Diabetes, also referred as Diabetes mellitus is a chronic condition caused due to failure in the regulation of blood glucose levels in the body. Diabetes and obesity have become a major challenge to the global healthcare. Over the past few decades, diabetes incidences have significantly doubled. In the USA alone, nearly 30 million people are being diagnosed with this condition. The most prevalent form of diabetes is the type 2 diabetes (95%). In Type 1 diabetes, the body's immune system mistakenly destroys the beta cells and thereby, prevents the production of insulin. While, Type 2 diabetes is due to numerous causes such as genetics, heredity, lifestyle or a combination of these factors leading to insulin resistance i.e., one’s body fails to use insulin. There are many possible causes of diabetes, such as the genetic makeup, ethnicity, family history, health factors and environmental conditions.
The success of the Human Genome Project has promoted the beginning of genomic medicine, where knowledge of the genetic data is used to predict, identify and treat human diseases. It has helped clinical practitioners to diagnose diseases accurately, prescribe safer and more effective drugs to treat many human diseases, including cancer, cardiovascular diseases, diabetes, etc., based on the genetic profile of a patient. In recent times, the precision medicine for diabetes is evolving with the accessibility of genomics and health data of a large population. With the advent of DNA sequencing and genome wide association (GSA) studies have contributed in revealing the genetic base for diabetes susceptibility and its progress, though the genetic markers still remain elusive.
In genome wide association (GWA) findings are carried out to identify over several hundreds of thousands minor- and low-allele frequency (MAF) variants across the noncoding (intron) and protein coding (exon) sections of the genome. These studies have resulted in the recognition of over 50 genetic loci linked to different glycemic traits, and at least 90 genetic loci linked to type 2 diabetes. For instance, a variant TCF7L2 (rs7903146) in the intronic region is strongly related with type 2 diabetes, with a 37% higher risk per copy of the variant allele. Type 2 diabetes is also predicted using the Genotype Score (GS) in addition to common risk factors. In a Framingham Offspring study, single-nucleotide polymorphisms (SNPs) were genotyped at 18 loci linked to diabetes and created a GS from the number of risk alleles to predict the risk of diabetes. However, the GS failed to identify new cases in the population, but improved the prediction of common risk factors.
The knowledge about the biological modifications and genetic data could be very useful in predicting diabetes. For example, the risk of type 2 diabetes is linked to DNA methylation occurring at the CpG islands which is a vital mechanism of gene expression regulation. Likewise, metabolic profiles of small molecules such as amino acids may also associate with diabetes, especially in young adolescents. Nevertheless, these genomics data may require more assessment to effectively adopt and validate their clinical utility.
Although, genetic markers are not very useful in predicting or preventing diabetes, they play a significant role in differentiating type 1 and type 2 diabetes. The prevalence of obesity poses a challenge in distinguishing diabetes type as many kids, teen-agers and adults having type 1 diabetes are also obese. Thus, an inappropriate classification may lead to substantial health risks. For example, wrong analysis of type 2 diabetes might lead to incorrect intake of glucose-lowering drugs orally. Likewise, an inappropriate diagnosis of type 1 diabetes will lead to unnecessary intake of insulin. Diabetes genetic risk scores (GRS) generated from the genetic loci, HLA genotypes, clinical factors and autoimmune antibody tests have assisted in distinguishing between type 1 and type 2 diabetes, and assisted in predicting the requirement of insulin treatment to patients within 2-3 years of diagnosis. Genetic testing is a helpful tool to diagnose certain forms of diabetes affected by a single gene defects, such as mutations in HNF1A and KCNJ11 genes lead to maturity onset diabetes of the young (MODY) and neonatal diabetes, respectively (both are monogenic forms of diabetes). Other genetic variants, such as glucokinase (GCK), HNF4A, HNF1B, NEUROD1 and IPF1 also contribute to a MODY form of diabetes. Likewise, neonatal form of diabetes is also identified due to PLAG1, ABCC8, GCK and INS genetic variants. Studies have shown that about 50% of the genetic risk for type 1 diabetes associated with the Major Histocompatibility Complex (MHC) region positioned on chromosome 6 (6p21.3).
There are many approved drugs for treating diabetes. Most of them have comparable effects on hemoglobin A1c (HbA1c), however they vary in their treatment response and adverse effects depending on the individual. Thus, pharmacogenomics will be useful in testing genetic variants to define subgroups of diabetes patients, i.e., segregate patients predictable to have more benefits from or less harmful effects of a specific drug. This could be one of the most potential practice of curing diabetes using genomics. Pharmacogenomics investigations have been initiated to locate candidate genes linked to absorption, circulation, metabolism, biologic effect and elimination of drugs on their targets. Some of the selected pharmacogenomic findings for metformin and sulfonylureas are given in figure 1.
Figure-1: Selected pharmacogenomic findings for sulfonylureas and metformin.
(Adopted from Floyd and Psaty 2016; https://doi.org/10.2337/dc16-0738)
Genetic variants of CYP2C9, TCF7L2, KCNJ11 and ABCC8 genes are shown to greatly lower the glucose levels and increase hypoglycemia risks. The remarkable success of precision medicine is due to the fact that the designed drugs target pathogenic mutations exactly. The gene, ADRA2A encodes for α2A adrenergic-receptor (α2AAR) and any defects/variants in this gene result to overexpression of α2AAR and leads to diminished secretion of insulin and type 2 diabetes. The use of indole alkaloid, yohimbine (a naturally occurring compound from Pausinystalia johimbe) is known to block this receptor and increases insulin secretion.
Some of the challenges faced by diabetes genomic research include the restricted detection of variants and associations, the degree of the consequence to be discovered and the sample size. Though, large genome wide association studies involving large numbers of patients have assisted in revealing the biology behind many intricate diseases, most of the identified genetic loci have insignificant effects. Overall, genomic findings for diabetes have suggested the possible application of personalized medicine in near future. However to make it ready for implementation certainly requires the large populations genomic data. Genomic discoveries and health data from pharmacomedicine offer to identify the limitations, and help to lead the future direction of research. Nonetheless, till date, drug therapy in diabetic patients do not consider about the patient’s genetic diversity, and pharmacogenetics application in diabetes diagnosis and cure is still in initial stages. The recent advancements in the omics-allied approaches (transcriptomics, metabolomics, proteomics, and epigenenomics) added with improved studies involving a large number of patients are expected to improve understanding on the genetic aspects of diabetes patients and help to make use of personalized genomic medicine for diagnosing and treating diabetes.