Antiallodynic effects of alpha lipoic acid in an optimized RR-EAE mouse model of MS-neuropathic pain are accompanied by attenuation of upregulated BDNF-TrkB-ERK signaling in the dorsal horn of the spinal cord

Neuropathic pain may affect patients with multiple sclerosis (MS) even in early disease. In an experimental autoimmune encephalomyelitis (EAE)-mouse model of MS, chronic alpha lipoic acid (ALA) treatment reduced clinical disease severity, but MS-neuropathic pain was not assessed. Hence, we investigated the pain-relieving efficacy and mode of action of ALA using our optimized relapsing-remitting (RR)-EAE mouse model of MS-associated neuropathic pain. C57BL/6 mice were immunized with MOG35-55 and adjuvants (Quil A and pertussis toxin) to induce RR-EAE; sham-mice received adjuvants only. RR-EAE mice received subcutaneous ALA (3 or 10 mg kg−1 day−1) or vehicle for 21 days (15–35 d.p.i.; [days postimmunization]); sham-mice received vehicle. Hindpaw hypersensitivity was assessed blinded using von Frey filaments. Following euthanasia (day 35 d.p.i.), lumbar spinal cords were removed for immunohistochemical and molecular biological assessments. Fully developed mechanical allodynia in the bilateral hindpaws of vehicle-treated RR-EAE mice was accompanied by marked CD3+ T-cell infiltration, microglia activation, and increased brain-derived neurotrophic factor (BDNF)-tyrosine kinase B (TrkB) signaling in the dorsal horn of the lumbar spinal cord. Consequently, phospho-ERK, a marker of central sensitization in neuropathic pain, was upregulated in the spinal dorsal horn. Importantly, hindpaw hypersensitivity was completely attenuated in RR-EAE mice administered ALA at 10 mg kg−1 day−1 but not 3 mg kg−1 day−1. The antiallodynic effect of ALA (10 mg kg−1 day−1) was associated with a marked reduction in the aforementioned spinal dorsal horn markers to match their respective levels in the vehicle-treated sham-mice. Our findings suggest that ALA at 10 mg kg−1 day−1 produced its antiallodynic effects in RR-EAE mice by reducing augmented CD3+ T-cell infiltration and BDNF-TrkB-ERK signaling in the spinal dorsal horn.


Introduction
Multiple sclerosis (MS) is a chronic inflammatory-demyelinating disease of the central nervous system (CNS). Apart from motor impairment, MS is often associated with sensory changes (Compston and Coles 2008). Central neuropathic pain (CNP), that may affect patients with MS even early in the disease course, is mostly persistent in nature and unaffected by the clinical disease course (Kalia and O'Connor 2005;Osterberg et al. 2005). Similarly, the time course for development of CNP behaviors in experimental autoimmune encephalomyelitis (EAE)rodent models is discordant with that for development of EAE-induced motor impairments . EAE-induced motor deficits are predominantly correlated with pathobiologic changes in the ventral horn of the spinal cord and/or dorsal funicular white matter (Wu et al. 2008). By contrast, neuropathic pain behaviors in rodent models of CNP are associated with pathobiologic mechanisms in sensory neurons in the dorsal horn (laminae I-II) of the spinal cord Watson et al. 2014). Recommended drug treatments for the relief of peripheral neuropathic pain conditions (Dworkin et al. 2010) are often ineffective for alleviating MS-associated CNP . Hence, new and well-tolerated analgesic treatments for improved relief of this condition, are an unmet medical need.
Alpha lipoic acid (ALA), a nutraceutical compound known for antioxidant properties has been shown recently to produce significant alleviation of peripheral neuropathic pain in rodent models of painful diabetic neuropathy (Morgado et al. 2011) and cancer chemotherapy-induced peripheral neuropathy (CIPN), when administered by chronic and acute dosing regimens, respectively (Trevisan et al. 2013). In other work using EAE rodent models, chronically administered ALA attenuated clinical disease severity (Morini et al. 2004;Wang et al. 2013), but its effect on MS-associated CNP was not investigated.
Hence, the present work was designed to assess the potential pain-relieving efficacy and mode of action of ALA administered chronically according to an intervention protocol, using an optimized relapsing-remitting experimental autoimmune encephalomyelitis (RR-EAE) mouse model of MS-induced neuropathic pain developed by our group . Notably, our findings are novel showing that once-daily subcutaneous (s.c.) administration of ALA at 3 or 10 mg kg À1 to RR-EAE mice for 21 consecutive days, evoked dose-dependent alleviation of mechanical hypersensitivity in the bilateral hindpaws. The pain-relieving mode of action involved a significant reduction in augmented CD3 + T-cell infiltration and microglial activation in the dorsal horn of the lumbar (L4-L6) spinal cord. This in turn reduced EAEinduced upregulation of expression levels of BDNF, tyrosine kinase B (TrkB), and phosphorylated-p44/p42 mitogen-activated protein kinase (pp44/pp42 MAPK; also called pERK), a marker of BDNF/TrkB downstream signaling, in the lumbar spinal dorsal horn of RR-EAE mice to match the respective levels in vehicle-treated shamcontrol mice. To the best of our knowledge, this is the first study to evaluate the potential antiallodynic effects of ALA in a rodent model of MS-induced neuropathic pain, and we show that pain-relief appears to be mediated by attenuation of upregulated BDNF-TrkB-ERK signaling in the spinal dorsal horn of RR-EAE mice.

