- Email: [email protected]
Modulation of cytokeratin subtype, EGF receptor, and androgen receptor expression during progression of prostate cancer
Modulation of Cytokeratin Subtype, EGF Receptor, and Androgen Receptor Expression During Progression of Prostate Cancer SIXTINA GIL-DIEZ DE MEDINA, MD, PHD, LAURENT SALOMON, MD, MARC COLOMBEL, MD, PHD, CLAUDE C, ABBOU, MD, JACQUELINE BELLOT, MD, JEAN PAUL THIERY, PHD, FRAN(~OIS RADVANYI, PHD, THEODORUS H, VAN DER KWAST, MD, AND DOMINIQUE K, CHOPIN, MD
•
i
After initial regression in response to androgen deprivation, most prostate cancers develop resistance to endocrine therapy. Identification of cellular and molecular changes occurring during endocrine therapy-induced regression and subsequent hormone insensitivity may point to mechanisms underlying the transition to hormoneindependent prostate cancer. A series of untreated (n = 24), regressed (n = 15), and endocrine therapy-resistant (n = 10) prostatic adenocarcinomas were analyzed using immunohistochemistry with regard to cytokeratin 5 and 18, androgen receptor (AR), and epidermal growth factor receptor (EGF-R) expression in tumor cells. Using semiqumltitative reverse transcription-polymerase chain reaction, the amount ofAR mRNA also was determined. In regressed and therapyresistant prostate cancers, an increase in cytokeratin 5-positive tumor cells was noted when compared with untreated carcinomas. Similarly, the proportion of EGF-R-positive tumor cells increased in the treated cases, whereas the proportion of AR-positive tumor cells dropped in regressed carcinomas and increased in hormone-refractory cancers.
In the latter group, an eightfold higher level of AR mRNA was observed when compared with the other cases. Changes in the proportion of cytokeralin 5 and EGF-R-positive tumor cells suggests that during androgen deprivation an enlarged subpopnlation of tumor cells with combined features of basal and secretory phenotypes arises. The increased proportion of AR-positive tumor ceils during the transition from the regression phase to the hormone escape phase points to an important role ofAR overexpression in this process. HUM PAWHOL29:1005-1012. Copyright © 1998 by W.B. Saunders Company Key words: prostate cancer, growth factors, androgen receptor, androgen deprivation, tumor progression. Abbreviations: PSA, prostate-specific antigen; AR, androgen receptor; EGF-R, epidermal growth factor receptor; TGF-~, transforming growth factor-alpha; PBS, phosphate-buffered saline; APAAP, alkaline phosphatase anti-alkaline phosphatase; TBS, TRIS-buffered saline; RT-PCR, reverse transcription polymerase chain reaction; LH-RH, luteinizing hormone-releasing hormone; TBP, TAT.&binding protein.
Development of androgen independence is a major clinical problem in the management of nonlocalized prostate cancer. Despite extensive progress in the knowledge of the molecular basis of this disease, 1,2 our understanding of the mechanisms underlying endocrine therapy resistance is quite limited. This subject is under scrutiny for the identification of molecular and cellular alterations associated with tumor progression. Normal prostatic glands are lined by a peripheral layer of basal cells and a luminal layer of secretory epithelial cells. In addition to prostate-specific antigen (PSA), the secretory cells express androgen receptor (AR),S cytokeratins 8 and 18, 4,5 and lack a receptor for
epidermal growth factor (EGF-R). 6 In contrast, basal cells do not express PSA or AR3 but do express EGF-R6 and cytokeratins 5 and 14. 7 It is well established that the development and maintenance of the prostate gland depends on androgens. 8 The difference in both AR and EGF-R expression may reflect the differential response of basal and secretory cells to androgen deprivation. In fact, after 3 months of androgen blockade, basal cells from prostatic glands may show hyperplasia, whereas secretory cells regress or undergo apoptosis. 9 Several lines of evidence suggest that polypeptide growth factors and their cognate receptors are also key modulators of prostate epithelial and stromal cell growth and homeostasis. 1° Androgen-mediated stimulation of growth factor production such as keratinocyte growth factor by stromal cells may subsequently affect the function of glandular epithelia! cells and adenocarcinoma cells through a paracrine pathway. ~°-13The epidermal growth factor (EGF) family o f ligands including EGF and transforming growth factor-~ (TGF-00 and their corresponding receptors represent the beststudied nonsteroidal growth factor system involved in normal prostatic epithelial maintenance and growth. 14,15 EGF and TGF~ are detected in normal and hyperplastic glands, 14,16and their expression is modulated by androgens. 17,1sConversely, variable patterns of EGF-R expression alongside TGF~ have been found in prostate cancer. 19,2°The interpretation of these findings is com-
From the Groupe d'Etude des Tumeurs Urologiques, Centre de Recherches Chirurgicales, Hopital Henri Mondor, Cr~teil, France; UMR 144, CNRS/Institut Curie, Paris, France; and the Department of Pathology, Erasmus University, Rotterdam, The Netherlands. Accepted for publication February 19, 1998. Supported by the Association Claude Bernard, Universit6 Paris XII, the D~ltgation/t la Recherche Clinique AP-HP, the Association pour la Recherche des Tumeurs Prostatiques (ARTP), Ipsen-Biotech. S.G.D.M. was the recipient of a fellowship from the Ligue contre le Cancer-Comit~ du Val de Marne. Address correspondence and reprint requests to Dominique K. Chopin, MD, Groupe d'Emde des Tumeurs Urologiques, Centre de Recherches Chirurgicales C.H.U. Henri Mondor, 8 rue du G4ntral Sarrai194010 Cr4teil Cedex, France. Copyright © 1998 by W.B. Saunders Company 0046-8177/98/2909-001858.00/0
1005
HUMANPATHOLOGY Volume29, No. 9 (September 1998) p l i c a t e d by t h e fact t h a t t h e r e is n o c o r r e l a t i o n b e t w e e n p r o t e i n a n d m R N A e x p r e s s i o n f o r EGF-R. R e c e n t studies i n d i c a t e t h a t EGF-R a n d TGFoL e x p r e s s i o n by prostatic a d e n o c a r c i n o m a m a y b e r e l a t e d to t u m o r p r o g r e s sion a n d t r a n s i t i o n to a n d r o g e n i n d e p e n d e n c e . 162° Most prostatic adenocarcinomas display the phenotype o f p r o s t a t i c s e c r e t o r y cells as r e f l e c t e d by t h e e x p r e s s i o n o f c y t o k e r a t i n s 8 a n d 18, PSA, a n d A R Y These observations suggest that prostatic adenocarcinoma may share a number of the functional properties o f s e c r e t o r y e p i t h e l i a l cells with r e s p e c t to a n d r o g e n r e s p o n s i v e n e s s a n d to s t r o m a - d e r i v e d g r o w t h - m o d u l a t i n g factors. At t h e initial stage a f t e r a n d r o g e n d e p r i v a tion, b e f o r e p r o s t a t i c c a r c i n o m a cells b e c o m e refractory to h o r m o n e d e p r i v a t i o n , a p r o p o r t i o n o f t u m o r cells u n d e r g o a p o p t o s i s , w h e r e a s a n o t h e r s u b p o p u l a t i o n o f t u m o r cells persist. 8,21 C h a n g e s in e x p r e s s i o n levels o f a n u m b e r o f d i f f e r e n t i a t i o n - a s s o c i a t e d m o l ecules during the transition from an androgen-sensitive to a n a n d r o g e n - i n s e n s i t i v e c a r c i n o m a m a y p o i n t to mechanisms responsible for the development of androg e n i n d e p e n d e n c e . T h e p u r p o s e o f this study is to analyze p h e n o t y p i c a l t e r a t i o n s o f p r o s t a t i c a d e n o c a r c i n o m a cells b o t h d u r i n g s h o r t - t e r m a n d r o g e n a b l a t i o n a n d at t h e stage o f h o r m o n a l failure, with t h e a i m o f d e f i n i n g m o l e c u l a r targets f o r h o r m o n a l e s c a p e .
MATERIALS AND METHODS
Clinical Data and Samples Tissue samples were obtained from radical prostatectomy specimens, transrectal needle biopsies, or transurethral resections. A representative sample was taken from each tissue for histopathological and immunohistochemical assessment, and an adjacent piece was placed in liquid nitrogen for RNA extraction when available. Clinical and pathological characteristics are presented in Table 1. Three groups of patients were distinguished: group 1 patients without any hormonal manipulation (n = 24/: group 2 patients with hormonal deprivation (n = 15): group 3 panents with patent hormonal failure (n = 10); using objective criteria as defined by the National Prostatic Cancer P r o j e c t P In group 2, hormonal blockade was obtained using classical androgen deprivation with a luteinizing hormone-releasing h o r m o n e (LH-RH) analog in 13 patients or by orchidectomy in two patients (Table 2). In group 3, six patients had hormonal deprivation using LH-RH analogs, and four had an orchidectomy. Tissue samples were
TABLE 1. Clinical and Pathological Features of Tumor Specimens Used
Stage and Grade
Group 1 Untreated CaP (N= 24)
Group 2 CaP Hormonal Deprivation (N= 15)
T2 T3 T4 TX Gleason <6 Gleason >6 Gleason X NqM+
14 5 5 0 8 16 0 2 1
5 6 3 1 2 11 2 6 3
Group 3 CaP Hormonal Failure (N = 10) 0 2 2 6 0 9 1 4 8
TABLE 2. HormonalTreatment in Patients of Groups 2 and 3 Treatment
Group 2 CaP Hormonal Deprivation
Group 3 CaP Hormonal Failure
LH-RH Orchidectomy
13 2
6 4
obtained between 3 and 36 months (average, 6 months), and four tissue samples after 60 months of treatment. For all patients in group 2. there was no clinical evidence of tumor progression. This included a persistently low PSA level in patients with at least two PSA determinations at 4-week intervals.
