Mutagen Treatment As a Means for Selecting Immunogenic Variants from Otherwise Poorly Immunogenic Malignant Murine Tumors1

Home , Mutagen

[CANCER RESEARCH 43, 125-132, January 1983] 0008-5472/83/0043-OOOOS02.00 Mutagen Treatment as a Means for Selecting Immunogenic Variants from Otherwise Poorly Immunogenic Malignant Murine Tumors1

Philip Frost,2 Robert S. Kerbel,3 Elaine Bauer,2 Rose Tartamella-Biondo,4 and William Cefalu

Department of Medicine, Long Beach Veterans Administration Hospital, Long Beach. California 90822 [P. F., E. B.. W. C.J: Departments of Medicine and Microbiology, University of California. Irvine. California Â¡P.F.. R. T-B.j; and the Cancer Research Laboratories. Department of Pathology, Queens University, Kingston, Ontario, Canada [R. S. K.]

ABSTRACT found to be nontumorigenic (turn ) and grew progressively only in highly immunosuppressed recipients such as 600-R X-irra- The selection of nontumorigenic (turn'), highly immunogenic diated hosts (4-6, 27-29). Three different tumors were used, variants from four different tumorigenic but poorly immuno namely, the PCC4 teratocarcinoma (4, 5), the P815 mastocy- genic murine tumors by mutagen treatment and cloning is toma (6, 27), and the Lewis lung carcinoma (29). In all cases, described. Several factors were found to determine the suc cloning of these tumors after treatment with MNNG resulted in cessful nature of this selection procedure including: the nature the selection of highly immunogenic tum~ variants at an ex of the tumor used; the nature of the mutagen; the number of traordinarily high frequency. These variants were shown to mutagen treatments; and the time at which cloning is performed share antigens with the parent tumor as well as each other; in after treatment. addition, they expressed a newly induced unique or private In some cases, e.g., the TA3 adenocarcinoma or the BALB/ antigen. We have recently discussed the considerable impor e SS1 spontaneous mammary adenoacanthoma, a single treat tance and implications of these results in relation to recognition ment with ethyl methanesulfonate or A/-methyl-A/'-nitro-/v-nitro- of tumor antigens by the immune system as well as for immu- soguanidine led to a very high frequency of turn" clones, notherapy of poorly immunogenic tumors, such as those of whereas in others, e.g., the MDAY-D2 tumor line, no stable turn" clones were obtained. The immunogenic clones selected spontaneous origin (20). We herein describe our own experience with additional mu were always immunologically cross-reactive with the parent rine tumors (including one nonimmunogenic spontaneous tumor from which they were derived and were found to protect BALB/c carcinoma) and a different mutagen, namely, EMS. the murine host against challenge with the parent tumor in vivo. Our findings confirm some of those of Boon and his colleagues Thus, the cloned immunogenic variants share an antigen with but also point toward additional complexities in this model the parent tumor. Additional evidence, however, suggested relating to the nature of the mutagen used, the tumor cell type that each clone also expresses a new private or unique antigen. used, and the timing of cloning after mutagen treatment. The frequency of immunogenic variant selection ranged from a low of 6% to a high of 95%. In some cases in which the MATERIALS AND METHODS frequencies were very high, cloning was not required to reveal the turn" phenotype. Finally, we also noted that selection for Animals. DBA/2 and A strain mice, 5 to 6 weeks old, were pur drug resistance, e.g., resistance to 6-thioguanine, after muta- chased from The Jackson Laboratory, Bar Harbor, Maine. BALB/c mice were originally obtained from the NIH, Bethesda, Md., and are genesis could have an enhancing effect on the generation of highly immunogenic turn" clones. The results show that the currently bred in our colony at the Long Beach Veterans Administration Hospital. BALB/c-nu/nu mouse breeding pairs were originally ob immunogenicity of poorly immunogenic tumors, including those tained from the NIH and are currently maintained in our own nude of spontaneous origin, can be dramatically enhanced by ap mouse colony established at the Long Beach Veterans Administration propriate mutagen treatments but that there is considerable Hospital. In all experiments aimed at assessing the growth potential of variation in the ease with which highly immunogenic variants tutrr variants, both nu/nu and nu/+ heterozygous littermates were can be obtained. challenged with tumor. Tumors. The origin and characteristics of the MDAY-D2 tumor have been described in detail elsewhere (11, 19, 21). The P815 mastocy- INTRODUCTION toma was obtained from Dr. R. Herberman (NIH). The TA3 mammary adenocarcinoma was obtained through Dr. Joseph G. Mayo, NIH, and In a series of papers, Boon et al. (4-6, 27-29) reported that Dr. Arthur Bogden, Mason Research Institute Tumor Bank, Worcester, treatment of murine tumor cells with the mutagen MNNG5 Mass. The SS1 adenoacanthoma is a spontaneous mammary tumor results in the generation of highly immunogenic variants. The derived from a BALB/c female in the colony maintained at the Clinical increased immunogenicity was so marked that the clones were Research Center, Harrow, England. SS1 grows in BALB/c mice de rived from the NIH colony. It does not grow in histoincompatible strains. 1Supported by the Medical Research Council of Canada. Although this tumor was isolated 8 years ago, it was stored in liquid 3 Recipient of a grant from the Veterans Administration and Research Grant nitrogen after its third in vivo passage. This stock serves as the current CA 28060 from the USPHS. To whom requests for reprints should be addressed source of the SS1 tumor used in these experiments. Tumor cells were at Long Beach Veterans Administration Hospital, 9501 East 7th Street N-S, Long Beach, CA 90082. maintained in culture or by weekly i.p. passage. All cell lines were 3 Research Associate of the National Cancer Institute of Canada. routinely tested to assure that they are free of Mycoplasma by assess ' Recipient of a grant from the National Cancer Institute of Canada. ing for growth in pleuropneumonia-like organism agar plates and 5 The abbreviations used are: MNNG, W-methyl-W-nitro-N-nitrosoguanidine; modified Hayflick's broth medium under aerobic and anaerobic condi turn', nontumorigenic in normal immunologically intact syngeneic hosts; EMS, ethyl methanesulfonate; RPMI, Roswell Park Memorial Institute; CMC, cell-me tions. diated cytotoxicity; turn*, tumorigenic; met , nonmetastatic; met*, metastatic. Treatment of Cells with Mutagens and Selection for Drug Resist Received May 6, 1982; accepted October 12, 1982. ance. Treatment of tumor cells with MNNG was performed according

JANUARY 1983 125

Protocol I ParentTumor -TreatedwithEMSorMNNG-

Selectedfor UnclonedPopulation OuabainResistance InjectedS.C.

TreatedwithEMS NoGrowth InVivo InVivo Chart 1. Schematic outline of treatment Select'!edfor6-Thioguanine with EMS of cells selected for resistance to Resistance, ouabain and 6-thioguanine. Only clones which SurvivorsareCloned Reinjectedx3* failed to grow after 3 sequential injections into different animals were considered to be turn". andRandomClonesare UnetonedPopulation Cloned Injected NoGrowth Grows Injected InVivo InVivo Individual NoGrowth Grows Clones In Vivo h Vivo Injected Grows NoGrowth 3;rowth x I X InVivo InVivo Reinjectediti Grows | Not Grows 1 mVivo Reflected In\flvo InVivo NoGrowth NoGrowth\ x3Grows mvivo InVivo NoGrowth InVivo InVivoReinjected

-AssessedforAbilitytoGenerateCMCResponseandforCross ReactivitywithParentTumorInVivoandInVitro

to the methods of Boon and Kellerman (4). Cells (3 x 106in 5 ml) were Fc Receptor Assay. Fc receptors were assessed using the method exposed to MNNG at 3 fig/ml for 60 min. The cells were washed and described by Kerbel (18). The antiserum used was a monoclonal lgG2B placed in culture. Treatment with EMS was according to the method of anti-sheep RBC antibody derived from a hybrid myeloma designated Gillin et al. (13) as modified by Wasmuth.6Five ml of tumor cells at 6 Sp 2/HL by Kohler and Milstein (22). This antiserum was purchased x 105cells/ml were placed in small sterile Retri dishes and allowed to from Sera Labs (Crawley Downs, Sussex, England). equilibrate to a 5% CO2atmosphere in RPMI without fetal calf serum. Radiolabeling of Tumor Cells. The radiolabeling of tumor cells with After 1 hr, 2.5 /il (0.5 fil/ml) of EMS were added, and the cells were ['"ln]indium oxine is based on methods that we have described previ allowed to incubate for 2 hr, after which they were centrifuged and ously (12). Briefly, cells were labeled with 30 fiCi [11'ln]indium oxine at incubated in RPMI with 10% fetal calf serum. Under these conditions, a concentration of 107 cells/0.5 ml in RPMI Medium 1640 with 10% cell viability is reduced to approximately 30% after 24 hr. This is fetal calf serum and 20 mw N-2-hydroxyethylpiperizine-W-2-ethane- consistent with the findings of others (4).6 For drug selection, the cells sulfonic acid buffer for 15 min at 37Â°and washed 3 times in 15 ml of were exposed to either 6-thioguanine (8.5 fig/ml) or 1 rriMouabain 4 medium. to 5 days after EMS treatment. The drug dose was increased each time CMC Assay. CMC was assessed by a method described previously confluent cell growth occurred until resistance to 6-thioguanine (15 (19, 30). In initial experiments, individual spleens from 5 mice were jug/ml) or 3 rriMouabain was achieved. Cells selected for resistance to tested for their CMC activity. The results demonstrated almost identical both drugs were treated for a second time with EMS and carried reactivity. Subsequent experiments have therefore used pools of 3 to through the same procedure. A schematic outline of these procedures 5 immune spleens. is shown in Chart 1. The CMC assay was performed in round-bottomed microplates (Fal Assessment of MÃ©tastases.TheMDAY-D2 tumor metastasizes in con) by mixing effector and target cells at desired ratios in triplicate in a widespread and aggressive manner (11, 21). Clearly defined meta- a final volume of 0.2 ml RPMI Medium 1640 plus 10% fetal calf serum. static foci are easily visualized in the liver, spleen, kidney, and lung. All Immediately prior to incubation at 37Â°for 24 hr in 5% CO2,the plates animals bearing tumors were autopsied and evaluated visually for the were centrifuged at 300 rpm for 3 min. Assays were terminated by presence of mÃ©tastases. centrifugaron of the plates at 1500 rpm for 10 min at roomtemperature. Antisera. Anti-H-2 alloantisera (anti-H2* and anti-H2") were obtained One hundred-/il aliquots of supernatant were removed from each mi- through Dr. John Ray, Transplantation and Immunology Branch, Na crowell for counting in a Beckman Model 300 gamma counter. By tional Institute of Allergy and Infectious Disease, Bethesda, Md. Mon counting all experimental and control tubes at the same time and by oclonal anti-Ly6.2 antiserum was a generous gift of Dr. U. Hammerling incorporating internal standards for each experimental group (not of the Sloan-Kettering Cancer Institute, New York, N. Y. exceeding 50 samples),we automatically correct for radioactive decay. Serological Assay. The cytotoxic effect of the alloantisera was Calculation of Specific Isotope Release. The percentage of release assessed with the use of a rabbit complement-dependent cytotoxicity (SR) or spontaneous release of [1"ln]indium oxine was calculated by assay as described previously (16). The rabbit complement was ob subtracting isotope release in the presence of normal lymphocytes tained from a New Zealand White rabbit from our own animal colony. (RNL) from that released in the presence of sensitized lymphocytes The serum was pretested for complement activity and stored at -70Â°. (RSL) and dividing by 0.8 x total counts (TC). The 0.8 constant Prior to use, all complement was absorbed with the appropriate target represents the maximum release of isotope from ['"Injindium oxine- cell and used at a final dilution of 1:10. Calculation of the percentage labeled cells (12). of killing by antisera and complement was as described previously (16). RSL - RNL %SR x 100 sJ. Wasmuth, personal communication. 0.8X TC

126 CANCER RESEARCH VOL. 43

% of spontaneous release Table 1 RNL - cpm present in supernatant at initiation of the assay Effect of mutagen treatments) on the selection of immunogenic variants of x 100 murine tumors TC Tumor cells were treated with a single mutagen exposure with either EMS(E) or MNNG(N). Thus, e.g., MDAY-D2 cells treated with EMS were designated as Cold Target Inhibition of CMC. The method originally described by MDE, whereas those treated with MNNG were called MDN. Repeated treatments de Landazuri and Herberman (8) was used. Each inhibition of CMC of MDE or MDN cells with either EMS or MNNG resulted in tumor cell lines called assay was first calibrated by using various inhibitortarget ratios. Op MDEE and MDEN or MDNN and MDNE. All tumor lines selected for drug resistance were treated with EMS prior to their exposure to the individual drugs. timal inhibition ratios were then used. Thus, PETEO is P815 treated once with EMS and selected for 6-thioguanine (T) Cloning. Limiting dilution cloning was performed in all these exper resistance, and treated a second time with EMS and selected for ouabain (O) iments as described previously (17). resistance. PETE was not selected for ouabain resistance after the second EMS treatment. Similarly. MDOT1-4 cells were MDAY-D2 cells treated with EMS, selected for ouabain resistance, treated with EMS, and selected for 6-thioguanine RESULTS resistance. of Our original efforts were aimed at expanding and confirming clones injectedturn*22 met22whole the observations of Boon and his colleagues. In addition, we TumorTA3-ESS1-ESS1-NP815PEPEEPENPETEOPETEEarlyLateMDAY-D2MDEMDEEMDENMDNMDNEMDNNMDOT1MDOT2MDOT3MDOT4No. chose to use EMS as well as MNNG in our mutagen treatment 1Not cloned;Not immunogenicwholepopulation protocols. This was done to determine if the previously reported cloned;16 immunogenic03a1a011e3433a1a1a2a2a6'103population findings were restricted to the use of MNNG. We have to date 1610 studied 4 different murine tumors, the A strain TA3 mammary 710 912 70 12b15"1 adenocarcinoma, the spontaneous BALB/c SS1 mammary 1215 adenoacanthoma, the P815 mastocytoma, and the MDAY-D2 437 undifferentiated metastatic tumor, both of DBA/2 mice. 318 3719 1424 Mutagen Treatment of A Strain TA3 Mammary Adenocarci 2154 5"42 11"26 noma Cells 5347 4629 2Ã¬"-a'e16 2727 13b16 A single treatment of TA3 cells with EMS at 0.5 fil/ml resulted 2726 11b17 in the emergence of a cell population which failed to grow in 2418 7"12 syngeneic A strain hosts. This initial observation was unex 6Whole12turn21met* cloned15 population Â¡mmunogenic; not pected since we felt that the parent population, even after 935 6926 treatment with EMS, would contain at least some tumorigenic 2617 9910 79 clones which should have grown in vivo. To our surprise, the 106979 a Unstable; reverted to turn* phenotype. injection of this uncloned population resulted in tumors which Unstable; reverted to met* phenotype. initially grew but eventually regressed. This implied that the c Stability not assessed. d H. . great majority of clones of the TA3 tumor treated once with " ThreePlease clones see text. have remained stable for turn*, mer phenotype. EMS (TA3-E) were immunogenic. This is in marked contrast to Threeree clones have remained stable .uiuic foriui turnium K/ncnwiyKÂ»c.phenotype. the control animals given injections of parent TA3 tumor cells. 9 Stability of clones with turn*, met~ phenotype not assessed All such animals developed tumors and died within 28 days. We subsequently cloned the mutagen-treated TA3-E cells and Table 2 found that 21 of 22 clones failed to establish tumors in the A Induction of CMC responses by EMS- or MNNG-treated SS 1 tumor cells strain host (Table 1). This protocol was repeated and identical Five BALB/c mice were given injections of 5 X 10s SS1-E, SS1-N. or SS1 cells s.c. on Day 0. On Day 14, the spleen cells were harvested, pooled, and results were obtained. Of additional interest was the observa placed in culture with mitomycin C-treated SS1, SS1-E, or SS1-N cells. Five days tion that 15 animals which had rejected the TA3-E tumor was later, the spleen cells were used as effectors (at a 50:1 effectortarget ratio) subsequently found to be resistant to challenge with the parent against SS1, SS1-E, and SS1-N targets in an 18-hr release assay. SS1-E and SS1-N refer to cells treated once with either EMS (E) or MNNG (N) (one of 3 (tumorigenic) TA3 tumor cell population. This finding correlated experiments). with a demonstrable CMC response detectable in TA3-E-challenged animals.7 In vivo2ndtion 1st In vitro CMCSS1410010010019and % of Â«on1EEENNN" Mutagen Treatment of SS1 Spontaneous Adenoacanthoma Cells of BALB/c Mice

Treatment of SS1 cells with either EMS or MNNG (SS1-E Â±0.667 Â±2.032 3.1Â± and SS1-N, respectively) resulted in tumor cell progeny which Â±1.218 Â±0.943 0.4Â± failed to grow in syngeneic BALB/c hosts (Table 1). Once Â±1.4SS1-ENT"10010010040Â±1.3SS1-NNT100100100302427Â±1.3 again, this occurred prior to cloning. Control animals given NT, nottested6 injections of 104 or 10s SS1 cells all succumbed to their tumors Mean Â±S.D.1EN1NE1Target within 35 and 30 days, respectively. The uncloned populations that generated by SS1-N cells, no matter which cell line was of mutagen-treated cells generate a cytotoxic response de used for the secondary in vitro stimulation, or as a target in the monstrable 2 weeks after tumor challenge (Table 2). It should be noted that the CMC response generated when SS1-E was CMC assay. The decreased cytotoxic potential generated by SS1-N cells does not appear to be relevant in vivo, since SS1- used for the initial in vivo priming was consistently greater than N cells are not tumorigenic in the BALB/c host. Such a dis 7 P. Frost and W. Cefalu, unpublished observations. crepancy between in w'fro-measured CMC responses and in

JANUARY 1983 127

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1983 American Association for Cancer Research. P. Frasi ef al. vivo effectiveness of the same response has been reported Table 3 previously (32). Furthermore, all 7 animals which rejected SS1 - Cross-reactive cytotoxic responses of PETE and P8I5 E cells were resistant to subsequent challenge with 105 SS1 Five DBA/2 mice were challenged with 5 X 10s PETE or P815 tumor cells s.c. on Day 0. Fourteen days later, the spleens were harvested, pooled, and tumor cells. This demonstrates that SS1-E and SS1 share cultured with either mitomycin C-treated PETE or mitomycin C-treated P815 common antigens. We have not as yet evaluated the effective tumor cells. Five days later, the restimulated spleen cells were used as effector ness of SS1-N immunization in the protection against SS1 cells (at a 50:1 effectortarget ratio) in a cytotoxic assay using either P815 or challenge. Finally, the injection of SS1-E and SS1-N cells into PETE as a target (one of 4 experiments). BALB/c-nu/rtu mice results in growth of the tumor in a manner In vivo 1st im vitro 2nd im munizationP815P815P815P815PETEPETEPETEInTargetP815PETEPETEP815PETEP815PETEP815P815PETEPETEPETEPETEP815%munization ofcytotoxicity0017 indistinguishable from that generated by the parent SS1 tumor.

Mutagen Treatment of the P815 Mastocytoma Â±10100053

Our work with the P815 mastocytoma provided the first evidence of the complexity of this model. Since Boon and his Â±4.6a1 colleagues had already extensively evaluated and documented Mean Â±S.D. the use of MNNG for selection of immunogenic variants from the P815 mastocytoma (6, 27), we chose to limit our experi animals. When cloned, only 3 of 17 clones expressed the turn" ments to the use of EMS. A single treatment of P815 cells with phenotype. While these 3 clones have remained stable for 2 EMS resulted in the generation of a parent population (PE) months, this finding points toward the necessity for repeated which grew in a manner indistinguishable from that of the assessments of the stability of mutagen-treated and selected original P815 cells (Table 1). Clones of PE were subsequently turn" clones. injected, and of 10 clones 7 behaved identically to the parent tumor, whereas 3 appeared to be immunogenic, i.e., did not Mutagen Treatment of the Undifferentiated, Highly Meta- grow. However, subsequent reinjection of these 3 apparently static MDAY-D2 Tumor tum~ variants demonstrated that they were unstable, and they Our most extensive investigations to date have involved the grew in a pattern identical to that of the parent tumor. A second MDAY-D2 tumor. Findings obtained with this tumor have re treatment of PE cells with EMS (PEE) did not enhance our vealed the complexity of this system and have confirmed that ability to select tum~, highly immunogenic variants from this the selection of highly immunogenic turn" variants is dependent tumor. Further treatment of PE cells with MNNG (PEN) was on the mutagen used and on whether selection for drug resist also unsuccessful in enhancing for the selection of immuno ance is also performed. genic variants. Thus, in our hands, EMS treatment of P815 cells was ineffective in generating stable immunogenic turrr Cloning MDA Y-D2 Prior to Mutagen Treatment variants. The MDAY-D2 tumor is an anaplastic, highly metastatic For reasons unrelated to these experiments, we did, how DBA/2 tumor. The precise cellular origin of this tumor is ever, subject P815 cells to a single EMS treatment and sub sequently selected these cells for 6-thioguanine resistance. unknown, although some evidence indicates that It may have derived from the lymphoreticular series (11, 19, 21). The The cells were then treated with EMS once again and split into injection of as few as 10 MDAY-D2 cells results in growth and 2 flasks, only one of which was selected for ouabain resistance. extensive metastasis of tumor cells, resulting in the death of These 2 cell lines, PETE and PETEO, were both found to the host. Cloning by itself has never led to the selection of predominantly express the turrr phenotype when reinjected stable immunogenic variants from the parent tumor. The most into DBA/2 mice (Table 1). Our experiments with these tumors recent experiment of this kind used 24 clones of MDAY-D2 have concentrated primarily on the PETE variant cell popula which were injected s.c. into individual DBA/2 hosts (Table 1). tion. While all clones derived from the parent P815 tumor were turn*, 34 of 37 PETE clones expressed the turn" phenotype Of the 24 clones injected, 21 behaved in a manner character istic of the MDAY-D2 tumor. Three of the clones did not grow (Table 1). These cells generate a potent cytotoxic response in the DBA/2 host (Table 3) and furthermore protect the synge- after the initial injection but, upon subsequent reinjection of the same clones (which had been maintained in culture), growth neic host against challenge with the parent P815 tumor. How ever, the cytotoxic response generated by PETE-primed spleen and mÃ©tastasesin a pattern identical to that of the parent tumor were observed. This experience is similar to the previously cells is best directed to PETE rather than to P815 targets. reported finding that cloning of MDAY-D2 resulted in nonmet- Furthermore, spleen cells from animals primed with P815 fail astatic clones which proved to be highly unstable (21). to generate any significant cytotoxic response whether they are secondarily stimulated with P815 or PETE. The demonstra Use of EMS ble differences in CMC response between these tumors had little relevance to the in vivo protective effect of PETE cells but Treatment of MDAY-D2 tumor cells with EMS (0.5 /Â¿I/ml) could indicate that PETE cells, while sharing an antigen with resulted in 60 to 70% cell death at 24 hr. The remaining tumor P815, also express a private antigen best recognized by PETE- cells replicated at a rate comparable to untreated cells. Treat primed spleen cells. ment of MDAY-D2 cells with higher doses of EMS (0.7 to 1.8 In an effort to assess the stability of the PETE variant, our jul/ml) resulted in their failure to survive in culture. early experiments with PETE were repeated 3 months later Single EMS Treatment. The in vivo injection of the "whole" with PETE cells that had been maintained in vitro. The injection population of such cells (called MDE) resulted in a tumor of the uncloned population resulted in tumor growth in 4 of 6 indistinguishable from the parent MDAY-D2 tumor (Table 1). A

128 CANCER RESEARCH VOL. 43

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1983 American Association for Cancer Research. Selection of Immunogenic Tumor Variants by Mutagen Treatment total of 54 clones of MDE were independently injected into Table 4 syngeneic hosts. Of these, 53 were tum+ and 11 were non- Cross-reactive CMC responses between selected MDAY-D2 variants metastatic (met"). The single turn" clone proved not to be Five DBA/2 mice were given s.c. injections of 5 x 105 cells from each of MDNN clones B. J, or F or from MDW1 cells. Fourteen days later, the spleens stable, for upon reinjection into secondary DBA/2 hosts on 2 were removed, pooled, and restimulated in vitro with mitomycin C-treated cells independent occasions it behaved in a manner identical to that identical to the initial priming clone. Five days later, the spleen cells were of the turn* clones. The stability of the tum+, met" clones was harvested and were used as effectors against 5 different targets. The MDW1 also assessed. None of these clones retained their met" phe- cells were selected from an MDAY-D2 parent tumor different from that from which the MDNN clones were obtained [MDAY-D20) versus MDAY-D2Ã•3)) (one notype and they all metastasized upon reinjection. of 3 experiments). Multiple EMS Treatment. Two sequential treatments of targetsB93 of cytotoxicity of different MDAY-D2 with EMS (MDEE cells) did not enhance the fre Cells used in quency of selection of turn" variants. Of 47 MDEE clones immunizationBF Â±3.4a injected, 46 were turn*, 1 was tum~, 26 were met", and 21 Â±2.9 Â±2.8 Â±3.3 Â±0.4 were met~ (Table 1). The turn" clone proved to be unstable 46 Â±2.3 100 76 Â±4.6 50 Â±1.9 1 Â±0 J 2 Â±0.1 20 Â±1.0 100 17 Â±0.1 35 Â±1.8 and reverted to the turn*, met* phenotype. Of the 21 met" MDW1% 33 Â±1.6F70 21 Â±0.4J62 23 Â±1.9MDAY-D26392 Â±2.6MDW1696 Â±4.4 clones, 3 have retained the turn*, met" phenotype to date. Mean Â±S.D.

These clones could therefore represent stable nonmetastatic Table 5 variants of MDAY-D2. Cross-inhibition of CMC responses between selected immunogenic MDA Y-D2 variants Use of MNNG Five DBA/2 mice were given s.c. injections of 5 x 10s cells from MDNN A single treatment of MDAY-D2 tumor cells with MNNG (3 clones B, F, or J. Fourteen days later, the spleens were harvested, pooled, and /ig/ml) for 60 min (MDN cells) did not generate any turn" restimulated in vitro with mitomycin C-treated cells identical to the initial priming clone. Five days later, the spleen cells were harvested and were used as effectors variants. Of the 27 MDN clones injected, all were phenotypi- against their respective clonal target. In addition, unlabeled MDNN clone B, J, or cally tum+ (Table 1). While 11 of these clones were met", they F; MDAY-D2; or MDW1 cells were added to the effectortarget mixture at an inhibitoreffector ratio of 12.5:1 (one of 3 experiments). proved to be unstable upon reinjection. In contrast, a second Cells used in im of inhibi treatment of MDN cells with MNNG (MDNN cells) resulted in munizationMDW1MDW1MDW1BBBBBBFFFFFFJJJJJJInhibitorMDW1MDNN-BMDNN-FMDNN-JMDAY-D2BFJMDW1MDAY-D2FBJMDW1MDAY-D2JBFMDW1MDAY-D2%ofcytotoxicity100 tion96681249029492940854562135392923259528 the selection of 3 turn", met" clones from the 18 clones Â±0a4 injected. The initial number of selected turn" clones was 6, but Â±0.294 3 of these were unstable and reverted to the turn"1"phenotype Â±4.892 Â±3.488 after 3 sequential injections in vivo in different groups of DBA/ Â±2.696 2 mice. However, MDNN clones B, F, and J have remained Â±2.482 turn" and met",/.e., highly immunogenic after being maintained Â±4.43 in culture for 3 months. Eight repeated s.c. and/or i.p. injec Â±0.458 tions of 5 x 105 cells from these clones into DBA/2 hosts Â±4.142 Â±3.258 failed to produce tumors in animals observed for periods of up Â±2.249 to 60 days. However, all 3 clones grew (but metastasized Â±1.610015 poorly) in athymic BALB/c-nu/nu mice or 600-fl-irradiated DBA/2 mice. In contrast with this finding is the observation Â±0.655 over many years that 1 x 103 parent MDAY-D2 tumor cells Â±1.838 Â±2.487 grow in 100% of DBA/2 mice. Â±3.647 All 3 MDNN clones share surface characteristics and a Â±1.21008 karyotype identical to those of the original MDAY-D2 host. They are all H-2d positive, express the Ly6.2 surface marker, Â±0.868 and have Fc receptors. Karyotyping reveals a mean chromo Â±2.341 some number of 38 Â± 2 (S.D.) with 2 metacentric chromo Â±1.195 Â±4.672 somes characteristic of the MDAY-D2 tumor (11 ). These clones Â±3.4% appear, therefore, to be true variants of the MDAY-D2 tumor. That these clones are truly immunogenic is demonstrated in directed toward clone B targets more effectively than did clone Table 4. Individual MDNN clones were used to immunize DBA/ F or J cells (Table 5). Similar results are observed with spleen 2 hosts. In addition, a known immunogenic variant of MDAY- cells directed towards clone J or F. This implies that, while D2 (MDW1) was used to determine if it shared an antigen(s) each clone has antigens in common with each other, they also with MDNN-B, -J, or -F. Spleen cells were restimulated in vitro express private antigens. with the same tumor used in the initial injection. Five days later, 2. Cytotoxic spleen cells directed toward the MDNN clones the spleen cells were harvested and used as effectors against will also kill MDAY-D2, but with varying effectiveness. Thus, 5 targets: MDNN-B, MDNN-F, MDNN-J, MDW1, and MDAY- while T-cells sensitized to clones B and F, respectively, kill 63 D2. The MDNN clones were also used as inhibitors. The results and 50% of MDAY-D2 targets, clone J-primed spleen cells kill revealed the following. only 17%. All MDNN clone-primed spleen cells are inhibited to 1. All MDNN clones generate a CMC response which is best some degree by the parent MDAY-D2 tumor cells. Clone J is directed toward themselves. Thus, T-cells sensitized to clone inhibited least, as would be predicted by the relative inability of B kill 93% of clone B targets but 70 and 62% of clones F and clone J-sensitized spleen cells to kill MDAY-D2 cells. These J, respectively (Table 5). This is further confirmed in the inhi results imply once again that the MDNN clones share some antigens with MDAY-D2 but that each expresses a "new" bition assays where clone B cells inhibit a CMC response

JANUARY 1983 129

antigen(s) not expressed by the parent tumor. Table 7 3. The MDW1 immunogenic variant generates a cytotoxic Cross-reactive CMC responses between mutagen-treated and drug-selected MDAY-D2 variants response which is weakly reactive with the MDNN clones. In Five DBA/2 mice were given injections of 5 x 10s MDOT-1, -2. -3, or -4 tumor addition, MDW1 cells are poor inhibitors of the response gen cells. Fourteen days later, the spleens were removed, pooled, and restimulated erated against the MDNN clones, nor can the MDNN clones in vitro with mitomycin-treated cells identical to the priming tumor. Five days inhibit a CMC response directed toward MDW1 at a 12:1 later, the cells were harvested and were used as effectors against various targets (one of 3 experiments). inhibitortarget ratio. This inhibition improves to 60% at a 50:1 Cells used in immuniza ratio (data not shown). No such improvement occurs when the tionMDOT1MDOT1MDOT1MDOT1MDOT2MDOT2MDOT2MDOT3MDOT3MDOT3MDOT4ofcytotoxicity95 MDNN clones are used at a 50:1 ratio. Â±3.1a43 Â±1.436 Effect of Combined EMS and MNNG Treatments Â±1.234 (current)6MDOT2MDOT1M Â±0.922 Combined treatments of MDAY-D2 tumor cells with EMS and Â±0.436 Â±2.632 MNNG failed to produce immunogenic variants. The MDEN and DOT3MDOT3MDOT1MDOT2MDOT4 Â±1.830 MDNE cell lines have been injected on 2 occasions, and no Â±0.615 stable immunogenic clones were isolated from these lines Â±0.431 Â±1.910090 (Table 1). (early)MDOT4MDOT4 (early)MDOT1MDOT4 In Vivo Effectiveness of Immunization with the turn' Â±3.834 MDNN (current)TargetMDOT1MDOT2MDOT3MDOT4(current)% Â±2.4 Clones ' Mean Â±S.D. 0 MDOT-4 (early), cells used within 4 weeks of selection; MDOT-4 (current), Groups of DBA/2 mice immunized with MDNN-J, -F, or -B cells used after a minimum of 10 weeks in culture. are totally resistant to challenge with 2 x 103 MDAY-D2 paren tal cells but not resistant to challenge with 2 x 103 MDW-23a, antigen loss variant cells, or 5 x 10" P815 mastocytoma cells in growth of a tumor which is indistinguishable from the parent MDAY-D2 tumor. Uncloned MDOT-4 tumor cells behave errat (Table 6). This confirms the specificity of these responses. It ically after injection (see below). However, repeated cloning of further demonstrates that in vitro CMC activity of low magni MDOT-2 and MDOT-3 cells has resulted in the isolation of a tude, e.g., MDNN-J versus MDAY-D2, may still be highly effec number of stable immunogenic variants. For example, MDOT- tive in vivo (32). 3 has been cloned on 3 individual occasions. A total of 35 individual clones were injected, 9 of which were found to be Effect of Drug Selection after Mutagen Treatment on the stable immunogenic turn" variants. MDOT-2 was cloned on Selection of Immunogenic Variants one occasion, and of 15 clones 6 were found to be stable turn" The original immunogenic variants obtained from the MDAY- variants. Table 7 lists the cross-reactive cytotoxic potential of D2 tumor were derived from EMS-treated parent tumor cells uncloned populations of the mutagen-treated and drug-se lected MDAY-D2 variants. The turn" clone MDOT-1 generates which were selected for either wheat germ agglutinin resistance (MDW1, MDW3) (17) or resistance to ouabain and thioguanine a response approaching 100% killing against MDOT-1, but less effective killing against the mixed population (turn" and (MDOT-1) (Table 7) (10). The MDW1 and MDW3 wheat germ- resistant variants of MDAY-D2 have remained stable immuno tum+) MDOT-2 and MDOT-3 tumors. MDOT-2 and MDOT-3 genic variants after 4 years in culture. MDOT-1 has remained also generate CMC responses, but these are moderate when stable for 1.5 years. More recently, we have selected additional compared to those of MDOT-1. variants of MDAY-D2 for both ouabain and 6-thioguanine re The situation with MDOT-4 is as follows. Initial injections of sistance. These lines, called MDOT-2, -3, and -4, were each the parent population (5 x 105 cells s.c. per animal) resulted treated initially with EMS and selected for ouabain resistance. in no growth. However, reinjection of MDOT-4 after 4 weeks in The ouabain-resistant cells were then retreated with EMS and culture resulted in growth of the tumors of 5 of 10 mice. We selected for thioguanine resistance. All these lines are capable can only surmise that the parent MDOT-4 population was of growing in thioguanine (15 /ig/ml) and 3 mM ouabain. The initially dominated by highly immunogenic clones. The clonal injection of whole populations of MDOT-2 and MDOT-3 results make-up of the parent tumor changed in vitro, resulting in an altered clonal dominance, with turn" clones becoming less Table 6 prominent. Unfortunately, the early population was not cloned; In vivo protection against MOA Y-D2 by prior immunization with MDNN B. J. or F but of 17 clones from the current stock of MDOT-4 (6 weeks Groups of DBA/2 mice were immunized by a s.c. injection of 5 x 105 MDNN later), 3 have to date remained stable tum~ variants. CMC data B, J, or F cells 2 weeks prior to challenge with either 2 x 103 MDW4-23a antigen loss variant cells, 1 x 105 cells from the unrelated P815 syngeneic mastocytoma, with MDOT-4 seem to support the changing clonal dominance or 2 x 103 MDAY-D2 tumor cells. The animals were then observed for survival of this variant with time. The initial MDOT-4 cells generated a (one of 2 experiments). strong CMC response (100% killing at a 50:1 effectortarget No. surviving/no, injected after tumor challenge ratio). The current MDOT-4 stock generates only 34% killing with under the same conditions. The latter level of responsiveness Immunization MDAY-D2 MDW4-23a P815 is consistent with that observed with the "mixed-clone" MDOT- NoneNoneNoneMDNN-BMDNN-JMDNN-F0/1016/1615/1515/150/100/120/120/120/100/60/60/62 and MDOT-3 cells. At the other extreme is MDOT-1 which has not been cloned but remains turn", generates a strong CMC response (after 1.5 years in culture), and continues to protect in vivo against a challenge of MDAY-D2 cells. It was worth noting that MDOT-2, -3, and -4 all retain a karyotype

130 CANCER RESEARCH VOL. 43

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1983 American Association for Cancer Research. Selection of Immunogenic Tumor Variants by Mutagen Treatment virtually identical to that of MDAY-D2. MDOT-1, however, has colleagues in showing that, while these clones share antigens consistently karyotyped at 60 to 65 chromosomes.8 with each other and with the parent MDAY-D2 tumor, they also express private antigens. In contrast, the MDNN clones appear DISCUSSION not to share any antigens with MDW1, a turn" clone selected In confirmation of the results of Boon ef al. (4-6, 27-29), it over 4 years ago from a different population of MDAY-D2 cells. is apparent from these experiments that the treatment of tumor The fact that both MDW1 and MDNN clones are cross-reactive cells with mutagens enhances the selection of stable immuno- with MDAY-D2 implies that the parent tumor expresses more genic variants which are immunologically cross-reactive with than one antigen. This question will be best answered using the parent tumor from which they were derived. However, our recently isolated long-term cytotoxic T-cell lines which are several variables including the nature of the tumor, the mutagen directed to MDW1 and the MDNN clones. used, the number of treatments, and whether the treated cells Another issue which must be addressed is the relative fre quency of turn" clones selected. Boon ef al. have reported are subsequently selected for drug resistance all determine the relative success of this selection. frequency rates after mutagen treatment of 12 of 55 or 22% A single treatment of the TA3 and SS1 mammary tumors of for the PCC4 teratocarcinoma, 49 of 55 or 89% for the Lewis A and BALB/c mice, respectively, with EMS results in the lung carcinoma, and 11 of 76 (14%) for the P815 mastocytoma. conversion of the uncloned parent population to the turn" In some cases, these frequency rates are based on observed phenotype. TA3-E and SS1-E both generate strong CMC re "decreased tumorigenicity," rather than on the absolute failure sponses and are protective against challenge with the parent of individual clones to grow. In contrast, we have chosen to tumor. However, these newly derived uncloned populations are use the criterion of progressive growth or no progressive relatively unstable (not yet proven for SS1-E) and selection of growth as the singular estimation of the turn" phenotype. truly stable turn" variants results only after cloning mutagen- Furthermore, in most cases, we have cloned mutagen-treated treated cells maintained in culture for 4 to 6 weeks. Similar populations after 1 month in culture (this is particularly true of the MDAY-D2 tumor), when any unstable turn" clones will have findings with the PETE variant of the P815 mastocytoma and the MDOT-4 variant of MDAY-D2 confirm this observation. reverted to the turn* phenotype as shown for the TA3-E, PETE, Both EMS and MNNG appear to generate turn" variants, but and MDOT-4 variants. Boon ef al. do not describe any delay in this effect differs depending on the tumor. SS1 cells treated cloning after mutagen treatment. In addition, we have purposely delayed calling a clone turn" until it has been shown to: (a) with EMS generate stronger CMC responses than do those retain the turn" phenotype after 3 in vivo injections of 5 x 10s treated with MNNG, but neither (uncloned) population grew in vivo. In contrast, neither EMS nor MNNG was successful in cells; (b) generate a strong CMC response; and (c) protect generating turn" clones from MDAY-D2 after a single treatment. against in vivo challenge by the parent tumor. Multiple treatments of the MDAY-D2 cells with EMS did select Despite these rigorous criteria, we experienced frequency for stable tum+, met" variants, but no turn" variants. This was rates for the selection of turn" variants ranging from 6 to 95%. only achieved when MDAY-D2 was treated twice with MNNG. Even the lowest frequency at 6% is a 600-fold increase over These findings were further complicated by experiments in the known frequency of selection for 6-thioguanine resistance which MDAY-D2 or P815 cells were treated with EMS twice (10""). The mechanisms for this inordinate frequency are cur but were selected in addition for resistance to ouabain and 6- rently unknown. However, previous work by Preud'Homme ef thioguanine. Only when drug selection was performed in as al. (25) and Yelton ef al. (33) in detecting antibody structural sociation with multiple EMS treatments were we successful in variants in myeloma cells after mutagen treatment reported obtaining turn" variants with EMS. The precise role that drug rates of 2 to 6%. Adair ef al. (1 ) and Siciliano ef al. (26) have selection plays in the system is currently unclear. However, 6- reported phenotypic variants of 1% or more in the expression thioguanine itself has been shown to cause chromosome dam of isozymes in Chinese hamster ovary cells. Fisher and Gitone age (23) and thus could be mutagenic. Therefore, a combina (9) have recently demonstrated that fibrosarcoma cells ex tion of EMS plus 6-thioguanine may effectively act as a double posed to UV in vitro will produce more experimental mÃ©tastases mutagen treatment. This possibility is certainly supported by than do counterpart untreated cultures. Mutation after UV the work of the Bonmassars, Goldin, and their colleagues (3, treatments was assessed by demonstrating a 10-fold increase 7), who have selected turn" variants of the L1210 leukemia in the conversion of UV-treated cells to ouabain resistance. after prolonged exposure of these cells in vivo to potentially These findings, while in agreement with our observations re mutagenic drugs such as 5-(3,3'-dimethyl-1-triazeno)imida- garding frequency rates, seem at first to contradict our findings zole-4-carboxamide. of more turn" variants, since these authors demonstrated an Our success with the P815 mastocytoma was not nearly as increase in malignancy (conversion to the metastatic pheno clear as that reported by Boon ef al. (6, 27). This could relate type) rather than diminished malignancy. Without detracting to the fact that we used a different mutagen (EMS) or that the from their basic observation, we would point out that the cell P815 cell line that we used was different than that used by lines were not cloned after UV treatment. We would therefore Boon ef al. However, we were successful in obtaining a turn" predict that the UV-treated cultures contain a mixture of turn* and turn" clones, with the tum+ clones dominating. This could population (PETE) derived from P815, confirming the basic efficacy of this approach. obviously be proved only by cloning the UV-treated cells im An additional question which we have attempted to resolve mediately as well as 2 months after the final UV exposure. It is whether the turn" clones that we have selected from any should also be noted that Ramshaw ef a/.9 have reported the particular tumor share antigens. Our experiments with the MDNN clones lend support to the findings of Boon and his 91. A. Renshaw, R. C. Warrington. and S. Carlsen. A 6-thioguanine variant of MAT 13762 cell line which is no longer tumorigenic or metastatic, submitted lor ' P. Frost, unpublished observations. publication.

JANUARY 1983 131

induction of a turn" variant from a rat mammary tumor following P815 II: T lymphocyte mediated cytolysis. J. Exp. Med., Ã•52. 1184-1193, treatment of the cells with 6-thioguanine. 1980. 7. Contessa, A. R., Bonmassar, A., Giampietri, A., Circolo, A., Goldin. A., and It is apparent that, although precedents for the observed Fioretti, M. D. In vitro generation of a highly immunogenic subline of L1210 high frequency rates exist, a clear explanation for the mecha leukemia following exposure to 5-(3,3'-dimethyl-1-triazeno)imidazole-4-carboxamide. Cancer Res., 41: 2476-2482. 1981. nisms is not available. Because these rates are beyond what could be explained by "point mutation" alone, we chose to 8. de Landazuri, M. O., and Herberman, R. B. Specificity of cellular immune reactivity to virus induced tumors. Nat. New Biol., 238. 18-19. 1972. describe our experiments as mutagen treatment. This broader 9. Fisher, M. S., and Gitone, M. N. Enhanced metastatic potential of murine fibrosarcomas treated in vitro with ultraviolet radiation. Cancer Res., 41: term allows for broader effects of the mutagens used. The 3018-3023, 1982. possibility of "regulator gene" control of tumor antigen expres 10. Frost, P., and Kerbel, R. S. Mutagenesis can result in the generation of sion, while feasible, remains speculative. immunogenic variants cross reactive with the parent tumor. Clin. Res., 29. The high frequency of the selection of turn" clones after 7739, 1981. 11. Frost, P., Kerbel, R. S., and Tartamella-Biondo, R. Generation of highly mutagen treatment is not at variance with the Nowell hypothesis metastatic tumors in DBA/2 mice: oncogenicity of A strain tumor cells. Invasion Met., 1: 22-33, 1981. of tumor progression (24). Nowell postulated that tumor cells 12. Frost, P., Smith, J., and Frost, H. The radiolabeling of lymphocytes and are genotypically unstable and are constantly mutating in a tumor cells with Ill-Indium. Proc. Soc. Exp. Biol. Med., Â»57:61-65, 1978. random fashion. If one uses the expression of tumor antigens 13. Gillin, G. D., Raufa, D. J., Beaudet. A. L., and Caskey, G. T. 8-Azaguanine resistance in mammalian cells: I: Hypoxanthine-guanine phosphoribosyl as an example, then this genetic instability results in the de transferase. Genetics, 72. 239-253, 1972. velopment of a more malignant tumor, since any new randomly 14. Hewitt, H. B. A critical examination of the foundations of immunotherapy of produced variants which express antigens which generate a cancer. Clin. Radiol., 30. 361-364, 1979. 15. Hewitt, H. G., Blake, E. R., and Walder, A. S. A critique of the evidence of host response are removed by immunological mechanisms. active host defense against cancer based on personal studies of 27 murine However, the prediction from Nowell's hypothesis is that, de tumors of spontaneous origin. Br. J. Cancer, 33. 241-259. 1976. 16. Kerbel, R. S. Surface antigen changes accompanying lymphocyte activation: spite their removal, new additional turrr clones are constantly selective reactivity of certain anti-H2 antisera with activated lymphocytes of being produced. We would postulate that mutagen treatment unrelated H-2 haplotype. Scand. J. Immunol.. 6. 1029-1039. 1977. in vitro could stabilize individual turn" clones and prevent some 17. Kerbel, R. S. Immunologie studies of membrane mutants of a highly meta static murine tumor. Am. J. Pathol., 97. 609-622, 1979. of them from reverting to the malignant phenotype. 18. Kerbel, R. S. Increased sensitivity of rosetting assay for Fc receptors Finally, we would offer an optimistic note regarding sponta obtained by using non-hemagglutinating monoclonal antibodies. J. Immunol. neous murine tumors which Hewitt, and others, have implied Methods, 34: 1-10, 1980. 19. Kerbel, R. S.. Florian, M., Man, M. S., Dennis, J., and McKenzie, I. F. C. may not express tumor-associated antigen (14, 15). SS1 is a Carcinogenicity of tumor cell populations. Origin of a putative H-2 isoanti- spontaneous BALB/c tumor which is not immunogenic by any genic loss variant tumor. J. Nati. Cancer Inst., 64. 1221-1230, 1980. available assessment. SS1-E is, however, a strongly immuno 20. Kerbel, R. S.. and Frost, P. Heritable alterations in tumor-cell immunogenicity Immunol. Today, 3. 34-35, 1982. genic variant of SS1 selected after a single mutagen treatment 21. Kerbel, R. S., Twiddy, R. R., and Robertson, D. M. Induction of a tumor with with EMS. Furthermore, the strong CTL populations induced greatly increased metastatic growth potential by injection of cells from a low against SS1-E were found to kill the parental SS1 cells. This metastatic H-2 heterozygous tumor cell line into an H-2 incompatible paren tal strain. Int. J. Cancer, 22. 583-594, 1978. shows that SS1 cells are indeed antigenic but that the presence 22. Kohler, G. L., and Milstein, C. Derivations of specific antibody-producing of the putative tumor antigen is not sufficient to render the cells tissue culture and tumor lines by cell fusion. Eur. J. Immunol.. 6. 511 -519, 1976. immunogenic. Boon and his colleagues have obtained similar 23. Maybaum. J., and Mandel, H. G. Differential chromatid damage induced by results using one of Hewitt's nonimmunogenic tumor lines (2, 6-thioguanine in CHO cells. Exp. Cell Res., )35: 465-467, 1981. 28). These results are clearly of the utmost importance to the 24. Nowell. P. The clonal evolution of tumor cell populations. Science (Wash. D. C.), 794. 23-28, 1976. controversial status of the immunogenicity of spontaneous 25. Preud'Homme. J. L., Buxbaum, J.. and SchÃ¤rft, M. D. Mutagenesis of mouse tumors (14) and point a way toward effective restoration of myeloma cells with melphalan. Nature (Lond.). 245: 320-322, 1973. immunogenicity of such tumors (2). 26. Siciliano, M. J., White, B. F., and Humphrey, R. M. Electrophoretically detectable mutations induced in CHO cells by varying doses of ultraviolet radiation, in press, 1983. 27. Uyttenhove, C., Van Snick, J., and Boon, T. Immunogenic variants obtained REFERENCES by mutagenesis of mouse mastocytoma P815 I: rejection by syngeneic mice. J. Exp. Med., 752: 1175-1183, 1980. 1. Adair, G. M.. Siciliano. M. J.. and Humphrey. R. M. Induction of electropho- 28. Van Pel, A., and Boon, T. Protection against a non-immunogenic mouse retic mutants at multiple gene loci in CHO cells: differential mutability of leukemia by an immunogenic variant obtained by mutagenesis. Proc. Nati. gene loci and apparent "hot spots" for mutation. J. Cell Biol.. 91: 382A, Acad. Sei. U. S. A., 79: 4718-4722, 1982. 1981. 29. Van Pel, A.. Georlette, M.. and Boon, T. Tumor cell variants obtained by 2. Baldwin. R. W., and Klein. G. Sapporo Cancer Seminar: Escape of tumor mutagenesis of a Lewis lung carcinoma cell line: immune rejection by cells from immune controls. Cancer Res., 42: 1608-1609, 1982. syngeneic mice. Proc. Nati. Acad. Sei. U. S. A., 76: 5282-5285, 1979. 3. Bonmassar. E., Bonmassar. A., Vadlamudi, S., and Goldin. A. Immunological 30. Wiltrout, R. H., and Frost, P. Cell mediated cytotoxic responses induced in alterations of leukemia cells in vivo after treatment with an antitumor drug. vivo and in vitro by a metastatic murine tumor. J. Immunol., 124: 2254-2263, Proc. Nati. Acad. Sei. U. S. A., 66: 1089-1095, 1970. 1980. 4. Boon, T., and Kellerman. O. Rejection by syngeneic mice of cell variants 31. Wiltrout, R. H., Frost, P., and Cummings. G. D. Isotope release cytotoxicity obtained by mutagenesis of a malignant teratocarcinoma line. Proc. Nati. assay with the use of Ill-Indium: advantage over chromium-51 in long term Acad. Sei. U. S. A., 74: 272-275, 1977. assays. J. Nati. Cancer Inst., 61: 183-188, 1978. 5. Boon, T., and Van Pel, A. Teratocarcinoma cell variants rejected by synge 32. Wiltrout, R. H., Frost, P., and Morrison, M. K. Immune regression of viceral metastasis in athymic mice. Correlation of "low level" in vitro CMC reaction neic mice: protection of mice immunized with these variants against other variants and against the original malignant cell line. Proc. Nati. Acad. Sei. U. with allogroft rejection in vivo. Transplantation (Baltimore), 29. 283-286, S. A., 75. 1519-1523. 1978. 1980. 6. Boon. T., Van Snick. J., Van Pel, A., Uyttenhove, C., and Marchand, M. 33. Yelton, D. E., Cook, W. D., and SchÃ¤rft, M. D. Somatic variants in mouse Immunogenic variants obtained by mutagenesis of mouse mastocytoma myeloma and hybridoma cell lines. Transplant. Proc.. 12: 439-442. 1980.

132 CANCER RESEARCH VOL. 43

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1983 American Association for Cancer Research. Mutagen Treatment as a Means for Selecting Immunogenic Variants from Otherwise Poorly Immunogenic Malignant Murine Tumors

Philip Frost, Robert S. Kerbel, Elaine Bauer, et al.

Cancer Res 1983;43:125-132.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/43/1/125

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/43/1/125. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.