HA! The sweet smell of progress with hyaluronan and heart failure

In this issue of Hypertension, Hackert et al make the initial observation that hyaluronan synthase 2 (HAS2) expression is increased in the heart in response to transverse aortic constriction (TAC), a well-established experimental model of cardiac fibrosis and HF. To evaluate the role of hyaluronan synthesis in adverse cardiac remodeling, the authors use 4-methylumbelliferone (4-MU), a naturally occurring coumarin that inhibits hyaluronan deposition purportedly through depletion of UDP-GlcA, which is required for hyaluronan formation. Incidentally, coumarin is responsible for the sweet smell of hay and can be found in bison grass vodka. The authors demonstrate that preventive oral treatment of C57BL/6 mice with 4-MU markedly attenuates adverse cardiac remodeling in response to cardiac pressure overload induced by TAC. 4-MU reduced perivascular and interstitial LV fibrosis, cardiomyocyte hypertrophy and the numbers and frequency of innate immune cells present in the stressed hearts. Not surprisingly, LV contractile function was improved in 4-MU treated mice compared to control¹.

Heart failure (HF) is a leading cause of hospitalization and death, which affects over 25 million people worldwide². Cardiac fibrosis, defined as the excessive deposition of extra cellular matrix (ECM) in the myocardium, and cardiac inflammation due to immune cells infiltration in the myocardium, are both hallmarks of HF. Evolutionarily, the fibrotic and inflammatory responses have worked in tandem to facilitate healing after tissue destruction and as a result of injury or pathogen clearance. However, these responses in a non-regenerative organ as the heart, do not fully resolve and contribute to myocardial stiffness, which impacts contractility, relaxation, and conduction of the heart and results in HF, even in cardiac conditions associated with chronic inflammation that are not directly caused by overt tissue injury or infection.

The cardiac ECM consists of a panoply of molecules that provide structural support and maintain cardiac homeostasis. These include collagen, fibronectin, proteoglycans, and hyaluronan (HA). hyaluronan is a broadly produced, non-sulfated glycosaminoglycan composed of repeating disaccharide of D-glucuronic acid and N-acetyl-D-glucosamine that form polymers of different molecular weights. The high molecular weight hyaluronan is present in intact tissues and organs and is thought to be immunosuppressive, and maintain tissue hydration, whereas the smaller molecular weight hyaluronan, a product of degradation by hyaluronidases, is thought to induce inflammation³. However, abundant conflicting literature exists. In the heart, hyaluronan synthesis is presumably via cardiac fibroblasts, and, to a lesser extent, by resident macrophages and endothelial cells. Hyaluronan and ECM changes occur during fibroblast activation.

Alongside fibrotic and ECM remodeling, a coordinated innate and adaptive immune response results in the sequential infiltration of myeloid cells and T lymphocytes in the heart, which sustain cardiac inflammation and fibrosis. Release of pro-inflammatory and pro-fibrotic cytokines and growth factors, as well as direct cell-cell communication, all contribute to ECM deposition. Hyaluronan can, in-turn, further induce the production of pro-fibrotic cytokines and collagen by eliciting trans-differentiation of cardiac fibroblasts⁴. All these factors take part in the complexity of adverse cardiac remodeling and HF, and hyaluronan sits in the midst of this complexity as an active player that bridges the inflammatory and fibrotic response to cardiac stress.

From the basic science perspective, this work provides insights into the role that hyaluronan plays in cardiac inflammation and heart disease. But perhaps the most significant aspect of this study is that it provides a platform for additional mechanistic studies that may contribute to set the stage to therapeutically using 4-MU in non-ischemic HF, given that 4-MU is an already approved as the drug hymecromone for liver and autoimmune diseases⁵.

