A synonymous KCNH2 polymorphism and methadone trough level influence QTc prolongation in Kelantanese Malay recipients of methadone maintenance therapy (MMT) in Malaysia

Abstract

Potassium voltage-gated channel subfamily H member 2 (KCNH2) polymorphisms have been found to influence the heart-rate adjusted QT (QTc) intervals. We investigated the association between KCNH2 polymorphisms and QTc intervals among Malay opioid-dependent methadone maintenance treatment (MMT) recipients. A cross-sectional study was conducted involving 111 patients with stable methadone dosage for at least 6 months attending several methadone clinics in Kelantan, Malaysia between March 2011 and October 2012. Those with cardiac structural defects, recipients of other QTc-prolonging pharmacotherapeutic agents, had aggressive behavior or other active psychiatric illnesses, chronic medical and surgical ailments and who were unable to communicate in Malay and English were excluded. The Fridericia-corrected QTc intervals were recorded using a 12-lead electrocardiogram. DNA samples were extracted from peripheral blood leukocytes and genotyped using nested allele-specific polymerase chain reaction for these four KCNH2 polymorphisms: 1539C > T, 1956T > C, 2350C > T, and 2690A > C. Mean QTc interval is 408 ms (SD: 24). Molecular docking was performed on all four KCNH2 polymorphisms to investigate the impact of the nucleotide changes on methadone binding. Based on multiple regression analysis, only 1539T > C polymorphism (β_adjusted: 10.506 (95% CI:0.846, 20.166), p = 0.033; recessive model), serum methadone trough (β_adjusted: 0.025 (95% CI: 0.006, 0.043), p = 0.009), potassium (β_adjusted: -8.756 (95% CI: -15.938, -1.575), p = 0.017) and magnesium (β_adjusted: -106.226 (95% CI: -159.291, -53.161), p < 0.001) levels were significantly associated with mean QTc. Molecular docking analysis resulted in good binding-energy values between the 1539C > T and methadone, with the formation of hydrophobic and π–π stacking interactions, suggesting that 1539C > T was the newly discovered SNP involved in QTc prolongation. In conclusion, the 1539C > T KCNH2 polymorphism is associated with QTc prolongation in our MMT recipients, necessitating QTc monitoring to prevent methadone-associated cardiotoxicity in this Malay MMT population.

Introduction

Methadone is widely used in opioid dependence treatment and was introduced in Malaysia in 2005 to curb HIV transmission [1]. This initiative has shown positive outcomes, as HIV incidence in Malaysia declined from 28.4 per 100,000 population in 2002 to 11.7 per 100,000 in 2014 [2]. However, methadone is linked to serious adverse effects, including QTc prolongation and torsade de pointes (TdP) [3,4]. Factors such as high methadone doses, stimulant use [5], liver dysfunction, CYP3A4 inhibitors, and electrolyte imbalances contribute to QTc variability, though unidentified factors may also be involved [6].

Pharmacogenomics has identified genetic variations influencing drug response, including polymorphisms in ion channel genes linked to long QT syndromes (LQTS) and TdP [7–9]. One such gene, the potassium voltage-gated channel subfamily H member 2 (KCNH2) gene, encodes the human ether-à-go-go related gene (hERG) potassium channel, which regulates cardiac repolarization. Polymorphisms in KCNH2, such as 1039C > T (Pro347Ser) and 2350C > T (Arg784Trp), are associated with drug-induced TdP [10–13]. Methadone and other hERG-blocking drugs may pose higher LQTS risk in individuals carrying these variants [14,15].

KCNH2 polymorphisms in opioid-dependent individuals remain underexplored in Southeast Asia, with most studies focusing on general populations or small cohorts [13,16]. Research on their interaction with methadone is especially limited, including data on their frequency in Malay populations [17]. Identifying these variants could improve methadone safety by enabling personalized dosing strategies [16].

This study hypothesizes that KCNH2 variations contribute to methadone-related QTc prolongation. Understanding these genetic risk factors could help tailor MMT dosing, enhancing patient safety and treatment efficacy [18]. To date, only Hajj et al. have investigated KCNH2 polymorphisms in MMT recipients, linking the 2690A > C polymorphism and methadone dosage to QTc prolongation in a French cohort [16]. However, no study has examined this association in the Kelantanese Malay population, making this the first endeavour to explore the relationship between KCNH2 polymorphisms and QTc intervals in opioid-dependent Kelantanese Malay MMT recipients.

Methods

Study design and participants

This cross-sectional study involved 111 Malay opioid-dependent patients who are free from cardiac structural defects and attended the Hospital Universiti Sains Malaysia (HUSM) Psychiatric Clinic and four other MMT clinics (SAHABAT-Kota Bharu, Wakaf Bharu, Ketereh and Selising), Kelantan, Malaysia between 1^st of March 2011 and 31^st of October 2012. The reason for such racial selectivity is to minimize cross-ethnic variations in the KCNH2 polymorphism frequency. The study enrolled participants who were taking prescribed similar daily doses of methadone as described in previous research [19].

The study’s inclusion criteria are (a) patients aged 18 years and above; (b) had stable methadone dosage for at least six months during the MMT; (c) of Kelantanese Malay ancestry of up to three generations; (d) a history of good compliance to the Directly Observed Therapy (DOT)/ takeaway program; and (e) consented to study participation. We excluded the patients if they had cardiac structural defects (e.g., atrial or ventricular septal defects (ASD/VSD), valvular heart diseases, aortic coarctation), received other QTc-prolonging pharmacotherapeutic agents (e.g., haloperidol, amitriptyline, amiodarone, nelfinavir, efavirenz), had aggressive behavior or other active psychiatric illnesses, chronic medical and surgical ailments and patients who were not English or Malay proficient. The ethics approval for the research protocol was granted by the Universiti Sains Malaysia (USM)’s Human Research Ethics Committee (HREC), Kelantan, Malaysia (Reference number: USMKK/PPP/ Ethics.Com/2010(19), approval date: 01/02/2010). This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki for research involving human participants and all participants provided written informed consent prior to their inclusion in the study.

Initially, 140 eligible MMT patients regularly attended the recruitment centres and were, therefore, available for study participation. However, 29 patients were excluded for the following reasons: participation refusal (n = 10), mixed ancestry (n = 9), incomplete follow-up data (n = 6), and more than 20% missing clinical and genotyping data (n = 4). The participants were recruited using convenience sampling due to the paucity of suitable patients available for study recruitment. The flow of participant recruitment is summarized as a Strengthening the Reporting of Observational studies in Epidemiology (STROBE) [20] diagram in Fig 1.

[Figure omitted. See PDF.]

This study is registered with the ClinicalTrials.gov registry (ID: NCT03603158), and the full dataset is available from the Harvard Dataverse repository (doi: https://doi.org/10.7910/DVN/WMK7HL). The Strengthening the Reporting of Genetic Association Studies (STREGA) guideline was followed to improve the quality and clarity of reporting for this study [21].

Sample size calculation

The sample size for detecting the prevalence of hERG polymorphisms was calculated using a single-proportion formula [22]:

(1)

Where p is the proportion of subjects with the SNPs, z = 1.96, and Δ is the precision. Based on the information obtained from Koo et al., the largest sample size required is for detecting 1539T > C SNP [17]. The proportion of Malay subjects with C allele for 1539T > C SNP is 0.441, and Δ is fixed at 0.095. Hence, the biggest sample size required is therefore 105 subjects. Sample size calculation for detecting the differences in mean QTc between different genotypes in all KCNH2 polymorphisms under study could not be performed due to the absence of satisfactory information from previous literature to aid such an endeavor.

Clinical data collection

The consented patients were interviewed to gather socio-demographic information, history of drug addiction, drug dependency patterns, other drug usage, psychiatric illnesses, and relevant treatment-related issues. Subsequently, a validated Malay version of the Subjective Opioid Withdrawal Scale (SOWS) questionnaire [23] was administered to assess opioid withdrawal symptoms.

Five ml of venous blood was collected from each participant to ascertain serum biochemical profiles (K⁺, Mg²⁺, Ca²⁺), methadone trough levels, and KCNH2 genotyping. Urine dipstick (FastScreenTM Drug Combo Test 6-in-1. Reszon Diagnostics, Subang Jaya, Malaysia) tests were performed for methamphetamine, cannabis or marijuana, methylenedioxyl methamphetamine (MDMA) and benzodiazepines. Patients were required to provide samples at the urine collection point, accompanied by the clinic staff. The color and temperature of the specimen were recorded.

A standard 12-lead electrocardiogram (ECG) was performed by an experienced internist (NM) using validated Welch Allyn CP 50™ ECG (Welch Allyn Australia Pty Ltd., New South Wales, Australia) machines. All ECGs were recorded by the same instrument and printed at a paper speed of 25 mm/s and voltage of 10 mm/mV. QT interval was quantified in milliseconds (ms) utilizing the method elucidated in Wedam et al. [24] and Ehret et al. [6]. Fridericia’s formula (QTc = QT/(RR)^0.33) was then employed to calculate the QTc interval [25]. No patients reported any prior history of TdP, syncope, or seizure. All personnel involved in QT_c measurement were masked to information on other biochemical parameters and serum methadone trough concentrations.

Serum methadone trough level measurement

The quantification of serum methadone trough level (methadone C_trough) was executed based on the previously described protocol [26]. Serum methadone trough concentration Methadone C_trough was measured using a validated Methadone ELISA kit invented by the researchers of Pharmacogenetic and Novel Therapeutic Cluster, INFORMM that has optimal sensitivity [27]. The quantification limit of this method is 50 ng/ml and it can be thus typically employed in a routine clinical context.

DNA extraction

DNA extraction and KCNH2 genotyping were conducted at the Pharmacogenetic Research Laboratory, Institute for Research in Molecular Medicine (INFORMM), Kelantan, Malaysia. About 2.5 ml blood samples for genotyping were collected in sodium citrate tubes from all participants after QTc measurement, and these were rapidly transferred to -20 °C freezer in ice boxes. Genomic DNA from peripheral leucocytes was extracted from 200 µl of each participant’s whole blood using the QIAamp DNA Blood Mini Kit (Qiagen Gmbh, Hilden, Germany) according to the manufacturer’s instructions. The quantity and quality of the extracted genomic DNA were evaluated by using the NanoDrop ND-1000 Spectrophotometer (NanoDrop Technologies, Inc. Wilmington, USA) with measurements done at 260 and 280 nm, and these were stored in -20 °C freezer.

Design of allele-specific PCR for KCNH2 polymorphism detection

All samples were subsequently genotyped for the following KCNH2 polymorphisms; rs1805120 (1539C > T, F513F), rs1137617 (1956T > C, Y652Y), rs12720441 (2350C > T, R784W), and rs1805123 (2690A > C, K897T) using a two-step nested allele-specific polymerase chain reaction (PCR). The reason why 1539C > T, 1956T > C, and 2690A > C were chosen is that they are the most commonly encountered KCNH2 SNPs in the general Malay population, and 2690A > C is directly associated with QTc prolongation in methadone recipients [16,17]. For 2350C > T, several studies verified its presence and association with drug-associated QT prolongation and TdP [13,28].

In the first PCR (PCR1), we amplified exons 6, 8, 9 and 11 of KCNH2. We subsequently utilized the PCR1 products to identify the presence of 4 SNPs using allele-specific primers in the second PCR (PCR2) performed in 2 parallel sets/reactions to ensure optimal product size; set A (2690A > C and 1956T > C) and B (2350C > T and 1539C > T). For PCR2, the primers were meticulously designed to possess a mismatch at specific 3′-ends and were specific to both the variant and wild-type DNA sequence at a particular locus. The primers were also manipulated to distinguish single nucleotide alteration/allele at a particular locus during PCR amplification. We utilized the Basic Local Alignment Search Tool (BLAST) tool, which is available at http://www.ncbi.nlm.nih.gov/blast, to verify the primers’ specificities [29]. All primers were synthesized by Invitrogen© USA, and purchased from a local vendor, and the melting temperature (T_m) was approximated as follows: T_m = 2(A + T) + 3(C + G), and the optimal T_m was identified via gradient PCR method employing the Veriti 96-well Thermal Cycler® (Applied Biosystem, Foster City, CA). S1 Table provides further information on the primers used in PCR1 and PCR2 reactions.

PCR protocols

The final reaction mixture for PCR1 consists of 2 µl of DNA, 10.5 µl of primer mix (volume and concentrations of each primer were shown in S1 Table plus 2.5 µ L of dH₂O), and 12.5 µl of ready-made commercial MyTaq™ Mix (Bioline®, London, UK) mastermix, a total volume of 25 µ L per reaction. The PCR amplifications of exons 6, 8, 9 and 11 were performed using GeneAmp® PCR system 9700 Perkin Elmer (Applied Biosystems, Foster City, CA) under the following thermocycle profiles; an initial pre-denaturation phase for 95 °C for 1 minute and this is succeeded by 25 cycles of denaturation at 95 °C for 1 minute, annealing at 65 °C for 15 seconds and extension 72 °C for 10 seconds. For the last step, the final extension phase at 72 °C for 7 minutes was also incorporated.

The PCR1 products were separated using ethidium bromide-stained 2% agarose gel (Promega Corporation, Madison, USA) electrophoresis in mini electrophoresis tanks (Primo Submarine Gel system^TM. ThermoQuest, Holbrook, New York, USA) at 130V for 90 minutes. The PCR products were subsequently visualized under ultraviolet (UV) light and photographed. The laboratory personnel who conducted the experiments were blinded to QTc measurements and other biochemical profiles of study participants.

Two µl of PRC1 amplicons, which were diluted to 1:20, served as PCR2 DNA templates. The total volume of the master mix for each set A and B reaction is 10 µl and this consists of dH₂O (set A: 6.0 µl; set B:5.75 µl), 1x reaction buffer (2.5 µl; both sets), MgCl₂ (set A: 0.5 µl; set B: 0.75 µl) and dNTP (0.5 µl; both sets) and Taq polymerase (0.5 µl; both sets). For the primer mix, the total volume is 13 µl for both sets A and B, which comprises wild-type and mutant primers (S1 Table) and dH₂O (9 µl). The final volume for each parallel PCR reaction is 25 µ L.

The thermocycler profile for PCR2 is similar to PCR1 except for the following: annealing at 69 °C for 30 seconds and extension at 72 °C for 4 seconds. The PCR2 products were separated and visualized using the same modality as PCR1 (S1 Table).

Direct DNA sequencing of KCNH2 polymorphism

The PCR1 products were submitted for direct sequencing. QIAquick PCR purification kit (Qiagen, USA) was used to purify PCR products. Sequencing was performed using 3130XL genetic analyzer DNA sequencer (ABI, USA). The published sequences for KCNH2 in the NCBI [accession number: NC_000007.13] were subsequently used to compare the DNA sequencing results.

Data and statistical analysis

The frequencies of the KCNH2 allele and genotype were calculated using the gene counting technique based on the samples’ observed genotype numbers. The Hardy–Weinberg equilibrium (HWE) departure was statistically tested by comparing the observed and expected frequency of homozygous genotypes using the Levene and Haldane’s exact test based on the sum equally likely or more extreme samples (SELOME) [30–32]. An R package, HardyWeinberg version 1.59, implemented on the R platform (R Core Team, 2017. Vienna, Austria) was deployed to achieve this. The magnitude of HWE deviation was calculated and represented using Lindley’s α [33], and if the value was more than 0.5, the SNP was definitely dropped from further analyses. The formula for Lindley’s α is:

(2)

where P_aa, P_AA, and P_Aa represent the proportion of each genotype [33,34].

Simple linear regression analysis was used to screen for variables (KCNH2 polymorphisms, serum methadone trough levels, serum potassium, magnesium, calcium levels, patient’s age and gender) that may be potentially associated with QTc intervals. The independent predictors with p values below 0.25 in the univariable analysis were selected for multivariable model building using backward and forward stepwise multiple linear regression (p_entry = 0.05, p_removal = 0.10). The presence of multicollinearity and effect modifiers were also respectively examined by using variance inflation factor (VIF)’s threshold of less than 10 and assessing the significance of the 2-way interaction term for KCNH2 genotypes and serum methadone trough level. Model assumptions (normality, homoscedasticity, and independence of residuals) were assessed using Studentized residuals versus predicted value plots and a Q-Q plot of residuals. The IBM Statistical Package for the Social Sciences (SPSS) Version 29 (Released September 2022. Armonk, NY: IBM Corp) was utilized to carry out statistical analyses. All estimates were reported with their 95% confidence intervals (Cis) and significance threshold were fixed at 0.05 (two-tailed).

Identification of functional domains in KCNH2 protein sequences

Proteins comprise various domains, each associated with distinct functions. Therefore, examining SNPs in the KCNH2 at the domain level may offer valuable insights into molecular processes that may affect QTc prolongation among MMT recipients. Before constructing the protein structure model, an integrated database of protein families, domains, and functional sites called InterPro was utilized to determine the domain architecture of KCNH2 protein sequences [35]. The location of the mutations was subsequently mapped across the functional domains and interdomains of the KCNH2 protein using Illustrator for Biological Sequences version 2.0 [36].

Protein structure modeling of KCNH2 polymorphisms

To analyze the impact of mutations on the protein structure, protein structure modeling was carried out on the KCNH2 protein sequence for identified SNPs, such as 1539C > T, F513F (synonymous); 1956T > C, Y652Y (synonymous); 2350C > T, R784W (missense); and 2690A > C, K897T (missense). Firstly, the protein sequence of KCNH2 was obtained from the UniProt database [37]. The sequences of KCNH2 proteins were then mutated by manually altering amino acids of both missense variants R784W and K897T from arginine (R) and lysine (K) to tryptophan (W) and threonine (T), respectively. However, the alteration in the coding sequences KCNH2 (1539C > T and 1956T > C) did not result in any change to the amino acid. Therefore, the KCNH2 F513F and Y652Y proteins were regarded as wild-type (WT) proteins since their structures remained unaltered. To model the protein structures, a Support Vector Machine (SVM) was employed in conjunction with the SWISS-MODEL server [38]. The SVM algorithm integrates interface conservation, structural clustering, and various template features to generate a Quaternary Structure Quality Estimate (QSQE) and a Global Model Quality Estimate (GMQE) [38]. The accuracy of the structural assessment was measured through GMQE and QSQE scores, which both range between 0 and 1. After the assessment, SWISS-MODEL was employed to conduct a Ramachandran plot analysis on the models to predict the stereochemical quality of the predicted protein structures. Finally, the protein structures were visualized and analyzed using UCSF Chimera software [39].

Mutational effect prediction of missense variant 2350C > T (R784W) and 2690A > C (K897T)

Missense variants can result in either benign or deleterious effects, as they may influence protein stability and potentially disrupt the binding between the proteins and ligands (drugs). Therefore, by employing statistical and ML-based servers like Predicting the Effects of Mutations on Protein Stability (PremPS) [40], SVM-based approach DUET [41], and mutation Cutoff Scanning Matrix (mCSM) [41], we measured the mutational effects of the missense variants 2350C > T (R784W) and 2690A > C (K897T) on protein stability, which may support their impact on methadone binding compared to 1539C > T (F513F) and 1956T > C (Y652Y).

