Overall, the data presented here indicate that human basophils generate IL-3 with parameters identical to those described for their production of IL-4 (11
). In particular, increases in IL-3 mRNA were seen as early as 15 minutes following stimulation with anti-IgE Ab, with evidence of IL-3 secretion within the first hour and peaking by 4h after activation. Of course, the rapid secretion of IL-3 only became apparent after blocking its absorption using IL-3R Ab (). It therefore goes to reason that the rapid production of IL-3 plays an important role in allowing for autocrine priming for subsequent effector functions (as discussed below). Another feature of IL-3 production by basophils that bears remarkable similarity to that of IL-4 is its strong link to IgE/FcεRI-dependent activation. Basophils responding well to anti-IgE stimulation (with regard to histamine release) up-regulated IL-3 mRNA by as much as several thousand-fold after 1h and were also more likely to secrete protein for this cytokine ( and ). In contrast, common basophil secretagogues such as C5a, FMLP and PMA that are all quite capable of activating basophils for mediator release did not induce detectable IL-3 protein, nor did these stimuli greatly affect IL-3 mRNA levels compared to cells cultured in medium alone (). This dissociation of IL-3 secretion from histamine release has long been observed for IL-4 and once again emphasizes that little if any of these two cytokines is pre-formed in basophils and released during degranulation (1
We predict that changes in cytosolic calcium are important for IL-3 generation (just as they are for IL-4), since ionomycin markedly induced secretion of this cytokine (). In fact, IL-3 secretion exceeded 4000 pg/106 basophils when using this stimulus during a 24h incubation, and likely eliminated the need for IL-3R Ab to block absorption, since receptors were most certainly saturated at these levels. Based on these findings further demonstrating the similarities in the parameters important for IL-3 and IL-4 secretion by basophils, we also predict that IL-3 is controlled pharmacologically much like what has been reported for IL-4, even though studies to definitively show this are yet to be done.
We have also previously reported that TLR2 ligands, such as PGN, enhance both IgE-dependent and IgE-independent responses in basophils much like IL-3 (9
). We therefore considered that this mode of innate immune stimulation might prime basophils by inducing IL-3 that would then mediate autocrine effects. However, our results indicated that PGN (and LPS) only poorly induced IL-3 mRNA compared to those seen with IgE-dependent activation. In addition, basophils did not secrete detectable IL-3 when cultured with these TLR ligands (). It seems appropriate to conclude that the priming of basophil responses reported for TLR2 agonists is not the result of autocrine IL-3 effects. It is also important to note that we also considered the possibility that IL-5 and GM-CSF could mediate basophil priming by inducing autocrine IL-3. However, IL-3 mRNA was not induced in basophils treated with these cytokines (JTS, unpublished data). Like the TLR2 ligands, these substances appear to augment basophil responses independently of IL-3 generation.
The concept that IL-3 production is linked to IgE/FcεRI-dependent activation was first suggested nearly 20 years ago using murine mast cell lines (13
). In those studies, mast cells were reported to up-regulate IL-3 mRNA as well as the expression of message for IL-4, IL-5 and IL-6 when subsequently stimulated through FcεRI or activated with calcium ionophore. A more recent study reports that IL-3 is secreted by mouse mast cells even in response to monomeric IgE alone, which acts in an autocrine fashion to extend survival (14
). However, it is important to note that the bone marrow-derived mast cells used in these studies were in fact first exposed to exogenously added IL-3 in order to induce their development from precursor cells. It therefore seems possible that this added IL-3 may have played a role in regulating its own production. Moreover, unlike the mouse data which implicates the importance of IL-3 in mast cell development and function, recent evidence indicates that IL-3 does not affect the differentiation of human mast cells (15
). Thus, these issues raise questions regarding the translatability of the two models, particularly as it pertains to the role of IL-3 in mast cell vs. basophil function.
Nonetheless, the results presented in the current study are not unlike those seen with murine mast cells in that they clearly show that human basophils also rapidly generate mRNA and secrete protein for IL-3 within the first 4h following IgE-dependent activation. However, the basophil studies are clearly more relevant to clinical disease in demonstrating that they utilize the IL-3 they produced. For example, in vitro
studies have long shown IL-3 to be the most potent cytokine to affect human basophil function and that it does so by binding receptors densely expressed on their cell surface. It is therefore plausible that these cells may very well regulate their own priming in vivo
. The importance of this hypothesis is underscored by the long-standing belief that in vitro
basophil releasability is clinically relevant (16
). In fact, many studies conducted during the past 30 years have shown evidence that basophils from allergic subjects are primed both functionally and phenotypically when compared to cells from normal subjects.
To investigate the possibility that basophils mediate their own priming, we explored whether the IL-13 generated by these cells following IgE-dependent activation is, indeed, secondary to IL-3 secretion. It seemed possible that the prolonged kinetics (~24h) originally reported for optimal IL-13 secretion (17
) vs. the time course for IL-4 (~4h) (12
) in response to IgE-dependent activation is actually due to the actions of autocrine IL-3. By neutralizing IL-3 activity using IL-3R Ab, we were able to demonstrate that ~30% of the IL-13 secreted during a 16h incubation was dependent on autocrine activity (). Of note, we accurately predicted that neutralizing IL-3 would have less of an effect on IL-4 (~19% reduction) and histamine (no change), since release of these mediators, respectively, occur either simultaneously or after IL-3 is produced. However, it is also fair to conclude that much of the IL-13 produced during IgE-dependent activation is in fact independent of IL-3. This finding thus supports our long-held belief that at least 2 pathways exist in basophils for the production of IL-13 –one that is FcεRI-mediated and one dependent on IL-3 (1
). Therefore, the IL-3 produced by basophils likely functions, in part, to augment and/or prolong the secretion of IL-13 that is initiated upon IgE-dependent activation.
Evidence that basophils might mediate their own priming is also supported by the finding that autocrine IL-3 induces the expression of CD69 following IgE-dependent activation. First, it is important to note that Suzukawa, et al. have recently shown that low-level FcεRI-dependent activation up-regulates CD69 expression on basophils, but only under conditions that involved the addition of IL-3 along with the anti-FcεRIα Ab used in these studies (18
). Our results confirm that basophils express CD69 following IgE-dependent activation, but additionally show that IL-3 does not need to be added exogenously to do so (). In observing the relatively slower kinetics for CD69 induction (≥1h) compared to the minutes for the up-regulation of degranulation-specific markers (i.e. CD63 and CD203c), we then hypothesized that the former is regulated indirectly by autocrine IL-3. Indeed, this was our conclusion after confirming that the addition of IL-3R Ab markedly inhibited CD69 induction resulting from IgE-dependent activation ().
Overall, these in vitro
findings demonstrating that autocrine IL-3 modulates basophil IL-13 secretion and CD69 expression could help interpret clinical observations. Several studies have shown that basophils produce IL-13 in direct response to IL-3 (3
) and that this reaction is greater using cells from allergic vs. normal controls (19
). Circulating basophils have also been shown to spontaneously secrete IL-13 following experimental allergen challenge in the nose compared to pre-challenge responses (20
). Likewise, Yoshimura, et al. first reported constitutive expression of CD69 on the basophils of asthmatics compared to normal subjects (21
). This has since been extended to include basophils prepared from allergic rhinitics and subjects with chronic idiopathic urticaria (22
). We have also shown that circulating basophils from subjects allergic to insect venom express greater levels of CD69 following a controlled insect sting compared to pre-sting levels (23
). Therefore, it seems feasible that the basophils from these subjects are primed for greater IL-13 secretion and expression of CD69 having come in contact with IL-3. More significantly, the data presented here would imply that the source of this IL-3 could very well be the basophil itself, and that low-level IgE/FcεRI cross-linking resulting from allergen exposure is stimulating this autocrine response at a systemic level. Support for this hypothesis may be forthcoming from studies that investigate whether IgE depletion using omalizumab reduces these markers (i.e. IL-13 and CD69) of basophil priming.
In conclusion, we have shown that human basophils rapidly generate and secrete IL-3 upon IgE-dependent stimulation. Through autocrine activity, this IL-3 is capable of priming basophils for functional and phenotypic properties characteristic of basophils isolated from allergic individuals. We predict that autocrine IL-3 activity resulting from low-level IgE/FcεRI cross-linking by specific allergen represents an important mechanism behind the hyper-reactive nature of basophils that has long been observed in allergic disease.