Phosphorylation of CRMP-2 correlates with axonal degeneration in EAE. (A) Immunoprecipitation of pThr555 CRMP-2 using our polyclonal anti-pThr555 CRMP-2 antibody, of lumbosacral spinal cord lysates, dissected from wild-type MOG_35–55-induced EAE, uninjected control or adjuvant injected alone control mice, followed by western immunoblot analysis using the monoclonal anti-CRMP-2 antibody. Control wells include non-specific rabbit IgG polyclonal antibody and pThr555 CRMP-2 (20 µg) peptide spiked in lysate buffer alone. The membranes were then reprobed using the monoclonal anti-CRMP-2 antibody. (B) Western immunoblot for CRMP-2 from spinal cord lysate samples loaded pre-immunoprecipitation (5% input of immunoprecipitation sample) using the monoclonal anti-CRMP-2 antibody. (C) Densitometric quantification (AU) of total CRMP-2 and pThr555 CRMP-2B (after immunoprecipitation) from spinal cord lysates of control and EAE-induced mice (n = 4 mice per group and per time-point; *P < 0.05 one-way ANOVA). (D) Co-immunoprecipitation of CRMP-2-associated tubulin was performed from control and EAE-induced spinal cord lysates using monoclonal anti-CRMP-2 antibody followed by western immunoblotting with rat anti-mouse α-tubulin. Total loading levels of tubulin for each spinal cord sample pre-immunoprecipitation were demonstrated by loading 5% input of total protein (from the same samples illustrated in this figure pre-immunoprecipitation) and then probing the western transfer membrane with the rat anti-mouse α-tubulin monoclonal antibody. (E) Polyclonal anti-pThr555 CRMP-2 antibody reactivity was demonstrated by double immunofluorescence staining on 10 µm cryostat sections showing co-localization with degenerating spinal cord axons from EAE-induced mice immunostained with the monoclonal βIII-tubulin antibody (disassembled, monomeric tubulin; arrows) in axons. DAPI stain shows mononuclear cell infiltrate surrounding lesion (arrowheads). Scale bar = 20 µm. dpi = days post-injection; IP = immunoprecipitation.

Localization of pThr555CRMP-2 in degenerating neuron somata and axons in chronic-active multiple sclerosis lesions. (A) Chronic-active demyelinating multiple sclerosis lesion in the dorsolateral white matter of the lumbar spinal cord stained with Luxol fast blue and Periodic acid Schiff (scale bar = 50 µm). (B) Axonal degeneration (blue box and arrowhead) in the dorsolateral white matter of the lumbar spinal cord from the same chronic-active multiple sclerosis case stained with Bielschowsky stain (scale bar = 50 µm). (C) Axonal retraction bulb (red arrowhead) present in the dorsal horn of the same chronic-active multiple sclerosis case stained with Bielschowsky (scale bar = 50 µm). (A–C) Drawing of a cross-section through the lumbar spinal cord to illustrate the white matter tracts analysed (solid red circle) and the ventral grey matter lesions (solid black circle; D–F). (D) Ventral horn degenerative neurons of the lumbar spinal cord stained with Luxol fast blue and Periodic acid Schiff, (E) Bielschowsky stain and (F) Luxol fast blue and Periodic acid Schiff staining, showing perivenular cuffing and infiltrates in the ventral grey matter (arrow). Scale bar = 50 µm. (G–J) Chronic-active demyelinating multiple sclerosis lesion of the lumbar spinal cord counterstained with (G) DAPI, immunostained for (H) NF200 and (I) pThr555CRMP-2; and the merged image (J) (scale bar = 50 µm). (K–M) The same chronic-active demyelinating multiple sclerosis lesion of the lumbar spinal cord with a closer view of the ventral horn grey matter, counterstained with (K) DAPI, immunostained for (L) NF200, (M) pThr555CRMP-2 and (N) the merged image (arrow showing axonal retraction bulb). Scale bar = 50 µm. (O and P) The same chronic-active demyelinating multiple sclerosis lesion of the lumbar spinal cord with (O) secondary anti-mouse Alexafluor® 488 antibody control or (P) secondary anti-rabbit Alexafluor® 555 antibody control (scale bar = 50 µm). (Q–Si) Degenerative axons in the dorsolateral spinal cord in longitudinal (arrow) and cross-section (arrowheads) immunostained for (Q) hyperphosphorylated tau AT8, (R) pThr555CRMP-2 and (S) merged image and (Si) magnified image of the retraction bulb (scale bar = 50 µm). (T–W) Ventro-lateral region of the same chronic-active demyelinating multiple sclerosis lesion of the lumbar spinal cord with (T) DAPI counterstain, (U) CD3-positive T cell infiltrates near (V) pThr555CRMP-2 positive axons; (W) merged image.

