Microparticles mediate hepatic ischemia-reperfusion injury and are the targets of Diannexin (ASP8597) - PubMed (original) (raw)

Microparticles mediate hepatic ischemia-reperfusion injury and are the targets of Diannexin (ASP8597)

Narci C Teoh et al. PLoS One. 2014.

Abstract

Background & aims: Ischemia-reperfusion injury (IRI) can cause hepatic failure after liver surgery or transplantation. IRI causes oxidative stress, which injures sinusoidal endothelial cells (SECs), leading to recruitment and activation of Kupffer cells, platelets and microcirculatory impairment. We investigated whether injured SECs and other cell types release microparticles during post-ischemic reperfusion, and whether such microparticles have pro-inflammatory, platelet-activating and pro-injurious effects that could contribute to IRI pathogenesis.

Methods: C57BL6 mice underwent 60 min of partial hepatic ischemia followed by 15 min-24 hrs of reperfusion. We collected blood and liver samples, isolated circulating microparticles, and determined protein and lipid content. To establish mechanism for microparticle production, we subjected murine primary hepatocytes to hypoxia-reoxygenation. Because microparticles express everted phosphatidylserine residues that are the target of annexin V, we analyzed the effects of an annexin V-homodimer (Diannexin or ASP8597) on post-ischemia microparticle production and function.

Results: Microparticles were detected in the circulation 15-30 min after post-ischemic reperfusion, and contained markers of SECs, platelets, natural killer T cells, and CD8+ cells; 4 hrs later, they contained markers of macrophages. Microparticles contained F2-isoprostanes, indicating oxidative damage to membrane lipids. Injection of mice with TNF-α increased microparticle formation, whereas Diannexin substantially reduced microparticle release and prevented IRI. Hypoxia-re-oxygenation generated microparticles from primary hepatocytes by processes that involved oxidative stress. Exposing cultured hepatocytes to preparations of microparticles isolated from the circulation during IRI caused injury involving mitochondrial membrane permeability transition. Microparticles also activated platelets and induced neutrophil migration in vitro. The inflammatory properties of microparticles involved activation of NF-κB and JNK, increased expression of E-selectin, P-selectin, ICAM-1 and VCAM-1. All these processes were blocked by coating microparticles with Diannexin.

Conclusions: Following hepatic IRI, microparticles circulate and can be taken up by hepatocytes, where they activate signaling pathways that mediate inflammation and hepatocyte injury. Diannexin prevents microparticle formation and subsequent inflammation.

PubMed Disclaimer

Conflict of interest statement

Competing Interests: Alavita Inc. gifted the Diannexin for this study; a therapeutic protein also known as Astellas compound, ASP8597, and developed by Alavita. This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials.

Figures

Figure 1

Figure 1. MP production is increased in hepatic IRI.

A. MP release (concentration expressed as phosphatidylserine, PS equivalent in nM) and Bi,ii serum ALT and hyaluronic acid (HA) after 60 min ischemia and indicated reperfusion times in naïve and Diannexin-treated mice (n = 10 per cohort). * p<0.05 all experimental groups vs. sham. ‡ p<0.05 IRI 2 hr vs. 30 min reperfusion. # p<0.05 IRI 24 hr vs. 15 min, 30 min and 2 hr reperfusion. § p<0.05 Diannexin vs. no Diannexin at 24 hr reperfusion. C. FACS plots demonstrating MPs released early in reperfusion are composed predominantly of endothelial-derived/CD144 remnants while in late reperfusion, MPs are largely from leukocytes and hepatocytes. (Ci) FACS data depicted as % bearing cell-specific markers (Table S1). Note, because individual MPs are cell fragments and may potentially fuse, MPs can bear more than 1 cell-marker and therefore, the sum of all expressed markers may feasibly exceed 100% (Materials and Methods S1). (Cii) Annexin V was utilized to analyze for double positive events together with cell-specific markers CD144, CD41, CD62P F4/80 and Ly6G. * p<0.05 IRI groups vs. sham. D. MPs bear ASPGR (hepatocytes) and VE-cadherin (SECs), VCAM-1, ICAM-1, E-selectin, P-selectin. MPs obtained from mice subjected to liver post-IR at indicated times. Blots representative of three similar experiments. E. Image analyses/quantification show upregulation of adhesion molecules especially at 2, 24 hr reperfusion. * p<0.05 IRI vs. sham. & p<0.05 IRI 2 and 24 hr vs. IRI 15 and 30 min reperfusion. %p<0.05 IRI 30 min vs. IRI 15 min reperfusion. # p<0.05 IRI 24 hr vs. 15,30 min, 2 hr reperfusion. § p<0.05 IRI 2 hr vs. 15 min reperfusion. ‡ p<0.05 IRI 2 hr vs. IRI 30 min reperfusion. F. MPs at 15 min and 2 hr reperfusion contain increased levels of F2-isoprostanes compared to those derived from sham-operated mice by GCMS and LC/MS/MS (normalised to total AA detected).

