To the editor

The recent paper by Partida-Sánchez et al., in which the generation of CD38 deficient mice is reported, shows that CD38 has an important role in the action of neutrophils combating infection1. However, we question whether the decreased ability of these knock-out mice to counter infection is the result of a defect in Ca++ signaling for neutrophil chemotaxis. A different interpretation of these results suggests that CD38 has a role in the Ca++ signaling triggered by a separate aspect of neutrophil behavior.

There are several reasons for suggesting an alternative explanation for increased infection in the CD38-deficient mice. First, although the authors provide evidence for an inhibition of neutrophil chemotaxis in response to formyl-methionyl-leucyl-phenylalanine (fMLP), they found that interleukin-8 (IL-8) signaling remains intact. Therefore IL-8 may have no role in signaling neutrophil chemotaxis in the infected mice. This would seem unlikely as these chemo-attractants, together with C5a, are usually involved in driving neutrophil chemotaxis. Also, given that the receptors for these three chemoattractants are of the same class of seven transmembrane-domain, G protein–linked receptors2, it is possible that they use similar intracellular signaling pathways.

Second, the evidence that Ca++ signaling triggered by fMLP is totally dependent on CD38 activity can only be explained by postulating an interplay with IP3, as there is a wealth of experimental data pointing to IP3 generation. For example, fMLP generates IP3 at intracellular concentrations sufficient to release Ca++ (ref. 3); blockade of IP3 receptors inhibits Ca++ release triggered by fMLP (ref. 4); and knock-out mice devoid of the IP3-generating enzymes phospholipase C (PLC) β2 and β3 are unable to generate Ca++ signals in response to fMLP (ref. 5). In other cell-types, IP3 interacts with additional second messengers, including the product of CD38 activity, calcium-mobilizing metabolite cyclic ADP-ribose (cADPR), to generate Ca++ signals6.

Third, the evidence that Ca++ signaling is crucial for chemotaxis by neutrophils is difficult to reconcile with the fact that fMLP–mediated chemotaxis can be readily dissociated from Ca++ signaling. For example, fMLP triggers chemotaxis at sub-nanomolar concentrations, whereas higher concentrations are required for Ca++ signaling; chemotaxis in response to fMLP occurs without changes in cytosolic free Ca++ concentration7; and chemotaxis by neutrophils in response to fMLP is unaffected in PLC β2/3-deficient mice despite a lack of Ca++-signaling in these mice5.

Although these authors clearly show that CD38 is important for an aspect of neutrophil physiology and signaling, we suggest the relevant process may be β2 integrin–mediated responses. The neutrophil β2 integrin is involved in two key stages in combating infection, both of which have been shown to be signaled through Ca++. β2 is implicated in neutrophil extravasation, whereupon β2 integrin engages intracellular adhesion molecule-1 on endothelial cells, and in phagocytosis of C3bi-opsonized bacteria. A defect in β2-signaling would therefore provide an explanation for both the lack of accumulation of neutrophils in lung and the inability of circulating neutrophils to clear bacteria in Cd38−/− mice (the most dramatic difference between wild-type and Cd38−/− animals Partida-Sánchez et al.1 found was in circulating bacteria). Furthermore, as the generation of IP3 in response to β2 integrin engagement is very low8, and the characteristics of the Ca++ signal are different from those induced by fMLP (ref. 9), an IP3-independent route to Ca++ signaling probably exists. Recently, we have shown that blockade of IP3 receptors by simple lipid-assisted micro-injection (SLAM)10 of intracellular heparin fails to inhibit β2 integrin-mediated phagocytosis and its accompanying Ca++ signal, despite preventing Ca++ signaling by fMLP (http://www.uwcm.ac.uk/uwcm/sr/demo3.gif). Experiments of this type point to a non-IP3-mediated mechanism for the induction of Ca++ signaling in neutrophils.

Clearly, identifying the intracellular pathways used by neutrophils in regulating their program of activity is important in understanding neutrophil behavior. Although many roles have been established for cytosolic Ca++ in neutrophils, a gap in understanding exists for β2 integrin signaling by Ca++. The results of Partida-Sánchez et al.1 suggest that a route involving CD38 must now be high on the list of important candidates to be investigated.

See Reply to “Does neutrophil CD38 have a role in Ca++ signaling triggered by β2 integrin?” by Lund et al.