One proline deletion in the fusion peptide of neurotropic mouse hepatitis virus (MHV) restricts retrograde axonal transport and neurodegeneration

Mouse hepatitis virus (MHV; murine coronavirus) causes meningoencephalitis, myelitis, and optic neuritis followed by axonal loss and demyelination. This murine virus is used as a common model to study acute and chronic virus-induced demyelination in the central nervous system. Studies with recombinant MHV strains that differ in the gene encoding the spike protein have demonstrated that the spike has a role in MHV pathogenesis and retrograde axonal transport. Fusion peptides (FPs) in the spike protein play a key role in MHV pathogenesis. In a previous study of the effect of deleting a single proline residue in the FP of a demyelinating MHV strain, we found that two central, consecutive prolines are important for cell–cell fusion and pathogenesis. The dihedral fluctuation of the FP was shown to be repressed whenever two consecutive prolines were present, in contrast to the presence of a single proline in the chain. Using this proline-deleted MHV strain, here we investigated whether intracranial injection of this strain can induce optic neuritis by retrograde axonal transport from the brain to the retina through the optic nerve. We observed that the proline-deleted recombinant MHV strain is restricted to the optic nerve, is unable to translocate to the retina, and causes only minimal demyelination and no neuronal death. We conclude that an intact proline dyad in the FP of the recombinant demyelinating MHV strain plays a crucial role in translocation of the virus through axons and subsequent neurodegeneration.


Mouse hepatitis virus (MHV; murine coronavirus) causes meningoencephalitis, myelitis, and optic neuritis followed by axonal loss and demyelination. This murine virus is used as a common model to study acute and chronic virus-induced demyelination in the central nervous system. Studies with recombinant MHV strains that differ in the gene encoding the spike protein have demonstrated that the spike has a role in MHV pathogenesis and retrograde axonal transport. Fusion peptides (FPs) in the spike protein play a key role in MHV pathogenesis.
In a previous study of the effect of deleting a single proline residue in the FP of a demyelinating MHV strain, we found that two central, consecutive prolines are important for cell-cell fusion and pathogenesis. The dihedral fluctuation of the FP was shown to be repressed whenever two consecutive prolines were present, in contrast to the presence of a single proline in the chain. Using this proline-deleted MHV strain, here we investigated whether intracranial injection of this strain can induce optic neuritis by retrograde axonal transport from the brain to the retina through the optic nerve. We observed that the prolinedeleted recombinant MHV strain is restricted to the optic nerve, is unable to translocate to the retina, and causes only minimal demyelination and no neuronal death. We conclude that an intact proline dyad in the FP of the recombinant demyelinating MHV strain plays a crucial role in translocation of the virus through axons and subsequent neurodegeneration. MHV 3 has been used extensively as a model to study acute and chronic virus-induced demyelination in the central nervous system for many years. Isogenic recombinant strains of MHV (RSA59, demyelinating strain; RSMHV2, nondemyelinating strain), which differ only in the spike gene generated by targeted RNA recombination between the parental demyelinating strain MHV-A59 and parental nondemyelinating strain MHV-2, revealed that the spike is important for pathogenesis (1). Intracranial inoculation of RSA59 and RSMHV2 has revealed their differential ability to induce macrophage infiltration, demyelination, and axonal loss in the spinal cord (2,3).
Interneuronal transport of viruses and cell-cell fusion can be critical checkpoints in mechanisms of demyelination and axonal loss in the spinal cord and in the optic nerve. The spike protein plays a critical role in virus-cell fusion, leading to infection, and pathological cell-cell fusion for spread, and the spike can be a suitable target for antiviral therapy and virus-induced demyelination (2,4,5). Preventing spike protein interaction with the host cell may lead to the design of therapeutic agents, but detailed studies of the fusogenic determinant (minimal motif) of the spike protein and its mechanistic pathways have yet to be performed. Previous studies have shown that the minimal motif, i.e. fusion peptide, plays an important role in cellcell fusion (6 -8) and that it includes two consecutive proline amino acids that putatively play a key role in the structure and function of the fusion peptide (8). Alteration of the central proline amino acids in the putative fusion peptide impairs the fusogenic property of RSA59, and subsequently alteration of fusogenicity alters the demyelination pathology in the spinal cord during the chronic stage of infection (8).
Site-directed mutagenesis of proline and generation of recombinant viruses revealed that one proline deletion, RSA59 (P) alters viral replication and also causes delayed syncytia formation from cell-cell fusion compared with RSA59 (PP), which contains the nonmutated parental strain MHV-A59 fusion protein (8). Altered fusogenicity consequently alters viral spread and pathogenicity. Acute-stage pathology demonstrated that RSA59 (PP) was able to spread widely during acute encephalitis from the brain to the spinal cord and cause demyelination, whereas, in the proline mutated strain RSA59 (P), viral antigen spread was mainly restricted to the inoculation site and meninges and showed reduced demyelination (8).
MHV has been reported to spread along neuronal axons in anterograde and retrograde directions (3,9,10). Viruses can reach and spread to proximal oligodendrocytes and myelin via anterograde axonal transport, resulting in demyelination in spinal cord white matter. Earlier studies have suggested that demyelinating MHV strains reach the optic nerve via retro-grade axonal transport (9, 10), but whether the fusogenicity of viruses mediates such transport is not well defined. In the optic nerve, the parental demyelinating strains MHV-A59 and RSA59 cause inflammation, demyelination, and axonal loss (i.e. optic neuritis), in contrast to the nondemyelinating parental MHV-2 and RSMHV2 strains (11,12), with clear evidence of retrograde viral spread in an MHV model of optic neuritis (11). The study demonstrated the ability of demyelinating MHV strains, but not nondemyelinating strains, to induce optic neuritis and axonal injury by retrograde axonal transport. Also, RSA59, in contrast with RSMHV2, can infect retinal ganglion cell (RGC) bodies in the retina following intracranial inoculation, and at the chronic stage, RSA59-infected mice show significant loss of RGCs (13,14).
Whether retrograde transport of MHV to induce optic neuritis and damage RGCs depends on the capability of a virus to cause cell-cell fusion is not known. Here we compared the incidence and phenotype of optic neuritis after inoculation with fusiogenic RSA59 (PP) and its fusion-deficient, prolinemutated RSA59 (P) strain and assessed retrograde axonal transport and RGC loss. RSA59 (PP), with direct capability of cellcell fusion, was able to reach the optic nerve and RGCs whereas fusion-deficient RSA59 (P) was not.

