METHOTREXATE RESISTANCE IN HUMAN UROEPITHELIAL CELLS WITH p53 ALTERATIONS

0022-5347/98/1592-0581$03.00/0 Vol. 159,581485, February 1998 Printed in U S A .

THE JOURNAL OF UR0U)GY

Copyright 0 1998 by AMERICAN UROLOCICAL ASS~CIATION, INC.

METHOTREXATE RESISTANCE IN HUMAN UROEPITHELIAL CELLS WITH p53 ALTERATIONS THOMAS R. YEAGER AND CATHERINE A. REZNIKOFF* From the University of Wisconsin Comprehensive Cancer Center, Department of Human Oncology, and the Program in Cell and Molecular Biology, Madison, Wisconsin

ABSTRACT

Purpose: Bladder cancers are frequently treated with combination chemotherapy that includes methotrexate (MTX). The development of drug resistance is a common problem in treatment of bladder cancers. We tested if the status of p53 and/or pRb affects the development of M l X resistance in bladder epithelial cell lines. Materials and Methods: We used two isogeneic sets of cell lines in which we manipulated the status of p53 and/or pRb by transformation with Human Papillomavirus (Hpv) E6 and/or E7 or with a transdominant TP53 mutant (TP53143).One series of isogeneic origin was derived from normal human uroepithelial cells (HUC), and the other was derived from a human transitional cell carcinoma (TCC). Cell lines with p53 and/or pRb alterations were cultured for six months while increasing the MTX concentration in each line, as resistance developed. Results: Two cell lines with both pRb and p53 alterations, aE6LE7-HUC and aE7-HUC$3" acquired the greatest resistance (750 nM) to MTX. One line with p53 loss, EG-TCC#lO, acquid intermediate resistance (500 nM), while two lines, aE7-HUC and E7-TCC#10, with altered pRb but wildtype p53, showed low levels of MTX resistance (125 nM and 80 nM, respectively). Two clear mechanisms of MTX resistance were identified. All five MTX resistant cell lines showed altered uptake of MTX. In addition, two of five MTX resistant cell lines, both with altered p53, showed dihydrofolate reductase (DHFR) amplification. Conclusions: p53 alteration increases the risk for development of drug resistance by both DHFR amplification and altered MTX transport in transformed human bladder epithelial cell lines. KEY WORDS:methotrexate resistance, p53 alterations,human uroepithelial cells

MTX resistance can develop through several mechanisms including; gene amplification and changes in either drug uptake or retention.3 Drug resistance has also been associated with increase in levels of two anti-apoptotic proteins, BCL-2 and BCL-X.6 Tumor cells, but not normal cells have been shown to develop drug resistance by gene amplification when exposed to genotoxic agents in vitro.6.7 The development of drug resistance has been associated with wildtype (wt) p53 loss in some systems. For example, several in vitro studies using human or mouse fibroblasts or hematopoietic cells show an association between loss of wt p53 function and amplification of genes such as dihydrofolate reductase (DHFR) and CAD that cause methotrexate (MTX) and phosphonacetyl-L-aspartate (PALA) resistance, respedive1y.S-11 In addition to p53 loss, the loss of p16 may impact positively on the development of drug resistance.12 Loss of wt p53 function is the most common genetic alteration identified to date in TCCs and associates with high grade, late stage and progressed disease.13 Two other genetic changes commonly observed in TCCs are pRb and p16 inactivations.13 Therefore, it would be of clinical relevance to determine if alterations in p53, pRb, and/or p16 impact on the development of drug resistance during tumorigenesis of human bladder epithelium. To accomplish this goal, we exa m i n e d two independent isogeneic sets of bladder cells lines. The first set were immortal human uroepithelial cells ((rHUC)transformed by Human Papillomavirus 16 E6 or E7 or a transdominant TP53 mutant, Tp53143.The second set Accepted for publication August 20, 1997. * Requests for reprints: Department of Human Oncology, Univer- was derived from a spontaneously immortalized human TCC, sity of Wisconsin Comprehensive Cancer Center, Madison, WI TCC#10, which was subsequently transformed by Hw E6 or 53792. Supported b NIH grants ROlCA29525-16 and ROlCA67158-01 E7.These lines represented cells with p53 and/or pRb alterations. In addition, some of these cell h e s contained altered and molecularbsciences training grant 5 "32 GM07215.

Bladder cancer is currently the sixth most common cancer and the third most prevalent cancer in men over sixty in the USA. Bladder cancer is three to four times more common in men than women. At present, an estimated 12,000 bladder cancer deaths occur annually in the USA.' Bladder cancers are heterogeneous with respect to their histopathologic classification, morphological growth patterns, biologic potentials for progression, and responses to clinical treatment protocols. In the USA, approximately 90% of bladder cancers are transitional cell carcinomas (TCCs). There are two distinct morphologic growth patterns for TCCs. 80-90% of superficial TCCs grow as papillary lesions, the remainder grow as flat TCC or carcinoma in situ (CIS). Papillary TCC frequently recur, but infrequently progress to myoinvasive tumors.2 Superficial papillary tumors have a 5 year survival rate greater than 90%,but this decreases to 49% and 6% for regional or distant disease, respedively.1 The poor survival rate of invasive TCC requires the implementation of aggressive treatment regimens that can include cystectomy, chemotherapy, radiotherapy or combinations of these modalities. A common combination chemotherapeutic protocol, (M-VAC) includes methotrexate, vinblastine, adriamycin and ~is-platin.~ Even with this potent combination of drugs, the response rate for myoinvasive TCC usually does not exceed 40%, and there are currently no markers to predict which individuals will fail treatment.4

581

582

METHOTREXATE RESISTANCE IN UROTHELIAL CELLS

p16. The results presented in this communication are consistent with a model in which p53 inactivation, but not pRb or p16 loss, plays a significant role in determining the probability of development of significant MTX resistance in bladder epithelial cells.

MATERIALS AND METHODS

Cell lines and cell culture. The 5637 and RT-4 bladder cancer cell lines were used as controls for high and low grade bladder cancer cells, respectively. The Hela S3 cervical cancer cell line was used as a positive control for cells that develop MTX resistance.14 The 5637, RT-4 and Hela S3 cell lines were received from the American Type Culture Collection. These three control lines were grown in RPMI medium (Sigma, St. Louis, Mo) with 10%dialyzed fetal bovine serum (FBS), 100 unitdml. penicillin, 100 unitdml. streptomycin, 2.7 mglml. dextrose, and 2 mM Gglutamine (Sigma). Two independent sets of cell lines of isogeneic origin established and characterized in our laboratory were used to address the primary question in this research. The three lines in the first set were all derived from one normal tissue source, namely aHUC. In this set, the p53 andor pRb status were manipulated by transformation with HPV 16 E6 andor E7 to generate aE7-HUC and aE6/E7-HUC.16 aE7HUCp63"" was generated by transformation of the immortal aE7-HUC line using a retroviral vector carrying a transdominant TP53 mutant (codon 143). The cell lines in the second set were also of isogeneic origin, all being derived h m one clinical biopsy of a T3 Grade I1 TCC (TCC#lO). TCC#10 spontaneously immortalized. The p53 and pRb status of the parental immortal TCC#10 were determined to be wildtype. These were manipulated for these experiments as above using HPV16 E6 or E7. The aHUC series and the TCC#10 series were both grown in RPMI medium as above with 1% dialyzed FBS, with growth supplements described for human uroepithelial cells added.16 Cells were grown on 100 111111. plastic petri dishes at 37C in a 5%CO, incubator. Western analysis. Exponentially growing cells were collected for western lysates using ECB lysis buffer (50 mM Tris [pH 8-01, 120 mM NaC1, 0.5% Nonidet P-40, 100 mM NaF, 200 p,M Na,VO,, 1 mM phenylmetylsulfonyl fluoride, 100 units of aprotiuin/ml. and 10 pg. of leupeptidml.) (Sigma). Protein lysates were quantified using the Bradford assay usually with 50 pg. of protein loaded per lane on a SDSPAGE gel. The p53 and pRb westerns were run on 7.5% gels, while the p16 western was run on a 15% gel. Gels were transferred to nitrocellulose and stained with ponseau S (Sigma) to determine the equality of loading. The proteins were detected using chemiluminescence (Pierce, Rockford, EL). p53 was detected using a monoclonal antibody (Ab-2, Oncogene Science, Cambridge, MA), pRb was detected using a monoclonal antibody (#14001 A, Pharmingen, San Diego, CAI, and p16 was detected using a polyclonal antibody (C-20, Santa Cruise Biotechnology, Santa Cruise, CAI. MTX resistance. Each cell culture was split into two groups, one for MTX treatment and one as an untreated control. Except for the MTX treatment, controls and the experimental groups were cultured using the same protocol for the six month duration of these experiments. The MTX (Sigma) concentrations used ranged from 5-1000 nM. The treatment groups were initially exposed to 5 nM MTX, a concentration shown to be nontoxic on all the cell lines in cytotoxicity studies. Then, the selection for cells that developed MTX resistance was done using a 'shallow" stepwise Specifically,the M I X concentrationwas gradually increased as cells were passaged, but only with cultures that showed resistance to the lower concentration. Thus, the concentration was increased faster in some cultures than in other cultures. The concentrations used during the experi-

ment were: 5, 10, 20, 30,40,60, 80, 100, 125, 150, 200, 250, 300,400,500, 750, and 1000 nm MTX. Southern analysis. DNA was collected for Southern analysis of DHFR copy number following the six month continuous treatment with increasing concentrations of MTX. DNA was collected from both the MTX treated and untreated control of each experimental group. 10 pg. of DNA was digested with EcoRI and run out on an 0.8% agarose gel. The DHFR copy number was detected using the 1.1Kb ApaI-EcoRI fragment from the pDHFR-1.8 plasmid.14 Actin was used as a loading control and as a reference point to calculate DHFR amplification levels. The probes were labeled fluorescently by random priming (Amersham, Arlington Heights, IL) and DNA was detected using a fluorimager (MolecularDynamics, Fluorimager SI). Flow cytometry. The cells growing in growth medium were treated for eight hours with 2 pM fluorescent-labeled MTX (F-MTX) (Molecular Probes,Eugene, OR) while in log phase growth. Then the cells were treated with 0.1 /J.M MTX for three hours.17 The cells were then washed with PBS, trypsinized,and resuspended in PBS with 1%FBS and analyzed on a Becton Dikinson FACScan cell sorter. The cells were labeled with propidium iodide (Sigma) and only the gaited, live cell population was analyzed. RESULTS

Alterations of p53 andlor pRb in isogeneic cell sets. The three control cell lines and the two isogeneic sets of uroepithelial cells to be used in this study were characterized by western analysis to confirm for each line, the expected alterations in three tumor suppressor genes, pRb, p53 and p16 (fig. 1,table). 5637 showed no pRb signal consistent with pRb loss and contained a stabilized p53 consistent with the two known mutations in TP53 (codons 72 and 28O).lS p16 was stabilized consistent with our findings that bladder cancer lines with pRb loss show elevated ~16.19920The RT-4 cell line showed signals on western consistent with both a w t TP53 and RB, and the p16 signal was lost, again as expected for a cell line with wt pRb. The HeLa S3, which was used as a control because it was able to become MTX resistant in response to selection,l3 contains integrated HPV18. Hela S3

Controls

TCC#10

a-HUC

n n 3

p53 b

p16 b FIG. 1. Western blot analysis demonstrating status of pRb, p53 and p16 in three sets of human uroepithelial cell lines, control, TCC#lO and aHUC, were used in these studies. All lanes were loaded with 50 pg., except p53 lanes for 5637 and aE7-HUCp53m" which were loaded with 10 pg. because of the stabilized signal. Results were reproduced at least twice using independent samples.

583

METHOTREXATE RESISTANCE IN UROTHELIAL CELLS Effect of p53 and pRb status on development of MTX resistance' Sets

Alterations in

Cell Line P53

PRb

~ 1 6

Final MTX Resistance (nM)

DHFR Amplification

Altered MTX Uptake

N.D. RT-4 -p16 30 HeLa -p53 -Rb 400 -25 copies Yes -Rb 5637 -p53 Yes lo00 S-TCC#lO -p16 30 None I1 Ei'-TCC#lO -Rb -p16 80 N.D. EG-TCC#lO -p53 -p16 500 Yes aE7-HUC -Rb 125 N.D. I11 aE7-HUCp63m" -p53 -Rb 750 -6 copies Yes -Rb 750 Yes aE6/E7-HUC -p53 * Western blot analysis was used to determine the status of p53, pRb, and p16. Data are shown in figure 1. The growth kineties in development of Ml'X resistance during the six month period of these experiments are shown in figure 2. Representative data for DHFFt amplification are shown in figure 3, and representative data for altered MTX uptake are shown in figure 4. The data summarized in this table represent the findings from experiments that were reproduced at least three times with the exception of figure 2. These six month experiments were done once. N.D. stands for not determined. I

shows the expected phenotype for an HPV positive cell lines, six-fold DHFR amplification compared to the untreated conwhich is p53 degradatiodoss, decreased pRb signal and sta- trol. None of the TCC series, including the p53 negative TCC bilized ~ 1 6 . ~ ~ 7 ~ 0 that showed resistance to 500 nM MTX, had DHFR amplifiThe TCC#10 parental spontaneously immortalized cell line cation (data not shown). showed a wt pRb and wt p53, but CDKNVpl6 was mutated Changes in cellular uptake in some MTX resistance cells. with loss of detectable p16 signal.21 E7-TCC#10 in which With three of the highly MTX resistant lines not amplifying pRb, but not p53, was manipulated using HPV16 genes, DHFR, we tested if altered uptake of the MTX in the cell lines showed a normal p53 signal. In contrast, EG-TCC#lO showed was the mechanism underlying resistance. Flow cytometry wt pRb signal and the expected p53 degradation (fig. 1 and was used to determine if there was a differential uptake (or table). The aE7-HUC, which was also generated by HPV16 retention) of F-MTX. Both the MTX resistant and the unE7 transformation, showed a degraded pRb signal, while the treated controls for 5637, HeLa, EG-TCC#lO, S-TCC#lO, aE7-HUC transformed with a transdominant TP53 mutant aE7-HUCpS3"" and aEG/E7-HUC were challenged with 2 pM showed both a degraded pRb and stabilized p53. aE7-HUC F-MTX and then 0.1 p M MTX. Following exposure to the transformed by E6 and E7 showed the expected p53 loss and F-MTXthe MTX resistant and control lines were examined pRb degradation. All the aE7 series showed elevated p16 as by flow cytometry for the retention of the F-MTX. It was expected.19.20 found that all of the resistant lines retained the F-MTX, Association of MTX resistance with p53 alteration. To de- while the untreated control lines did not retain the F-MTX termine the effects of losses of pRb, p53, and/or p16 on the (table). The F-MTX is taken up via a passive mechanism, development of MTX resistance, three sets of cell lines (con- while the unlabeled MTX that can compete with the F-MTX trols, TCCs and HUCs of isogeneic origin) were treated with is taken up by an active mechanism. Lines with impaired MTX continuously for six months in vitro using a shallow MTX uptake retain the F-MTX, while normal cells do not retain the F-MTX in this assay.17 Representative MTX u p stepwise selection protocol starting with 5 nM. The positive control cell line for development of MTX re- take data is shown for S-TCC#lO, which showed minimal sistance, Hela S3, acquired resistance to MTX, 400 nM at six MTX resistance, no DHFR amplification, and no alteration in months (fig. 2, A ) as expected from previous reports.14 The transport and for EG-TCC#lO, which showed significant MTX RT-4 from a low grade bladder cancer with wt pRb and wt resistance, no DHFR amplification, but altered MTX transp53 acquired minimal resistance, while 5637 containing two port (fig. 4, A and B). TP53 mutations and an RB mutation acquired the highest level resistance seen in this study (fig. 2, A, table). Thus, in DISCUSSION the control set the lines with both pRb and p53 acquired the We report here the development of MTX resistance in HUC highest resistance. Loss of p l 6 appeared not to potentiate the and TCC lines and show that high level MTX resistance development of MTX resistance. Similar results were obtained using the two sets (the correlates with loss of wt p53 function, but not with either TCC#10 series and the aE7-HUC series) of cell lines of iso- p16 or pRb loss. Furthermore, we demonstrate several mechgeneic origin (fig. 2, B and C ) .Neither the loss of p16 nor pRb, anisms, including gene amplification and altered transport, alone or in combination, potentiated the development of sig- by which HUC and TCC develop MTX resistance. This is the first demonstration of the development of MTX nificant MTX resistance. In contrast, the loss of p53 in the TCC#10 series alone led to high level resistance (500 nM) resistance using a human epithelial cell culture system in when compared to that in cell lines with only pRb loss (80 which the status of p53 and pRb are known. Furthermore, nM) (fig. 2, B). In the aE7-HUC series, p53 loss, whether by this study is unique in that p53 and pRb were manipulated, E6 degradation or by transformation with a transdominant using HPV16 E6 or E7 or a transdominant TP53 mutant as TP53 mutant resulted in development of similar high level of molecular tools, in both cultured normal HUCs and TCCs, thus generating isogeneic sets. It is noteworthy that similar MTX resistance, i.e., 750 nM (fig. 2 , C ) . DHFR amplification in some MTX resistance cells lines. TO results were obtained using both the sets derived from the test if the resistance of the cell lines to MTX was due to normal and tumor tissues. Thus, these results support a amplification of DHFR, Southern analysis for DHFR copy model in which the acquisition of a "53 mutation, whenever number was assayed on all of the cell lines, both MTX treated it occurs during tumorigenesis whether as an early or late and untreated controls (fig. 2 and table). The Southern anal- step, sensitizes cells to the development of drug resisysis reproducibly showed that two of the resistant lines, Hela tance.8-10 One possible mechanism for the MTX resistance in 53 and aE7-HUCG3"" acquired amplified DHFR, as com- the p53 compromised lines may be due to alteration in the . tion of players pared to the untreated controls (fig. 3). Quantitative analysis cell's ability to undergo apoptosis. Examma using the phosphoimager showed approximately 25 fold involved in the apoptosis pathway, such as Bcl-2 and Bax are DHFR amplification in MTX-treated Hela S3 compared to currently under investigation in thew resistant cell lines. Our findings are wnsistent with those from several other the untreated control. aE7-HUCflsmUshowed on the average

-

584

METHOTREXATE RESISTANCE IN UROTHELW, CELLS

comb

-5637 +Hela +RT-4

800 --

F 6

g

600

.-w 2

400 - -

HUC RT4 M T X - +

-

Help

+

5637

-

+

'-

aHUCS

E7

E7P53m E6/E:

+

-

+

-

+

DHFR +

--

n

FIG. 3. Southern analysis of resistant and parental cell lines to test for DHFR amplification. All lanes were equalized using an actin control (data not shown). Results were quantified using three indendent samples using the phos hoimager Hela was determined to am lified about 25 fold, whife aE7-HUCp53mU was shown to be ampliffed approximately six fold relative to corresponding untreated control.

v) v)

0

8

4

12

16

20

24

A0 0

Weeks

w

* 1000 +EG-TCC#lO +E7-TCC#lO +S-TCC#1

800 E

fa

Control

m

0

13

0

600

CI In

3 E

-

....

(Y 0

400

=

200 0

4

0 0

4

8

12

16

20

24

Weeks 0

n

c

!i

aoo

1

I

&a-E7p53mu +a-E6/E7

0

600

-

EG-TCC#lO Control

m

....

iz 0

U v)

w 3

400

E

200

0

0

0

4

8

12

16

20

24

Weeks FIG. 2. MTX resistance ac uired in three sets of cell lines with altered p53 during the 6 mon&s of MTX selection in vitro including: A, control cell lines; B, TCC#lO lines, and C, aHUC lines. Note that cells with p53 loss acquired highest levels of MTX resistance.

groups who, using human or rodent cell models, showed a correlation between the development of MTX resistance and loss of p53 function.s-"JInterestingly, tumor cells with p53 alteration not only developed resistance to MTX, but also to radiation and certain chemotherapeutic drugs, including

RlHFlTC

FIG. 4. Flow cytometric data showing alterations in uptake of fluorescent labelled methotrexate. A, S-TCC#lO cells do not develop MTX resistance and there is no difference in uptake between control cells and cells exposed to MTX for six months. B , EG-TCC#lO, in contrast, develop resistance to 500 nM MTX and these MTX resistant cells clearly show altered MTX uptake compared to untreated control.

adriamycin.10 Similar to the findings of these other investigators, we also showed that normal cells fail to develop drug re~istance.~. Furthermore, we demonstrated that drug re-

METHOTREXATE RESISTANCE IN UROTHELIAL CELLS sistance can occur by several mechanisms, a n d at least in our study, altered transport w a s a more common mechanism than DHFR a m p l i f i ~ a t i o n . ~ The findings of an association between p53 loss and development of d r u g resistance is particularly significant in hum a n bladder cancers. First, TP53 mutation is the most frequent genetic alterations present in h u m a n bladder cancers and correlates with late stage high grade cancers.13 The survival rate for patients with late stage bladder cancer is poor, usually less than lo%.' Second, chemotherapeutic regimens for bladder cancers often include combinations containing MTX andlor adriamycin. Third, even with combination therapy, approximately 60% fail treatment.4 Thus, the development of multidrug resistance is a major clinical problem for physicians and for patients with bladder cancer. These studies therefore have potential translational value. The present d a t a suggest that bladder cancer patients with p53 mutations are at high risk of failing chemotherapeutic treatments that contain certain classes of drugs, including MTX a n d adriamycin. Thus, knowledge of the p53 status is important in planning treatment strategies for these patients. Clearly, alternate therapies, whose effectiveness is not affected by p53 status would be a more rational choice for these patients.

Acknowledgments. We would like to thank Kathleen Schell for her help with the flow cytometry. REFERENCES

1. American Cancer Society: Cancer Facts & Figures, 1996. 2. Thrasher, J. and Crawford, E. D.: Current management of invasive and metastatic transitional cell carcinoma of the bladder. J. Urol., 149: 957, 1993. 3. Jolivet, J., Cowan, K. H., Curt, G. A., Clendeninn, N. J. and Chabner, B. A.: The pharmacology and clinical use of methotrexate. N. Engl. J. Med., 309: 1094, 1983. 4. Bane, B. J., Rou, J. Y. and Hamstreet, G. P.: Pathology and staging of bladder cancer. Semin. Oncol., 23: 546, 1996. 5. Tu,Y. P., Xu, F. H., Liu, J., Vescio, R., Berenson, J., Fady, C. and Lichtenstein, A,: Upregulated expression of BCG2 in multiple myeloid cells induced by exposure to doxorubicin, ebposide and hydrogen peroxide. Blood, 88.1805, 1996. 6. Wright, J. A., Smith, H. S., Wat, F. M., Hancock, M. C., Hudson, D. L. and Stark, G. R.: DNA amplification is rare in normal human cells. Roc. Natl. Acad. Sci. USA, 87: 1791, 1990. 7. Tlsty, T. D.: Normal diploid human and rodent cells lack a detectable frequency of gene ampliikation. Roc. Natl. Acad. Sci. USA, 87: 3132, 1990. 8. Yin, Y., Tainsky, M. A., Bischoff, F. Z., Strong, L. C. and Wahl, G. M.: Wild-type p53 restores cell cycle control and inhibits gene amdification in cells with mutant 1153 alleles. Cell, 70: i37,199i. 9. Livingstone, L. R., White, A., Sprouse, J., Livanos, E., Jacks, T.

585

and Tlsty, T. D.: Altered cell cycle arrest and gene amplification potential accompany loss of wild-type p53. Cell, '70: 923, 1992. 10. Groker, E., Waltham, M., Kheradpour, A., Trippett, T., Mazumdar, M., Elisseyeff, Y., Schnieders, B., Steinhen, P., Tan, C., Berman, E. and Bertino, J. R.: Amplification of the dihydrofolate reductase gene is a mechanism of acquired resistance to methotrexate in patients with acute lymphoblastic leukemia and is correlated with p53 gene mutations. Blood, 86: 677, 1995. 11. Shinohara, N., Libert, M., Wedemeyer, G., Chang, J. H. C. and Grossman, H. B.: Evaluation of multiple drug resistance in human bladder cancer cell lines. J. Urol., IM): 505, 1993. 12. Stone, S., Dayananth, P. and Kamb, A.: Reversible, p16 mediated cell cycle arrest as a proteetion from chemotherapy. Cancer Res., 56: 3199, 1996. 13. Reznikoff, C. A., Belair, C. D., Yeager, T. R., Savelieva, E., Blelloch, R. H., Puthenveettil, J. A. and Cuthill, S.:A molecular model of human bladder cancer pathogenesis. Semin. Oncol., 23: 571, 1996. 14. Sharma, R. C. and Shimke, R. T.: The propensity for gene amplification: a comparison of protocols, cell lines, and selection agents. Mut. Res., 304:243, 1994. 15. Reznikoff, C., Belair, C., Savelieva, E., Zhai, Y., Heifer, K., Yeager, T., Thompson, K. J., DeVries, S., Bindley, C., Newton, M. A., Sekhon, G. and Waldman, F.:Long-term genomic stability and minimal genotypic and phenotypic alterations on HW16 E7-, but not E6-, immortalized human uroepithelial cells. Genes & Dev., 8: 2227, 1994. 16. Reznikoff, C. A., Loretz, L. J., Pesciotta, T. D., Oberley, T. D. and Ignjatovic, M. M.: Growth kinetics and differentiation in vitro of normal uroepithelial cells on collagen gel substrates in defined medium. J. Cell. Physiol., 131: 285, 1987. 17. Assarif, Y. G. and Shimke, R. T.: Identification of methotrexate transport deficiencyin mammalian cells using fluoresceniated methotrexate and flow cytometry. F'roc. Natl. Acad. Sci. USA, &I: 7154,1987. 18. Cooper, M. J., Haluschak, J. J., Johnson, D., Schwartz, S., Morrison, L. J., Lippa, M., Hatzivassiliou, G. and Tan, J.: p53 mutations in bladder cancer cell lines. Oncology Res., 6: 569, 1994. 19. Yeager, T., Stadler, W., Belair, C., Puthenveettil, J., Olopade, 0. and Reznikoff, C.: Increased p16 levels correlates with pRb alterations in human urothelial cells. Cancer Res., 55: 493, 1995. 20. Reznikoff, C. A., Yeager, T. R., Belair, C. D., Savelieva, E., Puthenveettil, J. A. and Stadler, W. M.: Elevated p16 at senescence and loss of p16 a t immortalization in human papillomavirus 16 E6, but not E7 transformed human uroepithelial cells. Cancer Res., 56.2886, 1996. 21. Yeager, T. R., DeVries, S., Jarrard, D. F., Nakada, S.Y., Kao, C., Moon, T. D., Bruskewitz, R., Stadler, W. M., Meisner, L. F., Gilchrist, K. W., Newton, M. A, Waldamn, F. M. and Reznikoff, C. R.: Overcoming cellular senescence in human cancer pathogenesis. Genes &-Dev. (In Press).