Reduction in EAE disease, axonal degeneration and phosphorylation of CRMP-2 in ngr1^−/− mice. (A) EAE clinical scores after MOG_35–55-inducion in female C57Bl/6 wild-type littermate (ngr1^+/+, n = 21, black line) and ngr1^−/− mice (n = 14 ngr1^−/−, red line; ***P < 0.001 repeated measures ANOVA). (B) Histopathology of EAE in ngr1^+/+ and ngr1^−/− mice at pre-onset, onset, peak and chronic stages. Luxol fast blue (LFB) and Periodic acid Schiff (PAS) stain showing inflammatory demyelination and Bielschowsky stain to demonstrate axonal degeneration/loss (arrows show axonal swelling; arrowhead shows an axonal retraction bulb; scale bar = 20 µm). (C) Representative semi-thin (1 µm) lumbosacral spinal cord sections from EAE-induced ngr1^+/+ and ngr1^−/− mice demonstrating significant axonal degeneration in the dorsal column (DC) white matter tracts of ngr1^+/+ mice (arrows) at peak of disease. Inflammatory cells are also present in ngr1^+/+ mice (arrowheads). Scale bar = 50 µm. (D) Immunoprecipitation of pThr555CRMP-2 of spinal cord lysates from EAE-induced ngr1^+/+ and ngr1^−/− mice, followed by reprobing of the membranes with the monoclonal anti-CRMP-2 antibody. There is an increase in pThr555 CRMP-2 of both the CRMP-2A and CRMP-2B variants in EAE-immunized ngr1^+/+ mice from pre-onset to peak of disease. No such changes occurred in the ngr1^−/− mice at the same time-points after MOG_35–-55 challenge. Control lanes include non-specific rabbit IgG polyclonal antibody and pThr555 CRMP-2 (20 µg) peptide spiked in lysate buffer alone. A pre-immunoprecipitation 5% input of protein from spinal cord lysates show no discernable difference in total levels of CRMP-2. (E) Densitometric quantification of total CRMP-2 and pThr555 CRMP-2B (after immunoprecipitation) from spinal cord lysates of control, ngr1^+/+ and ngr1^−/− EAE-induced mice, represented as a percentage of basal levels (control uninjected ngr1^−/− mice; n = 4 mice per group and per time-point; *P < 0.05 one-way ANOVA). Representation of the percentage change in the levels of pThr555CRMP-2 demonstrates up to a 4-fold increase occurring in ngr1^+/+ EAE-induced mice when compared with ngr1^−/− mice at peak stage of disease. (F) Co-immunoprecipitation of CRMP-2 from spinal cord lysates of control, ngr1^+/+ and ngr1^−/− EAE-induced mice, showed a decreased association with tubulin in the ngr1^+/+ mice from pre-onset until peak stage of EAE. This finding was not replicated in the ngr1^−/− EAE-induced mice. (G) Densitometric quantification of the levels of CRMP-2-associated tubulin (after immunoprecipitation) from spinal cord lysates of control, ngr1^+/+ and ngr1^−/− EAE-induced mice, represented as a percentage of basal levels (n = 4 mice per group and per time-point; *P < 0.05 one-way ANOVA). Representation of the percentage change in the levels of CRMP-2-bound tubulin, demonstrates up to a 3-fold increase in the dissociation of tubulin from CRMP-2 in ngr1^+/+ when compared with ngr1^−/− EAE-induced mice at peak stage of disease (*P < 0.05). IP = immunoprecipitation.

