Gabapentinoids Suppress Lipopolysaccharide-Induced Interleukin-6 Production in Primary Cell Cultures of the Rat Spinal Dorsal Horn

Introduction: Gabapentin and pregabalin are drugs to treat neuropathic pain. Several studies highlighted effects on presynaptic terminals of nociceptors. Via binding to α2δ subunits of voltage-gated calcium channels, gabapentinoids modulate the synaptic transmission of nociceptive information. However, recent studies revealed further properties of these substances. Treatment with gabapentin or pregabalin in animal models of neuropathic pain resulted not only in reduced symptoms of hyperalgesia but also in an attenuated activation of glial cells and decreased production of pro-inflammatory mediators in the spinal dorsal horn. Methods: In the present study, we aimed to investigate the impact of gabapentinoids on the inflammatory response of spinal dorsal horn cells, applying the established model of neuro-glial primary cell cultures of the superficial dorsal horn (SDH). We studied effects of gabapentin and pregabalin on lipopolysaccharide (LPS)-induced cytokine release (bioassays), expression of inflammatory marker genes (RT-qPCR), activation of transcription factors (immunocytochemistry), and Ca2+ responses of SDH neurons to stimulation with substance P and glutamate (Ca2+-imaging). Results: We detected an attenuated LPS-induced expression and release of interleukin-6 by SDH cultures in the presence of gabapentinoids. In addition, a significant main effect of drug treatment was observed for mRNA expression of microsomal prostaglandin E synthase 1 and the inhibitor of nuclear factor kappa B. Nuclear translocation of inflammatory transcription factors in glial cells was not significantly affected by gabapentinoid treatment. Moreover, both substances did not modulate neuronal responses upon stimulation with substance P or glutamate. Conclusion: Our results provide evidence for anti-inflammatory capacities of gabapentinoids on the acute inflammatory response of SDH primary cultures upon LPS stimulation. Such effects may contribute to the pain-relieving effects of gabapentinoids.

studied effects of gabapentin and pregabalin on lipopolysaccharide (LPS)-induced cytokine release (bioassays), expression of inflammatory marker genes (RT-qPCR), activation of transcription factors (immunocytochemistry), and Ca 2+ responses of SDH neurons to stimulation with substance P and glutamate (Ca 2+ -imaging).Results: We detected an attenuated LPS-induced expression and release of interleukin-6 by SDH cultures in the presence of gabapentinoids.In addition, a significant main effect of drug treatment was observed for mRNA expression of microsomal prostaglandin E synthase 1 and the inhibitor of nuclear factor kappa B. Nuclear translocation of inflammatory transcription factors in glial cells was not significantly affected by gabapentinoid treatment.Moreover, both substances did not modulate neuronal responses upon stimulation with substance P or glutamate.Conclusion: Our results provide evidence for anti-inflammatory capacities of gabapentinoids on the acute inflammatory response of SDH primary cultures upon LPS stimulation.Such effects may contribute to the pain-relieving effects of gabapentinoids.

Introduction
Gabapentinoids, such as gabapentin and pregabalin, are drugs to treat symptoms of neuropathic pain [1].Initially, they were established as analogues of the inhibitory neurotransmitter gamma aminobutyric acid (GABA) [2].However, agonist-like effects for the GABA B receptor could not be proven [3].Instead, several studies provide evidence for a scenario, in which gabapentinoids bind to the α 2 δ subunits of voltage-gated calcium channels (Ca v ) and, therefore, modulate Ca 2+ currents at the presynaptic terminals of nociceptive neurons in the dorsal root ganglia (DRG) [4,5].In animal models of neuropathic pain, the α 2 δ-1 subunit of Ca v is upregulated in DRG neurons [6,7] and exhibits an enhanced plasma membrane expression [8].Intracellular trafficking of Ca v is attenuated by α 2 δ ligands, such as gabapentin or pregabalin [8,9].Together, these presynaptic effects may lead to an attenuated release of excitatory neurotransmitters, such as glutamate [10][11][12], and are one possible explanation for the anti-nociceptive properties of gabapentinoids.
More recent studies provide evidence that gabapentin and pregabalin not only influence neuronal excitability and neurotransmitter release, but further impact inflammatory processes in the spinal dorsal horn.In animal models of neuropathic pain, gabapentin or pregabalin attenuated the injury-induced production of pro-inflammatory cytokines, like tumor necrosis factor alpha (TNFα), interleukin (IL)-1β, or IL-6 [13][14][15].In other studies, this effect was accompanied by a reduced activation of glial cells, e.g., astrocytes or microglia [15][16][17].Interestingly, an impact of gabapentinoids on inflammatory processes was observed not only in the context of neuroinflammation and pain, but also in models of intestinal inflammation [18], ocular inflammation [19], carrageenan-induced peritonitis [20], or sepsis-related cardiotoxicity [21].These results indicate potential anti-inflammatory properties of gabapentinoids.
However, it remains unclear how gabapentinoids modulate (neuro-)inflammatory processes within the spinal cord and to which extent these effects contribute to the analgesic actions of these drugs.Glial cells within the superficial dorsal horn (SDH), such as astrocytes, oligodendrocytes, and microglial cells, play a key role in pathological pain states [22,23].Via release of pro-inflammatory mediators, they are able to modulate neuronal excitability and, therefore, the synaptic transmission of nociceptive information [24,25].We have recently shown that neuro-glial primary cell cultures of the rat SDH are a valuable tool to study effects of inflammatory stimulation on neuronal responses, production of inflammatory mediators, and activation of inflammatory transcription factors in vitro [26].Thus, this approach is also useful to study putative anti-inflammatory or anti-nociceptive capacities of novel therapeutic interventions, such as cellbased therapy [27].
In the present study, we aimed to investigate proposed anti-inflammatory properties of gabapentin or pregabalin applying the established model of SDH primary cultures.We present inhibitory effects of gabapentinoids on lipopolysaccharide (LPS)-induced IL-6 production.We further detected significant drug effects on relative expression of microsomal prostaglandin E synthase 1 (mP-GES-1) as well as the inhibitor of kappa B (IκB), a negative regulator of the nuclear factor kappa B (NFκB) cascade [28].However, LPS-induced nuclear translocation of transcription factors NFκB in microglial cells or signal transducer and activator of transcription 3 (STAT3) in astrocytes was not significantly affected by gabapentinoid treatment.Furthermore, we could not detect modulatory capacities of gabapentin or pregabalin on substance Pand glutamate-evoked neuronal Ca 2+ responses.

