Go with the flow: GEF-H1 mediated shear stress mechanotransduction in neutrophils

ABSTRACT

Neutrophils in circulation experience significant shear forces due to blood flow when they tether to the vascular endothelium. Biochemical and biophysical responses of neutrophils to the physical force of flowing blood modulate their behavior and promote tissue recruitment under pro-inflammatory conditions. Neutrophil mechanotransduction responses occur through mechanisms that are not yet fully understood. In our recent work, we showed that GEF-H1, a RhoA specific guanine nucleotide exchange factor (GEF), is required to maintain neutrophil motility and migration in response to shear stress. GEF-H1 re-localizes to flottilin-rich uropods in neutrophils in response to fluid shear stress and promotes spreading and crawling on activated endothelial cells. GEF-H1 drives cellular contractility through myosin light chain (MLC) phosphorylation downstream of the Rho-ROCK signaling axis. We propose that GEF-H1-dependent cell spreading and crawling in shear stress-dependent neutrophil recruitment from the vasculature are due to the specific localization of Rho-induced contractility in the uropod.

Introduction

Neutrophils are important early responders of the innate immune system that are recruited to sites of infection from the circulation. To make their way to the site of infection they must recognize specific signals on the inflamed vascular endothelium, roll, adhere, crawl and migrate across the endothelial barrier. Furthermore, neutrophil responses on the endothelial surface are exquisitely sensitive to shear forces due to blood flow. In our recent work, we uncovered an important role of the guanine nucleotide exchange factor (GEF), GEF-H1, in neutrophil recruitment to the inflamed peritoneal cavity.¹ Specifically, we demonstrated a role of GEF-H1 in neutrophil spreading and crawling in response to shear stress. Here we discuss our findings and review available evidence that provide insights in the regulation of mechanotransduction responses induced by shear forces.

Neutrophils, shear stress and mechanotransduction

Neutrophils are effector cells of the innate immune system that are critical for the response to inflammation. Their efficacy in providing protection against invading pathogens relies on their ability to rapidly get recruited from the blood to the site of infection. This recruitment is accomplished by adhesion to the activated vascular endothelium through a series of overlapping attachment processes that include selectin and integrin mediated capture. These events lead to neutrophil activation and their transmigration across the endothelial wall, which is followed by chemotactic migration through the interstitial space to the site of infection. At the site of infection neutrophils destroy invading pathogens through a variety of functions including phagocytosis, secretion of reactive oxygen species (ROS) and other cytotoxic molecules and secretion of neutrophil extracellular traps (NETs). Neutrophil activation and recruitment is a finely tuned process, as excessive neutrophil activity can result in significant tissue damage and increased morbidity.²

Mechanotransduction, the ability of cells to translate mechanical forces into biochemical responses, plays an essential role in many biologic processes, including embryonic³ and tissue development,⁴ cell fate determination,⁵ cell migration,⁶ cell morphology,⁷ cytokine activation,⁸ gene expression,^5,9 and cancer cell invasion.¹⁰ Neutrophils and other circulating leukocytes have evolved in the context of shear forces due to blood flow, consequently, they have adapted mechanosensitive molecular switches that trigger appropriate biologic responses in a shear force sensitive manner.^11-14 Leukocytes experience the strongest shear forces as they interact with and tether to the endothelial surface.

Evidence that appears to be in conflict suggests that mechanical and shear forces can serve to either activate or deactivate neutrophils.¹⁵ Shear forces have been shown to promote or limit neutrophil responses and recruitment to the endothelial surface depending on experimental context. In-vivo, in the normal healthy vasculature, in the absence of inflammation, shear was shown to induce pseudopod retraction,¹⁶ a response that depended on the presence of red blood cells.¹⁷ However, in a spontaneously hypersensitive rat (SPR) model fluid shear stress had the opposite effect.¹⁸ Centrifugation, which is known to stimulate neutrophils,¹⁹ and treatment with glucocoritcoids also reversed the shear stress response.^20,21 Neutrophils plated under inflammatory conditions (i.e. activated neutrophils) had elevated crawling and extravasation,¹¹ and increased neutrophil invagination into the apical endothelial interface,¹² in response to shear. HL-60 cells differentiated into neutrophils, showed retraction in response to shear stress, however when the cells had been treated with f-Met-Leu-Phe (f MLP) they remained spread.²² In contrast to the observations above, one study using passively isolated human neutrophils, indicated that even in the absence of priming neutrophils have a cell spreading response within minutes of exposure to shear stress.²³ Since shear stress helps to maintain the naïve phenotype of circulating neutrophils,¹⁶ it is possible that the lack of shear stress, after blood is removed from the vasculature, is sufficient to allow some neutrophil activation. Once these cells are re-exposed to shear stress they respond as activated cells. Alternatively, it is possible that some degree of activation occurred due to neutrophil interaction with glass coverslips or trace amounts of endotoxin.²⁴ Overall the literature suggests that shear stress promotes the rounded-up phenotype of naïve neutrophils in the circulation, acting to limit neutrophil recruitment in the absence of inflammatory cues. However, neutrophils that become activated as they encounter pro-inflammatory cues at the endothelial surface, or isolated neutrophils, which are likely to experience some degree of activation during purification,²⁴ exhibit the opposite response to shear stress, namely, increased spreading, crawling and extravasation.

Although further work is necessary to elucidate the receptors that are responsible for mechanosensation, evidence implicates selectins,²⁵ integrins,^26-28 stretch activated calcium channels,²⁹ and G-protein coupled receptors (GPCRs),³⁰ including the f MLP receptor.³¹ The nucleus has also been proposed to have mechanosensory potential.^32,33 The actin and microtubule (MT) cytoskeletons, polymeric structures that determine cell morphology, polarity and migrational capacity,³⁴ are key downstream effectors of mechanotransducive signaling. While integrins are implicated as mechanosensory organelles in some contexts, they are also likely to be important downstream targets of the mechanosensory response. Shear stress induced strengthening of integrins by anchoring to cytoskeletal components was shown to be an important mechanism to enhance leukocyte adhesion in response to force.³⁵

Rho-family small GTPases and cell motility

Successful extravasation of leukocytes to a site of infection requires highly orchestrated regulation of cell attachment, morphology, and polarity in response to extracellular cues. This is achieved through control of integrin based adhesions, cytoskeletal dynamics, and cell contractility. Maintaining traction for efficient cell migration depends on coordination between adhesion strengthening, protrusion at the leading edge, contraction of the cell body, and de-adhesion in the tail,³⁶ and requires cyclical regulation of integrins and their association with the contractile F-actin cytoskeleton. Rho family small GTPases, such as RhoA, Rac1, and Cdc42, and their activators, the GEFs, are central regulators of these processes and control cell migration in fast moving innate immune cells³⁷ and slow moving fibroblasts.³⁸ Rho family small GTPases are important regulators of the cytoskeleton, and their roles in directed cell migration, polarity and the forces of adhesion, propulsion and retraction have been studied in various cell types.^38,39 RhoA is an important regulator of the cellular contractile response. In its GTP bound form, RhoA activates its downstream effector ROCK, which in turn activates the myosin light chain (MLC) through direct phosphorylation,⁴⁰ or by inhibiting myosin light chain phosphatase (MLCP).⁴¹ ROCK can also activate LIM kinase (LIMK) through direct phosphorylation. In its active state, LIMK phosphorylates cofilin, thus inactivating it and preventing cofilin dependent depolymerization of actin.⁴² Another RhoA effector, mDia, directly catalyzes F-actin polymerization.⁴³ In fibroblasts and epithelial cells, the RhoA-induced contractile response, characterized by cell contraction, and the formation of actin stress fibers and focal adhesions, is induced by thrombin,⁴⁴ lysophosphatidic acid (LPA),⁴⁵ and disruption of the MT cytoskeleton.⁴⁶ Interestingly, cell contraction itself drives stress fiber and focal adhesion formation in fibroblasts cells.⁴⁷ Using transformed fibroblasts, RhoA dependent focal adhesion strengthening was demonstrated as a biochemical response to intracellular contractile tension, associated with the mechanosensory function of integrins.²⁸ Rho family small GTPases have also been implicated in the shear stress response. There is evidence demonstrating that RhoA, Rac1 and Cdc42 have significant roles in the regulation of shear stress induced cytoskeletal dynamics in osteoblasts,⁴⁸ chondrocytes⁴⁹ and endothelial cells.⁵⁰ In neutrophils exposed to shear forces, a reduction in active Rac and an increase in Rho activity have been shown to drive the cell rounding response.²²

Soluble factors that promote neutrophil migration include chemokines, formylated peptides of bacterial origin, such as f MLP, and the complement fragment, C5a. These factors signal through a cognate GPCR, which activates PI3K and leads to enrichment of phophatidylinositol 3,4,5-trisphosphate (PIP3) at the leading edge. Enrichment of PIP3 at the leading edge and polarized activation of Rho GTPases are necessary for maintaining polarity in neutrophil chemotaxis.⁵¹ Rac1 is activated at the leading edge where it promotes F-actin based protrusion, while active RhoA is enriched at the sides and the back of the cell, where it drives myosin based contractile activity.⁵² The mutual exclusion of Rac1 to the front of the cell and RhoA to the back is thought to help establish and maintain self-organizing polarity during migration.⁵² RhoA dependent cell contraction is necessary for retraction of the trailing cell body in migrating neutrophils,^53,54 and monocytes.⁵⁵ Although RhoA dependent contractility is necessary for detachment of the tail, it is also known to strengthen integrin based adhesion in neutrophils,⁵⁶ lymphocytes,^57,58 and fibroblasts.⁵⁹

GEF-H1 mediated mechanotransduction

GEF-H1 is a RhoA specific MT-associated GEF, and can be activated by MT-depolymerizing agents such as nocodazole and colchicine. In epithelial cells, GEF-H1 is necessary for RhoA-dependent contractility and stress fiber formation in response to MT depolymerization.⁶⁰ Hence, GEF-H1 serves to promote F-actin and actomyosin based phenomena as a direct consequence of its release from MTs.

Work from our laboratory and others, has demonstrated that the association of GEF-H1 with the MT cytoskeleton depends on its interaction with the dynein light chain, Tctex-1.⁶¹ Phosphorylation of serine 885 in the C-terminus of GEF-H1 by PKA⁶¹ promotes binding to 14-3-3 proteins, which maintains GEF-H1 in an inactive state on the MTs. In addition to activation by MT-depolymerizing agents, signaling to GEF-H1 can be achieved through stimulation of several different kinds of cell surface receptors. GEF-H1 is activated downstream of LPA signaling in fibroblasts,⁶¹ thrombin stimulation of endothelial cells,⁶² TNF-α and epidermal growth factor (EGF) signaling in tubular epithelial cells,⁶³ Wnt signaling in neuronal cells,⁶⁴ NOD-like receptor stimulation in macrophages,⁶⁵ and as a result of mechanosensory stimulation of integrins.⁶⁶ ERK mediated phosphorylation of GEF-H1 on threonine 678 promotes its GEF activity.⁶⁷ Recent evidence indicates that GEF-H1 and another GEF called LARG (leukemia-associated Rho GEF) are both necessary for RhoA-induced mechanical stiffening in response to force on integrins in fibroblasts.⁶⁶ In this system LARG was activated downstream of the Src family tyrosine kinase Fyn, while GEF-H1 was activated by a FAK-Ras-ERK signaling axis. Moreover, TGF-β induced epithelial-to-mesenchymal transition (EMT) is mediated through enhanced proteosomal degradation of LARG and GEF-H1, which leads to stiffness attenuation and increased invasion capacity.⁶⁸ Signaling induced phosphorylation or dephosphorylation of specific serine and threonine residues, release from the MT array, localized MT depolymerization, or some combination of these are all possible mechanisms of GEF-H1 activation that could potentially occur in different contexts. Elucidation of the relative contribution of these processes may be complicated by the fact that GEF-H1 itself stabilizes MTs.⁶⁹ In the case of thrombin induced GEF-H1-dependent responses in endothelial cells, partial depolymerization of MTs has been observed.⁷⁰ However, in the case of integrin induced mechanotransduction by GEF-H1, MT depolymerization is not necessary, since the MT stabilizing agent, taxol, did not block GEF-H1 activation.⁶⁶ Differential association of GEF-H1 with specific binding partners regulates its intracellular localization and activity. In confluent epithelial cells, GEF-H1 is sequestered to the tight junctions through interaction with cingulin,⁷¹ and paracingulin.⁷² In neurons, MT-associated GEF-H1 is released in response to membrane depolarization, and binds to neurabin and spinophilin in dendritic spines, where it regulates dendritic spine morphology.⁷³ GEF-H1 binds to ASAP1 (ArfGAP with SH3 domain, ankyrin repeat, and PH domain 1) in fibroblasts, where it negatively regulates podosomes.⁷⁴ Recently, it was shown that tensional-mechanical forces induce RhoA activation through the FAK/p52(Shc) complex and the activation of p115-RhoGEF and GEF-H1 in endothelial cells.⁷⁵

In our recent work, we have provided evidence that GEF-H1 is activated in response to shear stress and promotes neutrophil spreading, crawling and transmigration.¹ Our results indicate that upon exposure to shear stress GEF-H1 becomes dephosphorylated at serine 885 (S885), and relocalizes to Flotillin-rich uropods. Previously, we demonstrated that dissociation of GEF-H1 from the dynein light chain protein, Tctex-1, and dephosphorylation at S885 was sufficient to induce GEF-H1 exchange activity toward Rho.⁷⁶ Although it has not yet been determined whether Flotillin-associated GEF-H1 is dissociated from Tctex-1, our results suggest that localization of S885-dephosphorylated GEF-H1 to the uropod is sufficient to promote neutrophil spreading and crawling, likely through stimulation of the Rho-ROCK-pMLC (phospho-myosin light chain) signaling axis. In addition to being downstream of mechanical force, GEF-H1 induced cellular contractility, through the Rho pathway⁶⁰ and produced intracellular tension,⁷⁷ which could in theory generate feed forward amplification.

In addition to it's mechanosensory role in neutrophils, GEF-H1, p115-RhoGEF⁷⁵ and LARG⁷⁸ have all been implicated in mechanosensory mechanisms in endothelial cells. Mechanotransduction through GEF-H1 has been implicated in ventilator-induced vascular endothelial permeability in the lung.⁷⁹ Interestingly, tractional forces generated by crawling leukocytes induce stiffening of underlying endothelial cells through a LARG-RhoA induced pathway, resulting in enhanced transendothelial migration.⁷⁸ This illustrates the potential of GEF-dependent mechanosensory mechanisms to influence leukocyte recruitment by regulating endothelia.

Effects of nocodazole

Nocodazole is a MT depolymerizing agent that induces contractility and morphological effects in fibroblasts.⁸⁰ In cultured fibroblasts, MT-depolymerization induces focal adhesion and actin stress fiber formation and reduced locomotion.^81,82 However, in neutrophils nocodazole induces actomyosin contractility, polarization and spontaneous migration.^83-85 We and others have found that GEF-H1 is an essential factor contributing to the cross-talk between the MT cytoskeleton and the actomyosin system.^1,60 In neutrophils, we have shown that MT depolymerization with nocodazole stimulates GEF-H1 dependent activation of Rho-induced contractility. This stimulates small membrane blebs in the short term (5 minutes), which coincides with peak phosphorylation of MLC. After 30 minutes of stimulation with nocodazole a subset of wildtype neutrophils, but not GEF-H1^−/− neutrophils, develop contractile uropods and exhibit random crawling. The most striking effects of nocodazole, which include neutrophil polarization, contractile morphological contortions and crawling, which we observed by live cell imaging, occur significantly after the peak in nocodazole induced pMLC. One possible explanation for this lag is that early contractile events are responsible for establishing neutrophil polarity, with GEF-H1-dependent membrane blebbing events promoting membrane re-configuration and enrichment of Flottilin-rich membrane rafts, eventually consolidating and forming stable uropods after 30 minutes. This explanation seams feasible when one considers that the membrane constitution of blebs is likely to be different from the parts of the membrane that don't take part in bleb formation. Furthermore, we consistently observed membrane blebbing of neutrophil differentiated HL60 cells immediately before shear stress induced spreading (unpublished result). It is of interest that only 20-30% of neutrophils develop uropods after treatment with nocodazole, which suggests the presence of neutrophil subsets in varied stages of differentiation or priming.

Similar to neutrophils, nocodazole induced contractility of fibroblasts depended on GEF-H1, and caused a reduction in spread area (unpublished results). This effect of GEF-H1 induced contractility in fibroblasts is opposite to the cell spreading and migration effect that we observed in neutrophils, although some studies have suggested that GEF-H1 promotes migration of fibroblasts through actions at the leading edge.⁸⁶ One important factor that is likely to govern whether a cell contracts or spreads upon Rho-mediated contractility is the specific subcellular localization of the contractile events. Activated neutrophils possess an intrinsic polarity and chirality,⁸⁷ and in this context restriction of GEF-H1-dependent contractility to the Flotillin-rich uropod is sufficient to drive spreading and crawling in response to shear stress. On the contrary, a non-activated/non-polarized neutrophil or fibroblast is likely to round-up/retract as a result of intracellular contractility that is not limited to a specific subcellular domain. It is likely that GEF-H1 induced contractility in the uropod produces intracellular tension that can induce cell spreading, consistent with the observation that mechanical deformation of neutrophils into narrow channels is sufficient to induce pseudopod formation.⁸⁸ Use of new FRET-based techniques, which can measure localized intracellular force with piconewton sensitivity⁸⁹ could be used to demonstrate subcellular localization of GEF-H1 induced contractility in neutrophils.

GEF-H1 translocation to the uropod

Bulk depolymerization of the MT cytoskeleton with nocodazole is a crude way of activating GEF-H1, and our results show that shear stress can accomplish GEF-H1 activation without MT-depolymerization, since the process was not inhibited by taxol. Furthermore, the kinetics of the neutrophil response to shear stress were much faster than the response to nocodazole. This suggests that mechanotransduction signaling produces neutrophil polarity and polar localization of activated GEF-H1 to the uropod more efficiently than release of GEF-H1 consequent to MT-depolymerization. This could be due to the presence of preformed membrane raft complexes and/or uropods due to f MLP or ICAM interactions in our experimental setup, or faster delivery of GEF-H1 to the uropod by unknown signaling mechanisms and chaperone factors. The rapid kinetics of shear stress induced GEF-H1 activation are consistent with signaling mediated activation of Rho downstream of receptor stimulation in leukocytes, which occurs within seconds.

GEF-H1-dependent contractility in the uropod, through the Rho-ROCK-pMLC signaling axis is sufficient to promote cell locomotion and spreading through a contractile flowing and squeezing mechanism, which is likely to be highly dependent on strong integrin attachments for anchoring. More work will be necessary to elucidate the mechanism by which localized GEF-H1 mediated contractility in the uropod induces neutrophil spreading. Neutrophils can exhibit amoeboid and mesenchymal modes of locomotion, and it is possible that GEF-H1 promotes a transition from an amoeboid (Rho-dependent) mode of migration to a more spread and flattened mesenchymal (Rac-dependent) mode in response to shear stress (Fig. 1). While the importance of integrins for shear-induced neutrophil spreading is obvious, further experiments will be necessary to determine if integrins are induced secondary to GEF-H1 effects or if load bearing integrins are the mechanosensory organelles that lie upstream of GEF-H1 activation.

Figure 1.

Crawling neutrophils can have an amoeboid or mesenchymal morphology. Neutrophils roll on the endothelial surface (a) until they encounter the right combination of pro-inflammatory signals to induce adhesion. Once adherent, they are able to crawl in a partly rounded up amoeboid (b) or a spread out mesenchymal (c) mode. In the amoeboid mode they are more exposed to shear forces, which results in GEF-H1 activation/dephosphorylation and relocalization to the uropod, where it promotes Rho-induced contractility. Contractile squeezing of the uropod to push the cytosol down and forward, in combination with strongly anchored integrins, will promote cell flattening. The impetus to promote the transition from (b) to (c) in the context of shear force is biologically coherent.

Future

While we have demonstrated a role for GEF-H1 in neutrophil spreading and crawling in response to shear stress, other roles of GEF-H1 in mechanotransduction are possible. Future work will be necessary to determine if GEF-H1 is involved in the rounding up of naïve neutrophils in response to shear. Furthermore, since neutrophils experience mechanical stress in the 3-dimensional tissue matrix, it will be of interest to determine whether GEF-H1 plays a role in migration in this context. Finally, since other leukocyte subsets, including monocytes and lymphocytes, which also cross the endothelial barrier under shear forces, also form uropods during extravasation,⁹⁰ it will be of interest to determine if GEF-H1 plays a similar role in these cell types.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.