RGMa inhibition promotes axonal growth and recovery after spinal cord injury

Abstract

Repulsive guidance molecule (RGM) is a protein implicated in both axonal guidance and neural tube closure. We report RGMa as a potent inhibitor of axon regeneration in the adult central nervous system (CNS). RGMa inhibits mammalian CNS neurite outgrowth by a mechanism dependent on the activation of the RhoA-Rho kinase pathway. RGMa expression is observed in oligodendrocytes, myelinated fibers, and neurons of the adult rat spinal cord and is induced around the injury site after spinal cord injury. We developed an antibody to RGMa that efficiently blocks the effect of RGMa in vitro. Intrathecal administration of the antibody to rats with thoracic spinal cord hemisection results in significant axonal growth of the corticospinal tract and improves functional recovery. Thus, RGMa plays an important role in limiting axonal regeneration after CNS injury and the RGMa antibody offers a possible therapeutic agent in clinical conditions characterized by a failure of CNS regeneration.

Figure 1.

Effects of membrane bound/diffusible RGMa on neurons. (A) The membrane bound form of RGMa inhibits neurite outgrowth by a Rho-kinase–dependent mechanism. The cerebellar neurons were cultured for 24 h on control CHO cells (control) or RGMa-CHO cells (RGMa) in the presence or absence of 10 μM Y27632. (B) Mean length of the longest neurite per neuron. *, P < 0.01 compared with the control; **, P < 0.01 compared with RGMa (t test). (C) RGMa activates RhoA. The cerebellar neurons were treated for 10 min with conditioned media, derived from PI-PLC treatment of control CHO or RGMa-CHO cells. The active fraction of RhoA along with total RhoA was detected by Western blot. (D) Immunostaining of the supernatant from control/RGMa CHO cells treated with or without PI-PLC. The protein was detected by an anti-HA antibody. Coomassie staining was performed for the same samples. (E) The soluble form of RGMa inhibits neurite outgrowth. Neurons were cultured on PLL-coated chamber slides for 12 h in the conditioned media from untreated control CHO cells, PI-PLC–treated control CHO cells, untreated RGMa-CHO cells, or PI-PLC–treated RGMa-CHO cells. (F) Mean length of the longest neurite per neuron. *, P < 0.01 compared with the control CHO PI-PLC(+) (t test). (G) PI-PLC treatment reverses the inhibitory effect of RGMa-CHO cells. Neurons were cultured for 24 h on control/RGMa CHO cells pretreated with or without PI-PLC. (H) Mean length of the longest neurite per neuron. *, P < 0.01 compared with the control CHO PI-PLC(−); **, P < 0.01 compared with RGMa-CHO PI-PLC(−) (t test). There is no significant difference between the control CHO PI-PLC(+) and RGMa-CHO PI-PLC(+). (I–L) RGMa inhibits neurite outgrowth by an NgR-independent mechanism. (I) The cerebellar neurons were cultured for 24 h on control/RGMa CHO cells in the presence or absence of 1 μM NEP1-40. (J) Mean length of the longest neurite per neuron. *, P < 0.01 compared with the control; **, P < 0.01 compared with the control + NEP1-40 (t test). There is no significant difference between RGMa and RGMa+NEP1-40. (K) The neurons were cultured on PLL-coated chamber slides for 24 h in the conditioned media from PI-PLC–treated control/RGMa CHO cells in the presence or absence of 1 μM NEP1-40. (L) Mean length of the longest neurite per neuron. *, P < 0.01 compared with the control; **, P < 0.01 compared with the control + NEP1-40 (t test). There is no significant difference between RGMa and RGMa + NEP1-40. Data are represented as the mean ± SEM of three independent experiments.

Figure 2.

The expression of RGMa is induced in response to SCI. (A) Western blot showing the expression of RGMa in purified CNS myelin. (top) Purified myelin and the lysates from control/RGMa CHO cells were blotted with the anti-RGMa antibody. (bottom) CSPG, purified MAG-Fc, and purified myelin were blotted with the anti-RGMa antibody. (B–E) RGMa is expressed in oligodendrocytes and neurons in the rat spinal cord. Fresh frozen tissues were obtained for immunohistochemistry from an uninjured adult rat spinal cord. Parasagittal sections were stained with anti-RGMa (B) or double stained with anti-GFAP and anti-RGMa (C), anti-MOSP and anti-RGMa (D), or anti-Tuj1 and anti-RGMa (E). (C) Double labeling with GFAP and RGMa antibodies shows no colocalization. (D) Double staining with the anti-MOSP and anti-RGMa antibody demonstrates that RGMa is expressed in oligodendrocytes. (E) RGMa immunoreactivity was localized to the somata of Tuj1-positive neurons in the gray matter but not to their axons. (F–L) Western blot analysis and immunohistochemistry reveal up-regulated expression of RGMa after SCI. (F) Western blot demonstrates that expression of RGMa is enhanced in the spinal cord tissue 7 d after injury. The protein was detected by an anti-RGMa or anti-actin antibody. An uninjured animal served as the control. (G) Injured spinal cord is illustrated schematically. A dorsal hemisection was performed at Th9/10. Parasagittal fresh frozen sections of an uninjured control spinal cord and those at various postoperative intervals were obtained and used for a time-course study of the immunoreactivity to RGMa in the epicenter area (H) and in the white matter, 400–500 μm rostral to the epicenter area (I). (J) Double staining of the same area as H with IB4 and anti-RGMa reveals the presence of RGMa-expressing microglia/macrophages. (K) Double staining of the same area as I with anti-MOSP and anti-RGMa reveals the presence of double-labeled cells (arrows) and non–double-labeled cells (RGMa-positive and MOSP-negative cells; arrowheads). (L) Quantification of the number of the cells expressing RGMa and those coexpressing RGMa and MOSP. The expression was examined 7 d after SCI. The x axis indicates specific locations along the rostrocaudal axis of the spinal cord. Data are represented as the mean ± SEM of four animals. Bars: (B and H) 100 μm; (C–E and I–K) 20 μm.

Figure 3.

The anti-RGMa antibody neutralizes the effect of RGMa and CNS myelin in vitro. (A) The cerebellar neurons were cultured for 24 h on control/RGMa CHO cells in the presence or absence of 10 μg/ml of the control rabbit IgG or anti-RGMa antibody. (B) Mean length of the longest neurite per neuron. *, P < 0.01 compared with the control; **, P < 0.01 compared with the control + control IgG; ***, P < 0.01 compared with RGMa (t test). There is no significant difference between the control + anti-RGMa and RGMa + anti-RGMa. (C) Neurons were cultured on PLL-coated chamber slides for 24 h in conditioned media from PI-PLC–treated control CHO/RGMa CHO cells supplemented with or without 10 μg/ml of the control IgG or anti-RGMa antibody. (D) Mean length of the longest neurite per neuron. *, P < 0.01 compared with the control; **, P < 0.01 compared with the control + control IgG; ***, P < 0.01 compared with RGMa (t test). There is no significant difference between the control + anti-RGMa and RGMa + anti-RGMa. (E) Neurons were cultured on PLL-coated (control) or PLL + myelin-coated (myelin) chamber slides for 24 h in the presence or absence of 10 μg/ml of the anti-RGMa antibody. (F) Mean length of the longest neurite per neuron. *, P < 0.01 compared with the control; **, P < 0.01 compared with myelin; ***, P < 0.01 compared with the control + anti-RGMa. (G) Immunodepletion of purified CNS myelin. CNS myelin, depleted myelin, and the protein A beads after immunodepletion were blotted with the anti-RGMa antibody. (H) Neurons were cultured on PLL-coated (control), PLL + myelin-coated (myelin), and PLL + depleted myelin-coated (d-Myelin) chamber slides for 24 h. (I) Mean length of the longest neurite per neuron. *, P < 0.01 compared with the control; **, P < 0.03 compared with myelin; ***, P < 0.04 compared with the control. Data are represented as the mean ± SEM of three independent experiments.

Figure 4.

The anti-RGMa antibody promotes locomotor recovery after SCI. (A) Quantification of the lesion depth revealed that there was no statistically significant difference between the control and anti-RGMa antibody–treated animals. The significance was determined by t test. Mean ± SEM of nine anti-RGMa antibody–treated and 11 control rats. (B) The BBB score was determined at the indicated times after thoracic dorsal hemisection in the anti-RGMa antibody–treated, control antibody–treated, and sham-operated rats. Mean ± SEM of 9, 11, and 5 rats for each group, respectively. The anti-RGMa antibody–treated group is statistically different from the control group, as indicated. *, P < 0.05 compared with the control; **, P < 0.06 compared with the control.

Figure 5.

The anti-RGMa antibody promotes the regeneration/sprouting of CST axons after SCI. (A and B) Representative pictures of BDA-labeled CST fibers; rostral is on the left. Anterograde-labeled CST fibers in control IgG–treated (A) or anti-RGMa antibody–treated (B) spinal cord 10 wk after injury. The epicenter of the lesion is indicated by an asterisk. (C–G) Higher magnification of the boxed regions in A and B, showing increased collateral CST fiber sprouting rostrally (D), as well as regenerating fibers at the lesion site (F, arrows), in rats treated with the anti-RGMa antibody but not in the corresponding regions of the control IgG–treated rats (C and E). (H) Different sections from the same animal (B), showing regenerating/sprouting fibers in the caudal part of the spinal cord. (I–R) Serial microscopic images of an anti-RGMa antibody–treated animal. Asterisks indicate the epicenter of the lesion. Arrows denote BDA-positive fibers. Bars: (A, B, and I–R) 500 μm; (C–F and H) 100 μm; (G) 200 μm.

Figure 6.

Quantification of the extent of regeneration/sprouting in anti-RGMa antibody– and control IgG–treated rats. (A) The distance from the end of the main CST bundles to the lesion epicenter 10 wk after SCI in rats treated with the anti-RGMa antibody or control IgG (n = 6–8 per group). It was measured under low-magnification view of the longitudinal sections. *, P < 0.01 compared with the control. The anti-RGMa antibody suppresses the retraction of the injured main CST. (B) Quantification of the labeled CST fibers in six anti-RGMa antibody– and eight control IgG–treated animals. All the 50-μm-thick serial parasagittal sections were evaluated. The x axis indicates specific locations along the rostrocaudal axis of the spinal cord. The y axis indicates the ratio of the number of BDA-labeled fibers at the indicated site to those at 4 mm rostral to the lesion site. *, P < 0.05 compared with the control. Error bars indicate SEM.

Figure 7.

Distribution of the regenerating/sprouting CST fibers in the spinal cord. (top) The injured spinal cord is illustrated schematically. (A–H) Representative transverse sections of the spinal cord taken from 10 mm rostral (A and B), 10 mm caudal (F–H), or adjacent (C–E) to the lesioned site from control IgG–treated (A, C, and F) and anti-RGMa antibody–treated (B, D, E, G, and H) rats. Insets of A and B show the high-magnification view of the dorsal CST. (E) High-magnification view in D shows fibers in the developing scar tissue at the lesion site. Insets of F and G illustrate that no fibers are seen in the ventral part of the dorsal column (top right) or dorsolateral column (bottom right) 10 mm caudal to the lesion site. (H) High-magnification view in G shows increased regenerated fibers with tortuous appearance and unusual branching in the gray matter (inset) 10 mm caudal to the lesion site in rats treated with the anti-RGMa antibody. No labeled fibers are observed 10 mm caudal to the lesion site in the corresponding region of the control IgG-treated rats (F). (I) The number of labeled corticospinal axons 10 mm rostral to the lesioned site in rats treated with the control IgG or anti-RGMa antibody. No significant difference is observed. (J) The number of labeled corticospinal axons 10 mm caudal to the lesioned site in rats treated with the control IgG or anti-RGMa antibody. Cross-sections of the spinal cord taken from 10 mm caudal to the lesioned site were examined, and the number of the labeled fibers was estimated in the gray matter (GM), the normal locations of the dorsal CST (CST), or the white matter other than the normal locations of the dorsal CST (WM). Data based on 11 control and 9 anti-RGMa antibody–treated rats. Increase in the regenerating fibers caudal to the lesioned site in the anti-RGMa antibody–treated rats is seen mainly in the gray matter. A small amount of nonspecific staining accounts for a few fiber counts in the CST (no apparent BDA-positive fibers were detected in this area). *, P < 0.01 compared with the control. (K and L) Camera lucida drawings illustrating the distribution of BDA-labeled fibers in 10 serial sections from the cases illustrated in F and G, respectively. Bars: (A–D, F, and G) 500 μm; (A, B, F, and G, insets) 200 μm; (E) 50 μm; (H) 200 μm; (H, inset) 50 μm.

Figure 8.

Anatomical analysis of the spinal cords 3 and 5 wk after SCI. (A and B) Representative pictures of anterograde-labeled CST fibers in rat spinal cord treated with the control IgG (A) or anti-RGMa antibody (B) 5 wk after injury; rostral is on the left. Asterisks indicate the lesion epicenter. Higher magnification of the boxed regions in A and B, showing regenerating fibers, with typical irregular meandering growth patterns around the cysts at the lesion site (D) in rats treated with the anti-RGMa antibody but not in the corresponding regions of the control IgG–treated rats (C). In both the groups, no BDA-traced fibers could be observed in the caudal part of the spinal cord 5 wk after injury. (E) Quantification of the labeled CST fibers in the anti-RGMa antibody–treated animals 3 wk (n = 3) and 5 wk (n = 3) after injury and in the control IgG–treated animals 3 wk (n = 3) and 5 wk (n = 3) after injury. All the 50-μm-thick serial parasagittal sections were evaluated. The x axis indicates specific locations along the rostrocaudal axis of the spinal cord. The y axis indicates the ratio of the number of BDA-labeled fibers at the indicated site to those at 4 mm rostral to the lesion site. *, P < 0.05 compared with the control (5W). Bars: (A and B) 500 μm; (C and D) 100 μm.