A myelin basic protein fragment induces sexually dimorphic transcriptome signatures of neuropathic pain in mice

In the peripheral nerve, mechanosensitive axons are insulated by myelin, a multilamellar membrane formed by Schwann cells. Here, we offer first evidence that a myelin degradation product induces mechanical hypersensitivity and global transcriptomics changes in a sex-specific manner. Focusing on downstream signaling events of the functionally active 84-104 myelin basic protein (MBP(84-104)) fragment released after nerve injury, we demonstrate that exposing the sciatic nerve to MBP(84-104) via endoneurial injection produces robust mechanical hypersensitivity in female, but not in male, mice. RNA-seq and systems biology analysis revealed a striking sexual dimorphism in molecular signatures of the dorsal root ganglia (DRG) and spinal cord response, not observed at the nerve injection site. Mecha-nistically, intra-sciatic MBP(84-104) induced phospholipase C (PLC)-driven (females) and phosphoinositide 3-kinase-driven (males) phospholipid metabolism (tier 1). PLC/inositol trisphosphate receptor (IP3R) and estrogen receptor co-regulation in spinal cord yielded Ca 2 1 -dependent nociceptive signaling induction in females that was suppressed in males (tier 2). IP3R inactivation by intrathecal xestospongin C attenuated the female-specific hypersensitivity induced by MBP(84-104). According to sustained sensitization in tiers 1 and 2, T cell – related signaling spreads to the DRG and spinal cord in females, but remains localized to the sciatic nerve in males (tier 3). These results are consistent with our previous finding that MBP(84-104) – induced pain is T cell – dependent. In summary, an autoantigenic peptide endogenously released in nerve injury triggers multisite, sex-spe-cific transcriptome changes, leading to neuropathic pain only in female mice. MBP(84-104) acts through sustained co-activation of metabolic, estrogen receptor – mediated nociceptive, and autoimmune signaling programs.

In the peripheral nerve, mechanosensitive axons are insulated by myelin, a multilamellar membrane formed by Schwann cells. Here, we offer first evidence that a myelin degradation product induces mechanical hypersensitivity and global transcriptomics changes in a sex-specific manner. Focusing on downstream signaling events of the functionally active 84-104 myelin basic protein (MBP(84-104)) fragment released after nerve injury, we demonstrate that exposing the sciatic nerve to MBP(84-104) via endoneurial injection produces robust mechanical hypersensitivity in female, but not in male, mice. RNA-seq and systems biology analysis revealed a striking sexual dimorphism in molecular signatures of the dorsal root ganglia (DRG) and spinal cord response, not observed at the nerve injection site. Mechanistically, intra-sciatic MBP(84-104) induced phospholipase C (PLC)-driven (females) and phosphoinositide 3-kinase-driven (males) phospholipid metabolism (tier 1). PLC/inositol trisphosphate receptor (IP3R) and estrogen receptor co-regulation in spinal cord yielded Ca 21 -dependent nociceptive signaling induction in females that was suppressed in males (tier 2). IP3R inactivation by intrathecal xestospongin C attenuated the femalespecific hypersensitivity induced by MBP(84-104). According to sustained sensitization in tiers 1 and 2, T cell-related signaling spreads to the DRG and spinal cord in females, but remains localized to the sciatic nerve in males (tier 3). These results are consistent with our previous finding that MBP(84-104)-induced pain is T cell-dependent. In summary, an autoantigenic peptide endogenously released in nerve injury triggers multisite, sex-specific transcriptome changes, leading to neuropathic pain only in female mice. MBP(84-104) acts through sustained co-activation of metabolic, estrogen receptor-mediated nociceptive, and autoimmune signaling programs.
Seminal research implicated T cell activity in selective mediation of mechano-allodynia in female, but not male, mice with peripheral nerve trauma, as males employed innate immune spinal microglia activity (11,19). The mechanisms and the specific T cell epitopes released that underlie the female-specific T cell activity in mechano-allodynia remain obscure.
