key: cord-0804481-s7zeyrqa
authors: Galisova, Andrea; Zahradnik, Jiri; Allouche-Arnon, Hyla; Morandi, Mattia I.; Karam, Paula Abou; Avinoam, Ori; Regev-Rudzki, Neta; Schreiber, Gideon; Bar-Shir, Amnon
title: Genetically engineered MRI-trackable extracellular vesicles as SARS-CoV-2 mimetics for mapping ACE2 binding in vivo
date: 2022-03-28
journal: bioRxiv
DOI: 10.1101/2022.03.27.485958
sha: 6a48505b508c3a60911251c79139eed117c89f36
doc_id: 804481
cord_uid: s7zeyrqa

The elucidation of viral-receptor interactions and an understanding of virus-spreading mechanisms are of great importance, particularly in the era of pandemic. Indeed, advances in computational chemistry, synthetic biology, and protein engineering have allowed precise prediction and characterization of such interactions. Nevertheless, the hazards of the infectiousness of viruses, their rapid mutagenesis, and the need to study viral-receptor interactions in a complex in vivo setup, call for further developments. Here, we show the development of biocompatible genetically engineered extracellular vesicles (EVs) that display the receptor binding domain (RBD) of SARS-CoV-2 on their surface as coronavirus mimetics (EVsRBD). Loading EVsRBD with iron oxide nanoparticles makes them MRI-visible, and thus, allows mapping of the binding of RBD to ACE2 receptors non-invasively in live subjects. Importantly, the proposed mimetics can be easily modified to display the RBD of SARS-CoV-2mutants, namely Delta and Omicron, allowing rapid screening of newly raised variants of the virus. The proposed platform thus shows relevance and cruciality in the examination of quickly evolving pathogenic viruses in an adjustable, fast, and safe manner.

Virus-receptor recognition is the initial step in the infectious cycle and is considered to be a key stage in the induction of viral pathogenesis 1 . Therefore, the elucidation of the interactions of viruses with host cells' receptors is of paramount importance for a better understanding of pathology pathways and for the development of antiviral interventions. For example, it has been shown that the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which has caused the current and prolonged coronavirus disease 2019 (COVID-19) pandemic, specifically attacks cells expressing high levels of angiotensin-converting enzyme 2 (ACE2) 2 . The understanding of the interactions between the receptor binding domain (RBD) of the spike-S protein of the virus with ACE2 3 has resulted in the development of a wide range of efficient therapeutics and vaccines [4] [5] [6] [7] [8] . Unfortunately, SARS-CoV-2 has shown a remarkable ability to rapidly introduce mutations to the spike protein and the RBD for improved affinity and immune evasion [9] [10] [11] [12] , which have led to rapid spreading of more transmissible variants and compromised effectiveness of available vaccines 13 . Thus, it is clear that there is a need for the ability to characterize viruses and their evolving mutants quickly and safely and potentially even predict dangerous variants before they emerge. Indeed, in silico 14 and in vitro 15 examination of viruses provides crucial insights into virus-receptor interactions.

Nevertheless, these approaches are limited in the study of off-target binding events and are not applicable for spatial and real-time mapping of viral-receptor binding in deep tissues. This calls for a method with the ability to longitudinally and non-invasively monitor and map in vivo viral distribution and receptor binding in a safe and rapid way to enhance our ability to study emerging viruses and assess biological feedback to therapeutics.

Several types of non-viral nano-sized formulations have been proposed to elucidate viral-receptor interactions so far, including those for studying SARS-CoV-2 4, [16] [17] [18] . Among these, extracellular vesicles (EVs) offer several advantages over synthetic nanoparticles. First, as cellular content nanocarriers [19] [20] [21] , they are biological substances, suggesting that they can be introduced into the body without leading to the side effects often encountered with synthetic formulations. Second, they can be genetically engineered to present biomolecules on their surface, providing a rapid and general method for the display of peptides [22] [23] [24] . As such, EV targetability to a tissue of interest has been enhanced by displaying peptides that are not present on the surface of native EVs [25] [26] [27] [28] . Third, EVs share important similarities with enveloped viruses, including comparable sizes and host membrane compositions 29, 30 . EVs are thus attractive non-infectious vehicles with which to exploit viral uptake pathways for cellular cargo delivery 25, [31] [32] [33] , or for the development of EV-based vaccines 34 and related adjuvants 35 . For example, EVs presenting the coronavirus S protein or its RBD were proposed as potential vaccines already in the mid-2000s for SARS-CoV 36 , as well as for the current SARS-CoV-2 pandemic 37, 38 . Moreover they showed efficiency as decoys for neutralizing antibodies 39 and as systems for targeted delivery of antiviral agents 40 . In addition, the ease at which EVs can be genetically engineered makes these formulations ideal for rapid studies of emerging viral mutations as they appear.

Given that EVs can mimic viruses and can be labeled with imageable material 27, 41, 42 , EVs can potentially be used for non-invasive in vivo imaging of viral-receptor interactions. In fact, it has been shown that EVs can be fluorescently labeled and imaged in vivo; however, these fluorescent methods are unable to track EVs in deep tissues and offer limited spatial resolution 43 . In contrast, tracking of EVs with threedimensional imaging modalities (such as CT 41 and MRI 42 ) allows an assessment of their spatial distribution even in deep tissues. In this regard, MRI stands out due to its ability to provide spatial information from the introduced EVs that can be overlaid on high-resolution anatomical images of the same subject avoiding the need of using hybrid multimodal imaging approaches. Here, we show the design, development, and implementation of genetically-engineered EVs that display the RBD of SARS-CoV-2 (EVs RBD ) as coronavirus mimetics for studying RBD-ACE2 interactions. Magnetically-labeled EVs RBD allow mapping RBD-ACE2 binding in vivo and in real-time using a clinically translatable MRI setup. Moreover, we demonstrate the modifiability of the EV-based formulation by presenting the RBD of currently spreading SARS-CoV-2 variants -Delta and Omicron. The ability to monitor both in vivo biodistribution and the effect of different binding affinities of RBD to ACE2 in a fast and safe way highlights the potential of our approach in prolonged pandemic eras and for the study of other emerging viruses.

Several methods have already been implemented to genetically engineer EVs to display peptides on their surface as, e.g., in fusion with the Lamp2b EVs membrane protein 32,44 , the vesicular stomatitis virus G protein (VSVG) 40 , or the transmembrane domain of platelet-derived growth factor receptor (PDGFR) when using the pDisplay TM vector 27 . Although widely used, Lamp2b showed an inability to express the RBD of the SARS-CoV-2 on the membrane efficiently 40 . Starting from the pDisplay TM vector, which has been extensively used for heterologous expression and surface display of cell receptors in mammalian cells, we first engineered it for efficient expression of viral peptides on the surface of EVs. To this end, and with a purpose to create SARS-CoV-2 mimetics -EV RBD (Figure 1 ), we constructed a new pAGDisplay plasmid through a three-component assembly of fragments from the widely used pDisplay, pLVX-Puro, and pET26b plasmids. There are four main benefits of using our designed pAGDisplay plasmid over other alternatives. First, it uses the puromycin resistance marker, which is a more potent selective antibiotic compared to geneticin. Second, an IRES sequence, which was introduced downstream from the transmembrane domain to allow co-expression of the antibiotic resistance and the transmembrane domain under a single promoter restricting expression quenching. Third, the fluorescent protein eUnaG2 was fused with the C-terminus of PuroR, allowing for easy detection of transfected cells. Fourth, we introduced a RFnano protein, a MIRFPnano670 derivative (see the Materials and Methods section), at the C-terminus of the PDGFβ transmembrane domain to allow efficient selection of RBD-expressing cells using fluorescent activated cell sorting (FACS). Having designed the pAGDisplay plasmid for efficient and versatile surface display of cell receptors, the Wuhan RBD of SARS-CoV-2 was first cloned at the Nterminus preceding the transmembrane PDGFR domain. Specific tags were added to the obtained constructs for further validation of expression with an ALFA tag as a marker to the RBD construct and a Myc-tag to the control construct (referred to later in the text as noRBD). From left to right: A scheme of the pAGDisplay plasmid designed for this study. The released EVs display RBD (EV RBD ) and control EVs display no RBD (EV noRBD ) from HEK293 cells transfected with pAGDisplay-RBD and pAGDisplay-noRBD, respectively. On the right, a model of interaction between the RBD of the spike protein of the SARS-CoV-2 virus and the ACE2 receptor; RBD is highlighted in the black square). eUnaG2 -a green fluorescent protein; IRES -internal ribosomal entry site; RFnano -red fluorescent protein (MIRFPnano670); TM -transmembrane domain; L -linker; RBD -receptor binding site of SARS-CoV-2; ALFA -ALFA tag; Myc -Myc tag.

Human embryonic kidney 293 (HEK293) cells were transfected with pAGDisplay-RBD or pAGDisplay-noRBD followed by FACS to select cells associated with the highest expression levels of Having demonstrated the presence of RBD on the cell surface, we then tested its functionality. For this purpose, the recombinant extracellular portion of ACE2 protein was expressed, purified, and conjugated to a fluorophore CF640R (excitation 642 nm, emission 662 nm) followed by its incubation with either noRBD or RBD cells (Figure 2d -f). Fluorescence microscopy images showed a clear binding of the ACE2 protein only to RBD cells as a higher fluorescent signal compared to controls, confirming the presence of a functional RBD on the cell surface (Figure 2d) . Similarly, flow cytometry analysis showed a significantly higher fluorescent signal in RBD cells compared to noRBD cells (p < 0.001), corresponding to the bound ACE2 protein (Figure 2e and Figure S2 for FACS data). By incubating the cells with a range of different concentrations of fluorescently labeled ACE2 protein and performing a series of FACS experiments to determine its affinity to RBD at the cell surface, an equilibrium dissociation constant (KD) of 6.7 ± 1.3 nM was determined for RBD cells (Figure 2f ). This affinity of RBD expressed at the surface of HEK293 cells to the ACE2 protein is in accordance with the values reported for purified proteins or through a yeast-display assay 10, 45 .

Following the validation of the expression of functional RBD at the surface of HEK293 cells through ACE2 binding assessments (Figure 2 ), secreted EVs of these cells were obtained using a standard method of differential centrifugation of the cells' media (with an average 5×10 9 EVs/ one million cells) according to MISEV guidelines 46 . The isolated and purified EVs from RBD cells or control cells were referred to as EVs RBD , or EVs noRBD , respectively, and were further examined. Western blot analysis . Note here that the very low red fluorescence found in control cells after their incubation with the EVs RBD may be in accordance to the expected low expression levels of native ACE2 in HEK293 cells. Flow cytometry analysis showed that EVs RBD had significantly greater binding efficiency to ACE2-expressing cells (HEK ACE2 ) when compared to EV noRBD (p < 0.05), while the uptake into control cells (HEK) was similar for both EVs RBD and EV noRBD (Figure 3f-g and Figure S3 for FACS data). These results thus confirm the presence and functionality of the engineered RBD peptide on the EVs RBD surface and its targeting and binding capability to ACE2 receptors expressed on cells with no observable effect on their viability ( Figure S4 ). These findings show that the engineered EVs RBD can be used as SARS-CoV-2 mimetics with ACE2 binding capabilities with a similar size and shape to the SARS-CoV-2 virus. Importantly, this EV-based formulation, in contrast to a highly infectious SARS-CoV-2 virus 48 , can be used to study RBD-ACE2 interactions at a single-cell level under standard laboratory conditions without the need for strict safety regulations. This platform may, therefore, expand the research of viralreceptor interactions to research institutes and industrial setups that do not yet have access to dedicated facilities, which must be designated with the highest biosafety level required to work with SARS-CoV-2. 

