Prognostic Significance of b-MicroseminoproteinmRNA Expression in Prostate Cancer

Hideki Sakai,1* Toshifumi Tsurusaki,1 Shigeru Kanda,1 Takehiko Koji,2Jim W. Xuan,3 and Yutaka Saito1

1Department of Urology, Nagasaki University School of Medicine, Nagasaki, Japan2Department of Histology and Cell Biology, Nagasaki University School of Medicine,

Nagasaki, Japan3Department of Surgery, University of Western Ontario, London, Ontario, Canada

BACKGROUND. Human b-microseminoprotein (b-MSP or PSP94) is a small protein se-creted by prostatic epithelial cells. We recently reported the presence of low levels of b-MSPmRNA expression and protein in most prostate cancer tissues.METHODS. b-MSP and mRNA expression was examined by in situ hybridization in biopsyspecimens obtained from 92 patients with prostate cancer. All tissue specimens were obtainedby needle biopsies prior to treatment. All patients subsequently received endocrine therapy.To estimate the influence of b-MSP mRNA expression and three possible prognostic factors,i.e., patient age, clinical stage, and Gleason score, on time to progression under endocrinetherapy, univariate and multivariate analyses were performed using Coxs proportional haz-ards regression model.RESULTS. Multivariate survival analysis showed that clinical stage was the strongest prog-nostic factor (P =0.006) and that b-MSP mRNA expression was the second strongest factor (P= 0.038) in 92 patients with stage BD disease. Analysis of only 51 patients with stage Ddisease showed that b-MSP mRNA expression was the only significant prognostic indicatorfor progression under endocrine therapy (P = 0.003).CONCLUSIONS. The presence of cells that express the b-MSP transcript may be a novelindicator of potentially aggressive prostate cancer. Prostate 38:278284, 1999. 1999 Wiley-Liss, Inc.

KEY WORDS: prostate cancer; b-microseminoprotein; PSP94; mRNA; prognosis

INTRODUCTION

Prostate cancer has a high incidence and is the sec-ond leading cause of death from cancer among men inthe Western world [1]. Identification of factors thatpredict clinical outcome is important but can be diffi-cult. In general, the clinical stage and histologicalgrade correlate to the biological behavior of prostatecancer, and are considered as the most useful prog-nostic factors [2,3]. Several studies have examined therole of other possible prognostic indicators, such asage [2], extent of disease on a bone scan [4], DNAploidy [5], and the expression of prostate-specific an-tigen (PSA) and prostatic acid phosphatase (PAP) intumor cells [6,7]. Recently, other potential prognosticmarkers have also been reported [8].

Human b-microseminoprotein (b-MSP), alsoknown as prostatic inhibin peptide (PIP), b-inhibin, orprostatic secretory protein of 94 amino acids (PSP94),is one of the major proteins secreted by prostatic epi-thelial cells [911]. b-MSP is a small (14 kDa) cysteine-rich protein, abundant in the seminal plasma [911],and detectable in serum and urine [12,13]. In contrastto PSA and PAP, regulation of b-MSP expression isindependent of androgens [14]. The b-MSP gene is

Grant sponsor: Ministry of Education, Science, Sports, and Cultureof Japan; Grant number: 06404059.*Correspondence to: Hideki Sakai, M.D., Department of Urology,Nagasaki University School of Medicine, 1-7-1 Sakamoto, Nagasaki852-8501, Japan. E-mail: hsakai@net.nagasaki-u.ac.jpReceived 19 February 1998; Accepted 3 August 1998

The Prostate 38:278284 (1999)

1999 Wiley-Liss, Inc.

located on chromosome 10q11.2 [15]. Although b-MSPwas initially identified as being specific to the pros-tate, immunohistochemical and Northern analyseshave since detected b-MSP secretion in other mucus-producing cells, e.g., in tracheal epithelium [16]. Thus,b-MSP is not currently considered a prostate-specificprotein. Its biological function, however, remains un-known. It has been demonstrated that b-MSP can sup-press DNA synthesis and growth of prostate cells invitro and in vivo [17,18]. It has also been reported thatb-MSP can induce apoptosis of prostate cancer cells inan animal model system [17]. These findings suggestthat b-MSP may be useful therapeutically.

We recently analyzed the expression of b-MSP atboth the mRNA and protein levels in 104 biopsy speci-mens of untreated prostate cancer, using in situ hy-bridization (ISH) and immunohistochemical staining[19]. Although no significant correlation was detectedbetween the expression of b-MSP mRNA or proteinand Gleason grade, we found that levels of b-MSPmRNA expression and protein were low in most pros-tate cancer tissues. These results suggest that reducedexpression of b-MSP may play an important role in thedevelopment of prostate cancer.

In the present study, we attempted to correlateb-MSP mRNA expression with progression-free sur-vival of prostate cancer patients treated with endo-crine therapy. Multivariate survival analysis demon-strated that in patients with advanced prostate cancer,b-MSP mRNA expression was an independent prog-nostic indicator for progression during endocrinetherapy.

MATERIALS AND METHODS

Patients and Tissues

We selected 92 patients with prostate cancer from agroup treated with endocrine therapy at NagasakiUniversity Hospital. The inclusion criteria were theavailability of a sufficient tumor biopsy tissue andwell-documented clinical follow-up. All patients ana-lyzed in this study were also included in our previousstudy [19]. Diagnosis of prostate cancer was made be-tween June 1984March 1996. The mean age of ourpatients with untreated prostate cancer at diagnosiswas 72.7 years (range, 4790 years). In 12 patients,cancer was localized within the prostate (stage B, orT2N0M0 [20]), in 29 the cancer was locally advanced(stage C, or T34N0M0), and metastatic disease wasfound in the remaining 51 (stage D, or N13 or M1).

All patients received endocrine therapy after theinitial diagnosis. Fifty patients were treated with a lu-teinizing hormone-releasing hormone agonist and theremainder were treated with estrogen or a steroidal

antiandrogen, chlormadinone acetate. Disease pro-gression was estimated based on evidence of bio-chemical failure and/or evidence of clinical progres-sion. Biochemical failure was defined as an increase inPSA values on two or more consecutive follow-up vis-its. Clinical progression was defined as an increase inany previously measurable lesion by more than 25% intwo perpendicular diameters, or appearance of a newmetastatic tumor on chest X-ray, CT scan, or bonescintigraphy.

All tissue specimens were obtained by transperi-neal or transrectal needle biopsies using 16- or 18-gauge needles prior to administration of any treat-ment. A systematic sextant biopsy was performed inmost cases. In addition, five control samples compris-ing two benign prostatic hyperplasia (BPH) and threenormal prostate tissue samples were obtained frombiopsies performed due to a strong suspicion of pros-tate cancer based on elevated serum PSA levels and/or abnormal findings at either digital rectal examina-tion or transrectal ultrasonography. Informed consentwas obtained from every patient prior to biopsy. Tis-sue samples were fixed in 10% neutral bufferd forma-lin and embedded in paraffin. Sections (6 mm in thick-ness) were mounted on 3-aminopropyltriethoxysi-lane-coated slides. Biopsy specimens were stainedwith hematoxylin and eosin for histopathological ex-amination. The grade of malignancy was determinedin each sample by using the Gleason grading system[21], and the tissue area containing the primary sitewas selected for further investigation. These malig-nant tissue samples usually contained normal tissue ortissue with benign hyperplasia, which were also usedas positive controls.

Preparation of Oligo-DNA Probes

A 44-base pair sequence complementary to b-MSPmRNA (nucleotides 159202, part of exon 3) was se-lected due to the presence of a short form of b-MSPmRNA, which lacked exon 3 [22]. A computer-assistedsearch (GenBank Release 95.0, 1996) of the above an-tisense sequence and the sense sequence showed nosignificant homology with any known sequences.These antisense and sense sequences were synthesizedtogether with two and three TTA repeats at the 58- and38-ends of sequences, and used as probes after hap-tenization with a T-T dimer. The T-T dimer was intro-duced into oligo-DNAs by ultraviolet irradiation (254nm), using a dose of 12,000 J/m2. The generation of aT-T dimer was verified immunochemically using amouse monoclonal anti-T-T IgG (Kyowa Medex, To-kyo, Japan). Furthermore, a dot-blot hybridizationstudy showed that the antisense probe synthesized inour laboratory was specific and had adequate sensi-

b-MSP mRNA and Prognosis in Prostate Cancer 279

tivity to be useful in ISH studies [19]. Preliminary ex-periments were also conducted on benign tissue speci-mens to confirm the sequence specificity of the b-MSPmRNA signal.

In Situ Hybridization

ISH was performed according to the method de-scribed previously [23]. Briefly, sections were depar-affinized and rehydrated by standard procedures.This was followed by treatment with 0.3% H2O2 inmethanol for 15 min to inactivate endogenous peroxi-dase, 0.2 N HCl at room temperature for 20 min, and100 mg/ml proteinase K at 37C for 15 min. After fixa-tion with 4% paraformaldehyde in phosphate-buffered saline (PBS) for 5 min, sections were im-mersed in 2 mg/ml glycine in PBS for 30 min andstored in 40% deionized formamide in 4 SSC (1 SSC: 0.015 M sodium citrate, pH 7.0, supplementedwith 0.15 M NaCl) until used for hybridization. Hy-bridization was carried out overnight at 37C with 2mg/ml T-T-dimerized antisense oligo-DNA forb-MSP, dissolved in a hybridization medium contain-ing 10 mM Tris/HCl (pH 7.4), 0.6 M NaCl, 1 mMEDTA, 1 Denhardts solution, 250 mg/ml yeasttRNA, 125 mg/ml salmon testicular DNA, 10% dex-tran sulfate, and 40% deionized formamide. In thenext step, the slides were washed with 50% formam-ide in 2 SSC containing 0.075% Brij35 followed byPBS. The sections were stained immunohistochemical-ly as described previously [23], and the sites of per-oxidase activity were visualized using a solution con-taining 3,38-diaminobenzidine-4 HCl (DAB), H2O2,CoCl2, and NiSO4(NH4)2SO4 [24].

The hybridization signal was considered positivewhen the accumulated black deposits in individualcells exceeded the background level and more than20% of carcinoma cells showed such a distinct signal.When staining was similar to that of the negative con-trol, the sample was regarded as negative. On theother hand, weak positive staining was assignedwhen the signal was present, but localized to the peri-nuclear area only. Strong positive staining repre-sented a strong and clear signal in the cytoplasm. Anumber of consecutive tissue sections were hybridizedwith T-T-dimerized oligo-DNA complementary to 28SrRNA as a positive control [25] and also with T-T-dimerized b-MSP sense oligo-DNA as a negative con-trol in every run. All specimens were independentlyevaluated by two observers who were blind to theclinical outcome.

Statistical Analysis

A computer software package, StatView 4.5 (Aba-cus Concepts, Inc., Berkeley, CA), was used for statis-

tical analysis. Difference in mean age between groupswas analyzed by Students t-test. The x2 test was usedto analyze the correlation between b-MSP mRNA ex-pression and the clinical stage or Gleason score. Pro-gression-free survival curves were calculated accord-ing to the Kaplan-Meier method, and statistical sig-nificance was examined using the log rank test. Coxsproportional hazards regression model was also usedfor univariate and multivariate survival analyses.

RESULTS

Expression of b-MSP mRNA in prostate tissue wasexamined by ISH [19]. All five control specimens wereevaluated as strongly positive, i.e., the glandular andductal epithelial cells were positive for b-MSP mRNA,whereas other cells such as basal, stromal, endothelial,and vascular smooth muscle cells were negative.Staining for b-MSP mRNA was localized to the cyto-plasmic area of epithelial cells, predominantly in theperinuclear area. In contrast, prostate cancer speci-mens exhibited b-MSP mRNA staining which was het-erogenous or scattered in most of the malignant epi-thelia. Typical staining patterns of b-MSP mRNA-strongly positive and -negative malignant tissues areshown in Figures 1 and 2, respectively.

Of the 92 cancerous specimens, 63 were negative, 18weakly positive, and 11 strongly positive for b-MSPmRNA. Accordingly, patients were divided into two

Fig. 1. b-MSP mRNA expression of a grade 5 prostate cancertissue, using nonradioactive in situ hybridization. A: Hematoxylinand eosin staining. B: 28S rRNA antisense probe, as positive con-trol. C: b-MSP antisense probe. b-MSP mRNA showed strong butdiffuse expression in the cytoplasmic area of only malignant epi-thelia. D: b-MSP sense probes, as negative control (AD, 200).These photomicrographs also appeared in our previous report[19].

280 Sakai et al.

groups: 63 negative and 29 positive for b-MSP mRNAexpression. The correlations between b-MSP mRNAexpression and patient age, clinical stage, or Gleasonscore were analyzed. There was no significant differ-ence in age between the two groups (Students t-test, P= 0.91), and no significant correlation was noted be-tween b-MSP mRNA expression and clinical stage (P= 0.27), or between b-MSP mRNA expression andGleason score (x2 test, P = 0.45, Table I). As expected,however, the Gleason score significantly correlatedwith the clinical stage (P = 0.009).

The mean and median follow-up periods of pa-tients treated with endocrine therapy were 34 and 29months, respectively. Biochemical and/or clinical pro-gression was observed in 44 patients (48%). Univariateand multivariate survival analyses with Coxs regres-sion model were used to evaluate the influence ofb-MSP mRNA expression and three possible prognos-tic factors, i.e., patient age, clinical stage, and Gleasonscore, on time to progression. In the univariate analy-sis where all 92 patients treated with endocrinetherapy were included, the clinical stage (P = 0.0006)and Gleason score (P = 0.001) were considered impor-tant prognostic factors. Multivariate analysis of thefour variables showed that clinical stage had thestrongest effect on progression-free survival (P =0.006), while b-MSP mRNA expression (P = 0.038) wasthe second most significant factor associated with pro-gression-free survival (Table II).

Because the clinical stage correlated strongly withtime to progression in hormonally-treated prostatecancer patients, further survival analyses were per-formed in only 51 patients with stage D disease. Uni-variate analysis including this group of patientsshowed only b-MSP mRNA expression as the signifi-cant variable (P = 0.007), whereas patient age (P = 0.14)and Gleason score (P = 0.16) did not influence pro-gression-free survival. Multivariate analysis of thethree variables showed that only b-MSP mRNA ex-

pression was a significant determinant of prognosis (P= 0.003, Table III). The progression-free survivalcurves according to b-MSP mRNA expression in stageD prostate cancer patients are shown in Figure 3. Asignificantly worse progression-free survival rate wasnoted in patients with positive b-MSP mRNA expres-sion (P = 0.005 by log rank test). Thus, positive b-MSPmRNA expression significantly correlated with a shorttime to progression in advanced prostate cancer pa-tients treated with endocrine therapy.

DISCUSSION

Multivariate survival analyses in the present studyshowed that hormonally-treated stage D prostate can-cer patients with positive b-MSP mRNA expressionhad a significantly poorer outcome in terms of pro-gression-free survival than those with negative ex-pression (P = 0.003). In contrast, when all patients withstage BD disease were included in such analysis,b-MSP mRNA expression showed only a borderlineprognostic value (P = 0.038). The prognosis for pros-tate cancer is generally considered to depend on theclinical stage and tumor grade [2,3]. In this series ofpatients, clinical stage and tumor grade (Gleasonscore) also had strong prognostic value in the univari-ate analysis. However, the Gleason score was not asignificant independent prognostic factor in the mul-tivariate analysis, as can be explained by the strong

Fig. 2. A representative example of b-MSP mRNA-negativeprostate cancer (grade 4) (200).

TABLE I. Correlation Between b-MSP mRNAExpression and Age, Clinical Stage, and Gleason Score*

b-MSP mRNA

Negative (%) Positive (%)

Mean age (years) 72.8 72.6Clinical stage

B (T2N0M0) 10 (16) 2 (7)C (T34N0M0) 17 (27) 12 (41)D (N13 or M1) 36 (57) 15 (52)

Gleason score4 4 (6) 1 (3)5 2 (3) 0 (0)6 5 (8) 3 (10)7 16 (25) 5 (17)8 15 (24) 6 (21)9 17 (27) 8 (28)

10 4 (6) 6 (21)Total 63 (100) 29 (100)

*There was no significant difference in patient age between thetwo groups by Students t-test (P = 0.91), and no significantcorrelation was noted between b-MSP mRNA expression andclinical stage (P = 0.27), and between b-MSP mRNA expressionand Gleason score (P = 0.45) by x2 test.

b-MSP mRNA and Prognosis in Prostate Cancer 281

correlation between Gleason score and clinical stage.Thus, although the presence of metastasis had thestrongest predictive value in all patients treated withendocrine therapy, among those with metastasis, posi-tive b-MSP mRNA expression indicated a short inter-val to progression. Thus, our findings suggest that thepresence of cells that express the b-MSP transcriptmay be an indicator of potentially more aggressiveprostate cancer.

Although a number of prognostic factors for pros-tate cancer have been reported, e.g., proliferating cellnuclear antigen [26], p53, bcl-2 [27], E-cadherin,b-catenin [28], and vascular density [29], most of thesefactors correlated with tumor grade assessed by vari-ous criteria. Thus, such prognostic factors have onlyan incidental significance, because as mentionedabove, tumor grade generally has a strong prognosticvalue for prostate cancer. On the other hand, the pre-sent study demonstrated that b-MSP mRNA expres-sion did not correlate with tumor grade as assessed bythe Gleason grading system. Furthermore, multivari-ate analysis showed that b-MSP mRNA expressionwas a significant predictor for failure of endocrinetherapy in prostate cancer patients with metastasis.Therefore, b-MSP mRNA expression is considered tobe a truly independent prognostic indicator for ad-vanced prostate cancer patients treated with endo-crine therapy.

Initially we attempted to correlate the expression ofb-MSP protein and b-MSP mRNA with prognosis.However, we abandoned the idea of performing suchanalysis using b-MSP protein expression because of anextremely low frequency of b-MSP-positive prostatecancer. Only 6 patients (7%) of 92 included in thisstudy had immunohistochemically positive b-MSP ex-pression (data not shown).

The reason for the correlation between positive ex-pression of b-MSP mRNA and poor prognosis in hor-monally-treated prostate cancer patients is unknown

TABLE II. Univariate and Multivariate Survival Analyses for 92 Prostate Cancer Patients Treated With EndocrineTherapy in Coxs Regression Model

Variable

Univariate Multivariate

Hazard ratio95% confidence

interval P value Hazard ratio95% confidence

interval P value

Age 0.964 0.9301.000 0.049 0.971 0.9351.007 0.11Clinical stage 2.762 1.5484.930 0.0006 2.466 1.3004.678 0.006Gleason score 1.519 1.1811.952 0.001 1.238 0.9391.633 0.13b-MSP mRNA expression 1.711 0.9183.192 0.091 2.043 1.0424.008 0.038

TABLE III. Univariate and Multivariate Survival Analyses for 51 Patients With Stage D Prostate Cancer Treated WithEndocrine Therapy in Coxs Regression Model

Variable

Univariate Multivariate

Hazard ratio95% confidence

interval P value Hazard ratio95% confidence

interval P value

Age 0.971 0.9351.009 0.14 0.957 0.9151.000 0.050Gleason score 1.248 0.9161.701 0.16 1.169 0.8431.623 0.35b-MSP mRNA expression 2.883 1.3386.211 0.007 3.436 1.5027.856 0.003

Fig. 3. Kaplan-Meier progression-free survival curves accordingto b-microseminoprotein (b-MSP) mRNA expression in 51 hor-monally-treated prostate cancer patients with metastasis. Theb-MSP mRNA-positive group exhibited a significantly shorter pe-riod to progression than the b-MSP mRNA-negative group (P =0.005 by log rank test).

282 Sakai et al.

at present. We previously analyzed the expression ofb-MSP at both the mRNA and protein levels in 104biopsy specimens of untreated prostate cancer, andfound a lower level of expression of b-MSP in prostatecancer tissue compared with benign prostate tissue[19]. We also suggested that this phenomenon mightbe mainly due to the presence of reduced levels ofb-MSP mRNA [19]. Based on these findings, we ex-pected in the present study that patients with highexpression of b-MSP mRNA would show a betterprognosis than those with low levels of expression. Incontrast to our expectations, in the present study aworse progression-free survival rate was noted in pa-tients with positive b-MSP mRNA expression. Al-though most of the transformed prostatic epithelialcells lose the ability to express b-MSP mRNA and pro-duce its protein, some carcinoma cells with a highbiological potential may continue to express b-MSPmRNA through an aberrant intracellular signal trans-duction. Furthermore, our finding of a low frequencyof b-MSP-positive prostate cancer suggests that insuch malignant cells the mRNA may be hardly trans-lated into the wild type b-MSP. In a series of prelimi-nary immunohistochemical studies from our labora-tory, we recently analyzed the expression of b-MSP insurgical specimens obtained from patients treatedwith prostatectomy after neoadjuvant hormonaltherapy. Interestingly, although none of the prostatecancers treated with androgen deprivation therapyshowed positive PSA staining, a number of high-gradetumors (Gleason grade 4 and 5) showed positiveb-MSP expression (unpublished results). While thephysiological consequences of changes in b-MSP ex-pression in malignant cells of the prostate remain to beelucidated, we speculate that the presence of carci-noma cells expressing b-MSP mRNA may contributeto the malignant potential or the androgen-independent growth of the tumor. In this regard, as-sessment of b-MSP mRNA expression in prostate can-cer tissue samples may be useful for selection of pa-tients who should be treated aggressively, e.g., withcytotoxic chemotherapy. To establish the role ofb-MSP mRNA as a prognostic indicator, further stud-ies, including a prospective cohort study, are needed.

CONCLUSIONS

Multivariate survival analysis showed that in pros-tate cancer patients with metastasis, b-MSP mRNAexpression was an independent prognostic determi-nant of survival during endocrine therapy, and thatthis prognostic indicator was independent of histo-logical grade. Assessment of b-MSP mRNA expres-sion in prostate cancer tissue samples may be useful

for the identification of a subgroup of patients withaggressive malignant prostate cancer.

REFERENCES

1. Boyle P, Maisonneuve P, Napalkov P. Geographical and tem-poral patterns of incidence and mortality from prostate cancer.Urology [Suppl] 1995;46:4755.

2. Byar DP, Corle DK. Analysis of prognostic factors for prostaticcancer in the VACURG studies. In: Denis L, Murphy GP, ProutGR Jr, Schroder FH, editors. Controlled clinical trials in urologiconcology. New York: Raven Press; 1984. p 147169.

3. Partin AW, Steinberg GD, Pitcock RV, Wu L, Piantadosi S, Cof-fey DS, Epstein JI. Use of nuclear morphometry, Gleason histo-logic scoring, clinical stage, and age to predict disease-free sur-vival among patients with prostate cancer. Cancer 1992;70;161168.

4. Soloway MS, Hardeman SW, Hickey D, Raymond J, Todd B,Soloway S, Moinuddin M. Stratification of patients with meta-static prostate cancer based on extent of disease on initial bonescan. Cancer 1988;61:195202.

5. Pollack A, Troncoso P, Zagars GK, von Eschenbach AC, MakAC, Wu CS, Terry NHA. The significance of DNA-ploidy andS-phase fraction in node-positive (stage D1) prostate cancertreated with androgen ablation. Prostate 1997;31:2128.

6. Sakai H, Shiraishi K, Minami Y, Yushita Y, Kanetake H, Saito Y.Immunohistochemical prostatic acid phosphatase level as aprognostic factor of prostatic carcinoma. Prostate 1991;19:265272.

7. Sakai H, Yogi Y, Minami Y, Yushita Y, Kanetake H, Saito Y.Prostate specific antigen and prostatic acid phosphatase immu-noreactivity as prognostic indicators of advanced prostatic car-cinoma. J Urol 1993;149:10201023.

8. Gao X, Porter AT, Grignon DJ, Pontes JE, Honn KV. Diagnosticand prognostic markers for human prostate cancer. Prostate1997;31:264281.

9. Mbikay M, Nolet S, Fournier S, Benjannet S, Chapdelaine P,Paradis G, Dube JY, Tremblay R, Lazure C, Seidah NG, ChretienM. Molecular cloning and sequence of the cDNA for a 94-amino-acid seminal plasma protein secreted by the human prostate.DNA 1987;6:2329.

10. Seidah NG, Arbatti NJ, Rochemont J, Sheth AR, Chretien M.Complete amino acid sequence of human seminal plasma b-in-hibin. FEBS Lett 1984;175:349355.

11. Akiyama K, Yoshioka Y, Schmid K, Offner GD, Troxler RF,Tsuda R, Hara M. The amino acid sequence of human b-micro-seminoprotein. Biochim Biophys Acta 1985;829:288294.

12. Teni TR, Sheth AR, Kamath MR, Sheth NA. Serum and urinaryprostatic inhibin-like peptide in benign prostatic hyperplasiaand carcinoma of prostate. Cancer Lett 1988;43:914.

13. Garde SV, Sheth AR, Venkatesan VM, Panchal CJ, Porter AT,Grignon DJ. Prostate inhibin peptide (PIP) in prostate cancer: acomparative immunohistochemical study with prostate specificantigen (PSA) and prostatic acid phosphatase (PAP). CancerLett 1994;78;1117.

14. Hurkadh KS, Lokeshwar B, Sheth AR, Block NL. Detection ofprostatic-inhibin-like peptide in the cytoplasm of LNCaP cells, ahuman prostatic adenocarcinoma cell-line. Prostate 1994;24:285290.

15. Sasaki T, Matsumoto N, Jinno Y, Niikawa N, Sakai H, KanetakeH, Saito Y. Assignment of the human b-microseminoproteingene (MSMB) to chromosome 10q11.2. Cytogenet Cell Genet1996;72:177178.

16. Weiber H, Andersson C, Murne A, Rannevik G, Lindstrom C,

b-MSP mRNA and Prognosis in Prostate Cancer 283

Lilja H, Fernlund P. b-microseminoprotein is not a prostate-specific protein. Am J Pathol 1990;137:593603.

17. Mundle SD, Sheth NA. Suppression of DNA synthesis and in-duction of apoptosis in rat prostate by human seminal plasmainhibin (HSPI). Cell Biol Int 1993;17:587594.

18. Lokeshwar BL, Hurkadli KS, Sheth AR, Block NL. Human pros-tatic inhibin suppresses tumor growth and inhibits clonogeniccell survival of a model prostatic adenocarcinoma, the DunningR3327G rat tumor. Cancer Res 1993;53:48554859.

19. Tsurusaki T, Koji T, Sakai H, Kanetake H, Nakane PK, Saito Y.Cellular expression of beta-microseminoprotein (b-MSP)mRNA and its protein in untreated prostate cancer. Prostate1998;35:109116.

20. Schroder FH, Hermanek P, Denis L, Fair WR, GospodarowiczMK, Pavone-Macaluso M. The TNM classification of prostatecancer. Prostate [Suppl] 1992;4:129138.

21. Gleason DF. Histologic grading and clinical staging of prostaticcarcinoma. In: Tannenbaum M, editor. Urologic pathology: theprostate. Philadelphia: Lea & Febiger; 1977. p 171198.

22. Xuan JW, Chin JL, Guo Y, Chambers AF, Finkelman MA, ClarkeMW. Alternative splicing of PSP94 (prostatic secretory proteinof 94 amino acids) mRNA in prostate tissue. Oncogene 1995;11:10411047.

23. Koji T, Nakane PK. Recent advances in molecular histochemicaltechniques: in situ hybridization and southwestern histochem-istry. J Electron Microsc (Tokyo) 1996;45:119127.

24. Adams JC. Heavy metal intensification of DAB-based HRP re-action product. J Histochem Cytochem 1981;29:775.

25. Yoshii A, Koji T, Ohsawa N, Nakane PK. In situ localization ofribosomal RNAs in a reliable reference for hybridizable RNA intissue sections. J Histochem Cytochem 1995;43:321327.

26. Harper ME, Glynne-Jones E, Goddard L, Wilson DW, Matenhe-lia SS, Conn IG, Peeling WB, Griffiths K. Relationship of prolif-erating cell nuclear antigen (PCNA) in prostatic carcinomas tovarious clinical parameters. Prostate 1992;20:243253.

27. Matsushima H, Kitamura T, Goto T, Hosaka Y, Homma Y,Kawabe K. Combined analysis with bcl-2 and p53 immunostain-ing predicts poorer prognosis in prostatic carcinoma. J Urol1997;158:22782283.

28. Richmond PJM, Karayiannakis AJ, Nagafuchi A, Kaisary AV,Pignatelli M. Aberrant E-cadherin and b-catenin expression inprostate cancer: correlation with patient survival. Cancer Res1997;57:31893193.

29. Lissbrant IF, Stattin P, Damber J-E, Bergh A. Vascular density isa predictor of cancer-specific survival in prostatic carcinoma.Prostate 1997;33:3845.

284 Sakai et al.