Methylation and Detection

Bisulfite sequencing is a very simple, accurate method of quantifying cytosine methylation in DNA. Bisulfite treatment converts unmethylated cytosine residues into uracil and leaves methylated cytosines intact. This treated DNA is then amplified by PCR and sequenced. Only methylated cytosines in the initial sequence remain as Cs; the other unmethylated C sites appear as Ts. All remaining cytosine residues indicate a methylated cytosine in the original DNA template.

Cancer and Tumor Suppression

Investigators have begun looking for an effective method of reactivating tumor suppressor genes by inhibiting DNA methyltransferase. To researchers interested in identifying those most at risk for developing cancer, DNA methylation offers some tantalizing clues. Because methylation is very stable, methylated diagnostic markers are apparent well before cells begin to metastasize.

In theory, by identifying changes early, patients can begin treatment sooner, and the treatment with the greatest chance of success can be chosen.

The Role of BEX1 and BEX2 in Gliomas

Anup Madan, Ph.D., Gregory Foltz, M.D., and their colleagues at the Swedish Medical Centre and Institute for Systems Biology in Seattle are investigating the role that the BEX1 and BEX2 genes play as tumor suppressors in gliomas, one of the most common and most aggressive brain tumors, and tumors with a great deal of heterogeneity. Some success has already been demonstrated. “In our studies, the two novel brain-expressed genes BEX1 and BEX2 were silenced in all tumor specimens and exhibited extensive promoter hypermethylation,” Dr. Madan states. “Viral-mediated re-expression of these genes led to a heightened sensitivity to chemotherapy-induced apoptosis and potent tumor suppressor effects in vitro and in a xenograft mouse model.”

“it is important to discover which genes have an impact therapeutically. If we can reverse the methylation process, we have an opportunity to reactivate a gene. To date, we have identified a gene that was silenced and were able to reactivate it, which caused the tumor cells to die.”

Gregory Foltz, M.D., member of the neurosurgery faculty at the University of Washington, a member of the Seattle Neuroscience Institute at Swedish Hospital, and a practicing neurosurgeon in Seattle, specializing in neuro-oncology

Identifying Additional Tumor Suppressor Genes

In addition to BEX1 and BEX2, Dr. Madan, Dr. Foltz, and their colleagues identified various other novel genes that inhibit tumor suppression. The team began its investigation by performing genomewide screening using RNA expression profiling on the Applied Biosystems 1700 Chemiluminescent Microarray Analyzer and validating hits with TaqMan^® Gene Expression Assays to examine expression differences in tumor and normal cells.

After identifying 400–500 likely candidates, they then sequenced the promoter regions on the 3730xl DNA Analyzer to determine which genes exhibit hypermethylation. Suspect genes typically exhibit CpG-rich islands. A smaller group of 20–30 tumor suppressor genes was eventually selected, from which the five most promising were identified for further investigation. Some of the new genes identified by this process have never been fully characterized and are only now being investigated for their role in functional pathways.

Keys to Success—Accurate Primer Design and Bisulfite Conversion

Accurate primer design and complete bisulfite conversion are crucial to their research. Because the PCR process translates unmethylated Cs to Ts, Dr. Madan’s laboratory has written software to analyze high throughput bisulfite sequence data. The software relies on PHRED/PHRAP output files and calculates the percentage of C/T positions at various CpG sites. “This process provides a powerful new method for the identification of epigenetically silenced genes with potential function as tumor suppressors and biomarkers for disease diagnosis and detection,” Dr. Madan says.

As a clinician, Dr. Foltz finds that “it is important to discover which genes have an impact therapeutically. If we can reverse the methylation process, we have an opportunity to reactivate a gene. To date, we have identified a gene that was silenced and were able to reactivate it, which caused the tumor cells to die.”

CpG islands in different methylated states — if methylated, the expression of gene related to this CpG island is turned off.

Additionally, in a recent major advance, promoter hypermethylation and epigenetic silencing of the DNA repair gene O⁶ - methylguanine-DNA methyltransferase were used to identify patients with markedly improved survival at two years in response to combined treatment with chemotherapy and radiation treatment. If these studies can be replicated with similar success rates, they will provide oncologists with an important new therapeutic option.

Visit here and click on “methylation analysis” for more information on DNA methylation.

Dr. Foltz’s and Dr. Madan’s research on the BEX1 and BEX2 genes can be found here.

Anup Madan, Ph. D (left), and Gregory Foltz, M.D. (right)

Anup Madan, Ph.D and Gregory Foltz, M.D. currently direct the Neurogenomics Research Laboratory at the Seattle Neuroscience Institute at Swedish Hospital in Seattle. Greg Foltz is also a practicing neurosurgeon and is a principal founder of Neurosurgery International, a nonprofit educational organization for young neurosurgeons from developing countries. Both of them previously held the title of Assistant Professor in the departments of Neurosurgery and Neurology at the University of Iowa, College of Medicine. The Neurogenomics Research Laboratory (NGRL) is dedicated to the application of systems biology approaches to the molecular and genetic analysis of human brain function. In this regard, the laboratory is actively developing new technologies and computational tools to understand the role of epigenetically regulated genes or pathways in Glioblastoma multiforme (GBM).

Genomics in Cancer Research
The recent wave of genomics technologies has provided unprecedented advancements in the molecular classification of cancer. However, while it is clear that individual genomic differences often indicate a modified biology, it is also clear that large numbers of genes interact to cause the cellular changes that result in cancer.

Genomic cancer researchers are looking for a method of clinically classifying tumors that will enable physicians to choose the most appropriate treatment for their patients. Next-generation genomics tools can be a welcome augmentation to their arsenal.

Cancer of the Unknown Primary

One application of genomics technologies that could have an immediate clinical impact is the identification of tumors that are difficult to diagnose. Cancers of the unknown primary (CUP) are defined as malignant tumors for which the tissue of origin is unknown. While they represent approximately 5% of cancer diagnoses, they are among the top 10 causes of cancer death. Clinicians and pathologists routinely employ a battery of diagnostic methods in an attempt to locate the primary site, as this is important in treatment selection. However, in the majority of cases, the origin is not revealed. A more accurate diagnosis of CUP may lessen patient suffering, improve outcome, and reduce the time and cost of treatment.

Applying Genomics

Malignant tumors have been reported to maintain a gene expression profile characteristic of the site of origin. Thus, creating a genomic database of known tumor types for the purpose of determining their primary sites is feasible. However, the challenge is demonstrating clinical utility and developing an accurate test for pathologists to use. To address this, researchers at the Peter MacCallum Cancer Centre in Melbourne, Australia, began by profiling an extensive series of cancer specimens (of known origin) using whole genome gene expression microarrays. This was followed with validation on a subset of highly informative gene markers using TaqMan^® Low Density Arrays. The study generated a large training set from which a prediction model was created for identifying metastatic cancers of unknown origin.

While microarrays are powerful discovery platforms, they are not particularly suited for translation to a clinical test. The Melbourne group found that real-time PCR offered a number of benefits that makes it a better platform for the development of a clinical laboratory test. It has a broad dynamic range with good sensitivity at low copy levels and a turnaround time in hours. The technology uses less RNA than conventional microarrays, making it amenable to biopsy sampling, and functional data can be recovered from degraded samples such as formalin fixed paraffin embedded (FFPE) tissues.

Current studies involve screening more than 500 clinical tumor samples with 768 TaqMan^® assays to refine the classifier down to the most indicative markers. Extracting sufficient RNA from small biopsies and archival fixed specimens has been the largest obstacle faced.

Pilot tests using the TaqMan^® PreAmp Master Mix on non-preamplified and preamplified material indicated that although some minor systematic biases are introduced for particular genes, the overall correlation between nonpreamplified and preamplified specimens was high. According to Richard Tothill Ph.D., who led the study, “Preamplification is an attractive method. We see significant advantages for limited pathology specimens, and it’s been critical in developing profiles across our gene set.”

Future Directions

One obstacle to the development of an unknown primary classifier has been obtaining enough well-characterized tumor samples for training and validating the classifier. The Melbourne group has formed important collaborations with cancer and pathology centers in Australia and abroad to attain sufficient samples for validation. Genomic tools, such as an unknown primary diagnostic, must be tested in the clinical arena and benchmarked against histopathology and other existing methods. This will take time, but an important step has been taken in cancer genomics research.

Dr. Richard Tothill, Centre for Cancer Genomics and Predictive Medicine, Peter MacCallum Cancer Centre, Smorgon Family Building, St. Andrews Place, East Melbourne, Victoria, Australia.

Back to top