Animals
All animals used in experiments were obtained from an inhouse breeding colony.Parent animals originated from Charles River WIGA (Sulzfeld, Germany).Room temperature (22°C ± 1°C) and relative humidity (∼50%) were constantly monitored, and a daylight cycle of 12 h of light (7:00 a.m.-7:00 p.m.) with ad libitum access to standard laboratory chow and water was employed.Animal care, breeding, and experimental setup were performed according to the German Law on Animal Welfare, authorized by the Justus Liebig University Giessen (approval number GI 580_M), and registered by the regional authority of Hessia, Germany.

Preparation of SDH Primary Cell Cultures
Preparation of SDH primary cultures was performed as previously described in detail [26].On postnatal day 4-6, rat pups of either sex were sacrificed by decapitation.The vertebral column was dissected and transversally sliced into 1 mm sections.Using binoculars, the superficial parts of the dorsal horn, including laminae I and II (substantia gelatinosa), were extracted and collected in Hank's Balanced Salt Solution (HBSS, without Ca 2+ and Mg 2+ ; Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany).A total number of about 10-15 pairs of dorsal horns were obtained from one animal.Superficial parts of dorsal horns were transferred into an enzyme mix containing collagenase (CLS II, 2.5 mg/mL; Bio & Sell GmbH, Nuernberg, Germany) and dispase II (5 mg/mL; Sigma-Aldrich Chemie GmbH) diluted in HBSS and enzymatically digested for 40 min.The enzymes' activity was stopped using 1 mM EDTA in HBSS, and the remaining tissue was dissociated mechanically and resuspended in complete medium, consisting of Neuro-basal-A Medium supplemented with 2% B27, penicillin (100 U/ mL)/streptomycin (0.1 mg/mL), and 2 mM glutamine (all; Thermo Fisher Scientific GmbH, Langenselbold, Germany).Number of cells per milliliter of cell solution was calculated applying a Neubauer Improved Hemocytometer (NanoEntek, Seoul, South Korea) and equalized to a total number of 100,000 cells/mL by adding complete medium.The cell suspension was transferred into flexi-PERM © (micro12; Sarstedt AG & Co. KG, Nuembrecht, Germany) bounded chambers on poly-L-lysine (0.1 mg/mL; Bio & Sell GmbH) coated coverslips (Menzel, Braunschweig, Germany) with a volume of 350 μL ("wells").For Ca 2+ -imaging experiments, specific glass coverslips with a laser-edged grid were employed for later immunocytochemical cell-type classification (Fig. 4).Cells were cultivated for 4 h at 37°C in a humidified atmosphere of 5% CO 2 and 95% air.

Stimulation Protocols and Sample Collection
Four hours after preparation and cultivation of primary cell cultures, the medium was removed and complete medium containing phosphate-buffered saline (PBS; Capricorn Scientific GmbH, Ebsdorfergrund, Germany), gabapentin (100 μM; Sigma-Aldrich Chemie GmbH), or pregabalin (100 μM; Sigma-Aldrich Chemie GmbH) was added.Stock solutions of gabapentin and pregabalin were stored at −20°C (10 mM in PBS).After 18 h of preincubation, the media were exchanged by complete medium containing the respective drug and LPS (10 μg/mL, Escherichia coli O111:B4; Sigma-Aldrich Chemie GmbH), as inflammatory stimulus, or solvent in the same dilution (PBS).This procedure results in six treatment groups.All concentrations were chosen according to previous studies [26,29].
After performing the stimulation protocol, supernatants from all SDH primary cultures were collected and deep-frozen for subsequent cytokine measurements.Cultured cells were used for distinct methodological approaches.Per independent experiment, two wells per group were used for Ca 2+ -imaging experiments, while three wells per group were fixed in PFA (4%) to study transcription factor activation by means of immunocytochemistry.Cells from eight wells per group were lysed and pooled in 200 μL of lysis buffer for subsequent RNA extraction for RT-qPCR experiments.

Cytokine Measurements (TNFα, IL-6)
We applied specific bioassays to detect the rather low amounts of pro-inflammatory cytokines TNFα and IL-6 released by SDH primary cell cultures after two or four h of LPS stimulation.TNFα exerts a concentration-dependent cytotoxic effect on the fibrosarcoma cell line WEHI 164 subclone 13, which can be quantified using the dimethylthiazol-diphenyl tetrazolium bromide colorimetric assay [30].Applying an international standard (murine TNFα standard: code 88/532, National Institute for Biological Standards and Control [NIBSC], South Mimms, UK), the concentration of released TNFα can be calculated.To determine concentrations of released IL-6, we used a bioassay based on the concentration-dependent growth of B9 hybridoma cells [31].IL-6-dependent cell growth of B9 cells was quantified using the dimethylthiazol-diphenyl tetrazolium bromide assay, and concentrations of released IL-6 were calculated using an international standard (human IL-6 standard: code 89/548, NIBSC).The detection limits for both assays were proven to be 6.0 pg/mL for TNFα and 3.0 international units for IL-6.For detailed description of both assays, also see [32,33].

RT-qPCR
Four hours after LPS stimulation, supernatants of primary cultures were removed and cells were used for RNA extraction to determine relative expression of inflammatory target genes.Cells from eight wells of the same group were detached from glass cover slips and pooled in 200 μL of lysis buffer (RA1-buffer, NucleoSpin © RNA XS kit; Macherey-Nagel, Dueren, Germany).We extracted RNA applying the NucleoSpin © RNA XS kit (Macherey-Nagel) and adjusted RNA concentrations to 15 ng/μL.A total quantity of 120 ng of RNA was employed, and random hexamers (40 μM) and deoxynucleoside triphosphates (10 μM) were supplemented before denaturation at 65°C for 10 min.For subsequent reverse transcription (37°C, 60 min), RT-buffer, dithiothreitol (0.1 M), and murine leukemia virus reverse transcriptase (50 U) were added.Afterwards, the enzyme activity was abolished by heating to 90°C.Reverse transcription (including all substances) was performed according to the protocol provided by the manufacturer (Applied Biosystems, Foster City, CA, USA).DNA was stored at −20°C for subsequent RT-qPCR experiments.For relative quantification, the StepOnePlus Real-Time PCR System (Applied Biosystems) was used.All samples were investigated in triplicates applying a mixture of 1 μL cDNA with 0.5 μL primer (TaqMan © Gene Expression Assay; Applied Biosystems; see Table 1) and 5 μL TaqMan © Master Mix (Applied Biosystems).After polymerase activation (50°C for 2 min) and initial denaturation (95°C for 10 min), 40 cycles of denaturation (95°C for 15 s) and annealing/elongation (60°C for 1 min) were performed.The housekeeping gene GAPDH was implemented as reference after comparison of different possible housekeeping genes to normalize cDNA quantities.For relative quantification, the 2 −(ΔΔCt) method was applied.The presented results (Fig. 3) represent the x-fold increase in relation to a control with the lowest expression, given a value of 1.

Immunocytochemistry
We performed immunocytochemistry for cell-type identification after Ca 2+ -imaging experiments and to investigate activation of inflammatory transcription factors (NFκB, STAT3) in SDH glial cells.Therefore, cells were fixed in 4% paraformaldehyde in PBS (PFA; Sigma-Aldrich Chemie GmbH), pH 7.4, for 20 min and after three washing steps in PBS used for immunocytochemistry.To block unspecific binding sites, a blocking solution, containing 10% fetal calf serum (Capricorn Scientific GmbH) in PBS-T (0.05% Triton-X in PBS; Sigma-Aldrich Chemie GmbH), was applied for 2 h.Cells were incubated in the presence of primary monoclonal antibodies or polyclonal antisera diluted in blocking solution for 24 h at room temperature in a humidified atmosphere (see Table 2).
After three washing steps in PBS-T for 5 min, fluorophore-coupled secondary antisera diluted in PBS-T were added for 2 h (Table 2).Again, cells were washed three times in PBS-T.Cellular nuclei were stained with 2-(4-amidinophenyl)-1H-indole-6-carboxamidine (Thermo Fisher Scientific GmbH), and after three final washing steps, coverslips were embedded on microscope slides with a glycerol/PBS solution (Citifluor Ltd., London, UK).Immunoreactivity of SDH cells was examined using a fluorescence microscope (BX-50; Olympus Optical) equipped with the appropriate filter sets.Cells were photographed and analyzed with the MetaMorph microscopic imaging software (Molecular Devices, San Jose, CA, USA).To calculate nuclear intensities of a given transcription factor, nuclei of microglial cells (CD68-positive) or astrocytes (GFAP-positive) were marked as a region of interest and the mean signal intensity of the respective channel was quantified.Specificity of used antibodies was previously shown [26,28].

Ca 2+ -Imaging Experiments
After pre-incubation with PBS, gabapentin, or pregabalin and subsequent inflammatory stimulation with LPS or PBS as control, cells were incubated with 2 μM fura-2-AM (Thermo Fisher Scientific GmbH) in complete medium for 45 min in a humidified atmosphere of 5% CO 2 /95% air at 37°C.For Ca 2+ -imaging experiments, coverslips were transferred into a Teflon © chamber under an inverted microscope (IMT-2; Olympus GmbH, Hamburg, Germany).Cells were constantly superfused with Ca 2+ -imaging buffer consisting of 5 mM HEPES, 130 mM NaCl, 5 mM KCl, 1 mM MgCl 2 , 1.25 mM CaCl 2 , and 10 mMD-glucose (all; Sigma-Aldrich Chemie GmbH; pH 7.4) at a monitored temperature of 37°C and a flow rate of 2.0 mL/min.Fluorescence measurements were performed using a filter wheel-based excitation system and analyzed with the MetaFluor 7.7.8.0 software (Molecular Devices).Single cells were marked as region of interest (see Fig. 4b), and emitted fluorescence (>515 nm) was recorded over time using a Spot Pursuit digital CCD-camera (Model 23.0; Visitron GmbH) after alternating excitations at 340 and 380 nm.The 340/380 nm ratios are proportional to intracellular calcium concentrations [Ca 2+ ] i , and a stimulus-induced increase (Δratio 340/380 nm) represents a Ca 2+ response of the respective cell.Substance P (1 μM), glutamate (10 μM), and KCl (50 mM) were applied as excitatory stimuli for 180 s at constant temperature (37°C) and a superfusion rate of 2.0 mL/min.All doses were chosen according to preliminary studies [26,34].After Ca 2+ -imaging experiments, immunocytochemistry was performed to identify the investigated cells as neurons (MAP-positive cells).

Evaluation and Statistics
Cytokine Measurements Specific bioassays for TNFα and IL-6 were performed to measure the release of both cytokines upon stimulation with LPS or PBS after two or 4 h.Results originate from supernatants of 9-15 wells per group (Fig. 1: symbols in bars) of 3 (4 h) or 4 (2 h) independent experiments.

RT-qPCR
To detect changes in relative expression of inflammatory target genes upon LPS stimulation (4 h), cells from eight wells per group were pooled in 200 μL lysis buffer for subsequent RNA extraction, reverse transcription, and RT-qPCR.This procedure was repeated in three independent experiments (n = 3).

Immunocytochemistry
We investigated activation of inflammatory transcription factors, NFκB and STAT3, by means of immunocytochemistry.Nuclear signal intensities of the respective transcription factor were detected in single cells (CD68-positive microglia or GFAP-positive astrocytes) by selecting the nucleus as region of interest.We calculated the mean nuclear intensity of all investigated cells per well (∼50-100 cells/well) in relation to the control group (PBS-PBS), given a value of 1.The mean intensities per well are presented as symbols in Figure 3, while the bars represent the mean ± SEM from three independent experiments (n = 3).Ca 2+ -Imaging Effects of gabapentinoids and inflammatory stimulation on neuronal responses (substance P, glutamate, KCl) were examined by means of Ca 2+ -imaging.Per experiment, twelve wells were used for Ca 2+ -imaging experiments, resulting in two wells per group.In each well, ∼40-50 cells were examined.Overall, Ca 2+ responses of 1,320 neurons were calculated as stimulus-induced increase from baseline (Δratio 340/380 nm).A Δratio larger than 0.05 was assigned as a stimulus-induced Ca 2+ response.We determined the mean Δratio of all responsive neurons per well, presented as symbols in Figure 4. Bars in Figure 4 show the means ± SEM of six wells (n = 6).In all experiments, groups were compared using a two-way ANOVA and a Bonferroni post hoc test (# = main effect "treatment" [LPS], + = main effect "drug" [gabapentinoids], § = interaction, * = Bonferroni post hoc test).All columns represent the mean ± SEM with symbols presenting results of single wells from three independent experiments.For data analysis and graphic illustration, Microsoft Excel 2013 (Microsoft Corporation, Redmond, WA, USA) and Prism 5.0 and 7.0 (GraphPad Software, Inc., San Diego, CA, USA) software were applied.

Results
To evaluate effects of gabapentinoids on the inflammatory response of SDH primary cultures, cells were preincubated with PBS (control), gabapentin, or pregabalin and subsequently stimulated with PBS or LPS (10 μg/mL), resulting in six treatment groups in all experiments.We analyzed the release of inflammatory cytokines, TNFα and IL-6 (Fig. 1), relative expression of inflammatory target genes (Fig. 2), activation of inflammatory transcription factors (Fig. 3), and neuronal responsiveness upon stimulation with excitatory neurotransmitters, substance P and glutamate (Fig. 4).

LPS-Induced Activation of Inflammatory Transcription Factors
Inflammatory mediators, such as LPS, TNFα, or IL-6, activate several intracellular signaling cascades in distinct cellular populations, expressing the respective receptor (e.g., TLR-4, TNFR-1/2, gp130).Activation of a given transcription factor leads to its phosphorylation and translocation into the nucleus that can be visualized by means of immunocytochemistry [46].In the present study, we investigated activation of two important inflammatory transcription factors (NFκB and STAT3) that have previously been shown to be activated during spinal neuroinflammatory processes [47][48][49][50].For SDH primary cell cultures, we previously observed that LPS-induced NFκB translocation occurs predominantly in microglial cells, while STAT3 is activated in astrocytes [26,34].
Here, we aimed to examine effects of gabapentinoid treatment on LPS-induced activation of NFκB in spinal microglia (CD68-positive) and STAT3 in astrocytes (GFAP-positive) (Fig. 3).No morphological changes upon LPS or gabapentinoid treatment were detectable in these cells.

Effects of Gabapentinoids on Ca 2+ Responses in SDH Neurons
In a series of Ca 2+ -imaging experiments, we aimed to investigate effects of inflammatory stimulation and/or treatment with gabapentinoids on the neuronal responses of SDH neurons upon acute stimulation with neurotransmitter substance P and glutamate.Activation of primary nociceptors by noxious stimuli (e.g., heat or TRPV1-agonist capsaicin) results in a release of excitatory neurotransmitters by central nerve endings in the superficial laminae (I & II) of the spinal dorsal horn [51].This synaptic transmission of nociceptive information in the SDH can be modulated in states of spinal neuroinflammation [52,53] and in SDH primary cultures upon inflammatory stimulation [26].We, therefore, performed Ca 2+ -imaging experiments of SDH cultures after pre-incubation with gabapentinoids and subsequent inflammatory stimulation (LPS, 4 h).Selecting single cells as a region of interest, Ca 2+ responses upon stimulation with substance P and glutamate were recorded and analyzed.In Figure 4a, Ca 2+ responses of an exemplary cell, shown in Figures 4b and c, are presented.Neurons were identified by their responsiveness to a KCl stimulus (Fig. 4a) or by their immunoreactivity for the neuronal marker MAP2a+b (Fig. 4c).The mean stimulus-induced increase in [Ca 2+ ] i was calculated for all responsive neurons of six wells (n = 6).Overall, we investigated 1,320 KCl-responsive neurons, of which 695 also responded to substance P (∼53%) and 1,182 to glutamate (∼90%).Basal levels of intracellular Ca 2+ prior to any stimulation were not affected by pre-incubation with gabapentinoids or LPS stimulation (data not shown).Moreover, no Ca 2+ responses upon direct application of gabapentin or pregabalin were detectable.Co-stimulation of gabapentinoids together with excitatory stimuli (glutamate, KCl) did not alter stimulus-induced Ca 2+ responses in neurons (data not shown).Neither pre-incubation with gabapentin or pregabalin, nor stimulation with LPS significantly altered neuronal responses (Fig. 4d-f): substance P (Fig. 4d): PBS: PBS 0.163 ± 0.015 versus LPS 0.153 ± 0.012; Gaba: PBS 0.164 ± 0.012 versus LPS 0.174 ± 0.019; Preg: PBS 0.188 ± 0.025 versus LPS 0.165 ± 0.015; glutamate (Fig. 4e): PBS: PBS 0.148 ± 0.021 versus LPS 0.151 ± 0.013; Gaba: PBS 0.154 ± 0.013 versus LPS 0.153 ± 0.014; Preg: PBS 0.161 ± 0.026 versus LPS 0.136 ± 0.011; KCl (Fig. 4f): PBS: PBS 0.241 ± 0.029 versus LPS 0.240 ± 0.028; Gaba: PBS 0.236 ± 0.039 versus LPS 0.237 ± 0.019; Preg: PBS 0.279 ± 0.052 versus LPS 0.196 ± 0.011.

Discussion
The present study demonstrates that inflammatory processes in SDH primary cultures upon stimulation with LPS are attenuated in the presence of gabapentinoids, with the most eminent impact on IL-6 production.We further detected main effects of drug treatment on expression of mPGES-1 and IκB.However, LPS-induced activation of transcription factors in glial cells, as well as neuronal Ca 2+ responses upon stimulation with substance P or glutamate, was not significantly affected by gabapentinoids in our experiments.

Spinal Neuroinflammatory Processes in vivo and in vitro
Spinal neuroinflammatory processes are associated with symptoms of inflammatory and neuropathic pain [24,25].Via enhanced immune cell infiltration and glial activation (microgliosis, astrocytosis), the expression of several inflammatory mediators, such as cytokines, growth factors, or prostaglandins, is upregulated [22][23][24].Excitability of spinal dorsal horn neurons is modulated under inflammatory conditions, e.g., by intrathecal injection of LPS [38,52], TNFα [40,54], or IL-6 [39].Inhibition or blockade of pro-inflammatory mediators can effectively attenuate these effects and, thereby, reduce symptoms of hyperalgesia and allodynia [38,55,56].Therefore, attenuation of spinal neuroinflammatory processes is an attractive target for pharmaceutical interventions to treat symptoms of chronic pain.
Our recent studies provided evidence that primary cell cultures from structures of the peripheral [29,57] and central nervous system [26,34,58] represent useful tools to study effects of glia-neuron interactions under inflammatory conditions.In SDH primary cultures, stimulation with LPS resulted not only in an upregulation of pro-inflammatory cytokines, but also in an activation of microglial cells and astrocytes, indicated by nuclear translocation of inflammatory transcription factors NFκB or STAT3 [26].Moreover, neuronal Ca 2+ responses upon application of excitatory neurotransmitters (glutamate, substance P) were modulated by LPS stimulation.Based on these results, we showed that this in vitro DOI: 10.1159/000525657 tool can further be applied to study pathophysiological mechanisms including LPS tolerance [34], as well as immunomodulatory capacities of novel treatments such as co-cultivation with adipose tissue-derived medicinal signaling cells [27].
In the present study, we applied the established model of LPS stimulation in SDH primary cultures to examine suggested anti-inflammatory effects of gabapentin or pregabalin on the inflammatory response.In line with previously published data [26], we detected significant LPS effects on cytokine release (Fig. 1), expression of inflammatory mediators (Fig. 2), and activation of inflammatory transcription factors NFκB in microglial cells and STAT3 in astrocytes (Fig. 3).However, neuronal responses upon stimulation with excitatory neurotransmitters, substance P, and glutamate were not altered under inflammatory conditions (Fig. 4).

Anti-Inflammatory Capacities of Gabapentin and Pregabalin May Support Analgesic Effects
The modes of action responsible for the analgesic effects of gabapentinoids are still not completely identified.Initially, gabapentin was developed as an analogue of the endogenous inhibitory neurotransmitter GABA [2].However, GABA does not only act as inhibitor of synaptic transmission in the spinal dorsal horn, but can also modulate inflammatory processes via GABA receptors on microglial cells [59] and astrocytes [60].Thus, GABAergic function of gabapentin may also modulate spinal neuroinflammation.Indeed, several studies indicate that gabapentinoids do not directly interact with GABA receptors [3,61].Observed analgesic effects of gabapentin and pregabalin are mainly associated with their action on the α2δ-1 subunit of voltage-gated Ca 2+ channels at the presynaptic terminals of nociceptors and, thereby, an attenuated release of excitatory neurotransmitters [4,5].Compared to gabapentin, pregabalin showed improved pharmacokinetic and pharmacodynamic functions and overall enhanced potency at lower dosages [62].However, several studies implicate further capacities of gabapentinoids on inflammatory processes.In the context of spinal neuroinflammation, other researchers observed anti-allodynic effects accompanied by a reduced expression of inflammatory cytokines in the spinal cord (IL-6, IL1β, TNFα) upon gabapentin treatment over seven or 14 days in models of neuropathic pain such as spinal nerve ligation [13] or chronic constriction injury [15].It remains a matter of debate, how such immunomodulatory effects are mediated, and studies approaching the underlying mechanisms are rare.
In line with these anti-inflammatory effects observed in animal models of neuropathic pain, the presented results provide evidence for an impact of gabapentin and pregabalin on spinal inflammatory processes.In our experiments, drug effects were detectable in an early phase of the inflammatory response, 4 h after LPS stimulation, and mainly for IL-6, while TNFα expression and release were not impaired by treatment with gabapentin or pregabalin (Fig. 1, 2).In a comparable study in DRG primary cultures, we observed similar effects of gabapentinoid treatment on LPS-induced expression of IL-6, but not TNFα, 2 h after stimulation [29].In this initial phase of inflammatory stimulation, microglial cells are activated by LPS via the TLR4-NFκB cascade and represent the main source of pro-inflammatory mediators, e.g., TNFα and IL-1β [63].The presented data do not indicate a direct impact of gabapentinoids on this fast LPS-induced microglial activation, since nuclear translocation of NFκB (Fig. 3a) and expression of microglial-derived mediators, such as TNFα and IL-1β, are not affected by drug treatment.However, production of IL-6 is not restricted to microglia, but neurons can also be a source for IL-6.Indeed, several studies implicate expression of IL-6 by neurons from DRG [64], spinal cord [65,66], hippocampus [67], or cortex [68].It could be hypothesized that gabapentinoid action on voltage-gated calcium channels alters intracellular signaling in neurons and, thereby, neuronal production of IL-6.Some of the observed effects can be explained as secondary to released IL-6.Nuclear translocation of STAT3 in astrocytes is induced by IL-6 [69][70][71][72], and the JAK-STAT pathway is an important intracellular signaling cascade involved in pathological states of pain [49,55].In a model of spinal nerve injury, symptoms of neuropathic pain were accompanied by spinal astrogliosis and an enhanced nuclear translocation of STAT3 in dorsal horn astrocytes at the injury site, but not the contralateral site or in control animals [50].In that study, inhibition of STAT3 activation also attenuated symptoms of neuropathic pain, highlighting the role of astrocytic STAT3 activation.Nuclear translocation of STAT3 in astrocytes is also observed in SDH primary cultures upon inflammatory stimulation [26,34], but was not affected by pre-incubation with gabapentin or pregabalin (Fig. 3b).One study investigating the effects of gabapentinoids on paclitaxel-induced peripheral neuropathy showed ameliorative effects of pregabalin treatment on thermal hyperalgesia that was accompanied by an attenuated expression of pro-inflammatory cytokines (TNFα, IL-6) and STAT3 after 21 days [14].However, an altered protein expression of STAT3 does not necessarily translate in an Neuroimmunomodulation 2023;30:1-14 DOI: 10.1159/000525657 altered activation of the transcription factor as implicated by its nuclear translocation.We suggest that the released bioactive IL-6 by LPS-stimulated SDH primary cultures, even in the gabapentinoid-treated groups, is still sufficient to activate STAT3 in astrocytes and that measurement of fluorescent intensities may not be precise enough to detect rather small effects of gabapentin-or pregabalin-induced modulations of a strong LPS-induced nuclear STAT3 translocation.Via the STAT3-cascade, IL-6 modulates transcription of several target genes, including enzymes involved in the synthesis of PGE2, namely, COX-2 and mPGES-1 [35,36].While the LPS-induced increase of COX-2 expression was not significantly affected by pre-incubation with gabapentin and pregabalin (Fig. 2d), we detected a significant drug effect for mPG-ES-1 (Fig. 2e).This effect was more pronounced in pregabalin-treated groups.Attenuation of PGE2-synthesis may contribute to anti-inflammatory and analgesic effects, since prostaglandins are potent modulators of neuronal activity by either direct activation of neurons [26,29,33], modulation of responses upon stimulation with excitatory neurotransmitters [33], or altering firing rates of spinal neurons upon noxious stimulation [73].

Effects of Gabapentinoids on Neuronal Ca 2+ Responses
To study effects of gabapentin and pregabalin on Ca 2+ responses of SDH neurons, we performed a series of Ca 2+imaging experiments.We did not detect changes of [Ca 2+ ] i in SDH neurons in response to acute application of gabapentin or pregabalin.Moreover, neuronal responses upon excitatory stimulation were not affected by co-stimulation of cells with glutamate or KCl together with gabapentin or pregabalin.Applying the protocol of a pre-incubation with both drugs and subsequent inflammatory stimulation with LPS, no significant effects of gabapentin or pregabalin were detectable.Substance P and glutamate are two important neurotransmitters released by central nerve endings of nociceptors and acting on the respective receptors on spinal dorsal horn neurons [51].Inflammatory stimuli, e.g., LPS or cytokines, result in enhanced neuronal excitability, which is causative for symptoms of hyperalgesia [39,40,52,54].Some of these effects are related to enhanced NMDA-induced currents, highlighting the role of glutamatergic transmission [74].In SDH primary cultures, we previously observed effects of LPS stimulation on neuronal responsiveness [26].However, this effect was dose-and time-dependent.While a short-term stimulation (2 h) with a moderate dose of LPS (10 μg/mL) attenuated substance P responses, a prolonged stimulation (24 h) with a low dose (0.01 μg/mL) resulted in increased glutamate responses.In the present study, we performed Ca 2+ -imaging experiments after 4 h of LPS stimulation (10 μg/mL) according to the protocols applied for cytokine measurements (Fig. 1), RT-qPCR (Fig. 2), and immunocytochemistry (Fig. 3).At this point, gabapentin or pregabalin treatment did not affect Ca 2+ responses of SDH neurons.These results may further support the idea of a rather presynaptic than postsynaptic effect of gabapentinoids on neurons [75,76].

