Introduction

An estimated 35.9 million adults per year suffer from alcohol use disorder (AUD) in the United States [1]. Alcohol research has found that dysregulation of the stress system greatly contributes to the reinforcement of alcohol-related behaviors [2, 3]. Currently, there are no FDA-approved medications that target the stress system [4], however, pharmacotherapies targeting the noradrenergic system have been suggested as promising AUD therapeutic interventions [5,6,7].

Activation of the noradrenergic system has been implicated as a component of stress-induced anxiety [8] and alcohol-seeking behaviors [5, 9,10,11,12]. Yohimbine, an α2 adrenergic receptor antagonist, is commonly utilized in alcohol research to pharmacologically probe the noradrenergic system in preclinical [13,14,15,16] and in human laboratory studies [17, 18]. Importantly, the drug increases peripheral [17] and central [13] noradrenergic activity. As yohimbine can be safely administered in animals and humans, it has the potential to be a strong translational pharmacological probe in alcohol research. There are, however, several underdeveloped parameters that may account for reducing the translational efforts of the yohimbine for probing the noradrenergic systems across species. Two major limitations include: the lack of inclusion of sex as a biological variable (SABV) [19] and limited studies examining brain functioning after peripheral yohimbine administration [20, 21].

Although sex differences within alcohol-related behaviors [22, 23], AUD [24], and noradrenergic activation [25, 26] have been well-established, very few studies using yohimbine utilized females, or had the power to include SABV in the analysis. Preclinical and clinical studies suggest that sex differences are largely important in the regulation of alcohol-related behaviors, suggesting that SABV should be investigated in the noradrenergic regulation of AUD [23, 27, 28]. Similarly, although yohimbine has been utilized as a probe to evaluate the response in many brain regions associated with reinstatement of alcohol seeking and consumption in rodents, there are limited investigations that examined the central effect of yohimbine within a clinical context.

The goal of this systematic review is to evaluate published yohimbine literature in alcohol studies, with particular emphasis on findings that inform yohimbine’s peripheral administration on neural mechanistic actions, and assess the value of yohimbine as a translational pharmacological probe for AUD research. We included primary literature of AUD preclinical models and randomized controlled trials (RCTs) involving yohimbine as a pharmacological probe for noradrenergic activation, and provided a detailed overview of: (1) Experimental Characteristics and Methodology, (2) Sex Differences, (3) Neurochemical Systems and Brain Regions, and (4) Discussion of Applications for Medication Development.

While yohimbine has been extensively utilized within animal and human alcohol research for decades, to our knowledge, a systematic review comprehensively assessing the experimental characteristics and yohimbine-related findings to improve translational efforts has not been conducted. We hypothesize that yohimbine is a strong, safe, and titratable translational pharmacological probe, that despite extensive characterization in preclinical models, has not been thoroughly examined as a translational probe for alcohol human laboratory studies. The results from this study will inform future research on the yohimbine’s cross species properties, to improve its use as a translational pharmacological tool to employ in alcohol research with the purpose of facilitating the development of novel pharmacotherapies for AUD.

Methods

The systematic scoping review design was reported in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [29], and the plan was pre-registered with Open Science Frame (OSF, https://osf.io/6vjzd).

Search methods

A literature search was conducted by the Brown University librarians, who constructed comprehensive search strategies in MEDLINE, Embase, Web of Science Core Collection, CINAHL, PsycInfo, and Cochrane Central Register of Controlled Trials. The search strategies used a combination of controlled vocabulary terms and keywords to describe two parameters: “yohimbine” AND “alcohol OR ethanol”. Searches were conducted on February 8th, 2021 and updated on November 15th, 2021. All retrieved studies were exported to EndNote for Windows version X9.3.3 (Clarivate Analytics), and duplicates were removed [30]. The de-duplicated results were uploaded to Covidence for screening and were subsequently reviewed by two independent reviewers in accordance with the PRISMA [29]. Detailed descriptions of search methods are included within the Supplementary Materials.

Eligibility criteria and study selection

We included in vivo preclinical studies and RCTs that administered yohimbine to probe the noradrenergic system to study alcohol-related behaviors. Relevant studies were first screened by the title and abstract and we applied the following criteria: abstract written in English and study conducted in rodents or humans. If criteria determination could not be made based on the title and abstract, the full text of the study was reviewed. Studies were excluded for the following reasons: (1) not reported in English, (2) duplicate study results, (3) not an original study (i.e., review), (4) alcohol/ethanol not studied, (5) incorrect drug used (i.e., not yohimbine), (6) yohimbine not administered in a laboratory paradigm of alcohol research, and (7) not peer-reviewed manuscripts. Title and abstract screening, full text review, and data extraction were completed by two independent reviewers (DEC and TRV-K), with discrepancies resolved by two consensus reviewers for the rodent studies (RC, NC) and one consensus reviewer (CLH-K) for clinical studies. Full description of study inclusion/exclusion criteria are shown within the PRISMA flow diagram (Fig. 1).

