Prediction of functional consequences of single nucleotide substitutionsSIFT analysis

SIFT is a bioinformatics method that predicts how amino acid substitution affects protein function, prioritizing mutations to identify potentially harmful variants. The SIFT score of 0.05 distinguishes between harmful and tolerated amino acid substitutions; values below this level denote non-tolerant mutation, while scores above this level denote protein tolerance. The protein's capacity to function will be impacted by the D for G substitution at position 614. According to the SIFT study, its score is 0.05 indicating that this mutation has a damaging effect in (Supplementary Fig. 2) and Table S1 (Supplementary material tables).

PolyPhen-2

Human protein structure and function are affected by amino acid alterations, and PolyPhen-2 is a tool that forecasts these effects. A higher score indicates more harmful changes, whereas a lower score indicates benign ones. Its score runs from zero to one. The D614G mutation in the CFTR protein has a score of 1.000 (sensitivity: 0.00; specificity: 1.00), which suggests that it is probably damaging. A score close to 1 suggests a high probability of having a more damaging effect in Table S2 (Supplementary material tables).

PANTHER-PSEP

The PANTHER result categorized the variant as “probably damaging” and “possibly damaging” resulting in the consequence of mutation on protein. Variant D614G shows a value of 0.85 h means this mutation, is probably damaging in Table S2 (Supplementary material tables).

Prediction of harmful single amino acid substitutions on protein stabilityMUpro

To more fully comprehend genetic variants and protein functions, this machine learning method forecasts changes in protein stability brought on by single-site mutations. The delta G value obtained for mutation was less than zero i.e., − 1.734507 which signals towards decreased stability. The confidence score of the support vector machine method and the neural network was one and − 0.999645374156623 respectively, which shows that protein becomes unstable when substitution of Glycine takes place at 614 position in replace of Aspartic acid in Table S2 (Supplementary material tables).

DDGun

DDGun predicts changes in protein stability caused by mutations using evolutionary traits. A positive G value indicates a potential benefit, whereas negative G values indicate that this substitution is potentially harmful. The CFTR D614G G value of -0.2 suggests that there may be less stability because of the mutation in Table S2 (Supplementary material tables).

DUET

DUET uses the change in folding free energy (G) that comes from a protein mutation. A destabilizing mutation, according to DUET, may cause functional changes in the protein that may be significant for diseases. The mutation in CFTR is anticipated to have destabilizing consequences, as calculated by the DUET projected stability change (G), at a rate of 0.8966 kcal/mol (Supplementary Fig. 3) and Table S2 (Supplementary material tables).

SDM

A computer method called SDM (Site-Directed Mutator) was used to forecast how mutations affect the stability of proteins. It assesses the effects of single-point mutations on the stability of a protein sequence. The protein is said to be destabilized upon mutation, according to the SDM projected stability change for CFTR D614G, which was − 0.1 kcal/mol in Table S2 (Supplementary material tables).

MAESTRO web

MAESTRO is an online program that predicts alterations in protein stability caused by mutations using a multi-agent machine learning system. It was determined that the specific mutation would impact the stability of the protein. The MAESTRO site forecasted a destabilizing mutation, as indicated by the − 0.004 kcal/mol change in overall stability (G value). Furthermore, it was found that the cpred, or confidence estimation, was 0.860, indicating high reliability in Table S2 (Supplementary material tables).

DynaMut

DynaMut is a computational tool that forecasts the effects of mutations on proteins and offers information on their dynamics and stability. It analyzes the protein molecule dynamics using normal mode techniques to determine how mutations affect protein function and disease risk. Following a mutation, the expected change in Gibbs free energy was − 1.326 kcal/mol, suggesting a decrease in the stability of the CFTR protein. According to the modified protein's G ENCoM (Elastic Network Contact Model) value of − 0.147 kcal/mol, it is destabilizing. The CFTR protein may have become more flexible after mutation because of the smaller size of the amino acid substitutions. The difference in vibrational entropy energy between the wild-type and mutant forms of the CFTR protein, calculated as SVib ENCoM, was 0.184 kcal.mol-1 K-1 (Supplementary Fig. 4, 5, 6, 7,8) and Table S2 (Supplementary material tables). Calculations over the first 10 non-trivial modes of the molecule were used to calculate the atomic fluctuations, which yielded the amplitude of the absolute atomic motion. Visual representations of the atomic fluctuations are shown in Table S2 (Supplementary material tables). The atomic fluctuations of the first 10 important molecular modes of the CFTR protein were calculated. The deformation energy over the first 10 nontrivial molecular modes was calculated to determine the local flexibility of the CFTR protein.

Prediction of the structural and functional effect of single nucleotide substitutions

MutPred2 is an online tool, which is used to predict whether an amino acid substitution is pathogenic or benign. MutPred2 scored our variant D614G as 0.855, which shows this variant results in loss of helix and altered metal binding of the protein. This was ultimately altering the transmembrane protein.

Prediction of physiochemical properties of protein

The mutated as well as nonmutated CFTR proteins’ physiochemical properties were assayed using ProtParam tool, which provides valuable structural and functional properties of the protein such as molecular weight, instability index, theoretical pI, aliphatic index, and GRAVY. By adding the average isotopic masses of the given amino acids, ProtParam determines the molecular weight of the protein. The CFTR protein is projected to have a molecular weight of 168,141.57 for the normal protein and 168,083.54 for the mutant protein, which indicates that it is a rather large protein with both a complicated structure and significant biological functions. The instability index below 40 is typically regarded as a reliable indicator of protein stability. Both proteins' instability indices show that they are only moderately stable, with little variation in mutation upon projected stability. Due to the presence of aliphatic amino acids, the protein in question is likely to be thermostable, as shown by the Aliphatic Index value of 102.82 produced by ProtParam. The protein possesses a net positive charge at pH values below 8.91 and a net negative charge at pH values above 8.91, according to the theoretical pI value of 8.91 for non-mutated CFTR. The CFTR D614G variation would behave similarly, being positive at pH levels below 8.94 and negative at higher pH. This knowledge is essential for comprehending how the protein behaves in diverse biological settings. The GRAVY, which in our case is obtained as 0.024 and 0.026 for standard and mutant CFTR, respectively, shows the hydropathicity of protein structure. It implies that the protein is not strongly hydrophobic but does have a small affinity for hydrophobic interactions. All this information is crucial in predicting the functionality of protein in biological functions in Table S3 (Supplementary material tables).

Predicting the impact of single nucleotide substitution on PTM sites

MUsite Deep server was used to predict the posttranslational modification in the protein. At the 614 positions, there is no post-translation modification available.

Predicting the effect of deleterious single nucleotide substitutions on protein properties

The amino acid aspartic acid shifts to glycine at position 614, according to the HOPE server. The schematic depiction of the original (left) and mutant (right) amino acid is displayed in (Supplementary Fig. 9). The side chain, as seen in black, is different, but the backbone is the same. In comparison to the wild residue, the mutant residue is more hydrophobic and smaller. The remnant of the mutant. The NBD1 domain contains the altered residue, which is crucial for the binding of other molecules. The mutation may have an impact on the protein's ability to bind to other molecules, which could have an impact on the protein's functionality.

Analysis of structure validity of CFTR

The modeled CFTR proteins were subjected to structure validation using Ramachandran plots ERRAT and VERIFY 3D. The comparative outcomes of these validations are presented in Table S4 (Supplementary material tables). The majority of the protein's backbone torsion angles appear to be within optimum ranges given that more than 80% of the amino acid residues in both protein structures are located in this region. This is a potent sign of a protein structure of excellent quality (Supplementary Fig. 10) and in Table S4 (Supplementary material tables). Based on non-bonded atomic interactions, the ERRAT score, which is presented as a percentage, indicates the overall quality of the protein structure. Both CFTR mutant and wild-type protein structures have good levels of overall quality, according to the ERRAT score, which shows that most non-bonded atomic interactions fall within the predicted range. The majority of the CFTR mutant proteins can be seen below the substantial error levels in the residue-by-residue predicted graphs shown in Table S5 (Supplementary material tables) and (Supplementary Fig. 11). By examining whether the 3D model and the protein's own amino acid, sequence agree, Verify3D is a technique for evaluating the quality of protein structures. It shows how closely the model fits the experimental data. The relevant amino acid sequences could not be adequately matched by either of these protein structures, since less than 80% of the amino acids scored >  = 0.1 in the 3D/1D profile (Supplementary Figs. 12, 13).

Secondary structure validation

Using GOR IV, the secondary structure of both proteins was predicted. It was discovered that both proteins included 37.64 percent random coils, 18.18 percent extended strands, and 44.19 percent alpha helix. The protein sequence's projected secondary structure was unaffected by a single amino acid change. Figure 5 shows a thorough graphic representation of the secondary structure (Supplementary Fig. 14).

Modelling of variants of CFTR proteinTM-align

The average distance between matching atoms in two proteins’ aligned structures is measured by RMSD. The RMSD value in this alignment is 2.23, indicating that the average distance between atoms in the structures of the wild-type and mutant proteins is 2.23 angstroms. A lower RMSD value indicates that the proteins are more structurally similar or more comparable in their three-dimensional structure. The TM-score, which ranges from zero to one, is a numerical indicator of how structurally similar two proteins are, with higher values denoting greater similarity. The TM-score in this alignment is 0.88043. With 1.0 being a perfect match, the high TM score indicates that the wild-type and mutant proteins share plenty of structural similarities (Supplementary Fig. 14).

PyMOL

PyMOL tool is used for the visualization of the 3D structure of the wild and mutated type of CFTR protein. Below are the figures showing that in wild-type CFTR protein at 614 position aspartic acid amino acid shown in red is changed into glycine amino acid shown in yellow color in Fig. 1.

Fig. 1

a–b: CFTR protein structure: a Wild type CFTR represents green area. b Mutated CFTR (D614G) is represented in the yellow area

Molecular docking of multiple ligands with mutated CFTR variant (D614G)

To investigate the influence of a particular mutation on the structure and function of the CFTR protein, molecular docking was performed with three different natural compounds. This was done by calculating the binding energy of the ligands with the mutated structure of the CFTR protein. Below are the different ligands and their interaction with the mutated D614G variant. Auto-DOCK scores for these ligands against mutant human CFTR (D614G) are − 6.20 kcal/mol for Silibinins, − 6.5 kcal/mol for Curcumin, − 5.85 kcal/mol for Demethoxycurcumin, − 5.50 kcal/mol for the reference drug Trikafta. Among all the natural compounds we used silibinins compound shows the maximum binding affinity with the mutated D614G variant of CFTR. The docking results showed that the docking score of silibinin, curcumin, and demethoxycucumin were the highest among the natural ligands used in this study. The interaction of the silibinin in the active site of the mutated model protein showed that the silibinin was well established and had a significant and maximum interaction with the key amino acids of the protein. Investigation of the interaction of the silibinin with the CFTR showed that the carbonyl and hydroxyl group of silibinin could form hydrogen bond with the amino acid’s residues; Glu193, Arg 1097. The interaction of the Demethoxycurcumin and curcumin in the active site of the mutated model protein showed that the Demethoxycurcumin and curcumin were also having good interaction with the key amino acids of the protein. Investigation of the interaction of these two ligands Demethoxycurcumin and curcumin with the CFTR showed that the carbonyl and hydroxyl group of these ligands could form two and one hydrogen bond with the amino acid’s residues: Lys 1041, Arg 1078, and Glu 193 respectively, see Figs. 2, 3, 4, 5a, c and Table S6 (Supplementary material tables).

Fig. 2

a–c: Docking complex of mutated CFTR (D614G) structure with demothoxycurcumin: a Docking complex of the protein–ligand surface of human mutated CFTR (red surface) with (dark purple color). b Indicate interacting residue of mutated CFTR structure with ligand. Red and grey balls show the active binding site of proteins. c The 2D diagram represents protein–ligand interacting residues indicating hydrogen bonding

Fig. 3

a–c: Docking complex of mutated CFTR (D614G) structure with silibinin: a Docking complex of the protein–ligand surface of human mutated CFTR (red surface) with silibinins (dark green color). b Indicate interacting residue of mutated CFTR structure with ligand. Red and grey balls show the active binding site of proteins. c The 2D diagram represents protein–ligand interacting residues indicating hydrogen bonding

Fig. 4

a–c: Docking complex of mutated CFTR (D614G) structure with curcumin: a Docking complex of the protein–ligand surface of human mutated CFTR (red surface) with curcumin (purple bolls). b Indicate interacting residue of mutated CFTR structure with ligand. Red and grey balls show active binding sites of proteins. c The 2D diagram represents protein–ligand interacting residues that indicate hydrogen bonding

Fig. 5

a–c: Docking complex of mutated CFTR (D614G) structure with Trikafta: a Docking complex of protein–ligand surface of human mutated CFTR (red surface) with Trikafta (cyan, green color). b Indicate interacting residue of mutated CFTR structure with the reference drug Trikafta. Red and grey balls show active binding sites of proteins. c The 2D diagram represents protein drug interacting residues indicating hydrogen bonding

The three of them fail to meet the criteria of Lipinski's rule of five with regard to the specified parameters. For each ligand, the permissible ranges for molecular weight < 500, the number of hydrogen bond acceptors and donor are less than 5, and 10, respectively. Molecules with a TPSA of 140 Å2 or greater would exhibit low fractional absorption (< 10%), while those with a TPSA of 60 Å2 or less would demonstrate high absorption (> 90%). The topological Polar Surface Area (TPSA) and %age of the natural ligands showed the preference as demethoxycurcumin > curcumin > silibinin. A medication candidate with a lipophilicity value between 0 and 5 would be suitable. Demethoxycurcumin (3.0), silibinin (1.59), and curcumin (1.23), according to the lipophilicity (LogPo/w) study, are significantly more lipophilic than the other natural compounds. The projected results indicate that oral and intestinal absorptions are achievable, although more clinical research is needed. Understanding that a medicine is a substrate for CYP and P-gp is also important for its proper bodily metabolism. Together, the two tiny chemicals strengthen the body’s defences against organ and tissue damage. Almost 50 to 90 percent of medicinal compounds are either one of the five CYP substrates. CYP3A4 is the most significant of the five isoforms. A substance will have a hazardous effect on the body if it inhibits any of these enzymes. Whereas demethoxycurcumin and silibinin inhibit CYP3A4, curcumin does not inhibit it in Tables S7,S8 (Supplementary material tables).

Analysis of molecular dynamic simulation of mutated CFTR protein with different natural compounds

We were verified the ligand binding modalities and the stability of the protein–ligand complexes using MD simulations and the Desmond program. The 200 ns simulations were performed using the top dock scorer, which included the CFTR (D614G) complexes together with three natural ligands including silibinin, curcumin and demethoxycurcumin. The RMSD figure showed the root mean standard deviation of the protein–ligand complex’s C-α backbone during the simulation.

Using MD for 200 ns, the C-α backbone of the CFTR protein in the Silibinin complex was simulated. The ligand was depicted in pink, and the C-α backbone's RMSD within the light blue range. The protein–ligand complex was found to be stable throughout the simulation, with observations made up to 200 ns. The simulation demonstrated the persistence of the CFTR protein, as seen by its C-α backbone of 3.5 Å, which observed minor fluctuation between 2.5 nm and 3.6 nm after 60 ns and then remain constant. The simulation results' RMSF showed that the residues of ARG-550 and GLU-400 varied above 6.0 Å. During the simulation, there was no ligand association seen with these protein residues. In the Silibinin-CFTR complex, the crucial residue of THR-1064 was shown to be directly involved in forming a hydrogen bond with the OH-benzne ring. The protein's crystal structure demonstrates how the Silibinin-CFTR complex interacts with residues including ASP-572, GLN-573, SER-1375, and LYS-1060 to generate stable hydrogen bonds through water-mediated interactions. Furthermore, Silibinin engages in hydrophobic bonding with water-mediated contacts with TRP-1063 and PHE-494. During the simulation, additional interactions were seen with different amino acids. Nonetheless, throughout the simulation, the contact with the receptor's GLU-267, THR-460, and ALA-462 residues remained constant and strong as shown in Fig. 6a–c.

Fig. 6

a–c. MD simulation analysis of the Silibinin with CFTR (D614G) complex. a RMSD protein ligand graph. b CFTR protein RMSFs. c The protein–ligand contact histogram

The CFTR-Curcumin complex was shown to be stable throughout the simulation, with observations up to 200 ns, according to the RMSD plot. With a C-α backbone of 3.6 Å, the CFTR protein was shown to have remained stable over the simulation, which observed minor fluctuation between 1.2 nm and 3.5 nm after 60 ns and then remain constant. The simulation findings' RMSF showed that GLU-400 and GLY-600 residues fluctuated above 5.6 Å. Throughout the simulation, the CFTR-curcumin complex interacts with residues like ARG-1437 through water-mediated interactions, forming stable hydrogen bonds. Nonetheless, throughout the simulation, the contact with the receptor's residues remained constant and strong as shown in Fig. 7a–c.

