Biomedical research and practice, from cell biology to clinical pathology relies on imaging cells and tissues and the molecules they express. However, while sequencing is now used extensively to profile genomes, transcriptomes and interactions in bulk samples and single cells, it typically cannot resolve the spatial organization of these profiles. This leads to an ever-widening gap between genomics and cell biology and histopathology. There is thus an enormous need for methods that would collect spatially resolved genomics data. Unfortunately, even the most recent technological advances for spatial genomics still face major barriers and cannot be broadly adopted, as they either require costly, specialized and slow imaging equipment, yield data of limited quality or cannot handle idiosyncratic samples like tumors. Here, we propose a completely novel approach ? DNA microscopy ? as a new, scalable, cost-effective, general method for spatial genomics in cells, tissue sections, and whole tissues. DNA microscopy encodes spatial organization into a DNA library, sequences it using standard sequencing, and infers the relative position of RNA, DNA or other molecules using inference algorithms. DNA microscopy relies on a novel PCR-based approach that leads to encoded spatial information just based on the laws of diffusion: the closer two molecules were in the original sample, the more likely they are to have a joint product in the DNA microscopy reaction. Following sequencing, an image is recovered by computation on this information. In preliminary results we provide an end-to-end demonstration that DNA microscopy recovers accurate images, without any optical microscopy, and without any prior knowledge on tissues, cells or their organization. Here, we will develop, expand and disseminate DNA microscopy to a broad utility tool, especially with clinical pathology samples. We will develop DNA microscopy to read out the spatial distribution of sets of transcripts in a biological sample and validate it across diverse cell lines and primary cells (Aim 1). We will extend DNA microscopy for spatial profiling of whole transcriptome profiling with randomized and oligonucleotide-library priming strategies, and of epigenomic markers using antibody-oligonucleotide conjugates targeting methylated DNA cytosine and specific acetylated histones (Aim 2). We will maximize DNA microscopy's impact by adapting it for spatial transcript analysis of signatures and whole transcriptomes in 2D tissue sections and in whole mount (3D) tissue (Aim 3), demonstrating successful analysis in diverse tissues, including brain and human tumors. We will disseminate DNA microcopy broadly to users in research or pathology labs (Aim 4). The reagents we use and the protocols we invented are straightforward, and we will facilitate dissemination by distributing reaction chambers to other labs, releasing software, and conducting outreach. DNA microscopy does not require special or costly equipment, relies on PCR protocols that can be easily adopted in any lab, and will handle cells, tissue sections, and whole tissue, thus maximizing its transformative impact on science and the clinic.
Doctors use imaging of tissue sections to diagnose disease in patients and determine their treatment. Genomic measurements of RNA and DNA are important for this, but it is difficult to measure where they are present in tissue. DNA microscopy will allow to determine where each molecule is without any optical microscopes.