Animals
Female C57BL/6 mice aged 4-6 weeks were from the University of Queensland Biological Resources. Mice were housed in groups of 6-8 per cage in a temperature-controlled facility (22-23°C) with a 12/12 h light/dark cycle. Rodent chow and water were available ad libitum. Ethics approval was from the Animal Ethics Committee of the University of Queensland. Experiments complied with the requirements of the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes (NHMRC (National Health andMedical Research Council) 2004, 2013;) and the ARRIVE guidelines (Kilkenny et al. 2010).

EAE induction and clinical disease scoring
Our optimized RR-EAE mouse model of MS-associated neuropathic pain is novel and it is devoid of confounding motor deficits in the hindpaws . Briefly, mice were immunized with 200 lg of MOG 35-55 (Mimotopes, Clayton, Victoria, Australia) mixed with a solution of Quil A (45 lg) in 100 lL of phosphate-buffered saline (PBS) (Sigma-Aldrich, Sydney, New South Wales, Australia). Mice then received an intraperitoneal (i.p.) injection of pertussis toxin at 250 ng (Sigma-Aldrich) in PBS (1 ng lL À1 ) and this was repeated 48 h later . Sham-mice (control group) received adjuvants only (Quil A and pertussis toxin). Clinical disease scoring of RR-EAE and sham-mice was undertaken once-daily in a blinded manner using a 5-point scale with half-point gradations (Table 1) Peiris et al. 2007). RR-EAE clinical disease was classified as present by clinical scores ≥1, whereas clinical scores ≤0.5 were regarded as disease remission or absence. General health and body weights of all mice were assessed prior to immunization 2015 | Vol. 3 | Iss. 3 | e00137 Page 2 and once-daily thereafter in a blinded manner until the completion of study.

Test compound dosing regimen
ALA was from GeroNova Research Inc. (Richmond, CA) and supplied as the water-soluble sodium salt of the Renantiomer of ALA. Dosing solutions of ALA were prepared in sterile water for injection (Pfizer, West Ryde, New South Wales, Australia).
RR-EAE mice received once-daily s.c. injections of ALA at 3 or 10 mg kg À1 day À1 , or vehicle for 21 consecutive days commencing at 15 d.p.i., as the first episode of clinical disease had occurred in all EAE mice by day 15. Sham-control mice received once-daily s.c. vehicle injections for the same treatment period. Solutions of ALA or vehicle were prepared freshly each day and masked before administration to animals in a blinded manner. All drug treatments were administered between 09.00 am and 11.00 am. The total number of sham-and RR-EAE mice (in Cohorts 1 and 2) treated with vehicle or ALA is shown in Figure 1. To maintain blinding, equal numbers of animals were included in each group. Animals were randomized in a blinded manner and received masked solutions of ALA or vehicle by the first person.
The drug and molecular target nomenclature used herein is in accordance with that recommended by Alexander et al. (2013).