Fig. 1: PRISMA Flowchart.
figure 1

A systematic literature search of MEDLINE, Embase, Web of Science Core Collection, CINAHL, PsycInfo, and Cochrane Central Register of Controlled Trials was conducted on February 8th, 2021, and updated on November 15th, 2021. A combination of controlled vocabulary terms and keywords were used to describe two parameters: “yohimbine” AND “alcohol OR ethanol”. This produced 66 studies, with one additional study meeting the criteria was identified during peer review (this study was not returned through the systematic literature search as it did not contain the search terms within the abstract/title [32]). Studies were excluded for the following reasons: (1) study not reported in English, (2) duplicate study results, (3) not an original study (i.e., review), (4) alcohol/ethanol not studied, (5) incorrect drug used (i.e., not yohimbine), (6) yohimbine not administered in a laboratory paradigm of alcohol research, and (7) not peer-reviewed manuscripts.

Full size image

Study characteristics assessments and outcomes

For all studies, we reported: experimental paradigm, yohimbine dose and route of administration, central nervous system (CNS) response, neurochemical systems and brain regions probed, and sex differences. For rodent models, we also reported: rodent strain and genetic background, (Supplementary Table S1) and for human studies, participant description and AUD diagnosis (Supplementary Table S2).

Risk of bias assessments

Risk of bias assessments were conducted by two independent reviewers (DEC and TRV-K), with discrepancies resolved by consensus. Bias in rodent studies was assessed by examining allocation concealment and randomization, blinding, inclusion and exclusion criteria, sample size, ethical compliance, and statistical methods. Assessment findings were reported simply as “Yes” if the criteria were met and reported, or “No” (Table 1) [31]. RCTs were assessed for selection, performance, detection, attrition, reporting biases, and described as high-risk, low-risk, or unclear (Table 2) [29].

Table 1 Risk of bias in rodent studies examining yohimbine in alcohol research; Yes: Criteria was present and met; No: Criteria was not met (or not present to assess).
Full size table
Table 2 Risk of bias in human trials examining yohimbine in alcohol research.
Full size table

Results

Of the 1135 articles initially identified in the literature search, 67 studies satisfied the criteria for inclusion in this systematic review after full-text screening, including one additional preclinical study which was identified during peer review [32]. Sixty-two studies were conducted in rodents, and five studies were on humans (PRISMA diagram, Fig. 1). Excluded studies that were identified prior to the screening process are also reported in the (Supplementary Table S3).

Preclinical studies

Experimental characteristics and methodology

Of the 62 studies conducted in rodents, 53 included rats and nine studies included mice. Rodents of various genetic backgrounds were broadly categorized as alcohol-preferring (AP) or not. Of the studies utilizing rodents not genetically selected to prefer alcohol: eight used Sprague-Dawley [16, 33,34,35,36,37,38,39], 12 Wistar [14, 16, 40,41,42,43,44,45,46,47,48,49], and 12 Long-Evans [13, 50,51,52,53,54,55,56,57,58,59,60] rats; two used Swiss-Webster [61, 62], three Albino [63,64,65], three C57/BL6 [66,67,68] mice. Studies involving AP rodents: nine used Marchigian Sardinian Alcohol-Preferring (msP) rats [32, 69,70,71,72,73,74,75,76], five inbred alcohol-preferring (iP) rats [77,78,79,80,81], and one used rats descended from Finnish AA line [82]. Several studies utilized two rodent strains: two Wistar and Long-Evans rats [83, 84]; two studies had a non-preferring line as controls: one Wistar and alcohol-preferring (P) [85], and one Wistar and msP [86] rats; one study used two AP lines: one P and HAD-2 rats [15]; one used Long-Sleep and Short-Sleep mice [87]. Only one study did not specify the strain of the rats used [88] (Supplementary Table S1).

In most rodent studies (N = 59), yohimbine was administered through intraperitoneal (IP) injection, with dose from 0.001 mg/kg [88] to 10 mg/kg [64, 88] (Fig. 2). Other methods of yohimbine injection include subcutaneous (SC, 2 mg/kg) [82], intravenous (IV, 1 mg/kg) [39], and intracerebroventricular (ICV, 1 μg) [87] administration. Fifteen experiments investigated multiple doses of yohimbine [15, 33, 42, 44, 46, 59, 61,62,63, 65, 68, 84,85,86, 88] (Supplementary Table S2).

Fig. 2: Yohimbine doses administered in animal studies.
figure 2

Yohimbine was primarily administered intraperitoneally (IP) in the 59 studies included in this figure. Three studies, not included in this figure administered yohimbine in the following dosages and methods: subcutaneous injection of 2 mg/kg [82], venous cannula injection of 1 mg/kg [39], and intracerebroventricular injection of 1 μg [87].

Full size image

Yohimbine was primarily used to investigate the impact of noradrenergic activation on alcohol self-administration and reinstatement of alcohol seeking.

Sex differences

Of the 62 rodent studies, only seven included females [33, 35, 63, 68, 70, 76, 82], and only four examined sex differences [33, 35, 68, 70]. Alcohol self-administration was included in the studies, however, yohimbine administration was tested only in reinstatement of alcohol seeking.

In Sprague-Dawley rats, chronically exposed to corticosterone in adolescence, females were more sensitive to yohimbine-induced reinstatement than males [33]. Additionally, in a yohimbine administration paired to alcohol-related cues paradigm, females showed enhanced reinstatement of alcohol seeking compared to males [33]. However, chronic and acute hormone manipulation (female: estradiol and male: testosterone) did not affected the yohimbine + cue-induced reinstatement of alcohol seeking [35]. Recently, it was reported that in msP rats, females exhibited more pronounced yohimbine-induced reinstatement of alcohol seeking compared to males [70]. In C57BL/6 J mice, yohimbine-induced reinstatement of alcohol seeking was reduced by IP oxytocin, both in male and females [68].

Neurochemical systems and brain regions

Extensive rodent work utilized yohimbine to study neurochemical and receptor systems: corticotropin releasing factor (CRF), opioid, neurokinin1 (NK1), peroxisome proliferator-activated receptor γ (PPARγ), endocannabinoid, and orexin. Additionally, several studies included analysis of multiple brain regions evaluated by Fos expression profiles.

CRF system and hypothalamic–pituitary–adrenal (HPA) axis activation

A wealth of studies demonstrated that yohimbine-induced increase of alcohol drinking and seeking is associated with mechanisms involving the activation of the CRF system [14, 36, 45, 52, 58, 67, 78, 86, 89]. Administration of yohimbine significantly increases expression of CRF mRNA in the dorsal region of the bed nucleus of the stria terminalis (BNST) and the central nucleus of the amygdala (CeA) of Wistar rats [89], and the paraventricular nucleus (PVN) of C57BL/6 J mice [67]. Downstream of the HPA axis, administration of yohimbine increased plasma corticosterone levels in the alcohol-trained animals compared to controls [58]. Yohimbine administration was tested both in alcohol-self administration and reinstatement of alcohol seeking.

Self-administration

Studies in Long-Evans rats trained to self-administer alcohol, reported that viral-mediated downregulation of CRF binding protein (CRFBP) in the CeA reduced alcohol consumption compared to controls, however this manipulation was not sufficient to attenuate yohimbine-induced increases in alcohol self-administration [52]. Pretreatment with the CRF1 receptor antagonist, antalarmin, blocked yohimbine-induced increases in alcohol self-administration in both Wistar [14, 86] and msP rats [86].

Reinstatement

Yohimbine-induced reinstatement of alcohol seeking was attenuated by infusions of d-Phe CRF (non-specific CRF receptor antagonist) into the median raphe nucleus of Wistar rats [45]. In iP rats, yohimbine-induced reinstatement of alcohol seeking was reduced by infusions into the nucleus incertus (NI) of CP376395 (CRF1 receptor antagonist), but not by astressin-2B (CRF2 receptor antagonist) [78]. Glucocorticoids, that are downstream key regulators of the CRF system, were also investigated [13, 34] through yohimbine-induced reinstatement of alcohol seeking. In Long-Evans rats, systemic and intra-CeA, but not intra-basolateral amygdala (BLA), infusions of mifepristone (glucocorticoid receptor antagonist) suppressed yohimbine-induced reinstatement of alcohol seeking without changing plasma corticosterone level [13]. Administration of corticosterone in adolescent Sprague-Dawley rats was shown to enhance yohimbine stress-induced reinstatement in alcohol-reinforced rats, compared to controls [34]. Also, intra-BLA infusion of a G-protein coupled receptor kinase 2 (GRK2) inhibitor reduced yohimbine-induced alcohol seeking [34]. The only mouse study using yohimbine to probe the role of the CRF system in alcohol seeking indicated that overexpression of CRF1 receptor in calcium-calmodulin-dependent kinase II (αCaMKII) neurons in the CeA increased yohimbine-induced reinstatement of alcohol seeking [36].

Opioid, neurokinin, PPARγ, orexin, and endocannabinoid system

Within the investigation of the opioid, neurokinin, PPARγ, orexin, and endocannabinoid system, yohimbine administration was tested only in reinstatement of alcohol seeking.

