Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Aug 1.
Published in final edited form as: Drug Alcohol Depend. 2020 May 25;213:108076. doi: 10.1016/j.drugalcdep.2020.108076

GABAA Receptor Subtypes and the Reinforcing Effects of Benzodiazepines in Remifentanil-Experienced Rhesus Monkeys

Lais F Berro 1,*, James K Rowlett 1
PMCID: PMC7371532  NIHMSID: NIHMS1598394  PMID: 32474260

Abstract

Background:

Opioid-use disorder is associated with a high degree of co-abuse with benzodiazepines. While the mechanisms underlying the co-abuse of opioids and benzodiazepines remain unknown, α1 subunit-containing GABAA receptors may play a critical role in the reinforcing effects of benzodiazepine-type compounds, depending on whether the monkeys have a history of benzodiazepine or stimulant self-administration. The present study extended our prior research by comparing the reinforcing effects of a compound lacking activity at α1 subunit-containing GABAA receptors with the reinforcing effects of non-selective GABAA receptor positive allosteric modulators in monkeys with a history of opioid self-administration.

Methods:

The reinforcing effects of L-838,417 (partial intrinsic efficacy at α2, α3, and α5 subunit-containing GABAA receptors, but no efficacy at α1 subunit-containing GABAA receptors, i.e., “α1-sparing compound”) were compared with those of the non-selective GABAA receptor partial modulator MRK-696, and non-selective GABAA receptor full modulators, triazolam and lorazepam, in rhesus monkeys (n=3) experienced in remifentanil self-administration under a progressive-ratio schedule of intravenous drug injection.

Results:

Neither the partial modulator nor the α1-sparing compound were self-administered above vehicle levels. The full modulators triazolam and lorazepam were self-administered significantly above vehicle levels, albeit at lower levels than remifentanil.

Conclusions:

Our findings suggest that relatively high efficacy at one or more GABAA receptor subtypes is required for a compound to have reinforcing effects in monkeys with a history of remifentanil self-administration, in contrast to monkeys with benzodiazepine or stimulant self-administration histories.

Keywords: rhesus monkey, remifentanil, benzodiazepine, self-administration, GABAA receptor, drug history

1. Introduction

The co-abuse of opioids and benzodiazepines has been recognized since the 1970’s. Early human studies already reported high rates of opioid and benzodiazepine co-use, with benzodiazepines increasing the subjective-rating effects of opioids (Preston et al., 1984). Currently, in the United States, ~20% of individuals prescribed a benzodiazepine or a benzodiazepine-like drug (e.g., zolpidem) for an extended period of time are co-prescribed an opioid (Moore and Mattison 2017). Alarmingly, overdose deaths due to concurrent opioid and benzodiazepine use have increased at least two-fold since 1999 (Paulozzi et al. 2015), and for 2015-2016, young adults were the most likely of all age groups to engage in opioid and benzodiazepine abuse (Schepis et al. 2018). Although the negative consequences of opioid-benzodiazepine co-abuse are clear, the mechanisms underlying this form of polydrug abuse are understood poorly.

Benzodiazepines exert their pharmacological effects by binding to γ-aminobutyric acid type A (GABAA) receptors (Möhler, 2011), with their major targets being GABAA receptors containing α1, α2, α3 and α5 subunits (α1GABAA, α2GABAA, α3GABAA, and α5GABAA, respectively; for reviews, see Engin et al., 2018; Möhler, 2011). Previous studies pointed to both the α1GABAA and the α3GABAA receptors as key mediators of benzodiazepine self-administration in monkeys with a history of self-administration of the conventional benzodiazepine midazolam (Rowlett et al., 2005; Rowlett and Lelas, 2007; Svitek et al., 2008; Shinday et al., 2013; Meng et al., 2019). Importantly, our previous studies also suggest that α1GABAA, but not α3GABAA, receptors may have a prominent role in the reinforcing effects of benzodiazepine-type compounds in monkeys with a history of cocaine self-administration, whereas the α1GABAA subtype is not necessary for a benzodiazepine-type compound to have reinforcing effects in monkeys with a history of midazolam self-administration (Shinday et al., 2013). Therefore, the prior drug experience of the subject may be a critical determinant of the GABAA receptor subtype mechanisms underlying self-administration of benzodiazepines.

In the present study, we extended these studies to investigate the reinforcing effects of GABAA receptor positive modulators in monkeys with a history of opioid self-administration. Because α1GABAA receptors seem to play a critical role in the reinforcing effects of benzodiazepine-type compounds in monkeys with a history of barbiturate, benzodiazepine and stimulant self-administration (Rowlett et al. 2005; Shinday et al., 2013; Meng et al., 2019), we compared the reinforcing effects of a compound lacking activity at α1GABAA receptors with the reinforcing effects of conventional non-selective GABAA receptor positive allosteric modulators (triazolam, lorazepam) in monkeys with a history of opioid (remifentanil) self-administration. Remifentanil was chosen as the training drug because it maintains robust i.v. self-administration in rhesus monkeys under conditions used in this study (Woolverton et al., 2008) and has an “ultra-short” elimination half-life (generally <10 min after i.v. injection in human subjects), thereby minimizing the likelihood of physical dependence as a potential experimental factor (Stroumpos et al. 2010). As with our previous studies, we used a progressive-ratio (PR) schedule of reinforcement which has been used previously to study benzodiazepine and opioid reinforcement (e.g., Woolverton et al., 2008; Shinday et al. 2013), in the absence of physical dependence for either drug class (cf. Rowlett and Woolverton, 1997). The α1-sparing compound used was L-838,417 [3-(2,5-Difluorophenyl)-7-(1,1-dimethylethyl)-6-[(1-methyl-1H-1,2,4-triazol-5-yl)methoxy]-1,2,4-triazolo[4,3-b]pyridazine], which has partial intrinsic efficacy at α2, α3, and α5 subunit-containing GABAA receptors, but no efficacy at α1 subunit-containing GABAA receptors. Because L-838,417 has partial efficacy at α2/α3/α5GABAA receptors, its effects also were compared with the non-selective partial modulator MRK-696 [7-cyclobutyl-6-(2-methyl-2H-1,2,4-triazol-2-ylmethoxy)-3-(2-flurophenyl)-1,2,4-triazolo(4,3-b)pyridazine; no selectivity but partial intrinsic activity] (Atack, 2011; Shinday et al., 2013).

2. Methods

2.1. Subjects

Three adult (ages 10, 10 and 25 for subjects M-1170, M-1027 and M-98-007, respectively) male rhesus monkeys (Macaca mulatta) weighing 11-14 kg were housed individually and maintained on a 12h light/12h dark cycle (lights on at 6h), at a temperature of 21±2°C, with water available ad libitum and monkey diet available once/day, supplemented by fresh fruit. Monkeys were weighed every other week during physical examinations. Amount of chow for each monkey was determined in consultation with veterinary staff to be that which maintains healthy weights in rhesus monkeys. Animals were fed before the beginning of their self-administration sessions, and were given fruit/forage twice a day (before and halfway through their self-administration session).

All three subjects were experimentally naive prior to the beginning of the studies. The monkeys were prepared with a chronic indwelling venous catheter (femoral, brachial or jugular vein) according to the procedures described by Platt et al. (2011). The catheter was protected by a stainless steel tether and cloth jacket (Lomir Biomedical INC., Malone, NY, USA), and flushed daily with heparinized saline (150–200 U/ml). All of the procedures and animal maintenance was in accordance with the Guide for the Care and Use of Laboratory Animals (National Research Council, 2011), with review and approval via the Institutional Animal Care and Use Committees of the University of Mississippi Medical Center.

2.2. Drugs

Remifentanil hydrochloride was purchased commercially (ChemPartner Co., Ltd., Shangai, China) and solutions were prepared using 0.9% saline. L-838,417 [3-(2,5-Difluorophenyl)-7-(1,1-dimethylethyl)-6-[(1-methyl-1H-1,2,4-triazol-5-yl)methoxy]-1,2,4-triazolo[4,3-b]pyridazine] has partial intrinsic efficacy at α2, α3, and α5 subunit-containing GABAA receptors, but no efficacy at α1 subunit-containing GABAA receptors (α1-sparing compound). Its reinforcing effects were compared with the conventional benzodiazepines triazolam and lorazepam (non-selective GABAA receptor full modulators) as well as with a non-selective partial modulator, MRK-696 [7-cyclobutyl-6-(2-methyl-2H-1,2,4-triazol-2-ylmethoxy)-3-(2-flurophenyl)-1,2,4-triazolo(4,3-b)pyridazine; no selectivity but partial intrinsic activity]. The base forms of triazolam (0.0001—0.01 mg/kg/injection) and lorazepam (0.0003—0.03 mg/kg/injection), MRK-696 (0.001—0.03 mg/kg/injection) and L-838,417 (0.001—0.01 mg/kg/injection) were prepared in 100% propylene glycol and diluted using 25% propylene glycol and sterile water solutions. Triazolam and lorazepam were purchased commercially (Sigma-Aldrich, St Louis, MO; or Tocris-Cookson, Ellisville, MO). New remifentanil solutions were prepared every 7 days, and triazolam, lorazepam, MRK-696 and L-838,417 solutions were used for up to 30 days before a new solution and stock were prepared. All solutions were filtered with 200 μm filters before being administered intravenously. During self-administration sessions, all drugs were administered at an injection volume of 0.54 ml per injection. MRK-696 and L-838,417 were obtained from Merck, Sharp, and Dohme Research Laboratories (Harlow Essex, UK). All drugs were administered intravenously.

2.3. Self-Administration Training and Testing

Drug self-administration sessions occurred in each monkey’s home cage, which was custom-modified for these studies. In this apparatus, the self-administration panel is inserted into one side of the cage and prevents visual contact with other cages/monkeys on that side of the monkey. The monkeys do still have visual, auditory, and olfactory contact with others in the room via the front of their cage unit as well as the side opposite the self-administration panel. During self-administration sessions, access to the room was restricted to research staff. Sessions occurred Monday through Friday, sessions were not scheduled over weekends. All sessions started at 9h. Monkeys were given the opportunity to self-administer the opioid remifentanil (0.0003 mg/kg/injection) under a progressive-ratio (PR) schedule of reinforcement (Fischer et al., 2016).