Mechanosensory dorsal root ganglia (DRG) neurons of the leg (sciatic nerve) are among the longest cells in the body, spanning from foot to spine. Their lipid-rich, multilamellar myelin membrane sheath provides electrical and metabolic insulation and ensures fast, saltatory conduction of a tactile-stimulated impulse (20). In healthy nerves, myelin compaction and integrity are maintained via interactions of the intrinsically unstructured, cationic myelin basic protein (MBP) with the anionic membrane phospholipids (21,22), including phosphatidylinositol 4,5-bisphosphate (PIP2) (23). Disruption of structural and molecular integrity of myelin by peripheral nerve trauma contributes to mechano-allodynia (24)(25)(26).
MBP is a putative autoantigen implicated in the pathogenesis of autoimmune demyelinating states of the central (21,22) and peripheral (27) nervous systems. In female rats with peripheral nerve trauma, we have shown that inhibition of matrix metalloproteinase-mediated proteolysis of MBP attenuates mechanoallodynia at least in part by blocking the release of T cell epitopes of MBP and T cell homing into the nerve (26,28,29). Specifically, focal nerve trauma causes release of the immunodominant T cell epitope of MBP comprising the evolutionarily conserved, invariant a-helix 87 VVHFF 91 motif (26,(28)(29)(30)(31). To elucidate the downstream events of the MBP peptide exposure in mechano-allodynia, we identified the 84-104 fragment of MBP (MBP(84-104)) as the potent pro-algesic peptide in female rats (29).
In female rats, endoneurial delivery of the pure MBP(84-104), but not the control scramble, peptide by an adjuvant-free injection into an intact sciatic nerve (IS) induces persistent (2 weeks) mechano-allodynia (29,30,32). The MBP(84-104)induced mechano-allodynia in females is refractory to systemically delivered analgesics, such as ketorolac and lidocaine (32). The IS MBP(84-104) mechano-allodynia is T cell-dependent, as shown using athymic nude female rats lacking mature T cells (29) and by mutagenesis of the T cell-binding site of the peptide (30). In contrast to a focal nerve trauma, IS MBP(84-104) is not accompanied by major neuropathological changes (29,30), making it an attractive model to study sustained mechano-allodynia, without background effects of acute degenerative and reparative processes caused by nerve trauma. Based on these data obtained in female rats, we have proposed that the autoimmune pathogenesis of mechano-allodynia arises due to MBP(84-104)/T cell-mediated targeting of myelinated afferents (28)(29)(30)(31)(32)(33)(34)(35).
To determine the role of MBP(84-104) as an epitope initiating the sexual dimorphism (female specificity) of T cell action in mechano-allodynia in mice (11), we here comparatively assessed mechanical sensitivity after IS MBP(84-104) in female and male mice. Our comprehensive, comparative, genomewide transcriptomic analysis revealed molecular signatures in the injected nerve and ipsilateral DRG and spinal cord, with biological processes and signaling cascades under MBP(84-104) control. The present data provide the first evidence that nerve MBP(84-104) exposure is sufficient to induce mechano-allodynia in a sexually dimorphic (female-specific) manner. We here reveal an overarching 3-tier signaling program associated with sex-specific pain mediation and establish a comprehensive, open-access database for future data mining of the mechanisms leading to autoimmune mediation of neuropathic pain, specifically in females.

IS MBP(84-104) triggers female-specific allodynia
Release of MBP(84-104) drives T cell-dependent mechanoallodynia after focal nerve injury in female rodents (29,30,32). Whether a comparable process exists in male rodents is not known. The recent evidence of female-specific T cell mediation of mechano-allodynia signaling in C57BL6 mice (11,19) prompted us to test the algesic effect of MBP(84-104) in female versus male C57BL6 mice. Withdrawal thresholds to mechanical stimulation by von Frey filaments were measured in mice that received a single adjuvant-free IS bolus injection of pure MBP(84-104) or scramble (MBPS) peptides, both diluted in PBS (30 mg in 3 ml, each, Fig. 1a). PBS vehicle injections (3 ml) served as control.
