Clinical hypothermia temperatures increase complement activation and cell destruction via the classical pathway

Abstract

Background: Therapeutic hypothermia is a treatment modality that is increasingly used to improve clinical neurological outcomes for ischemia-reperfusion injury-mediated diseases. Antibody-initiated classical complement pathway activation has been shown to contribute to ischemia-reperfusion injury in multiple disease processes. However, how therapeutic hypothermia affects complement activation is unknown. Our goal was to measure the independent effect of temperature on complement activation, and more specifically, examine the relationship between clinical hypothermia temperatures (31-33°C), and complement activation.

Methods: Antibody-sensitized erythrocytes were used to assay complement activation at temperatures ranging from 0-41°C. Individual complement pathway components were assayed by ELISA, Western blot, and quantitative dot blot. Peptide Inhibitor of complement C1 (PIC1) was used to specifically inhibit activation of C1.

Results: Antibody-initiated complement activation resulting in eukaryotic cell lysis was increased by 2-fold at 31°C compared with 37°C. Antibody-initiated complement activation in human serum increased as temperature decreased from 37°C until dramatically decreasing at 13°C. Quantitation of individual complement components showed significantly increased activation of C4, C3, and C5 at clinical hypothermia temperatures. In contrast, C1s activation by heat-aggregated IgG decreased at therapeutic hypothermia temperatures consistent with decreased enzymatic activity at lower temperatures. However, C1q binding to antibody-coated erythrocytes increased at lower temperatures, suggesting that increased classical complement pathway activation is mediated by increased C1 binding at therapeutic hypothermia temperatures. PIC1 inhibited hypothermia-enhanced complement-mediated cell lysis at 31°C by up to 60% (P = 0.001) in a dose dependent manner.

Conclusions: In summary, therapeutic hypothermia temperatures increased antibody-initiated complement activation and eukaryotic cell destruction suggesting that the benefits of therapeutic hypothermia may be mediated via other mechanisms. Antibody-initiated complement activation has been shown to contribute to ischemia-reperfusion injury in several animal models, suggesting that for diseases with this mechanism hypothermia-enhanced complement activation may partially attenuate the benefits of therapeutic hypothermia.

Figure 1

Functional hemolytic assays of complement activation. (A) CH50-type antibody-initiated complement assays incubating antibody-sensitized erythrocytes (EA) with 0.2% human serum for 1 hour at 31-41°C showed that hemolysis of antibody-sensitized erythrocytes in human serum differed significantly as a function of temperature (ANOVA P < 0.01). There is 2-fold more hemolysis at 31°C vs. 37°C (*t-test P = 0.01) (P ≤ 0.05 for the comparisons 31 vs. 37, 31 vs. 39, 31 vs. 41, 33 vs. 37, 33 vs. 39, 33 vs. 41, 35 vs. 39, 35 vs. 41, with significantly more hemolysis at lower temperatures) Data are mean ± SEM for 4 independent experiments. (B) Evaluating antibody-initiated complement activation from 0-37°C showed that hemolysis of antibody-sensitized erythrocytes in human serum differed significantly as a function of temperature (ANOVA P < 0.01). There was 2.5-fold more hemolysis at 31°C vs. 37°C (*t-test P = 0.01). (P ≤ 0.05 for the comparisons 31 vs. 37, 24 vs. 37, 19 vs. 37, with significantly more hemolysis at lower temperatures). The trend of increased complement activation was reversed at 13°C. (No statistical difference between hemolysis at 13 vs. 37 or 7 vs. 37). Data are mean ± SEM for 4 independent experiments. (C) AP50-type alternative complement pathway assays, incubating rabbit erythrocytes with 4% human serum in Mg-EGTA-GVBS for 30 minutes at 31-41°C showed that temperature did not influence complement-mediated cell lysis over therapeutic hypothermia temperatures (ANOVA P = 0.45). Data are mean ± SEM for 4 independent experiments. (D) Evaluating alternative pathway activation from 0-37°C showed that alternative pathway activation was significantly inhibited at 24°C and below. There was no difference in the degree of hemolysis at 31°C vs. 37°C) Data are mean ± SEM for 4 independent experiments.

