Saturday, 10 January 2015

Mapping Metabolic Memory

Some experiences are harder to forget than others. 
The genome's ability to adapt to environmental changes also has a dark side. Emerging research is beginning to unravel an extraordinary phenomenon of cellular memory centred on the field of chromatin biology. One where periods of challenging metabolic conditions can initiate programs of gene expression that dramatically increase the risk for disease later in life. 

Your genome remembers the past

Living cells respond to environmental variations with rapid and stable adjustments to gene expression that are fundamental to many aspects of survival, regeneration, and metabolism. The ability to sense and adapt to local changes in nutrient availability, toxicity, stress, and other environmental factors centres on the gene-regulating dynamic assembly of DNA and histone proteins called chromatin.

Many chemical modifications occurring both on DNA and histones, such as methylation and acetylation, can influence if a gene is 'switched on' or 'off'. Specifically, they can determine whether or not the cellular machinery can physically access the biological information encoded in a gene by controlling transitions between relaxed and condensed chromatin configurations. By several recently discovered mechanisms, these modifications to the chromatin template sensitise the genome to changing environmental conditions.

Considerable interest surrounds the direct connections between metabolism and chromatin structure dynamics. Numerous chromatin-modifying enzymes and complexes respond to various metabolic signals. In particular, several components of the epigenetic machinery utilise intermediates of cellular metabolism for their enzymatic function.

For example, addition of epigenetically important methyl groups (CH3) to DNA and histones is influenced by dietary intake of methyl-donors such as folate. Similarly, many chromatin-modifying enzymes can respond to fluctuations in metabolites derived from the processing of fatty acids and glucose for energy. Furthermore, changes to intracellular metabolism can alter the expression of specific histone-modifying enzymes leading to widespread variations in chromatin patterns.

Indeed most modifications are transiently and dynamically written to and erased from chromatin. However it is increasingly appreciated that some modifications can persist beyond the initial metabolic stimulus. Depending on the type of modification and genomic location, such changes can cause a gene to remain activated or silenced, conferring a molecular memory of previous metabolic experiences.

Under certain conditions, maladaptive expression of particular genes contributes greatly to the development and progression of metabolic disease. Understanding the chromatin component of metabolic memory now represents an important conceptual challenge to explain many aspects of metabolic dysfunction.

The importance of first impressions

Our earliest metabolic experiences can have profound effects on health in adulthood.

Groundbreaking clinical observations associating low birth weight with risk for coronary heart disease later in life led Professor David Barker to propose the developmental origins of disease hypothesis in 1989 (often called the Barker hypothesis). Adaptive responses to the intrauterine environment, including over- and under-nutrition, in anticipation of post-natal conditions are hypothesised to confer increased risk of chronic disease in adulthood.

Because chromatin modifications established in early development can respond to local changes in nutrient availability, and in some cases are remarkably stable, gestational cues have enormous capacity to define and direct future gene expression. Importantly, the chromatin landscape of genes involved in metabolism can be altered by conditions in utero. And it is these changes that are speculated to underlie the elevated risk of developing disease.

A recent Austrian population study described a massive increase in the risk for developing diabetes in people born during and immediately after three major famines during the last century. Similar risk for metabolic disease was also observed in a cohort of individuals conceived in the winter famine of 1944-45 in Nazi-occupied Holland.

Among the best characterised epigenetic marks is the addition of methyl groups to the DNA, which stably modifies gene expression. DNA methylation is a key regulator of genomic imprinting. However its role in the development and progression of metabolic disease has only recently begun to emerge. Importantly, researchers examining the Dutch cohort were able to link maternal starvation around the time of conception with altered DNA methylation patterns in blood cells 60 years later. Specifically, the level of DNA methylation at the maternally-imprinted insulin-like growth factor 2 gene (IGF2) was reduced in people born during this period compared to their unexposed, same-sex siblings. Similar epigenetic changes at this growth-promoting gene were observed in rodent livers from models of intrauterine growth restriction and low protein diet as well as other metabolically important genes in the hippocampus, the pancreas, and skeletal muscle.

Indeed most findings linking transient environmental conditions in gestation to chromatin changes in adult disease are derived from experimental animal models of maternal dietary modulation. With important implications at the level of public health, studies have begun to link maternal over-nutrition with chromatin changes associated with adult disease. 

As an organ that is crucial to fat metabolism, the liver has drawn considerable attention in studies of maternal high fat diets. For example, DNA methylation-dependent silencing of the fatty acid desaturase 2 gene (Fads2) was observed in livers of adult offspring born to rats fed a diet of high fat content during pregnancy. Similarly, transcriptional silencing of a gene associated with mitochondria biogenesis, Pcg-1α, in mice born to dams fed a high fat diet was sustained in adulthood by a DNA-methyl-dependent mechanism.

