The Maternal Microbiome as a Map to Understanding the Impact of Prenatal Stress on Offspring Psychiatric Health

Abstract

Stress and psychiatric disorders have been independently associated with disruption of the maternal and offspring microbiome and with increased risk of the offspring developing psychiatric disorders, both in clinical studies and in preclinical studies. However, the role of the microbiome in mediating the effect of prenatal stress on offspring behavior is unclear. While preclinical studies have identified several key mechanisms, clinical studies focusing on mechanisms are limited. In this review, we discuss 3 specific mechanisms by which the microbiome could mediate the effects of prenatal stress: 1) altered production of short-chain fatty acids; 2) disruptions in TH17 (T helper 17) cell differentiation, leading to maternal and fetal immune activation; and 3) perturbation of intestinal and microbial tryptophan metabolism and serotonergic signaling. Finally, we review the existing clinical literature focusing on these mechanisms and highlight the need for additional mechanistic clinical research to better understand the role of the microbiome in the context of prenatal stress.

Pregnancy is a critical time during which insults can perturb fetal development and shape offspring long-term health outcomes (1,2). Specifically, maternal mental health during pregnancy is associated with alterations in offspring neurodevelopment (3). There is evidence that excessive perceived stress on top of genetic predisposition increases the risk for development of psychiatric disorders in the offspring (4-6). Proposed mechanisms by which maternal stress or psychiatric disorders can lead to disruptions in offspring behavior and mental health include epigenetic changes, disturbances of the hypothalamic-pituitary-adrenal axis, inflammation, and, most recently, alterations in the maternal microbiome (7-11).

The gut microbiome refers to the collection of microbes or microbiota that reside within the host intestinal tract as well as their functional output, including genetic elements and metabolites (12). The gut microbiome reflects aspects of an individual’s health and can be heavily influenced by diet and pregnancy status (13-15). The gut microbiome is essential for the host, as microbes metabolize complex carbohydrates for digestion, protect the host from pathogenic microbes, promote immune development, and provide essential vitamins and metabolites (16). Gut microbes communicate with the brain through factors such as host immune mediators and microbial metabolites, comprising the microbiota-gut-brain axis (17). The microbiota-gut-brain axis has been shown to be altered by psychiatric illness and chronic stress, including during pregnancy (8,18-24). Furthermore, changes to the maternal microbiome during pregnancy have lasting impacts on offspring neurodevelopment and behavior (25,26).

Clinical studies have demonstrated associations between psychosocial stress during pregnancy and changes to the maternal and offspring microbiome (27-35), with a focus on vertical transmission of perturbed microbial communities to the offspring as a primary mechanism of impacting neurodevelopment. However, there have been notable inconsistencies in specific microbiota that are impacted across the studies. Preclinical studies have explored additional mechanisms by which stress can alter the microbiota-gut-brain axis during pregnancy, including through microbial metabolites such as short-chain fatty acids (SCFAs) or tryptophan metabolites and through interactions with the immune system. The clinical studies examining these mechanistic underpinnings are limited.

Here, we first review the clinical studies on stress during pregnancy, the microbiota-gut-brain axis, and childhood psychiatric outcomes. We then review preclinical studies identifying potential mechanisms by which the microbiota-gut-brain axis can mediate impacts of maternal stress on offspring neurodevelopment. Finally, we highlight initial clinical work examining these mechanisms that supports the need to further delve into these mechanisms in clinical studies.

CURRENT CLINICAL RESEARCH ON STRESS DURING PREGNANCY, THE MICROBIOTA-GUT-BRAIN AXIS, AND CHILDHOOD BEHAVIORAL OUTCOMES: FOCUS ON VERTICAL TRANSMISSION OF MICROBES

Research on the impact of perceived stress during pregnancy on the maternal gut microbiome provided primarily associative rather than causal evidence. For example, a Dutch cohort found no association of measures of general stress and composition of the maternal gut microbiota during late pregnancy; however, several taxa at the genus level were found to contribute to a model predicting general anxiety during pregnancy (27). A study using two sociodemographically distinct cohorts in the United States investigated associations between perceived stress, measured by the Perceived Stress Scale, and the gut microbiota during the second and third trimesters (28). While alpha and beta diversity did not significantly differ based on levels of total maternal perceived stress, specific taxa were found to be associated with perceived stress. Dividing the Perceived Stress Scale into two factors (emotional distress and self-efficacy) showed that distress was associated with more pathogenic taxa, and self-efficacy was associated with more beneficial taxa. Bacteroides uniformis was positively associated with perceived stress and negatively associated with self-efficacy (28). B. uniformis has been hypothesized to be beneficial in some studies and pathogenic in others, with the presence of dietary fiber as a possible contributing factor (36,37). Self-efficacy was associated with increased alpha diversity during the first and second trimesters in another study (38). At the postpartum time point, perceived stress was negatively correlated with alpha diversity (38). Findings from nonpregnant individuals suggest that low alpha diversity of gut microbiota is correlated with markers of poor glucose tolerance and inflammation (39-41). Similarly, in pregnancy, decreased microbial diversity may be associated with these same markers of poor health, and, conversely, individuals with less perceived stress and greater microbial diversity could have healthier pregnancies.

