Amphetamines modulate fentanyl-depressed respiration in a bidirectional manner

Abstract

Background:

The opioid epidemic remains one of the most pressing public health crises facing the United States. Fentanyl and related synthetic opioid agonists have largely driven the rising rates of associated overdose deaths, in part, because of their surreptitious use as substitutes for other opioids and as adulterants in psychostimulants. Deaths involving opioids typically result from lethal respiratory depression, and it is currently unknown how co-use of psychostimulants with opioids affects respiratory toxicity. Considering psychostimulant overdoses have increased over 3-fold since 2013, and half of those co-involved opioids, this is a cardinal question.

Methods:

Naloxone, d-amphetamine (AMPH), and (±)-methamphetamine (METH) were evaluated for their effects on basal and fentanyl-depressed respiration. Minute volume (MVb) was measured in awake, freely moving mice via whole-body plethysmography to quantify fentanyl-induced respiratory depression and its modulation by dose ranges of each test drug.

Results:

Naloxone immediately reversed respiratory depression induced by fentanyl only at the highest dose tested (10 mg/kg). Both AMPH and METH exhibited bidirectional effects on MVb under basal conditions, producing significant (p ≤ 0.05) depressions then elevations of respiration as dose increased. Under depressed conditions the bidirectional effects of AMPH and METH on respiration were exaggerated, exacerbating and then reversing fentanyl-induced depression as dose increased.

Conclusions:

These results indicate that co-use of amphetamines with fentanyl may worsen respiratory depression, but conversely, monoaminergic components of the amphetamines may possibly be exploited to mitigate fentanyl overdose.

1. Introduction

Opioid overdose currently represents one of the most significant public health crises in the United States, with death rates increasing yearly for over a decade (CDC Injury Center, 2021). Since 2016 illicitly manufactured fentanyl and novel synthetic opioids (NSO, e.g., fentanyl analogs) have become the greatest contributors to these increasing opioid overdose death rates, which increased by 1040 % between 2013 and 2019 (CDC Injury Center, 2021; Mattson et al., 2021a). In contrast, over the same period rates of heroin and prescription opioid overdoses slowed and recently fell (CDC Injury Center, 2021; Mattson et al., 2021a). These trends indicate that measures such as responsible prescribing practices, prescription drug monitoring, and improved access to treatment and naloxone may be mitigating some of the mortality associated with misuse of typical morphinan opioids but are not satisfactorily curbing fentanyl-related overdoses (Fleming et al., 2020; Lai et al., 2021; NIDA, 2017). Fentanyl and NSOs likely pose a greater risk of mortality than morphinans due to several factors, including high potency, lipophilicity, speed of onset, and effects such as upper airway rigidity (Armenian et al., 2017; Hill et al., 2020; Kim et al., 2019; Merel et al., 2012; Torralva and Janowsky, 2019; van Dorp et al., 2007). Additionally, fentanyl and NSOs are more easily trafficked and frequently utilized as adulterants or clandestine substitutes for other misused drugs such as oxycodone and alprazolam (Cano and Huang, 2021; DEA, 2018; Fleming et al., 2020; Kariisa et al., 2019; McCall Jones et al., 2017; Singh et al., 2019). Evidence suggests decreased availability of heroin and prescription opioids that are unadulterated with fentanyl has promoted co-use of methamphetamine alongside opioids to manage undesirable sedation, augment subjective effects, and treat withdrawal symptoms (Ellis et al., 2018; Lopez et al., 2021). These factors have contributed to a new wave in the overdose epidemic, consisting of overdose deaths in which both fentanyl and psychostimulants (stimulants) are present due to adulteration or intentional co-consumption.

Combinations of morphinans and stimulants have a history of medical use and interact in both synergistic (analgesia) and antagonistic (somnolence) ways depending on the effect (Izenwasser and Kornetsky, 1988; Kauppila et al., 1992; Sasson et al., 1986; Sprague and Takemori, 1978). Considering the clinical relevance of respiratory modulation by recreationally used drugs there is a paucity of scientific evidence documenting these properties of stimulants and, more importantly, how they interact with known respiratory depressants (i.e., opioids). Our laboratory previously characterized the effects of fentanyl and its analogs in mouse models of nociception, locomotion, and respiratory depression (Varshneya et al., 2019, 2021, 2022a, 2022b). However, we are not aware of published studies that have reported the respiratory consequences of fentanyl and stimulant interactions. Such characterization could guide clinical treatment strategies and help identify therapeutic agents to modulate depressed respiration. In the present study, the dose- and time-effects of fentanyl on ventilatory measures in mice were recorded, and the ability of naloxone to reverse fentanyl-induced ventilatory depression was assessed. Next, d-amphetamine and (±)-methamphetamine were tested for their ability to influence basal ventilation. Finally, the effects of d-amphetamine and (±)-methamphetamine on fentanyl-depressed ventilation were assessed to determine the nature of their interactions.

