Journal of Clinical Oncology, Vol 22, No 14 (July 15), 2004: pp. 2790-2799
© 2004 American Society of Clinical Oncology.
DOI: 10.1200/JCO.2004.05.158
Gene Expression Alterations in Prostate Cancer Predicting Tumor Aggression and Preceding Development of Malignancy
Yan Ping Yu,
Douglas Landsittel,
Ling Jing,
Joel Nelson,
Baoguo Ren,
Lijun Liu,
Courtney McDonald,
Ryan Thomas,
Rajiv Dhir,
Sydney Finkelstein,
George Michalopoulos,
Michael Becich,
Jian-Hua Luo
From the Department of Pathology and Urology, University of Pittsburgh School of Medicine, and Biostatistics Center, University of Pittsburgh Cancer Institute and Department of Biostatistics, University of Pittsburgh, Pittsburgh, PA
Address reprint requests to Jian-Hua Luo, MD, University of Pittsburgh Cancer Institute and Department of Pathology, University of Pittsburgh, 3550 Terrace St, Pittsburgh, PA 15261; e-mail: luoj{at}msx.upmc.edu
 |
ABSTRACT
|
|---|
PURPOSE: The incidence of prostate cancer is frequent, occurring in almost one-third of men older than 45 years. Only a fraction of the cases reach the stages displaying clinical significance. Despite the advances in our understanding of prostate carcinogenesis and disease progression, our knowledge of this disease is still fragmented. Identification of the genes and patterns of gene expression will provide a more cohesive picture of prostate cancer biology.
PATIENTS AND METHODS: In this study, we performed a comprehensive gene expression analysis on 152 human samples including prostate cancer tissues, prostate tissues adjacent to tumor, and organ donor prostate tissues, obtained from men of various ages, using the Affymetrix (Santa Clara, CA) U95a, U95b, and U95c chip sets (37,777 genes and expression sequence tags).
RESULTS: Our results confirm an alteration of gene expression in prostate cancer when comparing with nontumor adjacent prostate tissues. However, our study also indicates that the gene expression pattern in tissues adjacent to cancer is so substantially altered that it resembles a cancer field effect.
CONCLUSION: We also found that gene expression patterns can be used to predict the aggressiveness of prostate cancer using a novel model.
|
INTRODUCTION
|
|---|
Prostate cancer is second only to skin cancer as the most commonly diagnosed malignancy in American men: at current rates of diagnosis, one man in six will be diagnosed with the disease during his lifetime. Approximately 30,000 men will die from this disease annually.1 Against the backdrop of a common malignancy, it is not clear what molecular events are responsible for the progression of prostate cancer to a lethal form of the disease. Completion of human genome sequencing has provided the basis for a comprehensive genetic profile of a disease. Application of high-throughput quantitative analysis of gene expression in prostate cancer samples should help to exploit the information of gene expression in prostate cancers and to advance our understanding of the disease.
In this report, we performed a comprehensive gene expression analysis on 152 human prostate samples, including prostate cancer (PC), prostate tissues adjacent to (AT) cancer, and donor (OD) prostate tissue totally free of disease, using the Affymetrix (Santa Clara, CA) U95a, U95b, and U95c chip sets. We identified a set of 671 genes whose expression levels were significantly altered in PCs compared with normal tissues. Interestingly, the expression patterns of histological benign prostate tissues were significantly overlapped with those of PC, and were distinctly different than donor prostate tissue. Separately, a "70-gene" model was developed to predict the aggressiveness of the disease. Collectively, these data suggest that genetic alterations in a gland with PC are not limited to the malignant cells, and these patterns of alteration may predict the population both at risk for the disease and for disease progression.
|
PATIENTS AND METHODS
|
|---|
Sample Preparation
Fresh prostate tissues, recovered immediately from the operating room after removal, were dissected and trimmed to obtain pure tumor (completely free of normal prostate acinar cells) or normal prostate (free of tumor cells) tissues. Microdissection was coupled with sandwich frozen and permanent section analyses to confirm the purity and homogeneity of the samples: gross and microscopic analyses were performed by board-certified genitourinary pathologists. For tumor tissues, only samples with less than 30% of stromal components were selected. For donor prostate tissues, obtained at the time of organ donation in brain-dead men, samples from peripheral zone of the prostate gland with at least 60% glandular components and free of any pathological alteration were selected (Table 1) . For prostate tissues adjacent to cancer, samples free of cancer cells, high-grade prostatic neoplasia, or any obvious neoplastic alterations, containing at least 60% glandular cells, were selected. Whenever possible, all tissues were processed and frozen within 30 minutes after removal. These tissues were then homogenized. All patients with PCs have at least a 4-year follow-up, with regular evaluations for relapse or the presence of metastasis. Protocols for tissue banking, tissue anonymization, and tissue processing, were approved by the institutional review board.
Affymetrix Chip Analysis
cRNA preparation.
Total RNA was extracted and purified with Qiagen RNeasy kit (Qiagen, San Diego, CA). Five micrograms of total RNA were used in the first strand cDNA synthesis with T7-day(T)24 primer (GGCCAGTGAATTGTAATACGACTCACTATAGGGAGGCGG-(dT)24) by Superscript II (GIBCO-BRL, Rockville, MD). The second strand cDNA synthesis was carried out at 16°C by adding Escherichia coli DNA ligase, E coli DNA polymerase I, and RnaseH in the reaction. This was followed by the addition of T4 DNA polymerase to blunt the ends of newly synthesized cDNA. The cDNA was purified through phenol/chloroform and ethanol precipitation. The purified cDNA were then incubated at 37°C for 4 hours in an in vitro transcription reaction to produce cRNA labeled with biotin using MEGAscript system (Ambion Inc, Austin, TX).
Affymetrix chip hybridization.
Between 15 and 20 µg of cRNA were fragmented by incubating in a buffer containing 200 mmol/L Tris-acetate, pH 8.1, 500 mmol/L KOAc, and 150 mmol/L MgOAc at 95°C for 35 minutes. The fragmented cRNA were then hybridized with a pre-equilibrated Affymetrix chip at 45°C for 14 to 16 hours. After the hybridization cocktails were removed, the chips were then washed in a fluidic station with low-stringency buffer (6x sodium chloride, sodium phosphate dibasic, and EDTA; 0.01% Tween 20; 0.005% antifoam) for 10 cycles (two mixes/cycle), and stringent buffer (100 mmol/L MES, 0.1 M NaCl and 0.01% Tween 20) for four cycles (15 mixes/cycle), and stained with Strepto-avidin Phycoerythrin (SAPE; Molecular Probe, Eugene, OR). This was followed by incubation with biotinylated mouse antiavidin antibody, and restained with SAPE. The chips were scanned in a HP ChipScanner (Affymetrix Inc) to detect hybridization signals. For quality assurance, all samples were run on Affymetrix test-3 chips to evaluate the integrity of RNA; samples with RNA 3'/5' ratios less than 2.5 were accepted for further analysis.
Data Analysis
Hybridization data were normalized to an average target intensity of 500 per chip, and were converted to Microsoft Excel spreadsheet text file (Redmond, WA). The primary comparison of OD to PC was conducted through the following steps: (1) Two-sample t tests of log-transformed gene expression values, (2) adjustment of P values through the Benjamini and Hochberg procedure,2 (3) selection of genes that meet both the critical P value and show at least a two-fold change in PC, (4) reduction of dimensionality through principal component analysis, (5) prediction of case status (ie, normal v cancer tissue) through logistic regression, and (6) evaluation of the classification rate using 10-fold cross-validation. Regarding the second step, the Benjamini and Hochberg procedure calculates a conservative P value to minimize the expected number of falsely significant results. For tests between PC and AT, the paired t test (of log-transformed expressions) was utilized to account for the matching. A sufficient number of principle components (in the fourth step) were retained to quantify at least 90% of the variability in these genes. For the cross validation procedure (sixth step), a separate logistic model is fit for each of the ten subsets used for training, and then used to predict the outcome for the remaining subset of validation data. After this process is implemented for classifying donors versus PC, the resulting model parameters (using the entire data set) were saved and utilized to predict case status of adjacent to tumor normal tissue. The fitted logistic model (again using the entire data set) was also used to classify separate validation data sets collected from other institutions. These analyses were all conducted using S-PLUS statistical software (Insightful Corp, Seattle, WA).
In addition to these analyses, tumor aggressiveness was predicted using a separate procedure, conducted using GeneSpring software (version 4.2; Silicon Genetics Inc, Redwood City, CA). This procedure starts with the ordered list of the five most differentially expressed genes, and sequentially adds subsets of five genes, where the classification rate is evaluated by leave-one-out cross-validation3 (using the automated algorithm in the GeneSpring program). Prediction of aggressiveness was made on the basis of a cutoff P value ratio of 0.3 of the expression profile of the "leave-out" sample, with the mean expression profiles of the remaining samples from the aggressive and nonaggressive tumors. This process was found to achieve better prediction (as compared with the procedure for donors v PC) for classifying tumor aggressiveness. Finally, clusters of gene expression were visually displayed using Michael Eisen’s cluster and tree view software3a (for both donors v PC and aggressive versus nonaggressive tumors).
Immunohistochemistry Staining and Tissue Array Analysis
Formalin-fixed and paraffin-embedded human prostate tissues including PC, OD prostate samples, and benign prostate tissue samples from prostate glands containing PC were used. For immunostaining, 4-µm-thick sections of tissue array were cut and mounted on glass slides. The sections were heated at 60°C for 12 hours and deparaffinized in xylene and ethanol. Antigen retrieval was performed using 25 mmol/L sodium citrate buffer (pH 9.0) at 90°C for 15 minutes, followed by treatment of 3% H2O2 to block endogenous peroxidase. The slides were incubated at room temperature for 2 hours with anti-glutathione-S-transferase pi (GSTpi) and anti- -methylacyl-CoA racemase (AMACR) antibodies at 1:1000 and 1:400 dilution, respectively. The sections were then incubated with horseradish peroxidase-conjugated antirabbit immunoglobulin G for 30 minutes at room temperature. This was followed by incubating the section with 3,3' diaminobenzidine solution (DAKO, Carpentaria, CA) to develop staining color. Hematoxylin was used for counterstaining. The specificities of immunostainings were verified by incubating the similar slides with preimmune sera.
|
RESULTS
|
|---|
To more accurately compare PC samples with morphologically normal prostate tissues, we used OD prostate samples free of any histological change, as our normal samples in the analysis. These OD prostate samples were obtained from men aged 13 to 63 years with no clinical or histological evidence of prostatic disease. They had no history of genitourinary illness. A second comparison was made between PC and histologically benign prostate tissue adjacent to the cancer.
For this study, we analyzed a total of 152 samples plus 23 validation samples from a separate institution. These samples include 66 PC, 23 OD, 60 AT, and three cell lines. The vast majority of AT samples had corresponding and matched PC samples. There were, however, some AT samples without a corresponding and matched PC sample (n = 5; 8%), likewise, a few PC samples lacked a corresponding and matched AT sample (n = 11; 17%). A separate validation set has also been included in this analysis. The 23 independent samples were obtained from another investigator at another institution. These samples therefore serve as a truly unbiased validation set.
Field Effect of PC
To maximize the difference between the three groups of tissues (PC, OD, and AT), we eliminated genes whose expression was very similar throughout all the samples (correlation coefficient 0.95 of expression mean of a gene). We also eliminated genes consistently displaying low levels of expression, because signal intensity below 200 arbitrary units contains a high error rate ( 40%), and lacks correlation with other quantifying methodologies such as Taqman reverse transcriptase-polymerase chain reaction or Northern blot analysis.4 Overall, this process eliminated 18,638 genes and ESTs. Analysis of the remaining 19,139 genes and ESTs used a cutoff P value less than .002 between OD and PC, leading to an expected maximum of 1% of the significant results being false-positives. Therefore, 99% of the observed differences were expected to be authentic. Further, we require at least a two-fold variation in expression between the two groups (Fig 1A). Obviously, restricting gene selection in this fashion can be considered somewhat arbitrary—eliminating many of the otherwise statistically significant genes—but this procedure insures verifiable results. Applying this rule, the majority of the 19,139 genes were excluded, leading to 671 genes and ESTs (P < .002 and a greater than two-fold variation) being statistically significant (Fig 1B and Table 1).
View larger version (39K):
[in this window]
[in a new window]
|
Fig 1. Differential gene expression between normal prostate donor and prostate cancer. (A) P values of 19,139 genes and ESTs were plotted with prostate cancer (PC) and organ donor (OD) data. Red diamonds represent genes with P < .002 and two-fold variation. (B) Expression profiles of 671 genes and ESTs in OD and PC. Red represents strong expression, green for weak, and black for less than 200 units. (C) Probability prediction of prostate cancer on prostate tissues adjacent to cancer samples.
|
|
Principle component analysis was performed to simplify the data into major expression patterns.5 A model was developed to predict normal prostate versus PC. The first 11 expression patterns (principle components) explained 90% of the variability in all 671 genes. Then, a logistic regression was fit using the 11 principle components as covariates.5 When using gene expression data (671 genes) on the 89 donor and tumor samples, all 66 PC samples were predicted to be tumor with very high predicted probabilities (> 0.999), and all 23 donor samples were predicted to be tumor with low predicted probabilities (< 0.001).
These results, however, are subject to resubstitution bias, as the same data are used to fit and evaluate (classification rates of) the model. To better validate this model, we used 10-fold cross-validation. That is, data were first randomly partitioned into 10 roughly equal sets (each of n = 8 to 9); classification of case status for each subset was then conducted by using the logistic model fitted with the other 90% of the data. Although this approach does leave some potential for bias (because the selection of significant genes and principle components is not repeated for each subset due to computational demands of doing so), such potential is minimized by using separate logistic models for the training and validation phases. Using this cross-validation approach, all 66 tumor samples were still classified as tumors (with predicted probabilities all above 0.999), and 22 of 23 (96%) donor samples were still classified as donors (with the 22 correct classifications all corresponding to predicted probabilities below 0.001). To ensure that the significant genes within each cross-validation set were representative of the overall list of significant genes, we repeated the tests of differential expression within each cross-validation set, including the t tests, fold-changes, and Benjamini-Hochberg adjustment. Results of this additional analysis indicate that most genes that are significant in the overall data set are still significant within a given cross-validation set (ie, more than 80% of those genes in most cross-validation sets); the converse is also true, as approximately 99% of the nonsignificant genes are still nonsignificant within each cross-validation set. To further validate our model, we tested our logistic regression model on an independent data set of 23 tumor samples obtained from another institute. Twenty-one (91.3%) of 23 PC samples were predicted as having a greater than 0.99 probability of PC, and one, as having a 0.86 probability. Only one sample was predicted as possibly normal (0.403). To rule out the potential age effect on our logistic model, we divided the OD group into a group of those younger than 45 years, and a group  45 years of age. We then fit a logistic model to the same 11 principle components using only the PC versus OD younger than 45 years. We then used this logistic model to obtain predicted probabilities for the 45-year-old group. All 11 older donors demonstrated a probability of "PC" less than 0.001. This clearly demonstrates similar genetic expression between the young- and older-age groups.
The same logistic regression model described above was then used to evaluate AT in terms of their relationship with OD and PC. Ninety percent (54 of 60) of AT tissues were predicted to be tumor with high ( 0.93) probability (Fig 1C), whereas 10% (6 of 60) of AT were predicted to be tumors with low probability ( 0.13). Not surprisingly, all three PC cell lines (PC-3, DU145, and LNCaP) were predicted to be tumor with high probability (> 0.99).
Only 25 genes and ESTs were found to be differentially expressed between AT (n = 55) and PC using the rigorous screenings (pair-wise t test P < .0002 as determined by Benjamini and Hochberg; > two-fold) described above, derived to compare PC with OD samples. However, when statistical stringency was relaxed to P < .05 and 1.5-fold variation, 53% (358) of the 671 genes and expression sequence tag were found differentially expressed between PC and AT. To test the hypothesis that the histological benign tissues adjacent to cancer may have undergone genetic changes similar to PC (the "field effect"), we tested the gene expression patterns in AT and PC versus OD. Genes and ESTs were selected for P < .002 (AT v OD or PC v OD). One thousand twenty-two genes were found differentially expressed in AT samples in comparison with OD. Interestingly, the expression of the majority (710 or 70%) of these genes were similarly altered in tumor samples (v donors), suggesting a general similarity of expression patterns between AT and PC. It appears, therefore, the patterns of gene expression in AT are much more similar to PC than donor prostate, supporting the "field effect" hypothesis. Not surprisingly, however, the patterns of gene expression in PC were unique, where 2,434 genes and ESTs were found to have P < .002 when compared with OD: 36% (874 of 2,434) are uniquely overexpressed, and 35% (850 of 2,434) were uniquely underexpressed in PC, compared with OD.
To identify the gene expression characteristic of AT, we searched the expression data for well-known PC-related genes. This exercise was performed to determine how similar the AT samples were to PC. We separated the AT samples into three groups based on the gene expression prediction metrics for PC defined previously (ie, low probability of tumor [< 0.2, AT1], inconclusive [0.2 to 0.9, AT2], and high probability [> 0.90, AT3]). As presented in Table 2, no cancerlike expression of PC-related genes were found in the low-probability group (AT1). The expression of several PC-related genes in AT3 group, however, was found to have a pattern overlapping significantly with those occurring in organ-confined PC. The magnitude and number of gene expression alterations in PC-related genes increased in parallel with the clinical progression of PC; so do the number of genes that show differences in expression from OD normal tissues (Fig 2A) .
View larger version (47K):
[in this window]
[in a new window]
|
Fig 2. Abnormality of gene expression of prostate cancer (PC) and prostate tissues adjacent to cancer (AT) samples. (A) Number of genes in AT1, AT2, AT3, organ-confined, minimally invasive, and aggressive PC having P value less than .002 when comparing with organ donor (OD) samples is plotted. (B) Immunohistochemistry staining of GSTpi and AMACR on OD, AT, and PC samples. Top panel represents immunostaining with anti-AMACR antisera, bottom with anti-GSTpi monoclonal antibody.
|
|
To observe the field effect of gene expression in AT samples, the protein expression of two PC signature genes (GSTpi and AMACR) were immunostained using antibodies against GSTpi and AMACR in OD, AT, and PC samples. As shown in Figure 2B, appreciable upregulation of AMACR and downregulation of GSTpi were seen in AT samples. Stronger alterations were seen in cancer samples. These results clearly support our field effect hypothesis.
Gene Expression Model for Prediction of Aggressiveness
Based on the increasingly altered pattern of gene expression observed in Figure 2, it is our hypothesis that gene expression pattern can predict aggressive behavior of PC. We divided the PC samples into two groups based on the observed clinical aggressiveness, where an aggressive tumor is defined by any of the following: cancer invasion into adjacent organs or seminal vesicles, clinical relapse evidenced by an increase of PSA level following radical prostatectomy, or distant metastasis. Nonaggressive tumors were defined by lacking the above features regardless of tumor differentiation. We reverse ranked the P value of gene expression between the two groups. Seventy-two genes and ESTs has a P value less than .002 (Fig 3A; Table 3). Starting with the top five genes, we found the expression profile model 48% accurate in predicting aggressiveness. We sequentially added five genes at a time to improve the prediction accuracy of the model, and found the model with gene number 70 being the best (Fig 3B). The "70-gene" model correctly predicted 27 of 29 aggressive tumors, and 32 of 37 nonaggressive tumors (Fig 3C), producing 86.5% specificity and 93% sensitivity. Sixteen (88%) of 18 PCs with postoperational PSA failure were also predicted as "aggressive."
View larger version (39K):
[in this window]
[in a new window]
|
Fig 3. Gene expression pattern of aggressive and nonaggressive prostate cancers (PCs) and prediction of aggressiveness by 70-gene "aggressiveness predictor" model. (A) Seventy-predictor gene expression profiles in PC samples. Red represents strong expression, green for weak, and black for less than 200 units. (B) Prediction of PC aggressiveness by gene expression profile. (C) Plot of predictor P value ratio for 66 tumor samples using "leaving-one-out model". P value ratio less than .3 is used to predict likely aggressive PC.
|
|
We performed a similar analysis based on tumor differentiation (Gleason score) to predict aggressiveness. Excluding four metastatic samples (in which Gleason score is not assigned), 45 samples had Gleason scores equal to or greater than 7 (poorly differentiated) and 17 samples 6 (well to moderately differentiated). As presented in Table 4, Gleason score has only limited accuracy in predicting true aggressiveness status, with 33 of 62 (53%) correctly classified. It appears that Gleason score is sensitive (22 of 25; 88% sensitivity), but is less specific (14 of 37; 38% specificity; P < .001 in McNemar’s test6) than the "70-gene" model.
To evaluate the validity of this model, we tested this "70-gene" model on an independent set of data from 23 cases of aggressive and nonaggressive primary PCs. The 70-gene model was more accurate overall (78% v 52%) than Gleason scores and had higher specificity (82% v 9%; P = .02 in McNemar’s test) in predicting aggressiveness. Although the 70-gene model is slightly less sensitive (75% versus 92%), the differences were not significant (P = .63 in McNemar’s test). Seven (77%) of nine cases of PC with subsequently recurrence were also predicted as "aggressive," suggesting similar effectiveness of this model to predict potential clinical relapse of PC.
Within the 70-gene aggressiveness predictor, ESTs or genes with unknown function dominate the list and comprise 64% (45 of 70) of the sequences, indicating the potentially important roles of these sequences in understanding the aggressive behavior of PC. Of the 25 known genes, p57kip2, a cyclin-dependent kinase inhibitor,7 seems downregulated in aggressive tumors versus nonaggressive tumor samples. This gene is implicated in sporadic cancers and a familiar cancer syndrome—Beckwith-Wiedemann syndrome. Cdc42 effector protein 3 is a protein regulating formation of F-actin-containing structures.8 Mutations and deletions of genes involved in actin-bundling functions are associated with PC metastases,9,10 implying the downregulation of Cdc42 effector protein 3 in similar role. Upregulation of TML1, a oncogene implicated in T-cell leukemia,11 is the most consistently overexpressed (over nonaggressive tumors) genes (28 of 29) in the "70-gene" model. Finally, the predictor list contains several genes in G protein signaling pathway (indeed, the largest set of genes associated with a single pathway), including G protein beta subunit, G protein coupled receptor 12, and Rab-escorted protein 1. This implies a role for G protein related signal transduction in aggressive PC.
|
DISCUSSION
|
|---|
As shown in this study, comparing gene expression patterns between histologically cancer-free tissues and neighboring malignancy can result in a significantly limited set of differences. Patterns of gene expression in benign prostate tissue adjacent to a PC have been significantly altered, and in some aspects, resemble cancerous changes. The implications for a clearer understanding of prostate carcinogenesis are obvious: gene expression patterns in histologically benign tissues must be more thoroughly examined. The clinical observation of multiple, diffuse yet geographically separate and distinct cancers arising with the same prostate gland led to the field effect hypothesis. The gene expression patterns described in this study further support the notion of a gland fertile for transformation.
Unlike the body of work described here, previous studies using cDNA array in PCs12-20 have examined the difference between PCs and adjacent nontumor prostate tissues, not unlike the AT described above. The current study further substantiates the conclusion that there is a unique pattern of gene expression alteration in PC. Indeed, when comparing the genelist of 671 genes of the current study with those previous ones,12-21 more than 70% of these gene and EST sequence were overlapped, even though the array platform, gene numbers, and number of cases have been varied widely. The large number of samples and availability of organ donor prostate samples in the current study allow a better subclassification of the prostate samples based on gene expression patterns. Our study clearly indicates the heterogeneity in gene expression is not limited to PCs. This finding is very unlikely to represent a coincidence, given the high level of similarity of quantitative gene expression alterations between a subset of AT and a subset of PCs. It is reasonable to speculate that genetic alterations occur even before cells are morphologically transformed. This finding is also consistent with the current concept of a step-wise progression of neoplasia seen in other cancers (colon22 and others). Furthermore, our study supports a "field effect" hypothesis in the prostate that develops cancer.
By performing comprehensive gene expression analysis on a large number of samples, we demonstrate the feasibility of predicting PC aggressiveness based on gene expression patterns. Since only a fraction of PCs are metastatic, identifying variables that predicting the behavior of this tumor should prove important in clinical management. Gene expression predictors may supplement the current morphology-based diagnosis systems by offering mechanistic insight, by eliminating the subjective nature of the tumor grading systems, and by providing potential therapeutic targets.
|
Authors’ Disclosures of Potential Conflicts of Interest
|
|---|
The authors indicated no potential conflicts of interest.
|
Acknowledgment
|
|---|
We thank Tracy Wagner and Mark Rubin for providing tissues necessary for this work; William L. Gerald for providing independent set of prostate cancer data files; and Uma Chandran and John Gilbertson for constructive comments.
|
NOTES
|
|---|
Supported by grants 1UO1CA88110-01 (G.M.) and R01 CA098249 (J.H.L.) from the National Cancer Institute.
Authors’ disclosures of potential conflicts of interest are found at the end of this article.
|
REFERENCES
|
|---|
1. Jemal A, Murray T, Samuels A, et al: Cancer statistics, 2003. CA Cancer J Clin 53:5-26, 2003[Abstract/Free Full Text]
2. Benjamini Y, Hochberg Y: Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Stat Soc B 57:289-300, 1995
3. Golub TR, Slonim DK, Tamayo P, et al: Molecular classification of cancer: Class discovery and class prediction by gene expression monitoring. Science 286:531-537, 1999[Abstract/Free Full Text]
3. Eisen MB, Spellman PT, Brown PO, et al: Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A. 95:14863-14868, 1998[Abstract/Free Full Text]
4. Luo JH, Yu YP, Cieply K, et al: Gene expression analysis of prostate cancers. Mol Carcinog 33:25-35, 2002[CrossRef][Medline]
5. Myers RH: Classical and Modern Regression With Applications. Boston, MA, PWS-KENT, 1990
6. Rosner B: Fundamentals of Biostatistics (ed 3). Boston, MA, PWS-KENT, 1990
7. Algar E, Brickell S, Deeble G, et al: Analysis of CDKN1C in Beckwith Wiedemann syndrome. Hum Mutat 15:497-508, 2000[CrossRef][Medline]
8. Hirsch DS, Pirone DM, Burbelo PD: A new family of Cdc42 effector proteins, CEPs, function in fibroblast and epithelial cell shape changes. J Biol Chem 276:875-883, 2001[Abstract/Free Full Text]
9. Lin F, Yu YP, Woods J, et al: Myopodin, a synaptopodin homologue, is frequently deleted in invasive prostate cancers. Am J Pathol 159:1603-1612, 2001[Abstract/Free Full Text]
10. Lutchman M, Pack S, Kim AC, et al: Loss of heterozygosity on 8p in prostate cancer implicates a role for dematin in tumor progression. Cancer Genet Cytogenet 115:65-69, 1999[CrossRef][Medline]
11. Pekarsky Y, Hallas C, Isobe M, et al: Abnormalities at 14q32.1 in T cell malignancies involve two oncogenes. Proc Natl Acad Sci U S A 96:2949-2951, 1999[Abstract/Free Full Text]
12. Luo J, Duggan DJ, Chen Y, et al: Human prostate cancer and benign prostatic hyperplasia: Molecular dissection by gene expression profiling. Cancer Res 61:4683-4688, 2001[Abstract/Free Full Text]
13. Dhanasekaran SM, Barrette TR, Ghosh D, et al: Delineation of prognostic biomarkers in prostate cancer. Nature 412:822-826, 2001[CrossRef][Medline]
14. Ernst T, Hergenhahn M, Kenzelmann M, et al: Decrease and gain of gene expression are equally discriminatory markers for prostate carcinoma: A gene expression analysis on total and microdissected prostate tissue. Am J Pathol 160:2169-2180, 2002[Abstract/Free Full Text]
15. Bull JH, Ellison G, Patel A, et al: Identification of potential diagnostic markers of prostate cancer and prostatic intraepithelial neoplasia using cDNA microarray. Br J Cancer 84:1512-1519, 2001[CrossRef][Medline]
16. Chaib H, Cockrell EK, Rubin MA, et al: Profiling and verification of gene expression patterns in normal and malignant human prostate tissues by cDNA microarray analysis. Neoplasia 3:43-52, 2001[CrossRef][Medline]
17. Welsh JB, Sapinoso LM, Su AI, et al: Analysis of gene expression identifies candidate markers and pharmacological targets in prostate cancer. Cancer Res 61:5974-5978, 2001[Abstract/Free Full Text]
18. Singh D, Febbo PG, Ross K, et al: Gene expression correlates of clinical prostate cancer behavior. Cancer Cell 1:203-209, 2002[CrossRef][Medline]
19. Stamey TA, Warrington JA, Caldwell MC, et al: Molecular genetic profiling of Gleason grade 4/5 prostate cancers compared to benign prostatic hyperplasia. J Urol 166:2171-2177, 2001[CrossRef][Medline]
20. Chetcuti A, Margan S, Mann S, et al: Identification of differentially expressed genes in organ-confined prostate cancer by gene expression array. Prostate 47:132-140, 2001[CrossRef][Medline]
21. LaTulippe E, Satagopan J, Smith A, et al: Comprehensive gene expression analysis of prostate cancer reveals distinct transcriptional programs associated with metastatic disease. Cancer Res 62:4499-4506, 2002[Abstract/Free Full Text]
22. Bronchud MH: Is cancer really a "local" cellular clonal disease? Med Hypotheses 59:560-565, 2002[CrossRef][Medline]
Submitted May 22, 2003;
accepted April 28, 2004.
 CiteULike  Complore  Connotea  Del.icio.us  Digg  Facebook  Reddit  Technorati  Twitter What's this?
This article has been cited by other articles:
|
|
|
|
 
A. P. Khan, L. M. Poisson, V. B. Bhat, D. Fermin, R. Zhao, S. Kalyana-Sundaram, G. Michailidis, A. I. Nesvizhskii, G. S. Omenn, A. M. Chinnaiyan, et al.
Quantitative Proteomic Profiling of Prostate Cancer Reveals a Role for miR-128 in Prostate Cancer
Mol. Cell. Proteomics,
February 1, 2010;
9(2):
298 - 312.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C. S. Moreno
The Sex-Determining Region Y-Box 4 and Homeobox C6 Transcriptional Networks in Prostate Cancer Progression: Crosstalk with the Wnt, Notch, and PI3K Pathways
Am. J. Pathol.,
February 1, 2010;
176(2):
518 - 527.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
F. Rizzi and S. Bettuzzi
The clusterin paradigm in prostate and breast carcinogenesis
Endocr. Relat. Cancer,
January 29, 2010;
17(1):
R1 - R17.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Y. Zhang, T. Z. Ali, H. Zhou, D. R. D'Souza, Y. Lu, J. Jaffe, Z. Liu, A. Passaniti, and A. W. Hamburger
ErbB3 Binding Protein 1 Represses Metastasis-Promoting Gene Anterior Gradient Protein 2 in Prostate Cancer
Cancer Res.,
January 1, 2010;
70(1):
240 - 248.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
A. Bhattacharya and R. K. De
Bi-correlation clustering algorithm for determining a set of co-regulated genes
Bioinformatics,
November 1, 2009;
25(21):
2795 - 2801.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. S. Arredouani, B. Lu, M. Bhasin, M. Eljanne, W. Yue, J.-M. Mosquera, G. J. Bubley, V. Li, M. A. Rubin, T. A. Libermann, et al.
Identification of the Transcription Factor Single-Minded Homologue 2 as a Potential Biomarker and Immunotherapy Target in Prostate Cancer
Clin. Cancer Res.,
September 15, 2009;
15(18):
5794 - 5802.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
D. E. Frigo, A. B. Sherk, B. M. Wittmann, J. D. Norris, Q. Wang, J. D. Joseph, A. P. Toner, M. Brown, and D. P. McDonnell
Induction of Kruppel-Like Factor 5 Expression by Androgens Results in Increased CXCR4-Dependent Migration of Prostate Cancer Cells in Vitro
Mol. Endocrinol.,
September 1, 2009;
23(9):
1385 - 1396.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. H. Muders, H. Zhang, E. Wang, D. J. Tindall, and K. Datta
Vascular Endothelial Growth Factor-C Protects Prostate Cancer Cells from Oxidative Stress by the Activation of Mammalian Target of Rapamycin Complex-2 and AKT-1
Cancer Res.,
August 1, 2009;
69(15):
6042 - 6048.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. E. Brennan, Y. Kuwano, N. Alkharouf, P. J. Blackshear, M. Gorospe, and G. M. Wilson
The mRNA-Destabilizing Protein Tristetraprolin Is Suppressed in Many Cancers, Altering Tumorigenic Phenotypes and Patient Prognosis
Cancer Res.,
June 15, 2009;
69(12):
5168 - 5176.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. Lupien and M. Brown
Cistromics of hormone-dependent cancer
Endocr. Relat. Cancer,
June 1, 2009;
16(2):
381 - 389.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
G. J. Kelloff, P. Choyke, D. S. Coffey, and for The Prostate Cancer Imaging Working Group
Challenges in Clinical Prostate Cancer: Role of Imaging
Am. J. Roentgenol.,
June 1, 2009;
192(6):
1455 - 1470.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
G. Wang, S. Haile, B. Comuzzi, A. H. Tien, J. Wang, T. M.K. Yong, A. E. Jelescu-Bodos, N. Blaszczyk, R. L. Vessella, B. A. Masri, et al.
Osteoblast-Derived Factors Induce an Expression Signature that Identifies Prostate Cancer Metastasis and Hormonal Progression
Cancer Res.,
April 15, 2009;
69(8):
3433 - 3442.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
G. Monami, V. Emiliozzi, A. Bitto, F. Lovat, S.-Q. Xu, S. Goldoni, M. Fassan, G. Serrero, L. G. Gomella, R. Baffa, et al.
Proepithelin Regulates Prostate Cancer Cell Biology by Promoting Cell Growth, Migration, and Anchorage-Independent Growth
Am. J. Pathol.,
March 1, 2009;
174(3):
1037 - 1047.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
K. D. Sorensen, P. J. Wild, A. Mortezavi, K. Adolf, N. Torring, S. Heeboll, B. P. Ulhoi, P. Ottosen, T. Sulser, T. Hermanns, et al.
Genetic and Epigenetic SLC18A2 Silencing in Prostate Cancer Is an Independent Adverse Predictor of Biochemical Recurrence after Radical Prostatectomy
Clin. Cancer Res.,
February 15, 2009;
15(4):
1400 - 1410.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. A. Koziol, A. C. Feng, Z. Jia, Y. Wang, S. Goodison, M. McClelland, and D. Mercola
The wisdom of the commons: ensemble tree classifiers for prostate cancer prognosis
Bioinformatics,
January 1, 2009;
25(1):
54 - 60.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Y. Zhang, D. Linn, Z. Liu, J. Melamed, F. Tavora, C. Y. Young, A. M. Burger, and A. W. Hamburger
EBP1, an ErbB3-binding protein, is decreased in prostate cancer and implicated in hormone resistance
Mol. Cancer Ther.,
October 1, 2008;
7(10):
3176 - 3186.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
E. N. Gal-Yam, G. Egger, L. Iniguez, H. Holster, S. Einarsson, X. Zhang, J. C. Lin, G. Liang, P. A. Jones, and A. Tanay
Frequent switching of Polycomb repressive marks and DNA hypermethylation in the PC3 prostate cancer cell line
PNAS,
September 2, 2008;
105(35):
12979 - 12984.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. Yan, H. Erdem, R. Li, Y. Cai, G. Ayala, M. Ittmann, L.-y. Yu-Lee, S. Y. Tsai, and M.-J. Tsai
Steroid Receptor Coactivator-3/AIB1 Promotes Cell Migration and Invasiveness through Focal Adhesion Turnover and Matrix Metalloproteinase Expression
Cancer Res.,
July 1, 2008;
68(13):
5460 - 5468.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
L. G. Wang, E. M. Johnson, Y. Kinoshita, J. S. Babb, M. T. Buckley, L. F. Liebes, J. Melamed, X.-M. Liu, R. Kurek, L. Ossowski, et al.
Androgen Receptor Overexpression in Prostate Cancer Linked to Pur{alpha} Loss from a Novel Repressor Complex
Cancer Res.,
April 15, 2008;
68(8):
2678 - 2688.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C. D. McCabe, D. D. Spyropoulos, D. Martin, and C. S. Moreno
Genome-Wide Analysis of the Homeobox C6 Transcriptional Network in Prostate Cancer
Cancer Res.,
March 15, 2008;
68(6):
1988 - 1996.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. Cao, C. Chiarelli, O. Richman, K. Zarrabi, P. Kozarekar, and S. Zucker
Membrane Type 1 Matrix Metalloproteinase Induces Epithelial-to-Mesenchymal Transition in Prostate Cancer
J. Biol. Chem.,
March 7, 2008;
283(10):
6232 - 6240.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
T M Murphy, A S Perry, and M Lawler
The emergence of DNA methylation as a key modulator of aberrant cell death in prostate cancer
Endocr. Relat. Cancer,
March 1, 2008;
15(1):
11 - 25.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
B. S. Taylor, M. Pal, J. Yu, B. Laxman, S. Kalyana-Sundaram, R. Zhao, A. Menon, J. T. Wei, A. I. Nesvizhskii, D. Ghosh, et al.
Humoral Response Profiling Reveals Pathways to Prostate Cancer Progression
Mol. Cell. Proteomics,
March 1, 2008;
7(3):
600 - 611.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. Yu, J. Yu, D. R. Rhodes, S. A. Tomlins, X. Cao, G. Chen, R. Mehra, X. Wang, D. Ghosh, R. B. Shah, et al.
A Polycomb Repression Signature in Metastatic Prostate Cancer Predicts Cancer Outcome
Cancer Res.,
November 15, 2007;
67(22):
10657 - 10663.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. Villar, M. I. Arenas, C. M. MacCarthy, M. J. Blanquez, O. M. Tirado, and V. Notario
PCPH/ENTPD5 Expression Enhances the Invasiveness of Human Prostate Cancer Cells by a Protein Kinase C{delta} Dependent Mechanism
Cancer Res.,
November 15, 2007;
67(22):
10859 - 10868.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Q. Xu, P. K. Majumder, K. Ross, Y. Shim, T. R. Golub, M. Loda, and W. R. Sellers
Identification of prostate cancer modifier pathways using parental strain expression mapping
PNAS,
November 6, 2007;
104(45):
17771 - 17776.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Y. P. Yu, G. Yu, G. Tseng, K. Cieply, J. Nelson, M. Defrances, R. Zarnegar, G. Michalopoulos, and J.-H. Luo
Glutathione Peroxidase 3, Deleted or Methylated in Prostate Cancer, Suppresses Prostate Cancer Growth and Metastasis
Cancer Res.,
September 1, 2007;
67(17):
8043 - 8050.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. Y. Wong, H. Haack, J. L. Kissil, M. Barry, R. T. Bronson, S. S. Shen, C. A. Whittaker, D. Crowley, and R. O. Hynes
Protein 4.1B suppresses prostate cancer progression and metastasis
PNAS,
July 31, 2007;
104(31):
12784 - 12789.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
D. Huang, G. P. Casale, J. Tian, N. K. Wehbi, N. A. Abrahams, Z. Kaleem, L. M. Smith, S. L. Johansson, J. E. Elkahwaji, and G. P. Hemstreet III
Quantitative Fluorescence Imaging Analysis for Cancer Biomarker Discovery: Application to {beta}-Catenin in Archived Prostate Specimens
Cancer Epidemiol. Biomarkers Prev.,
July 1, 2007;
16(7):
1371 - 1381.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. C. Smith, G. Oxford, A. S. Baras, C. Owens, D. Havaleshko, D. L. Brautigan, M. K. Safo, and D. Theodorescu
Expression of Ral GTPases, Their Effectors, and Activators in Human Bladder Cancer
Clin. Cancer Res.,
July 1, 2007;
13(13):
3803 - 3813.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
B. Ren, Y. P. Yu, G. C. Tseng, C. Wu, K. Chen, U. N. Rao, J. Nelson, G. K. Michalopoulos, and J.-H. Luo
Analysis of Integrin {alpha}7 Mutations in Prostate Cancer, Liver Cancer, Glioblastoma Multiforme, and Leiomyosarcoma
J Natl Cancer Inst,
June 6, 2007;
99(11):
868 - 880.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
I. A. Mawji, C. D. Simpson, R. Hurren, M. Gronda, M. A. Williams, J. Filmus, J. Jonkman, R. S. Da Costa, B. C. Wilson, M. P. Thomas, et al.
Critical Role for Fas-Associated Death Domain-Like Interleukin-1-Converting Enzyme-Like Inhibitory Protein in Anoikis Resistance and Distant Tumor Formation
J Natl Cancer Inst,
May 16, 2007;
99(10):
811 - 822.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
R. S. Turley, E. C. Finger, N. Hempel, T. How, T. A. Fields, and G. C. Blobe
The Type III Transforming Growth Factor-{beta} Receptor as a Novel Tumor Suppressor Gene in Prostate Cancer
Cancer Res.,
February 1, 2007;
67(3):
1090 - 1098.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Y. Li, J. Yang, J.-H. Luo, S. Dedhar, and Y. Liu
Tubular Epithelial Cell Dedifferentiation Is Driven by the Helix-Loop-Helix Transcriptional Inhibitor Id1
J. Am. Soc. Nephrol.,
February 1, 2007;
18(2):
449 - 460.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
R. J. Molinaro, B. K. Jha, K. Malathi, S. Varambally, A. M. Chinnaiyan, and R. H. Silverman
Selection and cloning of poly(rC)-binding protein 2 and Raf kinase inhibitor protein RNA activators of 2',5'-oligoadenylate synthetase from prostate cancer cells
Nucleic Acids Res.,
December 2, 2006;
34(22):
6684 - 6695.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. Chiosea, E. Jelezcova, U. Chandran, M. Acquafondata, T. McHale, R. W. Sobol, and R. Dhir
Up-Regulation of Dicer, a Component of the MicroRNA Machinery, in Prostate Adenocarcinoma
Am. J. Pathol.,
November 1, 2006;
169(5):
1812 - 1820.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
R. L. Parr, G. D. Dakubo, K. A. Crandall, J. Maki, B. Reguly, A. Aguirre, R. Wittock, K. Robinson, J. S. Alexander, M. A. Birch-Machin, et al.
Somatic Mitochondrial DNA Mutations in Prostate Cancer and Normal Appearing Adjacent Glands in Comparison to Age-Matched Prostate Samples without Malignant Histology
J. Mol. Diagn.,
July 1, 2006;
8(3):
312 - 319.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
B. S. Taylor, S. Varambally, and A. M. Chinnaiyan
A Systems Approach to Model Metastatic Progression
Cancer Res.,
June 1, 2006;
66(11):
5537 - 5539.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
D. K. Vanaja, K. V. Ballman, B. W. Morlan, J. C. Cheville, R. M. Neumann, M. M. Lieber, D. J. Tindall, and C. Y.F. Young
PDLIM4 Repression by Hypermethylation as a Potential Biomarker for Prostate Cancer
Clin. Cancer Res.,
February 15, 2006;
12(4):
1128 - 1136.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. Nanni, C. Priolo, A. Grasselli, M. D'Eletto, R. Merola, F. Moretti, M. Gallucci, P. De Carli, S. Sentinelli, A. M. Cianciulli, et al.
Epithelial-Restricted Gene Profile of Primary Cultures from Human Prostate Tumors: A Molecular Approach to Predict Clinical Behavior of Prostate Cancer
Mol. Cancer Res.,
February 1, 2006;
4(2):
79 - 92.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
G. Yu, G. C. Tseng, Y. P. Yu, T. Gavel, J. Nelson, A. Wells, G. Michalopoulos, D. Kokkinakis, and J.-H. Luo
CSR1 Suppresses Tumor Growth and Metastasis of Prostate Cancer
Am. J. Pathol.,
February 1, 2006;
168(2):
597 - 607.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
D. C. Malins, N. K. Gilman, V. M. Green, T. M. Wheeler, E. A. Barker, and K. M. Anderson
A cancer DNA phenotype in healthy prostates, conserved in tumors and adjacent normal cells, implies a relationship to carcinogenesis
PNAS,
December 27, 2005;
102(52):
19093 - 19096.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. A. Tomlins, D. R. Rhodes, S. Perner, S. M. Dhanasekaran, R. Mehra, X.-W. Sun, S. Varambally, X. Cao, J. Tchinda, R. Kuefer, et al.
Recurrent Fusion of TMPRSS2 and ETS Transcription Factor Genes in Prostate Cancer
Science,
October 28, 2005;
310(5748):
644 - 648.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H. Kim, J. Lapointe, G. Kaygusuz, D. E. Ong, C. Li, M. van de Rijn, J. D. Brooks, and J. R. Pollack
The Retinoic Acid Synthesis Gene ALDH1a2 Is a Candidate Tumor Suppressor in Prostate Cancer
Cancer Res.,
September 15, 2005;
65(18):
8118 - 8124.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
I. U. Agoulnik, A. Vaid, W. E. Bingman III, H. Erdeme, A. Frolov, C. L. Smith, G. Ayala, M. M. Ittmann, and N. L. Weigel
Role of SRC-1 in the Promotion of Prostate Cancer Cell Growth and Tumor Progression
Cancer Res.,
September 1, 2005;
65(17):
7959 - 7967.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Y. P. Yu, S. Paranjpe, J. Nelson, S. Finkelstein, B. Ren, D. Kokkinakis, G. Michalopoulos, and J.-H. Luo
High throughput screening of methylation status of genes in prostate cancer using an oligonucleotide methylation array
Carcinogenesis,
February 1, 2005;
26(2):
471 - 479.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
TOP
ABSTRACT
INTRODUCTION
