Endothelial hyaluronan synthase 3 aggravates acute colitis in an experimental model of inflammatory bowel disease

Highlights

• Systemic pharmacological blockade of hyaluronan synthesis by 4-Methylumbelliferone exacerbates acute colitis in the DSS model.

• Mice with selective deletion of hyaluronan synthase 3 (Has3) in endothelial cells are protected against DSS-induced colitis.

• Cell-type specific interference with HAS3-mediated HA synthesis may be further explored as potential treatment for human inflammatory bowel disease.

Abstract

The association between hyaluronan (HA) accumulation and increased inflammation in the colon suggests that HA is a potential therapeutic target in inflammatory bowel disease (IBD). However, whether patients with IBD would benefit from interference with HA synthesis is unknown. Here, we used pharmacological and genetic approaches to investigate the impact of systemic and partial blockade of HA synthesis in the Dextran Sodium Sulfate (DSS)-induced colitis model. To systemically inhibit HA production, we used 4-Methylumbelliferone (4-MU), whereas genetic approaches included the generation of mice with global or inducible cell-type specific deficiency in the Hyaluronan synthase 3 (Has3). We found that 4-MU treatment did not ameliorate but exacerbated disease severity characterized by increased body weight loss and enhanced colon tissue destruction compared to control mice without colitis. In contrast, global Has3 deficiency had a profound protective effect as reflected by a low colitis score and reduced infiltration of immune cells into the colon. To get further mechanistic insight into the proinflammatory role of HAS3, we deleted Has3 in a cell-type specific manner. Interestingly, while lack of Has3 expression in intestinal epithelial and smooth muscle cells had no effect or was rather proinflammatory, mice with Has3 deficiency in the endothelium were strongly protected against acute colitis. We conclude that endothelium-derived HAS3 plays a critical role in driving experimental colitis, warranting future studies on cell type-specific therapeutic interference with HA production in human IBD.

Introduction

Inflammatory bowel disease (IBD) is a chronic inflammatory disorder of the gastrointestinal tract comprising two major forms, Ulcerative Colitis (UC) and Crohn's Disease (CD). Prevalence rates of IBD are globally rising with an estimated 6.8 million people affected worldwide in 2017 [1]. IBD can be a debilitating condition to live with; the multitude of symptoms include abdominal discomfort, diarrhea, bloody stool, fever, fatigue and weight loss [2]. In addition, patients with IBD are at increased risk for developing colorectal cancer [3]. While it is widely recognized that a complex interplay between environmental factors, genetics and dysregulated immune responses results in compromised gut barrier function and chronic inflammation of the colon, the etiology of IBD is incompletely understood and there is no cure [4,5].

Interference with proinflammatory cytokine pathways such as the TNF-α, the IL-12/ IL-23 and the IL-6 pathway has been among the most successful therapeutic strategies to date; however, despite considerable success, unsatisfactory response rates and sometimes intolerable side effects highlight the need for the development of novel drugs [5,6].

Like many human pathologies including vascular diseases, fibrosis and cancer [7], [8], [9], IBD is characterized by dysregulation of components of the extracellular matrix (ECM). In search of novel therapeutic options, hyaluronan (HA), a major constituent of the ECM, has emerged as an attractive target molecule. HA is a linear polysaccharide of alternating units of D-glucuronic acid and N-acetyl-D-glucosamine, synthesized at the plasma membrane by three highly homologous isoenzymes (HA synthase (HAS)1-3) and extruded through porelike structures into the extracellular space [10]. In humans, HA is present in virtually all tissues in the body with high concentrations found in the umbilical cord, in synovial fluid, the vitreous of the eye and the dermis of the skin [11]. In addition to its role in providing tissue stability and hydration, HA is critically involved in many physiological processes including organ development, wound healing and angiogenesis [12], [13], [14]. At homeostasis, HA is synthesized predominantly as a high molecular weight molecule of over 1000 kD, however, under inflammatory conditions, HA is degraded into smaller fragments by reactive oxygen and nitrogen species and the activity of hyaluronidases (HYALs) [11]. Functionally, low molecular weight HA has been associated with proinflammatory effects due to activation of immune cells, although LPS contamination of HA preparations has been reported a confounding factor [13].

Importantly, HA accumulates in the serum or tissue of many human disorders including various cancers [15], autoimmune [16] and chronic inflammatory diseases like atherosclerosis [17] or IBD [18]. In the setting of IBD, seminal work from the group of de la Motte has shown that both in tissue sections from the inflamed colon of patients with IBD [18] and in the colon of mice with experimental colitis [19], HA deposition was increased. Interestingly, HA accumulation appeared to precede the influx of inflammatory cells into the colon [19], fueling the hope that targeting HA in IBD may interfere with early events of disease pathogenesis. Aside from this appealing thought, several studies also suggest that reduction of HA deposition may be beneficial in established disease. This is due to the potent immunomodulatory potential of HA through its receptors, in particular CD44 and RHAMM (Receptor for Hyaluronan Mediated Motility). As the principal receptor of HA, CD44 is expressed on many hematopoietic and non-hematopoietic cell types, regulating leukocyte adhesion to the endothelium of blood vessels, thereby controlling immune cell extravasation into inflamed tissue [20]. Notably, mice lacking the CD44 variant v7 in macrophages were shown to be protected against experimental colitis [21]. However, so far no anti-CD44 targeted therapy has been approved for clinical use, not least due to significant toxicity observed in many studies [22].

In contrast, HA synthesis can be inhibited by the clinically established therapeutic 4- Methylumbelliferone (4-MU), a coumarin-derivate mostly used to treat biliary dyskinesia [23]. 4-MU reduces HA production via several mechanisms; first, as a competitive substrate for UDP-glucuronosyltransferase (UGT), an enzyme involved in HA synthesis, and second, by reducing mRNA expression of the HA synthases and the enzymes UDP-glucosepyrophosphorylase and dehydrogenase [23]. The HA reducing and disease ameliorating effect of 4-MU has been demonstrated in multiple animal models of human disease including cancer [23], infectious [24,25] and non-infectious diseases [26,27] and autoimmunity [28,29]. Based on those studies, we decided to test 4-MU as treatment for experimental colitis. In addition, being aware that systemic blockade of such a widely distributed ECM component like HA bears the risk of unwanted side effects, we also aimed to target HA in a more selective manner. For this purpose, we employed transgenic mice competent to produce HA but lacking global or cell-type specific expression of Has3. We focused on HAS3 because work from our group had suggested a special role for this HAS isoenzyme in vascular pathology, an important hallmark of IBD [30,31]. Specifically, we showed that HAS3-mediated HA synthesis in vascular smooth muscle cells enhances neointimal hyperplasia [32] and that Has3 deficiency modulates the immune cell response in a mouse model of atherosclerosis [33]. Moreover, we demonstrated a role of HAS3 in endothelial cell function in a model of hind limb ischemia (accepted manuscript ATVB/2020/315478DR29), where in a severe form of the disease, an ongoing study in our laboratory suggests that HAS3 promotes tissue infiltration of neutrophils and macrophages (unpublished data). Therefore, since we have found HAS3 to affect major cell types which are also critical during colitis development, we decided to selectively delete Has3 in smooth muscle, endothelial and intestinal epithelial cells.

To our knowledge, the present study is the first to address the impact of cell-type specific interference with HA synthesis in the setting of a chronic inflammatory disorder.