2. Material and methods

2.1. Materials

Fentanyl citrate (#F3886; IUPAC: N-phenyl-N-[1-(2-phenylethyl) piperidin-4-yl]propanamide) and d-amphetamine hemisuflate salt (#A5880; IUPAC: (2S)– 1-phenylpropan-2-amine) were obtained commercially (Sigma-Aldrich, Inc., St. Louis, MO, USA). (±)-Methamphetamine hydrochloride (IUPAC: N-methyl-1-phenylpropan-2-amine) and (−)-naloxone HCl dihydrate (IUPAC: (4R,4aS,7aR,12bS)– 4a,9-dihydroxy-3-prop-2-enyl-2,4,5,6,7a,13-hexahydro-1H-4,12-methanobenzofuro[3,2-e]isoquinolin-7-one;hydrochloride) were provided by the National Institute on Drug Abuse (Bethesda, MD, USA) Drug Supply Program. All drugs were prepared in saline, sterilized by filtration through 0.2 μm filtration disks, and administered s.c. at a volume of 10 ml/kg body weight.

2.2. Subjects

Adult male mice [Swiss Webster, CFW(SW), Charles River Laboratories International, Raleigh, NC, USA] weighing approximately 25–50 g (mean weight ± SD; 39.0 ± 3.6 g) at the time of testing were housed four subjects per cage in Association for Assessment and Accreditation of Laboratory Animal Care-accredited facilities. Mice had ad libitum access to food (Teklad 7012 Rodent Diet; Envigo, Madison, WI, USA) and tap water. Vivaria were maintained at 22 °C ± 2 °C and 45–50 % humidity, with lights set to a reverse 12-h light/dark cycle (lights on at 10:00). All tests were conducted on weekdays during the dark period between 11:00 and 17:00 to ensure mice were active (i.e., not asleep). All subjects were acclimated to the vivarium for at least one week before the commencement of studies and were experimentally and drug-naive before testing. Subjects were tested once and were not used for any subsequent tests to preclude drug or testing history effects. All procedures were carried out in accordance with the National Research Council’s Guide for Care and Use of Laboratory Animals (2011). This experimental protocol was approved by the Institutional Animal Care and Use Committee at Virginia Commonwealth University.

2.3. Apparatus

Mice were tested using whole-body plethysmograph devices (FinePointe WBP Chamber with Halcyon Technology, Data Sciences International, St. Paul, MN, USA) while unrestrained and allowed free movement in individual isolated experimental vessels. Experimental vessels (0.5 L volume with adjustable 0.5 L/min room air bias flow) were housed in a laboratory illuminated by custom 660 nM-emitting T8-style ceiling-mounted light tubes each with 96, 0.2-watt Epistar 2835 SMD LEDs (Shenzhen Benwei Electronics Co., Ltd., Longhua District, Shenzhen, China). This wavelength is minimally visible to mice (Peirson et al., 2018) which enabled maintenance of subjects in the dark phase of their activity cycle during testing. All tests were conducted using a standardized gas mixture composed of 5 % CO₂, 21 % O₂ and balance N₂ to minimize variability between tests due to fluctuations in ambient air and to increase the sensitivity and capacity of the assay to detect pharmacologically-induced effects on respiration. This gas mixture has been evaluated and reported to be devoid of anxiogenic effects (Hill et al., 2016). Respiratory rate, tidal volume, and minute volume (MVb) were recorded using software (FinePointe Software Research Suite; Data Sciences International).

2.4. Three-phase whole-body plethysmography procedure

On the day prior to testing (24 h before the start of each test session), mice were habituated to the testing environment by placing them in the chambers under red light conditions for 30 min while they were supplied ambient air by the active plethysmograph (see Fig. 1). The day following habituation to testing conditions, naïve mice (n = 8 per group) were randomly assigned to treatment groups and administered three s.c. injections, each initiating a separate recording phase during testing: Phase I (Baseline Phase) was initiated by an injection of saline, the vehicle used whenever an opioid respiratory standard was used, after which respiratory measures were recorded for 20 min. Phase II (Agonist Phase) was initiated by an injection of an opioid agonist (fentanyl) or its vehicle and their effects on measures of respiration were recorded for 20 min. Phase III (Reversal Phase) was initiated by an injection of a potential reversal agent (e.g., naloxone) or its vehicle, a test compound (e.g., d-amphetamine or methamphetamine) or their respective vehicles (saline for all) and respiratory parameters were recorded for 1 h to measure their ability to alter MVb depression induced by administration of the opioid agonist given prior to Phase II. Each respective phase will henceforth be referred to by its numeric identifier (Phase I, II, or III) rather than descriptive identifiers (e.g., Agonist Phase) for clarity and consistency.

Fig. 1.

Illustration of the three-phase experimental protocol following the habituation session on the day prior to testing.

Fentanyl, d-amphetamine, and (±)-methamphetamine were evaluated at behaviorally active doses within the WBP protocol described above. Fentanyl (0.01, 0.1, 0.3, 1.0, 3.0, 10 mg/kg) was administered as the opioid agonist prior to Phase II, while d-amphetamine, (±)-methamphetamine, naloxone or their respective vehicles (saline) were administered as ‘test compounds’ at the start of Phase III. Naloxone was evaluated at three doses (0.1, 1.0, 10 mg/kg) for its ability to reverse the MVb depressant effects of 0.3 mg/kg fentanyl. Both amphetamines were tested at three identical doses (1.0, 3.0, 10 mg/kg) under basal and fentanyl-depressed conditions. Depressed conditions consisted of treatment with 0.3 mg/kg fentanyl prior to Phase II. This dose was chosen for reversal tests because it was the closest half-log unit to its ED₅₀ for suppressing respiration (see Results, Section 3.1), and was the lowest dose that was determined to consistently produce statistically significant depression by 20 min its post-administration and that maintained significant depression of MVb for at least 10 min during Phase III.

