Disulfiram Abrogates Morphine Tolerance—A Possible Role of µ-Opioid Receptor-Related G-Protein Activation in the Striatum

One of the key strategies for effective pain management involves delaying analgesic tolerance. Early clinical reports indicate an extraordinary effectiveness of off-label disulfiram—an agent designed for alcohol use disorder—in potentiating opioid analgesia and abrogation of tolerance. Our study aimed to determine whether sustained µ-opioid signaling upon disulfiram exposure contributes to these phenomena. Wistar rats were exposed to acute and chronic disulfiram and morphine cotreatment. Nociceptive thresholds were assessed with the mechanical Randal-Selitto and thermal tail-flick tests. µ-opioid receptor activation in brain structures important for pain processing was carried out with the [35S]GTPγS assay. The results suggest that disulfiram (12.5–50 mg/kg i.g.) augmented morphine antinociception and diminished morphine (25 mg/kg, i.g.) tolerance in a supraspinal, opioid-dependent manner. Disulfiram (25 mg/kg, i.g.) induced a transient enhancement of µ-opioid receptor activation in the periaqueductal gray matter (PAG), rostral ventromedial medulla (RVM), hypothalamus, prefrontal cortex and the dorsal striatum at day 1 of morphine treatment. Disulfiram rescued µ-opioid receptor signaling in the nucleus accumbens and caudate-putamen 14 days following morphine and disulfiram cotreatment. The results of this study suggest that striatal µ-opioid receptors may contribute to the abolition of morphine tolerance following concomitant treatment with disulfiram.


Introduction
Management of refractory pain often involves long-lasting exposure of patients to opioids, which have a long history of mainstay medicine for effective relief of moderate to severe pain [1][2][3]. A plethora of different factors influence the extent of opioid tolerance development. Among molecular mechanisms, downregulation of cell surface receptors, effector uncoupling and desensitization, impaired receptor trafficking, intrinsic drug efficacy and intensity of nociceptive input are the most frequently mentioned. Additionally, the influence of other factors such as dosing regimen, route of administration, drug halflife and plasma concentration were described as important contributing factors (see [4]). Psychological factors have also been shown to play a part in the phenomenon of analgesic tolerance. Unfortunately, a significant percentage of opioid-treated chronic pain patients do not adhere to the prescribed pain management regimen mostly because of their incapability to cope with pain-evoked physical and emotional distress. This misuse of opioids may precipitate the development of hyperalgesia or tolerance as recent studies have shown [5]. Strategies to overcome opioid tolerance include careful drug titration, dose escalation, opioid rotation or use of adjunctive drugs to reduce opioid consumption [6]. A broad selection of adjuncts are available for opioid therapy, including antidepressants [7], anticonvulsants [8], nonsteroidal anti-inflammatory drugs [9], calcium channel blockers [10], NMDA receptor antagonists [11], cholinesterase inhibitors and allosteric modulators of the glutamate receptor 7 (mGlu7) [12,13]. As adjuncts to opioid therapy are used with the key intention of potentiating analgesia and only indirectly affect the development of tolerance only by reducing opioid consumption, their use in a chronic opioid use setting may prove insufficient in the long run [14,15] Additionally, patients often display variability in response to available treatment schemes, thus, introducing novel adjuncts to opioid therapy is one step towards tailoring pain management to the patients' individual needs.
For decades, disulfiram (DSF) has been used for the pharmacological support of abstinence in patients with chronic alcohol use disorder [16]. By inhibiting acetaldehyde dehydrogenase, it impairs alcohol metabolism and causes buildup of toxic acetaldehyde which causes unpleasant hangover-like effects immediately following ingestion of even small amounts of alcohol [17]. Interestingly, off-label DSF was shown to produce clinically relevant enhancement of stimulation-induced analgesia in refractory pain patients that was insensitive to the development of tolerance for a period of 10 months [18]. The possible mechanistic underpinnings of this phenomenon have not yet been clarified, but several hypotheses were put forward. One hypothetical mechanism proposed by authors in early studies involves enhancement of dopamine release by inhibition of β-hydroxylase [19][20][21]. Of note, dopamine, along with the opioid system, was evidenced as an important player in nociceptive processing as the two neurotransmitter systems topographically overlap and interact in many ways [22][23][24]. Additionally, DSF is converted to many active metabolites that together affect the release of other neurotransmitters such as γ-aminobutyric acid (GABA) and glutamate, which are important mediators of opioid analgesia, withdrawal and tolerance [25][26][27][28]. Another more recent hypothesis assumes that opioid tolerance arises from biased signaling featuring enhanced µ-opioid receptor uncoupling from the G i/0 protein and its coupling to the G s protein [29].
Hence, in the present study, we aimed to investigate the possible influence of chronic DSF treatment in morphine (MRF)-tolerant rats on µ-opioid receptor associated G i/0 signaling in brain structures important for pain processing.