The in vivo targetability of the EVs RBD mimetics toward ACE2 receptors was then studied in an animal model. Several studies have already shown the ability of engineered EVs to interact and bind to receptors of interest in vivo, including acetylcholine receptors in the brain 32 , tumor cells 27 , immune cell surfaces 49 , and viral receptors, including ACE2 40 . Since human ACE2 receptors are not naturally present in mice, a few transgenic mice models had been proposed 50,51 , but they are not readily accessible, and therefore, are not yet extensively used. Therefore, we established a xenograft model to study the in vivo targetability of EVs RBD to ACE2 following their systemic administration. For this purpose, control Figure S6 ).

Next, we studied the biodistribution of fluorescently DiR-labeled EVs after their intravenous administration in a dose of 3×10 11 EVs per mouse. Both types of EVs, EVs RBD and EVs noRBD , were detected mostly in the liver (Figure 4d and 4e and Figure S7 ), in good agreement with previous studies demonstrating that this is the main organ through which EVs are cleared from the body 24 . In contrast to the liver, much milder fluorescent signals were also obtained from excised spleen and kidney, implying that other clearance pathways might be involved in EVs biodistribution, but with no significant difference when comparing EVs noRBD and EVs RBD . As no human ACE2 receptors are expressed in the lungs of mice 52 , the fluorescent signals of excised lungs following the injection of either EVs noRBD or EVs RBD was found to be similar and very low, as expected. Overall, the comparable biodistribution profiles of EVs noRBD and EVs RBD (Figure 4d and 4e) indicate, once again, the specific targetability of EVs RBD to ACE2-expressing cells even in vivo after their systemic administration (Figure 4a and 4b) . 

After showing the specific targetability of EVs RBD to ACE-expressing cells both in vitro ( Figure 3) and in vivo (Figure 4) , we aimed to examine the ability to map the accumulation in their target cells using a non-invasive and three-dimensional imaging modality, such as MRI, which was demonstrated to be applicable for monitoring of magnetically labeled EVs in vivo 42, [53] [54] [55] . To this end, and to load EVs with superparamagnetic iron oxide nanoparticles (SPIONs), parental HEK-293 cells stably expressing the RBD ( Figure 1) were incubated for 24 hours with SPIONs added to their culture medium (40 µg of iron per mL).

Then, the incubating medium was replaced with exosome-free medium and the released EVs RBD were collected and purified by differential centrifugation (Figure 5a ) and further characterized. Incorporation of SPIONs into the secreted EVs RBD was clearly visualized by cryo-TEM as multiple hypointense clusters inside the lumen of the EVs (Figure 5b ). Solutions containing different concentrations of SPIONs-labeled EVs RBD were then examined by MRI and showed concentration-dependent T2*-weighted MRI contrast ( Figure 5c) . Importantly, labeling of EVs with SPIONs did not compromise their size (84.8 ± 4.3 nm vs.

94.8 ± 3.4 nm, Figure 5d ), charge (-18.8 ± 3.9 mV vs. -16 ± 1.0 mV, Figure 5e ), or the expression of the RBD (Figure 5f ).

After successful labeling of the EVs with SPIONs, we aimed to examine the MRI-detectability of their ACE2 targetability in vivo. To this end, a solution containing 3×10 11 SPIONs-labeled EVs RBD was injected systematically through the mouse tail vein. The preferential accumulation of the SARS-CoV-2 mimetics in ACE2-expressing tissue was observed as a reduced MRI signal intensity in T2*-weighted images (Figure 5g ). To quantify the change in the MRI signal of the two tumorous-like tissues (HEK vs.

HEK ACE2 ), mice were scanned before and four hours after EVs administration and the difference in the tissue contrast-to-noise ratio (CNR) obtained before and after injection was calculated. As shown in Figure   5h , a significant increase in (p-value < 0.01, n=5) CNR difference (before vs. after SPIONs-labeled EVs RBD injection) was obtained in the region of interest (ROI) of the ACE2-expressing cells compared to control tissue. Moreover, Prussian blue staining of slices of the excised tissues clearly showed iron deposits only at the samples obtained from ACE2-expressing cells (Figure 5i) , confirming the delivery of the SPIONslabeled EVs RBD to HEK ACE2 , but not to the controls. Examined tumors: control (CTRL) or HEK ACE2 , and color-coded images of tumors overlaid on anatomical MR images of mice (gray). (h) Quantification of contrast-to-noise ratios (CNR) in control and ACE2 tumors after EVs RBD injection (n=5, p-value=0.009). CNR was calculated as the ratio between the signal of the tumor ROI and a muscle ROI. CNR before injection was set at 100%. (i) Histological analysis of tumor tissues after Prussian blue staining for iron deposition (blue color). Scale bar represents 100 µm. The inset shows a magnified image of the tissue with accumulated iron in blue. Data are presented as mean values ± s.d. Statistics: two-tailed unpaired Student's t-test with *p-value<0.05 and **p-value<0.01.

As for other RNA viruses, so for SARS-CoV-2, the rapid evolution and the introduction of diverse mutations at the RBD manipulates not only the binding affinity to the host cell receptors and the infectivity of the virus, but also reduces the efficiency of proposed vaccines and other therapeutics. In that regard, one key feature of the SARS-CoV-2 mimetics presented here is that suspicious mutations or those found in rapidly spread variants of a virus of interest can be easily displayed at the surface of the EVs through a onestep cloning procedure into the pAGDisplay plasmid. This means that the method can provide rapid data on novel mutations as they appear. To examine this, we have designed and tested several mimetics that represent different binding affinities to the ACE2 receptor, namely the Wuhan, Delta and Omicron variants of SARS-CoV-2 ( Figure 6 and Supplementary Sequences). The genes encoding for the mutated peptides were designed and cloned into the expression plasmid and the parental cells were transiently transfected. After EVs isolation to obtain EVs RBD-Wuhan , EVs RBD-Delta and EVs RBD-Omicron , the difference in their binding capabilities to ACE2 receptors expressed at the surface of live cells was examined using flow cytometry. As shown in Figure 6 , and as expected, HEK ACE2 cells incubated with engineered EVs displaying the Delta variant of RBD showed similar fluorescence when compared to cells incubated with the Wuhan RBD, implying on similar binding capabilities. In contrast, HEK ACE2 cells incubated with EVs presenting the Omicron variant on their surface (EVs RBD-Omicron ) significantly higher fluorescence was obtained, as expected for RBD with a stronger binding affinity to ACE2 56,57 . These results demonstrate that the proposed genetically engineered EVs, which are here used as coronavirus mimetics, can be used as a reliable platform for fast and safe examination of evolved mutations of SARS-CoV-2, with the potential to be extended to other viruses.

In summary, we showed the design and implementation of genetically engineered EVs as SARS-CoV Invitrogen (V66020). The pAGDisplay vector backbone was assembled by combining a pET28b fragment bearing KanR and origin of replication, a pLVX vector fragment bearing WPRE, PuroR, and IRES sequences, and a pDisplay CMV promoter with a PDGFRβ expression cassette by a restriction-free threecomponent assembly 58 . In the subsequent restriction-free cloning step, the PuroR was fused with eUnaG2 fluorescent protein at the C-terminus 59,60 . The full-length pAGDisplay was sequenced to verify its correct assembly.

pAGDisplay modification with RFnano near-infrared fluorescence protein: To increase our spatial resolution, we introduced, by restriction-free cloning, a modified near-infrared miRFP670nano 61 , an RFnano protein, as a cytoplasmic domain at the C-terminus of PDGFRβ. The preparation and design of RFnano is the subject of a publication currently in preparation. Briefly, the PROSS and Rosetta-based stabilization design was combined with S.cerevisiae EBY100 expression and sorting to obtain a brighter signal and the same spectral parameters. In total, 23 mutations were introduced into the protein.

Protein production, purifications, and labeling procedures: The designed ALFA-tag binding nanobody (DnbALFA) and its mNeonGreen fusion were expressed by using expression plasmid pET28bdSUMO 62 and E.coli BL21(DE3) cells as described previously 60 . Briefly, 200 ml of 2YT medium (16 g tryptone, 10 g yeast extract and 5 g NaCl, pH 7) was inoculated (1%), grown to the OD600 = 0.6 (37°C), and the Fluorescence microscopy: For visualization by fluorescence microscopy, cells were seeded on 14 mmdiameter coverslips in a 24-well plate. The wells were coated with fibronectin by incubation for 45 min.

Then, fibronectin was removed and the wells were washed with PBS before seeding the cells. For uptake of EVs in cells, the DiD-labeled EVs were dissolved in exosome-free medium and incubated with cells for three hours. Then, the medium was removed, and cells were washed two times with PBS. For staining of cell nuclei, DAPI was dissolved in 2.5% formaldehyde and the cells were incubated in the solution for 20 min. After two washes, the coverslips were carefully transferred and mounted on glass slides and the fluorescence was visualized using a wide-field microscope (Leica DMI8). Production of magnetically labeled EVs: HEK293 cells stably expressing RBD or without RBD (control) were seeded into 15 cm culture plates at 50% confluency. The next day, SPIONs (Molday ION Dye Free; 2 mg Fe/mL; BioPal, USA, CL-50Q02-6A-0) were added to the culture medium at a final concentration of 40 μg/mL. Then, 24 hours after, medium was discarded, the cells were washed three times with 10 mL of PBS and EV-depleted medium was added to the cells followed by another 24-hour incubation. Afterward, EVs were isolated from the medium according to the standard protocol and resuspended in PBS.

Animal model: Immunodeficient Hsd:Athymic Nude-Foxn1nu mice (Envigo) were used in the experiments. All animal studies were approved in accordance with the Weizmann Institute's Animal Care and Use Committee (IACUC) guidelines and regulations (approval number 00580120-3). All animals were kept in a daily controlled room at the Weizmann Institute of Sciences animal facility with a surrounding relative humidity level of 50 ± 10% and a temperature of 22 ± 1 °C, with a 12/12 cycle of dark and light phases. Subcutaneous xenograft tumors were induced by injection of HEK293T (left side) and ACE2expressing HEK293T cells (right side) under the skin above the mouse flank. Mice were kept under isoflurane anesthesia during the whole procedure. Each mouse received 10 mil of each cell type in 150 μL of PBS. The tumors were allowed to grow for two to three weeks until they reached a sufficient size for imaging (diameter around 0.5 -1 cm). If the tumors reached more than 1 cm or the tumor size was not equal, the animals were removed from the experimental groups.

MRI experiments were conducted on a horizontal 15.2 T horizontal scanner (in vitro and in vivo) ausing a 1 H volume coil rf with a 23 mm diameter (Bruker BioSpin, Germany).

Phantom measurements: Different concentrations of isolated EVs were added to glass tubes and imaged using a RARE (Rapid Acquisition with Relaxation Enhancement) sequence with the following parameters: time of repetition (TR) 2800 ms; echo time (TE) 42 ms; and resolution 0.23×0.23×1 mm 3 . MR images were analyzed by the ImageJ software or by a custom-made script written in MATLAB (MathWorks, USA).

In vivo MRI of mice: Prior to MR imaging, a cannula was inserted into a mouse tail. Mice with induced tumors were then measured on a 15.2T MRI scanner before and after injection of SPIONs-labeled EVs RBD at a dose of 3×10 11 EVs in 100 μL of PBS administered through a long tube without changing the position of the animal. The mice were measured under general isoflurane inhalation anesthesia (5% induction, 1% maintenance) up to four hours after EVs injection using a gradient echo Fast Low Angle Shot (FLASH) with the following parameters: TR = 300 ms; TE = 2 ms; resolution of 0.23×0.23×0.7 mm 3 . After tuning and matching to the 1 H frequency, shimming of the magnetic field, and B0 correction, the tumor area was measured with the following parameters: both axial and coronal images were acquired; T2 and T2* maps were reconstructed in the Paravision software; and were followed by analysis in the ImageJ software.

Regions of interest (ROI) were outlined around each tumor, muscle, and a noise area outside of the animals.

The contrast-to-noise ratios (CNR) were calculated as follows: CNR = (Stumor-Smuscle)/SDnoise CNR difference was calculated as a percentage of CNR before and after EV injection.

