Novel FISH methods to unveil genome architecture

Sammanfattning: The genome consists of incredibly long DNA strands that encode all the vital information for the cell to function. The DNA inclusion in the very tight nuclear space and, simultaneously, the establishment of a hierarchical organization of the chromatin that favors transcription of certain genes over others, sparked a long-lasting quest to understand the design principles that govern genome architecture. The lack of proper technologies to study nanoscale structures made the progress slow until the advent of high precision microscopy techniques, next-generation sequencing technologies and in situ methodologies, which revolutionized the field. Not only did the new methodologies unveil which chromatin regions have preferred interactions with other regions, they also revealed that individual DNA sequences have preferential locations inside the nucleus, such as being either closer to the center of the nucleus or to the nuclear periphery. When these methods were applied to different cell and tissue types, genes were found to have different nuclear localization and compaction and to be surrounded with distinctive constellations of genes, depending on the tissue type. Those observations point to a strong interplay between genome architecture and chromatin activity that results in variable cellular functions. Gene activity is modulated by epigenetic modifications of chromatin, which vary depending on nuclear environment and affect gene accessibility to transcriptional machinery. Thenceforth, increased gene accessibility promotes a higher association with the transcription machinery that ultimately transcribe a set of accessible genes. The following thesis papers follow the concept of first developing the needed technology to then be able to address biological questions that could otherwise not be tackled. Henceforth, the insights into the chromatin structure presented here are debated in the light of the new methodology. Paper I aims to empower genome organization studies with a larger and more accessible repertoire of DNA FISH probes, which allow us to visualize small (10 kb) DNA regions using fluorescence microscopy. The paper describes the iFISH technique that reduces the probe production cost by pooling thousands of oligos that belong to many different probes in one single tube. The probes are selectively amplified thanks to a combination of barcodes that can include a color barcode that will hybridize with fluorescently labelled oligos. The probes are very specific and produce high signal-to-noise ratio signals in 6 colors. The ability to image 6 colors simultaneously, together with nucleus staining, permitted a thorough analysis of chromosome intermingling in both embryonic and differentiated cells. Paper II describes a new method, GPSeq, which measures the radial position of DNA regions through genome-wide sequencing. The radial position is computed from samples that go through different, increasing, digestion times with restriction enzymes in fixed cells. The restriction enzymes gradually progress into the center of the nucleus homogeneously, forming concentric layers of digested chromatin. The longer the digestion time, the deeper the area cut by the restriction enzymes towards the nuclear center, which increases the number of different DNA sequences being detected through sequencing. To validate GPSeq in an independent manner, extensive DNA FISH was required to target a vast number of loci scattered across various chromosomes, to compare radiality estimates coming from GPSeq with the goldstandard DNA FISH approach. Paper III demonstrates how to enhance the imaging power of FISH by acquiring more loci simultaneously through multi-color probes. This expansion of iFISH into miFISH increases the number of probes that are individually detected, which empowers higher-throughout chromosome organization studies. The combining of different loci provided 120 pairwise distance measurements that were used to study megabase-scale models for chromosomal arrangement in space. This paper provides publicly available and open-access images that are coupled with detailed information of the multi-color probes datasets. Additionally, the iFISH image datasets used for the method validation are also available, together with miFISH datasets of one probe per channel, which helped to setup the method. Paper IV introduces a new FISH technique, FRET-FISH, for measuring the condensation level of a target locus. FRET is employed to quantify the DNA proximity within a locus of interest using Förster resonance energy transference between dye pairs. In case the dye pairs are within a short distance, the dye of higher quantum excitation energy transfers the energy to the lower energic dye that in turn emits the photons. Thus, the overall proximity between oligos labelled by either donor or acceptor dyes translates into a local compaction estimate, a metric that has been lacking thus far. For the technique optimization, FISH oligos were modified to include an additional sequence to stabilize the fluorescent dye interactions. Furthermore, the oligos were placed at a far enough distance to avoid instantaneous FRET with neighbor oligos. The resulting FRET efficiencies were compared with ATAC-seq and Hi-C results and showed a good correlation. The radial localization in the nucleus affects the locus structure in such a way that a locus closer to the lamina is more compacted than a locus closer to the center of the nucleus. Lastly, the same genes were used to study the influence of drug-induced alterations in compaction, cell cycle and genome instability.