TU Darmstadt

Results summary

Introduction

We, the iGEM TU Darmstadt team, developed a modular Virus-like particle (MVP) platform based on P22 Virus-like particles (VLPs). This platform can be used for faster development of vaccinations, drug delivery approaches, and many other purposes.

The real MVP platform is based on the following components:

P22 scaffold protein (SP)
P22 coat protein (CP)
CP with Sortase A recognition tag (CP-LPETGG)
Sortase A5M and Sortase A7M
a protein of interest (here fluorescence reporter proteins)

Assembled VLPs can be modified through a enzyme-mediated linkage to other proteins by sortase. We tested in vivo methods to control the modification ratio through reporter genes (superfolder green fluorescence protein (sfGFP) and mCherry) and developed a prototype bioreactor system for the automated in vivo production of VLPs with a specific modification ratio.

Successful Assembly of VLPs

We successfully assembled Virus-like particles (VLPs) based on SP and CP. SP was genetically fused to sfGFP for some experiments. CP was genetically fused to LPETGG-tag. This tag allowed us a sortase-mediated modification of the VLPs after autonomously assembly of SP and CP. A variety of modified VLPs were produced and characterized. We applied a lot of different methods, ranging from SDS-PAGE, transmission electron microscopy (TEM) and dynamic light scattering (DLS) to endotoxin-tests, for characterization. Altogether, we assessed physical and safety properties of our VLPs. The most important result is the vertification of difference between modified and unmodified VLPs (see Fig. 2).

Figure 1: Sediment containing P22-VLPs modified with sfGFP using Sortase A5M. Sediment was imaged in transmission electron microscope.

Figure 2: Dynamic light scattering analysis. Hydrodynamic diameters of different P22-VLP species.

Characterization of sortase

Enzyme kinetics

We aimed to modify the particles with an enzyme called Sortase A, in particular, two variants – Sortase A5M and Sortase A7M. These enzymes covalently link two molecules with specific tags and enable high modularity for the modification of the surface of our VLPs. Sortase A5M is one of the fastest sortases known, however, Sortase A7M is Ca²⁺-independent which can be useful in biological systems that do not provide constant Ca²⁺ levels.

Figure 3: Comparison of the reaction speed of Sortase A5M with Ca²⁺ and Sortase A7M without Ca²⁺ under optimized conditions. The enzyme kinetics were generated through the linking of TAMRA-LPETG and GGGG-sfGFP resulting in detectable a FRET.

Modeling

We modeled a 3D structure of Sortase A7M through homology modeling and investigated binding affinities as well as the molecular mechanisms using molecular dynamics (MD) and docking simulations. We postulated a possible Sortase A7M linking mechanism, concerning a flexible loop and its role during the reaction.

Animation 1: 3D-Structure of Sortase A7M. To find out more visit our modeling page.

Fluorescence resonance energy transfer (FRET)

In the lab, we studied the catalytic efficiencies of both sortases by performing SDS-PAGEs and FRET-assays. During the FRET-assays, we identified a new, superior FRET pair (5-Carboxytetramethylrhodamin-LPETG (TAMRA) and mCherry).

Figure 4: Depicted are the extinction and emission spectra of TAMRA and mCherry. Due to the large overlap of TAMRA emission and mCherry extinction it is possible to perform a FRET with this pair of fluorophores. The graph show the relative fluorescence unit (RFU[%]) in relation to the extincted/emitted wavelength [nm]. The peaks are normalized to 100 %.

Modification ratio

The VLP surface modification ratio is an important parameter. The modification ratio can be adjusted with a ratio of CP-LPETGG and CP without LPETGG. Therefore, we investigated the expression ratio of two fluorescent proteins using a dual expression plasmid. Depending on the inducer concentration, the ratio can be tightly controlled. We devised an equation that helps adjusting the expression ratio and thereby influences the modification ratio of the VLPs.

Figure 5: Proof of concept calibration curve for the adjustment of VLP surface modification. Test of our dual expression system with two reporter genes for fluorescent proteins: Various indication concentrations of AHT (blue) lead to this asymetric sigmoidal regression (red) of the ratio of mCherry to sfGFP. The samples were taken in triplicates after an overnight expression.

Industrial production

To achieve a potential industrial production of the VLPs, we developed a custom bioreactor system with an optical density (OD) sensor. This sensor is the foundation for an automated VLP production in the future.

Figure 6: 3D printed OD₆₀₀ SensorBrick.

Team:TU Darmstadt/Project/Results Summary