Introduction

Stretching is a widely applied therapy for preventing and reducing spasticity and contractures resulting from spinal cord injury (SCI) while attempting to maintain or increase joint range of motion in this population [1]. The rationale for the use of stretching following SCI is largely based on studies in intact rats demonstrating that muscle shortening and reduced joint range of motion following cast immobilization can be mitigated with intermittent stretching therapies [2]. While increased range of motion has been seen in spinally intact animals, studies from our lab indicate that daily hindlimb stretching does not improve joint contractures following contusive SCI, and furthermore, it results in deficits in locomotor function compared to unstretched SCI control animals [3]. These deficits are present when stretching is applied at acute or chronic timepoints [4] using either static or dynamic stretching protocols [5] without overt signs of muscle damage. Although the mechanisms behind this dysfunction are not entirely clear, we found that SCI animals depleted of TRPV1+ unmyelinated primary afferents (predominantly nociceptors) are less affected by daily stretching compared to SCI controls [6], showing that nociceptors contribute to stretch-induced locomotor deficits, likely by their activation.

We therefore sought to address the mechanisms by which this presumed nociceptive input may be acting on locomotor capacity. We considered the endogenous opioid system due to its known role in pain control, effects on locomotor output, and common clinical use after SCI. Opioid receptors (μ, δ, and κ) are expressed in DRG neurons and second-order neurons, with the highest levels expressed in the substantia gelatinosa of the dorsal horn [7]. All classes of opioid receptors are coupled to inhibitory G proteins which depress neuronal firing rates by affecting downstream targets such as adenylate cyclase and pre/post-synaptic ion channel properties [8]. Following SCI or any severe tissue injury, opioid systems in the spinal cord and brain facilitate an adaptive compensatory response to inhibit pain [9]. In addition, SCI can result in sensitization of the nociceptive response below the level of injury by several methods, including disruption of descending pathways, reduced spinal GABA levels, and altered Cl- equilibrium potential [10]. It has been suggested that endogenous opioids may counteract or mask a nociceptor hyperfunctional state, as the opioid antagonist naltrexone was shown to reinstate pain-like behaviors when administered months after nociceptor activation and sensitization by peripheral administration of complete Freund’s adjuvant [11]. Based on this background, we chose to use naltrexone to investigate the role of endogenous opioids in the stretching phenomenon. Naltrexone is a long-lasting antagonist of µ- and κ-opioid receptors, which are primarily located on peptidergic afferents and associated nociceptive circuitry [12, 13].

Methods

Experimental design

A summary of the experimental design is shown in Supplementary Fig. 1. Eight adult female Sprague-Dawley rats (225–250 g, Envigo, Indianapolis, IN) were acclimated in their home cages for two weeks then divided into two groups: drug-treated (n = 4) and vehicle-treated (n = 4). We performed stretching on weekdays during post-SCI weeks 6, 8, and 11. No stretching was performed during post-SCI weeks 7, 9, and 10. Stretch and non-stretch weeks were intercalated to assess the effect of drug on the known post-stretch recovery of locomotor function, as well as allow for sensory measurements in the absence of an acute effect of stretching.

Spinal cord injury

Animals were anesthetized with a ketamine (50 mg/kg)/xylazine (0.024 mg/kg) cocktail to provide a surgical level of anesthesia as confirmed by the absence of paw withdrawal reflexes in response to toe pinch. The back of each animal was shaved and treated to produce an aseptic incision area. Next, a midline incision was made over the lower thoracic spine, followed by a laminectomy of the T9 vertebra. A spinal cord contusion was delivered using the NYU Impactor set to 25 g-cm as previously described [14]. This magnitude of injury is generally considered to be “moderately-severe” and is selected for these studies because it is incomplete and spares a meaningful degree of motor function like many clinical injuries.

Drug dosing and administration

Naltrexone hydrochloride (Sigma–Aldrich Corporation, St. Louis, MO) was dissolved in saline and aliquoted for use in coded vials such that experimenters performing stretching, assessments, and analysis were blinded to the drug group. Prior to stretching or other sensory evaluations, experimental rats were given subcutaneous injections of the naltrexone-saline mixture (15 mg/kg) while control rats were given an equivalent volume of saline. Drug dose was selected based on its efficacy in other studies examining the impact of nociceptive sensory input on spinal motor function [15]. Drug/vehicle was given at least 15 min prior to stretching or evaluations based on evidence that the drug would be present in plasma at this time [16].

Stretching protocol and evaluation

Our daily stretching protocol has been thoroughly described previously [3,4,5,6, 17]. Briefly, the protocol consists of stretches applied to the 6 major hindlimb muscle groups bilaterally, each held at the end of range-of-motion for 1 min and repeated once for a total of ~24 min of stretching/day. During the stretching sessions, 4 researchers (blinded to experimental groups) were assigned such that the number of times a researcher stretched each rat was minimized while maintaining an approximately even distribution of rats from each experimental group that were stretched by an individual. Responses displayed after each stretch such as kicking, clonic-like spasticity, or withdrawal were noted. Responses were given a score of 1, 2, or 3 for mild/infrequent, moderate/frequent, and severe/very often as previously described [6]. To control for the effects of handling on locomotor function, animals were wrapped in a towel daily for 24 min, but not stretched, for 5 days (week 5 post-injury) when locomotor function had plateaued.

Locomotor assessments

Locomotor function was assessed using the BBB Open Field Locomotor Scale [18]. Baseline measurements were taken pre-injury followed by weekly post-injury assessments (weeks 1–4) to describe recovery. During week 5 (no stretching, towel wrapping), BBB scores were assessed Monday (pre-towel wrapping) and Friday (post-towel wrapping). During week 6, BBB scores were assessed 3 times per week at three daily time-points (Monday, Wednesday, and Friday) as follows: point 1 (pre-stretching, pre-injection), point 2 (pre-stretching, post-injection), and point 3 (post-stretching, post-injection). During week 8, BBB scores were assessed 3 times per week at two daily time-points as follows: point 1 (pre-stretching, pre-injection) and point 2 (post-stretching, post-injection). During week 11, animals were evaluated before stretching, and once after each stretching session (Table 1A, B, and D).

Table 1 Spasticity assessment during the (A) first, (B) second, and (D) third week of stretching, quantified on a scale of 1–3 during each stretch.
Full size table

Sensory assessments

A tail-flick assay was employed to assess thermal sensitivity using a radiant heat tail-flick device (Columbus Instruments, Columbus, Ohio, USA). A rat was placed on the platform and gently restrained, while the base of the tail was placed above the photocell for thermal stimulation. Intensity of the heat source was set to 10 out of 25, with a maximum duration of 10 s for each trial to avoid tissue damage. For each rat, a baseline tail-flick latency was measured three times with 30 s between measurements. Following baseline testing, each rat was administered a naltrexone dose (or saline) and re-tested 30 min later. Each sensory assessment was performed during week 9 post-injury when no stretching occurred.

Euthanasia and tissue histology

Animals were sacrificed ~4 h after the last stretching session using a ketamine (50 mg/kg)/xylazine (0.024 mg/kg) cocktail and transcardially perfused with 4% PFA. Spinal cords were dissected and post-fixed in 4% PFA for 2 h, then transferred to 30% sucrose for cryoprotection. The L1–L5 segments were separated from the rest of the spinal cord under a dissecting microscope and blocked in O.C.T. compound. Transverse sections were cut at 20 μm on a cryostat for immunohistochemical analysis into 6 blocks. For CGRP analysis, sections of L3 spinal cord were warmed and washed with PBS and 0.3% PBS-Triton (PBST). Slides were then blocked for 1 h with PBST, 5% bovine serum albumin (BSA), and 10% normal donkey serum (NDS). After blocking, slides were washed and incubated overnight with CGRP primary antibody (guinea pig polyclonal anti-CGRP, 1:1000, 20R-CP007, lot #P17101902, Fitzgerald), 5% BSA, and 5% NDS at 4 °C. Following incubation, sections were washed and incubated with a secondary antibody (Alexa Fluor 594-conjugated Donkey anti guinea pig, 1:200, 706–585–148, Lot #129041, Jackson ImmunoResearch Laboratories) for 1 h. Slides were coverslipped using Fluoromount.

For c-Fos analysis, sections of L2–L5 spinal cord were warmed and washed with PBS and 0.3% PBST for 30 min at room temperature. Sections were then incubated in antibody diluent reagent (Invitrogen # 00-3218, Lot1966331A) with c-Fos primary antibody (mouse monoclonal anti-c-Fos, 1:1000, cat#ab208942, Abcam) overnight at 4 °C. The next day, sections were washed once with 0.1% PBST and twice with PBS (10 min/wash at room temperature). Secondary antibody (donkey anti-mouse FITC, 1:250, cat#715-096-151, Jackson ImmunoResearch Laboratories) was combined with antibody diluent reagent (Invitrogen # 00-3218, Lot1966331A) for 1 h at room temperature. Following incubation, slides were washed once with 0.1% PBST and twice with PBS for 10 min/wash at room temperature. A secondary amplification process was performed using AlexaFluor 488 (mouse anti-FITC, 1:500, cat#200-542-037, Jackson ImmunoResearch Laboratories) in antibody diluent reagent (Invitrogen # 00-3218, Lot1966331A) for 1 h at room temperature. Sections were then washed once with 0.1% PBST for 10 min at room temperature, and once with Hoechst (1:1000) in PBST for 5 min. After a set of 3 washes (PBS for 10 min/wash at room temperature), Sudan Black B was dissolved (1 mg/ml) in 70% EtOH and applied for 15 min at room temperature. Slides were coverslipped using fluromount after a final set of 6 washes (PBS for 5 min/wash at room temperature).

Image analysis

Images of each section were acquired on an inverted microscope using a 20x objective for all histological analyses (resolution 0.33 µm/pixel). For CGRP analysis, a threshold was chosen based on control images, then applied to each image to quantify the CGRP-positive area of spinal cord cross section; the same threshold was applied to images from all animals. Total CGRP area was quantified for each dorsal horn excluding the dorsal root entry zones, which were removed manually from each image. Analysis of images was conducted using MatLab (MathWorks, Natick, MA).

For c-Fos analysis, a control brightness was established for c-Fos and DAPI, then this threshold was applied to each image using NIS-Elements (Nikon Corporation, Tokyo, Japan). Each c-Fos positive nucleus was manually identified and marked based on overlap between c-Fos and DAPI. Neuron locations were displayed as heatmaps using a custom MatLab program. To generate contour plots, neurons were first marked using Nikon Elements software. A custom-made MatLab program was then developed to reconstruct and normalize the position of labeled neurons across sections. A reference axis was created for each image with the origin centered on the central canal, the y axis parallel to the spinal cord midline, and the x axis orthogonal to the y axis. Contour/scatter plots were constructed using R. Distribution contours were created by calculating the two-dimensional kernel density (using the kde2d function in the MASS library), then connecting points of equal density values between 30% and 100% of the estimated density range in increments of 10%.

Statistical analyses

Data were analyzed using SPSS (IBM SPSS Statistics for Windows, Versions 26/27. IBM, Armonk, NY). The BBB locomotor scores, tail flick response latency, stretching spasticity and CGRP data were analyzed using repeated measures ANOVA. While this was a pilot study with a smaller sample size per group, overall the distributions were generally normal and when the variances were not equal both degrees of freedom were adjusted accordingly and strict post hoc t tests (Bonferroni t test for multiple comparisons) were employed. Nonparametric analysis of c–Fos data were compared between independent groups using Mann–Whitney U tests. No exclusion criteria were set a priori, and no animals or data points were removed from analysis.

Results

Acute response to stretching

Consistent with previous results [17], rats in both groups exhibited clonic-like responses and spasms in their hindlimbs during stretching. Upon the release of stretch, spasticity, air-stepping, and kicking again reemerge with greater intensity but usually last for only one or two seconds. It was noted that some rats in the current study exhibited stronger reactions during initial limb positioning and during stretch. In the final analysis stage, it was determined that those animals belong to the naltrexone group. Naltrexone treated animals also exhibited unusually prolonged spastic responses right after the release of a stretch and a writhing phenotype in the hindlimbs and tail that could last for almost a minute post-stretching (Supplementary Video 1). The most robust responses were seen following knee and hip stretches, although differences could be seen in ankle stretches as well. These reactions were evident during the first day of stretching, and the intensity of the responses persisted throughout all stretching sessions. Analysis of variance showed significant group differences in all stretches (Tibialis Anterior: F(1,22) = 11.7, p = 0.002, Gastrocnemius: F(1,30) = 63.6, p < 0.001, Hamstring: F(1,33) = 112.24, p < 0.001, Rectus Femoris: F(1,26) = 78.9, p < 0.001, Hip Abduction: F(1,32.2) = 67.7, p < 0.001, Hip Adduction: F(1,28) = 74.9, p < 0.001). Pairwise comparisons are shown in Table 1A, B, and D.

Locomotor function

Locomotor function as assessed by BBB scores plateaued at the beginning of week 5 post injury with no significant differences observed between groups. Any stress induced by wrapping the animals in towels alone (burrito) was insufficient to induce changes in locomotor function (Fig. 1A). Administration of naltrexone alone (without stretching) did not alter locomotor function at any timepoint tested. Analysis of variance confirmed significant differences between groups (F(1,91) = 3282.7, p < 0.001). There were no significant differences between the naltrexone and saline groups at any timepoint tested during the first week of stretching (Fig. 1B). However, naltrexone-treated animals had a significantly lower BBB scores following stretching during weeks 8 and 11 (Fig. 1C, D).

Fig. 1: BBB open field locomotor scores.
figure 1

Stretching causes a decrease in locomotor function following SCI. The locomotor dysfunction is accelerated following administration of naltrexone. A Baseline BBB measurements, as well as BBB measurements before and after the towel wrapping session without stretching. The first, second, and third weeks of stretching are shown in panels (B), (C), and (D), respectively. Significant differences were seen following 1 day of stretching during the second week of stretching (C) and following 2 days of stretching during the third week of stretching (D). Data are displayed as group mean (solid/dashed line) and individual values. All data were analyzed using a mixed model ANOVA between each time point, *p < 0.05.

Full size image

Response to afferent stimulation

Analysis of variance revealed a difference between saline and naltrexone groups (F(1,11) = 6.8, p = 0.024). However, post-hoc analysis showed no significant difference in tail-flick response latency between groups before and after drug/saline injections (Table 1C, p > 0.05).

Immunohistochemistry

Representative images of CGRP immunohistochemistry are shown for naltrexone-treated and control animals in Figs. 2A, B, respectively, while the CGRP+ area used for calculations is shown in Fig. 2C, D. There were no differences in CGRP-positive area between groups as demonstrated in Fig. 2E, F (F(1,36) = 0.058, p > 0.05).

Fig. 2: CGRP quantification and analysis.
figure 2

A, B show representative images from naltrexone and control animals, respectively. C, D images from A&B are shown in grayscale, with the area determined to be CGRP positive overlayed in red (E), (F) quantification of CGRP+ area in the dorsal horn of the spinal cord (L3 segment). Data shown as means + standard deviation (One-way ANOVA, Tukey HSD post hoc). Scale bars (white) for A–D = 200 μm, image resolution 0.33 μm/pixel.

Full size image

Representative images of c-Fos immunohistochemistry are shown for control and naltrexone-treated animals in Fig. 3A and B, respectively. Heatmap visualizations of total neuron distribution from spinal L2–L5 (Fig. 3C, D) demonstrate the spatial arrangement of the number of c-Fos positive nuclei (groupwise comparisons are in Fig. 3E, F). Due to differences between left and right sides, counts were split for Mann–Whitney analysis (n1 = n2 = 12, two-tailed significance for all tests). On the left side, differences were seen between groups at spinal segments L2 (U = 7, p < 0.001), L3 (U = 3, p < 0.001), L4 (U = 33.5, p < 0.05), while right sided differences were seen at spinal segments L2 (U = 20, p < 0.005), L3 (U = 22.5, p < 0.005), L4 (U = 13.5, p < 0.001), and L5 (U = 31.5, p < 0.05). Neuron distributions appear similar between both groups (Fig. 3G–N), although higher concentrations are seen in the naltrexone-treated group.

Fig. 3: c-Fos quantification and analysis.
figure 3

A, B Representative images from control and naltrexone animals (green:c-fos, blue: Hoechst, scale bar [white] = 200 μm), respectively, while (C) & (D) show heatmaps representing the overall distribution of neurons from L2–L5 (each “pixel-bin”:100 × 100 μm, scale bar [white] = 500 μm). E, F quantifications of neurons divided into spinal level, where each bar represents an average number of c-Fos positive neurons from 3 sections at each level. Data are displayed as group mean ± SD. All data were analyzed using a Mann–Whitney test, *p < 0.05, **p < 0.001. G–N Scatter-contour plots of for L2–L5 c-Fos neurons of control animals (G–J) and naltrexone animals (K–N), (scale bar = 500 μm).

Full size image

Discussion

Based on evidence that ablation of C-fibers reduces the detrimental locomotor effects of stretching after SCI [6], the goal of this study was to use a pharmacological approach to further elucidate the role of the nociceptive system in this phenomenon. Opioid receptors are widely distributed throughout nociceptive circuitry in the spinal cord as well as directly on sensory afferents, therefore endogenous opioids could play a role in modulating the information conveyed from nociceptors to populations of locomotor related interneurons. Consistent with previous studies, daily stretching led to a progressive decrease in locomotor function over the course of a single week. In addition, typical stretch responses such as air stepping, spasms, and kicking were present following stretching as seen in previous studies [17]. Administration of naltrexone prior to stretching increased the rate of locomotor decline following stretching and greatly increased the spasticity and spastic hindlimb responses during and following stretch. Previously, we found that nociceptor-depleted animals had significantly reduced clonic-like hindlimb responses to stretch as compared to nociceptor-intact stretched rats. In addition, the latter group of animals had a significant correlation between the number of c-Fos+ neurons in the intermediate and ventral horn gray matter and the clonic-like hindlimb response severity scores [6]. These data suggest that endogenous opioids mitigate the inhibitory effect on locomotion induced by clinically modeled hindlimb stretching.

Application of naltrexone in this study was systemic and could therefore be acting at multiple targets in the nociceptive circuitry, although the majority of µ-opioid receptors are expressed on central terminals of nociceptive afferents that primarily synapse in laminae I and II [19]. Endogenous opioid peptides activate μ-opioid receptors on these terminals and decrease neurotransmitter release by shortening repolarization time and duration of action potentials via inhibition of adenyl cyclase/reductions of cAMP and decreasing conductance of Ca2+ channels [20]. Blocking this effect with naltrexone would result in greater synaptic efficacy, therefore amplifying the effects of stretching. However, if this were the sole mechanism of action the nociceptors involved in the tail-flick reflex would also be altered and should lead to decreased response time, which was not the case. This discrepancy could simply be due to the difference in stimulus duration and scope; stretching is a prolonged, presumed-nociceptive stimulation over the course of 2 min for an entire muscle group, whereas the tail flick assay uses a brief and localized stimulus (2–3 s) with a pause between tests. This theory is supported by evidence demonstrating that opioid antagonists can modulate the response of deep dorsal horn interneurons to repetitive electrical stimulation of C-fibers, emphasizing the importance of timing and duration [21]. In addition, the dosage of naltrexone used here would antagonize not only μ-opioid receptors but also κ-opioid receptors. It has been demonstrated that κ-opioid receptors exert their effects post-synaptically by inhibiting NMDAR synaptic currents [22], which could provide an additional mechanistic difference between stretch and tail-flick induced changes. Furthermore, specific antagonism of κ-2 opioid receptors is sufficient to inhibit the sensory-dependent modification of motor output in a model of instrumental learning in the spinal cord [23].

Increased expression of c-Fos positive neurons in the naltrexone animals after two days of stretching corresponds with the quicker decrease in locomotor function seen in this group. The distribution of these neurons appears to be concentrated in the deep dorsal horn, intermediate gray matter, and ventral horn. This distribution of is similar to previous stretching induced expression of c-Fos [6], suggesting activation of the same circuitry. However, despite the increase in c-Fos positive neurons, administration of naltrexone prior to stretching did not appear to influence sprouting of peptidergic C-fibers below the level of injury compared to stretching without naltrexone. These results are in apparent contrast to our earlier findings showing that increases in c-Fos+ nuclei and CGRP+ afferent processes were both present in the lumbar dorsal horns of stretched animals [6]. It is possible that the naltrexone-exacerbated stretching effect was insufficient to cause a change in sprouting, or that the sprouting observed previously was only indirectly related to muscle stretching. For example, it has been demonstrated that forced exercise can reduce nonpeptidergic C-fiber sprouting [24], suggesting a likely role for inactivity in sprouting. In addition, the severity of injury has been shown to play a role in the amount of sprouting both rostral and caudal to the site of injury [25]. Further investigation into the relationship between nociceptor sprouting and the effects on the spinal circuitry following stretching are warranted. Emphasis should be placed on the action of specific opioid receptor subtypes (e.g., µ vs κ) and how their effects differ on pre- and post-synaptic nociceptive circuitry.

In conclusion, we have described the locomotor, behavioral, and histological effects of acutely blocking opioid receptors before hindlimb stretching post-SCI. Given generally negative effects associated with nociceptive afferent signaling following SCI, the evidence in the current study suggests that endogenous opioids play an important role in modulating nociceptive inflow and/or processing. Interestingly, in a clinical trial, patients with SCI who received naloxone infusions experienced increased frequency and duration of spasticity during the period of infusion and these effects were not seen in able-bodied subjects [26]. While we did not observe spontaneous spasticity in response to naltrexone over the course of this study, the responses to stretch following administration of naltrexone are like those described in humans. These findings highlight physiologically similar responses in human patients and animal models of SCI to opioid receptor blockade and thus emphasize the potential clinical relevance of our observations.

Whereas opioid receptor antagonism in the current study enhanced stretch-induced locomotor dysfunction, it is important to note that exogenous administration of opioid agonists post-SCI can also have a detrimental effect on locomotor recovery [27]. A potential physiological mechanism underlying this agonist-induced phenomenon is kappa opioid-mediated central sensitization resulting in excitotoxicity [28]. Other evidence suggests that endogenous opioids play a mixed role following SCI: pharmacological blockade can improve spinal cord perfusion and neurological function following SCI [29], however, endogenous opioids also play an important role to counter maladaptive hyperexcitability and states of hyperalgesia. While many of these published results appear contradictory, they highlight the complexity of the opioid system and our limited knowledge of the mechanisms of action, particularly in conditions like SCI which can broadly alter circuitry, receptor expression, and cell physiology.