Tobacco Documents Online

} ~ y h DrQ~'t 11171 MICROBIOLOGICAL ASSO E INC. RR'..4 Tumor Virus Genomes as Determinants of Chemically Induced Transformation In Vitro by Aaron E. Freeman, Paul J. Price, Eugene M. Zimmerman, Gary J. Kelloff, and Robert J. Huebner

In the preceding papers we have.seen data showing that chemical induction of neoplasms in animals is correlated with activation of the C-type RNA tumor viruses as demonstrated by isolation of the infectious particles or, more frequently, by demonstration of the virus-specific gs antigens. In this presentation, I shall discuss our experiments in a parallel field of study on the individual and combined effects of chemicals and leukemia viruses as inducers of cell transformation in vitro. Combined effects of chemical carcinogens and murine leukemia viruses on transformation of rat embryo cultures:--The general protocol for our first experiment is shown in Figure 1. We prepared frozen pools of cells derived from embryos of the Fischer, Osborne-Mendel, or Wistar strains of rat. Secondary cultures prepared from the frozen pools were treated for 21 days with 0.1 mM diethylnitrosamine (DEYA) and simultaneously infected with either the Rauscher (RLV) or CF-i (CF-lLV) strains of murine leukemia virus. All combinations of chemical and virus were tested so that, in cells from three kinds of rats, we were studying the individual and combined effects of diethylnitrosamine with both a known laboratory strain and with a newly isolated wild strain of mouse leukemia virus. Although there were no clear visual differences in the early phases of the experiment, we continued to subdivide the cultures as necessary. By the sixth subpassage, the cultures had begun to change morphologically and these changes continued until the twelfth subculture (Table I). Untreated cultures or cultures treated with DEArA alone were not changed. Cultures chronically infected with the murine leukemia viruses were changed but were not transformed by the criteria we shall discuss later. In every case, however, cultures which had been treated with DENA and virus were transformed (Figure 2). The fibroblastic and contact-inhibited C°~~ ~~~~~~~~~ 002-3-711

2 control cells retained characteristics of normal cellular orientation. Cultures treated only with DENA were similar to control cultures. The cells in the virus infected cultures crowded together more closely and appeared more epithelioid, biit were not considered transformed because they retained the normal characteristics of contact inhibition and cellular orientation. In sharp contrast, cultures which had been treated with both the chemical and a mouse leukemia virus lost both their contact inhibition and their cellular orientation and by these criteria were considered to be transformed. We continued to passage cell lines derived from this experiment for over sixty subcultures. During this time the morphologies of the cultures remained constant with the control cultures continuing to retain normal characteristics and the chemical-viral treated cultures continuing to appear transformed. Regardless of the differences in cellular morphology, all of the cultures were essentially diploid until the twenty-eighth subpassage. By the forty-third subpassage, however, there was a marked shift toward heteroploidy in the transformed cultures. This shift in the chromosome number did not occur in the non-transformed cultures even up to the sixty- fifth subpassage (1). In subsequent experiments, we tested 3-methylcholanthrene (3MC) (2), benzpyrene (BzP) (3), city smog residues (GSR) (3), and cigarette smoke condensates (CSC) (4). In similar experiments in his laboratory, Dr. Rhim tested benzpyrene (5) and dimethylber_zanthracene (DMBA) (6). Meanwhile, we had been inoculating isogeneneic newborn rats subcutaneously with 1-2 x 10 control culture cells or with cells derived from cultures transformed by each chemical agent we had worked with. In early experiments the DERA transformed cultures did not induce tumors, but in later experiments tumors

were obtained with cultures transformed by 3~K'C, BzP, DMBA, smog extracts and CSC. When the results of all these experiments were reviewed (Table II), it was apparent that we were able to demonstrate transformation only jn rat embryo cultures wcich were virus infected and not in control cultures or cultures treated with either virus or chemical as individual agents. Correlation of in vitro assay with in vivo activity of chemicals:--Th.e next set of experiments was designed to determine whether the transforming effects of chemicals in vitro had any correlation with the carcinogenic activity of the same compounds in vivo. For this purpose, Drs. John and Elizabeth Weisburger of the National Cancer Institute provided us with a series of carcinogenic families which included compounds of known carcino- genic activity as well as non-carcinogenic analogues. In the new experiments (Figure 3), we used a frozen pool of high passage rat embryo celLs! These were inoculated with Rauscher leukemia virus for one week and then exposed to varying dilutions of the test chemical for an additional 7 days. One set of the original bottles was set aside to be held indefinitely, the other set of bottles was subdivided after a 2-week interval to provide two new sets of cultures, one for holding indefinitely and the other for sub- division after two weeks. This reduplicating of experimental sets was continued for five subdivisions. We designated those cultures which were not subdivided as the holding series and those which were subjected to subdivision as the vert;cal series. Foci of transformed cells often appeared in the holding series, usually in the third subpassage. Transformation in the vertical series was much more difficult to read, took longer, and was not focal (Figure 4). Obviously, transformation occurred at low frecuency (at some early point in the experiments) and the number of altered colonies was magnified by two

