Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Apr 1.
Published in final edited form as: Behav Neurosci. 2019 Feb 4;133(2):247–254. doi: 10.1037/bne0000297

The effects of clavulanic acid and amoxicillin on cue-primed reinstatement of cocaine seeking

Allison R Bechard 1,*, Peter U Hamor 1, Lizhen Wu 1, Marek S Schwendt 1, Lori A Knackstedt 1
PMCID: PMC6508097  NIHMSID: NIHMS1016710  PMID: 30714803

Abstract

Research using the cocaine self-administration and reinstatement animal model of relapse finds that the beta-lactam antibiotic, ceftriaxone, attenuates cocaine-primed reinstatement of cocaine seeking and upregulates two proteins that regulate glutamate release and reuptake (xCT and GLT-1, respectively) in the nucleus accumbens core (NAc). We tested three compounds with beta-lactam rings for their ability to attenuate cue-primed reinstatement and increase GLT-1 and xCT expression in the NAc and prefrontal cortex (PFC). Rats self-administered intravenous cocaine for 1 hr/day for 7 days then 6 hrs/day for 10 days. Cue-primed reinstatement tests began after 8–10 days of extinction training. Rats received oral vehicle, clavulanic acid (CA), amoxicillin (AMX), or CA+AMX (Augmentin; AUG) for 5 days prior to testing. Only AMX-treated rats demonstrated a reduction of cocaine-seeking that trended towards significance, warranting future investigation of a wider range of doses. In the NAc, GLT-1a expression was reduced in vehicle-treated rats relative to cocaine-naïve controls and was not restored by AMX or AUG. CA-treated rats reinstated more than vehicle-treated rats and exhibited GLT-1a and xCT expression intermediate between cocaine-naïve controls and vehicle-treated cocaine rats. In agreement with our previous work, cocaine did not decrease PFC GLT-1a expression. Cocaine reduced xCT expression in the PFC that was unchanged by any of the three compounds. These results indicate that AMX may be another beta-lactam that attenuates cocaine relapse. Furthermore, the upregulation of both GLT-1 and xCT in the NAc may be needed to attenuate cocaine seeking.

Keywords: beta-lactam antibiotics, glutamate transporter 1 (GLT-1), xCT, nucleus accumbens, prefrontal cortex

Relapse to cocaine use after a drug free period remains a significant clinical problem. Cocaine seeking and relapse can be studied in animals using the “extinction-reinstatement” model in which rats are trained to self-administer intravenous cocaine in an operant chamber. After a period of self-administration, the response made to obtain cocaine is extinguished, followed by “reinstatement testing”. During such testing, stimuli that prompt relapse to cocaine use in humans (e.g. cocaine associated cues, stress) also promote the reinstatement of instrumental responding in rodents (Epstein, Preston, Stewart, & Shaham, 2006). Cocaine seeking in this rodent model has been demonstrated to be attenuated by the beta-lactam antibiotic ceftriaxone through the restoration of glutamate homeostasis mechanisms in the nucleus accumbens (NA) core (Bechard, Hamor, Schwendt, & Knackstedt, 2018; Knackstedt, Melendez, & Kalivas, 2010; Sari, Smith, Ali, & Rebec, 2009; Trantham-Davidson, LaLumiere, Reissner, Kalivas, & Knackstedt, 2012). Increased glutamate transporter 1 (GLT-1) and xCT (the catalytic subunit of system xC-) protein expression in the NA core and prefrontal cortex (PFC) accompanies the attenuation of cocaine-seeking following chronic treatment with ceftriaxone (Knackstedt et al., 2010; Sari et al., 2009), and the upregulation of both proteins in the NAc is necessary for ceftriaxone to attenuate reinstatement (LaCrosse et al. 2017).

While ceftriaxone has consistently been demonstrated to attenuate cocaine seeking by multiple labs, these preclinical studies used subchronic IP injections at high doses (e.g. 100–200 mg/kg) and such a route of administration could be a barrier to successful treatment in humans. Furthermore, the oral bioavailability of ceftriaxone is low and no oral formulations of ceftriaxone are currently on the market. Other beta-lactam antibiotics increase GLT-1 expression, potentially due to their beta-lactam ring component (Rothstein et al., 2005). Clavulanic acid (CA) is an FDA-approved beta-lactamase inhibitor that also possesses a beta-lactam ring. Clinically, CA is administered in conjunction with the beta-lactam antibiotic amoxicillin (AMX), in order to inhibit bacteria that inactivate beta-lactam containing antibiotics. CA, AMX, and their combination (sold as Augmentin; AUG) have been studied in rodent models of addiction, and results suggest that their ability to reduce alcohol consumption is associated with increased GLT-1 expression in the NA and PFC (Goodwani, Rao, Bell, & Sari, 2015; Hakami, Alshehri, Althobaiti, & Sari, 2017; Hakami, Hammad, & Sari, 2016; Hakami & Sari, 2017).

