Chronic pain has both sensory and emotional/affective components that, while distinct, share many of the same neurobiological substrates. Chronic pain often induces a persistent negative affective state, or “emotional pain,” likely as a result of neuroadaptations in the brain’s reward processing circuitry [39–41]. Chronic pain is also associated with the development of other disorders of impaired reward processing, including depression and substance use disorders [42]. The prevalence of suicidal ideation among pain patients suggests that it may be an expression of emotional pain. In fact, studies have found emotional pain to be the psychological variable most strongly associated with current suicidality, even more so than the presence of depressed mood or hopelessness [18, 43, 44].

The relationship between AUD and suicidality also suggests that suicidal ideation may be an expression of emotional pain. That AUD is often a key risk factor for suicidal behavior is likely due, at least in part, to alcohol-induced exacerbations in negative emotionality and alcohol-related negative consequences, particularly in interpersonal domains [45]. Hyper-negative emotional states and hyperalgesia are both consequences of repeated alcohol use [46] and may contribute to an increase in alcohol use as a compensatory mechanism [47]. Among heavy drinkers, the negative emotional components of pain (i.e., pain catastrophizing) can enhance alcohol craving more than physical pain [48].

The emotional pain caused by the disruption or loss of social attachments also plays an important role in suicidality. In particular, suicidal ideation shares neurobiological and psychological features with separation distress—the innate, emotionally painful, dysphoric response of animals and humans to social separation or rejection [49, 50]. Converging evidence from preclinical and clinical studies supports the link between suicidal ideation and separation distress. Opioids reduce separation distress behaviors in non-human mammals [51–53] and have recently shown promise for the treatment of suicidality [54, 55]. Suicidal acts are most common after interpersonal losses or rejections [56], and patients with borderline personality disorder are particularly susceptible to social rejection and often become suicidal after interpersonal rejections [57, 58].

Research in the field of affective neuroscience suggests that separation distress represents one of the ancestral primary-process emotional systems (referred to as PANIC/GRIEF) [49, 50, 59, 60]. Importantly, the PANIC/GRIEF system probably evolved from general pain mechanisms [59], and studies have found a link between separation distress and physical pain. Maternal separation in mouse neonates is capable of altering nociceptive behavior in adulthood [61, 62]. In humans, early life adversity in the form of both physical and psychological trauma (including familial separation) is associated with an increased risk of chronic pain in later life [63]. Studies show that parental bonding in adolescents is significantly associated with adolescent chronic pain and depression, where low maternal care contributes to increased pain via heightened depressive symptoms [64].

The neuroanatomy of the PANIC/GRIEF system overlaps with the brain’s system for processing physical pain [65], suggesting that both physically painful (e.g., injury) and emotionally painful (e.g., interpersonal rejection) stimuli may engage this shared neurocircuitry to produce distress and dysphoria (emotional pain), with separation distress representing a particular subtype of emotional pain. The PANIC/GRIEF system also represents a potential substrate linking emotional pain/distress with suicidal ideation. The similarities between suicidal ideation and separation distress discussed above suggest that suicidal ideation may result from increased activity in the brain’s PANIC/GRIEF network, whether initiated by social loss or physical injury. Thus, chronic pain patients may be particularly susceptible to suicidal ideation due to persistent activation of this system by painful sensory stimuli.

Sustained activation of the PANIC/GRIEF system can also lead to reduced activity in the brain reward SEEKING system (the system primarily responsible for motivation and arousal, particularly the mesolimbic dopamine pathway), perhaps as a means to protect against sustained emotional pain by reducing the overall arousal of emotions [49, 50, 59, 66]. Underactivity in the SEEKING system causes blunted reward processing, leading to an amotivational state characterized by the diminished experience of positive feelings (anhedonia). Thus, down-regulation of the SEEKING system in response to sustained emotional pain (dysphoria) may be one underlying cause of the anhedonia and blunted affect that is characteristic of depression.

This two-stage model then suggests that the anhedonia of depression may reflect, in part, an emotional shutdown that follows the behavioral agitation of separation distress or other emotional pain, with these behavioral states correlating with underactivity of the SEEKING network and overactivity of the separation distress PANIC/GRIEF network, respectively. Of course, depression is a complex disorder that likely has multiple etiologies involving changes in many neural substrates. However, this model suggests one possible framework for understanding the relationship between suicidal ideation and depression. Sustained activation of the PANIC/GRIEF network may produce both a state of dysphoria that leads to suicidal ideation as well as, indirectly, a reduction in motivation (anhedonia) through subsequent downregulation of the reward SEEKING system (Figure 1). In this way, suicidal ideation would often be comorbid with depression but could be the result of dysregulation of the PANIC/GRIEF circuitry independent of a disruption in reward processing. In support of this hypothesis, patients with borderline personality disorder, but not comorbid depression, typically experience brief durations of suicidality as a result of interpersonal stresses. In contrast, patients with comorbid borderline personality disorder and depression display more persistent symptoms of depression and suicidality, which also include a loss of interest in ordinarily pleasurable activities (anhedonia) [18].

Ultimately, painful experiences (whether the emotional pain of social rejection or the physical/sensory pain characteristic of chronic pain conditions) likely engage shared brain systems which produce negative affective states (distress/dysphoria–implicated in suicidal ideation) and may subsequently disrupt downstream reward processing (anhedonia–implicated in depression), leading to a complex relationship between pain, separation/social distress, and motivation [65, 67, 68]. The neurocircuitry and neurochemistry underlying each is discussed in more detail below.

Much of the same neurocircuitry described above is also engaged by separation distress, suggesting that the emotional pain elicited by separation distress is a result of the activation of circuits underlying physical pain. Specifically, the separation distress circuitry starts in the PAG and ascends through the dorsomedial thalamus, terminating in various basal forebrain regions including the anterior cingulate cortex [50, 59, 65] (Figure 2). Localized electrical stimulation of the anterior cingulate, dorsomedial thalamus, and PAG provoke separation cries in mammals [83, 84]. In humans, the experience of emotional distress induced by social rejection is associated with an increase in activity in the anterior cingulate cortex measured by fMRI [85]. Together, these studies suggest that psychological pain, particularly social rejection and intense loneliness, may share some of the same neural pathways that elaborate physical pain.

Reward and motivation, while often considered opponent processes to pain, are also processed within many of the same brain structures, in particular the mesolimbic system, which includes the VTA and nucleus accumbens (Figure 2). In neuropathic pain animals, functional connectivity is altered within the limbic system (including the nucleus accumbens) as well as between the limbic and nociceptive systems (including the thalamus, primary sensory cortices, insula, and PAG) [86]. In human clinical pain cohorts, connectivity is altered between the mesolimbic system and cortical structures [87–89]. Given that the mesolimbic system is responsible for the modulation of motivated behaviors and reinforcement learning [90, 91], altered activity in this system likely contributes to the negative affective component of pain. Pain can also directly impair general reward processing, leading to an anhedonic state [42]. Dopamine is a critical neurotransmitter within the mesolimbic system, and dopamine signaling in the nucleus accumbens may modulate the salience of painful experiences. While acute pain activates dopaminergic transmission to the nucleus accumbens [81, 92], chronic pain produces the opposite effect [93–95].

Together, the circuitry underlying pain, separation distress, and motivation provides a neuroanatomical substrate for the transition from physical or emotional pain to suicidality and possibly depression. Activity in pain/nociceptive circuitry (including the anterior cingulate, dorsomedial thalamus, and PAG) is correlated with the distress PANIC/GRIEF system, while activity in the mesolimbic circuit, particularly dopamine transmission in the nucleus accumbens, is correlated with the SEEKING system [50, 60, 66]. Sustained activity in the pain circuitry can reduce activity in the motivational circuitry through modulation of mesolimbic dopamine signaling, such that the distress caused by prolonged pain (sensory or emotional) may ultimately lead to decreased processing of rewarding stimuli. These neural substrates are further linked through shared activation by the opioid system, discussed in more detail below.

Opioids and their receptors can modulate both the sensory and emotional (bothersome) components of pain. For example, when injected into the dorsomedial thalamus (a key region for processing both the sensory and affective components of pain), the MOR agonist DAGO elevated rats’ sensory pain thresholds and induced a positive affective state, while the KOR agonist U50,488 reduced rats’ pain thresholds and induced a negative affective state [111]. However, studies have shown sex differences with respect to the functioning of the MOR and KOR systems, particularly with respect to pain and addiction. Several comprehensive reviews are available [112–114].

