TL;DR:

1. We aim to build a communication pathway between bacteria and mammalian cells by engineering bacteria to “stick and secrete.”

2. To accomplish this, we split into 3 teams:

a. Adhesion: engineer E. Coli bacteria to display ligands that bind to mammalian cell receptors.

b. Secretion: engineer E. Coli to secrete epidermal growth factor (EGF), which is involved in several cell signaling pathways.

c. Integration: develop co-culture experiments to evaluate the engineered E. Coli on HEK293T mammalian cells.

Project Aims

Our project aims to establish a modular pathway of interkingdom communication between bacterial and mammalian cells, with the primary objective of achieving bacterial adhesion to mammalian cells and subsequent secretion of a target ligand. This research holds promise for advancing current cancer therapies, addressing the limitations inherent in current labor-intensive and costly approaches such as CAR-T therapy.

The costs for various FDA-approved CAR-T therapies range from $373,000 to $475,000 per infusion, limiting this treatment option only to those who could afford it. ⁽¹⁾ Additionally, the process of patient T cell acquisition and modification can take weeks—during this time, many patients become ineligible for CAR-T therapy due to disease progression or declining organ function. ⁽²⁾ We wished to overcome these barriers and limitations with our bacterial communication system, which has the potential to make immunotherapy significantly simpler and less expensive, as bacteria are relatively inexpensive to produce and can be easily engineered to carry therapeutic payloads ⁽³⁾.

In parallel with the focus on bacterial adhesion, we hoped to achieve successful bacterial protein secretion upon close attachment to mammalian cells. Protein secretion by bacteria, once adhered to mammalian cells, presents an opportunity for precise and localized delivery of therapeutic molecules. In fact, Turkish researchers Uratu Seker and his associates had recently explored the release of HlyE toxins by nanobodies once they were bound to the HER-2 receptors (a common receptor found on metastatic tumor tissues). ⁽⁶⁾

Our Approach

To advance our "stick-and-secrete" paradigm, we divided ourselves into three sub-teams: Nanobody Adhesion, Secretion, and Interkingdom Integration. Each team focused on developing specific aspects of the project, aiming to contribute to the overall goal of achieving bacterial adhesion and secretion. We successfully demonstrated the effectiveness of nanobody and antigen bonding in facilitating intercellular communication through microscopy experiments, and we were able to observe a response from mammalian cells to protein secretion. The Interkingdom Integration team also worked on engineering mammalian cells with Synthetic Notch (SynNotch) receptors to elicit a cellular response in parallel with binding. Unfortunately, we were unable to conclusively explore our SynNotch approach, as we were unable to troubleshoot testing the receptor with the proper antigen.

Our project explores methods of interkingdom communication between bacterial and mammalian cells, with the ultimate goal of expanding immunotherapy strategies. Our desire to investigate this issue stems from the need for basic research into such communication mechanisms. We see our project as a basic technology that is pre-therapeutic: in clinical applications, our system has the potential to mitigate economic barriers to treatment, as well as certain efficacy limitations of the therapies themselves. In particular, we foresee our system being used as an alternative to traditional immunotherapy methods.

Adhesion Specific Overview

Goal: Successfully adhere DH10B E. Coli cells to HEK293T Cells via nanobody antigen binding.

Nanobody Antigen Pairs Tested

1. C-Tag Antigen, Anti C-Tag Nanobody

2. GFP Antigen and AntiGFP Nanobody

3. EGFR Antigen and EGFR Nanobody

The adhesion team's objective was to create specific connections between the DH10B strain of E. Coli and HEK293T mammalian cells, essentially testing the binding specificity between different antigen and nanobody pairs. Our experiments primarily referenced the work of David Glass, creator of a nanobody-antigen toolkit. Nanobodies are variable domains of camelid heavy-chain antibodies and unlike large antibodies can be produced cheaply and be expressed by bacteria in a resource-efficient manner to promote interkingdom communication ⁽⁴⁾ Because of their efficacy in binding as well as cheap cost, they quickly became a main focus of our adhesion strategies. Nanobodies displayed on bacteria can recognize an antigen that is either soluble or displayed on other bacteria. Furthermore, they are able to facilitate binding between bacteria via the interactions of the antigen and nanobody displayed on bacteria. The surface display mechanism we are using in this experiment is via the protein intimin. Intimin is an "autotransporter adhesin from enteropathogenic and enterohemorrhagic Escherichia coli that mediates attachment to gut epithelial cells" using its outer membrane protein. ⁽³⁾ Essentially it is a natural binding site for E. Coli for which different nanobodies can be attached. Via antigen binding, nanobody-antigen pairs form strong adhesive units, driving us forward with our project goal, which is to ultimately adhere the bacterial nanobodies to antigens on mammalian cells.

We contacted the coauthors Dr. David Glass and Dr. Ingmar Riedel-Kruse about our interest in using their proposed methodology to test cross-kingdom adhesion, they generously donated their plasmid constructs containing GFP nanobody and GFP antigen fused to the intimin autotransporter protein, which enables surface display. We ordered the C-tag nanobody and antigen constructs directly from AddGene that the authors had made publicly available. All of these plasmids had the same intimin backbone with the adhesin being the variable region in them. Additionally, we aimed to express EGFR nanobody (EgB4 and 7D12) on the intimin surface protein. For this, we designed primers to isolate the Intimin Core backbone from the Glass plasmids and inserted anti-EGFR geneblocks ^{(10, 12)} to them. These primers were designed using the method of Gibson Assembly.

The C-tag antigen/nanobody pair and GFP antigen/nanobody pair was used to demonstrate whether the bacteria attached to the intimin structure could enable bacterial adhesion. However, these two set of Ag-Nb pairs could also be used to run adhesion tests in different combinations between bacteria and mammalian cells by engineering both kinds of cells. The EGFR nanobody was intended to facilitate binding between the engineered E. Coli and HEK293T cells that expressed EGFR on their surface.

The first figure above shows the breakdown of intimin protein at DNA level⁽⁴⁾ Expression of intimin is repressed by the cell but the presence of the inducer (Anhydrous tetracycline and Arabinose) activate the gene. Intimin is an autotransporter protein with a consistent N-terminal domain that faces the inside of the cell and the C-terminal domain points away from the cell, as shown in the second figure above⁽⁹⁾. The protein overall is embedded in the surface of the cell. The adhesin is the variable region at the C-terminal end of intimin that can be either a antigen or nanobody and the linkage of this adhesin to the naturally occurring intimin protein in bacteria builds the foundation of the adhesion side of the project.

Secretion Specific Overview

Goal: Explore different signal peptides to find ones that successfully secrete the EGF ligand.

The primary objective of the secretion team was to investigate and evaluate strategies for the excretion of our model ligand, EGF, from our genetically modified bacterial cells. It's noteworthy that Escherichia coli (E. coli), which served as the bacterial host, belongs to the Gram-negative category and possesses a double membrane. Consequently, the successful secretion of our target proteins depended not only on their proper translocation into the bacterial periplasm but also on their subsequent passage through the bacterial outer membrane into the extracellular matrix.

Traditionally, the engineering of protein secretion into the extracellular environment or the outer membrane of Gram-negative bacteria has involved the manipulation of the Type II Protein Secretion system. This system comprises two essential components. The first component involves the efficient secretion of proteins from the cellular cytoplasm into the periplasm through two distinct protein secretory pathways, known as the SEC and TAT pathways. The SEC pathway involves the transportation of unfolded protein from the cytoplasm into the periplasm where it later undergoes conformational changes and the TAT pathway involves the transportation of folded protein from the cytoplasm into the periplasm. The second component involves the exportation of the protein of interest from the periplasm to the outer membrane through the utilization of Type II Protein Machinery (7,8).

By investigating different signal peptides that were known to be involved in the Type II secretory pathway our team hoped to successfully engineer optimal secretion of the EGF ligand from the DH10B strain of E. coli cells into the bacterial outer membrane and cellular environment in hopes that these proteins would successfully attach to Mammalian cells of interest and induce a downstream response.

Inter-Kingdom Specific Overview

Goal: Characterize the interactions of engineered bacteria constructs with Mammalian Cells via Confocal Microscopy.

The interkingdom integration team is responsible for carrying out co-culture experiments with human cells and our engineered E. coli. While the adhesion and secretion mechanisms have been engineered by the other subteams, our main responsibility is to tune these tools appropriately in order to obtain desirable behavior from the mammalian cells. We aim to characterize the interactions of the engineered bacteria in 2D co-culture.

Additionally, we explored the Synthetic Notch (SynNotch) system, as an alternative method of communication. The SynNotch system can combine adhesion with signaling due to its release of a transcription factor after binding its ligand and experiencing a mechanical force. We modeled our SynNotch approach after the SNIPR system from Roybal KT, ⁽¹¹⁾ where the receptor head can be switched for other receptors to manually change the specificity of binding. ⁽¹¹⁾

SynNotch Mechanism Specific Overview

Synthetic Notch (SynNotch) Receptors are a specialized class of receptors capable of recognizing user-defined membrane-bound antigens. Upon antigen attachment to the receptor, a mechanical force triggers intramembrane proteolysis, leading to ɣ-secretase mediated cleavage. This cleavage leads to the release of an intracellular transcription factor that triggers the activation of a targeted gene Due to their unique ability to regulate gene transcription, SynNotch receptors are currently being investigated for potential therapeutic applications. ⁽¹⁰⁾

Zhu et al. explored the design of Synthetic Intramembrane Proteolysis Receptors (SNIPRs) which area type of SynNotch system, creating a library of potential receptors capable of modulating gene expression. Leveraging these findings, we utilized referenced SNIPRs to bind to antigens on the DH10B strain of E. Coli. The antigens that we investigated for this method were EGFP and C-tag nanobodies which were expressed on the surface of bacteria. Specifically, we modified the pHR_PGK_SNIPR_Hinge Notch SNIPR plasmid (an anti-GFP SNIPR receptor plasmid that releases Gal4, a transcription factor) and the pHR_Gal4UAS_tBFP_PGK_mCitrine reporter plasmid (UAS > BFP), as described in Zhu's study ⁽¹⁰⁾.

Through Gibson Assembly, we inserted the SNIPR receptor and reporter genes into the piggybac backbone labeled pBR70, which was given to us by a mentor. See parts BBa_K4686061 and BBa_K4686062. This system constitutively produces mCitrine as a marker to select for which cells are properly transfected with the reporter. The system’s reporter plasmid causes the cell to be constitutively expressing mCitrine, and only when the SNIPR is bound by its ligand and experiences a sufficient mechanical force, the receptor will release Gal4 which will drive expression of tBFP.

Additionally, we engineered the SNIPR receptor gene from Zhu et al ⁽¹⁰⁾ to be linked via an IRES site to produce mCherry as a reporter that the SNIPR receptor is being made. While inserting the SNIPR receptor into the piggybac vector, we specifically chose to use the backbone from pBR102, again a lab stock from a mentor, which already had the IRES_mCherry sequence. The final plasmid was in a piggybac vector for transfection, and constitutively produces the SNIPR receptor and mCherry.

The desired genes for the receptor and reporter were cloned into piggybac vectors because the original plasmids required the use of lentiviruses for proper transduction in mammalian cells. While the risk related to lentiviruses is relatively low due to its common use, we wanted to reduce risks as much as possible. Using piggybac vectors also allows more accessibility for others if they wish to recreate our results, since they will not need as high of a Biosafety rating to work with a piggybac vector. Read in the human practices tab why piggybac vectors were used instead.

The two cloned plasmids were then cotransfected into a plate of mammalian cells. Individual cells were sorted using the markers we designed. They were then cultured and kept as a new cell line, but unfortunately we were unable to set up a microscopy experiment that paired the sorted, cotransfected cells with the matching bacteria that express surface GFP as we ran into issues with expressing GFP in the bacteria that we were unable to troubleshoot by the end of our research period.

Although the results of this method of adhesion proved unsuccessful, it was a major component of the work we conducted this summer.

References

⁽¹⁾ Borgert R. Improving outcomes and mitigating costs associated with CAR T-cell therapy. Am J Manag Care. 2021 Aug;27(13 Suppl):S253-S261. doi: 10.37765/ajmc.2021.88737. PMID: 34407361.

⁽²⁾ Choi G, Shin G, Bae S. Price and Prejudice? The Value of Chimeric Antigen Receptor (CAR) T-Cell Therapy. Int J Environ Res Public Health. 2022 Sep 28;19(19):12366. doi: 10.3390/ijerph191912366. PMID: 36231661; PMCID: PMC9566791.

⁽³⁾ Nguyen, DH., Chong, A., Hong, Y. et al. Bioengineering of bacteria for cancer immunotherapy. Nat Commun 14, 3553 (2023).

⁽⁴⁾ Glass, D. S., & Riedel-Kruse, I. H. (2018). A Synthetic Bacterial Cell-Cell Adhesion Toolbox for Programming Multicellular Morphologies and Patterns. Cell, 174(3), 649-658.e16.

⁽⁵⁾ Piñero-Lambea, C., Bodelón, G., Fernández-Periáñez, R., Cuesta, A. M., Álvarez-Vallina, L., & Fernández, L. Á. (2015). Programming controlled adhesion of E. coli to target surfaces, cells, and tumors with synthetic adhesins. ACS Synthetic Biology, 4(4), 463–473.

⁽⁶⁾ Shahid G., Ahan R., Ostaku J., Seker, U. (2023). A Bacterial Living Therapeutics with Engineered Protein Secretion Circuits To Eliminate Breast Cancer Cells. BioRvix The Preprint Server for Biology. A Bacterial Living Therapeutics with Engineered Protein Secretion Circuits To Eliminate Breast Cancer Cells. BioRvix The Preprint Server for Biology

⁽⁷⁾ Pei D, Dalbey RE. Membrane translocation of folded proteins. J Biol Chem. 2022 Jul;298(7):102107. doi: 10.1016/j.jbc.2022.102107. Epub 2022 Jun 4. PMID: 35671825; PMCID: PMC9251779.

⁽⁸⁾ Dammeyer T, Tinnefeld P. Engineered fluorescent proteins illuminate the bacterial periplasm. Comput Struct Biotechnol J. 2012 Nov 22;3:e201210013. doi: 10.5936/csbj.201210013. PMID: 24688673; PMCID: PMC3962181.

⁽⁹⁾ Weikum, J., Kulakova, A., Tesei, G., Yoshimoto, S., Jægerum, L. V., Schütz, M., Hori, K., Skepö, M., Harris, P., Leo, J. C., & Morth, J. P. (2020). The extracellular juncture domains in the intimin passenger adopt a constitutively extended conformation inducing restraints to its sphere of action. Scientific Reports, 10(1), Article 1.

⁽¹⁰⁾ Zeronian, M.R., Doulkeridou, S., van Bergen en Henegouwen, P.M.P. et al. Structural insights into the non-inhibitory mechanism of the anti-EGFR EgB4 nanobody. BMC Mol and Cell Biol 23, 12 (2022).

⁽¹¹⁾ Zhu I, Liu R, Garcia JM, Hyrenius-Wittsten A, Piraner DI, Alavi J, Israni DV, Liu B, Khalil AS, Roybal KT. Modular design of synthetic receptors for programmed gene regulation in cell therapies. Cell. 2022 Apr 14;185(8):1431-1443.e16. doi: 10.1016/j.cell.2022.03.023. PMID: 35427499; PMCID: PMC9108009.

⁽¹²⁾ Group, N. (n.d.). 4KRM: Nanobody/VHH domain 7D12 in complex with domain III of the extracellular region of EGFR, pH 3.5. Retrieved October 9, 2023, from https://www.ncbi.nlm.nih.gov/Structure/pdb/4KRM