Novel Nuclear Phosphoprotein pp32 Is HighlyExpressed in Intermediate- and High-Grade

Prostate CancerS.S. Kadkol, J.R. Brody, J.I. Epstein, F.P. Kuhajda, and G.R. Pasternack*

Divisions of Molecular Pathology and Surgical Pathology, Department of Pathology,The Johns Hopkins University School of Medicine, Baltimore, Maryland

BACKGROUND. pp32 is a differentiation-regulated nuclear phosphoprotein that is highlyexpressed in many cancers, but is restricted to self-renewing and long-lived normal cellpopulations. During murine embryogenesis, pp32 is expressed in primitive cell populations,diminishing as tissues terminally differentiate. Functionally, pp32 confers resistance to pro-grammed cell death and, paradoxically, inhibits transformation mediated in vitro by a broadrange of oncogenes, suggesting that pp32 is a multifunctional molecule with potentiallycomplex activities in cancer.METHODS. We studied pp32 expression in prostatic adenocarcinomas and benign prostatichyperplasia by in situ hybridization.RESULTS. In benign prostatic tissues, moderate pp32 expression occurs only in the basalcells. This study found elevated pp32 expression in 98% (54/55) of prostatic adenocarcinomasof Gleason score ù5 (P < 0.0001).CONCLUSIONS. These results suggest that pp32 may be diagnostically useful and maycontribute mechanistically to prostate tumor development. In comparison to other molecularalterations, increased pp32 expression is one of the most frequent events in primary prostatecancer. Prostate 34:231–237, 1998. © 1998 Wiley-Liss, Inc.

KEY WORDS: prostate cancer; in-situ hybridization; pp32

INTRODUCTION

Prostatic adenocarcinoma is the most frequent ma-lignancy in adult men, with approximately 317,000new cases diagnosed each year [1]. In spite of the ca-pabilities for early diagnosis and treatment [2], it rep-resents the second leading cause of cancer death inmen following lung cancer.

The molecular pathogenesis and evolution of pros-tate cancer remain poorly understood. Molecularstudies, like their histopathologic counterparts [3,4],can provide important clinical information while re-vealing key steps in tumor development or progres-sion. Characterization of biologically important mol-ecules in a clinical context can enable better and earlierdiagnosis, establishment of prognosis, and treatment.Nevertheless, study of alterations in specific genes hasnot been particularly rewarding in primary prostatecancer. Most alterations in the widely studied onco-genes and tumor suppressor genes occur in only 20–30% of primary prostate carcinomas, except for themyc gene, where overexpression has been observed in

as many as 50–60% of such cases [5]. Up to 40% ofprimary prostate cancers studied by comparative ge-nomic hybridization display chromosomal aberrations[6], although such alterations occur more frequently astumors recur and become refractory to hormonaltherapy. Characterization of candidate protoonco-genes or tumor suppressor genes at such altered locimay eventually shed light on tumor progression in theprostate.

In the present study, we extend previous prelimi-nary work [7] investigating the expression pattern ofpp32, a novel nuclear phosphoprotein in prostate can-cer. pp32 is a nuclear phosphoprotein that is differen-

Contract grant sponsor: USPHS; Contract grant number: ROI CA54404; Contract grant sponsor: W.W. Smith Charitable Trust.*Correspondence to: Gary R. Pasternack, M.D., Ph.D., Departmentof Pathology, 512 Ross, The Johns Hopkins University School ofMedicine, 720 Rutland Avenue, Baltimore, MD 21205. E-mail:gpastern@welchlink.welch.jhu.eduReceived 26 February 1997; Accepted 10 June 1997

The Prostate 34:231–237 (1998)

© 1998 Wiley-Liss, Inc.

tiation-regulated in most normal tissues during em-bryogenesis [Rebel et al., unpublished observations]and during differentiation of adult prostatic epithe-lium [8], but is highly expressed in many neoplasticcells [8,9]. While the detailed mechanism of pp32 ac-tion remains unknown, recent studies suggest thatpp32 may function both at the level of transformingevents [10] and through inhibition of programmed celldeath [11]. The human pp32 cDNA sequence (Gen-Bank U73477) is 1,052 bp in length and encodes a pro-tein of 249 amino acids. The protein is composed oftwo domains: an amino terminal amphipathic a-helical region containing a leucine zipper, and ahighly acidic carboxyl terminal region. The murineand human forms of pp32 are highly conserved, withover 90% nucleic acid homology and over 95% pro-tein-level homology. The purpose of the present studywas to determine the clinical significance of alteredpp32 expression in prostatic adenocarcinoma.

MATERIALS AND METHODS

Tissue Specimens

Fifty-five cases of formalin-fixed, paraffin-embed-ded tissues of prostatic adenocarcinoma and 8 cases ofbenign prostatic hyperplasia free of tumor were ob-tained from the Department of Pathology at the JohnsHopkins Hospital (Baltimore, MD).

Preparation of Labeled Probes

Nonradioactive in situ hybridization was per-formed with antisense and sense pp32 RNA probesgenerated by in vitro transcription. A 298-bp N-terminal sequence of the pp32 cDNA was subclonedinto the expression vector BlueScript by standard tech-niques [12]. Digoxigenin-labeled antisense and senseRNA probes were generated using a commerciallyavailable kit (Boehringer Mannheim, Indianapolis,IN). Vector DNA linearized with BamHI and XhoI, fol-lowed by phenol chloroform extraction and ethanolprecipitation, served as template for antisense andsense probe generation, respectively. In vitro tran-scription was performed for 2 hr at 37°C in a finalvolume of 20 ml, which contained 1 mg of templateDNA, 2 U/ml of either T3 or T7 RNA polymerase, 1U/ml ribonuclease inhibitor, 1 mM each of ATP, CTP,and GTP, 0.65 mM UTP, 0.35 mM digoxigenin-11-UTP, 40 mM Tris-HCl, pH 8.0, 10 mM NaCl, 10 mMDTT, 6 mM MgCl2, and 2 mM spermidine. The reac-tion was stopped by adding 2 ml of 0.2 M EDTA, pH8.0, and the synthesized transcripts were precipitated

for 30 min at −70°C with 2.2 ml of 4 M LiCl and 75 mlof prechilled ethanol. RNA was pelleted by centrifu-gation, washed with 80% ethanol, mildly dried, anddissolved in 100 ml of DEPC-treated water. Yields oflabeled probe were determined by an enzyme linkedimmunoassay, using a commercially available kit(Boehringer Mannheim, Indianapolis, IN).

Nonisotopic mRNA In Situ Hybridization

In situ hybridization was carried out essentiallyas described [7,13]. Sections (4-mm) of paraffin-em-bedded prostate tissue on poly-L-lysine-coated slideswere heated for 1 hr at 65°C, deparaffinized in twochanges of xylene for 10 min each, and washed twicein 100% ethanol for 5 min, followed by rehydration ina graded series of ethanol (95%, 70%, and 50%, 5 mineach). Sections were rinsed in DEPC water for 2 minand immersed in 0.1 M phosphate-buffered saline(PBS), pH 7.4, at room temperature. Subsequently, sec-tions were digested for 30 min with 10 mg/ml protein-ase K at 37°C in 100 mM Tris HCl, pH 8.0, and 50 mMEDTA. Proteolysis was stopped by immersing theslides in 4% paraformaldehyde at 4°C for 5 min, fol-lowed by three PBS washes. Sections were then acety-lated with freshly prepared 0.25% acetic anhydride in0.1 M triethanolamine HCl, pH 8.0, at room tempera-ture, for 10 min to prevent nonspecific probe binding,washed twice in 2 × SSC for 5 min, and prehybridizedfor 1 hr at 45°C with 50 ml of a solution containing 50%deionized formamide, 2 × SSC, 5% dextran sulfate, 1 ×Denhardt’s solution, and 400 mg/ml heat-denaturedsalmon sperm DNA. Hybridization was performed for18 hr at 45°C in 50 ml of hybridization solution (pre-hybridization solution + 15 ng/ml probe) per section.After washing off excess probe in 2 × SSC, unhybrid-ized single-stranded probe was removed by incubat-ing sections in 40 mg/ml RNAase A in 500 mM NaCl,10 mM Tris HCl, pH 8.0, and 1 mM EDTA. Finally, toachieve high-stringency hybridization, sections werewashed in 2 × SSC and 0.2 × SSC for 30 min each at50°C.

