Regulating Neutrophil PAD4/NOX-Dependent Cerebrovasular Thromboinflammation

Abstract

Background: Neutrophil extracellular trap (NET) production has been implicated in the pathogenesis of thromboinflammatory conditions such as Sickle Cell Disease (SCD), contributing to heightened risk for ischemic stroke. NETs are catalyzed by the enzyme Peptidyl Arginine Deiminase 4 (PAD4) and neutrophil derived reactive oxygen species (ROS), especially NADPH oxidase (NOX) which interacts with PAD4 and is therefore critical for neutrophil function. However, the role that NOX-dependent ROS and NETs play in the accelerated cerebral microvascular thrombosis associated with thromboinflammatory conditions, such as SCD, has not been fully elucidated and is the aim of this study. Methods: The in-vitro effects of targeting PAD4 and NOX were examined using physiologically relevant NET assays with neutrophils isolated from healthy volunteers (control) and SCD patients. In addition, in-vivo intravascular effects of targeting PAD4 and NOX in the cerebral microcirculation of C57BL/6 and sickle transgenic mice (STM) were assessed using a photoactivation thrombosis model (light/dye) coupled with real-time fluorescence intravital microscopy. Results: We found that targeting PAD4 and NOX in human neutrophils significantly inhibited ionomycin dependent H3cit⁺ neutrophils. Targeting PAD4 and NOX in-vivo resulted in prolonged blood flow cessation in cerebrovascular arterioles as well as venules. Moreover, we were able to replicate the effects of PAD4 and NOX targeting in a clinical model of accelerated thromboinflammation by increasing blood flow cessation times in cerebral microvessels in STM. These findings concurred with the clinical setting i.e. neutrophils isolated from SCD patients, which possessed an attenuation of H3cit⁺ neutrophil production on targeting PAD4 and NOX. Conclusions: Taken together, our compelling data suggests that PAD4 and NOX play a significant role in neutrophil driven thromboinflammation. Targeting PAD4 and NOX limits pathological H3cit⁺ neutrophils, which may further explain attenuation of cerebral thrombosis. Overall, this study presents a viable pre-clinical model of prevention and management of thromboinflammatory complications such as ischemic stroke.

Keywords: NADPH oxidase; Thromboinflammation; neutrophil extracellular traps; neutrophils; peptidyl arginine deiminase 4; thrombosis.

Figure 1

PAD4 and NOX inhibition causes decrease in H3Cit⁺ neutrophils. (A) Schematic of experimental design of neutrophil isolation and neutrophil extracellular trap (NET) analysis. (B) NETs were quantified by Sytox green intensity using a plate reader (BioTek; excitation = 485 nm, emission = 525 nm) from unstimulated and ionomycin (4 µM)-stimulated neutrophils isolated from human volunteers; unstimulated (n = 8), ionomycin stimulated (n = 8), GSK484 pre-treated ionomycin-stimulated neutrophils (n=5), VAS3947 pre-treated ionomycin-stimulated neutrophils (n=6), and GSK484 + VAS3947 pre-treated ionomycin-stimulated neutrophils (n=5). (C+D) In a similar manner, percentage of NETs hypercitrullinated at histone H3 (H3Cit⁺) quantified from unstimulated (n = 8), ionomycin stimulated (n = 7), GSK484 pre-treated ionomycin-stimulated neutrophils (n=5), VAS3947 pre-treated ionomycin-stimulated neutrophils (n=5), and GSK484 + VAS3947 pre-treated ionomycin-stimulated neutrophils (n=5). ^*p<0.05, ^***p<0.001, ^****p<0.0001 vs. control unstimulated neutrophils. ^####p<0.0001 vs. ionomycin-stimulated neutrophils. ^$p<0.05 vs. GSK484 pre-treated ionomycin-stimulated neutrophils. Graphs are expressed as mean±SEM from independent experiments. All imaging analysis was done in a double-blinded fashion.

Figure 2

NOX associated NET production is PAD4 independent. (A) Schematic of experimental design of neutrophil isolation and neutrophil extracellular trap (NET) analysis. (B) NETs were quantified by Sytox green intensity using a plate reader (BioTek; excitation = 485 nm, emission = 525 nm) from unstimulated and PMA (100 nM)-stimulated neutrophils isolated from human volunteers; unstimulated (n = 10), PMA stimulated (n = 9), GSK484 pre-treated PMA-stimulated neutrophils (n=10), VAS3947 pre-treated PMA-stimulated neutrophils (n=5), and GSK484 + VAS3947 pre-treated PMA-stimulated neutrophils (n=5). (C) Images of NETs with merged H3Cit (green/Alexa Fluor 488), NE and nucleus (Blue/49,6-diamidino-2-phenylindole). Bars in images represent 100 mm. (D) In a similar manner, percentage of NETs hypercitrullinated at histone H3 (H3Cit⁺) quantified from unstimulated (n = 9), PMA-stimulated (n = 9), GSK484 pre-treated PMA-stimulated neutrophils (n=9), VAS3947 pre-treated PMA-stimulated neutrophils (n=6), and GSK484 + VAS3947 pre-treated PMA-stimulated neutrophils (n=5). (E+F) DHR123 production was measured using a plate reader (BioTek; excitation = 485 nm, emission = 525 nm) from (E) unstimulated neutrophils, PMA-stimulated neutrophils, and VAS3947 treated PMA-stimulated neutrophils (n=5 in all groups), and from (F) unstimulated neutrophils, ionomycin-stimulated neutrophils, and VAS3947 treated ionomycin-stimulated neutrophils (n=5 in all groups). ^*p<0.05, ^**p<0.01, ^***p<0.001, ^****p<0.0001 vs. control unstimulated neutrophils. ^##=p<0.01 vs. PMA-stimulated neutrophils. ^$p<0.05 vs. GSK484 pre-treated PMA-stimulated neutrophils. Graphs are expressed as mean±SEM from independent experiments. All imaging analysis was done in a double-blinded fashion.

Figure 3

PAD4 and NOX targeting attenuate in-vivo cerebrovascular thrombosis. (A) Schematic representation of light/dye-induced thrombosis model induced with intravenous infusion of 10 mg/kg of 5% FITC-dextran followed by photoactivation of the cerebral microvessels of C57BL/6 mice. Time to flow cessation (minutes) was defined as the complete stop of blood flow for ≥ 30 seconds and was assessed in (B) arterioles and (C) venules in C57BL/6 mice treated with vehicle (saline. n=6 mice per group), GSK484 (10 μM. n=6 mice per group), VAS3847 (5 μM/ n=6 mice per group) or a combination of GSK484 + VAS3947 (n=5 mice per group). (D-E) Images of onset (start of platelet aggregation) and cessation (complete stop of flow for ≥30 seconds) in arterioles and venules respectively. ^**p<0.01, ^***p<0.001, ^****p<0.0001 vs. vehicle treated mice. Graphs are expressed as mean±SEM from 5-6 mice per group.

Figure 4

Clinical thromboinflammation can be attenuated by targeting PAD4 and ROS signalling in human neutrophils. (A) Schematic of experimental design of neutrophil isolation and neutrophil extracellular trap (NET) analysis. (B+E) Images of NETs with merged H3Cit (green/Alexa Fluor 488), NE and nucleus (Blue/49,6-diamidino-2-phenylindole). Bars in images represent 100 mm. (C) NETs were quantified by Sytox green intensity from unstimulated and ionomycin (4 µM)-stimulated neutrophils isolated from sickle cell disease (SCD) patients; unstimulated (n = 10), ionomycin stimulated (n = 10), GSK484 pre-treated ionomycin-stimulated SCD neutrophils (n=8), VAS3947 pre-treated ionomycin-stimulated SCD neutrophils (n=5), and GSK484 + VAS3947 pre-treated ionomycin-stimulated SCD neutrophils (n=5). (D) In a similar manner, percentage of H3Cit⁺ neutrophils were quantified from unstimulated (n = 8), ionomycin-stimulated (n = 8), GSK484 pre-treated ionomycin-stimulated SCD neutrophils (n=8), VAS3947 pre-treated ionomycin-stimulated SCD neutrophils (n=6), and GSK484 + VAS3947 pre-treated ionomycin-stimulated SCD neutrophils (n=6). (F) Further NETs were also quantified by Sytox green intensity from unstimulated and PMA (100 nM)-stimulated SCD neutrophils isolated from SCD patients; unstimulated (n = 5), PMA-stimulated (n = 5), GSK484 pre-treated ionomycin-stimulated SCD neutrophils (n=5), VAS3947 pre-treated ionomycin-stimulated SCD neutrophils (n = 5), and GSK484 + VAS3947 pre-treated ionomycin-stimulated SCD neutrophils (n = 5). (G) Percentage of H3Cit⁺ neutrophils were quantified from unstimulated (n = 8), PMA stimulated (n = 8), GSK484 pre-treated PMA-stimulated SCD neutrophils (n = 6), VAS3947 pre-treated PMA-stimulated SCD neutrophils (n = 5), and GSK484 + VAS3947 pre-treated PMA-stimulated SCD neutrophils (n = 5). ^*p<0.05 vs. control unstimulated SCD neutrophils. ^#p<0.05, ^##p<0.01, ^###p<0.001 vs. ionomycin-stimulated SCD neutrophils. ^$p<0.05, ^$$p<0.01, ^$$$$p<0.001 PMA stimulated SCD neutrophils. Graphs are expressed as mean±SEM from independent experiments. All imaging analysis was done in a double-blinded fashion.

Figure 5

Clinical thromboinflammation can be attenuated by targeting PAD4 and NOX in-vivo. (A) Schematic representation of light/dye-induced thrombosis model induced with intravenous infusion of 10 mg/kg of 5% FITC-dextran followed by photoactivation of the cerebral microvessels of sickle cell transgenic mice (STM). Time to flow cessation (minutes) was defined as the complete stop of blood flow for ≥ 30 seconds and was assessed in (B) arterioles and (C) venules in STM treated with vehicle (saline. n=6 mice per group), GSK484 (10 μM, n=5 mice per group), VAS3847 (5 μM, n=5 mice per group) or a combination of GSK484 + VAS3947 (n = 5 mice per group). (D-E) Images of onset (start of platelet aggregation) and cessation (complete stop of flow for ≥30 seconds) in arterioles and venules respectively. ^*p<0.05, ^**p<0.01, ^***p<0.001, ^****p<0.0001 vs. vehicle treated mice. Graphs are expressed as mean±SEM from 5-6 mice per group.