Spinal Cord Stimulation Attenuates Paclitaxel-Induced Gait Impairment and Mechanical Hypersensitivity via Peripheral Neuroprotective Mechanisms in Tumor-bearing Rats

Abstract

Background:

Taxanes such as paclitaxel (PTX) induce dose-dependent chemotherapy-induced peripheral neuropathy (CIPN), which is associated with debilitating chronic pain and gait impairment. Increased macrophage-related proinflammatory activities have been reported to mediate the development and maintenance of neuropathic pain. While spinal cord stimulation (SCS) has been used for a number of pain conditions, the mechanisms supporting its use for CIPN remain to be elucidated. Thus, we aimed to examine whether SCS can attenuate Schwann cell– and macrophage-mediated neuroinflammation in the sciatic nerve of Rowlette Nude (RNU) rats with PTX-induced gait impairment and mechanical hypersensitivity.

Methods:

Adult male tumor-bearing RNU rats were used for this study examining PTX treatment and SCS. Gait and mechanical hypersensitivity were assessed weekly. Cytokines, gene expression, macrophage infiltration and polarization, nerve morphology, and Schwann cells were examined in sciatic nerves using multiplex immunoassay, bulk ribonucleic acid (RNA)-sequencing, histochemistry, and immunohistochemistry techniques.

Results:

SCS (50 Hz, 0.2 milliseconds, 80% motor threshold) attenuated the development of mechanical hypersensitivity (20.93 ± 0.80 vs 12.23 ± 2.71 grams, p<0.0096) and temporal gait impairment [swing (90.41 ± 7.03 vs 117.27 ± 9.71%, p<0.0076), and single stance times (94.92 ± 3.62 vs 112.75 ± 7.27%, p<0.0245)] induced by PTX (SCS+PTX+Tumor vs Sham SCS+PTX+Tumor). SCS also attenuated the reduction in Schwann cells, myelin thickness and increased the concentration of anti-inflammatory cytokine interleukin (IL)-10. Bulk RNA sequencing revealed differential gene expression after SCS, with 607 (59.2%) genes upregulated while 418 (40.8%) genes were downregulated. Notably, genes related to anti-inflammatory cytokines and neuronal growth were upregulated, while genes related to proinflammatory-promoting genes, increased M2γ polarization, and decreased macrophage infiltration and Schwann cell loss were downregulated.

Conclusion:

SCS may attenuate PTX-induced pain and temporal gait impairment, which may be partly attributed to decreases in Schwann cell loss and macrophage-mediated neuroinflammation in sciatic nerves.

1. Introduction

Chemotherapy-induced peripheral neuropathy (CIPN) is a common dose-limiting side effect for cancer patients treated with taxanes such as paclitaxel (PTX). Approximately 30% to 40% of patients experience persistent CIPN pain or other sensorimotor impairment 6 months after treatment.¹ Cancer survivors also often describe sensorimotor impairment, including gait, balance, and fine motor movements, as the most significant contributor to poor quality of life after chemotherapy.² The severity of CIPN appears to correlate with chemotherapy-induced gait impairment, which can manifest as declines in gait velocity, kinematics, step length, and time; increased cognitive attention to gait; and increased falls.^{3 4} These motor changes may be due to structural degeneration of sensory peripheral nerves that impair the peripheral sensation necessary for movement.^{2 4 5} However, structural changes are not always present, and functional changes, such as those resulting in mechanical hyperalgesia, may also cause gait impairment.^{2 5 6} Though sensory neuron mitochondrial dysfunction and axonal degeneration are thought to contribute to PTX-induced peripheral neuropathy (PIPN),⁷ emerging evidence suggests that it may also be caused by dedifferentiation of myelinating Schwann cells and subsequently increased macrophage infiltration.^{8 9}

According to the American Society of Clinical Oncology, CIPN has no effective preventive measures and only one treatment, which is mildly effective.¹⁰ Spinal cord stimulation (SCS) is a minimally invasive procedure that is most commonly applied for the treatment of complex regional pain syndrome, persistent spinal pain syndrome type 2, non-surgical refractory back pain, and more recently, painful diabetic peripheral neuropathy. SCS also has been applied for other types of peripheral neuropathies in humans and animal models, with reports of its use in patients who have established CIPN pain. SCS for gait impairment is a burgeoning field of research, and recent studies have demonstrated its potential to prevent and restore motor function related to stroke and spinal cord injury.^{11 12} Our previous work revealed that SCS can prevent the development of CIPN-associated mechanical and cold hypersensitivity in Sprague-Dawley rats, with an accompanying change in gene networks mediating synaptic plasticity and innate immune function in dorsal root ganglia (DRG).¹³ More recently, we found that SCS enhanced PTX’s anti-tumor efficacy and prevented PIPN pain in Rowlette Nude (RNU) rats (Crl:NIH-Foxn1^rnu) by modulating macrophages, fractalkine (CX3CL1), and inflammatory cytokines in DRG.¹⁴ Previous clinical studies and reports indicate the potential of SCS to improve not only pain but also physical function, including gait.^15–17 However, no prior work has evaluated SCS for CIPN-induced gait impairment, its effects on CIPN-induced macrophage polarization and infiltration, or its effect on Schwann cells in peripheral nerves.

