The increasing prevalence of metabolic diseases and their associated complications is a major healthcare burden to both developed countries and developing nations. Diabetes in particular affects more than 415 million people worldwide in 2015, and the projection is alarming at more than 640 million by 2040. In the United States, diabetes afflicts 12.3% of the adult population, and accounts for an annual healthcare cost of $245 billion in 2012. Across the Pacific, the prevalence of diabetes in Asia displays a similarly bleak outlook: with 11.6% or 114 million people in China and 11.3% in Singapore [1].

The major type of diabetes is type 2 diabetes, which is due to reduced insulin secretion and impaired insulin sensitivity. As such, significant efforts, including our own, have been devoted to understanding the molecular regulation of insulin granule exocytosis and insulin-stimulated glucose transporter translocation [2-10], the basic cellular events that underlie glucose homeostasis. Of particular relevance to Asia, it is very interesting to note that the majority of the Asian diabetes patients are lean, in contrast to their obese counterparts in the Western countries. The biological and genetic basis for the lean vs the obese diabetics are beginning to emerge, and currently a leading model attributes this distinction to the poor plasticity and function of adipose tissue in Asians, which leads to ectopic fat accumulation in other metabolic organs, such as skeletal muscle and liver, and the ensuing insulin resistance and diabetes. Diabetes is a major risk factor for a number of complications, including peripheral neuropathy, nephropathy, hypertension, and cardiovascular diseases.

Not only Asians are more prone to the disease, they also appear to follow a different course of development in the associated complications. For example, heart failure (HF), which is highly associated with diabetes in Asia, hits Asians 10 years earlier compared with patients in the United States and Europe. Observations in ~1000 HF patients in Singapore reveal a unique Asian phenotype of lean diabetic and hypertensive HF (prevalence of diabetes 55% and up to 75% in Indians, compared to ~25-30% in similar Western cohorts), associated with endothelial dysfunction and high circulating inflammatory response biomarkers.

Both genetic and environmental factors are at play that Asians respond differently and show different disease patterns, which dictates that we cannot simply extrapolate the data from the US and Europe to the Asian patients. There are now concerted efforts to address this uniquely Asian situation. In particular national research institutes like A*STAR’s SBIC (Singapore Bioimaging Consortium) and GIS (Genome Institute of Singapore) have teamed up with their clinical partners such as SGH and NUH to establish a transboundary research program to tackle the heart failure problem. This project, ATTRaCT (Asian neTwork for Translational Research and Cardiovascular Trials), which spans across 8 ASEAN countries, brings together expertise across multiple disciplines such as cardiology, genetics, immunology and imaging, representing a more than $100 million private-public partnership. Biomedical imaging is a core theme in ATTRaCT, which is spearheaded by SBIC. The primary aim is to develop and apply advanced cardiometabolic imaging in the detection of novel structural and functional cardiovascular changes, prior to as well as from the onset of heart failure.

Cardiac magnetic resonance (CMR) imaging provides superior image resolution and clinical diagnostic biomarkers compared to the current standard echocardiography, and indeed it is the gold standard in the diagnosis of multiple cardiovascular diseases. The advent of novel MR technology including hyperpolarized carbon-13 further value-adds its diagnostic value by allowing real-time longitudinal measurements of myocardial energetics and substrate utilization [11-15]. In addition the combination of MRI & PET exploits the advantages of these imaging modalities, offering high biomolecular sensitivity with superior anatomical and functional imaging for the early detection of metabolic abnormalities in HF before clinical symptoms manifest. By imaging selected patient cohorts (lean vs obese diabetics; normotensive vs hypertensive) at different stages of HF progression, coupled with parallel longitudinal studies in pertinent animal models to understand the biological mechanisms involved in the pathology, valuable insights into novel metabolic targets may be identified.

To mimic the unique Asian lean diabetic phenotype and associated HF development, we have established a lean diabetic model by genetic manipulation of the BSCL2 gene. BSCL2 (Berardinelli-Seip Congenital Lipodystrophy type 2) is a genetic disorder characterized by the near complete loss of both mechanical and metabolic adipose tissue, severe insulin resistance, hyperglycemia, hepatomegaly and hyperinsulinemia. BSCL2 encodes the protein seipin, whose functional roles in adipogenesis and obesity are being studied [16-21]. Clinical studies indicate that BSCL2 is associated with hypertrophic cardiomyopathy and HF, with ~80% patients present as having preserved ejection fraction.

Phenotyping analysis shows that the whole-body BSCL2 KO mouse model recapitulates all the clinical symptoms of BSCL2 patients, including near complete loss of both subcutaneous and visceral adipose tissue, severe insulin resistance, and glucose intolerance. In fact hyperpolarized carbon-13 MRS reveals metabolic changes in carbohydrate oxidation at 6 months of age, before clinically-relevant structural and functional changes such as LV hypertrophy and ejection fraction occur. Ongoing research efforts are focused on characterizing cardiac phenotypes as HF develops.

The small animal models are integral to the iterative process in understanding the biological basis of HF, as novel candidates discovered from human studies can be validated by generating and phenotyping new animal models followed by molecular and mechanistic studies of the disease progression, and therapeutic development and validation. Besides the above lean diabetic mouse model, we are also establishing a platform to image large animal models, including a porcine model of HF.

Central Regulators of Energy Homeostasis and Obesity At the other end of the extreme of lipodystrophy is morbid obesity. A research program within SBIC is to discover novel regulators in the CNS by combining molecular physiology and molecular imaging approaches [22-29]. For example, we have generated a mouse line with deletion of two genes important for neurotransmitter release from the hypothalamus. These double knockout mice gain weight at a much faster pace and much more weight under normal chow diet, and can reach nearly 100 grams by 1 year of age. As a comparison, wildtype mice at 1 year of age typically weigh 40-50 grams. As expected, these mice show much higher body fat content and percentage of body weight. MRI clearly shows dramatically increased body fat at multiple compartments. In addition, metabolic imaging with hyperpolarized carbon-13 reveals increased hepatic gluconeogenesis in the fatty liver of a type-2 diabetes mouse model, which can be reversed upon therapeutic intervention with metformin.

We have also implemented other imaging techniques to study the changes in fat volumes, fat composition and their role in the progression of metabolic disease (34-41). In particular, our results revealed an extra-myocellular lipid signal that exists only in the spectra acquired from the heart of rats fed with a high-fat diet, contrary to Chow-fed rats. This study proposes that lipid accumulation in the myocardium might be an early biomarker of cardiovascular pathophysiology associated with type-2 diabetes. Furthermore computational algorithms are being developed to distinguish and quantitate white and brown fat depots.

The obesity pandemic propels the search for novel preventive and therapeutic interventions. The discovery of metabolically active brown adipose tissue (BAT) in human adults led to a paradigm shift towards its role as a regulator of energy metabolism. Studies have suggested that BAT quantity in humans inversely correlates with body mass index, thereby proposing the generation and pharmacological activation of BAT as a potential means to counter obesity. Concomitant studies have also demonstrated the transition of white adipocytes to brown-like adipocytes is feasible under certain conditions (“browning of WAT”). BAT detection through near infrared (NIR) fluorescence based-optical imaging has been proposed as a cost-effective method that may be useful in screening compounds that are modulators of BAT activity in vitro and in vivo. This high-throughput technique could aid in the discovery of BAT-stimulating drugs for the treatment of obesity and diabetes. In particular SBIC is examining the use of phytochrome-derived NIR fluorescent proteins and chemically-synthesized NIR probes to detect BAT and to measure the “browning” of white adipocytes. Additionally, we are also exploring MR based measures of diffusion and perfusion that exploit differences in morphology and vascularization to distinguish between BAT and WAT tissues.

Singapore contains three main ethnic groups: Chinese, Malay and Indian, which represent nearly two-thirds of the population in Asia, a region where the burden of type 2 diabetes is likely to grow rapidly over the next several decades. Through the Singapore Adult Metabolism Study (SAMS), we sought to understand mechanisms through which ethnicity modulates the association between body composition and metabolic health.

1. Investigation of ethnic differences in fat distribution and its effect on metabolic health

In this study, we sought to determine whether the propensity to accumulate fat in specific fat depots with increasing adiposity mediates the interaction between ethnicity, adiposity, and insulin sensitivity. A total of 268 subjects (101 Chinese, 82 Malays and 85 Indians) participated in the study. The results showed that ethnicity modulates the association between adiposity and insulin sensitivity, but this relationship cannot be purely explained by differences in body fat partitioning. Insulin sensitivity was inversely correlated with all the fat depots. Chinese had the highest insulin sensitivity followed by Malays and Indians. Malays tend to become more insulin resistant with increasing adiposity compared to Chinese and Indians [30].

2. Adipose Tissue Hydration - Noninvasive Imaging Marker for Hypertrophic Obesity

We also sought to investigate if ‘fat quality’ could further explain ethnic differences in the insulin sensitivity, in addition to fat partitioning. Subcutaneous adipocyte hypertrophy has been linked to impaired fat storage, inflammation, lipotoxicity, and systemic insulin resistance, independent of BMI. In this study, we investigated the use of adipose tissue hydration, measured noninvasively using 1H MRS based hydrolipidic ratio (HLR), as a potential noninvasive marker of adipocyte hypertrophy. In a sub-study, we analyzed adipose tissue samples extracted from high-fat diet fed rats, and established a negative association between HLR and adipocyte size. HLR was negatively associated with leptin in all three ethnicities and negatively associated with insulin sensitivity index (ISI), serum triglycerides, LDL-cholesterol in Chinese and Malays but not in South Asians. After controlling for the confounding effect of body fat %, there were ethnic differences in association between HLR and ISI, with Malays showing a strong effect of HLR on ISI.

3. Ethnic differences in the role of adipocytokines linking abdominal adiposity and insulin sensitivity among Asians

The inability of body fat partitioning to explain why ethnic differences that exist in ISI in lean individuals are lost in obese individuals, motivated us to examine other adipose related factors i.e. adipocytokines to explain the ethnic differences in insulin resistance. In this study, we tested the hypothesis that ethnic differences in the adipocytokines production might mediate the relationship between abdominal adiposity and insulin resistance among three major Asian ethnic groups. We found that the negative impact of visceral and subcutaneous abdominal adiposity on insulin sensitivity differs between Asian ethnic groups. While excess abdominal adiposity may result in abnormal production of adipocytokines, only a cluster of adipocytokines consisting of leptin, adiponectin and FGF21 mediates the relationship between abdominal adiposity and insulin resistance. This is observed only in South-Asians, but not in Chinese and Malays [32].

Liver cancer is a leading cause of cancer mortality worldwide. The incidence of liver cancer in East and South-East Asia is disproportionately high, which is largely attributable to chronic hepatitis virus infection. The majority of liver cancers in Asia are hepatocellular carcinomas (HCC). The high mortality of liver cancer is due to late diagnosis and limited treatment options. Overall, the 5-year survival rate is approximately 15%. The patients who are diagnosed early with resectable liver cancer have ~40% chance of survival. However, signs and symptoms often do not appear until liver cancer is at an advanced stage. For advanced, unresectable liver cancer, only two treatment options are available (i) TACE (trans-arterial chemical embolization), an interventional radiology procedure and (ii) Sorafenib, a chemical inhibitor of PDGF/VEGF and Raf kinases, the only FDA-approved biological drug for treating liver cancer. However, neither option offers significant extension of life. In effect, advanced liver cancer remains incurable and untreatable. Given this dismal outlook, our intention is to focus on the development of new strategies to treat advanced HCC.

The concept of cancer as a genetic disease has been the dominant hypothesis of the last two decades. This general notion has been supported by numerous studies documenting genetic mutations that are consistently found in specific signalling pathways in tumour cells. However, there is now a renewed emphasis on cancer as a disease intimately associated with altered metabolism. Compelling evidence supports that deregulated metabolism such as the well-described Warburg effect (aerobic glycolysis), is a hallmark of many cancers including HCC. Thus, the cancer phenotype is emerging as a complex interplay of genetic mutations, epigenetic deregulation, and metabolic reprogram­ming.

To better understand the evolutionary forces driving HCC, a near-saturating transposon mutagenesis screen in a mouse HBV model of HCC was performed. This screen identified 21 candidate early stage drivers and a very large number (2,860) of candidate later stage drivers that were enriched for genes that are mutated, deregulated or functioning in signalling pathways important for human HCC, with a striking 1,199 genes being linked to cellular metabolic processes. Quantitative changes in pyruvate metabolism as tumours develop were detected by hyperpolarized carbon-13 magnetic resonance imaging in vivo. This study showcases the multiplier effect gained in understanding HCC development by providing a comprehensive overview of the genetic landscape of HCC together with advanced metabolic MRI.

We also investigated the early metabolic changes in the liver at pre-neoplastic stage and during its malignant transformation in HCC rodent models. These models were either individually or synergistically comprised of AFB1 and the hepatitis component. Importantly, this model recapitulates the human HCC pathology. We identified acylcarnitine as a biomarker that can be detected with 1H MRS prior to the tumor formation. Acylcarnitine concentration increases with tumor growth in HCC mice, indicating elevated metabolic activity and increased cell-turnover of the tumor cells [33].

We believe that an understanding of the functional integration of the signalling pathways relevant to liver cancer with the altered metabolic network will open the door to new concepts and therapeutic approaches. Delineation of this functional integration can be further utilized to identify drug candidates and drug combinations that work at the level of genetic alterations and metabolic reprogramming in order to achieve synthetic lethality.