RSA59 (PP) and RSA59 (P) viruses are able to replicate in the optic nerve at day 3 and day 6 post-inoculation (p.i.)
Four-week-old male C57Bl/6 mice were intracranially inoculated with 50% LD 50 doses of RSA59 (PP) and its mutated recombinant strain RSA59 (P). Optic nerves and whole eyes were harvested, fixed, and sectioned. To detect the presence of viral antigens in the optic nerve, immunohistochemistry was performed with anti-nucleocapsid (anti-N) antibody using an avidin-biotin-peroxidase assay with the substrate 3,3Ј Diaminobenzidine. Images of immunostained optic nerves from mock-infected, RSA59 (PP)-infected, and RSA59 (P)infected mice were taken with a ϫ40 objective. At days 3 and 6 p.i., no viral staining was observed in mock-infected mice ( Fig. 1, a and d). At day 3 p.i., 100% of mice infected with RSA59 (PP) were positive for viral antigen staining in the optic nerve, whereas ϳ75% of RSA59 (P)-infected mice were positive for the same (Fig. 1, b and c). No significant difference was found in the area of viral antigen staining between both strains (Fig. 1g) at day 3 p.i. At day 6 p.i., 100% infection was found in both infected strains (Fig. 1, e and f); however, the area of viral antigen staining was higher in RSA59 (PP)compared with RSA59 (P)-infected mice (Fig. 1h). At day 6 p.i., viral antigen in RSA59 (PP)-infected mice was spread throughout the optic nerve, whereas in RSA59 (P)-infected mice, the viral antigen spread was restricted to individual cells and meninges.