Other dietary factors are also known to influence gestational epigenetic programming. As the universal substrate for methylation reactions, cellular concentrations of S-adenosyl methionine are fundamental to some of the most prevalent chromatin modifications. The importance of dietary methyl donors such as folate has been examined in rodent models as well as in humans, and the influences of maternal dietary intake are a key focus of recent studies. In fact the effects of maternal folate diet are quite complicated, with both low and high intake associated with increased disease risk in adult offspring. Strikingly, standard recommended maternal daily folate intake of 0.4mg/day was associated with decreased birth weight and altered DNA methylation at the IGF2 gene.

Finally, persistent changes to the expression  of chromatin modifying enzymes are implicated in chromatin modifications induced by maternal dietary fat over-consumption. The livers of non-human primates and rodents exposed to high fat during development exhibited changes in the expression of enzymes involved in the removal of histone acetylation in parallel with a distinctly altered chromatin landscape at genes important for fat metabolism.

The glucose legacy

This remarkable ability of chromatin to incorporate environmental information in to its program of gene regulation also holds tremendous importance beyond early developmental periods. The deleterious effects of periods of poor metabolic conditions in adulthood can persist well beyond the initial exposure. This is particularly relevant to diseases that can remain undiagnosed or unsatisfactorily treated in early stages. The most prominent example in humans is diabetes.

An average of 263 diabetic Australians register each day with the National Diabetes Services Scheme (September 2014). This devastating disease is characterised by loss of insulin producing pancreatic β-cells (type 1 diabetes) or progressive decline in insulin sensitivity and β-cell dysfunction (type 2 diabetes). Because insulin is necessary for the cellular uptake of glucose from the blood and normal metabolic functioning, uncontrolled diabetes leads to elevation of blood glucose concentrations to toxic levels (hyperglycaemia) as well as secondary factors such as high blood pressure and oxidative stress.  

The effects of hyperglycaemia on long-term health are pronounced. They are particularly associated with injury and inflammation of the blood vessels (known clinically as the vasculature). For this reason diabetes remains the leading cause of cardiovascular disease, amputation, kidney impairment, and vision loss in adults.

The deleterious effects of hyperglycaemia are associated with injury and inflammation across multiple organs and tissues. Diabetes is the leading cause of vascular complications in adults leading to cardiovascular disease, limb amputation, kidney impairment, and vision loss. Incorporates image by Mikael Häggström/Wikimedia commonsClick to enlarge

Pioneering studies of animal models first highlighted the difficulties of reversing the damage induced by hyperglycaemia. In 1987 Engerman and Kern reported persistent vascular damage to the retinas of diabetic dogs despite more than two years of insulin therapy and strict blood glucose control. Similarly, pancreatic islet transplantation and subsequent correction of blood glucose levels failed to prevent pathological changes in the kidneys of diabetic rats.

Large-scale epidemiological studies underscore the importance of early and strict metabolic control in preventing vascular complications of diabetes in humans. Moreover, these studies exemplify that early periods of poorly controlled diabetes can instigate the development of complications that present later in the progression of the disease.

The Diabetes Control and Complications Trial (DCCT) compared conventional and intensive regimens of blood glucose control in type 1 diabetes. At the conclusion of the DCCT, the same cohort of patients were all assigned intensive therapy as part of the Epidemiology of Diabetes Intervention and Complications (EDIC) 10-year follow-up study. Throughout this period, vascular complications were significantly reduced in patients that received intensive therapy during the DCCT. On the other hand, patients initially assigned to the conventional treatment group were found to be at significantly higher risk of developing vascular complications despite the tightening of blood glucose during the EDIC phase. Large-scale trials of type 2 diabetes therapies such as the United Kingdom Prospective Diabetes Study similarly point toward a deleterious memory of previous periods of inadequate metabolic control.

A memory of metabolic control. Significantly lower cumulative incidents of retinopathy were observed in patients that received intensive insulin therapy during the DCCT phase, despite normalisation of blood glucose in both groups during the EDIC phase. Click to enlarge

So what is the main driving force behind glucose memory? Clearly it is not a simple matter of poor metabolic or blood glucose control, as even with intensive intervention complications still occur. Further still, it is not simply having the wrong genes, as genetic studies have demonstrated that the DNA code explains only a fraction of the risk for diabetic complications. The most likely explanation is a complex interaction between the cellular environment and genes

Endothelial cells line the inner most surface of blood vessels. Because they are in constant contact with blood and are critical to vascular health, endothelial cells are a useful starting point to examine metabolic memory.  

