Clinical Chemistry AACC Online Job Center
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Clinical Chemistry 50: 826-835, 2004. First published February 26, 2004; 10.1373/clinchem.2003.028563
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Data Supplement
Right arrow All Versions of this Article:
clinchem.2003.028563v1
50/5/826    most recent
Right arrow Submit an electronic Letter to
the Editor about this paper
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (17)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by O’Hara, S. M.
Right arrow Articles by Terstappen, L. W.M.M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by O’Hara, S. M.
Right arrow Articles by Terstappen, L. W.M.M.
Related Collections
Right arrow Molecular Diagnostics and Genetics
Right arrow Proteomics and Protein Markers
Right arrow Automation and Analytical Techniques
(Clinical Chemistry. 2004;50:826-835.)
© 2004 American Association for Clinical Chemistry, Inc.

Molecular Diagnostics and Genetics

Multigene Reverse Transcription-PCR Profiling of Circulating Tumor Cells in Hormone-Refractory Prostate Cancer

S. Mark O’Hara1, Jose G. Moreno2, Daniel R. Zweitzig1, Steve Gross1, Leonard G. Gomella2 and Leon W.M.M. Terstappen1,a

1 Immunicon Corporation, Huntingdon Valley, PA. 2 Thomas Jefferson University Hospital, Department of Urology, Philadelphia, PA.

aAddress correspondence to this author at: Immunicon Corporation, 3401 Masons Mill Road, Suit 100, Huntingdon Valley, PA 19006. Fax 215-830-0751; e-mail Lterstappen{at}Immunicon.com.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Background: Circulating tumor cells (CTCs) represent a surrogate source of tissue and conceptually represent a "real-time" biopsy. We previously reported that the number of CTCs mirrors disease progression in hormone-refractory prostate cancer (HRPC). To improve characterization of CTCs we further investigated whether in vitro transcription-based multigene reverse transcription-PCR expression profiles could be obtained from CTCs in HRPC.

Methods: We evaluated the expression of 37 genes with potential utility for epithelial cell characterization from antisense RNA libraries constructed from immunomagnetically enriched CTCs from 7.5-mL blood samples from healthy donors and patients with HRPC.

Results: In the control group 13 of 37 genes were not expressed. The most notable of the genes expressed in CTCs of 23 blood specimens drawn from 9 patients with metastatic prostate cancer were prostate-specific antigen (20 of 23; 87%), prostate-specific membrane antigen (17 of 23; 74%), androgen receptor (16 of 23; 70%), human glandular kallikrein 2 (7 of 23; 30%), epidermal growth factor receptor (4 of 23; 17%), and prostate-specific gene with homology to G protein receptor (2 of 23; 9%). The number of CTCs in these samples ranged from 4 to 283 in 7.5 mL of blood (mean, 87; median, 89). Expression of some of the genes was low in the control samples and higher in the patient samples. In all 23 samples, cytokeratin 19, epithelial cell adhesion molecule, or mucin 1 was expressed. Because of background expression in the controls, expression of 13 of the 37 genes, including HER-2, p53, and BCL-2, could not be measured in CTCs.

Conclusion: Antisense RNA libraries can be constructed from CTCs and gene expression profiles of CTCs obtained from patients with HRPC. This could enhance the characterization of HRPC and facilitate the development of more effective therapies.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Evidence exists that the presence of minimal residual disease in patients with solid tumors has clinical significance (1)(2)(3)(4). Advancements in the detection, phenotyping, and genotyping of circulating tumor cells (CTCs) 1 may lead to therapies tailored to individual patients (5)(6)(7)(8)(9). We recently demonstrated that the number of CTCs mirrors disease progression in hormone-refractory prostate cancer (HRPC) (10). This assay was based on immunomagnetic enrichment of epithelial cells in concert with multiparameter flow cytometry (9)(10). This system allowed us to phenotype CTCs for HER-2 protein in breast cancer patients (9). Morphologic characterization of CTCs was achieved with an automated fluorescent microscope (11)(12).

In this study we expand the analytical capabilities of immunomagnetically enriched CTC enumeration to include molecular biological characterization of CTCs. Although immunomagnetic enrichment of CTCs can provide a four to five log reduction in leukocytes, the typical range for CTC-to-leukocyte ratios is 1–10 CTCs/103–104 leukocytes. The low number of CTCs and the leukocyte carryover during the CTC enrichment process pose significant detection restrictions in terms of the signal-to-noise ratio, which constrains the choice of genes and gene expression profiling methods. The low CTC number thus requires preamplification of the entire mRNA library to analyze a multigene reverse transcription-PCR (RT-PCR) panel without compromising the sensitivities of individual marker genes. Here we show that detection of the RNA encoded by numerous genes in CTCs is feasible. The method entails constructing and synthesizing an antisense RNA (aRNA) library from mRNA extracted from immunomagnetically enriched CTCs, which permits the analysis of numerous genes through RT-PCR. We investigated the expression of 37 genes, including the prostate- tissue-specific genes prostate-specific antigen (PSA), prostate-specific membrane antigen (PSMA), and human glandular kallikrein 2 (HK2). Also included were prostate-cancer-specific genes such as hepsin (HPN) and prostate-specific gene with homology to G protein receptor (PSGR), as well as therapy-relevant genes such as androgen receptor (AR), HER-2, BCL-2, multidrug resistance gene (MDR1), and epidermal growth factor receptor (EGFR). The ability to quantify, phenotype, morphologically image, and now profile the gene expression of CTCs allows improved biological characterization of HRPC in real time and may help expedite the development of effective patient-specific therapies.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
patients
Blood from 9 patients with prostate cancer (CaP) and 13 healthy volunteers (7 males and 6 females) was drawn into two 10-mL EDTA-containing Vacutainer Tubes (Becton Dickinson) and pooled. Of the nine patients, eight had HRPC with a sustained increase in PSA and/or a positive bone scan. The single hormone-sensitive patient initially presented with a positive bone scan, but he subsequently had a sustained complete response to androgen ablation with a negative bone scan and a PSA of 0.1 µg/L. The male controls were 27–73 years of age (mean, 45 years), and the female controls were 27–61 years of age (mean, 39 years). The patients ranged in age from 60 to 81 years (mean, 74 years). The treatments, bone scan status, PSA concentrations, and CTC counts are given in Table 1 . Of the nine patients, two had five longitudinal blood samples, one had four samples, three had two samples, and three had single samples. All participants signed an institutional informed consent form.


