• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of epigLink to Publisher's site
Epigenetics. Mar 2011; 6(3): 388–394.
Published online Mar 1, 2011. doi:  10.4161/epi.6.3.14056
PMCID: PMC3063331
NIHMSID: NIHMS269745

Downregulation of microRNA-29c is associated with hypermethylation of tumor-related genes and disease outcome in cutaneous melanoma

Abstract

Hypermethylation of the promoter region of tumor-related genes (TRGs) has been shown to silence gene expression during melanoma progression, whereas microRNA-29(miR-29) has been found to downregulate DNA methyltransferases DNMT3A and DNMT3B which were shown as essential to the methylation of TRGs. We hypothesized that the expression level of miR-29 is associated to TRG methylation status and may have prognostic utility in melanoma. AJCC stage I–IV cutaneous melanoma paraffin-embedded archival tissue (PEAT) specimens (n = 149) were assessed. Expression of miR-29 isoforms a, b and c were analyzed by reverse-transcription quantitative real-time polymerase chain reaction (RT-qPCR). Expression of DNMT3A and DNMT3B was assessed by immunohistochemistry (IHC) on defined clinically annotated tissue microarrays (TMA) of AJCC stage III melanoma lymph node metastases. Promoter region CpG island methylation status of RASSF1A, TFPI-2, RAR-β, SOCS, GATA4 and genomic sequences MINT17 and MINT31 were previously evaluated in melanoma tissues. miR-29c isoform expression was correlated to advancing AJCC stages in melanoma. miR-29c expression was significantly downregulated in AJCC stage IV melanoma tumors compared to primary melanomas. Hypermethylation status of TRGs and non-coding MINT loci in different stages of melanoma showed an inverse association with miR-29c expression. Overall, an increase in miR-29c expression inversely correlated to both DNMT3A and DNMT3B protein expression in melanomas. Expression of DNMT3B and miR-29c were significantly (p = 0.004 and p = 0.002, respectively) associated with overall survival (OS) in AJCC stage III melanoma patients by multivariate analysis. The studies demonstrated that both miR-29c and DNMT3B have significant roles in melanoma progression and may be useful epigenetic biomarkers for disease outcome.

Key words: DNMT3, melanoma, metastasis, methylation, miR-29

Introduction

Recent studies demonstrated that specific microRNAs (miRs) have functional roles in human cancer progression.13 miR is a class of naturally occurring small non coding RNAs which can function as regulators of gene expression. Mature miRs are typically 19–25 nucleotides in length4,5 and have been found to bind to the 3′UTR of target messenger RNA (mRNA) and inhibit translation by blocking mRNA transcription or cause degradation of mRNA.46 Studies have uncovered specific microRNA-29 (miR-29) isoforms as being associated with specific malignancies and have tumor suppressive roles.69 Members of the miR-29 family are located on chromosome 1q32.2 and 7q32 and have complementary sites to the 3′UTR of DNA methyltransferase (DNMT) 3A and DNMT3B genes, which can globally influence gene methylation status. Transfected miR-29s were demonstrated as effective inhibitory agents decreasing DNMT3 expression, can restore normal patterns of methylation and cause changes in cell apoptosis, proliferation, invasion, migration and colony formation.10

Cutaneous melanoma is the sixth most common cancer in the US with increasing incidence worldwide11 and methylation of gene promoter region has been found to epigenetically regulate cell functions in melanoma.1216 Recently, we have shown that CpG island methylator phenotype (CIMP) is associated with disease progression in melanoma.14 Tumor-related genes (TRGs) and other noncoding methylated-in-tumor (MINT) loci were found inactivated in a coordinated fashion through methylation of promoter CpG islands in melanoma.14,17,18 miR-29 has been suggested to affect DNMT3A and DNMT3B activity, and these methyltransferases have been found essential for de novo methylation. We therefore investigated the relationship among miR-29, DNMT3 and methylated TRGs and MINT in melanomas and their role in tumor progression.

Results

miR-29 expression.

To determine the significance of miR-29 in cutaneous melanoma, the expression of miR-29 isoforms a, b and c was assessed by reverse-transcription quantitative real-time polymerase chain reaction (RT-qPCR) in melanoma paraffin-embedded archival tissues (PEAT) of different stages of melanoma. The miR-29 RT-qPCR assay was optimized for sensitivity and verified for specificity. Assessment of miR-29 expression was performed on a training set of melanomas that included AJCC stage I and II primary tumors (n = 11) and stage III and IV metastatic tumors (n = 8). All three miR-29 isoforms were found to be expressed lower in stage III and IV metastases compared to stage I/II primary tumors. The differences in the expression of miR-29a and -b in primary versus metastatic melanoma were not significant; however, expression of miR-29c significantly decreased in metastatic tumors (p = 0.003) (Fig. 1). Based on this analysis, we then focused our efforts on investigating miR-29c in additional cutaneous melanoma tumors for a total of 15 primary, 16 lymph node (LN) metastases and 31 distant metastases. Significant decrease in miR-29c expression was found both in the LN metastases (p = 0.02) and in distant metastases (p = 0.006) as compared to primary melanomas (Fig. 2).

Figure 1
Expression of miR-29 in AJCC stage I and II primary melanomas (n = 11) versus stage III and IV metastatic melanomas (n = 8). Assessments of miR-29a (A) and miR-29b (B) show no significance between primary tumors and metastases. Expression of miR-29c (C) ...
Figure 2
Comparing expression levels of miR-29c in AJCC stage I and II primary melanomas (n = 15) versus stage III LN metastases (n = 16) (p = 0.02) or stage IV distant metastases (n = 31) (p = 0.006).

Correlation of miR-29c to DNMT3 expression.

We performed immunohistochemistry (IHC) analysis for DNMT3A and DNMT3B expression in the same melanoma tumors that were RT-qPCR assessed for miR-29c (Fig. 3) DNMT3A and DNMT3B expression was separately quantified and compared in AJCC stages I and II primary melanoma tumors (n = 6) versus AJCC stages III and IV melanoma metastases (n = 12) (Fig. 4). Both DNMT3A and DNMT3B were found significantly higher in stages III and IV metastatic melanomas compared to stage I and II primary melanomas (p = 0.0007) (Fig. 4). miR-29c expression was inversely correlated to DNMT3A (Pearson correlation coefficients-0.676, p = 0.002) and DNMT3B (Pearson correlation coefficients-0.628, p = 0.005) expression.

Figure 3
Representative IHC staining with DNMT3A and DNMT3B antibodies. DNMT3A (A) and DNMT3B (B) are highly expressed in stage IV metastatic melanoma. (C) is the no-primary antibody control for the stage IV metastatic melanoma. DNMT3A (D) and DNMT3B (E) are weakly ...
Figure 4
Box plots comparing IHC staining intensity of AJCC stage I and II primary tumors (n = 6), stage III metastatic tumors (n = 7) and stage IV metastatic tumors (n = 5) of DNMT3A (A) and DNMT3B (B).

Correlation of miR-29c expression to CIMP.

Within the miR-29c specimen analysis, a subset of melanoma tumors that included four primary tumors and 30 AJCC stage III and IV metastases were previously studied for methylation status of TRGs and MINT. The expression level of miR-29c was compared with the methylation status of RAS association family 1A (RASSF1A), tissue factor pathway inhibitor-2 (TFPI-2), retinoic acid receptor β (RAR-β), Wnt inhibitory factor 1 (WIF-1), suppressor of cytokine signaling proteins (SOCS), a transcription factor (GATA4), Methylated-In-Tumor (MINT) 17 and MINT 31, which in combination are related to CIMP of melanoma.14 miR-29c expression was found inversely correlated to CIMP of melanomas (p = 0.033); lower expression of miR-29c is associated with more frequent methylation of TRGs and MINT loci in melanoma. Of all of the biomarkers assessed, only RASSF1A hypomethylation was significantly correlated to miR29c expression (p = 0.045).

miR-29c and disease outcome.

The disease-free survival (DFS) and overall survival (OS) were evaluated relative to miR-29c expression in LN (n = 87) from melanoma patient with Stage III disease. Tumor cells were microdissected from each tissue section as to prevent non-cancerous cells from contaminating the analysis. The assessment of disease outcome in relationship to miR-29c was focused on a group of patients whose LN metastases were included in our tissue microarray (TMA) and whose tissue blocks were available and sufficient for miR-29c RT-qPCR analysis. Patients were divided into two groups based on their miR-29c expression levels: those with less than median miR-29c expression level and those with greater than or equal to median miR-29c expression level. DFS and OS were evaluated relative to miR-29c expression. Expression of miR-29c above the median level was found a significant predictor of improved disease-free survival (DFS) (p = 0.003) and overall survival (OS) (p = 0.0002) in univariate analyses (Fig. 5A, B). When analyzed in a multivariate Cox Regression model, greater than or equal to median miR-29c expression level is significantly correlated to improved OS (HR 0.324, 95% CI: 0.156–0.670, p = 0.002), but not DFS.

Figure 5
Kaplan-Meier curve analysis of AJCC stage III patients grouped by high and low miR-29c (n = 87) and DNMT3B (n = 81) expression in their LN metastases: DFS (A) and OS (B) of miR-29c expression as assessed by RT-qPCR and OS (C) of DNMT3B IHC staining levels. ...

DNMT3 correlation to disease outcome.

Using IHC analysis, we assessed DNMT3A and DNMT3B expression level in the LN metastases on the TMA and correlated IHC staining intensity with OS and DFS of the patients (n = 81). All of the metastatic melanoma cell lines in the TMA had strong IHC staining for DNMT3A and DNMT3B while the normal tissues (non-cancerous normal LNs) expressed minimal DNMT3A and DNMT3B. DNMT3B was found to be a significant (p = 0.006) prognostic factor for OS (Fig. 5C) with DFS approaching significance (p = 0.075) in univariate analyses. When analyzed in a Cox Regression multivariate model comparing other prognostic factors, DNMT3B was found significant (HR 2.63, 95% CI: 1.371–5.046, p = 0.004) in predicting 5 yr OS and not DFS. DNMT3A expression level was not significant for OS or DFS (data not shown).

Discussion

All isoforms of miR-29 can variably influence methylation and suppress tumorigenicity in different tumors,6,7,10 however only miR-29c isoform was found to inversely correlate to DNMT3 expression and CIMP in melanoma. Furthermore, miR-29c expression was found to correlate to OS in stage III melanoma patients. Based on these results, miR-29c expression may potentially provide important treatment decision making information for stage III melanoma patients by differentiating aggressive disease. A biomarker that can separate LN metastases for aggressiveness and prognosis may be helpful in stratifying patients for appropriate adjuvant therapy.

Multiple mechanisms of how miR-29c regulate cell functions in tumors have been investigated, including its anti-apoptotic association,7,19 suppressor gene activation20 and methyltransferase inhibition.10 Because we have previously found CIMP events in melanoma to be associated with tumor progression,14 we focused on whether miR-29 may be correlated to CIMP possibly through its influence on DNMT3 by first comparing miR-29c expression to the hypermethylation status of TRGs and MINT loci in melanoma tumors. Whether by downregulating DNMT3 or effects on other tumorigenic pathways, miR-29c decreases in more advanced melanoma tumors where more frequent methylation has also been reported. Further analyses will be needed to determine whether miR-29c expression level can be used as a potential surrogate CIMP biomarker.

Past studies have shown that both methyltransferases, DNMT3A and DNMT3B, are involved in de novo methylation,2124 and silencing these enzymes by knockdown or deletion leads to hypomethylation of TRGs.6,10,23,24 Having established that there is a relationship between miR-29c and CIMP in melanoma, we then demonstrated an inverse relationship between miR-29c expression level and presence of DNMT3 in melanoma. The inverse relationship between expression of miR-29c and DNMT3A and DNMT3B in melanoma corroborates the antagonistic role of miR29c in regulating DNMT3A and DNMT3B found in other malignancies. Further, DNMT3 family members have been demonstrated as gene-specific in catalyzing de novo methylation only in specific gene promoter regions,23,24 which may partially explain the lack of correlation of specific gene methylation to that of miR-29c expression. We did, however, find an inverse correlation between RASSF1A hypermethylation and miR-29c expression in melanoma. RASSF1A has been considered to have a role in cell growth and metastasis18,25 in various cancers including melanoma;26,27 RASSF1A hypermethylation has been reported as a mechanism by which the suppressor is negatively regulated in cancer and contributes to melanoma progression.14,17,18 The correlation between miR-29c and RASSF1A demonstrates another tier of tumor suppressor gene regulation, adding to the complexity of tumor progression.

The detection of hypermethylated melanoma-related TRGs has been correlated to poor OS and DFS,13,14 and we demonstrated in this study that miR-29c and DNMT3B can also independently predict disease outcome. To our knowledge, this is the first report to demonstrate miR-29c as a biomarker significantly predicting OS and DFS in cutaneous melanoma. The studies also demonstrate that miR-29c is a potential clinicopathological biomarker for differentiating between individual melanoma stages, and its downregulation may be indicative of aggressive disease. There are reports of DNMT3 overexpression being significantly associated in cancer progression.22,28 DNMT3B has been suggested to predict DFS while DNMT3A showed no significance in predicting DFS in breast cancer.28 Another study demonstrated that the expression of DNMT3A to be significantly associated with only OS, while no significant correlation was observed for DNMT3B in the same lung cancer patient group.22 In our analysis of the correlation of DNMT3A and DNMT3B to OS and DFS in LN metastases of AJCC stage III patients, only DNMT3B presence was predictive of 5 year OS while DNMT3A was not. The use of TMA with LN metastases stratified into clinically annotated disease outcomes provided an excellent platform for the rigorous assessment of DNMT3A and DNMT3B utility as biomarker for AJCC stage III melanoma.

Recently, studies have attempted profiling miR expression in cancers; however, few miRs have been identified as having significant prognostic utility in the assessment of tumor progression and disease outcome.29,30 Here, we reported significant correlations of miR-29c and DNMT3B to disease outcome. Along with TRG hypermethylation, these epigenetic biomarkers provide important information on melanoma, such as CIMP activity during tumor progression.

Materials and Methods

Patients.

The use of human subjects was approved by the Western Institution Review Board (WIRB). PEAT were obtained from 149 patients diagnosed with AJCC stage I–IV cutaneous melanoma: AJCC stage I (n = 11), II (n = 4), III (n = 103) and IV (n = 31). Eighty-seven of the stage III melanoma patients are those whose LN metastases are included in a TMA which was used for DNMT3 studies. The TMA contained LN metastases from a cohort of 160 AJCC stage III melanoma patients, but only 87 had the original PEAT blocks with sufficient tissue available to conduct microdissection for miR-29 RT-qPCR studies. TMA patients were categorized as either good overall survival (>5 years; n = 80) or poor overall survival (<2 years; n = 80). These melanoma patients used in the TMA had surgery between 1993 and 2006, with a median follow-up of 3.5 years.

Melanoma LN tissue microarray (TMA).

The melanoma TMA was constructed at Yale U. Dept of Pathology by Dr. D. Rimm31 using AJCC stage III melanoma LN metastases with defined clinical outcome and follow-up. Melanoma for each PEAT block was identified; cores measuring 0.6 mm in diameter were made and array was constructed with each core spaced out at 8.0 mm apart (Beecher Instruments).31 The TMA also included ten melanoma cell lines, 12 histopathology negative LN (cancer-free), and two normal liver tissues to be used as positive or negative controls. Duplicated cores of all specimens and controls were included in each TMA as a quality assurance to ensure that all of the specimens on the slides are uniformly represented.

Immunohistochemistry.

IHC was performed on 5 µm PEAT and TMA sections that had been incubated overnight at 37°C. DNMT3A and DNMT3B protein expression was assessed using DNMT3A and DNMT3B (accession NP072046 and NP008823 respectively) rabbit polyclonal Ab (Abgent, San Diego, CA) separately, at a dilution of 1:50. IHC was performed using an optimized protocol. Slides were deparaffinized, rehydrated and washed in 1x PBS as previously described.32 Antigen retrieval was performed with 1x citrate buffer at 100°C for 10 min and then incubated in H2O2 at room temperature to block endogenous peroxidase. Separate slides were incubated in primary Ab against DNMT3A or DNMT3B overnight in a 4°C humid chamber followed by 1 h incubation with secondary biotinylated link Ab and peroxidase; sections were then counterstained with Methyl Green (Vector Labs, Burlingame, CA). All tissues are expected to express some level of DNMT3A and DNMT3B, therefore tissues treated with only secondary Ab under the same conditions were used as negative control and non-cancerous normal human tissues on the TMA are checked for background level expression. A photograph of each IHC-stained section was taken for analysis using a Nikon Eclipse Ti microscope and NIS elements software (Nikon, Melville, NY). After the IHC, staining density was determined by ImageJ software (rsbweb.nih.gov/ij) following adjustment for background on each selected fi31eld of image of the stained slide, the density of the individual LN specimen was quantified and given a numerical value from 0–250. There were duplicates for each patient LN specimens, so the average of the two staining intensity values was used for correlation analysis.

RNA extraction and RT-qPCR.

Total RNA was extracted using our optimized protocol as previously described.32,33 Three 10 µm thick tissue sections were cut from each PEAT with a sterile microtome blade which were then mounted on non-charged slides and dried at 37°C overnight. Tumor tissues were identified and marked on the H&E stained sections. Manual needle micro-dissection was performed to isolate tumor tissues for total RNA extraction. Each tissue was deparaffinized, washed in ethanol and digested in proteinase K solution for 3 h. RNAWiz (Ambion, Austin, TX) was used to homogenize the tissue and extract RNA following the manufacturer's instructions. Pellet paint (Novagen, Madison, WI) was added during RNA precipitation. Total RNA was assessed for purity by UV spectrophotometry and quantified using the Quant-iT RiboGreen RNA Assay Kit (Invitrogen, Carlsbad, CA). If the RNA was low in concentration or poor in quality, specimens were not used in the study. Established melanoma cell line (M14) was used as a positive controls for the miR-29 RT-qPCR assays.

cDNA synthesis was performed using modified procedures outlined for miRCURY LNA microRNA PCR primer set (Exiqon, MA). Briefly, RT reactions were performed with 10 ng of total RNA using Maloney Murine Leukemia Virus reverse-transcriptase (Promega, Madison, WI) with specific miR primers from Exiqon (Exiqon, Woburn, MA). RNA was denatured at 70°C for 10 min, incubated at 37°C for 2.5 h and followed by a 95°C denaturing step for 5 min. Following Exiqon recommended protocol, the synthesized cDNA was diluted 10-fold before being used for RT-qPCR.

The target sequences for has-miR-29 were: miR-29a UAG CAC CAU CUG AAA UCG GUU A, miR-29b UAG CAC CAU UUG AAA UCA GUG UU and miR-29c UAG CAC CAU UUG AAA UCG GUU A with the PCR primers designed by Exiqon to work with a universal primer targeting the tag that was inserted during RT. The 5s endogenous control primers were from Exiqon. The PCR mixture consisted of 2.5 U/µl of Quanta PerfecTA Supermix (Quanta Bioscience, Gaithersburg, MD) with LNA forward and universal reverse primers, cDNA from 1 ng of total RNA and molecular grade water. The samples were initially heated at 95°C for 10 min, followed by a 30 cycle of 95°C for 60 sec and 60°C for 60 sec. The same conditions were used for miR-29 and the 5s endogenous control. A reference gene is used for normalizing the miR RT-qPCR and to ensure the quality of assays. Initially, three potential candidates, 5s, LET-7i and U6, were evaluated for consistent detection in melanoma PEAT. 5s was chosen for its ubiquitous expression in melanoma PEAT and consistency among various types of tissues.

Each RT-qPCR assay was performed in duplicate and included positive, negative and no template reagent controls. The relative miR expression level was obtained by subtracting the average of quantification cycle (Cq) values from the duplicate reaction of reference control from that of miR-29 to obtain the difference in Cq value (ΔCq).

Methylation status of CIMP in melanomas.

In a study previously conducted, methylation-specific PCR was used to obtain methylation status of TRGs, while absolute quantitative analysis of methylated alleles (AQAMA) methylation was used for MINT 17 and MINT 31 analysis.14 The U-index derived from AQAMA was further dichotomized into methylated or unmethylated by a cutoff that is one standard deviation above the mean U-index of 12 non-tumor skin specimens. The overall number of methylated gene or non-coding loci was compared to miR-29c expression level.

Biostatistics.

In analyzing continuous variables, Student's t-test and Pearson correlation coefficients were used. For categorical variables, Chi-square test or Fisher's exact test were used. The 5 year OS and DFS rates for patient groups were calculated using Kaplan-Meier methods and compared using log rank test. Cox proportional hazards (Cox-PH) models were used for multivariate analysis of OS and DFS incorporating the following standard prognostic variables along with miR29c, DNMT3A, DNMT3B: Breslow thickness, Clark level, ulceration, age and sex. In finding best predictive Cox PH regression model, stepwise method was used. p values ≤0.05 are considered significant.

Acknowledgements

This work was supported by the Dr. Miriam and Sheldon G. Adelson Medical Research Foundation and P0 CA029605 and P0 CA012582 from the NIH, NCI. The content is solely the responsibility of the authors and does not necessarily represent the official view of the National Cancer Institute or the National Institutes of Health. We thank Kana Rivera and Linhda Nguyen for editorial assistance, the Department of Molecular Oncology staff for support, Xing Ye, M.Sc., for biostatistical support and Division of Surgical Pathology staff at Saint John's Health Center for pathology support.

Abbreviations

AQAMA
absolute quantitative analysis of methylated alleles
CIMP
CpG island methylator phenotype
DFS
disease-free survival
DNMT
DNA methyltransferase
IHC
immunohistochemistry
LN
lymph node
mRNA
messenger RNA
MINT
methylated-in-tumor
MI
methylation index
miR
microRNA
miR-29
microRNA-29
OS
overall survival
PEAT
paraffin-embedded archival tissues
Cq
quantification cycle
RASSF1A
RAS association family 1A
RT-qPCR
reverse-transcription quantitative real-time polymerase chain reaction
TFPI-2
tissue factor pathway inhibitor-2
TMA
tissue microarray
TRGs
tumor-related genes

References

1. Iorio M V, Ferracin M, Liu CG, Veronese A, Spizzo R, Sabbioni S, et al. MicroRNA gene expression deregulation in human breast cancer. Cancer Res. 2005;65:7065–7070. [PubMed]
2. Volinia S, Calin GA, Liu CG, Ambs S, Cimmino A, Petrocca F, et al. A microRNA expression signature of human solid tumors defines cancer gene targets. Proc Natl Acad Sci USA. 2006;103:2257–2261. [PMC free article] [PubMed]
3. Dykxhoorn DM. MicroRNAs and metastasis: Little RNAs go a long way. Cancer Res. 2010;70:6401–6406. [PMC free article] [PubMed]
4. Bartel D P. MicroRNAs: genomics, biogenesis, mechanism and function. Cell. 2004;116:281–297. [PubMed]
5. Doench JG, Sharp PA. Specificity of microRNA target selection in translational repression. Genes Dev. 2004;18:504–511. [PMC free article] [PubMed]
6. Fabbri M, Garzon R, Cimmino A, Liu Z, Zanesi N, Callegari E, et al. MicroRNA-29 family reverts aberrant methylation in lung cancer by targeting DNA methyltransferases 3A and 3B. Proc Natl Acad Sci USA. 2007;104:15805–15810. [PMC free article] [PubMed]
7. Xiong Y, Fang JH, Yun JP, Yang J, Zhang Y, Jia WH, et al. Effects of MicroRNA-29 on apoptosis, tumorigenicity and prognosis of hepatocellular carcinoma. Hepatology. 2009;51:836–845. [PubMed]
8. Xu H, Cheung IY, Guo HF, Cheung NK. MicroRNA miR-29 modulates expression of immunoinhibitory molecule B7-H3: potential implications for immune based therapy of human solid tumors. Cancer Res. 2009;69:6275–6281. [PMC free article] [PubMed]
9. Cortez CC, Jones PA. Chromatin, cancer and drug therapies. Mutat Res. 2008;647:44–51. [PMC free article] [PubMed]
10. Pass HI, Goparaju C, Ivanov S, Donington J, Carbone M, Hoshen M, et al. hsa-miR-29c* is linked to the prognosis of malignant pleural mesothelioma. Cancer Res. 2010;70:1916–1924. [PMC free article] [PubMed]
11. Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ. Cancer statistics 2009. CA Cancer J Clin. 2009;59:225–249. [PubMed]
12. Baylin SB, Esteller M, Rountree MR, Bachman KE, Schuebel K, Herman JG. Aberrant patterns of DNA methylation, chromatin formation and gene expression in cancer. Hum Mol Genet. 2001;10:687–692. [PubMed]
13. de Maat MF, van de Velde CJ, Umetani N, de Heer P, Putter H, van Hoesel AQ, et al. Epigenetic silencing of cyclooxygenase-2 affects clinical outcome in gastric cancer. J Clin Oncol. 2007;25:4887–4894. [PubMed]
14. Tanemura A, Terando AM, Sim MS, van Hoesel AQ, de Maat MF, Morton DL, et al. CpG island methylator phenotype predicts progression of malignant melanoma. Clin Cancer Res. 2009;15:1801–1807. [PMC free article] [PubMed]
15. Gaudet MM, Campan M, Figueroa JD, Yang XR, Lissowska J, Peplonska B, et al. DNA hypermethylation of ESR1 and PGR in breast cancer: pathologic and epidemiologic associations. Cancer Epidemiol Biomarkers Prev. 2009;18:3036–3043. [PMC free article] [PubMed]
16. Glockner SC, Dhir M, Yi JM, McGarvey KE, Van Neste L, Louwagie J, et al. Methylation of TFPI2 in stool DNA: a potential novel biomarker for the detection of colorectal cancer. Cancer Res. 2009;69:4691–4699. [PMC free article] [PubMed]
17. Pfeifer GP, Yoon JH, Liu L, Tommasi S, Wilczynski SP, Dammann R. Methylation of the RASSF1A gene in human cancers. Biol Chem. 2002;383:907–914. [PubMed]
18. Spugnardi M, Tommasi S, Dammann R, Pfeifer GP, Hoon DS. Epigenetic inactivation of RAS association domain family protein 1 (RASSF1A) in malignant cutaneous melanoma. Cancer Res. 2003;63:1639–1643. [PubMed]
19. Mott JL, Kobayashi S, Bronk SF, Gores GJ. mir-29 regulates Mcl-1 protein expression and apoptosis. Oncogene. 2007;26:6133–6140. [PMC free article] [PubMed]
20. Park SY, Lee JH, Ha M, Nam JW, Kim VN. miR-29 miRNAs activate p53 by targeting p85alpha and CDC42. Nat Struct Mol Biol. 2009;16:23–29. [PubMed]
21. Bastian BC, LeBoit PE, Hamm H, Brocker EB, Pinkel D. Chromosomal gains and losses in primary cutaneous melanomas detected by comparative genomic hybridization. Cancer Res. 1998;58:2170–2175. [PubMed]
22. Okano M, Bell DW, Haber DA, Li E. DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development. Cell. 1999;99:247–257. [PubMed]
23. Kato Y, Kaneda M, Hata K, Kumaki K, Hisano M, Kohara Y, et al. Role of the Dnmt3 family in de novo methylation of imprinted and repetitive sequences during male germ cell development in the mouse. Hum Mol Genet. 2007;16:2272–2280. [PubMed]
24. Linhart HG, Lin H, Yamada Y, Moran E, Steine EJ, Gokhale S, et al. Dnmt3b promotes tumorigenesis in vivo by gene-specific de novo methylation and transcriptional silencing. Genes Dev. 2007;21:3110–3122. [PMC free article] [PubMed]
25. Mathe E. RASSF1A, the new guardian of mitosis. Nat Genet. 2004;36:117–118. [PubMed]
26. Dallol A, Agathanggelou A, Tommasi S, Pfeifer GP, Maher ER, Latif F. Involvement of the RASSF1A tumor suppressor gene in controlling cell migration. Cancer Res. 2005;65:7653–7659. [PubMed]
27. Whitehurst AW, Ram R, Shivakumar L, Gao B, Minna JD, White MA. The RASSF1A tumor suppressor restrains anaphase-promoting complex/cyclosome activity during the G1/S phase transition to promote cell cycle progression in human epithelial cells. Mol Cell Biol. 2008;28:3190–3197. [PMC free article] [PubMed]
28. Girault I, Tozlu S, Lidereau R, Bieche I. Expression analysis of DNA methyltransferases 1, 3A and 3B in sporadic breast carcinomas. Clin Cancer Res. 2003;9:4415–4422. [PubMed]
29. Segura MF, Belitskaya-Levy I, Rose AE, Zakrzewski J, Gaziel A, Hanniford D, et al. Melanoma MicroRNA signature predicts post-recurrence survival. Clin Cancer Res. 2010;16:1577–1586. [PubMed]
30. Satzger I, Mattern A, Kuettler U, Weinspach D, Voelker B, Kapp A, et al. MicroRNA-15b represents an independent prognostic parameter and is correlated with tumor cell proliferation and apoptosis in malignant melanoma. Int J Cancer. 2010;126:2553–2562. [PubMed]
31. Camp RL, Neumeister V, Rimm DL. A decade of tissue microarrays: progress in the discovery and validation of cancer biomarkers. J Clin Oncol. 2008;26:5630–5637. [PubMed]
32. Goto Y, Ferrone S, Arigami T, Kitago M, Tanemura A, Sunami E, et al. Human high molecular weight-melanoma-associated antigen: utility for detection of metastatic melanoma in sentinel lymph nodes. Clin Cancer Res. 2008;14:3401–3407. [PMC free article] [PubMed]
33. Takeuchi H, Morton DL, Kuo C, Turner RR, Elashoff D, Elashoff R, et al. Prognostic significance of molecular upstaging of paraffin-embedded sentinel lymph nodes in melanoma patients. J Clin Oncol. 2004;22:2671–2680. [PMC free article] [PubMed]

Articles from Epigenetics are provided here courtesy of Landes Bioscience
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...