Biology Reference
In-Depth Information
genes. There are two potential explanations for why a stronger differential expression signal
was not observed:
1. Bronchial epithelium itself is not a locus of molecular pathology in asthma.
2. Heterogeneity of gene expression within the group of asthmatics examined led to an
underestimation of the gene expression differences when asthmatics, taken together, were
compared to controls.
The former explanation is unlikely, as many studies have shown dysregulation of the
bronchial epithelium in asthma, implicating defects in bronchial epithelium as an initiating
factor in asthmatic airway inflammation [32] . The second explanation, given the known het-
erogeneity of infiltrating inflammatory cells in asthma, is more likely, as gene expression pat-
terns present in some asthma patients and not others may have a mutually cancelling effect
in the comparison examined (i.e., all asthmatics vs. all controls).
Among the significantly differentially expressed genes in asthmatic bronchial epithelial
samples were three genes that had previously been shown to be responsive to IL13: chloride
channel, calcium-activated, family member 1 (CLCA1); serine peptidase inhibitor, clade B, member 2
(serpinB2); and periostin . CLCA1 is an extracellular protein associated with chloride channels
that mediates IL13-mediated differentiation of ciliated bronchial epithelial cells into mucus-
producing goblet cells, a process known as mucous metaplasia [33-35] . SerpinB2, also known
as plasminogen activator inhibitor-2, is a serine protease inhibitor that may be a negative reg-
ulator of Th1 responses in the airway [36,37] . Periostin is a matricellular protein and integrin
ligand associated with fibrosis and remodeling that is inducible in stromal cells by various
stimuli including IL13 and TGF-β [38-42] .
As these three genes could be induced by IL13 in bronchial epithelial cells in vitro and IL13
had been implicated as a potential mediator of asthmatic airway inflammation and hence
was under investigation as a therapeutic target, we hypothesized that bronchial epithelial
gene expression patterns might be used to identify a subset of asthma patients more likely
to have elevated activity of IL13 in their airways. Importantly, IL13 itself does not appear
to be expressed by stromal cells such as bronchial epithelial cells [43] , therefore, by serving
as signal amplifiers, IL13-responsive genes may be better indicators of IL13 activity than the
cytokine itself. To test this hypothesis, we examined whether CLCA1, serpinB2, and periostin
were co-regulated with each other within samples from the subjects in the study. We found
that each of the three genes was expressed at elevated levels in some but not all asthmat-
ics, with substantial overlap between the asthmatic and control populations at lower expres-
sion levels. Furthermore, the expression of the three genes was highly intercorrelated in
individual samples; i.e., if one gene was expressed at a given level in an individual subject
within the range of values observed in the population, the other two genes were expressed at
proportionally similar levels in that subject relative to the population. Based on these obser-
vations, we performed hierarchical clustering of all the asthma patients and controls in the
study based on the expression levels of CLCA1, serpinB2, and periostin and found that the
population divided into two distinct clusters: one cluster had high expression of all three
genes, whereas the second cluster had low expression of all three genes. Importantly, these
clusters did not separate by diagnosis: the cluster with high expression comprised about
half of the asthma patients in the study, while the cluster with low expression comprised the
Search WWH ::




Custom Search