Characterisation of the role of NPAT in metformin regulation of body weight and glycaemic control; moving GWAS data to biological function

There is no cure for diabetes. The main drug prescribed for diabetes (taken by millions world-wide) is metformin, which improves insulin action to reduce blood glucose levels. However metformin does not reduce blood glucose in all patients, it can take 3-6 months of treatment to establish whether the drug is effective and its efficacy reduces with time. Our previous work identified genetic variations in people with diabetes, which were linked to the ability of metformin to improve glucose control. This implies that these gene sequences directly influence an aspect of the drug action in the diabetes population. The most significant genetic polymorphism focused to a region of the genome, which encodes 7 genes including the cell growth regulating genes ATM and NPAT. Further work, funded by the MRC, confirmed ATM and NPAT as the genes responsible for the variation in metformin action in human patients. The ultimate purpose of doing these very expensive genetic studies is to identify 'markers' that could help predict which patient will respond to which drug, without having to wait for months of treatment. The reality is that very few of these markers are strong or common enough to be used in this way. However the second big hope is that understanding how the 'hits' we identify (like ATM and NPAT) influence drug action may illuminate how the drug works and why it doesn't work in some people. This will help improve our ability to treat or prevent the disease. Our data argues that changing NPAT amount modifies the body response to metformin. We now propose to gain detailed biological and molecular insight into how NPAT influences metformin action.

We will perform investigations to elucidate the physiological, cellular and molecular connections between metformin action and the NPAT gene:
Firstly we will characterize glucose metabolism and hormone action in a mouse lacking NPAT and in a mouse lacking NPAT only in the liver (this is because metformin is thought to work mostly on the liver). We will use state of the art techniques to compare how the liver, muscle, fat tissue, pancreas and brain work together to control body weight and blood glucose and the response to hormones and metformin in these models with defective NPAT. This will tell us if NPAT is necessary (and if liver NPAT specifically is necessary) for basic regulation of glucose physiology, and in the clinical response to metformin.

Secondly we will perform studies on isolated liver cells lacking NPAT to learn how NPAT contributes to biology inside the cells. Similarly we will use molecular approaches to increase or decrease NPAT in immortal cell lines representative of different tissues. This allows us to compare effects of increased or decreased production of NPAT proteins on specific functions of liver, pancreas and brain. We will also make cells that produce the genetic variants in NPAT found in the human patients and study how this effects NPAT biology and metformin action.

Our overarching aim is to establish why NPAT gene sequences influence the response to therapy in people with diabetes. This information will help clinicians decide on the best early treatment options for people with diabetes and help scientists develop more effective therapies for this common disease.

Metformin is the most common diabetes treatment despite the fact its mechanism of action remains unclear and it does not reduce blood glucose in all patients. Our previous work produced compelling genetic evidence that variations in the genes for the cell growth-regulating gene ATM and the transcription factor NPAT are associated with metformin response. We now wish to gain biological and molecular insight into how NPAT actually influences metformin action, and whether alterations in NPAT function could contribute to the development of obesity and diabetes. We aim to 1) establish the detailed physiological and metabolic consequences of NPAT deletion in mice, 2) establish the detailed physiological and metabolic consequences of NPAT deletion specifically in the liver of mice, 3) investigate the functional consequences of the genetic variations in the coding regions of NPAT in cellular systems and 4) investigate the functional consequences on glucose homeostasis, metformin response and hormone action of increasing or decreasing NPAT protein in cellular systems.

The work will include metabolic phenotyping of 2 mouse models (NPAT-/-, and NPAT liver -/-). We will perform hyperinsulinemic-euglycemic and hyperglycemic clamps, glucose, pyruvate, leptin and insulin tolerance tests, food intake, and indirect calorimetry on mice placed on a high fat diet with or without metformin. This will establish the outcome of altered NPAT expression on energy metabolism and response to metformin in the intact animal. In cell-based studies where NPAT is deleted, mutated or overexpressed we will establish its influence on specific aspects of glucose metabolism, cellular bioenergetics and metformin action. This will include generation of stable cells that express each point mutant of NPAT identified in the human GWAS.

Finally, we will use the information gained to guide human studies on diabetes treatment (funded by the Wellcome Trust), which are ongoing within our group.

Who will benefit from this work?
The direct beneficiaries of this work are academic research groups and funding bodies working in the fields of diabetes, cancer, obesity, metabolism, and pathways to healthy ageing. In addition, small biotechnology companies and the pharmaceutical industry developing therapeutics for all aspects of healthy ageing. The project is primarily preclinical research aimed at improving mechanistic knowledge on drug action but it is likely that the outcomes will be sufficiently developed within the timeframe of the grant to engage with many non-specialised groups. There is clear translational potential in studies using the information generated to improve patient stratification and drug prescribing, which will allow more direct communication with interested public groups (charities, health workers, carers and patients).

How will they benefit from this research?
The outputs of this research proposal will yield novel results that address a number of important clinical and basic science areas: The mechanism of action of metformin; pharmacogenetics of metformin and potential for stratified medicine in diabetes; the role of the NPAT gene in glucose homeostasis and body weight control, and the potential for metformin use in cancer. A key output will be to unravel how NPAT influences metformin action. Our previous MRC funded human genetic studies identified NPAT and ATM as modulators of metformin response, yet we still don't know why variations in their sequence alter metformin response in humans. As such we aim to establish if these proteins influence additional aspects of energy homeostasis. The mouse studies are a fundamental step to translate these human genetic observations into mechanistic insights, which can then be converted back into patient benefit through improved patient stratification and novel therapeutic opportunities.

The discovery that the NPAT/ATM gene locus is involved in metformin action, links together two areas of research that are converging - cancer and diabetes. The recent reports that key 'cancer' genes, e.g. p53 and ATM, are involved in glucose regulation; that type 2 diabetes genes and some cancer genes overlap; and that metformin protects against both type 2 diabetes and cancer offers an exciting area for research and impact on both the basic science community and clinical care. As such this proposal is at the forefront of research in this field, and will yield important results that could potentially translate into change to clinical practice in the medium term. One potential area for early clinical impact resulting from this proposal will be in the possible role for metformin in the treatment with Ataxia-telangiectasia, who are homozygous for ATM mutations, develop early lymphoproliferative cancer and are insulin resistant.

Finally, we believe that the time is right to use the exciting developments of human genetic research to focus physiological and molecular research to provide a mechanistic basis for the genetic observations. The clinical utility of human genetic information will clearly be increased by improved mechanistic insights from follow up biological research of this type. As yet very few of the exciting discoveries linking genetic variation to disease have uncovered novel clinical benefits and therefore it is vital to develop the knowledge of how these genes regulate biology in order to fully justify collecting the genetic information.