Background
Peripheral neuropathy arises from lesion and disease affecting the peripheral nervous system. There are over 100 peripheral neuropathy types associated with metabolic disease, nutritional deficiencies, trauma and exposure to drugs, toxins, alcohol, and viral pathogens, with neurological symptoms including muscle weakness, numbness, loss of autonomic functions, and debilitating neuropathic pain [
1]. The mechanisms underlying pathological changes in peripheral neuropathy depend on the triggers and may be influenced by biological variables, including sex and gender. Recent experimental evidence highlighted distinct immune response mechanisms to peripheral nerve injury in the development of mechanical hypersensitivity in female and male rodents [
2]. These recent data added to the body of evidence of sexual dimorphism in the formation of collateral axonal sprouting, cortical connectivity, and the activities of brain-derived neurotrophic factor,
N-methyl-D-aspartate, and opioid receptors in the damaged nervous system (reviewed in [
3]).
The matrix metalloproteinase (MMP) family of calcium-dependent zinc endopeptidases includes soluble (collagenases, gelatinases, matrilysins, and stromelysins) and membrane-type MMPs [
4]. The structure of soluble MMP comprises an N-terminal inhibitory prodomain followed by an active site catalytic domain, a flexible linker region, and a C-terminal hemopexin domain. MMPs are synthesized as latent inactive proenzymes that require the proteolytic removal of their inhibitory prodomain to expose the active site and generate the catalytically active mature proteases. MMP proteolysis regulates the levels and the functionality of extracellular matrix components and cell surface signaling receptors [
5]. In the damaged adult peripheral nerve, pro-inflammatory MMP-9 (gelatinase B) is produced uniquely and immediately after injury by Schwann, endothelial, and immune cells to regulate the blood-nerve barrier breakdown, immune cell recruitment, glial activation, demyelination, remyelination, and pain [
6‐
16].
MMP activity is regulated by four tissue inhibitors of MMPs (TIMPs), each comprised of the N-terminal inhibitory and the C-terminal non-inhibitory domains. Among TIMPs, TIMP-1 is the most efficient inhibitor of the pro-inflammatory MMP-9 gelatinase [
4]. In addition, TIMP-1 via its C-terminal domain forms a unique stoichiometric (1:1), stable heterodimer complex with the hemopexin domain of MMP-9 proenzyme. This complex is significantly more resistant to activation relative to the TIMP-1-free MMP-9 proenzyme. Because TIMP-1 is highly expressed in the naïve rat nerve [
17,
18] and is further upregulated concomitantly to MMP-9 in the acute phase of nerve injury [
17,
18], MMP-9 is found predominantly as a latent inactive proenzyme [
7,
8,
17,
18].
Whether MMP-9 is expressed and active in late-phase painful peripheral nerve injury remains unknown. In light of the emerging evidence of sexual dimorphism in neuroimmune pathogenesis of neuropathic pain progression after peripheral nerve injury [
2], we here aimed to determine the patterns of MMP-9 expression, activity, and excretion utilizing a widely used model of painful peripheral neuropathy, rat sciatic nerve chronic constriction injury (CCI) in female and male rats. Our results evidence excessive, uninhibited proteolytic MMP-9 activity in late-phase (day 28) post-CCI and suggest that the roles of MMP-9 in peripheral neuropathy are universal in both sexes.
Discussion
Despite the high regenerative capacity of the peripheral nerve, nerve trauma often results in poor functional recovery and debilitating neuropathic pain [
36‐
38]. Chronic neuroinflammation is a major contributing factor in the development of neuropathic pain [
39‐
41]. Chronic constriction injury (CCI) to rat sciatic nerve is an animal model for the study of the mechanisms underlying clinically relevant pain-like behaviors from normally painless stimuli (allodynia) and exaggerated pain from painful stimuli (hyperalgesia) sustained for weeks after incitement of CCI [
26‐
29]. In the models of painful peripheral neuropathy, the MMP enzyme family has multiple established and critical roles in immune cell recruitment and demyelination and the development of neuropathic pain [
6‐
16,
19,
39,
42].
MMP-9 is a unique early-response proinflammatory MMP family member induced in adult nerve exclusively after injury, including CCI [
6,
8,
10‐
12,
14,
23,
43‐
45], as confirmed in the present study. Proinflammatory cytokines, including TNF-α and IL-1β, stimulate MMP-9 expression in Schwann cells and endothelial cells of the nerve [
11,
12]. Infiltrating immune cells, such as neutrophils and macrophages, serve as additional sources of MMP-9 in the damaged nerve. A number of the early-phase MMP-9 activities in nerve injury have been established, including immune cell infiltration, suppression of Schwann cell mitosis, and promoting demyelination by degradation of myelin protein [
7‐
12,
44]. In addition to tissue remodeling at the nerve injury site, MMP-9 is thought to contribute to pro-nociceptive dorsal spinal cord plasticity after peripheral nerve injury [
6,
46,
47]. In support of the net proinflammatory early-phase gelatinolytic activity in painful neuropathy, systemic acute MMP-9/2 inhibition at day 1 post-CCI prevents the development of neuropathic pain [
8]. In addition, broad-spectrum MMP inhibition [
44] and MMP-9 targeting using siRNA [
19] protect from nerve injury-induced pain.
The present study offers the first evidence that MMP-9 expression and activity are elevated in the nerve during late-phase nerve injury. Because of a multi-fold shift of the MMP-9:TIMP-1 ratio in favor of MMP-9, the excess MMP-9 activity was unencumbered by its endogenous inhibitor TIMP-1 specifically during the late-phase. These data are in contrast to the findings of the early-phase MMP-9 increase as inactive pro-enzyme and add significantly to the previous studies implicating the early-phase action of MMP-9 in pain [
8,
19,
44]. In addition to pain, the early-phase Schwann cell anti-mitogenic MMP-9 action during the first days post-nerve crush affects long-term nerve repair in part by regulating Schwann cell numbers in the regenerating nerve [
7,
9,
10]. Using MMP-9 knockout animals, we have established the essential role of MMP-9 in voltage-gated sodium (Nav) channel clustering in the nodes of Ranvier, and MMP-9 deletion resulted in excessively large Nav channel clusters at 1 month post-crush [
7]. This finding suggests that by regulating remyelination, MMP-9 may contribute to pain resolution. Insofar, the direct pro-regenerative role of gelatinases in the sciatic nerve has been attributed to MMP-2, and not MMP-9, via degradation of growth-inhibitory proteoglycans [
48,
49].
The peripheral nerve comprises a plethora of MMP-9 substrates. By regulation of the cleavage and the release of soluble cytokine and trophic factor ligand and receptors from their membrane isoforms, MMP-9 activity regulates activation and inactivation of cytokine and trophic pathway in the damaged nerve [
10,
39,
43]. By proteolytic fragmentation of myelin basic protein (MBP) and myelin-associated glycoprotein (MAG) of nerves, MMP-9 activity promotes demyelination and the release of the pro-algesic MBP [
8,
22,
50] and growth-inhibitory MAG [
51] fragments, respectively. MMP-9 controls Nav channel clustering in the injured nerve by proteolysis of laminin and dystroglycan, both required for Nav clustering [
7,
52], or by direct proteolysis of the extracellular loop region of Nav channels [
53]. The evidence of the active MMP-9 species in the nerve at the late-phase painful neuropathy offers a possibility for MMP-9 role in pain sustenance or resolution via control of nerve regeneration, demyelination, and ion channel functioning. Our studies also indicate multiple roles of catalytic MMP-9 (defined by inhibition using GM6001 inhibition) in Schwann cell signaling; it produced a bi-phasic ERK1/2 activation by proteolytic activation of insulin-like growth factor 1, epidermal growth factor, and platelet-derived growth factor signaling pathways [
10]. In turn, by inhibiting excess MMP-9 proteolysis, TIMP-1 is expected to promote nerve regeneration and remyelination [
7,
18] and to inhibit the development of neuropathic pain [
19,
54]. Our data suggest that insufficient TIMP-1 level to limit late-phase MMP proteolysis in the damaged nerve provides rationale for the use of late-phase TIMP-1-based therapy in painful neuropathy.
In addition to proteolysis of extracellular proteins and protein domains [
55,
56], MMP-9 regulates cell behavior by direct receptor binding [
25,
57‐
59]. The unique representation of the multiple MMP-9 isoforms, including the proenzyme, mature enzyme, homodimers, and heterodimers, in the injured nerve suggests its complex and multiple roles. At day 1 post-CCI, high levels of inactive MMP-9 proenzyme are consistent with high TIMP-1 level. In excess of latent MMP-9 over TIMP-1, MMP-9 proenzyme free from TIMP-1 is activated by plasmin or MMP-3 [
60]. In addition, MMP-9 proenzyme may be involved in signaling neurite growth or Schwann cell migration through interaction of the non-catalytic hemopexin domain with low-density lipoprotein receptor-related protein-1 [
61,
62] or through the cell surface complex with CD44, mediating TIMP-1-induced survival signaling [
63]. Thus, the spatial and temporal relation of the specific MMP-9 and TIMP-1 species and binding partners will determine the functional outcomes of their individual and joint activities in painful peripheral neuropathy.
MMP-9 is a glycoprotein with multiple
O-glycosylation sites in the linker region between the catalytic and hemopexin domains and, in addition, two well-defined
N-glycosylation sites—one in the prodomain and another in the catalytic domain [
33‐
35]. A majority of the secretory and cell surface proteins that are shed into the urine are glycosylated [
30,
64]. The urinary MMPs are elevated in multiple cancer and nephropathy types, osteoarthritis, and heart failure [
31,
65‐
68] and constitutively in male compared to female rat urine. Similarly, we recorded a significant increase in the total MMP cleavage activity as well as the presence of the elevated levels of the multiple MMP species in rat urine post-nerve injury. While the non-invasive experimental urine measurements readily distinguished animals with painful neuropathy from normal animals, the presence of non-neuropathic conditions with elevated urinary MMPs [
31,
65‐
68] cannot be excluded in the clinical setting. Nevertheless, characterization of urinary extracellular vesicles present considerable research interest due to their possible use as a source of biomarkers and have been shown to contain proteases in patients with diabetic neuropathy [
32].
Understanding the influence of biological variables, including subject’s sex, on pathophysiology of painful neuropathy is critical to the development of personalized therapeutic and diagnostic approaches. Because females are more common sufferers of chronic pain [
20,
21], our historical [
7‐
13,
23] and initial experiments in the present study employed female rat CCI. Comparative studies of immune response to nerve injury in female and male rodents indicated clearly that immunotherapeutic targeting of mechanical allodynia is sex-dependent [
2]. Because MMP-9 expression, activity, and excretion are universal in female and male rats with painful neuropathy, in the search for selective markers of nerve injury and neuropathic pain, MMP-9 presents a reliable, injury-specific surrogate biomarker regardless of subject’s sex.
Methods
Reagents
All reagents were purchased from MilliporeSigma unless indicated otherwise. The horseradish peroxidase (HRP)-conjugated goat anti-rat IgM (#3020-05) was purchased from Southern Biotech. The HRP-conjugated goat anti-rat IgG (#112-035-175) and the [(7-methoxycoumarin-4-yl)acetyl]-Pro-Leu-Gly-Leu-[N-3-(2,4-dinitrophenyl)-L-2,3-diamino-propionyl]-Ala-Arg-NH2 (Mca-PLGL-Dpa-AR-NH2) fluorescent MMP substrate were obtained from Jackson ImmunoResearch and R&D Systems, respectively. A 3,3′,5,5′-tetramethylbenzidine substrate (TMB/E) and protease-free BSA (a 30% solution) were from Surmodics and US Biological, respectively. Gelatin-Sepharose 4B beads and Micro Bio-Spin columns were from GE Healthcare and Bio-Rad, respectively. The Coomassie protein assay reagent and Novex 10% zymogram (0.1% gelatin) gels were purchased from Thermo Scientific.
Animal model and sample collection
Sixty adult female and 20 male Sprague-Dawley rats (200–225 g) were obtained from Envigo Labs and housed in a temperature-controlled room (22 °C) on a 12-h light/dark cycle with free access to food and water. Following anesthesia with 4% isoflurane in oxygen (Aerrane; Baxter), the common sciatic nerve was exposed unilaterally at the mid-thigh level. The nerve received three loosely constrictive chromic gut ligatures to produce CCI [
26]. At days 0–60 post-CCI, urine sample aliquots (0.2–0.4 ml) were collected, readily placed on ice for a few minutes, and then cleared by centrifugation (2,000×
g; 10 min; 4 °C). Cleared urine aliquots were equilibrated in 50 mM HEPES, pH 7.5, containing 10 mM CaCl
2, 0.5 mM MgCl
2, and 10 μM ZnCl
2 (MMP buffer, pH 7.5), using a desalting spin column and immediately used in the MMP activity tests. Sciatic nerve samples were collected and snap-frozen in liquid N
2 and stored at − 80 °C for protease activity assays and in RNA-later and stored at − 20 °C for the qRT-PCR analyses. For immunohistochemistry, sciatic nerves were isolated following transcardial perfusion in 4% paraformaldehyde in 0.2 M phosphate, post-fixed and embedded in paraffin. Animals were sacrificed using Beuthanasia (150 mg/ml; i.p., Schering-Plough Animal Health). All animal procedures were performed according to the PHS Policy on Humane Care and Use of Laboratory Animals and the protocol approved by the Institutional Animal Care and Use Committee at the VA San Diego Healthcare System.
von Frey testing
Rats were habituated to the testing environment prior to baseline tests. Testing was performed daily for 3 consecutive days prior to and then at the indicated time points after CCI. Rats were placed in individual Plexiglas compartments with wire mesh bottom, and von Frey filaments (0.41–15.2 g, Stoelting) were applied perpendicularly to the mid hind paw and held for 4–6 s. A positive response was noted if the paw was sharply withdrawn. The 50% probability of withdrawal threshold was determined by Dixon’s up-down method [
69].
Taqman qRT-PCR
Taqman primers and a probe containing 5′-FAM reporter for rat TIMP-1 (GenBank, NM_053819) were from Applied Biosystems (cat. # Rn01430873_g1). Primers and probes for MMP-9 (GenBank, NM_031055) and glyceraldehyde 3-phosphate dehydrogenase (GAPDH; GenBank, X02231) were from Biosearch Technologies [
8,
12]. Total RNA was extracted using TRIzol and purified on an RNeasy mini column (Qiagen). The RNA purity was estimated by measuring the A
260/280 and the A
260/230 ratios. The samples were treated with RNase-free DNAse I (Qiagen). cDNA was synthesized using a first-strand cDNA kit (Roche). Gene expression levels were measured in a Mx3005P (Agilent) using 50 ng cDNA and 2×Taqman Universal PCR Master Mix (Applied Biosystems) with a one-step program: 95 °C, 10 min; 95 °C, 30 s; 60 °C, 1 min for 50 cycles. Using the injured sciatic nerve cDNA samples, primers (Biosearch Technologies) and Taqman probes for MMP-9 (Roche) and TIMP-1 (Applied Biosystems) were optimized to reach the amplification efficiency of 100.1–100.3% [
12]. GAPDH was used as a normalizer; its expression changes were insignificant in the injured relative to naïve nerves. Samples without cDNA (a no template control) showed no contamination. Relative mRNA levels were quantified using the comparative delta Ct method [
70]. The fold change between CCI and naïve samples was determined using the Mx3005P software.
Protease activity assay
The cleavage assay was performed in a 0.2-ml total volume in wells of a 96-well plate (Thermo Scientific) using the fluorescent Mca-PLGL-Dpa-AR-NH2 peptide substrate (1 μM) and the 50-μl urine aliquots equilibrated in the MMP buffer, pH 7.5. Where indicated, GM6001 (10 μM) was co-incubated for 30 min at ambient temperature with the urine samples to inactivate MMPs. Initial reaction velocity was monitored continuously at λex = 320 nm and λem = 400 nm using a fluorescence spectrophotometer. Data are means ± SD from several independent experiments performed at least in duplicate. Protein concentrations in the dialyzed urine samples were determined using the Coomassie protein assay.
MMP-9 purification using gelatin-Sepharose beads
The proteins were extracted for 1 h at 4 °C from the sciatic nerve samples using 50 mM Tris-HCl buffer, pH 7.4, containing 150 mM NaCl, 1% Triton X-100, 0.1% SDS, 10 mM EDTA, the protease cocktail inhibitor, and 1 mM phenylmethylsulfonyl fluoride. The protein concentration of the extracts was determined using the Coomassie protein assay and adjusted to reach 1 mg/ml. The extract aliquots (100 μg total proteins, each) were 10-fold diluted using the above buffer lacking Triton X-100 and SDS and allowed to bind to gelatin-Sepharose 4B beads for 16–18 h at 4 °C. Following extensive washing, the bound material was eluted using non-reducing 2×SDS sample buffer (50 μl).
Gelatin zymography
The dialyzed rat urine samples equilibrated in MMP buffer, pH 7.5; the crude nerve extracts; and the MMP-9 samples isolated from sciatic nerve were analyzed by gelatin zymography in a 10% acrylamide-0.1% gelatin gel. Gels were next processed as described previously to visualize the clear gelatinolytic activity bands [
7,
8]. Where indicated, gels were incubated in 20 mM EDTA to inactivate MMP activity.
Immunohistochemistry
Transverse (10-μm-thick) sciatic nerve sections collected at 4 weeks post-CCI after transcardial perfusion in 4% paraformaldehyde in 0.2 M phosphate, post-fixed and embedded in paraffin, were deparaffinized in xylenes and rehydrated in graded ethanol. Following endogenous aldehyde group block (0.5% sodium borohydride in 1% sodium dibasic buffer for 5 min) and non-specific binding block (5% goat serum in PBS for 30 min at room temperature), nerve sections were incubated with a rabbit polyclonal antibody to MMP-9 (1:500; Torrey Pines Biolabs, 16–18 h at 4 °C) alone or with rabbit polyclonal antibody to MMP-9 (1:50; SantaCruz, 16–18 h at 4 °C) sequentially with mouse monoclonal antibody to S100 (1:2000, Sigma, 16–18 h at 4 °C), followed by the respective species-specific Alexa 594 (red) or Alexa 488 (green)-conjugated secondary antibody (ThermoFisher Scientific, 1 h, room temperature, each). PBS was used for rinsing. Slowfade Gold antifade reagent containing DAPI (4′,6-diamidino-2-phenylindole, ThermoFisher Scientific, blue) was used for mounting. Staining specificity was confirmed by a primary antibody omission. The images were acquired using All-In-One Fluorescence Microscope BZ-X700 (Keyence, Itasca, IL).
Statistical analyses
Statistical analyses were performed using Graph Prism 6 (Synergy Software) or SPSS 16.0 (SPSS) software by a two-tailed, unpaired Student’s t test or one-way analyses of variance (ANOVA) with multiple comparisons followed by Tukey’s post hoc test. p ≤ 0.05 values were considered significant.