Potential roles of the gut microbiota in the manifestations of drug use disorders

Drug use disorders (DUDs) not only cause serious harm to users but also cause huge economic, security, and public health burdens to families and society. Recently, several studies have shown that gut microbiota (GM) can affect the central nervous system and brain functions. In this review, we focus on the potential role of the GM in the different stages of DUDs. First, the GM may induce individuals to seek novel substances. Second, the gut microbiota is involved in the decomposition and absorption of drugs. Symptoms of individuals who suffer from DUDs are also related to intestinal microorganisms. Third, the effects of the GM and its metabolites on drug relapse are mainly reflected in the reward effect and drug memory. In conclusion, recent studies have preliminarily explored the relationship between GM and DUDs. This review deepens our understanding of the mechanisms of DUDs and provides important information for the future development of clinical treatment for DUDs.


Introduction
Drug use disorders (DUDs) are notable as a global public health problem. According to the latest report from the United Nations Office on Drugs and Crime, approximate 284 million people aged 15-64 years old worldwide have used drugs in 2020 (1). Notably, people with DUDs who are co-infected with the severe acute respiratory syndrome coronavirus 2, the etiological agent of the current coronavirus disease 2019 pandemic, are at risk of contracting multiple diseases. In addition, because of the hindered economic development in various countries due to the pandemic, the possibility of drug use by high-risk groups and the risk of relapse by addicts during the abstinence period will significantly increase (2,3). DUDs significantly harm the physical and mental health of drug abusers, and it is associated with public health and safety problems (e.g., AIDS) (4). Importantly, DUDs are considered the chronic and recurrent brain disease,

Gut microbiota and novelty seeking
The initiation of drug use is closely related to family environment and social factors (e.g., parental influences and deviant friends) (16). There is also a positive correlation between the frequency of addictive substance use and the level of sensation seeking, which could be one predictor of drug use (17). However, the development of DUDs mainly depends on neuropharmacology and neurobiological factors (18). The GM plays a role in the initial stage of DUDs, especially in the seeking and preference of novel substances.
Specifically, people who pursue high novelty tend to be more sensitive to novel, packaged addictive drugs. Individuals with cocaine and methamphetamine (MA) use disorder have significantly higher levels of drug seeking than healthy controls, which may impact drug use (19). There is also a positive correlation between novelty seeking and relapse of cocaine users (20). Novelty seeking can partially predict individual drug susceptibility and abuse trends, and this has clinical value in the prevention of DUDs (21). Interestingly, some changes in the GM can significantly enhance the novelty-seeking behavior of animals. For example, intervention with a high-dose probiotic mixture can increase the number of Bifidobacteria in ferrets, thus increasing the time spent interacting with the novelty kettlebell and "strange animals" (22). In addition, other studies have shown that antibiotic-treated mice have decreased gut microbial content, resulting in a significant increase in their preference for cocaine, suggesting that the GM and its metabolites might enhance individual preference for cocaine (23)(24)(25).
The GM could also indirectly increase the possibility of using addictive drugs through ghrelin, a hormone produced by the gastrointestinal tract. Positive associations have been observed between ghrelin and total bacteria, Clostridium, and Ruminococcus; a negative association between an increased Bacteroidets/Firmicutes ratio, Faecalibacterium, Prevotellaceae, and ghrelin levels has also been found (26). For example, the abundance of Proteus, Bacteroides, Clostridium, and Prevotella in rats fed a restricted diet was significantly higher than in rats fed an unrestricted diet, while the number of actinomycetes, thick-walled phyla, Lactobacillus, and Bifidobacterium decreased significantly (27). The single-generation metabolites or related derivatives (e.g., short-chain fatty acids; SCFAs) of GM may be the key inducer or driving force of gut-brain communication (28). Rahat-Rozenbloom et al. (29) also found that ghrelin secretion decreased with an acute increase in SCFAs. Notably, upregulation of the ghrelin system may also increase individual cravings for drug use (30). It is closely linked to the central dopamine system and can promote the expression of dopamine receptors in the striatum, enhance reward behavior, and enhance novelty-seeking behavior in rodents (31,32). Hansson et al. (33) also observed that the injection of ghrelin into rats increased their exploration of novel targets, and these rats showed a stronger preference for the new environment. Conversely, when the ghrelin receptor was inhibited, the novel response of rats was significantly weakened. These authors also collected venous blood from human subjects and grouped genes encoding ghrelin and gastrin receptors. The results showed that ghrelin Frontiers in Psychiatry 02 frontiersin.org receptor antagonists weakened individuals' preference for a new environment and decreased novelty activity. There was a significant negative correlation between ghrelin receptor single nucleotide polymorphism and novelty traits of the subjects. Therefore, GM and its metabolites could interfere with the level of ghrelin secretion, which may have a further effect on some external behaviors (e.g., novelty seeking) of individuals to some extent. That is, the effect of the GM on ghrelin may also be an important basis for novelty seeking in rodents and humans. Taken together, GM and its metabolites could change individual's preference for novel substances in difference degrees through both direct and indirect ways, especially for the people with high susceptibility to addictive drugs. In the future, GM interventions for susceptible drug users may reduce the possibility of first use of addictive drugs.

Gut microbiota and drug taking
Drug metabolism by gut microbiota A large number and variety of intestinal flora in the human body participate in the physiological activities of the host in many forms. With the progress of biotechnology, the intestinal microflora has been gradually regarded as another "metabolic organ" of the body, and its metabolic ability is comparable to that of the liver. Previous studies have shown that the intestinal flora could catabolize drugs (34,35). Specifically, the GM directly affects in vivo drug metabolism and induces biotransformation reactions (e.g., demethylation and dehydroxylation) (36). In some conjugate hydrolysis reactions to drug toxicity, the GM may indirectly affect drug toxicity levels by regulating the competition between bacterial metabolites in different metabolic pathways (36).
Moreover, the GM may affect the decomposition and absorption of MA. Salamanca et al. (37) reviewed that although MA metabolism mainly depends on the liver, its primary and secondary metabolites are absorbed by the gastrointestinal tract for further metabolic activities. Second, individuals who used MA showed symptoms of acute transient ischemic colitis the following day, indicating that MA can be decomposed and absorbed by the human GM and various digestive enzymes, causing damage to the intestinal environment, and resulting in symptoms of intestinal disease (38). In addition, Caldwell et al. (39) showed that Lactobacillus, Enterococcus, and Clostridium in guinea pig intestines can transform MA through N-demethylation and other ways.
Similarly, the GM is also involved in the catabolism of opioids. The plasma concentration of morphine is lower in cancer patients after oral morphine administration than in cancer patients after rectal morphine administration; this may be due to morphine catabolism by the GM (40). Wang et al. (41) also revealed that microbes (e.g., Bacteroides and Bifidobacterium) could regulate the reabsorption of morphine by expressing β-glucuronidase; the loss of catabolic metabolites (e.g., those of Bacteroides) regulated morphine metabolism and enterohepatic recycling. Therefore, the intervention of GM could decrease the absorption rate of morphine to weaken effect of analgesia. This is also the main reason why long-term use of morphine will lead to tolerance and drug addiction.
Taken together, these results suggest that the GM may have a high metabolic ability for addictive drugs. As many factors can affect microbial metabolism (e.g., drug type, host, and microbiome differences), researchers are also considering combining chemistry and toxicology to predict the effects of the GM on drug metabolism and toxicity. For example, Guthrie et al. (42) proposed a graph database called the MicrobeFDT, which clusters chemically similar drug and food compounds and links these compounds to microbial enzymes and known toxicities. This set could be used to study and predict the contribution of microbial N-demethylase to drug metabolism and toxicity. In addition, future research may reduce druginduced CNS damage by using the gut-brain axis through GM intervention. An in-depth understanding of the effect of the GM on the metabolism of addictive drugs is of great significance in guiding toxicological research and the clinical treatment of DUDs.