Recent studies have also demonstrated correlations between maternal stress and the infant microbiome. This highlights one potential mechanism by which maternal stress impacts offspring health outcomes—through vertical transmission of maternal microbes to the infant. Microbes that persist in the neonatal gut in the first few months of life originate from the maternal gut (42). Increased pregnancy-related anxiety was associated with increased diversity in newborn meconium (29). One study demonstrated an association between maternal composite psychosocial stress scores and infant gut microbiome diversity: Infants of mothers with higher levels of stress assessed at delivery had higher alpha diversity at 6 months of age with lower levels of Lactobacillus gasseri and Bifidobacterium pseudocatenulatum (30). A similar study found that several beneficial microbial species of Bifidobacterium and Lactobacillus in the infant stool within the first year of life negatively correlated with maternal anxiety, depression, and stress scores during pregnancy (35). Psychological stress during midgestation and late gestation was associated with an increase in abundance of genera from the Proteobacteria phylum and a decrease in Akkermansia and Lactobacillus in the offspring during the first several months of life (33,34). However, another study demonstrated that maternal stress during the third trimester was positively associated with abundance of Lactobacillus and Bifidobacterium in infant stool samples collected at 1 month of age and negatively associated with infant Bacteroides and Enterobacteriaceae (31). Ruminococcus was less abundant in fecal samples collected at 2 years of age from children of women who had higher levels of self-reported anxiety during the second trimester of pregnancy (32).

Alterations in the maternal and neonatal gut microbiome have also been associated with adverse child behavioral outcomes. Alpha diversity of the maternal fecal microbiome during the third trimester of pregnancy was associated with child internalizing behaviors at 2 years of age, while maternal Lachnospiraceae and Ruminococceae were positively associated with normative behavior in children (43). Additionally, another large-scale study demonstrated that the maternal stool microbiome during the third trimester was more strongly associated with the child’s neurodevelopment at 1 year of age compared with the child’s stool microbiome at 3 to 6 months or 1 year of age (44). Higher maternal perceived stress, depression, or anxiety during the third trimester of pregnancy was associated with increased alpha diversity in the neonate after birth and at 1 month of age (35). Furthermore, meconium Lactobacillus was negatively associated with prosocial behavior and positively associated with hyperactivity at 2 years of age (45). Abundance of an Intestinibacter taxon at 2 years of age was negatively associated with internalizing problems at 4 years of age, while abundance of a Streptococcus taxon was negatively associated with developmental problems at 2 years of age (32). Prevotella in fecal samples at 12 months of age was associated with internalizing problems at 2 years (46), and Bifidobacterium and Streptococcus in stool collected from 2.5-month-old infants were associated with positive emotionality (34). Decreased Bacteroides and increased Veillonella, Dialister, and Clostridiales were associated with increased fear behavior at 1 year of age (47).

Altogether, though these clinical studies reveal associations between maternal stress, changes in the maternal and infant gut microbiome, and offspring behavior, the microbiota identified in these studies are inconsistent and vary widely among the cohorts. This suggests that vertical transmission is unlikely to be the sole mechanism underlying the relationship between maternal stress and offspring psychiatric and behavioral outcomes and indicates the need to investigate additional mechanisms identified in preclinical studies.

PRECLINICAL MODELS OF MATERNAL STRESS AND THE MICROBIOTA-GUT-BRAIN AXIS IN PREGNANCY: POTENTIAL MECHANISMS

Preclinical models have shown that stress during pregnancy leads to depressive- and anxiety-like behavior in pregnant dams. Chronic variable stress, restraint stress, and social isolation during pregnancy have been shown to induce depressive-like and anxiety-like behavior in pregnant rodents (48-51) that is reversible with antidepressant treatment (48-50). Prenatal stress has shown variable behavioral programming effects in the offspring due to the interactions between type of stress, sex, and gestational timing (52-57).

