Fig. 3.
Increased susceptibility of CD59a-deficient mouse erythrocytes to induced complement lysis.
(A) Antibody-induced complement lysis of wild-type (filled circle; n = 3 mice) and CD59a knockout (open circle; n = 3) mouse erythrocytes by MAC assembled from human C5b-7 and mouse C8/C9. Antibody-sensitized cells were sequentially exposed to C8-depleted human serum (1:20) and mouse serum at the indicated dilutions.P < .05 for all serum dilutions by Student ttest. (B) Antibody-induced lysis by rat complement (1:20) of wild-type (WT), DAF knockout (DAF), CD59a knockout (CD59), and Crry/C3 knockout (Crry) mouse erythrocytes (n = 4 mice for each group). DAF and CD59a knockout cells were significantly more sensitive than wild-type cells (P < .05 andP < .001, respectively, by Student t test). Crry/C3 knockout erythrocytes were not more sensitive than wild-type cells. (C) Antibody-induced lysis by human complement (1:40) of wild-type and various knockout mouse erythrocytes (n = 4 mice for each group). All knockout erythrocytes were more sensitive than wild-type cells to human complement lysis (P < .001 for DAF and CD59a knockout;P < .01 for Crry/C3 knockout; Student t test). (D) Antibody-induced mouse C3 deposition on wild-type and various knockout mouse erythrocytes (n = 8 mice for each group). CD59a knockout mouse erythrocytes had levels of C3 deposition similar to those of wild-type mouse erythrocytes (P = .45, Student t test). Both DAF knockout and Crry/C3 knockout mouse erythrocytes had more C3 deposition than the wild-type cells (P < .01 for DAF knockout and P < .05 for Crry/C3 knockout mice, Student t test). C3 deposition on DAF knockout and Crry/C3 knockout mouse erythrocytes was not significantly different (P = .26, Student t test). (E) (F) CD59a-deficient mouse erythrocytes were more susceptible to CVF-induced autologous complement lysis. Injection of CVF (10 μg per mouse) caused systemic complement activation (data not shown) and increased plasma hemoglobin levels (measured by OD414) in knockout mice (panel F; n = 8 mice; P < .01, Student t test), but not in wild-type mice (panel E; n = 8 mice; P = .9, Student t test). Plasma samples were taken immediately before and 1 hour after CVF injections. Data presented are representative of at least 3 independent experiments.