Immunologic studies were performed on 27 blood specimens obtained at various intervals from 17 patients with acquired hemolytic anemia. RBC and serum agglutinin titers, hemolysins, and viability of coated RBC in serum and in saline were determined. Diseased cells, freshly separated from their parent sera, were highly resistant to incubation at 37 C, in buffered, isotonic saline. Heparin was found a most effective inhibitor of nonimmune hemolysis, as shown by its ability to preserve the integrity of normal and diseased RBC in serum or in saline.
Twenty-six RBC samples were eluted and the agglutinating substances compared with respect to solubility, pH, heat resistance, UV absorption spectra, and reactivity toward heparin and trypsin. Spectrophotometric data revealed the appearance and time-dependent increase of an absorption hand at 4100 Å upon incubation of normal RBC with eluted agglutinins. Reasons were given to indicate that this phenomenon was not hemolysis.
Coating of normal RBC could be inhibited by normal sera and by some patients’ sera. All fractions tested were inhibitory. In general, inhibition was independent of complement activity, but may have been due to (one of its) thermostable component(s). Eluates incubated with human sera could not be recovered, even with trypsinized RBC. Furthermore, no changes in serum agglutinin titers were obtained by the addition of eluates to human sera.
A critical analysis of the variables influencing the outcome of a Coombs’ reaction was made. Considering eluated agglutinins as "antibodies" with respect to RBC and "antigens" with respect to Coombs’ serum, the observed variables were separated into two independent groups according to their influence on: (a) RBC-eluate interaction; (b) agglutinin (or coated RBC)-Coombs’ serum interaction. The former was not affected by eluate concentration or "Group" (apparent immunologic specificity of eluates for different Coombs’ serum fractions), but varied with avidity of eluates (i.e., the speed necessary for optimal coating in saline, and the amount of agglutinin required for such coating). Avidity was, in turn, greatly influenced by the suspension media used. It was found that optimal coating with at least 2 agglutinins ("supersensitization") could be obtained, regardless of the type of sensitization (simultaneous or consecutive), when a saline suspension of RBC was used. While no damage to RBC resulted under these conditions, supersensitization in an albumin suspension could be carried to the point of total hemolysis. Thus, the attraction of eluted agglutinins to RBC was not (necessarily) mediated by antigenic groupings but by the surface as a whole, with theoretically no limit to the amount of agglutinin which could be fixed under suitable conditions. The agglutinin-Coombs’ serum interaction was subject to laws applicable to antigen-antibody reactions, with the complications common to cross-reacting, heterologous mixtures.
Examples of prozones in both eluate and Coombs' serum excess were given. A method was described for the determination of Coombs’ serum-eluate equivalence zone proportions (C/E ratios) in the absence of a precipitin test. During this procedure, most of the eluates were categorized into a gamma globulin and two non-gamma globulin "Groups." Cold hemagglutinins were excluded from the present study.
Statistical data involving eluates showed no correlation; however, direct Coombs’ titers on patients’ RBC vs. autoagglutinin titers of the respective parent sera gave a determination coefficient (r2) of 0.3662 with P(t) of 0.001.
The nature of hemolysins, RBC and serum agglutinins, and the possibility to define and classify them were briefly discussed. Emphasis was placed on artifacts and their possible overinterpretation. Three tentative conclusions were reached:
1) It is highly doubtful whether serum and RBC factors are identical or were so prior to adsorption on RBC. Hemolysis is dependent on the serum factor. The substance which can be eluted from RBC presumably represents a protective complex.
2) Despite superficial similarities, the eluted substances seemed to have individual characteristics for each patient studied. Evidence of their protein nature was overshadowed by properties common to nucleic acids.
3) While the value of a quantitative Coombs’ test is questionable since the number of RBC antigens (?) seems unlimited, much information about eluates can be obtained by employing identical RBC populations and the same Coombs’ serum.