Assessment of antiallodynic effects evoked by chronic ALA in RR-EAE mice
Calibrated von Frey filaments (Stoelting Co., Wood Vale, IL) were used according to the up-down method (Chaplan et al. 1994) to assess paw withdrawal thresholds (PWTs) in the bilateral hindpaws of RR-EAE mice rela-tive to sham-mice by a blinded tester. Briefly, baseline PWTs were measured prior to immunization on day 0 and once-weekly thereafter to define the time course for temporal development of mechanical allodynia. For assessment of the antiallodynic effects of ALA (or vehicle) in RR-EAE mice, hindpaw PWTs were measured onceweekly at 0.5-0.75 h post-dosing. This was the time of peak effect observed in pilot experiments and it is in agreement with the time of peak antinociception produced by single bolus doses of ALA in a rat model of CIPN (Trevisan et al. 2013). Hindpaw PWTs for shammice were assessed concurrently with the corresponding cohort of RR-EAE mice by a blinded tester.

Chronic ALA treatment: EAE clinical disease profiling
The effect of ALA on EAE clinical disease progression was assessed by measuring clinical disease scores once-daily at 0.5-0.75 h post-dosing. Clinical disease scores for shammice administered the chronic vehicle dosing regimen were also assessed concurrently with the corresponding cohort of RR-EAE mice in a blinded manner.

Ex-vivo analysis: tissue collection and preparation
RR-EAE mice administered the chronic ALA (10 mg kg À1 day À1 ) or vehicle treatment, as well as shammice (administered the chronic vehicle dosing regimen n = 4/group for each technique), were euthanized with an overdose (~300 mg kg À1 ; i.e. 1 mL/kg of 325 mg/mL, i.p.) of pentobarbitone (Lethabarb â ; Virbac, Milperra, New South Wales, Australia) at 0.5-0.75 h post-dosing on the last treatment day (35 d.p.i.). Lumbar (L4-L6) spinal cord tissues were removed for ex-vivo mode of action analyses using immunohistochemistry (IHC), western blotting, and quantitative real-time PCR (RTqPCR). For IHC, mice were perfused with 4% paraformaldehyde (Sigma-Aldrich) in ice-cold 1 9 PBS (pH 7.4;~100 mL/mouse) prior to removal of lumbar spinal cord tissues. Transverse cryosections (10-12 lm) were prepared and mounted on Superfrost Plus â slides (Thermo Fisher Scientific, Scoresby, Victoria, Australia) prior to immunostaining, as per our previous report . For western blot and RTqPCR, lumbar (L4-L6) spinal cord tissues were collected without perfusion and immediately stored at À80°C prior to further processing as described below.

Quantitative real-time PCR
Absolutely RNA Miniprep Kits (Agilent, Santa Clara, CA) were used to isolate total RNA individually from lumbar spinal cord tissue from each mouse. Complementary DNA (cDNA) was reverse-transcribed from total RNA using high-capacity cDNA reverse transcription kits (Applied Biosystems, Mulgrave, Victoria, Australia). Quantitative BDNF and TrkB messenger RNA (mRNA) expression levels were analyzed using SYBR â Green PCR master mix (Applied Biosystems) and specific primers. The specificity of the real-time PCR reaction was confirmed using melting curve analysis. Primers used were (1) 18S-forward: CCCTCCAATGGATCCTCGTT; 18Sreverse: TCGAGGCCCTGTAATTGGAA (2) BDNF Quan-tiTect â primer assay, Mm_Bdnf_1_SG (Qiagen, Pty Ltd., Victoria, Australia) and (3) TrkB QuantiTect â primer assay, Mm_Ntrk2_vb.1_SG (Qiagen) containing a mixture of forward and reverse primers.

Mouse plasma biochemistry and organ weights
Please see supplementary methods for method descriptions.

Data analyses
Behavioral data Graphical data are presented as mean AE standard error of the mean (AESEM). Clinical scores were the mean (AESEM) score for each group of mice at each time point.
Hindpaw PWT values are the mean (AESEM) of three measurements taken at least 5 min apart for both hindpaws combined.

