The Endocannabinoids-Microbiota Partnership in Gut-Brain Axis Homeostasis: Implications for Autism Spectrum Disorders

1.2 Gut Microbial Composition and eCB Signaling: The Powerful Role of Dietary Fat Composition

It is widely recognized that the large gut microbial population, generally named microbiota, is composed of about 100 trillions of microorganisms mainly including anaerobic bacteria, fungi and protozoa (Ley et al., 2006; Matijašić et al., 2020). Firmicutes, Bacteroidetes, Proteobacteria and Actinobacteria are the most copious and representative bacterial species, living in symbiotic “harmony” with the host and significantly contributing to its metabolic and immune functions as well as to the preservation of integrity of the intestinal epithelium (Burcelin, 2016). Such diversity and sensitive symbiotic ecosystem can be selectively perturbed by different lifestyles and nutritional challenges. The two main abundant phyla Firmicutes and Bacteroidetes are also the most susceptible to the shaping action of low- vs. high-fat diet, high-sugar diet and ingestion of different fatty acids, as showed by the incidence of dysbiosis and the higher representation of Firmicutes phyla (e.g., Lactobacillaceae, Veillonellaceae, and Lachnospiraceae) in obesity (Crovesy et al., 2020).

Since eCBs and NAEs are bioactive lipids, their cell and tissue availability can be directly changed by dietary habits and, in particular, by dietary fatty acids composition (Chianese et al., 2017). Indeed, while the n-6 carbon essential fatty acid linoleic acid (LA, 18:2, n-6) is the precursor for ARA biosynthesis (and then for AEA and 2-AG production), the n-3 carbon essential fatty acid α-linolenic acid (ALA, 18:3, n-3) is the precursor of the main n-3 PUFAs docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA) from which active metabolites such as N-docosahexaenoyl-ethanolamine (DHEA) and N-eicosapentaenoyl-ethanolamine (EPEA) are also produced (Berger et al., 2001; Maccarrone et al., 2010a). It is well-established that the n-6/n-3 ratio has been dramatically altered by the worldwide adoption of the Western diet (WD) lifestyle especially in the last 50 years, shifting from about 5:1 to the currently estimated 20:1 ratio (Simopoulos, 2006; Blasbalg et al., 2011). A general scheme of the gut-brain interface, including the major pathways of communication involved, the key components of the eCB system and the influence of dietary factors is showed in Figure 1.

FIGURE 1

The figure schematizes the bi-directional and reciprocal influence between brain and gut (i.e., the gut-brain axis) along the main routes of communication such as the HPA axis, the autonomic nervous system, the vagus nerve and endocrine signaling. The ability to sense microbiota metabolites, gut peptides and neuroactive compounds and to transfer to the brain this complex information about the ongoing status of gut microbial ecosystem is fundamental for the generation of adaptive or maladaptive responses, for instance by the HPA axis and cortisol secretion. Here, the main components of the eCB system machinery are depicted at both presynaptic and postsynaptic terminals as well as in microglia cells, and the attention is focused on the role of powerful external modifiers such as dietary composition to mutually affect gut microbiota composition and eCB signaling. created in moc.redneRoiB.

Increasing the intake of LA might alter the eCBs levels (Berger et al., 2001; Alvheim et al., 2013; Alvheim et al., 2014), and CB₁ receptor function in selected brain areas (prefrontal cortex and nucleus accumbens) can be permanently altered by feeding animals with a n-3 deficient diet during gestation and lactation (Lafourcade et al., 2011). Moreover, there is evidence that the beneficial impact on cognition, or the attenuation of cognitive decline provided by the decrease of dietary n-6/n-3 ratio via n-3 PUFAs dietary supplementation can be preserved both during adolescence and adulthood (Muldoon et al., 2010; Luchtman and Song, 2013), although negative results have also been reported (Stough et al., 2012). The anti-neuroinflammatory and anti-depressive potential of DHA and EPA has been studied in the recent literature in both animal models (Minogue et al., 2007; Moranis et al., 2012) and clinical reports (Rees et al., 2006; Liao et al., 2019), along with their use in dietary supplementation as nutritional guideline to slow down age-associated cognitive decline through preservation of neurogenesis and synaptic plasticity (Dyall, 2015). The levels of dietary LA can affect brain content of 2-AG and AEA and increase liability to obesity (Massiera et al., 2010; Alvheim et al., 2012) but, on the other hand, the different composition of dietary fatty acids can be exploited as anti-obesity agent by both reduction of 2-AG levels (Banni et al., 2011) and the modulation of NAEs concentrations (Pu et al., 2016). Most importantly, the composition of dietary n-3 and n-6 PUFAs and therefore ALA or LA is able to shape the huge microbial community composing the gut microbiota (Marrone and Coccurello, 2019; Wolters et al., 2019). Gut dysbiosis, low-grade inflammation and gut permeability are indeed leading causes for obesity development (Levy et al., 2017; Sun et al., 2018). In obesity, the systemic elevation of the eCB tone (brain included) as well as the upregulation of the CB₁ receptor in adipose tissue, liver and muscle have been extensively demonstrated (Matias and Di Marzo, 2006; Starowicz et al., 2008; Silvestri and Di Marzo, 2013). Accordingly, blockade of CB₁ receptors may revert both systemic inflammation and dysbiosis induced in obese mice by chronic high-fat diet (HFD) regimen (Mehrpouya-Bahrami et al., 2017).

In support of the idea of a mutual interconnection between endocannabinoids and microbiota, especially in the regulation of energy metabolism (Cani et al., 2016), it has been pointed out that human microbiota can encode N-acyl amides (i.e., eCBs congeners) that can bind G-protein-coupled receptors (GPCRs), thus modulating the intestinal function by the way of the GPR119 (Cohen et al., 2017). Dietary habits and composition are the most important modifiers of the intestinal microbial community (Rinninella et al., 2019), and the eCB machinery is high susceptible to changes in food composition and dietary patterns (Bisogno and Maccarrone, 2014; Passani and Coccurello, 2016; Chianese et al., 2017). In turn, food-induced changes in the eCB system can be paralleled by changes in gut microbiota composition. Switching for only 3 days to a high-fat/high-sucrose diet can produce in the small intestine an increase of AEA and 2-AG together with changes in the abundance of selected bacteria that were found to correlate to blood levels of eCB mediators such as AEA and DHEA (Lacroix et al., 2019). Both eCBs and NAEs plasma levels are directly modulated by nutrients and diet composition, as demonstrated by the switch to a Mediterranean dietary regimen in obese or overweight subjects (Tagliamonte et al., 2021) that reduced AEA levels while increased OEA/PEA and OEA/AEA ratios. Interestingly, the shift from Western-to-Mediterranean diet also produced an increase of faecal content of the mucin-degrading Gram-negative bacterium of the phylum Verrucomicrobia Akkermansia muciniphila (A. muciniphila), whose abundance appears inversely associated with gut dysbiosis and gut permeability (Lopetuso et al., 2020) and therefore is associated to healthy microbial ecology. As “gatekeeper” of intestinal mucosa and gut health, the abundance of A. muciniphila can be modulated by diet and nutrients as demonstrated by fructo-oligosaccharides prebiotic supplementation (Everard et al., 2013). However, not only the fluctuations in A. muciniphila gut abundance are paradigmatic of the tight relationship between diet, gut microbial ecology and (as afterwards discussed) eCB-like lipid mediators (Depommier et al., 2021), but also of the alterations found in gut microbiota of subjects with autism spectrum disorder (ASD). The direct, as well as indirect, impact of macronutrients (mostly fats) on the eCB system of the gastrointestinal (GI) tract and gut-brain signaling have been recently and comprehensively analyzed, especially for the implication in metabolic disease (Di Patrizio, 2021). In metabolic disease, but also in other clinical conditions, there is evidence that the alteration of selected microbial populations and the administration of eCB or eCB-like lipids may help to rebalance gut bacterial composition. As for instance, vitamin D deficiency was shown to exacerbate neuropathic pain, reducing both gut microbial diversity and 2-AG colon levels, while the supplementation of vitamin D and PEA was shown to attenuate allodynia and rebalance the Firmicutes/Bacteroidetes ratio, including the increase of gut Akkermansia species (Guida et al., 2020).

In this context, a novel route for a future investigation of the relationship between ASD, microbiota and eCB system might also be identified in the analysis of the eating problems/disorders reported in ASD subjects. Indeed, unhealthy eating habits are observed in subjects with ASD in which the incidence of eating disorders (EDs) is much greater than in the general population (Sharp et al., 2013). The higher incidence of EDs and alterations of eating behavior and food selection in ASD patients (Råstam et al., 2013; Sharp et al., 2013; Nickel et al., 2019), suggest to investigate whether in addition to cognitive endophenotypes (Zucker et al., 2007) there might be common gut microbial signatures in EDs and autism along with parallel changes in eCB signaling. A surprising variety of different eating-associated abnormalities afflict children and adolescents who have received a diagnosis of autism. The eating and feeding disorders described span from the more “canonical” EDs (i.e., AN, BN, and BED) to rumination disorder, picky eating, avoidant/restrictive food disorder (ARFID) as well as Pica (i.e., insistent eating of non-food substances) (American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), 2012; American Psychiatric Association, 2013; Fields et al., 2021; Inoue et al., 2021). For this reason, these aspects are suggested but not further developed in the present work, hoping that other studies will address in the next future the fascinating issue of multiple dietary approach in ASD (e.g., gluten-free diet or ketogenic diet) and parallel impact on eCB system/gut microbiota interplay.

