PROJECT SUMMARY/ABSTRACT
Allergic diseases, including atopic dermatitis (AD), food allergies (FA), and asthma are common in young children
and are often co-morbid, suggesting a shared allergic pathobiology. Supporting this, development of these
diseases often occurs in a sequential pattern across infancy known as the “allergic march,” with a diagnosis of
AD before 12-months resulting in increased risk of FA and asthma. However, only a subset of infants with AD
proceed down the allergic march for unknown reasons. We hypothesize that different pathobiological
mechanisms (i.e. endotypes) operating in the skin, underlie AD clinical subtypes and their propensity to give rise
to additional allergic diseases. Supporting this, we have found a type 2 inflammatory skin endotype that was
greatly increased in AD subjects with FA relative to AD subjects without FA, with the ADFA+ patients also
exhibiting skin barrier abnormalities. Moreover, animal models have shown that skin barrier disruption with
inflammation can drive sensitization to allergens and the development of AD, FA, and airway inflammation.
Genetic studies also support a shared genetic disposition among allergic diseases and implicate dysregulation
of skin genes in this risk. Among adults, heterogeneity in inflammatory skin endotype has been observed,
including Th22, Th17, IL36G, and Th1 pathway activation. The roles of these endotypes in progression of infants
through the allergic march is unexplored. We have pioneered skin tape stripping (STS) RNA-seq methods to
allow repeated, minimally invasive skin expression profiling. Using these methods, we will evaluate whether
early-life skin pathobiology leads to the development of the allergic march and other sequences of allergic
disease development. Our proposal relies on phenotyping and STS samples collected in the Systems Biology of
Early Atopy (SUNBEAM) birth cohort (n=2500). From a SUNBEAM allergic disease case-cohort (n~800) we will
generate RNA-seq data on STS samples collected at birth, 2, 5, 12, 24, and 36 months, paired with allergic
disease phenotyping. In Aim 1 we will determine inflammatory endotypes of AD in infancy and their associated
patterns of gene expression dysregulation. Cellular deconvolution of RNA-seq will identify immune cell types and
changes in epidermal cellular composition that underlie different AD endotypes. We will determine whether skin
expression and endotypes, at birth and with AD at 2 and 5 mos, are predictive of food allergy at one year of age.
In Aim 2 we will determine longitudinal patterns of early life skin pathobiology that underlie the allergic march
and other sequences of allergic disease development. We will define allergic disease developmental classes,
their associated longitudinal expression and cellular composition profiles, and prospective biomarkers for allergic
disease development. In Aim 3, whole genome- and RNA-seq data will be used to identify eQTLs that influence
skin expression at each timepoint and longitudinal skin development among allergic disease groups. Using
TWAS, genetic models of early-life skin gene expression will be used with GWAS data for allergic diseases to
identify genetic variants influencing allergic disease risk through modulation of early-life skin gene expression.
