Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Feb 2.
Published in final edited form as: Prostate. 2015 May 19;75(12):1312–1321. doi: 10.1002/pros.23011

Metastasis Initiating Cells in Primary Prostate Cancer Tissues From Transurethral Resection of the Prostate (TURP) Predicts Castration-Resistant Progression and Survival of Prostate Cancer Patients

Qinlong Li 1, Quanlin Li 2, Jill Nuccio 3, Chunyan Liu 1, Peng Duan 1, Ruoxiang Wang 1, Lawrence W Jones 3,**, Leland W K Chung 1,4,*, Haiyen E Zhau 1,*
PMCID: PMC4736544  NIHMSID: NIHMS751384  PMID: 25990623

Abstract

BACKGROUND

We previouslyreported that the activation of RANK and c-Met signaling components in both experimental mouse models and human prostate cancer (PC) specimens predicts bone metastatic potential and PC patient survival. This study addresses whether a population of metastasis-initiating cells (MICs) known to express a stronger RANKL, phosphorylated c-Met (p-c-Met), and neuropilin-1 (NRP1) signaling network than bystander or dormant cells (BDCs) can be detected in PC tissues from patients subjected to transurethral resection of the prostate (TURP) for urinary obstruction prior to the diagnosis of PC with or without prior hormonal manipulation, and whether the relative abundance of MICs over BDCs could predict castration-resistant progression and PC patient survival.

METHODS

We employed a multiplexed quantum-dot labeling (mQDL) protocol to detect and quantify MICs and BDCs at the single cell level in TURP tissues obtained from 44 PC patients with documented overall survival and castration resistance status.

RESULTS

PC tissues with a higher number of MICs and an activated RANK signaling network, including increased expression of RANKL, p-c-Met, and NRP1 compared to BDCs, were found to correlate with the development of castration resistance and overall survival.

CONCLUSIONS

The assessment of PC cells with MIC and BDC phenotypes in primary PC tissues from hormone-naïve patients can predict the progression to castration resistance and the overall survival of PC patients.

Keywords: prostate cancer patient survival, metastasis-initiating cells, dormant cells, RANK signaling, multiplexed quantum dot labeling

INTRODUCTION

One in six men in the United States will be diagnosed with prostate cancer (PC), and one in 36 will die from the disease [1]. Distinguishing indolent from aggressive PC remains an unmet medical need. As active surveillance rather than early treatment becomes accepted for managing PC patients, the need for reliable predictors of progression and survival becomes increasingly critical, especially if the use of serial transrectal biopsies is to be minimized. Our laboratory recently reported that metastasis-initiating cells (MICs), defined as a subpopulation of PC cells that can form tumors locally, can also recruit and reprogram neighboring bystander dormant or non-metastatic cancer cells (BDCs) to gain tumorigenic and metastatic competency in live mice via an activated RANK (receptor activator of NF-κB) mediated signal network [2]. Here we tested the hypothesis that the proportion of MICs and BDCs in primary human PC tissues has prognostic value for differentiating indolent from aggressive PC. In previous studies, we found that metastatic PC displayed osteomimicry, adopting a bone-like phenotype and overexpressing many bone-like proteins [3,4] including RANK and its ligand RANKL during disease progression [2,5–11]. We demonstrated that MICs with stem cell, neuroendocrine cell, mesenchymal cell, and osteomimetic properties can recruit and reprogram BDCs to express bone-like proteins and gain metastatic potential to bone and soft tissues in animal models [2]. To gain further insight into the significance of the activated RANKL-mediated cell signaling network for cancer metastasis in the clinical setting, we sought to validate RANK pathway-associated biomarkers in this study and explore the potential use of MICs in a well-documented cohort of clinical specimens from primary PC as predictors for the progression and survival of PC patients. We employed multiplexed quantum-dot labeling (mQDL)-based quantitative histopathology at the single cell level as reported previously by our group [8,9] to analyze the MIC and BDC phenotypes simultaneously in primary PC tissue specimens from patients with known overall survival, to test their predictive utility for lethal PC progression.

MATERIALS AND METHODS

Prostate Cancer Tissue Specimens

This study included patients under the care of Dr. L. W. Jones and his associates at Huntington Medical Research Institutes, Huntington Memorial Hospital, Pasadena, California. Dr. Jones and his colleagues have followed these patients for up to 15 years. The archival patient tissue specimens from Huntington Memorial Hospital were obtained in accordance with the ethical policies and procedures of the Huntington Medical Research Institutional Review Board. The single source of patient tissues reduced variability in sample handling and processing, patient data entry, pathological diagnosis by pathologists from different institutions, and patients’ ethnic backgrounds.

