Using the methylome to identify aggressive Barrett’s esophagus
360
Hypermethylation
G.I. and Other Associated Cancers Research Group
1
OVERALL STRATEGY: Our strategy will consist of using HumanMethylation450 arrays to identify methylation profiles and/or candidate methylated genes that distinguish BE from BE+LGD, BE+HGD and EAC (Aim 1). We will then assess whether these genes are predictive markers for aggressive BE (Aim 2)
Aim 1: To characterize the genome-wide methylation status of various histological subtypes of
esophageal tissues:
Aim 1a: To determine the methylome of BE, BE+LGD, BE+HGD, and EAC.
Aim 1b: To determine if a panel of methylated genes can be used to distinguish BE,
BE+LGD, BE+HGD, and EAC.
1) 1st quarter
1. Request tissue samples from EDRN biorepository
2. Obtain tissues from biorepository and assess quality
3. Extract DNA from tissue samples and treat with sodium bisulfite
2) 2nd quarter
1. Run HumanMethylation450 arrays using tissue samples
2. Analyze data and identify methylated genes that are specific for BE, BE+LGD, BE+HGD, and
EAC
3) 3rd quarter
1. Develop pyrosequencing assays and/or MethyLight assays for candidate genes identified
from array studies
2. Organize DNA samples in Grady lab tissue collection and bisulfite treat DNA samples
3. Assess methylation state of candidate genes in independent tissue set available in Grady lab
4. Identify methylated genes that show highest specificity for BE, BE+LGD, BE+HGD, and EAC
4) 4th quarter
1. Continue to assess methylation state of candidate genes in independent tissue set available
in Grady lab
2. Continue to identify methylated genes that show highest specificity for BE, BE+LGD,
BE+HGD, and EAC
3. Determine best methylated gene(s) to be used to identify aggressive BE vs. indolent BE
Aim 2: To conduct a pilot study to determine if methylated genes can be used to identify
aggressive BE vs. indolent BE.
1) 1st quarter
1. Arrange to obtain tissue samples from EDRN for initial analysis of methylated gene
biomarkers. This will be run on samples from 33 patients
2. Begin enrollment of study subjects and collection of tissues from these additional study
subjects enrolled in EDRN GLNE study
2) 2nd quarter
1. Continue enrollment and tissue collection of samples from EDRN GLNE study
2. Perform histologic quality assessment of tissues obtained from the initial patient cohort
(N=33)
2
3. Begin DNA extraction and bisulfite treatment of DNA samples
3) 3rd quarter
1. Run HumanMethylation450 arrays using tissue samples from initial patient cohort
2. Analyze data and identify methylated genes that are specific for aggressive BE vs. indolent
BE
3. Develop pyrosequencing assays and/or MethyLight assays for candidate genes identified
from array studies
4. Run pyrosequencing assays and/or Methylight assays on DNA samples from array studies to
validate array results
5. Identify methylated genes that have the highest specificity for aggressive BE vs. indolent BE
4) 4th quarter
1. Analyze samples from initial patient cohort using best candidate genes for identifying
aggressive BE vs. indolent BE
2. Obtain tissue samples from additional patients in EDRN GLNE study, assess quality of
samples, and extract DNA
3. Bisulfite treatment of DNA from tissue samples from 2nd patient cohort.
4. Assess best candidate genes in 2nd patient cohort to determine if results can be valid
Statistical Analysis: Dr. Chao-Jen Wong will direct the statistical analysis. We will assess genomewide
methylation patterns using HumanMethylation450 arrays in matched aggressive BE and indolent
BE cases. We will subsequently identify methylated gene profiles that are significantly methylated (ß
value >0.5) in tissues from both these groups using the selection criteria as described in Aim 1.
Next, we will compare candidate methylated genes identified in Aim 1 between the aggressive BE
and indolent BE groups using two-sided two sample t-test to identify transcripts deregulated specifically
in the aggressive BE population. Methylated genes will then be prioritized based on their consistent
methylation pattern in BE, HGD or EAC lesions in Aim 1. Finally, those methylated genes will be
considered as candidate biomarkers associated with aggressive BE, and will be subsequently validated
in an independent cohort of patients in future studies.
3) Power Analysis: Given the small initial sample size, we will initially assess for trends indicating a
methylated gene is a potential aggressive BE biomarker. In year 2, Dr Wong will conduct a power
analysis when we know how many people with aggressive vs. indolent BE have been collected.
There are currently no biomarkers annotated for this protocol.
No datasets are currently associated with this protocol.
