Modulation of Hyaluronan Signaling as a Therapeutic Target in Human Disease

Abstract

The extracellular matrix is an active participant, modulator and mediator of the cell, tissue, organ and organismal response to injury. Recent research has highlighted the role of hyaluronan, an abundant glycosaminoglycan constituent of the extracellular matrix, in many fundamental biological processes underpinning homeostasis and disease development. From this basis, emerging studies have demonstrated the therapeutic potential of strategies which target hyaluronan synthesis, biology and signaling, with significant promise as therapeutics for a variety of inflammatory and immune diseases. This review summarizes the state of the art in this field and discusses challenges and opportunities in what could emerge as a new class of therapeutic agents, that we term “matrix biologics”.

1. Introduction

The extracellular matrix (ECM) is not simply an inert filler of intercellular space; on the contrary, the ECM is an active participant, modulator, and mediator of cell fate, signaling and injury response. In a way, the ECM and the cells constitute a microscopic “host-environment” unit which ultimately dictates the macroscopically observed host-environment interactions that can be observed after injury and in disease processes. ECM biology is complex, derived in part from a dynamic mix of ECM components. One such prominent ECM component is hyaluronan (HA), a ubiquitously expressed glycosaminoglycan sugar. HA structure is deceptively simple, consisting of repeating disaccharide chains of N-acetyl-glucosamine and glucuronic acid, however recent research has shed light on its fascinatingly complex biology and has strongly suggested that HA plays a central role in fundamental biological processes that ultimately dictate homeostasis or disease development. Such processes include cell and organ development, tissue injury response, inflammation (including autoimmunity), cell migration, cancer formation and cancer resistance. Recent research has shed considerable light into the mechanistic effects of HA biology, which seems to be regulated by four major mechanisms: 1) physical properties, size and molecular weight of HA; 2) HA binding partners in a matrix and the formation of a macromolecular structure; 3) HA metabolism (synthesis and degradation); and 4) HA receptor signaling (1). These insights, in turn, have enabled the translation of HA biology into the human health arena, including the ability to intelligently design or modify HA molecules to specific therapeutic ends.

There are obvious benefits to using HA in medicine: it is naturally occurring, easily biodegradable, non-antigenic and highly hydrophilic. Consequently, HA has been in wide use in cosmetics (dermal filler), orthopedics (intra-articular injections in degenerative disease) and ophthalmology (as a filler of the corpus vitreum in eye surgery) (2–8). However, most of these applications are utilizing the biocompatibility and physical properties of HA. In this review, we will advance the argument that HA-related applications can be vastly expanded, taking advantage of HA biological (as opposed to, or in addition to its physical) properties. Furthermore, approaches that modify HA biology (e.g. synthesis, interaction with other proteins) can also be viewed as belonging to the same realm of pharmaceuticals. We will give examples of recent experimental evidence of the clinical utility of such approaches and highlight opportunities for further development in the future.

2. Mechanistic background

The following chapters will address the basic mechanistic steps of HA biology which can be targeted in a pharmacological setting: HA synthesis and degradation; HA association with extracellular proteins to form a HA matrix; and HA signaling.

2.1. Hyaluronan synthesis and metabolism

HA is synthesized at the plasma membrane (9) by one of three isozymes, Hyaluronan synthase (HAS) 1, 2, and 3. HAS are evolutionarily conserved, highly homologous (55–70% protein identity) (10, 11), and catalyze the addition of UDP-D-glucuronic acid (GlcA) and UDP-N-acetyl-D-glucosamine (GlcNAC) monomers in an alternating assembly to form HA (12, 13). HAS isoforms are different in their half-life, rate of HA synthesis, affinity for HA substrates (14) and molecular weight of synthesized HA: HAS1 and HAS2 synthesize larger polymers (2×10⁵ to 2×10⁶ Da, HAS1 product size being somewhat lower than HAS2), while HAS3 synthesized the shortest HA polymer sizes (1×10⁵ to 1×10⁶ Da) (14). It is still not clear how mammalian HAS regulate the length of their HA product, in contrast to bacterial HAS, where the length of the final product depends on the availability and stoichiometry of UDP-sugar precursors (15). In any case, HAS enzymes are the first modulator of HA biological activity, by way of producing HA of different sizes, which may have different biological functions, as we will see below.

At the tissue level, HAS gene expression and subsequent HA synthesis is regulated by a wide range of cytokines and growth factors (16, 17). HAS genes may respond similarly or differentially to a particular signal; for example, TGF-β1 induces downregulation of HAS3 but upregulation of HAS1 expression (18). Furthermore, cell- or tissue-specific HAS expression may lead to opposing effects. For example, HAS2 expression in airway epithelia leads to resistance against injury and decrease in experimental fibrosis (19, 20), while HAS2 expression in fibroblasts promotes fibrosis (21, 22). HA synthesis is also under tight metabolic control. For example, the half-life time of HAS2 is only 17 min; however, after the post-translational ligation of a single O-linked-β-N-acetylglucosamine (O-GlcNAc) on the hydroxyl group of its serine 221 residue, HAS2 half-life time is extended substantially to >5 hours (23, 24), thus linking HA synthesis to metabolic processes in the cell and the organism, like hyperglycemia (23). Conversely, HAS2 can be inhibited by phosphorylation by the AMP-activated protein kinase, a master metabolic regulator activated by low ATP/AMP ratios, and epigenetically by sirtuin 1, (SIRT1) another important energetic sensor. The SIRT1 action is very interesting and intricate, in that it suppreses expression of an antisense nucleotide (HAS2-Antisense1), which is known to stabilize HAS2 mRNA and induce HAS upregulation. Therefore, decreased HAS2-AS1 leads to HAS2 downregulation (25). In aggregate, mounting evidence points towards an intricate and complex regulatory network that can influence HA synthesis, underscoring the importance of this molecule for biological processes.

