Page 32 - Read Online

Page 8 of 11 Zhang et al. J Transl Genet Genom 2018;2:18. I https://doi.org/10.20517/jtgg.2018.22

DISCUSSION
Asthma is a heterogeneous disease that exemplifies the need for personalized approaches to medicine.
Because different asthma phenotypes may be driven by different mechanisms and thus respond differently to
specific therapies, it is crucial to develop diagnostic approaches to identify and classify these subgroups.

In a previous study, we identified a panel of ~40 miRNAs that were differentially expressed in asthmatics vs.
[7]
non-asthmatics . Our previous work also demonstrated that miRNAs may provide information about the
presence or absence of eosinophilic asthma, and that miRNAs upregulated in eosinophilic asthma may play
roles in disease pathogenesis. In this study, we sought to further characterize these relationships to determine
whether miRNA expression patterns define subgroups of asthmatics that differ by atopy, eosinophilia,
severity, and clinical characteristics. Our findings demonstrate that blood miRNA expression profiles define
distinct asthma phenotypes, and that expression patterns differ based on severity, BMI, and eosinophila. The
importance of these findings is discussed herein.

Association of miRNA expression patterns with mild, non-eosinophilic asthma
Cluster 1 was associated with mild asthma, low ICS use, and low atopy (as assessed by number of allergens
that tested positive and blood eosinophilia). In terms of miRNA expression, it correlated with low levels of
the Let7 family, which also includes miR-98 (we classified these miRNAs as “group 1” based on expression
pattern), as well as low levels of group 2 miRNAs, which included miR-21 and miR-155 [Figure 3]. All
of these miRNAs have been shown to play roles in asthmatic inflammation. It was demonstrated that
increased miR-98 in peripheral blood may contribute to pathogenesis of asthma by lowering the frequency
[12]
of immune regulatory B cells . There is some evidence to suggest that Let-7 has a pro-inflammatory role
in asthma, primarily by promoting type 2 inflammation [13,14] . MiR-155 and miR-21 are both known to be
involved in asthma pathogenesis, particularly in asthma that is associated with eosinophilia [15-18] . Inhibition
of the miRNA and/or ablation of genes coding for the miRNA in mice attenuated airway inflammation and
reduced tissue eosinophilia. Furthermore, miR-21 has been shown to convey steroid resistant asthma in a
[19]
mouse model, by amplifying the effects of phosphoinositide 3-kinase . As such, low expression of these
miRNAs is congruent with the clinical observations that these subjects in this cluster have mild asthma, low
eosinophil levels, and little need for corticosteroids.

Identification of miRNAs associated with increased asthma severity and eosinophilia
Based on differences in asthma severity and eosinophilia between asthma clusters, our results identified
miRNA signatures associated with clinical features of disease. Expression of Let7 family and miR-98 (i.e.,
group 1 miRNAs) appeared to be most representative of severity; expression was the lowest in Cluster 1 (mild
asthma), higher in Cluster 2 (mild-moderate asthma) and the highest in severe asthma (Cluster 3 and 4).
Regarding an eosinophilic specific miRNA signature, expression of group 2 miRNAs (including miR-155
and miR-21) appeared to be the most useful to define eosinophilia, independent of severity. Expression of
these miRNAs were the highest in the eosinophilic Clusters (2 and 3), and the lowest in the non-eosinophilic
cluster (Cluster 1). The patterns of miRNA expression and clinical correlations are consistent with the
findings in mouse studies discussed above, suggesting conserved pathways between mouse and humans. A
major implication of these findings is that measurement of blood expression levels of these miRNAs at the
time of asthma diagnosis could be useful in predicting outcomes and selection of treatment. For instance,
patients with high levels of group 1 and group 2 miRNAs may require more aggressive treatment targeted to
eosinophils, whereas those with low levels of these miRNAs may have better outcomes with need for lower
doses of ICS. Prospective studies will be essential to validate these hypotheses and determine whether these
miRNA signatures have predictive properties.

MiRNA expression patterns distinguish subsets of severe asthmatics
Severe asthma continues to be challenging to treat, in part because it is not clear whether there are distinct
sub-phenotypes of severe asthma that require different treatment approaches. Our miRNA expression