Moving beyond the glial scar for spinal cord repair

Abstract

Traumatic spinal cord injury results in severe and irreversible loss of function. The injury triggers a complex cascade of inflammatory and pathological processes, culminating in formation of a scar. While traditionally referred to as a glial scar, the spinal injury scar in fact comprises multiple cellular and extracellular components. This multidimensional nature should be considered when aiming to understand the role of scarring in limiting tissue repair and recovery. In this Review we discuss recent advances in understanding the composition and phenotypic characteristics of the spinal injury scar, the oversimplification of defining the scar in binary terms as good or bad, and the development of therapeutic approaches to target scar components to enable improved functional outcome after spinal cord injury.

Fig. 1

Cellular and extracellular composition of the spinal injury scar. Traumatic spinal cord injury triggers a complex cascade of events that culminate in the spinal injury scar which consists of multiple cell types as well as extracellular and non-neural components. a In the acute post-injury phase (0–72 h), cell death and damage lead to release of a number of cellular and blood-derived DAMPs (damage associated molecular patterns). These are powerful activating and inflammatory stimuli for stromal cells, astrocytes, NG2 + OPCs and microglia. Fibroblast-like cells proliferate from perivascular origin. Activated cells increase deposition of extracellular matrix molecules such as chondroitin sulfate proteoglycans (CSPGs) and stromal-derived matrix. Circulating immune-responders (neutrophils, monocytes) are recruited, their relative expression of cytokines, chemokines and matrix metalloproteinases becomes shaped by the early injury environment, and a mixed immune cell phenotype (M1, pro-inflammatory; M2, pro-resolving) is initially adopted. This becomes increasingly proinflammatory. b In the chronic spinal injury scar, monocyte-derived macrophages/microglia adopt a predominantly M1 phenotype. Rather than entering a phase of resolution, responding innate immune cells present DAMPs to circulating adaptive immune cells and pathology spreads. Reactive astrocytes hypertrophy, upregulate expression of intermediate-filament associated proteins and secrete matrix CSPGs. Fibroblast-like cells contribute to fibrotic tissue remodelling and deposition of stromal-derived matrix. Innate immune cells become unable to process cellular and matrix debris effectively and become synonymous with lipid-rich foam cells. Scar-forming reactive astrocytes organise into a barrier-like structure which separates spared tissue from a central region of inflammation and fibrosis where wound-healing fails to undergo resolution. In most mammalian species a chronic cystic cavity develops. Wallerian degeneration of injured axonal projections contributes to continued extracellular deposition of axonal and myelin debris, which is ineffectively processed by immune cells, and along with CSPGs, acts to inhibit neuronal regeneration and neuroplasticity long-term

Fig. 2

From injury to scar: time course of progressive scar pathology showing interlinked relationships between different components of the spinal injury scar. Following traumatic spinal cord injury, acute cell death and damage triggers release of cell-derived and blood-derived DAMPs, ATP release, dysregulated ionic homeostasis oxidative stress and excitotoxicity, which represent potent stimuli for triggering glial cell activation, stromal cell proliferation, deposition of extracellular matrix (ECM), and recruitment of circulating innate immune cells. Within a few days following injury, monocyte-derived macrophage/microglia adopt a predominantly M1 phenotype which do not favour resolution and tissue remodelling becomes fibrotic. Proinflammatory innate responders also present DAMP-derived antigens (such as MBP) to T and B-cells. B cells, in turn, may present antigens to T-cells, triggering their expansion. During this time, reactive astrocytes proliferate, hypertrophy and overlap in order to isolate this zone of non-resolving pathology from spared tissue. They also secrete matrix CSPGs, which are known to downregulate neuronal plasticity. Wallerian degeneration of degenerating axonal tracts contributes to continued deposition of axonal and myelin debris, which is ineffectively processed by immune cells and leads to the deposition of myelin-associated molecules (MAG, Nogo, OMgp) which are known inhibitors of neuronal regrowth. Ongoing Wallerian degeneration at later post-injury stages further triggers gliosis and neuroinflammation. Dashed grey arrows show cross talk between different components of the spinal injury scar, which is usually bidirectional. For example, CSPGs released by reactive astrocytes are thought to activate receptors on macrophages/microglia to induce a proinflammatory phenotype and in turn increasing inflammation induces further astrocytic reactivity and CSPG deposition. Fibroblast-like cells also synthesise type 1 collagen, implicated in the induction of astrogliosis and further deposition of matrix molecules. Cross talk between the innate and adaptive immune response also propagates inflammatory pathology and further influences glial activation and CSPG production. The dynamic interactions between inflammation, dramatic tissue and ECM remodelling and reactive cellular and extracellular changes drive the progressive, propagating pathology that culminates in the spinal injury scar