Detection of Hybridization

After the posthybridization washes, sections wereimmersed in TBS I (100 mM Tris HCl, pH 7.5, and 150mM NaCl) at room temperature for 10 min. Nonspe-cific antibody binding was blocked by incubation inTBS I containing 10% normal sheep serum and 0.3%Triton X-100 for 30 min. Fifty ml of 1.5 U/ml of sheepanti-digoxigenin antibody-alkaline phosphatase con-jugate (Boehringer Mannheim, Indianapolis, IN) werethen applied to each section in TBS I with 5% normalsheep serum and 0.3% Triton X-100 for 2 hr. This was

232 Kadkol et al.

followed by three washes in TBS I to remove excessantibody. Sections were then immersed in TBS II (100mM Tris HCl, pH 9.5, 100 mM NaCl, and 50 mMMgCl2) for 5 min to activate the alkaline phosphatase.The anti-digoxigenin alkaline phosphatase conjugatewas visualized by incubating sections in dark for 3 hrwith BCIP/NBT color solution, freshly prepared ac-cording to the manufacturer’s directions (BoehringerMannheim, Indianapolis, IN). All detection steps werecarried out at room temperature. After color develop-ment (about 3 hr), sections were rinsed in distilledwater, counterstained for 1 min in nuclear fast redsolution, and mounted with aqueous medium (glyc-erol-gelatin).

Analysis of Tissue Sections

Each tissue section was graded by the Gleason sys-tem and scored for signal intensity, ranging from

negative to 1+ (weak expression) through 4+ (intenseexpression), taking into consideration the predomi-nant staining pattern. Sections were considered posi-tive if more than 20% of tumor cells expressed pp32.

Statistical Methods

Chi-square analysis was used to determine statisti-cal significance.

RESULTS

pp32 Expression in Benign ProstaticHyperplasia (BPH)

In noncancerous prostate, pp32 expression was ob-served only in the basal cells of benign glands. The

Fig. 1. pp32 expression in benign prostatic hyperplasia (×100).Fig. 2. High-power view of pp32 expression in benign prostatichyperplasia (×400). Note that pp32 expression is restricted tobasal cells. Fig. 3. pp32 expression in prostatic intraepithelial neo-plasia (×250). Note that pp32 expression is higher in basally lo-cated cells. Fig. 4. Low-level pp32 expression in Gleason score2–4 prostatic adenocarcinoma (×400). Fig. 5. High-level pp32expression in Gleason score 5–7 prostatic adenocarcinoma

(×250). Fig. 6. High-level pp32 expression in Gleason score 8–10prostatic adenocarcinoma (×400). Fig. 7. Heterogeneity of pp32expression in prostatic adenocarcinoma (×400). Fig. 8. Low-levelpp32 expression in main mass of prostatic adenocarcinoma. Inset:High-level pp32 expression in those malignant glands on the ad-vancing edge of the same lesion, close to the capsule. Fig. 9.Representative negative control section hybridized to a sensepp32 probe.

pp32 Expression in Prostate Cancer 233

glandular secretory cells did not express detectablepp32 mRNA (Figs. 1, 2).

pp32 Expression in Prostatic Adenocarcinoma

Overall, 98% (54/55) cases of prostate cancer ex-pressed pp32. When present, prostatic intraepithelialneoplasia (PIN) expressed pp32 in a heterogeneousmanner; cells close to the basement membrane highlyexpressed pp32, whereas cells towards the luminal as-pect showed weak pp32 expression (Fig. 3). Most PINlesions were identified on the same sections as tumors.Low Gleason score adenocarcinomas (Gleason score2–4) predominantly expressed pp32 mRNA at low lev-els (7/9 cases, 78%, Fig. 4). In contrast, prostatic ad-enocarcinomas of Gleason score ù5 generally ex-pressed high levels of pp32 (40/46 cases, 87%, Figs. 5,6). Only 6/40 (13%) cases above Gleason score ù5expressed low levels of pp32. Seventy to eighty per-cent of tumor cells expressed pp32, although tumorsin the Gleason score 5–7 range did so with some het-erogeneity (Fig. 7). Gleason score 8–10 carcinomas ex-pressed pp32 uniformly. These results are summa-rized in Table IA.

