Analysis of the Na,K-ATPase α- and β-subunit expression profiles of bladder cancer using tissue microarrays

Authors

  • Cromwell Espineda B.S.,

    1. Department of Pathology and Laboratory Medicine, University of California-Los Angeles, Los Angeles, California
    Search for more papers by this author
    • The first two authors contributed equally to this work.

    • Cromwell Espineda, Yunda Huang, and Tao Shi are graduate students.

  • David B. Seligson M.D.,

    1. Department of Pathology and Laboratory Medicine, University of California-Los Angeles, Los Angeles, California
    2. Jonsson Comprehensive Cancer Center, University of California—Los Angeles, Los Angeles, California
    Search for more papers by this author
    • The first two authors contributed equally to this work.

  • William James Ball Jr. Ph.D.,

    1. Department of Pharmacology and Cell Biophysics, University of Cincinnati Medical Center, Cincinnati, Ohio
    Search for more papers by this author
  • JianYu Rao M.D.,

    1. Department of Pathology and Laboratory Medicine, University of California-Los Angeles, Los Angeles, California
    2. Jonsson Comprehensive Cancer Center, University of California—Los Angeles, Los Angeles, California
    Search for more papers by this author
  • Aarno Palotie M.D., Ph.D.,

    1. Department of Pathology and Laboratory Medicine, University of California-Los Angeles, Los Angeles, California
    2. Department of Human Genetics, University of California-Los Angeles, Los Angeles, California
    Search for more papers by this author
  • Steve Horvath Ph.D.,

    1. Department of Human Genetics, University of California-Los Angeles, Los Angeles, California
    2. Department of Biostatistics, University of California-Los Angeles, Los Angeles, California
    Search for more papers by this author
  • Yunda Huang B.S.,

    1. Department of Biostatistics, University of California-Los Angeles, Los Angeles, California
    Search for more papers by this author
    • Cromwell Espineda, Yunda Huang, and Tao Shi are graduate students.

  • Tao Shi,

    1. Department of Human Genetics, University of California-Los Angeles, Los Angeles, California
    2. Department of Biostatistics, University of California-Los Angeles, Los Angeles, California
    Search for more papers by this author
    • Cromwell Espineda, Yunda Huang, and Tao Shi are graduate students.

  • Ayyappan K. Rajasekaran Ph.D.

    Corresponding author
    1. Department of Pathology and Laboratory Medicine, University of California-Los Angeles, Los Angeles, California
    2. Jonsson Comprehensive Cancer Center, University of California—Los Angeles, Los Angeles, California
    • Department of Pathology and Laboratory Medicine, Room 13-344 CHS, University of California-Los Angeles, Los Angeles, CA 90095
    Search for more papers by this author
    • Fax: (310) 267-2410


Abstract

BACKGROUND

The purpose of this study was to determine the clinical significance of Na,K-ATPase α- and β-subunit expression in a histopathologically well-characterized group of patients representing a wide spectrum of tumor grades and disease stages with transitional cell carcinomas (TCC).

METHODS

Na,K-ATPase α- and β-subunit protein expression patterns were analyzed using immunohistochemistry on urothelial cancer tissue microarrays (TMA) of 146 patients diagnosed with urothelial carcinoma. For each subunit, the maximum staining intensity and the percentage of positive cells staining at the maximal intensity were analyzed.

RESULTS

Compared with the benign fields, the mean protein expression for both Na,K-ATPase α- and β-subunits were found to be decreased overall in in situ and invasive tumors, as well as in tumor-adjacent dysplastic fields. When Na,K-ATPase α- and β-subunit expression levels were dichotomized into distinct groups, they were both found to be significant predictors of recurrence risk in multivariate logistic regression analysis (P = 0.0062, odds ratio [OR] = 2.6 and P = 0.013, OR = 0.43, for Na,K-ATPase α- and β-subunits, respectively). The authors also found that patients with high α- and low β-subunit expression had a high risk for early recurrence, whereas patients with a low α- and high β-subunit expression had a significantly longer median recurrence-free time (17 months and 125 months, respectively, log rank statistics P = 0.0005).

