Maternal IL-17A in autism

I. Immune dysregulation in autism spectrum disorder

Autism spectrum disorder (ASD) constitutes a group of neurodevelopmental disorders with overlapping but variable behavioral and cognitive symptoms, severity, and comorbid conditions. The core diagnostic features of ASD are impaired social interaction, communication challenges, and expression of repetitive or perseverative behaviors (Kanner, 1943; Moldin et al., 2006). Although ASD has a strong genetic basis, its etiology is complex, with several genetic factors likely to be involved as well as interactions with environmental influences (Betancur, 2011; DiCicco-Bloom et al., 2006; Goines and Ashwood, 2013; Risch et al., 1999). ASD therefore represents a collection of conditions with heterogeneous causes, but they may converge on common molecular pathways to bring about the features that characterize this disorder. Immune dysregulation has gained significant attention as a pathway for the pathogenesis of ASD (Depino, 2013; Korvatska et al., 2002).

Immune system abnormalities have been observed widely in the brain and periphery of ASD individuals. Studies have found chronic neuroinflammation in ASD, indicated by increased activation of microglia and astrocytes and production of cytokines and chemokines in the brain (Li et al., 2009; Morgan et al., 2010; Vargas et al., 2005). Transcriptome analyses of ASD brains also have indicated an upregulation of immune response genes (Garbett et al., 2008; Voineagu et al., 2011; Ziats and Rennert, 2011). Similarly, in the periphery, studies have found elevated expression of pro-inflammatory cytokines and other immune factors in the blood and gastrointestinal tract of ASD individuals (Ashwood et al., 2004; Ashwood et al., 2011; Enstrom et al., 2010). Additionally, ASD has been associated with inflammatory disorders and autoimmunity in not only the affected individuals but also their mothers (Atladottir et al., 2009; Buie et al., 2010; Jonakait, 2007). Moreover, prenatal exposure to maternal immune activation (MIA) has been implicated as an environmental risk factor for ASD.

Several epidemiological studies have found that maternal infection during pregnancy correlated with an increased frequency of ASD in the children (Atladottir et al., 2010; Brown et al., 2014; Chess, 1977; Wilkerson et al., 2002). The infection activates a maternal immune reaction in which the molecular signaling cascades are thought to adversely affect neurodevelopment and subsequent behavior manifested as ASD in the offspring. Supporting this idea, animal models have shown that MIA results in offspring with behavioral, neurological, and immunological abnormalities like those observed in ASD (Abrahams and Geschwind, 2010; Depino, 2013). For example, offspring of pregnant mice that had been exposed to viral infection or injected with synthetic double-stranded RNA poly(I:C), mimicking viral infection, exhibited ASD-like behaviors. Notably, they demonstrate abnormal ultrasonic vocalization mirroring communication challenges in ASD, reduced social approach mirroring impaired social interaction in ASD, and increased marble burying mirroring repetitive behaviors in ASD (Choi et al., 2016; Malkova et al., 2012; Schwartzer et al., 2013; Shi et al., 2003; Smith et al., 2007). The offspring have been reported also to exhibit ASD-like alterations in cortical and cerebellar growth and function (Choi et al., 2016; De Miranda et al., 2010; Meyer et al., 2008; Shi et al., 2009) as well as in immune responses (Arrode-Bruses and Bruses, 2012; Garay et al., 2013; Hsiao et al., 2012; Meyer et al., 2006; Meyer et al., 2008). How can MIA exposure lead to ASD in susceptible individuals?

II. Induction of inflammation

A. Modeling MIA in mice

To model MIA as a risk factor for ASD, inflammation has been induced in pregnant mice using several approaches, including direct infection or exposure to lipopolysaccharide (LPS) to mimic bacterial infection or poly(I:C) to mimic viral infection (Choi et al., 2016; Oskvig et al., 2012; Shi et al., 2003; Shi et al., 2009; Smith et al., 2007). The resulting phenotypes in the offspring depend on the trigger and the degree and temporal window that prenatal inflammation is induced. One frequently used protocol is a high single-dose intraperitoneal injection of poly(I:C) in pregnant mice on embryonic day (E)12.5 (Choi et al., 2016; Schwartzer et al., 2013; Shi et al., 2009; Smith et al., 2007). This gestational stage resembles the late first trimester in humans (Clancy et al., 2007), during which viral infections have been correlated with increased incidence of ASD in the offspring (Atladottir et al., 2010). Poly(I:C) evokes a pro-inflammatory antiviral response in the mother that is similar to the immune response occurring after activation of Toll-like receptor 3 (TLR3) by viral infection (Alexopoulou et al., 2001; Datta et al., 2003; Traynor et al., 2006). Mimicking MIA during viral infection in this way results in offspring that demonstrate many ASD-like phenotypes, as described above.

B. IL-6 in inflammation and ASD

The pro-inflammatory cytokine interleukin-6 (IL-6) has been identified as a causal factor in ASD-like phenotypes associated with MIA (Samuelsson et al., 2006; Smith et al., 2007). Increased expression of IL-6 is often noted in ASD (Ashwood et al., 2011; Li et al., 2009; Vargas et al., 2005), and animal models suggest that this increase is pathogenic in MIA-associated ASD. In the poly(I:C) induction model of MIA, ASD-like phenotypes in the offspring can be prevented if pregnant mice are deficient in IL-6, implicating the necessity of IL-6 signaling (Smith et al., 2007). Importantly, injecting pregnant mice with IL-6 alone was sufficient to recapitulate many of the ASD-related behavioral and immunological features in offspring exposed to MIA (Garbett et al., 2012; Smith et al., 2007). Therefore, MIA induces maternal IL-6, which can cross the placenta (Dahlgren et al., 2006; Hsiao and Patterson, 2011) and has been found upregulated in the fetal brain (Meyer et al., 2006; Wu et al., 2015). Besides its function in inflammation, IL-6 has a known role in normal neurodevelopment, including regulation of neural cell proliferation, differentiation, migration, axonal pathfinding, and synaptic connections (Deverman and Patterson, 2009; Islam et al., 2009). The presence of higher IL-6 levels due to MIA can therefore directly alter fetal brain development that results in ASD. However, IL-6 is a pleiotropic signal capable of triggering the expression of other cytokines and immune regulatory genes that may additionally impact the developing brain or are responsible for precipitating ASD. Investigation into the molecular pathways through which IL-6 may act to promote ASD in the offspring following MIA has revealed an important contribution from interleukin-17A (IL-17A) signaling (see Figure 1).

Figure 1.

