| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
The Urologic Laboratory (R.M.A., J.G.B., M.R.F.), Department of Urology, Children’s Hospital, and the Department of Surgery, Harvard Medical School, Boston, Massachusetts 02115; Division of Urology (K.R.L.), Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts 02115; and Department of Urology (B.J.W., J.A.E.), Louisiana State University Medical Center, Shreveport, Louisiana 71130
Address all correspondence and requests for reprints to: Dr. Michael R. Freeman, Enders Research Laboratories, 1161, Children’s Hospital, 300 Longwood Avenue, Boston, Massachusetts 02115. E-mail: freeman_m{at}a1.tch.harvard.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, amphiregulin (ARG), HB-EGF,
betacellulin, and epiregulin, and the related molecule heregulin-,
in a series of 10 prostate tissue specimens. Only EGF showed a
disease-specific association, with increased mRNA levels in four of
five PCa specimens in comparison to matched normal tissue from the same
subject. In contrast, ARG and HB-EGF mRNAs showed a coordinate pattern
of expression in 7/10 specimens that was distinct from all other growth
factor or receptor genes examined and from mRNAs for prostate specific
antigen, the androgen receptor and GAPDH, a housekeeping enzyme.
Analysis of an additional series of benign prostatic hyperplasia and
prostate cancer specimens from 60 individuals confirmed that ARG and
HB-EGF mRNA levels varied in a highly coordinate manner (r = 0.93;
P < 0.0001) but showed no association with
disease. ARG was immunolocalized largely to interstitial smooth muscle
cells (SMC), previously identified as the site of synthesis of HB-EGF
in the prostate, while the cognate ARG and HB-EGF receptor, ErbB1, was
localized exclusively to ductal epithelial cells and carcinoma cells.
Although ARG was a relatively poor mitogen for Balb/c3T3 cells in
comparison to HB-EGF, it was similar in potency to HB-EGF in
stimulating human prostate epithelial cell growth, suggesting that
prostate epithelia may be a physiologic target for ARG in
vivo. Expression of both ARG and HB-EGF mRNAs was induced in
cultured prostate SMC by fibroblast growth factor-2, a human prostate
SMC mitogen linked to prostate disease. These findings indicate that
ARG and HB-EGF are likely to be key mediators of directional signaling
between SMC and epithelial cells in the human prostate and appear to be
coordinately regulated. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Stromal-epithelial interactions have also been proposed to determine the natural history of tumors arising from epithelial organs (4, 8, 9, 10, 11, 12). This hypothesis has been explored extensively using cell recombination models, which have demonstrated that cells present in tumor stroma, such as undifferentiated fibroblasts of several types, are capable of regulating growth rates, phenotypic differentiation and hormonal sensitivity of carcinoma cells (4, 13, 14, 15). Descriptive observations of carcinomas likewise support the physiological relevance of a stromal-epithelial interaction in tumor progression. Tumor stroma has been documented to "react" to the presence of associated carcinoma cells by altering (usually increasing) the expression levels of certain secreted proteins. Proteins capable of paracrine signaling which have been identified as being up-regulated in the reactive stroma of carcinomas include transforming growth factor (TGF)-ß, stromelysins (16, 17, 18, 19) and vascular endothelial growth factor (20). Functional effects of stromal cell products on carcinoma growth in vivo have also been demonstrated. For example, aromatase activity produced locally by the tumor stroma is capable of promoting estrogen-dependent growth of human breast carcinomas in vivo (21).
In a series of recent reviews (22, 23, 24), Cunha and colleagues have pointed out that the physiologically relevant stromal-epithelial interaction in the prostate under normal conditions is likely to be an epithelial-SMC interaction. This argument derives primarily from correlative data in which epithelial and SMC differentiation occur coordinately within epithelial-mesenchymal tissue recombinants studied in model systems (22, 23, 24) and from the close apposition between SMC and ductal epithelial cells observed microscopically (25) in tissue sections. SMC are thus anatomically positioned to secrete paracrine factors capable of activating signaling cascades within the epithelial cells of the prostatic ducts. Although a few candidate stroma-derived soluble factors have been identified, which may act as mediators of epithelial cell growth and phenotypic differentiation in the prostate (26, 27, 28), persuasive evidence for the hypothetical paracrine factors that mediate epithelial-SMC interactions specifically has not been reported.