The authors report that the preserved cardiac function, decreased fibrosis and cardiomyocyte hypertrophy observed when mice are treated with 4-MU, compared to when they receive vehicle controls, are due to decreased infiltration and proliferation of macrophages in the heart, and thus, decreased inflammation. The conclusions are based on data from flow cytometric and histological analysis, gene expression levels of chemoattractant and BrdU incorporation in macrophages in vivo. The striking inhibition of cardiac fibrosis and cardiac hypertrophy in mice treated with 4-MU is attributed to effects of 4-MU on macrophage recruitment and proliferation, and the authors discuss the possibility that 4-MU may exert direct effects on cardiomyocytes or fibroblasts directly, something that has not been evaluated in the current study. They report that lymphocyte infiltration is not impaired by 4-MU, but the analysis is performed in pooled B and T lymphocytes, and it is therefore possible that B cells, present in high numbers in the resting heart⁶, may obscure the T cell presence. If this was the case, the observed decreased collagen deposition induced by 4-MU could indeed be as a result of decreased T cell infiltration and interactions with cardiac fibroblasts, a step that is preceded by macrophages⁷, the cell type elegantly showed to be altered by 4-MU in this study. Additionally, CD44, a hyaluronan receptor, is expressed broadly in the heart by many different cell types, including neutrophils and T helper 1 cells, which use it as an endothelial selectin ligand that facilitates their adhesion to the vascular endothelium during extravasation to sites of inflammation^8,9. One intriguing possibility is that once in the heart, Th1 cells are “kidnapped” through interactions with hyaluronan, having an opportunity to establish cell-cell communication with cardiac fibroblasts. Thus, 4-MU may prevent these associations and limit cardiac fibrosis, or acts directly on fibroblasts, endothelial cells or myocytes beyond the direct effects on macrophage proliferation and invasion.

There are, however, several important caveats to consider regarding the translational potential of the present findings. The mechanism of inhibition of hyaluronan synthesis by 4-MU is not resolved. It may work by inhibiting key transferases that form the sugar donors for hyaluronan, or 4-MU may function as a sacrificial scavenger of UDP-glucuronate, or other mechanisms altogether. Whichever mechanism may be true, 4-MU does not appear to directly inhibit hyaluronan synthase. Hence, potential reductions in hyaluronan would not be the only effect of 4-MU, and because of the ubiquitous nature of hyaluronan production, the impact of 4-MU would be expected in most organs. In addition, the dose used in the present study is ~500 times higher than doses used in humans, and the present study uses a prevention, but not a therapeutic strategy. Can 4-MU reverse adverse cardiac remodeling once it has developed? Therapeutic studies using the same experimental model will provide new insights that may position the use of hymecromone one step closer to the possibility of being used in patients with HF.

If, however, we take the conclusions at face value, a multitude of questions remain. First, the cellular source of hyaluronan is not known in the heart. It is presumed to arise from fibroblasts; however, this question is not resolved in the failing heart. In addition, the factors regulating hyaluronan accumulation in both the healthy and failing heart are not known. Another important issue that is relevant to hyaluronan biology but not addressed in the present study is the relative size of hyaluronan. Nascent hyaluronan is typically in the MDa (i.e., 10⁶ Da) range; catabolism of this higher molecular weight hyaluronan occurs through physical degradation or the action of a growing number of hyaluronidases. As mentioned earlier, various sizes of hyaluronan impact immune cells differently. Other important questions include how changing hyaluronan abundance in the ECM may change the content of binding partners, such as hyaladerins, or prominent hyaluronan receptors besides CD44, such as LYVE-1 (lymphatic vessel endothelial hyaluronan receptor 1) and HMMR (hyaluronan-mediated mobility receptor).

Once these and other questions are answered, we could conceive of a new generation of agents that target hyaluronan-metabolizing enzymes. Ideally, these strategies would be tissue specific, rather than global, such as 4-MU. In addition, such a new generation of hypothetical agents would more directly target hyaluronan production. But would such a strategy prove successful? Limiting hyaluronan production could remove a potential metabolic “relief valve”. And in the context of the failing heart, interventions that impinge on aspects of intermediary metabolism should be considered carefully. What if, however, we were to target the enzymes responsible for degradation of hyaluronan—increase or decrease their activity, if needed? Perhaps such a strategy could be pursued in addition to strategies such as 4-MU. Either way, Hackert et al remind us that there is a sweet smell of progress with 4-MU and biologically active sugars—we just need to make HA while the sun shines.