The Effect of Disulfiram on Mechanical and Thermal Thresholds
As shown by two-way ANOVA, a single delivery of disulfiram weakly but significantly increased nociceptive thresholds in response to both mechanical (F 3,160 = 28.9; p < 0.001, Figure 1A) and thermal (F 3,160 = 50.9; p < 0.001, Figure 1B) stimuli. The effect amounting to less than 10% maximum possible effect (%MPE) was dose-independent, surfaced 30 min postdelivery and remained stable for 90 min (mechanical stimulus) or 120 min (thermal stimulus).

The Effect of Disulfiram on Morphine-Induced Antinociception
When disulfiram was coadministered with morphine, augmentation of the antinociceptive effect was observed in response to mechanical (F 4,200 = 57.9; p < 0.001, Figure 2A) and thermal (F 4,200 = 119.7; p < 0.001, Figure 2B) stimuli. However, the effect was significant (30-90 min postadministration) for all DSF doses, only when mechanical stimulation was used. When the area under curve (AUC) values were calculated, each dose of DSF potentiated MRF-induced antinociception by 1.4-, 1.6-and 1.7-fold, respectively ( Figure 2D). However, when animals were exposed to thermal noxious stimulation, only the highest DSF dose elicited a response at 30-60 min postadministration with AUC values only 1.4-fold higher than in the MRF alone group ( Figure 2C). In the Randal-Selitto test, DSF augmented MRF antinociception in a synergistic manner at any dose tested (12.5-50 mg/kg) with CI values of 0.69, 0.61 and 0.66, respectively (Table 1). In the tail-flick test, very weak synergy was observed for the highest dose of DSF (CI = 0.9). When MRF was administered together with 12.5 and 25 mg/kg DSF, no synergistic or additive effect was observed (CI = 1.12 and 1.04, respectively).

