Substance P spinal signaling induces glial activation and nociceptive sensitization after fracture

Abstract

Tibia fracture in rodents induces substance P (SP)-dependent keratinocyte activation and inflammatory changes in the hindlimb, similar to those seen in complex regional pain syndrome (CRPS). In animal pain models spinal glial cell activation results in nociceptive sensitization. This study tested the hypothesis that limb fracture triggers afferent C-fiber SP release in the dorsal horn, resulting in chronic glia activation and central sensitization. At 4 weeks after tibia fracture and casting in rats, the cast was removed and hind paw allodynia, unweighting, warmth, and edema were measured, then the antinociceptive effects of microglia (minocycline) or astrocyte (LAA) inhibitors or an SP receptor antagonist (LY303870) were tested. Immunohistochemistry and PCR were used to evaluate microglia and astrocyte activation in the dorsal horn. Similar experiments were performed in intact rats after brief sciatic nerve electric stimulation at C-fiber intensity. Microglia and astrocytes were chronically activated at 4 weeks after fracture and contributed to the maintenance of hind paw allodynia and unweighting. Furthermore, LY303870 treatment initiated at 4 weeks after fracture partially reversed both spinal glial activation and nociceptive sensitization. Similarly, persistent spinal microglial activation and hind paw nociceptive sensitization were observed at 48 hours after sciatic nerve C-fiber stimulation and this effect was inhibited by treatment with minocycline, LAA, or LY303870. These data support the hypothesis that C-fiber afferent SP signaling chronically supports spinal neuroglia activation after limb fracture and that glial activation contributes to the maintenance of central nociceptive sensitization in CRPS. Treatments inhibiting glial activation and spinal inflammation may be therapeutic for CRPS.

1. Introduction

Distal limb fracture is the most common cause of complex regional pain syndrome (CRPS)(Sandroni et al. 2003, de Mos et al. 2007) and we have developed a rodent fracture model that closely resembles CRPS. Tibia fractured rats treated with 4-weeks cast immobilization exhibit hindpaw allodynia, unweighting, warmth, edema, spontaneous protein extravasation, increased spinal Fos-immunoreactivity, facilitated substance P (SP) and calcitonin gene-related peptide (CGRP) neuropeptide signaling in the injured limb, and SP dependent keratinocyte proliferation and epidermal thickening, with increased keratinocyte expression of tumor necrosis factor-α (TNF), interleukin-1β (IL-1), interleukin-6 (IL-6), chemokine (C-C) motif ligand 2 (CCL2), and nerve growth factor (NGF) inflammatory mediators in the skin of the fractured limb (Guo et al. 2004, Guo et al. 2006, Sabsovich et al. 2008, Sabsovich et al. 2008, Li et al. 2009, Wei et al. 2009, Wei et al. 2009, Li et al. 2010).

Recently we observed increased mRNA and protein levels of these same inflammatory mediators (TNF, IL-1, IL-6, CCL2, and NGF) in the lumbar spinal cords of tibia fracture rats at 4 weeks postfracture (Shi et al. 2015). Fracture mice exhibited similar increases in spinal inflammatory mediatory expression and up regulated spinal expression of SP and the SP neurokinin 1 (NK1) receptor. Fracture-induced increases in spinal inflammatory mediators were attenuated in transgenic fracture mice lacking substance P (SP) and these mice had reduced postfracture nociceptive sensitization. Intrathecal injection of selective receptor antagonists or inhibitors for SP, TNF, IL-1, IL-6, CCL2, or NGF each reduced pain behaviors in the fracture rats. Collectively, these data support the hypothesis that facilitated spinal SP signaling up regulates the expression of spinal inflammatory mediators contributing to nociceptive sensitization in the rodent fracture CRPS model. Interestingly, several studies have reported elevated cerebrospinal fluid levels of IL-1 and IL-6 in chronic CRPS patients (Alexander et al. 2005, Alexander et al. 2007), but other investigators have been unable to confirm these findings (Munts et al. 2008). Elevated cerebrospinal fluid levels of cytokines and neurotrophic factors have also been observed in other chronic pain states (Bjurstrom et al. 2014).

CRPS patients frequently exhibit widespread regional pain and allodynia extending beyond the initial injury site. Furthermore, the intensity of pain in the first week after wrist fracture is the best predictor of the development of CRPS in the ensuing 4 months (Moseley et al. 2014). Intense and sustained peripheral sensory nociceptive afferent activation can induce increased excitability in the spinal cord neurons, leading to the spread of pain sensitivity beyond the site of injury (secondary hyperalgesia) and the generation of pain in response to low threshold stimuli (allodynia). These changes are commonly referred to as central sensitization (Woolf and Costigan 1999, Woolf and Salter 2000). It is widely believed that spinal microglia and astrocytes are key contributors to the development of central nociceptive sensitization in a variety of rodent pain models, including neuropathic, inflammatory, and bone cancer (Watkins et al. 2001, 2003, Ren and Dubner 2010, Ji et al. 2013, Mika et al. 2013). These spinal immunocytes, when activated after injury, release proinflammatory cytokines and growth factors capable of enhancing spinal neuronal excitability. Reciprocal interactions between neurons and glia are postulated to play a crucial role in the perpetuation of injury-induced pain.

The current study tested the hypothesis that post-fracture pain is mediated by ongoing C-fiber nociceptor discharges triggering SP release from their central afferent terminals in the lumbar dorsal horn, with subsequent glial activation, central nociceptive sensitization, allodynia and unweighting.

2. Experimental Procedures

These experiments were approved by the Veterans Affairs Palo Alto Health Care System Institutional Animal Care and Use Committee (Palo Alto, CA) and followed the animal subjects guidelines of the National Institute of Health Guide for the Care and Use of Laboratory Animal. One hundred forty five adult (9-month-old) male Sprague-Dawley rats (Simonsen Laboratories, Gilroy, CA) were used in these experiments. The animals were housed individually in 30 × 30 × 19 cm isolator cages with solid floors covered with 3 cm of wood chip bedding with a Bio-Serv gummy bone for enrichment, and were given food and water ad libitum. During the experimental period the animals were fed Lab Diet 5012 (PMI Nutrition Institute, Richmond, IN), which contains 1.0% calcium, 0.5% phosphorus, and 3.3 IU/g vitamin D3 and were kept under standard conditions with a 12-h light-dark cycle (6 am – 6 pm). All in vivo experiments were performed between 10 am and 4 pm in the Veterinary Medical Unit of the Veterans Affairs Palo Alto Health Care System.

2.1 Surgery

Tibia fracture was performed under 2–4% isoflurane to maintain surgical anesthesia as we have previously described (Guo et al. 2014). The right hind limb was wrapped in a stockinet (2.5cm wide) and the distal tibia was fractured using pliers with an adjustable stop (Visegrip, Newell Rubbermaid, Atlanta, GA) that had been modified with a three-point jaw. The hind limb was wrapped in casting tape (Delta-Lite, Johnson & Johnson, New Brunswick, NJ) so the hip, knee, and ankle were flexed. The cast extended from the metatarsals of the hind paw up to a spica formed around the abdomen. The cast over the paw was applied only to the plantar surface; a window was left open over the dorsum of the paw and ankle to prevent constriction when post-fracture edema developed. To prevent the animals from chewing at their casts, the cast material was wrapped in galvanized wire mesh. The rats were given subcutaneous saline and buprenorphine immediately after procedure (0.03 mg/kg) and on the first day after fracture for postoperative hydration and analgesia. At 4 weeks the rats were anesthetized with isoflurane and the cast removed with a vibrating cast saw. All rats used in this study had union at the fracture site after 4 weeks of cast immobilization.

2.2 Hindpaw nociception, temperature and edema measurements

All hindpaw behavior testing was performed in the following sequential order; 1) von Frey allodynia, 2) unweighting, 3) temperature, and 4) paw thickness. To measure hindpaw plantar allodynia in the fracture rats an up-down von Frey testing paradigm(Chaplan et al. 1994) was used as we have previously described (Guo et al. 2004, Guo et al. 2006).

An incapacitance device (IITC Inc. Life Science, Woodland Hills, CA) was used to measure hind paw unweighting. The rats were manually held in a vertical position over the apparatus with the hind paws resting on separate metal scale plates, and the entire weight of the rat was supported on the hind paws. The duration of each measurement was 6s, and 10 consecutive measurements were taken at 60s intervals. Eight readings (excluding the highest and lowest) were averaged to calculate each hind paw’s weight bearing mean value (Guo et al. 2006, Sabsovich et al. 2008).