Chronic administration of CA (1, 10 mg/kg IP for 7 days) has also been found to reduce motivation to seek intravenous cocaine in mice (Kim et al. 2016). In this study, both doses of CA were increased GLT-1 expression in the NA of cocaine-naïve mice (Kim et al. 2016). CA did not reduce cocaine self-administration when mice were trained using a fixed-ratio 1 (FR1) schedule of reinforcement. However, CA-treated mice displayed a reduction in break point, suggesting CA specifically reduced the motivation to work for cocaine (Kim et al. 2016).

A reduction in alcohol consumption in alcohol-preferring (P) rats has also been associated with increased GLT-1 in the NA (Goodwani et al. 2015; Hakami et al., 2016, 2017). Following 5 days of treatment with either AMX (100 mg/kg IP) or AUG (100 mg/kg IP), alcohol consumption in P rats was reduced and GLT-1 and xCT in the NA was increased (Goodwani et al., 2015; Hakami et al., 2016). This same treatment regimen with AUG (i.e. 100 mg/kg for 5 days), but not AMX, was also successful in upregulating GLT-1 and xCT in the prefrontal cortex (PFC) (Goodwani et al., 2015; Hakami et al., 2016). When administered orally, AUG (100 mg/kg) again significantly attenuated alcohol consumption and upregulated GLT-1 and xCT in the NA and PFC (Hakami et al., 2017).

The ability of these compounds to attenuate the reinstatement of cocaine seeking has not yet been assessed. While we have repeatedly shown that ceftriaxone attenuates the reinstatement of cocaine-seeking after short-access to cocaine self-administration (2 hr/day), we recently demonstrated that following long-access to self-administration (6 hr/day) and extinction training, ceftriaxone (200 mg/kg IP) attenuates cue-primed reinstatement of cocaine seeking and restores expression of both xCT and GLT-1 in the NA (Schwendt et al., 2018). Here we assessed the effects of CA, AMX, and AUG on cue-primed reinstatement of cocaine seeking after long-access to intravenous cocaine self-administration. We also examined GLT-1 and xCT protein expression in the NA core and PFC. We have previously shown both splice variants of GLT-1 (a and b) to be decreased by cocaine and increased by ceftriaxone (LaCrosse et al. 2017). However, GLT-1a is the predominant isoform, accounting for over 90% of total brain GLT-1 expression (Rimmele and Rosenberg, 2016). We hypothesized that chronic administration of CA, AMX and their combination, CA+AMX (AUG), would attenuate cue-primed cocaine seeking relative to vehicle-treated rats, an outcome associated with increased expression of GLT-1 and xCT within the NA core. We further hypothesized that only AUG would increase xCT and GLT-1 in the PFC. We also hypothesized that the effects of CA and AMX on reinstatement would be additive or synergistic, in that rats treated with both (AUG) would display reduced reinstatement relative not only to vehicle-treated rats, but to CA and/or AMX treated rats.

Methods

Animals

Male Sprague-Dawley rats (Charles River LLC, NC, USA, N = 42) spent one week habituating to the colony. A subset (n = 35) underwent jugular catheter surgery; the remaining 7 served as cocaine-naïve controls for western blotting. All animal procedures followed the Guidelines for the Care and Use of Laboratory Animals and were approved by the University of Florida’s Institutional Animal Care and Use Committee.

Surgeries

Methods for surgical procedures are described in (Bechard et al., 2018). Briefly, rats were anaesthetized with a mixture of ketamine (87.5 mg/kg, IP) and xylazine (5 mg/kg, IP). A jugular catheter was inserted that emerged from the skin at the center point between the shoulder blades. A harness was attached to the external part of the catheter that contained a cannulae for intravenous drug delivery. Post-operative care included antibiotic (Cefazolin, 100 mg/mL, IV) and pain medication (Carprofen, subcutaneous) for 2 days after surgery, and a total recovery period of 5–7 days. Heparinized saline (0.2 mL of 100 units/mL, IV) was used to flush catheters daily to maintain their integrity; catheter patency checks were completed at regular intervals with sodium brevitol (10 mg/mL IV; Eli Lilly, Indianapolis, IN, USA).

Self-administration of cocaine and cue-primed reinstatement testing

Rats were trained to self-administer intravenous cocaine using a fixed ratio 1 schedule of reinforcement. Two levers were presented, but only the “active” lever was associated with cocaine delivery (1 mg/kg/infusion in 0.1 mL 0.9% physiological saline) and simultaneous 5 s presentation of a cue complex (tone and stimulus light over the active lever). Presses on the “inactive lever” had no programmed reinforcement but were recorded. Rats self-administered cocaine for 1 hr/day for 7 days, then for 6 hrs/day for 10 days. Animals that experienced a failed catheter were excluded (n = 3). Rats experienced 8–9 days of instrumental extinction training during which both levers are extended but result in no associated consequence (i.e. drug delivery, tone, light). When presses on the previously active lever were less than 20 for two consecutive days of extinction, rats were given a 2-hour cue-primed reinstatement test (Cue Test 1). During this test, presses on the previously active lever delivered drug associated cues (tone and light), but no cocaine. The day after Cue Test 1, rats experienced 1–2 additional extinction training sessions until they achieved the same criteria of 20 or fewer presses on the previously active lever. The following day rats experienced a second cue-primed reinstatement test (Cue Test 2). Rats were sacrificed by rapid decapitation 2 days after reinstatement testing to allow for a return of baseline protein levels. The brains were frozen at −80C for later dissection of the PFC and NA core for Western blotting.

CA, AMX, and AUG administration

Drugs were dispensed onto animal crackers for oral consumption. Animals were observed to ensure that the entire cracker was consumed. In our preliminary studies, all rats consumed the cracker in less than 5 min, and this method was found to be significantly less stressful than oral gavage for administration of drug. CA (5 mg/kg of Potassium clavulanate cellulose, Med Chem Express, Cat no.#HY-19964), AMX (100 mg/kg of Amoxicillin, Sigma-Aldrich, CAS#26787–78-0), and AUG (80 mg/kg CA + 20 mg/kg AMX) were mixed with sterile water. The CA dose was selected based on a previous study that found CA administered at this dose elicited signs of arousal in male rats (Sanna, Melis, Angioni, & Argiolas, 2013). AMX was administered at a dose of 100 mg/kg as this regimen increases GLT-1 and xCT expression in the NA (Goodwani et al., 2015, Hakami et al., 2016), and when combined with CA, in the NA and PFC (Hakami et al., 2016). Note that we did not purchase Augmentin itself, but rather combined CA and AMX at a ratio found in commercial AUG and previously used in studies to assess alcohol consumption (Hakami et al., 2016, 2017). Rats assigned to the Vehicle group received animal crackers treated with water. One rat in the AMX group and one rat in the AUG group did not eat their crackers, and thus were eliminated from the study leaving a final group size of 7 rats each in the CA-, AMX-, and AUG-treated groups, and 9 rats in the Vehicle group. All treatments were administered subchronically for 5 days prior to the first reinstatement test; treatment was administered immediately following the extinction session, as this timing of ceftriaxone administration has proven effective in our previous studies (Knackstedt et al., 2010, Schwendt et al. 2018). We conducted a second reinstatement test to assess whether the acute delivery of CA, AMX or AUG could attenuate reinstatement. During the extinction sessions immediately prior to Cue Test 2, post-session treatment continued. On the day of Cue Test 2, treated crackers were given 15 minutes prior to the start of the test session. Drug treatment resumed the next day and continued until rats were killed.

Home-cage behavioral observations

In male rats, CA induces yawning and erections in a dose-dependent manner when given intraperitoneally (IP, 0.05–5 mg/kg), intracereboventricularly (ICV, 0.01–5 ug/kg), and orally (0.1–5 mg/kg). In order to verify that oral consumption of CA was sufficient, we conducted home cage observations to count arousal behaviors as previously reported (Sanna et al. 2013). Home cage behaviors to assess CA, AMX, and AUG acute effects on arousal were scored during one 45 minute session, beginning 15 minutes after the drug was administered. Scored behaviors included: inactivity, activity (e.g. ambulation, sniffing, rearing, etc.), yawning/stretching, erections, and grooming. Scan sampling collection methods were used and for each rat resulted in 1 observation every 2 minutes, for a total of 22 observations/45 min. Scan sampling occurred for each rat on the day prior to Cue Test 2.

Western-blotting

Proteins were separated using 10% SDS-PAGE and transferred to PVDF membrane. The membranes were probed overnight at 4°C with primary antibodies diluted in 5% milk/Tris-buffered saline with 0.1% Tween-20. Anti-GLT-1a antibody (generous donation from Paul Rosenberg, Harvard University) was used at a dilution of 1:80,000 and anti-xCT (Novus) was used at a 1:5,000 dilution. After incubation with HRP-conjugated secondary antiserum (Jackson Immuno; 1:10,000–1:50,000), immunoreactive bands on the membranes were detected by enhanced chemiluminescence (ECL Plus; GE Healthcare Bio-Sciences). Band density was measured using NIH ImageJ software. Blots were re-probed with anti-calnexin (Millipore, 1:40,000) as a loading control.

