VIROLOGY

190, 86 l-865

(1992)

Synergism

between

Bovine Papillomavirus Type 4 and the Flavonoid in Cell Transformation in Vitro WILLIAM

The Beatson

Institute

D. PENNIE

M.

AND

SAVERIA

CAMPO’

for Cancer Research, CRC Beatson Laboratories, Garscube Glasgow G61 lBD, Scotland, United Kingdom Received

May

1, 1992;

accepted

Quercetin

June 23,

Estate,

Bearsden,

1992

Bovine papillomavirus type 4 (BPV-4) morphologically transforms primary bovine ceils in vitro only in the presence of an activated ras gene. The transformed cells are capable of anchorage-independent growth, but are not immortal and are incapable of inducing tumors in nude mice, suggesting that other events are needed to convert the cells to the fully transformed phenotype. We show here that treatment of the cells with a single dose of the flavonoid quercetin leads to full oncogenic transformation of cells transfected with BPV4 and ras. Quercetin is one of the most potent mutagens found in bracken fern, the environmental cofactor in BPV-Cassociated carcinogenesis of the upper alimentary canal of cattle. Our results point to quercetin as the probable in vivo cocarcinogen synergizing with BPV-4 in malignant progression. 0 1992 Academic Press, Inc.

wide spectrum of chromosomal abnormalities (9). In addition, quercetin activates protein kinases and competitively inhibits ATP binding by phosphatases (10). Quercetin alone does not act as a carcinogen (1 I- 13), but can act as an initiator in a two-stage transformation assay in mammalian cells in vitro, with TPA as a promoter (14). The effects of quercetin as an initiating agent in PalF was examined by treating cell cultures with a single dose of quercetin. Quercetin (Sigma) was dissolved in ethanol and added to the cells at final concentrations of 5, 20, and 45 PM. After 48 hr, the cells were washed and transfected by calcium phosphate precipitation (6) with recombinant pBPV4 (15) and/or pT24, which contains the activated ras gene (16). In all cases pZipneo (17) was cotransfected to allow selection of cells in G418 (6). The use of a single dose of quercetin in this assay ensured that any possible effects on cell morphology, growth characteristics, or tumorigenicity were due to long-term mutational damage and not to short-term phenomena such as protein kinase activation (10). Treatment of PalF with quercetin at the three different concentrations caused no observable changes in the morphology of cells transfected with no DNA, with pZipneo only, with pZipneo and pBPV4, or with pZipneo and pT24 (Fig. 1). Only small (~5 mm diameter) contact-inhibited G418-resistant colonies were observed, which could not be expanded and quickly senescenced, even when pooled (Table 1). Thus, quercetin does not induce any significant long-term changes in cell morphology even in the presence of

Bovine papillomavirus type 4 (BPV-4) infects the mucous epithelium of the alimentary canal of cattle and induces epithelial papillomas (I), which can progress to malignancy in cattle feeding on bracken fern (2). Bracken is known to contain both immunosuppressants (3) and mutagens (4) and the first effects of a bracken diet are chronic immunosuppression and widespread persistent papillomatosis of the alimentary canal (5). Following one or more decades, squamous cell carcinomas develop from papillomas in approximately 30% of the affected animals (2). The cooperation between BPV-4 and bracken has been experimentally reproduced (5), but the molecular mechanisms of the synergism between virus and chemicals are still unclear. We have recently reported that BPV-4 is incapable of transforming primary bovine fibroblasts (PalF) and partial morphological transformation is achieved only in cooperation with an activated ras gene (6). The partially transformed PalF have acquired an extended life span compared to parental cells but are not immortal, are anchorage independent, but not tumorigenic in nude mice, indicating that additional events are needed for full transformation. One of the most potent mutagens in bracken fern is the 5,7,3’,4’-tetrahydroxyflavone quercetin (4). Quercetin binds DNA and induces a variety of genetic lesions in bacteria and cultured mammalian cells (see Ref. (7) for review), including clastogenic damage (8). This observation is of particular significance as bracken-grazing cattle have a

’ To whom dressed.

correspondence

and reprint

requests

should

be ad-

861

0042.6822/92

$5.00

Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

862

SHORT

COMMUNICATIONS

FIG. 1. Comparison of colony morphology and growth in methocel between quercetrn-treated and nontreated cells transformed by BPV-4/ras. (A) Control PalF cells, (B) PalF cells treated with 20 NM quercetin; (C) PalF cells transformed by BPV-4/ras; (D) 20 PLM quercetin-treated PalF cells transformed by BPV-4/ras, all cells were G418 selected; (E) growth of BPV-4/rastransformed cells in methocel, (F) growth of 20 pM quercetintreated BPV-4/ras-transformed PalF cells in methocel. Magnification is 60X in all cases.

BPV-4 or activated ras. Neither BPV-4 alone nor r-as alone is capable of inducing morphological transformation of PalF (6) and this limitation is clearly not overcome by exposure of the cells to quercetin. Cotransfection of PalF with pBPV4, pT24, and pZipneo together gave rise to morphologically transformed colonies, as previously described (6). No significant difference was observed in the number of G41 a-resistant colonies obtained following transfection of cultures treated with 5, 20, or 45 FM quercetin, although numbers were generally higher in Experiments 1 and 2 (Table 1). Thus, quercetin does not enhance the overall transformation efficiency of BPV-4/ras and both BPV-4 and ras are necessary for transformation, even in the

presence of an initiator. However, qualitative differences were noticeable in cells exposed to the higher quercetin concentrations. While there was no detectable difference in morphology between nontreated cells and cells treated with 5 pM quercetin (not shown), in cells exposed to 20 or 45 yM quercetin, approximately 3-5% of the colonies were strikingly different from the rest (Table 1). The cells were very elongated and refractile, were not contact inhibited, and adopted a “crisscross” morphology, piling up to form dense foci (Fig. 1). The colonies produced in each transfection in experiment 3 were pooled and the different pools were tested for anchorage-independent growth in methocel

SHORT TABLE

1

EFFECT OF QUERCETIN ON COLONY

FORMATION

None pZipne0 pBPV4 pT24 + pBPV4 (Expt. pBPV4 (Expt. pBPV4 (Expt.

OfiM

+ pZipneo pZipneo + pT24 + pZipneo 1) + pT24 + pZipneo 2) + pT24 + pZipneo 3)

5PM

0 0 0 0

BY BPV-4

AND ras

20 pM

45 pM

0 0 0 0

0 0 0 0

34 (0)

nd

45 (3)

33 (2)

nd

nd

46 (1)

27 (0)

17

18

20

ivofe. Colony formation was assessed by G418 resistance assay: only colonies of >5 mm and with a transformed appearance were scored. Numbers represent total colonies: numbers in parentheses are colonies with a highly transformed morphology. These were not scored in Experiment 3. nd, not done.

and tumorigenicity in nude mice. There were no differences in anchorage-independent growth between nontreated cells and cells treated with 5 PM quercetin, in agreement with the absence of morphological differences between these two cell populations (Table 2). TABLE

2

EFFECT OF QUERCETIN TREATMENT ON ANCHORAGE INDEPENDENCE IN PalF CELLS Quercetrn DNA

(44

0 0 5 5 20 20 45 45 45 45

NIH-BPV2 pZipne0 pBV4 + pT24 + pZipne0 pZipne0 pBV4 + pT24 + pZipne0 pZipne0 pBV4 + pT24 + pZipne0 pZipne0 pBV4 + pZipneo pT24 + pZipneo pBV4 + pT24 + pZipne0

Growth in methocel

Efficiency

Positive Negative Positive

4.41

(t0.6)

Negative Positive

4.08

(t0.32)

X 1 O-4

Negative Positive

6.95

(kO.45)

X 1 O-4

Negative Negative Negative Positive

1.54 (f0.21)

x 1 o-4 x 1O-4

-

6.75

(~0.15)

X 1O-4

/Vote. NIH-BPV2 are NIH-3T3 cells transformed by BPV-2 [18]. Efficiency of methocel colony formation was determined by plating 1 O6 cells in duplicate in a 60.cm’ petri dish and incubating for 7-10 days. Each plate was scored for colonies by counting six l-cm’ areas from each plate, averaging this number and then multiplying by 60 to give an estimate of total numbers of colonies. The duplicates were then averaged and the result is given as number of colonies per 10” cells. i, Standard deviation.

3

EFFECT OF QUERCETIN TREATMENT ON TUMORIGENICITY Quercetin (44

of quercetin

0 0 0 0

17

863 TABLE

Concentration DNA

COMMUNICATIONS

0 5 20 20 20 45 45 45

DNA pBPV4 + pT24 + pBPV4 + pT24 + pBPV4 + pZipneo pT24 + pZipneo pBPV4 + pT24 + pBPV4 t pZipneo pT24 + pZipneo pBPV4 + pT24 + NIH-BPV2

Note. Three mice were used for each were subcutaneously injected into each Table 2.

Tumors pZipneo pZipneo

pZipneo

pZipneo

o/3 o/3 o/3 o/3 313 o/3 o/3 313 313

batch of cells and 1 O7 cells mouse. NIH-BPV2 are as in

In contrast, cells exposed to the higher quercetin doses grew in methocel with 50% higher efficiency than nontreated cells (Table 2), and the resultant colonies were of a much larger size (Fig. 1). It has to be noted that irrespective of the presence of quercetin, transformed PalF grew in methocel with a higher efficiency than NIH3T3 cells transformed by BPV-2 (18). Quercetin-treated BPV-4lrastransformed PalF were subcutaneously injected into nude mice. The cells which had been exposed to the higher doses of the initiator induced tumors approximately 4 weeks after injection (Table 3). Transformation of primary cells by papillomavirus does not lead immediately to tumorigenicity (19~21), and additional steps are required to achieve tumor induction (22, 23), pointing to the critical role of other events in the oncogenic progression (see Ref. (7) for review). In earlier reports, tumorigenicity was achieved by continuously culturing human papillomavirus type 18 (HPV-18)immortalized cells for several passages (23) or by introducing activated ras into cells previously immortalized by HPV-16 (22). In our system quercetin-treated cells rapidly converted to malignancy, emphasizing the synergism between the chemical, BPV-4, and the r-as oncogene. The tumors initially grew to a larger size than those induced by the highly tumorigenic BPV-2-transformed NIH3T3 cells (18) attaining a maximum size of approximately 2 cm in diameter by Weeks 6 to 7. Afterward, the tumors regressed to a considerable extent (approximately 0.5 cm), although they were still present at 12-l 4 weeks when they were excised. This partial tumor regression implies that either the tumors were susceptible to the vestigial immune response of the nude mice and/orthe injected cells lost their capacity for growth while in the animal. However, both the original cells and the cells explanted from the nude mice tumors had a high ca-

864

SHORT

COMMUNICATIONS

pacity for growth in vitro, having been passaged weekly for several months. During this period they showed no reduction in growth rate, whereas PalF similarly transfected but without quercetin treatment senesced. The pronounced morphological differences, the enhanced anchorage independence, and the tumorigenicity of the quercetin-treated cells indicate that exposure to quercetin, albeit transient, gives rise to a population of cells which can be transformed to a higher degree by BPV-4 and ras. Quercetin therefore acts as an initiator also in our system as had previously been reported for mouse cells (14). For quercetin to exert its effect, it needs to be administered before the promoter: the addition of quercetin to cells after initiation with methylcholanthrene and promotion with 12-O-tetradecanoylphorbol-13-acetate (TPA) inhibited the promoting effect of TPA (14). We do not know whether quercetin treatment of bovine cells after transfection with BPV-4 and ras would lead to the same results. Quercetin induces chromosomal damage (8) and cattle feeding on bracken fern present with chromosomal abnormalities (9). The karyotype of quercetintreated and control cells was examined, but no chromosomal abnormality was apparent in either group of cells (R. Fries and S. Solinas, unpublished observations). It should be noted, however, that for quercetin to act as a clastogenic agent, considerably higher doses are needed than utilized in these studies (8). In conclusion, in our in vitro system, which exploits the ability of BPV-4 to partially transform primary bovine cells in cooperation with ras, quercetin acts as an initiator, allowing full conversion of cells to a malignant phenotype. This strongly supports the hypothesis that quercetin is a natural initiator of cancer in bracken-eating cattle. BPV-4 appears to provide the necessary proliferative stimulus, possibly by binding plOSRb (6), which leads to expansion of the initiated cells, and subsequent events, typified by ras mutations (24), convert cells to malignancy, in agreement with the requirement for activated ras in in vitro transformation. In our experiments exposure to quercetin in vitro was transient; therefore, it is most likely that quercetin had acted at the genome level by introducing mutations, as has been shown in bacteria where quercetin induces point mutations by intercalating with the DNA (25). It has to be kept in mind, however, that cattle feeding on bracken fern are exposed to quercetin and other potentially mutagenic chemicals for a long time, therefore increasing the chance of mutations or other epigenetic events such as activation of protein kinases and inhibition of phosphatases (10). Our current efforts are directed at defining the critical changes induced by quer-

cetin which allow the tumorigenic BPV-4-transformed primary cells.

transformation

of

ACKNOWLEDGMENTS Thanks are due to all our colleagues the development of this work; to Dr. Wyke and W. F. H. Jarrett for critical the Cancer Research Campaign for the recipient of a CRC studentship

for helpful discussions during M. Jackson and Professors .I. A. reading of the manuscript and to continuous funding. W.D.P. was and M.S.C. is a CRC Fellow.

REFERENCES 7. CAMPO, M. S., MOAR, M. H., JARREIT, W. F. H., and LAIRD, H. M., A new papillomavirus associated with alimentary tract cancer in cattle. Nature 286, 180-l 82 (1980). 2. JARRETT, W. F. H., MCNEIL, P. E., GRIMSHAW, W. T. R., SELMAN, I. E., and MCINTYRE, W. I. M., High incidence area of cattle cancer with a possible interaction between an environmental carcinogen and a papilloma virus. Nature 274, 215-217 (1978). 3. EVANS, I. A., PROROK, J. H., COLE, R. C., AL-SAMANI, M. H., AL-SAMARRAI, A. M., PATEL, M. C., and SMITH, R. M. N., The carcinogenic, mutagenic and teratogenic toxicity of bracken. Proc. R. Sot. Edinburgh 81, 65-78 (1982). 4. EVANS, W. C., PATEL, M. C., and KOOHY, Y., Acute bracken poisoning in homogastric and ruminant animals. Proc. R. Sot. Edinburgh 81, 29-64 (1982). 5. CAMPO, M. S., and JARRETT, W. F. H., Papillomavirus infection in cattle: Viral and chemical cofactors in naturally occurring and experimentally induced tumours. In “Papillomaviruses, CIBA Foundation Symposium” (D. Evered and S. Clark, Ed.), Vol. 120, pp. 1 17-l 31. Wiley, New YorkIChichester, 1986. 6. JAGGAR, R. T., PENNIE, W. D., SMITH, K. T., JACKSON, M. E., and CAMPO, M. S., Co-operation between bovine papillomavirus type 4 and ras in the morphological transformation of primary bovine fibroblasts. /. Gen. Viral. 71, 3041-3046 (1990). 7. JACKSON, M. E., CAMPO, M. S., and GAIJKROGER, J. M., Coperation between papillomavirus and chemical cofactors in carcinogenesis. Crif. Rev. Oncogenesis, 4(3) (1993). 8. ISHIDATE, M., “Data Book on Chromosomal Aberration Tests in Vitro.” Elsevier, Amsterdam/New York, 1988. 9. MOURA, J. W., STOCCO DOS SANTOS, R. C., DAGLI, M. L. Z., D’ANGELINO, J. L., BIRGEL, E. H., and BECAK, W., Chromosome abberations in cattle raised in bracken fern pasture. Experientia 44,785-788 (1988). 70. VAN WART-HOOD, J., LINDER, M. E., and BURR, J. G., TPCK and quercetin act synergisticallywith vandate to increase proteintyrosine phosphorylation in avian cells. Oncogene 4, 12671271 (1989). II. AMBROSE, A. M., ROBBINS, D. J., and DEEDS, F., Comparative toxicities of quercetin and quercetrin. /. Am. Pharmacol. Assoc. 41,119-122 (1951). 12. HIRONO, I., OGINO, H., FUJIMOTO, M., YAMADA, K., YOSHIDA, Y., IKAGAWA, M., and OKUMARA, M., Induction of tumours in ACI rats given a diet containing ptaquiloside, a bracken carcinogen. J. Nafl. Cancer inst. 79, 1143-l 149 (1987). 13. MORINO, K., MATSUKURA, N., KAWACHI, T., OHGAKI, H., SUGIMURA, T., and HIRONO, I., Carcinogenicity test of quercetin and rutin in golden hamsters by oral administration. Carcinogenesis 3, 93-97 (1981). 14. SAKAI, A., SASASKI, K., MIZUSAWA, H., and ISHIDATE, M., Effects of quercetin, a plant flavonol on two-stage transformation in vi-

SHORT

15.

76.

77.

18.

79.

20.

COMMUNICATIONS

tro. Teratogenesis, Carcinogenesis, Mutagenesis 10, 333340 (1990). CAMPO, M. S., and COGGINS, L. W., Molecular cloning of bovine papillomavirus genomes and comparison of their sequence homologies by heteroduplex mapping. J. Gen. L’irol. 63, 255264 (1982). SANTOS, E., TRONICK, S. R., AARONSON, S. A., PIJLCIANI, S., and BARBACID, M., T24 human bladder oncogene is an activated form of the normal homologue of BALE- and Harvey-MSV transforming genes. lvature 298, 343-347 (1982). CEPKO, C. L., ROBERTS, B. E., and MULLIGAN, R. C., Construction and applications of a highly transmissible murine retrovirus shuttle vector. Cell 37, 1053-l 062 (1984). CAMPO, M. S.. and SPANDIDOS, D. A., Molecularly cloned bovine papillomavirus DNA transforms mouse fibroblasts in virro. J. Gen. Viral. 64, 549-557 (1983). CERNI, C., BINETRUY, B.. SCHILLER, 1. T., Lowv, D. R., MENEGUZZI G., and CUZIN, F., Successive steps in the process of immortalization identified by transfer of separate bovine papillomavirus genes into rat fibroblasts. Proc. Nat/. Acad. Sci. USA 86, 3266-3270 (1989). MATLASHEWSKI, G., OSBORN, K., BANKS, L., STANLEY, M., and CRAWFORD, L., Transformation of primary human fibroblast

27.

22.

23.

24.

25.

865 cells with human papillomavirus type 16 DNA and EJ-ras. ht. J. Cancer 42, 232-238 (1988). PIRISI, L., YASUMOTO, S., FELLER, M., DONIGER, J., and DIPAOLO, 1. A., Transformation of primary human fibroblasts and keratinocytes with human papillomavirus type 16 DNA. J. tirol. 61, 1061-1066 (1987). DIPAOLO, J. A., WOODWORTH, C. D., POPESCU, N. C.. NOTARIO, V., and DONIGER, J., Induction of human cervical squamous cell carcinoma by sequential transfection with human papillomavirus 16 DNA and viral Harvey ras. Oncogene 4, 395-399 (1989). HURLIN, P. H., KAUR, P., SMTH, P. P., PEREZ-REYES, N., BLANTON, R. A., and MCDOUGALL, J. K., Progression of human papillomavirus type 18.immortalized human keratinocytes to a malignant phenotype. Proc. Natl. Acad. Sci. USA 88, 570-574 (1991). CAMPO, M. S., MCCAFFERY, R. E., DOHERTY, I., KENNEDY, I. M., and JARRED, W. F. H., The Harvey r-as gene 1 is activated in papillomavirus-associated carcinomas of the upper alimentary canal in cattle. Oncogene 5, 303-308 (1990). RHAMAN, A., SHAHABUDDIN, HADI, S. M., and PARISH, J. H., Complexes involving quercetin, DNA and Cu(ll). Carcinogenesis 11,2001-2003 (1990).