Seminal findings published in 2008 and 2009 were the first to implicate persistent chromatin modifications in the sustained activation of inflammatory genes in endothelial models of glycemic variability. Professor Assam El-Osta and colleagues at Baker IDI Heart and Diabetes Research Institute in Melbourne demonstrated that the high glucose-mediated induction of a gene critically associated with vascular inflammation persisted in cultured human endothelial cells despite restoration of normal glucose conditions. Moreover, a specific histone methyl modification written by the Set7 lysine methylase was found to be essential to this metabolic memory.

Glucose memory in vascular cells. Cultured human vascular endothelial cells transiently exposed to high glucose concentrations exhibited strong activation of inflammatory genes that persisted despite restoration of normal glucose conditions. Click to enlarge

The same research group reported that restoring normal glycemic conditions per se does not reduce atherosclerosis in animal models of diabetes, and also that transient induction of hyperglycaemia in vivo was associated with persistent up-regulation of genes implicated in vascular disease, resulting from chromatin changes.

Currently it is unclear if transient spikes in blood glucose are sufficient to confer a vascular legacy in human diabetes, and the level of hyperglycaemic exposure required to establish future cell memories is yet to be clearly defined.

Very recently however, the same Set7-driven chromatin signatures were observed in peripheral blood monocytes isolated from patients with type 2 diabetes, providing strong clinical evidence for the importance of this enzyme in diabetic vascular disease. Similarly, epigenome-wide analysis recently revealed substantial increases in monocyte histone acetylation in the conventional treatment group compared to the intensive treatment group subjects of the DCCT/EDIC, indicating a further possible mechanism of metabolic memory.

Future strategies to counter memories of the past

Modification of the chromatin template can sensitise the genome to changing environmental conditions to establish diverse functional states of gene expression. Despite recent advances, scientists are only beginning to scratch the surface of this evidently complex system. The rapidly increasing prevalence of obesity and metabolic dysfunction underscores the imperative challenge of mapping the metabolically responsive chromatin landscape.

There is now strong evidence that chromatin-dependent gene regulation underlying phenotypic determinants of adult metabolic health is influenced by maternal and early postnatal diet. Numerous studies emphasise the importance of dietary balance during development and researchers are increasingly focussing their attention on this important period of foetal development and nutrition. In particular the past few years have seen a surge in the number of published studies examining the epigenetic effects of maternal diet. What we learn from historical human famine cohorts and experimental animal models is pivotal to our ability to cope with the after-effects of recent, current, and future famines around the globe. Of equal importance to public health are the findings that link maternal fat over-consumption with future metabolic dysfunction.

This adaptive agency is also known to influence gene expression in adulthood, where chromatin-dependent mechanisms drive persistent pathophysiological changes. Vascular complications remain the major cause of mortality and morbidity in diabetes with increasing evidence that prior glycemic exposure is a major determinant of susceptibility and progression of these disorders. Despite the promise of the genetic revolution, progress using genome-wide association approaches has thus far been limited. It is becoming increasingly clear that genetic factors simply do not explain diabetes susceptibility and its clinical sequelae. 

Generation of epigenetic maps using newly developed deep-sequencing technologies will greatly advance our understanding of gestationally-programmed chromatin changes as risk factors for adult disease. Equally, accurately defining the molecular events underlying diabetic complications will offer new strategies and specific targets for the development of therapies to prevent, retard, or reverse the long-term deleterious end-organ effects of chronic, intermittent, and prior hyperglycaemia.

To this end, molecular therapeutic approaches aimed at directly inhibiting enzymes that modify DNA and histones have shown promise in various pre-clinical models of disease. However several key challenges delay a comprehensive understanding and impede the efficacy of such treatments.

Distinguishing persistent and transient responses to environmental change remains a key challenge to defining metabolic memory. Precisely how are metabolic changes signalled to the chromatin modifying machinery? Specific intermediates of cellular metabolism such as acetyl-CoA, NAD+, and α-ketoglutarate are mechanistically associated with gene regulation by their use in chromatin-modifying reactions. Fluctuations in the cellular concentrations of these metabolites could therefore connect the chromatin with environmental signalling.

Furthermore the chromatin modifying machinery comprises a multifaceted network of enzymes capable of modifying both the chromatin as well as each other. Understanding the metabolically contextualised effects of inter-enzymatic modifications on the chromatin holds immense promise for greater understanding of environment-gene interactions and the development of effective intervention strategies to treat persistent metabolic dysfunction.