Limitations and Outlook
Overall, our results provide evidence for anti-inflammatory capacities of gabapentin and pregabalin in an initial phase of LPS-induced inflammation in SDH primary cell cultures, with the most eminent effect on IL-6 production.These observations are in line with previous reports in animal models of neuropathic pain [13][14][15][16][17] as well as other models of inflammation [18][19][20][21].It remains to be elucidated, how these effects are mediated on a cellular level, and to what extent they contribute to the analgesic effects of gabapentinoids.Further studies, investigating the underlying mechanisms of gabapentinoids' immunomodulatory capacities, may focus on cell-typespecific activation of intracellular signaling cascades and cytokine expression, e.g., by specific inhibition or activation of distinct receptors and pathways.Primary cell cultures can be a useful tool for more extensive studies on a cellular level and may, thereby, contribute to a better understanding of gabapentinoids' action within the spinal dorsal horn.At the same time, the use of such in vitro models may contribute to a reduction of the number of experimental animals used for scientific purposes according to the 3R principle [77].

Fig. 1 .
Fig.1.LPS-induced release of pro-inflammatory cytokines TNFα and IL-6 after 2 or 4 h of stimulation.SDH primary cultures were pre-incubated with PBS, gabapentin (Gaba), or pregabalin (Preg) and subsequently stimulated with PBS (white bars) or LPS (gray bars) and the respective drug for 2 or 4 h.Supernatants of SDH cultures were investigated by means of specific bioassays for TNFα (a) and IL-6 (b).Stimulation with LPS resulted in a highly significant main effect "treatment" ( ### p < 0.0001) for both cytokines and both time points (a + b).After 4 h of stimulation, a significant main effect "drug" ( + p = 0.017) was observed for IL-6 release with a significant interaction between both effects ( § p = 0.0167).LPS-induced IL-6 production was significantly suppressed in the presence of pregabalin (***p < 0.001).Bars represent the mean ± SEM with symbols presenting results of 9-15 samples from 3 (4 h) or 4 (2 h) independent experiments.

Fig. 2 .
Fig. 2. Effects of gabapentinoid treatment on the expression of inflammatory target genes after 4 h of LPS stimulation.SDH cells from eight primary cultures per group were lysed after pretreatment with PBS, gabapentin (Gaba), or pregabalin (Preg) and inflammatory stimulation for 4 h (LPS: gray bars vs. PBS: white bars).After extraction of mRNA and reverse transcription, RT-qPCR was performed to evaluate effects of gabapentinoids on LPS-induced expression of inflammatory target genes.a-f Stimulation with LPS resulted in a highly significant main effect "treatment"

Fig. 3 .
Fig.3.LPS-induced activation of inflammatory transcription factors NFκB in microglial cells and STAT3 in astrocytes is not modulated by treatment with gabapentinoids.After stimulation with LPS or PBS for 4 h, cells were fixed in PFA (4%) and used for subsequent immunocytochemical detection of inflammatory transcription factors NFκB and STAT3.LPS-induced nuclear translocation of NFκB was observed in CD68-positive microglial cells (a1 + a2), while STAT3 was mainly detectable in GFAP-positive astrocytes (b1 + b2).For relative quantification, we calculated the mean signal intensity of the respective channel (transcription factor; red) within the area of the nucleus (2-(4-amidinophenyl)-1H-indole-6-carboxamidine; blue) in distinct cellular populations (CD68 or GFAP; green).The mean intensity of all investigated cells of one experiment was calculated and is presented in relation to the respective PBS control group, given a value of 1. Inflammatory stimulation resulted in a highly significant main effect "treatment" ( ## p < 0.01) in all groups (a3 + b3).Pre-incubation with gabapentin (Gaba) or pregabalin (Preg) did not significantly modulate LPSinduced activation of transcription factors.Scale bar in b2 represents 10 μm applicable for all images.Bars in a3 and b3 show the mean ± SEM with symbols presenting single results of three distinct experiments.

Fig. 4 .
Fig.4.Ca 2+ responses of SDH neurons upon stimulation with excitatory neurotransmitters (substance P, glutamate).After the stimulation protocol of 18 h of pre-incubation with PBS, gabapentin (Gaba), or pregabalin (Preg) and subsequent inflammatory stimulation, SDH primary cultures were loaded with FURA-2AM (B) for Ca 2+ -imaging experiments.The ratio [340/380 nm] was calculated over time (a), and substance P (SP) and glutamate (Glut) were applied as excitatory stimuli.KCl was used as vitality test, and only neurons with a clear KCl-induced increase were included for later analysis (d-f).An increase in the ratio [340/380 nm] upon stimulation with the respective substance (Δratio) is regarded as stimulus-induced increase in the intracellular Ca 2+ level [Ca 2+ ] i .a-c Show one exemplary cell examined in Ca2+ -imaging experiments with a presenting the ratio [340/380 nm] over time of one neuron with clear responses to stimulation with SP, Glut, and KCl and a picture of the same FURA-loaded neuron at 340 nm (b), as well as the respective cell after subsequent immunocytochemistry for neuronal marker MAP.d-f The mean Δratios of all responsive cells (substance P, glutamate, KCl) per well were calculated and analyzed.However, no significant effect of inflammatory stimulation or pre-incubation with gabapentin or pregabalin was detectable.Bars represent the mean ± SEM of six wells from three independent experiments.Scale bar in c represents 10 μm and is applicable to b.