The temperature of the hindpaw was measured using a fine wire thermocouple (Omega Engineering, Stamford, CT) applied as previously described (Guo et al. 2004, Guo et al. 2006). Temperature testing was performed over the hindpaw dorsal skin between the first and second metatarsals (medial), the second and third metatarsals (central), and the fourth and fifth metatarsals (lateral). The measurements for each hindpaw were averaged for the mean paw temperature.

Hindpaw edema was determined by measuring the hindpaw dorsal-ventral thickness over the midpoint of the third metatarsal with a LIMAB laser measurement sensor (Goteborg, Sweden) while the rat was briefly anesthetized with isoflurane (Guo et al. 2004, Guo et al. 2006).

Hindpaw von Frey thresholds, temperature, and thickness data were analyzed as the difference between the fracture side and the contralateral untreated side. Right (fracture side) hindpaw weight bearing data were analyzed as a ratio between the average right hindpaw weight bearing value and the mean of right and left hindpaws’ values ((2R/(R + L)) · 100%).

2.3 Drugs

Minocycline (Sigma, St Louis, MO) is a semisynthetic tetracycline derivative with multiple anti-inflammatory effects independent of its antimicrobial activity. Previous studies demonstrated that systemic application of minocycline (40 mg/kg/day, i.p.) for 3 days profoundly suppressed spinal microglial activation and attenuated neuropathic pain in several nerve-injury models (Raghavendra et al. 2003, Ledeboer et al. 2005, Piao et al. 2006, Guasti et al. 2009).

L-2-aminoadipic acid (LAA, Sigma, St Louis, MO) is a selective astrocytic toxin (Khurgel et al. 1996). Previous studies have demonstrated that intrathecal application of LAA (50-150 nmol) inhibited spinal astrocyte activation and exerted analgesic effects on spinal nerve ligation induced neuropathic pain (Zhuang et al. 2006, Mei et al. 2011).

The SP NK1 receptor antagonist LY303870 (LY) was a generous gift from Dr. L. Phebus (Eli Lily Co., Indianapolis, IN, USA). This compound has nanomolar affinity for the rat NK₁ receptor, has no affinity for 65 other receptors and ion channels, has no sedative, cardiovascular, or core body temperature effects in rats at systemic doses up to 30 mg/kg, and is physiologically active for 24 h after a single systemic dose of 10 mg/kg (Gitter et al. 1995, Hipskind et al. 1996, Iyengar et al. 1997).

2.4 Intrathecal injection

Under light isoflurane anesthesia (1-2%), a 25-gauge needle was inserted between the L5 and L6 spinal vertebra. When a brisk tail flick was observed, confirming that the needle had entered the subarachnoid space, 10 μl of solution was injected. Intrathecal injection of the 0.9% saline vehicle had no effect on hindpaw von Frey allodynia or unweighting in the fracture rats.

2.5 Sciatic nerve electrical stimulation

The aims of this experiment were to examine whether electrical stimulation of sciatic nerve could induce spinal glia activation and to determine the nature of the afferent signals triggering this activation. Electrical stimulation was performed as described by Hathaway et al (Hathway et al. 2009). Briefly, rats were anaesthetized under 2–4% isoflurane and body temperature maintained using a homeothermic-heating blanket (Harvard Apparatus, Holliston, MA). Under sterile conditions, the right sciatic nerve was exposed via an incision through the thigh, and completely isolated from the surrounding connective tissue. A plexiglas-platinum stimulating electrode was gently secured around the nerve (Harvard Apparatus, Holliston, MA). Care was taken to ensure that the electrodes were only in contact with the nerve and that the sciatic nerve was never stretched. Trains of electrical stimuli were delivered to the sciatic nerve for 5 min using either C-fiber stimulation parameters (10mA, 5Hz, 0.5 msec) or Aδ/Aβ-fiber stimulation parameters (5 mA, 10 Hz, 0.15 msec). A sham control group that underwent surgery and electrode placement without electrical stimulation was also employed. Following stimulation the electrodes were carefully removed, the muscle and skin sutured shut using 5/0 Mersilk, and the animals returned to their home cage to recover, with free access to food and water. Nociceptive testing was performed at baseline and at 3, 24, and 48 h after electrical stimulation.

2.6 Experimental designs

The first experiment tested the hypothesis that spinal microglia activation contributes to the nociceptive changes observed in the fracture model. Rats underwent distal tibia fracture and after 4 weeks cast immobilization the cast was removed and the following day the rats underwent baseline behavioral testing (hindpaw von Frey thresholds, weight bearing, skin temperature, and paw thickness). After baseline testing, the microglia inhibitor minocycline (40 mg/kg/day/i.p.) or vehicle were administered for 3 days (n = 8 per cohort), after which the behavioral testing was repeated and the rats were then euthanized for spinal cord tissue collection. The minocycline dose was chosen on the basis of previously published studies and our preliminary dose–response studies demonstrating analgesic effects in fracture rats. An additional study used a one time intrathecal injection of minocycline (100μg/10μl/i.t.) in 4-week fracture rats and the hindpaw von Frey thresholds and unweighting were tested at baseline and at 1, 3, 6, and 24 h after injection.

The second experiment tested the hypothesis that astrocyte activation contributes to the nociceptive changes observed in the rat fracture model. At 4 weeks post-fracture the cast was removed and the following day baseline behavioral testing was performed. After baseline behavioral testing, the astrocyte inhibitor L-2-aminoadipic acid (150nmol/10μl/i.t., LAA) or vehicle was intrathecally injected and behavioral testing was repeated at 3, 6, and 24 h after the injection (n = 8 per cohort). The rats were then euthanized for spinal cord tissue collection. The LAA dose was chosen on the basis of previous published studies (Zhuang et al. 2006, Mei et al. 2011) and our preliminary dose-response studies.

The next experiment determined whether SP signaling contributes to spinal glial cell activation in the fracture model. The 4 weeks postfracture rats underwent baseline behavioral testing and then were treated with the SP NK1 receptor antagonist LY303870 (30 mg/kg/day/i.p.) or vehicle for 3 days, after which the behavioral testing was repeated and the rats were then euthanized for spinal cord tissue collection (n = 8 per cohort). The dose was chosen on the basis of our previous experiments demonstrating analgesic and anti-inflammatory effects in fracture rats.(Wei et al. 2012)

Another experiment tested the hypothesis that peripheral C-fiber nociceptor excitation can induce the development of CRPS-like nociceptive and vascular changes similar to those seen after tibia fracture. After baseline behavioral testing (hindpaw von Frey allodynia, unweighting, warmth, and edema) intact (no fracture) rats were anesthetized with isoflurane (2-4%) and under sterile conditions, the right sciatic nerve was exposed via an incision through the thigh, and completely isolated from the surrounding connective tissue. A plexiglas-platinum stimulating electrode was gently secured around the nerve and 5 min of sciatic nerve electrical stimulation at C-fiber intensity (10mA, 5Hz, 0.5msec) (Hathway et al. 2009) was applied, then the stimulator was removed and the wound was closed. Behavioral testing was repeated at 3, 24, and 48 h after electrical stimulation, then the animals were euthanized and spinal cord tissues were collected. Sham stimulated control rats also were anesthetized with isoflurane and the right sciatic nerve was exposed via an incision and the stimulating electrodes were placed around the sciatic nerve, but no stimulation was performed.

Additional control experiments were performed to confirm that C-fiber activation mediates hindpaw nociceptive sensitization. In one control experiment baseline behavioral testing was performed, then under isoflurane anesthesia and sterile conditions the right sciatic nerve was exposed via an incision through the thigh and a plexiglas-platinum stimulating electrode was secured around the nerve. Bupivacaine (0.5%, Sigma) soaked cotton pellets were placed around the sciatic nerve 1 cm proximal to the electrical stimulation site. Parafilm was placed over the surrounding musculature to prevent bupivacaine absorption. After 10 min of bupivacaine exposure the hindpaw toes were pinched to confirm anesthetic block, then the pellets were removed and the sciatic nerve was stimulated for 5 minutes at C-fiber intensity, then the incision was closed and behavioral testing was repeated at 3, 24, and 48 h after electrical stimulation. Another control experiment tested the effects of 5 min of sciatic stimulation at Aδ/Aβ-fiber strength (5mA, 10Hz, 0.15msec) (Hathway et al. 2009) with behavioral testing at baseline and 3, 24, and 48 h after electrical stimulation.

