Mold levels may modify the effect of variants in the chitinase gene, CHIT1, on emergency department visits and hospitalizations from asthma. We also found that mold levels do not modify the association between other variants in CHIT1, and variants in both CHIA and CHI3L1 and childhood asthma or asthma-related phenotypes. To our knowledge, this was the first study to examine the effect of mold levels on the association of SNPs in the genes of both chitinases and chitinase-like proteins with asthma and allergy-related phenotypes. Strengths of our study include the availability of mold levels in a well-defined clinical trial, the availability of outcomes over a 4-year time period, and a family-based design that avoids issues with population stratification.
Our results support increasing evidence that CHIT1
, which is primarily expressed in the lung, plays an important role in the pathophysiology of asthma in the proper environmental context of exposure to chitin, which was approximated by mold levels (24
). Our results are supported by a study of workers in the snow crab–processing industry who are exposed to high levels of chitin (found in the exoskeletons of crustaceans), which found that cumulative exposure to snow crab allergens is associated with prevalence of occupational asthma and allergy in a dose–response manner, even after adjusting for age, sex, and smoking (25
). Furthermore, intranasal administration of chitin to mice induces the accumulation of IL-4 expressing innate immune cells, inducing eosinophils and basophils (26
), further lending support to the notion that exposure to chitin induces an allergic phenotype. Chitinases seem to be able negatively to regulate the tissue infiltration of eosinophils and basophils (26
). Thus, the level of enzymatic activity of chitinases may be protective against development of allergies or asthma by breaking down chitin. Additional evidence supporting our hypothesis is a pilot study that found that subjects with the CHIT1
genotype that correlates with decreased levels of chitotriosidase had increased susceptibility to filarial infection (27
). These findings in the pathophysiology of asthma support our finding that CHIT1
may be associated with hospitalizations and emergency department visits from asthma in the setting of varying mold exposures.
Previous studies have found conflicting results on whether genes of chitinases and chitinase-like proteins are not associated with asthma-related phenotypes. Some studies have suggested that variants in CHIT1
, and CHI3L1
are not associated with asthma or other asthma phenotypes (13
). On the other hand, previous studies found that polymorphisms in CHIA
are associated with asthma and IgE levels (30
). Ober and coworkers (32
) concluded that SNPs in CHI3L1
are associated with bronchial hyperresponsiveness in the Hutterite population. Because this is a relatively isolated population, with similar environmental exposure, this finding may avoid some confounding effects of genetic and environmental heterogeneity (32
). One potential reason for the conflicting findings in the literature is environmental heterogeneity in exposure to sources of chitin between populations, and we show that accounting for environmental exposure to mold levels may help to clarify these genetic associations.
Despite the strengths of our study, a few caveats deserve mention. First, our sample size of 395 subjects was relatively small. Nevertheless, we did find that mold exposure significantly modified the relation between three SNPs in CHIT1
and one or more emergency department visits or hospitalizations from asthma. Second, our analysis was limited to one population and we did not have a replication population for study; thus, our results may not be generalizable to other populations. To our knowledge, no other longitudinal clinical trials of asthma have measured mold levels as an exposure; thus, we do not have other populations to replicate our findings. A recent review article on gene-by-environment interaction in asthma mentioned that the study of gene–environment interaction in relation to asthma is in its infancy (33
). Thus, our findings provide support for future studies examining gene–environment interactions in asthma, and should encourage evaluation of environmental exposures in these studies. In addition, we only included mold measurement from the randomization visit. Although a second mold measurement was attempted at the 3-year visit, 23% of subjects did not have a mold measurement, and FBAT does not allow repeated measurements for the environment variable. Finally, it is likely that mold exposure is only one source of chitin exposure. A comprehensive evaluation of all sources of environmental chitin exposure was beyond the scope of this study.
Both in vitro
and in vivo
studies have demonstrated that chitin and chitin derivatives have important immunologic effects and play an important role in pulmonary inflammation (34
). The literature suggesting the importance of chitinases in the pathophysiology of asthma is strong (4
), and chitinases may play a role in future targets for asthma therapy (24
). In future genetic studies of asthma, measurements of fungal levels could contribute important knowledge on the pathophysiology of asthma.
In conclusion, fungal levels may modulate the effect of variants in the chitinase gene, CHIT1, on emergency department visits and hospitalizations. This finding supports the important role that chitinases have in asthma.