A tissue’s extracellular matrix (ECM) microenvironment has significant effects on cell behavior, as first revealed in studies from Mina Bissell’s lab (Bissell 2016). Their data showed that there is a dynamic reciprocity between cells and their surrounding ECM that impacts their biomechanical properties (Roskelley and Bissell 1995; Nelson and Bissell 2005; Nelson and Bissell 2006). This leads to changes in gene expression that can alter cell function to drive disease processes. More recently, it has become clear that, in response to wounding, both the composition and physical properties of the matrix produced to promote tissue repair/regeneration can become pro-fibrotic, as well as contribute to cancer progression (Lu et al., 2011; Kim et al., 2018; Basta et al., 2021; Correa-Gallegos and Rinkevich 2021; Tolg et al., 2021). These pro-fibrotic, post-wounding matrix microenvironments promote fibrosis in part by inducing cells that regulate wound-repair to acquire and maintain a myofibroblast phenotype (Hinz 2016; Karppinen et al., 2019; Tai et al., 2021).

Wound-induced provisional matrix proteins like fibronectin and collagen I are among the components of the wound-healing microenvironment that have been linked to promoting fibrosis. We show that specific expression of these matrix proteins by mesenchymal leader cells that populate the wound edge in response to cataract surgery wounding and their transition to myofibroblasts is dependent on TGFβ (Basta et al., 2021). These studies were performed with an ex vivo post-cataract surgery explant model that mimics both lens epithelial wound repair and the pathological fibrotic outcome of Posterior Capsule Opacification (PCO) (Walker et al., 2007; Walker et al., 2010; Menko et al., 2014; Walker et al., 2018). These wounded lens explant cultures provide an ideal reductionist model for investigating how provisional matrices organized to promote cell migration in response to wounding can lead to fibrotic disease progression. They are created by performing a mock cataract surgery on chick embryo lenses, ex vivo. This microsurgery removes the lens fiber cell mass and leaves behind the lens epithelial cell monolayer closely linked to the basement membrane capsule that surrounds the lens. Interdigitated among the lens epithelial cells are a subpopulation of mesenchymal cells that we have identified as resident immune cells. Resident immune cells are among the earliest responders to a wound site (Oishi and Manabe 2018), with a primary function of maintaining tissue homeostasis (Lech et al., 2012), earning them the moniker of the “sentinels of the immune system” (Davies et al., 2013). Our studies demonstrate that resident immune cells are immediate responders to lens wounding, rapidly populating the wound edges in both the chick lens mock cataract surgery explants and in human pediatric post-cataract surgery explants (Menko et al., 2021). These vimentin-rich mesenchymal leader cells express CD45, CD44 and MHCII (Walker et al., 2018; Menko et al., 2021), and display properties of professional phagocytes (Walker and Menko 2021), features they share with resident immune cells in other tissues (Arandjelovic and Ravichandran 2015; Lim et al., 2017; Oishi and Manabe 2018). Their presence in the lens and their response to lens injury involve important roles in directing lens wound repair (Menko et al., 2014; Bleaken et al., 2016). Like other tissue resident immune cells, the resident immune cells would also induce activation of an adaptive immune response, as occurs in response to lens injury (Jiang et al., 2018) and dysgenesis (Logan et al., 2017).

While the primary function of tissue resident immune cells is to maintain homeostasis, they can become agents of pathogenesis (Boyman et al., 2007; Davies et al., 2013; Richmond and Harris 2014; Ginhoux and Guilliams 2016; Lu et al., 2019; Masopust and Soerens 2019; Ardain et al., 2020). In studies with human post-cataract surgery explants we show that the CD45 ⁺ resident immune cells but not the lens epithelial cells in these explants, acquire a myofibroblast phenotype (Menko et al., 2021). Cell tracking studies using antibody to CD44 that binds exclusively to the resident immune cells identified that the resident immune cells in the chick mock cataract surgery explant cultures were the progenitors of the αSMA+ myofibroblasts that appear in the mock cataract surgery explant cultures (Walker et al., 2018). The transition to αSMA+ myofibroblasts is a hallmark of PCO, a common fibrotic pathological outcome of cataract surgery, which also is characterized by cell proliferation, migration, and deposition of matrix proteins (Wormstone et al., 2009; Wormstone and Eldred 2016; Menko et al., 2020; Walker and Menko 2021; Wormstone et al., 2021). While the matrix proteins in the wound microenvironment are key to signaling the acquisition of a fibrotic phenotype, we have also identified a requisite role for the intermediate filament protein vimentin associated with the mesenchymal leader cell population. These functions include directing the collective migration of lens epithelial cells (Menko et al., 2014), and the transition of these leader cells to a myofibroblast phenotype (Walker et al., 2018).

