Opioid-Like Adverse Effects of Tianeptine in Male Rats and Mice

Abstract

Rationale

Tianeptine is a mu-opioid receptor (MOR) agonist with increasing reports of abuse in human populations. Preclinical data regarding the abuse potential and other opioid-like adverse effects of tianeptine at supratherapeutic doses are sparse.

Objectives

The present study evaluated tianeptine in a rat model of abuse potential assessment and in mouse models of motor, gastrointestinal, and respiratory adverse effects.

Methods

Abuse potential was assessed in adult male Sprague-Dawley rats using an intracranial self-stimulation (ICSS) procedure to determine effects of acute and repeated tianeptine on responding for electrical brain stimulation. Male ICR mice were used to determine the effects of tianeptine in assays of locomotor behavior and gastrointestinal motility. Male Swiss-Webster mice were monitored for respiratory changes using whole-body plethysmography.

Results

In rats, acute tianeptine produced weak and delayed evidence for abuse-related ICSS facilitation at an intermediate dose (10 mg/kg, IP) and pronounced, naltrexone-preventable ICSS depression at a higher dose (32 mg/kg, IP). Repeated 7-day tianeptine (10 and 32 mg/kg/day, IP) produced no increase in abuse-related ICSS facilitation, only modest tolerance to ICSS depression, and no evidence of physical dependence. In mice, tianeptine produced dose-dependent, naltrexone-preventable locomotor activation. Tianeptine (100 mg/kg, SC) also significantly inhibited gastrointestinal motility and produced naloxone-reversible respiratory depression.

Conclusions

Tianeptine presents as a MOR agonist with resistance to tolerance and dependence in our ICSS assay in rats, and it has lower abuse potential by this metric than many commonly abused opioids. Nonetheless, tianeptine produces MOR agonist-like acute adverse effects that include motor impairment, constipation, and respiratory depression.

INTRODUCTION

Tianeptine is an atypical antidepressant first patented in the early 1970s (Malen et al., 1973) alongside many other tricyclic compounds. A number of clinical trials demonstrated that tianeptine was effective as an antidepressant, comparable to other clinically-approved antidepressants including tricyclics and serotonin-selective reuptake inhibitors (SSRIs) (Kasper & Olié, 2002; Wagstaff et al., 2001); however, though it is similar in structure to classic tricyclic antidepressants, it is mechanistically distinct as it does not exhibit affinity for monoamine transporters or receptors. Although several alternative mechanisms have been proposed over the years, Gassaway et al. (2014) discovered that tianeptine acts as a low-affinity but high-efficacy mu opioid receptor (MOR) agonist. Subsequent experiments in rats suggested that tianeptine’s agonist activity at MOR is necessary for its antidepressant and other behavioral effects such as antinociception (Samuels et al., 2017). However, the degree to which tianeptine produces opioid-like abuse potential and other common opioid side effects remains unclear. Clinical and preclinical testing of tianeptine originally indicated a favorable safety profile without strong indicators of abuse liability (Wilde & Benfield, 1995), but in the years following the identification of tianeptine as an MOR agonist, poison control centers in the United States have seen a sharp increase in the number of reports related to tianeptine exposure (El Zahran et al., 2018; Rushton et al., 2021). Tianeptine has been easily obtainable in recent years by purchasing over the Internet or in local retail establishments, leading to scheduling actions at the state level in Michigan (2019) and Alabama (2021). Although these states have controlled tianeptine as C-II, tianeptine is not controlled at the federal level in the United States as of January 2022. Its increasing use has prompted warnings from the Food and Drug Administration and the Centers for Disease Control, and case reports have been published implicating tianeptine in overdoses, some of which have been fatal (Bakota et al., 2018; Dempsey et al., 2017; El Zahran et al., 2018; Proença et al., 2007; Rushton et al., 2021; U.S. Food and Drug Administration, 2018). The current opioid crisis and the emergence of tianeptine as a novel abused MOR agonist prompts a need for pharmacological characterization (Baumann et al., 2018).

Given the clinical status of tianeptine as an antidepressant with a tricyclic structure, most existing data address the stimulant-like abuse liability of tianeptine. For example, one study in human patients compared psychostimulant-like subjective effects of methylphenidate with tianeptine at a dose of 75 mg, twice the therapeutic dose for the treatment of depression. Methylphenidate produced significant stimulant effects, but tianeptine effects were not different from placebo (Bernard et al., 2011). Notably, this study was conducted prior to the discovery of tianeptine as a MOR agonist, and although they used a supratherapeutic dose, it is considerably smaller than the doses reported in cases of tianeptine abuse, which can exceed 1000 mg per day (Lauhan et al., 2018; Smith et al., 2021). In another study, the subjective effects produced by tianeptine in rats were evaluated using a drug-discrimination procedure, a type of procedure that can be used in abuse liability assessment; rats successfully learned to discriminate 10 mg/kg tianeptine from saline, but the only comparator drugs tested were monoaminergic antidepressant medications with low abuse potential (e.g. fluoxetine) and caffeine, and none of these produced full substitution for the tianeptine training stimulus (Alici et al., 2006). In the only preclinical evidence for tianeptine abuse potential, 30 mg/kg tianeptine induced a MOR-dependent conditioned place preference in rats similar to that of 5 mg/kg morphine (Samuels et al., 2017). This finding is consistent with the hypothesis that tianeptine has opioid-like abuse potential, but additional assessments of abuse liability and other MOR-mediated adverse effects would aid in making evidence-based decisions about drug regulation.

