The Role of Rac and Pak in Neutrophil Histone Hypercitrullination and Neutrophil Extracellular Traps Formation

Under specific activating conditions, polymorphonuclear neutrophils (PMNs) release neutrophil extracellular traps (NETs) composed of decondensed chromatin lined with microbicidal protein such as neutrophil elastase and myeloperoxidase. NETs contribute to innate immunity but can also foster autoimmune diseases and thrombus formation. NET formation (NETosis) requires reactive oxygen species (ROS) production by NADPH oxidase and histone hypercitrullination by peptidylarginine deiminase 4 (PAD4), allowing for chromatin decondensation. Rac GTPases are expressed in three isoforms: Rac1 is ubiquitously expressed and plays a role in PMN migration and oxidase function; Rac2 is hematopoietic-specific and the major isoform in PMNs and Rac3 is mostly neuronal. Rac1 and Rac2 regulate the cytoskeleton in PMNs, controlling actin polymerization, cell shape, adhesion and migration and are essential components of the NADPH oxidase complex. The present study aimed to explore the role of the Rac pathway on NETosis in PMNs, including the upstream guanosine exchange factor (GEF) activator, Vav, and a downstream effector of Rac, p21 activated kinase, Pak. We developed a flow cytometry-based quantification of H3 hypercitrullination (H3Cit). In response to phorbol myristate ester (PMA) stimulation, H3Cit is increased to 136% of basal in WT cells, compared with 103% in Rac2-/- (P<0.01) (Table). H3Cit levels observed by flow were confirmed in a NET formation assay. Rac2-/- PMNs formed significantly fewer NETs both spontaneously and after PMA stimulation (WT unstimulated 2.79%, Rac2-/- unstimulated 0.72%, WT+PMA 10.84%, Rac2-/-+PMA 1.39%, P< 0.05 for all pair comparisons). Furthermore, Rac2-/- mice demonstrated a trend towards reduced frequency of provoked thrombosis in an in vivo vena cava stenosis model (WT 78% and Rac2-/- 56% of mice with thrombus). Deletion of floxed Rac1 sequences in a Rac2-/- background in vivoallows generation and purification of PMNs lacking both Rac isoforms. Rac1Δ/Δ,Rac2-/- PMNs, which are defective in actin polymerization, had reduced basal H3Cit and a nearly complete lack of PMA-induced increase in H3Cit (136% vs 69%, WT vs Rac1Δ/Δ,Rac2-/-, P<0.01) (Table). Null knockouts of the GEFs Vav1 (hematopoietic-specific), Vav2, Vav3 or both Vav1 and 3 did not impair H3Cit response to PMA (Table). We next studied downstream effectors of Rac. Group A Paks include Pak1, 2 and 3 isoforms. Pretreatment of wild-type PMNs with either PF3758309 or IPA-3, two group A Pak inhibitors with distinct mechanisms of action, led to reduced H3Cit after PMA stimulation (induction reduced from 36% to 11% for both PF3758309 and IPA-3 treated (Table). To validate this in a genetic model, we studied Pak2Δ/Δ PMNs, since we have recently demonstrated the dependence of hematopoietic stem cell migration on Pak2. Pak2Δ/Δ demonstrated a reduced basal level of H3Cit and a significantly reduced PMA-induced increase in H3Cit (136% vs 94%, WT vs Pak2Δ/Δ, P<0.05, Table). In summary, we describe a flow-based assay that quantitates the early processes of NET formation and validated that this assay reliably predicts agonist-induced NET formation in a genetic model. The results establish that both Rac1 and Rac2, and the downstream effector Pak2, regulate histone H3 hypercitrullination and NET formation in PMNs, while suggesting that Vav does not activate the Rac pathway in PMA-induced NET formation. These data further delineate the role of the Rac pathway in NETosis, linking cytoskeleton and oxidase functions. Furthermore, these data indicate Pak could represent a therapeutic target for a wide array of pathological processes related to NETosis such as thrombosis and numerous autoimmune diseases.