Experimental evidence suggests that HAS genes play important roles in disease and injury. In the naked mole rat, expression of a particularly high molecular weight HA by HAS2 was shown to be the reason behind the longevity and cancer resistance of this rodent (26). In cancer, overexpression of HAS influences tumor growth, metastatic potential, and progression in prostate, colon, breast, and endometrial tumors (27). New evidence suggests that during cancer-induced tissue remodeling, an abundant HA matrix is produced, based on a dramatic metabolic reprogramming of cancel cells, with overabundance of HA substrate saccharides and overexpression of HAS, particularly HAS2 (reviewed in (28). Ectopic expression of HAS genes may also functionally alter the biological responses of cells to injury in vivo (19, 21). Thus, HAS activity may have opposing effects dependent on context and tissue. HAS activity can be chemically inhibited (29, 30) and this can be utilized for therapeutic purposes, as we will see below.

HA turnover is achieved either non-enzymatically or enzymatically. Reactive oxygen species such as superoxide, hydrogen peroxide, nitric oxide, peroxynitrite, and hypohalous acids, either generated during inflammation or introduced via environmental exposures, degrade HA (31, 32). Enzymatically, hyaluronidases specifically induce HA digestion. Several HA degrading enzymes exist in mammals: hyaluronidases 1–3 (HYAL1–3), PH20 (33, 34), the newly discovered CEMIP, also known as KIAA1199 (which may not be directly degrading HA but facilitate HA uptake leading to degradation (35)) and CEMIP2, also known as TMEM2 (36–39). HYALs are highly homologous endoglycosidases, which hydrolyze the β−1,4 linkage of the HA molecule, which is a linear polysaccharide composed of repeating β−1,4-linked D-glucuronic acid (GlcA) and β−1,3-linked N-acetyl-D-glucosamine (GlcNAc) disaccharide units (34, 40, 41). HYAL activity is regulated by environmental cues, such as acidity: for example, HYAL1–3 are primarily active at an acidic pH supporting their activity in lysosomes, while PH20 optimum activity is at a neutral pH (33, 42) and TMEM2 activity is at pH5–8 (39). TMEM2 is found in all organs, HYAL2 is found in heart, skeletal muscle, colon, spleen, kidney, liver, placenta, and lung, while HYAL1 is more limited in scope but is high in liver (which is a primary location of HA degradation), while also found in plasma and urine. PH20 is specific to the acrosome, an organelle of sperm cells and has a role in fertilization (33). HA degradation is accomplished by TMEM2, HYAL1 and HYAL2 acting in concert. TMEM2 and HYAL2 appear to be active on the cell membrane, possibly in an organ- and cell-specific fashion. The acidic environment necessary for HYAL2 activity is provided by NHE2 (43). Thus generated HA fragments are internalized into endolysosomes, where HYAL1 further degrades the HA into tetrasaccharides (44).

Perturbations in HA catabolism have significant biological effects. Hyaluronidase deficiency in humans leads to elevated HA levels in plasma (45) and a lysosomal storage disorder called mucopolysaccharidosis IX characterized by accumulation of HA, short stature, and multiple soft tissue masses in the joints (46). In mice, HYAL2 deficiency also leads to craniofacial and hematological defects (47, 48). CEMIP deficiency enhanced the inflammatory response to cutaneous infection (49). On the other end of the spectrum, local overexpression of HYAL1 in mouse skin induced dendritic cell (DC) migration and maturation, which in turn muted the response to contact hypersensitization (CHS) via activation of the innate immune receptor Tlr4 by HA oligosaccharides (50), while overexpression of TMEM2 resulted in increased lifespan, ER stress resistance and infection resistance in the C. elegans model (51). Further, increased hyaluronidase levels have been found in several cancers, and tend to correlate with more invasive and metastatic phenotypes (42).

2.2. Hyaluronan association with extracellular molecules

HA does not exist in the ECM as an isolated molecule. Instead, it has been known for decades that ECM HA is found in covalent binding with other proteins or glycoproteins which have been grouped under the portmanteau term “hyaladherins”. Importantly, many signaling properties of HA only come to be when HA is bound to hyaladherins, but not with free, unbound HA (52, 53). We will discuss 4 major hyaladherins, Inter-α-inhibitor (IαI), Tumor necrosis factor stimulated gene 6 (TSG-6), Pentraxin-3 (PTX3), and Versican, because their interactions with HA have been well described and clear therapeutic targets identified.

IαI is a family of composite proteins comprised of a common light chain and several homologous heavy chains (HCs). The light chain consists of a chondroitin 4-sulfate domain and a core protease inhibitory moiety, bikunin, which lends IαI its name. One bikunin molecule is usually linked to 1 or 2 HCs via a unique ester bond, producing pre-α-trypsin inhibitor (Pαl, HC3-bikunin in humans) and lαI (HC1·bikunin-HC2 in humans) (54–56). IαI are produced in the liver and coupled with bikunin before release into the circulation, but also can be produced in other solid organs including lung, kidney and brain (57, 58). IαI HCs can be bound to HA in a transesterification reaction catalyzed by tumor necrosis factor-stimulated gene 6 (TSG-6), a 35-kDa secreted ECM hyaladherin. TSG-6 is induced by TNFα, and secreted by immune and structural cells (59, 60). Thus, during inflammation, induction of TSG-6, and extravasation (as well as local de novo production) of IαI promote the deposition of a “pathological” HA-HC matrix (61–63).

TSG-6 and IαI demonstrate very interesting paradoxical effects in the pathogenesis of inflammation. Both proteins are increased in diseases such as asthma, COPD, cystic fibrosis, lung fibrosis, atherosclerosis, kidney disease, and many others (58, 61–69). On the one hand, both proteins are necessary for the generation of a pathological HA matrix and the development of airway inflammation and hyperresponsiveness in allergic and TLR4-mediated lung injury (62, 63, 68, 70). On the other hand, both have prominent anti-inflammatory activities in other disease models such as lung inflammation after endotoxin (71) and bleomycin exposure (72), pancreatitis, arthritis and many others (58, 69). TSG-6 is thought to mediate the beneficial effects of stem cells (73–75) in regenerative models. IαI binds to coagulation factor IX (76) and is a strong factor XI inhibitor (77). IαI inhibits plasmin (78) and significantly ameliorates disseminated intravascular coagulation and lung injury in LPS-induced sepsis (79). Furthermore, IαI inhibits complement activation (80–82) and prevents complement-induced lung injury (80). Finally, IαI may directly bind pathogen-associated molecular patterns such as the dengue virus (83), adenovirus and HIV (84, 85), or danger-associated molecular patterns like histones (86, 87) and HMGB1 (88), thus sequestering infectious and danger-associated molecular patterns and inhibiting inflammation.

IαI HC in the HA matrix are crosslinked through PTX3 (89). IαI, TSG-6 and PTX3 are indispensable for the normal development of a functional HA matrix in the cumulus oophorus complex in the ovulating ovary, and their absence results in severe female infertility (89–92). During inflammation, PTX3 can be induced via TLR activation (93) and has a role in injury repair (94), antimicrobial defense (95–102), and allograft survival (103). PTX3 also ameliorates allergic lung inflammation (104), LPS-induced lung injury (like TSG-6 and IαI) (105), and promotes injury repair, fibrin deposition and fibrosis in a lung injury model of acid aspiration (94). In further parallels with IαI, PTX3 can bind viral antigens, such as the influenza hemagglutinin (106) and interacts with the complement, but with conflicting effects on activation (107, 108).

These complex and partly conflicting reports on the role of TSG-6, IαI and PTX3 in inflammation may be best explained if we take into consideration the inflammatory mechanism underlying a given disease or disease model: when pathological HA matrix formation is necessary for the development of disease, for example airway hyperresponsiveness after lung injury or allograft dysfunction (63–65, 70, 109–112), then PTX3, IαI and TSG-6 promote disease by mediating the formation of pathological HA matrix. However, in non-HA-mediated processes (71, 73), PTX3-, TSG-6- and IαI-supported HA matrices may promote an anti-inflammatory role, by ameliorating complement- or coagulation factor-dependent pathways. Generally, acute processes appear to benefit from the accumulation of LMW-HA matrix which promotes inflammation and clearance of the offending agents, while chronic or degenerative processes, which lack homeostasis, may be adversely affected by the accumulation of HA matrix.

HA is also decorated by the proteoglycan Versican, and this serves to further modulate the inflammatory response in a feedback loop (113, 114). Versican is induced by TLR activation, infection or aseptic injury (115–118) and regulates the activity of chemokines (113), the retention of immune cells, and chemotaxis (115, 119). As we have seen with HAS2 above, Versican also has opposing effects depending on cell type: anti-inflammatory (macrophage expression) (117) or pro-inflammatory (mesenchymal cell expression (115, 119–121). Taken together, available research suggests that a HA matrix, including cross-linked TSG-6, PTX3 and IαI HC, and Versican, modulates the immune response to injury in complex and cell-specific ways. Indeed, HA biology cannot be appreciated without including HA interactions with hyaladherins. The hyaluronan matrix is effectively a canvas, which is dynamically decorated by hyaladherins, and it is this interaction that may dictate hyaluronan-specific effects and roles after injury (114).

2.3. Hyaluronan signaling

There are several cell surface receptors for HA such as CD44, RHAMM, HARE, LYVE1, layilin, and the innate immune receptors TLR2, TLR4 and TLR5 (19, 122–132). Importantly, HA size is a crucial determinant for receptor signaling, perhaps because signal transduction is dependent on receptor clustering on the membrane (19, 133). Although there is no generally accepted nomenclature describing HA sizes, for the purposes of this review we will be referring to high molecular weight HA (HMW-HA, 1 million Da or higher, the naturally occurring form), low-molecular-weight hyaluronan (LMW-HA, between 100–500 kDa, occurs during inflammation and tissue injury) and HA oligosaccharides (oHA, 10 HA disaccharides or less, the minimum size able to engage receptors but cannot induce receptor clustering (134). This is probably because longer chains of HA possess multivalent sites for CD44 binding while oHA have only 1 or 2 binding sites (135–137). It is important to stress that HA biological effects depend on HA size. HMW-HA is anti-inflammatory and anti-angiogenic (17), while LMW-HA is pro-inflammatory (138) and pro-angiogenic, promoting tumor progression (133, 139). In general, LMW-HA seems to be pro-inflammatory and rather deleterious while HMW-HA appears to be, by and large, beneficial. Interestingly, oHA of can either stimulate or inhibit inflammation depending on cell type and disease model. oHA can increase angiogenesis during wound healing (140) and promote inflammation in synovial fibroblasts (141), but have been also shown to ameliorate TLR3-mediated inflammation (142), inhibit HA/CD44-mediated activation of intracellular kinases (143, 144), retard tumor growth and sensitize resistant cancer cells to chemotherapeutics (133).

An important aspect of HA signaling is its ability to signal through toll-like receptors (TLRs). LMW-HA engages a receptor complex of CD44 and TLRs and induces cytokine release (145) and airway hyperresponsiveness (AHR) (70, 109) in macrophages and naïve mice, respectively. In bleomycin-induced acute lung injury, LMW-HA and HMW-HA signal through TLR2 and TLR4, but not CD44 (19). Again, HA size appears to be a major determinant of TLR-mediated signaling pathways. In general, LMW-HA induces TLR-mediated inflammation (19, 109, 146), while HMW-HA is inert (or antagonizes LMW-HA effects) (68, 70). However, HMW-HA can also signal through TLR in some contexts, although the effect is not pro-inflammatory. For example, overexpression of HAS2, which produces HMW-HA (14, 147), promotes epithelial resiliency to injury and alveolar progenitor cell renewal through TLR4 and TLR2 activation (19, 20). Thus, both HMW-HA and LMW-HA activate TLR signaling, but their effects differ (inflammation vs. epithelial resilience). The reasons and mechanisms underlying these differences are unclear but may involve different receptor clusters in each cell type. For example, LMW-HA induces activation of dendritic cells via TLR4, independently of TLR2 (123), while it activates macrophages via TLR2 pathway independently of TLR4 (148); in another example oHA activate TLR4 and CD44 in chondrocytes (149–151), but TLR2 and TLR4, and not CD44, in synovial fibroblasts (141). oHA can signal through CD44 or TLRs (152) but not both at the same time (123, 149, 153, 154). A recent study suggested that the N-acetyl groups on HA are important for TLR4 signaling, and that partial de-acetylation or selective butyrylation of HA led to inert or anti-inflammatory molecules along the TLR4 axis (155). However, much remains unknown about HA signaling through TLR, including an understanding of how the signaling occurs, since no one has been able to show direct binding of HA to these receptors. Indeed, recent studies suggest that some of the original findings of TLR-mediated LMW-HA signaling may be due to contamination by the TLR ligands, notably lipopolysaccharide (endotoxin), which activates TLR4. Some reports suggest that endotoxin-free LMW-HA does not have TLR4-activating properties (156); other work has shown that although LMW-HA causes inflammation, TLR4 actually ameliorates this inflammatory response (157). These contradictory findings could be due to differences in cells studied, cell provenance (e.g. which organism), their activation states, and the endpoints that were assayed (152). It may also be that LMW-HA does indeed activated TLR4 under some conditions, but not others, and that some of the action of HA is not direct, but due to either coordinate action by different-sized fragments (158, 159). Finally, it could be that the effects of some HA fragments are indirect: it is possible that differently-sized HA fragments compete for receptor engagement, and it is well understood that clustering of receptors is dependent on HA fragment size (134). Overall, it is probable that different mechanisms apply in different contexts and conditions.

HA signaling is normally initiated at the cell membrane. However, HA has also been detected intracellularly. It may be there as part of a physiologic pathway; for example, HA associates with the mitotic spindle, microtubules, and the receptor RHAMM during mitosis (160, 161). However, there is also the possibility that HA is produced aberrantly during pathological processes; for example, in the setting of hyperglycemia hyaluronan is produced in the Golgi and initiates an endoplasmic reticulum stress and autophagy response (162–166).

3. Therapeutic applications of hyaluronan signaling

In aggregate, the available literature may suggest the following (necessarily simplified) general principles regarding HA signaling: HMW-HA is the prevailing molecule during health and homeostasis, and it promotes cellular resilience and longevity, aided by its hyaladherin matrix. During acute inflammation, HMW-HA is degraded into LMW-HA fragments, which together with the associated hyaladherins promote the inflammatory response with the biological goal of removing the offending agents and restoring homeostasis. The process can be derailed in two major ways: first, chronic inflammation (e.g. in disease) or aberrant inflammation (e.g. through exposure to pollutants) may transform the LMW-HA pathway into a maladaptive response which supports pathology instead of homeostasis, because it fails to resolve; secondly, aberrant cells (e.g. in cancer) may co-opt the HA signaling pathway to aid local survival at the expense of the organism. In such cases, disrupting the LMW-HA signaling pathway, or restoring the HMW-HA milieu would promote disease resolution. This, then, is the guiding principle for therapeutic applications of HA that will be explored in this chapter.

Here, we will first summarize the known associations of HA with human disease and then introduce principles of therapeutic applications for HA signaling. As the preceding chapters suggest, HA biology is far-reaching, and the spectrum of potential therapeutic targets is vast. Therefore, rather than enumerating every possible disease that can be addressed with HA signaling axis, we will focus on mechanistic considerations, based on the above principles, that can guide intelligent design of HA-targeting drugs and careful selection of disease processes, with the ultimate goal to maximize therapeutic benefit and minimize side effect profiles. Clearly, more than one pathway can be invoked for given diseases. For example, inhibiting the formation of HA complexes or HA synthesis will also impair HA signaling downstream of these processes. Nevertheless, we believe that the mechanistic approach is more conducive to a thoughtful drug design process.

3.1. Associations and roles for hyaluronan matrix in human disease

Dysregulation of HA metabolism is a hallmark of many inflammatory and degenerative diseases. Increased levels of circulating or tissue HA have been found in many lung disorders including asthma, COPD, interstitial lung disease, and pulmonary hypertension (167, 168), as well as inflammatory, fibrotic and degenerative diseases like rheumatoid arthritis (169), hepatitis (170, 171), cirrhosis (172, 173), chronic kidney disease (174), atherosclerosis (175), cardiac remodeling in failure (176), diabetic pancreas (177), allograft rejection (110, 178, 179), virtually every type of solid cancer (133, 180–182), etc. In most, or all these diseases, elevated serum levels of HA have been associated with adverse outcomes and worse prognosis. Whenever HA size has been evaluated, it was most commonly LMW-HA. Furthermore, abnormal HAS expression is often found in malignancies and tissue injury and is also associated with clinical outcomes. Hematologic malignancies (monoclonal gammopathy of undetermined significance, multiple myeloma and Waldenström’s macroglobulinemia) and solid cancers (bladder cancer) contain cells with aberrant HAS1 splice variants, which are associated with poor prognosis (183). Increased HAS2 expression also negatively correlates with outcomes in breast, oral, endometrial and brain cancers (184). The prevailing theory is that increased HA expression in the tumor stroma promotes cancer progression and metastasis by enhancing pro-proliferation signaling, promoting evasion of apoptosis, inducing angiogenesis and lymphangiogenesis, promoting invasivity and metastatic potential, reprogramming energy metabolism and aiding immune evasion (184).

Abnormal HA matrix deposition (HA associated with hyaladherins) is observed in many pathological processes associated with tissue remodeling like pulmonary hypertension (185), inflammatory bowel disease (186) and asthma (61). Furthermore, in conditions of endoplasmic reticulum stress cells produce so-called “HA cables” (HA associated with IαI HC, and sometimes TSG-6, that appears in cable-like structures) that originate in the perinuclear area and protrude into the extracellular space, where they create an adhesive matrix entrapping inflammatory cells (187). Versican deposition is increased in airways of patients with idiopathic pulmonary fibrosis (188), severe ARDS (189) and asthma (190, 191). In human lung transplant patients, increased PTX3 levels are associated with graft dysfunction (i.e. ischemia reperfusion injury) (192) and genetic variability in PTX3 that leads to increased serum PTX3 levels, predisposes to allograft dysfunction (103). On the opposite side of the spectrum, genetic variability in PTX3 that leads to decreased PTX3 function, is associated with increased susceptibility to invasive aspergillosis in stem cell transplant patients (95). TSG-6 levels in synovial fluid correlate with disease progression in patients with osteoarthritis (193). Finally, there is a wealth of literature suggesting that circulating levels of IαI are associated with survival in human infections/inflammatory diseases like severe sepsis in adults, necrotizing enterocolitis in newborns and severe dengue in children (194–198).

3.2. Established applications of HA therapeutics in medicine

In this chapter we will briefly summarize established applications of HA in clinical medicine. As mentioned above, many or most of these applications are based on viscoelastic and physical properties of HA, and thus will not be a focus of this review, which is targeting biological translation of HA signaling. Nevertheless, this list highlights and supports the translational potential, safety profile and applicability of HA pharmaceuticals in clinical practice. It should also be noted that the physical and signaling effects of HA are likely to be linked, in that specific HA sizes demonstrate distinct biological and physical behavior that go in parallel. Thus, many future applications of HA will integrate its biology with its physics (199–202), as in, for example, the production of biomaterials that promote stem cell growth (HA biology) along with tissue integration and compatibility (HA physics).

Dermatology:

Perhaps the most common application of HA preparations is in cosmetic dermatology, both used superficially (included in many skin cosmetic creams) and injected as filler for wrinkle or scar treatment. HA-containing gels are used in skin wounds (including burn wounds (203)) and ulcer care (2).

ENT:

Along similar lines, injected HA is used for augmentation or repair of vocal cords. HA sprays are also used successfully as adjunct (hydrating) agents after sinus surgery to promote healing, reduce crusting and expedite post-operative recovery (204, 205).

Ophthalmology:

HA fillers are used in eye surgeries, in particular as replacement of corpus vitreum lost during surgical manipulation in cataract surgery, lens implantation and corneal transplantation. HA is also a very common lubricant in “artificial tear” eye drops.

Rheumatology/Orthopedics:

Rheumatoid and degenerative (osteo-) arthritis are both characterized by loss and degradation of HMW-HA in the synovial fluid and cartilage destruction. Intra-articular injections with HMW-HA have been shown to improve symptoms.

Gynecology:

HA preparations are used in two main applications: either as vaginal suppositories or creams to alleviate symptoms of vaginal atrophy or injury after menopause, surgery, irradiation etc. (206–209), or as surgically applied gels to prevent adhesions after hysteroscopic lysis (210).

Tissue Engineering:

HA biodegradability, biocompatibility and immune tolerance also make it an ideal candidate as a tissue engineering scaffold material. Skin, bone and soft tissue grafts have been engineered with the use of HA, and the range of applications is ever increasing (2, 211, 212).

3.3. Novel applications of HA Therapeutics: Modulation of HA signaling

3.3.1. HA oligosaccharides

Hyaluronan oligosaccharides (oHA) bind monovalently to HA receptors (135) and participate in irreversible binding with ECM molecules like the IαI HC (213), thus disrupting HA signaling. oHA inhibit HA activation of the PI3/Akt pathway and the complex formation between HA receptor CD44 and tyrosine kinases as well as CD44 and Emmprin, and induce PTEN (143, 214–217), thus inhibiting cancer cell growth. The inhibition of CD44/Emmprin association further blocks plasma membrane localization of monocarboxylate transporters MCT1 and MCT4, which results in the inhibition of lactate efflux and induces metabolic reprogramming of the cancer cell (218, 219). Finally, oHA inhibit the association of CD44 with drug transporters ABCB1 and ABCG2, thus decreasing the chemoresistance of tumor cells to chemotherapy drugs (215, 218, 220). These affects raised interest in the use of oHA in cancer. Indeed, studies in preclinical models have demonstrated that oHA can arrest or delay the growth of grafted tumors in vivo (214, 215, 217, 221, 222). Modification of oHA, such as sulfation of their side chains, may add to their activity by inducing inhibition of HA synthesis (223). oHA are attractive because they are small molecules (in general less than 10 disaccharide units) that can be easily formulated and delivered, and, like all other HA molecules, are non-immunogenic.

3.3.2. LMW-HA

Although most literature suggests that LMW-HA fragments are pro-inflammatory, there is some evidence that they can be utilized as therapeutics, primarily from the lung disease literature. LMW-HA promotes motile cilia beating through activation of its receptor RHAMM (224) and induces the expression of mucins (225, 226), thus potentially improving mucociliary clearance. There is a substantial body of literature suggesting that LMW-HA may protect from the development of emphysema, possibly by ameliorating elastic fiber degradation as was shown in animal models of emphysema (227–232). Inhaled LMW-HA was safe in patients with COPD and preliminary data suggested that it was effective in reducing breakdown of elastin fragments even after only 2 weeks of treatment (232, 233). In the intestine, LMW-HA induces the expression of antibacterial defensins and promotes epithelial integrity ((234) more about this below). These results suggest that there is still much to be found about mechanisms of HA signaling depending on organs, and that HA effects in disease (e.g. in COPD) or in conditions of heavy exposure burden (e.g. in the microbiome-filled intestine) may be very different from these seen in isolated cell models or exposures in naïve animals and sterile conditions.

3.3.3. HMW-HA

Preclinical and clinical studies support a role for HMW-HA as a treatment in airway disease (19, 229, 235). In several preclinical models, HMW-HA ameliorates epithelial injury, inflammation and airway hyperresponsiveness (19, 53, 68, 70, 109, 236–242). In allergic asthma, HMW-HA promotes regulatory T-cell activity and suppression of adaptive immunity (238, 239), but in all models HMW-HA can prevent repolarization and calcium flux-mediated contraction of myocytes (243), suggesting utility in preventing airway constriction across diseases. In humans, HMW-HA protects against exercise-induced airway hyperresponsiveness in patients with bronchial asthma (235). In COPD patients, HMW-HA ameliorates severe acute disease exacerbations, reduced the duration of non-invasive ventilation treatment and shortened the hospital length of stay (244).

It is worth noting that inhaled HMW-HA has been used for several years in Europe, especially in patients with cystic fibrosis, with a remarkably good safety profile (245–248) and evidence suggests that it does not undergo fragmentation or promote inflammation (249). Based on these data, HMW-HA can be considered a mature treatment candidate for inflammatory airway disease.

Further applications of HMW-HA in tissue injury and fibrosis are possible, based on available evidence. HMW-HA (induced by the anti-inflammatory cytokine IL-10) improved renal fibrosis after injury (250) and skin injury (251) in a mouse model, while the HA-production inhibitor 4MU was deleterious in these conditions. In another mouse model HMW-HA improved allograft survival by competitively inhibiting the effects of LMW-HA (110). Taken together, these results all point towards HMW-HA as being a powerful candidate agent protecting against epithelial injury, fibrosis and chronic inflammation.

3.4. Novel applications of HA Therapeutics: Inhibition of synthesis (4-Methylumbelliferone)

Inhibition of HA synthesis may be an attractive modality when exuberant HA production is part of the disease pathogenesis. In this context we may benefit from the existence of a well-known drug, 4-methylumbelliferone (4-MU), which is a coumarin derivative that inhibits HA synthesis in two ways. First, by acting in competition with uridine diphosphate (UDP) as a substrate for UDP-glucuronyltransferase (an enzyme involved in HA synthesis), it leads to a reduction in the availability of the UDP-glucuronyltransferase product UDP-glucuronic acid (UDP-GlcUA) in the cytosol. Since the abundance of UDP-GlcUA is one of the determinants of HAS synthetic activity, this in turn leads to a reduced HA abundance. Another, more recently discovered and less well understood effect of 4-MU isthe downregulatiion of HAS expression (252, 253). 4-MU has been in clinical use for over 30 years in the treatment of bile duct disorders (254–256) and is very well tolerated. This may be somewhat counterintuitive, since we just described multiple beneficial effects of HA, which 4-MU inhibits. It should be noted however, that 4-MU is not a complete inhibitor of HA production. Thus, when given in conditions of high LMW-HA production, like inflammation, it is possible that 4-MU preferentially inhibits pro-inflammatory HA expression, while still permitting the expression of native, HMW-HA which may be metabolized at a lower rate.

Aberrant HA signaling has been implicated in several autoimmune diseases, including Type 1 Diabetes mellitus. HMW-HA, in complex with TSG-6 and IαI, is abundant in the extracellular matrix if healthy pancreatic islets, where it may support normal cell function (257, 258). However, HMW-HA and its matrix are degraded to LMW-HA molecules during the development of autoimmune insulitis in humans and animal models of the disease (177, 259), and inhibition of HA synthesis with 4-MU has been effective in preventing autoimmune destruction of pancreatic islets, and diabetes progression (260).

Primary Sclerosing Cholangitis is an autoimmune disease of the biliary ducts resulting in fibrosis and ultimately cirrhosis, without known curative treatment short of transplantation. Serum HA levels are associated with PSC progression (261) and 4-MU was shown to be beneficial in preclinical models of PSC and liver fibrosis (22, 262). Because 4-MU has been known to ameliorate bile duct spasm for many years (263), it is now investigated as a potential therapeutic in PSC. A dose-finding study of 4-MU (NCT02780752) has just been completed and results are eagerly anticipated, as this could be the first step towards the study of 4-MU in PSC. Other autoimmune diseases, such as multiple sclerosis, thyroiditis and rheumatoid arthritis have been associated with elevated HA levels in serum and ECM, and may similarly benefit from inhibition of HA synthesis (264).

Given the roles of HA in cancer, described above, inhibition of HA synthesis is an attractive treatment option in cancer (265). 4-MU induces apoptosis and inhibits cancer growth in vitro (266, 267) while promoting the immune response in cancer tissue, inhibiting angiogenesis and improving survival in vivo (268–271).

3.5. Novel applications of HA Therapeutics: Hyaluronidase use

Hyaluronidase has been used for several years as an adjunct to local injection treatment (for example with local anesthetics or subcutaneously administered systemic medications), since the degradation of skin HA aids in drug absorption thus improving effectiveness (272–274). A similar use of hyaluronidase as adjunct in clinical care explored the use of hyaluronidase-assisted subcutaneous resuscitation in severely dehydrated children (aged 2 months or older) who could not get intravenous fluids (275).

Based on preclinical data (276, 277), hyaluronidase treatment has been mostly considered in cancer. In a study using recombinant human hyaluronidase in association with chemotherapy in pancreatic cancer, a significant improvement in survival was observed with hyaluronidase (278). Unfortunately these results were not replicated in a follow up, larger study, where there was no improvement in survival, although there was an improvement in objective response rate (decrease of cancer lesion size) (279). However, questions about patient selection and adequacy of dosing were raised (280), suggesting that hyaluronidase treatment may have utility in carefully selected cancer patients.

Non-cancer disease processes may also be candidates for hyaluronidase treatment. For example, hyaluronan deposition occurs after intraventricular hemorrhage in premature infants, resulting in inflammation, arrested oligodendrocyte maturation, and reduced white matter myelination. In a rabbit model of intraventricular hemorrhage, hyaluronidase treatment reduced CD44 and TLR4 expression and inflammation, and promoted oligodendrocyte maturation and myelination (281).

3.6. Novel applications of HA Therapeutics: Inhibition of HA-hyaladherin matrix formation

As mentioned in Chapter 2, a pathological HA matrix is deposited during chronic inflammation and mediates the development of disease. Disruption of this pathological matrix may thus ameliorate the disease. Most approaches have taken advantage of the fact that this pathological HA matrix seems to be reversible and amenable to resolution by the addition of exogenous HA molecules which remove hyaladherins from the deposited LMW-HA. In a mouse model of allergic asthma, instillation of HMW-HA into the airways led to reduction of pathological HA-HC complexes and amelioration of cellular and humoral inflammation (53). In the rabbit intraventricular hemorrhage model discussed above, HA-HC complexes contribute to inflammation and intraventricular injection of oHA reduced inflammation and enhanced myelination by depleting HC-HA levels (281).

3.7. Novel applications of HA Therapeutics: Therapeutic applications of hyaladherins (IαI, TSG-6)

In the Introduction to Chapter 3 we suggested that deposition of HA matrix is a physiological response to acute inflammation and has protective effects in this context. Along those lines, hyaladherins TSG-6, IαI and PTX-3 have been extensively found to be protective in many models of acute disease. TSG-6 may be one of the main mediators of the beneficial effects of mesenchymal stem cells whenever they are applied therapeutically, and was successful in ameliorating disease in many preclinical models e.g. myocardial infarction, skin and eye wound healing, colitis, traumatic brain injury and acute lung injury (69). IαI is perhaps the closest to pharmacological application. Based on promising results in preclinical models (87, 282–285), a pharmaceutical company is currently producing IαI for applications in severe pneumonia, adult and neonatal severe sepsis, and adult and neonatal ischemic brain injury (286). Overall, there is a promising future in the utilization of hyaladherins, either alone or in a HA matrix composition for medical purposes. Hyaladherins may affect the conformation of HA, thus its ability to cluster and engage receptors, they may also vastly broaden the spectrum of receptors and other ECM molecules that the HA matrix can interact with (114), and lastly they can also affect the metabolism and degradation of HA matrices.

3.8. HA in host-environment interfaces (lung, skin, intestine) think interactions with mucins and rheology, infections

HA is abundant in skin (287), in the luminal surface of the conducting airways (288), and other interfaces like the urinary bladder (289). HA can also be ingested under natural conditions: for example, HA of different molecular weights (ranging from oligosaccharides to MHW-HA) can be found in human breast milk (290), and human milk HA can promote epithelial antimicrobial defenses in vitro and in vivo via activation of CD44, ERK1/2 and the expression of β-defensins (158). Specifically-sized HA molecules have shown comparable effects. A 750 kDa HMW-HA molecule (given intraperitoneally) protects against intestinal epithelial injury in vivo via activation of TLR4 and COX2 (291, 292). A much smaller 35 kDa LMW-HA molecule, given orally, was able to promote β-defensin expression and antibacterial response in a TLR4-dependent but CD44 independent fashion (293) while also promoting epithelial integrity via activation of the HA receptor layilin and protecting against infectious and necrotizing colitis in vivo (234, 294–296). Oral administration of this HA fragment was well tolerated by healthy adult humans (297), supporting the potential utility of HA as a dietary supplement.

In the skin, emerging evidence suggests that HA fulfills a complex role linking innate and adaptive immunity (298). Degradation of HA into oHA fragments may serve to fight cutaneous infections, but may promote or ameliorate cutaneous sensitization to allergens depending on the timing of HA exposure in relation to allergen exposure: oHA exacerbate allergy when present at the time of sensitization, but may prevent it if present prior to sensitization (298). While these data provide tantalizing clues for a utility of HA as a treatment in skin infection and inflammatory diseases (like atopic dermatitis), it is still too early to fully appreciate the mechanisms and potential applications.