Distribution of microglia/macrophages in RSA59 (PP)-and RSA59 (P)-infected optic nerves
To determine whether the two strains could induce inflammation in the optic nerve, paraffin sections of the optic nerve were stained with H&E to evaluate the presence of inflammatory cells as a sign of optic neuritis. Both RSA59 (PP) (75% of infected mice) and RSA59 (P) (50% of infected mice) virus strains induced optic neuritis (Fig. 2, b and c), which was not present in mock-infected mice (Fig. 2a). In previous studies, most inflammatory cells in the optic nerve were found to be microglia/macrophages (11). To confirm similar findings here, optic nerve sections were immunostained with anti-Iba-1 antibody. In optic nerves from mice infected with RSA59 (PP) or RSA59 (P) (Fig. 2, e and f), a significant number of Iba1ϩ inflammatory cells (a microglia/macrophage marker) were present compared with controls (Fig. 2d). There was no significant difference in the number and distribution of Iba1ϩ cells between mice infected with RSA59 (PP) and RSA59 (P) (Fig. 2g).

RSA59 (PP), but not RSA59 (P), is able to infect the retina
Because viral antigen was detected in the optic nerve, we further investigated whether both viral strains have the capability to traffic to RGC bodies from the site of inoculation (brain) in a retrograde manner. Whole eyes from mock-, RSA59 (PP)-, and RSA59 (P)-infected mice were isolated, serially sectioned, immunostained with anti-N antibody, and counterstained with hematoxylin. Images were captured at ϫ40 magnification to look for viral antigen-positive staining in the RGC layer. Interestingly, viral antigens were detected in ϳ75% of RSA59 (PP)infected mouse eyes at day 3 p.i. and in all RSA59 (PP)-infected mouse retinas at day 6 p.i. (Fig. 3, b and e), whereas no viral antigen was detected in RSA59 (P)-infected (Fig. 3, c and f) mouse retinas, similar to mock infection (Fig. 3, a and d), at 3 and 6 days p.i. At day 3 p.i., RSA59 (PP) viral antigen was predominantly found in the ganglion cell layer (GCL), whereas at day 6 p.i., it spread through all layers of the retina. A significant difference was found in the area of viral antigen staining between RSA59 (PP) and RSA59 (P) strains at days 3 p.i. and 6 p.i. (Fig. 3, g and h).

Microglia/macrophage numbers increase significantly in the RGC layer at day 6 p.i.
Microglia constitute a prominent part of the residential glial population in the retina. During maturation, they acquire their topographical distribution between the inner plexiform layer (IPL) and outer plexiform layer (OPL) in the retina (15). To examine potential topographical changes after infection, 20-m-thick cryopreserved sections were cut and immunofluorescently labeled with anti-Iba-1. Retinas from mock-infected mice showed resting ramified microglia in the IPL and OPL with only a minimal presence in the GCL (Fig. 4a). In eyes from RSA59 (PP)-infected mice, the population of microglia/macrophages was found to migrate and increase specifically in the GCL (Fig. 4b). However, in RSA59 (P)-infected mouse retina, the microglia/macrophages seen in the GCL were comparatively less than in RSA59 (PP)-infected mice (Fig. 4, c and d). Across all retinal layers, there was a trend toward fewer total numbers of microglia/macrophages in mice infected with RSA59 (P) compared with RSA59 (PP)-infected mice at day 6 p.i., although this difference was not statistically significant ( Fig. 4e).

Fusion peptide modulates retrograde axonal transport RSA59 (PP) induces more optic nerve demyelination and inflammation than RSA59 (P) at the chronic stage of optic neuritis
In previous studies, mice infected with RSA59 showed chronic-stage demyelination in the optic nerve (11). To evaluate demyelination in this study, optic nerve sections were stained with LFB. At the chronic stage (day 30 p.i.), no demyelination was observed in mock-infected mouse optic nerves (Fig. 5a). However, 62.5% of optic nerves showed prominent demyelination (Fig. 5d) when mice were infected with RSA59 (PP), whereas little or no demyelination was observed in RSA59 (P)-infected mouse optic nerves (Fig. 5, g and j). To evaluate inflammation at the chronic stage, optic nerve sections were stained with H&E. RSA59 (PP)-infected mouse optic nerves exhibited inflammation at day 30 p.i. (Fig. 5e), whereas little or no inflammation was found in mice infected with RSA59 (P) (Fig. 5h) or in mock-infected mice (Fig. 5b). We characterized the inflammatory cells by staining optic nerve sections with anti-Iba1. At day 30 p.i., resting ramified microglia were found in mock-infected mouse optic nerves (Fig. 5c). However, phagocytotic microglia/macrophage accumulation was found in numerous RSA59 (PP)-infected (Fig. 5f) mouse optic nerves, but in RSA59 (P)-infected (Fig. 5i) mice, it was only found occasionally. Quantification of this staining showed that there was a significant increase in the total number of microglia/macrophages in RSA59 (PP)-infected compared with mock-infected and RSA59 (P)-infected optic nerves (Fig. 5k). The amount of anti-N staining was quantified as stated under "Materials and methods." Staining was higher in RSA59 (PP)-infected mouse optic nerves compared with mock-infected ones on days 3 (g) and 6 (h) p.i. A significant difference in staining was observed between RSA59 (PP) and RSA59 (P) at day 6 p.i. Level of significance was determined by one-way ANOVA and Tukey's multiple comparisons test. *, p Ͻ 0.05; **, p Ͻ 0.01.