Influence of addictive drugs on gut microbiota
Many animal and human experiments have also indicated that addictive drugs can alter the diversity of the GM. Scorza et al. (43) showed that the abundance of Spirochetaceae and Desulfovibrionaceae in cocaine-treated rat feces diminished significantly, while the abundance of Lachnospiraceae and Prevotellaceae increased. There is a significant ecological imbalance in the fecal microbial population of chronic opioid users. Another study found that the relative abundance of Bacteroidaceae in the gut of patients with cirrhosis receiving opioid treatment was significantly lower than that in patients with cirrhosis not receiving opioids (44). Additionally, Barengolts et al. (45) detected Bifidobacterium in fecal samples from male patients with type 2 diabetes and found that their abundance in opioid users was significantly higher than that in non-users. This may be because opioids affect the abundance of bifidobacteria when they are used as organic cation transporter 1 inhibitors. Studies on drug users have also shown that there are differences in GM diversity between opioid users and healthy individuals. For example, the relative abundance of Roseburia and Bilophila was lower in participants who used opioid agonists compared to participants who used neither opioid agonists nor antagonists (46). Therefore, cocaine and opioids may cause dysbiosis in the intestinal environment.
These gut microbial changes further affect brain function and have long-term effects on the CNS of the user. Of note, various symptoms associated with DUDs (e.g., emotional disorders, increased susceptibility to drugs, and brain damage) are directly or indirectly related to GM. First, changes in the GM are closely related to the various mental symptoms associated with DUDs. Forouzan et al. (47) found that after injection of MA, the fecal samples of rats in the experimental (MA) group had a higher diversity of Actinomycetes and a lower diversity of Bacteroides than in the fecal samples of rats in the control (saline) group. The relative abundances of Bifidobacterium and Lactobacillus in the experimental group were higher than in the control group, and the exploratory behavior of rats in the MA group decreased during the withdrawal period, accompanied by depressive behavior. Ning et al. (48) investigated the GM of rats with conditioned location preference in the MA and control groups using 16SrRNA and high-throughput sequencing. Their results showed that the abundance of Coccidiaceae in the control group was higher than that in the experimental group, while that of Verruciaceae and Bacillus in the experimental group was higher than that in the control group. Verrucous microflora in the intestinal microflora is associated with anxiety (49). Recently, Yang et al. (50) detected the fecal samples of MA users and found that the overall Shannon diversity index of GM in the addiction group was lower than that in the healthy group; the Enterobacteriaceae diversity in the addiction group was positively correlated with delusions, suspicions, and other mental symptoms, and the total general psychopathology scale was negatively correlated with the abundance of Collinsella and Faecalibacterium. Therefore, the long-term mental symptoms of DUDs are closely related to GM and its metabolites.
Second, GM changes are related to the preference and sensitivity for addictive drugs. For example, Yang et al. (51) revealed that after the MA-induced conditioned place preference (CPP) paradigm, the CPP score was positively correlated with the relative abundance of Verrucomicrobia (phylum) and Verrucomicrobiaceae (family). They also showed that the CPP score of antibiotic-treated rats was significantly higher than that of rats treated with distilled water, and the former had an increased preference for MA. Repeated use of morphine could cause the decrease of Olsenella and Rothia, and the increase of Helicobacter, which may have a higher risk of addictive behaviors (52). The differential relative abundance of these taxa may be the nature of rats with high/low sensitivity to morphine. The disorder of intestinal microbiota may improve the sensitivity of users to drugs and increase their preference for drugs. Therefore, the GM may be an important intrinsic factor in addictive drug-induced behavioral changes and DUDs.
Finally, the GM changes induced by DUDs are also associated with neurotoxicity and brain injury in DUD patients. The fecal samples of mice treated with multiple high doses of MA had a higher diversity of GM species than those of the control group, particularly, a reduced relative abundance of Lactobacillaceae and Prevodiaceae and an increased relative abundance of Pseudomonas and Enterobacteriaceae; compared with the control group, the expression of monoamine oxidase in the striatum of MA-treated mice increased significantly, while the expression of tyrosine hydroxylase decreased, which indicates that MA could induce dopamine terminal neurotoxicity (53). Drug use can change the diversity of intestinal flora and increase the permeability of the bloodbrain barrier. In certain cases, these have pathogenic intestinal microflora and its secretions can enter the brain and induce neurotoxicity. Choi et al. (54) also showed that an increase in the abundance of Enterobacteriaceae may be involved in the damage of dopaminergic neurons and inflammation in the substantia nigra and striatum. Cook et al. (55) sequenced the 16SrRNA gene from human male rectal swab samples and found that MA abuse was associated with significant changes in overall composition of the gastrointestinal microbiome (e.g., Parvimonas, Butyricicoccus, and Faecalibacterium), which also included some pathogenic bacteria with neural activity potential. Therefore, after addictive drug intake, pathogenic GM and its metabolites may enter the CNS through damaged barriers and signaling pathways, resulting in brain dysfunction.
To some extent, many of the above studies indicate that the decomposition and absorption of addictive drugs can affect the intestinal microbiota of the body. The frequent use of addictive drugs intensifies their influence on GM, which will have lasting effects on users' psychological and physiological functions. In addition, the gut sends signals to the brain through the production of neuroactive metabolites, signaling via the vagus nerve, and interactions with the immune system (56,57). This is a new way to understand the relationship between the GM and DUDs. Further elucidation of the interaction between the GM and the immune system may contribute to a deeper exploration of the mechanisms of DUDs.

Gut microbiota and drug relapse
Compulsive drug use is one of the core features of DUDs, causing some individuals to relapse after withdrawal. The neurobiochemical mechanism of relapse after withdrawal is mainly reflected in changes in synapses in the brain caused by using drugs (58)(59)(60). These changes mainly include the reward mechanism and drug memory, which are closely related to relapse. The abnormal connection between drugs and the reward mechanism is manifested in the activity of neurons in multiple brain regions. In addition, after abstinence, drugrelated episodic memory and cues can activate dopamine neurons and promote an individual's sense of craving and seeking behavior. In the relapse process of drug addiction, GM and its derivatives also play an invisible role in the memory and reward mechanism of drug addiction.

Gut microbiota and its metabolites in drug reward
The GM and its metabolites are directly involved in the reward mechanisms of addictive drugs. The rewarding effect is activated by dopamine neurons located in the ventral tegmental area, which project to areas such as the nucleus accumbens. The release of dopamine and other neurotransmitters promotes euphoria (61). The GM may act on the brain area of dopaminergic nerve transmission, and the dopamine circuit is sensitive to these changes (62). There is increasing evidence that the microbial-gut-brain axis may be a key factor in regulating the reward mechanism and is closely related to the occurrence of related diseases (63). Lee et al. (25) found that there was a causal relationship between the changes in GM and neuroinflammation and impaired reward response in mice treated with antibiotics. Moreover, the normal reward behavior of mice could be restored by fecal microbial transplantation. The GM and its metabolites are also necessary for the morphine reward mechanism. Hofford et al. (24) showed that the diversity of the GM and the SCFA levels decreased in antibiotic-treated mice, showing a persistent weakening of the reward effect of morphine because a decrease in diversity changes the transcriptional response of morphine in the nucleus accumbens. However, supplementation with SCFAs can reverse morphine reward defects caused by antibiotics. Therefore, the GM and its metabolites play important roles in the reward mechanism of addictive drugs.
In addition, some previous work has also suggested that the GM and its metabolites could indirectly regulate drugrelated reward pathway via glucagon-like peptide 1 (GLP-1). GLP-1, produced in the gastrointestinal tract, is encoded by the glucagon gene, and its receptor (GLP-1R) is widely found in the CNS. The change of GLP-1 levels was correlated with the disturbances experienced by different families, genera, and species of the microbiota. For example, some families, genera, and species of the phylum Actinobacteria and Firmicutes showed positive correlations with GLP-1 levels, while the families and genera of the phylum Bacteroidetes and the species Blautia producta have the opposite correlations with GLP-1 levels (64). Neurons expressing GLP-1 can project to areas of the brain associated with reward (e.g., the ventral tegmental area and nucleus accumbens) (65). The metabolite SCFAs of GM may stimulate the release of GLP-1 through the phospholipase C signaling pathway (66). Breton et al. (67) found that infusion of Escherichia coli proteins into the rat colon could also stimulate the secretion of GLP-1 and increase its concentration in the plasma. Moreover, GLP-1 and its analogs could regulate abnormal reward effects caused by drugs (e.g., cocaine, amphetamine); its receptors are expressed in the reward-related areas (68). GLP-1 and GLP-1R enhance the behavioral response of mice to cocaine, and the loss of GLP-1R could regulate the anxiety-related behavior (69). Another study showed that activation of GLP-1R in the ventral tegmental area attenuates cocaine intake in rats (70). GLP-1R antagonists reduced the self-administration behavior and recurrence behavior of MA withdrawn rats (71). The above evidence shows that GLP-1 and GLP-1R play important roles in drug relapse.
During relapse, the GM directly or indirectly enhances the reward effect of drug use. Therefore, the GM contributes to the neural mechanism of the individual pursuit of pleasure. Drugs change the GM composition and act on the CNS, and this may be the potential connection between the GM and relapse after withdrawal; microbiological therapy for this abnormal connection may reduce DUDs relapses.

Gut microbiota and drug memory
The drug-related stimuli can trigger memories of addiction and lead to re-use of the drug, so drug memory is an important factor for relapse (72,73). Although there is not a lot of evidence on the direct relationship between GM and drug addiction memory, the existing research is still suggestive.
The following studies suggest that gut microbes may influence memory function through lactic acid produced by astrocytes. Astrocytes can regulate the neuronal activity, synaptic transmission, and plasticity by providing energy and growth factors and producing neurotransmitters (74). The lactic acid is transported to the interstitial fluid via monocarboxylic acid transporter (MCT)1/4 and transported into the neuron via MCT2 (74,75). GM and its metabolites (e.g., SCFAs, glutamate) affect the production and transport of lactic acid in astrocytes (28,76). Lactic acid release in astrocytes is necessary for the development and maintenance of long-term memory. Lactic acid is involved in synaptic plasticity, memory formation, and signal transduction in DUDs (77). For example, lactic acid in astrocytes regulates synaptic plasticity, and its release contributes to the formation of cocaine memory (78). Zhang et al. (79) also found that as the concentration of lactic acid in the basolateral amygdala decreased, cocaine use also decreased in rats. This suggests that the transport of lactic acid regulated by GM and its derivations is important for the drug memory.
Taken together, as a signaling molecule, lactic acid in astrocytes is important to the drug memory (80). Although little research provides the evidence that GM can directly affect drug memory, a previous study has shown that memory disorders in rats with elevated levels of lactic acid, an important biomarker, could be improved by correcting intestinal microbial disorders (e.g., Lactobacilllus, Bacteroidales, and Bacteroides) (81,82). Therefore, future research on drug memory and reducing relapse may be carried out through GM therapy to interfere with the production and secretion of lactic acid.

Conclusion
It can be seen from this review that GM and its metabolites play a considerable role in the different stages of drug addiction. First, when the GM of individuals with high susceptibility to addictive drugs changes, they tend to seek novel substances. Second, the GM could interfere with the breakdown and adsorption of drugs in the body at the beginning of using. Third, during the withdrawal period, the intervention of intestinal microbes may interfere with the formation of drug memory to a certain extent, which is of great significance for the prevention of relapse among addictive patients.
The limitations of current technology challenge research progress but also bring opportunities. Importantly, it is helpful to explore the process of microecological circulation in the human body and reveal the mechanisms employed by the GM in individual physiology and psychology. Some researchers have pointed out that microbiological medicine is a new medical model in the 21st century (83). The microbiota will become the frontier and center of disease prevention and treatment (84). Microbiological therapy may be a more comprehensive and multi-effect approach for some chronic and recurrent diseases (85,86). This is beneficial to the development of drugs for the treatment of DUDs. Therefore, future research may reduce drug-induced damage by using the gut-brain axis through GM intervention.

Author contributions
ZW contributed to conceptualizing the work, drafting the manuscript, and managing the project. WL, MZ, CH, and LC provided support for conceptualizing and drafting the manuscript. WL and MZ supervised the project. All the authors reviewed and edited the manuscript.