mechanisms. First, the holding of cultures for two weeks between subdivisions gave the non-contact inhibited transformed cells a selective advantage over normal cells and the entire population tended to shift toward transformation. Secondly, subdivision of the cultures set up multiple satzllite transformed colonies. Each compound was tested at several doses, at concentrations which did not cause visible toxic effects. In general, transformation increased as the dose increased until an optimal dose was reached. Higher doses of chemical reduced the number of transformed foci (Figure 5). This optimal dose phenomenon, which has been reported previously by Berwald and Sachs (7) and by Chen and Heidelberger (8), probably means that toxicity at the molecular level occurs even at doses that show no visible effects and that some factor needed for transformation, perhaps Ihl'A synthesis, is inhibited. We have now demonstrated four characteristics of cell transformation which can be quantitated. These are speed of transformation, number of foci, optimal dose, and ability to transplant as tumors in animals. Using these characteristics, we have studied a series of azo dyes, aromatic amines and hydrocarbons (Table III). In general, compounds which are not carcinogenic in vivo are inactive in vitro. However, some chemicals, such as 3,3'-dichlor- obenzidine, which are weakly active in vivo are strong transforming agents in vitro. Due to a lack of defense mechanisms under cell culture conditions, we expect to find transforming activity in a certain number of chemicals which are inactive or relativeZy inactive in vivo. For this reason, we are suggesting that in vitro ormation should be used as a preliminary screening of potential carcinogens. Those agents which appear to be carcinogens by this test should then be further studied in the whole animal. C.

Combined effects of chemical carcinogens and hamster leukemia virus on the transformation of hamster embrvo cells:--T_n vitro transformation of cell cultures by chemical carcinogens is not new since it has been reported many times in the Ii_erature (7-L3). In those earlier studies, viral technology had not progressed to the point where the endogenous C-type RNA flora of each cell culture could be easily identified. We wondered if carefully controlled experiments would reveal that C-type RNA viruses were generally ima?:cated in cellular transformation by chemical carcinogens. To this end,we repeated our transformation studies using hamster embryo cultures, hamster leukemia virus (HaLV) and extracts of city smog residue. We found that control cultures or cultures chronically infected with HaLV did not become transformed for at least 12 subpassages. Within 6 subpassages, however, cultures infected with HaLV were transformed by 0.007 or 0.0007 gg units/mI; HaLV free cultures were transformed by 0.007 but not by 0.0007 gg units/m2 of smog residue (Figure 6). Thus, the HaLV infected cultures were 10 times more sensitive to transformation than HaLV free cultures, and yet, hamster embryo cells which were apparently virus free were readily transformed by the chemicals (14). Syrian hamster embryo cultures were transformed by 3MC and certain fractions of cigarette smoke residue (Figure 7). These transformed cell cultures produced t=ors in newborn hamsters 43-140 days after a subcutaneous inoculation of 2 x106 ce1Zs/mi. In each case, we were unable to demonstrate HaLV antigen by the CF test in the original or transformed cell lines. However, 6 of 9 ce?2 lines derived from the tumors were positive for HaLV antigen. The virus and its infectious nature were further demonstrated by electron microscopy and by the CF antigen induction (CoFAL) test (15). As additional evidence of infectious virus, we labelled the cultures with tritiated uridine, analyzed the 24 hour supernatant fluid, and demonstrated

radioactivity at bouyant density of the virus in sucrose gradients (-1.16 g/w). HaLV was also found in cell lines derived from tumors induced in vivo by chemical carcinogens, but not in cell lines derived from SV40 induced tumors (Table IV). Thus, isolation of virus was correlated with chemical treatment and not merely with tumor induction. Murine leukemia viruses as transformation inducinQ agents.--We have repeatedly reported that RLV and other C-type M viruses are non-transforming agents. This was true under the conditions of our early experiments which empasized utilization of low passage (less than 50 population doublings) rapidly growing mass cultures of rat embryo cells. other workers, however, have reported that murine leukemia virus transformed mouse (16-17) and hamster (18-19) cultures. In addition, Dr. Rhim reported that one subline of RLV chronically infected rat embryo cells became transformed after 14 subcultures (20). emphasized that this was a single observation and that multiple sublines He chronically infected with RLV did not become transformed (21). In order to ascertain whether Rhim's observed transformation was indeed induced by RLV or represented merely a random spontaneous event, we went back to our frozen inventory of control and RLV infected cultures and plated out low cell concentrations in order to see colony types. Five (5J5) different virus free control and seven (7/9) RLV infected cell lines produced normal colonies for at least 100 population doublings; but after only 30 population doublings, a small percentage of abnormal colonies (Figure 9) were seen (2J9) separate RLV infected sublines of Fischer rat embryos. Further evidence that RLV is a transforming agent was obtained from the virus infected control of the h'_gh passage cell line used in the holding-vertical experimentsi. After 50-60 days of holding the $-I0th vertical passage, these cultures also became transformed and subsequently produced tumors in newborn rats.

F40T10TE #l In an accompanying paper ( ), Rhim describes a new antigen found in rat embryo cells t=aasformed by Polyoma virus. This antigen reacts with certain mouse sarcoma virus antisera prepared in rats, but has been shown not to be a gs antigen of murine leukemL°a or aurine sarcoma viruses. Certain, but not all, high passage rat embryo cell lines develop this new antigen spontaneously and then acquire more sensitive characteristics regarding susceptibility to transformation. One of these antigen positive cell lines (FI7Q6 P$8) has been used for all of the holding-vertical experiments. Although :I706 controls do not transform under our e mental conditions, spontaneous transformation of individual cultures has happened on occasion after 20 additional subpassages. Treatment with either MuLV or 3MC cuts the transformation time from 20 subpassages to 10 subpassages; treatment with MuLV and 3MC leads to the results presented here, i.e. transformation in 1-5 subpassages. i ~

T.r-~.B L.: HaLV Detection in CelZ Lin es Derived from Tumors ~ Source tr t_ liptake of Demonstracion CF Antio< Hamster ~ of \o. -FJSo. I Subculture j"iter a) ; Triated ~ or Farcicles Inductior Strair. iumor ~ ! ;,o. wi,ern i O+ ~::Lor { m uridiiae by Li2C~~t~n T2st ~ r ~ Tested ~ Tested I Cell ;.ines t ~:Iicrosco?y Titer oi Cells Trans- ~ ~ LSii formed by 1/2 11 3 ~ 8 ~ yes \,c} Ni 1. 0 ug/ml ,f6 c} # , Ceil Line I Cells Trans- 2/ 2 {:839} 1foraed by Lir.e 9 8 ~ yes yes (({ IC4.5 ~ l.GugJul il9 c} ~-Line 2:; 9 16 ~ yes N: ~~ 1/3 ' 3 ~ 4 ! yes N~ NIH Ce11s Trans- 2/2 r Celi. Line formed by Line i:1 18 J 8 yes ~ yes `~ {, b95} 4.Iug/raI 3;1C 2:' I8 iiI 4 f yes ~~ \T Trans lants ~ ~ N I:i iof SV40 0/26 ~ 8-22 0 i no ao 1C)C Ce1l Line Hamster (F2241) } ~ Tu,zor ~ ~ Tumor Induced 1/3 ~ 3 ~ 4 h'T \T \T I In Vivo by 1/4 11 4 ~ -,, 'T \T NT LSH DMBA ~ 26 ~ 4 ! yes \T R'T Tumor Induced f In Viyo by ~ 0/1 ~ 3 i'Ae} ~ fiT 1i INT 3iiC ' 1 29 NA Tumor Induced In Vivo by 1/4 I ~ 4 1i NIH I?uiBA Tumor Induced In Vivo by 0/1 ~ 4 \A ~ :~'T ~ ti: ~ \L 1 3i4C ~ 25 \-% ~ 1~ N7 ITumor Induced ' ` : 1 In Vivo by ~ 1/2 ~ 10 > Ib j Ni N iDYtBA ~ ffi G 27 >_ 4 = ";T ra ITumor Induced 'In Vivo by {3.:C 1/3 ; 29 >4 \~ a) Reciprocai of CF titer. 0 = < 3•4. c) Se=ected cigarette smoke condensate b) Logs of infectivity. fractions (28), d) Not tested. e) Not Applicable

Discussion Our interpretations of these results are based on the axiom that cells in culture transform spontaneously, the rate of transfor-...ation being dependent upon the g e n u s, species, strain, ar.d individual paenot:Jpe of the cell donor. Therefore, transformation induced by the agents under study represent acceleration of a natural process. the role of the virus?:--Sde bel%eve- that the C-type RiVA viruses provide the basic information for the transformation event. First of all, it has already been demonstrated that these and closely related sarcoma viruses induce leukemia, lymphoma, and sarcoma in chickens ( ) ( ), mice ( ) ( ) and cats ( ) ( Secondly, the C-type R,U virsses, on rare occasion, produce sarcoma-like trans- formation in cell cultures derived from mice ( amster ( ) and rats ( } . , .. Thirdly, the C-type particules are ubiquitous in mice ( ) and chickens () and can be activated from apparently virus free cell cultures by aging (}, irradiation (} or treatment with certain chemicals () ( ). According to the oncogene hypothesis of Huebner and Todaro (), the nascent viral genome is present in every cell, but cannot be expressed because of cell mediated repressor systems. Thus, the virus is unable to multiply either by producing more infectious particles or by producing more copies of the cell integrated viral information through uncontrolled cell reproduction (orccogene). Either of these methods of viral synthesis can be derepressed independently by the factors already listed (aging, irradiation, chemicals). Thus, spontaneous transformation results from derepression of an endogenous nascent viral oncogene. When we add exogenous infectious virus, we usually do not cause transformation because, although we have by-passed the cells repressorsfor synthesis of complete viral particles, the cell still is able to repress the oncogene. However, the addition of exogenous virus alone does accelerate the rate of transformation somewhat because multiple copies of oncogenic information are more difficult to repress than fewer copies. ~~~~~