Respiratory depressant effects of fentanyl analogs are opioid receptor-mediated

Abstract

Opioid-related fatalities involving synthetic opioids have reached unprecedented levels. This study evaluated the respiratory depressant effects of seven fentanyl analogs that have either emerged in the illicit drug supply or been identified in toxicological analyses following fatal or non-fatal intoxications. Adult male Swiss Webster mice were administered fentanyl analogs (isobutyrylfentanyl, crotonylfentanyl, para-methoxyfentanyl, para-methoxybutyrylfentanyl, 3-furanylfentanyl, thiophenefentanyl, and benzodioxolefentanyl) and their effects on minute volume as compared to mu-opioid receptor (MOR) agonist standards (fentanyl, morphine, and buprenorphine) were measured using whole body plethysmography (WBP). All drugs elicited significant (p ≤ 0.05) hypoventilation relative to vehicle for at least one dose tested: morphine (1, 3.2, 10, 32 mg/kg), buprenorphine, (0.032, 0.1, 0.32, 1, 3.2 mg/kg), fentanyl (0.0032, 0.01, 0.032, 0.1, 1, 32 mg/kg), isobutyrylfentanyl (0.1, 0.32, 1, 3.2, 10 mg/kg), crotonylfentanyl (0.1, 0.32, 1, 3.2, 10 mg/kg), para-methoxyfentanyl (0.1, 0.32, 1, 3.2, 10 mg/kg), para-methoxybutyrylfentanyl (0.32, 1, 3.2, 10 mg/kg), 3-furanylfentanyl (0.1, 0.32, 1, 3.2, 10 mg/kg), thiophenefentanyl (1, 3.2, 10, 32, 100 mg/kg), and benzodioxolefentanyl (3.2, 10, 32, 100 mg/kg). The ED₅₀ values for hypoventilation showed a rank order of potency as follows: fentanyl (ED₅₀ = 0.96 mg/kg) > 3-furanylfentanyl (ED₅₀ = 2.60 mg/kg) > crotonylfentanyl (ED₅₀ = 2.72 mg/kg) > para-methoxyfentanyl (ED₅₀ = 3.31 mg/kg) > buprenorphine (ED₅₀ = 10.8 mg/kg) > isobutyrylfentanyl (ED₅₀ = 13.5 mg/kg) > para-methoxybutyrylfentanyl (ED₅₀ = 16.1 mg/kg) > thiophenefentanyl (ED₅₀ = 18.0 mg/kg) > morphine (ED₅₀ = 55.3 mg/kg) > benzodioxolefentanyl (ED₅₀ = 10,168 mg/kg). A naloxone pretreatment (10 mg/kg) attenuated the hypoventilatory effects of all drugs. These results establish that the respiratory depressant effects of these fentanyl analogs are at least in part mediated by the MOR.

1. Introduction

Synthetic opioids, including fentanyl and its FDA-approved analogs, are known to produce their effects by activating the mu-opioid receptor (MOR), and can elicit life-threatening respiratory depression [1–8]. Medical use of fentanyl and its FDA-approved analogs (alfentanil, remifentanil, and sufentanil) is common when analgesic and anesthetic agents are indicated [9–12]. Abuse of fentanyl and structurally related-substances can produce euphoria and attenuate withdrawal symptoms in physically dependent persons with opioid use disorder (OUD) [13,14]. Fatal and non-fatal intoxications involving fentanyl and its analogs have steadily increased in recent years [15–23]. From 2013 to 2019, there was a >10-fold increase in deaths in the United States involving synthetic opioids other than methadone (3,105–36,359) [24]. In 2019, 72.9% of the 49,860 opioid-related deaths in the United States involved synthetic narcotics like fentanyl and its analogs [24]. Given the limited or non-existent in vivo data for fentanyl-related substances emerging in the illicit drug supply and their increasing involvement in drug overdose deaths, this study sought to evaluate the respiratory depressant effects of fentanyl and its analogs and their mechanisms of action.

Several studies evaluated fentanyl analogs for their in vitro affinity, efficacy, and functional selectivity at the MOR and demonstrated substantial variability in G-protein vs. beta-arrestin recruitment, even amongst drugs with similar binding affinity for the MOR [25–28]. Our laboratory group previously reported the in vivo pharmacological effects of fourteen emerging fentanyl-related substances in mice including isobutyrylfentanyl, crotonylfentanyl, para-methoxyfentanyl, para-methoxybutyrylfentanyl, 3-furanylfentanyl, thiophenefentanyl, benzodioxolefentanyl, beta-methylfentanyl, para-methylfentanyl, para-methoxyfentanyl, fentanyl carbamate, 3-furanylfentanyl, phenylfentanyl, and beta^′-phenylfentanyl [29,30]. We demonstrated that these fentanyl-related substances elicit dose-dependent antinociception, but only a subset elicited dose-dependent hyperlocomotion. We also established that MOR-selective antagonist pretreatments resulted in significant rightward shifts in antinociceptive dose effect curves indicating that the effects were at least partly MOR-mediated. Other studies examined fentanyl-related substances for their therapeutic viability in vivo following Janssen Pharmaceutica’s initial exploration of the 4-anilidopiperidine class and identified MOR agonists with minimal respiratory depression [31–33], pH-dependent MOR agonists [34], MOR agonists with weak beta-arrestin recruitment [35], and even MOR antagonists [36,37]. Here we report that fentanyl-related substances will elicit respiratory depression in mice and that their effects are at least in part mediated by the MOR. The respiratory depressant effects of these fentanyl analogs had not been previously characterized and their relative potencies, efficacies, and mechanisms of action were previously unknown.

This study was designed to test two hypotheses related to the respiratory depressant effects of fentanyl-related substances. First, that fentanyl-related substances will elicit hypoventilation in mice similar to prototypical MOR agonists. Second, that recruitment of MOR activity is sufficient to establish hypoventilation in mice. To evaluate these hypotheses, seven representative fentanyl-related substances and three MOR agonist standards were tested to determine if they would elicit prototypical opioid-like effects including hypoventilation, as measured by decreases in minute volume in ventilation tests. To further investigate the opioid-like mechanism responsible for these effects, the ability of a relatively MOR-selective antagonist, naloxone, was administered as a pretreatment to attenuate hypoventilation. The MOR agonist standards tested were morphine, buprenorphine, and fentanyl. The fentanyl-related substances tested were isobutyrylfentanyl, crotonylfentanyl, para-methoxyfentanyl, para-methoxybutyrylfentanyl, 3-furanylfentanyl, thiophenefentanyl, and benzodioxolefentanyl. These substances were selected as representative for their chemical and pharmacological diversity and because of their recent emergence in the illicit drug supply. These substances were deemed imminent hazards to public safety by drug chemistry laboratories, toxicology laboratories, medical examiners’ and coroners’ offices, and law enforcement. Consequently, several of these fentanyls were placed into the list of Schedule I drugs as defined by the Controlled Substances Act, the most restricted drug class in the United States [29,30,38–46]. This study clarifies whether these fentanyl analogs, which were identified in the illicit drug supply and have only limited or no existing in vivo evaluations previously reported, display significant hypoventilatory effects and whether those effects are MOR-mediated.

2. Materials and methods

2.1. Drugs

(1) Morphine, (4R,4aR,7S,7aR,12bS)-3-methyl-2,3,4,4a,7,7a-hexahydro-1H-4,12-methanobenzofuro[3,2-e]isoquinoline-7,9-diol sulfate pentahydrate, item no. 9300–001, was provided by the National Institute on Drug Abuse (Bethesda, MD, USA) Drug Supply Program. (2) Buprenorphine, (4R,4aS,6R,7R,7aR,12bS)-3-(cyclopropylmethyl)-6-(2-hydroxy-3,3-dimethylbutan-2-yl)-7-methoxy-1,2,3,4,5,6,7,7a-octahydro-4a,7-ethano-4,12-methanobenzofuro[3,2-e]isoquinolin-9-ol hydrochloride, item no. B1637, was obtained from Spectrum Chemical Manufacturing Corporation (New Brunswick, NJ, USA). (3) Fentanyl, N-(1-phenethylpiperidin-4-yl)-N-phenylpropionamide citrate, item no. 22659, and fentanyl-related substances: (4) isobutyrylfentanyl, N-(1-phenethylpiperidin-4-yl)-N-phenylisobutyramide hydrochloride, item no. 18584; (5) crotonylfentanyl, (E)-N-(1-phenethylpiperidin-4-yl)-N-phenylbut-2-enamide, item no. 22801; (6) para-methoxyfentanyl, N-(4-methoxyphenyl)-N-(1-phenethylpiperidin-4-yl)propionamide hydrochloride, item no. 20035; (7) para-methoxybutyrylfentanyl, N-(4-methoxyphenyl)-N-(1-phenethylpiperidin-4-yl)butyramide hydrochloride, item no. 18089; (8) 3-furanylfentanyl, N-(1-phenethylpiperidin-4-yl)-N-phenylfuran-3-carboxamide hydrochloride, item no. 21213; (9) thiophenefentanyl, N-(1-phenethylpiperidin-4-yl)-N-phenylthiophene-2-carboxamide hydrochloride, item no. 22802; (10) benzodioxolefentanyl, N-(1-phenethylpiperidin-4-yl)-N-phenylbenzo[d][1,3] dioxole-5-carboxamide, item no. 20858; and (11) naloxone, (4R,4aS,7aR,12bS)-3-allyl-4a,9-dihydroxy-2,3,4,4a,5,6-hexahydro-1H-4,12-methanobenzofuro[3,2-e]isoquinolin-7(7aH)-one hydrochloride, item no. 15594, were provided by Cayman Chemical Company (Ann Arbor, MI, USA). All drugs were dry powders and either dissolved in sterile saline or suspended in 0.5% methylcellulose, item no. M0430, (Sigma-Aldrich, St. Louis, MO, USA) in deionized water and were injected subcutaneously (SC) in a volume equivalent to 10 ml/kg body weight. Structures for substances 1–10 are shown in Fig. 1. Spectral data and analyses for drugs provided by Cayman Chemical Company are available at https://osf.io/57489/.

Fig. 1.

Chemical structures of (1) morphine, (2) buprenorphine, (3) fentanyl, (4) isobutyrylfentanyl, (5) crotonylfentanyl, (6) para-methoxyfentanyl, (7) para-methoxybutyrylfentanyl, (8) 3-furanylfentanyl, (9) thiophenefentanyl, and (10) benzodioxolefentanyl.

2.2. Subjects

Adult male Swiss Webster mice (N = 264; Crl:CFW(SW), Charles River Laboratories, Raleigh, NC, USA) weighing ~ 25–50 g at the time of testing were housed four subjects per cage in facilities accredited by the Association for Assessment and Accreditation of Laboratory Animal Care. Subjects had ad libitum access to food (Teklad 7012 Rodent Diet; Envigo, Madison, WI, USA) and tap water. The vivarium was maintained at 22 °C ± 2 °C and 50% ± 5% humidity, with lights set to a 12-hour light/dark cycle (lights off at 1000) and testing occurred during the dark cycle to ensure subjects were active (i.e., not sleeping). Subjects were typically tested on weekdays between the hours of 1100 and 1700. Subjects were acclimated to the vivarium for at least one week before the commencement of studies and were drug-naïve before testing. All procedures were carried out in accordance with the Guide for Care and Use of Laboratory Animals as adopted and promulgated by the National Institutes of Health [47]. The experimental protocol was approved by the Institutional Animal Care and Use Committee at Virginia Commonwealth University.