Figure 2

Figure 2. MPs are directly pathogenic to primary hepatocytes, incite neutrophil recruitment and activate platelets.

Ai. Hepatic neutrophils were derived by liver perfusion and chemotaxis measured in ThinCert chambers. Lower compartments were seeded with MPs from mice subjected to 2 hr post-ischemia-reperfusion (IR), IL-8 or TNF-α as chemoattractants (IL-8,TNF:positive controls) and incubated (1 hr, 37°C). Co-treatment of MPs with 1 mM NAC reduced neutrophil transmigration. To ascertain effect of _N_-acetylcysteine (NAC) on transmigration potential, MPs were pre-treated with 1 mM NAC for 1 hr 37°C; thereafter, MPs were resuspended in PBS then recovered by prolonged centrifugation (see Materials and Methods S1 for detailed MP isolation protocol) prior to the transwell migration assay. Aii. Addition of MPs generated by IRI from specific cell types (significantly, activated platelets and endothelial cells, CD62P and macrophages, F4/80; not resting SECs, CD144) to these chambers enhanced transmigration of leukocytes. * p<0.05 experimental groups vs. control. B. Platelets from sham-operated mice treated with 60 nM MPs (from mice subjected to 2 hr post-IR) for 30 min, labelled with CD62P (activated platelet and endothelial cell marker) and subjected to FACS. * p<0.05 experimental groups vs. control. C. MPs from mice subjected to 2 hr post-IR co-incubated with primary hepatocytes. By 30 min, MPs adhered to hepatocytes and were endocytosed at 1 hr (first row/top panel of figures: 0 min, 30 min, 60 min); engulfment was inhibited by 0.45 M sucrose (second row/lower panel of figures: 30 min, 60 min). D. Primary hepatocytes treated with increasing concentrations of MPs derived from mice subjected to 2 hr post-IR. LDH leakage and cell viability (MTT assay) in MP-treated cells compared with PBS-controls. (E) Primary hepatocytes treated with increasing concentrations of MPs derived from specific cell types (CD62P: activated platelets and endothelial cells, CD144: SECs, F4/80: macrophages) obtained from mice subjected to 2 hr post-IR. Experiments performed in triplicate in MP-preparations from 4 mice (n = 12). *p<0.05 experimental groups vs. control.

Figure 3

Figure 3. MPs invoke mitochondrial permeability transition, oxidative stress, adhesion molecule expression and activate NF-κB in hepatocytes.

A. Hepatocytes loaded with 50 nM TMRM (red fluorescence), 1 µM DCFH2-DA (green fluorescence) and placed in chambers at 37°C for 15 min. MPT occurred with increasing MP concentrations, blocked by 1 µM CyA. B. 1 mM of NAC inhibited oxidative stress generated by MPs. C, D. Primary hepatocytes were exposed to 60 nM of MPs derived from liver 2 hr post-IR and pro-inflammatory molecules COX-2 and PKC-δ determined (WB) as well as (E) JNK-1/2 (46 kDA, 54 kDA) activation (phospho-JNK). * p<0.05 experimental groups vs. control. + p<0.05 30 and 60 nM vs. 12 nM MPs. † p<0.05 MPs 60 nM vs. 30 nM MPs. F. MPs activate NF-κB in primary hepatocytes, by IκB-α degradation (immunoblots). Blots represent three experiments conducted in triplicate (n = 9). * p<0.05 experimental vs. control.

Figure 4

Figure 4. Hypoxia-reoxygenation and oxidative stress trigger MP release.

A. Primary hepatocytes were exposed to 4 hr hypoxia followed by 24 hr reoxygenation (Methods). Pretreatment with 1 µM CyA (MPT-inhibitor) for 2 hr prior to HR attenuated MP release and (B) LDH leakage from hepatocytes subjected to HR. Controls: cells incubated in normoxic conditions. Experiments performed in triplicate. * p<0.05 experimental vs. control. & p<0.05 hypoxia-reoxygenation vs. CyA. C. H2O2 causes MP primary hepatocytes to form MPs, which is attenuated by calcium chelation (EGTA). Primary murine hepatocytes were incubated with H2O2 for 24 hr. At 10 µM or higher, H2O2 causes generation of MPs, while MP release is abolished by co-administration of EGTA. Assays performed in triplicate. * p<0.05 experimental vs. control. +p<0.05 EGTA vs. H2O2.

