l arti
opo
ros
, Mia
nd Mil
ond Mi
17 Sep
Ab
P tin and
M EPO a
car a (PIN
30
R igh-gr
BP R scor
EPOR or EPO score and Gleason score in PCa.
Conclusions: Up-regulation of EPOR has a more important role in comparison with EPO in prostate carcinogenesis. Differential
ove
bet
Key
1.
ing
tio
adv
tum
mo
vas
car
tiss
lev
clo
[1,
(05
653
riginal
107
doi
ARTICLE IN PRESS
r-expression of EPOR and EPO in benign and malignant prostatic tissue is ascribed to different mechanisms involved in up-regulation
ween EPO and EPOR. © 2009 Elsevier Inc. All rights reserved.
words: Prostate carcinoma; Erythropoietin; Erythropoietin receptor; Hypoxia
Introduction
Hypoxia is a common phenomenon in solid tumors, result-
from inadequate blood supply with impaired vascular func-
n. Hypoxia stress has not always been considered to be
erse to the survival and growth of tumors. Actually, once
or cells adapt to the hypoxia environment, they become
re aggressive as hypoxia directs the remodeling of tumor
culature or phenotypic changing of tumor cells.
Hypoxia regions has been demonstrated in human prostate
cinoma (PCa) by analyzing the extent of hypoxia in tumor
ues with Eppendorf pO2 microelectrode, and increasing
els of hypoxia in prostate carcinomas were reported to be
sely associated with increasing clinical stage and patient age
2]. Moreover, over-expression of hypoxia-inducible fac-
tor-1� (HIF-1�), the surrogate marker of hypoxia, was
shown in human PCa, and its expression was considered as
an early event for prostate carcinogenesis [3].
Once the transcription factor HIF-1� is activated, more
than 60 putative target genes will actively participate in
tumor cells’ response to hypoxia stress. Of them, erythro-
poietin (EPO) with its receptor (EPOR), a member of the
type I cytokine receptor superfamily, has recently been
reported to be expressed in human PCa [4]. Previously, we
have demonstrated up-regulation of EPOR in human PCa
and high-grade prostate intraepithelial neoplasia (high-
grade PIN) [5]. In the present study, we further investigated
the differential up-regulation of EPO and EPOR in normal,
benign and malignant human prostatic tissue.
2. Materials and methods
2.1. Patients and samples
� This project was funded by a grant from Shanghai PuJiang Program
PJ14005).
* Corresponding author. Tel.: �86-21-25070586; fax: �86-21-
38288.
Origina
Differential up-regulation of erythr
and malignant p
Chuanliang Xu, M.D.a,1, Tie Zhou, M.D.a,1
a Department of Urology, Shanghai Hospital, The Seco
b Department of Pathology, Shanghai Hospital, The Sec
Received 22 June 2008; received in revised form
stract
urpose: To investigate differential up-regulation of erythropoie
aterials and methods: An immunohistochemical analysis of
cinoma (PCa) with 16 high-grade prostate intraepithelial neoplasi
normal prostatic tissue samples as control.
esults: Over-expression of EPOR was only shown in PCa and h
H tissue. There was significant relationship between EPO and EPO
Urologic Oncology: Seminars and O
pro
E-mail address: sunyh@medmail.com.cn (Y. Sun).
1 These authors contributed equally to this work.
8-1439/09/$ – see front matter © 2009 Elsevier Inc. All rights reserved.
:10.1016/j.urolonc.2008.09.023
cle
ietin and its receptor in benign
tatic tissue�
oxia He, M.D.b, Yinghao Sun, M.D.a,*
itary Medical University, Shanghai, P. R. China
litary Medical University, Shanghai, P. R. China
tember 2008; accepted 22 September 2008
its receptor in benign and malignant prostatic tissue.
nd EPOR expression was performed on 30 cases of prostate
), 50 cases of benign prostatic hyperplasia (BPH) lesions, and
ade PIN tissue but over-expression of EPO was also shown in
e in BPH, high-grade PIN, and PCa but no association between
Investigations xx (2009) xxx
Study protocols involving human materials were ap-
ved by the institutional ethics committee at Changhai
Ho
the
tw
fro
(av
pro
fro
ye
ng
tat
ob
ye
an
the
sco
Gl
So
som
tox
rev
pa
2.2
for
sec
no
ma
scr
po
no
sli
gra
in
bu
an
4°C
de
En
30
bu
DA
zid
aft
sta
sam
(ne
sec
pa
bra
ba
2.3
po
tat
wi
sta
ma
for
pe
sys
als
sta
tita
pre
int
str
be
inc
an
wi
im
DU
we
wh
sam
sta
2.4
EP
wa
EP
sco
for
NC
be
3.
H
ch
fou
mu
gla
Tab
Ch
PSA
Me
15.
P
2 C. Xu et al. / Urologic Oncology: Seminars and Original Investigations xx (2009) xxx
ARTICLE IN PRESS
spital. All the samples were from specimens archived in
Department of Pathology at the Changhai Hospital be-
een 2003 and 2004. PCa tissue samples were obtained
m 30 patients with an age range from 60 to 77 years
erage 70 years), who had undergone radical retropubic
statectomy in the hospital. BPH tissue samples were
m 50 patients (aged from 62 to 79 years, average 72
ars) with prostate-specific antigen (PSA) level less than 4
/ml, who underwent transurethral resection of the pros-
e. A total of 30 normal prostate tissue samples were
tained by autopsy from the patients aged from 20 to 30
ars (average 24 years). Patients with PCa did not receive
y treatment such as radiation, chemotherapy, or hormonal
rapy before the removal of the prostate. Tumors were
red with the Gleason system. PSA, tumor stage, and
eason score of patients with PCa are shown in Table 1.
me PCa samples had adjacent high-grade PIN lesions and
e BPH samples contained inflammation lesions. Hema-
ylin and eosin (H & E) stained slides of all cases were
iewed, and the diagnosis was confirmed by two senior
thologists.
. Immunohistochemical staining
Four-�m-thick paraffin sections were serially cut from
malin-fixed, paraffin-embedded tissues. One of these
tions was stained with H & E for histopathologic diag-
sis. Immunohistochemical assays were performed on for-
lin-fixed, paraffin-embedded sections as previously de-
ibed [6]. The primary antibodies used were rabbit
lyclonal antibodies purchased from Santa Cruz Biotech-
logy, Inc. (Santa Cruz, CA). Sections of 4 �m on glass
des were deparaffinized in Hemo-D and rehydrated in
ded alcohols, followed by endogenous peroxidase block
3% H2O2 and antigen retrieval in boiling 10% citrate
ffer. Then, slides were incubated with the polyclonal
tibody against EPO, EPOR (1:200 dilution) overnight at
, and subsequently with horseradish peroxidase labeled
xtran polymer coupled to anti-rabbit antibody (DAKO
vision� System HRP, DAKO Inc., Carpinteria, CA) for
minutes at room temperature after 3 washes with tris
ffered saline containing Tween 20 (TBST, pH 7.6,
KO). Finally, slides were developed with diaminoben-
ine for 10 minutes and counterstained with hematoxylin
er washing three times with TBST. The specificity of
ining was confirmed by processing sections from the
le 1
aracteristics of patients with PCa, n � 30, 2003–2004
(ng/ml) Tumor stages Gleason score
dian range pT2 pT3a pT3b 2–4 5–7 8 9
6 4.1–23.6 16 8 6 7 16 6 1
SA � prostate-specific antigen; PCa � prostate carcinoma.
e paraffin block with omission of the primary antibody of
gative control). As a positive control, reactions with
tions of breast cancer archived in the pathological de-
rtment of our hospital were used. Cytoplasmic or mem-
ne staining that was clearly distinguishable from the
ckground was considered positive.
. Semiquantitative analysis and interpretation of staining
At least 500 epithelial cells within each area showing
sitive immunoreactivity were evaluated in normal pros-
e, BPH, high-grade PIN, and PCa. The percentage of cells
th no staining (0) or weak (1), moderate (2), or intense
ining (3) was analyzed by visual inspection under �100
gnification and a staining score was calculated using the
mula: weighted mean of stain intensity � (�intensity �
rcentage of cells)/total percentage of cells. The scoring
tem took into account not only the staining intensity but
o the percentage of the cells that exhibit EPO, EPOR
ining. EPO and EPOR expression was graded semiquan-
tively according to the results of staining score. In the
sent study, EPO and EPOR staining were also classified
o overexpression, which was defined as moderate or
ong staining shown in any prostatic epithelium within
nign or malignant tissues, and normal expression, which
luded weak or negative staining of EPO or EPOR. These
alyses were performed using a Nikon E-400 microscope
th computer-aided image analysis system, and digital
ages were captured using a digital camera (Nikon
100; Tokyo, Japan) at �200 magnification [7]. Slides
re evaluated twice at different times by 3 investigators
o were unaware of the pathologic characteristics of the
ples, and the resulting mean levels were used for the
tistical analyses.
. Statistics
Kruskal-Wallis test was used to compare differences in
O, EPOR staining scores among groups. Spearman test
s used to analyze the relationship between EPO and
OR score for each group, and to relate EPO or EPOR
re to Gleason score in PCa. Computations were per-
med using SAS 9.1.3 software (SAS Institute Inc., Cary,
). A two-sided P value less than 0.05 was considered to
statistically significant.
Results
PCa, high-grade PIN, and BPH lesions were identified on
& E staining sections according to morphological
anges (Fig. 1). A total of 16 high-grade PIN lesions were
nd in 16 PCa samples, and no PCa sample showed
ltiple high-grade PIN lesions.
EPO and EPOR expression were observed mainly in the
ndular epithelium. A predominantly cytoplasmic pattern
staining for EPO and EPOR was observed, but membrane
im
an
tis
ma
ov
PIN
sta
pro
EP
fur
gra
tre
bu
an
EP
score 4
Tab
EP
nor
Imm
EP
�
�
�
�
EP
EP
H
pro
ma
ove
a
b
3C. Xu et al. / Urologic Oncology: Seminars and Original Investigations xx (2009) xxx
ARTICLE IN PRESS
munoreactivity was also noted for EPOR. Staining level
d median score of EPO and EPOR in different prostate
sues are shown in Table 2. Over-expression of EPOR was
inly shown in high-grade PIN and PCa (Fig. 2), whereas
er-expression of EPO was shown not only in high-grade
and PCa but also in BPH tissue (Fig. 3). No positive
ining for EPO and EPOR was found for all the normal
static tissues tested.
PCa and high-grade PIN had significantly higher median
OR score than BPH (P � 0.0001; P � 0.0096) and
thermore, PCa had higher median EPOR score than high-
de PIN (P � 0.0072). With regard to EPO, an increased
Fig. 1. Morphology of BPH (A), high-grade PIN (B), PCa (Gleason
le 2
O and EPOR immunohistochemistry in different prostate lesions and
mal prostatic epithelia
unoreactivity PCa High-grade PIN BPHa Normalb
O EPOR
� 22 10 0 0
� 0 2 22 0
� 8 4 0 0
� 0 0 28 30
O Score 2.38 1.75 0.98 —
OR Score 2.50 1.65 0.73 —
igh-grade PIN � high-grade prostatic intraepithelial neoplasia; PCa �
state carcinoma; BPH � benign prostatic hyperplasia; Normal � nor-
l prostate; � � normal expression (negative or weak expression); � �
rexpression (moderate or strong expression).
10 BPH samples contained inflammation lesions.
whNegative expression of EPO and EPOR in normal prostate.
nd was observed among BPH, high-grade PIN, and PCa,
t only the difference of median EPO score between PCa
d BPH was statistically significant (P � 0.0064).
There was a significant correlation between EPO and
OR score in BPH, high-grade PIN, and PCa (P � 0.031,
� 0.014, P � 0.003, respectively). The relationship be-
een Gleason score and EPO or EPOR score in PCa was
t observed. Statistic analysis showed that there was no
ociation between EPOR/EPO score and Gleason score in
a (P � 0.085, P � 0.616).
Comment
EPO, a 30.4 kDa glycoprotein, is the principal regulator of
blood cell production in mammals by binding EPOR [8],
produced primarily in the adult kidney under the control of
oxygen-sensing mechanism. Although there are several
tors, such as hypoglycemia, increased intracellular calcium,
ulin release, estrogen, and androgenic steroids, and various
okines can modulate EPO production, low tissue oxygen
sion is still the main driver that induces EPO gene expres-
n through both transcriptional activation and mRNA stabi-
ation [9,10]. The hypoxia dependent up-regulation of EPO
ectly results from the activation of HIF-1�, which binds a
poxia responsive element in the 3= flanking region of the
O gene as a transcription factor.
Recently, expression of EPOR and EPO in tumor cell
es as well as primary cancers has been reported [11–18],
) (C), and PCa (Gleason score 7) (D). H & E staining (�200).
P
tw
no
ass
PC
4.
red
and
an
fac
ins
cyt
ten
sio
liz
dir
hy
EP
lin
ich suggests the potential for generation of an autocrine
or
cel
ate
inc
ch
EP
Fig
mo leason
Fig
in h
4 C. Xu et al. / Urologic Oncology: Seminars and Original Investigations xx (2009) xxx
ARTICLE IN PRESS
paracrine growth-stimulatory EPO-EPOR loop in cancer
ls. Thus, the activation of EPO-EPOR loop may be associ-
d with modulation of various aspects of tumor biology,
. 2. Immunohistochemical staining of EPOR in high-grade PIN and PCa.
derate in high-grade PIN (�400); (C) PCa (Gleason score 4); (D) PCa (G
. 3. Immunohistochemical staining of EPO in BPH, high-grade PIN, and
igh-grade PIN (�400); (C) PCa (Gleason score 4); (D) PCa (Gleason score 7) e
luding cellular proliferation, apoptosis, and sensitivity to
emoradiation therapy. Furthermore, the expression of
OR in vascular endothelium in tumors has suggested
gative staining for EPOR in BPH (�200); (B) expression of EPOR is
score 8) exhibiting strong EPOR immunoreactivity (�400).
) Moderate EPO staining in BPH (�400); (B) moderate EPO staining
PCa. (A
(A) Ne
xhibiting strong EPO immunoreactivity (�400).
po
ge
an
[19
of
ten
inv
sio
en
ma
cli
bre
fou
hy
EP
can
ex
sie
Th
tor
filt
au
na
can
sig
les
log
EP
via
EP
pre
cri
tho
rel
hy
reg
ma
fou
wi
ov
hy
an
mi
on
of
eri
sug
BP
Fu
wa
BP
po
EP
wa
eff
firm
fin
5.
ha
co
EP
asc
be
Re
[1
[2
[3
[4
[5
[6
[7
[8
[9
[10
[11
[12
[13
[14
5C. Xu et al. / Urologic Oncology: Seminars and Original Investigations xx (2009) xxx
ARTICLE IN PRESS
tential effects of EPO on the stimulation of tumor angio-
nesis.
EPOR expression in human prostate has been reported,
d the expression of EPOR in prostate was functional
,20]. In the previous study, we also found overexpression
EPOR in PCa and high-grade PIN. Considering the po-
tial existence of EPO-EPOR loop in PCa, we further
estigated, in this study, whether there was overexpres-
n of EPO in PCa and explored the significance of differ-
tial up-regulation of EPO and its receptor in benign and
lignant prostate tissue.
Acs et al. reported high levels of EPO and EPOR in 50
nical specimens of breast carcinoma but not in normal
ast, benign papilloma, or fibrocystic tissue [21]. They
nd the highest levels of EPO and EPOR expression in
poxic tumor regions and thought that enhanced EPO-
OR signaling contributed to the promotion of human
cer by tissue hypoxia. Mohyeldin et al. investigated the
pression and function of EPO and EPOR in tumor biop-
s and cell lines from human head and neck cancer [16].
ey reported the highest expression of EPO and its recep-
in malignant cells, particularly within hypoxic and in-
rating tumor regions. Their data supported a role for
tocrine or paracrine EPO-EPOR signaling in the malig-
nt progression and local invasiveness of head and neck
cer. More importantly, Yasuda et al. found EPO-EPOR
naling pathway in 24 malignant human cell lines regard-
s of their origins, types, genetic characteristics, and bio-
ical properties [13]. Their findings suggested that EPO-
OR signaling was indispensable for the growth and
bility of malignant tumor.
Similar to previous studies, up-regulation of EPO and
OR was also found in PCa and high-grade PIN in the
sent study, which suggested a potential role of the auto-
ne EPO-EPOR signaling in prostate tumorigenesis. Al-
ugh the mechanism involved in EPO-EPOR signaling-
ated prostate tumorigenesis is still unknown, undoubtedly
poxia plays an important and indispensable role.
Up to now, to our best knowledge, the differential up-
ulation of EPO and its receptor (EPOR) in benign and
lignant prostatic tissue has not been investigated. We
nd an over-expression of EPOR in all PCa tissue tested,
th the majority of them over-expressing EPO. Although
er-expressed EPO and EPOR have been ascribed to be
poxia-related, the differential over-expression of EPO
d EPOR in PCa tissue suggested non-hypoxia mechanism
ght be also involved in up-regulation of EPOR. Actually,
cogenic mechanism has been stressed in the up-regulation
EPOR [22].
Over-expressed EPO was also shown in BPH. Consid-
ng that EPO was one of the target genes of hypoxia, we
gest that hypoxia may play a role in development of
H, even malignant transition of benign prostatic tissue.
rthermore, given the fact that over-expression of EPOR
s observed only in high-grade PIN and PCa but not in
H tissue, up-regulation of EPOR may have a more im-
rtant role in prostate carcinogenesis.
In this study, we did not find any relationship between
O/EPOR score and Gleason score. Gleason score is al-
ys related to prognosis of PCa, so EPO or EPOR as an
ective prognostic factor for PCa has not yet been con-
ed. Considering that the current sample is small, the
al conclusion needs a further large-sample study.
Conclusions
Our data have demonstrated that up-regulation of EPOR
s a more important role for prostate carcinogenesis in
mparison with EPO. Differential over-expression of
OR and EPO in benign and malignant prostatic tissue is
ribed to different mechanisms involved in up-regulation
tween EPO and EPOR.
ferences
] Movsas B, Chapman JD, Horwitz EM, et al. Hypoxic regions exist in
human prostate carcinoma. Urology 1999;53:11–8.
] Movsas B, Chapman JD, Greenberg RE, et al. Increasing levels of
hypoxia in prostate carcinoma correlate significantly with increasing
clinical stage and patient age. Cancer 2000;89:2018–24.
] Zhong H, Semenza GL, Simons JW, et al. Up-regulation of hypoxia-
inducible factor 1alpha is an early event in prostate carcinogenesis.
Cancer Detect Prev 2004;28:88–93.
] Arcasoy MO, Amin K, Vollmer RT, et al. Erythropoietin and eryth-
ropoietin receptor expression in human prostate cancer. Mod Pathol
2004;18:421–30.
] Zhou T, Xu C, He M, et al. Up-regulation of erythropoietin receptor
in human prostate carcinoma and high-grade prostate intraepithelial
neoplasia. Prostate Cancer Prostatic Dis 2008;11:143–7.
] Acs G, Zhang PJ, Rebbeck TR, et al. Immunohistochemical expres-
sion of erythropoietin and erythropoietin receptor in breast carci-
noma. Cancer 2002;95:969–81.
] Muller W. The reliability of immunohistochemistry as a prescreening
method for the diagnosis of hereditary nonpolyposis colorectal cancer
(HNPCC)—results of an international collaborative study. Fam Can-
cer 2001;1:87–92.
] Krantz SB. Erythropoietin. Blood 1991;77:419–34.
] Jelkmann W. Erythropoietin: Structure, control of production, and
function. Physiol Rev 1992;72:449–89.
] Ebert BL, Bunn HF. Regulation of the erythropoietin gene. Blood
1999;94:1864–77.
] Batra S, Perelman N, Luck LR, et al. Pediatric tumor cells express
erythropoietin and a functional erythropoietin receptor that promotes
angiogenesis and tumor cell survival. Lab Invest 2003;83:1477–87.
] Arcasoy MO, Jiang X, Haroon ZA. Expression of erythropoietin
receptor splice variants in human cancer. Biochem Biophys Res
Commun 2003;307:999–1007.
] Yasuda Y, Fujita Y, Matsuo T, et al. Erythropoietin regulates tumor
growth of human malignancies. Carcinogenesis 2003;24:1021–9.
] Kumar SM, Acs G, Fang D, et al. Functional erythropoietin autocrine
loop in melanoma. Am J Pathol 2005;166:823–30.
[15] Lai SY, Childs EE, Xi S, et al. Erythropoietin mediated activation of
JAK-STAT signaling contributes to cellular invasion in head and
neck squamous cell carcinoma. Oncogene 2005;24:4442–9.
[16] Mohyeldin A, Lu H, Dalgard C, et al. Erythropoietin signaling pro-
motes invasiveness of human head and neck squamous cell carci-
noma. Neoplasia 2005;7:537–43.
[17] Dagnon K, Pacary E, Commo F, et al. E