Beyond the ivory tower: bridging the disconnect between DNA methylation biomarker discovery and industry for precision diabetes medicine

For decades, Type 2 Diabetes (T2D) has been a leading cause of morbidity and premature death worldwide. It is a chronic condition that affects the way the body processes blood sugar (glucose) which results in too much glucose circulating in the bloodstream. Over time, high blood glucose levels can lead to life-threatening complications, such as heart disease and kidney damage. Thanks to recent medical advances, this is not inevitable. When detected early, high-quality interventions could prevent or delay the onset of T2D. Unfortunately, this promise hasn’t been realized due to our inability to identify those who are asymptomatic and could benefit from treatment. Yet, despite the advancements in diabetes drug development, tests which diagnose and assess the risk for condition still relies on insufficient and aging technology that fails to identify patients in the early stages of the disease when treatment is more effective. The result: the prevalence of T2D continues to escalate while accruing billions of dollars in healthcare expenses.

That isn’t to say that attempts to improve diagnostics haven’t been made, early screening tests for T2D have been in development for decades. Scientists have examined almost every organ and have dived into the depths of our genome to find the root cause of diabetes; however, we none-the-less have not yet substantially improved preventive care. Through this long process, we have learned that the conditions which eventually result in T2D are present years before high blood glucose levels. In other words, patients are already developing complications before our glucose-based tests are able to detect it. Secondly, we have learned that neither lifestyle factors nor genetics alone are enough to cause T2D. We know that diabetes is mediated by dozens of lifestyle factors and hundreds of genetic variants, each with small effects on susceptibility. This implies that it is more complicated than a one-to-one relationship between a risk-factor and disease (i.e., smoking and lung cancer or BRCA1 and breast cancer). Therefore, the unexpected layer of complexity in diabetes development lies in how the genome and human behavior work together synergistically to cause illness. If we are to predict the early stages of T2D, we must first understand the complex interactions between genetics and lifestyle risk factors that underpin its development and progression.

One way to do so is to extend the work that was pioneered by these prior studies by incorporating epigenetics. Epigenetics is the study of how your environment and behaviors can affect the way your genes work. Epigenetic changes are modifications to DNA that regulate whether genes are “turned on or off” and can influence the production of proteins in cells. Unlike genetic changes, epigenetic changes are reversible and do not change the sequence of DNA. Since behaviors (i.e., poor diet and exercise), the cellular environment (i.e., high blood glucose levels) and genetic variation can influence the epigenome, epigenetic-based technology holds promise for revolutionizing diabetes prevention. For example, epigenetic tests can detect disease development prior to clinical signs and symptoms, and can identify which patients would benefit from therapy. This would allow physicians to provide focused treatment for the patients who need it most, when they need it most. Physicians can also leverage epigenetic technology to monitor disease progression and treatment response.

Although the idea of using epigenetics to detect the early stages of T2D is compelling, there have been no attempts by other researchers to demonstrate if the scientific advancements made in epigenetics could actually be implemented into diabetes medicine. Because our laboratory group has successfully developed and commercialized epigenetic technology for other disorders, I sought to determine the feasibility and usefulness of epigenetic information for T2D. My goal is to use this knowledge to develop an epigenetic-based screening test for T2D to address some of the limitations of existing glucose-based methods.

There are several different types of epigenetic changes, including DNA methylation (DNAm) and histone modifications. Our laboratory group studies DNAm because it would be the easiest epigenetic change to implement into current clinical practice. DNAm works by adding a chemical group to specific places in the DNA called CpG sites. Adding and removing DNA methylation from a CpG site is influenced by genetic variation, and environmental and lifestyle factors. Traditional methods of DNAm biomarker discovery often fail because they do not take into account the influence of both genetic variation and environmental factors on the epigenome simultaneously. Typically, methylation turns genes “off” and removal of methylation turns genes “on”.

Several research groups have shown that the methylation of cg19693031, a CpG site in the TXNIP gene, is lower in people with T2D. This CpG site is a promising biomarker because TXNIP plays a role in the development of diabetes and its complications. Unfortunately, the methylation difference between those with and without T2D is not big enough to be used clinically, as it could lead to inappropriate treatment decisions and poor patient outcomes. My research was the first to investigate the influence of genetic variation and HbA1c (long-term blood glucose levels) on cg19693031 methylation. My aim was to determine whether the ability of cg19693031 methylation to distinguish between those with and without T2D would improve if I corrected for the genetic influence on the methylation. My research showed that there is genetic influence on cg19693031 methylation; however, some of these genetic variants only influence the methylation in those with T2D. These influencing genes are key players in pathways that lead to diabetes and other diseases that also involve TXNIP, such as cancer, Alzheimer’s and coronary heart disease. These results suggest that influences on cg19693031 methylation may not be specific to T2D and that even when corrected for genetic influences, cg19693031 is not able to detect the early stages of the disease.

The main cause of pre-mature death in T2D patients stems from coronary artery disease. Considering that several of the conditions which lead to T2D and coronary heart disease overlap, xii I sought to determine if cholesterol measures influenced the methylation of cg19693031 and if these influences were strong enough to hinder its use in diabetes medicine. My research showed that the influence of cholesterol measures on cg19693031 methylation was dependent on the blood glucose levels. This aligns with the clinical observation that diabetes is a risk factor for coronary heart disease and accelerates its progression. We also demonstrated that in T2D patients, cg19693031 methylation was lower in those with a higher risk for heart disease than those with a lower risk. Taken together, our data suggests that due to non-specificity of cg19693031 methylation for T2D that it is unable to be used in diabetes medicine. Simply, cg19693031 methylation seems to measure disease in general, not just T2D. To extend upon this idea, we used cg19693031 and other CpG sites specific for smoking, drinking and heart disease to predict mortality. Future studies will look at harnessing cg19693031 to develop a mortality index to determine a patient’s risk of all-cause mortality and which of these conditions is contributing most to their risk.