A compelling report, recently released on the bioRxiv* preprint server, describes unexpectedly powerful inhibition of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) by thiol-based mucolytic compounds P2119 and P2165, by preventing the virus spike from binding to its host cell receptor, the angiotensin-converting enzyme 2 (ACE2).

The mechanism of action of these compounds appears to be via disulfide reduction at the receptor binding domain (RBD) of the spike protein that mediates virus-receptor binding and viral entry. The disulfides affected here appear to play an allosteric role, acting via a hydrophobic binding pocket in the RBD that is conserved between multiple coronaviruses that infect humans.

Background

While an array of strategies have been explored to identify and develop effective and safe antivirals to prevent and treat SARS-CoV-2 infection in the ongoing coronavirus disease 2019 (COVID-19) pandemic, many have centered around blocking protease and RNA polymerase, viral enzymes critical to viral replication. Despite promising results, few have been clinically effective.  

Similarly, many therapies that aim to block virus RBD-ACE2 binding depend on developing molecules that show a specific 3-D fit to the viral epitopes involved in this process.

Disulfide reducing activity may have an antiviral effect

In the current study, the researchers used a simpler approach, exploiting the central role of disulfide bonds in multiple viral proteins required for viral binding and entry, especially in coronaviruses. These bonds are formed first in the endoplasmic reticulum (ER) and are stabilized in the presence of oxidant molecules outside the cell membrane.

In the case of the SARS-CoV-2 RBD, four pairs of disulfides were found: Cys480–Cys488, at the ACE2 binding surface, Cys336–Cys361, Cys379–Cys432, and Cys391–Cys525. The latter three stabilize the β sheet structure of the RBD.

The mucolytics P2119 and P2165 are in preclinical development and have the unique feature that they are found only extracellularly. Both are composed of thiol units bound to glucose or mannose sugars, respectively. While the first is a monothiol, the second contains two thiol groups.

The hydrophilic nature of the sugars accounts for their remaining outside the cell. Despite their inability to enter the cell, they reduce viral titer, hinting that they block viral entry rather than post-entry processes like replication.

What were the findings?

P2119 and P2165 reduced SARS-CoV-2 titers by >3 log and >6 log, respectively, at a concentration of 30 mM. On direct comparison, NAC, which is the gold standard for clinical thiol-based mucolytics, failed to show virucidal activity against this virus, at up to 100 mM, while P2165 inhibited the virus at over 3 mM, reducing it by 1-3 logs.

Against other coronaviruses, including SARS-CoV and NL63-CoV, a similar 3-4 log reduction of viral titers was seen at P2119/P2165 concentrations between 10-30 mM. The former has a spike protein that is 75% identical to the SARS-CoV-2 spike, with three of the above disulfides. The latter also uses ACE2 as its host receptor, with disulfides on the RBD that stabilize it and promote virus-receptor binding.

These compounds were tested in other types of cells, showing comparable (1-4 log) reductions in viral titers at the same range of concentrations. However, P2165 has more potent inhibitory activity than the other, presumably because it is a dithiol and the other a monothiol. Both seem to share the same mode of action.

In vitro experiments showed that the recombinant SARS-CoV-2 spike RBD binding to ACE2 was effectively blocked by both agents via their reduction of the disulfides at this region. The half-maximal effective concentration (EC50) with P2165 and P2119 was reduced by approximately 6-fold and 4-fold, respectively. Tris(2-carboxyethyl) phosphine (TCEP), which is a powerful reducing agent, completely inhibits RBD-ACE2 binding, further supporting this mechanism of action.

Spike targeting by disulfide reduction

Moreover, approximately 850 extracellular human peptides containing cysteine were found to be reactive to these agents, P2119 and P2165, respectively, while there were 49 corresponding viral peptides. With respect to mixed disulfide formation, seen only with monothiols, there were 12 and over 280 viral and human peptides, respectively. Among the mucolytic-affected viral peptides, 67% to 75% were in the spike protein.

The Cys480–Cys488 pair at the loop region of the RBD forms a stable disulfide bond, which is not alkylated during infection. In contrast, Cys432 and Cys525 are hyper-reactive cysteines and therefore form semi-stable disulfide bonds with their cognate cysteines.

The Cys480–Cys488 pair was the most accessible to solvent, along with Cys391–Cys525, in the isolated or open-RBD simulations. Cys336–Cys361 and Cys379–Cys432 were buried disulfides.

Rare allosteric disulfides

Molecular docking analysis studies indicated that these three disulfides (other than Cys480–Cys488) regulate virus-ACE2 binding since their reduction causes the RBD to undergo conformational changes. Since the Cys379–Cys432 and Cys391–Cys525 pairs are located away from the ACE2 binding site, this suggests that they are allosteric modulators of the RBD conformation.

This makes them part of a very rare category of disulfide bonds, unlike the more usual structural and catalytic disulfides. The Cys379–Cys432 disulfide bond, when reduced, causes a series of changes in the region of interaction of the CR1 region of the receptor binding motif (RBM).

While the changes were fewer with reduction of the Cys391–Cys525 disulfide, they were more widespread, also extending to CR2 and CR3. The resulting RBD conformation was quite different from the native RBD, such as shrinkage of the RBM in the vicinity of CR1 and CR3, possibly affecting virus-ACE2 recognition. Similar findings were found with Cys480–Cys488 as well, though to a lesser extent.

In fact, the latter pair, when reduced, caused structural changes in the RBD that resulted in very different conformations for the RBM. These pairs, therefore, are key in the dynamic conformation of the RBD, and are targets of thiol-based mucolytics.

The Cys379–Cys432 disulfide was found to enter a hydrophobic binding pocket at the interface of neighboring RBD subunits.

When this pocket is bound by hydrophobic molecules like linoleic acid, it stabilizes the closed spike conformation, as shown by the in vitro inhibition of RBD-ACE2 binding by polyunsaturated fatty acids or fat-soluble vitamins, e.g., vitamin D.

However, this pocket is a hidden one, with breathing-like alternations between the accessible and inaccessible state for this Cys379–Cys432 disulfide. This binding pocket may be accessible only when the spike is in the open conformation.

Further experiments with molecular docking showed that both P2119 and P2165 bound to this pocket, with the benzene group buried and the reducing thiol group pointing straight towards the Cys432. This also allows it to interact with neighboring residues, both aromatic and hydrophobic, while stabilizing the polyol bond through polar bonds and hydrogen bonds.

What are the implications?

The study demonstrates that human coronaviruses were potently inhibited by both agents because they reduce key disulfides in the SARS-CoV-2 spike RBD. Triggering conformation changes, this inhibits spike-receptor binding and hence infection.

The inhibition is independent of covalent bond formation, but rather depends on disulfide reduction, particularly of mucolytic sensitive pairs, namely, Cys379–Cys432 and Cys391–Cys525, sparing the Cys480–Cys488 pair, during infection. The latter is both conserved among coronaviruses and stable, thus slow to undergo reduction.

Moreover, the findings suggest these are allosteric in their function, causing changes in protein conformation when they form or break. The thiol agents bind to similar hydrophobic pockets as the Cys379–Cys432 disulfide, and in fact, Cys432 forms a covalent bond with P2165. This could help hone the agent still further to enhance RBD interactions.

These results may indicate an alternative mechanism of inhibition of virus-ACE2 binding by neutralizing antibodies, other than physically blocking the RBM – instead, by disrupting disulfides near the binding epitopes of these antibodies. In fact, the thiol-based agents were effective against the D614G variant of SARS-CoV-2. This will be an important advantage in view of the rapid emergence of new variants of concern.

Other potential benefits of this approach include the ability to restore intracellular antioxidant activity by boosting glutathione levels to normal. Thiol-based mucolytics also reduce the secretion of mucus by breaking down disulfides formed between molecules within mucus, thus helping clear airway obstruction and reduce airway inflammation and infection.  

They are also safer, with a higher therapeutic index than the approved drug N-acetyl cysteine (NAC).

These collective findings establish the vulnerability of human coronaviruses to repurposed thiol-based mucolytics and lay the groundwork for developing these compounds as a potential treatment, preventative and/or adjuvant against infection,” conclude the researchers.

*Important notice

bioRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.

Written by

Dr. Liji Thomas

Dr. Liji Thomas is an OB-GYN, who graduated from the Government Medical College, University of Calicut, Kerala, in 2001. Liji practiced as a full-time consultant in obstetrics/gynecology in a private hospital for a few years following her graduation. She has counseled hundreds of patients facing issues from pregnancy-related problems and infertility, and has been in charge of over 2,000 deliveries, striving always to achieve a normal delivery rather than operative.