We recently identified heparin-binding epidermal growth factor
(EGF)-like growth factor (HB-EGF) as a product predominantly of
interstitial and vascular SMC of the human prostate (29). HB-EGF is one
of six known activating ligands for the epidermal growth factor
receptor (EGF-R)/ErbB1 receptor tyrosine kinase expressed by human
cells (30). In the present study we demonstrate that a structurally
related ErbB1 ligand, amphiregulin (ARG), is also expressed
predominantly in the SMC of the prostatic stroma and, on the basis of
an analysis of a large series of human prostate tissue specimens,
appears to be coordinately expressed with HB-EGF. ErbB1 has been
identified as a positive regulator of normal and transformed prostate
epithelial cells and is expressed predominantly by basal epithelial
cells in the ductal network and by prostate carcinoma cells (31, 32).
The basal location of ErbB1 in the prostatic ducts, combined with the
observation that HB-EGF and ARG are synthesized by the SMC of the
interstitial stroma, suggests that these growth factors form a
unique class of mediators of directional SMC 
epithelial
signaling in the prostate.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, heparin-binding EGF-like
growth factor (HB-EGF), amphiregulin (ARG), betacellulin (BTC), and
heregulin- (HRG-) were obtained from R+D Systems (Minneapolis,
MN). Antipeptide antibodies for amphiregulin immunostaining were a
generous gift from Dr. Donna Davies (University of Southampton, UK).
Antiamphiregulin antibody, Ab-1, and anti-ErbB1 antibody were obtained
from Neomarkers (Fremont, CA). Antibodies against CD44, anti--smooth
muscle actin and antidesmin antibodies were from Sigma
Immunochemicals, and antibody against high molecular weight (HMW)
cytokeratin (34ßE12), was from DAKO Corp., (Carpenteria,
CA). The Vectastain ABC kit was from Vector Laboratories, Inc. (Burlingame, CA). FBS was from HyClone Laboratories, Inc. (Logan, UT). Normal prostate epithelial
cells, dissociation reagents, serum-free medium and supplements were
obtained from Clonetics Corp. (San Diego, CA).
Methods
Processing and RNA extraction of human prostate tissue
samples. Prostate tissue biopsy specimens comprising normal
prostate, tissue from benign prostatic hyperplasia, and tissue from
prostate carcinoma were obtained at the time of surgery (Department of
Urology, Brigham and Women’s Hospital, Boston, MA; and Department of
Urology, Louisiana State University Medical Center, Shreveport, LA) as
per IRB approval. Tissue specimens were snap frozen in liquid nitrogen
and stored at -80 C until required. Tissue samples were thawed in
TRIzol reagent and minced finely with razor blades to disperse tissue
fragments, before proceeding with the extraction procedure, according
to the manufacturer’s protocol. Microcarrier gel was added to samples
at the start of the procedure to assist in extraction of small amounts
of RNA and maximize yield. RNA pellets were reconstituted in
dH2O rendered RNase-free by diethylpyrocarbonate treatment
and RNA yield and purity were determined following measurement of
absorbance at 260 and 280 nm.
Complementary DNA (cDNA) preparation and amplification.
Expression of messenger RNAs (mRNAs) of interest was assessed by RT-PCR
following first-strand synthesis and precipitation of cDNA. Briefly, 3
µg total RNA was reverse-transcribed using oligo-dT primer and the
reverse transcriptase Superscript II, according to the manufacturer’s
protocol. First-strand synthesis was allowed to proceed at 42 C for 50
min, followed by denaturation of reverse transcriptase at 70 C for 15
min. cDNA was precipitated from the reaction mix by addition of 0.1 vol
linear acrylamide, 1 vol 4 M ammonium acetate, and 4 vol
absolute ethanol at room temperature. cDNA was pelleted by
centrifugation, pellets washed with 80% ethanol and reconstituted in
40 µl 10 mM Tris-Cl, 0.1 mM EDTA, pH 8. For
subsequent amplification, 2 µl of cDNA and 0.4 µM each
primer pair were used with 23 µl PCR Supermix, in the presence of
0.05 µl/reaction of 32P--dCTP. PCR cycling parameters
were as follows: 1 cycle of 94 C for 5 min, 30 cycles of 94 C for 30
sec, denaturation), 50–58 C for 30 sec (annealing; see Table 1
for specific annealing temperature for
primer pair), 72 C for 60 sec (extension), and 1 cycle of 72 C for 7
min. To ensure integrity and equivalence of RNA, separate reactions
were performed using primers specific for the housekeeping gene,
glyceraldehyde-3-phosphate dehydrogenase (GAPDH). All PCR products were
analyzed by gel electrophoresis through a 5.1% acrylamide gel in
1 x Tris-borate-EDTA (TBE) buffer and bands were identified
following exposure of x-ray film to dried gels. Densitometric analysis
was performed using an IS1000 Digital Imaging System (Alpha Innotech
Corp.) and band densities were normalized to that of GAPDH. Specific
primer sequences are presented in Table 1.