Fluorescence imaging: Prior to imaging, mice with induced subcutaneous tumors were retro-orbitally injected with DiR-labeled EVs RBD or EVs noRBD (dose of 3×10 11 EVs in 100 μL of PBS). Six hours after injection, the mice were intracardially perfused with 2.5% formaldehyde solution under general anesthesia induced by ketamine (80 mg/kg) and medetomidine (0.6 mg/kg). Organs (liver, kidneys, spleen, intestine, heart, lungs, brain) and tumors were excised, fixed with 2.5% formaldehyde solution overnight, and then kept in PBS. A day after, fixed organs were measured with IVIS Lumina XR optical images (Perkin Elmer, USA) with the FOV and exposure time of 1 s (liver), 10 s (kidney, lungs, spleen, heart), 30 s (brain), or 60 s (tumors). For intravital microscopy, an Olympus microscope MVX was used with exposure times of 700 ms (Cy7 filter for detection of the DiR signal).

The fixed tumor tissues were kept in 1% formaldehyde solutions and then embedded in paraffin blocks according to a standard protocol. Tissue slices were cut on a microtome and stained by hematoxylin&eosin and Prussian blue. The slides were imaged on a Leica DMI8 wide-field fluorescent microscope using a standard bright-field filter.

Statistical analysis: All numerical data are presented as mean±standard deviation (s.d.). Statistical analysis was performed using GraphPad Prism 8.0 software (GraphPad Software Inc., USA). Comparison of two groups was analyzed by a two-tailed Student's t-test. A p-value of 0.05 and below was considered significant: *p-value<0.05, ** p-value <0.01, ***, p-value <0.001 and **** p-value <0.0001.

Virus entry: open sesame

Cell entry mechanisms of SARS-CoV-2

SARS-CoV-2 Receptor ACE2 Is an Interferon-Stimulated Gene in Human Airway Epithelial Cells and Is Detected in Specific Cell Subsets across Tissues

Cell-mimicking nanodecoys neutralize SARS-CoV-2 and mitigate lung injury in a non-human primate model of COVID-19

Targeting ACE2-RBD Interaction as a Platform for COVID-19

Therapeutics: Development and Drug-Repurposing Screen of an AlphaLISA Proximity Assay

Inhibition of SARS-CoV-2 Infections in Engineered Human Tissues Using Clinical-Grade Soluble Human ACE2

Engineering human ACE2 to optimize binding to the spike protein of SARS coronavirus 2

BNT162b2 vaccine induces divergent B cell responses to SARS-CoV-2 S1 and S2

Detection of a SARS-CoV-2 variant of concern in South Africa

Deep Mutational Scanning of SARS-CoV-2 Receptor Binding Domain Reveals Constraints on Folding and ACE2 Binding

Enhanced fusogenicity and pathogenicity of SARS-CoV-2 Delta P681R mutation

An infectious SARS-CoV-2 B.1.1.529 Omicron virus escapes neutralization by therapeutic monoclonal antibodies

mRNA booster immunization elicits potent neutralizing serum activity against the SARS-CoV-2 Omicron variant

In silico comparison of SARS-CoV-2 spike protein-ACE2 binding affinities across species and implications for virus origin

In Vitro and In Vivo Models for Studying SARS-CoV-2, the Etiological Agent Responsible for COVID-19 Pandemic

Quantum Dot-Conjugated SARS-CoV-2 Spike Pseudo-Virions Enable Tracking of Angiotensin Converting Enzyme 2 Binding and Endocytosis

Nanotechnology for COVID-19: Therapeutics and Vaccine Research

Nanoparticles Mimicking Viral Cell Recognition Strategies Are Superior Transporters into Mesangial Cells

Pre-metastatic niches: organ-specific homes for metastases

Specificities of secretion and uptake of exosomes and other extracellular vesicles for cell-to-cell communication

Extracellular vesicles in parasite survival

Visualization and tracking of tumour extracellular vesicle delivery and RNA translation using multiplexed reporters

Reprogramming Exosomes as Nanoscale Controllers of Cellular Immunity

Dynamic Biodistribution of Extracellular Vesicles in Vivo Using a Multimodal Imaging Reporter

Extracellular Vesicles Exploit Viral Entry Routes for Cargo Delivery. Microbiol

Microvesicle-associated AAV vector as a novel gene delivery system

Systemically injected exosomes targeted to EGFR deliver antitumor microRNA to breast cancer cells

A doxorubicin delivery platform using engineered natural membrane vesicle exosomes for targeted tumor therapy

Virus-Mimetic Fusogenic Exosomes for Direct Delivery of Integral Membrane Proteins to Target Cell Membranes

Extracellular vesicles and viruses: Are they close relatives?

Pseudotyping exosomes for enhanced protein delivery in mammalian cells

Delivery of siRNA to the mouse brain by systemic injection of targeted exosomes

Human cytomegalovirus-infected cells release extracellular vesicles that carry viral surface proteins

Release of Staphylococcus aureus extracellular vesicles and their application as a vaccine platform

Exosome-based tumor antigens-adjuvant co-delivery utilizing genetically engineered tumor cell-derived exosomes with immunostimulatory CpG DNA

Exosomal vaccines containing the S protein of the SARS coronavirus induce high levels of neutralizing antibodies

A bacterial extracellular vesicle-based intranasal vaccine against SARS-CoV-2 protects against disease and elicits neutralizing antibodies to wild-type virus and Delta variant

Extracellular vesicle-based vaccine platform displaying native viral envelope proteins elicits a robust anti-SARS-CoV-2 response in mice

Extracellular vesicles carry SARS-CoV-2 spike protein and serve as decoys for neutralizing antibodies

Tagged extracellular vesicles with the RBD of the viral spike protein for delivery of antiviral agents against SARS-COV-2 infection

In vivo Neuroimaging of Exosomes Using Gold Nanoparticles

Highly efficient magnetic labelling allows MRI tracking of the homing of stem cellderived extracellular vesicles following systemic delivery

In vivo imaging and tracking of exosomes for theranostics

Exosomes Engineered to Express a Cardiomyocyte Binding Peptide Demonstrate Improved Cardiac Retention in Vivo

SARS-CoV-2 variant prediction and antiviral drug design are enabled by RBD in vitro evolution

Minimal information for studies of extracellular vesicles 2018 (MISEV2018): a position statement of the International Society for Extracellular Vesicles and update of the MISEV2014 guidelines

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

Characteristics of SARS-CoV-2 and COVID-19

Genetically Engineered Cell-Derived Nanoparticles for Targeted Breast Cancer Immunotherapy

A humanized mouse model of chronic COVID-19

SARS-CoV-2 infection of human ACE2-transgenic mice causes severe lung inflammation and impaired function

Emerging SARS-CoV-2 variants expand species tropism to murines

Magnetic Resonance Imaging of Ultrasmall Superparamagnetic Iron Oxidelabeled Exosomes from Stem Cells: a new method to obrain labeled exosomes

Magnetic Resonance Imaging of Melanoma Exosomes in Lymph Nodes

NRP-1 targeted and cargo-loaded exosomes facilitate simultaneous imaging and therapy of glioma in vitro and in vivo

SARS-CoV-2 Omicron-B.1.1.529 leads to widespread escape from neutralizing antibody responses

Broadly neutralizing antibodies overcome SARS-CoV-2 Omicron antigenic shift

Application of the Restriction-Free (RF) cloning for multicomponents assembly

Fusion of fluorescent protein to puromycin N-acetyltransferase is useful in Drosophila Schneider S2 cells expressing heterologous proteins

A Protein-Engineered, Enhanced Yeast Display Platform for Rapid Evolution of Challenging Targets

Smallest near-infrared fluorescent protein evolved from cyanobacteriochrome as versatile tag for spectral multiplexing

Flexible regions govern promiscuous binding of IL-24 to receptors IL-20R1 and IL-22R1

A new set of highly efficient, tag-cleaving proteases for purifying recombinant proteins

Automated electron microscope tomography using robust prediction of specimen movements

This project received funding from the following sources: European Research Council (ERC-StG no.