Immunogenomics advances point to new biomarkers, therapies

Next-generation gene-sequencing technology and new data-analysis tools are pointing the way to fresh diagnostic and treatment approaches for autoimmune diseases, cancer and many other conditions. That was the message at Immunogenomics 2014, a recent conference hosted by Huntsville’s HudsonAlpha Institute for Biotechnology and Science magazine for researchers studying the interaction between genes and the immune system. The event was sponsored in partnership with UAB and its Comprehensive Arthritis, Musculoskeletal and Autoimmunity Center (CAMAC).

Investigators from major national and international research institutions described how detailed profiles of immune cells could improve response to influenza vaccines and accelerate new treatments for emerging infectious diseases. They explained new immune-mediated links between microbial populations and cancer risk, and highlighted progress in understanding the pathogenesis of complex diseases such as multiple sclerosis.

“We now have the tools to examine these gene-disease associations in finer detail,” said S. Louis Bridges Jr., M.D., Ph.D., director of UAB’s Division of Clinical Immunology and Rheumatology and the CAMAC. Bridges is using genomic techniques to study the autoimmune condition rheumatoid arthritis. Bridges presented his research at Immunogenomics 2014, and served as chair of a session on the genetics of complex disease. “We’ve started to refine our analysis to home in on cells with an increasing degree of specificity,” Bridges said. “Technology is now allowing us to analyze in more detail specific subsets of cells, including ultimately at the single-cell level.”

From Associations to Biomarkers

Genomewide association studies have identified a host of genetic changes linked with disease. “Now investigators are looking at the functional effects of these polymorphisms,” Bridges said. “It could be that a polymorphism affects expression of a certain gene, or it may only affect expression of that gene in a certain cell type.” Bridges’ project, being performed in collaboration with UAB epidemiology chair Donna Arnett, Ph.D., and HudsonAlpha investigator Devin Absher, Ph.D., is examining genetic risk factors in African-Americans with RA.

"We now have the tools to examine these gene-disease associations in finer detail," Bridges said. "We've started to refine our analysis to home in on cells with an increasing degree of specificity."

In his talk at Immunogenomics 2014, Bridges demonstrated that overexpression of the interferon gamma 2 receptor gene is strongly linked with severity of disease in African-American patients with RA. “Our next step is to see in which cells that particular expression occurs,” Bridges said. This work could ultimately point the way to biomarkers that tell clinicians which of several known signaling pathways is active in a patient with RA, guiding treatment decisions.

Epigenetics and Therapeutics

Researchers are also focusing increasing attention on the ways gene expression is regulated dynamically in cells through epigenetic changes, says Robert P. Kimberly, M.D., director of the UAB Center for Clinical and Translational Science. Epigenetics refers to mechanisms that alter gene expression without changes in the actual DNA sequence. One of the most common epigenetic changes is methylation. When a methyl group attaches to the DNA base cytosine, it blocks the ability of the neighboring gene to be expressed.

Tracking and analyzing epigenetic markers implicated in a particular disease — such as systemic lupus erythematosus, one of Kimberly’s own research interests — could give clinicians crucial information on “if to treat, when to treat and also how to treat” that condition, he said.

For example, if a key risk gene is hypo-methylated in a patient — increasing the likelihood of the gene’s being expressed — “that could mean the patient is poised for a flare-up of disease,” Kimberly said. Another patient, who would look exactly the same to a clinician, would be much less likely to have a flare-up if that gene is hyper-methylated, he adds. “Understanding this epigenetic regulation, and eventually manipulating it to therapeutic advantage, is very exciting.” This research is a focus of several investigative teams at UAB, Kimberly says.

A “tree map” depicting the immune diversity in a patient diagnosed with Parkinson’s disease. Each rectangle represents a unique antigen receptor detected in the sample, and the size of each rectangle represents the relative frequency of that receptor within the sample. (Color is arbitrary.) Image courtesy iRepertoire.

Profiling Immune Signatures

One advance highlighted by several presenters at Immunogenomics 2014 was immune repertoire sequencing. Researchers now understand that an individual’s immune response is based greatly on the specific cell populations, or “repertoire,” present in that individual. For both T and B cells, for example, millions of distinct variants are possible. The exact mix is determined by a person’s encounters with microbes, disease and other environmental exposures over a lifetime, explains HudsonAlpha investigator Jian Han, M.D., Ph.D., a 1991 graduate of UAB’s medical genetics doctoral program.

Everyone produces T cells, for example, Han says. “But which one of those naive T cells gets used is determined by if it met, and was activated by, its antigen.” By analyzing the variety of T and B cells present in a particular patient, or mapping the repertoire commonly found in a particular disease, researchers can identify biomarkers to aid in diagnosis and treatment, Han notes.