Preclinical models have implicated the microbiome in mediating offspring neurodevelopment. Studies using germ-free (GF) mice, which have developed in the absence of microbes, show that microbes directly impact multiple systems critical to neurodevelopment and stress response. GF mice have an elevated corticosterone response to restraint stress (58) and novel environment stress (59), increased blood-brain barrier permeability (60), increased prefrontal cortex myelination (61), increased hippocampal neurogenesis in adulthood (62), decreased BDNF (brain-derived neurotrophic factor) in the brain (58,59), and a male-specific increase in hippocampal serotonin (59). The microbial-dependent programming of the stress response and serotonin signaling occur during a critical period, as these changes cannot be rescued by colonization of microbes after weaning (58,59), although blood-brain barrier function can be recovered in GF mice after colonization in adulthood (60). Together, these preclinical studies identify how commensal microbes are important for neurodevelopment.

Other promising mechanisms beyond vertical transmission include altered production of microbial metabolites with a focus on SCFAs and tryptophan metabolism and interaction of the microbiome with the immune system. The microbiome plays a key role in training and development of various components of the immune system (63). Furthermore, there is evidence that stress impacts the immune system (64), and further that prenatal stress impacts the maternal and offspring immune system (8,65). Clinical studies have demonstrated that stress increases circulating proinflammatory cytokines in maternal circulation during pregnancy (19,66). While preclinical studies also demonstrate maternal inflammation with stress (67), some studies show decreased maternal inflammation with stress (68). Furthermore, studies in rodents have shown that prenatal stress induces inflammation in the placenta and fetal brain (52-55). Nonetheless, stress impacts the maternal and offspring immune system, with the maternal gut microbiome as a proposed intermediary.

These mechanisms are summarized in Figure 1, which shows that prenatal stress alters the maternal gut microbiome, leading to changes in SCFA production and tryptophan metabolism and modulation of the maternal immune system, specifically through TH17 (T helper 17) cell activation. These factors circulate through the maternal bloodstream and can impact placental tryptophan metabolism and immune function. These factors can alter fetal brain development through actions of SCFAs and indoles, a tryptophan metabolite, on microglia, the innate immune cell of the brain, and astrocytes as well as by direct actions on serotonergic neurons or cytotoxic effects of quinolinic acid through the NMDA receptor (NMDAR). These specific mechanisms are discussed in more detail in the following sections.

Figure 1.

Summary of preclinical research depicting how prenatal stress may influence fetal neurodevelopment via maternal microbes, microbe-related metabolites, and immune factors. Prenatal stress has been linked to alterations in the maternal gut microbial community and causes placental and fetal brain inflammation and placental metabolism dysfunction through microbe- and CCL2-dependent mechanisms. In the maternal gut, microbes also drive TRP metabolism, produce SCFAs, and interact with the TH17 component of the host immune system. Microbial metabolites borne through maternal circulation further affect fetal neurodevelopment by directly interacting with AhRs, serotonergic neurons, and NMDA receptors in the developing brain; alternatively, SCFAs do so by diffusing directly into microglia. Postnatally, maternal gut microbes are directly transferred to the offspring, forming the basis of its own microbiome. 5-HT, serotonin; AhRs, aryl hydrocarbon receptors; EC, enterochromaffin; KA, kynurenic acid; KYN, kynurenine; QA, quinolinic acid; SCFAs, short-chain fatty acids; TH, T helper; TRP, tryptophan.

Dietary Fiber and SCFA Production Are Important in Microglia Development

SCFAs have emerged as an important potential mechanism by which the microbiome mediates the impact of maternal stress on offspring neurodevelopment. The microbiome is responsible for the production of several bioactive metabolites, including SCFAs derived from microbial fermentation of dietary fibers. SCFAs mostly consist of acetate, propionate, and butyrate, which are readily transported into the circulation and brain parenchyma. Furthermore, SCFAs are absent in GF mice, indicating the importance of microbes in their production (69).

In the context of pregnancy, SCFAs can impact the brain both prenatally and postnatally, as maternal microbes produce SCFAs that can enter the maternal circulation and reach the fetal brain during pregnancy (70). Early pregnancy stress resulted in fewer early colonizers of the offspring microbiome with the ability to produce SCFAs (71). Chronic restraint stress in rodents decreases SCFAs in the stool (72), though social defeat stress increases stool SCFAs, specifically acetate (73). However, this study also showed that SCFA supplementation ameliorated social defeat stress–induced deficits in rewardseeking behavior, suggesting a potentially beneficial effect of SCFAs.

SCFAs can also affect the developing fetus through impacts on developing microglia. Microglia are the resident innate immune cells of the central nervous system, and evidence suggests that they have unique neuromodulatory functions in development and adulthood (74,75). Microglia help facilitate fetal brain development through regulation of neural precursor cells (76), modulation of cortical interneuron migration and wiring of inhibitory circuits (77,78), facilitating axonal guidance (77,79), regulating synapse formation (80,81), and supporting proper myelination (82). Microglia acquire the ability to respond to environmental signals during embryonic development, and the absence of maternal microbial metabolites perturbs their function and maturation (83). Male fetuses from GF dams have increased microglia density during the perinatal period (84) and into adulthood (85). While microglia do not express SCFA receptors, SCFAs can diffuse into microglia to regulate gene expression through inhibition of histone deacetylase activity (86). Supplementation of GF mice with SCFAs largely rescues the increased microglia proliferation and their perturbed morphology (85). Supplementing mice with acetate recapitulates this amelioration of microglia density and morphology in GF mice (87). Acetate supplementation is capable of restoring normal mitochondrial function and reversing most of the differentially expressed genes seen between microglia isolated from GF and conventionally colonized mice (87).

Commensal microbes are also necessary for microglia to mount an appropriate response to environmental signals. Mice born from GF dams fail to increase the expression of inflammatory cytokines in microglia that is seen in conventionally colonized mice after birth (84,88). Further, neonatal GF mice develop increased cerebellar synapses during the postnatal period with changes to microglial function, suggesting decreased microglial synaptic pruning (88). Similarly, GF microglia in adulthood have a blunted cytokine response to the bacterial endotoxin lipopolysaccharide (85). Acetate supplementation is capable of reversing most of the differentially expressed genes between microglia isolated from GF mice and conventionally colonized mice (87). Promoting SCFA production through increased dietary fiber also decreases the lipopolysaccharide-induced inflammatory response in microglia (86). Together, these studies suggest that SCFAs could mediate the effect of maternal stress on offspring neurodevelopment by perturbing developing microglia.

Microbial Impacts on TH17 Are Key to Understanding the Balance Between Protection Against Pathogens and Tolerance of Commensal Bacteria

In addition to influencing microglial development, maternal microbial metabolites also play a key role in helping the maternal immune system tolerate the fetus and in regulating maternal intestinal and peripheral immune cells. Specifically, TH17 cells are a subset of T cells that produce the cytokine interleukin 17 (IL-17; also known as IL-17A) to induce inflammation and protect mucosa from pathogens (89,90). Furthermore, TH17 cells primarily reside in the terminal ileum mucosa and depend on intestinal macrophages and commensal microbes to induce their differentiation (91). Additionally, a careful balance of TH17 cells and a different subset of antiinflammatory T cells, regulatory T cells, is necessary for maternal immune tolerance of the fetus during pregnancy (92). As in vitro studies show that SCFAs such as butyrate induce T cells to become regulatory T cells rather than TH17 cells (93), alterations to the maternal microbiome and metabolites could impact the fetus through interacting with the maternal immune system.

Notably, animal models have shown that IL-17 is a factor in disrupted fetal brain development, though the majority of the data are derived from maternal immune activation (MIA) instead of prenatal stress models. MIA in mice with the viral mimetic poly(I:C) (polyinosinic:polycytidylic acid) increased IL-17A in the maternal serum (94). Furthermore, MIA induced abnormal patches of fetal brain cortical organization and perturbed offspring social behavior, which was reversed by blocking maternal IL-17A signaling or inhibiting differentiation of maternal TH17 cells (94). These cortical and behavioral effects were recapitulated with intraventricular injection of recombinant IL-17A into the fetal brain.

The differentiation of intestinal TH17 cells is largely dependent on segmented filamentous bacteria (SFB) (95). Mice of the C57BL/6 strain from The Jackson Laboratory (JAX) are devoid of SFB, resulting in minimal intestinal TH17 cells compared with mice from other vendors that have SFB. This deficiency in TH17 cells in JAX mice can be reversed with SFB colonization (95). Increased maternal IL-17A and fetal brain cortical disruptions and behavioral deficits are absent in JAX pregnant dams that undergo MIA. Additionally, colonizing these JAX mice with SFB before MIA recapitulates the fetal brain disruption (96). Adult JAX mice are mostly resistant to acquiring learned helplessness, but colonizing these mice with SFB leads to increased helplessness (97). These results suggest that gut colonization by SFB is necessary for TH17 cell–mediated effects in MIA offspring and in learned helplessness models of depression. However, it is also important to consider the impact of attenuating TH17 cell function on gut immunity, as JAX mice lacking SFB and TH17 cells are more susceptible to infection (95). In pregnancy, TH17 cells are increased in the decidua of the uterus, and this is thought to serve a role in protecting against pathogenic microbes, though excessive TH17 cells could contribute to inflammation for mother and fetus (98,99).

