key: cord-1036354-kd6cn4qd
authors: Gourdelier, Manon; Swain, Jitendriya; Arone, Coline; Mouttou, Anita; Bracquemond, David; Merida, Peggy; Saffarian, Saveez; Lyonnais, Sébastien; Favard, Cyril; Muriaux, Delphine
title: Optimized production and fluorescent labelling of SARS-CoV-2 Virus-Like-Particles to study virus assembly and entry
date: 2022-03-27
journal: bioRxiv
DOI: 10.1101/2022.03.23.485575
sha: c3d66ee60eaa80e4c7acd7e864508b1534418f49
doc_id: 1036354
cord_uid: kd6cn4qd

SARS-CoV-2 is an RNA enveloped virus responsible for the COVID-19 pandemia that conducted in 6 million deaths worldwide so far. SARS-CoV-2 particles are mainly composed of the 4 main structural proteins M, N, E and S to form 100nm diameter viral particles. Based on productive assays, we propose an optimal transfected plasmid ratio mimicking the virus RNA ratio allowing SARS-CoV-2 Virus-Like Particle (VLPs) formation composed of the viral structural proteins M, N, E and S. Furthermore, monochrome, dual-color fluorescent or photoconvertible VLPs were produced. Thanks to live fluorescence and super-resolution microscopy, we quantified VLPs size and concentration. It shows a diameter of 110 and 140 nm respectively for MNE-VLPs and MNES-VLPs with a minimum concentration of 10e12 VLP/ml. SARS-CoV-2 VLPs could tolerate the integration of fluorescent N and M tagged proteins without impairing particle assembly. In this condition, we were able to establish incorporation of the mature Spike in fluorescent VLPs. The Spike functionality was then shown by monitoring fluorescent MNES-VLPs docking and endocytosis in human pulmonary cells expressing the receptor hACE2. This work provides new insights on the use of non-fluorescent and fluorescent VLPs to study and visualize the SARS-CoV-2 viral life cycle in a safe environment (BSL-2 instead of BSL-3). Moreover, optimized SARS-CoV-2 VLP production can be further adapted to vaccine design strategies.

nm respectively for MNE-VLPs and MNES-VLPs with a minimum concentration of 10e12 23 VLP/ml. SARS-CoV-2 VLPs could tolerate the integration of fluorescent N and M tagged 24 proteins without impairing particle assembly. In this condition, we were able to establish 25 incorporation of the mature Spike in fluorescent VLPs. The Spike functionality was then 26 shown by monitoring fluorescent MNES-VLPs docking and endocytosis in human pulmonary 27 cells expressing the receptor hACE2. This work provides new insights on the use of non- 28 fluorescent and fluorescent VLPs to study and visualize the SARS-CoV-2 viral life cycle in a 29 safe environment (BSL-2 instead of BSL-3). Moreover, optimized SARS-CoV-2 VLP production 30 can be further adapted to vaccine design strategies. 31 

plasmid N and inserts were digested and purified. Ligation between these digested products 124 was performed using T4 DNA ligase (M0202S, NEB). 50µl of competent bacteria were next 125 transformed with 5µl of the ligation reaction product. Plasmids were extracted from various 126 colonies with the NucleoSpin plasmid mini-kit (740588.50, Macherey-Nagel). Validated 127 plasmids were then sequenced by Eurofins. HEK293T cells were seeded into 6-well plate or 10 mL dishes 24h before transfection. 131 Plasmids expressing viral proteins were transfected with CaCl 2 /HBS2X (50mM HEPES pH 7.1, 132 280mM NaCl, 1.5mM Na 2 HPO 4 ) (v/v) into HEK293T cells. A total of 1.4µg of mixed plasmids 133 was applied for the transfection of a well (for 6-well plate) and 0.7µg/well for individual 134 plasmid. Co-transfection of double, triple or quadruple plasmids was conducted with M, N, E 135 and S with plasmid ratio 3:3:3:5 respectively (0.3µg:0.3µg:0.3µg:0.5µg per well to 6-well 136 plate for example) in the first assay as shown in Figure 1B . Optimization of SARS-CoV-2 VLPs 137 was performed by co-transfecting M, N, E and S with a plasmid ratio of 3:12:2:5 respectively 138 corresponding to the molecular viral RNA ratio found in infected cells [17] [18] , 139 corresponding, for an example, to a total of 2,2µg/well of 6 well plate (with M:N:E:S 140 corresponding to 0.3µg:1.2µg:0.2µg:0.5µg of transfected plasmid) as shown in Figure 1C . Collected VLPs and purification 154 VLPs were obtained from culture medium of transfected cells by clarification at 5000 rpm for 155 5 min at 4°C. Then the VLPs were collected from clarified supernatants after 156 ultracentrifugation, through a 25% sucrose cushion in TNE buffer (10 mM Tris-HCl pH 7.4, 157 100 mM NaCl, 1 mM EDTA), at 28 000 rpm or 30 000 rpm for 3h at 4°C in Beckman SW41Ti 158 or SW55Ti rotors respectively depending on the volume. VLPs from ultracentrifugation were 159 resuspended overnight at 4°C in TNE buffer. The purified VLPs were stored at 4°C. Proteins from cell lysates (20µg of total proteins) or VLPs (volume to volume) were loaded 172 and separated on 8% and 12% SDS-PAGE gel and transferred onto a polyvinylidene difluoride 173 transfer membrane (Thermo Fisher). Immunoblotting was performed by using the 174 corresponding antibodies. Horse Radish Peroxidase signals were revealed by using 175 Amersham ECL Prime (Sigma Aldrich). Images were acquired using Chemidoc Imaging system 176 (Bio-Rad). Each band intensity on the immunoblot were quantified using ImagJ software. with a constant approach/retract speed of 25 µm/s (z-range of 100 nm). Using the JPK SPM-199 data processing software, images were flattened with a polynomial/histogram line fit. Low-200 pass Gaussian and/or median filtering was applied to remove minor noise from the images.

The Z-color scale in all images is given as relative after processing. Particle height analysis, 202 based on the height (measured) channel of the QI mode, was performed using the cross-203 section tool of the analysis software to calculate the maximal central height on each particle. in order to obtain the diffusion time D and the number of particle N (eq. 1). All particles 222 were assumed to have similar brightness.

The VLPs concentration (in particle number/ml) is immediately obtained by equation 2

The VLPs diameters (d) were calculated using the following expression:

.µ. ² Amsterdam, the Netherlands) using a NA 1.49 SR-100x TIRF objective at room temperature. ThunderSTORM plugin in Fiji [23] . The module DBSCAN of the super-resolution 243 quantification software SR Tesseler [24] was used to quantify cluster diameters. 283 We first explored the minimal system requirements for SARS-CoV-2 VLPs production. We proportionality (Western Blot Figure 1C , compare S2 in lanes 1 and 3 "VLP"). This suggests 323 that the production of mature SARS-CoV-2 particles is most probably fine-tune regulated by 324 viral RNA ratios during infection [17] .

The purified MNE and MNES VLPs were next imaged by AFM in buffer for size and shape 326 characterization ( Figure 1D ), in the conditions previously used for the wild-type virus [19] . impairing VLP assembly 339 We next aimed at producing fluorescent SARS-CoV-2 VLPs in HEK293T cells, by co- In order to accurately measure particle size with single molecule localization microscopy, we Figure 4B ). Although, since some VLPs were not GFP labelled, "red" dots also 407 appeared. In order to prove that these dots were more likely VLPs rather than 408 exosomes/extracellular vesicles (EVs) containing S, we checked the colocalization between 409 these "red" S dots and CD81 with an anti-CD81 exosomal marker ( confirming the results seen by immunoblotting ( Figure 1B) . We also checked that these VLPs VLPs production, requiring MNE/S with an optimized transfected plasmid ratio based on the 471 RNAseq ratio measured in infected cells [17] [18] . We show that the transfection and expression of M, N, E, or S alone are not sufficient to produce VLPs, as well as the 473 combination of only two structural proteins. Our conclusion is that M, E, and N is the best 474 condition, at a 3:12:2 ratio, required for optimal VLPs production (as shown by Kumar et al.

[10] and Boson et al. [14] ). We also observed that the addition of S reduces the 476 incorporation of N and, to a lesser extent M (as seen by immunoblots, Figure 1 and by dual 477 color labelled VLP, Figure 3 ) suggesting that the incorporation of the transmembrane Spike 478 in the VLPs reduces N incorporation in favor of S. Moreover, S incorporation was favored in 479 the optimized plasmid ratio and its maturation was much greater. Apparently, the presence 480 of the packageable genomic RNA was not required for VLPs formation since all these VLPs 481 were produced without. Some of these findings could be in discordance with other studies.

As a matter of fact, Xu et al. [5] showed that M alone is able to exit the cell, with this protein 483 being the driver and M and E the minimal system. Indeed, in Figure 1B al. [9] ) and previously in mammalian cells in the case of SARS-CoV [12] .

With the intention of optimizing both VLPs formation and production, we performed an 498 assay with an optimal ratio of the different structural proteins based on the viral RNA ratio 499 coding for the 4 main structural proteins found in infected pulmonary Calu3 or intestine 500

Caco2 cells [17] and in human pulmonary A549hACE2 cells [18] . Consequently, RNAseq data 501 analysis allowed us to establish a plasmid ratio of 3:12:2:5 for M:N:E:S. This ratio increases 502 VLPs production by 1.5 fold for MNE and MNES, as compare with the 3:3:3:5 ratio, with more S incorporation and more mature S2 on the VLP ( Figure 1C ), mimicking the wild-type 504 virus. 505 We confirmed that particles analyzed by AFM were genuinely VLPs and rather not 506 extracellular vesicles ( Figure 2C ). To confirm that aspect, we checked one exosomal marker, 507 CD81, using immunospotting on GFP-VLP and confirm very low (less than 5%) or no 508 colocalization between VLP and CD81(+) EVs (Supplementary Figure 3) . This suggests that 509 SARS-CoV-2 VLPs are exiting through a secretory pathway that is different from EVs. with AFM results ( Figure 1D ). It is to notice that VLPs size are slightly bigger than wild-type 519 virus particles as it has been described using AFM [19] , TEM [19] or cryoET [27] [28] [29] . for the wild-type virus [14] . These differences between our data and the ones from electron 528 microscopy could arise from the dehydration process required during TEM sample 529 preparation or from the absence of genomic RNA shaping the inside nucleocapsid N-RNA 530 core that might change particle assembly. The incorporation of other viral or cellular 531 proteins could also be considered for tuning particle assembly. In another study, McMahon 532 et al. [15] used high-throughput super-resolution microscopy to measure the size of the Interestingly, a recent article [29] is demonstrating that this type of VLPs can package viral 557 RNAs to transduce and express genes in target cells, which reinforces the strong potential of 558 SARS-CoV2 VLPs in vaccine development.

Overall, we report that the minimal system required for VLPs formation is M, N, and E 560 structural proteins incorporating S and that VLPs can be optimized in production and 

Influenza Virus Like Particles (VLPs): Opportunities for H7N9 Vaccine Development

HIV-1-Based Virus-like Particles that Morphologically Resemble Mature, Infectious HIV-1 Virions

Involvement of an Arginine Triplet in M1 Matrix Protein Interaction with Membranes and in M1 Recruitment into Virus-Like Particles of the Influenza A(H1N1)pdm09 Virus

Betacoronavirus Assembly: Clues and Perspectives for Elucidating SARS-CoV-2 Particle Formation and Egress

Construction of SARS-CoV-2 Virus-Like Particles by Mammalian Expression System

SARS-CoV-2 viral budding and entry can be modeled using BSL-2 level virus-like particles

Minimal system for assembly of SARS-CoV-2 virus like particles

Production of SARS-CoV-2 Virus-Like Particles in Insect Cells

Construction of SARS-CoV-2 virus-like particles in plant

Assembly and Entry of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV2): Evaluation Using Virus-Like Particles

The SARS-CoV-2 envelope and membrane proteins modulate maturation and retention of the spike protein, allowing assembly of virus-like particles

The M, E, and N Structural Proteins of the Severe Acute Respiratory Syndrome Coronavirus Are Required for Efficient Assembly, Trafficking, and Release of Virus-Like Particles

Single molecule localisation microscopy reveals how HIV-1 Gag proteins sense membrane virus assembly sites in living host CD4 T cells

Full assembly of HIV-1 particles requires assistance of the membrane curvature factor IRSp53

Illuminating the nanoscopic world of viruses by fluorescence super-resolution microscopy

Monitoring HIV-1 Assembly in Living Cells: Insights from Dynamic and Single Molecule Microscopy

A high-resolution temporal atlas of the SARS-CoV-2 translatome and transcriptome

Reorganization of F-actin nanostructures is required for the late phases of SARS-CoV-2 replication in pulmonary cells

Atomic force microscopy analysis of native infectious and inactivated SARS-CoV-2 virions

SARS-CoV-2 Unrevealed: Ultrastructural and Nanomechanical Analysis

Low Selectivity Indices of Ivermectin and Macrocyclic Lactones on SARS-CoV-2 Replication In Vitro

PyCorrFit-generic data evaluation for fluorescence correlation spectroscopy

ThunderSTORM: a comprehensive ImageJ plug-in for PALM and STORM data analysis and super-resolution imaging

SR-Tesseler: a method to segment and quantify localization-based super-resolution microscopy data

Topography, Spike Dynamics, and Nanomechanics of Individual Native SARS-CoV-2 Virions

Structures and distributions of SARS-CoV-2 spike proteins on intact virions

Morphometry of SARS-CoV and SARS-CoV-2 particles in ultrathin plastic sections of infected Vero cell cultures

A structural view of the SARS-CoV-2 virus and its assembly

Three-dimensional reconstruction by electron tomography for the application to ultrastructural analysis of SARS-CoV-2 particles

Highthroughput super-resolution analysis of influenza virus pleomorphism reveals insights into viral spatial organization

Rapid assessment of SARS-CoV-2-evolved variants using virus-like particles

Size and concentration of fluorescent and photoconvertible SARS-CoV-2 VLPs 616 (A) Western blot of cell lysates and VLPs of HEK293T cells transfected with M/M(GFP), N, E 617 ±S with the optimal plasmid ratio of M/M(GFP):N:E±S as 2/2:12:2±5. (B) TIRF microscopy 618 images of SARS-CoV-2 M(GFP)NE and M(GFP)NES VLPs deposited on a glass coverslip. Scale 619 bar is 5µm. (C) Correlative AFM and wide-field fluorescence images of fluorescent SARS-CoV-620

Scale bars represents 1µm. (D) Size and concentration of

VLPs determined by FCS. (E) Images and quantitative size

Figure 4 : Incorporation of the Spike S on fluorescent SARS-CoV-2 M(GFP)NES VLPs using 635 immuno-spotting coupled to TIRF-Microscopy. (A) TIRF images of M(GFP)NES VLPs

One point in the 641 graph represents the % of S incorporation in the VLP(GFP), i.e. % of colocalization in each 642 frame. t test was calculated and p-value between VLP M(GFP)NES and VLP M(GFP)NE is 643 0.0001 and p-value between VLP M(GFP)NES and No VLP is <0.0001. 644 conditions (n=16 cells) are respectively plotted as green and blue lines (mean) surrounded 653 by rectangle (sd)

01; **** p < 1.e10 -15 ; nd non 655 significative. (D) Confocal images of M(GFP)NES VLPs (in green) and A549-hACE2-mScarlet1 656 (in red) at 37°C (upper panel) and at 4°C (lower panels) showing internalization of VLPs

37°C (E) Percentage of colocalization of hACE2-mScarlet1 in M(GFP)NES VLPs (in red) and of

VLPs in hACE2-mScarlet1 (in green) showing that most of the labelled VLPs are 659 surrounded by hACE2-mScarlet1 receptor. (F) Image of a cell expressing

corresponding to Figure 1B. (B) Cell viability of assays based on mimicking WT SARS-CoV-2 674 mRNA ratio corresponding to Figure 1C

CD81(+) EVs and SARS-CoV-2 M(GFP)NES VLPs are distinct entities 677 as revealed by immuno-spotting coupled to TIRF-M. (A) Imaging of incorporation of the

with a neutralizing antibody anti-S 679 coupled with secondary AF555 antibody and an antibody anti-CD81 coupled with secondary 680

AF647using immuno-spotting coupled to TIRF-M, showing that M(GFP)NES can contain CD81 681 but CD81 exosomes are not containing M(GFP) or S. Scale bar is 5µm. (B) Percentage of 682 incorporation of M(GFP) on CD81(+) EVs showing that CD81(+) exosomes are not containing

M(GFP) or S. Scale bar is 5µm