key: cord-0911294-yjpcyui0
authors: Lima, Marcelo A.; Skidmore, Mark; Khanim, Farhat; Richardson, Alan
title: Development of a nano-luciferase based assay to measure the binding of SARS-CoV-2 spike receptor binding domain to ACE-2
date: 2020-11-17
journal: Biochem Biophys Res Commun
DOI: 10.1016/j.bbrc.2020.11.055
sha: 1b1c2aeb659a2ac49f319f581b3831cf7fc905b7
doc_id: 911294
cord_uid: yjpcyui0

To identify drugs that could potentially be used to treat infection with SARS-CoV-2, a high throughput 384-well assay was developed to measure the binding of the receptor binding domain (RBD) of the viral S1 protein to its main receptor, angiotensin converting enzyme 2 (ACE2). The RBD was fused to both a HiBIT tag and an IL6 secretion signal to enable facile collection from the cell culture media. The addition of culture media containing this protein, termed HiBIT-RBD, to cells expressing ACE2 led to binding that was specific to ACE2 and both time and concentration dependant, Binding could be inhibited by both RBD expressed in E. coli and by a full length S1 - Fc fusion protein (Fc-fused S1) expressed in eukaryotic cells. The mutation of residues that are known to play a role in the interaction of RBD with ACE2 also reduced binding. This assay may be used to identify drugs which inhibit the viral uptake into cells mediated by binding to ACE2.

Zoonotic viruses make up the vast majority of emerging, human diseases owing to their ability to cross-spread between species from animals to humans [1] . SARS-CoV-2, the seventh coronavirus known to infect humans [2] , hit the world rapidly and uncontrollably, resulting in the death of over 1 million people as of October 2020, alongside devastating social-economical burdens. To-date, no widely effective treatment or vaccines are available and a significant drive towards SARS-CoV2 management relies on the effective track and trace of infected individuals in concert with restrictions on personal freedom. There is, therefore, an urgent need to identify effective therapies that ameliorate the significant health and socioeconomic burdens caused by SARS-CoV-2.

The viral tropism is determined by the interactions that occur between viral surface proteins and host cell receptors, and as such, these interactions are vital for effective cross-species and intra-species transmission [3] . Coronaviruses (CoVs) express a surface glycoprotein, termed spike, that binds to cell-surface receptors and mediate the entry of the virus into cells to initiate an infection [4] . Many host receptors for coronaviruses have been identified [4, 5] ;

in the case of SARS-COV2, the viral surface glycoprotein S1 contains a receptor binding domain (RBD) that interacts with angiotensin converting enzyme 2 (ACE2) and this is thought to mediate viral uptake into cells [6] . Thus, blocking this interaction could potentially limit viral infections. More recently, it has also been suggested that neuropilin can also mediate SARS-COV-2 uptake [7] .

The current state of the pandemic demands a rapid response. Drug repurposing screens are a convenient and attractive route to the identification of drugs that could address unmet medical needs. This is particularly attractive in the case of SARS-COV-2 because drugs that are already licensed for the treatment of other diseases could be rapidly redeployed to treat the current pandemic [8] . Many marketed drugs are readily available in large quantities and their safety profiles are well understood. When compared to traditional drug discovery paradigms, drug repurposing significantly reduces the time required for the identification of drug candidates that may treat emerging viral infections. Nevertheless, suitable assays are required for the rapid identification of candidate SARS-COV2 therapeutic drugs from the vast list of approved medicinal drugs. Computational screens provide an alternative strategy, although they lack a full consideration of the complex physiological environment [9] that dictates how molecules interact in vivo and, in any case, still require biochemical assays to validate the in silico results.

To develop an assay to measure the binding of SARS-COV-2 spike protein to ACE2, here the HiBIT system has been used. HiBIT is an 11 amino acid peptide that binds with high affinity to LgBIT, a fragment of an engineered form of luciferase termed nanoluc [10] . The binding of HiBIT to LgBIT renders the complex catalytically competent, creating a luminescence signal when an appropriate substrate is provided. The binding of HiBIT to LgBIT is of sufficiently high affinity that additional proteins are not required to colocalize the two proteins and create a bioactive complex. Thus, LgBIT can be utilised for the detection of a HiBIT-tagged protein. In this study, SARS-COV-2 RBD was fused inframe with a sequence encoding HiBIT to create a reagent that was anticipated to bind ACE2 and which could be detected by addition of LgBIT and the nanoluc substrate. An IL6 secretion signal was also included to allow HiBIT-tagged RBD secretion into cell culture medium of transfected cells.

The resultant culture medium, containing recombinantly expressed HiBIT-tagged RBD, was subsequently added to COS cells ectopically expressing ACE 2 to measure RBD-ACE2

binding. This was used in a proof-of-concept screen of a small library of marketed drugs.

Materials. The plasmid pCDNA3 encoding ACE2 was obtained from Genescript (OHu20260); the plasmid encoding prolactin was obtained from Sino biological (HG10275-CY). pRSETA encoding the SARS-CoV-2 S1 protein was obtained from GeneArt Gene Synthesis. Optimem and lipofectamine 2000 were obtained from Thermofisher. HiBIT J o u r n a l P r e -p r o o f detection reagent was obtained from Promega. The flexicloning transfer system (C8820), pFN39K and HiBIT detection reagent were obtained from Promega. The Q5 mutagenesis kit was obtained from New England Biolabs. SARS CoV-1 spike S1 protein was obtained from the Centre for AIDS Reagents, NIBSC, UK: Recombinant SARS CoV-1 spike S1 protein, consisting of a combination of the Fc fusion (25 kDa) and the SARS CoV-1 S1 domain (70 kDa) was obtained from Prof. Ian M Jones.

RBD constructs were prepared using the flexicloning transfer system. The region encoding amino acids 328-536 of the SARS-CoV2 spike S1 were amplified using Pfu polymerase and pRSETA SARS-CoV-2 S1 as the template. Forward The desired clones were identified by sequencing. The wild-type and mutated pF4ACMV-RBD clones and pFN39K, which includes a HiBIT tag and an IL6 secretion signal, were subsequently digested with flexiblend, heat inactivated and ligated. pFN39K clones containing the RBD were selected using kanamycin and identified by restriction digest.

Plasmids for transfection were prepared using a Qiagen plasmid purification kit.

Residues 319−597 of the SARS-CoV2 spike S1 RBD (GenBank: MN908947) were cloned upstream of an N-terminal 6XHisTag in the pRSET A expression vector and transformed into J o u r n a l P r e -p r o o f Western blotting.

Cells were transfected as described above with pFN39K HiBIT-RBD. Cells in a 12 well plate were lysed with a 250 µL modified RIPA as described [12] , separated on a 4-12% SDSpolyacrylamide gel, transferred to PVDF and analysed by immunoblotting with anti ACE2 (ab15348, 1 µg / mL).

The receptor binding domain of SARS-CoV-2 was cloned into the pFN39K expression vector in frame with an IL6 secretion signal and the HiBIT tag (Fig 1a) . This allowed the HiBITtagged protein (HiBIT-RBD) to be collected from the culture medium of cells transfected with the tag. The expression of the protein was verified by measuring luminescence after adding this culture medium to the HiBIT detection reagent (Fig 1b) . conducted at either at 37°C or at 0°C to minimize the potential uptake of the receptor-ligand complex by endocytosis. After incubation at 0°C, significantly more HiBIT-RBD was detected bound to cells expressing ACE2 than cells expressing PRL (Fig 1d) . In preliminary experiments, when the assay was conducted at 37°C, less specific binding was observed (data not shown) so all further experiments were conducted at 0°C.

Next, the kinetics of HiBIT-RBD binding to ACE2 were investigated. COS cells expressing ACE2 or PRL were incubated with HiBIT-RBD at 0°C for different periods and the specific binding, defined as binding to cells expressing ACE2 minus binding to cells expressing PRL, was determined (Fig 2a) . To provide evidence that the HiBIT-RBD provides a model of the viral spike protein binding to ACE-2, competition experiments were performed. In these experiments, SARS-CoV-2 S1

RBD and Fc-fused S1 (recombinantly produced in E. coli and eukaryotic cells, respectively) were used to give confidence that the binding mode was not unique to reagents prepared in COS cells. Cells transiently expressing ACE-2 were incubated with HiBIT-RBD in the presence of either RBD or with the full-length Fc-Fused S1. Both proteins were able to almost completely suppress the specific binding of HiBIT-RBD (Fig 3) .

To provide further evidence that HiBIT-RBD bound ACE2 in a manner that reflected the virus binding to ACE2, we made use of reports [13] [14] [15] that identified crucial amino acids in the RBD that are involved in the RBD-ACE2 interaction. It has also been reported that mutation of L455, F486 and Q493 in the SARS-COV-2 RBD to the corresponding residues in SARS-COV-1 substantially reduces the binding of the SARS-COV-2 RBD to ACE2 when measured in an ELISA assay [16] . Consequently, mutations encoding L455Y, F486L and Q493N were engineered into the HiBIT-RBD construct. When these were tested in the HiBIT binding assay, HiBIT-RBD L455Y and HiBIT-RBD F486L exhibited diminished binding to cells expressing ACE-2 and, in the case of HiBIT-RBD L455Y the specific binding was almost completely blocked. In the case of HiBIT-RBD Q493N , binding to ACE-2 expressing cells was reduced more modestly (although significantly). HiBIT-RBD F486L,Q493N , which contained both F486L

and Q493N mutations, also bound weakly to cells expressing ACE2, although not significantly more so than did HiBIT-RBD F486 . There was no measurable effect of any of the mutations on the non-specific binding of the HiBIT-RBDs to control cells expressing prolactin.

We have developed an assay that specifically measures the binding of SARS-COV2 to ACE2. The assay was implemented in a 384-well format allowing reasonable throughput.

The assay can be conducted using widely available scientific instrumentation (a plate reader capable of measuring luminescence) and reagents, and it is envisaged that this technology could be deployed in many laboratory settings.

For this assay ACE2, the main SARS-COV-2 receptor, was transiently expressed in COS cells because these cells can be transfected with high efficiency. HEK-293 cells were also investigated, but owing to their propensity to become easily dislodged during washing it was deemed unlikely that reproducible data could be obtained with them. COS cells were also preferred over Ovcar-8 cells, which can also be transfected with relatively high efficiency, because of the higher level of expression observed in the former cells. In the future, it may be useful to develop a cell line stably expressing the receptor. However, since we hoped to identify compounds which could inhibit RBD binding to ACE2 as quickly as possible to assist in the SARS-COV-2 pandemic, that option was not pursued here.

Several lines of evidence suggest that the described assay faithfully measures RBD binding to ACE2. As expected, binding of HiBIT-RBD was both time dependant and saturable, consistent with binding to ACE2. Importantly, significantly more binding of HiBIT-RBD was measured in cells expressing ACE2 than to cells transfected with a negative control (PRL).

This indicated that that the measured binding was specific to ACE2. Furthermore, both RBD and full length S1 proteins obtained from distinct biological sources could compete with

HiBIT-RBD for ACE2 binding thereby giving confidence that the binding of HiBIT-RBD reflected binding of the viral RBD to its receptor in its authentic binding mode. Finally, the mutation of several different residues that have previously been shown to inhibit RBD binding to ACE2 in an ELISA assay [16] also reduced it the binding measured in the HiBIT assay. Taken together, these data strongly suggest that an assay has been developed that J o u r n a l P r e -p r o o f replicates and measures the binding of the viral RBD to ACE2 through its authentic binding mode.

The three mutations introduced into the RBD mimicked the corresponding amino acids in SARS-COV-1 and have been previously shown to inhibit SARS-COV-2 binding to ACE2 in an ELISA assay [16] . These residues have also been shown from studies of the structure of ACE-2-RBD complex [13] [14] [15] to play a significant role at the interaction interface of the two proteins. In this study, two of the mutations (L455Y and F486L) substantially inhibited HiBIT-RBD binding to ACE2. However, a third mutation (Q493N) had a very modest effect, although it substantially inhibited binding in the ELISA assay [16] . To explore this further, an RBD was generated containing both Q493N and F486L mutations and this also bound more weakly than the wild-type protein and HiBIT-RBD Q493N but not significantly more than HIBIT-RBD F486L The relatively modest effect is perhaps not surprising when considering the conservative nature of the glutamine to asparagine mutation. Furthermore, the crystal structure of ACE2 bound to the RBD shows Q493 in two possible conformations [15] , suggesting there is some flexibility in the binding mode.

The assay was employed to screen a library of approximately 100 approved drugs [17] .

Unfortunately, no drugs were identified that reproducibly inhibited binding. Some hits appeared to inhibit the binding, but further exploration of these compounds revealed that they also inhibited the luminescence generated by adding HiBIT-RBD to the detection reagent directly, suggesting that the drugs inhibited nanoluc reporter reaction, rather than the HiBIT-RBD-ACE2 binding interaction. This emphasizes the importance of confirming hits identified in drug screens through validation in biochemical assays based on distinct technologies or by using other functional assays such as pseudotyped virus-like particle and/or live virus assays.

One issue with the assay is that the whole assay must be conducted on ice. In preliminary experiments (not shown) binding was not detected when the assay was performed at 37ºC.

Consequently, the assay was conducted on ice to prevent endocytosis of RBD bound to J o u r n a l P r e -p r o o f ACE2. To avoid the plate warming appreciably during reading luminescence, a very short integration time per well (0.1s) was used. Even so, some increase in the signal measured in control samples was observed as the plate was read and to control for this, multiple controls were included across the plate. It would be preferable to avoid this problem by using a plate reader that allowed the sample plate to be cooled. Alternatively, it might be possible to conduct the assay at a slightly higher temperature to minimize the effect of warming. It may also be possible to configure the assay to use cell membranes or fixed cells.

The assay developed here is suitable for the identification of drugs that inhibit the binding of SARS-COV-2 to ACE2. It may also be used to validate hits identified by other workers using binding assays based on other technologies. Although the assay has some limitations, these could be ameliorated with further work. 

Epidemiology, control, and prevention of emerging zoonotic viruses

The proximal origin of SARS-CoV-2

Virus-Receptor Interactions: The Key to Cellular Invasion

SARS-CoV-2 Infection Depends on Cellular Heparan Sulfate and ACE2

SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

Neuropilin-1 facilitates SARS-CoV-2 cell entry and infectivity

Artificial intelligence in COVID-19 drug repurposing

To cope with an increasing disease burden, drug discovery needs biologically relevant and predictive testing systems

Application of a novel HiBiT peptide tag for monitoring ATF4 protein expression in Neuro2a cells

Heparin inhibits cellular invasion by SARS-CoV-2: structural dependence of the interaction of the surface protein (spike) S1 receptor binding domain with heparin

Inhibition of cell spreading by expression of the C-terminal domain of focal adhesion kinase (FAK) is rescued by coexpression of Src or catalytically inactive FAK: a role for paxillin tyrosine phosphorylation

Structural basis for the recognition of SARS-CoV-2 by full-length human ACE2

Structural basis of receptor recognition by SARS-CoV-2

Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor

Key residues of the receptor binding motif in the spike protein of SARS-CoV-2 that interact with ACE2 and neutralizing antibodies

Highlights: 1. A high-throughput, 384-well plate assay was developed to measure the binding S1 RBD to ACE2

HiBIT-RBD binds to cells expressing ACE2 specifically and in a time dependant fashion

The binding of HiBIT-RBD to ACE2 can be inhibited using recombinantly expressed SARS-CoV-2 RBD and full-length

Site specific mutations within the RBD demonstrate the specificity of this assay

This research was funded by Keele University (AR) and the BBSRC (MAL & MAS; NIBB HVB Business Interaction Voucher and E3B Proof of Concept schemes).

The authors declare no conflict of interest.

AR conceived the study and performed the experiments with the assay. MAL and MS designed and prepared the RBD expressed in E. Coli and FK designed the drug library. All authors reviewed and contributed to the manuscript.

J o u r n a l P r e -p r o o f

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.☐ The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:Please note that all Biochemical and Biophysical Research Communications authors are required to report the following potential conflicts of interest with each submission. If applicable to your manuscript, please provide the necessary declaration in the box above.(1) All third-party financial support for the work in the submitted manuscript.(2) All financial relationships with any entities that could be viewed as relevant to the general area of the submitted manuscript.(3) All sources of revenue with relevance to the submitted work who made payments to you, or to your institution on your behalf, in the 36 months prior to submission. (4) Any other interactions with the sponsor of outside of the submitted work should also be reported. (5) Any relevant patents or copyrights (planned, pending, or issued). (6) Any other relationships or affiliations that may be perceived by readers to have influenced, or give the appearance of potentially influencing, what you wrote in the submitted work. As a general guideline, it is usually better to disclose a relationship than not.