2.5. Statistical analysis

The primary dependent measure, normalized minute volume (MVb = frequency x tidal volume), was expressed as a percentage of baseline MVb collected during Phase I for each individual subject (within-subject normalization). Normalized MVb values were calculated for individual subjects at each time point by dividing the average MVb value obtained from each 5-min bin by the average MVb value recorded during the 20-min baseline assessment in Phase I x 100%. The mean of each test groups’ normalized MVb values was then calculated from the average of individual subjects’ normalized MVb values at each time point. Normalized group MVb values were analyzed using a 2-way mixed model ANOVA with time as the within-subjects factor and treatment as the between-subjects factor. Normalization to baseline was performed in order to control for any variation between treatment groups that may have arisen from differences in baseline MVb. ANOVAs with significant treatment effects were followed by Holm-Şídák post-hoc multiple comparisons tests to compare MVb values of treatment groups with control groups at individual time points. Control groups received three injections of saline (vehicle), one injection each prior to the initiation of Phases I, II, and III respectively. To determine if between-group differences existed during the baseline assessment in Phase I, raw MVb data (ml/min) were analyzed using one-way ANOVA, followed by Holm-Şídák post-hoc multiple comparisons tests to determine if a significant group effect was detected. Fentanyl’s ED₅₀ (95 % CI) for depression of MVb was calculated by nonlinear fit of the difference between baseline and normalized MVb means at 10 min post-administration (i.e., the time to peak effect). ANOVAs and post-hoc comparisons were considered statistically significant if p < 0.05. Complete reversal of fentanyl-induced depression was defined as occurring when treatment groups did not differ significantly from controls following the initiation of Phase III. Partial reversal was defined as an increase in the MVb of a treatment group compared to the fentanyl-treated group following the initiation of Phase III that still significantly differed from vehicle controls. Breath frequency and tidal volume were analyzed via identical methods to those used for MVb to identify their respective contributions to observed effects on MVb. All analyses were performed using software (GraphPad Prism 9 for Macintosh; GraphPad Software, San Diego, CA, USA).

3. Results

3.1. Depression of ventilation by fentanyl

Fig. 2 shows the effects of fentanyl (0.01–10 mg/kg) on MVb. Following fentanyl administration, MVb was significantly altered [F(96, 784) = 3.833; p< 0.0001]. Post-hoc analysis indicated that 0.01 mg/kg fentanyl, the lowest dose tested, did not alter MVb, whereas doses of 0.1 mg/kg and higher significantly (p ≤ 0.05) decreased ventilation beginning within 5 min of their administration. The magnitude and duration of depression occurred in a dose-dependent manner, with maximal depression occurring at 10 min post-administration for 10 mg/kg, the highest dose tested, and between 10 min and 15 min post-administration for all other doses. Fentanyl-induced depression of MVb was driven primarily by decreases in breath frequency (Freq) for all doses (data not shown). Decreases in tidal volume (TVb) were relatively small following doses of 0.01 – 1.0 mg/kg (100 – 86.67 % of baseline), with higher doses of 3.0 and 10 mg/kg significantly depressing TVb to 80.25 % and 61.50 % of baseline within 10 min of administration, respectively. Fentanyl’s ED₅₀ value (95 % CI) for suppressing ventilation was 0.3969 mg/kg [0.2410–0.6650]. As fentanyl dose increased, the duration of MVb suppression increased, with 10 mg/kg fentanyl still significantly decreasing MVb 55 min post-administration. Normalized MVb values were below baseline at the end of recording for all treatment groups with final MVb values between 59.4 % (1.0 mg/kg fentanyl) to 87.2 % (saline control). Mean percent MVb values of the control group ranged from 83.2 % to 108.0 % across the duration of the test session. Control group MVb values increased momentarily 5 min after each injection (e.g., at 5 min following the end of Phase I), possibly resulting from the handling and injection process. Subject weights were similar with respect to mean and variability across all experimental conditions (Sections 3.1 - 3.4), with a mean (± SD) of 39.0 ± 3.6 g at the time of testing.

Fig. 2. :

Effects of fentanyl (FENT) on ventilation in mice. Left ordinate: mean raw MVb (ml/min) indexing values of symbols only during baseline (B) of Phase I. Right ordinate: normalized (percent baseline) MVb indexing values of symbols during the 80-min test session following Phase I baseline. These symbols indicate mean MVb expressed as a percentage of baseline MVb of 8 mice per treatment group. Filled symbols indicate significant differences compared to the MVb of saline-treated controls at individual timepoints (p ≤ 0.05). Abscissa labels: B = mean baseline MVb, F or S = fentanyl or saline (SAL) injection, respectively, S = saline injection. Legend labels correspond to dose in mg/kg. Error bars represent the standard error of normalized MVb mean values within individual 5 min bins. Points lacking error bars represent instances where the range of standard error is smaller than the given symbol. No significant differences were detected at baseline across experimental conditions when raw MVb values were compared via one-way ANOVA [F(6, 49) = 0.6546; p = 0.6862].

3.2. Reversal of fentanyl-induced depression by naloxone

Fig. 3 shows that the MVb depression produced by 0.3 mg/kg fentanyl is reversed by naloxone (0.1, 1.0, 10 mg/kg) in a dose-dependent manner [F(64, 560) = 4.692; p < 0.0001]. Post-hoc analysis indicated that 0.3 mg/kg fentanyl significantly depressed MVb relative to saline-treated controls for 30 min post-administration, with peak depression of MVb occurring at t = 15 min (47.5 %). Naloxone reversed the effects of fentanyl on MVb in a dose-dependent manner, with MVb values at 5-min post-administration (i.e., t = 25 min) of 78.8 %, 83.6 %, and 94.3 % for 0.1, 1.0 and 10 mg/kg respectively, compared with an MVb of 59.5 % in subjects who received saline prior to Phase III. Naloxone at 10 mg/kg completely reversed depression of MVb, increasing mean MVb to a level that did not significantly differ from that of vehicle-treated control subjects (p = 0.09) at t = 25 min. Naloxone-induced increases in MVb were driven primarily by elevation of Freq with lesser contributions of TVb (approximately 40% elevation vs 12 %, respectively). Groups that received any dose of naloxone did not significantly differ from controls by 10 min post-administration of naloxone (t = 30), while subjects receiving fentanyl and saline recovered to non-significant levels by t = 35 min. Normalized MVb values for all fentanyl treated subjects (64.9–76.7 %) remained nonsignificantly lower than those of controls (87.2 %) at the end of the recording period.

Fig. 3. :

Dose- and time-effects of naloxone (NLX) on fentanyl-depressed Minute Volume. Abscissa labels B = mean baseline MVb, F or S = fentanyl or saline injection, respectively, N or S = naloxone or saline injection, respectively. No significant differences were detected at baseline across experimental conditions when raw MVb values were compared via one-way ANOVA [F(4, 35) = 0.4731; p = 0.7551]. All other details same as Fig. 2.

3.3. Effects of d-AMPH and METH on ventilation

Fig. 4 shows the dose-dependent effects of AMPH (1.0, 3.0, 10 mg/kg) on basal MVb in subjects who received saline prior to Phase II. MVb values for saline-treated controls ranged from 76.4 % to 118.0 % over the duration of the test session, with transient increases seen after each injection. Basal MVbs were significantly lower in all treatment groups (1.0, 3.0, 10 mg/kg) than controls at 5 min following saline injections and this was sustained until 10 min in two treatment groups (1.0, 3.0 mg/kg). No other differences in basal MVb were observed between groups prior to the start of Phase III. Administration of AMPH significantly affected MVb [F(48, 448) = 21.61; p < 0.0001], with 1.0 and 3.0 mg/kg inducing significant depression of MVb and 10 mg/kg significantly increasing MVb relative to saline controls. AMPH at low and moderate doses (1.0, 3.0 mg/kg) significantly depressed MVb within 5 min of administration, with peak depression to 69.6 % (p = 0.0088) and 71.2 % (p = 0.0065) of baseline occurring at 15 min post-administration, respectively. In contrast, 10 mg/kg AMPH significantly increased MVb to 118.8 % of baseline (t = 35, p = 0.0015) at 15 min post-administration and MVb values increased further throughout the duration of Phase III, with a final MVb value of 151.09 % (t = 80) that was significantly (p = 0.0019) greater than controls. AMPH-induced changes in basal MVb were due to opposing effects on Freq and TVb. All AMPH doses elevated Freq within 15 min of administration. Conversely, administration of AMPH depressed TVb according to a U-shaped dose response, with the greatest reductions observed after 3.0 mg/kg which significantly depressed TVb within 15 min, and both 1.0 and 10 mg/kg producing comparatively minor reductions.

Fig. 4. :

Dose- and time-effects of d-amphetamine (AMPH) on basal minute volume following saline (SAL) pretreatment. B = mean baseline MVb, A = AMPH injection, S = saline injection. N = 8 per group. No significant differences were detected at baseline across experimental conditions when raw MVb values were compared via one-way ANOVA [F(3, 28) = 2.532; p = 0.0773]. All other details same as Fig. 2.

Fig. 5 shows dose-dependent effects of METH (1.0, 3.0, 10 mg/kg) on basal MVb following treatment with saline. Significant differences in basal MVb were observed between two treatment groups and controls at 5 min (1.0, 3.0 mg/kg; p = 0.0049 – 0.0155), 10 min (1.0, 10 mg/kg; p = 0.0003 – 0.0222), and 15 min (3.0 mg/kg; p = 0.0405) following saline injections. Administration of METH significantly affected basal MVb [F(48, 448) = 11.81; p < 0.0001] in a manner similar to the effects seen with AMPH. Specifically, 1.0 mg/kg METH significantly depressed MVb compared to saline controls from 5 to 30 min post-administration, with peak depression to 71.8% of baseline occurring 15 min post-administration (t = 35; p = 0.0439). In contrast to AMPH, administration of 3.0 mg/kg METH had no significant effect on MVb values at any time point relative to controls. Following administration of METH at 10 mg/kg MVb values increased significantly beginning at 15 min post-administration to 101.5 % of baseline (t = 35; p = 0.0028) and increased continuously throughout the duration of Phase III, reaching a peak of 150.1 % at the end of the recording session compared to 81.6 % in controls (t = 80; p = 0.0176). Finally, the contributions of both Freq and TVb to METH’s effects on basal MVb were similar to those observed with AMPH. Namely, following administration of METH dose-dependent elevations in Freq were observed within 10 min, with peak elevation of Freq to 146.77 % of baseline occurring 25 min after administration of 10 mg/kg METH. Effects on basal TVb similarly followed a U-shaped dose-response with the greatest depression of TVb occurring after administration of 3.0 mg/kg. However, in contrast to AMPH, METH at the lowest dose of 1.0 mg/kg had significant depressant effects on TVb while 10 mg/kg METH produced greater elevations of TVb relative to AMPH.

Fig. 5. :

Dose- and time-effects of (±)-methamphetamine (METH) on basal Minute Volume following saline pretreatment. B = mean baseline MVb, M = METH injection, S = saline injection. N = 8 per group. No significant differences were detected at baseline across experimental conditions when raw MVb values were compared via one-way ANOVA [F(3, 28) = 1.661; p = 0.1979]. All other details same as Fig. 2.

3.4. Effects of d-AMPH and METH on fentanyl-depressed ventilation

Fig. 6 displays the dose-dependent, bidirectional effects of AMPH on fentanyl-depressed respiration. Administration of fentanyl depressed MVb across treatment groups to 44.8–57.8 % of baseline by 15 min post-administration. There was a significant main effect of treatment on respiration [F(64, 560) = 16.72; p < 0.0001]. Administration of 1.0 mg/kg AMPH augmented the depressant effects of fentanyl on MVb by maintaining significantly depressed MVb values between 46.2 % and 53.6 % for the duration of the test session, with a final MVb of 53.6% (t = 80; p = 0.0002). Administration of 3.0 mg/kg AMPH similarly augmented the effects of fentanyl on MVb by maintaining significantly depressed MVb values between 52.1 % and 61.6 % for the duration of the test session, with a final MVb of 52.2 % (t = 80; p = 0.0006). In contrast, subjects who received saline following fentanyl pretreatment showed a steady recovery of MVb to a level not significantly lower than saline controls by 50 min (t = 50; p = 0.0526), with a final MVb of 71.2 % of baseline (t = 80; p = 0.2651). Enhanced depression of both Freq and TVb contributed to the effects of 1.0 mg/kg AMPH on fentanyl-depressed MVb while 3.0 mg/kg had opposing effects on the two component measures; moderately elevating depressed Freq but significantly exacerbating decreases in TVb post-administration. In contrast, administration of 10 mg/kg AMPH to fentanyl-pretreated subjects produced a significant increase in MVb, with complete reversal of MVb depression occurring at 15 min post-administration (t = 35; p = 0.5248). Thereafter, MVb values continued to steadily rise throughout the duration of Phase III in this treatment group, becoming significantly greater than control values by 50 min post administration when average MVb was 109.0 % of baseline (t = 70; p = 0.0194) and this was sustained for the duration of the recording period culminating in a peak MVb value of 118.8 % at the final timepoint (t = 80; p = 0.0348). As seen in basal tests, 10 mg/kg AMPH primarily elevated MVb via large increases in Freq without any corresponding loss or gain of TVb.

Fig. 6. :

Dose- and time-effects of d-amphetamine (AMPH) on Minute Volume following pretreatment with 0.3 mg/kg fentanyl (FENT). B = mean baseline MVb, F = fentanyl injection, A = AMPH injection, S = saline injection. N = 8 per group. No significant differences were detected at baseline across experimental conditions when raw MVb values were compared via one-way ANOVA [F(4, 35) = 2.598; p = 0.0529]. All other details same as Fig. 2.

Fig. 7 displays the bidirectional effects of METH (1.0, 3.0, 10 mg/kg) on depressed MVb following treatment with 0.3 mg/kg fentanyl. Administration of 0.3 mg/kg fentanyl significantly depressed MVb across treatment groups to 43.0–53.9% of baseline, during the window of peak effects at 15–20 min post-administration. Administration of METH to fentanyl-pretreated subjects significantly [F(64, 560) = 15.90; p < 0.0001] affected MVb in a bidirectional manner. The depressant effects of fentanyl were further increased by 1.0 mg/kg METH to 42.7% of baseline by 30 min post-administration (t = 50; p < 0.0001) and this significant depression lasted for the duration of the test session resulting in a final MVb of 54.08 % (t = 80; p < 0.0001). Administration of METH at 3.0 mg/kg similarly extended the length of significant MVb depression with values between 58.3 % and 66.8 % of baseline lasting until the end of the test session and a final MVb of 58.3% (t = 80; p = 0.0004). In both cases METH’s effects on Freq and TVb closely tracked those observed with AMPH under fentanyl-depressed conditions, with elevation of Freq at 3.0 mg/kg and enhanced depression of TVb predominating at 1.0 and 3.0 mg/kg. Administration of METH at 10 mg/kg significantly increased MVb in fentanyl-pretreated subjects, achieving complete reversal of MVb depression to 80.4 % of baseline at 10 min post-administration (t = 30; p = 0.2047). Elevation of depressed MVb by METH was achieved primarily via sharp increases in Freq. However, transient decreases in TVb were observed between 20 and 40 min postadministration before increasing rapidly to the level of controls over the following 10 min (t = 60 – 70). Additionally, 10 mg/kg METH continued to increase MVb values over the course of the recording period to 125.9–128.9 % of baseline by 50 and 55 min post-administration, both of which were significantly higher than controls that had not received fentanyl injections (t = 70 and 75; p = 0.0355 – 0.0354, respectively).

Fig. 7. :

Dose- and time-effects of (±)-methamphetamine (METH) on Minute Volume (MVb = Frequency x Tidal Volume) following pretreatment with 0.3 mg/kg fentanyl (FENT). B = mean baseline MVb, F = fentanyl injection, M = METH injection, S = saline injection. N = 8 per group. No significant differences were detected at baseline across experimental conditions when raw MVb values were compared via one-way ANOVA [F(4, 35) = 1.006; p = 0.4176]. All other details same as Fig. 2.

4. Discussion

The present results show that fentanyl potently and rapidly depresses ventilation in male Swiss-Webster mice with an ED₅₀ of 0.3969 mg/kg and T_max of 10 min. These results also confirm the ability of fentanyl to significantly depress respiration to a clinically relevant degree at doses as low as 0.1 mg/kg (Fig. 2). These doses are clinically relevant when compared via allometric scaling, corresponding to doses of approximately 0.6 – 6 mg in a 70 kg person (i.e. 600 – 6000 –g/70 kg or 8.5 – 85 μg/kg), and are in line with doses that produce life-threatening respiratory depression in opioid-naïve individuals (CDER, 2018; DEA, 2021; Grell et al., 1970). Despite the species differences in sensitivity to opioid-induced respiratory toxicity, the depression of MVb produced by fentanyl at the doses tested in this WBP model provide a translational basis for the subsequent determination of reversal and polydrug toxicity (Shook et al., 1990).

Following the characterization of fentanyl’s dose-effects, naloxone was assessed for its ability to reverse fentanyl-induced respiratory depression over a comparably translational dose-range (Armenian et al., 2017). The results shown in Fig. 3 display the limitations of naloxone to reverse fentanyl-induced decrements in MVb. Following parenteral doses of naloxone corresponding to moderate to high doses in humans (~0.6–6.0 mg/70 kg; Rzasa Lynn and Galinkin, 2018), mice pretreated with 0.3 mg/kg fentanyl displayed a partial reversal of respiratory depression by 5-min with effects dissipating entirely by 40-min post-administration. Only at a dose of naloxone (10 mg/kg) that would correspond to a supratherapeutic dose in humans (~60 mg/70 kg; Rzasa Lynn and Galinkin, 2018) was complete reversal of fentanyl’s depressant effects achieved immediately after administration. For reference, the most widely distributed formulation of intranasal naloxone, Narcan, is currently supplied as 4 mg single dose nasal spray applicators (Moss et al., 2020). Additionally, administration of high dose naloxone (≥10 mg) to humans can precipitate severe adverse effects including life-threatening cardiac arrhythmias and pulmonary edema, especially in persons who chronically use opioids (Clark et al., 2014; Clarke et al., 2005; Rzasa Lynn and Galinkin, 2018). Evidence from several other laboratories corroborate these findings and suggest that fentanyl’s effects on respiration may not be solely attributable to its activity at mu opioid receptors, but also involve muscle rigidity mediated via adrenergic activity (Hill et al., 2020; Torralva et al., 2020; Torralva and Janowsky, 2019). Taken together these results demonstrate that naloxone, while effective at extremely high doses, lacks both sufficient potency and duration to fully reverse the respiratory effects of fentanyl in a life-threatening situation. Finally, as a competitive and selective antagonist of opioid receptors, naloxone can only ameliorate respiratory toxicity arising from activation of those receptors (Handal et al., 1983; Tam, 1985). Presently a large percentage of overdose deaths involve two or more substances some of which are often non-opioid and nonresponsive to naloxone administration (Lyndon et al., 2017; Mattson et al., 2021b). These cases of polysubstance overdose require additional treatment strategies to actively stimulate respiration rather than simply blocking a subset of offending receptors. Taken together, these limitations of naloxone support the development of respiratory stimulants (analeptics) to deploy alongside antagonist-based therapeutics to address the challenges posed by both polysubstance overdose and fentanyl’s unique non-opioid effects.

Starting in 2015, polysubstance use deaths and associated overdoses involving both stimulants and opioids deaths began to exhibit yearly increases (Mattson et al., 2021b). Unlike opioid agonists such as heroin and oxycodone, which elicit their effects via direct activation of central and peripheral opioid receptors, amphetamine-type stimulants have more complex pharmacological profiles. For instance, d-amphetamine and methamphetamine are both prototypical monoamine releasers that rapidly increase synaptic concentrations of dopamine (DA), norepinephrine (NE) and serotonin (5-hydroxytryptamine; 5-HT) via multiple mechanisms. Specifically, the amphetamines act as substrates and competitive inhibitors of monoamine transporters (DAT, NET and SERT) that normally function to rapidly remove neurotransmitters from the synapse after they are released by vesicles (Docherty and Alsufyani, 2021; McCreary et al., 2015; Rothman et al., 2001; Rothman and Baumann, 2003). Through intracellular actions in the presynaptic neuron at vesicular monoamine transporter 2 (VMAT2), trace amine associated receptor 1 (TAAR1), and other target sites, amphetamines increase cytosolic concentrations of monoamines and cause the reversal of transporter activity to promote efflux of transmitters into the synapse. Through both efflux and inhibition of transporter activity the synaptic concentrations of monoamine transmitters are greatly increased, leading to indirect activation of numerous receptor classes and subtypes. The subjective, behavioral, and physiological effects produced by amphetamines have been extensively studied and their myriad of effects have been well documented to include euphoria, wakefulness, improved psychomotor performance, increased blood pressure, increased heart rate, and hyperthermia among others (Hassan et al., 2016; Mendelson et al., 2006). The ability of amphetamines to promote wakefulness and cardiovascular activity via non-opioid mechanisms in conjunction with their increasing rates of co-use alongside fentanyl (either intentional or unintentional) provided a compelling rationale for evaluating their effects on normal and compromised respiration.

Results shown in Fig. 4 and Fig. 5 demonstrate that the administration of 10 mg/kg d-amphetamine or (±)-methamphetamine can robustly increase basal MVb in this mouse WBP procedure. These results are in line with both their established clinical pharmacology and recent evidence indicating that stimulants are increasingly used in combination with fentanyl in an attempt to reduce unwanted adverse effects such as respiratory depression and sedation (Ellis et al., 2018; Lopez et al., 2021; Mendelson et al., 2006; Moody et al., 2020). However, to our knowledge, this is the first time that divergent, dose-dependent effects of amphetamine-type stimulants on measures of respiration have been reported in the scientific literature. When tested under basal conditions, both low (1.0 mg/kg) and moderate (3.0 mg/kg) doses of d-amphetamine significantly depressed MVb for 25 min post-administration, in contrast to the gradual and robust increase in MVb seen in the high dose (10 mg/kg) group (Fig. 4). (±)-Methamphetamine displayed nearly identical bidirectional effects on MVb with a similar degree of depression in the low dose (1.0 mg/kg) condition, but without significant effects in subjects who received a moderate dose (3.0 mg/kg; Fig. 5). Under both basal and fentanyl-depressed conditions the onset of respiratory modulation was rapid (<10 min) following administration of d-amphetamine or methamphetamine, while maximal effects occurred at 30 – 60 min post-administration depending on the dose. In the case of both amphetamines the time course of respiratory modulation (i.e. onset and time to peak effect) was generally similar to what is reported for other behavioral effects such as locomotor activity (Yates et al., 2007). Both the depressant effects seen at lower doses and the gradual elevations of MVb seen following higher doses were found to be the result of opposing effects on breath frequency and tidal volume when those component measures were examined separately (data not shown). Lower doses of both amphetamines primarily depressed tidal volume, with peak depressant effects occurring at 3.0 mg/kg, while higher doses (10 mg/kg) primarily increased frequency without further depression of tidal volume. The differential dose-related effects on frequency and tidal volume seen under both basal and depressed conditions are expected to be the result of the specific pharmacological mechanisms that predominate at individual doses of each amphetamine.

The slight differences in respiratory modulation observed following treatment with d-amphetamine or methamphetamine likely result from the similar, but divergent pharmacology of the two amphetamines. The nearly equipotent effects of d-amphetamine and (±)-methamphetamine on measures of respiration reported here likely reflect the use of enantiopure d-amphetamine vs. racemic methamphetamine as methamphetamine has been demonstrated to be more potent than amphetamine across many behavioral models in rodents (Cho and Segal, 1994; Hall et al., 2008; Rebec, 1998). Relative to methamphetamine, d-amphetamine is more selective for catecholamine release over 5-HT, is slightly less potent at releasing DA, and is slightly more potent at releasing NE in the CNS (Rothman et al., 2001; Rothman and Baumann, 2003). Likewise, the bidirectional effects of both amphetamines on basal and depressed respiration are likely a result of a complex interplay between various monoaminergic receptor subtypes arising from their indirect activation as synaptic levels of DA, NE, and 5-HT are rapidly increased. Amphetamines influence monoamine neurotransmission in a dose-dependent manner due to their effects on DAT, NET and to a lesser extent SERT (Robertson et al., 2009; Rothman et al., 2001). The modest increase in synaptic monoamines produced by lower doses is more reliant on synaptic accumulation after release via vesicular fusion and may activate high-affinity autoreceptors (e.g. D₂ and α₂ receptors) preferentially over post-synaptic heteroreceptors. The preferential activation of inhibitory autoreceptors would be expected to dampen presynaptic activity and may have an overall effect of reducing transmission in these neural pathways (Benoit-Marand et al., 2001; Rougé-Pont et al., 2002). Conversely, as dose increases amphetamines promote monoamine efflux from vesicles, further efflux of monoamines into the synapse via membrane transporters, competitive reuptake inhibition, and trafficking of transporters away from the synaptic membrane (Robertson et al., 2009). The overall effect of these convergent mechanisms is to drastically increase synaptic levels of monoamines, allowing for greater activation of post-synaptic heteroreceptors (e.g., D₁ and α₁ receptors) and subsequent increases in downstream activity. A body of evidence suggests that direct activation of post-synaptic DA receptors stimulates respiration in animal models, lending support to the notion that changes in synaptic transmission of DA may positively correlate with respiratory activity (Lalley, 2008, 2005, 2004).