IL-17A and immunity: Infection causes increased expression and secretion of the proinflammatory cytokine interleukin-6 (IL-6). In the presence of the cytokine transforming growth factor β (TGFβ), IL-6 can stimulate the differentiation of naïve CD4⁺ lymphocytes into T helper 17 (Th17) cells. A key downstream effector of IL-6 involved in Th17 differentiation is the transcription factor retinoic acid receptor-related orphan receptor γt (RORγt). RORγt promotes transcription of the pro-inflammatory cytokine IL-17A. IL-17A signaling plays a protective role in adaptive immunity. Dysregulation of Th17 cells and IL-17A production are associated with autoimmune diseases and inflammatory disorders.

IL-17A and ASD: In the case of maternal immune activation (MIA), IL-6 is induced and stimulates the differentiation of Th17 cells in the mother, leading to increased IL-17A secretion. IL-17A in turn crosses the placental barrier or is induced in the fetus, where it can act on cells expressing the receptor IL-17RA in the developing nervous system. Stimulation of IL-17RA can activate several intracellular signaling pathways mediated by Act1, such as NF-κB, ERK, p38 MAPK, and TRAF2/5. Consequently, various developmental processes can be altered, including neural stem cell (NSC) proliferation, gliogenesis, microglial function, blood-brain barrier activity, and neuronal connectivity. In these ways, inappropriate Th17 cell and IL-17A activity may adversely impact prenatal development, resulting in cortical dysplasia and the manifestation of behaviors associated with ASD in the offspring (Choi et al. 2016).

In the peripheral immune system, major downstream effectors of IL-6 are the transcription factors signal transducer and activator of transcription 3 (STAT3) and retinoic acid receptor-related orphan nuclear receptor γt (RORγt). RORγt is expressed exclusively in lymphoid cells that secrete the pro-inflammatory cytokine IL-17A (Ivanov et al., 2006; Zhou et al., 2007). Most prominent are the Th17 cells, a subset of CD4⁺ T helper (Th) lymphocytes that produce large quantities of IL-17A to mediate inflammatory processes (Ivanov et al., 2006). IL-6 drives RORγt activity in naïve CD4⁺ T cells, which is required for their differentiation into the Th17 subset (Zhou et al., 2007; Zhou et al., 2008). Following infections or stimulation by poly(I:C), IL-6 is released early in the acute phase of the immune response to downregulate immunosuppressive regulatory T (Treg) cells and promote Th17 cells (Bettelli et al., 2006; Jonakait, 2007). In the presence of the immunoregulatory cytokine transforming growth factor β (TGFβ), naïve CD4⁺ T cells co-express RORγt and the transcription factor forkhead box P3 (FOXP3), which inhibits RORγt activity (Zhou et al., 2008). FOXP3 is required for generating Treg cells, but IL-6 suppresses FOXP3 expression and therefore limits Treg differentiation while relieving inhibition of RORγt (Bettelli et al., 2006; Zhou et al., 2008). Additionally, IL-6 activates STAT3 in naïve CD4⁺ T cells to induce further expression of RORγt, which in turn cooperates with STAT3 to induce IL-17A expression (Chen et al., 2009; Manel et al., 2008; Voo et al., 2009). RORγt and IL-17A are gene expression signatures of the Th17 lineage. Thus, the combination of IL-6 with TGFβ directs T cell development into the Th17 subset. To maintain and amplify the pool of Th17 cells, IL-6 also induces STAT3- and RORγt-dependent expression of IL-1 receptors by these cells, making them responsive to IL-1 (Chung et al., 2009; Ikeda et al., 2014). IL-1 reinforces RORγt expression and downregulates FOXP3 expression, thereby enhancing Th17 differentiation (Chung et al., 2009; Ikeda et al., 2014). STAT3 also induces Th17 production of IL-21, which has an autocrine positive feedback effect on STAT3 activity (Korn et al., 2007; Wei et al., 2007; Zhou et al., 2007). Concurrently, RORγt induces Th17 expression of IL-23 receptors for stabilizing the commitment of these cells to the Th17 lineage (Stritesky et al., 2008). IL-23, enhanced by IL-1, then stimulates the terminal differentiation of Th17 cells through maintenance of lineage signature gene expression and activation of effector genes such as IL-22 in mature Th17 cells (Chung et al., 2009; Stritesky et al., 2008).

C. Th17 cells

Th17 cells have pro-inflammatory actions that are normally beneficial to host defense (Cai et al., 2016; Chung et al., 2003; Conti et al., 2009; Huang et al., 2004; Huppler et al., 2014; Korn et al., 2009; Ye et al., 2001). Through the secretion of effector cytokines, Th17 cells potently induce tissue inflammation as a protective response against infection. IL-17A, also known as IL-17, is the signature cytokine produced by Th17 cells. In turn, IL-17A promotes the production of cytokines such as granulocyte-macrophage colony-stimulating factor (GM-CSF), granulocyte colony-stimulating factor (G-CSF), IL-6, IL-1β, and tumor necrosis factor α (TNFα), and chemokines such as CCL20 and CXCL1 (or GROα) from different cell types to recruit neutrophils and mediate inflammation (Fossiez et al., 1996; Huang et al., 2007; Jones and Chan, 2002; Kawaguchi et al., 2001; Laan et al., 1999; Laan et al., 2001; Moseley et al., 2003). In addition to IL-17A, Th17 cells produce IL-17F, another member of the IL-17 cytokine family with the highest homology to IL-17A (Hymowitz et al., 2001; Liang et al., 2007). IL-17A and IL-17F can be secreted as homodimers and heterodimers which have overlapping but distinct proinflammatory functions and differing potency, with more strength and effects attributed to IL-17A (Ishigame et al., 2009; Liang et al., 2007; Wright et al., 2007; Yang et al., 2008). As described earlier, other Th17 effectors are IL-21, which has a self-amplifying effect, and IL-22, which has paracrine effects such as remodeling of endothelial and epithelial barriers, involved in inflammation (Aggarwal et al., 2001; Caruso et al., 2007; Dumoutier et al., 2000; Monteleone et al., 2006). Interestingly, Th17 cells can themselves secrete IL-6 (Langrish et al., 2005), which can result in a positive feedback loop, as well as GM-CSF (Codarri et al., 2011) and TNFα (Langrish et al., 2005), which can synergize with IL-17A to enhance the inflammatory response. Paradoxically, Th17 cells have been reported also to co-express interferon γ (IFNγ), a proinflammatory cytokine that has classically been the signature effector of Th1 lymphocytes and inhibits Th17 differentiation (Acosta-Rodriguez et al., 2007; Chen et al., 2007; Wilson et al., 2007). These Th17 cells with simultaneous expression of IL-17A and IFNγ have been associated with greater resistance to Treg regulation and greater pathogenic potential (Annunziato et al., 2007).