The Effect of Disulfiram on Morphine Tolerance
As indicated by two-way ANOVA, DSF altered nociceptive thresholds in MRF-treated rats when measured in the Randal-Selitto (F 3,580 = 399.4; p < 0.001, Figure 4) and tail-flick (F 3,580 = 137.1; p < 0.001, Figure 5) tests. The antinociceptive effect of MRF declined within 9 or 10 days of treatment (depending on the nociceptive test used), announcing the onset of tolerance. However, all DSF rescued the antinociceptive effect of MRF throughout the treatment regimen regardless of the dose used. In the Randal-Selitto test, rats receiving the lowest (12.5 mg/kg) and intermediate (25 mg/kg) doses of DSF, had higher mechanical thresholds from day 9 until day 22 of treatment ( Figure 4A). Mechanical thresholds of rats receiving MRF with the highest dose of DSF (50 mg/kg) were higher within the first two days of treatment and from day 9 until day 23 when compared with MRF-treated rats. When the time-course area under curve (AUC) were calculated, mechanical nociceptive thresholds in MRF rats cotreated with 12.5, 25 and 50 mg/kg of DSF were higher by 2.7-, 3.5-and 6-fold, respectively, when compared to MRF alone ( Figure 4B). Control animals received either water for injection (MRF group control) or 0.1% methylcellulose (MRF + DSF group control). Baseline thresholds were measured at time 0. Drugs were delivered consecutively from day 1 until day 21. Drugs were discontinued on day 22 and nociceptive thresholds were measured until day 28. Results from the time-course curves were expressed as control-subtracted means ± SEM and analyzed with two-way ANOVA followed by Bonferroni's post hoc test. Area under the curve (AUC) data were analyzed with the Kruskal-Wallis test. Statistical differences between MRF-treated and MRF + DSF cotreated groups were indicated as follows: * p < 0.05; ** p < 0.01; *** p < 0.001 (MRF vs. MRF + DSF 12.5); # p < 0.05; ## p < 0.01; ### p < 0.001 (MRF vs. MRF + DSF 25); $$$ p < 0.001 (MRF vs. MRF + DSF 50).
In the tail-flick test, rats treated with a combination of MRF and the lowest dose of DSF (12.5 mg/kg) showed elevated pain thresholds from day 13 until day 22 of treatment. Administration of 25 mg/kg DSF together with MRF resulted in a similar effect that surfaced from day 12 and lasted until day 23 ( Figure 5A). Coadministration of the highest DSF dose tested (50 mg/kg) with MRF, resulted in a transient elevation of thermal thresholds in the first two days of cotreatment. Further enhancement of the antinociceptive effect of MRF was noted from day 10 post-treatment that persisted until day 22. Calculation of the AUC values yielded results showing that MRF groups cotreated with 12.5, 25 and 50 mg/kg of DSF, showed higher thermal pain thresholds by 0.63-, 1.06-and 1.18-fold than rats treated with MRF alone ( Figure 5B). MRF + DSF 50. Control animals received either water for injection (MRF group control) or 0.1% methylcellulose (MRF + DSF group control). Baseline thresholds were measured at time 0. Drugs were delivered consecutively from day 1 until day 21. Drugs were discontinued on day 22 and nociceptive thresholds were measured until day 28. Results from the time-course curves were expressed as control-subtracted means ± SEM and analyzed with two-way ANOVA followed by Bonferroni's post hoc test. Area under the curve (AUC) data were analyzed with one-way ANOVA. Statistical differences between MRF-treated and MRF + DSF cotreated groups were indicated as follows: * p < 0.05; ** p < 0.01; *** p < 0.001; (MRF vs. MRF + DSF 12.5); # p < 0.05; ### p < 0.001 (MRF vs. MRF + DSF 25); $ p < 0.05; $$ p < 0.01; $$$ p < 0.001 (MRF vs. MRF + DSF 50).

