Biomedical research faces a ubiquitous and fundamental challenge: Because the goal of the research is to improve human health, the human body is logically the best place to perform experiments. However, except in rare, controlled cases, experimenting on humans is both unethical and impractical. Thus, the biomedical scientist must turn to more practicable models, usually animals or cultured cells, to carry out their work. Doing so is a compromise: however hard we try to pick the best experimental model, its fundamental un-humanness means that we can never be sure how well what we observe in the model will translate to humans, nor what aspects of the model may affect this translatability. Consider, for example, the work of Dr. Jesse Bloom, Professor in Fred Hutch’s Basic Sciences and Public Health Sciences Divisions. Since the start of the Covid-19 pandemic, Dr. Bloom’s laboratory has focused on understanding how mutations of SARS-CoV-2 affect our immune system’s ability to recognize and neutralize the virus. Nothing would be more informative in this endeavor than to experimentally mutate the virus, infect a cohort of people, and examine how their bodies respond. But even setting aside the most significant impediment to such a plan – its moral and ethical failings – this experiment is impractical, because comprehensively addressing this question requires the testing of thousands of different mutations. To overcome both of these challenges, Dr. Bloom’s lab has established cell culture-based neutralization assays to study how effectively antibodies can impair the infectiousness of SARS-CoV-2 variants. But how well does such an assay represent the human condition? In a new research article published in Viruses, Dr. Bloom and colleagues, led by research technician Ariana Farrell and staff scientist Bernadeta Dadonaite, identify a new factor with important impacts on the results of their assay.
“Neutralization assays are the most widely used experimental method to assess immunity elicited by SARS-CoV-2 vaccination and infection,” the authors explain. Importantly, however, “the neutralization measured in the lab depends on the details of the assay…previous studies have shown that some antibodies can have markedly different neutralizing activities depending on which viral systems or target cells are used.” It is unsettling to any scientist to know that decisions in experimental design could impact the translational impact of the results. Thus, Dr. Bloom’s group set out to better understand how differences in assay design affect study outcomes. They chose to focus on one variable: the level ACE2 receptor expression in target cells. Binding of SARS-CoV-2 to the ACE2 receptor is crucial to its ability to infect our cells, and antibody design has been heavily targeted to the ACE2-binding region (the so-called receptor-binding domain or RBD) of the virus’s spike protein. Thus, all target cells used in neutralization assays must express ACE2. But the level of that expression varies. As the authors explain, some cells natively express ACE2 (like Vero cells), while some can be engineered to overexpress it (like 293T cells) and both cell types might be used in a neutralization assay. Could such differences in receptor expression, the group asked, affect the ability of some antibodies to neutralize the virus?
To address this question, the group exploited a cell line created by Dr. Kenny Matreyek at Case Western University, engineered to express ACE2 at four different, carefully controlled levels ranging from very low to high. They used the cells to examine how well polyclonal sera – blood fluid containing antibodies against SARS-CoV-2 – collected from immunized human patients could neutralize the infection. The goal of this experiment was not to understand how well the serum worked, but whether different antibodies in the serum had different neutralization abilities based on ACE2 expression. Thus, they focused on two types of antibodies – those targeting the RBD, and those targeting other regions of the spike protein. To distinguish between these, they tested either unaltered serum, or serum in which RBD-targeting antibodies had been depleted. Importantly, they observed that on high ACE2 expressing cells, RBD-targeting antibodies were responsible for nearly all of the neutralization. However, their relative importance decreased with ACE2 expression, so that on lower ACE2 expressing cells, non-RBD antibodies made a greater contribution to neutralization. Finally, the group performed the neutralization assay using monoclonal antibodies with defined targets. Again, they found that antibodies targeting outside the RBD more effectively protected cells with lower ACE2 expression. “Our study shows that target cell ACE2 expression is an important experimental variable in neutralization assays, and ACE2 expression will influence the relative potency of different antibodies targeting SARS-CoV-2 spike,” says Farrell. Perhaps most importantly, the authors indicate, is that studies exclusively using high-expressing cells may underestimate the value of antibodies targeting outside the RBD to immunity.
An idealized outcome of a study such as this is to create a “better” experimental model – one whose results most accurately inform us of what will happen in humans. But the reality is often not so simple: creating an assay which is both feasible in the laboratory and representative of true infection involves balancing variables. Human cells, for instance, express ACE2 at a variety of levels (although the airway cells infected by SARS-CoV-2 tend to express it at low levels). Ultimately, the authors conclude, “our study is unable to definitively answer the most important question it raises: What target cell ACE2 expression provides the most biologically relevant measure of SARS-CoV-2 neutralization?” Nevertheless, muses Farrell, “while it is hard to say which ACE2 cells are “best” to use for neutralization assays, it is clear this is an important variable to consider and report in experiments.”
Finally, Farrell acknowledged the value of cross-institutional collaboration to this work. “This project hinged on using serum which we acquired through our fantastic collaboration with Helen Chu and the HAARVI [Hospitalized or Ambulatory Adults with Respiratory Viral Infections] study. The close alliance of Fred Hutch with UW enabled our access to this central resource for this study.”
This work was supported by the National Institutes of Health, the European Molecular Biology Organization, and the Howard Hughes Medical Institute.
Farrell AG, Dadonaite B, Greaney AJ, Eguia R, Loes AN, Franko NM, Logue J, Carreño JM, Abbad A, Chu HY, Matreyek KA, Bloom JD. Receptor-Binding Domain (RBD) Antibodies Contribute More to SARS-CoV-2 Neutralization When Target Cells Express High Levels of ACE2. Viruses. 2022 Sep 16;14(9):2061. doi: 10.3390/v14092061. PMID: 36146867; PMCID: PMC9504593.