Effect of androgen deprivation therapy on the expression of prostate cancer biomarkers MSMB and MSMB-binding protein CRISP3

Abstract

We have investigated the effects of short-term neoadjuvant and long-term androgen deprivation therapies (ADTs) on β-microseminoprotein (MSMB) and cysteine-rich secretory protein-3 (CRISP3) expression in prostate cancer patients. We also studied if MSMB expression was related to genotype and epigenetic silencing. Using an Affymetrix cDNA microarray analysis, we investigated the expression of MSMB, CRISP3, androgen receptor (AR), KLK3 and Enhancer of Zeste Homologue-2 (EZH2) in tissue from prostate cancer patients receiving (n=17) or not receiving (n=23) ADT before radical prostatectomy. MSMB, CRISP3 and AR were studied in tissue from the same patients undergoing TURP before and during ADT (n=16). MSMB genotyping of these patients was performed by TaqMan PCR. MSMB and KLK3 expression levels decreased during ADT. Expression levels of AR and CRISP3 were not affected by short-term ADT but were high in castration-resistant prostate cancer (CRPC) and metastases. Levels of EZH2 were also high in metastases, where MSMB was low. Genotyping of the MSMB rs10993994 polymorphism showed that the TT genotype conveys poor MSMB expression. MSMB expression is influenced by androgens, but also by genotype and epigenetic silencing. AR and CRISP3 expression are not influenced by short-term ADT, and high levels were found in CRPC and metastases.

Introduction

Prostate cancer is the most common form of cancer and a leading cause of cancer-related deaths in men in Western countries.1 Patient prognosis will vary from a rapidly progressing disease with a high probability of death in a minority of patients, to a relatively indolent prostate cancer that can be controlled for the remainder of the patients life with little intervention in a majority of cases. Despite identification of a number of promising new biomarkers, there is still no absolute way of determining disease prognosis at the time of diagnosis. Inevitably, this leads to overtreatment of a large number of men for the benefit of the few who need it.

Attention has been focused on the predictory abilities of β-microseminoprotein (MSMB); also known as prostate-specific protein of 94 amino acids (PSP94). This protein is second only to PSA as the most predominant protein in seminal plasma.2, 3 The function of MSMB is largely unknown, but in vitro and in vivo studies suggest it may be involved in apoptosis, cell mobility, vascularization and tumor suppressor functions, all recently reviewed by Whitaker et al.4 The suitability of MSMB as a biomarker for prostate cancer has been reported by several groups during the past two decades. It has been reported that MSMB mRNA and protein expression is reduced in malignant prostatic epithelia compared with benign epithelia.5 In prostate cancer, the MSMB gene may be subjected to transcriptional silencing by methylation, mediated by enhancer of zeste homologue-2 (EZH2), a Polycomb group member which is often overexpressed in castration-resistant prostate cancer (CRPC).6, 7

MSMB expression has been reported by several groups to be negatively associated with disease-free survival8 and outcome.9 MSMB expression has also been proposed to be a significant prognostic indicator for prostate cancer progression under endocrine therapy.10 Recently, the MSMB gene has gained further attention as one of the primary candidate prostate cancer susceptibility genes,11, 12 and several causal risk alleles have been identified in the region upstream of the coding sequence.13, 14, 15 Serum levels of MSMB, its binding protein PSP94-binding protein (PSPBP) and the ratio free/bound MSMB, have all been suggested as independent prognostic factors in prostate cancer.16, 17, 18, 19

We have previously reported that the cysteine-rich secretory family (CRISP) family member CRISP3 also forms a high-affinity complex with MSMB in human seminal plasma.20 CRISP3 has been found to be one of the most upregulated genes in prostate cancer compared to benign tissue,21, 22 and was proposed to be a useful biomarker for prostate cancer.23, 24 Recently we reported that in a tissue microarray (TMA) with samples from 945 prostate cancer patients undergoing radical prostatectomy (RP), high CRISP3 and low MSMB expression were associated with poor outcome.9

As the impact of androgen availability on the expression of certain tumor biomarkers is currently unclear, we wanted to investigate how the candidate markers MSMB and CRISP3 are regulated by androgens during androgen deprivation therapy (ADT) and in progressive prostate cancer. First, we investigated the effect of short term neoadjuvant ADT on CRISP3 and MSMB gene expression in primary prostate cancer, as well as in a limited set of prostate cancer metastases. CRISP3 and MSMB protein levels were then evaluated in serial tissue samples from a small but unique cohort of men with prostate cancer undergoing TURP before and during long-term ADT. Finally, as MSMB expression may be differentially regulated in primary prostate cancers and in progressive disease, we also studied whether MSMB expression levels were related to a single nucleotide polymorphism (SNP) in the MSMB promotor region and associated with epigenetic silencing of MSMB by EZH2.