Clinical records were obtained from the Huntington Memorial Hospital Tumor Registry. The tissues were accrued between 1981 and 1991, generally prior to the clinical availability of PSA blood assay. A total of 44 formalin-fixed, paraffin-embedded (FFPE) tissues from patients who underwent transurethral resection of the prostate (TURP) primarily for relief of bladder outlet obstruction, and were subsequently diagnosed with PC, were used for this study under IRB# Pro 00021228. Twelve patients underwent radiation or hormonal therapy, and no patient in this series underwent subsequent radical prostatectomy. Table I summarizes the age, race, pathological Gleason score, status of distant metastasis and castration resistance, overall survival, and treatments prior to TURP for all 44 patients.

TABLE I.

Clinical Characterization of the Prostate Cancer Patients

Case Age (year) Gleason Score Eventa (month) Othersb (month) Metsc CRd Therapy prior to TURP
Radf HTg ORCHh
1 66 9 32 Yes Yes Yes No No
2 77 7 77 Yes NA No No No
3 76 9 61 Yes Yes No No No
4 72 8 116 Yes Yes No No No
5 76 8 19 No Yes No No No
6 69 3 39 Yes Yes No No No
7 85 10 70 NA No No No No
8 79 8 33 Yes Yes No No No
9 82 8 20 NAe NA No No No
10 60 7 81 Yes Yes No No No
11 66 6 73 Yes Yes No No No
12 83 8 11 NA Yes No Yes No
13 61 5 94 Yes Yes No No Yes
14 73 9 62 No Yes No No No
15 56 9 45 Yes Yes No No No
16 80 8 34 No Yes No No No
17 57 7 34 Yes Yes No No No
18 77 7 48 Yes Yes No No No
19 74 8 51 No NA No No No
20 72 5 16 No NA No No No
21 85 8 104 NA Yes No Yes No
22 67 9 25 Yes Yes No Yes No
23 84 7 80 Yes Yes No Yes Yes
24 84 7 74 Yes Yes No Yes No
25 69 6 41 NA No No No No
26 86 7 49 Yes Yes No No No
27 78 5 114 No No No No No
28 88 4 68 NA NA No No No
29 77 6 77 NA No No No No
30 58 7 82 No No No No No
31 69 6 37 No NA No No No
32 68 3 32 No NA No No No
33 77 5 41 Yes No No No No
34 90 10 43 NA Yes No No No
35 82 8 22 NA No No No No
36 74 6 53 No No No No No
37 75 5 23 No No No No No
38 70 6 31 NA NA No No No
39 92 6 19 No No No No No
40 69 9 86 No No Yes No No
41 80 7 53 No Yes No Yes No
42 81 7 206 Yes Yes Yes Yes No
43 72 9 82 Yes Yes No Yes No
44 71 8 114 Yes Yes Yes Yes No
Open in a new tab
a

Patients died from prostate cancer since diagnosed.

b

Patients died from other causes or lost to follow up since diagnosed.

c

Metastasis.

d

Castration resistance.

e

Not available.

f

Radiation therapy.

g

Hormonal therapy.

h

Hormonal therapy by orchiectomy.

Antibodies and Reagents

Goat polyclonal antibody to neuropilin-1, a co-receptor for VEGF, (NRP1, at 1:50), C-19, sc-7239, rabbit polyclonal antibodies to RANK (1:100), H-300, sc-9072 and RANKL (1:100), FL-317, sc-9073 were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Rabbit polyclonal antibody to p-c-Met (pYpYpY1230/1234/1235 at 1:100), 44888G, was obtained from Invitrogen (Carlsbad, CA). All antibodies were diluted in 1% horse serum (Vector Laboratories, Inc., Burlingame, CA) and 20% Biotin Block Reagent (Invitrogen). Secondary antibodies were prepared in a cocktail of biotinylated antibodies to mouse, rabbit, and goat IgG (Vector Laboratories). Phosphate-buffered saline (PBS) and streptavidin-conjugated QDs at 565-, 585-, 605-, and 625 nm wavelengths as 1 μM stock solutions were from Invitrogen. Rinse buffer or PBST (0.4% Triton X-100 in PBS, pH 7.4) was routinely used as a 3 × 5 min rinse. Streptavidin-conjugated QDs (565-, 585-, 605-, and 625 nm) were prepared as 10 nM in PBS with 6% IgG free, protease-free BSA (Jackson Immunoresearch, West Grove, PA). The blocking reagents used were Biotin Block-1% horse serum (Vector Laboratories), 20% Biotin Block Reagent (Invitrogen) in PBS, Avidin Block-2.5% horse serum and 20% Streptavidin Block Reagent (Invitrogen)in PBS.

Multiplexed Quantum-Dot Labeling (mQDL)

To perform mQDL, a single tissue section from each study case was subjected to sequential labeling for multiple biomarkers. Each biomarker labeling started with a streptavidin block, followed by primary antibody reaction and biotinylated secondary antibody incubation, and reaction with specific streptavidin-conjugated QD. De-paraffinized and antigen-retrieved FFPE specimens were incubated in Streptavidin Block Reagent for 1 hr, followed by reaction with primary antibody in Biotin Block at 4°C overnight or at 37°C for 2 hr. After rinsing, specimens were incubated with a secondary antibody cocktail at 37°C for 1 hr and then incubated with the respective streptavidin-QD at 37°C for 1.5 hr. At the end of the incubation, the specimens were subjected to a routine 3 × 5 min rinse and mounted in 4’6-diamidino-2-phenylindole (DAPI) mounting media (Vector Laboratories) for optical imaging.

The reaction sequence of the paired antibody and QD was as follows: 1) anti- RANK antibody at 37°C, 2 hr; biotinylated secondary antibody at 37°C, 1 hr; streptavidin-QD605 at 37°C, 1.5 hr; 2) anti-neuropilin-1 antibody at 4°C, overnight; biotinylated secondary antibody at 37°C, 1 hr; streptavidin-QD625 at 37°C, 1.5 hr; 3) anti p-c-Met antibody at 37°C, 2 hr; biotinylated secondary antibody at 37°C, 1 hr; streptavidin-QD565 at 37°C, 1.5 hr; and 4) anti- RANKL antibody at 4°C overnight; biotinylated secondary antibody at 37°C, 1 hr; streptavidin-QD585 at 37°C, 1.5 hr.

Positive and negative control specimens were subjected to mQDL in parallel with the test tissue slides. ARCaPM, an invasive bone-metastatic human PC cell line established in our laboratory [12,13], was used as a positive control. To mimic FFPE tissues, the cells were prepared as cell pellets, formalin-fixed, and paraffin-embedded. Negative controls were performed by replacing each primary antibody with species- and subtype-matched immunoglobulin prepared at the same dilution and applied to an immediately adjacent tissue section. We performed optical imaging acquisition and deconvolution or unmixing using a CRi multi-spectral camera with built-in Nuance software. The individual QD-labeled protein or cell-based biomarker intensity can be detected and quantified by inForm software (Perkin Elmer, Waltham, MA). Detailed information on mQDL was previously reported [8,9]. Metastasis initiating cells, or MICs, are defined as cells with above-median levels of cell-based expression of RANKL, NRP1, and p-c-Met. Bystander or dormant cells, or BDCs, are defined as having below-median levels of expression of RANKL, NRP1, and p-c-Met.

Statistical Analyses

The primary outcome was overall survival (OS). Variables measured were age, Gleason score, and the cell-based biomarker expression intensity of RANKL, RANK, p-c-Met, and NRP1 in the cytoplasm (C), nucleus (N), and cytoplasm plus nucleus (C +N). Biomarker measurements for each patient were the mean intensity measured from 13 images (range 7–20) taken from different tissue sites at TURP. The average number of cells analyzed was 7,106 with a range between 2,956 and 11,070. Total sample size was 44 cases.

Statistical methods are described in the relevant legends to figures and tables. The Kaplan–Meier method was used to estimate OS and a log-rank test was used to compare groups. Logistic regression was used to examine the association between biomarker intensities and the clinical status of metastasis or the development of castration-resistant prostate cancer (CRPC). The level of significance was set at 5%.

RESULTS

Table I summarizes the age, race, pathological Gleason score, status of distant metastasis and castration resistance, any radiation or hormonal therapy prior to TURP, and overall survival of the 44 patients. The mean age of the patients was 75 years (range 56–92 years). The ethnic background included 39 Caucasian-Americans, one African-American, and four whose race was undetermined. The metastasis status was 20 positive, 14 negative, and 10 data not available. The number of patients who died of other causes or were lost to follow up after diagnosis was 18 and mean survival was 50.7 months (range 19–114 months). The number of events was 26 with a mean overall survival of 63.8 months (range 11–206 months). Castration-resistant PC (CRPC), defined as disease progression documented radiographically in either surgically or chemically castrated patients, was 25 positive, 11 negative, and eight data not available. The number of patients who underwent radiation, hormonal treatment, or orchiectomy was 4, 9, and 2, respectively. We found that neither radiation nor hormonal therapy impacted the statistically significant correlation between biomarker expression and overall patient survival (data not included).

We used correlograms to show pair-wise Pearson correlations between Gleason score and biomarker expression or between two biomarkers (Fig. 1). Although Gleason score did not correlate with the expression of biomarkers, expression of the pathway-associated biomarker pairs correlated significantly. Figure 2 shows higher expression detected by image (Panel A) and labeling intensity (Panel B) of RANK-pathway mediators, RANKL, RANK, phospho-c-Met (p-c-Met) and neuropilin-1(NRP1) in tissues from a PC patient who survived 114 months (A, top panels) compared to a patient who survived 11 months (B, bottom panels) after first diagnosis. We used the Kaplan–Meier method to estimate overall survival and a log-rank test to compare biomarker expressions categorized in the two groups, using the median as a cutoff point. While the overall survival of PC patients does not correlate with patient age or clinical Gleason score, cell-based total RANKL, p-c-Met and NRP1 expression correlated significantly with overall survival of patients with PC (P <0.05) (Fig. 3A). These results are in agreement with our previous report using a different cohort of PC tissues obtained from patients from another institution [9]. A significant correlation was also found when bio-marker expression was analyzed separately in cytoplasm or nucleus (data not included) when compared with combined cytoplasm plus nucleus.

Fig. 1.

Fig. 1

Open in a new tab

Correlogram with pairwise correlations between Gleason score and RANK-mediated biomarker expression or between biomarkers involved in RANK-mediated signal network. The main diagonal has the covariate name. At the horizontal and vertical intersection of each covariate, the Pearson correlation coefficient (center) and the associated P-value (top right corner, **P <0.01, ***P <0.001) are shown. Although Gleason score did not correlate with any biomarkers, there are significant correlations between most of the biomarker pairs.

Fig. 2.

Fig. 2

Open in a new tab

mQDL assessment of differential biomarker protein expression in clinical primary prostate cancer specimens. A. Unmixed mQDL images of RANKL, RANK, p-c-Met, and NRP1 protein expression from tissues of two patients with contrasting survival data. Scale bar, 200 μm. B. Quantification of protein expression of biomarkers using the inForm software.

Fig. 3.

Fig. 3

Open in a new tab

Log-rank test of: A. overall survival by age, Gleason score and biomarker protein expression in cytoplasm (C) plus nucleus (N) categorized in two groups, high and low, using the median as a cutoff point. Although age and Gleason score did not significantly correlate with overall survival, RANKL, p-c-Met, and NRP1 correlated significantly with overall survival. B. Overall survival of prostate cancer patients by number of MICs and DCs.

We further analyzed the association of individual biomarker expression and the clinical status of PC metastasis and development of CRPC. The results of logistic regression showed that RANKL or NRP1 expression alone was significantly associated with the clinical status of CRPC (P-value = 0.0251 and 0.0255, respectively) but not with metastatic status. We next analyzed if the percentage of MICs or BDCs could predict cancer metastasis or the development of CRPC. MICs are defined as single cells with high, or above median, levels of expression of RANKL, NRP1, and p-c-Met. BDCs are defined as low, or below median, levels of expression of the above three biomarkers. These definitions are based on the evaluation of these gene expression profiles in experimental cell models of MICs and BDCs where we observed that MICs can recruit and reprogram BDCs and alter their phenotypes permanently, including conferring metastatic potential to BDCs in mice [2]. For each tissue specimen, we analyzed 3,000–11,000 cells and found a distribution of MICs in each specimen ranging from 0 to 2,000 cells representing 0–60% of total cells and a distribution of BDCs ranging from 100 to 10,000 cells representing 4–99% of total PC epithelial cells analyzed (Table II). We found that a high ratio of MICs to BDCs predicted poor overall survival of PC patients (P = 0.0043 for both; Fig. 3B). Furthermore, the percentage of MICs or BDCs analyzed on their nuclear (N), cytoplasmic (C), or combined (C +N) co-expression of RANKL, p-c-Met, and NRP1 correlated with the development of castration resistance (Table III).

TABLE II.

Distribution and Percent of Prostate Cancer Cells Expressing RANK-Mediated Pathway Biomarkers

Case Na p-c-Met RANKL NRP1 MICsc BDCsd
Numberb Percent Number Percent Number Percent Number Percent Number Percent
1 8,730 33 0.38 3,646 41.76 3,185 36.48 23 0.26 4,182 47.90
2 5,608 52 0.93 1,763 31.44 2,613 46.59 48 0.86 2,683 47.84
3 5,297 140 2.64 963 18.18 1,176 22.20 113 2.13 3,724 70.30
4 6,592 202 3.06 621 9.42 194 2.94 5 0.08 5,646 85.65
5 4,108 350 8.52 654 15.92 80 1.95 8 0.19 3,115 75.83
6 6,427 439 6.83 670 10.42 153 2.38 2 0.03 5,257 81.80
7 5,933 72 1.21 7 0.12 8 0.13 0 0.00 5,851 98.62
8 3,372 1,575 46.71 2,329 69.07 2,757 81.76 1,490 44.19 448 13.29
9 7,336 11 0.15 2,298 31.32 52 0.71 4 0.05 5,035 68.63
10 6,478 471 7.27 6,210 95.86 5,507 85.01 460 7.10 250 3.86
11 6,093 796 13.06 1,326 21.76 1,258 20.65 390 6.40 4,108 67.42
12 7,330 1,352 18.44 4,844 66.08 4,627 63.12 1,164 15.88 1,808 24.67
13 6,137 1,594 25.97 4,417 71.97 4,622 75.31 1,207 19.67 788 12.84
14 6,477 2,873 44.36 4,426 68.33 2,735 42.23 2,083 32.16 1,805 27.87
15 9,639 957 9.93 2,872 29.80 1,143 11.86 510 5.29 6,416 66.56
16 8,163 4,216 51.65 2,441 29.90 520 6.37 314 3.85 3,028 37.09
17 9,038 2,392 26.47 3,576 39.57 1,345 14.88 1,098 12.15 5,019 55.53
18 8,975 292 3.25 42 0.47 33 0.37 9 0.10 8,640 96.27
19 7,717 40 0.52 102 1.32 4 0.05 0 0.00 7,576 98.17
20 9,127 2,208 24.19 1,341 14.69 1,211 13.27 269 2.95 5,648 61.88
21 7,455 150 2.01 935 12.54 1,471 19.73 10 0.13 5,299 71.08
22 6,855 108 1.58 926 13.51 1,226 17.88 17 0.25 5,184 75.62
23 4,325 1,367 31.61 3,889 89.92 2,833 65.50 1,289 29.80 349 8.07
24 7,544 77 1.02 125 1.66 445 5.90 1 0.01 6,969 92.38
25 6,879 284 4.13 211 3.07 110 1.60 7 0.10 6,340 92.16
26 10,408 128 1.23 153 1.47 3 0.03 0 0.00 10,152 97.54
27 9,720 331 3.41 758 7.80 313 3.22 201 2.07 8,879 91.35
28 8,808 11 0.12 234 2.66 42 0.48 0 0.00 8,541 96.97
29 6,628 660 9.96 212 3.20 1 0.02 0 0.00 5,910 89.17
30 8,639 102 1.18 185 2.14 459 5.31 14 0.16 8,006 92.67
31 4,205 4 0.10 258 6.14 43 1.02 1 0.02 3,918 93.17
32 2,956 2,191 74.12 2,588 87.55 2,380 80.51 1,817 61.47 134 4.53
33 6,042 3,123 51.69 5,287 87.50 2,886 47.77 2,236 37.01 531 8.79
34 6,760 272 4.02 201 2.97 38 0.56 4 0.06 6,283 92.94
35 8,013 927 11.57 70 0.87 71 0.89 30 0.37 7,046 87.93
36 8,860 2,032 22.93 316 3.57 116 1.31 3 0.03 6,469 73.01
37 9,403 164 1.74 5 0.05 5 0.05 0 0.00 9,229 98.15
38 7,205 150 2.08 3 0.04 66 0.92 0 0.00 6,989 97.00
39 6,905 18 0.26 497 7.20 65 0.94 2 0.03 6,364 92.17
40 3,835 2 0.05 439 11.45 194 5.06 0 0.00 3,233 84.30
41 11,074 4,727 42.69 1,401 12.65 887 8.01 183 1.65 5,338 48.20
42 9,348 27 0.29 66 0.71 8 0.09 4 0.04 9,267 99.13
43 6,815 2,286 33.54 5,694 83.55 3,377 49.55 1,822 26.74 1,020 14.97
44 7,034 97 1.38 368 5.23 110 1.56 5 0.07 6,519 92.68
Open in a new tab
a

Total number of cells analyzed.

b

Number of cells positive for biomarker.

c

Metastasis initiating cells, high in RANKL, p-c-Met and NRP1.

d

Bystander dormant cells, low to none in RANKL, p-c-Met and NRP1.

TABLE III.

Association of Percent MICs and BDCs With the Status of Castration Resistance in PC Patients

Positivea Negativeb P-values
Mean STD Mean STD
MICs (N) 7 11 3 8 0.0267
BDCs (N) 57 31 79 25 0.0323
MICs (C) 8 12 4 13 0.0274
BDCs (C) 61 32 87 25 0.0152
MICs (C+N) 8 13 4 11 0.0278
BDCs (C+N) 58 32 83 25 0.0235
Open in a new tab

Castration resistance status: patients with (a) or without (b) castration resistance.

MICs and BDCs analyzed by biomarker subcellular localization: N, nucleus; C, cytoplasm (see Materials and Methods).

DISCUSSION

We showed that the signaling pathway involving RANKL-RANK, mediated by downstream p-c-MET and NPR1, can predict the overall survival of PC patients [8,9]. The present study used TURP tissues without reference to the histopathologic grading of the primary cancer to predict the lethal progression of PC by analyzing the RANKL, p-c-MET, and NPR1 molecular markers found to contribute to prostate cancer bone and visceral organ metastases in animal models. In addition to evaluating the RANKL–RANK-mediated signaling pathway, we also assessed the relative abundance of MICs versus BDCs, shown to contribute to skeleton and soft tissue metastases upon intracardiac inoculation in animal models. Analyzing the number of MICs and BDCs, we found that a higher abundance or proportion of MICs to BDCs predicted not only survival but also the development of CRPC. Increased RANK–RANKL signaling in PC cells could enhance normal bone turnover [14–17], stem cell renewal, and differentiation [16,18,19] and these factors together contribute to prostate and breast cancer bone and soft tissue metastases [5,10,20–23]. The concurrence of these two sets of data strongly supports the conclusion that MICs, which over-express the RANK–RANKL-p-c-MET and NPR1 signaling pathway, predict the development of castration resistance and the survival of PC patients.

Correlogram analyses of a statistical relationship among the downstream signaling components, suggest that these signaling components are highly related in a statistically significant manner to each other as predictors distinguishing indolent from aggressive PC. There are several underlying reasons supporting the concept that the RANK–RANKL signaling pathway plays pivotal roles conferring aggressive human PC. 1) RANKL–RANK downstream signaling components are directly linked to a large number of transcription factors that regulate over 1,644 genes broadly classified to confer osteomimicry, epithelial-to-mesenchymal transition, stem cell and neuroendocrine cell phenotypes, and cancer growth and survival related genes in PC [2]. 2) PC cells overexpressing RANKL also transition to a more mensenchymal phenotype allowing access to blood circulation through increased cell motility, migration and invasion, and subsequently enter into a metastatic environment in bone and soft tissues. RANKL-producing PC cells could elicit positive interactions with osteoclasts in bone to greatly increase their bone colonizing capability through increased osteoclastogenesis [2]. 3) Since a stem cell or neuroendocrine phenotype confers PC cell resistance to host immune defenses or chemotherapy-induced cytotoxicity [24–26], PC cells with the MIC phenotype and increased stem cell or neuroendocrine cell properties have growth advantages at both primary and metastatic sites, with an increased propensity for developing chemotherapeutic resistance or escaping immune surveillance for preferential survival [27–29]. At the present time, it is speculative to state the origin of MICs and their relationship to the development of castration resistance in PC patients. Since the clinical specimens were collected from most hormonal naïve patients, derivation of MICs, therefore, could be age-related and influenced by the genetics and hormonal status of the patients. Because our data suggest that this unique subpopulation of MICs preferentially exists in PC patients with defined castration resistance, analysis of the phenotypic expression of MICs and BDCs could hold prognostic promise for differentiating aggressive from indolent PC.

Our studies showed that the mQDL technique has the unique capability of detecting multiple bio-markers at the single cell level on the same clinical specimen, with the advantages of reduced sample variability and conservation of scarce clinical materials. The inForm software allows cell-based gene expression quantification for profiling cell characteristics, and is perfectly suited for the analysis of MICs and BDCs. Our ultimate goal is to customize antibody-QD by conjugating them directly, reducing the labor intensive protocol and streamlining future sample analyses.

In summary, our study of clinical specimens from hormone naïve patients suggests that MICs with activated RANK-RANKL signaling, mediated by downstream p-c-MET and NRP1, confer castration resistant lethal progression of human PC. We found the proportion of MICs and BDCs predicts the status of castration resistance and the survival of PC patients. Future studies are needed to determine the etiology and evolution of MICs and BDCs in the prostate gland to gain further insights of the meta-static processes and therapeutic resistance in PC patients.

Acknowledgments

Grant sponsor: Cedars-Sinai Board of Governors Cancer Research Chair; Grant numbers: NCI 2-P01CA-98912; NCI 1-R01-CA122602; Grant sponsor: Prostate Cancer Foundation Challenge Award; Grant sponsor: DoD Idea Award.

This work is supported in part by funding from NCI 2-P01CA-98912, NCI 1-R01-CA122602, the Cedars-Sinai Board of Governors Cancer Research Chair, Prostate Cancer Foundation Challenge Award to LWKC and the DoD Idea Award to HYEZ. We thank Dr. S. Ashraf Imam for valuable discussions and Mr. Gary Mawyer for editing. We dedicate this study to the late Dr. John A. Arcadi for his inspiration and dedication to translational urologic research.

