High Prevalence of Human Cytomegalovirus in Prostatic Intraepithelial Neoplasia and Prostatic Carcinoma

0022-5347/03/1703-0998/0 THE JOURNAL OF UROLOGY® Copyright © 2003 by AMERICAN UROLOGICAL ASSOCIATION

Vol. 170, 998 –1002, September 2003 Printed in U.S.A.

DOI: 10.1097/01.ju.0000080263.46164.97

HIGH PREVALENCE OF HUMAN CYTOMEGALOVIRUS IN PROSTATIC INTRAEPITHELIAL NEOPLASIA AND PROSTATIC CARCINOMA MINU SAMANTA, LUALHATI HARKINS, KATRIN KLEMM, WILLIAM J. BRITT AND CHARLES S. COBBS* From the Surgical Service, Veterans Affairs Medical Center (MS, LH, CSC) and Departments of Surgery (CSC), Pathology (KK) and Pediatrics (WJB), University of Alabama at Birmingham, Birmingham, Alabama

ABSTRACT

Purpose: Recent epidemiological data indicate that a history of increased exposure to sexually transmitted diseases is associated with an increased risk of prostate cancer. Human cytomegalovirus (HCMV) is a member of the herpesvirus family, is sexually transmitted in adults and can persistently infect prostatic epithelium in nonimmunocompromised hosts. Based on increased awareness of the oncogenic potential of this virus, we decided to reexplore the issue of whether HCMV might be involved in prostate cancer pathogenesis. Materials and Methods: Paraffin embedded biopsy specimens from 22 randomly selected patients with prostatic intraepithelial neoplasia (PIN) lesions and prostatic carcinoma were analyzed by immunohistochemistry, in situ hybridization, polymerase chain reaction and DNA sequencing to detect HCMV nucleic acids and determine whether HCMV gene products were specifically associated with neoplastic cells. Results: We detected HCMV proteins and/or nucleic acids in all 22 of the 22 preneoplastic and neoplastic prostate lesions evaluated. HCMV proteins were specifically and often highly expressed in basal cell hyperplasia and PIN lesions, and to a lesser degree in carcinoma cells. Results: To our knowledge these data demonstrate for the first time the specific localization of HCMV nucleic acids and proteins in a high percent of PIN and prostate carcinoma lesions, and raise the possibility that HCMV might contribute to the natural history of prostatic cancer. KEY WORDS: prostate, prostatic neoplasms, cytomegalovirus virus infections, sexually transmitted diseases

Prostate cancer is the most common male cancer in the United States as well as the Western world and the second leading cause of male cancer death in the United States.1 While the genetic and environmental factors responsible for the high incidence of prostate cancer are largely unknown, recent epidemiological studies indicate a significant association between prostate cancer incidence and increased exposure to sexually transmitted diseases, implying that a sexually transmissible agent or agents increase the risk of prostate cancer.2, 3 To our knowledge no specific infectious agents have been causally linked to prostate cancer to date. Herpesviruses (eg Epstein Barr virus and human herpesvirus-8) are currently implicated in the pathogenesis of several human malignancies.4 Human cytomegalovirus (HCMV) is a ␤-herpesvirus that persistently infects 50% to 90% of the adult American population and in immunocompromised patients epithelial lesions may occur in multiple organs, including the prostate.5, 6 In adults sexual contact is a major route of HCMV transmission and increased HCMV seropositivity rates in adults are associated with an increased history of sexually transmitted diseases.5 Recent data indicate that HCMV has multiple oncogenic properties and HCMV gene expression can promote mutagenesis, cell cycle progression, angiogenesis, cell invasion and immune evasion in cells.7, 8 Although several studies in the 1970s sought to establish a link between HCMV and

prostate cancer, no definitive causal association was determined9, 11 and to our knowledge this area of investigation was not pursued. We recently described the presence of HCMV nucleic acids and viral gene products in human gliomas12 and decided to use sensitive molecular techniques developed at our laboratory to reexamine the question of whether HCMV may be specifically associated with prostate cancer lesions. MATERIALS AND METHODS

Clinical samples. Formalin fixed, paraffin embedded surgical biopsy specimens of prostatic neoplastic lesions from nonimmunocompromised patients were randomly collected from the pathology archives of the Birmingham Veterans Affairs Hospital. A genitourinary pathologist (KK) reviewed all cases to confirm the diagnosis. Immunohistochemical analyses of paraffin sections. Paraffin sections were cut (4 ␮m), hydrated, digested with pepsin (37C and 4 minutes), subjected to antigen retrieval in citrate buffer (BioGenex, San Ramon, California), pH 6.0, (37C for 12 hours) and blocked for peroxidase (3% H2O2 for 12 minutes). They were then blocked with Fc receptor blocker for IE1-72, pp65 and Ki-67 monoclonal antibodies (20C for 10 minutes) or background buster for Innovex CMV cocktail and cytokeratin-20 mAbs (Innovex Biosciences, Richmond, California) (20C for 20 minutes). Subsequently they were incubated with mAbs specific for IE1-72 (BioGenex) (1:20), pp65 (Novocastra, Newcastle upon Tyne, United Kingdom) (1:40), a CMV cocktail (1:90) (Innovex Biosciences), Ki67 (1:20), cytokeratin 20 (BioGenex) (1:20) or no antibody in tris buffered saline (TBS), pH 7.6, with 0.05% Tween 20 (4C for 12 hours). Detection was performed using a horseradish peroxidase system (BioGenex) with the chromogen diaminobenzidine (Innovex Biosciences). A pathologist (KK) graded all

Accepted for publication March 14, 2003. Study received institutional review board approval. Supported by a Merit Award for Cancer Research by the United States Veterans Administration. * Corresponding author: Departments of Surgery and Cell Biology, University of Alabama at Birmingham Medical Center, Faculty Office Tower, Suite 1034B, 510 20th St. South, Birmingham, Alabama 35294 (telephone 205-934-1674; FAX: 205-975-6088; e-mail: [email protected]). 998

999

HUMAN CYTOMEGALOVIRUS AND PROSTATE CANCER IHC, ISH and UL73 PCR results in biopsy specimens of PIN and carcinoma lesions from 22 patients Pt No.

Diagnosis

Gleason Grade

IHC IE1

pp65

ISH Innovex

HCMV

HPV

UL73 PCR (fig. 3)

1 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ ⫹ 2 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ ⫺ ⫹ DNA band present 3 Normal epithelium, PIN ⫹ Ca 4⫹3 ⫹ ⫹ ⫹ 4 Normal epithelium ⫹ PIN 3⫹3 ⫹ DNA band present 5 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ ⫹ ⫹ 6 Normal epithelium ⫹ PIN No tumor ⫹ ⫹ 7 Normal epithelium, PIN ⫹ Ca 3⫹4 ⫹ ⫹ ⫹ DNA band present 8 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ ⫹ ⫹ ⫺ 9 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ ⫹ DNA band present 10 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ ⫹ ⫹ ⫺ DNA band present 11 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ ⫹ ⫹ 12 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ ⫹ ⫺ DNA band present 13 Normal epithelium ⫹ PIN No tumor ⫹ ⫹ ⫹ 14 Normal epithelium ⫹ PIN No tumor ⫹ ⫹ 15 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ ⫺ ⫹ 16 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ ⫹ 17 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ ⫺ DNA band present 18 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ 19 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ ⫹ 20 Normal epithelium ⫹ PIN No tumor ⫹ DNA band present 21 Normal epithelium ⫹ PIN No tumor ⫹ ⫹ 22 Normal epithelium, PIN ⫹ Ca 3⫹3 ⫹ ⫺ For IHC mAbs included antiIE1-72 (IE1), antipp65 (pp65) and Innovex HCMV cocktail (Innovex) and for ISH biotinylated probes specific for HCMV and HPV (negative control) were used.

specimens in blinded fashion. Specimens were scored as positive (cell specific immunoreactivity was present) or negative (none was present). In situ hybridization (ISH) of paraffin sections. To detect HCMV RNA we used a biotinylated 21 base oligonucleotide (5⬘-GTGGTGGCGCTGGGGGTGGCG-3⬘) specific for highly expressed HCMV early gene nucleic acids and a biotinylated positive (specific for poly-A mRNA) control. For negative controls we used a cocktail containing 6 different biotinylated oligonucleotide probes specific for different human papillomavirus (HPV) strains (InnoGenex, San Ramon, California) or no probe. Paraffin sections (4 ␮m) were dehydrated and digested with pepsin (Zymed Corp., South San Francisco, California) (37C for 20 minutes) and denatured (90C for 15 minutes). The slides were hybridized (37C for 12 hours). A DNA/mRNA alkaline phosphatase detection system with the chromogen nitroblue tetrazolium (BioGenex) was used. Infection of PC-3 cells in vitro, immunohistochemistry (IHC) and ISH. Human prostate carcinoma derived PC-3 cells (American Type Culture Collection, Rockville, Maryland) were cultured at 37C in 5% CO2 in Dulbecco’s modified Eagle’s medium with 10% fetal bovine serum (BioWhittaker, Walkersville, Maryland). Cells were infected (or mock infected) with HCMV Towne strain (American Type Culture Collection) by adding virus at a multiplicity of infection of 1. After 1 hour cells were washed with phosphate buffered saline and medium was re-added. After 48 hours cells were fixed with methanol (⫺20C for 20 minutes). For double immunostaining cells were blocked for endogenous peroxidase (3% H2O2 for 8 minutes), incubated in Fc receptor blocker (BioGenex) (20C for 10 minutes) and blotted dry. Primary mAb to IE1-72 (1:300) or pp65 (1:300) was added (4C for 12 hours). Slides were rinsed with TBS and detected with a horseradish peroxidase system using black chromogen for IE1-72 (Zymed Corp.) and brown for pp65 (Innovex Sciences). For Bcl-2 and proliferating cell nuclear antigen (BioGenex) double immunostaining (PCNA) to the IE1-72 immunostained cells was added mAb to Bcl-2 (Novocastra) (1:200) or PCNA (1:300) (4C for 12 hours). Detection was done with an alkaline phosphatase system using fast red chromogen. For ISH HCMV infected PC-3 cells were infected and fixed as described. HCMV or control probe was added as described, and the cells were denatured (90C for 15 minutes) and hybridized (37C for 90 minutes). Slides were soaked in TBS, probe wash (BioGenex) was added (10 minutes) and hybridization was detected as described with chromogen nitroblue tetrazolium.

Polymerase chain reaction (PCR) and DNA sequencing. DNA was purified from 3 to 6, 10 ␮m paraffin sections from a subset of the same biopsy specimens described using the DNeasy Tissue System (Qiagen, Valencia, California) according to manufacturer instructions. DNA (500 ng) was amplified by PCR primers specific to the HCMV UL73 open reading frame (forward 5⬘-CGTCGACTGCTAGCACACCGCCTCCC-3⬘ and reverse 5⬘-GAGAGCGTTAAGCATAGTCCACCAGGC-3⬘) including 94C for 5 minutes, followed by 40 cycles of 94C for 30 seconds, 58C for 45 seconds and 72C for 1 minute, followed by 72C for 10 minutes. PCR buffer B, 400 ␮M of each deoxynucleoside triphosphate and 1.25 U Taq polymerase (Epicentre, Madison, Wisconsin) were used. PCR was performed in a 25 ␮l volume. Amplified products were visualized on agarose gels with ethidium bromide. Bands were cut out, and DNA was extracted using a gel extraction kit (Qiagen) and analyzed by automated sequencing using a Model 377 DNA Sequencer (ABI, Foster City, California). Confirmation of the HCMV UL73 sequence was performed using a National Center for Biotechnology Information BLAST (National Library of Medicine, Bethesda, Maryland) search of published sequences.13 RESULTS

To determine if HCMV viral encoded proteins were expressed in hyperplastic, preneoplastic (PIN) and neoplastic (carcinoma) prostatic epithelial lesions we performed IHC for HCMV in randomly selected human biopsy specimens using 3 mAbs. We used an antIE1-72 mAb for initial screening. We detected IE1-72 immunoreactivity in hyperplastic, preneoplastic and neoplastic epithelium in all 20 of the 20 biopsy specimen analyzed (see table, fig. 1). IE1-72 immunoreactivity was often prominent in areas of basal cell hyperplasia and in PIN lesions. In PIN lesions IE1-72 immunoreactivity was often characterized by a coarse granular cytoplasmic appearance (fig. 1, C, G and J), while the pattern of immunoreactivity in areas of carcinoma was often more diffusely cytoplasmic (fig. 1, D and H). In areas of basal cell hyperplasia hyperplastic basal cells were often strongly immunoreactive, while luminal epithelial cells and adjacent stroma was negative (fig. 1, B). Areas of normal prostatic epithelium were generally minimally immunoreactive (fig. 1, A) or nonimmunoreactive (fig. 1, E) for IE1-72. Stromal cells within tumor specimens were not immunoreactive for IE1-72 (fig. 1, A to J).

1000

HUMAN CYTOMEGALOVIRUS AND PROSTATE CANCER

FIG. 1. IHC staining of normal prostate, PIN and prostate carcinoma lesions with different antiHCMV mAbs. HCMV antiIE1-72 mAb IHC from 4 representative fields of same patient biopsy specimen shows limited immunoreactivity in normal epithelium (brown areas) (A), intense immunoreactivity in cells in areas of basal cell hyperplasia (B) and PIN (C), and less immunoreactivity in invading carcinoma cells (D). AntiIE1-72 IHC showed no immunoreactivity in normal epithelium (E), areas of PIN (F and G), and carcinoma (H) in another patient. In a third patient there was intracellular localization of IE1-72 protein in PIN lesion (I and J). Antipp65 immunoreactivity was shown in PIN lesions from another biopsy specimen (K and L). PIN (M) and carcinoma cells (N) were immunoreactive with antiHCMV cocktail mAbs. Positive control antiIE1-72 immunoreactivity was present in intranuclear inclusions in pneumocytes from CMV-positive patient with AIDS (O). Isotype control immunoreactivity occurred in scattered nuclei in PIN lesion with antiKi-67 mAb (P). A, F, H, I, K, M and N, reduced from ⫻100. B to E, G and L, reduced from ⫻200. J, reduced from ⫻400. O, reduced from ⫻400. P, reduced from ⫻200.

FIG. 2. ISH for HCMV early gene nucleic acids. HCMV early gene nucleic acid hybridization (purple areas) was detected in PIN but not in stromal cells from same patient (A and B). Invasive prostate carcinoma cells (arrows) show evidence of HCMV nucleic acid hybridization in another patient (C and D). B, 400X). E–F. No hybridization signal was detected in areas of PIN or carcinoma in 3 of 4 biopsy specimens when HPV cocktail probe was used (E (100⫻) and F). A, C and E, reduced from ⫻100. B, E and F, reduced from ⫻400.

HUMAN CYTOMEGALOVIRUS AND PROSTATE CANCER

1001

FIG. 3. HCMV UL73 PCR and DNA sequencing. Paraffin embedded, biopsy derived DNA was used for HCMV UL73 gene PCR (A). PCR products (262 bp) were noted in 9 of 10 reactions and water control was negative lane M, DNA ladder. Lanes 2, 4, 7 to 10, 12, 15, 17 and 20, patients 2, 4, 7 to 10, 12, 15, 17 and 20, respectively (see table). Representative DNA sequencing data shows HCMV UL73 gene sequence derived from patient 15 biopsy specimen (B).

Immunostaining with antipp65 mAb was used to determine whether the delayed HCMV tegument protein pp65 was expressed in tumors. We detected pp65 immunoreactivity in preneoplastic and neoplastic epithelium (fig. 1, K and L) from 10 of the same 13 prostate biopsy specimens that were immunoreactive for IE1-72. With a cocktail of mAbs specific for HCMV we also detected HCMV antigens in 10 of the 10 specimens specifically in preneoplastic and neoplastic epithelium (fig. 1, M and N). The patterns of immunoreactivity of pp65 and with the HCMV cocktail antibody were similar to the pattern of IE1-72 immunoreactivity except pp65 and HCMV cocktail immunoreactivity were present in the nuclei and cytoplasm of cells in neoplastic prostatic lesions. As a positive control for our antiIE1-72 mAb, we used paraffin sections of HCMV infected lung from a patient with AIDS, in which characteristic nuclear inclusion bodies were observed to be strongly immunoreactive (fig. 1, O). Isotype identical negative control mAbs served as controls in all 20 specimens. As IE1-72 and pp65 controls, we used an IgG1 isotype identical mAb specific for Ki-67 (a cell cycle marker). As a control for HCMV cocktail mAbs, we used an IgG2a isotype identical mAb specific for cytokeratin-20. Only rare focal or no immunoreactivity was present when either of these antibodies was used (fig. 1, P). To show that HCMV nucleic acids were present in the same cellular distribution as HCMV antigens we used mRNA ISH to detect HCMV IE gene mRNA in several of these specimens. We detected HCMV nucleic acids specifically in preneoplastic and neoplastic epithelial cells in PIN and carcinoma lesions from 4 of 4 patients (see table, fig. 2, A to D). No ISH signal was detected in 3 of these 4 biopsy specimens using the HPV cocktail probe (fig. 2, E and F), while discrete HPV hybridization was detected in 1 specimen in rare epithelial cells (data not shown). No ISH signal was detected when the probe was eliminated from hybridization (data not shown). The ISH signal pattern was similar to the pattern of HCMV immunoreactivity observed in the specimens, in that only epithelial and not stromal cells demonstrated evidence of HCMV nucleic acid hybridization. We extracted DNA from the same paraffin embedded biopsy specimens to amplify and sequence HCMV viral DNA. Of 10 specimens we amplified the HCMV UL73 gene (glycoprotein N) in 9 (fig. 3, A). No DNA was amplified from 5 water or blank paraffin negative controls. The HCMV UL73 gene encodes a variable region of the HCMV genome and UL73 sequences can be used to distinguish clinical isolates from laboratory strains of HCMV.13 We cut amplified UL73 PCR products from agarose gels, extracted DNA and performed direct DNA sequencing (fig. 3, B). DNA sequences of all PCR

FIG. 4. IHC and ISH of PC-3 cells 48 hours after infection with HCMV Towne strain. Immunostaining of HCMV infected PC-3 cells with mAbs specific to IE1-72 (black areas) (A) and pp65 (brown areas) (B) shows that only small percent of PC-3 cells expressed these HCMV proteins and these cells had rounded morphology compared with flattened morphology of most cells on monolayer. Double immunostaining with mAbs specific for IE1 (black areas) and Bcl-2 (red areas) demonstrate that PC-3 cells expressing IE1-72 (arrows) co-express Bcl-2 (C). (400X). Similarly cells expressing IE1-72 (black areas) are also advancing through cell cycle, as demonstrated by co-expression of PCNA (red areas in cells, arrows) (D). ISH with oligonucleotide probe specific for HCMV IE1 demonstrates hybridization in cell nuclei throughout monolayer (E) 100X) with increased nucleic acid hybridization in rounded cells (F), 400X), similar to pattern of IE1-72 and pp65 protein expression. Note ISH positive controls (probe for poly-A mRNA) (G) and negative control (no probe) (H). A, black chromagen, reduced from ⫻100. Inset, reduced from ⫻400. B, brown chromagen, reduced from ⫻100. Inset, reduced from ⫻400. C and D, black and red chromagens, reduced from ⫻400. E, reduced from ⫻100. F, reduced from ⫻400. G and H, reduced from ⫻100. Inset, reduced from ⫻400.

1002

HUMAN CYTOMEGALOVIRUS AND PROSTATE CANCER

specimens were determined to be homologous to a clinical isolate of HCMV, which is genetically distinct from the Towne and AD169 laboratory strains. To determine whether we could infect prostate carcinoma cells with HCMV and whether it might result in the induction of cellular pathways relevant to prostate cancer pathogenesis we infected the PC-3 prostate carcinoma cell line with the HCMV Towne strain. To confirm that cells were infected with HCMV we performed IHC with the IE1-72 and pp65 mAbs. These experiments demonstrated that the IE1-72 and pp65 mAbs were specific for HCMV since discrete cellular immunoreactivity occurred only in HCMV infected and not in mock infected PC-3 cells (fig. 4, A and B). Since we detected HCMV IE1-72 and pp65 in only a subset of PC-3 cells, it suggested that only a subset of cells was infected or viral protein expression was likely restricted to tumor cells in a specific state of differentiation or in a specific stage of the cell cycle. To determine if HCMV protein expression was indeed restricted to a subset of HCMV infected PC-3 cells we performed ISH on HCMV infected cells. We detected punctate HCMV nucleic acid hybridization in cell nuclei throughout the confluent PC-3 monolayer, suggesting that most PC-3 cells were infected with HCMV (fig. 4, E). We observed robust HCMV nucleic acid hybridization in many cells that were rounded and appeared to be dividing (fig. 4, E and F). We did not observe HCMV nucleic acid hybridization in mock infected cells (data not shown). These findings suggested that while most PC-3 cells were infected with HCMV, only a subset of cells, which appeared to be actively dividing cells, were expressing HCMV protein products. To determine whether HCMV protein expression was restricted to PC-3 cells that were undergoing cell cycle progression we performed double labeling IHC using mAbs specific to IE1-72 and to PCNA, which is only expressed in cells during cell cycle progression. We observed PCNA immunoreactivity restricted to cells that were also IE1-72 immunoreactive (fig. 4, D), suggesting that HCMV protein expression may be limited to cells undergoing cell cycle progression or those in a specific state of differentiation. Since the antiapoptotic protein Bcl-2 protein can be induced by HCMV infection14 and it is over expressed in low and high grade PIN lesions, we also performed double labeling experiments in HCMV infected PC-3 cells with mAbs specific for HCMV IE1-72 and Bcl-2 (fig. 4, C). Like PCNA, Bcl-2 expression in PC-3 cells was observed specifically in cells infected by HCMV (fig. 4, C, arrows). DISCUSSION

The data presented demonstrate that HCMV gene products were specifically expressed in a high percent of PIN and prostate carcinoma lesions in our study population. In vitro findings suggest that HCMV protein expression may be restricted to cells during a specific stage of the cell cycle and may promote anti-apoptotic pathways in prostate cancer cells. Together these findings imply that persistent HCMV infection of prostate epithelial cells may be common, and preneoplastic and neoplastic prostatic epithelium might serve as a unique reservoir for persistent viral infection and gene expression. Recent data on oncogenic pathways mediated by HCMV indicate that HCMV can dysregulate multiple important cellular pathways involved in neoplasia. Many similarities exist between the functions of the major immediate-early proteins of HCMV and similar proteins encoded by other DNA tumor viruses, including targeting of members of the Rb and p53 families, and the ability of these viral factors to promote cell cycle progression in quiescent cells, induce DNA mutations and block apoptotic pathways.7 HCMV infection can stimulate a host of cell signaling pathways, including activation of

c-myc, p38 mutagen activated protein kinase and Bcl-2 expression, which can influence cell differentiation, proliferation and sensitivity to apoptotic signals.14, 15 HCMV infection can also promote angiogenic, invasive and immune evasive pathways in cells.16, 17 Thus, persistent infection of prostate epithelium by HCMV could potentially promote oncogenic pathways that lead to transformation. CONCLUSIONS

Prostate carcinomas develop during many years through the acquisition of somatic DNA mutations and they occur at a higher rate in individuals with increased exposure to sexually transmitted diseases.18 The data presented indicate that persistent infection by HCMV, which can be sexually transmitted and has oncogenic properties, occurs in a high percent of hyperplastic, preneoplastic and neoplastic prostatic epithelial lesions. These findings are consistent with a model in which persistent infection of prostatic epithelium by HCMV might contribute to the pathogenesis of prostate cancer. REFERENCES

1. Nomura, A. M. and Kolonel, L. N.: Prostate cancer: a current perspective. Epidemiol Rev, 13: 200, 1991 2. Dennis, L. K. and Dawson, D. V.: Meta-analysis of measures of sexual activity and prostate cancer. Epidemiology, 13: 72, 2002 3. Hayes, R. B., Pottern, L. M., Strickler, H., Rabkin, C., Pope, V., Swanson, G. M. et al: Sexual behaviour, STDs and risks for prostate cancer. Br J Cancer, 82: 718, 2000 4. Howley, P. M., Ganem, D. and Kieff, E.: DNA viruses. In: Cancer: Principles and Practice of Oncology, 6th ed. Edited by V. T. De Vita, S. Hellman and S. A. Rosenberg. Philadelphia: Lippincot Williams and Wilkins Co., pp. 168 –173, 2001 5. Britt, W. J. and Alford, C. A.: Cytomegalovirus. In: Fields Virology, 3rd ed. Edited by B. N. Fields, D. M. Knipe and P. M. Howley. New York: Raven Press, pp. 2493–2523, 1996 6. Benson, P. J. and Smith, C. S.: Cytomegalovirus prostatitis. Urology, 40: 165, 1992 7. Castillo, J. P. and Kowalik, T. F.: Human cytomegalovirus immediate early proteins and cell growth control. Gene, 290: 19, 2002 8. Doniger, J., Muralidhar, S. and Rosenthal, L. J.: Human cytomegalovirus and human herpesvirus 6 genes that transform and transactivate. Clin Microbiol Rev, 12: 367, 1999 9. Sanford, E. J., Geder, L., Laychock, A., Rohner, T. J., Jr. and Rapp, F.: Evidence for the association of cytomegalovirus with carcinoma of the prostate. J Urol, 118: 789, 1977 10. Geder, L., Sanford, E. J., Rohner, T. J. and Rapp, F.: Cytomegalovirus and cancer of the prostate: in vitro transformation of human cells. Cancer Treat Rep, 61: 139, 1977 11. Rapp, F., Geder, L., Murasko, D., Lausch, R., Ladda, R., Huang, E. S. et al: Long-term persistence of cytomegalovirus genome in cultured human cells of prostatic origin. J Virol, 16: 982, 1975 12. Cobbs, C. S., Harkins, L., Samanta, M., Gillespie, G. Y., Bharara, S., King, P. H. et al: Human cytomegalovirus infection and expression in human malignant glioma. Cancer Res, 62: 3347, 2002 13. Pignatelli, S., Dal Monte, P. and Landini, M. P.: gpUL73 (gN) genomic variants of human cytomegalovirus isolates are clustered into four distinct genotypes. J Gen Virol, 82: 2777, 2001 14. Cinatl, J., Jr., Cinatl, J., Vogel, J. U., Kotchetkov, R., Driever, P. H., Kabickova, H. et al: Persistent human cytomegalovirus infection induces drug resistance and alteration of programmed cell death in human neuroblastoma cells. Cancer Res, 58: 367, 1998 15. Hagemeier, C., Walker, S. M., Sissons, P. J. and Sinclair, J. H.: The 72K IE1 and 80K IE2 proteins of human cytomegalovirus independently trans-activate the c-fos, c-myc and hsp70 promoters via basal promoter elements. J Gen Virol, 73: 2385, 1992 16. Speir, E., Modali, R., Huang, E. S., Leon, M. B., Shawl, F., Finkel, T. et al: Potential role of human cytomegalovirus and p53 interaction in coronary restenosis. Science, 265: 391, 1994 17. Loenen, W. A., Bruggeman, C. A. and Wiertz, E. J.: Immune evasion by human cytomegalovirus: lessons in immunology and cell biology. Semin Immunol, 13: 41, 2001 18. Abate-Shen, C. and Shen, M. M.: Molecular genetics of prostate cancer. Genes Dev, 14: 2410, 2000