Materials and methods

Sample preparation and data analysis of Affymetrix U95 human gene array

Microarray tissue samples were obtained as previously described.25 In brief, tissue samples were collected from patients undergoing therapeutic or diagnostic procedures at Memorial Sloan-Kettering Cancer Center, New York, NY. All samples were snap-frozen, histologically analyzed, and tissues containing approximately 60–80% prostate cancer were manually dissected from the frozen block. Samples included 23 primary prostate cancer tissues from patients not receiving adjuvant therapy before RP, 17 primary prostate cancer tissues from patients collected at RP after three months of neoadjuvant ADT (monthly injections of 3.6 mg of goserelin and 250 mg flutamide three times daily). A limited set of 9 prostate cancer metastases (the secondary sites were in bone (two samples); lung (one sample) and lymph node (three samples)) and metastatic CRPC (three samples, sites not specified) were also included. Details of patient characteristics have previously been described.25

Gene expression analysis were performed as described previously.25 In brief, total RNA was extracted from frozen tissue by Trizol (Invitrogen, Carlsbad, CA, USA), purified and evaluated for integrity. Complementary DNA (cDNA) was synthesized from total RNA, and gene expression analysis was performed using Affymetrix U95 human gene arrays with 63 175 probes for genes and expressed sequence tags as described by the manufacturer (Affymetrix Inc., Santa Clara, CA, USA). Samples were analyzed using Affymetrix Microarray Suite version 4.0 as previously reported.25

Tissue specimens and clinical data from patients undergoing TURP

Formalin-fixed paraffin-embedded human tissue samples were obtained from 16 prostate cancer patients (mean age 70 years, range 52–78) undergoing repeated therapeutic TURP at the University Hospital Malmö, Sweden. Immunohistochemistry was performed as previously described,9 using antibodies and dilutions as described in Table 1. Descriptive characteristics of the patients are given in Table 2. Samples were collected between 1971 and 2001. From each patient, samples were collected at diagnosis, before ADT and from a later TURP due to local tumor progression. Nine patients were subjected to medical castration with GnRH analogue, one patient was treated with flutamide monotherapy and five patients underwent bilateral orchiectomy. Mean follow-up time was 73 months (range 12–167 months) and mean time of hormonal treatment was 37 months (range 3–118 months). Patients were considered to have progressive disease and CRPC at the second TURP date as levels of serum PSA or acid phosphatase were rising despite ongoing ADT. The study was approved by the local ethics committee at Lund University, Sweden.

Table 1 Antibody characteristics
Table 2 Patient description and tumor sample characteristics

Genotype analysis

Paraffin-embedded tissues from 13 of the 16 patients undergoing TURP were available for genotype analysis. Fresh sections were microscopically studied, and tissue was manually punched out of the paraffin block with a 2 mm biopsy needle. DNA was isolated using the QIAamp DNA FFPE tissue kit (Qiagen, Hilden, Germany) according to the manufacturer's protocol. The rs10993994 polymorphism was determined using TaqMan primers and probes (Applied Biosystems, Foster City, CA, USA), using 1 ng per μl template DNA in a reaction volume of 25 μl. The analysis was performed on a 7900HT Fast Real-Time PCR System (Applied Biosystems) according to the manufacturer's instructions.

Results

MSMB but not CRISP3 transcript levels decrease in response to short-term neoadjuvant ADT

Changes in expression of MSMB and CRISP3 genes due to androgen deprivation were analyzed using data from Affymetrix U95 gene expression microarrays. The clinical and pathological features related to the patients and tissue samples included in the study (patient age, preoperative serum PSA levels, Gleason score, pathological TNM stage and recurrence) have been described earlier.25 MSMB expression was found to be high in primary prostate cancer from patients not receiving adjuvant ADT, and expression was significantly decreased on ADT (Figure 1; Mann–Whitney P<0.001). No significant reduction in gene expression levels was observed in either CRISP3 or androgen receptor (AR) transcript levels in patients receiving neoadjuvant ADT. However both CRISP3 and AR were highly expressed in metastatic prostate cancer lesions (Figure 1). There was considerable interpatient variability in expression of both MSMB and CRISP3. Interestingly, the MSMB expression level and decrease during ADT were equal to those of KLK3, a well-known target of androgen signaling (Figure 1). As expected, KLK3 gene expression levels decreased in the group receiving neoadjuvant ADT compared with the nontreated group (Mann–Whitney P<0.001). However, as in the case of MSMB, considerable interpatient variability was detected.

Figure 1
figure1

Transcription profile changes before and after androgen deprivation therapy (ADT) in primary prostate cancer tumors and in metastases. Patients with primary prostate cancer undergoing routine surgery either received (n=17) or did not receive (n=23) neoadjuvant ADT before surgery. Transcription is represented as graphs and box plots for β-microseminoprotein (MSMB), cysteine-rich secretory protein-3 (CRISP3), androgen receptor (AR) (reproduced from Am J Pathol 2004, 164:217–227 with permission from the American Society for Investigative Pathology), KLK3 and Enhancer of Zeste Homologue-2 (EZH2). The cDNA array included two probe sets for AR, and the corresponding boxplot is based on the average.

Prostate cancer metastases express low transcript levels of MSMB, and high levels of CRISP3

A majority of prostate cancer patients develop CRPC within a few years following ADT. Such tumors often show highly upregulated levels of AR though the signaling may be dysfunctional.26, 27 To investigate MSMB and CRISP3 transcriptional levels in metastases, we studied gene expression in nine remote tumors, out of which three were from patients with CRPC.

Although caution is a prerequisite when interpreting small sample numbers, we found that MSMB expression is substantially decreased in metastatic tumors compared with primary prostate cancer, whereas CRISP3 is expressed at high levels in metastases (Figure 1). Interestingly, CRISP3 expression pattern appears to be somewhat similar to that of AR in that both transcipts are expressed at the same level and neither AR nor CRISP3 is affected by ADT (Figure 1).

Protein expression of MSMB and CRISP3 during ADT and prostate cancer progression

To evaulate the link between long-term ADT, progressing disease and the expression levels of MSMB and CRISP3 proteins, we used a small but unique set of serially collected tissue samples from 16 patients undergoing TURP before and during ADT. At the time of the second TURP, all patients were in biochemical progress and considered to be in a CRPC stage (Table 2). Tissue sections were immunhistochemically stained for MSMB, CRISP3 and AR, and protein expression was evaluated by a pathologist for both intensity and percentage of positive tumor cells. Representative staining is depicted in Figure 2.

Figure 2
figure2

Immunohistochemical staining showing expression pattern of β-microseminoprotein (MSMB), cysteine-rich secretory protein-3 (CRISP3) and androgen receptor (AR) in TURP specimens derived from the same patient undergoing diagnostic TURP before androgen deprivation therapy (ADT, upper panel), and later during ADT when TURP was performed due to local tumor progression at a castration-resistant prostate cancer stage (lower panel). Images were generated at an original magnification of × 20.

A majority of the patients had very low MSMB levels already at the time of the first TURP, and decreased expression was therefore not readily detected (Figure 3). In the few patients with >50% tumors cells positive for MSMB at the time of the first TURP, there was a dramatic decrease in expression at the time for the second TURP. No patient had more than 45% tumor cells positive for MSMB at a CRPC stage. Tumor tissue from these CRPC patients during long-term ADT showed higher levels of CRISP3 and AR compared with tissue from primary tumors from the same patient at the time of diagnosis (Figure 3).

Figure 3
figure3

Bar charts representing the fraction of cancer cells positively immunostained for β-microseminoprotein (MSMB), cysteine-rich secretory protein-3 (CRISP3), and androgen receptor (AR) in TURP samples from patients undergoing TURP before receiving any treatment (gray bars), and later at tumor progression during androgen deprivation therapy (ADT, black bars). The rs10993994 single nucleotide polymorphism genotype for each patient is inserted below the upper chart, illustrating MSMB protein expression. *Tumor material missing.

