Acriflavine is traditionally employed as a topical antiseptic, though has more recently been implicated as an antiviral agent at low doses, potentially by stimulating the production of interferon. In a new study recently uploaded to the bioRxiv* preprint server, the application of the drug to SARS-CoV-2 infection is investigated, demonstrating a broad activity towards a range of other betacoronaviruses.
Finding drug leads against SARS-CoV-2
Repurposing known drugs could be of vital importance in the current coronavirus disease 2019 (COVID-19) pandemic. Because they already bear thoroughly investigated toxicity profiles, they have the potential to be speedily deployed without having to adhere to the normally extremely lengthy clinical development timeline. In an effort to discover suitable drug leads, the group identified several essential and highly conserved proteins encoded by the SARS-CoV-2 genome. The cysteine papain‐like protease (PLPRO) belongs to non-structural protein 3, and is essential to virus protein maturation while being indicated to attenuate host interferon response, making it an appealing drug target.
A high-throughput screening approach towards discovering inhibitors of PLPRO was employed, exposing a fluorescently tagged protease assay to nearly 6,000 known and approved small molecule compounds. At this stage, acriflavine was already noted to generate the greatest inhibition of the compounds tested, and the effect of the drug on enzymatic activity was further explored utilizing fluorescently tagged acriflavine against natural PLPRO, giving an IC50 of 1.66 µM and 1.46 µM in each experiment, respectively. The drug’s specificity towards PLPRO was also demonstrated by comparison with a structurally and functionally related protease, which exhibited less than 50% inhibition at high concentrations of 100 µM.
Further characterization by NMR and X-ray crystallography illustrated the way in which acriflavine binds with PLPRO, not affecting the overall conformation of the protein and thus likely binding in a spatially localized site. Two molecules of dechlorinated acriflavine (proflavine) are able to stack within the binding pocket of PLPRO, inducing a large conformational change characterized by rotation and translocation of the nearby BL2 loop, a conserved region across coronaviruses that when distorted in this way prevents entrance into the enzyme active site. Commercial acriflavine contains a mixture of structurally similar molecules, some of which bear the aforementioned chloride ion, while others may bear methylation. A number of other binding sites were found to contribute towards enzyme inhibition, both in more hydrophilic and hydrophobic pockets that were thus occupied by the more or less polar forms of the drug. The compound mixture was noted by the group to be a significantly more potent inhibitor than pure proflavine, suggesting that it is the combined distortion from multiple binding sites on PLPRO that best disrupts protein function.
Testing antiviral activity
To further test the efficacy and toxicity of the drug against SARS-CoV-2 infection in vitro assays were performed in human cells. Cellular cytotoxicity 50% concentrations were found to be in the range of around 3-12 µM across the variety of human cells exposed, while IC50 values towards SARS-CoV-2 were around 50 nM. Applying the drug at later time points proved equally effective in virus inhibition, further supporting that inhibition takes place during the virus replication phase and does not affect cell entry.
A SARS-CoV-2 infected human airway ex vivo model then compared acriflavine to remdesivir, a broad-spectrum antiviral, showing significantly greater inhibition at concentrations as low as 500 nM compared with 10 µM of the latter. Two betacoronaviruses other than SARS-CoV-2 – MERS-CoV and HCoV-OC43 – in addition to two alphacoronaviruses, were then tested against acriflavine. Each of the betacoronaviruses demonstrated comparable IC50 values to SARS-CoV-2, while the drug was completely ineffective in slowing the replication of the alphacoronaviruses.
Finally, a murine model was utilized to compare acriflavine with remdesivir, administered in similar quantities over 6 days, with SARS-CoV-2 challenge introduced on the second. Subsequent RT-qPCR demonstrated that the virus was present in the lungs and brains of the untreated group, common due to the high incidence of the ACE2 receptor at these sites. Remdesivir lessened the viral load in the brain by around 2 orders of magnitude, while acriflavine did so around 5 fold. Viral load was less significantly reduced in the lungs by any method, though oral administration of acriflavine proved more effective than intramuscular injection.
Acriflavine could potentially be deployed as a useful broad-spectrum betacoronavirus drug despite a relatively short retention time in vivo. The group points out that the aromatic ring structure that allows the drug to stack within the PLPRO binding pocket and prevent protein function is generally associated with DNA intercalation and mutagenesis, so there could be some concerns regarding long-term use. This said, that history has yet to indicate concerns over limited use in humans.
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.