Discussion
In the present study, we aimed to verify the hypothesis that disulfiram (DSF) enhances µ-opioid receptor G-protein signaling, which could in part contribute to the augmentation of morphine (MRF) antinociception and abolition of morphine tolerance development. A single bolus of DSF alone produced meager but significant antinociceptive action (Figure 1), which does not seem to be related to direct opioid receptor activation. As evidenced by our recent observations, although DSF alone produced potent enhancement of G-protein activity in the [ 35 S]GTPγS assay, this action was not reversed by NTX ( Figure S1A). Moreover, DSF altered neither potency nor efficacy of MRF-induced receptor stimulation in brain homogenates from naïve rats ( Figure S1B). Thus, the pharmacological interplay between DSF and MRF could involve indirect mechanisms. The conducted behavioral assays confirmed that DSF rescued the antinociceptive effectiveness of chronically administered MRF (Figures 4 and 5). This observation correlated with results from the functional ex vivo assay, where µ-opioid receptor signaling was not fully abolished following chronic MRF treatment (Table 2, Figure 7).
The first report describing the potential use of DSF as an adjunctive medication to pain management therapy was released in the late 1970s by Hosobuchi and Wemmer [18]. In that particular investigation, intractable pain patients experienced potentiation of PAG stimulation-induced analgesia (SIA) when concomitantly receiving DSF. Additionally, DSF provided long-standing protection against the development of analgesic tolerance. Moreover, DSF alone potently decreased opioid demand when SIA was discontinued [30]. These findings are in conformity with our observations as we also saw synergistic potentiation of MRF antinociception even after a single dose of DSF ( Figure 2, Table 1). Moreover, the effect of both drugs was NTX-reversible, confirming the contribution of opioid receptor signaling (Figure 3).
The notion that µ-opioid receptors may play a vital role in the antinociceptive effect of DSF is derived from the well-acknowledged fact that they modulate the activity of pain-inhibitory neurons in the ventrolateral PAG (vlPAG). Namely, these neurons are under tonic inhibition from GABAergic interneurons, which is suppressed by µ-opioid receptor activation. This subsequently activates glutamatergic transmission in the RVM, leading to pain suppression [31,32]. Our investigation revealed a significant increase in µ-opioid receptor G-protein activation 23 h following drug delivery in the PAG of DSF and MRF cotreated rats compared with MRF alone. A similar trend was observed when rats receiving concomitant DSF and MRF treatment were compared with control animals. Similar results were also obtained for the RVM and prefrontal cortex, which both send their projections to the PAG [33] (Table 2, Figure 6). One might thus expect enhanced µ-opioid-dependent suppression of GABA release following DSF treatment and the ability of naltrexone to negate this effect when a mechanical stimulus is applied.
It seems that potentiation of MRF antinociception by acute DSF delivery was restricted to supraspinal sites, which are responsible for mechanical pain transmission. Morphine, apart from having a direct inhibitory effect on primary afferent neuron excitability, also increases output from the PAG to the RVM by decreasing GABAergic transmission. As confirmed recently, mechanical pain is modulated by presynaptic enkephalin and GABA release from dorsal horn interneurons located mainly in laminae II and III. Their suppression increases mechanical but not thermal sensitivity due to a limited number of heat-sensitive C-fibers receiving input from enkephalinergic interneurons in lamina I [34,35]. Thus, it is possible that DSF enhances enkephalin release from dorsal horn interneurons, causing presynaptic inhibition of mechanosensory neurons. Especially, when DSF was previously shown to increase plasma Met-enkephalin levels in dogs subjected to an acute ethanol challenge [36].
As mentioned earlier, our study has shown major changes in µ-opioid receptor Gprotein activation as early as on day 1 of MRF treatment (Table 2, Figure 6). Structures that were mostly affected included the thalamus, hypothalamus, PAG, nucleus accumbens and to a lesser extent the caudate-putamen. The first three structures are involved in nociceptive processing and their stimulation has long been acknowledged to trigger pain-relief in an opioid-related manner [37][38][39]. Moreover, as we have shown earlier, hypothalamic, thalamic and PAG µ-opioid receptor G-protein activity is decreased in a mouse line refractory to opioid-induced analgesia [40]. Of note, the early enhancement of G-protein activity in these structures does not contribute significantly to the long-lasting effect of DSF, as the differences in G-protein activity between MRF-treated and DSF and MRF cotreated rats diminished at day 14. A noticeable decline in DAMGO-stimulated G i/0 -protein activation upon chronic MRF exposure was evidenced before in many of the brain structures studied. This implicates biased functional signaling as the mechanism responsible for the appearance of MRF tolerance [41,42].
As evidenced before, chronic MRF treatment induces µ-opioid receptor adaptive modifications responsible for the decline in drug efficacy over time. It has been reported that chronic MRF treatment induces tolerance that is at least in part owed to desensitization effected by weak receptor internalization and recycling as well as serine 375 phosphorylation [43][44][45]. The phenomenon of tolerance is also thought to be directly linked to regionally specific G i/0 uncoupling, where a progressive reduction in signal transduction is observed without major changes in receptor density [42,46,47]. Stimulation of cAMP production (or cAMP overshoot) by MRF speaks in favor of G i/0 uncoupling and preferential coupling of the receptor to the Gs stimulatory subunit [48,49]. Additionally, a 10-day chronic MRF exposure was shown to increase the G s /G i/0 protein ratio in forebrain cell membrane fractions, which correlated with a trend toward lower DAMGO efficacy [50]. This observation is in line with our behavioral data, where MRF efficacy was compromised at this timepoint.
Recently, the role of the striatal nuclei comprising the nigrostriatal dopamine pathway were indicated in nociceptive processing and pain perception owing to their high density of opioid receptors [51][52][53]. Similarly as in the PAG, chronic DSF exposure was evidenced to attenuate GABA release and decrease glutamate uptake in the striatum [27,54]. Our study provided further functional evidence that upregulation of µ-opioid receptor activity by DSF in the nucleus accumbens and caudate-putamen may correlate with sustained MRF efficacy. Especially, it was shown that µ-opioid receptor-mediated G-protein activation in the striatum is more resistant to desensitization than in the brainstem nuclei [55,56]. Observations that some forms of antinociception rely on nucleus accumbens opioid receptors only when µ-opioid receptors have not yet undergone downregulation further supports their importance of the striatum in chronic opioid efficacy [57].
As the most meaningful changes in µ-opioid receptor signaling were detected in structures of the dopamine circuit, the role of this monoamine in the behavioral effects of DSF cannot be ruled out. Especially when DSF and its active metabolites inhibit β-hydroxylase activity, elevated dopamine levels suppress GABA A receptor signaling in the PAG and trigger upregulation of µ-opioid receptor mRNA in the nucleus accumbens [27,54,[58][59][60][61]. It is possible that in our behavioral investigation, NTX executes its pronociceptive effects by reducing DSF-induced dopamine release. This effect was previously observed in the nucleus accumbens in male rats either exposed to a pleasurable stimulus or challenged with amphetamine [62,63]. However, further studies are needed to confirm the role of dopamine signaling in the antinociceptive effect of DSF. Additionally, other clinically important implications such as increased likelihood of addiction to stimulants or the risk of drug-induced psychosis in schizophrenic patients need to be addressed as well [20,64].
Interestingly, the amygdala of chronic MRF-treated rats was the only structure where an increase in µ-opioid receptor signaling was detected. It is possible that sustained µopioid receptor activation in this brain region perpetuates the development of opioid dependence, as shown before [65,66]. As described in our earlier behavioral study [26], disulfiram abrogated MRF-induced dependence and withdrawal. The mechanistic underpinnings of this phenomenon are not yet known; however, the current study has shown that disulfiram pretreatment causes downregulation of µ-opioid receptor signaling in the amygdala (Table 2, Figure 7). Of course, other mechanisms may also contribute to the reversal of MRF preference and dependence. Interestingly, activation of amygdala neurokinin-1 (NK1) receptors was linked to MRF craving [67] and disulfiram was previously reported to reduce substance-P (SP) release [68].
It may seem that the side-effect profile of DSF, including hepatic, psychiatric and neurotoxic complications [69], may potentially render this drug unsuitable for long-term support of opioid-based pain management-especially when the most recent investigation points to the possibility of liver damage in MRF-treated, alcohol-naïve animals [70]. Firstly, however, most of the adverse reactions were seen in patients with alcohol use disorder. Secondly, the most recent study employed a high-dose DSF treatment scheme, as only the 100 mg/kg dose produced meaningful extension of MRF antinociception when delivered intraperitoneally. Such high doses of DSF carry a greater risk of hepatotoxicity and may additionally cause disturbances in catecholamine levels, leading to neurotoxic effects [21], whereas our study demonstrated that 25 mg/kg DSF is sufficient enough to exert an effect of similar magnitude when given orally, which potentially lowers the risk of hepatitis and/or neurotoxicity. Thus, low-dose oral DSF may still be considered a valuable asset to MRF-based pain relief strategy owing to its robust and long-lasting suppression of analgesic tolerance and lower risk of side-effects.

