Studies suggesting a role for maternal antibodies in the etiology of autism reported that mothers of children with autism present serum patterns of immunoreactivity to prenatal rat brain, which are not observed in control mothers. These antibodies can be detected after delivery (Zimmerman et al. 2007) or at midpregnancy (Croen et al. 2008a) . Actually, mothers of an ASD child are four times more likely to have antibodies against brain proteins than control women (Brimberg et al. 2013) .

The potential role of a heightened or activated maternal immune response in the risk for ASD is further strengthened by epidemiological data from large population-based studies that show increased rates of autoimmune disorders in the families of individuals with ASD. Independently or coincidentally, the presence of specific anti-fetal brain antibodies in approximately 12 % of mothers of children with ASD, which are absent in mothers of children who are typically developing or mothers of children with developmental delays, suggests a potential inflammatory process occurring in mothers of children with ASD that leads to the production of antibodies directed to the developing brain (Onore et al. 2012; Croen et al. 2008a; Careaga and Ashwood 2012) .These antibodies can be detected after delivery (Zimmerman et al. 2007) or at midpregnancy (Zimmerman et al. 2007; Croen et al. 2008a). Actually, mothers of an ASD child are four times more likely to have antibodies against brain proteins than control women (Zimmerman et al. 2007; Brimberg et al. 2013) .

Others evidences suggest that the presence of maternal antibodies against fetal brain proteins of 37 and 73 kDa of molecular weight increases the risk of autism (Zimmerman et al. 2007; Braunschweig et al. 2008, 2012, 2013; Braunschweig and Van de Water 2012) or, at least, impairs verbal and nonverbal language acquisition (Piras et al. 2014) . A recent study showed that sera from mothers of autistic children react specifically to proteins highly expressed during brain development (Braunschweig et al. 2013; Bauman et al. 2013) . Moreover, sera from autism mothers recognize fetal brain rat proteins, but not adult ones (Zimmerman et al. 2007) , suggesting a specific vulnerability during embryonic development. However, other studies characterizing the presence of autoantibodies in autistic patients suggest that they play a postnatal role (see below).

Correlational studies linking the occurrence of maternal antibodies and specific symptoms of ASD are still largely lacking. Recently, it has been shown that the presence of IgG reactive to fetal brain tissue in the maternal serum is associated with an enlarged frontal lobe in autistic children, when compared with autistic children born from mothers without these antibodies (Nordahl et al. 2013) .

Although, it is not clear how these antibodies can result in the behavioral symptoms of ASD, some evidences from experimental animals show that they can affect neuronal function and alter behavior of the offspring. For example, the injection of the serum of a mother of an autistic child into pregnant mice resulted in cerebellar and behavioral alterations in the offspring. This serum contained antibodies binding to cerebellar Purkinje cells (Courchesne 1997; Fatemi et al. 2012) . Similarly, the injection of purified IgGs from 63 autism mothers into pregnant mice resulted in offspring with reduced sociability (Singer et al. 2009) . Finally, a single low dose gestational exposure to IgG from autism mothers results in mice showing impaired sensory and motor development and altered behavior (Braunschweig et al. 2012) .

In addition, the injection of purified IgGs from 21 mothers of at least one child with autism into pregnant rhesus monkeys resulted in offspring showing atypical social behavior and increased stereotypies (Martin et al. 2008) . Moreover, monkeys prenatally exposed to IgGs from autism mothers showed enlarged brain volume, especially due to increased white matter volume (Bauman et al. 2013) .

Autoantibodies directed against different CNS proteins have not only been detected in mothers, but also in autistic patients. The first report on autoantibodies against brain proteins was almost 30 years ago (Todd and Ciaranello 1985) , and they were shown to bind to proteins expressed in various rat brain regions (Singh and Rivas 2004b, 2004a) (Singh and Rivas 2004b) and human brain proteins (Cabanlit et al. 2007; Singer et al. 2006, 2009) . Numerous antigens have been identified, including glial and neuron-axon filament proteins (Singh et al. 1997b) , myelin basic protein (MBP)(Singh and Rivas 2004b, 2004a; Libbey et al. 2008) , serotonin receptors (Todd and Ciaranello 1985) , nerve growth factor (Connolly et al. 2006) , brain-derived neurotrophic factor (Connolly et al. 2006), and brain endothelial (Vojdani et al. 2002, 2004; Wills et al. 2007, 2011, 2009; Rout et al. 2012; Libbey et al. 2008; Kirkman et al. 2008) . It is also important to note that it is not clear whether autistic children are positive for more than one antibody.

Ganglioside M1 is the most abundant ganglioside in neural membranes and specific antibodies have been detected in some autoimmune diseases with neurological involvement. Autistic patients show increased serum levels of anti-ganglioside M1 antibodies and their titer correlated with the severity of the disease (Mostafa and Al-Ayadhi 2011b; Vojdani et al. 2002, 2004; Wills et al. 2011) . Nevertheless, some authors had found lack of association between these autoantibodies and autism (Moeller et al. 2013) .

A study, characterizing the prevalence and specificity of autoantibodies against cerebellar tissue (Wills et al. 2009; Vojdani et al. 2002), 21 % of the ASD children presented autoantibodies reactive to a protein expressed at cerebellum, specifically by Golgi cells, while another study identified antibodies against different cerebellar proteins in autistic children (Wills et al. 2009; Vojdani et al. 2002; Goines et al. 2011) , and showed that antibodies against a 45 kDa cerebellar protein are associated in an specific manner with autism,whereas, antibodies reacting to a 62 kDa cerebellar protein was associated with other disorders under the umbrella of ASD. Therefore, the specificity of the autoantibodies in ASD could underlay the different symptoms observed in autism, nevertheless, the pathological relevance of these must be clarified (Hegvik et al. 2014) . Some authors reporting antibodies reactive to fetal brain proteins in autism, suggest that it may predict better behavioral and emotional problems than diagnosis (Rossi et al. 2011) , while others reports claim that these antibodies do not predict autism (Morris et al. 2009) and that autoantibodies in children with autism appear to react more with the fully developed brain than with specific structure in the fetal brain unlike their mothers (Cabanlit et al. 2007; Singer 2009) ,as well as the previous reported immunoreactivities have been challenged (Mostafa and Al-Ayadhi 2012; Moeller et al. 2013) .

Worth noticing, another study has revealed in the sera of ASD children the presence of antibodies against GABAergic neurons throughout the central nervous system, and not specifically in the cerebellum (Wills et al. 2011; Rout et al. 2012) .

Moreover, the presence of neuronal protein-specific autoantibodies are associated with increased behavioral impairments and more severity in children with ASD (Mostafa and Al-Ayadhi 2011b, 2012) , suggesting a link between the autoimmune processes and behavioral dysfunction. So, autoantibodies directed against a 45 kDa protein present in the cerebellum were not only found more frequently in children with ASD but were also associated with lower adaptive and cognitive function, as well as increased aberrant behaviors (Wills et al. 2011; Goines et al. 2011) . However, replication studies of antibody-specific antigen targets, such as MBP and Glial Fibrillar Acid Protein (GFAP), have been inconsistent, suggesting that further studies are needed to identify the target or targets and/or associated autoimmune phenomena (Libbey et al. 2008; Kirkman et al. 2008) . It would be interesting to clarify whether the presence of autoantibodies in children can be or not useful to predict ASD, and which are the pathological mechanisms resulting in the core symptoms of the disease.