SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References
  9. Supporting Information

A new diagnostic marker for urothelial carcinoma (UC) is needed to avoid painful cystoscopy during the initial diagnosis and follow-up period. However, the current urine markers are useless because of the low sensitivities and specificities for UC detection. MiR-96 and miR-183 were differentially upregulated microRNA in our previous microRNA screening for UC. The expression levels of miR-96 and miR-183 in the urine samples were significantly higher in 100 UC than in healthy controls (miR-96, P = 0.0059; and miR-183, = 0.0044). The receiver-operating characteristic curve analyses demonstrated that each microRNA had good sensitivity and specificity for distinguishing UC patients from non-UC patients (miR-96, 71.0% and 89.2%; and miR-183, 74.0% and 77.3%). Our cohort included 78 UC patients who had undergone urinary cytology. MiR-96 was positively detected in 27 of 44 patients who had had a “negative” urinary cytology diagnosis. We combined the miR-96 detection data with the urinary cytology data, and diagnosed 61 of 78 cases as UC; sensitivity rose from 43.6% to 78.2%. We found significant stepwise increases in miR-96 and miR-183 expression with advancing tumor grade (miR-96, = 0.0057; and miR-183, = 0.0036) and pathological stage (miR-96, = 0.0332; and miR-183, = 0.0117). The expression levels of the microRNA were significantly lower in urine collected after surgery (miR-96, = 0.0241; and miR-183, = 0.0045). In conclusion, miR-96 and miR-183 in urine are promising tumor markers for UC. In particular, miR-96 may be a good diagnostic marker in combination with urinary cytology. (Cancer Sci 2011; 102: 522–529)

Urothelial carcinoma (UC) is among the five most common malignancies worldwide, and it is the second most common tumor of the genitourinary tract and the second most common cause of death in patients with genitourinary tract malignancies.(1) The current standard of diagnostic tool for UC depends on urethro-cystoscopy. This approach is costly, invasive and uncomfortable. The endoscopic approach using nephro-ureteroscope is also invasive for patients with upper urinary tract UC (renal pelvic and ureter UC). Urinary cytology is a reliable urine marker for UC diagnosis because of its high specificity (90–95%). In contrast, it has low sensitivity (30–40%) and patients are forced to undergo a painful cystoscopy to confirm diagnoses.(2) For these reasons, many new urine-based tests for UCC have been developed, and Bladder Tumor Antigen (BTA), Nuclear Matrix Protein 22 (NMP22), Urine fibrin fibrinogen Degradation Products (FDP), ImmunoCyt and FISH (UroVysion), etc., have been approved for clinical use.(3,4) However, the specificities of these new urine markers are rather low (60–80%) in comparison with urinary cytology, although they have higher sensitivities (50–70%). This implies that specificity may come at the cost of sensitivity, conventional urinary cytology being a good example of this.(3) Because of their insufficient sensitivity and specificity, none of the new urine markers can replace cystoscopy or urinary cytology as of now.(5,6) Hence, patients with suspected UC continue to be subjected to painful cystoscopy during the initial diagnosis and follow-up examination after undergoing an endoscopic surgery, including trans-urethral resection of bladder tumor (TUR-BT). A new useful diagnostic marker is really needed in UC patients.

