CD59 blocks not only the insertion of C9 into MAC but inhibits ion channel formation by homologous C5b-8 as well as C5b-9

Abstract

Activation of the complement system on the cell surface results in the insertion of pore forming membrane attack complexes (MAC, C5b-9). In order to protect themselves from the complement attack, the cells express several regulatory molecules, including the terminal complex regulator CD59 that inhibits assembly of the large MACs by inhibiting the insertion of additional C9 molecules into the C5b-9 complex. Using the whole cell patch clamp method, we were able to measure accumulation of homologous MACs in the membrane of CD59(-) human B-cells, which formed non-selective ion channels with a total conductance of 360 +/- 24 pS as measured at the beginning of the steady-state phase of the inward currents. C5b-8 and small-size MAC (MAC containing only a single C9) can also form ion channels. Nevertheless, in CD59(+) human B-cells in spite of small-size MAC formation, an ion current could not be detected. In addition, restoring CD59 to the membrane of the CD59(-) cells inhibited the serum-evoked inward current. The ion channels formed by the small-size MAC were therefore sealed, indicating that CD59 directly interfered with the pore formation of C5b-8 as well as that of small-size C5b-9. These results offer an explanation as to why CD59-expressing cells are not leaky in spite of a buildup of homologous C5b-8 and small-size MAC. Our experiments also confirmed that ion channel inhibition by CD59 is subject to homologous restriction and that CD59 cannot block the conductivity of MAC when generated by xenogenic (rabbit) serum.

Figure 1. Flow cytometry of NCU-2 and NCU-1 cells after incubation with anti-CD59 (dashed line), anti-MCP (dashed-dotted line) and anti-DAF (dotted line) antibodies

FACS analysis revealed the presence of complement regulatory molecules in the membrane of NCU-2 cells (A) whereas NCU-1 cells expressed DAF and MCP but not CD59 (B). The continuous line represents the FITC control.

Figure 2. Flow cytometry of NCU-2 (A) and NCU-1 (B) cells after human serum treatment and incubation with anti-MAC antibody

Both cell types harbour MAC in their membranes (dotted line). The fluorescence signal of cells incubated with anti-MAC antibody without serum treatment (dashed line) does not differ significantly from the signal of the FITC control (continuous line).

Figure 3. Whole cell clamp electrophysiology of NCU-1 and NCU-2 cells during homologous (human, A-J) or heterologous (rabbit, K-L) serum treatment

Inward current was evoked in NCU-1 cells as a result of extracellular serum administration (A). However, similar serum application could not trigger current pulses in NCU-2 cells expressing CD59 (B). A background current was observed only when heat-inactivated serum was applied to the NCU-1 cells (C), showing that the inward current evoked by normal human serum was due to complement activation. The calcium channel blocker cobalt chloride diminished but did not abolish the amplitude of the current in the NCU-1 cells (D). When NaCl was replaced with N-methyl-d-glucamine in the extracellular solution, the ion current with decreased amplitude was still recorded (E) suggesting the presence of non-selective channels in the membrane of NCU-1 cells, probably MACs. This was further supported by the linear current-voltage relationship demonstrated both in individual NCU-1 cells (F) and on averaged amplitudes of currents followed by linear regression (G). Human serum prepared from B-type blood also evoked current pulses in NCU-1 cells (H), but not in NCU-2 cells (I). Involvement of CD59 in blocking the serum-evoked current was further confirmed when CD59 was incorporated in the membrane of NCU-1 cells (J). Heterologous (rabbit) serum triggered current pulses in cells which were unrelated to the expression (NCU-2 cells; L) or absence of expression (NCU-1 cells; K) of CD59 on the membrane. Except G, each graph shows representative recording of 10 measurements.