Sunday, April 27, 2014

DOK7 identified as potential Epigenetic indicator of Breast Cancer

        Lauren and I were intrigued by the role of epigenetics in cancerous growth and chose to further investigate this correlation in our Cancer Project, so imagine my excitement after an epigenetic cancer study was introduced in my general psychology class!

This studyconducted by Bellvitge Biomedical Research Institute; University of Coimbra Centre for Neurosciences and Cell Biology; Kings College London Department of Twin Research and Genetic Epidemiology; University of Iceland Molecular and Cell Biology Research Laboratory; and University of Barcelona School of Medicine, looked at 15 monozygotic (MZ) twin pairs discordant for breast cancer, with further examination into the different degrees of methylation affecting gene rich segments within the cancerous cells. In other words, despite the genetic equivalence within each twin pair, an extraneous variable, presumably differences in epigenetics, resulted in one twin exhibiting breast cancer while the other was cancer free. By comparing isolated DNA from blood and cell samples obtained from the cancer positive twin with that of the cancer negative twin, this study isolated 403 differentially methylated CpG sites, or Cytosine-Guanine DNA regions, many of which contain specific genes referred to as “novel breast cancer genes,” genes highly correlated with breast cancer incidence.


After each blood and tumor/normal tissue sample was macroscopically examined, DNA regions were isolated from the whole blood and triple negative breast tumors of the cancer positive twins and normal breast tissue of the cancer negative twins (n=15). The DNA samples were then randomly distributed into a 96-well plate to eliminate experimenter bias, after which, an EZ DNA methylation kit was used to determine methylation degrees of specific regions per sample. DNA methylation levels were normalized, then presented in decimal form ranging from 0 to 1 and referred to as the Beta-value, Beta-values falling outside the significance value (p > 0.01) were thereby eliminated from the sample. Following Beta-value attribution, a Wilcoxon rank sum paired test was performed to identify consistently differentially methylated sites between each cancer positive and cancer negative twin pair with a p value < 0.05. This test identified the statistically significant 403 differentially methylated sites, demonstrating data much in line with the hypothesis that epigenetic regulation plays a role in cancer development.

Figure 1. Differentially methylated CpG sites within MZ twin pairs discordant for breast cancer. (E) bcDMP varying epigenetically despite the identical genetic background of the twins identified by multivariate filter analysis. The bcDMP DNA methylation level is color coded (yellow: sample with lowest methylation level; red: sample with highest methylation level.) (Heyn, Holger, et al. "DNA methylation profiling in breast cancer discordant identical twins identifies DOK7 as novel epigenetic biomarker." Carcinogenesis (2012): 102-08.)
Epigenetic regulation of genes and corresponding CpG sites accounts for activation and suppression of functionality and transcription via alterations in DNA methylation, offering more insight into controllable influencers of cancer. Thus, much research has focused more recently on methylation patterns in these novel breast cancer genes, as apposed to predetermined genetic susceptibility, with the aim of determining potential epigenetic indicators of Breast Cancer. This study aimed to narrow down the original 403 differentially methylated CpG sites to a handful of consistent regions known as bcDMPs, or “differentially methylated CpG positions in breast cancer.” Regions identified as bcDMPs, hold the promise of serving as potential epigenetic indicators of breast cancer, in both early and late states of the disease, and eventually helping to better tailor the understanding of effective treatment and pathology.

 So, following the first round of results, the study conducted a genome-wide DNA methylation technique to analyze whole blood samples from an additional sample of monozygotic twin pairs (n=21), again discordant for breast cancer. However, this time a new variable was introduced: point of diagnosis, 16 twin pairs incorporated a cancer twin that had already been diagnosed with breast cancer, the other 5 incorporated a twin suspected of breast cancer but not yet formally diagnosed. Using locus-specific pyrosequencing for gene alteration, 14 CpG sites were isolated as potential bcDMPs. 

             One of these 14 CpG sites, Docking Protein 7 (DOK7) a known receptor tyrosine kinase activator, presented statistically significant methylation variation between the cancer positive and cancer negative twin after completing the Wilcoxon signed rank test. The cancer twin exhibited hypermethylation of the DOK7 region, a finding consistent across all the twin pairs, with exactly four additional CpG sites (directly upstream of DOK7) also presenting clear patterns of significant hypermethylation. The DOK7 regions and same additional four CpG sites were isolated in six healthy breast samples; interestingly enough, these cancer-free cells displayed clear hypomethylation, that is, decreased levels of methylation. This whole 5-site loci, DOK7 plus the additional 4 CpG sites, was thus confirmed to be a bcDMP.
Figure 2. DOK7 is hypermethylated in different breast cancer contexts. (B) Intra-pair difference (cancer-healthy) of the DOK7 associated bcDMP in eight twins (identification set) postdiagnosis assessed by pyro-sequencing. (C) Intra-pair difference (cancer-healthy) of the DOK7 associated bcDmP in 16 twins (validation set) postdiagnosis assessed by pyrosequencing. 2 (Heyn, Holger, et al. "DNA methylation profiling in breast cancer discordant identical twins identifies DOK7 as novel epigenetic biomarker." Carcinogenesis (2012): 102-08.)
But what about the yielded differences in pre versus post diagnosis cancer twins? In fact, three of the five suspected cancer twins displayed blood and tissue samples with a significant gain of DOK7 methylation, and these samples were taken prior to actual diagnosis, with a mean of 4.7 years before concrete diagnosis. In other words, DOK7 displayed significant enough methylation to be detected as a breast cancer indicator in patients about 4.7 years before their cancer was symptomatic and fully diagnosable. While this represents a very small sample size, such visible DOK7 methylation prior to actual diagnosis of breast cancer presents a very promising avenue of early detection.

This is why this study, though not fully perfect, is incredibly exciting. This study has first identified 403 differentially methylated regions within important DNA regions, highlighting a significant epigenetic difference in cancer positive and cancer negative twins, independent of genetic influencers. Furthermore, these 403 regions were narrowed down to 14 possible bcDMPs, and finally 1 confirmed bcDNP, DOK7. DOK7 represents one of many possible bcDMPs that not only demonstrates significant methylation differences between tumor samples and normal tissues post diagnosis, but also unlocks the possibility for pre-symptomatic detection, epigenetic monitoring of disease development, and eventually epigenetic treatment of cancer!

Source Citation
Heyn, Holger, et al. "DNA methylation profiling in breast cancer discordant identical twins identifies
           DOK7 as novel epigenetic biomarker." Carcinogenesis (2012): 102-08.