Here, we aim to determine whether SCS can prevent PIPN-induced gait impairment by modulating Schwann cell- and macrophage-induced neuroinflammation in the peripheral nervous system. We used our established xenograft model of non-small cell lung cancer (NSCLC) in RNU rats.¹⁴ We used reflexive pain behavior; automated gait analysis; RNA sequencing; light microscopy; a multiplex immunoassay; and immunohistochemistry approaches to investigate these potential mediators of PIPN pain and gait impairment after SCS and PTX. We applied preemptive SCS (during PTX treatment) and late SCS (after PTX treatment). We hypothesized that SCS attenuates Schwann cell injury and promotes anti-inflammatory macrophage polarization to reduce PIPN-induced peripheral neuroinflammation and demyelination and thereby prevent PIPN-induced gait impairment and pain.

2. Materials and methods

2.1. Animals

Experiments were conducted with adult male athymic nude rats (RNU#316; Crl:NIH-Foxn1^rnu) aged 6 to 7 weeks (n=75; 208–252 g starting weight; Charles River Laboratories, Wilmington, MA, USA). The animals were housed at the Johns Hopkins University under a 12-hour light/dark cycle at 25 ± 2°C with 60% humidity and free access to food and water. All experimental protocols were approved by the Johns Hopkins University Animal Care and Use Committee and were carried out in accordance with the NIH Guide for the Care and Use of Laboratory Animals.

2.2. Study rigor and reproducibility

Animal cages were arranged randomly. Animals and group treatment were selected and performed at random. All replicates in these experiments refer to biological replicates. Experiments were conducted and data analyzed in a blinded fashion. We performed at least 2 independent repeated experiments for the biochemical and immunohistochemical assays, and all data were included in analyses.

2.3. Experimental design

The study was divided into two batches. Batch one (modeling experiments), rats were randomized to 1 of 4 groups: (1) no treatment (naïve; n = 5), (2) tumor only (Tumor; n = 5), (3) PTX only (PTX; n = 5), and (4) PTX with tumor (PTX+Tumor; n = 5). Batch two (SCS experiments), rats were randomized to 1 of 3 groups: (1) no treatment (naïve, n = 5), (2) sham SCS with PTX and tumor (sham SCS+PTX+Tumor; n = 11), and (3) SCS with PTX and tumor (SCS+PTX+Tumor; n = 11) (Fig. 1A; Supplementary Fig. 3). The sample size was determined based on our previous studies and experience with spinal cord stimulation in rats.^{13 14} Baseline gait and behavior were assessed in all animals at Week 0. Rats were then implanted with tumor cells and SCS electrodes as determined by their grouping. Rats designated to the PTX groups received a 2 mg/kg dose intraperitoneally (i.p.) every other day during Week 2 for a total PTX dose of 8 mg/kg. SCS (50 Hz, 0.2 ms, constant current, bipolar, continuous 8 hours/session per day at an intensity of 80% motor threshold, passive recharge) or sham SCS was applied during Weeks 2 (concurrently with PTX treatment) and 4 (after PIPN induction). We applied the 4th week SCS treatment to assess the prevention and treatment effects of SCS on gait impairment and mechanical hypersensitivity. In order to also study the more independent effects of SCS on gait impairment, we did not apply SCS during Week 3 to allow a wash-out period greater than 5 half-lives of PTX,¹⁸ as previously described.¹⁴ We selected a tonic SCS paradigm of 50Hz and 80% motor threshold to activate low-threshold A fibers and remain within tolerance threshold, as described in previous pre-clinical studies.^{13 14 19 20} All animals were euthanized with an overdose of isoflurane by Week 6, and the sciatic nerves were harvested.

Figure 1. Experimental protocol.