Immunostaining of Cytokeratin-5 (RCK-103) and Cytokeratin-18 (RGE-53) Fresh frozen cryosections (4 pm) were air dried and fixed in chilled acetone for 10 minutes. Slides were incubated in blocking solution (nonfat dry milk, 5 % i n phosphate-buffered saline [PBS]) for 30 minutes. Subsequently, slides were incubated with mouse monoclonal antibody RCK-103 (anticytokeratin 5) or RGE-53 (anti-cytokeratin 18) hybfidoma culture supernatant for 60 minutes at room temperature. 7 Slides were washed in PBS and incubated with horseradish peroxidase-conjugated rabbit anti-mouse immunoglobulins (Dako, Copenhagen-Denmark) for 45 minutes at room temperature. Next, sections were washed in PBS and incubated in 3,3' diaminobenzidine tetrahydrochloride (Sigma, France) for 5 minutes or APAAP complex with appropriate bridge antibody system (Dako, Copenhagen-Denmark). Counterstaining was performed with Harris hematoxylin. Sections were dehydrated and mounted with i m m u n o m o u n t (Shandon, UK) or resin Eukitt (Kindler GmbH, Freiburg, Germany).
immunostaining of EGF receptor Fresh frozen cryosections (4 lain) were air-dried and fixed in chilled acetone for 10 minutes and incubated in blocking solution (nonfat dry milk, 5% in TRIS-buffered saline [TBS], pH 7.2) for 30 minutes. Then the slides were incubated with a mouse anti-EGF receptor monoclonal antibody (clone 528) (10 p g / m L ) , which recognizes an extracellular epitope (Oncogene Science. Paris, France), for 60 minutes at room temperature, washed in TBS, and incubated with a secondary bridge rabbit anti-mouse antibody (Dako, Copenhagen, Denmark) for 30 minutes. Slides were washed in TBS and incubated with a mouse APAAP complex (Dako, Copenhagen, Denmark) rinsed in TBS, and consequently the reaction was developed with the fast red chromogen substrate (Sigma, France) for 30 minutes. Slides were counterstained with Mayer's hematoxilin and mounted with immunomount.
Immunostainin~ With an Antihuman Androgen Receptor (F-39-4-1) Frozen sections (4 pm) were air dried and fixed for 5 minutes with paraformaldehyde (4% in PBS), incubated in chilled methanol (-20°C, 4 minutes), dehydrated in cold acetone (-20°C) for 2 minutes, and incubated with a mouse monoclonal anti-human androgen receptor antibody (1/ 2.000) (ascites) overnight at 4°C. 3 The indirect APAAP procedure was performed as described.
Immunostaining Scoring For each slide, the percentage of stained tumor cells and the intensity (0 -= negative, 1 = low, 2 = moderate, 3 - in-
1006
PHENOTYPE OF PROSTATE CANCER PROGRESSION (GiI-Diez de Medina et al)
FIGURE 1. Expression of cytokeratin 18 and 5 in human prostate carcinoma. Frozen sections (4 IJm) of prostatic adenocarcinoma (A, D) without hormonal treatment, (B, E) with hormonal deprivation, and (C, F) with hormonal failure. (A, B, C) Immunostaining with RGE-53 (CK-18) shows positive staining of tumor cells, in all cases. (D, E, F) Immunostaining with RCK-103 (CK-5) (D) shows negative immunostaining of tumor cells from patients without hormonal treatment. (E, F) Expression of CK-5 in the tumor cells after hormone deprivation and hormonal failure. (A, B, D, E, original magnification x200; C and F, original magnification x500.) tense) of staining was recorded using a light microscope equipped with an eyepiece reticule at 400× magnification. The score was determined in each case after counting at least 100 to 500 tumor cells in three to five selected fields.
RNA Extraction The RNA was extracted from frozen tissue samples for a selected group of specimens (six peripheral zone [PZ] from normal prostates, 11 adenocarcinomas of group 1; patients without any hormonal manipulation, eight adenocarcinomas of group 2; patients with hormonal deprivation and 10 adenocarcinomas of group 3; patients with patent hormonal
failure). In all selected adenocarcinomas, histological adjacent sections from the samples used for RNA extraction were composed of more than 60% tumor cells. RNA was prepared according to Chirgwin et al, 23 using ultracentrifugation on a 5.7 m o l / L cesium chloride cushion after tissue homogenization in 4 m o l / L guanidium thiocyanate. RNA was subsequently extracted twice with pheno! chloroform.
RT-PCR The level ofAR mRNA24was determined by semiquantitafive reverse transcription-polymerase chain reaction (RTPCR), by comparison with an internal control as previously
1007
HUMAN PATHOLOGY
Volume29, No, 9 (September 1998)
described. 25,~6GAPDH or the TATA binding protein (TBP), a ubiquitous transcription factor, was used as an internal control. 27 cDNA was prepared using 1 lag RNA from each sample. The semiquantitative polymerase chain reaction was performed according to Gil Diez de Medina et al27 using 2 laL (1/50) of the reverse transcription product in a final volume of 50 laL containing all four dNTPs (each at 100 lamol/L), with 1 lamol/L of each primer (a pair for GADPH, together with a pair specific for AR or TBP) and 2 pCi of ~[~2P]dCTE The number of cycles was chosen to be in the exponential phase of the two PCR reactions. We used 21 cycles for the co-amplification of AR and GAPDH and 23 cycles for the co-amplification of TBP and GAPDH. The primers' sequences were: AGTGAAGAACAGTCCAGACTG and CCAGGAAATAACTCTGGCTCAT (TBP); CTGCACCACCAACTGCT'FAG and AGGTCCACCACTGACACGTT (GAPDH); GACTTCACCG~ ACCTGATG and CTGGCAGTCTCCAAACGCAT (AR). All primers are given in the sense 5' to 3' direction. The primers for the AR cDNA were chosen in the first and second exons34 The amplification reactions were executed in a PTC-200 programmable thermocycler (MJ Research, Watertown, MA) with an initial cycle of 95°C for 5 minutes before the addition of Hi-Taq DNA thermostable polymerase (Bioprobe, France). Thereafter, each cycle was as follows: 94°C fbr 1 minute, 57°C for 1 minute, and 72°C for 80 seconds. These cycles were followed by a final incubation step at 72°C for 10 minutes. The PCR-amplified products were loaded in duplicate and electrophoresed in 8% polyacrylamide gels, fixed in 7% acetic acid, and vacuum-dried. Autoradiograms showed two bands corresponding to the co-amplified fragments. The signals were quantified with a Molecular Dynamics 300 Phosphorlmager (Molecular Dynamics, Sunnyvale, CA). Each measurement was repeated for three independent PCR reactions and found to be identical within a 15% margins. Negligible amplification was observed when reverse transcriptase was omitted from the reverse transcription reaction.
Statistics Analysis of variance was performed, using the KruskalWallis nonparametric ANOVA test, to compare the immune phenotype of tumor cells from the three groups of patients and for the RT-PCR results.
RESULTS Expression of Cytokeratins 5 a n d 18 in Prostatic C a r c i n o m a Cells Most carcinoma cells (Fig 1A, B), in b o t h u n t r e a t e d (group 1) and androgen-deprived specimens (group 2), expressed cytokerafin 18 as defined by their reactivity with RGE-53. In the tumor-treated group, we f o u n d a significant decrease in p e r c e n t a g e of RGE-53 staining c o m p a r e d with g r o u p 1. A similar descending trend was observed for the intensity of staining during h o r m o n a l failure (group 3) (Table 3). However, in this group, the staining was m o r e h e t e r o g e n e o u s (Fig 1C). T h e basal cell marker, RCK-103, directed against cytokeratin 5 was positive for only a small p e r c e n t a g e of t u m o r cells (14.3% + 3.4) f r o m u n t r e a t e d g r o u p 1 tumors (Fig 1D). A m o r e distinct staining reaction was observed in g r o u p 2 (Fig 1E) a n d g r o u p 3 (Fig 1F)
TABLE $. Percentage of Positive Tumor Ceils and Intensity for RGE 53 (Cytokeratin 18) and for RCK 103 (Cytokeratin 5) in Relation to the Hormonal Status of the Tumor CK18 (RGE 53)
CK5 (RCK103)
Percentage Percentage (%) Positive (%) Positive N Ceils -+ SEM Intensity Cells + SEM Intensity Untreated CaP 24 85.3 + 2.1 CaP hormonal deprivation 15 66.7 ___6.0 CaP hormonal failure 10 63.3 +_10.3
3
14.3 _ 3.4
1 to 2
2 to 3
47.0 + 6.2
1 to 3
0 to 3
38.3 -+ 10.5
2 to 3
NOTE. Nonparametric ANOVA test: RGE-53 P = .002 and RCK-103P = .0004.
tumors (respectively, 47.0% + 6.2 a n d 38.3% - 10.5) with an increased staining intensity (Table 3). In groups 2 a n d 3, we n o t e d in adjacent sections the coexpression of CK-8-18 a n d CK-5 in t u m o r cells, suggesting an intermediate basal a n d secretory cells phenotype.