In agreement with our prior reports in female rats (29,30,32), female mice displayed a highly significant reduction in the mechanical force required to evoke hind paw withdrawal (i.e. mechano-allodynia) after IS MBP(84-104), but not PBS or MBPS. The allodynia persisted over the 7-day testing period (Fig. 1b). In contrast, male mice after IS MBP(84-104) exhibited mechanical thresholds comparable with control MBPS and PBS injection over the 7-day assessment (Fig. 1c). No delayed mechanical sensitivity was observed in males over additional testing out to 24 days (data not shown). No hypersensitivity was observed in contralateral paws in either sex (data not shown). We conclude that MBP(84-104) mediates mechano-allodynia selectively in female, but not male, mice.
Based on our extensive behavioral, histological, and molecular, including transcriptomic (GEO ID GSE34868), findings in naive, post-IS PBS, MBPS, MBP(84-104), and other MBP peptide, both ipsilateral and contralateral to injection (28-30, 32, 37), as well as sham-operated and post-nerve injury rodents (37,42), IS PBS was selected as control for  assessing sexually dimorphic changes induced by IS MBP (84-104) (algesic) relative to MBPS (not algesic), each solubilized in PBS. We note that physical nerve injury due to injection, represented in the IS PBS group of both female and male tissues, is not a source of sexual dimorphism in pain behavior or transcriptomics response.
Normalized gene expression was computed as transcripts per kilobase of transcript/million mapped reads (TPM). Transcripts with low expression (TPM , 0.05) across 18 samples were excluded from further analysis. A total of 21,900 protein coding genes were identified and further considered for bioinformatic analysis (Fig. 2b). Differentially expressed (DE) protein-coding genes with fold-change (FC) 2 (log 2 FC 1, p 0.05) of TPM counts were grouped in MBP(84-104) and MBPS groups relative to PBS. SN samples displayed the greatest transcriptional response specific to MBP(84-104) in both sexes. Transcription patterns in female and male nerves exhibited a significant overlap. In contrast, fewer DE genes were identified in DRG and DSC, whereas the majority exhibited sexually dimorphic expression. In female DRG and DSC, we recorded a higher number of up-regulated relative to down-regulated genes, whereas males demonstrated a higher number of down-regulated relative to up-regulated genes in response to MBP(84-104).
The principal component analysis of the top DE genes (Fig.  2c) indicates that in sciatic nerve, MBP(84-104) was responsible for 96.2% of the variance of the top DE genes; only 3.5% variance was due to sex. Conversely, in DRG, sex was responsible for 41.4% variance, whereas MBP(84-104) contributed to 30% variance of the top DE genes. In the spinal cord, neither sex nor MBP(84-104) treatment contributed significantly to the variance of the top DE genes.
The hierarchic clustering analysis of the top DE genes in SN, DRG, and DSC ( Fig. 2d) exhibited notable sexual dimorphism in response to the injection (IS PBS group) in all tissues. It further confirmed that the vast transcriptome changes in female and male SN in response to MBP(84-104) were absent in the PBS group. Over 1,180 top up-regulated DE genes in SN, regardless of sex, were co-expressed in the corresponding DRG. Gene list enrichment analysis using ToppGene Suite (43) predicted that the SN/DRG co-expressed genes related to anterograde transsynaptic and ion channel signaling (data not shown). To further explicate the site-specific changes caused by IS MBP(84-104), we conducted predictive analysis of the biological functions and canonical signaling pathways in SN, DRG, and DSC using the curated content of the ingenuity pathway analysis (IPA) knowledge base (Qiagen Bioinformatics, Redwood City, CA, USA).
Robust signaling changes with minor sexual dimorphism in nerve exposed to MBP(84-104) At day 7 post-injection, IS MBP(84-104), but not MBPS, stimulated robust signaling activity in male and female SN (Figs. 3 and 4). Sexually dimorphic regulation of biological processes was detected by gene enrichment analysis (Fig. 3a) that revealed induction of transcriptional activity related to cell projection morphogenesis and vesicular localization preferentially active in female SN, whereas neurotransmitter secretion and transport were more dominant in male SN (Fig. 3a). The canonical path-ways in SN, induced selectively by IS MBP(84-104), but not MBPS (fold-change relative to PBS, Fig. 3b) displayed mild sexual dimorphism related to nociceptive, immune, bioenergetics, and lipid metabolism signaling, detailed below.
Immune pathways-Immune pathways, including T cellrelated interleukin (IL)-22 signaling and 4-1BB signaling in T lymphocytes, were activated in response to MBP(84-104) in both sexes, with higher activation scores in female SN (Fig. 3b). Innate immune pathways, related to inflammasome, neuroinflammation, Toll-like receptor, and pro-nociceptive cytokine signaling were not significantly regulated at day 7 after IS MBP (84-104) relative to MBPS in female or male SN (Fig. 3b). These data confirm our finding of T cell homing in SN after IS MBP (84-104) but not MBPS (29,32).
Vesicle localization-The vesicle localization process was selectively active in female SN exposed to MBP(84-104) (  nociceptive signaling pathways in DRG and DSC regulated by IS MBP(84-104).
Immune system pathways-The immune system pathways were induced in female DRG (Fig. 5) and DSC (Fig. 6) after IS MBP(84-104). In female DRG, leukocyte differentiation and activity GO processes (Fig. 5a) correspond to leukocyte extravasation signaling (Fig. 5b) related to leukocyte rolling and transmigration related genes (Fig. 5c). Female DRG displayed pro-inflammatory p38/MAPK, iNOS, and TNFR1 signaling activity. Selectively in female DSC, IS MBP(84-104) induced CD28 signaling in T helper cells, protein kinase C (PKC) signaling in lymphocytes, B cell receptor signaling, IL-8, iNOS, neutrophil, and MAPK signaling (Fig. 6b). In male DRG (Fig. 5b) and DSC (Fig. 6b), immune pathways were inactive in all groups, with the exception of macrophage migration inhibitory factor (MIF) regulation of innate immunity and nuclear factor of activated T cells (NFAT) in the immune response induced in DSC after IS MBP(84-104) (Fig. 6b). In male DRG (Fig. 5b) and DSC (Fig. 6b), IS MBP (84-104)-induced phosphoinositide 3-kinase (PI3K)/AKT signaling, suggesting an active regulatory process may exist in males that is absent in females.  Transcriptome of sexually dimorphic MBP pain in a sex-dependent manner. This network includes VGCC, CAMKII/calcineurin, and glutamate AMPA/mGLU receptor signaling, co-regulated (induced in females and suppressed in males) by estrogen receptor ESR1 and PLC-activated endoplasmic reticulum inositol trisphosphate receptor (IP3R) signaling cascades.
These findings support the multicellular "shared amplifier" paradigm of the ESR1/IP3R system contribution to female-prevalent hyperalgesic priming via IP3-mediated Ca 21 release from endoplasmic reticulum (ER) and Ca 21 signal amplification in spinal cord (47,48). According to this paradigm, mechanical allodynia in females is therapeutically sensitive to intrathecal (IT) administration of xestospongin C, a specific and reversible IP3R inhibitor that inhibits IP3-mediated Ca 21 release from ER stores (47,48). To test this model, we used von Frey testing to confirm mechano-allodynia at day 10 after IS MBP(84-104) in female mice. Administration of IT xestospongin C (1 nmol), but not 10% DMSO vehicle, mitigated IS MBP(84-104) allodynia for ;2 h (Fig. 7b). Intraperitoneal (i.p.) gabapentin (100 mg/kg), used as control anti-allodynia therapy effective in a clinical setting (49) and IS MBP(84-104) in female rats (32), effectively attenuated IS MBP(84-104) allodynia in female mice (Fig. 7b). Thus, the nerve exposure of autoantigenic myelin peptides initiates ESR1-nociceptive signaling in females (presumably due to high estrogen levels) that can be targeted by the endoplasmic reticulum IP3R inactivation in cells of spinal cord.

Discussion
Despite the high regenerative capacity of damaged peripheral nerve, an intractable neuropathic pain, mechano-allodynia, may persist for years after nerve repair. In evidence for autoimmune pathogenesis of mechano-allodynia pain specifically in females, we here demonstrate that nerve exposure to a pure autoantigenic MBP peptide, endogenously released post-injury (26,(28)(29)(30)(31)50), initiates female-specific mechano-allodynia. Our unbiased, comparative genome-wide transcriptional view of the respective sciatic nerve, DRG, and spinal cord of both sexes reveals striking multisite, sexually specific molecular changes after nerve MBP(84-104) exposure, leading to mechano-allodynia in females and no change in mechanical sensitivity in males. This dimorphism represents several organizing principles revealing a 3-tier signaling program, discussed below and illustrated in Fig. 8.

Tier 1: sex-specific utilization in lipid and cell energy metabolism
Myelin is a lipid-rich multilamellar membrane formed by Schwann cells and maintained via interactions of cationic MBP with anionic membrane phospholipids, such as PIP2 (23), and anionic proteins, including actin, tubulin, and Ca 21 -calmodulin (21,22). In agreement, we find the MBP(84-104)-induced mechano-allodynia is diminished with charge reduction by substitution of basic histidine 89 for amphoteric glycine 89 (30).

Strong transcriptional response in nerves of both sexes suggests
Transcriptome of sexually dimorphic MBP pain efficient MBP(84-104) uptake regardless of sex. Potential sex differences in cell membrane lipid composition or function may underlie sexual dimorphism of nociception induced by MBP. The cell type specificity of signaling activity obscured by bulk RNA-seq remains to be elucidated. In the injected nerve, Schwann cells constitute over 80% of cell population, whereas immune cells (particularly, macrophages and lymphocytes), fibroblasts, and endothelial cells may contribute to the present findings (51,52).
According to the established MBP-PIP2 interactions (23), MBP(84-104) causes an increase in PLC (which hydrolyzes PIP2 to IP3) in females and increase in PI3K (which mediates PIP2 to PIP3 synthesis) in males (tier 1). These changes may underlie subsequent changes, discussed in tier 2, below and illustrated in Fig. 8.
MBP(84-104) seems to induce a high level of ergogenic metabolic activity in male compared with female nerves, including Krebs/TCA cycle, glycolysis and oxidative phosphorylation signaling relative to females. In contrast, female nerves show increased superpathway of cholesterol biosynthesis signaling and reduced cholesterol efflux LXR/RXR signaling. Generally,   (50), MBP(84-104) induces PLC-driven PIP2 hydrolysis to IP3 in females and PI3K-driven PIP2 to PIP3 synthesis in males. Tier II, ESR1/nociceptive co-regulation: in female spinal cord, sustained ESR1/IP3 receptor (IP3R) co-activation (red arrows, due presumably to high estrogen levels relative to male) results in IP3R-mediated Ca 21 release from ER yielding Ca 21 -dependent nociceptive signaling via VGCC, AMPAR, mGLUR, and CaMKII increase. This ESR1/IPR3/nociception loop is repressed in male spinal cord (green arrows). Tier III, autoantigenicity: the autoantigenic MBP(84-104) epitope exposure causes T cell (T) activation in nerves of both sexes. T cellrelated activity advances to DRG and spinal cord in females, whereas stays localized to nerves in males. The three-tier signaling program can be interrupted by spinal IP3R inactivation (Fig. 7b) and (48) and T cell deletion (29).
peripheral nerves resemble other tissues that in response to stress, utilize carbohydrate and protein as an energy source in male cells, and lipid in females (53). In DRG and Schwann cell cultures, we find that MBP(84-104) controls cell energy metabolism and the lactate/glucose ratio via potential interaction with lactate dehydrogenase and mitochondrial voltage-dependent anion-selective channel protein (VDAC)-1 (34), and by modulating the ATPase activity of cyclin-dependent kinases (30).
Tier 2: sex-specific co-regulation of estrogen receptor, ER stress, and nociceptive signaling Presumably due to inherently high estrogen level in females relative to males, the female-specific PLC activity in tier 1, the estrogen receptor ESR1 co-activation with IP3R causes IP3Rmediated release of calcium ions from ER stores, leading to imbalance in Ca 21 homeostasis after nerve MBP(84-104) exposure. This further potentiates the vast Ca 21 -dependent nociceptive signaling via VGCC, AMPAR, mGLUR, and CAMKII/ calcineurin up-regulation exclusively in the female spinal cord, as illustrated in Fig. 8. The cells responsible for the induction of the predicted pathways remain to be elucidated. The femalespecific hyperalgesic priming has been previously attributed to co-regulation of ESR1/IP3R activity via imbalance in Ca 21 homeostasis (48).
Conversely, the ESR1/IP3R-nociceptive genes are suppressed in the male spinal cord presumably due to inherently low estrogen levels in males relative to females. Interestingly, many such nociceptive genes are locally up-regulated in male and female nerves exposed to MBP(84-104). This effect is not seen after PBS or MBPS injection, suggesting that in males, MBP(84-104) may stimulate an active analgesia (or an adverse hypoalgesia) mechanism. The present data set provides rich information on the transcriptional landscapes in MBP(84-104)-induced nerve changes, which could be useful for further data mining of the analgesia mechanism in males to develop new anti-allodynia approaches in both sexes. In contrast to the PLC-driven PIP2 hydrolysis to IP3 and activation of the IP3R/ESR1-nociceptive signaling, males induce PI3K activity (expected to mediate PIP2 to PIP3 synthesis) in the nerve, DRG, and spinal cord (Fig. 8). The cell specificity of the predicted signaling activity remains to be determined. Xestospongin C, whereas deemed a selective IP3R inhibitor, has the ability to inhibit the Ca 21 -ATPase of the ER causing depletion of both IP3-and ryanodine-sensitive Ca 21 stores (54, 55). Furthermore, mobilization of Ca 21 stores in glia (microglia, astrocytes) and neurons may result in distinct pro-and anti-nociceptive activities.
Male nerves exhibited no exosomal and vesicular localization gene activity seen in female nerves. This may suggest that MBP (84-104) exposure induces a pro-nociceptive retrograde axonal vesicular transport from the nerve to DRG in females that is absent in males. The transcriptomic analysis does not reflect activity of pro-and anti-nociceptive proteins, lipids, and metabolites. Future studies must elucidate the cellular, subcellular, and molecular mechanism that link estrogen receptor, ER stress, and nociceptive changes leading to sexual dimorphism in mechanical hypersensitivity.
IS MBP(84-104) does not model mechano-allodynia in male mice. In male mice with nerve injury, mechano-allodynia is thought to depend on activity of spinal microglia (11). To some degree, mechanical-allodynia in males may respond to T cell targeting (15)(16)(17)(18), including by immunization with an altered/ mutant MBP ligand (56), depending on species, strains, and types of neuraxial injury. In males, T cell-related signaling was apparent in nerve, at the site of nerve exposure, but not in DRG or spinal cord. These data suggest a more complex T cell function in neuropathic pain. Indeed, prior data using both sexes has implicated T helper (Th)1 and Th17 cells infiltrating the injured peripheral nerve and the segmental spinal cord in the development of mechano-allodynia (15)(16)(17)(18). The studies elucidating sex differences in T cell regulation after IS MBP(84-104) are ongoing.
The bulk RNA-seq that generated data from mixed cell types and the changes were not reliably attributed to different cell types. It is plausible that T cells access DRG lacking a neurovascular barrier (57). Non-immune cells may contribute to adaptive immune pathway activity, particularly in spinal cord with intact blood-spinal cord barrier. In female rats exposed to IS MBP(84-104), we find that astrocytes and deep dorsal horn neurons of ipsilateral spinal cord contribute to immune/IL-6related production (32). Using female athymic nude rats lacking mature T cells, we have conclusively established that maintenance of IS MBP(84-104) allodynia is T cell-dependent (29). It is presumed that via MBP(84-104) antigen-presenting nerve Schwann cells and macrophages assist in T cell homing to mechanosensory neurons leading to cytokine-mediated sensitization (29). The site and mechanisms of T cell function remain to be determined.
