ASD is a complex neurodevelopmental disorder with multifactorial etiology, characterized by the impairment of two main functional areas: (1) continuous communication and social interaction deficits, and (2) restrictive and repetitive behaviors and interests. The incidence among sexes is a male/female rate of 3:1 ^[1][2][1,2].

ASD diagnostic criteria are available in the Diagnostic and Statistical Manual of Mental Disorders-5 [3], describing persistent deficits in social communication and social interaction across multiple contexts, such as deficits in social–emotional reciprocity; nonverbal communicative behaviors; and developing, maintaining and understanding relationships. Moreover, ASD children present behavior restricted and repetitive patterns of interests or activities, as manifested by at least two of the following: (a) repetitive or stereotyped motor movements, use of objects or speech; (b) insistence on sameness, inflexible adherence to routines or ritualized patterns of verbal or nonverbal behavior; (c) fixated and restricted interests that are abnormal in intensity or focus; and (d) hyper- or hypo-reactivity to sensory input or unusual interest in sensory aspects of the environment. Such symptoms must be present in the early period of development, be generally evident at the age of 3 and cause significant clinical impairment of global functioning. In 2014 in the United States, the Centers for Disease Control (CDC) estimated 1 case out of 68 children of the age of 8 [4], and in 2018, the CDC reported an ASD prevalence rate of 1 in 44 or an incidence rate of 2.3% [5].

Considering the great geographical variability and the methodologic differences employed in prevalence studies, the actual prevalence of ASD is estimated to be higher than 2.5% in the United States and 1.5% in Denmark, Finland and Sweden. In Italy, the prevalence of children diagnosed with ASD is estimated to be 1 out of 77 among children between 9 and 17 years old, with a higher risk for the male sex; the number of male patients is four times higher than the number of female patients.

The disease is now defined as a complex disorder with multifactorial etiopathology because it seems to be determined by the contribution of many risk factors, such as genetics, epigenetics and environmental factors.

Much evidence supports genetic factors as the predominant cause of ASD. Emphasis on the genetic component from epidemiological data derives from familiarity and the high incidence of the autistic behavioral phenotype in the context of genetic diseases with well-known etiology. The amount of ASD cases with a genetic base represents 10–20% of the overall number of autistic patients. Concordance studies on twins have shown very suggestive data; the concordance among monozygotic twins has been shown to be variable from 70% to 90%, whereas that between dizygotic twins is between 0–10% [6]. Moreover, ASD incidence is estimated to be 2% among siblings, with a 100 times higher risk than the normal population (0.02%). Furthermore, the recurrence risk is significantly higher in families with a first diagnosis of ASD than it is in the normal population. In particular, families with a first child diagnosed with ASD have greater probabilities of having an autistic second child depending on the child’s sex, as follows: 15–25% if the child is male, and 5–15% if the child is female. Last, the presence of many first-grade relatives diagnosed with ASD testifies to the importance of the genetic hypothesis ^[7][8][7,8]. There are ASD forms that are similar to genetic syndromes, and these forms represent 10% of all ASD diagnoses. Beyond the findings of specific alterations that cause specific syndromes, genetic anomalies implied in ASD can be caused by mutations in one gene or in the total number of copy variations (CNVs). Besides specific genetic alterations that cause well-known clinical pictures, there are uncommon genetic variants (with a documented presence of less than 1% of the general population). These mutation sites are in correspondence with genes of great importance in the process of neurodevelopment, and the most documented are CNVs, such as microdeletion or microduplication; nonsense mutation with the insertion of a stop codon; and missense mutation with the creation of aberrant products, such as inactive proteins or proteins with reduced biological activity [9].

Some relevant examples include genes that code for synaptic transmembrane proteins, such us neuroligine 3 and 4 and neurexine 1 and 3, which are crucial for synaptic function; others are SHANK family genes (SHANK 1, 2, 3) that codify for proteins involved in synapsis formation and dendritic spine maturation. In detail, SHANK 3 is involved in dendritic development and in a pathway with reelin, a protein that is essential for the stabilization and the laminar organization of the cerebral cortex [10]. Even if the scientific community well accepts the role of genetic anomalies, many studies have shown similar associations with environmental risk factors.

