Eosinophilia is a common clinical problem associated with numerous disorders such as atopic diseases, parasitic infections, hypereosinophilic syndrome (HES), and cancer. A marked decrease in circulating eosinophils, or eosinopenia, has long been associated with acute bacterial infections. Eosinopenia has been shown to be a sensitive and reliable marker for distinguishing between non-infectious and infection-associated sepsis in the intensive care unit setting. Eosinopenia with acute infection had been assumed to be secondary to the release of adrenal glucocorticoids in response to the stress of infection. However, animal studies performed in the 1970s demonstrated that inhibition of parasite-associated eosinophilia with acute bacterial infection was independent of adrenal stimulation; the investigator speculated that it was due to a direct effect on the eosinophil precursor. To date, the mechanism of infection-associated eosinopenia, and its potential for suppressing IL-5-mediated eosinophilia, has not been investigated.
To study eosinophil development, we established an ex vivo liquid culture system in which we can follow the differentiation of murine eosinophil lineage-committed progenitors (EoPs) to mature functional effector eosinophils that express known eosinophil surface markers and granule proteins and respond to eosinophil-active chemoattractants. We demonstrate that murine eosinophil precursors express mRNA for six TLRs. Ligands for TLR2 heterodimers, TLR4 and TLR7, but not TLR3, inhibited eosinophil ex vivo growth by attenuating EoP proliferation in response to IL-5 stimulation. Exposure to heat-killed E. coli also resulted in diminished EoP proliferation. The developing eosinophils further responded to TLR activation with production of a subset of inflammatory cytokines, including IL-6, IL-10, TNF-alpha and CXCL10. Co-culturing of untreated EoPs with LPS-exposed EoPs resulted in reduced IL-5-mediated proliferation of the untreated EoPs, suggesting that EoPs secrete mediators in response to TLR activation that regulate their proliferation in an autocrine/paracrine manner. Neutralization of IFN-beta, but not IFN-alpha or IFN-gamma, partially protected developing eosinophils from the inhibitory effects of LPS. In vivo studies revealed that endotoxemia and bacteremia reduced numbers of EoPs in the bone marrow of wild-type mice. Further, LPS dose-dependently reduced hypereosinophilia in IL-5 transgenic mice, highlighting the potential therapeutic value of this approach even in an extreme IL-5-driven clinical state. Suppression of eosinophil production by TLR4 ligands is conserved between mice and humans as yield of eosinophils from human EoPs was reduced following LPS exposure. Taken together, these findings identify a mechanistic explanation for eosinopenia following bacterial infections and identify a novel therapeutic strategy for inhibiting eosinophil production and peripheral eosinophilia in eosinophil-associated diseases.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.