Prat, Alexandre, Ph.D.

Research axis

• Cellular Defense
• Auto-Immune Diseases
• Diseases of the Immune System
• Immune Mediators: Cytokines and Chemokines
• Multiple Sclerosis
• COVID-19
• Cell
• Cellular Degeneration

Research description

Our role in CanProCo is to understand MS disease progression at the gene-to-cell level to uncover novel prognostic and predictive biomarkers, and future targeted therapies. We have made significant progress in validating our experimental approaches, establishing analysis pipelines, and discovering novel biomarker targets. We have performed sc-seq on PBMCs from >400 patients using the BD Rhapsody platform. We have optimized the sc-seq platform to enrich for live cells, prepare sc-seq libraries, quality control, and NGS sequencing. More than 1,500,000 single PBMCs have been sequenced, equating to over 15T of data. In collaboration with Bastian Rieck (Institute of AI for Health, Helmholtz) we have built an automated analysis pipeline for sequence alignment, quality control, data normalization, clustering, cell annotation, and batch correction. A UMAP representation of the PBMCs sequences show the identification of putative immune cell populations. Comparing the proportion of each cell type, we observed a relative reduction of NK cells number in MS compared to controls; properties of the various immune subsets are being analysed.
In CanProCo I, we conducted a series of sc-seq experiments to identify new novel prognostic and predictive biomolecular MS biomarkers. These experiments revealed new promising targets, but also uncover a high degree of complexity underlying the heterogenous presentations and trajectories of MS. In CanProCo II, we aim to overlay the high-quality PBMCs collected yearly, matching clinical and MRI data, and omic data to identify and validate characterize novel MS biomarkers. These combined approaches will generate a comprehensive omic atlas representing a stepping stone toward precision medicine and better prognostics for MS patients. Our aim it to assess the transcriptomic changes in defined peripheral blood cell populations pertaining to MS progression (Unbiased, hypothesis-free approach).
The first phase of CanProCo I focussed on baseline (study entry) samples for sc-seq due to the impossibility of predicting participant trajectories. As CanProCoI is now well underway, we can now select worst (severe progression or significant relapses) and best outcome (no progression, stable remission) trajectories among RIS, RRMS and PPMS participants and conduct transcriptomic analyses on these important clinical groups. To follow how peripheral cell transcriptomes evolved over time, we will leverage the Antibody Sequencing (AbSeq) technology. This technology marks surface protein markers with specific antibody conjugated with nucleotide barcodes compatible with the BD Rhapsody sc-seq platform. This approach was used to reveal broad and specific disease-specific signatures. To avoid biased preconceptions, we will used the standard 30-marker AbSeq Immune Discovery panel, thus providing a balanced labelling of PBMCs. The single-cell transcriptome landscape of barcoded subpopulations at the timepoint of new worst onset will be pairwise compared with autologous sample from one year prior. Conversely, we will compare best outcome samples (no progression) taken one year apart to identify protective transcriptomic signatures. As pairwise autologous comparison in extreme groups should be statistically powerful, we plan to perform 20 samples per subgroups (20 best and 20 worst RRMS and 20 best and 20 worst PPMS). AIM 1’s ML techniques will be used to extract the most relevant and predictive biomarkers predicting at new onset. In addition, worst or best outcome could be MRI parameters, including brain volume loss, number of T2 lesion, T2 lesion volume. Custom panel design is an available AbSeq option. We plan to revisit baseline timepoints by sc-seq dataset, this time introducing custom barcoded antibodies targeting molecules identified in CanProCo I. This will serve as a powerful confirmation of their cell-based biomarkers value.
This exciting and ambitious project is possible through a tight network of neuroscientists expert. We have the bioinformatic expertise (CITADEL) to ensure guidance. But we need a student in molecular biology to take care of the sample preparations and processing prior sequencing, as well as for validation experiments, we plan to perform by qPCR and western blotting.