Prostatic Intraepithelial Neoplasia 21 in ...

Allelic Loss on Chromosome 8p12−21 in Microdissected Prostatic Intraepithelial Neoplasia Michael R. Emmert-Buck, Cathy D. Vocke, Rudy O. Pozzatti, et al. Cancer Res 1995;55:2959-2962. Published online July 1, 1995.

Updated Version

Citing Articles

E-mail alerts Reprints and Subscriptions Permissions

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/55/14/2959

This article has been cited by 38 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/55/14/2959#related-urls

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].

Downloaded from cancerres.aacrjournals.org on July 13, 2011 Copyright © 1995 American Association for Cancer Research

[CANCER RESEARCH 55, 2959-2962,

July 15, 1995]

Advances in Brief

Allelic Loss on Chromosome 8pl2-21 Intraepithelial

in Microdissected

Prostatic

Neoplasia

Michael R. Emmert-Buck, Cathy D. Vocke, Rudy O. Pozzatti, Paul H. Duray, Scott B. Jennings, Charles D. Florence, Zhengping Zhuang, David G. Bostwick, Lance A. Liotta, and W. Marston Linchan1 Laboratory of Pathology Â¡M.R. E-B., P. H. D., Z. Z., L. A. L] and Surgical Branch Â¡S.B. J.. C. D. V., R. O. P.. C. D. F., W. M. L/, National Cancer Institute, Belhesda. Maryland 20892, and Department of Pathology, Mayo Clinic. Rochester. Minnesota 55905 ID. G. B.Â¡

Abstract The development and progression of human prostate cancer is associ ated with genetic abnormalities in tumor cells. Inactivation of tumor suppressor genes due to allelic loss is thought to be an important mech anism of gene alteration in prostatic neoplasms. In this study we examined allelic loss on chromosome Sp 12-21 in microdissected samples of normal prostatic epithelium, high grade prostatic intraepithelial neoplasia (PIN), and invasive prostate carcinoma from the same patients. Tissue microdis section under direct microscopic visualization procures pure populations of cells of interest, including small lesions such as PIN. Among 30 patients with concomitant cancer and PIN, we found loss of heterozygosity on chromosome 8pl2-21 in 63% (34 of 54) of foci of PIN examined and 90.6% (29 of 32) of tumors, suggesting that abnormalities on chromosome 8pl2-21 may be important in the early stages of prostatic carcinoma development. Several cases in which multiple foci of PIN from the same patient were sampled showed different patterns of allelic loss. Fifty-five % (16 of 29) of the prostate carcinomas contained a potential precursor PIN focus based on allelic loss pattern. Our results are consistent with the hypothesis that PIN arises multifocally within the prostate gland, and that a subset of these lesions progress to become carcinoma.

Introduction Human prostate cancer has been proposed to progress through an in situ tumor phase called PIN2 before evolving into overtly invasive cancer (1). PIN is frequently found in association with prostate car cinoma, and the cells of PIN have several histological features similar to those of invasive prostate cancer cells. However, there are currently few molecular studies examining the relationship of PIN and invasive prostate cancer. PIN typically arises in the peripheral zone of the prostate and has a spatial association with prostate cancer (2). Previ ous reports have shown that PIN is frequently aneuploid, and Sakr and Crissman (3) found LOH of chromosome 8p in 2 of 7 cases of PIN. These results favor the classification of PIN as a neoplastic lesion, but the precise relationship between PIN and invasive carcinoma remains unclear. The classification of PIN as a neoplastic entity has relevance both to the understanding of the fundamental genetic alterations that occur early in the development of human prostate cancer as well as to the potential use of PIN as a clinical marker of malignant transfor mation before the development of prostate cancer. In this study we used tissue microdissection to study LOH on chromosome 8pl2-21 in patients with both prostatic carcinoma and adjacent foci of PIN. Our results suggest that loss of DNA on

chromosome 8pl2-21 is important in the development of PIN and supports the hypothesis that PIN is a precursor of prostate cancer. Materials and Methods Tumor Samples. Frozen normal and tumor prostate samples from over 100 patients treated with transurethral prostatectomy or radical prostatectomy were obtained from the Mayo Clinic (Rochester, MN). Thirty cases that contained cancer and at least one focus of high-grade PIN were selected for study. Fourteen cases contained more than one focus of PIN. The tumor grade varied from well differentiated to poorly differentiated. Microdissection. Microdissection of selected populations of normal epi thelial cells, cells from PIN, and cancer cells from frozen tissue sections was performed under direct light microscopic visualization with the use of a method developed previously (4-6). Specific cells of interest were microdis sected and procured from unstained 8-^m frozen tissue sections. In each case normal epithelium, PIN, and tumor cells from the same patient were analyzed. DNA Extraction. Procured cells were immediately resuspended in a 20-/il solution containing 10 mM Tris-HCl (pH 8.0J-1 mM EDTA-1% Tween 20-0.1 mg/ml proteinase K and incubated overnight at 37Â°C.The mixture was boiled for 5 min to inactivate the proteinase K, and 0.5-2% of this solution was used for PCR. Detection of LOH. The oligonucleotide markers used in this study were purchased sized on an Applied Biosystems DNA D8S137, and NEFL, located on chromosome

primers for the polymorphic DNA from Research Genetics or synthe synthesizer. Loci were D8SI36, 8pl2-21. Reactions with D8S137

and NEFL were performed in an MJ Research thermal cycler as follows: 2 min at 95Â°C;followed by 40 cycles of 95Â°Cfor 30 s, 62Â°Cfor 30 s, and 72Â°Cfor 30 s; followed by a final 2-min incubation at 72Â°C.Reactions with D8S136 were cycled as follows: 2 min at 95Â°C;40 cycles with 95Â°Cfor 30 s, 55Â°Cfor 30 s, and 72Â°Cfor 30 s; followed by a final 2-min incubation at 72Â°C.PCR was performed in 12.5-fil reactions with 200 mM dNTP, 0.8 /J.Mprimers, 2 uCi of [a-32P]dCTP (DuPont New England Nuclear), and 1 unit of Taq polymerase (Boehringer Mannheim). Labeled amplified DNA was mixed with an equal volume of formamide-loading dye (95% formamide; 20 mm EDTA; 0.05% bromophenol blue, and 0.05% xylene cyanol). The samples were denatured for 5 m at 95Â°C and loaded onto a gel consisting of 7% acrylamide (49:1, acrylamide:bis), 5.6 Murea, 32% formamide, and 1x TBE (0.089 MTris, 0.089 M borate, 0.002 M EDTA). Samples were electrophoresed at 95 W for 2-4 h. Gels were transferred to 3 M Whatman paper, and autoradiography was performed with Kodak X-OMAT film. The criterion for loss of heterozygosity was complete or near complete loss of one alÃ-eleas determined by visualiza tion. Cases with LOH showed two alÃ-elesin the normal epithelium control and one alÃ-elein the tumor or PIN, all with similar intensities. Cases with complete or near complete loss (i.e., very faint band) of one alÃ-elein tumor or PIN were scored as LOH at that marker.

Results Received 4/19/95; accepted 6/2/95. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ' To whom requests for reprints should be addressed, at Urologie Oncology Section, Surgery Branch, Building 10, Room 2B47, National Cancer Institute, 20892-1502. 2 The abbreviations used are: PIN, prostatic intraepithelial neoplasia; heterozygosity; HPF, high-power field.

We used tissue microdissection to study 30 patients with concom itant high grade PIN and prostate cancer. In each case normal epithe lium, invasive tumor and at least one focus of PIN from the same Bethesda, MD patient were examined. In 14 cases multiple foci of PIN were exam ined. In all cases each focus of PIN and the corresponding invasive LOH, loss of tumor were selectively microdissected free of adjacent stroma, normal 2959


