Immune abnormalities have been described in some individuals with autism spectrum disorders (ASD) as well as their family members. However, few studies have directly investigated the role of prenatal cytokine and chemokine profiles on neurodevelopmental outcomes in humans. In the current study, we characterized mid-gestational serum profiles of 22 cytokines and chemokines in mothers of children with ASD (N=415), developmental delay without ASD (DD) (N=188), and general population (GP) controls (N=428) using a bead-based multiplex technology. The ASD group was further divided into those with intellectual disabilities (DQ<70) (ASD+ID, N=184) and those without (DQ≥70) (ASD-noID, N=201). Levels of cytokines and chemokines were compared between groups using multivariate logistic regression analyses, adjusting for maternal age, ethnicity, birth country, and weight, as well as infant gender, birth year, and birth month. Mothers of children with ASD+ID had significantly elevated mid-gestational levels of numerous cytokines and chemokines, such as GM-CSF, IFN-γ, IL-1α, and IL-6, compared to mothers of children with either ASD-noID, those with DD, or GP controls. Conversely, mothers of children with either ASD-noID or with DD had significantly lower levels of the chemokines IL-8 and MCP-1 compared to mothers of GP controls. This observed immunologic distinction between mothers of children with ASD+ID from mothers of children with ASD-noID or DD suggests that the intellectual disability (ID) associated with ASD might be etiologically distinct from DD without ASD. These findings contribute to the ongoing efforts toward identification of early biological markers specific to sub-phenotypes of ASD.

Objectives To assay if plasma antibody levels in children with autism or developmental delays (DD) differ from those with typical development as an indicator of immune function and to correlate antibody levels with severity of behavioral symptoms. Methods Plasma was collected from children with autistic disorder (AU; n=116), DD but not autism (n=32), autism spectrum disorder but not full autism (n=27), and age-matched typically developing (TD) controls (n=96). Samples were assayed for systemic levels of immunoglobulin (IgG, IgM, IgA, and IgE) by enzyme-linked immunosorbent assay. Subjects with autism were evaluated using the Autism Diagnostic Observation Schedule and the Autism Diagnostic Interview—Revised, and all subjects were scored on the Aberrant Behavior Checklist (ABC) by the parents. Numerical scores for each of the ABC subscales as well as the total scores were then correlated with Ig levels. Results Children with AU have a significantly reduced level of plasma IgG (5.39±0.29 mg/mL) compared to the TD (7.72±0.28 mg/mL; P<0.001) and DD children (8.23±0.49 mg/mL; P<0.001). Children with autism also had a reduced level of plasma IgM (0.670.06mg/mL) compared to TD (0.79±0.05 mg/mL; P<0.05). Ig levels were negatively correlated with ABC scores for all children (IgG: r=−0.334, P<0.0001; IgM: r=−0.167, P=0.0285). Conclusion Children with AU have significantly reduced levels of plasma IgG and IgM compared to both DD and TD controls, suggesting an underlying defect in immune function. This reduction in specific Ig levels correlates with behavioral severity, where those patients with the highest scores in the behavioral battery have the most reduced levels of IgG and IgM.

Inflammation and asthma have both been reported in some children with autism spectrum disorder (ASD). To further assess this connection, peripheral immune cells isolated from young children with ASD and typically developing (TD) controls and the production of cytokines IL-17, −13, and −4 assessed following ex vivo mitogen stimulation. Notably, IL-17 production was significantly higher following stimulation in ASD children compared to controls. Moreover, IL-17 was increased in ASD children with co-morbid asthma compared to controls with the same condition.In conclusion, children with ASD exhibited a differential response to T cell stimulation with elevated IL-17 production compared to controls.

The contribution of peripheral immunity to autism spectrum disorders (ASDs) risk is debated and poorly understood. Some mothers of children with ASD have autoantibodies that react to fetal brain proteins, raising the possibility that a subset of ASD cases may be associated with a maternal antibody response during gestation. The mechanism by which the maternal immune system breaks tolerance has not been addressed. We hypothesized that the mechanism may involve decreased expression of the MET receptor tyrosine kinase, an ASD risk gene that also serves as a key negative regulator of immune responsiveness. In a sample of 365 mothers, including 202 mothers of children with ASD, the functional MET promoter variant rs1858830 C allele was strongly associated with the presence of an ASD-specific 37+73-kDa band pattern of maternal autoantibodies to fetal brain proteins (P=0.003). To determine the mechanism of this genetic association, we measured MET protein and cytokine production in freshly prepared peripheral blood mononuclear cells from 76 mothers of ASD and typically developing children. The MET rs1858830 C allele was significantly associated with MET protein expression (P=0.025). Moreover, decreased expression of the regulatory cytokine IL-10 was associated with both the MET gene C allele (P=0.001) and reduced MET protein levels (P=0.002). These results indicate genetic distinction among mothers who produce ASD-associated antibodies to fetal brain proteins, and suggest a potential mechanism for how a genetically determined decrease in MET protein production may lead to a reduction in immune regulation.

Maternal pregnancy levels of the inflammatory marker C-reactive protein (CRP) has been previously associated with autism spectrum disorder (ASD) in the offspring. We conducted a population-based nested case–control study with 500 children with ASD, 235 with developmental delay (DD) and 580 general population (GP) controls to further investigate whether elevated CRP during pregnancy increases the risk of ASD. Maternal CRP concentration was measured in archived serum collected during 15–19 weeks of pregnancy and genome-wide single-nucleotide polymorphism (SNP) data were generated. The levels of CRP were compared between ASD vs GP and DD vs GP. The genetic associations with CRP were assessed via linear regression. Maternal CRP levels in mid-pregnancy were lower in mothers of ASD compared with controls. The maternal CRP levels in the upper third and fourth quartiles were associated with a 45 and 44% decreased risk of ASD, respectively. Two SNPs at the CRP locus showed strong association with CRP levels but they were not associated with ASD. No difference was found between maternal CRP levels of DD and controls. The reasons for the lower levels of CRP in mothers of ASD are not known with certainty but may be related to alterations in the immune response to infectious agents. The biological mechanisms underlying this association remain to be clarified.