We now examine the impact of hyaluronic acid/hyaluronan (HA) and its receptors CD44 and Receptor for Hyaluronan Mediated Motility (RHAMM) post-wounding in studies with our clinically relevant, ex vivo mock cataract surgery wound repair/fibrosis model. HA is a glycosaminoglycan ubiquitously expressed in the ECM that also has been localized to the cytoplasm and the nucleus (Ripellino et al., 1988; Evanko and Wight 1999; Dicker et al., 2014). In the eye, HA is the principal component of the vitreous humor (Bremer and Rasquin 1998; Slevin et al., 2007; Theocharis et al., 2008), from where it was first purified. HA is found from early stages of development in the basement membranes of many eye tissues, including the lens (Peterson et al., 1995), and the interphotoreceptor matrix of the retina (Inatani and Tanihara 2002; Tanihara et al., 2002). A salient feature of HA is its expression in the provisional matrix that is assembled in response to tissue wounding. HA functions in promoting cell migration, a central element in modulating tissue repair and regeneration (Maytin 2016). Important to our investigations, the presence of HA in the post-wounding microenvironment is also linked to the development of fibrosis (Albeiroti et al., 2015; Maytin 2016; Tai et al., 2021). This property could, in part, reflect the reported role for HA in fibronectin fibrillogenesis (Assunção et al., 2021). HA associates either directly or indirectly with many different proteins in the ECM. The relationship between HA and the accumulation of fibronectin in the matrix environment is highlighted in studies of the trabecular meshwork that showed blocking HA synthesis reduced the presence of fibronectin (Keller et al., 2012). The complex, and sometimes antithetical, roles of HA in biological processes, pro-inflammatory and anti-inflammatory, promoting wound-repair/regeneration and promoting fibrosis, is determined by many different factors. These features include its size, modifications, structural organization, and binding to different receptors (Amorim et al., 2021).

HA synthesis and turnover are tightly regulated (Jiang et al., 2007; Aya and Stern 2014; Garantziotis and Savani 2019; Kobayashi et al., 2020). Hyaluronan synthases (HASs), are the transmembrane enzymes responsible for producing HA, while hyaluronidases are the enzymes critical to breakdown of HA (Jiang et al., 2007; Aya and Stern 2014; Garantziotis and Savani 2019; Kobayashi et al., 2020). There are three HASs; HAS1, HAS2 and HAS3, which can exhibit distinct subcellular locations, expression patterns and regulation (Itano and Kimata 2002; Heldin et al., 2019; Kobayashi et al., 2020). Typically, HASs are localized at the plasma membrane where HA is secreted to the extracellular environment (Itano and Kimata 2002; Heldin et al., 2019; Kobayashi et al., 2020).