Fusion peptide modulates retrograde axonal transport RSA59 (PP) induces RGC loss whereas RSA59 (P) does not
Demyelinating optic neuritis induced by RSA59 has been shown previously to induce neuronal damage with RGC loss (13). To examine whether the proline-mutated strain RSA59 (P) can cause RGC loss at day 30, retinas from mock-infected, RSA59 (PP)-infected, and RSA59 (P)-infected mice were isolated and labeled with anti-Brn3a (an RGC marker), and RGCs were counted in 12 standardized fields. Mice infected with RSA59 (PP) had a significant decrease in RGC number compared with mock-infected mice (Fig. 6, a and b), whereas infection with RSA59 (P) led to no significant difference in RGC numbers compared with mock-infected mice (Fig. 6, c and d).

Discussion
A previous study demonstrated that deletion of one proline in the fusion peptide of spike in the fusogenic MHV strain RSA59 reduced its ability to cause cell-cell fusion in vitro (8). In addition, the proline-mutated strain RSA59 (P) showed reduced acute-stage dissemination of viruses in the brain and spinal cord and reduced chronic-stage demyelination in the white matter of the spinal cord (8). The results of this study, comparing optic nerve and RGCs, reveal that the proline-mutated strain RSA59 (P) leads to reduced optic nerve inflammation and demyelination, suggesting significant consequences of differential virus replication and distribution compared with RSA59 (PP) infection. It is also evident from this study that RSA59 (PP), which contains the WT fusion peptide, causes severe RGC loss, similar to previous studies of RSA59 (8), whereas RSA59 (P) fails to induce RGC loss. Thus, these results demonstrate a key role of dual prolines in the fusion peptide of spike in mediating pathology in optic neuritis. This is likely due to effects of the fusion peptide on the ability of the virus to be transported along neuronal axons. Indeed, the presence of viral antigen in the eyes of RSA59 (PP)-infected mice suggests that retrograde axonal transport of the virus from the site of inoculation occurred. In contrast, the lack of viral antigen in the retina of RSA59 (P)-infected mice suggests that a single-proline mutant strain of RSA59 was not capable of traveling into the Figure 2. a-g, differential optic nerve inflammation at day 6 (acute stage) p.i. by H&E staining (a-c) and Iba-1 staining (d-f). Optic nerve sections from mock-, RSA59 (PP)-, and RSA59 (P)-infected mice (n ϭ 4 mice/virus) harvested on day 6 p.i. were stained with H&E. RSA59 (PP)-infected (b) and RSA59 -infected (P) (c) optic nerves showed accumulation of inflammatory cells (arrows) compared with mock-infected mice (a). Similarly, sequential sections stained with anti-Iba-1 antibody (a microglia/macrophage marker) showed a few resident Iba-1ϩ cells in optic nerves of mock-infected mice (d), whereas in RSA59 (PP)-infected (e) and RSA59 (P)-infected (f) mice, higher numbers of Iba1ϩ cells were found. Iba1ϩ cells were counted in ϫ40 images. Significantly more Iba1ϩ cells were present after infection with RSA59 (PP) and RSA59 (P) compared with mock infection. No significant difference was observed in inflammation levels between RSA59 (PP)-and RSA59 (P)-infected mice (g). Level of significance was determined by one-way ANOVA and Tukey's multiple comparisons test. *, p Ͻ 0.05.