Figure 2

Complement cascade activation assays. (A) Heat-aggregated IgG 1% normal human serum for 1 hour at 31-41°C demonstrated that C5a generation differed significantly as a function of temperature (ANOVA P < 0.01). There was a 2.5 fold increase in C5a generation at 31°C vs. 37°C (*t-test P =0.05). Data are mean ± SEM for 4 independent experiments. C3/C4-analysis samples were generated by incubating 1% C8-deficient serum (to prevent hemolysis) with 0.25 mL EA for one hour at 31-41°C. (B) Total C3-fragment opsonization did not differ significantly as a function of temperature (ANOVA P = 0.5). (C) Cell membrane bound iC3b differed significantly as a function of temperature (ANOVA P < 0.01), with a 2-fold increase at 31°C vs. 37°C (t-test P = 0.04). Data are mean ± SEM for 4 independent experiments. (D) A representative Western blot analysis of membrane-bound C3-fragments confirmed that iC3b was the predominant form. (E) C4-fragment opsonization differed significantly as a function of temperature (ANOVA P = 0.04). There was an almost 2-fold increase in C4 generation at 31°C vs. 37°C (*t-test P = 0.05). Data are mean ± SEM for 4 independent experiments. (F) A representative Western blot analysis demonstrated a mixture of C4b, iC4b and C4d fragments.

Figure 3

C1s activation by heat-aggregated IgG for increasing temperatures. One μg of partially purified human C1 was incubated with heat-aggregated IgG at 31-41 C for 0–90 minutes. C1s activation, as measured by generation of C1s heavy and light chains by Western blot methodology, was quantified by Odyssey imaging and demonstrated decreased activation of C1s at therapeutic hypothermia temperatures (ANOVA P < 0.01), with a 5-fold decrease at 31°C vs. 37°C (*t-test P < 0.01). Data are mean ± SEM for 3 independent experiments.

Figure 4

C1/C1q binding at therapeutic hypothermia temperatures. C1q binding was tested by incubating antibody-sensitized erythrocytes with (A) purified C1q, (B) normal human serum, or (C) C8-deficient serum for one hour at 31-41°C. In each case, there was a consistent trend of increased C1q binding to EA at lower temperatures. Data are mean ± SEM for 4 independent experiments.

Figure 5

Peptide inhibitor of C1 (PIC1) and hypothermia-enhanced complement-mediated hemolysis. Human serum (16.5%) was pre-incubated with 0.98 mM, 0.82 mM, 0.65 mM, 0.48 mM or 0.32 mM PIC1, DMSO vehicle control, or buffer control before performing the CH50-type assay at 31° or 37°C. PIC1 inhibited hypothermia-enhanced complement activation and hemolysis at 31°C for all doses tested up to 60% inhibition for 0.98 mM PIC1 compared with 31°C untreated (*ANOVA P <0.01). Increasing concentrations of PIC1 showed a significant dose dependent inhibition of hypothermia-enhanced complement activation at 31°C for all doses tested. At 31°C, 0.98 mM of PIC1 inhibited complement mediated cell lysis to a level equivalent with euthermia (37°C) (t-test P = 0.44) Data are mean ± SEM for 3 independent experiments.

Figure 6

Model of antibody-initiated complement activation in ischemia reperfusion injury and hypothermia effects on complement activation. Hypoxic insult induces expression of 'neoantigens' on the surface of vascular endothelial cells. These neoantigens are recognized by natural antibodies (IgM) initiating complement activation leading to downstream inflammatory effectors. Therapeutic hypothermia temperatures were shown to increase C1/C1q binding, increase opsonization with C4-fragments and C3-fragments, increase C5a anaphylatoxin generation, and increase eukaryotic cell lysis via membrane attack complex (MAC) formation. Increases in complement function demonstrated in this study are shown in grey. PIC1 inhibits complement activation at C1 preventing C4 activation.