Endothelial targeting with C1-inhibitor reduces complement activation in vitro and during ex vivo reperfusion of pig liver

Abstract

Tissue damage during cold storage and reperfusion remains a major obstacle to wider use of transplantation. Vascular endothelial cells and complement activation are thought to be involved in the inflammatory reactions following reperfusion, so endothelial targeting of complement inhibitors is of great interest. Using an in vitro model of human umbilical vein endothelial cells (HUVEC) cold storage and an animal model of ex vivo liver reperfusion after cold ischaemia, we assessed the effect of C1-INH on cell functions and liver damage. We found that in vitro C1-INH bound to HUVEC in a manner depending on the duration of cold storage. Cell-bound C1-INH was functionally active since retained the ability to inhibit exogenous C1s. To assess the ability of cell-bound C1-INH to prevent complement activation during organ reperfusion, we added C1-INH to the preservation solution in an animal model of extracorporeal liver reperfusion. Ex vivo liver reperfusion after 8 h of cold ischaemia resulted in plasma C3 activation and reduction of total serum haemolytic activity, and at tissue level deposition of C3 associated with variable level of inflammatory cell infiltration and tissue damage. These findings were reduced when livers were stored in preservation solution containing C1-INH. Immunohistochemical analysis of C1-INH-treated livers showed immunoreactivity localized on the sinusoidal pole of the liver trabeculae, linked to sinusoidal endothelium, so it is likely that the protective effect was due to C1-INH retained by the livers. These results suggest that adding C1-INH to the preservation solution may be useful to reduce complement activation and tissue injury during the reperfusion of an ischaemic liver.

Fig. 1

Binding of C1-INH to endothelial cells in vitro. (a) shows that the binding of C1-INH (120 μg/ml) was maximum after 48 h of cold storage. (Shown are means ± s.d. of three experiments in triplicate.) (b) the immunostaining for C1-INH in HUVEC treated with UW preservation solution or peroxidase conjugated C1-INH. HUVEC receiving C1-INH showed a diffuse pattern of staining. (Sections were counterstained with eosin.)

Fig. 2

Inhibition of C1s activity by C1-INH bound to HUVEC. Shown are the changes in amidolytic activity of exogenous C1s induced by HUVEC or HUVEC treated with C1-INH (120 μg/ml). Means ± s.d. of three experiments run in triplicate for each group. AU = arbitrary unit.

Fig. 3

Immunohistochemical detection of C1-INH and C3 in liver tissue.Before reperfusion, sections of UW-treated livers showed no staining using rabbit anti-C1-INH (a), whereas definite immunoreactivity was evident at the sinusoidal pole of livers treated with UW plus C1-INH (b). After 2 h of reperfusion, C3 deposition was present at the sinusoidal borders of the trabeculae in UW-treated livers (c), while it was not detected in C1 INH-treated livers (d). (× 1000).

Fig. 4

Inhibition of complement activation during liver reperfusion by cell-bound C1-INH. At specific time points CH100 and C3 activation were measured in samples from the portal vein (in) and hepatic vein (out) of pig livers cold stored in the presence of UW (U25CF) or of UW plus C1-INH (○). The percentage of cleaved C3 is reported at the bottom of each lane of the immunoblotting membrane.

Fig. 5

Histological pictures after 2 h reperfusion of an untreated pig liver showing diffuse necro-inflammatory activity (a) compared with C1INH-treated liver featuring only mild increase in sinusoidal neutrophils (b). (Haematoxilin-eosin, 400 ×.) (c) The injury scores in basal and after 2 h reperfusion biopsies, in the group of untreated (UW) and C1INH-treated animals (UW + C1INH). Reperfusion injury was scored from 0 to 4 as reported in Material and Methods. AU = arbitrary unit.