Immunohistochemistry data
IIHC images of at least 3-4 lumbar spinal dorsal horn sections (>100 lm apart) per animal from each treatment group (n = 3-4/group) were used for quantitative analysis or colocalization studies in vehicle-treated RR-EAE mice as appropriate. Laminae I-II of the spinal dorsal horn were targeted for all IHC analyses (Fig. S2). Densitometric counts were quantified using Axiovision Rel. v4.8 software (Carl Zeiss, G€ ottingen, Germany) in a blinded manner ). Data are expressed as fold-changes in fluorescence intensity for sections from RR-EAE mice treated with ALA (10 mg kg À1 day À1 ) or vehicle, relative to the corresponding data for vehicle-treated sham-mice.

Western blot data
Image Studio Lite v4.0 software (LI-COR) was used to quantify immunoreactive band intensities in western blots for each treatment group (n = 3/group). For each protein of interest, band intensity was normalized relative to GAPDH (Jain et al. 2012) to give relative expression levels and fold-changes between ALA-and vehicle-treated RR-EAE mice compared with the vehicle-treated sham-group.

RTqPCR data
For RTqPCR, mRNA expression levels were normalized relative to 18S, for each treatment group (n = 3-4/ group). For the ALA-and vehicle-treated RR-EAE mice, mRNA levels were quantified relative to that for vehicletreated sham-mice using the DDC T method (Schmittgen and Livak 2008). The fold-changes for each target of interest were determined using the 2 ÀDDCT method (Schmittgen and Livak 2008).

Mouse plasma biochemistry and organ weights
Mouse plasma biochemistry parameters are reported as mean (AESEM). Organ weights (kidneys and liver) are presented as relative to respective body weights on the day of euthanasia (Lee et al. 2014) ( Table 2).

Statistical analyses
Statistical analyses were performed using GraphPad Prism TM v6.04 (GraphPad Software, La Jolla, USA). Repeated measures two-way analysis of variance (ANO-VA) followed by the Bonferroni test was used to analyze pain behavioral end points. The Kruskal-Wallis test with the multiple comparison post hoc Dunn's test, was used to analyze clinical scores for the various groups of mice, consistent with previous work by others (Fleming et al. 2005;Amor and Baker 2012). The one-way ANOVA followed by Tukey's multiple comparison test, was used to compare intergroups differences for IHC, western blot and RTqPCR data, plasma biochemistry data, and organ weights. The statistical significance criterion was P < 0.05.  (Fig. 1).

RR
Chronic ALA treatment alleviates hindpaw hypersensitivity in RR-EAE mice In RR-EAE mice, once-daily s.c. administration of ALA at 3 or 10 mg kg À1 day À1 for 21 consecutive days (15-35 d.p.i.) evoked dose-dependent alleviation of mechanical hypersensitivity in the bilateral hindpaws. Specifically, mean (AESEM) hindpaw PWTs for RR-EAE mice (Cohorts 1 and 2) administered the chronic dosing regimen of ALA at 10 mg kg À1 day À1 , did not differ significantly (P > 0.05) from the corresponding groups of vehicle-treated sham-mice at 35 d.p.i. However, by 3 weeks after cessation of ALA at 10 mg kg À1 day À1 in Cohort-2 RR-EAE mice at 55 d.p.i., mechanical allodynia was again fully developed in the bilateral hindpaws of these mice (Fig. 1).
Chronic ALA treatment arrests clinical disease relapse in RR-EAE mice RR-EAE mice (Cohorts 1 and 2) treated with ALA at 3 or 10 mg kg À1 day À1 for 21 consecutive days commencing on day 15 d.p.i did not exhibit clinical disease relapses. Specifically, the clinical disease patterns in Cohorts 1 and 2 of RR-EAE mice treated with ALA at 10 mg kg À1 day À1 were comparable (P > 0.05) to those from the corresponding vehicle-treated sham-mice at the end of treatment (35 d.p.i.) (Fig. 1). However, by 9 days after cessation of ALA treatment at 10 mg kg À1 day À1 in Cohort 2 RR-EAE mice (44 d.p.i.), there was a recurrence of relapsing-remitting clinical disease that persisted until study completion at 55 d.p.i. (Fig. 1).

