Chapter IV Specificity of Rheumatoid Factor with Gamma Globulin of Different Species Chirmun: JOHN H. VAUGHAN,M.D. DR.VAUGHAN:In discussing the reaction of rheumatoid factor with animal r-globulins, I would first like to center your attention on a comparison of the activity of rheumatoid factor with animal 7-globulins to that with human 7-globulin. The data reported by Dr. Butler1 on the reactivity of bisdiazotized benzidine aggregates of 7-globulins with rheumatoid sera can illustrate this point. Figure 1 shows the results of tests of the rate of nitrogen absorption from rheumatoid sera by animal and human 7-globulins made insoluble through cross-linkage by bis-diamino-benzidine ( BDB ) . One finds that BDB-aggregated 7-globulins made from human 7-globulins absorb far more nitrogenous material from rheumatoid serum than did BDB aggregates made from rabbit or bovine 7-globulins. We like to consider these curves as a way of seeing roughly what proportions of material in a rheumatoid serum are reactive with the various types of 7-globulins. All agglutinating activity was removed from this serum by absorption with human 7-globulin aggregates a bit beyond the last point indicated on the curve. Activity against rabbit, bovine, and equine 7-globulins was all removed as well as activity against human 7-globulins. By contrast, the same projected point on the rabbit or bovine 7-globulin curves gave supernatants which were devoid of activity against rabbit or bovine 7-globulin, respectively, but not against human 7-globulin and not completely devoid of activity against the other types of 7-globulin in the case of the bovine curve and in reverse. Recovery of agglutinating activities from the insolubilized BDB 7-globulins could be affected by exposing the washed insoluble material to a glycine buffer at pH 3.7. From each of the three types of 7-globulins used for absorption, agglutinating activity against cells coated with human, rabbit, bovine, or equine 7-globulins was recovered. This was true of all three of the types of 7-globulin used in the absorption. This suggested to Dr. Butler and me that the relationship between these anti-7-globulin activities should be demonstrable in agar diffusion analyses. In this experiment, soluble BDB aggregates of human 7-globulin and solnble BDB aggregates of rabbit 7-globulin formed precipitates with rheumatoid factor isolated from the bottom fractions of a sucrose density gradient ultracentrifugation of the serum illustrated in figure 2. Fairly diffuse lines were formed which did not cross each other. Although spurs were not seen, as are characteristic of clinical cross-reacting systems, these data did indeed suggest that the BDB aggregates of human and rabbit 7-globulins were interacting with common anti-7-globulin molecules in the rheumatoid factor. 446 SPECIFICITY OF RHEUMATOID FACTOR FOR OTHER GG pg w 447 CC-BDB Fig. 1.- Adsorption of nitrogen to insoluble complexes. Turning to the isospecih serum discussed earlier, we used the same technics in analysis. With this serum, human 7-globulin regularly absorbs more nitrogen from the serum than does either rabbit or bovine 7-globulin. Doing the absorptions illustrated in figure 3, we were using very large amounts of nitrogenous material and getting relatively small additions to the insolubIe 7-globulin, so reproducing these curves exactly has been difficult. In the absorption with human 7-globulin aggregates all the anti-Gm activity [if one accepts the agglutination of Gm(a+) coated cells as evidence of anti-Gm(a ) activity in this particular serum] was gone from the supernatant after 180 pg, N of BDB aggregate human y-globulin had been added to the serum. After exhaustive absorption (more than 2000 pg. N of BDB aggregate human 7-globulin), all agglutinating activity was removed from this serum. Exhaustive absorption with rabbit or bovine y-globulin could remove only the agglutinating activity against the respective type of 7-globulin. The autogenous 7-globulin prepared from the patient was made into insoluble 7-globulin by BDB and used for absorption. It can be noted that its curve fell intermediately among the others. The fact that the autogenous aggregated y-globulin curve was quantitatively less high than the RDB aggregate human 7-globulin (F-I1) curve was interesting and certainly must have represented somehow the isospecificity of this serum. We could not tell from this whether there was a difference because some of the groupings in BDBlaggregated human YLglobulin were missing from 448 IMXfUNOLOGIC ASPECTS OF RHEUMATOID ARTHRITIS AND SLE Fig. 2.-Precipitates formed in agar diffusion between ultracentrifugally isolated rheumatoid factor and soluble y-globulin aggregates. the autogenous BDB aggregates or whether they were just fewer in the autogenous aggregated material. We could not carry these curves out any further with larger amounts of BDB 7-globulin because the errors of the experiment became enormous in that area. When this serum was exhaustively absorbed with repeated exposure to autogenous 7-globulin, essentially all the agglutinating activity was removed from the serum. These data are consistent with the assumption that the apparent isospecificity that exists for anti-animal 7-globulin activities in the rheumatoid factors in sera containing anti-7-globulin activities are part of a total immune response to autogenous denatured 7-globulins. D R . FUDENBERG: Dr. Vaughan, in your experiments the specificity against the patient’s own 7-globulin may have already been partially saturated. This might explain why the autogenous 7-globulins absorbed less than others. It would be perhaps wise to repeat these experiments using purely the 19s rheumatoid factor antibodies. DR.SINGER:In collaboration with Drs. Rafter, Plotz, Halberstam, and Schuller, we attempted to produce in experimental animals an antibody which would have two main characteristics of rheumatoid factor: a large molecular size and the ability to react with human 7-globulin. Two experimental lines were followed. The first utilized heterologous 7-globulin and the second utilized autologous 7-globulin. Latex particles with absorbed 7-globulin were used as particulate carriers. Our choice of this method was based on the fact that human 7-globulin coated on latex particles of 0.2 p served as a potent stimulator of antibody production. It also eliminated the individual variation in antibody production in different animals of the same species. Our decision was justified by a study of goats immunized with latex particles 449 SPECIFICITY OF RHEUMATOID PACTOR FOR OTHER GG 75 ROC ug. N absorbed 25 500 )rg. Fig. 3.-Adsorption 1000 1500 BDB-GC of nitrogen to insoluble complexes. alone, human 7-globulin alone, latex particles with absorbed human r-globulin, and finally, human 7-globulin in Freund's adjuvant. When the goat serum was tested for agglutinating antibodies, high titers were obtained both with human 7-globulin coated on latex particles and with 7-globulin in Freund's adjuvant. Most precipitating activity was in the system of 7-globulin in Freund's adjuvant, less activity being exhibited by latex 7-globulins. Human 7-globulin showed low titers, low precipitation, and negative results were obtained with the latex particles alone, as expected. Following these findings, latex particles were coated with human 7-globulin and injected into cows, horses, sheep, goats, and rabbits. All primary and successive bleedings were studied by DEAE column chromatography, agglutinating activity, precipitation curves, sodium chloride density gradients, analytic ultracentrifugation, double agar gel diffusion, and immunoelectrophoresis. Prior to immunizations all animals showed a 19s component in their sera as detected by analytic ultracentrifugation. We found that all animals immunized with latex particles coated with human 7-globulin produced two antibodies, one a 7s component and the other a 19s macroglobulin. However, the pattern of distribution of antibody of low and high molecular weight and the time of appearance of this antibody varied among the different animals tested. Figure 4' shows the chromatographic pattern of cow sera obtained at varying intervals during the immunization periods. The location and titers of 450 IhlhlUn'OLOCIC ASPECTS OF HHEXJMATOID ARI'HRI'rIS AND SLE FII s.c TITER R 0 302 1 n 20 60 100 20 60 I00 Fig. 4.-Chromatography of cow sera (10 ml.) on diethylaminoethyl cellulose. Cone sphere gradient 0.02M,pH8-0.3M, pH5-1M NaC1. #301-309. serologically active fractions as determined by tanned sensitized sheep cells are shown in relation to protein peaks. In cow sera, peak I of low molecular weight coincided with the 7 2 region and peak I1 was located within the macroglobulins. The initial bleeding revealed serologic activity only within the second, or macroglobulin, peak. After booster immunization activity was noted in the y2 region, while at the end of 6 months the only serologically active material again was a macroglobulin. In horses, sheep and rabbits the distribution of antibodies appeared differently. For example, figure 5 illustrates the chromatographic pattern of horse sera at varying intervals during the immunization period. Antibody appears in both the y1 and 7 2 fractions and continues in both fractions throughout. The control percentage is from 2 to 5 per cent of the total y-globulin. Figure 6 shows the precipitation curves obtained by adding human 7-globulin in varying amounts to aliquots of experimentally produced macroglobulin obtained by column chromatography. The shape of the curve is similar to that obtained with the precipitation of the rheumatoid factor and denatured 7-globulin. The reactivity of various animal macroglobulins with various animal globnlins coated on tanned sheep cells was tested. As expected, in every case the specificity was direct to human 7-globulin with less specificity shown to 7globulin of other species. The cow macroglobulin produced the greatest agglutination reaction with human 7-globulin and cross-reacted with rabbit, horse and sheep but not cow sera. Similarly, sheep macroglobulin reacted with human, rabbit and cow 7-globulins, but not with that of sheep. In every instance there was no serologic activity with autologous 7-globulin. 451 SPECIFICITY OF RHEUMATOID FACTOR FOR 0Tm.R GG FlI S.C.,TITER 1 /05 I:"," 20 60 20 60 loo 100 PER C€NT €LUAT€ Fig. 5.-Chromatography of horse sera ( 10 ml.) on diethylaminoethyl cellulose. Cone sphere gradient 0.02M, pH8-0.3M, pH5-1M NaCl. Following this line of investigation, rabbits and goats were immunized with latex particles coated, respectively, with autologous rabbit and goat y-globulins. Animals immunized by this procedure produced agglutination titers not only with the autologous 7-globulins but also with human 7-globulin. However, very low titers-1:32 and l:6P-were obtained. While these titers may have been significant as being indicative of autoantibody production, they were too low for quantitative data to be obtained. We are considering developing this technic further by utilizing autologous 7-globulin on latex particles for the production of autoantibody. In order to further characterize these antibodies we postulated that if two major characteristics of rheumatoid factor were present, other similarities with rheumatoid factor might exist. Table 1 compares rheumatoid factor with the macroglobulin antibody to human 7-globulin produced in various animals. They share the following characteristics: both are 19s macroglobulins which react strongly in the tanned sheep cell and latex fixation tests with human y-globulin; the agglutinating activity of both is destroyed by mercaptoethanol; both exhibit cross-reactivitywith other species of 7-globulin; and they produce similar precipitation curves. The only difkrence between them is that rheumatoid factor reacts with autologous denatured y-globulin while none of the animal macroglobulins so far produced shows such autologous activity. In summary, the injection of heterologous y-globulin into animals produced macroglobulin antibodies similar in most physical and immunochemical 452 IMMUNOLOGIC ASPECTS OF RHEUMATOID ARTHRITIS AND SLE 751 y601 PEAK Il 308 REACTION VOL. 4mf $45 OV 540 ’ 7h0 960 ’ Idso’ I k O ’ d40 MICROGRAMS OF HUMAN GAMMA GLOBULIN N ADDED lb0 ’ 3 b Fig. 6.-Precipitation curve of a 19s macroglobulin cow serum immunized with human y-globulin. (Vertical column = ppt. y N per ml.) properties to rheumatoid factor. This strongly suggests that rheumatoid factor is a true antibody. The most distinctive feature of rheumatoid factor is its ability to react with denatured autologous 7-globulin. DR.MILGROM: The data presented by Dr. Vaughan indicate that the rheumatoid factor may be considered an antibody primarily directed against human y-globulin and that the reactions which are obtained with y-globulins of other species origin should be considered as cross-reactions. On the basis of Dr. Vaughan’s experiments as well as on the basis of experiments published by one would have to conclude that there exist essentially other two factors: one which combines with human y-globulin only and a crossreacting factor which combines with both human and foreign species 7-globulin whose existence was rather well established by Dr. Vaughan’s experiments. I will present the data only briefly. InasAs our findings were publi~hed,~ much as they were obtained by quite different technics, our results do not contradict those of Dr. Vaughan. We wished to explore whether in addition to these two factors there was another factor directed only against foreign species 7-globulin (represented by rabbit 7-globulin in our investigations). We confined our experiments to the agglutination of sensitized red blood cells. Three indicator systems were Iised. ( 1 ) A system elaborated by Cohen et ala5which consisted of alligator red blood cells sensitized by subagglutinating doses of rabbit anti-alligator erythrocyte serum. This system was designated Ar. (2) This was elaborated in our laboratory6 on the basis of the discovery by Kellner and Hedal.’ It consisted of rabbit blood group G red blood cells sensitized by incomplete anti-G serum of rabbit origin. We will call this indicator system Rr. ( 3 ) A system elaborated by Waller and Vaughan8 consisted of Rh-positive red blood cells sensitized by incomplete anti-Rh serum. The anti-Rh serum “Ripley” used in this investigation is known to combine with most rheumatoid sera. This was designated Hh. 453 SPECIFICITY OF RHEUMATOID FACTOR FOR OTHER GG Table 1.-Comparison between Animal 19s HGG Antibodies and Rheumatoid Factors /9 s A m HGG. /9 s #EUM. FACTORS + In the experiments presented in figure 7, a mixture of Ar Rr was added to rheumatoid sera and examined for mixed agglutination. With all rheumatoid sera tested, positive results were obtained and the clumps found on microscopic examination appeared as those presented in figure 8. However, when Hh was added to rheumatoid sera, in most instances the mixture of Ar homologous agglutinates were recognized consisting of separate alligator erythrocyte clumps and human erythrocyte clumps (fig. 9 ) . Only a minor proportion of rheumatoid sera, about 1 out of 5, produced mixed agglutination of the Ar Hh mixture. These results already indicated the existence of a separate factor for rabbit 7-globulin because if such a factor did not exist the mixture of Ar Hh should have always been agglutinated in mixed clumps as was the case in the mixture of Ar Rr, Further experiments were designed to compare the serologic properties of mixed agglutinates composed of Ar Rr and Ar Hh (the latter were produced by exceptional rheumatoid sera capable of agglutinating Ar Hh in mixed clumps). The results of this experiment are presented in table 2. As can be seen, the Ar Rr agglutinates were not affected by 1 per cent F-I1 of human serum but were completely dispersed by 1 per cent F-I1 of rabbit serum. Mixed agglutinates composed of Ar Hh were affected quite difhrently by F-I1 preparations. The addition of human F-I1 caused the dispersion of human red blood cells leaving the alligator red blood cells agglutinated, and the addition of rabbit F-I1 resulted in the dispersion of alligator erythrccytes leaving human erythrocytes in clumps. These experimental data can be interpreted by assuming the existence of three types of rheumatoid factor: ( 1 ) one directed against human 7-globulin only; ( 2 ) one capable of combining with both human and rabbit 7-globulin; + + + + + + + + + 454 IMMUNOLOGIC ASPECTS OF RHEUMATOID ARTHRITIS AND SLE I Appearance Mixed Atglutination I Ar +Rr P06. I 0 0. o O O Neq with mod rheum.sera &s with some rheum.sera I AY+Ho 0 0 0 00 0 o0 0 @ :: Neq. 0 Fig. 7.-Ar agglutination. + Rr mixture added to rheumatoid sera and examined for mixed and ( 3 ) one directed against rabbit 7-globulin only. We believe that factors 1 and 3 appear in much larger quantities than factor 2. We could demonstrate factor 2 only in 1 out of 5 rheumatoid sera (those which were capable of Hh). Most probably in the remaining producing mixed agglutination of Ar sera the titer of the cross-reacting factor was not high enough to produce positive mixed agglutination reactions. + SPECIFICITY OF RHEUMATOID FACTOR FOR OTHER GG Fig. 8.-Clumps found upon microscopic examination of agglutination of Ar 455 + Rr. DR.VAUGHAN:Is this postulating a heteroligating type of antibody activity? DR. MILGROM:Not many antibodies but many reactive sites. I think, according to Dr. Kunkel, there are seven reactive sites in rheumatoid factor. DR.SINGER:Did you test with other than the rabbit 7-globulin? DR.MILGROM:In this system bovine would not be active in the dispersion, but I did not have an incomplete bovine antibody system which could be applied for this type. I think, as Dr. Milgrom indicated, there is no inconsistency DR.MCDUFFIE: between his findings and those of Dr. Vaughan. They certainly coincide with the observations of Drs. Fudenberg and K ~ n k e l They .