Purpose: The 14-3-3 family proteins are highly conserved over many mammalian species. The σ isoform (also called HME-1 or stratifin) is expressed in epithelial cells. Loss of 14-3-3σ is associated with failure to arrest the cell cycle at the G2-M phase checkpoint after DNA damage that leads to increased G2-type chromosomal aberrations. The role of 14-3-3σ in prostatic carcinogenesis is uncertain.
Experimental Design: We studied one hundred and eleven specimens of invasive prostate adenocarcinoma with paired, adjacent high-grade prostatic intraepithelial neoplasia and normal prostate epithelium. Immunohistochemistry was used to detect the expression of 14-3-3σ. The findings were correlated with various clinical pathological parameters.
Results: 14-3-3σ is ubiquitously expressed at high levels in normal prostate epithelium. Its expression is significantly decreased in prostatic intraepithelial neoplasia and prostatic adenocarcinoma. Ninety percent of samples of prostatic intraepithelial neoplasia had no or low 14-3-3σ expression. Ninety-seven percent of invasive adenocarcinomas had no or low 14-3-3σ expression. In most specimens (90%), suppression of 14-3-3σ expression occurred during the development of prostatic intraepithelial neoplasia from normal epithelium.
Conclusions: Our data suggest that loss of 14-3-3σ contributes to the development of prostate adenocarcinoma. 14-3-3σ expression is significantly decreased during the progression of normal prostatic epithelium to prostatic intraepithelial neoplasia and invasive cancer.
The 14-3-3 family of proteins consists of seven isoforms, which are highly conserved over many eukaryotic organisms (1 , 2) . Much recent interest has focused on the σ isoform (also called HME-1 or stratifin), which is expressed in human epithelial cells (3, 4, 5) . A crucial role of 14-3-3σ is its control of the G2 cell cycle checkpoint (6) . At G2 phase, cdc2-cyclin B1 normally enters the nucleus to initiate mitosis. In response to DNA damage, 14-3-3σ is induced in a p53-dependent manner and prevents the cdc2-cyclin B1 complex from entering the nucleus. These changes provide an opportunity for DNA repair of damage before further cell cycle progression (6 , 7) . Indeed, cells lacking 14-3-3σ function have impaired cell cycle control after DNA damage and increased genomic instability. Suppression of 14-3-3σ expression has been documented in transformed cell lines, including v-Ha-ras (ras)-transformed mammary cells (5) , SV40-transformed human keratinocytes (8) , and several cancers (9, 10, 11, 12) . Loss of 14-3-3σ expression is caused by DNA hypermethylation rather than gene deletion or mutation (9, 10, 11, 12, 13) and importantly, restoration of 14-3-3σ expression by the DNA demethylation agents (e.g., 5-aza-2′-deoxycytidine) normalize control of the cell cycle (11) .
To determine the role of 14-3-3σ in prostate cancer, we examined its expression in primary invasive prostate adenocarcinoma, high-grade prostatic intraepithelial neoplasia (PIN), and in adjacent normal prostate epithelium.
MATERIALS AND METHODS
Radical prostatectomy specimens (n = 111) containing invasive prostatic adenocarcinoma with adjacent high-grade PIN and normal prostate epithelium were obtained from the surgical pathology files of Indiana University Medical Center from 1990 to 1996. These cases were selected to represent the full spectrum of Gleason grade and pathological stages. The patients ranged in age from 44 to 77 years (mean = 63 years). Grading of the primary tumor from radical prostatectomy specimens was performed according to the Gleason system (14) . The Gleason scores ranged from 4 to 10. Pathological staging was performed according to the 1997 Tumor, Lymph Nodes, and Metastasis System (15) . The final pathological stages included T2a (11 patients), T2b (46 patients), T3a (33 patients), and T3b (21 patients). Seventeen patients had lymph node metastases at the time of surgery. This research was approved by the Indiana University Institutional Review Board.
Generation of 14-3-3σ Antibody.
Goat polyclonal antibody specific to 14-3-3σ was generated by immunizing goat with a peptide mapping near the NH2 terminus of 14-3-3σ of human origin. The antisera was affinity purified. The specificity of the purified IgG antibody was confirmed by Western blotting and immunoprecipitation assay. No cross-reactivity with other 14-3-3 isoforms was observed.
Serial 5 μm-thick sections prepared from formalin-fixed, paraffin-embedded slices of prostate adenocarcinoma specimens were used for the study. Tissue blocks that contained the maximum amount of tumor and highest Gleason score were selected for each case. One representative slide from each case was analyzed. We recognized the limitation of sample variation. Slides were deparaffinized in xylene twice for 5 min and rehydrated through graded ethanol solutions to distilled water. Antigen retrieval was carried out by heating sections in 1 mm of EDTA (pH 8.0) for 30 min. Endogenous peroxidase activity was inactivated by incubation in 3% H2O2 for 15 min. Nonspecific binding sites were blocked using Protein Block (DAKO Corp) for 20 min. Tissue sections were then incubated with the goat polyclonal antibody against human 14-3-3σ (1:200 dilution, Santa Cruz Biotechnology, Santa Cruz, CA) at 4°C overnight, followed by incubation with biotinylated donkey antigoat antibody, and peroxidase-labeled streptavidin. Diaminobenzidine was used as the chromogen in the presence of hydrogen peroxide. Positive and negative controls were run in parallel with each batch and demonstrated that the procedure functioned properly.
Evaluation of 14-3-3σ Expression.
The extent and intensity of immunoreactivity for 14-3-3σ were evaluated in benign epithelium, PIN, and adenocarcinoma from the same slide for each case. Microscopic fields with the highest degree of immunoreactivity were chosen for analysis. At least 1000 cells were analyzed in each case. The percentage of cells exhibiting staining in each case was evaluated semiquantitatively on a 5% incremental scale ranging from 0 to 95%. A numeric intensity score was set from 0 to 3 (0, no staining; 1, weak staining; 2, moderate staining; and 3, strong staining). These methods were described previously (16, 17, 18) .
The mean percentage of immunoreactive cells in benign epithelium, high-grade PIN, and adenocarcinoma were compared using the Wilcoxon-paired signed rank test. The intensities of staining for 14-3-3σ in benign epithelium, high-grade PIN, and adenocarcinoma were compared using Cochran-Mantel-Haenszel tests for correlated ordered categorical data. A P < 0.05 was considered significant, and all Ps were two-sided.
Patient characteristics were illustrated in Table 1⇓ . Significant differences in 14-3-3σ immunoreactivity distinguished normal prostate epithelium, PIN, and prostate adenocarcinoma. In normal prostate epithelium, expression of 14-3-3σ was detected in all samples (111 of 111; Table 2⇓ ). The distribution of negative, weak, moderate, and strong staining for 14-3-3σ was 0% (0 of 111), 3% (3 of 111), 35% (39 of 111), and 62% (69 of 111), respectively. The median percentage of cells in each specimen that stained positively for 14-3-3σ expression was 86% in normal epithelium.
The intensity and percentage of cells that reacted with 14-3-3σ antibodies were significantly lower in PIN and invasive prostate adenocarcinoma (P < 0.001; Fig. 1⇓ ; Table 2⇓ ). The majority of PIN cells (90%) showed negative or weak 14-3-3σ immunoreactivity. Notably, none of the PIN cells reacted strongly with 14-3-3σ antibodies. An even more dramatic decrease in 14-3-3σ immunoreactivity was observed in invasive adenocarcinoma cells. In those specimens, 76% of adenocarcinoma cells lacked any 14-3-3σ immunoreactivity. The little 14-3-3σ immunoreactivity that was observed was primarily defined as weak staining (20%; 22 of 111) with some moderate staining in a few samples (4%; 4 of 111) and strong staining in 1% (1 of 111). The mean percentages of the specimens with 14-3-3σ staining in PIN and invasive cancer were 4% (SD, 5.9) and 2% (SD, 4.8), respectively.
In most cases, decreased 14-3-3σ expression related to the progression from normal epithelium to PIN. Ninety percent (100 of 111) of samples of PIN had either no (55%) or low (35%) levels of 14-3-3σ expression. The frequency and intensity of 14-3-3σ expression further decreased from PIN (90% of negative or weak expression) to invasive cancer (97% of negative or weak expression).
We also assessed whether other clinical parameters related to 14-3-3σ immunoreactivity. Adenocarcinomas with high Gleason scores (>7) had significantly higher staining intensities (0.5 ± 0.7 versus 0.1 ± 0.2; P = 0.008) and higher percentages of 14-3-3σ immunoreactive cells (4.5 ± 7.1 versus 0.3 ± 1.2; P < 0.001) than adenocarcinomas with low Gleason scores (<7; Fig. 2⇓ ). Adenocarcinomas with lymph node metastases had higher percentages of 14-3-3σ expression (5.6 ± 8 versus 1.5 ± 3.5; P = 0.001), when compared with adenocarcinomas without lymph node metastases. Overall, all invasive prostate adenocarcinomas had low intensities and low percentages (<5%) of 14-3-3σ staining. There was no significant correlation between the level of 14-3-3σ expression and other clinical and pathological features, including patient age, extraprostatic extension, vascular invasion, surgical margin, perineural invasion, or the presence of high-grade PIN.
This is the first report that 14-3-3σ expression is suppressed during the transformation of prostatic epithelium to adenocarcinoma. Considering its function in the regulation of the G2-M phase checkpoint and in the maintenance of genomic integrity (6 , 7) , decreased 14-3-3σ expression likely contributes to increased mutability and well-established ability of malignant cells to overcome microenvironmental and therapeutic challenges, which frequently characterize aggressive tumor cells.
In breast cancer, suppression of 14-3-3σ arises as malignant behavior progresses from late atypical hyperplastic lesions to ductal carcinoma in situ (12) . Our data also indicate that decreased 14-3-3σ expression similarly represents an early event during the formation of prostate adenocarcinoma. Notably, 14-3-3σ was not always uniform within a particular specimen. These islands of prostate tumor cells, which varied with regard to 14-3-3σ expression, are likely representative of the multifocal nature of prostatic adenocarcinoma (19) . Alternatively, these differences may represent tumor cell clones that have decreased 14-3-3σ expression as they progress toward a malignant phenotype. Nonetheless, we should emphasize that our data cannot discriminate between a causal role for 14-3-3σ in prostate cancer and the alternative that the decreased 14-3-3σ expression is an effect of the carcinogenic process. We also note a paradoxical increase of 14-3-3σ expression as tumor progress (Fig. 2)⇓ . In some cases, islands of tumor cells with and without 14-3-3σ expression coexisted in the same specimens (Fig. 1L)⇓ . This suggested that clones of cells without 14-3-3σ expression may arise after cancer has developed. Cells with decreased 14-3-3σ expression may develop mitotic failure as a result of DNA damage. However, the cells that do survive have a high probability of mutation and escape the growth restriction of their own mitotic clocks and surrounding microenvironments. This may explain the phenomenon that cancer populations are not homogeneous (20 , 21) . For metastatic cancer, single-modality treatment usually fails to eradicate cancer cells because resistant cells will arise during treatment (22) .
In conclusion, our findings are novel, in part, because they show that 14-3-3σ expression significantly decreases during the progression from normal prostate tissue to PIN and further to invasive cancer. We postulate that decreased 14-3-3σ expression contributes to the deregulation of genetic stability, which frequently epitomizes invasive prostate cancer.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Requests for reprints: Liang Cheng, Department of Pathology and Laboratory Medicine, Indiana University Medical Center, University Hospital 3465, 550 North University Blvd., Indianapolis, IN 46202. Phone: (317) 274-1756; Fax: (317) 274-5346; E-mail:
- Received November 28, 2003.
- Revision received January 12, 2004.
- Accepted January 20, 2004.