View this table:
[in this window]
[in a new window]
  		Table 1. Patient characteristics.

cell lines
The breast cancer cell line SKBR-3 and prostate cancer cell line LNCaP were cultured in RPMI-1640 supplemented with 100 mL/L fetal calf serum.

immunomagnetic sample preparation
Immunomagnetic enrichment of epithelial cells has been described in detail previously (10)(11). Blood samples were divided into two 7.5-mL aliquots processed within 3 h. Briefly, the enrichment step involves adding magnetic nanoparticles coated with a monoclonal antibody with affinity for epithelial cell adhesion molecule (EpCAM) to a 7.5-mL blood sample, incubating the sample, and placing the tube inside a magnetic separator for immunomagnetic separation of the magnetically labeled CTCs. After separation, the enriched CTCs from one tube are used for RNA extraction, and the second tube is used for CTC enumeration.

ctc enumeration
Enriched cells containing CTCs and contaminating leukocytes were stained with pan-anti-cytokeratin-phycoerythrin to detect epithelial cells, a pan-leukocyte marker (anti-CD45 PerCP; Becton Dickinson), and a nucleic acid dye as described previously (10). Leukocytes were defined as nucleic acid+, CD45+, cytokeratin– events with typical forward and orthogonal light scatter signals in flow cytometry. CTCs were defined and differentiated from leukocytes as nucleic acid+, CD45–, cytokeratin+ events with forward light scatter and orthogonal light scatter signals at least as large as those of leukocytes.

rna preparation
After immunomagnetic enrichment of CTCs, the cells were lysed by adding 100 µL of Trizol reagent (Trizol reagent total RNA isolation; cat. no. 15586-018; Invitrogen). During Trizol RNA isolation, coprecipitate was formulated according to the formula 100 µL of 200 ng/µL linear acrylamide (cat. no. 9520; Ambion) plus 25 µL of pellet paint (cat. no. 69049; Novagen) and added to the aqueous phase before alcohol precipitation. RNA pellets were resuspended in 15 µL of RNase-free water. Isolated RNA was treated with 1 U/µL DNase I (amplification grade; cat. no. 18068-015; Invitrogen) in a 20-µL reaction volume according to the manufacturer’s instructions. DNase-treated RNA was purified by extraction with 180 µL of TRIzol and 2 µL of coprecipitate, and the pellet was resuspended in 5 µL of RNase-free water. To evaluate the quantity and quality of the isolated total RNA, we electrophoresed 1 µL (20%) of the RNA on a 1% denaturing agarose gel and performed a Northern blot for hybridization with a 10-pmol mixture of ribosomal 18S (5'-GCCCTCCAATGGATCCTCGTTAAAGG-3') and 28S (5'-GCTCTTCCCTGTTCACTCGCCGTTA-3') oligomer probes 32P-end-labeled with polynucleotide kinase in ExpressHyb Hybridization Solution (cat. no. 8015-1; Clontech) at 42 °C for 2 h, according to the manual for the probes. The probed blot was phosphorimaged on a Cyclone PhosphorImager (Packard Instruments) to determine RNA integrity (18S:28S ratio) and total mass relative to high-integrity total RNA calibrators of known mass (0.25, 0.5, 1.0, and 2.0 ng) run in parallel on the blot. The RNA mass from each sample was defined as the 7.5-mL blood-donor-equivalent for sample normalization by volume. The detailed procedure for RNA isolation can be found in the Data Supplement that accompanies the online version of this article athttp://www.clinchem.org/content/vol50/issue5/ .

transcript library preamplification by smart-Arna
The SMART-aRNA procedure uses the switching mechanism at the 5' termini of RNA templates in the SMARTTM PCR cDNA Synthesis Kit (cat. no. K1052-1; Clontech). The SMART-aRNA procedure was used because it simplifies the second-strand synthesis for constructing double-stranded cDNA libraries. The SMART primer was modified on the 3' terminus to contain a functional T7 RNA polymerase promoter for subsequent aRNA amplification reactions. In addition, a minimum number of PCR cycles was used to boost cDNA library mass before T7 RNA polymerase linear amplification with minimal negative effects on mRNA library representation. The remaining 80% of each RNA sample was used in the synthesis of first-strand cDNA followed by 10 cycles of PCR performed according to the SMART Uses Manual (part no. PT3041-1; Clontech) along with the following three modifications:

We then performed 10 cycles of PCR amplification with the following cycling profile: 95 °C for 1 min followed by 10 cycles of 95 °C for 5 s, 65 °C for 15 s, and 68 °C for 6 min, with a final 20-min step at 72 °C. The PCR reaction mixture was processed through a Sephadex G-50 Quick Spin column (TE; cat. no. 1 523 023; Roche Diagnostics), and the eluate was concentrated under reduced pressure to 7 µL. T7 aRNA polymerase amplification reactions using the AmpliScribe reagent set (cat. no. AS2607; Epicenter Technologies) were assembled and incubated at 37 °C for 12–16 h. The aRNA was purified by use of 180 µL of TRIzol with 1.25 µL of coprecipitate added to the aqueous phase, and the pellet was resuspended in 10 µL of RNase-free water. We denatured 20% (2 µL) of the purified aRNA reaction mixture, using 1 µL of 6x formamide–formaldehyde and 3 µL of water at 65 °C for 15 min. The denatured mixture was electrophoresed (2% agarose gel) along with 50 and 500 ng of RNA molecular weight and mass calibrators (0.24–9.5 Kb RNA ladder; 1 µg/µL; cat. no. 15620-016; Invitrogen) for 15 min at 5 V/cm and then was stained with SYBR Gold (cat. no. S-11494; Molecular Probes). The quality and quantity of the aRNA library were determined by AlphaImager densitometry relative to calibrators by use of AlphaEaseTM software, Ver. 5.04. (Alpha Innotech Corp.). Sample quantities were normalized by volume according to the equation:

where "Total RNA mass/7.5-mL sample x 0.015" indicates that mRNA accounts for 1.5% of total RNA mass; and "3" indicates that the mean molecular weight of the aRNA library ~3-fold less than that of the total mRNA.

The detailed procedure for transcript library preamplification by SMART-aRNA can be found in the online Data Supplement.

multigene rt-pcr analysis
We reverse-transcribed 25 ng of aRNA in a 10-µL reaction volume containing 50 ng of random 9mer primers, 1 µL of SuperScript (Invitrogen), and 2 µL of 5x reaction buffer (Invitrogen); the mixture was incubated at 25 °C for 10 min, 37 °C for 10 min, 42 °C for 20 min, and 50 °C for 60 min. For PCR, cDNA was then diluted to 1 donor-equivalent/µL. Ten donor-equivalents (10 µL) per cDNA sample (mass range, 50–1300 pg) were used in each subsequent 50-µL PCR reaction containing 1 U of platinum Taq (Invitrogen). We generated individual PCR kinetics curves from single PCR reaction tubes by pausing the thermocycler at 31, 35, and 40 cycles and aliquoting 15 µL to a separate tube for subsequent gel densitometry analysis. The PCR program was as follows in a PE-9700 thermal cycler: 1 min at 95 °C, followed by 40 cycles of 95 °C for 15 s, 65 °C for 15 s, and 72 °C for 1 min, followed by 20 min at 72 °C. RT-PCR samples were electrophoresed on 2% agarose gels (5 V/cm for 20 min) in parallel with DNA size and mass calibrators (Low Range Quantitative DNA ladder; cat. no. 12373-031; Invitrogen), stained with ethidium bromide (5 g/L), and analyzed by an AlphaImager Gel Documentation and Image Analysis System with AlphaEaseTM software, Ver. 5.04, to determine molecular weight and spot densitometry.

primer design
All gene-specific RT-PCR primer sets were designed to amplify within the last 500 nucleotides of each transcript’s 3'-untranslated region upstream from poly(A) and optimized to give single bands of specific molecular weights (Table 2 ). Sequence identity was confirmed by internal oligoprobe blot hybridization.


View this table:
[in this window]
[in a new window]
  		Table 2. Multigene RT-PCR sequence information and gene expression in SMART-aRNA libraries from immunomagnetically enriched CTC from healthy donors.

scaling of rt-pcr results relative to an external calibration curve
Relative gene expression values of 0, 1, 2, 3, and 4 were assigned to samples based on comparisons of kinetics curve band intensities relative to an external calibration curve, prepared in parallel, of cytokeratin 19 (CK19) cDNA, as shown in Fig. 1 . The lower limit of RT-PCR detection was evaluated with a CK19 in vitro-transcribed RNA construct (CK19 cRNA) that contained the 3'-most 800 nucleotides of GenBank accession no. NM_002276.



View larger version (60K):
[in this window]
[in a new window]
  		Figure 1. Gene expression scaling was performed relative to a CK19 cDNA external calibration curve.

RT-PCR for CK19 was performed on dilutions of CK19 cDNA containing 25 000, 2500, 250, and 25 copies, each added to 2 ng of leukocyte total RNA. Reactions were run for 31, 35, and 40 cycles, and typical band intensities corresponding to CK19 cDNA copy number are shown here. For multigene RT-PCR scaling, comparisons were made to the CK19 cDNA external calibration curve run in parallel, and approximate 10-fold differences in relative gene expression were assigned: 0, nondetectable; 1, ~25–50 copies; 2, ~250–500 copies; 3, ~2500–5000 copies; 4, >25 000 copies.

lower limit of detection for the smart-Arna preamplification and rt-pcr analysis system
The lower limit of mRNA transcript detection for the entire SMART-aRNA preamplification and RT-PCR analysis system was determined in triplicate with a test sample set. The test RNA samples were composed of a tissue culture cell line RNA containing predetermined copy numbers of PSA, PSMA, CK19, AR, and HPN mRNA, which was added to a background of 2000 leukocyte-equivalents of total RNA. The test set was then subjected to SMART-aRNA preamplification and RT-PCR analysis.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
ctc enumeration and rna analysis
In the blood samples from 13 healthy donors, the number of leukocytes carried over in the immunomagnetic selection ranged from 655 to 5560 (median, 1450; mean, 1759). In the 23 samples from CaP patients, the number of leukocytes carried over ranged from 813 to 92 000 (median, 4350; mean, 12 300). The events that were classified as CTCs in blood samples from the control group ranged from 0 to 4 CTCs/7.5 mL [mean (SD), 0.8, (1.2)]. In blood samples from HRPC patients, the CTCs ranged from 4 to 283/7.5 mL (median, 89; mean, 87; Table 1Up ). The total RNA from the 13 healthy donors as assessed by Northern blot had a mean 28S:18S band ratio of 1.4 (range, 1.2–1.8) with a mean yield of 3.5 ng (range, 0.8–11.12 ng). In CaP patients, this ratio was 1.2 (range, 0.6–1.8) with a mean yield of 7.2 ng (range, 0.8–35.12 ng). All samples subsequently produced aRNA libraries with masses closely proportional to the starting total RNA values.

characterization of smart-Arna library amplification
The modified aRNA amplification method produced libraries with a mean 10 000-fold yield above the original mRNA mass. The original amount of mRNA was estimated to be 1.5% of the total RNA mass. The median transcript length of the libraries was 600 nucleotides (range, 550–800 nucleotides). Individual transcript lengths within each library ranged from 300 to 3000 nucleotides. We investigated the distortion effect of this modified aRNA amplification method on the relative mRNA abundance by comparing the expression of CK8, CK19, mammaglobin 1 (MGB1), mammaglobin 2 (MGB2), prolactin-inducible protein (PIP), EpCAM, PSA, and PSMA in RNA libraries obtained from tumor cell lines with and without amplification. As shown in Fig. 2 , there was no dropout of any mRNA target sequence attributable to the SMART-aRNA method. In addition, we saw no dramatic differences in band intensity ratios for all eight mRNA sequences when we compared the amounts of RNA after 32 and 40 PCR cycles.



View larger version (36K):
[in this window]
[in a new window]
  		Figure 2. Relative abundance of individual mRNA sequences in mRNA libraries.

Multigene RT-PCR analyses for PSA, PSMA, MGB1, MGB2, PIP, CK8, CK19, and EpCAM on native unamplified and SMART-aRNA-amplified mRNA libraries. A mixture of 15 cell-equivalents of LNCaP and SKBR3 was added to 2 ng of total leukocyte RNA (1000 cell-equivalents) to generate libraries.

We assessed the effect of the SMART-aRNA preamplification procedure on the lower limit of detection for RT-PCR with a model system mimicking immunomagnetically enriched CTC samples. In triplicate samples, the combination of SMART-aRNA followed by RT-PCR amplification showed reproducible detection of mRNA from all six targets—PSA, PSMA, AR, HPN, EpCAM, and CK19—down to at least 50 transcripts (data not shown). We evaluated the lower limit of detection for RT-PCR with in vitro-transcribed CK19 cRNA. The results from 10 separate RT-PCR reactions showed that the CK19 cRNA was reproducibly detectable in 10 of 10 separate reactions down to 25 molecules of RNA (data not shown). RT-PCR contamination controls containing all components except cDNA sample template were run in parallel with each batch and did not show any detectable signals, indicating the absence of crossover and carryover contamination during these studies.

multigene rt-pcr profiling of smart-Arna from samples from healthy donors immunomagnetically enriched for ctcs
An external calibration curve constructed with CK19 cDNA was run in parallel with samples. The band intensities from the calibration curve were used to assign relative gene expression values of 0, 1, 2, 3, and 4 to the samples shown in Fig. 1Up . All blood samples contained leukocytes carried over from the CTC enrichment process. The {alpha}1-globin gene was found to be specifically expressed in leukocytes and was therefore used as the functional system control for indicating successful RNA processing and amplification. All 36 samples gave gene expression values >4 for {alpha}1-globin (Table 2Up ; data not shown for patient samples). Expression of 38 genes was assessed in the aRNA libraries of the 13 healthy controls. The expression profiles of these genes and the aRNA-RT-PCR design characteristics are listed in Table 2Up . Expression of the genes in CTC-enriched samples from healthy controls represented the expression of these genes in the leukocytes that were carried over in the CTC enrichment procedure. We observed no expression of AR, carcinoembryonic antigen (CEA), cytokeratin 5 (CK5), CK19, EGFR, estrogen receptor-ß (ER-b), HK2, MGB1, MGB2, PSA, PSGR, PSMA, or tumor-associated calcium signal transducer 2 (Trop2) in any of the libraries from the controls. The expression of these genes in CTCs enriched from cancer patients could therefore be attributed to the CTCs and not the leukocytes carried over in the CTC enrichment procedure. We found no or low expression for EpCAM, macrophage-inhibitory cytokine-1 (Mic1), matrix metalloproteinase 2 (MMP2), mucin 1 (Muc1), an androgen-regulated gene (NKX3A), and topoisomerase II{alpha} (Topo2A) in controls. Expression of these genes in CTCs can therefore be attributed to CTCs if they are expressed in increased amounts. Genes more highly expressed in the controls were not used in the evaluation of CTCs with this CTC enrichment procedure. Within this group were the genes encoding for the cytoskeletal proteins cytokeratins 8, 10, and 18, which are generally restricted to nonhematopoietic cells, and genes that are potential candidates for therapeutic agents: BCL2, HER-2, multidrug resistance-associated protein (MRP), and p53. In all samples, the score for CK10 was 3, which implied a housekeeping function that potentially can be used for sample normalization.

multigene rt-pcr profiling of smart-Arna from samples from hrpc patients immunomagnetically enriched for ctcs
The relative expression of CK19, PSA, PSMA, AR, HPN, HK2, PSGR, MGB1, and MGB2 in the mRNA libraries from CTC-enriched samples obtained from 23 samples from 9 patients with metastatic CaP are shown in Fig. 3 . The numbers for individual samples from Table 1Up are used to indicate the relative expression of each gene. A multigene expression profile can be obtained from each patient and each sample by examining the positions of the sample numbers across all genes. CK19 was expressed in 18 of 23 (78%) samples and 6 of 9 (67%) CaP patients. PSA was expressed in 20 of 23 (87%) samples and 8 of 9 (89%) CaP patients. PSMA was expressed in 17 of 23 (74%) samples and 7 of 9 (78%) CaP patients. AR was expressed in 16 of 23 (70%) samples and 7 of 9 (78%) CaP patients. HPN was expressed in 13 of 23 (57%) samples and 6 of 9 (78%) CaP patients. It must be noted that HPN was expressed in the RNA library of one of the female controls. HK2 was expressed in 7 of 23 (30%) samples and 4 of 9 (44%) CaP patients. PSGR was expressed in 2 of 23 (9%) samples and 2 of 9 (22%) CaP patients. Unexpectedly, both "breast-specific" genes MGB1 and MGB2 scored positive in 1 of 23 (4%) and 2 of 23 (9%) CaP samples, respectively. The multigene combination of PSA and PSMA was positive in 20 in 23 (87%) samples, and addition of HPN produced positive results in 21 of 23 (91%) samples. Only one sample, patient 9, scored negative for all of the tested genes shown in Fig. 3 . The genes EGFR (4 of 23; 17%), Trop2 (2 of 23; 9%), ER-b (1 of 23; 4%), CK5 (1 of 23; 4%); and CEA (0 of 23; 0%) were infrequently expressed in patient samples and were not expressed in the leukocytes. The genes EpCAM, Mic1, MMP2, Muc1, NKX3A, and Trop2A were expressed in low background amounts in leukocytes.



View larger version (23K):
[in this window]
[in a new window]
  		Figure 3. Multigene mRNA expression profiles of immunomagnetically enriched CTCs in 23 samples from CaP patients based on SMART-aRNA preamplification followed by RT-PCR.

Patients are numbered 1–9, and their serial samples are numbered x.1–x.5 (Table 1Up ). With the unique identification numbers the multigene profile can be traced. The x axis shows 9 of the 13 genes that did not show any leukocyte background in Table 2Up , except for HPN, which was positive in one of the female control samples. The y axis shows the gene expression values based on RT-PCR relative to an external calibration curve.

When the threshold was increased to 2 or higher, EpCAM scored positive in 17 of 23 (74%), Mic1 in 15 of 23 (65%), MMP2 in 0 of 23 (0%), Muc1 in 11 of 23 (48%), NKX3A in 2 of 23 (9%), and Topo2A in 5 of 23 (22%) patient samples. MDR1 expression was scored as 0, 1, or 2 in the control samples, and this expression was exceeded in 5 of 23 (22%) patient samples. The multigene combination of CK19, EpCAM, and Muc1 yielded a sensitivity of 100% (23 of 23).

serial ctc and Mrna profiles during treatment of three hrpc patients
Longitudinal patient profiles of CTC counts and mRNA expression for PSA, PSMA, AR, and HPN in blood samples drawn over the course of 18–26 weeks from three patients with HRPC are shown in Fig. 4 . Patient 1 (Fig. 4A ) was treated with long-term hormone ablation (Lupron). Patients 2 and 3 (Fig. 4, B and C , respectively) were also on long-term Lupron treatment and, in addition, were begun on chemotherapy consisting of Taxane/estramustine. The CTC values changed by ~10-fold for all three patients during this survey period, and some trends were noted as follows. For the patient treated with Lupron alone (Fig. 4A ), who had four longitudinal samples drawn at 0, 5, 10, and 18 weeks, we observed a dramatic progressive increase in CTC counts from 45 up to 300 during the survey period. In contrast, the two patients treated with Lupron and adjuvant chemotherapy (Fig. 4 , B and C) showed a significant increase in CTC counts followed by a decrease. For the patient in Fig. 4C , we observed a dramatic increase in CTCs after the chemotherapy regimen was discontinued. mRNA concentrations during serial sampling showed that PSMA expression paralleled CTC changes, whereas PSA expression remained high throughout the survey period. The amounts of AR mRNA in the patients represented in panels A and B of Fig. 4 remained relatively constant. AR expression was detected in blood draws from the patient represented in Fig. 4C only during treatment with Taxane/estramustine and was not detected before or after treatment. We detected a dramatic change in HPN mRNA in the patients represented in panels B and C of Fig. 4 , from high expression when the patients were untreated to complete elimination during the Taxane/estramustine treatment. HPN expression fluctuated in the blood samples from the patient represented in Fig. 4A . The changes in mRNA concentrations were independent of the number of CTCs and leukocytes.



View larger version (31K):
[in this window]
[in a new window]
  		Figure 4. Longitudinal monitoring of number of CTCs (•) and relative amounts of PSA ({diamond}), PSMA (), AR ({square}), and HPN ({circ}) mRNA in HRPC patients during therapy.

(A), samples 1.1–1.4 from patient 1, who was treated with Lupron and Casodex. (B and C), samples 2.1–2.5 and 3.1–3.5 from patients 2 and 3, respectively, who were treated with Lupron, estramustine (ES), Taxotere (TX), and Taxol (TL). The numbers of CTCs in samples from healthy donors (mean + 2 SD, 2.8 CTCs/7 mL of blood) is indicated by the dashed line with no symbols in each panel.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Clinical endpoints to measure HRPC disease burden have traditionally been limited to survival. Recently, serum PSA has gained some acceptance as a viable endpoint (13), and positive signals for PSA RT-PCR, signifying the presence of CTCs in HRPC, have also been suggested to predict survival (3). We have shown that the quantity of CTCs trends with metastatic disease progression in a manner that is independent of PSA concentration in select cases (10). Our goal in this study was to determine the feasibility of profiling gene expression in CTCs with the ultimate aim of better characterizing advanced CaP biology. However, a substantial enrichment of CTCs over leukocytes is needed to overcome signal-to-noise factors. A modified T7 RNA polymerase preamplification method (SMART-aRNA) was engineered to produce aRNA CTC libraries from immunomagnetically enriched CTCs targeting the epithelial cell surface antigen EpCAM. The preamplification of mRNA produced a 10 000-fold increase in mRNA, whereas the representation of genes in the aRNA library was not compromised compared with the native, unamplified mRNA library (Fig. 2Up ). mRNA was detectable down to amounts as low as 50 molecules (data not shown). The small linear dynamic range obtained with the semiquantitative RT-PCR detection system used in this study can be readily improved by use of "real-time" quantitative RT-PCR. This approach to profile mRNA gene expression in the CTCs present in 7.5 mL of blood is limited by a background of 103–104 leukocytes. The unavoidable leukocyte carryover translates into significant signal-to-noise restrictions that limit the genes that can be adequately evaluated. For example, the expression of p53, TS, MRP, HER-2, and BCL-2 in leukocytes precludes unambiguous detection of their overexpression in CTCs. It might, however, be possible to develop a method to subtract the leukocyte mRNA background from patient-specific CTCs based on the leukocyte profile obtained from patient-matched CTC-depleted leukocyte fractions or CTCs enriched from healthy donors. Alternatively, the leukocyte background can be significantly reduced by use of fluorescence-activated cell sorting. However, cell sorting inevitably leads to losses of CTCs and is practical only in a research setting.

Another method to overcome this limitation of gene expression is to assess the amount of mRNA through protein production. For example, although HER-2 mRNA analysis is not possible in immunomagnetically enriched CTCs because of the background expression of HER-2 mRNA in leukocytes, the production of HER-2 protein by CTCs has been assessed successfully in breast cancer patients (9). Conversely, not all proteins can be assessed on CTCs. Some secreted proteins, such as PSA, migrate out of the CTCs during permeabilization and can no longer be detected. Hence the limitations of CTC detection by flow cytometry or microscopy can be overcome by complementary aRNA analysis. This was illustrated in one of the nine patients. Case 9 was the only androgen-sensitive case in this study and had 30 CTCs/7.5 mL of blood, which was confirmed morphologically by fluorescence microscopy. However, these CTCs were negative on RT-PCR for PSA, PSMA, HK2, and PSGR, suggesting that they might express different genes reflecting their different disease biologies. However, at this time we cannot rule out the possibility that these CTCs might not be derived from the prostate. mRNA concentrations for Muc1 and Trop2A were higher in samples from patients than in control samples, suggesting that these CTCs could be derived from epithelial tissue representing another clinically occult neoplasm. CTCs ranged from 4 to 283/7.5 mL blood for 23 specimens from 9 prostate cancer patients. In contrast the CTC range in the 13 control cases was 0–4 CTCs/7.5 mL (mean, 0.8). One of the epithelial cell-specific genes (CK19, EpCAM, or Muc1) was expressed in all patient samples, whereas 21 of 23 samples from 8 of 9 patients showed expression of one of the prostate-specific genes (PSA, PSMA, or HPN). Expression of these prostate-tissue-specific genes was completely absent in controls. The relative expression of CK19, PSA, and PSMA was comparable to the values reported in the literature (14)(15). For example, the proportion of samples that were positive for HK2 expression (30%) was similar to the 25% reported by Ylikoski et al. (16). Although the proportion of samples in which PSGR expression was detectable (2 of 23; 9%) may be lower than previously published values, it is possible that PSGR is expressed in higher amounts in CTCs of patients with hormone-naive prostate disease (17)(18). HPN has not been studied extensively in patients with HRPC, and the proportion of samples with detectable expression (57%) in this report suggests that this may be a viable drug therapy target (19). Moreover, the efficacy of anti-HPN agents may be assessed by their ability to decrease HPN expression in CTCs. EGFR was expressed in only 17% of the specimens, which is low compared with the 57% reported in prostate samples from 16 patients with HRPC (20). This discrepancy may be attributable to several factors, including differential EGFR expression in CTCs vs prostate cancer cells within the organ of origin. We found a similar discrepancy in Topo2A expression (22%) (21). The expression of AR in 70% of the samples is in agreement with the results reported by Linja et al. (22), who found a correlation between androgen independence and high AR expression in 13 patients with HRPC. Many hormone-dependent features are known to be common to breast and prostate cancers, which may explain the observed expression of PSA in breast cancer (23) and the first reported expression of the breast-specific genes MGB1 and MGB2 in CTCs from CaP patients.

Expression of the AR, HPN, PSMA, and PSA genes in CTCs was surveyed serially over an ~8- to 26-week period in three patients (Fig. 4Up ). Two of the three patients had received chemotherapy. The results shown in Fig 4Up demonstrate that the CTC counts reflect disease progression more accurately than semiquantitative PSA and PSMA RT-PCR results, which showed uniformly high expression throughout the monitoring period. This may be limited by the semiquantitative RT-PCR assay used in this study, however, because a relative gene expression score of 4 for PSA and PSMA can not be differentiated from saturation in this detection system. Our data imply that previous PSA RT-PCR studies that attempted to predict survival may have been hampered by an inability to differentiate between high and low CTC counts. The results further support research into AR as a therapeutic target because the amount of AR expression was sustained and found to be minimally affected by chemotherapy.

In conclusion, we believe that gene expression profiling of CTC mRNA is feasible. Our data further validate that the immunomagnetically enriched CTCs enumerated by flow cytometry or fluorescence microscopy originated from the CaP tissue. The prostate CTCs exhibited heterogeneous expression of genes such as PSA, PSMA, and HPN. In the HRPC cases studied, AR expression appeared to be consistently increased, suggesting AR as a viable therapeutic target. The ability to quantify, phenotype, morphologically examine, and now molecularly profile CTCs could lead to improved characterization of HRPC and ultimately to development of more effective, personalized novel therapeutic strategies.


   Footnotes
 
1 Nonstandard abbreviations: CTC, circulating tumor cell; HRPC, hormone-refractory prostate cancer; RT-PCR, reverse transcription-PCR; aRNA, antisense RNA; PSA, prostate-specific antigen; PSMA, prostate-specific membrane antigen; HK2, human glandular kallikrein 2; HPN, hepsin; PSGR, prostate-specific gene with homology to G-protein receptor; AR, androgen receptor; MDR1, multidrug resistance gene; EGFR, epidermal growth factor receptor; CaP, prostate cancer; EpCAM, epithelial cell adhesion molecule; CK, cytokeratin; MGB1 and MGB2, mammaglobin 1 and 2; PIP, prolactin-inducible protein; CEA, carcinoembryonic antigen; ER-b, estrogen receptor ß; Trop2, tumor-associated calcium signal transducer 2; Mic1, macrophage inhibitory cytokine-1; MMP2, matrix metalloproteinase 2; Muc1, mucin 1; Topo2A, topoisomerase II{alpha}; and MRP, multidrug resistance-associated protein.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Braun S, Pantel K, Muller P, Janni W, Hepp F, Kentenich CR, et al. Cytokeratin-positive cells in the bone marrow and survival of patients with stage I, II or III breast cancer. N Engl J Med 2000;341:525-533.
  2. Janni W, Hepp F, Rjosk D, Kentenich C, Strobl B, Schindlbeck C, et al. The fate and prognostic value of occult metastatic cells in the bone marrow of patients with breast carcinoma between primary treatment and recurrence. Cancer 2001;92:46-53.[CrossRef][ISI][Medline] [Order article via Infotrieve]
  3. Halabi S, Small EJ, Hayes DF, Vogelzang HN, Kantoff PW. Prognostic significance of reverse transcriptase polymerase chain reaction for prostate-specific antigen in metastatic prostate cancer: a nested study within CALGB9583. J Clin Oncol 2003;21:490-495.[Abstract/Free Full Text]
  4. Cristofanilli M, Reuben JM, Fritsche HA, Kau SW, Lara J, Arun B, et al. Circulating tumor cells (CTCs): a novel prognostic factor for newly diagnosed metastatic breast cancer (MBC). Proc Am Assoc Cancer Res 2003;44:1349.
  5. Ellis WJ, Pfitzenmaier J, Colli J, Arfman E, Lange PH, Vessella RL. Detection and isolation of prostate cancer cells from peripheral blood and bone marrow. Urology 2003;61:277-281.[CrossRef][ISI][Medline] [Order article via Infotrieve]
  6. van’t Veer LJ, Dai H, van de Vijver MJ, He YD, Hart AA, Mao M, et al. Gene expression profiling predicts clinical outcome of breast cancer. Nature 2002;415:530-536.[CrossRef][Medline] [Order article via Infotrieve]
  7. Van Gelder RN, von Xastrow ME, Yool A, Dement DC, Barchas JD, Eberwine JH. Amplified RNA synthesized from limited quantities of heterogeneous cDNA. Proc Natl Acad Sci U S A 1990;87:1663-1667.[Abstract/Free Full Text]
  8. Gordon GJ, Jensen RV, Hsiao LL, Gullans SR, Blumenstock JE, Richards WG, et al. Using gene expression ratios to predict outcome among patients with mesothelioma. J Natl Cancer Inst 2003;95:598-605.[Abstract/Free Full Text]
  9. Hayes DF, Walker TM, Singh B, Vitetta ES, Uhr JW, Gross S, et al. Monitoring expression of HER-2 on circulating epithelial cells in patients with advanced breast cancer. Int J Oncol 2002;21:1111-1118.[ISI][Medline] [Order article via Infotrieve]
  10. Moreno JG, O’Hara SM, Gross S, Doyle G, Fritsche H, Gomella LG, et al. Changes in circulating carcinoma cells in patients with metastatic prostate cancer correlate with disease status. Urology 2001;58:386-392.[CrossRef][ISI][Medline] [Order article via Infotrieve]
  11. Kagan M, Howard D, Bendele T, Mayes J, Silvia J, Repollet M, et al. A sample preparation and analysis system for identification of circulating tumor cells. J Clin Ligand Assay 2002;25:104-110.
  12. Kagan M, Trainer M, Bendele T, Rao C, Horton A, Tibbe A, et al. CellTracks cell analysis system for rare cell detection. Cohn GE eds. Clinical diagnostic systems: technologies and instrumentation 2002;4625:20-28 SPIE Washington. .
  13. Bubley GJ, Carducci M, Dahut W, Dawson N, Daliani D, Eisenberger M, et al. Eligibility and response guidelines for phase II clinical trials in androgen-independent prostate cancer: recommendations from the Prostate-Specific Antigen Working Group. J Clin Oncol 1999;17:3461-3467.[Abstract/Free Full Text]
  14. Gomella LG, Raj GV, Moreno JG. Reverse transcriptase polymerase chain reaction for prostate specific antigen in the management of prostate cancer. J Urol 1997;158:326-337.[CrossRef][ISI][Medline] [Order article via Infotrieve]
  15. Moreno JG, Gomella LG. Circulating prostate cancer cells detected by reverse transcription-polymerase chain reaction (RT-PCR): what do they mean?. Cancer Control 1998;5:507-512.[Medline] [Order article via Infotrieve]
  16. Ylikoski A, Karp M, Pettersson K, Lilja H, Lovgren T. Simultaneous quantification of human glandular kallikrein 2 and prostate-specific antigen mRNAs in peripheral blood from prostate cancer patients. J Mol Diagn 2001;3:111-122.[Abstract/Free Full Text]
  17. Xu LL, Stackhouse BG, Florence K, Zhang W, Shanmugam N, Sesterhenn IA, et al. PSGR, a novel prostate-specific gene with homology to a G protein-coupled receptor, is overexpressed in prostate cancer. Cancer Res 2000;60:6568-6572.[Abstract/Free Full Text]
  18. de Kok JB, Verhaegh GW, Roelofs RW, Hessels D, Kiemeney LA, Aalders TW, et al. DD3(PCA3), a very sensitive and specific marker to detect prostate tumors. Cancer Res 2002;62:2695-2698.[Abstract/Free Full Text]
  19. Magee JA, Araki T, Patil S, Ehrig T, True L, Humphrey PA, et al. Expression profiling reveals HPN overexpression in prostate cancer. Cancer Res 2001;61:5692-5696.[Abstract/Free Full Text]
  20. Autorino R, D’Armiento M, De Laurentiis M, De Placido S, Catalano G, Bianco AR, et al. Expression of epidermal growth factor receptor correlates with disease relapse and progression to androgen-independence in human prostate cancer. Clin Cancer Res 2002;11:3438-3444.
  21. Willman JH, Holden JA. Immunohistochemical staining for DNA topoisomerase II-{alpha} in benign, premalignant, and malignant lesions of the prostate. Prostate 2000;42:280-286.[CrossRef][ISI][Medline] [Order article via Infotrieve]
  22. Linja MJ, Savinainen KJ, Saramaki OR, Tammela TL, Vessella RL, Visakorpi T. Amplification and overexpression of androgen receptor gene in hormone-refractory prostate cancer. Cancer Res 2001;61:3550-3555.[Abstract/Free Full Text]
  23. Monne M, Croce CM, Yu H, Diamandis EP. Molecular characterization of prostate-specific antigen messenger RNA expressed in breast tumors. Cancer Res 1994;54:6344-6347.[Abstract/Free Full Text]



The following articles in journals at HighWire Press have cited this article:


Home page
 Clin. Cancer Res.Home page
J. S. de Bono, G. Attard, A. Adjei, M. N. Pollak, P. C. Fong, P. Haluska, L. Roberts, C. Melvin, M. Repollet, D. Chianese, et al.
Potential Applications for Circulating Tumor Cells Expressing the Insulin-Like Growth Factor-I Receptor
Clin. Cancer Res., June 15, 2007; 13(12): 3611 - 3616.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Cancer Res.Home page
D. R. Shaffer, M. A. Leversha, D. C. Danila, O. Lin, R. Gonzalez-Espinoza, B. Gu, A. Anand, K. Smith, P. Maslak, G. V. Doyle, et al.
Circulating Tumor Cell Analysis in Patients with Progressive Castration-Resistant Prostate Cancer
Clin. Cancer Res., April 1, 2007; 13(7): 2023 - 2029.
[Abstract] [Full Text] [PDF]

Home page
 CarcinogenesisHome page
M. Mimeault and S. K. Batra
Recent advances on multiple tumorigenic cascades involved in prostatic cancer progression and targeting therapies
Carcinogenesis, January 1, 2006; 27(1): 1 - 22.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Res.Home page
D. A. Smirnov, D. R. Zweitzig, B. W. Foulk, M. C. Miller, G. V. Doyle, K. J. Pienta, N. J. Meropol, L. M. Weiner, S. J. Cohen, J. G. Moreno, et al.
Global Gene Expression Profiling of Circulating Tumor Cells
Cancer Res., June 15, 2005; 65(12): 4993 - 4997.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Chem.Home page
K. Koyanagi, C. Kuo, T. Nakagawa, T. Mori, H. Ueno, A. R. Lorico Jr, H.-J. Wang, E. Hseuh, S. J. O'Day, and D. S.B. Hoon
Multimarker Quantitative Real-Time PCR Detection of Circulating Melanoma Cells in Peripheral Blood: Relation to Disease Stage in Melanoma Patients
Clin. Chem., June 1, 2005; 51(6): 981 - 988.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Cancer Res.Home page
W. J. Allard, J. Matera, M. C. Miller, M. Repollet, M. C. Connelly, C. Rao, A. G. J. Tibbe, J. W. Uhr, and L. W. M. M. Terstappen
Tumor Cells Circulate in the Peripheral Blood of All Major Carcinomas but not in Healthy Subjects or Patients With Nonmalignant Diseases
Clin. Cancer Res., October 15, 2004; 10(20): 6897 - 6904.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Data Supplement
Right arrow All Versions of this Article:
clinchem.2003.028563v1
50/5/826    most recent
Right arrow Submit an electronic Letter to
the Editor about this paper
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (17)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by O’Hara, S. M.
Right arrow Articles by Terstappen, L. W.M.M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by O’Hara, S. M.
Right arrow Articles by Terstappen, L. W.M.M.
Related Collections
Right arrow Molecular Diagnostics and Genetics
Right arrow Proteomics and Protein Markers
Right arrow Automation and Analytical Techniques

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS