Interleukin-30 Expression in Prostate Cancer and Its Draining Lymph Nodes Correlates with Advanced Grade and Stage

Published OnlineFirst November 25, 2013; DOI: 10.1158/1078-0432.CCR-13-2240

Clinical Cancer Human Cancer Biology Research

Serena Di Meo1,2, Irma Airoldi3, Carlo Sorrentino1,2, Alessia Zorzoli3, Silvia Esposito1,2, and Emma Di Carlo1,2

Abstract Purpose: The interleukin (IL)-27 cytokine subunit p28, also called IL-30, has been recognized as a novel immunoregulatory mediator endowed with its own functions. These are currently the subject of discussion in immunology, but completely unexplored in cancer biology. We set out to investigate the role of IL-30 in prostate carcinogenesis and its effects on human prostate cancer (hPCa) cells. Experimental Design: IL-30 expression, as visualized by immunohistochemistry and real-time reverse transcriptase PCR on prostate and draining lymph nodes from 125 patients with prostate cancer, was correlated with clinicopathologic data. IL-30 regulation of hPCa cell viability and expression of selected gene clusters was tested by flow cytometry and PCR array. Results: IL-30, absent in normal prostatic epithelia, was expressed by cancerous epithelia with Gleason 7% of 21.3% of prostate cancer stage I to III and 40.9% of prostate cancer stage IV. IL-30 expression by tumor infiltrating leukocytes (T-ILK) was higher in stage IV that in stage I to III prostate cancer (P ¼ 0.0006) or in control tissue (P ¼ 0.0011). IL-30 expression in prostate draining lymph nodes (LN)-ILK was higher in stage IV than in stage I to III prostate cancer (P ¼ 0.0031) or in control nodes (P ¼ 0.0023). The main IL-30 sources þ þ þ þ were identified as CD68 macrophages, CD33 /CD11b myeloid cells, and CD14 monocytes. In vitro, IL- 30 stimulated proliferation of hPCa cells and also downregulated CCL16/LEC, TNFSF14/LIGHT, chemokine-like factor (CKLF), and particularly CKLF-like MARVEL transmembrane domain containing 3 (CMTM3) and greatly upregulated ChemR23/CMKLR. Conclusions: We provide the first evidence that IL-30 is implicated in prostate cancer progression because (i) its expression by prostate cancer or T- and LN-ILK correlates with advanced disease grade and stage; and (ii) IL-30 exerts protumor activity in hPCa cells. Clin Cancer Res; 20(3); 585–94. 2013 AACR.

Introduction presenting cells, such as dendritic cells (DCs; refs. 3, 4), and Prostate cancer is the second most common cause of is biologically active (5), independent of the other cytokine male cancer-related deaths (1). Mortality for prostate can- receptor–like component, namely Epstein–Barr virus- cer is related to metastatic disease driven by both genetic induced gene 3 (EBI3; refs. 6, 7). Thus, it has been recently and epigenetic alterations and multiple signals delivered recognized as a novel cytokine endowed with its own within the tumor microenvironment, which are critical properties (7–10). However, although the immunoregula- factors in skewing cancer toward dormancy or progression tory functions of IL-27 are fairly well known (11, 12) and (2). Discrimination of molecular pathways driving tumor the mechanisms of its antitumor effects are becoming growth and progression is thus of crucial importance to progressively clear (13, 14), the involvement of IL-30 in identify novel prognostic markers and targets for advanced cancer biology has not been explored, and its biologic treatments. functions are currently a matter of controversy. The IL-27 cytokine subunit p28, also known as IL-30, is a IL-30 has so far been shown to act as a natural antagonist 28 kDa protein that may be secreted by activated antigen- of gp130-mediated signaling in response to IL-6 and IL-27 and thus resulting in anti-inflammatory effects (6), whereas more recently it has been reported to signal via IL-6 a-receptor (IL-6R) by recruiting a gp130 homodimer (15). Authors' Affiliations: 1Department of Medicine and Sciences of Aging, Section of Anatomic Pathology and Molecular Medicine; 2Ce.S.I. Aging IL-6R (gp80) and gp130 are both expressed in human Research Center, "G. d'Annunzio" University Foundation, Chieti; and prostate cancer (hPCa; ref. 16) and increase during prostate 3 Laboratory of Oncology, Istituto Giannina Gaslini, Genova, Italy carcinogenesis (17). Corresponding Author: Emma Di Carlo, Department of Medicine and As we and others have found that prostate cancer usually Sciences of Aging, Section of Anatomic Pathology and Molecular Medi- cine, "G. d'Annunzio" University, "SS Annunziata" Hospital, Via dei Vestini, houses tumor-infiltrating leukocytes (T-ILK) in its stromal 66100, Chieti, Italy. Phone: 39-0871-357395; Fax: 39-0871-540079; compartment (18–20), we first asked whether IL-30 is E-mail: [email protected] expressed in this context, and then assessed its effects doi: 10.1158/1078-0432.CCR-13-2240 in vitro on hPCa cell viability and expression of selected 2013 American Association for Cancer Research. gene clusters.

www.aacrjournals.org 585

Di Meo et al.

The cases were divided into two groups on the bases of Translational Relevance pathologic TNM classification (21): (i) those without (stage Mortality for prostate cancer is related to metastatic I–III; 103 cases) and those (ii) with metastases to the pelvic disease driven by both genetic and epigenetic alterations lymph nodes (stage IV; 22 cases; Table 1). and multiple signals delivered within the tumor micro- Normal prostates were obtained from 12 untreated environment. The specific ways in which the microen- patients ages 54 to 62 following prostatectomy for bladder vironment regulates prostate cancer progression are still cancer (control patients). They were histologically negative poorly understood. In this article, we provide evidence for prostate cancer or benign prostatic hyperplasia. In that a newly discovered cytokine, namely interleukin addition, we obtained pelvic lymph nodes (control lymph (IL)-30, displays cancer-promoting effects in vitro, and nodes) from autopsies of 15 men, ages 51 to 65, who died that endogenous IL-30 expression is tightly linked with for reasons other than cancer and were histologically free advanced prostate cancer grade and stage. Our findings from prostate cancer. reveal this cytokine as a novel molecule shaping the Patients entering the study had not received hormone or tumor and lymph node microenvironment, and hence immunosuppressive treatments or radiotherapy, and were one to be targeted by modern integrated therapeutic free from immune system diseases. Clinicopathologic stages approaches to metastatic disease. were determined according to the seventh edition of the TNM classification of malignant tumors (22). Tumor grade was assessed according to the Gleason scoring system from the prostate biopsies (23). One-half of each normal or neoplastic tissue sample was Materials and Methods fixed in 4% formalin and embedded in paraffin. The other Patients and samples was embedded in Killik frozen section medium (Bio- We collected biologic samples (cancer and normal pros- Optica), snap frozen in liquid nitrogen, and preserved at tate samples, and draining lymph nodes), clinicopathologic 80C. data of 125 patients with prostate cancer, ages 54 to 73, For histology, paraffin-embedded samples were sec- treated by radical prostatectomy for prostate cancer between tioned at 4 mm and stained with hematoxylin and eosin 2009 and 2012 at the S.S. Annunziata Hospital (Chieti, (H&E). Italy). Twenty-two of them were diagnosed with lymph Written informed consent was obtained from patients. node metastasis at surgery. Preoperative androgen depriva- The study has been approved by the Ethical Committee for tion had not been used. Biomedical Research of the Chieti University and Local Prostate cancer samples were graded as Gleason score 5 Health Authority no. 2 Lanciano-Vasto-Chieti in PROT (n ¼ 22), 6 (n ¼ 19), 7 (n ¼ 37), 8 (n ¼ 32), and 9 (n ¼ 15), 1945/09 COET of July 14, 2009, and performed in accor- and staged as pT2, organ-confined cancer [n ¼ 69 (15 dance with the principles outlined in the Declaration of T2aN0M0, 28 T2bN0M0, 21 T2cN0M0, and 5 T2cN1M0)], Helsinki. and pT3, capsular penetration [n ¼ 56 (22 T3aN0M0, 7 T3aN1M0, 17 T3bN0M0, and 10 T3bN1M0)]. Immunohistochemistry The cases were then subdivided into two groups: (i) those For immunohistochemistry (IHC), formalin-fixed, par- with a Gleason score of <7 (41 cases) and (ii) those with a affin-embedded sections were treated with H2O2/3% for 5 Gleason score of 7 (84 cases). minutes to inhibit endogenous peroxidase and then washed

Table 1. IL-30 expression by prostatic epithelia

Prostate cancer

Controls (n ¼ 12) Stage I–III (n ¼ 103) Stage IV (n ¼ 22) Weakly Weakly Weakly Negativea positiveb Positivec Negativea positiveb Positivec Negativea positiveb Positivec 12 81 16 6 13 6 3

NOTE: P < 0.05 Fisher exact test for comparisons between two classes, within the same category (negative, weakly positive, or positive). aStage I to III prostate cancer versus controls ¼ 0.1188; stage IV prostate cancer versus controls ¼ 0.0132; stage IV prostate cancer versus stage I to III prostate cancer ¼ 0.0624. bStage I to III prostate cancer versus controls ¼ 0.2116; stage IV prostate cancer versus controls ¼ 0.0691; stage IV prostate cancer versus stage I to III prostate cancer ¼ 0.2191. cStage I to III prostate cancer versus controls > 0.999; stage IV prostate cancer versus controls ¼ 0.2941; stage IV prostate cancer versus stage I to III prostate cancer ¼ 0.3586.

586 Clin Cancer Res; 20(3) February 1, 2014 Clinical Cancer Research

IL-30 and Prostate Cancer Progression

* in H2O. Antigen was unmasked with heat-induced epitope distinct, when (i) the staining involved >50% 70% of retrieval in EDTA buffer at pH 8. The slices were then held leukocytes and its strength was scarce (), or (ii) the for 20 minutes at room temperature. After washing in PBS/ staining involved 50% of leukocytes and its strength range Tween-20, sections were incubated for 30 minutes with the scarce () to strong (þþ); primary antibody [polyclonal rabbit anti-IL-30 (anti-IL- * scanty, when the staining involved 50% of leukocytes 27p28, catalog: ab118910); Abcam] and immunocom- and its strength was scarce () to absent (). plexes were detected using the Bond Polymer Refine Detec- tion Kit (Leica Biosystems) according to the manufacturer’s Immunostained sections were examined by two pathol- protocol. Negative controls were carried out by replacing ogists with very good agreement (k value ¼ 0.82, 0.75, and the primary antibody with 10% nonimmune serum. 0.79 for evaluations of IL-30 staining in tumors, T-ILK, or LN-ILK, respectively; ref. 25). Double and triple IHC The rate of cancer cell positive for Ki-67, in the primary For double and triple IHC, formalin-fixed, paraffin- tumor or in lymph node metastasis, was assessed as embedded sections were deparaffinized, treated with H2 reported previously (26). The proliferation rate was mea- O2/3% for 5 minutes to inhibit endogenous peroxidase, sured by quantifying the fraction of Ki-67 antigen-positive and then washed in H2O. cells in immunostained tissue sections. The mean fraction Double staining was performed with anti-IL-30 antibody of positive nuclei was estimated, and when one or more in combination with anti-CD11b (clone EP1345Y; Abcam), positive nuclei were present, it was estimated at 1%. For the anti-CD14 (clone 7; Leica Biosystems), anti-CD33 (clone analysis, the Ki-67 was grouped into two categories: 0% to PWS44; Leica Biosystems), or anti-CD68 (clone PG-M1; 5% and >5% (low and high frequency). Dako) antibodies and, triple immunostaining was performed with anti-IL-30 antibody in combination with both Laser-capture microdissection and real-time reverse anti-CD33 and anti-CD11b antibodies as we previously transcriptase PCR reported (24). For laser-capture microdissection (LCM), 10-mm frozen sections from cancer and normal prostate specimens (of both Morphometric analyses control and prostate cancer patients) were mounted on poly- IL-30 expression by primary tumors or lymph node ethylene naphthalate membrane–covered slides (P.A.L.M. metastases was evaluated using the following criteria based Microlaser Technologies), thawed at room temperature, and on (i) the widening of the staining expressed as the per- immersed in cold acetone (5 minutes). Immediately after centage of tumor or metastasis stained, i.e., <50%, 50% H&E staining, sections were used for LCM. Two sections per 70%, and >70%, and (ii) the strength of the staining: sample were analyzed. From 1,000 to 1,500 epithelial cells defined as absent (), slight (), distinct (þ), or strong were cut and catapulted intact into the cap of a laser pres- (þþ). sure catapulting (LPC) microfuge tube (P.A.L.M. Microlaser Thus, IL-30 immunostaining was defined as Technologies), and RNA was immediately isolated with the RNeasy Plus Micro Kit (Qiagen). Tissue sections for micro- * positive, when (i) the widening was >70% and its strength dissection of the stroma were labeled with a monoclonal range slight () to strong (þþ), or (ii) the widening was antibody (mAb) that identifies fibroblasts and myofibroblasts >50% 70% and its strength range distinct (þ) to strong (and excludes leukocytes, endothelial, and epithelial cells; (þþ); clone TE-7; Millipore). The stroma was isolated among the * weakly positive, when (i) the widening was >50% 70% glands of low- and high-grade prostate cancer, or in the histo- and its strength was slight (), or (ii) the widening was ¼ logically normal zones far from the prostate cancer foci. 50% and its strength range slight () to strong (þþ); The real-time reverse transcriptase PCR (RT-PCR) was * negative, when the widening was 50% and its strength carried out as reported previously (24). Primers for IL-30 was slight () to absent (). and EBI3 were purchased from Qiagen (product number QT00236250 and QT01014104, respectively), whereas the T-ILK or lymph node (LN)-ILK expression of IL-30 was primers for the housekeeping gene hypoxanthine phosphor- evaluated using the following score based on (i) the per- ibosyltransferase 1 (HPRT1) were designed and synthesized 0 centage of leukocyte expressing the cytokine, i.e., <50%, by Sigma-Aldrich Corporation: HPRT forward 5 -AGACT- 0 0 50% 70%, and >70%, and (ii) the strength of the TTGCTTTCCTTGGTCAGG-3 and HPRT reverse 5 -GTCTG- 0 cytokine staining that was defined as absent (), scarce GCTTATATCCAACACTTCG-3 . The sizes of the amplified (), distinct (þ), or strong (þþ). cDNA fragments were 148 bp for IL-30, 88 bp for EBI3, and Thus, IL-30 expression by T-ILK or LN-ILK was defined as 101 bp for HPRT. The samples were processed in triplicate, and wells without added cDNA served as negative controls. * strong, when (i) the staining involved more than 70% of leukocytes and its strength range scarce () to strong Cell culture, antibodies, reagents, flow cytometry, and (þþ), or (ii) the percentage of positively stained immunocytochemistry leukocytes was >50% 70% and the strength of the The human PC3 (Interlab Cell Line Collection, CBA/IST staining range distinct (þ) to strong (þþ); San Martino), 22Rv1, and LNCaP prostatic carcinoma cell

www.aacrjournals.org Clin Cancer Res; 20(3) February 1, 2014 587

Di Meo et al.

lines (both from the American Type Culture Collection) Statistical analysis were cultured in RPMI-1640 with 10% fetal calf serum (FCS; Differences in IL-30 protein expression between control Seromed-BiochromKG). Cell lines were obtained directly prostates or lymph nodes and stage I to III or stage IV from the above-mentioned cell banks that performed cell prostate cancer and tumor draining lymph nodes were line characterizations by short tandem repeat profile anal- assessed by the Fisher exact test. Between-group differences ysis. PC3, 22Rv1, and LNCaP were passaged in our labo- in the relative expression of IL-30 mRNA, by real-time RT- ratory for fewer than 6 months after resuscitation. PCR, were assessed by one-way ANOVA and the difference Human recombinant (hr) IL-30 (IL-27p28 Recombi- between each pair of means was evaluated with the Tukey nant Protein, catalog: H00246778-P01; Abnova) was honestly significant difference (HSD) test. Differences in used at 100 ng/mL, following titration experiments using proliferating cell percentages between primary cancers and 10 to 200 ng/mL. The expression of gp130 and IL-6Ra correspondent lymph node metastases were assessed by the were analyzed using phycoerythrin-conjugated specific Student t test. The Spearman rank correlation coefficient (r) mAb (R&D Systems). Isotype-matched antibodies of irrel- was used to examine the correlation between IL-30 protein evant specificity (Caltag) were used as controls. Cells were expression and immunohistochemical staining for Ki-67 in run on Gallios flow cytometer (Beckman Coulter), acquir- primary prostate cancer and lymph node metastases. The ing at least 104 events. Data were analyzed using Kaluza SPSS software, version 11.0 (IBM), was used, with P < 0.05 analysis software (Beckman Coulter). For immunocyto- as the significance cutoff. chemical staining on PC3 cells, cytospin slides were fixed in acetone for 10 minutes and then incubated for 30 Results minutes with rabbit anti-IL-30 (Abcam) antibody or IL-30 expression by prostate cancer epithelia correlates mouse anti-EBI3 (clone EL8; Leica Biosystems) antibody with high-grade and advanced-stage prostate cancer and immunocomplexes were detected using the Bond To determine whether IL-30 is expressed in hPCa, we Polymer Refine Detection Kit (Leica Biosystems) accord- first performed IHC with a mAb specific against this ing to the manufacturer’s protocol. Negative controls subunit of IL-27 in large sets of prostate samples from were carried out by replacing the primary antibody with patients who underwent radical prostatectomy for pros- 10% nonimmune serum. tate cancer, at different stages of disease, and from control patients. Cell proliferation IL-30 expression was absent in normal prostatic epi- The human PC3, LNCaP, and 22Rv1 cells were cultured thelia (from both prostate cancer, n ¼ 125; and control for 24, 48, and 72 hours with or without 10 to 200 ng/mL patients, n ¼ 12) and in high-grade prostatic intraepithe- hrIL-30. Cells were incubated with 2 mmol/L carboxy– lial neoplasia, whereas it was detected, ranging positive to fluorescein diacetate succinimidyl ester (CFSE) in RPMI weakly positive, in the cancerous epithelia of 22 of 103 1% FCS for 15 minutes at 37 C, washed in RPMI 10% FCS, prostate cancer stage I to III (21.3%) and 9 of 22 meta- plated, and analyzed by flow cytometry at the above men- static prostate cancer stage IV (40.9%; Table 1 and Fig. tioned time points. 1A). In addition, we analyzed IL-30 mRNA expression levels by real-time RT-PCR and confirmed data obtained PCR array from tissue section immunostainings. IL-30 expression in Total RNA was extracted, using the RNeasy Micro Kit normal prostate epithelium from patients with prostate (Qiagen), from PC3 and 22Rv1 cells cultured overnight cancer was comparable with that found in prostate epi- with 100 ng/mL hrIL-30 or medium alone. Contaminant thelium from control patients. A significant difference genomic DNA was removed by Dnase treatment (Qia- (P ¼ 0.0132) was disclosed by the Fisher exact test in the 2 gen). RNA was retrotranscribed by the RT First Strand expression of IL-30 between control tissues and prostate cDNA Synthesis Kit (SABioscience). Human tumor metas- cancer stage IV because the percentages of IL-30–negative tasis (code #PAHS-028Z) and chemokines and receptors cases were 100% and 59%, respectively. The strength of 2 2 (code #PAHS-022Z) RT PCR Arrays and RT Real-Time IL-30 expression in lymph node metastasis was usually SyBR Green/ROX PCR Mix were from SABioscience. PCR comparable with that observed in the primary tumor (Fig. was done on the ABI Prism 7700 Sequence Detector 1A). All the 29 IL-30–positive prostate cancers were (Applied Biosystems). Gene expression of hrIL-30–trea- graded as Gleason score 7. ted and control samples was analyzed separately in dif- þ ferent PCR array plates. For each plate, results were IL-30 expression by T-ILK, particularly CD68 þ normalized on the median value of a set of housekeeping macrophages and CD33 myeloid cells, correlates with genes. Then, changes in gene expression between hrIL- advanced-stage prostate cancer 30–treated and control samples were calculated using the Analyses of the prostate cancer stromal compartment DDCt formula. Results from hrIL-30–treated and control revealed that IL-30 expression was lacking in malignant samples, performed in duplicate, were pooled and ana- fibroblasts as in the normal counterpart, as assessed by real- lyzed by the software provided by the manufacturer. A time RT-PCR analyses of microdissected prostate cancer significant threshold of 4-fold change in gene expression stroma, whereas IHC clearly localized IL-30 in the T-ILK. corresponded to P < 0.001. Its expression was particularly evident in T-ILK of metastatic

588 Clin Cancer Res; 20(3) February 1, 2014 Clinical Cancer Research

IL-30 and Prostate Cancer Progression

A Normal prostate epithelial compartment Primary PCa stromal compartment LN Metastasis IL-30

B Primary PCa IL-30 CD68

C Primary PCa IL-30 CD33

Figure 1. IL-30 expression in the prostate. A, IHC showed that expression of IL-30 (brown) was absent in normal epithelia, whereas it was evident in prostate þ cancer epithelial and stromal compartments and in lymph node metastasis. B, IL-30 expression (brown) colocalized with CD68 macrophages þ (fuchsia; magnification in the inset), as indicated by arrows in the image to the right of the panel and, in part (C), with CD33 myeloid cells (fuchsia; magnification in the inset), as indicated by arrows in the image to the right of the panel (A, 400; B and C, images to the left, 400; images to the right, 630; insets, 1,000). prostate cancer, as assessed by the Fisher exact test because prevailed in stage I to III prostate cancer (P ¼ 0.0031) and the percentage of cases endowed with a distinct pattern of control tissue (P ¼ 0.0129; 83.3% and 64.0%, respectively, IL-30 staining was significantly higher in stage IV (63.6%) versus 22.7% in stage IV prostate cancer; Table 2). Double than in stage I to III prostate cancer (29.1%; P ¼ 0.0006) or immunostainings revealed that IL-30 production was main- þ þ in control tissue (16.6%; P ¼ 0.0011), whereas the percent- ly attributable to CD68 macrophages and CD33 myeloid age of cases showing a scanty IL-30 staining significantly cells infiltrating the prostatic stroma (Fig. 1B and C).

Table 2. IL-30 expression by prostate inﬁltrating leukocytes

Prostate cancer

Controls (n ¼ 12) Stage I–III (n ¼ 103) Stage IV (n ¼ 22) Scantya Moderateb Strongc Scantya Moderateb Strongc Scantya Moderateb Strongc 10 2 66 30 7 5 14 3

NOTE: P < 0.05 Fisher exact test for comparisons between two classes, within the same category (scanty, moderate, or strong). aStage I to III PCa versus controls ¼ 0.2176; stage IV PCa versus controls ¼ 0.0011; stage IV PCa versus stage I to III PCa ¼ 0.0006. bStage I to III prostate cancer versus controls ¼ 0.5056; stage IV prostate cancer versus controls ¼ 0.0129; stage IV prostate cancer versus stage I to III prostate cancer ¼ 0.0031. cStage I to III prostate cancer versus controls ¼ 0.6084; stage IV prostate cancer versus controls ¼ 0.2941; stage IV prostate cancer versus stage I to III prostate cancer ¼ 0.3790.

www.aacrjournals.org Clin Cancer Res; 20(3) February 1, 2014 589

Di Meo et al.

A Prostate draining LN PCa draining LN Metastatic PCa draining LN IL-30

B Metastatic PCa draining LN IL-30 CD11b CD33

Figure 2. IL-30 expression in the prostate draining lymph nodes. A, IL-30 expression (brown) was scanty in lymph nodes draining normal prostate, scanty to distinct in prostate cancer stage I to III draining lymph nodes, whereas it was strong (one half of the cases) in lymph nodes draining metastatic prostate cancer stage IV, both without (second-last image of the panel) or with (last image of the panel) metastatic lesion. B, double IHC revealed that expression of IL-30 þ þ þ (brown) in lymph node draining metastatic prostate cancer was mostly attributable to CD68 macrophages (fuchsia) and CD14 monocytes (fuchsia). CD33 þ myeloid cells (fuchsia) contribute to this IL-30 production, and CD11b cells (fuchsia), to a lesser extent. All insets show, in brick red staining, a magniﬁcation of IL-30 expressing immune cells. Triple immunostaining (image at the bottom right of the panel) showed IL-30 (brown) colocalization with CD33 (fuchsia), indicated by the arrowhead (brick red staining), and also with CD11b (blue), indicated by arrows and showed in the inset (dark staining). A and B, 400; insets in B, 1,000.

Prostate draining lymph nodes express IL-30 in the * In lymph nodes draining prostate cancer stage IV, IL-30 þ metastatic stage of prostate cancer progression production colocalized with CD68 macrophages, þ þ Assessment of IL-30 production, by IHC, in lymph nodes CD14 monocytes, and CD33 myeloid cells, part of þ draining normal prostate, prostate cancer stage I to III, and which were also CD11b , as assessed by triple metastatic prostate cancer stage IV, revealed as follows: immunostaining (Fig. 2B).

