Itchy skin like what people get in dermatitis. The main crux (without digging into the weeds on the second crux regarding JAK signalling probably only interesting/comprehensible for immunologists) of the article is that they weren't getting asthma in their mice that were genetically engineered to get asthma like a human with a similar genetic background. This was because the lab environment is too clean and the mice weren't getting exposed to the allergen. When they introduced an allergen to these mice then they got asthma as expected. So in this case humans are unreliable models of this disease in mice because humans live in highly variable and dirty environments where they might get exposed to environmental conditions that drive certain phenotypes based on their genotype. so when you go and model these genotypes in the lab without an understanding of the environmental background, you might not get the phenotypes you expect to find based on those genotypes.