Featured,  Health,  Science

Maternal Antibodies, Brain Development and Autism

By Roberta Attanasio

The world of autism is a diverse one — autism and autism spectrum disorder are general terms for a group of complex and extremely heterogenous life-long neurodevelopmental conditions. Autism is also one of medicine’s greatest challenges. The current general consensus is that autism has multiple causes, and these causes can be as varied as the types of autism, which are not very well defined, yet. It is almost certain, though, that autism has its roots in very early brain development, development that can be influenced by a variety of factors.

In a subset of childrem with autism, one of these factors may involve production of antibodies by the mother — not the antibodies that defend us from infection — rather, antibodies that bind our own molecules and therefore may lead to damage of our own cells. These antibodies are called auto-antibodies and, in the case of some autistic children, may induce damage of brain cells.

Credit: Petteri Sulonen from Helsinki, Finland, licensed under the Creative Commons Attribution 2.0 Generic license.

During pregnancy, maternal antibodies are normally transferred to the fetus through the placenta to protect the infant from infection. However, the same placental transfer system can also allow auto-antibodies to reach the fetus — indeed, the placental transfer system does not discriminate between “good” and “bad” antibodies. If, before or during pregnancy, the mother produces auto-antibodies that bind the fetal brain, and these auto-antibodies reach the fetal brain, normal brain development can be hindered.

Results from a recent study (July 2013) by Judy Van de Water (UC Davis MIND Institute) and collaborators show that mothers of children with autism produce auto-antibodies that bind seven proteins involved in different aspects of brain development. These proteins are found throughout the body and are present at significant levels in the human fetal brain, where they play established roles in neurodevelopment.

The study, entitled “Autism-specific maternal autoantibodies recognize critical proteins in developing brain” and published in the scientific journal Translational Psychiatry, involved 246 mothers with children affected by autism and 149 mothers with normally developing children. Twenty-three percent of the mothers with children affected by autism produced auto-antibodies that recognized two or more of the seven proteins, whereas only 1 percent of the mothers with normally developing children produced such auto-antibodies.

The investigators coined a special designation for the type of autism possibly linked to these auto-antibodies — maternal autoantibody-related (MAR) autism. Van de Water said “It is important to note that women have no control over whether or not they develop these autoantibodies, much like any other autoimmune disorder and, like other autoimmune disorders, we do not know what the initial trigger is that leads to their production.”

The autoantibody/MAR hypothesis appears to be supported by a variety of studies carried out in animals during the past few years. For example, results from a study published in 2009 show that antibodies collected from mothers of children with autism and experimentally transferred to pregnant mice produce neurobehavioral alterations in the offspring.

This hypothesis applies to only a subset of children with autism spectrum disorder. A few months ago, a study pointed out a possible link between autism and air pollution. Indeed, there are multiple genetic and environmental factors currently under study.

For children affected by MAR autism, the identification of specific brain proteins representing the target of maternal autoantibodies may represents a much-needed pathway to treatment. Finding out why women produce auto-antibodies against proteins present at significant levels in the fetal brain will provide another much-needed pathway, the pathway to prevention.


  • chelton

    The importance of researching sources of autism is increasingly apparent when considering the prevalence of autism spectrum disorder. The statistics from the CDC and Van de Water’s article are staggering, stating one in eighty-eight children has ASD. When considering this posting on maternal auto-antibody related autism, I immediately begin thinking about measures for prevention from this horrible circumstance. The adult immune systems’ mechanism of clonal deletion protects self-antigens from antibody production. However, fetus self-antigens are susceptible to maternal antibodies that cross the placental barrier that is indiscriminant to “good” or “harmful” antibodies. I think a possible source for the onset of these maternal auto-antibodies could be found in the examination of these fetal neurodevelopment proteins. Each of these functions in a pivotal role to establish metabolism and form of fetal brain structures. Though, these seven proteins regularly exist throughout the body. Perhaps, the elevated concentration of these individual proteins coupled with the combined presence of them during neurodevelopment allows them to naturally diffuse across the blood-brain barrier and come into contact with the maternal immune system. Thus, the proteins elicit the onset of the maternal auto-antibodies and delay typical brain development.
    Furthermore, I noticed in the Van de Water study LDH was a significant factor in the combinations found in the mothers with children with ASD, and LDH is related to an inflammatory response. I would be curious to see further research about this enzyme.