2.3 ASD and the eCB Signaling-Gut Microbiota Reciprocal Influence

If ASD is a NDD accompanied by comorbid GI disorders and intestine inflammation, then we also know that eCBs and diet are reciprocally and metabolically linked and that the eCB system may be modulated not only by dietary factors and lifestyles (Bisogno and Maccarrone, 2014; Passani and Coccurello, 2016; Coccurello and Maccarrone, 2018), but also by gut microorganisms and selected bacterial taxa that can change circulating eCBs and regulate immune function and adaptive immunity (Castonguay-Paradis et al., 2020; Zheng et al., 2020). The eCB system is widely distributed all along the GI tract (Izzo et al., 2015) and the eCBs of the intestinal epithelium are functionally involved in the control of gut-brain signaling, thus influencing and being influenced by gut microbiota composition.

An elegant demonstration of this reciprocal influence is for instance showed via the association between anhedonia, as key diagnostic symptom in depression (Coccurello, 2019), and gut microbiota composition both in an animal model of depression-like behavior (Chevalier et al., 2020) and in a population cohort of volunteer twins (Minichino et al., 2021). The authors of this latter study used a model of analysis and a working hypothesis through which microbiota diversity can be exploited to predict changes in faecal and/or serum levels of eCBs and occurrence of anhedonia as trait of personality in the general population (Minichino et al., 2021). Notably, they found that PEA faecal (but not serum) levels mediates such association between microbial diversity and anhedonia incidence, further supporting the critical importance of the endocannabinoids-microbiota partnership for the understanding of symptoms that may anticipate the onset of a large set of neuropsychiatric and NDDs, including depression, substance abuse and ASD (Marrone and Coccurello, 2019). Paradigmatic is the case of A. muciniphila, a mucin-degrading Gram-negative bacterium able to regulate tight junctions and preserve gut health, whose activity opposes different pathogenetic mechanisms underlying metabolic diseases among which gut permeability, and for this also baptized “gut gatekeeper” (de Vos, 2017; Chelakkot et al., 2018). Indeed, A. muciniphila supplementation was shown to improve gut permeability, previously degraded by obesity-associated inflammation, and produce an increase of 2-AG intestinal levels (Everard et al., 2013). Later, the A. muciniphila beneficial effects have been attributed to a couple of eCB-derived lipids of the 2-AcGs family (i.e., 1-Palmitoyl-glycerol (1-PG) and 2-PG), and to their ability to bind the PPAR-α receptor (Depommier et al., 2021). Studies exploring changes of A. muciniphila density in the gut of ASD children have reported conflicting results. A. muciniphila abundance was found decreased according to the results by Wang et al. (2011), while was reported increased in another investigation (De Angelis et al., 2013). Nevertheless, here the main point is that important changes in the abundance of a bacterium that utilizes colonic mucin as substrate and protect from gut inflammation are frequently described in ASD, and that alterations of A. muciniphila can be mirrored by changes of eCB signaling.

Besides several studies reporting the beneficial effects of PEA supplementation on neurodegeneration (e.g., hippocampal damage) and social and non-social behaviors in different ASD-like rats and mice models (Bertolino et al., 2017; Cristiano et al., 2018; Herrera et al., 2018), an improvement of language skills following PEA supplementation has also been reported in two children with autism (Antonucci et al., 2015). Interestingly, PEA treatment in the ASD-like BTBR mouse model (Cristiano et al., 2018) also reduced pro-inflammatory cytokine production and intestinal permeability (“leaky” gut) at intestinal/colonic level. The PEA potential against clinical conditions involving GI inflammatory disorders has been shown in different models, as for experimental colitis in which the positive impact on inflammation and intestinal permeability has been attributed to the mediation of GPR55, PPAR-α and CB₂ receptors (Borrelli et al., 2015). Among the interesting animal models of autistic-like behaviors (e.g., fmr1 knockout) there is the exploitation of fragile X syndrome (FXS) and ASD comorbidity (Abbeduto et al., 2014) through the expansions mutations of FMR1 gene, which greatly increase the incidence of ASD disabilities (Bailey et al., 2008). The seminal investigation of the eCB system in different animal models of FXS (Maccarrone et al., 2010b) has provided evidence that in mice lacking both the FMR1-encoding FXS protein FMRP (FMRP KO mice) and the regulatory brain cytoplasmic 1 RNA (BC1 KO mice) there is a severe alteration of the functional coupling between metabotropic glutamate 5 receptors (mGlu5Rs) and eCB signaling. Indeed, in double KO mice FMRP-BC1 was observed a marked disinhibition of mGlu5R-mediated activity leading to an increased activity of 2-AG-synthetizing enzyme diacylglycerol lipase (DAGL) and enhanced 2-AG levels (Maccarrone et al., 2010b). Of note, striatal levels of 2-AG were not found changed in FMRP KO mice but an increased activity of the enzyme responsible for 2-AG degradation (i.e., MAGL) was instead observed, thus suggesting a decrease of 2-AG tone/signaling in a brain region whose function is considered of key importance for autism pathophysiology (Fuccillo, 2016). Along the same line of thought, FMRP KO mice do not display the correct functional coupling between mGlu5R-dependent long-term depression and mGlu5-DAGL-dependent 2-AG release at the postsynaptic excitatory synapse, and inhibition of glutamate release upon 2-AG stimulation of presynaptic CB₁ receptors (Jung et al., 2012)_. Ultimately, this means that 2-AG production is downscaled, and blockade of 2-AG degradation through MAGL inhibition reinstated eCB-LTD in the striatum and normalized abnormal behavioral disinhibition in the elevated plus maze (Jung et al., 2012). More recently, a couple of studies showed that the increase of AEA signaling via pharmacological FAAH inhibition can counteract either social deficit (Wei et al., 2016) or aversive memory (Qin et al., 2015), not only in FXS animal model (i.e., fmr1/FMRP KO mouse), but also induce an improvement of social impairment in the BTBR mouse model of ASD-like behavior (Wei et al., 2016). To further corroborate the functional association between ASD and eCB signaling, and in line with the dysregulation of 2-AG metabolism found in different animal models of FXS (Maccarrone et al., 2010a; Jung et al., 2012), also the use of neurodevelopmental models of autism such as the prenatal exposure to valproic acid (VPA) has provided evidence for reduced DAGL-α activity in the cerebellum and concomitant increase of MAGL-dependent 2-AG degradation in the hippocampus (Kerr et al., 2013). The same study also shown that social exposure was able to induce an enhancement of hippocampal levels of NAEs such as PEA and OEA (Kerr et al., 2013). This study appears to match the framework emerging from the analysis of gut microbiota dysbiosis in depression-like behaviors in which has been described a reduced synthesis of 2-AG (Chevalier et al., 2020), as well as the aforementioned increase of 2-AG intestinal levels induced by A. muciniphila supplementation (Everard et al., 2013) and, although not always reported, the decrease of A. muciniphila abundance in children with autism (Wang et al., 2011). A brain decrease of NAPE-PLD expression and a parallel increase of FAAH-dependent AEA degradation following exposure to VPA prompts for the idea that defective AEA metabolism could, at least in part, account for the ASD-like deficits in social communication induced by VPA exposure (Servadio et al., 2016). Incidentally, the reduction of 2-AG (Kerr et al., 2013) and particularly of AEA (Servadio et al., 2016) brain levels described after exposure to VPA is reminiscent of the decrease of AEA serum levels observed in children with ASD (Aran et al., 2019), while pharmacological FAAH inhibition was shown to mitigate VPA-induced deficits in social behavior (Kerr et al., 2016).

Although with a significant variability of the bacterial taxa involved, gut microbial diversity is severely disrupted in ASD children. Overall, a reduced ratio of Bacteroidetes to Firmicutes in ASD patients is more frequently portrayed (De Angelis et al., 2013; Strati et al., 2017), as well as the fact that Clostridium and Desulfovibrio are generally found as pathologically dominant species (Finegold et al., 2010; de Angelis et al., 2015). By contrast, beneficial bacteria such as Bifidobacterium genera are commonly underrepresented in the altered bacterial profile of children with ASD (Adams et al., 2011; de Angelis et al., 2015).

Interestingly, changes in gut microbiota dysregulates the intestinal eCB system, and probiotic supplementation with Lactobacillus acidophilus induced an increase of CB₂ receptor mRNA expression in mice colon (Rousseaux et al., 2007), which may be of importance for the above mentioned data concerning the upregulation of CB₂ receptors in PBMCs of autistic patients (Siniscalco et al., 2013). The use of different Lactobacillus strains, in both experimental models of ASD-like behavior and clinical trials conducted on children with ASD, has provided increasing evidence that selected probiotics can ameliorate ASD symptoms. Here, it is worth remember the improvement of social deficits and the recovery of oxytocin levels in the paraventricular nucleus (PVN) of the hypothalamus obtained in offspring from obese mothers through the selective administration of Lactobacillus reuteri (Buffington et al., 2016). As noted, the alterations of the gut microbiota ecosystem in ASD patients include marked deficiencies in some bacterial species, such as Bifidobacterium species, which can be regulated by a class of probiotics/“psychobiotics” (Dinan et al., 2013) able to shape the diversity of commensal gut bacteria and affect several CNS functions and psychiatric diseases through different gut-to-brain routes of communication (e.g., enteric nervous system, vagal signaling, bacteria-derived metabolites and microbiota-neuroimmune axis) (Sarkar et al., 2016). The oral treatment with Bacteroides fragilis in offspring from mothers underwent immune activation improves both integrity of epithelial barrier and social-communicative deficits, restoring physiological gut composition and serum metabolites (Hsiao et al., 2013). Some of these Bifidobacterium species that are modulated by probiotic administration have been found strongly reduced in ASD pathophysiology, as demonstrated for the significant depletion of Bifidobacterium longum (B. longum) in young children with ASD (Coretti et al., 2018). Interestingly, the shift from Western diet to the consumption of Mediterranean diet was found to increase plasma OEA/PEA ratio, and such increase was observed to positively correlate with the growth of several microorganisms, included B. longum (Tagliamonte et al., 2021). Moreover, the administration of B. longum (under the form of probiotic mix) was shown to increase in zebrafish the intestinal mRNA expression of Cnr1 and Cnr2 genes thus providing molecular evidence of the contribution of eCB signaling, and in particular of the involvement of CB₂ receptor-mediated signaling, in the anti-inflammatory effects achieved by probiotics administration (Gioacchini et al., 2017). Among other living bacteria and in addition to B. longum, the probiotic mix used included the Lactobacillus acidophilus and the Bifidobacterium infantis, and the study also showed a significant decrease of faah and mgll gene expression, thus suggesting a reduced degrading activity for both the enzymes and the potential increase of AEA- and 2-AG-mediated signaling (Gioacchini et al., 2017).

Bifidobacterium species in general, and B. longum in particular, recruit multiple pathways and mechanisms to provide their well-known health benefit in humans (e.g., in a variety of GI disorders). Indeed, the impact of B. longum on intestinal ecology is triggered via different mechanisms among which the stimulation of the activity of butyrate-producing bacteria via acetate production (Sugahara et al., 2015). Butyrate (or butyric acid) is a short chain fatty acid (SCFA) that, likewise other SCFAs (e.g., acetate and propionate), is produced by the gut microbiota upon fermentation of non-digestible carbohydrates (Morrison and Preston, 2016). By exploiting its features of histone deacetylases (HDAC) inhibitor, butyrate treatment has been used in ASD-like mice models such as in the prenatal exposure to VPA (Takuma et al., 2014) and in the BTBR model of autism (Kratsman et al., 2016), in both cases describing improvements either in recognition memory or in social behavior. However, data incoming from ASD population are more controversial. On one hand, there are data collected in Chinese ASD patients in which are described lower than normal levels of faecal butyrate and reduced abundance of butyrate-producing bacteria (e.g., Lachnospiraceae) (Liu et al., 2019). On the other hand, there are data analysis in which a drastic reduction in B. longum microbiota colonization is reported together with an increase of the butyrate-producing bacterium Faecalibacterium prausnitzii (Coretti et al., 2018). Moreover, the severity of ASD symptoms appears to inversely correlate to Faecalibacterium genus (Ding et al., 2020), and a further confirmation of the increase of butyrate faecal level in children with ASD has been previously reported (Wang et al., 2012). Recently, it has been provided evidence that gut microbial metabolites, and specifically butyrate, can directly modulate the eCB tone at gut epithelial level (Hwang et al., 2021). Using different concentrations, the study showed that both eCBs synthetizing enzymes (i.e., NAPE-PLD and DAGL) were reduced by butyrate treatment in dose-dependent manner, and that in parallel butyrate was able to upregulate MAGL (but not FAAH) expression in the gut epithelial cells (Hwang et al., 2021). Accordingly, it seems conceivable that an excessive increase of butyrate level in children with ASD (Wang et al., 2012; Coretti et al., 2018) could contribute to reduce both AEA and 2-AG synthesis and, in particular, cause a depletion of 2-AG tone in the gut epithelium. Moreover, this point of view appears compatible with the decrease of AEA, PEA and OEA serum levels in children with ASD (Karhson et al., 2018; Aran et al., 2019). A recent investigation has compared the potential derangement of the eCB machinery in children with autism as well as in the ASD-like VPA animal model (Zou et al., 2021). Data from this study account for a general decrease of eCB signaling, which is corroborated by the increased expression of both mRNA and protein levels of CB₂ receptors and eCB degrading enzymes FAAH and MAGL in PBMCs from autistic children (Zou et al., 2021). Of particular interest, the study not only confirmed the reduced blood concentration of AEA, PEA and OEA (Karhson et al., 2018; Aran et al., 2019) and the upregulation of CB₂ receptors at PBMCs level (Siniscalco et al., 2013), but also reported a reduction of 2-AG serum levels in ASD patients (Zou et al., 2021). Here, Figure 3 can provide a snapshot of the reciprocal influence between eCB signaling and gut microbiota in ASD, as well as of the therapeutic potential that results from the boosting of eCB tone through the involvement of the gut microbial community.

FIGURE 3

The figure portraits selected evidence in support of eCB signaling-gut microbiota relationship and representative mechanisms underlying the link between ASD pathophysiology and the eCB machinery. On the left panel (A), A. muciniphila supplementation can improve gut permeability and recover gut integrity as well as increase 2-AG signaling. Supplementation of Lactobacillus acidophilus increase both 2-AG and AEA signaling and CB₂ receptor mRNA expression in mice colon and in human epithelial cells. B. Longum supplementation is associated with increase of Cnr1 and Cnr2 genes intestinal expression. Below, an overall growth in Bacteroides population can produce an increase of eCB-like lipids (e.g., OEA and 2-OG) acting as GPR119 ligands. Recovery of gut microbial balance (e.g., Firmicutes to Bacteroides ratio) can be associated with PEA supplementation and an increase of 2-AG signaling cause an improvement of sociability. On the right panel (B), downregulation of CB₁ receptors found in brains of subjects with autism, and decrease of serum AEA, PEA and OEA levels in ASD patients. Below, some evidence of dysregulation of gut microbial community and reduced bacteria diversity. Gut dysbiosis in ASD is characterized by disrupted gut barrier integrity and increased gut permeability as well as by a general Firmicutes to Bacteroides imbalance. Next, the increase of Clostridia and Desulfovibrio bacteria as well as of Erysipelotrichaceae family are responsible of excessive propionate- and butyrate-producing bacteria, respectively. Boosting eCB tone and 2-AG and AEA signaling via prebiotics supplementation emerges as potential therapeutic via changes of gut microbial communities found defective in ASD patients.

