In this protocol, quantum dots conjugated to recombinant SARS-CoV-2 spike enable cell-based assays to monitor spike binding to hACE2 at the plasma membrane and subsequent endocytosis of the bound proteins into the cytoplasm.
The development of new technologies for cellular fluorescence microscopy has facilitated high-throughput screening methods for drug discovery. Quantum dots are fluorescent nanoparticles with excellent photophysical properties imbued with bright and stable photoluminescence as well as narrow emission bands. Quantum dots are spherical in shape, and with the proper modification of the surface chemistry, can be used to conjugate biomolecules for cellular applications. These optical properties, combined with the ability to functionalize them with biomolecules, make them an excellent tool for investigating receptor-ligand interactions and cellular trafficking. Here, we present a method that uses quantum dots to track the binding and endocytosis of SARS-CoV-2 spike protein. This protocol can be used as a guide for experimentalists looking to utilize quantum dots to study protein-protein interactions and trafficking in the context of cellular physiology.
Fluorescence microscopy enables researchers to peer into the inner workings of the cell using specialized dyes1, genetically encoded fluorescent proteins2, and fluorescent nanoparticles in the form of quantum dots (QDs)3. For the severe acute respiratory syndrome coronavirus of 2019 (SARS-CoV-2) global pandemic, researchers have employed fluorescence microscopy to understand how the virus interacts with the cell both at the plasma membrane and in the cytoplasm. For example, researchers have been able to gain insights into the binding of the SARS-CoV-2 Spike protein on the virion's surface to human angiotensin-converting enzyme 2 (hACE2) on the surface of human cells, subsequent internalization via fusion at the plasma membrane, and endocytosis of the Spike:hACE2 protein complex4,5. Great insights have also been gained into the SARS-CoV-2 egress from cells via the lysosome using cellular fluorescence imaging, a unique feature of coronaviruses previously thought to occur via traditional vesicle budding from the Golgi, as it is with many other viruses6. A mainstay of almost all aspects of biological research, the cellular fluorescence microscopy technique has necessarily advanced in its breadth and scope of applications from super-resolution imaging of whole animals to automated high-content multi-parametric imaging for drug screening. Here, automated high-content confocal microscopy is applied to the study of SARS-CoV-2 cell entry using fluorescent QDs conjugated to the viral spike protein.
High-content analysis of images generated by biological imaging platforms allows for greater extraction of valuable biological insights than single parameters such as whole-well intensity, that one would obtain using a multi-modal plate reader7. By separating the objects in a field of view using automated segmentation algorithms, each object or a population of objects can be analyzed for parameters such as intensity, area, and texture in each available fluorescence channel8. Combining many measurements into multivariate datasets is a useful approach for phenotypic profiling. When the desired phenotype is known, such as QD internalization in the form of puncta, one can use the measurements related to puncta such as size, number, and intensity to assess the efficacy of a treatment.
Cloud-based high content imaging analysis software can accommodate a large variety of instrument data outputs, including the high content imaging platform. By using a cloud-based server for image storage and online analysis, the user is able to upload their data from either the imaging instrument or from the network drive where the data is stored. The analysis portion of the protocol is conducted within the cloud software environment, and data can be exported in a variety of file formats for downstream data visualization.
The SARS-CoV-2 virus is composed of nonstructural and structural proteins that aid in its assembly and replication. SARS-CoV-2 spike has two domains called S1 and S2, with S1 containing the receptor-binding domain responsible for hACE2 interactions at the plasma membrane9. Spike has also been found to interact with other molecules at the plasma membrane that may act as co-receptors in addition to hACE210,11. Throughout the spike protein sequence and particularly at the S1/S2 interface, there are protease cleavage sites that enable fusion at the membrane after the transmembrane serine protease 2 (TMPRSS2)12. Various recombinant SARS-CoV-2 Spike proteins have been produced from individual receptor binding domains, to S1, S2, S1 with S2, and whole spike trimers from multiple commercial vendors for use in research activities13.
In this work, the surface of QDs was functionalized with recombinant spike trimers that contain a histidine tag (QD-Spike). The QDs produced by Naval Research Laboratory Optical Nanomaterials Section contain a cadmium selenide core and a zinc sulfide shell14,15. The zinc on the QD surface coordinates the histidine residues within the recombinant protein to form a functionalized QD that resembles a SARS-CoV-2 viral particle in form and function. The generation of the nanoparticles and protein conjugation was previously described using the QD-conjugated receptor binding domain15. This method describes the cell culture preparations, QD treatment, image acquisition, and data analysis protocol that can guide a researcher in studying SARS-CoV-2 Spike activity in the physiological context of a human cell.
The HEK293T cell line used in this study is an immortalized cell line. No human or animal subjects were used in this study.
1. Cell culturing and seeding
2. Treatment of cells with QD-Spike
3. Fixation and nuclei staining
4. Acquisition set-up and imaging
5. Data analysis
6. Export data
7. Analyze the data in a spreadsheet
Upon treatment, the QDs will be internalized as the nanoparticle will bind to ACE2 on the plasma membrane and induce endocytosis. Using an ACE2-GFP expressing cell line, translocation of both QDs and ACE2 can be visualized using fluorescence microscopy. Once internalized, the two QD and ACE2 signals show strong colocalization. From these images, image segmentation and subsequent analysis can be performed to extract relevant parameters such as spot count (Figure 1, Figure 2B). Figure 1A is a montage of cells treated with different concentrations of QD-Spike or media only as a control. Three channels were used, including digital phase contrast, FITC, and the custom QD channel with excitation at 405 nm and emission at 608 nm. DPC provides an image of the general cell shape. The DPC image provides an approximate indication of the entire cell morphology. The area of interest overlaps with the DPC signal and is sufficient for detecting internalized QDs. The FITC channel shows the ACE2-GFP changing localization and accumulating at regions that colocalize with QD-Spike. As the concentration of QD-Spike decreases, the QDs are no longer visible and the ACE2-GFP signal is similar to control. The merge channel demonstrates the colocalization of ACE2-GFP with QD-Spike. These objects are a mixture of spots and larger accumulations that can be segmented with the analysis software. Fine-tuning of the software analysis method used for image segmentation can be used to segment the objects of interest.
Here a concentration-response experiment with six different concentrations of QD-Spike was performed, starting at 20 nM, to determine optimal concentrations of QD (Figure 2A). QDs can be used as low as 2.22 nM and still show binding and internalization. However, it is recommended to use concentrations of 20 nM or higher to ensure a robust response.
During conjugation to the QD, protein aggregation may occur. The main issue with aggregates is that they will accumulate on top of cells and cause artifacts in the image segmentation step. Aggregates will not be able to enter cells, and the nanoparticles as a whole will no longer resemble a viral particle (Figure 3). Aggregates can be spotted in the QD solution as bright, clumped precipitates.
This assay can also be used to assess biologics, such as neutralizing antibodies that block viral entry. QDs were incubated with neutralizing antibodies (Figure 4A), starting at 30 µg/mL, for 30 min at room temperature before addition to cells. Antibodies raised against the reference Washington strain, SARS-CoV-2 RBD, were used. They blocked binding, internalization, and caused a reduction in the measured spot count compared to cells treated with QD-only (Figure 4B).
Figure 1: High-content image analysis and segmentation of ACE2-GFP cells treated with QD608-Spike. Representative images of accurate segmentation. First, nuclei are segmented from the channel containing the nuclear marker to create an ROI population. From the ROI population, an ROI region outlining the cell is segmented using the ACE2-GFP channel. Lastly, QD608-Spike spots are segmented from the Cell ROI region. Please click here to view a larger version of this figure.
Figure 2: QD608-Spike binds hACE2 and is endocytosed in hACE2-GFP HEK293T cells. (A) Image montage of hACE2-GFP (yellow) HEK293T cells treated with several concentrations of QD608-Spike (magenta) or media-only. Digital phase contrast (cyan) was used to visualize the cell body. Scale bar: 20 µm. (B) High content analysis of QD608-Spike Spot Counts normalized to 20 nM QD608-Spike (100%) and the control (0%). N = triplicate wells. Error bars indicate standard deviation (S.D.). Please click here to view a larger version of this figure.
Figure 3: Aggregated QD608-Spike is not internalized into hACE2-GFP HEK293T cells. Image montage of hACE2-GFP HEK293T cells treated with 10 nM QD608-Spike or media-only. Scale bar: 20 µm. Please click here to view a larger version of this figure.
Figure 4: Neutralizing antibody blocks endocytosis of QD608-Spike in hACE2-GFP HEK293T cells. (A) Image montage of hACE2-GFP (yellow) HEK293T cells treated with QD608-Spike (magenta) preincubated with decreasing concentrations of neutralizing antibodies, using digital phase contrast (cyan) to identify cell bodies. Scale bar: 20 µm. (B) High content analysis of QD608-Spike Spot Counts normalized to media-only control (100%) and 10 nM QD608-Spike only (0%). N = triplicate wells. Please click here to view a larger version of this figure.
The method described in this article provides the necessary steps for imaging functionalized QDs in human cells using high-throughput confocal microscopy. This method is best suited for cells where endocytosis is the main route of viral entry rather than the activity of TMPRSS2 and membrane fusion, as it enables the study of SARS-CoV-2 Spike and hACE2 endocytosis. Because of the nature of the QD model and the C-terminal His-tag on the commercially available Spike trimer, any TMPRSS2 cleavage of Spike S1 and S2 domains would leave the QD attached to the S2 domain only12. This may prevent internalization, given that the RBD is found in S1. Therefore, if the sequence of events at the cell surface were precise, where hACE2 is bound and then TMPRSS2 cleaves Spike, a negative signal with no internalization is expected.
As this protocol deals with imaging cellular processes, the cultured cells must be permissive to viral infection with the expression of hACE2. hACE2 may be transiently transfected into cells, or a stable cell line expressing hACE2 may be generated17. It is recommended to verify hACE2 expression and localization using immunofluorescence with antibodies against hACE2 unless the hACE2 has a fluorescent protein tag such as GFP that can be observed using microscopy. GFP-tagged hACE2 confers the added benefit of visualizing hACE2 trafficking following QD-Spike endocytosis. Some cell lines with endogenous hACE2 expression may be used, but this should be confirmed using immunocytochemistry. In some cases, the endogenous expression does not provide enough hACE2 binding partners for QD-Spike and may result in a signal that is poorly detected by high-throughput confocal microscopy.
One critical step in the protocol includes the procurement of high-quality QDs conjugated to the recombinant His-tagged SARS-CoV-2 Spike. The prerequisite for this is a properly purified protein that can be conjugated to the QDs without causing aggregation. Aggregated QDs will have several problems that prevent a successful experiment. Therefore, testing of QD-Spike using analysis tools (e.g., UV-visible spectroscopy, transmission electron microscopy, or dynamic light scattering) prior to full experimentation is highly recommended to preserve valuable resources if testing precious reagents16. During the labeling of QDs with spike, specific concentrations of QD and Spike are mixed to achieve a specific ratio of molecules. Only QD and Spike are mixed, and both solutions are highly purified. To assess the labeling of QDs, an acrylamide gel can be cast and loaded with QD alone as well as QD-conjugated to Spike, which results in heavier molecular weight bands.
The QDs used in this protocol have a fluorescence excitation/emission spectrum tuned to 608 nm emission following UV excitation (≤405 nm) since most QDs have high absorption near the UV range producing high photoluminescence18. This is an unconventional excitation/emission combination that requires a microscope to have customizable channels. Many traditional confocal microscopes can be set up with the right filters and laser lines to achieve this excitation/emission. Alternatively, excitation of the QD at the first absorption maximum, approximately 10 to 20 nm away from the emission peak (e.g., 592 nm for QD608 used here), will also be able to produce sufficient photoluminescence.
The cloud-based software used in this high-content analysis protocol uses a naming scheme that builds on previous steps. For example, the first objects that are segmented are nuclei, which create a population called Nuclei. Following this, the cell or cytoplasm can be identified as an ROI region within the population Nuclei. The terminology used in the image analysis software sets the name of the population of objects segmented using the nucleus building block as Nuclei. However, they do not necessarily have to be nuclei and can be cell bodies if no nuclear dye is available. This naming scheme can also be changed and customized within each building block output name.
Our protocol did not account for membrane interactions, but this could be done by adding an additional membrane stain that is independent of ACE2-GFP trafficking. To block nonspecific binding, 0.1% BSA is included in the assay media (DMEM + 0.1% BSA), and the cells are grown in 10% FBS as well. Mutant spike proteins could be used, but we were limited to commercially available spike proteins. In this case, a SARS-CoV-2 mutant non-ACE2 binding Spike protein was not available.
The QD nanoparticle method presents a powerful technology to study the binding and internalization of viruses that rely on Spike-mediated entry. This assay can be used for high-throughput screening in an analogous manner, as shown in Figure 4, to repurpose and identify potent antivirals that block cellular entry. While this protocol only demonstrated the use of QDs conjugated to the Washington WA-1 reference strain of SARS-CoV-2, this assay can be easily adapted to study the binding properties of newly emerging variants such as Alpha, Beta, Gamma, and Delta through the use of their respective spike proteins conjugated to QDs.
The authors have nothing to disclose.
This research was supported in part by the Intramural Research Program of the National Center for Advancing Translational Sciences, NIH. Naval Research Laboratory provided funding via its internal Nanoscience Institute. Reagent preparation was supported via the NRL COVID-19 base fund.
32% Paraformaldehyde | Electron Microscopy Sciences | 15714 | Used for fixing cells after quantum dot treatment, final concentration 3.2% Used for stabilizing QDs in Optimem I and preventing non-specific interactions, final concentration 0.1% |
7.5% Bovine Serum Albumin | Gibco | 15260-037 | Used as a cell viability dye for fluorescence cell counting |
Acridine Orange / Propidium Iodide Stain | Logos Biosystems | F23001 | Microwell plates used for seeding cells and assaying QD-Spike |
Black clear bottom 96 well coated plate coated with poly-D-lysine | Greiner | 655946 | Used to support cell culture, DMEM supplement |
Characterized Fetal Bovine Serum | Cytiva/HyClone | SH30071.03 | Cloud-based high-content image analysis software; V2.9.1 |
Columbus Analyzer | Perkin Elmer | NA | Used for labeling cell nuclei and cell bodies after fixation, deep red nuclear dye |
DRAQ5 (5 mM) | ThermoFisher Scientific | 62252 | Basal media for HEK293T cell culture |
Dulbecco's Minimal Essential Media, D-glucose (4.5g/L), L-glutamine, sodium pyruvate (110 mg/L), phenol red | Gibco | 11995-065 | Used for arranging data after export from Columbus; V2110 Microsoft 365 |
Excel | Microsoft | NA | Used to continue selection of hACE2-GFP positive cells, DMEM supplement |
G418 | InvivoGen | ant-gn-5 | Human embryonic kidney cell line stably expression human angiotensin converting enzyme 2 tagged with GFP |
HEK293T hACE2-GFP | Codex Biosolutions | CB-97100-203 | Automated cell counter |
Luna Automated Cell Counter | Logos Biosystems | NA | Used for fluorescence cell counting |
Luna Cell Counting Slides | Logos Biosystems | L12001 | High-content imaging platform |
Opera Phenix | Perkin Elmer | NA | Imaging media, used for incubating cells with quantum dots |
Opti-MEM I Reduced Serum Medium | Gibco | 11058-021 | Phosphate-buffered saline without calcium or magnesium used for washing cells during passaging and assaying |
PBS -/- | Gibco | 10010-023 | Used to prevent bacterial contamination of cell culture, DMEM supplement |
Penicillin Streptomycin | Gibco | 15140-122 | Used for graphing, data visualization, and statistical analysis;V9.1.0 |
Prism | GraphPad | NA | Used for assaying SARS-Cov-2 Spike binding to hACE2 and monitoring Spike endocytosis |
Quantum Dot 608 nm-Spike (QD608-Spike) | custom made by Naval Research Laboratory | Used for inhibition of SARS-Cov-2 Spike binding to hACE2 | |
SARS-CoV-2 (2019-nCoV) Spike Neutralizing Antibody, Mouse Mab | Sino Biological | 40592-MM57 | Used to dissociate cells from flask during passaging |
TrypLE Express | Gibco | 12605-010 |