Skip to main content
PMC home page
Endocrinology logoLink to Endocrinology
. 2009 Jun 11;150(9):3991–4002. doi: 10.1210/en.2009-0573

Epigenetic Alterations in Human Prostate Cancers

William G Nelson 1, Angelo M De Marzo 1, Srinivasan Yegnasubramanian 1
PMCID: PMC2736081  PMID: 19520778

Abstract

Human prostate cancer cells carry a myriad of genome defects, including both genetic and epigenetic alterations. These changes, which can be maintained through mitosis, generate malignant phenotypes capable of selective growth, survival, invasion, and metastasis. During prostatic carcinogenesis, epigenetic changes arise earlier than genetic defects, linking the appearance of epigenetic alterations in some way to disease etiology. The most common genetic defect thus far described, leading to fusion transcripts between the androgen-regulated gene TMPRSS2 and genes from the ETS family of transcription factors, likely endows prostate cancer cells with the ability to co-opt androgen signaling, the major prostate differentiation pathway, to support the malignant phenotype. Whether epigenetic changes promote the appearance of TMPRSS2-ETS family fusion transcripts or collaborate with fusion transcript expression in the pathogenesis of prostate cancer has not been established. However, a growing list of epigenetic alterations has provided new opportunities for clinical tests that might aid in prostate cancer screening, detection, diagnosis, staging, and risk stratification. The epigenetic changes appear to be more attractive than genetic changes as prostate cancer biomarkers because epigenetic alterations are present in a greater fraction of prostate cancer cases than any of the known genetic defects. In addition, an emerging generation of assay strategies for detection of specific DNA sequences carrying 5-meC, the major epigenetic genome mark, has pushed somatic epigenetic alterations to the forefront of molecular biomarker assay development for cancer. Finally, a growing portfolio of epigenetic drugs, capable of reversing the phenotypic consequences of somatic epigenetic defects, has entered clinical trials for prostate cancer in the search for a new rational therapy for the disease.

Somatic epigenetic defects, which appear earlier and more consistently during prostate cancer development than genetic defects, have provided new insights into mechanisms of prostatic carcinogenesis and a rich collection of molecular biomarkers for the disease.

Somatic epigenetic alterations are an increasingly recognized phenomenon in the development of human cancers, including prostate cancer (1,2,3). These genome changes, which affect chromatin structure and function, can be maintained through mitosis, providing a selective advantage for growth and survival during cancer pathogenesis. Epigenetic defects in cancer cells reported thus far include 1) transcriptional silencing of caretaker genes and tumor suppressor genes, 2) reactivation of embryonic genes, 3) loss of imprinted gene partitioning into active and inactive alleles, 4) redirection of transcription promoter use, 5) disordered micro-RNA gene expression, 6) activation of retrotransposition, and 7) increased genetic recombination at repeat elements. For human prostate cancer, epigenetic changes arise at the earliest recognizable steps of transformation and persist through invasion, metastasis, and life-threatening malignant progression (4,5). Growing knowledge of the breadth of epigenetic alterations in prostate cancer has created new opportunities for the discovery of biomarkers useful for prostate cancer screening, detection, diagnosis, staging, and risk stratification. In addition, because epigenetic processes do not directly change genome sequence, therapeutic modulation of the epigenome, using drugs that affect dynamic chromatin function, has great promise as a rational approach to prostate cancer prevention and treatment.

Epigenome states, which are established in normal cells as part of development, are maintained through somatic cell mitosis as a consequence of marks present in genomic DNA and in chromatin proteins. The major DNA mark is symmetric methylation of cytosine bases in the self-complementary nucleotide sequence CpG. Most CpG dinucleotides carry this mark; unmethylated CpG dinucleotides, clustered into approximately 1-kb regions encompassing the transcription start sites of many genes, have been termed CpG islands (6). Somatic changes in DNA methylation at CpG islands affect gene activity. Genes with unmethylated CpG islands are competent for regulated transcription, responding to signaling cues by recruiting trans-activating factors that modify nucleosome and chromatin structure, using histone acetyltransferases and histone methyltransferases to promote transcript synthesis by RNA polymerases. In contrast, genes with methylated CpG islands tend to be incompetent for expression, tightly wound around nucleosomes in a repressive chromatin structure maintained by histone deacetylases (HDACs) and other enzymes. Thus, somatic increases in CpG island methylation in cancer cells have been associated with gene silencing and heterochromatinization, although somatic decreases in CpG island methylation have been implicated in illegitimate gene activation, retrotransposition, and recombination.

The CpG dinucleotide methyl marks are established and maintained by DNA methyltransferases (DNMTs), which catalyze the transfer of a methyl group from S-adenosyl-methionine to cytosine bases in CpG dinucleotides. For this reason, DNMTs have been suspected to contribute in some way to cancer development. Three DNMTs have been described: DNMT1, DNMT3a, and DNMT3b. DNMT1 acts on hemi-methylated DNA substrates generated during DNA synthesis to maintain CpG methylation patterns through genome replication and mitosis and through genome damage and repair (7,8). DNMT3a and DNMT3b can create de novo CpG dinucleotide methylation marks during embryonic development (9,10). Of all of the DNMTs, DNMT1 has attracted the most attention for possible roles in cancer pathogenesis in model systems. Forced DNMT1 expression in normal cells triggers DNA hypermethylation and gene silencing (11,12,13). By causing or maintaining increased DNA methylation, the enzyme has been found to be necessary for c-fos transformation of rodent fibroblasts, for intestinal polyp formation in ApcMin/+ mice, and for lung cancer development in mice exposed to tobacco carcinogens (14,15,16,17). In the mouse prostate, forced expression of SV40 T antigen under control of a rat Probasin promoter (the transgenic adenocarcinoma of the mouse prostate, or TRAMP model), also produces early hypermethylation of specific genes accompanying tumorigenesis, with emerging epigenomic heterogeneity upon cancer progression (18,19,20). These somatic epigenetic changes were likely critical to prostatic carcinogenesis in TRAMP mice, because treatment with a DNMT inhibitor completely prevented tumor development by 24 wk (21). Inadequate DNMT1 function, which leads to hypomethylation, can also contribute to cancer development. Mice with a defective Dnmt1 allele resulting in 10% of normal enzyme activity show increased genetic instability and a tendency toward developing T-cell lymphomas (22,23). These findings in model systems recapitulate many phenomena seen in human cancers, although in most advanced life-threatening human cancers, including prostate cancer, the cancer cell genomes simultaneously display both hypermethylation and hypomethylation (4,5).

It is tempting to attribute the evolution of both hypermethylation and hypomethylation in human cancer to some sort of dysfunction in DNMT regulation or action, in which inappropriate de novo methylation activity is accompanied by poor fidelity of methylation pattern maintenance during genome replication. In support of such a mechanism, increased expression of DNMT1 has been found in many human cancer cells; most often, increased DNMT1 levels are a consequence of defective enzyme degradation after completion of the S-phase of the cell cycle, a process normally mediated by ubiquitin conjugation and proteasome destruction managed by the anaphase-promoting complex (24,25). In addition, immunohistochemical staining for DNMT1 in advanced cancers has revealed striking heterogeneity in enzyme content among cancer cells, with many of the cells in a single tumor deposit appearing to contain excess enzyme and others seemingly devoid of enzyme polypeptides (26). Finally, posttranslational modifications of DNMTs may affect enzyme function. Treatment of mammalian cells with inflammatory cytokines promotes DNA methylation-associated silencing of FMR1 and HPRT, a process attenuated by inhibitors of nitric oxide synthases (27). S-Nitrosation of critical DNMT1 cysteines by nitric oxide is likely what is responsible for the augmented enzyme activity. In contrast, methylation of K1096 in Dnmt1, catalyzed by Set7/9 and antagonized by Lsd1, reduces DNA methylation (28).

Chromatin protein marks, which collaborate with DNA methylation marks in managing gene activity, are also altered in cancer cells. As an example, the polycomb complex of proteins usually functions during development to establish chromatin protein marks that permit directed gene expression. One of the polycomb complex components, enhancer of zeste homolog 2 (EZH2), a histone H3-K27 methyltransferase, is overexpressed in metastatic prostate cancers (29). The polycomb complex containing EZH2 likely acts to repress genes upon recruitment to specific genome sites by creating histone H3-K27 marks, which promote heterochromatinization (30). In an example of this mechanism, DAB2IP, encoding a GTPase-activating protein that can affect Ras signaling and TNF-associated apoptosis, appears to be epigenetically silenced by an EZH2-dependent mechanism in prostate cancer cells (31). In cell culture studies, EZH2 overexpression decreases DAB2IP levels in normal prostate cells, whereas EZH2 knockdown increases DAB2IP production (32). In the cells with silenced DAB2IP promoters, the transcriptional regulatory region carried histone H3-K27 marks, consistent with EZH2 action (32). Thus, just as inappropriate DNMT expression or activity can trigger repression of specific genes, dysregulation of EZH2 function also can drive epigenetic gene silencing. The interplay between targeted repression complexes, capable of placing heterochromatin marks, and DNMTs, capable of establishing and maintaining DNA methylation marks, may be at the heart of somatic epigenetic silencing of cancer genes. Genome regions carrying polycomb-associated chromatin marks appear especially vulnerable to de novo DNA methylation (33,34).

Genes carrying densely methylated transcriptional regulatory sequences are maintained in a transcriptionally inactive state by 5-mCpG-binding domain (MBD) family proteins (35). 5-mCpGs clustered at gene promoters marks the genes for heterochromatinization; MBD family proteins transduce these marks into signals for transcription repression. One of the MBD family members, MeCP2, not only contains a 5-mCpG-binding motif but also carries a transcriptional repression domain that mediates interactions with Sin3 and HDACs (36). When bound to promoter sequences, MeCP2 can thus silence critical genes in a manner dependent on HDAC activity, a phenomenon potentially antagonized by HDAC inhibitors such as trichostatin A and others (36). Another MBD family protein, MBD2, which also binds selectively to DNA containing 5-mCpG, recruits a 1-megadalton transcription repression and chromatin remodeling complex, MeCP1, which contains MBD3, HDAC1 and HDAC2, histone-binding proteins RbAp46 and RbAp48, the SWI/SNF helicase/ATPase domain-containing protein Mi-2, MTA2, and other components (37). When associated with methylated promoter sequences, MBD2, like MeCP2, can trigger gene silencing. However, unlike MeCP2, MBD2-mediated transcriptional repression is often not antagonized by HDAC inhibitors (38). MBD2 function is critically important for epigenetic gene silencing in cancer cells and for cancer development. Treatment of human cancer cells with small interfering RNA targeting MBD2 mRNA triggers reactivation of epigenetically silenced genes (38). In addition, when mice carrying defective Apc genes, prone to intestinal tumorigenesis, are crossed to an Mbd2−/− background, the mice grow far fewer intestinal tumors than mice capable of Mbd2 expression (39).

Changes in DNA Methylation Marks in Prostate Cancer

Changes in DNA methylation marks, accompanied by epigenetic gene silencing, appear to be the earliest somatic genome changes yet recognized in human prostate cancer (Fig. 1). The most studied gene affected by de novo methylation during prostatic carcinogenesis is GSTP1, which encodes an enzyme responsible for detoxifying electrophiles and oxidants, including those that threaten cell and genome damage (40). Hypermethylation of GSTP1 transcriptional regulatory sequences has been consistently detected in more than 90% of prostate cancers in more than 50 independent analyses, far more frequently than any other known somatic gene defect (3). Loss of glutathione S-transferase-π (GSTP1) function likely occurs at the initiation of prostatic carcinogenesis, with GSTP1 methylation evident in some 5–10% of proliferative inflammatory atrophy (PIA) lesions, the earliest prostate cancer precursors, and in more than 70% of prostatic intraepithelial neoplasia lesions (41,42). Human prostate cancer cells devoid of GSTP1 better activate heterocyclic amine carcinogens, such as those found in overcooked meats, to mutagenic species than cells capable of GSTP1 expression (43). Perhaps, GSTP1 silencing early during the pathogenesis of prostate cancer, with the resultant loss of enzymatic protection against reactive chemical species, may offer an explanation for the well-known sensitivity of human prostatic carcinogenesis to dietary and lifestyle habits (43,44,45). In support of this notion, mice carrying disrupted Gstp1/2 genes are more prone to develop skin tumors upon exposure to a topical carcinogen than wild-type mice (46). Alternatively, loss of GSTP1 may afford some other selective advantage: mice with defective Gstp1/2 genes show accelerated prostatic carcinogenesis in response to Pb-c-Myc transgenes (unpublished data).

Figure 1.

Figure 1

Open in a new tab

Cells with abnormal DNA methylation marks arise in PIA lesions as a consequence of an epigenomic catastrophe and accumulate selectively in prostatic intraepithelial neoplasia and prostate cancer lesions.

As of yet, the mechanism by which GSTP1 is subjected to de novo methylation is not known. However, the appearance of GSTP1 hypermethylation in PIA lesions hints that some sort of chronic or recurrent inflammatory process may play a critical role in epigenetic gene silencing enforced by DNA methylation marks. Of interest, most epithelial cells in PIA lesions express very high levels of GSTP1, likely in response to induction of GSTP1 transcription, triggered by stress signaling pathways (47,48). Rare epithelial cells, in only a few PIA lesions, tend to show loss of GSTP1 expression. Could the GSTP1 transcriptional regulatory region be more vulnerable to heterochromatinization when in an induced state? The appearance of epigenetically silenced genes in epithelial cells in response to inflammation, such as is seen in PIA, may be a recurrent theme in epithelial carcinogenesis generally, with abnormally methylated genes thus far described in inflammatory bowel disease, hepatitis, gastritis, and other inflammatory conditions associated with cancer development (49). The manner by which the epigenome of epithelial cells becomes corrupted in response to inflammatory stresses constitutes what may be the major unresolved problem in epithelial carcinogenesis. Provocatively, although the mechanism by which an inflammatory microenvironment might promote epigenetic gene silencing in epithelial cells has not been established, inflammatory cytokines such as IL-1β have been found to trigger gene silencing in certain cells by promoting nitric oxide generation (27). In another study using mice genetically prone to intestinal inflammation and carcinogenesis, genome-wide assessment of DNA methylation in inflamed ileum epithelium revealed new DNA methylation marks in some 250 genes, with 70% of the genes representing targets of polycomb complex-mediated repression (50).

In addition to GSTP1, more than 40 genes have been reported to be targets of epigenetic gene silencing in prostate cancers (3). Early insights from genome-wide assessments of abnormal DNA methylation marks in prostate cancer cells suggest that the number of genes affected may number in the hundreds to thousands (unpublished data). Nonetheless, the DNA methylation marks appear to be established in at least two waves. The first changes, which arise in response to some sort of methylation catastrophe in prostate cancer precursors, may be what are needed for transformation; a second set of changes may promote malignant cancer progression. Yegnasubramanian et al. (4) reported that although hypermethylation of at GSTP1, APC, RASSF1a, COX2, and MDR1 could be detected both in localized and in metastatic prostate cancers, hypermethylation at ERα, hMLH1, and p14/INK4a tended to appear only later during disease progression. Of note, most de novo methylation changes in prostate cancer arise before clonal evolution of life-threatening disease. In men dying of prostate cancer subjected to autopsy, DNA hypermethylation profiles were 5-fold more variable case to case than metastatic site to metastatic site (P < 0.0001) (4).

Loss of DNA methylation marks, although not studied as intensively as gain of DNA methylation marks, also occurs during prostatic carcinogenesis (Fig. 2). Analyses of 5-meC content in genomic DNA have revealed undermethylation mostly in association with disease progression, with low 5-meC levels more evident in metastatic prostate cancer cases (51). Decreased LINE-1 repeat methylation, found in as many as 53% of prostate cancer cases, was detected in 67% of cases with lymph node metastases but in only 8% of cases without lymph node metastases (52). In animal models, reduced DNA methylation has been accompanied by genetic instability (22,23). Although increased recombination frequency might be expected in the setting of reduced LINE-1 methylation, whether undermethylation of repeat sequences contributes to copy number gains and losses in prostate cancer has not been established. Nevertheless, a correlation between DNA hypomethylation and losses or gains of sequences on chromosome 8 has been described in localized prostate cancer cases (22,53). In addition, lethal prostate cancers recovered at autopsy show both strikingly high levels of recombinations, with associated copy number alterations, and strikingly low levels of DNA methylation (5). The progressive genome undermethylation appears most easily attributed to lack of DNA methylation maintenance fidelity during DNA replication. In men with lethal prostate cancer who underwent autopsy, although DNA hypermethylation profiles tended to be very similar in different metastatic sites in individual cases, hypomethylation profiles were highly variable (P < 0.0001) (5). In fact, hypomethylation, and associated reactivation of epigenetically repressed developmental genes, was highly variable cell to cell in individual metastatic deposits (5).

Figure 2.

Figure 2

Open in a new tab

Loss of DNA methylation marks worsens during prostate cancer progression, probably as a result of poor DNA methylation maintenance fidelity, and contributes to marked cell-to-cell and lesion-to-lesion phenotypic heterogeneity.

Detection of DNA Methylation Marks for Prostate Cancer Screening, Early Detection, Diagnosis, Staging, Risk Stratification, and Treatment Monitoring

Prostate cancer screening, using serum testing for prostate-specific antigen (PSA) along with digital rectal examination has become commonplace in the developed world, with most men over 50 yr of age in the United States having been screened. PSA, a protein secreted by prostatic epithelial cells to become a component of the ejaculate, appears in the circulation only as a result of epithelial damage or dysfunction in the prostate. As such, although PSA testing is frequently used as a screening tool, in this setting, serum PSA levels are not biomarkers of prostate cancer per se. Rather, the serum PSA can be elevated in a variety of noncancerous prostate diseases and conditions, including prostate infections, symptomatic and asymptomatic nonbacterial prostatitis, and benign prostatic hyperplasia/hypertrophy (BPH). Because inflammatory conditions and/or BPH affect more than half of men in the prostate cancer screening age group, the significance of an isolated elevation in serum PSA is often not clear. In one study, only 22% of men with elevated serum PSA values from 4.0–9.9 ng/ml were found to have prostate cancer upon prostate biopsy (54). Also, a normal serum PSA may not be very comforting either. In the Prostate Cancer Prevention Trial, 24.4% of men on the placebo treatment arm who entered the study with normal serum PSA values and underwent prostate biopsies at the end of the trial were found to have prostate cancer (55,56). Prostate cancers, particularly life-threatening prostate cancers, tend to grow over time. For this reason, perhaps, serial serum PSA testing may be more informative than isolated PSA tests in discriminating prostate cancer from benign prostate conditions (57). A PSA velocity of 0.35 ng/ml increase per year evident some 10–15 yr before prostate biopsy may have a higher predictive value for the presence of life-threatening prostate cancer than any single elevated serum PSA level (58). Even with its imperfections, serum PSA testing likely saves lives (59,60). Certainly, the widespread use of serum PSA assays has changed the natural history of prostate cancer. The number and fraction of men presenting with symptomatic metastastic prostate cancer has declined significantly, with a commensurate increase in the number and fraction of men with localized disease (61). Nonetheless, new tests, with better positive and negative predictive values, could greatly improve prostate cancer screening, directing prostate biopsy procedures toward the men most likely to have prostate cancer.

If screening for prostate cancer is in need of refinement, the current approach to detection and diagnosis of prostate cancer leaves even more to be desired. The most commonly employed biopsy strategy features ultrasound-guided core needle sampling of less than 0.3% of prostate tissues. The major unmet need is the lack of definitive prostate imaging that can discriminate likely carcinoma from focal inflammation, atrophy, scarring, etc. Because the biopsies are not directed at suspicious lesions, arguments continue to rage over the pattern of sampling, the zonal regions of the prostate to be sampled, the number of core samples to be taken, and when or whether repeat biopsy procedures should be pursued if prostate cancer is not detected (62,63). Of interest, serum testing for bound vs. free serum PSA was approved as a tool for reducing unnecessary prostate biopsies in men with an elevated serum PSA (64). A man with an abnormally low free PSA (indicative perhaps of BPH) has a 3.2-fold increased chance of having prostate cancer (with a 95% confidence interval of 2.5–4.1 and P < 0.001). Novel biomarkers, such as urine testing for PCA3 RNA and for TMPRSS2-associated fusion transcripts, are also under development for this indication. Another appetite for prostate-cancer-specific biomarkers has arisen as surgical pathologists have been confronted with millions of prostate biopsy cores to inspect. Frequently, cancerous prostate glands are sampled obliquely during core biopsy procedures, yielding microscopic lesions that are suspicious for, but not diagnostic of, prostate cancer. To aid in the accurate assessment of such abnormalities, a number of biomarker tools, including immunohistochemical staining for p63 or for cytokeratins (expressed by basal epithelial cells that are absent in carcinomas), and staining for α-methylacyl-coenzyme A-racemase (AMACR; selectively expressed by prostatic carcinoma cells) have been developed and readily adopted (65,66).

The greatest biomarker need in prostate cancer medicine may be for risk stratification. In the United States and in most of the developed world, prostate cancers may be present in more than half of all men over age 50 yr, yet few appear to threaten morbidity and/or mortality. For this reason, the entire rationale for prostate cancer screening, early detection, and early treatment has been questioned (67,68). By one estimate, as many as 1410 men may need to be screened to detect one prostate cancer, and an additional 48 men may need to be treated with definitive local therapy (surgery or radiation therapy) to save one life (60,68). The best single risk stratification tool described so far, Gleason grading of the microscopic appearance of prostate cancer lesions, has several limitations. Gleason grades (1,2,3,4,5) are assigned the most common histological patterns exhibited by the cancers; the two most prevalent grades are summed to create a Gleason score (e.g. Gleason 3 + 4 = 7). When used by expert prostate pathologists in examination of prostate tissue removed at radical prostatectomy, Gleason grading is consistent and highly correlated with outcome; in a recent study of more than 2500 men who underwent resection of Gleason 6 or less prostate cancer, none were found to have died of the disease (69). Despite this astonishing performance, two problems have limited the full utility of Gleason grading for prostate cancer risk stratification. First, there remains some variability in the capability of surgical pathologists distributed throughout the United States and the world to accurately assign Gleason grades. Second, when used for prostate biopsy cores, a Gleason score of 6 or better, which might otherwise define a prostate cancer case best treated with expectant management/watchful waiting, may not carry the same predictive value as when used for surgical resection specimens. There remains the chance that a region of the prostate that was not sampled by core biopsy might contain Gleason score 7 or higher cancer. For this reason, the most urgent need for a new prostate cancer biomarker may be for one that can sample the entire prostate (perhaps via testing of blood, urine, or expressed prostate secretions) and can provide prognostic information as good as or better than Gleason grading.

Changes in DNA methylation, the major DNA marks of somatic epigenome alterations, constitute a superb source of cancer biomarkers for several reasons. For one, such changes can be detected using PCR methods at near single-copy sensitivity, and small DNA fragments may be more stable in blood and body fluids than RNA or protein species. Next, acquired DNA methylation differences have been reported for nearly every human cancer. Furthermore, for somatic hypermethylation of CpG island sequence, such changes are often more consistently present case to case for a given cancer than genetic changes. For example, although fusion transcripts containing sequences from androgen-regulated genes such as TMPRSS2, SLC45A3, HERV-K, CANT1, and KLK2 have been found cumulatively in as many as 60–70% or more of prostate cancer cases, the complexity of genome lesions responsible for generating such transcripts appears considerable (70). As a result, it is likely that dozens or more assays may need to be designed to detect such changes with significant sensitivity. In contrast, GSTP1 hypermethylation, present in more than 90% of prostate cancer cases, can be detected at astonishing sensitivity using a single assay (4,71,72,73,74). Finally, for a given prostate cancer, different patterns of DNA methylation marks appear likely to discriminate aggressive vs. nonaggressive disease and to predict responsiveness to specific treatments.

Historically, there have been two general strategies for detection of DNA methylation changes at specific DNA sequences in cancer cells. The older approach exploits the use of methylation-sensitive restriction endonucleases, which cut recognition sites differently depending on whether 5-meCpG is present. The digested DNA can be subjected to Southern blot analysis, with cut fragments migrating farther during agarose gel electrophoresis than intact fragments. To increase assay sensitivity, the differential cutting by methylation-sensitive restriction enzymes has been adapted for use with PCR. When added before PCR amplification, the enzymes ideally cut unmethylated target sequences but leave methylated sequences intact. As a consequence, the appearance of a product during PCR is interpreted as evidence for methylation of the target sequence. Unfortunately, although quite sensitive, this PCR approach tends to be bedeviled by false-positive product amplification, attributable to inefficient suppression of amplification from unmethylated sequences (74,75). The second major strategy for selective detection of genome sequences carrying methylation marks features modification of DNA using sodium bisulfite, a procedure that promotes C deamination to U but spares 5-meC. After PCR, U appears as T, whereas 5-meC appears as C, creating a DNA sequence difference out of a DNA methylation difference. The bisulfite modification procedure has been used in two modes: for sequencing of 5-meC in specific genome regions and for sensitive detection of sequences containing 5-meC vs. C using specific primers, a procedure referred to as methylation-specific PCR (76,77). The performance of bisulfite modification assays can be limited both by undertreatment, with insufficient conversion of C to U, and by overtreatment, with damage to target DNA templates that undermines sensitivity (3). The major limitation of bisulfite modification assays for the future is that they are poorly suited for genome-wide analyses. By converting C to T throughout the genome, the coding complexity is reduced, restricting the dimensionality of methylation-specific PCR assays to a handful of genes and rendering useless next-generation sequencing approaches that depend on assignment of short DNA sequence reads to specific genome locations. The genome-wide assays using bisulfite modification employed thus far have focused on only a small fraction of the genome, featuring hybridization of amplified bisulfite-treated DNA to CpG island microarrays (78).

A new generation of 5-meCpG capture approaches appears poised to replace these previous technologies. Although various 5-meC-binding proteins and anti-5-meC antibodies have been used to selectively capture methylated DNA sequences for various analyses, the DNA-binding fragment of MBD2 [MDB2 DNA-binding domain (DBD)] is remarkably well suited for this task. MBD2-DBD capture has formed the basis for the construction of assays for the detection of DNA methylation changes at individual genes, for collections of genes, or throughout the entire genome (74). The MBD2-DBD quantitatively captures double-stranded DNA containing symmetrically methylated CpG dinucleotides over a concentration range of nearly five orders of magnitude while not binding hemi-methylated or unmethylated DNA at all. With this high degree of sensitivity and specificity for capturing methylated DNA, the captured DNA can then be analyzed in variety of ways, including by using PCR or multiplex PCR (the COMPARE-MS assay) or by using almost any whole genome assay platform, including whole genome tiled arrays, high-density single-nucleotide polymorphism arrays, or high-coverage next-generation sequencing.

The most substantial epigenomic biomarker development activity thus far has been for the introduction of clinical tests for somatic methylation marks at GSTP1, which are nearly universally present in prostate cancer cell genomes but are absent from normal cell DNA. Since the initial report of this epigenome change in 1994, well over a thousand prostate cancer cases have been assessed using a variety of detection strategies (3). A PubMed query using search terms GSTP1, methylation, and prostate yields more than 120 published manuscripts. The major commercial use for the assays that can detect GSTP1 methylation marks has been as an adjunct to diagnostic histopathology, with tests available for DNA isolated from formalin-fixed paraffin-embedded biopsy cores. In this setting, the biomarker serves as a surrogate for prostate cancer cell genomes, increasing the accuracy of diagnosis for suspicious lesions and permitting analysis of greater amounts of tissues than can easily be inspected through the microscope (in principle, the whole biopsy core can be assessed without cutting large numbers of tissue sections). The commercial tests introduced so far have used bisulfite modification technology; however, the sensitivity of assays for GSTP1 methylation marks vary markedly depending on the detection strategy used and the specific region of the GSTP1 CpG island targeted (3). The new-generation capture assays, such as COMPARE-MS, will offer improved performance when adapted for clinical use. For example, the COMPARE-MS assay for GSTP1 methylation marks in prostate cancer shows 99.2% sensitivity and 100% specificity for cancer vs. normal prostate tissue (74). The major future questions for clinical tests of GSTP1 methylation marks as surrogates for prostate cancer DNA are whether such tests can be used for prostate cancer screening and early detection, using DNA from blood or urine, or for refinement of prostate cancer staging, using DNA from lymph nodes or bone marrow. Provocatively, GSTP1 methylation marks have been detected in each of these types of specimens (79,80,81,82). What is not yet clear is whether the predictive value of a negative test is sufficient to change the course of medical care. If a negative blood or urine test for GSTP1 methylation carried only a 50–75% negative predictive value for the presence of prostate cancer upon prostate biopsy, then the test would not likely reassure men that no further evaluation for prostate cancer was needed. As for improvement of prostate cancer staging, detection of GSTP1 methylation marks in serum DNA from men with localized prostate cancer was correlated with an increased risk of cancer recurrence after radical prostatectomy (82). Finally, in addition to GSTP1 methylation marks, a growing collection of hypermethylated gene sequences appear to function reasonably well as cancer DNA surrogates, including APC, RASSF1a, and MDR1 (4). As new such somatic methylation marks are discovered, new clinical tests for prostate cancer cell DNA may follow.

In addition to detection of prostate cancer DNA using tests for methylation marks, DNA methylation biomarkers for prostate cancer prognosis may constitute a fertile area for the discovery and development of new clinical assays. Somatic methylation marks at EDNRB, RARβ, RASSF1a, ERβ, and TIG1 may be suitable for this unmet need. Each has been correlated with known prognostic factors for primary prostate cancer such as tumor stage and/or Gleason grade (4,83,84,85,86). Testing for such methylation marks may even offer improvements over Gleason grading. In one study, PTGS2 methylation marks in the localized prostate cancer predicted prostate cancer recurrence after radical prostatectomy, independently of tumor stage and Gleason grade (4). Thus, using GSTP1 CpG island methylation assays for molecular staging, the detection of DNA with GSTP1 CpG island hypermethylation in the serum of men with localized prostate cancer was associated with an increased risk of prostate cancer recurrence after radical prostatectomy (82).

Epigenetic Gene Silencing as a Therapeutic Target for Prostate Cancer Prevention and Treatment

Of all the somatic genome changes that accumulate during the pathogenesis of human prostate cancers, only changes in DNA methylation appear to occur consistently (virtually all cases), to arise early (first appearing in preneoplastic lesions), and to be potentially reversible (the DNA sequence remains intact). Attempts to therapeutically reverse epigenetic gene silencing are at the cusp of clinical translation. One strategy, which has led to two new drugs approved within the last 2 yr by the U.S. Food and Drug Administration, features the use of inhibitors of DNMTs, such as azacitidine (Vidaza), decitabine (Dacogen), zebularine, procainamide, or hydralazine, to reduce 5-mCpG density at CpG island sequences in dividing cancer cells (87,88,89,90). Another approach, also under active clinical development, has been the use of inhibitors of HDACs, such as vorinostat (Zolinza), MS-275, valproic acid, and many others, to limit the formation of repressive chromatin conformation near the genes carrying abnormally methylation marks (91,92,93). And a third possibility, not yet in the clinic, involves selectively small-molecule targeting the MBD family proteins that mediate repression of transcription at genes with hypermethylated CpG islands (unpublished data).

Although each of the FDA-approved drugs, azacitidine (Vidaza), decitabine (Dacogen), and vorinostat (Zolinza), was found to provide a treatment benefit for a specific indication when used as a single agent [myelodysplasia for the DNMT inhibitors (87) and cutaneous T-cell lymphoma for the HDAC inhibitor (94)], combination treatments involving epigenetic drugs appear likely to be needed for most cancers, including prostate cancer. For prostate cancer, the results of a small phase 2 clinical trial (n = 14 men) of decitabine, given iv every 8 h at a dose of 75 mg/m2 every 5–8 wk, for androgen-independent metastatic prostate cancer has been reported (95). In the trial, two of 12 men who could be assessed for a treatment response exhibited stable disease, progressing after 10 wk or so (95). Thus, the single-agent activity of epigenetic drugs for prostate cancer, and for most other solid organ cancers, is likely to be limited. Certainly, combinations of DNMT and HDAC inhibitors have been found to more effectively reverse epigenetic gene silencing in cancer cells in vitro when used together vs. when used alone (96); such combinations appear to have intriguing anticancer activity in preclinical models and in early clinical studies in hematological malignancies and in lung cancer (17,97). Even more provocatively, there have been sporadic reports of DNMT and/or HDAC inhibitors triggering reactivation of epigenetically silenced genes that encode known drug targets, such as the estrogen receptor in breast cancer cells (98,99,100,101); components of the retinoic acid signaling pathway in prostate, breast, and kidney cancer cells and in melanoma cells (102,103,104,105,106); and cyclooxygenase 2 in many different types of cancer cells (4,107). In the case of the silenced ER gene in estrogen-independent breast cancer cells, induction of receptor expression in tamoxifen-resistant breast cancer cells by DMNT and HDAC inhibitors promotes sensitivity to growth suppression by tamoxifen (108). Similarly, for retinoic acid signaling in prostate cancer cells and other cancer cells, epigenetic drugs reactivate the signaling pathway and restore sensitivity to growth suppression by istoretinoin (102,103,104,105,106).

What is the future of epigenetic drugs used in combination for prostate cancer and other cancers? Transcriptome profiling studies of cancer cells, carrying disrupted DNMT genes and/or exposed to DNMT and HDAC inhibitors, have suggested that as many as 400 genes might be affected by somatic epigenetic inactivation in different cancer cells and thus might be candidates for reactivation by epigenetic drugs (109,110). As a result, the phenotype of cancer cells treated with epigenetic drugs is likely to be substantially different from the phenotype existing before treatment, exposing new and unexpected vulnerabilities and therapeutic opportunities (Fig. 3). For this reason, even if epigenetic drugs do not directly kill cancer cells or prevent cancer cell growth as single agents, the phenotype induced by drug treatment might create a new sensitivity to cytotoxicity or growth inhibition by a second drug targeting the product of a reactivated gene. Furthermore, the phenotype induced by epigenetic drugs in cancer cells will likely differ from normal cell phenotypes, perhaps providing a theoretical rationale for epigenetic drugs augmenting the therapeutic index of targeted drugs. This phenomenon, in which two drugs, one an epigenetic drug and the other targeting the product of an epigenetically silenced gene, are not deleterious to a cancer cell when given alone but antagonize cancer cell growth when given together, is reminiscent of genetic synthetic lethality, in which two gene deletions that are otherwise well tolerated alone give rise to a lethal phenotype when present together (111). Synthetic lethal screens, in which a query gene is deleted that does not impact viability and a systematic series of gene deletions are surveyed for loss of viability in combination with the deleted query gene, have been used to unmask and identify interactions of different gene products and of different cellular pathways (112). A systematic hunt for candidate drug combinations using a screen for agents exhibiting synthetic lethal-like activity against prostate cancer cells when administered with epigenetic drugs might yield a translational research pipeline for new prostate cancer treatment development.

Figure 3.

Figure 3

Open in a new tab

Epigenetic drugs drastically alter the phenotypes of cancer cells, perhaps rendering the cells sensitive to various targeted drugs that can be used in combination treatments.

Future Directions

Over the next few years, genome-wide approaches for assessing DNA methylation marks and chromatin protein marks will be systematically applied to prostate cancer cases, providing a myriad of new insights into prostate cancer cell phenotypes and countless new molecular biomarkers for prostate cancer. Already, the MBD2-DBD capture assays for methylated DNA are improving the accuracy and throughput of methylation mark detection in genomic DNA. In addition, new epigenetic drugs will be available for prostate cancer treatment development. Hopefully, rational strategies for discovering effective combination treatments using these drugs will permit timely testing of such combinations in clinical trials, ultimately leading to improvements in prostate cancer outcomes. Finally, as the complexity of gene regulation in normal and neoplastic prostate cells becomes better and better understood, other epigenomic determinants to prostate cancer cell phenotypes, including micro-RNAs, will become more thoroughly characterized, providing even more opportunities for biomarker discovery and new treatment development (113,114,115,116,117,118).

Footnotes

Each of the authors has been supported by National Institutes of Health/National Cancer Institute Grants R01 CA70196, GSTP1 Promoter Hypermethylation in Prostate Cancer, and P50 CA58236, SPORE in Prostate Cancer.

Disclosure Summary: W.G.N. is an inventor on U.S. Patent 5,552,277, entitled Genetic Diagnosis of Prostate Cancer, which has been licensed to OncoMethylome Sciences, Inc., and is entitled to receive royalty payments upon the sale of related products. W.G.N. and S.Y. are inventors on pending U.S. Patents 11/076,095, Epigenetic Tests for Prostate Cancer, and 60/755,980, COMPARE-MS. W.G.N. and S.Y. receive royalties from Oncomethylome Sciences; S.Y. has consulted for Veridex. A.M.D.M. has nothing to disclose.

First Published Online June 11, 2009

Abbreviations: BPH, Benign prostatic hyperplasia/hypertrophy; DBD, DNA-binding domain; DNMT, DNA methyltransferase; EZH2, enhancer of zeste homolog 2; GSTP1, glutathione S-transferase-π; HDAC, histone deacetylase; MBD, 5-mCpG-binding domain; PIA, proliferative inflammatory atrophy; PSA, prostate-specific antigen.

References

  1. Jones PA, Baylin SB 2002 The fundamental role of epigenetic events in cancer. Nat Rev 3:415–428 [DOI] [PubMed] [Google Scholar]
  2. Jones PA, Baylin SB 2007 The epigenomics of cancer. Cell 128:683–692 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bastian PJ, Yegnasubramanian S, Palapattu GS, Rogers CG, Lin X, De Marzo AM, Nelson WG 2004 Molecular biomarker in prostate cancer: the role of CpG island hypermethylation. Eur Urol 46:698–708 [DOI] [PubMed] [Google Scholar]
  4. Yegnasubramanian S, Kowalski J, Gonzalgo ML, Zahurak M, Piantadosi S, Walsh PC, Bova GS, De Marzo AM, Isaacs WB, Nelson WG 2004 Hypermethylation of CpG islands in primary and metastatic human prostate cancer. Cancer Res 64:1975–1986 [DOI] [PubMed] [Google Scholar]
  5. Yegnasubramanian S, Haffner MC, Zhang Y, Gurel B, Cornish TC, Wu Z, Irizarry RA, Morgan J, Hicks J, DeWeese TL, Isaacs WB, Bova GS, De Marzo AM, Nelson WG 2008 DNA hypomethylation arises later in prostate cancer progression than CpG island hypermethylation and contributes to metastatic tumor heterogeneity. Cancer Res 68:8954–8967 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bird AP 1986 CpG-rich islands and the function of DNA methylation. Nature 321:209–213 [DOI] [PubMed] [Google Scholar]
  7. Leonhardt H, Page AW, Weier HU, Bestor TH 1992 A targeting sequence directs DNA methyltransferase to sites of DNA replication in mammalian nuclei. Cell 71:865–873 [DOI] [PubMed] [Google Scholar]
  8. Mortusewicz O, Schermelleh L, Walter J, Cardoso MC, Leonhardt H 2005 Recruitment of DNA methyltransferase I to DNA repair sites. Proc Natl Acad Sci USA 102:8905–8909 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Jeltsch A 2002 Beyond Watson and Crick: DNA methylation and molecular enzymology of DNA methyltransferases. Chembiochem 3:274–293 [DOI] [PubMed] [Google Scholar]
  10. Okano M, Bell DW, Haber DA, Li E 1999 DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development. Cell 99:247–257 [DOI] [PubMed] [Google Scholar]
  11. Graff JR, Herman JG, Myöhänen S, Baylin SB, Vertino PM 1997 Mapping patterns of CpG island methylation in normal and neoplastic cells implicates both upstream and downstream regions in de novo methylation. J Biol Chem 272:22322–22329 [DOI] [PubMed] [Google Scholar]
  12. Vertino PM, Yen RW, Gao J, Baylin SB 1996 De novo methylation of CpG island sequences in human fibroblasts overexpressing DNA (cytosine-5-)-methyltransferase. Mol Cell Biol 16:4555–4565 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Feltus FA, Lee EK, Costello JF, Plass C, Vertino PM 2003 Predicting aberrant CpG island methylation. Proc Natl Acad Sci USA 100:12253–12258 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Bakin AV, Curran T 1999 Role of DNA 5-methylcytosine transferase in cell transformation by fos. Science 283:387–390 [DOI] [PubMed] [Google Scholar]
  15. Laird PW, Jackson-Grusby L, Fazeli A, Dickinson SL, Jung WE, Li E, Weinberg RA, Jaenisch R 1995 Suppression of intestinal neoplasia by DNA hypomethylation. Cell 81:197–205 [DOI] [PubMed] [Google Scholar]
  16. Eads CA, Nickel AE, Laird PW 2002 Complete genetic suppression of polyp formation and reduction of CpG-island hypermethylation in Apc(Min/+) Dnmt1-hypomorphic mice. Cancer Res 62:1296–1299 [PubMed] [Google Scholar]
  17. Belinsky SA, Klinge DM, Stidley CA, Issa JP, Herman JG, March TH, Baylin SB 2003 Inhibition of DNA methylation and histone deacetylation prevents murine lung cancer. Cancer Res 63:7089–7093 [PubMed] [Google Scholar]
  18. Camoriano M, Kinney SR, Moser MT, Foster BA, Mohler JL, Trump DL, Karpf AR, Smiraglia DJ 2008 Phenotype-specific CpG island methylation events in a murine model of prostate cancer. Cancer Res 68:4173–4182 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Morey Kinney SR, Smiraglia DJ, James SR, Moser MT, Foster BA, Karpf AR 2008 Stage-specific alterations of DNA methyltransferase expression, DNA hypermethylation, and DNA hypomethylation during prostate cancer progression in the transgenic adenocarcinoma of mouse prostate model. Mol Cancer Res 6:1365–1374 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Morey SR, Smiraglia DJ, James SR, Yu J, Moser MT, Foster BA, Karpf AR 2006 DNA methylation pathway alterations in an autochthonous murine model of prostate cancer. Cancer Res 66:11659–11667 [DOI] [PubMed] [Google Scholar]
  21. McCabe MT, Low JA, Daignault S, Imperiale MJ, Wojno KJ, Day ML 2006 Inhibition of DNA methyltransferase activity prevents tumorigenesis in a mouse model of prostate cancer. Cancer Res 66:385–392 [DOI] [PubMed] [Google Scholar]
  22. Eden A, Gaudet F, Waghmare A, Jaenisch R 2003 Chromosomal instability and tumors promoted by DNA hypomethylation. Science 300:455 [DOI] [PubMed] [Google Scholar]
  23. Gaudet F, Hodgson JG, Eden A, Jackson-Grusby L, Dausman J, Gray JW, Leonhardt H, Jaenisch R 2003 Induction of tumors in mice by genomic hypomethylation. Science 300:489–492 [DOI] [PubMed] [Google Scholar]
  24. Agoston AT, Argani P, De Marzo AM, Hicks JL, Nelson WG 2007 Retinoblastoma pathway dysregulation causes DNA methyltransferase 1 overexpression in cancer via MAD2-mediated inhibition of the anaphase-promoting complex. Am J Pathol 170:1585–1593 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Agoston AT, Argani P, Yegnasubramanian S, De Marzo AM, Ansari-Lari MA, Hicks JL, Davidson NE, Nelson WG 2005 Increased protein stability causes DNA methyltransferase 1 dysregulation in breast cancer. J Biol Chem 280:18302–18310 [DOI] [PubMed] [Google Scholar]
  26. De Marzo AM, Marchi VL, Yang ES, Veeraswamy R, Lin X, Nelson WG 1999 Abnormal regulation of DNA methyltransferase expression during colorectal carcinogenesis. Cancer Res 59:3855–3860 [PubMed] [Google Scholar]
  27. Hmadcha A, Bedoya FJ, Sobrino F, Pintado E 1999 Methylation-dependent gene silencing induced by interleukin 1β via nitric oxide production. J Exp Med 190:1595–1604 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Wang J, Hevi S, Kurash JK, Lei H, Gay F, Bajko J, Su H, Sun W, Chang H, Xu G, Gaudet F, Li E, Chen T 2009 The lysine demethylase LSD1 (KDM1) is required for maintenance of global DNA methylation. Nat Genet 41:125–129 [DOI] [PubMed] [Google Scholar]
  29. Varambally S, Dhanasekaran SM, Zhou M, Barrette TR, Kumar-Sinha C, Sanda MG, Ghosh D, Pienta KJ, Sewalt RG, Otte AP, Rubin MA, Chinnaiyan AM 2002 The polycomb group protein EZH2 is involved in progression of prostate cancer. Nature 419:624–629 [DOI] [PubMed] [Google Scholar]
  30. Gibbons RJ 2005 Histone modifying and chromatin remodelling enzymes in cancer and dysplastic syndromes. Hum Mol Genet 14(Spec No 1):R85–R92 [DOI] [PubMed] [Google Scholar]
  31. Chen H, Toyooka S, Gazdar AF, Hsieh JT 2003 Epigenetic regulation of a novel tumor suppressor gene (hDAB2IP) in prostate cancer cell lines. J Biol Chem 278:3121–3130 [DOI] [PubMed] [Google Scholar]
  32. Chen H, Tu SW, Hsieh JT 2005 Down-regulation of human DAB2IP gene expression mediated by polycomb Ezh2 complex and histone deacetylase in prostate cancer. J Biol Chem 280:22437–22444 [DOI] [PubMed] [Google Scholar]
  33. Ohm JE, McGarvey KM, Yu X, Cheng L, Schuebel KE, Cope L, Mohammad HP, Chen W, Daniel VC, Yu W, Berman DM, Jenuwein T, Pruitt K, Sharkis SJ, Watkins DN, Herman JG, Baylin SB 2007 A stem cell-like chromatin pattern may predispose tumor suppressor genes to DNA hypermethylation and heritable silencing. Nat Genet 39:237–242 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. McCabe MT, Lee EK, Vertino PM 2009 A multifactorial signature of DNA sequence and polycomb binding predicts aberrant CpG island methylation. Cancer Res 69:282–291 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Hendrich B, Bird A 1998 Identification and characterization of a family of mammalian methyl-CpG binding proteins. Mol Cell Biol 18:6538–6547 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Jones PL, Veenstra GJ, Wade PA, Vermaak D, Kass SU, Landsberger N, Strouboulis J, Wolffe AP 1998 Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription. Nat Genet 19:187–191 [DOI] [PubMed] [Google Scholar]
  37. Feng Q, Zhang Y 2001 The MeCP1 complex represses transcription through preferential binding, remodeling, and deacetylating methylated nucleosomes. Genes Dev 15:827–832 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Lin X, Nelson WG 2003 Methyl-CpG-binding domain protein-2 mediates transcriptional repression associated with hypermethylated GSTP1 CpG islands in MCF-7 breast cancer cells. Cancer Res 63:498–504 [PubMed] [Google Scholar]
  39. Sansom OJ, Berger J, Bishop SM, Hendrich B, Bird A, Clarke AR 2003 Deficiency of Mbd2 suppresses intestinal tumorigenesis. Nat Genet 34:145–147 [DOI] [PubMed] [Google Scholar]
  40. Lee WH, Morton RA, Epstein JI, Brooks JD, Campbell PA, Bova GS, Hsieh WS, Isaacs WB, Nelson WG 1994 Cytidine methylation of regulatory sequences near the π-class glutathione S-transferase gene accompanies human prostatic carcinogenesis. Proc Natl Acad Sci USA 91:11733–11737 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Nakayama M, Bennett CJ, Hicks JL, Epstein JI, Platz EA, Nelson WG, De Marzo AM 2003 Hypermethylation of the human glutathione S-transferase-π gene (GSTP1) CpG island is present in a subset of proliferative inflammatory atrophy lesions but not in normal or hyperplastic epithelium of the prostate: a detailed study using laser-capture microdissection. Am J Pathol 163:923–933 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Brooks JD, Weinstein M, Lin X, Sun Y, Pin SS, Bova GS, Epstein JI, Isaacs WB, Nelson WG 1998 CG island methylation changes near the GSTP1 gene in prostatic intraepithelial neoplasia. Cancer Epidemiol Biomarkers Prev 7:531–536 [PubMed] [Google Scholar]
  43. Nelson CP, Kidd LC, Sauvageot J, Isaacs WB, De Marzo AM, Groopman JD, Nelson WG, Kensler TW 2001 Protection against 2-hydroxyamino-1-methyl-6-phenylimidazo[4,5-b]pyridine cytotoxicity and DNA adduct formation in human prostate by glutathione S-transferase P1. Cancer Res 61:103–109 [PubMed] [Google Scholar]
  44. Nelson WG, De Marzo AM, Isaacs WB 2003 Prostate cancer. N Engl J Med 349:366–381 [DOI] [PubMed] [Google Scholar]
  45. Lin X, Tascilar M, Lee WH, Vles WJ, Lee BH, Veeraswamy R, Asgari K, Freije D, van Rees B, Gage WR, Bova GS, Isaacs WB, Brooks JD, DeWeese TL, De Marzo AM, Nelson WG 2001 GSTP1 CpG island hypermethylation is responsible for the absence of GSTP1 expression in human prostate cancer cells. Am J Pathol 159:1815–1826 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Henderson CJ, Smith AG, Ure J, Brown K, Bacon EJ, Wolf CR 1998 Increased skin tumorigenesis in mice lacking π-class glutathione S-transferases. Proc Natl Acad Sci USA 95:5275–5280 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. De Marzo AM, Marchi VL, Epstein JI, Nelson WG 1999 Proliferative inflammatory atrophy of the prostate: implications for prostatic carcinogenesis. Am J Pathol 155:1985–1992 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Putzi MJ, De Marzo AM 2000 Morphologic transitions between proliferative inflammatory atrophy and high-grade prostatic intraepithelial neoplasia. Urology 56:828–832 [DOI] [PubMed] [Google Scholar]
  49. Suzuki H, Toyota M, Kondo Y, Shinomura Y 2009 Inflammation-related aberrant patterns of DNA methylation: detection and role in epigenetic deregulation of cancer cell transcriptome. Methods Mol Biol 512:55–69 [DOI] [PubMed] [Google Scholar]
  50. Hahn MA, Hahn T, Lee DH, Esworthy RS, Kim BW, Riggs AD, Chu FF, Pfeifer GP 2008 Methylation of polycomb target genes in intestinal cancer is mediated by inflammation. Cancer Res 68:10280–10289 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Bedford MT, van Helden PD 1987 Hypomethylation of DNA in pathological conditions of the human prostate. Cancer Res 47:5274–5276 [PubMed] [Google Scholar]
  52. Santourlidis S, Florl A, Ackermann R, Wirtz HC, Schulz WA 1999 High frequency of alterations in DNA methylation in adenocarcinoma of the prostate. Prostate 39:166–174 [DOI] [PubMed] [Google Scholar]
  53. Schulz WA, Elo JP, Florl AR, Pennanen S, Santourlidis S, Engers R, Buchardt M, Seifert HH, Visakorpi T 2002 Genomewide DNA hypomethylation is associated with alterations on chromosome 8 in prostate carcinoma. Genes Chromosomes Cancer 35:58–65 [DOI] [PubMed] [Google Scholar]
  54. Catalona WJ, Smith DS, Ratliff TL, Dodds KM, Coplen DE, Yuan JJ, Petros JA, Andriole GL 1991 Measurement of prostate-specific antigen in serum as a screening test for prostate cancer. N Engl J Med 324:1156–1161 [DOI] [PubMed] [Google Scholar]
  55. Thompson IM, Pauler DK, Goodman PJ, Tangen CM, Lucia MS, Parnes HL, Minasian LM, Ford LG, Lippman SM, Crawford ED, Crowley JJ, Coltman Jr CA 2004 Prevalence of prostate cancer among men with a prostate-specific antigen level ≤4.0 ng per milliliter. N Engl J Med 350:2239–2246 [DOI] [PubMed] [Google Scholar]
  56. Thompson IM, Goodman PJ, Tangen CM, Lucia MS, Miller GJ, Ford LG, Lieber MM, Cespedes RD, Atkins JN, Lippman SM, Carlin SM, Ryan A, Szczepanek CM, Crowley JJ, Coltman Jr CA 2003 The influence of finasteride on the development of prostate cancer. N Engl J Med 349:215–224 [DOI] [PubMed] [Google Scholar]
  57. Carter HB, Pearson JD, Metter EJ, Brant LJ, Chan DW, Andres R, Fozard JL, Walsh PC 1992 Longitudinal evaluation of prostate-specific antigen levels in men with and without prostate disease. JAMA 267:2215–2220 [PMC free article] [PubMed] [Google Scholar]
  58. Carter HB, Ferrucci L, Kettermann A, Landis P, Wright EJ, Epstein JI, Trock BJ, Metter EJ 2006 Detection of life-threatening prostate cancer with prostate-specific antigen velocity during a window of curability. J Natl Cancer Inst 98:1521–1527 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Andriole GL, Crawford ED, Grubb 3rd RL, Buys SS, Chia D, Church TR, Fouad MN, Gelmann EP, Kvale PA, Reding DJ, Weissfeld JL, Yokochi LA, O'Brien B, Clapp JD, Rathmell JM, Riley TL, Hayes RB, Kramer BS, Izmirlian G, Miller AB, Pinsky PF, Prorok PC, Gohagan JK, Berg CD 2009 Mortality results from a randomized prostate-cancer screening trial. N Engl J Med 360:1310–1319 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Schröder FH, Hugosson J, Roobol MJ, Tammela TL, Ciatto S, Nelen V, Kwiatkowski M, Lujan M, Lilja H, Zappa M, Denis LJ, Recker F, Berenguer A, Määattanen L, Bangma CH, Aus G, Villers A, Rebillard X, van der Kwast T, Blijenberg BG, Moss SM, de Koning HJ, Auvinen A 2009 Screening and prostate-cancer mortality in a randomized European study. N Engl J Med 360:1320–1328 [DOI] [PubMed] [Google Scholar]
  61. Soh S, Kattan MW, Berkman S, Wheeler TM, Scardino PT 1997 Has there been a recent shift in the pathological features and prognosis of patients treated with radical prostatectomy? J Urol 157:2212–2218 [PubMed] [Google Scholar]
  62. Makhlouf AA, Krupski TL, Kunkle D, Theodorescu D 2004 The effect of sampling more cores on the predictive accuracy of pathological grade and tumour distribution in the prostate biopsy. BJU Int 93:271–274 [DOI] [PubMed] [Google Scholar]
  63. de la Taille A, Antiphon P, Salomon L, Cherfan M, Porcher R, Hoznek A, Saint F, Vordos D, Cicco A, Yiou R, Zafrani ES, Chopin D, Abbou CC 2003 Prospective evaluation of a 21-sample needle biopsy procedure designed to improve the prostate cancer detection rate. Urology 61:1181–1186 [DOI] [PubMed] [Google Scholar]
  64. Catalona WJ, Partin AW, Slawin KM, Brawer MK, Flanigan RC, Patel A, Richie JP, deKernion JB, Walsh PC, Scardino PT, Lange PH, Subong EN, Parson RE, Gasior GH, Loveland KG, Southwick PC 1998 Use of the percentage of free prostate-specific antigen to enhance differentiation of prostate cancer from benign prostatic disease: a prospective multicenter clinical trial. JAMA 279:1542–1547 [DOI] [PubMed] [Google Scholar]
  65. Luo J, Zha S, Gage WR, Dunn TA, Hicks JL, Bennett CJ, Ewing CM, Platz EA, Ferdinandusse S, Wanders RJ, Trent JM, Isaacs WB, De Marzo AM 2002 α-Methylacyl-CoA racemase: a new molecular marker for prostate cancer. Cancer Res 62:2220–2226 [PubMed] [Google Scholar]
  66. DeMarzo AM, Nelson WG, Isaacs WB, Epstein JI 2003 Pathological and molecular aspects of prostate cancer. Lancet 361:955–964 [DOI] [PubMed] [Google Scholar]
  67. Albertsen PC 2005 What is the value of screening for prostate cancer in the US? Nat Clin Pract Oncol 2:536–537 [DOI] [PubMed] [Google Scholar]
  68. Barry MJ 2009 Screening for prostate cancer: the controversy that refuses to die. N Engl J Med 360:1351–1354 [DOI] [PubMed] [Google Scholar]
  69. Hernandez DJ, Nielsen ME, Han M, Trock BJ, Partin AW, Walsh PC, Epstein JI 2008 Natural history of pathologically organ-confined (pT2), Gleason score 6 or less, prostate cancer after radical prostatectomy. Urology 72:172–176 [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Kumar-Sinha C, Tomlins SA, Chinnaiyan AM 2008 Recurrent gene fusions in prostate cancer. Nat Rev Cancer 8:497–511 [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Herman JG, Baylin SB 2003 Gene silencing in cancer in association with promoter hypermethylation. N Engl J Med 349:2042–2054 [DOI] [PubMed] [Google Scholar]
  72. Sidransky D 2002 Emerging molecular markers of cancer. Nat Rev Cancer 2:210–219 [DOI] [PubMed] [Google Scholar]
  73. Laird PW 2003 The power and the promise of DNA methylation markers. Nat Rev Cancer 3:253–266 [DOI] [PubMed] [Google Scholar]
  74. Yegnasubramanian S, Lin X, Haffner MC, DeMarzo AM, Nelson WG 2006 Combination of methylated-DNA precipitation and methylation-sensitive restriction enzymes (COMPARE-MS) for the rapid, sensitive and quantitative detection of DNA methylation. Nucleic Acids Res 34:e19 [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Lee WH, Isaacs WB, Bova GS, Nelson WG 1997 CG island methylation changes near the GSTP1 gene in prostatic carcinoma cells detected using the polymerase chain reaction: a new prostate cancer biomarker. Cancer Epidemiol Biomarkers Prev 6:443–450 [PubMed] [Google Scholar]
  76. Clark SJ, Harrison J, Paul CL, Frommer M 1994 High sensitivity mapping of methylated cytosines. Nucleic Acids Res 22:2990–2997 [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Herman JG, Graff JR, Myöhänen S, Nelkin BD, Baylin SB 1996 Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci USA 93:9821–9826 [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Gitan RS, Shi H, Chen CM, Yan PS, Huang TH 2002 Methylation-specific oligonucleotide microarray: a new potential for high-throughput methylation analysis. Genome Res 12:158–164 [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Harden SV, Sanderson H, Goodman SN, Partin AA, Walsh PC, Epstein JI, Sidransky D 2003 Quantitative GSTP1 methylation and the detection of prostate adenocarcinoma in sextant biopsies. J Natl Cancer Inst 95:1634–1637 [DOI] [PubMed] [Google Scholar]
  80. Gonzalgo ML, Pavlovich CP, Lee SM, Nelson WG 2003 Prostate cancer detection by GSTP1 methylation analysis of postbiopsy urine specimens. Clin Cancer Res 9:2673–2677 [PubMed] [Google Scholar]
  81. Gonzalgo ML, Nakayama M, Lee SM, De Marzo AM, Nelson WG 2004 Detection of GSTP1 methylation in prostatic secretions using combinatorial MSP analysis. Urology 63:414–418 [DOI] [PubMed] [Google Scholar]
  82. Bastian PJ, Palapattu GS, Lin X, Yegnasubramanian S, Mangold LA, Trock B, Eisenberger MA, Partin AW, Nelson WG 2005 Preoperative serum DNA GSTP1 CpG island hypermethylation and the risk of early prostate-specific antigen recurrence following radical prostatectomy. Clin Cancer Res 11:4037–4043 [DOI] [PubMed] [Google Scholar]
  83. Jerónimo C, Henrique R, Hoque MO, Ribeiro FR, Oliveira J, Fonseca D, Teixeira MR, Lopes C, Sidransky D 2004 Quantitative RARβ2 hypermethylation: a promising prostate cancer marker. Clin Cancer Res 10:4010–4014 [DOI] [PubMed] [Google Scholar]
  84. Liu L, Yoon JH, Dammann R, Pfeifer GP 2002 Frequent hypermethylation of the RASSF1A gene in prostate cancer. Oncogene 21:6835–6840 [DOI] [PubMed] [Google Scholar]
  85. Zhang J, Liu L, Pfeifer GP 2004 Methylation of the retinoid response gene TIG1 in prostate cancer correlates with methylation of the retinoic acid receptor β gene. Oncogene 23:2241–2249 [DOI] [PubMed] [Google Scholar]
  86. Zhu X, Leav I, Leung YK, Wu M, Liu Q, Gao Y, McNeal JE, Ho SM 2004 Dynamic regulation of estrogen receptor-β expression by DNA methylation during prostate cancer development and metastasis. Am J Pathol 164:2003–2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  87. Kaminskas E, Farrell A, Abraham S, Baird A, Hsieh LS, Lee SL, Leighton JK, Patel H, Rahman A, Sridhara R, Wang YC, Pazdur R 2005 Approval summary: azacitidine for treatment of myelodysplastic syndrome subtypes. Clin Cancer Res 11:3604–3608 [DOI] [PubMed] [Google Scholar]
  88. Lin X, Asgari K, Putzi MJ, Gage WR, Yu X, Cornblatt BS, Kumar A, Piantadosi S, DeWeese TL, De Marzo AM, Nelson WG 2001 Reversal of GSTP1 CpG island hypermethylation and reactivation of π-class glutathione S-transferase (GSTP1) expression in human prostate cancer cells by treatment with procainamide. Cancer Res 61:8611–8616 [PubMed] [Google Scholar]
  89. Cheng JC, Matsen CB, Gonzales FA, Ye W, Greer S, Marquez VE, Jones PA, Selker EU 2003 Inhibition of DNA methylation and reactivation of silenced genes by zebularine. J Natl Cancer Inst 95:399–409 [DOI] [PubMed] [Google Scholar]
  90. Segura-Pacheco B, Trejo-Becerril C, Perez-Cardenas E, Taja-Chayeb L, Mariscal I, Chavez A, Acuña C, Salazar AM, Lizano M, Dueñas-Gonzalez A 2003 Reactivation of tumor suppressor genes by the cardiovascular drugs hydralazine and procainamide and their potential use in cancer therapy. Clin Cancer Res 9:1596–1603 [PubMed] [Google Scholar]
  91. Carducci MA, Gilbert J, Bowling MK, Noe D, Eisenberger MA, Sinibaldi V, Zabelina Y, Chen TL, Grochow LB, Donehower RC 2001 A Phase I clinical and pharmacological evaluation of sodium phenylbutyrate on an 120-h infusion schedule. Clin Cancer Res 7:3047–3055 [PubMed] [Google Scholar]
  92. Kelly WK, O'Connor OA, Krug LM, Chiao JH, Heaney M, Curley T, MacGregore-Cortelli B, Tong W, Secrist JP, Schwartz L, Richardson S, Chu E, Olgac S, Marks PA, Scher H, Richon VM 2005 Phase I study of an oral histone deacetylase inhibitor, suberoylanilide hydroxamic acid, in patients with advanced cancer. J Clin Oncol 23:3923–3931 [DOI] [PMC free article] [PubMed] [Google Scholar]
  93. Kelly WK, Richon VM, O'Connor O, Curley T, MacGregor-Curtelli B, Tong W, Klang M, Schwartz L, Richardson S, Rosa E, Drobnjak M, Cordon-Cordo C, Chiao JH, Rifkind R, Marks PA, Scher H 2003 Phase I clinical trial of histone deacetylase inhibitor: suberoylanilide hydroxamic acid administered intravenously. Clin Cancer Res 9:3578–3588 [PubMed] [Google Scholar]
  94. Mann BS, Johnson JR, He K, Sridhara R, Abraham S, Booth BP, Verbois L, Morse DE, Jee JM, Pope S, Harapanhalli RS, Dagher R, Farrell A, Justice R, Pazdur R 2007 Vorinostat for treatment of cutaneous manifestations of advanced primary cutaneous T-cell lymphoma. Clin Cancer Res 13:2318–2322 [DOI] [PubMed] [Google Scholar]
  95. Thibault A, Figg WD, Bergan RC, Lush RM, Myers CE, Tompkins A, Reed E, Samid D 1998 A phase II study of 5-aza-2′deoxycytidine (decitabine) in hormone independent metastatic (D2) prostate cancer. Tumori 84:87–89 [DOI] [PubMed] [Google Scholar]
  96. Cameron EE, Bachman KE, Myöhänen S, Herman JG, Baylin SB 1999 Synergy of demethylation and histone deacetylase inhibition in the re-expression of genes silenced in cancer. Nat Genet 21:103–107 [DOI] [PubMed] [Google Scholar]
  97. Gore SD, Baylin S, Sugar E, Carraway H, Miller CB, Carducci M, Grever M, Galm O, Dauses T, Karp JE, Rudek MA, Zhao M, Smith BD, Manning J, Jiemjit A, Dover G, Mays A, Zwiebel J, Murgo A, Weng LJ, Herman JG 2006 Combined DNA methyltransferase and histone deacetylase inhibition in the treatment of myeloid neoplasms. Cancer Res 66:6361–6369 [DOI] [PubMed] [Google Scholar]
  98. Ferguson AT, Lapidus RG, Baylin SB, Davidson NE 1995 Demethylation of the estrogen receptor gene in estrogen receptor-negative breast cancer cells can reactivate estrogen receptor gene expression. Cancer Res 55:2279–2283 [PubMed] [Google Scholar]
  99. Ottaviano YL, Issa JP, Parl FF, Smith HS, Baylin SB, Davidson NE 1994 Methylation of the estrogen receptor gene CpG island marks loss of estrogen receptor expression in human breast cancer cells. Cancer Res 54:2552–2555 [PubMed] [Google Scholar]
  100. Yang X, Ferguson AT, Nass SJ, Phillips DL, Butash KA, Wang SM, Herman JG, Davidson NE 2000 Transcriptional activation of estrogen receptor α in human breast cancer cells by histone deacetylase inhibition. Cancer Res 60:6890–6894 [PubMed] [Google Scholar]
  101. Yang X, Phillips DL, Ferguson AT, Nelson WG, Herman JG, Davidson NE 2001 Synergistic activation of functional estrogen receptor (ER)-α by DNA methyltransferase and histone deacetylase inhibition in human ER-α-negative breast cancer cells. Cancer Res 61:7025–7029 [PubMed] [Google Scholar]
  102. Kato Y, Salumbides BC, Wang XF, Qian DZ, Williams S, Wei Y, Sanni TB, Atadja P, Pili R 2007 Antitumor effect of the histone deacetylase inhibitor LAQ824 in combination with 13-cis-retinoic acid in human malignant melanoma. Mol Cancer Ther 6:70–81 [DOI] [PubMed] [Google Scholar]
  103. Pili R, Kruszewski MP, Hager BW, Lantz J, Carducci MA 2001 Combination of phenylbutyrate and 13-cis retinoic acid inhibits prostate tumor growth and angiogenesis. Cancer Res 61:1477–1485 [PubMed] [Google Scholar]
  104. Qian DZ, Ren M, Wei Y, Wang X, van de Geijn F, Rasmussen C, Nakanishi O, Sacchi N, Pili R 2005 In vivo imaging of retinoic acid receptor β2 transcriptional activation by the histone deacetylase inhibitor MS-275 in retinoid-resistant prostate cancer cells. Prostate 64:20–28 [DOI] [PubMed] [Google Scholar]
  105. Sirchia SM, Ren M, Pili R, Sironi E, Somenzi G, Ghidoni R, Toma S, Nicolò G, Sacchi N 2002 Endogenous reactivation of the RARβ2 tumor suppressor gene epigenetically silenced in breast cancer. Cancer Res 62:2455–2461 [PubMed] [Google Scholar]
  106. Wang XF, Qian DZ, Ren M, Kato Y, Wei Y, Zhang L, Fansler Z, Clark D, Nakanishi O, Pili R 2005 Epigenetic modulation of retinoic acid receptor β2 by the histone deacetylase inhibitor MS-275 in human renal cell carcinoma. Clin Cancer Res 11:3535–3542 [DOI] [PubMed] [Google Scholar]
  107. Song SH, Jong HS, Choi HH, Inoue H, Tanabe T, Kim NK, Bang YJ 2001 Transcriptional silencing of cyclooxygenase-2 by hyper-methylation of the 5′ CpG island in human gastric carcinoma cells. Cancer Res 61:4628–4635 [PubMed] [Google Scholar]
  108. Sharma D, Saxena NK, Davidson NE, Vertino PM 2006 Restoration of tamoxifen sensitivity in estrogen receptor-negative breast cancer cells: tamoxifen-bound reactivated ER recruits distinctive corepressor complexes. Cancer Res 66:6370–6378 [DOI] [PMC free article] [PubMed] [Google Scholar]
  109. Schuebel KE, Chen W, Cope L, Glöckner SC, Suzuki H, Yi JM, Chan TA, Van Neste L, Van Criekinge W, van den Bosch S, van Engeland M, Ting AH, Jair K, Yu W, Toyota M, Imai K, Ahuja N, Herman JG, Baylin SB 2007 Comparing the DNA hypermethylome with gene mutations in human colorectal cancer. PLoS Genet 3:1709–1723 [DOI] [PMC free article] [PubMed] [Google Scholar]
  110. Costello JF, Frühwald MC, Smiraglia DJ, Rush LJ, Robertson GP, Gao X, Wright FA, Feramisco JD, Peltomäki P, Lang JC, Schuller DE, Yu L, Bloomfield CD, Caligiuri MA, Yates A, Nishikawa R, Su Huang H, Petrelli NJ, Zhang X, O'Dorisio MS, Held WA, Cavenee WK, Plass C 2000 Aberrant CpG-island methylation has non-random and tumour-type-specific patterns. Nat Genet 24:132–138 [DOI] [PubMed] [Google Scholar]
  111. Dobzhansky T 1946 Genetics of natural populations. XIII. Recombination and variability in populations of Drosophila pseudoobscura. Genetics 31:269–290 [DOI] [PMC free article] [PubMed] [Google Scholar]
  112. Tong AH, Evangelista M, Parsons AB, Xu H, Bader GD, Pagé N, Robinson M, Raghibizadeh S, Hogue CW, Bussey H, Andrews B, Tyers M, Boone C 2001 Systematic genetic analysis with ordered arrays of yeast deletion mutants. Science 294:2364–2368 [DOI] [PubMed] [Google Scholar]
  113. Ambs S, Prueitt RL, Yi M, Hudson RS, Howe TM, Petrocca F, Wallace TA, Liu CG, Volinia S, Calin GA, Yfantis HG, Stephens RM, Croce CM 2008 Genomic profiling of microRNA and messenger RNA reveals deregulated microRNA expression in prostate cancer. Cancer Res 68:6162–6170 [DOI] [PMC free article] [PubMed] [Google Scholar]
  114. Lin SL, Chiang A, Chang D, Ying SY 2008 Loss of mir-146a function in hormone-refractory prostate cancer. RNA 14:417–424 [DOI] [PMC free article] [PubMed] [Google Scholar]
  115. Louro R, Nakaya HI, Amaral PP, Festa F, Sogayar MC, da Silva AM, Verjovski-Almeida S, Reis EM 2007 Androgen responsive intronic non-coding RNAs. BMC Biol 5:4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  116. Shi XB, Xue L, Yang J, Ma AH, Zhao J, Xu M, Tepper CG, Evans CP, Kung HJ, deVere White RW 2007 An androgen-regulated miRNA suppresses Bak1 expression and induces androgen-independent growth of prostate cancer cells. Proc Natl Acad Sci USA 104:19983–19988 [DOI] [PMC free article] [PubMed] [Google Scholar]
  117. Sun T, Wang Q, Balk S, Brown M, Lee GS, Kantoff P 2009 The role of microRNA-221 and microRNA-222 in androgen-independent prostate cancer cell lines. Cancer Res 69:3356–3363 [DOI] [PMC free article] [PubMed] [Google Scholar]
  118. Volinia S, Calin GA, Liu CG, Ambs S, Cimmino A, Petrocca F, Visone R, Iorio M, Roldo C, Ferracin M, Prueitt RL, Yanaihara N, Lanza G, Scarpa A, Vecchione A, Negrini M, Harris CC, Croce CM 2006 A microRNA expression signature of human solid tumors defines cancer gene targets. Proc Natl Acad Sci USA 103:2257–2261 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Endocrinology are provided here courtesy of The Endocrine Society

RESOURCES