Dexmedetomidine Induces Microglial Activation and Modulates Microglial M1/M2 Polarization in Attenuation of Chronic Morphine Tolerance in Cancer Pain

The pro-inammatory (M1) and anti-inammatory (M2) status of microglial determines the outcome of neuroinammation, which contributes to the pathogenesis of chronic morphine tolerance. Studies report that α2-adrenoceptor agonist dexmedetomidine exerts anti-inammatory effects in inhibiting morphine tolerance in normal and neuropathic pain animals, which has not been studied in cancer pain. Therefore, we investigate the effect of intrathecal DEX on morphine tolerance in cancer pain, and whether dexmedetomidine functions via modulating microglial activation and M1/M2 polarization. 54 Wistar rats with intrathecal catheterization were treated by morphine for 10 days. Test groups received intrathecal α2-adrenoceptor agonist dexmedetomidine or antagonist MK-467. The mRNA levels of TLR4 and NF-κB were tested by RT-PCR. The protein levels of TLR4, NF-κB, Iba-1, iNOS, CD206 were quantifed using Western blotting, and IL-10 and TNF-α were examined by ELISA. Dexmedetomidine attenuates mechanical threshold and thermal latency, and increased the expression of TLR4 and NF-κB in morphine tolerance of cancer pain. Dexmedetomidine attenuates mechanical and thermal nociception in morphine tolerance in cancer pain rats. Intrathecal DEX pre-treatment signicantly increased the protein levels of microglia maker Iba-1, M2 marker CD206 and anti-inammatory factor IL-10, while had no evident inuence on the pro-inammatory factor TNF-α and M1 marker iNOS in morphine tolerance. Our ndings suggest that intrathecal dexmedetomidine attenuates morphine tolerance in cancer pain via α2-adrenoceptor pathway. Furthermore, dexmedetomidine upregulates TLR4/NF-κB pathway and induces microglia activation in chronic morphine tolerance of cancer pain. The anti-inammatory effect of dexmedetomidine might be exerted by inducing microglia M2 polarization and increasing anti-inammatory The data are expressed as the mean fold-change in mRNA expression relative to the mean±SD (n=6). Vehicle+saline group served as control. The mRNA expression levels were analyzed via a one-way analysis of variance ANOVA (P=0.001) and then followed by multiple comparisons using Student t test with Bonferroni correction (P<0.05/7=0.007143 was dened as statistical signicance). ※ P<0.0071 compared with vehicle+saline group, #P<0.007 compared with vehicle+morphine group.


Introduction
Cancer pain is one of the most serious conditions that severely impairs patients' quality of life [1].
Morphine is a prototypical opiate analgesic commonly used in the treatment for cancer pain. However, with long-term use, the development of analgesic tolerance becomes a major problem [2]. In addition to neuronal mechanisms, microglial activiation has emerged as a potentially signi cant new mechanism of morphine tolerance. Recent studies demonstrate that microglia-mediated neuroin ammation plays a double-edged role in the pathological processes of chronic morphine treatment [3]. There are basically two polarized states of activated microglia, the classical deleterious 'M1' phenotype and the alternative neuroprotective 'M2' phenotype [4]. Imbalanced microglial polarization, in the form of excessive activation of M1 microglia and dysfunction of M2 microglia, promotes the development of neuroin ammatory response in sustained morphine exposure. Thus, regulating the microglia polarization provides us a new target in the treatment for morphine tolerance.
A number of signaling molecules, including Toll-like receptor 4 (TLR4) and Nuclear Factor-kappa B (NF-κB), are critical to the modulation of microglial activation and neuroin ammation. TLR4 expresses in a wide variety of cells and recognizes the invariant molecular structures of pathogens and participates in the innate immunity [5]. The central immune signaling initiated by TLR4, resulting in the activation of downstream mediators, including the transcription factor nuclear factor (NF)-κB, which increases the production of pro-in ammatory cytokines, such as tumor necrosis factor (TNF)-α, interleukin (IL)-1β, and anti-in ammatory cytokines such as IL- 10 [6]. Accumulating evidence indicates that TLR4-mediated NF-κB activation in the spinal cord is implicated in opioid tolerance, hyperalgesia, and physical dependence [7][8].
Microglial cells express mRNA encoding α2-adrenoceptor (α2-AR) linked to the signal transduction cascade TLR4/NF-κB pathway. In microglia, TLR4 pathway mediates the activation of iNOS induced by LPS. It has been found that the in ammatory response of microglia could be weakened by inhibiting TLR4 pathway [9]. Dexmedetomidine (DEX) is a highly selective agonist of α2-adrenoceptor (α2-AR) with sedative properties and analgesic potency. Numerous studies have shown that DEX exerts an antiin ammatory effect in many conditions of neural disorders [10] [11]. Furthermore, in vitro, a recent study demonstrates that DEX alleviates neuroin ammation through in uencing microglial M1/M2 polarization [12]. It has been reported that administration of α2-AR agonists attenuated the progress of morphine analgesic tolerance in normal rats and neuropathic pain rats, possibly by preventing the morphineinduced neuroin ammatory activities [13] [14]. Moreover, the role of α2-AR DEX in neuroin ammatory response of non-cancer pain have been investigated involving down-regulating TLR4/NF-κB signaling pathway [15][16]. However, the mechanism underlying the effect of DEX on chronic morphine tolerance in cancer pain has not been studied.
In this study, we aim to investigate the effect of intrathecal DEX injection on the development of morphine tolerance in cancer pain treatment, and whether this DEX-induced effect is exerted by modulating microglia polarized activities and propose that TLR4/NF-κB signaling pathway might be involved in this process.

Animals
Adult male Wistar rats (6-7 weeks, 180-200g, n=60) obtained from Harbin Medical University Laboratories (Harbin, China) were housed at a constant ambient temperature of 22±1℃ under a 12-h light-dark cycle with ad libitum access to water and food. All animal experiments complied with the policies and recommendations of the International Association for the Study of Pain, and the National Institutes of Health (NIH) guidelines for the handling and use of laboratory animals. The experimental protocol was approved by the Animal Care and Use Committee of Harbin Medical University (Harbin, China).

Intrathecal catheterization
Intrathecal (i.t.) catheter implantation was performed following the procedures described by previous studies with modi cations [17]. After inhalation anesthetized with sevo urane (2-3% in oxygen, Abbott Pharmaceutical Co. Ltd., Lake Bluff, IL, USA), the animal was placed in a prone position on the experimental table, and the bilateral spurs were found. A midline skin incision was made in the lumbar region (L3-L5), and the intervertebral membrane between L3 and L4 was exposed. A 25G needle was then used to puncture the membrane, and the catheter (Polyethylene PE 10, Becton Dickinson, Franklin Lake, NJ) was inserted into the subarachnoid space about 2.0 cm. The catheter was xed to the muscle, and the other end was tunneled rostrally underneath the skin to exit 1.5 cm in the occipital region. The incisions were then closed in layers using 4-0 silk. China) on the plantar region of right hindpaw. The Walker 256 cell culture procedure was conducted as described previously [19]. On day 5 post-inoculation, a signi cant proliferation of tumor cells was detected with a development of hyperalgesia and allodynia of the hind paw, as indicated by comparing the TWL and MWT to the values before inoculation in all rats. This indicates a successful establishment of cancer pain model.
A total of 54 tumor-bearing rats were randomly allocated to 6 groups (n=9 for each group). All drugs were administered twice daily (9:00 am and 5:00 pm) for 10 days (day 1 to day 10). To induce morphine analgesic tolerance, a 10-day cumulative dosing regimen was used. The rats in morphine group were injected with intrathecal (i.t.) normal saline 10µl and subcutaneous (s.c.) morphine sulfate (Northeast Pharmaceutical Group Co., Ltd., Harbin, China) 10mg/kg/mL (vehicle+morphine group). The analgesic effects of morphine were measured by the tail-ick test. Drugs were delivered as follows: in separate group, rats received i.t. injection of dexmedetomidine (10µg/kg, 20µl, Heng Rui Pharmaceutical Co., Ltd. Jiangsu, China) and s.c. injection of morphine (DEX+morphine group), or i.t. injection of α 2 -AR antagonist MK-467 (0.25mg/kg, 10µl, USA) 30 min prior to DEX and s.c. injection of morphine (MK-467+DEX+morphine group). DEX was injected 30 min prior to morphine. The rats received i.t. 10µl and s.c. 1ml/kg injection of normal saline served as control group (vehicle+saline group). Furthermore, to test the effects of DEX on cancer pain, the rats received i.t. injections of DEX (DEX+saline group) or MK-467

