Research

2. Understanding lung Cancer Susceptibility, Prognosis and Metastasis

Lung cancer is found to be the most prevalent malignancy with the highest mortality rates worldwide (Bray, Ferlay et al. 2018). It is the fourth most prevalent malignancy ranked at the third position in causing cancer-related mortalities in India (Bray, Ferlayetal. 2018). Environmental and lifestyle changes have contributed a lot to the increase in the incidences of lung cancer in the world, including India, which has become a serious concern to the clinicians and healthcare providers considering theprognosis and survival rate of lung cancer is very poor. Lung carcinogenesis is amultistep process with a series of progressive pathological changes, such as pre-neoplastic changes initiated by hyperplasia (Choy, Pass et al. 2006). The conventional methods of treatment for lung cancer have not been entirely successful in increasing the survival of the patients.Therefore, more precise and targeted approaches such as chemoprevention (Sabaand Khuri 2005) and precision genomics (Ahmad and Gadgeel 2016, Dammacco and Silvestris 2019, Von Hoff andHan 2019) are necessary for effective treatment of the patients to ensure better survival. The identification of molecular targets for early detection, treatment, and management of lung cancer is essential in this regard. I am working in a three-pronged approach to understand the molecular genetic basis of lung cancer susceptibility, progression and pathogenesis: 

(a) Understanding the inherited susceptibility among cigarette smokers

It has been reported that about 15-20% of tobacco smokers develop lung cancer while the rest somehow evade the disease (Bach, Kattan et al. 2003, Wang, Chorleyet al. 2010), which indicates an operant genetic predisposition to lung cancer. Thus it is imperative to identify the precise genetic markers contributing to this differential susceptibility to lung cancer among individuals and across populations, which would be essential for better understanding, prediction, and management of lung cancer through personalised treatment and precision medicine. Lung carcinogenesis involves aberrant expression of xenobiotic metabolism and DNA repair genes as critical modulators of differential susceptibility (Spira, Beane etal. 2004, Baglole, Sime et al. 2008, Steiling, Kadar et al. 2009, Van Dyck, Nazarov etal. 2014) among individuals. Such differential expression of genes could be attributed to variants in the cis-regulatory regions regulating the transcription of the respective genes (Spira, Beaneetal. 2004, Baglole, Simeetal.2008, Steiling, Kadaretal.2009, VanDyck, Nazarovet al.2014). My team in the Department of Genetics, University of Calcutta has been engaged in identification, prioritisationand validation of variants as cis-regulatory markers regulating the differential expression of the candidate xenobiotic metabolism and DNA repair genes conferring risk to lung cancer in smokers. To-date, we have been able to identify 2984 variants within 531 gene-expression correlated DNaseI hyper sensitive sites (DHS) associated with 23 DNA repair and 25 xenobiotic metabolizing genes that have the potential of acting as cis-regulatory markers for lung cancer. The 2984 variants were prioritized further based ontheir transcription factor binding site altering potential, and being in cis-eQTL in lung tissue.The exercise identified 22 regulatory SNVs (rSNVs) for 7 xenobiotic metabolismand 7 DNA repair genes through a pipeline developed in our laboratory. To our hypothesis, the presence of these SNVs would confer an inherent risk or protection towrdas development of lung cancer in the smokers. Case-control analysis in lung cancer smokers and non lung cancer smokers has so fa rrevealed rs3764821 of ALDH3B1 and rs3748523 of RAD52 to be strongly associated associated with cigarette smoke induced lung cancer in the Indian population, as also with lung cancer modified by pack-years of smoking, tobacco and betel quid chewing. Interestingly, the rSNVs were also found to be associated with different histological subtypes and TNM stages that provide an insight into personalised therapy with the application of precision medicine in a population-specific manner. Depending upon the frequency of the risk genotypes present, we have assigned unweighted risk score to the 1000 genome populations. Incorporating more genes of critical lung cancer regulatory pathways would enable us to construct a comprehensive, personalized genomic map of individuals across different populations for assessing their lung cancer risk profiles to design personalized therapy based on precision medicine and formulating of effective tobacco control policies and genetic counselling for the containment of the disease.

