Blocking agents that bind to ACE2 receptor
The first strategy would consist of administering to patients an agent that would bind to ACE2. The key advantage here is that the host ACE2 protein will not change, so there is no concern about escape from binding the therapeutic agent. Moreover, the virus will not have the ability to mutate and bind an entirely new host receptor in the time frame of this outbreak; such functional relationships are established by evolution over long periods. By analogy, the influenza virus changes the mutations on its surface to escape antibody neutralization every year, but it always focuses on using sialic acid on the cell surface as an entry receptor 43.
There are two known options for agents to bind to ACE2. The first is using the small receptor-binding domain (RBD) from the SARS S protein that has been shown to be the key domain that binds the ACE2 receptor during entry 44. Administration of this domain, 193 amino acids in size, has been shown to effectively block the entry of SARS in cell culture 44. It is well within reason that SARS RBD could be given to patients, thereby binding their ACE2 proteins on target cells, preventing infection ( Figure 1). There is also the potential for the equivalent RBD of 2019-nCoV to be produced and used as a therapy as well. This strategy assumes SARS and 2019-nCoV share the same binding site on ACE2, which is highly likely given the similar ACE2 binding sites of SARS and NL63 coronavirus The small size of the therapy, similar in size in nanobody domains from camelid antibodies, would enhance the perfusion of the biologic into tissues to more effectively bind to viral entry receptors 45. In regards to the outbreak situation that is ongoing, the small protein facilitates the rapid production of the therapy in bacteria potentially, which would help production yields 19. Moreover, bacterial production would allow RBD proteins to be produced in a wide range of production facilities today in China, which already has numerous contract research organization operations 46. Alternatively, the RBD protein could be attached to an Fc fragment for extended circulation, which was done for an equivalent 212 amino acid domain from MERS. The MERS RBD-Fc fusion demonstrated the ability to block viral infection toward cell receptors, as well as to stimulate an immune response against that specific viral domain in mice 47. Of note, since the RBD-Fc fusion would bind to normal cells, one would want to eliminate cytotoxic Fc domain functions through mutations that eliminate Fc receptor binding 48.
Figure 1.  Therapeutic agents that could be used to block 2019-nCoV from infecting cells.
Target cells expressing ACE2 include lung and gastrointestinal tissues in the human body. The large spike protein on the surface of the coronavirus binds to ACE2 on infected cells, leading to cell entry. Three proposed strategies would block this interaction would abrogate infection. In the first, the receptor-binding domain (RBD) of the spike protein from SARS or 2019-nCoV would be administered, thereby binding ACE2 and saturating available sites. Alternatively, an antibody or single chain antibody fragment (scFv) could be administered against ACE2 to accomplish the same. A third strategy would target the coronavirus virions directly by using the ACE2 extracellular domain as bait to bind to spike protein. An Fc domain fused to ACE2 would facilitate prolonged circulation of the biologic (ACE2-Fc). A second, similar strategy would be to administer an antibody that would bind to ACE2 protein, thereby preventing 2019-nCoV infection ( Figure 1). This strategy was shown to effectively block SARS entry and replication in experiments 42. While no ACE2 antibody sequences are published in literature indexes, monoclonal antibodies do exist and the associated hybridoma sequences could be cloned in a matter of days. There would be no concern for any viral escape from an ACE2 binding antibody, which is an advantage over neutralizing approaches against the S protein. There are a couple of design considerations when thinking about how to employ the ACE2 antibody strategy. Any effector functions would need to be removed from the Fc domain 49, such that inflammation would not be caused in different tissues expressing ACE2. This would retain the long-half life endowed by the Fc domain without any of the side effects. The downside of including the Fc domain is the need to use a more expensive mammalian cell production system to preserve proper glycosylation, which would decrease the turnaround time for getting the drug to patients in the outbreak scenario. Alternatively, one could just administer a single chain variable fragment (scFv) that binds to ACE2. A nanobody or VHH domains from camelids are another option as well 50, 51. These could be produced in bacteria, and its small size would allow for rapid permeation into different tissues. The downside is the shorter half-life of these molecules without the Fc domain.
There are several limitations to these two options. Regarding the SARS RBD strategy, the body would likely develop an immune response to the SARS protein eventually, although the key intervention period of infection to combat 2019-nCoV would fall under this window of time, where after an immune response for both viruses would develop. Alternatively, if one were to use the homologous RBD from 2019-nCoV itself, this immune response would likely be very advantageous since it could yield both a blocking effect and a vaccination effect 52. For both strategies, the dose that would be needed to block ACE2 receptors in the body across different organs is unknown, and as is the percentage of ACE2 receptors that would need to be saturated in order to slow the infection. The number of ACE2 receptors in the body, which are found in lung and gastrointestinal organs along with vascular endothelial cells among other tissues 53, could ultimately prove prohibitive for this strategy. Moreover, the turnover of the ACE2 receptor on the cell surface would also influence how often the therapeutic protein would need to be administered. To solve this issue, one could increase the concentration of anti-ACE2 therapy at the crucial site of infection in the lungs, via local administration to lungs via nebulization. Lastly, there is the possibility that binding ACE2 directly could paradoxically worsen lung physiology and clinical symptoms. A study found that a fusion protein of SARS RBD to Fc domain bound ACE2 in murine lung tissue after administration, exacerbating alveolar edema via ACE2 interaction, which normally counteracts acute lung injury 54. This suggests that if one were use an ACE2 binding strategy, it would be best employed early during infection or as a prophylaxis to block the initial viral infection. Ultimately, clinical trials in patients would need to investigate these potential issues.