Behavioral Tests
Rats were acclimated within plexiglass enclosures on a clear glass plate maintained at 26±0.5˚C.
Behavioral assays were performed before drug administration (at baseline) on day 0 and 30 min after morning drug delivery from day 1 to day 10. Mechanical withdrawal threshold (MWT) was measured by applying von Frey laments (Stoelting Company, Wood Dale, IL, USA) and used a staircase method. The MWT was de ned as the lowest lament in grams when the lament aroused a positive reaction (licking or lifting of hindpaw), and a 60 g maximum was imposed as a cutoff to avoid tissue damage. Three measurements were performed on each rat randomly beginning with the left or right paw, and the mean value of MWT from both hind paws was used. Thermal withdrawal latency (TWL) was determined for each rat using a hot plate apparatus in a plastic cylinder (Technology & Market CORP, Chengdu, China). Rats were individually placed on the hotplate (52°C), and latency was de ned is the interval between exposure to the hotplate and positive reaction (licking or lifting of the unilateral hindpaw) due to heat stimulation. The maximal cut-off time was 30 s in order to avoid damage to the paw. The average latency was calculated from 3 measurements taken at 5 min intervals.

Enzyme-linked immunosorbent assay
After the rats were sacri ced under deep anesthesia with pentobarbital sodium (1% in NS, 50 mg/kg, i.p.), the lumbar spinal cord segments were collected. The supernatants of tissue homogenates were collected and analyzed using enzyme-linked immunosorbent assay (ELISA) kits (Senbeijia, Nanjing, China) for IL-10 and TNF-α according to the manufacturer's instructions.

Immunohistochemistry
The rats were deeply anesthetized and perfused through the ascending aorta with normal saline, followed by 4% paraformaldehyde in 0.1 mol/L phosphate buffer. Subsequently, the lumbar (L3-5) SC was removed and post-xed in the same xative at 4℃ for 24h, and then the xative was replaced by 30% sucrose in PBS over two nights. Transverse spinal sections (20 µm) were cut on a cryostat and prepared for immuno uorescence staining. Sections, after a randomly selection and PBS wash, were blocked with 5% goat serum in 0.3% Triton X-100 for 1h at 37°C, and incubated with primary antibodies for TLR4 (ABclonal, 1:100) and NF-κB (Immunoway, 1:100) overnight at 4°C. After washing with PBS for 3 times, the sections were incubated in goat anti-rabbit IgG (1:100) for NF-κB p65 or goat anti-mouse IgG (1:100) for TLR4 for 1h at 37˚C. The morphological details were examined under an inverted microscope (Olympus Corporation, Tokyo, Japan). The positive area of the images was digitized and subjected to color threshold analysis using NIH ImageJ software version 2.1 (National Institutes of Health, Bethesda, MD, USA).

Statistical analyses
Data were analyzed by the Statistical Product for Social Sciences (SPSS version 23.0). Data were presented as the mean±SD. For behavioral data, the difference between groups was determined by Twoway Repeated Measures Anova followed by Student t test with Bonferroni correction. TLR4, NF-κB, Iba-1, iNOS, CD206, TNF-α and IL-10 expression were analyzed using one-way ANOVA followed by Student t test with Bonferroni correction for multiple comparisons. For the data from behavioral tests, RT-PCR, western blotting and ELISA, we performed 7 comparisons: vehicle+saline versus the other 5 groups, DEX+morphine versus vehicle+morphine, and MK-467+DEX+morphine versus vehicle+morphine. A P value less than the Bonferroni-corrected threshold of 0.0071 (0.05/7) was de ned as statistically signi cant. For the data from immunohistochemistry, we performed 5 comparisons: vehicle+saline versus DEX+morphine, vehicle+morphine, and MK-467+DEX+morphine group. A P value less than the Bonferronicorrected threshold of 0.01 (0.05/5) was de ned as statistically signi cant.

Establishment of Morphine Tolerance in Cancer Pain
Consistent with the results of our previous research [18], 5 days after injection of Walker 256 tumor cells, a signi cant development of mechanical and thermal sensitivity indicated painful behaviors caused by cancer was observed in rats. Based on the 7-day morphine treatment in our previous research [18], chronic morphine analgesic tolerance in cancer pain was established using a 10-day morphine administration. As shown in

Discussion
In this study, we assessed the effect of DEX on chronic morphine tolerance in a rat model of cancer pain. Intrathecal injection of DEX enhanced morphine analgesia and attenuated the development of chronic morphine tolerance. DEX promoted microglial activation and upregulated the expression of TLR4/NF-κB signaling pathway. Moreover, DEX exhibited anti-in ammatory effect in attenuation of chronic morphine tolerance by modulating microglial activity towards the M2 phenotype, characterized by higher expression of anti-in ammatory factor IL-10 and M2 maker CD206. All these effect of DEX could be blocked by a selective α2-AR antagonist MK-467.
Morphine-induced microglia-mediated neuroin ammation is a complex molecular system involving various signaling pathways. Accumulating evidence demonstrates that TLR4/NF-κB pathway plays a critical role in morphine-induced microglia activation [7], and inhibition of TLR4/NF-κB has been proven to be effective on attenuating morphine tolerance [16]. However, to date, almost all these researches on TLR4/NF-κB pathway in morphine tolerance have been performed based on normal animals or noncancer pain models. Few studies have investigated the role of TLR4/NF-κB signaling in morphine tolerance on cancer pain model. Surprisingly, inconsistent with most of previous studies based on noncancer pain models, our study revealed that intrathecal injection of DEX promoted microglial activation and increased the expression of TLR4 and NF-κB in DRG of morphine tolerant rats with cancer pain, and this effect could be reversed by α 2 -AR antagonist MK-467. These ndings showed that DEX-induced attenuation in morphine tolerance of cancer pain did not exert via suppressing TLR4/NF-κB pathway.
Numerous studies indicate that TLR4 promotes in ammatory responses that are correlated with the pathology of pain [20] [21]. It has been reported that DEX signi cantly decreased upregulation of TLR4 expression in spinal cord of rats with monoarthritis [22]. Similarly, in our study, DEX administration (DEX+saline group) decreased the mRNA and protein expression of TLR4 in DRG compared with nontreated rats (vehicle+saline group), suggesting that DEX could suppress the in ammatory responses via down-regulating TLR4 expression in cancer pain. However, in the state of morphine tolerance, our results showed that intrathecal administration of DEX signi cantly increased TLR4 and NF-κB expression in DRG of the rats with cancer pain on both mRNA and protein level. This nding suggested that DEX increased the morphine-induced microglial activation characterized in upregulating TLR4/NF-κB pathway. Previous studies have shown TLR4 antagonist LPS-RS alleviates chronic neuropathic pain caused by nerve chronic constriction injury at the rst administration, while fails at a subsequent administration [23]. In the post-in ammatory stage, LPS-RS also fails to inhibit the TLR4 activation that occurred after the stimulus ceased [24]. These results indicate that TLR4 may not play a major effect during the whole chronic neuroin ammatory response. On the other hand, a previous study has reported that in a morphine tolerance model, microglial activation is caused by a TLR4-independent mechanism [25]. Moreover, Mattioli and colleagues [26] demonstrated that TLR4 was not required for the development of morphineinduced analgesic tolerance, hyperalgesia, or physical dependence. Herein, we speculate that the inhibiting effect of DEX on morphine tolerance in cancer pain may not regulated via TLR4/NF-κB pathway, whereas involve other mechanism of central neuro-in ammatory responses.
Long-term application of morphine has been found to enhance neuroimmune reactivity, involving microglial activation and proin ammatory cytokine production, and then resulted in analgesic tolerance. Activated microglial cells exist in M1 phenotype and M2 phenotype. M1 microglia aggravates tissue damage by producing destructive pro-in ammatory factors such as IL-1β, TNF-α, while M2 microglia promotes injury repair by releasing anti-in ammatory factors such as IL-4, IL-10 and some neurotrophic factors [27]. Furthermore, the ratio between microglial M1 and M2 phenotypes affects the development trend of neuroin ammatory response, which could be intensi ed by increase of the M1/M2 ratio [28]. Recently, a study assessed the effects of DEX on microglia polarization [12], revealing that treatment with DEX increased the expression of anti-in ammatory factor IL-10 and M2 phenotypic marker CD206, while decreased the levels of pro-in ammatory factors TNF-α and M1 phenotypic marker iNOS. IL-10 plays a pivotal role in morphine tolerance and regulates the production of proin ammatory cytokines, including IL-1β, IL-6, TNF-α [29] [30]. Consistently, in this study of cancer pain, we found that intrathecal DEX increased the expression of anti-in ammatory factor IL-10 and M2 phenotypic marker CD206, while had no signi cant in uence in the levels of pro-in ammatory factors TNF-α and M1 phenotypic marker iNOS, indicating that DEX-modulated microglial activation may due to microglia polarisation to the M2 phenotype and promotion of the anti-in ammatory cytokines IL-10, consequently inhibiting morphineinduced chronic neuro-in ammatory responses of analgesic tolerance. M1/2 microglia polarization has not been studied previously in the context of chronic morphine administration. In cancer pain management, prolong morphine exposure stimulated the microglial activities and induces analgesic tolerance. In our investigation, the increased expression of M2-microglia in the spinal cord could be interpreted as an effort to resolve the neuroin ammation caused by chronic morphine administration.
The mechanisms of microglia polarization involve various molecules and signaling pathways. Parthenolide, known to affect intracellular microglial in ammation pathways, like p-38 and ERK1/2 has promoted spinal M2 microglia/macrophage polarization and relieved pain in a rat model of neuropathy [31]. A recent study demonstrated that dexmedetomidine induces microglial polarization towards M2 phenotype by inhibiting phosphorylation of ERK1/2 in vivo [32]. Dexmedetomidine decreased the expression of TNF-α and inhibited activation of ERK1/2 in the lung tissues of cecal ligation and puncture (CLP)induced septic rats [33]. Dexmedetomidine also inhibited the expression of purinergic receptor 7 (P2X7R) and ERK phosphorylation that attenuated chronic neuropathic pain [34]. Considering the role of activated microglia in both neuropathic pain and opioid-induced neuroin ammation, drugs that promote microglia M2-conversion could be bene cial in attenuating morphine tolerance. In our study, further researches need to be carried out to explore the molecular mechanisms and signaling pathways underlying DEX-induced M2 polarization and anti-in ammatory cytokine releasing.
There are some limitations to this study. Firstly, we only measured the expression of TLR4/NF-κB signaling pathway and the activation of central immune signals at a single time point (day 10). It could not fully re ect the in uence of DEX on the central immune system during the whole process of morphine tolerance development in cancer pain. Secondly, our research for the rst time showed that in the development of morphine treatment in cancer pain, DEX inhibited morphine tolerance by promoting microglial activation and M2 polarization, and meanwhile DEX exerted the anti-in ammatory effects by releasing anti-in ammatory factor IL-10. However, we only detected the expression levels of M1 and M2 marker proteins in microglia from a molecular perspective. A cytological method would be carried out to better detect the changes in the ratio of microglia M1 and M2 phenotypes. Third, we did not investigate the central immune signaling pathways and molecular mechanisms involved in the DEX-induced attenuation of morphine tolerance in cancer pain. Further studies would be required to address these limitations.
In conclusion, the present study demonstrated that intrathecal injection of DEX augmented the analgesic effect of morphine and attenuated the development of morphine tolerance in cancer pain via the α2-AR pathway. The anti-in ammatory effect of DEX in morphine tolerance of cancer pain involved in modulating microglia M2 phenotype and promoting releases of IL-10. Moreover, co-administration of DEX and morphine induced microglial activation and increased the expression of TLR4/NF-κB pathway in morphine tolerant rats with cancer pain. It provides an new insight into the fact that the inhibitory effect of DEX on morphine tolerance in cancer pain may be regulated not through suppressing TLR4/NF-κB pathway but by modulating microglia M2 polarization.

Declarations
Ethics approval and consent to participate The experimental protocol was approved by the Animal Care and Use Committee of Harbin Medical University (Harbin, China).

Consent for publication
Not applicable Availability of data and materials Data sharing not applicable to this article as no datasets were generated or analysed during the current study.    Effects of i.t. injection of DEX on Ibal-1 protein expression in ipsilateral spinal cord of morphine tolerant rats with cancer pain. (A) protein expression, (B) analysis of gray value of protein bands. Western blotting was performed using the ipsilateral lumbar L3/L4/L5 segments of DRG. The data are expressed as the mean fold-change protein expression relative to the mean±SD (n=3). Vehicle+saline group served as control. The protein expression levels were analyzed via a one-way analysis of variance ANOVA and then followed by multiple comparisons using Student t test with Bonferroni correction (P<0.05/7=0.007143 was de ned as statistical signi cance). #P<0.007 compared with vehicle+morphine group. i.t., intrathecal; DEX, dexmedetomidine; DRG, dorsal root ganglia.  Effects of i.t. injection of DEX on TLR4 (Fig. 4-1) and NF-κB ( Fig. 4-2) protein expression in ipsilateral DRG of morphine tolerant rats with cancer pain. (A) protein expression, (B) analysis of gray value of protein bands. Western blotting was performed using the ipsilateral lumbar L3/L4/L5 segments of DRG. The data are expressed as the mean fold-change protein expression relative to the mean±SD (n=6).
Vehicle+saline group served as control. The protein expression levels were analyzed via a one-way analysis of variance ANOVA and then followed by multiple comparisons using Student t test with Bonferroni correction (P<0.05/7=0.007143 was de ned as statistical signi cance). #P<0.007 compared with vehicle+morphine group. i.t., intrathecal; DEX, dexmedetomidine; DRG, dorsal root ganglia.  Effects of i.t. injection of DEX on CD206 (Fig. 6-1) and iNOS (Fig. 6-2) protein expression in ipsilateral SC of morphine tolerant rats with cancer pain. (A) protein expression, (B) analysis of gray value of protein bands. Western blotting was performed using the ipsilateral lumbar L3/L4/L5 segments of spinal cord.
The data are expressed as the mean fold-change protein expression relative to the mean±SD (n=3). Vehicle+saline group served as control. The protein expression levels were analyzed via a one-way analysis of variance ANOVA and then followed by multiple comparisons using Student t test with Bonferroni correction (P<0.05/7=0.007143 was de ned as statistical signi cance). #P<0.007 compared with vehicle+morphine group. i.t., intrathecal; DEX, dexmedetomidine; SC, spinal cord.

Figure 8
Effects of i.t. injection of DEX on IL-10 ( Fig. 7-1) and TNF-α ( Fig. 7-2) protein expression in ipsilateral SC of morphine tolerant rats with cancer pain. ELISA was performed using the ipsilateral lumbar L3/L4/L5 segments of spinal cord. The data are expressed as the mean fold-change protein expression relative to the mean±SD (n=3). Vehicle+saline group served as control. The protein expression levels were analyzed via a one-way analysis of variance ANOVA and then followed by multiple comparisons using Student t test with Bonferroni correction (P<0.05/7=0.007143 was de ned as statistical signi cance). #P<0.007 compared with vehicle+morphine group. i.t., intrathecal; DEX, dexmedetomidine; SC, spinal cord.