(A) A miniaturized SCS electrode and tumor cells were implanted at Week 0, followed by PTX treatment 2 weeks later (green bars). SCS was applied during Weeks 2 and 4 (purple bars). The sham SCS group had the SCS electrode implanted but did not receive stimulation (dashed purple bars). Baseline pain behavior and gait were assessed before electrode and tumor implantation and then weekly after PTX treatment. Sciatic nerves were excised at Week 6. (B) Schematic representation of the electrode implantation and mechanical hypersensitivity assessment. (C) Schematic representation of electrode. IHC, immunohistochemistry; i.p., intraperitoneal; PTX, paclitaxel; SCS, spinal cord stimulation.

2.4. Mechanical sensitivity test

Using the up-and-down method, we assessed the mechanical hypersensitivity in rats with von Frey filaments using the Dixon method.^{13 14} Paw withdrawal, flicking, or licking was counted as a positive response to the mechanical stimulus, monofilaments of varying force (grams).

2.5. Gait assessment

Catwalk XT system (Noldus Information Technology, Netherlands) was used to assess gait and pain in rats. Three compliant runs with at least 3 to 4 individual paw placements were recorded for each animal. A compliant run was defined as having a run duration between 1 and 5 seconds and maximum speed variation of 60%. Analysis was performed using the Catwalk XT 10.6 software. Sciatic functional index was determined using the following equation:

$$\mathrm{S}\mathrm{F}\mathrm{I}=-38.3\times \frac{\left(PLe-PLn\right)}{PLn}+109.5\times \frac{\left(TSe-TSn\right)}{TSn}+13.3\times \frac{\left(ITSe-ITSn\right)}{ITSn}$$

$\mathrm{P}\mathrm{L}$ = Print Length; $\mathrm{T}\mathrm{S}$ = Toe Spread; $\mathrm{I}\mathrm{T}\mathrm{S}$ = Intermediate Toe Spread; $\mathrm{n}$ = normal; $\mathrm{e}$ = experiment²¹

2.6. Bulk RNA sequencing

One hundred nanograms of total RNA isolated from frozen sciatic nerve segments were used to generate mRNA-seq libraries. Libraries were dual-indexed, pooled, and sequenced to an estimated depth of 60 million reads per sample. After alignment to the rn6 transcriptome, raw gene counts were normalized, log2 transformed, and analyzed using the default procedures in DESeq2. Complete details of RNA isolation, library preparation, sequencing, and analysis can be found in the supplementary methods.

2.7. Statistical analyses

Data were analyzed with GraphPad Prism 10.0.0 (GraphPad Software, La Jolla, CA). Two-way ANOVA followed by Tukey’s post hoc test was used for mechanical hypersensitivity and gait impairment tests. Unpaired two-tailed t tests with Welch’s correction were used to analyze sciatic nerve morphology, multiplex immunoassay, and immunohistochemistry data. All data are presented as mean ± SEM. P <0.05 was considered statistically significant.

3. Results

3.1. SCS attenuates PTX-induced mechanical hypersensitivity and temporal gait impairment

As a surrogate measure for neuropathic pain, we performed the von Frey filament test to determine mechanical hypersensitivity in the hind paws. Compared with the naïve group (19.88 ± 1.84 g), rats in the PTX (2.23 ± 0.13 g, p<0.0001) and PTX+Tumor groups (3.20 ± 0.54 g, p<0.0001), but not those in the Tumor group (17.80 ± 2.48 g, p<0.8165), had reduced mechanical paw withdrawal thresholds to von Frey filaments (Fig. 2A). Compared to the sham SCS+PTX+Tumor group, RNU rats in the SCS+PTX+Tumor group had less of a decrease in mechanical thresholds from baseline (20.93 ± 0.80 g vs 12.23 ± 2.71 g, p<0.0096) (Fig. 2B), suggesting that SCS may attenuate the development of PTX-induced mechanical hypersensitivity.

Figure 2. SCS attenuates PTX-induced mechanical hypersensitivity and temporal gait impairment in tumor-bearing RNU rats.

(A) PTX decreases the paw withdrawal threshold to von Frey filaments. (B) SCS attenuates mechanical hypersensitivity resulting from PTX treatment in tumor-bearing RNU rats. PTX increases percent change in (C) swing duration of hind limbs, step cycle duration, ground contact duration for a single hind paw (single stance), and sciatic functional index (a measure of hind limb function with sciatic nerve injury). (D) Gait representation of naïve, PTX, PTX+Tumor, and Tumor rats at Week 6 showing the footprint view (top panels, depicting spatial gait properties) and timing view (bottom panels, depicting temporal gait properties). PTX induced spatial (sciatic functional index) and temporal (swing, step cycle, and single stance durations) gait impairment. (E) SCS attenuates PTX-induced impairment of swing, step cycle, and single stance durations, but not sciatic functional index. (F) Gait representation of the naïve, sham, and SCS rats at Week 6 showing the footprint view (top panels) and timing view (bottom panels). SCS prevented PTX-induced temporal gait impairment but not spatial gait impairment. (A-C, E) PTX was administered during Week 2 (green bar). (B, E) SCS treatment occurred during Weeks 2 and 4 (purple bars). (A-C) Two-way ANOVA and Tukey’s post hoc test. *P<0.05, **P<0.01, ***P<0.001, Tumor, PTX, or PTX+Tumor versus naïve (n=5 per group); ##P<0.01, SCS+PTX+Tumor (n=8) versus sham SCS+PTX+Tumor (n=9). (E) Two-way ANOVA and Tukey’s post hoc test. #P<0.05, ##P<0.01, ###P<0.001, SCS+PTX+Tumor (n=7) or sham SCS+PTX+Tumor (n=6) versus naïve (n=4). Data represent mean + SEM. PTX, paclitaxel; SCS, spinal cord stimulation; g, grams.

To detect gait alterations, we analyzed spatiotemporal gait performance using a quantitative camera-based system (Catwalk XT, version 10.6 software, Noldus Information Technology, Netherlands). Compared with the naïve group, rats in the PTX and PTX+Tumor groups, but not those in the Tumor group, showed increased swing, step cycle, and single stance times, as well as increased sciatic functional indices (Fig. 2C–D). The PTX and PTX+Tumor groups did not differ in paw placement intensity (maximum contact mean intensity, mean intensity, and maximum contact maximum intensity) when compared with the naïve group (Fig. 2D). When compared with rats in the naïve group, rats in the sham SCS+PTX+Tumor group had increased swing (117.27 ± 9.79% vs 99.16 ± 3.64%, p<0.049), step cycle (106.38 ± 6.27% vs 84.29 ± 5.09%, p<0.025), and single stance times (112.75 ± 7.27% vs 89.96 ± 3.67%, p<0.009), and an increased sciatic functional index (275.59 ± 56.97% vs 59.6 ± 13.23%, p<0.0009) (Fig. 2E). Importantly, rats in the SCS+PTX+Tumor group had swing, step cycle, and single stance times that were similar to those of the naïve group but had an increased sciatic functional index similar to that of the sham SCS+PTX+Tumor group (Fig. 2E). Paw placement intensities of SCS- and sham SCS–treated rats were similar to those of the naïve rats (Fig. 2F). These findings suggest that SCS may attenuate PTX-induced temporal gait impairment in RNU rats.

3.2. Differential gene expression after SCS and PTX in tumor-bearing RNU rats

To determine how gene expression is altered in peripheral nerves after SCS, we compared RNA-seq data obtained from the sciatic nerve at 1 week after SCS or sham SCS and 3 weeks after PTX (Week 6). We identified 1025 (5.0%) DEGs in the sciatic nerve after SCS, compared to sham SCS. Of these 1025 DEGs, 607 (59.2%) genes were upregulated and 418 (40.8%) genes were downregulated (Fig. 3A). GO analysis of the 607 upregulated genes indicated that immune-mediated processes, including cytokine signaling, regulation of T-cell activation, and regulation of immune response (including B-cells, NK-cells, and macrophages) were among the biologic processes that were statistically enriched (Fig. 3B, C; Supplementary Table 1). The anti-inflammatory cytokine–promoting genes IL-10 receptor subunit alpha (IL10rα), IL-1 receptor antagonist (IL1rap), IL-13 receptor subunit alpha-2 (IL13ra2), early growth response 1 (Egr1), oncostatin M (Osm), and transforming growth factor beta 1 (Tgfb1) were upregulated after SCS (Fig. 3D; Supplementary Table 1). The 418 downregulated genes were involved with branched-chain amino acids, anterograde dendritic/axon transport, and localization to the presynaptic membrane (Supplementary Table 2). The proinflammatory cytokine-promoting genes lipocalin-2 (Lcn2), early growth response 2 (Egr2), atlastin-1 (Atl1), and S100 calcium-binding protein B (S100b) were among downregulated genes after SCS (Fig. 3D).

Figure 3. SCS increases anti-inflammatory gene expression and decreases neuroinflammation in the sciatic nerves of tumor-bearing RNU rats.

(A-D) Changes in gene expression in the sciatic nerve after PTX administration and SCS treatment. (A) Volcano plot of differential gene expression between SCS+PTX+Tumor (n=3) and sham SCS+PTX+Tumor (n=4) groups. Points highlighted in red indicate genes with differential expression at an unadjusted P value <0.05. (B) Bar plot comparing the significant GO biologic processes identified from the list of upregulated genes that are differentially expressed. (C) Scatterplot of GO biologic processes enriched in differentially expressed genes that were upregulated after SCS versus sham SCS. All the enriched terms are plotted, and each is represented by an individual point. Significant processes are outlined in black with clusters identified as immune activation and inflammatory response (gray cluster), cell proliferation (blue cluster), and neural regeneration (yellow cluster). (D) Heatmap of differentially expressed genes that are commonly expressed by Schwann cells and/or macrophages. (E) Multiplex immunoassay expression levels of the anti-inflammatory cytokine IL-10, MIP-3a, and proinflammatory cytokine GRO/KC at 1 week after SCS. Sham SCS+PTX+Tumor (n=4) versus SCS+Tumor+PTX (n=5). Unpaired t test with Welch’s correction. *P<0.05. Data represent mean + SEM. GO, gene ontology; IL, interleukin; PTX, paclitaxel; SCS, spinal cord stimulation; TNF, tumor necrosis factor.

Importantly, Schwann cell genes associated with positive regulation of myelination (S100b, Egr2) were downregulated while Schwann cell genes associated with negative regulation of myelination (Jun, Junb) were upregulated after SCS, which indicates decreased myelin damage and decreased need for remyelination (Fig. 3D). Neuroprotective and regeneration-promoting genes, aquaporin 1 (Aqp1) and ephrin type-B receptor 2 (Ephb2), respectively, were upregulated after SCS (Fig. 3D).

3.3. SCS alters expression of pro- and anti-inflammatory cytokines and chemokines in PTX-treated RNU rats

Based on findings from our RNA-seq study, we further investigated the role of SCS in modulating neuroinflammation. We assessed cytokine and chemokine levels in sciatic nerves 1 week after SCS or sham SCS and 3 weeks after PTX (Week 6). The levels of IL-10, MIP-3α, and GRO/KC were significantly increased in the SCS group, in comparison to the sham SCS group (Fig. 3E).

3.4. SCS prevents PTX-induced peripheral nerve injury

We further assessed peripheral nerve morphology using a toluidine blue staining technique to quantify myelination and axonal degeneration, and classified axons into small (1–4 μm), medium (4–7 μm), and large diameter (7–10 μm). When compared with the naïve group, rats in the PTX+Tumor group had decreased myelin thickness of small- and large-diameter axons. The small-diameter axons of the PTX+Tumor rats had an increased g-ratio (ratio of the inner axonal to the inner axonal plus outer myelinated area), but the medium-diameter axons had a decreased g-ratio (Fig. 4A–D), which may be due to increased myelin thickness or axonal degeneration. These results indicate that PTX induced gross morphologic injury across all small-, medium-, and large-diameter axons. In comparison with the sham SCS+PTX+Tumor group, rats that received SCS had greater myelin thickness of small-, medium-, and large-diameter axons and a smaller g-ratio of medium- and large-diameter axons (Fig. 4E–H). These results indicate that SCS may reduce PTX-induced demyelination.

Figure 4. SCS attenuates PTX-induced axonal degeneration and demyelination in tumor-bearing RNU rats.

(A-H) Representative light micrographs (10×, 100×) and analyses of sciatic nerve sections stained with toluidine blue. (A) Toluidine blue–stained cross-sectional micrographs of naïve and PTX+Tumor rats. (B) Magnified ROI from A showing myelination of individual axons. Myelin sheaths surrounding axons are shown in violet. (C) Automated segmentation of the ROI shown in B with axons depicted in yellow. (D) PTX induced axon degeneration and demyelination of small-, medium-, and large-diameter nerve axons. PTX reduced myelin thickness of small (1–4 μm) and large (7–10 μm) axons and increased the g-ratio of small and medium (4–7 μm) axons. (E) Toluidine blue–stained cross-sectional micrographs of SCS+PTX+Tumor and sham SCS+PTX+Tumor rats. (F) Magnified ROI from E showing myelination of individual axons. (G) Automated segmentation of the ROI shown in F with axons depicted in yellow. (H) SCS increased myelin thickness of small, medium, and large axons and decreased the g-ratio of medium and large axons. (A and C) Unpaired t test with Welch’s correction. **P<0.01, ***P<0.001, ****P<0.0001, PTX+Tumor (n=5) versus naïve (n=4); #P<0.05, ###P<0.001, ####P<0.0001, SCS+PTX+Tumor (n=5) versus sham SCS+PTX+Tumor (n=6). Data represent mean + SEM. PTX, paclitaxel; ROI, region of interest; SCS, spinal cord stimulation.

3.5. SCS promotes M2γ polarization of PTX-induced macrophage infiltration

Based on our RNA sequencing and multiplex immunoassay findings which suggested macrophage involvement after PTX treatment, we then assessed the infiltration and polarization of macrophages in the sciatic nerve using immunohistochemistry. CD68 is mostly expressed by M1 macrophages,^{22 23} whereas CD206 is expressed by M2a macrophages.²² Rats in the PTX and PTX+Tumor groups, but not those in the tumor-only group, had higher expression of M2a (CD206+CD68−), M2γ (CD206+CD68+), and M1 (CD206−CD68+) macrophages than did naïve rats (Supplementary Fig. 1A–B). However, the ratio of M2a or M2γ to M1 macrophages was decreased for all groups in comparison with naïve rats (Supplementary Fig. 1B). Thus, although tumor alone did not increase macrophages, both PTX and tumor appeared to independently decrease the ratio of M2a/M2γ to M1 macrophages. In comparison with the sham SCS+PTX+Tumor group, SCS-treated rats had decreased M2a, M2γ, and M1 macrophages (Fig. 5A–B). SCS did not alter the ratio of M2a to M1 macrophages but did increase the ratio of M2γ to M1 macrophages (Fig. 5B).

Figure 5. SCS inhibits PTX-induced Schwann cell loss, macrophage infiltration, and M1 polarization in tumor-bearing RNU rats.

(A) Representative confocal images of CD206+CD68− M2A, CD206+CD68+ M2γ, and CD206−CD68+ M1 macrophage expression in sciatic nerves of SCS- and sham SCS–treated rats. (B) SCS attenuated PTX-induced increases in proinflammatory (M1) and anti-inflammatory (M2A and M2γ) macrophages, with an overall shift toward M2γ polarization. (C) Representative confocal images of CX3CR1 and S100B expression in sciatic nerves of SCS- and sham SCS–treated rats. (D) SCS reduced CX3CR1 expression and attenuated Schwann cell loss. Unpaired t test with Welch’s correction; #P<0.05, ###P<0.001, SCS+PTX+Tumor versus sham SCS+PTX+Tumor (n=4 rats/group with 3 sections/rat). Data represent mean + SEM. PTX, paclitaxel; SCS, spinal cord stimulation.

3.6. SCS attenuates PTX-induced Schwann cell loss and macrophage infiltration

To further investigate the degree of macrophage infiltration and their relation to Schwann cells in sciatic nerves, we used immunohistochemical techniques to examine CX3CR1 and S100B expression. S100B is a common marker of Schwann cells in peripheral nerves. S100B is a common marker of Schwann cells in peripheral nerves. In comparison with the naïve group, rats in the PTX and PTX+Tumor groups had decreased S100B immunoreactivity (Supplementary Fig. 2A–B). Rats in the PTX, PTX+Tumor, and tumor-only groups had increased expression of CX3CR1, which indicates the degree of macrophage infiltration (Supplementary Fig. 2A–B). These decreases in S100B expression correlated with the increases in CX3CR1 expression, suggesting an inverse relationship between Schwann cells and infiltrating macrophages. Rats that received SCS had greater expression of S100B and less macrophage infiltration than did rats in the sham SCS group (Fig. 5G–H).

4. Discussion

Our findings show that, in PTX-treated tumor-bearing RNU rats, temporal gait impairment and mechanical hypersensitivity are due to PTX since the presence of tumors in the absence of PTX did not alter temporal gait or mechanical sensitivity. Collectively, our gene expression findings suggest that SCS may promote anti-inflammatory biologic processes in the sciatic nerve in PTX-treated tumor-bearing RNU rats, by upregulating anti-inflammatory cytokine-promoting genes and downregulating proinflammatory cytokine-promoting genes, possibly associated with Schwann cells and macrophages. They also indicate that SCS may exert a peripheral neuroprotective effect by attenuating PTX-induced demyelination and neuronal degeneration. The decreased neuroinflammation observed after SCS may be partly due to increased IL-10 levels in peripheral nerves. Our immunohistochemical findings in the sciatic nerve suggest that PTX, but not tumor, increases both proinflammatory (M1) and anti-inflammatory (M2a, M2ƴ) macrophage levels and that SCS may reduce macrophage infiltration, while increasing macrophage polarization to an M2ƴ phenotype. Furthermore, SCS may preserve myelination partly by decreasing PTX-induced Schwann cell loss and macrophage infiltration.