Animals and Husbandry
The study was conducted in compliance with the guidelines published in the European directive 2010/63/EU on the protection of animals used for scientific purposes. The protocol was approved by the Ethical Committee for Experiments on Small Animals, Medical University of Warsaw (permit numbers: 64/2013 and 101/2016). Outbred, male Cmdb:Wi rats (270-320 g) were housed in a room maintained at a temperature of 20 ± 2 • C, under 12-12 h light-dark cycle. Experimental groups consisted of 6 rats. Animals had free access to food and water.

Chemicals and Supplies
Disulfiram (DSF), morphine sulfate and water for injection were purchased from Polfa (Warsaw, Poland).

Drug Administration
All drugs were administered by oral gavage in a volume of 1 mL/kg. Morphine (MRF, 25 mg/kg), naltrexone (NTX, 25 mg/kg) were dissolved in water for injection. Disulfiram (DSF, 12.5, 25 and 50 mg/kg) was suspended in 0.1% methylcellulose solution as described previously [26]. In experiments where DSF was delivered as a single bolus, control animals received a 0.1% methylcellulose solution. In acute cotreatment experiments, control animals received water for injection (MRF and MRF + NTX group control) or 0.1% methylcellulose (control for groups treated and cotreated with DSF). DSF or NTX were administered 10 min before MRF. In the NTX + DSF + MRF group, NTX was delivered 10 min before DSF and MRF was administered 10 min following DSF. The following experimental groups were included in the chronic behavioral assays: MRF (water + 25 mg/kg morphine), MRF + DSF 12.5 (25 mg/kg morphine + 12.5 mg/kg disulfiram), MRF + DSF 25 (25 mg/kg morphine + 25 mg/kg disulfiram), MRF + DSF 50 (25 mg/kg morphine + 50 mg/kg disulfiram). Control animals received 0.1% methylcellulose (MRF + DSF group control) or water for injection (MRF group control). In behavioral experiments, all drugs were administered for 21 consecutive days. All drugs were discontinued on day 22 and nociceptive thresholds were measured until day 28. In experiments involving tissue harvest for the functional [ 35 S]GTPγS assay, drugs or vehicle (0.1% methylcellulose) were administered either for 1 or 14 days and animals were sacrificed 23 h later.

Behavioral Testing
Nociceptive thresholds were measured in the Randal-Selitto and tail-flick tests as described previously [71]. In the single bolus experiments mechanical a thermal nociceptive thresholds were measured before drug/vehicle administration and every 30 min for 3.5 h. A 1 min interval was ensured between both types of tests. Animals were returned back to their home cages between testing at each timepoint. In chronic drug delivery experiments, single measurements were taken in individual animals every day, starting at 10.00 a.m. at 23 h postadministration. After discontinuation of drug treatment at day 21, behavioral testing continued until day 28. The tail-flick test was performed first. Changes in pain thresholds for each animal were calculated as % maximum possible effect (%MPE) using the following formula (Formula (1)): where T 1 -baseline latency before treatment; T 2 -post-treatment latency; C off -cut-off time.

Measurement of Thermal Thresholds
Thermal thresholds were measured in the tail-flick test according to the thermal stimulation of the tail method by D'Amour and Smith [72] using the I.R. Tail-Flick Unit (Ugo Basile, Comerio, Italy). Briefly, rats were gently restrained by hand. Then, radiant heat was directed onto the ventral aspect of the tail, 4-5 cm from the tip. A positive nocifensive reaction was a vigorous removal of the tail from the heat source, which automatically terminated the test and recorded tail-flick latency (in seconds). Light intensity was set to elicit a baseline response of ca. 6 s. Cut-off time was set to 20 s to avoid burns to the tail.

Measurement of Mechanical Thresholds
Mechanical thresholds were determined in the Randal-Selitto paw pressure test originally described previously [73] with some modifications. Briefly, rats were restrained by hand and increasing pressure was applied to the dorsal surface of the hindpaw at a rate of 32 g/s. Nociceptive thresholds were defined as the force (in grams) that evoked an attempt to withdraw the paw. The cut-off pressure was set to 480 g.

Homogenate Preparation
Rats from the MRF and MRF + DSF 25 groups from day 1 and day 14 of treatment were sacrificed by decapitation under deep isoflurane anesthesia. Their brains were carefully removed and selected CNS structures (frontal cortex, prefrontal cortex, thalamus, hypothalamus, nucleus accumbens, caudate-putamen, rostral ventromedial medulla (RVM), periaqueductal gray matter (PAG) and amygdala) were dissected on ice using the Alto Stainless Steel Coronal 0.5 mm Brain Matrix. The tissue was immediately frozen on dry ice following dissection and stored at −80 • C until use. After thawing, tissues were homoge-

[35. S]GTPγS Binding Assay
The assay was carried out as previously described [40]. Homogenates (15µg/mL) were incubated in triplicate with 0.8 nM [ 35  potency (EC 50 ) of DAMGO stimulation in various brain structures were calculated and expressed as means ± SEM from 6 animals.
A scheme representing the study design is included below (Figure 8).

Statistical Analysis
Behavioral data from chronic cotreatment experiments along with data involving NTX-reversal of acute MRF and DSF effects, were reported as mean control-subtracted values ± SEM. Other acute treatment data were shown along with respective controls and expressed as mean ± SEM. Results from time-course curve experiments were analyzed with two-way ANOVA, followed by Bonferroni's post hoc test. Area under the time-course curve (AUC) data were analyzed with either the nonparametric Kruskal-Wallis test or one-way ANOVA followed by Dunnet's post hoc test. The nature of MRF + DSF combination index (CI) was calculated at peak effect corresponding to 60 min postadministration according to the Bliss Independence model [74] (2): where E A -fractional effect of drug A; E B -fractional effect of drug B; E AB -fractional effect of drug combination. A CI < 1 indicates synergy, CI = 1 an additive effect and CI > 1 antagonism. Binding data were fitted with a one-site nonlinear regression model provided by GraphPad Prism, version 5.03 software for Windows (GraphPad Software, San Diego, CA, USA, www.graphpad.com) (accessed on 20 June 2019). Efficacy (E max ) and potency (EC 50 ) were calculated using Equation where Y-% stimulation, B-% stimulation at bottom plateau, T-% stimulation at top plateau, X-ligand concentration in log units, logEC 50 -ligand concentration at halfmaximal stimulation. E max and pEC 50 values were expressed as means ± SEM from averaged data collected from 6 rats. Differences in potency and efficacy of DAMGO in different CNS structures in MRF + DSF, MRF and control rats at day 1 and 14 were evaluated with one-way ANOVA. As G-protein stimulation efficacy was comparable in both MRF and MRF + DSF controls (water for injection and 0.1% methylcellulose), the results were combined.

Conclusions
Our study is the first to describe the possible mechanism contributing to the maintenance of morphine antinociception by low-dose DSF. We proposed the importance of the partial rescue of supraspinal µ-opioid G-protein signaling by DSF in the striatum. Of note, the mechanism presented in the current study constitutes only a small piece of the puzzle and other yet unknown mechanisms are certainly involved. It should be emphasized that DSF affects a plethora of physiological pathways. Nevertheless, the presented research is a first but significant step towards considering low-dose DSF as a possible adjunct agent for opioid pain management.