PerspectiveNeuronal Regeneration

A New Role for RPTPσ in Spinal Cord Injury: Signaling Chondroitin Sulfate Proteoglycan Inhibition

See allHide authors and affiliations

Science Signaling  23 Feb 2010:
Vol. 3, Issue 110, pp. pe6
DOI: 10.1126/scisignal.3110pe6

Abstract

It has been known for more than two decades that chondroitin sulfate proteoglycans (CSPGs) inhibit axonal growth and regeneration. In the adult nervous system, CSPGs are enriched in perineuronal nets, and their abundance is increased in reactive astrocytes following injury to brain or spinal cord. Degradation of chondroitin sulfate (CS) sugar moieties by the local infusion of the bacterial enzyme chondroitinase ABC (ChaseABC) enhances experience-dependent neuronal plasticity in the adult visual cortex and results in substantially improved behavioral outcomes after spinal cord injury (SCI). Although the positive effects of ChaseABC treatment on neuronal plasticity have been known for some time, the underlying mechanisms remained enigmatic. The receptor protein tyrosine phosphatase sigma (RPTPσ) has now been identified as a receptor for inhibitory CSPGs. Similarly to ChaseABC treatment, functional ablation of Ptprs, the gene encoding RPTPσ, promotes neurite outgrowth in the presence of CSPGs in vitro and enhances axonal growth into CSPG-rich scar tissue following SCI in vivo. The discovery of neuronal RPTPσ as a receptor for inhibitory CSPGs not only provides important mechanistic clues about CSPG function, but also identifies a potential new target for enhancing axonal growth and plasticity after nervous system injury.

In the adult mammalian central nervous system (CNS), the capacity of severed axons to undergo spontaneous regeneration is very limited (1, 2). A grim example of the regenerative failure of injured CNS neurons is the poor prognosis for individuals suffering from spinal cord injury (SCI). Traumatic injury to the spinal cord often results in substantial damage to spinally projecting ascending and descending fiber tracts. Although most neuronal elements proximal and distal to the injury site survive for a long time, injury-induced disruption of neuronal connections blocks the propagation of electric impulses across the lesion and results in permanent neurological deficits. One important goal of SCI research is to restore lost neuronal connectivity in order to wire electric impulses past the lesion site. This may be achieved by long-distance axonal regeneration, followed by synapse formation on appropriate (preinjury) target cells. An alternative repair strategy is short-distance axonal growth and synapse formation on neuronal elements that form relays to neuronal targets located distal to the injury site. Because the majority of human patients suffering from complete loss of function following SCI nonetheless exhibit spared axons in rimes of white matter that bypass the injury site (3), the latter strategy may be a more realistic goal than true long-distance axon regeneration.

Neuronal growth following SCI, even over short distances, remains a formidable challenge. Although forms of injury-induced neuronal plasticity are observed in the CNS, growth attempts are transient in nature and often fail to restore substantial recovery of lost function. The reasons for the poor regenerative growth of injured CNS neurons are manifold. Cell-intrinsic growth programs in injured CNS neurons are poorly reactivated, and even when provided with a permissive environment, adult CNS neurons are far less capable of extending processes than their embryonic counterparts (46). To make matters worse, the environment of injured adult CNS tissue is “rich” in growth inhibitory cues and “poor” in growth-promoting factors and, as a result, fails to support regenerative growth of severed axons or compensatory sprouting of spared fibers. Much attention has focused on the molecular characterization of the cues that contribute to the growth-hostile CNS environment. Important players are thought to include the prototypic myelin inhibitors Nogo, oligodendrocyte-myelin glycoprotein, and myelin-associated glycoprotein (MAG) (2, 79); inhibitory axon guidance molecules (10); and chondroitin sulfate proteoglycans (CSPGs) (11, 12). The coupling of these inhibitory cues with a relative paucity in growth factor availability or with gradients of growth factors that are supportive of axonal regeneration (13, 14) makes the CNS environment a major hurdle for neuronal regeneration.

Important roles for CSPGs in limiting neuronal plasticity have long been recognized. CSPGs are a diverse group of extracellular molecules composed of a core protein covalently linked to specific types of glycosaminoglycan (GAG) side chains. Neural CSPGs with known inhibitory activities include members of the lectican family (aggrecan, brevican, neurocan, and versican) and phosphacan (12). Many CSPGs are distributed throughout the brain and spinal cord, often condensed into perineuronal nets surrounding somata and dendrites of various types of neurons (12). After SCI, the abundance of aggrecan and neurocan is increased in reactive astrocytes associated with glial scar tissue (Fig. 1). Enzymatic digestion of the chondroitin sulfate (CS)–GAG chains with ChaseABC largely abrogates the inhibitory effect of CSPGs on neurite outgrowth in vitro (1517). Moreover, intrathecal administration of ChaseABC in the mature CNS facilitates neuronal plasticity and repair (18, 19). For example, ChaseABC promotes axonal regeneration of severed corticospinal and dorsal column axons in rats with SCIs (Fig. 1) (20) and regeneration of serotonergic axons following dorsal hemisection in the cat (21). In ChaseABC-treated animals, anatomical regeneration following SCI correlates with improved functional outcomes. New evidence shows that ChaseABC treatment opens a window of enhanced plasticity during which rehabilitative training leads to much greater functional improvement after SCI than either ChaseABC treatment or rehabilitative training alone (22). This suggests that enhanced plasticity alone is not sufficient to substantially improve skilled motor function, but needs to be combined with task-specific training. Notably, performance improved only on tasks that were specifically trained, whereas performance on tasks for which the injured animals were not trained worsened compared with that of animals that received no training at all (22). Presumably, increased neuronal plasticity coupled with task-specific training is necessary to drive activity-dependent strengthening of newly formed and functionally meaningful connections and, at the same time, to weaken or eliminate unused connections. The findings by García-Alías and colleagues are reminiscent of previous studies in the rat primary visual cortex where local infusion of ChaseABC is sufficient to restore experience-dependent plasticity after cessation of the critical period (Fig. 1) (18). Thus, use-dependent activation of specific neural circuits, either by task-specific training (22) or after monocular deprivation (18), in a network with increased plasticity allows formation and strengthening of functionally meaningful neuronal connections.

Fig. 1

(A) CSPGs belonging to the lectican family bind to RPTPσ. The ectodomain of LAR family members LAR, RPTPδ, and RPTPσ is composed of three immunoglobulin-like (Ig) repeats and eight fibronectin type III–like (FN3) repeats. The cytoplasmic portion contains a tandem pair of phosphatase domains (D1 and D2), of which only the membrane proximal domain (D1) is catalytically active. As does RPTPσ, mouse LAR and RPTPδ contain a cluster of conserved lysine residues in the first Ig-like domain, which suggests that they may all participate in CS-GAG binding. (B) Schematic of the mammalian visual system, showing the visual pathway of the left eye in purple and of the right eye in green through the lateral geniculate nuclei to the primary visual cortex (V1 and V2). The binocular zone receives input from both eyes. Accumulation of CSPGs in perineuronal nets correlates temporally with the closure of the critical period. Local infusion of ChaseABC near the binocular zone of the rat visual cortex (+ChaseABC) is sufficient to restore ocular dominance plasticity after the closure of the critical period. This indicates that CSPGs in perineuronal nets restrict experience-dependent neuronal plasticity in adulthood. Enzymatic degradation of CS-GAG chains removes nonpermissive substrates surrounding synapses and presumably facilitates rearrangement and formation of new synapses in favor of the nondeprived eye (18). (C) Injury to the brain or spinal cord results in axonal damage and formation of a glial scar at the injury site. The glial scar is composed of reactive astrocytes, microglia and macrophages, and meningeal fibroblasts and forms a barrier for regenerating axons. Injury to the dorsal column severs ascending axons of DRG neurons. In wild-type mice (Ptprs+/+), sensory fibers stall at the periphery of CSPG-rich scar tissue and show dystrophic end bulbs at their leading tips. In mutant mice (Ptprs−/−), reactive astrocytes produce inhibitory CSPGs; however, severed DRG axons grow deeper into glial scar tissue, which indicates that they are less sensitive to inhibitory CSPGs present at the lesion site in an injured mouse spinal cord (23).

Clearly, these are very exciting advances in the field of CNS regeneration. However, the underlying mechanisms by which degradation of CS-GAG chains results in enhanced neuronal plasticity remained largely elusive. The laboratories of John Flanagan and Jerry Silver now report on the identification of RPTPσ as a neuronal cell surface receptor that signals CSPG inhibition in vitro and that limits growth of severed axons into CSPG-enriched scar tissue after SCI in vivo (23). RPTPσ is a member of the leukocyte antigen–related (LAR) subfamily composed of immunoglobulin (Ig)–like and fibronectin type III (FNIII)–like repeats (Fig. 1). Additional vertebrate family members include RPTPδ and LAR (Fig. 1). The avian homolog of RPTPσ promotes intraretinal axon growth (24) and binds through its first Ig-like domain to heparan sulfate (HS)–GAG chains of agrin and collagen XVIII (25). Drosophila LAR (DLAR) functions as a synaptic receptor for the fly heparan sulfate proteoglycans (HSPGs) syndecan and dallylike (26, 27). These observations prompted Shen and colleagues to examine whether RPTPσ not only associates with HS-GAG chains, but also with structurally related CS-GAG chains. They struck gold, as RPTPσ bound with high affinity to the inhibitory CSPGs aggrecan and neurocan. Binding is sensitive to ChaseABC treatment and, similarly to the HS-GAG interaction, depends on the presence of a cluster of positively charged amino acids in the first Ig-like domain of RPTPσ (23, 25). Large amounts of RPTPσ are present in dorsal root ganglion (DRG) neurons throughout life. Functional studies showed that Ptprs mutant DRG neurons, but not wild-type DRG neurons, grow long neurites in the presence of purified neurocan, a mixture of neural CSPGs, or mouse astrocytes with high surface abundance of neurocan. In addition, RPTPσ is a receptor that specifically mediates the inhibitory responses of CSPGs, because MAG inhibition of Ptprs wild-type and mutant DRG neurons was strong and indistinguishable in vitro (23).

To examine whether RPTPσ participates in CSPG inhibition in vivo, Shen et al. performed a dorsal column crush injury in Ptprs wild-type and mutant mice. Two weeks after injury, the position of labeled sensory axons in the fasciculus gracilis was analyzed. In Ptprs mutant mice, sensory axons grew deeper into the proteoglycan-rich scar tissue than in wild-type mice, but failed to show robust regeneration beyond the lesion (23). This is reminiscent of the enhanced growth reported for corticospinal axons in animals having SCI treated with ChaseABC (20) and probably reflects the action of additional growth inhibitors that function in an RPTPσ-independent manner (8). Collectively, the study by Shen et al. convincingly demonstrates that neuronal RPTPσ is a receptor for CSPGs that mediates inhibitory responses toward glial scar tissue after SCI. Consistent with this conclusion, previous studies reported enhanced axonal regeneration in RPTPσ knockout (Ptprs–/–) mice following injury to the sciatic or facial nerve in the peripheral nervous system (PNS) (28, 29) or following injury to the optic nerve (30) and corticospinal tract in the CNS (31). Because the abundance of inhibitory CSPGs is increased near the site of injury in both the PNS and CNS, the work by Shen et al. now provides a mechanistic basis for the improved regeneration observed in various neuronal systems of Ptprs–/– mutant mice.

Is RPTPσ the only CSPG receptor? Because CSPGs are a large and diverse family of proteins, this appears unlikely. Other CSPG-binding proteins (32, 33) and guidance molecules regulated by CSPG association (34) have been identified; however, RPTPσ is the first receptor shown to directly mediate CSPG inhibition. It will be interesting to examine whether other members of the LAR family also participate in CSPG inhibition. Similar to RPTPσ, a cluster of conserved lysine residues is present in the first Ig-like domain of LAR and RPTPδ (Fig. 1), which indicates that these proteins may also interact with CSPGs. Do CSPGs and HSPGs compete for binding to RPTPσ, and does the protein core of proteoglycans influence binding specificity or receptor function? Moreover, functional redundancy in axonal pathfinding has been reported for RPTPσ and RPTPδ, suggesting that these two receptors may also collaborate in mediating growth inhibition following CNS injury (35).

More work is needed to examine whether loss of Ptprs mimics the effects of ChaseABC treatment when combined with task-specific training after SCI (22) or in the adult visual cortex following monocular deprivation (18). The LAR family RPTPs not only regulate axonal pathfinding (36), but also function in synaptogenesis (37) and activity-dependent synaptic plasticity (38). Considerable progress has been made in defining downstream signaling mechanisms of LAR family members. Some of the known pathways link LAR RPTPs to the neuronal cytoskeleton and thus are poised to mediate CSPG inhibitory responses. The cytoplasmic portion of LAR family members contains a conserved tandem pair of phosphatase domains. Only the membrane proximal phosphatase domain (D1) is catalytically active, whereas the membrane distal domain (D2) associates with (i) the guanine nucleotide exchange factor trio, (ii) the tyrosine kinase Abl, (iii) Abl’s substrate enabled, (iv) β-catenin, and (v) liprin-α (39). How exactly RPTPσ signaling is linked to molecules already implicated in CSPG inhibition, such as RhoA (40) or protein kinase C isoforms (41), is the subject of future studies.

Because of promising results with ChaseABC in experimental models of SCI, the identification of RPTPσ as a CSPG receptor is of great interest, not only biologically, but also clinically. New tools to antagonize CSPG inhibition may now be developed, including RPTPσ function–blocking antibodies, soluble PRTPσ ectodomain constructs, or small molecular compounds that block CS-GAG binding to the first Ig-like domain of RPTPσ. The phosphatase activity of RPTPσ, if involved in growth inhibition, is potentially an attractive target for drug design. Some of these new tools will likely be more specific and less invasive than intrathecal infusion of bacterial ChaseABC and could be combined with strategies to overcome myelin inhibitors or to activate neuron intrinsic growth programs. Any of these treatment strategies will need to be combined with task-specific rehabilitative training, as activity is essential to shape and reinforce functionally meaningful connections. RPTPσ is evolutionarily conserved, and it will be important to define the role of RPTPσ in limiting human CNS plasticity. If a drug can be found to effect transient blockage of human RPTPσ, this route may provide a new strategy to enhance neuronal plasticity and regeneration following CNS injury or disease.

Acknowledgments

R.J.G. is funded by the National Institute of Neurological Disorders and Stroke, NIH (R01-047333); the U.S. Department of Veterans Affairs; and the New York State Department of Health, Spinal Cord Injury Research Program.

References and Notes

View Abstract

Stay Connected to Science Signaling

Navigate This Article