Two types of cases afforded an interesting oppor-tunity to study low- and high-grade prostatic adeno-carcinoma from the same specimen but in differentblocks. In these cases, the low-grade tumor expressedpp32 weakly, whereas the high-grade tumor stronglyexpressed pp32. pp32 expression level and tumorGleason score ù5 showed a statistically significantcorrelation (Table IB). Among the adenocarcinomasstudied, 23/55 cases (42%) showed malignant glandsor cells close to the prostatic capsule and well sepa-rated from the main mass. These glands or cellsshowed a much higher level of pp32 expression in8/32 cases (30%, Fig. 8) than the main mass of thetumor. These lesions may represent either aggressiveclones derived from main tumor, or else multifocaltumorlets of independent origin. Primary prostatic ad-enocarcinomas were grouped into clinically aggres-sive and nonaggressive categories based on focal or

extensive capsular penetration, seminal vesicle inva-sion, and surgical margin involvement. Overall therewas no difference in the frequency of pp32 expressionlevels between tumors that showed clinically aggres-sive features and tumors that did not exhibit thesefeatures. However, all clinically invasive adenocarci-nomas had a high level of pp32 expression. These re-sults are summarized in Table II.

Additional normal structures in each case of pros-tatic adenocarcinoma were evaluated for pp32 expres-sion. A moderate to high level of pp32 expression wasalways observed in lymphocytes, in the ganglion cellsof the periprostatic neural ganglia, and occasionally inthe urothelium, vascular endothelium, vessel wall,and stromal nodules. An occasional case showed weakexpression of pp32 in the benign prostatic glandularepithelium close to the tumor. Whether this was a fieldeffect of the tumor or is treatment-related is unknownat present, since the glandular cells are negative forpp32 expression in BPH without malignancy.

DISCUSSION

The major finding of this study is that all primarytumors of Gleason score ù5 expressed high levels ofpp32. An initial study of pp32 expression in 11 cases ofprostatic adenocarcinomas showed a significant corre-lation of high-level pp32 expression with clinical in-

TABLE IB. Correlation of pp32 Expression WithGleason Score in Prostatic Adenocarcinoma*

Gleason score2–4

Gleason scoreù5

Low or no expression 89% 13%High expression 11% 87%

*Primary prostatic adenocarcinomas were graded according tothe Gleason scoring system and analyzed for pp32 expression. Achi-square analysis revealed a statistically significant correlationof pp32 expression level with tumor Gleason score ù5 (P <0.0001).

TABLE IA. pp32 Expression in Human Prostatic Adenocarcinomas According to Gleason Score*

Gleason 2–4 (n = 9) Gleason 5–6 (n = 15) Gleason 7 (n = 18) Gleason 8–10 (n = 13)

High expression 1/9 (11%) 14/15 (93%) 15/18 (84%) 11/13 (85%)Low expression 7/9 (78%) 1/15 (7%) 3/18 (16%) 2/13 (15%)No expression 1/9 (11%) 0/15 (0%) 0/18 (0%) 0/13 (0%)

*Nonisotopic in situ hybridization was performed on formalin-fixed, paraffin-embedded sections with digoxigenin-11-UTP-labeledantisense RNA probe generated by in vitro transcription of a 298-bp cDNA insert in the expression vector BlueScript. Hybridizationwas detected colorimetrically with BCIP/NBT after incubation sections with anti-digoxigenin antibody linked to alkaline phosphatase.The signal intensity was scored from negative to 1+, 2+ (low expression) through 3+, 4+ (high expression). Hybridization with a senseprobe served as a negative control in each case.

234 Kadkol et al.

vasiveness as well as Gleason score ù5 [7]. In contrast,the present study of 55 cases found that high-levelexpression still correlated with Gleason score ù5, butalso found no correlation with features of invasion,including focal and extensive capsular penetration,seminal vesicle involvement, and surgical margin in-volvement. This apparent discrepancy most likely re-lates to the small size of the initial study.

The frequency of increased pp32 expression in pri-mary prostate cancer is striking in comparison to therelative paucity of quantitative or qualitative alter-ations of other molecules in this disease. More typi-cally, protooncogene activation or overexpression andtumor suppressor gene loss occur in about 15–40% ofprimary prostatic adenocarcinomas [14–19]. From amechanistic standpoint, alterations present in only aminority of primary prostatic adenocarcinomas areunlikely to play a significant role in the pathogenesisof most prostate cancers. Molecular alterations foundonly in a relatively small proportion of cases are un-likely to be general etiologic factors. Along with theloss of glutathione-S-transferase pi expression througha methylation-dependent mechanism and telomeraseactivation [20,21], pp32 expression is one of the mostfrequently observed molecular changes, occurring innearly all prostate cancers studied.

Our results, demonstrating the high prevalence ofpp32 expression in human prostatic adenocarcinoma,suggest that pp32 might be exploited for determina-tion of clinical diagnosis and treatment. Because pp32is expressed in the vast majority of cases of prostaticadenocarcinoma of Gleason score ù5, and becausepp32 expression does not correlate with surrogatemarkers of disease progression, pp32 expression doesnot appear to be inherently prognostic. Although pp32is expressed in normal cells (basal cells in the prostaticglands and lymphocytes), the regular and pronouncedquantitative differences in pp32 expression in clini-cally significant cases might prove useful for tumordiagnosis, especially in situations where limited bi-opsy tissue is available for examination. Moreover, ifpp32, like certain other nuclear proteins, proves to bedetectable in serum, it could also serve as a serologictumor marker. Finally, unlike other molecular eventswhich occur in only a fraction of clinical prostatic can-cers, the high prevalence of pp32 expression suggeststhat it could also be a potential therapeutic target, ei-ther by chemotherapeutic or gene-targeting methods,as its mechanism of action becomes clarified.

The present study reveals higher levels of pp32mRNA in tumor tissue, but does not provide a mecha-nism for this occurrence. The high levels of pp32 ex-pression in prostatic adenocarcinoma may be due tomutations or alternative splicing that increase steady-state pp32 levels, or to changes in pp32 regulation that

TA

BL

EII

.pp

32E

xpre

ssio

nin

Rel

atio

nto

Clin

ical

Inva

sive

ness

of

Pro

stat

icA

deno

carc

ino

mas

*

FCP

(n=

34)

No

FCP

(n=

18)

EC

P(n

=3)

No

EC

P(n

=52

)SM

(n=

16)

No

SM(n

=39

)SV

I(n

=8)

No

SVI

(n=

47)

No

OC

(n=

34)

OC

(n=

21)

Hig

hex

pres

sion

27/

34(8

0%)

14/

18(7

7%)

3/3

(100

%)

42/

52(8

0%)

14/

16(8

7.5%

)31

/39

(80%

)7/

8(8

7.5%

)39

/47

(82%

)27

/34

(79%

)16

/21

(76%

)

Low ex

pres

sion

6/34

(17%

)4/

18(2

3%)

0/3

(0%

)9/

52(1

8%)

2/16

(12.

5%)

7/39

(18%

)1/

8(1

2.5%

)7/

47(1

4.8%

)6/

34(1

7.6%

)5/

21(2

4%)

No ex

pres

sion

1/34

(3%

)0/

18(0

%)

0/3

(0%

)1/

52(2

%)

0/16

(0%

)1/

39(2

%)

0/8

(0%

)1/

34(3

%)

1/34

(3%

)0/

21(0

%)

*Pri

mar

ypr

osta

tic

aden

ocar

cino

mas

wer

eex

amin

edfo

rcl

inic

alin

vasi

vene

ssan

dpp

32ex

pres

sion

.FC

P,fo

calc

apsu

lar

pene

trat

ion;

EC

P,ex

tens

ive

caps

ular

pene

trat

ion;

SM,

surg

ical

mar

gin

invo

lvem

ent;

SVI,

sem

inal

vesi

cle

inva

sion

;OC

,org

an-c

onfi

ned

.No

stat

isti

cally

sign

ific

ant

corr

elat

ions

wer

eid

enti

fied

(P>

0.05

).

pp32 Expression in Prostate Cancer 235

increase its synthesis or decrease its degradation. Ex-periments to distinguish between these possibilitiesare beyond the scope of the study. The method em-ployed in the present study detects pp32 through hy-bridization with a probe corresponding to roughly onethird of the pp32 cDNA sequence. Southern blotanalysis of genomic DNA with pp32-derived probesyields a complex pattern suggestive of multiple,closely related loci in the genome. It is, however, un-certain which and how many pp32-related loci are ex-pressed, since at least one appears to be a pseudogene[Rebel et al., unpublished observations]. Stringent hy-bridization with a 298-bp probe, as performed here,cannot exclude a small number of minor differences inthe target mRNA; nor does it rule out differences oc-curring outside the region encompassed by the probe.It is thus possible that the mRNA hybridizing with thepp32 probe in prostate cancer might represent mutantpp32, alternatively spliced pp32, or expression of aclosely related gene. These possibilities are currentlybeing addressed in a separate study characterizingpp32 mRNA in benign and neoplastic prostatic tissuesby RT-PCR, single-strand conformational polymor-phism analysis, and sequencing of the resultantcDNA. Regarding the question of whether the presentstudy was performed at the proper hybridizationstringency, it should be noted that the temperaturedependence of the hybridization signal in both normaland neoplastic tissues was identical. Thus, any poten-tial differences between normal and neoplastic pp32mRNA would not be resolvable by this method.