High molecular weight (HMW) HA accumulates at sites of injury and plays roles in promoting tissue remodeling and wound repair (Jiang et al., 2007; Aya and Stern 2014; Kobayashi et al., 2020). In the wound environment, hyaluronidases and free radicals can fragment HMW-HA into low molecular weight forms (LMW-HA) that are associated with promoting inflammation and disease (Jiang et al., 2007; Aya and Stern 2014; Kobayashi et al., 2020). Many studies support a role for HA in the development of fibrosis in tissues such as the liver (Andreichenko et al., 2019; Yang et al., 2019) and lung (Li et al., 2011). HA was found to mediate fibroblast transition to a scar-inducing myofibroblast phenotype associated with fibrosis (Meran et al., 2007; Webber et al., 2009). TGFβ-induced differentiation of dermal fibroblasts to a myofibroblast phenotype is linked to HA production and assembly into a pericellular coat (Meran et al., 2007). Interestingly, in this same study, nonscarring oral fibroblasts, were found resistant to TGFβ-induced myofibroblast differentiation. This alternative response was associated with a lack of HA production and HA pericellular coat formation (Meran et al., 2007). HA synthesis is also required for TGFβ-induced differentiation of lung fibroblasts to myofibroblasts (Webber et al., 2009). Surprisingly, providing lung fibroblasts with exogenous HA blocked myofibroblast differentiation, which was associated with the relocation of the HA receptor CD44 and TGFβR ALK5 from lipid raft to non-lipid raft regions of the membrane (Webber et al., 2009). This paradoxical HA response lends support to the concept that HA presentation and organization are critical to how HA modulates cellular function (Webber et al., 2009). These studies emphasize a key role for cell-associated HA in inducing a fibrotic response, and demonstrate that HA interactions with receptors at the membrane is essential to its functional outcomes.

Principal among the many receptors for HA, are CD44 and RHAMM; others include LYVE-1, TLR2 and TLR4 (Garantziotis and Savani 2019; Abatangelo et al., 2020; Amorim et al., 2021). The interaction between HA and these receptors provide HA with the ability to activate many different downstream signaling effectors and impact cell behavior. Its cell-surface receptor CD44 is a single pass transmembrane glycoprotein expressed by a number of cell types including bone marrow mesenchymal cells, immune cells, embryonic stem cells and cancer stem cells (Levesque and Haynes 1996; Chen et al., 2018; Lee-Sayer et al., 2018). CD44 has many functions including the regulation of cell migration and cell proliferation (Krolikoski et al., 2019; Chen et al., 2020). The HA receptor RHAMM is even more complex. It is a multifunctional protein that localizes to both extracellular and intracellular sites. RHAMM has important roles both during development and in the wound-repair response that include mediating cell migration, from which it derives its name (Savani et al., 1995; Leng et al., 2019; Tolg et al., 2020). RHAMM localizes to the cell surface where it mediates binding to HA in the extracellular matrix environment, and to the cell nucleus where it associates with transcriptional complexes to regulate gene expression (Meier et al., 2014). Studies also show localization of RHAMM to the cytoplasm and its association with both microtubules and actin filaments (Assmann et al., 1999), as well as to the centrosome where it functions in regulating spindle pole stability (Maxwell et al., 2003). Following wounding, RHAMM has been specifically localized to mesenchymal cells with a fibroblastic morphology that localize to the leading edge of the wound in response to injury (Hardwick et al., 1992; Savani et al., 1995). Like their ligand HA, both CD44 and RHAMM have been linked to the development of fibrosis (Tolg et al., 2012; Cui et al., 2019; Govindaraju et al., 2019; Wu et al., 2021). In this study, we show that HA is expressed in the matrix microenvironment of the mesenchymal cells that are recruited to the leading edge of the wound where it is associated with both CD44 and RHAMM and induces cell migration and the transition of the mesenchymal leader cells to αSMA+ myofibroblasts.

