Acute antibody-mediated complement activation mediates lysis of pancreatic islets cells and may cause tissue loss in clinical islet transplantation

Abstract

Background: Clinical islet transplantation is associated with loss of transplanted islets necessitating tissue from more than one donor to obtain insulin independence. The instant blood-mediated inflammatory reaction (IBMIR) is one explanation to the tissue loss. Complement activation is an important cytotoxic component of the IBMIR, and in the present study, we have investigated this component in detail.

Methods: Isolated human islets were analyzed by large particle flow cytometry and confocal microscopy after incubation in human ABO-compatible hirudin-plasma.

Results: After incubation in plasma, the islets bound IgG and IgM, CIq, C4, C3 and C9. The binding of C3b/iC3b was evident already after 5 min. The binding of C3b/iC3b and the generation of C3a and sC5b-9 were inhibited by the complement inhibitor Compstatin. Lysis as reflected by propidium iodide (PI) staining and release of C-peptide was also inhibited by Compstatin. There were significant correlations between IgM/IgG versus C3b/iC3b and between sC5b-9 and C-peptide.

Conclusion: The conclusion is that complement is activated by natural IgG and IgM antibodies already after 5 min. The complement activation leads to lysis of cells of the pancreatic islets. This very rapid reaction may be an essential entity of the damage induced by the IBMIR in clinical islet transplantation.

FIGURE 1

Human islets (n=5) incubated in ABO-compatible hirudin-treated plasma (n=5) for 30 min. The islets were stained for IgG, IgM, MBL, C1q, C4, C3b/iC3b, C9, and negative control. Controls consisting of islets incubated without (islets alone) and with mouse isotype IgG1 (negative control) were included. The islets were analyzed by (A) COPAS and (B) confocal microscopy. All values were compared with values obtained from the negative control are given as mean±SEM (*P<0.05, **P<0.01).

FIGURE 2

Human islets (n=5) incubated in ABO-compatible hirudin-treated plasma (n=5) for 30 min in the presence (open symbols) or absence (filled symbols) of 10 μM Compstatin. (A) The binding of C3b/iC3b as a percentage of the maximum value; values for the generation of (B) C3a and (C) sC5b-9 are given as mean±SEM (*P<0.05, **P<0.01, ***P<0.001).

FIGURE 3

Confocal microscopic images of islets incubated in ABO-compatible hirudin-treated plasma human in the presence or absence of 10 μM Compstatin at 5 and 30 min. The islets were stained for C3b/iC3b. One hundred islets from five different donors/recipient combinations were investigated.

FIGURE 4

Human islets incubated in ABO-compatible hirudin-treated ABO-compatible plasma for 30 min in the presence (dark grey) or absence (light grey) of 10 μM Compstatin. The islets were stained with PI (A) and analyzed by COPAS, and the presence of C-peptide in the supernatants was determined by enzyme-linked immunosorbent assay (B) (n=7; mean±SEM; *P<0.05, **P<0.01). In (C) islets incubated in the presence (squares) and absence (diamonds) of ABO-compatible plasma for 30 min were challenged by glucose and the release of insulin assessed (n=6; ns=non significant).

FIGURE 5

Correlation between the binding of (A) IgM and C3b/iC3b (ρ=0.67; P=0.01); (B) IgG and C-peptide (ρ=0.83; P=0.01); and (C) sC5b-9 and C-peptide (ρ=0.62; P=0.04) to the islets.