In addition to their role in the sensory/affective components of pain, it is hypothesized that opioids constitute a major neurochemical underpinning of social bonding and isolation distress [115, 116]. For example, opioid peptides are decreased in the midbrain of rat pups following social isolation [117], and endogenous opioid peptides acting at MORs have been shown in animal models to alleviate distress behaviors following social separation [115]. The MOR agonist morphine decreases distress vocalizations in rat pups isolated from their mother [118, 119], while the KOR agonist U50,488 increases isolation-induced ultrasonic vocalization [118]. Thus, it appears that dynorphins are responsible for mediating negative affect within the neurocircuitry underlying social distress, similar to their role in the modulation of physical pain (Figure 2).

High expression levels of KOR have been also detected in brain areas responsible for reward and motivation, including the VTA and nucleus accumbens [120, 121] (Figure 2). The ability of KOR agonists to negatively modulate mesolimbic dopamine has significant implications for motivated behavior. Since reduced nucleus accumbens dopamine signaling is associated with a loss of motivation, KOR modulation of dopamine circuitry may link the acute distress of physical or emotional pain with the subsequent onset of negative motivational states and affect.

Studies in rat brain slices show that KOR agonists (U69,593) are capable of suppressing mesolimbic dopamine release via receptors expressed on dopamine neuronal terminals as well as neuronal cell bodies [122–125], which may contribute to the dysphoric effects of KOR activation. KOR agonists (including U50,488, spiradoline, U69,593) also inhibit dopamine signaling when applied directly into the nucleus accumbens of rats, as measured by microdialysis in intact animals [126, 127] or by superfused brain slices [128]; however, changes in dopamine signaling in the nucleus accumbens did not correlate with KOR agonist (U50,488) induced conditioned place aversion (CPA) in mice [129]. Furthermore, morphine-evoked increases in extracellular dopamine within the nucleus accumbens were blocked by the administration of a KOR agonist (U50,488) into this brain region in mice [130]. Finally, the expression of KOR-mediated aversion (U69,593-induced CPA) requires the activity of medium spiny neurons expressing dopamine receptors within the nucleus accumbens of rats [131] and mice [132].

Activation of KORs also contributes to the dopamine hypofunction observed in chronic pain states. Given that hypo-dopaminergic states contribute to chronic pain [133, 134] and mood disorders comorbid with chronic pain [135], KOR antagonism to recover dopamine may hold promise as a novel therapeutic for treating chronic pain and associated mood disorders. Indeed, reduced motivation for food (sucrose) reward induced by inflammatory pain was recovered by KOR antagonism (with norBNI) or silencing of dynorphin neurons within the ventral striatum of rats [136]. A comprehensive review of KOR function in chronic pain and its relationship with drug-seeking behavior is available [137].

The role of KORs in dysfunctions of reward and motivation is particularly well-characterized with respect to AUD. KOR-mediated reductions in dopamine release in the nucleus accumbens have been hypothesized to mediate negative emotional states associated with alcohol withdrawal, particularly pain associated with acute alcohol withdrawal [138]. At the preclinical level, alcohol-preferring rats show increased dynorphin mRNA expression in the central amygdala and hypothalamus compared to non-preferring rats after voluntary drinking [139]. The KOR antagonist norBNI also attenuates withdrawal-related anxiety-like behaviors in alcohol-dependent mice [140]. In addition to alleviating alcohol-induced negative affective states, pharmacotherapies with KOR antagonist properties, including buprenorphine, reduce binge-like alcohol drinking [141, 142], alcohol self-administration [143, 144], and block escalation of compulsive-like drinking after dependence induction [145, 146] and stress exposure [147] in rodent.

The ability of opioids to modulate both sensory/affective pain circuits as well as reward/motivation circuits suggests these compounds have the capability to counteract multiple features of suicidality and depression. Opioids are able to produce dopamine-independent positive affective states through their influence on pain and social distress circuits, including areas such as the anterior cingulate cortex, PAG, and dorsomedial thalamus. Opioids can also counteract negative affective states by promoting increased motivational drive through their downstream influence on dopaminergic reward circuits, especially in the nucleus accumbens. Thus, since opioids can restore deficits in both pain and reward circuits, they may be particularly useful in treating suicidal ideation, as they would be able to blunt suicidality whether it was caused primarily by a disruption in dopaminergic motivational circuity or a disruption in non-dopaminergic pain/social distress circuits. Other treatments which primarily target reduced motivation (anhedonia) may not be effective in treating the subset of individuals experiencing suicidal ideation without a concomitant reduction in motivation (i.e., patients with or without comorbid depression).

Given the above understanding of the opioid system, there has been a renewed interest in the use of opioid analgesics, particularly buprenorphine, for the treatment of suicidality. A retrospective 24-months study reported buprenorphine had the lowest incidence of suicide intent and deaths compared to other opioid analgesics [148]. Buprenorphine has demonstrated rapid antidepressant effects in humans [149–152], including those with treatment-resistant depression, and has shown particular promise in reducing suicidal ideation [153].

Buprenorphine is an analgesic derived from oripavine that acts as a partial MOR agonist and KOR antagonist, as well as an antagonist at the DOR and an agonist at the NOR [154]. Buprenorphine has similar affinities for the MOR and KOR, but a 10-fold lower affinity for the DOR [154]. For the purposes of this review, we have largely limited our discussion to the MOR agonist and KOR antagonist effects as these are well-characterized, although note that some work has shown that buprenorphine may have partial agonist activity at KORs [155–157]. Importantly, due to its partial MOR agonist properties, buprenorphine has a lower overdose risk compared to full MOR agonists such as morphine [158].

The potential anti-suicidal effects of buprenorphine were first described in a case report of a 61-year-old woman suffering from treatment-resistant depression, chronic back pain, severe opioid use disorder, and ongoing suicidal ideation [159]. While treatment with buprenorphine (16 mg/4 mg buprenorphine/naloxone) was prescribed to treat the patient’s opioid use disorder, she reported that her suicidal ideation completely disappeared after 1 week of treatment, and suicidal ideation remained absent up to 3 months after the initial treatment. In another case report, a patient with cannabis-induced psychotic disorder and opioid depressive disorder with severe suicidal thoughts was treated successfully with a single high dose (96 mg) of buprenorphine [160]. Another case report showed that buprenorphine/naloxone (8 mg/2 mg) was effective in reducing pain and suicidal ideation in a 39-year-old male with a history of bipolar disorder, multiple suicide attempts, and polysubstance abuse [161]. Chart reviews of suicidal adult depressed patients with comorbid chronic pain and opioid use disorder who received off-label buprenorphine also found some support for the anti-suicidal properties of buprenorphine [162]. The presence of chronic pain, depression, and substance use in these case reports and studies suggests that buprenorphine may effectively mitigate suicidal ideation by targeting multiple overlapping neurocircuits, which underlie these often-comorbid disorders.

Based on these case reports and anecdotal findings, clinical trials have begun to explore the anti-suicidal potential of buprenorphine. A multisite randomized double-blind placebo-controlled trial of ultra-low-dose (initial dosage, 0.1 mg once or twice daily; mean final dosage, 0.44 mg/day) buprenorphine found that severely suicidal patients showed a reduction in Beck Scale for Suicidal Ideation (BSSI) scores after 2 and 4 weeks of treatment compared to patients that received placebo [54]. Another randomized clinical trial tested the efficacy of one of three single high doses of buprenorphine (32, 64, 96 mg) in suicidal opioid-dependent patients [163]. The researchers found that BSSI scores were significantly reduced in patients across all three buprenorphine doses. While these findings suggest that buprenorphine has particular promise as an anti-suicide treatment option, more research is needed to determine the conditions under which buprenorphine treatment is most effective and tolerable, including whether low or high doses are more successful (the wide range of effective doses in these clinical trials - from 0.44 to 96 mg - is interesting and will be an important area for future studies) and whether treatment is affected by comorbidities including substance use disorder or chronic pain.

Other studies have examined the effects of buprenorphine in combination with other drugs. A randomized double-blind placebo-controlled trial in adults with treatment-resistant depression showed that treatment with 2 mg/2 mg buprenorphine/samidorphan significantly improved scores on multiple depression measures compared to placebo controls [164]. Antidepressant activity was also demonstrated in another study using a 1:1 ratio of buprenorphine:samidorphan [165]. Samidorphan is a MOR antagonist, thus this combined treatment leaves the KOR antagonist activity of buprenorphine intact while blocking the MOR agonist activity of buprenorphine. Blocking the subjective and objective MOR effects likely mitigates the potential addictive properties of buprenorphine in opioid-naïve individuals. While these studies did not directly measure suicidal ideation independent of depression, they provide important insight into the potential mechanisms of buprenorphine’s behavioral effects. In both cases, the effects of buprenorphine were not blocked by a MOR antagonist, suggesting that activity at this receptor may not be required for the anti-suicidal effects of buprenorphine.

In addition to the clinical findings described above, pre-clinical studies also support the potential anti-suicidal treatment effects of buprenorphine. While no behavioral assays for suicidal ideation exist for rodents, several behavioral assays have been validated to screen for depressive-like behaviors, including the forced swim test and the novelty-induced hypophagia test. The forced swim test is considered a measurement of behavioral despair [166] and is one of the gold-standard screens for depressive behavior as it is reliably reversed by antidepressants [167]. The novelty-induced hypophagia test is a conflict-based behavioral task that assesses the impact of an environmental stressor on animals’ conditioned approach toward a palatable food reward, with longer approach latencies indicating greater depressive or anxious behavior [168]. Treatment with antidepressants or benzodiazepines reduces approach latencies in the novelty-induced hypophagia test [168–170].

In mice, administration of buprenorphine produced significant reductions in forced swim test immobility (at doses ranging from 0.065–2 mg/kg) and reduced approach latencies in the novel environment of the novelty-induced hypophagia test (at a dose of 0.25 mg/kg) [171]. In Wistar Kyoto (WKY) rats, a strain which has an exaggerated depressive phenotype and is resistant to certain antidepressants, buprenorphine (2.25 mg/kg) significantly reduced immobility in the forced swim test [172]. This effect was specific to the WKY strain, suggesting that buprenorphine may be more effective in individuals with certain treatment-resistant depressions. In general, the doses which produced behavioral effects in these preclinical studies were lower than those used in clinical trials; however, one trial showed that an ultra-low dose of buprenorphine (0.44 mg/day, comparable to the preclinical doses) significantly reduced suicidal ideation [54].

A better understanding of the specific pharmacological underpinnings of the anti-suicidal and anti-depressive effects of buprenorphine would be valuable as it could inform the development of even more targeted therapeutics which might avoid side effects, including abuse potential. As discussed previously, buprenorphine is both a partial agonist at the MOR and an antagonist at the KOR, but which receptor type is primarily responsible for the anti-suicidal and anti-depressive effects of the drug is uncertain. A review of preclinical studies can help to shed light on this question. In general, the data are conflicting, and both MOR partial agonist and KOR antagonist activity likely play a role in the anti-suicidal effects of buprenorphine. However, the data supporting the importance of KOR antagonism appear slightly more consistent (discussed in more detail below).

Studies combining buprenorphine with opioid antagonists suggest that the MOR is not necessary for the behavioral effects of buprenorphine. In mice, the anti-depressive effects of buprenorphine described above (reduction in forced swim test immobility and reduced approach latency in the novelty-induced hypophagia test) were maintained when buprenorphine (1 mg/kg) was co-administered with the opioid antagonist naltrexone (1 mg/kg), suggesting that activation of MORs is not necessary for the expression of buprenorphine’s anti-depressive effects [173]. Co-administration of buprenorphine (0.1 mg/kg) with the MOR antagonist samidorphan (0.3 mg/kg) in Wistar Kyoto rats did not alter the drug’s efficacy in the forced swim test [174], again suggesting that activation of MOR is not necessary for the anti-depressive effects of buprenorphine.

The use of selective knockout mice provides evidence that the behavioral effects of buprenorphine may be mediated by both MOR and KOR. In mice with genetic deletion of the MOR (Oprm1^−/−) or KOR (Oprk1^−/−), buprenorphine (0.25 mg/kg)-induced decreases in latency in the novelty-induced hypophagia test were blocked in Oprm1^−/− but not Oprk1^−/− mice [175], suggesting that buprenorphine’s activity at MOR, but not KOR, is required for its anti-depressive effects. Consistent with this idea, a mouse model of the A118G polymorphism (associated with less opioid receptor expression and lower signaling efficiency) in the MOR gene (OPRM1) also disrupted the effects of buprenorphine on this behavior [176]. These data are further supported by human studies which have found that the A118G polymorphism was associated with treatment onset suicidal ideation [177] and more severe depression following a recent targeted rejection major life event [178].

