Construction of MSP
1.1 Mono-MSP
After reviewing a significant amount of literature, we identified four types of membrane scaffold proteins (MSPs) and expressed them in E. coli for nanodisc fabrication. To enhance the expression of monomeric MSPs in E. coli, we conducted a gradient experiment to determine the optimal IPTG concentration and induction time for expression (details can be found in the Engineering and Results sections). Additionally, we refined our purification protocol for MSPs, particularly for spNW50, which has a higher molecular weight. This iterative approach allowed us to efficiently purify MSPs in large quantities with high purity (further details are available in Engineering). Ultimately, we successfully expressed and purified all four monomeric MSPs (see Figure 1).
Figure 1 presents the SDS-PAGE analysis of the monomeric MSPs. Lane 1 contains the molecular weight marker. (a) shows the molecular sieve effluent of spNW15, (b) depicts the effluent of spMSP1D1, (c) illustrates the refined result of spNW50, and (d) displays the molecular sieve effluent of MSP1E3D1.
1.2 Multi-polymerized MSP
During our literature review, we found that the aforementioned monomeric MSPs had limitations in length extension and could not meet the requirements for preparing larger nanodiscs. To address this issue, we designed multi-polymerized MSPs (for details, see the Design section) and expressed each component in E. coli. We conducted a gradient experiment to determine the optimal IPTG concentration for inducing the expression of each part of the multi-polymerized MSP (details can be found in the Engineering section) and refined our protein purification process (see Engineering for specifics). As a result, we successfully extracted and stored the three components of the multi-polymerized MSP in significant quantities (see Figure 2). We also detected mCherry fluorescence, supporting our hypothesis that we successfully constructed the multi-polymerized MSP.
Figure 2. (a) SDS-PAGE analysis of purified SCSdC-mCh [1-10] (one of fragments of the multi-polymerized MSP). The molecular weight of SCSdC-mCh [1-10] is 76.3 kDa. (b) SDS-PAGE analysis of purified SnCSdT (one of fragments of the multi-polymerized MSP). The molecular weight of SnCSdT is 42.1 kDa. (c) Western Blot analysis of purified SnTST- mCh [11] (one of fragments of the multi-polymerized MSP). The molecular weight of SnTST- mCh [11] is 33.6 kDa.
Figure 3. The mCherry fluorescence observation chart (10×10) under green light excitation. It was observed using a fluorescent Inverted microscope and photographed with an ordinary mobile phone.
Membrane Protein Polymerization
The infection of many viruses does not rely on a single receptor (as seen in HIV[1]), prompting us to develop nanodiscs that incorporate membrane protein dimers for functional validation. We acquired monomers of membrane proteins and designed homodimers utilizing the strong interaction of the streptavidin-biotin system along with the self-assembly properties of split GFP.
To confirm the accuracy of our plasmid construction, we performed colony PCR (see Figure 4), which showed successful introduction of the plasmid into BL21(DE3). Following this, we induced protein expression at low temperatures over an extended period using IPTG and carried out protein purification. We separately purified soluble proteins and inclusion body proteins, followed by SDS-PAGE analysis (see Figure 5). The results confirmed the successful isolation of the desired proteins with a relatively high level of purity.
Figure 4. PCR result. Figure A shows the PCR results for the sGFP1-10 tether, with a base length of 1071 bp. Figure B presents the PCR results for the sGFP11 tether, with a base length of 477 bp.
Figure 5. SDS-PAGE result of sGFP tethers. The data for each lane is as indicated in the figure.
After successfully obtaining the proteins, we conducted ELISA activity verification and binding incubation. The comparison of absorbance data with the actual color indicated that the purified soluble protein demonstrated significant activity, while the inclusion body protein remained inactive despite attempts at renaturation (see Figure 6).
Upon incubating the purified soluble protein in a mixed environment for 24 hours, we observed fluorescence under a fluorescence microscope, confirming successful binding and activity (see Figure 7). Furthermore, by adjusting the ratio of the sGFP tether and optimizing the incubation duration—either shortening or extending the time—we were able to increase the binding of additional proteins, enhancing the overall effectiveness of the nanodisc assembly.
Figure 6. ELISA result. Row A, wells 1-9 contain a gradient of mSA standard solution, wells 10 and 11 are 0.5% BSA negative controls; Row B, wells 1-10 contain a gradient of mSA standard solution, wells 11 and 12 are 0.5% BSA negative controls; Row C, wells 1 and 4 contain sGFP1-10 tether protein refolding samples, wells 2 and 5 contain sGFP11 tether protein refolding samples; Row D, well 1 contains sGFP1-10 tether protein purification sample, well 2 contains sGFP11 tether protein purification sample; Rows E and F, well 1 contains sGFP11 tether inclusion body purification sample, well 2 contains sGFP1-10 tether inclusion body purification sample.
Figure 7 presents the combined sGFP observations under a fluorescence microscope. Figures A and C illustrate the background fluorescence, serving as a control. In contrast, Figures B and D display the fluorescence signals after the addition of the samples at their respective locations. By comparing these results, we can accurately identify the corresponding fluorescence signals and effectively eliminate any potential false positives.
In the future, we will achieve the construction of membrane protein dimers by co-incubating the well-bound sGFP tether with biotinylated membrane proteins, and subsequently insert them into nanodiscs for subsequent verification.
Assembly of Nanodiscs
We have successfully prepared nanodiscs using spMSP1D1, spNW15, spNW50, and DOPC at protein-to-lipid ratios of 1:60, 1:100, and 1:600 based on reference literature and our own exploration, and verified them using negative staining electron microscopy.
For the spMSP1D1 nanodiscs, our electron microscopy results are approximately 6-8nm, close to the literature reference value of 11nm (Figure 8a，Figure 8c). Furthermore, we have incorporated the membrane protein Low-Density Lipoprotein Receptor (LDL-R) to prepare spMSP1D1 nanodiscs with LDL-R for our subsequent cell validation experiments. And observed membrane protein structures similar to LDL-R. (Figure 9) Delightfully, we were fortunate to observe the side view of the nanodiscs, which clearly included two spMSP1D1 proteins.(Figure 8b)
Figure 8. The negative staining electron microscopy image of spMSP1D1 nanodiscs shows that the black circular objects in the image are the nanodiscs (a). We observed the nanodiscs from the side (b). According to the quantitative statistics of electron microscope results, the particle sizes of our nanodiscs are close to the value in the literature (c).
Figure 9. The negative staining electron microscopy image of spMSP1D1 nanodiscs with LDL-R.
For the spNW15 nanodiscs, our electron microscopy results show that their particle size is around 15nm, which is basically consistent with the literature values. (Figure 10)
Figure 10. The negative staining electron microscopy image of spNW15 nanodiscs shows that the black circular objects in the image are the nanodiscs(ab). By analyzing the quantization results of electron microscope images, their diameter is approximately 15nm, which is close to the theoretical value (c).
For the spNW50 nanodiscs, our electron microscopy results indicate that their particle diameter is around 300nm, which is larger than the reference values from the literature. Subsequently, we used dynamic light scattering (DLS) for further characterization and found the results to be similar to those from the electron microscopy images. (Figure 11)
Figure 11. The transmission electron microscope image of spMSP1D1 nanodiscs shows that the black circular objects in the image are the nanodiscs(abc). The frequency distribution histogram of particle size measured by DLS shows that the peak particle size is around 200-300 nm(d).
In the end, we successfully obtained spMSP1D1 nanodiscs, spNW15 nanodiscs, and spNW50 nanodiscs, and also obtained nanodiscs with LDL-R. These results provide valuable experience and theoretical support for the development of trimer nanodiscs, and nanodiscs with LDL-R will be used in future cell experiments to verify their functionality. (Figure 12)
Figure 12. Three monomer MSPS, pMSP1D1 (a), spNW15 (b) and spNW50 (c), have successfully produced nanodisks with different particle sizes. The difference in particle size (d) can be highlighted by the box plot.
Effective model construction for nanodiscs
Our model successfully predicted protein structures, streamlined experimental processes, enhanced membrane protein performance, and confirmed the effectiveness of our nanodiscs. Additionally, we simulated the concentration profiles of nanodiscs in the human body over time, paving the way for future practical applications.
We utilized AlphaFold for protein structure predictions and applied statistical design of experiments to guide nanodisc assembly, performing variance and regression analyses to ensure efficiency and reliability. Using EVcouplings, we modeled the effects of directed evolution on membrane proteins and employed HDOCK for molecular docking between membrane proteins and antigens, aiming to produce more stable, higher-affinity variants. To validate the therapeutic potential of the nanodiscs, we used Cellular Automata to simulate interactions between viruses, nanodiscs, and human cells, providing clear visual insights. Lastly, we conducted pharmacokinetic modeling to predict nanodisc concentration changes in the body, offering valuable data for future clinical applications.
Functional Verification
In the novel SEND system we developed, the plasmid cargoRNA is derived from our innovative design. After sequencing the plasmid, we confirmed the rationality of the cargoRNA design and the successful construction of the system. (Figure 13)
Figure 13. CargoRNA sequencing results.
After transfecting cells with the three plasmids—capsid, fusogen, and cargoRNA,we observed that the cells expressed green fluorescent protein. (Figure 14.A) We collected the supernatant by centrifugation and verified the presence of proteins similar in size to PEG10 and VSV-G using SDS-PAGE. However, protein bands also appeared at the same positions in the control medium and in the supernatant of normal cells. (Figure 14.B) These results suggest that the transfected cells may have expressed exosomes containing PEG10 VLPs.
Figure 14. Cells expressed exosomes containing PEG10 VLPs
A. After transfection of the three plasmids, cells emit green fluorescence (red circles);
B. SDS-PAGE of supernatant (VLP: supernatant suspected to contain VLP; Medium: cell culture-medium; Cell supernatant: supernatant of untransfected cells).
In further experiments, we infected normally growing HEK293 FT cells with supernatant suspected to contain PEG10 VLP exosomes. We found that the infected cells emitted green fluorescence. (Figure 15) This indicates that the cells acquired the mRNA of the GFP protein and expressed it. The only possible source of this mRNA is the supernatant we added. This result confirms that the novel SEND system can produce PEG10 VLP exosomes capable of effectively transporting mRNA, leading to its expression in target cells.
Figure 15. After being infected by supernatant suspected to contain PEG10 VLP exosomes, the interior of the cells fluoresced green.
In the aforementioned experiment, we also conducted two control groups: one with only culture medium and another with PEG10 VLP exosomes plus nanodiscs. Fluorescence observations showed that the cells with only culture medium did not express fluorescence. Cells treated with VLPs alone expressed more GFP protein compared to those that also received nanodiscs. (Figure 16) This result partially validates the blocking effect of nanodiscs on the pseudovirus VLPs. Unfortunately, due to time constraints, we were unable to collect more data. We hope to incorporate more rigorous experiments for validation in future studies.
Figure 16. Fluorescence of HEK293 FT cells infected with VLP exosomes and nanodiscs over time.
References
[1]. Sun, G. and C. Wu, Application of GA-BP Neural Network in Online Education Quality Evaluation in Colleges and Universities. Mobile Information Systems, 2022. 2022: p. 7846247.