MicroRNA are small non-coding RNA of 20–22 nucleotides and they are involved in crucial biological processes, including development, differentiation, apoptosis and proliferation,(7–9) through imperfect pairing with target messenger RNA (mRNA) of protein-coding genes and the transcriptional or post-transcriptional regulation of their expression.(10–12) MicroRNA are aberrantly expressed or mutated in cancers, suggesting that they may be a novel class of oncogenes or tumor suppressor genes, depending on the targets that they regulate.(13) We previously screened 156 microRNA using 14 clinical UC specimens, five normal bladder epitheliums (NBE) and three UC cell lines and identified a profile of 27 microRNA including 19 downregulated and eight upregulated ones, which were differentially expressed in UC compared with NBE.(14) On the basis of the profile, we studied the functional role of the downregulated microRNA, that is, miR-145, miR-133a and miR-1, and their target genes.(15,16) Recent studies demonstrated that several cancer-specific microRNA are detectable in the serum from patients with breast, gastric colorectal and ovarian cancers.(17–21) Hankeet al. demonstrated that miR-126 and miR-182 might be used as tumor markers to distinguish bladder cancer (BC) from healthy control (HC) and/or urinary tract infection (UTI) patients.(22) These studies demonstrated that circulating microRNA might be detectable in serum and urine samples even if no cytosolic fraction is involved in a sample. Of the eight upregulated microRNA in our profiles of UC, miR-96, miR-183 and miR-190 were the top three upregulated microRNA.(15) We hypothesized that these microRNA are detectable in the urine from UC patients and are useful tumor markers for diagnosing UC. To test this hypothesis, we measured the expression levels of these microRNA in more than 100 clinical UC specimens by using stem-loop RT-PCR. Following that, we tested the microRNA expression in urine samples from another series of 100 UC, 49 HC and 25 UTI, and evaluated the correlations of their expression and clinicopathological features.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References
  9. Supporting Information

Patients and clinical samples.  To evaluate microRNA expression in clinical samples, we used tissue specimens from 104 UC patients who had undergone cystectomy or TUR-BT at Kagoshima University Hospital and three affiliated hospitals between 2003 and 2007. The NBE were derived from patients with prostate cancer and were used as the controls. Table S1 summarizes the patients’ backgrounds and clinicopathological characteristics.

Urine samples were collected from another series of UC patients (BC, renal pelvic and ureter UC) who had undergone cystectomy, TUR-BT or nephrouretectomy at Kagoshima University Hospital and three affiliated hospitals between 2008 and 2010 (Table 1). The present study was approved by the Bioethics Committee of Kagoshima University; written prior informed consent and approval were given by these patients. These samples were staged according to the American Joint Committee on Cancer-Union Internationale Contre le Cancer tumor-node-metastasis classification and histologically graded.(23) Our cohort did not include patients with carcinoma in situ or a metastasis region. Urine samples from 49 HC and 25 UTI were also available for study. We also analyzed 34 HC (30 male and four female) from older patients with a median age of 74.5 (range, 52–87) for evaluating age-related microRNA expression.

Table 1.   Patient’s characteristics
  1. TUR-BT, trans-urethral resection of bladder tumor.

Urothelial carcinoma
Total number100
Median age (range) (years)75 (47–94)
Gender
 Male67
 Female33
Stage (pTa)
 Non-invasive (pTa)27
 Invasive (≥pT1)73
Grade
 G19
 G235
 G356
Surgery
 TUR-BT66
 Cystectomy19
 Nephrouretectomy15
106
Follow-up period (range) (days)347 (2–709)
Healthy control
Total number49
Median age (range) (years)36 (28–53)
Gender
 Male45
 Female4
Urinary tract infection
Total number25
Median age (range) (years)64 (20–76)
Gender
 Male6
 Female19

Sample collection and RNA extraction.  A total of 135 tissue specimens (104 UC and 31 NBE) were immersed in RNA later (QIAGEN, Valencia, CA, USA) and stored at −20°C until RNA extraction. Total RNA was extracted from frozen fresh tissues using an ISOGEN kit (Nippon Gene, Tokyo, Japan) in accordance with the manufacturer’s protocol. The concentrations of RNA were determined spectrophotometrically; integrity was checked by gel electrophoresis. The RNA quality was confirmed in an Agilent 2100 bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). On the other hand, a total of 30 mL of voided urine samples was collected just before surgery and mixed with 15 mL of 30% polyethylene glycol 8000 (MP Biomedicals, Solon, OH, USA) in 3 M NaCl. The mixture was then centrifuged at 2000g for 15 min. After decanting the supernatant, a precise volume of 1 mL urine with the pelleted urine sediment was stored at −80°C until RNA extraction. Total urine RNA was isolated with mirVana PARIS kit (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s protocol. Each urine sample was frozen within 4 h of sampling, and we confirmed that the concentration and quality of a class of small RNA including microRNA was maintained 24 h after sampling.

MicroRNA detection by stem-loop PCR.  Stem-loop RT-PCR (TaqMan MicroRNA Assays; Applied Biosystems) was used to quantitate microRNA according to previously published conditions.(14) Because the concentration of total RNA in several urine samples (especially in HC) was too low to achieve an equal concentration per sample, a fixed volume of 5 μL total RNA was used for cDNA preparation, as previously described.(17) To prepare cDNA specific to the microRNA, each 15 μL RT reaction contained 5 μL purified total RNA, 50 nM stem-loop RT primer, 1× RT buffer, 0.25 mM each of dNTP, 3.33 U/μL MultiScribe reverse transcriptase and 0.25 U/μL RNase inhibitor. The reactions were incubated in a 96-well plate for 30 min at 16°C, 30 min at 42°C and 5 min at 85°C, and then held at 4°C. Each stem-loop RT-PCR for each microRNA assay was carried out in duplicate, and each 20 μL reaction mixture included 1.33 μL of diluted RT product, 10 μL of 2× TaqMan Universal PCR Master Mix and 1 μL of 20× TaqMan MicroRNA Assay Mix. The reaction was incubated in a 7300HT Fast Real-Time PCR System in 96-well plates at 95°C for 10 min, followed by 40 cycles at 95°C for 15 s and 60°C for 1 min.

Data normalization of tissue specimens and quantification of microRNA in urine samples.  RNA U6B small nuclear (RNU6B) (42 bp, a small RNA) served as the endogenous control for the clinical BC and NBE specimens. We used the manufacturer’s primary data across several human tissues and cell lines (P/N: 4373381; Applied Biosystems). Expression levels of microRNA were evaluated using the comparative CT method. The expression level of each tissue specimen was the relative value to the average expression of NBE. The microRNA in the urine samples can be detected in the urine fluid as well as in the urine sediment, which includes the cytosolic fraction. Because the amount of urine sediment differs between samples, there seems to be no appropriate endogenous control that will equally normalize both the urine fluid and urine sediment fractions. Therefore, we measured the absolute amount of microRNA, as previously described,(17,24,25) and performed the real-time RT-PCR assay with known amounts of synthetic human miR-96 and miR-183 (Hokkaido System Science Co., Ltd, Sapporo, Japan). In the presence of1 × 104 copies of synthetic miR-96 (CT = 39.3) and synthetic miR-183 (CT = 37.4) to 1 × 1010 copies (CT = 15.4 and 14.3, respectively), we observed a linear correlation (r2 > 0.99) between the logarithm of the amount of input RNA and the CT value (Fig. 1). No signal was detected even after 40 cycles of real-time PCR from either miR-96 or miR-183. We set 40 cycles as the baseline because the limit CT value for reliably detecting synthetic miR-96 and miR-183 was close to 40 but not beyond it. The relative concentration of microRNA in each sample was expressed as 2^(40-CT).

image

Figure 1.  Standard curves for stem-loop PCR assays used in the present study. Standard curves were generated for each microRNA assay using a dilution series of known input amounts of synthetic microRNA oligonucleotide in duplicate. The dilution series samples were run using common RT and PCR enzyme master mixes and on the same plate as the experimental samples. No signal was detected from either miR-96 or miR-183 even after 40 cycles of real-time PCR.

Download figure to PowerPoint

Oligonucleotide microarray analysis of the permanent miR-96- and miR-183-transfected UC cell lines.  Mir-96 and miR-183 vectors were constructed by inserting cloning sequences including the full-length of the mature microRNA sequences into the pBApo-CMV NeoTM vector (Takara Bio, Otsu, Japan). The entire sequence used for each microRNA transfection is described in Table S2. The microRNA and control vectors (vehicle) were transfected into KK47 and T24 cells by calcium phosphate co-precipitation. The transfectants were split and grown in selective medium with 1000 mg/L of G418. G418-resistant colonies were chosen and expanded in medium containing 200 mg/L of G418. DNA sequences of all constructs were confirmed by DNA sequencing (BIO MATRIX RESEARCH, INC, Tokyo, Japan), and microRNA overexpression was confirmed by the stem-loop PCR. Oligo-microarray Human 44K (Agilent) was used for expression profiling in miR-96- and miR-183-transfected BC cell lines (KK47 and T24) in comparison with miR-negative control transfectant, as previously described.(15) Briefly, hybridization and washing steps were performed in accordance with the manufacturer’s instructions. The arrays were scanned using a Packard GSI Lumonics ScanArray 4000 (Perkin Elmer, Boston, MA, USA). The data obtained were analyzed with DNASIS array software (Hitachi Software Engineering, Tokyo, Japan), which converted the signal intensity for each spot into text format. The Log2 ratios of the median subtracted background intensity were analyzed. Data from each microarray study were normalized by global normalization.

Statistical analysis.  The relationship between two variables and the numerical values obtained by real-time RT-PCR was analyzed using the Mann–Whitney U test. The relationship between three variables and the numerical values was analyzed using the Bonferroni-adjusted Mann–Whitney U test or the Kruskal Wallis test. The analysis software was Expert StatView (version 4; SAS Institute Inc., Cary, NC, USA); for the comparison test among the three variables, a nonadjusted statistical level of significance of < 0.05 corresponds to a Bonferroni-adjusted level of < 0.0167. The cut-off scores for the analysis were obtained by analyzing the receiver-operating characteristic (ROC) curve using MedCalc software (Med-Calc Software, Mariakerke, Belgium). The molecular function of the downregulated genes in the microRNA transfectants was classified into eight groups as referenced in the Gene Ontology (http://www.geneontology.org/index.shtml): metabolic process, response to stimulus, apoptosis, cell differentiation, regulation of transcription, anatomical structure development, regulation of signal transduction, and others.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References
  9. Supporting Information

Verification of upregulated microRNA expression in clinical specimens.  Because miR-96, miR-183 and miR-190 were the top three upregulated microRNA in our previous microRNA screening test of UC,(14) we subjected these microRNA to stem-loop RT-PCR using 104 clinical UC and 31 NBE. The expression levels of miR-96 and miR-183 in UC were significantly higher than in NBE (miR-96, 3.271 ± 0.321 vs 0.838 ± 0.153, < 0.0001; and miR-183, 4.225 ± 0.414 vs 1.224 ± 0.224, < 0.0001), while there was no significant difference in miR-190 expression between clinical UC and NBE (1.597 ± 0.310 vs 1.081 ± 0.219, = 0.512) (Fig. 2).

image

Figure 2.  Stem-loop PCR analysis of miR-96, miR-183 and miR-190 expression in tissue specimens from 104 urothelial carcinomas (UC) and 31 normal bladder epitheliums (NBE). The relative microRNA expression was normalized to the amount of RNU6B. Statistical significance was determined by the Mann–Whitey U test.

Download figure to PowerPoint

MiR-96 and miR-183 detection in urine samples from UC, HC and UTI patients.  Because the difference in miR-190 expression between clinical UC and NEB was not significant, we focused on detecting miR-96 and miR-183 in the urine samples (100 UC, 49 HC and 25 UTI). The stem-loop PCR revealed that the expression levels of miR-96 and miR-183 were significantly higher in UC than in HC (miR-96, 3895 ± 1018 vs 127 ± 20, = 0.0059; and miR-183, 25 541 ± 6617 vs 172 ± 36, = 0.0044), whereas there were trends to significance in the microRNA expression between UC and UTI (miR-96, 3895 ± 1018 vs 200 ± 57, = 0.0344; and miR-183, 25 541 ± 6617 vs 1195 ± 319, = 0.0320) (Fig. 3). There were no significant differences in the microRNA expression between HC and UTI (miR-96, = 0.9695; and miR-183, = 0.9341). The age distribution of our HC was considerably younger than that in the UC patients (Table 1). Therefore, we performed an additional experiment for evaluating the microRNA expression levels in HC of older patients (median age, 74.5 years), and there were no significant differences in the microRNA expression between young and older HC (Fig. S1).

image

Figure 3.  Stem-loop PCR analysis of miR-96 and miR-183 expression in urine samples from 100 urothelial carcinoma (UC) patients, 49 healthy controls (HC) and 25 patients with urinary tract infections (UTI). The relative microRNA concentration of each sample was expressed as 2^(40-CT), based on the standard curves for the absolute amount of each microRNA. Statistical significance was determined by the Bonferroni-adjusted Mann–Whitney U-test.

Download figure to PowerPoint

Evaluation of microRNA as a diagnostic marker: ROC curve analysis for distinguishing UC patients from non-UC patients.  We compared the microRNA expression level in UC (= 100) with that in non-UC (n = 74) that is the sum of HC and UTI and found that the miR-96 and miR-183 expressions were significantly higher in UC than in non-UC (miR-96, 3895 ± 1018 vs 152 ± 24, < 0.0001; and miR-183, 25 541 ± 6617 vs 518 ± 123, < 0.0001) (Fig. 4a). The ROC curve analyses demonstrated that each microRNA had good sensitivity and specificity with optimal cut-off values for distinguishing UC patients from non-UC patients as follows: miR-96, 71.0% (sensitivity), 89.2% (specificity), 284 (cut-off); and miR-183, 74.0% (sensitivity), 77.3% (specificity), 466 (cut-off) (Fig. 4b). The miR-96 detection yielded seven false positive cases out of 74 (9.6%), including three HC and four UTI, whereas the miR-183 detection yielded 17 false positive cases out of 74 (23.3%), including four HC and 13 UTI.

image

Figure 4.  (a) Stem-loop PCR analysis of miR-96 and miR-183 expression in urine samples from 100 urothelial carcinoma (UC) patients, 74 non-UC patients (49 healthy controls [HC] + 25 urinary tract infections [UTI]). Statistical significance was determined by the Mann–Whitey U test. (b) Diagnostic performance of miR-96 and miR-183 detection. The receiver-operating characteristic curves for miR-96 and miR-183 detected in the urine samples from 100 UC patients in comparison with 74 non-UC patients are shown.

Download figure to PowerPoint

Diagnosing UC in combination with urinary cytology.  In our cohort, 78 UC patients had undergone urinary cytology. Among these patients, 34 of 78 (43.6%) were given a “positive” urinary cytology diagnosis (Fig. 5, top). On the other hand, miR-96 was positively detected in 54 of 78 (69.2%) (Fig. 5, middle), and it was positively detected in 27 of 44 patients who had been given a “negative” diagnosis by urinary cytology (Fig. 5, middle). However, miR-96 was not detected in seven of 78 (9.0%) cases, although these seven patients had been given a “positive” diagnosis by urinary cytology (Fig. 5, middle). When we combined the miR-96 detection data with urinary cytology data, 61 of 78 cases were given a diagnosis of UC; the sensitivity for diagnosing UC rose from 43.6% to 78.2% (Fig. 5, bottom). Our cohort included 32 cases with low-grade cancer (G1 and G2). Among these, there were only seven cases given a positive urinary cytology, whereas miR-96 was positively detected in 16 of 32 (50.0%). Moreover, a remarkable finding was that our cohort included 18 patients with non-invasive and low-grade UC, and miR-96 was positively detected in nine of 18 (50%), whereas urine cytology was detected in only two (11%) (Table S3).

image

Figure 5.  Detection sensitivities of urinary cytology and miR-96 detection. Top: among the patients, 34 of 78 (43.6%) were given a “positive” diagnosis by urinary cytology. Middle: MiR-96 was positively detected in 54 of 78 (69.2%). MiR-96 was also positively detected in 27 of 44 patients who had been given a “negative” diagnosis by urinary cytology. Bottom: when we combined the miR-96 detection data with the urinary cytology data, 61 of 78 (78.2%) were given a diagnosis of urothelial carcinoma.

Download figure to PowerPoint

Correlation of clinicopathological features with miR-96 and miR-183 expression levels.  We found significant stepwise increases in miR-96 and miR-183 expression with advancing tumor grade (miR-96, 794 ± 480 [G1], 3577 ± 1743 [G2], 4555 ± 1457 [G3], = 0.0057; and miR-183, 2257 ± 1167 [G1], 17 789 ± 8187 [G2], 33 896 ± 10 555 [G3], = 0.0036) (Fig. 6a). The expression levels of the microRNA were significantly lower in non-invasive tumors (pTa) than in invasive tumors (≥pT1) (miR-96, 3476 ± 2166 vs 4067 ± 1167, = 0.0332; and miR-183, 12 816 ± 7057 vs 30 456 ± 8760, = 0.0117) (Fig. 6b). Among the 34 patients who underwent radical surgery, 17 (11 cystectomies and six nephrouretectomies) were available for evaluating microRNA expression; urine samples were collected at both pre- and post-surgery. The expression levels of the microRNA significantly decreased in the post-surgery urine compared with the pre-surgery urine (miR-96, 695 ± 304 vs 1411 ± 474, = 0.0241; and miR-183, 1522 ± 754 vs 7462 ± 3162, = 0.0045) (Fig. 6c). There was no significant correlation between UC recurrence and the microRNA expressions in the post-surgery urine. We found no difference between the microRNA expressions and the age/gender/tumor size of patients.

image

Figure 6.  Correlation of miR-96 and miR-183 expression levels with clinicopathological features. (a) The expression levels of miR-96 and miR-183 were classified into (a) tumor grade (G1, G2 and G3) and (b) pathological stage (pTa and ≥pT1). (c) They were also evaluated before and after the operation.

Download figure to PowerPoint

Gene expression analysis of miR-96 and miR-183 transfectants.  To gain further insight into which genes are downregulated by miR-96 and miR-183 in UC, we performed gene expression analysis of miR-96 and miR-183 transfectants (Fig. 7). We identified 587 genes that were generally downregulated more than 2.0-fold in miR-96 transfectants compared with the control transfectants; and 183 genes were down-regulated in miR-183 transfectants (Fig. 7a). The downregulated genes in miR-96 and miR-183 transfectants were involved in metabolic process (19.4% and 15.4%), response to stimulus (8.9% and 7.5%), apoptosis (6.8% and 9.3%), cell differentiation (5.8% and 6.4%), regulation of transcription (5.3% and 16.4%), anatomical structure development (3.9% and 6.4%), regulation of signal transduction (3.7% and 2.5%), and other functions (46.2% and 36.1%, respectively). We found that nine genes involved in activating apoptosis were commonly downregulated in both miR-96 and miR-183 transfectants (Fig. 7b). Entries from the current microarray data were approved by the Gene Expression Omnibus (GEO) and were assigned GEO accession number, GSE19717.

image

Figure 7.  Distribution of altered expressions of functionally categorized genes in miR-96 and miR-133a transfectants compared with the control transfectant. The functional features of the 587 and 183 genes, which were more than 2.0-fold downregulated in miR-96 and miR-133a transfectants, were classified into eight categories.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References
  9. Supporting Information

Several investigators have reported upregulated microRNA in UC in comparison with NBE by using a microRNA array (Table 2).(14,22,26–28) They used different microRNA array platforms to screen for specific microRNA, and there were no common upregulated microRNA except for miR-129 among the studies. They did not validate actual expression levels of the microRNA by using real-time PCR-based testing in a large number of clinical samples. Our microRNA profile, which contains 27 microRNA differentially expressed in UC, was established by screening microRNA using clinical UC specimens.(14) We previously found that the downregulated microRNA in our profile were indeed downregulated in more than 100 UC specimens and found their tumor suppressive function through regulating certain oncogenic genes.(14–16) In the present study, we conducted upregulated microRNA in our profile and validated their expression levels in both tissue and urine from UC patients. This strategy brought with it good results whereby miR-96 and miR-183 expressions in urine were well correlated with tumor stage and grade; these microRNA are thus promising diagnostic tumor markers to distinguish UC patients from non-UC patients. In addition, the expression of these microRNA significantly decreased after radical surgery, suggesting that they can be used as a prognostic marker of recurrence. Unfortunately, we could find no significant difference in prognosis between patients because our follow-up period was too short to make such an evaluation. MiR-96 and miR-183 have been found to be upregulated in various human cancers such as breast, lung, colon, liver, ovary, prostate, testis cancer and lymphoma.(29–35) Interestingly, miR-96 and miR-183 are in the same cluster; they are separated by just a short distance (212 bp) on chromosome 7q32. Therefore, it is plausible that their expressions are synchronized for targeting the same genes. In the present study, 14 cases did not show a synchronized expression pattern of miR-96 and miR-183 although they are closely located (Table S3). Post-transcriptional modifications by argonaute protein might account for these phenomena.(36) A previous study demonstrated that miR-96 and miR-183 were overexpressed in breast and endometrial cancers and simultaneously regulated the Forkhead Box O subfamily of transcription factors (FOXO), which is a tumor suppressor gene promoting G1 cell arrest and cell death.(30,35) Another study demonstrated that miR-183 directly regulated programmed cell death 4 (PDCD4), which is a proapoptotic molecule involved in TGF-β1-induced apoptosis in human hepatocellular carcinoma (HCC) cells.(34) In the present study, the gene expression profile of miR-96 and miR-183 transfectants demonstrated that the downregulated gene categories seem to include tumor suppressive categories, for example, anatomical structure development, cell differentiation and apoptosis. BCL2-associated X protein (BAX) was the top of the common downregulated genes in both miR-96 and miR-183 transfectants. These results implied that miR-96 and miR-183 are onco-microRNA and might be a potential target of gene therapy of some human malignancies. However, FOXO and PDCD4 were not strongly repressed by these microRNA in UC. In contrast, a recent paper demonstrated that miR-96 was downregulated in pancreatic cancers and plays as a tumor suppressor via KRAS gene regulation.(37) Thus, functional roles of these microRNA and their target genes may be different among malignancies. Further study is necessary to elucidate the precise functional role of these microRNA in UC.

Table 2.   Previously reported upregulated microRNA in UC compared with NBE
#microRNASample numberReference
  1. HC, healthy control; NBE, normal bladder epithelium; UC, urothelial carcinoma; UTI, urinary tract infection.

1miR-965/14 (NBE/UC)Ichimi et al.(14)
2miR-183
3miR-190
4miR-130b
5miR-124b
6miR-215
7miR-224
8miR-106a
9miR-1269/9/9 (urine: HC/UTI/UC)Hanke et al.(22)
10miR-182
11miR-17-5p2/25 (NBE/UC)Gottardo et al.(26)
12miR-23a-b
13miR-26b
14miR-103-1
15miR-185
16miR-203
17miR-205
18miR-221
19miR-223
20miR-519a11/106 (NBE/UC)Dyrskjøt et al.(27)
21miR-193a-3p
22miR-21
23miR-20a
24miR-198
25miR-510
26miR-184
27miR-492
28miR-129
29miR-1297/7 (matched NBE/UC)Wang et al.(28)
30miR-141
31miR-494
32miR-498
33miR-500
34miR-513

In several studies, small RNA, RNU6B or RNA U6 were used as endogenous controls to normalize the expression data of microRNA in tissue specimens. However, we think that these endogenous controls are not suitable for normalization of urine sample data because the expression levels of RNU6B or U6 would be influenced by the amount of urine sediment. Because microRNA can be detected in urine fluid as well as in urine sediment, and the amount of urine sediment differs between samples, there seems to be no appropriate endogenous control that will equally normalize both the urine fluid and urine sediment fractions. Therefore, the absolute calibration method, which is often used for microRNA detection in serum samples,(17,24,25) would be a better choice for measuring the expression levels of microRNA in whole urine. Hankeet al. suggested that miR-126 and miR-182 could be tumor markers for distinguishing UC from HC and/or UTI with a sensitivity of 72% and specificity of 82% using 29 urine samples from BC patients, but these microRNA were not correlated with clinicopathological features.(22) However, when they conducted microRNA screening they used RNU6B and RNA U6 as an endogenous control for normalizing the urine sample. This might be the reason why there are no common upregulated microRNA between their profile and our microRNA profile in UC. On the basis of the absolute calibration method, we used a simple formula, 2^(40-CT), to evaluate the expression level of the microRNA in urine, as previously described.(24) Our simple method might contribute towards widespread use of microRNA detection as tumor markers. Moreover, microRNA in urine is readily detectable by stem-loop PCR: it takes approximately 3.5 h from harvesting a urine sample to reaching a diagnosis as UC or non-UC, and it costs approximately $10 (US) per sample.

We found more false positive cases in miR-183 detection compared with miR-96 detection. Because the majority of false positive cases were UTI patients, miR-183 might be upregulated and function in UTI as well as UC. We therefore assumed that miR-183 might be useful as a staging marker but not as a diagnostic marker. MiR-96 seems to be a good tumor marker for distinguishing UC patients from non-UC patients with high sensitivity and specificity (approximately 70% and 90% in the present study). These data might be superior to the current urine markers, that is, BTA, NMP22 and FDP. In the current study, neither miR-96 nor miR-183 was detected in nine invasive UC including four low-grade and five high-grade UC (Table S3), and these data suggest that some invasive phenotypes might not require miR-96 or miR-183 expression for their development. A major drawback of current urinary cytology is its low sensitivity, and we believe the diagnostic accuracy for UC may be improved by a concurrent examination involving miR-96 detection and urinary cytology (approximately 80% accuracy in the present study). Furthermore, miR-96 detection in urine is a non-invasive tumor marker, and it might be a good replacement for the current painful urethro-cystoscopy during the initial diagnosis and follow-up examination after TUR-BT. A prospective study of using miR-96 for UC diagnosis is now underway in our affiliate hospitals.

We concluded that miR-96 and miR-183 in urine are promising tumor markers for UC. In particular, miR-96 might be a useful diagnostic marker in combination with urinary cytology.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References
  9. Supporting Information

This research was partially supported by the Ministry of Education, Science, Sports and Culture, Grants-in-Aid for Scientific Research, 20390427 and 20591861, 2008. The authors thank Ms Mutsumi Miyazaki for her excellent laboratory assistance.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References
  9. Supporting Information

Fig. S1. Stem-loop PCR analysis of miR-96 and miR-183 expression in urine samples from 100 urothelial carcinoma (UC) patients, 49 healthy controls (HC) (median age, 36 years) and 34 older healthy controls (HC) (median age, 74.5 years).

Table S1. Patient’s characteristics of clinical specimens.

Table S2. Entire sequence used for each microRNA transfection.

Table S3. MicroRNA detection in a patient who had a urine cytology test.

FilenameFormatSizeDescription
CAS_1816_sm_FigureS1.TIF563KSupporting info item
CAS_1816_sm_TableS1.doc61KSupporting info item
CAS_1816_sm_TableS2.doc29KSupporting info item
CAS_1816_sm_TableS3.doc144KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.