used as a salt tolerant and sensitive rice genotype, respectively. FPKM (fragments per kilobase of transcript per million mapped reads) values were extracted and heatmaps were generated in TBtools. Furthermore, the rice genotype IR-6 was used in this study under drought and salt stresses. Plants were grown under normal conditions for 2 weeks. Then drought and salt stresses were applied to two batches of plants and one batch was kept as a control. Three biological replicates for each treatment were used. Gene specific primers of selected OsDNA2 genes along with drought and salt reported genes (OsEm1 and bZIP23) for qRT-PCR are presented in Supplementary Table S1. For this purpose, 1 g leaf tissues from control and treated samples were grinded in liquid nitrogen and used for RNA extraction. Total RNA was extracted with the help of TRIzol method. Complementary DNA (cDNA) was also synthesized from the 800 ng extracted RNA with the help of reverse transcriptase-III, first strand cDNA Synthesis Kit (K1691, Thermo Scientific Revert Aid). StepOne RT-PCR (Applied Biosystems^® 7900 H T Fast RT-PCR) was used for quantification. OsActin was used as a reference gene and ^2−ΔΔCT method was used for expression calculations (Uzair et al., 2021).

Wheat DNA2 protein (TraesCS2A02g301600) and Arabidopsis (AT2G03270) sequences were used as queries to identify DNA2 genes in the rice genome. From these analysis, 17 OsDNA2 members were found (Figure 1A, Supplementary Table S2). OsDNA2 was distributed on the basis of its physical position on eight chromosomes (2, 3, 4, 6, 7, 9, 10, and 11). Of these chromosomes, Chr4 had the maximum number of six OsDNA2 genes, followed by Chr3, which had three genes. Chr2 and Chr10 each had two members, whereas the rest of the chromosomes had a single member. Chr9 was the shortest one which had OsDNA2_5. This structure plays an important role in the expression of genes. For this purpose, we also checked the gene structures and found that the number of introns varied from 1 to 32 (Figure 1B). The maximum number of introns was found in OsDNA2_1, whereas the minimum was found in OsDNA2_11. The CDS length extended from 399 bp (OsDNA2_11) to 3624 bp (OsDNA2_1). Similarly, the protein length, protein molecular weight, and number of exons varied from 133 aa (OsDNA2_11) to 1208 aa (OsDNA2_1), 132.72854 KDa (OsDNA2_1) to 14.19668 KDa (OsDNA2_11), and 2 (OsDNA2_11 and OsDNA2_3) to 33 (OsDNA2_1), respectively (Table 1). A total of five genes (OsDNA2_1, OsDNA2_2, OsDNA2_4, OsDNA2_7, and OsDNA2_17) out of 17 were found on the positive strand. The isoelectric points varied between 4.2510 (OsDNA2_11) and 11.6011 (OsDNA2_3). The charge on a protein molecule depends on the ionizable groups and their pKa values. The protein becomes negatively charged when the pH becomes higher than the pI. In the present study, only four OsDNA2 genes (OsDNA2_2, OsDNA2_8, OsDNA2_10, and OsDNA2_11) were negatively charged (Table 1). All members of the OsDNA2 gene family, except OsDNA2_7, showed negative GRAVY, indicating that they are hydrophilic in nature (Table 1). Additionally, the subcellular localization of all OsDNA2 proteins was determined. The results showed that most of the genes were located in the nucleus, cytoplasm, mitochondria, and chloroplasts (Figure 1C). Meanwhile, the subcellular localization of OsDNA2_16 was predicted in the peroxisomes (Figure 1C).

When characterizing newly identified proteins, it is very important to understand the motifs and domains of that specific protein. In the present study, motifs were predicted for all OsDNA2 genes using MEME (Bailey et al., 2009). For this purpose, we used the protein sequence of OsDNA2 and found ten conserved motifs (Figure 2A). The lengths of the predicted motifs varied between 20 and 39 amino acids (Supplementary Figure S1). Motif five was present in all genes except OsDNA2_2, OsDNA2_6, and OsDNA2_12. Motifs 1, 2, 3, and six were conserved among all members. Similarly, we also examined the domains in OsDNA2 members (Figure 2B). The conserved domain in the OsDNA2 genes of rice is the AAA_11 superfamily (Pfam: PF13086), which belongs to the DEAD-like helicase superfamily involved in the unwinding of ATP-dependent RNA or DNA.

In the current study, the evolutionary associations among OsDNA2s, ZmDNA2s, SbDNA2s, HvDNA2s, and AtDNA2s were assessed (Figure 3). The results revealed that 99 DNA2 molecules were clustered into six main clusters (C1 = pink, C2 = blue, C3 = yellow, C4 = green, C5 = brown, and C6 = purple). Cluster six contained a maximum of six OsDNA2 genes (OsDNA2_2 and OsDNA2_4–8). Interestingly, Cluster five contained only one, OsDNA2_1. Overall, OsDNA2 showed a closer association with ZmDNA2, HvDNA2, and SbDNA2 than with AtDNA2 in each cluster.

To estimate the evolutionary relationship among DNA2 members of Oryza sativa, a synteny analysis of DNA2 protein sequences was conducted (Figure 4A and Supplementary Table S2). These analyses were performed to study gene duplication using TBtools for the 17 predicted OsDNA2 (Chen et al., 2020). For gene duplication, Ka/Ks values were calculated, and the results revealed that six gene pairs, including OsDNA2_1-OsDNA2_8, OsDNA2_2-OsDNA2_14, OsDNA2_5-OsDNA2_6, OsDNA2_10-OsDNA2_11, OsDNA2_12-OsDNA2_13, and OsDNA_16-OsDNA2_17 could be duplicated in the rice genome (Supplementary Table S3). Similarly, we also performed collinearity analysis among Oryza sativa, Sorghum bicolor, Zea mays, and Hordeum vulgare. The results revealed that rice DNA2 genes were more collinear with sorghum than with maize and barley, indicating that whole genome or segmental duplication was involved in OsDNA2 gene family progression (Figure 4B).

It was previously reported that the promoter region is the control center for the expression and regulation of genes (Rasool et al., 2021). Promoters are also known as cis-regulatory elements in DNA. The 2-kb upstream region of each OsDNA2 gene was subjected to the PLACE database, and the results revealed that more than 85 different types of cis-acting elements and nine unnamed types of elements were detected (Figure 5, Supplementary Table S4). CAAT-box, TATA-box, and unnamed were the most identified elements for the OsDNA2 genes. Furthermore, different stress-related regulatory elements, such as CGTCA-motif, G-box, Sp1, GATA-motif, I-box, GT1-motif, and AT-rich, were detected. Similarly, hormone-related cis-regulatory elements, such as TATC-box, ABRE, CGTCA-motif, P-box, and TGACG-motif, were also detected. ABRE was detected in all members of OsDNA2, except for OsDNA2_6 (Figure 5). The GCN4-motif, GC-motif, O2-box, and ARE were detected in the different OsDNA2 members, indicating that these genes are involved in cellular development. Notably, it has been reported that the GCN4-motif and O2-box are involved in the expression of endosperm and zein metabolism, respectively.

GO annotation was used for the practical investigation of OsDNA2 genes. In silico characterization based on functions was conducted, which revealed three types of biological processes (BPs), cellular components (CCs), and molecular functions (MFs) (Figure 6, Supplementary Table S5). Further analysis of the BPs’ annotations revealed that most of the terms were related to DNA replication. Similarly, the MFs also showed DNA helicase, and catalytic and ATP-dependent activities. Based on these findings, we concluded that OsDNA2 genes play an essential role in DNA replication.

Over the past few decades, many studies have shown that miRNAs play essential roles in the regulation of genes in specific environments. Thus, to understand the post-transcriptional modification of OsDNA2 genes, we identified 653 miRNAs targeting 17 OsDNA2 genes (Figure 7, Supplementary Table S6). The results indicated that at the maximum, OsDNA2_1 and OsDNA2_4 were targeted by 74 miRNAs each, and OsDNA2_9 was less targeted (20 miRNAs). Nineteen members of the osa-miR164 family targeted four genes (OsDNA2_1, OsDNA2_4, OsDNA2_6, and OsDNA2_14). Furthermore, the single miRNAs, such as osa-miR6255, osa-miR6245, osa-miR6246, and osa-miR6248, targeted OsDNA2_10, OsDNA2_14, OsDNA2_10, and OsDNA2_4, respectively. Future studies will be required to validate the functions of these miRNAs and their target genes to understand their biological interactions in the rice genome.

Understanding the structure of proteins is very difficult due to their complexity, and the fact that they contain a different number of atoms and convoluted topology. In this study, the SWISS_MODEL online server was used to predict the three-dimensional (3D) structures of 17 OsDNA2 proteins (Figure 8, Supplementary Table S7). The results revealed the successful prediction of the 17 OsDNA2 protein models. Many reports have shown that >30% identity with the template is acceptable (Xiang, 2006; Rahim et al., 2022), and our findings also showed an average of 40% similarity with the template (5ean. 1.A, 6ff7. 1.c, 5mzn. 1.A, 2wjv. 1.A, 4b3f. 1.A, 4b3g. 2.A, 2gjk. 1.A, 2wjv. 1.A, and 2xzl. 1.A). Spiral shapes represent α-helices, thick arrows depict ß-sheets, and thin lines indicate loops and turns. Most of the OsDNA2 members had similar structures (Figure 8). Furthermore, these predicted 3D models were confirmed using Ramachandran plots with the help of diahedral angles of the OsDNA2 proteins (Anderson et al., 2005). The results of the Ramachandran plots revealed that >95% of the regions of OsDNA2 proteins showed highly favorable regions. This confirmed the better quality and stability of the predicted OsDNA2 protein structures.

RNA-seq data for drought and salt stress were retrieved from publicly available data sites. Circular heat maps for these stresses were generated from the FPKM values, showing the expression of OsDNA2 members at the seedling stage (Figures 9A,B, Supplementary Table S8). OsDNA2_11, OsDNA2_14, and OsDNA2_15 were downregulated and showed low expression under drought and salt stress conditions (Figures 9A,B). Quantitative real-time PCR (qRT-PCR) analysis of five randomly selected OsDNA2 genes was performed to study the transcription profile (Figure 9C). OsDNA2_2 and OsDNA2_5 were up-regulated under both stress conditions, whereas OsDNA2_4 and OsDNA2_7 were up-regulated under salt and drought stress, respectively. Moreover, OsDNA2_15 expression was unchanged after the treatments. These findings indicate that these genes might be involved in mitigating abiotic stresses (drought and salt), and they may provide useful information for the functional characterization of these genes in the future.