The influence of the host microbiome on 3,4-methylenedioxymethamphetamine (MDMA)-induced hyperthermia and vice versa

Hyperthermia induced by 3,4-methylenedioxymethamphetamine (MDMA) can be life-threatening. Here, we investigate the role of the gut microbiome and TGR5 bile acid receptors in MDMA-mediated hyperthermia. Fourteen days prior to treatment with MDMA, male Sprague-Dawley rats were provided water or water treated with antibiotics. Animals that had received antibiotics displayed a reduction in gut bacteria and an attenuated hyperthermic response to MDMA. MDMA treated animals showed increased uncoupling protein 1 (UCP1) and TGR5 expression levels in brown adipose tissue and skeletal muscle while increased expression of UCP3 was observed only in skeletal muscle. Antibiotics prior to MDMA administration significantly blunted these increases in gene expression. Furthermore, inhibition of the TGR5 receptor with triamterene or of deiodinase II downstream of the TGR5 receptor with iopanoic acid also resulted in the attenuation of MDMA-induced hyperthermia. MDMA-treatment enriched the relative proportion of a Proteus mirabilis strain in the ceca of animals not pre-treated with antibiotics. These findings suggest a contributing role for the gut microbiota in MDMA-mediated hyperthermia and that MDMA treatment can trigger a rapid remodeling of the composition of the gut microbiome.


Quantification of Intestinal Bacteria by qPCR.
To determine if changes in intestinal microbiota influence the thermogenic response to MDMA, animals were provided a cocktail of antibiotics (ABX): bacitracin, neomycin, and vancomycin, via their drinking water, for 14 days before MDMA treatment. With the exception of the first day of exposure to the antibiotics ( Fig. 1A-two-tailed t-test: p = 0.0019; t = 3.535), daily fluid intake and body weight ( Fig. 1B-two-tailed t-test: p = 0.2144; t = 1.278) did not differ between groups over the course of the experiment. To ensure the ability of mixed ABX in reducing gut bacteria, total DNA was purified from 200 mg of the cecum of each animal and they were pooled together, followed by qPCR analysis using 200 ng of pooled DNA and universal primers specific for the bacterial 16S rRNA gene. Based on the quantification cycles (Cq) values, a significant reduction in apparent bacterial number was evident in the ABX groups compared to the H 2 O control groups ( Fig. 2-two-tailed t-test: p = 0.0016; t = 4.296). Prolonged treatment with antibiotics resulted in markedly enlarged ceca (data not shown), similar to the previous observations of others 26 . Analysis of Cultivatable Cecal Bacteria. qPCR analysis (Fig. 2) established that ABX treatment reduced bacterial numbers in the cecum relative to untreated controls, evident by the increased Cq value for the ABX-treated animals (Fig. 2). Our working assumption was that ABX treatment would alter the composition of the cecal microbiota by differentially selecting against antibiotic sensitive bacterial strains. To determine if this occurred, we evaluated kanamycin resistance in cecal samples harvested at necropsy from each of the four groups. Resistance to kanamycin would be indicative of the selection of strains capable of producing aminoglycoside modifying enzymes by the ABX regimen. Comparison of the relative proportions of kanamycin resistant culturable bacterial verified that such alterations did occur. For the six saline-challenged animals that received no antibiotics (H 2 O control), only 15-48% (mean = 27%) of the culturable cecal bacteria grew in the presence of kanamycin, whereas for the six saline-challenged ABX treated animals (ABX control) 53-100% (mean = 78%) of the culturable cecal bacteria grew in the presence of kanamycin. A similar proportion of culturable cecal bacteria (46-83%; mean = 64%) able to grow in the presence of kanamycin were found in the six MDMA-challenged ABX treated animals (  www.nature.com/scientificreports www.nature.com/scientificreports/ For the six non-ABX-treated MDMA-challenged animals (H 2 O MDMA), the proportion of kanamycin-resistant culturable bacteria could not be determined, as the colonies were obscured by the presence of swarming bacteria on plates lacking kanamycin. Bacterial swarming on a nutrient-rich medium is a phenotype associated with the genus Proteus, a common inhabitant of mammalian digestive systems 27 . The swarmers were subsequently isolated as single colonies on MacConkey agar, which contains bile salts that inhibit swarming behavior. Colony PCR of eight such isolates using primers specific for 16S rRNA produced amplimers corresponding with a sequence to a 314-base pair region of the 16S rRNA genes (GeneBank accession number MK033607) that is shared by both Proteus mirabilis and certain isolates of Salmonella enterica, including a serovar Typhimurium strain, all with 100% sequence identity and minimum E-values of 2.30 E −158 (Fig. 4A). Isolates were compared with a well-documented strain of S. enterica serovar Typhimurium (ATCC 19585) by plating on another bile salt-containing medium, Hektoen enteric (HE) agar, that distinguishes salmonella from other enteric bacteria based on the production of hydrogen sulfide, as visualized by the formation of black ferric sulfate precipitates in the inner region of the colony. Such precipitates were not evident in swarmer isolates grown on HE agar, but were apparent for the S. enterica strain (Fig. 4B). Swarmer isolates were re-plated onto a nutrient rich medium (LB) where swarmer behavior was again evident, in contrast to S. enterica, which did not swarm (Fig. 4C). Collectively, these data suggest that the swarming bacteria are Proteus mirabilis.   www.nature.com/scientificreports www.nature.com/scientificreports/ that the effects of the oral antibiotic treatment on MDMA-induced hyperthermia were due to pharmacodynamic interactions between MDMA and the antibiotics, a second experiment was performed, with animals injected with a single intraperitoneal dose of antibiotics (1.67 mg/kg vancomycin, 20 mg/kg neomycin, and 293 U/kg bacitracin in PBS) thirty minutes prior to treatment with MDMA. These animals did not display an attenuated thermogenic response to MDMA. Instead, these animals displayed a similar maximum change in temperature as the control group (data not shown-two-tailed t-test: p = 0.4525; t = 0.781), supporting the interpretation that components of the microbiota, not the antibiotics, were responsible for the attenuated thermogenic response to MDMA in the ABX MDMA group.

Discussion
Several studies have found that microbiome-generated heat contributes to the maintenance of body temperature in animals (for a review see 28 ). Oral treatment of rabbits 29 and rodents 30,31 with non-absorbable antibiotics lowers core body temperature. Recent findings using Brandt's voles suggest that changes in core body temperature changes the host gut microbial community, which in turn regulates the host's thermal homeostasis 32 . The present findings suggest that the microbiome-gut-brain axis may play a contributory role in the hyperthermia mediated by MDMA. ABX treatment for 14 days prior to MDMA treatment not only reduced the number and altered the composition of the cecal bacterial population, but also attenuated MDMA-induced hyperthermia. Recently, Althobailti et al. 33 , demonstrated that the β-lactam antibiotic, ceftriaxone, administered for three days following methamphetamine treatment blocked the methamphetamine-induced hyperthermia. Those authors did not examine the effects of ceftriaxone treatment on the gut microbiome, but linked the effects to ceftriaxone regulation of the central expression of the glutamate transporter. Intraperitoneal injection of minocycline for three days has been shown to attenuate the hyperthermic effects of MDMA 34 , but had no effect on body temperature when given for two days 35 . Previous studies by Rawls et al. 36 found that ceftriaxone given for seven days prior to hyperthermic doses of morphine also blocked hyperthermia. They suggested a potential role for the glutamate transporter, but did not examine the effects of ceftriaxone on the gut microbiome. This present study is the first to examine the potential role of the bacterial component of the gut microbiome in the hyperthermia associated with drugs of abuse. Although the results of the present study demonstrate a potential relationship between MDMA-induced hyperthermia and the gut microbiome, they do not suggest that the microbiome is a major contributor to this hyperthermia in that the effects of the ABX treatment were slight yet significant.
We have recently ascertained that UCP1 and UCP3 play complementary roles in the onset (UCP1) and maintenance (UCP3) of sympathomimetic-induced hyperthermia 37 . In the present study, we observed an increase in UCP1 expression in both BAT and skeletal muscle following MDMA challenge while UCP3 expression increased www.nature.com/scientificreports www.nature.com/scientificreports/ only in skeletal muscle. ABX treatment blocked these changes in UCP expression levels, consistent with our previous studies regarding the role of UCP in MDMA-mediated hyperthermia 37 . The TGR5 receptor has been suggested to play a role in cold-induced thermogenesis through its regulation of the thyroid-catalyst protein deiodinase II (D2) and the ultimate regulation of UCP expression 14,38 . Similarly, in both BAT and skeletal muscle of our rat models, MDMA increased TGR5 expression levels. Conversely, prior ABX treatment rendered TGR5 gene expression refractory to this MDMA-induced change.
Recent studies have suggested that TM inhibits the TGR5 receptor 24 . Those authors reported that TM was able to dose-dependently inhibit the increase in glucose uptake mediated by TGR5 agonists in Chinese hamster ovary cells (CHO-K1 cells). Moreover, glucagon-like peptide-1 secretion and increased cAMP levels, which are induced by TGR5 activation, were both shown to be dose dependently reduced by TM 24 . Additional animal studies by the same researchers using streptozotocin-induced diabetic rats supported these findings, indicating that TM is an effective TGR5 antagonist 24 . In the present study, we demonstrate and confirm that TM functions as an antagonist of the TGR5 receptor to attenuate the hyperthermia mediated by MDMA. Additionally, the D2-inhibitor IOP significantly attenuated MDMA-mediated hyperthermic response. IOP was first shown to inhibit D2 activity and the enzymatic conversion of T4 to T3 by Larsen et al. 25 . T3 has been demonstrated to increase UCP1 and UCP3 mRNA expression in BAT and skeletal muscle, respectively 39,40 . TGR5 receptor activation facilitates cAMP production that ultimately leads to an increase of D2 production 14 . The attenuated MDMA-mediated hyperthermic response seen after D2 inhibition is consistent with previous studies from our laboratory, demonstrating a role for thyroid hormone in the hyperthermia mediated by MDMA 6 .
This study suggests that the resident gut flora influences the physiological response of their host to the challenge of MDMA. The nature of this influence, and the specific members of the resident microbial community responsible, remains to be discerned, but clearly merits further consideration. Identifying mechanisms by which the microbiota modulate hyperthermic responses provides alternative strategies to dissect this and other complex regulatory pathologic processes. The fact that MDMA changed the composition of the gut microbiota, coupled with the observation that altering the gut microbiota with antibiotics modified the thermogenic response to MDMA in the rodent model, suggests a potential relationship between human gut microbiota and body temperature regulation of relevance to our understanding of MDMA-induced pathological thermogenesis in humans.
The hyperthermia mediated by sympathomimetic agents such as MDMA involves both peripheral and central mechanisms. The communication between the gut-brain-axis is complex and involves many of the same neurotransmitters that are targets for the pharmacological and toxicological actions of MDMA. For example, serotonin and norepinephrine play key roles in regulating the communication between gut-brain-axis. Subsequently, an association has been made between intestinal bacteria and the pathophysiology of CNS disorders such as autism, depression and substance-use disorders (for a review, see 41 ). In the present study, we did not examine the potential link between the gut-brain-axis and studies in this area are therefore warranted.
This present study was optimized to provide for a consistent pathological response to afford greater precision to evaluate the potential of individual components (such as the composition of the gut microbiota) to enhance www.nature.com/scientificreports www.nature.com/scientificreports/ MDMA-induced thermogenesis. To this end, we used a 20 mg/kg dose of MDMA. This is the standard dose that our lab has used to induce MDA induced hyperthermia 6 . Additionally, toxicity studies with MDMA typically use a dosage range of 10 to 80 mg/kg (for a review, see 42 ). Ambient temperature has been shown to influence the thermogenic response to amphetamines 4,43,44 . O'Shea et al. 45 demonstrated that rats treated with MDMA in environmental temperatures of 30 °C or higher became hyperthermic. However, if the ambient temperature is reduced to 20-22 °C, MDMA mediates a hypothermic response 46 . In present study, we maintained an ambient temperature 24-26 °C. Finally, we had previously demonstrated that the amount of fat in the diet further enhanced the thermogenic response to MDMA 47 . We therefore fed the animals a 10% fat diet.
Intriguingly, this study found the presence of a Proteus population of sufficient relative number to obscure colony counting, but only in non-ABX animals treated with MDMA. This indicated that MDMA itself either directly or indirectly stimulates rapid alterations in the composition of the resident bacterial population of the cecum. This suggests the possibility that the repeated use of MDMA and similar compounds may result in long-term alterations in the composition of resident intestinal bacterial populations. This possibility and its potential consequences also merit further consideration.  www.nature.com/scientificreports www.nature.com/scientificreports/ accordance with the eighth edition of the Guide for the Care and Use of Laboratory Animals as adopted and circulated by the National Institutes of Health, and protocols were approved by the Bowling Green State University Animal Care and Use Committee. All animals were allowed to acclimate to the facility for one week prior to the start of any treatments. ABX Study Design. Animals were randomly assigned to four treatment groups (n = 6): H 2 O saline, ABX saline, H 2 O MDMA, and ABX MDMA. Antibiotics were administered via the drinking water for fourteen days in accordance with the methods described in the study by Kiraly et al. 26 . Antibiotic doses were: Neomycin 2 mg/mL, Vancomycin 0.2 mg/mL, and Bacitracin 0.5 mg/mL. Fluid intake was measured daily, and body weight was measured every five days. On the fourteenth day, animals were injected subcutaneously with either saline or MDMA (20 mg/kg). Temperatures were taken rectally just prior to treatment (baseline) and every 30 min for an hour post-treatment using a Physiotemp Thermalert TH-8 termocouple (Physitemp Instruments, Clifton, NJ) attached Each value is the mean ± SEM (n = 6). * indicates significantly different between treatment groups (p < 0.05). (B) Temperature Area Under the Curve (TAUC) following MDMA treatment 30 minutes after iopanoic acid pretreatment. Each value is the mean ± SEM (n = 6). ** indicates significantly different between treatment groups (p < 0.01). (C) Max Change in Temperature (ΔTmax) following treatment of MDMA 30 minutes after iopanoic acid pretreatment. Each value is the mean ± SEM (n = 6). *** indicates significantly different between treatment groups (p < 0.001).

Methods
www.nature.com/scientificreports www.nature.com/scientificreports/ to a RET-2 rectal probe. The animals were then euthanized by carbon dioxide asphyxiation and BAT, SKM, and cecum were dissected out for qPCR analysis and for bacterial count (cecum only), UCP 1&3 and TGR5 quantification (BAT and SKM only). To rule out the systemic effects of the antibiotics, six animals were injected intraperitoneally with 1.67 mg/kg vancomycin, 20 mg/kg neomycin, and 293 U/kg bacitracin in PBS 26 thirty minutes before the administration of MDMA (20 mg/kg). Temperatures were recorded just prior to antibiotic administration (baseline) and every 30 minutes for an hour post-treatment of MDMA.
RNA isolation and qRT-PCR. SKM and BAT were collected and preserved at −80 °C. Total RNA was isolated after homogenizing the tissues using PureZOL ™ RNA Isolation reagent (Biorad). The concentration and quality of the RNA was determined using a NanoDrop Spectrophotometer (Thermo) and by 1% agarose gel electrophoresis, respectively. cDNA was synthesized from 200 ng of total RNA using the iScript ™ Select cDNA   Table 1. The qPCR cycle parameters were 3 min at 95 °C; 40 cycles of 95 °C for 10 s, 55 °C for 20 s, and 68 °C, 10 s. Each qPCR reaction was run in triplicate along with reactions without DNA (no template) as negative controls. Quantification cycles (Cq) threshold was placed above baseline of the background noise of the fluorescence dye present in negative controls within the linear portion of the curve and were further used for statistical analysis.
Agar Plate Assay and Antibiotic Screening. Cecal contents from all animals were collected in sterile tubes and adjusted to 200 mg/ml with 4 °C phosphate buffered saline (PBS; 0.1 M, pH 7.4). The resultant cecal slurries were homogenized and centrifuged at 400 × g for 2 minutes 49 . Serial 10-fold dilutions of the cecal slurry were spread on LB agar (Miller formulation; Difco-BRL) with or without 50 µg/ml kanamycin (LB kan50 ), incubated at 37 °C overnight, then manually scored for colony forming units (CFU), with the percent of kanamycin resistant (kan r ) determined by dividing the number of CFU on the LB kan50 by the number of CFU on LB, and multiplying by 100.
Isolation and Identification of Swarming Bacteria. Swarmers were isolated from the initial LB agar plates by subculture onto MacConkey agar (Difco-BRL), where the presence of bile salts inhibited swarming behavior. Plates were incubated for 18 hrs at 37 °C. Resultant colonies were picked and streaked for isolation onto fresh MacConkey agar and incubated as above. Isolated swarmer colonies were subjected to colony PCR, with the 16S rRNA gene amplified using the primer set Eub-16S F (Table 1) 50 corresponding to bp of 331 to 799 of 16S rRNA of Escherichia coli. Sequencing of the PCR amplicons was performed at the University of Chicago DNA sequencing and genotyping facility, Chicago, Illinois, USA. Sequences were evaluated for similarity to known sequences in the National Center for Biotechnology Information (NCBI) nucleotide sequence database using BLAST 51 .
Because the resultant rRNA sequences were identical to those of both P. mirabilis and S. enterica, the phenotype of the swarming isolate was compared to that of S. enterica, by plating on the differential medium Hektoen enteric agar (Difco-BRL) and by the ability to swarm on LB agar.  24 , TM (50 mg/kg) was administered via an ip injection 30 minutes before treatment with MDMA (20 mg/kg, sc). Animals in the IOP/SAL and IOP/MDMA groups were pretreated with IOP (50 mg/kg/day, ip) daily for seven days, with the IOP dose size in accordance with Larsen et al. 25 . On day 7, animals were treated with MDMA (20 mg/kg, sc) 30 minutes after treatment with IOP or its vehicle, and core rectal temperature was measured at 30-minute intervals for 90 minutes post-treatment.