| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Review |
1
Oncology Research Centre, Prince of Wales Hospital, High Street, Randwick, New South Wales, Australia, 2031 and Division of Medicine, University of New South Wales, Kensington, New South Wales 2052, Australia.
2
Department of Urology, St. Vincent's Hospital, 438
Victoria St., Darlinghurst, New South Wales 2010, Australia.
a Address correspondence to this author at: Oncology Research Centre, Villa 1, Prince of Wales Hospital, High Street, Randwick, New South Wales 2031, Australia. Fax 61 2 9382 2629; e-mail P.Russell{at}usnw.edu.au.
 |
Abstract |
|---|
 Top Abstract The Natural History of...Biological Properties of the...The AR in Prostate...The Role of Growth...Models for Studies of...Growth Factor Families Important...Concluding RemarksReferences |
|---|
, retinoic acid, vitamin D3, and the
transforming growth factor ß families. We review their role in normal
prostate development and in cancer progression, using evidence from
clinical specimens and models of PC cell growth. |
The Natural History of Prostate Cancer |
|---|
|
TopAbstractThe Natural History of...Biological Properties of the...The AR in Prostate...The Role of Growth...Models for Studies of...Growth Factor Families Important...Concluding RemarksReferences |
|---|
Treatment for clinically detected prostate cancer includes watchful waiting, prostate surgery, or targeted irradiation. The latter two treatments can cure only cancer that is organ confined, yet up to 60% of patients present with metastases, particularly to the bone. Because determining the biological potential of localized cancers detected with any degree of certainty through screening is not possible, observation alone will result in a lost opportunity for cure in many patients; on the other hand, treatments such as radical prostatectomy or radiation therapy in such cases will necessarily overtreat many patients (8). Advanced prostate cancer with metastases to bone and other soft tissues also presents a difficult therapeutic problem (4). Management alternatives for advanced prostate cancer are palliative at best. Treatment for such cancers depends on the androgen requirement of prostate epithelial cells for growth and survival. Endocrine therapy leads to substantial periods of remission. However, it is ineffective once the tumors progress from androgen-dependent (AD), through androgen-sensitive (AS; these tumors do not require androgen for growth but proliferate more rapidly in its presence), to hormone refractory or androgen-independent (AI). This progression is almost inevitable in men after androgen ablation therapy. Total androgen ablation used in hormonally naive patients results in a median remission of about 24 months (4). Therapeutic alternatives in men who have progression of the disease after androgen ablation are very limited, with a median survival between 8 and 12 months.
The need to develop more effective therapy for these patients exists. A better understanding of growth factor pathways may provide an additional target for therapy in patients with advanced prostate cancer. This review focuses on the current state of knowledge of growth factor pathways in prostate cancer. The majority of data presented were derived from cancer cell lines without functional AR and from animal models. We thus warn the reader that such data may not be directly relevant to human prostate cancer. In addition, some of the data are based on single literature reports, and they need confirmation before definitive conclusions can be drawn.
|
Biological Properties of the Prostate Gland |
|---|
|
TopAbstractThe Natural History of...Biological Properties of the...The AR in Prostate...The Role of Growth...Models for Studies of...Growth Factor Families Important...Concluding RemarksReferences |
|---|
Various hypotheses to explain how malignant but not benign prostatic epithelial cells become AI have been proposed [reviewed in(13)]. These hypotheses are based on: (a) clonal selection of AI cells from a heterogeneous population of AD and AI cells by androgen ablation; (b) an adaptive theory, which proposes the presence of AD stem cells that can adapt to self-renewal in the absence of androgen; (c) the potential effects of small residual amounts of androgen, which can stimulate AS cells; and (d) changes in the AR, such as mutation, overexpression, or loss, that occur in some but not all cases of AI prostate cancer. Currently, determining which hypothesis is correct is not possible, but given that most prostate cancers are heterogeneous, it is likely that both clonal selection of preexisting AI cells and adaptive processes may contribute to AI progression. However, the findings described below make it clear that AI progression must also be mediated by growth-regulatory factors that function independently of androgen.
Both autocrine and paracrine growth factors are up-regulated in AI prostate tumors and may replace androgen as the primary growth-stimulatory factors in cancer progression. This up-regulation may represent an adaptive response to androgen ablation in the growth regulation of AI tumors and is the basis for discussion in this review.
|
The AR in Prostate Cancer |
|---|
|
TopAbstractThe Natural History of...Biological Properties of the...The AR in Prostate...The Role of Growth...Models for Studies of...Growth Factor Families Important...Concluding RemarksReferences |
|---|
The AR is a member of the steroid/thyroid hormone/retinoic acid family of receptors that bind to specific hormone-responsive elements in target genes, thus regulating transcription [reviewed in (15)]. In the developing prostate, only the stromal cells express ARs, suggesting that androgen regulation of prostatic development is mediated via the stromal cells [reviewed in (16)(17)]. In the normal adult prostate, AR expression is found mainly in the epithelium, but it is also identified in the stromal cells (16)(17). In contrast, in prostate cancer specimens, AR staining in the epithelium is heterogeneous, with a marked decrease in AR-positive cells occurring in less differentiated tumors, corresponding to the reported insensitivity to androgen observed in advanced prostate cancers (18)(19). The reason for this altered expression of the AR and why some cells lose their AR is still not known.
Sequence analysis of the AR gene in prostate cancers has revealed that many tumors contain mutations (20)(21)(22)(23)(24). The resulting mutant proteins may be unable to bind androgen but remain constitutively activated, or they may bind androgen but be nonfunctional (15)(25). AR mutations in prostate cancers are commonly point mutations but may also reflect microsatellite instability. Point mutations have been identified in many prostate cancers, including the AS cell line LNCaP (20)(21)(22). In particular, a hot spot for mutation exists at codon 887 (ACT-GCT, Thr-Ala) in the hormone-binding domain of the gene. In LNCaP cells, this mutation enables the AR to bind progestogenic and estrogenic steroids, but it also causes a decreased affinity for androgen (23). Such abnormal binding may induce the transcription of genes out of context, thus upsetting the delicate balance of growth factors in the prostate and promoting cancer development.
Prostate cancer development has also been linked to a change in the number of repeats in the polymorphic CAG and GGC microsatellite in exon 1 of the AR gene (24)(25)(26). How this leads to prostate cancer development is not yet clear.
Prostate cancer cells have derived complex mechanisms that allow them to grow in the absence of androgen stimulation or in the presence of mutated or lost AR. Exposure of DU-145 AI prostate cancer cells transfected with an AR expression vector and a chloramphenicol acetyltransferase (CAT) reporter gene to keratinocyte growth factor (KGF), epidermal growth factor (EGF), or insulin-like growth factor 1 (IGF-1) could activate CAT gene expression in the absence of androgen (27). These data suggest that growth factors may activate AR in an androgen-deprived environment.
|
The Role of Growth Modulators in Prostate Cancer |
|---|
|
TopAbstractThe Natural History of...Biological Properties of the...The AR in Prostate...The Role of Growth...Models for Studies of...Growth Factor Families Important...Concluding RemarksReferences |
|---|
) (28)(29)(30)(31)(32)(33)(34)(35)(36)(37)(38)(39)(40)(41)(42)(43)(44)(45)(46)(47)(48)(49)(50)(51)(52)(53)(54)(55)(56)(57)(58)(59)(60)(61)(62)(63)(64)(65)(66)(67)(68)(69)(70)(71)(72)(73)(74)(75)(76)(77)(78)(79). This renders the prostate gland very
sensitive to any aberrations in growth factor or androgen expression.
Therefore, it is not surprising that in prostate cancer the expression,
regulation, and cell-related production of many of these growth
modulators are altered (Table 1
).
|
Alterations may take the form of up- or down-regulation of growth
factors or their receptors or a change from paracrine to autocrine
mediation of growth factor pathways. Alternatively, because many growth
factor pathways send their messages to the cells via common signal
transduction pathways, any mutations affecting signal transduction may
affect several growth factor pathways simultaneously. Functional
analysis of these potential changes is outside the scope of this
review. Our focus is on growth factor families for which there is
evidence to support a major role in normal and cancerous growth of the
prostate. These families include the following: the fibroblast growth
factor (FGF) family, the IGF family, EGF, and transforming growth
factor (TGF-), all of which are predominantly stimulators of
proliferation; retinoic acid, which causes differentiation and
invasiveness; and the TGF-ß family and vitamin D3,
which are predominantly inhibitors of prostatic growth. Using evidence
from human prostate cancer cell lines and animal models in vivo and in
vitro, we will discuss the role that these factors play in normal
prostate development and in cancer progression. Evidence obtained from
clinical findings is often controversial, and many different models of
prostate cancer have been studied in vivo and in vitro in an attempt to
define more precisely the roles particular growth factors play in
prostate cancer progression.
|
Models for Studies of Prostate Cancer |
|---|
|
TopAbstractThe Natural History of...Biological Properties of the...The AR in Prostate...The Role of Growth...Models for Studies of...Growth Factor Families Important...Concluding RemarksReferences |
|---|
(81)(82)(83)(84)(85)(86)(87)(88)(89)(90)(91)(92)(93)(94)(95)(96)(97)(98)(99)(100)(101)(102)(103)(104)(105)(106)(107)(108)(109)(110)(111).
|
rodent models
The rat prostate differs from the human gland in that it is
divided into distinct individually encapsulated lobes (dorsal, ventral,
and anterior), each of which has a separate set of compound ducts
(111). The best known model of prostate cancer is the
Dunning R-3327 rat prostatic adenocarcinoma model, derived by the
passage of a spontaneous prostate tumor discovered at autopsy in a
Copenhagen rat (81). Subcutaneously implanted tumors become
palpable in ~60 days and histologically are well-differentiated
adenocarcinomas with both glandular and stromal elements. Multiple
sublines indicative of cancer progression have been developed
[described in detail in (82)], including AS lines
(H) and AI tumors (A subline), which lack 5-reductase and
ARs (83), and metastatic (84) sublines (Table 1
).
Other rodent models include the hormone- or N-methyl-N-nitrosourea-induced Noble rat prostatic adenocarcinoma model (85)(86), Pollard rat tumors in Lobund-Wistar (L-W) rats derived from a spontaneous tumor (90)(91), spontaneous ACI rat prostate cancers in aging August x Copenhagen hybrids (93), and the Shionogi mouse mammary carcinoma model (95)(96), which produces both AD and AS wild-type tumors. In Shionogi mice, the proportion of AI stem cells in recurrent AI tumors has been shown to increase 500-fold, lending support for self-renewal of stem cells in the absence of androgen (113). These models have been described elsewhere in detail (13).
Thompson et al. (97) have developed a mouse prostate reconstitution (MPR) model that exploits the ability of fetal epithelial and stromal cells from the urogenital sinus to form a mature prostate gland when implanted under the renal capsule of adult mice. This model allows the study of paracrine interactions between cell compartments by introducing candidate genes for growth factors, oncogenes, or suppressor genes into the epithelial or fibroblast compartments derived from the fetal prostate gland and then combining and engrafting them under the renal capsule of syngeneic male mice.
human prostate cancer cell lines
Most human prostate cancer cell lines have been established from
metastatic deposits, with the exception of PC-93 (98), grown
from an AD primary tumor. However, PC-93 and other widely used lines,
including PC-3 (99), DU-145 (100), and TSU-PR1
(114), are all AI; all lack ARs (with the possible exception
of PC-93), PSA, and 5-reductase; and all produce poorly
differentiated tumors if inoculated into nude mice. The paucity of cell
lines that are AD has made studies of the progression of prostate
cancer using human material very difficult. However, metastatic
sublines of PC-3 have been developed by injection of cells into nude
mice via different routes, especially orthotopically (115).
The LNCaP cell line, established from a metastatic deposit in lymph node (100), is the only human prostate cancer cell line that demonstrates androgen sensitivity but not androgen dependence. After its initial characterization (100), several laboratories found that this line was poorly tumorigenic in nude mice unless coinoculated with tissue-specific mesenchymal or stromal cells (116) or Matrigel (117), suggesting that extracellular matrix and paracrine-mediated growth factors play a role in prostate cancer growth and site-specific metastasis (118). LNCaP cells grown in castrated mice that had progressed to the AI state were cultured to obtain new cell lines. The C-4 LNCaP (119) line produces PSA and a factor that stimulates PSA production, and the C4-2 line metastasizes to lymph nodes and bone after subcutaneous or orthotopic inoculation (102)(103). Another subline of LNCaP, LNCaP 104-R2, cultured in androgen-depleted medium for >100 passages, is stimulated by finasteride, causing some concern over the use of antiandrogens for the treatment of late-stage prostate cancer (104).
human xenograft models
The CWR22 xenograft line is highly AD in vivo and relapses to an
AI line, CWR22R, after androgen withdrawal (105), thus
providing a useful model for studies of the progression of human
prostate cancer. PC-82 is one of several xenograft lines established in
Rotterdam. PC-82 and PC-EW are AD prostate cancer xenograft lines
(106)(107) that are useful for studying AR
regulation (119). Honda and LuCap xenografts are also both
AS (108)(109). The UCRU-PR-2 xenograft
line, established from a patient with prostatic adenocarcinoma, is a
small cell carcinoma of the prostate that secretes
pro-opiomelano-corticotropin-derived peptides
(110)(111). Cell lines have not been established
in vitro from these lines.
transgenic mouse models
Fusion of tissue-specific promoter elements to oncogenes has been
used to target expression of the oncoprotein to a given organ,
sometimes with the development of cancer in that organ. The use of a
viral promoter with the oncogene INT2 causes benign
hyperplasia of the prostate in transgenic mice (49), whereas
TGF- expressed under the control of a metallothionine gene promoter
produced prostate epithelial hyperplasia and focal dysplasia resembling
carcinoma in situ (121). Regulatory elements of the rat
probasin gene have been shown to target hormonally regulated
expression of heterologous genes in the prostates of transgenic mice
(122). Two new transgenic models have excellent potential
for studies of prostate cancer progression. These are C3(1)/SV40 large
T antigen transgenic mice, which show a progression to cancer from
intraepithelial neoplasia (123), and the TRAMP (transgenic
adenocarcinoma mouse prostate) model, in which metastases develop
(124).
|
Growth Factor Families Important in Prostate Cancer |
|---|
|
TopAbstractThe Natural History of...Biological Properties of the...The AR in Prostate...The Role of Growth...Models for Studies of...Growth Factor Families Important...Concluding RemarksReferences |
|---|
. Some of these families appear to play a more
major role in prostate cancer progression than others. Their
characteristics are described in Table 3
(51, 125–144).
|
the TGF-ß FAMILY
Three proteins of the TGF-ß superfamily, TGF-ß1, TGF-ß2, and
TGF-ß3, are expressed during prostate development and in the adult
prostate in both normal and malignant tissue
(125)(126). More distantly related
peptides, including the activins and inhibins, are not discussed here.
The TGF-ß family is highly conserved between species
(145). TGF-ß1 predominates in all tissues, whereas the
expression of TGF-ß2 and TGF-ß3 is more tissue-restricted
(126)(127)(146). However, all three
isoforms share a multiplicity of biological effects (126).
TGF-ßs induce angiogenesis in wound healing (41);
stimulate the synthesis of extracellular matrix components such as
collagen, fibronectin, proteoglycans, and integrins; and also can
inhibit extracellular matrix formation through down-regulation of a
wide variety of proteases (42)(145). TGF-ßs also can
induce proliferation of mesenchyme cells and can act as growth
inhibitors of epithelial cells (147) and as important
immunoregulatory molecules (43).
How TGF-ß1 works on prostrate cancer cells is unclear. TGF-ß1 binds
to the TGF-ß2 receptor (TbetaR-II), which recruits the TGF-ßI
receptor (TbetaR-I) to initiate a signal transduction cascade. Although
TbetaR-I and TbetaR-II are transmembrane serine-threonine kinases
(129), they can trigger decreases in the expression of
members of the Src family of tyrosine kinases (130),
affecting protein tyrosine kinase signaling and hence growth regulation
(131). TGF-ß3 receptor (TbetaR-III) plays a more indirect
role because it delivers ligands to the signaling receptors
(132). TGF-ß1 prevents phosphorylation of the
retinoblastoma gene product (149) that is involved in
cellular proliferation and should, therefore, inhibit proliferation.
Thus, in the nondiseased prostate, TGF-ß is believed to play a role
in regulating cell growth through its antiproliferative effects because
it can inhibit the mitogenic effects of EGF/TGF- on epithelial cells
(28) and of basic FGF (bFGF) on stromal cells
(29).
TGF-ß1, TGF-ß2, and TGF-ß3 are important for fetal prostate development and are expressed at high concentration in 17-day murine urogenital sinus mesenchyme but not in the epithelium. In adult mice, only TGF-ß1 is increased, with the highest concentrations observed during epithelial duct formation (30). The expression of TGF-ß1 and its receptor are negatively regulated by androgen in the prostate. In apparently healthy rats, expression of TGF-ß and TbetaR-I and TbetaR-II is up-regulated within 24 h after castration and has been linked to programmed cell death in the prostate (31). In tumors, androgen withdrawal results in increased TGF-ß1 and receptor concentrations in both rat Dunning R3327 PAP (31) and human PC-82 prostatic cancer cells (32). Increasing expression of TGF-ß1 appears to be important in prostate cancer progression, but its exact role remains uncertain (125). In humans, increased mRNA and protein expression of both epithelial cell-specific and, to a lesser extent, stromal cell expression of intracellular TGF-ß1 is associated with prostate cancer progression (33)(34). Moreover, serum TGF-ß1 concentrations in patients with lymph node and/or distant metastases were markedly higher than in patients with localized disease but did not differ substantially among localized cancers as to tumor extension (35). Similarly, increased TGF-ß1 expression is associated with increasing malignancy in the mouse MPR model (36) and, in particular, in metastatic versus primary cell lines (37). Moreover, transfection of TGF-ß1 into rat R3327-MATLyLu prostate cancer cells resulted in larger, less necrotic, and more metastatic cells than controls (38).
Not only is expression of TGF-ß1 increased with prostate cancer progression, but secretion also occurs. A factor involved in TGF-ß1 secretion (150)(151), the latent TGF-ß1 high molecular weight complex, is associated with a latent binding protein (150) that is not expressed in prostate cancer (150). This suggests the possibility that progression of this disease is associated with TGF-ß1 switching from an autocrine/paracrine to a juxtacrine mode of action. This is reflected by secretion of TGF-ß1 by the human AI cell lines DU-145 and PC-3 but not by the AS LNCaPs (152). Less information is available on the other TGF-ß isoforms in prostate cancer. The PC-3 cells do not respond to TGF-ß2 (153), indicating that an autocrine pathway is not present. Studies in the MPR model show that TGF-ß3 but not TGF-ß2 concentrations are increased in carcinomas (36). These observations may reflect increased TGF-ß1 concentrations given that TGF-ß1 has been shown to up-regulate the expression of TGF-ß2 and TGF-ß3 in many epithelial and stromal cell lines (154).
Studies of human prostate cancer cell lines suggest that changes in sensitivity to TGF-ß1 may play a role in prostate cancer progression, but different results are obtained in clinical samples and in rodent cells. TGF-ß1 inhibits the proliferation of AI PC-3 and DU-145 cells in a dose-dependent manner but not the proliferation of AS LNCaP cells (39). TGF-ß insensitivity in LNCaP cells has been attributed to a genetic change in their TGF-ß receptor I gene (39) and can be reversed by transfection with the wild-type type I receptor gene (39). However, in clinical samples of prostate cancer, an inverse correlation between the loss of expression of TbetaR-I and TbetaR-II and tumor grade is observed (40). This could provide a potential mechanism for prostate cancer cells to escape the growth-inhibitory effects of TGF-ß. In the rodent models, sensitivity to TGF-ß growth inhibition is lost with tumor progression. In the R3227 Dunning model, functional TbetaR-I and TbetaR-II, as well as decreasing sensitivity to TGF-ß1, are found in advanced prostate carcinoma cells (41), whereas metastatic cell lines derived from the MPR model secrete but do not respond to TGF-ß1 (37). This suggests the role of TGF-ß in the progression of prostate cancer cell lines from rodent and human cancers.
TGF-ß can modulate extracellular matrix proteins and has effects on
bone-derived cells, suggesting the possibility that it may promote the
spread of prostate cancer cells and provide a suitable milieu in bone
for metastatic growth. TGF-ß can induce type IV collagenase matrix
metalloproteinase-9 and plasminogen activator inhibitor 1 in mouse
prostate-derived cell lines (37) and procollagen I-chain
mRNA in human osteoblast-like cells (131). TGF-ß1 also
stimulates adhesion of PC-3 cells to bone matrix proteins, possibly via
the 2 ß1 integrin receptor (155) and the migration of
human osteoblasts (44).
In summary, the role of TGF-ß in the prostate is highly complex (Fig. 1
). In the nondiseased prostate, TGF-ß can counterbalance the
mitogenic effects of various growth factors, thus having a role in
growth regulation. Moreover, its expression and that of its receptors
is associated with castration-induced prostate cell apoptosis. In
cancer, TGF-ß expression is increased as prostate cancer progresses.
It can be secreted and may show autocrine rather than paracrine
regulation, although its autocrine role is as yet unexplained. The
sensitivity of prostate cancer cells from different species to respond
to TGF-ß varies, in one case, because of genetic changes in the
receptor. TGF-ß also has the capacity to modulate matrix
metalloprotease production and to stimulate adhesion of prostate cancer
cells to bone cells, providing a possible role in prostate cancer cell
metastases. Given its role in angiogenesis and as an immunoregulatory
molecule, the secretion of TGF-ß by prostate cancer cells could have
important effects on the cancer cell environment.
|
EGF AND TGF-
EGF and TGF- are two structurally and functionally related
peptides (Table 3
) that signal through the same 170-kDa EGF receptor, a
transmembrane tyrosine kinase (133)(134).
Consequently, their biological activities overlap and include roles in
embryogenesis, cell differentiation, and angiogenesis (135).
Prostatic fluids from nondiseased human prostates contain large amounts
of EGF (64), which appears to be an important regulator for
normal growth in both rat and human prostate (156). Human
prostate epithelial cells require EGF in serum-free medium for growth
in primary culture (157), and nondiseased human fetal
prostatic fibroblasts can replicate in response to this cytokine
(65). Immunohistochemical studies on nondiseased and benign
prostatic tissue have shown that TGF- expression occurs
predominantly in the stroma, whereas its receptor is expressed by
epithelial cells, suggesting a paracrine/juxtacrine mode of regulation
(72).
Prostatic cell EGF expression is regulated by androgen. Castration of
mice or rats results in a marked decrease in EGF expression
(66)(67), which can be restored by testosterone
administration. There is some evidence that TGF- may also be
androgen-related. Castration of rats, followed by androgen-induced
regrowth, results several days later in increased TGF- mRNA
(67); because of the timing, this may reflect the induction
of other regulatory factors by androgen.
Increased expression of EGF/TGF- has been linked to prostate cancer
development. A slight but important rise in EGF protein expression is
observed in the epithelial cells of prostate cancer specimens
(68)(69), and similarly, TGF- protein
concentrations are raised in human prostate cancers in comparison with
benign tissue (70)(71). In many tumors, TGF-
and epithelial growth factor receptor (EGFR) are coexpressed in the
epithelial cells, suggesting a switch from paracrine to autocrine
regulation (72).
These studies are supported by findings in vitro using human prostate
cancer cell lines. LNCaP cells are stimulated in vitro (158)
and in vivo (159) by EGF and TGF-; in LNCaP cells
undergoing AI progression, the EGFR is up-regulated (160).
Both LNCaP and DU-145 cell lines secrete EGF, with 14-fold higher
expression in the DU-145 line (161)(162).
Interestingly, although both lines express single EGF-binding sites of
similar high affinities, LNCaP cells exhibit markedly enhanced DNA
synthesis in response to exogenous EGF compared with DU-145 cells,
suggesting that locally produced EGF may be an important regulator of
cell proliferation in the AS cells compared with AI cancer cells
(163). Similarly, changes in TGF- expression are observed
in human prostate cancer cell lines. The AI lines PC-3 and DU-145
express higher concentrations of TGF- mRNA than do LNCaP cells
(164), but at the protein level, PC-3 cells express the
least TGF-, whereas DU-145 cells express the most (165).
However, PC-3 cells are more sensitive to TGF--induced proliferation
than DU-145 cells, although they express lower concentrations of EGFR
(165). The differential sensitivity of prostate cancer cells
to TGF- suggests that the autocrine loop involving TGF- in
prostate cancer cell lines may be further regulated by other unknown
factors.
The importance of these findings may relate to the ability of EGF to enhance prostate tumor cell invasion. EGF increased the invasive capacity of PC-3 cells in a Boyden chamber microinvasion assay. This was associated with increased expression of urokinase plasminogen activator, a serine protease, mRNA, and protein (73).
Because both EGF and TGF- bind to the EGFR, regulation of receptor
expression is mandatory to an understanding of prostate cancer
development and progression. In the nondiseased prostate, EGFR
expression is largely confined to the basal cell layers
(166)(167)(168), whereas in human cancer cell lines, EGFR
expression is increased with increasing malignancy. Thus, the AI DU-145
prostate cancer cells express 10-fold more EGFR than do AS LNCaP cells
(163). Results from immunohistochemistry of human prostatic
tumors are inconclusive. Some studies reported no difference in EGFR
expression between nondiseased and malignant cells (168);
others noted a decrease in EGFR expression in advanced malignancy
compared with nondiseased tissue (166)(167),
whereas still other reports described an increased EGFR mRNA and
protein expression in secretory epithelial cells that is associated
with advanced cancer (166) or with increasing tumor grade
(169) but not with patient survival (170). Such
discrepancies may relate to differences in methodology, or they may
reflect heterogeneity between prostate tumors in relation to EGFR
expression.
EGFR expression in nondiseased prostate tissues appears to be under
negative androgen regulation. Prostate biopsies from patients with
benign hyperplasia show substantially increased EGFR expression after
androgen withdrawal (171). Likewise in rats, castration
induces increased EGFR expression with a return to normal
concentrations after administration of dihydrotestosterone
(65), whereas exposure of LNCaP cells to the synthetic
androgen, R1881, up-regulates EGFR expression from 11 500 to 28 500
sites/cell (172). However, the LNCaP data may be affected by
the presence of a mutated AR in these cells (23). Taken
together, these data suggest the possibility that the regulation of
EGFR by androgen may be disrupted in prostate cancer. What impact this
has on the actions of the EGFR ligands, EGF and TGF-, remains to be
elucidated.
In summary, in the nondiseased prostate, EGF appears to be an important
regulator of growth (173), and its expression is positively
regulated by androgen, whereas that of EGFR is negatively regulated by
androgen. TGF- is predominantly expressed in a paracrine fashion by
nondiseased prostate stroma. In prostate cancer, EGFR expression is
up-regulated with progression as judged from prostate cancer cell
lines, but evidence from clinical trials leaves its role controversial.
However, up-regulation of EGF, TGF-, and EGFR suggest their
autocrine expression in advanced cancer. Increased EGF expression
appears to be associated with the invasive ability of prostate cancer
cells.
FGFS
The FGF family consists of nine structurally related
heparin-binding peptides that share 40–55% amino acid homology
(174). Four genes have been identified that encode
distinct high affinity (Kd =
10-11mol/L) receptors for FGFs, fibroblast growth factor
receptor 1 (FGFR-1), FGFR-2, FGFR-3, and FGFR-4 (175). Each
encodes transmembrane receptor tyrosine kinases, derived from
alternative mRNA splicing, that give rise to the potential of multiple
binding combinations for each FGF peptide (176). These
receptors also differ in their affinity for each member of the FGF
family. The FGFs show different cellular locations. Some forms are
secreted, and others are located in the nucleus
(176)(177). Three members of this family, bFGF
(FGF-2), acidic FGF (aFGF or FGF-1), and KGF (FGF-7) have been
implicated in prostate cancer.
bFGF
bFGF is encoded by a 36-kb single copy gene
(178)(179) with four variants from
alternative splicing (Table 3
). Differing intracellular locations
suggest different biological functions (175). Despite no
signal sequence (178), bFGF has been found cell-associated
or deposited into the basement membrane (176), with its
release possibly due to cell injury in vivo.
bFGF is synthesized by a wide variety of cell types, including epithelial cells, stromal cells (45), macrophages (180), and endothelial cells (181). Its biological functions include a role in angiogenesis, tissue development and differentiation (182)(183), and the ability to modulate neural function (184).
Prostatic stromal and epithelial cells actively synthesize bFGF (45). However, in the nondiseased prostate, only the stromal cells express the bFGF receptor and thus respond to this growth factor, suggesting that bFGF plays an important role in maintaining prostatic mesenchymal homeostasis (46). Human fetal prostatic fibroblasts have also been shown to proliferate in response to bFGF (185), which can overcome the TGF-ß1 inhibition of these cells (186). Some conflict concerning the response of the bFGF pathway to androgen exists. In 7-day-old rats, castration results in decreased bFGF expression, with increased bFGF mRNA 16 h after exposure of regressed prostate to androgen (187). In AS LNCaP cells, androgen causes an increase in bFGF expression (188), whereas other human prostatic cancer cell lines produce bFGF independently of androgen (189). Similarly, AI cells from Shionogi mice produce a bFGF-like protein (190). These data suggest that production of bFGF becomes AI as cancer progresses.
Further evidence of a role for the bFGF pathway in prostate cancer progression comes from studies in the Dunning tumor model (48). A switch in exon IIIb (which has a high affinity for KGF) of the FGFR gene to exon IIIc (with a high affinity for bFGF and aFGF) occurs in malignant epithelial cells as they become independent of their requirement for stroma (48). This suggests that prostate cancer progression may be associated with a switch from stromal autocrine bFGF regulation to epithelial autocrine regulation. It has been proposed that bFGF may enable the epithelial cell to metastasize by promoting angiogenesis in the malignant tumor. These studies are supported by findings in human prostate cancer cell lines. The AS LNCaP cell line produces very low concentrations of bFGF and FGFR (47) compared with the AI cell lines PC-3 and DU-145 (47). However, the data are confusing; although LNCaP and DU-145 cells proliferate in response to bFGF, the more malignant PC-3 cells do not (47). As with the rat model, this suggests that some form of autocrine regulation of bFGF by epithelial cells may be an important step in prostate cancer progression.
Because bFGF is angiogenic, its increased production in late stage prostate cancer may promote angiogenesis, allowing tumors to grow and metastasize (50). The acquisition of metastatic ability in prostate cancer has been correlated with increasing microvessel density (191)(192), implying that angiogenic ability is necessary for metastasis to occur.
In summary, bFGF appears to be produced in an autocrine fashion by nondiseased prostate stromal cells and is important for maintaining their homeostasis. As cancer occurs and progresses, the production of bFGF becomes independent of androgen and becomes regulated in an autocrine fashion by prostate cancer epithelial cells. The ability to produce bFGF may stimulate angiogenesis, allowing the cells to metastasize.
aFGF
aFGF (Table 3
) is important in the development of the prostate in
rats, but it has not been detected in human prostatic tissues. In 6- to
8-week-old rats, expression is confined to epithelial cells and is
increased, but it declines at 14 weeks and cannot be detected by 35
weeks (51). Epithelial and mesenchymal cells from
nondiseased rat prostate and various grades of rat prostate tumors
exhibit specific aFGF membrane receptor sites (52). In the
Dunning model, aFGF alone is expressed only by stromal cells of the
slow-growing AD R3327PAP tumor, whereas fast-growing metastatic AT-3
cells express both aFGF and bFGF (52). This suggests that,
in rat models of prostate cancer, progression is associated with a
switch in regulation of aFGF from paracrine to autocrine control.
KGF
Another member of the FGF family, KGF (140), is
important in prostate development (Table 3
). In the rodent prostate,
KGF is expressed and secreted by nondiseased stromal cells, but only
epithelial cells express the BEK/FGFR-2 receptor gene that binds KGF,
suggesting that this cytokine controls epithelial cell proliferation in
a paracrine manner (53). The BEK receptor has been shown to
require an interaction with heparan sulfate proteoglycans to facilitate
binding to its ligands (54). Serum-free organ culture of
neonatal rat ventral prostates, which express both KGF and BEK receptor
mRNA, has been developed as a model to study prostate development
(193). The addition of an anti-KGF antibody could inhibit
testosterone-induced branching in this model, and in testosterone-free
conditions, KGF could mimic the effects of the hormone. These data
suggest that, in the developing prostate, KGF may act as a paracrine
mediator of androgen action. An important and similar role for KGF has
also been implicated in the human prostate. KGF and its receptor are
expressed in the stromal and epithelial cells, respectively, and in
both fetal and adult nondiseased prostates (55). This
cytokine is also a potent mitogen for nondiseased human prostatic
epithelial cells in vitro (141)(194) and
promotes the growth of these cells under serum-free conditions
(195).
In human prostatic carcinomas, in situ hybridization studies have shown increased expression of the KGF gene and receptor in epithelial cells of high-grade carcinomas but not in benign hyperplasia of the prostate, suggesting a switch from a paracrine to an autocrine loop (55)(56).
In summary, multiple autocrine and potentially intracrine ligand-receptor loops result from alterations within the FGF-FGFR family, which may underlie the autonomy of malignant tumor cells. bFGF and KGF appear to be produced by nondiseased prostate stromal cells in an autocrine and paracrine fashion, respectively, and are important for maintaining prostate homeostasis. As cancer occurs and progresses, the production of bFGF becomes independent of androgen and becomes regulated in an autocrine fashion by prostate cancer epithelial cells. The ability to produce bFGF may stimulate angiogenesis, allowing the cells to metastasize. Although aFGF appears to have a role in the rat prostate, its importance in the human prostate has not been ratified. The production of KGF by prostate stromal cells appears to control epithelial cell proliferation in a paracrine manner, but in human prostate cancer, autocrine production of KGF accompanies the progression to AI disease.
IGFS
IGFs are polypeptide growth factors with functional homology to
insulin, but in contrast to insulin, these proteins are locally
produced by a wide variety of tissues. Regulation of their production
and function is extremely complex. There are two IGF peptides, IGF-1
and IGF-2, two cell surface receptors (the type 1 IGF receptor family
and the type 2 IGF receptor), and at least six specific high-affinity
binding proteins, IGF-BP-1 through IGF-BP-6, that regulate IGF
availability and are in turn regulated by a group of IGF-BP proteases
that cleave IGF-BPs to modulate IGF action (Table 3
).
In the nondiseased prostate, IGFs are produced only by stromal cells, but normal epithelial cells express IGF-1 receptors and secrete predominantly IGF-BP-4. However, they also secrete IGF-BP-2, IGF-BP-3, and IGF-BP-6, suggesting a paracrine mode of regulation (58)(59). Some controversy concerning the IGF loop in prostate cancer exists. Some authors (74) have shown that DU-145 cells can proliferate in response to IGF-1 but do not produce this protein, suggesting maintenance of the paracrine mode of action of IGFs in prostate cancer. However, others (196) describe autocrine production of IGF-1 in PC-3, DU-145, and a subline of LNCaP cells, all of which were reported to grow in serum-free medium, secrete IGF-1, and display constitutively autophosphorylated IGF-1 receptors. The LNCaP cell line also proliferates in response to IGF-1 but does not produce it; however, this only occurs in synergy with dihydrotestosterone (197). IGF-1 may act directly through the AR pathway (27) and in turn may be regulated through an EGF autocrine growth regulatory loop (74). IGF-2 protein secretion has been detected in all three human prostate cancer cell lines, and under serum-free conditions, as described above, each can produce IGF-1 (196). This suggests that the capacity for autocrine production of EGFs by prostatic epithelial cells exists, and that this production may play a role in prostate cancer development.
This situation is even more complicated, however, because changes in
the expression of the IGF-BPs are observed in all of these lines. For
instance, PC-3 cells express large amounts of IGF-BP-2 and IGF-BP-3 and
less IGF-BP-4, whereas LNCaP cells express only IGF-BP-2
(60). Thus, dysregulation of the IGF-BP system may be
associated with prostate cancer development (Fig. 2
). Constitutively autophosphorylated IGF-1 receptors are also
displayed by the PC-3, DU-145, and LNCaP cell lines (196),
but proliferation in response to IGF in these cell lines appears to be
regulated by the autocrine secretion of IGF-BPs (74). In
serum-free culture, DU-145 cells produce IGF-BP-1, but the addition of
an antibody that binds this protein inhibits the effects of IGF-1
(74). There is also evidence that some IGF-BPs may be under
androgen regulation. PC-3 cells stably transfected with an active AR
construct do not produce IGF-BP-3 and proliferate in response to IGF-1
or IGF-2, in contrast to untransfected cells. PSA, a serine protease
that is up-regulated by androgen, can cleave IGF-BP-3 (61),
and the nerve growth factor 
-subunit, which has high sequence
homology with PSA, also has this capacity (198). This could
release IGFs locally to stimulate prostate cancer cell growth. These
data have been interpreted to suggest that androgen may indirectly
modulate IGF-induced proliferation of prostate cancer cells by
regulating IGF-BP-3 production (62).
|
In prostate cancer patients, serum IGF-BP-2 is increased; this is related to a rise in PSA concentrations, suggesting that the prostate is the source of IGF-BP-2 production (61). Because PSA is a protease for IGF-BP-3, it likely that the rise in PSA may be related to lowered concentrations of IGF-BP-3 in prostate cancer patients (61).
A potential role for the IGFs in prostate cancer progression is in the development of bone metastasis. Both IGF-1 and IGF-2 mRNA transcripts have been detected in nondiseased human osteoblast-like cells (199) and appear to have an important role in bone formation (63). Interestingly, studies have shown that factors that decrease the activity of IGF-BP-3, such as dexamethasone, also inhibit bone formation (63), suggesting an important role for IGF-BP-3 in the formation of bone metastasis in advanced prostate cancer.
In summary, the IGFs appear to have an important role in the development of prostate cancer. This can be achieved by modulation of paracrine pathways, which occur in the nondiseased prostate, to autocrine pathways, seen in prostate cancer cell lines, and also by modulation of the concentrations of different IGF-BPs that are differentially expressed in the normal condition vs cancer. Androgen appears to regulate the expression of some of these IGF-BPs, and PSA, which is androgen-regulated, particularly can cleave IGF-BP-3. The local release of IGFs in distant tissues as a result of cleavage of IGF-BP-3 may allow prostate cancer growth as metastatic deposits.
retinoic acid (RA) AND VITAMIND3
Strong evidence suggests that dietary factors can affect the
incidence of prostate cancer (200). Two such factors
are RA (vitamin A) and 1,25-dihydroxyvitamin D (vitamin
D3). RA can induce the differentiation of a wide variety of
cells and elicit changes in growth factor protein and receptor
expression (201)(202)(203). The active form of vitamin
D3, 1,25-dihydroxyvitamin D, affects calcium and phosphate
homeostasis (204)(205) and is an important
modulator of proliferation and differentiation in both nondiseased and
malignant cells (206)(207). Recent evidence
suggests a role for these chemical hormones in regulating the
proliferation and differentiation of prostate cancer cells.
RA has marked but different effects on different human prostate cancer cells in vitro. LNCaP cells are induced to differentiate in response to RA in the presence or absence of androgens (208). DU-145 cells are growth-inhibited by 13-cis-retinoic acid (209). In contrast, RA induces increased invasiveness of PC-3 cells by up-regulating urokinase plasminogen activator expression, suggesting that in more advanced tumors, RA may promote disease progression (210). The different cell responses to RA may relate to their differences in AR expression. RA can inhibit the binding capacity of the LNCaP AR (which is mutated) by 30–40% (211), whereas PC-3 cells lack an AR.
A deficiency in vitamin D has been proposed to increase the risk of prostate cancer (212). This result is still controversial because a comparison of serum concentrations of vitamin D metabolites between prostate cancer patients and age-matched controls showed no substantial differences (213). However, a striking difference between black and white men in the allelic frequencies of the gene encoding a vitamin D3-binding protein has been reported, raising the possibility that this protein may be important in indicating risk of prostate cancer development (214).
Most evidence concerning the role of vitamin D3 in prostate cancer comes from studies of prostate cancer cell lines. The human lines DU-145, PC-3, and LNCaP all express receptors for vitamin D3, with PC-3 showing the greatest binding capacity (215). Proliferation of LNCaP and PC-3 cells is inhibited by vitamin D3 (215)(216), and proliferation of DU-145 is inhibited only by analogs of vitamin D3 (217). In LNCaP cells, exposure to vitamin D3 can neutralize the proliferative effects of androgens, suggesting that it is a strong inhibitor of epithelial cell proliferation (208). Vitamin D3 can up-regulate expression of IGF-BP-6 mRNA in a dose-dependent manner in all three human prostate cancer cell lines (216), suggesting that it may modulate growth via the IGF axis. In Lobund/Wistar rats, growth of a nontumorigenic AS epithelial cell line derived from the dorsal-lateral prostate is inhibited by vitamin D3, whereas its AI tumorigenic counterpart is insensitive to these effects (217). Taken together, these results suggest that vitamin D3 may act as a growth modulator of prostate epithelial cell proliferation, but that its inhibitory effects may be lost in late-stage prostate cancer.
A recent study has shown that prostate cancer may be associated with vitamin D receptor gene polymorphism. Race-adjusted combined analysis showed that men who were homozygous for the t allele (shown to correlate with higher serum concentrations of the active form of vitamin D) have only one-third of the risk of developing prostate cancer requiring prostatectomy compared with men who were heterozygotes or homozygous for the T allele (218). Because this is a single study, credence awaits confirmation, but this appears to suggest that vitamin D is an important determinant of prostate cancer risk and could lead to strategies for chemoprevention.
The vitamin D receptor is thought to act as a heterodimer with the retinoid X receptor, suggesting functional interactions between 1,25-dihydroxyvitamin D3 and retinoids (219). The combination of 1,25-dihydroxyvitamin D3 and 9-cis-retinoic acid was shown to act synergistically in inhibiting the growth of LNCaP cells.
In summary, different activities of RA on different cancer cell lines leave its potential role in controlling prostate cancer in doubt. The ability of RA to differentiate LNCaP cells has exciting potential for control of the disease, but its ability to increase invasiveness of PC-3 cells would argue against its use. However, Phase I clinical trials on the use of liarozole, which binds to the cytochrome P450-dependent hydroxylating enzymes involved in RA catabolism (220), and Phase II trials of all-trans-RA for control of hormone-refractory prostate cancer are in progress (221). Moreover, the use of liarozole fumarate, a compound that blocks the cytochrome P450-dependent catabolism of RA and which has been successful in reducing both AD and AI tumor growth in the Dunning rat model, has been proposed (222). The progression of prostate cancer to androgen independence is accompanied by a loss in sensitivity to the antiproliferative effects of vitamin D3. Hence this vitamin could be of therapeutic benefit in patients with early disease but not once progression to late-stage disease has occurred.
interactions of growth factor pathways in prostate cancer
The IGF axis appears to be under the control of other growth
factor pathways. EGF, FGF, and TGF-ß have been shown to regulate the
expression of the IGFs and their binding proteins. Antibodies to EGFR
inhibit the secretion of IGF-BP (74) and the
growth-promoting effects of IGF-1. The addition of bFGF to the human
osteoblast cell line MC3T-E1 inhibits IGF-1 and IFG-2 mRNA and
IGF-BP-2, IGF-BP-4, IGF-BP-5, and IGF-BP-6 concentrations
(223), whereas TGF-ß can increase the expression of IGF-1
mRNA in nondiseased human osteoblast-like cells (224).
Moreover, as mentioned previously, vitamin D3 can
up-regulate expression of IGF-BP-6 mRNA in a dose-dependent manner in
LNCaP, PC-3, and DU-145 prostate cancer cells (216). This
suggests that the interactions of the IGFs and their binding proteins
with other cytokines may in some way regulate prostate growth.
Additionally, any changes to this complex IGF regulatory system may
promote prostate cancer development. Moreover, several growth factor
pathways including IGF-1, KGF, and EGF can aberrantly activate the AR
in the absence of androgen, suggesting that the androgen-signaling
chain may be activated by growth factors in an androgen-depleted
environment (27).
|
Concluding Remarks |
|---|
|
TopAbstractThe Natural History of...Biological Properties of the...The AR in Prostate...The Role of Growth...Models for Studies of...Growth Factor Families Important...Concluding RemarksReferences |
|---|
), apparently bypassing any requirement for stromal cells, may
represent an adaptive response to androgen ablation in the growth
regulation of AI tumors. Future studies may lead to targeted therapy
for patients with advanced prostate cancer, using an understanding of
the changes that occur in growth factor pathways as prostate cancer
progresses (225). A recent example of a possible antigrowth
factor treatment is the use of suramin (225), which has
antiproliferative effects against prostate cancer cells in vitro. The
mode of action of suramin is unclear, but it may be mediated by its
ability to disrupt the cellular energy balance of prostate cancer cells
(226)(227). There are conflicting reports about
the ability of suramin to interfere with growth factor pathways; it can
inhibit the growth-stimulating effects of exogenous testosterone and
bFGF (228), but it is not counteracted by 10-fold excesses
of exogenous growth factors in vitro (229), and the effects
are reversible. Suramin is a polysulfonated naphthylurea with
substantial activity in prostate cancer. When high dosages of suramin
were used in 38 patients with hormone-refractory prostate cancer,
declines in serum PSA of >75% were obtained in 38% of patients, and
measurable disease response was observed in 35% of patients
(225)(230). Although preliminary studies on the effects of
growth factor receptor inhibitors and neural peptide inhibitors on
prostate cancer cells in vitro are very impressive, no definite
conclusions regarding the efficacy or clinical applicability can yet be
made. Understanding growth factors will help novel and innovative
therapeutic approaches, but it will obviously require much refinement
before being translated from bench to bedside.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractThe Natural History of...Biological Properties of the...The AR in Prostate...The Role of Growth...Models for Studies of...Growth Factor Families Important...Concluding RemarksReferences |
|---|
