Discussion
In childhood, asthma is more common in boys than girls but after puberty
the prevalence shifts such that twice as many women have asthma as men
[47]. The timing of this shift
suggests a role for rising levels of female sex hormones such as
estrogens, shown to mediate type 2 cytokine expression
[32,
48].
We found that women with severe
asthma had more circulating Th2 cells than men despite similar amounts
of ICS. We reasoned that in these women Th2 cells may be either less
sensitive to the anti-inflammatory effects of GC or more readily
activated. Our in vitro data support both hypotheses, indicating
that in vivo estrogens may interfere with the ability of GC
therapies to eliminate Th2 cells as well as promote CRTh2 expression,
mediating type 2 inflammation.
Women are more likely than men to have severe asthma and to experience
asthma exacerbations and emergency department visits
[27, 29,
49]. Our study suggests these
longstanding observations may relate to differences in type 2
inflammation since women with severe asthma had more circulating Th2
cells and hospitalizations than severe asthmatic men despite similar ICS
usage. This finding contrasts with Aw et al. who found no
difference in type 2 inflammation between severe asthmatic women and men
[50]. The reason for this discrepancy
may lie in the severity of the asthma population and/or asthma
management. Our study cohort had only a few patients on OCS for more
than 50% of the year (7.5%), while all the patients in Aw et
al. required OCS for 6 months or more
[50], which may have suppressed
differences in systemic inflammation
[51]. Studies have shown women with
mild asthma to have more sputum ILC2s than men prior to and after
allergen challenge [50] and women
with severe asthma to have more blood ILC2s than men
[52]. Neither, however, reported
differences in the proportion of circulating Th2 cells nor assessed the
relationship between type 2 inflammation and asthma severity across a
heterogeneous asthma population. Our findings that blood Th2 cells,
eosinophils and CRTh2 mRNA correlated with clinical characteristics
associated with asthma severity (ICS and FEV1) in women
but not men suggests differences in sensitivity to ICS therapy and/or
propensity for type 2 inflammation.
Several human studies have reported females experiencing less
improvement on ICS than males (reviewed in
[53]). This may be driven by
crosstalk between GR and ERα, known to result in antagonistic or
cooperative effects in a gene-dependent manner
[54,
55]. Our in vitro studies
suggest both mechanisms are at play. The ability of GC to induce
apoptosis was decreased by ERα signaling. This was related to ERα
enhancing the anti-apoptotic response, upregulating BCL-2, and suggests
that in vivo estrogens may antagonize this
anti-inflammatory aspect of GC function, perpetuating type 2
inflammation. On the other hand, our study is the first to report that
GC and ERα agonist cooperate to upregulate CRTh2 expression. The
CRTh2 promoter contains putative GC response elements (GRE) and so the
ERα enhancing effect may be related to its ability to stabilize GR
binding [55,
56], increasing CRTh2 transcriptional
activity. That increases in CRTh2 mRNA and protein were observed only
with co-treatment suggests the timing of analysis may have failed to
capture a GC-mediated effect and/or that ERα also influences mRNA
stability and translation of CRTh2. Although we focused on Th2 cells,
CRTh2 is expressed by eosinophils, basophils and ILC2
[10,
11] and so CRTh2 mRNA levels detected
within the blood may collectively reflect expression by all these cell
types.
The GC and ERα mediated increase in CRTh2 levels enhanced Th2 cell
response to PGD2, significantly increasing IL-5 and
IL-13 release. Since CRTh2 activation also mediates chemotaxis
[10, 13,
57], this interaction may lead to more
CRTh2 expressing cells infiltrating the airways following mast cell
release of PGD2 [15,
58, 59].
Though further work is needed to determine whether type 2 inflammation
within the airways is also higher in women compared to men with severe
asthma, this paradoxical pro-type 2 effect of GC treatment has been
previously reported within the airways of female mice
[60,
61]. Together these studies suggest
that in vivo estrogen and GC interaction could drive persistence
of type 2 inflammation by upregulating the CRTh2 pathway, leading to
asthma that is more difficult to treat in women.
Studies examining sex differences in autoimmune disease consistently
report that females have higher CD4+ T cell counts,
CD4+ T cell activity and greater antibody responses
[62]. As such, our finding likely
represents only the tip of the iceberg in terms of how immune responses
in asthma differ based on biological sex
[63]. Mouse models show females
mounting stronger type 2 responses than males
[52, 64,
65] and that sex hormones have opposing
effects on type 2 inflammation. In female mice estrogens induced type 2
inflammation through ERα signaling
[66], while in males androgen
hormones inhibited development of type 2 inflammation
[52,
64]. A head-to-head comparison of
gonadectomy showed that loss of androgens in males lead to higher type 2
inflammation, while loss of ovarian hormones in females lowered levels
of type 2 inflammation [65]. Sex
hormones may also have cell type-specific effects, since androgen
receptor mutant mice showed reduced IL-13 or GATA levels in ILC2s, but
not in Th2 cells [65], while estrogen
receptor signaling increased T cell production of IL-4 and GATA3
[33]. Ultimately, these studies
suggest type 2 inflammation may develop in a sex- and cell type-specific
manner involving different pathways, similar to mechanical pain
hypersensitivity shown to be mediated by microglial cells in male mice
and T cells in female mice [67].
Our study reveals the capacity of estrogen to enhance GC-induction of
CRTh2 and the propensity of Th2 cells to release IL-5 and IL-13.In vivo the higher estrogen levels in women than men (nM vs pM)
interacting with ICS, could explain our clinical observation of
persistent type 2 inflammation in severe asthmatic women. This induction
of CRTh2 would likely increase infiltration of both Th2 cells and ILC2,
which may synergize locally with the ability of ERα to control ILC2s
through epithelial expression of IL-33
[68]. Other mechanisms, however, may
also be involved. Androgens have been shown to suppress type 2
inflammation [64], their levels to
correlate with better lung function in both men and women
[69] and to be influenced by genetic
variation within 3β-hydroxysteroid dehydrogenase-1 (3β-HSD1), an
important enzyme for androgen metabolism
[70]. As such, we cannot rule out the
possibility that severe asthma in women could also involve having lower
androgen levels and/or AR signaling. Future studies in human asthmatics
are required to determine the relationship between sex hormone levels,
type 2 cells and response to GC therapy and will need to be specifically
designed to account for age/stage of life (menstruation to menopause),
phase of the monthly menstrual cycle and oral contraceptive use. Our
findings call for examination of whether women have higher ICS
requirements to control type 2 inflammation. Moreover, studies assessing
CRTh2 levels as a systemic indicator of persistent type 2 inflammation
could help optimize ICS therapy and/or identify patients most likely
responsive to anti-type 2 biologics.
In summary, we report that women with severe asthma had higher
circulating levels of Th2 cells than men with severe asthma. Ourin vitro data show this could be attributed to estrogen and GC
interactions enhancing CRTh2-mediated type 2 inflammation.
Sex-stratified approaches may be necessary to fully unravel the
complexity of type 2-high severe asthma.