SAN FRANCISCO (GenomeWeb) – At Cambridge Healthtech Institute's Molecular Medicine Tri-Conference here this week, a pair of Boston-based researchers presented novel approaches for the detection of circulating nucleic acid biomarkers for potential use in cancer diagnosis and disease monitoring.
The scientists — Chen Song, a research fellow with Dana-Farber Cancer Institute and Harvard Medical School, and Leonora Balaj, a research fellow at Massachusetts General Hospital and Harvard Medical School — discussed their methods in the "PCR for Molecular Medicine" track on Monday.
In her presentation, Song discussed the use of a new mutation enrichment technology developed by her lab called nuclease-assisted mutation enrichment using probe overlap (NaME-Pro) to greatly enhance the sensitivity of various downstream assays to detect circulating cell-free tumor DNA.
The key to the NaME-Pro method is a duplex-specific nuclease extracted from the hepatopancreas, an organ of the digestive tract of various arthropods, mollusks, and fish. The enzyme has a strong cleavage preference for wild-type DNA duplexes, and as such, Song and her colleagues can create partially overlapping probes, apply them along with the enzyme to a mixture of wild-type and mutant DNA to selectively digest the wild-type, then perform PCR to enrich the mutant DNA.
As proof of principle, the group performed single-target NaME-Pro on a KRAS exon 2 mutation, then used digital PCR to verify the level of enrichment they achieved. They were able to enrich the mutant DNA to about 80 percent mutation abundance from a starting point of 0.5 percent. Serial dilution experiments revealed that the technique's detection sensitivity on its own hovers in the 0.01 percent to 0.1 percent range, depending on the mutant type. The absolute technical detection limit of the technique, when combined with PCR and Sanger sequencing, is one mutant DNA copy in about 3 million wild-type copies.
One of the major advantages of the technique, according to Song, is that it can be applied to original genomic material in front of a number of other molecular detection techniques such as high-resolution melt analysis, ICE-COLD PCR, or Sanger sequencing to greatly enhance detection sensitivity. For instance, she noted that the group conducted a proof-of-principle experiment where they attempted to use Sanger sequencing to pick up extremely low-level mutations in MEK1, JAK2, and KRAS. On its own, Sanger sequencing was not able to detect the mutations, but easily did so when combined with NaME-Pro.
Song and her colleagues also successfully applied the method to a Thermo Fisher Scientific Ion Torrent AmpliSeq cancer panel targeted sequencing workflow, greatly enhancing its ability to detect specific BRAF mutations from genomic DNA.
Finally, Song noted that the technique was able to detect three different hot spot mutations at different positions in the KRAS gene; showed the potential for mutation scanning of TP53 exons 6-9; and can be multiplexed for up to 50 targets at once. On this latter point, Song said that her group is now embarking on experiments to test the method for massively parallel enrichment for hundreds or thousands of targets.
Balaj, meanwhile, discussed her group's efforts to use digital PCR to increase the detection rate of rare cancer mutations in exosome-derived nucleic acids from the blood of brain cancer patients.
Balaj's lab is conducting the work in the context of glioblastoma multiforme, which is the most common malignant brain tumor and has extremely poor prognosis and survival, with four subtypes that differ in their mutational status.
She and her colleagues decided to look to exosomes, or what they more generally call extracellular vesicles (EVs), as sources for tumor-derived nucleic acids. EVs are extremely attractive for mutational analysis of these nucleic acids because they are secreted by all living cells into all biofluids, and are known to be involved in a number of cancerous pathways such as angiogenesis, tumor growth, blunt immune response, and intra-cellular communication. Furthermore, EVs contain a plethora of biomarkers including messenger RNA, microRNA, non-coding RNA, various proteins, and often times DNA, either inside them or embedded on their surfaces. In addition, EVs are able to preserve the integrity of the biomarkers they contain over long periods of time, Balaj said.
Balaj and her colleagues have been specifically examining mutations in the IDH1 gene, which are found in about 80 percent of gliomas and secondary glioblastomas, and are the subject of many phase II and phase III clinical trials for therapeutics. In general, the mutations have been suggested as an independent marker of improved overall survival.
When they began examining these point mutations a few years ago, they used BEAMing technology, which was developed by Bert Vogelstein and his colleagues at Johns Hopkins University and involves the creation of numerous microdroplets containing nucleic acid targets, PCR amplification, and FACS analysis for detection and enumeration.
However, as has been widely noted, the BEAMing technique is relatively cumbersome to use compared to newer, more automated digital PCR methods. Balaj and her colleagues initially explored RainDance droplet-based digital PCR as an alternative method, and found that the technique worked very well. However, this week at Tri-Con, Balaj noted that the group has made the switch to Bio-Rad's Droplet Digital PCR platform for several reasons, including the fact that it is now a less expensive and widely available commercial platform.
Balaj presented early data from a study of eight glioblastoma patient plasma samples in which IDH1 mRNA and IDH1 DNA did not correlate. In collaboration with Exosome Diagnostics, which has a proprietary platform for isolating EV-derived nucleic acids, the researchers were able to use the Droplet Digital PCR approach to detect IDH1 mutations in both exosomal RNA and cell-free DNA in plasma samples with a high degree of sensitivity.
Another study that Balaj and her colleagues have recently undertaken is a retrospective analysis of EGFR amplifications in patients enrolled in a specific clinical trial because they had these amplifications and had failed standard therapy. The researchers have collected serum samples from every month of treatment until the point of treatment failure, and will now use their approach to measure levels of wild-type EGFR, specific known EGFR mutations, and perhaps other novel mutations.
In collaboration with the lab of Xandra Breakefield at MGH, they will examine RNA and DNA in circulating EVs, and then will combine this work with diagnostic NMR for protein analysis with the laboratory of Ralph Weissleder at the MGH Center for Systems Biology, Balaj said.