Pronociceptive autoantibodies in the spinal cord mediate nociceptive sensitization, loss of function, and spontaneous pain in the lumbar disc puncture model of chronic back pain

Abstract

Previously we observed that B cells and autoantibodies mediated chronic nociceptive sensitization in the mouse tibia fracture model of complex regional pain syndrome (CRPS) and that CRPS patient antibodies were pronociceptive in muMT fracture mice. The current study used a lumbar spinal disc puncture (DP) model of low back pain in wildtype (WT) and muMT mice (lacking B cells and antibodies) to evaluate pronociceptive adaptive immune responses. Spinal discs and cords were collected 3 weeks after DP for PCR and immunohistochemistry analyses. WT DP mice developed 24 weeks of hindpaw mechanical allodynia and hyperalgesia, grip weakness, and a conditioned place preference response indicative of spontaneous pain, but pain responses were attenuated or absent in muMT DP mice. Spinal cord expression of inflammatory cytokines, immune cell markers, and complement components were increased in WT DP mice and in muMT DP mice. Dorsal horn immunostaining in WT DP mice demonstrated glial activation and increased neuronal C5a receptor expression. Serum collected from WT DP mice and injected into muMT DP mice caused nociceptive sensitization, as did intrathecal injection of IgM collected from WT DP mice, and IgM immune complexes were observed in lumbar spinal discs and cord of WT DP mice. Serum from WT tibia fracture mice was not pronociceptive in muMT DP mice and vice versa, evidence that each type of tissue trauma chronically generates its own unique antibodies and targeted antigens. These data further support the pronociceptive autoimmunity hypothesis for the transition from tissue injury to chronic musculoskeletal pain state.

Summary

Pronociceptive autoimmune responses in the spinal cord mediate pain behaviors, functional loss, and spontaneous pain in the lumbar disc puncture model of low back pain.

1. Introduction

Low back pain (LBP) is the most common cause of years lived with disease (YLD) world-wide according to statistics from the 2017 Global Burden of Disease Study [15]. The economic costs of LBP are estimated to amount to greater than $100 billion per year in the US, with most costs attributable to disability and lost productivity [26]. Although consensus guidelines for the management of LBP are available, most recommended therapies help only about half of those treated, gains in terms of pain control and functional recovery tend to be modest, and the durability of treatment effects is often limited [42; 45].

Most LBP is nonspecific and may have contributions from vertebral endplates, degenerative facet joints, spinal ligaments, axial musculature and intervertebral discs [1; 14; 37]. Of these, it is the intervertebral disc that has received the most attention, and many therapies including pharmacological agents, interventional procedures, and surgeries have all been used to address this presumed cause, although with limited effectiveness [1; 7; 9; 39; 41; 57]. In fact, back pain correlates poorly with the findings from structural imaging studies such as disc bulges, herniations, or loss of disc height [4; 5; 8; 24], challenging investigators to consider mechanisms beyond simple mechanical or structural changes as the cause of back pain [13].

Observational evidence suggests that LBP may be an autoimmune disease. There is an increased prevalence of LBP in females and significant heritability, similar to most autoimmune diseases[2; 38]. The intervertebral disc is the largest avascular structure in the body and lacks lymphatic vessels, thus isolating it from the immune system, but after disc herniation or degeneration there is ingrowth of vascular and lymphatic vessels from the surrounding connective tissues that could present disc antigens to the immune system and potentially trigger an adaptive immune response [25; 54]. Cerebrospinal fluid levels of IgG and IgM were found to be elevated in nearly half of LBP patients with herniated discs, and immunostaining of herniated and degenerative discs from LBP patients frequently demonstrated IgG and IgM pericellular deposition[22; 46; 49]. Furthermore, antibodies obtained from patient degenerative discs selectively bound to various extracellular matrix proteins, including collagens II, V, and aggrecan [6]. These data suggest that an adaptive immune response can potentially develop in human degenerative discs, but whether this response is pronociceptive is unclear.

A major challenge for the field, therefore, is to develop preclinical models suitable for both mechanistic and translational studies of pronociceptive autoimmune responses as a possible cause of chronic low back pain. The current study tests the hypothesis that lumbar disc puncture in mice induces B lymphocytes to produce pronociceptive autoantibodies that bind to injury induced neoantigens in the spinal discs and corresponding spinal cord, resulting in antibody-antigen complexes capable of activating complement and thus sensitizing nociceptive neurons with subsequent hindlimb mechanical allodynia and hyperalgesia, grip weakness, and spontaneous pain behavior.

2. Methods

2.1. Animals

These experiments followed the animal subjects guidelines of the International Association for the Study of Pain and the ARRIVE guidelines, and these studies were approved by the Veterans Affairs Palo Alto Health Care System Institutional Animal Care and Use Committee (Palo Alto, CA, USA). The majority of experiments used 3 months old male muMT mice lacking mature B cells and immunoglobulin [28; 50], on a C57BL/6J congenic background (#002288, Jackson Laboratory, Bar Harbor, ME), with wildtype C57BL/6J controls (#000664, Jackson Laboratory). One series of experiments used 3 months old female muMT mice with wildtype C57BL/6J controls. Data collection was conducted blind to group assignment and all experimental designs complied with the ARRIVE guidelines.

2.2. Surgery

The lumbar intervertebral disc puncture mouse model replicates the nucleus pulposus extrusion, loss of disc height, proteoglycan depletion, and annular fibrotic disorganization observed in LBP patient herniated/degenerated intervertebral discs [51]. To generate the spinal disc puncture (DP) model, under deep isoflurane anesthesia (2% at 1L/min) the mice were placed in a supine position, and the abdominal skin was shaved and sterilized. A 1.5 cm left lateral skin incision was made and the ventral peritoneum incised and retracted. The space between the dorsal peritoneum and the left psoas major muscle was exposed and the peritoneum incised to allow access to the ventral spine. The L4/5, L5/6 and L6/S1 discs were identified utilizing the pelvic rim landmark and the spinal discs were punctured at midline with a microscalpel to a depth of 0.7–0.8 mm, guided by a polyethylene stopper sleeve. A repeat puncture through the same incision was performed using a curved tip microscalpel to destroy and extrude the nucleus pulposus tissue. The ventral peritoneum and skin were then closed and the mice treated with subcutaneous buprenorphine SR (1mg/kg, s.c. once), enrofloxacin (5mg/kg, s.c. once), and saline (1ml, s.c. once). Sham operated mice underwent ventral skin incision and peritoneal incision and retraction, without spinal disc puncture, then the incisions were closed and the mice treated with subcutaneous buprenorphine SR, enrofloxacin, and saline.

One set of experiments utilized the tibia fracture mouse model of complex regional pain syndrome (CRPS) [3]. The fracture model was performed in 3 month-old male mice as previously described [20]. Under isoflurane anesthesia a hemostat was used to make a closed fracture of the right tibia just distal to the middle of the tibia. The hindlimb was then wrapped in casting tape (Delta-Lite, BSN Medical, Hamburg, Germany) so the hip, knee and ankle were all fixed. After fracture and casting, the mice were given subcutaneous buprenorphine SR, enrofloxacin, and saline. At 3 weeks after surgery the mice were anesthetized with isoflurane and the cast removed. All mice had union at the fracture site by manual inspection.

2.3. Hindpaw nociceptive testing

Hindpaw plantar mechanical allodynia was assayed using von Frey filaments according to the “up-down” algorithm as we have previously described [31]. Estimation of the mechanical withdrawal threshold by data-fitting algorithm permitted the use of parametric statistics for analysis [44].

Hindpaw pinch hyperalgesia was assayed using a Rodent Pincher algometer (Bioseb, Pinellas Park, FL) utilizing a calibrated forceps to apply linearly increasing mechanical stimulation over the mid-paw metatarsals (starting pressure 5 g, cutoff pressure 400 g, testing duration to cutoff 10 s, measure tolerance 10 %). Testing was repeated 3 times at 5 min intervals and the withdrawal thresholds averaged. This test is similar to the Randall-Sellito assay and is used to measure mechanical hyperalgesia and analgesic drug effects with minimal restraint and variability[36]. The delta (%) for pinch withdrawal thresholds was calculated as ((3wk post surgery value – baseline value)/baseline value) x 100), so a negative value represents mechanical hyperalgesia.

Hotplate heat hyperalgesia was measured using a Hotplate Analgesia Meter (Columbus Instruments, Columbus, OH. The surface temperature was set at 52 C with a cut off of 30 s to avoid skin damage. A positive response was recorded when the mouse exhibited a lick, flick, or jump response. The test was repeated 3 times at 3 min intervals and the response latencies averaged.

Tail immersion testing was performed using a water bath set at 48 C and the distal 1/3 of each mouse’s tail was immersed into the water bath and the withdrawal latency recorded. The test was repeated 3 times at 3 min intervals and the response latencies averaged.

2.4. Hindpaw grip strength testing

Grip strength was measured using a Grip Strength Meter (Bioseb, Pinellas Park, FL), utilizing a digital strain gauge to measure tension force applied to a metal grid. Each mouse was lowered by its tail onto the grid and when the mouse grasped the grid with its forepaws and hindpaws the tail was slowly pulled in a direction parallel to the grid until the mouse released its grip. The force applied to the grid just before the mouse released its grip was recorded as the peak tension. This procedure was repeated 3 times at 3 min intervals and the grip strength measurements were averaged. The delta (%) for grip strength was calculated as ((3wk post surgery value – baseline value)/baseline value) x 100), so a negative value represents grip weakness.

2.5. Burrowing behavior

The burrowing assay measures the mouse’s instinctual burrowing activity. A plastic tube 20cm long and 10cm in diameter was filled with 140 g of cage bedding material, then closed on each end and placed into each mouse’s cage after all the bedding in the cage had been removed. After the mouse acclimated for 5 min to the tube being in its cage the tube was opened on one end and that end of the tube was placed in a slightly elevated (3cm) position off the floor of the cage. The mouse was allowed 10 minutes to spontaneous burrow into the tube full of cage bedding, then the amount of bedding material on the cage floor was weighed. Two consecutive burrowing test sessions (30 min interval) were averaged for the final result.

2.6. Conditioned placed preference testing

A single trial conditioned place preference (CPP) test was used to assess affective component of pain (tonic pain) after lumbar disc puncture surgery, as described previously [56]. The experiments were done using a standard 3-compartment CPP box placed inside sound attenuating cabinets. The cabinets had controlled lighting to eliminate chambers bias. Initial 3 days were for habituation where the mice had free access to all three chambers for 30 minutes each day. On day 3, the exploration time of each mouse for the first 15 minutes was recorded by a video camera to determine if any animal had an inherent chamber bias. The amount of time spent in each of the outside chambers was determined as pre-conditioning place preference. Any mouse that spent more than 80% or less than 20% of the total experimental time in either of the outside chambers was excluded from further study. On day 4, mice received saline injections first, and were randomly assigned to one of the outside chambers and remained there for 50 min. After a 4-hour break, mice were placed in the opposite outside chamber (conditioning chamber) after getting a low-dose morphine (0.3 mg/kg, i.p.) for 50 minutes. On Day 5, mice were again allowed free access to all three chambers for 15 minutes. Time spent in the conditioning chamber was recorded and compared to that measured during the pre-conditioning phase. The preference score for each mouse was calculated as: Time spent in the drug compartment – Time spent in vehicle compartment. Chamber assignments were randomized and counterbalanced between groups and video recordings were analyzed by TopScan (Clever Sys Inc., Reston, VA) for time spent in each of the two active association chambers. The delta (s) for CPP was calculated as time spent in the morphine chamber - time spent in the vehicle chamber, so a positive value represents spontaneous pain.

2.7. Real-time polymerase chain reaction (PCR) assays for inflammatory cytokines, immune cell markers, and complement components and receptors

At 3 weeks after DP the mice were euthanized and the L4/5, L5/6, and L5/6 lumber discs and the L4, 5 lumber spinal cord tissues were collected. Total RNA was extracted with the RNeasy Mini Kit (Qiagen, Hilden, Germany), and the purity and concentration were determined spectrophotometrically. Then complementary DNA was synthesized from 1 µg RNA using RT² First Strand Kit (Qiagen). Real-time polymerase chain reactions were performed on an ABI 7900HT sequencing detection system (Applied Biosystems, Waltham, MA) using RT² qPCR Primer Assay and the RT² SYBR Green qPCR Mastermix (Qiagen, Hilden, Germany). Interleukin 1 beta (IL-1β), interleukin 6 (IL-6), tumor necrosis factor (TNF-α), nerve growth factor (NGF), B-lymphocyte antigen CD20, complement component 1, q subcomponent, alpha polypeptide (C1qa), complement component 1, q subcomponent, beta polypeptide (C1qb), complement component 1, q subcomponent, C chain (C1qc), hemolytic complement (HC, C5), complement component 5a receptor 1 (C5aR1), complement component 5a receptor 2 (C5aR2) and 18S primer sets were validated on dissociation curves to document single product formation and agarose gel analysis was conducted to confirm the size. The data from real-time polymerase chain reaction experiments were analyzed as described in the manufacturer’s manual for the ABI 7900HT sequencing detection systems. All results were confirmed by repeating the experiment three times.

