The KYN pathway, also derived from tryptophan and modulated by gut microbes, largely depends on indoleamine-2,3-dioxygenase (IDO) and, to a lesser degree, tryptophan-2, 3-dioxygenase (TDO) for metabolization
[78]. IDO, expressed in all body tissues, is typically activated in the presence of pro-inflammatory cytokines, whereas TDO, expressed primarily in liver tissues, is activated by glucocorticoids
[78][79][78,79]. Once transformed from tryptophan, KYN metabolizes into two downstream metabolites, neuroprotective kynurenic acid (KA) and neurotoxic quinolinic acid (QA)
[78]. Recent evidence suggests that altered KYN metabolism is indicative of greater tryptophan depletion and an impaired serotonergic pathway in ASD
[80]. In a study investigating the role of the KYN pathway in ASD, Bryn et al.
[81] showed that children with ASD had significantly lower KA serum levels, higher KYN/KA ratios, and higher QA serum concentrations than TD children. These findings are consistent with those of Gevi et al.
[73], who found that tryptophan was disproportionately metabolized into QA, with significantly decreased levels of KA, in children with ASD. Both studies demonstrate an increased potential for neurotoxicity in children with ASD, which is thought to be involved in the pathophysiology of the disorder
[73][81][73,81]. Interestingly, Xiao et al.
[55] found increased KA in mice following FMT from children with ASD. These levels correlated with specific bacteria (e.g., genera in the orders
Clostridiales and
Bacteroidetes), supporting their modulatory role in tryptophan metabolism, but demonstrating a need for further research on how microbiota alter KYN-pathway products
[55].
Although there is minimal literature associating gut microbiota with the KYN pathway in human ASD populations, studies in other clinical and normative populations have provided evidence to support this relationship
[78]. Interestingly, Luna et al.
[82] found in their study of ASD microbiome-neuroimmune signatures that along with tryptophan and serotonin levels, inflammatory cytokine levels correlated with certain bacterial species in children with ASD and functional GI disorder comorbidities. No direct link was made to the KYN pathway; however, because IDO is typically activated in response to cytokines
[78][79][78,79], there is reason to investigate whether the abnormal microbial profile of individuals with ASD may be implicated in the dysregulation of the KYN pathway.
4.1.3. Serotonin Pathway and ASD
Serotonin (also referred to as 5-HT) is important for mood regulation, higher order cognition, and neurodevelopment of both the CNS and ENS
[83][84][83,84]. Although the majority (>90%) of serotonin comes from enterochromaffin cells in the GI tract, serotonin is also synthesized in the neurons of the ENS and CNS, particularly the raphe nuclei in the brainstem
[85]. Gut microbiota and their metabolites can influence central and peripheral serotonin production and metabolism through a variety of mechanisms
[71][86][71,86]. Because only a small percentage of tryptophan is converted into serotonin, any alterations to its metabolism and availability can pose a significant risk to one’s health
[87].
Approximately 30% of ASD patients have hyperserotonemia, or elevated whole-blood serotonin levels
[88], which is believed to be due in part to increased serotonin production in enterochromaffin cells in the gut
[89]. Based on similar and replicated findings, it has been postulated that hyperserotonemia may represent a highly heritable biomarker of ASD and that the serotonin pathway as a whole may be dysfunctional in at least a subgroup of ASD individuals
[89][90][89,90]. In preclinical models, hyperserotonemia has been linked to social-behavioral deficits characteristic of ASD
[49][91][92][49,91,92]. Tanaka et al.
[92], for example, found that a tryptophan-depleted diet, which decreases brain serotonin levels and regulates gene expression inside the serotonin system, improved social impairments of genetically modified ASD mouse models. Lim et al.’s
[91] report of elevated serum serotonin levels in environmental risk factor mouse models of ASD that were associated with changes in bacteria known to stimulate serotonin production suggests that alterations in serotonin and hyperserotonemia itself may have a microbial origin. The connection between serotonin and the microbiome has been made in humans as well, as demonstrated by a link between increased GI symptom severity and hyperserotonemia in ASD youth
[93]. Other studies investigating serotonin-related dysfunction in children with ASD and co-occurring GI symptoms have implicated fecal metabolites in the metabolic network of various neurotransmitters, including serotonin
[33], and have found increased levels of serotonergic metabolites, including 5-HIAA, the main metabolite of serotonin, in the rectal tissue of ASD youth with co-occurring functional GI disorders
[82]. These metabolite levels correlated with the dysbiosis of several bacterial species, demonstrating a potential microbiome profile for ASD
[82].
SERT Ala56, the most common variant of the serotonin-selective transporter responsible for serotonin reuptake in both the brain and intestines, has been found to be overexpressed in ASD patients and linked to neurobiological and GI symptoms in a genetically modified murine ASD model
[94]. SERT Ala56 mice are also known to exhibit serotonin-related dysfunction, including excess clearance of central serotonin, augmented serotonin receptor sensitivity, and hyperserotonemia
[95]. Research supporting connections between an altered serotonin system and ASD pathophysiology has demonstrated a positive relationship of serotonin and SERT levels with autism symptom severity in humans
[96]. Furthermore, numerous animal studies have implicated gene polymorphisms of SERT, as well as genetic and surface transporter expression and function, in the underlying repetitive behaviors and social behavior deficits of ASD, for a review, see
[97].
Contributing to the link between serotonin, gut microbiota, and ASD, the BTBR inbred strain has been shown to display (1) reduced SERT density and binding throughout the brain and increased serotonin activity in the hippocampus (a brain region involved in learning, social, and emotional processing, and found to be abnormal in ASD)
[98][99][100][98,99,100]; (2) changes in intestinal microbiota associated with slowed GI motility and impaired intestinal serotonin production
[48]; and (3) increased sociability following brief exposure to serotonin reuptake inhibitors
[98][101][98,101] and tryptophan supplementation
[102]. Taken together, these studies support the hypothesis that altered gut microbiota are involved in the tryptophan-serotonin metabolic pathway in ASD and provide a framework for future studies aiming to alleviate the GI and, consequently, behavioral symptoms of ASD patients.
A recent study by Fung et al.
[103] showed that the gut bacterium
Turicibacter sanguinis expresses a neurotransmitter sodium symporter-related protein with sequence and structural homology to mammalian SERT. This microbe imports serotonin through a mechanism that, like its host homologue, is inhibited by the selective serotonin reuptake inhibitor, fluoxetine. Serotonin reduces expression of sporulation factors and membrane transporters in
T. sanguinis, which is reversed by fluoxetine exposure. Treating
T. sanguinis with serotonin or fluoxetine modulates its competitive colonization in the GI tract of antibiotic-treated mice. In addition, fluoxetine reduces membership of
T. sanguinis in the gut microbiota of conventionally colonized mice. One may speculate that genetic variants exist for the microbial SERT-like mechanism and that alterations in the bidirectional host-microbe interactions in tryptophan metabolites play a role in ASD pathophysiology, including gut symptoms.
4.2. Serotonin in the Brain and Relationships with Behavior
Several functional neuroimaging studies indicate that the microbiota-modulated serotonergic system affects neural functioning in ASD. For example, positron emission tomography (PET) ligand studies have observed atypical serotonin functioning throughout the brain that was associated with greater social deficits in individuals with ASD
[104][105][104,105]. One study found that ASD participants had significantly less binding of thalamic serotonin receptors than age-matched controls and that binding potential in the ASD group was negatively associated with social communication impairments
[104]. Another PET study found a global reduction in SERT binding in adults with ASD
[103]. Moreover, reduced SERT binding in the anterior and posterior cingulate cortices (ACC; PCC) was associated with impaired social cognition and reduced binding in the thalamus was associated with repetitive behaviors
[105]. A more recent PET study found lower serotonin transporter availability in the total gray matter and brainstem of adults with ASD, relative to matched controls
[106]. Serotonin transporter availability in the nucleus accumbens, ACC, and putamen was positively correlated with social cognition, indicating that the serotonin transporter may be a marker that can be targeted in pharmacological interventions
[106].
Additionally, a polymorphism of the serotonin transporter gene (serotonin-transporter-linked-promoter region; 5-HTTLPR) has been shown to affect brain function in ASD. Short variants of 5-HTTLPR, which reduce serotonin transporter expression
[107], are associated with impairments in social communication and interactions in individuals with ASD
[108][109][108,109]. This is in line with other findings that have linked 5-HTTLPR to the default mode network (DMN), a neural network engaged during passive self-referential cognitive processing and associated with social cognition
[110]. Wiggins et al.
[111] found that while youth with ASD who had low expressing 5-HTTLPR alleles had stronger posterior-anterior DMN connectivity than those with high expressing genotypes, the opposite was true for TD controls. Thus, resting-state connectivity was shown to be influenced by 5-HTTLPR in a different pattern in ASD than typical controls, suggesting that individuals with low expressing 5-HTTLPR genotypes may comprise a subtype of ASD.
The same researchers showed that ASD youth with low expressing 5-HTTLPR genotypes also had atypical amygdala functioning while performing social tasks
[112][113][112,113]. During an observational face-processing task, youth with low expressing 5-HTTLPR genotypes failed to show amygdala habituation to repeated observation of sad faces compared to TD controls and ASD youth with high expressing genotypes. Building on these findings, Velasquez et al.
[112] showed that when viewing happy faces, the ASD group with the low expressing genotypes had abnormally high rates of functional connectivity between the amygdala and subgenual ACC than ASD individuals with higher expressing genotypes and TD groups of higher and lower expressing genotypes. ASD participants with greater amygdala-subgenual ACC connectivity, key regions involved in emotion arousal and regulation, also showed higher expressing genotypes of 5-HTTLPR and less social dysfunction. Together, these findings indicate that, in ASD, it is possible that some socioemotional functioning during rest and social processing may be influenced by atypicalities in the serotonergic system
[112][113][112,113]. These results further support the importance of the 5-HTTLPR genotype in examining the heterogeneity of social function in ASD. Nevertheless, a recent meta-analysis found no support for a direct effect of 5-HTTLPR polymorphism on risk of ASD, indicating that patterns across studies may be difficult to find given the heterogeneity of the disorder. Further analyses in larger sample sizes with more homogeneous subgroups of ASD participants may be necessary
[114].
Research also suggests that there are group differences between ASD and TD neural responses to serotonin availability. One fMRI study found that increased serotonin (via a selective serotonin reuptake inhibitor) was associated with sustained neural activation in emotion-related brain regions during a negative facial emotion-processing task in ASD adults compared to neurotypical controls, which exhibited an expected habituation response
[115]. The authors concluded that homoeostatic control of these regions is altered by serotonin in ASD. Two other fMRI studies found that acute tryptophan depletion (which lowers serotonin levels) was associated with deficits in facial affect
[116] and inhibitory processing
[117] in ASD. These findings may help explain some of the social and restricted/repetitive behaviors that are hallmarks of the disorder
[116][117][116,117]. Further, many of the brain regions found to be modulated by the tryptophan depletion have previously been reported to have serotonergic abnormalities in people with ASD (e.g., synthesis differences, lower numbers of receptors and transporters)
[116].
Despite the extensive evidence showing alterations in the tryptophan pathway and the serotonergic system in ASD, there is no consensus in current literature on tryptophan availability and its influence on behavior in patients with ASD. While studies have suggested that reduced levels of tryptophan
[118][119][120][118,119,120] and decreased tryptophan metabolism are prominent features of ASD
[121], a recent study showed both increased and decreased levels of tryptophan in ASD youth
[122], which, again, may be related to ASD heterogeneity and marks the need for ASD sub-grouping. The potential roles of oxytocin and other upstream markers in ASD have also been discussed
[123] and it is likely that interactions between oxytocin and serotonin may further influence the pathophysiological processes of ASD
[124]. More research is necessary to fully understand the complex neurochemical processes underlying the relationship between the BGM system, tryptophan metabolism, and ASD. Below, we discuss specific findings in BGM connections at the structural and functional levels and how changes in the microbiome impact neural networks.