Year: 2018
Type: dissertation
Abstract: Epigenome regulates gene expression without changing the order of nucleotide sequences in DNA and has been confirmed to be involved in enormous numbers of biological processes including cell proliferation and differentiation, DNA replication and repair, and disease development. It is of medical relevance to further our understanding of the complex network of epigenomic regulation and apply the knowledge in disease diagnosis, prognosis, and therapy development. Eventually, profiling and screening personal epigenome can make a crucial part in precision medicine. Current investigating methods of epigenome usually start with a large number of cells and rely on manual processing of massive numbers of sample preparation prior to data readout, unrealistic in clinical settings where sample sizes may be limited, user qualification may vary, and human resources may also be inefficient. Signals from single cells are also buried due to bulk processing, limiting the application of these powerful assays in heterogeneous epigenome mapping. Thus, automating epigenomic assays using droplet microfluidics were proposed, aiming to utilizing the favorable features of droplets to approach flexible sample-sized, automated epigenomic assays starting directly from cells for qualified clinical usages. The first trial was using droplet microfluidics to automate samples preparation for an assay probing nucleosome positioning called MNase-seq. Using a droplet microfluidic device to encapsulate cells with lysis and digestion reagent, cellular chromatin was digested extensively to mononucleosomes (yield at ~ 80%) with sample sizes ranging from 125,000 cells to 2,500 cells. Next-generation sequencing of purified mononucleosomal DNA from these samples validated the efficacy of this platform to generate qualified samples for nucleosome mapping with results in consistency with literatures. Clinical samples from 5 pairs of septic shock patients and matched non-septic controls along with 3 healthy controls (buffy coat extracted directly from whole blood) were also successfully processed by the same strategy and the sequencing results conveyed diagnostic values from the identified nucleosome positioning. More patients need to be enrolled to continue the clinical study to further validate the clinical relevance of nucleosome mapping from samples generated droplet microfluidically. To push forward the droplet microfluidic sample preparation, another device capable of preparing two samples for two different assays from the same cell sample was designed and characterized. Using a bifurcating channel design, droplets were aliquoted randomly to two directions allowing digestion of chromatin to different extents. Nucleosome-depleted regions as well as nucleosomal regions were both recovered for the detection of accessible regions and nucleosome locations, respectively. qPCR quality control results confirmed the enrichment of nucleosome-depleted, linker regions where regulatory sequences are located from the samples prepared on the device (fold change ≥5). Both cell line cells and clinical samples were processed with the optimal design for sequencing to achieve a final validation. Besides chromatin processing and related assay automation, chromatin immunoprecipitation (ChIP), the gold standard probing DNA-protein interaction, was also transformed to droplet microfluidics. The preliminary results confirmed that the conditions on the droplet microfluidic device were compatible with ChIP. Further optimization was required to improve the enrichment of targeted regions from ChIP DNA collected out of the droplet microfluidic device. The automation level of the entire workflow also needed enhancement by integrating the immunoprecipitating device with the immuno-bead conjugate wash device.
Author: Yi Xu
Language: English
Subfield: Fluid Flow and Transfer Processes
Field: Chemical Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W3187317929