2.8. Immunohistochemistry methods in the lumbar spinal cord.

Lumbar spinal cord tissues were collected, embedded, sectioned (20-um), permeabilized and blocked, exposed to primary and secondary antibodies, mounted on slides, and the sections were imaged using confocal microscopy as previously described [30; 32]. Spleen and popliteal lymph nodes were collected and prepared as positive controls for CD20 and CD19 immunostaining. Primary antibodies included rabbit anti Iba1, 1:500 (WAKO, Osaka, Japan), rabbit anti GFAP, 1:2500 (DAKO), mouse anti neuronal nuclei (NeuN), 1:1000 (EMD Millipore Corp., Burlington, MA), rat anti C5aR1 monoclonal antibody, 1:500 (Abcam, Cambridge, UK), rabbit anti-CD20 monoclonal antibody, 1:200 (Abcam, Cambridge, UK), and rat anti-CD19 monoclonal, 1:400 (Invitrogen, Carlsbad, CA). Secondary antibodies included Alexa Fluor 594-conjugated affinity pure donkey anti-rabbit IgG, 1:1000, Alexa Fluor 594-conjugated affinity pure donkey anti-mouse IgG, 1:1000, and Alexa Fluor 488-conjugated affinity pure donkey anti-rat IgG, 1:1000 (Jackson ImmunoResearch, West Grove, PA), incubated with respective primary antibodies. The anti-CD20 and anti-CD19 immunostaining experiments were counterstained with DAPI 1:3000, (Thermo Scientific, Waltham, MA) to identify nuclei. Control experiments included incubation of slices in primary and secondary antibody-free solutions and primary antibody pre-absorption control, all of which led to low intensity non-specific staining patterns in preliminary experiments (data not shown). The number of Iba1+ and C5aR+NeuN+ immunostained cells per high-power field (HPF, x400) and the intensity of GFAP staining per HPF were assessed in the dorsal horn of the spinal cord by a blinded observer. The fields for each HPF were randomly selected and 4–6 HPFs were analyzed per sample.

2.9. Serum transfer experiments in spinal disc puncture and tibia fracture mice

This experiment examined the pronociceptive effects of wildtype male mouse serum, collected at 3 weeks after spinal disc puncture (DP), when injected intraperitoneally into 3 weeks post-DP muMT mice, respectively. Blood was collected by transcardial puncture in isoflurane anesthetized 3 weeks post DP or sham operated wildtype mice. The blood was left undisturbed at room temperature for 60 min to allow clotting, then the blood samples were centrifuged at 1,500g for 15 min at 4°C and the serum supernatants were aliquoted and frozen at − 80°C. After baseline hindpaw von Frey fiber threshold testing muMT mice were anesthetized and underwent DP surgery. At 3 weeks post-DP the hindpaw von Frey testing was repeated and then the muMT DP mice were injected with the 3 weeks post-DP or sham operated wildtype mouse serum (0.5ml). Further von Frey testing was performed at 1, 3, 5, 7, 14, and 21 days after injection. As an additional control, 3 weeks sham operated muMT mice were injected with 3 weeks DP wildtype serum. Additional experiments examined the pronociceptive effects of 3 weeks post DP wildtype serum in 3 weeks post fracture muMT mice and the pronociceptive effects of 3 weeks post-fracture wildtype mouse serum in 3 weeks post DP muMT mice.

2.10. IgM and IgG antibody injection experiments in spinal disc puncture mice

These experiments examined the pronociceptive effects of disc punctured wildtype mouse IgM or IgG antibodies when injected intrathecally into post DP muMT mice. Under isoflurane anesthesia, transcardial puncture was performed in 3 weeks post DP wildtype male mice and the serum was collected. IgM and IgG from serum were purified with IgM and IgG purification kits (Advanced BioReagents, Hayward, CA) following the manufacturer’s instructions. Briefly, diluted mouse serum was applied to the IgM affinity column. Subsequently, the flow through was applied to the protein G column for IgG purification. After the columns were washed with binding buffer, the bound IgM and IgG were eluted using the acidic elution buffers included in the kits. Coomassie Blue Protein Assay (Bio-rad, Hercules, CA) was used to identify and combine elution fractions that contain the purified antibody followed by dialysis (Slide-A-Lyzer Dialysis Devices 3.5K MWCO, Thermo Scientific, Rockford, IL) in PBS. The dialyzed IgM and IgG concentrations were adjusted with PBS to 1ug/ul and 2ug/ul, respectively, prior to injection. The yield of IgM and IgG were approximately, 220 ug/ml and 1.0 mg/ml serum, respectively. The purity of IgM and IgG was confirmed by SDS-PAGE followed by Coomassie Blue staining.

The dose of DP wildtype mouse IgM used for intrathecal and intraplantar injections in the current study (5ug/5ul, IT or IPL) is the same as the CRPS patient IgM dose we previously demonstrated was pronociceptive after intrathecal or intraplantar injections in the muMT tibia fracture mouse model of CRPS [21]. The IgG dose used for intrathecal injections in the current study (10ug/5ul) was twice the dose used for IgM intrathecal injections. To test the pronociceptive spinal effects of the IgM and IgG immunoglobulin isotypes, muMT mice underwent DP surgery and 3 weeks later baseline hindpaw von Frey threshold testing was performed, then the mice were intrathecally injected with either IgM (5ug/5ul) or IgG (10ug/5ul) collected from 3 weeks post DP wildtype male mice, or IgM (5ug/5ul) collected from 3 weeks post sham surgery wildtype mice. Additional von Frey testing was performed at 0.5, 1, 3, 6, and 24 hours and 7 days after injection. Another experiment tested the effects of intraplantar injection of 3 weeks post DP wildtype IgM (5ug/5ul) into 3 weeks post DP muMT mice.

2.11. IgM and IgG western blot analysis of spinal discs and cords from disc puncture mice

Mouse lumbar and cervical discs and spinal cord were harvested at 3 weeks after disc puncture or sham surgery and stored at −80°C. All tissues were homogenized in ice cold Tris buffer with 0.7% (v/v) β-mercaptoethanol and 10% glycerol. Lysates were centrifuged at 13,000g for 15 min at 4°C. Equal amounts of protein were size fractionated by SDS–PAGE and transferred onto a polyvinylidene difluorided membrane. The blots were blocked overnight with 5% milk in Tris-buffered saline with 0.5% Tween-20 (TBST), and incubated with primary antibodies against IgM, IgG or β-actin (Bio-Rad Laboratories, Hercules CA) for 1 hr on a rocking platform at room temperature. After washing in TBST, the blots were incubated with secondary antibody for 1 h at room temperature. The membrane was then washed again, and proteins were detected using ECL chemiluminescence reagent (ThermoFisher Scientific, Waltham, MA). Images were obtained using ChemiDoc MP Image Systems (Bio-Rad Laboratories, Hercules CA) and analyzed by National Institutes of Health ImageJ.