2.3. Measurement of ventilation

Ventilatory parameters were captured for freely moving mice in individual test chambers using whole-body plethysmography (FinePointe WBP Chamber with Halcyon Technology, Data Sciences International, St. Paul, MN, USA). The test chambers (0.5 L volume with adjustable 0.5 L/min room air bias flow) were housed in a room illuminated by custom-built, 660 nM-emitting T8-style ceiling-mounted light tubes each with 120, 0.2-watt Epistar 2835 SMD LEDs (Benwei Electronics Co., Ltd., Shenzhen, China), a wave-length with limited visibility to mice [48], to enable maintenance of the dark cycle during testing. The test chambers were continuously supplied 5% CO₂, 21% O₂, and balance N₂ (AirGas, Radnor, PA, USA) to minimize variability of baseline ventilatory activity, and to increase the sensitivity and capacity of the assay to detect meaningful differences in ventilation as determined in preliminary tests as well as reported and used by others [49–54]. Subjects were tested no earlier than at least 1 h after the start of the dark phase to further enhance the capacity to detect perturbations to ventilatory parameters [55]. Respiratory rate, tidal volume, and minute volume as a function of dose and time were recorded using software (FinePointe Software Research Suite; Data Sciences International, St. Paul, MN, USA).

2.3.1. Acute dosing

Subjects (n = 8 mice per dose) received SC injections of vehicle (saline for fentanyl) and were placed into the test chambers. After 20 min of recording, subjects were removed from the test chambers, administered a dose of fentanyl (0.01, 0.1, 0.32, 1, 3.2, or 10 mg/kg) and then immediately returned to the chambers for a minimum of 10 min. Subjects in this cohort were habituated in test chambers for 30 min under ambient air conditions 24 h prior to testing.

2.3.2. Cumulative dosing

Subjects (n = 8 mice per drug) were administered two SC injections—(1) the vehicle for the test drug, and (2) either naloxone (10 mg/kg), or saline, naloxone’s vehicle—and were placed into the test chambers. After 10 min of recording, subjects were removed from the test chambers, administered the first dose of test drug (e.g., 0.001 mg/kg fentanyl), and then immediately returned to the chambers for another 10-min recording period. The process of administering the next dose of test drug proceeded with additional acute administrations within-subjects (continuing the example, 0.0022, 0.0068, and 0.022 mg/kg fentanyl) to yield cumulative fentanyl doses of 0 (vehicle), 0.001, 0.0032, 0.01 and 0.032 mg/kg. In separate experiments, drug-naïve subjects were tested to evaluate the ability of a relatively selective MOR-antagonist to attenuate the respiratory-depressant effects of each drug. Naloxone (10 mg/kg) was co-administered with the respective drug vehicle, then testing proceeded as previously described. Initial proposed dose ranges and route of administration were based on previous antinociceptive and locomotor activity tests conducted in our laboratory and were expected to include inactive doses to doses that elicit significant decreases in minute volume [29,30].

2.4. Data analysis

The primary dependent measures expressed as a percentage of vehicle control were as follows: (1) respiratory rate [defined as the number of breaths in one minute (BPM)], (2) tidal volume [defined as the lung volume representing the normal volume of air displaced between inhalation and exhalation], and (3) minute volume [defined as the volume of air inspired or expired in a minute]. Data from the last 3 min of each 10-min period were binned for analysis. Each dependent measure was analyzed using a one-way repeated measures ANOVA with Geisser-Greenhouse correction (Gaussian distribution of residuals assumed; sphericity not assumed). A significant ANOVA was followed by post-hoc analyses using Bonferroni correction, with individual variances computed for each comparison. In addition, ED₅₀ values with 95% confidence intervals were determined for each drug to produce decreases in ventilation using a non-linear regression model with the following equation: [Inhibitor] vs. normalized response – Variable slope, Y = 100/(1 + (IC50/X)ĤillSlope). Constraints were applied to all models such that IC₅₀ values were presumed to be greater than zero. This model does not assume that the standard slope is equal to one and therefore produces a curve that most closely fits the shape of the dataset allowing for precise interpolation (or extrapolation based on the slope of the dose–effect curve) of ED₅₀ values for hypoventilation. Additionally, the hypoventilatory potency of each fentanyl analog was compared to its antinociceptive potency as determined in previous studies by our laboratory group to compute a measure of relative safety, a protective index (PI = TD₅₀/ED₅₀) to model therapeutic performance [29,30]. Results from the highest cumulative dose tested for each drug with a saline (SAL) or 10 mg/kg naloxone (NLX) pretreatment were analyzed with multiple unpaired two-tailed t-tests with Welch correction (Gaussian distribution assumed; sphericity not assumed). G*Power 3.1.9.7 was used to a priori calculate the sample sizes required to detect statistically significant differences for a large effect (F-test for ANOVA: 0.40; t-test for means: 0.80) given alpha (0.05) and power (0.8) for all tests [56–58]. A sample size of n = 8 subjects per dose–effect curve was determined to be sufficient. Comparisons were considered statistically significant if p ≤ 0.05. Analyses were performed with software (GraphPad Prism 9.2.0 (332) for Microsoft Windows 10 ×64; GraphPad Software, San Diego, CA, USA).

3. Results

3.1. Results from ventilation tests

Fig. 2 shows ventilation test results (minute volume expressed as % control) for fentanyl as a function of dose in acute dosing (unfilled circles) and cumulative dosing (filled squares) experiments. The ED₅₀ values for fentanyl-induced hypoventilation in acute dosing (between-subjects design) and cumulative dosing (within-subjects design) tests were 1.02 mg/kg (95% CI: 0.57–1.95) and 0.96 mg/kg (95% CI: 0.50–2.24) respectively and not significantly different when compared using the extra-sum-of-squares F-test [F(1, 124) = 0.014, p = 0.9076] and therefore could be modeled by a single function with an ED₅₀ of 1.00 mg/kg (95% CI: 0.65 to 1.60).

Fig. 2.

Results for fentanyl in acute vs. cumulative dose ventilation tests. Symbols indicate mean minute volume as a percent of control (± SEM) for fentanyl in between-subjects, acute dosing tests (unfilled circles) or within-subjects, cumulative dosing tests (filled squares) for n = 8–16 mice per dose.

Fig. 3 shows ventilation test results (respiratory rate, tidal volume, and minute volume expressed as % control) for morphine, buprenorphine, and fentanyl as a function of dose with a saline (filled circles) or 10 mg/kg naloxone (unfilled circles) pretreatment. Morphine, with a saline pretreatment, elicited significant decreases in respiratory rate [1, 3.2, 10, 32 mg/kg; F(1.846, 12.92) = 11.70, p = 0.0015] and minute volume [1, 3.2, 10, 32 mg/kg; F(1.644, 11.51) = 8.706, p = 0.0066], but not tidal volume [F(2.208, 15.45) = 0.5561, p = 0.6011]. Morphine, with a naloxone pretreatment, elicited a significant decrease in minute volume at a dose of 32 mg/kg [F(1.871, 13.10) = 4.552, p = 0.0334], but not respiratory rate [F(1.681, 11.77) = 3.231, p = 0.0823] or tidal volume [F(1.851, 12.96) = 0.7795, p = 0.4697]. Buprenorphine, with a saline pretreatment, elicited significant decreases in respiratory rate [0.032, 0.1, 0.32, 1, 3.2 mg/kg; F(2.713, 18.99) = 20.95, p < 0.0001] and minute volume [0.032, 0.1, 0.32, 1, and 3.2 mg/kg, F(1.741, 12.19) = 13.68, p = 0.0010], but not tidal volume [F(2.153, 15.07) = 3.105, p = 0.0713]. Buprenorphine, with a naloxone pretreatment, did not elicit significant changes in respiratory rate [F(2.214, 15.50) = 3.259, p = 0.0619], tidal volume [F(1.692, 11.85) = 2.248, p = 0.1529], or minute volume [F(2.414, 16.89) = 2.658, p = 0.0912]. Fentanyl, with a saline pretreatment, elicited significant decreases in respiratory rate at all doses tested [F(3.030, 21.21) = 39.53, p < 0.0001] as well as significant decreases in minute volume following the administration of all doses except for 0.001 mg/kg [F(3.724, 26.06) = 35.31, p < 0.0001], but did not elicit significant changes in tidal volume [F(2.586, 18.10) = 1.170, p = 0.3433]. Fentanyl, with a naloxone pretreatment, again elicited significant decreases in respiratory rate at all doses tested [F(4.049, 28.34) = 5.982, p = 0.0012] as well as significant decreases in minute volume at doses of 0.01, 0.032, 0.1, 1, and 3.2 mg/kg [F(3.771, 26.40) = 3.905, p = 0.0140], and left tidal volume unaffected [F(1.604, 11.23) = 1.060, p = 0.3628]. In ventilation tests with a saline (SAL) or naloxone (NLX) pretreatment, mean baseline minute volume values (± SEM) in ml/min for positive controls were as follows: 0.01–0.32 mg/kg morphine (SAL: 167 ± 15; NLX: 165 ± 13), 1–32 mg/kg morphine (SAL: 157 ± 7; NLX: 174 ± 4), 0.001–0.032 mg/kg buprenorphine (SAL: 188 ± 17; NLX: 154 ± 5), 0.1–3.2 mg/kg buprenorphine (SAL: 193 ± 7; NLX: 152 ± 6), 0.001–0.032 mg/kg fentanyl (SAL: 176 ± 11; NLX: 178 ± 11), and 0.1–3.2 mg/kg fentanyl (SAL: 157 ± 8; NLX: 161 ± 10).

Fig. 3.

Results for positive controls in cumulative dose ventilation tests. Symbols indicate respiratory rate, tidal volume, or minute volume as a percentage of control (± SEM) for (1) morphine [MOR], (2) buprenorphine [BUP], and (3) fentanyl [FEN] with a saline [SAL] (filled circles) or 10 mg/kg naloxone [NLX] (unfilled circles) pretreatment for n = 8–16 mice per dose. Significant differences between a drug’s dose and within-subject vehicle condition are indicated by asterisks: *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.

Fig. 4 shows ventilation test results (minute volume expressed as % control) for fentanyl-related substances as a function of dose with a saline (filled circles) or 10 mg/kg naloxone (unfilled circles) pretreatment. With a saline pretreatment, all fentanyl-related substances elicited significant decreases in minute volume relative to vehicle control at least at one dose tested: isobutyrylfentanyl [0.1, 0.32, 1, 3.2, 10 mg/kg; F (1.789, 12.52) = 19.21, p = 0.0002], crotonylfentanyl [0.1, 0.32, 1, 3.2, 10 mg/kg; F(2.214, 15.50) = 136.5, p < 0.0001], para-methoxyfentanyl [0.1, 0.32, 1, 3.2, 10 mg/kg; F(2.628, 18.40) = 195.8, p < 0.0001], para-methoxybutyrylfentanyl [0.32, 1, 3.2, 10 mg/kg; F(2.670, 18.69) = 65.16, p < 0.0001], 3-furanylfentanyl [0.1, 0.32, 1, 3.2, 10 mg/kg; F (2.390, 16.73) = 72.85, p < 0.0001], thiophenefentanyl [1, 3.2, 10, 32, 100 mg/kg; F(2.246, 15.73) = 89.62, p < 0.0001], and benzodioxolefentanyl [3.2, 10, 32, 100 mg/kg; F(3.166, 22.16) = 30.22, p < 0.0001]. With a naloxone pretreatment, fentanyl-related substances elicited significant (p ≤ 0.05) decreases in minute volume at the following doses: isobutyrylfentanyl [3.2 mg/kg; F(2.139, 14.97) = 4.553, p = 0.0267], crotonylfentanyl [1, 3.2 mg/kg; F(2.942, 20.59) = 5.606, p = 0.0059], para-methoxyfentanyl [0.32, 1, 3.2, 10 mg/kg; F (2.480, 17.36) = 18.25, p < 0.0001], para-methoxybutyrylfentanyl [3.2, 10 mg/kg; F(3.088, 21.62) = 6.616, p = 0.0023], 3-furanylfentanyl [0.32, 3.2, 10 mg/kg; F(1.931, 13.52) = 11.24, p = 0.0014], thiophenefentanyl [32, 100 mg/kg; F(1.859, 13.02) = 9.237, p = 0.0036], and benzodioxolefentanyl [3.2, 10, 32, 100 mg/kg; F(3.209, 22.47) = 8.178, p = 0.0006]. In ventilation tests with a saline (SAL) or naloxone (NLX) pretreatment, mean baseline minute volume values (± SEM) in ml/min for fentanyl-related substances were as follows: isobutyrylfentanyl (SAL: 181 ± 8; NLX: 176 ± 11), crotonylfentanyl (SAL: 178 ± 7; NLX: 165 ± 7), para-methoxyfentanyl (SAL: 214 ± 7; NLX: 195 ± 9), para-methoxybutyrylfentanyl (SAL: 194 ± 14; NLX: 195 ± 6), 3-furanylfentanyl (SAL: 197 ± 5; NLX: 168 ± 7), thiophenefentanyl (SAL: 170 ± 6; NLX: 161 ± 6), and benzodioxolefentanyl (SAL: 185 ± 6; NLX: 148 ± 7).

Fig. 4.

Results for fentanyl analogs in cumulative dose ventilation tests. Symbols indicate mean minute volume as a percentage of control (± SEM) for (4) isobutyrylfentanyl [IB-FEN], (5) crotonylfentanyl [CTO-FEN], (6) para-methoxyfentanyl [PMO-FEN], (7) para-methoxybutyrylfentanyl [PMOB-FEN], (8) 3-furanylfentanyl [3FU-FEN], (9) thiophenefentanyl [SFU-FEN], and (10) benzodioxolefentanyl [BDX-FEN] with a saline [SAL] or vehicle [VEH] (filled circles) or 10 mg/kg naloxone [NLX] (unfilled circles) pretreatment for n = 8 mice per dose. Significant differences between a drug’s dose and within-subject vehicle condition are indicated by asterisks: *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.

With a saline pretreatment, fentanyl-related substances elicited significant decreases in respiratory rate at all doses tested: isobutyrylfentanyl [F(1.911, 13.38) = 67.74, p < 0.0001], crotonylfentanyl [F(2.193, 15.35) = 100.5, p < 0.0001], para-methoxyfentanyl [F (2.686, 18.80) = 148.7, p < 0.0001], para-methoxybutyrylfentanyl [F(2.499, 17.50) = 54.70, p < 0.0001], 3-furanylfentanyl [F(2.176, 15.23) = 84.08, p < 0.0001], thiophenefentanyl [F(1.786, 12.50) = 86.54, p < 0.0001], and benzodioxolefentanyl [F(3.424, 23.97) = 31.05, p < 0.0001]. With a saline pretreatment, crotonylfentanyl [1, 3.2, 10 mg/kg; F(2.304, 16.13) = 25.63, p < 0.0001] and 3-furanylfentanyl [3.2, 10 mg/kg; F(2.555, 17.89) = 14.04, p < 0.0001] significantly decreased tidal volume. In contrast, benzodioxolefentanyl [1, 3.2, 10, 32, 100 mg/kg; F(2.987, 20.91) = 10.85, p = 0.0002] significantly increased tidal volume. Several drugs did not show significant differences in tidal volume according to post-hoc analyses with Bonferroni correction (even though the overall ANOVA rejected the null hypothesis of equal group means) including isobutyrylfentanyl [F(1.755, 12.28) = 4.191, p = 0.0450], para-methoxyfentanyl [F(3.000, 21.00) = 7.191, p = 0.0017], para-methoxybutyrylfentanyl [F(2.778, 19.44) = 13.93, p < 0.0001], and thiophenefentanyl [F(2.180, 15.26) = 11.06, p = 0.0009].

Table 1 shows the maximum suppression as a percentage of control (the smaller the percentage, the greater the suppression; i.e. 100% = zero suppression, 0% = complete suppression), potency estimates for hypoventilation (ED₅₀ values), and potency ratios to both morphine and fentanyl for each drug. The %E_max values (± SEM) for hypoventilation showed a rank order of efficacy as follows: para-methoxyfentanyl (41.5 ± 2.53) > 3-furanylfentanyl (42.9 ± 3.34) > thiophenefentanyl (43.5 ± 2.38) > crotonylfentanyl (44.9 ± 2.30) > fentanyl (51.6 ± 4.73) > morphine (56.6 ± 5.42) > isobutyrylfentanyl (58.8 ± 5.30) > para-methoxybutyrylfentanyl (59.5 ± 2.56) > buprenorphine (60.5 ± 3.65) > benzodioxolefentanyl (76.5 ± 1.98). The ED₅₀ values [95% CI] for hypoventilation showed a rank order of potency as follows: fentanyl (ED₅₀ = 0.96 mg/kg [0.50–2.24]) > 3-furanylfentanyl (ED₅₀ = 2.60 mg/kg [1.80–3.94]) > crotonylfentanyl (ED₅₀ = 2.72 mg/kg [1.99–3.88]) > para-methoxyfentanyl (ED₅₀ = 3.31 mg/kg [2.52–4.52]) > buprenorphine (ED₅₀ = 10.8 mg/kg [3.6–70.8]) > isobutyrylfentanyl (ED₅₀ = 13.5 mg/kg [6.28–46.8]) > para-methoxybutyrylfentanyl (ED₅₀ = 16.1 mg/kg [10.9–27.1]) > thiophenefentanyl (ED₅₀ = 18.0 mg/kg [10.8–32.9]) > morphine (ED₅₀ = 55.3 mg/kg [19.3–393]) > benzodioxolefentanyl (ED₅₀ = 10,168 mg/kg [1,790– 276,521]).

Table 1.

Results from cumulative dose ventilation tests.

#	Drug	%Emax ± SEM	ED50 mg/kg (95% CI)	Potency ratio to morphine	Potency ratio to fentanyl
1	Morphine	56.6 ± 5.42	55.3 (19.3–393)	1.00	0.02
2	Buprenorphine	60.5 ± 3.65	10.8 (3.6–70.8)	5.10	0.09
3	Fentanyl	51.6 ± 4.73	0.96 (0.50–2.24)	57.6	1.00
4	Isobutyrylfentanyl	58.8 ± 5.30	13.5 (6.28–46.8)	4.10	0.07
5	Crotonylfentanyl	44.9 ± 2.30	2.72 (1.99–3.88)	20.3	0.35
6	Para-methoxyfentanyl	41.5 ± 2.53	3.31 (2.52–4.52)	16.7	0.29
7	Para-methoxybutyrylfentanyl	59.5 ± 2.56	16.1 (10.9–27.1)	3.44	0.06
8	3-Furanylfentanyl	42.9 ± 3.34	2.60 (1.80–3.94)	21.4	0.37
9	Thiophenefentanyl	43.5 ± 2.38	18.0 (10.8–32.9)	3.08	0.05
10	Benzodioxolefentanyl	76.5 ± 1.98	10,168 (1,790–276,521)	0.0054	0.000094

Efficacy estimates are expressed as %E_max (maximum suppression as percentage of control; the smaller the percentage, the greater the suppression, i.e., 100% = zero suppression, 0% = complete suppression).

Potency estimates are expressed as ED₅₀ (mg/kg) values for drug alone, drug potency ratio to morphine, and drug potency ratio to fentanyl.

Data are mean ± SEM or 95% confidence intervals for n = 8 mice per dose.

Fig. 5 shows the effects of 10 mg/kg naloxone pretreatment on the hypoventilatory effects of the highest cumulative dose tested for each drug. Pretreatment with naloxone attenuated the hypoventilatory effects of all drugs, except for buprenorphine (SAL: 68.0% ± 4.07%; NLX: 83.0% ± 5.71%; t(12.64) = 2.141, p = 0.052392). Significant increases in minute volume with 10 mg/kg naloxone pretreatment were observed for 32 mg/kg morphine (SAL: 56.6% ± 5.42%; NLX: 83.1% ± 3.68%; t (12.33) = 4.049, p = 0.001529), 3.2 mg/kg fentanyl (SAL: 51.6% ± 4.73%; NLX: 78.6% ± 3.56%; t(13.01) = 4.570, p = 0.000524), 10 mg/kg isobutyrylfentanyl (SAL: 58.8% ± 5.30%; NLX: 85.1% ± 6.06%; t (13.75) = 3.270, p = 0.005696), 10 mg/kg crotonylfentanyl (SAL: 44.9% ± 2.30%; NLX: 90.1% ± 3.25%; t(12.60) = 11.36, p < 0.000001), 10 mg/kg para-methoxyfentanyl (SAL: 41.5% ± 2.53%; NLX: 73.5% ± 3.87%; t(12.05) = 6.915, p = 0.000016), 10 mg/kg para-methoxybutyrylfentanyl (SAL: 59.5% ± 2.56%; NLX: 86.1% ± 3.54%; t(12.75) = 6.106, p = 0.000041), 10 mg/kg 3-furanylfentanyl (SAL: 42.9% ± 3.34%; NLX: 81.1% ± 4.58%; t(12.79) = 6.741, p = 0.000015), 100 mg/kg thiophenefentanyl (SAL: 47.0% ± 3.67%; NLX: 80.3% ± 5.22%; t (12.56) = 5.220, p = 0.000185), and 100 mg/kg benzodioxolefentanyl (SAL: 79.6% ± 1.55%; NLX: 90.3% ± 2.02%; t(13.14) = 4.173, p = 0.001068).

Fig. 5.

Results for the highest cumulative doses of each drug tested with and without naloxone pretreatment in cumulative dose ventilation tests. Bars represent minute volume as a percentage of control (± SEM) for (1) 32 mg/kg morphine, (2) 3.2 mg/kg buprenorphine, (3) 3.2 mg/kg fentanyl, (4) 10 mg/kg isobutyrylfentanyl, (5) 10 mg/kg crotonylfentanyl, (6) 10 mg/kg para-methoxyfentanyl, (7) 10 mg/kg para-methoxybutyrylfentanyl, (8) 10 mg/kg 3-furanylfentanyl, (9) 100 mg/kg thiophenefentanyl, and (10) 100 mg/kg benzodioxolefentanyl with a pretreatment of either vehicle [VEH] (open bars) or 10 mg/kg naloxone [NLX] (filled bars) for n = 8 mice per dose. Significant differences between drug + vehicle and drug + naloxone conditions are indicated by asterisks: *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.