The involvement of mechanical allodynia in gait impairment has gained attention recently and presents CIPN-induced gait dysfunction as a sensory disorder rather than a motor disorder.⁵ Studies of chemotherapy- or diabetes-induced peripheral neuropathy have proposed a correlation between mechanical hypersensitivity and gait impairment.^{3 6 24} Indeed, temporal gait impairment may represent a compensatory reaction to altered mechanical sensation that ensures careful placement of the limbs during locomotion.³ We delineated gait impairment into spatial and temporal gait components after PTX treatment. Similar to previous studies,^{3 24} we observed that the temporal properties of gait (e.g., swing, step cycle, and single stance times) were impaired after PTX treatment. Unlike other neuropathy models (eg, chronic constriction injury, spinal cord injury)^25–27, we did not observe spatial gait changes (maximum contact mean intensity, mean intensity, maximum contact maximum intensity) except for sciatic functional index after PTX treatment This suggests that sensory axons were impaired while motor axons were spared and supports similar findings from another study of PTX-induced sensory polyneuropathy.²⁴ The sparing of motor axons and spatial gait in our model may correlate with the preserved myelination and g-ratio we observed in medium-diameter axons. Additionally, the temporal characteristics of gait seemed to worsen alongside increases in mechanical hypersensitivity after PTX treatment. The beneficial effect of SCS seems to relate to the temporal rather than spatial gait properties. SCS attenuated PTX-induced temporal gait impairment, and the improvement persisted as mechanical hypersensitivity resolved. Yet, sciatic functional index, a parameter of spatial gait that represents functional recovery of the sciatic nerve, was not improved with SCS. Because most parameters of spatial gait were spared with PTX and tumor induction, we were not able to discern the effects of SCS on spatial gait. Other studies have shown that SCS facilitates motor function through propriospinal and supraspinal input by engaging afferent neurons that provide excitatory input to motoneurons and interneurons.^{11 12}

Chemotherapy-induced gait impairment has been attributed to structural (eg, axonal degeneration)³ and functional disorders (eg, sensorimotor disability associated with sensory processing dysfunction).^{2 5} Our findings revealed Schwann cell loss with increased macrophage infiltration after PTX treatment. Schwann cells promote axon myelination that is crucial for nerve conductivity and impulse propagation. Because Schwann cells play an active role in mechanical perception owing to the presence of mechanosensitive ion channels (PIEZO1 and 2),²⁸ Schwann cell loss in the sciatic nerve may affect sensory transmission of gait information and result in temporal gait impairment. SCS may prevent temporal gait impairment by attenuating Schwann cell loss associated with demyelination and dysfunction of sensory transmission (mechanotransduction in particular). Schwann cell preservation may be associated with increased expression of Aqp1 and Ephb2 genes, which promote and maintain Schwann cell survival and function, and neuronal regeneration.^{29 30}

PTX-induced demyelination mostly impacted small- and large-diameter axons in sciatic nerves. Large and medium axons are known to transmit kinematic and kinetic impulses to the central nervous system for processing and generation of proprioception and tactile perception in response to mechanical stimuli. Hence, we propose that demyelination of large axons (A-fibers [Aα and Aβ] that conduct kinematic and kinetic sensation of body movement and posture to the central nervous system) mediates temporal gait impairment after PTX. Alternatively, demyelination and axonal degeneration may lead to mechanical hypersensitivity and impaired somatosensation required for motor function and gait.^{2 5} Accordingly, SCS may also attenuate temporal gait impairment by reducing demyelination and axonal degeneration. The upregulation Schwann cell genes associated with positive regulation of myelination (S100b and Egr2) and downregulation of Schwann cell genes associated with negative regulation of myelination (Jun and Junb) we observed with PTX, but attenuated by SCS, is indicative of demyelination and similarly occurs in other conditions associated with nerve injury.^31–33 This was further confirmed by decreased expression of S100B protein (a marker of Schwann cells) following PTX treatment.

Macrophage–Schwann cell interactions appear to mediate neuroinflammation involved in the development and maintenance of CIPN pain and gait impairment. Upregulation of S100b and Lcn2 gene expression has been implicated in the development of neuroinflammation.^{34 35} Under pathologic conditions, macrophages release S100B, which enhances proinflammatory activities,³⁵ and LCN2 promotes thermal and mechanical hypersensitivities by inducing chemokine expression.³⁶ In our current study, PTX induced upregulation of S100b and Lcn2 genes (possibly via Schwann cell injury), which may have subsequently enhanced secretion and release of chemokines that increased macrophage infiltration into the sciatic nerve. PTX may increase the expression of CX3CR1, indicating macrophage infiltration in response to nerve injury.

SCS may intervene by attenuating Schwann cell loss and S100b, Egr2, and Lcn2 gene expression, thereby promoting an anti-inflammatory milieu (possibly via M2ƴ) and reducing pro-inflammatory macrophage infiltration. The ratio of M2 to M1 macrophages is most likely key to the effects of injury-induced inflammation on tissues and organs, including peripheral nerves. At the onset of injuries leading to neuroinflammation, the M1 phenotype surges²² and releases proinflammatory mediators that can lead to mechanical hypersensitivity. However, the M2 phenotype increases late after onset and elevates the release of anti-inflammatory mediators that are involved with pain resolution.³⁷ In our study, PTX increased both M1 and M2 macrophages. However, the M2-to-M1 ratio decreased, indicating an overall shift toward a pro-inflammatory phenotype. In contrast, SCS increased the M2-to-M1 ratio, indicating an overall shift toward an anti-inflammatory phenotype. Our findings indicate that M2γ macrophages may be more critical to SCS effects than the M2a subtype because they were the primary contributors to the increased M2-to-M1 ratio in the sciatic nerve. The M2a macrophage subtype is associated with wound healing, is activated by IL-13, and contributes to anti-inflammation by releasing IL-10.³⁷ The M2γ macrophage subtype mediates tissue regeneration and anti-inflammation by inhibiting Th1 cell activities and producing anti-inflammatory cytokines.³⁸ Increases in the expression of anti-inflammatory promoting cytokine genes (IL-10, IL-1rn, IL-13, Tgfb1) and IL-10 after SCS may prevent the neuroinflammatory effects of PTX in the sciatic nerve. The IL-10 and TGFβ1 released by M2 macrophages are thought to be critical for inflammation suppression, tissue regeneration, and pain resolution.^{22 37} Thus, SCS may attenuate mechanical hypersensitivity and temporal gait impairment partly by increasing M2γ macrophages and anti-inflammatory cytokines.As developments in SCS miniaturization and less invasive modalities are rapidly emerging, our findings may help guide early human experimental studies of less costly and invasive modalities of electrical stimulation for CIPN prevention and treatment. An increased mechanistic understanding of factors mediating myelination, Schwann cell proliferation, M2ƴ polarization, and neuroinflammation in the peripheral nervous system may serve to identify novel molecular targets for new CIPN therapeutics. Objective measures of these CIPN pain-related mechanisms could also help predict CIPN pain development and assess treatment efficacy. This work will aid future investigations of electrical stimulation for other circumstances associated with a risk of developing chronic pain or gait impairment.

One limitation of our study is that we did not examine the sexual dimorphic effects of SCS. Additionally, we did not examine Schwann cell infiltration or specific Schwann cell subtypes (ie, myelinated and unmyelinated). We also did not examine the effect of variable SCS amplitude, frequency, pulse width, or recharge paradigms. While reflexive measures of pain and gait assessment can provide insight into “pain-like” behaviors or nociception, we will incorporate operant measures in future studies to better assess non-stimulus-evoked nociception. Because we focused on neuroimmune changes at 3 weeks post-PTX treatment, additional investigation into the temporal changes of macrophage infiltration and polarization in relation to CIPN will help identify the best time for SCS intervention.

5. Conclusion

SCS attenuates PTX-induced mechanical hypersensitivity and temporal gait impairment, which may involve inhibition of neuroinflammation and Schwann cell loss in the sciatic nerve (Fig. 6). Furthermore, SCS promotes the polarization of macrophages to an M2γ subtype that promotes its anti-inflammatory response to PTX-induced neuroinflammation in the sciatic nerve.

Figure 6. Schematic diagram of working hypothesis.

SCS attenuates PTX-induced mechanical hypersensitivity and gait impairment via decreased Schwann cell loss and neuroinflammation, and increased anti-inflammatory (M2γ) macrophage polarization in the sciatic nerve.

What is already known on this topic –

Spinal cord stimulation (SCS) has been reported to attenuate painful paclitaxel (PTX)-induced peripheral neuropathy but its effects on PTX-induced gait impairment and the peripheral mechanisms by which SCS may be mediated are yet to be elucidated.

What this study adds –

This study in rats revealed that SCS attenuates PTX-induced gait impairment and mechanical hypersensitivity through increased anti-inflammatory promoting gene expression, anti-inflammatory cytokines and M2ƴ polarization, as well as decreased Schwann cell loss in the sciatic nerves.

How this study might affect research, practice or policy –

This study adds to the growing literature on SCS for gait dysfunction and highlights its potential peripheral neuroprotective mechanisms, particularly in the setting of painful chemotherapy-induced peripheral neuropathy.

Data availability

The data that support the findings of this study are available, upon reasonable request, from the corresponding author.