2.12. Enzyme immunoassay screening of IgM and IgG antibody binding to potential neoantigens

Native or recombinant proteins were purchased for enzyme immunoassay (EIA) determination of specific IgM and IgG binding reactivity in the sera of 3 weeks post DP wildtype male mice and 3 weeks post sham operated control mice. The protein candidates screened included; 1) aggrecan (bovine cartilage, Sigma, St. Louis, MO), 2) collagen II (native, Novus Biologicals, Littleton, CO), 3) collagen V (native, Novus Biologicals, Littleton, CO), 4) glial fibrillary acidic protein (GFAP, Aviva Systems Biology, San Diego, CA), 5) N-methyl-D-aspartate receptor B2 (NMDAR B2, Aviva Systems Biology, San Diego, CA), and 6) Actin (Cytoskeleton, Denver, CO). All proteins were either native or full-length. These potential autoantigens were previously identified in a proteomic survey of the mouse intervertebral disc [40] and are also present in spinal cord tissues. In addition, a previous study demonstrated the presence of IgG antibodies in cultured human degenerated disc media that preferentially bind to aggrecan, collagen II, and collagen V [6]. Each native or recombinant protein (250ng/50ul/well) was applied to an EIA plate, incubated at 4 °C overnight. After the plate was further blocked for 2h with Sea Block (250ul/well, Thermofisher Scientific, Waltham, MA), the plate was incubated with human sera (1:500 dilution in Sea Block 5x diluted with 0.05% Tween-20 PBS (TPBS)) at 4 °C overnight, followed by incubation with anti-mouse IgM-HRP or IgG-HRP secondary antibody (1:10000 dilution in Sea Block 5x diluted with TPBS) for 1 hour. After 4 washes with TPBS, optical densities at 450nm were developed with HRP substrate (50 ul/well, R&D, Minneapolis, MN) followed by addition of 2N H₂SO₄. (50ul/well).

2.13. Statistical analysis

Statistical analysis was performed using a two-way repeated measures ANOVA or a one-way ANOVA with Sidak multiple comparisons test for post-hoc contrasts. Data are presented as the mean ± standard error of the mean, and differences are considered significant at a P value less than 0.05 (Prism 5, GraphPad Software, San Diego, CA).

3. Results

3.1. B cells contributed to the development of chronic nociceptive sensitization, functional impairment, and spontaneous pain behavior after lumbar disc puncture in male mice.

Figure 1 illustrates that lumbar disc puncture (DP) in male wildtype C57BL6 male mice, compared to sham operated wildtype male mice, initiated the development of chronic hindpaw von Frey allodynia, pinch hyperalgesia, grip strength weakness, and a positive conditioned place preference (CPP) response indicative of spontaneous pain behavior. Hot plate withdrawal latencies, spontaneous burrowing behavior, and tail immersion withdrawal latencies were unchanged in the wildtype DP mice, relative to wildtype sham surgery mice. MuMT mice lacking mature B cells and antibodies failed to develop hindpaw allodynia, grip strength weakness, or CPP responses after DP surgery. Hindpaw pinch hyperalgesia did develop in muMT mice after DP surgery, but was attenuated. Critically, pinch testing thresholds in the forepaws of DP wildtype mice at 1 and 3 weeks post surgery were unchanged relative to presurgical baseline thresholds (Supplemental Fig.1), indicating that DP induced nociceptive sensitization was regionally restricted to the hindlimbs innervated by the lumbar spinal cord.

Figure 1. Lumbar disc puncture caused the development of chronic nociceptive sensitization, functional impairment, and spontaneous pain behavior after disc puncture in male mice.

After lumbar disc puncture (DP) male wildtype (WT) mice exhibited chronic hindpaw von Frey allodynia (A), pinch hyperalgesia (B), grip strength weakness (C), and a positive conditioned place preference (CPP) response indicative of spontaneous pain behavior (D). DP surgery in 3 weeks post-surgery male muMT mice lacking mature B cells and antibodies did not alter hindpaw von Frey thresholds, grip strength, or CPP responses, compared to sham operated mice. The muMT DP mice exhibited attenuated pinch hyperalgesia response, relative to WT DP mice, at 3 weeks post-surgery. Hot plate withdrawal latencies (E), spontaneous burrowing behavior (F), and tail immersion withdrawal latencies (G) were unchanged in the WT DP mice, relative to WT sham surgery mice. The delta (%) for pinch (B) and grip strength (C) was calculated as ((3wk post surgery value – baseline value)/baseline value) x 100) and the delta (s) for CPP (D) was calculated by subtracting the time spent in the vehicle chamber from time spent in the morphine chamber. A two-way repeated measures ANOVA was used to test the effects of each treatment group on the dependent variables over time, a one-way ANOVA was used to test for effects between groups at a single time point, and a Sidak correction test was used to test for post hoc contrasts. Data are expressed as mean values ± SEM, n = 12 per cohort. * P < 0.05, ** P < 0.01, *** P < 0.001, and **** P < 0.0001 for differences between the surgery and sham groups, and in (B) # P<0.05 and ## P < 0.01 for differences between DP muMT and DP WT groups. WT: wildtype mice, muMT: mice lacking mature B cells and antibodies, DP: disc puncture surgery, Sham: sham DP surgery, BL: baseline, 3wk: 3 weeks after DP surgery, 12wk: 12weeks after DP surgery

3.2. B cells also contributed to the development of chronic nociceptive sensitization, functional impairment, and spontaneous pain behavior after lumbar disc puncture in female mice.

Spinal disc puncture in female mice had behavioral effects similar to those observed after DP surgery in male mice. Female DP wildtype mice exhibited chronic hindpaw von Frey allodynia, pinch hyperalgesia, grip strength weakness, and a positive conditioned place preference (CPP) response indicative of spontaneous pain behavior (Fig. 2). DP surgery in female muMT mice lacking mature B cells and antibodies had no effect on hindpaw von Frey thresholds, grip strength, or CPP responses and attenuated the hindpaw pinch hyperalgesia response, mirroring the results observed in male muMT DP mice and indicating a crucial pronociceptive role for B cells and autoantibodies in the development of disc injury induced chronic nociceptive sensitization in both sexes. The primary difference between the effects of DP in male and female mice was the duration of nociceptive sensitization, functional impairment, and spontaneous pain behavior after DP surgery, with these behaviors persisting for at least 12 weeks after surgery in male wildtype mice and for only 8 weeks in female wildtype mice.

Figure 2. Disc puncture also caused the development of chronic nociceptive sensitization, functional impairment, and spontaneous pain behavior in female mice.

Similar to male DP mice, lumbar disc puncture (DP) in female wildtype (WT) mice initiated the development of chronic hindpaw von Frey allodynia (A), pinch hyperalgesia (B), grip strength weakness (C) and a positive conditioned place preference (CPP) response indicative of spontaneous pain behavior (D). DP surgery in female muMT mice had no effect on hindpaw nociceptive thresholds, grip strength, or CPP responses, indicating a crucial pronociceptive role for B cells and autoantibodies in the development of disc injury induced chronic nociceptive sensitization. The primary difference between male and female mice was the duration of nociceptive sensitization, functional impairment, and spontaneous pain behavior after DP surgery, with these behaviors persisting for up to 16 weeks after surgery in male mice and for only 8 weeks in female mice. A two-way repeated measures ANOVA was used to test the effects of each treatment group on the dependent variables over time, a one-way ANOVA was used to test for effects between groups at a single time point, and a Sidak correction test was used to determine post hoc contrasts. Data are expressed as mean values ± SEM, n = 12 per cohort. * P < 0.05, ** P < 0.01, *** P < 0.001, and **** P < 0.0001 for differences between the surgery and sham groups. WT: wildtype mice, muMT: mice lacking mature B cells and antibodies, DP: disc puncture surgery, Sham: sham DP surgery, BL: baseline, 3wk: 3 weeks after DP surgery, 12wk: 12 weeks after DP surgery

3.3. Disc puncture did not affect expression of inflammatory mediators, immune cell markers, or complement components or their receptors in the spinal discs of wildtype mice.

Three weeks after DP in wildtype male mice there were no changes in mRNA expression, relative to 3 weeks sham-operated mice, of IL-1β, IL-6, TNF-α, NGF, CD20, C1qa, C1qb, C1qc C5, C5aR1, and C5aR2 in the lumbar spinal discs (Fig. 3).

Figure 3. Disc puncture had no effect on spinal disc expression of inflammatory mediators, immune cell markers, or complement components.

Three weeks after DP in wildtype male mice there were no changes in lumbar spinal disc IL-1β, IL-6, TNF-α, NGF, CD20, C1qa, C1qb, C1qc C5, C5aR1, and C5aR2 mRNA expression levels, relative to 3 weeks sham operated male mice. A one-way ANOVA was used to test for effects between groups and a Sidak correction test was used to determine post hoc contrasts. Data are expressed as mean values ± SEM, n = 12 per cohort. DP: disc puncture surgery, Sham: sham DP surgery

3.4. Disc puncture induced up-regulated expression of inflammatory mediators, immune cell markers, and complement components in the lumbar spinal cord.

Disc puncture surgery induced the up-regulated expression of IL-1β, IL-6, TNF-α, CD20, C1qb, and C5aR1 mRNA in the lumbar spinal cord of wildtype male mice at 3 weeks post surgery, relative to sham operated wildtype male mice at 3 weeks post surgery (Fig. 4). Similarly, DP muMT mice also exhibited increased spinal cord expression of inflammatory mediators, immune cell markers, and complement components at 3 weeks post surgery, paralleling the changes observed in WT mice and indicating that these inflammatory responses to disc injury did not require the presence of mature B cells and autoantibodies.

Figure 4. Disc puncture evoked increased spinal expression of inflammatory mediators, immune cell markers, and complement components.

Lumbar spinal cord expression of TNF-α, IL-1β, IL-6, CD20, C1qb, and C5aR1 mRNA was up-regulated at 3 weeks after DP in wildtype male mice, relative to sham operated wildtype male mice. Similarly, DP muMT mice also exhibited increased spinal cord expression of inflammatory mediators, immune cell markers, and complement components at 3 weeks post surgery, paralleling the changes observed in WT mice, indicating that these responses to disc injury did not require the presence of mature B cells and autoantibodies. A one-way ANOVA was used to test for effects between groups and a Sidak correction test was used to determine post hoc contrasts. Data are expressed as mean values ± SEM, n = 12 per cohort. P < 0.05, ** P < 0.01, and *** P < 0.001 for differences between the surgery and corresponding sham groups. WT: wildtype mice, muMT: mice lacking mature B cells and antibodies, DP: disc puncture surgery, Sham: sham DP surgery

3.5. Disc puncture induced microglial activation and proliferation in the lumbar spinal cord with increased C5a receptor expression observed in dorsal horn neurons.

Previously we observed that tibia fracture in mice induced spinal microglial proliferation and increased C5a receptor expression in microglia and dorsal horn spinal neurons at 3 weeks post injury [52]. Figure 5 illustrates that 3 weeks after lumbar DP in wildtype mice there were increased numbers of Iba1 immunostained activated microglia exhibiting extensive ramification and increased cell body size, but C5a receptor co-labeling was not detected in the spinal microglia. Figure 6 illustrates that DP surgery had no effect on dorsal horn astrocyte GFAP immunostaining intensity and there was only minimal C5a receptor immunostaining in astrocytes after DP surgery. C5a receptor immunostaining was increased in the dorsal horn neurons after DP (Fig. 7), consistent with the spinal cord PCR results demonstrating increased C5a receptor expression in the lumbar spinal cord at 3 weeks after DP surgery in wildtype mice vs sham operated controls (Fig. 4).

Figure 5. Disc puncture induced microglial activation and proliferation in the lumbar spinal cord, but there was minimal C5a receptor expression in spinal microglia.

Previously we observed that tibia fracture in mice induced spinal microglial proliferation and increased C5a receptor expression in microglia. This figure illustrates that 3 weeks after lumbar disc puncture in wildtype male mice there were increased numbers of Iba1 immunostained activated microglia exhibiting extensive ramification and increased cell body size, but C5a receptor co-staining was not detected in spinal microglia. The top row shows representative confocal images of Iba1 (red, a microglia marker) immunostaining in the L4,5 dorsal horn sections of Sham (A) and DP (B) operated mice. There was a dramatic increase in large bodied ramified Iba+ microglia and a 3-fold increase in the absolute numbers of Iba1 stained microglia per high powered field at 3 weeks after disc puncture (C). The bottom row includes representative images of Iba1 (D) and C5aR (E, green) immunostaining showing minimal co-staining (yellow) in the merged image (F). Scale bars are 25 um (A, B) and 50 um (D, E, F). An unpaired t-test was used to test differences between treatment groups and data is presented as mean values ± SEM, n = 5 per cohort. ** P < 0.01 for DP vs Sham. Sham: sham operated mice, DP: disc puncture mice, C5aR: C5a complement receptor, HPF: high powered field

Figure 6. Disc puncture had no effect on astrocyte GFAP staining intensity in the lumbar spinal cord and only minimal C5a receptor expression was observed in spinal astrocytes.

Previously we observed that tibia fracture in mice induced spinal astrocyte proliferation and activation. This figure presents GFAP (red, an astrocyte marker) and C5a receptor (C5aR, green) immunostaining results in the L4,5 dorsal horn sections from 3 weeks post sham surgery (A,B,C) and lumbar disc puncture (D,E,F) wildtype male mice. There was no change in the intensity of GFAP immunostaining in the lumbar dorsal horns (G) and only minimal C5a receptor co-staining was observed in GFAP labeled astrocytes (C, F). Scale bar is 50 um. An unpaired t-test was used to test differences between treatment groups and data is presented as mean values ± SEM, n = 5 per cohort. Sham: sham operated mice, DP: disc puncture mice, C5aR: C5a complement receptor, HPF: high powered field

Figure 7. Disc puncture increased C5a receptor expression in dorsal horn neurons.

