Expression of steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion
Aim: To analyze an expression pattern of the steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion; and to examine a putative correlation between gene expression and clinical characteristics, to define...
Gespeichert in:
| Datum: | 2018 |
|---|---|
| Hauptverfasser: | , , , , , , |
| Format: | Artikel |
| Sprache: | English |
| Veröffentlicht: |
Інститут експериментальної патології, онкології і радіобіології ім. Р.Є. Кавецького НАН України
2018
|
| Schriftenreihe: | Experimental Oncology |
| Schlagworte: | |
| Online Zugang: | http://dspace.nbuv.gov.ua/handle/123456789/145575 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Zitieren: | Expression of steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion / G.V. Gerashchenko, L.V. Mevs, L.I. Chashchina, M.V. Pikul, O.P. Gryzodub, E.O. Stakhovsky, V.I. Kashuba // Experimental Oncology. — 2018. — Т. 40, № 2. — С. 101–108 — Бібліогр.: 30 назв. — англ. |
Institution
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
irk-123456789-145575 |
|---|---|
| record_format |
dspace |
| spelling |
irk-123456789-1455752019-01-25T01:23:40Z Expression of steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion Gerashchenko, G.V. Mevs, L.V. Chashchina, L.I. Pikul, M.V. Gryzodub, O.P. Stakhovsky, E.O. Kashuba, V.I. Original contributions Aim: To analyze an expression pattern of the steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion; and to examine a putative correlation between gene expression and clinical characteristics, to define the molecular subtypes of prostate cancer. Materials and Methods: The relative gene expression (RE) of 33 transcripts (27 genes) and the presence/absence of the TMPRSS2/ERG fusion were analyzed by a quantitative PCR. 37 prostate cancer tissues (T) paired with conventionally normal prostate tissue (CNT) and 21 samples of prostate adenomas were investigated. RE changes were calculated, using different protocols of statistics. Results: We demonstrated differences in RE of seven genes between tumors and CNT, as was calculated, using the 2−ΔCT model and the Wilcoxon matched paired test. Five genes (ESR1, KRT18, MKI67, MMP9, PCA3) showed altered expression in adenocarcinomas, in which the TMPRSS2/ERG fusion was detected. Two genes (INSR, isoform B and HOTAIR) expressed differently in tumors without fusion. Comparison of the gene expression pattern in adenomas, CNT and adenocarcinomas demonstrated that in adenocarcinomas, bearing the TMPRSS2/ERG fusion, genes KRT18, PCA3, and SCHLAP1 expressed differently. At the same time, we detected differences in RE of AR (isoform 2), MMP9, PRLR and HOTAIR in adenocarcinomas without the TMPRSS2/ERG fusion. Two genes (ESR1 and SRD5A2) showed differences in RE in both adenocarcinoma groups. Fourteen genes, namely AR (isoforms 1 and 2), CDH1, OCLN, NKX3-1, XIAP, GCR (ins AG), INSR (isoform A), IGF1R, IGF1R tr, PRLR, PRL, VDR and SRD5A2 showed correlation between RE and tumor stage. RE of four genes (CDH2, ESR2, VDR and SRD5A2) correlated with differentiation status of tumors (Gleason score). Using the K-means clustering, we could cluster adenocarcinomas in three groups, according to gene expression profiles. A specific subtype of prostate tumors is characterized by the activated ERG signaling, due to the presence of TMPRSS2/ERG fusion, and also by high levels of the androgen receptor, prolactin, IGF, INSR and PCA3. Conclusions: We have found the specific differences in expression of the steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes, depending on the presence/absence of the TMPRSS2/ERG fusion in prostate adenocarcinomas, CNT and adenomas. We showed three different gene expression profiles of prostate adenocarcinomas. One of them is characteristic for adenocarcinomas with the TMPRSS2/ERG fusion. Further experiments are needed to confirm these data in a larger cohort of patients. Key Words: prostate tumors, TMPRSS2/ERG fusion, gene expression patterns, steroid receptors, peptide receptors, EMT regulation. 2018 Article Expression of steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion / G.V. Gerashchenko, L.V. Mevs, L.I. Chashchina, M.V. Pikul, O.P. Gryzodub, E.O. Stakhovsky, V.I. Kashuba // Experimental Oncology. — 2018. — Т. 40, № 2. — С. 101–108 — Бібліогр.: 30 назв. — англ. 1812-9269 http://dspace.nbuv.gov.ua/handle/123456789/145575 en Experimental Oncology Інститут експериментальної патології, онкології і радіобіології ім. Р.Є. Кавецького НАН України |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| language |
English |
| topic |
Original contributions Original contributions |
| spellingShingle |
Original contributions Original contributions Gerashchenko, G.V. Mevs, L.V. Chashchina, L.I. Pikul, M.V. Gryzodub, O.P. Stakhovsky, E.O. Kashuba, V.I. Expression of steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion Experimental Oncology |
| description |
Aim: To analyze an expression pattern of the steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion; and to examine a putative correlation between gene expression and clinical characteristics, to define the molecular subtypes of prostate cancer. Materials and Methods: The relative gene expression (RE) of 33 transcripts (27 genes) and the presence/absence of the TMPRSS2/ERG fusion were analyzed by a quantitative PCR. 37 prostate cancer tissues (T) paired with conventionally normal prostate tissue (CNT) and 21 samples of prostate adenomas were investigated. RE changes were calculated, using different protocols of statistics. Results: We demonstrated differences in RE of seven genes between tumors and CNT, as was calculated, using the 2−ΔCT model and the Wilcoxon matched paired test. Five genes (ESR1, KRT18, MKI67, MMP9, PCA3) showed altered expression in adenocarcinomas, in which the TMPRSS2/ERG fusion was detected. Two genes (INSR, isoform B and HOTAIR) expressed differently in tumors without fusion. Comparison of the gene expression pattern in adenomas, CNT and adenocarcinomas demonstrated that in adenocarcinomas, bearing the TMPRSS2/ERG fusion, genes KRT18, PCA3, and SCHLAP1 expressed differently. At the same time, we detected differences in RE of AR (isoform 2), MMP9, PRLR and HOTAIR in adenocarcinomas without the TMPRSS2/ERG fusion. Two genes (ESR1 and SRD5A2) showed differences in RE in both adenocarcinoma groups. Fourteen genes, namely AR (isoforms 1 and 2), CDH1, OCLN, NKX3-1, XIAP, GCR (ins AG), INSR (isoform A), IGF1R, IGF1R tr, PRLR, PRL, VDR and SRD5A2 showed correlation between RE and tumor stage. RE of four genes (CDH2, ESR2, VDR and SRD5A2) correlated with differentiation status of tumors (Gleason score). Using the K-means clustering, we could cluster adenocarcinomas in three groups, according to gene expression profiles. A specific subtype of prostate tumors is characterized by the activated ERG signaling, due to the presence of TMPRSS2/ERG fusion, and also by high levels of the androgen receptor, prolactin, IGF, INSR and PCA3. Conclusions: We have found the specific differences in expression of the steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes, depending on the presence/absence of the TMPRSS2/ERG fusion in prostate adenocarcinomas, CNT and adenomas. We showed three different gene expression profiles of prostate adenocarcinomas. One of them is characteristic for adenocarcinomas with the TMPRSS2/ERG fusion. Further experiments are needed to confirm these data in a larger cohort of patients. Key Words: prostate tumors, TMPRSS2/ERG fusion, gene expression patterns, steroid receptors, peptide receptors, EMT regulation. |
| format |
Article |
| author |
Gerashchenko, G.V. Mevs, L.V. Chashchina, L.I. Pikul, M.V. Gryzodub, O.P. Stakhovsky, E.O. Kashuba, V.I. |
| author_facet |
Gerashchenko, G.V. Mevs, L.V. Chashchina, L.I. Pikul, M.V. Gryzodub, O.P. Stakhovsky, E.O. Kashuba, V.I. |
| author_sort |
Gerashchenko, G.V. |
| title |
Expression of steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion |
| title_short |
Expression of steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion |
| title_full |
Expression of steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion |
| title_fullStr |
Expression of steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion |
| title_full_unstemmed |
Expression of steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion |
| title_sort |
expression of steroid and peptide hormone receptors, metabolic enzymes and emt-related genes in prostate tumors in relation to the presence of the tmprss2/erg fusion |
| publisher |
Інститут експериментальної патології, онкології і радіобіології ім. Р.Є. Кавецького НАН України |
| publishDate |
2018 |
| topic_facet |
Original contributions |
| url |
http://dspace.nbuv.gov.ua/handle/123456789/145575 |
| citation_txt |
Expression of steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion / G.V. Gerashchenko, L.V. Mevs, L.I. Chashchina, M.V. Pikul, O.P. Gryzodub, E.O. Stakhovsky, V.I. Kashuba // Experimental Oncology. — 2018. — Т. 40, № 2. — С. 101–108 — Бібліогр.: 30 назв. — англ. |
| series |
Experimental Oncology |
| work_keys_str_mv |
AT gerashchenkogv expressionofsteroidandpeptidehormonereceptorsmetabolicenzymesandemtrelatedgenesinprostatetumorsinrelationtothepresenceofthetmprss2ergfusion AT mevslv expressionofsteroidandpeptidehormonereceptorsmetabolicenzymesandemtrelatedgenesinprostatetumorsinrelationtothepresenceofthetmprss2ergfusion AT chashchinali expressionofsteroidandpeptidehormonereceptorsmetabolicenzymesandemtrelatedgenesinprostatetumorsinrelationtothepresenceofthetmprss2ergfusion AT pikulmv expressionofsteroidandpeptidehormonereceptorsmetabolicenzymesandemtrelatedgenesinprostatetumorsinrelationtothepresenceofthetmprss2ergfusion AT gryzodubop expressionofsteroidandpeptidehormonereceptorsmetabolicenzymesandemtrelatedgenesinprostatetumorsinrelationtothepresenceofthetmprss2ergfusion AT stakhovskyeo expressionofsteroidandpeptidehormonereceptorsmetabolicenzymesandemtrelatedgenesinprostatetumorsinrelationtothepresenceofthetmprss2ergfusion AT kashubavi expressionofsteroidandpeptidehormonereceptorsmetabolicenzymesandemtrelatedgenesinprostatetumorsinrelationtothepresenceofthetmprss2ergfusion |
| first_indexed |
2025-07-10T21:59:06Z |
| last_indexed |
2025-07-10T21:59:06Z |
| _version_ |
1837298917696339968 |
| fulltext |
Experimental Oncology 40, 101–108, 2018 (June) 101
EXPRESSION OF STEROID AND PEPTIDE HORMONE RECEPTORS,
METABOLIC ENZYMES AND EMT-RELATED GENES IN PROSTATE
TUMORS IN RELATION TO THE PRESENCE OF THE TMPRSS2/ERG
FUSION
G.V. Gerashchenko1, #, *, L.V. Mevs1, #, L.I. Chashchina1, M.V. Pikul2, O.P. Gryzodub3, E.O. Stakhovsky2,
V.I. Kashuba1
1Institute of Molecular Biology and Genetics NAS of Ukraine, Kyiv 03680, Ukraine
2National Cancer Institute, Ministry of Health of Ukraine, Kyiv 03022, Ukraine
3Institute of Urology, NAMS of Ukraine, Kyiv 04053, Ukraine
Aim: To analyze an expression pattern of the steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes
in prostate tumors in relation to the presence of the TMPRSS2/ERG fusion; and to examine a putative correlation between gene
expression and clinical characteristics, to define the molecular subtypes of prostate cancer. Materials and Methods: The relative gene
expression (RE) of 33 transcripts (27 genes) and the presence/absence of the TMPRSS2/ERG fusion were analyzed by a quan-
titative PCR. 37 prostate cancer tissues (T) paired with conventionally normal prostate tissue (CNT) and 21 samples of prostate
adenomas were investigated. RE changes were calculated, using different protocols of statistics. Results: We demonstrated differences
in RE of seven genes between tumors and CNT, as was calculated, using the 2−ΔCT model and the Wilcoxon matched paired test.
Five genes (ESR1, KRT18, MKI67, MMP9, PCA3) showed altered expression in adenocarcinomas, in which the TMPRSS2/ERG
fusion was detected. Two genes (INSR, isoform B and HOTAIR) expressed differently in tumors without fusion. Comparison of the
gene expression pattern in adenomas, CNT and adenocarcinomas demonstrated that in adenocarcinomas, bearing the TMPRSS2/
ERG fusion, genes KRT18, PCA3, and SCHLAP1 expressed differently. At the same time, we detected differences in RE of AR
(isoform 2), MMP9, PRLR and HOTAIR in adenocarcinomas without the TMPRSS2/ERG fusion. Two genes (ESR1 and SRD5A2)
showed differences in RE in both adenocarcinoma groups. Fourteen genes, namely AR (isoforms 1 and 2), CDH1, OCLN, NKX3-1,
XIAP, GCR (ins AG), INSR (isoform A), IGF1R, IGF1R tr, PRLR, PRL, VDR and SRD5A2 showed correlation between RE and
tumor stage. RE of four genes (CDH2, ESR2, VDR and SRD5A2) correlated with differentiation status of tumors (Gleason score).
Using the K-means clustering, we could cluster adenocarcinomas in three groups, according to gene expression profiles. A specific
subtype of prostate tumors is characterized by the activated ERG signaling, due to the presence of TMPRSS2/ERG fusion, and
also by high levels of the androgen receptor, prolactin, IGF, INSR and PCA3. Conclusions: We have found the specific differences
in expression of the steroid and peptide hormone receptors, metabolic enzymes and EMT-related genes, depending on the pre-
sence/absence of the TMPRSS2/ERG fusion in prostate adenocarcinomas, CNT and adenomas. We showed three different gene
expression profiles of prostate adenocarcinomas. One of them is characteristic for adenocarcinomas with the TMPRSS2/ERG
fusion. Further experiments are needed to confirm these data in a larger cohort of patients.
Key Words: prostate tumors, TMPRSS2/ERG fusion, gene expression patterns, steroid receptors, peptide receptors, EMT
regulation.
Alterations in expression of the androgen recep-
tor (AR) are often associated with development
of prostate cancer. It is known already that the AR gene
expression is regulated by quite many molecular
pathways [1]. Another example of frequent altera-
tions in prostate tumors is formation of gene fusions
of androgen dependent gene TMPRSS2 (transmem-
brane protease, serine 2) with the ETS (E26 transfor-
mation-specific) family in particular with ERG (ETS
related gene) [2]. Previously, we have shown that the
TMPRSS2/ERG fusion is present in prostate adeno-
carcinoma and even in conventionally normal prostate
tissue (CNT) in a group of patients of the Ukrainian
population [3]. Therefore, we may speculate that the
presence or absence of the gene fusions could be the
cause of development of various prostate cancer
types with different sensitivity to therapy, recurrence
and metastasizing, despite the similar histological
characteristics [4].
One of the important characteristics of normal
functioning of prostate epithelial cells is sensitivity
to steroid and peptide hormones. In the process of cell
transformation, tumor cells often lose the sensitivity
to hormones and growth factors and also change their
metabolism. The AR is a key element of prostate func-
tioning and is involved in malignant transformation.
As was shown already, AR signaling plays a primary role
in development of androgen resistant and castration-
resistant prostate cancer [1]. There are few isoforms
of ARs. Some of them are prostate specific. AR expres-
Submitted: December 22, 2017.
*Correspondence: E-mail: g.v.gerashchenko@imbg.org.ua
#These authors contributed equally to this work
Abbreviations used: A — prostate adenomas; AR — androgen
receptor; CNT — conventionally normal prostate tissue; CPC —
clinical and pathological characteristics; EMT — epithelial-to-me-
senchymal cell transition; FDR — false discovery rate; N — normal/
conventionally normal prostate tissue; PSA — prostate-specific
antigen; qPCR — quantitative polymerase chain reaction; RE —
relative gene expression; RNA — ribonucleic acids; T — pros-
tate cancer, adenocarcinoma; TNM — International System
of Classification of Tumors, based on tumor-node-metastasis;
WHO — World Health Organization.
Exp Oncol 2018
40, 2, 101–108
102 Experimental Oncology 40, 101–108, 2018 (June)
sion can change during prostate carcinogenesis. Thus,
the overexpression of AR isoform A (1 isof) decreases
proliferation but accelerates invasion of prostate
cancer cells, compared with overexpression of AR iso-
form B (2 isof) [5]. Also, it was proposed that forma-
tion of the fusion between TMPRSS2 and ERG might
be controlled by androgens [6].
In prostate cells, the most potent AR ago-
nist is dihydrotestosterone. This is a metabolite
of testosterone, and the reaction of conversion
is catalyzed by SRD5A1 (5α-reductase, type 1)
and SRD5A2 (5α-reductase, type 2). The latter are
expressed at low levels in normal prostate tissues,
but upon prostate cancer progression expression
of these enzymes is altered [7]. Noteworthy, pros-
tate cancer is a complex pathology and many other
hormone receptors and corresponding pathways
are involved in tumor development, especially GCR
(glucocorticoid receptor, NR3C1 nuclear receptor
subfamily 3 group C member 1), IGF1R (insulin like
growth factor 1 receptor), ESR1 and ESR2 (estrogen
receptors 1 and 2), PRLR (prolactin receptor), VDR
(vitamin D receptor) and others.
Of note, GCR and AR share several transcriptional
targets [8]. All of the three isoforms of GCR (alpha (A),
beta (B) and gamma (G)) are very important in de-
velopment and progression of prostate cancer [9].
In initiation and also in progression of the prostate
cancer the IGF network, including INSR (insulin recep-
tor) — (subtypes INSR A and B), IGF1R and IGF2R
plays an important role [10–12].
Both estrogen receptors, alpha (ESRα, ESR1)
and beta (ESRβ, ESR2) are associated with deve-
lopment of prostate cancer [13]. It was shown, that
the increased expression of ESRα is observed upon
progression, metastasizing, and in androgen resistant
phenotype; ESRα could be involved in regulation of ex-
pression of the TMPRSS2-ERG fusion [14].
PRL (prolactin) can induce growth and survival
of prostate cancer cells [15]. The PRL expression cor-
relates with the disease severity.
It was shown that vitamin D (calcitriol) influences
on prostate cancer cells growth [16]. Furthermore, the
TMPRSS2-ERG fusion expression is increased upon
activation of VDR and AR. Consequently, expression
of TMPRSS2/ERG leads to inactivation of the VDR
signaling [17].
We have shown earlier that several genes, regu-
lating the epithelial-to-mesenchymal cell transi-
tion (EMT), such as CDH1, CDH2, NKX3-1, FN1 and
VIM, are expressed differently in prostate tumors [18].
In a present work, we aimed to analyze the expres-
sion pattern of a group of the cancer-related genes, de-
pending on the presence or absence of the TMPRSS2/
ERG fusion in prostate tumors. Also, we wanted to find
the putative correlations between gene expression pat-
terns and clinical and pathological characteristics (CPC)
to define the molecular subtypes of prostate cancer.
MATERIALS AND METHODS
A collection of prostate tissues samples.
Samples of cancer tissue and CNT (at an opposite side
of tumor) were frozen in liquid nitrogen immediately
after surgical resection at the National Cancer Insti-
tute (Kyiv, Ukraine). Benign prostate tumors (prostate
adenoma samples) were collected at the Institute
of Urology (Kyiv, Ukraine) after radical prostatectomy
and frozen, as described above. The samples were
collected in accordance with the Declaration of Hel-
sinki and the guidelines issued by the Ethic Committee
of the Institute of Urology, the National Cancer Institute
and an Ethic Committee of the Institute of Molecular
Biology and Genetics. Experimental studies were
conducted on 37 prostate adenocarcinoma samples
of different Gleason score and stages; 37 paired CNT
samples; 21 samples of benign prostate tumors (ade-
nomas). Tumor samples were characterized, according
to an International System of Classification of Tumors,
based on the tumor-node-metastasis (TNM) and the
World Health Organization (WHO) criteria. CPC and the
presence/absence of the TMPRSS2/ERG fusion, that
we have detected earlier [3] are presented on Table 1.
Table 1. CPC and TMPRSS2/ERG status (T/ERG) of prostate adenocarci-
noma samples
Sample number T/ERG Stage Gleason score TNM PSA, ng/ml
1 – ІІ < 7 T2bN0M0 12.8
2 – ІІ < 7 T2сNxM0 27.3
3 – ІІІ < 7 T3bNхM0 23.6
4 – ІІ < 7 T2bNxM0 6.5
5 – II < 7 T2cNxM0 25.2
6 + ІІ < 7 T2аNxM0 18.6
7 + ІІ < 7 T2аN0M0 9.3
8 + ІІ < 7 T2aN0M0 6.0
9 + II < 7 T2pN0M0 5.0
10 + ІІ < 7 T2аN0M0 13.3
11 + II < 7 T2cN0M0 29.1
12 – ІІ 7 T2аNxM0 11.7
13 – ІІ 7 T2сNxM0 13.9
14 – ІІ 7 T2сNxM0 19.8
15 + ІІ 7 T2аNxM0 7.1
16 + І 7 T1сNxM0 8.2
17 + ІІ 7 T2сNxM0 19.3
18 + ІІ 7 T2аNхM0 5.6
19 + ІІ 7 T2cN0M0 14.3
20 + ІІІ 7 T2bN0M0 24.6
21 – ІІІ > 7 T3bNхM0 86.3
22 – ІV > 7 T3aN0M1 37.8
23 – IV > 7 T2сN0M1 22.6
24 – ІІІ > 7 T2сN1M0 2.3
25 – ІІ > 7 T2bNxM0 6.9
26 – III > 7 T3bNxM0 51.0
27 – ІІІ > 7 T2bNxM0 0.5
28 – ІІ > 7 T2bN0M0 20.3
29 + ІІ > 7 T2cN0M0 9.7
30 + IІІ > 7 T3bN0M0 12.1
31 + III > 7 T3aN0M0 25.1
32 + ІІІ > 7 T3bNхM0 16.0
33 + ІІI > 7 T3bN0M0 84.2
34 + ІІІ > 7 Т3bNхМ0 20.9
35 + IV > 7 T2cN1M0 17.0
36 + ІІ > 7 T2bNxM0 33.0
37 + ІІІ > 7 T3bNxM0 106.0
Note: + presence of TMPRSS2/ERG fusion; − absence of TMPRSS2/ERG
fusion.
Total RNA isolation and cDNA synthesis.
50–70 mg of frozen prostate tissues were mashed
to a powder in the liquid nitrogen. Total RNA was
extracted by TRI- reagent (SIGMA), according to the
manufacturer’s protocol. Total RNA concentration was
analyzed by a spectrophotometer (NanoDrop Tech-
Experimental Oncology 40, 101–108, 2018 (June) 103
nologies Inc., USA). The quality of the total RNA was
determined in a 1% agarose gel by band intensity of 28S
and 18S rRNA (28S/18S ratio). cDNA was synthesized
from 1 µg of the total RNA, that was treated with the
RNase free DNase I (Thermo Fisher Scientific, USA),
using RevertAid H-Minus M-MuLV Reverse Transcrip-
tase (Thermo Fisher Scientific, USA), according to the
manufacturer’s protocol.
Quantitative quantitative polymerase chain re-
action (qPCR). Relative gene expression (RE) levels
of 27 genes (33 transcripts) were detected by qPCR,
using Maxima SYBR Green Master mix (Thermo Fisher
Scientific, USA) and Bio-Rad CFX96 Real-Time PCR
Detection System (USA) under the following condi-
tions: 95 °C — 10 min, following 40 cycles of 95 °C —
15 s, 60 °C — 30 s, elongation 72 °C — 30 s. Primers
for the different transcripts of INSR and IGF1R and
various isoforms of GCR were as published earlier [9,
19]. Primers for others genes were selected, using
qPrimerDepot (https://primerdepot.nci.nih.gov/).
Four reference genes — TBP, HPRT, ALAS1 and
TUBA1B — were used for normalization of the gene
expression [20]. The two main models (2-ΔCT and 2-ΔΔCT
methods), described earlier [18, 21], were used for the
RE level calculation and analysis.
Statistical analysis. The Kolmogorov — Smirnov
test was used to analyze the normality of distribution.
The Kruskal — Wallis test was used to determine
differences by multiple comparison between experi-
mental groups. The Wilcoxon Matched Pairs test was
performed to compare RE in prostate adenocarcinoma
and paired CNT. RE fold differences in 2-ΔΔCT model
were considered significant when expression changes
were more, than 2 fold. The Fisher exact test was per-
formed to monitor differences between these sample
groups [21]. The Benjamini — Hochberg procedure
with false discovery rate (FDR) 0.10–0.25 was used
when multiple comparisons were performed [22].
The Dunn — Bonferroni post hoc test was performed
to determine RE differences between pairs of prostate
samples. The Spearman’s rank correlation test was
used to find the putative correlations between RE and
CPC of prostate tumors and also correlations between
RE of investigated genes. The K-Mean clustering was
applied for prostate cancer subtyping and also for
the specific gene expression profiles, following by the
Kruskal — Wallis and Dunn — Bonferroni post hoc tests
for detection of inter-cluster differences in RE.
RESULTS
Expression of 17 transcripts (11 genes), represent-
ing the receptors and metabolic enzymes and also
16 EMT-related transcripts/genes (3 from them are
lncRNAs) were studied in prostate adenocarcinomas,
CNT and adenomas.
Earlier, we have shown that the TMPRSS2/ERG fu-
sion was expressed in 21 out of 37 adenocarcinomas [3].
In this group, in 16 paired CNT the TMPRSS2/ERG fusion
was detected, and 5 CNT did not show this fusion. Thus,
we have 3 groups in a set of the paired adenocarcinomas/
CNT: 1) T–/N– group — the TMPRSS2/ERG fusion was
not detected neither in adenocarcinomas nor in CNT
(n = 16); 2) T+/N+ group — the TMPRSS2/ERG fusion
was found in both, cancer and CNT (n = 16); 3) T+/N–
group — the TMPRSS2/ERG fusion was present in ad-
enocarcinomas, but not in CNT (n = 5).
The Wilcoxon Matched paired test in the 2-ΔCT
model showed the differences in RE of 7 genes, when
the paired adenocarcinoma (T) and CNT (N) were
compared, regardless presence or absence of the
TMPRSS2/ERG fusion (Table 2).
Table 2. RE differences between prostate adenocarcinoma samples
and paired CNT with and without fusion status detection (dependent sam-
pling, 2−ΔCT model)
Gene
Pairs with differences
without fusion status
detection
p-value&
Pairs with dif-
ferences with
fusion status
p-value&
ESR1 T/N 0.010 T+/N+ 0.038
T+/N– 0.043
INSR
(B isof)
T/N 0.037 T–/N– 0.039
KRT18 T/N 0.000 T+/N+ 0.007
MKI67 T/N 0.017 T+/N+ 0.003
MMP2 T/N 0.011 no –
MMP9 T/N 0.014 T+/N+ 0.011
VIM T/N 0.010 no –
HOTAIR T/N 0.007 T–/N– 0.027
PCA3 no – T+/N+ 0.049
Note: &Wilcoxon Matched Pairs test significant with FDR = 0.1.
The following five genes were upregulated in ad-
enocarcinomas, when T/N pairs with the fusion in both,
tumor and CNT were analyzed: ESR1 (p = 0.038),
Table 3. Frequency of RE fold changes (2−ΔΔCT) in prostate adenocarcinoma (T) in comparison with paired CNT (N) in groups with different TMPRSS2/ERG
status and statistical significant differences in paired T/CNT in 2−ΔCT model
Group N RE fold
changes
AR
(1isof)
AR
(2isof) ESR1 ESR2 GCR
(AG isof)
GCR
(in AG)
GCR
(in B)
INSR
(A isof)
INSR
(B isof) IGF1R IGF1R tr PRLR PRL SRD5A1 SRD5A2 VDR
1 T–/
N–
16 < 0.49 1 4 3 2 0 1 0 0 0 1 3 4 4 2 1 3
> 2.10 1 2 7$ 4 0 0 2 2 4# 2 1 1 2 0 0 1
2 T+/
N+
16 < 0.49 3 2 1 3 0 1 1 3 0 3 1 1 0 3 4 2
> 2.10 1 2 9$ # 1 1 1 1 0 1 1 0 1 1 2 1 3
3 T+/
N–
5 < 0.49 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0
> 2.10 0 1 3# 2 0 0 0 0 2 1 0 0 0 1 0 0
Group N RE fold
changes CDH1 CDH2 FN1 VIM OCLN MMP2 MMP9 NKX3-1 PSA KRT18 MKI67 CASP3 XIAP PCA3 HOTAIR SCHLAP1
1 T–/N– 16 < 0.49 4 2 3 1 2 0 1 6 5 2 3 1 0 6 2 6
> 2.10 0 3 3 3 3 1 5 1 3 2 2 0 2 4 9$ # 6
2 T+/N+ 16 < 0.49 5 6 0 0 3 1 0 5 2 0 1 1 0 4 2 3
> 2.10 0 2 2 3 4 3 9$ # 2 1 5# 7$ # 2 0 9$ # 7$ 7$
3 T+/N– 5 < 0.49 1 2 0 0 0 0 1 0 1 0 1 0 0 0 1 0
> 2.10 2 0 0 1 2 0 1 1 1 1 2 0 0 4$ 1 2
Notes: $statistical significant differences between adenocarcinoma and CNT groups by Fisher exact test (p < 0.05) (2–ΔΔCT);
#statistical significant differences between adenocarcinoma and CNT groups by Wilcoxon Matched Pairs test (p < 0.05) (2–ΔCT).
104 Experimental Oncology 40, 101–108, 2018 (June)
KRT18 (p = 0.007), MKI67 (p = 0.003), MMP9 (p = 0.011)
and PCA3 (p = 0.049). In adenocarcinomas without fu-
sion INSR (B isof) (p = 0.039) and HOTAIR (p = 0.027)
were expressed at the higher levels, than in the paired
CNT. Only one gene, ESR1, showed significant chang-
es in RE in adenocarcinomas with the presence of the
fusion, compared with CNT without fusion (p = 0.043).
When the 2-ΔΔCT model was used, we found 6 genes
with significant differences in RE between adenocarci-
nomas and CNT (Table 3). Three genes (MMP9, MKI67,
and SCHLAP1) where expressed at the higher levels
in tumors, compared with CNT (the T+/N+ group)
(p < 0.05), two genes (ESR1 and HOTAIR) have shown
increased RE in T+/N+ and T–/N– groups (p < 0.05).
Only one gene, the PCA3 was significantly increased
in T+/N+ and T+/N– groups (p < 0.05).
Hence, the data obtained by the two abovemen-
tioned models are only partially overlapping. This could
be due to different statistical calculations.
Earlier, we have discussed that CNT isolated from
patients with prostate tumors do not represent the
normal tissue, therefore they can’t be considered
as an adequate control [18]. In order to avoid working
with inadequate controls, adenomas were used as the
control instead. Noteworthy, the TMPRSS2/ERG fusion
was detected in 4 adenomas as well. No differences
in the gene expression patterns were found in these
Table 4. RE differences between pairs of groups with different TMPRSS2/
ERG status
Gene/transcript p-value* Pairs with differences p-value**
AR (2 isof) 0.024 T–/A 0.017
ECR1 < 0.001 T–/A 0.002
T+/A < 0.001
N–/A 0.040
PRLR 0.017 T–/A 0.009
SRD5A2 0.002 T–/A 0.039
T+/A 0.003
N+/A 0.020
KRT18 0.007 T+/A 0.008
MMP9 0.001 T–/A 0.003
N–/A 0.012
OCLN 0.021 no –
VIM 0.045 no –
PCA3 0.001 T+/A 0.001
N+/A 0.001
HOTAIR 0.003 T–/A 0.002
SCHLAP1 0.010 T+/A 0.011
Notes: *Kruskal — Wallis test data significant with FDR = 0.1;
**Dunn — Bonferroni post hoc method for multiple comparisons.
Fig. 1. RE of genes with significant differences between 5 groups with presence (+) or absence (–) of fusion transcript.
*p < 0.05 in comparision with adenomas group (A) (Dunn — Bonferoni post hoc test for multiple comparisions)
AR (2 isof)
Median
25%-75%
Min-Max
T- T+ N- N+ A
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0.009
0.010
0.020
0.030
0.040
0.050
0.060
0.070
0.080
0.090
0.100
RE
, 2
-∆
C T,r
.u
. *
ESR1
T- T+ N- N+ A
0.0050
0.0075
0.0250
0.0500
0.0750
0.2500
0.5000
0.7500
2.5000
RE
, 2
-∆
C T,r
.u
.
*
*
*
PRLR
T- T+ N- N+ A
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
2.00
RE
, 2
-∆
C T,r
.u
.
*
SRD5A2
T- T+ N- N+ A
0.04
0.05
0.06
0.07
0.08
0.090.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.901.00
2.00
3.00
4.00
RE
, 2
-∆
C T,r
.u
.
* **
KRT18
T- T+ N- N+ A
0.50
0.75
2.50
5.00
7.50
25.00
50.00
75.00
RE
, 2
-∆
C T,r
.u
.
*
MMP9
T- T+ N- N+ A
0.05
0.50
5.00
RE
, 2
-∆
C T,r
.u
.
*
*
PCA3
T- T+ N- N+ A
0.0005
0.0050
0.0500
0.5000
5.0000
50.0000
RE
, 2
-∆
C T,r
.u
.
*
*
HOTAIR
T- T+ N- N+ A
5E-5
0.0005
0.0050
0.0500
0.5000
RE
, 2
-∆
C T,r
.u
.
*
SCHLAP1
T- T+ N- N+ A
0.0005
0.0050
0.0500
0.5000
5.0000
RE
, 2
-∆
C T,r
.u
.
*
Experimental Oncology 40, 101–108, 2018 (June) 105
4 adenomas, compared with adenomas without fusion.
For further comparison, only the group of adenomas
without fusion was analyzed (n = 17). Also, CNT samples
without fusion (n = 5) from adenocarcinoma pairs with
the fusion were attributed to total CNT fusion negative
(N–) group after verification of RE differences in CNT
sample groups for all investigated genes.
The Kruskal — Wallis test (with FDR = 0.1) has shown
significant differences of RE in 11 out of 33 trans cripts/
genes between 5 investigated groups (T+, T–, N+, N–
and A–), while the Dunn — Bonferroni post hoc method
of the multiple comparisons has confirmed changes
only for 9 transcripts/genes (Table 4, Fig. 1).
Increased RE levels in the adenocarcinoma
and CNT groups, compared to the adenoma group
was shown for 6 genes 1) ECR1 T–/A– group
(p = 0.002), T+/A– (p < 0.001), N–/A– (p = 0.040);
2) KRT18 T+/A– (p = 0.008); 3) MMP9 T–/A– (p = 0.003),
N–/A– (p = 0.012); 4) PCA3 T+/A– (p = 0.001),
N+/A– (p = 0.001); 5) HOTAIR T–/A– (p = 0.002), and
6) SCHLAP1 T+/A– (p = 0.011). Decreased RE levels
in the adenocarcinoma and CNT groups, compared
to the adenoma group, was detected for 3 genes: 1) AR
(2 isof) T–/A– (p = 0.0172), 2) PRLR T–/A– (p = 0.0088)
and 3) SRD5A2 T–/A– (p = 0.0393), T+/A– (p = 0.0034),
N+/A– (p = 0.0203).
Correlations of CPC with RE levels. The Spear-
man Rank Order Correlations (rs) analysis of CPC
characteristics and RE of a set of genes in prostate
adenocarcinomas has revealed a number of positive
and negative correlations (Table 5A). For example,
there is the reverse correlation between the Gleason
score and RE of ESR2, VDR and SRD5A2: rs = –0.354,
rs = –0.382 (p < 0.05), rs = –0.520 (p < 0.01), respec-
tively. Also, RE of GCR (in AG) and PRL showed the
direct correlations with a tumor stage, and 8 genes —
AR (1 isof), AR (2 isof), INSR (A isof), IGF1R, IGF1R tr,
PRLR, VDR and SRD5A2 showed the negative corre-
lation with the tumor stage. Levels of the PSA in serum
correlate negatively with RE of VDR and SRD5A2.
Correlations of RE levels between investigated
genes. Investigation of RE correlations in prostate
adenocarcinomas have shown 131 significant correla-
tions (from p < 0.0001 to p < 0.05) (Table 5B). Among
them 34 have the highest score rs = |0,524–0,936|
(from p < 0.0001 to p < 0.05). A maximal number
of strong RE correlations showed INSR (A isof) —
7 correlations, AR (1 isof), GCR (in AG), IGF1R,
Table 5. Spearman Rank Order Correlations (rs) of CPC with genes RE (5A) and rs among genes RE in prostate adenocarcinomas (5B)
5A
AR
(1 isof)
AR
(2 isof) ESR1 ESR2 GCR
(AG isof)
GCR
(in AG)
GCR
(in B)
INSR
(A isof)
INSR
(B isof) IGF1R IGF1R tr PRLR PRL VDR SRD5A1 SRD5A2
GL 0.023 0.033 -0.238 -0.354 0.157 0.178 0.073 -0.157 -0.304 -0.146 -0.086 -0.009 0.007 -0.382 0.051 -0.520
Stage -0.381 -0.390 -0.235 0.142 0.261 0.377 -0.087 -0.478 -0.187 -0.441 -0.486 -0.326 0.437 -0.444 -0.005 -0.395
PSA, ng/ml -0.147 -0.097 -0.029 -0.271 -0.001 -0.016 -0.304 -0.178 -0.315 -0.319 -0.317 -0.254 -0.168 -0.409 0.067 -0.461
Age -0.154 -0.069 0.113 -0.011 0.211 0.187 0.357 -0.058 -0.066 -0.004 -0.090 -0.056 -0.052 0.261 -0.156 0.032
5B
Gene/
transcript T/ERG AR
(1 isof)
AR
(2 isof) ESR1 ESR2 GCR
(AG isof)
GCR
(in AG)
GCR
(in B)
INSR
(A isof)
INSR
(B isof) IGF1R IGF1R tr PRLR PRL VDR SRD5A1 SRD5A2
AR (1 isof) 0.331 1.000
AR (2 isof) 0.486 0.665 1.000
ESR1 0.156 0.383 0.291 1.000
ESR2 -0.294 -0.049 -0.096 0.252 1.000
GCR
(AG isof)
-0.118 0.017 0.043 0.461 0.297 1.000
GCR
(in AG)
-0.239 -0.071 -0.025 0.267 0.457 0.855 1.000
GCR (in B) -0.007 0.204 0.322 0.307 0.322 0.539 0.584 1.000
INSR
(A isof)
0.074 0.518 0.405 0.435 0.301 0.060 0.060 0.217 1.000
INSR
(B isof)
-0.367 0.217 0.202 0.445 0.418 0.340 0.388 0.258 0.589 1.000
IGF1R 0.349 0.532 0.488 0.382 0.119 0.071 0.023 0.303 0.751 0.486 1.000
IGF1R tr 0.332 0.560 0.461 0.324 0.043 0.008 -0.051 0.230 0.755 0.450 0.936 1.000
PRLR 0.097 0.464 0.443 0.322 -0.079 0.087 -0.130 0.162 0.462 0.315 0.446 0.458 1.000
PRL -0.314 -0.263 -0.401 -0.293 0.093 -0.053 -0.051 -0.260 -0.391 -0.050 -0.371 -0.325 -0.038 1.000
VDR -0.001 0.291 0.280 0.340 0.241 0.063 0.025 0.247 0.536 0.278 0.248 0.237 0.232 -0.250 1.000
SRD5A1 0.194 -0.033 -0.041 -0.091 0.134 -0.029 0.125 -0.163 0.256 0.079 0.249 0.304 -0.227 -0.136 -0.035 1.000
SRD5A2 -0.222 0.319 0.058 0.328 0.505 -0.038 -0.023 0.080 0.424 0.491 0.422 0.398 0.329 -0.029 0.406 0.091 1.000
CDH1 0.125 0.439 0.374 0.124 0.116 -0.136 -0.153 0.311 0.601 0.296 0.716 0.631 0.367 -0.335 0.304 0.118 0.429
CDH2 -0.358 0.272 0.155 0.118 0.125 0.518 0.530 0.289 0.120 0.311 -0.007 -0.064 0.253 0.156 0.049 -0.138 0.116
CASP3 -0.045 0.492 0.389 0.130 0.104 0.261 0.314 0.340 0.332 0.348 0.248 0.239 0.220 -0.161 0.172 0.174 0.118
FN1 -0.250 0.069 0.099 0.343 0.168 0.499 0.520 0.362 0.039 0.364 -0.104 -0.140 0.001 -0.057 0.070 -0.229 -0.049
KRT18 0.449 -0.064 0.103 -0.313 -0.263 -0.512 -0.527 -0.328 0.063 -0.381 0.220 0.221 -0.051 -0.153 -0.055 0.258 -0.174
OCLN 0.467 0.631 0.516 0.150 -0.177 -0.190 -0.246 0.044 0.634 0.079 0.619 0.619 0.276 -0.268 0.326 0.267 -0.009
MKI67 0.317 0.356 0.430 -0.094 -0.128 -0.248 -0.161 -0.085 0.315 -0.090 0.140 0.186 0.087 -0.129 0.408 0.174 0.032
MMP2 -0.141 -0.053 -0.185 0.149 0.365 0.313 0.374 -0.124 -0.072 0.089 -0.350 -0.305 -0.331 0.056 0.137 0.205 0.205
MMP9 -0.027 -0.278 -0.175 0.462 0.355 0.213 0.099 -0.201 0.136 0.109 0.000 0.014 -0.078 0.014 0.208 0.162 0.211
NKX3-1 0.231 0.540 0.538 0.071 -0.038 -0.104 -0.105 0.366 0.562 0.212 0.632 0.592 0.495 -0.293 0.351 0.064 0.307
PSA 0.338 0.346 0.304 -0.330 -0.434 -0.658 -0.717 -0.307 0.168 -0.305 0.299 0.356 0.202 -0.215 -0.087 0.160 0.037
VIM -0.235 -0.235 -0.444 0.209 0.034 0.268 0.203 -0.316 -0.101 0.226 -0.201 -0.125 -0.164 0.244 -0.116 0.296 0.027
XIAP 0.077 0.572 0.524 0.235 0.101 0.177 0.191 0.427 0.562 0.349 0.416 0.439 0.389 -0.178 0.355 0.064 0.084
PCA3 0.344 0.151 0.380 -0.092 -0.315 -0.341 -0.388 0.042 0.193 -0.111 0.408 0.402 0.095 -0.304 -0.124 0.054 -0.257
HOTAIR -0.290 0.025 0.020 0.133 0.006 0.549 0.469 0.293 -0.053 0.203 -0.136 -0.163 0.300 0.260 0.019 -0.509 -0.157
SCHLAP1 0.338 -0.047 0.071 -0.421 -0.536 -0.477 -0.545 -0.440 -0.169 -0.587 -0.118 -0.098 -0.148 -0.057 -0.258 0.148 -0.445
Notes: p < 0.0001 (dark blue bold type), p < 0.001 (dark blue bold+italic type), p < 0.01 (red bold type), p < 0.05 (red).
106 Experimental Oncology 40, 101–108, 2018 (June)
IGF1R tr — 6 correlations. This big number of correla-
tions confirms robust relationships between gene ex-
pression profiles and the close connections pathways,
where these genes belong to.
Expression profiling of adenocarcinomas.
To determine the putative molecular subtypes of the
prostate adenocarcinomas, showing the certain gene
expression profile, the K-means clustering was per-
formed, with analysis of RE of all of the studied genes
and CPC (Gleason score and tumor stages) in the
adenocarcinoma group. We found three specific clus-
ters (Fig. 2, Table 6), that included 33 out of 37 cancer
samples (89%). These clusters showed the significant
differences in RE of 21 out of 33 transcripts. The larg-
est distance is between clusters 1 and 3. All three
clusters consist of tumors with the various Gleason
scores (6, 7, 9).
The cluster 1 contains 12 samples with the
TMPRSS2/ERG fusion. Also, in this group the highest
expression of AR, epithelial markers (CDH1, NKX3-1,
OCLN) and prostate cancer markers (PSA, PCA3,
KRT18, SCHLAP1) is detected.
The cluster 3 contains the tumors with the highest
Gleason score and a tumor stage index. By other words,
the cluster 3 consists of the most aggressive tumors.
This assumption is supported by the RE data. For exam-
ple, in this group we found the lowest expression of AR,
epithelial markers (CDH1, OCLN, NKX3-1), SRD5A2,
INSR (A and B) and IGF1R, and the high levels of PRL,
lncRNA SCHLAP1 and HOTAIR, and also of mesenchy-
mal markers (VIM, FN1, MMP2). We have to mention,
however, that the cluster 3 contains the lowest number
of samples with the fusion — only 2 out of 8.
The gene expression profile in cluster 2 has a mixed
pattern. For example, several epithelial and luminal
markers, such as KRT18, PCA3 and PSA show the
lowest expression, and other genes, namely mesen-
chymal markers CDH2, MMP2, FN1 and VIM are highly
expressed.
DISCUSSION
The TMPRSS2-ERG fusion transcript iso-
form 2 (EF194202.1) was first detected in prostate
tumor samples by Lapointe et al. [23]. It is known that
formation of this fusion transcript leads to overex-
pression of the ERG protein, which is involved in the
signaling pathways associated with prostate cancer
development [24, 25]. We wanted to enlighten the
influence of this fusion on expression of some prostate
cancer-associated receptors, enzymes and EMT-
associated genes. Thus, in paired adenocarcinoma/
CNT samples we have found the specific changes
in RE in cancers with the fusion for 5 genes, whereas
RE alterations for tumors without fusion were found
only for 2 genes. The high level of ESR1 in tumors
where the fusion was detected was associated with
faster cancer progression [14].
In the present work we found among adenocar-
cinomas, CNT and adenomas that the ESR1 and
SRD5A2 genes showed altered expression regard-
Table 6. Prostate adenocarcinomas RE means of clusters and statistical significant differences between them
Cl
us
te
r
nu
m
be
r
Nu
m
be
r
of
c
as
es
Pe
rc
en
t-
ag
e
(%
)
T/
ER
G
st
at
us
(+
)
St
ag
e
G
le
as
on
sc
or
e
T/
ER
G
AR
(1
is
of
)
AR
(2
is
of
)
CD
H1
NK
X3
-1
KR
T1
8
PS
A
O
CL
N
XI
AP
IG
F1
R
IG
FR
tr
PC
A3
SC
HL
AP
1
CD
H2
FN
1
VI
M
1 12 36.4 12 2 7 0.94 2.10 0.03 4.92 0.82 31.78 537.05 1.76 0.48 4.32 4.33 22.61 0.55 0.09 4.5 9.50
2 13 39.4 4 2 6 0.00 1.57 0.02 3.03 0.34 15.11 166.95 0.51 0.38 3.21 3.26 0.22 0.01 0.32 9.42 13.04
3 8 24.2 2 3 9 0.01 0.97 0.01 1.92 0.21 26.15 299.15 0.42 0.23 1.02 0.94 7.98 0.67 0.14 10.8 14.10
p-value
< 0.05
*,
***
*** *** *** *** * * *,
***
**,
***
**,
***
* *,
**
*
Cl
us
te
r
nu
m
be
r
M
M
P2
M
M
P9
HO
TA
IR
ES
R1
ES
R2
IN
SR
(A
is
of
)
IN
SR
(B
is
of
)
VD
R
SR
D5
A1
SR
D5
A2
G
CR
(A
G
is
of
)
G
CR
(B
is
of
)
G
CR
(in
A
G
)
G
CR
(in
B
)
PR
LR
PR
L
M
KI
67
CA
SP
3
1 8.22 1.18 0.01 0.44 0.03 0.48 0.35 0.16 0.10 0.46 2.49 0.00 2.41 8.45 0.12 0.00 0.25 0.35
2 16.43 2.52 0.03 0.81 0.05 0.46 0.60 0.23 0.07 0.82 3.26 4.73 3.65 2.50 0.12 0.00 0.14 0.38
3 16.15 4.28 0.02 0.22 0.03 0.15 0.23 0.09 0.07 0.25 4.15 0.00 4.88 0.85 0.05 0.01 0.23 0.28
p-value
< 0.05
* ** **,
***
*,
**
**,
***
** *** ***
Notes: *differences between 1 and 2 clusters; **differences between 2 and 3 clusters; ***differences between 1 and 3 clusters (Dunn — Bonferroni post
hoc method for multiple comparisons).
Fig. 2. Prostate adenocarcinomas RE profiling by K-means clustering
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
No
rm
al
ize
d
m
ea
ns
Cluster 1 Cluster 2 Cluster 3
TMPRSS2/ERG
AR (1
iso
f)
AR (2
iso
f)
CDH1
NKX3-1
KRT18
PSA
OCLN
IGF1R
IGFR tr
PCA3
SCHLAP1
CDH2
FN1 VIM
MMP2
MMP9
HOTAIR
ESR1
ESR2
INSR (A
iso
f)
INSR (B
iso
f)
SRD5A1
SRD5A2
GCR (A
G iso
f)
GCR (B
iso
f)
GCR (in
AG)
GCR (in
B)
PRLR
MKI67
CASP3
XIAP
VDR PRL
Experimental Oncology 40, 101–108, 2018 (June) 107
less presence of the TMPRSS2-ERG fusion, while
AR, MMP9 and HOTAIR were affected only in cases
with no fusion, and expression of KRT18, PCA3 and
SCHLAP1 were changed in adenocarcinomas with the
fusion. Noteworthy, in adenomas we have detected
the highest SRD5A2 RE levels. It is known that in-
creased SRD5A2 in adenomas provokes hyperplasia
extension through NF-kB and AR isoform 7 conferring
5α-reductase inhibitors resistance [26]. From other
hand decreased levels of SRD5A2 in adenocarcinomas
is associated with the enhanced cell migration and
invasion [7]. Moreover, when SRD5A2 gene was re-
introduced, cell migration and invasion was inhibited,
due to F-actin reorganization [27].
The high RE of lncRNA SCHLAP1 adenocarcino-
mas with the fusion predict unfavorable prognosis
of disease [28]. The other lncRNA, HOTAIR when
it expressed at the high levels in adenocarcinomas
without fusion enhances proliferation and invasion
at late stages of prostate cancer [29].
Earlier, we could not find a correlation between
frequency of the fusion transcript detection and CPC,
such as the Gleason score and stage [3], therefore
we didn’t analyze TMPRSS2/ERG dependent changes
for investigated genes in sample groups with different
CPC. Now we found many correlations between CPC
and RE of the genes, encoding receptors/enzymes
in total group of adenocarcinomas. Eight out of ten
significant (p < 0.01 to p < 0.05) correlations were
negative, i.e. expression of these genes was decreas-
ing upon cancer progression. Furthermore, large
quantity of the RE correlations of investigated genes
allow us to perform the clustering of patients with ad-
enocarcinomas. We clustered prostate adenocarcino-
mas in three groups, based on the RE of 33 transcripts
and also CPC characteristics.
Our experimental data on the RE profiles in pros-
tate adenocarcinomas are in concordance with the
literature data [1, 2, 4]. It is widely accepted, that high
expression of epithelial and luminal markers is usu-
ally accompanied by low expression of mesenchymal
markers, and that we have observed in cluster 1.
Noteworthy, we showed simultaneous high expression
of the fusion transcript, PCA3 and NKX3-1 in one clus-
ter. It seems, that the fusion transcript and PCA3 do not
influence negatively on expression of the tumor sup-
pression gene NKX3-1 and vice versa, as they belong
to different pathways [1, 4]. At the other hand, the
oncogenic PCA3 pathway [28, 30], probably, acts
in parallel with the ERG pathway [23, 24].
To summarize subtyping data, it is essential to note
specific cluster features. Cluster 1, which contains all
fusion positive adenocarcinomas, has the most char-
acteristic expression profile namely fusion positive
androgen dependent luminal subtype 1. Probably on-
cogenic pathways in this group are ERG and PCA3 [25,
30] with high sensitivity to androgens, prolactin, IGF,
INS stimulation oncogenic signaling.
We suppose that cluster 3 is another luminal
prostate cancer subtype most of all it is fusion
negative with androgen independent and castra-
tion resistant characteristics [30] (fusion negative
androgen independent luminal subtype 2). It has
molecular characteristic properties as the lowest
expression of AR, epithelial markers (CDH1, OCLN,
NKX3-1), SRD5A2, INSR (A and B) and IGF1R, high
levels of mesenchymal markers (VIM, FN1, MMP2)
and lncRNAs SCHLAP1 and HOTAIR. Moreover, in-
creasing RE of HOTAIR may cause the resistance for
enzalutamide [29]. It is unique cluster with the highest
PRL level, which could promote cancer progression
through the PRL/STAT5 signaling pathway [15]. This
is could mean prolactin administration of this cluster
carcinogenesis.
We assume that cluster 2 is mixed stem-like an-
drogen dependent subtype. The lowest expression
of some epithelial and luminal markers KRT18, PCA3,
PSA and high expression for mesenchymal markers
CDH2, MMP2, and tendency to RE growth of FN1, and
VIM are characteristics of stem-like (basal) prostate
cancer, in spite of high AR, CDH1, NKX3-1 RE. The
highest RE levels of ESR1, SRD5A2, INSR B, PRLR
and lncRNA HOTAIR give to this cluster peculiar car-
cinogenic property.
CONCLUSIONS
We have analyzed RE of 33 transcripts from
27 genes to find alterations in prostate tumors, de-
pending on the presence or absence of the TMPRSS2/
ERG fusion. The significant differences of RE
(p < 0.05) for 7 genes were detected, when compared
adenocarcinomas and corresponding CNTs, using the
2-ΔCT model. Five genes (ESR1, KRT18, MKI67, MMP9,
PCA3) showed differential expression, when the paired
samples with compared that were bearing the fusion;
and only two genes (INSR (B isof) and HOTAIR) —
when samples did not expressed the fusion product.
When the 2-ΔΔCT model was used, the number of the
differentially expressed genes were six (MMP9, MKI67,
PCA3, SCHLAP1) and two (ESR1, HOTAIR) when the
tissues expressed the fusion or regardless the pres-
ence of the fusion, respectively.
When adenomas, CNT and adenocarcinomas were
compared, the KRT18, PCA3 and SCHLAP1 genes
showed significant differences in RE in adenocarcino-
mas with the fusion. In adenocarcinomas without the
fusion, such properties were shown by the AR (2 isof),
MMP9, PRLR the HOTAIR genes. The ESR1 and
SRD5A2 gene expression was altered in both types
of adenocarcinomas.
Using the statistical analysis, we created three
clusters of adenocarcinomas, based on gene RE and
CPC characteristic. One of clusters was represented
by adenocarcinomas with the TMPRSS2/ERG fusion.
Further experiments are needed to confirm these data
in a larger patient cohort.
REFERENCES
1. Zhou Y, Bolton E, Jones J. Androgens and androgen
receptor signaling in prostate tumorigenesis. J Mol Endocrinol
2015; 54: 15–29.
108 Experimental Oncology 40, 101–108, 2018 (June)
2. Cancer Genome Atlas Research Network. The Mo-
lecular Taxonomy of Primary Prostate Cancer. Cell 2015;
163: 1011–25.
3. Mevs L, Gerashchenko G, Rosenberg E, et al. Detection
of prostate specific ETS fusion transcripts in cancer samples.
Biopolym Cell 2017; 33: 256–67.
4. Shtivelman E, Beer T, Evans C. Molecular pathways
and targets in prostate cancer. Oncotarget 2014; 5: 7217–59.
5. Zeng R, Liu Z, Sun Y, et al. Differential expression
and function of AR isoforms in prostate cancer. Oncol Rep
2012; 27: 492–8.
6. Lin C, Yang L, Tanasa B, et al. Nuclear receptor-
induced chromosomal proximity and DNA breaks underlie
specific translocations in cancer. Cell 2009; 139: 1069–83.
7. Audet-Walsh Е, Yee T, Tam I, et al. Inverse regula-
tion of DHT synthesis enzymes 5α-reductase types 1 and
2 by the androgen receptor in prostate cancer. Endocrinol
2017; 158: 1015–21.
8. Isikbay M, Otto K, Kregel S, et al. Glucocorticoid re-
ceptor activity contributes to resistance to androgen-targeted
therapy in prostate cancer. Horm Cancer 2014; 5: 72–89.
9. Kach J, Long T, Selman P, et al. Selective glucocorticoid
receptor modulators (SGRMs) delay castrate-resistant prostate
cancer growth. Mol Cancer Ther 2017; 16: 1680–92.
10. Sharon M, Pozniak Y, Geiger T, et al. TMPRSS2-ERG
fusion protein regulates insulin-like growth factor-1 receptor
(IGF1R) gene expression in prostate cancer: involvement
of transcription factor Sp1. Oncotarget 2016; 7: 51375–92.
11. Samani A, Yakar S, Le Roith D, et al. The role of the
IGF system in cancer growth and metastasis: overview and
recent insights. Endocr Rev 2007; 28: 20–47.
12. Heidegger I, Kern J, Ofer P, et al. Oncogenic functions
of IGF1R and INSR in prostate cancer include enhanced
tumor growth, cell migration and angiogenesis. Oncotarget
2014; 5: 2723–35.
13. Jia M, Dahlman-Wright K, Gustafsson J. Estrogen
receptor alpha and beta in health and disease. Best Pract Res
Clin Endocrinol Metab 2015; 29: 557–68.
14. Setlur S, Mertz K, Hoshida Y, et al. Estrogen-depen-
dent signaling in a molecularly distinct subclass of aggressive
prostate cancer. J Natl Cancer Inst 2008; 100: 815–25.
15. Sackmann-Sala L, Goffin V. Prolactin-induced
prostate tumorigenesis. Adv Exp Med Biol 2015; 846: 221–42
16. Kambale P, Haldar D, Kabi B, et al. Study of vitamin
D receptor gene polymorphism (FokI, TaqI and ApaI) among
prostate cancer patients in North India. J Clin Diagn Res 2017;
11: BC05–8.
17. Roberts J, Martin R, Piyarathna D, et al. Vita-
min D receptor activation reduces VCaP xenograft tumor
growth and counteracts ERG activity despite induction
of TMPRSS2:ERG. Oncotarget 2017; 8: 44447–64.
18. Gerashchenko G, Mankovska O, Dmitriev A, et al.
Epithelial-mesenchymal transition related gene expression
in prostate tumours. Biopolym Cell 2017; 33: 335–55.
19. Flannery C, Rowzee A, Choe G, et al. Development
of a quantitative PCR assay for detection of human insulin-like
growth factor receptor and insulin receptor isoforms. Endo-
crinol 2016; 157: 1702–8.
20. Schmidt U, Fuessel S, Koch R. Quantitative multi-
gene expression profiling of primary prostate cancer. Prostate
2006; 66: 1521–34.
21. Rosenberg E, Gerashchenko G, Hryshchenko N, et al.
Expression of cancer-associated genes in prostate tumors. Exp
Oncol 2017; 39: 131–7.
22. Benjamini Y, Hochberg Y. Controlling the false discov-
ery rate: a practical and powerful approach to multiple testing.
J Royal Statist Society 1995; 57: 289–300.
23. Lapointe J, Kim Y, Miller M, et al. A variant TMPRSS2 iso-
form and ERG fusion product in prostate cancer with implications
for molecular diagnosis. Mod Pathol 2007; 20: 467–73.
24. Luo J, Liu S, Zuo Z, et al. Discovery and classification
of fusion transcripts in prostate cancer and normal prostate
tissue. Amer J Pathol 2015; 185: 1834–45.
25. Feng F, Brenner C, Hussain M, et al. Molecular
pathways: targeting ETS gene fusions in cancer. Clin Cancer
Res 2014; 20: 4442–8.
26. Austin D, Strand D, Love H, et al. NF-κB and andro-
gen receptor variant 7 induce expression of SRD5A isoforms
and confer 5ARI resistance. Prostate 2016; 76: 1004–18.
27. Aggarwal S, Singh M, Kumar A, et al. SRD5A2 gene
expression inhibits cell migration and invasion in prostate
cancer cell line via F-actin reorganization. Mol Cell Biochem
2015; 408: 15–23.
28. Smolle M, Bauernhofer T, Pummer K, et al. Current
insights into long non-coding RNAs (LncRNAs) in prostate
cancer. Int J Mol Sci 2017; 22: E473–90.
29. Zhang A, Zhao J, Kim J, et al. LncRNA HOTAIR
enhances the androgen-receptor-mediated transcriptional
program and drives castration-resistant prostate cancer. Cell
Rep 2015; 13: 209–21.
30. Teixeira A, Marchit S, Emmanuel Dias-Neto E, et al.
Going viral? Linking the etiology of human prostate cancer
to the PCA3 long noncoding RNA and oncogenic viruses.
EMBO Mol Med 2017; 9: 1327–30.
Copyright © Experimental Oncology, 2018
|