Additional experiments were performed to confirm that sciatic electrical stimulation evoked nociceptive behaviors are dependent on glial cell activation and SP signaling. One hour prior to electrical stimulation the rats were pretreated with vehicle or one of the following drugs: the microglia inhibitor minocycline (40mg/kg/i.p.), the astrocyte toxin LAA (150nmol/10ul/i.t.), or the SP NK1 receptor antagonist LY303870 (30mg/kg/i.p.), and at 48 h after electrical stimulation the rats underwent behavioral testing (hindpaw von Frey allodynia, unweighting, warmth and edema) and then were euthanized and spinal cord tissue was collected (n = 8 per cohort). Minocycline was administered 3 times (1 h prior and 23 and 47 h after electrical stimulation) and the LAA and LY303870 were administered only once at 1 h prior to electrical stimulation.

2.7 Quantitative real time PCR

These experiments tested the hypothesis that tibia fracture up regulates the expression of biochemical markers of glial cells in the lumbar spinal cord. Total RNA from normal and fracture rat spinal cord was extracted using the RNeasy Mini Kit (Qiagen, Venlo, Netherlands) and the purity and concentration were determined spectrophotometrically. Next cDNA was synthesized from 1 μg RNA using an iScript cDNA Synthesis Kit (Bio-Rad Laboratories, Hercules, CA). Real-time polymerase chain PCR reactions (PCRs) were conducted using the SYBR Green PCR master mix (Applied Biosystems, Foster City, CA). Real-time PCR amplification of chemokine (C-C motif) ligand 2 (CCL2), matrix metallopeptidase 3(MMP3), CD68, CD163, interleukin 6 (IL-6), colony stimulating factor 1(CSF1), colony stimulating factor 1 receptor (CSF1R), glial fibrillary acidic protein (GFAP), nestin, nuclear receptor co-repressor 1(NcoR), ciliary neurotrophic factor receptor (CNTFR), SRY (sex determining region Y)-box 2(SOX2), vimentin (VIM), epidermal growth factor receptor (EGFR), tumor necrosis factor α (TNF), interleukin 1β (IL-1), interleukin 6 (IL-6), nerve growth factor (NGF), substance P (TAC1), SP neurokinin1 receptor (TACR1), alpha-calcitonin gene-related peptide (CALCA), beta-calcitonin gene-related peptide (CALCB), calcitonin gene-related peptide type 1 receptor (CALCRL), receptor activity modifying protein 1, the CGRP receptor chaperone gene (RAMP1), activating transcription factor 3(ATF3), and 18S was performed on an ABI 7900HT sequencing detection system (Applied Biosystems, Waltham, MA). To validate the primer sets used, we performed dissociation curves to document single product formation, and agarose gel analysis was conducted to confirm the size (Table 1). The data from real time PCR experiments were analyzed by the comparative CT method as described in the manual for the ABI prism 700 real time systems. All results were confirmed by repeating the experiment 3 times.

Table 1. Primers used for real-time PCR.

Gene	GenBank Accession #	Forward primer	Reverse primer	Product Size(bp)
TNF-α	NM_012675	ctcccagaaaagcaagcaac	cgagcaggaatgagaagagg	210
IL-iβ	NM_031512	agtctgcacagttccccaac	agacctgacttggcagagga	230
IL-6	NM_012589	cacaagtccggagaggagac	acagtgcatcatcgctgttc	168
NGF	XM_227525	acctcttcggacactctgga	gtccgtggctgtggtcttat	168
TAC1	NM_012666	tttgcagaggaaatcggtgccaac	ggcattgcctccttgatttggtca	83
TACR1	NM_012667	ctggaaagaggagccttgtg	ctgagacggaaaggaacagc	205
CALCA	NM_017338	agaagagatcctgcaacactgcca	ggcacaaagttgtccttcaccaca	94
CALCB	NM_138513	cccagaagagatcctgcaac	agttcctcagacccgaaggt	158
CALCRL	NM_012717	tcattgtggtggctgtgttt	aatgggaccatggatgatgt	176
RAMP1	NM_031645	ggcaaacaagattggctgtt	aatggggagcacaatgaaag	154
ATF3	NM_012912	ccctcctagggaagatggag	ctgatgaaactcccggaaaa	194
CCL2	NM_031530	ccagaaaccagccaactctc	ccgactcattgggatcatct	192
CD68	NM_001031638	caagcagcacagtggacatt	ggcagcaagagagattggtc	218
CD163	NM_001107887	tgtggcgtggctattaatgc	acatgaactccgagcagaca	184
MMP3	NM_133523	tcagcggatcttcacagttg	tgaggttgactggtgccata	229
CSF1	NM_023981	gaccctcgagtcaacagagc	tgtcagtctctgcctggatg	221
CSF1R	NM_001029901	gcccagagctggttgtagag	cgcatagggtcttcaagctc	196
VIM	NM_031140	agatcgatgtggacgtttcc	cacctgtctccggtattcgt	206
NOCR	NM_001271103	tgagcgtgaaaggatcactg	tggcatgattcgtagctgag	220
CNTFR	NM_001003929	aacgagatggctgcttctgt	tttaccctccaggtcacagc	185
EGFR	NM_031507	accgtggagagaatcccttt	ttgttgctaaatcgcacagc	176
GFAP	NM_017009	gaagaaaaccgcatcaccat	gcacacctcacatcacatcc	190
NES	NM_012987	aaccacaggagtgggaactg	tctggcattgactgagcaac	219
SOX2	NM_001109181	aagggttcttgctgggtttt	gccctaaacaagaccacgaa	160
18S	X01117	cgcggttctattttgttggt	agtcggcatcgtttatggtc	219

2.8 Immunohistochemistry, confocal microscopy and quantification

Animals were euthanized and transcardially perfused with phosphate buffered saline (PBS) followed by 4% paraformaldehyde (PFA) in PBS, pH 7.4, via the ascending aorta; the lumbar enlargement of the spinal cord was removed, post-fixed in 4% PFA overnight, and then the tissues were cryoprotected with 30% sucrose in PBS at 4°C before embedding in TissueTek optimal cutting temperature compound (Sakura Finetek USA). Following embedding, 20-μm thick serial frozen sections were cut on a coronal plane using a cryostat, mounted onto Superfrost microscope slides (Fisher Scientific), and stored at −80°C. A microglia specific marker, ionized calcium-binding adapter molecule 1 (Iba-1), or an astrocyte specific marker, glial fibrillary acidic protein (GFAP), were used to examine spinal microglial or astrocyte activation and minocycline/LLA/LY303870 effects at 4 weeks post-fracture or after electrical nerve stimulation. Frozen spinal cord sections were permeabilized and blocked with PBS containing 10% donkey serum and 0.3% Triton X-100, followed by exposure to a polyclonal rabbit anti-Iba1 primary antibody (Wako Chemicals, Richmond, VA, diluted 1:1000) overnight at 4 °C in PBS containing 2% serum. Sections were then rinsed in PBS and incubated with donkey anti-rabbit IgG (1:500) conjugated with cyanine dye 3 secondary antibody (Jackson ImmunoResearch Laboratories, Westgrove, PA) for 1h at room temperature. After three washes, the sections were mounted with anti-fade mounting medium (Invitrogen, Carlsbad, CA). For GFAP immunostaining, frozen spinal cord sections were permeabilized and incubated with a mouse anti-GFAP conjugated with cyanine dye 3 (EMD Millipore, Billerica, MA, diluted 1:1000). Images were obtained using LSM/710 confocal microscope (Carl Zeiss, Jena, Germany) and stored on digital media. Images were quantified for fluorescent intensity using ImageJ software (National Institutes of Health) by a blinded investigator, who quantified 4 to 8 sections of each spinal cord, and 4 to 5 high power (400× magnification) fields in the superficial layer of each section to derive a mean score for that spinal cord. The total area quantified in the dorsal horn of each section was 0.18-0.22 mm². The individual mean scores were then used to calculate the mean intensity values and standard error of the mean for the intact control (n=6), fracture (n=7), fracture treated with minocycline (n=6), fracture treated with LAA (n=7), fracture treated with LY (n=7), sham electrical nerve stimulation (n=8), C-fiber electrical stimulation (n= 7), C-fiber stimulation with minocycline (n=5), and C-fiber stimulation with LY (n=5). The immunohistochemistry data quantified in the figures is presented as mean intensity per high powered field. Control experiments included incubation of slices in primary and secondary antibody-free solutions, which exhibited low intensity non-specific staining patterns in preliminary experiments (data not shown).

