Aging, Environment, & DiseaseDevelopmental Biology & Stem CellsDNA Methylation and HydroxymethylationTools & Technology

Using Saliva to Measure the Brain: Translating A Salivary Diagnostic for Fetal Alcohol Exposure

shutterstock_49830313The consumption of any amount of alcohol during pregnancy can cause fetal alcohol spectrum disorders (FASD). FASD is an umbrella term covering a wide range of exposures from prenatal alcohol exposure, with the most severe outcome termed Fetal Alcohol Syndrome (FAS). However, the spectrum is wide and with no lower bound in sight, it represents only the tip of the iceberg. It is currently impossible to assess the amount of the population affected with FASD but even the most modest numbers suggest an unprecedented public health burden. Children with the behavioural and learning and memory deficits of FASD that do not show the craniofacial features do not receive the proper treatment. Early treatment is critical as it attenuates a large amount of the associated socio-behavioural and learning and memory problems by providing the unique educational aids required for these children. Unfortunately, many children with FASD are overlooked by current diagnostics that include the measurement of the distance between the eyes or tests of the babies’ first stool for whether the mother consumed extremely large amounts of alcohol during the later stages of pregnancy. Also, these children are often misdiagnosed with ADD/ADHD, as current psychological tests cannot differentiate between the two and this leads to improper and ineffective treatment, including unnecessary psychiatric medication.

Earlier studies have shown that the (endo)phenotypic heterogeneity of prenatal alcohol exposure stems from a combination of dosage, timing, and genetic background. Interestingly, FASD appears to be the first ‘disease of the epigenome’ and effects human embryonic stem cells. Unlike other drugs, alcohol is an extremely simple molecule, which allows it to bypass a number of cellular defence mechanisms. Alcohol (ethanol) directly impacts one-carbon metabolism, which establishes the epigenome and also interferes with synaptic signaling in the brain. Previously, we have shown in a mouse model that moderate prenatal alcohol exposure results in long-term dysregulation of DNA methylation and ncRNA of brain, with most marks appearing to represent disturbances to past developmental events. In our latest project we compared alcohol-associated alterations in mice to those seen in young human children. Rather than examining peripheral blood, we choose to examine buccal epithelial cells as spit can be obtained in a less invasive fashion. Furthermore, the buccal epithelium and neurons also share an intimate relationship via stem cell precursors.  These stem cells often shift between states resembling either the epithelium or neurons and give rise to both the brain and buccal epithelium. Alterations to regulatory regions in these stem cells during development may remain in both neurons and buccal epithelium. Thus, profiling changes in buccal epithelium may reveal possible perturbations in neurodevelopment

Given the heterogeneity of the disease we sought to initially discover these alterations in a ‘homogenous’ human cohort with confirmed exposure so that we could examine the effects without the confounding interactions that are common to the general population. Using CpG methylation arrays we found epigenomic dysregulation of functions strikingly enriched for neurodevelopmental and ‘online’ brain processes. The canonical pathways affected were also related to stem-cell signalling and synaptic signaling. Both these ontologies and pathways closely mirrored the findings in adult mouse brains.

Next, we examined whether specific loci were affected. We found increased methylation in the clustered protocadherins, which almost identically resembled the methylation pattern seen in our mouse model. The clustered protocadherins are involved in establishing individual neuronal identity – essentially serving as a barcode for each neuron to distinguish itself from its neighbours. These cellular adhesion molecules play an enormous role in formation of the complex networks of the brain.  A locus with a function so complex comes with regulation to match. The complexity of transcripts originating from the clustered protocadherins parallels the complexity of the immune system. However, rather than using genetic mechanisms, this diversity is accomplished through DNA methylation. In this region, there are several alternative promoters that lead to alternative transcript choices from the clusters. DNA methylation prevents binding of the genomic insulator CTCF and subsequent expression of the proper combination of isoforms. Thus, the increased methylation in FASD at this locus restricts the diversity of the developing brain’s neural networks.

However, the clustered protocadherins were not the only loci identified. We confirmed other equally important candidates, related to other disorders, via pyrosequencing. This confirmation was critical since it allows for children to be analyzed as individuals without the need for advanced statistical models and (carefully) matched controls. Finally, we analyzed a more diverse cohort and found that sex, age, and other drug exposure alter the profile in some, but not all of the discovered sites.

These findings offer a number of novel insights. First, they provide a proof-of-principle that salivary DNA methylation diagnostics are informative for FASD at a level much more sensitive than current diagnostics. Importantly, this principle may hold true for other neurodevelopmental exposures. Second, it offers a new hope for those with FASD, as it may be possible to use the FDA approved pyrosequencing in the clinic, much like how it is currently used for tumours. The development of such a sensitive diagnostic across a panel of the identified CpGs would ensure proper educational resources arrive to children with FASD and alleviate an enormous burden on humanity.


Original Article:

Laufer BI, Kapalanga J, Castellani CA, Diehl EJ, Yan L, & Singh SM (2015). Associative DNA methylation changes in children with prenatal alcohol exposure. Epigenomics, 1-16 PMID: 26178076



1. Khalid O, Kim JJ, Kim HS, Hoang M, Tu TG, Elie O, Lee C, Vu C, Horvath S, Spigelman I, & Kim Y (2014). Gene expression signatures affected by alcohol-induced DNA methylomic deregulation in human embryonic stem cells. Stem cell research, 12 (3), 791-806 PMID: 24751885

2. Laufer BI, Mantha K, Kleiber ML, Diehl EJ, Addison SM, & Singh SM (2013). Long-lasting alterations to DNA methylation and ncRNAs could underlie the effects of fetal alcohol exposure in mice. Disease models & mechanisms, 6 (4), 977-92 PMID: 23580197

Previous post

Epigenetics, Aging, and Neurodegeneration

Next post

Epigenetic Programming of Stress

Ben Laufer

Ben Laufer

Ben hails from the snowy land of Canada where he’s finishing a Ph.D. focused on neuroepigenomics and the environment. When he’s not consumed with his love for cats or researching and communicating epigenetics, you can find him outside hiking, trying to photograph nature, and California Dreamin'.