Fig. 7

a–c. MD simulation analysis of the Curcumin with CFTR (D614G) complex. a RMSD protein ligand graph. b CFTR protein RMSFs. c The protein–ligand contact histogram

The RMSD figure indicated that the CFTR-Desmethoxycurcumin combination was not found to be stable during the simulation, with observations up to 200 ns. With a C-α backbone of 3.6 Å, the CFTR protein was shown to have remained stable over the simulation, which observed minor fluctuation between 2.4 nm and 3.6 nm after 125 ns and then remain constant. The ligands of Desmethoxycurcumin were observed major fluctuations in the complex structure and did not shown stability within protein CFTR mutated structure. The simulation results showed that the residues of GLU-400, GLY-600, and GLY-801 fluctuated over 5.6 Å, as shown by the RMSF. During the simulation, it was observed that these protein residues were not connected to a ligand. When the CFTR-Curcumin complex interacts with residues like GLU-267, THR-1064, and HIS-1348 through water-mediated interactions, stable hydrogen bonds are formed. Moreover, during the simulation, Curcumin interacts with TRP-496 and PHE-494 through hydrophobic bonding and water-mediated interactions; (see Fig. 9). But throughout the simulation, the contact with the receptor's ILE-177, SER-263, GLU-264, and ASN-268 residues remained constant and strong. The best MD simulation results indicated Silibinin and Curcumin with CFTR protein indicate both complex stable throughout the simulation and no significant fluctuations in (supplementary Fig. 15a–c).

Analysis of radius of gyration (rGyr) of mutated CFTR protein with different natural compounds

The whole molecular dynamics (MD) simulation trajectory set was used to compute the rGyr parameter, which was used to assess the stability of the Silibinins—CFTR (D614G) complex. When the protein maintains its stability throughout the simulation, it has a stable rGyr value of 5.4 Å. During the 200 ns simulation, the structural compactness, or folding changes of silibinins with protein complexes were demonstrated using the RGyr parameter. Interestingly, no discernible variations were found for the complicated see Fig. 9a. The stability of the Demethoxycurcumin-CFTR (D614G) complex was evaluated by using the rGyr parameter, which was obtained from the complete molecular dynamics (MD) simulation trajectory set across the 200 ns simulation. The protein has a steady rGyr value of 3.72 Å when it was not stable throughout the simulation. Figure 9(b) for the complex see showed no discernible differences. Based on a simulation lasting 200 ns, the stability of the Curcumin—CFTR (D614G) complex was assessed using the rGyr parameter. A consistent rGyr value of 4.8 Å suggests that the protein maintained its stability following some modest fluctuations during the experiment. Interestingly, there were no discernible variations for the complex. These findings imply that following binding, protein–ligand complexes stay stable, as seen in Fig. 8c.

Fig. 8

a–c. Gyration radii of the complex. a. Silibinin ligand with CFTR protein gyration radii complex graph. b Curcumin ligand with CFTR protein gyration radii complex graph. c Demethoxycurcumin with CFTR protein gyration radii complex graph

Calculation of MM-GBSA Binding free energy calculation

After molecular simulation, the ΔGbind values were determined using the MM-GBSA method and shows the degree of binding strength between the protein and ligand complexes. The complexes generated by CFTR with Silibinins, Demethoxycurcumin, and Curcumin had average ΔGbind values of − 124.55, − 62.637497, and -39.111, respectively. As can be seen in Table S9 (Supplementary material tables), a significantly negative number indicates a larger binding affinity between the two.

Principle component analysis of mutated CFTR protein with compounds

Silibinins, and Curcumin were the most stable ligands among the selected compounds during the MDS process, according to PCA data. The percentage of variance for each component is displayed in three separate parts of PCA eigenvalue plots. For instance, different regions of the silibinins-CFTR protein in variants in PC1, PC2, and PC3 contribute, respectively, 29.6%, 11.74%, and 7.2% of the total variance. A graphic representation of these data may be found in Fig. 9. The PCA results for Curcumin; 32.9%, 13.39%, and 6.09%. The PCA result indicating silibinins and curcumin with the CFTR protein had the highest score in Fig. 9a–c.

Fig. 9

a–c. Principle Component Analysis. a PCA of Silibinin with CFTR (D614G) complex. b PCA of Curcumin with CFTR (D614G) complex c PCA of Demethoxycurcumin with CFTR (D614G) complex