Leveraging the Power of Droplets Microfluidics to Profile the Epigenome

Abstract

The human epigenome is a vast genome-wide body of proteins, chemical structures, and nucleic acids that orchestrate mechanistic gene regulation. Monitoring epigenetic modifications as biomarkers can be beneficial in clinical settings to track disease development and diagnose patients with certain conditions. Conventional methods for surveilling the epigenome are chromatin and epigenetic profiling assays that probe for chromatin segments where specific epigenetic modifications are localized. These assays incorporate next-generation sequencing (NGS) which outputs a wealth of information. While epigenetic assays are clinically relevant and generate large sets of data, their utility is mostly constricted to research settings because they require skilled technicians to perform the intricate steps as well as an abundant cell population, rendering them impractical for profiling in small, or rare, clinical samples. In order for screening of the epigenome to be systematically implemented in clinical and diagnostic settings, profiling assays must become automated to be reliable, reproducible, and operatable by non-experts. Herein, we describe work completed to automate three profiling assays, ChIP-Seq, CUT&RUN, and CUT&Tag. Chromatin Immunoprecipitation with sequencing (ChIP-seq) is the gold standard for identifying loci associated with a specific target of interest, such as transcription factors or histone tail modifications. In collaboration with the Mayo Clinic’s Epigenetic Translation Program, this thesis presents consecutive droplet microfluidic modules that address different sections of the conventional protocol to automate ChIP-Seq. Our initial microfluidic module simultaneously lysis whole cells, enzymatically digests chromatin to nucleosomal length, and delivers antibody functionalized magnetic beads that target the epigenetic mark of interest. The second microfluidic module focuses on washing the magnetic beads with the bound target of interest through a co-flow of four distinct wash buffers. Our current work is aimed at demonstrating the versatility of the droplet integrated ChIP system by optimizing the workflow to profile a diverse range of histone tail modifications. Using quantitative polymerized chain reactions (qPCR) and calculating ΔΔCt for a positively and negatively associated gene, we demonstrate effective enrichment of histone tail modifications through our droplet integrated ChIP system. Enzymatic tethering assays, CUT&RUN and CUT&Tag, identify genomic sequences associated with targets of interest, such as chromatin proteins and histone tail modifications. Low-cost, high-resolution methods, like CUT&Tag, are the techniques needed in clinical and medical research. However, improved automation would allow for higher throughput processing. This thesis presents droplet microfluidic platforms for CUT&RUN and CUT&Tag (DropCUTT), capable of permeabilizing whole cells, extracting nuclei, attaching magnetic beads, and assembling a multi-unit immunoprecipitation construct to identify genetic regions of interest, all in droplets. This expands our lab’s droplet microfluidic toolbox by reconfiguring a previously published microfluidic module to condense incubation steps and decrease the amount of manual pipetting steps, all reducing user variation. This semiautomated procedure supports a range of whole-cell inputs (250K-10K cells per sample) while yielding quality tagmented DNA fragments. Preliminary sequencing data confirms accurate genomic profiling of the target of interest, H3K27me3, using the condense assay. Taken together, the work described in this thesis begins to solve some of the main challenges for integration of epigenetic assays into clinical settings. We fully automated the conventionally manual ChIP-Seq procedure and developed microfluidic platforms for CUT&RUN and CUT&Tag that decrease the required cell inputs and condenses protocol steps. In summary, this work expands the epigenetic profiling field by providing methods and microfluidic modules to decrease user variation and required sample input.