Gynecologic Oncology 122 (2011) 648–655
Contents lists available at ScienceDirect
Gynecologic Oncology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / y g y n o
Differential expression patterns of CXCR3 variants and corresponding CXC chemokines in clear cell ovarian cancers and endometriosis Mitsuko Furuya a,⁎, Tomona Yoneyama a, Etsuko Miyagi b, Reiko Tanaka c, Kiyotaka Nagahama a, Yohei Miyagi d, Yoji Nagashima a, Fumiki Hirahara b, Yoshiaki Inayama e, Ichiro Aoki a a
Department of Pathology, Yokohama City University Graduate School of Medicine, Yokohama 236–0004, Japan Department of Gynecology, Yokohama City University Graduate School of Medicine, Yokohama 236–0004, Japan Medical Mycology Research Center, Chiba University, Chiba 260–8673, Japan d Molecular Pathology and Genetics Division, Kanagawa Cancer Center, Yokohama, 241–0815, Japan e Department of Pathology, Yokohama City University Hospital, Yokohama 236–0004, Japan b c
a r t i c l e
i n f o
Article history: Received 15 March 2011 Accepted 27 May 2011 Keywords: Ovarian cancer Endometriosis Chemokine CXCR3 variants Tumor microenvironment
a b s t r a c t Objectives. Carcinogenesis of the ovary is often associated with endometriosis. We previously demonstrated that antitumor chemokine receptor CXCR3 was upregulated both in endometriosis and ovarian cancers. Currently, little is known about the roles of CXCR3 variants in these ovarian diseases. In this study, we investigated the expression of CXCR3 variants and their corresponding ligands in endometriosis and ovarian cancers. Methods. The expression patterns of CXCR3 variants (CXCR3A, CXCR3B and CXCR3-alt) and their corresponding ligands were investigated by quantitative RT-PCR, Western blot and in situ hybridization in normal ovaries (n = 16), endometriosis (n = 12), and clear cell ovarian cancers (n = 22) including endometriosis-coexisting cases (n= 11). Results. Sequence analysis of purified RT-PCR products confirmed the presence of three CXCR3 variants in human ovaries. Quantitative RT-PCR analysis revealed differential expression patterns of these variants depending on conditions. CXCR3A was upregulated both in endometriosis and cancers. On the other hand, CXCR3-alt was upregulated and CXCR3B was downregulated in cancers compared with endometriosis. The corresponding ligand CXCL11 was upregulated only in the cancers with elevated CXCR3-alt. Another ligand CXCL4 was downregulated in the cancers with suppressed CXCR3B. In situ hybridization demonstrated preferential expression of CXCR3A in cancer cells and infiltrating lymphocytes. CXCR3B and CXCR3-alt were detectable mainly in microvessels. Conclusions. Collective data suggest that differential expression patterns of CXC chemokines and CXCR3 variants are involved in specific inflammatory microenvironment of ovarian cancers. Altered balance of CXCR3 variants may become helpful information for better understanding of the pathogenesis of ovarian cancers arising from endometriosis. © 2011 Elsevier Inc. All rights reserved.
Introduction Ovarian cancers show the highest fatality among malignancies of the female genital tract. Several molecular events are thought to occur in carcinogenesis of the ovary in a stepwise fashion such as germline mutation of BRCA gene and loss of heterozygosity (LOH) of several chromosome arms. Apart from genetic events that promote de novo cancers, the development of some types of ovarian cancers is closely
⁎ Corresponding author at: Department of Pathology, Yokohama City University Graduate School of Medicine, 3-9 Fuku-ura, Kanazawa-ku, Yokohama 236-0004, Japan. Fax: +81 45 786 0191. E-mail address:
[email protected] (M. Furuya). 0090-8258/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2011.05.034
associated with pre-malignant conditions, such as serous cystadenomas for low-grade serous cancers [1]. With regard to endometrioid and clear cell cancers, many studies have supported the notion that endometriosis becomes a precursor lesion of these types of malignancies [2–6]. Molecular studies reveal that epithelia of endometriosis gland expand monoclonally [7,8], and that common LOH events occur both in endometriosis and ovarian cancer cells contiguously localized in the same cyst [9]. These studies strongly suggest that endometriosis possesses pre-cancerous potential [10–12]. On the other hand, it should be noted that many endometriosis patients display chronic courses without malignant conversion. Inflammatory microenvironment is expected to play important roles in suppression and promotion of ovarian carcinogenesis, but the difference and similarity of inflammatory milieu of these two conditions are poorly understood. Various cytokines and chemokines are known to be upregulated in the patients with endometriosis as well as those with ovarian cancers
M. Furuya et al. / Gynecologic Oncology 122 (2011) 648–655
[13,14]. Among subsets of chemokines, CXC chemokines work as important regulators of angiogenesis and lymphocytes recruitment [15–17]. CXC chemokines with Glu–Leu–Arg (ELR) motif (ELR +) such as CXCL1, CXCL5 and CXCL8 (interleukin-8 [IL-8]) generally induce endothelial migration and proliferation [18,19]. In contrast, members of interfelon-γ (IFN-γ)-inducible CXC chemokines that lack the ELR motif (ELR −) such as CXCL9 (Mig), CXCL10 (IP-10) and CXCL11 (I-TAC) promote anticancer effects by attracting cytotoxic T lymphocytes [18,20] and inhibiting endothelial migration and proliferation [21]. The main receptor for these ELR − chemokines is CXCR3. This receptor is expressed in various immune cell subsets [22,23], and in the other cell types such as mesangial cells, myofibroblasts and vascular cells [21,24,25]. In addition, CXCR3 is detectable in several types of tumor cells [26–29]. We previously demonstrated that CXCR3 was upregulated in renal and ovarian cancers, and that the expression of CXCL11 was upregulated in these cancer tissues [30,31]. The results indicated that upregulated CXCL11–CXCR3 axis might not exert efficient anticancer properties, or that CXCR3 + cancer cells might disturb host anticancer effects by competing with CXCR3 signaling in immune cells and vascular cells. Biochemical studies have revealed that there are at least three CXCR3 variants; CXCR3A, CXCR3B and CXCR3-alt. CXCR3A represents classical CXCR3 that plays pivotal roles in IFN-γ-inducible immune responses. CXCR3B is reported to induce apoptosis and inhibits cellular proliferation [32]. A few recent studies suggest that CXCR3B may have anticancer properties [33,34]. At present, however, the expression balances of CXCR3A, CXCR3B and CXCR3-alt in human tissues are not well understood. In the present study, we investigated these three CXCR3 variants in clear cell type ovarian cancers including endometriosis-associated ovarian cancers (EAOC), endometriosis and normal ovaries. Our data showed downregulated CXCL4–CXCR3B axis and upregulated CXCL11–CXR3A/CXCR3-alt axis in cancers, suggesting that altered expression balances of CXCR3 variants and their corresponding ligands might contribute to specific inflammatory conditions of clear cell cancers arising in endometriosis.
Materials and methods Samples 22 ovarian clear cell cancers from the patients without the history of former therapies, 12 ovarian endometriosis and 16 normal ovaries were obtained from surgical specimens of the patients operated at Yokohama City University Hospital and Kanagawa Cancer Center between 2002 and 2009. Histological typing of the tumors was done based on the WHO classification. According to the histological investigations of clear cell cancers in some studies [6,35], we designated endometriosis-associated ovarian cancer (EAOC) as endometriosis within the tumor or in ipsilateral residual non-tumor ovarian tissue. All the clear cell cancer cases were reviewed by two pathologists (M. F. and Y. M.). Among 22 cancers, 11 cases were confirmed to have co-existing endometriosis in adjacent lesions, and were defined as EAOC. In the other 11 cases, endometriosis lesion was not detectable in the diseased ovary, thus defined as non-EAOC. Written informed consent was obtained from each patient, and the experiment design was approved by both institutional ethics committees. These samples were immediately frozen with liquid nitrogen. The remaining tissues were fixed with 10% formalin.
RNA isolation Total RNAs from the frozen tissues were obtained using RNeasy Mini kit (QIAGEN, Tokyo, Japan) according to the manufacturer's instructions.
649
RT-PCR and real-time RT-PCR The primers of reverse transcriptase-mediated polymerase chain reaction (RT-PCR) to examine the expression levels of human CXCR3 variants are as follows: forward (F) 5′-ACCCAGCAGCCAGAGCACC-3′ and reverse (R) 5′-TCATAGGAAGAGCTGAAGTTCTCCA-3′ for CXCR3A [30], (F) 5′-TGCCAGGCCTTTACACAGC-3′ and (R) 5′-TCGGCG TCATTTAGCACTTG3′ for CXCR3B [32], and (F) 5′-CCAATACAACTTCCC -ACAGGGGT-3′ and (R) 5′-GTCTCAGACCAGGATGAATCCCG-3′ for CXCR3-alt (135 bp). These primer sets were designed to amplify each variant specifically, indicated as black arrowheads in Fig. 1(a). The other primers used in this study are as follows: (F) 5′GCTATAGCCTTGGCTGTGATATTGTG-3′ and (R) 5′-CTGCCACTTTCACTGCTT -TTACC-3′ for CXCL11 [30], (F) 5′-TGAAGAATGGAAGGAAAATTTGC-3′ and (R) 5′-GCAGCTAGTAGCTAACTCTCCAAAAGT-3′ for CXCL4, (F) 5′-CCACCCATGGCA- AATTCC-3′ and (R) 5′-TGATGGGATTTCCATTGATGAC-3′ for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) [30]. Conditions for PCR using fast cycling PCR kit (QIAGEN) were as follows; at 95 °C for 5 min, 35 cycles at 95 °C for 5 s, 60 °C for 5 s, 68 °C for 3 s, with an extension step of 1 min at 72 °C. QuantiTect SYBR Green PCR kit (QIAGEN) was used for real-time RT-PCR analysis according to the method previously described [31]. In brief, a solution of SYBR Green PCR Master Mix (10 μl) and primers (final concentration is 0.3 μM each) were prepared and aliquoted into individual MicroAmp Optical Plate (Applied Biosystems) and cDNA (100 ng — 32 pg/reaction) was added to give a final volume of 20 μl. Conditions for PCR were at 50 °C for 2 min, at 95 °C for 15 min, 40 cycles at 95 °C for 30 s, at 60 °C for 30 s. GAPDH was used as an internal control gene. mRNA levels were expressed as the absolute number of copies normalized against GAPDH mRNA. Difference in amplification was determined using standard curve method. Gel extraction and sequence analysis The PCR products were loaded on 1.5% agarose gel, and they were cut and purified using QIAquick Gel Extraction Kit (QIAGEN) according to the manufacturer's instructions. Primers used for sequence analysis were similar to those used for in situ hybridization. They were designed to detect specific sequence for each variant (indicated as white arrowheads in Fig. 1(a)). After purification, DNA was labeled with Big Dye Terminator v1.1 Cycle Sequencing Kit (Applied Biosystems, Cleveland, OH), and DNA sequencing was done using a sequencer ABI Prism 3100 Genetic Analyzer (Applied Biosystems). In situ hybridization (ISH) For non-radioactive in situ hybridization (ISH), digoxigenin (DIG)labeling sense/anti-sense RNA probes were generated by PCR amplification of each CXCR3 variant products according to the method previously described [36]. The primers for ISH probes are as follows; (F) 5′-ACCCAGCAGCCAGAGCACC-3′ and (R) 5′-CTGCGTAGAAGTTGATGTTGAAGAG-3′ for CXCR3A (440 bp), (F) 5′-TGGGCCGATGGGATCA-3′ and (R) 5′-GGAAGGCCCGGTCGAA-3′ for CXCR3B (392 bp). With regard to the primers for CXCR3-alt probe, PCR was performed using (F) 5′-TTCATGCCACCCAGCTCTAC-3′ and (R) 5′-TGAGGTCTCAGACCAGGATGA-3′. These primers produce both common sequence for CXCR3A/CXCR3B (622 bp) and a specific sequence for CXCR3-alt (285 bp) [25]. The lower band (285 bp) was extracted and purified. After CXCR3-alt sequence was confirmed, TA-cloning was done and a specific probe was generated. Formalin-fixed tissues were sectioned at 10 μm in RNase-free condition. All sections were permeabilized with pepsin, and incubated with RNA probe overnight at 55 °C. After washing twice with 2xSSC and TNE buffer, sections were incubated at 37 °C with RNase mixture (Ambion, Austin, TX). Slides were stringently washed at 55 °C (DAKO, Carpinteria, CA). For signal amplification, a horseradish
650
M. Furuya et al. / Gynecologic Oncology 122 (2011) 648–655
a)
CXCR3A CXCR3A
NH2
5’
3’
COOH
CXCR3B
NH2
CXCR3B
51 aa 5’
3’
COOH
CXCR3-alt CXCR3-alt 5’
3’
NH2 Stop
337 bp deletion Frameshift
b)
79bp
CXCR3A CXCR3B
common to CXCR3A/3B
differential sequences for CXCR3A and CXCR3B 695bp CXCR3A
CXCR3-alt 337 bp deletion
frameshift in CXCR3-alt, resulting in stop codon at 269th amino acid.
1033bp
Fig. 1. Schemas of CXCR3 variants and primers design. (a): N-terminal of CXCR3B mRNA has a longer extracellular domain (blue box) than that of classical CXCR3A mRNA (black box), and the rest sequences are common to both variants (dark gray) [32]. CXCR3-alt misses 337 bp in the region common to CXCR3A and CXCR3B [37]. CXCR3-alt mRNA forms stop codon at the predicted 6th transmembrane region [37]. The primers for RT-PCR are designed to amplify the specific sequences of each variant (black arrowheads). The primers for in situ hybridization probes are designed to amplify longer regions (white arrowheads). (b): Representative results of sequence analysis for each variant using purified RT-PCR products of human ovarian tissues. Refer to GenBank accession no. NM_001504 (CXCR3A) and AF469635 (CXCR3B) for sequence numbers.
peroxidase (HRP)-conjugated rabbit anti-DIG antibody (DAKO) was used to catalyze the deposition of biotin-tyramide (DAKO). Signal was detected with 3,3′-diaminobenzidne. Western blot The mouse monoclonal antibody (mAb) against CXCR3 (clone 1 C6), which recognizes full-size CXCR3 and CXCR3-alt [37], was purchased from BD Biosciences. Mouse mAbs against CXCR3B and β-actin were purchased from Creative BioMart (Shirley, NY), and Sigma (St. Louis, MO), respectively. 12.5 μg proteins were mildly warmed at 37 C° for 1 h and subjected to 12.5% gel electrophoresis. HRP-conjugated goat antimouse IgG was used as the secondary antibody. Statistical analysis Statistical analyses were performed in Statview 5.0. One-way ANOVA was used for statistical evaluation. When a significant difference was seen, post-hoc analysis was performed. Statistical significance was assumed when p b 0.05 was obtained.
Results Expression of CXCR3 variants mRNAs in controls, endometriosis and cancers First, we investigated the sequence of each CXCR3 variant in normal controls, endometriosis and clear cell cancer tissues. Clinical information of the cancer patients is summarized in Table 1. In RT-PCR using the primer set amplifying full-sized CXCR3A, PCR products of CXCR3A and CXCR3-alt were produced, respectively (Supplement Fig. 1(a)). When we used the primer set specific for CXCR3B, full-sized CXCR3B was produced (Supplement Fig. 1(b)). In sequence analysis, we confirmed that the sequences of these gel-extracted PCR products were matched with these CXCR3 variants, respectively, and no specific codon mutation was detected in the current study (Figs. 1(a) and (b)). Next, we investigated expression levels of three CXCR3 variants. Our previous study demonstrated that classical CXCR3 (CXCR3A) was upregulated in endometriosis and cancer tissues [31], but no information has been reported about the expression patterns of CXCR3 variants in ovarian cancers. The quantification analysis of CXCR3 variants by real-time RT-PCR showed that CXCR3-alt mRNAs were upregulated (Fig. 2(a)) and that CXCR3B mRNAs were
M. Furuya et al. / Gynecologic Oncology 122 (2011) 648–655 Table 1 Summary of clinical information. Case
Age
FIGO stage
Endometriosis
Prognosis (months)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
39 54 62 50 42 55 48 49 52 59 47 57 61 57 56 46 56 72 59 59 55 37
IIIc Ia Ic IIc Ia Ic Ic Ia Ia IIIb Ia Ia IIIc IIIb IIIc IIIc Ic IIc Ic Ic IIIc Ia
Yes Yes Yes Yes Yes No Yes No No No Yes Yes No Yes No No Yes No No No No Yes
NED (104) NED (92) UK (12) NED (81) NED (76) NED (73) NED (66) NED (41) NED (35) NED (34) NED (28) NED (79) UK (36) DOD (7) DOD (26) DOD (13) NED (39) NED (38) DOD (4) NED (26) AWD (23) NED (12)
downregulated in clear cell cancers (Fig. 2(b)). The expression levels of these variants were statistically different between cancers and normal controls, but not between cancers and endometriosis (p = 0.2 in CXCR3-alt and p = 0.3 in CXCR3B). CXCR3A mRNAs tended to be upregulated both in endometriosis and cancers (Fig. 2(c)), which supported the result of our previous study [31]. In the current study focused on clear cell cancers, statistical significance was not obtained among three groups in CXCR3A expression.
b)
CXCR3-alt/GAPDH
0.4
p