These emerging concepts open new perspectives to combat the rising global epidemic of metabolic disorders. The extent that pathological chromatin changes can be reversed remains controversial. However novel research approaches are aspiring to ensure that our metabolic past is not our destiny.

References and further reading

Originally published in The Australia Times Science magazineVol.3 No.1 (January 2015)
  1. Keating ST & El-Osta A, (2013) Glycemic memories and the epigenetic component of diabetic nephropathy. Current Diabetes Reports 8(4):361-72
  2. Keating ST & El-Osta A, (2013) Epigenetic changes in diabetes. Clinical Genetics 84(1): 1-10
  3. Barker DJ et al., (1989) The intrauterine and early postnatal origins of cardiovascular disease and chronic bronchitis. Journal of Epidemiology and Community Health 43(3): 237-40
  4. Thurner S et al., (2013) Quantification of excess risk for diabetes for those born in times of hunger, in an entire population of a nation, across a century. PNAS 110(12): 4703-7
  5. Heijmans BT et al., (2008) Persistent epigenetic differences associated with prenatal exposure to famine in humans. PNAS 105(44): 17046-9
  6. Gong L et al., (2010) Gestational low protein diet in the rat mediates Igf2 gene expression in male offspring via altered hepatic DNA methylation. Epigenetics 5(7): 619-26
  7. Hoile SP et al., (2013) Maternal fat intake in rats alters 20:4n-6 and 22:6n-3 status and the epigenetic regulation of Fads2 in offspring liver. Journal of Nutritional Biochemistry 24(7): 1213-20
  8. Laker RC et al., (2014) Exercise prevents maternal high-fat diet-induced hypermethylation of the Pcg-1α gene and age-dependent metabolic dysfunction in the offspring. Diabetes 63(5): 1605-11
  9. Vanhees K et al., (2014) You are what you eat, and so are your children: The impact of micronutrients on the epigenetic programming of offspring. CMLS 71(2): 271-85
  10. Steegers-Theunissen RP et al., (2009) Periconceptional maternal folic acid use of 400 microg per day is related to increased methylation of the IGF2 gene in the very young child. PLoS One 4(11): e7845
  11. Aagard-Tillery KM et al., (2008) Developmental origins of disease and determinants of chromatin structure: Maternal diet modifies the primate feral epigenome. Journal of Molecular Endocrinology 41(2): 91-102
  12. Suter MA et al., (2012) A maternal high-fat diet modulates feral sirt1 histone and protein deacetylase activity in nonhuman primates. FASEB 26(12): 5106-14
  13. Strakovsky RS et al., (2011) Gestational high fat diet programs hepatic phosphoenolpyruvate carboxykinase expression and histone modification in neonatal offspring rats. Journal of Physiology 589(11): 2707-17
  14. Suter MA et al., (2014) In utero exposure to a maternal high-fat diet alters the epigenetic histone code in a murine model. American Journal of Obstetrics and Gynecology 210(5): 463
  15. Engerman RL & Kern TS, (1987) Progression of incipient diabetic retinopathy during good glycemic control. Diabetes 36(7): 808-12
  16. Gøtzsche O et al., (1981) Irreversibility of glomerular basement membrane accumulation despite reversibility of renal hypertrophy with islet transplantation in early experimental diabetes. Diabetes 30(6): 481-5
  17. White NH et al., (2008) Prolonged effect of intensive therapy on the risk of retinopathy complications in patients with type 1 diabetes mellitus: 10 years after the Diabetes Control and Complications Trial. Archives of Ophthalmology 126(12): 1707-15
  18. Aschner P et al., (2010) Practical steps to improving the management of type 1 diabetes: recommendations from the Global Partnership for Effective Diabetes Management. International Journal of Clinical Practice 64(3): 305-15
  19. Holman RR et al., (2008) 10-year follow-up of intensive glucose control in type 2 diabetes. The New England Journal of Medicine 359(15): 1577-89
  20. El-Osta A et al., (2008) Transient high glucose causes persistent epigenetic changes and altered gene expression during subsequent normoglycemia. The Journal of Experimental Medicine 205(10): 2409-17
  21. Brasacchio D et al., (2009) Hyperglycemia induces a dynamic cooperativity of histone methylase and demethylase enzymes associated with gene-activating epigenetic marks that coexist on the lysine tail. Diabetes 58(5): 1229-36
  22. Paneni F et al., (2014) Adverse epigenetic signatures by histone methyltransferase Set7 contributes to vascular dysfunction in patients with type 2 diabetes. Circulation. Cardiovascular Genetics [Epub ahead of print]
  23. Miao F et al., (2014) Evaluating the role of epigenetic histone modifications in the metabolic memory of type 1 diabetes. Diabetes 36(5): 1748-62