Microbiome research in the past decade has revealed an astounding prevalence of bacteria in various tissues in the human body. Concurrent progress in synthetic biology has generated a converging interest in the genetic programming of bacteria to locally produce therapeutic payloads and supplant physiological niches. This dissertation presents the development of bioengineering tools that address several key challenges for the clinical translation of therapeutic bacteria. In particular, we focus on the engineering of bacteria for tumor and granuloma applications. Bacteria have been demonstrated to selectively grow within solid tumors, primarily due to the reduced immune surveillance in the necrotic and hypoxic cores. This natural tropism to tumors presents a unique opportunity to engineer bacteria as drug delivery vehicles for cancer therapy. While the recent advancement in microbial engineering has constructed ranges of therapeutic bacteria, a universal bottleneck for clinical development is the lack of tools to rapidly characterize therapeutic candidates in a complex physiological environment. To recapitulate bacterial tumor colonization in vitro, we developed a method that selectively grows bacteria within the necrotic core of tumor spheroids. This platform enabled high-throughput cocultures and predicted in vivo therapeutic outcomes, identifying potent anticancer proteins deliverable by tumor-homing Salmonella typhimurium.
To ensure safety when using bacteria that produce cytotoxic payloads, we prevented bacterial spread to unintended locations by confining bacterial growth in a tumor-specific environment. We constructed hypoxia, pH, and lactate sensors and regulated bacterial growth based on sensor activation. To improve tumor specificity, we engineered gene circuits to sense hypoxia and lactate in an AND-logic gate manner. Leveraging the coculture platform, we characterized sensor activities and circuit functionalities in tumor spheroids. This engineered strain showed improved tumor specificity in an animal tumor model.
Moving towards clinical applications, a key challenge is to ensure bacterial delivery to tumors without activating adverse immune responses. Approaches such as surface decoration can evade immune systems, but static modification may result in bacterial overgrowth. We developed a genetically-encoded microbial encapsulation system with a tunable, dynamic expression of capsular polysaccharides. We constructed an inducible gene circuit to regulate encapsulation, which exhibited tunable protection of the probiotic Escherichia coli Nissle 1917 (EcN) from host immune factors. By dynamically balancing low immunogenicity and protection, transient encapsulation increased the maximum tolerated dose of bacteria by approximately 10-fold when systemically injected in vivo. This strategy enhanced antitumor efficacy in multiple tumor models.
Building on our work of therapeutic bacteria for cancer, we explored the use of engineered bacteria to infiltrate other pathogenic regions in the body. Specifically, we discovered that probiotic EcN colonizes granulomas, pathological features that develop at infection sites including tuberculosis. Granulomas share key similarities with solid tumors, including hypoxia and necrosis, and pose significant challenges for delivering therapeutic agents to eradicate the pathogen Mycobacterium tuberculosis within. We engineered the probiotics to locally produce antimicrobial proteins against Mycobacterium within granulomas. We developed a novel dual lysis mechanism to simultaneously enhance therapeutic protein release and limit bacterial overgrowth. To improve specificity, we constructed hypoxia-dependent bacterial growth coupled with quorum-mediated gene activation. Finally, we showed that our engineered probiotics reduced levels of Mycobacterium strains.
Altogether, the presented technologies utilize a multiscale framework from circuit design to in vitro and in vivo models and advance bacteria as next-generation drug delivery vehicles capable of sensing and responding to diseases in the body.
Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/9axy-5n15 |
Date | January 2022 |
Creators | Harimoto, Tetsuhiro |
Source Sets | Columbia University |
Language | English |
Detected Language | English |
Type | Theses |
Page generated in 0.0021 seconds