Actin powers the neutrophil traps

In this issue of Blood, Sprenkeler et al¹ demonstrate that actin polymerization is essential for the formation of neutrophil extracellular traps (NETs). The actin cytoskeleton powers a myriad of cellular processes in eukaryotic cells. In no other cell type is actin more relevant than in neutrophils. These essential phagocytes rapidly mobilize to sites of infection to capture and kill pathogens. Over their lifespan, neutrophils must leave the bone marrow, revert to a quiescent state in the circulation, polarize once again to invade tissues, and hunt small microbes or stretch to great lengths to tackle large microbes. As a testament to the potent actin dynamics under the hood, neutrophils move faster than any other human cell and can wield particles many times their size. Eventually, most neutrophils depolarize and die or journey to other distal sites. This complex lifestyle relies on robust actin dynamics and their precise spatiotemporal regulation. However, whether and how actin influences the release of NETs, one of the neutrophil’s most unique cell biological feats, remained incompletely understood.

NETs are decondensed extracellular chromatin structures deployed by neutrophils to combat pathogens but also contribute to pathology. NET release occurs primarily via a cell death process termed NETosis that progresses through the disassembly of the nuclear envelope, cell lysis, and chromatin decondensation of epic proportions. Chromatin posttranslational modifications, such as histone citrullination, are implicated in chromatin decondensation as well as the translocation of the protease neutrophil elastase (NE) from the azurophilic granules to the nucleus, where it partially cleaves histones to disrupt nucleosome organization.² The release of NE from granules is triggered by reactive oxygen species (ROS) via the dissociation from the membrane-associated azurosome complex and inflammasome-dependent gasdermin D activation.^3,4 In the cytosol, NE interacts with and degrades the actin cytoskeleton, driving cell depolarization, facilitating cell lysis, arresting phagocytosis, and committing cells to NETosis.⁴ An early report, however, suggested that pharmacological inhibition of actin dynamics blocked NETosis.⁵ While the limited amount of data demonstrated that cytochalasin D blocked histone citrullination, the inhibitor did not affect chromatin decondensation.⁶ Moreover, the upstream regulator of actin polymerization Wiscott-Aldrich syndrome protein is reportedly required for NETosis but is also able to suppress spontaneous NETosis.^6,7

To clarify these discrepancies, Sprenkeler et al systematically interrogated the requirement for actin polymerization in NET formation using pharmacological and genetic approaches. The authors first evaluated how actin cytoskeleton inhibitors impacted F-actin dynamics in activated neutrophils. These studies indicated that only latrunculin B and jasplakinolide completely disrupted polarized actin polymerization at the leading edge. Other compounds such as CK-666, an inhibitor of the actin filament nucleator Arp2/3, cytochalasin-B, or SMIFH-2, reduced F-actin accumulation and partially decreased the formation of the leading and trailing edge. In contrast, except for CK-666, all compounds blocked NETosis induced by the soluble protein kinase C agonist phorbol myristate acetate (PMA), or the complement component 5a (C5a). Fungal filament–induced NETosis was abrogated by all compounds, although this may also be attributed to a substantial reduction in interactions between neutrophils and hyphae. Importantly, neutrophils derived from donors with mutations in the Arp2/3 subunit ARPC1B were unable to form NETs in response to PMA, C5a, MSU crystals, and hyphae although they polarized and attached to extracellular particles. Similar findings were obtained in MKL1-deficient cells that have low actin cytoskeleton gene expression.

To understand how actin polymerization supports NETosis, the authors dissected how the inhibitors affected processes implicated in NETosis. One possible scenario involved a role for the actin cytoskeleton in the assembly of the NADPH oxidase, the multi-subunit enzyme producing superoxide. In addition, the authors examined chemotaxis and NE nuclear translocation. Overall, the ability of actin polymerization inhibitors to interfere with ROS production depended on the stimulus, but most compounds did not block ROS production. Only jasplakinolide blocked ROS induced by PMA or yeast, whereas ROS induced by hyphae was sensitive to jasplakinolide or SMIFH2. In fact, mobilization of most neutrophil granules increases upon pharmacological inhibition of actin, in line with the idea that actin filaments interfere with access to the plasma membrane. On the contrary, ROS is required for actin polymerization.⁷ All compounds blocked chemotaxis, except for CK-666. Similarly, Arp2/3 deficiencies did not affect ROS production and, although they decreased neutrophil chemotaxis rates, neutrophils were still able to phagocytize yeasts and bind crystals and hyphae. Importantly, jasplakinolide and latrunculin B, the two most potent inhibitors of actin polymerization, blocked NE translocation to the nucleus.

Hence, different stimuli and cellular processes appear to have specific dependencies on actin polymerization as reflected by the variable sensitivities to actin cytoskeleton inhibitors. Actin polymerization may influence several steps in NE translocation, such as the release from granules or entry into the nucleus. Both processes occur in the absence of additional actin components in cell-free systems.² However, intact cells are more complex. The interaction of NE with F-actin may have a greater mechanistic significance because it may act as a cytosolic sink to sequester NE out of granules.⁴ Jasplakinolide stabilizes F-actin in vitro but can disrupt actin networks in vivo. The fact that these inhibitors reduce F-actin largely rules out that the observed effects are due to sequestration of NE into stabilized disorganized actin filaments. Actin polymerization is also required for entry of other proteins such as PKC into the nucleus. In addition, the authors demonstrate a requirement for actomyosin contractility and discuss how it may be implicated in nuclear envelope disassembly, as observed during mitosis. Another interesting hypothesis might be that it is not cytoplasmic but rather nuclear actin polymerization that plays the critical role by supporting vital cellular functions such as DNA repair, which has been implicated in NETosis.^8,9 Alternatively, actin polymerization may potentiate inflammasome activation¹⁰ or membrane signaling or remove other cellular roadblocks to NE translocation.

While these exciting mechanistic questions remain to be addressed, the systematic approach undertaken by Sprenkeler et al demonstrates a requirement for actin polymerization in NETosis independently of ROS production and cell migration. Moreover, the study hints that actin networks influence neutrophil cell biological processes in distinct ways that can be dissected by different pharmacological strategies.