Abstract
Neutrophils (PMN) are a critical cell in inflammation. In response to external stimuli, they activate various signaling pathways to move rapidly to a site of microbial invasion and perform phagocytosis, cytokine and reactive oxygen species release. Rho GTPases, Rac1, Rac2, CDC42 and Rho, are central regulators of cell movement via actin rearrangement. We have shown the specific role of Rac1 and Rac2 in PMN functions (Gu, Science 2003; Filippi, Nat Immunol, 2004) which raises the question of the role of other Rho GTPases in PMN functions. CDC42 primarily regulates filopodia formation and controls cell polarity and migration in non-hematopoietic cells and some hematopoietic cell lines. Most of previous studies have used dominant active or negative mutants which lack specificity and cannot be used to define in vivo cell biology. Here, we used mice genetically deficient in the CDC42 negative regulator CDC42 GTPase Activating Protein (GAP) to study the role of CDC42 in PMN functions in vitro and in vivo. PMN deficient in CDC42GAP (CDC42GAP−/−) displayed a 2-fold increase in CDC42 activity. In vivo recruitment of PMN in peritoneal cavities was significantly higher in CDC42GAP−/− animals than in WT mice (4.5 ± 0.3x106 vs 3.4 ± 0.2x106, p<0.05) indicating that CDC42 plays a physiological role in neutrophil migration. We examined F-actin assembly upon integrin ligation. Podosome-like structures identified by a vinculin ring surrounding F-actin that are present at the leading edge in WT PMN were significantly reduced in frequency in the mutant cells (15% vs 3%). In addition, CDC42GAP−/− PMN showed increased lateral filopodia-like formation and abnormally elongated uropod with tail filopodia. Thus, CDC42GAP−/− PMN appeared less polarized than WT PMN (50% vs 16%). This abnormal F-actin assembly was associated with abnormal cell motility. In vitro, CDC42GAP−/− PMN showed increase random movement (chemokinesis) compared with WT PMN. By contrast but similar to the loss of CDC42 activity, CDC42GAP−/− PMN displayed defective directed migration towards fMLP suggesting that CDC42 activity plays a critical role in both chemokinesis and directed migration. These functions may be regulated by podosome-like and filopodia formation respectively. To further understand this correlation at a mechanistic level, we examined MAPK signaling. CDC42GAP−/− PMN showed sustained ERK phosphorylation at 15min compared to WT PMN. By contrast, p38MAPK was significantly decreased in CDC42GAP−/− PMN compared to WT at both 5 and 15min. Pharmacological inhibition of ERK activity in CDC42GAP−/− PMN using U0126 rescued the abnormal increased chemokinesis to level similar to WT and was associated with partial rescue of podosome-like formation at the leading edge of the cells. Inhibition of p38MAPK activity in WT PMN using SB203580 reduced directed migration and was associated with increased tail filopodia that mimicked CDC42GAP−/− PMN. Taken together, these results suggest that CDC42GAP plays an important role in PMN chemokinesis and directed migration likely via distinct signaling pathways. CDC42GAP may control chemokinesis via ERK-mediated podosome-like turnover at the leading edge. CDC42GAP may regulate directed migration by inhibiting filopodia at the uropod via p38MAPK and subsequently by restraining filopodia to the leading edge. This reinforces the importance of turnover of attachment structures during cell movement and suggests a new role for CDC42 in attachment structures in neutrophils and for p38MAPK in CDC42-mediated directed migration.
Mechanism and cellular function of Bud6 as an actin nucleation–promoting factor
