Friday, 7 February 2014

Anti-brain antibodies and behavioural profiles of autism

Anti-brain antibodies have been quite a recurrent theme in autism research circles over the past few years. I've covered research in this area a few times on this blog; primarily from the perspective of an intriguing bank of work looking at how the presence of circulating maternal auto-antibodies to foetal brain protein might heighten the risk of offspring autism (see here). If I also mention the words 'MAR autism' you can get a flavour as to how far this research has come, as well as the potential pitfalls which still await this area (see here).
Meeting Solomon @ Wikipedia 

Running concurrent to the maternal aspect to anti-brain antibodies is the suggestion that certain types of such autoantibodies may also be present in some people diagnosed with an autism spectrum disorder (ASD)* and indeed, whether there may be other important correlates to be further investigated (see here). The idea being that where such antibodies are present (for whatever reason**), different areas of the brain may be susceptible to our voracious immune system as a function of a blurring of the lines between what is 'self' and what is 'foreign' (see here for a better overview). Autoimmunity being the net result (see here).

The paper by Piras and colleagues*** joins the emerging research in this area with their findings pointing to "evidence of a pathomorphic role for anti-brain antibodies in autism". Once again I have to thanks Natasa from bringing this work to my attention.

The keen reader will already have recognised at least one familiar name attached to the Piras paper in Judy Van de Water affiliated to the MIND Institute. Indeed, mention of Prof. VdW provides me an ideal opportunity to link to quite a nice review paper titled 'Maternal and Fetal Anti-brain Antibodies in Development and Disease'**** which she co-authored and which provides about a good a review as one could hope for on the background to this area of investigation.

The Piras paper furthers this area of research insofar as looking at both mothers, offspring with autism and unaffected siblings (unaffected by a diagnosis of autism) with regards to the presence of anti-brain antibodies. That also their participant group is drawn from an Italian population offers something of a geographical comparison with the work so far (which has almost exclusively been done with mums and children living in the US).

The authors looked for the presence of a few things: (a) "patient-produced 45 and 62kDa autoantibodies to adult brain" (specifically to the cerebellum) and (b) "maternal 37, 39 and 73 kDa anti-fetal brain antibodies". Drawing on my very limited knowledge of all things chemistry, kDa refers to a kiloDalton, and a Dalton (named very aptly after John Dalton) is another name for the atomic mass unit. So a kDa is equivalent to 1000 atomic mass units, and when talking about kDa and antibodies, we're talking about the mass of antibodies (as proteins).

The results: well, actually quite interesting but not necessarily in the way you might expect. If I'm reading the paper right, the percentage of participants (those with autism and their unaffected siblings) presenting with those 45 and 62kDa anti-brain antibodies seemed to be skewed towards a greater frequency in siblings as opposed to those with autism (although there were no significant group differences). Certainly with respect to the 45kDa autoantibody, it looks like upwards of 7% of those with autism (27 / 355) were positive compared with over 12% of siblings (18 / 142). The frequencies for the 62kDa antibody were about the same (6-7%).

Further, the authors then reported on any connections between the presence of those two autoantibodies and a variety of different variables linked to both behavioural presentation of autism (assessed by VABS, ADOS, ADI-R) and other "Developmental, clinical and family history variables". Bearing in mind the fairly low positive rates recorded for the selected antibodies, various significant associations are reported with regards to IQ and "autism severity". An IQ less than or equal to 70 is reported in all the 45kDa antibody positive participants compared with just shy of 70% of antibody negative ones. That being said, I'd like to see some replication of these associations before coming down as too conclusive about them as a function of the small antibody positive group(s). Indeed with a significance value of p=0.012, "Number of sisters" (a greater number of sisters) comes out as the most significant item when correlating positive antibody status (again to the 45kDa antibody) and all that behavioural and background information in participants.

When it came to analysing the maternal antibodies (in mothers!), there is likewise quite a lot of data reported as a function of which antibody was looked at and whether mums were positive or not for detecting the antibody and how that related to participants with autism. First, "no difference in prevalence of maternal anti-fetal brain Abs [antibodies] was recorded between autistic individuals and unaffected siblings". That being said, birth order was potentially a factor for one antibody. Second, purely going on the best p-value, it was in relation to the maternal presence of the 73 kDa that participants with autism were more likely to have a familial presence of 'autoimmune disorder in a I degree relative' than those whose mothers were negative for this antibody (p=0.005). Personally, I think this could be an important finding.

There were lots of other correlations and associations with regards to sleep and the presence (or not) of non-verbal language which might also be important, although I'm minded to say that issues with 'kissing' behaviour for example, are not necessarily the most important part of this study (p=0.043). Finally, when combining the data on participant and maternal autoantibodies and their presence or not, there are some interesting data on the overlap between findings. Again, significance-wise was the observation that "the presence of the 62kDa autoantibody in an autistic individual predicts the presence of either the 39 or 73kDa antibodies in his/her mother".

I know I've gone on in this post a bit but I do feel it was important to do so. Not least because whilst this is potentially important data for autism and the anti-brain antibodies research, the results from Piras and colleagues were not exactly overwhelming in my opinion. Certainly the pretty low levels of participants with autism reporting the presence of the autoantibodies in question combined with the lack of an external control group outside of just unaffected siblings has left me, unfortunately, a little bit wanting.

Having also talked about all that recent work from Elaine Hsiao and colleagues (congratulations to her by the way for making this list) on the characteristics of mouse offspring following maternal immune activation during pregnancy, I'm also minded to start asking a few other questions about what might have been accomplished from this study with a few extra parameters. So, if that favourite topic of mine abnormal gut permeability had been looked at, could we have expected some differences in those presenting with anti-brain antibodies whether participants themselves or as a function of the various combinations of maternal antibodies? Delving further into the data, would permeability be a factor where there is for example, a family history of autoimmune conditions (bearing in mind that gut permeability and autoimmunity seem to be developing something of a relationship*****)? And then there is all that work on antibody cross-reactivity to consider****** which is a whole extra ball-game...

I'm not going to get too obsessed over these additional factors, but it certainly would be useful to see them in experimental play at some point.

Music... how about The Killers and Human.


* Singer HS. et al. Antibrain antibodies in children with autism and their unaffected siblings. J Neuroimmunol. 2006 Sep;178(1-2):149-55.

** Hornig M. The role of microbes and autoimmunity in the pathogenesis of neuropsychiatric illness. Curr Opin Rheumatol. 2013 Jul;25(4):488-795.

*** Piras I. et al. Anti-brain antibodies are associated with more severe cognitive and behavioral profiles in Italian children with Autism Spectrum Disorder. Brain Behav Immun. 2014 Jan 2. pii: S0889-1591(13)00605-3.

**** Fox E. et al. Maternal and fetal antibrain antibodies in development and disease. Dev Neurobiol. 2012 Oct;72(10):1327-34.

***** Fasano A. Zonulin and its regulation of intestinal barrier function: the biological door to inflammation, autoimmunity, and cancer. Physiol Rev. 2011 Jan;91(1):151-75.

****** Vojdani A. et al. Immune response to dietary proteins, gliadin and cerebellar peptides in children with autism. Nutr Neurosci. 2004 Jun;7(3):151-61.

----------- Piras IS, Haapanen L, Napolioni V, Sacco R, Van de Water J, & Persico AM (2014). Anti-brain antibodies are associated with more severe cognitive and behavioral profiles in Italian children with Autism Spectrum Disorder. Brain, behavior, and immunity PMID: 24389156