Environmental determinants of and impact on childhood asthma by the bacterial community in household dust

Abstract

Asthma increased dramatically in the last decades of the 20th century and is representative of chronic diseases that have been linked to altered microbial exposure and immune responses. Here we evaluate the effects of environmental exposures typically associated with asthma protection or risk on the microbial community structure of household dust (dogs, cats, and day care). PCR-denaturing gradient gel analysis (PCR-DGGE) demonstrated that the bacterial community structure in house dust is significantly impacted by the presence of dogs or cats in the home (P = 0.0190 and 0.0029, respectively) and by whether or not children attend day care (P = 0.0037). In addition, significant differences in the dust bacterial community were associated with asthma outcomes in young children, including wheezing (P = 0.0103) and specific IgE (P = 0.0184). Our findings suggest that specific bacterial populations within the community are associated with either risk or protection from asthma.

FIG. 1.

CCA triplot analysis of DGGE gel OTU banding patterns, comparing exposure and control microbial communities in household dust samples as a function of a single environmental variable. (a) Dogs (P = 0.0190); (b) cats (P = 0.0029); (c) day care attendance (P = 0.0037). Dust samples from the same households were used as the control (N) group for the dog and cat exposure analyses, and six out of seven of these dust samples were used as the control group for the day care analysis. No household dust samples were in common between the dog and cat or the day care and cat exposure (Y) groups. One household was in common between the dog and day care exposure (Y) groups. The red squares in each triplot represent the seven control (N) samples (e.g., no dog), and the blue circles represent the seven exposure (Y) samples (e.g., dog). The gray triangles in each triplot represent the centroids, which are the average location of the exposure or control sample. A permutation test was used to evaluate the null model of no relationship between the exposure and controls (P ≤ 0.05). The vertical distribution of sample points above and below the centroids (the residual axis) represents the variation in profile OTU banding patterns that is unrelated to variation between the exposure and control averages (centroids). The green numbers in each panel represent the OTUs from the DGGE gel profiles that have >20% of the maximum abundance. Each number represents a unique vertical OTU location on the DGGE gel. In numerical order, beginning at the top of the gel, a number was assigned to each possible vertical location for a band (OTU) among the 14 lanes analyzed. Thus, the location of these numbers can be used to distinguish the relative frequency of occurrence of an OTU in control and exposure samples. From the CCA plot then, the horizontal distribution of each green number (OTU) provides information about whether it is found more often in the exposure (located closer to the exposure centroid) or control (located closer to the control centroid) samples or is shared equally between them.

FIG. 2.

CCA triplot analysis of DGGE gel OTU banding patterns, comparing outcome and control microbial communities in household dust samples as a function of a single outcome variable. (a) Wheeze (P = 0.0103); (b) Sp-IgE (P = 0.0184). In these analyses, one of the seven household dust samples was the same for the asthma and Sp-IgE control (N) groups. Two out of seven households were the same for the wheeze and specific IgE outcome (Y) analyses. For a detailed explanation of the symbols on the triplot, see the legend for Fig. 1.