Periodic Reporting for period 1 - EpiDetect (Detecting epigenetic biomarkers in the blood for non-invasive precision oncology)
Reporting period: 2022-09-01 to 2025-02-28
The EpiDetect project aims at developing more universal non-invasive cancer detection using epigenetic markers. The project is organized in 3 interconnected aims comprising
AIM1- the use of DNA methylation alterations that can be detected from cell-free DNA (cfDNA)
AIM2- cfDNA fragmentation signatures originating from cancer cells and which indicate the cell-of- origin
AIM3- the detection of circulating transcripts using extracellular vesicles taking advantage of their protective function to better detect RNA molecules
AIM1- the use of DNA methylation alterations that can be detected from cell-free DNA (cfDNA)
AIM2- cfDNA fragmentation signatures originating from cancer cells and which indicate the cell-of- origin
AIM3- the detection of circulating transcripts using extracellular vesicles taking advantage of their protective function to better detect RNA molecules
In these 3 aims, we started by focusing on the signatures associated to the reactivation of retrotransposons which occurs in multiple forms of cancer and which should provide universal markers to cover the heterogeneous profiles of cancer patients and detect multiple types of cancer.
During the last 2 years, we were able to demonstrate that indeed targeting hypomethylation of evolutionary young LINE-1 elements (L1PA in humans) from cfDNA is a specific and sensitive multicancer marker (publication #1 cited in §1.3 related to aim 1). We are now in the process of optimizing the DIAMOND assay by switching from a PCR-based targeting of L1PA to a capture- based targeting. We have generated promising preliminary results using a new technology provided by Agilent (through a grant obtained in the frame of the Agilent ACT-UR program which includes supply of these specific reagents). We obtained an extremely good specificity of targeting (87% on target) and detected cancer-specific hypomethylation. This is now ready to profile a wide range of patient samples.
In the frame of AIM 2, we have identified cancer-specific cfDNA fragmentation signatures at these L1PA regions by analyzing publicly available data. A new PhD student in computational biology has just started working on this project and will investigate this on our own data generated in the lab (including data generated with the new DIAMOND capture-base method which preserves the fragmentation patterns. In parallel, we have published our study demonstrating the power of lncRNAs to identify the tissue of origin using cfDNA fragmentation patterns (publication #2 cited in §1.3).
To explore the potential of EVs as circulating biomarkers of cancer, and in particular to analyze EV- transcriptomes as stated in AIM 3, we are setting up an optimized workflow for isolating these plasma EVs. An optimized protocol is critical to provide the best enrichment in EVs with the least contamination in lipoproteins and platelets. To achieve this, we have been testing several parameters at different stages of EV isolation to define the most appropriate protocol. We have now acquired all necessary equipment for this and are currently finishing some tests to finalize the optimized protocol. Once it has been defined, it will be possible to analyze various parameters associated with EVs, including the composition of their membrane-associated or cytosolic proteins as well as their RNA content.
During the last 2 years, we were able to demonstrate that indeed targeting hypomethylation of evolutionary young LINE-1 elements (L1PA in humans) from cfDNA is a specific and sensitive multicancer marker (publication #1 cited in §1.3 related to aim 1). We are now in the process of optimizing the DIAMOND assay by switching from a PCR-based targeting of L1PA to a capture- based targeting. We have generated promising preliminary results using a new technology provided by Agilent (through a grant obtained in the frame of the Agilent ACT-UR program which includes supply of these specific reagents). We obtained an extremely good specificity of targeting (87% on target) and detected cancer-specific hypomethylation. This is now ready to profile a wide range of patient samples.
In the frame of AIM 2, we have identified cancer-specific cfDNA fragmentation signatures at these L1PA regions by analyzing publicly available data. A new PhD student in computational biology has just started working on this project and will investigate this on our own data generated in the lab (including data generated with the new DIAMOND capture-base method which preserves the fragmentation patterns. In parallel, we have published our study demonstrating the power of lncRNAs to identify the tissue of origin using cfDNA fragmentation patterns (publication #2 cited in §1.3).
To explore the potential of EVs as circulating biomarkers of cancer, and in particular to analyze EV- transcriptomes as stated in AIM 3, we are setting up an optimized workflow for isolating these plasma EVs. An optimized protocol is critical to provide the best enrichment in EVs with the least contamination in lipoproteins and platelets. To achieve this, we have been testing several parameters at different stages of EV isolation to define the most appropriate protocol. We have now acquired all necessary equipment for this and are currently finishing some tests to finalize the optimized protocol. Once it has been defined, it will be possible to analyze various parameters associated with EVs, including the composition of their membrane-associated or cytosolic proteins as well as their RNA content.
We established the DIAMOND method, a new non-invasive multicancer detection tool. DIAMOND is an innovative ctDNA-based method detecting hypomethylation of evolutionary young LINE-1 elements (L1PA), a shared feature of multiple cancers, which already showed an extremely performant ability to discriminate between healthy and tumor plasmas from 6 different types of cancers, including colorectal, breast, lung, ovarian, gastric cancers and uveal melanoma (see our preprint, which is now accepted for publication by Clinical Cancer Research). Our method allows to dramatically increase the sensitivity of ctDNA detection in a cost-effective manner, providing an optimal trade-off between the number of targeted regions and sequencing depth. A patent has been filed for this method (PCT/EP2023/074092 – cited in §1.3).
Demonstrating that transposable elements are innovative non-invasive multicancer biomarkers is major breakthrough and can change the use of liquid biopsies in clinics if we manage to develop and test usable in clinics. This was my working hypothesis in this ERC project, thus is not unexpected findings.
Other studies are also now reporting related findings. We have highlighted this in the discussion of our manuscript undergoing publication (publication #1 cited in §1.3 related to aim 1).
Demonstrating that transposable elements are innovative non-invasive multicancer biomarkers is major breakthrough and can change the use of liquid biopsies in clinics if we manage to develop and test usable in clinics. This was my working hypothesis in this ERC project, thus is not unexpected findings.
Other studies are also now reporting related findings. We have highlighted this in the discussion of our manuscript undergoing publication (publication #1 cited in §1.3 related to aim 1).