Multilineage differentiating stress-enduring cells alleviate neuropathic pain in mice through TGF-β and IL-10-dependent anti-inflammatory signaling

TL;DR

Multilineage-differentiating stress-enduring (Muse) cells reduce neuropathic pain in mice by suppressing inflammation through TGF-β and IL-10.

Contribution

Muse cells show superior analgesic effects compared to BMSCs via TGF-β and IL-10-dependent anti-inflammatory mechanisms.

Findings

Muse cells effectively reduce neuropathic pain in mouse models better than low-dose BMSCs.

Muse cells inhibit spinal cord neuroinflammation via TGF-β and IL-10 secretion.

Muse cells migrate to injured dorsal root ganglion via the CCR7–CCL21 chemotactic axis.

Abstract

Neuropathic pain is a chronic condition characterized by damage to and dysfunction of the peripheral or central nervous system. There are currently no effective treatment options available for neuropathic pain, and existing drugs often provide only temporary relief with potential side effects. Multilineage-differentiating stress-enduring (Muse) cells are characterized by high expansion potential, a stable phenotype and strong immunosuppression. These properties make them attractive candidates for therapeutics for neuropathic pain management. Muse cells from different species demonstrated analgesic potential by reversing chronic constriction injury model (CCI)-induced neuropathic pain. Protein profiling revealed a high degree of similarity between Muse cells and bone marrow stromal cells (BMSCs). The intrathecal injection of Muse cells effectively reduced neuropathic pain in various mouse models, resulting in better analgesic effects than the administration of equivalent low doses of BMSCs. Immunohistochemical analysis and quantitative real-time PCR revealed the ability of Muse cells to inhibit spinal cord neuroinflammation caused by spared nerve injury model. In addition, Transwell and ELISA revealed that Muse cells migrated through the injured dorsal root ganglion via the CCR7–CCL21 chemotactic axis. In addition, the secretion of transforming growth factor-beta (TGF-β) and interleukin 10 (IL-10) by Muse cells was identified as the mechanism underlying the analgesic effect of Muse cells. The capacity of Muse cells to mitigate neuroinflammation and produce analgesic effects via the modulation of TGF-β and IL-10 underscores their potential as promising therapeutic approaches for the treatment of neuropathic pain.

Figures

B: F (2, 6) = 71.43, *p* < 0.0001, sham *versus* PBS, *p* < 0.0001, sham *versus* Muse cells, *p* = 0.0019, PBS *versus* Muse cells, *p* < 0.0001; Figure 4C: F (2, 6) = 187.7, *p* < 0.0001, sham *versus* PBS, *p* < 0.0001, PBS *versus* Muse cells, *p* = 0.0002). *D*, qPCR showing the expression levels of IL-1β, IL-6, and TNF-α mRNAs in the spinal cord dorsal horn of sham and SNI mice with or without Muse cells treatment (IL-1β: F (2, 6) = 25.81, *p* = 0.0011, sham *versus* PBS, *p* = 0.0009, sham *versus* Muse cells, *p* = 0.0384, PBS *versus* Muse cells, *p* = 0.0189; IL-6: F (2, 6) = 16.45,

Full text

Neuropathic pain caused by multiple insults to the nervous system is a refractory intractable pain (1). Current treatments for neuropathic pain are insufficient. Bone marrow stromal cells (BMSCs), a population of progenitor cells that exist in adult bone marrow, have emerged as a major source for cell–based therapies because they are easy to collect and have high proliferation potential and strong immunosuppressive properties (2). Mounting evidence suggests that the transplantation of BMSCs can achieve long-term relief of neuropathic pain (3, 4), which provides a new method for the clinical treatment of neuropathic pain. Although the analgesic effects of BMSC transplantation on animal pain models are encouraging, there are still problems to overcome before their clinical application. The greatest barrier is the difficulty in establishing quality control standards for BMSCs that are suitable for clinical use.

BMSCs are a group of cells with nonuniform antigenic phenotypes, and their entire composition has not been fully identified (5). Therefore, it is unclear which cell subsets of BMSCs have analgesic effects. During the process of culture and amplification, the proportions of different subsets of BMSCs always change. Because the BMSCs used in different laboratories may contain different proportions of different cell subsets, inconsistent research results on the analgesic effects of BMSCs in different laboratories may occur. Even in the same laboratory, BMSCs may undergo significant changes in cell characteristics, such as loss of multilineage differentiation and analgesic effects, after multiple passages. (6). To date, no commercial BMSCs have been used to treat chronic pain in the clinic. At present, most clinical trials of BMSC transplantation have used acute isolation of autologous BMSCs or monocytes, and patients have to undergo considerable pain to extract enough bone marrow for transplantation (7). In addition, BMSC-derived products that include multiple cell subsets not only increase production and use costs but also may reduce analgesic effects due to the antagonistic effects of certain unknown cell subsets (8, 9). In summary, in a mixture of BMSCs, obtaining a single type of cell subset that is safe, stable, and durable is the key to promoting the clinical translation of the research of BMSCs in the treatment of neuropathic pain.

Multilineage-differentiating stress-enduring (Muse) cells, which were first described in 2010, constitute a small percentage of BMSCs, can generate cells representative of all three germ layers from a single cell, and they are nontumorigenic and self-renewable (10, 11). Muse cells are particularly unique compared with other stem cells in that they efficiently migrate and integrate into damaged tissue when supplied into the bloodstream and spontaneously differentiate into cells compatible with the homing tissue (12, 13). Therefore, current research on Muse cells has focused on their ability to differentiate and repair. Muse cells have emerged as a novel source for cell-based therapies, with a diverse spectrum of potential clinical applications. Due to their high expansion potential, genetic stability, stable phenotype, and strong immunosuppressive properties, Muse cells can be exploited for successful autologous and heterologous transplantation without the need of immune suppressants (10). However, whether Muse cells, like BMSCs, play an analgesic role by secreting a variety of anti-inflammatory factors has not yet been reported. In this study, we investigated whether Muse cells can be used to treat neuropathic pain.

Results

Characterization of Muse cells in human and rodents

As previously described (11), Muse cells were obtained from cultured BMSCs by long–term trypsin incubation. Eight hours of trypsin treatment resulted in many dead cells, and the surviving cells were collected and cultured under suspension culture conditions (11). These cells generated cell clusters after 1 week and could be subcultured many times (Fig. 1A). These cells were confirmed to be Muse cells by their specific cell surface markers (SSEA-3 and CD105; Fig. 1B). Flow cytometry analysis revealed that they accounted for 0.58% of the BMSC population (Fig. S1A). Muse cells can also be cultured in adherent media and have a cell shape similar to that of BMSCs (Fig. 1C). In this study, we cultured human, rat, and mouse Muse cells and found that they had the same characteristic features under suspension culture and adherent culture conditions (Fig. 1, A–C).Figure 1**Characterization of Muse cells.**A, representative images of single Muse cell from human, rat, and mouse formed cell cluster after 1 week of suspension culture. The scale bar represents 100 μm. B, human Muse cell cluster expressed SSEA-3 and CD105. The scale bar represents 100 μm. C, representative images of Muse cell from human, rat, and mouse in adherent culture for 1 week. The scale bar represents 100 μm. Muse, multilineage-differentiating stress-enduring; SSEA-3, stage–specific embryonic antigen–3; CD105, endoglin, cell membrane glycoprotein.

Inhibition of neuropathic pain in mice by a single intrathecal injection of Muse cells

To test the hypothesis that Muse cells alleviate neuropathic pain, various mouse models including nerve injury, diabetes, and chemotherapy-induced neuropathic pain were used. First, the analgesic efficacy of Muse cells from different species was examined in a sciatic nerve chronic constriction injury (CCI)-induced persistent neuropathic pain mouse model (14, 15). Twelve days after CCI, the same number (2.5 × 10^5^ cells) of human, rat, or mouse Muse cells were intrathecally injected, all of which rapidly (<1 day) and long lastingly (>42 days) inhibited CCI-induced mechanical allodynia (Fig. 2A) and thermal hyperalgesia (Fig. 2B). There was no significant difference in the analgesic effects among the three types of Muse cells. Notably, these experiments were performed in wild-type mice, and no immunosuppressive agents were used.Figure 2**A single intrathecal injection of Muse cells inhibit neuropathic pain in mice.**A and B, intrathecal injection of the same amount (2.5 × 10^5^ cells) of human, rat, or mouse Muse cells, 12 days after CCI, all produced a rapid (<1 day) and long-lasting (>42 days) inhibition of CCI-induced mechanical allodynia (A) and thermal hyperalgesia (B). n = 4 or 5 mice/group. (Fig. 2A: F (33,180) = 2.627, p < 0.0001; 1 day, Vehicle versus hMuse cells, 2.5 × 10^5^, i.t., p = 0.0069; 3 days, Vehicle versus rMuse cells, 2.5 × 10^5^, i.t., p = 0.0017, Vehicle versus hMuse cells, 2.5 × 10^5^, i.t., p = 0.0338; 7 days, Vehicle versus mMuse cells, 2.5 × 10^5^, i.t., p = 0.0481, Vehicle versus hMuse cells, 2.5 × 10^5^, i.t., p = 0.0025; 10 days, Vehicle versus mMuse cells, 2.5 × 10^5^, i.t., p = 0.0061, Vehicle versus hMuse cells, 2.5 × 10^5^, i.t., p = 0.0064; 14 days, Vehicle versus rMuse cells, 2.5 × 10^5^, i.t., p = 0.0003, Vehicle versus hMuse cells, 2.5 × 10^5^, i.t., p = 0.022; 17 days, Vehicle versus mMuse cells, 2.5 × 10^5^, i.t., p = 0.0385, Vehicle versus hMuse cells, 2.5 × 10^5^, i.t., p = 0.0141; 22 days, Vehicle versus mMuse cells, 2.5 × 10^5^, i.t., p = 0.0208, Vehicle versus rMuse cells, 2.5 × 10^5^, i.t., p = 0.0035,Vehicle versus hMuse cells, 2.5 × 10^5^, i.t., p = 0.0013; 28 days, Vehicle versus mMuse cells, 2.5 × 10^5^, i.t., p = 0.0013, Vehicle versus rMuse cells, 2.5 × 10^5^, i.t., p < 0.0001, Vehicle versus hMuse cells, 2.5 × 10^5^, i.t., p < 0.0001; 35 d/42d, Vehicle versus mMuse cells, 2.5 × 10^5^, i.t., p < 0.0001, Vehicle versus rMuse cells, 2.5 × 10^5^, i.t., p < 0.0001, Vehicle versus hMuse cells, 2.5 × 10^5^, i.t., p < 0.0001; Figure 2B: F (33,192) = 4.248, p < 0.0001; 1 d/3 d/5 d/7 d/10 days, Vehicle versus mMuse cells, 2.5 × 10^5^, i.t., p < 0.0001, Vehicle versus rMuse cells, 2.5 × 10^5^, i.t., p < 0.0001, Vehicle versus hMuse cells, 2.5 × 10^5^, i.t., p < 0.0001; 14 d/21 days, Vehicle versus mMuse cells, 2.5 × 10^5^, i.t., p < 0.0001, Vehicle versus rMuse cells, 2.5 × 10^5^, i.t., p = 0.0002, Vehicle versus hMuse cells, 2.5 × 10^5^, i.t., p < 0.0001; 28 days/35 days, Vehicle versus mMuse cells, 2.5 × 10^5^, i.t., p < 0.0001, Vehicle versus rMuse cells, 2.5 × 10^5^, i.t., p < 0.0001, Vehicle versus hMuse cells, 2.5 × 10^5^, i.t., p < 0.0001; 42 days, Vehicle versus mMuse cells, 2.5 × 10^5^, i.t., p < 0.0001, Vehicle versus rMuse cells, 2.5 × 10^5^, i.t., p = 0.0003, Vehicle versus hMuse cells, 2.5 × 10^5^, i.t., p < 0.0001). C and D, intrathecal injection of human Muse cells inhibited STZ-induced mechanical allodynia (C) and thermal hypoalgesia (D). n = 7 mice/group. (Vehicle versus Muse cells, 2.5 × 10^5^ cells, i.t., Fig. 2C: F (14,180) = 54.34, p < 0.0001; 14 days: p = 0.0036, 21 d/28 d/33 d/37 d/47 d/56 d/63 d/71 d/85 days: p < 0.0001; Figure 2D: F (14,180) = 5.09, p < 0.0001; 10 d/14 d/21 d/28 d/33 d/37 d/47 days: p < 0.0001; 56 days: p = 0.0156). E, intrathecal injection of human Muse cells inhibited paclitaxel-induced mechanical allodynia. n = 6 mice/group. (Vehicle versus Muse cells, 2.5 × 10^5^ cells, i.t., F (11, 120) = 6.166, p < 0.0001;5 h, p = 0.0196; 1 d/8 d/14 d/21 d/35 d/42 d, p < 0.0001; 56 days, p = 0.0227). Statistical significance was indicated by ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, as determined by two-way ANOVA, followed by Bonferroni’s post hoc test. All data are expressed as the mean ± SD. BL, baseline. Muse, multilineage-differentiating stress-enduring; CCI, chronic constriction injury; BMSC, bone marrow stromal cell; STZ, streptozotocin.

Diabetes associated with peripheral neuropathy is a serious health problem. The streptozotocin (STZ)-induced diabetes model elicits mechanical allodynia and heat hypoalgesia and is recognized as one of the most difficult types of sensorial neuropathy to treat (16). Behavioral signs of mechanical allodynia and thermal hypoalgesia were evident 2 weeks after diabetes model induction. Mechanical allodynia was maintained for 8 weeks, at which time diabetic mice showed a gradual loss of mechanical sensitivity, while thermal hypoalgesia persisted during the experimental period of 15 weeks. To determine whether Muse cells induce therapeutic effects on diabetic sensory neuropathy, mice were treated with human Muse cells (2.5 × 10^5^) (treated with PBS as vehicle group) via the intrathecal (i.t.) route 20 days after diabetes induction, when sensory neuropathy was fully established. Two weeks after administration, neuropathic mice treated with Muse cells exhibited antinociceptive effects in response to mechanical stimuli (Fig. 2C). The antinociceptive effect of Muse cells was progressive, peaking 3 weeks after treatment, when complete reversal of mechanical allodynia was achieved. Importantly, the progression of sensory neuropathy, indicated by the late loss of mechanical sensitivity, was completely prevented in Muse cell-treated mice. In addition, Muse cell treatment reversed the thermal hypoalgesia of neuropathic model mice from 10 days after administration until 8 weeks after treatment (Fig. 2D).

Paclitaxel-induced peripheral neuropathy is a neuropathic pain model that represents a frequent and serious consequence of chemotherapy agents (17). In this model, paclitaxel–induced mechanical allodynia in mice reached the maximum level 1 day after the induction of chemotherapy, lasted for approximately 3 weeks, and then gradually returned to normal levels 8 weeks after induction (Fig. 2E). To determine whether Muse cells induce therapeutic effects in a chemotherapy-induced neuropathic pain model, mice were treated with human Muse cells (2.5 × 10^5^) via the i.t. route 1 day after paclitaxel (2 mg/kg, 4 times every other day) treatment. The antinociceptive effect of Muse cells on mechanical stimuli was observed 5 h after administration and gradually peaked at 5 weeks before being maintained until the end of the evaluation period (Fig. 2E). At 1 day following paclitaxel administration, two doses of human BMSCs or human Muse cells (1 × 10^5^ and 2.5 × 10^5^) were administered intrathecally to mice. The results demonstrated that i.t. delivery of the same doses of BMSCs or Muse cells elicited comparable analgesic effects, with no significant difference observed between the two cell types (Fig. S2).

Taken together, these results indicate that the i.t. injection of Muse cells has long-term analgesic effects on neuropathic pain model mice.

iTRAQ analysis of Muse cells and BMSCs

Proteomic strategies represent powerful tools for the global investigation of many cellular proteins. To explore the potential functions of Muse cells, quantitative proteomic analysis based on isobaric tags for relative and absolute quantitation (iTRAQ) labeling was conducted to obtain an unbiased view of the proteomic profiles of human Muse (hMuse) cells and hBMSCs. With a false discovery rate of less than 1%, a total of 4368 proteins were identified. Among the 4368 identified proteins, 3624 proteins were identified by two or more unique peptides, and the remaining 744 proteins were identified by one unique peptide. To characterize the differences between Muse cells and BMSCs, altered proteins were determined according to their relative protein expression (Muse cells/BMSCs). According to the threshold values for downregulated and upregulated proteins (≤0.20 and ≥ 4.44, respectively and 90% confidence interval), a total of 105 downregulated and 59 upregulated proteins were found in Muse cells (Fig. 3A). Thus, the protein maps of Muse cells and BMSCs were 96.25% similar. The differentially expressed proteins were classified into 14 categories according to their main biological functions collected from relevant literature in PubMed, including metabolism, cell proliferation and growth, transcription and translation, transport, migration, cytoskeleton, differentiation, cell adhesion, apoptosis, immune response, signal transduction, development, and the cell cycle (Fig. 3B).Figure 3**Muse cells have a stronger and more stable analgesic effect than BMSCs.**A, distribution of mean ratios of 4368 identified proteins, as measured by three independent iTRAQ experiments. Ratios were calculated as human Muse cells versus human BMSCs. The 90% confidence intervals are indicated by vertical lines in the plot. A total of 105 down-regulated and 59 up-regulated proteins were found in Muse cells. B, functional categories of differentially expressed proteins between human Muse cells versus human BMSCs. C, intrathecal injection of different number of human BMSCs or Muse cells all inhibited SNI-induced mechanical allodynia. The analgesic effects of injection of low- (1 × 10^4^) and middle-does (5 × 10^4^) Muse cells were stronger and more durable than the same dose of BMSCs. n = 5 mice/group. (F (48, 252) = 6.617, p < 0.0001; 7 days: Vehicle versus BMSCs, 2.5 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus Muse cells, 2.5 × 10^5^ cells, i.t., p < 0.0001; 14 days, Vehicle versus Muse cells, 5 × 10^4^ cells, i.t., p = 0.0002, Vehicle versus BMSCs, 1 × 10^5^ cells, i.t., p = 0.0019, Vehicle versus Muse cells, 1 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus BMSCs,2.5 × 10^5^ cells,i.t., p < 0.0001, Vehicle versus Muse cells, 2.5 × 10^5^ cells, i.t., p < 0.0001; 21 days: Vehicle versus BMSCs,5 × 10^4^ cells, i.t., p = 0.0003, Vehicle versus Muse cells, 5 × 10^4^ cells, i.t., p = 0.0002, Vehicle versus BMSCs, 1 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus Muse cells, 1 × 10^5^ cells,i.t., p < 0.0001, Vehicle versus BMSCs, 2.5 × 10^5^ cells,i.t., p < 0.0001, Vehicle versus Muse cells, 2.5 × 10^5^ cells, i.t., p < 0.0001; 28 days: Vehicle versus BMSCs, 5 × 10^4^ cells, i.t., p < 0.0001, Vehicle versus Muse cells, 5 × 10^4^ cells, i.t., p = 0.0002, Vehicle versus BMSCs, 1 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus Muse cells, 1 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus BMSCs, 2.5 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus Muse cells, 2.5 × 10^5^ cells, i.t., p < 0.0001; 35 days: Vehicle versus BMSCs, 5 × 10^4^ cells, i.t., p = 0.004, Vehicle versus Muse cells, 5 × 10^4^ cells, i.t., p = 0.0002, Vehicle versus BMSCs, 1 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus Muse cells, 1 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus BMSCs, 2.5 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus Muse cells, 2.5 × 10^5^ cells, i.t., p < 0.0001, BMSCs, 5 × 10^4^ cells, i.t. versus Muse cells, 5 × 10^4^ cells, i.t., p = 0.0319; 42 days: Vehicle versus Muse cells, 5 × 10^4^ cells, i.t., p < 0.0001, Vehicle versus Muse cells, 1 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus BMSCs, 2.5 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus Muse cells, 2.5 × 10^5^ cells, i.t., p < 0.0001, BMSCs, 5 × 10^4^ cells, i.t. versus Muse cells, 5 × 10^4^ cells, i.t., p = 0.0003, BMSCs, 1 × 10^5^ cells, i.t. versus Muse cells, 1 × 10^5^ cells, i.t., p = 0.0017). D, the analgesic effects of different passages (P0 and P20) of human BMSCs and Muse cells on SNI-induced mechanical allodynia. P20 Muse cells produced effective and long-term analgesic effects, whereas P20 BMSCs only produced mild and transient analgesic effects. n = 4 or 5 mice/group. (F (40, 198) = 5.394, p < 0.0001; 7 days: Vehicle versus Muse cells p0, 2.5 × 10^5^ cells, i.t., p = 0.047; 14 days: Vehicle versus BMSCs p0, 2.5 × 10^5^ cells, i.t., p = 0.0135, Vehicle versus Muse cells p0, 2.5 × 10^5^ cells, i.t., p = 0.0008, Vehicle versus Muse cells p20, 2.5 × 10^5^ cells, i.t., p = 0.001; 21 days, Vehicle versus BMSCs p0, 2.5 × 10^5^ cells, i.t., p = 0.0004, Vehicle versus Muse cells p0, 2.5 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus Muse cells p20, 2.5 × 10^5^ cells, i.t., p < 0.0001, BMSCs p0, 2.5 × 10^5^ cells, i.t. versus Muse cells p20, 2.5 × 10^5^ cells, i.t., p = 0.0021; 28 d/35 d/49 days, Vehicle versus BMSCs p0, 2.5 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus Muse cells p0, 2.5 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus Muse cells p20, 2.5 × 10^5^ cells, i.t., p < 0.0001, BMSCs p0, 2.5 × 10^5^ cells, i.t. versus Muse cells p20, 2.5 × 10^5^ cells, i.t., p < 0.0001; 70 d/85 d, Vehicle versus BMSCs p0, 2.5 × 10^5^ cells, i.t., p = 0.0003, Vehicle versus Muse cells p0, 2.5 × 10^5^ cells, i.t., p < 0.0001, Vehicle versus Muse cells p20, 2.5 × 10^5^ cells, i.t., p < 0.0001, BMSCs p0, 2.5 × 10^5^ cells, i.t. versus Muse cells p20, 2.5 × 10^5^ cells, i.t., p < 0.0001). Statistical significance was indicated by ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, as determined by two-way ANOVA, followed by Bonferroni’s post hoc test. All data are expressed as the mean ± SD. BL, baseline. Muse, multilineage-differentiating stress-enduring; BMSC, bone marrow stromal cell; SNI, spared nerve injury; iTRAQ, isobaric tags for relative and absolute quantitation.

Muse cells have a stronger and more stable analgesic effect than BMSCs

To compare the analgesic effects of BMSCs and Muse cells on neuropathic pain in mice, three doses (1 × 10^4^, 5 × 10^4^ and 2.5 × 10^5^) of primary cultured human BMSCs or human Muse cells were intrathecally injected into the mice 5 days after a spared nerve injury (SNI). As shown in Figure 3C, intrathecal injections of high-dose (2.5 × 10^5^) BMSCs or Muse cells produced a similar, rapid and long-lasting analgesic effect, and there was no significant difference between the two treatments. However, the analgesic effects of injections of low-dose (1 × 10^4^) and middle-dose (5 × 10^4^) Muse cells were stronger and more durable than those of the same dose of BMSCs.

Next, we compared the analgesic effects of cultured human BMSCs and Muse cells at different passages. Passage 0 (P0) and passage 20 (P20) BMSCs and the same passage number of Muse cells (Fig. S1B) were intrathecally injected into SNI model mice on day 5. At the same time, the PBS treatment group was used as the vehicle group. Similar to the above results, both P0 BMSCs and P0 Muse cells effectively reversed SNI-induced mechanical allodynia for a long period of time (Fig. 3C). Interestingly, the intrathecal treatment of P20 Muse cells into SNI mice also produced effective and long-term analgesic effects, whereas P20 BMSCs produced only mild and transient analgesic effects (Fig. 3D). Notably, there was no significant difference in the analgesic effects of P0 Muse cells and P20 Muse cells, which indicates that the analgesic ability of Muse cells is very stable. In summary, these results suggest that Muse cells have stronger and more stable analgesic effects than BMSCs do, especially after long-term culture and passaging.

Intrathecal Muse cells inhibited SNI–induced DRG neuron damage and spinal glial activation

SNI-induced axonal injury elicits dorsal root ganglion (DRG) neuron injury and spinal glial activation, which contribute to the induction and persistence of neuropathic pain (15, 18). First, we investigated whether the intrathecal injection of Muse cells can protect injured DRG neurons in SNI mice. Activating transcription factor 3 (ATF3) is elicited in various tissues in response to stress and can be used as an indicator to assess the degree of damage. Compared with the positive rate of less than 1% in the sham–operated mice, SNI-induced ATF3 expression in L4-L5 DRG neurons 10 days after nerve injury significantly increased to 45% (Fig. 4A), and a single intrathecal injection of Muse cells 5 days after SNI reduced the percentage of ATF3-positive neurons to 19% at the same time point (Fig. 4A).Figure 4**Intrathecal Muse cells inhibited SNI-induced DRG neurons damage and spinal glial activation.**A, Left: Double staining of the DRG neuron damage marker ATF3 (red) and the nucleus marker Hoechst (blue) in ipsilateral L5 DRG 10 days after SNI. Right: Quantification results of ATF3 staining in DRGs. Compared the PBS group, intrathecal injection of Muse cells (2.5 × 10^5^ cells), given 5 days after SNI, reduced the ATF3 expression. The scale bars represent 50 μm. n = 3 mice/group (F (2, 6) = 40.57, p = 0.0003, sham versus PBS, p = 0.0003, sham versus Muse cells, p = 0.0218, PBS versus Muse cells, p = 0.0048). B and C, intrathecal Muse cells (2.5 × 10^5^ cells), given 5 days after SNI, reduced SNI-induced GFAP (B, the astrocyte marker) and IBA-1 (C, the microglial marker) expression in the L4–L5 spinal cord dorsal horn. Bottom panels are enlarged images of the top panels. The scale bar represents 50 μm. Graph in B and C shows the quantification of GFAP and IBA-1 staining. n = 3 mice/group (Figure  4B: F (2, 6) = 71.43, p < 0.0001, sham versus PBS, p < 0.0001, sham versus Muse cells, p = 0.0019, PBS versus Muse cells, p < 0.0001; Figure  4C: F (2, 6) = 187.7, p < 0.0001, sham versus PBS, p < 0.0001, PBS versus Muse cells, p = 0.0002). D, qPCR showing the expression levels of IL-1β, IL-6, and TNF-α mRNAs in the spinal cord dorsal horn of sham and SNI mice with or without Muse cells treatment (IL-1β: F (2, 6) = 25.81, p = 0.0011, sham versus PBS, p = 0.0009, sham versus Muse cells, p = 0.0384, PBS versus Muse cells, p = 0.0189; IL-6: F (2, 6) = 16.45, p = 0.0037, sham versus PBS, p = 0.0037, PBS versus Muse cells, p = 0.0135; TNF-α: F (2, 6) = 60.61, p = 0.0001, sham versus PBS, p < 0.0001, sham versus Muse cells, p = 0.0029, PBS versus Muse cells, p = 0.0047). Statistical significance was indicated by ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, as determined by One way ANOVA, followed by Bonferroni’s post hoc test. All data are expressed as the mean ± SD. Muse, multilineage-differentiating stress-enduring; SNI, spared nerve injury; DRG, dorsal root ganglion; IL, interleukin; ATF3, activating transcription factor 3; GFAP, glial fibrillary acidic protein; IBA-1, ionized calcium binding adapter molecule 1; qPCR, quantitative real-time PCR; TNF-α, tumor necrosis factor-alpha.

Next, we investigated the inhibitory effects of intrathecal Muse cells on spinal glial cell activation (19). Immunoreactivity of both glial fibrillary acidic protein (GFAP) (an astrocyte marker) and ionized calcium binding adapter molecule 1 (IBA-1) (a microglial marker) were markedly increased in the dorsal horn of spinal cord 10 days after SNI (Fig. 4, B and C) and were strongly suppressed at the same time point after Muse cell injection at 5 days after SNI (Fig. 4, B and C). Notably, the morphological changes in spinal astrocytes and microglia induced by SNI were also inhibited by Muse cell treatment (Fig. S3). Since activated spinal glial cells participate in the regulation of neuropathic pain through the production of several inflammatory cytokines (20, 21), such as interleukin (IL)-1β, IL-6 and tumor necrosis factor-alpha (TNF-α), we further examined the mRNA levels of these inflammatory cytokines in the spinal cord dorsal horns of sham surgery and SNI model mice with or without Muse cell treatment. The quantitative real-time PCR (qPCR) results revealed that intrathecal Muse cells inhibited the increase in IL-1β, IL-6 and TNF-α levels induced by SNI in the dorsal horn of spinal cord (Fig. 4D). Collectively, these data indicate that the intrathecal injection of Muse cells reduces SNI-induced DRG neuron damage and spinal glial activation.

Muse cells inhibit neuropathic pain via TGF-β and IL-10 secretion

To explore the molecular mechanism of analgesia in Muse cells, we tested the secretion of anti-inflammatory factors by Muse cells. The cytokines transforming growth factor-beta (TGF-β) and IL-10 have been widely demonstrated to have anti-inflammatory effects and participate in the beneficial effects of BMSC treatment (3, 4). Our previous report revealed that BMSCs release TGF-β but not IL-10 to inhibit neuropathic pain (15). In this study, we compared the expression and secretion of TGF-β and IL-10 in Muse cells and BMSCs. The TGF-β ELISA results showed lower levels of TGF-β expression but similar levels of TGF-β secretion in Muse cells compared with BMSCs (Fig. 5A). Importantly, the IL-10 ELISA results revealed that, compared with that in BMSCs, the expression level of IL-10 in Muse cells was 3-fold greater, and the secretion level was 40-fold greater (Fig. 5A). Next, we examined the release of TGF-β and IL-10 in cerebrospinal fluid (CSF) collected from sham surgery and SNI mice with or without Muse cell treatment. TGF-β and IL-10 release in the CSF did not increase in SNI mice but significantly increased 4 days after they received an intrathecal administration of Muse cells (Fig. 5B).Figure 5**Muse cells release IL-10 and TGF-β to inhibit neuropathic pain in SNI mice.**A, ELISA analysis showing IL-10 and TGF-β release in BMSC and Muse cells culture medium. Compared with BMSCs, a lower expression but similar secretion levels of TGF-β were observed in Muse cells. However, the expression level of IL-10 in Muse cells was 3-fold higher and the secretion level was 40-fold higher than BMSCs. Statistical significance was determined by multiple unpaired t test, n = 3 or 4 separate cultures from different mice (BMSCs versus Muse cells: IL-10: content, p = 0.000932, release, p = 0.005331; TGF-β: content, p = 0.002463). B, ELISA analysis showing increased IL-10 and TGF-β release in CSF 8 days after SNI and 4 days after i.t. delivery of 2.5 × 10^5^ Muse cells. Statistical significance was determined by one-way ANOVA, followed by Bonferroni’s post hoc test. n = 4 mice/group (IL-10: F (2, 9) = 9.359, p = 0.0063, sham versus Muse cells, p = 0.007; TGF-β:F (2, 9) = 19.22, p = 0.0006, sham versus Muse cells, p = 0.001). C, reversal of Muse cells-induced inhibition of mechanical allodynia by IL-10 and TGF-β–neutralizing Abs (4 μg, i.t.), but not by control IgG (4 μg, i.t.). Statistical significance was determined by two-way ANOVA, followed by Bonferroni’s post hoc test. n = 5 or 6 mice/group. (F (21, 136) = 3.223, p < 0.0001; 1 h, Muse cells, i.t.; control IgG, 4 μg versus Muse cells, i.t.; TGF-β Ab, 4 μg, p = 0.0017, Muse cells, i.t.; control IgG, 4 μg versus Muse cells, i.t.; IL-10 Ab, 4 μg, p = 0.0276, Muse cells, i.t.; control IgG,4 μg versus Muse cells, i.t.; TGF-β Ab + IL-10 Ab, p < 0.0001; 3 h, Muse cells, i.t.; control IgG, 4 μg versus Muse cells, i.t.; TGF-β Ab, 4 μg, p < 0.0001, Muse cells, i.t.; control IgG, 4μg versus Muse cells, i.t.; IL-10 Ab,4 μg, p = 0.0007, Muse cells, i.t.; control IgG, 4 μg versus Muse cells, i.t.; TGF-β Ab + IL-10 Ab, p < 0.0001; 5 h, Muse cells, i.t.; control IgG, 4 μg versus Muse cells, i.t.; TGF-β Ab, 4 μg, p < 0.0001, Muse cells, i.t.; control IgG, 4 μg versus Muse cells, i.t.; IL-10 Ab, 4 μg, p = 0.0012, Muse cells, i.t.; control IgG, 4 μg versus Muse cells, i.t.; TGF-β Ab + IL-10 Ab, p < 0.0001; 7 h, Muse cells, i.t.; control IgG, 4 μg versus Muse cells, i.t.; TGF-β Ab, 4 μg, p = 0.01, Muse cells, i.t.; control IgG, 4 μg versus Muse cells, i.t.; IL-10 Ab, 4 μg, p = 0.0459, Muse cells, i.t.; control IgG, 4 μg versus Muse cells, i.t.; TGF-β Ab + IL-10 Ab, p < 0.0001). Statistical significance was indicated by ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. All data are expressed as the mean ± SD. Muse, multilineage-differentiating stress-enduring; BMSC, bone marrow stromal cell; SNI, spared nerve injury; TGF-β, transforming growth factor-beta; IL-10, interleukin 10; CSF, cerebrospinal fluid; IgG, immunoglobulin G.

Behavioral tests further revealed that TGF-β and IL-10 are involved in the analgesic effect of Muse cells. As shown in Figure 5C, the intrathecal injection of TGF-β- and IL-10-neutralizing antibodies partially reversed the analgesic effect of Muse cells, and when these two antibodies were injected together, the effect of reversing the analgesic effect was more prominent.

CCL21/CCR7 axis controls Muse cell migration

Finally, we investigated the distribution of Muse cells after their intrathecal injection into CCI mice. In accordance with our previous methods, the fluorescent dye CM-Dil was used to label the injected Muse cells. Four weeks after injection, we examined the distribution of Muse cells on both sides of the L5 DRG, which is the most severe segment of DRG damage caused by the CCI model. Similar to the previous results of BMSC injection, very few Muse cells were detected in the L5 DRG on the normal control side (data not shown), and many Muse cells were detected on the injured side (Fig. 6A). The number of Dil-labeled Muse cells peaked on day 3, followed by a gradual decrease. Importantly, Dil-labeled Muse cells were still detectable in the DRGs up to day 84 post implantation, supporting their ability to survive in the DRG for an extended period of time (Fig. 6, A and B).Figure 6**Long-term survival of CM-Dil–labeled Muse cells in ipsilateral L5 DRGs following i.t. injection in CCI mice.**A, localization of CM-Dil–labeled Muse cells in ipsilateral L5 DRGs 3 to 84 days after i.t. injection. The scale bars represent 100 μm. Bottom panels are enlarged images of the top panels. Note that Muse cells were mainly localized at DRG borders. B, number of CM-Dil–labeled Muse cells in L5 DRGs 3 to 84 days after i.t. Muse cell injection, given 4 days after CCI. n = 3 mice/group. Muse, multilineage-differentiating stress-enduring; CCI, chronic constriction injury; DRG, dorsal root ganglion

To study the mechanism of Muse cell chemotaxis in CCI mice, we first measured the expression of all known chemokine receptors, including CXCR1-6, CCR1-10, XCR1, and CX3CR1, in Muse cells. Among all the examined receptors, the expression of C–C chemokine receptor type 7 (CCR7) was unexpectedly much greater than that of the other receptors (Fig. 7A). Since the ligands of CCR7 are CCL19 and C–C motif chemokine ligand 21 [CCL21] (22, 23, 24), we next detected the expression of these two ligands in DRG neurons. The qPCR results revealed that in the DRGs of mice 5 days after CCI, the expression of CCL21 in the ipsilateral DRG was increased 1000-fold compared with that in the sham group. In contrast, very low expression of CCL19 mRNA was detected in DRG tissues (Fig. 7B). To further examine the role of the CCL21/CCR7 axis in regulating the migration of Muse cells, a Transwell migration assay was used, and BMSCs were used as a control. Interestingly, the number of migrating Muse cells was three times greater than that of BMSCs under normal culture conditions (Fig. 7, C and D). C–X–C motif chemokine ligand 12 (CXCL12) induced the migration of only BMSCs but not Muse cells, while CCL21 induced both Muse cell and BMSC migration (Fig. 7, C and D). These results indicate that the CCL21/CCR7 axis controls Muse cell migration.Figure 7**CCL21/CCR7 axis controls Muse cells migration.**A, qPCR showing the expression levels of all known chemokine receptors in Muse cells, including CXCR1-6, CCR1-10, XCR1, and CX3CR1. Note, the expression of CCR7 was unexpectedly much higher than other receptors. n = 3 cultures. B, qPCR showing CCL19 and CCL21 mRNA expression in ipsilateral L5 DRGs on days 5 after CCI. The expression of CCL21 in CCI mice was increased more than 1000-fold than the sham group. Statistical significance was determined by Multiple unpaired t test. n = 3 mice/group (con versus CCI 5 days: CCL19, p = 0.000268; CCL21, p = 0.00007). C, chemotaxis (Transwell invasion) assay showing the effects of CXCL12 and CCL21 on the migration of BMSCs and Muse cells. Note, CXCL12 only induced migration of BMSCs but not Muse cells, while CCL21 induced both Muse cells and BMSCs migration. Statistical significance was determined by One way ANOVA, followed by Bonferroni’s post hoc test. n = 5 wells from separate cultures (BMSCs: F (2, 21) = 34.76, p < 0.0001; Vehicle versus CXCL12, p < 0.0001, Vehicle versus CCL21, p = 0.0010; Muse cells: F (2, 12) = 94.64, p < 0.0001; Vehicle versus CCL21, p < 0.0001). Statistical significance was indicated by ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. All data are expressed as the mean ± SD. Muse, multilineage-differentiating stress-enduring; CCI, chronic constriction injury; BMSC, bone marrow stromal cell; DRG, dorsal root ganglion; CCL21, C–C motif chemokine ligand 21; qPCR, quantitative real-time PCR; CCR7, C–C chemokine receptor type 7; CXCL12, C–X–C motif chemokine ligand 12.

Discussion

Neuropathic pain is a serious and growing health problem. Currently, there are no drugs or treatments that can completely and definitively alleviate neuropathic pain (25). Accumulating evidence indicates that the transplantation of BMSCs can relieve neuropathic pain (15). However, BMSCs are heterogeneous and contain many different cell subtypes, making it unclear which cell subsets play an analgesic role (26). Here, we identified Muse cells, a subpopulation of BMSCs with potential value in treating neuropathic pain in various mouse models, including nerve injury, diabetes, and chemotherapy–induced neuropathic pain. We also provided evidence that this analgesic effect is produced by the secretion of TGF-β and IL-10 from Muse cells. In addition, the analgesic effect of Muse cells did not change after 20 passages.

The most valuable aspect of this study is the discovery of a subpopulation of BMSCs that can relieve pain for a long period of time after intrathecal injection. BMSCs are mixed cells containing various subpopulations, and previous studies have rarely focused on a single subpopulation. Muse cells are a newly discovered cell subgroup, and many functions of Muse cells are still unclear (27). In this study, mass spectrometry (MS) analysis revealed that the protein compositions of Muse cells and BMSCs were more than 95% similar, suggesting that Muse cells may have physiological functions similar to those of BMSCs. Previous studies of Muse cells have focused mainly on their stem cell characteristics, in which Muse cells were found to be able to repair damaged tissues by differentiating into specific types of cells (12, 13), and there are few studies describing their anti-inflammatory ability. Here, we compared the anti-inflammatory and analgesic abilities of Muse cells and BMSCs. Intrathecal injection of both types of cells reduced nerve injury-induced neuropathic pain for a long time, and both reduced the inflammatory response and glial cell activation, protecting damaged neurons. In addition, Muse cells had three advantages over BMSCs in terms of analgesic effects. First, Muse cells had stronger analgesic effects than did low-dose BMSCs (5 × 10^4^ cells, i.t.), effectively reducing the number of cells needed for clinical treatment. Second, Muse cells simultaneously secreted TGF-β and IL-10 for analgesia, which has more applications in clinical therapy (28, 29, 30). Third, the analgesic ability of Muse cells remained stable after long-term subculture, whereas the analgesic ability of BMSCs appeared to be significantly weakened. This advantage of Muse cells will provide a key benefit and convenience in the development of commercial cellular drugs with stable analgesic effects.

The main purpose of this study was to facilitate the application of BMSCs in the treatment of neuropathic pain. Currently, because of the safety of cell culture, most clinical treatments for BMSCs are simple separation by flow cytometry after bone marrow extraction, followed by autologous BMSC transplantation (31). Patients need to endure considerable pain to extract a sufficient amount of bone marrow to obtain the number of BMSCs required for transplantation (32). Compared with patients with other serious and dysfunctional diseases, such as nerve injury, liver disease, and arthropathy, patients with chronic pain are less willing to receive autologous BMSC transplantation. BMSCs are a group of cells consisting of various subtypes (33), and no unified criterion has been established to identify BMSCs and no specific marker molecule for BMSCs has been identified to date (34). Because BMSCs obtained under different culture conditions are different, the data from different laboratories are not comparable. In addition, even in the same laboratory, BMSCs may undergo significant changes in cell characteristics (35), such as loss of multilineage differentiation and analgesia, after multiple passages (36). Therefore, these characteristics of BMSCs make it difficult to establish strict quality control standards for commercially available BMSCs. In contrast, Muse cells constitute a homogeneous cell group with consistent characteristics (11). In this study, our results confirmed that single Muse cells rapidly proliferated into pellets in suspension culture and that Muse cells still had stable and strong analgesic effects after multiple passages. Combined with other characteristics of Muse cells, such as their nontumorigenic and strong immunosuppressive properties, Muse cells are good candidates for commercial analgesic cell therapy.

Although mesenchymal stem cells generally exhibit low immunogenicity, they may still elicit immune rejection by the host. In contrast, Muse cells demonstrate unique immunomodulatory properties. Their high expression of human leukocyte antigen-G (HLA-G) and indoleamine 2,3-dioxygenase (IDO) enables them to effectively evade both humoral and cellular immune responses (37). Clinical studies have demonstrated that Muse cells derived from HLA-mismatched donors can avoid immune rejection and persist in recipient tissues for extended periods without requiring concomitant immunosuppressive therapy (38). In a rabbit model of myocardial infarction, 87.5% of transplanted Muse cells expressed HLA-G, compared to only 20% of mesenchymal stem cells, and Muse cell-derived differentiated cells were shown to survive in host tissue for over 6 months (39). Therefore, in the present study, Muse cells exhibited a significant analgesic effect in mouse models of neuropathic pain.

Previous studies have reported that Muse cells transplanted via the systemic route can effectively migrate to damaged tissues and can spontaneously differentiate into cells compatible with homing tissues (40). In this study, we demonstrated a paracrine mechanism by which intrathecal Muse cells target CCL21-producing DRGs to achieve long-term relief of neuropathic pain via the secretion of TGF-β and IL-10. A small intrathecal injection space allows a small amount of cytokines secreted by Muse cells to reach therapeutic concentrations. Some clinical trials have shown that intrathecal BMSCs are safe and do not cause health problems for at least 12 months (41). The intrathecal route not only requires far fewer cells than the intravenous route but also avoids systemic immune responses to transplanted cells because of the blood–spinal/brain barrier. Our previous results revealed that intrathecally injected BMSCs act directly on the spinal cord and DRG and can survive in these tissues for a long time, providing sustained analgesic and neuroprotective effects. In comparison, transplanted BMSCs via the systemic route are mostly confined to the lung and survive for only a few days (42).

Whereas the direct administration of TGF-β and IL-10 provides transient analgesia primarily through anti-inflammatory action, their effects are short-lived and lack the capacity to repair neural pathology or reverse underlying neuropathological changes (19, 28, 43). In contrast, Muse cells actively migrate to injury sites, where they persistently survive and function (39, 44). By continuously secreting a combination of neurotrophic factors (e.g., BDNF and GDNF) and anti-inflammatory cytokines (including TGF-β and IL-10), they exert a multifaceted mechanism encompassing anti-inflammation, neuroprotection, and promotion of regeneration (45). This integrated action enables Muse cells to achieve more sustained and comprehensive analgesic outcomes.

Although this study demonstrates the potent analgesic effects of Muse cells, all experiments were performed in murine models. Further validation in nonhuman primates, such as monkeys, is necessary before these findings can be translated into clinical practice. Overall, Muse cells can be used as a stable, standardized commercial stem cell source for allogeneic transplantation for the treatment of neuropathic pain. The intrathecal injection of Muse cells may provide an efficient, long-term, safe and inexpensive way to treat chronic pain in future clinical trials.

Experimental procedures

Animals and surgery

Adult male Institute of Cancer Research (ICR) mice (25 ± 2 g) were purchased from the Laboratory Animal Center of Nantong University and used for behavioral studies and primary cultures of mouse Muse cells. Adult male Sprague–Dawley (SD) rats (180 ± 20 g) were used for primary cultures of Muse cells. All animal experiments were approved by the Society for Animal Ethics of Nantong University (permission no. S20220303–005). We used several mouse models of neuropathic pain, including a nerve injury model (CCI and SNI), a STZ-induced diabetes model and a chemotherapy-induced neuropathic pain model. To establish the CCI model, the mice were anaesthetized with isoflurane (RWD Life Science, R510–22–10), the left sciatic nerve was exposed, and three ligatures (7–0 prolene) were placed around the nerve proximal to the trifurcation with 1 mm between each ligature. The ligatures were loosely tied until a short flick of the ipsilateral hind limb was observed. The mice in the sham group underwent surgery but not nerve ligation. For SNI surgery, the mice were anaesthetized with isoflurane, and 5.0 silk tight ligation of the tibial and common peroneal nerves was performed, followed by transection and removal of a 3- to 5-mm portion of the nerve. However, the third peripheral branch of the sciatic nerve, the sural nerve, was left intact, and any contact with or stretching of this nerve was carefully avoided. For the diabetes model, the animals were injected intraperitoneally with STZ (Sigma-Aldrich, catalog: S0130) (80 μg/g in 0.1 M citrate-buffered saline, pH 4.2) for three consecutive days. After 2 weeks, we conducted behavioral and blood glucose tests. Mice with blood glucose above 30 mg (16.7 mmol/L) and mechanical pain sensitivity were used for subsequent experiments. For the chemotherapy-induced neuropathic pain model, the animals were injected intraperitoneally with paclitaxel (Selleck, catalog: S1150) (2 mg/kg in saline) four times every other day.

Behavioral analysis

The mice were habituated to the testing environment for at least 2 days before baseline testing. The room temperature and humidity remained stable throughout all the experiments. All the behavioral tests were conducted in a blinded manner. To test mechanical sensitivity, we confined the mice to boxes placed on an elevated metal mesh floor and stimulated their hind paws with a series of von Frey hairs with logarithmically increasing stiffness (0.02–2.56 gf; Stoelting), which were presented perpendicularly to the central plantar surface. We determined the 50% paw withdrawal threshold by Dixon’s up–down method. Mechanical allodynia after SNI was also assessed according to the frequency of response (expressed as the percentage of response) to a low–threshold von Frey hair (0.16 gf, 10 times). Thermal sensitivity was tested using a Hargreaves radiant heat apparatus (IITC Life Science), with the basal paw withdrawal latency adjusted to 9 to 12 s and a cutoff of 20 s to prevent tissue damage.

Drugs and drug administration

TGF-β–neutralizing Ab, AMD3100 (CXCR4 antagonist), human CCL21/6Ckine and CXCL12/SDF–1α were purchased from R&D Systems; LPS from Invitrogen; TNF-α from Novus; IL-10–neutralizing Ab from BioLegend; and normal rabbit IgG from Cell Signaling Technology. For intrathecal injection, a spinal cord puncture was made with a 29G needle between the L5 and L6 levels to deliver reagents (10 μl) or cells (1 or 2.5 × 10^5^ cells in 10 μl of PBS) to the CSF. Before injection, Muse cells were washed 3 times with 0.01 M PBS, centrifuged for 5 min at 1000 rpm, and then resuspended in PBS. In some cases, Muse cells were incubated with Vybrant CM-Dil Cell-Labeling Solution (Molecular Probes, Life Technologies) for 10 min at 37 °C. The cells were then washed three times with PBS and resuspended.

Cell culture

Human Muse cells were generated from human-MSC, which was provided from Genesis Stem Cell Regenerative Medicine Engineering Co., LTD, China. Muse cell culture was performed according to previous reports (11, 46). In brief, Muse cells were obtained from passage 3 or 4 human, mouse or rat BMSCs. Primary cultures of human BMSCs were isolated from bone marrow that was voluntarily donated by healthy adults. Primary cultures of rodent BMSCs were isolated from adult mice and rats. The BMSCs were subjected to long-term trypsin incubation for 8 h (37 °C, 100% humidity, and 5% CO2 in air), followed by vortexing at 2000 rpm for 3 min (BioCote) and centrifugation at 740g for 15 min. To produce M-clusters, individual cells were cultured in MC (MethoCult H4100, StemCell Technologies) culture. For MC culture, culture dishes were purchased from Corning (Ultra Low Attachment Plates) to avoid the attachment of cells to the bottom of the dish. For suspension culture, the culture medium was minimum essential medium Eagle-alpha modification (α-MEM)/20% fetal bovine serum (FBS), and MC was diluted in 20% (vol/vol) FBS in α-MEM to a final concentration of 0.9%. For adherent culture, the culture medium was α-MEM/20% FBS.

Transwell migration assay

The migratory ability of the BMSCs and Muse cells was determined using Transwell plates (6.5 mm in diameter with 8 μm pore filters; Corning Costar). In brief, 5 × 10^4^ cells in 100 μl of serum-free medium were added to the upper well, and 600 μl of CXCL12-or CCL21-containing medium and mitomycin C-containing medium were added to the lower well of a Transwell plate. Following incubation for 10 h (37 °C, 100% humidity, 5% CO_2_ in air), the number of cells that had migrated to the lower side of the filter was counted under a light microscope (data are presented as the average number of migratory BMSCs and Muse cells in five randomly selected fields). Each experiment was performed in triplicate, and the data were averaged for statistical analysis.

Protein extraction and protein mass spectrometry

Protein extraction, digestion, and labeling with iTRAQ reagents

Primary human BMSCs and human Muse cells were washed thoroughly with ice-cold PBS and lysed in buffer containing 50 mM Tris–HCl (pH 7.6), 5 mM EDTA, 50 mM NaCl, 30 mM sodium pyrophosphate, 50 mM NaF, 0.1 mM Na3VO4, 1% (v/v) Triton X-100, 1 mM PMSF, and a protease inhibitor mixture (Roche Applied Science). The lysates were then clarified by centrifugation at 13, 300g for 30 min at 4 °C; the supernatant was collected, and a BCA assay (Beyotime Biotechnology) was used to determine the protein concentration.

Dithiothreitol (DTT) was added to the protein solution to a final concentration of 10 mM, followed by incubation at 55 °C for 30 min. The sample was then cooled to room temperature, and iodoacetamide was added to a final concentration of 20 mM. The mixture was incubated in the dark at room temperature for 30 min. Subsequently, the solution was diluted with triethylammonium bicarbonate (TEAB) buffer and digested with trypsin at an enzyme-to-protein ratio of 1:50 (w/w). Digestion was carried out overnight at 37 °C with shaking and terminated by adding 1% formic acid.

The resulting peptides were labeled with iTRAQ Reagents (Applied Biosystems) as follows: hBMSCs, iTRAQ Reagent 113&115; hMuse cells, iTRAQ Reagents 117&119. The labeled digests were mixed and dried.

2D-Liquid chromatography with tandem mass spectrometry

The mixed peptide sample was fractionated by strong cation exchange chromatography on a 20AD HPLC system (Shimadzu) using a PolySULFOETHYL A column (200 × 4.6 mm, 5 μm, 200 Å; PolyLC). The mixed peptides were diluted with the loading buffer (5 mM KH2PO4 in 25% acetonitrile, pH 2.8), and loaded onto the column. Buffer A composition was identical to the loading buffer, while Buffer B was composed of Buffer A with the addition of 350 mM KCl. Separation was performed using a linear binary gradient of 0 to 100% buffer B for 105 min. The absorbance at 214 nm and 280 nm was monitored.

Each strong cation exchange fraction was first desalted on a HyperSep C18 column (Thermo Fisher Scientific). Subsequently, the sample was dried and finally reconstituted in buffer C (5% acetonitrile and 0.1% formic acid). Briefly, peptides were resuspended in 0.1% formic acid and analyzed on an Orbitrap Fusion Lumos mass spectrometer coupled to an EASY-nLC 1200 system. Peptides were loaded on a trap column (100 μm × 2 cm, 5 μm) and separated on an analytical column (50 μm × 15 cm, 2 μm) using a 120-min gradient (7–40% acetonitrile/0.1% formic acid). MS parameters: MS1 resolution 60,000; MS2 resolution 30,000; scan range m/z 350 to 1200; high energy collision dissociation fragmentation at 30% normalized collision energy. iTRAQ labeling followed by 2D-LC-MS/MS analysis was repeated twice (different protein lysates) to diminish the effect of experimental variation in the results of a proteomics analysis.

Data analysis

Raw MS files were searched against a Homo sapiens protein database (Uniprot https://www.uniprot.org/; 20405 entries) using MaxQuant (2.4.7.0). The following parameters were used for data processing: enzyme (trypsin; max missed cleavages:2); mass tolerance was set to 5 ppm for precursor ion, and 0.02 Da for the fragment ion; The amino methyl group on cysteine represents a fixed modification, whereas Kcr modification and methionine oxidation are classified as variable modifications. The Percolator algorithm (47) was employed to ensure that the peptide-level false discovery rate remained below 1% and that the protein identification q-value was less than 0.01. In addition, the posterior error probability was controlled to remain below 0.25 (i.e., confidence level >75%). All raw mass spectrometry data files have been deposited to the ProteomeXchange Consortium via the iProX repository with the data set identifier PXD070082.

ELISA

Human TGF-β1 and IL-10 ELISA kits were purchased from R&D Systems. ELISAs were performed using BMSC and Muse cell culture media or cell lysates. Cultured cells were homogenized in lysis buffer containing protease and phosphatase inhibitors. The ELISA was performed according to the manufacturer’s instructions.

Trafficking of Muse cells to DRGs

To examine the distribution of transplanted CM-Dil–labeled Muse cells following i.t. injection, lumbar spinal cord segments and L4–L6 DRGs were collected. For quantitative analysis of engrafted cells in DRGs, 10 sections (12 μm) from each DRG were examined for the labeled Muse cells.

Immunohistochemical staining

The animals were deeply anaesthetized with isoflurane and perfused through the ascending aorta with PBS, followed by 4% paraformaldehyde. After perfusion, the spinal cord segments and DRGs were removed and postfixed in the same fixative overnight. Spinal cord and DRG sections were cut with a cryostat and processed for immunofluorescence. The sections were incubated overnight at 4 °C with the following primary Abs: mouse GFAP (1:1000; Millipore Cat# MAB360) and rabbit IBA-1 (1:1000; Wako Chemicals USA Inc. Cat# 019-19741). In some cases, Hoechst (Sigma-Aldrich) was used to stain the cell nuclei. The stained sections were examined under a Nikon fluorescence microscope or a Zeiss confocal microscope. The intensity of fluorescence was analyzed using Photoshop CS6 (PS).

Quantitative real-time PCR

Total RNA was isolated from spinal dorsal horn tissues and DRGs of the L4–L5 segments using the TRIzol RNA isolation system (Sigma-Aldrich). Complementary DNA was synthesized from total RNA using a PrimeScript RT Master Mix kit (TAKARA) following the supplier’s instructions. We performed gene-specific mRNA analyses using the SYBR Premix Ex TaqTM Real-Time PCR system (TAKARA). The primer sequences are described in Table S1. The primer efficiency was obtained from the standard curve and integrated for the calculation of relative gene expression, which was based on the real–time PCR threshold values of different transcripts and groups.

Statistical analysis

All mice were randomly assigned to different groups. The data were analyzed by two independent researchers in a double-blind manner. All data from mice were included except for the screening of mice with successful STZ model construction. For immunohistochemistry or behavioral analysis, we conducted two or three independent replications. The experimental data were analyzed using GraphPad Prism 8.0 software. The measured data are expressed as the mean ± SD. Student’s t test (two groups) or ANOVA (one-way or two-way) were used to compare the differences between groups, followed by the Bonferroni correction. A value of p < 0.05 was considered statistically significant.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. All data generated or analyzed during this study are included in this published article and its supplementary information files.

Supporting information

This article contains supporting information.

Conflict of interest

The authors declare that they have no conflicts of interest with the contents of this article.