2.9 Statistical analysis

Sample sizes were based on a power analysis of preliminary and previously published data generated from using each of the proposed assays in fracture animals. Based on this analysis we calculated that the proposed experiments would require 8 animals per cohort to provide 80% power to detect 25% differences between groups. Animals were randomized to experimental groups using computer generated random numbers and all testing was performed in a blinded fashion when possible. There were no adverse events and no animals were excluded after enrollment into the experimental cohorts. Normal distribution of the data was confirmed using the D’Agostino-Pearson omnibus normality test.

Statistical analysis was accomplished using a one-way analysis of variance (ANOVA) with post-hoc Bonferonni correction test to evaluate changes in biochemical markers (i.e., CCL2, MMP3, CD68, CD163, IL-6, CSF1, CSF1R, GFAP, nestin, NcoR, CNTFR, SOX2, vimentin, EGFR) of glial activation after fracture (Fig. 1). A one-way ANOVA followed by post hoc Bonferroni test was performed for the effects of drugs (i.e., minocycline, LY, and LAA) on von-Frey mechanical allodynia, weight-bearing, warmth, edema, and glial activation after fracture (Figs. 2-5) or after electrical stimulation (Figs. 7-9). Behavioral data collected over time after electrical nerve stimulation with and without drug application (i.e., bupivacaine, minocycline, LAA, and LY) were analyzed by a two-way ANOVA on data for each test time point, comparing with sham or vehicle treatments (Fig. 6). For simple comparison of 2 means, unpaired Student t-tests were performed. All 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 6, GraphPad Software, La Jolla, CA, USA).

Figure 1.

Real-time PCR measurement of changes in postfracture gene expression by spinal microglia and astrocytes. The top row of graphs illustrate up regulation of the microglia marker genes CCL2, MMP3, CD163, IL-6, CD68, and CSF1R at 1 week after fracture (A) and increased mRNA levels of CCL2, MMP3, and IL-6 at 4 weeks postfracture (B). The bottom row of graphs demonstrate increased levels for the astrocyte marker genes VIM and GFAP at 1 week after fracture (C) and no increase in mRNA levels of astrocyte activation genes at 4 weeks postfracture (D). Data were analyzed using a one-way analysis of variance (ANOVA) with Bonferonni correction test for post-hoc contrasts, error bars indicate SEM (n=6 per cohort). *p <0.05, **p <0.01, ***p < 0.001 vs. control values. FX: fracture. CCL2: chemokine (C-C motif) ligand 2, MMP3: matrix metallopeptidase 3, IL-6: interleukin 6, CSF1R: colony stimulating factor 1 receptor, CSF1: colony stimulating factor 1, VIM: vimentin, GFAP: glial fibrillary acidic protein, CNTFR: ciliary neurotrophic factor receptor, NCOR: nuclear receptor co-repressor 1, EGFR: epidermal growth factor receptor, SOX2: SRY (sex determining region Y)-box 2, NES: nestin.

Figure 2.

Treatment with the microglia inhibitor minocycline attenuated fracture-induced nociceptive and vascular changes. Rats underwent tibia fracture (FX) and were casted for 4 week, then the cast removed and the next day they were treated with minocycline, either systemically (40mg/kg/day, i.p. for 3 days, Figs. A-D) or intrathecally (100μg in 10μl saline, IT, E-F), intact controls treated with vehicle served as controls. Hindpaw nociceptive testing and assessment of warmth and edema were performed prior to drug administration, after 3 days of systemic minocycline treatment, or 1, 3, 6, and 24 h after intrathecal injection. Measurements presented in panels A, C, D, and E represent the difference between the fracture side and the contralateral paw. Thus, a negative value represents a decrease in nociceptive thresholds on the fracture side, and a positive value represents an increase in temperature or thickness on the fracture side. The data presented in panels B and F represent weight bearing on the fracture hind limb, with 100% representing an even distribution of weight between the fracture and contralateral hind paws, a percentage lower than 100% represents hindpaw unweighting. Systemic minocycline treatment partially reversed the hindpaw allodynia (A), unweighting (B), and warmth (C) that developed after fracture but had no effect on hindpaw edema (D). Intrathecal minocycline treatment also reduced postfracture von Frey mechanical allodynia (E) and unweighting (F), with analgesia lasting between 1 and 6 h after drug administration. Data were analyzed using a one-way analysis of variance (ANOVA) with Bonferonni correction test for post-hoc contrasts, error bars indicate SEM. (n=8 per cohort). * P < 0.05, **P <0.01 vs. control, # P< 0.05, ## P < 0.01 vs. FX-BL. FX: fracture, BL: baseline, Mino: minocycline.

Figure 5.

Treatment with a substance P NK1 receptor antagonist (LY303870) alleviated fracture-induced nociceptive and vascular changes, and reversed spinal microglia and astrocyte activation. Rats underwent tibia fracture (FX) and were casted for 4 week, then the cast removed and the next day they were treated with LY303870 (LY, 30mg/kg/day, IP for 3 days). Hindpaw nociceptive testing and assessment of warmth and edema were performed prior to fracture, at 4 weeks postfracture upon cast removal, and after 3 days of LY303870 administration. The animals were euthanized after the last behavioral tests and the spinal cords were collected for Iba-1 (microglia marker), and GFAP (astrocyte marker) immunostaining. LY303870 treatment reversed fracture-induced allodynia (A), unweighting (B), and warmth (C) (n=8 per cohort). Panels E and G represent fluorescence photomicrographs of Iba-1 (red) and GFAP (red) proteins in the spinal dorsal horn at L5 segment from a nonfractured control, a fractured rat ipsilateral spinal cord, and a fractured rat ipsilateral cord treated with LY303870. Scale bars = 25μm. The quantification of Iba-1 (F) and GFAP (H) immunostaining demonstrated that LY303870 inhibited fracture-induced microglia and astrocyte activation (n = 6-7 per cohort). Data were analyzed using a one-way analysis of variance (ANOVA) with Bonferonni correction test for post-hoc contrasts, error bars indicate SEM. Immunostaining data is presented as mean fluorescence intensity per high powered field. * P < 0.05, **P <0.01 vs. control, # P< 0.05, ## P < 0.01, ### P < 0.001 vs. FX-BL. FX: fracture, BL: baseline, LY: LY303870.

Figure 7.

Treatment with the microglia inhibitor minocycline reversed C-fiber stimulation-induced nociceptive sensitization and spinal microglia activation. Rats underwent surgery and electrode placement without (Sham) or with 5 minutes of sciatic nerve electrical stimulation at C-fiber intensity, and one group of rats was pretreated with minocycline (C-fiber+Mino, 40 mg/kg, i.p. given 1 hour before C-fiber stimulation and at 23 h and 47 h after stimulation). Sciatic C-fiber stimulation induced allodynia (A) and unweighting (B) at 48 h after stimulation, and treatment with minocycline blocked this effect. (C) At 48 h after sciatic stimulation there was an increase in immunostaining for Iba-1 (a microglia marker) and this effect was blocked by minocycline treatment (n = 6-7 per cohort). The mRNA levels of several microglia activation markers (IL-6, CCL2) were significantly elevated in the lumbar cord at 3 h after sciatic stimulation (D), but by 48 h the mRNA levels of all the microglia markers had returned to normal (E). (F) Representative photomicrographs of Iba-1 staining (red) in the spinal dorsal horn at the L5 segment from a sham stimulated control rat (left), a sciatic C-fiber stimulated rat (middle), and a C-fiber stimulated rat treated with minocycline (right). Bottom panels show enlarged view of the boxed regions. These images demonstrate that sciatic nerve stimulation at C-fiber intensity induced microglia activation, and treatment with minocycline blocked this response. Scale bar in D = 100 μm in top panels, 25 μm in bottom panels. Data were analyzed using a one-way analysis of variance (ANOVA) with Bonferonni correction test for post-hoc contrasts, error bars indicate SEM. *p <0.05, **p <0.01, ***p <0.001 vs. Sham stimulation, #p <0.05, ##p <0.01, ###p <0.001 vs. C-fiber stimulation. Immunostaining data is presented as mean fluorescence intensity per high powered field. C-fiber: electrical stimulation at C-fiber intensity, Mino: minocycline, IL-6: interleukin 6, CCL2: chemokine (C-C motif) ligand 2, MMP3: matrix metallopeptidase 3, CSF1R: colony stimulating factor 1 receptor, CSF1: colony stimulating factor 1.

