Clinical Chemistry Siemens Point of Care - Urinalysis
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Clinical Chemistry 53: 2070-2077, 2007. First published October 5, 2007; 10.1373/clinchem.2007.091363
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow 091363.Supplemental Data
Right arrow All Versions of this Article:
clinchem.2007.091363v1
53/12/2070    most recent
Right arrow Alert me when this article is cited
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 Google Scholar
Google Scholar
Right arrow Articles by Hanke, M.
Right arrow Articles by Warnecke, J. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hanke, M.
Right arrow Articles by Warnecke, J. M.
Related Collections
Right arrow Molecular Diagnostics and Genetics
(Clinical Chemistry. 2007;53:2070-2077.)
© 2007 American Association for Clinical Chemistry, Inc.

Molecular Diagnostics and Genetics

Detailed Technical Analysis of Urine RNA-Based Tumor Diagnostics Reveals ETS2/Urokinase Plasminogen Activator to Be a Novel Marker for Bladder Cancer

Merle Hanke1,2, Ingo Kausch3, Gerlinde Dahmen4, Dieter Jocham3 and Jens M. Warnecke1,2,a

1 Kompetenzzentrum fuer Drug Design und Target Monitoring, Luebeck, Germany.
2 Institut fuer Molekulare Medizin,3 Klinik und Poliklinik für Urologie, and4 Institut fuer Medizinische Biometrie und Statistik, UK-S-H, Campus Luebeck, Luebeck, Germany.

aAddress correspondence to this author at: Institut fuer Molekulare Medizin, Universitaetsklinikum Schleswig-Holstein, Campus Luebeck, Ratzeburger Allee 160, 23538 Luebeck, Germany. Fax 49-451-500-2729; e-mail warnecke{at}imm.uni-luebeck.de.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Background: The noninvasive detection of RNA tumor markers in body fluids represents an attractive diagnostic option, but diagnostic performance of tissue-derived markers is often poorer when measured in body fluids rather than in tumors. We aimed to develop a procedure for measurement of tumor RNA in urine that would minimize donor-dependent influences on the results.

Methods: RNA isolated from urinary cell pellet, cell-depleted fraction, and whole urine was quantified by reverse transcription quantitative–PCR. The donor-dependent influence of urine background on individual steps of the standardized procedure was analyzed using an external RNA standard. Using a test set of samples from 61 patients with bladder cancer and 37 healthy donors, we compared 4 putative RNA tumor markers identified in whole urine with 5 established, tissue-derived RNA tumor markers for the detection of bladder cancer.

Results: Of the markers analyzed by this system, the RNA ratio of v-ets erythroblastosis virus E26 oncogene homolog 2 (avian; ETS2) to urokinase plasminogen activator (uPA) enabled the most specific (100%) and sensitive (75.4%) detection of bladder cancer from whole urine, with an area under the curve of 0.929 (95% CI 0.882–0.976).

Conclusions: The described methodology for RNA marker detection in urine appears to be clinically applicable. The ratio of ETS2 mRNA to uPA mRNA in urine is a potential marker for bladder cancer.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Gene expression signatures of tumor cells have been investigated for a variety of human cancers(1) to gain insight into tumor development and progression. Microarray gene expression analysis by Dyrskjøt et al.(2), for example, revealed stage-specific gene clusters whose expression pattern allowed a classification of bladder cancer into 3 distinct histopathological groups: noninvasive Ta tumors, invasive T1 tumors, and T2–T4 tumors. However, validation of these consolidated findings via reverse transcription quantitative–PCR (RT-qPCR)1 failed although the same sample material was used(3). The statistical basis, the RNA isolation procedure, and the stability of PCR amplicons, as well as the amount and particularly the quality of RNA, are some factors that influence the outcome of gene expression studies. The development of one general standard operating protocol that should be used without restrictions in any laboratory institution—including the sampling, isolation, quantification (use of identical PCR amplicons), and normalization of RNA—is therefore mandatory to allow a comparison of gene expression data.

To avoid differences in processing, the use of body fluids without any pretreatment (e.g., centrifugation) is most desirable for a noninvasive diagnostic test. This application is challenging, since body fluids from different donors differ with respect to the amount, origin, and integrity of cells and nucleic acids. The cells may contain RNA that is fragmented by necrotic and apoptotic processes that also give rise to cell-free RNA in urine(4). Cell-free RNA is resistant to the activity of urinary RNases because of packaging into apoptotic bodies. In addition, cell death–independent processes have been described for the generation of extracellular RNA(5). To avoid exclusion of patients, however, a diagnostic routine application based on body fluids ideally is not dependent on the integrity of the RNA.

Despite these obstacles, the detection of RNA tumor markers in urine has been reported as an emerging tool for noninvasive tumor diagnosis(6). One of the most frequently studied markers for bladder cancer is the mRNA of human telomerase reverse transcriptase (hTERT),2 because its concentration correlates with telomerase activity, which is absent in most human somatic cells but detectable in 85% of human tumors(7). RNA extracts from tumor material(8)(10), bladder washings(11)(12), and exfoliated cells(13) have been used for RT-qPCR–based hTERT mRNA quantification for the detection of bladder cancer, resulting in reported diagnostic sensitivities of 50% to 100%. When urine sediment was used as sample material in a larger prospective study, however, hTERT mRNA expression was often found to be near the detection limit, resulting in a sensitivity of 55%(14).

In this study, we aimed to develop a RT-qPCR–based test of whole urine for robust and sensitive diagnosis of bladder cancer in a clinical setting. Toward this goal, we 1st analyzed factors influencing the RNA isolation and quantification procedure. We then applied a standardized process to compare RNA tumor markers originating from our screening experiments with established markers for bladder cancer.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
clinical samples
The study was approved by the local research ethics committee. All 139 samples were obtained with written informed consent of the participants. For the determination of assay characteristics, urine of 21 donors (13 men, 8 women; median age 66 years) was used for preparation of urine fractions. For the investigation of daytime profiles, spontaneously voided urine of 10 participants (6 men, 4 women; median age 31 years) was collected in the morning (1st void of the day), at midday, and in the evening. We selected 10 further urine samples for enrichment experiments (5 men, 5 women; median age 30 years).

To investigate selected tumor markers, we used whole urine from 98 donors. For a control group, we chose healthy donors who best matched the age and sex distribution of the bladder cancer group. Detailed clinical information is shown in Table 1 in the Data Supplement that accompanies the online version of this article athttp://www.clinchem.org/content/vol53/issue12. An overview of the experimental procedures is given in Fig. 1 .


Figure 1
View larger version (45K):
[in this window]
[in a new window]

  		Figure 1. Experimental setup.

Urine was collected from 139 participants. Based on availability and sample volume, 41 attendants were selected for the assay characterization (left box). To determine the influence of urine composition on individual steps of the process, a defined amount of RNALUC was added as "spike" at individual steps and subsequently quantified by RT-qPCR (black arrows). We collected 98 urine samples for the investigation of tumor markers (gray box, right side). Hexagonal boxes denote standard operating procedure-based steps. RNALUC was added to each sample as external standard before RNA isolation (white arrow).

preparation of the external rna standard (rnaluc)
We synthesized truncated luciferase RNA (RNALUC) by in vitro runoff transcription using T7 RNA-polymerase. The T7-transcript was prepared from a purified EcoRV-restricted pet24a(+) vector (Promega) fragment containing the Photinus pyralis luciferase gene. The reaction mixture (50 µL) for T7 transcription contained 1x reaction buffer (MBI Fermentas), 2 mmol/L each nucleotide triphosphate, 40 units RNase inhibitor (Ambion), 30 units T7 RNA-Polymerase (MBI Fermentas), and 0.8 µg DNA template. After transcription of RNALUC (1570 nucleotides), the DNA template was digested using DNase I (Ambion). The reaction mixture was gel-filtrated through a NAP-5 column (Amersham), extracted by chloroform-phenol, and precipitated with ethanol. Integrity and size of RNALUC were controlled by denaturing agarose gel electrophoresis. RNA concentration was determined by absorbance at 260 nm (1 A260 = 40 mg RNA/L) of a dilution series of the purified RNALUC in vitro transcript using NanoDrop ND-1000 UV-Vis Spectrophotometer (NanoDrop Technologies). We used mean values to calculate RNALUC copy numbers.

processing of urine samples
Spontaneously voided urine was adjusted to a final concentration of 3 mol/L guanidinium thiocyanate (GTC), 0.025 mol/L sodium acetate, and 0.25% N-lauroylsarcosine. For the monovette-based sample collection, we transferred 3.54 g GTC powder into a 10-mL urine monovette (Sarstedt) and adjusted the stamp to the 7-mL mark of the monovette. Further handling of the monovette was performed according to manufacturer’s instructions. After the GTC powder was dissolved in the collected urine, the sample was transported to the analytic laboratory. Adjustment to 0.025 mol/L sodium acetate and 0.25% N-lauroylsarcosine and addition of 1 mol/L HEPES (pH 7) to a final volume of 10 mL were carried out before storage at –80 °C.

preparation of urinary cell pellet
The cellular fraction was obtained by centrifugation of 3 mL voided urine at 400g for 5 min. We discarded the supernatant and dissolved the pellet in a special lysis buffer(15). The lysate was frozen in liquid nitrogen and stored at –80 °C until RNA isolation.

preparation of cell-depleted urine
Cell-depleted urine was obtained by passing voided urine through a 5-µm filter (Sartorius). The urine filtrate was treated as described for whole urine.

standard procedure for the isolation and reverse transcription of total rna
Before RNA isolation, 107 copies of RNALUC were added. Total RNA was isolated using the RNeasy Midi Kit (Qiagen) according to the manufacturer’s instructions, except that special lysis buffer was used instead of RLT buffer. The procedure included an on-column digestion of genomic DNA with DNase I. RNA was eluted twice with 160 µL nuclease-free H2O and lyophilized. The RNA pellets were resolved in 20 µL nuclease-free H2O. We used 10 µL RNA extract for cDNA synthesis and non-reverse transcription reaction. Additionally, 107 copies of RNALUC were directly reverse-transcribed to verify applied copies. A detailed description of the cDNA synthesis protocol is provided in the online Data Supplement.

quantitative pcr and data analysis
TaqMan-based quantification of cDNA was always run in triplicate, and the non–reverse transcription reaction was run in duplicate. The protocol is described in detail in the online Data Supplement, together with primer sequences and amplicon characteristics (see Table 2 in the online Data Supplement). We performed data analysis with SDS 2.1 software (Applied Biosystems). The threshold cycle values of amplified targets were transformed into absolute RNA copy numbers using standard curves, allowing an absolute quantification of target RNA copy numbers.

assay characterization
RT-qPCR.
RNALUC (107 copies) were directly reverse-transcribed, and we quantified RNALUC cDNA by use of qPCR in triplicate. We performed independent qPCR assay runs using a standard curve for every run to determine the interrun variation of RNALUC quantification. To determine the donor-dependent influence of the urine background on the qPCRs, 5 x 105 copies of RNALUC were added to RNA extracts from different donors. Copy numbers of RNALUC were determined by the standard RT-qPCR procedure.

Interrun imprecision of the complete RNA isolation and quantification procedure.
We added 107 copies of in vitro–transcribed RNALUC to 8 aliquots of spontaneously voided urine of a single donor after adjustment to 3 mol/L GTC. For each sample, we determined copy numbers of RNALUC, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), ribosomal protein large P0 (RPLP0), and ubiquitin C (UBC) by independent isolation and quantification procedures.

Intradonor variability.
For daytime profiles, spontaneously voided urine was collected in the morning, at midday, and in the evening. We added 107 copies of RNALUC to whole urine after adjustment to 3 mol/L GTC. RNA was isolated in duplicate and quantified in triplicate via RT-qPCR. Copy numbers of RNALUC, GAPDH, RPLP0, and UBC were determined by standardized RT-qPCR procedure.

Interdonor variability.
For the investigation of donor-dependent influence of urine background on RNA recovery, we added 107 copies of RNALUC to whole urine, cell-depleted urine, or the cellular fraction before RNA isolation and after adjustment to 3 mol/L GTC. RNA was isolated in duplicate, and we determined copy numbers of RNALUC by the standard procedure for RNA isolation and quantification. In addition, we determined copy numbers of endogenous GAPDH.

statistical analysis and software
The statistical analysis of RNA marker ratios is described in detail in the online Data Supplement. The diagnostic power of selected markers was analyzed by ROC curves. For the comparison of 2 datasets (comparison of RNALUC yields and housekeeping gene ratios), we used the nonparametric paired Wilcoxon test. For all statistical tests, 2-sided P values ≤0.05 were considered statistically significant.

Statistical analyses were performed using SAS statistical software (version 9.1, SAS Institute), SPSS statistical software (version 13.0, SPSS GmbH Software), R language and environment for statistical computing (version 2.3.1, R Foundation for Statistical Computing), and STATA/SE statistical software (version 9.1, Stata).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
RNA tumor marker quantification in a clinical setting requires a characterized process to allow precise and reproducible processing. We developed a monovette-based system containing GTC powder to ensure fast, standardized processing of urine samples. For the systematic characterization of the complete RNA isolation and quantification process, an external RNA standard was added before RNA isolation and quantified by TaqMan-based RT-qPCR. Reproducible processing was achieved using standard operating procedures for each step, including sample collection. An overview of the experiments that were performed to determine the assay characteristics is shown in Fig. 1Up (left panel).

assay characterization
The assay characteristics are summarized in Tables 1 and 2 .


View this table:
[in this window]
[in a new window]

  		Table 1. Interrun imprecision of the RNA isolation and quantification process.


View this table:
[in this window]
[in a new window]

  		Table 2. Donor-dependent influence on the RNA isolation and quantification process.

interrun imprecision
Interrun imprecision (CV) for the isolation and quantification process of RNALUC from 8 aliquots of whole urine was 0.16 (Table 1Up ). Endogenous RNAs for GAPDH, RPLP0, and UBC were detected with a comparable imprecision (CVGAPDH 0.15; CVRPLP0 0.18; CVUBC 0.16).

interdonor variability
We next determined the donor-dependent influence of urine background on assay imprecision (Table 2Up ). The overall RNALUC yield of the RNA isolation and quantification process was quantified for different fractions (cell-depleted, cellular fraction, and whole urine) prepared from urine of different donors (n = 21). The yield of RNALUC was best for the cellular fraction and lowest for cell-depleted urine (see Fig. 1 in the online Data Supplement). RNALUC yields for the cellular fraction (mean 27%) were significantly higher (Wilcoxon test P = 0.002) than for whole urine (mean 16%). The highest variability is introduced by the RNA isolation step, since downstream steps turned out to be robust (CVRT-qPCR 0.14; Table 2Up ). Accordingly, for all subsequent experiments, RNALUC-normalized copy numbers were determined to allow a comparison of urine samples from different donors.

intradonor variability
The amount of endogenous housekeeping genes such as GAPDH RNA in urine varied by 4 orders of magnitude (Table 2Up ). To allow a comparison between individual samples, we calculated the ratios of housekeeping gene RNAs. Systematic differences in RNA composition were analyzed by investigation of housekeeping gene ratios at 3 times of the day (n = 10).

Ratios were calculated from RNA copy numbers of GAPDH, RPLP0, and UBC. The intradonor variability for GAPDH:RPLP0 (Fig. 2A ) was smaller (mean CV 0.25; range 0.09–0.61) than for GAPDH:UBC (mean CV 0.41; range 0.07–0.71). Pairwise comparison using nonparametric Wilcoxon test provides no evidence for a significant influence of time of day on the GAPDH:UBC ratio (P = 0.13). Only the comparison of the GAPDH:RPLP0 ratio from morning and midday showed a significant difference (P = 0.05). To obtain conclusive data independently of patients’ compliance with instructions, we used only spontaneously voided urine (midday) instead of morning urine for further experiments.


Figure 2
View larger version (41K):
[in this window]
[in a new window]

  		Figure 2. Detection of housekeeping gene RNA in whole urine.

(A), intraday fluctuations of GAPDH:RPLP0 RNA ratios in whole urine. Morning, midday, and evening urine was collected from 10 donors, and total RNA was isolated in duplicate for each donor. RNA was reverse-transcribed, and copy numbers were determined in triplicate by qPCR. Ratios were calculated for each experiment separately. Bars represent mean values of 2 independent RNA isolations and quantification procedures. SDs are depicted by error bars. (B), GAPDH and UBC RNA copy numbers/mL urine for whole urine and cell-depleted fractions of 21 donors as evaluated by qPCR. Total RNA isolation and the reverse transcription reactions were performed in duplicate; the qPCR was run in triplicate. Detected copy numbers were normalized with RNALUC to compensate for variations in total RNA isolation. White bars indicate GAPDH copy numbers/mL whole urine; white hatched bars indicate UBC copy numbers/mL whole urine. GAPDH copy numbers in the cell-depleted fraction are represented by black bars and UBC copy numbers by gray hatched bars. Bars represent mean values of 2 independent RNA isolations and quantification procedures. SDs are depicted by error bars. The term RNA instead of mRNA is used to indicate that RNA integrity was not determined.

rna composition in urine fractions
The GAPDH:UBC ratios in total differed significantly from the cellular fraction (Wilcoxon test, P = 0.01; n = 21) in contrast to the GAPDH:RPLP0 RNA ratios (Wilcoxon test, P = 0.07; n = 21). The cell-depleted fraction differed significantly from whole urine and the cellular fraction for the 2 RNA ratios (Wilcoxon test, P <0.001 for GAPDH:UBC and GAPDH:RPLP0, n = 21).

We next determined the amount of cell-free RNA that contributes to the detected RNA in whole urine (Fig. 2BUp ). Quantification of housekeeping gene RNA in the cell-depleted urine fractions (n = 21) revealed high amounts of GAPDH RNA (mediancell-depleted = 1.06 x 106 copy numbers/mL; medianwhole urine = 1.89 x 106 copy numbers/mL) and UBC RNA (mediancell-depleted = 5.12 x 105 copy numbers/mL; medianwhole urine = 5.58 x 106 copy numbers/mL). In some samples, the copy numbers of GAPDH RNA detected in whole urine consisted almost exclusively of cell-free RNA (see Fig. 2BUp , samples 2, 11, 15, and 21). The cell-free RNA portion of the whole urine fraction was higher for GAPDH RNA (mean 46.9%) compared with UBC RNA (mean 16.9%).

These results provided a rationale for the reinvestigation of different cell-based tumor markers for their applicability using whole urine. Their diagnostic performance was analyzed together with markers derived from RT-qPCR–based screening of RNA isolated from whole urine (data not shown).

investigation of rna tumor markers in whole urine
RNA was isolated from urine of 37 healthy donors and 61 patients with bladder cancer. A set of 8 different RNA tumor markers (Table 3 ) containing putative as well as established bladder cancer markers and the housekeeping gene GAPDH were quantified as specified in Fig. 1Up , right panel.


View this table:
[in this window]
[in a new window]

  		Table 3. Selected targets for RNA tumor marker analysis in whole urine.

Multivariate analysis of individual marker ratios was performed to test the ability to separate the 98 donors into those with cancer and those without. We selected 3 important marker ratios using classification trees. In a logistic regression model with backward selection including the selected variables and all interactions, only the ratio of v-ets erythroblastosis virus E26 oncogene homolog 2 (avian; ETS2) to urokinase plasminogen activator (uPA) allowed a statistically significant separation. To confirm the result of the logistic regression model, we performed 2 nonparametric diagonal linear discriminant analyses, one with the 3 important marker ratios and another with ETS2:uPA. The classification results of the diagonal linear discriminant analyses were not different (P >0.05). Therefore, the RNA ratio of ETS2:uPA was shown to be the only independent marker for the detection of bladder cancer. The ETS2:uPA ratios in whole urine of patients with bladder cancer and healthy controls are presented in Fig. 3A . ROC curves were calculated that represent the diagnostic power of the RNA marker combinations. ROC analysis (Fig. 3B ) of the study population by use of the ETS2:uPA ratio revealed an area under the curve (AUC) of 0.929 (95% CI 0.882–0.976), indicating the strong diagnostic power of the test. Setting the specificity at 100% (cutoff value 0.96), a sensitivity of 75.4% was achieved. For the group of low-grade tumors, we determined a sensitivity of 53.9%. Sensitivity can be enhanced to 79.9% for low-grade tumors and 89.1% for the group of all tumors, if the specificity is reduced to 89.2%. As demonstrated by the ROC analysis in Fig. 3B , the use of GAPDH-normalized ETS2 RNA and GAPDH-normalized uPA RNA as individual markers did not turn out to be informative. The AUC values for the other marker combinations tested are listed in Table 3 in the online Data Supplement.


Figure 3
View larger version (17K):
[in this window]
[in a new window]

  		Figure 3. Diagnostic performance of the ETS2:uPA RNA ratio.

(A), box plot of ETS2:uPA RNA ratios in whole urine from healthy donors compared with patients with bladder cancer stratified according to tumor grade. The line inside each box denotes median, whereas the boxes mark 25th and 75th percentiles. Error bars mark 5th and 95th percentiles. Symbols indicate outlying data points. (B), ROC curves for selected tumor markers on the basis of 37 healthy donors and 61 patients with bladder cancer. The thick black line represents the ROC of the ETS2:uPA RNA ratio, the gray line indicates the ROC for GAPDH-normalized ETS2 RNA, and the thin black line indicates the ROC of GAPDH-normalized uPA RNA.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
RT-qPCR–based marker detection in urine sediment allows a noninvasive, sensitive, and specific detection of bladder cancer. The feasibility of the diagnostic application has been demonstrated by prospective studies(14), but different authors report substantially different diagnostic accuracies, even if the same markers were used. Using the monovette-based sampling of whole urine presented here provides several technical advantages.

The initial step of sampling, i.e., the transfer of urine into the monovette containing GTC powder, does not require any laboratory equipment. It is therefore simple to implement into hospital routine, thereby avoiding increased processing time. Inactivation of RNA-degrading enzymes occurs rapidly by the dissolved GTC. The urine samples were thus stabilized immediately, transported, and frozen within 6 h after voiding. This is an important step toward the development of a standard method for the sampling of urine, comparable to EDTA monovettes for the collection of blood samples. To obtain a high informative rate, the RNA isolates have to be concentrated. We have recently demonstrated the feasibility of ethanol precipitation for this purpose(15). To allow maximum standardization, this step was substituted by lyophilization.

The use of the "1st void of the day" (morning) urine is quite common in diagnostic studies because of the higher content of nucleic acids. Investigation of intraday variance indicated that the RNA ratio of GAPDH:RPLP0 showed a significant difference between morning and midday urine. This RNA ratio–dependent intraday variation indicates an advantage of spontaneously voided urine (midday), since the results are independent of the patient’s compliance with instructions.

Using whole urine, differences in yield of the complete RNA isolation and quantification process ranged from 0.7% to 47.9%. Accordingly, RNA tumor markers with low abundance have to be excluded from this application. This holds particularly for the analysis of hTERT RNA, in which our analysis indicated insufficient abundance for reliable quantification (data not shown). This is in line with results from the prospective study of Weikert et al.(14), in which hTERT mRNA expression in exfoliated cells was often found to be near the detection limit. In contrast to hTERT mRNA, ETS2 and uPA RNA copy numbers/mL were sufficiently high in whole urine (medianETS2 = 4.64 x 105 copy numbers/mL; medianuPA= 4.84 x 105 copy numbers/mL).

The emerging interest in using RNA-tumor marker detection in whole urine for the diagnosis of other tumor entities(16)(17) accentuates the need to investigate the variability and source of RNA species in whole urine. Our experiments indicate that RNA tumor markers derived from gene expression analysis of cells or tumor material (e.g., hTERT, UPK1A, HTATIP2) cannot be transferred unconditionally to RT-qPCR–based analysis of whole urine. Based on our data, one major reason is the presence of cell-free RNA in significant amounts. When GAPDH is used for normalization, the high concentrations of cell-free GAPDH RNA, in addition to the well-known overexpression of GAPDH RNA in tumor tissue, mask the diagnostic power of cellular markers, as demonstrated here for uPA/GAPDH (AUC 0.359; see Fig. 3BUp ). In contrast to extracellular DNA in urine(18)(19), the presence of specific extracellular RNA has not been reported. In addition to apoptosis and necrosis, cell death–independent generation of extracellular RNA has been demonstrated(5).

RT-qPCR array–based screening experiments (data not shown) using whole urine and the subsequent evaluation of putative tumor markers revealed the ratio of ETS2:uPA to be a suitable marker for the detection of bladder cancer. When the specificity was set at 100%, a sensitivity of 75.4% was achieved. This result outperforms the sensitivity (28%–76%) of urinary cytology(20) as the standard noninvasive method and most of the molecular bladder cancer markers when the specificity is set at 100%(21)(22). Stratification by tumor grade (Fig. 3AUp ) indicated a higher sensitivity for high-grade tumors (81.3%, grade 2–3), an observation that has been described for other urine-based RNA markers(14)(23)(24).

ETS2 is a member of the ETS family of transcription factors that regulate a variety of biological processes(25). ETS2 RNA and protein concentrations were found to be increased in breast cancer samples compared with normal tissue(26). Buggy et al.(26) showed a significant correlation of protein expression between ETS2 and uPA, which contains ETS2 binding sites in its promoter region. The role of the uPA system in tumorigenesis has been intensively studied(27). In bladder cancer, urinary protein concentrations of uPA [e.g., (28)] as well as the uPA mRNA content in tissue [e.g.,(29)], have been suitable for tumor diagnostics. The experimental data of our study population indicate a higher ETS2 RNA concentration compared with uPA in the case of bladder cancer, resulting in an increased ETS2:uPA RNA ratio.

In conclusion, we describe a fully standardized process for the isolation and quantification of RNA tumor markers from total urine. The new tumor marker ratio of ETS2:uPA is promising for the detection of bladder cancer with high specificity and sensitivity in a clinical setting.


   Acknowledgments
 
Grant/funding support: M.H. is supported by European Union Grant 3ASH2000/32/19535.

Financial disclosures: None declared.

Acknowledgments: We thank Drs. C. Höppner, M. Horn, T. Menke, S. Thomas, and J. Träder for collecting samples, A. Krüger and B. Thode for technical assistance, and G. Sczakiel for critically reading the manuscript.


   Footnotes
 
1 Nonstandard abbreviations: RT-qPCR, reverse transcription quantitative–PCR; GTC, guanidinium thiocyanate; AUC, area under the curve.

2 Human genes: hTERT, human telomerase reverse transcriptase; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; RPLP0, ribosomal protein, large, P0; UBC, ubiquitin C; ETS2, v-ets erythroblastosis virus E26 oncogene homolog 2 (avian); uPA, urokinase plasminogen activator; HTATIP2, HIV-1 Tat interactive protein 2, 30kDa; UPK1A, uroplakin 1A.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Wadlow R, Ramaswamy S. DNA microarrays in clinical cancer research. Curr Mol Med 2005;5:111-120.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  2. Dyrskjøt L, Thykjaer T, Kruhoffer M, Jensen JL, Marcussen N, Hamilton-Dutoit S, et al. Identifying distinct classes of bladder carcinoma using microarrays. Nat Genet 2003;33:90-96.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  3. Schultz IJ, Wester K, Straatman H, Kiemeney LA, Babjuk M, Mares J, et al. Prediction of recurrence in Ta urothelial cell carcinoma by real-time quantitative PCR analysis: a microarray validation study. Int J Cancer 2006;119:1915-1919.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  4. Bryzgunova OE, Skvortsova TE, Kolesnikova EV, Starikov AV, Rykova EY, Vlassov VV, et al. Isolation and comparative study of cell-free nucleic acids from human urine. Ann N Y Acad Sci 2006;1075:334-340.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  5. Stroun M, Anker P, Beljanski M, Henri J, Lederrey C, Ojha M, et al. Presence of RNA in the nucleoprotein complex spontaneously released by human lymphocytes and frog auricles in culture. Cancer Res 1978;38:3546-3554.[Abstract/Free Full Text]
  6. Goessl C. Noninvasive molecular detection of cancer: the bench and the bedside. Curr Med Chem 2003;10:691-706.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  7. Nakamura TM, Morin GB, Chapman KB, Weinrich SL, Andrews WH, Lingner J, et al. Telomerase catalytic subunit homologs from fission yeast and human. Science 1997;277:955-959.[Abstract/Free Full Text]
  8. Ito H, Kyo S, Kanaya T, Takakura M, Inoue M, Namiki M. Expression of human telomerase subunits and correlation with telomerase activity in urothelial cancer. Clin Cancer Res 1998;4:1603-1608.[Abstract]
  9. De Kok JB, Schalken JA, Aalders TW, Ruers TJ, Willems HL, Swinkels DW. Quantitative measurement of telomerase reverse transcriptase (hTERT) mRNA in urothelial cell carcinomas. Int J Cancer 2000;87:217-220.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  10. de Kok JB, Ruers TJ, van Muijen GN, van Bokhoven A, Willems HL, Swinkels DW. Real-time quantification of human telomerase reverse transcriptase mRNA in tumors and healthy tissues. Clin Chem 2000;46:313-318.[Abstract/Free Full Text]
  11. Isurugi K, Suzuki Y, Tanji S, Fujioka T. Detection of the presence of catalytic subunit mRNA associated with telomerase gene in exfoliated urothelial cells from patients with bladder cancer. J Urol 2002;168:1574-1577.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  12. de Kok JB, van Balken MR, Roelofs RW, van Aarssen YA, Swinkels DW, Klein Gunnewiek JM. Quantification of hTERT mRNA and telomerase activity in bladder washings of patients with recurrent urothelial cell carcinomas. Clin Chem 2000;46:2003-2007.[Free Full Text]
  13. Bowles L, Bialkowska-Hobrzanska H, Bukala B, Nott L, Razvi H. A prospective evaluation of the diagnostic and potential prognostic utility of urinary human telomerase reverse transcriptase mRNA in patients with bladder cancer. Can J Urol 2004;11:2438-2444.[Medline] [Order article via Infotrieve]
  14. Weikert S, Krause H, Wolff I, Christoph F, Schrader M, Emrich T, et al. Quantitative evaluation of telomerase subunits in urine as biomarkers for noninvasive detection of bladder cancer. Int J Cancer 2005;117:274-280.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  15. Menke TB, Boettcher K, Kruger S, Kausch I, Boehle A, Sczakiel G, et al. Ki-67 protein concentrations in urothelial bladder carcinomas are related to Ki-67-specific RNA concentrations in urine. Clin Chem 2004;50:1461-1463.[Free Full Text]
  16. Groskopf J, Aubin SM, Deras IL, Blase A, Bodrug S, Clark C, et al. APTIMA PCA3 molecular urine test: development of a method to aid in the diagnosis of prostate cancer. Clin Chem 2006;52:1089-1095.[Abstract/Free Full Text]
  17. Bai VU, Kaseb A, Tejwani S, Divine GW, Barrack ER, Menon M, et al. Identification of prostate cancer mRNA markers by averaged differential expression and their detection in biopsies, blood, and urine. Proc Natl Acad Sci U S A 2007;104:2343-2348.[Abstract/Free Full Text]
  18. Tong YK, Lo YM. Diagnostic developments involving cell-free (circulating) nucleic acids. Clin Chim Acta 2006;363:187-196.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  19. Mao L, Schoenberg MP, Scicchitano M, Erozan YS, Merlo A, Schwab D, et al. Molecular detection of primary bladder cancer by microsatellite analysis. Science 1996;271:659-662.[Abstract]
  20. Konety BR. Molecular markers in bladder cancer: a critical appraisal. Urol Oncol 2006;24:326-337.[Web of Science][Medline] [Order article via Infotrieve]
  21. Wang XS, Zhang Z, Wang HC, Cai JL, Xu QW, Li MQ, et al. Rapid identification of UCA1 as a very sensitive and specific unique marker for human bladder carcinoma. Clin Cancer Res 2006;12:4851-4858.[Abstract/Free Full Text]
  22. Weikert S, Christoph F, Schrader M, Krause H, Miller K, Muller M. Quantitative analysis of survivin mRNA expression in urine and tumor tissue of bladder cancer patients and its potential relevance for disease detection and prognosis. Int J Cancer 2005;116:100-104.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  23. Christoph F, Muller M, Schostak M, Soong R, Tabiti K, Miller K. Quantitative detection of cytokeratin 20 mRNA expression in bladder carcinoma by real-time reverse transcriptase-polymerase chain reaction. Urology 2004;64:157-161.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  24. Schultz IJ, Kiemeney LA, Willems JL, Swinkels DW, Witjes JA, de Kok JB. Survivin and MKI67 mRNA expression in bladder washings of patients with superficial urothelial cell carcinoma correlate with tumor stage and grade but do not predict tumor recurrence. Clin Chem 2006;52:1440-1442.[Free Full Text]
  25. Foulds CE, Nelson ML, Blaszczak AG, Graves BJ. Ras/mitogen-activated protein kinase signaling activates Ets-1 and Ets-2 by CBP/p300 recruitment. Mol Cell Biol 2004;24:10954-10964.[Abstract/Free Full Text]
  26. Buggy Y, Maguire TM, McDermott E, Hill AD, O’Higgins N, Duffy MJ. Ets2 transcription factor in normal and neoplastic human breast tissue. Eur J Cancer 2006;42:485-491.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  27. Duffy MJ, Duggan C. The urokinase plasminogen activator system: a rich source of tumour markers for the individualised management of patients with cancer. Clin Biochem 2004;37:541-548.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  28. Casella R, Shariat SF, Monoski MA, Lerner SP. Urinary levels of urokinase-type plasminogen activator and its receptor in the detection of bladder carcinoma. Cancer 2002;95:2494-2499.[Medline] [Order article via Infotrieve]
  29. Bhuvarahamurthy V, Schroeder J, Denkert C, Kristiansen G, Schnorr D, Loening SA, et al. In situ gene expression of urokinase-type plasminogen activator and its receptor in transitional cell carcinoma of the human bladder. Oncol Rep 2004;12:909-913.[Web of Science][Medline] [Order article via Infotrieve]
  30. Reed JC. Proapoptotic multidomain Bcl-2/Bax-family proteins: mechanisms, physiological roles, and therapeutic opportunities. Cell Death Differ 2006;13:1378-1386.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  31. Le Bras M, Rouy I, Brenner C. The modulation of inter-organelle cross-talk to control apoptosis. Med Chem 2006;2:1-12.[CrossRef][Medline] [Order article via Infotrieve]
  32. Ito M, Jiang C, Krumm K, Zhang X, Pecha J, Zhao J, et al. TIP30 deficiency increases susceptibility to tumorigenesis. Cancer Res 2003;63:8763-8767.[Abstract/Free Full Text]
  33. Hsu T, Trojanowska M, Watson DK. Ets proteins in biological control and cancer. J Cell Biochem 2004;91:896-903.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  34. de Kok JB, Roelofs RW, Giesendorf BA, Pennings JL, Waas ET, Feuth T, et al. Normalization of gene expression measurements in tumor tissues: comparison of 13 endogenous control genes. Lab Invest 2005;85:154-159.[Web of Science][Medline] [Order article via Infotrieve]
  35. Mistry SJ, Atweh GF. Stathmin expression in immortalized and oncogene transformed cells. Anticancer Res 1999;19:573-577.[Web of Science][Medline] [Order article via Infotrieve]
  36. Kageyama S, Yoshiki T, Isono T, Tanaka T, Kim CJ, Yuasa T, et al. High expression of human uroplakin Ia in urinary bladder transitional cell carcinoma. Jpn J Cancer Res 2002;93:523-531.[CrossRef][Web of Science]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow 091363.Supplemental Data
Right arrow All Versions of this Article:
clinchem.2007.091363v1
53/12/2070    most recent
Right arrow Alert me when this article is cited
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 Google Scholar
Google Scholar
Right arrow Articles by Hanke, M.
Right arrow Articles by Warnecke, J. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hanke, M.
Right arrow Articles by Warnecke, J. M.
Related Collections
Right arrow Molecular Diagnostics and Genetics

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS