News & Highlights

The curious case of the intestinal eosinophil

I was taught that eosinophils were hallmarks of immunoglobulin E–mediated immune responses—allergy, atopy, and the like. If they appeared in the intestine in any numbers, this was a result of worm infection or some rare form of eosinophilic gastroenteritis. Of course, like most things we were told at medical school, this was an oversimplification, and it has always been clear that eosinophils could also be a component of other forms of inflammation in the gut, including inflammatory bowel diseases.^{1, 2, 3} Nevertheless, along with others in the field, I have recently been struck by the large numbers of eosinophils that appear to be present in preparations from normal small intestine and colon. Like most “new” findings, this one is not entirely novel—eosinophils have previously been reported in resting human and rodent intestine, although they are not obvious on histology.⁴ However, it seems to be only in the past two or three years that many of us have stumbled on larger numbers of these cells than we anticipated; indeed, recent work suggests that the small intestine contains the largest population of eosinophils in the body.⁵ How has this come about? Have the husbandry and/or diets of laboratory animals around the world changed in some dramatic way that has led to eosinophils becoming a bigger part of the mucosal environment? This seems unlikely, especially when one considers that animal facilities are generally much cleaner than ever and thus much less likely to harbor worms and other infections that might drive intestinal eosinophilia. In fact, our own experience suggests that the numbers of eosinophils obtainable from mouse colon are not at all related to the health status of the mice. Similarly, the methods for isolating mucosal cells, which have not changed significantly in at least the past 20 years, are unlikely to account for marked differences in the kind of cells being obtained. My own theory is that the increase in intestinal eosinophils is a consequence of the increasing sophistication of flow cytometry. Not only does it enable the use of many more markers to identify relatively small populations of cells, but there is less need for enrichment steps such as density gradients or for rigid “mononuclear” gates to exclude debris. As a result, granular cells such as eosinophils are more obvious than before. It is also likely that mucosal eosinophils have been masquerading as other cell types for many years, because they can express macrophage/dendritic cell markers, including CD11c, CD11b, and F4/80 (refs. 5, 6). Hence, they may have gone unrecognized when only a few surface markers were analyzed. In our laboratory, around one-third of F4/80⁺ cells in mouse colon have the appearance of eosinophils, as defined by forward and side scatter and expression of Siglec-F, a marker that is highly specific for eosinophils⁷ (Figure 1). Others have shown that these also express CCR3, the receptor for the eosinophil-selective chemokine CCL11 (eotaxin 1).⁵ Thus, although the numbers found in isolated cell preparations may overestimate their true abundance in situ, because the isolation protocols may allow them to be released more easily, there is clearly an unsuspected abundance of eosinophils in the normal mucosa.

Figure 1

Eosinophils make up a significant component of the resting colon lamina propria in mice and express the F4/80 macrophage marker. Around 30% of the live-gated CD45⁺ cells that can be isolated from colonic lamina propria are F4/80^int class II MHC⁻ cells (dark gray). These are also CD11b⁺, but they have the forward and side scatter properties of granulocytes and express Siglec-F. In comparison, conventional colonic macrophages (light gray) are F4/80⁺ class II MHC⁺, CD11b⁺ Siglec-F⁻. 7-AAD, 7-amino-actinomycin D; FSC, forward scatter; MHC, major histocompatibility complex; SSC, side scatter. (Figure courtesy of C.C. Bain)

What is the biological significance of this? It important to note that eosinophils have a variety of functions in addition to their classic roles in immunoglobulin E–mediated allergy and protective immunity. They contain and release a wide variety of preformed cytokines and other mediators, in addition to those classically associated with T-helper cell type 2–driven inflammation. Indeed, their major role in the body may be in tissue remodeling and repair.⁸ In the gut, they may contribute to epithelial renewal and barrier integrity or perhaps are a source of the “conditioning” factors that maintain local macrophages and dendritic cells in their usual hyporesponsive state. A homeostatic role for mucosal eosinophils is supported by their abundance and long life span in normal gut in vivo.⁵ Eosinophils in the gut are distinct, both phenotypically and in turnover kinetics, from those in other tissues,⁵ making it difficult to predict their functions on the basis of what we know from tissues such as the lung. A final caution raised by the new awareness of this population is that single phenotypic markers should no longer be used to identify cells such as macrophages and dendritic cells in the gut.

All in all, what was once considered an exotic and specialized cell type may turn out to be much more important than we were originally led to believe.