Sex‐specific role of the circadian transcription factor NPAS2 in opioid tolerance, withdrawal and analgesia

Abstract Opioids like fentanyl remain the mainstay treatment for chronic pain. Unfortunately, opioid's high dependence liability has led to the current opioid crisis, in part, because of side‐effects that develop during long‐term use, including analgesic tolerance and physical dependence. Both tolerance and dependence to opioids may lead to escalation of required doses to achieve previous therapeutic efficacy. Additionally, altered sleep and circadian rhythms are common in people on opioid therapy. Opioids impact sleep and circadian rhythms, while disruptions to sleep and circadian rhythms likely mediate the effects of opioids. However, the mechanisms underlying these bidirectional relationships between circadian rhythms and opioids remain largely unknown. The circadian protein, neuronal PAS domain protein 2 (NPAS2), regulates circadian‐dependent gene transcription in structure of the central nervous system that modulate opioids and pain. Here, male and female wild‐type and NPAS2‐deficient (NPAS2−/−) mice were used to investigate the role of NPAS2 in fentanyl analgesia, tolerance, hyperalgesia and physical dependence. Overall, thermal pain thresholds, acute analgesia and tolerance to a fixed dose of fentanyl were largely similar between wild‐type and NPAS2−/− mice. However, female NPAS2−/− exhibited augmented analgesic tolerance and significantly more behavioral symptoms of physical dependence to fentanyl. Only male NPAS2−/− mice had increased fentanyl‐induced hypersensitivity, when compared with wild‐type males. Together, our findings suggest sex‐specific effects of NPAS2 signaling in the regulation of fentanyl‐induced tolerance, hyperalgesia and dependence.

Hallmarks of opioid dependence and opioid use disorder are profound disruptions to sleep and circadian rhythms that persist during abstinence and withdrawal. [4][5][6][7] In line with this, previous evidence suggests opioid tolerance, dependence and analgesia [8][9][10][11][12] are modulated by sleep and circadian rhythms. Altered circadian rhythms and sleep disruptions are commonly observed in patients being treated with opioids. 13 For example, sleep disturbances are a known indicator of opioid withdrawal symptoms in patients attempting to discontinue their opioid treatment. 14 Sleep and circadian rhythm disruptions may also contribute to relapse risk in opioid abstinent patients. 15,16 However, our understanding of the molecular pathways that underlie the relationship between circadian rhythms and opioid tolerance and withdrawal is limited.
Nearly every cell in the body expresses the necessary machinery that controls circadian rhythms at the molecular level. 17 The molecular clock is comprised of a series of interacting transcriptional and translational feedback loops. Transcription is driven by heterodimers of CLOCK (Circadian Locomotor Output Cycles Kaput Protein), or NPAS2 (neuronal PAS domain protein 2), with BMAL1 (brain muscle aryl nuclear translocase like-1), that bind to enhancer elements in gene promoters. These heterodimers of circadian transcription factors (CLOCK or NPAS2 and BMAL1) drive the transcription of circadian genes, Cry1,2 (Cryptochrome) and Per1,2 (Period). Accumulation of CRYs and PERs in the cytoplasm over 24-h initiate the formation of dimers that eventually translocate to the nucleus to inhibit their own transcription via interactions with BMAL1 heterodimers. 18 Molecular clocks are critical to the function of most peripheral organs and many regions in the brain. 19 In peripheral tissues and the central nervous system, the molecular clock modulates the circadian expression of endogenous opioid peptides and opioid receptors. [20][21][22][23] Opioids also impact the expression of circadian genes in the brain. For example, acute and chronic administration of opioids alters circadian genes in the hypothalamus, [24][25][26] while withdrawal from opioids alters the rhythmic expression of circadian genes in the midbrain and striatum. [26][27][28] Recent findings show circadian genes, including Per1 and Per2, regulate opioid reward, 29 tolerance and withdrawal. 30 In a tissue-dependent manner, Per1 and Per2 transcription are driven by BMAL1 dimerization with CLOCK or NPAS2. [31][32][33] Notably, the circadian transcription factor NPAS2 seems to be predominantly expressed in the central nervous system 34 and enriched in major neural substrates of opioid-induced tolerance, 35 dependence, [36][37][38] and hyperalgesia, 39 including the spinal cord, 34 primary sensory neurons, 40 and within the subregions of the striatum, including the nucleus accumbens (NAc). 41 However, to our knowledge, the involvement of NPAS2 signaling in the emergence of opioid sideeffects has not yet been studied.
In the present study, we investigated the role of NPAS2 in opioid tolerance, dependence and analgesia, using the synthetic opioid, fentanyl. Fentanyl is a widely prescribed opioid with high potential for dependence, found in most drug overdose deaths in the United States. 3 Thus, further investigation into the potential pathways mediating the development of fentanyl-induced tolerance and dependence is imperative, as new therapeutic approaches are designed to improve opioid treatments and mitigate secondary effects of opioids. To explore the potential relationship between NPAS2 and opioids, we assayed behavioral phenotypes of fentanyl-induced tolerance, dependence and hyperalgesia in NPAS2 deficient male and female mice (NPAS2À/À). These transgenic mice carry a LacZ reporter that replaces exon 2 of the Npas2 gene and effectively deletes the basic Helix-loop-Helix (bHLH) domain required for NPAS2 to directly bind DNA. The lack of the bHLH domain impedes NPAS2-dependent transcription. 32 Therefore, NPAS2-deficient (NPAS2À/À) mice retain the expression of the NPAS2 protein, while lacking the ability to bind DNA and drive transcription dependent on NPAS2. Overall, our findings reveal sex-specific effects of NPAS2 deficiency on thermal nociceptive thresholds, acute anti-nociception and development of tolerance to fentanyl, as well as behavioral markers of physical dependence and pain hypersensitivity.

| General procedures and experimental timeline
Throughout the experiments, experimenters were blind to both the treatment and genotype groups. Prior to each behavioral assay, mice were habituated to experimental rooms and gently handled by experimenters for at least 1 h per day for three consecutive days. We used 9-11 mice per genotype per sex. The experimental timeline is described in Figure 1 Testing of TFL was performed before and 15 min after the ZT2 administration to test development of hyperalgesia and tolerance, respectively. On day 11, a second dose-response was conducted using the following fentanyl doses: 160, 320, 640 and 1280 μg/kg. Finally, on day 12, mice underwent naloxone-precipitated withdrawal, as described below.

| Tail-immersion test for fentanyl-induced analgesia and tolerance
The tail-immersion behavioral assay was used to measure TFL, and assess thresholds of thermal nociception, analgesia and tolerance to fentanyl. A container filled with water maintained at 50 ± 0.5 C using a thermoregulated water bath. Part of the mouse's tail was immersed in the water and the TFL was recorded with a maximum immersion limit of 10 s to avoid tissue damage. After testing, mice were immediately returned to their home cage. TFL measurements were repeated 2-3 times per mouse at 2-min intervals.

| Statistical analyses
A combination of statistical analyses was used for behavioral assays.
Tolerance measures were analyzed using two-way analysis of variance (ANOVA) (Time and Treatment) followed by Tukey's multiple comparison tests. Thermal nociceptive thresholds, EC50s and withdrawal behaviors were analyzed using one-way ANOVA, followed by Tukey's post-hoc multiple comparison tests, when appropriate. Doseresponse curves were generated using a non-linear regression of logtransformed values compared with normalized values from 0 to 100.
Rightward shifts in the dose-response curves between mouse genotypes were compared by calculating the ratio between EC50 obtained on day 11 versus EC50 obtained on day 1. Unpaired t test comparisons were performed to determine difference of rightward shift between NPAS2À/À and WT mice of the same sex. Data were analyzed using GraphPad 9.0 and considered statistically significant if p ≤ 0.05.

| RESULTS
3.1 | Impact of NPAS2 deficiency on thermal thresholds, fentanyl acute analgesia, fentanylmediated tolerance and hypersensitivity The impact of NPAS2 deficiency on thermal thresholds and development of tolerance and hypersensitivity with chronic fentanyl was F I G U R E 1 Experimental timeline. All animals underwent the same procedures. On day 0, animals were tested for thermal nociceptive thresholds (baselines) with the tail flick (TFL) assay. On day 1, a first dose-response of the fentanyl analgesic effect on TFL was performed. The 5-day fentanyl administration regimen involving two fentanyl i.p. injections per day was conducted from day 4 to day 8. Tolerance and OIH development were evaluated simultaneously. On day 11, a second fentanyl dose-response was conducted. This series of experiments concluded with the naloxone-precipitated withdrawal assay on day 12 evaluated by measuring TFLs in NPAS2À/À and WT mice before and

| Impact of NPAS2 deficiency on sex-specific changes in fentanyl potency
To further test the impact of NPAS2 deficiency on fentanyl analgesic potency, we performed two fentanyl dose-response procedures on male and female NPAS2À/À and WT mice. The first dose-response was conducted on day 1 when mice were naïve to opioids ( Figure 3A), and the second dose-response was conducted on day 11, after mice had received the tolerance-inducing 5-day fentanyl administration regimen ( Figure 3B). Dose-responses performed on day 11, revealed a rightward shift of the fentanyl dose-response curves in all groups, as compared with dose-responses measured on day 1 (Males, Figure 3A, Figure 3B). Therefore, our regimen induced a robust state of analgesic tolerance in all mice. Based on best-fit values of these curves, we calculated effective concentration 50 (EC50) values, which represent the concentration of fentanyl that gives half-maximal anal-

| DISCUSSION
Long-term opioid use for the treatment of chronic pain is hampered by analgesic tolerance, 42 physical dependence, 43 and disruptions of sleep and circadian rhythms. [4][5][6][7] Previous studies implicated circadian regulation of pain, tolerance and physical dependence, 11,12 suggesting bidirectional interactions between circadian rhythms and opioids. 9 NPAS2 is a circadian gene enriched in spinal cord and brain regions involved in pain and opioids, and previously shown to be involved in psychostimulant reward. 44  Our results are consistent with prior studies which also evaluated the impact of genetic deletion of circadian genes on pain and opioids. 29,30 In these studies, mice with global deletion of mPer1 (mPer1-KO) 29 or mPer2 (mPer2-KO) 30 genes showed no change in thermal pain thresholds and acute analgesia. Similarly, NPAS2À/À mice exhibited no measurable difference in thermal pain thresholds and acute analgesia, suggesting these circadian genes may not be directly involved in these behaviors. Conversely, mPer1 and mPer2 clock genes were differentially involved in tolerance and dependence, as mPer2-KO promoted morphine tolerance and mitigated physical dependence, 30 while mPer1-KO did not affect tolerance or physical dependence. 29 In our study, which included both males and females, NPAS2À/À mice did show altered tolerance development to a fixed dose of fentanyl across both sexes. In addition, we also evaluated fentanyl potency, by conducting dose-response tests before and after tolerance development. Strikingly, NPAS2À/À mice developed higher analgesic tolerance compared with controls, an effect only observed in females. These behaviors are thought to be mediated by different mu-opioid receptor (MOR) signaling mechanisms, possibly in a cell type-specific and/or region-specific manner. 46 Importantly, tolerance can develop to all MOR-mediated behaviors, although at different rates. 1,47 The apparent dichotomy between the decreased analgesic potency of fentanyl while withdrawal behaviors were increased in NPAS2À/À females, could be explained by either a distinct involvement of NPAS2 in analgesia and physical dependence, or perhaps pronounced differential development of tolerance to analgesia and to physical dependence in these mice. Nevertheless, these data could suggest that circadian genes are differentially involved in tolerance, with mPer2 being involved in its development, Npas2 in its expression and mPer1 likely not involved. However, caveats of studies investigating the roles of Per genes include the absence of testing opioid potency in tolerant animals with a dose-response assay, along with these studies including only male mice. Because Npas2 and Per genes are both expressed in structures involved in tolerance such as the spinal cord 34,48 or the NAc, 49 and NPAS2 can directly regulate the transcription of Per genes, involvement of each of these circadian genes in tolerance remains a possibility.
Sex specific effects of NPAS2 deficiency were also observed with physical dependence symptoms. NPAS2À/À females showed more withdrawal behaviors than controls, also more than NPAS2À/À males. Conversely, when testing fentanyl-induced hypersensitivity, a symptom also emerging during opioid withdrawal, NPAS2À/À females were like WT females, while NPAS2À/À males had more pronounced hypersensitivity than WT males. Together, these results suggest that tolerance, physical dependence and opioid-induced hypersensitivity behaviors could be modulated by NPAS2 signaling in a sex specific manner.
Involvement of NPAS2 in these behaviors could be supported by the fact that NPAS2 expression is enriched in the NAc. 32,41 The NAc is involved in tolerance, 49,50 physical dependence, 37,51,52 and opioidinduced hyperalgesia. 12 NPAS2 modulates dopaminergic and glutamatergic neurotransmission in the striatum, 53 both of which are altered during opioid tolerance 49,50 and withdrawal. [54][55][56] Interestingly, changes in the expression of circadian genes in the NAc were shown to occur in rodents with opioid-induced hyperalgesia during a state of withdrawal. 12 Thus, together with our current findings, NPAS2 may modulate opioid-related behaviors and involve dopaminergic and glutamatergic signaling in the NAc. In addition, NPAS2 may also play a role in peripheral tissues 57,58 and NPAS2 is expressed in primary sensory neurons, 40 which are known to be essential in opioid tolerance and hyperalgesia behaviors. 39 Ongoing and future studies are exploring possible involvement of NPAS2 in the NAc dopaminergic and glutamatergic neurotransmission in opioid tolerance and dependence.
Our results illustrate the importance of examining interactions between sex, opioid behaviors and circadian genes. This is also supported by several previous studies, which examined whether sex could have a differential impact on opioid behaviors. Analgesic effect of opioids was shown to be variable depending on sex, with a higher and longer-lasting effect in male than female rodents, [59][60][61] although, other studies did not find a sex difference in opioid analgesia. 62,63 In addition, sexual dimorphism in opioid tolerance has not been extensively studied in rodents, yet the studies that have, report higher tolerance in males than females, 64 with tolerance developing faster in females than males. 65,66 However, these findings have not been supported by other studies. [67][68][69] Finally, sexual dimorphism in opioid dependence-mediated withdrawal behaviors were also reported in rodent studies, with overall more dependence in WT males than WT female rats. 64,70 Inconsistent findings between men and women have also been reported in humans. [71][72][73][74] In our current study, we did not observe sex-related differences in opioid analgesia, tolerance development and expression, dependence and hyperalgesia, between WT male and female mice. Collectively, our data illustrates the lack of consensus on the impact of sex on opioid behaviors, and thus requires further investigation.
Nevertheless, in our study, we specifically examined intersectional consequences of NPAS2 deficiency and sex on opioid analgesia, tolerance and dependence behaviors. Interestingly, NPAS2À/À females developed markedly more physical dependence behaviors and marginally more profound tolerance than female WT littermates, which was not observed in males. However, NPAS2 deficiency had no consequences on hyperalgesia development in females while it exacerbated that symptom in males. Thus, our data indicates that sex differences in our study are related to interacting effects between NPAS2 deficiency and sex. This could be explained by the fact that sex differences are also known to exist in circadian rhythms 75,76 and in circadian genes rhythmicity between males and females in brains of humans 77 and of rodents. 78 Our results are also consistent with prior studies which examined intersectional consequences between sex and circadian genes such as Clock or Npas2. 44,79 Interestingly, Npas2 deletion had higher impact on cocaine reward and self-administration behaviors in female mice. 44 More profound consequences on females than males could be explained by levels of circulating hormones, as sex differences in cocaine self-administration were abolished in ovariectomized females. 44 This was consistent with the fact that circulating estrogens were shown to be essential in orchestrating rhythms of circadian genes in the SCN. 80 In addition, estrogen signaling has been shown to influence opioid tolerance and dependence behaviors. 81 Thus, female circulating hormones could also be involved in the sexually dimorphic consequences of NPAS2 deficiency on opioid tolerance and dependence. Further studies examining the mechanisms of interaction between NPAS2, opioids sexual hormones are now warranted.
Overall, our study and prior studies examining interactions between circadian genes and opioid-related behaviors, show a differential involvement of Npas2 and Per genes. 29,30 However, Npas2 and Per genes bi-directionally regulate their expression levels, leading to variations in expression that follow circadian rhythmicity. 33 Importantly, rhythmic expression of these circadian genes follow different circadian phases. 82,83 Therefore, suggesting that involvement in behavior of these genes could vary at different times of day. In our study, we examined the role of NPAS2 deficiency at a similar time of day (ZT2) as the mPer1-KO and mPer2-KO studies (ZT3-5). 29,30 Whether these mutations could impact opioid-related behaviors via alterations of the circadian clock remains unknown. 31,32,45 However, other behaviors involving NAc circuitry, like feeding, and food or cocaine self-administration, were shown to be differentially altered at varying times of day in NPAS2À/À mice. 44,45 Therefore, ongoing studies are now also evaluating whether impact of NPAS2-defficiency on opioid tolerance and physical dependence vary at different times of day.
While prior studies used morphine to evaluate the impact of mPer1 and mPer2 deletion, we used fentanyl to evaluate impact of NPAS2 deficiency. Although both opioids modulate analgesia via MORs, 84,85 they can activate different MOR downstream signaling pathways. [86][87][88] Therefore, fentanyl and morphine may recruit different circadian genes and downstream signaling pathways involving MOR.
In conclusion, our study provides evidence for a differential role of NPAS2 signaling in fentanyl mediated behaviors, with high impact on physical dependence and marginal effect on tolerance. Importantly, NPAS2 deficiency modulated these behaviors in a sexually dimorphic manner, with female mice more profoundly affected than males. Identification of NPAS2-controlled genes and signaling pathways that modulate opioid behaviors and that may interact with substrates of opioid tolerance and dependence, could provide better insight on understanding the impact of clock genes and circadian rhythmicity on the chronic use of prescription opioids in patients.

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