The role of the probiotic Akkermansia muciniphila in brain functions: insights underpinning therapeutic potential

Abstract

The role of Akkermansia muciniphila, one of the most abundant microorganisms of the intestinal microbiota, has been studied extensively in metabolic diseases, such as obesity and diabetes. It is considered a next-generation probiotic microorganism. Although its mechanism of action has not been fully elucidated, accumulating evidence indicates the important role of A. muciniphila in brain functions via the gut-brain axis and its potential as a therapeutic target in various neuropsychiatric disorders. However, only a limited number of studies, particularly clinical studies, have directly assessed the therapeutic effects of A. muciniphila interventions in these disorders. This is the first review to discuss the comprehensive mechanism of A. muciniphila in the gut-brain axis via the protection of the intestinal mucosal barrier and modulation of the immune system and metabolites, such as short-chain fatty acids, amino acids, and amino acid derivatives. Additionally, the role of A. muciniphila and its therapeutic potential in various neuropsychiatric disorders, including Alzheimer’s disease and cognitive deficit, amyotrophic lateral sclerosis, Parkinson’s disease, and multiple sclerosis, have been discussed. The review suggests the potential role of A. muciniphila in healthy brain functions.

Keywords:

• Akkermansia muciniphila
• probiotic
• neuropsychiatric disorders
• gut-brain axis
• therapeutic target

1. Introduction

Akkermansia muciniphila, a gram-negative, anaerobic, host-derived mucin-degrading bacterium, is by far the best characterised species of the Akkermansia genus and was first identified in 2004 by Derrien et al. (2004); Belzer and De Vos (2012). Akkermansia, named after the renowned microbial ecologist Antoon D. L. Akkermans, is a unique representative and culturable genus that belongs to the family Akkermansiaceae, Verrucomicrobiae class I, and phylum Verrucomicrobia (Derrien et al. 2008). A. muciniphila resides mostly within the mucus layer and colonizes organs of the vertebrate gastrointestinal (GI) tract, such as the caecum, colon, and ileum (Geerlings et al. 2018). It has also been isolated from the upper digestive tract (Ye et al. 2016) and breast milk (Collado et al. 2012; Aakko et al. 2017). A. muciniphila may represent 3%–5% of the microbial community (Derrien et al. 2004; Belzer and De Vos 2012) and 1%–4% of the faecal microbiome (Collado et al. 2007; Everard et al. 2011).

A. muciniphila protects the mucus layer and associates with various intestinal disorders, such as inflammatory bowel disease (Png et al. 2010), ulcerative colitis (Earley et al. 2019), and is inversely related to the severity of acute appendicitis (Swidsinski et al. 2011). Studies also confirm the beneficial effects of A. muciniphila in metabolic diseases, such as diabetes (Hanninen et al. 2018) and obesity (Lukovac et al. 2014; Plovier et al., 2016). Additionally, A. muciniphila could act as a biomarker of health status and a “next-generation probiotic” (Zhang et al. 2019b). Moreover, the benefits of this probiotic are not limited to the improvement of metabolic functions and conditioning of the immune system (Belzer and De Vos 2012; Reunanen et al. 2015; Plovier et al., 2016; Van Herreweghen et al. 2017; Ottman et al. 2017b; Geerlings et al. 2018; Naito et al. 2018; Zhai et al. 2019a; Zhang et al. 2019c). Recently, A. muciniphila was considered as biomarker of response to immune checkpoint inhibitors (ICIs) and idenpendently associated with overall survival in non-small-cell lung cancer (NSCLC) (Routy et al., 2018; Derosa et al., 2022). As shown in investigations of the “Microbiome-Gut-Brain axis (MGB),” the critical roles of A. muciniphila have been suggested in multiple neuropsychiatric disorders, such as depression and anxiety, Alzheimer’s disease (AD) and cognitive impairment, substance use disorders (SUDs), amyotrophic lateral sclerosis (ALS), multiple sclerosis (MS), autism spectrum disorders (ASDs), epilepsy, Parkinson’s disease (PD), and stroke. This is the first review to summarize these findings and discuss the possible mechanisms underlying the role of A. muciniphila in the gut-brain axis, as well as its therapeutic potential. Although more direct and strong evidence is required, A. muciniphila, with its safety and tolerability in clinical applications, might be considered as a potential probiotic in the MGB, especially as a promising, long-term, and moderate treatment for neuropsychiatric disorders.

2. Mechanism of action of A. muciniphila in brain functions

Since few studies have explored the direct impact of A. muciniphila on the nervous system, most hypotheses discussing the mechanisms of action of A. muciniphila in brain functions are largely based on the functions of A. muciniphila in protecting the intestinal mucosal barrier and modulating the immune system, metabolic system, and endocannabinoid system through metabolites such as short-chain fatty acids (SCFAs), amino acids, and amino acid derivatives, which are produced by A. muciniphila from mucin glycolipids, mucin proteins, and mucin hexoses or oligosaccharides. Various parts of A. muciniphila, including live or pasteurized A. muciniphila, an outer membrane protein Amuc_1100, as well as extracellular vesicles (AmEVs), have shown to regulate metabolic system and intestinal barrier integrity through up-regulating tight-junction protein expression and reducing LPS leakage, thus further reducing inflammation (Yan et al. 2021).

2.1 A. Muciniphila protects the integrity of the mucosal barrier

A. muciniphila counteracts mucosal barrier dysfunction by increasing mucus-producing goblet cell levels (Grander et al. 2018) and tight-junction protein expression (Chelakkot et al. 2018), thereby enhancing mucus production (Hanninen et al. 2018) and preserving mucosa thickness (Ottman et al. 2017a). This leads to the decrease of intestinal permeability (Bian et al. 2019) and strengthening of epithelial barrier function (Reunanen et al. 2015). Oral administration of A. muciniphila upregulated the expression of ZO-1 (Huck et al. 2020) and Occludin (Wu et al. 2017) and effectively thickened the mucosal barrier (Everard et al. 2013). Moreover, co-culture of Caco-2 cells in vitro, A. muciniphila adhered to the intestinal epithelium, improved the enterocyte monolayer integrity of Caco-2 cells, (Reunanen et al. 2015) and increased in the transepithelial electrical resistance (TEER) (Reunanen et al. 2015; Ottman et al. 2017b). A. muciniphila also increased the levels of gatekeeping molecules, such as 2-arachidonoylglycerol, 2-oleoylglycerol (2-OG), and 2-palmitoylglycerol, which helped preserve the gut barrier and reduce intestinal inflammation (Everard et al. 2013). Furthermore, A. muciniphila is known to regulate the production of glucagon-like peptide-1 and 2 (GLP-1 and GLP-2), whose release can be stimulated by 2-OG (Everard et al. 2013). These two gut peptides are not only involved in glucose regulation but also in gut barrier modulation (Drucker 2001). Amuc-1100, a highly abundant outer membrane pill-like protein, elevated the development of TEER in Caco2-cells (Ottman et al. 2017b). Pasteurized A. muciniphila and Amuc-1100 increased the expression of tight junction proteins and improved intestinal barrier function (Plovier et al., 2016). In addition, A. muciniphila-derived extracellular vesicles (AmEVs) were shown to decrease the gut permeability by inducing AMPK activation (Chelakkot et al. 2018). These findings confirmed that A. muciniphila plays an important role in protecting the intestinal mucosal barrier.

2.2. Impact of A. muciniphila on the immune system

Complimentary to its protective effect on intestinal permeability, A. muciniphila also plays an important role in alleviating systemic inflammation. It was shown that a reduction in A. muciniphila abundance is associated with higher inflammatory scores (Earley et al. 2019). Moreover, the oral administration of live A. muciniphila or AmEVs helped reduce inflammatory infiltration (Kang et al. 2013a; Bian et al. 2019). Concurrently, live A. muciniphila administration suppressed the overall infiltration of mononuclear leukocytes and reduced the levels of TLR2, TLR4 (Ottman et al. 2017b), and proinflammatory cytokines (including TNF-α, IL1α, IL6, and IL12A), and upregulated IL-10 expression (Bian et al. 2019; Zhai et al. 2019b). A. muciniphila administration also tended to increase the expression of the antibacterial peptide Reg3γ (Hanninen et al. 2018), which is associated with antibacterial and anti-inflammatory functions. In diabetes, A. muciniphila treatment was found to increase the levels of anti-inflammatory (type 2) macrophage, Ym1, and restore the number of regulatory Foxp3+ Treg cells (Shin et al. 2014; Hanninen et al. 2018). Furthermore, Amuc-1100 protein is identified as the major factor inducing specific cytokines, such as IL-10, through Toll-like receptor (TLR) 2 and 4 activation (Ottman et al. 2017b). Amuc-1100 treatment improves colitis by reducing infiltrating macrophages and CD8 cytotoxic T lymphocytes (CTLs) in the colon (Wang et al. 2020a). Furthermore, Amuc-1100 increases kynurenine (Kyn) levels, decreases 2-picolinic acid (PIC) levels, and affects the PIC/Kyn ratio without regulating any of the genes involved in Trp metabolism, suggesting that it could suppress the Kyn pathway (KP), independent of colon tissue (Gu et al. 2021). Additionally, Amuc-1100 could promote the expression of the 5-HT synthesis rate-limiting enzyme (Tph1) in RIN-14B cells and reduce the expression of the serotonin reuptake transporter (SERT) in Caco-2 cells through direct interaction with TLR2, thereby improving 5-HT biosynthesis and extracellular availability (Wang et al. 2021). A secreted protein of A. muciniphila, P9, induced both GLP1and IL-6 thus regulated glucose homostasis (Yoon et al. 2021). So far, few studies have focussed on the role of A. muciniphila in the immune responses of the brain. Treatment with A. muciniphila was found to prevent high-fat diet (HFD)-induced microgliosis and proinflammatory cytokine expression in the hippocampus (Yang et al. 2019). These findings collectively suggest an anti-inflammatory role of A. muciniphila on the immune system in both the gut and brain of the host.

However, it is interesting to note that A. muciniphila significantly exacerbated Salmonella typhimurium-induced intestinal inflammation in gnotobiotic mice formed from germ-free C3H mice, compared to the absence of A. muciniphila; whereas the A. muciniphila colonization did not exhibit an increase in pro-inflammatory cytokines as seen in S. typhimurium infection. This indicates the complex interactions within the microbial ecosystem (Ganesh et al. 2013). Furthermore, the pathobiont role of A. muciniphila in a genetically susceptible host has been suggested (Seregin et al. 2017). A. muciniphila also showed increased in mice that with higher rates of colitis-associated tumorigenesis, simultaneously, genes linked to mucin degradation were also positively correlated with tumour incidence (Baxter et al. 2014). Similarly, A. muciniphila was significantly decreased in immunodeficient Rag–/– mice and bone marrow transfer specified suppressed the colonization of the A. muciniphila in adult Rag–/– mice (Zhang et al. 2015). These evidences suggested that A. muciniphila might be a biomarker for immunodeficiency. The contradictory change of A. muciniphila colonization and the anti-inflammatory role of A. muciniphila could be result from self-compensatory mechanism. However, the underlying mechanism might be complicated, but it reminds us to think carefully about the function of the microbiome regarding homeostasis integrity; the balance within microbial ecosystems, among protective, butyrate-producing populations and inflammatory, mucin-degrading populations; and the balance between microbiome and the host with genetic diversity (Alberdi et al. 2021; Nichols and Davenport 2021).

2.3. The relationship between A. muciniphila metabolites and brain function

Studies have demonstrated that one of the molecules produced by A. muciniphila is an oligosaccharide chain-degrading enzyme (e.g. glycosidases and sulfatases) that degrades gut mucin into various products. It was speculated that A. muciniphila and Faecalibacterium prausnitzii play an anti-inflammatory role by upregulating anti-inflammatory factors and downregulating inflammatory factors through their metabolites and amino acid derivatives (Demirci et al. 2019). SCFAs are important metabolites produced by the microbiome. A study elucidated the immunomodulatory function of SCFAs, which are associated with human health and gut/immune homeostasis (Tan et al. 2014). In this part of the review, we focus on how SCFAs and amino acid derivatives produced by A. muciniphila affect the central nervous system through the MGB.

2.3.1. SCFAs

To date, several studies have provided direct and indirect evidence on the role of A. muciniphila-derived SCFAs in the MGB. Among several clinical studies on the traditional Mediterranean diet, one showed its protective effect on glucose metabolism, as indicated by an increase in the postprandial plasma butyric acid levels as well as in the abundance of A. muciniphila; another study showed improvements in the levels of AD biomarkers in the cerebrospinal fluid, along with increased A. muciniphila abundance, faecal propionate, and butyrate, in a typical Mediterranean diet group (Vitale et al. 2021). An in vitro study showed that A. muciniphila abundance and the levels of propionate and butyrate, but not the levels of acetate, affected the expression of Fiaf, Gpr43, and histone deacetylases (HDACs) in a mouse gut organoid model, indicating that the function and metabolites compounds of A. muciniphila overlap with propionate and butyrate, but not acetate (Lukovac et al. 2014). A thesis from Prof. W.M. de Vos’s group showed that propionate and acetate were produced in the culture medium of A. muciniphila when mucin and glucose were used as substrates (Ottman 2015). Another study conducted in 2019 also showed that both the type strain and an A. muciniphila^sub clinical isolate produced acetic acid, propionic acid, and isovaleric acid, with isovaleric acid showing the highest concentration in in vitro analysis (Wu et al. 2020b). These results indicate the role of A. muciniphila in the production of SCFAs. These SCFAs, primarily butyrate, have been considered as regulators of HDACs (Waldecker et al. 2008), and are known to participate in brain development and the induction of several diseases, including depression (Lv et al. 2020), schizophrenia (Joseph et al. 2017), and AD (Doifode et al. 2021). Therefore, these findings indicate the crucial role of A. muciniphila-derived SCFAs in the MGB.

2.3.2. Amino acid derivatives

In a study by Dooling et al., A. muciniphila acted synergistically to reduce the gamma-glutamylation of amino acids and increase the hippocampal gamma-aminobutyric acid (GABA)/glutamate ratio, which eventually helped prevent seizures (Dooling and Costa-Mattioli 2018). HFD administration can lead to the elevation of serum tryptophan levels, which was shown to be reduced by A. muciniphila subtype gavage. In addition, the serum serotonin levels in HFD-fed mice were significantly lower than those in mice fed a normal diet (p < 0.05), indicating that HFD administration significantly decreased the serum serotonin concentration, which was not altered by A. muciniphila subtype gavage (Wu et al. 2020b). Furthermore, Sod1-Tg mice that were administered A. muciniphila were found to exhibit A. muciniphila-associated nicotinamide accumulation in the central nervous system, and systemic nicotinamide supplementation in Sod1-Tg mice improved their motor symptoms and spinal cord gene expression patterns (Blacher et al., 2019). The oral administration of A. muciniphila for 21 days in chronic restraint stress mice induced recovery from abnormal variations in the levels of corticosterone, dopamine, serotonin, and brain-derived neurotrophic factor, which are closely associated with the stress behavioural response (Ding et al. 2021). These findings support the essential role played by A. muciniphila in the MGB via the regulation of hormones, neurotransmitters, and neurotrophic factors.

3. Role of A. muciniphila in neuropsychiatric diseases

Accumulating evidence suggests a correlation between the gut microbiome, including A. muciniphila, and neuropsychiatric disorders. However, alterations in the abundance of A. muciniphila vary in different neuropsychiatric disorders. The abundance was shown to decrease significantly in depression and anxiety, AD and cognitive deficits, SUDs, epilepsy, and ALS; whereas it was shown to increase significantly in MS and PD. Here, we summarize findings from studies on the correlation between neuropsychiatric disorders and alterations in A. muciniphila abundance.

3.1. Depression and anxiety

It is well known that stress, both physical and psychological, affects the gut microbiome (Guo et al. 2016; Wong et al. 2016; Mcgaughey et al. 2019; Song et al. 2019) by altering its composition, and consequently, modulates depressive-like behaviour via the hypothalamic-pituitary-adrenal axis (Cryan and Dinan 2012; Foster and Mcvey Neufeld 2013).

In animal models, chronic adrenocorticotrophic hormone (ACTH) administration induced depressive-like behaviour, increased the abundances of Ruminococcus and Klebsiella, and reduced the abundances of Akkermansia and Lactobacillus; further, the abundances of Akkermansia and Lactobacillus were negatively correlated with integrated metabolomic signatures in ACTH-induced depression model rats (Song et al. 2019). Chronic paradoxical sleep-deprived rats displayed depressive-like behaviour and changes in the metabolic profile and microbial composition, including a significant reduction in the Akkermansia abundance (Park et al. 2020). Additionally, mice exhibiting depressive- and anxiety-like behaviour following social defeat showed decreased Akkermansia abundance, and the reduction in the Akkermansia spp. abundance was negatively correlated with anxiety- and depressive-like behaviour in sucrose-preference tests and open-field tests (Mcgaughey et al. 2019). Meanwhile, in mice subjected to minocycline administration under restraint stress (Wong et al. 2016), prebiotics, including fructo-oligosaccharides and galacto-oligosaccharides (Burokas et al. 2017), successfully attenuated depressive- and anxiety-like behaviour, besides increasing the abundance of Akkermansia. Fish oil ameliorated depressive-like behaviour induced by chronic mild stress. In the same study, olive oil significantly increased the Akkermansia abundance, but did not ameliorate depressive-like symptoms (Tung et al. 2019). Recently, in the FinnBrain Birth Cohort Study, the association between prenatal psychological distress (PPD) in 398 mothers and the faecal microbial composition of their 2.5-month-old infants was analysed; the result showed a negative correlation between Akkermansia abundance and the symptoms of maternal PPD (Aatsinki et al. 2020). A preclinical study showed that A. muciniphila significantly ameliorated depressive-like behaviour induced by chronic restraint stress (Ding et al. 2021). Taken together, both animal and human studies consistently reported a negative correlation between depressive behaviour and the abundance Akkermansia, suggesting that increasing the A. muciniphila abundance could be a potential method for treating depressive- and anxiety-like behaviour. However, direct evidence from clinical studies is necessary to explore the therapeutic potential of A. muciniphila in depression and anxiety.

3.2. AD and cognitive deficit

Accumulating evidence shows the impact of the microbiome on AD and cognitive functions, and the gut microbiome is considered a novel treatment target in AD (Jiang et al. 2017; Davidson et al. 2018; Chu et al. 2019; Bulgart et al. 2020). A study on the gut microbiome in an APP/PS FAD transgenic mice model at different time points since 1 month old showed that before obvious amyloid plaque deposition and microglial activation, signs of divergence in the microbiota were observed, including a significant increase in the abundance of Akkermansia, which helped identify potential microbiome biomarkers, particularly Akkermansia, for AD (Chen et al. 2020). However, three different studies conducted by two research groups consistently showed that a modified Mediterranean ketogenic diet or ketogenic diet (KD), which also boosted the abundance of Akkermansia, helped to prevent the decline in cognitive function in patients with AD by improving the AD biomarker profile and brain vascular functions as well by altering the abundances of individual gut microbes (Ma et al. 2018) or Akkermansia (Nagpal et al. 2019; Neth et al., 2020). Two clinical studies from the same group reported that the larger proportions of Verrucomicrobia were significantly correlated with better performances on cognitive tests (Anderson et al. 2017; Manderino et al. 2017). Furthermore, recent direct evidence from four laboratory-based studies published at a similar time demonstrated the beneficial effects of A. muciniphila in patients with AD and cognitive deficits (Yang et al. 2019; Ou et al. 2020; Wu et al. 2020a; Higarza et al. 2021). Ou et al. showed that in AD animal models, intervention with A. muciniphila reduced the Aβ 40–42 levels in the cerebral cortex and alleviated cognitive impairments, such as dyslipidemia and spatial learning and memory deficits (Ou et al. 2020). The other three studies showed that the oral administration of A. muciniphila or its subtype for 4 weeks was sufficient to prevent HFD-induced memory decay, as indicated by the performance in the Y-maze test (Wu et al. 2020b), contextual fear conditioning test and Barnes circular maze test (Yang et al. 2019), and novel objective recognition and spatial working memory (Higarza et al. 2021). The authors also found that A. muciniphila reversed microgliosis, proinflammatory cytokine expression, neuronal development, and long-term potentiation in the hippocampus of HFD-fed mice, besides restoring the expression of the GluA1 and GluA2 subunits (Yang et al. 2019). Higarza et al. also reported the beneficial role played by A. muciniphila in improving the oxidative metabolic activity in the brain by restoring the activity of the mitochondrial enzyme cytochrome C oxidase, which indicated ATP production and brain energy demand (Higarza et al. 2021). Studies on the beneficial effects of A. muciniphila in AD and cognitive functions have been conducted only recently, after the safety of A. muciniphila was confirmed in clinical applications; as a prospective disease treatment/prevention strategy, A. muciniphila can be administered long-term at a moderate dose in advance to prevent disease progression in patients with a higher risk of AD or mild cognitive impairments. However, such applications require more direct evidence for validation, particularly evidence from human studies.

3.3. SUDs

The role of the gut microbiome in SUDs is well studied (Meckel and Kiraly 2019). The role of Akkermansia, or A. muciniphila specifically, might vary in different SUDs. Acute-on-chronic alcohol feeding altered the gut microbiome at different taxonomic levels, and the loss of Akkermansia was observed as an early marker of alcohol-induced gut dysbiosis (Lowe et al. 2017). Patients with alcohol use disorder presented a unique gut microbiome, with low abundance of Akkermansia but an increased abundance of Bacteroides, with an accuracy of 93.4% (Addolorato et al. 2020). A study showed the decreased A. muciniphila abundance in patients with alcoholic hepatitis, which correlated with the severity of alcoholic liver disease (ALD). Furthermore, using an animal model, the authors showed that A. muciniphila supplementation prevented alcohol-induced hepatic injury and ameliorated ALD symptoms by restoring the intestinal barrier integrity (Grander et al. 2018). This evidence directly demonstrated the beneficial role of A. muciniphila in ALD. However, the role of A. muciniphila in other SUDs is debateable. Morphine tolerance is related to microbial ecological disorders. The addition of the probiotic VSL#3 reinstated morphine analgesic tolerance by restoring the abundances of Bifidobacteria and Lactobacillaeae while decreasing the abundance of A. muciniphila (Zhang et al. 2019a). A study conducted by our group (Yang et al. 2020) showed that the Akkermansia abundance was significantly higher in rats with higher methamphetamine (MA) reward response and positively correlated with the MA-induced conditioned place preference (CPP) scores, which is a widely used experimental index for measuring the rewarding effects of drugs on laboratory animals. Meanwhile, contrasting patterns were observed for Acetivibrio. Treatment with an antibiotic cocktail reduced the abundance of the different strains in the gut microbiome, but increased the abundance of Akkermansia, and also boosted MA-induced CPP (Yang et al. 2020). To summarize, patients with ALDs might benefit from A. muciniphila supplementation, whereas the role of A. muciniphila in other SUDs remains unclear.

3.4. ALS

ALS is a complicated neurodegenerative disorder characterized by premature irreversible degeneration of the upper and lower motor neurons. In a study by Blacher et al., the serum levels of key molecules of the tryptophan–nicotinamide metabolism pathway, including indole acetic acid, kynurenine, 5-hydroxytryptamine, and cyclic nicotinamide, changed significantly in patients with ALS (Blacher et al. 2019). Furthermore, 37 patients with ALS showed a notable reduction in the nicotinamide levels compared to 29 healthy controls, and also showed dysbiosis (Blacher et al. 2019). In Sod1-Tg mice, an ALS animal model, maintenance under germ-free conditions and treatment with broad-spectrum antibiotics exacerbated the motor symptoms. Eleven commensal bacterial strains were selected for the treatment of Sod1-Tg mice, as the abundances of these strains correlated with the severity of ALS symptoms, but only the administration of A. muciniphila ameliorated ALS symptoms and improved survival by effectively increasing the nicotinamide levels in the central nervous system. In addition, systemic nicotinamide supplementation improved the motor symptoms and gene expression patterns in the spinal cord of Sod1-Tg mice (Blacher et al. 2019). This finding indicated the therapeutic action of A. muciniphila mediated via metabolic pathways in patients with ALS. However, additional studies conducted using other genetic ALS models and investigating the interplay among the microbiome, immune system, metabolism, and genetic differences are warranted (Gotkine et al. 2020)

3.5. MS

The symptoms of MS, a putative autoimmune disease of the central nervous system, have been shown to be strongly associated with Akkermansia abundance in recent years (Berer et al. 2017). Studies have consistently reported the significant increase in the relative abundance of A. muciniphila (Jangi et al. 2016; Berer et al. 2017; Cekanaviciute et al., 2017; Tankou et al. 2018; Al-Ghezi et al. 2019) in patients with MS compared to that in healthy controls; the same was reported in twins with MS compared to that in healthy twins in a study conducted on 34 pairs of monozygotic twins discordant for MS (Berer et al. 2017) as well as in a mouse model of experimental autoimmune encephalomyelitis (EAE) compared to that in control mice (Al-Ghezi et al. 2019). Further, faecal microbiome transplant from patients with MS to mice induced a pro-inflammatory environment, which consequently aggravated the severity of EAE (Berer et al. 2017; Cekanaviciute et al. 2017). The exposure of peripheral blood mononuclear cells to total bacterial extracts from patients with MS with detectable A. muciniphila in vitro significantly boosted their differentiation into IFNγ⁺ Th1 lymphocytes, whereas the same was not observed when the cells were exposed to bacterial extracts devoid of A. muciniphila (Cekanaviciute et al. 2017). This evidence indicated the crucial role of Akkermansia in the pathogenesis of MS mediated via pro-inflammatory responses. Moreover, treatments that reverse MS-induced inflammation decreased the abundance of Akkermansia. For example, probiotic supplementation with Lactobacillus, Bifidobacterium, and Streptococcus weakened the pro-inflammatory microenvironment, which also decreased the abundance of Akkermansia, in nine patients with MS (Tankou et al. 2018). Combined treatment with the cannabinoids delta-9-tetrahydrocannabinol and cannabidiol effectively reduced the abundance of A. muciniphila in the gut and the LPS levels in the brains of EAE mice, which helped mitigate EAE (Al-Ghezi et al. 2019). These findings indicated the crucial role played by Akkermansia in MS pathogenesis by inducing pro-inflammatory responses. A. muciniphila was suggested as a biomarker in the EAE model (Al-Ghezi et al. 2019). However, evidence from a very recent study conducted by the Howard L. Weiner group seems to contradict this view, as the authors showed the compensatory beneficial role of A. muciniphila in MS. They found that A. muciniphila abundance was negatively correlated with the age-adjusted disability score and MRI indicators of disease severity. Furthermore, they isolated three subtypes of A. muciniphila from patients with MS, inoculated mice with these, and isolated a new A. muciniphila strain that inhibited RORγT and IL-17A expression in CD4⁺ or CD8⁺ T cells and ameliorated the symptoms of EAE in mice (Cox et al. 2021). To summarize, the increased abundance of Akkermansia in patients with MS might be a compensatory effect, whereas the abundances of other bacteria were highly correlated with enhanced EAE progression and MS pathogenesis. Treatments that increase the abundance of certain A. muciniphila strains may be effective in reversing the symptoms of EAE and MS. However, more direct investigations are necessary to identify the biomarkers and treatments targeting the gut bacteria in patients with MS.

3.6. ASDs

ASDs are characterized by difficulties in social interaction, communication, and repetitive and restricted behaviours (Rabe-Jabłońska and Bieńkiewicz 1994). In addition to their effects on mental health, ASDs have been reported to be associated with GI problems in 23%–70% of affected individuals. Individuals with ASDs without GI problems are more prone to GI problems than healthy individuals (Chaidez et al. 2014). An altered gut microbiome has also been found in children with ASDs (Wang et al. 2011; De Angelis et al. 2013). Manipulations of the gut microbiome in mice can be induced either by genetic modification (Chaidez et al. 2014) or in response to the diet-influenced symptoms of ASDs (Newell et al. 2016).

The alteration of the gut microbiome has been studied in patients with ASDs; however, inconsistent patterns of changes in Akkermansia abundance have been reported in different studies on ASDs (Xu et al. 2019). Some researchers have reported an increase in the relative faecal abundance of Akkermansia in patients with ASD compared with that in healthy controls (Finegold 2008; Finegold et al. 2010; Inoue et al. 2016; Lee et al. 2017), whereas other studies have shown a decrease in the same parameter in patients with ASDs (Kang et al. 2013b; Strati et al. 2017; Zou et al. 2020) A meta-analysis of five trials showed that the overall percentage of Akkermansia abundance in patients with ASDs is lower than that in controls, whereas the effect size of Akkermansia is relatively small (Xu et al. 2019). Another study showed that a gut microbiome transplant from human donors with ASD to germ-free mice sufficiently reproduced the typical human autistic behaviours in mice. Specifically, an increase in the abundance of A. muciniphila and reduction in the abundances of Bacteroidetes, Bacteroides, and Parabacteroides were observed in mice inoculated with the microbiome of patients with ASD (Sharon et al. 2019). These findings demonstrated that the faecal microbiome contributes to the development of ASD symptoms; however, given the multifarious results, the relationship between Akkermansia abundance and the occurrence or prognosis of ASD remains elusive. Further studies are needed to confirm this aspect.

3.7. Epileptic seizures

Epilepsy is a chronic disease in which the neurons of the brain fire abruptly and abnormally, causing transient brain dysfunction. An effective treatment protocol for epilepsy is KD administration (Dooling and Costa-Mattioli 2018). In an elegant preclinical study by Olson et al., the abundances of Akkermansia and Parabacteroides increased significantly when conventionally colonized SPF Swiss Webster mice were fed a KD with fat and protein in a 6:1 ratio (Olson et al. 2018) The therapeutic mechanism of KD in epilepsy may involve the regulation of the abundance of intestinal microbiota. Moreover, KD administration or direct intervention with A. muciniphila and Parabacteroides decreased the levels of gamma-glutamylated ketogenic amino acids in the colon and serum and suppressed the gamma-glutamyl transpeptidase (GGT) activity in the faeces (Olson et al. 2018); the inhibitors to GGT alone can suppress seizures. In addition, this study showed that the levels of the inhibitory neurotransmitter GABA as well as glutamate (which promotes the release of GABA) were increased in response to the treatment. The study also suggested that KD administration and microbial treatment protected the brain function during seizures by increasing the GABA levels in the central nervous system (Dooling and Costa-Mattioli 2018). Interestingly, A. muciniphila and Parabacteroides exerted no obvious effect when used individually (Olson et al. 2018), which suggested the irrefutable cooperative relationship between the two. However, children with cerebral palsy and epilepsy were found to show higher Akkermansia abundance than healthy children (Huang et al. 2019). In conclusion, the supplementation of A. muciniphila combined with Parabacteroides might be a potential treatment for epilepsy, although the beneficial role of Akkermansia in epilepsy remains debateable.

3.8. PD

PD is a progressive protein aggregation disease. Lewy bodies (LBs), which are composed of misfolded α-synuclein and other proteins, is the key neuropathological marker in PD (Luna and Luk 2015). Of note, LBs are present not only in the central nervous system but also in the enteric nervous system throughout the GI tract, which indicates the potential role of the intestinal microbiota in PD (Wakabayashi et al. 1990).

Recently, findings from 26 studies on PD, 15 of which reported an increase in the abundance of A. muciniphila in patients with PD, were summarized in different systematic analyses (See Parkinson’s Disease section in Table 1). These studies were conducted in different countries, including the USA, Germany, China, Finland, Russia, Italy, Australia, and Ireland, with 1743 patients with PD and 1340 healthy controls evaluated (Nishiwaki et al. 2020; Romano et al. 2021; Shen et al. 2021). Further, a study that used three machine learning algorithms to analyse metagenomic results from 472 patients with PD and 374 healthy controls identified 22 bacterial families that helped distinguish between controls and predicted patients with PD. Among the bacteria from the 22 families, Akkermansia was highly effective in distinguishing patients from controls and ranked fourth in terms of feature importance, although it had a low relative abundance (Pietrucci et al. 2020). These findings suggested the important role of Akkermansia in PD pathogenesis. Additionally, the increased abundance of Akkermansia could be considered a potential early biomarker for PD diagnosis.

Table 1. Akkermansia muciniphila and its association with different neuropsychiatric diseases.

Neuropsychiatric Disorders	Subjects and sample size	Changes in the microbiome of subjects compared with that in healthy controls	Changes in other organ systems	A. muciniphila intervention and its effect	Citation
Depression and anxiety	Infants aged 2.5 months (n = 446)	After maternal prenatal psychological distress: ↑ Proteobacteria ↓ Akkermansia	↑ levels of depression and anxiety (Score of chronic EPDS, PRAQ-R2, SCL, and Daily Hassles negative subscale)	NA	(Aatsinki et al. 2020)
	Male C57BL/6J mice, n = 20; wild-type mice treated with minocycline, n = 27.	After genetic deficiency or pharmacological inhibition of caspase-1: ↑Akkermansia ↑ Blautia spp. ↑ Lachnospiracea	After genetic deficiency or pharmacological inhibition of caspase-1: Ameliorated stress-induced depressive-like behaviour, increased locomotor activity and skills	NA	(Wong et al. 2016)
	Male C57BL/6J mice n = 69 administered prebiotics FOS, GOS, FOS + GOS.	GOS and FOS + GOS treatment: ↑ Lachnospiraceae ↑Akkermansia ↑ Oscillibacter ↓ Lactobacillus ↓ Bifidobacterium	GOS and FOS + GOS treatment: ↓ Stress-induced corticosterone release	NA	(Burokas et al. 2017)
	Male C57BL/6J mice Social defeat, n = 20; Control, n = 19.	Group with social defeat: ↓Overall diversity ↓Akkermansia ↑ Actinobacteria ↓ Deferribacteres ↓ Proteobacteria	NA	NA	(Mcgaughey et al. 2019)
	Male Wistar rats paradoxical sleep deprivation group, n = 10; control group, n = 10.	Paradoxical sleep deprivation group: ↓Akkermansia ↑ Oscillospira ↑Parabacteroides ↑Ruminococcus ↑Phascolarctobacterium ↑Aggregatibacter	After paradoxical sleep deprivation: Hyper-activated hypothalamic-pituitary-adrenal (HPA) axis.	NA	(Zhang et al. 2019a)
	Male Wistar rats ACTH-treated TRD model group, n = 10; Control group, n = 10.	ATCH-treated group: ↑ Ruminococcus ↑ Klebsiella ↓ Lactobacillus ↓Akkermansia	ATCH-treated group: ↓ Peripheral serotonin ↑ HPA axis activation.	NA	(Song et al. 2019)
	Male Sprague-Dawley rats (n = 43, 6 weeks old).	After feeding with olive oil-rich diet and induction of chronic mild stress: ↑Akkermansia ↑ Romboutsia	Chronic mild stress: ↓ weight gain during growing period ↑ Anxiety-like and anxiety-like behaviour ↑ Corticosterone Not relieved by olive oil.	NA	(Tung et al. 2019)
	Control group, n = 6; Chronic restraint stress (CRS) group, n = 6; CRS + AKK, group n = 6; AKK group, n = 6.	NA	NA	Treatment with AKK: Amelioration of depressive-like behaviour; Restoration of abnormal variations in corticosterone, dopamine, serotonin, and brain-derived neurotrophic factor.	(Ding et al. 2021)
Alzheimer’s Disease and cognitive deficit	Patients with mild cognitive impairment, n = 11; Mild cognitively normal participants, n = 6. With administration of a modified Mediterranean-ketogenic diet (MMKD).	After MMKD for 6 weeks: ↑Akkermansia ↑ Enterobacteriaceae ↑ Slackia ↑ Christensenellaceae ↑ Erysipelotriaceae ↓ Bifidobacterium ↓ Lachnobacterium (compared to microbiome before MMKD)	After MMKD for 6 weeks: Improved AD biomarker profile and brain vascular functions	NA	(Nagpal et al. 2019)
	Wild-type mice at 1, 2, 3, 6, and 9 months, n = 14,17, 17, 31, and 18, respectively; APP/PS mice at 1, 2, 3, 6, and 9 months, n = 21, 24, 24, 34, and 18, respectively.	Shifted with age alpha and beta diversity Ratio of Firmicutes/Bacteroidetes ↑ Enterobacteriaceae 1, 6 m ↑ Verrucomicrobiaceae 2, 6, 9 m ↑ Actinobacteria 2, 9 m ↑ Proteobacteria 2, 9 m ↑ Prevotellaceae 2 m ↑ Erysipelotrichaceae 2 m ↑ Bifidobacteriaceae 2 m ↓ Bacteroidaceae 6 m ↓ Rikenellaceae 6 m ↑ Desulfovibrionaceae 9 m ↑ Clostridiales 9 m	Detection by Aβ staining at 3 months; Detection of Aβ plaques at 6–9 months; Increased microglial activation at 6–9 months.	NA	(Chen et al. 2020)
	C57BL/6 male mice Ketogenic diet group (KD), n = 10; control regime group, n = 10.	After KD: ↑Akkermansia ↑ Lactobacillus ↓ Clostridiales ↓ Clostridium ↓ Dorea	After KD: Enhanced neurovascular functions, modulated blood ketone and glucose, ↓ Bodyweight	NA	(Ma et al. 2018)
	Juvenile C57BL/6J male mice fed a high-fat diet (HFD).	After HDF in early life: ↑ Firmicutes ↓ Bacteroidetes ↑↑ A. muciniphila	After HDF in early life: HFD impaired hippocampus-dependent contextual/spatial learning and memory.	A. muciniphila treatment: ↓ Hippocampal microgliosis ↑ Gut permeability ↓ Immune response ↑ Synapse plasticity ↑ Neuronal development ↓ Defects in learning and memory	(Yang et al. 2019)
	APP/PS1 mice: four groups (n = 10 per group) fed either a normal chow diet or HFD; Wild-type mice: two groups (n = 6 per group).	NA	NA	A. muciniphila treatment: ↑ Glucose homeostasis ↓ Damage to the intestinal barrier ↓ Hyperlipidemia ↓ Whitening of brown adipose tissue ↓ Deposits of Aβ plaque and increased Aβ levels in the brain Improved impaired cognitive, anxiety-related behaviours.	(Ou et al. 2020)
	Eight-week-old female C57BL/6 mice HFD-PBS, n = 10; HFD-AKK^sub, n = 10; NCD-PBS, n = 10; NCD-AKK^sub, n = 10.	NA	NA	A. muciniphila treatment of HFD-fed mice: ↓ Body weight and food consumption; ↑ Blood glucose control; ↑ Memory decay; ↓ Systemic inflammation; ↑ Tryptophan metabolism; ↑ Nissl bodies in neurons of the hippocampus; ↓ Liver steatosis	(Wu et al. 2020b)
	Normal chow (NC) NC + enriched environment (EE) High-fat, high cholesterol (HFHC) HFHC + EE, n = 8; NC HFHC + PBS HFHC + LGG (Lacticaseibacillus rhamnosus GG) HFHC + AKK (A. muciniphila).	HFHC vs NC ↓ Lactobacillaceae ↓ Ruminococ acceae ↑ Enterobacteriaceae ↑ Bacteroidaceae ↑ Peptostreptococ caceae Unique in NC + EE ↑ Micrococcaceae ↑ Christensenellaceae ↑ Ruminococcaceae Unique in HFHC + EE ↑ Firmicutes phyla ↑ Akkermansiaceae	EE restored HFHC-induced cognitive deficits, but further decreased brain metabolic activity	A. muciniphila restored HFHC-induced cognitive deficits and brain metabolic activity.	(Higarza et al. 2021)
Substance use disorders	Patients with AUD, n = 36; Controls, n = 36.	Patients with AUD: ↓Akkermansia ↑ Bacteroides	Patients with AUD showed increased serum levels of lipopolysaccharide (LPS), TNF-α, IL-1β, MCP-1, and IL-6.	NA	(Addolorato et al. 2020)
	Patients with alcoholic steatohepatitis (ASH), n = 21; Patients with severe ASH, n = 15; Healthy individuals without obesity, n = 16.	Patients with ASH compared with healthy individuals: ↓ A. muciniphila	NA	A. muciniphila treatment: ↓ Hepatic injury and established alcoholic liver disease	(Grander et al. 2018)
	Wild-type C57BL/6 6-to 8-week-old female mice	After alcohol induction: ↓Akkermansia ↓ Verrucomicrobiaceae ↓ Lachnospiraceae ↑ Eubacteriaceae	NA	NA	(Lowe et al. 2017)
	Male Sprague–Dawley rats: MA group, n = 17; Saline group, n = 5.	MA exposure: ↑ A. muciniphila ↑ Acetivibrio	NA	NA	(Yang et al. 2020)
Amyotrophic Lateral Sclerosis (ALS)	Antibiotic-treated mSod1G93A transgenic mice.	mSod1 ALS mice (before or after antibiotic treatment): ↓ A. muciniphila, Parabacteroides with diseases progression	NA	A. muciniphila treatment: ↓ ALS in mice by increasing NAM levels ↑ Survival of ALS mice (improved motor function)	(Blacher et al. 2019)
Multiple Sclerosis (MS)	MS mice, n = 60; healthy controls, n = 43.	Experimental autoimmune encephalomyelitis (EAE) mice: ↑ Methanobrevibacter ↑Akkermansia ↓ Butyricimonas	↑ Interferon, NF-κB, Toll-like receptor, and IL-6 signalling pathways; ↓ PPARα/RXRα.	NA	(Jangi et al. 2016)
	34 Monozygotic twin pairs discordant for MS	Untreated MS twins: ↑Akkermansia	NA	NA	(Berer et al. 2017)
	Germ-free mice with microbiome transplant twins with MS, n = 22–26, or healthy donors, n = 20–26.	Mice with microbiome transplant from twins with MS: ↓ Sutterella	Mice with microbiome transplant from twins with MS: ↑ Autoimmune ↓ IL-10
	MS patients not undergoing treatment, n = 71; healthy controls, n = 71.	MS patients: ↑ Acinetobacter ↑ A. muciniphila ↓ Parabacteroides	NA	NA	(Cekanaviciute et al. 2017)
	GF C57BL/6 mice Three MS and control donor pairs; n = 6–8 mice per group.	MS mice model: ↓ Sutterella ↑ Ruminococcus	↓ IL-10 + Treg induction in mesenteric lymph nodes from MS microbiome-colonized mice
	Patients with MS, n = 9; controls, n = 13 orally administered a probiotic.	Patients with MS: ↑Akkermansia ↑ Blautia ↑ Dorea	NA	NA	(Tankou et al. 2018)
	Female C57BL/6 mice Naive mice, n = 4; EAE-VEH, mice n = 5; EAE-(delta-9-tetrahydrocannabinol (THC)+ cannabidiol (CBD), n = 6.	MS mice model treated with THC + CBD: ↓ A. muciniphila ↓ Tenericutes ↑ Firmicutes	MS mice model treated with THC + CBD: ↑ Anti-inflammatory response ↑ Levels of LPS	NA	(Al-Ghezi et al. 2019)
	Relapsing remitting MS (RRMS), n = 135; Progressive MS mice, n = 31; Healthy controls, n = 38.	In both progressive and RRMS vs. healthy controls ↓ Blautia wexlerae ↓ Dorea formicigenerans ↑ Romboutsia timonensis ↑ Bacteroides Unique to RRMS ↓ Erysipelotrichaceae CCM ↓ Ruminococcaceae ↑ Bacteroides uniformis ↑ Pseudoflavonifractor Unique to progressive MS ↑Akkermansia ↑ Streptococcus ↓ Blautia ↓ Agathobaculum	Several Clostridium species in progressive MS (n = 27) and C. bolteae, C. leptum, and C. scindens in RRMS (n = 95) were associated with higher EDSS and fatigue scores. Akkermansia abundance was negatively correlated with EDSS and MRI measures of disease severity in patients with progressive disease.	MS-derived Akkermansia strains BWH-J5, BWH-H3, and BWH-I7 colonized C57/BL6 mice, lowered disease in the MOG/C57 model of EAE, with the strongest protective effect exerted by strain H3, which also reduced RORγT-positive γδ T cells and IL-17-producing γδ T cells.	(Cox et al. 2021)
Autism Spectrum Disorder (ASD)	Children with typically developed ASD, n = 23; Siblings of children from the ASD cohort, n = 22; Unrelated controls, n = 9.	Children with ASD: ↓ Bifidobacterium spp. ↓ A. muciniphila	NA	NA	(Wang et al. 2011)
	Healthy children, n = 10; Pervasive Developmental-Disorder Not Otherwise Specified, n = 10; autism, n = 10.	Children with ASD: ↓ Fusobacteria ↓ Verrucomicrobia	NA	NA	(De Angelis et al. 2013)
	Patients with autism with gastrointestinal symptoms, n = 33; Siblings not showing autistic symptoms, n = 7; Non-sibling control subjects, n = 8.	Patients with ASD: ↑ Desulfovibrio species ↑ Bacteroides vulgatus ↑ Verrucomicrobia	NA	NA	(Finegold 2008; Finegold et al. 2010)
	Infants with ASD, n = 6; Healthy infants, n = 6.	Patients with ASD: ↑ Faecalibacterium ↓ Blautia ↑Akkermansia	NA	NA	(Inoue et al. 2016)
	Individuals with ASD, n = 20; healthy subjects, n = 28.	Patients with ASD: ↑ Verrucomicrobia ↑ Firmicutes ↓ Proteobacteria ↓ Cyanobacteria	NA	NA	(Lee et al. 2017)
	Children with ASD, n = 20; neurotypical children, n = 20.	Children with ASD: ↑ Prevotella ↓Akkermansia	NA	NA	(Kang et al. 2013b)
	Individuals with autism, n = 40; Neurotypical controls, n = 40.	Patients with ASD: ↓Akkermansia ↓ Alistipes ↓ Bilophila, ↓ Dialister ↓ Parabacteroides ↑ Collinsella ↑ Corynebacterium ↑ Dorea ↑ Lactobacillus	NA	NA	(Strati et al. 2017)
	2-to-7-year-old children with ASD (n = 48).	↑Bacteroidetes/Firmicutes ratio ↓Akkermansia ↓ Dialister invisus	NA	NA	(Zou et al. 2020)
	C57BL/6 (B6) mice n = 21; BTBR^T + tf/j mice n = 25.	BTBR^- mice: ↓Akkermansia	↑ Abdominal pain; ↑ Inflammation; ↓ Gastric motility	NA	(Newell et al. 2016)
Epileptic Seizures	Healthy children, n = 20; Patients with cerebral palsy (CP) and epilepsy, n = 25.	Patients with CP: ↑ Bifidobacterium, ↑ Streptococcus, ↑Akkermansia, ↑ Enterococcus ↑ Prevotella ↑ Veillonella ↑ Rothia ↑ Clostridium IV ↓ Bacteroides, ↓Faecalibacterium, ↓Blautia ↓Ruminococcus, ↓Roseburia, ↓Anaerostipes ↓Parasutterella	NA	NA	(Huang et al. 2019)
	Kcna1^−/− SPF C3HeB/FeJ mice fed ketogenic diet (KD) or standard chow.	After KD: ↑Akkermansia ↑ Parabacteroides ↑ Sutterella ↑ Erysipelotrichaceae	After treatment with Akkermansia and Parabacteroides: ↓ Systemic gamma-aminobutyric acid (GABA) ↑ Hippocampal GABA ↑Hippocampal glutamate	Treatment with both Akkermansia and Parabacteroides: ↑ Acute electrically-induced seizure protection; Seizure protection was not observed with A. muciniphila alone	(Olson et al. 2018)
Parkinson’s Disease (PD)	Male patients with PD, i = 31; Male healthy controls, n = 28.	Patients with PD: ↑Akkermansia ↑ Firmicutes ↓ Prevotella copri ↓ Eubacterium biforme	NA	NA	(Bedarf et al. 2017)
	Patients with PD, n = 34; Healthy controls, n = 34.	Patients with PD: ↑Akkermansia ↑ Bifidobacterium ↑ Enterobacteriaceae ↓ Bacteroidetes ↓ Prevotellaceae ↓Lactobacillaceae	NA	NA	(Unger et al. 2016)
	Patients with PD, n = 72; Healthy control, n = 72.	Patients with PD: ↑ Lactobacillaceae ↑ Verrucomicrobiaceae ↑ Bradyrhizobiaceae ↑ Ruminococcaceae ↓ Prevotellaceae ↓ Clostridiales i.s. IV	NA	NA	(Scheperjans et al. 2015)
	Patients with PD, n = 38; Healthy controls, n = 34.	Patients with PD: ↑Akkermansia ↑ Ralstonia ↑Proteobacteria. ↑Bacteroidaceae, ↑Clostridiaceae, ↓ Blautia ↓ Coprococcus ↓ Roseburia ↓ Faecalibacterium ↓ Lachnospiraceae	NA	NA	(Keshavarzian et al. 2015)
	Patients with PD, n = 80; Healthy controls, n = 72.	Patients with PD: ↑Verrucomicrobiaceae ↑Lactobacillaceae ↑Enterobacteriaceae ↑Enterococcaceae ↓Lachnospiraceae	Patients with PD: ↑ Fatty acid and butanoate metabolism ↓Amino acid metabolism	NA	(Pietrucci et al. 2019)
	Patients with PD, n = 193; Patients with multiple system atrophy, n = 22; Patients with progressive supra-nuclear palsy, n = 22; Healthy controls, n = 113.	Patients with PD: ↑Akkermansia ↑ Proteobacteria ↑ Bifidobacteriaceae ↑ Christensenellaceae ↑ Coriobacteriaceae ↑ Lactobacillaceae ↑ Enterobacteriaceae ↑ Lactobacillaceae ↓ Lachnospiraceae ↓Prevotellaceae ↓Streptococcaceae	NA	NA	(Barichella et al. 2019)
	Dataset 1: Patients with PD, n = 212; Controls, n = 136; Dataset 2: Patients with PD, n = 323; Controls, n = 184.	Patients with PD: ↑Akkermansia ↑ Bifidobacterium ↑ Lactobacillus ↑ Prevotella ↓ Faecalibacterium ↓ Roseburia	NA	NA	(Wallen et al. 2020)
	Typical PD, n = 147 cases; controls, n = 162.	Patients with PD: ↑Akkermansia ↑Lactobacillus ↓Turicibacter	NA	NA	(Baldini et al. 2020)
	Patients with PD, n = 14; Healthy controls, n = 7.	Patients with PD: ↑Akkermansia ↑Gammaproteobacter	NA	NA	(Gorecki et al. 2019)
	Patients with PD, n = 51; Healthy controls, n = 48.	Patients with PD: ↑Akkermansia ↑Ruminococcaceae ↓Lactobacillus ↓Streptococcus ↑Prevotellaceae	NA	NA	(Li et al. 2019a)
	Patients with PD, n = 75; Healthy controls, n = 45.	Patients with PD: Slight changes ↑Akkermansia ↑Bifidobacteriaceae ↑Eubacteriaceae, ↓Lachnospiraceae ↓Firmicutes ↓Tenericutes, ↓Euryarchaeota. ↓Streptococcaceae,	NA	NA	(Lin et al. 2018)
	Cohort 1: Patients with PD, n = 80; Healthy controls, n = 77. Cohort 2: Patients with PD, n = 120; Healthy controls, n = 120.	Patients with PD: ↑Verrucomicrobia ↑Lactobacillus ↑Odoribacteraceae ↓Prevotellaceae	A correlation between Bacteroides abundance and plasma levels of TNF-α, and a correlation between Verrucomicrobia abundance and plasma levels of IFN-γ was observed	NA	(Lin et al. 2019)
	Patients with PD, n = 197; Healthy controls, n = 130.	Patients with PD: ↑Akkermansia ↑Lactobacillus ↑Bifidobacterium ↑Christensenellaceae ↑Ruminococcaceae ↑Tissierellaceae ↓Lachnospiraceae ↓Pasteurellaceae	Patients with PD: ↑Acetyl-CoA synthetase ↓Butyrate kinase	NA	(Hill-Burns et al. 2017)
	Patients with PD, n = 76; Patients with idiopathic rapid eye movement sleep behaviour disorder, n = 21; Healthy controls, n = 78.	Patients with PD: ↑Akkermansia ↑Actinomycetales, ↑Flavonifractor ↓Prevotella	NA	NA	(Heintz-Buschart et al. 2018)
	Twenty-eight 6–8-week-old male C57BL/6 mice were randomized into four groups evenly. Controls (10 mg/kg/day body weight, 4% carboxymethylcellulose and 1.25% chloroform), Rotenone (10 mg/kg/day body weight), Restraint stress (RS), RS + rotenone.	RS: ↓Lactobacillus RS + rotenone (compared to RS): ↑Verrucomicrobiaceae ↑ A. muciniphila ↓Coriobacteriaceae	Rotenone, RS, and RS + rotenone: increased immune cell infiltration in the lamina propria	NA	(Dodiya et al. 2020)
	Control group, n = 13; MPTP-injected group, n = 14; MPTP-injected plus KRG-treated group, n = 14; Vehicle-injected plus KRG-treated group, n = 14.	After MPTP injection: ↑Verrucomicrobia ↑Firmicutes ↑Tenericutes ↓Bacteroidetes ↓Proteobacteria ↓A. muciniphila	After MPTP injection: ↑Motor dysfunction ↑Iba-1 expression ↑aSyn expression in the SN ↑IL-1β and TNF-α ↓Number of TH-positive cells in the SN and optical density in the striatum ↓Occludin expression	NA	(Jeon et al. 2021)
Stroke	Healthy controls, n = 200; Patients with acute cerebral infarction, n = 198.	Patients with stroke: ↓Akkermansia ↓Mucispirillum ↑Aerococcaceae ↑Flavobacterium	NA	NA	(Chang et al. 2021)
	Healthy controls, n = 30; Patients with cerebral ischaemic stroke, n = 30.	Patients with stroke: ↑Odoribacter ↑Akkermansia ↑Ruminococcaceae_UCG_005 ↓Enterobacter	NA	NA	(Li et al. 2019b)
	Post-stroke animals; Sham-operated animals.	Post-stroke animals: ↑Akkermansia ↑Parabacteroides ↑Bacteroides	NA	NA	(Stanley et al. 2018)

However, the abundance of Akkermansia did not increase in most mouse models of PD. In the same study, Guadalupe et al. showed a significant increase in the Akkermansia abundance in patients with PD, but not in parkinsonian A53T aSyn and MSA CNP-aSyn Tg mouse models (Vidal-Martinez et al. 2020). The abundance of Akkermansia did not increase in a 6-OHDA-induced PD mice model either (Hou et al. 2021). Studies conducted using the PD transgenic mice model established by inducing Thy1-human-α-syn, A53T, or CNP overexpression did not report an increase in the relative abundance of Akkermansia (Gorecki et al. 2019; Vidal-Martinez et al. 2020). A GLP-1 engineered strain protected mice from MPTP-induced motor deficit with increased Akkermansia abundance and decreased Enterococcus and Proteus abundances (Fang et al. 2020). However, only a few studies have reported an increase in the abundance of Akkermansia, especially A. muciniphila, in the PD mice model. A mouse model of MPTP-induced PD was found to exhibit increased Akkermansia abundance, and treatment with Korean red ginseng could suppress the high A. muciniphila abundance and inhibit dopaminergic neuronal death and inflammation in the substantia nigra and colon (Jeon et al. 2021) Hemraj et al showed that the Akkermansia abundance increased in a rotenone-treated mice model (Dodiya et al. 2020). The inconsistent alterations observed between patients in clinical settings and PD mice models might be attributed to the variations in the different species or genus. However, the following issues remain unaddressed: (1) whether animal models can mimic the human disease; (2) whether the mouse gut microbiome can be evaluated to predict PD occurrence in humans; (3) whether several taxa, rather than a single taxon, collectively contribute to PD; (4) the likelihood of a compensatory beneficial effect exerted by A. muciniphila in PD. Extensive investigations are needed to answer these questions.

3.9. Stroke

Emerging evidence indicates a typical intestinal microbiome signature in patients after stroke. In a study conducted in 2020, A. muciniphila was proposed as a gut microbial marker for post-ischaemic stroke (Xiang et al. 2020). Further investigations showed a considerable decrease in the abundance of Akkermansia in 198 patients with acute cerebral infarction who suffered from ischaemic stroke (Chang et al. 2021). Furthermore, Puerariae Lobatae Radix and Chuanxiong Rhizoma, which are commonly used to treat cerebrovascular diseases, significantly improved the intestinal colonization of Akkermansia (Chen et al. 2019). However, previous studies have shown that Akkermansia is enriched in the gut of patients with cerebral ischaemic stroke (Li et al. 2019b) and animals with stroke (Stanley et al. 2018). As the results are inconsistent, the effect of the probiotic strain A. muciniphila in patients with stroke remains elusive. Further research is needed to confirm the role of A. muciniphila in patients with stroke.

4. Dietary interventions, drugs, and lifestyle factors that affect the relative abundance of A. muciniphila as well as brain functions

Dietary interventions or drug treatments could alter the relative abundance of A. muciniphila; meanwhile, these interventions also significantly affect brain functions, such as learning, memory, and social interaction. Metformin, the primary treatment in clinical practice for metabolic disorders, such as obesity and type 2 diabetes, increased the abundance of Akkermansia both in vitro and in vivo (Shin et al. 2014; De La Cuesta-Zuluaga et al. 2017; Ji et al. 2019). Meanwhile, several studies have shown that metformin considerably improves the spatial learning and memory deficit in animal models of AD (Ou et al. 2018; Farr et al. 2019; Lu et al. 2020; Syal et al. 2020; Wang et al. 2020b; Wu et al. 2020a). A. muciniphila was found to be one of the most abundant microorganisms after vancomycin treatment (Hansen et al. 2012; Basolo et al. 2020), and vancomycin was shown to drastically improve autism symptoms related to behaviour, cognition, and GI function. However, only one case report (Rodakis 2015), one small open-label study (Finegold 2011), and one clinical trial on 11 children with regressive-onset autism (Sandler et al. 2000) showed the beneficial effects of vancomycin in autism. Additionally, other vancomycin-resistant bacteria may also contribute to the improvement (Finegold 2011). Interestingly, the abundance of A. muciniphila increased significantly upon starvation or underfeeding (Basolo et al. 2020), whereas a KD, which mimics starvation, provided protection against cognitive decline in several neuropsychiatric diseases, such as AD (Ma et al. 2018; Nagpal et al. 2019; Neth et al., 2020) and ASD (Newell et al. 2016). Certain foods, such as 2% fish oil, 2% soybean oil (Jackson et al. 2012; Tung et al. 2019), blackberry leaf and fruit extracts (Meireles et al. 2016; Park et al. 2019), and cranberries (Anhê et al. 2017; Shukla et al. 2018) increased the A. muciniphila abundance, and fish oil treatment improved the symptoms of psychosis and anxiety-/depression-like behaviour. These studies provided subtle yet irrefutable evidence on the connection between A. muciniphila and brain functions.

Numerous lifestyle factors have been recognized to play an important role in positively modifying neuropsychiatric disorders by altering the abundance of A. muciniphila. These include the intake of healthy food, increased physical activity, and avoidance of smoking and intake of alcohol and illicit drugs. The influence of HFD has been studied widely, and HFD intake has been shown to be negatively correlated with the abundance of A. muciniphila in mice (Everard et al. 2013; Cox et al. 2014; Nobel et al. 2015), rats (Carmody et al. 2015; Fåk et al. 2015), and humans (Karlsson et al. 2012; Teixeira et al. 2013). Physical activity was also shown to influence the increase in the abundance of Akkermansia in mice (Liu et al. 2017) and women (Bressa et al. 2017). A 75-day lifestyle intervention, involving the intake of a low-saturated-fat diet, low energy intake with functional foods, and physical activity, helped increase the abundance of A. muciniphila in individuals with metabolic syndrome (Guevara-Cruz et al. 2019).

5. Association of Akkermansia with healthy ageing and longevity

Emerging evidence strongly suggests a link between the gut microbiome and ageing (Bana and Cabreiro 2019). A systematic analysis of 27 empirical human studies and 4 articles showed that Akkermansia was positively correlated with individual lifespan and health (Badal et al. 2020). Although A. muciniphila levels were decreased in forty-five 80—82 years old elderly subjects compared to fifty-five 25—35 year old healthy adults (Collado et al. 2007), a substantial increase in Akkermansia levels were consistently reported in centenarians from different regions. In twenty-four 105–109 year old semi-supercentenarians from Italy, two well-known health-associated genera, Akkermansia and Bifidobacterium, were boosted compared to fifteen adults, elderly, and centenarians, respectively (Biagi et al. 2016). In a Chinese study, Akkermansia levels were significantly higher in sixty-seven ≥90 healthy long-lived elderly, compared to fifty-four elderly and forty-seven young adults (Kong et al. 2016). A study from South Korea compared nine adults and thirteen elderly living in the same longevity village and found that Akkermansia levels were significantly increased in twenty-five centenarians with a similar sufficient and diverse food intake (Kim et al. 2019). Further evidence indicates the beneficial effects of Akkermansia in the maintenance of health in ageing and extending lifespan. The abundance of Akkermansia was three-fold greater in individuals showing healthy ageing than in individuals showing non-healthy ageing (Singh et al. 2019), and was also positively and independently correlated with better sleep quality and Stroop performance in the healthy elderly (Anderson et al. 2017; Manderino et al. 2017).

The correlation was notable in an animal experiment where fecal microbiome transplantation (FMT) from healthy donors or oral gavage of A. muciniphila was found to improve health and extend lifespan in two Hutchinson–Gilford progeria syndrome mouse models (Bárcena et al. 2019). Furthermore, A. muciniphila is capable of ameliorating the age-related decline in colonic mucosal thickness and attenuating immune activation (Van Der Lugt et al. 2019); also, ageing itself alters the gut microbiome, which involves a reduction in Akkermansia abundance, thereby contributing to systematic inflammation (Fransen et al. 2017). Two recent studies from different groups reported that oral administration of Akkermansia for 9 months or even one month could significantly promote healthy ageing, by influencing behaviour response, anxiety-like behaviours, immune functions, redox state, cognitive function, muscle atrophy, and lifespan in ageing mice; indicating a potential strategy to promote healthy longevity (Cerro et al. 2021; Shin et al. 2021).

In summary, these results suggest that Akkermansia could be a signature of longevity and healthy ageing by playing an important role in health maintenance during ageing and possibly extending longevity. In addition to the beneficial effects exerted by Akkermansia in the metabolic system, immune system, and brain function, the health- and longevity-promoting effects of this microorganism is expected to garner considerable attention in future studies (Figure 1).

Figure 1. The primary mechanism of action of Akkermansia muciniphila in the gut-brain axis. Besides the peripheral circulation system, the vagus nerve is an important structure involved in the bidirectional communication between the gut and the brain, as it is distributed in the major parts of the intestines, with 80% as afferent nerves and 20% as efferent nerves (Goswami et al. 2018). Mechanism of action in the (A): mucous membrane and (B): immune system. (C) and (D) show the relationship between A. muciniphila metabolites and brain functions: (C) mechanism of action of SCFAs: SCFAs produced by A. muciniphila and their transport pathway. a. Passive diffusion; b. MCT1pathway: through the MCT receptor coupled with H⁺ ions, one molecule of SCFA can be transported simultaneously by the delivery of one H⁺ ion into intestinal cells at a time; c. SMCT1 pathway: through the SMCT receptor coupled with Na⁺, two molecules of Na⁺ and one molecule of SCFA can be transported into intestinal cells simultaneously; d. Exchange with HCO₃^– through an unknown exchanger, followed by partial oxidation to carbon dioxide for additional cellular energy generation in the form of ATP. (D) Mechanism of action of amino acid derivatives. SCFA, short-chain fatty acid; HCO₃^–, bicarbonate; MCT1, monocarboxylate transporter 1; SMCT1, sodium-dependent monocarboxylate transporter 1; TER, transepithelial resistance; HPA, hypothalamic-pituitary-adrenal axis; HDAC, histone deacetylase; BBB, blood-brain barrier; BDNF, brain-derived neurotrophic factor; 5-HT, 5-hydroxytryptamine, GABA, gamma-aminobutyric acid.

6. Final remarks

Since its identification in 2004, A. muciniphila has been shown to play an essential role in metabolic and immune systems. It is one of the most well-studied gut bacterial species and is considered a “next-generation probiotic” for metabolic diseases, such as diabetes and obesity, and the clinical outcomes of cancers. Furthermore, the safety and tolerability of live A. muciniphila in clinical applications has been confirmed, indicating its prospects as a therapeutic agent. Accumulating evidence shows the correlation between A. muciniphila abundance and several neuropsychiatric disorders, as shown in Figure 2. Although the exact mechanisms remain unclear, A. muciniphila is thought to modulate the gut-brain axis primarily by exerting its protective effects on the intestinal mucosal barrier, immune system, and metabolic system, as well as through its metabolites, such as SCFAs and amino acid and derivatives. The application of A. muciniphila in preclinical animal models of neuropsychiatric disorders highlighted its impact on immune and metabolic responses systemically and in the central nervous system.

Figure 2. Types of diseases associated with the abundance of Akkermansia muciniphila: This figure summarizes the diseases associated with the abundance of A. muciniphila and the gut-brain axis, as discussed in this review. (A. muc: A. muciniphila;↑: the relative abundance of A. muciniphila increased in patients with the disease compared to that in the control;↓: The relative abundance of A. muciniphila decreased in patients with the disease compared to that in the control;? : limited evidence is available.↑↓: inconsistent results from studies published to date.).

According to the evidence summarized here, the exploration of the probiotic role of A. muciniphila in the gut-brain axis and its therapeutic potential in neuropsychiatric disorders, especially in depression and anxiety, AD, and cognitive impairment, is an attractive topic for future research. For various neuropsychiatric disorders, characterized by an undetectable onset and gradual progression of symptoms, early prevention is the most important step in disease management. Long-term A. muciniphila administration before the onset of symptoms, which would help moderately modulate the gut microbiome homeostasis in individuals at a higher risk of disease development, could be beneficial for preventing disease onset or progression. Additionally, A. muciniphila modulates the metabolic and immune systems and promotes overall health and longevity, which further strengthen its beneficial effects in the treatment of neuropsychiatric disorders, via comprehensive and integrity mechanisms. However, to date, most studies conducted are either based on animal models or are correlation studies, and comprehensive pre-clinical analysis and well-designed clinical studies are necessary to obtain direct and compelling evidence of the effects and interventional efficiency of A. muciniphila in neuropsychiatric disorders.

Author contributions

XZ designs the overall idea. XZ, RX, YZ, SC and TC performed the review of the literature and wrote the manuscript; YZ, XF and SL participated in reviewing the manuscript.

Acknowledgment

We are grateful to the Mental Health Institute of Central South University for technical assistance.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Hunan Provincial Natural Science Foundation (2021JJ40894), Changsha Natural Science Foundation Project (kq2014239) and Scientific Research Project of Hunan Provincial Health Commission 202103090528 for X.Z., Scientific Research Project of Hunan Provincial Health Commission 202211004855 for T.C. Central South University Undergraduate Research and Innovation (URI) Program S2020105330520 for R.X.