Fusion peptide modulates retrograde axonal transport
eye. Although viral antigen was found in the optic nerve of RSA59 (P)-infected mice, the absence of viral antigen in the eye suggests a reduced ability of RSA59 (P) to cause cell-cell fusion, further reducing the degree of retrograde axonal transport.
Demyelinating strain RSA59 (PP) successfully induced optic nerve inflammation as well as demyelination (11) and RGC loss (13), in accordance with our previous studies. Optic nerve inflammation was also evident in RSA59 (P)-infected optic nerves in this study, but the degree of demyelination and RGC loss was not significant in RSA59 (P)-infected mice compared with mock-infected mice. In addition, inflammation in the ganglion cell layer of RSA59 (PP)-infected mouse retinas at the acute stage was comparatively more than in RSA59 (P)infected mice. Thus, the results further support a key role of the fusion peptide in virus-induced neuroinflammation and demyelination. This may be a secondary effect because of reduced axonal transport of the virus or due to other direct effects of the mutated fusion peptide, which remain to be elucidated in future studies.
Together, our results suggest that lack of cell-cell fusion in the case of RSA59 (P) could play a major role in RGC loss induced during optic neuritis. To our knowledge, this is the first report showing that a mutation known to reduce the ability of RSA59 (P) to cause cell-cell fusion also reduces its trafficking

Fusion peptide modulates retrograde axonal transport
and infection to the RGC layer. This loss of pathologic function is likely governed by loss of proline, altering the rigidity and stability of the spike protein structure during fusion, as shown in a previous study (8). The fusion property of the virus is important for its pathogenicity. From our previous study, it is evident that deletion of one proline from the fusion peptide may destabilize the spike protein, which may result in impaired fusogenicity (8). Also, this single proline mutation results in reduced replication and spread of RSA59 (P). The interaction of the virus with axonal transport machinery needs to be further investigated. It has been suggested to either hijack cargo proteins to move along microtubules or to be transported inside vesicles or directly interact with microtubules. How the virus enters and exits neurons is also not yet clear, but there is a possibility that the virus may gain access to nerve endings due to traumatic disruption at the site of inoculation. The presence of viral antigen in the RGC layer found here is consistent with our prior study suggesting that the virus travels through retrograde axonal transport (11). Although infection in RGCs alone cannot exclude hematogenous spread as an alternative mechanism for retrograde transport, the distinct timing and amount of viral spread between the optic nerve (seen by day 3 p.i.) and the retina (seen by day 6 p.i.) suggest retrograde axonal transport to be the more likely mechanism of transport.
From the above proline insertion-deletion study, it is evident that proline, a single amino acid, may serve as an impor-tant entity in the fusion peptide of MHV, required for efficient retrograde axonal transport and demyelination. Dissecting the minimal essential motif of the spike protein fusion peptide may enable us to design a mimetic peptide to set the stage for competition to reduce virus-induced neuroinflammation and RGC loss. Overall, our data suggest that the presence of two consecutive prolines plays a major role in preventing spread of the virus to postsynaptic targets (RGCs).

Mice
Four-week-old MHV-free C57BL/6J male mice (In Vivo Biosciences) were used for the experiment. All procedures were carried out in accordance with ethical guidelines approved by the Institutional Animal Care and Use Committee at the Indian Institute of Science Education and Research Kolkata. The animal protocols adhered to the guidelines of the Committee for the Purpose of Control And Supervision of Experiments on Animals (CPCSEA), India.

Viruses
A proline-mutated recombinant virus RSA59 (P) engineered from RSA59 (PP), both expressing enhanced GFP, was used as described in a previous study (8). Mice were monitored up to day 30 p.i. The few Iba1ϩ cells in mock-infected eyes were scattered across retinal layers, including the IPL and GCL (arrows), and in RSA59 (P)-infected mice (c), similarly few Iba-1ϩ cells were found in the GCL (arrows). A higher number of Iba-1ϩ cells was found in the GCL of RSA59 (PP)-infected (b) retina (arrows). d, total Iba-1ϩ cells present in the GCL of the retina per section were counted. The number of Iba1ϩ cells in the GCL was higher in RSA59 (PP)-than in mock-infected mice. In RSA59 (P)-infected mice, there was no significant increase in Iba-1ϩ cells. e, similarly, total Iba1ϩ cells present in all layers of the retina were counted. Despite the appearance of more Iba1ϩ cells present after infection with RSA59 (PP) compared with mock infection in the representative sections, this trend was not significant, and no significant difference was observed between RSA59 (PP)-and RSA59 (P)-infected mice. Level of significance was determined by one-way ANOVA and Tukey's multiple comparisons test. **, p Ͻ 0.01; ***, p Ͻ 0.001; ns, not significant (n ϭ 3-6 mice/group). INL, inner nuclear layer, ONL, outer nuclear layer. The percentage area of myelin (LFB stain) in the optic nerve was calculated by densitometry and showed greater myelin loss in RSA59 (PP)-infected mice compared with mock-infected optic nerves. RSA59 (P)-infected mice had no significant myelin loss compared with mock-infected controls (j). k, total Iba1ϩ cells were counted and showed a significant difference between mock-infected control and RSA59 (PP) as well as RSA59 (PP) versus RSA59 (P). Level of significance was determined by one-way ANOVA and Tukey's multiple comparisons test. *, p Ͻ 0.05 (n ϭ 3-6 mice/group).

Fusion peptide modulates retrograde axonal transport
Inoculation of mice 20,000 pfu (50% LD 50 ) of RSA59 (PP) and RSA59 (P) was used to inoculate 4-week-old MHV-free C57BL/6J (2,8) mice. The desired viruses were diluted from stock solution using 0.75% PBS/BSA, and a final volume of 20 l of diluted virus was inoculated intracranially. The virus was injected directly into the brain by syringe needle through the skull posterior to the orbits near the lateral geniculate nucleus (LGN). Upon intracranial inoculation, the virus infected neurons in regions containing RGC axonal projections, including the LGN in the thalamus (Fig. 7) (11). Animals were euthanized using isoflurane and perfused with 1ϫ PBS followed by 4% paraformaldehyde (PFA), and tissues were harvested at days 3 and 6 p.i. for acute-stage analysis and at day 30 p.i. for chronic-stage study. 4 -6 mice were used per infection type per day post-infection, as indicated in the figure legends. For mock infection, 3 mice per day postinfection were used and inoculated with 0.75% PBS/BSA.

Histopathology
At days 3, 6, and 30 p.i., optic nerves and eyes were isolated from mock-infected as well as MHV-infected mice. For paraffin sectioning, optic nerves and eyes were post-fixed with 4% PFA for 15 min, processed, and embedded with paraffin. For cryosectioning, eyes were post-fixed with 4% PFA for 15 min, processed in 10% and 30% sucrose solution, embedded in OCT medium (Tissue Tek, Hatfield, PA), sectioned sagittally with the help of a cryotome (Thermo Scientific) to 20-m thickness, and mounted on charged glass slides. Right eyes were collected for immunohistochemistry on paraffin sections with viral nucleocapsid antigen, whereas left eyes were collected for immunofluorescence on cryosections with Iba1. 5-m-thick optic nerve sections were cut and stained with H&E, and additional sections of day 30 p.i. optic nerves were stained with LFB to detect myelin damage.

Immunofluorescence of frozen sections
Frozen tissue sections were washed with PBS at room temperature to remove cryomatrix. Tissues then were incubated for 1 h at room temperature with 1 M glycine in PBS to reduce nonspecific cross-linking, followed by 10-min incubation at room temperature with 1 mg/ml NaBH 4 in PBS to reduce autofluorescence. Slides were washed with PBS and incubated with blocking serum containing PBS with 0.5% Triton X-100 and 2.5% goat serum. The sections were incubated overnight at 4°C with a primary anti-Iba1 antibody diluted in blocking serum, washed, and subsequently incubated with Alexa Fluor 568conjugated goat anti-rabbit IgG secondary antibody (Thermo Figure 6. a-c, comparative RGC loss in RSA59 (PP)-and RSA59 (P)-infected retina during the chronic stage of infection. Significant RGC loss was found in RSA59 (PP)-infected (b) eyes at day 30 p.i. compared with mock-infected mice (a). Mice inoculated with RSA59 (P) (c) did not show any significant RGC loss compared with mock-infected control mice (a). RGCs were counted from 12 standardized retinal fields. d, the average number of RGCs surviving per eye shows that RSA59 (PP) induced a significant decrease in RGC number compared with mock-infected mice, whereas RSA59 (P) did not cause any significant decrease in RGC number. Level of significance was determined by one-way ANOVA and Tukey's multiple comparisons test. **, p Ͻ 0.01 (n ϭ 5 mice/group).

Fusion peptide modulates retrograde axonal transport
Fisher Scientific) diluted in PBS with goat serum for 2 h at room temperature. All incubations were carried out in a humidified chamber. After washing with PBS, sections were mounted with 4Ј,6-diamidino-2-phenylindole containing mounting medium (Vectashield) and imaged using a Nikon eclipse Ti2 microscope. The images were processed with Fiji (ImageJ 1.52g) software.

Quantification of histological slides
Anti-viral antigen staining was quantified as described previously using Fiji (ImageJ 1.52g) software (8,16). Briefly, five fields of each optic nerve tissue section were captured at ϫ40 magnification and analyzed. Four to five optic nerve sections per mouse per infection were used to measure the area of staining. Similarly, four to five sections of retina, each from medial to peripheral per mouse per infection, were scanned at the highest magnification of ϫ40. The RGB images were color-deconvoluted into three different colors to separate 3,3Ј Diaminobenzidine-specific staining. To ensure that all labeled cells were selected, a threshold value was defined for each image. The magnitude of staining was defined as the area of staining and plotted. Each dot in a graph represents the average value of the captured frames from each mouse. To ensure error-free data collection, the entire quantification procedure was performed by two investigators and read in a blinded manner.
The magnitude of Iba1ϩ microglia/macrophage activation was defined as the number of Iba1ϩ cells present in the total area. Four to five optic nerve tissue sections per mouse per infection were randomly selected to count the total number of Iba1ϩ cells. Similarly, four to five slices of immunofluorescently labeled retina, from medial to peripheral per mouse per infection, were selected to count the total number of Iba1ϩ cells and number of Iba1ϩ cells present in the GCL layer.
To determine the total white matter area and areas with myelin loss in day 30 RSA59 (PP) and RSA59 (P) p.i. mice, four to five LFB-stained optic nerve sections from each mouse were randomly selected and analyzed using Fiji software (ImageJ 1.52g) (8,17). The total area of the demyelinating plaque was outlined and measured. The percentage of demyelination per section per mouse was obtained by dividing the total area of the demyelinating plaque over the total area of the optic nerve section and then multiplied by 100.

Quantification of RGC numbers
RGC immunolabeling and quantification were performed as described in our previous study (13). Briefly, both eyes at day 30 p.i. were removed and fixed with 4% PFA. Retinas were then isolated and washed three times with 1ϫ PBS and permeabilized with 0.5% Triton X-100 at Ϫ70°C. Retinas were incubated in blocking buffer (PBS containing 2% goat serum and 2% Triton X-100) for 4 h at room temperature. Then they were incubated overnight with anti-Brn3a antibody (Synaptic Systems, catalog no. 411003) diluted 1:500 in blocking buffer at 4°C. After washing, retinas were incubated for 2 h at room temperature with Texas Red-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch Laboratories, West Grove, PA) diluted 1:200. Photographs of RGCs were taken in 12 standardized fields at 1/6, 3/6, and 5/6 of the retinal radius from the center of the retina in four quadrants at ϫ40 magnification. RGCs were counted in each field by a blinded investigator using Image-Pro Plus 5.0 (Media Cybernetics, Silver Spring, MD) software.

Statistical analysis
The level of significance for immunohistochemistry staining and LFB staining in spinal cord sections was calculated using one-way ANOVA and Tukey's multiple comparisons test. All data were scatterplotted with standard deviation and analyzed using GraphPad Prism 6.01 software. Each dot in a graph represents the average value acquired from each mice. p Ͻ 0.05 was considered statistically significant.

Data availability
Details of all experimental procedures are available under "Material and methods," and all experimental data, including a schematic presentation of the inoculation site, are available under "Results."