Harpagophytum procumbens Extract Ameliorates Allodynia and Modulates Oxidative and Antioxidant Stress Pathways in a Rat Model of Spinal Cord Injury

Abstract

Spinal cord injury (SCI) is a deliberating disorder with impairments in locomotor deficits and incapacitating sensory abnormalities. Harpagophytum procumbens (Hp) is a botanical widely used for treating inflammation and pain related to various inflammatory and musculoskeletal conditions. Using a modified rodent contusion model of SCI, we explored the effects of this botanical on locomotor function and responses to mechanical stimuli, and examined possible neurochemical changes associated with SCI-induced allodynia. Following spinal cord contusion at T10 level, Hp (300 mg/kg, p.o.) or vehicle (water) was administered daily starting 24 h post-surgery, and behavioral measurements made every-other day until sacrifice (Day 21). Hp treatment markedly ameliorated the contusion-induced decrease in locomotor function and increased sensitivity to mechanical stimuli. Determination of Iba1 expression in spinal cord tissues indicated microglial infiltration starting 3 days post-injury. SCI results in increased levels of 4-hydroxynonenal, an oxidative stress product and proalgesic, which was diminished at 7 days by treatment with Hp. SCI also enhanced antioxidant heme oxygenase-1 (HO-1) expression. Concurrent studies of cultured murine BV-2 microglial cells revealed that Hp suppressed oxidative/nitrosative stress and inflammatory responses, including production of nitric oxide and reactive oxygen species, phosphorylation of cytosolic phospholipases A₂, and upregulation of the antioxidative stress pathway involving the nuclear factor erythroid 2-related factor 2 and HO-1. These results support the use of Hp for management of allodynia by providing resilience against the neuroinflammation and pain associated with SCI and other neuropathological conditions.

Introduction

The toll of spinal cord injury (SCI) is extraordinary. Estimates place the yearly global incidence at 40 to 80 cases per million (Rahimi-Movaghar et al. 2013; Lee et al. 2014). Following SCI, individuals are faced with a multitude of challenging symptoms and complications, including impaired motor function of varying severities, development of central and peripheral pain syndromes, autonomic imbalances, and bowel/bladder dysfunction (Widerstrom-Noga 2012). Of these, the pain syndromes associated with SCI are often the most detrimental to quality of life and ability to function (Jensen et al. 2007). Neuropathic pain caused by a lesion or disease of the somatosensory nervous system (Jensen et al. 2011), is the most incapacitating post-SCI pain syndrome reported, often manifesting allodynia (pain produced by non-nociceptive stimuli) and hyperalgesia (increased response to painful stimuli) at or below the level of spinal cord injury (Baastrup and Finnerup 2012; Finnerup 2013).

Central nervous system injury, including SCI, is divided into primary (immediate and irreversible) and secondary (subsequent and potentially preventable) phases. The irreversibility of the primary phase is inherent to the tissue destruction caused by axonal transection, shearing, and compression during the initial traumatic event (Tator and Fehlings 1991). The secondary phase is characterized by continued neuronal death attributable to ischemia and complex inflammatory cascades. It is generally acknowledged that oxidative processes and neuroinflammation following SCI may contribute to both the development of neuropathic pain and poor neurological recovery (Gosselin et al. 2010; Kwon et al. 2011).

The extent of the secondary phase is potentially modifiable, and some current therapeutic strategies are effective for the short-term management of neuropathic pain. However, long-term use of non-steroidal anti-inflammatory drugs may lead to gastrointestinal lesions or renal/hepatic failure (Rao and Knaus 2008). Opiate analgesics are not beneficial in treating neuropathic pain (Gaskell et al. 2016), and their use adds fuel to the fire of the current opioid crisis. In light of urgency to establish novel treatments for SCI related neuropathic pain, some botanicals offer the potential for new treatment options (Quintans et al. 2014).

One candidate botanical is Harpagophytum procumbens (Hp, or Devil’s Claw), that has long been used to treat a variety of inflammatory conditions including degenerative rheumatoid arthritis, osteoarthritis and tendonitis (Mncwangi et al. 2012). Previous work suggests its ability to attenuate multiple components of the inflammatory cascade including the NF-κB pathway, which promotes inducible nitric oxide synthase (iNOS) (Huang et al. 2006; Kaszkin et al. 2004), important cytokine mediators (IL-1β, IL-6, and TNF-α), and enzymes for prostaglandin synthesis, such as cyclooxygenase-1/2 (COX-1/2) (Fiebich et al. 2012; Mncwangi et al. 2012). Harpagoside, harpagide, 8-p-coumaroylharpagide and related metabolites, as well as phenylpropanoid glycosides have been implicated as primary anti-inflammatory effectors; however in purified form, these have variable effects (Abdelouahab and Heard 2008; Zhang et al. 2011; Mncwangi et al. 2012; Georgiev et al. 2013).

Neuroinflammatory and oxidative changes accompanying SCI include microglial activation, alterations in neuronal firing, and long-term changes in synaptic plasticity and regulation (D’Angelo et al. 2013). We hypothesized that administration of Hp following SCI may attenuate the oxidative/inflammatory cascade associated with the secondary phase of injury, and thus suppress responses to thermal and mechanical stimuli, as well as promote functional recovery. To study this, we modified an existing spinal cord contusion model by dispersing impact forces around the dorsal surface of the spine in order to unmask contusion-induced delayed sensory deficits, particularly allodynia, rather than the severe locomotor dysfunction. With this model, we examined the ability for Hp to ameliorate SCI-induced allodynia as well as to mitigate underlying neurochemical changes.

Materials and Methods

Materials

Dulbecco’s modified Eagle’s medium (DMEM) and penicillin/streptomycin were obtained from GIBCO (Gaithersburg, MD). 4-hydroxyhexenal (4-HHE, 1 mg in 0.1 mL), 4-hydroxynonenal (4-HNE, 1 mg in 0.1 mL), 4-hydroxyhexenal-d₃ (4-HHE-d₃, 100 μg in 0.1 mL in methyl acetate), fetal bovine albumin (FBS), Greiss reagent (sulfanilamide and N-1-napthylethylenediamine dihydrochloride), 1,3-cyclohexanedione (CHD, 97%), ammonium acetate (HPLC grade), acetic acid (ACS grade) and formic acid (mass spectrometry grade) were purchased from Sigma-Aldrich (St. Louis, MO). For reactive oxygen species (ROS) detection, CM-H₂DCFDA was purchased from Invitrogen (Eugene, OR).

Antibodies: Iba-1, 016–20001 (Wako,VA), erythroid 2-related factor 2 (Nrf2), GTX103322 (GeneTex, CA), HO-1 (sc-10789) and NQO1 (F8 sc-393736) from Santa Cruz Biotechnology (Santa Cruz, CA), phospho-cytosolic phospholipases A₂ (p-cPLA₂) and cPLA₂ from Cell Signaling, 2831S and 2832S (Beverly, MA), goat anti-mouse or rabbit IgG-horseradish peroxidase from Santa Cruz Biotechnology, sc-2005 and sc-2004, respectively (Santa Cruz, CA), and monoclonal anti-β-actin peroxidase from Sigma-Aldrich, A3854 (St. Louis, MO).

C18 Sep-Pak cartridges (1 mL, 100 mg) were obtained from Waters Corporation (Milford, MA). Phospholipid removal cartridges (Phree, 1 mL) were purchased from Phenomenex (Torrance, CA). All solvents (HPLC grade) used for LC and MS analysis were obtained from Thermo Fisher Scientific (Fair Lawn, NJ).

Animals and Experimental Plan

Animals

Male Sprague–Dawley rats (250–305 g upon arrival) purchased from Charles River Laboratories were housed two per cage in a temperature controlled (20 ± 1 °C) animal facility with a 12-h/12-h light/dark cycle, and allowed free access to standard chow (Teklad) and water. Rats were randomly distributed into treatment or sham control groups. Experimental procedures were approved by the University of Missouri Animal Care and Use Committee, and complied with the National Institute of Health Guide for the Care and Use of Laboratory Animals and ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines.

The behavior study used a 2 × 2 between-group design with four independent groups of 12 rats per group (Fig. 1). (1) Rats received SCI surgery and administered Hp extract (300 mg/kg body weight, q.d.) as the SCI + Hp group; (2) SCI + vehicle (water); (3) Rats received sham surgery and administered Hp (Sham + Hp), and (4) Sham + vehicle. In a parallel study, an additional group of SCI animals were treated with the commonly used ibuprofen (20 mg/kg body weight, b.i.d.) in order to compare with the effects of Hp.

Fig. 1.

Scheme of the spinal cord injury (SCI) experimental design. a Time course for pre-surgery baseline behavior evaluation followed by surgery and post-surgery behavior evaluation for 21 days. Rats were sacrificed on Day 22. b Dissection of the segmented spinal cord regions at 1, 3, and 7 days post-injury for mass spectrometry and neurochemical analyses

Spinal Surgery

Upon arrival, rats were handled for at least 3 days before evaluation of baseline behavioral assessments (Day-9 to Day-1). Surgery was performed on Day 0 and Hp or vehicle treatment began 24 h after surgery (Day 1), and continued once-daily with concomitant evaluation of behavior through Day 21, when all rats were euthanized. Spinal cord contusion was applied to the SCI group, and sham surgery was limited to laminectomy alone (Krishna et al. 2013). Animals were anesthetized with isoflurane inside a sealed anesthesia chamber to achieve a proper depth of anesthesia, and then transferred to a silicon heating pad (HS-3 × 2.5 Heater, Cell MicroControls, Norfolk, VA) to maintain a constant 37 °C temperature. Isoflurane was continuously administered for the duration of the procedure.

Bony landmarks were palpated to determine the T10 level, and the respective region was prepped with iodine before shaving a small area. Under aseptic conditions, an approximately 20-mm midline incision was made over the T10 level to expose the spinous process and lamina. Using an operating microscope, a laminectomy was performed at the T10 level, exposing the dorsal spinal cord. For each animal, a 2–3 mm small polystyrene disc was inserted over the exposed dorsal cord surface prior to impact in order to dissipate the force of impact evenly along the dorsal surface of the spinal cord. A MATLAB-controlled electromagnetic impactor (Leica Impact One Stereotaxic Impactor, St. Louis, MO) with a 3.0-mm diameter tip was centered perpendicularly over the plastic disc, and delivered a contusion injury with a velocity of 3.0 m/s with dwell time of 100 ms, at a depth of 2.0 mm. This combination of velocity, depth, and placement of the plastic disc produced a moderate and consistent spinal cord contusion in the animals during pilot studies. Following the impact, the plastic disc was removed and the muscular plane and skin were stitched together, and triple antibiotic ointment was applied to the incision. Animals were released from anesthesia, returned to their cages, and observed until fully awake.

After operation, bladders of all rats with SCI were gently pressed twice-daily to aid reflex on voiding functions. A standardized pain and distress scale was used to score animals twice-daily on Days 1–7, and then twice weekly thereafter through Day 21. Observations, including behavior, weight, appetite, and respiratory rate were recorded by trained observers blinded to treatment. For the sham rats, anesthesia and laminectomies were performed in a similar fashion. Following cord exposure, no impact was delivered and the muscular plane and skin were sutured closed. Following recovery from anesthesia, no sham animals were observed to exhibit motor deficits.

Preparation of H. Procumbens Extract

The aqueous Hp extract from which pro-inflammatory mediators are removed by precipitation with ethanol was prepared as described in U.S. Patent 6,280,737B1 and is compliant with NIH and FDA guidances for botanical products. The extract was analyzed by ultra-high performance liquid chromatography (UHPLC) coupled with both UV photodiode assay detection and time-of-flight tandem mass spectrometry (TOF–MS/MS) with lockspray ionization for candidate bioactive secondary metabolites characteristic of Hp (Fig. 2). UHPLC-MS/MS analyses were performed on a Bruker maXis impact quadrupole-time-of-flight mass spectrometer coupled to a Waters ACQUITY UHPLC system. Separation was achieved on a Waters C18 column (2.1 × 100 mm, BEH C18 column with 1.7-μm particles) using a linear gradient and mobile phase A (0.1% formic acid) and B (B: acetonitrile). Gradient condition: B increased from 5 to 70% over 30 min, then to 95% over 3 min, held at 95% for 3 min, then returned to 5% for equilibrium. The flow rate was 0.56 mL/min and the column temperature was 60 °C. Mass spectrometry was performed in the negative electrospray ionization (ESI) mode with the nebulization gas pressure at 43.5 psi, dry gas of 12 L/min, dry temperature of 250 C and a capillary voltage of 4000 V. Mass spectral data were collected from 100 to 1500 m/z and were auto-calibrated using sodium formate after data acquisition. Quantitation of harpagoside concentration was performed using UV at 280 nm. Standard curve was generated using authentic harpagoside standard of 6 different concentrations. Harpagoside in the samples was quantified using the peak area at UV280 nm and the calibration curve.

Fig. 2.

UHPLC analysis of H. procumbens extract. Hp extract was used for UHPLC analysis using protocol as described in text. Results show signature iridoid and phenylpropanoid metabolites identified based upon measured retention time (Rt) and accurate mass (m/z) relative to authentic standards

The content of harpagoside in the extract was determined to be 5.3%. This extract was administered daily by gavage, and the dose administered to each rat was approximately 300 mg/kg/day or 16 mg/kg/day harpagoside. This dose was chosen because it resembles the dose used by humans and has been shown to alleviate hypersensitivity in the spared nerve injury rat model (Lim et al. 2014). Concurrent toxicological studies in rats (data not shown) and reports of 28 clinical trials of a variety of H. procumbens products (Vlachojannis et al. 2008) indicate the extract causes no clinical pathology and the equivalent dose of other Hp products is well-tolerated by humans.

Evaluation of Behaviors

Motor Function

Motor function was evaluated pre- (Days-9, −7, −5, −3 and −1) and post-surgery (Days 1, 3, 5, 7, 9, 11, 13, 15 and 21) using a standard scale (Basso et al. 2002). Rats were individually observed in an open field environment (120 cm × 60 cm) for at least one minute (min) by two trained observers, who were blinded to the animal groups. Scores on the scale ranged from 0 to 21. Briefly, 0 was defined as absolutely no hind limb movement, and 21 was defined as full mobility in hind limbs, consistent with elevation of the tail, trunk stability, and parallel lift and placement of the paws. Analyses were performed on data from the pre- and post-surgery periods using the Kruskal–Wallis test. Group differences at each day were elucidated via the Mann–Whitney test when appropriate (p < 0.05), with correction of the significance level to account for multiple comparisons.

Locomotor Activity

Locomotor activity was assessed in acrylic monitor boxes (ENV-515, Med Associates’ Open Field Test Environments, St Albans VT; 43.2 × 43.2 × 30.5 cm) that were housed in sound-resistant cubicles. On Days-3 and 15, rats were placed in the apparatus and behavior was recorded for 60 min by three sets of infrared sensors located on the x, y, and z-axes. Open Field Activity Software (SOF-811) was used to calculate the distance traveled (in cm) by measuring the amount of sensor breaks that occurred. The monitor boxes were cleaned with a mild soap solution between sessions. Distance traveled data from the pre- and post-surgery tests were analyzed via analysis of variance with surgery (SCI and sham) and treatment (H. procumbens and vehicle) as between-groups factors and time as a within-subjects factor. Tukey post hoc comparisons were performed when appropriate (p < 0.05).

Response to Mechanical Stimulus

The response to a mechanical stimulus was assessed with von Frey filaments (Touch-Test Sensory Evaluator, North Coast Medical, Gilroy CA). On each pre-surgery (Days-9, −7, −5, −3 and −1) and post-surgery (Days 1, 3, 5, 7, 9, 11, 13, 15 and 21) test day, rats were acclimated for 15 min within individual Plexiglas chambers (12 × 20 × 17 cm) on the surface of a metal grid (approximately 0.6 × 0.6 cm) platform. A filament of 2 g force was applied to the plantar surface of one hind paw for approximately 2 s. If withdrawal to the stimulus was observed, then a filament of a lower weight (1.4–0.4 g) was applied. If withdrawal to the 2 g stimulus was not observed, then a filament of a higher weight (4–15 g) was applied. The dependent variable was the lowest weight that produced a full response (Lim et al. 2014). There was a 5-min inter-stimulus interval, and then the other plantar surface was tested, as described, beginning with the 2 g filament. After both hind limbs were examined, rats were returned to the home cage and the apparatus was cleaned with a mild soap solution. Responses from the left and right limbs were averaged. Data were handled with a similar statistic program as described in Sect. 2.4.1.

Neurochemical Determinations

Preparation of Spinal Cord Homogenate

For neurochemical assays, frozen spinal cord tissues were weighed, homogenized in Laemmli lysis buffer, and centrifuged at 10,000×g for 15 min at 4 °C. Aliquots of the homogenate were taken for protein determination using the Pierce BCA Protein Assay kit (Thermo Scientific). Samples were stored in Ependorf tubes and stored at −80 °C until use for neurochemical assays.