At the beginning of each session, one white stimulus light above a response lever was illuminated (Med Associates, St. Albans, VT). Upon completion of a response requirement, the white light was extinguished and a red stimulus light was illuminated for 3 s, coinciding with a 3-s infusion. Each trial ended with either an injection or the expiration of a 30-min limited hold, and trials were separated by a 30-min timeout period. All sessions consisted of 5 components made up of 4 trials each. The response requirement remained constant for each of the 4 trials within a component, and doubled during each successive component. The session ended when a monkey self-administered a maximum of 20 injections or when the response requirement was not completed for two consecutive trials. The PR schedule consisted of a sequence of response requirements: either 60, 120, 240, 480, 960 (subject M-98-007) or 100, 200, 400, 800, 1600 (subjects M-1170 and M-1027) responses per injection. Once performance was stable under these conditions (no increasing or decreasing trend in the number of injections per session for three consecutive sessions), remifentanil or vehicle (saline) was made available on alternating days. Self-administration sessions alternating remifentanil and vehicle (saline) were conducted until a consistent difference between drug and saline self-administration was obtained (number of injections during saline sessions ≤ 35% of remifentanil sessions). The amount of time required for each subject to reach these criteria varied, with one subject requiring 3 months of training (subject M-1027), one subject requiring 4 months (subject M-98-007) and one subject requiring 5 months (subject M-1170). The test phase began immediately after reaching these criteria. Therefore, animals had 3, 4 or 5 months of remifentanil/saline self-administration experience before the test phase began.

Test sessions (T) were added to the alternating sequence of remifentanil (R) and saline (S) sessions according to the following sequence: RTRST or STSRT. During test (T) sessions, a dose of a GABAA receptor modulator (triazolam, lorazepam, MRK-696 or L-838,417) was made available in random order and alternating with its vehicle, although all doses of one drug/compound were completed prior to moving to another drug/compound. Doses of triazolam (0.0001-0.01 mg/kg/injection), lorazepam (0.0003-0.03 mg/kg/injection), MRK-696 (0.001-0.03 mg/kg/injection) and L-838,417 (0.001-0.01 mg/kg/injection) were chosen based on their ability to maintain self-administration in monkeys experienced in barbiturates, midazolam and/or cocaine (Rowlett et al., 2005; Shinday et al., 2013).

Based on previous research with response requirement manipulations with our PR schedule (e.g., Rowlett et al., 2002), drug taking (i.e., consumption) tends to decrease with higher response requirements (i.e., costs). Therefore, in order to investigate whether the reinforcing effects of MRK-696 or L-838,417 were obscured because the response requirements did not result in completion of responding (i.e., price was too high), another set of tests was conducted for which the initial response requirement (IRR) was manipulated. For these tests, when a drug/compound was self-administered above vehicle levels, the maximally effective unit dose (EDMax, i.e., the unit dose that engendered highest number of injections/session) was determined, and tests were repeated for that dose at half the original IRR (30 for IRR 60, subject M-98-007, and 50 for IRR 100, subjects M-1170 and M-1027). IRR manipulation tests began after all dose-effect curves had been determined. If a test compound was not self-administered above vehicle levels at the original IRR, we assessed the extent to which self-administration was evident at a lower IRR, using a dose known to engender the highest number of injections/session in animals with a history of barbiturate (Rowlett et al., 2005), benzodiazepine, or cocaine (Shinday et al., 2013) self-administration.

Finally, a dose-response function of remifentanil self-administration was established at the end of the study. A dose of a remifentanil (0.00003—0.0003 mg/kg/injection) was made available in random order and alternating with vehicle (saline), and introduced to the alternating sequence of remifentanil (0.0003 mg/kg/injection, maintenance dose) and saline (S) sessions as a test (T) session.

2.4. Data Analysis

For all test conditions (dose-effect curves and IRR manipulation tests), each dose was determined twice (once after a remifentanil session and once after a saline session) and averaged together for analysis. The number of injections/session and the last response requirement completed in a session (break point, BP) were determined for individual monkeys for each test drug/dose. When a test drug was self-administered above vehicle levels, the maximally effective unit dose of each drug (EDMax, i.e., the unit dose that engendered highest number of injections/session) and one-half log-step unit dose below (−0.5 EDMax) and above (+0.5 EDMax) the EDMax dose were identified. The injections/session data were analyzed by repeated measures analysis of variance (ANOVA) with dose as the factor. A dose of drug was etermined to be self-administered significantly above vehicle levels by comparing mean injections/session for each dose to the corresponding vehicle control value (Bonferroni t-test, alpha level equal to P<0.05).

The maximum BP (BPmax) was calculated as the highest BP, irrespective of dose, for each test drug that was self-administered significantly above vehicle levels. The BPmax measure provides an index of reinforcing strength that takes into account individual differences in peak BP values. Because monkeys had different IRRs, BPmaxdata for compounds that were self-administered significantly above vehicle levels were normalized based on the BPmax for remifentanil. BPmax data were analyzed by repeated measures ANOVA with drug/compound as the factor. Multiple comparisons of BPmax means were conducted using Bonferroni t-test (P<0.05).

3. Results

Under remifentanil maintenance conditions, the dose of 0.0003 mg/kg/injection maintained an average of 15 injections/session (SEM=0.2), whereas saline availability resulted in an average of 4.3 injections/session (SEM=0.75). No substantial changes in baseline (maintenance) responding occurred during the ~1.5 year duration of this study. A dose-response function of remifentanil self-administration established at the end of the study showed that the doses of 0.0001 and 0.0003 mg/kg/injection engendered number of injections/session significantly above saline levels in all 3 subjects (Figure 1) [F (3, 5) = 39.7, p<0.001], The individual subject BPmax’s for remifentanil were 1600 (M-1170), 800 (M-1027) and 480 (M-98-007).

Figure 1.

Figure 1.

Open in a new tab

Individual (a) and grouped (b) data for the dose-response function for the number of i.v. remifentanil injections in male rhesus monkeys (n=3) responding under a progressive-ratio schedule of reinforcement. Monkeys were trained and maintained on 0.0003 mg/kg/injection of remifentanil. Each remifentanil dose as well as saline test were determined twice (once after a remifentanil maintenance session and once after a saline session) and averaged together for analysis. Data are mean number of injections/session ± SEM. *p < 0.05 vs. saline (Sal) (Bonferroni t-tests).

Both triazolam and lorazepam were self-administered significantly above vehicle levels (Figure 2). The EDMax dose for triazolam varied between subjects, with the dose of 0.0003 mg/kg/injection being the EDMax dose for two subjects, and the dose of 0.003 mg/kg/injection being the EDMax for the third subject (Figure 2a). Both remifentanil alone and the EDMax dose of triazolam maintained mean number of injections/session that was significantly above vehicle (see Figure 2c [F (4, 8) = 28.7, p<0.0001], Bonferroni t-tests vs. vehicle, p<0.05). Similarly, for lorazepam the dose of 0.003 mg/kg/injection was the EDMax for two subjects, and the dose of 0.01 was the EDMax for the third subject (Figure 2b), with the EDMax dose maintaining mean number of injections/session significantly above vehicle levels (see Figure 2d, [F (4, 8) = 42.87, p<0.0001], Bonferroni t-tests vs. vehicle, p<0.05).

Figure 2.

Figure 2.

Open in a new tab

Self-administration of the conventional benzodiazepines triazolam and lorazepam (GABAA receptor full modulators) by male rhesus monkeys (subjects M-1170, M-1027 and M-98-007) trained under a progressive-ratio schedule of i.v. remifentanil (R, 0.0003 mg/kg/injection) injection. Self-Administration sessions alternating remifentanil and its vehicle (saline, S) were conducted until a consistent difference between drug and saline self-administration was obtained. Once self-administration was stable, test sessions with benzodiazepine compounds and their vehicle (V) were added to an alternating sequence of remifentanil and saline sessions. Individual (a, b) and grouped (c, d) data for triazolam and lorazepam self-administration. Grouped data are expressed as the unit dose that engendered highest number of injections/session (EDMax) and one-half log-step unit dose below (−0.5 EDMax) and above (+0.5 EDMax) the EDMax dose. Data are represented as mean number of injections/session ± SEM, out of a total of 20 injections available in a daily session. For panels a and b, data for R, S and V represent responding during the study of the drug, and filled symbols represent a meaningful difference (i.e., drug reinforcement) compared to responding under the V condition. Note that symbols obscure error bars in some instances. *p < 0.05 vs vehicle (V) (Bonferroni t-tests).

Contrary to the non-selective full modulators, the non-selective partial modulator MRK-696 did not maintain average injections/session above vehicle levels at any dose (Figures 3a, 3c) [F (3, 6) = 1.3, p>0.05], even at doses that were self-administered significantly above vehicle levels in midazolam-and cocaine-experienced rhesus monkeys (0.003–0.03 mg/kg/injection, see Shinday et al., 2013). Similarly, the α1-sparing compound L-838,417 did not maintain self-administration at any of the doses tested (Figures 3b, 3d) [F (3, 6) = 2.59, p>0.05], even at a dose that was self-administered significantly above vehicle levels in methohexital-experienced rhesus monkeys (0.01 mg/kg/injection, see Rowlett et al., 2005). Evaluation of BPmax values (% remifentanil BPmax) showed that triazolam had half (50%) the reinforcing effectiveness of remifentanil. Comparisons between the different GABAA receptor modulators revealed significant effects [F (3, 6) = 14.09, p<0.01] that were due to % BPmax values for MRK-696 and L-838,417, but not lorazepam, being lower than the % BPmax values obtained with triazolam (Figure 4).

Figure 3.

Figure 3.

Open in a new tab

Self-administration of MRK-696 (GABAA receptor partial modulator) and L-838,417 (partial intrinsic efficacy at α2, α3, and α5 subunit-containing GABAA receptors, but no efficacy at α1 subunit-containing GABAA receptors – α1-sparing compound) by male rhesus monkeys (subjects M-1170, M-1027 and M-98-007) trained under a progressive-ratio schedule of i.v. remifentanil (R, 0.0003 mg/kg/injection) injection. Animals were trained to discriminate remifentanil from saline (S), and test sessions with benzodiazepine compounds were added to an alternating sequence of remifentanil and saline sessions. Individual (a, b) and grouped (c, d) data for MRK-696 (0.001-0.01 mg/kg/injection) and L-838,417 (0.001-0.01 mg/kg/injection) self-administration. A higher dose of MRK-696 (0.03 mg/kg/injection) was tested for subject M-1170 to ensure no self-administration was observed even at a higher dose. Data are represented as mean number of injections/session ± SEM, out of a total of 20 injections available in a daily session. For panels a and b, data for V, R and S represent responding during the study of the drug, and filled symbols represent a meaningful difference (i.e., drug reinforcement) compared to responding under the V condition. Note that symbols obscure error bars in some instances.

Figure 4.

Figure 4.

Open in a new tab

Relative reinforcing effectiveness of GABAA receptor modulators in monkeys trained under a progressive-ratio schedule of remifentanil injection. Data are derived from break points, i.e., the highest response requirement completed in a session, and expressed in terms of BPmax, which is the highest break point obtained irrespective of dose. BPmax data for individual compounds were normalized based on the BPmax for remifentanil (% Remifentanil BPmax). The individual subject BPmax’s for remifentanil were 1600 (M-1170), 800 (M-1027) and 480 (M-98-007). *p < 0.05 vs % Remifentanil BPmax for triazolam (Bonferroni t-tests).

Because MRK-696 and L-838,417 were not self-administered above vehicle levels, IRR manipulation tests were conducted at the dose shown previously to engender the highest number of injections/session in animals experienced in benzodiazepine, barbiturate, and/or cocaine self-administration (dose of 0.01 mg/kg/injection for both drugs, Rowlett et al., 2005; Shinday et al., 2013). Similar results were obtained for the IRR manipulation tests, with triazolam and lorazepam, but not MRK-696 and L-838,417, being self-administered above vehicle levels when the IRR was reduced by half (Figure 5) [F (9, 18) = 18.75, p<0.0001], The number of injections/session did not change for any of the test drugs under IRR 30 or 50 compared to the original IRR conditions (IRR 60 or 100).

Figure 5.

Figure 5.

Open in a new tab

Self-administration of vehicle and selective or non-selective GABAA receptor modulators by male rhesus monkeys (n=3) trained under a progressive-ratio schedule of i.v. remifentanil (R, 0.0003 mg/kg/injection) injection. Comparison between tests under original initial response requirement (IRR) conditions (60 for subject M-98-007, 100 for subjects M-1170 and M-1027) vs tests with IRR manipulations, in which specific doses of each compound were tested under half the original IRR (30 for subject M-98-007, 100 for subjects M-1170 and M-1027). Triazolam and lorazepam were tested at the EDMax doses (see Figure 2). For MRK-696 and L-838,417, IRR manipulation tests were conducted at the dose known to engender the highest number of injections/session in animals trained under barbiturate, benzodiazepine and/or cocaine self-administration (Rowlett et al., 2005; Shinday et al., 2013) (0.01 mg/kg/injection for both drugs). Data are represented as mean number of injections/session ± SEM, out of a total of 20 injections available in a daily session. *p < 0.05 vs vehicle (Bonferroni t-tests).

4. Discussion

The present study shows that the full GABAA receptor modulators triazolam and lorazepam have reinforcing effects in rhesus monkeys with a history of self-administration with the opioid agonist remifentanil. The partial modulator MRK-696 and the α1-sparing partial modulator L-838,417, on the other hand, did not have reinforcing effects in remifentanil-experienced rhesus monkeys. Our findings suggest that, in contrast to monkeys with benzodiazepine, barbiturate, or cocaine self-administration histories, for a compound to have reinforcing effects in monkeys with a history of remifentanil self-administration, relatively high efficacy at one or more GABAA receptor subtypes is required.

Previous research from our group has extensively investigated GABAA receptor subtypes and the reinforcing effects of benzodiazepine-type drugs in benzodiazepine- and cocaine-experienced rhesus monkeys. We have demonstrated that both α1GABAA and α3GABAA receptor subtypes play a key role in benzodiazepine self-administration depending on the subjects’ past drug experiences. Specifically, we have shown that ligands with functional selectivity for α2/α3GABAA or α2/α3/α5GABAA receptors were self-administered in monkeys with a history of midazolam, but not cocaine, self-administration under a progressive-ratio schedule of reinforcement (Shinday et al., 2013). In other words, α1-sparing compounds were not self-administered by monkeys experienced in cocaine, consistent with the hypothesis that α1GABAA receptors have a critical role in the reinforcing effects of benzodiazepine-type compounds proposed by Tan et al. (2011) and Engin et al. (2018), albeit in monkeys with a history of cocaine self-administration.

More recently, we demonstrated that YT-III-31, a compound with functional selectivity at α3GABAA receptors, was self-administered by monkeys experienced in midazolam self-administration (Meng et al., 2019), suggesting that positive modulation of α3GABAA receptors is sufficient for a ligand to have benzodiazepine-like reinforcing effects. However, the relative reinforcing effectiveness (as measured by “BPmax”, the highest break point achieved, irrespective of dose) of YT-III-31 was significantly lower than conventional benzodiazepines, such as midazolam (Meng et al., 2019). Because of the robust self-administration observed with the α1GABAA modulator zolpidem (Griffiths et al., 1992; Rowlett et al., 2005) and the research implying α1GABAA mechanisms in benzodiazepine reward using transgenic mouse models (for review see Engin et al., 2018), these findings suggest that α1GABAA receptor efficacy determines the degree of reinforcing effectiveness of benzodiazepine-like ligands, whereas α3GABAA activity is sufficient for benzodiazepine ligands to have reinforcing effects in monkeys with a history of benzodiazepine self-administration. The present findings offer an additional complication, because activity at α1GABAA receptors was not predictive of self-administration under an opioid baseline.

Although under some conditions α1-sparing compounds have reinforcing effects, these conditions are limited (i.e., only happened in methohexital- and benzodiazepine-experienced monkeys, (Rowlett et al., 2005; Shinday et al., 2013). Moreover, α1-sparing compounds consistently have relatively lower reinforcing effectiveness that conventional benzodiazepines, as evident in the present study (BPmax results, see also Rowlett et al., 2005; Shinday et al., 2013; Meng et al., 2019). Overall, these findings suggest that these compounds may be relatively weak reinforcers compared with non-selective benzodiazepines. Strikingly, the partial modulator MRK-696 was not self-administered in remifentanil-experienced monkeys, even though it was a reinforcer in benzodiazepine- and cocaine-experienced animals (Shinday et al., 2013). At present, it is unclear why MRK-696 lacked reinforcing effects in the monkeys with a history of remifentanil self-administration. Irrespective of differences among the training and maintenance conditions, it appears that low efficacy GABAA modulators are weaker reinforcers, and in some cases lack reinforcing effects entirely. However, it is important to note that lack of self-administration does not necessarily mean that the compound lacks effects in the context of reinforcement. In this regard, in a two-lever choice procedure rhesus monkeys chose mixtures of L-838,417 plus cocaine over cocaine alone, even though L-838,417 lacked reinforcing effects when tested lone (Huskinson et al., 2019).

In the present study, rhesus monkeys were trained under a higher initial response requirement (IRR or 60 or 100) compared to our previous reports (IRR of 40; Rowlett et al., 2005; Shinday et al., 2013). A higher IRR was used in this study because it became difficult for remifentanil-experienced animals to maintain saline self-administration levels using our criteria for acceptable self-administration performance (i.e., monkeys responded more under the saline condition in the present study compared to previous studies from our laboratory using midazolam or cocaine as the training drug). Based on previous research with response requirement manipulations in general (Rowlett et al., 2002; for review see Bentzley et al., 2013), drug taking (i.e., consumption) tends to decrease with higher response requirements (i.e., costs). For behavior maintained under a progressive ratio schedule, drugs with higher relative reinforcing effectiveness, such as the opioid agonist alfentanil, are less susceptible to IRR changes than drugs with lower reinforcing effectiveness, such as the partial opioid agonist nalbuphine (Rowlett et al., 2002). In the present study, an IRR manipulation was conducted in which the peak dose (or predicted peak dose, when a compound did not have reinforcing effects) of each compound was tested at half the original IRR. Contrary to predictions, decreasing the IRR did not change our original findings: non-selective or α2/α3/α5GABAA-selective partial modulators did not function as reinforcers at lower IRRs. Therefore it appears unlikely that “hidden” reinforcing effects of MRK-696 or L-838,417 were obscured because the response requirements were too high across the progressive ratio sessions.

Although the conventional benzodiazepines largely prescribed as anxiety-reducing drugs are full non-selective GABAA receptor modulators, non-selective partial modulators have also been shown to have anxiety-reducing effects in rodent models (Cole and Rodgers, 1993; Griebel et al., 1996; Li et al., 2006). We have determined previously that the non-selective partial GABAA receptor modulators bretazenil and MRK-696 have anxiolytic-like effects in a rhesus monkey conflict model (unpublished data). Of note, the α1-sparing compound L-838,417 also had anxiolytic-like effects in rhesus monkeys (Rowlett et al., 2005). Given the alarming increase in the number of patients being co-prescribed opioids and benzodiazepines (Moore and Mattison 2017), our present and previous findings suggest that both non-selective as well as selective partial GABAA receptor modulators could have anxiety-reducing effects without exerting abuse potential, or showing a lower abuse liability, in opioid-taking patients. It is important to note, however, that studies are needed to investigate the extent to which partial modulators have anxiolytic-like effects in monkeys with a history of opioid self-administration. Also, Weed et al. (2017) previously reported that in a choice procedure, rhesus monkeys preferred i.v. injections of remifentanil when it was combined with the benzodiazepine midazolam over injections of remifentanil alone. These and other related findings (e.g., Preston et al. 1984) suggest that the abuse-related effects of opioids when combined with benzodiazepines may be more reinforcing than the single drugs, which may account in part for the prevalence of this type of polydrug abuse in patients with opioid use disorders. Therefore, future studies investigating the co-administration of opioids and non-selective or subtype-selective partial modulators in progressive ratio or choice procedures are warranted.

The neurobiological basis underlying the differential mechanisms of action of benzodiazepine-type compounds remains unknown. Because of their localization within the mesolimbic system, α1GABAA receptors have been proposed as the main mediators of the abuse-related effects of benzodiazepine drugs (Tan et al., 2011). In the ventral tegmental area (VTA), α1GABAA receptors are localized on inhibitory interneurons (Heikkinen et al, 2009; Tan et al, 2010). According to Tan et al. (2011), benzodiazepines would decrease firing of these interneurons via α1GABAA receptors, therefore disinhibiting dopamine neurons and leading to an increase in dopamine release in the nucleus accumbens—a neurobiological phenomenon commonly associated with drug self-administration. On the other hand, α1-sparing compounds would not activate the inhibitory interneurons, instead binding to and activating α3GABAA subtypes which are localized on dopamine neurons originating in the VTA (Heikkinen et al., 2009; Tan et al., 2010). Activation of α3GABAA receptors would be predicted to decrease dopamine release in the nucleus accumbens (Tan et al, 2011). However, this theory does not explain the results obtained in monkeys with a history of benzodiazepine self-administration, and also only partially explain the findings in cocaine-experienced monkeys (Rowlett et al., 2005; Shinday et al., 2013; Huskinson et al., 2019; Meng et al., 2019), suggesting a more complex “circuit pharmacology” for benzodiazepine reinforcement (cf. Engin et al., 2018). Our present results raise the possibility that yet another, separate circuit pharmacology that does not involve α1GABAA receptors may be involved in benzodiazepine effects with remifentanil-experienced monkeys. Alternatively, our findings may be best explained in terms of intrinsic efficacy: For a compound to have reinforcing effects in monkeys experienced in an opioid, relatively high efficacy at one or more subtypes is required (cf. Huskinson et al., 2019). Reconciliation of these two views and further understanding of the different baseline conditions awaits experimentation with additional pharmacological and molecular tools.

Highlights.

  • Full GABAA modulators have reinforcing effects in monkeys trained with remifentanil

  • A partial GABAA modulator was not self-administered by remifentanil-trained monkeys

  • An α1-sparing partial modulator was not self-administered by these monkeys

  • High efficacy GABAA modulators have reinforcing effects in opioid-trained monkeys

Acknowledgements

The authors thank Meagan Follett and Tanya Pareek for technical assistance, and Dr. Sally Huskinson for surgical training and guidance during this study.

Role of Funding Source

This work was supported by the National Institutes of Health [DA011792, DA011792-S1, and DA043204-02], The funding source had no role in the study design, in the collection, analysis and interpretation of data, in the writing of the report or in the decision to submit the article for publication.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of Interest

No conflict declared.