Han’s HudsonAlpha lab has pioneered the multiplex PCR technology needed to gather the massive amounts of data required for repertoire sequencing, and the analytical tools required to resolve that data into meaningful reports. His presentation at Immunogenomics 2014 focused on Repertoire10K, a HudsonAlpha-funded project to sequence the immune repertoires of 10,000 patients: 100 with each of 100 critical diseases. The goal is to identify a signature in the immune repertoire for each disease. UAB researchers have been key contributors of the genetic samples that are critical to the project, Han says. In return, the investigators have access to state-of-the-art sequencing data that can advance their own studies.

Team-based Science

Collaborations between investigators at HudsonAlpha and UAB have taken place since the institute first opened in 2008, Kimberly says. But the new UAB–HudsonAlpha Center for Genomic Medicine, launched this summer, will increase these research partnerships and speed new discoveries in immunology, cancer, cardiovascular disease and many other fields, he notes.

Leveraging the strengths of each institution is critical as the scale of the research challenges becomes ever greater, Kimberly adds. “To understand what’s really happening in disease states, we’re going to have to be able to take all the data on genomics, epigenetics and more and figure out how to pull it all together,” he said.

That’s why UAB is also creating a new Informatics Institute. It will work in tandem with the UAB–HudsonAlpha Center for Genomic Medicine and a third initiative, the UAB Personalized Medicine Institute, to build the infrastructure and recruit the data scientists needed to succeed in this new era of research.

“It’s a major frontier right now,” Kimberly said. “The algorithms to combine all this data for the most part haven’t even been formulated yet. But it’s clear that the institutions that succeed in the future will be the innovators in this area.”

Discovery route: Path to potential diabetes drugs began with a simple question

More than 12 years of research led Anath Shalev, M.D. (right, with Junqin Chen, Ph.D.) from a basic discovery to the first human trial of a new type of diabetes drug.

In 2002, diabetes researcher Anath Shalev, M.D., asked a basic question: What gene in the insulin-producing islets of the human pancreas is most turned on by high levels of glucose, a hallmark of diabetes?

The answer has led the UAB endocrinologist to discover new cellular pathways in beta cells of the islets, pathways that are a key to diabetes progression or protection. Those discoveries have now opened the door to the first human trial of a potential diabetes drug with a mode of action different from any current diabetes treatment. (Learn more about the trial, which will begin in early 2015, in this story.)

Anath Shalev explains verapamil's protective effects against diabetes, and a new human clinical trial of the drug at UAB, in this video.

Beta cells are critical in type 1 and type 2 diabetes. In both diseases, the cells are lost gradually due to programmed cell death (apoptosis); but the trigger for that programmed death was unknown. The loss of beta cells contributes to the progression of diabetes, a growing worldwide epidemic that affects more than 20 million people in the United States, making it the seventh leading cause of death and the source of complications like blindness and more than 40,000 lower limb amputations a year.

From Molecular Mechanisms to New Treatments

The beta-cell gene that responded to the high glucose in Shalev’s 2002 experiment produces TXNIP (pronounced "ticks-nip"), a protein normally involved in controlling oxygen radicals in many types of cells but never known to be important in beta-cell biology. Its response to glucose was intriguing because TXNIP (thioredoxin-interacting protein) was already recognized as a regulator of thioredoxin. Overexpression of thioredoxin had previously been shown to prevent experimentally induced diabetes by inhibiting the programmed death of islet beta cells. Since TXNIP inhibits thioredoxin, and because Shalev had discovered that islet TXNIP was highly regulated by glucose, Shalev realized that TXNIP might have major implications for beta-cell biology.

What does it take to go from a basic microarray gene discovery to a human trial of a completely novel drug to treat diabetes?

A dozen years of elegant research unraveling the control and function of a protein called TXNIP.

Over the next dozen years, Shalev — who left the University of Wisconsin–Madison to head the UAB Comprehensive Diabetes Center in 2010 — set out to reveal how TXNIP acts in cells at the molecular level, knowing that an understanding of those molecular mechanisms might point to possible new diabetes treatments. The payoff has been substantial: Using cell cultures, mouse models and pancreatic islets isolated from humans, the Shalev lab team has shown that manipulating TXNIP levels up or down in beta cells could exacerbate or protect against experimental diabetes.

Details about the research journey show the incremental steps that basic science takes, and how those connected steps sometimes lead to potential clinical impacts.

Controlling TXNIP to Treat Diabetes

In 2005, the Shalev lab team found that beta-cell TXNIP levels are higher in mouse diabetes models, and that experimentally increasing TXNIP levels in rat beta cells in vitro led to increased programmed cell death, by means of a well-known trigger signal of apoptosis. The Shalev team also found that sugars in general, whether metabolized or not, turn the TXNIP gene on. This clue led them to a newly identified carbohydrate response element (ChoRE) in the TXNIP promoter that acts as a regulator of TXNIP.

In 2008, the Shalev lab developed mice that had little or no TXNIP in their beta cells. These lower levels protected against experimental diabetes. The team also discovered that the lower levels sent a known signal that inhibited mitochondrial beta-cell death. Shalev wrote, “These results suggest that lowering beta-cell TXNIP production could serve as a novel strategy for the treatment of type 1 and type 2 diabetes by promoting endogenous beta-cell survival.”

An Approved Drug Offers Protection

In 2012, the Shalev group tested an already approved oral drug that they had earlier found to reduce levels of TXNIP in heart cells. The drug — verapamil — is a calcium channel blocker used primarily to treat high blood pressure, but also to treat migraine headaches. Shalev’s team found that exposing in vitro beta cells or isolated human islets to verapamil reduced TXNIP expression, and halted programmed apoptotic death of beta cells. Furthermore, mice that were fed verapamil in their drinking water were protected from experimentally induced diabetes, and verapamil rescued mice that already had diabetes. The verapamil mice had lower TXNIP levels and less programmed beta-cell death, as well as better levels of insulin

"I actually went down to the mouse house to see if the mice were getting diabetes," Shalev told The Birmingham News in 2012. When she found normal glucose levels, "We were dancing."

In those studies, the group also revealed how verapamil lowers TXNIP — the decreased intracellular level of calcium ions caused by verapamil led to phosphorylation of the ChoRE binding protein that normally responds to glucose to control TXNIP transcription at the ChoRE. This phosphorylation prevented the binding protein from entering the beta-cell nucleus and interacting with the TXNIP gene. Shalev noted that these verapamil results identified, for the first time, “… an effective pharmacological means … to inhibit pancreatic beta-cell expression of proapoptotic TXNIP, enhance beta-cell survival and function, and thereby prevent and even improve overt diabetes and shed light on the mechanisms involved.”

Another Role for TXNIP, Another Drug Target?

In 2013, TXNIP was shown to play another crucial role in beta-cell biology when the Shalev laboratory team discovered that high levels of TXNIP directly blocked insulin production in beta cells, acting through a newly identified pathway. TXNIP, they found, induced a microRNA called miR-204, which in turn down-regulated the MAFA transcription factor involved in promoting transcription of the insulin gene.

This means that miR-204 may offer another target for a future RNA drug, an area that is currently also being actively pursued by the Shalev lab. MicroRNAs, with 20 to 24 noncoding nucleotides, have rapidly gained prominence as regulators of gene expression in health and disease. Researchers are beginning to explore whether silencing targeted microRNAs may lead to a treatment for cancers or other diseases.

TXNIP's Vicious Cycle

This year Shalev reported that TXNIP — surprisingly — can induce its own transcription. Her UAB research team found that TXNIP does this by affecting the same ChoRE binding protein (ChREBP) that was previously found to be key in the response to the drug verapamil. The researchers experimentally elevated TXNIP levels in beta cells and found this caused decreased phosphorylation of ChREBP, which led to its increased entry into the nucleus and its increased binding to the TXNIP promoter to boost transcription. This creates a harmful positive-feedback loop.

"These findings support the notion,” Shalev wrote in this 2014 paper, “that TXNIP levels rise over time, not only as a result of elevated blood glucose levels and/or endoplasmic reticulum stress, but also as part of a vicious cycle by which increased TXNIP levels lead to more TXNIP expression and thereby amplify the associated detrimental effects on beta-cell biology including oxidative stress, inflammation, and ultimately beta-cell death and disease progression.”

First Human Trial

Get a quick overview of the science behind UAB's verapamil
trial in this animation

The story doesn’t end here. Shalev’s long trail of laboratory research has now led to the first human trial to see if verapamil has an effect in patients who have developed type 1 diabetes within the previous three months. Adult volunteers, ages 19-45, will be treated with verapamil or a placebo for one year, as their insulin and blood glucose levels are continuously monitored. The three-year, $2.2 million trial will be conducted by the UAB Comprehensive Diabetes Center with funding from JDRF, the largest charitable supporter of type 1 diabetes research.

Meanwhile, a UAB partnership with the Southern Research Institute — called the Alabama Drug Discovery Alliance — is already working to develop small therapeutic molecules that mimic the diabetes-protecting effect produced by verapamil and inhibit TXNIP, but have a greater selectivity and efficacy. [Learn more about this work, and other high-potential projects in the Alabama Drug Discovery Alliance, in a new feature from UAB Magazine.]

So Shalev’s simple question — what gene in insulin-producing beta cells is most turned on by glucose? — has thus led the research out of her laboratory to possible new drugs, acting against a novel target to alleviate or reverse diabetes.

— Jeff Hansen