A provisional matrix is formed in the wound microenvironment to promote cell migration across the injured area and close the wound (Barker and Engler 2017; Chester and Brown 2017; Johnson et al., 2018; McKay et al., 2019; Das et al., 2021). In addition to serving as the substrate for movement of the wound-activated cells, this ECM can also be a sink for cytokines and growth factors that impact both cell movement and cell fate (Schultz and Wysocki 2009; Wilgus 2012; Frangogiannis 2017). Induction of many of the components of this provisional matrix post-wounding, including collagen I, fibronectin and HA, have been linked to activation of TGFβ (Roberts et al., 1986; Ignotz et al., 1987; Albeiroti et al., 2015). These matrix proteins provide ligands for receptors expressed on the cell surface that participate directly in cell adhesion and in the transmission of signals from the matrix that promote cell migration. In the post-wounding microenvironment, fibronectin, collagen I, HA and TGFβ are all also linked to promoting fibrosis, a pathological outcome to their function in wound repair (Leask and Abraham 2004; Albeiroti et al., 2015; Walraven and Hinz 2018). The post-cataract surgery fibrotic disease PCO is characterized by both cell migration onto the cell-denuded posterior lens capsule and the subsequent appearance of myofibroblasts (Spalton 1999; Wormstone et al., 2009; Wormstone and Eldred 2016; Walker et al., 2018; Wormstone et al., 2021). Collagen I, fibronectin and TGFβ have been linked to fibrosis and fibrotic PCO (Uitto et al., 1982; Serini et al., 1998; Gabbiani 2003; Muro et al., 2008; Herrera et al., 2018; Basta et al., 2021). However, the only evidence that HA may play a role in induction of fibrosis following lens wounding comes from a study with a canine ex vivo cataract surgery model where the inclusion of exogenous HA promoted fibrotic PCO (Chandler et al., 2012). Our findings now showed that HA is produced in response to cataract surgery wounding and that its presence in the lens wound environment promotes cell migration and also creates conditions permissive to transitioning the cells involved in lens wound repair to a myofibroblast phenotype.

These findings support developing a therapeutic approach targeting HA production to mitigate the development of the lens fibrotic disease PCO. The use of a small molecular inhibitor to block HA production at the time of cataract surgery, applied directly to the post-cataract surgery lens capsular bag is expected to have limited effects on surrounding eye tissues. The inhibitor used in our study, 4-MU, blocks global HA synthesis by depleting cellular pools of the HAS substrate UDP- glucuronic acid (UDP-GlcUA), which is required for HA synthesis (Kakizaki et al., 2004). 4-MU also was shown to decrease HAS mRNA levels (Kultti et al., 2009). Therefore, treatment with 4-MU will interfere with the function of all three HA synthases, HAS1, HAS2 and HAS3, with the potential to impact both fibrotic and regenerative repair. A better approach would be to identify and target the specific injury activated HAS(s) lead to the lens fibrotic phenotype post-cataract surgery wounding. A likely candidate is HAS2, which is linked to driving fibrosis in other tissues, such as the liver and lung (Li et al., 2016; Yang et al., 2019). The ex vivo post-cataract surgery model will provide an ideal reductionist model for future studies in which to determine the function of individual HASs in promoting the fibrotic outcome to cataract surgery wounding.

In studies with the ex vivo chick mock cataract surgery explant cultures and human post-cataract surgery explant cultures we now identify the mesenchymal leader cells that rapidly populate the wound edges of the explant and acquire a myofibroblast phenotype as tissue resident immune cells (Menko et al., 2021). Our earlier studies demonstrated that this mesenchymal leader cell population directs the wounded lens epithelium to migrate off the capsule onto and across the surrounding tissue culture substrate (Walker et al., 2010; Menko et al., 2014; Walker et al., 2015; Bleaken et al., 2016), and that their acquisition of myofibroblast phenotype continues to expand over time (Walker et al., 2018). We show that the properties of these resident immune cells includes expression of CD45, MHCII, and the HA receptor CD44 (Menko et al., 2021; Walker and Menko 2021). The interaction between CD44, a hallmark of the leader cell population, and HA is considered a driving force in the recruitment and migration of immune cells (McDonald and Kubes 2015; Nagy et al., 2019).

Studies of the cataract surgery response in CD44 knockout mice shows that the development of fibrotic PCO is not impaired in the absence of this HA receptor (Desai et al., 2010). Our new findings suggest that this role is played by the HA receptor RHAMM. Its pattern of localization in the mesenchymal leader cells post-cataract surgery wounding to lamellipodia processes extended at the leading edge, and along actin stress fibers are consistent with functions in leader cell migration, transition to a myofibroblast, and myofibroblast persistence in fibrosis. Since HA signaling through RHAMM can mediate the induction of cell migration by TGFβ (Samuel et al., 1993), it is possible that RHAMM function is coordinated with TGFβ in promoting a fibrotic outcome. The intracellular population of RHAMM have distinct, and important functions that include roles as a transcriptional regulator (Meier et al., 2014) and in its association with different elements of the cytoskeleton (Assmann et al., 1999). In mitotic cells, RHAMM is a stabilizer of the spindle poles and there is evidence that of a stabilizing role along microtubules and actin filaments (Maxwell et al., 2003). While we find no evidence of a microtubule association in the mesenchymal leader cells post-cataract surgery wounding, we found that RHAMM is localized along the αSMA+ stress fibers of myofibroblasts during the later culture times of our study. A recent study shows that Hyaluronidase-2 (HYAL2) localizes along F-actin rich stress fibers and associates with αSMA in TGFβ-induced myofibroblasts (Midgley et al., 2020). In this study, HYAL-2 is shown to interact with and activate RhoA to regulate myofibroblast migration, contraction, and expression of pro-fibrotic genes such as FN and collagen I (Midgley et al., 2020). The initial localization of RHAMM to lamellipodial extensions of the mesenchymal cells that locate to the leading-edge post-wounding in our studies suggests that RHAMM may be concentrated at integrin focal adhesion complexes, sites where active Rho/ROCK signaling involved in inducing the assembly of both the focal adhesions and the actin stress fibers that directly link to them. We speculate that RHAMM localization to these sites may play a role in the initial formation of actin stress fibers, and the mechanotransduction signaling involved in acquisition of a myofibroblast phenotype. Furthermore, localization of RHAMM along αSMA+ stress fibers formed after the transition of the leader cells to myofibroblasts may function in a similar manner to HYAL2 to regulate their contractile, pro-fibrotic functions. Rho-Rho kinase signaling is critical to the acquisition of a myofibroblast phenotype in the lens (Korol et al., 2016) and for capsule contraction linked to fibrosis in a mouse lens injury model (Ichikawa et al., 2020). Our new findings that RHAMM localizes initially to leader cells in regions rich in integrin focal adhesion complexes and then along the αSMA+ contractile machinery after their transition to myofibroblasts open promising new areas of study for investigating whether there is a functional link between RHAMM and Rho-ROCK signaling in response to cataract surgery wounding. RHAMM’s distinct extracellular and intracellular subcellular patterns observed with the ex vivo post-cataract surgery cultures suggest that RHAMM plays multiple location-specific functions to mediate the fibrotic outcome to cataract surgery wounding.

Our PLA studies provide the first evidence that the localization of RHAMM to the lamellipodial processes of mesenchymal leader cells post-wounding reflects its presence in a complex together with both HA and vimentin. Vimentin is an intermediate filament protein with essential functions in both cell migration and wound closure (SundarRaj et al., 1992; Eckes et al., 1998; Eckes et al., 2000; Mendez et al., 2010; Menko et al., 2014). While it is best known as a cytoskeletal filamentous network that provides cells with resistance to mechanical stresses, non-filamentous forms of vimentin have been identified with both intracellular and extracellular functions (Mor-Vaknin et al., 2003; Teshigawara et al., 2013; Menko et al., 2014; Shigyo et al., 2015; Shigyo and Tohda 2016; Walker et al., 2018). The extracellular form of vimentin has been linked to the activation of latent TGFβ (Nishida et al., 2009). Our previous studies show that extracellular vimentin is produced in response to cataract surgery wounding and that an extracellular, cell surface-linked, population of vimentin is involved in signaling the transition of the mesenchymal leader cells to a myofibroblast phenotype (Walker et al., 2018). The ex vivo lens-wound explant model closely parallels the complexity of typical wound environments, including the organization of an HA-containing provisional matrix that is formed to promote cell migration and close the wound that also can induce repair-modulating wound-response cells to acquire a myofibroblast phenotype. Our discoveries, including that of an HA/extracellular vimentin/RHAMM axis that is likely linked to both cell migration and the transition of mesenchymal leader cells to myofibroblasts provides a previously unknown key to the mysteries of the antithetical outcomes of healing and fibrosis in the wound environment.