Magnesium and Morphine in the Treatment of Chronic Neuropathic Pain–A Biomedical Mechanism of Action

The effectiveness of opioids in the treatment of neuropathic pain is limited. It was demonstrated that magnesium ions (Mg2+), physiological antagonists of N-methyl-D-aspartate receptor (NMDAR), increase opioid analgesia in chronic pain. Our study aimed to determine the molecular mechanism of this action. Early data indicate the cross-regulation of µ opioid receptor (MOR) and NMDAR in pain control. Morphine acting on MOR stimulates protein kinase C (PKC), while induction of NMDAR recruits protein kinase A (PKA), leading to a disruption of the MOR-NMDAR complex and promoting functional changes in receptors. The mechanical Randall-Selitto test was used to assess the effect of chronic Mg2+ and morphine cotreatment on streptozotocin-induced hyperalgesia in Wistar rats. The level of phosphorylated NMDAR NR1 subunit (pNR1) and phosphorylated MOR (pMOR) in the periaqueductal gray matter was determined with the Western blot method. The activity of PKA and PKC was examined by standard enzyme immunoassays. The experiments showed a reduction in hyperalgesia after coadministration of morphine (5 mg/kg intraperitoneally) and Mg2+ (40 mg/kg intraperitoneally). Mg2+ administered alone significantly decreased the level of pNR1, pMOR, and activity of both tested kinases. The results suggest that blocking NMDAR signaling by Mg2+ restores the MOR-NMDAR complex and thus enables morphine analgesia in neuropathic rats.


Introduction
Despite the fact that the last few decades have brought progress in the treatment of many diseases, the relief of neuropathic pain is still one of the major challenges in medicine. The barrier in the alleviation of this type of pain is the limited efficacy of classic analgesics, including opioids [1][2][3]. A reduction in opioid analgesia appears to be associated with the increased activation of pronociceptive N-methyl-D-aspartate receptors (NMDARs) during the development of neuropathic pain [4,5].
One of the solutions is the administration of opioids in high doses [6], which unfortunately increases the risk of side effects, such as respiratory depression, opioid hyperalgesia, constipation, rapid development of tolerance, and spreading in recent years addiction risk [7][8][9][10].
Data available in the literature indicate that synthetic NMDAR antagonists not only alleviate neuropathic pain [11][12][13][14][15] but also enhance the analgesic effect of opioids [16][17][18]. It is worth noting that magnesium ions (Mg 2+ ) are the physiological antagonists of the NMDAR physiological antagonists, on MOR activity. The level of activation and inactivation of MOR and NMDAR after the tested compounds administration was investigated by analyzing phosphorylation of MOR Ser375 and NMDAR NR1 subunit Ser896 using the Western blot method. The activity of PKA and PKC involved in the bidirectional interaction between tested receptors was determined using standard enzyme immunoassays.

The Influence of Mg 2+ on the Analgesic Effect of Morphine in Streptozotocin-Induced Hyperalgesia after Mechanical Stimulation
Streptozotocin (STZ)-treated rats developed mechanical hyperalgesia within 2 weeks after single intramuscular administration of STZ. In turn, we recorded no changes in the nociceptive threshold in the control (healthy animals) group (data not shown). Mg 2+ and MRF were administered daily for seven days from day 18 to day 24. Mg 2+ applied alone to neuropathic animals caused an increase in the nociceptive threshold observed from day 23 until day 26 of the experiment (Figure 1). On the 18th day of the experiment (the 1st day of application), Mg 2+ alone did not modify STZ-induced hyperalgesia ( Figure 2A). However, on day 24 of the experiment (the 7th day of application), Mg 2+ increased the nociceptive threshold over a period of 120 min after administration ( Figure 2B). In turn, MRF administered alone on seven consecutive days did not change STZ-induced hyperalgesia ( Figure 1). Similarly, no significant changes in nociceptive thresholds were observed on the 1st day of the drug's application (day 18 of the experiment) over a period of 120 min after MRF administration ( Figure 2A). However, on the 7th day (day 24 of the experiment), MRF alone induced a slight statistically significant analgesic effect 15 min after injection (p < 0.05) ( Figure 2B). As it is shown in Figure 1, the mechanical hyperalgesia was significantly reduced in rats receiving the combination of Mg 2+ and MRF starting from day 21 of the experiment (the third day of drugs application) compared to MRF-treated animals. After cessation of tested drugs (Mg 2+ and MRF) application, animals' pain sensitivity was reported to increase. On day 18 of the experiment (the 1st day of administration), simultaneous injection of Mg 2+ and MRF increased the nociceptive threshold in comparison to rats receiving only MRF starting from 15 min after administration (166.66 ± 3.33 g and 149.167 ± 5.23 g for STZ + Mg 2+ + MRF and STZ + MRF, respectively; Figure 2A). It is noteworthy that this effect was significantly higher on day 24 of the experiment (the 7th day of application) at the same time point (200 ± 2.23 g and 150.83 ± 3.27 for STZ + Mg 2+ + MRF and STZ + MRF, respectively; Figure 2B).

Changes in the Expression of Phosphorylated Proteins Measured at the Serine Residues of N-methyl-D-aspartate Receptor NR1 Subunit and µ Opioid Receptor
To determine changes in the level of protein phosphorylation at the Ser375 MOR ( Figure 3) and at the Ser896 NMDAR NR1 subunit (Figure 4), we used the Western blot technique in PAG lysate. As shown in Figure 3B, phosphorylation of Ser375 at MOR for STZ-treated rats was significantly higher compared to the control group (healthy animals). Importantly, Mg 2+ administered alone in rats with STZ-induced diabetic neuropathic pain resulted in a statistically significant decrease in the protein phosphorylation on Ser375 MOR in relation to the STZ group (0.35 ± 0.08 vs. 1 for STZ group; p < 0.001). MRF injected alone showed no changes in the phosphorylation of tested protein compared with the STZ group (1.07 ± 0.1 vs. 1 for the STZ group; p > 0.05). In turn, simultaneous administration of Mg 2+ and MRF caused a significant reduction in the phosphorylation of MOR compared to the STZ group receiving MRF alone (0.74 ± 0.09 vs. 1.07 ± 0.1; p < 0.05). In the case of ionotropic NMDAR ( Figure 4B), high phosphorylation of the Ser896 NR1 subunit was observed in STZ-treated rats compared to the control group (p < 0.05). Treating STZ animals with MRF did not affect the phosphorylation of the NR1 in relation to the STZ group (0.94 ± 0.13 vs. 1, p > 0.05). In contrast, coadministration of Mg 2+ and MRF generated a statistically significant decrease in the level of NR1 phosphorylation in rats with STZ-induced diabetic neuropathic pain in comparison to the non-treated STZ Figure 1. Influence of magnesium (Mg 2+ ) on the analgesic activity of morphine (MRF) in streptozotocin (STZ)-treated rats. Days 18-24-drugs administration; days 19-25-measurements of prolonged activity of investigated drugs; days 26-33-measurements of pain threshold after discontinuation of drugs administration. Nociceptive thresholds, expressed in grams, were measured before STZ administration (baseline) and then before tested substances administration in rats with STZ-induced diabetic neuropathic pain from day 19 until day 25. Data are presented as means ± SEM. Statistical analysis was performed by two-way ANOVA followed by Bonferroni's post-hoc test. *** p < 0.001, * p < 0.05 (STZ + Mg 2+ + MRF vs. STZ), ### p < 0.001, ## p < 0.01, # p < 0.05 (STZ + Mg 2+ + MRF vs. STZ + MRF), $$$ p < 0.001, $$ p < 0.01 (STZ + Mg 2+ vs. STZ); nn = 6 rats for each group.

Changes in the Activity of Protein Kinase A and Protein Kinase C in the Streptozotocin-Treated Rats after Administration of Tested Compounds
As shown in Figure 5B, an increase in the activity of PKC was observed in STZ-treated rats compared to the control animals. In turn, no statistically significant changes were recorded in levels of PKA in these two groups ( Figure 5A). However, for both PKA and PKC, the 7-day application of Mg 2+ led to a reduction in their activities (0.49 ± 0.08 vs. 1, p < 0.01 for PKA and 0.48 ± 0.05 vs. 1, p < 0.001 for PKC). Importantly, the application of MRF to animals with chronic neuropathic pain resulted in an increase in the activity of both kinases compared to the STZ group (1.51 ± 0.16 vs. 1 for PKA and 1.31 ± 0.04 vs. 1 for PKC; p < 0.01). Simultaneous administration of Mg 2+ and MRF showed no changes in the level of PKA and PKC (1.08 ± 0.1 for PKA and 0.8 ± 0.11 for PKC, p > 0.05) in relation to the STZ group. However, statistically significant changes in the activity of both enzymes were observed after injection of Mg 2+ and MRF in relation to MRF alone (1.08 ± 0.1 vs. 1.51 ± 0.16 for PKA, p < 0.05 and 0.8 ± 0.11 vs. 1.31 ± 0.04 for PKC, p < 0.001). There were no statistically significant differences in total MOR expression (p > 0.05; graph C); nn = 6 rats/group. There were no statistically significant differences in total NR1 (p > 0.05; graph C); nn = 6 rats/group.

Changes in the Activity of Protein Kinase A and Protein Kinase C in the Streptozotocin-Treated Rats After Administration of Tested Compounds
As shown in Figure 5B, an increase in the activity of PKC was observed in STZ-treated rats compared to the control animals. In turn, no statistically significant changes were recorded in levels of PKA in these two groups ( Figure 5A). However, for both PKA and PKC, the 7-day application of Mg 2+ led to a reduction in their activities (0.49 ± 0.08 vs. 1, p < 0.01 for PKA and 0.48 ± 0.05 vs. 1, p < 0.001 for PKC). Importantly, the application of MRF to animals with chronic neuropathic pain resulted in an increase in the activity of both kinases compared to the STZ group (1.51 ± 0.16 vs. 1 for PKA and 1.31 ± 0.04 vs. 1 for PKC; p < 0.01). Simultaneous administration of Mg 2+ and MRF showed no changes in the level of PKA and PKC (1.08 ± 0.1 for PKA and 0.8 ± 0.11 for PKC, p > 0.05) in relation to the STZ group. However, statistically significant changes in the activity of both enzymes were observed after injection of Mg 2+ and MRF in relation to MRF alone (1.08 ± 0.1 vs. 1.51 ± 0.16 for PKA, p < 0.05 and 0.8 ± 0.11 vs. 1.31 ± 0.04 for PKC, p < 0.001).

Discussion
Despite medical advances of the last few decades, opioids are still one of the powerful groups of analgesics. These compounds are used not only in the treatme cancer-related pain and acute postoperative pain but also, with uncertain effectivene the treatment of neuropathic pain [47]. In clinical practice, a regimen know co-analgesia is employed to increase the analgesic efficiency of opioids as well as duce their doses and thereby avoiding the risk of side effects. NMDAR antagonis Results are presented as means ± SEM. Statistical analysis was performed using one-way ANOVA followed by Dunnett's post-hoc test. *** p < 0.001; ** p < 0.01; * p < 0.05 vs. STZ; ### p < 0.001; ## p <0.01, # p < 0.05; vs. STZ + MRF; &&& p < 0.001, && p < 0.01 vs. STZ + Mg 2+ ; n = 6 rats/group.

Discussion
Despite medical advances of the last few decades, opioids are still one of the most powerful groups of analgesics. These compounds are used not only in the treatment of cancer-related pain and acute postoperative pain but also, with uncertain effectiveness, in the treatment of neuropathic pain [47]. In clinical practice, a regimen known as co-analgesia is employed to increase the analgesic efficiency of opioids as well as to reduce their doses and thereby avoiding the risk of side effects. NMDAR antagonists are widely studied co-analgesics, and coadministration of Mg 2+ with opioids has received particular interest. However, despite many reports regarding the beneficial interaction between Mg 2+ and opioids [48], the mechanism of this phenomenon still remains to be elucidated.
In our studies, we focused on PAG structure, which is a part of the descending pain modulatory pathway and is involved in the opioid control of nociception [49]. Moreover, since MOR and NMDAR colocalize on the cell membranes in PAG, but in neuropathic pain, NMDAR activity negatively affects the signaling capacity of MOR, we decided to investigate the activities of both mentioned receptors after administration of Mg 2+ and MRF.
The starting point for characterization of this interaction was a behavioral study investigating the analgesic potential of coadministration of tested drugs in the model of STZ-induced diabetic neuropathic pain. We studied the antihyperalgesic effect of tested drugs after the chronic administration as well as an analgesic response over a period of 120 min directly after the drugs' application on day 18 and day 24 of the experiment using mechanical stimulus. Some data indicate that PAG stimulation was effective in mechanical allodynia in the animal model of neuropathic pain [50]. As shown in our previous studies [25,51], the gradual development of mechanical hyperalgesia was observed in STZtreated animals (data not shown). We demonstrated a lack of antinociceptive effect upon MRF application alone over 7 consecutive days in animals with persistent neuropathic pain. Moreover, the nociceptive threshold data measured over a period of 120 min after MRF administration on the 1st day of application did not modify STZ-induced hyperalgesia ( Figure 2A). However, on day 7 of consecutive MRF administration, the antinociceptive activity of the opioid occurred 15 min after the injection ( Figure 2B). The weak analgesic activity of MRF has also been observed in our previous studies in different kinds of neuropathic pain [24,25]. Moreover, other authors have shown that MRF administered alone induced no change in the nociceptive threshold in diabetic rats or fairly modest antihyperalgesic effect in mononeuropathic animals [23]. In contrast, Mg 2+ applied alone elevated the mechanical pain threshold in the group of animals with diabetic hyperalgesia (Figure 1). This is probably the result of chronic use of subtherapeutic doses of Mg 2+ , which gradually inhibit the NMDAR hyperactivity. Additionally, on day 7 of application, this analgesic effect lasted over 2 h after Mg 2+ administration ( Figure 2B). The results of these studies are not surprising since similar effectiveness of Mg 2+ , as well as synthetic NMDAR antagonists, has been demonstrated by other laboratories in neuropathic pain [11,19,23,52]. Importantly, it was also demonstrated that coadministration of Mg 2+ with MRF produced a significant reduction in mechanical hyperalgesia in STZ-treated rats after systemic few days of administration. This effect was also maintained for several days after cessation of drug administration. Similar results have been observed in previous studies in a vincristine neuropathic model of pain [24]. This is potentially associated with the action of morphine-6glucuronide (M6G), the analgesic active metabolite of MRF, which affects MOR [53]. M6G, during long-term MRF administration, may reach sufficiently high concentrations in CNS and cause longer-lasting pain relief compared to MRF [53,54]. We can speculate that Mg 2+ blocking of the NMDAR signaling and the likely unblocking of MOR activity allows for the prolonged action of MRF metabolites. Comparing Mg 2+ cotreatment with MRF to the inferior analgesic effect observed after separate administration of both compounds, the coadministration of these substances appears to be highly effective in the alleviation of neuropathic pain. Results of this study confirm previous observations on the efficacy of such combination in relieving various types of neuropathic pain [23][24][25][26]. Thus, the intensification of opioid analgesia by Mg 2+ can provide an opioid-sparing dose, which in turn allows for the reduction in side effects. Unfortunately, as we mentioned above, the nature of synergistic analgesic interactions between opioids and Mg 2+ remains unclear.
One of the most important processes regulating the receptor's activity is the covalent modification of receptors by their phosphorylation or dephosphorylation [27]. In the case of the examined receptors, their phosphorylation results in metabotropic MOR inactivation and ionotropic NMDAR stimulation. Therefore, in our studies, the activity of the abovementioned receptors was determined by evaluation changes in protein phosphorylation. It was observed that the phosphorylation of both, Ser375 at the C-terminus MOR ( Figure 3B) and Ser896 of the NMDAR NR1 subunit ( Figure 4B), was significantly higher in STZ-treated rats than in control animals. Our results are in agreement with those of Rondon et al., who also observed NMDARs hyperactivity based on increased Ser896 phosphorylation in STZ-treated rats [19]. The high levels of an inactive, phosphorylated form of MOR detected in STZ-treated rats explain the lack of effectiveness of MOR agonists, such as MRF, in treating neuropathic pain. In addition, the phosphorylation of Ser375 at MOR after chronic administration of MRF to STZ-treated animals was comparable to the diabetic rats. These results are in line with our behavioral data showing poor analgesic properties of MRF in the neuropathic model of pain.
Importantly, our data showed that chronic administration of Mg 2+ in rats with STZinduced diabetic neuropathic pain significantly reduced the level of Ser375 phosphorylation at MOR in relation to STZ-treated animals ( Figure 3B). Moreover, we demonstrated significantly reduced NMDAR phosphorylation after chronic application of Mg 2+ in STZ-treated rats to levels observed in control animals ( Figure 4B). These results, together with behavioral observations, indicate that blocking NMDAR signaling by Mg 2+ induces the MOR-NMDAR re-association and restores the analgesic efficacy of MRF in neuropathic pain. Furthermore, our data revealed that the simultaneous administration of MRF and Mg 2+ caused a reduction in Ser375 phosphorylation in comparison to rats receiving MRF alone. However, the Ser375 phosphorylation was greater than in the group treated only with Mg 2+ . The data presented above demonstrate the concurrent activity of the signaling pathways involving both receptors. MOR indirectly dephosphorylated due to the activity of Mg 2+ can bind MRF and transmit the signal. Subsequently, after signal transmission, the phosphorylation of MOR is again elevated. Thus, this observation seems to explain the mechanism of unblocking morphine's efficacy by Mg 2+ .
In our studies, we also examined the activity of PKA and PKC involved in the bidirectional regulation between MOR and NMDAR. Surprisingly, we found that chronic administration of MRF to STZ-treated rats significantly increased the activity of both kinases compared to untreated diabetic animals, although the level of MOR phosphorylation suggested its inactivation in a state of neuropathic pain. Nevertheless, we note that MRF is a nonselective agonist, and it influences the activity of kinases through other receptor subtypes. Moreover, it was established that MORs and δ opioid receptors (DORs) colocalize in the same axonal terminals of some CNS structures, including PAG and spinal cord [55][56][57]. A number of studies have shown that MOR and DOR undergo cross-regulation, and the DOR has a negative influence on MOR function [58]. Particularly, the administration of DOR antagonists decreases the development of MRF-induced tolerance and dependence [59]. According to Garzon et al., MRF preferentially influences MOR, but DOR can also be involved in the chronic responses to MRF [46]. Moreover, data described by Gucker and Bidlack [60] showed that PKC can play a significant role in DOR activity. They revealed that MRF in the presence of a PKC activator increases DOR down-regulation in vitro. Additionally, in our behavioral studies, a slight analgesic effect was observed 15 min after injection on day 7 of the chronic MRF administration. Thus, we suspect that in our studies increase in PKC and, consequently, PKA activity compared to the STZ group could be partially associated with the activation of DOR.
Importantly, our study revealed that chronic administration of Mg 2+ causes a significant reduction in the activities of PKA and PKC in STZ-treated animals in relation to diabetic rats ( Figure 5). It could be speculated that NMDAR inhibition promoted the reduction in PKA level and consequently decreased PKC activity. Data available in the literature indicate that activation of PKC and PKA may cause NMDAR NR1 Ser896 and Ser897 residues phosphorylation [61,62], further supporting the evidence that intracellular mechanisms activate NMDAR. Moreover, simultaneous induction of both kinases results in the transfer of newly synthesized NMDAR subunits from the cytosolic fraction to the membrane leading to central sensitization [63]. Thus, reduced activity of both enzymes after chronic Mg 2+ stimulation may inhibit the transfer of NMDAR subunits from the cell interior to the cell membrane.
In the group receiving the combination of both tested drugs, the activity of PKA and PKC was statistically significantly lower than after MRF administration alone but significantly higher compared to diabetic rats treated only with Mg 2+ . Thus, it can be concluded that the NMDAR inhibition by Mg 2+ and consequently the reduction in PKA level restores the activity of MOR. In addition, PKA-induced heterologous MOR phosphorylation is reduced by Mg 2+ , which corresponds to reduced PKC activity observed in our study. The combination of Mg 2+ and MRF enables MRF to stimulate MOR, and consequently, a greater PKC activation was observed, representing the beginning of the cascade leading to heterologous MOR phosphorylation and its inactivation.
Interestingly, in our studies, we used the systemic route of drug administration and demonstrated the antinociceptive effect of the simultaneous use of Mg 2+ and MRF. As both of investigated drugs cross the blood-brain barrier [64,65] and MOR and NMDAR colocalize in PAG, we suggest that the mechanism of this interaction is direct in PAG neurons.