Similar mechanisms may be at play in the context of NE and 5-HT transmission and their impact on the dose-dependent effects of amphetamines reported here. Evidence from studies in humans and animals suggest that direct activation of α₂-adrenergic receptors with an agonist such as clonidine can produce moderate depression of respiration (Penon et al., 1991). α₂ receptors predominantly serve as presynaptic autoreceptors on NE neurons of the central and peripheral nervous systems where they regulate the release of this transmitter into synapses. Their activation by an exogenous agonist would be expected to inhibit the release of NE and reduce adrenergic signaling. Considering the amphetamines have an equal or greater effect on NE than DA, the differential effects on NE levels in these synapses conferred by low versus high doses of amphetamines may indicate a role of NE signaling in the modulation of respiration. This is particularly relevant in the face of recent evidence that suggests NE receptors may play a role in mediating acute fentanyl toxicity associated with muscle rigidity and laryngospasm (Torralva et al., 2020; Torralva and Janowsky, 2019). Finally, evidence from animal models of apnea, primarily in neonates, suggests a limited role for 5-HT receptors in the modulation of compromised respiration (Corcoran et al., 2014, 2009; Stettner et al., 2008). The synthetic 5HT_1A receptor agonist 8-OH-DPAT counteracts opioid-induced respiratory depression in rats, but a study in humans using the FDA approved 5HT_1A agonist buspirone was unable to demonstrate any antagonism of morphine’s respiratory effects (Guenther et al., 2009, 2002; Oertel et al., 2007). Although d-amphetamine and methamphetamine predominantly affect catecholamine transmission, their lesser effects on 5-HT release may impinge upon these receptor mechanisms at high doses.

Finally, the bidirectional influence of d-amphetamine and racemic methamphetamine on fentanyl-depressed respiration may be caused by pharmacokinetic interactions rather than pharmacodynamic ones, or a combination thereof. Both d-amphetamine and methamphetamine are primarily metabolized via the actions of CYP450 enzymes CYP2D6 and CYP3A4 while the metabolism of fentanyl is mediated primarily by CYP3A4, and to a lesser extent CYP2D6, (Feierman and Lasker, 1996; Lin et al., 1997). Considering both amphetamines tended to enhance fentanyl-depressed respiration via a prolongation of peak depression rather than an increase in its magnitude it is possible that this effect is due to competitive inhibition of fentanyl metabolism preventing its clearance and the corresponding reduction in respiratory effects. However, based on the degree of metabolic overlap, dosage, and similarity of both amphetamines’ effects on MVb observed under both basal and depressed conditions, it is likely that pharmacodynamic factors are primarily responsible for the modulation of fentanyl’s effects (Fig. 4 - Fig. 7). Additionally, the doses of d-amphetamine and methamphetamine that produced the greatest augmentation of fentanyl’s depressant effects were likely not high enough to appreciably inhibit CYP3A4 and CYP2D6 such that fentanyl would be unable to be metabolized (Wu et al., 1997). Further studies of fentanyl metabolism and clearance in combination with amphetamines are needed to clarify the degree to which pharmacokinetic interactions contributed to the reported respiratory effects.

While the present results provide much room for theoretical speculation, it is important to recognize the limitations of the experimental protocol and techniques employed in the present studies. Measurement of ventilatory parameters via WBP has inherent limitations in that it does not allow for the identification of the physiological mechanisms underlying observed changes in ventilation, such as decreased chest wall compliance or airway obstruction that may be induced by high fentanyl doses. Furthermore, WBP alone cannot detect how the observed changes in physical ventilation (rate and depth of inspiration/expiration) affect vital physiological parameters such as blood oxygenation and CO₂ clearance. A related limitation of the experimental protocol was the administration of fentanyl via s.c. injection, which does not reproduce acute toxicity associated with rapid muscle rigidity and laryngospasm (Grell et al., 1970; Torralva and Janowsky, 2019). Although the primary aim of this study was the characterization of respiratory modulation by the selected drugs, the lack of rigidity may reduce the ability to generalize results to situations where muscle rigidity is a contributing factor, as with intravenous use or very high dose toxicity.

Based on the demonstrated ability of d-amphetamine and (±)-methamphetamine to completely reverse fentanyl-induced respiratory depression (Figs. 6-7), future studies to identify the receptor targets underlying these reversal effects are warranted. Furthermore, related investigations into the mechanisms mediating amphetamine-augmentation of fentanyl-induced depression are perhaps just as, if not more important considering the role such an interaction is likely to play in polydrug toxicity.

Abbreviations:

AMPH

d-amphetamine

FENT

fentanyl

METH

methamphetamine

MVb

minute volume

NLX

naloxone

NSO

novel synthetic opioids

SAL

saline

TVb

tidal volume

WBP

whole body plethysmography