* The production of IL-30, particularly localized in the IL-30 stimulates in vitro proliferation of hPCa cells and lymphatic sinuses, was wider and stronger in lymph its expression in vivo by primary prostate cancer and nodes draining metastatic prostate cancer stage IV, lymph node metastasis is associated with higher cancer both those harboring the metastasis and those simply cell proliferation draining metastatic prostate cancer, than in lymph nodes Because IL-30 expression in prostate cancer microenvi- draining prostate cancer stage I to III or control prostate ronment and, particularly in draining lymph nodes, corre- draining lymph nodes (Fig. 2A) because the percentage of lates with advanced stages of disease, we next try to clarify the lymph nodes endowed with a strong IL-30 staining was mechanisms involved in the supposed tumor promoting significantly higher in the cohort of prostate cancer stage activity of this cytokine, through in vitro experiments with IV (50%) than in that of prostate cancer stage I to III hPCa cell lines. (18.4%; P ¼ 0.0031) or the controls (0%; P ¼ 0.0023). It has been found that in the presence of EBI3 IL-30 binds to Inversely, the percentage of scantly stained lymph nodes a gp130/WSX-1 heterodimer, otherwise it binds to the receptor was higher in the cohort of controls (66.6%; P ¼ 0.0498) complex composed of IL-6R and a gp130 homodimer (15). or prostate cancer stage I to III (66.9%; P ¼ 0.0034) than Therefore, we first assessed the expression of gp130 and IL-6R in that of prostate cancer stage IV (31.8%; Table 3); in hPCa cell lines PC3, LNCaP, and 22Rv1, by flow cytometry.

590 Clin Cancer Res; 20(3) February 1, 2014 Clinical Cancer Research

IL-30 and Prostate Cancer Progression

Table 3. IL-30 expression by lymph node inﬁltrating leukocytes

Prostate cancer

Controls (n ¼ 15) Stage I–III (n ¼ 103) Stage IV (n ¼ 22) Scantya Moderateb Strongc Scantya Moderateb Strongc Scantya Moderateb Strongc 10 5 69 15 19 7 4 11

NOTE: P < 0.05 Fisher exact test for comparisons between two classes, within the same category (scanty, moderate, or strong). aStage I to III prostate cancer versus controls > 0.9999; stage IV prostate cancer versus controls ¼ 0.0498; stage IV prostate cancer versus stage I to III prostate cancer ¼ 0.0034. bStage I to III prostate cancer versus controls ¼ 0.1313; stage IV prostate cancer versus controls ¼ 0.4382; stage IV prostate cancer versus stage I to III prostate cancer ¼ 0.7439. cStage I to III prostate cancer versus controls ¼ 0.1253; stage IV prostate cancer versus controls ¼ 0.0023; stage IV prostate cancer versus stage I to III prostate cancer ¼ 0.0031.

As shown in Fig. 3A, PC3 cells (left) express both gp130 that they expressed IL-30, but were negative for EBI3 (not (top) and IL-6Ra (bottom) at surface level, and hence may shown), which should thus be absent in PC3 cell culture. respond to IL-30. In contrast, the LNCaP (middle) and 22Rv1 We next looked to see whether hrIL-30 regulates PC3 cell cells (right) express IL-6Ra only. Immunocytochemical proliferation, metastasis-related gene expression, and che- assessment of IL-30 and EBI3 expression in PC3 cells revealed mokine/chemokine receptor gene expression.

A PC3 LNCaP 22Rv1 C IL-30 Ki-67 20 60 20 Anti-gp130 50 Isotype 15 15 40

30 10 10

20 5 5 10 Primary PCa 0 0 0 gp130 30 Anti-IL-6Rα 60 60 Isotype 20 40 40

10 LN metastasis 20 20

0 0 0 IL-6Rα 200 B D IL-30 60 25 40 100 Medium 20 30 CCL16 CCRL1 CMTM3 CXCR3 CKLF TNFSF14 40 CXCR5 15 0 20 10 20 10 5 −100 CMTM1 Fold differences Fold 0 0 0 CMKLR1 − CFSE 200 (IL-30/medium treated cells)

Figure 3. Expression of gp130 and IL-6Ra and modulation of cell proliferation and chemokine/chemokine receptor expression in hPCa cell lines by IL-30. A, flow cytometry of gp130 (top) and IL-6Ra (bottom) expression in PC3, LNCaP, and 22Rv1 cell lines. These experiments were performed at least in triplicate. B, flow cytometry assessment of PC3 cell proliferation induced by hrIL-30. CFSE staining in PC3 (left), LNCaP (middle), and 33Rv1 (right) cells cultured for 48 hours in the presence (dark profile) or absence (light gray profile) of 100 ng/mL IL-30. C, cancer cell proliferation, as assessed by Ki-67 staining (fuchsia), was higher in IL-30–positive (brown) prostate cancer (image on the top right) and its lymph node metastasis (image on the bottom right), than in IL-30–negative prostate cancer (image on the top left) and its lymph node metastasis (image on the bottom left). D, gene expression profiling of chemokine/chemokine receptor genes by PCR array in PC3 cells cultured in the presence or absence of hrIL-30. Pooled results SD from two experiments performed in duplicate are shown. Histogram represents fold differences of individual mRNA between PC3 cells cultured in the presence or absence of IL-30. A significant threshold of 4-fold change in gene expression corresponded to P < 0.001. C, 400.

www.aacrjournals.org Clin Cancer Res; 20(3) February 1, 2014 591

Di Meo et al.

PC3, LNCaP, and 22Rv1 cells were cultured in the pres- the tumor metastasis PCR array (not shown) in PC3 nor ence or absence of hrIL-30 (from 10 to 200 ng/mL) for 72 in 22Rv1 cells. hours. An aliquot was harvested every 24 hours for CFSE intracellular staining. These experiments revealed that the optimal concentration of hrIL-30 that induced PC3 cell Discussion proliferation was 100 ng/mL. This event was clearly detected This study provides the first evidence that the newly starting from 48 hours of treatment (Fig. 3B, left), as identified cytokine IL-30 (4, 6, 7), corresponding to the demonstrated by the higher CFSE intensity in untreated IL-27p28 subunit, may be expressed in both the epithelial PC3 cells compared with hrIL-30–treated cells at this and stromal compartments of prostate cancer. In the for- time point. About 50 ng/mL of hrIL-30 induced PC3 cell mer, IL-30 expression is a hallmark of poorly differentiated, proliferation although to a lower extent compared with 100 high-grade prostate cancer and is observed in about 41% of ng/mL (not shown). hrIL-30 did not affect LNCaP and cases that have metastatized to the regional lymph nodes. In 22Rv1 proliferation at the same time points (Fig. 3B, middle the latter, IL-30 is basically lacking in malignant fibroblasts, and left, respectively), as expected considering the lack of as revealed by real-time PCR, whereas it is clearly produced the gp130 receptor in these prostate cancer cell lines. by infiltrating leukocytes in approximately 77% of meta- Because the expression of IL-30 in patents’ prostate sam- static prostate cancer. Endogenous IL-30, irrespective of its ples was particularly frequent in metastatic prostate cancer, cellular source, is thus implicated in tumor progression and we performed double immunostainings with anti-Ki-67 likely conditions tissue-specific "niche" microenvironment and anti-IL-30 antibodies in IL-30–positive primary tumors of cancer stem cell subsets and thus their metastatic poten- and related lymph node metastasis (total n ¼ 9) versus IL- tial (36). This assumption is corroborated by the frequency 30–negative prostate cancer and related metastasis (total n of a strong IL-30 expression in the regional lymph nodes ¼ 13). Cancer cell proliferation was higher in IL-30 expres- from stage IV metastatic prostate cancer when compared sing tumors and metastasis (8 of 9; 89%) than in IL-30– with those from stage I to III prostate cancer or control lacking samples (4 of 13; 31%; r ¼ 0.574, P < 0.005226, by lymph nodes. the Spearman rank correlation coefficient; Fig. 3C). The rate The intriguing finding that leukocyte expression of IL-30 of cancer cell positive for Ki-67 was not significantly differ- in metastasis-free lymph nodes draining metastatic prostate ent between the primary tumor and related metastasis. cancer is comparable or even stronger than in metastasis homing lymph nodes led us to suppose that locally released hrIL-30 regulates the expression of various genes IL-30 paves the way for prostate cancer seeding to regional þ þ encoding chemokines or their receptors in the PC3 line lymph nodes. Indeed, CD68 macrophages, CD14 mono- þ þ To find out whether IL-30 also regulates cancer cell cyte, and CD33 CD11b myeloid cell populations, firmly expression of metastasis-related genes or inflammatory recognized as main actors in tumor promotion (37–39), chemokine/chemokine receptor–related genes that may seem to be the major sources of IL-30 in both the primary drive toward cancer progression, we next performed PCR tumor and the regional lymph node microenvironment. arrays, after coculture with hrIL-30, of PC3 cells responsive The possibility that the availability of EBI3 in the tumor to IL-30, and 22Rv1 cells, as negative control. or lymph node microenvironment allows IL-30 to engage As shown in Fig. 3D, the chemokine/chemokine recep- IL-27R on locally available leukocytes, and thus act like IL- tors PCR array demonstrated that, in PC3 cells, hrIL-30 27, is quite low because our immunohistochemical and downmodulates the expression of C–C chemokine ligand PCR analyses (not shown) have demonstrated that EBI3 is 16 (CCL16), also known as liver-expressed chemokine (7- almost absent in the epithelia of both primary and meta- fold downregulation; ref. 27), tumor necrosis factor ligand static lesions and barely detected in T- or LN-ILK, but far superfamily member 14 (TNFSF14), also known as LIGHT from IL-30. (9.7-fold downregulation; refs. 28, 29), and chemokine-like Gp130 and IL-6Ra expression has been well documented factor (CKLF; 13.4-fold downregulation; ref. 30). Other in prostate cancer epithelia and increases during progres- downregulated genes were those coding for chemokine sion (16, 17), suggesting that endogenous IL-30, via auto- receptors, C–X–C chemokine receptor 5 (CXCR5; 30-fold crine or paracrine signaling, may directly affect prostate downregulation), C–X–C chemokine receptor 3 (CXCR3; cancer cells. We addressed this issue by assessing the via- 31.5-fold downregulation), and C–C chemokine receptor– bility and expression profiles of selected genes in hPCa cell like 1 (CCRL1), also known as CCX–CKR (37-fold down- lines cocultured with hrIL-30. regulation). The most downregulated gene was the tumor PC3 cells are endowed with gp130 and IL-6R. They alone suppressor and androgen corepressor CKLF-like MARVEL respond to IL-30 stimuli with a significant increase of their transmembrane domain containing 3 (CMTM3; 134-fold proliferation and a quite distinctive regulation of specific downregulation; refs. 31–33). chemokine/chemokine receptor genes. IL-30, in fact, was hrIL-30 also upregulated two molecules, CMTM1 unable to affect the expression of canonical metastasis- (34), and chemokine-like receptor 1 (CMKLR1; 146- and related genes. Furthermore, it downregulated the expression 120-fold increase) the multifuctional receptor, also of the chemokine receptor genes CXCR3, CXCR5, and known as chemerinR23 (35). No significant modulation CCRL1, which may favor cancer cell migration (40–42). was observed of the gene expression profile included in Instead, the main effects of IL-30 on prostate cancer cells are

592 Clin Cancer Res; 20(3) February 1, 2014 Clinical Cancer Research

IL-30 and Prostate Cancer Progression

suppression of leukocyte chemoattractant expression and of cancer cells, as they were leukocytes, in response to an dramatic modulation of the expression of multifunctional inflammatory tumor or lymph node chemerin-rich molecules of the CMTM family. microenvironment. In particular, IL-30 significantly downregulated pros- Taken as a whole, our results, by revealing (i) that IL-30 tate cancer cell expression of immunoregulatory media- displays cancer-promoting effects in vitro and (ii) that tors such as CCL16 (27), TNFSF14 (28, 29), and CKLF endogenous IL-30 expression is tightly linked with (30) that may recruit and activate different leukocyte advanced prostate cancer grade and stage, strongly nomi- populations at the tumor site. The most downregulated nate this cytokine as a novel molecule shaping the tumor gene (134-fold) is that coding for CMTM3, which is and lymph node microenvironment and hence one to be physiologically highly expressed in the testes and deeply targeted by modern integrated therapeutic approaches to involved in male reproductive system maturation (31), metastatic disease. inhibits prostate-specific antigen(PSA)expressionand represses androgen receptor (AR) transactivation in Disclosure of Potential Conflicts of Interest LNCaP cells (33). Thus, CMTM3 downregulation may No potential conflicts of interest were disclosed. result in PSA increase and AR transactivation with related boosting of prostate cell proliferation. Moreover, Authors' Contributions CMTM3, which functions as cancer cell growth inhibitor Conception and design: E.D. Carlo Development of methodology: S.D. Meo, I. Airoldi, A. Zorzoli by inducing apoptosis, has been reported to be silenced Acquisition of data (provided animals, acquired and managed patients, by aberrant promoter methylation in many carcinomas provided facilities, etc.): C. Sorrentino, E.D. Carlo Analysis and interpretation of data (e.g., statistical analysis, biosta- (32). This epigenetic phenomenon may constitute the tistics, computational analysis): S.D. Meo, C. Sorrentino, E.D. Carlo mechanism by which IL-30 regulates CMTM3 expression Writing, review, and/or revision of the manuscript: I. Airoldi, E.D. Carlo and eventually boosts prostate cancer cell proliferation. Administrative, technical, or material support (i.e., reporting or orga- nizing data, constructing databases): S. Esposito Basically, two chemokine/chemokine receptor-relat- Study supervision: E.D. Carlo ed genes were highly upregulated and greatly suscept- ible to the effect of IL-30. They code for CMTM1 Grant Support (146-fold increase; ref. 34), whose role is still unclear, This work was supported by grants from the Associazione Italiana Ricerca and CMKLR1/chemR23 (120-fold increase). The latter is Cancro (AIRC; investigator grant no. 13134), Milano, Italy, and the a multifuctional receptor, usually highly expressed by "Umberto Veronesi" Foundation for the Progress of Sciences, Milano, Italy (to E.D. Carlo); and grants from AIRC (investigator grant no. 13018), Ricerca monocyte-derived human macrophages and immature Finalizzata Collaboratore Estero Ministero della Salute (grant no. RF-2010- plasmacytoid DCs (35), leading to their chemerin-medi- 2308270), and from Cinque per mille e Ricerca Corrente, Ministero della Salute (to I. Airoldi). ated migration. It has also been observed on acute The costs of publication of this article were defrayed in part by the monocytic leukemia cells and human glioblastoma cells payment of page charges. This article must therefore be hereby marked (43) to mediate activation of calcium-triggered down- advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. stream signaling after interacting with specific chemerin isoforms. Though its functional role in this context Received August 13, 2013; revised October 22, 2013; accepted November remains to be investigated, it may drive the migration 11, 2013; published OnlineFirst November 25, 2013.

References 1. Thorne H, Willems AJ, Niedermayr E, Hoh IM, Li J, Clouston D , et al. functions of IL-27 and suppresses anti-allogeneic immune responses. Decreased prostate cancer-specific survival of men with BRCA2 Immunology 2009;128:e816–25. mutations from multiple breast cancer families. Cancer Prev Res 8. Dibra D, Cutrera J, Xia X, Kallakury B, Mishra L, Li S. Interleukin-30: a 2011;4:1002–10. novel antiinflammatory cytokine candidate for prevention and treat- 2. Prendergast GC, Jaffee EM. Cancer immunologists and cancer biol- ment of inflammatory cytokine-induced liver injury. Hepatology 2012; ogists: why we didn't talk then but need to now. Cancer Res 2007; 55:1204–14. 67:3500–4. 9. Garbers C, Hermanns HM, Schaper F, Muller-Newen€ G, Grotzinger€ J, 3. Pflanz S, Timans JC, Cheung J, Rosales R, Kanzler H, Gilbert J, et al. IL- Rose-John S, et al. Plasticity and cross-talk of interleukin 6-type 27, a heterodimeric cytokine composed of EBI3 and p28 protein,induces cytokines. Cytokine Growth Factor Rev 2012;23:85–97. proliferation of naive CD4(þ) T cells. Immunity 2002;16:779–90. 10. Liu X, Wang Z, Ye N, Chen Z, Zhou X, Teng X, et al. A protective role of 4. Liu J, Guan X, Ma X. Regulation of IL-27 p28 gene expression in IL-30 via STAT and ERK signaling pathways in macrophage-mediated macrophages through MyD88- and interferon-gamma-mediated path- inflammation. Biochem Biophys Res Commun 2013;435:306–12. ways. J Exp Med 2007;204:141–52. 11. Vignali DA, Kuchroo VK. IL-12 family cytokines: immunological play- 5. Stumhofer JS, Laurence A, Wilson EH, Huang E, Tato CM, Johnson makers. Nat Immunol 2012;13:722–8. LM, et al. Interleukin 27 negatively regulates the development of 12. Hunter CA, Kastelein R. Interleukin-27: balancing protective and path- interleukin 17-producing T helper cells during chronic inflammation ological immunity. Immunity 2012;37:960–9. of the central nervous system. Nat Immunol 2006;7:937–45. 13. Batten M, Ghilardi N. The biology and therapeutic potential of inter- 6. Stumhofer JS, Tait ED, Quinn WJ III, Hosken N, Spudy B, Goenka R, leukin 27. J Mol Med 2007;85:661–72. et al. A role for IL-27p28 as an antagonist of gp130-mediated signaling. 14. Murugaiyan G, Saha B. IL-27 in tumor immunity and immunotherapy. Nat Immunol 2010;11:1119–26. Trends Mol Med 2013;19:108–16. 7. Shimozato O, Sato A, Kawamura K, Chiyo M, Ma G, Li Q, et al. The 15. Garbers C, Spudy B, Aparicio-Siegmund S, Waetzig GH, Sommer J, secreted form of p28 subunit of interleukin (IL)-27 inhibits biological Holscher€ C, et al. An interleukin-6 receptor-dependent molecular

www.aacrjournals.org Clin Cancer Res; 20(3) February 1, 2014 593

Di Meo et al.

switch mediates signal transduction of the IL-27 cytokine subunit p28 30. Han W, Lou Y, Tang J, Zhang Y, Chen Y, Li Y, et al. Molecular cloning (IL-30) via a gp130 protein receptor homodimer. J Biol Chem and characterization of chemokine-like factor 1 (CKLF1), a novel 2013;288:4346–54. human cytokine with unique structure and potential chemotactic 16. Hobisch A, Rogatsch H, Hittmair A, Fuchs D, Bartsch G Jr, Klocker H, activity. Biochem J 2001;357:127–35. et al. Immunohistochemical localization of interleukin-6 and its recep- 31. Zhong J, Wang Y, Qiu X, Mo X, Liu Y, Li T, et al. Characterization and tor in benign, premalignant and malignant prostate tissue. J Pathol expression profile of CMTM3/CKLFSF3. J Biochem Mol Biol 2006; 2000;191:239–44. 39:537–45. 17. Culig Z, Steiner H, Bartsch G, Hobisch A. Interleukin-6 regulation of 32. Wang Y, Li J, Cui Y, Li T, Ng KM, Geng H, et al. CMTM3, located at the prostate cancer cell growth. J Cell Biochem 2005;95:497–505. critical tumor suppressor locus 16q22.1, is silenced by CpG methyl- 18. Di Carlo E, Magnasco S, D'Antuono T, Tenaglia R, Sorrentino C. The ation in carcinomas and inhibits tumor cell growth through inducing prostate-associated lymphoid tissue (PALT) is linked to the expression apoptosis. Cancer Res 2009;69:5194–201. of homing chemokines CXCL13 and CCL21. Prostate 2007;67: 33. Wang Y, Li T, Qiu X, Mo X, Zhang Y, Song Q, et al. CMTM3 can affect 1070–80. the transcription activity of androgen receptor and inhibit the expres- 19. Di Carlo E, D'Antuono T, Pompa P, Giuliani R, Rosini S, Stuppia L, et al. sion level of PSA in LNCaP cells. Biochem Biophys Res Commun 2008; Clin Cancer Res 2009;15:2979–87. 371:54–8. 20. Bostwick DG, de la Roza G, Dundore P, Corica FA, Iczkowski KA. 34. Wang L, Wu C, Zheng Y, Qiu X, Wang L, Fan H , et al. Molecular cloning Intraepithelial and stromal lymphocytes in the normal human prostate. and characterization of chemokine-like factor super family member 1 Prostate 2003;55:187–93. (CKLFSF1), a novel human gene with at least 23 alternative splicing 21. Cheng L, Montironi R, Bostwick DG, Lopez-Beltran A, Berney DM. isoforms in testis tissue. Int J Biochem Cell Biol 2004;36:1492–501. Staging of prostate cancer. Histopathology 2012;60:87–117. 35. Yoshimura T, Oppenheim JJ. Chemokine-like receptor 1 (CMKLR1) 22. Sobin LH, Gospodarowicz MK, Wittekind C. TNM classificazione dei and chemokine (C-C motif) receptor-like 2 (CCRL2); two multifunc- tumori maligni, settima edizione. Milano, Italy: Raffaello Cortina Edi- tional receptors with unusual properties. Exp Cell Res 2011;317: tore; 2009. p. 241–5. 674–84. 23. Young RH, Srigley JR, Amin MB, Ulbright TM, Cubilla AL. Tumors of the 36. Baccelli I, Trumpp A. The evolving concept of cancer and metastasis prostate gland, seminal vesicles, male urethra, and penis. Washington, stem cells. J Cell Biol 2012;198:281–93. DC: Armed Forces Institute of Pathology; 2000. p. 129. 37. Gabrilovich DI, Ostrand-Rosenberg S, Bronte V. Coordinated regula- 24. Sorrentino C, Musiani P, Pompa P, Cipollone G, Di Carlo E. Androgen tion of myeloid cells by tumours. Nat Rev Immunol 2012;12:253–68. deprivation boosts prostatic infiltration of cytotoxic and regulatory T 38. Lu T, Gabrilovich DI. Molecular pathways: tumor-infiltrating myeloid lymphocytes and has no effect on disease-free survival in prostate cells and reactive oxygen species in regulation of tumor microenvi- cancer patients. Clin Cancer Res 2011;17:1571–81. ronment.Clin Cancer Res 2012;18:4877–82. 25. Landis JR, Koch GG. The measurement of observer agreement for 39. Biswas SK, Allavena P, Mantovani A. Tumor-associated macro- categorical data. Biometrics 1977;33:159–74. phages: functional diversity, clinical significance, and open questions. 26. Aaltomaa S, Karj€ a€ V, Lipponen P, Isotalo T, Kankkunen JP, Talja M , Semin Immunopathol 2013;35:585–600. et al. Expression of Ki-67, cyclin D1 and apoptosis markers correlated 40. Murakami T, Kawada K, Iwamoto M, Akagami M, Hida K, Nakanishi Y, with survival in prostate cancer patients treated by radical prostatec- et al. The role of CXCR3 and CXCR4 in colorectal cancer metastasis. tomy. Anticancer Res 2006;26:4873–78. Int J Cancer 2013;132:276–87. 27. Youn BS, Zhang S, Broxmeyer HE, Antol K, Fraser MJ Jr, Hangoc G, 41. El-Haibi CP, Sharma P, Singh R, Gupta P, Taub DD, Singh S, et al. et al. Isolation and characterization of LMC, a novel lymphocyte and Differential G protein subunit expression by prostate cancer cells and monocyte chemoattractant human CC chemokine, with myelosup- their interaction with CXCR5. Mol Cancer 2013;12:64. pressive activity. Biochem Biophys Res Commun 1998;247:217–22. 42. Feng LY, Ou ZL, Wu FY, Shen ZZ, Shao ZM. Involvement of a novel 28. Mauri DN, Ebner R, Montgomery RI, Kochel KD, Cheung TC, Yu GL , chemokine decoy receptor CCX-CKR in breast cancer growth, metas- et al. LIGHT, a new member of the TNF superfamily, and lymphotoxin tasis and patient survival. Clin Cancer Res 2009;15:2962–70. alpha are ligands for herpesvirus entry mediator. Immunity 1998; 43. Yamaguchi Y, Du XY, Zhao L, Morser J, Leung LL. Proteolytic cleavage 8:21–30. of chemerin protein is necessary for activation to the active form, 29. Yu P, Fu YX. Targeting tumors with LIGHT to generate metastasis- Chem157S, which functions as a signaling molecule in glioblastoma. clearing immunity. Cytokine Growth Factor Rev 2008;19:285–94. J Biol Chem 2011;286:39510–9.

594 Clin Cancer Res; 20(3) February 1, 2014 Clinical Cancer Research

Interleukin-30 Expression in Prostate Cancer and Its Draining Lymph Nodes Correlates with Advanced Grade and Stage

Serena Di Meo, Irma Airoldi, Carlo Sorrentino, et al.

Clin Cancer Res 2014;20:585-594. Published OnlineFirst November 25, 2013.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-13-2240

Cited articles This article cites 41 articles, 12 of which you can access for free at: http://clincancerres.aacrjournals.org/content/20/3/585.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/20/3/585.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/20/3/585. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.