Inappropriate Th17 activity has been linked to the pathogenesis of several autoimmune and inflammatory disorders, including rheumatoid arthritis, asthma, and multiple sclerosis (MS) (Korn et al., 2009). Uncontrolled Th17 activation can lead to excessive IL-17A levels and a hyperinflammatory state in the periphery. In the brain, IL-17A has been shown to disrupt the blood-brain barrier, facilitating infiltration of peripheral factors and even Th17 cells that injure the central nervous system in MS (Kebir et al., 2007). IL-17A is produced by other immune cells as well, such as neutrophils in the periphery (Ferretti et al., 2003) and microglia in the brain (Kawanokuchi et al., 2008). Because these cells can also be stimulated by IL-17A, a propagative cycle could result from an interaction with Th17-released IL-17A. Under normal brain conditions, IL-17A can induce neuroinflammation as a protective response by activating glia to produce inflammatory cytokines and chemokines (Hu et al., 2014; Kawanokuchi et al., 2008). However, at high levels, IL-17A has been associated with brain pathology not only in MS and experimental autoimmune encephalomyelitis (EAE), a mouse model for MS, but also in epilepsy and ischemia (Das Sarma et al., 2009; He et al., 2013; Kawanokuchi et al., 2008; Kebir et al., 2007; Komiyama et al., 2006; Wang et al., 2009). Dysregulation of Th17 cells and subsequent IL-17A signaling has also been implicated in ASD pathogenesis (Figure 1).

III. The role of IL-17A signaling in ASD

A. Th17 cells and IL-17A in ASD

Accumulating evidence support a role for Th17 cells and their effector cytokine IL-17A in ASD. The gene for IL-17A was identified in a genome-wide analysis to have enriched copy number variants associated with ASD (van der Zwaag et al., 2009). In subsets of ASD individuals, IL-17A has been found at elevated levels in the blood and correlated with the severity of behavioral symptoms (Akintunde et al., 2015; Al-Ayadhi and Mostafa, 2012; Suzuki et al., 2011). In the MIA mouse model of ASD, poly(I:C) treatment has been found to increase IL-17A levels in maternal blood and the postnatal brain (Choi et al., 2016; Garay et al., 2013). Poly(I:C) treatment also increases mRNA levels of IL-17A in the placenta (Choi et al., 2016). Correspondingly, lymphocytes isolated from the placenta and offspring of poly(I:C)-treated mothers differentiate significantly more into Th17 cells or secrete more IL-17A upon stimulation (Choi et al., 2016; Hsiao et al., 2012; Mandal et al., 2010; Mandal et al., 2011). To determine whether this concordant increase in IL-17A expression may be symptomatic or pathogenic in ASD, a recent study inhibited IL-17A signaling using antibody blockade of the cytokine in poly(I:C)-treated pregnant mice and found that ASD-like phenotypes in the offspring were prevented (Choi et al., 2016). Specifically, mouse behaviors modeling the core diagnostic features of ASD and cortical malformation mirroring abnormal cortical growth in ASD individuals (Casanova et al., 2013; Stoner et al., 2014) were normalized in the MIA-exposed offspring. The source of aberrant IL-17A levels was ascribed to maternal Th17 cells because genetic removal of the transcription factor RORγt selectively from CD4⁺ T cells in pregnant mice, thereby suppressing Th17 differentiation, also was able to prevent these ASD-like phenotypes in MIA-exposed offspring (Choi et al., 2016). Thus, IL-17A dysregulation may play a causal role in the development of ASD (Figure 1).

B. IL-17A pathway in MIA-associated ASD

IL-17A sources during pregnancy

Immune challenges in the mother are relayed to the fetus via the placenta. The finding that mRNA levels of IL-17A are elevated in the placenta in response to poly(I:C)-evoked MIA suggests that immune cells are activated within the placenta to produce IL-17A. This is supported by the other finding that lymphocytes in poly(I:C)-treated pregnant mice secrete more IL-17A but was a response specific to cells isolated from the placenta (Choi et al., 2016). During normal pregnancy, the placenta promotes maternal immune tolerance to prevent rejection of the fetus while still retaining the capacity to detect and defend against infection. Treg cells in the placenta play an important role maintaining the immunosuppressive uterine environment to permit fetal development (Zenclussen et al., 2006). However, under normal conditions, Th17 cells are also present in the decidua, the maternally derived component of the placenta (Nakashima et al., 2010; Wu et al., 2014), which indicates a delicate coordination of regulatory and pro-inflammatory T cell activity in the placenta. MIA shifts this balance by altering the cytokine milieu, enhancing IL-6 levels in particular, to favor activation of maternal Th17 cells. Indeed, poly(I:C)-evoked MIA increases Th17 cells in the placenta but not Treg cells (Choi et al., 2016). As a result, Th17 cells in the mother may compromise placental function, with detrimental effects on the developing fetus.

Maternal Th17 cells release effectors like IL-17A which may impact fetal development indirectly by regulating placental function and production of intermediary soluble factors that enter fetal circulation or directly by transport into the fetus. Maternal Th17 cells have been reported also to cross to the fetus themselves (Wienecke et al., 2012). In the case of MIA during pregnancy, IL-17A may be acting directly on the fetal brain at E14.5 to alter neurodevelopment resulting in ASD-like phenotypes. Direct injection of IL-17A cytokine into the fetal brain at this stage in the absence of MIA was shown to phenocopy poly(I:C)-induced MIA effects of ASD-like behaviors and cortical disorganization in the offspring (Choi et al., 2016). IL-17A mRNA was not detected in E14.5 fetal brain, indicating that IL-17A cytokine is not produced locally at this stage (Choi et al., 2016). Maternal sources likely provide the input IL-17A signal to the fetal brain during MIA, but it remains to be demonstrated directly that IL-17A or even Th17 cells from the mother traverse the placenta under these conditions.