C5a receptor immunostaining was increased in dorsal horn neurons at 3 weeks after DP surgery, consistent with the spinal cord PCR results demonstrating increased C5a receptor expression in the lumbar spinal cord of 3 weeks post-DP wildtype mice (Fig. 4). The top row shows representative images of NeuN (A, red, a neuron marker) and C5aR (B, green) immunostaining in the L4,5 dorsal horn sections of sham operated mice showing minimal co-staining (yellow) in the merged image (C). The bottom row presents immunostaining for NeuN (D, red) and C5aR (E, green) in the DP mice and demonstrates extensive co-staining (yellow) in the merged image (F). There was a 1-fold increase in NeuN + C5aR co-labeled neurons in DP mice, vs sham operated controls (G). Scale bar is 50 um. An unpaired t-test was used to test differences between treatment groups and data is presented as mean values ± SEM, n = 5 per cohort. * P < 0.01 for DP vs Sham. Sham: sham operated mice, DP: disc puncture mice, NeuN: neuron marker, C5aR: C5a complement receptor, HPF: high powered field

Interestingly, CD20+ mRNA expression in the spinal cord was increased at 3 weeks after DP surgery, relative to sham operated controls, in both wildtype and muMT mice (Fig.4). To determine whether B cells infiltrated the lumbar spinal cord tissues after spinal disc injury, spinal cord sections were immunostained using used several antibodies targeting CD20 and CD19 cell surface proteins, using the spleen and lymph node tissues used as positive controls. CD20 is a cell surface protein B cell marker that is observed in late pre-B cells, but not pro-B cells or plasma blasts or plasma cells [43]. Supplementary Figure 2 illustrates the abundance of CD20+ immunostained cells in the spleen (and lymph nodes, data not shown) and the complete absence of CD20+ cells in the spinal cord of 3 weeks post DP mice. Immunostaining for CD19+ cells (a specific B cell marker that is present in the majority of plasma cells [27]) also demonstrated an abundance of CD19+ cells in the spleen (and lymph nodes, data not shown) and an absence of CD19+ cells in the spinal cord of 3 weeks post DP mice (Supplemental Fig.3). The discrepancy between the PCR and immunohistochemistry data for the presence of CD20 in the spinal cords of sham and DP operated mice may reflect the sensitivity of the PCR assay or possible contamination at the time of tissue collection, but the CD20 and CD19 immunostaining data conclusively demonstrate that B cells did not infiltrate the spinal cord tissues after disc puncture in mice.

3.6. Injecting serum from disc puncture wildtype mice into disc puncture muMT mice, but not into sham operated muMT mice or tibia fracture muMT mice, had pronociceptive effects.

Intraperitoneal injection of serum (0.5 ml) from 3 weeks post DP wildtype male mice (but not from sham operated mice) into 3 weeks post DP muMT male mice caused hindpaw von Frey allodynia that peaked by 7 days and resolved by 14 days after injection (Fig. 8A). Intraperitoneal injection of serum from 3 weeks post DP wildtype male mice into 3 weeks post sham operated muMT male mice had no effect on von Frey thresholds (Fig. 8A). Intraperitoneal injection of serum from 3 weeks post tibia fracture wildtype male mice into 3 weeks post DP muMT male mice had no effect (Fig. 8B), but when 3 weeks fracture wildtype mouse serum was injected into 3 weeks fracture muMT mice it caused increased hindpaw von Frey allodynia in the fracture limb, peaking at 7 days and resolving by 14 days after injection (Fig. 8C). Furthermore, when 3 weeks DP serum was injected into 3 weeks fracture muMT mice it had no effect on von Frey allodynia (Fig. 8C). These results support the hypothesis that the pronociceptive systemic autoantibodies and target neoantigens formed after disc injury or tibia fracture are unique to each type of injury and that the pronociceptive effects of serum transfer require the presence of injury induced neoantigens.

Figure 8. Pronociceptive serum effects were observed after intraperitoneal injection of disc puncture wildtype mouse serum into disc puncture muMT mice.

Intraperitoneal injection of serum (0.5 ml) from 3 weeks post DP wildtype male mice (but not from sham operated mice) into 3 weeks post DP muMT male mice caused hindpaw von Frey allodynia that peaked by 7 days and resolved by 14 days after injection (A). Intraperitoneal injection of serum from 3 weeks post DP wildtype mice into 3 weeks post sham operated muMT male mice (A) or into 3 weeks post tibia fracture muMT male mice (C) had no effect on hindpaw von Frey thresholds. Intraperitoneal injection of serum from 3 weeks post tibia fracture wildtype male mice into 3 weeks post DP muMT male mice had no effect, but when 3 weeks fracture wildtype mouse serum was injected into 3 weeks fracture muMT mice it caused increased hindpaw von Frey allodynia in the fracture limb, peaking at 7 days and resolving by 14 days after injection (B). These results support the hypothesis that the pronociceptive systemic autoantibodies and target neoantigens formed after disc injury or tibia fracture are unique to each type of injury. A two-way repeated measures ANOVA was used to test the effects of each treatment group on the dependent variables over time and a Sidak correction test was used to determine post hoc contrasts. Data are expressed as mean values ± SEM, n = 8 per cohort. * P < 0.05, ** P < 0.01, and *** P < 0.001 for differences between the surgery and sham groups (A, D) or for differences between fracture and DP groups (B, C). WT: wildtype mice, muMT: mice lacking mature B cells and antibodies, DP: disc puncture surgery, Sham: sham DP surgery, BL: baseline, 3wk: 3 weeks after DP surgery, IP: intraperitoneal injection

3.7. Intrathecal injection of IgM antibodies from disc puncture wildtype mice had pronociceptive effects in disc puncture muMT mice and disc puncture caused IgM deposition in the lumbar discs and spinal cord of wildtype mice.

Intrathecal injection of 3 weeks post DP wildtype male IgM (5ug/5ul) into 3 weeks post DP muMT male mice caused hindpaw von Frey allodynia that peaked at 1 hour after injection and resolved over 24 hours (Fig.9A). Intrathecal injection of IgM from 3 weeks post sham surgery wildtype mice had no effect on von Frey thresholds and similarly, IgG (10ug/5ul) from 3 weeks post DP wildtype mice had no effect on von Frey thresholds in 3 weeks post DP muMT male mice (Fig. 9A). Intraplantar injection of 3 weeks post DP wildtype IgM (5ug.5ul) into the hindpaw of 3 weeks post DP muMT mice had no effect on hindpaw von Frey thresholds (Supplemental Fig.4). IgM deposition in lumbar discs and spinal cord was observed at 3 weeks post DP surgery in wildtype mice, but not in sham operated mice (Fig 9B). Significantly, there was no IgM deposition in the cervical spinal discs or cord at 3 weeks after lumbar disc puncture, demonstration of a regionally restricted IgM antibody-antigen response to lumbar disc injury (Fig.9B). There was increased IgG deposition in the lumbar discs at 3 weeks post DP surgery in wildtype mice, but only minimal IgG was observed in the lumbar spinal cord at 3 weeks after disc injury (Fig.9C).

Figure 9. Intrathecal injection of IgM antibodies collected from wildtype disc puncture mice were pronociceptive in muMT disc puncture mice and IgM immune complex deposition was observed in lumbar spinal discs and cord of wildtype disc puncture mice.