Accordingly, in this study, we evaluated tianeptine in a preclinical models of abuse potential and other opioid adverse effects that have been used to examine other opioids. An intracranial self-stimulation (ICSS) procedure in rats was used to evaluate the abuse potential of tianeptine. ICSS procedures have been used to evaluate abuse potential of both therapeutic and illicit drugs from a broad range of pharmacological classes, and the predictive validity of ICSS to identify drugs with abuse potential is similar to that of preclinical drug self-administration procedures (Baird et al., 2021; Negus & Miller, 2014). In ICSS procedures, which are usually conducted in rats, an electrode is surgically implanted into a brain reward area, and the subject is able to depress a lever to self-administer pulses of electrical stimulation via the electrode. Many drugs with abuse liability increase rates of ICSS responding, a phenomenon known as “ICSS facilitation.” MOR agonists have been extensively studied in these procedures and typically present with predominant ICSS depression after acute dosing but robust abuse-related ICSS facilitation after repeated administration (Altarifi et al., 2012, 2013, 2017, 2020; Miller et al., 2015; Negus & Moerke, 2019). We also evaluated tianeptine effects on three other endpoints of opioid adverse effects in mice. These effects included (1) locomotor behavior, in which MOR agonists typically stimulate locomotor activity as a sign of motor disruption (Diester et al., 2021; Rethy et al., 1971; Varshneya et al., 2019, 2021), (2) gastrointestinal motility, which is typically inhibited by MOR agonists (Ross et al., 2008), and (3) respiration, which is typically depressed by MOR agonists (Hill et al., 2018, 2020; Varshneya et al., 2022).

METHODS

Subjects.

Adult male Sprague-Dawley rats (N=25; Envigo, Indianapolis, IN, USA) were individually housed with free access to food and water in the home cage and maintained on a 24-hour light/dark cycle with lights on from 6:00 AM to 6:00 PM. ICR (N=28) or Swiss-Webster (N=16) male mice (Harlan Laboratories, Frederick, MD) were 6–8 weeks old upon arrival to the laboratory and were group housed with ad libitum access to food (Teklad LM-485 Mouse/Rat Diet, Harlan Laboratories) and water in a temperature-controlled room with a 12-hour light/dark cycle. For studies of locomotion and gastrointestinal inhibition, ICR mice were maintained on a conventional light:dark cycle with lights on from 6:00 AM to 6:00 PM. For studies of respiration, Swiss-Webster mice were maintained under a reverse light:dark cycle with lights on from 6:00 PM to 6:00 AM. The housing facility was maintained at a temperature of 22 ± 2 °C and humidity of 50 ± 5%. Behavioral experiments were performed between the hours of 9:00 AM and 6:00 PM. Animal-use protocols were approved by the Institutional Animal Care and Use Committee and complied with the National Research Council Guide for the Care and Use of Laboratory Animals.

Drugs.

Tianeptine sodium salt (Cayman Chemical, Ann Arbor, MI, USA), naltrexone hydrochloride, naloxone hydrochloride, and morphine sulfate (National Institute on Drug Abuse Drug Supply Program, Bethesda, MD) were dissolved in bacteriostatic saline (Hospira, Inc., Lake Forest, IL, USA) for intraperitoneal (IP) or subcutaneous (SC) administration. Injection volumes were held constant at 1 mL/kg.

Intracranial Self-Stimulation in Rats

Surgical Procedure.

Rats were anesthetized by inhalation of 2.5–3.0% isoflurane (Zoetis Inc., Parsippany, NJ, USA) in oxygen until the rat was unresponsive to toe pinch, and anesthesia was maintained throughout surgery. A stainless-steel electrode was implanted using a stereotaxic surgical procedure. The cathode targeted the medial forebrain bundle at the level of the lateral hypothalamus (2.8 mm posterior to bregma, 1.7 mm lateral to the midsagittal suture, and 8.8 mm ventral to the exterior surface of the skull), and the anode was grounded by coiling around one of the screws anchored into the dorsal surface of the skull. The electrode, grounding wire, and screws were embedded in acrylic resin and permanently affixed to the skull. Ketoprofen (5 mg/kg, IP) was administered as an analgesic during surgery as well as 24 h post-surgery. Rats were monitored and allowed to recover for at least one week prior to ICSS training.

Apparatus.

ICSS sessions occurred inside modular operant test chambers (29.2 × 30.5 × 24.1 cm) (Med Associates, St. Albans, VT, USA) made of stainless steel and clear polycarbonate with a grid floor. Test chambers were contained within opaque sound-attenuating cabinets. The test chamber contained a response lever positioned 3 cm above the floor and 7.6 cm below three stimulation lights (red, yellow, and green), a 2 W house light, and an ICSS stimulator. A bipolar cable and swivel commutator (Model SL2C; Plastics One, Roanoke, VA, USA) connected the ICSS stimulator to the electrode. Data collection and ICSS programming were controlled using a computer system running Med-PC IV software (Med Associates, St. Albans, VT, USA).

Training.

Rats were trained to depress a lever under a fixed-ratio 1 (FR 1) schedule to receive electrical brain stimulation consisting of a 0.5-s train of 0.1 ms electrical pulses. Electrical stimulation was accompanied by the simultaneous illumination of the stimulus lights above the lever. Additional lever presses during each 0.5-s period of electrical stimulation had no scheduled consequences. Initial training consisted of 30- to 60-minute sessions, during which the frequency of electrical stimulation was held constant at 126 Hz and the amplitude was adjusted to a magnitude sufficient to maintain responding in excess of 30 stimulations per minute. Once sufficient rates of responding were acquired, rats entered the second phase of training in which frequency manipulations were introduced. These sessions consisted of 3 to 6 consecutive 10-minute components. Each component was divided into ten one-minute frequency trials, and the frequency of stimulation decreased across trials from 158 to 56 Hz in 0.05 log increments. Each frequency trial began with a 10-s time-out period, during which five non-contingent presentations of electrical stimulation at the frequency for that trial were presented noncontingently, and was followed by a 50-s response period, during which lever presses resulted in contingent electrical stimulation under the FR1 schedule. Amplitudes were adjusted until individual rats responded at high rates for the highest 3 to 4 frequencies and at low rates for the remaining frequencies. Once responding was stable during 3-component training sessions for at least three consecutive days, amplitudes were fixed for the remainder of the study (100–280 μA).

Acute testing procedures.

Tianeptine effects on ICSS were evaluated in a series of three experiments conducted in separate cohorts of rats, and the experimental design was similar to that used previously for other opioid agonists (Altarifi et al., 2017, 2020). In the first experiment, the potency, effectiveness, and time course of acute tianeptine effects on ICSS were determined in a group of six rats. Test sessions consisted of three daily baseline components followed by IP injection of tianeptine. Test components began 10, 20, 30, 40, and 100 minutes post-administration. Doses tested were 1.0, 3.2, 10, and 32 mg/kg as well as saline, and dose order was randomized across rats with a Latin-square design. These doses were selected based upon preliminary studies used to identify behaviorally active doses. These pilot studies also indicated increased variability in drug effects for the SC route of administration in rats, therefore IP was chosen as the preferred route for ICSS experiments. Testing was performed on Tuesdays and Fridays, and three-component training sessions were conducted on all other weekdays.

Naltrexone antagonism.

The first experiment indicated that tianeptine produced transient and weak ICSS facilitation at a dose of 10 mg/kg and more robust and sustained ICSS depression at 32 mg/kg. The weak ICSS facilitation produced by 10 mg/kg tianeptine was considered too small for reliable detection of antagonism, and as a result, the second experiment evaluated naltrexone antagonism of the more prominent ICSS rate-decreasing effects of 32 mg/kg tianeptine. Naltrexone was selected as the most appropriate agent for antagonist blockade of opioid-induced adverse effects due to its long duration of action and its role as a clinically approved treatment for opioid use disorder. Six rats were used for this study. Test sessions consisted of three daily baseline components followed first by SC injection of either saline or 0.1 mg/kg naltrexone and then 10 min later by IP injection of either saline or 32 mg/kg tianeptine. Test components began 10 and 20 minutes after tianeptine administration. Thus, rats received a total of four different treatments in a 2×2 experimental design (saline or naltrexone + saline or tianeptine), and treatment order was randomized across rats using a Latin-square design. As with acute-dosing studies, testing occurred on Tuesdays and Fridays, and training sessions were conducted on other weekdays.

Repeated daily tianeptine treatment.

Previous studies have found that many MOR agonists produce primarily ICSS rate-decreasing effects in opioid-naïve rats, but that repeated daily treatment can produce tolerance to rate-decreasing effects and enhanced expression of abuse-related ICSS facilitation (Altarifi et al., 2013; Miller et al., 2015; Moerke & Negus, 2019, 2021, 2021). Accordingly, a third experiment evaluated effects of repeated daily tianeptine treatment using a 9-day protocol as described previously with other MOR agonists. On Day 1, rats underwent a three-component baseline session, followed by four consecutive 30-minute test periods. Each test period consisted of a 10-min timeout followed by two 10-min test components. Saline was administered IP at the beginning of the first test period, and increasing doses of 3.2, 10, and 32 mg/kg tianeptine were administered IP at the start of each subsequent test period. On Days 2–7, tianeptine was administered by IP injection at a dose of 10 mg/kg/day in one set of six rats, and 32 mg/kg/day in a second set of seven rats. On Days 2, 3, 4, and 7, daily tianeptine injections were administered in the context of ICSS test sessions that consisted of three baseline components followed first by a 10-min timeout period during which the tianeptine dose was administered and then by two test components. On Days 5 and 6 (i.e. over a weekend), tianeptine was administered but no ICSS sessions were conducted. On Day 8, the effects of saline and increasing tianeptine doses were redetermined as on Day 1. On Day 9, a similar sequential dosing regimen was conducted for morphine, with the exceptions that test-period timeout durations were extended to 30 min to accommodate the slower onset of morphine effects. As a result, the total duration of each test period was 50 min (30 min timeout followed by two 10-min test components). Thus, saline and doses of 0.32, 1.0, and 3.2 mg/kg morphine were administered at the start of each sequential test period. We have shown previously that this morphine dosing regimen produces only ICSS depression in opioid-naïve rats but produces ICSS facilitation in opioid-experienced rats (Altarifi et al., 2017, 2020; Miller et al., 2015; Negus & Moerke, 2019). Here, we tested whether repeated tianeptine could produce cross tolerance to rate-decreasing effects of morphine and enhanced expression of ICSS facilitation by morphine.

Data Analysis.

The first baseline component of each session was considered an acclimation component, and data were discarded. The primary dependent variable for all remaining components was the reinforcement rate in stimulations per trial during each frequency trial. To normalize these raw data, reinforcement rates from each trial in each rat were converted to percent maximum control rate (% MCR), which was defined as the mean of the maximal rates observed during the second and third ‘baseline’ components for that rat on that day. Thus, % MCR = ((rate during a frequency trial)/(MCR)) × 100. Normalized ICSS rates at each frequency were averaged across test components within each rat and then across rats to yield a ‘frequency-rate’ curve for each experimental manipulation. Two-way ANOVA was used to compare frequency-rate curves, with ICSS frequency as one variable and dose as the second variable. ANOVA results are reported below only for the main effect of dose/treatment and the frequency × dose/treatment interaction, because the main effect of frequency was always significant. To provide an additional summary measure of ICSS performance in these studies, the total number of stimulations per component was calculated as the average of the total stimulations delivered across all 10 frequency trials of each component. Data are expressed as a percentage of the total stimulations per component earned during the daily baseline. Thus, the percentage of baseline total stimulations was calculated as (mean total stimulations during test components/mean total stimulations during baseline components) × 100. Data were analyzed by one or two-way ANOVA as appropriate. For all analyses, the Geisser-Greenhouse correction was used to adjust for unequal standard deviations. A significant ANOVA was followed by Dunnett’s post hoc test, and the criterion for significance was set at p < 0.05.

Opioid-Like Side Effects in Mice

Locomotor Behavior.

Horizontal locomotor activity was assessed in male ICR mice as described previously (Diester et al., 2021). Locomotor activity chambers (16.8 × 12.7 cm² floor area × 12.7 cm high) were housed in sound-attenuating cabinets (Med Associates, St. Albans, VT). Each box had black plexiglass walls, a clear plexiglass ceiling equipped with a house light, bar floors, and six photobeams arranged at 3 cm intervals across the long wall and 1 cm above the floor. The primary dependent variable was the total number of beam breaks during each 60-min session, excluding consecutive interruptions of the same beam. Test sessions were conducted twice a week with at least 48 h between sessions. Different groups of n=6 mice were used to test morphine (vehicle, 1.0–100 mg/kg) and tianeptine (vehicle, 10–100 mg/kg). The doses chosen for testing were identified in preliminary locomotor experiments, and the highest dose (100 mg/kg) was used for subsequent studies of gastrointestinal and respiratory function in mice. Within each group, all mice received all doses, dose order was counterbalanced across mice using a Latin square design, and test drug was administered SC 5 min before the start of each 60-min session. After completion of the tianeptine dose-effect curve, this group was also used to assess effects of naltrexone (vehicle, 0.1, and 1.0 mg/kg, 10 min pretreatment) on locomotor activation induced by 100 mg/kg tianeptine.

Gastrointestinal Inhibition.

Tianeptine effects on gastrointestinal transit were determined as described previously (Ross et al., 2008). Male ICR mice were fasted for 24 h with free access to water and had access to 5% dextrose for the first 8 h of the fasting period. Two groups of mice (n=8 per group) were treated with SC injection of either saline or tianeptine (100 mg/kg) and 15 min later given an oral gavage consisting of 5% aqueous suspension of charcoal (10 μl/g body weight; Sigma, C7606-125G) in a 10% gum Arabic solution. At 30 min after the administration of the charcoal meal, mice were euthanized by cervical dislocation, and the small intestine from the jejunum to the caecum was dissected and placed in cold saline to stop peristalsis. The distance traveled by the leading edge of the charcoal meal was measured relative to the total length of the small intestine, and the percentage of intestinal transit for each animal was calculated as percentage transit (charcoal distance) / (small intestinal length) × 100.

Respiratory Depression.

Ventilation was assessed in male Swiss-Webster mice using procedures similar to those described previously (Hill et al., 2018, 2020; Varshneya et al., 2022) using whole-body plethysmography chambers (Data Sciences International, St. Paul, MN) to measure respiratory rate (breath frequency), breath tidal volume, and minute volume. Respiratory studies were conducted under 660nm illumination, a wavelength with limited visibility to mice, in order to promote maintenance of subjects in the dark phase of their cycle. Testing mice during their active phase (dark cycle) promoted higher basal respiration rates and prevented mice from sleeping during the procedure. Mice were placed in the chambers for a 30-min habituation session the day before testing under ambient air conditions. On the test day, plethysmography chambers were supplied with 5% CO₂, 21% O₂, and a balance of N₂. This CO₂-enriched mixture has been found to minimize ventilatory variability over time, improve sensitivity for detecting opioid-induced respiration depression, and is devoid of anxiogenic effects (Crowley et al., 2021; Hill et al., 2016; Varshneya et al., 2022). 100-min test sessions were divided into three phases. The initial “baseline” phase lasted 20 min and always began with a SC saline injection. The second “agonist” phase also lasted 20 min and began with SC injection of either saline or 100 mg/kg tianeptine. The final “naloxone-reversal” phase lasted 60 min and began with SC injection of either saline or 1.0 mg/kg naloxone. Naloxone was used in place of naltrexone for this study as it is the MOR antagonist used clinically to reverse respiratory depression associated with opioid overdose. Each mouse was tested only once, and four different groups of mice were used to test four different treatment combinations in a 2×2 experimental design (Saline or Tianeptine + Saline or Naloxone; n=6–8 per group). The primary dependent measures recorded were breath frequency per minute, tidal volume per breath, and minute volume (i.e., breath frequency per minute x tidal volume) during 5 min bins, and the primary dependent measure for data analysis was minute volume expressed as a percentage of each mouse’s mean minute volume during the final 5 min of the baseline period.

Data Analysis.

Data were analyzed by t-test, one-way ANOVA, or two-way ANOVA as appropriate. All tests included corrections for unequal standard deviations (Welch’s correction for t-test, Geisser-Greenhouse for ANOVAs). A significant one-way ANOVA was followed by Dunnett’s post-hoc test, and a significant interaction in a two-way ANOVA was followed by the Holm-Sidak post hoc test, to compare drug effects with vehicle effects. The criterion for significance was p < 0.05 for all analyses.

RESULTS

Intracranial Self-stimulation in Rats.

Under baseline conditions, electrical brain stimulation maintained frequency-dependent increases in ICSS rates. Across all rats, the mean MCR ± SEM was 57.4 ± 2.0 stimulations per trial and mean total number of stimulations ± SEM was 261.4 ± 15.6 stimulations per component. Figure 1 shows that acute administration of tianeptine produced dose- and time-dependent changes in ICSS with a mixed profile of facilitation and depression. ICSS was stable over time after saline administration, and 1.0 and 3.2 mg/kg tianeptine did not significantly affect ICSS rates at any time compared to saline. A higher dose of 10 mg/kg tianeptine decreased high rates of ICSS responding maintained by high brain-stimulation frequencies (141 and 158 Hz) 10 and 20 min after tianeptine administration; however, after 30 min, these rate decreasing effects were no longer apparent, and significant ICSS facilitation was observed at intermediate brain-stimulation frequencies (89–100 Hz). The highest dose of 32 mg/kg tianeptine nearly eliminated ICSS responding after 10 and 20 min. Responding started to recover after 30 min, and rate-decreasing effects were no longer significant after 40 min. Figure 2 shows that 0.1 mg/kg naltrexone blocked ICSS depression induced by 32 mg/kg tianeptine. Naltrexone alone did not affect ICSS.

Figure 1.

Dose- and time-dependent effects of acute IP tianeptine on ICSS. Panel A shows the time course of drug effects on ICSS summarized as percent baseline number of stimulations per component delivered across all stimulation frequencies for all doses tested. Abscissa: Time post-administration of IP tianeptine. Ordinate: Percent baseline number of stimulations per component. Each data point represents the mean ± SEM of six rats, and filled points represent significantly different from saline vehicle at that time point using a repeated-measures two-way ANOVA followed by a Dunnett post-hoc test (p < 0.05). Two-way ANOVA for significant time x dose interaction: F(2.384, 11.92) = 5.798; P=0.0144. Panels B-D show frequency-rate curves after IP treatment with saline, 10 mg/kg, and 32 mg/kg tianeptine for test components that began 10, 20, and 30 minutes after tianeptine administration. Abscissae: Frequency of electrical brain stimulation in Hz (log scale). Ordinates: Number of stimulations per trial expressed as a percentage of daily maximum control rate (MCR). Each data point represents the mean ± SEM from six rats, and filled points represent frequencies at which ICSS rates were different compared to saline as determined by a repeated-measures two-way ANOVA followed by a Dunnett post-hoc test (p < 0.05). Two-way ANOVA for significant frequency x dose interactions were as follows: 10 min (F(3.511, 17.55) = 5.095; p = 0.0081), 20 min (F(4.001, 20.00) = 5.561; p = 0.0035), 30 min (F(3.669, 18.34) = 4.637; p = 0.0105).

Figure 2.

Antagonism of IP-tianeptine-induced ICSS depression by SC naltrexone pretreatment. Panel A shows the effect of treatment on ICSS expressed as a percent of baseline number of total stimulations per component. Bars represent mean ± SEM (n=6), and asterisks signify a significant difference between treatments. Two-way ANOVA (F(1, 5) = 86.77; p = 0.0002)) indicates a NTX x TIA interaction, and Dunnett’s post-hoc test indicates significant effects of all treatment conditions compared to SAL/TIA treatment (**p < 0.01, ****p < 0.0001). Panel B shows frequency-rate curves for each treatment condition. Abscissa: Frequency of electrical brain stimulation in Hz (log scale). Ordinate: Number of stimulations per trial expressed as a percentage of daily MCR. Each point shows the mean ± SEM of six rats, and filled points represent frequencies at which ICSS rates were different compared to SAL-SAL as determined by a repeated-measures two-way ANOVA followed by Dunnett’s post-hoc test (p < 0.05). Two-way ANOVA indicated a significant Frequency x Treatment interaction (F(4.197, 20.98) = 21.66; p < 0.0001)SAL: saline (SC or IP); NTX: 0.1 mg/kg naltrexone (SC); TIA: 32 mg/kg tianeptine (IP).

Figure 3 shows that 7-day exposure to 10 or 32 mg/kg/day tianeptine failed to enhance tianeptine-induced ICSS facilitation. On Day 1, sequential doses of 3.2 and 10 mg/kg tianeptine produced little or no effect on ICSS, whereas 32 mg/kg tianeptine decreased ICSS as in acute studies. During repeated dosing, daily baseline sessions were conducted approximately 24 hours after each drug administration, and these baseline frequency-rate curves were stable while tianeptine continued to produce little effect or ICSS depression at doses of 10 and 32 mg/kg/day, respectively (data not shown). After seven days of daily exposure to 10 or 32 mg/kg tianeptine, the sequential-dosing procedure was repeated. On Day 8, neither 3.2 nor 10 mg/kg tianeptine altered rates of responding for either group. The higher dose of 32 mg/kg tianeptine still robustly depressed total ICSS stimulations; however, this effect was significant at fewer individual brain stimulation frequencies on Day 8 compared to Day 1, suggesting minor tolerance to the rate-decreasing effects of 32 mg/kg tianeptine. There was no significant facilitation of ICSS at any frequency or at any dose following repeated treatment of 10 or 32 mg/kg/day tianeptine. Figure 4 shows no significant effect of either 1 or 3.2 mg/kg doses of morphine compared to saline following 8 days of tianeptine administration.

Figure 3.

Effects of IP tianeptine on ICSS before and after repeated tianeptine administration. Panels A and D display summary data from Days 1 and 8 in rats receiving repeated 10 mg/kg (A) and 32 mg/kg (D) tianeptine, indicating significant depression of ICSS rates at 32 mg/kg. Abscissae: Dose of tianeptine. Ordinates: Percent baseline number of stimulations per component. Each data point shows the mean ± SEM of 7 rats, and asterisks represent significantly different ICSS rates compared to saline vehicle. Two-way ANOVA of summary data for A indicated a significant effect of Dose (F(1.324, 7.944) = 23.27; p = 0.0009), but not of Day, and the Dose x Day interaction was also not significant (p>0.05). Dunnett’s post-hoc test on data collapsed across days indicated a significant difference between saline vehicle and 32 mg/kg (** p < 0.01). Two-way ANOVA of summary data for D indicated a significant effect of Dose (F(1.355, 8.133) = 9.222; p= 0.0119), but not of Day, and the Dose x Day interaction was also not significant (p > 0.05). Dunnett’s post-hoc test on data collapsed across days indicated a significant difference between saline vehicle and 32 mg/kg (** p < 0.01). Panels B, C, E, and F display frequency-rate curves for sequential dosing of tianeptine before and after repeated daily administration of 10 mg/kg (B & C) or 32 mg/kg (E & F) tianeptine. Abscissae: Frequency of electrical brain stimulation in Hz (log scale). Ordinates: Number of stimulations per trial expressed as a percentage of daily MCR. Each data point represents the mean ± SEM of 7 rats, and filled points represent frequencies at which ICSS rates were different compared to saline as determined by a repeated-measures two-way ANOVA followed by Dunnett’s post-hoc test (p < 0.05). Two-way ANOVA for significant frequency x dose interactions were as follows: 10 mg/kg Day 1 (F(3.695, 22.17) = 9.230; p = 0.0002), 10 mg/kg day 8 (F(3.547, 21.28) = 6.337; p = 0.0021), 32 mg/kg Day 1 (F(3.965, 23.79) = 10.92; p < 0.0001), 32 mg/kg Day 8 (F(3.334, 20.01) = 4.928; p = 0.0085).

Figure 4.

Effects of SC morphine on ICSS after repeated daily IP tianeptine administration. Panels A and C display summary data from Days 9. Abscissae: Dose of morphine. Ordinates: Percent baseline number of stimulations per component. Each data point represents the mean ± SEM (n=7). One-way ANOVA of summary data for A did not indicate a significant effect of Dose (p > 0.05). One-way ANOVA of summary data for C also did not indicate a significant effect of Dose (p > 0.05). Panels B and D display frequency-rate curves for sequential dosing of morphine before and after repeated daily administration of 10 mg/kg (B) or 32 mg/kg (D) tianeptine. Abscissae: Frequency of electrical brain stimulation Ordinates: Number of stimulations per trial expressed as a percentage of daily MCR. Each data point represents the mean ± SEM (n=7). Two-way ANOVA did not indicate significant frequency x dose interactions for either B or D.

Opioid-Like Adverse Effects in Mice.

Figure 5 shows effects of tianeptine on locomotor activity, gastrointestinal transit, and respiration in mice. Tianeptine (10–100 mg/kg) produced a dose-dependent and naltrexone-preventable increase in locomotor activity, and relative to morphine, tianeptine was less potent but produced a similar maximum effect (Figure 5A, 5B). Tianeptine (100 mg/kg) also significantly inhibited gastrointestinal transit (Figure 5C). Lastly, 100 mg/kg tianeptine produced a sustained respiratory depressant effect that could be transiently reversed by 1.0 mg/kg naloxone (Figure 5D). The depression of respiration determined by minute volume was effected by decreases in both frequency and tidal volume (Supplemental Figure S1).

Figure 5.

Effects of tianeptine on unconditioned behaviors in mice. Panel A shows dose-dependent increases in locomotion by SC tianeptine and morphine. Abscissa: Dose of tianeptine or morphine in mg/kg. Ordinate: Total movement counts as number of photobeam breaks per 60-minute session. Bars represent mean ± SEM (n=6), and filled points indicate a significant difference from saline by one-way ANOVA followed by Dunnett’s post-hoc test (p < 0.05). One-way ANOVA indicated significant effects of tianeptine (F(1.88, 20.72)=86.70; p < 0.0001) and morphine (F(2.05, 10.26)=11.18; p=0.0025). Panel B shows dose-dependent antagonism by SC naltrexone of locomotor activation induced by SC tianeptine (100 mg/kg). One-way ANOVA indicated a significant effect of naltrexone (F(1.86, 18.55)=31.80; p<0.0001). Asterisks indicate a significant difference from saline vehicle by Dunnett’s post-hoc test (*p<0.05; **p<0.01). Panel C shows inhibition of gastrointestinal transit by tianeptine (100 mg/kg). Asterisks indicate a significant difference from saline by an unpaired t test with Welch’s correction (t=8.74, df=9.30; p<0.0001). Panel D shows naloxone-reversible depression of respiration by tianeptine. Abscissa: Time in minutes post-administration of tianeptine (100 mg/kg). Ordinate: Minute volume as percentage of baseline. Bars indicate mean ± SEM (n=6–8), and filled points indicate a significant difference from saline-saline by two-way ANOVA followed by a Holm-Sidak post-hoc test (p<0.05). Two-way ANOVA indicated a significant treatment x time interaction (F(48, 416)=13.07; p<0.0001). NX = naloxone. Mean ± SEM (mL/min) baseline data for each group is as follows: Sal/Sal: 151.89 ± 5.42; Sal + 1.0 NX: 153.24 ± 11.27; 100 Tianeptine + Sal: 169.73 ± 3.66; 100 Tianeptine + NX: 169.37 ± 7.30.

DISCUSSION

This study used rodent models to evaluate opioid-like adverse effects of tianeptine. There were three main findings. First, tianeptine produced significant but weak, delayed, and transient evidence for abuse potential in the ICSS procedure. Moreover, in contrast to results reported previously with other MOR agonists, repeated tianeptine treatment failed to increase expression of abuse-related ICSS facilitation. Rather, the predominant effect after both acute and repeated tianeptine was non-specific ICSS depression by a high dose of 32 mg/kg. Second, tianeptine produced MOR agonist-like changes in locomotor activity, gastrointestinal transit, and respiration in mice. Lastly, both tianeptine-induced ICSS depression in rats and locomotor activation in mice were blocked by naltrexone pretreatment, and tianeptine-induced respiratory depression in mice was reversed by naloxone treatment. This sensitivity to opioid antagonists is consistent with MOR mediation of these tianeptine effects. Overall, these results suggest that tianeptine has low abuse potential relative to many other MOR agonists, but that it can nonetheless produce problematic MOR-mediated motor impairment, constipation, and respiratory depression.

Relative to many other MOR agonists tianeptine produced weak evidence for abuse potential in the ICSS procedure. Acute tianeptine in tianeptine-naïve rats produced facilitation that was weak, delayed, and transient, with facilitation emerging only 30 min after drug administration. The more prominent effect was naltrexone-preventable ICSS depression by a high dose of tianeptine. This is consistent with ICSS profiles produced by other acutely administered MOR agonists such as morphine and tramadol (Altarifi et al., 2020; Altarifi & Negus, 2011). However, tianeptine effects diverged from effects of other MOR agonists in the chronic-treatment study. We have reported previously that repeated treatment with many other MOR agonists produces tolerance to ICSS rate-decreasing effects and enhancement of abuse-related ICSS facilitation produced both by the chronically administered opioid and by other MOR agonists (Moerke & Negus, 2019, 2021; Negus & Moerke, 2019). Although modest tolerance to the rate-decreasing effects of tianeptine on ICSS was observed in the present study, repeated administration of high doses of tianeptine produced only a trend toward enhanced ICSS facilitation for tianeptine and morphine, but neither of these effects reached the criterion for statistical significance. The evidence for only weak tolerance to the rate-decreasing effects of tianeptine is consistent with preclinical results using a hot-plate test in mice, where repeated exposure to tianeptine did not produce antinociceptive tolerance (Samuels et al., 2017).

Repeated opioid administration can also produce physical dependence that can manifest in ICSS procedures as opioid withdrawal-induced rightward shifts in ICSS frequency-rate curves (Negus & Moerke, 2019). In this study, however, baseline frequency-rate curves were stable during repeated tianeptine treatment, indicating that the tianeptine dosing regimen did not lead to withdrawal-induced rightward shifts of the frequency-rate curve. Although somatic withdrawal-related endpoints were not formally monitored, no overt withdrawal signs were apparent (e.g. diarrhea, loss of body weight). This lack of withdrawal signs is consistent with other preclinical evidence for a lack of somatic withdrawal signs after chronic tianeptine (Samuels et al., 2017). Possible explanations for the weak evidence for tolerance and physical dependence after repeated tianeptine treatment may involve tianeptine’s high MOR efficacy and short duration of action. Because of its high efficacy and short duration of action, tianeptine would be expected to produce MOR-mediated effects at relatively low levels of MOR occupancy and for relatively short periods of time during our once-daily dosing regimen. Together, these factors likely resulted in low probability of ligand binding to any given receptor and hence few opportunities for agonist-induced receptor desensitization and/or downregulation. Overall, the high efficacy and short duration of action of tianeptine suggests that frequent administration of high doses may be required to produce tolerance and dependence. Both the daily dosing regimen in our study and the twice-daily dosing regimen used by Samuels et al. (2017) was insufficient to achieve either tolerance or dependence. The low potency of tianeptine (~3- to 10-fold less potent than morphine) is an additional practical obstacle to sufficient dosing for tolerance and dependence.

The absence of tolerance to tianeptine-induced rate-decreasing effects appears to have limited the effectiveness of tianeptine to produce abuse-related ICSS facilitation; however, tianeptine did produce a range of other opioid-like adverse effects. As noted above, naltrexone-preventable ICSS depression in rats is one sign of these opioid-like adverse effects, and the present study also demonstrated other typical opioid-like adverse effects in mice. First, tianeptine produced a naltrexone-preventable increase in locomotor activity in this study, which is in agreement with another recent study that observed increases in locomotor activity at similar doses (Samuels et al., 2017). This effect was similar to that produced by morphine in this study and to the effects of morphine and other MOR agonists reported previously (Diester et al., 2021; Varshneya et al., 2019; Osborn et al., 2010; Frischknecht et al., 1983). It should be noted that locomotor activation is especially prominent as a sign of MOR-mediated motor disruption in mice, and insofar as this effect appears to require activation of mesolimbic dopaminergic pathways also involved in reward (Botz-Zapp et al., 2021; Severino et al., 2020; Urs & Caron, 2014; Funada et al., 1993), it may also be suggestive of abuse potential for tianeptine. In other species, MOR agonist-induced motor disruption typically manifests as a decrease in motor activity. In particular, rats did not show increased locomotion during tianeptine treatment in this study; instead, rats appeared cataleptic and immobile at high tianeptine doses. Thus, tianeptine-induced ICSS depression was not driven by an allocation of behavior toward competing perseverative locomotion but rather by an overall decrease in behavior. It should also be noted that the present report of locomotor activation by 10–100 mg tianeptine contrasts with a previous report of depressed locomotion in mice by a higher dose of 160 mg/kg (Uzbay et al., 2007). MOR agonists produce inverted-U shaped dose-effect curves on locomotor activity in mice (Varshneya et al., 2019, 2021), and 160 mg/kg may lie on the descending limb of the tianeptine dose-effect curve.

In addition to locomotor activation, a high dose of tianeptine (100 mg/kg) also produced an MOR agonist-like decrease in gastrointestinal transit. Decreases in gastric motility can lead to constipation, a common side-effect of MOR agonist analgesics and drugs of abuse (Ross et al., 2008; Webster, 2015). Bolton et al. (2008) noted tianeptine-induced inhibition of neuronally mediated contractions in the isolated rat stomach and colon, an effect consistent with MOR agonists. Finally, tianeptine produced robust and naloxone-reversible respiratory depression. Respiratory depression is a severe adverse effect responsible for MOR agonist-induced lethality in humans, and the present findings suggest that tianeptine-induced respiratory depression could contribute to tianeptine overdose deaths. Findings at autopsy in reported fatal tianeptine overdoses indicate the presence of pulmonary congestion and edema, which is a frequent observation in deaths due to opioid-induced respiratory depression (Bakota et al., 2018; Duberstein & Kaufman, 1971; Proença et al., 2007). In some of these fatal overdose cases, tianeptine was the only remarkable toxicological finding, although low concentrations of alcohol or benzodiazepines may have had a minor contributory role in depressing respiration (Bakota et al., 2018; Proença et al., 2007). In the only previous published study to examine tianeptine effects on respiration, lower doses of 2–10 mg/kg tianeptine were reported to increase respiration in rats (Cavalla et al., 2015). The reason for this discrepancy with the present results requires further study and may be a function of the different doses tested; however, the present results clearly show that high tianeptine doses can produce significant and naloxone-reversible respiratory depression typical of MOR agonists.

While nonmedical tianeptine use has been observed in human populations, the full extent of its use is not known. Initial reports of tianeptine’s MOR agonist effects (Gassaway et al., 2014; Samuels et al., 2017) were followed by a spike in case of tianeptine misuse (El Zahran et al., 2018; Rushton et al., 2021); however, despite its nonscheduled status and ready availability, its incidence of abuse appears to be far lower than that of other abused opioids. For example, in 2017, the National Poison Data System reported 4186 heroin exposures compared to only 81 tianeptine exposures during the same time frame (El Zahran et al., 2018; Gummin et al., 2018). A recent study compiled data from social media in which users described their experiences (Smith et al., 2021). Reported symptoms included positive subjective effects as well as adverse effects typical of other MOR agonists, and these effects were generally dose-dependent: on average, lower doses were associated with positive effects and 10-fold higher doses were associated with adverse effects. Interestingly, and contrary to the findings of the present study, users reported strong and rapid tolerance to tianeptine effects as well as dependence and severe withdrawal symptoms. This is further supported by data collected from poison control centers, which indicate that tianeptine-associated withdrawal can be severe enough to require hospitalization (El Zahran et al., 2018; Rushton et al., 2021). However, it is important to note that data from self-reports can be unreliable, and a role for tianeptine in these cases has not been analytically confirmed. Nonetheless, these reports suggest that liberal tianeptine availability many enable consumption of sufficiently high and frequent doses to produce tolerance and dependence.

CONCLUSIONS

Tianeptine presents as a MOR agonist with a resistance to tolerance and withdrawal in our ICSS assay in rats, and it appears to have lower abuse potential than many other commonly abused opioids, consistent with the low incidence of reported abuse in human populations relative to other opioids. Nonetheless, tianeptine is effective to produce MOR agonist-like acute adverse effects that include motor impairment, constipation, and respiratory depression. Assessment of tianeptine abuse potential using other procedures such as drug self-administration may facilitate decisions regarding its regulatory control.