Axonal and myelin damage in EAE-induced ngr1^−/− mice is reduced and corresponds with abrogated CRMP-2 phosphorylation. (A) Semi-thin (1 µm) optic nerve sections of ngr1^+/+ and ngr1^−/− mice at pre-onset, onset, peak and chronic stages of EAE or adjuvant only injected control mice. Axonal degeneration and demyelination in the ngr1^+/+ EAE-induced mice are present throughout disease, becoming more prevalent at peak (arrows), with significant axonal loss at chronic stage. Perineurial inflammatory infiltrates are a ubiquitous finding in the optic nerves of ngr1^+/+ EAE-mice (arrowhead). These pathological findings are not as evident in the ngr1^−/− mice. Scale bar = 50 µm. (B) Increased demyelination of optic nerves in ngr1^+/+ mice during the course of EAE as determined by G-ratios. At pre-onset and onset of EAE, respectively in ngr1^−/− mice (n = 3 mice per genotype, per disease stage and per animal), there is a reduction nearing basal levels (adjuvant injected controls). By peak stage of EAE, there is an increase in the G-ratios (i.e. axonal degeneration/demyelination) of ngr1^+/+ mice. The percentage of demyelination is significantly reduced in the ngr1^−/− after EAE-induction, except at pre-onset stage of disease. Axonal diameters in the optic nerves are increased in ngr1^+/+ mice following EAE-induction. The greatest axonal diameter was demonstrated by peak stage of EAE in ngr1^+/+ mice. In the ngr1^−/− mice, axonal calibre remains relatively unaltered, with only a modest increase in the axonal diameter by peak stage of disease (***P < 0.001, **P < 0.01 and *P < 0.05 for ngr1^+/+ versus ngr1^−/− at all time points, analysed by ANOVA with Tukey’s post hoc test). (C) Number of viable myelinated axons in optic nerve semi-thin sections from ngr1^+/+ and ngr1^−/− mice during the course of EAE (n = 3 mice per genotype, per disease stage and per animal). A significant reduction in axonal viability is demonstrated in the ngr1^+/+ by peak stage of EAE (**P < 0.01, analysed by one-tailed student’s t-test). (D) Immunoprecipitation of pThr555CRMP-2 of pooled optic nerve lysates from EAE-induced ngr1^+/+ and ngr1^−/− mice (n = 6 per group), followed by reprobing of the membranes with the monoclonal anti-CRMP-2 antibody. There is an increase in pThr555 CRMP-2 of both the CRMP-2A and CRMP-2B variants in EAE-immunized ngr1^+/+ mice at peak of disease. No such changes occurred in the ngr1^−/− mice at the same time-points after MOG_35–55 challenge. Control lanes include non-specific rabbit IgG polyclonal antibody and pThr555 CRMP-2 (20 µg) peptide spiked in lysate buffer alone. A pre-immunoprecipitation, 5% input of protein from spinal cord lysates show no discernible difference in total levels of CRMP-2. Densitometric quantification of total CRMP-2 and pThr555 CRMP-2B (after immunoprecipitation) from optic nerve lysates of control, ngr1^+/+ and ngr1^−/− EAE-induced mice, represented as a percentage of basal levels (control uninjected ngr1^−/− mice; n = 6 mice per group; *P < 0.05 one-way ANOVA, **P < 0.01). Representation of the percentage change in the levels of pThr555CRMP-2 demonstrates up to a 2-fold increase occurring in ngr1^+/+ EAE-induced mice when compared with ngr1^−/− mice at peak stage of disease. (E, top) Co-immunoprecipitation of CRMP-2 from optic nerve lysates of control, ngr1^+/+ and ngr1^−/− EAE-induced mice, showed a decreased association with tubulin in the ngr1^+/+ mice at peak stage of EAE. This finding was not replicated in the ngr1^−/− EAE-induced mice. Left: Densitometric quantification of the levels of CRMP-2-associated tubulin (after immunoprecipitation) from spinal cord lysates of control, ngr1^+/+ and ngr1^−/− EAE-induced mice, represented as a percentage of basal levels (n = 6 mice per group; *P < 0.05 one-way ANOVA). Right: Representation of the percentage change in the levels of CRMP-2-bound tubulin, demonstrates an increase in the dissociation of tubulin from CRMP-2 in ngr1^+/+ when compared with ngr1^−/− EAE-induced mice at peak stage of disease (*P < 0.05).

Limiting the phosphorylation of CRMP-2 through a rAAV2 intravitreal delivery system in EAE-induced mice maintains optic nerve axonal integrity. (A) Map of the vector construct inserted into a rAAV2 (GFP and Flag-T555ACRMP-2, respectively). (B) Seven days following intravitreal injections of rAAV2-GFP or rAAV2-Flag-CRMP2-GFP (n = 8 mice per group), or in uninjected control mice (n = 5), EAE was induced using the MOG_35–55 peptide and then followed until all mice reached peak stage of disease (*P < 0.05 analysed by Friedman’s non-parametric repeated measures). (C and D) Representative retinal whole mounts immunolabelled for GFP, in rAAV2-GFP and rAAV2-Flag-CRMP2-GFP injected mice, respectively. (E) The percentage of retinal ganglion cells (RGC) transduced with either rAAV2-GFP or rAAV2-Flag-CRMP2-GFP. (Fi) rAAV2-GFP transduced optic nerve axons demonstrating localization of GFP particularly in degenerative/transected axons with evident retraction bulbs (arrows). (Fii) Immunolabelling of pThr555CRMP-2 of axons (arrows) in the same optic nerve. (Fiii) Co-labelling of GFP and pThr555CRMP-2 degenerative axons in the same optic nerve (arrows) near DAPI-positive inflammatory infiltrates (arrowheads). (Gi) rAAV2-Flag-CRMP2-GFP transduced optic nerve axons demonstrating localization of Flag in normal appearing axons with (arrow). (Gii) Immunolabelling for pThr555CRMP-2 of axons (arrows), in the same optic nerve. (Giii) Lack of co-labelling of Flag and pThr555CRMP-2-positive axons in the same optic nerve near DAPI-positive inflammatory infiltrates (arrowheads). (H) Estimated numbers of transduced and morphologically degenerative axons per mm² of optic nerve tissue in either rAAV2-GFP or rAAV2-Flag-CRMP2-GFP transduced mice (left, **P < 0.002, analysed by one-tailed student’s t-test). The percentage of degenerative transduced axons in optic nerves of rAAV2-GFP or rAAV2-Flag-CRMP2-GFP transduced mice demonstrating a 10-fold reduction in the later (right, **P = 0.008, analysed by one-tailed student’s t-test). (I) Profound axonal retraction bulbs/spheroids in rAAV2-GFP (arrows) compared with more linear intact axons in rAAV2-Flag-CRMP2-GFP transduced nerves. (J and K) Amyloid precursor protein-positive axons in optic nerves from EAE-induced mice, with and without rAAV2-GFP injection (arrows), co-labelling with pThr555 CRMP-2 (arrows), near DAPI-positive inflammatory infiltrates. (L) Reduced amyloid precursor protein-immunopositive axons in rAAV2-Flag-CRMP2-GFP transduced nerves (arrows). Scale bar = 1 mm for low magnification and 50 µm for high magnification.

Therapeutic administration of anti-Nogo-A antibodies prevents clinical progression, axonal degeneration and CRMP-2 phosphorylation in EAE. (A) Combination of intravenous (Days 8, 11 and 14) and intraperitoneal (Days 9 and 16) administration of anti-Nogo(623-640) antibody (1 mg per injection), non-specific IgG control antibody or no injection in EAE-induced mice (n = 8 mice per group, ***P < 0.001, *P < 0.05 analysed by Friedman’s non-parametric repeated measures). (B) Histopathological analysis of inflammatory demyelination [Luxol fast blue and Periodic acid Schiff (LFB/PAS) stain] and axonal degeneration (Bielschowsky silver stain) at chronic stage of EAE after either, no treatment regime, IgG or anti-Nogo(623–640) antibody treatment. Staining with Luxol fast blue and Periodic acid Schiff highlights significant sparing of myelin in the Nogo(623–640) antibody treated spinal cord white matter tracts compared to the demyelination in EAE-induced mice with non-specific IgG or without treatment. Importantly, axonal degeneration is also substantially reduced in the Nogo(623–640) antibody treated group. Scale bar = 20 µm. (C) Immunoprecipitation of pThr555CRMP-2, demonstrating increased levels in spinal cord lysates from EAE-mice without treatment or treated with IgG antibody. However, substantial reduction is evident following Nogo(623–640) antibody treatment. Co-immunoprecipitation of CRMP-2 shows reduced association of tubulin to CRMP-2 in EAE-induced mice treated with non-specific IgG or without treatment, but CRMP-2-bound tubulin levels are higher in the Nogo(623–640) antibody treatment group. (D) Optical density measurements of protein levels show a significant reduction in pThr555CRMP-2 (*P < 0.05) that correlate with an increase in tubulin association to CRMP-2 in the Nogo(623–640) antibody treatment group (n = 8 mice per group).