Culture of BV-2 Microglial Cells

For confirmation of oxidative/nitrosative stress, inflammatory responses, and anti-oxidative pathways, BV-2 microglial cells, between 14 and 25 passages, were prepared as previously described (Shen et al. 2005; Chuang et al. 2016,2014, 2015; Song et al. 2016). Initially, cells were cultured in 75 cm² flasks with DMEM supplemented with 5% FBS, and with 100 units/mL penicillin and streptomycin (100 μg/mL). Cells were maintained in a 5% CO₂ incubator at 37 °C. During experiment, cells were subcultured in required dishes or plates until 80% confluent. Cells were serum starved for 3 h prior to application of Hp extract for 1 h, and subsequently exposed to lipopolysaccharide (LPS, 100 ng/mL) for the designated time.

Nitric Oxide (NO) Determination in Culture Medium

NO released from cells was converted to nitrite in the culture medium, which was determined using the Griess reagent (Sheng et al. 2011; Qu et al. 2014). In these experiments, cells were cultured in phenol red free DMEM for 24 h. After being serum starved for 4 h, cells were treated with different amounts of Hp extract for 1 h and then stimulation with LPS (100 ng/mL) for 16 h. For assay of NO, aliquots (50 μL) of culture medium were transferred to 96-well plates and incubated with 50 μL of reagent A [1%, w/v, sulfanilamide in 5% phosphoric acid] for 10 min at room temperature, and followed by incubation with 50 μL of reagent B [0.1%, w/v, N-1-napthylethylenediamine dihydrochloride] for 10 min at room temperature in the dark. The absorbance of samples were measured at a wavelength of 565 nm using the Synergy4 Plate Reader (BioTek Instruments, Winooski, VT). Sodium nitrite (0–100 μM), diluted in culture media, was used to prepare the nitrite standard curve.

Measurement of ROS Production

ROS production was measured using the ROS detection reagent CM-H₂DCFDA (DCF) as described previously (Chuang et al. 2013). Briefly, BV-2 cells were cultured in 96-well plate for 4 h in serum free DMEM prior to adding Hp extracts for 1 h and LPS for 12 h. This was followed by adding DCF (for 1 μM final concentration) to each well for 1 h, and the fluorescence intensity of DCF was measured using the Synergy4 Plate Reader with an excitation wavelength of 490 nm and an emission wavelength of 520 nm.

Western Blot Analysis

Western blot analysis was carried out either with the spinal cord homogenates or with BV-2 cells (Chuang et al. 2015). Samples were treated with Laemmli lysis buffer and centrifuged at 10,000×g for 15 min at 4 °C to remove cellular debris. After protein quantification with the BCA protein assay kit, samples together with protein standards (Dual color, BioRad, Hercules, CA) were loaded into sodium dodecyl sulfate–polyacrylamide gel electrophoresis gels and resolved at 100 V. After electrophoresis, proteins were transferred to 0.45 μm nitrocellulose membranes at 100 V for 1.5 h. Membrane strips were blocked in Tris-buffered saline, pH 7.4, with 0.1% Tween 20 (TBS-T) containing 5% non-fat milk for 1.5 h at room temperature. The blots were incubated with anti-Iba-1 (1:1000 dilution), anti-Nrf2 (1:500 dilution), anti-HO-1 (1:800 dilution), anti-NQO1 (1:200 dilution), anti-p-cPLA₂ (1:1000 dilution) or anti-cPLA₂ (1:1000 dilution) antibodies overnight at 4 °C. After repeated washing with TBS-T, blots were incubated with goat anti-rabbit IgG-horseradish peroxidase (1:6000 dilution) for 1 h at room temperature and then washed three times with TBS-T. Immuno-labeling was detected by SuperSignal chemiluminescent substrates (Thermo Scientific, Rockford, IL). For loading control, blots were incubated with anti-β-actin (1:30,000 dilution) and goat anti-mouse IgG-horseradish peroxidase (1:6000 dilution). Films were scanned, and the optical density of protein bands was measured using the QuantityOne software program (BioRad, Hercules, CA).

LC–MS/MS Analysis of 4-HNE and 4-HHE in Spinal Cord

For LC–MS/MS analysis, spinal cord tissue sections above the injured site were used (see Fig. 1). These segments were weighed, and homogenized in double distilled water using a weight to water ratio of 1:8 (wt/vol). The homogenate was centrifuged at 4 °C at 10,000×g for 20 min, and 30 μL of supernate was transferred to an Eppendorf tube. Samples were processed similar to that described earlier (Yang et al. 2018). Briefly, equal volume of internal standard (4-HHE-d₃, 1000 ng/mL) was added to the cell suspension together with acetonitrile (0.5 mL) containing 1% formic acid. Solid phase extraction (SPE) was carried out using a Phree cartridge. 4-HHE, 4-HNE, and 4-HHE-d₃ were derivatized by adding 200 μL of freshly prepared acidified 1,3-cyclohexanedione (CHD) reagent at 60 °C for 1 h. The derivatized 4-HHE and 4-HNE were desalted using a C18 SPE cartridge. The eluate was evaporated to dryness under a stream of nitrogen gas. An aliquot of the reconstituted solution was injected into a Waters Xevo TQ-S triple quadrupole mass spectrometer. The multiple reaction monitoring transitions m/z 326.3 > 216.1 Da, 284.2 > 216.1 Da and 287.2 > 216.1 Da were chosen for simultaneous monitoring of 4-HNE, 4-HHE and 4-HHE-d₃ derivatives, respectively. MassLynx software (v4.1, Waters) was used for all data acquisition.

Statistical Analysis for Neurochemical Studies

To quantify neurochemicals in spinal cords from sham and SCI animals, results were expressed as mean ± SEM and analyzed using one-way ANOVA with Fisher-LSD for multiple comparisons by Prism v7.03 (GraphPad Software, San Diego, CA). For studies measuring 4-HHE and 4-HNE levels using LC–MS/MS, experiments were carried out with three biological replicates and two analytical replicates. Results are expressed as the mean ± SEM, and analyzed by one-way ANOVA with Fisher-LSD for multiple comparisons. For studies to assess oxidative and inflammatory responses in activated microglial cells, triplicate analyses were performed on a given condition, and at least three independent experiments with different passages were performed for each condition. Results are analyzed by one-way ANOVA followed by Tukey’s post-tests (v7.03; GraphPad Prism Software, San Diego, CA). Differences were considered significant at p < 0.05 for all analyses.

Results

Locomotor Function

Motor function scores from the standard scale ranging from 0 to 21 (with 0 defined as no limb movement and 21 as full limb mobility) are presented in Fig. 3a. Significant differences were observed among groups on Days 1 to 21 (T-values > 19.3, p < 0.002). There were no significant differences between the Sham + Vehicle and Sham + Hp groups on any day. However, the SCI + Vehicle group showed significantly lower scores than the Sham + Vehicle group on Days 1 to 11. From Day 1 through Day 21, scores for the SCI + Vehicle group increased systematically, such that scores on Days 13 to 21 were significantly greater than those on Day 1. The SCI + H. procumbens group had lower scores than the Sham + Vehicle group only on Days 1 to 5. Moreover, rats in the SCI + Hp group demonstrated a more rapid increase in motor function relative to rats in the SCI + Vehicle group. For the SCI + Hp group, scores were greater on Days 7 to 21 than on Day 1.

Fig. 3.

Impairment in motor function in SCI is attenuated with H. procumbens treatment. a Rats were placed in an open field and their behavior was observed and rated on a standard scale by trained observers blinded to treatment. Data were analyzed via the Kruskal–Wallis test followed by Mann–Whitney comparisons. Asterisks indicate a significant (p < 0.001) difference between the Sham + Vehicle and SCI + Vehicle group in the respective test day. Plus signs indicate a significant (p < 0.001) difference from the within-group response on Day 1. b SCI produces impairment in locomotor activity that is attenuated with H. procumbens treatment. Four groups of rat (n = 12 per group) on Days-3 and 15 were placed in an automated locomotor activity apparatus, and behavior was measured for 60 min. Analyses were performed from the pre- and post-surgery periods using the Kruskal–Wallis test. Group differences at each day were elucidated via the Mann–Whitney test when appropriate (p < 0.05)

Locomotor activity was evaluated by assessing total distance traveled for the four groups of rats (n = 12 per group) in pre-surgery (Day-3) and post-surgery (Day 15). There were no significant differences among the four groups on the pre-surgery test (overall mean = 7411 cm/60 min, SEM = ± 313 cm/60 min). Figure 3b displays data from the post-surgery test for total distance traveled from the entire 60-min session. Although not reaching significance, rats in the SCI + Vehicle group were less active than rats in the Sham + Vehicle group; and rats in the SCI + Hp group appeared to be more active than the SCI + Vehicle group (Fig. 3b). There was a trend of decreasing in distance traveled across the entire 60-min session in SCI + Vehicle group.

Response to a Mechanical Stimulus

The response to a mechanical stimulus is presented in Fig. 4. All groups were tested with Von Frey filaments on alternating days before and after surgery. No significant differences were found among groups on the 5 days before surgery (Day-9 to −1). However, there were significant differences among groups after surgery (T-values > 8.75, p < 0.03). There were no differences between the Sham + Vehicle and Sham + H. procumbens group on any test day. On Days 9, 11, 13 and 15, rats in the SCI + Vehicle group exhibited a lower mechanical withdrawal threshold compared to rats in the Sham + Vehicle group. However, rats in the SCI + H. procumbens group differed from the Sham + Vehicle group only on Day 15. There were no significant differences between the SCI + H. procumbens group and Sham + Vehicle group on any other day.

Fig. 4.

SCI produces an enhanced response to a mechanical stimulus that is attenuated by H. procumbens. Von Frey filaments were applied to the plantar surface of each hind limb. The minimal intensity that produced an observable withdrawal response was measured. Data were analyzed via Kruskal–Wallis test followed by Mann–Whitney comparisons. Asterisks indicate a significant (p < 0.001) difference from the Sham + Vehicle control group at the respective test day

Neurochemical Studies

Experimental Design

In this study, animals were divided into sham, SCI, and SCI + Hp groups, and spinal cord tissue sections were collected at 1 day (n = 3), 3 days (n = 3) and 7 days (n = 6) after injury. In the 7-day study, an additional SCI group (n = 6) was treated with Ibuprofen. In animals involving Hp, Hp was administered to SCI rats by gavage daily; the 1 day group was administered Hp the next morning 2 h prior to euthanization. Spinal cord tissues were dissected (see Fig. 1), weighed and homogenize and protein concentration was determined as described in “Materials and Methods” section.

SCI-Induced Increase in Iba1 Expression and Microglial Infiltration

Western blot analysis of Iba-1 expression was carried out with spinal cord homogenates to indicate microglial infiltration. Results showed significant increases in Iba-1 expression in all animals with SCI in the 7-day samples, but no significant difference among the groups treated with Hp or ibuprofen (Fig. 5a). In the experiment in which spinal cords were obtained from 1 and 3 days after SCI, increases in Iba-1 expression were observed only in the 3-day groups and not in the 1-day groups (Fig. 5b).

Fig. 5.

SCI induces increase in Iba-1 expression. Sham and SCI animals were treated daily with H. procumbens extract, ibuprofen or vehicle (water) p.o. as described in text. Spinal cord tissues were dissected 1, 3 and 7 days after surgery, and homogenate was used for Western blot analysis for quantification of Iba-1 expression. Immunoblots for the levels of β-actin were performed as loading control. a Results for 7 day samples are mean ± SEM and analyzed by one-way ANOVA with Fisher-LSD for multiple comparisons. **p < 0.01, and ***p < 0.001. N = 6 for sham, SCI + vehicle (water), SCI + H. procumbens (Hp) groups, and N = 5 for SCI + Ibuprofen (Ibu) group. b Results for 1 and 3 day samples are mean ± SE, N = 3, ** p < 0.01, *** p < 0.001 comparing SCI with sham

SCI and Hp Treatment Altered Levels of 4-HNE and 4-HHE

Although oxidative stress is important to the damage associated with SCI (Jia et al. 2012; Visavadiya et al. 2016) and the concomitant neuropathic pain (De Logu et al. 2017; Grace et al. 2016; Raoof et al. 2018), mechanisms for this are not well established. Lipid peroxidation has been regarded an oxidative process, and recent studies recognized production of 4-HNE from arachidonic acid (ARA) and 4-HHE from docosahexaenoic acid (DHA) (Sun et al. 2018). Recently, a modified LC–MS/MS protocol was validated to simultaneously determine levels of 4-HNE and 4-HHE in microglial cells (Yang et al. 2018) and in mouse brain tissue (Yang et al. 2019b). In the present study, we observed increases in 4-HNE and decreases in 4-HHE in the SCI group at 7 day after SCI (Fig. 6a, b). Interestingly, these changes were mitigated upon treatment with Hp.

Fig. 6.

H. procumbens extract and ibuprofen suppressed SCI-induced increase in 4-HNE (a) and decrease in 4-HHE (b) in spinal cord. Spinal cord tissues from sham and SCI animals were obtained at 7 days post-surgery as described in text. Two aliquots of tissue homogenates from three animals per group were obtained for the LC–MS/MS analysis which was carried out in three analytical replicates. Results are mean ± SEM and analysis by one-way ANOVA with Fisher-LSD for multiple comparisons shows * p < 0.05, ** p < 0.01, and *** p < 0.001 for sham, SCI + vehicle (water), and SCI + H. procumbens (Hp), and SCI + Ibuprofen (Ibu) group

In subsequent study, we determined effects of Hp on levels of 4-HNE and 4-HHE in spinal cord samples at 1 and 3 days after SCI. Although not reaching significant levels, there is a trend for increases in 4-HNE levels in the SCI groups in both 1 and 3 days, and Hp treatment suppressed 4-HNE levels in the 3-day group (Table 1). Similar to the data from the 7-day samples, SCI appeared to decrease 4-HHE levels, and Hp treatment tended to reverse this change. Overall, determination of the 4-HNE/4-HHE ratios indicated increase in 4-HNE/4-HHE ratios due to SCI and reversal by Hp treatment for all three time groups (Table 1).

Table 1.

LC-MS/MS analysis of 4-HNE and 4-HHE levels and 4-HNE/4-HHE ratios in spinal cord homogenates after SCI

Ng/g wt	4-HNE (1 day)	4-HNE (3 day)	4-HHE (1 day)	4-HHE (3 day)
Sham	234.5 ± 27.8	244.9 ± 35.9	337.5 ± 54.4	357.8 ± 67.8
SCI	267.3 ± 46.1	265.2 ± 35.4	274.3 ± 49.5	314.3 ± 66.9
SCI + Hp	285.1 ± 45.9	246.3 ± 30.1	420.2 ± 36.7a,b	402.0 ± 70.4
4-HNE/4-HHE	1 day		3 day	7 day
Sham	0.73 ± 0.2		0.74 ± 0.2	0.87 ± 0.06
SCI	1.02 ± 0.1a		0.85 ± 0.1	1.29 ± 0.35a
SCI + Hp	0.73 ± 0.2b		0.56 ± 0.1	0.73 ± 0.06

Spinal cord tissues (sham, SCI, and SCI + Hp, n = 3 per group) were obtained at 1, 3 and 7 days post-surgery as described in text. Three aliquots of the tissue homogenate from each animal were obtained for the LC-MS/MS analysis which was carried out in two analytical replicates. Results are mean ± SEM and analysis by one-way ANOVA with Tukey multiple comparisons, and showing significants:

^acomparing SCI or SCI + Hp with sham p < 0.05;

^bcomparing SCI + Hp with SCI p < 0.0001

Effects of SCI and Hp Extract on HO-1 Expression in Rat Spinal Cord

Some botanicals with electrophilic properties have been shown to upregulate the antioxidant stress pathway involving Nrf2 and HO-1 (Ajit et al. 2016; Sun et al. 2015). In this study, immunoblotting analysis of Nrf2 and HO-1 expression in spinal cord samples indicated increases in HO-1 expression in SCI and SCI + Hp groups as compared to the sham group. In the study using samples from the 7-day group, the Hp treated group appeared to show higher expression as compared with the SCI group (Fig. 7). In this study, attempts to test for expression of Nrf2 in these samples were not successful, possibly due to lack of reactivity of the antibodies used against rat samples.

Fig. 7.

H. procumbens extract administration induced increase in HO-1 expression in injured spinal cord. Spinal cord homogenates were obtained from sham and SCI animals treated daily with H. procumbens extract, ibuprofen or vehicle (water) p.o. as described in Fig. 4. Western blot analysis for quantification of HO-1 and NQO1 expression, and immunoblots for the levels of β-actin were used as loading control. Analysis by one-way ANOVA with Fisher-LSD for multiple comparisons shows *p < 0.05, and **p < 0.01. N = 6 for sham, SCI + vehicle (water), and SCI + H. procumbens (Hp), and N = 5 for SCI + Ibuprofen (Ibu) group

The Nrf2 pathway is known to induce a number of antioxidant enzymes, and besides HO-1, corresponding induction of NAD(P)H quinone dehydrogenase 1 (NQO1) has been shown in a number of studies (Riego et al. 2018; Li et al. 2016; Mao et al. 2011). In the present study, the 7-day samples used for HO-1 expression were tested for expression of NQO1. Despite a higher base line expression for NQO1, expression profile for NQO1 appeared similar to HO-1 in the SCI and SCI + Hp groups (Fig. 7).

Effects of H. procumbens on Microglial Cell Oxidative and Inflammatory Responses

In recent years, BV-2 microglial cells have been used as a model to test effects of botanicals on oxidative and inflammatory pathways (Dai et al. 2019; Lv et al. 2019). In this study, we examined effects of Hp extract on LPS-induced production of NO, ROS, and p-cPLA₂, as well as on the antioxidant pathway involving Nrf2 and synthesis of HO-1. Hp extract dose-dependently suppressed LPS-induced NO (Fig. 8a) and ROS production (Fig. 8b). In agreement with our previous studies showing that LPS stimulated phosphorylation of cPLA₂, enzyme responsible for production of ARA (Chuang et al. 2015, 2016), results here also show that Hp extract could suppress the LPS-induced increase in p-cPLA₂ in BV-2 cells (Fig. 9).

Fig. 8.

H. procumbens extract suppressed LPS-induced NO (a) and ROS (b) production in BV-2 microglial cells. Microglial cells were cultured in 96-well plates as described in text. During experiments, cells were serum starved for 3 h prior to addition of H. procumbens extract for 1 h and then LPS (100 ng/mL) for 16 h for NO assay and for 12 h for ROS assay. Protocols for NO and ROS assays are described in text. Results are mean ± SEM from experiments with 4 passages, and in each passage, samples were carried out in triplicates. For NO determination, samples with or without LPS were carried out with different concentrations of H. procumbens (from 0 to 80 μg/ mL). The net concentrations of NO in the samples from the four passages as expressed as μM (mean ± SEM) are: 1.85 ± 0.50, 2.24 ± 0.30, 6.07 ± 0.82, and 2.78 ± 0.12, and the mean ± SEM is 3.23 ± 1.92 μM. The value for NO induced by LPS in each passage was then converted to 100% for testing effects of H. procumbens extract. Analysis by one-way ANOVA with Bonferonni post-tests **p < 0.01 and ***p < 0.001, comparing treated samples with controls

Fig. 9.

H. procumbens extract suppressed LPS-induced p-cPLA₂ in BV-2 microglial cells. Microglial cells in 12-well plates were serum starved for 3 h prior to addition of H. procumbens extract for 1 h and followed by with or without LPS (100 ng/mL) for 4 h. p-cPLA₂ and PLA₂ expressions were assessed by Western blot as described in text. Immunoblots for the levels of β-actin were used as loading control. Results are mean ± SEM of p-cPLA₂/cPLA₂ ratios from 4 passages, and data analyzed by one-way ANOVA with Bonferonni post-tests indicated *p < 0.05 comparing treated samples with controls

Botanicals, such as quercetin, have been shown to upregulate the antioxidant stress pathway involving Nrf2 and synthesis of HO-1 (Sun et al. 2015). In this study, Hp dose-dependently stimulated Nrf2 and HO-1 expression in microglial cells (Fig. 10). In addition, Hp did not suppress levels of HO-1 in cells stimulated with LPS despite that LPS itself also increased Nrf2 and HO-1 expression in the cells (Fig. 10).

Fig. 10.

H. procumbens extract enhances the Nrf2/HO-1 pathway in microglial cells. Microglial cells in 12-well plates were incubated in DMEM with 5% FBS as described in Fig. 7. During experiments, cells were serum starved for 3 h prior to addition of different concentrations of H. procumbens extract. Cells were incubated with or without LPS and with different concentrations of H. procumbens for 6 h. a A representative Western blots analysis of cells treated with different levels of H. procumbens. Expression of b Nrf2 and c HO-1 from 4 cell passages with β-actin as loading control. Results are mean ± SEM from 4 passages and analyzed by one-way ANOVA, *p < 0.05 and **p < 0.01

Discussion

Neuropathic pain developed following SCI causes significant distress. In many instances, SCI is also marked by increased oxidative stress and inflammation, although the mechanism(s) linking neuropathic pain and such pathways are only beginning to be understood (De Logu et al. 2017; Grace et al. 2016; Hackel et al. 2013; Raoof et al. 2018; Kartha et al. 2018). In this study following a contusion-induced SCI, we observed behavioral deficits associated with neuropathic pain, and demonstrated the ability for Hp extract to mitigate such behavioral deficits. Initially, we tested two primary indicators of neuropathic pain, an enhanced response to a mechanical stimulus and an enhanced response to a thermal stimulus. The former simulates the allodynia experienced by humans after SCI, while the latter refers to hyperalgesia. In our experiments, contusion-induced SCI produced a heightened response to a mechanical stimulus, but did not produce a marked alteration in the response to the thermal stimulus (data not shown).

As expected, spinal cord contusion caused a marked impairment in locomotor behavior after surgery. Over the course of the post-surgery period, rats showed recovery in function, although there were significant differences between rats in the SCI + Vehicle and Sham + Vehicle groups throughout the treatment.

Treatment with Hp extract attenuated the increased sensitivity to the mechanical stimulus: i.e., rats in the SCI + Hp group required a stronger tactile stimulus to display a response than rats in the SCI + Vehicle group. However, this difference was not evident immediately after contusion, but became obvious between Days 9 and 21. There are several possible reasons for this delayed response. As noted, the contusion-induced an impairment of motor function that was pronounced in the first week. Motor function impairment can mask the group differences in the sensation and/or perception of the tactile stimulus. Subsequently, sub-chronic or chronic treatment with Hp extract to gain efficacy reflect that therapeutic conditions in humans. In this experiment, Hp extract was administered to rats at a dose level equivalent for human consumption, which may differ for rats. These results suggest future studies to test whether different dose levels of the extract may modify the sensory/motor circuits and time for recovery of behavior.

Neuropathic pain likely develops a multitude of cellular responses following spinal cord injury, and elevations in neuroinflammatory mediators (ROS, IL-1β, IL-6, TNF-α) and glial activation are likely involved in the development of neuropathic pain (Hackel et al. 2013; Raoof et al. 2018). Together with several mechanisms likely at play, a broad approach to the treatment of SCI pain is essential. Ibuprofen is known to temporarily reduce fever, inflammation and pain, but high doses and chronic use may cause stomach and intestinal problems. Hp reduces inflammation but the mechanism may differ from Ibuprofen. To this end, a variety of Hp products have been shown to reduce production of numerous inflammatory mediators including ROS, iNOS, IL-1β, IL-6, TNF-α, COX-1/2, and to increase the stress-mediated anti-oxidative responses through increasing heme oxygenase in different animal models (Parenti et al. 2015; Mncwangi et al. 2012; Huang et al. 2006; Kaszkin et al. 2004). In our present study, SCI-induced increase in Iba1 expression at 3 days and not 1 day, suggesting that microglial infiltration is delayed. Treatment with Hp extract did not appear to suppress Iba-1 expression in the injured spinal cord. The possibility that Hp treatment might change the phenotype of the injured microglia will be an important subject for future exploration. Our modified contusion model in rats should allow studies to investigate neurochemical factors contributing to the development of neuropathic pain and potentially explain the effects observed with Hp extract.

Increase in oxidative stress is a major secondary event linked to brain injuries (Grace et al. 2016). Phospholipids in the central nervous system are enriched in polyunsaturated fatty acids, namely DHA and ARA, which are susceptible to lipid peroxidation. Oxidized phospholipids can modify channel activity and membrane functions, and induce acute and chronic inflammatory pain (Oehler et al. 2017). Recent studies have focused on 4-HNE, a lipid peroxidation product and a proalgesic derived from the inflammatory pathway causing pain though cPLA₂ and ARA (Csala et al. 2015). cPLA₂ is ubiquitous in different brain cells (Sun et al. 2014), and activation of this enzyme has been shown to play an important role in neurodegenerative diseases and injury, including SCI (Liu et al. 2014a, b). In turn, interaction of ARA by COX and LOX results in the production of a large number of oxylipins, lipid intermediates involved in inflammatory processes and pain (Sun et al. 2018). Our earlier studies have linked the cPLA₂/ARA pathway to inflammatory responses for microglial activation (Chuang et al. 2015). Botanical polyphenols have been shown to mitigate microglial activation and reduce neurotoxic effects due to cPLA₂ (Chuang et al. 2016).

Increase in oxidative stress is accompanied by free radical production and associated lipid peroxidation in biological systems (Anthonymuthu et al. 2016; Yadav et al. 2018). Our previous study with microglial cells demonstrated differences in signaling pathways for metabolism of ARA and DHA, and subsequently the production of 4-HNE and 4-HHE, respectively (Yang et al. 2018; Yang et al. 2019a). LPS is known to stimulate p-cPLA₂ and in turn, release of ARA which is linked to the increase in 4-HNE (Yang et al. 2018). This pathway well explains the increase in 4-HNE due to SCI, and ability for Hp extract to mitigate SCI-induced increase in 4-HNE. This is the first study to demonstrate effects of this botanical on SCI and 4-HNE, an important proalgesic.

Many botanicals with electrophyllic properties can not only suppress the oxidative and inflammatory pathways involving NF-kB, but can also upregulate the anti-oxidative stress pathway involving Nrf2 and synthesis of HO-1 (Sun et al. 2015; Ajit et al. 2016). In a recent study, polydatin, a glycoside of resveratrol, was shown to attenuate spinal cord injury through up- and down-regulating these two pathways in microglial cells (Lv et al. 2019). There are other studies demonstrating involvement of Nrf2 and HO-1 from phytochemicals and neuropathic pain (Diaz et al. 2019; Rahbardar et al. 2018; Ma et al. 2018). In the present study, Hp extract dose-dependently increased Nrf2 and HO-1 expression in the murine BV-2 microglial cells. However, although Hp appeared to enhance HO-1 expression in injured spinal cord, and similarly, expression of NQO1 appeared to parallel that for HO-1, more studies are needed to ascertain the role of Hp on Nrf2 pathway in SCI.

Conclusion

In summary, this study demonstrated the ability for Hp to ameliorate SCI-induced allodynia in a rat model. SCI-induced increase in expression of Iba-1, suggesting increase in microglial cell infiltration. SCI also induced increases in 4-HNE, a lipid peroxidation product derived from cPLA₂ and ARA, and a key proalgesic. Administration of Hp extract was able mitigate SCI-induced production of 4-HNE. Finally, administration of Hp extract was shown to increase in HO-1 in the injured spinal cord. Study with BV-2 microglial cells confirms the anti-oxidative and anti-inflammatory properties of Hp extract, and suggests ability for this herbal product to upregulate the antioxidant stress pathway involving the Nrf2 pathway.

Abbreviations

4-HHE

4-Hydroxyhexenal

4-HNE

4-Hydroxynonenal

ARA

Arachidonic acid

CHD

1,3-Cyclohexanedione

COX-1/2

Cyclooxygenase-1/2

cPLA₂

Cytosolic phospholipases A₂

DHA

Docosahexaenoic acid

DCF

ROS detection reagent CM-H2DCFDA

DMEM

Dulbecco’s modified Eagle’s medium

ESI

Electrospray ionization

FBS

Fetal bovine albumin

HO-1

Heme oxygenase-1

Hp

Harpagophytum procumbens, or Devil’s Claw

HPLC

Ammonium acetate

Ibu

Ibuprofen

iNOS

Nitric oxide synthase

LPS

Lipopolysaccharide

Nrf2

Erythroid 2-related factor 2

NQO1

NAD(P)H quinone dehydrogenase 1

ROS

Reactive oxygen species

SCI

Spinal cord injury

SPE

Solid phase extraction

TBS-T

Tween 20

TOF–MS/MS

Time-of-flight tandem mass spectrometry

UHPLC

Ultra-high performance liquid chromatography