Although pp32 and certain cytokeratins are bothdifferentially expressed in various prostatic cellularcompartments in development and in neoplasia, per-tinent differences exist. At one level, pp32 expressionis generally similar to that of CK5 and CK14, whichare expressed only by the basal cells in benign pros-tates and PIN lesions. Unlike CK5 and CK14, however,prostatic adenocarcinomas express high levels ofpp32, similar to the pattern found for CK 8 and CK 19[22]. Antibody 34bE12, which labels basal cells, reactsspecifically with high molecular weight cytokeratins 1,5, 10, and 14 [23,24]. In benign prostatic glands, pp32expression is similar to the staining pattern foundwith antibody 34bE12, but, unlike 34bE12, expressionpersists and is not lost in cancer. To summarize, pp32thus provides information distinct from that providedby cytokeratin analysis and may thereby offer an at-tractive method of diagnosing malignant prostateglands in doubtful cases.

The expression pattern of pp32 in the human pros-tate supports previous work associating pp32 expres-sion with putative prostatic epithelial stem cells [8]. Aswith pp32, other molecules normally found only inprostatic stem cells, such as telomerase [21,25] and

bcl-2 [26–28], are expressed at high levels in prostatecancer. Unlike bcl-2, found predominantly in hor-mone-refractory prostate tumors, pp32 is expressed inalmost all primary prostate cancers. Together withtelomerase [22] and the loss of glutathione-S-trans-ferase pi [21], pp32 expression constitutes one of themost frequently observed molecular alterations in pri-mary prostate cancer.

CONCLUSIONS

The results of this study suggest that pp32 may bediagnostically useful through its ability to mark pros-tatic adenocarcinoma of Gleason score ù5. Further-more, pp32 may contribute mechanistically to prostatetumor development, since in comparison to other mo-lecular alterations, increased pp32 expression is one ofthe most frequent events in primary prostate cancer.

REFERENCES

1. Parker S, Tong T, Bolden S, Wingo P: Cancer statistics. CA Can-cer J Clin 1996;46:8–27.

2. Potosky AL, Miller BA, Albertsen PC, Kramer BS: The role ofincreasing detection in the rising incidence of prostate cancer.JAMA 1995;273:548–552.

3. Ware JL: Prostate cancer progression: Implications of histopa-thology. Am J Pathol 1994;145:983–993.

4. Gleason DF: Histologic grading of prostate cancer: A perspec-tive. Hum Pathol 1992;23:273–279.

5. Fleming WH, Hamel A, McDonald R, Ramsey E, Pettigrew NM,Johnston B, Dodd JG, Matusik RJ: Expression of c-myc proto-oncogene in human prostate carcinoma and BPH. Cancer Res1986;46:1535–1538.

6. Visakorpi T, Kallioniemi AH, Syvanen AC, Hyytinen ER, KarhuR, Tammela T, Isola JJ, Kallioniemi OP: Genetic changes in pri-mary and recurrent prostate cancer by comparative genomichybridization. Cancer Res 1995;55:342–347.

7. Kadkol SS, Rebel JM, Brody JR, Kuhajda FP, Pasternack GR:Expression of pp32 in human benign prostatic hyperplasia andprostatic adenocarcinoma. Cancer Mol Biol 1996;3:905–913.

8. Walensky LD, Coffey DS, Chen T-H, Wu T-C, Pasternack GR: Anovel M(r) 32,000 nuclear phosphoprotein is selectively ex-pressed in cells competent for self-renewal. Cancer Res 1993;53:4720–4726.

9. Malek SN, Katumuluwa AI, Pasternack GR: Identification andpreliminary characterization of two related proliferation-associated nuclear phosphoproteins. J Biol Chem 1990;265:13400–13409.

10. Chen T-H, Brody JR, Romantsev FE, Yu J-G, Kayler AE, VoneiffE, Kuhajda FP, Pasternack GR: Structure of pp32, an acidicnuclear protein which inhibits oncogene-induced formation oftransformed foci. Mol Biol Cell 1996;7:2045–2056.

11. Chen T-H, Romantsev FE, Furuya Y, Isaacs JT, Voneiff E, Ku-hajda FP, Pasternack GR: pp32 is a nuclear phosphoproteinwhich inhibits ras-myc transformation and drug-induced apop-tosis. Mol Biol Cell [Suppl] 1994;5:146.

12. Sambrook J, Fritsch EF, Maniatis T: ‘‘Molecular Cloning. ALaboratory Manual,’’ 2nd ed. New York: Cold Spring HarborLaboratory Press, 1989.

236 Kadkol et al.

13. Ehrlein J, Wanke R, Weis S, Brem G, Hermanns W: Sensitivedetection of human growth hormone mRNA in routinely for-malin-fixed, paraffin-embedded transgenic mouse tissues bynon-isotopic in situ hybridization. Histochemistry 1994;102:145–152.

14. Gumerlock PH, Poonamallee UR, Meyers FJ, deVere White RW:Activated ras alleles in human carcinoma of the prostate arerare. Cancer Res 1991;51:1632–1637.

15. Kuhn EJ, Kurnot RA, Sesterhenn IA, Chang EH, Moul JW: Ex-pression of the c-erbB-2 (HER-2/Neu) oncoprotein in humanprostatic carcinoma. J Urol 1993;150:1427–1433.

16. Kallioniemi OP, Visakorpi T: Genetic basis and clonal evolutionof human prostate cancer. Adv Cancer Res 1996;68:225–255.

17. Bookstein R, MacGrogan D, Hilsenbeck SG, Sharkey F, AllredDC: p53 is mutated in a subset of advanced-stage prostate can-cers. Cancer Res 1993;53:3369–3373.

18. Isaacs WB, Bova GS, Morton RA, Bussemakers MJG, Brooks JD,Ewing CM: Genetic alterations in prostate cancer. Cold SpringHarbor Symp Quant Biol 1994;59:653–659.

19. Bookstein R, Rio P, Madreperla SA, Hong F, Allred C, GrizzleWE, Lee W-H: Promoter deletion and loss of retinoblastomagene expression in human prostate carcinoma. Proc Natl AcadSci USA 1990;87:7762–7766.

20. Lee WH, Morton RA, Epstein JI, Brooks JD, Campbell PA, BovaGS, Hsieh WS, Isaacs WB, Nelson WG: Cytidine methylation ofregulatory sequences near the pi-class glutathione S-transferasegene accompanies human prostatic carcinogenesis. Proc NatlAcad Sci USA 1994;91:11733–11737.

21. Sommerfeld HJ, Meeker AK, Piatyszek MA, Bova GS, Shay JW,Coffey DS: Telomerase activity: A prevalant marker of malig-nant human prostate tissue. Cancer Res 1996;56:218–222.

22. Yang Y, Hao J, Liu X, Dalkin B, Nagle RB: Differential expres-sion of cytokeratin mRNA and protein in normal prostate, pros-tatic intraepithelial neoplasia, and invasive carcinoma. Am JPathol 1997;150:693–704.

23. Gown AM, Vogel AM: Monoclonal antibodies to human inter-mediate filament proteins: II. Distribution of filament proteinsin normal human tissues. Am J Pathol 1984;114:309–321.

24. Wojno KJ, Epstein JI: The utility of basal cell-specific anti-cytokeratin antibody (34 betaE12) in the diagnosis of prostatecancer. A review of 228 cases. Am J Surg Pathol 1995;19:251–260.

25. Meeker AK, Sommerfeld HJ, Coffey DS: Telomerase is activatedin the prostate and seminal vesicles of the castrated rat. Endo-crinology 1996;137:5743–5746.

26. Rouleau M, Leger J, Tenniswood M: Ductal heterogeneity ofcytokeratins, gene expression, and cell death in the rat ventralprostate. Mol Endocrinol 1990;4:2003–2013.

27. McDonnell TJ, Troncoso P, Brisbay SM, Logothetis C, ChungLW, Hsieh JT, Tu SM, Campbell ML: Expression of the protoon-cogene bcl-2 in the prostate and its association with emergenceof androgen-independent prostate cancer. Cancer Res 1992;52:6940–6944.

28. Colombel M, Symmans F, O’Toole KM, Chopin D, Benson M,Olsson CA, Korsmeyer S, Buttyan R: Detection of the apoptosis-supressing oncoprotein bcl-2 in hormone-refractory humanprostate cancers. Am J Pathol 1993;143:390–400.

pp32 Expression in Prostate Cancer 237