EGF Receptor Expression In the 24 u n t r e a t e d specimens (group 1), basal cells of n o r m a l and hyperplastic glands r e p r e s e n t e d an internal positive control (arrow) (Fig 2A). In these cases only a few prostatic c a r c i n o m a cells expressed EGF receptor (14.7% ___4.5) with low intensity (score 1). In contrast, a significant increase in the n u m b e r of positive cells with somewhat e n h a n c e d labeling intensity was observed in g r o u p 2 a n d 3 samples (Fig 2B, C) with 34.5% ± 6.8 (score 1 to 2) a n d 46.3% - 10.8 (score 1 to 2), respectively (Table 4).
A n d r o g e n Receptor Expression In u n t r e a t e d prostatic adenocarcinomas, m o s t tum o r cells (64.9% +_ 3.9) displayed a u n i f o r m nuclear staining intensity equivalent to the staining observed in n o r m a l secretory cells (Fig 2D). In treated tumors (group 2), the fraction of i m m u n o s t a i n e d t u m o u r cells (Fig 2E) was markedly r e d u c e d (32.3% - 8.4) with a h e t e r o g e n e o u s pattern of nuclear expression (staining intensity f r o m 1 to 3). At the stage o f h o r m o n a l failure (group 3), a significant increased p e r c e n t a g e (66,3 +__ 4.1%, 2 to 3) of stained t u m o r cells a n d intensity was detected (Table 5). Remarkably, cytoplasmic staining was occasionally seen in cases o f h o r m o n a l refractory tumors in addition to typical nuclear staining (Fig 2F).
RT-PCR T h e AR m R N A level was d e t e r m i n e d by semiquantitative RT-PCR. 27 T h e means of a n d r o g e n receptor m R N A expression in the g r o u p 1 a n d 2 prostatic carcinomas were c o m p a r a b l e to n o r m a l peripheral prostates (average = 0.075 ___0.01). However, we observed high expression in the h o r m o n a l failure g r o u p
1008
PHENOTYPE OF PROSTATE CANCER PROGRESSION (GiI-Diez de Medina et al)
FIGURE 2. Expression of the epidermal growth factor and the androgen receptor in prostate carcinoma ceils. Frozen sections (4 IJm) of prostatic adenocarcinoma (A, D) without hormone manipulation, (B, E) with hormone deprivation, and (C, F) with hormonal failure. (A, B, C) Immunostaining with an anti-EGF-R antibody (A)-negative immunostaining of tumor cells, and preservation of staining in basal cells from adjacent benign glands (arrow), (B, C) positive staining of the tumor cells. (D, E, F) Immunostaining with an anti-AR antibody (39-4-1) (D) shows positive immunostaining in the nucleus of the tumor cells, (E) exhibits decreased immunostaining after hormone deprivation, and (F) expression of the AR in the nucleus of tumor cells in this subset of hormone-refractory prostatic cancer cytoplasmic staining was observed in many different tumor cells (A, D, E, F, original magnification ×200; B and C, original magnification ×500,)
(group 3). The increase in mRNA expression (average = 0.62 + 0.2) was eightfold when compared with other groups and normal peripheral prostatic tissue (P = .0003). A similar eighffold increase was obtained when another internal standard (TBP) was used instead of GADPH (data not shown).
In benign tissue cell-type separated, 2s using an enzymatic digestion combined with percoll gradient centrifugation (data not shown), there was an equivalent number of AR transcripts expressed by the epithelial cell fractions when compared with the stromal cell fractions.
1009
HUMAN PATHOLOGY
Volume 29, No. 9 (September 1998) TABLE 5.
1.8
1.6
E
1.4
•I" a.
1.2
.< 0 ~.
1.0
Androgen Receptor Immunostaining
Untreated CaP CaP hormonal deprivation CaP hormonal failure
N
Percentage (%) PositiveCells
Intensity
24 15 10
64.9 +- 3.9 32.3 -+ 8.4 66.3 -+ 4.1
2 to 3 1 to 3 2 to 3
NOTE. Nonparametric ANOVAtest P = .007.
0.8 9,
< Z
0.6
E
0.4
n,, <[~
0.2 0
+•
+ Normal Prostate n = 6
I
+
Group 1
Group 2
Group 3
n =11
n = 8
n =10
Semiquantitative mRNA expression of androgen receptor. The expression level of androgen receptor mRNA was determined by means of semiquantitative RT-PCRusing GADPH as internal control: in normal peripheral prostatic tissue (n = 6); group 1 (n= 11); group 2 (n= 8); and group 3 (n = 10) of prostatic adenocarcinomas. The bar in each group represents the means of value. The eightfold increase in AR mRNA levels in hormone-refractory population compared with the other groups is statistically significant (P : .0003) using nonparametric ANOVA. FIGURE 3,
DISCUSSION
According to the prostatic stem cell model, the prostatic glandular basal cell population comprises a stern cell population committed to differentiate into amplifying cells that may rapidly proliferate and terminally differentiate into secretory cells29 In this model, it was also hypothesized that the differentiated androgend e p e n d e n t secretory cells are the progeny o f a subpopulation of self-replicating cells that do not require androgens for their survival. It has been suggested that this amplifying cell population shares differentiation features with basal and secretory cells. In particular, the demonstration in rat prostatic tissue of cells expressing basal cell type and secretory cell type cytokeratins is in agreement with this concept29,3° Thus, cells with a combined expression of basal and secretory type cytokeratins could be considered as candidates of amplifying ceils. In this study, we confirmed that prostatic carcin o m a cells co-express cytokeratins characteristic of both basal and secretory cells. 7 This observation agrees with the plasticity of basal cells that give rise to different TABLE 4.
FGF Receptor Immunostaining
Untreated CaP CaP hormonal deprivation CaP hormonal failure
N
Percentage (%) Positive Ceils - SEM
Intensity
24 15 10
14.7 + 4.5 34.5 --- 6.8 46.3 --- 10.8
1 1 to 2 1 to 2
NOTE. Nonparametric ANOVAtest P = .003. 1010
epithelial cell lineages in neoplastic h u m a n prostate. ~1 In addition we showed that the proportion of the cytokeratin 5-positive subpopulation was increased after androgen deprivation during both the regression stage and after h o r m o n a l escape. This increased percentage of cytokeratin 5-positive cells can be considered as the outcome of a selection process based on the androgen i n d e p e n d e n c e of a less mature putative amplifying cell subpopulation. Conversely, cellular adaptation to an altered endocrine environment may account for the observed increase in population o f cytokeratin 5-positive cells. EGF-R is present in basal cells, but not in secretory epithelial cells of the prostatic glands. Previous studies in the rat have shown an increase in EGF-R expression in normal or hyperplastic prostate tissue after androgen ablation. ~2 (Immunohistochemical studies have suggested that this increase was not due to overexpression of the EGF-R, by secretory cells. In fact, e n r i c h m e n t or positive selection for the (EGF-R positive) basal cell population underlies the observed increase in EGF-R in castrated animals). Several studies have shown the presence of EGF-R in untreated primary prostate cancers, with conflicting results with regard to expression levels and tumor grade. 16,32Differences in results can be attributed to variation in immunohistochemical assays, antibodies used (eg, directed against extracellular v intracellular domain of EGF-R) and tissue fixation. We p e r f o r m e d our study on frozen sections, gently fixed in acetone, and f o u n d a comparatively low percentage (average, 14.7%) of EGF-R-positive t u m o r cells in primary untreated prostatic adenocarcinomas. It should be n o t e d that other studies r e p o r t e d the percentage of EGF-R-positive tumors or intensity scores of immunostaining but not the percentage of immunolabeled t u m o r ceils. 16,32 Notably, we f o u n d a significant increased percentage of EGF-R-positive t u m o r cells in androgen-sensitive carcinomas shortly after androgen deprivation (group 2) with a further rise in endocrine therapy-resistant tumours. Similarly, Scher et a.119 recently r e p o r t e d h o m o g e n o u s EGF-R expression in metastases of endocrine therapy-resistant prostatic cancers. Furthermore, both in primary carcinomas and in metastatic a n d r o g e n - i n d e p e n d e n t tumours, the presence of TGFot suggests the existence o f an autocrine loop in these t u m o r ceils. 16,19 Again, these data can be interpreted as a selection process for EGF-R-positive tumor cells, or as an adaptation to a low androgen environment. It has been r e p o r t e d that distant metastases of prostate cancer express AR protein. 3a In addition, endocrine therapy-resistant prostatic carcinomas dis-
PHENOTYPE OF PROSTATECANCER PROGRESSION (GiI-Diez de Medina et al)
play high AR expression levelsP 4 This level is increased when compared with untreated primary prostatic carcinomas. Both genetic 35 and epigenetic mechanisms may be involved in the control of AR expression in androgenindependent tumorsP 6 In this study, the observed decreased in percentage of AR-positive tumor cells early after androgen deprivation when compared with the proportion of AR-positive tumor cells found in untreated carcinomas is in agreement with recent preliminary observations on another group of patients, s7 We hypothesize that AR-positive cell populations are maintained while the AR-negative tumor cells are eliminated after long-term androgen deprivation. These results suggest that androgen-depleted environments may select for AR-positive prostatic tumor cells. 3s The increased proportion of AR-positive tumor cells in the hormonotherapy-resistant prostate cancers was associated with an increase in AR mRNA as compared with untreated cancers. Using mRNA in situ hybridization, Koivisto et al so recently reported a similar observation. Cytoplasmic staining of the androgen receptoi, as observed in this study, may be related to the very high concentration of the receptor in the cell or to interactions of the AR receptor with co-regulatory proteins. This finding merits further investigations. Amplification of the AR gene and multiple X chromosomes was shown in a large proportion of endocrine therapy-resistant prostate cancers. 35,39 It is not clear whether a gene dosage effect alone accounts for the observed increase in mRNA coding AR. In conclusion, we have shown that progression of untreated prostatic carcinoma to local endocrine therapy-resistant carcinoma is associated with differentiationassociated features of the tumor cell population. Doublelabeling experiments are required to define more accurately the changes in the various cell populations during the transition to androgen independence. The inappropriate expression of EGF-R and cytokeratins observed during hormone failure is already apparent during the early androgen-sensitive stage of prostate cancer. AR overexpression arises in hormonal refractory prostate cancer after a downregulation in hormoneresponsive tumors. This observation suggests that increased expression of AR may play a role in this process. Further experimental studies should be designed to evaluate whether a relationship exists between AR and EGF-R expression during the progression of prostate cancer.
