Diagnostic and Therapeutic Microbial Circuit with Application to Intestinal Inflammation

Creators: Merk, Liana N.¹; Shur, Andrey S.²; Jena, Smrutiti³; Munoz, Javier³; Brubaker, Douglas K.^{4, 5}; Murray, Richard M.²; Green, Leopold N.^{2, 3}

1. Harvard University
2. California Institute of Technology
3. Purdue University West Lafayette
4. Case Western Reserve University
5. University Hospitals Case Medical Center

Abstract

Bacteria genetically engineered to execute defined therapeutic and diagnostic functions in physiological settings can be applied to colonize the human microbiome, providing in situ surveillance and conditional disease modulation. However, many engineered microbes can only respond to single-input environmental factors, limiting their tunability, precision, and effectiveness as living diagnostic and therapeutic systems. For engineering microbes to improve complex chronic disorders such as inflammatory bowel disease, the bacteria must respond to combinations of stimuli in the proper context and time. This work implements a previously characterized split activator AND logic gate in the probiotic Escherichia coli strain Nissle 1917 (EcN). Our system can respond to two input signals: the inflammatory biomarker tetrathionate and a second input signal, anhydrotetracycline (aTc), for manual control. We report 4-6 fold induction with a minimal leak when the two chemical signals are present. We model the AND gate dynamics using chemical reaction networks and tune parameters in silico to identify critical perturbations that affect our circuit's selectivity. Finally, we engineer the optimized AND gate to secrete a therapeutic anti-inflammatory cytokine IL-22 using the hemolysin secretion pathway in the probiotic E. coli strain. We used a germ-free transwell model of the human gut epithelium to show that our engineering bacteria produce similar host cytokine responses compared to recombinant cytokine. Our study presents a scalable workflow to engineer cytokine-secreting microbes driven by logical signal processing. It demonstrates the feasibility of IL-22 derived from probiotic EcN with minimal off-target effects in a gut epithelial context.

Copyright and License

Acknowledgement

We thank Miki Yun, William Poole, Ayush Pandy, John Marken, Andy Halleran, Reed McCardell, Mark Prator, Chelsea Hu, and Boo Tsang for technical support and all other Murray Lab and Tseng Lab members for insightful discussions. We thank Henry Schreiber and Sarkis Mazmanian for their collaboration and discussions regarding this project. Special gratitude to Caltech NSF AGEP Fellowship, Rosen Fellowship, and NSF by grant #23057066 for supporting the work led by Leopold Green. We thank Caltech CEMI for supporting this work’s directions. Justin Bois has provided excellent discussions regarding data analysis and availability. Martin Buck and Baojun Wan provided logic gate strains. Julie Haseman (Purdue University) designed figures; others were created with BioRender.com and modified in Adobe Illustrator.

Contributions

L.N.M.: Conceptualization, construction of tetrathionate and AND gates, investigation, simulation of tetrathionate and AND gates, visualization, analysis, writing–original draft, writing–review and editing.A.S.S.: Construction of tetrathionate and AND gates, investigation, simulation of tetrathionate and AND gates, analysis.S.J./J.M./D.K.B.: Investigation, transwell in vitro model, comparison of cytokine response, visualization, analysis, writing–review, and editing.R.M.M.: Supervision, funding acquisition, writing–review, and editing·L.N.G.: Conceptualization, supervision, funding acquisition, visualization, analysis, writing–original draft, writing–review, and editing.

Supplemental Material

Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acssynbio.3c00668.

Supplementary Tables contain (1) chemical reactions governing the two-component system model, (2) parameters for this model, (3) a generalized linear model all pairwise comparison, and (4) compared to control. The remaining Supporting Information pertains to two-component model sensitivity analysis Figure S1, design space exploration and RNAP binding tuning Figure S2, AND gate screening Figure S3, optimization of EcN secreting IL-22 Figure S4, coupled to the AND gate Figure S5, Caco2 culturing Figure S6, and cytokine analysis of the transwell model (PDF)

Files

sb3c00668_si_001.pdf

Files (38.9 MB)

Name	Size	Download all
sb3c00668_si_001.pdf md5:3cb5e6cc5ac268acfc1555804ea0261a	38.9 MB	Preview Download

	All versions	This version
Views	0	0
Downloads	0	0
Data volume	0 Bytes	0 Bytes