References

  1. Atack JR, 2011. GABAA receptor subtype-selective modulators. I. α2/α3-Selective agonists as nonsedating anxiolytics. Curr. Top. Med. Chem 11, 1176–1202. [DOI] [PubMed] [Google Scholar]
  2. Bentzley BS, Fender KM, Aston-Jones G, 2013. The behavioral economics of drug self-administration: a review and new analytical approach for within-session procedures. Psychopharmacology (Berl). 226, 113–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Cole JC, Rodgers RJ, 1993. An ethological analysis of the effects of chlordiazepoxide and bretazenil (Ro 16-6028) in the murine elevated plus-maze. Behav. Pharmacol 4, 573–580. [PubMed] [Google Scholar]
  4. Engin E, Benham RS, Rudolph U, 2018. An emerging circuit of pharmacology of GABAA receptors. Trends Pharmacol. Sci 39, 710–732. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Fischer BD, Platt DM, Rallapalli SK, Namjoshi OA, Cook JM, Rowlett JK, 2016. Antagonism of triazolam self-administration in rhesus monkeys responding under a progressive-ratio schedule: In vivo apparent pA2 analysis. Drug Alcohol Depend. 158, 22–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Griebel G, Sanger DJ, Perrault G, 1996. The use of the rat elevated plus-maze to discriminate between non-selective and BZ-1 (omega 1) selective, benzodiazepine receptor ligands. Psychopharmacology (Berl). 124, 245–54. [DOI] [PubMed] [Google Scholar]
  7. Griffiths RR, Sannerud CA, Ator NA, Brady JV, 1992. Zolpidem behavioral pharmacology in baboons: self-injection, discrimination, tolerance and withdrawal. J. Pharmacol. Exp. Ther 260, 1199–1208. [PubMed] [Google Scholar]
  8. Heikkinen AE, Moykkynen TP, Korpi ER, 2009. Long-lasting modulation of glutamatergic transmission in VTA dopamine neurons after a single dose of benzodiazepine agonists. Neuropsychopharmacology. 34, 290–8. [DOI] [PubMed] [Google Scholar]
  9. Huskinson SL, Freeman KB, Rowlett JK, 2019. Self-administration of benzodiazepine and cocaine combinations by male and female rhesus monkeys in a choice procedure: role of α1 subunit-containing GABAA receptors. Psychopharmacology (Berl). 236, 3271–3279. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Li X, Morrow D, Witkin JM, 2006. Decreases in nestlet shredding of mice by serotonin uptake inhibitors: comparison with marble burying. Life Sci. 78, 1933–9. [DOI] [PubMed] [Google Scholar]
  11. Meng Z, Berro LF, Sawyer EK, Riiedi-Bettschen D, Cook JE, Li G, Platt DM, Cook JM, Rowlett JK, 2019. Evaluation of the anti-conflict, reinforcing, and sedative effects of YT-III-31, a ligand functionally selective for α3 subunit-containing GABAA receptors. J. Psychopharmacol [in press], [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Möhler H, 2011. The rise of a new GABA pharmacology. Neuropharmacology. 60, 1042–1049. [DOI] [PubMed] [Google Scholar]
  13. Moore TJ, Mattison DR, 2017. Adult Utilization of Psychiatric Drugs and Differences by Sex, Age, and Race. JAMA Intern. Med 177, 274–275. [DOI] [PubMed] [Google Scholar]
  14. Paulozzi LJ, Strickler GK, Kreiner PW, Koris CM, 2015. Centers for Disease Control and Prevention (CDC). Controlled Substance Prescribing Patterns-Prescription Behavior Surveillance System, Eight States, 2013. MMWR Surveill. Summ 64, 1–14. [DOI] [PubMed] [Google Scholar]
  15. Platt DM, Carey G, Spealman RD, 2011. Models of neurological disease (substance abuse): Self administration in monkeys. Curr. Protoc. Pharmacol Chapter 10:Unit 10.5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Preston KL, Griffiths RR, Stitzer ML, Bigelow GE, Liebson IA, 1984. Diazepam and methadone interactions in methadone maintenance. Clin. Pharmacol. Ther 36, 534–41. [DOI] [PubMed] [Google Scholar]
  17. Rowlett JK, Lelas S, 2007. Comparison of zolpidem and midazolam self-administration under progressive-ratio schedules: consumer demand and labor supply analyses. Exp. Clin. Psychopharmacol 15, 328–337. [DOI] [PubMed] [Google Scholar]
  18. Rowlett JK, Platt DM, Lelas S, Atack JR, Dawson GR, 2005. Different GABAA receptor subtypes mediate the anxiolytic, abuse-related, and motor effects of benzodiazepine-like drugs in primates. Proc. Natl. Acad. Sci. USA 102, 915–920. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Rowlett JK, Rodefer JS, Spealman RD, 2002. Self-administration of cocaine, alfentanil, and nalbuphine under progressive-ratio schedules: consumer demand and labor supply analyses of relative reinforcing effectiveness. Exp. Clin. Psychopharmacol 10, 367–75. [DOI] [PubMed] [Google Scholar]
  20. Rowlett JK, Woolverton WL, 1997. Self-administration of cocaine and heroin combinations by rhesus monkeys responding under a progressive-ratio schedule. Psychopharmacology. 133, 363–371. [DOI] [PubMed] [Google Scholar]
  21. Schepis TS, Teter CJ, Simoni-Wastila L, McCabe SE, 2018. Prescription tranquilizer/sedative misuse prevalence and correlates across age cohorts in the US. Addict. Behav 87, 24–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Shinday NM, Sawyer EK, Fischer BD, Platt DM, Licata SC, Atack JR, Dawson GR, Reynolds DS, Rowlett JK, 2013. Reinforcing effects of compounds lacking intrinsic efficacy at α1 subunit-containing GABAA receptor subtypes in midazolam- but not cocaine-experienced rhesus monkeys. Neuropsychopharmacology. 38, 1006–1014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Stroumpos C, Manolaraki M, Paspatis GA, 2010. Remifentanil, a different opioid: Potential clinical applications and safety aspects. Exp. Opin. Drug Safety 9, 355–364. [DOI] [PubMed] [Google Scholar]
  24. Svitek J, Heberlein A, Bleich S, Wiltfang J, Kornhuber J, Hillemacher T, 2008. Extensive craving in high dose zolpidem dependency. Prog. Neuropsychopharmacol. Biol. Psychiatry 32, 591–592. [DOI] [PubMed] [Google Scholar]
  25. Tan KR, Brown M, Labouebe G, Yvon C, Creton C, Fritschy M, Rudolph U, Luscher C, 2010. Neural bases for addictive properties of benzodiazepins. Nature. 463, 769–774. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Tan KR, Rudolph U, Luscher C, 2011. Hooked on benzodiazepines: GABAA receptor subtypes and addiction. Trends. Neurosci 34, 188–197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Weed PF, France CP, Gerak LR, 2017. Preference for an opioid/benzodiazepine mixture over an opioid along using a concurrent choice procedure in rhesus monkeys. J. Pharmacol. Exp. Ther 362, 59–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Woolverton WL, Wang Z, Vasterling T, Carroll FI, Tallarida R, 2008. Self-administration of drug mixtures by monkeys: Combining drugs with comparable mechanisms of action. Psychopharmacology. 196, 575–582. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES