A proper wound healing response is vital for tissue regeneration and involves discrete stages of inflammation, tissue remodeling, and their resolution (1). The inflammatory response is critical for the removal of pathogens and debris from the injury site. It also recruits and activates tissue specific effector cells required for remodeling. Immune and effector cells deposit a variety of extracellular matrix (ECM) molecules to reconstitute matrix lost to injury (1). Successful wound healing recapitulates the tissue environment and restores function (2). Dysregulation of wound healing following repeated, chronic, or pronounced single injury leads to fibrotic scarring due to excessive build-up of cells and ECM (1, 3).

Fibrosis is characterized by increased tissue stiffness and disrupted tissue architecture ( Figure 1 ) (3). It is identified by the accumulation of ECM components, altered protease expression, increased levels of pro-fibrotic signaling molecules, and the presence of pro-fibrotic cells such as fibroblasts and their differentiated forms, myofibroblasts (1). Even in conditions not classically associated with fibrosis, disease outcomes such as hypoxia and epigenetic reprogramming of cells are described with fibrosis-linked responses (4) such as progressive scarification and exacerbated tissue injury (5).

In the central nervous system (CNS) fibroblasts occupy border regions such as the meninges and perivascular space (6). Following injury they become elevated in the parenchyma where they, along with astrocytes, microglia, infiltrating immune cells and potentially pericytes contribute to tissue reorganization and ECM accumulation (7, 8). This neurofibrosis response occurs commonly in CNS pathologies such as spinal cord injury (SCI), stroke and multiple sclerosis (MS) (9). Here we discuss general wound healing and fibrosis related responses so as to instruct concepts of neurofibrosis, and we then consider fibrosis in the CNS with a focus on MS. Finally, we evaluate the therapeutic potential of targeting neurofibrosis to improve outcomes from CNS injuries.

Fibrosis occurs in many organs and has shared and unique tissue specific characteristics (4). The initiating injury varies based on anatomical location and trigger (5). For instance, liver and kidney fibrosis can occur due to viral hepatitis and diabetes, respectively. Resident and infiltrating immune cells contribute to fibrosis during initial innate and later adaptive responses (3) ( Figure 2 ). Two primary signaling axes, interleukin (IL)-4/IL-13 and IL-1/IL-17A/transforming growth factor-β (TGFβ), are core pro-fibrosis pathways (5), although IL-17 may also be anti-fibrotic depending on the triggering insult and the organ affected (10). Macrophages, the most abundant immune cell type in fibrosis, are recruited by damage associated molecular patterns (DAMPs), pathogen associated molecular patterns (PAMPs), and chemokines (11). Early arriving macrophages possess an inflammatory phenotype and secrete pro-inflammatory tumor necrosis factor-α (TNFα) and IL-1β which fuel inflammation (11). Later stage macrophages assume a tissue remodeling regulatory phenotype characterized by the expression of TGFβ, IL-10, and mannose receptor 1(MRC1/CD206) (11). Inflammatory macrophages contribute to fibrosis by exacerbating the injury, while regulatory macrophages stimulate effector cells to deposit ECM and express other tissue remodeling genes (5). It is not clear what dictates IL-4/IL-13 or IL-1 mediated fibrosis although the chronicity of the injury may be a determinant. Single injections of bleomycin produce fibrosis in an IL-1/IL-17/TGFβ dependent manner; conversely, repeated injections of bleomycin over a longer period elicits IL-4R signaling (5). This has repercussions for signaling to effector cells and the transition to the tissue remodeling stage of repair.

Effector cells responsible for tissue remodeling originate from resident and/or recruited fibroblasts, epithelial cells, endothelial cells, and other tissue resident cells and immune cells (4). Effector cells are recruited by chemokines and growth factors elaborated at sites of injury by the initial-arriving immune cells where they then proliferate and upregulate ECM, proteases, and growth factors (12). Positive feedback loops between effector and immune cells and the fibrotic environment promote further fibrosis by elevation of pro-fibrotic genes (13, 14). For example, fibroblasts upregulate ECM genes when cultured on ECM derived from idiopathic pulmonary fibrosis tissue (14). This can lead to greater scarring of the tissue and disruption of function.

Organ failure and increased morbidity are outcomes of fibrosis-related disorders such as cardiac fibrosis following myocardial infarction (5). Prominently, fibrosis-related disorders are responsible for 45% of fatalities in the United States of America (5). This highlights the severity of fibrosis for tissue function and recovery which is made more significant in a tissue environment that is not prone to regeneration, such as the CNS.

