Limits...
Identification of epigenetically regulated genes that predict patient outcome in neuroblastoma.

Carén H, Djos A, Nethander M, Sjöberg RM, Kogner P, Enström C, Nilsson S, Martinsson T - BMC Cancer (2011)

Bottom Line: Expression was analyzed using whole-genome expression arrays to identify genes activated by the treatment.Differential methylation was observed for the genes SCNN1A (p < 0.001), PRKCDBP (p < 0.001) and KRT19 (p < 0.01).Among these, the mRNA expression of KRT19 and PRKCDBP was significantly lower in patients that have died from the disease compared with patients with no evidence of disease (fold change -8.3, p = 0.01 for KRT19 and fold change -2.4, p = 0.04 for PRKCDBP).

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Clinical Genetics, Institute of Biomedicine, University of Gothenburg, Sahlgrenska University Hospital, SE-41345 Gothenburg, Sweden. helena.caren@clingen.gu.se

ABSTRACT

Background: Epigenetic mechanisms such as DNA methylation and histone modifications are important regulators of gene expression and are frequently involved in silencing tumor suppressor genes.

Methods: In order to identify genes that are epigenetically regulated in neuroblastoma tumors, we treated four neuroblastoma cell lines with the demethylating agent 5-Aza-2'-deoxycytidine (5-Aza-dC) either separately or in conjunction with the histone deacetylase inhibitor trichostatin A (TSA). Expression was analyzed using whole-genome expression arrays to identify genes activated by the treatment. These data were then combined with data from genome-wide DNA methylation arrays to identify candidate genes silenced in neuroblastoma due to DNA methylation.

Results: We present eight genes (KRT19, PRKCDBP, SCNN1A, POU2F2, TGFBI, COL1A2, DHRS3 and DUSP23) that are methylated in neuroblastoma, most of them not previously reported as such, some of which also distinguish between biological subsets of neuroblastoma tumors. Differential methylation was observed for the genes SCNN1A (p < 0.001), PRKCDBP (p < 0.001) and KRT19 (p < 0.01). Among these, the mRNA expression of KRT19 and PRKCDBP was significantly lower in patients that have died from the disease compared with patients with no evidence of disease (fold change -8.3, p = 0.01 for KRT19 and fold change -2.4, p = 0.04 for PRKCDBP).

Conclusions: In our study, a low methylation frequency of SCNN1A, PRKCDBP and KRT19 is significantly associated with favorable outcome in neuroblastoma. It is likely that analysis of specific DNA methylation will be one of several methods in future patient therapy stratification protocols for treatment of childhood neuroblastomas.

Show MeSH

Related in: MedlinePlus

Differential methylation analysis. (A) Analysis of differences in methylation beta values from patients with no evidence of disease (NED) compared to those with an adverse outcome (DOD) for the most statistically significant genes. One of the CpG sites from the Illumina methylation array is shown. Box plot explanation; upper and lower hinges of the box represent the 75th percentile and 25th percentile respectively; whiskers indicate the highest and lowest values that are not outliers or extreme values; thick horizontal line within box, median. Open circles represent outliers and asterisks represent extremes. The p-value at gene-by-gene level is indicated in the left lower corner in each graph. (B) Analysis of methylation frequencies for tumors with different chromosomal profiles. For definition of chromosomal groups, see Carén et al [44]. CpG sites for SCNN1A and DUSP23 are shown. The highest methylation frequencies in SCNN1A are found in NB tumors with an unfavorable chromosomal profile. Methylation of DUSP23 is almost mutually exclusive in MYCN-amplified NBs.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3045360&req=5

Figure 3: Differential methylation analysis. (A) Analysis of differences in methylation beta values from patients with no evidence of disease (NED) compared to those with an adverse outcome (DOD) for the most statistically significant genes. One of the CpG sites from the Illumina methylation array is shown. Box plot explanation; upper and lower hinges of the box represent the 75th percentile and 25th percentile respectively; whiskers indicate the highest and lowest values that are not outliers or extreme values; thick horizontal line within box, median. Open circles represent outliers and asterisks represent extremes. The p-value at gene-by-gene level is indicated in the left lower corner in each graph. (B) Analysis of methylation frequencies for tumors with different chromosomal profiles. For definition of chromosomal groups, see Carén et al [44]. CpG sites for SCNN1A and DUSP23 are shown. The highest methylation frequencies in SCNN1A are found in NB tumors with an unfavorable chromosomal profile. Methylation of DUSP23 is almost mutually exclusive in MYCN-amplified NBs.

Mentions: In order to see whether the methylation also varied among biologically different subgroups of NB, we compared the methylation beta values from the Illumina arrays with different patient variables. We compared patients who are alive with no evidence of disease (NED) five years after diagnosis (5-year overall survival; OS) with those who were dead of disease (DOD), known prognostic chromosomal aberrations such as 1p deletion, MYCN amplification, 11q deletion and 17q gain, age at diagnosis, as well as chromosomal profiles obtained from array copy number data analysis [2]. Differential methylation based on 5-year OS was seen for the genes SCNN1A, PRKCDBP and KRT19 with a higher methylation frequency found in tumors from patients with an unfavorable outcome, see Figure 3A and Table 2. For SCNN1A, PRKCDBP, KRT19, TGFBI and DUSP23, significantly higher methylation frequencies found in MYCN-amplified tumors, see Figure 3B and Table 3.


Identification of epigenetically regulated genes that predict patient outcome in neuroblastoma.

Carén H, Djos A, Nethander M, Sjöberg RM, Kogner P, Enström C, Nilsson S, Martinsson T - BMC Cancer (2011)

Differential methylation analysis. (A) Analysis of differences in methylation beta values from patients with no evidence of disease (NED) compared to those with an adverse outcome (DOD) for the most statistically significant genes. One of the CpG sites from the Illumina methylation array is shown. Box plot explanation; upper and lower hinges of the box represent the 75th percentile and 25th percentile respectively; whiskers indicate the highest and lowest values that are not outliers or extreme values; thick horizontal line within box, median. Open circles represent outliers and asterisks represent extremes. The p-value at gene-by-gene level is indicated in the left lower corner in each graph. (B) Analysis of methylation frequencies for tumors with different chromosomal profiles. For definition of chromosomal groups, see Carén et al [44]. CpG sites for SCNN1A and DUSP23 are shown. The highest methylation frequencies in SCNN1A are found in NB tumors with an unfavorable chromosomal profile. Methylation of DUSP23 is almost mutually exclusive in MYCN-amplified NBs.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3045360&req=5

Figure 3: Differential methylation analysis. (A) Analysis of differences in methylation beta values from patients with no evidence of disease (NED) compared to those with an adverse outcome (DOD) for the most statistically significant genes. One of the CpG sites from the Illumina methylation array is shown. Box plot explanation; upper and lower hinges of the box represent the 75th percentile and 25th percentile respectively; whiskers indicate the highest and lowest values that are not outliers or extreme values; thick horizontal line within box, median. Open circles represent outliers and asterisks represent extremes. The p-value at gene-by-gene level is indicated in the left lower corner in each graph. (B) Analysis of methylation frequencies for tumors with different chromosomal profiles. For definition of chromosomal groups, see Carén et al [44]. CpG sites for SCNN1A and DUSP23 are shown. The highest methylation frequencies in SCNN1A are found in NB tumors with an unfavorable chromosomal profile. Methylation of DUSP23 is almost mutually exclusive in MYCN-amplified NBs.
Mentions: In order to see whether the methylation also varied among biologically different subgroups of NB, we compared the methylation beta values from the Illumina arrays with different patient variables. We compared patients who are alive with no evidence of disease (NED) five years after diagnosis (5-year overall survival; OS) with those who were dead of disease (DOD), known prognostic chromosomal aberrations such as 1p deletion, MYCN amplification, 11q deletion and 17q gain, age at diagnosis, as well as chromosomal profiles obtained from array copy number data analysis [2]. Differential methylation based on 5-year OS was seen for the genes SCNN1A, PRKCDBP and KRT19 with a higher methylation frequency found in tumors from patients with an unfavorable outcome, see Figure 3A and Table 2. For SCNN1A, PRKCDBP, KRT19, TGFBI and DUSP23, significantly higher methylation frequencies found in MYCN-amplified tumors, see Figure 3B and Table 3.

Bottom Line: Expression was analyzed using whole-genome expression arrays to identify genes activated by the treatment.Differential methylation was observed for the genes SCNN1A (p < 0.001), PRKCDBP (p < 0.001) and KRT19 (p < 0.01).Among these, the mRNA expression of KRT19 and PRKCDBP was significantly lower in patients that have died from the disease compared with patients with no evidence of disease (fold change -8.3, p = 0.01 for KRT19 and fold change -2.4, p = 0.04 for PRKCDBP).

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Clinical Genetics, Institute of Biomedicine, University of Gothenburg, Sahlgrenska University Hospital, SE-41345 Gothenburg, Sweden. helena.caren@clingen.gu.se

ABSTRACT

Background: Epigenetic mechanisms such as DNA methylation and histone modifications are important regulators of gene expression and are frequently involved in silencing tumor suppressor genes.

Methods: In order to identify genes that are epigenetically regulated in neuroblastoma tumors, we treated four neuroblastoma cell lines with the demethylating agent 5-Aza-2'-deoxycytidine (5-Aza-dC) either separately or in conjunction with the histone deacetylase inhibitor trichostatin A (TSA). Expression was analyzed using whole-genome expression arrays to identify genes activated by the treatment. These data were then combined with data from genome-wide DNA methylation arrays to identify candidate genes silenced in neuroblastoma due to DNA methylation.

Results: We present eight genes (KRT19, PRKCDBP, SCNN1A, POU2F2, TGFBI, COL1A2, DHRS3 and DUSP23) that are methylated in neuroblastoma, most of them not previously reported as such, some of which also distinguish between biological subsets of neuroblastoma tumors. Differential methylation was observed for the genes SCNN1A (p < 0.001), PRKCDBP (p < 0.001) and KRT19 (p < 0.01). Among these, the mRNA expression of KRT19 and PRKCDBP was significantly lower in patients that have died from the disease compared with patients with no evidence of disease (fold change -8.3, p = 0.01 for KRT19 and fold change -2.4, p = 0.04 for PRKCDBP).

Conclusions: In our study, a low methylation frequency of SCNN1A, PRKCDBP and KRT19 is significantly associated with favorable outcome in neuroblastoma. It is likely that analysis of specific DNA methylation will be one of several methods in future patient therapy stratification protocols for treatment of childhood neuroblastomas.

Show MeSH
Related in: MedlinePlus