A much more mature arena is the role of HA at the airway interface. HA interacts with the cystic fibrosis transmembrane conductance regulator (299) which has emerged as a pathogenic factor not only in cystic fibrosis but also in COPD (300–304), and forms complexes with IαI, which blocks epithelial sodium channel activation, thus promoting mucus fluidity (305). As discussed above, LMW-HA promotes ciliary beating (224) and induces mucin expression (225, 226), while in COPD cells, HMWHA promotes mucociliary clearance without affecting ciliary beating frequency or airway surface layer hydration (244). In children with recurrent upper respiratory tract infections, aerosolized HMW-HA improved symptoms and ciliary motility while reducing adenoid hypertrophy, neutrophilic inflammation and microbial growth (306). In patients with nasal polyposis undergoing sinus surgery, intranasal HMW-HA improved mucociliary clearance and nasal obstruction (307). In aggregate, these data support a potentially fundamental role of HA in airway clearance. Furthermore, HMW-HA has bacteriostatic (308, 309), antibacterial (310) and antiviral (311) properties, and can sterically inhibit the diffusion of particles (such as micro-organisms) through mucin (312). HMW-HA is thus a well-studied modality in airway disease and is ripe for investigation in novel applications and disease settings (313, 314).

HMW-HA can also impact infections by interfering with the development of bacterial biofilms, i.e. bacteria attached to surfaces (either organic or medical devices) and held together by self-produced polymer matrices. Approximately 65–80% of microbial infections in the body, including almost all chronic infections, involve bacterial biofilms (315). Important biofilm-producing bacteria are well known pathogens like Streptococcus pneumoniae, S. pyogenes, Haemophilus influenzae, Moraxella catarrhalis and Pseudomonas aeruginosa (316–318). In studies of children with chronic or recurrent otitis and sinusitis media, mucosal biofilm was found in 54–95% of cases (319, 320). Biofilms insulate bacteria against antibiotics, thus fostering antibiotic resistance (315, 321, 322). Thus, bacterial biofilm inhibition contributes to antibacterial properties. HMW-HA inhibits bacterial adhesion to cellular substrates and has notable antiadhesive properties and high antibiofilm activity (323). In patients with recurrent or refractory cystitis, bladder instillation of HA significantly reduced the prevalence of recurrent infections and symptoms or irritation in affected patients. (324–326). Together, available data support the consideration of topical HA as a treatment agent in diseases with inflammatory and infectious etiology.

3.9. HA as delivery system

Hyaluronan has gained significant utility as a delivery system for cancer therapeutics but also other drugs, and there are several reviews on this functionality (2, 327–329). The bioavailability, degradability, non-immunogenicity of HA, and ability to target tissues with an abundance of HA receptors (e.g., cancer tissue which overexpresses the HA receptor CD44), make HA an attractive delivery system. Several applications are either in development or in clinical testing and have been reviewed elsewhere (327, 329). For reasons of expedience, we will only mention one example: intravesical application of a conjugate of paclitaxel and hyaluronic acid has shown promise in patients with carcinoma in situ (CIS) of the bladder that is not responsive to traditional treatment with bacillus Calmette-Guérin (BCG) (330). Undoubtedly, HA-linked delivery systems are an intense area of development in pharmaceutics (329).

4. Conclusions and Future Directions

The sum of available information suggests that HA is an attractive target for pharmaceutical development. HA (and its hyaladherins) are natural, cheap, biodegradable, biocompatible molecules. There is a long history of use with an outstanding safety profile. Furthermore, novel findings strongly suggest that there are exciting new horizons in HA applied biology. The HA matrix, in its many forms, is shown to participate in the maintenance of homeostasis and the development of disease in many areas of clinical medicine. There are several challenges to be overcome in the process. Size and viscosity of HMW-HA preparations can complicate the formulation of easily dispensable preparations. Also, because effects of the HA matrix can differ based on cell type, disease process and pathogenesis, the candidate diseases for a clinical translation must be carefully selected. HA is easily degradable in its native form, and formulations need to be explored which prolong half-life time without compromising efficacy. Finally, large-scale, cost-efficient production of high-purity HA, without bacterial contamination, will have to be implemented. These challenges, however, are not insurmountable. Thus, the time seems ripe to explore HA as a pharmaceutical not simply based on its physical characteristics, but by exploiting its unique and versatile biology. HA can be a prominent member of a new drug family, that may be called Matrix Biologics, which will unleash the potential of the extracellular matrix in shaping health and disease.

Figure 1.

HA metabolism and signaling pathways that may be amenable to therapeutic targeting. HA synthases produce HA which is then bound to hyaladherins (TSG-6, PTX-3, IαI, Versican) in the extracellular matrix, and indirectly to other extracellular molecules like complement, fibronectin, collagen etc. In disease, enzymatic (hyaluronidases) and non-enzymatic degradation of HA generates a LMW-HA matrix which can mediate disease progression or development through interactions with surface HA receptors such as CD44, RHAMM and toll-like receptors, which mediate its biological effects via kinase activation and receptor signaling. There are several pharmacological targets and interventions in this cycle (highlighted in bold letters), including: inhibition of HA synthases with 4-methylumbelliferone (4-MU); targeting hyaluronidases; and using HMW-HA or oHA to disturb LMW-HA matrix formation and signaling. Furthermore, in select diseases use of hyaladherins like TSG-6 or IαI may be of therapeutic benefit.