Further studies are needed to determine what SFB are beneficial versus pathogenic and when TH17 cell responses are required or excessive in the context of prenatal stress. Pregnancy is particularly challenging, as the immune system must be suppressed to allow fetal growth during the second trimester (100). Furthermore, the third trimester requires tight regulation, as increasing inflammation is required for delivery but excessive inflammation may result in obstetric complications such as preeclampsia (101).

Microbial Metabolism of Tryptophan and Serotonergic Signaling as Key Regulators

A third potential mechanism by which the maternal microbiome could mediate the effects of prenatal stress on neurodevelopment is through altered maternal intestinal metabolism of tryptophan, with downstream impacts on the developing offspring. Tryptophan comes from the maternal diet, and the majority (95%–99%) is metabolized along the kynurenine pathway, while only 1%–2% is shunted through the serotonin pathway (102). During stress, proinflammatory cytokines increase both IDO (indoleamine 2,3-dioxygenase) and TDO (tryptophan 2,3-dioxygenase) (103-105), which catalyze tryptophan to kynurenine. Therefore, the small proportion of tryptophan that is metabolized along the serotonin pathway at baseline is further reduced by stress. Prenatally stressed mice also have microbes driving tryptophan metabolism (71). Maternal production of IL-17A also perturbs host tryptophan metabolism in the pregnant dam and in the offspring (106).

Within the kynurenine pathway, tryptophan is metabolized to kynurenine, which is then converted to either kynurenic acid or quinolinic acid (107). Quinolinic acid is a neuroactive agonist of glutamate NMDAR, and excessive activation of the NMDAR is neurotoxic (108). Quinolinic acid also inhibits the uptake of glutamate by astrocytes (108), which leads to underactivity of the prefrontal cortex, with diminished reward and cognitive impairment (109). When inflammation is fleeting, the system can be shunted back to production of kynurenic acid, which attenuates the agonist activity of the NMDAR and decreases glutamate levels (110,111).

In addition to intestinal metabolism of tryptophan, there is evidence that gut microbes alter tryptophan metabolism in the context of prenatal stress. Prenatal restraint stress reduced tryptophan-metabolizing taxa Parasutterella and Bifidobacterium in both the maternal colon and the postnatal offspring colon. Restraint stress also altered expression of genes related to tryptophan metabolism, tryptophan uptake from the intestinal tract, and tryptophan signaling in the pregnant dams (112). Prenatal restraint stress increased placental tryptophan and serotonin concentrations and was associated with placental and fetal brain inflammation. Microbes are critical to this relationship, as GF mice exposed to prenatal stress did not have increases in placental tryptophan or serotonin (52).

Altered tryptophan metabolism then plays a role in serotonergic signaling by increasing or decreasing peripheral serotonin production in the host and by altering serotonin availability (113). Other key considerations include the presence and level of activity of the bacterial β-glucuronidase enzyme in the intestine, which can alter the amount of free serotonin in the blood (114). Serotonin production and signaling are especially important in the context of stress and psychiatric disorders, as selective serotonin reuptake inhibitors (SSRIs) are frequently prescribed to treat depression and anxiety disorders, including during pregnancy. One study showed that although administration of fluoxetine, an SSRI, during pregnancy did not alter the maternal gut microbiome in mice, pretreatment of the mice with antibiotics to deplete the microbiome modified the fetal brain transcriptional response to maternal fluoxetine treatment (115). Interestingly, in a model of pregnancy following early life stress, fluoxetine treatment modulated gut microbiome community structure and fecal amino acid concentrations during pregnancy and lactation in mice exposed to the stressor (116). Not only are SSRIs clinically relevant, but also their use in preclinical studies shows complex interactions between the maternal microbiome and offspring outcomes.

Microbes can also metabolize tryptophan into an indole, which can enter circulation and cross the blood-brain barrier (69). Indoles modulate immune function through the aryl hydrocarbon receptor (AhR), a transcription factor expressed in several populations of immune cells, including microglia. AhR controls immune cell function through directly inhibiting nuclear factor-κB. Consequently, loss of AhR function in microglia impairs their ability to regulate activation of proinflammatory mediators and minimize damage to the central nervous system (117). Further, AhR signaling in astrocytes is important for attenuating central nervous system inflammation (117). Nontargeted metabolomic analysis comparing GF and specific pathogen–free dams has revealed distinct metabolomic profiles in the absence of microbes including lower indole and 3-indolepropionic acid in the fetal brain, fetal intestine, and the placenta (70). Together, these data suggest a critical interplay between stress, intestinal and microbial metabolism of tryptophan, and the immune system, which may shape offspring behavior.

CLINICAL MICROBIOME MECHANISTIC STUDIES AND CLINICAL TRIALS: VERY EARLY STAGES

While limited, there is some evidence from clinical studies of the mechanisms identified in preclinical studies. IL-17A has been associated with pregnant women in the third trimester who have severe depression and anxiety (118) as well as in children with autism spectrum disorder (119). A combination of cytokine and tryptophan markers in maternal serum during the second trimester predicted the development of depression during the third trimester (120). When the cytokine IL-6 was combined with quinolinic acid, kynurenine, and the kynurenine-to-tryptophan ratio from the second trimester, both severity of depressive symptoms and higher likelihood of major depression could be predicted (120). The chemokine CCL2 was associated with higher scores of anxiety symptoms in both the anxiety-alone group and the anxiety and comorbid depression group during the third trimester (121). Additionally, maternal fecal metabolites analyzed during the third trimester were associated with neurodevelopment in children at 1 year of age (44).

Some evidence can also be gleaned from research of commercially available probiotics, such as Bifidobacterium and Lactobacillus, though their use has yielded mixed results in studies to treat depressive symptoms and to improve mood (122-124). In a double-blind randomized pilot trial, pregnant women who were administered a probiotic containing species of Bifidobacterium, Lactobacillus, and Lactococcus during the third trimester had no differences in depressive symptoms, anxiety, stress, and maternal bonding than women who were administered a placebo (125). While an in-depth analysis of probiotics is outside the scope of this review, it is worth noting that patients who received the probiotic Lactobacillus plantarum concomitantly with an SSRI had lower kynurenine concentrations and improved cognitive functioning compared with patients receiving a placebo with an SSRI (126). Furthermore, it is important to note that improving maternal mental health is a critical area of research independent of its impact on fetal programming. Figure 2 provides a theoretical model of the interplay of the mechanisms described in the context of prenatal stress and how they may mediate offspring neurodevelopment, with the aim of providing guidance for future clinical studies.

Figure 2.

Theoretical illustration of microbial community composition, immune factors (particularly the TH17 arm of the immune system), and microbe-related metabolites that may be important components to the microbiota-gut-brain axis in the third trimester of pregnancy for further study and might differentiate individuals as follows: 1) an individual with low stress and no depression or anxiety with a diverse microbiome; 2) an individual with high stress that is buffered by a diverse microbial community and high fiber intake; and 3) an individual at a high risk of maternal depression and anxiety in the presence of elevated and chronic toxic stress, low microbial diversity, and low fiber intake. 5-HT, serotonin; CNS, central nervous system; KA, kynurenic acid; KYN, kynurenine; NMDAR, NMDA receptor; QA, quinolinic acid; SCFAs, short-chain fatty acids; TH, T helper; TReg, T regulatory; TRP, tryptophan.

FUTURE DIRECTIONS AND CONCLUSIONS

Overall, we have reviewed the current state of clinical research on maternal stress, the gut microbiome, and offspring behavioral outcomes. Furthermore, we have identified key mechanisms identified in preclinical studies that may explain how the microbiome may mediate the effects of maternal stress on offspring neurodevelopment, including 1) altered production of SCFAs, 2) interaction with TH17 cells, and 3) metabolism of tryptophan and serotonin signaling. Given the inconsistencies evident in the current literature in identifying specific microbes that are impacted by stress during pregnancy, future studies may benefit from focusing on mechanisms in addition to specific microbes. For example, future studies could investigate the effects of dietary fiber or SCFA supplementation in pregnant women exposed to psychological stressors on behavioral outcomes in their children. Additional metabolomic studies are also warranted to examine the impact of stress on tryptophan metabolites in the mother and child. Ultimately, this research is also necessary to help elucidate potential targets for intervention during pregnancy to prevent adverse outcomes in the children.