2.1 Structure and phylogenetic analyses of the AtRecQl genes

The genomic DNA, coding, and amino acid sequences of previously reported (Hartung et al., 2000) AtRecQl genes were retrieved from the Arabidopsis Information Resource (TAIR) (https://www.arabidopsis.org/) database in FASTA format. The corresponding AtRecQl protein sequences were used as queries in BLASTP searches to examine their phylogenetic relationships with RecQls in other crop plants. The retrieved protein sequences were also used for multiple sequence alignments using ClustalW with default parameters. A phylogenetic tree was constructed using the MEGA X software and the unrooted neighbor-joining method.

2.2 Analysis of cis-elements and gene annotation

The in silico promoter analysis considered one kilobase (kb) upstream of the translation start site of each AtRecQl gene, with sequences obtained from the TAIR database. The cis-elements were identified using three online cis-elements tools—Arabidopsis Gene Regulatory Information Server (AGRIS; http://arabidopsis.med.ohio-state.edu/AtcisDB/), PlantCare (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/), and Plant Cis-Acting Regulatory DNA Elements (PLACE; https://www.dna.affrc.go.jp/PLACE/)—and compared among tools. The AtRecQl genes were annotated with gene ontology (GO) terms using the TAIR search tool (https://www.arabidopsis.org/tools/bulk/go/index.jsp).

2.3 Protein-protein interaction predictions

An interaction network was constructed for the AtRecQl proteins using the STRING website (http://www.string-db.org). The protein-protein interaction (PPI) map of the AtRecQls was predicted with the confidence parameter set to 0.15 using the threshold/molecular action mode with K means clustering.

2.4 Transmembrane, solubility, and thermodynamic properties

Transmembrane topology was predicted using DeepTMHMM (https://dtu.biolib.com/DeepTMHMM). Protein solubility was predicted using Protein-Sol (https://protein-sol.manchester.ac.uk). Thermal resilience was assessed using the SCooP server (http://babylone.3bio.ulb.ac.be/SCooP/index.php).

2.5 Molecular modeling and validation

The secondary structural characteristics of the AtRecQl protein sequences were computed using SOPMA (https://npsa-prabi.ibcp.fr/NPSA/npsa_sopma/) with 17 window widths, 8 similarity thresholds, and 4 conformational states. A hypothetical 3D model was predicted for the protein sequence via comparative homology modeling using SWISS-MODEL (https://swissmodel.expasy.org/). The quality of the modeled protein structures was assessed using MolProbity (http://molprobity.biochem.duke.edu/) assessing. Inaccuracies within the predicted 3D structures were detected based on the model’s atomic coordinates using ProSA-web (https://prosa.services.came.sbg.ac.at/prosa.php).

2.6 MD simulations

The AtRecQl proteins were assessed with time-dependent biological conditions for 100 ns using the Yasara Structure software. The protein residue ionizations were set at pH 7.4. A cubic box, located 10 Å away from all atoms, was solvated with the TIP3P water model and 0.9% NaCl to simulate a physiological environment. The MD simulations used the AMBER14 force field for the solute, GAFF and AM1-BCC charges for the ligands, and TIP3P for water molecules. The temperature was maintained at 298 K under a pressure of 1 atm (NPT ensemble) using a Berendsen barostat and thermostat, with time-averaged temperature and density coupling. Non-bonded long-range interactions were computed using the Particle Mesh Ewald algorithm.

3.1 Evolutionary analysis of RecQls

The domains of AtRecQl proteins and structures of AtRecQl genes with chromosomal locations were analyzed to obtain information about their evolution (Supplementary Fig. S1). An unrooted phylogenetic tree was constructed among the model plant A. thaliana and nine crop plants (Brassica napus, Camelina sativa, Raphanus sativus, Gossypium hirsutum, Citrus sinensis, Momordica charantia, Solanum tuberosum, Glycine max, and Oryza sativa) to understand the evolutionary connection and predict the possible functions of RecQls in different crops (Supplementary Fig. S2 and Supplementary Table S1). The 44 RecQl proteins of ten plants were partitioned into four groups representing RecQl1, RecQl2, RecQl3, and RecQl4. Two copies of RecQl4 (like AtRecQl4A and AtRecQl4B) were presented only in three crops: B. napus, C. sativa, and R. sativus. Phylogenetic analysis revealed that RecQl2 and RecQl4 were closer among the four groups, consistent with the Simple Modular Architecture Research Tool (SMART) domain structure analysis (Supplementary Fig. S1A). As predicted, the O. sativa RecQls were more distant from all other plant RecQls, except the RecQl2 groups.

3.2 Cis-element analysis of the promoters of AtRecQls

Both reverse and forward strands 1 kb upstream of the five AtRecQl genes were scanned for cis-elements using three programs: PlantCare, AGRIS, and PLACE. Among all AtRecQl genes, development and stress-responsive elements were the most common (Supplementary Fig. S3A). Cis-elements associated with different stress responses, such as dehydration-responsive element, low-temperature-responsive element, early responsive to dehydration ACGTATERD1, and heat shock element (HSE) CCAAT BOX1, were significantly evident in the AtRecQl promoter regions . Drought- and abscisic acid-related MYC and different stress and cell-cycle-related MYB cis-elements were found in all AtRecQl genes. Salicylic acid-induced WRKY DNA binding site W-box/WB-box, abscisic acid-responsive elements (ABREs), auxin response element ARFAT were also common in the AtRecQl promoter regions. Furthermore, light-responsive cis-elements GT1, GATA box, cytokinin-regulated transcriptional activator ARR1, and various organ-specific cis-elements were found in the AtRecQl promoter regions. The AtRecQls promoter regions contained stress, hormone, cell cycle, and development-responsive cis-elements, indicating that these genes are involved in various events in plant growth and development.

3.3 GO annotation of AtRecQls

The five AtRecQl genes were annotated with GO terms using the TAIR database. All three categories of GO enrichment terms (cellular component, molecular function, and biological process) were examined (Supplementary Fig. S3B). Four GO terms were predicted under cellular component categories, of which cytoplasm, nucleus, and other intracellular components were identified for all five AtRecQl genes, whereas other cellular components was identified for only one gene. Four GO terms were also predicted under molecular function categories for the five AtRecQl genes, of which catalytic activity and hydrolase activity were identified for all five genes, whereas nucleic acid binding was identified for four and protein binding for only one. The AtRecQls are involved in various biological processes. The GO annotations revealed that the five AtRecQls are dedicated to DNA metabolic, biosynthetic, cellular components organization, and other cellular and metabolic processes. GO terms related to different stress responses, such as responses to stress, abiotic stimuli, chemicals, and endogenous stimuli, were significantly represented. These results indicate that AtRecQls are involved in various abiotic stresses.

3.4 PPI predictions

The PPI prediction results for the AtRecQls are shown in Fig. 1. The network was divided into three groups. According to the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, the 25 proteins constituting the network were present in three main KEGG pathways: 11 in HR, 4 in non-homologous end joining, and 2 in mismatch repair (Supplementary Fig. S4). The top ten predicted functional partners were FANCM, MUS81, TOP3α, UVH1, AT5G19950, ATR, XRCC3, MSH5, AT1G29630, and MRE11 (Supplementary Fig. S5). The FANCM had the highest score and interacted with all examined AtRecQl proteins. The AtRecQls exhibited significant co-expression with FANCM, TOP3α, ATR, MSH5, and AT1G29630 (Supplementary Fig. S6). The GO terms of these 25 proteins were also examined (Supplementary Fig. S7, S8, and S9), indicating that most are DNA-related. Some proteins in this network are also involved in cellular responses to various abiotic stresses stimulus.

Close

Fig. 1

The interaction network of RecQl proteins in A. thaliana. The predicted protein–protein interaction map of AtRecQls was generated with 3 clusters of K-means clustering (clusters-1 is brown, clusters-2 is dark blue, and clusters-3 is green). The “minimum required interaction score” was set to 0.9 (highest confidence); other boxes used default settings.

Export to PPT

3.5 Transmembrane, solubility, and thermodynamic properties of the AtRecQls

The transmembrane analysis indicated the absence of transmembrane α-helices and signal regions, while all AtRecQl proteins exhibited an internal localization. The distribution of protein types and regions suggested a predominant protein localization pattern within cellular compartments.

The predicted solubility was 0.521 for AtRecQl1, 0.500 for AtRecQl2, 0.447 for AtRecQl3, 0.447 for AtRecQl4A, and 0.395 for AtRecQl4B (Table 1); proteins with scaled solubility values >0.45 are expected to be soluble, while those with values <0.45 are expected to be less soluble (Niwa et al., 2009). Therefore, AtRecQl1 and AtRecQl2 are expected to be soluble since their scaled solubility values are above this threshold. However, the lower predicted solubilities for AtRecQl3, AtRecQl4A, and AtRecQl4B imply potential structural instability or aggregation tendencies, which could impact their functionality.

The changes in enthalpy (ΔHm) for the AtRecQls were −17.7 to −48.8 kcal/mol (Table 1). Similarly, their changes in heat capacity (ΔCp) were −0.53 to −1.35 kcal/(mol K). Their melting temperatures (Tm) and changes in Gibbs free energy (ΔGr) were 58.7°C to 64.3°C and −0.8 to −3.6 kcal/mol, respectively. The temperature ranges over which these proteins underwent ΔGr were −6.07–77.52°C for AtRecQl1, −22.29°C–74.49°C for AtRecQl2, −14.72°C–73.67°C for AtRecQl3, −6.12°C–78.99°C for AtRecQl4A, and −53.56°C–79.28°C for AtRecQl4B. AtRecQl3 exhibited the highest ΔHm of −48.8 kcal/mol and ΔGr of −2.6 kcal/mol, suggesting its structure is more stable than the other examined proteins (Pucci and Rooman, 2016). AtRecQl4B also showed a high ΔHm of −46.5 kcal/mol) and ΔGr of −3.6 kcal/mol, indicating high thermodynamic stability. In contrast, AtRecQl1 and AtRecQl2 had lower ΔHm of −17.7 and −26.5 kcal/mol, respectively, and relatively smaller ΔGr of −0.8 and −1.5 kcal/mol, respectively, implying comparatively lower stability. AtRecQl4A fell between these groups, with a moderate ΔHm of −22 kcal/mol and ΔGr of −1.0 kcal/mol. Moreover, AtRecQl4A and AtRecQl4B had higher Tm of 64.0°C and 64.3°C, respectively, compared to 62.5 for AtRecQl1, 59.5 for AtRecQl2, and 58.7 for AtRecQl3, suggesting greater thermal stability (Pucci and Rooman, 2016). The temperature range also varied, with AtRecQl4B exhibiting the widest range of −53.56°C–79.28°C, followed by AtRecQl1, AtRecQl4A, AtRecQl3, and AtRecQl2, in descending order.

3.6 Structural analysis of AtRecQl proteins

The alpha helix content was highest for AtRecQl2 (48.51%) and lowest for AtRecQl4A (37.04%) (Fig. 2 and Table 2). The beta-sheet content was highest for AtRecQl1 (14.85%) and lowest for AtRecQl4A (9.09%). The turn and coil contents also varied among the proteins, with AtRecQl4A having the highest coil content (48.23%) and AtRecQl1 the lowest (35.64%).

Close

Fig. 2

The secondary structure of retrieved AtRecQl amino acid sequences by SOPMA server. The blue, red, green and purple color represents the helix, sheet, turn and coil structure of the proteins.

Export to PPT

Regarding all-atom contacts, the Clashscore values indicated relatively low steric clashes, with AtRecQl2 having the highest and AtRecQl1 the lowest values (Chen et al., 2015) (Supplementary Fig. S10 and Supplementary Table S2). Protein geometry analysis revealed that all proteins had high percentages of favored rotamers, ranging from 92.20% to 96.03%. AtRecQl2 had the lowest percentage of poor rotamers (1.09%) and the highest percentage of Ramachandran favored (94.58%). AtRecQl4A had the highest percentage of Ramachandran outliers (1.09%), while AtRecQl1 had the highest percentage of favored rotamers (96.03%). Ramachandran analysis highlighted minor differences among the proteins, with AtRecQl2 having no outliers, while AtRecQl4A and AtRecQl4B had slightly higher percentages of Ramachandran outliers (Ravikumar et al., 2019). The Rama distribution Z-scores indicated acceptable backbone conformations for all proteins (Supplementary Fig. S11).

3.7 MD simulation and principal component analysis

The root means square deviations (RMSD) for the five AtRecQls were calculated to assess their structural variability. The average RMSD varied from 0.3399 to 0.4579 nm, with AtRecQl1 exhibiting the lowest average RMSD of 0.3399 nm and AtRecQl4 showing the highest average RMSD of 0.5422 nm (Fig. 3). The maximum RMSD varied considerably across proteins, with AtRecQl2 having the highest maximum RMSD of 1.3171 nm at 7.5 ns and AtRecQl1 having the lowest maximum RMSD of 0.3884 at 29.75 ns. Similarly, the minimum RMSD values also varied, with AtRecQl2 having the lowest minimum RMSD of 0.1238 nm at 5 ns and AtRecQl1 having the highest minimum RMSD of 0.1528 nm at 7.5 ns. This observation indicated that the AtRecQl proteins had stable carbon backbone conformations throughout the simulation period (Sargsyan et al., 2017). The average radius of gyration (RG) ranged from 2.219 to 2.764 nm across the proteins, with AtRecQl2 having the highest average RG of 2.842 nm and AtRecQl1 having the lowest average RG of 2.219 nm.

Close

Fig. 3

The structural characteristics of RecQ proteins. The MD trajectories were scrutinized utilizing RMSD for assessing structural stability and RG for analyzing rigidity and compactness. Visualization of MD plots was accomplished using Matplotlib v3.7.

Export to PPT

A principal component analysis (PCA) score plot was used to measure and identify related motions and dynamic regions within the protein structure to assess fluctuations, changes in atomic locations, and flexibility (Lang et al., 2009). The contributions of eigenvectors to the overall conformational variability of AtRecQl1, AtRecQl2, AtRecQl3, AtRecQl4A, and AtRecQl4B were determined to be cumulative variance ratios of 45.65% (PC1 = 35.37%, PC2 = 11.24%), 79.95% (PC1 = 58.29%, PC2 = 21.66%), 50.8% (PC1 = 36.13%, PC2 = 14.68%), 56.53% (PC1 = 37.8%, PC2 = 18.73%), and 40.24% (PC1 = 22.42%, PC2 = 17.82%), respectively (Fig. 4). These values suggested limited coordinate movements, indicating stability throughout the MD simulation.

Close

Fig. 4

The PCA and the PDF plot of RecQ proteins. A, B, C, D and E represent the RecQl1, RecQl2, RecQl3, RecQl4A and RecQl4B, respectively. Left panel represents PCA where each point on the scatter plot represents the conformation of the complexes along the axes. The color gradient from purple (initial timestep) to teal to yellow (final timestep) corresponds to different time points within the simulation. The right panel shows the PDF plots depicting the distribution of RMSD and RG values for RecQ proteins.

Export to PPT

The free energy landscape (FEL) represents the intricate interplay of elements. It reveals the complexes’ most energetically stable states, transition pathways, and widespread dynamic motion (Parra-Cruz et al., 2018). Regions with lower RG and RMSD demonstrated packed and stable structure conformations (Fig. 5); when associated with reduced Gibbs free energy (GFE), these regions signified thermodynamically favorable states. Conversely, higher RMSD suggested structural fluctuations, potentially corresponding to less stable conformations. Increased GFE in such regions indicated higher energy stateand less favorable circumstances. Energy minima, depicted as valleys, represented considerably stable conformations.

Close

Fig. 5

3D and 2D FEL depictions of RecQ, alongside RMSD and RG structural parameters. A, B, C, D and E represent the RecQl1, RecQl2, RecQl3, RecQl4A and RecQl4B, respectively. The left panel shows the two dimensional and the right panel illustrates their corresponding three-dimensional energy surface.

Export to PPT