Dynamic CRISPRa/I regulation of gene expression in CFS and E.Coli
Advisor: James Carothers (Chemical Engineering)
Effective control of gene expression underlies many modern biotechnological applications from metabolic engineering to diagnostics and bio-computation. Historically, a paucity of orthogonal and engineerable regulators has stymied efforts to increase the scale and complexity of gene regulatory networks. The development of CRISPR-based transcriptional regulators has enabled the generation of increasingly sophisticated and practical regulatory networks. The advent of transcriptional networks operating through the regulated expression of guideRNAs (gRNAs) has further expanded the composability of CRISPR-regulation, allowing gRNAs to serve as a standardized information carrier, greatly simplifying the level matching process in mult-layer circuits. Despite these rapid advances, until recently effective CRISPR-regulation in prokaryotes has been limited to CRISPR inhibition due to the lack of versatile activating domains. To meet the needs of increasingly ambitious undertakings we have developed a prokaryotic CRISPRa/i control system programmable through the regulated expression of gRNAs. In this work, we first establish design principles allowing the formation of multi-layer CRISPRa/i regulatory circuits to provide complex and dynamic control of gene expression. To improve upon CRISPRa/i network function we subsequently engineered expression characteristics of CRISPRa/i nodes to provide increased output dynamic ranges, enabling formation of CRISPRa/i expression programs with increased complexity. Finally, we discuss the application of this modular control system to provide dynamic regulation of gene expression in CFS and E.coli towards optimization of bioproduction. The dynamic regulatory capabilities afforded by the CRISPRa/i control system greatly expands the design space of genetically encoded expression programs. This expanded set of capabilities will enable the rapid generation of genetically encoded, dynamic, multi-gene programs providing access to new avenues for the optimization of metabolic engineering as well as complex signal processing in biological systems.
Ben is now the Lead Applications Engineer at Wayfinder Biosciences, a startup co-founded by fellow molecular engineering alumni.
Laser refrigeration of ytterbium-doped alkali-rare-earth-fluoride nanostructures and applications for radiation-based lasers
Advisor: Peter Pauzauskie (Materials Science & Engineering)
Solid-state laser refrigeration is a process that is able to cool a solid lattice using laser radiations. The material absorbs low-entropy photons produced by a laser and then emits new photons through spontaneous luminescence with the emitted photons having both a higher average energy and net entropy. Solid-state laser refrigeration could be used to cool materials in the absence of electrical connections or vibrations. These unique features have made laser refrigeration useful in many practical applications. A variety of laser refrigeration materials have been developed. Most of them are synthesized using bulk crystal growth methods, which are expensive and time-consuming. In this work, low-cost, rapid hydrothermal synthesis was used to develop new laser refrigeration materials. X-ray irradiation was performed with laser refrigeration materials to study their durability for space applications and new radiation balanced fiber laser designs - that could help reduce the photothermal heating effect during lasing and enable higher laser output - were explored.
After completing her Ph.D. in 2022, Xiaojing joined Dr. Emory Chan's team at Lawrence Berkeley National Laboratory as a postdoc scholar. Xiaojing's current research interest is experiment automation. She is working on AI-accelerated high-throughput synthesis and characterization of upconverting nanoparticles.
Developing multiplexed molecular assays for synthetic biology and DNA data storage with nanopore sensing technology
Advisor: Jeff Nivala (Computer Sciences & Engineering)
Multiplexed molecular assays have become widespread as the number of outputs that are tracked or read out in a reaction in parallel increases. Multiplexed assays offer significant advantages over singleplex assays, including time, reagent costs, sample requirements, and the amount of data that can be generated. In molecular biology, genetically encoded reporter proteins have expanded the toolbox of researchers to track biological phenomena. While they are widely used to measure many biological activities, the current number of uniquely addressable reporters that can be used together for one-pot multiplexed tracking is small due to overlapping detection channels such as fluorescence. Similarly, in DNA data storage, the ability to selectively target data files (i.e “randomly access”) in a multiplexable manner would lessen decoding latency and cost and enable deployment of practical DNA data storage architectures. Here, we develop new multiplexable molecular assays for synthetic biology circuits and DNA data storage random access readout using nanopore sensor arrays. To overcome genetically encoded reporter protein multiplexing limitations, we built an expanded library of orthogonally-barcoded Nanopore-addressable protein Tags Engineered as Reporters (NanoporeTERs), which can be read and demuxed by nanopore sensors at the single-molecule level. Subsequently, to improve upon previous random access architectures, we demonstrate a new random access approach in which files can be selected in multiplex using a CRISPR-Cas9 target address and then decoded using a nanopore sequencer. This work presents a new class of reporter proteins that permit multiplexed, real-time tracking of gene expression along with a new random access DNA data storage strategy that increases one-pot multiplexing and decreased time-to-decoding.
Epigenomic profiling of human tissues at single-cell resolution
Advisor: Steven Henikoff, Fred Hutch Research Center
Traditional methods for profiling DNA-protein binding interactions have been limited by low signal-to-noise, false positives, and high costs. To overcome these barriers we developed a simple assay, Cleavage Under Targets & Tagmentation (CUT&Tag), that leverages a transposon based fusion enzyme to map in situ DNA-protein interactions in small samples of cells at high resolution. We then automated CUT&Tag to generate hundreds of chromatin profiles for a multitude of histone modifications across different diseases. Furthermore, we are able to model their cell-type specific gene expression by integrating the data across multiple histone modifications.
CUT&Tag is characterized by an exceptionally high signal-to-noise ratio and we reasoned that the method could be used to resolve single-cell chromatin profiles. As a proof-of-concept we demonstrated that single-cell CUT&Tag resolves both active and repressive chromatin marks in cell lines. We then leveraged single-cell CUT&Tag to profile thousands of single-cells to uncover the heterogeneity in stem cell development, primary liquid, and solid tumors pre- and post-treatment. Our work is part of a large-scale effort to build a comprehensive map of all cell types to better understand human health and improve disease diagnosis and treatment.
Steven joined the San Francisco based biotechnology company Freenome as a Research and Development Computational Biologist where he will be working on epigenetic modeling.
A molecular engineering approach: multi-functional binders design for lithium-sulfur battery
Advisor: Alex Jen (Emeritus, Materials Science & Engineering)
Lithium-sulfur (Li-S) batteries have attracted great attention as the next-generation battery due to advantages, including high energy density and low cost. Despite its theoretical promise, the practical application of Li-S battery is significantly hindered by the technical challenges of sulfur cathode, including dissolution of LiPSs (Li2Sx, 3 ≤ x ≤ 8) in organic electrolyte, volume expansion cracking, etc.
In my research, I designed multi-functional PENDI-based binders to solve these problems with a molecular engineering strategy. To mitigate volume expansion cracking, we not only designed a self-healing polymer system based on reversible pi-pi interactions, but also proposed a general strategy to modulate self-healing and mechanical properties without changing polymer structure. By utilizing dipole-dipole, ion-dipole interactions, and redox mediation effect of NDI, LiPSs were trapped efficiently. Finally, we integrated these molecular designs into polymer binder for S cathode to achieve long-cycle-life, high-loading S cathode.
Jiaxu is now working in the private sector as a software engineer to design data platforms.
Design, development, and processing of perovskite nanocrystals for optical devices
Advisors: Daniel Gamelin (Chemistry), Christine Luscombe (Materials Science & Engineering), and Devin MacKenzie (Materials Science & Engineering)
Perovskite nanocrystals (NCs) are an exciting new class of luminescent materials with impressive photoluminescence quantum yields, highly tunable properties, and remarkable defect tolerance. Such properties make them ideal candidates for applications in light emitting diodes, luminescent solar concentrators, and quantum photonic devices. However, more work is needed to understand their fundamental properties and leverage these properties in real-world devices. This thesis presents three use-inspired studies that draw on an understanding of fundamental perovskite NC material properties to develop these materials for working device platforms. The first section explains the fundamental perovskite NC properties and potential device considerations necessary to explore the use of these materials in various applied research areas. Then, as a largely fundamental experimental investigation of these materials, I discuss the high potential of quantum-cutting ytterbium-doped perovskite NCs for high-performance luminescent solar concentrators. This study produces design constraints for perovskite NC/polymer composites that inform the development of modular polymers to suspend various perovskite NC compositions in the solid composites. Finally, this understanding of NC ligand chemistry and stability is leveraged to develop methods to process perovskite NCs with electrohydrodynamic inkjet printing for photonic device integration. These studies are important steps towards translating perovskite NC research from fundamental innovations towards the numerous potential device applications of these materials.
Ted is currently a postdoctoral fellow in the MacKenzie lab, where he is developing methods of scalable quantum optical device manufacturing.
Multiplex molecular recording of biological signals land events
Advisor: Jay Shendure (Genome Sciences)
Learning from the past can guide us to make better decisions. Similarly, knowing the past history of each cell within an organism during its development can help us understand what determined its current state and to predict its future behavior. Cells are constantly receiving signals from their environment throughout their lives, altering their current state in response. Yet, when I began my Ph.D., we lacked a technology to record and recover signals that each cell received prior to its sampling. To address this problem, I developed ENGRAM, a DNA-based memory device that utilizes CRISPR to record specific signals in living cells to their genome, analogous to a flight-data recorder. ENGRAM can not only record cellular history but also has the potential to program future cell fates.
A versatile TAL-effector (TALE) protein platform for sequence-specific nucleic acid imaging and detection
Advisors: Michael Jensen (Bioengineering) and John Stamatoyannopoulos (Genome Sciences)
Detection and quantification of genetic material is fundamental for many research and diagnostic applications. We have devised a platform technology based on a single-protein probe architecture compatible with detection of all types of genetic materials (RNA, ssDNA, dsDNA). In this dissertation, I describe the development of a universal platform for sequence-specific nucleic acid imaging and detection based on Transcription Activator-Like Effector (TALE) protein with versatile functional domains, characterization of the platform, and its application in live cells, fixed cellular specimens, and cell-free context.
In fixed and live cell specimens, I used TALE probes to selectively label individual loci under most preservation of nucleic acid condition in the cell environment. The simplicity and robustness of probing system render TALEs an attractive and minimally invasive platform for revealing native genomic structure and function by imaging. TALE platform is also expanded to ex vivo setup in detection of nucleic acid in cell-free context. I deployed TALE probes with various reporter units to implement sequence-specific nucleic acid detection in biological samples. The detection is highly sensitive and comparable detection limit with contemporary assays with robust expansion to multiplexing targeting.
Utilization of TALE platform in nucleic acid detection precipitates in understanding of genomic elements spatial and temporal locations in the cell context and providing a new probing technique in nucleic acids detection in clinical samples.
Redd is a Research Scientist at the Altius Institute for Biomedical Sciences in Seattle.
Removal of Toxins in Kidney Dialysis through Molecularly Imprinted Polymers and Membranes
Advisor: Buddy Ratner (Bioengineering)
Traditional hemodialysis disturbed end-stage renal disease (ESRD) patients’ personal lives while providing poor outcomes (5-year survival rate < 50%). Many studies demonstrated that more frequent hemodialysis with longer time can improve patients’ health conditions. Portable or wearable hemodialysis machines could provide continuous hemodialysis similar to human kidneys, which have the potential to liberate ESRD patients and improve the treatment outcomes. However, a key challenge towards the development of portable dialysis devices is to decrease the consumption of the dialysate (120 L dialysate needed for one traditional hemodialysis session). This problem could be solved by removing specific toxins from used dialysate to achieve dialysate recycle, for which molecularly imprinted polymers (MIPs) and membranes are the perfect materials. MIPs are synthetic “antibody mimics” with high specificity, excellent stability, and low cost. Therefore, they have the potential for targeted removal of key toxins with little effect on other components of the dialysate (e.g. essential nutrients). This dissertation developed MIPs and membranes that are able to specifically remove essential uremic toxins, namely trimethylamine nitro-oxide (TMAO), indoxyl sulfate (IS), p-cresol sulfate (PCS) with enough capacities for portable kidney dialysate devices. This dissertation also developed membranes to specifically filter glucose from urea solutions, enabling urea removal through electro-oxidation for portable dialysis. Overall this dissertation provides materials and devices for some key challenges hindering portable dialysis.