COMPARATIVE STUDIES ON THE ANTIGENS OF THREE CYTOPLASMIC FRACTIONS FROM THE HEPATOMA AND LIVER OF RATS III.AGAR-DIFFUSION TEST WITH SPECIAL REFERENCE TO THE PURIFICATION OF ANTIGENS

It has been demonstrated(Chen et al.,1961;Ku et al.,1961)that there are anti- genic differences between hepatoma and liver of rats.In order to identify the antigens responsible for such differences,method of agar diffusion,which permits the analysis of several antigen-antibody systems in one reaction and the demonstration identical and nonidentical antigens in different tissues,was employed. Three experiments were performed as follows: The supernatant,microsomal and mitochondrial fractions,used as antigens,of hepa- toma and liver,and the rabbits' antisera of these fractions were prepared as described by Chen et al.(1961).Ouchterlony's(1958)agar-diffusion method and its modification by Abelev(1960)were used. 1.Comparisons of antigenic compositions of homologous fractions of hepatoma and liver. Fig.1 shows that the supernatant fraction of tumour both gains some antigens,as shown by b,and losses others,as shown by c,which means that the liver antigens can not be revealed in tumour.Beside these,the two tissues have a large number of common antigens.Similar results are obtained in the majorities of reactions with different anti- supernatant fraction sera(Table 1,6—9/10 cases).Chen et al.(1961)using the comple- ment fixation method reached the same conclusion;and Ku et al.(1961)by means of anaphylaxis with desensitization stated that the supernatant fraction of hepatoma gained some antigen(s). The question of whether tumour gains or losses some antigen(s)has not yet been settled(Green,1958;Weiler,1959;and Zilber,1958).This might be partly due to the fact that different authors studied with different tumours and on different antigens.The tumour either gains or losses specific antigen(s)with respect to its original tissues was proposed by Zilber(1959).Abelev and Avenirova(1960),Abelev and Zvetkov(1960) studied on the supernatant fraction of mice hepatoma and reached the same conclusion. The fact presented here is more or less in accordance with this,but we are not sure that the antigens gained or lost are specific,though they may contain specific antigens. There is no obvious antigenic differences between microsomal fractions of hepatoma and liver(Fig.2,Table 2,0—3/9 cases)and only slight difference between mitochondrial fractions of turnout and normal tissue(Fig.3,Table 3,2—3/10 cases). Zilber et al.(1959)using agar diffusion method concluded that the specific tumour antigen of mice hepatoma was present in mitochondria-microsomal fraction.Later, Abelev and Zvetkov(1960)studied on the same material and mentioned that the amount of the specific tumour antigen present in mitochondria-microsomal fraction was only one- fourth of that in supernatant.Our results also indicate that the antigenic differences be- tween particulate fractions are less than that between supernatant fraction of hepatoma and liver. 2.The isolation of the "gained" antigens of hepatoma and the liver antigens which can not be revealed in tumour. The unequivocal answer to the question of the presence of specific tumour antigens will be their isolation and identification.Anti-supernatant fraction sera were purified at 50% saturation of ammonium sulphate,the precipitate were dialysed,the partially purified antisera thus obtained were added into the heterologous antigenic solutions(su- pernatant fractions),kept in ice-box overnight,and the precipitates centrifuged off.Four or five times of precipitation were carried out with increasing amount of the partially purified antisera until the common antigens were eliminated "completely"(Fig.8)or de- creased to a great extent(Fig.4 and 6).The antigenic solutions so obtained were purified with ammonium sulphate at 50% saturation in order to remove the remaining antisera.These solutions possess either no common antigens or very little of them so as to escape the detection by.agar diffusion method,their"gained" antigen of hepatoma and the liver antigens which can not be revealed in tumour,however,are kept(Fig.5 and 7).Similar results are obtained with antisera of other rabbits. In considering the competition of antigens during immunization,(Korngold,1957) and the less sensitive of the agar diffusion than that of other immunological method,etc., the "gained" antigens so far purified may still contain the common antigens.The liver antigens,which can not be revealed in tumour,thus obtained may also be the case.Me- thod of further elimination of the common antigens which may be present in the above two antigenic solutions is proposed and discussed. 3.Comparisons of the antigens among different fractions of the same tissues. Figs.9,10 and 11 demonstrate that(1):each fraction of liver has its antigens which can not be demonstrated in other fraction of the same tissue,(2)the large number of common antigens are present in two different fractions and the quantity of common anti- gens in each fraction varies.These two features are also observed with different frac- tions of hepatoma as illustrated in Fig.12.Vogt(1960b)employing quantitative preci- pitation method,and Perlmann et al.(1959)using agar diffusion technique also made the same observation in fractions from rat liver and interpreted them as indicating that the common antigens presented in different fractions were not necessary as the result of. contamination during the preparation of antigens.(3)in cross reaction there is a precipita- tion line which is not found in homologous reaction(Fig.10). Two additional experiments were carried out as follows: 1.The relationship between the precipitation line and titer of complement fixation of anti-liver fractions sera. It can be seen(Fig.13)that there is a more or less "positive correlation" between precipitation lines(number and thickness)and titers of anti-supernatant fraction sera.It may be interpreted by the unitarian hypothesis of antibody.Fig.13 also shows that different animals respond differently to the same component of a mixture of antigens. There is,however,no definite correlation be ween precipitation lines and titers of antesera of the particulate fractions.It is suggested that the "insoluble" and corpscular antigens may not be detected in agat diffusion,but will be determined out with complement fixa- tion. 2.Comparisons of the freshly prepared factions of liver and those stored at-20℃ for 9 months. Figs.14,15 and 16 show that freshly prepared and aged fractions have quite similar patterns.However,the former give an additional weak fine(Fig.14)or have slightly thicker lines(Fig.16)than that of the latter.Aged preparations still retain nearly all of the antigens as the freshly prepared ones may be due to the fact that(1)they are well preserved in the hypertonic solution(0.88 M.sucrose)and kept at-20℃,(2)the agar diffusion method is not sensitive enough to detect the small changes of antigens,and(3) the high stability of the antigens.The high stability of specific turnout antigens of the supernatant fraction of mice hepatoma and of the microsomal fraction of rat liver were mentioned by Abelev and Avenirova(1960)and Vogt(1960a)respectively.