~ showed there was no relationship between the ability to agglutinate anti-D coated cells and the ability to agglutinate sensitized sheep cells. Everything depends on the type of coat being used. We have been studying the reaction of rheumatoid factor with rabbit y-globulin using different coats and an inhibition technic. Figure 10 illustrates what happens when one tests the ability of human 7-globulin to inhibit the agglutination of tanned sheep cells by rheumatoid serum. The first five lines shew the results obtained with tanned cells coated with human 7s 7-globulin, and the second five lines the results when cells coated with rabbit y-globulin are used. Human u-globulin inhibits both systems equally well. Figure 11 shows the reverse experiment. If rabbit y-globulin is used as an inhibitor, it does not show any activity against cells coated with human 7-globulin. On the other hand, it strongly inhibits the agglutination of cells coated with rabbit 7-globulin. The failure of rabbit 7-globulin to inhibit the reaction of cells coated with human 7-globulin appears to be due to the absence of much antibody-like activity in most rheumatoid sera. 456 IMMUNOLOGIC ASPECTS OF RIIEUhlATOID ARTHRITIS AND SLE Fig. 9.-Clumps found on microscopic examination of agglutination of Ar + Hh. Next we investigated the effect of the denaturation and aggregation of the reactants. The first part of figure 12 shows the results of inhibition tests against tanned cells coated with 7s rabbit 7-globulin isolated by chromatography.Both aggregated (20 minutes at 68 C.) and unaggregated rabbit 7-globulin appear to be equally effective inhibitors. Somewhat different findings are observed when aggregated rabbit 7-globulin is used to coat the cells, however. In the second part of this figure it can be seen that unaggregated rabbit 7-globulin is a rather poor inhibitor, whereas aggregated inhibits well. It may also be noted that the absolute titer of the serum against aggregated coated cells is slightly higher than against cells coated with unaggregated rabbit 7-globulin. It does seem possible that the process of heating and aggregation has altered the 7-globulin is some way so as to uncover or produce new antigenic groups. We attempted by cross-absorption to see whether activity against the aggregated coated cells might remain after absorption with cells coated with unaggregated rabbit 7-globulin. Unfortunately, using such cells to absorb the sera we found that the coat comes off the cell and prozones occur when the supernatants are tested. Thus the results were inconclusive on this point. We are now using both aggregated and unaggregated anti-sheep cell antibodies to coat the cells, but have no results to report as yet. We have also compared the behavior of rheumatoid sera with that of a sheep anti-rabbit 7-globulin serum (fig. 13); they are identical with those obtained with rheumatoid serum. When unaggregated rabbit 7-globulin is used to coat the cells, both aggregated and unaggregated rabbit 7-globulin act as equally good inhibitors. If the cells are coated with aggregated rabbit 7-globulin, aggregated rabbit 7-globulin behaves as a much more potent inhibitor. We believe this phenomenon of enhanced reactivity following aggregation or denaturation may be characteristic of a number of antigen-antibody systems. 457 SPECIFICITY OF RHEUMATOID FACTOR FOR OTHER GG Table 2.-Dispersion of Mixed Agglutination: Effect of Fraction ZZ of Human and Rabbit Origin on Mixed Agglutinates Composed of Ar 4- RT and AT Hh + Microscopic Appearance of Ar + Rr agglutinates incubated with: lr;r,FIIof pooled human serum ______-__-____ Alligator erythrocytes Mammalian erythrocytes l%FIIof pooled rabbit serum saline + Hh agglutinatea incubated with: -_ 1% F I1 of pooled human serum 1% F I1 of pooled rabbit serum Aaplutinated Free .._____. ._______ Free Mixed apglutinates -_____-__-__ Ar Mixed agglutinates Free Free Agglutinated saline - __ Mixed agglutinates ____-_______.-_ Since the adsorption of 7-globulin onto the surface of a tanned cell h a y produce some structural alteration in the molecule, the sensitized sheep cell test was also employed. The results of agglutination and inhibition tests are recorded in figure 14. Both 7 s and aggregated rabbit 7-globulin as well as normal rabbit serum appear to be potent inhibitors; 7s human 7-globulin and whole human serum were only slightly less reactive. Surprisingly, aggregated human 7-globulin was much less reactive in the rheumatoid factor-rabbit y globulin system than was unaggregated human 7-globulin. The explanation of this apparent paradox is not clear. In summary, we believe that our results are most consistent with the hypothesis that the reacticn of rheumatoid factor with rabbit 7-globulin is analogous to that of an immunologic cross-reaction. Although human 7-globulin contains “antigenic” groups not present in rabbit 7-globulin, all the “antigenic” sites of rabbit y-globulin are present in human 7-globulin. There may be quantitative differences. The enhanced reactivity of rabbit 7-globulin after aggregation and denaturation by heating may result from a change in antigenicity but this has not yet been proved. DR.VAUGHAN: I would like to call your attention to the serum I discussed in an earlier session in which autologous y-globulin from an individual had no inhibitory activity for his own agglutinating activity, but acquired it after the alkaline denaturation. This raises the major possibility that the. differences between Dr.Milgrom’s data and ours are that we were dealing with partially denatured 7-globulin in our BDB aggregates, whereas Dr. Milgrom was dealing with a much more nearly native 7-globulin in his sensitized cells. Dr. Vaughan, we have not tried your particular aggregate DR. MCDUFFIE: but have tested a number of insoluble conjugated y-globulins in an attempt to measure rheumatoid factor quantitatively. We have had great difficulty getting reliable quantitative data because of high blanks and nonspecific adsorption of serum proteins to the conjugates. I am a little surprised at the large amounts of rheumatoid factor that appear to be present in rheumatoid sera according to your data. I would also like to ask about the blanks with normal sera. All our preparations have removed so much nitrogen from normal sera containing no detectable rheumatoid factor that it has become a major difficulty. DR.VAUGHAN: These sera, first of all, were selected for being especially high 458 TMMUKOLOGTC ASPECTS OF R3EUMATOID ARTHRITIS AND SLE . Coated Conc with Inhibitor mg/ml Human I1 None fiuman I1 11 I1 11 II 11 It None Human It I1 II I1 11 I1 Rabbit 0 Dilution (reciprocal) of RF pool 1 7 50 100 200 400 800 1600 3200 6400 12800 25600 - - * * * * * - i-t - - - - * i+ * * * 0.05* * * * * o . o 1 - * * * * u * - - * * -+ t- i +- *- * 0.70 0.10 * * - O . O 5 - + + + t t e 0.70 0.10 -I+ i+ 0 . 0 1 - * + t i + * - * * * i+ -* --- -- -* u u - --- --- -- . H i + + - - - + --- - Fig. 10.-Inhibition of rheumatoid agglutination of tanned sheep cells coated with human and rabbit y-globulin by human y-globulin. Coated with Human II Conc Inhibitor mg/ml None Rabbit II ,I I1 II Rabbit II None Rabbit ,I ,I II I, 0 50 Dilution (reciprocal) of RF pool #7 100 200 400 800 1600 3200 6400 12800 25600 - - * * * * * * * * +t U i+++ 0 . 1 - * * * * * * * * u - * * * * * * * 4- 4 +1.0 - - - - .- - 0.5 - - - - - 0.1 - - - - - - - 1.0 i+ ++ U I+ i+i+ ++ 0 . 5 - i + i + U + f t + + + ft ft * * * i+ - Fig. 11.-Inhibition of rheumatoid agglutination of tanned sheep cells coated with human and rabbit y-globulin by rabbit y-globulin. titer. They are not casual rheumatoid sera. Second, we have found it useful to dilute these sera 1:6 or 1:lO and allow them to stand awhile. A spontaneous precipitation occurs here, very small in quantity, but one which is quite large in relation to the nitrogen added to the aggregates. Hence what we are examining is essentially whole sera that have been diluted up in buffer, allcwed to stand, exhaustively spun and then analyzed with the inscjlubilized BDB aggregates. I think this circumvents the problem you mention. DR.KAPLAN: What kind of plot do you get if you graph the absorbed nitrogen to the amount of BDB-labeled F-II?Is this a changing proportion? If you noticed, there was quite a difference in the amount of DR.VAUGHAN: nitrogen added per pg. of BDB-aggregated 7-globulin between the two sera shown. We have yet to clarify in our minds whether this indeed represents some difference in combining ratio for the BDB aggregates or whether this is a difference in the way we prepare one bxtch of aggregates to another. In terms of reproducibility, I would emphasize that these are orders of differences and are reproducible in that way. Human y-globulin has consistently absorbed more than rabbit 7-globulin. But the absolute figures for quantities absorbed vary by 10 to 20 fig. of nitrogen. DR.WITEBSKY: May I call attention to the fact that the method discussed by Dr. Milgrom of dispersion cf agglutinates is a very fine quantitative procedvre. I feel I should recommend this method because it discloses many more 459 SPECIFICITY OF RHEUMATOID FACTOR FOR OTHER GC: Rheumatoid serum pool 1 7 Sheep c e l i s coated with 7s RGC (0.8 mg/ml) or aggregated RGG (0.8 mg/ml) Coated with 7s RGG Inhibitor AggRGG 0.8 0.5 0.1 0.8 0.5 0.1 None 75RcG 0.8 0.5 0.1 AggRGG Reciprocal of d i l u t i o n s of rheumatoid serum 20 40 80 160 320 640 1,280 2,560 5,120 - * * t + * c c * * None 7sRGG Agg RGG Conc. mglml None 0.8 0.5 0.1 - ++ tF ft -- --- * - * - -* * - * * - - . I * + - --- - - f -- - -- - - -- -- -- - - * * * * + ) * ** * -I+t+ +* - + k + t . H + , U f . + * * * * M i + . * f - ++ + -- *++ + -- -- -- * * - - * - - -- 10,240 -- - *- - Fig. 12.-Inhibition of rheumatoid serum by 7s rabbit y-globulin and aggregated rabbit y-globulin. differences than does the absorption experiment with denatured material. I believe the information gained this way is more accurate. Of course, the material to be used in a dispersion experiment can be denatured also but I believe one wants to get material as undenatured as possible. Is there any relationship between the observation that the DR. EPSTEIN: precipitin reaction occurs best with dilution of the rheumatoid serum and the fact that the prozone reaction, as seen in the Gm specific systems, will disappear as you dilute? Do you think that the failure to have as good a precipitin reaction without dilution relates to the removal of a prozone reaction between the 7s and 19s of the individual rheumatoid serum? Do you have an explanation for the observation made on the increase in precipitin reaction in large volume? DR.VAUGHAN: Our thought was, as you indicated, that this was a matter of diluting out inhibitor 7s more rapidly than the agglutinating or precipitating activity, for whatever physicochemical reason this would be. I think the major problem with accepting this unreservedly is Dr. Harboe’s finding, confirmed by Dr. Marion Waller and others, that some of the sera with prozones in the anti-Rh (Gm) system actually exhibit isospecificity. Thi; would say that the y-globulin which should be there to inhibit and thereby to induce the prozone is not there, or is not there in very large quantity. DR.EPSTEIN: The material that Dr. Vaughan spins off after dilution of the serum theoretically should contain on dissociation the two constituents in small quantity. Have you looked at the material spun off, dissociated and seen its specificity? DR.VAUGHAN: We have not actually looked at them. We assumed that we 460 IMMUNOLOGIC ASPECTS OF RHEUMATOID ARTE-IRITIS AND SLE (0.8 q h l ) Sheep c e l l s coated with IS I&% or aggregated RGC (0.8 mglml) Coated with Inhibitor kCiDrOCa1 Conc. lllglml Sheep a n t i PCO-S-1228596 None 20 Of dilutions of Anti BcG 5,120 80 160 320 640 1,280 2.560 40 10,240 ~ 7s RGG None 7SRGG - + u - I + c + t + * * * * 0.8 0.5 0.1 AgsRcG Agg RGG 0.8 0.5 0.1 None 7SRGG Ag*m 0.8 0.5 0.1 0.8 0.5 0.1 --- ** * * * - - - +6 - - + + + + * * * * + f . -- * * - + k + + i + * * * * * * -- * * -- * ++ * -* ++ * - * - * - +t ++ +t f+ -- - -- ++ +c + + + + * f f + + f + + + i + + b - * * * * * * * +c ++ i+i+ - * i + * + * i + i + + t 20,480 -~ - +t * * * * * ++ * -- -- -- -- - - - - - Fig. 13.--Inhibition of sheep anC-rabbit-y-globuliii by 7s rabbit 7-globulin and aggregated rabbit y-globulin. would find, much as you did, that there was agglutinating activity there. This is something we should do with the various types of specificities. DR.WITEBSKY: We always refer to an antibody against a protein of a foreign species as “heteroantibody.”So, we would call the antibcdy effective against rabbit 7-globulin a heteroantibody and we would use the word isoantibody to denote an antibody against human 7-globulin. DR. VAUCHAN: This is a very difficult problem of terminology. Indeed, is this an antibody to denatured autogenous 7-globulin? DR.WITEBSKY: From our point of view and I believe from the general immunologist’s point of view, an antibody found in human serum against rabbit protein is a “heteroantibody”while an antibody against protein from different human individuals, but not from the antibody-producer, is an isoantibody. Only the antibody active against the protein of the individual himself would be called autoantibody. Let us at least agree on that basic concept of terminology; otherwise we will never understand each other. DR. VAUGHAN: This is acceptable except for one possibility. Did you mean by heteroantibody antibody that is evolved from being immunized with a heterologous antigen? This is a problem. If we have antibody activity against rabbit 7-globulin, does it mean we have got rabbit 7-globulin in us as an antigen? DR. MILGROM:Antibody is defined as a serologic reaction evolved by immunizing stimulus. DR. BENACERRAF:I share your view on this, Dr. Vaughan. Usually the customary thing considered in specificity of the antibody is “what is the immunologic stimulus?” The cross-reactivity is what you see in vitro according to what you add to your antibody, but the total specificity is what the 461 SPECIFICITY OF RHEUMATOID FACTOR FOR OTHER GG Sheep c e l l s coated with rabbit A 9 anti-SC serum Inhibitor Conc. rnglrnl - None 7s RGG 1.0 0.01 Agg RGG 1.0 0.01 Normal R serum 1:lO 1:LOOO 7s HGG 1.0 0.01 Agg = 1.0 0.01 Normal H serum 1:lO 1:looo * All (-) a t 28 None - ----- + f D i 1 u t i o n J o r. 56 112 224 448 896 c * - * * - . , * f * - * - * + - - - - * * + - # * * * . - - 3500 +* * - * * - I + * * * * - * * * - - * + +* * * * + *-I+ t - 1792 - - + - - - - * - - * * - - + - -* - + * * * * - - - * - ** ** +* * * * - - 1:7000 Fig. 14.-Agglutination of sensitized sheep cells by Jor.Inhibition by human and rabbit y-globulin. antibodies have been designed for by the immunized animal. The trouble in this situation is that one assumes that this is some form of denatured autologous 7-globulin; since you don’t know what it is. That gives rise to this degree of confusion. DR. HEIMER:I have great difficulty understanding one aspect of Dr. Milgrom’s experiment which deals with the dispersion of the mixed agglutinate from alligator-rabbit sensitized cell and the rabbit-rabbit sensitized cell on adding human 7-globulin to it. If you add sufficient amounts of human 7-globulin to this mixed agglutinate you generally should get a dispersion because, if you reverse the procedure and absorb out the serum first with human 7-globulin, you would not get a reaction with any of the rabbit systems. DR. MILGROM:You are right. However, these experiments are mutally controlled in a quantitative way. One per cent human y-globulin does not disperse agglutinates of red blood cells sensitized by rabbit antibody and the rabbit 7-globulin will disperse these agglutinates not only at a 1 per cent concentration but also when much more diluted. For the agglutinates of red blood cells sensitized by human antibodies, the reverse is true. Admittedly if one used very concentrated human 7-globulin, it would also disperse the agglutinates of red blood cells sensitized by rabbit antibodies but I don’t think that this is the purpose of experiments of this type. DR.VAUGHAN: I think it would be very interesting if you would denature some of your human 7-globulin and use it in the normal quantities to see whether this now would disperse. DR. MILGROM:On the basis of your data I am almost positive that it would produce the dispersion. 462 IMMUNOLOGIC ASPECTS OF RHEUMATOID ARTHRITIS A N ) SLE DR. WILLIAMS: We have been interested in a little different approach to this same problem. We have utilized polyamino-polystyrene particles diazotized to different forms of 7-globulin after the methodology developed by Gyenes and Sehon.lo When one takes polyamino-polystyrenecolumns the particles of which have been diazotized to human F-I1 (the total column containing probably 120 mg. of human F-11), and then passes 16 cc. of sheep cell-absorbed and heat-inactivated rheumatoid serum over the column, one notices that the early effluent material, which is pretty much like serum, has retained sensitized sheep cell activity but completely lost latex activity. Thereafter, fractions in the effluent column retain both sheep cell and latex activity. The material which has been adsorbed on the column in this rather complicated form of adsorption is then eluted with reagents that are known to dissociate antigen-antibody complexes such as 6M urea or acid buffers at pH 3.5 or 4.5 and immediately dialyzed back to normality against buffered saline. In the eluted fractions one gets some relative concentration of rheumatoid factors. In this particular instance the material that was eluted from the column had a very high titer in the latex test but also a relatively high titer in the sheep cell reaction. The interesting finding in this column work which is a rather crude preparative method in contradistinction to the exquisite method previously described by Dr. Milgrom is that the material which is eluted from the column is both 7s and 19s y-globulin on immunoelectrophoresis. There are no other serum proteins noted. The finding that is still unexplained is what the 7s material constitutes. When the same serum was run over two different types of antigen column diazotized to polyamino-polystyrene particles-rabbit 7-globulin and human aggregates-different patterns were obtained. When the serum was passed over a rabbit 7-globulin column where 120 mg. of rabbit F-I1 had been diazotized to the polyamino-polystyrene particles, there was a striking ahsence of sheep cell activity in the first five fractions, whereas latex activity in titers of 1:10,240 was still retained. It was only until the sixth fraction that minute amounts of sheep cell activity did appear in the effluent. When this material was eluted from the column with 6M urea and dialized immediately to get rid of the urea, tests in the sensitized sheep cell reaction showed titers of 1:2O,OOOin the sheep cell reaction and very low titers ( I:% 1:80)in the latex fixation test. Again, this material was both 7s and 19s on immunoelectrophoresis and on analytic ultracentrifugation about 60 per cent was 19s material and 40 per cent was 7s material. The same serum was then passed over an aggregated human 7-globulin column and there was a striking absence of reactivity to human 7-globulin in the effluent material up t o about the sixth fraction where sheep cell-like activity was retained to titers of 1:2,560 and the latex activity was completely void in the first six fractions. These first fractions of the column effluent were similar to human serum in terms of protein concentration and patterns on immunoelectrophoresis. SPECIFICITY OF RHEUMATOID FACTOR FOR OTHER GG 463 This then is a demonstration of the possible application of such antigen columns to the problem of trying to separate out the specific reactivity of rheumatoid factors, some of which may react with only human 7-globulin or only rabbit 7-globulin. However, the majority of rheumatoid factors isolated by these procedures reacted with both human and rabbit 7-globulins. DR. CHRISTIAN: Regarding the fractions that had high sheep cell activity and very low latex titers, were other systems utilizing human 7-globulin comparable to the latex tests? DR.WILLIAMS: Yes, the latex agglutination followed most closely the anti-Rh coat testing in the ‘effluentsor the material eluted. One thing that concerned us was that when we diazotized an antigen-like 7-globulin to columns, how much of the diazotized material would possibly be eluted with such things as 6M urea. So we tested this by taking some of Dr. Harboe’s serum, kindly lent by him, which was Gm( a+b- ). This was passed over a human aggregate column. If any of the aggregates were eluted by the 6M urea, the effluents should be Gm(a+b+), but instead effluents retained the specific Gm activity of the testing system. We are fairly sure at this time that the material which we elute from the column is not what we diazotized to it originally. Have you tested for 7s rheumatoid factors with density gradDR. EPSTEIN: ients? DR.WILLIAMS: No,these experiments have not yet been done. We do know that in every experiment to date the ubiquitous 7s 7-globulin appears. It does not make any digerenee which high titer rheumatoid serum we test. We thought perhaps in the 19s rheumatoid factor reaction with the diazotized 7-globulin that somehow 7s may be trapped as a complex and comes off at thc same time as the eluting material is applied. The other possibility is that this is active 7s rheumatoid factor or anti-7-globulin antibody. DR. FUDENBERG: May I suggest that you use a control serum other than Dr. Harboe’s. The data in the literature indicate that Gm(b-) does not become Gm(b+ ) upon degradation; rather, Gm( b+ ) becomes Gm( b- ). Gm(a+) does not lose it. So, if you wanted a G m ( a - b f ) put on, then aggregated by urea and come off, you would have Gm ( a+ ) rather than Gm (a- ) and Gm(b-) rather than Gm(b+). DR. HARBOE: The key question here is whether some of the aggregates get through on elution. The material used to diazotize on the column was Gm( b+ ) and what got through the column was Gm( b- ). DR. MZLGROM: In 1956 we published a paper on an agglutinating factor found in 0.5 per cent of normal and pathologic human sera.ll We called it anti-antibody because it specifically combined with the antibody portion of the antigen-antibody complex. I was recently able to dernonstratel2that sera of 2 to 3 per cent of apparently normal rabbits contain a factor with serologic properties that are very similar to those of the factor previously demonstrated in human sera. The most convenient way to detect rabbit anti-antibody was the agglutination of rabbit blood group G red blood cells that had been sensitized by incomplete anti-G 464 IMMUNOLOGIC ASPECTS OF RHEUMATOID ARTHRITIS AND SLE Table 3.-Znhibition of Agglutinating Activity of Rabbit Sera Containing (A) Anti-Antibody and ( B ) Serum Group RGGZZ Zsoantibody Inhibiting specimen Agglutination of blood group G erythrocytea (sensitized by incomplete anti-G serum #log?) after addition of 10 agglutinating units d A or B mixed with the inhibiting specimen at a dilution of 1 to: 1 3 ___-___---___-- ++ - ++-I- 9 +++ - 27 81 +-I-+ 243 +++ +++ - 729 -I-++ 2187 ++-I- 4%F~tionII of pooled rabbit serum A B RabbitRGGI serum A RabbitRGGII serum A B Pre-immune sample of serum #lo97 A B +++ -I-++ -H-+ +-kk -I-++ +-k-k +++ +++ +++ +++ +++ +++ +++ +++ f4-f +++ +f+ +f+ S-f-k - ++-I- +f+ - +++ ff+ +++ +++ +++ fS.+ -H--k +++ +++ Saline A B +++ +++ B - - - +U -t++ f f - k serum of rabbit origin. The same indicator system was also agglutinated by rabbit antisera containing serum group antibodies. This gave us the possibility of comparing the serologic properties of anti-antibody on the one hand and serum group isoantibodies on the other. We had antisera against serum group antigens available which were kindly supplied by Dr. Dray. They were anti-RGGI and anti-RGGII sera.13 Only the anti-RGGII serum agglutinated the red blood cells sensitized by the anti-G serum under investigation. Obviously the latter antiserum was of RGGII serum group, which could be confirmed by double diffusion gel precipitation procedures. In table 3 an inhibition experiment is presented. Anti-antibody containing serum and the anti-RGGII serum were used at concentrations corresponding to 10 agglutinating units. Various specimens were tested for their activity in inhibition of the agglutination of sensitized blood group G erythrocytes. F-I1 of pooled rabbit serum did not inhibit anti-antibody but it strongly inhibited the anti-RGGII serum. Normal rabbit Serum of RGGI serum group did not inhibit any of them. Rabbit serum of RGGII group did not inhibit anti-antibody but it strongly inhibited the anti-RGGII serum. We also tested a pre-immune serum sample obtained from the same rabbit which, after proper immunization, served as a source of anti-G serum used in this experiment. As is to be seen, this serum sample did not inhibit the anti-antibody but it strongly inhibited the serum group isoantibodies. In table 4 an experiment is presented in which the anti-antibody-containing serum was mixed with anti-G serum. Both sera were used at various increasing dilutions and the experiment was performed as a box titration. The mixture of the two sera was left for several hours and then a suspension of blood group G erythrocytes was added. As is to be seen, the erythrocytes were strongly agglutinated; the mixed serum behaved as if it contained complete G antibodies. An experiment which was performed under identical conditions but in which 465 SPECIFICITY OF RHEUMATOID FACTOR FOR OTHER GG Table I.-AggEutination of Rabbit Blood Group G Erythrocytes by a Mixture of Anti-Antibodg and Anti-G Dilutions of anti-antibody containing serum 1 to: Dilutions of G antiserum 1 to: 10 128 +++ +++ +++ +++ +++ +++ ++ + Saline Zk 1 2 4 8 I6 32 64 80 40 20 +++ +++ +++ +++ +++ ++ +- +++ +++ +++ +++ ++ + saline 160 + + +- + Table 5.-Absorption of Anti-Antibody Containing Serum by Immune Agglutinates Agglutination of rabbit blood group G erythrocyten (sensitized by incomplete anti-G serum) at serum XU01 dilutions of 1: Serum P1201 absorbed with sheep red blood cells 2 Agglutinated by rabbit anti-sheep erythroqteserurnatadilution of 1 : 76 160 300 600 Agelutinatedbyhuman infectious mononucleosis serum at dilution of 1 : 10 Nonagglutinatsd 20 - 4 _____-- - + + +++ ++ fff -kff +++ f f - t +++ +++ 8 '' ++ 16 - ++ +f+ +++ 64 128 - + * -+ $+ -t rl: 32 - - -- +++ +++ f+ + * +++ +++ ++ ++ + anti-RGCII serum was used instead of anti-antibody gave completely negative results. The findings presented indicate that anti-antibody combined with a different antigenic determinant of the G antibody than the serum group antibody did. The latter antibody acted upon a serum group antigen RGGII which is also present in some whole rabbit sera and in the 7-globulin fraction of pooled rabbit serum. In contrast, anti-antibody acted upon the antigenic structure that is revealed on the antibody molecule during its reaction with the antigen and which resulted from the transformation of this molecule. The only way to remove anti-antibody from the serum was by absorption with an immune complex containing rabbit antibody. In the experiment presented in table 5, anti-antibodywas absorbed by sheep red blood cells agglutinated by the corresponding rabbit antiserum. However, the same sample of sheep red blood cells was incapable of removing anti-antibody if it was agglutinated by human infectious mononucleosis sera. We also obtained positive absorption experiments using other agglutinates, e.g., Proteus bacilli agglutinated by rabbit anti-Proteus serum. Similarly, immune precipitates consisting of any antigen and the corresponding antibody of rabbit origin were capable of absorbing anti-antibody. 466 IMMUNOLOGIC ASPECTS OF RHEUMATOID ARTHRITIS AND SLE Rabbit anti-antibody may be considered as a quite specific reagent capable of distinguishing between rabbit immunoglobulin that suffered denaturation in the serologic reaction and normal or immune rabbit 7-globulin as it appears free in serum or in the solution of the chemically isolated fraction. Similarly, as the affinity of human anti-antibody was restricted to the transformed human 7-globulin, the affinity of rabbit anti-antibodywas restricted to the transformed rabbit immunoglobulin. DR. HARBOE: We have found that the identical substance (anti-antibody) in human serum is 19S, and it reacts strongly with the B piece of the antibody on the cells because if one coats cells with pepsin there is no longer agglutination of this kind. REFERENCES 1. Butler, V. P., Jr., Gleich, G. J., and Vaughan, J. H.: An immunochemical basis for the cross-reactivity of rheumatoid factor with animal gamma globulin. Arth. & Rheumat. 5:104, 1962. 2. LoSpalluto, J., and Ziff, M.: Chromatographic studies of the rheumatoid factor. J. Exper. Med. 110:169, 1959. 3. Heimer, R., and Schwartz, E. R.: ISOlation of rheumatoid factors. Arth. & Rheumat. 4:153, 1961. 4. Milgrom, F., Witebsky, E., Goldstein, R., and Loza, U.: Studies on the rheumatoid and related serum factors. 11. Relation of anti-human and antirabbit gamma globulin factors in rheumatoid arthritis sera. J.A.M.A. 181:476, 1962. 5. Cohen, E., Neter, E., Mink, I., and Norcross, R. M.: Use of alligator erythrocytes for demonstrating agglutination activating factor in rheumatoid arthritis. Am. J . Clin. Path. 30:32, 1958. 6. Milgrom, F., Witebsky, E., Abeyounis, C . J., and Wilson, D. M.: Red blood cells sensitized by incomplete rabbit antibodies as a reagent for detecting the rheumatoid factor. Acta rheumat. scandinav. In press. 7. Kellner, A., and Hedal, E.: Experimental erythroblastosis fetalk in rabbits. I. Characterization of a pair of nllelic blood group factors and their specific immune isoantibodies. J. Exper. Med. 97:33, 1953. 8. Waller, M. V., and Vaughan, J. H. Z7se of anti-Rh sera for demonstrating agglutination activating factor in rheumatoid arthritis. Proc. SOC. Exper. Biol. & Med. 92:198, 1956. 9. Fudenberg, H. H., and Kunkel, H. G.: Specificity of the reaction between rheumatoid factors and gamma globulin. J. Exper. Med. 114:257, 1961. Prepara10. Gyenes, L., and Sehon, A .H.: tion and evalutation of polystyrene antigen conjugates for the isolation of antibodies. Canad. J. Biochem. & Physiol. 38: 1235, 1960. 11. Milgrom, F., Dubiski, S., and Womiczko, G.: Human sera with “antiantibody.” Vox Sang. 1: 172, 1956, 12. -: Rabbit sera with “anti-antibody.” Vox Sang. 7:545, 1962. 13. Dray, S., and Young, G.: Differences in the antigenic components of sera of individual rabbits as shown by induced isoprecipitins. J. Immunol. 81: 142, 1958.