A plethora of environmental risk factors have been taken into consideration, and most of them refer to the pre/peri-natal period because the maximum development of the central nervous system (CNS) is in this period. Risk factors that influence neurodevelopment and provoke long-term alterations in the brain’s physiology include pre-natal exposition to viral infections (e.g., Cytomegalovirus and Rubivirus); environmental toxic substances, such as pesticides, phthalates, solvents, environmental pollutants, and heavy metals; stress; alcohol intake; and diet [11]. An association between maternal conditions and ASD risk has been demonstrated.

The most common predisposing factors of ASD are neurodevelopment anomalies during the first and the second trimester of pre-natal life [12]. Other causes are less frequent but not completely negligible. Among these, cerebellar damage has been identified in the peri-natal period, which increases the risk of developing autism by 30 times. Neuropathology post-mortem studies on histological samples of CNS taken from autistic patients have shown the presence of cytoarchitectonic anomalies, which can involve various brain regions. Among these, reduced apoptosis and/or enhanced cellular proliferation (particularly evident in macrocephalic patients), neuronal migration alteration or anomalies in the process of cellular maturation and differentiation have been described [13]. From a functional point of view, neuropathological anomalies bring the formation of an atypical neural network characterized by reduced long-distance connectivity and exceeding local connectivity [14].

A bidirectional interaction between the gastrointestinal (GI) tract, gut microenvironment and CNS, called the ‘microbiota–gut–brain axis’, regulates intestinal and neurological homeostasis. An impairment of this complex system can promote, in the presence of other contributing factors, the pathogenesis of nervous-system-related diseases, such as ASD.

Enteric symptoms (including constipation, diarrhea, recurrent abdominal pain/bloating and gastroesophageal reflux) are frequent among ASD patients, who often present alterations in intestinal motility and dysfunction of the epithelial barrier.

2. Intestinal Permeability in ASD

Children with ASD frequently exhibit GI tract problem symptoms. These illnesses’ underlying causes, though, are still poorly understood. It is speculated that the pathophysiology of ASD may be influenced by the GM and its metabolites ^[15][16][17][15,16,17]. Numerous articles have identified the impact of GI alteration, GM and CNS function, as well as the potential participation of the microbiome–gut–brain axis [18]. Given that the prevalence of GI symptoms in ASD children may reach 70%, microbiome and gut–brain connections are likely to play a significant role in ASD [19]. Additionally, the severity of ASD is correlated with the prevalence of GI symptoms [20], demonstrating the role of the gut in the pathogenesis of ASD ^[21][22][21,22]. ‘Leaky Gut syndrome’ refers to a situation in which the small or large intestine’s epithelial barrier function is compromised, resulting in increased types and quantities of molecules and cells that can flow from the gut to the circulatory system and vice versa [23]. The intestinal microbiota, mucous layer, intestinal epithelium, elements of innate and acquired immunity, hormonal and neuroenteric systems, vascular–lymphatic system and digestive enzymes comprise the functional unit of intestinal permeability. It serves as the body’s first line of protection against toxic, immunogenic and pro-inflammatory substances by maintaining a delicate balance between the intestinal lumen’s antigenic charge and the intricate structure of the intestinal mucosa. It is crucial for preserving good health, stopping systemic and intestinal inflammation and suppressing the immune system. Only trace amounts of antigens can pass through the normal intestinal barrier to engage with the innate and adaptive immune systems. A change in its function may promote bacterial and antigen transit, which may then result in pathogenic diseases [24]. The core of intestinal permeability is the occlusive, tough intracellular connections. These tight junctions comprise a system of numerous proteins in the paracellular space between each cell in the gut lining. They are indeed responsible for the epithelial barrier’s functionality. This insurmountable but selective barrier is reinforced by a thick mucus coating and interacts steadily with luminal contents and enteric bacteria. It is well known that the gut and the brain have a strong relationship and regularly interact. Neuropeptides that allow two-way communication between the gut and the brain include substance P, calcitonin gene-related peptide, neuropeptide Y and vasoactive intestinal polypeptide [25]. Cortisol, a key player in developing anxiety and depressive disorders, is also released by the hypothalamic–pituitary–adrenal axis and controls intestinal motility, integrity and hypersecretion [26]. The GM can sequentially affect the function of the CNS through neuronal, endocrine, immunological and metabolic processes because communication is bidirectional [27]. The relationship between the gut and the brain in the etiology of autism is assumed to be increased gut permeability, which has been linked to ASD. For instance, it was demonstrated that the injection of propionic acid (that is produced by intestinal bacteria) in rats’ brains [28] results in neuroinflammation and symptoms resembling those of ASD [29]. This might explain why children with ASD experience worsened symptoms when exposed to food preservatives containing propionic acid. In this particular case, increased gut permeability would allow propionic acid to enter the bloodstream and eventually leak into the blood–brain barrier. Intestinal permeability prevents intestinal contents from entering the bloodstream and suppressing subsequent immunological inflammatory responses and GI illnesses [30]. As a result, an intact gut barrier decreases inflammatory responses. De Magistris et al. showed that 36.7% of ASD children have aberrant intestinal permeability compared to less than 5% of control children [31]. Similar data were reported by D’Eufemia et al., who reported that 43% of ASD children with GI symptoms have impaired intestinal permeability [32]. Recently, to investigate the association between intestinal permeability and behavior, Teskey et al. measured the intestinal fatty acid binding protein as a marker of intestinal epithelial damage in the plasma of children with ASD and found that an increase in this protein correlates with a severe deficit in communication, social interaction and maladaptive behavior [15].

3. Microbiota–Gut–Brain Axis Involvement in ASD

Two millennia ago, Hippocrates stated that “All disease begins in the gut” [33]. A growing interest in systems and organs closely related to the CNS has been triggered by the high prevalence of some specific medical comorbidities, such as GI disorders, in individuals with ASD compared to peers with typical development. Over the past decade, the bidirectional communication between the gut and the brain, the so-called “gut–brain axis,” has been the focus of preclinical and clinical research, investigating its possible role in the etiopathogenesis of some neuropsychiatric conditions, including ASD [34]. This interplay of bidirectional communication connecting mind and body provides a physiological rationale for interpreting these conditions within the biopsychosocial model. The biopsychosocial model examines the reciprocal and complex interactions among biological, psychological and environmental factors contributing to disease [35], mainly due to the growing knowledge of this axis [36]. The enteric nervous system (ENS) is a well-defined entity capable of regulating the intestinal functions of mobility, secretion and mucosal transport entirely autonomously from the CNS [27]. As demonstrated in the animal model, the ENS, even when the gut is entirely denervated by the CNS, it can function on its own. However, it maintains a bidirectional communication pathway with the CNS [27]. The CNS, after integrating a variety of information regarding internal and external environmental changes, performs parasympathetic control through the vagus nerve with cholinergic efferents acting on the myenteric plexus (motor movements) and Meissner’s plexus (secretions of the submucosal glands), and with sympathetic control through the splanchnic nerves that reduce the motility of the intestine and blood supply to the splanchnic circulation [27]. The pronounced synergy and continuous exchange of information along this axis is possible because of the vast neurochemical assets available to the enteric nervous system. In in vivo studies, the peripheral stimulation of vagal fibers has been shown to lead to dopamine release in the reward system [37]. This type of neuronal network, which connects the GI tract with different levels of the CNS involving the aforementioned neural pathways, humoral signaling molecules and hormones ^[22][38][22,38], is the functional basis of the gut–brain axis [39]. Notably, the hypothalamic–pituitary–adrenal axis is also involved in this network. It is responsible for coordinating the body’s adaptive responses following stressful events, modulating the composition of the GM and the integrity of the intestinal barrier through the secretion of norepinephrine and dopamine at the neuroendocrine level. The resulting dysbiosis state and altered permeability can induce the translocation of bacterial components, which promotes the secretion of adrenocorticotropic hormone, corticosterone, prostaglandins and proinflammatory cytokines [26]. Increasing evidence has demonstrated the existence of a complex and still not well-understood two-way connection among the GM, intestine and CNS. In recent years, as microbiological and neuroscientific knowledge has advanced and the role played by the GM in host physiology has become more evident, there has been a shift in the conception of the gut–brain axis, and the term has been renamed the “microbiota–gut–brain axis” [40]. Indeed, communication between gut microbes and the gut–brain axis occurs through multiple pathways and mechanisms, including immune, neural, metabolic and endocrine pathways that arbitrate bidirectional signaling locally in the gut and peripherally ^[22][41][22,41]. Interest in the microbiota–gut–brain axis was ignited when Lozupone et al. demonstrated an increased hypothalamic–pituitary–adrenal axis response to stress in germ-free mice compared with non-germ-free mice [42]. Bravo et al. showed that supplementation with Lacticaseibacillus rhamnosus can modify GABA receptor expression in cortical regions, the hippocampus and the amygdala, with subsequent reduction in anxiety- and depression-related behaviors and stress-induced corticosterone levels, suggesting the involvement of the neuroendocrine axis. Interestingly, these effects were reversed after vagotomy, suggesting a crucial role of the peripheral nervous system in the gut–brain connection [43]. Interactions with gut microbes occur in the intestinal barrier, which is an essential and highly dynamic interface between the host and the outside world consisting of several structures, including secretory immunoglobulin (Ig) A molecules, antimicrobial peptides, lysozyme and secretory phospholipase A2, intestinal epithelial cells and adaptive immune cells (macrophages, T cells, B cells and dendritic cells) [22]. Understanding the mechanisms that cause dysfunction in gut–brain communication has made essential contributions to understanding the basic pathophysiology of the microbiota–gut–brain axis in patients with ASD and has encouraged the proposal of new therapeutic perspectives [44]. Indeed, microorganisms harbored in the GI tract can modulate the activation of cells of the immune system and intestinal epithelium, transducing inflammatory or anti-inflammatory signals to the enteric nervous system and, consequently, to the CNS [45]. Several authors have hypothesized that alterations in the GM may contribute to the expression of the autistic phenotype or exacerbate the severity of symptoms in individuals genetically predisposed to ASD ^[46][47][48][46,47,48]. In fact, the microbiota covers many functions; it supports nutrient digestion, regulates metabolism, processes hazardous substances, participates in detoxification and organizes control of the immune system. Microbiota composition is influenced by many factors, including genes, the maternal microbiome, nutrition, brain activity and mood. This means that something that starts as an emotion in the brain affects the gut and the signals generated by the microbiota. These signals are, in turn, transmitted to the brain, often making that emotional state more intense and prolonged [49]. It has been hypothesized that ASD can result from any disruption that can alter the balance of the microbiota and the gut and that the disruption of a single part of this delicate mechanism can potentially impact any link in the chain [50]. Recent studies have shown that some bacteria belonging to the phylum Bacteroidetes [Barnesiella, Parabacteroides, Bacteroides, Odoribacter, Prevotella, Proteobacteria (e.g., Proteus and Parasutterella) and Alistipes] are more abundant in ASD patients as compared to the general population, whereas Actinobacteria, (Bifidobacterium species) are often less abundant in ASD patients ^[51][52][51,52]. Alterations in gut microbial composition can lead to altered levels of neuroactive molecules, such as short-chain fatty acids (SCFAs), particularly propionic acid, acetic acid and butyric acid, and lipopolysaccharides that can induce changes in the CNS through the endocrine pathway [53], and they have all been found to be overexpressed in ASD populations [54], although the results remain mostly contradictory ^{[55][56][57][58]}[55,56,57,58]. Studies in animal models have shown an overgrowth of Firmicutes species (spp.), a reduction in Bacteroidetes spp. and increased butyric acid levels in male mice with autistic-like behaviors ^[59][60][59,60]. Several studies have demonstrated that SCFAs can permeate the blood–brain barrier [61] and modulate the neural characteristics of brain cells ^[62][63][64][62,63,64]. In fact, the administration of propionic acid in mouse models, pre-natally through a pregnant mother ^[65][66][65,66] and in the early years of life ^{[65][66][67][68]}[65,66,67,68], as well as increasing dietary propionic acid in children [69], have facilitated the onset of autism-like behaviors in all animal and human studies. SCFAs of bacterial species act on specific GI and immune pathways, and this may impair gut metabolic function and increase immune response and mitochondrial dysfunction, resulting in increased oxidative stress. Sustained oxidative stress may, in turn, enhance intestinal permeability and increase inflammation. Microglial activation in the brain can further increase inflammation, resulting in the malfunction of synapses and manifestation of behavioral abnormalities and neuropathology ^{[70][71][72][73][74]}[70,71,72,73,74]. Like mammalians, also in Drosophila, an alteration in the GM can cause an epithelial oxidative burst, causing changes in gut permeability and affecting longevity and behaviors ^[75][76][77][75,76,77]. The GM can release metabolites that can modulate levels of psychoactive compounds in the CNS or produce these neuroactive substances on their own ^[22][78][22,78]. Among the different neurotransmitters involved in ASD that appear to be regulated by the microbiome are serotonin, glutamate and dopamine ^[79][80][79,80]. For example, Bifidobacterium spp. and Lactobacillus spp. are producers of γ-aminobutyric acid (GABA) ^[79][81][82][79,81,82]; Candida spp., Escherichia spp., Enterococcus spp. and Streptococcus spp. are producers of serotonin [81]; Escherichia spp. and Saccharomyces spp. generate norepinephrine; Lactobacillus spp. is a producer of acetylcholine; and Bacillus spp. and Serratia spp. are producers of dopamine [83]. In addition, elevated levels of norepinephrine have been detected with increased amounts of Bacillus, Enterococcus, Escherichia, Saccharomyces or Streptococcus spp. in the intestine ^[79][81][79,81]. Serotonin is a neurotransmitter that plays a fundamental role in mood regulation through its influence on microglial cells in the CNS [84]. Enterochromaffin cells distributed along the intestinal mucosa produce 95% of it [85]. In the study by Yano et al., it was shown that indigenous spore-forming bacteria in mouse and human microbiota promote serotonin biosynthesis from enterochromaffin cells and have a significant impact on host physiology by modulating GI motility and platelet function, suggesting direct metabolic signaling from gut microbes to enterochromaffin cells [84]. The GM possesses enzymes that regulate tryptophan metabolism pathways, leading to the production of serotonin, kynurenine or indole derivatives. Therefore, by controlling the amount of tryptophan, the microbiota can influence the brain’s amount of serotonin [86]. Hyper- and hypo-glutamine patterns at different developmental stages underlie the neurotransmitter communication hypothesis in the pathogenesis of ASD. Various studies on the antagonists of NMDARs or AMPARs have shown clinical benefits in ASD [87]. Similarly, the excitatory glutamate pathway is likely one of the factors involved in the etiopathogenesis of ASD. It is part of gut–brain communication, as it mediates trans-synaptic signaling. It is also implicated in cell adhesion, linking pre- and postsynaptic neurons, and it shapes neural networks by specifying synaptic functions [88]. In addition, a correlation between ASD phenotypes and glutamate/glutamine levels in various brain areas has been displayed through the use of in vivo neuroimaging of ASD individuals [87]. Furthermore, altered dopamine signaling has been associated with ASD in both mice and humans. A study by DiCarlo et al. suggested that mice that are homozygous for the T356M DNA variant of the SLC6A3 gene, which encodes the dopamine transporter, manifests altered dopamine signaling and metabolic dysfunction, weigh less and have reduced body fat [89]. The authors found a significant decrease in Fusobacterium abundance at the oral level. Moreover, there is a positive association among Fusobacterium abundance, better glucose management and decreased body fat [89]. However, neurotransmitters produced in the gut are unlikely to reach the brain because of the presence of the blood–brain barrier. A likely exception is GABA because its transporters are present in the blood–brain barrier. However, the CNS can be indirectly affected by neurotransmitters produced in the gut because they can act on the enteric nervous system ^[90][91][90,91]. In fact, the blood–brain barrier is another crucial anatomo-functional structure of the microbiota–gut–brain axis. Indeed, it modulates the trafficking of specific molecules and contributes to the maintenance of normal neuronal activity. It is also implicated in immunological functions and protects the brain from bacteria and microbial molecules during the CNS developmental phase and in adulthood ^[92][93][92,93]. Balanced GM is necessary to develop and maintain a normal blood–brain barrier. In addition, structural changes, including increased activation of microglial cells, have been observed post-mortem in the brains of autistic individuals ^[58][76][58,76]. All of this underlies the hypothesis that ASD is a condition caused or at least accompanied by immune activation in the brain that leads to a neuroinflammatory state and could then lead to malfunctioning synapses [76]. In the inflammatory phase, arginine vasopressin is released from the brain, and it is a metabolite known to act on social behavior and is considered a biomarker for ASD [76]. Furthermore, there is a reduced number of Purkinje cells in the cerebellum of ASD patients ^[35][94][35,94], which makes them susceptible to the tetanus neurotoxin produced by Clostridia tetani [94]. In fact, a high amount of Clostridium spp. was found in subjects with ASD, which could explain the decrease in Purkinje cells in the cerebellum of these subjects [94].