In contrast, knockout of the KOR in mice blocked buprenorphine’s reduction of immobility in the forced swim test while knockout of the MOR did not disrupt the behavioral effects of buprenorphine (0.25–0.5 mg/kg) [179], suggesting that buprenorphine’s activity at KOR, but not MOR, is required for its anti-depressive effects. Together, these studies indicate that the role of different opioid receptor types in buprenorphine’s anti-suicidal and anti-depressive effects are likely mediated by multiple factors including behavioral assay, genetic background, and drug dose.

We can also examine the behavioral effects of other KOR and MOR drugs to further explore the potential role of opioid receptor subtype in the behavioral effects of buprenorphine. Since buprenorphine is an antagonist at the KOR, if activity at this receptor is primarily responsible for the anti-suicidal effects of buprenorphine, we would expect manipulations that increase activity at KOR to be pro-depressive while manipulations that decrease activity at KOR to be anti-depressive (Table 1, dark blue cells). Conversely, buprenorphine is a partial agonist of the MOR, so if activity at this receptor is primarily responsible for the anti-suicidal effects of buprenorphine, we would expect manipulations that increase activity at MOR to be anti-depressive while manipulations that decrease activity at MOR to be pro-depressive (Table 2, yellow cells).

In general, KOR agonists produce behaviors indicative of depressed mood or dysphoria. An increase in intra-cranial self-stimulation (ICSS) threshold is a commonly used measure of dysphoria in rodents, and the KOR agonists U69,593 ¹⁰⁸ and salvinorin A [107] have been shown to increase ICSS thresholds in rats; however, another KOR agonist nalfurafine had no effect on this behavior [180]. Note that this effect could be due to the sedative effects of U69,593; however, U69,593 also produced a conditioned place aversion (CPA) in mice [132] and rats [131, 181], an effect consistent with the induction of a negative affective state. The KOR agonist U50,488 also induced a CPA [109], a behavior that we recently showed was exacerbated in chronic pain animals [182]. Given that the vast majority of preclinical studies find that KOR agonism produces depressive-like behaviors (see Table 1), there is strong support for the hypothesis that decreased activity at this receptor underlies the anti-depressant and anti-suicidal properties of buprenorphine.

Conversely, KOR antagonists tend to have the opposite effect, producing antidepressant effects in multiple behavioral assays. In particular, a number of studies indicate that KOR antagonists, including norBNI [109, 171, 179, 183–187], JDTic [187], and MCL-144B [188], reduce immobility on the forced swim test in rats and mice. Numerous studies support the anti-depressive effects of KOR anatgonists, particularly norBNI and JDTic (see Table 1), further supporting the role of KOR antagonism in the behavioral effects of buprenorphine.

MOR agonists have been shown to produce antidepressant-like effects, though their efficacy depends on the particular behavioral assay employed (see Table 2). For example, the MOR agonist morphine produced antidepressant-like effects in the learned helplessness model [189] and the tail suspension test [190], but did not have an effect in the forced swim test [171] or in the novelty-induced hypophagia test [175]. Morphine was effective at recovering pain-reduced behaviors including a pain-induced reduction in ICSS responding (increased ICSS threshold) [106] and pain-reduced nesting behavior [191] (although it did not block reduced nesting behavior caused by the KOR agonist U69,593) [191].

The effects of MOR antagonism on depressive-like behaviors are less conclusive (see Table 2). The selective MOR antagonist cyprodime reduced approach latencies in the novelty induced hypophagia test in mice [175], an antidepressant-like effect which is inconsistent with the hypothesis that a decrease in activity at the MOR should promote depressive behaviors. Furthermore, MOR knockout mice have shown reduced immobility in the forced swim test [192] and reduced anxiogenic and depressive-like responses [193]. These data suggest that activity specifically at the MOR is less likely to underlie the anti-suicidal and antidepressant properties of buprenorphine, since either increases or decreases in activity at the MOR are capable of producing antidepressant behaviors.

Overall, it appears that the evidence for MOR activation underlying buprenorphine’s anti-suicidal effects is less convincing than the evidence for KOR antagonism underlying these effects. This conclusion is based primarily on: 1) the inability of MOR antagonists to block the anti-depressive and anti-suicidal effects of buprenorphine in humans [159, 164, 165] and 2) rodents [173, 174], and 3) the anti-depressive effects of some manipulations which decrease MOR activity [175, 192, 193].