    • IJ Oh

      Chelton, I think you bring up some great points, particularly about the fetus self-antigens that are susceptible to maternal antibodies that are able to infiltrate the placenta. Would it then be possible to create a vaccine or some sort of treatment for the fetus (or even child, but would that be too late?) to produce a defensive immune response? Maybe we prep the fetus’s immune system for some type of B and T lymphocyte memory cells so that the body has some recognition of the maternal Ab? I don’t even think that would be possible since at this point the immune system of the fetus is developing and is passive at best.

      Also, this made me wonder.. is there a correlation between the concentration of maternal Ab and at what point on the Autism spectrum the child will be on? Autism is demonstrated in a multitude of ways depending on where the child is on the spectrum. Could there be some sort of relationship between prolonged exposure, the amount of maternal Ab that can cross, and how severity of impairment from Autism.
      Just thinking and typing.

      • chelton

        LDH is an acronym for lactate dehydrogenase, which was one of the seven protein antigens related to maternal autoantibody autism tested in this study. LDH is found in the rodent fetal brain to function as a component of cellular metabolism. It is typically found in high levels when released from damage and breakdown in tissues in several places in the human body with toxic exposures and viral infections. According to the study, while it’s antibodies do not seem to directly alter neurodevelopment, the antibodies for the antigen are found in individuals exposed to the industrial solvent trichloroethylene. Trichloroethylene is not in use in the United States any longer, but it is still used in some other countries, and some trace amounts are still found in ground water. Since LDH levels appear with this tissue damage and due to inflammation, the LDH could elicit antibody production in the form of a maternal immune response and show something about the development of these autoantibodies. As related to these factors and the 28% prevalence of LDH autoantibodies among the mothers of children with autism spectrum disorder from the study, further investigation is needed.

  • tha5

    Because I have two close friends with an autistic younger sister, I have also been very particularly fascinated in the topic of autism. My initial reaction to reading this blog post is how do the critical proteins become upregulated and expressed in higher levels in fetal brains? In an earlier study by Nordahl et al, the researchers find that maternal-autoantibodies also correspond with abnormal brain enlargement. This is most likely due to the upregulation of the seven proteins found by Judy Van de Water.

    Antibodies can either activate the complement system or activate effector cells. When the complement system is activated, the antibody can either flag the antigen for phagocytic killing by binding to certain complement molecules or directly kill the antigen with the help of a membrane attack complex. Binding of an antibody to an antigen can also induce effector cells like NK cells or phagocytes. It will be interesting to discover the exact mode of action of these maternal auto-antibodies to see whether they cause an antibody-antigen interaction or simply flag the antigen for destruction.

  • MeganNWynegar88

    I have had a long standing fascination with autism and this article has brought up some very interesting ideas about the disorder’s epidemiology. This process of auto-antibody transfer to the fetal brain was mentioned in an article by Moises Velasquez-Manoff in The New York Times (http://www.nytimes.com/2012/08/26/opinion/sunday/immune-disorders-and-autism.html?pagewanted=all) last year. Velasquez-Manoff’s article, however, propositions that inflammation during pregnancy, not just auto-antibody presence, increases the risk of offspring become autistic. Autoimmune diseases, allergies, asthma, arthritis, celiac disease, metabolic syndrome and pre-natal viral and bacterial infections are all cited as inflammation causing afflictions that increase the presence of autism in the offspring of affected mothers. Velasquez-Manoff further states that LACK of environmental pathogens in developed parts of the world are a root cause of autoimmune and other autism associated diseases, whereas regions that experience greater amounts of pathogen exposure have decreased incidence of autoimmune disease and autism.

    Perhaps development of tests and testing procedures to detect auto-antibodies in early pregnancy, implementation of early testing for autoimmune other autism-associated diseases, more rapid detection and control of pre-natal infection and future developments in inflammation regulating medications may go a long way in preventing or at least halting the increase of autism development.

  • IJ Oh

    This was an interesting article about a new approach when looking at Autism.
    In response to Christina K, I don’t think that the adaptive immune system would come into play. At this point, the fetus has inherited antibodies from the mother. Typically whenever a pathogen is encountered the mother’s immune response should handle it. But after the child is born and encounters pathogens for him/herself, then the child is able to develop memory B and T cells for the target response of adaptive immunity. Key point is adaptive immune response needs to be triggered by dendritic presentation of antigen peptides.

    To follow along with the article, the multifaceted cause of autism was mentioned and included environmental factors. I think it would be interesting to examine how environmental factors (like pollution, etc) could affect the development of auto-antibodies in potential mothers and/or how these dynamics would affect transference to child.

    • Mira

      This is very interesting, and I am enjoying reading this over and over as well as the comments from others. Again, this is another example of how an external factor may influence the immune system, and also have an effect on the offspring. Like the vast majority, a lot of questions triggered by the readings, and comments from others came to me. I am curious to know if auto antibodies in the above described case is the same as autoimmunity given that the immune system is binding proteins from the fetus which is part of the mother rather than foreign antigens.
      Also, an interesting study will be to find the exact stage of fetal development when this binding occurs. Also, does anybody know if the fetal proteins to which the auto-antibodies bind are localized only in the brain?
      Last, to respond to IJ OH, i think in order to reach antibody production, the adaptive immune system has to be involved, and of course thru the innate. Not in the fetus, but in the mother who produces antibodies, and transfers the antibodies to the fetus.

    • SJ

      I was wondering what other environmental factors other than air pollution would be the most important to take into account when discussing the development of auto-antibodies? What would you think would be the most important?

  • Samantha S.

    Autism is a disorder that I can willingly admit I have very little knowledge about and a large part of that rests in the seemingly vague description and understanding of it.

    The notion that auto-antibodies transferred from the mother is a very intriguing notion and one that will hopefully shed more light on this disorder.

    Another study conducted by M.D. Bauman et. al and published in “Translational Psychiatry” (2013) http://www.nature.com/tp/journal/v3/n7/full/tp201347a.html looked at the effects of human maternal auto-antibodies thought to cause autism on the rhesus monkey. They also found that some infant and adolescent monkeys exposed exhibited behavioral abnormalities compared to normal monkeys.

    As interesting as these findings are I find myself pondering over a few questions. (1) With such a broad scale for autism diagnosis I would like to know at what “stage” of autism where the children at whose mothers were selected for these studies? Are the levels of maternal auto-antibodies the same for all those diagnosed regardless of severity? (2) Ethically these experiments could not be done in humans. However, the antibodies used in the animal experiments are human and not an analogous animal variant. Could the cross-species testing have an effect on the observed results?

  • MEE

    I found this article very interesting and I enjoyed reading the other comments and ideas. I have a family friend that has quadruplets that are all boys and all have autism. Therefore, I found the comment about autism being higher in males very interesting. Also, it makes me wonder if the quads have MAR autism and their mom’s auto-antibodies binded to proteins in all 4 fetal brains.

    I read an article by Ed Yong that also references Van de Water’s group. It said they are working to develop a test to predict a child’s risk of developing autism based on the mother’s antibodies. But the major point it made is that, “[The test] would allow mothers to plan.” It’s sad because it seems like (at this point) this is a hard problem to prevent. But, at best, mother’s will be able to plan to get their autistic children in educational programs early on. It is known that early intervention is best for the child, rather than waiting until the child enters school at age 4 or 5.

    I do think having a test to predict a child’s risk for MAR is smart. But, I also think continued research on the proteins and why mother’s make auto-anitbodies against the proteins is very important for treatment and prevention.

  • Christina K.

    Correlating autism with maternal auto-antibodies is an interesting notion. In the article, “Autism-specific maternal auto-antibodies critical proteins”, highlights six proteins that are expressed in the brain during fetal development that have specific IgG reactivity to LDH, YBX1, cypin, STIP1, CRMP1, and CRMP2. The indication that maternal auto-antibodies are so specific, could lead to a discovery of how auto-antibodies react to the fetal brain. It is unfortunate that during fetal development, there is no distinction between maternal and auto-antibodies. The normal act of antibodies crossing the placenta to add protection from infection can have altering effects. The act of not being able to defend one from harm demonstrates the importance of the innate system and adaptive system. For example in developed adults, positive and negative cell selection allows identification of self T-cells that bear antigen receptors that work well with self-MHC I and II and those that do not work well with self-MHC I and II, respectively.

    Questions that may be posed, is there any way to counteract the auto-antibodies once they have reached the fetal brain? Can the adaptive immune system respond? Unfortunately, this does not seem a likely question that could be answered with positive results. Besides pollution, another factor to consider is the amount of male’s to female’s ratio in those who have autism. There are a larger number of males that have autism and have an increase negative effect in the change in brain development. What factors make males more likely to have MAR autism opposed to females? There are a number of questions that need answers. From experience, working with autistic children can be challenging and rewarding; hopefully the link between autism and auto-antibodies will stimulate positive results in future research.

    • mcotton6

      Christina K- In response to your question about understanding the difference in males to females diagnosed with autism, I found an article on The National Autistic Society website whcih had some interesting insight on this topic. One point mentioned was that females tend to show fewer and less severe symptoms of autisim than males do which leads to less frequent diagnosis of autism in females. Many autistic females could be missing from the statistics making it seem that there is a significant difference in diagnosed sex ratios. There are also sex differences to take in to consideration which could keep autistic females from being diagnosed. Females tend to posess stronger verbal skills than males which can also mask the symptoms of autism.

      There was an interesting hypothesis mentioned from a paper by D. H. Skuse which tied the lower percentage of autism in females to the X chromosome received from their fathers. Skuse proposed that since males and females both receive an X chromosome from their mother, this chromosome could contain the autism determining gene. Further, the X chromosome received from the father could contain a protective gene against autism leading to a lower likelihood of autism in females. This hypothesis has not been proved but it seemed to be a very interesting thought that could lead to more concrete answers concerning the difference in autism between genders.

      • Christina K.

        To mcotton6:

        Thank you for the insight. I think you bring up some very interesting points that could contribute to why there is a larger ratio of males then females with autism. Another factor that could be considered is social stigma.

  • aa_alidina

    This article gives an excellent insight to MAR autism. There is a significant amount of evidence suggesting that some forms of autism are associated with dysfunction of the immune system. People with autism have an abnormality with their immune system which leads to feeble immune response to pathogen .
    As written in the article, maternal autoantibody related autism is baffling because mothers have no control on whether or not they develop these autoantibodies and scientist are still seeking answers to why these autoantibodies develop. It will be good for researchers to study the development of MAR autism using different factors such as environment and genetics. Also it would interesting to research the phase of fetal development when these maternal autoantibodies first appear. And if we can figure that out then may be there can be an option of doing a blood test done prenatally to assess the risk of having a baby with ASD?

  • myshell3

    While there are likely many factors that cause the broad range of autism spectrum disorder, I find the case for maternal auto-antibodies to be a compelling one. Interestingly, this post also mentions a study correlating air pollution and autism. Could these two correlations be at all related? Could the general innate immune response invoked by certain pollutants be, in some cases, causing the development of auto antibodies? After all, the adaptive immune response that is responsible for antibody production is only activated when the innate immune response is trying unsuccessfully to clear an infection. It seems feasible that in the immune system’s attempts to rid itself of multiple foreign pollutant “pathogens”, could lead to the production of pro-inflammatory mediators (like cytokines, interferon etc), that are needed for antibody production. However, with no pathogen present, is it possible for the cell debris caused by the tissue damage associated with the pollution induced inflammation to be presented to lymphocytes that ultimately develop auto-antibodies? I am certainly curious although it is disheartening to think that such studies are necessary to make people concerned about pollution.

    • Roberta Attanasio

      This is a very interesting and plausible hypothesis, and one that can probably be tested in experimental settings. The antigens recognized by these auto-antibodies are present in many different tissues, however they can reach the brain only during fetal development. Production of these auto-antibodies by the mothers is most likely unrelated to brain-specific responses and could result from mechanisms similar to those you mentioned.

      • myshell3

        There are so many cytokines that can affect B cells, it seems like it would be difficult to develop an experiement in correlating pollution exposure to cytokine levels, especially considering that cytokines often act in paracrine and autocrine ways. It would probably be easier to do a broader population based study in which pollution exposure is compared to prevalence of autoantiboies. Although when considering such a study, I would be interested in how auto antibodies were detected.

  • edreamer91

    “The world of autism is a diverse one”. Yes, I fully agree. I think autism is just a big pot where you can put anything in, anything that involves difficulties to relate to those around you, that you can’t diagnose. And the diagnosis of autism is a global problem, it’s called the global epidemics of misdiagnosis.

Leave a Reply

Your email address will not be published. Required fields are marked *