Figure 9.

The SP NK1 receptor antagonist LY303870 blocked C-fiber stimulation-induced nociceptive sensitization, spinal microglia activation, spinal inflammatory mediator expression, and up-regulated spinal neuropeptide and cognate receptor expression. Rats underwent surgery and electrode placement without (sham) and with electrical stimulation of the sciatic nerve at C-fiber strength for 5 min, and one group of rats was treated with LY303870 (30 mg/kg, i.p.) at 1 hour before C-fiber stimulation (C-fiber+LY). Nociceptive changes were accessed at 48 h after stimulation (n = 8 per cohort). Lumbar spinal cord collected 48 h post stimulation was assessed by immunostaining for Iba-1, a microglia marker (n = 6-7 per cohort). Lumbar cord collected at 3 h post stimulation underwent PCR analysis (n = 8 per cohort). Sciatic C-fiber stimulation induced allodynia (A) and unweighting (B), and this response was blocked by LY303870 pretreatment. (C) Sciatic C-fiber stimulation also induced microglia activation at 48 h after stimulation that was blocked by LY303870 pretreatment. (D) mRNA levels of several inflammatory mediators (TNF, IL-1, IL-6, and NGF) were significantly higher in the lumbar cord at 3 hours after sciatic stimulation, but pretreatment with LY303870 prevented this increase. Similarly, sciatic stimulation also increased spinal cord mRNA levels for TAC1, TACR1, CALCA, CALCB, and RAMP1, and pretreatment with LY303870 blocked these increases (Fig. 9D). (E) Representative fluorescence photomicrographs of Iba-1 protein (red) in the ipsilateral spinal dorsal horn at L5 segment from a sham stimulated rat (left), an C-fiber stimulated rat ipsilateral spinal cord (middle), and a C-fiber stimulation-treated rat pretreated with LY303870 (right). Boxed regions are enlarged at the bottom of each panel. Scale bar in D =100μm in top panels, 25 μm in bottom panels. Data were analyzed using a one-way analysis of variance (ANOVA) with Bonferonni correction test for post-hoc contrasts, error bars indicate SEM. Immunostaining data is presented as mean fluorescence intensity per high powered field. *p <0.05, **p <0.01, ***p <0.001 vs. Sham stimulation, #p <0.05, ##p <0.01, ###p <0.001 vs. C-fiber stimulation. TNF: tumor necrosis factor α, IL-1: interleukin 1β, IL-6:interleukin 6, NGF:nerve growth factor, TAC1:substance P, TACR1: SP neurokinin1 receptor, CALCA:alpha-calcitonin gene-related peptide, CALCB:beta-calcitonin gene-related peptide, CALCRL: calcitonin gene-related peptide type 1 receptor, and RAMP1:receptor activity modifying protein 1, CGRP receptor chaperone gene.

Figure 6.

Brief sciatic nerve electrical stimulation at C-fiber intensity, but not Aδ, Aβ-fiber intensity, induced prolonged nociceptive sensitization, and this effect was blocked by proximal sciatic nerve bupivacaine pretreatment (prox block). Rats underwent surgery and sciatic nerve electrode placement without (sham) or with 5 min of sciatic nerve electrical stimulation at C-fiber (5Hz, 0.5msec, and 10mA) or Aδ/Aβ fiber (10 Hz, 0.15 msec, and 5 mA) intensity. One group of rats was pretreated with local application of 0.5% bupivacaine to the sciatic nerve proximal to the stimulation site for 10 minutes before stimulation. Hindpaw nociceptive testing and assessment of warmth and edema were performed prior to stimulation, and at 3, 24, and 48 h after stimulation. Graphs A and B illustrate that sciatic nerve stimulation at C-fiber intensity caused von Frey allodynia and unweighting lasting at least 48 h, but there was no allodynia or unweighting observed after stimulation at Aδ/Aβ fiber intensity, or with sham stimulation. Furthermore, local application of bupivacaine 10 min before stimulation blocked C-fiber stimulation induced nociceptive sensitization. Neither C-fiber nor Aδ/Aβ fiber stimulation could induce hindpaw warmth and edema. Five minutes of sciatic nerve stimulation at C-fiber intensity had no effect on L4 DRG ATF3 (neuronal injury marker) mRNA levels at 48 hours after stimulation, indicating that this duration and intensity of sciatic stimulation did not induce axonal loss (E). Behavioral data were analyzed using a two-way repeated measures ANOVA with post-hoc Bonferroni correction to compare differences between rats at indicated time points. ATF3 data was analyzed with a one-way analysis of variance (ANOVA) with post-hoc Bonferroni correction, error bars indicate SEM. * P < 0.05, **P <0.01, *** P <0.001 vs. sham, # P< 0.05, ## P < 0.01 vs. baseline. BL: baseline.

3. Results

3.1 Fracture activated microglia and astrocytes in the lumbar dorsal horn

First we examined the effects of fracture on spinal glial cell activation by using real-time PCR to analyze lumbar cord mRNA levels for gene markers of glial activation. Specific primers for microglial cell and astrocyte activation markers (Table 1) were used (Tanga et al. 2004, Hathway et al. 2009, Yang et al. 2012). Figure 1A demonstrates that mRNAs levels of all microglia markers except CSF1 were significantly higher in spinal dorsal horn at 1 week after fracture compared to the control, and levels CCL2, MMP3, and IL-6 remained elevated at 4 weeks after fracture (Fig. 1B). Figure 1C shows that mRNA levels of the astrocyte markers vimentin (VIM) and GFAP were significantly upregulated at 1 week after fracture, and there was an insignificant elevation of GFAP at 4 weeks after fracture (Fig. 1D). These data indicate that spinal microglia and astrocyte activation occurred in the spinal cord at 1 week after fracture and that microglia activation persisted at 4 weeks postfracture.

3.2 Minocycline treatment inhibited postfracture nociceptive and vascular changes

These experiments tested the hypothesis that spinal microglia activation contributes to CRPS-like changes in the rat fracture model. Beginning at 4 weeks postfracture, rats were treated intraperitoneally with the microglia inhibitor minocycline (40mg/kg/day) for 3 days. Hindpaw nociceptive testing and assessment of warmth and edema were performed before and after treatment. Figures 2A-D demonstrate that systemic administration of minocycline partially reversed the hindpaw von Frey allodynia, unweighting, and warmth that developed after fracture, but had no effect on hindpaw edema. Systemic minocycline had no effect on von Frey thresholds in the contralateral (normal) hindpaw (data not shown), indicating that this treatment only inhibited allodynia and unweighting, not basal pain thresholds. To confirm that the spinal cord was the primary site of minocycline analgesic activity, the effects of intrathecal administration were also evaluated. Intrathecal minocycline (100μg/10μl/i.t., once) also reduced post-fracture hindpaw mechanical allodynia and unweighting, with anti-allodynic effects lasting up to 6 hours (Figs. 2E,F). Intrathecal minocycline treatment did not alter von Frey thresholds in the contralateral hindpaw, indicating no effect on basal pain thresholds.

3.3 Minocycline inhibited postfracture microglia activation in spinal dorsal horn

This experiment evaluated whether fracture induced-microglia activation is suppressed by minocycline treatment. Microglia activation in the lumbar spinal dorsal horn was assessed by immunohistochemistry using polyclonal antibodies to ionized calcium-binding adapter molecule 1 (Iba-1), a biomarker for resting and activated microglia. As shown in Figure 3A, Iba-1 labeled microglia in the dorsal horn of the control rats were evenly distributed with small cell bodies bearing long, thin processes. At 4 weeks after fracture Iba-1 labeled microglia were more abundant, with enlarged rounded bodies (i.e., amoeboid shape) and fewer and shorter processes, when compared to dorsal horn sections from the contralateral side or to control rats, indicating microglial activation. In contrast, fracture animals treated with minocycline (40mg/kg/day for 3 days) exhibited a clear decrease in microglia activation at 4 weeks after fracture (Fig. 3A). Quantitative analysis of the mean Iba-1 immunostaining intensity in the superficial layer of dorsal horn revealed that Iba-1 expression increased 2- and 1.4-fold compared to controls and to the side contralateral to fracture, respectively, and that minocycline treatment blocked the fracture-induced increase in Iba-1 immunoreactivity (Fig. 3B). Collectively, these results indicate that minocycline inhibited fracture-induced microglia activation in the lumbar dorsal horn.

Figure 3.

Fracture-induced spinal microglia activation was inhibited by minocycline treatment. Rats underwent tibia fracture (FX) and were casted for 4 weeks, then the cast was removed and then they were treated with minocycline (40mg/kg/day, i.p. daily for 3 days) or vehicle. Additional controls were intact animals treated with vehicle. After the third day of injections the animals were euthanized and the spinal cords were removed to assess microglial cell activation by immunohistochemistry for the microglia marker Iba-1. (A) Representative fluorescence photomicrographs of Iba-1 immunostaining (red) in the L5 dorsal horn segment are from an intact control rat, the dorsal horn ipsilateral to fracture, contralateral to fracture, and ipsilateral to fracture in a rat treated with minocycline. Boxed regions are enlarged at the bottom of each panel. These images demonstrate extensive dorsal horn microglia activation at 4 weeks postfracture that was reversed after minocycline treatment. Scale bars = 50 μm in top panels, 25 μm in bottom panels. (B) The quantification of Iba-1 positive cells revealed that minocycline significantly inhibited fracture-induced microglia activation (n = 6-7 per cohort). Data were analyzed using a one-way analysis of variance (ANOVA) with Bonferonni correction test for post-hoc contrasts, error bars indicate SEM. Immunostaining data is presented as mean fluorescence intensity per high powered field. *p<0.05, **p < 0.01 vs. control values. ##p <0.01 vs. FX-ipsi. Error bars indicate SEM. FX: fracture, ipsi: ipsilateral, contra: contralateral, Mino: minocycline.

3.4 LAA inhibited postfracture nociceptive sensitization and spinal astrocyte activation

This experiment tested the hypothesis that spinal astrocyte activation contributes to nociceptive sensitization observed at 4 weeks after fracture. Fracture rats were injected intrathecally with the astrocyte specific toxin LAA, (150 nmol/10 μl/i.p.) at 4 weeks postfracture. Hindlimb nociceptive testing was performed prior to fracture and again at 4 weeks postfracture, then the fracture rats underwent intrathecal injection and were retested 3, 6, and 24 h after injection. Figures 4A, 4B illustrate that LAA treatment reversed hindpaw von Frey allodynia and alleviated unweighting, and that these antinociceptive effects lasted for at least 24 h after drug administration. Intrathecal LAA treatment did not alter von Frey thresholds in the contralateral hindpaw (data not shown), indicating no effect on basal pain thresholds. At 4 weeks after fracture GFAP labeled astrocytes were more abundant in both the ipsilateral and contralateral L4/5 segment dorsal horn, when compared to dorsal horn sections from control rats, indicating astrocyte activation (Fig. 4C). In contrast, LAA treatment reversed the fracture-induced increase in immunoreativity (Fig. 4C). Figure 4D shows representative confocal immunofluorescence microscopy results for the astrocyte activation marker GFAP in the dorsal horn. Collectively, these results indicate that LAA treatment inhibited fracture-induced astrocyte activation and central nociceptive sensitization.

Figure 4.

Spinal astrocyte activation contributes to nociceptive sensitization after fracture. Rats underwent tibia fracture (FX) and were casted for 4 week, then the cast removed and the next day they were treated intrathecally once with L-α-aminoadipate (LAA, astrocyte specific toxin, 150 nmol in 10 μl saline, i.t.). Hindpaw nociceptive testing was performed prior to fracture, 4 weeks after fracture upon cast removal (BL), and 3, 6, and 24 h after drug administration. The animals were euthanized after the last behavioral test and the spinal cords were collected for immunostaining for glial fibrillary protein (GFAP), an astrocyte marker. The methods for calculating the graphed values are described in Fig. 2. LAA treatment reversed postfracture allodynia (A) and the hindlimb unweighting (B) and these analgesic effects lasted at least 24 h after drug administration (n=8 per cohort). (C) The quantification of GFAP immunostaining revealed that LAA significantly inhibited fracture-induced astrocyte activation. Panel D presents representative fluorescence photomicrographs of GFAP immunostaining (red) in the L5 dorsal horn segment in controls, ipsilateral to the fracture, contralateral to the fracture, and ipsilateral to fracture in a rat treated with LAA. Scale bars=100 μm. Data were analyzed using a one-way analysis of variance (ANOVA) with Bonferonni correction test for post-hoc contrasts, error bars indicate SEM (n = 6-8 per cohort). Immunostaining data is presented as mean fluorescence intensity per high powered field. * P < 0.05, **P <0.01 vs. control, # P< 0.05, ## P < 0.01 vs. FX-BL. Error bars indicate SEM. FX: fracture, BL: baseline, FX: fracture, ipsi: ipsilateral, contra: contralateral, LAA: L-α-aminoadipate.

3.5 LY303870 inhibited fracture-induced nociceptive and vascular changes and glial activation

Using transgenic SP deficient mice we recently observed that SP spinal signaling contributes to spinal inflammatory mediator expression and nociceptive sensitization in the fracture CRPS model (Shi et al. 2015). Based on these results we postulated that exaggerated SP spinal signaling would activate lumbar cord glial cells contributing to the nociceptive sensitization observed after fracture. At 4 weeks postfracture rats were treated with either the selective SP NK1 receptor antagonist LY303870 (30 mg/kg/day/i.p.) for 3 days or with saline. Hindlimb nociceptive and vascular changes were assessed prior to fracture, at 4 weeks postfracture (baseline), and after 3 days of LY303870 administration. Lumbar spinal cord tissue was collected after behavioral tests for Iba-1 or GFAP immunostaining. Figures 5A-D illustrate that systemic LY303870 treatment alleviated fracture-induced hindpaw allodynia, unweighting, and warmth. LY303870 treatment also inhibited fracture-induced microglia and astrocyte activation (Figs 5E-H). There was no effect on von Frey thresholds in the contralateral limb (data not shown), indicating that LY303870 treatment only inhibited allodynia and unweighting, not basal pain thresholds. Collectively, these data indicate that SP signaling promotes spinal microglia and astrocyte activation and contributes to the nociceptive and vascular changes observed in the fracture CRPS model.

3.6 Sciatic nerve stimulation at C-fiber intensity induced prolonged nociceptive sensitization

We postulated that nociceptor excitation in the fracture limb triggers SP release from the central afferent terminals in the lumbar dorsal horn, with subsequent spinal glial activation and nociceptive sensitization. To test this hypothesis, we first identified the primary afferent fiber activity contributing to nociceptive sensitization. Control (no fracture) rats underwent 5 minutes of sciatic nerve electrical stimulation at C-fiber (10mA, 5Hz, 0.5msec) or Aδ/Aβ-fiber (5mA, 10Hz, 0.15msec) intensity (Hathway et al. 2009) and behavioral testing for hindpaw von Frey allodynia, unweighting, warmth and edema was performed before and 3, 24, and 48 hours after stimulation. Five minutes of C-fiber stimulation caused hindpaw von Frey allodynia and unweighting lasting 48 h after stimulation (Figs. 6A, 6B), but no hindpaw warmth or edema were observed (Figs. 6C, 6D). No allodynia or unweighting were observed after 5 min of sciatic nerve stimulation at Aδ/Aβ fiber intensity, or in sham stimulated rats (Figs. 6A, 6B). Local application of bupivacaine to the sciatic nerve proximal to the stimulation site blocked C-fiber stimulation induced hindpaw allodynia and unweighting (Figs. 6A, 6B), indicating that the pro-nociceptive effects of C-fiber stimulation were mediated by proximal action potential propagation. Five minutes of sciatic nerve electrical stimulation at C-fiber intensity had no effect on L4 dorsal root ganglia (DRG) ATF3 mRNA levels at 48 h after stimulation, indicating that this duration and intensity of sciatic stimulation did not induce axonal loss (Fig. 6E) (Hathway et al. 2009). This was in marked contrast to a sciatic axotomy group (positive control), where levels of ATF-3 mRNA were increased 4-fold in the ipsilateral L4 DRG relative to sham or sciatic stimulated rats (Fig. 6E), reflecting retrograde changes in the cell bodies after axotomy.

3.7 Minocycline treatment blocked sciatic stimulation induced microglia activation and nociceptive sensitization

The microglia inhibitor minocycline (40 mg/kg/day/i.p.) was administered 1 h prior to 5 min of sciatic C-fiber stimulation, and at 23 and 47 h after stimulation. Minocycline treatment reversed sciatic stimulation induced hindpaw allodynia and unweighting at 48 h after sciatic stimulation (Figs. 7A,7B). There was no effect on von Frey thresholds in the contralateral limb (data not shown), indicating that minocycline treatment only inhibited allodynia and unweighting, not basal pain thresholds. Figure 7D illustrates that mRNA levels of several microglia activation markers (IL-6, CCL2) were significantly higher in the lumbar cord at 3 h after sciatic stimulation, compared to controls, but by 48 h after stimulation the mRNA levels of all the microglia markers had returned to normal (Fig. 7E). Quantification studies revealed a 73% increase in L4/5 dorsal horn Iba-1 immunoreactivity at 48 h after sciatic stimulation, and this effect was blocked by minocycline treatment (Fig. 7C). Figure 7F presents representative fluorescence photomicrographs of Iba-1 protein in the spinal dorsal horn at L5 segment from a sham stimulated control rat (left), a C-fiber stimulated rat (middle), and a C-fiber stimulated rat treated with minocycline (right). These images demonstrate that minocycline inhibits C-fiber stimulation-induced microglia activation.

3.8 LAA treatment blocked sciatic stimulation induced nociceptive sensitization

The astrocyte toxin LAA (150 nmol/10 μl saline/i.t.) was administered 1 h prior to 5 min of sciatic C-fiber intensity stimulation. LAA treatment inhibited the development of sciatic stimulation induced hindpaw allodynia and unweighting at 48 h after sciatic stimulation (Figs. 8A, 8B). There was no effect on von Frey thresholds in the contralateral limb (data not shown), indicating that LAA treatment only inhibited allodynia and unweighting, not basal pain thresholds. Figure 8D illustrates that mRNA levels of most astrocyte activation markers (VIM, NOCR, CNTFR, EGFR) were significantly higher in the lumbar cord at 3 h after sciatic stimulation, compared to controls, but by 48 h after stimulation the mRNA levels of all the astrocyte markers had returned to normal (Fig. 8E). Consistent with the mRNA results, 5 minutes of sciatic stimulation did not increase GFAP immunoreactivity in the L4/5 dorsal horn at 48 h post stimulation (Fig. 8C). These data suggest that 5 min of sciatic C-fiber activity could induce astrocyte activation and nociceptive sensitization.

Figure 8.

Intrathecal L-α-aminoadipate (LAA, astrocyte toxin) injection blocked C-fiber stimulation-induced nociceptive sensitization. Rats underwent surgery and sciatic nerve electrode placement without (sham) or with 5 minutes of electrical stimulation at C-fiber (5Hz, 0.5msec, intensity, and one group of rats was pretreated with LAA (C-fiber+LAA, 150 nmol in 10 μl saline, i.t., given 1 hour before C-fiber stimulation). Nociceptive changes were assessed at 48 h after LAA treatment. Sciatic nerve stimulation induced allodynia (A) and unweighting (B), whereas LAA intrathecal pretreatment attenuated this pronociceptive effect. At 48 h after sciatic stimulation the spinal cords were collected and immunostaining for GFAP (an astrocyte marker) was unchanged after sciatic stimulation (C, n=6-7 per cohort). The mRNA levels of several astrocyte activation markers (VIM, NOCR, CNTFR, EGFR) were significantly higher in the lumbar cord at 3 h after sciatic stimulation (D), but by 48 h after stimulation these had returned to normal (E). Data were analyzed using a one-way analysis of variance (ANOVA) with Bonferonni correction test for post-hoc contrasts, error bars indicate SEM. Immunostaining data is presented as mean fluorescence intensity per high powered field. *p <0.05, **p <0.01, ***p <0.001 vs. sham stimulation, # P< 0.05, ## P< 0.01 vs. C-fiber stimulation. C-fiber: electrical stimulation at C-fiber intensity, LAA: L-α-aminoadipate, VIM: vimentin, NCOR: nuclear receptor co-repressor 1, CNTFR: ciliary neurotrophic factor receptor, EGFR: epidermal growth factor receptor, GFAP: glial fibrillary acidic protein, NES: nestin, SOX2: SRY (sex determining region Y)-box 2.

3.9 LY303870 treatment blocked sciatic stimulation induced microglia activation and nociceptive sensitization

The SP NK1 receptor antagonist, LY303870 (40 mg/kg/i.p.) was administered 1 h before 5 min of sciatic C-fiber stimulation. LY303870 completely blocked the stimulation-induced allodynia and unweighting at 48 h after stimulation (Figs 9A, 9B), but had no effect on contralateral hindlimb von Frey thresholds (data not shown). Figure 9D illustrates that mRNA levels of several inflammatory mediators (TNF, IL-1, IL-6, and NGF) were significantly higher in the lumbar cord at 3 h after sciatic stimulation, compared to controls, but pretreatment with LY303870 completely blocked this increase. Similarly, sciatic stimulation also increased spinal cord mRNA levels for TAC1 (SP gene), TACR1 (SP neurokinin1 receptor gene), CALCA (alpha-CGRP gene), CALCB (beta-CGRP gene), and RAMP1 (receptor activity modifying protein 1, CGRP receptor chaperone gene), and pretreatment with LY303870 completely blocked these increases (Fig. 9D). No changes were observed in mRNA expression for CALCRL (calcitonin gene-related protein type 1 receptor, CGRP receptor gene). Sciatic stimulation increased Iba-1 (microglia marker) immunoreactivity in the L4/5 dorsal horn at 48 hours post stimulation and LY303870 pretreatment blocked this effect (Fig. 9C). Figure 9E presents representative fluorescence photomicrographs of Iba-1 protein in the spinal dorsal horn at the L5 segment from a sham stimulated control rat (left), a C-fiber intensity stimulated rat (middle), and a C-fiber stimulated rat pretreated with LY303870 (right). These images demonstrate that LY303870 inhibits C-fiber stimulation-induced microglia activation and nociceptive sensitization.

4. Discussion

Both microglia and astrocytes express SP NK1 receptors (Torrens et al. 1986, Lin 1995, Lai et al. 2000, Rasley et al. 2002, Marriott 2004, Block et al. 2006) and SP stimulated glia express TNF, IL-1, and IL-6 pro-inflammatory cytokines (Martin et al. 1993, Luber-Narod et al. 1994, Fiebich et al. 2000, Zhou et al. 2010, Zhu et al. 2014). Previously we observed a postfracture increase in SP mRNA and protein expression in the ipsilateral L4,5 dorsal root ganglia and sciatic nerve, and increased expression of the SP NK1 receptor in the lumbar cord, indicating amplification of SP spinal signaling after fracture (Wei et al. 2009, Shi et al. 2015). We also observed an increase in spinal inflammatory mediators (TNF, IL-1, IL-6, CCL2, and NGF) after fracture and by using transgenic SP deficient fracture mice we demonstrated that this increase in spinal inflammatory mediators was dependent on SP signaling. Using intrathecal injections of SP, TNF, IL-1, IL-6, CCL2, and NGF inhibitors in the fracture rats, we demonstrated that these spinal inflammatory mediators contributed to nociceptive sensitization after fracture (Shi et al. 2015). Based on these data and the accumulating evidence indicating that spinal glial cell activation drives the facilitation of nociceptive transmission in the dorsal horn (Watkins et al. 2001, 2003, Nakagawa and Kaneko 2010, Ren and Dubner 2010, Ji et al. 2013, Clark et al. 2015), we postulated that exaggerated postfracture SP signaling in the dorsal horn activates spinal microglia and astrocytes, resulting in inflammatory mediator expression and chronic central nociceptive sensitization.

The current study demonstrated chronic spinal glia activation after tibia fracture. At 1 week after fracture there was an increase in the expression of most microglia activation markers in the ipsilateral lumbar cord, and CCL2, MMP3, and IL-6 mRNA levels remained elevated at 4 weeks (Fig. 1). The astrocyte activation markers vimentin and GFAP were up regulated at 1 week after fracture, but by 4 weeks they had returned to baseline levels (Fig. 1). These PCR data indicate early spinal microglia and astrocyte gene activation at 1 week and persistent microglia activation at 4 weeks postfracture.