In this respect we are also looking into the association of mitochondrial haplogroups and independent SNPs, if any, with lung cancer pathogenesis in the Indian context.

(b) Identifying the molecular basis of long non-coding RNA-mediated regulation of ROS metabolism, cytoprotection and induction of stemness in lung cancer

It has been well-established that the progression of lung cancer depends on the balance of ROS generation and adequate clearance of ROS by phase I and phase II detoxification pathway. When produced in excess amount, ROS results in damage to DNA triggering lung carcinogenesis. It has been suggested that lung cancer cells hijack the ROS metabolism pathway, which clears out intracellular ROS effectively and enhances the survival of the lung tumour cells (Thai, Statt et al. 2013). A lncRNA, SCAL1 was found to play a critical role inNRF2/KEAP1-mediated detoxification and ROS clearance conferring cytoprotection to the lung cancer cells (Thai, Statt et al. 2013). This lncRNA, SCAL1 has also been found to induce cell proliferation by epigenetic silencing of p21 and p57 in NSCLC cells (Sun, Jin et al. 2017)). Therefore, it is evident that the lncRNA SCAL1 plays a critical regulatory role in tumour cell cytoprotection and growth in lung cancer pathogenesis. Cytoprotection and tumour growth is the principal characteristic of the small population of cancer stem cells in the tumour bulk. Thus, we are trying to elucidate of the role of SCAL1 in the induction and maintenance of cancer stem cells leading to tumorigenic transformation in normal human bronchial epithelial (BEAS-2B) cells under cigarette smoke exposure as an aqueous extract. Real-time quantitative PCR analysis revealed the lncRNA to be significantly overexpressed in BEAS-2B cells under the exposure to 20% cigarette smoke extract (CSE).CSE exposure was also found to induce the upregulation of cancer stem cell markers of lung cancer, such as CD44, CD133, as revealed by western blot analysis. Interestingly, analysis by fluorescence spectrophotometer has showed that 20% CSE exposure to SCAL1-expressing BEAS-2B cells for higher incubation time resulted in a significant decrease in intracellular ROS compared to lower incubation time. The cell survival was also found to be significantly enhanced at higher incubation time. Further, SCAL1 was also found to play a role in P53/TCTP-mediated tumorigenic transformation of non-tumorigenic BEAS-2B cells, by upregulating CTP and downregulating P53 expression, as revealed by western blot analysis siRNA-mediated silencing of SCAL1 resulted in the significant downregulation of cancer stem cell markers, accumulation of intracellular ROS even at higher incubation time, decrease in cell survival even at higher incubation time, upregulation of P53 and low expression of TCTP, which is indicative of inhibition of tumorigenic transformation of BEAS-2B cells by 20% CSE exposure. We have been able to establish the reversible expressional regulation of TCTP and P53, as reported in the study by Amson et al. (Amson, Pece et al. 2011). Thus, the data indicate a cytoprotective role of SCAL1 along with its ability to induce and maintain the expression of cancer stem cell markers paving the way for the tumorigenic transformation of non-cancerous healthy bronchial epithelial cells (BEAS-2B) under 20% CSE exposure. The findings are suggestive of SCAL1 to be a prognostic biomarker and a critical pharmacologic target for effective treatment and management of lung cancer.

(c) Identifying and validating mutations responsible for alteration in the molecular docking dynamics of Robo/Slit

Cancer morbidity occurs mainly due to tumour invasion and metastases to secondary tissues causing great distress in the treatment and management of the disease. The continually evolving nature of the cancer cells compels the researchers to look for new molecular targets for the containment of the disease. Subsequent research following the discovery of SLIT/ROBO indicated that the pathway plays vital roles outside in the modulation of chemokine activation and cellular migration of different lineages (Wu, Feng et al. 2001, Prasad, Qamri et al. 2007, Geutskens, Hordijk et al.2010,Brantley-Sieders, Dunawayetal.2011), morphogenesis (Stella, Trusolino et al. 2009), leukocyte chemotaxis (Geutskens, Hordijk et al. 2010), metastasis (Kim, Zhang et al. 2008, Kong, Yi et al. 2015), and angiogenesis (Marlow, Binnewies et al.2010, Koch, Mathivet et al. 2011). The recognition of SLIT2 by ROBO1 is mediated through proper interaction of the respective domains of the proteins, such as SLIT2.D2 (Seiradake, vonPhilipsborn et al.2009) and ROBO1.Ig1domains facilitated by heparin sulfate units (Morlot, Thielens et al.2007, Fukuhara, Howittetal. 2008, Hohenester 2008). In lung cancer, proper interaction of ROBO.IG1 and SLIT2.D2 induce a tumour suppressive effect by inhibiting metastasis. Though less in formation regarding Robo4 and Slit2 interaction is available, the anti-angiogenic effect of Robo4 is reported (Park, Morrison et al. 2003, Koch, Mathivet et al. 2011).Thus, my team is also engaged in the characterization of the missense mutations of ROBO1, ROBO4, and SLIT2 to identify their role in the docking dynamics between the ligand SLIT2 to its cognate receptors ROBO1 and ROBO4.Therefore,following an in silico mutation prioritization through a sequence and structure-based analyses using the freely available web servers, the missense mutations were prioritized and we are assessing the role of the prioritised mutations on the disruption of molecular docking between Robo1, Robo4 and Slit2 homology models. Domain-domain molecular docking was done with the prioritised mutations localized within the interacting domains of the respective proteins. The whole protein-protein docking is also being done with the prioritised mutations localised outside the interacting domains. A weighted scoring system has been developed in our laboratory to grade the mutations as docking disruptors, non-disruptors, and neutral mutations based on particular parameters compared between the wild type: wild type docked complex and the mutant: wild typed ocked complexes, like change in free energy, change in binding affinity, change inthe gain in solvation energy, change in bond numbers, and change in RMSD value from the mean RMSD. The analysis revealed 2 mutations of Robo1 and 14 mutations of Slit2 to be high disruptors while 6 mutations of Robo1 and 4 mutations of Slit2 to be low disruptors for the docking of Robo1.Ig1 domain with the Slit2.D2 domain. Similarly, 1 mutation of Robo4 and 11 mutations of Slit2 were found to be high disruptors while 3 mutations of Robo4 and 7 mutations of Slit2 to be low disruptors for the docking of Robo4.Ig1-2 domain with the Slit2.D2 domain.The study helps to identify the potential molecular candidates of metastasis and angiogenesis in lung cancer mediated by the Robo/Slit pathway. To our knowledge, this study is the first of its kind on the structural dynamics of Robo4.Ig1-2 and Slit2.D2 interaction in a computational model. The molecular dynamics of Robo4.Ig1-2 interaction with Slit2.D2 could provide evidence of Robo4 mediated angiogenesis in lung cancer pathogenesis.Thus, the damaging mutations identified could serve as potential candidates of Robo/Slit-mediated metastasis and angiogenesis in lung cancer. Now, we would like to extend the study to perform combinatorial docking of the highest and lowest grade score mutations to assess the combinatorial effect of mutations, simultaneously on both receptor and ligand interaction dynamics. The conformational changes stated above would then be correlated to the in vitro biochemical properties of the mutants on protein-protein interaction and subsequent alteration in cell migration and angiogenic potential in wet lab for the validation of our computational prediction in real-time bio systems. These experiments will reveal why and how even a single residue alteration could lead to disruption of native dynamics of mutants leading to invasive and angiogenic phenotype of lung cancer.

(d) Meta-Analysis of genetic association studies pertaining to Lung Cancer

Reports of genetic association of polymorphisms with lung cancer in the Indian subcontinent are often conflicting. To summarise and replicate published evidence for association with lung cancer and its subgroups. We performed a meta-analysis of candidate associations on lung cancer, its histological subtypes and smoking status in the Indian subcontinent following PRISMA guidelines. Multiple testing corrections were done by the Benjamini-Hochberg method through assessment of significance at a false discovery rate of 10%. We genotyped and investigated rs1048943/CYP1A1 in a case-control sample from eastern India, followed by its global meta-analysis using a similar protocol. Meta-analysis of 18 variants of 11 genes reported in 39 studies (7,630 cases and 8,169 controls) showed significant association of rs1048943/CYP1A1 [2.07(1.49-2.87)] and rs4646903/CYP1A1 [1.48(1.93-1.95)] with overall lung cancer risk at 10% FDR, while nominal association (p < 0.05) was observed for del1/GSTT1, del2/GSTM1, rs1695/GSTP1 and rs17037102/ DKK2. Subtype analysis showed a significant association of del1/GSTT1 with adenocarcinoma, rs4646903/CYP1A1 with squamous carcinoma, and rs1048943/CYP1A1 with both. Association of rs4646903/CYP1A1 in smokers and effect modification by meta-regression analysis was observed. Genotyping of rs1048943/CYP1A1 that presented significant heterogeneity (p < 0.1) revealed an association with adenocarcinoma among eastern Indian smokers, while a global meta-analysis in 10458 cases and 10871controls showed association with lung cancer and its subgroups. This study identified the susceptibility loci for lung cancer and its covariate-subgroups.

Following the same protocol, we have also attempted meta-analysis for Breast cancer and cervical cancer. 

3. Movement Disorders: Wilson and Parkinson's

Human pigmentation and related disorders

Skin pigmentation in human is one of the most complex quantitative traits ever studied. Epidemiological studies had shown that skin pigmentation is a polygenic trait with high heritability (Frändberg et al., 1998). The key molecule driving variations in pigmentation is melanin and its different forms and the rate limiting enzyme to melanin biosynthesis is Tyrosinase enzyme. Apart from Tyrosinase different other proteins working as channel proteins or differentiation factors, regulate numerous avenues of melanin biosynthesis and its distribution. Thus, a plethora of available opportunities are there to dissect out the intricate regulation of melanin biogenesis. After the advent of genome wide screening studies, it became evident to the scientific world that a lot of hidden players are there which needed to be explored in terms of their roles in Human Pigmentation. Keeping in mind of the above facts my team has been exploring different avenues of human pigmentation like –

(a)  Finding the genetic bases of oculocutaneous albinism in Indian population

Oculocutaneous albinism (OCA) is a heterogeneous group of autosomal recessive disorders characterized by partial or complete loss of melanin in the skin, eyes and hair of the affected individuals. OCA is caused due to the defects in the genes involved in melanin biosynthesis and/or distribution. Based on the mutated gene (TYR, OCA2, TYRP1, SLC45A2, Locus 4q24 for OCA5, SLC24A5, LRMDA), OCA has been classified into seven sub-types (OCA1-7). OCA patients present with various developmental vision related problems, extremely poor visual acuity and thus are often considered to be legally blind. Due to lack of the photo-protective melanin, OCA patients are at a high risk of developing skin cancer. Being a congenital disorder, identification of ethnicity specific mutations and carrier detection followed by pre-marital genetic screening is the only option to contain the disease. With this precise motive, for the last 15 years, I have been actively associated with OCA research, and have pursued mutation screening of ~150 OCA patients from more than 89 families from different places of India especially from the eastern and southern part of the country. Our studies revealed OCA1, caused by TYR gene as the most preponderant subtype (>50%) in part of the globe, followed by OCA2 and OCA4 (Chaki et al., 2005, Sengupta et al., 2007, Ray et al., 2007; Mondal et al., 2012, Ganguly et al., 2017, Ganguly et al., 2020). Haplotype analyses indicated founder effect to be the cause of preponderance of homozygous mutations in the Indian context. In fact, the frequency of OCA1 in the eastern Indian ethnic group ‘Tili’ was estimated to be significantly higher than the world average due to founder effect. Interestingly, in spite of rigorous screening by targeted sequencing method, ~40% of all the alleles screened remain uncharacterized for OCA, worldwide. These uncharacterised mutations (UCMs) have been hypothesised to be located in the regulatory regions of the candidate gene(s), which are not screened in most of the studies, or these may be indicative of new locus/loci precipitating the disease (Ray et al., 2007). Thus, presently I am carrying out Exome Sequencing of the samples with missing heritability.

(b) Assessing the functional bases of tyrosinase related albinism in terms of TYROSINASE protein processing and post translational regulations

Human Tyrosinase enzyme is a monophenol mono-oxidase which catalyses the first two steps of melanin synthesis converting Tyrosine into L-DOPA. Our cell-based studies (Chaki et al., 2010; Mondal and Sengupta et al., 2016) have indicated that retention of mutant Tyrosinase enzyme within Endoplasmic Reticulum to be the reason for pathogenicity of OCA1. Currently through molecular dynamics and cell-based assays, my team is trying to delineate if there is any correlation between physical association of mutant Tyrosinase proteins with cellular chaperons like Calnexin in a time-dependent manner that allows the release of retention of tyrosinase variants from the ER, which could broaden our knowledge further into the protein dynamics (especially post-translational processing) of mutant tyrosinase proteins.

(c) Identifying key pigmentation regulating players for oculocutaneous albinism and vitiligo through bioinformatic analyses of publicly available datasets and literature mining

Although more than a hundred of human genes and their single nucleotide variants are currently known to be associated with human pigmentation, only a handful of them have been examined with respect to their functional insights. The recent boom in publicly available OMICS data (HapMap, 1000 genome, ENCODE, GTEx, Human Protein Atlas project etc.) has paved the way in finding answers to hitherto unanswered/unexplored regions of biological science including pigmentation. My team explores the public domain of all these datasets and bioinformatically assess relevant functional and statistical association data to fish out hitherto unexplored genes and SNPs that might have an implication in human pigmentation (Ganguly et al., 2019) and related disorders like vitiligo, a very common acquired depigmentation problem characterized by white patches due to the loss of the melanocytes distributed over the skin. GWAS and Candidate gene studies have successfully implicated the association between genetic variants and the vitiligo risk, but the pathogenesis of Vitiligo is still not clearly understood. Again, candidate gene association studies have often reported contradictory results with increased risk of false positive associations. Thus, to overcome the contradictory outcomes of functional candidate gene association studies, which has failed to determine the actual effect of the target variants towards Vitiligo risk, we are performing a global meta-analysis and are trying to discern the functional contribution of the genomic variants that have been found to be significantly associated with vitiligo using the predictive in-silico pipeline based on our previous work on skin pigmentation variation (Ganguly et al., 2019).

(d) Elucidating the role of mitochondrial F1-F0-ATPase in melanogenesis

Little is known about the role of mitochondria in pigmentation. Recent studies revealed a few interesting data regarding the potential role of mitochondrial players in melanogenesis. We are trying to explore the potential contribution of F1F0-ATPase and Prohibitin, a mitochondria-nucleus shuttle protein towards melanogenesis.

(e) Identifying genetic variants associated with skin pigmentation in eastern Indian tribal popualtions

Globally, skin pigmentation shows a strong correlation with UVR intensity and latitude as one its surrogates. The dual selective pressures of photoprotection and vitamin D synthesis have created two clines of skin pigmentation: the first cline of deeply melanized skin, from the equator to the poles, with greater need for photoprotection and the second cline with light skin as it facilitates the cutaneous synthesis of vitamin D in presence of UVB ray, abundant in polar region. In recent years, a number of genes and the allelic variants that affect human skin pigmentation have been identified, mostly in European populations. Indian populations, being at the cross-roads of migration, have undergone complex and ancient admixture events over a long period and have been the melting-pot of disparate ancestries originating from different parts of Eurasia and South-East Asia, however, very less is known about their pigmentation genetics. In this context, we aim to assess to assess the human skin pigmentation variation quantitatively with regard to latitude, longitude and altitude in some of the particularly vulnerable tribal populations of eastern India and correlate the same with the genetic polymorphisms. This woud help us assess fixation index as measure of population differentiation due to genetic structure with skin colour as biomarker.

B. Parkinson's Disease

Parkinson's disease (PD) is a multifaceted illness affecting ~0.3% of the world population. The complex genetic bases of PD have not been fully elucidated. Several studies suggest that mitochondrial DNA variants are associated with PD. In this respect, we are trying to explore the possibility of genetic association between mitochondrial haplogroups as well as independent SNPs with PD in east Indian population. Our initial efforts with the common Asian mtDNA haplogroups: M, N, R, B, D, M7, however, did not show any significant differences between patients and controls.

4. Diabetic Retinopathy

Increasing magnitude of diabetes mellitus (DM) in the past decades has become a public health concern in India (Huang et.al 2011). According to a study conducted by All India Ophthalmological Society of India (AIOS) in 2014, the prevalence of diabetic retinopathy was found to be 21.7%, out of which 61.2% were males. Nearly 88.6% were between 40-80 years of age. Our aim is to understand the underlying molecular genetic basis of diabetic retinopathy susceptibility, and pathogenesis through a 3-pronged approach: 

(a) Prioritization of important candidate genes (and SNPs) through meta-analysis of worldwide GWAS and candidate gene association studies from India, followed by testing the prioritized SNPs in eastern Indian patient pool

Candidate gene association studies as well as GWAS have identified different polymorphic variants to be associated with diabetic retinopathy. However, lack of replication studies, small sample size in the candidate study approach, underlying genetic differences between study populations, differences in study design have often led to contradictory results. Moreover, not all the polymorphic variants have been functionally annotated and ascribed a molecular basis towards disease susceptibility. Hence it is difficult to estimate the overall effect of the polymorphic variants found to be associated with DR. We are trying to address this apparent discrepancy through a meta-analysis approach. For meta-analysis, thorough text mining is being done using keywords to narrow down the significant searches from PubMed and then hypothesis constructed based on the extracted data. Data obtained from meta-regression analysis (example: R-4.0.2) are then being analysed for statistical significance by established methods (Sengupta et.al 2017). This would help in evaluation of our hypothesis, followed by selection of the top stand-out SNPs and their genotyping in a representative eastern Indian population for a case-control association study. We plan to do functional studies to validate the statistical association, if any, through cell-based assays.

(b) Identifying cis-regulatory variants conferring risk towards developing DR and testing them in eastern Indian patient pool

The pathophysiology of diabetic retinopathy incorporates many biochemical and inflammatory pathways. Through keywords-based PubMed search like “up/down-regulation AND diabetic retinopathy, microarray AND diabetic retinopathy” we are procuring data of differential expression of genes in diabetic retinopathy. We aim to identify cis-regulatory variants for the differentially expressed genes with special emphasis on the genes directly or indirectly associated with glucose metabolism, by checking their regulatory potential through ENCODE based web-servers. We hypothesise that the presence of these variants confer susceptibility to retinopathy among the diabetics through transcriptional alteration of the target genes. These variants with high regulatory potential would then be genotyped in an eastern Indian case-control cohort to assess their association, followed by functional validation. 

(c) Studying the contribution of mitochondrial genome variations towards the development of DR in eastern Indian population

Studies have shown mitochondrial dysfunction in the pathogenesis of diabetes and eye-related disorders (Nanjaiah et.al 2019). Mitochondrial haplogroups have been shown to be associated with ophthalmologic diseases but the underlying mechanisms are not well understood (Bregman et.al 2017). The role of mitochondria in increased oxidative stress may play a role in the diabetic retinopathy progression. It would be interesting to find out if variability in bioenergetics is a mechanism by which particular haplogroups are more susceptible to advanced forms of this disease. To elucidate the possible contribution of mitochondrial haplogroup in this disease, we plan to genotype specific mitochondrial haplogroups among the eastern Indian patient pool. Increased understanding of the biologic basis for these haplogroup associations may identify new targets for intervention for the treatment of this disease.