Besides the effects of CBD in the Shank3 mouse model of autism (Poleg et al., 2021), and the ASD-like neurophatological phenotype and deficits in social behavior described in models of prenatal exposure to VPA, in BTBR (Kratsman et al., 2016) and in FXS fmr1 KO mice (Kazdoba et al., 2016), there is another link between mutations in Synapsin (SYN) genes and patients with comorbid ASD and epilepsy that is worth exploring (Fassio et al., 2011). Here, the relevance comes from the role that synapsins play in the regulation of synaptic transmission, plasticity, and neural development (Fornasiero et al., 2010). In particular, as demonstrated by the sophisticated assessment of ultrasonic vocalizations, synapsin II KO mice (i.e., Syn2) exhibited more marked deficits in social communication and therefore a substantial ASD-like phenotype (Michetti et al., 2017). Notably, from a functional point of view, synapsins also participate in eCB-induced neural plasticity so that the stimulation of CB₁ receptors was shown to modulate the increase of synaptic vesicles by controlling synapsins phosphorylation and downstream activation of synaptic transmission (Patzke et al., 2021). Remarkably, restoration of gut dysbiosis by FMT was shown to normalize the reduced expression of presynaptic synapsin-1 expression in the APP/PS1 mouse model of Alzheimer’s disease (Wang et al., 2021), thus corroborating the use of ASD-like synapsin II KO mice for the study of the relationship between eCB signaling and gut microbiota in ASD development.

A further point worth of consideration, briefly outlined in the introduction, involves the deleterious effects of vitamin D deficiency on the reduction of microbial diversity were briefly mentioned as a convincing example of the eCB-microbiota reciprocal contamination, especially in the light of the beneficial results produced by PEA supplementation on the recovery of a balanced Firmicutes/Bacteroidetes ratio (Guida et al., 2020). Mice bearing a deletion of vitamin D receptor show dysbiotic microbial profile with a marked reduction of A. muciniphila, thus strengthens the importance of vitamin D signaling for the maintenance of intestinal integrity and eubiosys (Su et al., 2016). However, these examples could not be limited to the detrimental effects induced by vitamin D deficiency on dysbiosis but encompass the study of ASD pathogenesis. Indeed, accumulating evidence are in line with the notion that vitamin D deficiency during pregnancy can increase the risk of NDDs including ASD development (Lee et al., 2021), as well as that vitamin D supplementation might improve the expression of ASD symptoms (Principi and Esposito, 2020). Moreover, vitamin D deficiency during development has been associated to the risk of NPDs such as schizophrenia (McGrath et al., 2010). In support of a functional interplay between eCB signaling and vitamin D metabolism, there are the disruptive effects induced by high-dose systemic Δ9-THC administration in tasks simulating schizophrenia-like aberrant processing that were boosted by developmental vitamin D deficiency (Burne et al., 2014). A recent study addressing on astrocytes specific markers of neuronal aging and the anti-inflammatory potential of PEA (e.g., against reactive oxygen species and nitric oxide production) reported greater neuroprotective effects through the use of an innovative drug formulation consisting in PEA, alpha lipoic acid and vitamin D combination (Morsanuto et al., 2020). Not only the PEA/alpha lipoic acid/vitamin D formulation showed better cell viability, anti-inflammatory and anti-oxidant effects but also higher potency in the activation of CB₂ receptors (Morsanuto et al., 2020), which have also been recently described to have a fundamental role in the preservation of social memory (Komorowska-Müller et al., 2021). Thus, to further support the importance of the endocannabinoids-microbiota reciprocal contamination for the understanding of ASD there are also the studies describing an association between vitamin D deficiency and risk of autism (Principi and Esposito, 2020) as well as the fact that PEA provided therapeutic beneficial in subjects with ASD (Khalaj et al., 2018) and the existence of a specific interaction between vitamin D metabolism, eCB signaling and changes in gut microbiota (Su et al., 2016; Guida et al., 2020; Morsanuto et al., 2020). Several key studies discussed in subchapters from Section 2.1 to Section 2.3 are summarized in Table 1.

TABLE 1

Summary table of the key studies showing involvement of eCB signaling in ASD.

eCB signaling in ASD	Subjects/System model	Major effects		Study
Circulating AEA, PEA and OEA	Children with ASD	Reduced serum levels unchanged ASD symptoms		Aran et al. (2019)
MAGL inhibition and increase of eCB signaling in BLA-NAc circuit	SHANK3 mouse model	Decreased deficits in social behavior and social avoidance		Folkes et al. (2020)
CBD-enriched oil	SHANK3 mouse model	Decreased ASD-like behavior (e.g., social anxiety)		Poleg et al. (2021)
CBD-enriched oil	Children with ASD	Improvement of symptoms (hyperactivity, social abilities)		Ponton et al., (2020), Loss et al. (2021), Bar-Lev Schleider et al. (2019)
PEA supplementation	Children with ASD	Improvement language skills		Antonucci et al. (2015)
Ultramicronized PEA + Luteolin coadministration	VPA-induced ASD-like mice	Improvement of social behavior		Bertolino et al. (2017)
Ultramicronized PEA + Luteolin coadministration	10 year old male children	Stereotypies decrease Improvement of ASD symptoms		Bertolino et al. (2017)
FAAH inhibition/increase of AEA-signaling	fmr1/FMRP KO mouse model/ASD-like BTBR mouse model	Reversion of social deficit (both models)		Wei et al. (2016)
Prenatal VPA exposure/effects in adolescent rats	ASD-like VPA rat model	Decreased DAGL-α activity/reduced 2-AG synthesis (cerebellum) Hippocampal enhancement of MAGL activity Social exposure induces AEA, PEA and OEA increase		Kerr et al. (2013)
Prenatal VPA exposure	ASD-like VPA rat model	Decrease of NAPE-PLD expression/Increase of AEA degradation/Deficits social play and communication		Servadio et al. (2016)
Pharmacological FAAH inhibition	ASD-like VPA rat model	Reduced VPA-induced deficits in social behavior		Kerr et al. (2016)
PBMCs	ASD patients	Up-regulation of CB2 receptors		Siniscalco et al. (2013)

Open in a separate window

3.1 Summary and Conclusion

As for the understanding of neurodegenerative diseases (Armstrong, 2014), causes and effects are tightly intertwined in ASD, and no conclusive statements can be made on the primary or causal events underlying the eCBs-microbiota partnership in ASD etiology. Indeed, in spite of the considerable body of investigation, our current knowledge does not allow to determine whether the disruption of microbial ecosystem could be considered a primary event causing derangement of eCB signaling or, vice-versa, whether the unbalance of eCB system might trigger the disruption of gut microbial diversity descripted in ASD patients. The aspects so far examined might support both the lines of cause-effect relationship, thus either that a “major” gut derangement of microbial population produces downstream abnormal changes of eCB signaling and ASD-associated brain alterations (e.g., microglia-dependent neuroinflammation), or that dyshomeostasis of the eCB system contributes to alteration of gut microbial composition. Nevertheless, if the current updated appraisal of the endocannabinoids-gut microbiota partnership in ASD pathophysiology has provided sufficient evidence, this is a no way-out thinking.

Instead, adding the factor “eCB signaling” to the equation “ASD and gut microbiota” should help to disclose the role of eCB system as one of the major factors responsible of brain homeostasis. Loss of control exerted by the eCB system and ASD-associated abnormal variations in AEA, 2-AG and NAEs are factors that can help to fill the empty space between neurodevelopment, gut microbiota and brain inflammation. We reported extensive evidence (e.g., paragraphs in Sections 1.2, 2, 2.1, and 2.3) according to which ASD can be viewed as a NDD with a prominent state of brain inflammation and immune dysfunction. The detailed description of altered composition of intestinal microbiota in ASD-like models such as SHANK3, MIA, BTBR and VPA exposure provides several turning points where the study of the relationship between gut ecosystem and eCB signaling can be more elaborated.

Here, only a list of selected key evidence is reported, namely: 1) the recovery of social behavior in the Shank3B ^–/ mouse is achieved by the increase of 2-AG mediated eCB signaling (via MAGL inhibition) or CBD-enriched cannabis oil (Folkes et al., 2020; Poleg et al., 2021); 2) the abnormal increase of selective bacteria family (e.g., Erysipelotrichaceae) in ASD patients with lower oxytocin levels and the parallel decrease of Lactobacillaceae and Erysipelotrichaceae families in animals with increased levels of 2-AG and 2-MAGs (Dione et al., 2020); 3) the possibility to predict eCBs changes such as PEA faecal levels through microbiota diversity (Minichino et al., 2021); 4) the association between gut permeability improvement, A. muciniphila supplementation, increase of 2-AG and 2-AcGs family intestinal levels (Everard et al., 2013; Depommier et al., 2021) and decreased A. muciniphila abundance in ASD children (Wang et al., 2011); 5) the beneficial effects of PEA supplementation on neuroinflammation and social behaviors in ASD-like BTBR mouse model (Cristiano et al., 2018); 6) the possibility to counteract social deficit by increasing AEA signaling both in FXS (i.e., fmr1/FMRP KO mouse) and BTBR models of ASD-like behavior (Wei et al., 2016); 7) the evidence for reduced DAGL-α activity and increase of 2-AG degradation in the neurodevelopmental VPA model (Kerr et al., 2013); 8) the increase AEA degradation following exposure to VPA (Servadio et al., 2016) and the mitigation of social communication deficits after FAAH inhibition in the same model (Kerr et al., 2016); 9) the possibility to increase colon expression of CB₂ receptor mRNA after L. acidophilus supplementation (Rousseaux et al., 2007) mirroring the upregulation of CB₂ receptors in PBMCs of ASD subjects (Siniscalco et al., 2013); 10) the possibility to increase the intestinal mRNA expression of Cnr1 and Cnr2 genes as well as decrease of faah and mgll gene expression after B. longum, L. acidophilus and Bifidobacterium infantis supplementation (Gioacchini et al., 2017); 11) the existence of a functional interaction between eCB signaling, vitamin D metabolism, and changes in gut microbiota in ASD (Su et al., 2016; Guida et al., 2020; Morsanuto et al., 2020) 12) the ability of selected probiotics such as B. infantis to improve KP metabolism (Desbonnet et al., 2008), and also stimulate the expression of Cnr1 and Cnr2 genes in zebrafish intestinal epithelium (Gioacchini et al., 2017) and improve autistic core symptoms (Grossi et al., 2016); 13) the microglia alterations in density and morphology observed in ASD patients and ASD-like MIA mouse model (Hadar et al., 2017; Bilbo et al., 2018), and the possibility to recover social behavior, microglia activation and M2 phenotype by enhancing the eCB machinery; 14) the fact that intestinal SFB expression stimulates T_H17 cells and increase ASD risk (Kim et al., 2017) but enhancement of eCB signaling improves T_H17-driven diseases (Kozela et al., 2013; Jackson et al., 2014) and Lactobacillus plantarum administration decreases SFB presence and contribute to the expression of Cnr1 and Cnr2 genes (Gioacchini et al., 2017); 15) the fact that SCFA-producing bacteria are stimulated by physical exercise and that the parallel increase of AEA, OEA and PEA mediate the butyrate-dependent effects on anti-inflammatory activity (Vijay et al., 2021); 16) lastly, the fact that selected gut microbiota bacteria (e.g., Bacteroides) have been found capable to produce eCB-like molecules such as commendamide (N-acyl-3-hydroxypalmitoyl-glycine), which is an agonist at GPR132 (Cohen et al., 2015), and N-oleoylserinol that is an agonist of the eCB receptor GPR119 (Cohen et al., 2017).

Finally, a still unexplored area of investigation might emerge from the more extensive understanding of the role of astrocytes in ASD pathogenesis (Gzielo and Nikiforuk, 2021), and in particular by focusing the study on astroglial CB₁ receptors and lactate signaling in astrocyte-neuron communication (i.e., astrocyte-neuron lactate shuttle, ANLS) as well as on the perturbation of lactate levels and the alterations of lactate-producing bacteria in gut microbiota of subjects with ASD. There is indeed evidence that these aspects could be re-examined at the light of their functional relationship with the ASD pathophysiology. Lactate is known to act not only as signaling molecule between astrocytes and neurons where is required to meet energy needs and maintain brain homeostasis (i.e., metabolic coupling and ANLS) (Pellerin and Magistretti, 1994; Mason, 2017), but also as signaling molecule involved in memory formation (Suzuki et al., 2011) and in schizophrenia-like aberrant behavior (Sun et al., 2021). Several clinical surveys report higher blood lactate levels and higher lactate-to-pyruvate ratio in ASD (Oh et al., 2020), which are biochemical changes that could at least in part attributed to the dysregulation of the WNT/β-catenin pathway (Vallée and Vallée, 2018) and/or to reduction of gut bacteria responsible for lactate fermentation such as the decrease of Veillonellaceae described in subjects with ASD (Kang et al., 2013). Moreover, in some cases, the higher presence of lactate-producing bacteria might contribute to the reduction of SCFAs levels, and in particular of butyrate, whose consequences (e.g., inflammation susceptibility) have been emphasized earlier. Although with important exceptions (Coretti et al., 2018), in ASD has been reported not only a reduction of SCFA- and butyrate-producing bacteria, but in particular a reduction of the genera Anaerostipes and Eubacterium, (Liu et al., 2019; Cao et al., 2021), which are both essential for lactate-to-SCFAs conversion and prevention of lactate accumulation (Duncan et al., 2004).

In conclusion, the idea that gut microbiota can determine and underlie ASD pathophysiology is increasingly accepted, and in this critical review we have provided extensive evidence for gut dysbiosis, reduced microbial diversity, increased inflammation of gut epithelium, selective derangement of microbial population and gut ecosystem in ASD patients, aberrant immune profile, role of SCFAs and microbiota-derived metabolites and effects of probiotic intervention. In parallel, we discussed the literature focused on the emerging issue of the involvement of the eCB machinery in ASD and ASD-associated alterations of social behavior, as well as on the demonstration in ASD patients and in ASD-like animal models of a general decline in eCB signaling, and the potential therapeutic applications thereof in terms of both achievements and failures. A great attention has been addressed to the perspective to consider together the two aspects, and the possibility that the eCBs-microbiota partnership might offer novel pathways of investigation and synergistic options for therapeutic intervention, both pharmacological and nutritional, especially with the aim to improve the quality of life of patients and their caregivers.