The experiments of the preceding communication showed that the therapeutic action of antipneumococcic serum depends to a considerable extent upon a certain non-antibacterial factor. The experiments reported in the present communication had two main objects; the first was to determine the distribution of the non-antibacterial factor among the various protein fractions of the serum, and incidentally to correlate this property of the factor with that of certain known antibodies, as well as to learn whether or not the protein fractions commonly excluded from refined preparations of antipneumococcic serum, have any therapeutic value; the second was to determine the rôle of the non-antibacterial factor in the therapy of pneumococcus infection, as exemplified by the experimental, dermal, pneumococcus infection in rabbits.
To determine the distribution of the non-antibacterial factor, Type I antipneumococcic serum was fractionated with (NH4)2SO4, and the antibacterial bodies were absorbed by concentrated suspensions of heat-killed pneumococci. The activity of the non-antibacterial factor in the different protein fractions was then tested for by adding varying amounts of the absorbed supernatant liquids to a certain constant, subeffective dose of serum. Even though the method of titration has certain inherent faults it was possible to ascertain that the non-antibacterial factor was apparently associated with all the globulin fractions to a similar degree, and to a certain extent with the albumen as well. In this respect the non-antibacterial factor resembles neither the antibacterial bodies of antipneumococcic serum, nor, for example, the known diphtheria or tetanus antitoxins.
The determination of the relative importance of the antibacterial and non-antibacterial factors in the therapy of pneumococcus infection in rabbits, was rendered possible by the fact that the various globulin fractions differ in their relative content of the two factors. Thus whereas the water-insoluble fraction of the globulin, precipitated when the saturation of (NH4)2SO4 is raised from 30 per cent to 50 per cent, contains about 90 per cent of the antibacterial (mouse protective) bodies, it apparently contains no more of the non-antibacterial factor than the water-soluble fraction. The minimal effective doses as well as the mouse protective unit content of the different fractions were determined. Before the tests were performed, it seemed that if the therapeutic effect of a given dose were to depend chiefly upon its content of mouse protective antibody, the amounts required for 1 M.E.D. would be determined by the number of mouse protective units it contained; on the other hand, if it depended primarily on the non-antibacterial factor, one would expect the amounts required for 1 M.E.D. to be of approximately the same volume. Actually the therapeutic effect was found not to depend entirely upon either factor, alone; the M.E.D.'s varied both in volume and m.p.u. content, the absolute concentration of either factor being inconstant and almost proportional to the relative concentration of the other. Thus a relative abundance of non-antibacterial factor made up for a scarcity of mouse protective antibody (as in the water-soluble globulin), and vice versa (as in the water-insoluble globulin). It appears, therefore, that when a relatively larger amount of antibacterial bodies are acting, less of the non-antibacterial factor is necessary; conversely, when more of the non-antibacterial factor is available, less of the antibacterial bodies are necessary.
The last experiment dealt with the practical question as to whether any protein fraction or combination of fractions is as good a therapeutic agent as the whole, unrefined serum from which it was derived. Two considerations were in mind in making the comparison; one had to do with the quantitative recovery of potency, the other with the qualitative effect of the various fractions as regards the lesion, course, and duration of the experimental disease. The results may be summarized as follows: (a) the total globulin was not only as good, but in this instance, definitely better than the original serum; (b) the total water-insoluble globulin, although almost as potent as the original serum, was qualitatively and quantitatively less effective than the total globulin; (c) the removal of the 30 per cent (NH4)2SO4 fraction not only diminished the potency of the remaining globulin to a considerable extent (at least 25 per cent), but also seems to have disturbed a certain equilibrium, which resulted in an undesirable zone phenomenon. As a result of this zone phenomenon, which, it is important to note, was not observed in the agglutination or mouse protection tests on the same fractions, the range of effective dosage is very narrowly limited. Although the results of agglutination tests correlated well with those of the mouse protection tests, both of these showed no parallelism with the therapeutic effects of the various fractions on the experimental, dermal, pneumococcus infection in rabbits.