Table 2 shows values for safety of each fentanyl analog tested relative to positive controls morphine, buprenorphine, and fentanyl. A measure of relative drug safety, identified as a protective index (PI), was calculated by dividing the TD₅₀ for a given drug, or the amount of drug required to produce a half-maximal effect in hypoventilatory tests, by its ED₅₀, or the amount of drug required to produce a half-maximal effect in antinociceptive tests. PI values showed a rank order of safety (high safety to low safety) as follows: benzodioxolefentanyl (PI = 220) > isobutyrylfentanyl (PI = 176) > para-methoxybutyrylfentanyl (PI = 152) > buprenorphine (PI = 102) > crotonylfentanyl (PI = 12.1) > fentanyl (PI = 12.0) > para-methoxyfentanyl (PI = 7.74) > morphine (PI = 7.07) > 3-furanylfentanyl (PI = 5.05) > thiophenefentanyl (PI = 3.85).

Table 2.

Results from safety analyses.

#	Drug	Hypoventilation TD50 (mg/kg)	Antinociception ED50 (mg/kg)	Protective Index (PI)
1	Morphine	55.3	7.82	7.07
2	Buprenorphine	10.8	0.11	102
3	Fentanyl	0.96	0.08	12.0
4	Isobutyrylfentanyl	13.5	0.08	176
5	Crotonylfentanyl	2.72	0.23	12.1
6	Para-methoxyfentanyl	3.31	0.43	7.74
7	Para-methoxybutyrylfentanyl	16.1	0.11	152
8	3-Furanylfentanyl	2.60	0.51	5.05
9	Thiophenefentanyl	18.0	4.66	3.85
10	Benzodioxolefentanyl	10,168	46.3	220

Protective Indices (PI) were calculated as PI = TD₅₀/ED₅₀.

4. Discussion

This study represents the first known pharmacological assessments of the respiratory depressant effects of seven fentanyl-related substances in freely moving mice and provides essential, previously unreported, in vivo potency, efficacy, and mechanistic data. We present a cumulative dosing methodology for rapid and efficient screening of putative opioid receptor ligands for their respiratory depressant effects. We demonstrate that the fentanyl results from this within-subjects cumulative dosing design, including ED₅₀ values for hypoventilatory effects as interpolated (or extrapolated based on the slope of the dose–effect curve) by non-linear regression, are concordant with results from a conventional between-subjects acute dosing methodology. Accordingly, there were two main findings. First, subcutaneously administered fentanyl-related substances (isobutyrylfentanyl, crotonylfentanyl, para-methoxyfentanyl, para-methoxybutyrylfentanyl, 3-furanylfentanyl, thiophenefentanyl, and benzodioxolefentanyl), like known opioids such as morphine and fentanyl, induced hypoventilation to varying degrees in the mouse. Second, hypoventilation induced by fentanyl-related substances was attenuated with a 10 mg/kg naloxone pretreatment indicating that these effects are at least in part mediated by the MOR. Moreover, separations in potencies required to elicit hypoventilation and antinociception [29,30] were apparent, most notably for those of isobutyrylfentanyl and para-methoxybutyrylfentanyl. These findings have implications for medications development, specifically for the development of novel pharmacotherapeutics that produce significant analgesia, but lack serious adverse effects, such as respiratory depression, at identical doses. Taken together, these findings elucidate the mechanisms of action of previously uncharacterized fentanyl-related substances and support the further use of structure activity relationships to study the pharmacological determinants of the therapeutic vs. toxic effects of these drugs.

Consistent with the existing literature, morphine and fentanyl, the MOR agonist standards tested in cumulative dose ventilation tests, elicited hypoventilation in mice. These tests demonstrate that fentanyl (ED₅₀ = 0.96 mg/kg; 95% CI: 0.50–2.24) was 57.6x more potent than morphine (ED₅₀ = 55.3 mg/kg; 95% CI: 19.3–393). Consistent with the findings of Hill et al. (2019), both morphine and fentanyl suppressed ventilatory activity in the mouse, as measured by decreases in respiratory rate and minute volume [49]. Interestingly, however, in contrast to Hill et al. (2019), this study found that changes in tidal volume were not a significant contributor to fentanyl’s effects on minute volume—respiratory rate was the primary driver of decreases in ventilation. One plausible explanation for this difference is the route of administration (SC in this study, IV in the Hill study) as well as the subject strain used (Swiss Webster in this study, CD1 in the Hill study). Existing literature on the pharmacokinetics of fentanyl indicates that a dose of 15 μg/kg fentanyl administered IV can result in peak plasma concentrations that are >10× higher than when administered orally or sublingually [59]. Furthermore, studies by Elmer et al. (1995) using etonitazene on various strains of mice indicated the importance of strain on the minute volume in the mouse [60]. Overall, the MOR agonist standards in these tests produced effects on ventilatory endpoints in mice that were consistent with previous reports.

Buprenorphine is a relatively high-affinity MOR partial agonist and partial NOP agonist [61–63]. In this study, buprenorphine decreased ventilation at low doses and appeared to trend toward the baseline at higher doses. The U-shaped buprenorphine dose-effect curve relating minute volume to dose may be attributed to the increasing production of buprenorphine’s primary metabolite, norbuprenorphine, a high efficacy MOR agonist, that is then blocked at higher doses by an excess of the parent drug that shifts the equilibrium at the receptor, or the unique residency time of buprenorphine on the MOR that prevents binding of norbuprenorphine [64–66]. Several studies have described ceiling effects of buprenorphine on respiration in conscious Sprague Dawley rats (0.1, 1, 10 mg/kg, IA) as measured by arterial pCO₂ and pO₂ [63], in rhesus monkeys (0.1, 0.3, 1 mg/kg, IM) as measured by respiratory rate and minute volume [67], and in man (16, 32 mg, SL), as measured by respiratory rate and arterial oxygen saturation [68]. Moreover, attempts to antagonize the hypoventilatory effects of buprenorphine (0.003, 0.01, 0.03, 0.1 mg/kg, IM) with naltrexone (0.03 mg/kg, IM), a selective MOR antagonist, were effective as indicated by right-ward shifts in the respiratory rate and minute volume dose effect curves, but failed with J-113397 (0.1 mg/kg, IM), a selective NOP antagonist [67]. It is unknown, however, if higher doses of J-113397 (>0.1 mg/kg, IM) would have disrupted the hypoventilation ceiling exhibited by higher doses of buprenorphine (>1 mg/kg, IM) in rhesus monkeys. Overall, effects by buprenorphine on ventilation in Swiss Webster mice in this study are consistent with previous reports.

We previously reported the effects of the fentanyl-related substances tested in this study on locomotion and nociception (with and without a 1 mg/kg naltrexone pretreatment) in mice. We demonstrated that these fentanyls elicit their effects, at least in part, by the MOR [29,30]. Here, these drugs elicited significant, dose-dependent hypoventilation in mice that was attenuated by pretreatment with naloxone (10 mg/kg) suggesting that their respiratory depressant effects also involve the MOR. In safety analyses, ED₅₀ values for antinociception (from [29,30]) and TD₅₀ values for hypoventilation (in this study) were used to calculate protective indices (PI = TD₅₀/ED₅₀) where greater values indicate higher relative safety. The MOR agonist standard buprenorphine (PI = 102) was about a log unit safer than both fentanyl (PI = 12.0) and morphine (PI = 7.07). Interestingly, two fentanyl-related substances with significant analgesic activity, isobutyrylfentanyl (PI = 176) and para-methoxybutyrylfentanyl (PI = 152), had greater safety than even that of buprenorphine. Crotonylfentanyl (PI = 12.1) was approximately as safe as fentanyl, and para-methoxyfentanyl (PI = 7.74) was approximately as safe as morphine. Both 3-furanylfentanyl (PI = 5.05) and thiophenefentanyl (PI = 3.85) were less safe than morphine. A protective index for benzodioxolefentanyl (PI = 220) was computed and is unsurprisingly safe given its in vitro MOR efficacy [26].

It seems plausible that differences in the effects observed for these fentanyl analogs may be attributed to differences in efficacy and functional selectivity at the MOR. However, in a [³⁵S]-GTP-gamma-S assay using MOR-expressing CHO cells, the E_max values for isobutyrylfentanyl (94%), para-methoxybutyrylfentanyl (55%), and buprenorphine (27%) [26,69], suggest that the relative differences in safety for these substances is likely not influenced by MOR efficacy. While the MOR functional selectivity of isobutyrylfentanyl has to our knowledge not yet been characterized, it has been established that para-methoxybutyrylfentanyl is more G-protein biased than is fentanyl, possibly owing to its improved safety profile [27]. It is notable that naloxone shows some inability to completely reverse fentanyl-induced decreases in minute volume, which others have suggested may be due to rapid closure of the vocal cords followed by rigidity in the diaphragm, chest wall, and upper airway (wooden chest syndrome) induced by a non-opioid receptor mechanism [70,71]. Lastly, there is some evidence that several structural substituents may alter the binding pose of fentanyl analogs in the MOR binding pocket, with bulkier N-acyl substitutions lending to decreases in ligand potency and efficacy (e.g., benzodioxolefentanyl which is devoid of activity at the MOR) [26].

Several limitations were identified. First, the route of administration used in this study (subcutaneous) is different from the routes commonly used by persons with OUD to consume fentanyl containing products [45]. It is possible that the subcutaneous route lends to an underestimation in respiratory risk relative to intravenously administered fentanyl analogs due to potentially slower absorption, but likely represents a close approximation to the intranasal route for which drugs are taken up via blood vessels in the mucosa of the nasal cavity. Second, the floor effects inherent to measuring respiratory depression in mice, even when testing drug-naïve subjects under 5% CO2 and dark phase conditions, necessitate extremely high doses of fentanyl analogs to produce complete cessation of breathing [54]. However, the ED₅₀ values for fentanyl and morphine in this study are consistent with doses producing 50% suppression of ventilation as reported by others (see Hill et al., 2020) [49], even though the ED₅₀ values for some drugs were derived from dose-effect curves with E_max values close to, but not quite, achieving 50% levels. Third, the naloxone dose (10 mg/kg) used in this study is relatively high, but was selected to reduce the risk of false negatives. While naloxone is likely to have bound, at least to some degree, to delta-opioid receptors (DOR) and kappa-opioid receptors (KOR) at 10 mg/kg, it is unlikely that these receptors contribute to the respiratory depressant effect of the ligands tested in this study given previous research by others demonstrating an absence of opioid-induced respiratory depression in MOR-knockout mice [72]. Future studies may include pA₂ analyses using competitive, selective, reversible antagonists, the use of irreversible antagonists, and or receptor knockout models, to empirically test the contributions of receptors other than the MOR to respiratory depression induced by fentanyl analogs.

This study demonstrates that subcutaneously administered fentanyl-related substances like the MOR agonist standards elicit hypoventilation in mice with varying potencies and efficacies and that their effects are mediated, at least in part, by the MOR. Moreover, these studies demonstrate that, under the conditions for these tests, two of the fentanyl-related substances tested are >10× better in safety analyses than both fentanyl and morphine, that is, they exhibit a separation in antinociceptive and hypoventilatory potencies by more than two logarithmic units as compared to about one logarithmic unit for both fentanyl and morphine. These findings have significant implications for the viability of fentanyl-related substances as anesthetics and analgesics. In future studies, it will be paramount to assess alternative mechanisms with which isobutyrylfentanyl and para-methoxybutyrylfentanyl exert their antinociceptive and hypoventilatory effects (e.g., nociceptin opioid peptide (NOP) receptor and/or biased agonism), and their pharmacokinetics via different routes of administration, especially the via the intravenous route. Overall, fentanyl-related substances, although problematic for their abuse, represent a diverse class of drugs with potential clinical utility and should be explored further for their effects.

Abbreviations:

MOR

mu-opioid receptor

SAL

saline

NLX

naloxone