CONCLUSIONS

The results suggested that Na,K-ATPase α- and β-subunit expression levels may be useful predictors of clinical outcomes such as recurrence-free time of bladder cancer patients. Cancer 2003;97:1859–68. © 2003 American Cancer Society.

DOI 10.1002/cncr.11267

Bladder cancer is an important clinical problem. In the United States, an estimated 54,300 new cases of bladder cancer were diagnosed during 2001.1 Although 80% of cases are diagnosed early and can be effectively treated, approximately 12,400 patients die annually.2 Approximately 90% of malignant tumors arising in the bladder are of epithelial origin, the majority being transitional cell carcinomas (TCC).3 TCC of the bladder is the second most frequent disease of the genitourinary tract and the second most prevalent cause of death of all genitourinary tumors.4 Low-grade superficial lesions often recur yet seldom invade. High-grade lesions are more often present as invasive tumors, with significantly more life-threatening outcomes.5 More than 70% of treated tumors recur, and 30% of recurrent tumors advance to invasive tumors.6 Improved prognostication and surveillance are critical to the management of these patients.

The Na,K-ATPase catalyzes an ATP-dependent transport of three sodium ions out and two potassium ions into the cell per pump cycle, thereby generating a transmembrane sodium gradient. The activity of the Na,K-ATPase is involved in the control of cellular pH, osmotic balance, and the Na+-coupled transport of nutrients such as amino acids and vitamins into cells. The Na,K-ATPase consists of two noncovalently linked α- and β-subunits. The α1-subunit (approximately 112 kilodaltons [kD]) contains the catalytic and ligand binding sites of the enzyme.7 Four α-isoforms have been described in mammals;8–10 the α1 isoform is expressed in most of the tissue types.11, 12 The β1-subunit (approximately 55 kD)13 is a glycosylated protein, and its role in Na,K-ATPase enzyme function remains somewhat obscure. It may modulate the transport of Na+ and K+ across the membrane14 and facilitate the insertion of the αβ-complex into the cell membrane.15–17 Of the three isoforms described,18 the β1 isoform is expressed in the most tissues.11, 12

We have shown earlier that in renal clear-cell carcinoma, the protein levels of the Na,K-ATPase β-subunit but not that of the α-subunit are highly reduced, suggesting that reduced expression of β-subunit might be associated with the invasive phenotype of renal clear-cell carcinoma.19 Subsequently, we found that oncogenic transformation of Madin-Darby Canine Kidney (MDCK) cells with Moloney sarcoma virus (MSV-MDCK) resulted in highly reduced Na,K-ATPase β-subunit levels.20 Earlier studies correlated the invasive phenotype of MSV-MDCK cells to reduced expression of E-cadherin.21 We demonstrated that in E-cadherin-expressing MSV-MDCK cells, ectopic expression of Na,K-ATPase β-subunit was necessary to suppress invasiveness of these cells.20 These studies, for the first time to our knowledge, demonstrated that Na,K-ATPase β-subunit function might play an important role in the suppression of invasiveness of kidney carcinoma cells.

Tissue microarrays (TMAs) provide a convenient high-throughput tissue-based tool for in situ gene dosage and protein expression studies. Introduced by Kononen et al.,22 the TMA technique has been utilized for rapid profiling of molecular tumor markers and compares well to results obtained by standard methods.23–25 TMAs are being widely used to obtain valuable information regarding the expression pattern of molecular markers in cancers and particularly in bladder carcinomas.26

In the present study, we performed immunohistochemical analyses on urothelial TMAs to test the clinical significance of Na,K-ATPase α- and β-subunit expression in a histopathologically well-characterized group of patients with TCC representing a wide spectrum of tumor grades (predominantly Grade 2 and above) and disease stages (I–IV). The Na,K-ATPase subunit expression patterns were correlated with clinicopathologic parameters and patient outcome to determine their potential prognostic value.