ALLELIC

LOSS IN MICRODISSECTED

epithelium, and inflammatory cells. Essentially pure populations of cells of interest were procured (Fig. 1). LOH on chromsome 8p 12-21 occurred in at least one focus of PIN in 26 of 29 (89.7%) informative cases (Table 1 and Fig. 2). Fourteen cases contained more than 1 focus of PIN, and 11 of these showed different allelic loss patterns among the foci of PIN, including loss of opposite alÃ-elesas seen in case 311 for example. Table 1 illustrates the 8p 12-21 LOH patterns. Cases 277, 283, 298, 275, 311, 267, 718, 319, 276, 297, and 2779 all showed foci of PIN with variable allelic loss patterns. In total, 8p 12-21 LOH was seen in 63% (34 of 54) of foci of PIN studied. In this study we did not attempt to determine the frequency of 8pl2-21 LOH in low-grade PIN due to the difficulty in confidently identifying low-grade PIN in frozen tissue sections. Allelic loss on chromosome 8pl2-21 was seen in 29 of 32 (90.6%) tumor samples. Multiple cancer samples from the same patient were only counted as separate tumors if they showed different allelic loss patterns. Samples from the same patient with identical allelic loss patterns were counted as one tumor. Microdissection of each invasive tumor was confined to a region less than or equal to 2 HPFs (X400; Olympus CK 2 inverted microscope). Pure or near pure populations of tumor cells were microdissected and procured in each case. In 6 cases, geographically separate tumor foci (separated by a minimum of 5 HPFs) were sampled (Table 1, cases 277, 278, 302, 812, 549, and 318). Three cases showed identical allelic loss patterns, and 3 cases showed differences in LOH profiles among the tumor foci. Chromosome 8pl2-21 LOH was observed reproducibly. Reactions with each marker were run a minimum of two times for each sample. Normal epithelium consistently showed two alÃ-elesat informative loci, and LOH in tumor cells and PIN cells occurred reproducibly.

A

HUMAN PROSTATE

TUMORS

Artifactual allelic dropout was not observed in the normal samples. Thus, the two alÃ-elesin the microdissected samples were reliably amplified in the PCR reaction. Allelic instability was seen at two loci (D8SÃŒ37and D8SÃŒ36)in case 297 (Table 1). This case was consid ered to have no LOH at these two markers. Discussion The development and progression of prostate cancer is thought to be due to dysregulation of multiple oncogenes and tumor suppressor genes (7). However, little is understood of the molecular changes which occur in prostate carcinoma as compared to many other human malignancies. Prostate cancer is challenging to study as the tumors are composed of a heterogeneous collection of stromal, epithelial, and inflammatory cells, and long-term cultures of prostate cell lines are difficult to establish (7). Tissue microdissection is a useful technique in the study of prostate cancer. Investigators can recover essentially pure populations of cells from normal epithelium, PIN, or invasive carcinoma. Thus, specific genetic profiles can be generated and com pared for each of the cell types. Allelic loss has been reported in prostate cancer on chromsomes 7q, 8p, lOq, and 16q, suggesting the presence of a tumor suppressor gene(s) at these loci (3, 8-12). LOH studies of chromosome 8p have identified 2 separate areas of deletion on this chromosome located at 8pl2-21 and 8p22 (13, 14). Bova et al. (15) reported a homozygous allelic deletion at the MSR locus in a single case of prostate cancer. The short arm of chromosome 8 has also been implicated in other human malignancies, including colorectal carcinoma, malignant fi brous histiocytoma, bladder cancer, hepatocellular carcinoma, and

B

-. ^f

* Â¿5p

-r ^

Fig. I. Sequential microdissection of PIN showing selective procurement of one PIN focus is shown before (A) and after (B) microdissection. Sequential microdisseclion of invasive prostate tumor showing procurement of individual nests of invasive tumor cells is shown before (C) and after (D) microdissection.

2960


ALLELIC


HUMAN PROSTATE

Table 1 Analysis of loss of heterozygosity at 3 loci on chromosome 8pl2-21

TUMORS

in 30 patients with PIN and prostate carcinoma

Multiple cancer samples from the same patient were only counted as separate tumors if they demonstrated different allelic loss patterns. Cancer samples from the same patient with identical allelic loss were considered to be part of the same tumor. Case No.277283298275274278311267302718DiagnosisPIN-1PIN-2Cancer

No.3198122762972779549318315272299306253343009415350931300639DiagnosisPIN-1PIN-2CancerPINCan

1Cancer

2 Cancer 3 LU LU LULU 1 Cancer 4 LULULUniniLULUniniLUniniLUniLUniLUninini LU Cancer 2PIN-1PIN-2CancerPIN-1PIN-2CancerPIN-1PIN-2CancerPINCancer LULUniniAIAIAInininiLUniL LUniniLUniLUniniLuniniLUniLLninininini Cancer 5Cancer LULUniniLLLUniLULULULLLLLULLnininiLLD8S136LUniLULU 6PIN-1PIN-2PIN-3PIN-4PIN-5PIN-6CancerPIN-1PIN-2PIN-3PIN-4CancerPIN-1PIN-2PIN-3CancerPIN-1

1Cancer 2PINCancer 1Cancer 2PINCancerPIN

PIN-2 ni PIN-3CancerPIN-1PIN-2Cancer niLLLULULULULULLLLLLLU

ni CancerPINCancerPINCancerPINCancerPINCancerPINCancerPINCancerPIN niLLLLLLLLLLLULLLLLLniniLULULL

1Cancer 2PIN-1PIN-2PIN-3CancerPIN-1PIN-2PIN-3CancerPIN-1

PIN-2Cancer LULLLUninininiNEFLLUniLULU 1Cancer 2PIN-1PIN-2PIN-3CancerD8SÃŒ37LU"niLULU

LLLULLniniLLLLCase

CancerPINCancerPINCancerPINCancerD8S137niniLLAIniLULUniLLLLniLUniLLniniLU LULULUniniLLLUNEFLLUniLUniniLUniLUninininiLLniniLUnin LULLLLD8SI36LULULULULU

LU, loss of upper alÃ-ele;LL. loss of lower alÃ-ele;ni, no loss; AI, allelic instability. Blank spaces, noninformative

non-small cell lung cancer (16-20).

Thus, two separate areas of

chromosome 8p may contain important tumor suppressor genes the functional inactivation of which is important in many human malig nancies. The present study shows a significantly higher rate of 8p 12-21 LOH in invasive prostatic tumors than reported previously. This is consistent with several PCR-based studies of tumor LOH we have

loci.

indicates that the tumor cells procured from within each 2 HPF region are identical with respect to 8pl2-21 alÃ-eleloss, suggesting that each

sampled region represents a monoclonal tumor. If the tumor sample represented a mixture of several independent clones, each with a 50% chance of losing a specific alÃ-ele,then one would expect frequent masking of allelic loss which would not approach the high rate of 8pl2-21 LOH we observed. Thus, we can be reasonably confident that each 2 HPF tumor focus is representative of a single tumor clone. performed using tissue microdissection. Slight contamination of tu However, the results from the six cases in which geographically mor cell samples with nontumor cells can frequently result in sub stantial amplification of the contaminating normal alÃ-eleand mask distinct tumor regions were sampled suggest that prostate cancer can be composed of separately arising tumors. Case 302 shows 2 separate tumor specific allelic loss that is occurring in the tissue. In our study, 2 HPF tumor foci procured from the same frozen tissue section with microdissection of each invasive tumor focus was confined to an area opposite patterns of allelic loss, indicating they arose independently of less than or equal to 2 HPFs. Small nests of tumor cells from within each other, assuming that chromosome 8pl2-21 LOH is an early each region were microdissected free of contaminating stremai and event in the formation of these tumors. Thus, it appears that in genetic inflammatory cells and procured for analysis. Invasive tumors showed 8p 12-21 LOH in 90.6% (29 of 32) of samples. The high rate of LOH analyses of prostate cancer it is necessary to utilize microdissection of 2961


ALLELIC


A B C D E

HUMAN PROSTATE

TUMORS

frequent event that occurs early in the progression of prostate carci noma. The data are consistent with the hypothesis that chromosome 8p 12-21 contains a tumor suppressor gene(s) that is important in the development of prostate cancer. Characterization of genes important in the progression of PIN may be helpful in the identification of patients with PIN who are likely to develop carcinomas. References

Fig. 2. Denaturing gel showing cases of LOH on chromosome 8p 12-21 (marker D8S137) in PIN and prostate carcinoma in case 302. Lane A, normal prostate epithelium showing the presence of two alÃ-eles;Lanes B and C, individual PIN foci showing loss of the upper alÃ-ele;Lane D, cancer region 1 showing loss of the lower alÃ-ele;Lane E, cancer region 2 showing loss of the upper alÃ-ele.