Functional regeneration of the CNS does not occur effectively in adult humans (15) due to factors including lack of available stem cells, age-related changes in neural cells and the tissue environment, and the formation of inhibitory scar tissue after injury (16, 17). Scarring of the CNS results from the accumulation of astrocytes, microglia and macrophages, and fibroblasts and potentially pericytes. The formation of neurofibrosis contributes to impairment of axonal regrowth due to elevated deposition of collagens and chondroitin sulfate proteoglycans (CSPGs), amongst others, into the lesion (7, 8).

In SCI astrocytes become reactive and form a protective boundary to prevent spread of the injury while microglia/macrophages and fibroblast-like cells occupy the center ( Figure 3 ). In MS lesions microglia/macrophages and astrocytes are more intermingled while fibroblast-like cells are expanded in the perivascular compartments (18). Conflicting descriptions of the beneficial and harmful roles of reactive astrocytes exist and have been elaborated on elsewhere (17). Though reactive astrocytes restrict the spread of injury and attempt to maintain homeostatic functions (17), they have also been described to contribute to chronic neuroinflammation and they upregulate many ECM components detrimental for regeneration (19). Conversely, depletion of reactive astrocytes impairs the ability of axons to regrow after SCI (20). Instructively, studies such as these suggest that certain levels of regeneration of the CNS are possible under permissive conditions.

CNS-associated fibroblasts identified by PDGFRβ and Col1a1 driven reporter in mice and antibody labeling are increasingly implicated in neurofibrosis (8, 21). Their complete depletion in SCI results in an open wound defect (8) while their partial depletion leads to greater axon regeneration (22, 23). The identification and ontogeny of fibroblasts in neurofibrosis remain to be better resolved, and the use of different reporter mouse lines is partly responsible for the controversy. Due to the use of PDGFRβ based reporter mice and PDGFRβ as a common marker across several cell types, it is likely that fibroblasts, pericytes, and smooth muscle cells have been conflated with each other (24). Transcriptional profiling of murine mural cells has shown that all these three cell types express common markers including PDGFRβ and NG2 (5, 24). While these studies identified homeostatic gene expression patterns that differentiate pericytes (Kcnj8), fibroblasts (Col1a1), and smooth muscle cells (Acta2), it is unclear how expression of these are affected by injury or inflammation. Further adding to the controversy, type A pericytes identified by a GLAST driven reporter line have been described in SCI (8, 25) but GLAST also labels populations of astrocytes. As well, Dorrier et al. showed using NG2 and SMA inducible reporters that cells basally positive for these genes do not significantly contribute to fibrotic cells in neuroinflammatory lesions (21, 22). However, this does not preclude subpopulations of pericytes or other cell types from contributing to neurofibrosis. Additionally, other lines of evidence suggest that pericytes and smooth muscle cells do not contribute to the fibroblast populations in neurofibrosis although the data is inconclusive (8, 21). Based on the uncertainty of the ontogeny or type of fibroblasts in CNS injuries, we shall refer to them as fibroblast-like cells where indicated.

The molecular impediments to CNS regeneration within fibrotic lesions include inhibitory molecules, reduced levels of growth signals and products of particular inflammatory cells (8, 17). Microglia are CNS resident macrophages and are critical for maintaining tissue homeostasis. Monocyte derived macrophages become elevated following injury and, together with microglia, participate in cellular and lipid debris removal. As well, they upregulate inflammatory and tissue remodeling genes during early and late stages of injury. The ability of microglia/macrophages to metabolize debris becomes impaired with age leading to formation of foamy cells that contribute to chronic inflammation and impaired tissue remodeling (16). In neonatal mice infiltration by peripheral macrophages is resolved rapidly compared to adult mice. Depleting microglia using PLX3397 or CX3CR1-CSF1R^fl/fl reversed spontaneous neonatal axonal regeneration implying that it is microglia that act to promote recovery following optic nerve crush injury (26). Additionally, there is a clear interconnectedness of the cells within CNS lesions as highlighted by altered levels of astrocytes, macrophages, and fibroblasts following depletion of each individually (22, 25, 27). Indeed, both astrocytes and fibroblasts interact reciprocally with immune cells informing each cell’s phenotype ( Figure 3 ) (21, 28). This emphasizes the importance of the cellular components of CNS injury to the success of regeneration.

A theme that has been topical in peripheral fibrosis is that of fibroblast senescence and resistance to apoptosis (29, 30). Thus far, a systematic analysis of age-related neurofibrosis has not been thoroughly conducted. Studies focusing on the effects of age on traumatic brain injury described increased collagen levels in the injured aged meninges (31). Whether this is consistent in humans is not clear though proteomic analysis of CSF in people found increased levels of collagen with aging (32). Further determinations of changes with aging would include fibroblast density and susceptibility to apoptosis in lesions, accumulation of neural ECM in neurological disorders, and tissue rigidity in healthy aging or disease CNS.

While tissue stiffness generally increases in non-CNS fibrosis-related disorders, many forms of CNS injury are reported to result in reduced stiffness (33). One explanation for this phenomenon may be that most animal models of MS are studied during acute stages of injury and tissue stiffness increases during chronic stages of injury. Indeed, CNS lesions become increasingly stiff with chronicity (34). Thus, neurofibrosis is akin to fibrosis present in other organs and presents further questions about the role of fibrosis-related responses in CNS pathologies including MS.

MS is an inflammatory disease characterized by demyelination and neuroaxonal degeneration in the brain, spinal cord, and optic nerve (35). Demyelination occurs in the context of inflammation involving CD8+ T lymphocytes, CD4+ T lymphocytes, B cells, plasma cells, macrophages, microglia and reactive astrocytes (35–37). Lesions are categorized as active, chronic active, inactive, and remyelinating based on immune and remyelinating phenotypes (16). Found in the brain and spinal cord white and gray matter, the location of lesions contributes to symptom presentation and onset (35).

Axonal regeneration can occur under ideal conditions, but it is not common and has only been minimally examined in MS. Conversely, remyelination occurs spontaneously to varying degrees in MS, in all types of lesions except inactive ones, and has been extensively examined (16, 38). During remyelination, oligodendrocyte progenitor cells (OPCs) migrate to the site of injury, proliferate, and differentiate into mature oligodendrocytes capable of producing new compact myelin (16). Remyelination promotes axonal neuroprotection and functional recovery making it an important and regularly occurring regenerative process (16, 38).

Fibrosis affects remyelination in MS (7). OPC migration, proliferation, and maturation are positively and negatively regulated by particular ECM components deposited in lesions (7). Specific effects of different ECM components have been reviewed in detail previously (39). Examples include beneficial effects of some isoforms of laminins on OPC proliferation while ECM molecules such as CSPGs and fibronectin impair OPC activity (7, 40). Inhibition of CSPG synthesis following injury increased oligodendrocyte numbers and remyelination (41). Furthermore, depletion of astrocytes and fibroblasts in models of MS leads to greater density of oligodendrocytes (21, 22). Thus, several elements of fibrosis impair regenerative processes in MS, to which we now turn our attention.

Many processes dysregulated in fibrosis are present in MS lesions including chronic inflammation, effector cell recruitment, and excessive ECM deposition (5, 16). As well, outcomes of fibrosis such as tissue stiffness, hypoxia, cellular senescence, and impaired repair processes impact MS pathology (42, 43). Here we describe neurofibrosis in MS focusing on the inflammation, and remodeling and progression stages.

No therapies currently are used to directly affect fibrosis in MS. Thus, this section ( Figure 5 ) begins with an overview of medications that treat fibrosis in non-CNS disorders, and addresses whether these could be applied for MS. We then discuss whether disease modifying therapies (DMTs) used in MS have unintended effects on ameliorating fibrosis in MS, and end with a forward-looking view of potential therapeutics that could be applied to counter fibrosis in MS.

Only two medications are approved to treat fibrosis-related disorders, specifically for lung fibrosis: pirfenidone and nintendanib (5). Pirfenidone acts as an immunomodulator and antagonist of the TGFβ pathway that promotes fibrosis (4, 63). In people with idiopathic pulmonary fibrosis pirfenidone reduced disease progression, improved lung function, and increased exercise tolerance (64). Pirfenidone has undergone phase I and II clinical trials in secondary progressive MS reporting reduced incidence of relapse and improved bladder function (65, 66). However, these trials were small, and a larger study is needed to determine the potential of pirfenidone for people with progressive MS.

When inhibiting TGFβ signaling, it is important to consider its role as a regulatory cytokine and its function for the maintenance of microglia homeostasis in the CNS (44). Another consideration for drugs to overcome CNS fibrosis is the additional challenge that these compounds will have to cross the blood-brain barrier.

The other approved drug for lung fibrosis, nintendanib, is a tyrosine kinase inhibitor (4, 67). Pharmacological inhibition of tyrosine kinases by other drugs in animal models of MS reduced clinical scores, demyelination, inflammation, astrocyte reactivity, and vascular injury (68–71). However, tyrosine kinases are broadly expressed making off-target effects of systemic administration difficult to isolate from potential anti-fibrotic effects within the CNS. Nonetheless, two clinical trials using the tyrosine kinase inhibitor masitinib in progressive MS have been conducted (72, 73). The first, a Phase I trial, showed masitinib to be safe (72). The second, a Phase IIb/III trial, reported reduced elevation of EDSS disability scores and delayed the time to EDSS of 7.0 (73). The extent to which the promising clinical result could be attributed to amelioration of CNS fibrosis is unknown.

Next, we consider whether the approximately (country-dependent) 20 DMTs used in MS have potential impact on evolution of fibrosis ( Figure 5 ). While such data is lacking, the immunomodulatory nature (74, 75) of MS DMTs conceivably can alter the immune stimulation of fibrogenesis. Either by reducing the content of pro-inflammatory cytokines or elevating levels of T_h2 cells (e.g. IFNβ, glatiramer acetate), sequestering leukocytes in secondary lymphoid tissues (e.g. sphingosine-1-phosphate receptor modulators such as siponimod), inhibiting leukocyte trafficking across the blood-brain barrier (natalizumab), or by depleting B cells (anti-CD20 monoclonal antibodies) and the larger leukocyte populations (alemtuzumab, cladribine), the effect would be reduced recruitment and activation of effector cells of CNS fibrosis. Thus, MS DMTs may have unintended and indirect effects on reducing neurofibrosis, although the possibility of this occurring will have to be established. Future work may consider whether the extent of neurofibrosis is ameliorated in people with MS treated with DMTs, and such studies will be highly reliant on evolution of biomarkers with the capacity to detect neurofibrosis.

Recently, inhibitors targeting the intracellular signaling enzyme, Bruton’s tyrosine kinase (BTK), have grown in interest in the treatment of MS (76). Inhibition of BTK in models of graft versus host disease resulted in reduced dermal fibrosis. As well, the first generation BTK inhibitor, ibrutinib, has been approved for treatment of graft versus host disease (77). BTK inhibitors affect B cells but also microglia proliferation, activation, and survival (76). BTK inhibitors have not been trialed for fibrotic disorders, and the preclinical literature is heterogeneous. Some evidence suggests that loss of BTK can reduce fibrosis following cardiac injury while other studies show that BTK inhibitors either have no effect or detrimental effects in kidney and lung fibrosis (78, 79). However, BTK inhibitors have been approved for treatment of several cancers such as chronic lymphocyte leukemia and mantle-cell lymphoma (77). As well, they have been tested in a number of autoimmune diseases leading to many clinical trials including in MS. Five BTK inhibitors have entered late-stage clinical trials including evobrutinib, fenebrutinib, tolebrutinib, remibrutinib and orelabrutanib (76). Unfortunately, evobrutinib failed to meet its primary endpoint in the phase 3 EVOLUTION trials (80). At this time it is unknown whether BTK inhibitors have a role for treatment of fibrosis in MS.

Finally, another means to reduce CNS fibrosis is to prevent or limit the deposition of ECM molecules that occurs after injury. Pharmacological manipulation of ECM and its production have improved recovery in both SCI and MS models (41, 81). Both impairing CSPG synthesis with fluorosamines (4-F- or 4,4-difluoro-N-acetylglucosamine) and digestion of existing CSPGs with chondroitinase-ABC led to significant recovery (41, 81). In EAE, 4,4,-difluoro-N-acetylglucosamine reduced the frequency of cytotoxic T_h17 and improved remyelination (41). Altering expression of other common ECM components such as hyaluronan improved EAE outcomes (82).

In summary, targeting fibrosis-related responses and outcomes in preclinical studies shows promise as therapeutic strategies for MS. Translation from preclinical into clinically relevant therapies requires more work to understand the implications of these mechanisms on MS and to determine safety and efficacy.

Tissue regeneration is necessary to restore function (2). Altered immune and tissue remodeling results in fibrosis blocking functional recovery, and increases morbidity (5). The adult CNS does not regenerate well leading to accumulation of injury and disability (16, 83). Inability to repair successfully is partly due to remnant cells and ECM components of fibrosis within the injured CNS (9, 18, 41, 84). Similarities exist between CNS pathologies such as MS and peripheral fibrosis and the benefits of targeting these fibrosis-related processes in MS have been highlighted by a growing body of preclinical research (9, 21, 41, 68) as well as some early clinical trials (65, 66). Fibroblasts are beginning to be recognized as components of neurofibrosis and deserve more studies. Questions remain regarding safety and efficacy of targeting neurofibrosis, but its successful treatment would represent an important step forward in the promotion of CNS regeneration and recovery.

BL: Conceptualization, Writing – original draft, Writing – review & editing. SG: Writing – original draft, Writing – review & editing. VY: Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Writing – original draft, Writing – review & editing.