Design Overview
Goal: Our ultimate goal was to develop a scalable and efficient method through which to produce antimicrobial peptides (AMPs) in bacteria while preventing toxicity of the AMPs to the expressing bacteria.
Plan Overview
- Express an AMP toxic to bacteria in E. coli
- Show that the encapsulin system mitigated AMP toxicity during expression
- Purify AMP from the E. coli and encapsulin system
Conceptual Design
First, we had the option of chemically fusing our AMP directly to an encapsulin monomer or fusing the AMP to a targeting peptide (TP) that would be recognized by the encapsulin, causing it to encapsulate the AMP and orient the construct inside it. We chose the targeting peptide fusion, as we felt that encapsulin monomer fusion to an encapsulin N or C terminus may place the AMP in an unnatural cargo location where it may disrupt normal encapsulin structure and function.
Second, encapsulins are difficult to open. In fact, we could not find an instance of an encapsulin disassembling under mild conditions without special engineering. Most often, extreme pH environments or high levels of urea are required to disassemble the structure. Fortunately, a lab on the University of Michigan campus headed by Dr. Tobias Giessen, engineered an encapsulin known as T4GALA that disassembles in acidic conditions around pH 6 [1]. This milder disassembly condition reduces the risk of destruction of the AMP as the encapsulin is opened and enables safer, more accessible lab conditions to experiment with encapsulins.
Third, all existing research only used a single toxic protein per encapsulin monomer [2]. Encapsulins are made of several large monomers and create a nanometer-scale space which can fit many AMPs per monomer. Producing multiple encapsulin monomers creates a significant metabolic burden for the expression organism, thus it would be energetically advantageous to fill the assembled encapsulins with as much protein product as possible. This yields less encapsulins within the cell, but more AMPs per encapsulin. Therefore, we laid plans to create a chain of AMPs, separated by protease sites, which would increase AMP yield per encapsulin. However, we also recognized that our team should start on a smaller scale in our initial experimental attempts to ensure that E. coli can successfully express a single AMP per encapsulin monomer before increasing the AMP ratio. Thus, we began with a 1:1 encapsulin monomer-AMP ratio and designed DNA constructs to be ordered after perfecting the simpler design.
Fourth, AMPs often have post-translational modifications. AMPs are naturally generated by a wide variety of organisms, many of which are composed of eukaryotic cells that have the cellular machinery to modify peptide products. AMPs from these sources would require complex engineering to perform in a prokaryotic organism, so we sought an AMP that required no post-translational modifications. Similarly, if we were to use E. coli as our expression organism, we needed an AMP with strong potency against gram negative bacteria. Fortunately, research by Tek-Hyung Lee et al. had identified an AMP called HMBC2 that fit such a description [2]. HMBC2 is a hybrid AMP with two fused components, one a calcium pump inhibitor from bee venom called melittin [3], and the other a membrane disruptor found in moths called cecropin [4].
Experimental Design
To test our system, we planned a series of experiments to compare expression of our encapsulin+AMP fusion vs the encapsulin or AMP alone. We expressed our interest to the Giessen Lab, and they generously donated T4GALA encapsulin in a PetDuet expression vector with ampicillin resistance under inducible regulation by the lac operon. This plasmid served as our negative control, as work from the Giessen lab has proven that this construct is an efficient way to produce encapsulin protein without causing harm to the expression bacteria. We then designed and ordered a short segment of DNA which contained our AMP connected to the targeting peptide with a TEV site for later separation and a short, flexible, hydrophilic linker to allow for independent protein movement of these two functional domains. Our construct design oriented the targeting peptide immediately upstream of the encapsulin to promote translation of the two in close proximity. We took this construct and inserted it into an empty pETDuet vector to create a positive control as well as the pETDuet vector containing the encapsulin. The AMP construct in pETDuet alone served as our positive control, as we would expect the bacteria to experience cell death or restricted growth as a consequence of expressing the toxic peptide. The AMP construct in the pETDuet vector containing the encapsulin served as our experimental construct to test whether or not the presence of the encapsulin would protect the expression bacteria from negative repercussions of the toxic peptide.
Figure 1: Three Plasmid Constructs
Polymerase chain reaction (PCR), restriction digest, and Gibson assembly reactions were then used to clone our constructs into pETDuet. Gibson assembly was our chosen method of ligation due to its high efficiency and compatibility with our 1-2 moderately sized inserts. A restriction digest was chosen to linearize pETDuet due to conveniently located cloning sites and to avoid hard-to-detect off-site mutations that may have resulted from PCR linearization. We chose to digest our vector in two places with different restriction enzymes to prevent rapid re-ligation of compatible ends.
After confirming our assemblies had been correctly created with DNA sequencing, we tested several methods to measure the growth rates of bacteria expressing our proteins before settling on conducting the experiment in a plate reader that took consistent OD600 measurements of bacteria with all three constructs, plus empty pETDuet vector, with and without IPTG induction of protein expression. We chose to compare our AMP + targeting peptide to encapsulin + AMP + target peptide in order to avoid confounding, as it was a possibility that the target peptide could reduce or prevent AMP toxicity. Thus our design allowed us to isolate the effect of the encapsulin on mitigating toxicity. Additionally, this was consistent with findings from our literature review which suggested that the attachment of a short sequence to the end of our AMP would not affect its toxicity.
In parallel, we planned to express, purify, and test AMPs produced with our system. To do this, we designed experiments to transform our AMP into an expression strain of E. coli, induce protein expression through induction of the lac operon with IPTG, and purify encapsulin + AMP. Without easy access to FPLC or similar chromatography, we relied on the thermostability of our encapsulin and its solubility properties for purification. Literature had shown our AMP to be stable inside a thermostable protein structure under conditions up to 85 degrees C, and we planned a two-step purification of heating and precipitating with ammonium sulfate to purify our encapsulin and AMP from E. coli proteins and cell material from expression.
We are currently developing protocols to test AMP toxicity in an E. coli expression strain following separation from the encapsulin via purification and TEV protease digestion. These steps currently remain under development.
References
- Triggered reversible disassembly of an engineered protein nanocage Jesse A. Jones, Ajitha S. Cristie-David, Michael P. Andreas, Tobias W. Giessen bioRxiv 2021.04.19.440480; doi: https://doi.org/10.1101/2021.04.19.440480
- Lee, T. H., Carpenter, T. S., D'haeseleer, P., Savage, D. F., & Yung, M. C. (2020). Encapsulin carrier proteins for enhanced expression of antimicrobial peptides. Biotechnology and bioengineering, 117(3), 603–613. https://doi.org/10.1002/bit.27222
- Ceremuga M, Stela M, Janik E, Gorniak L, Synowiec E, Sliwinski T, Sitarek P, Saluk-Bijak J, Bijak M. Melittin-A Natural Peptide from Bee Venom Which Induces Apoptosis in Human Leukaemia Cells. Biomolecules. 2020 Feb 6;10(2):247. doi: 10.3390/biom10020247. PMID: 32041197; PMCID: PMC7072249.
- Huan, Y., Kong, Q., Mou, H., & Yi, H. (2020). Antimicrobial Peptides: Classification, Design, Application and Research Progress in Multiple Fields. Frontiers in microbiology, 11, 582779. https://doi.org/10.3389/fmicb.2020.582779References