Opioid

Naltrexone (non-selective opioid receptor antagonist) did not significantly impact yohimbine-induced reinstatement in msP rats [74]. However, studies conducted in Long-Evans rats indicated that pretreatment with nor-BNI (κ-opioid) [50], and SoRI-9409 (δ-opioid) [53] receptor antagonists, reduced yohimbine-induced reinstatement without increasing plasma corticosterone levels [53]. Yohimbine-induced reinstatement of alcohol seeking was significantly blocked by administration of MT-7716 (potent and selective NOP receptor agonist) in msP rats [32], and by administration of SR-8993 (NOP receptor agonist) in Wistar rats [40]. Intriguingly, it was later reported that LY2940094 and LY2817412 (potent and selective NOP antagonists), also attenuated yohimbine-induced reinstatement of alcohol seeking in msP rats [70, 76]. LY2817412 was efficacious when administered in the ventral tegmental area (VTA), and CeA, but not the accumbens (NAc), both in male and female msP rats [70].

Neurokinin

Consistent with literature linking the NK1 receptor system to the motivation for alcohol, it was found that administration of NK1 receptor antagonists L822429 reduced yohimbine-induced reinstatement of alcohol seeking in msP and Wistar rats [69], [47]. Similarly, in Long-Evans rats, L822429 attenuated yohimbine-induced reinstatement of alcohol seeking, but not alcohol intake [84]. Interestingly, in P rats (with an innate upregulation of the NK1 system), intra-CeA infusion of L822429 attenuated reinstatement of alcohol seeking [85]. Whereas, in Wistar rats viral-mediated overexpression of NK1 in the CeA enhanced the sensitivity to yohimbine-induced increases in alcohol drinking and seeking [85].

PPARγ

The PPARγ pathway was examined in msP rats in four studies utilizing yohimbine with the administration of pioglitazone (selective PPARγ agonist) [71, 72, 74], and the extracts from andrographis paniculata plant and andrographolide (its major active compound) [75]. Pioglitazone attenuated yohimbine-induced reinstatement of alcohol seeking, but not by alcohol cue [72, 74]. Pioglitazone’s actions were shown to be region specific in the CeA, VTA, RMTg, but not the NAc shell, injections of pioglitazone inhibited yohimbine-induced reinstatement [71]. Finally, extracts from andrographis paniculata plant and andrographolide also significantly reduced yohimbine-induced reinstatement [75].

Orexin

First, it was shown that in iP rats, orexin neurons in the NI were activated by yohimbine-induced reinstatement of alcohol seeking [77]. Then, NI bilateral microinjections of TCS-OX2-29 (orexin2 receptor antagonist) attenuated yohimbine-induced reinstatement of alcohol seeking, while intra-NI injection SB-334867 (orexin 1 receptor antagonist) showed no significant effect [77]. Systemic administration of SB-334867 in Long-Evans rats similarly attenuated yohimbine-induced reinstatement of alcohol seeking [54].

Endocannabinoid

CB1-receptor-mediated activity was investigated using a selective fatty acid amide hydrolase (FAAH) inhibitor (URB597), which had no effect on yohimbine-induced relapse in Wistar rats [49].

Brain regions

Several studies examined yohimbine-induced changes in Fos expression throughout the brain [16, 41, 67, 81, 89] in different rat strains and mostly evaluated in reinstatement of alcohol seeking induced by yohimbine. Only one study evaluated both yohimbine-induced alcohol self-administration and then reinstatement induced by yohimbine after extinction of the operant response. In Wistar rats, dopaminergic and noradrenergic neuron lesions of the dorsal or ventral NA bundles by neurotoxin 6-hydroxydopamine (6-OHDA) did not impact yohimbine-induced alcohol self-administration or reinstatement [42].

In iP rats, Fos immunohistochemistry analysis across 42 brain regions following yohimbine-induced reinstatement of alcohol seeking showed that the yohimbine robustly activated neurons in the prefrontal cortex (PFC), extended amygdala, hypothalamus, thalamus, and the hippocampus [81]. In another study, yohimbine-induced reinstatement of alcohol seeking increased Fos activation in the CeA, which was attenuated by intra‐CeA injections of a selective Relaxin Family Peptide Receptor 3 antagonist (R3(B1‐22)R) [79]. Yohimbine-induced reinstatement, also led to increased Fos expression in cocaine and amphetamine regulated transcript (CART) containing neurons in the capsular (CeC) and lateral (CeL) part of the CeA [80]. Yohimbine-induced reinstatement of alcohol seeking was attenuated by direct injection of the neutralizing CART antibody into the CeA [80].

In Wistar rats yohimbine-induced neuronal activation in the CeA and NAc, but not in the BLA [41]. In situ hybridization analysis showed that yohimbine significantly increased cFos expression in brain regions associated with alcohol and drug seeking, such as the anterior cingulate cortex (ACg), orbitofrontal cortex (OFC), NAc core, shell, dorsal and ventral BNST, PVN, BLA, and CeA [89]. Many of those brain regions overlapped in C57BL/6 J mice [67].

In Long-Evans rats, yohimbine blocked alcohol-induced cFos expression in the Edinger-Westphal nucleus in a dose-dependent manner [66]. Additional studies investigating the role of yohimbine in discrete brain regions included electrolytic and chemical lesioning. Lesions in the lateral habenula were found to block yohimbine-induced reinstatement [51]. Whereas lesions of the RMTg had no significant impact on yohimbine-induced reinstatement of alcohol seeking [56].

Finally, in Sprague-Dawley rats, yohimbine significantly increased Fos immunoreactive expression in the NAc [16], while injection with prazosin, an α1 antagonist [16], prior to yohimbine resulted in significantly fewer Fos-immunoreactive cells in the NAc.

Medication development

Several drugs targeting the noradrenergic system were investigated in conjunction with yohimbine to evaluate both in alcohol-self administration and reinstatement.

Self-administration

In Wistar rats, baclofen (γ-aminobutyric acid (GABA)-B agonist) blocked yohimbine-induced increases in alcohol self-administration [48], and WAY 100,635 (selective serotonin (5-HT) 1 A receptor antagonist) reduced yohimbine-induced self-administration [42]. Clonidine (non-selective α2 receptor agonist), however, was not sufficient to attenuate yohimbine-induced increases in alcohol self-administration [42].

Reinstatement: Prazosin (α1 receptor antagonist) blocked yohimbine-induced reinstatement of alcohol seeking, and guanfacine (α2 A receptor agonist) successfully attenuated yohimbine-induced reinstatement of alcohol seeking both in Wistar and Long-Evans rats [83]. The yohimbine-induced reinstatement of alcohol seeking was successfully reversed by administration of clonidine in: Wistar rats [42, 46], selectively bred Long-Sleep and Short-Sleep mice [87], alcohol-preferring Finnish AA rats [82], and Sprague-Dawley rats [38]. WAY 100,635 also attenuated yohimbine-induced reinstatement of alcohol seeking [42]. Systemic and intra-CeA infusion of mifepristone, suppressed yohimbine-induced reinstatement of alcohol seeking in Long-Evans [13]. Finally, the administration of AP-202 (selective nicotinic α4β2-receptor antagonist), was unable to reverse yohimbine-induced reinstatement of alcohol seeking in Sprague-Dawley rats [37].

Clinical studies

Experimental characteristics and methodology

Individuals with diagnosis of AUD were included in the RCTs analyzed, except in one study in which all participants were healthy individuals [17]. Three RCTs [90,91,92] were conducted in an inpatient setting with the same 22 patients, from the West Haven Veterans Administration Medical Center and one with 25 patients at the NIH Clinical Research Center [18]. Two RCTs [91, 92] included control subjects. Ethnic background (56% white) of the participants were only provided in one study [18]. The mean age ranged from 39 [90,91,92] to 44 [18] years. None of the RCTs administered yohimbine combined to a behavioral paradigm (e.g., alcohol self-administration or cue reactivity). Nor did any study use yohimbine for brain imaging or medication development. The route of administration of yohimbine was reported in all RCTs (IV, 0.4 mg/kg for 10 min) [17, 18, 90,91,92].

Sex differences

Two clinical studies included women; a study that evaluated the craving response (without cue reactivity) after yohimbine challenge (N = 24, 12.5% women) [18] and a study that evaluated the co-administration of alcohol and yohimbine on severity of acute intoxication (N = 12, 71% women) [17]. However, neither of these studies addressed sex differences. The remaining three RCTs analyzed were conducted with only men [90,91,92].

Neurochemical systems and brain regions

Within the RCTs examined, no study utilized imaging paradigms to investigate the brain activities following a pharmacological challenge with yohimbine. Similarly to the rodent studies, plasma cortisol levels (in four out of the five studies) were measured to determine the impact of yohimbine on the downstream of HPA axis [17, 18, 90, 91]. Administration of yohimbine robustly activated the HPA axis, with increase of cortisol level both in individuals with AUD [17, 18, 90, 91] and controls [17], as well as increased anxiety [17, 90, 91], and alcohol craving [18]. Only one study [18] demonstrated that yohimbine had an effect on alcohol craving on patients going through withdrawal.

Medication development

In a RCT conducted in treatment seeking individuals with AUD, administration of yohimbine was compared to the serotonergic compound meta-chlorophenylpiperazine (mCPP), and placebo (double blind, counterbalanced) to assess the effect of acamprosate administered for 14 days on craving. Both yohimbine and mCPP successfully induced craving compared to placebo, with stronger effect in patients with severe AUD [18]. The study did not observe the effect of acamprosate in reducing craving in this experimental condition.

Safety and tolerability

No serious adverse reactions to yohimbine administration were reported in any rodent or human studies. Three RCTs [17, 90, 91] examined the effect of yohimbine on hemodynamic parameters, with increase of systolic and diastolic blood pressure and heart rate [91]. Four of the five studies [17, 18, 90, 91] measured yohimbine’s effect on anxiety and stress related feelings, with yohimbine either increasing nervousness [91] and anxiety [17] or having no significant effect [18].

Risk of bias assessments

The rodent studies on average met six out of the 13 criteria to assess bias. Selection bias for all five human studies was determined to be low-risk, as all studies were randomized, and allocation was concealed (Table 2). Literature sources, process, and limitations are fully described in the Supplementary (Tables S2, S3, and S4).

Discussion

In the alcohol field, yohimbine as a pharmacological agent to investigate the noradrenergic system has been employed in numerous studies with greatly varied aims and methodology. To our knowledge, this is the first systematic review evaluating primary literature on yohimbine in alcohol research, in rodents and humans. Sixty-two studies in rodents and five RCTs have been evaluated. Summarized below are key findings, and assessment of limitations to inform future research using yohimbine and improve the robustness of translational efforts and alignment of findings.

Main findings

Yohimbine is a well validated preclinical pharmacological probe for AUD research

The evaluated Experimental Characteristics and Methodology Results of this review showed that extensive preclinical research supports yohimbine as an efficacious pharmacological probe for the activation of noradrenergic system to investigate alcohol-related behaviors. There are however, limited human laboratory studies that have implemented similar research endeavors.

First, recent advocacy for preclinical models that have a higher capability for clinical translation is reflected in the studies analyzed in this review, which included rodent models genetically predisposed to consume alcohol. It has been well established that risk for AUD in humans has significant genetic contributions, which supports the necessity of preclinical models in AUD research that more closely model genetic contributions to the disorder [93, 94]. Only two studies included in this review directly compared yohimbine’s actions in an experimental paradigm between AP (AUD model) and not genetically-selected rats (control) [85, 86], and one study compared two AP strains [15].

In the highly characterized msP rats line, researchers extended the evaluation of yohimbine effect to specifically test the two msP-derived lines carrying the wildtype (GG) and the point mutation (AA) in the CRFR1 causing the overexpression of this receptor [86]. Administration of multiple doses of yohimbine, elicited increased operant alcohol self-administration and reinstated alcohol seeking in both AA and GG msP rats and in Wistar controls. Interestingly, however, the highest dose tested (2.5 mg/kg) failed to reinstate alcohol seeking in AA msP rats [86]. Further, despite earlier studies indicating P rats express greater sensitivity to alcohol seeking and consumption, a study within the review, showed that yohimbine elicited reinstatement of alcohol seeking in both in P and HAD-2 rats, however the HAD-2 rodents consumed more alcohol [15]. Similarly, studies using yohimbine have been conducted in different mice strains. Yohimbine activated the HPA axis response in Long-Sleep mice, as indicated by increased plasma corticosterone levels, but not in Short-Sleep mice; however, yohimbine did not alter alcohol-induced adrenocortical responses in either line [15]. The only study that showed specific differences between AP and Wistar rats reported that after yohimbine administration, P rats showed increased CeA neuronal activation and higher sensitivity to yohimbine-induced reinstatement compared to Wistar controls [85]. Overall, findings from reviewed studies generally support that yohimbine exerts similar actions in AP and control rodents in inducing alcohol self-administration and reinstatement of alcohol seeking (although in rats with an overactive CRF system, yohimbine seems to work at lower doses) [86], with one study observing greater yohimbine sensitivity in P rats compared to Wistar [85].

Regarding the study population in human laboratories, two clinical studies included individuals with AUD and healthy controls [91, 92]. In both studies, both groups showed similar responses to yohimbine, as measured through hemodynamic parameters. However, yohimbine-induced changes in plasma cortisol, prolactin, and plasma 3-methoxy-4-hydroxyphenylglycol (MHPG) were greater in individuals with AUD compared to healthy controls. From translational perspective, the cortisol effect is in line with preclinical data that have shown that administration of yohimbine increased plasma corticosterone levels in the alcohol-trained animals compared to controls [58].

Unfortunately, alcohol behavioral paradigms paired with yohimbine administration have been extensively investigated only in preclinical research. To this end, there are two ongoing RCTs that are examining the effect of yohimbine paired with alcohol cue reactivity, in conjunction with pharmacological treatment of an α1 receptor blocker (doxazosin: NCT02243709) and a glucocorticoid receptor blocker (mifepristone: NCT04135846).

Administration of yohimbine in preclinical studies was primarily through IP injections, while all RCTs administered yohimbine through IV. Only rodent models allow for ICV administration of yohimbine, which is an administration method essential for evaluating the mechanistic effects of the drug within the CNS, and no rodent or human study administered yohimbine orally. In preclinical models, yohimbine was most often utilized at a dosage of 1.25 mg/kg, with dosages ranging from 0.0001–10 mg/kg. Several studies comparing multiple doses of yohimbine similarly found that the dose of 1.25 mg/kg was ideal to produce the desired yohimbine-induced reinstatement of alcohol seeking within AP and non-preferring rodent strains [15, 33, 42, 44, 59, 84, 85]. Conversely, only one dosage of yohimbine, 0.4 mg/kg, was investigated in the clinical studies, highlighting the need for investigation of various doses within RCTs. From a translational perspective, a study on oral yohimbine in non-human primates showed increase in motor activation and affective response, but not in a dose-response manner [95]. Yohimbine oral dose of 32.4 mg is currently under investigation in the two RCTs mentioned above (NCT02243709 and NCT04135846).

Sex differences exist in yohimbines effects on alcohol-related behaviors

Numerous preclinical and clinical studies have indicated that sex differences exist in AUD and alcohol-related behaviors [24, 28, 96, 97], particularly in stress-related alcohol consumption [23, 98]. The three studies within this review that examine sex differences support previous evidence indicating that differences exist in yohimbine effects on alcohol-related behaviors [33, 35, 68], with female rodents exhibiting greater baseline susceptibility compared to males, and increased sensitivity to yohimbine-induced reinstatement of alcohol seeking, as well as other drugs [99,100,101].

There is also a dearth of clinical research examining sex differences in the utilization of yohimbine within human AUD research. Although no analysis of sex differences was conducted in any studies that met the criteria for the systematic review, two studies [17, 18] included female participants. However, the sample size was not large enough to analyze SABV. Currently there are several active clinical trials utilizing yohimbine in healthy individuals (NCT04180969) and other substance use disorders (NCT04231708, NCT04051619, NCT04181515), which may be able to further elucidate the role that sex differences play in yohimbine’s mechanistic actions.

From translational perspective, studies examining cocaine use disorder, found that high levels of endogenous progesterone attenuate drug cues after yohimbine administration in women, compared to women with low progesterone levels [102, 103]. These findings taken together illustrate the need for future research to include SABV, specifically in the investigation of stress-induced alcohol-related behaviors, and further examine sex differences within preclinical and clinical AUD models to better inform personalized medicine approaches.

Neurochemical systems and brain regions

Numerous neurochemical and receptor systems have been examined within the context of stress and alcohol-related behaviors, with this relationship further explored in several notable reviews [104,105,106]. It has been shown that chronic alcohol treatment selectively increases α2 adrenergic receptor densities [107]. Rodent studies, however, have reported conflicting results on the receptor expression at peripheral and central level. For example, IP yohimbine administration in male Wistar-Kyoto rats increases α2 receptor density in the kidneys [108], but ICV injection of yohimbine in male Wistar rats decreases α2 receptor density in hippocampal neurons [109]. Future studies, however, could further elucidate this relationship by examining yohimbine-induced alterations in α2 receptor density within an alcohol research paradigm.

Additionally, the HPA axis response in animal models (corticosterone) has been closely aligned with the neuroendocrine response (cortisol) in individuals with AUD [110,111,112]. As shown in the included studies, yohimbine administration in both preclinical and clinical settings demonstrated increases in corticosterone and cortisol levels respectively. Interestingly, it has been suggested that yohimbine’s effects on alcohol-related behaviors are independent of yohimbine’s actions on the HPA axis [14], which was further supported through findings in another study, which proposed that the primary action of yohimbine in reinstatement studies is to increase salience of the selected cues [113]. Taken together, those neuroendocrine results, support the translational efforts of measuring glucocorticoids response both in preclinical and clinical studies, ideally using aligned paradigms. In fact, studies evaluating the relationship between the HPA axis’s effect on increasing salience of alcohol cues and noradrenergic activation by yohimbine may utilize these peripherally measured glucocorticoids to elucidate the CNS mechanistic actions, which are extremely difficult to study in human laboratory studies.

Medication development

The noradrenergic system, with projections through key limbic and forebrain areas involved in the arousal, reinforcement, and stress processes involved with the development and maintenance of AUD [114], has been proposed as a pharmacological target for the stress contributing to AUD [5, 115].

Under normal conditions, noradrenergic activation has been associated with the regulation of stress, anxiety, vigilance, and arousal through the release of norepinephrine from the locus coeruleus [5, 116]. However, in individuals with AUD, noradrenergic dysregulation resulting in elevated norepinephrine levels has been suggested as a contributor to alcohol-related behaviors, like alcohol craving and consumption [115, 117]. Further, preclinical [16, 83] and clinical [118,119,120,121] AUD studies utilizing α1 noradrenergic receptor antagonists have been effectively used to reduce alcohol consumption and seeking. Thus, targeting of the noradrenergic system with yohimbine to increase available norepinephrine provides researchers with a valuable pharmacological tool to study the role of the noradrenergic system within AUD. One RCT was the first to utilize yohimbine and compare to mCPP, and placebo to challenge the FDA-approved medication for AUD, acamprosate [18], highlighting the importance of understanding the craving component and noradrenergic activation in developing pharmacotherapies for AUD [122].

Rodent models have provided the building blocks for AUD medication development. Therefore, investigating interspecies differences between human and rodent α2 adrenoceptor genetic variation is critical to fully utilize yohimbine as translational tool. Studies evaluating the effects of mutations on ligand selectivity in mouse and human α2A receptors demonstrated that point mutation at the extracellular loop impacts the binding preferences of yohimbine for the human (K1 increased 6-fold) and mouse α2A receptors [123]. In terms of physiological response, however, human research showed that genetic variation of α2 receptor loci (α2A, α2B, and α2C), did not affect yohimbine-induced increase in blood pressure [124].

Limitations

The wide variety in preclinical rodent models used to study the effect of yohimbine on alcohol drinking and seeking limits the generalizations of the findings highlighted by the present analysis. It is important to acknowledge that yohimbine, as pharmacological stressor has also been utilized in preclinical [125] and clinical [126, 127] setting outside the AUD field with translational aspect that are not reported here (not AUD research).

Limited preclinical investigation of SABV and a complete lack of clinical studies examining sex differences represents a significant limitation of yohimbine-related AUD research. Additionally, generalization of the clinical findings to the overall population is hindered in that three of the five RCTs were conducted within the exact same participant group [90,91,92]. In the preclinical studies examined, very few studies looked at neural network analysis or imaging. This made it difficult to examine the actions of peripheral yohimbine administration in the brain.

Also, there are seminal studies on yohimbine that were not included in this systematic review because they did not meet the inclusion criteria (e.g., paradigm not tested for an alcohol outcome) that provide additional information for future research. For example, a study evaluating the dose-response of yohimbine’s effects on reinstatement of food/drug seeking and operant self-administration, demonstrated that yohimbine’s effect may be independent of the food/drug reward during training sessions, and therefore may not be stress-associated pathways [113].

Translational summary

Despite the reported safety and tolerability data and the common exploration of alcohol research domains (consumption, craving, and neural response), none of the reviewed studies included a translational paradigm across preclinical and clinical setting. Despite the several alignments between rodent and human studies (e.g., corticosterone/cortisol; AUD models and AUD diagnosis), the lack of integration between preclinical and clinical paradigms provided fragmented and often incomplete results. Experimental characteristics and methods with parallel examination of yohimbine in AUD preclinical models and in alcohol human laboratory studies would reduce the knowledge gap when evaluating sex differences, action on neural response, and of applications for medication development.

Future directions

Yohimbine as a pharmacological probe within translational alcohol research would greatly benefit from the development of future studies, within integrative parallel paradigms of AUD. Such studies would serve to reduce the translational gap, allowing for more rapid identification of promising clinical targets for AUD and the development of pharmacotherapies through clinical trials. Further, although development of genetically-selected rodent models of AUD has allowed for preclinical models that more closely represent the genetic variances contributing to AUD within humans, further studies directly comparing the rodent strains used within AUD research is necessary to gain a more in depth understanding of the role genetic alterations play in AUD.

Additionally, this area of research would benefit from further study investigating the interaction between specific neurotransmitter systems and yohimbine, to further parse the role of noradrenergic activation in alcohol seeking behaviors. This topic has been highlighted in a recent review, which provides a thorough overview of the neurobiology of alcohol seeking behavior, providing essential insight on neurobiological mechanisms mediating alcohol seeking, and the role of respective neurotransmitter systems [104].

Conclusion

In summary, this systematic scoping review broadly examines the alcohol research literature utilizing yohimbine. Findings support the use of yohimbine as a safe, titrable pharmacological probe for AUD. Further, considerations must be made to reduce the translational gap between preclinical and clinical studies. To that end, utilization of males and females, genetic rodent models of AUD vs non-selected control lines, clinical AUD trials involving a range of yohimbine dosages, and studies conducting preclinical and clinical research in parallel would serve to greatly advance the utility and understanding of yohimbine as a pharmacological probe.