Increased expression levels of CRISP3 and AR were found in terms of an increased fraction of positive cells. MSMB and CRISP3 staining intensity did not change with progressing disease, but AR staining was stronger at the time for the second TURP. The decreased MSMB protein expression on long-term ADT is corroborating the studies at the transcriptional level, and is also emphasizing that low MSMB expression is a feature of aggressive disease, whereas CRISP3 and AR show the opposite with high expression levels in CRPC.

Low MSMB protein levels are associated with the TT genotype of SNP rs10993994

The MSMB promoter contains an SNP reported to significantly affect MSMB gene expression. To investigate the impact of this SNP on MSMB expression in our experimental setup, we performed TaqMan genotyping of the rs10993994 locus on 13 of the 16 patients previously immunohistochemically examined for MSMB expression. TaqMan genotyping identified the TT genotype previously associated with lower MSMB expression in two patients. Both patients were showing low levels of MSMB. The remaining 11 patients all had CC or CT genotypes, and taken as a group, a higher level of MSMB expression (Figures 3 and 4).

Figure 4
figure4

Correlation of the rs10993994 genotype and the corresponding percentage of β-microseminoprotein (MSMB)-positive tumor cells in tissue collected from patients undergoing TURP before initiation of androgen deprivation therapy.

High transcriptional levels of the EZH2 gene are associated with low MSMB levels in progressive disease and metastases

MSMB expression before and during ADT varies considerably between patients, and cannot be explained by either androgen availability or genotype. Previous studies have shown that methylation, mediated by the Polycomb group member EZH2, is yet another way by which MSMB expression may be regulated. Therefore, we analyzed the transcriptional levels of EZH2 in the Affymetrix U95 gene expression microarray described above. Although this is a limited sample set, we found that the EZH2 gene was highly expressed in metastatic lesions from prostate cancer, where MSMB expression levels were very low. EZH2 gene expression does not appear to be affected by neoadjuvant ADT (Figure 1).

Discussion

In this study, we wanted to examine impact of short- and long-term ADT on prostate cancer outcome predictors MSMB and the MSMB-binding protein CRISP3. We show that short-term ADT significantly reduces both transcript and protein levels of MSMB, whereas CRISP3 levels do not change. Conversely, CRISP3 is upregulated in parallell with AR and KLK3 in progressive disease and during long-term ADT.

In concordance with studies conducted by others, the results of our study support that MSMB expression decrease in patients receiving neoadjuvant ADT indicating androgen dependent expression.28 Importantly, and in contrast to KLK3, MSMB expression is continuously low in progressive and metastatic disease (Figure 1). Rising level of PSA is considered a hallmark for disease progression. This indicates that despite similarities in androgen effect on KLK3 and MSMB expression in the normal prostate and primary prostate cancer, it is obvious that they are differentially regulated in progressive disease. This could be due to silencing of MSMB expression by EZH2-mediated methylation of the MSMB promotor.6 We find high levels of EZH2 in metastatic lesions, which is consistent with our findings of low MSMB expression in those tumors (Figure 1). ADT did not appear to affect EZH2 levels, thus methylation is unlikely linked to the rapid MSMB downregulation observed in hormonally treated primay prostate cancer.

Yet another way by which MSMB can be regulated is by an SNP in the promotor region. In our study, the TT allele was only observed in 2 of 13 patients (Figures 3 and 4). Both patients had very low levels of MSMB, but not lower than many patients carrying the CC or CT alleles. Despite much recent interest in the rs10993994 SNP genotype, our findings indicate that it does not fully account for the low MSMB levels generally observed in prostate cancer patients with progressive disease.

Our study data showed that neoadjuvant therapy does not appear to affect expression of CRISP3. However, the high CRISP3 expression in metastatic and recurrent tumors may be indicative of a role for CRISP3 in the progression of prostate cancer.

In this study, we find that MSMB expression may be regulated by several mechanisms, including androgen deprivation. In all, MSMB being suggested to function as a candidate tumor suppressor, it is most thought-provoking to note that according to this study, MSMB is downregulated by hormonal treatment and subsequently it may be silenced by different mechanisms. The results of our study favor the view of MSMB being a suitable marker for prostate cancer disease progression in CRPC.

References

  1. 1

    Jemal A, Siegel R, Ward E, Hao Y, Xu J, Murray T et al. Cancer statistics, 2008. CA Cancer J Clin 2008; 58: 71–96.

  2. 2

    Lilja H, Abrahamsson P-A . Three predominant proteins secreted by the human prostate gland. Prostate 1988; 12: 29–38.

  3. 3

    Abrahamsson PA, Andersson C, Bjork T, Fernlund P, Lilja H, Murne A et al. Radioimmunoassay of beta-microseminoprotein, a prostatic-secreted protein present in sera of both men and women. Clin Chem 1989; 35: 1497–1503.

  4. 4

    Whitaker HC, Warren AY, Eeles R, Kote-Jarai Z, Neal DE . The potential value of microseminoprotein-beta as a prostate cancer biomarker and therapeutic target. Prostate 2009; 70: 333–340.

  5. 5

    Tsurusaki T, Koji T, Sakai H, Kanetake H, Nakane PK, Saito Y . Cellular expression of beta-microseminoprotein (beta-MSP) mRNA and its protein in untreated prostate cancer. Prostate 1998; 35: 109–116.

  6. 6

    Varambally S, Dhanasekaran SM, Zhou M, Barrette TR, Kumar-Sinha C, Sanda MG et al. The polycomb group protein EZH2 is involved in progression of prostate cancer. Nature 2002; 419: 624–629.

  7. 7

    Beke L, Nuytten M, Van Eynde A, Beullens M, Bollen M . The gene encoding the prostatic tumor suppressor PSP94 is a target for repression by the Polycomb group protein EZH2. Oncogene 2007; 26: 4590–4595.

  8. 8

    Girvan AR, Chang P, van Huizen I, Moussa M, Xuan JW, Stitt L et al. Increased intratumoral expression of prostate secretory protein of 94 amino acids predicts for worse disease recurrence and progression after radical prostatectomy in patients with prostate cancer. Urology 2005; 65: 719–723.

  9. 9

    Bjartell AS, Al-Ahmadie H, Serio AM, Eastham JA, Eggener SE, Fine SW et al. Association of cysteine-rich secretory protein 3 and beta-microseminoprotein with outcome after radical prostatectomy. Clin Cancer Res 2007; 13: 4130–4138.

  10. 10

    Sakai H, Tsurusaki T, Kanda S, Koji T, Xuan JW, Saito Y . Prognostic significance of beta-microseminoprotein mRNA expression in prostate cancer. Prostate 1999; 38: 278–284.

  11. 11

    Thomas G, Jacobs KB, Yeager M, Kraft P, Wacholder S, Orr N et al. Multiple loci identified in a genome-wide association study of prostate cancer. Nat Genet 2008; 40: 310–315.

  12. 12

    Eeles RA, Kote-Jarai Z, Giles GG, Olama AA, Guy M, Jugurnauth SK et al. Multiple newly identified loci associated with prostate cancer susceptibility. Nat Genet 2008; 40: 316–321.

  13. 13

    Chang BL, Cramer SD, Wiklund F, Isaacs SD, Stevens VL, Sun J et al. Fine mapping association study and functional analysis implicate a SNP in MSMB at 10q11 as a causal variant for prostate cancer risk. Hum Mol Genet 2009; 18: 1368–1375.

  14. 14

    Lou H, Yeager M, Li H, Bosquet JG, Hayes RB, Orr N et al. Fine mapping and functional analysis of a common variant in MSMB on chromosome 10q11.2 associated with prostate cancer susceptibility. Proc Natl Acad Sci USA 2009; 106: 7933–7938.

  15. 15

    Kote-Jarai Z, Leongamornlert D, Tymrakiewicz M, Field H, Guy M, Al Olama AA et al. Mutation analysis of the MSMB gene in familial prostate cancer. Br J Cancer 2009; 102: 414–418.

  16. 16

    Wu D, Guo Y, Chambers AF, Izawa JI, Chin JL, Xuan JW . Serum bound forms of PSP94 (prostate secretory protein of 94 amino acids) in prostate cancer patients. J Cell Biochem 1999; 76: 71–83.

  17. 17

    Reeves JR, Xuan JW, Arfanis K, Morin C, Garde SV, Ruiz MT et al. Identification, purification and characterization of a novel human blood protein with binding affinity for prostate secretory protein of 94 amino acids. Biochem J 2005; 385 (Part 1): 105–114.

  18. 18

    Reeves JR, Dulude H, Panchal C, Daigneault L, Ramnani DM . Prognostic value of prostate secretory protein of 94 amino acids and its binding protein after radical prostatectomy. Clin Cancer Res 2006; 12 (20 Part 1): 6018–6022.

  19. 19

    Nam RK, Reeves JR, Toi A, Dulude H, Trachtenberg J, Emami M et al. A novel serum marker, total prostate secretory protein of 94 amino acids, improves prostate cancer detection and helps identify high grade cancers at diagnosis. J Urol 2006; 175: 1291–1297.

  20. 20

    Udby L, Lundwall A, Johnsen AH, Fernlund P, Valtonen-Andre C, Blom AM et al. beta-Microseminoprotein binds CRISP-3 in human seminal plasma. Biochem Biophys Res Commun 2005; 333: 555–561.

  21. 21

    Asmann YW, Kosari F, Wang K, Cheville JC, Vasmatzis G . Identification of differentially expressed genes in normal and malignant prostate by electronic profiling of expressed sequence tags. Cancer Res 2002; 62: 3308–3314.

  22. 22

    Udby L, Bjartell A, Malm J, Egesten A, Lundwall A, Cowland JB et al. Characterization and localization of cysteine-rich secretory protein 3 (CRISP-3) in the human male reproductive tract. J Androl 2005; 26: 333–342.

  23. 23

    Kosari F, Asmann YW, Cheville JC, Vasmatzis G . Cysteine-rich secretory protein-3: a potential biomarker for prostate cancer. Cancer Epidemiol Biomarkers Prev 2002; 11: 1419–1426.

  24. 24

    Bjartell A, Johansson R, Bjork T, Gadaleanu V, Lundwall A, Lilja H et al. Immunohistochemical detection of cysteine-rich secretory protein 3 in tissue and in serum from men with cancer or benign enlargement of the prostate gland. Prostate 2006; 66: 591–603.

  25. 25

    Holzbeierlein J, Lal P, LaTulippe E, Smith A, Satagopan J, Zhang L et al. Gene expression analysis of human prostate carcinoma during hormonal therapy identifies androgen-responsive genes and mechanisms of therapy resistance. Am J Pathol 2004; 164: 217–227.

  26. 26

    Chen CD, Welsbie DS, Tran C, Baek SH, Chen R, Vessella R et al. Molecular determinants of resistance to antiandrogen therapy. Nat Med 2004; 10: 33–39.

  27. 27

    Tran C, Ouk S, Clegg NJ, Chen Y, Watson PA, Arora V et al. Development of a second-generation antiandrogen for treatment of advanced prostate cancer. Science 2009; 324: 787–790.

  28. 28

    Imasato Y, Xuan JW, Sakai H, Izawa JI, Saito Y, Chin JL et al. PSP94 expression after androgen deprivation therapy: a comparative study with prostate specific antigen in benign prostate and prostate cancer. J Urol 2000; 164: 1819–1824.

Download references

Acknowledgements

We thank Jeff Holzbeierlein and the American Society for Investigative Pathology for permission to reprint the microarray data on AR expression; Elise Nilsson for excellent technical skills in the field of immunohistochemistry; Lene Udby for providing the CRISP3 antibody and Per Fernlund for the MSMB antibody.

Supporting grants: European Union 6th Framework (P-Mark) (Grant number LSHC-CT-2004-503011), Swedish Cancer Society, Swedish Research Council (Medicine), Cancer and Research Foundation at University Hospital Malmö, Gunnar Nilsson Cancer Foundation and the Crafoord Foundation.

Author information

Affiliations

Authors

Corresponding author

Correspondence to A Bjartell.

Ethics declarations

Competing interests

The authors declare no conflict of interests.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Dahlman, A., Edsjö, A., Halldén, C. et al. Effect of androgen deprivation therapy on the expression of prostate cancer biomarkers MSMB and MSMB-binding protein CRISP3. Prostate Cancer Prostatic Dis 13, 369–375 (2010). https://doi.org/10.1038/pcan.2010.25

Download citation

Keywords

  • microseminoprotein
  • PSP94
  • neoadjuvant
  • castration
  • tissue biomarker

Further reading