Matrix metalloproteinases, purinergic signaling, and epigenetics, hubs in the spinal neuroglial network following nerve injury.

The neuroglial network characterizes synaptic transmission and accounts for both cellular elements (neurons and glia) and neural extracellular matrix (nECM) roles. Glial cells, neuron, and nECMnetwork is strongly interconnected, in physiological and pathological conditions as shownin several neurodegenerative diseases. Purinergic activation and matrix metalloproteinases (MMPs) remodeling of the spinal cord is pivotal in maladaptive plastic changes following peripheral nerve injury (PNI). To understand how purinergic and MMPs inhibition may modulate and potentially reverse the neuroglial network failure, we used the spared nerve injury (SNI) model of the sciatic nerve. Molecular and morphological analysis of astrocytic and microglial activation, purinergic and neurotrophic receptors, Histone Deacetylase (HDAC)1, HDAC2 were analyzed to dene the pathways in response to the purinergic and MMPs inhibition. The data suggest complex protein interconnections, which are not passively inuenced by epigenetics but actively contribute to modify the transcriptomics machinery. The present study contributes to unveiling the spinal network consistency and ultimately encourages new paths for targeted treatments in neurological diseases with benets of neuroprotection, plasticity, and functional recovery.


Introduction
Glial cells are a well-recognized component of the synaptic complex, together with neurons and the surrounding matrix, constituting a more exhaustive model of synapse than the "classic" neuron-based structure, both in physiology [1] and pathology [2]. The matrix-neuroglial network response to noxious stimuli, early, involves both glial cells and the neural extracellular matrix (nECM), the matrisome [3]. glial cells overexpress purinergic receptors during pathological conditions, which affect the in ammasome and the lesion remodeling in combination with growth factors [11,12].
Further, MMPs are a pivotal enzyme family for nECM formation, adaptability, and pathological remodeling and recent data suggest a key role of these proteins in CNS diseases [13]. MMPs are secreted as pro-enzymes (pro-MMPs) by various cell types, including CNS residents. Afterward, pro-MMPs are activated by proteolytic cleavage and eventually blocked by tissue inhibitors of MMPs (TIMPs) [14].
Based on the data literature and our previous research, in the present work, we intend to deepen how P2XRs and MMPs are involved in the spinal cord plasticity following spared nerve injury (SNI) of the sciatic nerve, and if they share a modulatory impact. We inject intraperitoneally (i.p.) the oxidized ATP (OxATP), a nonselective P2XRs antagonist, alone or in combination with intrathecal (i.t.) infusion of GM6001 (a broad spectrum MMPs inhibitor). We focused our study on the expression of markers of astrogliosis, as glial brillary acidic protein (GFAP), the microglia and macrophages activation, as the ionized calcium-binding adaptor molecule 1 (Iba1), the purinergic (P2X4) and neurotrophic (TrkA and p75) receptors. However, MMPs and P2XRs modulation is not restricted to the extracellular and cytoplasmic level, more, it can in uence gene transcription processes, altering chromatin accessibility.
Histone deacetylases (HDACs) and DNA methylation could both regulate and in turn be affected by MMPs and P2XRs activation [15,16].
The present study analyzes molecular hubsin the spinal cord following nerve injury and modulation of MMPs and P2XRs, increasing our level of understanding of neuroglial networking between the nECM, channel receptors, up to the chromatin organization level.

Animals
Adult (250-300 g; Charles River, Calco, Italy) Sprague Dawley rats (n = 58) were used. Animal care was in compliance with the Italian (D.L. 116/92) and EC (O.J. of E.C. L358/1 18/12/86) regulations on the care of laboratory animals.Each animal was allowed free access to food and water, under a 12/12 h light/dark cycle. Animals were kept in pathogen-free iron-sheet cages with solid oor covered with 4-6 cm of sawdust. We did not use cages with thin-plate oors to avoid exacerbation of the discomfort from the affected hind paw [17].

SNI Model
Sciatic spared nerve injury (SNI) was made according to Decosterd and Woolf[18]. Brie y, each rat (n = 58) was anesthetized with chlorohydratetiletamine (30 mg/kg) during surgery. The sciatic nerve and its three terminal branches (the sural, common peroneal, and tibial nerves) were exposed on the lateral surface of the thigh. The SNI procedure comprised axotomy and tight ligation of the tibial and common peroneal nerves leaving the sural nerve intact. For the sham-operated control (CTR) group (n = 10), nerves were exposed but not truncated. Muscle and skin were closed in layers. Great care was taken to avoid any contact with or stretching of the intact sural nerve.

Drug Delivery
OxATP treatment was performed according to our protocol[6]: from the day after surgery, animalswere treated with i.p. OxATP (6 mg/kg), dissolved in 100 µl of sterile distilled water (dH2O) or dH20 alone.
The intrathecal lumbar spinal catheter for GM6001 or ACSF administration was positioned during SNI procedure to reduce the discomfort bias, according to Decosterd and Woolf[18]. Brie y, a small opening was made through the laminas of the lumbar spine and a catheter [polyethylene (PE) 10 tubing attached to PE 60 tubing for connection to an osmotic pump] was inserted into the subarachnoid space and directed to the lumbar enlargement of the spinal cord. After anchoring the catheter across the careful apposition of a glass ionomer luting cement triple pack (KetacCem radiopaque; 3 M ESPE, Seefeld, Germany), the wound was irrigated with saline and closed in two layers with 3-0 silk (fascial plane) and surgical skin staples. On recovery from surgery, lower body paralysis was induced by intrathecal lidocaine (2%) injection to con rm proper catheter localization. Each rat was placed on a table, and the gait and posture of the affected hind paw were carefully observed for 2 min. Only animals exhibiting appropriate, transient paralysis to lidocaine, as well as lack of motor de cits, were used for treatments; the free extremity of the catheter was connected to an osmotic minipump and the pump was implanted subcutaneously. Osmotic pumps attached to intrathecal lumbar spinal catheters were lled with GM6001 (VII -VIII) D3-D8control (CTR) groups (n = 10), sham-operated animals.

Tissue Preparation
Rats were deeply anesthetized on day 3 and 8 with an i.p. injection of chloral hydrate (300 mg/kg body weight) and perfused transcardially with saline solution (TrisHCl 0.1 M/ EDTA 6 mM). Spinal cords for western blotting (n = 29) were removed. Spinal cord for immunohistochemistry(n = 29) continued for xation by 4% paraformaldehyde added to 0.1% glutaraldehyde in 0.01 M phosphate-buffered saline (PBS), pH 7.4 at 4°C. They were removed and post xed 2 h in the same xative, then soaked in 30% sucrose phosphate-buffered saline (PBS), and frozen in chilled isopentane on dry ice, as previously described [19]. Serial sections were cut at the slide microtome (25 µm thickness) and collected in cold PBS.

Measurements and Statistical Analysis
Measurements of markers in the whole lumbar spinal cord were accomplished using computer-assisted image analysis system (MCID 7.1; Imaging Res. Inc., Canada). All data were collected in a blinded manner; the observer making the measurements was not aware of the group. Data were exported and converted to a density distribution histogram using the Sigma-Plot 10.0 program (SPSS-Erkrath) and presented as the mean ± SEM for all quantitative analyses. Data were checked for normal distribution and homogeneity of variance by the Kolmogorov-Smirnov's and Levene's meantests, respectively. Normal distributions with equal variances were statistically analyzedby using one-way ANOVA for multiple comparisons followed by all pairwise Holm-Sidak post hoc test. The level of signi cance was always set at p = 0.05 (*p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001).
Individual images were assembled, and the same adjustments were made for brightness, contrast, and sharpness using Adobe Photoshop (Adobe Systems).

Results
Reactive glial populations are selectively modulated by the purinergic inhibitor in combination with the metalloproteinases block.
Taken together these data showed that while damage-related astrocytosis is demodulated by purinergic inhibition without a clear contribution of GM6001 co-administration, the microglial reaction was affected by MMPs and P2XRs simultaneous inhibition.
Class I HDACs are differentially affected by nerve injury and purinergic signaling inhibition, even combined with MMPs modulation.
OxATP and GM6001 treatments successfully downregulate the expression of the purinergic receptors.
We considered P2X4R, as one of the most expressed receptors in both CNS and peripheral nerves, especially on microglia, to drive in ammation in pathological situations such as spinal cord damage [12,23]. We found that P2X4R was increased in D8ACSF group (0.798 ± 0.04, p ≤ 0,001) compared to the sham-operated one (D8 CTR 0.350 ± 0.06), with a signi cant decrease of the expression in the two groups of D8treatments (OxATP 0.586 ± 0.04, p = 0.008;OxATP + GM6001 0.524 ± 0.05,p = 0.002) (Fig. 3a-b). These data provide further support to the anti-in ammatory effect of the purinergic inhibition, alsoin combination with MMPs modulation.
Neurotrophin receptors are differentially modulated by nerve injury,and purinergic signaling inhibition, combined with MMPs modulation.
Neurotrophic growth factorsplay a key role in the spinal cord remodeling after lesion, along with purinergic receptors and MMPs [11,8].
However, both the treatments induced a sharp upregulation of the expression level (D8 OxATP 2.001 ± 0.108, p = 0.024; D8 OxATP + GM6001 2.891 ± 0.190, p ≤ 0,001), con rmingWB results.These ndings suggestan opposite role played by the high and low-a nity NGF receptors during the spinal neuroglial remodeling following nerve injury and the differential modulation of the purinergic and MMPs systems.

Discussion
Substantial modi cations occur in the nervous system following nerve injury, depending on the cell features, type of stimulus, and time [25,8,26]. However, interactions between glial cells and neurons through the neural extracellular matrix drive the pathological changes after the nerve damage [27]. Anti-in ammatory and antigliotic activityproduced by OxATP and GM6001 treatment has beenpreviouslyreportedby our group [30,6,31,8], hence, examining the expression of the pivotal protein in the neuro-in ammatory network, we aim to nd one or more keystones that could support the interaction between the matrisome and the channel receptors inhibition. We reported that OxATP reduced GFAP and Iba1 expression, diminishing astrocytosis and microglial activation. However, an interesting outcome is given by the relevant reduction of microglial reaction in the D8by the administration of both OxATP and GM6001. Molecular and morphological analyses have shown that Iba1 is downregulated by inhibiting the purinergic signaling, while GM6001 co-administration seems to further reduce Iba1 expression and partially restore microglial morphology after one week of treatment (Fig. 1). This effect may be due to the contemporary inhibition of intracellular Ca 2+ in ux (mediated by OxATP) and the blockage of nECM maladaptive remodeling (mediated by GM6001), having a selective effect on microglial cells rather than astrocytes.
Activated microglia has been reported to speci cally upregulate P2X4R, which drives the in ammation in neuropathic pain models [32,33].We demonstrated a decrease of P2X4R with OxATP and a possible additional effect given by MMPs inhibition (Fig. 2), supporting a pathway connection between purinergic receptor expression and the nECM modulation, to be in future fully elucidated.
Cells'reaction to insult and in ammatory response involves epigenetic changes. Histones deacetylases or demethylases of lysine residues regulate chromatin access to transcriptional factors with repression or activation of speci c genes [34]. While HDAC1 protein expression was not modi ed by the nerve injury, HDAC2 was upregulated, suggesting that class I HDACs are not functionally equivalent [35]. The spinal upregulation of HDAC2 following SNI was consistently associated with astrocytes and may be involved in the downregulation of glial glutamate transporters GLAST and GLT-1, causing the glutamate spillover in the synaptic cleft and the synaptic dysfunction [6,36]. Moreover, glutamate metabolism abnormalities have been shown in MECP2 (methyl-CpG-binding protein 2) null astrocytes, a transcriptional co-repressor thatcan bind HDACs [37]. HDAC2 expression was downregulated by purinergic antagonism,as also occurred after the MMPs modulation [8], and theconcomitant administration of the two drugs produceda higher effect, reverting the increase reported in the D8 ACSF group (Fig. 2) Neurotrophic support is determinant for the suitable activity of the neural circuitry, especially after injuries [7,39], while MMPs inhibition changes neurotrophins availability and their activation or degradation [31], P2XRs modulation could prevent MMPs activation or affect precursor neurotransmitter release and nerve growth factor signaling [40,41,42].
The NGF precursor proNGF is proteolyzed by plasmin tomature NGF and then inactivated by MMPs. In previous studies, we demonstrated an NGF enrichment of the ECM by administering GM6001 [5,31].We studied the NGFreceptors, high-a nity TrkA, and low-a nity p75 expression ( Fig. 2-3), andwe found an opposite trend concerning the expression of the receptors. The upregulation of TrkA following SNI was foundas previously reported [24,8].OxATP treatment alone and in combination with GM6001 reduced TrkA expression while determining p75 upregulation. TrkA decrease could be adaptive to the NGF matrisome enrichment, while p75 can generate pro-survival or pro-apoptotic signals depending on co-receptors and different intracellular pathways in the neuronal and glial cell types [43,7]. Therefore, we previously showed that a p75 upregulation in the dorsal horn was associated with a reduction of neuropathic behavior as well as adaptive plasticity modi cation to the PNI [24].
This work analyzed the expression of key molecules in the matrix-neuroglial network,contributing to de ne the impact ofthe purinergic and MMPs inhibition in the spinal cord following PNI. The resulting data suggest complex protein interconnections, which are not passively in uenced by epigenetics but actively contribute to modify the transcriptomics machinery. Purinergic and MMPs inhibitory modi cations affected cellular/matrix balance at various levels, suggesting a connection between the extracellular matrix and the chromatin rearrangement.The relations between P2XRs, MMPs, and epigenetics need to be further studiedin future experiments. These may help to identify targeted treatment in neurological diseases with bene ts for neuroprotection, spinal plasticity, and functional recovery. Authors' contributions All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Ciro De Luca, Assunta Virtuoso, and Michele Papa. The rst draft of the manuscript was written by Ciro De Luca and Assunta Virtuoso. All authors commented on previous versions of the manuscript. All authors read and approved the nal manuscript. Reactive gliosis is reducedby purinergic inhibition, even in combination with MMPs modulation. a Low and high magni cation representative images of immuno uorescence staining for Iba1 and GFAP in the dorsal horn of lumbar spinal cord in sham-operated (CTR) and SNI animals treated with ACSF or OxATP or OxATP+GM6001(GM) for 8 days.(Scale bar: 50 µm). b-c Densitometric quantitation of Iba1 and GFAP uorescence versus the ACSF group. Data are shown as the mean value of the data distribution (bars) ± SEM, n=4/group (*p≤0.05; **p≤0.01; ***p≤0.001; One way ANOVA followed by post hoc Holm-Sidak correctionfor multiple pairwise comparisons).

Figure 2
Reactive gliosis and epigenetic targets are affected by OxATP chronic treatment or in combination with GM6001. a Representative immunoblot of Iba 1, GFAP, and HDAC2 in the lumbar spinal cord from shamoperated (CTR) and SNI animals treated with ACSF or OxATP or OxATP+GM6001(GM) at D3 or D8. b-d Corresponding quantitation of Iba1, GFAP, and HDAC2 immunoblotsversus the ACSF group. Normalized data relative to the β-actin content are presented as the mean value of the data distribution (bars) ± SEM, n=4-5/group (*p≤0.05; **p≤0.01; ***p≤0.001; One way ANOVA followed by post hoc Holm-Sidak correction for multiple pairwise comparisons).

Figure 4
Neurotrophic receptors are differentially modulated by OxATP, also in combination with GM6001. a Representative image of immuno uorescence staining for TrkA and p75 in the dorsal horn of lumbar spinal cord in CTR rats, after SNI and OxATP treatment with or without GM6001 (GM) for 8 days. (Scale bar: 100 µm).b-c Densitometric quantitation of TrkA and p75 uorescence versus the ACSF group. Data are shown as the mean value of the data distribution (bars) ± SEM, n=4/group (*p≤0.05; **p≤0.01; ***p≤0.001; One way ANOVA followed by post hoc Holm-Sidak correction for multiple pairwise comparisons).