IS MBP(84-104) recapitulates focal nerve injury IS MBP(84-104) partially models events induced by endoneurial release of central MBP peptides during focal nerve injury (26,(28)(29)(30)(31)50). In contrast to nerve injury, IS MBP(84-104) allodynia is not accompanied by the acute degenerative or reparative processes (29,32), making it a more specific model of this intractable pain state. It is logical to assume that nerve injury induces a plethora of additional signaling cascades not recapitulated by IS MBP(84-104) that affect both males and females and mediates allodynia in males.
The pathogenic action of the MBP peptides is distinct from that of the full-length MBP in healthy tissues. After peripheral nerve injury, the MBP peptides translocate from the myelin internode toward the node (26,(28)(29)(30)(31) and escape macrophage phagocytosis (28), potentially by evading the scavenger lowdensity lipoprotein receptor-related protein 1 (LRP-1), which binds the full-length MBP (34). Nerve injury, required to release the cryptic epitopes concealed within the intact myelin from immunosurveillance (29), likely produces unique MBP peptide interactors, not captured by our IS MBP(84-104) model. Given the robust IS MBP(84-104) allodynia (produced by IS injection in intact nerve), we argue the injury-specific interactors are not essential to allodynia. The pro-algesic action of the MBP peptides may employ the full-length MBP interactors, including phospholipids.
From a clinical perspective, in addition to focal nerve trauma, MBP(84-104) is released in painful neurodegenerative and autoimmune demyelinating states, including multiple sclerosis (30,31). In support of autoimmune pathogenesis of mechanoallodynia, we find anti-MBP(84-104) autoantibodies in the serum of female rats with focal nerve injury and women with multiple sclerosis (31). The pathogenic role of these autoantibodies in pain is under current investigation. Other demyelinating triggers, including drugs, toxins, or pathogens, may produce MBP-mediated pain states. In addition, MBP epitope release via targeted proteolysis is seen before demyelination and after remyelination (29)(30)(31), potentially implicating MBP in highly prevalent idiopathic pain states. Based on its high structural homology with muscarinic M2 acetylcholine receptor, MBP(84-104) may contribute to complex regional pain syndrome (33).
In conclusion, the present study implicates for the first time a myelin autoantigen in sexually dimorphic regulation of neuropathic pain. The female-specific T cell mediation of mechano-allodynia (11) may represent at least in part an autoimmune myelin-targeted sensitization of myelinated mechanosensitive afferents caused by integrative dysregulation of lipid/cholesterol metabolism and calcium-dependent nociception via increased estrogen receptor activity. Our rigorous molecular, cellular, and functional validation of the predicted pathways is ongoing. Our robust data, new concepts, and the unique comprehensive open-access database aim to improve diagnostic and therapeutic strategies of neuropathic pain management based on sex.
Animal model C57BL/6 mice (8-10-week-old adult, Envigo Laboratory, Huntingdon, United Kingdom) were housed in a temperaturecontrolled room (;22°C) on a 12-h light/dark cycle with free access to food and water. All procedures and testing were conducted during the light cycle. Under isoflurane anesthesia (Isothesia, Henry Schein, Melville, NY, USA), the common sciatic nerve was exposed unilaterally at the mid-thigh level through a gluteal muscle-splitting incision. A single endoneurial bolus injection of the MBP(84-104) and MBPS peptides (30 mg in 3 ml of PBS, each) was performed into the sciatic nerve fascicle (IS) using a 33-gauge needle on a Hamilton syringe. Sciatic nerve (5-mm injection site fragment), lumbar (L)4-5 DRG (pooled), and L1-L6 spinal cord, dorsal quarter (DSC), and ipsilateral to injection were excised and stored in RNA-later (Thermo Fisher Scientific) at 220°C. Animals were sacrificed using Beuthanasia IP (Merck Schering-Plough Animal Health, Kenilworth, NJ, USA). All animal procedures were performed according to the PHS Policy on Humane Care and Use of Laboratory Animals, ethical guidelines of the International Association for the Study of Pain and the protocol approved by the Institutional Animal Care and Use Committee at the University of California, San Diego.

von Frey testing
Behavioral testing was performed by the experimenter blinded to treatment groups. The animals were randomized and stratified to treatment groups and habituated to the testing environment prior to baseline tests. For assessment of mechanical withdrawal threshold, mice were placed in individual plexiglas compartments with wire mesh bottom and von Frey filaments (2.44 to 4.31 (0.04-2.00 g), Stoelting, Wood Dale, IL, USA) were applied perpendicularly to the mid-hind paw and held for 4-6 s. Testing was performed daily for 3 days prior and then daily after IS MBP(84-104) and MBPS injection. A positive response was noted if the paw was sharply withdrawn. The 50% probability of withdrawal threshold was determined by Dixon's up-down method (61). Data were analyzed using GraphPad Prism 4.04 (GraphPad, San Diego, CA, USA) by two-way analysis of variance (ANOVA) for repeated measures and Dunnett's post hoc test. p 0.05 values were considered significant.

RNA-seq
Total RNA was extracted from mouse sciatic nerve (n = 12, pooled), L4/L5 DRG (n = 6, pooled), and L1-6 DSC (n = 6, pooled) ipsilateral to IS injection using TRI Reagent (Zymo Research, Irvine, CA, USA) and purified using Direct-zol RNA purification reagents (Zymo Research). Samples were treated with DNase I during purification. The RNA purity was estimated by measuring the OD260/280 ratio using a ND-1000 Spectrophotometer (Thermo Fisher Scientific). RNA integrity was confirmed using 2100 Bioanalyzer (Agilent Technologies, San Diego, CA, USA), and RNA samples with RIN 7.0 were processed further. Poly(A) RNA was isolated using the NEB-Next ® Poly(A) mRNA Magnetic Isolation Module and barcoded libraries were made using the NEBNext ® Ultra II TM Directional RNA Library Prep Kit for Illumina ® (New England Biolabs, Ipswich, MA). Libraries were pooled and single end sequenced (1X75) on the Illumina NextSeq 500 using the High output V2 kit (Illumina Inc., San Diego, CA, USA).

RNA-seq data processing
RNA-seq read data were processed using BaseSpace software (Illumina Inc.). Reads were aligned to the Mus musculus genome (mm10 assembly, RRID:SCR_002344) using STAR aligner (63) with default settings. Transcript and gene level expression was computed as transcripts per kilobase of TPM using Salmon software (64). Differential gene expression was determined using the Cufflinks package (65) on the Galaxy platform (66). Global DE analysis was conducted according to the previously established bioinformatic procedures (67)(68)(69). FC for each transcript with p 0.05 was determined as ratios of TPM of IS MBP(84-104) or MBPS relative to the respective PBS control. Thresholds log 2 FC . 0.585 (up-regulated) or log 2 FC , -0.585 (down-regulated) relative to IS PBS were set to filter out insignificant changes in expression. Venn diagrams were plotted using R Limma package (70). Principal component analysis and hierarchical clustering were conducted using ClustVis engine (71). False discovery rate was controlled using Benjamini-Yekutieli p value correction (72). Canonical pathway analysis was performed using Ingenuity IPA (Qiagen Bioinfor-matics, Redwood City, CA, USA). p Values were calculated using right-tailed Fisher's exact test. Canonical pathways with p 0.05 were considered for analysis. The activation status of the canonical pathways was predicted by calculating IPA Zscores. Canonical pathways with p 0.05 and Z-score 2.0 (or 22.0) were considered significantly activated (or inhibited). Gene list enrichment analysis to determine top GO Biological Processes (RRID:SCR_002811) was done using Top-pGene Suite (43). GO biological processes with Bonferroni corrected p 0.05 were considered for analysis. Gene and protein classification was conducted using PANTHER (73).

Data availability
The full transcriptomics data of the nerve, DRG, and spinal cord samples of all treatment groups of both sexes are available in a public repository (GEO ID GSE107159). Conflict of interest-The authors declare that they have no conflicts of interest with the contents of this article.