Acknowledgment. T h e authors t h a n k Dr Frans C.S. Ramaekers (University of Maastricht, T h e Netherlands) for fruitful advice and for generous d o n a t i o n of m o n o c l o n a l antibodies RCK-103 and RGE-53. They also thank Drs Aria Mafia Vallts and Ivan Coulter for helpful discussions and for critical reading of the manuscript. T h e authors thank Domin i q u e M o r i n e a u for e x p e r t p h o t o g r a p h i c assistance, A n n i e Maurette, and Catherine Petit for secretarial assistance. REFERENCES 1. Bostwick DG, Pacelli A, Lopez-Beltran A: Molecular biology of prostatic intraepithelial neoplasia. Prostate 29: l 17-134, 1996
2. Isaacs WB, Bova GS, Morton RA, et al: Molecular biology of prostate cancer. Semin Oncol 21:514-521, 1994 3. Ruizeveld de WinterJA, Trapman J, Vermey M, et al: Androgen receptor expression in human tissues: An immunohistochemical study. J Histochem Cytochem 39:927-936, 1991 4. Sherwood ER, Berg LA, Mitchell NJ, et al: Differential cytokeratin expression in normal, hyperplastic and malignant epithelial cells from human prostate. J Uro1143:167-171, 1990 5. Sherwood ER, Theyer G, Steiner S, et al: Differential expression of specific cytokeratin polypeptides in the basal and luminal epithelia of the human prostate. Prostate 18:303-314, 1991 6. Maddy SO~ Chisholm GD, Hawkins RA, et al: Localization of epidermal growth factor receptors in the human prostate by biochemical and immunocytochemical methods. J Endocrinol 113:147-153, 1987 7. Verhagen APM, Ramaekers FCS, Aalders TW, et 9_1:Colocalization of basal and luminal cell-type cytokeratins in human prostate cancer. Cancer Res 52:6182-6187, 1992 8. Bruchovsky N, Lesser B, Van Doorn E, et al: Hormonal effects on cell proliferation in rat prostate. Vitamins Horm 33:61-102, 1975 9. Vaillancourt L, Tttu B, Fradet Y, et al: Effect of neoadjuvant endocrine therapy (combined androgen blockade) on normal prostate and prostatic carcinoma: A randomized study. AmJ Surg Pathol 20:86-93, 1996 10. Cunha GR: Growth factors as mediators of androgen action during male urogenital development. Prostate 6:22-25, 1996 11. Thomson AA, Foster BA, Cunha GR: Analysis of growth factors and receptor mRNA levels during development of the rat seminal vesicle and prostate. Development 124:2431-2439, 1997 12. Fukabori Y, Yah G, Yamanaka H, et al: Rapid induction of keratinocyte growth factor (FGF-7) and beta-actin after exposure of prostate stromal cells to androgen. In Vitro Cell Dev Bio130A:745-746, 1994 13. Culig Z, Hobisch A, Cronauer MV, et al: Androgen receptor activation in prostatic turnout cell lines by insulin-like growth factor-I, keratinocyte growth factor and epidermal growth factor. Cancer Res 54:5474-5478, 1994 14. Ibrahim GK, Kerns BJM, MacDonald JA, et al: Differential immunoreactivity of epidermal growth factor receptor in benign, dysplastic and malignant prostatic tissues. J Uro1149:170-173, 1993 15. Taylor TB, Ramsdell JS: Transforming growth factor-~ and its receptor are expressed in the epithelium of the rat prostate gland. Endocrinology 133:1306-1311, 1993 16. Glynne-Jones E, Goddard L, Harper ME: Comparative analysis of mRNA and protein expression for epidermal growth factor receptor and ligands relative to the proliferative index in human prostate tissue. HUM PATHOL27:688-694, 1996 17. Fiorelli G, De Bellis A, Longo A, et al: Epidermal growth factor receptors in human hyperplastic prostate tissue and their modulation by chronic treatment with a gonadotropin-releasing hormone analog.J Clin Endocrinol Metab 68:740-743, 1989 18. Hofer DR, Sherwood ER, Bromberg WD, et al: Autonomous growth of androgen-independent hmnan prostatic carcinoma cells: Role of transforming growth factor a. Cancer Res 51:2780-2785, 1991 19. Scher HI, Sarkis A, Reuter V, et al: Changing pattern of expression of the epidermal growth factor receptor and transforming growth factor alpha in the progression of prostatic neoplasms. Clin Cancer Res 1:545-550, 1995 20. Yang Y, Chisholm GD, Habib FK: Epidermal growth factor and transforming growth factor alpha concentrations in BPH and cancer of the prostate: Their relationships with tissue androgen levels. BrJ Cancer 67:152-155, 1993 21. Bruchovsky N, Rennie PS, Coldman AJ, et al: Effects of androgen withdrawal on the stern cell composition of the Shionogi carcinoma. Cancer Res 50:2275-2282, 1990 22. Schmidt JD, Scott WW, Gibbons R, et al: Chemotherapy programs of the National Prostatic Cancer Project (NPCP). Cancer 45:1937-1946, 1980 23. Chirgwin JM, Przybyla AE, MacDonald RJ, et al: Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18:5294-5299, 1979 24. TilleyWD, Marcelli M, WilsonJD, et al: Characterization and expression of a cDNA encoding the human androgen receptor. Proc Natl Acad Sci U S A 86:32%331, 1989
1011
HUMAN PATHOLOGY
Volume 29, No. 9 (September 1998)
25. Kandel J, Bossy-Wetzel E, Radvanyi F, et al: Neovascularization is associated with a switch to the export of bFGF in the multistep development of fibrosarcoma. Cell 66:1095-1104, 1991 26. Radvanyi F, Christgau S, Baekkeskov S, et al: Pancreatic beta cells cultured from individual preneoplastic foci in a multistage tumourigenesis pathway: A potentially general technique for isolating physiologically representative cell lines. Mol Cell Biol 13:4223-4232, 1993 27. Gil Diez de Medina S, Chopin D, E1 Marjou A, et al: Decreased expression of keratinocyte growth factor receptor in a subset of human transitional cell bladder carcinomas. Oncogene 14:323-330, 1997 28. Kozlowski MJ, McEwan R, Keer H, et al: Prostate cancer and the invasive phenotype: Application of new in vivo and in vitro approaches, in Fidler IJ, Nicholson G (eds): Tumour Progression and Metastasis. NewYork, NY, Alan R. Liss, 1988, pp 189-231 29. Verhagen APM, Aalders TW, Ramaekers FCS, et al: Differential expression of keratins in the basal and luminal compartments of rat prostatic epithelium during degeneration and regeneration. Prostate 13:25-38, 1988 30. HsiehJT, Zhan HE, Wang XH, et al: Regulation of basal and luminal cell-specific cytokeratin expression in rat accessory sex organs: Evidence for a new class of androgen-repressed genes and insight into their pairwise control.J Biol Chem 267:2303-2310, 1992 31. BonkhoffH, Stein U, Remberger K: Multidirectionnal differentiation in the normal, hyperplastic and neoplastic human prostate: Simultaneous demonstration of cell-specific epithelial markers. HUM PATHOL25:42-46, 1994
32. Traish AM, Wotiz HH: Prostatic epidermal growth factor receptors and their regulation by androgens. Endocrinology 121:14611467, 1987 33. Hobisch A, Culig Z, Radmayr C, et al: Distant metastases from prostatic carcinoma express androgen receptor protein. Cancer Res 55:3068-3072, 1995 34. Van der Kwast TH, SchalkenJA, Ruizeveld de WinterJA, et al: Androgen receptors in endocrine therapy resistant human prostate cancer. IntJ Cancer 48:189-193, 1991 35. Koivisto P, Hyytinen E, Palmberg C, et al: Analysis of genetic changes underlying local recurrence of prostate carcinoma during androgen deprivation therapy. AmJ Patho1147:1608-1614, 1995 36. Kokontis J, Takakura K, Hay N, et al: Increased androgen receptor activity and altered c-myc expression in prostate cancer cell after long-term androgen deprivation. Cancer Res 54:1566-1573, 1994 37. Van der Kwast TH, T~tu B, Fradet Y, et al: Androgen receptor modulation in benign human prostatic tissue and prostatic adenocarcinoma during neoadjuvant endocrine combination therapy. Prostate 28:227-231, 1996 38. Culig Z, Hobisch A, Cronauer MV, et al: Mutant androgen receptor detected in an advanced stage prostatic carcinoma is activated by adrenal androgens and progesterone. Mol Endocrinol 7:1541-1550, 1993 39. Koivisto P, KononenJ, Palmberg C, et al: Androgen receptor gene amplification: A possible molecular mechanism for androgen deprivation therapy failure in prostate cancer. Cancer Res 57:314-319, 1997
1012
•
i
After initial regression in response to androgen deprivation, most prostate cancers develop resistance to endocrine therapy. Identification of cellular and molecular changes occurring during endocrine therapy-induced regression and subsequent hormone insensitivity may point to mechanisms underlying the transition to hormoneindependent prostate cancer. A series of untreated (n = 24), regressed (n = 15), and endocrine therapy-resistant (n = 10) prostatic adenocarcinomas were analyzed using immunohistochemistry with regard to cytokeratin 5 and 18, androgen receptor (AR), and epidermal growth factor receptor (EGF-R) expression in tumor cells. Using semiqumltitative reverse transcription-polymerase chain reaction, the amount ofAR mRNA also was determined. In regressed and therapyresistant prostate cancers, an increase in cytokeratin 5-positive tumor cells was noted when compared with untreated carcinomas. Similarly, the proportion of EGF-R-positive tumor cells increased in the treated cases, whereas the proportion of AR-positive tumor cells dropped in regressed carcinomas and increased in hormone-refractory cancers.
In the latter group, an eightfold higher level of AR mRNA was observed when compared with the other cases. Changes in the proportion of cytokeralin 5 and EGF-R-positive tumor cells suggests that during androgen deprivation an enlarged subpopnlation of tumor cells with combined features of basal and secretory phenotypes arises. The increased proportion of AR-positive tumor ceils during the transition from the regression phase to the hormone escape phase points to an important role ofAR overexpression in this process. HUM PAWHOL29:1005-1012. Copyright © 1998 by W.B. Saunders Company Key words: prostate cancer, growth factors, androgen receptor, androgen deprivation, tumor progression. Abbreviations: PSA, prostate-specific antigen; AR, androgen receptor; EGF-R, epidermal growth factor receptor; TGF-~, transforming growth factor-alpha; PBS, phosphate-buffered saline; APAAP, alkaline phosphatase anti-alkaline phosphatase; TBS, TRIS-buffered saline; RT-PCR, reverse transcription polymerase chain reaction; LH-RH, luteinizing hormone-releasing hormone; TBP, TAT.&binding protein.
Development of androgen independence is a major clinical problem in the management of nonlocalized prostate cancer. Despite extensive progress in the knowledge of the molecular basis of this disease, 1,2 our understanding of the mechanisms underlying endocrine therapy resistance is quite limited. This subject is under scrutiny for the identification of molecular and cellular alterations associated with tumor progression. Normal prostatic glands are lined by a peripheral layer of basal cells and a luminal layer of secretory epithelial cells. In addition to prostate-specific antigen (PSA), the secretory cells express androgen receptor (AR),S cytokeratins 8 and 18, 4,5 and lack a receptor for
epidermal growth factor (EGF-R). 6 In contrast, basal cells do not express PSA or AR3 but do express EGF-R6 and cytokeratins 5 and 14. 7 It is well established that the development and maintenance of the prostate gland depends on androgens. 8 The difference in both AR and EGF-R expression may reflect the differential response of basal and secretory cells to androgen deprivation. In fact, after 3 months of androgen blockade, basal cells from prostatic glands may show hyperplasia, whereas secretory cells regress or undergo apoptosis. 9 Several lines of evidence suggest that polypeptide growth factors and their cognate receptors are also key modulators of prostate epithelial and stromal cell growth and homeostasis. 1° Androgen-mediated stimulation of growth factor production such as keratinocyte growth factor by stromal cells may subsequently affect the function of glandular epithelia! cells and adenocarcinoma cells through a paracrine pathway. ~°-13The epidermal growth factor (EGF) family o f ligands including EGF and transforming growth factor-~ (TGF-00 and their corresponding receptors represent the beststudied nonsteroidal growth factor system involved in normal prostatic epithelial maintenance and growth. 14,15 EGF and TGF~ are detected in normal and hyperplastic glands, 14,16and their expression is modulated by androgens. 17,1sConversely, variable patterns of EGF-R expression alongside TGF~ have been found in prostate cancer. 19,2°The interpretation of these findings is com-
From the Groupe d'Etude des Tumeurs Urologiques, Centre de Recherches Chirurgicales, Hopital Henri Mondor, Cr~teil, France; UMR 144, CNRS/Institut Curie, Paris, France; and the Department of Pathology, Erasmus University, Rotterdam, The Netherlands. Accepted for publication February 19, 1998. Supported by the Association Claude Bernard, Universit6 Paris XII, the D~ltgation/t la Recherche Clinique AP-HP, the Association pour la Recherche des Tumeurs Prostatiques (ARTP), Ipsen-Biotech. S.G.D.M. was the recipient of a fellowship from the Ligue contre le Cancer-Comit~ du Val de Marne. Address correspondence and reprint requests to Dominique K. Chopin, MD, Groupe d'Emde des Tumeurs Urologiques, Centre de Recherches Chirurgicales C.H.U. Henri Mondor, 8 rue du G4ntral Sarrai194010 Cr4teil Cedex, France. Copyright © 1998 by W.B. Saunders Company 0046-8177/98/2909-001858.00/0
1005
HUMANPATHOLOGY Volume29, No. 9 (September 1998) p l i c a t e d by t h e fact t h a t t h e r e is n o c o r r e l a t i o n b e t w e e n p r o t e i n a n d m R N A e x p r e s s i o n f o r EGF-R. R e c e n t studies i n d i c a t e t h a t EGF-R a n d TGFoL e x p r e s s i o n by prostatic a d e n o c a r c i n o m a m a y b e r e l a t e d to t u m o r p r o g r e s sion a n d t r a n s i t i o n to a n d r o g e n i n d e p e n d e n c e . 162° Most prostatic adenocarcinomas display the phenotype o f p r o s t a t i c s e c r e t o r y cells as r e f l e c t e d by t h e e x p r e s s i o n o f c y t o k e r a t i n s 8 a n d 18, PSA, a n d A R Y These observations suggest that prostatic adenocarcinoma may share a number of the functional properties o f s e c r e t o r y e p i t h e l i a l cells with r e s p e c t to a n d r o g e n r e s p o n s i v e n e s s a n d to s t r o m a - d e r i v e d g r o w t h - m o d u l a t i n g factors. At t h e initial stage a f t e r a n d r o g e n d e p r i v a tion, b e f o r e p r o s t a t i c c a r c i n o m a cells b e c o m e refractory to h o r m o n e d e p r i v a t i o n , a p r o p o r t i o n o f t u m o r cells u n d e r g o a p o p t o s i s , w h e r e a s a n o t h e r s u b p o p u l a t i o n o f t u m o r cells persist. 8,21 C h a n g e s in e x p r e s s i o n levels o f a n u m b e r o f d i f f e r e n t i a t i o n - a s s o c i a t e d m o l ecules during the transition from an androgen-sensitive to a n a n d r o g e n - i n s e n s i t i v e c a r c i n o m a m a y p o i n t to mechanisms responsible for the development of androg e n i n d e p e n d e n c e . T h e p u r p o s e o f this study is to analyze p h e n o t y p i c a l t e r a t i o n s o f p r o s t a t i c a d e n o c a r c i n o m a cells b o t h d u r i n g s h o r t - t e r m a n d r o g e n a b l a t i o n a n d at t h e stage o f h o r m o n a l failure, with t h e a i m o f d e f i n i n g m o l e c u l a r targets f o r h o r m o n a l e s c a p e .
MATERIALS AND METHODS
Clinical Data and Samples Tissue samples were obtained from radical prostatectomy specimens, transrectal needle biopsies, or transurethral resections. A representative sample was taken from each tissue for histopathological and immunohistochemical assessment, and an adjacent piece was placed in liquid nitrogen for RNA extraction when available. Clinical and pathological characteristics are presented in Table 1. Three groups of patients were distinguished: group 1 patients without any hormonal manipulation (n = 24/: group 2 patients with hormonal deprivation (n = 15): group 3 panents with patent hormonal failure (n = 10); using objective criteria as defined by the National Prostatic Cancer P r o j e c t P In group 2, hormonal blockade was obtained using classical androgen deprivation with a luteinizing hormone-releasing h o r m o n e (LH-RH) analog in 13 patients or by orchidectomy in two patients (Table 2). In group 3, six patients had hormonal deprivation using LH-RH analogs, and four had an orchidectomy. Tissue samples were
TABLE 1. Clinical and Pathological Features of Tumor Specimens Used
Stage and Grade
Group 1 Untreated CaP (N= 24)
Group 2 CaP Hormonal Deprivation (N= 15)
T2 T3 T4 TX Gleason <6 Gleason >6 Gleason X NqM+
14 5 5 0 8 16 0 2 1
5 6 3 1 2 11 2 6 3
Group 3 CaP Hormonal Failure (N = 10) 0 2 2 6 0 9 1 4 8
TABLE 2. HormonalTreatment in Patients of Groups 2 and 3 Treatment
Group 2 CaP Hormonal Deprivation
Group 3 CaP Hormonal Failure
LH-RH Orchidectomy
13 2
6 4
obtained between 3 and 36 months (average, 6 months), and four tissue samples after 60 months of treatment. For all patients in group 2. there was no clinical evidence of tumor progression. This included a persistently low PSA level in patients with at least two PSA determinations at 4-week intervals.
Immunostaining of Cytokeratin-5 (RCK-103) and Cytokeratin-18 (RGE-53) Fresh frozen cryosections (4 pm) were air dried and fixed in chilled acetone for 10 minutes. Slides were incubated in blocking solution (nonfat dry milk, 5 % i n phosphate-buffered saline [PBS]) for 30 minutes. Subsequently, slides were incubated with mouse monoclonal antibody RCK-103 (anticytokeratin 5) or RGE-53 (anti-cytokeratin 18) hybfidoma culture supernatant for 60 minutes at room temperature. 7 Slides were washed in PBS and incubated with horseradish peroxidase-conjugated rabbit anti-mouse immunoglobulins (Dako, Copenhagen-Denmark) for 45 minutes at room temperature. Next, sections were washed in PBS and incubated in 3,3' diaminobenzidine tetrahydrochloride (Sigma, France) for 5 minutes or APAAP complex with appropriate bridge antibody system (Dako, Copenhagen-Denmark). Counterstaining was performed with Harris hematoxylin. Sections were dehydrated and mounted with i m m u n o m o u n t (Shandon, UK) or resin Eukitt (Kindler GmbH, Freiburg, Germany).
immunostaining of EGF receptor Fresh frozen cryosections (4 lain) were air-dried and fixed in chilled acetone for 10 minutes and incubated in blocking solution (nonfat dry milk, 5% in TRIS-buffered saline [TBS], pH 7.2) for 30 minutes. Then the slides were incubated with a mouse anti-EGF receptor monoclonal antibody (clone 528) (10 p g / m L ) , which recognizes an extracellular epitope (Oncogene Science. Paris, France), for 60 minutes at room temperature, washed in TBS, and incubated with a secondary bridge rabbit anti-mouse antibody (Dako, Copenhagen, Denmark) for 30 minutes. Slides were washed in TBS and incubated with a mouse APAAP complex (Dako, Copenhagen, Denmark) rinsed in TBS, and consequently the reaction was developed with the fast red chromogen substrate (Sigma, France) for 30 minutes. Slides were counterstained with Mayer's hematoxilin and mounted with immunomount.
Immunostainin~ With an Antihuman Androgen Receptor (F-39-4-1) Frozen sections (4 pm) were air dried and fixed for 5 minutes with paraformaldehyde (4% in PBS), incubated in chilled methanol (-20°C, 4 minutes), dehydrated in cold acetone (-20°C) for 2 minutes, and incubated with a mouse monoclonal anti-human androgen receptor antibody (1/ 2.000) (ascites) overnight at 4°C. 3 The indirect APAAP procedure was performed as described.
Immunostaining Scoring For each slide, the percentage of stained tumor cells and the intensity (0 -= negative, 1 = low, 2 = moderate, 3 - in-
1006
PHENOTYPE OF PROSTATE CANCER PROGRESSION (GiI-Diez de Medina et al)
FIGURE 1. Expression of cytokeratin 18 and 5 in human prostate carcinoma. Frozen sections (4 IJm) of prostatic adenocarcinoma (A, D) without hormonal treatment, (B, E) with hormonal deprivation, and (C, F) with hormonal failure. (A, B, C) Immunostaining with RGE-53 (CK-18) shows positive staining of tumor cells, in all cases. (D, E, F) Immunostaining with RCK-103 (CK-5) (D) shows negative immunostaining of tumor cells from patients without hormonal treatment. (E, F) Expression of CK-5 in the tumor cells after hormone deprivation and hormonal failure. (A, B, D, E, original magnification x200; C and F, original magnification x500.) tense) of staining was recorded using a light microscope equipped with an eyepiece reticule at 400× magnification. The score was determined in each case after counting at least 100 to 500 tumor cells in three to five selected fields.
RNA Extraction The RNA was extracted from frozen tissue samples for a selected group of specimens (six peripheral zone [PZ] from normal prostates, 11 adenocarcinomas of group 1; patients without any hormonal manipulation, eight adenocarcinomas of group 2; patients with hormonal deprivation and 10 adenocarcinomas of group 3; patients with patent hormonal
failure). In all selected adenocarcinomas, histological adjacent sections from the samples used for RNA extraction were composed of more than 60% tumor cells. RNA was prepared according to Chirgwin et al, 23 using ultracentrifugation on a 5.7 m o l / L cesium chloride cushion after tissue homogenization in 4 m o l / L guanidium thiocyanate. RNA was subsequently extracted twice with pheno! chloroform.
RT-PCR The level ofAR mRNA24was determined by semiquantitafive reverse transcription-polymerase chain reaction (RTPCR), by comparison with an internal control as previously
1007
HUMAN PATHOLOGY
Volume29, No, 9 (September 1998)
described. 25,~6GAPDH or the TATA binding protein (TBP), a ubiquitous transcription factor, was used as an internal control. 27 cDNA was prepared using 1 lag RNA from each sample. The semiquantitative polymerase chain reaction was performed according to Gil Diez de Medina et al27 using 2 laL (1/50) of the reverse transcription product in a final volume of 50 laL containing all four dNTPs (each at 100 lamol/L), with 1 lamol/L of each primer (a pair for GADPH, together with a pair specific for AR or TBP) and 2 pCi of ~[~2P]dCTE The number of cycles was chosen to be in the exponential phase of the two PCR reactions. We used 21 cycles for the co-amplification of AR and GAPDH and 23 cycles for the co-amplification of TBP and GAPDH. The primers' sequences were: AGTGAAGAACAGTCCAGACTG and CCAGGAAATAACTCTGGCTCAT (TBP); CTGCACCACCAACTGCT'FAG and AGGTCCACCACTGACACGTT (GAPDH); GACTTCACCG~ ACCTGATG and CTGGCAGTCTCCAAACGCAT (AR). All primers are given in the sense 5' to 3' direction. The primers for the AR cDNA were chosen in the first and second exons34 The amplification reactions were executed in a PTC-200 programmable thermocycler (MJ Research, Watertown, MA) with an initial cycle of 95°C for 5 minutes before the addition of Hi-Taq DNA thermostable polymerase (Bioprobe, France). Thereafter, each cycle was as follows: 94°C fbr 1 minute, 57°C for 1 minute, and 72°C for 80 seconds. These cycles were followed by a final incubation step at 72°C for 10 minutes. The PCR-amplified products were loaded in duplicate and electrophoresed in 8% polyacrylamide gels, fixed in 7% acetic acid, and vacuum-dried. Autoradiograms showed two bands corresponding to the co-amplified fragments. The signals were quantified with a Molecular Dynamics 300 Phosphorlmager (Molecular Dynamics, Sunnyvale, CA). Each measurement was repeated for three independent PCR reactions and found to be identical within a 15% margins. Negligible amplification was observed when reverse transcriptase was omitted from the reverse transcription reaction.
Statistics Analysis of variance was performed, using the KruskalWallis nonparametric ANOVA test, to compare the immune phenotype of tumor cells from the three groups of patients and for the RT-PCR results.
RESULTS Expression of Cytokeratins 5 a n d 18 in Prostatic C a r c i n o m a Cells Most carcinoma cells (Fig 1A, B), in b o t h u n t r e a t e d (group 1) and androgen-deprived specimens (group 2), expressed cytokerafin 18 as defined by their reactivity with RGE-53. In the tumor-treated group, we f o u n d a significant decrease in p e r c e n t a g e of RGE-53 staining c o m p a r e d with g r o u p 1. A similar descending trend was observed for the intensity of staining during h o r m o n a l failure (group 3) (Table 3). However, in this group, the staining was m o r e h e t e r o g e n e o u s (Fig 1C). T h e basal cell marker, RCK-103, directed against cytokeratin 5 was positive for only a small p e r c e n t a g e of t u m o r cells (14.3% + 3.4) f r o m u n t r e a t e d g r o u p 1 tumors (Fig 1D). A m o r e distinct staining reaction was observed in g r o u p 2 (Fig 1E) a n d g r o u p 3 (Fig 1F)
TABLE $. Percentage of Positive Tumor Ceils and Intensity for RGE 53 (Cytokeratin 18) and for RCK 103 (Cytokeratin 5) in Relation to the Hormonal Status of the Tumor CK18 (RGE 53)
CK5 (RCK103)
Percentage Percentage (%) Positive (%) Positive N Ceils -+ SEM Intensity Cells + SEM Intensity Untreated CaP 24 85.3 + 2.1 CaP hormonal deprivation 15 66.7 ___6.0 CaP hormonal failure 10 63.3 +_10.3
3
14.3 _ 3.4
1 to 2
2 to 3
47.0 + 6.2
1 to 3
0 to 3
38.3 -+ 10.5
2 to 3
NOTE. Nonparametric ANOVA test: RGE-53 P = .002 and RCK-103P = .0004.
tumors (respectively, 47.0% + 6.2 a n d 38.3% - 10.5) with an increased staining intensity (Table 3). In groups 2 a n d 3, we n o t e d in adjacent sections the coexpression of CK-8-18 a n d CK-5 in t u m o r cells, suggesting an intermediate basal a n d secretory cells phenotype.
EGF Receptor Expression In the 24 u n t r e a t e d specimens (group 1), basal cells of n o r m a l and hyperplastic glands r e p r e s e n t e d an internal positive control (arrow) (Fig 2A). In these cases only a few prostatic c a r c i n o m a cells expressed EGF receptor (14.7% ___4.5) with low intensity (score 1). In contrast, a significant increase in the n u m b e r of positive cells with somewhat e n h a n c e d labeling intensity was observed in g r o u p 2 a n d 3 samples (Fig 2B, C) with 34.5% ± 6.8 (score 1 to 2) a n d 46.3% - 10.8 (score 1 to 2), respectively (Table 4).
A n d r o g e n Receptor Expression In u n t r e a t e d prostatic adenocarcinomas, m o s t tum o r cells (64.9% +_ 3.9) displayed a u n i f o r m nuclear staining intensity equivalent to the staining observed in n o r m a l secretory cells (Fig 2D). In treated tumors (group 2), the fraction of i m m u n o s t a i n e d t u m o u r cells (Fig 2E) was markedly r e d u c e d (32.3% - 8.4) with a h e t e r o g e n e o u s pattern of nuclear expression (staining intensity f r o m 1 to 3). At the stage o f h o r m o n a l failure (group 3), a significant increased p e r c e n t a g e (66,3 +__ 4.1%, 2 to 3) of stained t u m o r cells a n d intensity was detected (Table 5). Remarkably, cytoplasmic staining was occasionally seen in cases o f h o r m o n a l refractory tumors in addition to typical nuclear staining (Fig 2F).
RT-PCR T h e AR m R N A level was d e t e r m i n e d by semiquantitative RT-PCR. 27 T h e means of a n d r o g e n receptor m R N A expression in the g r o u p 1 a n d 2 prostatic carcinomas were c o m p a r a b l e to n o r m a l peripheral prostates (average = 0.075 ___0.01). However, we observed high expression in the h o r m o n a l failure g r o u p
1008
PHENOTYPE OF PROSTATE CANCER PROGRESSION (GiI-Diez de Medina et al)
FIGURE 2. Expression of the epidermal growth factor and the androgen receptor in prostate carcinoma ceils. Frozen sections (4 IJm) of prostatic adenocarcinoma (A, D) without hormone manipulation, (B, E) with hormone deprivation, and (C, F) with hormonal failure. (A, B, C) Immunostaining with an anti-EGF-R antibody (A)-negative immunostaining of tumor cells, and preservation of staining in basal cells from adjacent benign glands (arrow), (B, C) positive staining of the tumor cells. (D, E, F) Immunostaining with an anti-AR antibody (39-4-1) (D) shows positive immunostaining in the nucleus of the tumor cells, (E) exhibits decreased immunostaining after hormone deprivation, and (F) expression of the AR in the nucleus of tumor cells in this subset of hormone-refractory prostatic cancer cytoplasmic staining was observed in many different tumor cells (A, D, E, F, original magnification ×200; B and C, original magnification ×500,)
(group 3). The increase in mRNA expression (average = 0.62 + 0.2) was eightfold when compared with other groups and normal peripheral prostatic tissue (P = .0003). A similar eighffold increase was obtained when another internal standard (TBP) was used instead of GADPH (data not shown).
In benign tissue cell-type separated, 2s using an enzymatic digestion combined with percoll gradient centrifugation (data not shown), there was an equivalent number of AR transcripts expressed by the epithelial cell fractions when compared with the stromal cell fractions.
1009
HUMAN PATHOLOGY
Volume 29, No. 9 (September 1998) TABLE 5.
1.8
1.6
E
1.4
•I" a.
1.2
.< 0 ~.
1.0
Androgen Receptor Immunostaining
Untreated CaP CaP hormonal deprivation CaP hormonal failure
N
Percentage (%) PositiveCells
Intensity
24 15 10
64.9 +- 3.9 32.3 -+ 8.4 66.3 -+ 4.1
2 to 3 1 to 3 2 to 3
NOTE. Nonparametric ANOVAtest P = .007.
0.8 9,
< Z
0.6
E
0.4
n,, <[~
0.2 0
+•
+ Normal Prostate n = 6
I
+
Group 1
Group 2
Group 3
n =11
n = 8
n =10
Semiquantitative mRNA expression of androgen receptor. The expression level of androgen receptor mRNA was determined by means of semiquantitative RT-PCRusing GADPH as internal control: in normal peripheral prostatic tissue (n = 6); group 1 (n= 11); group 2 (n= 8); and group 3 (n = 10) of prostatic adenocarcinomas. The bar in each group represents the means of value. The eightfold increase in AR mRNA levels in hormone-refractory population compared with the other groups is statistically significant (P : .0003) using nonparametric ANOVA. FIGURE 3,
DISCUSSION
According to the prostatic stem cell model, the prostatic glandular basal cell population comprises a stern cell population committed to differentiate into amplifying cells that may rapidly proliferate and terminally differentiate into secretory cells29 In this model, it was also hypothesized that the differentiated androgend e p e n d e n t secretory cells are the progeny o f a subpopulation of self-replicating cells that do not require androgens for their survival. It has been suggested that this amplifying cell population shares differentiation features with basal and secretory cells. In particular, the demonstration in rat prostatic tissue of cells expressing basal cell type and secretory cell type cytokeratins is in agreement with this concept29,3° Thus, cells with a combined expression of basal and secretory type cytokeratins could be considered as candidates of amplifying ceils. In this study, we confirmed that prostatic carcin o m a cells co-express cytokeratins characteristic of both basal and secretory cells. 7 This observation agrees with the plasticity of basal cells that give rise to different TABLE 4.
FGF Receptor Immunostaining
Untreated CaP CaP hormonal deprivation CaP hormonal failure
N
Percentage (%) Positive Ceils - SEM
Intensity
24 15 10
14.7 + 4.5 34.5 --- 6.8 46.3 --- 10.8
1 1 to 2 1 to 2
NOTE. Nonparametric ANOVAtest P = .003. 1010
epithelial cell lineages in neoplastic h u m a n prostate. ~1 In addition we showed that the proportion of the cytokeratin 5-positive subpopulation was increased after androgen deprivation during both the regression stage and after h o r m o n a l escape. This increased percentage of cytokeratin 5-positive cells can be considered as the outcome of a selection process based on the androgen i n d e p e n d e n c e of a less mature putative amplifying cell subpopulation. Conversely, cellular adaptation to an altered endocrine environment may account for the observed increase in population o f cytokeratin 5-positive cells. EGF-R is present in basal cells, but not in secretory epithelial cells of the prostatic glands. Previous studies in the rat have shown an increase in EGF-R expression in normal or hyperplastic prostate tissue after androgen ablation. ~2 (Immunohistochemical studies have suggested that this increase was not due to overexpression of the EGF-R, by secretory cells. In fact, e n r i c h m e n t or positive selection for the (EGF-R positive) basal cell population underlies the observed increase in EGF-R in castrated animals). Several studies have shown the presence of EGF-R in untreated primary prostate cancers, with conflicting results with regard to expression levels and tumor grade. 16,32Differences in results can be attributed to variation in immunohistochemical assays, antibodies used (eg, directed against extracellular v intracellular domain of EGF-R) and tissue fixation. We p e r f o r m e d our study on frozen sections, gently fixed in acetone, and f o u n d a comparatively low percentage (average, 14.7%) of EGF-R-positive t u m o r cells in primary untreated prostatic adenocarcinomas. It should be n o t e d that other studies r e p o r t e d the percentage of EGF-R-positive tumors or intensity scores of immunostaining but not the percentage of immunolabeled t u m o r ceils. 16,32 Notably, we f o u n d a significant increased percentage of EGF-R-positive t u m o r cells in androgen-sensitive carcinomas shortly after androgen deprivation (group 2) with a further rise in endocrine therapy-resistant tumours. Similarly, Scher et a.119 recently r e p o r t e d h o m o g e n o u s EGF-R expression in metastases of endocrine therapy-resistant prostatic cancers. Furthermore, both in primary carcinomas and in metastatic a n d r o g e n - i n d e p e n d e n t tumours, the presence of TGFot suggests the existence o f an autocrine loop in these t u m o r ceils. 16,19 Again, these data can be interpreted as a selection process for EGF-R-positive tumor cells, or as an adaptation to a low androgen environment. It has been r e p o r t e d that distant metastases of prostate cancer express AR protein. 3a In addition, endocrine therapy-resistant prostatic carcinomas dis-
PHENOTYPE OF PROSTATECANCER PROGRESSION (GiI-Diez de Medina et al)
play high AR expression levelsP 4 This level is increased when compared with untreated primary prostatic carcinomas. Both genetic 35 and epigenetic mechanisms may be involved in the control of AR expression in androgenindependent tumorsP 6 In this study, the observed decreased in percentage of AR-positive tumor cells early after androgen deprivation when compared with the proportion of AR-positive tumor cells found in untreated carcinomas is in agreement with recent preliminary observations on another group of patients, s7 We hypothesize that AR-positive cell populations are maintained while the AR-negative tumor cells are eliminated after long-term androgen deprivation. These results suggest that androgen-depleted environments may select for AR-positive prostatic tumor cells. 3s The increased proportion of AR-positive tumor cells in the hormonotherapy-resistant prostate cancers was associated with an increase in AR mRNA as compared with untreated cancers. Using mRNA in situ hybridization, Koivisto et al so recently reported a similar observation. Cytoplasmic staining of the androgen receptoi, as observed in this study, may be related to the very high concentration of the receptor in the cell or to interactions of the AR receptor with co-regulatory proteins. This finding merits further investigations. Amplification of the AR gene and multiple X chromosomes was shown in a large proportion of endocrine therapy-resistant prostate cancers. 35,39 It is not clear whether a gene dosage effect alone accounts for the observed increase in mRNA coding AR. In conclusion, we have shown that progression of untreated prostatic carcinoma to local endocrine therapy-resistant carcinoma is associated with differentiationassociated features of the tumor cell population. Doublelabeling experiments are required to define more accurately the changes in the various cell populations during the transition to androgen independence. The inappropriate expression of EGF-R and cytokeratins observed during hormone failure is already apparent during the early androgen-sensitive stage of prostate cancer. AR overexpression arises in hormonal refractory prostate cancer after a downregulation in hormoneresponsive tumors. This observation suggests that increased expression of AR may play a role in this process. Further experimental studies should be designed to evaluate whether a relationship exists between AR and EGF-R expression during the progression of prostate cancer.
Acknowledgment. T h e authors t h a n k Dr Frans C.S. Ramaekers (University of Maastricht, T h e Netherlands) for fruitful advice and for generous d o n a t i o n of m o n o c l o n a l antibodies RCK-103 and RGE-53. They also thank Drs Aria Mafia Vallts and Ivan Coulter for helpful discussions and for critical reading of the manuscript. T h e authors thank Domin i q u e M o r i n e a u for e x p e r t p h o t o g r a p h i c assistance, A n n i e Maurette, and Catherine Petit for secretarial assistance. REFERENCES 1. Bostwick DG, Pacelli A, Lopez-Beltran A: Molecular biology of prostatic intraepithelial neoplasia. Prostate 29: l 17-134, 1996
2. Isaacs WB, Bova GS, Morton RA, et al: Molecular biology of prostate cancer. Semin Oncol 21:514-521, 1994 3. Ruizeveld de WinterJA, Trapman J, Vermey M, et al: Androgen receptor expression in human tissues: An immunohistochemical study. J Histochem Cytochem 39:927-936, 1991 4. Sherwood ER, Berg LA, Mitchell NJ, et al: Differential cytokeratin expression in normal, hyperplastic and malignant epithelial cells from human prostate. J Uro1143:167-171, 1990 5. Sherwood ER, Theyer G, Steiner S, et al: Differential expression of specific cytokeratin polypeptides in the basal and luminal epithelia of the human prostate. Prostate 18:303-314, 1991 6. Maddy SO~ Chisholm GD, Hawkins RA, et al: Localization of epidermal growth factor receptors in the human prostate by biochemical and immunocytochemical methods. J Endocrinol 113:147-153, 1987 7. Verhagen APM, Ramaekers FCS, Aalders TW, et 9_1:Colocalization of basal and luminal cell-type cytokeratins in human prostate cancer. Cancer Res 52:6182-6187, 1992 8. Bruchovsky N, Lesser B, Van Doorn E, et al: Hormonal effects on cell proliferation in rat prostate. Vitamins Horm 33:61-102, 1975 9. Vaillancourt L, Tttu B, Fradet Y, et al: Effect of neoadjuvant endocrine therapy (combined androgen blockade) on normal prostate and prostatic carcinoma: A randomized study. AmJ Surg Pathol 20:86-93, 1996 10. Cunha GR: Growth factors as mediators of androgen action during male urogenital development. Prostate 6:22-25, 1996 11. Thomson AA, Foster BA, Cunha GR: Analysis of growth factors and receptor mRNA levels during development of the rat seminal vesicle and prostate. Development 124:2431-2439, 1997 12. Fukabori Y, Yah G, Yamanaka H, et al: Rapid induction of keratinocyte growth factor (FGF-7) and beta-actin after exposure of prostate stromal cells to androgen. In Vitro Cell Dev Bio130A:745-746, 1994 13. Culig Z, Hobisch A, Cronauer MV, et al: Androgen receptor activation in prostatic turnout cell lines by insulin-like growth factor-I, keratinocyte growth factor and epidermal growth factor. Cancer Res 54:5474-5478, 1994 14. Ibrahim GK, Kerns BJM, MacDonald JA, et al: Differential immunoreactivity of epidermal growth factor receptor in benign, dysplastic and malignant prostatic tissues. J Uro1149:170-173, 1993 15. Taylor TB, Ramsdell JS: Transforming growth factor-~ and its receptor are expressed in the epithelium of the rat prostate gland. Endocrinology 133:1306-1311, 1993 16. Glynne-Jones E, Goddard L, Harper ME: Comparative analysis of mRNA and protein expression for epidermal growth factor receptor and ligands relative to the proliferative index in human prostate tissue. HUM PATHOL27:688-694, 1996 17. Fiorelli G, De Bellis A, Longo A, et al: Epidermal growth factor receptors in human hyperplastic prostate tissue and their modulation by chronic treatment with a gonadotropin-releasing hormone analog.J Clin Endocrinol Metab 68:740-743, 1989 18. Hofer DR, Sherwood ER, Bromberg WD, et al: Autonomous growth of androgen-independent hmnan prostatic carcinoma cells: Role of transforming growth factor a. Cancer Res 51:2780-2785, 1991 19. Scher HI, Sarkis A, Reuter V, et al: Changing pattern of expression of the epidermal growth factor receptor and transforming growth factor alpha in the progression of prostatic neoplasms. Clin Cancer Res 1:545-550, 1995 20. Yang Y, Chisholm GD, Habib FK: Epidermal growth factor and transforming growth factor alpha concentrations in BPH and cancer of the prostate: Their relationships with tissue androgen levels. BrJ Cancer 67:152-155, 1993 21. Bruchovsky N, Rennie PS, Coldman AJ, et al: Effects of androgen withdrawal on the stern cell composition of the Shionogi carcinoma. Cancer Res 50:2275-2282, 1990 22. Schmidt JD, Scott WW, Gibbons R, et al: Chemotherapy programs of the National Prostatic Cancer Project (NPCP). Cancer 45:1937-1946, 1980 23. Chirgwin JM, Przybyla AE, MacDonald RJ, et al: Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18:5294-5299, 1979 24. TilleyWD, Marcelli M, WilsonJD, et al: Characterization and expression of a cDNA encoding the human androgen receptor. Proc Natl Acad Sci U S A 86:32%331, 1989
1011
HUMAN PATHOLOGY
Volume 29, No. 9 (September 1998)
25. Kandel J, Bossy-Wetzel E, Radvanyi F, et al: Neovascularization is associated with a switch to the export of bFGF in the multistep development of fibrosarcoma. Cell 66:1095-1104, 1991 26. Radvanyi F, Christgau S, Baekkeskov S, et al: Pancreatic beta cells cultured from individual preneoplastic foci in a multistage tumourigenesis pathway: A potentially general technique for isolating physiologically representative cell lines. Mol Cell Biol 13:4223-4232, 1993 27. Gil Diez de Medina S, Chopin D, E1 Marjou A, et al: Decreased expression of keratinocyte growth factor receptor in a subset of human transitional cell bladder carcinomas. Oncogene 14:323-330, 1997 28. Kozlowski MJ, McEwan R, Keer H, et al: Prostate cancer and the invasive phenotype: Application of new in vivo and in vitro approaches, in Fidler IJ, Nicholson G (eds): Tumour Progression and Metastasis. NewYork, NY, Alan R. Liss, 1988, pp 189-231 29. Verhagen APM, Aalders TW, Ramaekers FCS, et al: Differential expression of keratins in the basal and luminal compartments of rat prostatic epithelium during degeneration and regeneration. Prostate 13:25-38, 1988 30. HsiehJT, Zhan HE, Wang XH, et al: Regulation of basal and luminal cell-specific cytokeratin expression in rat accessory sex organs: Evidence for a new class of androgen-repressed genes and insight into their pairwise control.J Biol Chem 267:2303-2310, 1992 31. BonkhoffH, Stein U, Remberger K: Multidirectionnal differentiation in the normal, hyperplastic and neoplastic human prostate: Simultaneous demonstration of cell-specific epithelial markers. HUM PATHOL25:42-46, 1994
32. Traish AM, Wotiz HH: Prostatic epidermal growth factor receptors and their regulation by androgens. Endocrinology 121:14611467, 1987 33. Hobisch A, Culig Z, Radmayr C, et al: Distant metastases from prostatic carcinoma express androgen receptor protein. Cancer Res 55:3068-3072, 1995 34. Van der Kwast TH, SchalkenJA, Ruizeveld de WinterJA, et al: Androgen receptors in endocrine therapy resistant human prostate cancer. IntJ Cancer 48:189-193, 1991 35. Koivisto P, Hyytinen E, Palmberg C, et al: Analysis of genetic changes underlying local recurrence of prostate carcinoma during androgen deprivation therapy. AmJ Patho1147:1608-1614, 1995 36. Kokontis J, Takakura K, Hay N, et al: Increased androgen receptor activity and altered c-myc expression in prostate cancer cell after long-term androgen deprivation. Cancer Res 54:1566-1573, 1994 37. Van der Kwast TH, T~tu B, Fradet Y, et al: Androgen receptor modulation in benign human prostatic tissue and prostatic adenocarcinoma during neoadjuvant endocrine combination therapy. Prostate 28:227-231, 1996 38. Culig Z, Hobisch A, Cronauer MV, et al: Mutant androgen receptor detected in an advanced stage prostatic carcinoma is activated by adrenal androgens and progesterone. Mol Endocrinol 7:1541-1550, 1993 39. Koivisto P, KononenJ, Palmberg C, et al: Androgen receptor gene amplification: A possible molecular mechanism for androgen deprivation therapy failure in prostate cancer. Cancer Res 57:314-319, 1997
1012