**4. Dysregulation of the gut-brain axis. Evidences from experiments on animals**

Alterations in the microbial contain of the GIT are considered to contribute to inflammatory and functional bowel disorders and psychiatric comorbidities. The results of recent studies using various strains of mice and rats, various strains of probiotics and various experimental paradigms reported a number of microbial bowel modulation effects in emotional behavior [57–59], learning and memory [60], social interactions [61] and nutritional behavior [62].

#### **4.1 Experiments on mood changes in disturbed gut-microbiota in rats**

The first experiments in young GF mice which confirmed the influences of gut microbiota on postnatal development of the hypothalamic-pituitary response to stress were performed by Sudo et al. in 2004. It is interesting, that the GF mice expressed reduced anxiety-like behavior in the elevated plus maze (EPM), a reliable behavioral test that examines approach and avoidance behavior in mice, compared to specific pathogen free (SPF) mice [63].

Desbonnet et al. evaluated the potential antidepressant properties of probiotics through testing of rats chronically treated with Bifidobacteria infantis in the forced swim test [5]. Probiotic administration in naive rats had no effect on swimming behaviors. However mitogen stimulation in probiotic treated rats lead to substantial reduction of IFN-γ, TNF-α and IL-6 cytokines compared to controls (*p* < 0.05). In addition, the plasma concentrations of tryptophan (*p* < 0.005) and kynurenic acid (*p* < 0.05) were significantly elevated in the rats, treated with bifidobacteria [20]. Treatments with Bifidobacteria also lead to reduced 5-HIAA concentration in the frontal cortex and a decrease in DOPAC in the amygdaloid cortex [5].

Schroeder et al. provided evidences for production of benzodiazepine ligands in a rat model of encephalopathy or butyrate acting as a histone D-acetylesterase that was shown to have an antidepressant effect [64, 65].

The study of Arseneault-Breard et al. gave the first evidences for beneficial effect of probiotics *L. helveticus* R0052 and *B. longum* R0175 on post-myocardial infarct depression in rats. This positive probiotic influence was engaged in maintaining of the gut barrier integrity, which is possibly associated with the host' inflammatory state after MI [84].

The association of increased HPA axis responses and reduced anxiety-like behaviors observed in several of the studies performed in GF mice suggests that HPA axis and nonhypothalamic (anxiety-like behavior) components of central stress circuits may be affected on different ways according to the GF conditions, depending on species and mouse strain. These findings suggest that the increased HPA axis activity in GF animals may represent a response of the organism to the loss of microbiota-related energy sources [8].

Savignac et al. demonstrated that the two *Bifidobacterium* strains used in their study were able to improve the anxious phenotype of innately anxious BALB/c mice in a strain-specific manner and the effect was better than that from the administered antidepressant escitalopram. These findings support the statement that probiotics could be a reliable alternative for treatment of mood disorders [142].

**Figure 3.** *Impact of gut microbiota and external stress factors on behavior [66].*

On **Figure 3** is presented influence of both gut microbiota and external stressors on behavior.

### **4.2 Experiments on behavior changes in disturbed gut-microbiota in rats**

The increased hippocampal brain-derived neurotropic factor (BDNF) registered in the ATM-treated mice is corresponding with their gregarious behavior. A recent study found increased BDNF expression in the amygdala during fear learning [67, 68]. Over activation of the amygdala also has been implicated in depression and anxiety [67, 69]. Lower levels of BDNF in the amygdala of ATM treated mice were associated with increased exploratory behavior.

Bercik et al. found that SPF mice who received antimicrobial agents per os demonstrated enhanced exploratory behavior and hippocampal expression of BDNF. This finding was associated with temporary alteration of the representatives of their microbiota and was not accompanied by inflammatory status, alteration of gastrointestinal neurotransmitters levels, nor with vagal or sympathetic function. Intraperitoneal application of antimicrobial agents to SPF mice, similar to their oral administration in GF mice had no influence on behavior. Increased exploratory behavior and high hippocampal levels of BDNF were reported in GF BALB/c mice, colonized with microbiota from NIH Swiss mice. Suppression of exploratory behavior was demonstrated in GF NIH Swiss mice, colonized with BALB/c microbiota [2, 70].

The study of Bercik et al. did not provide proof for intestinal inflammation, as oppose to Verdú et al.' investigation [71], where administration of ATMs in a higher dose and for a longer period was made in NIH Swiss mice. In the Bercik's experiment embarrassment of the intestinal microbes did not change myeloperoxidase activity, histology or cytokine profile of the colon [8]. No differences in serotonin, dopamine, or noradrenaline content in the gut of ATM-al. treated mice were observed, suggesting that these neurotransmitters are not involved in mediating the behavioral changes observed in the model.

Li et al. and Bercik et al. reached similar results on memory and learning skills in adult mice [11], applying different nutritional supplements to animals at a very early age with the following disruption of the intestinal flora in very young age. Working and referred memory was better in the animals on rich in beef diet as opposed to the mice on standard meal [8].

Neufeld et al. supposed that the low anxiety-like phenotype was accompanied by long-term changes in plasticity-related genes in the hippocampus and amygdala. They found altered GF behavior, accompanied by a decrease in the N-methyl-D-aspartate receptor subunit NR2B mRNA expression in the central amygdala, increased BDNF expression and decreased serotonin receptor 1A (5HT1A) expression in the dentate granule layer of the hippocampus. It is the first work which demonstrated an altered behavioral phenotype related with lack of gut microbiota [59].

In their work Bravo et al. registered increased levels of GABAB1b mRNA in cingular and prelimbic areas in mice treated for a long time with *L. rhamnosus* (JB-1), while the concentration of these neurotransmitters was reduced in the hippocampus, amygdala and locus coeruleus in the same experimental animals. Furthermore, the GABAAα2 level was reduced in the prefrontal cortex and amygdala, and increased in the hippocampus. The observed mice expressed reduced response to stress, associated with releasing of corticosterone. Similar neurochemical and behavioral effects were not expressed in mice, who has underwent vagotomy [12, 73].

In their study Park et al. demonstrated that depressed-like behavior in mice that underwent bilateral olfactory bulbectomy (OBx) was associated with altered colonic motility and a shift in the microbiota profile. Their experiment also supposed that increased prokinetic neuropeptide, gut hormone and serotonin in the colonic wall are mediators of the altered motility [25]. Their finding was consistent with those of Rodes et al. who showed changed colonic transit and altered stability of the colonial microbial community [74].

Hsiao et al. demonstrated GI barrier defects and microbiota alterations in the maternal immune activation (MIA) mouse model who displayed ASD signs. MIA generation, who has received *Bacteroides fragilis* (human commensal microbe) per os, has evolved altered bacterial gut content which predisposes to impaired communication and manifestation of stereotypic, anxiety and sensorimotor behavior. The described experimental model showed change in profile of the serum metabolites and their levels. It is other evidence for the gut microbiota impact on human behavior through the gut microbiome-brain functional axis and it could help in searching of relevant probiotic treatment of behavior disturbances in neurodevelopment diseases in human [3, 75].

It was found that gut microbiota status reduce social interactions in GF mice and probiotics improve social interactions in a post-MI rat model. Desbonnet et al. examined whether the degree of information transfer during social interaction is disrupted in GF mice. In their experiment GF mice spent a decreased proportion of time engaged in social investigation and substantially greater proportion of time engaged in repetitive self-grooming behavior during social interaction. After GF bacterial colonization these behaviors were normalized, which is evidence for involvement of microbiota in modulation of such behaviors. However, the quality of information transfer during the interaction was not affected in GF mice, indicating that the ability to process information per se during social interaction was not affected in GF mice [76].

It is important to note that many of the psychologic deficits, registered in GF mice, are specific to males in which higher incidence of neurodevelopmental disorders was registered compared to females [59, 63, 77–79]. de Theije et al. demonstrated that gender-specific inflammatory conditions are present in the small intestines of VPA in utero-exposed mice and are accompanied by a disturbed serotonergic system both in the brain and in the intestinal tract [80]. Gut microbiota-associated behavioral changes were reported in different ASD mouse models using valproic acid administration or maternal infection; in the latter instance some behavioral disorders were favorably influenced by probiotic therapy with *Bacteroides fragilis* [8, 9].

Several studies proposed the influence of intestinal microorganisms on eating behavior [80], probably as a consequence of modified fatty acid receptors, gut receptors, responsible for taste, alteration of the intestinal transportation mechanisms or disturbed releasing of satiety hormones [9, 81, 82].

Crumeyrolle-Arias et al. found that lack of intestinal microbiota in sensitive to stress strain rats lead to neuroendocrine and behavior reactions of acute stress and significant changed degree of the dopaminergic turnover in the higher brain structures which modulate stress and anxiety—another support for the crucial impact of the gut microbiota on the higher brain activities [9, 15, 83, 84].

Recently it was reported that impaired permeability of the blood brain barrier in GF mice probably will restrict reaching of the liver bacterial metabolites to the brain [85]. Numerous remodeling experiments in GF animals confirmed that deviations of brain metabolism and behavior could be preserved through reconstitution of the gut microbial composition [1, 86].

Wong et al. found that genetically determined caspase-1 deficit in mice suppresses the anxiety-depressive like behavior and improves the motor activity and locomotor abilities, as well prevents manifestation of depressive symptoms after chronic

exposition at stressors. On the other hand, minocycline as pharmacological antagonist of caspase-1 alleviates the depressive like symptoms in wild type mice provoked by stress. Actually, both chronic stress and pharmacological inhibition of caspase-1 modify the composition of fecal microorganisms almost in the same way [3, 87].

The GF model has some limitations, which suppose that the investigators should be cautious in extrapolating the conclusions obtained in animals on people. Important is the fact that GF mice are born under aseptic conditions, such as separation from the mother via cesarean section and directly placement of the newborn in an special insulator in which the air, in which everything is sterilized, including the air, food and water. The biochemistry of brain and gut intestine is quite different [1, 81], HPA axis responses [63], in emotional, [58], social [75, 79], metabolic function, and ingestive behaviors [82] between GF animals and control animals which contain normal or pathogen-free flora obtained by colonization from the mother [8, 78]. However, up to date studies with animal model proved that the gut microbiota can influence the central nervous system in the absence of substantial changes in local or circulating cytokines or specific intestinal neurotransmitter.

It is unambiguous that bacterial products can get access to the brain via the bloodstream, they can act through the immune system via cytokine releasing by the mucosal immune cells, or through the endocrine system by triggering gut hormone release from enteroendocrine cells [9, 87]. Since GF animal models are not analogous to the development of the human brain, premature conclusions about the significance of these findings to humans should be avoided [88].