small defined regions of invasive tumors to avoid the possibility of inadvertently combining separate cancers. The high rate of chromosome 8pl2-21 LOH in PIN and matched cancers that we found suggests that PIN is a neoplastic lesion and a precursor of carcinoma. Trapman et al. (14) recently reported signif icant LOH of chromosome 8p 12-21 in prostate carcinoma. However, their investigation concluded that 8pl2-21 LOH was a late event in prostate cancer progression due to a much higher frequency of allelic loss detected in mÃ©tastases(83%) than in radical prostatectomy spec imens (33%). The present results suggest that 8p 12-21 LOH occurs early in the development of prostate cancer, i.e., in the in situ precur sor lesion. Thus, functional inactivation of one or more genes located on 8pl2-21 may be a prerequisite for the development of PIN. In contrast, chromosomes lOq and 16q, which have also been suggested to contain important prostate cancer-related tumor suppressor genes, showed less than 10% LOH in PIN foci.3 Eleven of the 14 cases with multiple foci of PIN showed differences in allelic loss profiles among the PIN foci, suggesting that these lesions arise independently and multifocally. In each case, the multi ple foci of PIN were from the same frozen tissue section and were individually microdissected. Comparison of 8pl2-21 LOH patterns between PIN and the corresponding cancer showed identical allelic loss, with at least one of the foci of PIN in 44.8% (13 of 29) of cases. If PIN with similar but smaller allelic loss is considered to be a potential precursor of the matched cancer [i.e., PIN may progress via loss of additional 8pl2-21 gene(s)], then 55.2% (16 of 29) of the cases showed a potential precursor PIN lesion. However, it is possible that the specific PIN lesion that has progressed to cancer in each case may no longer be identifiable in the tissue if the tumor has overgrown the PIN. Thus, it appears that multiple, independent PIN lesions arise in a field that precedes the development of prostate cancer. The subsequent progression of PIN to adenocarcinoma may require addi tional alterations in tumor suppressor genes or oncogenes, possibly involving a larger deletion on chromosome 8pl2-21. In summary, we report that LOH on chromosome 8pl2-21 is a 'C. D. Volke, R. D. Pozzatti, C. D. Florence, S. B. Jennings, S. E. Sirup, M. R. Emmert-Buck, P. H. Duray, D. G. Bostwick, L. A. Liotta, and W. M. Linehan. Analysis of 100 microdissected prostate carcinomas reveals high frequency of allelic loss on chromosome 8pl2-21, manuscripl in preparation.

1. Bostwick, D. G., and Brawer, M. K. Prostatic infra-epithelial neoplasia and early invasion in prostate cancer. Cancer (Phila.), 59: 788-794, 1987. 2. McNeal, J. E., and Bostwick, D. G. Intraductal dysplasia: a premalignant lesion of the prostate. Hum. Pathol., 17: 64-71, 1986. 3. Sakr, W. A., Macoska, J. A., Benson, P., Grignon, D. J.. Wolman, S. R., Pontes, J. E., and Crissman. J. D. Allelic loss in locally metastatic. multisampled prostate cancer. Cancer Res., 54: 3273-3277, 1994. 4. Emmert-Buck, M. R., Roth, M. J., Zhuang. Z., Campo, E.. Rozhin, J., Sloane, B. F., Liotta, L. A., and Stetler-Stevenson, W. G. Increased gelatinase A (MMP-2) and cathepsin B activity in invasive tumor regions of human colon cancer samples. Am. J. Pathol., 145: 1285-1290, 1994 5. Zhuang. Z.. Merino. M. J., Chuaqui, R., Liotta, L. A., and Emmert-Buck, M. R. Identical allelic loss on chromosome Ilql3 in microdissected in situ and invasive human breast cancer. Cancer Res., 55: 467-471, 1995. 6. Zhuang, Z, Bertheau, P., Emmert-Buck, M. R., Liotta, L. A.. Gnarra, J., Linehan, W. M., and Lubensky, I. A. A microdissection technique for archival DNA analysis of specific cell populations in lesions