INTRODUCTION:

Changes in the single base in human genome are known as single nucleotide polymorphism (SNP) and it is most frequent type of genetic variation^{1, 2}. Upto December 10, 2013 a total of 284103258 SNPs in humans have been identified and deposited in the NCBI dbSNP. When SNPs occur in a coding region and cause the modification in the amino acid in the protein is called nonsynonymous single nucleotide polymorphisms (nsSNPs)³. These nsSNPs occurring in a protein coding regions affects the protein function and causes various genetic diseases⁴.

Andersen–Tawil syndrome (ATS) is a rare inherited ion channel dysfunction often characterized by periodic paralysis, prolongation of QT interval with ventricular arrhythmias and characteristic physical features include clinodactyly, low set ears, micrognathia, a broad forehead cleft plate and hypertelorism^{5, 6}. Reports have suggested that ATS is caused mainly due to mutations in KCNJ2 gene which encodes an inward rectifying K+ channel protein, Kir2.1.⁶. The encoded protein has a greater tendency to allow potassium to flow into a cell rather than out of a cell, and participates in establishing action potential waveform and excitability of neuronal and muscle tissues. The concentration of extracellular potassium regulates their voltage dependence thus when external potassium is raised, it leads to more positive voltage ranges for channel opening. The inward rectification is principally owing to the blockage of outward current by internal magnesium. The Kir2.1 channels are predominantly expressed in the brain, heart and skeletal muscles⁷. These channels are important regulators of cellular excitability and resting membrane potential of both cardiac and skeletal muscles⁸.Structurally, alpha subunit is the pore-forming subunit of KIR channels. It contains one pore domain between two membrane spanning regions. A tetramer is formed by four alpha subunits, with the pore domain of each subunit contributing to the structure of the central pore. In this study we have tried to find out the effect of flecainide (a sodium channel blocker), on Kir2.1. Many clinical trials of this drug have been reported to be effective in controlling cardiac events^{9, 10}.Mutant sodium channels that are in closed inactivated state can offer a way to devise a new strategy in developing mutation specific treatment^{10, 11}. Therefore in this study we are looking at the binding efficiency of flecainide with mutant Kir2.1 channel to suggest mutation specific treatment for ATS. Hence, the present study has focused to understand the deleterious effect of genetic variations involved in the KCNJ2 with the help of bioinformatics. We have incorporated different computational algorithms to eliminate false positive results and to enhance the accuracy of the predictions.

MATERIAL AND METHODS:

Sequence data sets and polymorphism identification

The SNPs and their associated protein sequence for KCNJ2 gene were obtained from dbSNP (http://www. ncbi.nlm.nih.gov/SNP/) for our computational analysis.

Analysis of functional consequences of coding nsSNPs by sequence homology based method (SIFT) and Damaged nsSNP found by the PolyPhen-2 server:

SIFT (Sorting Intolerant from Tolerant) is a sequence homology based tool, which presumes that important amino acids will be conserved in the protein family¹². Hence, changes at well-conserved positions tend to be predicted as deleterious. The underlying principle of this program is that SIFT takes a query sequence and uses multiple alignment information to predict tolerated and deleterious substitutions for every position of the query sequence. SIFT is a multistep procedure that, given a protein sequence, (a) searches for similar sequences, (b) chooses closely related sequences that may share similar function, (c) obtains the multiple alignment of these chosen sequences, and (d) calculates normalized probabilities for all possible substitutions at each position from the alignment. Substitutions at each position with normalized probabilities less than a chosen cutoff are predicted to be deleterious, and those greater than or equal to the cutoff are predicted to be tolerated¹³. The cutoff value in SIFT program is a tolerance index of >0.05. Higher the tolerance index, less functional impact a particular amino acid substitution is likely to have. PolyPhen input is the amino acid sequence of a protein or the SWALL database ID or accession number together with sequence position and two amino acid variants characterizing the polymorphism. Sequence-based characterization of the substitution site, profile analysis of homologous sequences and mapping of the substitution site for a known protein's 3-dimensional structures are the parameters taken into account by PolyPhen server to calculate the score. It calculates position-specific independent counts (PSIC) scores for each of the two variants, and then computes the PSIC score difference between them. Higher the PSIC score difference, higher is the functional impact a particular amino acid substitution is likely to have¹⁴.

Analysis by SNPs and GO, PANTHER and PhD-SNP:

SNPs and GO collects in unique framework information derived from protein sequence, 3D structure, protein sequence profile, and protein function. The enormous number of human SNPs available in the data bases poses the question of relating protein variations to diseases. We propose a new server that uses different pieces of information, including that is derived from the Gene Ontology annotation to predict if a given variation can be classified disease-related or neutral. SNPs and GO is an SVM-based classifier consisting of a single SVM that takes in an input protein sequence, profile and functional information. The server also implements PHD-SNP method that takes in input different subsets of SNPs and GO's input features¹⁵. PANTHER algorithm is based on a library of Hidden Markov Models (HMMs) obtained from the multiple sequence alignments of different protein families. Thus it performs the evolutionary analysis of coding nsSNPs. PANTHER estimates the likelihood of a particular nsSNP causing a functional impact on the protein. The software computes the predicted free energy change value or sign (∆∆G) which is calculated from the unfolding Gibbs free energy value of the mutated protein minus unfolding Gibbs free energy value of the native protein (kcal/mol). A positive ∆∆G value indicates that the mutated protein possesses high stability and vice versa^16-18. PHD-SNP is based on SVM-based classifier. In the new version the authors have developed a predictor based on a single SVM trained and tested on protein sequence and profile information. PHD-SNP is optimized to predict if a given single point protein mutation can be classified as disease-related or as neutral polymorphism. The output consists of a table listing the number of the mutated position in the protein sequence, the wild-type residue, the new residue and if the related mutation is predicted as disease-related (Disease) or as neutral polymorphism (Neutral)^{15, 16,
19}.

Analysis by I-Mutant 2.0, iStable and Mupro:

Mutant 2.0 is a support vector machine (SVM)-based tool for the automatic prediction of protein stability changes upon single point mutations. I-Mutant 2.0 predictions are performed starting either from the protein structure or, more importantly, from the protein sequence. This program was trained and tested on a data set derived from ProTherm, which is presently the most comprehensive available database of thermodynamic experimental data on free-energy changes of protein stability upon mutation under different conditions.

Table 1: List of variants that were predicted to be functionally significant by SIFT and Polyphen 2.0 algorithm.

S.No.	SNPID	Nucleotide change	Amino acid Change	Score	Prediction	Score
1	rs375727662	C/G	G33R	0	Benign	0.154
2	rs375646186	A/C	K120T	0.73	Benign	0
3	rs375605948	G/T	A70S	0.36	Benign	0
4	rs375330016	A/G	E349K	0.01	Probably damaging	0.989
5	rs373799322	C/T	P415L	0.16	Benign	0.051
6	rs370571312	A/G	G206D	0.05	Probably damaging	0.982
7	rs368406938	G/T	C356F	0	Probably damaging	1
8	rs367560052	A/G	N318S	1	Benign	0
9	rs202149686	C/T	I406T	0.42	Benign	0
10	rs202067116	C/T	R325C	0	Probably damaging	0.999
11	rs201162707	C/T	V295A	0	Probably damaging	1
12	rs199473659	C/T	P351S	0.04	Probably damaging	0.965
13	rs199473658	A/G	M307I	0.01	Probably damaging	0.993
14	rs199473657	G/T	V227F	0	Probably damaging	1
15	rs199473656	C/T	L217P	0	Probably damaging	1
16	rs199473655	C/T	T192I	0	Probably damaging	1
17	rs199473654	A/G	G146S	0	Probably damaging	1
18	rs199473653	A/G	R82Q	0	Probably damaging	1
19	rs199473652	A/G	T74A	0	Possibly damaging	0.951
20	rs199473651	G/T	Y68D	0	Probably damaging	1
21	rs199473650	G/T	C54F	0	Probably damaging	1
22	rs199473390	G/T	R422L	0.01	Benign	0.425
23	rs199473389	C/T	R312C	0	Probably damaging	1
24	rs199473388	C/T	T309I	0.03	Probably damaging	0.997
25	rs199473387	A/C/G	T305P	0	Probably damaging	1
26	rs199473386	A/G	E303K	0	Probably damaging	0.997
27	rs199473385	C/G	R260P	0	Probably damaging	1
28	rs199473384	A/G	R218Q	0	Probably damaging	1
29	rs199473383	A/G	G215D	0	Probably damaging	1
30	rs199473382	A/G	T192A	0	Probably damaging	0.993
31	rs199473381	G/T	R189I	0	Probably damaging	1
32	rs199473380	G/T	C154F	0	Probably damaging	1
33	rs199473379	A/C/G	G146D	0	Probably damaging	1
34	rs199473378	A/G	G144S	0	Probably damaging	1
35	rs199473377	A/C/G	G144D	0	Probably damaging	1
36	rs199473376	C/T	S136F	0.1	Probably damaging	1
37	rs199473375	G/T	V123G	0	Probably damaging	0.999
38	rs199473374	C/T	C101R	0.3	Possibly damaging	0.614
39	rs199473373	C/T	R82W	0	Probably damaging	1
40	rs199473372	G/T	D78Y	0	Probably damaging	1
41	rs199473371	A/G	D78G	0	Probably damaging	1
42	rs199473370	A/G	T75A	0	Probably damaging	0.999
43	rs199473369	A/G	D71N	0	Probably damaging	1
44	rs199473368	A/G	R67Q	0	Probably damaging	1
45	rs151168735	C/T	T130M	0.39	Benign	0.386
46	rs150326473	A/G	I171V	0.34	Benign	0.448
47	rs147750704	A/G	V93I	0.47	Benign	0.004
48	rs144022753	C/T	T400M	0.13	Benign	0.025
49	rs141069645	A/G	N410S	0.76	Benign	0
50	rs141035459	A/G	G206S	0.01	Possibly damaging	0.852
51	rs140147979	A/G	V167M	0.07	Possibly damaging	0.797
52	rs140090605	A/G	R8H	0.15	Benign	0.192
53	rs140053197	A/G	D389N	0.21	Possibly damaging	0.6
54	rs111909178	A/T	N410Y	0.1	Benign	0
55	rs104894585	C/G/T	T75R	0	Probably damaging	1
56	rs104894584	A/G	D172N	0.18	Probably damaging	1
57	rs104894583	A/C	N216H	0.01	Probably damaging	1
58	rs104894582	A/G	V302M	0	Probably damaging	1`
59	rs104894581	C/T	P186L	0	Probably damaging	1
60	rs104894580	C/T	R67W	0	Probably damaging	1
61	rs104894579	A/C/G/T	G300D	0	Probably damaging	1
62	rs104894578	C/T	R218W	0	Probably damaging	1
63	rs104894575	A/T	D71V	0	Probably damaging	1
64	rs79650811	G/T	F98V	0.67	Benign	0.001

Note: Letters in bold indicates deleterious amino acid variants prediction by SIFT and Polyphen 2.0 program.

The output file shows the predicted free-energy change value or sign (∆∆G), which is calculated from the unfolding Gibbs free energy value of the mutated protein minus the unfolding Gibbs free energy value of the native type (kcal/mol). Positive ∆∆G values mean that the mutated protein posses high stability and vice versa²⁰. iStable²¹ was the next predictor which we used to predict protein stability changes which consists of grid computing architecture which uses protein sequence as input and integrates results from different element predictors. MUpro²¹is another tool used to predict stability changes due to single base mutations upon the input of sequence information and rarely protein structure. It predicts if a stability change will lead to destabilization or not without giving the actual DDG value.

Analysis by Align-GVGD, mCSM and nsSNP Analyzer:

Align-GVGD²² is an extension of Grantham Difference, which we used to predict the transactivation activity of each missense mutation. It is predicted by combining GV and GD scores which considers composition, polarity and volume of amino acids. These three parameters are taken as the three axes and when all amino acids are plotted, it forms a cloud of points at a given position in multiple sequence alignment . These points are enclosed in a box with maximum and minimum values of the parameters as its coordinates. GV is the Euclidian length of the main diagonal of the box which indicates the biochemical variation at a particular position and GD is calculated by plotting a mutation the graph and measuring the Euclidian distance between that mutation and a closest point on the box. GD=0, if the substitution lies in the box otherwise it is greater than 0. Another tool is Mcsm²³which we used to predict stability changes and it depends on graph based signatures which takes into account not only the protein stability but also the protein-protein and protein-nucleic acid interactions. Finally we used nsSNP Analyzer²⁴ which predicts nsSNP’s phenotypic effect. nsSNPAnalyzer uses a machine learning method called Random Forest to classify the nsSNPs. Random Forest is a type of classifier consisting of an entity of tree-structured classifiers²⁴. It takes a protein sequence and its nsSNPs as inputs and predicts if the nsSNP is disease associated or functionally neutral.

Modeling and Protein docking:

SWISS-MODEL is a fully automated protein structure homology-modeling server, accessible via the ExPASy web server. The purpose of this server is to make Protein Modeling easy. The amino acid sequence of a protein is submitted to construct a 3D model. Template selection, alignment and model building are done completely automated by the server²⁵. Interaction free energies are very important for analyzing binding tendency in proteins. We used the Docking Server²⁶ program which calculates the binding free energy between protein complexes. Docking server calculates the parameters needed at different steps for the docking, i.e. accurate ligand geometry optimization, energy minimization, charge calculation, docking calculation and protein-ligand complex representation. In order to calculate the binding affinity between KCNJ2 protein and Flecainide, we used SWISSMODEL to model the KCNJ2 protein structure and SWISSPDB viewer for performing mutation. The two dimensional structure of Flecainide drug is obtained from the PubChem in the Smiles format. Then the structure was submitted to corina to get the three dimensional structure of drug molecule. Finally the 3D structures of protein and flecainide were submitted to Docking Server to get them docked and then we got their free binding energy.

RESULTS AND DISCUSSION:

Screening of Deleterious Mutations by SIFT and PolyPhen-2.0 Server:

In disease associated gene studies, we need to distinguish between large number of neutral SNPs and SNPs of functional importance. This task could be easily achieved with the development computational techniques particularly with different genomic algorithms. In the present investigation, the initial screening was done using SIFT which predicted the tolerance indices of 64 SNP variants. As the tolerance level decreases, the functional importance of the amino acid substitution increases and vice versa. The results are shown in Table 1.

Among the 64 SNP variants, 47 variants were found to be deleterious having tolerance indices ≤ 0.05. Subsequently, the structural alterations of these variants were determined using PolyPhen-2.0 program. Protein sequence with mutational position and residues were submitted to the server and it predicted the level of damage for the structural alteration. Probably damaging stands the nearest chances to find mutations make the protein non-functional, possibly damaging are less likely prone in the category and benign come near the safe zone. In Table 1, 64 variants are listed, out of which 44 variants are predicted to be damaging by PolyPhen-2.0 server.

Prediction of Disease Probability of the Mutations by SNPs and GO, PANTHER and PhD-SNP:

Furthermore, SNPs and GO, PANTHER and PhD-SNP program were used to predict the disease and probability score of the nsSNPs and the results are shown in (Table 2).

Table 2: Prediction of functionally significant nsSNPs by SNPs and GO, PhD-SNP and PANTHER algorithm.

Mutation	SNPs and GO			PhD-SNP			PANTHER
	Prediction	Score	RI	Prediction	Score	RI	Prediction	Score	RI
E349K	Disease	4	0.719	Disease	1	0.539	Disease	7	0.873
G206D	Disease	6	0.818	Disease	9	0.93	Disease	5	0.765
C356F	Disease	6	0.803	Disease	8	0.881	Disease	8	0.877
R325C	Disease	6	0.815	Disease	8	0.883	Disease	10	0.99
V295A	Disease	3	0.668	Disease	5	0.727	Disease	6	0.806
P351S	Disease	3	0.639	Disease	5	0.725	Disease	3	0.639
V227F	Disease	1	0.574	Disease	5	0.735	Disease	4	0.722
L217P	Disease	7	0.828	Disease	8	0.923	Disease	9	0.972
T192I	Disease	5	0.739	Disease	7	0.829	Disease	10	0.989
G146S	Disease	7	0.848	Disease	8	0.904	Disease	10	0.989
R82Q	Disease	5	0.725	Disease	8	0.891	Disease	3	0.669
Y68D	Disease	8	0.894	Disease	9	0.944	Disease	9	0.953
C54F	Disease	7	0.856	Disease	8	0.919	Disease	10	0.99
R312C	Disease	5	0.768	Disease	8	0.877	Disease	10	0.983
T305P	Disease	5	0.755	Disease	7	0.867	Disease	9	0.944
E303K	Disease	8	0.879	Disease	8	0.919	Disease	10	0.977
R260P	Disease	6	0.813	Disease	6	0.808	Disease	2	0.598
R218Q	Disease	8	0.879	Disease	9	0.94	Disease	10	0.983
G215D	Disease	6	0.821	Disease	8	0.901	Disease	10	0.983
T192A	Disease	3	0.655	Disease	2	0.617	Disease	9	0.949
R189I	Disease	2	0.576	Disease	5	0.736	Disease	10	0.994
C154F	Disease	9	0.931	Disease	10	0.982	Disease	10	0.996
G146D	Disease	7	0.856	Disease	9	0.982	Disease	10	0.993
G144D	Disease	5	0.773	Disease	9	0.935	Disease	1	0.544
S136F	Disease	4	0.692	Disease	8	0.882	Disease	6	0.78
V123G	Disease	2	0.618	Disease	7	0.831	Disease	3	0.649
R82W	Disease	7	0.861	Disease	9	0.961	Disease	9	0.936
D78Y	Disease	8	0.922	Disease	9	0.97	Disease	10	0.98
D78G	Disease	9	0.927	Disease	9	0.953	Disease	8	0.923
T75A	Disease	8	0.903	Disease	8	0.886	Disease	9	0.955
D71N	Disease	7	0.869	Disease	9	0.869	Disease	10	0.982
R67Q	Disease	6	0.79	Disease	6	0.814	Disease	9	0.93
T75R	Disease	8	0.919	Disease	9	0.934	Disease	10	0.989
N216H	Disease	3	0.638	Disease	6	0.778	Disease	6	0.825
R67W	Disease	6	0.79	Disease	8	0.904	Disease	10	0.988
G300D	Disease	6	0.871	Disease	8	0.898	Disease	10	0.982
R218W	Disease	7	0.871	Disease	9	0.962	Disease	10	0.996
D71V	Disease	9	0.929	Disease	9	0.969	Disease	10	0.995

Fig.1. KCNJ2 protein-Flecainide docked complex structures

Table 3: Prioritization of nsSNPs by I Mutant 2.0, MUpro and iStable algorithm.

Complex Name	Prediction	ΔΔG (kcal/mol)	Prediction	Conf. Score	Prediction	Conf. Score
E349K	Decrease	-1.69	Decrease	-0.85	Decrease	0.85
G146S	Decrease	-1.47	Decrease	-0.22	Decrease	0.78
G206D	Decrease	-1.15	Decrease	-0.39	Decrease	0.81
C356F	Decrease	-0.1	Decrease	-0.06	Decrease	0.7
V295A	Decrease	-0.37	Decrease	-0.1	Decrease	0.86
P351S	Decrease	-2.2	Decrease	-1	Decrease	0.81
L217P	Decrease	-0.34	Decrease	-1	Decrease	0.9
T192I	Decrease	-1.54	Decrease	-0.8	Decrease	0.62
Y68D	Decrease	-0.6	Decrease	-0.09	Decrease	0.87
E303K	Decrease	-0.05	Decrease	-1	Decrease	0.74
R312C	Decrease	-0.12	Decrease	-0.14	Decrease	0.6
R260P	Decrease	-2.66	Decrease	-0.8	Decrease	0.72
R218Q	Decrease	-1.26	Decrease	-1	Decrease	0.74
T192A	Decrease	-2.15	Decrease	-0.97	Decrease	0.69
R189I	Decrease	-0.59	Decrease	-0.22	Decrease	0.73
C154F	Decrease	-0.17	Decrease	-1	Decrease	0.79
G144D	Decrease	-1.29	Decrease	-0.81	Decrease	0.73
V123G	Decrease	-3.62	Decrease	-1	Decrease	0.83
D78Y	Decrease	-0.97	Decrease	-0.41	Decrease	0.74
D78G	Decrease	-0.28	Decrease	-0.82	Decrease	0.79
T75A	Decrease	-1.06	Decrease	-1	Decrease	0.82
D71N	Decrease	-1.13	Decrease	-1	Decrease	0.83
R67Q	Decrease	-0.9	Decrease	-0.43	Decrease	0.73
T75R	Decrease	-0.05	Decrease	-0.23	Decrease	0.79
N216H	Decrease	-1.73	Decrease	-1	Decrease	0.67
G300D	Decrease	-1.04	Decrease	-0.71	Decrease	0.74
R218W	Decrease	-0.61	Decrease	-1	Decrease	0.7
D71V	Decrease	-1.3	Decrease	-0.09	Decrease	0.84

Table 4: Prioritization of nsSNPs by Align GVGD, mCSM and nsSNP algorithm.

Complex Name	GV	GD	Prediction	Score	DDG (kcal/mol)	Stability	nsSNP analyzer
G206D	0	93.77	65	79.1	-0.903	Destabilizing	Disease
C356F	0	204.39	65	44.3	-1.293	Destabilizing	Disease
V295A	0	65.28	65	0.1	-2.296	Highly destabilizing	Disease
L217P	0	97.78	65	5.6	-1.105	Destabilizing	Disease
Y68D	0	159.94	65	25.6	-1.807	Destabilizing	Disease
R312C	0	179.53	65	11.7	-1.617	Destabilizing	Disease
R260P	0	102.71	65	19.2	-0.353	Destabilizing	Disease
R189I	0	97.59	65	0.3	-0.48	Destabilizing	Disease
C154F	0	204.39	65	9.2	-1.282	Destabilizing	Disease
G144D	0	93.77	65	0	-2.32	Highly destabilizing	Disease
V123G	0	109.55	65	5.1	-2.683	Highly destabilizing	Disease
D78G	0	93.77	65	43.4	-0.624	Destabilizing	Disease
T75R	0	70.97	65	50.2	-0.686	Destabilizing	Disease
N216H	0	68.35	65	3.3	-1.254	Destabilizing	Disease
G300D	0	93.77	65	0.9	-2.91	Highly destabilizing	Disease
R218W	0	101.29	65	5.9	-0.656	Destabilizing	Disease
D71V	0	152.22	65	0.9	-0.824	Destabilizing	Disease

Table 5: Docking analysis of flecainide with native and mutant types KCNJ2 protein

Complex Name	Free energy of binding (kcal/mol)	Number of H- bonds
Native KCNJ2-Flecainide	-4.33	12
G206D- Flecainide	-5.05	11
C356F- Flecainide	-4.66	10
V295A- Flecainide	-4.09	9
L217P- Flecainide	-4.92	10
Y68D- Flecainide	-4.71	7
R312C- Flecainide	-4.32	10
R260P- Flecainide	-3.99	5
R189I- Flecainide	-5.43	9
C154F- Flecainide	-4.23	9
G144D- Flecainide	-4.45	10
V123G- Flecainide	-4.4	9
D78G- Flecainide	-4.68	8
T75R- Flecainide	-4.72	9
N216H- Flecainide	-4.65	8
G300D- Flecainide	-4.57	9
R218W- Flecainide	-4.98	6
D71V- Flecainide	-3.57	9

Fig. 2. Details of intermolecular hydrogen bonding network in (A) KCNJ2 protein-flecainide and (B) mutant (D71V) KCNJ2 protein-flecainide complex structures.

The disease probability score was predicted to be higher for the E349K, G206D, C356F, R325C, V295A, P351S, V227F, L217P, T192I, G146S, R82Q, Y68D, C54F, R312C, T305P, E303K, R260P, R218Q, G215D, T192A, R189I, C154F, G146D, G144D, S136F, V123G, R82W, D78Y, D78G, T75A D71N, R67Q, T75R, N216H, R67W, G300D, R218W, D71V.All the 38 variants showed a reliability index values between 0.539 to 0.996.

Screening of nsSNPs using I-Mutant 2.0, iStable and MUPRO:

All the 64 SNPs were submitted individually to I-Mutant 2.0 program which predicted the DDG value of each variant. The negative DDG value means that the point mutation is less stable which makes the variants significant in our study. Among the 64 SNPs, 28 were found to be less stable and thus functionally significant. Further, we used two more programs namely iStable and MUPRO for the prediction of stability changes which confirms 28 variants from the result of I-Mutant 2.0 to be less stable. The results are listed in Table 3.

Prediction of functionally significant mutation using Align-GVGD, mCSM and nsSNP Analyzer

The 28 SNPs identified to be less stable were then screened further using Align-GVGD and based on GVGD values predicted class 65 which is most likely significant were found to be 17 SNP variants listed in Table 4. Then again stability was predicted using mCSM as ‘Destabilizing’ and ‘Highly Destabilizing’. Furthermore, the 17 variants were finally confirmed as functionally significant by nsSNP. Analyzer as it was predicted using this program that all the 17 SNP variants listed by Align-GVGD and mCSM are disease causing SNPs. The screening efficiency was confirmed with the presence of experimental proof for mutations R218W⁶, D71V⁶, R312C⁷ and N216H⁷, which tells that these mutations lead to loss of function in Kir 2.1 channel.

Protein Docking:

Protein docking was carried out for the drug and the 17 SNP variants to compare the binding efficiency between native and mutant complexes. The free energy of the native complex is-4.33 kcal/mol whereas for the mutant complexes the free binding energy ranges from -3.57 to -5.43 kcal/mol. Minimum binding efficiency is shown by D71V mutation which suggest that the 3D conformation of flecainide does not fit effectively into the 3D space of the binding residues of the mutant as compared to other mutants. The number of hydrogen bond interaction in the native complex is 12 and for the mutant complexes the number of hydrogen bond interaction ranges from 5 to 11. The free energy and number of H bond interactions of the 17 variants are shown in Table 5. The docked complex structure is shown in Fig.1. Further intermolecular interactions were analysed and the results are shown in Fig. 2.

CONCLUSION:

KCNJ2 gene belongs to a family which produces potassium channels. These channels transports positive charge ions of potassium out of cell and also play a key role in a cell’s ability to generate and transmit electrical signals. In particular, a mutation in the KCNJ2 gene causes ATS disease. Hence, in the present investigation we made an attempt to detect the deleterious mutations in KCNJ2 gene by computational analysis. A total of 64 missense mutations were retrieved for our analysis, SIFT predicted 71.87% of SNPs as functionally significant, PolyPhen-2.0 predicted 56.25% of SNPs as damaging. In addition, SNPs and Go, PANTHER and PhD-SNP predicted the disease probability of the 38 nsSNP as high. I-Mutant 2.0, MUpro, iStable predicted 43.75% of SNPs to affect the stability, Align GVGD, mCSM, nsSNP predicted 17 mutations to be highly deleterious. Moreover, docking study was employed with the aid of flecainide molecule to improve the prediction accuracy. Finally, the data obtained from different algorithms were combined to eliminate the false positive in the computational prediction methods. The results suggest that V295A, R312C, R260P, C154F and D71V have dramatic impact in the KCNJ2 protein structure and its function. We certainly believe that this approach of screening deleterious mutations helps in predicting early prognosis and the response required for a particular therapeutic approach.

ACKNOWLEDGMENTS:

The authors express deep sense of gratitude to the management of Vellore Institute of Technology for all the support, assistance and constant encouragements to carry out this work.

REFERENCES:

1. Cargill M, Altshuler D, Ireland J, Sklar P, Ardlie K, Patil N, Shaw N, Lane CR, Lim EP, Kalyanaraman N, Nemesh J, Ziaugra L, Friedland L, Rolfe A, Warrington J, Lipshutz R, Daley GQ, Lander ES. Characterization of single-nucleotide polymorphisms in coding regions of human genes. Nat Genet. 22(3); 1999: 231-238.

2. Nadeau JH. Single nucleotide polymorphisms: tackling complexity. Nature. 420(6915); 2002:517-518.

3. Ramensky V, Bork P, Sunyaev S. Human non-synonymous SNPs: server and survey. Nucleic Acids Res. 30(17); 2002:3894-900.

4. Rennell D, Bouvier SE, Hardy LW, Poteete AR. Systematic mutation of bacteriophage T4 lysozyme. J Mol Biol. 222(1); 1991:67-88.

5. Choi BO, Kim J, Suh BC, Yu JS, Sunwoo IN, Kim SJ, Kim GH, Chung KW. Mutations of KCNJ2 gene associated with Andersen-Tawil syndrome in Korean families. J Hum Genet. 52(3); 2007:280-3.

6. Tristani-Firouzi M, Jensen JL, Donaldson MR, Sansone V, Meola G, Hahn A, Bendahhou S, Kwiecinski H, Fidzianska A, Plaster N, Fu YH, Ptacek LJ, Tawil R. Functional and clinical characterization of KCNJ2 mutations associated with LQT7 (Andersen syndrome). J Clin Invest. 110(3); 2002:381-8.

7. Limberg MM, Zumhagen S, Netter MF, Coffey AJ, Grace A, Rogers J, Böckelmann D, Rinné S, Stallmeyer B, Decher N, Schulze-Bahr E. Non dominant-negative KCNJ2 gene mutations leading to Andersen-Tawil syndrome with an isolated cardiac phenotype. Basic Res Cardiol. 108(3); 2013:353.

8. Schulze-Bahr E. Short QT syndrome or Andersen syndrome: Yin and Yang of Kir2.1 channel dysfunction. Circ Res. 96(7); 2005:703-704.

9. Haruna Y, Kobori A, Makiyama T, Yoshida H, Akao M, Doi T, Tsuji K, Ono S, Nishio Y, Shimizu W, Inoue T, Murakami T, Tsuboi N, Yamanouchi H, Ushinohama H, Nakamura Y, Yoshinaga M, Horigome H, Aizawa Y, Kita T, Horie M. Genotype-phenotype correlations of KCNJ2 mutations in Japanese patients with Andersen-Tawil syndrome.Hum Mutat. 28(2); 2007:208.

10. Jurkat-Rott K, Lehmann-Horn F. Muscle channelopathies and critical points in functional and genetic studies. J Clin Invest. 115(8); 2005:2000-2009.

11. Mohammadi B, Jurkat-Rott K, Alekov A, Dengler R, Bufler J, Lehmann-Horn F. Preferred mexiletine block of human sodium channels with IVS4 mutations and its pH-dependence. Pharmacogenet Genomics. 15(4); 2005:235-244.

12. Kumar P, Henikoff S, Ng PC. Predicting the effects of coding non-synonymous variants on protein function using the SIFT algorithm. Nat Protoc. 4(7); 2009: 1073-1081.

13. Ng PC, Henikoff S. SIFT: Predicting amino acid changes that affect protein function. Nucleic Acids Res. 31(13); 2003:3812-3814.

14. Adzhubei IA, Schmidt S, Peshkin L, Ramensky VE, Gerasimova A, Bork P, Kondrashov AS, Sunyaev SR. A method and server for predicting damaging missense mutations. Nat Methods. 7(4); 2010:248-249.

15. Calabrese R, Capriotti E, Fariselli P, Martelli PL, Casadio R. Functional annotations improve the predictive score of human disease-related mutations in proteins. Hum Mutat. 30(8); 2009:1237-1244.

16. Capriotti E, Altman RB, Bromberg Y. Collective judgment predicts disease-associated single nucleotide variants. BMC Genomics. 14 Suppl 3:S2. 2013.

17. Thomas PD, Campbell MJ, Kejariwal A, Mi H, Karlak B, Daverman R, Diemer K, Muruganujan A, Narechania A. PANTHER: a library of protein families and subfamilies indexed by function. Genome Res. 13(9); 2003:2129-2141.

18. Thomas PD, Kejariwal A, Campbell MJ, Mi H, Diemer K, Guo N, Ladunga I, Ulitsky-Lazareva B, Muruganujan A, Rabkin S, Vandergriff JA, Doremieux O. PANTHER: a browsable database of gene products organized by biological function, using curated protein family and subfamily classification. Nucleic Acids Res. 31(1); 2003:334-341.

19. Capriotti E, Calabrese R, Casadio R. Predicting the insurgence of human genetic diseases associated to single point protein mutations with support vector machines and evolutionary information. Bioinformatics. 22(22); 2006:2729-34.

20. Capriotti E, Fariselli P, Casadio R. I-Mutant2.0: predicting stability changes upon mutation from the protein sequence or structure. Nucleic Acids Res. 2005:33(Web Server issue):W306-10.

21. Chen CW, Lin J, Chu YW. iStable: off-the-shelf predictor integration for predicting protein stability changes. BMC Bioinformatics. 2013:14 Suppl 2:S5.

22. Mathe E, Olivier M, Kato S, Ishioka C, Hainaut P, Tavtigian SV. Computational approaches for predicting the biological effect of p53 missense mutations: a comparison of three sequence analysis based methods. Nucleic Acids Res. 34(5); 2006:1317-1325.

23. Pires DE, Ascher DB, Blundell TL. mCSM: predicting the effects of mutations in proteins using graph-based signatures. Bioinformatics. 30(3); 2014:335-42.

24. Bao L, Zhou M, Cui Y. nsSNPAnalyzer: identifying disease-associated nonsynonymous single nucleotide polymorphisms. Nucleic Acids Res. 2005:33(Web Server issue):W480-2.

25. Schwede T, Kopp J, Guex N, Peitsch MC. SWISS-MODEL: An automated protein homology-modeling server. Nucleic Acids Res. 31(13); 2003:3381-3385.

26. Bikadi Z, Hazai E. Application of the PM6 semi-empirical method to modeling proteins enhances docking accuracy of AutoDock. J Cheminform. 1:15; 2009.

Received on 10.09.2015 Modified on 25.09.2015

Research J. Pharm. and Tech. 8(11): Nov., 2015; Page 1540-1547

DOI: 10.5958/0974-360X.2015.00275.9