Alternatively, maternal IL-17A induces a trans-placental signaling cascade that triggers IL-17A production in the fetus peripherally, which then acts on the developing brain. At this time in the periphery of the fetus, progenitor blood cells are proliferating and maturing in the liver, which is the main site of hematopoiesis or the formation of blood cells such as lymphoid cells during fetal development (Crawford et al., 2010). T cell progenitors are the primary lymphoid cells in the mouse fetal liver around E12.5 but begins decreasing by E13.5 as their main site of hematopoiesis shifts to the thymus (Crawford et al., 2010). Thus, the fetus may be capable of producing Th17 cells and providing a source of IL-17A in the periphery this way. On the other hand, it remains possible that IL-17A is locally produced by resident cells in the fetal brain but at low levels which require more sensitive methods of detection. In contrast with the cytokine, mRNA of the IL-17A receptor IL-17RA is detectable in E14.5 fetal brain and, moreover, is upregulated by poly(I:C)-evoked MIA (Choi et al., 2016). In this way, low levels of IL-17A may be able to exert detrimental effects on neurodevelopment. It is important to note that directly injecting IL-17A into the fetal brain at E14.5 resulted in a thinning of the cortical plate not observed following MIA induced at E12.5 (Choi et al., 2016), which underscores distinctions in the timing and severity of immune challenge, stage of the inflammatory response, and contributions of effectors besides IL-17A to MIA. However, the current data taken together suggest that MIA leading to ASD critically involves induction of maternal IL-17A signaling that is relayed to the fetus via the placenta somehow and disrupts normal brain development.

IL-6 and links to IL-17A signaling

These findings extend earlier studies that support a pathogenic role for IL-6 in MIA-associated ASD by delineating a mechanistic pathway through activation of maternal Th17 cells and subsequent IL-17A signaling. Consistent with IL-17A acting downstream of IL-6, induction of IL-6 is early and acute following poly(I:C) treatment of pregnant mice while IL-17A has a slower onset (Arrode-Bruses and Bruses, 2012; Choi et al., 2016). In particular, strong upregulation of IL-17A cytokine levels in maternal blood and mRNA levels in the placenta was observed 48 hours post-poly(I:C) treatment, on E14.5, whereas IL-6 activity had returned to basal levels at this time point (Choi et al., 2016). Supporting this causal sequence of cytokine activity, IL-17A blocking antibody in pregnant mice prevents ASD-like phenotypes in the offspring induced by maternal IL-6 injection (Choi et al., 2016), which was shown previously to be sufficient to reproduce poly(I:C)-induced MIA effects (Smith et al., 2007). Conversely, pregnant mice lacking IL-6 cannot up-regulate IL-17A (Choi et al., 2016) and have been shown to bear phenotypically normal offspring following poly(I:C)-evoked MIA (Smith et al., 2007), but direct IL-17A injection into the fetal brain causes ASD-like behavior to reappear in these offspring (Choi et al., 2016). Moreover, direct injection of IL-6 cytokine into E14.5 fetal brain in the absence of MIA failed to cause ASD-like phenotypes in the offspring (Choi et al., 2016). Combined, these results indicate that IL-17A but not IL-6 acts directly on the fetal brain, at least during the E14.5 stage, to disrupt neurodevelopment and precipitate ASD-like phenotypes in the offspring following MIA. However, because IL-6 protein and mRNA levels in the placenta and the fetal brain are known to increase rapidly after MIA induction (Hsiao and Patterson, 2011; Meyer et al., 2006; Wu et al., 2015), these findings suggest critical windows for cytokine action or cytokine-specific functions in different phases of neurodevelopment.

IV. IL-17A targets in the brain

A. IL-17A receptors

The IL-17 receptor (IL-17R) family consists of five related subunits (A-E) (Gaffen, 2009). IL-17A cytokine targets the receptor IL-17RA to trigger downstream signaling. Binding of IL-17A induces IL-17RA to recruit the IL-17RC subunit to form the complete receptor complex (Ely et al., 2009; Toy et al., 2006; Yao et al., 1995). Assembly of the subunits is required for IL-17A-dependent functions (Gu et al., 2013; Toy et al., 2006; Yao et al., 1995). In the periphery, IL-17RA is expressed in several different cell types, with the greatest expression in the spleen, kidney and hematopoietic or immune cells (Gu et al., 2013; Yao et al., 1995). In the brain, several cell types have been reported to express IL-17RA but at low or undetectable levels basally (Das Sarma et al., 2009; He et al., 2013; Wang et al., 2009). Expression is strongly increased during inflammation associated with brain injury, disease and dysfunction. Resident cells with enhanced mRNA and protein levels of IL-17RA have been observed in ischemia and the oxygen-glucose deprivation model of ischemia (Wang et al., 2009), in MS and its EAE model (Das Sarma et al., 2009; Kebir et al., 2007), and in focal cortical dysplasia (He et al., 2013). Increased IL-17RA expression in the inflamed brain suggests a heightened responsiveness to IL-17A signaling.

Similarly, poly(I:C)-evoked MIA leads to an increase in mRNA levels of IL-17RA in the fetal brain at E14.5 (Choi et al., 2016). Whether this upregulation is mediated by a positive feedback loop through IL-17A binding in the fetal brain or by another effector mechanism is not known but is prevented by maternal treatment with IL-17A blocking antibody (Choi et al., 2016), which supports the necessity of IL-17A signaling in MIA effects. Consistent with these results, ASD-like phenotypes were prevented in MIA-exposed offspring if IL-17RA expression was reduced in the mother through genetic deletion of the receptor, thereby blocking IL-17A signal transduction (Choi et al., 2016). Interestingly, pregnant mice with reduced IL-17RA expression did not exhibit the increase of blood IL-17A levels following poly(I:C) treatment (Choi et al., 2016), which provides further support not only for a critical role of maternal IL-17A signaling but also for a positive feedback mechanism regulating expression of IL-17A and its receptor in MIA. On the other hand, this finding obscures IL-17A acting on receptors in the mother or the fetus to precipitate MIA effects in the offspring. Because injecting IL-17A directly into the fetal brain was able to reproduce MIA effects, this identifies a contribution from fetal IL-17RA signaling (Choi et al., 2016). Moreover, direct injection of IL-17A into the brain of fetuses lacking IL-17RA was unable to cause ASD-like phenotypes in the offspring (Choi et al., 2016). Therefore, MIA induction of IL-17A signals through the receptor IL-17RA to mediate downstream consequences on the developing fetus (see Figure 1).

B. IL-17RA expression during cortical development

Within the human brain, weak expression of IL-17RA has been reported in neurons, glia, and endothelial cells from control cortical tissue (He et al., 2013; Kebir et al., 2007). From the mouse brain, mRNA and protein expression of IL-17RA has been observed under basal conditions in microglia and astrocytes (Das Sarma et al., 2009; Kawanokuchi et al., 2008), oligodendrocytes (Paintlia et al., 2011), cerebellar granule cells (Hu et al., 2014) but not cortical and hippocampal neurons (Kawanokuchi et al., 2008; Wang et al., 2009), and neural stem cells (Li et al., 2013). Expression then increases dramatically under states of inflammation, including MIA (Choi et al., 2016). In some instances, infiltrating immune cells from the bloodstream partially account for the upregulation of IL-17RA signal in the brain (Das Sarma et al., 2009). In MIA however, the increased mRNA levels of IL-17RA within the fetal brain, specifically the cortex, have not yet been localized to a particular cell type, but there are several possibilities.

At the time of poly(I:C)-evoked MIA on E12.5, cortical neurogenesis has already begun in the mouse fetal brain. Around E10, the neuroepithelial cells lining the cerebral ventricles are generating a few early neurons as well as transforming into radial glia, which function as the primary progenitors or neural stem cells (NSCs) that give rise to glutamatergic neurons and later to astrocytes and oligodendrocytes within the cortex (Noctor et al., 2002; Qian et al., 2000). Radial glia also function as a scaffold for newborn neurons in the ventricular (VZ) and subventricular zones (SVZ) to migrate out toward the pial surface, forming cortical layers II-VI in an inside-out manner whereby each wave of newborn neurons will migrate past existing layers to form more superficial layers (Nadarajah et al., 2003). In parallel to this process, γ-aminobutyric acid (GABA)ergic interneurons are born outside of the cortex from ganglionic eminences in the fetal brain and migrate tangentially into the cortex to be incorporated there (Marin and Rubenstein, 2001). An early stream of interneurons that migrates into the cortex at the pial surface around E11.5 eventually become layer I Cajal-Retzius neurons (Marin and Rubenstein, 2001). By E12.5, cortical neurogenesis and neuronal migration to form the cortical layers are underway and lasts until E17-18, when synaptogenesis begins (Greig et al., 2013; Li et al., 2010; Semple et al., 2013). At E16, neurogenesis begins switching to gliogenesis, in which radial glia generate astrocytes and then oligodendrocytes instead of neurons (Qian et al., 2000; Semple et al., 2013; Stevens, 2008). Therefore, during the period of prenatal exposure to MIA initiated at E12.5 and continued through E14.5 by IL-17A signaling, astrocytes and oligodendrocytes are not yet present whereas NSCs and neurons are proliferating, differentiating, and migrating in the fetal brain.

Microglia also are present in the brain during this developmental period. Microglia are brain-resident macrophages known to have diverse functions, including active surveillance of their microenvironment, secretion of immune and trophic factors for homeostasis as well as repair, clearing dead or dying cells and debris, and shaping synaptic connections (Ferrer et al., 1990; Hanisch, 2002; Luo and Chen, 2012; Paolicelli et al., 2011; Sierra et al., 2010). In the embryonic brain, microglia have been implicated in regulating neurodevelopment. Microglia are derived from yolk sac progenitors of hematopoietic origin that colonize the embryonic mouse brain beginning around E10 (Ginhoux et al., 2010), which coincides with the emergence of radial glia. This timing puts microglia in a position to influence the other resident brain cells that have yet to develop. Indeed, embryonic microglia have been demonstrated to regulate NSC numbers by supporting not only proliferation (Antony et al., 2011) but also normal programmed cell death during cortical development (Cunningham et al., 2013). Additionally, they have been demonstrated to regulate NSC differentiation into astrocytes (Antony et al., 2011; Nakanishi et al., 2007), outgrowth of dopaminergic axons (Squarzoni et al., 2014), laminar positioning of cortical interneurons (Squarzoni et al., 2014), and angiogenesis (Fantin et al., 2010) in the embryonic brain. Microglia persist within the brain throughout life, developing as a unique population that self-renews and is distinct from peripheral macrophages (Bruttger et al., 2015). During embryonic development, microglia proliferate and migrate throughout the brain and become more ramified as they mature (Squarzoni et al., 2014; Swinnen et al., 2013). In the cortex, they accumulate at the pial surface, distribute sparsely near the ventricular walls, and are largely absent from the cortical plate that eventually becomes the cortical layers, prior to E16.5 in mice (Squarzoni et al., 2014; Swinnen et al., 2013).

Also during this developmental period, vasculature is present in the brain. Vascularization of the mouse cortex starts around E11 when endothelial cells invade the neural tissue to form blood vessels (Daneman et al., 2010). As these vessels grow, the blood-brain barrier (BBB) starts forming to create a physical barrier between peripheral blood circulation and the central nervous system so that the brain maintains and develops in a precise internal environment (Bauer et al., 1993; Ek et al., 2012). BBB properties include intercellular tight junctions, low rates of transcytosis, and expression of transporters in endothelial cells that selectively control exchange across the vascular wall (Ek et al., 2012). Tight junctions between endothelial cells can be observed shortly after they grow into the embryonic brain and already restrict the passage of several small molecules (Bauer et al., 1993; Daneman et al., 2010; Ek et al., 2012). In the adult brain, astrocytic end feet are an important component of the BBB, encircling the endothelial cells that form the blood vessels, but astrocytes are not yet generated when angiogenesis initially occurs in the embryonic brain (Daneman et al., 2010). In contrast, pericytes, another important BBB component that surrounds the endothelial cells, are present and have been found to regulate vascular permeability at this stage (Daneman et al., 2010). One way is by downregulating endothelial cell expression of proteins that promote BBB permeability and immune cell infiltration (Daneman et al., 2010). The BBB becomes fully functional where molecules do not leak across the vasculature by E15.5 in the mouse cortex (Ben-Zvi et al., 2014).

Therefore, during the period from E12.5 when MIA is evoked by poly(I:C) to E14.5 when IL-17A is significantly elevated following MIA, neurons, NSCs, microglia, and endothelial cells are present in the fetal brain. All these cell types have been reported to express IL-17RA, although it remains to be demonstrated in the fetal brain. Then any or all of these cells may upregulate IL-17RA expression in response to MIA, contributing to the MIA-induced increase of IL-17RA mRNA levels in the fetal brain (Choi et al., 2016). Furthermore, MIA induction of IL-17A signaling could potentially act on any or all of these cell types to disrupt cortical development, leading to cortical dysplasia and behavioral abnormalities associated with ASD in the offspring (Figure 1).

V. Potential mechanisms of IL-17RA activation on cortical development

A. Cell type-specific effects

How might MIA induction of excessive IL-17A signaling in the fetal brain lead to ASD-like cortical malformation and behaviors in the offspring? One study reported that IL-17RA expression in cultured embryonic neurons increases following ischemic stress after initially being undetectable at basal states, and stimulation with IL-17A exacerbates neuronal death (Wang et al., 2009). In contrast with neurons, IL-17A stimulation of cultured NSCs from embryonic mice does not increase cell death but does reduce NSC numbers along with astrocyte and oligodendrocyte numbers (Li et al., 2013). IL-17A acts instead to inhibit NSC proliferation and their subsequent differentiation into later stages of neural cells (Li et al., 2013). MIA-induced IL-17A signaling may have similar effects on neurons and NSCs in the fetal brain. Indeed, poly(I:C) has been shown to reduce embryonic NSC division and neuronal numbers in the superficial layers of the cortex (De Miranda et al., 2010). In this way, if MIA induces IL-17RA activation in cortical neurons and NSCs of the developing brain, network organization and function may be affected, resulting in the cortical dysplasia and behavioral abnormalities associated with ASD.

If microglia express IL-17RA at embryonic stages, then MIA-induced IL-17A signaling may also activate microglia in the fetus and alter their normal functions, which could contribute to the defects observed in the offspring. Besides carrying out phagocytosis, activated microglia can release a variety of soluble factors including cytokines like IL-6 and TNFα that affect neurodevelopment depending on the stimulus and context (Cacci et al., 2008; Cunningham et al., 2013). For example, addition of microglia-conditioned media to cultured neurons was found to increase neuronal proliferation (Morgan et al., 2004). Similarly, embryonic NSCs cultured from microglia-deficient mice showed impaired proliferation and astrogliogenesis (Antony et al., 2011). Conversely, LPS-activated microglia were found to secrete pro-inflammatory factors that induced NSC apoptosis, reduced neuronal differentiation and enhanced glial differentiation in culture (Cacci et al., 2008). In vivo, microglia activated under conditions of LPS-evoked MIA were found to decrease the NSC pool and thickness of the VZ and SVZ in the embryonic cortex (Cunningham et al., 2013) and induce premature entry of interneurons into the cortical plate, resulting in deficient distribution around layer V (Squarzoni et al., 2014). Embryonic microglia also have been shown to regulate vascularization in the developing brain through close contact with endothelial cells and secretion of soluble factors that stimulate angiogenesis (Fantin et al., 2010). Thus, activated microglia in the embryonic brain may promote excessive growth of some cell types, which may come at the expense of other types, disturb normal cell migration, and alter brain connectivity and vasculature.

One study reported that poly(I:C)-evoked MIA does not affect the density and activation state of microglia during embryonic development, suggesting that embryonic microglia might not be involved in the poly(I:C)-induced MIA effects on the fetal brain (Smolders et al., 2015). However, this result may reflect induction-dependent differences in inflammation, as poly(I:C) was administered intravenously at E11.5 or E15.5 (Smolders et al., 2015). Because the nature, intensity, and timing of MIA affects the resulting phenotypes observed in the offspring, it is possible that embryonic microglia are activated under other conditions of poly(I:C)-evoked MIA. Microglia also may be found to participate in poly(I:C)-evoked MIA effects on the fetal brain using other measures of microglial activity, such as the production of specific factors. Microglial cultures stimulated with IL-17A have been shown to increase production of cytokines such as IL-6, chemokines such as CXCL2, and neurotrophic factors such as brain-derived neurotrophic factor (Das Sarma et al., 2009; Kawanokuchi et al., 2008). MIA-induced IL-17A may similarly stimulate microglia in the fetal brain, which in turn affect neurodevelopmental processes that lead to the ASD-like phenotypes in offspring. Interestingly, microglia not only are a target of IL-17A, but also have been shown to produce IL-17A (Kawanokuchi et al., 2008), which can result in a positive feedback loop. At E14.5 however, IL-17A mRNA was not detected in the embryonic brain, which suggests that resident microglia are not producing IL-17A at this stage (Choi et al., 2016). It remains to be determined if and how IL-17A stimulation of microglia directly through IL-17RA activation or indirectly alters their functions during fetal brain development.

Activation of IL-17RA on endothelial cells in the fetal brain may also play a role in mediating ASD-like phenotypes associated with MIA. The BBB physically prevents the translocation of lymphocytes, including Th17 cells, into the brain where they can influence inflammation and exert harmful effects. For example, increased neuronal death was observed when neurons were co-cultured with Th17 cells, which secreted the cytolytic molecules perforin and granzyme B (Kebir et al., 2007). From E12.5 to E14.5, the BBB in the mouse cortex is not yet fully sealed but is capable of barrier functions that exclude many peripheral molecules and cells from the developing brain. However, IL-17A stimulation of endothelial IL-17RA have been shown to promote BBB permeability by causing the disruption of tight junctions and release of signals that recruit macrophages to the area (Huppert et al., 2010; Kebir et al., 2007). If endothelial cells in the fetal brain respond similarly to IL-17A signaling induced by poly(I:C)-evoked MIA, then this enables the infiltration of peripheral factors that can not only disturb neurodevelopment but also injure the brain, thereby precipitating defects in the offspring. Additionally, activated microglia can increase BBB permeability (Prat et al., 2001) and may thus intensify these effects in the fetal brain. Future studies investigating IL-17RA expression in the fetal brain can help identify the cell type-specific targets and effects of IL-17A following MIA.

B. IL-17RA signal transduction pathways

Act1 functions downstream of the IL-17A receptor

A critical effector of intracellular signaling downstream from IL-17RA is Act1 (also known as CIKS) (Leonardi et al., 2000; Li et al., 2000). In fact, loss of Act1 leads to many similar phenotypes as loss of IL-17RA (Boisson et al., 2013; Corneth et al., 2014). For example, Act1-deficient mice display attenuated EAE (Kang et al., 2010) as mice lacking IL-17RA do (Gonzalez-Garcia et al., 2009). Act1 mediates IL-17RA signal transduction through several identified protein domains, including a helix-loop-helix (HLH) domain, a pair of TNF receptor associated-factor (TRAF) domains, a U-Box E3 ligase domain, and finally a SEF/IL-17R (SEFIR) domain (Conti et al., 2009; Swaidani et al., 2009). Once IL-17A binds and activates IL-17RA to form the complete receptor complex with IL-17RC, Act1 is recruited via the SEFIR and U-box domains (Chang et al., 2006; Liu et al., 2009). Following Act1 recruitment to the IL-17A receptor, a number of signaling cascades have been shown to be activated. In the peripheral immune system, these cascades include the nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) pathway and mitogen-activated protein kinase (MAPK) pathways, which consist of the extracellular signal-regulated kinase (ERK), p38, and JUN N-terminal kinase (JNK) pathways (Arthur and Ley, 2013; Huang et al., 2009a). Act1 is expressed in cells throughout the nervous system (Kang et al., 2010; Kang et al., 2013) and based on its role mediating IL-17RA signal transduction in peripheral systems, is likely to function in this capacity in the brain as well. Indeed, IL-17A induction of glial cell death in EAE was found to function through Act1 (Kang et al., 2013).

NF-κB: a major effector IL-17RA/Act1 signaling

One of the primary signaling cascades activated via Act1 is the NF-κB pathway (Li et al., 2000). NF-κB is a transcription factor that is centrally involved in the production of chemokine, cytokine, and immune effectors for inflammation (Barnes, 1997; Ghosh and Dass, 2016). NF-κB normally resides in the cytoplasm unable to enter the nucleus because it is bound to an inhibitory regulatory protein called IkBα (Jacobs and Harrison, 1998). Act1 recruits Traf6 via its TRAF interaction domains and subsequently ubiquitinates the protein via its U-box E3 ligase domain. Ubiquitinated Traf6 activates TGFβ-activated kinase 1 (TAK1), which then phosphorylates and activates IkB kinase (IKK). In turn, IKK phosphorylates IκBα, disengaging it from NF-kB in the cytoplasm, which allows NF-κB to translocate into the nucleus to regulate the transcription of numerous gene products directly involved in immune function (Ghosh and Dass, 2016; Gilmore, 2006). For example, NF-κB can induce expression of the chemokine CCL20, which attracts Th17 cells (Fujiie et al., 2001). While its immunological roles are well-described, NF-κB is also known to function in the brain. NF-κB is widely expressed in the nervous system (Bhakar et al., 2002; O'Neill and Kaltschmidt, 1997; Yalcin et al., 2003) and is activated by neurotrophic factors (Carter et al., 1996; Hamanoue et al., 1999) and neurotransmitters (Guerrini et al., 1995; Kaltschmidt et al., 1995). NF-κB also has been shown to regulate different aspects of synaptic function, memory, and cognition in many species (Albensi and Mattson, 2000; Engelmann and Haenold, 2016; Freudenthal et al., 1998; Heckscher et al., 2007; Kaltschmidt et al., 2006; Levenson et al., 2004; Meffert et al., 2003; Yeh et al., 2002). Therefore, IL-17A may act through NF-κB to regulate brain function.

In MIA-associated ASD, aberrant IL-17A signaling might then influence the expression of cortical and behavioral abnormalities in the offspring through NF-κB dysregulation in the fetal brain. In agreement with this idea, NF-κB has been shown to play a role in neurodevelopment. In cultured sensory neurons from the embryonic and postnatal mouse, NF-κB signaling was required for normal size and complexity of neuritic arbors (Gutierrez et al., 2005). Similarly, in organotypic brain slices obtained from neonatal mice, neurons also required NF-κB function for normal dendritic arbors. Blocking NF-κB activity was found to cause significant reductions in branch point numbers and overall dendritic length of pyramidal neurons in the slices (Gutierrez et al., 2005). Some studies, however, have reported finding little or no nascent NF-κB activity in neurons (Jarosinski et al., 2001; Listwak et al., 2013). Other studies provide evidence for NF-κB activity in glial cell populations rather than neurons (Mao et al., 2009; Moerman et al., 1999) or from activation of transcriptional enhancers associated with NF-κB rather than NF-κB itself (Mao et al., 2007). Finally, reports of neuronal NF-κB studies being confounded by poor reagent specificity also cast some doubt on the role of NF-κB in the nervous system (Herkenham et al., 2011). Therefore, while Act1-mediated NF-κB signaling is a promising pathway through which IL-17A may influence brain development and function, more work is needed to firmly establish a link between IL-17A and NF-κB in the nervous system.

IL-17RA and ERK signaling

Another Act1-dependent intracellular signal cascade activated by IL-17A is the ERK pathway. Interestingly, Act1 activation of the ERK pathway is mediated by its U-box E3 ligase activity (Liu et al., 2009), which indicates that the ubiquitination crucial for NF-κB signaling may carry over to multiple pathways downstream of IL-17RA. IL-17A promotes ERK activation in a variety of epithelial cells, chondrocytes, fibroblasts, and stem cells (Andoh et al., 2002; Huang et al., 2009b; Jovanovic et al., 1998; Laan et al., 2001; Lee et al., 2008; Sebkova et al., 2004; Song et al., 2015). IL-17A also has been found to promote ERK activation in cells of the nervous system, such as neurons cultured from dorsal root ganglia (Richter et al., 2012; Segond von Banchet et al., 2013), oligodendrocyte precursors (Rodgers et al., 2015), and astrocytes (Zhang et al., 2016). ERK signaling features prominently in neuronal growth (Gomez and Cohen, 1991; Samuels et al., 2008; Wu and Bradshaw, 1996), differentiation (Rebay and Rubin, 1995; Samuels et al., 2008), process outgrowth (Perron and Bixby, 1999; Wu and Bradshaw, 1996; Yamazaki et al., 2001), neuronal and axonal migration (Hirai et al., 2002), synaptic plasticity (English and Sweatt, 1997; Impey et al., 1998), and cognition (Cesarini et al., 2009; Samuels et al., 2008; Schafe et al., 2000; Walz et al., 2000). Thus, MIA induction of excessive IL-17A activity in the fetal brain may precipitate ASD through the dysregulation of this critical signaling pathway. Indeed, ERK-related mutations and genetic variation have been found in ASD patient groups (Wen et al., 2016). Eccentric ERK activation has also been observed in peripheral lymphocytes from ASD patients (Erickson et al., 2017) and in IPSC-derived neurons from patients with 22q11.2 deletions, an intellectual disability disorder with features related to ASD (Zhao et al., 2015). In animal models, dysregulation of ERK signaling has been shown to promote the expression of ASD-like phenotypes (Faridar et al., 2014; Yufune et al., 2015). Given the enormous role for ERK-related signaling in brain development and post-developmental function, IL-17A dysregulation of this pathway may be a key mechanism linking inflammation to ASD.

Other downstream pathways of IL-17RA

IL-17RA/Act1 signaling can activate an additional MAPK pathway through p38. In the study where IL-17A stimulation of embryonic NSCs was found to reduce their proliferation and differentiation, these effects were in part rescued by the p38 MAPK inhibitor SB203580 (Li et al., 2013). This suggests that p38 functions downstream of IL-17RA in NSCs and may also mediate MIA-induced effects of excessive IL-17A signaling in these cells. Finally, Act1 also plays a role in mRNA stabilization. Upon binding IL-17A, IL-17RA can form a multi-protein complex with Act1, TRAF2, TRAF5, and arginine/serine-rich splicing factor (ASF) (Bulek et al., 2011; Sun et al., 2011). This complex can stabilize CXCL1 or CXCL5 mRNAs by preventing ASF cleavage of the mRNA and enhancing expression of the encoded protein (Herjan et al., 2013). Act1 may perform a similar function in neurons and other brain-resident cells to stabilize mRNAs with related structure to CXCL1/5. Related to its role in mRNA stabilization, a new role for Act1 in the nucleus also has been reported (Velichko et al., 2016). Thus, Act1 activity may function to modulate translation of specific subsets of mRNA critical to neural processes underlying ASD.

VI. Future directions

What are the relevant cellular targets for IL-17A signaling in ASD?

IL-17A is expressed by a variety of tissues in the body and can function in a broad spectrum of contexts, ranging from the familiar immunological effects to also having pronounced effects on cortical development. How IL-17A exerts these effects is complex, which depends on cell type expression of IL-17RA, downstream effectors, and parallel signaling pathways that may act in concert with IL-17A (e.g. TNFα). An important question moving forward in the context of MIA is the identification of the specific cell types that are affected by IL-17A signaling to promote ASD-related deficits. Does aberrant IL-17A signaling directly promote neuronal alterations that lead to ASD or are the effects secondary, resulting from some other physiological change such as blood-brain barrier activity or activation of microglia? It will also be important to determine how MIA-induced IL-17A signaling changes over the course of development. This review limited the discussion of potential IL-17A targets in the developing cortex relevant during the window from E12.5 to E14.5. This corresponds to the time when MIA was evoked by poly(I:C) or direct injection of IL-17A into the fetal brain could reproduce MIA-like effects in the offspring. It is not known how long IL-17A upregulation continues after E14.5 but it is likely to persist after this time based on what is seen with other pro-inflammatory cytokines such as TNFα and IL-1β (Choi et al., 2016). Continued IL-17A activity during cortical development may have effects that additionally contribute to ASD in the offspring. In the case of astrocytes and oligodendrocytes, which do not develop until after E14.5, IL-17A can already disrupt these cell types by adversely affecting their precursors. However, IL-17A is also known to affect astrocytes and oligodendrocytes directly. In the postnatal brain, astrocytes are a major source of soluble factors, including cytokines and chemokines during inflammation (Dong and Benveniste, 2001). IL-17A stimulation of astrocytes can enhance IL-6 signaling (Ma et al., 2010), release of nitric oxide, which increases glutamate release and may therefore promote excitotoxicity (Trajkovic et al., 2001), and infiltration of peripheral immune cells (Rodgers et al., 2010). With oligodendrocytes, IL-17A stimulation can exacerbate apoptosis (Paintlia et al., 2011). Therefore, prolonged IL-17A in the fetal brain could have other detrimental effects on astrocytes and oligodendrocytes besides their initial development. Future studies can address these questions.

Do known IL-17RA signaling pathways also function in MIA-associated ASD?

Intracellularly, IL-17RA activation can promote signaling through the NF-kB and MAPK pathways. Both can exert powerful effects on the developing nervous system. It is entirely possible that aberrant IL-17A signaling in the brain act through these pathways to promote the development of ASD. There is evidence for the involvement of both pathways in ASD pathophysiology. NF-κB signaling has been implicated by in silico studies identifying highly expressed ASD candidate genes (Ziats and Rennert, 2011). A study examining the blood of autistic patients revealed significantly elevated NF-KB activity (Naik et al., 2011). In the valproic acid model of ASD, excess NF-KB activity may be involved in reducing NSC loss at a developmentally critical time, thus contributing to autistic-related behavioral phenotypes (Go et al., 2011). Direct genetic evidence linking ASD and NF-KB homeostatic function has also been recently identified (Manzini et al., 2014). For MAPK, elevated activity has been observed in the brain and lymphocytes from the BTBR mouse model of autism-related behaviors (Faridar et al., 2014). Interestingly, blockade of MAPK with the MEK inhibitor SL237 causes ASD-like phenotypes in mouse models (Yufune et al., 2015). Future studies will be important to determine the molecular signaling pathways downstream of IL-17RA activation in MIA-associated ASD.

What role might postnatal IL-17A signaling play in ASD?

Cells in the brain develop primarily after birth, encompassing a wide range of activities that include gliogenesis, microglial proliferation and maturation, synaptogenesis, and activity-dependent circuit formation. Therefore, critically important is the question of what contribution, if any, postnatal IL-17A signaling has in the manifestation of ASD. Elevated IL-17A levels have been found in a subset of ASD patients (Akintunde et al., 2015; Al-Ayadhi and Mostafa, 2012; Suzuki et al., 2011), suggesting that IL-17A regulation may be perturbed and influence neural processes after prenatal development. Similarly, MIA-exposed offspring in animal models have been reported to show increased IL-17A levels postnatally (Garay et al., 2013). Together with other chronic symptoms of inflammation observed in the brain and periphery of ASD patients and animal models, these data indicate persistent immune dysregulation in ASD (Akintunde et al., 2015; Ashwood et al., 2004; Ashwood et al., 2011; Vargas et al., 2005). This feature may represent a self-propagating inflammatory cycle initiated from developmental stages or a long-term maladaptation to early environmental insults. Prenatal immune challenge may set up an environment in which the fetus develops abnormally, affecting subsequent function of both the postnatal immune and nervous systems. Additionally, both immune and neural cells may be primed by this prenatal event, resulting in exaggerated neuroimmune responses to homeostatic perturbations at postnatal stages. Given the importance of early life experience in shaping the developing brain, increased IL-17A activity due to maternal inflammation may represent a critical trigger for life-long maladaptation of neuroimmune function that in some patients manifests as ASD. Moving forward it will be important to determine how aberrant postnatal IL-17A signaling may further contribute to ASD phenotypes.

In conclusion, ASD is a highly complex neurodevelopmental disorder with equally complex etiology. It is clear that immune dysfunction can play a significant role in the etiology of this disorder. While much remains to be discovered about how the immune system contributes to ASD, a growing number of human and animal model studies implicate a specific role for IL-17A signaling in the manifestation of at least some ASD cases. To better understand IL-17A mechanisms in ASD, we believe that key areas of future investigation will be identifying (1) the brain-specific cellular targets and downstream effectors of IL-17A and (2) the pre- and postnatal contributions of aberrant IL-17A signaling in ASD-related phenotypes. Improved understanding of IL-17A cellular targets, signaling pathways, and developmental influences may allow for targeted therapies and more effective interventions that can counter the effects of immune dysregulation in ASD.

Highlights.

• Prenatal exposure to maternal immune activation (MIA) has been linked to ASD risk.

• Offspring in MIA mouse models exhibit ASD-like behaviors and cortical dysplasia.

• The pro-inflammatory cytokine IL-17A may play a causal role in MIA-associated ASD.

• IL-17A injection into the fetal brain can phenocopy MIA-induced effects.

• Potential mechanisms of IL-17A signaling effects on neurodevelopment are proposed.