Immunostaining for the microglia marker Iba-1 demonstrated a 2-fold increase in dorsal horn Iba-1 expression at 4 weeks after fracture, and this increase was blocked by treatment with the microglia inhibitor minocycline (Fig. 3). Systemic administration of minocycline partially reversed established postfracture allodynia, unweighting, and warmth (Fig. 2). Intrathecal minocycline also partially reversed post-fracture allodynia and unweighting, with anti-allodynic effects lasting up to 6 hours (Fig. 2). Collectively, these results indicate that fracture-induced microglia activation contribute to the maintenance of chronic central nociceptive sensitization.

Immunofluorescence labeling for the astrocyte activation marker GFAP in the lumbar spinal cord demonstrated a 2-fold increase in GFAP immunostaining at 4 weeks postfracture that was reversed by intrathecal treatment with the astrocyte toxin LAA (Fig. 4). Intrathecal LAA treatment also reversed hindpaw von Frey allodynia and alleviated unweighting, and that these antinociceptive effects lasted for at least 24 hours after drug administration (Fig. 4). These data demonstrate that fracture-induced astrocyte activation in the lumbar spinal cord contribute to the maintenance of chronic central nociceptive sensitization.

Minocycline analgesic trials in neuropathic pain patients have been disappointing (Vanelderen et al. 2015). Preclinical studies in neuropathic animal models suggest that minocycline analgesia is effective when started preemptively before nerve injury occurs, but when minocycline treatment is started after the establishment of neuropathic pain it is usually ineffective, suggesting that microglial activation contributes to the development, but not the maintenance of neuropathic pain (Raghavendra et al. 2003, Padi and Kulkarni 2008, Zhang et al. 2012, Yamamoto et al. 2015). Interestingly, there is preclinical evidence that LAA or fluorocitrate treatment can inhibit astrocyte activation and reverse established neuropathic pain, suggesting that astrocyte activation does play a role in the maintenance of neuropathic pain (Zhang et al. 2012, Ji et al. 2013). In contrast, in animal models of bone cancer pain minocycline treatment is effective when started weeks after the onset of tumor injection (Mao-Ying et al. 2012, Yang et al. 2015). Similarly, in the current study both intrathecal minocycline and intrathecal LAA, when started 4 weeks after fracture, partially reversed hind paw allodynia and unweighting, evidence suggesting that both microglia and astrocytes contribute to the maintenance of chronic post fracture pain (Fig. 2, 4). Neither of these drugs had any affect on von Frey thresholds in the contralateral intact hindlimb, indicating that their analgesic effects were restricted to the hypersensitive limb.

Using transgenic mice lacking SP, we recently observed that substance P spinal signaling contributes to up-regulated spinal inflammatory mediator (TNF, IL-1, IL-6, CCL2, NGF) expression and nociceptive sensitization after fracture (Shi et al. 2015). In the current study we found that treatment with the SP NK1 receptor antagonist LY303870 blocked fracture-induced microglia and astrocyte activation and alleviated fracture-induced hindpaw allodynia, unweighting, and warmth (Fig. 5). Collectively, these results suggest that SP signaling promotes the spinal microglia and astrocyte activation, spinal inflammatory mediator expression, and nociceptive sensitization observed in the fracture CRPS model.

On the basis of evidence that; 1) a brief period of low frequency C-fiber sciatic stimulation is sufficient to evoke prolonged microglial-induced central sensitization and spinal IL-6 and CCL2 expression (Hathway et al. 2009), 2) that neuropeptide SP is contained within a subpopulation of nociceptive (small) primary sensory neurons that project to the superficial dorsal horn of the spinal cord (Radhakrishnan and Henry 1995), and 3) that SP is released into spinal cord dorsal horn in response to stimulation of primary afferent C-fibers (Lever et al. 2001), we hypothesized that peripheral C-fiber nociceptor excitation releases SP in the dorsal horn, resulting in spinal glial activation and central nociceptive sensitization after fracture.

The current study demonstrated that brief low-frequency electrical stimulation of the sciatic nerve at C-fiber intensity caused hindpaw allodynia and unweighting lasting at least 48 h after stimulation, without evidence of nerve damage (Fig. 6). Immunostaining for the microglia marker Iba-1 was increased at 48 hours post stimulation and minocycline treatment blocked this increase and inhibited the development of hindpaw allodynia and unweighting (Fig. 7). Immunostaining for the astrocyte marker GFAP was not increased at 48 hours after stimulation, but LAA treatment reversed hindpaw allodynia and unweighting at 48 hours (Fig. 8). We postulate that a more chronic afferent bombardment of nociceptive impulses is required to induce the increase in dorsal horn GFAP immunoreactivity observed at 4 weeks postfracture (Fig. 4). Collectively, these results indicate that a brief burst of low frequency C-fiber discharges can induce prolonged lumbar dorsal horn microglia activation that contributes to central nociceptive sensitization.

Substance P signaling was required for electrically-evoked microglial activation, spinal inflammatory mediator expression, and nociceptive sensitization. Pretreatment with the SP NK1 receptor antagonist LY303870 completely blocked C-fiber evoked spinal microglia activation, increased mRNA expression of TNF, IL-1, IL-6, CCL2, and NGF in the lumbar cord, and hindpaw mechanical allodynia and unweighting (Figs. 7,9). Similarly, we previously had observed increased lumbar cord mRNA levels for TNF, IL-1, IL-6, CCL2, and NGF after tibia fracture in mice, but transgenic fracture mice lacking SP had diminished or no increases in spinal TNF, IL-1, CCL2, or NGF mRNA levels (Shi et al. 2015). When SP, TNF, IL-1, IL-6, CCL2, and NGF inhibitors were intrathecally administered, each of these agents attenuated post fracture allodynia and unweighting (Shi et al. 2015). These results support the hypothesis that a brief period of low frequency C-fiber discharges can induce SP release in the dorsal horn, initiating microglial activation, inflammatory mediator expression, and nociceptive sensitization, similar to the glial activation, up-regulated inflammatory mediator expression, and hypersensitivity observed at 4 weeks after fracture.

Sciatic stimulation also increased spinal cord mRNA levels for TAC1 (SP gene), TACR1 (the SP neurokinin1 receptor gene), CALCA (alpha-CGRP gene), CALCB (beta-CGRP gene), and RAMP1 (CGRP receptor chaperone gene), and pretreatment with LY303870 completely blocked these increases (Fig. 9). Similarly, we had previously observed increased lumbar cord mRNA levels for TAC1, TACR1, CALCA, CALCB, CALCRL (CGRP receptor gene), and RAMP1 in 3 weeks postfracture mice, but transgenic mice lacking SP had minimal or no postfracture increase in TACR1, CALCA, CALCB, CALCRL, or RAMP1 (Shi et al. 2015). These data suggest that a positive feedback loop may amplify SP spinal signaling after fracture. We postulate that discharging C-fibers release SP in the lumbar cord, stimulating microglia and astrocytes to express additional TAC1 receptors and thus amplifying the pronociceptive effects of SP (and perhaps CGRP) release in the dorsal horn.

It has been proposed that intense nociceptive afferent activation triggers neurotransmitter release in the dorsal horn, thus activating spinal glia to express and release inflammatory mediators (Watkins et al. 2001, 2003). This is the first study to demonstrate that substance P signaling plays a critical role in the maintenance of spinal glial activation and chronic central sensitization after limb injury. The results of this study provide preclinical support for CRPS treatments inhibiting central sensitization, such as NMDA antagonists, intrathecal drug treatments, and spinal cord stimulation.

Highlights.

• Substance P signaling contributed to spinal glial activation and nociceptive sensitization

• C-fiber afferent signaling induced spinal glial activation and hypersensitivity.

• Neuroglia activation contributed to the maintenance of chronic post fracture hypersensitivity.

Abreviations

CRPS

complex regional pain syndrome

SP

substance P

LAA

L-2-aminoadipic acid

TNF

tumor necrosis factor α

IL-1

interleukin 1β

IL-6

interleukin 6

NGF

nerve growth factor

CCL2

chemokine (C-C motif) ligand 2

NK1

the SP neurokinin 1 receptor

GFAP

glial fibrillary acidic protein