MATERIALS AND METHODS

Urothelial Cancer TMA Construction

Formalin-fixed, paraffin-embedded tissue samples derived from urothelial cancers from 1985–1995 were randomly chosen from the archives of the Department of Pathology at the University of California-Los Angeles (UCLA) Medical Center and utilized under IRB approval. The original Hematoxylin and Eosin (H & E)-stained slides were reviewed by a UCLA pathologist (J.R.), uniformly utilizing the 1997 TNM classification.27 TMA blocks were constructed following the technique described by Kononen et al.22 using 0.6-mm-diameter tissue cores arrayed into standard-sized histologic paraffin blocks to a density of approximately 454 cores per block. Where available, at least three tumor and one matching morphologically normal-appearing transitional epithelium sample were targeted for each case. Where available, separate three-core sets representing dysplasia, carcinoma in situ (CIS), distinct tumor grades, or metastases also were sampled from the cases. These same arrays have been used to evaluate other biomarkers including p53, Ki67, E-cadherin, and Gelsolin, and detailed information about the construction of these arrays has been previously described.28 An example of one of the urothelial TMAs used in this study is shown in Figure 1.

Figure 1.

Representative tissue microarrays (TMA) consisting of 0.6-mm-diameter tissue spots of transitional cell carcinomas and normal urothelial tissues (hematoxylin and eosin). Higher-power views illustrate tissue spot arrangement and an individual TMA spot.

Patients and Histopathology

Three urothelial arrays encompassing a total of 1363 tissue spots from 232 individual tumors from 167 patients with urothelial cancer were utilized. Among 167 patients, 123 patients had only 1 tumor, and 44 patients had multiple metachronous tumor samples (range, 2–6 per patient). H & E-stained array sections were histopathologically evaluated by an anatomic pathologist (D.B.S.) in a blinded fashion to validate the diagnostic morphology of each array spot. Cases comprising only metastatic tumors and tumors showing exclusive squamous cell carcinoma or adenocarcinoma differentiation were excluded from the analysis. Postexclusion material included 146 patients, 202 cases, and 1208 tissue spots. Of the 146 urothelial tumors, 140 tumors were from bladder (2 of which showed small cell differentiation, 1 showed signet ring features, and 1 accompanied concomitant renal pelvis urothelial carcinoma), 3 were from renal pelvis alone, and 3 were from ureter alone. The patient ages ranged from 33–94 years, (mean age, 67 years). The male to female ratio was 3.6:1. There were 57 Tis/Ta/T1 noninvasive or superficially (lamina propria) invasive tumors, and the remaining 89 were deeply invasive, including 44 T2, 35 T3, and 10 T4 tumors. There were 8 in situ, 6 Grade 1, 40 Grade 2, and 92 Grade 3 tumors.

Clinical and Pathology Database

Detailed retrospective demographic, pathology, and clinical history information, including treatment and follow-up data for at least 5 years, was incorporated into a correlative database linked to the tissue specimens in an anonymized fashion. Original data sources included surgical pathology reports from the UCLA Department of Pathology as well as tumor registry data obtained from the UCLA Cancer Program of the Jonsson Comprehensive Cancer Center.

Immunohistochemistry on TMA Sections

Mouse monoclonal antibodies (MoAbs) raised against Na,K-ATPase α- (M7-PB-E9) and β-subunit (M17-P5-F11) recognize epitopes that are common in human, sheep, and dog and have been characterized and described previously.19, 20, 29, 30 A standard two-step indirect avidin-biotin complex (ABC) method was used for immunohistochemical studies (Vector Laboratories, Burlingame, CA). Four-micron-thick tissue array sections were cut immediately prior to staining and were transferred using an adhesive slide system to maintain array integrity (Instrumedics, Hackensack, NJ) (Fig. 1A–C). They were first heated to 60 °C for 15 minutes, followed by deparaffinization in xylene. The sections were then rehydrated in graded alcohols and endogenous peroxidase quenched with 10% hydrogen peroxide in phosphate-buffered saline (PBS) at room temperature for 20 minutes. After washing, the sections were then placed in 95 °C solution of 0.01 M sodium citrate buffer for antigen retrieval. Protein blocking was accomplished through application of 1% normal horse serum, 5% bovine serum albumin (BSA) for 30 minutes. Primary mouse anti-Na,K-ATPase α- or β-subunit monoclonal immunoglobulin (Ig) G1 antibodies were applied at a 1:200 dilution for 60 minutes at 37 °C. After washing, biotinylated horse anti-mouse IgG was applied for 30 minutes at room temperature. The ABC complex was applied for 25 minutes, and diaminobenzidine (DAB) was used as the chromogen. The sections were counterstained with Harris hematoxylin, followed by dehydration and mounting. Complementary slides processed in the same manner minus primary antibody application served as negative controls. Whole-tissue sections of normal urothelial tissues served as positive tissue controls.

Histomorphologic Analysis and Scoring Criteria

The sections were analyzed with an Olympus BX-40 brightfield microscope (Olympus, Tokyo, Japan). Semiquantitative assessment of antibody staining was performed blinded to clinicopathologic variables. Scoring procedures were performed by a trained pathologist (D.B.S.). Metrics include both maximal membrane staining intensity (graded on a 0–4 scale: 0 = negative; 1 = weak staining; 2 = weak but distinct staining; 3 = moderate staining; 4 = strong staining) and the frequency of staining (proportion of the analyzed cells staining positively with the maximal intensity, 1–100%). Spots were considered informative if they were either 1) missing, 2) present but lacking target tissue, or 3) damaged, rendering them unreadable. If the patient had at least one evaluable tumor spot representing the patient's cancer grade, it was included in the analysis.

Statistical Analysis

The expression intensity and frequency distributions of the Na,K-ATPase α- and β-subunits were examined first by considering all informative tissue spots (662 spots for α and 678 spots for β) of 202 tumors from 146 patients. The Kruskal–Wallis test, which is a nonparametric version of one-way analysis of variance (ANOVA) for multigroup comparison, was used to compare the differences of Na,K-ATPase subunits expression between histologic categories. Association between the risk of tumor recurrence and Na,K-ATPase subunit expression was analyzed using multivariate logistic regression. Recurrence was defined as returning tumor growth postexcision, seen on reexcision material, or seen cystoscopically. Tumor spots in which spot grades did not match the overall case grades, CIS tumors, tumor spots that were not informative for both Na,K-ATPase subunits, tumors of patients with no disease-free interval, and those with incomplete follow-up data were excluded from the outcome analyses. The resulting data set encompassed primary tumors from 72 patients in our data set. The median age of diagnosis of this group was 67.4 years (range, 42–82 years) with a male to female ratio of 4.5:1. Twenty-nine patients had recurrences, and 43 had no recurrence. The median follow-up time (time to first recurrence in the recurring group or total follow-up time in the nonrecurring group) was 43.5 months (range, 2–152 months). This group contained 4 (5%) Grade 1, 20 (28%) Grade 2, and 48 (67%) Grade 3 patients. Pathology T-stage breakdown of this group included 13 Ta, 11 T1, 24 T2, 19 T3, and 5 T4. Pathology regional lymph node status included 26 NX, 41 NO, 4 N1, 1 N2, and 0 N3. Thirty-five percent of the patients were group Stage I, 29% were Stage II, 22% Stage III, and 14% Stage IV. No synchronous nor metachronous tumors were included in the outcome analyses. Na,K-ATPase expression intensities and frequencies from multiple spots representing the same tumor were pooled to form the mean value. Pearson correlation coefficients were used to measure correlations between the α- and β-subunit staining intensity and frequency. The staining intensity and frequency were highly correlated for Na,K-ATPase α but less so for β (Pearson correlation coefficient = 0.71, P < 0.0001 and 0.30, P = 0.01, respectively). Therefore, to prevent multicolinearity problems among the subunit predictors, either intensity or frequency alone was used in further analyses. For outcome analysis, optimal correlations using the frequency metric for α and intensity metric for β was used. Survival analysis involving recurrence-free time was conducted with recurrence time defined as the period (in months) from the date of first diagnosis to the first recurrence date or the censored date, which is the last date of negative clinical follow-up, or the date of cystectomy. Kaplan–Meier curves were used to estimate the recurrence-free time distribution. The log rank and Wilcoxon tests were used to test whether the recurrence-free time distributions differed. To assess which covariates affect recurrence-free time, we applied multivariate survival analysis using the Cox proportional hazards model. The proportional hazard assumption was tested using scaled Schoenfeld residuals. For each covariate, the relative hazard rate and the associated P value were examined. For all analyses, P < 0.05 was accepted as significant. Survival tree analysis was carried out with the rpart library in R. The analyses were carried out with the software packages R (url: http://cran.r-project.org/) and SAS (SAS Institute, Cary, NC).

RESULTS

Na,K-ATPase α- and β-Subunit Protein Expression in Normal and Cancer Bladder Tissues

Although Na,K-ATPase α- and β-subunit isoform expression pattern has been described for various tissue types,12 Na,K-ATPase isoforms expressed in human urinary bladder has not been reported. We found that in human urinary bladder transitional epithelium, Na,K-ATPase α1 and β1 isoforms are both expressed (Fig. 2). In most of the tissues, Na,K-ATPase α- and β-subunits were localized to the basolateral plasma membrane of polarized epithelial cells.31 In the tissue sections of morphologically normal urothelium, antibodies against Na,K-ATPase α- (Fig. 2A) and β-subunit (Fig. 2B) stained distinctly the basolateral plasma membrane of umbrella cells (see insets in Fig. 2A and B). In both basal cell and the intermediate transitional cell layers, Na,K-ATPase α- and β-subunit staining was uniformly distributed on the plasma membrane. The intensity of staining of the umbrella cells (arrowhead) and the basal cell layers (arrows) was greater compared with the intermediate transitional cell layers. In tumor tissue sections, Na,K-ATPase α- and β-subunits showed more varied staining intensities. Examples of low- and high-intensity staining of both subunits are shown in representative low- and high-grade transitional cell carcinomas (Fig. 2D,E and 2G,H, respectively). Both the α- and β-subunits distinctly stained the plasma membrane yet some cytoplasmic staining also was detected (insets in Fig. 2G and H).

Figure 2.

Staining of Na,K-ATPase α- (A,D,G) and β-subunits (B,E,H) and negative control (C,F,I) in matched morphologically normal (A–C) and transitional cell carcinoma (D–I) tissues. Note the intense staining of both α- and β-subunits in the umbrella cells (arrowheads) and the basal cells (arrows). Insets in (A) and (B) show distinct basolateral staining of Na,K-ATPase α- and β-subunit in umbrella cells. In (D) and (E), reduced staining intensity of α- and β-subunits in a Grade 1 tumor is shown. Insets in (D) and (E) show distinct plasma membrane staining of these subunits. In (G) and (H), increased staining of α- and β-subunits in a Grade 3 tumor is shown. Insets show both plasma membrane and cytoplasmic staining (arrowhead) of α- and β-subunits. Original magnification ×40; insets ×100.

Low-power views of the representative normal and tumor TMA spots showing the staining intensity of α- and β-subunits analyzed in this study are shown in Figure 3. In normal tissue TMA spots, localization of the α- and β-subunit is clearly seen to be transepithelial and strongly intense (Fig. 3A,B). In Grade 1 TMA spots (Fig. 3D–F), the intensity of the α- and β-subunit was generally less than that of the TMA spots representing normal tissues (Fig. 3, compare A and B with D and E). In these spots, β-subunit staining intensity was proportionally lower than that of α-subunit (Fig. 3, compare D and E). TMA spots representing Grade 3 showed an intensity approaching that in normal TMA spots (Fig. 3, compare G and H with A and B).

Figure 3.

Representative immunohistochemical staining of Na,K-ATPase α- (A,D,G) and β-subunits (B,E,H) and negative control (C,F,I) on tissue microarrays (TMA). (A)–(C), (D)–(F), and (G)–(I) represent matched morphologically normal, Grade 1 and Grade 3 TMA spots, respectively. Original magnification ×10.

In Figure 4, mean expression intensities with 95% confidence intervals (CI) upper limits of Na,K-ATPase α- and β-subunits from all TMA spots pooled by histologic category are shown. A reduced expression of the Na,K-ATPase α- and β-subunits is seen in dysplastic and in situ lesions, as well as in tumor and their metastases, as compared with morphologically normal matched tissues from these cancer patients. Both subunits showed reduced expression in the low-grade tumors (Grade 1) as compared with normal tissue protein expression values (P < 0.0001 by Kruskal–Wallis test for Na,K-ATPase α- and β-subunit mean expression levels). The trend of increasing expression is predominantly seen in the transition from Grade 1 to Grade 2, especially for the β-subunit, although expression always remains below normal levels. There was no association between expression levels of both subunits and tumor stage (data not shown).

Figure 4.

Mean intensity staining distributions of Na,K-ATPase α- (A) and β-subunits (B) in patients stratified by histologic categories. Numbers in parentheses indicate the number of tissue spots belonging to tumors of Grades 1, 2, and 3, carcinoma in situ (CIS), metastatic tumor (MET), morphologically normal matched urothelium (NL), and dysplastic urothelium (DYSP).

Evaluation of the Na,K-ATPase α- and β-Subunit Expression in the Prognosis of Bladder Carcinoma

The cohort of 72 patients, most of them (68) with Grade 2 and higher tumors as described above, was used to analyze whether Na,K-ATPase α- and β-subunit expression was associated with tumor recurrence. Because the distribution of the staining values is highly skewed, we decided to dichotomize them. We used survival tree analysis to determine appropriate cutoff values. We found the optimal cutoff value of 85% for Na,K-ATPase α- and an intensity of 3 for Na,K-ATPase β. These dichotomized Na,K-ATPase α- or β-subunit values showed significant segregation effects on the recurrence time in univariate analyses (P = 0.036, hazard ratio, 1.98, 95% CI, 1.05–3.76 for α- and P = 0.063, hazard ratio 0.55, 95% CI, 0.292–1.03 for β-subunit, respectively), indicating that α- and β-subunit expression in the high and low groups, respectively, is associated with a reduced time to recurrence.

In a multivariate Cox regression model utilizing the dichotomized Na,K-ATPase α- and β-subunit covariates as well as gender, age at diagnosis, and tumor grade and stage, the dichotomized Na,K-ATPase subunit expression levels were the only covariates that were significant (P = 0.045, hazard ratio 2.2, 95% CI, 1.02–4.65 for α and P = 0.033, hazard ratio 0.47, 95% CI, 0.24–0.94 for β). An interaction involving the α- and β-subunit expression was not significant in multivariate Cox regression, indicating that their association to recurrence time is not necessarily dependent on one another.

By using the same cutoff values, we were able to divide patients into groups with significantly different recurrence-free time, which we estimated with the Kaplan–Meier method. The median recurrence-free time of patients with low α-subunit expression was 82 months, whereas that of patients with high expression was 49 months (P = 0.031 by log rank test). In contrast, we found that high expression of Na,K-ATPase β-subunit was protective. The median recurrence-free time of patients with Na,K-ATPase β-intensity < 3 was 33 months, whereas that of patients with a high intensity was 78 months (the Wilcoxon test between the two groups yields P = 0.026 and the log rank test P = 0.061).

Next, we combined α- and β-subunit expression levels together to divide the patients into four groups. The first group contained patients with high α and low β (n = 8), the second group contained patients with low α and high β (n = 18), the third group contained patients with high α and high β (n = 12), and the fourth group contained patients with low α and low β (n = 34). Most distinctly, the median recurrence-free time was 17 months (range, 8–57 months) and > 126 months (range, 3–126 months) for patients belonging to the first group and second group, respectively, with a highly significant log rank test (P = 0.0005), indicating that these groups form prognostically distinct patient groups (Fig. 5). The median recurrence-free time for the fourth group and third group of patients was 42 months (range, 3–152 months) and 71 months (range, 2–87 months), respectively. When Na,K-ATPase expression Groups 1 and 2 were stratified into low (noninvasive/Group Stage 0a–I) and high (invasive/Group Stage II–IV) stage subgroups, a similar time to recurrence curve was obtained. The high-stage subgroup includes 22 patients with 17 belonging to the high β–low α and 5 patients belonging to the high α–low β groups. Low β–high α subgroup showed a recurrence-free time of 48 months (range, 8–50), whereas the high β–low α group showed a recurrence-free time of >126 months (range, 3–126 months), (log rank test P = 0.0053) (Fig. 6). Although this same pattern also held for the low-stage (0a–I) subgroup, the number of patients was too few to generate a meaningful statistical value. These data suggest that increased α-subunit expression is associated with decreased recurrence-free time, whereas increased β-subunit expression is associated with increased recurrence-free time. However, these results are based on a rather small number of patients and larger studies are necessary to further validate these findings.

Figure 5.

Kaplan–Meier curves illustrating the recurrence-free interval for patient subgroups (any stage 0a–IV; n = 36) with low Na,K-ATPase α- and high Na,K-ATPase β-subunit expression, or high Na,K-ATPase α- and low Na,K-ATPase β-subunit expression. Log rank analysis comparing these subgroups showed that the high Na,K-ATPase α- and low Na,K-ATPase β-subunit expression subgroup had a significantly higher recurrence rate (P = 0.0005).

Figure 6.

Kaplan–Meier curves illustrating the recurrence-free interval for patient subgroups (high-stage II–IV; n = 22) with low Na,K-ATPase α- and high Na,K-ATPase β-subunit expression, or high Na,K-ATPase α- and low Na,K-ATPase β-subunit expression. Log rank analysis comparing these subgroups showed that the high Na,K-ATPase α- and low Na,K-ATPase β-subunit expression subgroup had a significantly higher recurrence rate (P = 0.0053).

Furthermore, we found that high Na,K-ATPase α-subunit and low Na,K-ATPase β-subunit, measured either separately or together in the 72 patient group, led to an increased risk of recurrence using multivariate logistic regression. Both Na,K-ATPase α- and β-subunit values were significant (P = 0.013, odds ratio [OR] = 5.38, 95% CI, 1.420–20.368 for α and P = 0.044, OR = 0.33, 95% CI, 0.111–0.972 for β) when examined together as the only covariates. When other covariates such as gender, age, grade, and stage were added into the logistic regression model, the Na,K-ATPase α- and β-subunit expression levels lost significance (P = 0.06, OR = 3.98, 95% CI, 0.943–16.824 for α and P = 0.066, OR= 0.32, 95% CI, 0.092–1.077 for β).

DISCUSSION

Utilizing TMA technology, we have studied the protein expression patterns of Na,K-ATPase α- and β-subunits in urothelial cancer patients. Our study revealed a distinctive biphasic protein expression pattern of the Na,K-ATPase α- and β-subunits in urothelial cancer. Expression intensity was found to be highest for both subunits in morphologically normal tissues. Significant decreases in expression were seen in the limited number of adjacent dysplasia and Grade 1 papillary tumors, suggesting that reduced Na,K-ATPase subunit levels might have occurred at the early stage of tumor development. The lowest expression was seen in Grade 1 tumors, with tumors of higher grade displaying increasing expression. This biphasic pattern of Na,K-ATPase subunit expression in bladder cancer suggests that Na,K-ATPase subunit levels and possibly the Na,K-ATPase activity are modulated during bladder cancer progression. We have shown earlier that in renal clear-cell carcinoma, although near normal levels of α-subunit were present, reduced levels of β-subunit in tumor tissues correlated with significantly reduced Na,K-ATPase activity.19 Reduced Na,K-ATPase activity has been shown during the progression of colon cancer in animal models.32 It is possible that during bladder cancer progression, reduced β-subunit levels during the early stages of tumor development might lead to reduced Na,K-ATPase activity, which might result in events that favor the progression of bladder cancer. Recently we have shown that the intracellular sodium homeostasis, regulated by Na,K-ATPase, is crucial for the development of tight junctions and induction of polarity in epithelial cells.33 Tight junctions are crucial in maintaining the polarized phenotype of epithelial cells.34 Reduced Na,K-ATPase activity during the early stages of bladder cancer might lead to loss of tight junctions and polarity in urothelial cells. Consequently, the basolaterally localized proteins, such as epidermal growth factor receptor (EGFR), might be aberrantly expressed at the apical plasma membrane. Apical expression of EGFR should allow its association with EGF present in the urine35 and activation of EGF-mediated signaling pathways. Alternatively, the lumenal EGF might seep through the tight junctions and activate EGFR localized to the basolateral domain.36 Recent studies have shown that inhibition of Na,K-ATPase can activate EGFR in a ligand-independent fashion.37 Altered Na,K-ATPase α- and β-subunit levels in bladder cancer might lead to reduced Na,K-ATPase activity and activation of EGFR, which might further contribute to the progression of bladder cancer.

In renal clear-cell carcinoma, we found that Na,K-ATPase β-subunit expression is highly reduced in low-grade tumors.19 In this study, we found that Na,K-ATPase α- and β-subunit expression are the lowest in low-grade tumors. Both Na,K-ATPase α- and β-subunits are known to be regulated transcriptionally13, 38–40 as well as translationally41–43 in a wide variety of cell types. It is possible that factors induced during the early stage of tumor development may transcriptionally or translationally reduce the expression of Na,K-ATPase subunits. Increased expression of both subunits in higher-grade tumors suggests that the mechanisms that led to reduced Na,K-ATPase α- and β-subunit expression are likely inactivated either by newly induced factors or activation of existing factors in higher-grade tumors. Alternatively, reduced Na,K-ATPase activity in low-grade tumors might lead to an increase in the intracellular sodium concentration. It has been shown that increased intracellular sodium can increase the transcription of both α- and β-subunits,44–46 thus increasing the levels of these mRNAs and proteins in higher-grade cancers. Experiments are in progress in our laboratory to understand the mechanism(s) resulting in the biphasic nature of the Na,K-ATPase subunit expression in bladder carcinoma.

In patients with low α-subunit and high β-subunit expression levels, we found a significant decrease in recurrence risk and an increase in the recurrence-free time. Conversely, in patients with higher expression of α-subunit and low expression of β-subunit, the recurrence-free time was significantly reduced, suggesting that increased β-subunit expression has a protective effect against the recurrence of bladder cancer. In contrast, increased α-subunit expression appears to have an unfavorable influence resulting in the increased and earlier recurrence of this cancer. We have shown earlier that in E-cadherin-expressing MSV-MDCK cells, expression of the β-subunit induced the formation of junctional complexes such as tight junction and desmosomes and significantly reduced the motility and invasiveness of these cells.20 These studies revealed that β-subunit of Na,K-ATPase might be involved in the mechanisms leading to the suppression of invasiveness of carcinoma cells. It is possible that increased β-subunit expression might reduce invasiveness of tumor cells, thus reducing the chance of tumor spread and consequently its recurrence.

Both α- and β-subunits can be regulated by independent mechanisms.39, 45, 47, 48 Significantly different levels of α- and β-subunit in bladder carcinoma patients, as well as their statistical lack of interaction in multivariate analyses, suggest that in human urothelium, these subunits are probably differentially regulated. Identification of these mechanisms should facilitate novel therapeutic approaches to treat bladder carcinoma.

When patients were dichotomized into high and low Na,K-ATPase α- and Na,K-ATPase β-subunit expression subgroups, these markers were more significant predictors of recurrence risk and recurrence time than either the stage or grade of the tumor, indicating that these markers could provide a significant addition as clinically useful prognosticators. These findings indicate that Na,K-ATPase α- and Na,K-ATPase β-subunits may form a set of potentially useful tumor markers that could help guide therapeutic decision making and serve as promising therapeutic targets. Because TMA-based studies are best used as screening analyses, these findings should be further evaluated in large scale by conventional means.

Acknowledgements

The authors thank Mervi Eeva and Sheila Tze for their excellent technical assistance. The authors also thank Sigrid Rajasekaran for critical reading of the article and members of the Rajasekaran laboratory for helpful discussions.

Ancillary