609
3,. gen. Virol. (1979), 42, 6o9-614 Printed in Great Britain
Virus Carrier State Suppresses Tumorigenicity of Tumor Cells in Athymic (Nude) Mice (Accepted 2 October I978 ) SUMMARY
Nude mice injected subcutaneously with normal uninfected BHK 21 cells or HeLa cells regularly develop large, rapidly-growing tumours at the subcutaneous site of inoculation. However, these same tumour cell lines when persistently infected with VSV or other enveloped RNA viruses are either rejected or form small nodules in nude mice. This rejection phenomenon probably involves some type of immunocyte since heavily-irradiated nude mice (500 rads) cannot reject persistently infected cells but develop large, rapidly-growing tumours which shed virus and defective interfering virus (DI) and which do not exhibit the lymphocytic infiltration observed in the nodules of unirradiated mice given persistently infected cells. Finally, it was possible to select a subline of B H K 2I-VSV carrier cells which regularly produces large rapidly-growing tumours in normal unirradiated nude mice, although all these carrier cells express virus antigen and shed large amounts of mature infectious virus and DI both in vivo and in vitro. Due to its T-cell deficiency, the athymic nude mouse is unable to reject allogeneic and xenogeneic tissue transplants (Pantelouris, I968; Rygaard, 1973) and is, therefore, an excellent host for assay of tumorigenicity of most cell types (Klinger et al. 1976; Freedman & Shin, I974; Stiles et al. 1976). However, despite this deficiency ofT-cell rejection mechanisms, the nude mouse is not totally immunologically incompetent (Koene et al. I974). In fact, the mice usually exhibit approximately normal levels of spontaneous tumour formation (Rygaard, I973), an enhanced activated macrophage response to intracellular bacterial infections and certain strong but, as yet, ill-defined, cellular responses to foreign cells and infectious agents (Maguire et al. 1976; Haller et al. 1977; Miller et al. 1977; Lefkowits, 1978). It is known that both tumour virus-induced and chemically-induced tumour cells may express strong tumour-specific transplantation antigens at the cell surface, but these do not normally prevent tumorigenesis by the tumour cells in the athymic nude mouse (Rygaard, I973; Klinger et al. I976; Freedman & Shin, I974; Stiles et al. I976; Reid & Shin, I978). However, we demonstrate here that persistent, non-cytopathic infection of tumour cells by a number of enveloped RNA viruses suppresses tumorigenicity of the tumour cells to a remarkable degree in the nude mice. This suppression of tumorigenicity is apparently due to a strong cellular and/or humoral response by the nude mice against virus-encoded cell surface antigens or virus-altered tumour cell antigens, or other induced neoantigens since cells cured of persistent virus infection are again strongly tumorigenic and because whole body irradiation of the nude mouse abrogates the suppression of tumorigenesis of the virus infected cells. It has been shown elsewhere (Holland & Villarreal, 1974) that BHK 21 cells, HeLa and other cultured cells can become persistently infected by vesicular stomatitis virus (VSV), rabies virus, and other enveloped RNA viruses for indefinite periods of many years if defective interfering viruses (DI) are present in the original input virus inoculum. These oo22-1317/79/oooo-3384 $02.00 (~) I979 SGM
Short communications
6IO
Table i. Tumorigenicity of uninfected B H K 2z and HeLa cells, and lack of
tumorigenicity of their viral carrier cultures in B A L B / c nude mice Results at six weeks after injection ___z
r
Cell line BHK 2I
Persistently infected with Nothing found Benign nodule None (control)* VSV?*
HeLa
--
--
1/22
21/22
Tumour I4/14 --
NWS FLU§ Mumps None (control) Measles
--t/4
3/3 -3/4
-5/5 --
gag
4/I2
8/I2
--
--
I9/19
--
* Two x IO~ cells were injected subcutaneously into each mouse in all experiments above. When a dosage curve was done, all dosages down to IO cells produced tumours. "~ VSV = vesicular stomatitis virus. :~ When a dosage curve was done, no dosage up to I "5 × Io7 cells produced a rapidly-growing tumour, and a minimum of IO4 cells was necessary to produce a visible nodule. § NWS FLU = influenza virus A (NWS strain). persistently infected cells grow in vitro at normal or near normal rates even though nearly every cell is infected and expressing viral antigen at the plasma membrane as measured by fluorescent antibody and by quantitative 131I-labelled immunoglobulin binding assays (Holland et al. 1976; Villarreal & Holland, I976). Apparently the continuous presence o f D I and the accumulation o f mutations in the virus allow the establishment o f this permanent non-cytocidal carrier state by normally virulent R N A viruses. Surprisingly we f o u n d that these carrier cells do not express any tumorigenic potential when inoculated subcutaneously into n u / n u h o m o z y g o u s mice even t h o u g h they grew rapidly in cell culture. B H K 21 cells always give rise to large and highly-vascularized tumours within three weeks o f the injection, whereas the same cells persistently infected with VSV failed to produce any tumours within six weeks and longer, regardless o f the n u m b e r o f cells injected. The persistently infected cells either formed a small benign nodule or there was no detectable growth at all of the injected cells. This is summarized in Table I which shows that this p h e n o m e n o n was obtained with B H K 2I cells persistently infected with either VSV or influenza virus or m u m p s virus, or by H e L a cells persistently infected with either VSV or measles virus. In all cases the uninfected t u m o u r cells formed rapidly-growing tumours in the nude mice and the corresponding virus carrier cells were either rejected (no nodule or t u m o u r growth) or they formed a very slowly-growing or non-growing benign nodule. These benign nodules are typically small, hard and poorly vascularized; histologically they show a fibrous reaction and encapsulation. The nodules persist in mice for six months or more, and the B H K 2I-VSV carrier cells within them remain viable indefinitely, since the nodules can be excised, minced and dispersed and their c o m p o n e n t cells immediately re-cultured as rapidly growing in vitro cell cultures. In every case these nodule cells were still B H K 2I-VSV carrier cells as evidenced by their hamster karyotype and by the fact that they continue to shed infectious virus and D I particles, and are resistant to challenge by VSV. We have shown earlier that B H K 2 I-VSV carrier cells can be ' c u r e d ' o f persistent VSV infection by cloning in the presence o f antiviral antibody (Holland et al. I976 ). These cells are ' c u r e d ' o f detectable VSV expression by the following criteria: (I) no detectable infectious virus or D I are released; (2) no VSV antigens are detectable within these cells by fluorescent antibody techniques; (3) no specific resistance to VSV homologous challenge
Short communications
6I I
infection can be detected. We cannot rule out the presence of non-expressed virus genes, however. It should be noted that B H K 2I cells ' c u r e d ' in vitro after more than 3 years of persistent VSV infection are once again highly tumorigenic in nude mice. Furthermore, mice immunized 2 weeks previously by infection with live virus can still support the growth of turnouts from injections of HeLa cells or B H K 2I cells. Finally, when mice carrying a benign nodule of B H K 2I-VSV carrier cells were injected weeks later on the contralateral side with from IOa to Io 7 normal B H K 2I cells they always developed a rapidly-growing, soft turnout. This suggests that the anti-tumour response to B H K 21-VSV carrier cells is directed at virus antigens or virus-altered tumour cell antigens or other virus-induced target molecules. This interpretation is corroborated by results from experiments in which Io 6 B H K 2I cells were co-injected with Io 6 B H K 2I-VSV carrier cells. The mixture of the two cell lines in the same inoculum produced tumours associated with, or surrounding benign nodules. Thus, even when localized to the same site, the host responds differently to the carrier cell line and its uninfected parental cell line (virus release from the carrier cells is too low (Holland et al. I976) to infect all associated normal B H K 2I cells). The above results indicate that some form of cellular or humoral 'immune response' is controlling growth of tumour cells persistently infected by viruses. Since such a response by the nude mouse might be expected to be radiation-sensitive, as are most immunological responses, we examined the effect of high dosage whole body irradiation on the phenomenon. Radiation sensitivity of the response could rule out a trivial explanation for the difference in tumorigenicity of the parental and carrier cell lines: that the persistent virus infections of the tumour cells somehow damages their ability to grow in vivo although they grow at near normal rates in vitro. Nude mice that were whole body irradiated with 5oo rads of caesium la5 and subsequently injected with BHK-VSV carrier cells all developed rapidly growing, soft, well-vascularized tumours. Tumours removed from irradiated mice and passaged in culture shed infectious virus and were resistant to challenge by homologous virus infection. Furthermore, infectious virus could be isolated from the spleens of these mice. When mice were injected 12 days after irradiation, the VSV carrier cells produced tumours within three weeks after injection, but in mice injected with the carrier cells on the day of irradiation, the tumours appeared later (within 6 to 7 weeks after injection). Thus, there is an immune mechanism continuing to function for up to two or three weeks after irradiation. The virus carrier cells survive this residual immune response and then express their tumorigenic potential when the immune cells turn over and fail to be replaced after host irradiation. Little is known about radiation effects on nude mouse immunocytes, although some are rather radiation-resistant (Bonmassar, 1978). Presumably, our results are due to immunocyte depletion, but this large dosage of 5oo rads could have altered control of infected cells in other ways (by reduced interferon production, etc). Histological sections of control uninfected B H K 21 or H e L a cell tumours showed less host response than that seen with B H K 2 I-VSV intected cells. Histological sections of the benign nodules formed by B H K 2I persistently infected with VSV showed a strong fibrous reaction and encapsulation, with extensive lymphocytic infiltration into the tumour mass. There is no fibrous reaction or infiltration and only a slight encapsulation at the boundary of the tumour cell growth either with tumours from uninfected cells in normal nude mice, or with the virus-infected carrier cells in the irradiated nude mice. Clearly, there is a cellular reaction to persistently-infected cells which can limit tumour growth in normal mice and which does not occur in heavily irradiated nude mice.
612
Short communications Table 2. Tumorigenicity in B A L B / c nude mice of a selected, tumour-producing ' variant' subline of B H K 2I cells persistently infected with V S V
Cell line Standard BHK 21-VSV carrier line (CAR4) Selected tumour-producing subline of CAR4 VSV carrier cells, (CAR4-P)
Tumour production at 6 to 7 weeks after injection of 2 X IO ~ cells
% of carrier cells with VSV antigen
VSV cell Mature surface virus Mature DI antigen production production
0/22*
> 99 %t
Weak:l:
Low§
Low§
8/8*
> 99 %t
Strong:l:
High§
High§
* CARcP cells produced large, soft, invasive tumours whereas the parental CAR4 VSV carrier BHK 2I cells produced only small encapsulated nodules or nothing. t Intracellular VSV antigen was determined on fixed cells by fluorescent antibody (indirect) test. :~ VSV surface glycoprotein antigen on the surface of cultured CAR4 or CAR4-P cells was determined on living unfixed cells grown in culture on glass slides (indirect fluorescent antibody test). § Mature infectious virus production ranged from Io-2 to Io-5 p.f.u./cell for CAR4 and from i01 to I03/cell for CAR4-P (as determined on I0 % minced tumour suspensions & vivo, or on cell culture fluids in vitro). Mature DI production was determined by one stage amplification assays (Holland et al. I976) on I0 % suspensions of minced tumour suspensions. After periods of several months or more (4 to 6), a small percentage of the nude mice carrying small nodules of persistently-infected cells eventually did develop soft, rapidlygrowing tumours. Almost all of these, upon excision, proved to be B H K 2I cells which had become ' c u r e d ' of virus infection. However, a few of these late-arising tumours did shed virus, and when passaged in cell culture were shown to be carrier cultures shedding infectious virus and exhibiting strong immunity to homologous VSV challenge infection. We excised one such virus-shedding tumour which arose from a nodule of B H K 2I-VSV carrier cells after it had maintained benign nodular characteristics for 7 months. The persistently-infected B H K 21 cells from this tumour rapidly re-established growth in vitro and they shed both infectious virus and D I particles continuously in cell culture. U p o n re-injection into another nude mouse they formed a large soft tumour similar to uninfected B H K 21 cells but which was highly invasive. Once again, excision of this second passage tumour after several months provided cultured cells which shed virus and D I and which were resistant to VSV challenge. This selected 'virus-positive' tumour has repeatedly been serially passaged in nude mice, and it continues to produce large amounts of virus and D I in vivo (Table 2). Virus and D I have been isolated regularly from the blood, spleens, kidneys and livers of these mice. This tumour provides an excellent model for cellular and biochemical studies of virus persistence in vivo. Surprisingly, this virus-positive tumour sheds much higher levels of virus and mature D I in vivo and in vitro than standard carrier cells, and it expresses higher levels of virus antigen both intracellularly and at the cell surface (Table 2). The mechanism by which it is able to escape the above reported tumour suppressive ability of nude mice for virus infected cells is under investigation. The nature of tumour antigens is poorly understood even in isologous host systems, and there is probably considerable heterogeneity in the host anti-tumour response (Reid & Shin, I978). The nature of the ' a n t i - t u m o u r ' response mounted by the nude mice against virusinfected tumour cells is being explored and is probably cellular, since it is radiations sensitive. This response might be important in well-known difficulties of heterografting
S h o r t communications
613
certain types of human tumours such as prostatic carcinoma into the nude mice (Reid & Shin, I978). Moreover, studies trying to correlate in vitro cellular phenotypes with malignancy where malignant transformation of a cell line is assayed by its tumorigenicity in nude mice (Klinger et al. 1976; Stiles et al. I976), should be re-evaluated since some cell lines found to be non-tumorigenic may, in fact, be malignant but immunologically suppressed by the nude mice. Finally, the ability to maintain persistently infected cells in nodules in vivo for periods exceeding six months (and in certain selected cells as tumours, indefinitely) followed by rapid re-establishment in cell culture should allow better analysis of cellular and biochemical parameters of persistent virus infection. Recent in vitro studies (Minato et al. 1979) indicate strongly that natural killer (NK) cells (and not B or T cells) are the immunocytes responsible for the nude mouse rejection of persistently-infected cells reported above. An auxiliary role of humoral antibody cannot be ruled out since we find low levels of virus-neutralizing antibody in most nude mice with VSV carrier cell nodules or tumours. Note added i n p r o o f After this work was completed we discovered that a similar rejection of tumour cells persistently infected by Sendai virus was shown by Yamada & Hatano (1972) in normal hamsters.
This work was supported by grant number AI 14627 from the U.S. Public Health Service. Biology Department University o f California at San Diego L a Jolla, California 92093, U . S . A .
LOLA M . REID* CHARLOTTE L. JONES JOHN HOLLAND
* Present address: Department of Molecular Pharmacology, Albert Einstein College of Medicine, 13o0 Morris Park Avenue, Bronx, New York IO46I, U.S.A. REFERENCES BONMASSAR,E. 0978). Radioresistant inhibition of tumor growth. In Symposium on the Use of Athymic (Nude) Mice in Cancer Research (in the press). Edited by D. Houchens and T. Ovejera. New York: Gustav Fisher. FREEDMAN, V. H. & SHIN, S. (I974). Cellular tumorigenicity in nude mice: correlation with cell growth in semi-solid medium. Cell 3, 355-359. HALLER, O., HANSON,M., KIESSLING,R. & WIGZELL, H. (1977). Role of non-conventional natural killer cells in resistance against syngeneic tumor cells in vivo. Nature, London 270, 6o9-61 I. HOLLAND, J.J. & VlLLARREAL, L.P. (I974). Persistent noncytocidal vesicular stomatitis virus infections mediated by defective T particles that suppress virion transcriptase. Proceedings of the NationaIAcademy of Sciences of the United States of America 7 x, 2956-296o. HOLLAND, J. J.~ VILLARREAL,L. P., WELSH~R. M., OLDSTONE~M. B. A., KOHNE~D., LAZZAKINI,R. & SCOLNICK,E. (I976). Long-term persistent vesicular stomatitis virus and rabies virus infection of cells in vitro. Journal of General Virology 33, I93-211. KLINGER, H. P., SHIN, S. & FREEDMAN,V. H. (1976). Enhanced anchorage independence and tumorigenicity of aneuploid Chinese hamster cells with nearly doubled chromosome complements. Cytogenetics and Cell Genetics 17, I85-I99. KOENE, a., GERLAG,J., JANSEN,J., HAGEMANN,J. & WIJDEVELD, P. (I974). Rejection of skirt grafts in the nude mouse. Nature, London 25I, 69-7o. LEFKOWITS, I. (1978). In Symposium on the Use of Athymic (Nude) Mice in Cancer Research (in the press). Edited by D. Houchens and T. Ovejera. New York: Gustav Fisher. MAGUIRE, H. JR., OUTZEN, H. C., CUSTER, R. P. & PREHN, R. T. (1976). Brief communication: invasion and metastasis of a xenogeneic tumor in nude mice. Journal of the National Cancer Institute 57, 439-442. MILLER, R. G., SCHILLING, R. M. & PHILLIPS, R. A. (I977). Requirements for non T-cells in the generation of cytotoxic T lymphocytes (CTL) in vitro: II. Characterization of the active cells in the spleen of nude mice. Journal of Immunology xx8, I66-174.
614
Short communications
(1979). M e c h a n i s m o f rejection o f v i r u s persistently infected tumour cells by athymic nude mice. Journal of Experimental Medicine (in the press). PANTELOURIS, E. M. (1968). Absence of thymus in a mouse mutant. Nature, London 217, 37o. REID, L. & SHIN, S. (1978). Transplantation of endocrine tumors in nude mice. In The Nude Mouse in Experimental and Clinical Research, pp. 313-35I. Edited by J. Fogh and B. Giovanella. New York: Academic Press. RYGAARD, J. (I973). In Thymus and Self: lmmunobiology of the Mouse Mutant Nude. New York: John Wiley and Sons. STILES, C. D., DESMOND, W. O., SATO, G. & SAIER, M. H., JUN. (1976). Relationship of cell growth behavior in vitro to tumorigenicity in athymic nude mice. Cancer Research 36, 33oo-3305. VILLARREAL,L . P . &~OLLAND, ~', J. (I976). R N A synthesis in BHK, 2: cells persistently infected with vesicular stomatitis virus and rabies virus. Journal of General Virology 33, 213-224 • YAMADA,T, & HATANO, M. (I972). Lowered transplantability of cultured tumour cells by persistent infection with paramyxovirus. (HVJ) GANN: The Japanese Journal of Cancer Research 63, 647-655.
MINATO, N., BLOOM, B. R., JONES, C., HOLLAND, J. J. & REID, L. M.
(Received : 3 July I 9 7 8 )
3,. gen. Virol. (1979), 42, 6o9-614 Printed in Great Britain
Virus Carrier State Suppresses Tumorigenicity of Tumor Cells in Athymic (Nude) Mice (Accepted 2 October I978 ) SUMMARY
Nude mice injected subcutaneously with normal uninfected BHK 21 cells or HeLa cells regularly develop large, rapidly-growing tumours at the subcutaneous site of inoculation. However, these same tumour cell lines when persistently infected with VSV or other enveloped RNA viruses are either rejected or form small nodules in nude mice. This rejection phenomenon probably involves some type of immunocyte since heavily-irradiated nude mice (500 rads) cannot reject persistently infected cells but develop large, rapidly-growing tumours which shed virus and defective interfering virus (DI) and which do not exhibit the lymphocytic infiltration observed in the nodules of unirradiated mice given persistently infected cells. Finally, it was possible to select a subline of B H K 2I-VSV carrier cells which regularly produces large rapidly-growing tumours in normal unirradiated nude mice, although all these carrier cells express virus antigen and shed large amounts of mature infectious virus and DI both in vivo and in vitro. Due to its T-cell deficiency, the athymic nude mouse is unable to reject allogeneic and xenogeneic tissue transplants (Pantelouris, I968; Rygaard, 1973) and is, therefore, an excellent host for assay of tumorigenicity of most cell types (Klinger et al. 1976; Freedman & Shin, I974; Stiles et al. 1976). However, despite this deficiency ofT-cell rejection mechanisms, the nude mouse is not totally immunologically incompetent (Koene et al. I974). In fact, the mice usually exhibit approximately normal levels of spontaneous tumour formation (Rygaard, I973), an enhanced activated macrophage response to intracellular bacterial infections and certain strong but, as yet, ill-defined, cellular responses to foreign cells and infectious agents (Maguire et al. 1976; Haller et al. 1977; Miller et al. 1977; Lefkowits, 1978). It is known that both tumour virus-induced and chemically-induced tumour cells may express strong tumour-specific transplantation antigens at the cell surface, but these do not normally prevent tumorigenesis by the tumour cells in the athymic nude mouse (Rygaard, I973; Klinger et al. I976; Freedman & Shin, I974; Stiles et al. I976; Reid & Shin, I978). However, we demonstrate here that persistent, non-cytopathic infection of tumour cells by a number of enveloped RNA viruses suppresses tumorigenicity of the tumour cells to a remarkable degree in the nude mice. This suppression of tumorigenicity is apparently due to a strong cellular and/or humoral response by the nude mice against virus-encoded cell surface antigens or virus-altered tumour cell antigens, or other induced neoantigens since cells cured of persistent virus infection are again strongly tumorigenic and because whole body irradiation of the nude mouse abrogates the suppression of tumorigenesis of the virus infected cells. It has been shown elsewhere (Holland & Villarreal, 1974) that BHK 21 cells, HeLa and other cultured cells can become persistently infected by vesicular stomatitis virus (VSV), rabies virus, and other enveloped RNA viruses for indefinite periods of many years if defective interfering viruses (DI) are present in the original input virus inoculum. These oo22-1317/79/oooo-3384 $02.00 (~) I979 SGM
Short communications
6IO
Table i. Tumorigenicity of uninfected B H K 2z and HeLa cells, and lack of
tumorigenicity of their viral carrier cultures in B A L B / c nude mice Results at six weeks after injection ___z
r
Cell line BHK 2I
Persistently infected with Nothing found Benign nodule None (control)* VSV?*
HeLa
--
--
1/22
21/22
Tumour I4/14 --
NWS FLU§ Mumps None (control) Measles
--t/4
3/3 -3/4
-5/5 --
gag
4/I2
8/I2
--
--
I9/19
--
* Two x IO~ cells were injected subcutaneously into each mouse in all experiments above. When a dosage curve was done, all dosages down to IO cells produced tumours. "~ VSV = vesicular stomatitis virus. :~ When a dosage curve was done, no dosage up to I "5 × Io7 cells produced a rapidly-growing tumour, and a minimum of IO4 cells was necessary to produce a visible nodule. § NWS FLU = influenza virus A (NWS strain). persistently infected cells grow in vitro at normal or near normal rates even though nearly every cell is infected and expressing viral antigen at the plasma membrane as measured by fluorescent antibody and by quantitative 131I-labelled immunoglobulin binding assays (Holland et al. 1976; Villarreal & Holland, I976). Apparently the continuous presence o f D I and the accumulation o f mutations in the virus allow the establishment o f this permanent non-cytocidal carrier state by normally virulent R N A viruses. Surprisingly we f o u n d that these carrier cells do not express any tumorigenic potential when inoculated subcutaneously into n u / n u h o m o z y g o u s mice even t h o u g h they grew rapidly in cell culture. B H K 21 cells always give rise to large and highly-vascularized tumours within three weeks o f the injection, whereas the same cells persistently infected with VSV failed to produce any tumours within six weeks and longer, regardless o f the n u m b e r o f cells injected. The persistently infected cells either formed a small benign nodule or there was no detectable growth at all of the injected cells. This is summarized in Table I which shows that this p h e n o m e n o n was obtained with B H K 2I cells persistently infected with either VSV or influenza virus or m u m p s virus, or by H e L a cells persistently infected with either VSV or measles virus. In all cases the uninfected t u m o u r cells formed rapidly-growing tumours in the nude mice and the corresponding virus carrier cells were either rejected (no nodule or t u m o u r growth) or they formed a very slowly-growing or non-growing benign nodule. These benign nodules are typically small, hard and poorly vascularized; histologically they show a fibrous reaction and encapsulation. The nodules persist in mice for six months or more, and the B H K 2I-VSV carrier cells within them remain viable indefinitely, since the nodules can be excised, minced and dispersed and their c o m p o n e n t cells immediately re-cultured as rapidly growing in vitro cell cultures. In every case these nodule cells were still B H K 2I-VSV carrier cells as evidenced by their hamster karyotype and by the fact that they continue to shed infectious virus and D I particles, and are resistant to challenge by VSV. We have shown earlier that B H K 2 I-VSV carrier cells can be ' c u r e d ' o f persistent VSV infection by cloning in the presence o f antiviral antibody (Holland et al. I976 ). These cells are ' c u r e d ' o f detectable VSV expression by the following criteria: (I) no detectable infectious virus or D I are released; (2) no VSV antigens are detectable within these cells by fluorescent antibody techniques; (3) no specific resistance to VSV homologous challenge
Short communications
6I I
infection can be detected. We cannot rule out the presence of non-expressed virus genes, however. It should be noted that B H K 2I cells ' c u r e d ' in vitro after more than 3 years of persistent VSV infection are once again highly tumorigenic in nude mice. Furthermore, mice immunized 2 weeks previously by infection with live virus can still support the growth of turnouts from injections of HeLa cells or B H K 2I cells. Finally, when mice carrying a benign nodule of B H K 2I-VSV carrier cells were injected weeks later on the contralateral side with from IOa to Io 7 normal B H K 2I cells they always developed a rapidly-growing, soft turnout. This suggests that the anti-tumour response to B H K 21-VSV carrier cells is directed at virus antigens or virus-altered tumour cell antigens or other virus-induced target molecules. This interpretation is corroborated by results from experiments in which Io 6 B H K 2I cells were co-injected with Io 6 B H K 2I-VSV carrier cells. The mixture of the two cell lines in the same inoculum produced tumours associated with, or surrounding benign nodules. Thus, even when localized to the same site, the host responds differently to the carrier cell line and its uninfected parental cell line (virus release from the carrier cells is too low (Holland et al. I976) to infect all associated normal B H K 2I cells). The above results indicate that some form of cellular or humoral 'immune response' is controlling growth of tumour cells persistently infected by viruses. Since such a response by the nude mouse might be expected to be radiation-sensitive, as are most immunological responses, we examined the effect of high dosage whole body irradiation on the phenomenon. Radiation sensitivity of the response could rule out a trivial explanation for the difference in tumorigenicity of the parental and carrier cell lines: that the persistent virus infections of the tumour cells somehow damages their ability to grow in vivo although they grow at near normal rates in vitro. Nude mice that were whole body irradiated with 5oo rads of caesium la5 and subsequently injected with BHK-VSV carrier cells all developed rapidly growing, soft, well-vascularized tumours. Tumours removed from irradiated mice and passaged in culture shed infectious virus and were resistant to challenge by homologous virus infection. Furthermore, infectious virus could be isolated from the spleens of these mice. When mice were injected 12 days after irradiation, the VSV carrier cells produced tumours within three weeks after injection, but in mice injected with the carrier cells on the day of irradiation, the tumours appeared later (within 6 to 7 weeks after injection). Thus, there is an immune mechanism continuing to function for up to two or three weeks after irradiation. The virus carrier cells survive this residual immune response and then express their tumorigenic potential when the immune cells turn over and fail to be replaced after host irradiation. Little is known about radiation effects on nude mouse immunocytes, although some are rather radiation-resistant (Bonmassar, 1978). Presumably, our results are due to immunocyte depletion, but this large dosage of 5oo rads could have altered control of infected cells in other ways (by reduced interferon production, etc). Histological sections of control uninfected B H K 21 or H e L a cell tumours showed less host response than that seen with B H K 2 I-VSV intected cells. Histological sections of the benign nodules formed by B H K 2I persistently infected with VSV showed a strong fibrous reaction and encapsulation, with extensive lymphocytic infiltration into the tumour mass. There is no fibrous reaction or infiltration and only a slight encapsulation at the boundary of the tumour cell growth either with tumours from uninfected cells in normal nude mice, or with the virus-infected carrier cells in the irradiated nude mice. Clearly, there is a cellular reaction to persistently-infected cells which can limit tumour growth in normal mice and which does not occur in heavily irradiated nude mice.
612
Short communications Table 2. Tumorigenicity in B A L B / c nude mice of a selected, tumour-producing ' variant' subline of B H K 2I cells persistently infected with V S V
Cell line Standard BHK 21-VSV carrier line (CAR4) Selected tumour-producing subline of CAR4 VSV carrier cells, (CAR4-P)
Tumour production at 6 to 7 weeks after injection of 2 X IO ~ cells
% of carrier cells with VSV antigen
VSV cell Mature surface virus Mature DI antigen production production
0/22*
> 99 %t
Weak:l:
Low§
Low§
8/8*
> 99 %t
Strong:l:
High§
High§
* CARcP cells produced large, soft, invasive tumours whereas the parental CAR4 VSV carrier BHK 2I cells produced only small encapsulated nodules or nothing. t Intracellular VSV antigen was determined on fixed cells by fluorescent antibody (indirect) test. :~ VSV surface glycoprotein antigen on the surface of cultured CAR4 or CAR4-P cells was determined on living unfixed cells grown in culture on glass slides (indirect fluorescent antibody test). § Mature infectious virus production ranged from Io-2 to Io-5 p.f.u./cell for CAR4 and from i01 to I03/cell for CAR4-P (as determined on I0 % minced tumour suspensions & vivo, or on cell culture fluids in vitro). Mature DI production was determined by one stage amplification assays (Holland et al. I976) on I0 % suspensions of minced tumour suspensions. After periods of several months or more (4 to 6), a small percentage of the nude mice carrying small nodules of persistently-infected cells eventually did develop soft, rapidlygrowing tumours. Almost all of these, upon excision, proved to be B H K 2I cells which had become ' c u r e d ' of virus infection. However, a few of these late-arising tumours did shed virus, and when passaged in cell culture were shown to be carrier cultures shedding infectious virus and exhibiting strong immunity to homologous VSV challenge infection. We excised one such virus-shedding tumour which arose from a nodule of B H K 2I-VSV carrier cells after it had maintained benign nodular characteristics for 7 months. The persistently-infected B H K 21 cells from this tumour rapidly re-established growth in vitro and they shed both infectious virus and D I particles continuously in cell culture. U p o n re-injection into another nude mouse they formed a large soft tumour similar to uninfected B H K 21 cells but which was highly invasive. Once again, excision of this second passage tumour after several months provided cultured cells which shed virus and D I and which were resistant to VSV challenge. This selected 'virus-positive' tumour has repeatedly been serially passaged in nude mice, and it continues to produce large amounts of virus and D I in vivo (Table 2). Virus and D I have been isolated regularly from the blood, spleens, kidneys and livers of these mice. This tumour provides an excellent model for cellular and biochemical studies of virus persistence in vivo. Surprisingly, this virus-positive tumour sheds much higher levels of virus and mature D I in vivo and in vitro than standard carrier cells, and it expresses higher levels of virus antigen both intracellularly and at the cell surface (Table 2). The mechanism by which it is able to escape the above reported tumour suppressive ability of nude mice for virus infected cells is under investigation. The nature of tumour antigens is poorly understood even in isologous host systems, and there is probably considerable heterogeneity in the host anti-tumour response (Reid & Shin, I978). The nature of the ' a n t i - t u m o u r ' response mounted by the nude mice against virusinfected tumour cells is being explored and is probably cellular, since it is radiations sensitive. This response might be important in well-known difficulties of heterografting
S h o r t communications
613
certain types of human tumours such as prostatic carcinoma into the nude mice (Reid & Shin, I978). Moreover, studies trying to correlate in vitro cellular phenotypes with malignancy where malignant transformation of a cell line is assayed by its tumorigenicity in nude mice (Klinger et al. 1976; Stiles et al. I976), should be re-evaluated since some cell lines found to be non-tumorigenic may, in fact, be malignant but immunologically suppressed by the nude mice. Finally, the ability to maintain persistently infected cells in nodules in vivo for periods exceeding six months (and in certain selected cells as tumours, indefinitely) followed by rapid re-establishment in cell culture should allow better analysis of cellular and biochemical parameters of persistent virus infection. Recent in vitro studies (Minato et al. 1979) indicate strongly that natural killer (NK) cells (and not B or T cells) are the immunocytes responsible for the nude mouse rejection of persistently-infected cells reported above. An auxiliary role of humoral antibody cannot be ruled out since we find low levels of virus-neutralizing antibody in most nude mice with VSV carrier cell nodules or tumours. Note added i n p r o o f After this work was completed we discovered that a similar rejection of tumour cells persistently infected by Sendai virus was shown by Yamada & Hatano (1972) in normal hamsters.
This work was supported by grant number AI 14627 from the U.S. Public Health Service. Biology Department University o f California at San Diego L a Jolla, California 92093, U . S . A .
LOLA M . REID* CHARLOTTE L. JONES JOHN HOLLAND
* Present address: Department of Molecular Pharmacology, Albert Einstein College of Medicine, 13o0 Morris Park Avenue, Bronx, New York IO46I, U.S.A. REFERENCES BONMASSAR,E. 0978). Radioresistant inhibition of tumor growth. In Symposium on the Use of Athymic (Nude) Mice in Cancer Research (in the press). Edited by D. Houchens and T. Ovejera. New York: Gustav Fisher. FREEDMAN, V. H. & SHIN, S. (I974). Cellular tumorigenicity in nude mice: correlation with cell growth in semi-solid medium. Cell 3, 355-359. HALLER, O., HANSON,M., KIESSLING,R. & WIGZELL, H. (1977). Role of non-conventional natural killer cells in resistance against syngeneic tumor cells in vivo. Nature, London 270, 6o9-61 I. HOLLAND, J.J. & VlLLARREAL, L.P. (I974). Persistent noncytocidal vesicular stomatitis virus infections mediated by defective T particles that suppress virion transcriptase. Proceedings of the NationaIAcademy of Sciences of the United States of America 7 x, 2956-296o. HOLLAND, J. J.~ VILLARREAL,L. P., WELSH~R. M., OLDSTONE~M. B. A., KOHNE~D., LAZZAKINI,R. & SCOLNICK,E. (I976). Long-term persistent vesicular stomatitis virus and rabies virus infection of cells in vitro. Journal of General Virology 33, I93-211. KLINGER, H. P., SHIN, S. & FREEDMAN,V. H. (1976). Enhanced anchorage independence and tumorigenicity of aneuploid Chinese hamster cells with nearly doubled chromosome complements. Cytogenetics and Cell Genetics 17, I85-I99. KOENE, a., GERLAG,J., JANSEN,J., HAGEMANN,J. & WIJDEVELD, P. (I974). Rejection of skirt grafts in the nude mouse. Nature, London 25I, 69-7o. LEFKOWITS, I. (1978). In Symposium on the Use of Athymic (Nude) Mice in Cancer Research (in the press). Edited by D. Houchens and T. Ovejera. New York: Gustav Fisher. MAGUIRE, H. JR., OUTZEN, H. C., CUSTER, R. P. & PREHN, R. T. (1976). Brief communication: invasion and metastasis of a xenogeneic tumor in nude mice. Journal of the National Cancer Institute 57, 439-442. MILLER, R. G., SCHILLING, R. M. & PHILLIPS, R. A. (I977). Requirements for non T-cells in the generation of cytotoxic T lymphocytes (CTL) in vitro: II. Characterization of the active cells in the spleen of nude mice. Journal of Immunology xx8, I66-174.
614
Short communications
(1979). M e c h a n i s m o f rejection o f v i r u s persistently infected tumour cells by athymic nude mice. Journal of Experimental Medicine (in the press). PANTELOURIS, E. M. (1968). Absence of thymus in a mouse mutant. Nature, London 217, 37o. REID, L. & SHIN, S. (1978). Transplantation of endocrine tumors in nude mice. In The Nude Mouse in Experimental and Clinical Research, pp. 313-35I. Edited by J. Fogh and B. Giovanella. New York: Academic Press. RYGAARD, J. (I973). In Thymus and Self: lmmunobiology of the Mouse Mutant Nude. New York: John Wiley and Sons. STILES, C. D., DESMOND, W. O., SATO, G. & SAIER, M. H., JUN. (1976). Relationship of cell growth behavior in vitro to tumorigenicity in athymic nude mice. Cancer Research 36, 33oo-3305. VILLARREAL,L . P . &~OLLAND, ~', J. (I976). R N A synthesis in BHK, 2: cells persistently infected with vesicular stomatitis virus and rabies virus. Journal of General Virology 33, 213-224 • YAMADA,T, & HATANO, M. (I972). Lowered transplantability of cultured tumour cells by persistent infection with paramyxovirus. (HVJ) GANN: The Japanese Journal of Cancer Research 63, 647-655.
MINATO, N., BLOOM, B. R., JONES, C., HOLLAND, J. J. & REID, L. M.
(Received : 3 July I 9 7 8 )
