Computational Detection of Homologous Recombination Hotspots in X-Chromosome Autism Associated Genes

Background: Autism prevalence rates in the US and UK began increasing temporally close to the time that the MMR vaccine was switched from the type manufactured using animal cells to a type manufactured using human cells.    A yet unstudied possible environmental cause for autism is residual human DNA contaminant in the vaccines produced using human cells.    Debate about the dangers of residual human DNA in vaccines has been going on for 50 years, mainly related to cancer initiation.    Clinical gene therapy (1996) of male infants has induced childhood leukemia due to improper  DNA insertion.   To minimize the possibility of oncogenic DNA integrating into a vaccine recipient’s genome, the WHO and FDA have recommended DNAse treatments to reduce the size of contaminant DNA fragments to less than genic lengths, ~200-1000bp.   These recommendations were developed before the sequencing of the human genome.   DNA cellular diffusion studies have shown that short DNA fragments (<250bp) have higher probability of entering the nucleus than longer lengths.   Gene therapy studies have also demonstrated that naked DNA, injected intramuscularly, can remain  intact and be transported to the brain via retrograde axonal transport in motor neurons.   Once inside the nucleus, exogenous DNA can  integrate via homologous recombination, potentially in genomic regions called ‘recombination hotspots’.  Intra-species DNA integration, e.g., by homologous recombination, occurs with a probability a billion times greater  than inter-species homologous recombination.   Integration of short human DNA fragments has the potential to contribute to various human diseases, including autism.

Objectives: To verify the lengths of residual human DNA in vaccines and to determine if recombination hotspots occur near or in X-chromosome genes known to be associated with autism.

Methods: Human DNA from inactivated vaccines was isolated and characterized using standard procedures. A list of recombination hotspots, computationally derived by using Hapmap Phase II data, was downloaded.   A list of 238 genes associated with autism (AAGs) was downloaded from the ACGMAP website.   Gene coordinates, including transcription start and end sites, were downloaded from the UCSC Human Genome website.   Software was written to automate the location of overlaps between autism associated genes and recombination hotspots.

Results: The average human DNA fragment length in rubella vaccine was 220 base pairs.  Out of 1145 hotspots in the X-chromosome, 25 hotspots are located in 5 of 15 X-chromosome AAGs, between the transcription start and end sites.   These genes are NLGN3 and NLGN4X (neuroligins involved in synapse formation), AFF2 and  IL1RAPL1 (involved in X-linked mental retardation), and GRPR (gastrin releasing peptide receptor).

Conclusions: Autism-associated genes in the X-chromosome contain multiple regions where potential insertion of short, non-host homologous DNA can occur. With new knowledge due to the human genome project, particularly in regards to SNPs and epigenetics, further work must be done to understand the implications of integrated residual human DNA to the etiology of autism.