Laboratory Animals
Experiments were carried out on male Wistar rats weighing 250 to 280 g. The animals were housed in rooms maintained at a temperature of 22 ± 2 • C and humidity of 55% ± 10%. The rooms were equipped with an efficient mechanical ventilation system (15-20 air changes/h) and a 12 h light/12 h dark cycle. The rodents had free access to food and water. However, in an animal model of streptozotocin (STZ)-induced diabetic neuropathic pain, food was removed 16 h before STZ administration. Experimental groups consisted of at least 6 animals.

Drug Administration
Diabetes was induced by a single intramuscular administration of STZ (Cayman Chemical, Ann Arbor, MI, USA) at a dose of 40 mg/kg of body weight according to the method described previously [25,66]. Morphine sulfate (Polfa, Warsaw, Poland) and magnesium sulfate (MgSO 4 ) (Polfa, Warsaw, Poland) were dissolved in 0.9% saline immediately before injection. Both tested compounds were applied intraperitoneally (i.p.). MRF was injected at a dose of 5 mg/kg, while MgSO 4 at a dose of 40 mg/kg (i.e., 8 mg/kg Mg 2+ ). The doses of MRF and MgSO 4 were chosen based on our previous studies [25]. Control (healthy) rats were administered intraperitoneally with 0.9% saline.

Time Schedule
In behavioral experiments, all the tested drugs (except STZ administered only on day 1) were injected on 7 consecutive days (from day 18 to 24 after STZ was given) to rats with developed hyperalgesia. Mg 2+ was applied daily 30 min before MRF injection.
In experiments involving tissue harvest for Western blot analysis and enzyme-linked immunosorbent assays, the animals were killed by decapitation on day 24 of the experiment. Then, PAG was separated from the isolated brain and immediately immersed in liquid nitrogen.

Measurement of the Pain Threshold
The changes in pain thresholds in the experimental animals after administration of tested substances were estimated using a mechanical stimulus, the Randall-Selitto paw withdrawal test as described previously [67]. The pain threshold obtained for each rat before STZ administration constituted the baseline. Nociceptive thresholds for the STZ-treated rats were measured each day after STZ administration (data not shown). Measurements of prolonged activity of the tested substances were carried out daily in rats with developed hyperalgesia for 7 consecutive days before drugs administration (from day 19 until day 25 of the experiment) and then after cessation of compounds application until the end of the experiment (day 33). The changes in nociceptive thresholds were also assessed at 15,30,45,60,90, and 120 min after the application of the investigated drugs (on the 1st and the 7th day of administration; day 18 and 24 of the experiment) ( Figure 6).

Preparation of Lysates for Western blot Analysis
The research material was prepared based on the method originally developed previously [68,69] with some modifications. All reagents were purchased from Sigma-Aldrich (St Louis, MO, USA) unless otherwise stated. PAG structures were homogenized on ice in 10 volumes of homogenization buffer (25 mM Tris-HCl (pH = 7.4), 1 mM EGTA, 0.32 M sucrose, 20 µM H-89 () as well as phosphatase inhibitor cocktail (P2850) and protease inhibitor mixture (P8340)). The homogenate was then centrifuged for 10 min at 1000× g to eliminate the nuclear fraction (pellet). The supernatant (S1) was centrifuged (20 min, 20,000× g) to obtain the second supernatant (S2) containing the cytoplasmic fraction and the pellet (P2). P2 was resuspended in 200 µL of homogenization buffer and centrifuged for 20 min at 20,000× g. The supernatant (S3) was discarded, and the pellet (P3), containing the membrane fraction, was resuspended in 25 mM Tris-HCl (pH 7.4) with protease inhibitors (0.2 mM PMSF, 2 µg/mL leupeptin, 0.5 µg/mL aprotinin). The P3 solutions containing 500 µg of protein were centrifuged (20 min, 20,000× g), and the obtained pellet was resuspended in 200 µL of an extraction buffer with the following composition: 50 mM Tris-HCl (pH 7.7), 50 mM NaCl, 1% Nonidet P-40, 20 µM H-89, phosphatase, and protease inhibitor cocktail. The samples were shaken for 12 h at 4 °C, centrifuged (10 min, 4 °C, 10,000× g) and the obtained supernatant was stored at −80 °C until further use. Before Western blot analysis, protein concentration was measured using the BCA method (Pierce™ BCA Protein Assay Kit, Thermo Fisher Scientific, Rockford, IL, USA). Changes in pain thresholds after mechanical stimulation for each rat were expressed in grams as means ± SEM.

Preparation of Lysates for Western blot Analysis
The research material was prepared based on the method originally developed previously [68,69] with some modifications. All reagents were purchased from Sigma-Aldrich (St Louis, MO, USA) unless otherwise stated. PAG structures were homogenized on ice in 10 volumes of homogenization buffer (25 mM Tris-HCl (pH = 7.4), 1 mM EGTA, 0.32 M sucrose, 20 µM H-89 () as well as phosphatase inhibitor cocktail (P2850) and protease inhibitor mixture (P8340)). The homogenate was then centrifuged for 10 min at 1000× g to eliminate the nuclear fraction (pellet). The supernatant (S1) was centrifuged (20 min, 20,000× g) to obtain the second supernatant (S2) containing the cytoplasmic fraction and the pellet (P2). P2 was resuspended in 200 µL of homogenization buffer and centrifuged for 20 min at 20,000× g. The supernatant (S3) was discarded, and the pellet (P3), containing the membrane fraction, was resuspended in 25 mM Tris-HCl (pH 7.4) with protease inhibitors (0.2 mM PMSF, 2 µg/mL leupeptin, 0.5 µg/mL aprotinin). The P3 solutions containing 500 µg of protein were centrifuged (20 min, 20,000× g), and the obtained pellet was resuspended in 200 µL of an extraction buffer with the following composition: 50 mM Tris-HCl (pH 7.7), 50 mM NaCl, 1% Nonidet P-40, 20 µM H-89, phosphatase, and protease inhibitor cocktail. The samples were shaken for 12 h at 4 • C, centrifuged (10 min, 4 • C, 10,000× g) and the obtained supernatant was stored at −80 • C until further use. Before Western blot analysis, protein concentration was measured using the BCA method (Pierce™ BCA Protein Assay Kit, Thermo Fisher Scientific, Rockford, IL, USA).

Western blot Analysis
The obtained lysates were electrophoresed in 10% SDS polyacrylamide gels in an electrophoresis buffer (25 mM Tris-HCl, 143 mM glycine, 0.1% SDS) using the Mini-Protean ® Tetra System by Bio-Rad (Hercules, CA, USA) at a constant voltage of 100 V for about 90 min. All reagents were purchased from Sigma-Aldrich (St Louis, MO, USA) unless otherwise stated. In the next step, proteins were transferred (100 V, 100 min) to polyvinylidene fluoride membrane (PVDF, Immobilon ® -FL Transfer Membranes, Darmstadt, Germany). Membranes were blocked in blocking buffer (5% non-fat dried milk in Tris-buffered saline-0.1% Tween 20 (TBS-T), pH 7.6) for 1 h at room temperature and incubated at 4 • C overnight with primary antibodies for phosphorylated NMDAR1 (rabbit anti-NMDAR1 phospho Ser896; 1:500; cat. no. ab 75680, Abcam, Cambridge, UK) and phosphorylated MOR (rabbit anti-MOR phopsho Ser375; 1:1000; cat. no. 3451, Cell Signaling Technology, Danvers, MA, USA) diluted in blocking buffer. Then, the membranes were washed four times in TBS-T buffer and incubated for 1 h at room temperature with a secondary anti-rabbit antibody (1:3000; cat. no. #7074, Cell Signaling Technology, Danvers, MA, USA) conjugated to horseradish peroxidase. The chemiluminescent detection process was performed using WesternBright™ ECL reagents from Advansta (San Jose, CA, USA). The obtained results were recorded on the Carestream X-ray film (Rochester, NY, USA).
Then, in order to investigate total NMDAR and total MOR, the membranes were stripped using RestoreTM PLUS Western Blot Stripping Buffer from Thermo Scientific (Rockford, IL, USA). The effectiveness of stripping was confirmed at the beginning of the study by developing the stripped membranes. As no bands were detected, the stripping procedure was used over the study. Nevertheless, occasionally the randomly chosen stripped membranes were developed to confirm that the stripping procedure was effective during the whole study. Next, the membranes were reprobed with primary antibodies for NMDAR (1:1000; cat. no. ab 109182, Abcam, Cambridge, UK) and MOR (1:500; cat. no. ab 177898, Abcam, Cambridge, UK). The further procedure was as above.

Enzyme-Linked Immunosorbent Assays (ELISA): Quantification of Protein Kinase A and Protein Kinase C Activity
Changes in the level of PKA and PKC in the STZ-treated rats after administration of tested compounds were assessed using commercially available kits from Abcam, Cambridge, UK (PKA (ab 139435) and PKC (ab 139437)). The assay procedure was strictly following the instructions of the kits.

Data Analysis
All statistical calculations were performed using GraphPad Prism 5 (GraphPad, San Diego, CA, USA). Results from behavioral experiments were expressed as means ± SEM. Statistical analysis was performed by two-way ANOVA followed by the Bonferroni posthoc test. For Western blot assay, optic density was measured using the Image LabTM Software version 5.2 from Bio-Rad (Hercules, CA, USA). Results were expressed as the ratio of the phosphorylated form to the total form of the tested proteins (pMOR/MOR and pNR1/NR1) related to the STZ-treated rats. Changes in the activity of PKA and PKC were presented as relative values compared to the STZ-treated rats. Statistical analyses were conducted using one-way ANOVA followed by Dunnett's post-hoc test. Differences were considered significant at * p < 0.05, ** p < 0.01, and *** p < 0.001.

Conclusions
The processes presented in our study, relating to two opposing receptors, may have a significant effect on designing the bifunctional drugs containing the NMDAR antagonist and the MOR agonist. We suggest that blocking the NMDAR channel by Mg 2+ restores the MOR-NMDAR complex crucial for MRF analgesia in neuropathic rats. It is worth noting that Mg 2+ has a high safety profile compared to synthetic NMDAR antagonists. The mechanisms of action described in the paper are a justification for the inclusion of Mg 2+ into opioid therapy, especially during the treatment of opioid-resistant neuropathic pain.