Spatial transcriptome and epigenome analysis with focus on prostate cancer
Time: Fri 2022-12-16 13.00
Location: Ragnar Granit, Biomedicum, Karolinska Institutet, Solnavägen 9, Solna
Subject area: Biotechnology
Doctoral student: Maja Marklund , Genteknologi, Science for Life Laboratory, SciLifeLab, Spatial Research - Lundeberg lab
Opponent: Associate Professor Marc Friedländer, Department of Molecular Biosciences, The Wenner-Gren Institute, Stockholm University
Supervisor: Professor Joakim Lundeberg, Science for Life Laboratory, SciLifeLab, Genteknologi; Universitetslektor Patrik Ståhl, Science for Life Laboratory, SciLifeLab, Genteknologi
Each cancer is unique, and therefore the use of general treatments are often suboptimal. If we can understand the mechanisms of cancer development, we might be able to develop effective treatments tailored to each patient. Our bodies are complex three-dimensional structures and how things are organized correlate with proper functioning. Technologies for biological research have escalated enormously in the last years. Going from bulk analysis of tissues to the advent of single cell sequencing and spatially resolved transcriptomics has initiated a new era in biological research. The technology Spatial Transcriptomics (ST) combines histology with next-generation sequencing, making it possible to map which genes that are active at thousands of sub-areas in a tissue section.
In Paper I, ST was combined with an in-house developed artificial intelligence method to explore the landscape of prostate cancer tissue. We identified a gene expression-based tumor signature in healthy tissue areas not possible to recognize through visual assessment, indicating that the genotype changes before phenotype. A gradient of the tumor microenvironment was also identified. In Paper II, prostate cancer tissue from three patients were investigated before and after androgen deprivation therapy using ST. All patients treated with this therapy long enough will reach a clinically defined stage called castration-resistant prostate cancer. We could see that only a set of cancer cells across the tissue responded to the treatment, which allowed comparison of gene expression program in responding versus non-responding cells. By understanding the underlying mechanisms to resistance, it might be possible to target these cells and decrease relapse risk. In Paper III, we inferred copy number variation from ST data allowing for the generation of genome integrity maps in cancerous tissue of prostate, breast, brain, and skin, and in a lymph node. This allowed us to identify tumor clones not recognizable histologically, indicating how genomic instability can be initiated and spread before visible for the naked eye. In Paper IV, we developed a method for spatial ATAC-seq by fusing the ST-technology with ATAC-seq, enabling the analyses of accessible chromatin while preserving histological information. The Visium platform by 10x Genomics was used and we demonstrate a similar capture efficiency to single-cell ATACseq.