Molecular docking analysis

To associate SNPs with QTc prolongation among MMT recipients, the binding affinity between methadone and KCNH2 proteins with 2350C > T (R784W), 2690A > C (K897T), 1539C > T (F513F) and 1956T > C (Y652Y) was assessed by performing protein-ligand docking analysis using CB-DOCK2 [42,43]. CB-DOCK2 is a docking server that employs CurPocket, an algorithm based on protein surface curvature, to identify cavities and facilitate molecular docking using AutoDock Vina [43]. Before docking, ligand preparation involved obtaining the structure of methadone (CID: 4095), dextromethadone (CID: 643985), and levomethadone (CID: 22267) from PubChem [44], followed by converting the structure from SDF to the MOL2 file format using OpenBabel [45]. Following protein modeling, the CASTp Server was deployed to predict the active sites/cavities in the KCNH2 protein [46]. To assess the impact of KCNH2 polymorphisms on drug binding, the predicted structures of the MT and WT KCNH2 proteins were docked with methadone, dextromethadone, and levomethadone using CB-DOCK2, with contact residues for docking specified to the predicted active sites/cavities. All three drug structures were docked within the cavity using the same docking size and region to compare their binding affinities to the pocket of the MT and WT KCNH2 proteins. The binding energy (kcal/mol) of the MT and WT KCNH2 complexes with methadone, dextromethadone, and levomethadone was subsequently evaluated. The binding affinity of WT KCNH2 served as the control, while the binding affinity of MT KCNH2 was compared to that of WT KCNH2 to determine the impact of the polymorphism on drug-protein binding stability.

Results

Patients’ clinical characteristics

A total of 111 subjects were included in this study. The average age of the MMT recipients was 36.02 (SD: 7.02, range: 20 − 59) years old. Nearly all study participants were males (male: n = 109 (98.2%), female: n = 2 (1.8%)). Consequently, the gender-stratified analysis could not be conducted due to an inadequate number of female participants.

Based on urinary drug screening, 62 (55.9%) subjects were drug-free, and in those with positive drug screen, the most frequently abused drug was morphine (n = 5, 4.5%), although 7 (6.3%) subjects were multiple drug abusers, 36 (32%) subjects were injectable drug users and 1 (1.1%) subject with unknown drug screening status. No other significant co-morbidities that may affect QTc prolongation (e.g., hypertension, thyroid dysfunction, renal and hepatic dysfunction) were observed in the participants. The summary of relevant clinic-demographic characteristics and SOWS scores is given in Table 1.

[Figure omitted. See PDF.]

KCNH2 genotype and allele frequencies

The genotyping of DNA samples from the 111 drug users was successfully accomplished for the four KCNH2 SNPs. The genotype frequencies observed in this study were similar to those predicted by the HWE for 1539C > T (p_{exact(SELOME)} = 0.4513) and 2690A > C (p_{exact(SELOME)} = 0.1101) only, whilst the observed genotype frequencies for 1956T > C (minor allele frequency (MAF): 0.730) and 2350C > T (MAF: 0.036) significantly deviated from the expected genotype frequencies under HWE (p_{exact(SELOME)} = < 0.001 (Lindley’s α = 1.077) and 0.004 (Lindley’s α = 0.860), respectively. Full results are available upon request). Consequently, both 1956T > C and 2350C > T were excluded from further statistical analyses. Allele frequencies for the remaining KCNH2 polymorphisms and comparisons with results from prior studies are shown in Table 2. Examples of amplified SNPs are available in Fig 2.

[Figure omitted. See PDF.]

[Figure omitted. See PDF.]

M: DNA ladder; Wt: Wild type; Mt: Mutant.

KCNH2 polymorphisms and QTc intervals

Based on the genetic additive model (Table 3), there are significant differences between 1539C > T polymorphisms and QTc. However, only CT and TT genotypes have significant QTc differences based on post-hoc multiple comparison whilst the other comparisons were insignificant. No significant differences in QTc were found among 2690A > C genotypes. Due to inconclusive findings, the analyses were modified, and genetic recessive and dominant models were used instead to further examine the associations between the two SNPs and QTc intervals.

[Figure omitted. See PDF.]

Using simple linear regression analysis, only 1539C > T SNP, methadone trough levels, serum potassium, and magnesium concentrations reached statistical significance. These covariates were subsequently selected for multivariable model building using multiple linear regression. The effects of 2690A > C SNP and other variables on QTc are not statistically significant and, therefore, were not considered. Table 4 indicates that all 4 variables retained their significance and, therefore, were included in the predictive model for mean QTc. No significant multiplicative interaction effects or multicollinearity were found between the covariates.

[Figure omitted. See PDF.]

Model assumption check revealed that all multiple linear regression assumptions were fulfilled, and no interaction terms were deemed significant for inclusion in the final predictive model. The predictive model for mean QTc is thus given by the regression equation below:

(3)

We can thus interpret that a subject with TT genotype experiences 10.506 ms longer mean QTc compared to TT or CT genotype after adjustment was made for the effects of serum methadone trough concentration and serum potassium and magnesium levels. This QTc prolongation exceeds the 5–10 ms threshold that regulatory agencies, such as the US Food And Drug Agency (FDA) and International Council for Harmonisation (ICH), consider for potentially concerning QT prolongation risk in drug safety studies [47]. Although this threshold is typically applied to drug-induced effects, even small QTc increases may be clinically relevant, particularly in individuals with additional genetic risk factors for arrhythmias.

Protein structure models of WT and MT KCNH2

Protein structure models of WT KCNH2 (1539C > T (F513F) and 1956T > C (Y652Y)) demonstrated 96.37% sequence identity with potassium voltage-gated channel subfamily H member 2 (O35219.1.A) structure generated by AlphaFold v2, whereas both MT KCNH2 (2350C > T (R784W) and 2690A > C (K897T)) indicated 96.29% identity to the similar template. The homology models exhibited an exceptionally conserved structure with a high degree of sequence identity (>96%) and high structural similarity, as indicated by the root mean square deviation (RMSD) values of 0.001 Å after superimposing WT and MT KCNH2 to the template, respectively. Hence, it is suggested that the predicted benign missense SNPs, 2350C > T (R784W) and 2690A > C (K897T), do not cause any conformational changes in the KCNH2 protein. Structure assessment revealed 76.40% of all residues in Ramachandran favored regions, with 4.52% being rotamer outliers and 12.53% Ramachandran outliers. Although the favored regions were less than 90%, the models were reliable enough to perform virtual screening, as the models showed conserved features, especially in functional domain regions (Fig 3).

[Figure omitted. See PDF.]

The structure also highlighted the positions of four SNPs: 1539C > T (F513F), 1956T > C (Y652Y), 2350C > T (R784W), and 2690A > C (K897T).

The KCNH2 protein comprised the Per-Arnt-Sim (PAS) domain (37–132 aa), ion transport domain (409–668 aa), cyclic nucleotide-binding domain (742–860 aa), and cytoplasmic domain (664–1159 aa). Both synonymous SNPs, 1539C > T (F513F) and 1956T > C (Y652Y), were located within the ion transport domain region. Additionally, a substitution from arginine to tryptophan at position 784 was observed within the cyclic nucleotide-binding domain, and a substitution from lysine to threonine occurred at position 897 within the cytoplasmic domain region.

Effects of missense mutations on KCNH2 protein stability

Previously, the lack of differences between favorable residues of WT and MT KCNH2 indicates that the core structure of the aforementioned domains was not affected by the 2350C > T (R784W) and 2690A > C (K897T), and none of these variants could cause conformational changes that would affect the functions of KCNH2. However, both 2350C > T (R784W) and 2690A > C (K897T) have been demonstrated to impact the stability of the KCNH2 protein. The missense SNPs were highly predicted to destabilize the surface of the KCNH2 protein, as indicated by the predicted stability change (ΔΔG) values (Table 5). Thus, despite both missense SNPs being benign, they appeared to destabilize the protein structure of KCNH2, which could potentially disrupt the contact residue of drug binding in MMT recipients with these SNPs, leading to shorter QTc intervals.

[Figure omitted. See PDF.]

Drug binding affinity evaluation between WT and MT KCNH2 with methadone

To infer the possible role of KCNH2 polymorphisms in affecting QTc prolongation among MMT recipients, the binding affinity of methadone, levomethadone, and dextromethadone to each KCNH2 protein cavity was predicted via a molecular docking study. The structural analysis of KCNH2 identified 14 cavities, with Phe513 (F513F), Tyr652 (Y652Y), and both Arg784 (R784) and Trp784 (W784) being determined as binding pocket residues, except for Lys897 and Thr897 (K897T). Among these cavities, the target cavities of 1539C > T (F513F) (13, 50, 52, and 75) and 1956T > C (Y652Y) (5, 15, 40, and 76) were predicted to be situated within the ion transport domain, while 2350C > T (R784W) was found in cavities 1, 12, 20, 68, 69, and 90, which are located within the cyclic nucleotide-binding domain (Table 6).

[Figure omitted. See PDF.]

Methadone was docked and redocked with levomethadone and dextromethadone into a similar cavity or region of interest using a similar docking center and size. K897T was excluded from molecular docking analysis as the mutation was not determined to be located at the binding pocket of KCNH2. All three types of methadone exhibited more negative binding energy values to 1539C > T (F513F) compared to 1956T > C (Y652Y), 2350C > T (R784), and 2350C > T (W784). Overall, the drug binding energy score of F513F ranges from 6.5 to 6.9 kcal/mol, followed by Y652Y (5.8–6.1 kcal/mol), and R784 (3.7–3.8 kcal/mol). Out of the four mutations, W784 was also shown to have the highest binding energy score, suggesting a destabilizing effect on the KCNH2 protein, leading to ineffective drug binding (Table 5). In this study, we observed the interaction between Phe513, Tyr652, Arg784, and Trp784 with methadone as a ligand to decipher the possible role of polymorphisms in affecting the binding affinity of the drugs (Fig 4).

[Figure omitted. See PDF.]

Interestingly, the racemic methadone mixture, levomethadone, and dextromethadone were all found to interact with Phe513, forming three hydrophobic interactions and a π–π stacking in all predicted cavities (Fig 5). The strongest binding affinity is observed in cavity 13, with binding energies of -6.9 kcal/mol, -6.9 kcal/mol, and -6.8 kcal/mol, respectively.

[Figure omitted. See PDF.]

On the other hand, Tyr652 only forms hydrophobic interactions with all three methadone compounds (racemic mixture, levomethadone and dextromethadone), with energy scores lower than the Phe513-methadone binding. Meanwhile, both Arg784 (WT KCNH2) and Trp784 (MT KCNH2) had the lowest binding energy score with a range between -3.7 to -4.8 kcal/mol by forming a hydrophobic bond with all methadone within cavity 90 (Fig 6). Based on our structural and docking study, it is strongly suggested that the 1539C > T (F513F) polymorphism is associated with QTc prolongation in MMT recipients. However, more in-depth validation is required to confirm this association in future studies.

[Figure omitted. See PDF.]

Discussion

This research primarily aims to evaluate the link between KCNH2 genetic variants and QTc intervals in opioid-dependent MMT recipients who received stable daily methadone dosages. Of the four genetic polymorphisms we studied, one was identified frequently in our Malay MMT population (1539C > T (F513F), and one was less frequent (2690A > C (K897T)). No definitive conclusion could be about the prevalence of 1956T > C (Y652Y) and 2350C > T (R784W) minor alleles since both SNPs violated the HWE principle, which can be attributed to sampling bias secondary to genotyping error, population stratification, or population inbreeding [48]. This is supported by the high positive values of Lindley’s α (exceeding 0.5) for both 1956T > C and 2350C > T SNPs, signifying excess proportions of homozygotes [33,34].

Besides, this is also the first report confirming the association between rs1805120 (1539C > T, F513F) SNP and QTc interval prolongation in methadone recipients of the Kelantanese Malay descent, a subethnic group of Malay population living in the North-eastern region of Peninsular Malaysia that has genetic similarities to the Semang (one of the Malaysian aboriginal ethnic) and Indian populations [49], based upon the recessive model. We believe this finding has not been documented by any previous study conducted in other populations. We also established the first predictive model for QTc prolongation in Malay MMT recipients that also incorporates other significant clinical covariates such as methadone trough level, serum potassium and magnesium levels. We also found no significant multiplicative interaction effects among significant QTc predictors (1539C > T, methadone trough level, serum potassium, and magnesium concentrations), thus further refining the findings of Hajj and colleagues [16]. In contrast, Zerdazi and colleagues demonstrated significant multiplicative association between the intronic variant, rs11911509 in KCNE1 encoding the beta subunit of the potassium channel, and methadone dosage and QT_c prolongation in opioid-dependent French subjects on MMT [50].

Furthermore, based on the recessive genetic model, we also established that MMT patients with CC genotype (2690A > C SNP) have more reduced mean QTc intervals than MMT recipients having AA/AC genotypes (406.333 ms vs 408.287 ms), albeit being statistically insignificant (p = 0.892). Nevertheless, our results still parallel the findings of Hajj et al. [16], who demonstrated that based on the recessive model (AA + AC vs AC), the mean QTc for the CC group was much shorter than the AA + AC group (383 ms vs 419 ms, p = 0.004). This is further corroborated by Gouas and co-workers [51], who demonstrated that the C allele of 2690A > C polymorphism is more frequently encountered in French healthy subjects with short QTc intervals. Besides, these observations were strengthened by Bezzina et al. [52] and Marjamaa et al. [53], who verified such associations. Our insignificant finding is attributed to the small study sample size, which is evident from the low proportion of the minor C alleles in our MMT subjects (MAF: 0.108 vs 0.341, Table 1), resulting in the reduction of our study power. Furthermore, this discrepancy can also be explained by the different formulas used for calculating QTc (Bazett’s formula) by Hajj et al. [16]. Vandenberk and colleagues [25] demonstrated that Fredericia’s method is superior to Bazett’s formula in QTc computation since the former had enhanced capacity for predicting 30-day and 1-year all-cause fatality. Hence, the predictive potential of 2690A > C on QTc prolongation remains an open debate for our MMT population.

The mechanisms explaining the basis of our observed association between 1539C > T (F513F) SNP and QTc interval are, however, still unclear. We speculate that a synonymous SNP may alter co-translational protein folding, which results in protein structural modification and substrate specificity [54]. Furthermore, synonymous KCNH2 nucleotide alterations that decrease GC content, scarce codon, and the sum of mRNA secondary structure result in mRNAs that are less stable, diminished half-life, and translated at a lower rate into hERG ion channel proteins compared to the native KCNH2 mRNA [55]. This is corroborated by Wang et al [56], who discovered a significant relationship that links the synonymous 1539C > T SNP to increased predisposition to atrial fibrillation in individuals of Han Chinese descent, which they attributed to left atrial enlargement. Hence, we postulate that the 1539C > T (F513F) synonymous substitution produces QTc lengthening effects via the above mechanisms, and thus, greater emphasis is required to further characterize the definitive link between synonymous SNPs and QTc interval prolongation.

Furthermore, there are 4 transcript variants for KCNH2. Using UCSC Genome Browse (https://genome.ucsc.edu/), all four SNPs are present in transcript variant 1 of KCNH2, which is its major transcript variant. Upon translation, this yields the isoform a of α subunit of I_Kr potassium channel, the longest isoform of KCNH2 protein. In other transcript variants, however, not all four SNPs are present. For instance, rs1805123, which is located at exon 6 of KCNH2, is absent in transcript variant 2. Therefore, the precise functionality of non-major KCNH2 splice variants should be examined to understand their effects on QTc prolongation.

Coupled with a molecular docking study, the 1539C > T (F513F) mutation produces best binding results by forming π-π stacking and three hydrophobic interactions with methadone, levomethadone and dextromethadone compared to other SNPs. Hydrophobic interactions are crucial for stabilizing drugs at binding sites [57] as they compensate for enthalpy-entropy balance, enabling hydrophilic residues to shed water molecules and facilitating hydrogen bonding and electrostatic attraction among them at the binding site [58]. Hydrophobic interactions result from water molecules occupying lipophilic binding sites, thus preventing hydrogen bonding with the receptor. Upon release from the hydrophobic pocket, the water molecules can form strong hydrogen bonds with bulk water [59], establishing that hydrophobic interactions of Phe513 drive HCNH2 conformational changes upon methadone binding. Besides, π-π stacking facilitates diverse drug delivery and serves as a pivotal force in loading drugs into systems without altering their structural or functional properties [60]. The stacking is stabilized by the aromatic side chains present in the ligand-binding site due to the presence of phenylalanine, tryptophan, histidine, and tyrosine residues. Despite being considered a weak, non-covalent dispersion force, π–π interactions are critical in protein folding, thermal stability, and ligand binding [61]. Aromatic rings not only enhance drug design but also improve affinity and specificity, serving as the foundation for lead compound skeletons that can be refined to fulfill both target and off-target binding criteria [62]. Based on these arguments, our discoveries further supplement the findings by Titus-Lay and co-workers, who showed QT_c prolongation was caused by both methadone isomers [63], instead of only S-isomers [64].

There are several study limitations that may bias the validity of our findings. Firstly, the gender effect on QTc could not be sufficiently adjusted in view of the restricted pool of female participants in our study cohort. This is highly inevitable due to most opioid-dependent users are male in Malaysia [65]. Next, we were unable to control the effects of population stratification and other genetic confounders (e.g., assortative mating, gene-environment interaction) due to our small study sample size. A study with a much larger sample size based on the genome-wide association study (GWAS) design is thus required to statistically adjust for the effects of genetic confounders adequately. Apart from those, the intra-observer reliability for QTc interval measurements was not assessed. Future studies should consider evaluating intra-observer variability using intraclass-correlation coefficient (ICC) to ensure QTc measurement consistency, especially when manual QTc measurements are performed by a single assessor. Besides, two of our studied SNPs violated Hardy-Weinberg equilibrium, and this calls our genotyping quality and our relatively small sample size into question. Furthermore, our small sample size and the cross-sectional design of this study may preclude the establishment of the causal relationship between KCNH2 polymorphism and QT_C prolongation due to lack of temporal order of the cause and effect. A prospective cohort design therefore may serve as a more reliable way of ascertaining the causality of such an association. Apart from that, the predictive model (equation 3) obtained is not properly validated. Ideally, a predictive model should be validated using a fresh population sample, but several internal validation techniques such as leave-one-out cross-validation (LOOCV), k-fold cross-validation, and bootstrapping are also good alternatives. These are now addressed in our ongoing research work. The generalizability of this study’s findings is also restricted to the Kelantanese Malay ethnicity due to the exclusive racial composition of our study sample. However, our findings might be relevant for future studies conducted within similar ethnicity, which is also a major ethnicity in other South-East Asian (SEA) countries, for example, Indonesia, Brunei, Singapore, and Southern Thailand.

Our findings have several therapeutic ramifications. First, our discoveries offer a considerable impetus for recommending personalized treatment approaches for MMT patients who are homozygous carriers of the mutant-type genotype (1539TT) in managing their opioid dependence. The identified KCNH2 polymorphisms may function as a critical predictor to tailor the treatment strategies of opioid dependence in MMT recipients to optimize the benefits and mitigate the adverse impacts associated with MMT. Therefore, in addition to regular electrolye and serum methadone trough level monitoring and correction for hypokalemia and hypomagnesaemia, clinicians should also screen for variations in the SNPs 1539C > T in the Kelantanese Malay MMT patients and those harbouring the TT genotypes should receive lower starting methadone dose and be closely monitored through routine ECG screening for QTc prolongation. Besides, our findings can also be incorporated with other existing prediction models [66] for predicting methadone-associated cardiotoxicity that can be specifically tailored for our Kelantanese Malay MMT patients, a population living in a state experiencing the highest number of drug abuse-related hotspots in Malaysia [67].

Conclusion

In conclusion, this study identifies the synonymous 1539C > T polymorphism in KCNH2 as a contributor to QTc prolongation in opioid-dependent MMT recipients. Molecular docking suggests that this variant may alter methadone binding within the hERG channel, potentially affecting cardiac repolarization. The developed predictive model, integrating genetic and clinical factors, may aid in risk stratification and individualized methadone dosing to prevent methadone-associated cardiotoxicity. These findings underscore the need for preemptive QTc monitoring and stringent pharmacogenetic screening to enhance MMT safety in high-risk populations.

Supporting information

S1 Table. Primers for the 1st and 2nd (Set A and Set B) PCR reactions.

https://doi.org/10.1371/journal.pone.0322724.s001

(DOCX)

S1 Fig. The original gel electrophoresis image for Fig 2 in the main manuscript.

https://doi.org/10.1371/journal.pone.0322724.s002

S2 Fig. Additional gel electrophoresis image for KCNH2 polymorphisms.

https://doi.org/10.1371/journal.pone.0322724.s003

Acknowledgments

We posthumously thanked the late Professor Rusli Ismail and Dr. Muslih Abdulkarim Ibrahim for the initial conceptualization of this research.

References

1. 1. Reid G, Kamarulzaman A, Sran SK. Malaysia and harm reduction: the challenges and responses. International Journal of Drug Policy. 2007;18(2):136–40. pmid:17689356

* View Article

* PubMed/NCBI

* Google Scholar

2. 2. Ministry of Health Malaysia. National Strategic Plan for Ending AIDS 2016-2030. (Ministry of Health Malaysia, Putrajaya, 2015). [cited 21 June 2024. ]. Available from: https://www.moh.gov.my/moh/resources/auto%20download%20images/58510b70ca319.pdf

* View Article

* Google Scholar

3. 3. Alinejad S, Kazemi T, Zamani N, Hoffman RS, Mehrpour O. A Systematic Review of the Cardiotoxicity of Methadone. EXCLI Journal. 2015;14:577–600. pmid:26869865

* View Article

* PubMed/NCBI

* Google Scholar

4. 4. Mohamad N, Abdul Jalal MI, Hassan A, Abdulkarim Ibrahim M, Salehuddin R, Abu Bakar NH. Pattern of QTc Prolongation in Methadone Maintenance Therapy (MMT) Subjects Receiving Different Methadone Dosages: A Prospective Cohort Study. Pakistan Journal of Medical Sciences. 2013;29(5):1132–6. pmid:24353706

* View Article

* PubMed/NCBI

* Google Scholar

5. 5. Mayet S, Gossop M, Lintzeris N, Markides V, Strang J. Methadone Maintenance, QTc and Torsade de Pointes: Who Needs an Electrocardiogram and What Is the Prevalence of QTc Prolongation? Drug and Alcohol Review. 2011;30(4):388–96.

* View Article

* Google Scholar

6. 6. Ehret GB, Voide C, Gex-Fabry M, Chabert J, Shah D, Broers B, et al. Drug-induced long QT syndrome in injection drug users receiving methadone: high frequency in hospitalized patients and risk factors. Arch Intern Med. 2006;166(12):1280–7. pmid:16801510

* View Article

* PubMed/NCBI

* Google Scholar

7. 7. Johansson I, Ingelman-Sundberg M. Genetic polymorphism and toxicology--with emphasis on cytochrome p450. Toxicol Sci. 2011;120(1):1–13. pmid:21149643

* View Article

* PubMed/NCBI

* Google Scholar

8. 8. Wilke RA, Lin DW, Roden DM, Watkins PB, Flockhart D, Zineh I, et al. Identifying genetic risk factors for serious adverse drug reactions: current progress and challenges. Nature Reviews Drug Discovery. 2007;6(11):904–16. Erratum in: Nature Reviews Drug Discovery. 2008;7(2):185.

* View Article

* Google Scholar

9. 9. Campuzano O, Sarquella-Brugada G, Brugada R, Brugada J. Genetics of channelopathies associated with sudden cardiac death. Glob Cardiol Sci Pract. 2015;2015(3):39. pmid:26566530

* View Article

* PubMed/NCBI

* Google Scholar

10. 10. Hasegawa K, Ohno S, Itoh H, Makiyama T, Aiba T, et al. A Rare KCNE1 Polymorphism, D85N, as a Genetic Modifier of Long QT Syndrome. Journal of Arrhythmia. 2014;30(3):161–6.

* View Article

* Google Scholar

11. 11. Zhang Y, Chang B, Hu S, Wang D, Fang Q, Huang X, et al. Single nucleotide polymorphisms and haplotype of four genes encoding cardiac ion channels in Chinese and their association with arrhythmia. Annals of Noninvasive Electrocardiology. 2008;13(2):180–90. pmid:18426444

* View Article

* PubMed/NCBI

* Google Scholar

12. 12. Roden DM. Long QT syndrome: reduced repolarization reserve and the genetic link. Journal of Internal Medicine. 2006;259(1):59–69. pmid:16336514

* View Article

* PubMed/NCBI

* Google Scholar

13. 13. Paulussen AD, Gilissen RA, Armstrong M, Doevendans PA, Verhasselt P, Smeets HJ, et al. Genetic variations of KCNQ1, KCNH2, SCN5A, KCNE1, and KCNE2 in drug-induced long QT syndrome patients. Journal of Molecular Medicine. 2004;82(3):182–8.

* View Article

* Google Scholar

14. 14. Zhang W, Zhou HH, Liu ZQ, Chen XP, He YJ. Pharmacogenomics in China. In: Padmanabhan S (Ed.). Handbook of pharmacogenomics and stratified medicine. London: Academic Press; 2014. p. 999–1013.

15. 15. Lazarczyk MJ, Bhuiyan ZA, Perrin N, Giannakopoulos P. Selective acquired long QT syndrome (saLQTS) upon risperidone treatment. BMC Psychiatry. 2012;12:220. pmid:23216910

* View Article

* PubMed/NCBI

* Google Scholar

16. 16. Hajj A, Ksouda K, Peoc’h K, Curis E, Messali A, Deveaux LL, et al. KCNH2 polymorphism and methadone dosage interact to enhance QT duration. Drug Alcohol Depend. 2014;141:34–8. pmid:24875677

* View Article

* PubMed/NCBI

* Google Scholar

17. 17. Koo SH, Ho WF, Lee EJ. Genetic polymorphisms in KCNQ1, HERG, KCNE1 and KCNE2 genes in the Chinese, Malay and Indian populations of Singapore. British Journal of Clinical Pharmacology. 2006;61(3):301–8.

* View Article

* Google Scholar

18. 18. Aydin A, Bähring S, Dahm S, Guenther UP, Uhlmann R, Busjahn A, et al. Single nucleotide polymorphism map of five long-QT genes. Journal of Molecular Medicine. 2005;83(2):159–65. pmid:15599693

* View Article

* PubMed/NCBI

* Google Scholar

19. 19. Mohamad N, Bakar NH, Musa N, Talib N, Ismail R. Better retention of Malaysian opiate dependents treated with high dose methadone in methadone maintenance therapy. Harm Reduct J. 2010;7:30. pmid:21167035

* View Article

* PubMed/NCBI

* Google Scholar

20. 20. Vandenbroucke JP, von Elm E, Altman DG, Gøtzsche PC, Mulrow CD, Pocock SJ, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. Ann Intern Med. 2007;147(8):W163–94. pmid:17938389

* View Article

* PubMed/NCBI

* Google Scholar

21. 21. Little J, Higgins JP, Ioannidis JP, Moher D, Gagnon F, von Elm E, et al. STrengthening the REporting of Genetic Association studies (STREGA): an extension of the STROBE Statement. Ann Intern Med. 2009;150(3):206–15. pmid:19189911

* View Article

* PubMed/NCBI

* Google Scholar

22. 22. Fosgate GT. Practical sample size calculations for surveillance and diagnostic investigations. Journal of Veterinary Diagnostic Investigation. 2009;21(1):3–14. pmid:19139495

* View Article

* PubMed/NCBI

* Google Scholar

23. 23. Mohamed Nazar NI. Therapeutic Drug Monitoring in Methadone Maintenance Therapy (MMT): An Evaluation of Genetic Factors Influencing Clinical Outcomes and Serum Concentrations of Methadone. Pusat Pengajian Sains Perubatan: Ph.D Thesis. (Universiti Sains Malaysia, Kelantan, 2013). [cited 21 June 2024. ]. Available from: http://eprints.usm.my/36393/

* View Article

* Google Scholar

24. 24. Wedam EF, Bigelow GE, Johnson RE, Nuzzo PA, Haigney MC. QT-interval effects of methadone, levomethadyl, and buprenorphine in a randomized trial. JAMA Internal Medicine. 2007;167(22):2469–75.

* View Article

* Google Scholar

25. 25. Vandenberk B, Vandael E, Robyns T, Vandenberghe J, Garweg C, Foulon V, et al. Which QT Correction Formulae to Use for QT Monitoring? Journal of the American Heart Association. 2016;5(6):e003264. Erratum in: Journal of the American Heart Association. 2018;7(16):e004252. https://doi.org/10.1161/JAHA.117.004252

* View Article

* Google Scholar

26. 26. Zahari Z, Lee CS, Ibrahim MA, Musa N, Mohd Yasin MA, Lee YY, et al. Relationship between ABCB1 polymorphisms and serum methadone concentration in patients undergoing methadone maintenance therapy (MMT). Am J Drug Alcohol Abuse. 2016;42(5):587–96. pmid:27284701

* View Article

* PubMed/NCBI

* Google Scholar

27. 27. Mohamed Nazar NI, Liang S, Tan S, Ismail R, Musa R, Abu Bakar AZ. Daily Clinical Dose (D) Poorly Predicts Trough Serum Methadone Concentration (Ctrough) in Patients Undergo Methadone Maintenance Therapy (MMT) with Good Adherence. International Journal of Pharmaceutical Science and Health Care. 2014;2:139–45.

* View Article

* Google Scholar

28. 28. Yang P, Kanki H, Drolet B, Yang T, Wei J, Viswanathan PC, et al. Allelic variants in long-QT disease genes in patients with drug-associated torsades de pointes. Circulation. 2002;105(16):1943–8. pmid:11997281

* View Article

* PubMed/NCBI

* Google Scholar

29. 29. Boratyn GM, Camacho C, Cooper PS, Coulouris G, Fong A, Ma N, et al. BLAST: a more efficient report with usability improvements. Nucleic Acids Res. 2013;41(Web Server issue):W29–33. pmid:23609542

* View Article

* PubMed/NCBI

* Google Scholar

30. 30. Graffelman J. The number of markers in the HapMap project: some notes on chi-square and exact tests for Hardy-Weinberg equilibrium. American Journal of Human Genetics. 2010;86(5):813–8.

* View Article

* Google Scholar

31. 31. Haldane JBS. An exact test for randomness of mating. Journal of Genetics. 1954;52:631–35.

* View Article

* Google Scholar

32. 32. Levene H. On A Matching Problem Arising in Genetics. Ann Math Statist. 1949;20:91–4.

* View Article

* Google Scholar

33. 33. Lindley D. Statistical Inference Concerning Hardy-Weinberg Equilibrium in Bayesian Statistics (eds. Bernardo JM, DeGroot MH, Lindley DV, Smith AFM). Oxford University Press; 1988. p. 307–26.

34. 34. Minelli C, Thompson JR, Abrams KR, Thakkinstian A, Attia J. How should we use information about HWE in the meta-analyses of genetic association studies? International Journal of Epidemiology. 2008;37(1):136–46.

* View Article

* Google Scholar

35. 35. Paysan-Lafosse T, Blum M, Chuguransky S, Grego T, Pinto BL, Salazar GA, et al. InterPro in 2022. Nucleic Acids Res. 2023;51(D1):D418–27. pmid:36350672

* View Article

* PubMed/NCBI

* Google Scholar

36. 36. Xie Y, Li H, Luo X, Li H, Gao Q, Zhang L, et al. IBS 2.0: an upgraded illustrator for the visualization of biological sequences. Nucleic Acids Res. 2022;50(W1):W420–426. pmid:35580044

* View Article

* PubMed/NCBI

* Google Scholar

37. 37. UniProt Consortium. UniProt: a worldwide hub of protein knowledge. Nucleic Acids Res. 2019;47(D1):D506–15. pmid:30395287

* View Article

* PubMed/NCBI

* Google Scholar

38. 38. Waterhouse A, Bertoni M, Bienert S, Studer G, Tauriello G, Gumienny R, et al. SWISS-MODEL: homology modelling of protein structures and complexes. Nucleic Acids Res. 2018;46(W1):W296–303. pmid:29788355

* View Article

* PubMed/NCBI

* Google Scholar

39. 39. Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, et al. UCSF Chimera--a visualization system for exploratory research and analysis. Journal of Computational Chemistry. 2004;25(13):1605–12. pmid:15264254

* View Article

* PubMed/NCBI

* Google Scholar

40. 40. Chen Y, Lu H, Zhang N, Zhu Z, Wang S, Li M. PremPS: Predicting the impact of missense mutations on protein stability. PLOS Computational Biology. 2020;16(12):e1008543.

* View Article

* Google Scholar

41. 41. Pires DE, Ascher DB, Blundell TL. DUET: a server for predicting effects of mutations on protein stability using an integrated computational approach. Nucleic Acids Research. 2014;42(Web Server issue):W314–9.

* View Article

* Google Scholar

42. 42. Liu Y, Grimm M, Dai W-T, Hou M-C, Xiao Z-X, Cao Y. CB-Dock: a web server for cavity detection-guided protein-ligand blind docking. Acta Pharmacol Sin. 2020;41(1):138–44. pmid:31263275

* View Article

* PubMed/NCBI

* Google Scholar

43. 43. Liu Y, Yang X, Gan J, Chen S, Xiao Z-X, Cao Y. CB-Dock2: improved protein-ligand blind docking by integrating cavity detection, docking and homologous template fitting. Nucleic Acids Res. 2022;50(W1):W159–64. pmid:35609983

* View Article

* PubMed/NCBI

* Google Scholar

44. 44. Kim S, Chen J, Cheng T, Gindulyte A, He J, He S, et al. PubChem 2019 update: improved access to chemical data. Nucleic Acids Res. 2019;47(D1):D1102–9. pmid:30371825

* View Article

* PubMed/NCBI

* Google Scholar

45. 45. O’Boyle NM, Banck M, James CA, Morley C, Vandermeersch T, Hutchison GR. Open Babel: An open chemical toolbox. J Cheminform. 2011;3:33. pmid:21982300

* View Article

* PubMed/NCBI

* Google Scholar

46. 46. Tian W, Chen C, Lei X, Zhao J, Liang J. CASTp 3.0: computed atlas of surface topography of proteins. Nucleic Acids Res. 2018;46(W1):W363–7. pmid:29860391

* View Article

* PubMed/NCBI

* Google Scholar

47. 47. International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH). The Clinical Evaluation of QT/QTc Interval Prolongation and Proarrhythmic Potential for Non-Antiarrhythmic Drugs (E14) [Internet]. 2005 May 12 [cited 2025 Feb 28. ]. Available from: https://database.ich.org/sites/default/files/E14_Guideline.pdf

* View Article

* Google Scholar

48. 48. Almon R, Engfeldt P, Tysk C, Sjöström M, Nilsson TK. Prevalence and trends in adult-type hypolactasia in different age cohorts in Central Sweden diagnosed by genotyping for the adult-type hypolactasia-linked LCT-13910C > T mutation. Scandinavian Journal of Gastroenterology. 2007;42(2):165–70.

* View Article

* Google Scholar

49. 49. Loo KW, Gan SH. Genetic, historical and linguistic perspectives on the origin of the Kelantanese Malays. Gene. 2014;545(1):1–4. pmid:24815047

* View Article

* PubMed/NCBI

* Google Scholar

50. 50. Zerdazi E-H, Vorspan F, Marees AT, Naccache F, Lepine J-P, Laplanche J-L, et al. QT length during methadone maintenance treatment: gene × dose interaction. Fundam Clin Pharmacol. 2019;33(1):96–106. pmid:30086202

* View Article

* PubMed/NCBI

* Google Scholar

51. 51. Gouas L, Nicaud V, Berthet M, Forhan A, Tiret L, Balkau B, et al. Association of KCNQ1, KCNE1, KCNH2 and SCN5A polymorphisms with QTc interval length in a healthy population. European Journal of Human Genetics. 2005;13(11):1213–22.

* View Article

* Google Scholar

52. 52. Bezzina CR, Verkerk AO, Busjahn A, Jeron A, Erdmann J, Koopmann TT, et al. A common polymorphism in KCNH2 (HERG) hastens cardiac repolarization. Cardiovasc Res. 2003;59(1):27–36. pmid:12829173

* View Article

* PubMed/NCBI

* Google Scholar

53. 53. Marjamaa A, Newton-Cheh C, Porthan K, Reunanen A, Lahermo P, Väänänen H, et al. Common candidate gene variants are associated with QT interval duration in the general population. J Intern Med. 2009;265(4):448–58. pmid:19019189

* View Article

* PubMed/NCBI

* Google Scholar

54. 54. Kimchi-Sarfaty C, Oh JM, Kim I-W, Sauna ZE, Calcagno AM, Ambudkar SV, et al. A “silent” polymorphism in the MDR1 gene changes substrate specificity. Science. 2007;315(5811):525–8. pmid:17185560

* View Article

* PubMed/NCBI

* Google Scholar

55. 55. Bertalovitz AC, Badhey MLO, McDonald TV. Synonymous nucleotide modification of the KCNH2 gene affects both mRNA characteristics and translation of the encoded hERG ion channel. Journal of Biological Chemistry. 2018;293(31):12120–36. pmid:29907571

* View Article

* PubMed/NCBI

* Google Scholar

56. 56. Wang QS, Wang XF, Chen XD, Yu JF, Wang J, Sun J, et al. Genetic polymorphism of KCNH2 confers predisposition of acquired atrial fibrillation in Chinese. Journal of Cardiovascular Electrophysiology. 2009;20(10):1158–62.

* View Article

* Google Scholar

57. 57. Patil R, Das S, Stanley A, Yadav L, Sudhakar A, Varma AK. Optimized hydrophobic interactions and hydrogen bonding at the target-ligand interface leads the pathways of drug-designing. PLoS One. 2010;5(8):e12029. pmid:20808434

* View Article

* PubMed/NCBI

* Google Scholar

58. 58. Yang L, Li J, Guo S, Hou C, Liao C, Shi L, et al. SARS-CoV-2 Variants, RBD Mutations, Binding Affinity, and Antibody Escape. International Journal of Molecular Sciences. 2021;22(22):12114.

* View Article

* Google Scholar

59. 59. Schaeffer L. The Role of Functional Groups in Drug–Receptor Interactions in The Practice of Medicinal Chemistry (eds. Wermuth CG, Raboisson P, Aldous D, Rognan D). Elsevier; 2008. p. 359–78.

60. 60. Zhuang W-R, Wang Y, Cui P-F, Xing L, Lee J, Kim D, et al. Applications of π-π stacking interactions in the design of drug-delivery systems. Journal of Controlled Release. 2019;294:311–26. pmid:30550939

* View Article

* PubMed/NCBI

* Google Scholar

61. 61. Stornaiuolo M, De Kloe GE, Rucktooa P, Fish A, van Elk R, Edink ES, et al. Assembly of a π-π stack of ligands in the binding site of an acetylcholine-binding protein. Nature Communications. 2013;4:1875. pmid:23695669

* View Article

* PubMed/NCBI

* Google Scholar

62. 62. Lanzarotti E, Defelipe LA, Marti MA, Turjanski AG. Aromatic clusters in protein-protein and protein-drug complexes. J Cheminform. 2020;12(1):30. pmid:33431014

* View Article

* PubMed/NCBI

* Google Scholar

63. 63. Titus-Lay EN, Jaynes HA, Tomaselli Muensterman E, Walroth TA, Ott CA, Desta Z, et al. Methadone-associated QT interval prolongation in patients undergoing maintenance therapy in an urban opioid treatment program. Pharmacotherapy. 2021;41(3):238–46.

* View Article

* Google Scholar

64. 64. Katchman AN, McGroary KA, Kilborn MJ, Kornick CA, Manfredi PL, Woosley RL, et al. Influence of opioid agonists on cardiac human ether-a-go-go-related gene K(+) currents. Journal of Pharmacology and Experimental Therapeutics. 2002;303(2):688–94.

* View Article

* Google Scholar

65. 65. Manan MM, Ali SM, Khan MA, Jafarian S, Hameed MA. Review on the demographic and social impact of methadone-medication therapy on Malaysian patients. Pakistan Journal of Pharmaceutical Sciences. 2013;26(4):841–6.

* View Article

* Google Scholar

66. 66. Shi M, Bouwmeester H, Rietjens IMCM, Strikwold M. Integrating in vitro data and physiologically based kinetic modeling-facilitated reverse dosimetry to predict human cardiotoxicity of methadone. Archives of Toxicology. 2020;94(8):2809–27. pmid:32367273

* View Article

* PubMed/NCBI

* Google Scholar

67. 67. Ismail R, Shafurdin NS, Shukor MS, Mohammed Nawi A, Abdul Manaf MR, Ibrahim N, et al. Predictors of drug and substance abuse among school-going adolescents living in drug hotspot in Malaysia. PLoS One. 2024;19(6):e0305460. pmid:38875293

* View Article

* PubMed/NCBI

* Google Scholar

Citation: Abdul Jalal MI, Abdullah-Zawawi M-R, Musa N, Ghazali B, Zahari Z, Mohamad N (2025) A synonymous KCNH2 polymorphism and methadone trough level influence QTc prolongation in Kelantanese Malay recipients of methadone maintenance therapy (MMT) in Malaysia. PLoS One 20(5): e0322724. https://doi.org/10.1371/journal.pone.0322724

About the Authors:

Muhammad Irfan Abdul Jalal

Roles: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Supervision, Visualization, Writing – original draft, Writing – review & editing

E-mail: [email protected]

Affiliation: UKM Medical Molecular Biology Institute (UMBI), Cheras, Kuala Lumpur, Malaysia

ORICD: https://orcid.org/0000-0001-6911-3335

Muhammad-Redha Abdullah-Zawawi

Roles: Formal analysis, Methodology, Visualization, Writing – original draft, Writing – review & editing

Affiliation: UKM Medical Molecular Biology Institute (UMBI), Cheras, Kuala Lumpur, Malaysia

ORICD: https://orcid.org/0000-0002-4404-6014

Nurfadhlina Musa

Roles: Formal analysis, Investigation, Validation, Writing – original draft

Affiliation: Human Genome Centre, School of Medical Sciences, Universiti Sains Malaysia, Kubang Kerian, Malaysia

ORICD: https://orcid.org/0000-0002-7533-8474

Basyirah Ghazali

Roles: Formal analysis, Methodology

Affiliation: Institute for Research in Molecular Medicine (INFORMM), Universiti Sains Malaysia, Kubang Kerian, Kelantan, Malaysia

ORICD: https://orcid.org/0000-0002-8129-7189

Zalina Zahari

Roles: Writing – original draft, Writing – review & editing

Affiliation: Faculty of Pharmacy, Universiti Sultan Zainal Abidin, Besut Campus, Besut, Terengganu, Malaysia

ORICD: https://orcid.org/0000-0003-1459-8958

Nasir Mohamad

Roles: Supervision, Writing – original draft, Writing – review & editing

Affiliation: Faculty of Medicine, Sultan Zainal Abidin University, Medical Campus, Kuala Terengganu, Terengganu, Malaysia

ORICD: https://orcid.org/0000-0002-4370-1187

[/RAW_REF_TEXT]

[/RAW_REF_TEXT]

[/RAW_REF_TEXT]

References

1. Reid G, Kamarulzaman A, Sran SK. Malaysia and harm reduction: the challenges and responses. International Journal of Drug Policy. 2007;18(2):136–40. pmid:17689356

2. Ministry of Health Malaysia. National Strategic Plan for Ending AIDS 2016-2030. (Ministry of Health Malaysia, Putrajaya, 2015). [cited 21 June 2024. ]. Available from: https://www.moh.gov.my/moh/resources/auto%20download%20images/58510b70ca319.pdf

3. Alinejad S, Kazemi T, Zamani N, Hoffman RS, Mehrpour O. A Systematic Review of the Cardiotoxicity of Methadone. EXCLI Journal. 2015;14:577–600. pmid:26869865

4. Mohamad N, Abdul Jalal MI, Hassan A, Abdulkarim Ibrahim M, Salehuddin R, Abu Bakar NH. Pattern of QTc Prolongation in Methadone Maintenance Therapy (MMT) Subjects Receiving Different Methadone Dosages: A Prospective Cohort Study. Pakistan Journal of Medical Sciences. 2013;29(5):1132–6. pmid:24353706

5. Mayet S, Gossop M, Lintzeris N, Markides V, Strang J. Methadone Maintenance, QTc and Torsade de Pointes: Who Needs an Electrocardiogram and What Is the Prevalence of QTc Prolongation? Drug and Alcohol Review. 2011;30(4):388–96.

6. Ehret GB, Voide C, Gex-Fabry M, Chabert J, Shah D, Broers B, et al. Drug-induced long QT syndrome in injection drug users receiving methadone: high frequency in hospitalized patients and risk factors. Arch Intern Med. 2006;166(12):1280–7. pmid:16801510

7. Johansson I, Ingelman-Sundberg M. Genetic polymorphism and toxicology--with emphasis on cytochrome p450. Toxicol Sci. 2011;120(1):1–13. pmid:21149643

8. Wilke RA, Lin DW, Roden DM, Watkins PB, Flockhart D, Zineh I, et al. Identifying genetic risk factors for serious adverse drug reactions: current progress and challenges. Nature Reviews Drug Discovery. 2007;6(11):904–16. Erratum in: Nature Reviews Drug Discovery. 2008;7(2):185.

9. Campuzano O, Sarquella-Brugada G, Brugada R, Brugada J. Genetics of channelopathies associated with sudden cardiac death. Glob Cardiol Sci Pract. 2015;2015(3):39. pmid:26566530

10. Hasegawa K, Ohno S, Itoh H, Makiyama T, Aiba T, et al. A Rare KCNE1 Polymorphism, D85N, as a Genetic Modifier of Long QT Syndrome. Journal of Arrhythmia. 2014;30(3):161–6.

11. Zhang Y, Chang B, Hu S, Wang D, Fang Q, Huang X, et al. Single nucleotide polymorphisms and haplotype of four genes encoding cardiac ion channels in Chinese and their association with arrhythmia. Annals of Noninvasive Electrocardiology. 2008;13(2):180–90. pmid:18426444

12. Roden DM. Long QT syndrome: reduced repolarization reserve and the genetic link. Journal of Internal Medicine. 2006;259(1):59–69. pmid:16336514

13. Paulussen AD, Gilissen RA, Armstrong M, Doevendans PA, Verhasselt P, Smeets HJ, et al. Genetic variations of KCNQ1, KCNH2, SCN5A, KCNE1, and KCNE2 in drug-induced long QT syndrome patients. Journal of Molecular Medicine. 2004;82(3):182–8.

14. Zhang W, Zhou HH, Liu ZQ, Chen XP, He YJ. Pharmacogenomics in China. In: Padmanabhan S (Ed.). Handbook of pharmacogenomics and stratified medicine. London: Academic Press; 2014. p. 999–1013.

15. Lazarczyk MJ, Bhuiyan ZA, Perrin N, Giannakopoulos P. Selective acquired long QT syndrome (saLQTS) upon risperidone treatment. BMC Psychiatry. 2012;12:220. pmid:23216910

16. Hajj A, Ksouda K, Peoc’h K, Curis E, Messali A, Deveaux LL, et al. KCNH2 polymorphism and methadone dosage interact to enhance QT duration. Drug Alcohol Depend. 2014;141:34–8. pmid:24875677

17. Koo SH, Ho WF, Lee EJ. Genetic polymorphisms in KCNQ1, HERG, KCNE1 and KCNE2 genes in the Chinese, Malay and Indian populations of Singapore. British Journal of Clinical Pharmacology. 2006;61(3):301–8.

18. Aydin A, Bähring S, Dahm S, Guenther UP, Uhlmann R, Busjahn A, et al. Single nucleotide polymorphism map of five long-QT genes. Journal of Molecular Medicine. 2005;83(2):159–65. pmid:15599693

19. Mohamad N, Bakar NH, Musa N, Talib N, Ismail R. Better retention of Malaysian opiate dependents treated with high dose methadone in methadone maintenance therapy. Harm Reduct J. 2010;7:30. pmid:21167035

20. Vandenbroucke JP, von Elm E, Altman DG, Gøtzsche PC, Mulrow CD, Pocock SJ, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. Ann Intern Med. 2007;147(8):W163–94. pmid:17938389

21. Little J, Higgins JP, Ioannidis JP, Moher D, Gagnon F, von Elm E, et al. STrengthening the REporting of Genetic Association studies (STREGA): an extension of the STROBE Statement. Ann Intern Med. 2009;150(3):206–15. pmid:19189911

22. Fosgate GT. Practical sample size calculations for surveillance and diagnostic investigations. Journal of Veterinary Diagnostic Investigation. 2009;21(1):3–14. pmid:19139495

23. Mohamed Nazar NI. Therapeutic Drug Monitoring in Methadone Maintenance Therapy (MMT): An Evaluation of Genetic Factors Influencing Clinical Outcomes and Serum Concentrations of Methadone. Pusat Pengajian Sains Perubatan: Ph.D Thesis. (Universiti Sains Malaysia, Kelantan, 2013). [cited 21 June 2024. ]. Available from: http://eprints.usm.my/36393/

24. Wedam EF, Bigelow GE, Johnson RE, Nuzzo PA, Haigney MC. QT-interval effects of methadone, levomethadyl, and buprenorphine in a randomized trial. JAMA Internal Medicine. 2007;167(22):2469–75.

25. Vandenberk B, Vandael E, Robyns T, Vandenberghe J, Garweg C, Foulon V, et al. Which QT Correction Formulae to Use for QT Monitoring? Journal of the American Heart Association. 2016;5(6):e003264. Erratum in: Journal of the American Heart Association. 2018;7(16):e004252. https://doi.org/10.1161/JAHA.117.004252

26. Zahari Z, Lee CS, Ibrahim MA, Musa N, Mohd Yasin MA, Lee YY, et al. Relationship between ABCB1 polymorphisms and serum methadone concentration in patients undergoing methadone maintenance therapy (MMT). Am J Drug Alcohol Abuse. 2016;42(5):587–96. pmid:27284701

27. Mohamed Nazar NI, Liang S, Tan S, Ismail R, Musa R, Abu Bakar AZ. Daily Clinical Dose (D) Poorly Predicts Trough Serum Methadone Concentration (Ctrough) in Patients Undergo Methadone Maintenance Therapy (MMT) with Good Adherence. International Journal of Pharmaceutical Science and Health Care. 2014;2:139–45.

28. Yang P, Kanki H, Drolet B, Yang T, Wei J, Viswanathan PC, et al. Allelic variants in long-QT disease genes in patients with drug-associated torsades de pointes. Circulation. 2002;105(16):1943–8. pmid:11997281

29. Boratyn GM, Camacho C, Cooper PS, Coulouris G, Fong A, Ma N, et al. BLAST: a more efficient report with usability improvements. Nucleic Acids Res. 2013;41(Web Server issue):W29–33. pmid:23609542

30. Graffelman J. The number of markers in the HapMap project: some notes on chi-square and exact tests for Hardy-Weinberg equilibrium. American Journal of Human Genetics. 2010;86(5):813–8.

31. Haldane JBS. An exact test for randomness of mating. Journal of Genetics. 1954;52:631–35.

32. Levene H. On A Matching Problem Arising in Genetics. Ann Math Statist. 1949;20:91–4.

33. Lindley D. Statistical Inference Concerning Hardy-Weinberg Equilibrium in Bayesian Statistics (eds. Bernardo JM, DeGroot MH, Lindley DV, Smith AFM). Oxford University Press; 1988. p. 307–26.

34. Minelli C, Thompson JR, Abrams KR, Thakkinstian A, Attia J. How should we use information about HWE in the meta-analyses of genetic association studies? International Journal of Epidemiology. 2008;37(1):136–46.

35. Paysan-Lafosse T, Blum M, Chuguransky S, Grego T, Pinto BL, Salazar GA, et al. InterPro in 2022. Nucleic Acids Res. 2023;51(D1):D418–27. pmid:36350672

36. Xie Y, Li H, Luo X, Li H, Gao Q, Zhang L, et al. IBS 2.0: an upgraded illustrator for the visualization of biological sequences. Nucleic Acids Res. 2022;50(W1):W420–426. pmid:35580044

37. UniProt Consortium. UniProt: a worldwide hub of protein knowledge. Nucleic Acids Res. 2019;47(D1):D506–15. pmid:30395287

38. Waterhouse A, Bertoni M, Bienert S, Studer G, Tauriello G, Gumienny R, et al. SWISS-MODEL: homology modelling of protein structures and complexes. Nucleic Acids Res. 2018;46(W1):W296–303. pmid:29788355

39. Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, et al. UCSF Chimera--a visualization system for exploratory research and analysis. Journal of Computational Chemistry. 2004;25(13):1605–12. pmid:15264254

40. Chen Y, Lu H, Zhang N, Zhu Z, Wang S, Li M. PremPS: Predicting the impact of missense mutations on protein stability. PLOS Computational Biology. 2020;16(12):e1008543.

41. Pires DE, Ascher DB, Blundell TL. DUET: a server for predicting effects of mutations on protein stability using an integrated computational approach. Nucleic Acids Research. 2014;42(Web Server issue):W314–9.

42. Liu Y, Grimm M, Dai W-T, Hou M-C, Xiao Z-X, Cao Y. CB-Dock: a web server for cavity detection-guided protein-ligand blind docking. Acta Pharmacol Sin. 2020;41(1):138–44. pmid:31263275

43. Liu Y, Yang X, Gan J, Chen S, Xiao Z-X, Cao Y. CB-Dock2: improved protein-ligand blind docking by integrating cavity detection, docking and homologous template fitting. Nucleic Acids Res. 2022;50(W1):W159–64. pmid:35609983

44. Kim S, Chen J, Cheng T, Gindulyte A, He J, He S, et al. PubChem 2019 update: improved access to chemical data. Nucleic Acids Res. 2019;47(D1):D1102–9. pmid:30371825

45. O’Boyle NM, Banck M, James CA, Morley C, Vandermeersch T, Hutchison GR. Open Babel: An open chemical toolbox. J Cheminform. 2011;3:33. pmid:21982300

46. Tian W, Chen C, Lei X, Zhao J, Liang J. CASTp 3.0: computed atlas of surface topography of proteins. Nucleic Acids Res. 2018;46(W1):W363–7. pmid:29860391

47. International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH). The Clinical Evaluation of QT/QTc Interval Prolongation and Proarrhythmic Potential for Non-Antiarrhythmic Drugs (E14) [Internet]. 2005 May 12 [cited 2025 Feb 28. ]. Available from: https://database.ich.org/sites/default/files/E14_Guideline.pdf

48. Almon R, Engfeldt P, Tysk C, Sjöström M, Nilsson TK. Prevalence and trends in adult-type hypolactasia in different age cohorts in Central Sweden diagnosed by genotyping for the adult-type hypolactasia-linked LCT-13910C > T mutation. Scandinavian Journal of Gastroenterology. 2007;42(2):165–70.

49. Loo KW, Gan SH. Genetic, historical and linguistic perspectives on the origin of the Kelantanese Malays. Gene. 2014;545(1):1–4. pmid:24815047

50. Zerdazi E-H, Vorspan F, Marees AT, Naccache F, Lepine J-P, Laplanche J-L, et al. QT length during methadone maintenance treatment: gene × dose interaction. Fundam Clin Pharmacol. 2019;33(1):96–106. pmid:30086202

51. Gouas L, Nicaud V, Berthet M, Forhan A, Tiret L, Balkau B, et al. Association of KCNQ1, KCNE1, KCNH2 and SCN5A polymorphisms with QTc interval length in a healthy population. European Journal of Human Genetics. 2005;13(11):1213–22.

52. Bezzina CR, Verkerk AO, Busjahn A, Jeron A, Erdmann J, Koopmann TT, et al. A common polymorphism in KCNH2 (HERG) hastens cardiac repolarization. Cardiovasc Res. 2003;59(1):27–36. pmid:12829173

53. Marjamaa A, Newton-Cheh C, Porthan K, Reunanen A, Lahermo P, Väänänen H, et al. Common candidate gene variants are associated with QT interval duration in the general population. J Intern Med. 2009;265(4):448–58. pmid:19019189

54. Kimchi-Sarfaty C, Oh JM, Kim I-W, Sauna ZE, Calcagno AM, Ambudkar SV, et al. A “silent” polymorphism in the MDR1 gene changes substrate specificity. Science. 2007;315(5811):525–8. pmid:17185560

55. Bertalovitz AC, Badhey MLO, McDonald TV. Synonymous nucleotide modification of the KCNH2 gene affects both mRNA characteristics and translation of the encoded hERG ion channel. Journal of Biological Chemistry. 2018;293(31):12120–36. pmid:29907571

56. Wang QS, Wang XF, Chen XD, Yu JF, Wang J, Sun J, et al. Genetic polymorphism of KCNH2 confers predisposition of acquired atrial fibrillation in Chinese. Journal of Cardiovascular Electrophysiology. 2009;20(10):1158–62.

57. Patil R, Das S, Stanley A, Yadav L, Sudhakar A, Varma AK. Optimized hydrophobic interactions and hydrogen bonding at the target-ligand interface leads the pathways of drug-designing. PLoS One. 2010;5(8):e12029. pmid:20808434

58. Yang L, Li J, Guo S, Hou C, Liao C, Shi L, et al. SARS-CoV-2 Variants, RBD Mutations, Binding Affinity, and Antibody Escape. International Journal of Molecular Sciences. 2021;22(22):12114.

59. Schaeffer L. The Role of Functional Groups in Drug–Receptor Interactions in The Practice of Medicinal Chemistry (eds. Wermuth CG, Raboisson P, Aldous D, Rognan D). Elsevier; 2008. p. 359–78.

60. Zhuang W-R, Wang Y, Cui P-F, Xing L, Lee J, Kim D, et al. Applications of π-π stacking interactions in the design of drug-delivery systems. Journal of Controlled Release. 2019;294:311–26. pmid:30550939

61. Stornaiuolo M, De Kloe GE, Rucktooa P, Fish A, van Elk R, Edink ES, et al. Assembly of a π-π stack of ligands in the binding site of an acetylcholine-binding protein. Nature Communications. 2013;4:1875. pmid:23695669

62. Lanzarotti E, Defelipe LA, Marti MA, Turjanski AG. Aromatic clusters in protein-protein and protein-drug complexes. J Cheminform. 2020;12(1):30. pmid:33431014

63. Titus-Lay EN, Jaynes HA, Tomaselli Muensterman E, Walroth TA, Ott CA, Desta Z, et al. Methadone-associated QT interval prolongation in patients undergoing maintenance therapy in an urban opioid treatment program. Pharmacotherapy. 2021;41(3):238–46.

64. Katchman AN, McGroary KA, Kilborn MJ, Kornick CA, Manfredi PL, Woosley RL, et al. Influence of opioid agonists on cardiac human ether-a-go-go-related gene K(+) currents. Journal of Pharmacology and Experimental Therapeutics. 2002;303(2):688–94.

65. Manan MM, Ali SM, Khan MA, Jafarian S, Hameed MA. Review on the demographic and social impact of methadone-medication therapy on Malaysian patients. Pakistan Journal of Pharmaceutical Sciences. 2013;26(4):841–6.

66. Shi M, Bouwmeester H, Rietjens IMCM, Strikwold M. Integrating in vitro data and physiologically based kinetic modeling-facilitated reverse dosimetry to predict human cardiotoxicity of methadone. Archives of Toxicology. 2020;94(8):2809–27. pmid:32367273

67. Ismail R, Shafurdin NS, Shukor MS, Mohammed Nawi A, Abdul Manaf MR, Ibrahim N, et al. Predictors of drug and substance abuse among school-going adolescents living in drug hotspot in Malaysia. PLoS One. 2024;19(6):e0305460. pmid:38875293