CD3 + T-cell infiltration
As the role of reactive T-cells in MS-associated neuropathic pain is unknown (Khan and Smith 2014), we examined CD3 + T-cell IF in the lumbar spinal dorsal horn of RR-EAE mice. Specifically, CD3 + T-cell IF was 4.7-fold higher (F (3,44) = 26.44; P < 0.05) in vehicletreated RR-EAE mice exhibiting neuropathic pain behavior at 35 d.p.i. c.f. the respective levels in vehicle-treated sham-mice. For RR-EAE mice administered ALA at 10 mg kg À1 day À1 for 21-days, CD3 + T-cell IF in the lumbar spinal dorsal horn did not differ significantly (P > 0.05) from that for vehicle-treated sham-mice. However, by 3 weeks after cessation of ALA treatment at 10 mg kg À1 day À1 in Cohort-2 RR-EAE mice at 55 d.p.i., CD3 + T-cell IF in the spinal dorsal horn was again increased to~2.5-fold higher (F (3,44) = 26.44; P < 0.05) than in vehicle-treated sham-mice (Fig. 2).

BDNF and TrkB expression
BDNF-TrkB signaling in the spinal cord has been well established in peripheral neuropathic pain conditions (Vanelderen et al. 2010). Hence, we investigated expression levels of BDNF and TrkB in vehicle-treated RR-EAE mice exhibiting robust mechanical allodynia at 35 d.p.i. The mean IF levels for BDNF and TrkB in lumbar spinal dorsal horn of vehicle-treated RR-EAE mice were~2.5-(F (2,33) = 15.84; P < 0.05) and~3.2-(F (2,33) = 71.52; P < 0.05) fold higher c.f. their respective levels in vehicletreated sham-mice. By contrast, for RR-EAE mice treated chronically with ALA at 10 mg kg À1 day À1 , mean levels of BDNF and TrkB IF in lumbar spinal dorsal horn did not differ significantly (P > 0.05) from the respective mean levels of BDNF and TrkB IF for vehicle-treated sham-mice (Fig. 3).

Immunohistology: colocalization of BDNF, TrkB, and pERK with CD3 + T-cells
Using specific IF antibodies, we show herein that BDNF, TrkB, and pERK were highly colocalized with CD3 + T cells in the dorsal horn of lumbar spinal cord sections from RR-EAE mice . These markers were also colocalized with subsets of neurons (NeuN) and microglia/macrophages (CD11b or Iba1) (Figs. 6-8). BDNF and TrkB were minimally colocalized with astrocytes (GFAP), whereas pERK was colocalized with a subset of astrocytes (Figs. 6-8). BDNF was highly colocalized with its receptor, TrkB, in the dorsal horn of lumbar spinal cord sections from these animals (Figs. 6-8).

Discussion
Here, we show novel findings that once-daily chronic treatment with ALA (3 or 10 mg kg À1 day À1 ) for 3 weeks attenuated the development of bilateral hindpaw hypersensitivity in an optimized RR-EAE mouse model of MSinduced neuropathic pain in a dose-dependent manner. Specifically, at 35 d.p.i., mean (AESEM) PWTs for RR-EAE mice administered chronic ALA at 10 mg kg À1 day À1 did not differ significantly (P > 0.05) from the corresponding PWTs for vehicle-treated sham-mice. By contrast, for RR-EAE mice administered vehicle for 3 weeks, mechanical allodynia was fully developed in the bilateral hindpaws at 26 d.p.i. (P < 0.05), that persisted until study completion (Fig. 1).
Additionally, we showed that the efficacious chronic ALA dosing regimen at 10 mg kg À1 day À1 in RR-EAE mice was well tolerated. There were no gross abnormalities at necropsy and biochemical indices of hepatic and renal function were within the normal ranges (Table 2). Interestingly, 3 weeks after cessation of the chronic ALA dosing regimen at 10 mg kg À1 day À1 in Cohort 2 RR-EAE mice, mechanical allodynia recurred in the bilateral hindpaws at 55 d.p.i. in a manner similar to that exhibited by vehicle-treated RR-EAE mice from 26 d.p.i. onward. These findings clearly suggest that ongoing chronic ALA treatment is required to prevent development of persistent MS-associated neuropathic pain (Fig. 1).
2015 | Vol. 3 | Iss. 3 | e00137 Page 10 ALA (10 mg kg À1 day À1 ), show that CD3 + T-cell infiltration into the lumbar spinal dorsal horn was markedly reduced relative to that for vehicle-treated RR-EAE mice exhibiting mechanical allodynia in their bilateral hindpaws (Fig. 2). The augmented CD3 + T-cell infiltration into the lumbar spinal dorsal horn of vehicle-treated RR-EAE mice was accompanied by microglial activation, mirroring previous reports of activated microglia in the CNS of EAE mice exhibiting clinical disease (Murphy et al. 2010). Importantly, relief of hindpaw hypersensitivity in RR-EAE mice by the efficacious chronic ALA dosing regimen (10 mg kg À1 day À1 ) was accompanied by marked attenuation of microglial activation in the spinal dorsal horn (Figs. 3, 4). Our findings agree with previous work showing that ALA inhibits T-cell migration into the spinal cord of EAE mice (Morini et al. 2004;Wang et al. 2013).
Increased expression of BDNF and its high-affinity receptor, TrkB, has been widely implicated in the pathobiology of peripheral neuropathic pain conditions leading to the development of central sensitization in the dorsal horn of the spinal cord (Geng et al. 2010). Indeed, this is regarded as a pathobiologic hallmark of peripheral neuropathic pain, characterized by increased responsiveness of nociceptive neurons in the CNS to normal or subthreshold afferent input (Merskey and Bogduk 1994). However, the Figure 5. Expression levels of mRNA for BDNF and TrkB in lumbar (L4-L6) spinal cord of RR-EAE mice administered either ALA at 10 mg kg À1 day À1 or vehicle for 3 weeks did not differ significantly (F (2,8) = 3.8; BDNF and F (2,7) = 3.59; P > 0.05) from the respective data for the lumbar spinal cord of vehicle-treated sham-mice (oneway ANOVA followed by Tukey's multiple comparison test). EAE, experimental autoimmune encephalomyelitis, Veh, vehicle; ALA, alpha lipoic acid; BDNF, brain-derived neurotrophic factor; TrkB, tyrosine kinase B; ANOVA, analysis of variance. possibility that upregulated BDNF-TrkB-ERK signaling in the spinal dorsal horn may contribute to the pathobiology of MS-associated CNP, has not hitherto been investigated.
Herein, we showed that for vehicle-treated RR-EAE mice with fully developed mechanical allodynia in the bilateral hindpaws, lumbar spinal dorsal horn expression levels of BDNF and TrkB were markedly increased compared with the corresponding expression levels in vehicletreated sham-mice (Figs. 3, 4). However, lumbar spinal cord expression levels of mRNA for BDNF and TrkB in our RR-EAE-mouse model of MS-neuropathic pain did not differ significantly (P > 0.05) from the respective levels for vehicle-treated sham-mice (Fig. 5). Our findings are aligned with reports by others of unchanged mRNA levels for BDNF and TrkB in the CNS of EAE mice exhibiting clinical disease (Zhu et al. 2012(Zhu et al. , 2014. Elevated expression levels of BDNF and TrkB in the lumbar spinal cord of RR-EAE mice may be due to increased translation of BDNF and TrkB from mRNA in spinal cord neurons and/or glia. Alternatively, elevated spinal cord expression levels of BDNF and TrkB may be via a mechanism not involving enhanced local synthesis (Tonra et al. 1998). This latter notion is supported by reports of upregulated BDNF expression by infiltrating Tcells due to antigen stimulation and/or secondary to demyelination of CNS neurons (Stadelmann et al. 2002;Gielen et al. 2003). Although infiltrating T-cell-derived BDNF was originally proposed to be neuroprotective in the CNS of EAE mice (Moalem et al. 2000;Stadelmann et al. 2002), more recent work suggests that T-cell-derived BDNF is not critical for neuroprotective effects in CNS of EAE mice (Lee et al. 2012;Xin et al. 2012). Our present findings of differential changes in expression levels of BDNF isoforms in the lumbar spinal cord of vehicle-treated RR-EAE mice c.f. vehicle-treated sham-mice, are novel (Fig. 4). Specifically, there was significant upregulation of pro-BDNF and truncated-BDNF expression in lumbar spinal cord of vehicle-treated RR-EAE mice exhibiting neuropathic pain behavior. However, although there was a trend for reduced mature-BDNF expression levels in vehicle-treated RR-EAE mice, it was not statistically significant (P > 0.05) relative to the respective levels for vehicletreated sham-mice. Notably, the dysregulated expression of BDNF as well as TrkB isoforms was normalized in RR-EAE mice administered chronic ALA treatment at 10 mg kg À1 day À1 for 3-weeks (Fig. 4).
Our finding of upregulated total BDNF (Figs. 3, S1) in the lumbar spinal dorsal horn of vehicle-treated RR-EAE mice is consistent with previous reports of upregulated levels of BDNF in EAE mice as well as CSF and/or plasma levels from patients with RR-MS (Sarchielli et al. 2002;Frota et al. 2009). Our novel findings herein of differential changes in BDNF isoforms in the lumbar spinal cord of RR-EAE mice are aligned with a recent report of dysregulated serum levels of BDNF isoforms in patients with RR-MS (Tongiorgi et al. 2012). Hence, future investigation on the extent to which differential changes in serum levels of the various BDNF isoforms is a marker for MSassociated neuropathic pain in humans, is warranted.
A biological role for truncated-BDNF is currently unclear, whereas pro-and mature-BDNF are considered 'Yin and Yang' molecules due to their opposing biological roles (Chao and Bothwell 2002). Mature-BDNF signaling via TrkB has a role in cell survival and long-term potentiation in the CNS, whereas pro-BDNF is associated with apoptosis and long-term depression mediated by p75 NTR activation in the CNS (Chao and Bothwell 2002;Stadelmann et al. 2002). A pronociceptive role for pro-BDNF signaling via its main receptor, p75 NTR , is suggested by the presence of both pro-BDNF and p75 NTR in the superficial layers of the spinal dorsal horn together with its anterograde transport to nerve terminals (King et al. 2000;) and co-release with mature-BDNF from activated microglia (Srinivasan et al. 2004).
Our data showing concurrent increases in pERK and pro-BDNF expression levels in vehicle-treated RR-EAE mice (Figs. 3,4) are aligned with previous work by others showing that concentration-dependent activation of TrkB by pro-BDNF, albeit with lower affinity than mature-BDNF, elicits prototypical TrkB downstream signaling including ERK activation (Fayard et al. 2005;Boutilier et al. 2008;Sakuragi et al. 2013). As pERK is regarded as a marker of central sensitization in peripheral neuropathic pain conditions (Zhuang et al. 2005;Gao and Ji 2009), it may also have a pathobiologic role in CNP conditions such as MS-associated neuropathic pain. In vehicle-treated RR-EAE mice herein, upregulated expression levels of BDNF, TrkB, and pERK were colocalized primarily with infiltrated CD3 + T cells in the lumbar spinal dorsal horn. These markers were also colocalized with subsets of neurons and to a lesser extent with microglia/macrophages. However, there was minimal colocalization with astrocytes for BDNF and TrkB (Figs. 6, 8). Hence, it is plausible that augmented levels of pro-BDNF signaling, at least in part, via full-length TrkB to increase pERK expression levels in the spinal dorsal horn of vehicle-treated RR-EAE mice, may contribute to the development of hindpaw hypersensitivity in these mice.

Conclusions
Our findings demonstrate inhibition of the development of mechanical allodynia in the bilateral hindpaws of RR-EAE mice by once-daily treatment with ALA at 10 mg kg À1 day À1 for 3 weeks. The cellular and molecular mechanisms underpinning the antiallodynic effects of ALA appear to involve a marked reduction in upregulated expression levels of BDNF-, TrkB-, and pERK into the dorsal horn of the lumbar spinal cord of RR-EAE mice, to match the respective levels for vehicle-treated sham-mice.