Intrathecal injection of IgM antibodies (5ug/5ul) from 3 weeks post DP wildtype male mice reduced the hindpaw von Frey thresholds in 3 weeks post DP muMT male mice (A). IgM from sham operated wildtype mice had no effect on von Frey thresholds in the muMT DP mice and IgG (10ug/5ul) from 3 weeks post DP wildtype mice also had no pronociceptive effects in the muMT DP mice (A). IgM (B) and IgG (C) levels in 3 weeks post disc puncture mice were determined by western blot analysis of lumbar and cervical spinal discs and spinal cord tissues. Increased IgM deposition was observed in the lumbar spinal discs and cord, but not in the cervical discs and cord. IgG levels were only elevated in the lumbar spinal discs, not in the lumbar spinal cord or in the cervical discs or cord. Collectively, these results support the hypothesis that IgM (but not IgG) antigen-antibody complexes in the lumbar spinal cord initiate the pronociceptive effects observed with the serum transfer experiments presented in Figure 8. A two-way repeated measures ANOVA was used to test the effects of each treatment group on the dependent variable over time, a one-way ANOVA was used to test for effects between groups at a single time point, and a Sidak test was used to determine post hoc contrasts. Data are expressed as mean values ± SEM, n = 8 per cohort for intrathecal injection group and n = 4 per cohort for western blot analyses. *P< 0.05, **P< 0.01, ***P< 0.001. ****P<0.0001 vs sham. WT: wildtype mice, muMT: mice lacking mature B cells and antibodies, DP: disc puncture surgery, Sham: sham DP surgery, 3wk: 3 weeks after DP surgery, IT: intrathecal injection

3.8. IgM antibodies from disc puncture wildtype mice exhibited enhanced binding to lumbar disc and spinal cord proteins.

Preliminary sera IgM and IgG binding studies were performed using 5 recombinant proteins identified as potential autoantigen candidates from review of the literature. The recombinant protein candidates screened included; 1) aggrecan, 2) collagen II, 3) collagen V), 4) glial fibrillary acidic protein (GFAP), and 5) N-methyl-D-aspartate receptor B2 (NMDAR B2). Actin was used as a negative control. Full-length proteins were utilized whenever commercially available. These potential autoantigens were previously identified in a proteomic survey of the mouse intervertebral disc [40], are also present in spinal cord tissues, and a previous study demonstrated the presence of IgG antibodies in cultured human degenerated disc media that preferentially bound to the extracellular matrix proteins aggrecan, collagen II, and collagen V [6]. Four of the candidate autoantigens (aggrecan, collagen II, GFAP, and NMDAR B2) exhibited increased IgM binding when probed with 3 weeks post DP wildtype mouse sera, relative to 3 weeks post sham surgery wildtype mice (Fig.10A), consistent with the pronociceptive effects of IgM intrathecal injection in muMT DP mice (Fig.9A) and the increase in IgM deposition in the lumbar spinal cord of 3 week post DP mice (Fig. 9B). Only one of the candidate autoantigens (collagen II) exhibited minimally increased IgG binding when probed with 3 weeks post DP wildtype mouse sera (Fig.10B), consistent with the lack of IgG pronociceptive effects after intrathecal injection in DP muMT mice (Fig.9A) and the increase in IgG deposition in the spinal discs but not the spinal cord of 3 week post DP mice (Fig. 9C).

Figure 10. Disc puncture in wildtype mice enhanced IgM binding to spinal disc antigens.

DP and sham operated wildtype mouse sera IgM and IgG binding studies were performed using 5 recombinant proteins identified as potential autoantigen candidates from review of the literature. The recombinant protein candidates screened included; 1) aggrecan, 2) collagen II, 3) collagen V), 4) glial fibrillary acidic protein (GFAP), and 5) N-methyl-D-aspartate receptor B2 (NMDAR B2). Actin was used as a negative control. Four of the candidate autoantigens (aggrecan, collagen II, GFAP, and NMDAR B2) exhibited increased IgM binding when probed with 3 weeks post DP wildtype mouse sera, relative to 3 weeks post sham surgery wildtype mice (A), consistent with the pronociceptive effects of IgM intrathecal injection in muMT DP mice (Fig.9A) and the increase in IgM deposition in the lumbar spinal cord of 3 week post DP mice (Fig. 9B). Only one of the candidate autoantigens (collagen II) exhibited minimally increased IgG binding when probed with 3 weeks post DP wildtype mouse sera (B), consistent with the lack of IgG pronociceptive effects after intrathecal injection in DP muMT mice (Fig.9A) and the increase in IgG deposition in the spinal discs but not the spinal cord of 3 weeks post DP wildtype mice (Fig. 9C). Statistical significance was determined using a Sidak multiple comparison test. Data expressed as mean values ± SEM. ***P<0.001, **P<0.01, and *P<0.05 vs Sham. DP: disc puncture surgery, Sham: sham DP surgery

4. Discussion

Despite extensive investigation, the pathophysiology of chronic low back pain remains undefined in up to 85% of patients [1; 14] and there is continuing controversy regarding clinical management for chronic LBP [1; 7; 41; 57]. Clearly current hypotheses for the progression of tissue injury to chronic painful disability have not generated effective and safe treatments for many LBP patients. Innovative approaches are very much needed and position statements highlight the need for improved understanding of basic pain mechanisms and the development of scientifically rigorous trials of target based treatments[13]. Towards that goal, the principal findings of the current study were: 1) murine lumbar disc injury leads to long-lasting hindpaw mechanical allodynia, pinch hyperalgesia, reduced grip strength upon extension of the spine, and a spontaneous pain state, 2) muMT mice lacking mature B lymphocytes and antibody production failed to develop the full pain-related phenotypic changes observed in the wildtype mice after disc injury, 3) female mice developed the same pain-related phenotypic changes after disc injury as their male counterparts, albeit of shorter duration, 4) expression of spinal cord inflammatory mediators, immune cell markers, and complement factors were all up-regulated after disc injury, 5) disc injury induced spinal glial cell activation and proliferation and also up-regulated complement C5a receptors in dorsal horn neurons, 6) disc puncture mouse serum was pronociceptive when systemically injected into muMT disc puncture mice lacking antibodies, but was not pronociceptive when injected into muMT tibia fracture mice, 7) intrathecal injection of disc puncture mouse IgM, but not IgG, was pronociceptive in muMT disc puncture mice, 8) lumbar disc injury induced IgM deposition in lumbar discs and spinal cord, but not in cervical discs and spinal cord, and 9) potential disc autoantigens exhibited increased IgM (but not IgG) binding when probed with disc puncture mouse sera, relative to sham surgery sera. Collectively, these results further the novel hypothesis that spinal cord pronociceptive autoimmune responses to intervertebral disc injury are important contributors to chronic low back and possibly radicular pain, and establish a murine model in which to study those changes.

It is important to stress that the DP mouse LBP model exhibited chronic changes on multiple pain behavioral assays assessing hindpaw mechanical allodynia, hyperalgesia, grip strength, and spontaneous pain behavior in both male and female mice. Critically, pinch testing thresholds in the forepaws of wildtype mice after DP surgery were unchanged relative to pre-surgical baseline thresholds (Supplemental Fig.1) and tail flick thresholds were also unchanged (Fig.1G), indicating that lumbar DP induced nociceptive sensitization was regionally restricted to the hindlimbs innervated by the lumbar spinal cord. Furthermore, the segmental innervations of the L4/5, L5/6 and L6/S1 intervertebral discs undergoing DP surgery originate in the lumbar spinal cord [16]. Consistent with a postulated spinal pronociceptive autoimmune response segmentally localized to the lumbar cord after disc puncture injury, IgM immune complex deposition was limited to the lumbar cord and not observed in the cervical spinal cord of DP wildtype mice (Fig. 9B).

Serum and antibody transfer experiments in the tibia fracture mouse model of CRPS [19] and the DP mouse model of LBP (Fig. 8A, 9A) have demonstrated the systemic presence of pronociceptive autoantibodies in these post traumatic chronic pain models. The localized formation of neoantigens in a specific spinal cord segment would allow for the development of a spinal pronociceptive autoimmune response capable of mediating chronic regional nociceptive sensitization and pain. It has been proposed that spinal disc degeneration or herniation causes the loss intervertebral disc immune privilege resulting in the generation of autoantibodies targeting self-reactive antigens in the intervertebral disc [54]. Consistent with this hypothesis, we observed that candidate disc autoantigens exhibited increased IgM binding when probed with 3 weeks post-DP wildtype mouse sera, relative to 3 weeks post-sham surgery wildtype mice (Fig.10A). Disc injury results in cellular apoptosis and while immune system activation participates in the repair of damaged discs [10; 54; 55], the caspase-cleaved fragmented proteins generated after cellular death are potent novel antigenic signals to immune cells and have been linked to the development of neoantigens and autoimmunity [18; 35].

Both the tibia fracture and the DP trauma models develop spinal cord pronociceptive autoimmune responses after peripheral tissue injuries, with intrathecally injected IgM having pronociceptive effects (Fig. 9A) [52] and IgM immune complexes forming in the lumbar spinal cord (Fig.9B) [19]. It is still unclear how neoantigens form in the spinal cord after peripheral tissue injuries, but interestingly, fracture mouse sera was pronociceptive in muMT fracture mice, as we have previously observed [19], and not in muMT DP mice (Fig.8B). Contrariwise, DP mouse sera was only pronociceptive in muMT DP mice but not in muMT fracture mice (Fig. 8C), clearly demonstrating that the neoantigens targeted by serum autoantibodies are unique to each trauma model. These results further support the hypothesis that each type of post-traumatic chronic musculoskeletal pain has its own unique autoantibodies and targeted neoantigens.

Our prior investigations utilizing complex regional pain syndrome (CRPS) patient antibodies in the rodent fracture model of CRPS led us to postulate that trauma generated pronociceptive autoantibodies contribute to pain chronification. Limb fracture is the most common cause of CRPS [11] and we previously identified innate and adaptive pronociceptive immune responses in the rodent tibia fracture model closely resembling those observed in CRPS patients [3]. Recently we observed that tibia fracture injury in mice induced lymphadenopathy and a germinal center reaction in the ipsilateral popliteal lymph node, thus creating plasma cells expressing pronociceptive IgM antibodies [33; 34]. When serum IgM (but not IgG) autoantibodies from wildtype fracture mice were injected systemically into muMT fracture mice lacking mature B cells and immunoglobulin, they induced a gradually increasing hindpaw allodynia and unweighting, peaking at 7 days and resolving by 14 days after injection [19], similar to the time course we observed after intraperitoneal injection wildtype DP mouse serum into muMT DP mice (Fig.9A).

Remarkably, the pronociceptive effects of wildtype mouse fracture serum or IgM injections into muMT fracture mice were replicated when CRPS patient serum or IgM were injected into muMT fracture mice, but normal subject serum and CRPS patient IgG antibodies had no effect[21]. Furthermore, CRPS patient IgM antibodies had unilateral pronociceptive effects when injected intraplantarly into the fracture limb skin or intrathecally in the muMT fracture mice, indicating a regionally limited unilateral expression of neoantigens targeted by pronociceptive autoantibodies.

Antibody effects are primarily mediated by complement activation. C1 complement (which lacks Fc receptors) can bind to Fc fragments of antibody-antigen immune complexes to activate the complement pathway, resulting in the formation C5a complement proteins that bind to their cognate receptors (C5aR) on dermal macrophages and spinal microglia to induce the production and secretion of interleukin 1 (IL-1), interleukin 6 (IL-6), tumor necrosis factor (TNF), and nerve growth factor (NGF)[47; 53; 59; 60]. Systemic treatment with a C5aR antagonist in wildtype fracture mice transiently reversed post fracture hindpaw allodynia and unweighting and C5aR deficient fracture mice had attenuated hindpaw nociceptive sensitization [52]. Wildtype fracture mice also developed post fracture increases in C5a complement and C5aR expression in skin and cord, predominantly on dermal macrophages and spinal cord microglia, whereas in the DP injury model we observed C5aR immunostaining primarily on dorsal horn neurons (Fig.7). Intraplantar or intrathecal injections of CRPS patient IgM induced ipsilateral hindpaw nociceptive sensitization within an hour after injection in muMT fracture mice and induced increased expression of C1q complement and inflammatory cytokines. Intrathecal (but not intraplantar) injection of IgM from DP wildtype mice into DP muMT mice also induced hindpaw nociceptive sensitization within an hour after injection (Fig.9A, Supplemental Fig.4).

Based on these data we postulate that tissue trauma can induce the localized expression of novel neoantigens in the injured tissue and innervating spinal cord, with the resultant localized formation of T follicular helper cell dependent germinal center responses that generate plasma cells secreting circulating IgM autoantibodies that bind to their target antigens and form IgM immune complexes in the injured tissue and corresponding spinal cord. After fracture IgM immune complexes initiate C5a signaling in macrophages and microglia, resulting in the localized release of soluble pronociceptive inflammatory mediators with nociceptive sensitization and pain, but in the DP model it appears that C5aRs are primarily expressed by dorsal horn neurons and not by glial cells (Figs. 5–7). Previously we observed that C5a signaling directly activates nociceptive sensory neurons in vitro, inducing spontaneous C-fiber activity and sensitization [23], and we postulate that pronociceptive autoantibody effects in the DP injury model may be mediated by C5a signaling in dorsal horn neurons (Fig.7).

The pronociceptive autoimmune hypothesis for LBP is consistent with several recent previous reports and our emerging understanding of autoimmune contributions to chronic pain in localized musculoskeletal[21; 34; 52], neurologic[12; 29; 48], and widespread pain conditions [17; 58]. The results of the current study further support the pronociceptive autoimmunity hypothesis for the transition from tissue injury to chronic pain state and provide a foundation for future translational trials using LBP patient antibodies in the DP mouse model and clinical trials in LBP patients using FDA approved agents currently being used to treat other autoimmune diseases.