|
Culture and characterization of primary prostate epithelial and
smooth muscle cells. Normal human prostate epithelial cells (PrEC,
strain 4428) were obtained at first passage (Clonetics Corp.) and propagated in serum-free growth medium supplemented
with bovine pituitary extract (BPE), recombinant human EGF,
hydrocortisone, gentamicin-amphotericin (GA-1000), tri-iodothyronine,
insulin, transferrin, retinoic acid, and epinephrine; complete medium
was termed PrEGM. Cells were maintained in a humidified atmosphere at
37 C, 5% CO2 and used between passages 4 and 6.
Characterization of cells included RT-PCR for androgen receptor (AR)
and prostate-specific antigen (PSA) expression and immunocytochemical
analysis of high molecular weight (HMW) cytokeratin and CD44
expression. Normal human prostate smooth muscle cells (PrSMC) were
isolated from prostate tissue, obtained at the time of radical
prostatectomy, by enzymatic dispersion with collagenase. Cells were
expanded and maintained in MCDB105 medium supplemented with 10% FBS.
Cells were demonstrated to express a smooth muscle cell phenotype by
immunocytochemical staining with antibodies against -smooth muscle
actin and desmin. Experiments were conducted on PrSMC between passages
5 and 7.
Balb/c 3T3 thymidine incorporation assay. To evaluate the mitogenic potency of recombinant ARG and HB-EGF, the ability of these growth factors to promote tritiated thymidine (3H TdR) incorporation into Balb/c 3T3 fibroblasts was determined. Briefly, Balb/c 3T3 fibroblasts were seeded in DMEM/10% Colorado calf serum (CCS) at a density of 2 x 103 cells/well in microtitre plates and grown to confluence. Recombinant ARG, HB-EGF, or EGF was added to cells at the indicated concentrations together with 0.5 µCi/well 3H TdR, and cells were incubated for 48 h at 37 C. At the end of incubation, media were removed and cells were trypsinized, before harvesting on glass fiber filters and determination of incorporated radioactivity using a Betacount scintillation counter (Model 1450, Wallac, Inc., Gaithersburg, MD).
Clonal growth assay of primary prostate epithelial cells. The response of prostate epithelial cells to exogenous ARG and HB-EGF was determined in a clonal growth assay. Cells were seeded in 35-mm dishes at a density of 70 cells per dish in PrEGM lacking hEGF and allowed to plate down overnight. Recombinant ARG or HB-EGF was added to the medium at the indicated concentrations and cells were incubated at 37 C for 10 days. At the end of the incubation period, growth medium was aspirated and cells were fixed in 3.7% formaldehyde for 15 min. To visualize colony formation, fixed cells were stained with 0.1% crystal violet for 15 min and excess dye was removed by thorough washing in tap water.
ARG and HB-EGF gene expression in prostate smooth muscle cells. Regulation of ARG and HB-EGF mRNA expression in response to stimulation with fibroblast growth factor (FGF)-2 or TGF-ß1 was assessed in primary cultures of human prostate smooth muscle cells. Briefly PrSMC were seeded at 4–6 x 104 cells/well in six-well dishes and grown to 90% confluence in MCDB105/10% FBS. Before challenge with growth factors, cells were switched to serum-reduced medium (MCDB105/0.5% FBS) for at least 24 h. Cells were treated with 25 ng/ml FGF-2 or TGF-ß1 in MCDB105/0.5% FBS, for the indicated times; control cultures receiving only MCDB105/0.5% FBS were also included. At each time point, medium was aspirated and cells were scraped into 0.5 ml TRI reagent; samples were snap frozen in liquid nitrogen and stored at -80 C. RNA extraction, cDNA preparation and RT-PCR for GAPDH, ARG, and HB-EGF were performed as described above. To ensure amplification was proceeding in the linear range, serial dilutions of the cDNA were performed, allowing semiquantitative conclusions to be inferred from the data.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(HRG-), and TGF-, was demonstrated to a variable extent in all
specimens, however none of the samples expressed epiregulin (EPR),
consistent with its identity as a growth factor expressed primarily in
early development (33, 34).
Most of the receptors or ligands appeared not to be expressed in a
disease-specific manner, with the exception of EGF mRNA, which was
demonstrably higher in the tumor tissue in four of five subjects in
comparison to the matched controls. We also noted that the expression
patterns for ARG and HB-EGF varied in a coordinate manner in a majority
(7/10) of the tissue specimens. The epithelial cell component of the
tissue specimens was likely to be similar, based on comparable PSA mRNA
expression levels, suggesting that the apparent coordinate expression
pattern of ARG and HB-EGF mRNAs was not a reflection of variable
cellular content, but rather may be of biological significance.
Quantitative evaluation of densitometric data indicated that HB-EGF and
ARG mRNA levels only exhibited a positive correlation with each other.
BTC and EGF mRNA levels also exhibited a positive correlation with each
other (r = 0.72, with a positive slope of 0.94); however, a
disease-specific association for BTC was not evident as was the case
with EGF. EGF is known to be a secretory product of prostate epithelial
cells and is present at high levels in prostatic fluid (35).
Consequently, correlation of BTC and EGF mRNA levels suggests that BTC
is likely to be an epithelial cell product in the prostate. Consistent
with this possibility, we did not detect BTC mRNA in primary human
prostate SMC cultured in vitro, whereas expression was
detected in primary human prostate epithelial cells (PrEC) (data not
shown). Positive correlations were also observed between BTC and
HRG- (r = 0.47) and BTC and TGF- (r = 0.5). HRG- and
TGF- are typically epithelial cell products (30).
These results suggest that ARG and HB-EGF may represent a unique class
of prostatic ErbB1 ligand in that their expression appears to be
coordinately regulated in a manner not observed with other members of
the family. To further examine this possibility, ARG and HB-EGF mRNA
expression analysis was performed on an additional sixty human prostate
specimens comprising tissue from benign prostatic hyperplasia (BPH) and
from prostate carcinoma (PCa). Expression of mRNAs encoding ARG and
HB-EGF was normalized to that of the housekeeping gene, GAPDH,
following densitometric analysis of PCR products. The data are
presented in graphical form in Fig. 2. As
seen with the first set of specimens, the level of expression of the
ARG gene varied when plotted against the level of GAPDH mRNA, which
remained relatively constant (Fig. 2A); the line of best fit through
the data points exhibited a shallow slope, with a numerical value of
0.18. A similar plot was obtained for HB-EGF vs. GAPDH (data
not shown). In marked contrast, however, the expression patterns for
ARG and HB-EGF varied in a highly coordinate manner; when the band
densities for ARG and HB-EGF were plotted against each other, the line
of best fit displayed a much steeper slope, with a numerical value of
0.8 and a high degree of correlation (r = 0.930; P
< 0.0001). When plotted separately, both BPH and PCa specimens
displayed similarly steep positive slopes, of numerical values 0.786
(r = 0.935) and 0.790 (r = 0.925), respectively (Fig. 2B). No
significant difference was detected between the BPH and PCa specimens
with regard to the level or range of ARG and HB-EGF band
intensities.
|
, light
positive staining for ARG was evident in epithelial cells comprising
the prostatic ducts, with more intense staining in the adjacent stroma
(Fig. 3D). Significantly, stromal staining for ARG was localized
primarily to the interstitial and vascular smooth muscle cells and
displayed a pattern similar to that previously observed by us for
HB-EGF (29). Consistent with previous reports (32, 36, 37, 38, 39), ErbB1
staining was localized to the basal epithelial cells of normal
prostatic ducts (Fig. 3B), with less intense staining for ErbB1 evident
in the pseudoacini characteristic of prostate carcinoma. Little or no
stromal staining for ErbB1 was demonstrated, also consistent with
previous observations.
|
) and Western blot (data not shown).
ErbB2 and ErbB3 mRNAs were also detected in these cells, suggesting the
possibility for receptor coactivation following ligand binding to
ErbB1. Mitogenic activity of ARG and HB-EGF was assessed using PrEC as
a target cell. Both growth factors demonstrated similar potency as PrEC
mitogens (Fig. 4B). This result was somewhat surprising because ARG has
been described in the literature as being significantly less potent as
an epithelial or fibroblast cell mitogen in comparison to EGF (40) or
to HB-EGF (41). To verify this, we examined the relative potency of
EGF, ARG, and HB-EGF in a standard thymidine incorporation assay using
Balb/c 3T3 fibroblasts as the target cell, as described in
Materials and Methods. As shown in Fig. 4C, whereas HB-EGF
was found to be as potent as EGF in promoting DNA synthesis in Balb/c
3T3 cells, ARG was 100-fold less active, consistent with its activity
described previously (40, 41).
|
). To determine whether
independent evidence for coordinate regulation of these growth factors
might be obtained, we analyzed the expression of ARG and HB-EGF mRNAs
following treatment of primary cultures of PrSMC with FGF-2 or
TGF-ß1, as described in Materials and Methods. Both FGF-2
and TGF-ß1 are known to be expressed in the prostatic stroma and
altered expression of these molecules has been implicated in benign and
malignant prostate pathologies (42, 43). Expression levels for ARG and
HB-EGF mRNAs were normalized to that of GAPDH following densitometric
analysis of autoradiographs. Figure 5, B and C, illustrates that the
expression of both ARG and HB-EGF could be up-regulated in response to
treatment with FGF-2 and/or TGF-ß1. In response to treatment with
FGF-2, ARG expression increased up to 3.5-fold, with peak expression
observed at 8 h following treatment; in contrast, no difference in
ARG expression in response to TGF-ß1 stimulation, was observed
relative to control. HB-EGF expression in PrSMC was found to increase
up to 2.7-fold and 2.8-fold in response to treatment with FGF-2 and
TGF-ß1, respectively; peak expression of HB-EGF mRNA in response to
FGF-2 occurred at 24 h following treatment, whereas HB-EGF mRNA
expression in response to TGF-ß1 stimulation was maximal at 4–8 h
and decreased thereafter.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
epithelial
paracrine mechanism has been hypothesized to exist within the prostate
by Cunha and co-workers (22, 23, 24). These investigators have suggested
that SMC-epithelial signaling may be the predominant means of
stromal-epithelial interaction in the fully differentiated organ. This
hypothesis was based on the anatomical relationship between SMC and the
ductal network in rodent prostate, on results from other organ systems,
such as the uterus, and on experimental models in which urogenital
mesenchyme and epithelia (typically from rodents) are recombined and
their differentiation state evaluated after engraftment under the renal
capsule. However, before this report, evidence that this paracrine
system actually exists in the adult human prostate was indirect and
limited, although several candidate growth factors have been proposed
as potential mediators of such an interaction. We now present a variety of independent lines of evidence in support of the existence of this paracrine signaling mechanism in humans. Using immunohistochemistry, we have identified the SMC compartment of the human prostate as a site of expression of amphiregulin, an activating ligand for ErbB1. A series of previous reports have localized ErbB1 exclusively to normal and malignant prostatic epithelial cells in the sexually mature organ (32, 36, 37, 38, 39). Basal epithelial cells appear to be the major sites of ErbB1 synthesis in normal epithelial ducts, based on immunohistochemical and in situ hybridization analyses (32, 44). Our own evaluation of ErbB1 immunolocalization in this study is consistent with the conclusion that ErbB1 expression is restricted to the epithelial tissue compartments. This cellular location indicates that ErbB1 is present at a privileged site, i.e. in the basal layer of the ductal epithelium, for regulatory control by one or more of the receptor’s cognate ligands were they to originate from the stroma. Our finding that ARG is expressed by prostatic SMC in vivo is consistent with an earlier report from our laboratory in which we demonstrated that a growth factor that is structurally related to ARG, HB-EGF, is also synthesized predominantly by SMC in the prostate (29). Therefore, ARG and HB-EGF expression in vivo is likely to result in diffusion of the soluble forms of the growth factors from the stromal to the epithelial compartments.
Using in vitro clonal growth assays, we demonstrated that both ARG and HB-EGF are potent mitogens for normal human prostate epithelial cells, consistent with a potential paracrine role for both growth factors. Notably, the epithelial cells we used for the mitogenic assays express what appears to be a basal cell phenotype, based on their expression of CD44 and high molecular weight cytokeratins (45, 46, 47). A basal cell phenotype has also been ascribed to cultured prostatic epithelial cells by others (48, 49, 50). Therefore, although the limitations of in vitro models should be considered, this experiment may be physiologically relevant, in that normal basal PrEC, which express ErbB1 in vivo, are shown to be highly responsive to both mitogens. ARG has been described in the literature as a growth factor typically exhibiting one or two orders of magnitude lower potency than either EGF or HB-EGF (40, 41). This observation was confirmed by us using a standard Balb/c 3T3 3H thymidine incorporation assay, in which we showed ARG to be significantly less mitogenic, on a molar basis, in comparison to HB-EGF and EGF. In light of these findings, the equivalent molar potency of ARG and HB-EGF in the PrEC clonal growth assay was unexpected but may identify prostate epithelial cells as a specific physiologic target for the actions of ARG.
In an analysis of a large series of benign and malignant human prostate
tissues, HB-EGF and ARG mRNAs were found to exhibit highly coordinate
expression, a result that is consistent with the conclusion that they
are synthesized predominantly within the same tissue compartments.
Moreover, we also observed that both HB-EGF and ARG mRNAs could be
induced in human prostate SMC by FGF-2 (basic FGF), a growth factor
synthesized within the stromal compartment that has been linked to
pathologic cell growth in the prostate (42). FGF-2 has been shown
previously to be a human prostate SMC mitogen (51). HB-EGF mRNA
expression was also induced by TGF-ß1, which has been linked to both
prostate cancer and BPH (43). These data indicate that the HB-EGF and
ARG genes are under partial coordinate regulation in prostate SMC, a
finding that is also likely to account, at least in part, for the high
degree of coordinate expression we observed in vivo.
Consistent with this interpretation, coordinate expression of ARG and
HB-EGF has been demonstrated previously in several published studies
employing cultured cells (52, 53, 54). EGF mRNA levels were increased in
four of five prostate carcinoma specimens in comparison to matched
normal tissue from the same patient, and EGF mRNA levels were
positively correlated with mRNA for BTC. EGF is principally a secretory
product of epithelial cells in the prostate (35). ARG and HB-EGF mRNA
levels, in contrast, did not show a disease-specific association and
were observed only to correlate with each other. These data suggest a
distinct mechanism of regulation and/or cellular localization for ARG
and HB-EGF in comparison to other members of the ligand family. The
most recently identified ErbB1 ligand, epiregulin, was not detected in
the prostate, consistent with previous reports of a restricted pattern
of expression in placenta, uterus, peripheral blood cells and during
early development (33, 34). ErbB2 and ErbB3 receptor mRNAs, and the
ErbB3/ErbB4 ligand, HRG-, were detectable, but their expression
pattern did not resemble that seen for HB-EGF and ARG. ErbB4 mRNAs were
faint or undetectable, consistent with previous reports that this
receptor is not expressed in the prostate (55).
Early studies on the role of ErbB1 in paracrine and autocrine growth
regulation in the prostate assessed the possible function of the ErbB1
activating ligand, transforming growth factor- (TGF-). TGF- is
known to be expressed in the prostate (28) and potential sites of
TGF- synthesis have been identified in the epithelial and stromal
compartments by in situ hybridization (44, 56) and
immunohistochemical analyses (32, 36, 38). In the normal or benign
adult prostatic tissues, TGF- protein expression has typically been
observed to increase in prostate carcinoma cells in comparison to
normal epithelial cells (32, 37, 39), reminiscent of the expression
pattern in early development (32). Co-expression of TGF- and ErbB1
within epithelial cells has been proposed to enable the "switch"
from paracrine to autocrine ErbB1 activation and to promote the
uncontrolled carcinoma cell proliferation characteristic of tumor
progression (32, 39). From our mRNA expression analysis, TGF- mRNA
was present in normal and tumor tissue, however, it was not
coordinately expressed with HB-EGF and ARG, suggesting that these
molecules are subject to different regulatory mechanisms in
vivo. Furthermore, because TGF- has been identified in the
prostatic epithelium, as well as the stroma, our data suggest that this
molecule acts primarily as an autocrine growth factor in the prostate
and less so as a mediator of stromal-epithelial interactions.
Taken together, we conclude from these observations that ARG and HB-EGF
act as physiologic SMC-derived paracrine regulators of prostatic
epithelium and may play a similar functional role in vivo.
The significance of the apparent coordinate regulation of these factors
in the prostate is unknown; however, from their defined biological
activities, ARG and HB-EGF may function as mediators of epithelial cell
growth, differentiation or survival. In addition, their roles in
regulation of cancer cell behavior may differ from those performed in
normal tissue. Lin et al. (57) recently identified several
ErbB1 ligands, including HB-EGF, as cell survival factors, distinct
from their role as mitogens, for human prostate carcinoma cells. These
investigators also observed that the cell survival pathways operating
in prostate cancer cells were distinct from those in normal PrECs.
HB-EGF and ARG have a similar domain structure and exhibit similar
biochemical affinities for immobilized heparin. In addition, their
membrane-anchored, precursor forms have been shown to interact
functionally with CD9 (58), a membrane protein that belongs to the
tetraspanin protein family. CD9 is capable of altering the juxtacrine
activities of membrane HB-EGF and ARG. Interestingly, however, the
membrane form of TGF- appears not to interact with CD9, suggesting
that HB-EGF and ARG may play similar functional roles distinct from
those performed by TGF- (58). The localization of ARG and HB-EGF
synthesis in prostate SMC is also consistent with a role for these
growth factors as "andromedins," hypothetical peptide mediators of
the androgenic signaling characteristic of the prostate gland. ARG
synthesis has been reported previously to be under the control of
androgens in the anaplastic human prostate cell line, LNCaP (59).
Notably, Prins et al. (60) found that although stromal
fibroblasts and basal epithelial cells in the prostate were generally
found to lack the androgen receptor (AR), strong AR staining was
evident in SMC, supporting the proposal that AR-positive smooth muscle
cells could mediate stromal-epithelial interactions. Whether the AR
lies upstream of ARG and HB-EGF synthesis in vivo remains to
be examined; however, it is conceivable that these growth factors play
a role in the maintenance of the functional integrity of the gland in
the normal physiologic state. Recently, Levine et al. (61)
reported that androgen stimulated increased gene expression and
synthesis of VEGF by isolated prostatic stromal cells, consistent with
the hypothesis that androgen-dependent growth factor signaling can be
stroma-mediated.
A number of other growth factors have been localized to the prostate and have been suggested as mediators of stromal-epithelial interactions, including members of the fibroblast growth factor (FGF)- (42, 62), TGF-ß- (43, 63) and insulin-like growth factor (IGF)-families (64, 65). Of these, only the fibroblast growth factor, FGF-7/keratinocyte growth factor (KGF) has emerged as an unambiguous candidate, based on localization of the mRNA and protein, the expression pattern for the receptor and regulation by androgen. KGF mRNA has been localized to stromal cells by in situ hybridization analysis in the human (66, 67) and rat (68, 69) prostate, whereas the high-affinity KGF receptor is localized exclusively to prostatic epithelial cells (66, 67, 68, 69), thus presenting the opportunity for directional stromal-epithelial paracrine interactions. Furthermore, KGF expression has been shown to be androgen-regulated in isolated prostatic stromal and epithelial cells in co-culture (70). Also KGF was found to partially substitute for androgen in branching morphogenesis of the rat ventral prostate (68) and in development of the rodent seminal vesicle (71), strongly implicating KGF action in the androgen-stimulated phenotype and behavior. However, more recent evidence suggests that KGF expression in the rat prostate is unresponsive to changing androgen levels (68, 72) in vivo. This suggests further studies are required to confirm the identity of KGF as a true andromedin.
In summary, we have presented the first evidence that two structurally similar ligands for the ErbB1 receptor tyrosine kinase, ARG and HB-EGF, are directional mediators of ErbB1-dependent signaling between the SMC and epithelial compartments of the human prostate and may be regulated in a coordinate manner in vivo. These findings are potentially relevant to the normal function of the prostate gland as a secretory organ as well as to several prostate diseases, including prostate cancer and benign prostatic hyperplasia (BPH), a condition of aberrant prostatic enlargement and urethral obstruction that afflicts most men with age. Our results provide further evidence for the critical importance of the stroma in regulating epithelial cell growth and behavior.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received April 19, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and the EGFR in human
prostate tissues. J Urol 152:2120–2124[Medline]
in the progression of
prostatic neoplasms. Clin Cancer Res 1:545–550[Abstract]
in human hyperplastic prostate tissue: expression and cellular
localization. J Clin Endocrinol Metab 81:4148–4154This article has been cited by other articles:
 |
|
 |
|
  N. Ali and V. Knauper Phorbol Ester-induced Shedding of the Prostate Cancer Marker Transmembrane Protein with Epidermal Growth Factor and Two Follistatin Motifs 2 Is Mediated by the Disintegrin and Metalloproteinase-17 J. Biol. Chem., December 28, 2007; 282(52): 37378 - 37388. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Heldring, A. Pike, S. Andersson, J. Matthews, G. Cheng, J. Hartman, M. Tujague, A. Strom, E. Treuter, M. Warner, et al. Estrogen Receptors: How Do They Signal and What Are Their Targets Physiol Rev, July 1, 2007; 87(3): 905 - 931. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Q. Yu, E. M. C. Chow, H. Wong, J. Gu, O. Mandelboim, S. D. Gray-Owen, and M. A. Ostrowski CEACAM1 (CD66a) Promotes Human Monocyte Survival via a Phosphatidylinositol 3-Kinase- and AKT-dependent Pathway J. Biol. Chem., December 22, 2006; 281(51): 39179 - 39193. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Bavik, I. Coleman, J. P. Dean, B. Knudsen, S. Plymate, and P. S. Nelson The Gene Expression Program of Prostate Fibroblast Senescence Modulates Neoplastic Epithelial Cell Proliferation through Paracrine Mechanisms Cancer Res., January 15, 2006; 66(2): 794 - 802. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Tanaka, S. Miyamoto, S. O. Suzuki, E. Oki, H. Yagi, K. Sonoda, A. Yamazaki, H. Mizushima, Y. Maehara, E. Mekada, et al. Clinical Significance of Heparin-Binding Epidermal Growth Factor-Like Growth Factor and A Disintegrin and Metalloprotease 17 Expression in Human Ovarian Cancer Clin. Cancer Res., July 1, 2005; 11(13): 4783 - 4792. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D M Peehl Primary cell cultures as models of prostate cancer development Endocr. Relat. Cancer, March 1, 2005; 12(1): 19 - 47. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Miyamoto, M. Hirata, A. Yamazaki, T. Kageyama, H. Hasuwa, H. Mizushima, Y. Tanaka, H. Yagi, K. Sonoda, M. Kai, et al. Heparin-Binding EGF-Like Growth Factor Is a Promising Target for Ovarian Cancer Therapy Cancer Res., August 15, 2004; 64(16): 5720 - 5727. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. D. Abbott, T.-M. Lin, N. T. Rasmussen, R. M. Albrecht, J. E. Schmid, and R. E. Peterson Lack of Expression of EGF and TGF-{alpha} in the Fetal Mouse Alters Formation of Prostatic Epithelial Buds and Influences the Response to TCDD Toxicol. Sci., December 1, 2003; 76(2): 427 - 436. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. M. Adam, J. Kim, J. Lin, A. Orsola, L. Zhuang, D. C. Rice, and M. R. Freeman* Heparin-Binding Epidermal Growth Factor-Like Growth Factor Stimulates Androgen-Independent Prostate Tumor Growth and Antagonizes Androgen Receptor Function Endocrinology, December 1, 2002; 143(12): 4599 - 4608. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Pan, K. Sengoku, K. Goishi, N. Takuma, T. Yamashita, K. Wada, and M. Ishikawa The soluble and membrane-anchored forms of heparin-binding epidermal growth factor-like growth factor appear to play opposing roles in the survival and apoptosis of human luteinized granulosa cells Mol. Hum. Reprod., August 1, 2002; 8(8): 734 - 741. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Kim, R. M. Adam, and M. R. Freeman Activation of the Erk Mitogen-activated Protein Kinase Pathway Stimulates Neuroendocrine Differentiation in LNCaP Cells Independently of Cell Cycle Withdrawal and STAT3 Phosphorylation Cancer Res., March 1, 2002; 62(5): 1549 - 1554. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. W. Davis II, F. T. Lauer, A. D. Burdick, L. G. Hudson, and S. W. Burchiel Prevention of Apoptosis by 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) in the MCF-10A Cell Line: Correlation with Increased Transforming GrowthFactor {{alpha}} Production Cancer Res., April 1, 2001; 61(8): 3314 - 3320. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
), HB-EGF(
) or ARG (•). Data points represent the
mean of duplicate determinations and the data are representative of at
least three experiments.
4.