Statistical analysis

GraphPad Prism 7.0 (San Diego, CA, USA) was used for data analysis. Cocaine self-administration variables (infusions, active and inactive lever presses) were analyzed as a repeated measures (RM) ANOVA with treatment Group (CA, AMX, AUG, or VEH) as the between-subject factor and Time as the within-subjects (RM) factor. A RM model was also used to compare active and inactive lever presses (separately) during the extinction session preceding Cue Test 1 to the number of presses during cue-primed reinstatement Test 1 and Test 2 with treatment Group as a between-subject factor and Test as a within-subjects (RM) factor. The number of infusions across the self-administration period was summed and the resulting total number of infusions compared between Groups using a one-way ANOVA. Home cage behavioral data were first calculated as a proportion of observation and then active behaviors (groom, ambulation, sniff, rear, drink, etc.) were summed and arousal behaviors (erections, yawning, stretching) were summed. Separate analyses for inactive, active, and arousal behaviors were conducted using a one-way ANOVA with treatment Group as a factor in the model. Upon detection of a significant interaction or main effect in the ANOVAs for behavioral data, post-hoc Tukey tests were utilized to interrogate group differences. We were interested in potential additive or synergistic effects of CA and AMX on reinstatement in the AUG group, and thus post-hoc tests were used to examine group differences in reinstatement. Separate one-way ANOVAs were used to assess differences in GLT-1a and xCT protein expression in the NA core and PFC. Upon detection of a main effect in protein analyses, post-hoc Dunnett tests were utilized. An alpha value of less than 0.05 was considered significant.

Results

We examined cocaine self-administration behavior prior to the initiation of drug treatment. There was an effect of Time on the number of infusions (F (16, 416) = 349.1, p<0.0001), active lever presses (F (16, 416) = 99.2, p<0.0001), and inactive lever presses (F (16, 416) = 2.9, p=0.0001; Figure 1) exhibited during self-administration. There were no main effects of Group in the number of infusions, active and inactive lever presses. However, a Group × Time interaction was found for the number of infusions across the self-administration period (F (48, 416) = 1.4, p = 0.02). Post-hoc analysis indicates a difference in number of infusions between CA and AUG-treated rats on Day 8 (p = 0.01), and Vehicle and AMX (p = 0.001) and Vehicle and AUG-treated (p = 0.01) rats on Day 9, such that AUG- and AMX-treated rats had fewer infusions on these days. However, when we examined the total cocaine intake (mg/kg) for the entire experiment, we found no Group differences. The overall increase in infusions indicates that escalation of cocaine intake occurred, in agreement with previous studies using this method (Knackstedt and Kalivas 2007).

Figure 1.

Figure 1.

Open in a new tab

Self-administration behaviors during the long-access to cocaine paradigm. (A) general timeline of methods. (B) There were differences in the number of infusions across self-administration on Day 8 and 9. (C) No differences were found for active lever presses or (D) inactive lever presses across the 17 days of cocaine self-administration. (Vehicle (VEH) n = 9, Clavulanic Acid (CA) n = 7, Amoxicillin (AMX) n = 7, AUG n = 7). * p<0.05 indicates CA compared to AUG; # p<0.05 indicates VEH compared to AMX and AUG

Analysis of the number of presses made on the previously active lever during the extinction training period revealed a main effect of Time (F (7, 182) = 36.4, p<0.0001), as all groups reduced responding, but no main effect of Group or Group × Time interaction. Responses on the previously inactive lever also decreased with Time (F (1, 182) = 7.9, p<0.0001) with no effect of Group or Group × Time interaction. A RM ANOVA conducted on active lever pressing during the last day of extinction, Cue Test 1 and Cue Test 2 found a Group × Time interaction (F (6, 52) = 2.7, p=0.02). Post-hoc analyses examined differences between lever presses during extinction vs. Cue Tests to identify whether reinstatement of the cocaine-seeking response occurred (e.g. significant differences between extinction and test). This analysis found that reinstatement of cocaine-seeking occurred during both Cue Test 1 and Cue Test 2 for all groups (CA: Test 1 p<0.0001, Test 2, p=0.003; VEH: Test 1 p<0.0001, Test 2 p=0.01; AMX: Test 1 p=0.0007, Test 2 p=0.002; AUG: Test 1 p<0.0001, Test 2 p=0.001; see Figure 2). Only AMX rats did not attenuate cocaine-seeking across Cue Tests 1 and 2 (CA: p<0.0001; VEH: p=0.003; AMX: p=0.90; AUG: p=0.01). CA-treated rats displayed greater lever presses during Cue Test 1 than AMX-treated (p=0.0008) and Vehicle rats (p=0.04; see Figure 2). To explore the effects of AMX on reinstatement relative only to Vehicle rats, we conducted a 2-way RM ANOVA with only AMX and Vehicle rats and found a marginal interaction (F (2, 28) = 3.0, p=0.06).

Figure 2.

Figure 2.

Open in a new tab

Cocaine seeking behaviors during extinction training and cue-primed reinstatement testing. (A) The number of lever presses on the previously active lever decreased over time. (B) All groups reinstated lever pressing during Cue Test 1 and Cue Test 2, and an interaction between Test and Group resulted. Only AMX-treated rats did not reduce lever pressing during Cue Test 2 compared with Cue Test 1. * p<0.05 compared to Extinction; # p<0.05 compared to Test 1

The proportion of time rats spent engaged in active, inactive, or arousal (summation of erections, yawning, and stretching) behaviors did not differ across Groups. Descriptively, we observed arousal behaviors an average of 2.1, 0.5, 0.7, and 2.1 times in the CA-, Vehicle -, AMX-, and AUG-treated rats, respectively. Although not statistically significant, rats treated with CA (i.e. CA and AUG) were observed performing arousal behaviors at a 4-fold increased rate compared to vehicle rats.

GLT-1a expression in the NA core was different across treatment Groups (F (4, 27) = 5.3, p=0.002; Figure 3). Post-hoc analysis showed that cocaine-naive rats had greater expression of GLT-1a as compared to Vehicle (p=0.004), AMX- (p=0.01), and AUG- (p=0.007), but not CA- (p=0.78) treated rats. Expression of xCT in the NA core also differed across treatment Group (F = (4, 28) = 2.7, p=0.04); however, only a non-significant trend for Vehicle rats to have reduced xCT (p=0.07) was revealed by further analysis.

Figure 3.

Figure 3.

Open in a new tab

Protein expression levels in the nucleus accumbens (NA). (A) Expression of GLT-1 in the NA was reduced in Vehicle, AMX-, and AUG-treated rats relative to cocaine-naive controls and restored by Clavulanic Acid (CA). (B) Expression of xCT in the NA differed by Group. * p<0.05 compared to cocaine-naive controls

GLT-1a expression in the PFC was also affected by treatment Group (F (4, 29) = 5.18, p=0.002; Figure 4). Post-hoc tests showed that cocaine-naive control rats had greater levels of GLT-1a compared to AMX- (p=0.001) and AUG- (p=0.002) treated rats, a non-significant trend in CA-treated rats (p=0.05), and no difference in Vehicle rats (p=0.17). Levels of xCT in the PFC differed by Group (F (4, 29 = 2.81, p=0.04), and post-hoc analysis showed reduced expression in Vehicle (p=0.02) and a trend for reduced expression in CA-treated rats (p=0.06) compared to Controls.

Figure 4.

Figure 4.

Open in a new tab

Protein expression levels in the prefrontal cortex (PFC). (A) In AMX and AUG rats, expression of GLT-1a in the PFC was decreased compared with cocaine-naive controls. (B) Expression of xCT in the PFC differed by Group. * p<0.05 compared to cocaine-naive controls

Discussion

Here we present the first study to examine the effects of orally administered CA, AMX, and AUG on cue-primed reinstatement of cocaine-seeking, finding that none of these compounds attenuated reinstatement when rats were tested undrugged in Cue Test 1, nor when the beta-lactam was administered acutely prior to a relapse test in Cue Test 2. In Cue Test 1, CA treated rats displayed greater reinstatement lever pressing than did Vehicle rats, whereas AMX-treated rats displayed a trend for less lever presses than Vehicle rats. Furthermore, rats treated with AUG displayed lever pressing on Cue Test 1 that was intermediate to AMX and CA rats, suggesting the potential washing out of CA’s effects on brain and behavior when the two drugs are combined. In Cue Test 2, all groups similarly reinstated cocaine seeking.

An extinction effect from Cue Test 1 to Cue Test 2 was seen in all groups except AMX, which had the lowest number of active lever presses in Cue Test 1 but did not decrease presses from Test 1 to Test 2. Potentially, there was a floor effect in the AMX-treated rats due to the lower presses during Cue Test 1. A similar reduction in lever pressing from one cue-primed reinstatement test to the next has been reported previously when utilizing two cue-primed reinstatement tests that were separated by only a short period of time (e.g. less than one week) (Buffalari et al. 2013). The presence of such an extinction effect confounds the interpretation of Cue Test 2 results and future work should explore more thoroughly the ability of acute beta-lactam treatment to attenuate reinstatement. Another factor to consider in the interpretation of the present results is the presence of group differences in behavior prior to the reinstatement tests. Ideally, all groups would display self-administration and extinction behavior that does not differ. Here we observed group differences in the number of cocaine infusions earned in the AMX and AUG groups only on Days 8 and 9 of self-administration. As the total number of cocaine infusions across the entire experiment did not differ between the four treatment groups, it is likely that this drop in intake only on two days of self-administration did not influence later seeking. However, it remains possible that this reduction influenced seeking in the reinstatement tests.

Contrary to previous literature suggesting reduced alcohol consumption following administration of CA and AUG (Goodwani et al., 2015; Hakami et al., 2016, 2017), we found no benefits of CA and AUG on cue-primed reinstatement of cocaine-seeking. This finding is in agreement with our previous work with the beta-lactam antibiotic cefazolin, which we showed to attenuate the cue primed reinstatement of alcohol seeking but not cocaine (Weiland, Garcia, & Knackstedt, 2015). Thus, alcohol seeking and consumption is reduced by a number of beta-lactam compounds, but cue-primed reinstatement of cocaine seeking is reduced only by ceftriaxone.

During Cue Test 1, the CA-treated rats displayed greater presses on the active lever relative to AMX and VEH treated rats. Thus, CA facilitated cocaine-seeking in the present work. This is in contrast to the ability of CA to reduce the motivation to seek cocaine on a progressive ratio schedule (Kim et al., 2016). Both dopamine receptor 2 (D2) and oxytocin receptor antagonists given 15 min prior to CA reduce CA-induced yawning and erections by ~55% (Sanna et al. 2013). Even greater reductions of CA responses (~80%) were found following administration of morphine (an opioid agonist), and mianserin (a serotonin) receptor antagonist, whereas naloxone (an opioid antagonist) had no effect (Sanna et al. 2013). Thus, while we found that chronic administration of CA increased GLT-1 expression, it may also act on central dopamine and serotonin neurotransmission to increase motivation to respond for cocaine-associated cues during a relapse test, but decrease seeking on a progressive ratio (Kim et al., 2016). Interestingly, we observed this behavioral effect of chronic CA in the absence of acute behavioral effects on arousal and yawning. This may have occurred because the rats in the present study had a history of cocaine self-administration, and thus CA’s effect’s on dopamine, oxytocin and/or serotonin were altered by that history. It is also possible that we missed the peak occurrence of such behaviors because we observed them on Day 7 of CA treatment or because the cracker we administered with the drug altered the time course of responses reported in previous studies that orally administered the drug via liquid injections (e.g. Sanna et al., 2013).

Here, cocaine self-administration reduced levels of GLT-1a in the NA core of vehicle-treated rats relative to cocaine-naïve controls, in agreement with a number of studies, including those from our laboratory (Bechard et al., 2018; Knackstedt et al., 2010; LaCrosse et al. 2017; Reissner et al. 2015; Kim et al. 2016; Sari et al. 2009). GLT-1a expression continued to be reduced after AMX and AUG treatment, likely accounting for the inability of these compounds to attenuate reinstatement. An identical treatment regimen (dose, route of administration, length) of these compounds increased NA GLT-1 following alcohol consumption (Goodwani et al., 2015; Hakami et al., 2016, 2017). Of note, the studies investigating AMX and AUG in alcohol-seeking used alcohol preferring “P-rats”, which are bred to consume alcohol, and have altered gene expression from outbred rats, possibly explaining discrepancies between previous reports and the current study, which employed the outbred Sprague-Dawley rat strain. Another important distinction is that the studies in P rats assessed GLT-1 and xCT expression in tissue samples that included both the core and shell subcompartments of the NA, where here we only assessed expression in the NA core, in line with our previous research. Only CA rats had restored GLT-1a expression levels in the NA core, which agrees with previous results in drug-naive mice that examined expression in the entire NA (Kim et al., 2016). However, despite restored GLT-1 in CA rats, cue-primed reinstatement was not attenuated, potentially due to its effects on other neurotransmitters as discussed above. This finding supports our previous work demonstrating that GLT-1 upregulation in the NA core is necessary, but not sufficient to attenuate cued reinstatement of cocaine-seeking (LaCrosse et al., 2017; Logan, LaCrosse, & Knackstedt, 2018). xCT expression in the NA core of CA-, AMX- and AUG-treated rats was unchanged relative to both cocaine-VEH and cocaine-naive rats, replicating earlier findings for AMX and AUG in the NA of alcohol-drinking rats (Goodwani et al., 2015; Hakami et al., 2016, 2017). Thus, the absence of increased GLT-1 and xCT expression in the NA core likely accounted for the inability of these compounds to attenuate reinstatement. An alternative interpretation of the xCT expression results for the NA core is that while none of the beta-lactam compounds tested increased xCT expression, expression was not decreased in these treatment groups relative to cocaine-naïve controls. While xCT upregulation in the NA core has been demonstrated to be necessary for ceftriaxone to attenuate cued cocaine-seeking, it has not yet been shown to be sufficient. The present work indicates that such upregulation may not be sufficient to do so, and the attenuation of cocaine relapse may require other adaptations in either the NA core itself, or other regions of the reward circuitry.

In the PFC, GLT-1 was reduced in AMX- and AUG-treated rats, but not in VEH- or CA-treated rats, compared to Controls. This is in contrast to findings in P-rats where AUG increases GLT-1 in the PFC after alcohol consumption (Goodwani et al., 2015; Hakami et al., 2016, 2017). In a similar manner as the NAc, xCT expression in the PFC was reduced in Vehicle-treated rats relative to Controls, but not in any of the other conditions. In P-rats, AUG but not AMX, has been demonstrated to increase xCT expression (Goodwani et al., 2015; Hakami et al., 2016, 2017). CA has not been assessed for its ability to alter xCT or GLT-1 expression in the PFC. Thus, our results in regard to AMX effects on GLT-1 are in agreement with the effects of AMX in P-rats, including one study utilizing an oral route of AMX administration as was done here (Hakami et al., 2017). While Ceftriaxone consistently attenuates cue-primed reinstatement, only one study found accompanying increases in GLT-1 expression in the PFC (Sari et al., 2009), while another did not (Knackstedt et al., 2010). The necessity of GLT-1 or xCT upregulation in the PFC to the ability of Ceftriaxone to attenuate cocaine-seeking has not yet been investigated.

Conclusion

We tested the beta lactam antibiotic amoxicillin and the beta-lactamase inhibitor, clavulanic acid and their combination, Augmentin, for their ability to attenuate cocaine-seeking after extinction training, when likelihood of relapse is high. In spite of clavulanic acid’s ability to upregulate GLT-1 in the NAc and PFC, and xCT in the NA core, cue-primed reinstatement of cocaine-seeking was not attenuated, and in fact was increased relative to amoxicillin-treated rats. Amoxicillin and Augmentin did not upregulate GLT-1 in the NA core or PFC. However, xCT expression in the NA core and PFC were similar to cocaine-naive controls. We saw a significant reduction in cue-primed reinstatement of cocaine-seeking only in amoxicillin-treated rats compared to clavulanic acid-treated rats. As xCT was not reduced in the amoxicillin-treated group relative to cocaine-naïve controls, there may be a key role for xCT as a mediator of the attenuation of cocaine-seeking. We conclude that clavulanic acid is not an effective prevention for cocaine relapse, and may increase relapse probability. Amoxicillin-treated rats showed a marginal reduction in lever pressing during cued cocaine-seeking, and future research should examine additional amoxicillin doses.

Acknowledgments

Role of Funding Source: Funding for the study was received from the National Institutes of Health grants #DA033436 and DA037270 awarded to L.A.K. This work was also supported by Institute on Molecular Neuroscience Subcontract 8738sc and 9250sc awarded to LK and MS. The content of the information does not necessarily reflect the position or the policy of the Government, and no official endorsement should be inferred. Award Number: W81XWH-12–2-0048. The U.S. Army Medical Research Acquisition Activity, 820 Chandler Street, Fort Detrick, MD 21702–5014 is the awarding and administering acquisition office.

Bibliography

  1. Bechard AR, Hamor PU, Schwendt M, & Knackstedt LA (2018). The effects of ceftriaxone on cue-primed reinstatement of cocaine-seeking in male and female rats: estrous cycle effects on behavior and protein expression in the nucleus accumbens. Psychopharmacology, 235(3), 837–848. doi: 10.1007/s00213-017-4802-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Buffalari DM, Feltenstein MW, & See RE (2013). The effects of varied extinction procedures on contingent cue-induced reinstatement in Sprague-Dawley rats. Psychopharmacology, 230(2), 319–327. doi: 10.1007/s00213-013-3156-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Epstein DH, Preston KL, Stewart J, & Shaham Y (2006). Toward a model of drug relapse: an assessment of the validity of the reinstatement procedure. Psychopharmacology, 189(1), 1–16. doi: 10.1007/s00213-006-0529-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Goodwani S, Rao PSS, Bell RL, & Sari Y (2015). Amoxicillin and amoxicillin/clavulanate reduce ethanol intake and increase GLT-1 expression as well as AKT phosphorylation in mesocorticolimbic regions. Brain Research, 1622, 397–408. doi: 10.1016/j.brainres.2015.07.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Hakami AY, Alshehri FS, Althobaiti YS, & Sari Y (2017). Effects of orally administered Augmentin on glutamate transporter 1, cystine-glutamate exchanger expression and ethanol intake in alcohol-preferring rats. Behavioural Brain Research, 320, 316–322. doi: 10.1016/j.bbr.2016.12.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Hakami AY, Hammad AM, & Sari Y (2016). Effects of Amoxicillin and Augmentin on Cystine-Glutamate Exchanger and Glutamate Transporter 1 Isoforms as well as Ethanol Intake in Alcohol-Preferring Rats. Frontiers in Neuroscience, 10, 171. doi: 10.3389/fnins.2016.00171 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Hakami AY, & Sari Y (2017). β-Lactamase inhibitor, clavulanic acid, attenuates ethanol intake and increases glial glutamate transporters expression in alcohol preferring rats. Neuroscience Letters, 657, 140–145. doi: 10.1016/j.neulet.2017.08.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Kim J, John J Langford D, Walker E Ward S, Rawls SM (2016). Clavulanic acid enhances glutamate transporter subtype I (GLT-1) expression and decreases reinforcing efficacy of cocaine in mice. Amino Acids, 48, 689–696. doi: 10.1007/s00726-015-2117-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Kim R, Sepulveda-Orengo MT, Healey KL, Williams EA, & Reissner KJ (2018). Regulation of glutamate transporter 1 (GLT-1) gene expression by cocaine self-administration and withdrawal. Neuropharmacology, 128, 1–10. doi: 10.1016/j.neuropharm.2017.09.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Knackstedt LA, & Kalivas PW (2007). Extended access to cocaine self-administration enhances drug-primed reinstatement but not behavioral sensitization. The Journal of Pharmacology and Experimental Therapeutics, 322(3), 1103–1109. doi: 10.1124/jpet.107.122861 [DOI] [PubMed] [Google Scholar]
  11. Knackstedt LA, Melendez RI, & Kalivas PW (2010). Ceftriaxone restores glutamate homeostasis and prevents relapse to cocaine seeking. Biological Psychiatry, 67(1), 81–84. doi: 10.1016/j.biopsych.2009.07.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. LaCrosse AL, O’Donovan SM, Sepulveda-Orengo MT, McCullumsmith RE, Reissner KJ, Schwendt M, & Knackstedt LA (2017). Contrasting the Role of xCT and GLT-1 Upregulation in the Ability of Ceftriaxone to Attenuate the Cue-Induced Reinstatement of Cocaine Seeking and Normalize AMPA Receptor Subunit Expression. The Journal of Neuroscience, 37(24), 5809–5821. doi: 10.1523/JNEUROSCI.3717-16.2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Logan CN, LaCrosse AL, & Knackstedt LA (2018). Nucleus accumbens GLT-1a overexpression reduces glutamate efflux during reinstatement of cocaine-seeking but is not sufficient to attenuate reinstatement. Neuropharmacology, 135, 297–307. doi: 10.1016/j.neuropharm.2018.03.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Reissner KJ, Gipson CD, Tran PK, Knackstedt LA, Scofield MD, & Kalivas PW (2015). Glutamate transporter GLT-1 mediates N-acetylcysteine inhibition of cocaine reinstatement. Addiction Biology, 20(2), 316–323. doi: 10.1111/adb.12127 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Rimmele TS, Rosenberg PA (2016). GLT-1: The elusive presynaptic glutamate transporter. Neurochemistry International, 98, 19–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Rothstein JD, Patel S, Regan MR, Haenggeli C, Huang YH, Bergles DE, … Fisher PB (2005). Beta-lactam antibiotics offer neuroprotection by increasing glutamate transporter expression. Nature, 433(7021), 73–77. doi: 10.1038/nature03180 [DOI] [PubMed] [Google Scholar]
  17. Sanna F, Melis MR, Angioni L, & Argiolas A (2013). Clavulanic acid induces penile erection and yawning in male rats: comparison with apomorphine. Pharmacology, Biochemistry, and Behavior, 103(4), 750–755. doi: 10.1016/j.pbb.2012.12.001 [DOI] [PubMed] [Google Scholar]
  18. Sari Y, Smith KD, Ali PK, & Rebec GV (2009). Upregulation of GLT1 attenuates cue-induced reinstatement of cocaine-seeking behavior in rats. The Journal of Neuroscience, 29(29), 9239–9243. doi: 10.1523/JNEUROSCI.1746-09.2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Schwendt M, Shallcross J, Hadad NA, Namba MD, Hiller H, Wu L, … Knackstedt LA (2018). A novel rat model of comorbid PTSD and addiction reveals intersections between stress susceptibility and enhanced cocaine seeking with a role for mGlu5 receptors. Translational Psychiatry, 8(1), 209. doi: 10.1038/s41398-018-0265-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Trantham-Davidson H, LaLumiere RT, Reissner KJ, Kalivas PW, & Knackstedt LA (2012). Ceftriaxone normalizes nucleus accumbens synaptic transmission, glutamate transport, and export following cocaine self-administration and extinction training. The Journal of Neuroscience, 32(36), 12406–12410. doi: 10.1523/JNEUROSCI.1976-12.2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Weiland A, Garcia S, & Knackstedt LA (2015). Ceftriaxone and cefazolin attenuate the cue-primed reinstatement of alcohol-seeking. Frontiers in Pharmacology, 6, 44. doi: 10.3389/fphar.2015.00044 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES