From the experiments reported here it is evident that the amount of each complement component which combines with specific immune aggregates depends upon a number of factors, including the age of the complement; the concentrations of antigen, antibody, and complement; hydrogen ion and electrolyte concentrations; and time and temperature of incubation for fixation.
The experiments also reveal the following:
1. C'4 must be considered a combining component of complement because it is invariably inactivated or adsorbed by specific aggregates.
2. C'3, although necessary for the final action of complement, e.g., hemolysis or bactericidal action, is fixed only partially or not at all by antigen-antibody compounds.
3. C'2 is fixed together with C'4, as is evident from those experiments in which reactivations were conducted with a CO2-insoluble fraction which contained no C'4 activity.
4. Mid-piece, which contains C'1, generally stated to be the single combining component of complement, is adsorbed by immune aggregates in varying amounts depending on the experimental conditions employed.
5. No inactivation or adsorption of C'4 and of C'3 occurs upon the addition to specific aggregates of serum which has been heated at 56°C. for 30 to 50 minutes. This indicates that certain thermolabile constituents of serum are necessary for the fixation of C'4. Inactivation of C'4 by hydrazine or the removal of C'3 by the insoluble carbohydrate from yeast does not markedly influence the fixation of the other components.
6. The assumption that the mid-piece is the single combining component of complement is further questioned for the following reasons.
First, C'4 in fresh, untreated serum combines with specific aggregates, but C'4 in heat-inactivated serum does not. It is, therefore, apparent that a part of the combining complement must be attributed to C'4 and its carrier C'2. Second, it will be shown in a subsequent paper that C'1, heated at 56°C. for 45 minutes, combines with specific immune complexes, and in doing so may in fact inhibit the further combination of the components of untreated complement. This effect, first noted by Ehrlich and Sachs (27), is termed a "complementoid" action. Thirdly, it will also be shown on the basis of quantitative nitrogen data (28), that the combining nitrogen, in those instances in which large amounts of serum are added to a constant amount of aggregate, should for the most part be attributed to C'4 and its carrier C'2.
7. The nature and the molecular size of the antigen does not influence the qualitative picture of fixation of the complement components.
8. While it is tempting to speculate on the mechanism of complement fixation, it probably is advisable that such speculation be postponed until experiments similar to those reported here are performed with purified components. However, certain differences between the fixation of the complement components to specific aggregates and the adsorption of these components to non-specific agents, are apparent. For example, in specific immune fixation C'4, C'2, and varying amounts of C'1 are fixed, while in non-specific adsorption both to inorganic adsorbents (29) and to untreated bacteria (30, 31), no fixation of C'4–C'2 occurs while all of the other components of complement are adsorbed. Furthermore, C'3 is bound only partially or not at all to specific aggregates, while it is completely adsorbed or inactivated by non-specific adsorbents. It appears, therefore, that the highly reactive component, C'4, is directly and immediately involved in specific immune fixation.
9. A very small amount of specific immune aggregate combines with a large amount of complement; whereas a large amount of non-specific agent adsorbs only a small amount of complement. This difference is undoubtedly due to the marked chemical affinity of C'4–C'2 and C'1 for the surfaces presented by immune aggregates.
10. It can be now stated that when complement exerts its activity in hemolytic, bacteriolytic, or bactericidal reactions, C'4–C'2 and varying amounts of C'1 must first combine (fix) with the antigen-antibody compound in question; and that any secondary manifestation is dependent both on the adjunctive action of the unbound C'3 (26) and on the nature of the substrate employed.
The chemical and immunological implications of these results will be further elaborated in subsequent papers.