When cholesterinized antigen is dropped into an excess of water, the rapid flocculation of cholesterin crystals is prevented by the fact that, as tiny aggregates form, they adsorb a protective surface of hydrophilic lecithin (i.e., antigen) which endows the particles with its own stable surface properties and thus prevents further aggregation. The colloidally dispersed antigen-cholesterin particles have approximately the same isoelectric point (pH 1.9), critical potential (1 to 5 millivolts) and coagulation value (0.75 M NaCl) as pure antigen particles of the same concentration, while the corresponding values for cholesterin are pH 2.1 to 3.4 (probably due to an associated impurity), >100 millivolts, and <0.001 N NaCl, respectively.
Presumably, this adsorption of antigen by the cholesterin nucleus is determined by the fact that the former has a lower surface tension against water. At any rate, many surface active substances (serum; alcoholic extract of milk, egg or any animal tissue; Na-oleate; Na-glycocholeate; Na-taurocholate) cause a similar stable dispersion of cholesterin; and conversely, many otherwise water-insoluble substances of the most diverse chemical structure can be made to form a stable colloidal suspension by adding antigen to their alcoholic solutions before dropping into water.
The colloidal suspension formed by antigen alone is very finely dispersed: only a few of the particles exceed the limits of dark field visibility. Cholesterin causes a marked increase in the number of these particles, out of all proportion to its mass; thus, one part of cholesterin to five of antigen causes a ten-fold increase in such visible particles, at the expense of the submicroscopic micellae formed by antigen alone. At the same time, the suspension becomes much more turbid. The particles remain discrete until the cholesterin: antigen ratio exceeds 1:1, when slight microscopic aggregation is observed; microscopic flocculation is seen only when this ratio exceeds 5:1, when there is not sufficient antigen to act as an efficient protective colloid.
Cholesterin therefore causes a coarsened dispersion of antigen by forming a relatively large nucleus upon which antigen is adsorbed. As shown in the text, the larger the antigen particle the greater is its avidity for reagin per unit surface or mass. Thus, the coarse sol formed by dropping water-into-antigen is about twice as efficient as a finely dispersed antigen-into-water sol of the same concentration. The coarsened dispersion caused by cholesterin completely explains the greater sensitivity of the cholesterinized antigen in complement fixation.
The same factor obtains in the flocculation reactions. In addition, the coarsened dispersion acts as a preliminary quasi-aggregation, facilitating by just so much the subsequent formation of visible clumps (or sedimenting aggregates) upon the addition of syphilitic serum; moreover, there is less surface in a coarse sol, with more reagin per unit surface, and correspondingly more efficient flocculation.
The foregoing would be of purely academic interest were it not for the following considerations. From several points of view cholesterin is unsatisfactory as a sensitizer for antigen. Its solubility in alcohol is small. Even the 0.6 per cent concentration used in the Kahn test is difficult to keep in solution. Yet, as our experiments show, its sensitizing action increases indefinitely with its concentration. If it were sufficiently soluble, even 3 per cent could be used to advantage, increasing the sensitivity of 1½ per cent antigen for complement fixation some 200 to 400 per cent, instead of about 50 per cent, as does 0.2 per cent cholesterin.
Since, as we have shown, the sensitizing action of cholesterin upon antigen is due solely to the coarse dispersion it causes, and since it is quite inert during the actual combination of the lipoid particles with reagin, it can be replaced by any substance with similar physical properties. The problem in hand was therefore to find a water-insoluble substance, very soluble in alcohol, with so high an interfacial tension against water that, as in the case of cholesterin, microscopic particles would adsorb antigen when the alcoholic solution of the two is dropped into water. Given such a substance, it would be possible to obtain a more sensitive antigen for both complement fixation and flocculation, but particularly for the former.
These theoretical expectations have been realized in a group of substances shortly to be reported: they make possible an antigen which is from 2 to 10 times as efficient in the Wassermann test as any now available.