Figure 5

Figure 5. MPs stimulate TNF release and are themselves potent agonists of MP vesiculation in vivo.

A. Serum hyaluronic acid (HA) release increases in mice subjected to 60 min ischemia and 15 min reperfusion; this is exacerbated by administration of TNF in vivo (mice injected with TNF 5 µg/kg iv vs. saline controls), 5 min prior to 60 min ischemia. * p<0.05 experimental vs. control. B. TNF is produced by primary hepatocytes following addition of MPs derived from mice subjected to 2 hr post-ischemic reperfusion. * p<0.05 experimental vs. control. † p<0.05 60 nM vs. 30 nM MPs. C. Primary hepatocytes were incubated with 2 nM TNF for 1 hr and 24 hr in normoxic conditions. SP600125 (20 µM in 0.1% DMSO) significantly attenuated TNF-stimulated MP release. * p<0.05 experimental vs. control. ‡ p<0.05 TNF vs. TNF+SP600125. D. WT CL57BL6 mice were injected i.v. with TNF 5 µg/kg (saline for controls), 5 min prior to 60 min ischemia. MP production was measured in plasma at 15 min, 30 min, 2 hr reperfusion. Liver IRI triggered MP production; TNF injection further augmented MP release reaching a maximum at 2 hr. * p<0.05 experimental vs. control. ± p<0.05 IRI groups vs. sham TNF. # p<0.05 30 min+TNF vs. IRI 30 min reperfusion. § p<0.05 2 hr+TNF vs. 2 hr reperfusion.

Figure 6

Figure 6. Diannexin attenuates release of MPs after hepatic IRI and decreases their pro-inflammatory and deleterious potential.

MPs were obtained from mice pre-treated with 1 mg/kg iv Diannexin 5 min prior to 60 min liver IR and results compared with vehicle-treated/naïve mice subject to IR. A. FACS revealed significantly reduced resting endothelial cell (CD144), resting platelet (CD41) and activated platelet and endothelial cell (CD62P) subsets compared to naïve (see Fig.1C). Bi. Serum ALT after 60 min ischemia and indicated reperfusion times in naïve and Diannexin-treated mice (n = 10 per cohort). * p<0.05 all experimental groups vs. sham. § p<0.05 Diannexin vs. 24 hr reperfusion. Bii. Platelets were isolated from sham-operated mice and exposed to 60 nM MPs (derived from mice subjected to 60 min ischemia, 2 hr reperfusion) in the absence or presence of 4 µg/ml Diannexin. FACS then performed using CD62P to quantify activated platelet subset as a proportion of total platelets. * p<0.05 experimental vs. control. # p<0.05 Diannexin vs. 60 nM MPs. C, D. Hepatocyte-specific ASPGR, E-selectin and pro-inflammatory VCAM, ICAM-1 protein expression diminished by Diannexin compared to naïve mice subjected to 60 min ischemia and indicated reperfusion times. + p<0.05 Diannexin vs. sham. & p<0.05 Diannexin vs. naïve. E. Pre-treatment of MPs with Diannexin 4 µg/ml inhibited neutrophil transmigration in ThinCert chambers. To ascertain effect of test compounds on transmigration potential, MPs were pre-treated with Diannexin (4 µg/ml) for 1 hr 37°C; thereafter, MPs were resuspended in PBS then recovered by prolonged centrifugation twice (see Materials and Methods S1 for detailed MP isolation protocol) prior to the transwell migration assay. * p<0.05 experimental vs. control. # p<0.05 Diannexin vs. 60 nM MPs.

References

    1. Jaeschke H, Woolbright BL (2012) Current strategies to minimize hepatic ischemia-reperfusion injury by targeting reactive oxygen species. Transplant Rev 26: 103–114. - PMC - PubMed
    1. Teoh N, Farrell G (2003) Hepatic ischaemia reperfusion injury. Pathogenic mechanisms and basis for protection. J Gastroenterol Hepatol 18: 891–902. - PubMed
    1. Jaeschke H, Farhood A (2002) Kupffer cell activation after no-flow ischemia versus haemorrhagic shock. Free Rad Bio Med 33: 210–219. - PubMed
    1. Samarasinghe D, Farrell GC (2000) Role of oxidative stress in hypoxia-reoxygenation injury to cultured rat hepatic sinusoidal endothelial cells. Hepatology 31: 160–165. - PubMed
    1. Vollmar B, Menger MD (2009) The hepatic microcirculation: mechanistic contributions and therapeutic targets in liver injury and repair. Physiol Rev 89: 1269–1339. - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources