We thank Drs. Pillay, Ulfman, and Koenderman for their comments and are pleased that they appear to accept our major conclusion that certain soluble ligands of αMβ2, including fibrinogen, delay neutrophil apoptosis by engaging both subunits of the integrin. The role of soluble fibrinogen in supporting neutrophil survival also is in agreement with other reports (1, 2). The letter indicates that we cannot “exclude that the marked ERK1/2 and Akt/PKB signaling is a result of priming of PMA-induced signaling.” We have not stated nor implied that soluble integrin ligands alone would trigger appreciable intracellular signaling in resting cells, as they do not bind with sufficient affinity to αMβ2. However, the dose of PMA used itself also induces very little ERK1/2 and Akt/PKB phosphorylation in the absence of fibrinogen. We believe that our observations are consistent with a well-established paradigm in integrin signaling: low concentrations of an agonist, in this case PMA, activate the integrin via PKC-dependent inside-out signaling, allowing soluble fibrinogen to bind to αMβ2 (3, 4, 5). ERK1/2 and Akt/PKB phosphorylation depends on outside-in signaling arising from occupancy of the integrin since inhibitors of integrin clustering, a consequence of ligand binding, blocked PMN survival. This scenario is similar to that observed upon fibrinogen binding to the major platelet integrin αIIbβ3, where activation, occupancy, and clustering of the integrin are all required for such signaling events. We do not dispute that PMA at higher levels that those used in our study can induce ERK1/2 (6), although direct activation of Akt/PKB does not appear to occur in human neutrophils. Thus, we believe that both an integrin-activating stimulus and a soluble ligand that engages both subunits of αMβ2, such as fibrinogen and several others, are necessary to elicit the prosurvival signals.
The second issue raised by Drs. Pillay, Ulfman, and Koenderman is that contact of primed circulated neutrophils “with physiological concentrations of soluble fibrinogen…would amplify rather than inhibit the systemic inflammatory response, which is an undesired clinical condition.” Because neutrophil functions are fine-tuned by many interactions between various ligands (matrix/plasma proteins, chemokines, cytokines) and their receptors, caution needs to be exercised in making global predictions about pro- and anti-inflammatory consequences in the in vivo setting. Certainly, it can be envisioned that only a small subpopulation of circulating neutrophils might respond to a stimulus rather than the entire population of circulating neutrophils. Despite a heightened inflammatory state in patients with acute myocardial infarction or cardiac surgery, αMβ2 activation was only slightly augmented on circulating leukocytes (∼10% of neutrophils were positive for an activation-dependent αM epitope) (7, 8). This subpopulation of neutrophils could very well have αMβ2 occupied by fibrinogen or other ligands that would selectively prolong their survival. We certainly agree that fibrin, an immobilized ligand, might be a preferred ligand of αMβ2, but not to the complete exclusion of fibrinogen or other soluble ligands of the integrin.