References

  • 1.Edwards BK, Noone AM, Mariotto AB, Simard EP, Boscoe FP, Henley SJ, Jemal A, Cho H, Anderson RN, Kohler BA, Eheman CR, Ward EM. Annual Report to the Nation on the status of cancer, 1975–2010, featuring prevalence of comorbidity and impact on survival among persons with lung, colorectal, breast, or prostate cancer. Cancer. 2014;120:1290–1314. doi: 10.1002/cncr.28509. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Chu GC, Zhau HE, Wang R, Rogatko A, Feng X, Zayzafoon M, Liu Y, Farach-Carson MC, You S, Kim J, Freeman MR, Chung LW. RANK- and c-Met-mediated signal network promotes prostate cancer metastatic colonization. Endocr-Relat Cancer. 2014;21:311–326. doi: 10.1530/ERC-13-0548. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Huang WC, Xie Z, Konaka H, Sodek J, Zhau HE, Chung LW. Human osteocalcin and bone sialoprotein mediating osteomimicry of prostate cancer cells: Role of cAMP-dependent protein kinase A signaling pathway. Cancer Res. 2005;65:2303–2313. doi: 10.1158/0008-5472.CAN-04-3448. [DOI] [PubMed] [Google Scholar]
  • 4.Nomura T, Huang WC, Zhau HE, Josson S, Mimata H, Chung LW. Beta2-Microglobulin-mediated signaling as a target for cancer therapy. Anti-Cancer Agents Med Chem. 2014;14:343–352. doi: 10.2174/18715206113139990092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Chu GC, Chung LW. RANK-mediated signaling network and cancer metastasis. Cancer Metast Rev. 2014;33:497–509. doi: 10.1007/s10555-013-9488-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Chung LW, Huang WC, Sung SY, Wu D, Odero-Marah V, Nomura T, Shigemura K, Miyagi T, Seo S, Chi C, Molitierno J, Elmore J, Anderson C, Isotani S, Edlund M, Hsieh CL, Wang R, Shehata B, Zhau HE. Stromal-epithelial interaction in prostate cancer progression. Clin Genitourin Cancer. 2006;5:162–170. doi: 10.3816/CGC.2006.n.034. [DOI] [PubMed] [Google Scholar]
  • 7.Josson S, Matsuoka Y, Chung LW, Zhau HE, Wang R. Tumor-stroma co-evolution in prostate cancer progression and metastasis. Semin Cell Dev Biol. 2010;21:26–32. doi: 10.1016/j.semcdb.2009.11.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Hu P, Chu GC, Zhu G, Yang H, Luthringer D, Prins G, Habib F, Wang Y, Wang R, Chung LW, Zhau HE. Multiplexed quantum dot labeling of activated c-Met signaling in castration-resistant human prostate cancer. PloS ONE. 2011;6:e28670. doi: 10.1371/journal.pone.0028670. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Hu P, Chung LW, Berel D, Frierson HF, Yang H, Liu C, Wang R, Li Q, Rogatko A, Zhau HE. Convergent RANK- and c-Met-mediated signaling components predict survival of patients with prostate cancer: An interracial comparative study. PloS ONE. 2013;8:e73081. doi: 10.1371/journal.pone.0073081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Odero-Marah VA, Wang R, Chu G, Zayzafoon M, Xu J, Shi C, Marshall FF, Zhau HE, Chung LW. Receptor activator of NF-kappaB Ligand (RANKL) expression is associated with epithelial to mesenchymal transition in human prostate cancer cells. Cell Res. 2008;18:858–870. doi: 10.1038/cr.2008.84. [DOI] [PubMed] [Google Scholar]
  • 11.Zhau HE, Odero-Marah V, Lue HW, Nomura T, Wang R, Chu G, Liu ZR, Zhou BP, Huang WC, Chung LW. Epithelial to mesenchymal transition (EMT) in human prostate cancer: Lessons learned from ARCaP model. Clin Exp Metastas. 2008;25:601–610. doi: 10.1007/s10585-008-9183-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Xu J, Wang R, Xie ZH, Odero-Marah V, Pathak S, Multani A, Chung LW, Zhau HE. Prostate cancer metastasis: Role of the host microenvironment in promoting epithelial to mesenchy-mal transition and increased bone and adrenal gland metastasis. The Prostate. 2006;66:1664–1673. doi: 10.1002/pros.20488. [DOI] [PubMed] [Google Scholar]
  • 13.Zhau HY, Chang SM, Chen BQ, Wang Y, Zhang H, Kao C, Sang QA, Pathak SJ, Chung LW. Androgen-repressed phenotype in human prostate cancer. Proc Natl Acad Sci USA. 1996;93:15152–15157. doi: 10.1073/pnas.93.26.15152. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Burgess TL, Qian Y, Kaufman S, Ring BD, Van G, Capparelli C, Kelley M, Hsu H, Boyle WJ, Dunstan CR, Hu S, Lacey DL. The ligand for osteoprotegerin (OPGL) directly activates mature osteoclasts. J Cell Biol. 1999;145:527–538. doi: 10.1083/jcb.145.3.527. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Kong YY, Feige U, Sarosi I, Bolon B, Tafuri A, Morony S, Capparelli C, Li J, Elliott R, McCabe S, Wong T, Campagnuolo G, Moran E, Bogoch ER, Van G, Nguyen LT, Ohashi PS, Lacey DL, Fish E, Boyle WJ, Penninger JM. Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature. 1999;402:304–309. doi: 10.1038/46303. [DOI] [PubMed] [Google Scholar]
  • 16.Kong YY, Yoshida H, Sarosi I, Tan HL, Timms E, Capparelli C, Morony S, Oliveira-dos-Santos AJ, Van G, Itie A, Khoo W, Wakeham A, Dunstan CR, Lacey DL, Mak TW, Boyle WJ, Penninger JM. OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature. 1999;397:315–323. doi: 10.1038/16852. [DOI] [PubMed] [Google Scholar]
  • 17.Lacey DL, Timms E, Tan HL, Kelley MJ, Dunstan CR, Burgess T, Elliott R, Colombero A, Elliott G, Scully S, Hsu H, Sullivan J, Hawkins N, Davy E, Capparelli C, Eli A, Qian YX, Kaufman S, Sarosi I, Shalhoub V, Senaldi G, Guo J, Delaney J, Boyle WJ. Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell. 1998;93:165–176. doi: 10.1016/s0092-8674(00)81569-x. [DOI] [PubMed] [Google Scholar]
  • 18.Duheron V, Hess E, Duval M, Decossas M, Castaneda B, Klopper JE, Amoasii L, Barbaroux JB, Williams IR, Yagita H, Penninger J, Choi Y, Lezot F, Groves R, Paus R, Mueller CG. Receptor activator of NF-kappaB (RANK) stimulates the proliferation of epithelial cells of the epidermopilosebaceous unit. Proc Natl Acad Sci USA. 2011;108:5342–5347. doi: 10.1073/pnas.1013054108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Asselin-Labat ML, Vaillant F, Sheridan JM, Pal B, Wu D, Simpson ER, Yasuda H, Smyth GK, Martin TJ, Lindeman GJ, Visvader JE. Control of mammary stem cell function by steroid hormone signalling. Nature. 2010;465:798–802. doi: 10.1038/nature09027. [DOI] [PubMed] [Google Scholar]
  • 20.Sigl V, Penninger JM. RANKL/RANK - from bone physiology to breast cancer. Cytokine Growth Factor Rev. 2014;25:205–214. doi: 10.1016/j.cytogfr.2014.01.002. [DOI] [PubMed] [Google Scholar]
  • 21.Schramek D, Leibbrandt A, Sigl V, Kenner L, Pospisilik JA, Lee HJ, Hanada R, Joshi PA, Aliprantis A, Glimcher L, Pasparakis M, Khokha R, Ormandy CJ, Widschwendter M, Schett G, Penninger JM. Osteoclast differentiation factor RANKL controls development of progestin-driven mammary cancer. Nature. 2010;468:98–102. doi: 10.1038/nature09387. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Blake ML, Tometsko M, Miller R, Jones JC, Dougall WC. RANK expression on breast cancer cells promotes skeletal metastasis. Clin Exp Metastas. 2014;31:233–245. doi: 10.1007/s10585-013-9624-3. [DOI] [PubMed] [Google Scholar]
  • 23.Garraway IP. Targeting the RANKL pathway: Putting the brakes on prostate cancer progression in bone. J Clin Oncol. 2013;31:3838–3840. doi: 10.1200/JCO.2013.50.1544. [DOI] [PubMed] [Google Scholar]
  • 24.Bonkhoff H. Neuroendocrine cells in benign and malignant prostate tissue: Morphogenesis, proliferation, and androgen receptor status. The Prostate Supplement. 1998;8:18–22. [PubMed] [Google Scholar]
  • 25.Bonkhoff H. Neuroendocrine differentiation in human prostate cancer. Morphogenesis, proliferation and androgen receptor status. Ann Oncol. 2001;12:S141–S144. doi: 10.1093/annonc/12.suppl_2.s141. [DOI] [PubMed] [Google Scholar]
  • 26.Terry S, Beltran H. The many faces of neuroendocrine differentiation in prostate cancer progression. Front Oncol. 2014;4:60. doi: 10.3389/fonc.2014.00060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Crea F, Danesi R, Farrar WL. Cancer stem cell epigenetics and chemoresistance. Epigenomics. 2009;1:63–79. doi: 10.2217/epi.09.4. [DOI] [PubMed] [Google Scholar]
  • 28.Dean M, Fojo T, Bates S. Tumour stem cells and drug resistance. Nat rev Cancer. 2005;5:275–284. doi: 10.1038/nrc1590. [DOI] [PubMed] [Google Scholar]
  • 29.Yan J, Tang D. Prostate cancer stem-like cells proliferate slowly and resist etoposide-induced cytotoxicity via enhancing DNA damage response. Exp Cell Res. 2014;328:132–42. doi: 10.1016/j.yexcr.2014.08.016. [DOI] [PubMed] [Google Scholar]

RESOURCES