Establishment of an in vitro ACE2-expressing cell capturing system
To establish an in vitro ACE2-expressing cell capturing system, we first generated a cell line that stably expresses human ACE2. For this purpose, HEK293T cells were transfected with an ACE2-GFPfusion plasmid, allowing the visualizing of the expression of ACE2. As shown in Fig. 1, ACE2-GFP could be detected on the surface of the transfected HEK293T, and the expression of ACE2 was further confirmed by western blotting.
In the next step, we investigated whether HEK293T/ACE2-GFP cells can be captured by immobilized S protein of SARS-CoV-2. Different amounts of S protein ranging from 0 to 1 µg were immobilized on microplate surfaces, and HEK293T/ACE2-GFP cells captured by the S protein-coated surface were quantified using the CCK8 test. As shown in Fig. 2, the number of HEK293T/ACE2-GFP cells captured by S protein-coated surface increased accordingly with the increase of immobilized S protein, and it reached a plateau when the coated proteins were higher than 0.5 µg. By contrast, no HEK293T/ACE2-GFP cell was captured by uncoated surface. To better characterize the in vitro system, we determined the time kinetics of the cell capturing process. CapturedHEK293T/ACE2-GFP cells were detectable 5 min after the incubation, and the number of captured cells increased with the incubation period and reached a relatively stable level when the incubation period was more than 45 min (Supplementary Fig. 1).
To determine the specificity of the in vitro system, we utilized the receptor-binding domain (RBD) of S protein which is critical for the recognition of ACE2 to inhibit the cell capturing process. As shown in Fig. 3, RBD could efficiently inhibit the cell capturing process, with maximal inhibition of approximately90% and half-maximal inhibitory concentration (IC50) of 2.0 µg/ml.
Since HEK293T/ACE2-GFP cells overexpress ACE2, we next investigated cells without exogenous ACE2 expression can be captured by immobilized S protein of SARS-CoV-2. Seven cell lines were utilized in this experiment, including H1299, H460, HUH7, HepG, MRC5, U251 and HEK293T. We first determine the expression of ACE2 in those cells at both protein and mRNA levels. As shown in Fig. 4A, expression of ACE2 HEK293T/ACE2-GFP cells could be detected using western blotting, while levels of ACE2 in H1299, H460, HUH7, HepG, MRC5, U251 and HEK293T were under the detection limit. We then determined the expression of ACE2 at mRNA level using quantitative PCR. Among the seven cell lines, Huh7 showed the highest expression levels of ACE2, followed by H1299, HepG2 (Fig. 4B). Accordingly, the number of captured Huh7 cells also was higher than that of the other six cell lines (Fig. 4C). Also, the number of captured cells was highly correlated to the expression of ACE2 genes (R2 = 0.82, Fig. 4D)
Taken together, ACE2-expressing cells could be specifically captured by immobilized S protein, providing an in vitro system for investigating the interaction between S protein of SARS-CoV-2 and host cells.
Application of the in vitro cell capturing system
A potential application of this in vitro cell capturing system is evaluating the binding ability of mutant S protein variants to ACE2-expressing cells. To access this possibility, we compared the cell capturing ability of wild type S protein and S protein with the amino acid change from aspartate to a glycine residue at position 614(D614G) which is associated with enhanced infectivity and increased ACE2-binding affinity[11, 12]. As shown in Fig. 5, the S protein with D614G mutation showed a significantly higher cell capturing ability than wild type S protein.
Another potential application of the immobilized S protein-based cell capturing system is isolating and purifying host cells of SARS-CoV-2. We first tested whether this in vitro system can be used to purified ACE2-expressing cells. For this purpose, 293-T cells transiently transfected with ACE2-GFP were captured by immobilized S protein and then purified with a 2-step elution method (Fig. 6A). In this 2-step elution protocol, firstly captured cells were invertedly centrifuged to remove the weakly-bound cell and then eluted by treatment of trypsin. Compared with cells before the purification, cells eluted with the 2-step protocol contained a drastically higher percentage of cells that expressed higher levels of ACE2 (Fig. 6C). Next, we first determined whether the captured cells can be collected for further investigation. Captured cells were eluted after treatment with trypsin and accessed for cell viability. As shown in Fig. 6C, the majority of eluted cells were living cells, and cell viability was comparable between cells before and after the process of purification, In addition, the eluted cells were able to be further cultured and recaptured by S protein (data not shown). This finding suggests that cells with high binding affinity to S protein can be separated from those with low binding affinity.
The third potential application of this in vitro system is to identify neutralizing antibodies or screen inhibitors. To test this potential, we generated polyclonal antibodies against S1 protein in B6 mice and determined whether the polyclonal antibodies can block the cell capturing ability of immobilizing S1 protein. As shown in Fig. 7, 1:200 diluted serum from mice immunized with S1 protein completely inhibited the cell capturing process, while control murine serum did not.
Therefore, these results demonstrated that this in vitro cell capturing system is useful for determining neutralizing antibodies.