Defending Antiviral Cationic Amphiphilic Drugs (CADs) That May Cause Drug-Induced Phospholipidosis

Abstract

A recent publication in Science has proposed that cationic amphiphilic drugs repurposed for COVID-19 typically use phosholipidosis as their antiviral mechanism of action in cells but will have no in vivo efficacy. On the contrary, our viewpoint, supported by additional experimental data for similar cationic amphiphilic drugs, indicates that many of these molecules have both in vitro and in vivo efficacy with no reported phospholipidosis and therefore this class of compounds should not be avoided but further explored as we continue the search for broad spectrum antivirals.

Graphical Abstract

Over approximately 100 years we have seen many respiratory viruses emerge, such as the recently identified SARS-CoV-2, that have led to millions of deaths as well as catastrophic financial consequences for the countries affected ¹. Our ability to resist these and future viruses of different classes rests with our ability to quickly identify and develop vaccines or drugs to help prevent infection as well as treat the infected patients. Antiviral drug discovery is not without its experimental challenges, but is often also stifled by the lack of financial return on investment ² caused by limited market size ³ for each virus in turn limiting the number of companies willing to take them on. COVID-19 reawakened the biotech and pharmaceutical industries alongside public-private efforts,⁴ which in record time have developed effective vaccines that have prevented suffering and death on a massive, unprecedented scale. In their shadow, drug discovery has been limited to repurposing of very few drugs like remdesivir ^5-7, dexamethasone ⁸ or combinations of these two drugs ⁹, which have shown some utility (in hospital settings and for those that are severely infected patients). In a short period of time there have been noteworthy efforts to screen large libraries of drugs (either computationally or experimentally) versus individual SARS-CoV-2 targets as well as phenotypic screens in cells infected with this virus, in order to repurpose FDA-approved as well as other clinical candidates ^{10, 11}. Only limited numbers of molecules have emerged from these screens that have also reached clinical trials. This highlights how drug discovery for this virus is particularly challenging and efforts to improve our odds of finding molecules that can progress to the clinic and succeed in the future will require an unbiased approach. Recently, several molecules have been described that showed promising in vitro and in vivo efficacy in mouse or hamster models including plitidepsin ¹², clofazimine ¹³, diABZI-4 ¹⁴ and masitinib ¹⁵ (Table 1), however, some of these may already be attracting undue attention due to their physicochemical properties.

Table 1.

Inhibitors of SARS-CoV-2 in vitro and in vivo. pKa and LogP calculated with ChemAxon software (Budapest, Hungary).

Structure	Name	pKa	LogP	In vitro activity	In vivo activity	Reference
	diABZI-4	6.96	0.41	0.1 μM leads to > 2 log PFU decrease in A549-ACE2 cells.	0.5mg/kg increased survival of mice dosed 3h pre-treatment or 12 post treatment intranasally.	14
	Clofazimine	6.63	7.3	EC50 0.31μM in VeroE6 cells.	25mg/kg oral dose lowered viral load > 1 log PFU in hamster lung when dosed prophylactically or therapeutically.	13
	Plitidepsin	11.01	3.98	IC90 1.76nM in VeroE6 IC90 0.88nM in hACE2-HEK293T cells.	Lowered mouse lung viral load 2 log PFU after 0.3mg/kg daily dosing for 3 days.	12
	Masitinib	7.84	4.97	EC50 3.2 μM in A549-ACE2 cells.	25mg/kg twice a day Lowered mouse lung viral load >2 log PFU and improved survival.	15

In their recent article in Science, Tummino et al., ¹⁶ used compounds they originally identified from a repurposing screen as sigma receptor ligands, but had showed no relationship between receptor potency and SARS-CoV-2 antiviral activity; instead they found that cationic amphiphilic drugs (CADs) had antiviral activity. They went on to show that phospholipidosis in cells for 16 drugs correlated strongly with their in vitro antiviral activity. They also identified approximately 60% of the hits (310 drugs) coming out of SARS-CoV-2 repurposing screens were likely CADs based purely on their cLogP (≥ 3) and pKa (≥ 7.4) values. Interestingly, it is widely known that compounds with a basic pKa (> 6.5) and cLogP of > 2 tend to be lysosomotropic and accumulate in the lysosomes ¹⁷, a hallmark of many phospholipidosis-inducing compounds. Some of these molecules active against SARS-CoV-2 had also previously shown in vitro antiviral activity against a range of other viruses (e.g., Ebola (EBOV), Marburg (MARV), Hepatitis C and Dengue). Yet, 4 phospholipidosis-inducing drugs (amiodarone, sertraline, PB28 and tamoxifen) that showed in vitro activity were tested in a 3-day mouse efficacy model for SARS-CoV-2 infection and these did not show efficacy ¹⁶. Tummino et al., concluded that phospholipidosis-causing CADs were therefore wasting resources and should be avoided by counter-screening ¹⁶.

In our opinion, the study by Tummino et al universally disparage all such CADs by using just four examples of repurposed, phospholipidosis-inducing molecules that are active against SARS-CoV-2 in vitro. Phospholipidosis may not even be relevant in the mouse model for SARS-CoV-2 as the authors showed amiodarone offers neither antiviral protection nor hallmarks of phospholipidosis.

Many of the observations from Tummino et al., ¹⁶ were perhaps not surprising as much has already been written about CADs, such as their intracellular accumulation in different subcellular compartments (e.g. mitochondria and lysosomes) ^{18, 19}. Basic amines like chloroquine may also lead to accumulation in these compartments and change the pH, which ultimately inhibits entry for some viruses. The alkalinization of acidic components, like the lysosome, in such host cells has previously been proposed as a strategy to decrease SARS-CoV-2 viral infection ²⁰. Kitagawa et al., also discussed the importance of pH on the antimalarial CAD structure which may also impact their antiviral activity ²¹. These observations clearly suggest that not all CADs are created equal, and their biological activity may be influenced by other factors. It is well established that for numerous areas of drug discovery the in vitro activity rarely translates to in vivo activity. This could be simply because the plasma levels are inadequate ²², which could be caused by numerous factors including high protein binding, increased drug metabolism, species differences (from human cells to animal), efflux out of cells, etc. In our own antiviral drug discovery efforts, we have previously reviewed 7 molecules (chloroquine ^{23, 24}, azithromycin ²³, amiodarone ²⁵, iminosugars ²⁶, BGB324 ²⁷, NCK8 ²⁷ and 17-DMAG ²⁷) with promising activity against Ebola in vitro which did not demonstrate efficacy when tested in guinea pig ²⁸. Similar to the finding of Tummino et al. regarding inhibition of SARS-CoV-2 ¹⁶, we found a strong correlation between anti-EBOV and MARV activity and lysosomotropic activity ²⁹. 21 of 23 lysosomotropic compounds inhibited these viruses, with IC_50’s in the nM to low μM range, strongly supporting the role of lysosomotropic characteristics in the in vitro antiviral activity ²⁹. The lysosomotropic characteristic had previously been described for EBOV ^{23, 30}, though CADs concentrated in lysosomes being the mechanism of EBOV inhibition had not been completely resolved (e.g., cholesterol accumulation in endosomes and lysosomes, lysosomal membrane stability, or inhibition of acid sphingomyelinase ^{31, 32}). CADs could also bind to hydrophobic pockets of viral proteins, such as the EBOV-Glycoprotein, with example molecules of 1-Benzyl-3-cetyl-2-methylimidazolium Iodide (NH125) ³³ and diazachrysene analogs, which showed both potent in vitro and in vivo activity against EBOV ³⁴. Others have commented on the large number of CADs that have antiviral activity and also induce phospholipidosis during chronic treatment, however their use is well tolerated, and any toxicity is reversible ³¹.

We ²⁹ and others ¹⁷ have also described how lysosomotropic compounds can be identified through the inhibition of lysotracker accumulation in vitro and had shown that the CADs like pyronaridine, quinacrine and tilorone all inhibit lysotracker ²⁹ while also possessing activity against EBOV in vitro ³⁵ and in vivo in a mouse-adapted EBOV infection model ^36-38 (and in the case of pyronaridine in guinea pig as well ²⁸). A machine learning approach was also described to learn from the literature in vitro data; ^{17, 39} it was used to identify such lysosomotropic compounds and find other molecules with this mechanism, with several already showing utility as broad-spectrum antivirals ²⁹. Pyronaridine, quinacrine and tilorone have also been identified to possess other in vitro activities against additional viruses ^{36, 38, 40-42} and were demonstrated to have sub-micromolar IC₅₀ values in A549-ACE2 cells infected with SARS-CoV-2 ⁴³. It has been recently reported that the combination of pyronaridine and artesunate used in a Phase II clinical trial had some promising effects in those with severe illness ⁴⁴ although we await the more definitive publication of these results in a peer reviewed journal. One caveat of this clinical study is the use of a combination of drugs as opposed to using pyronaridine alone.

CADs are potentially of interest for other reasons, such as their potential for interfering with human transporters which could have potential undesirable effects. Several cationic compounds with SARS-CoV-2 antiviral activity were found to interact with OCT and MATE transporters in vitro ⁴⁵ and in some cases this was influenced by the substrate used (e.g. for OCT1 and 2 but not for MATE1). Chloroquine, hydroxychloroquine and quinacrine possessed IC₅₀ values between 1-10 μM for all these transporters. Several of the drugs tested had C_{unbound max}/IC₅₀ values that would predict potential drug-transporter interactions in vivo ⁴⁵. It is unlikely that these would impact SARS-CoV-2 viral entry, however it should also be pointed out that some transporters can be leveraged by viruses for entry, such as the bile acid transporter sodium taurocholate co-transporting polypeptide (NTCP) ⁴⁶. We can also target transporters in order to reach viral sanctuary sites such as the testes, for example, using human equilibrative nucleoside transporters (ENT) whose structure-inhibitor and structure-substrate relationships we are just beginning to elucidate ⁴⁷.

Therefore, perhaps before advocating for the abrogation of FDA-approved drugs or novel compounds that are CADs demonstrating phospholipidosis from SARS-CoV-2 in vitro screens (and ultimately from in vivo testing), we should consider the potential for missing promising clinical candidates which represent the needle in the proverbial haystack (Figure 1). As an example, out of the compounds in Table 1 that demonstrated in vitro and in vivo activity against SARS-CoV-2 we would suggest several may meet the logP and pKa criteria as likely CADs, however we are not aware of any published reports of phospholipidosis for these compounds. Therefore, we would suggest the need for a more careful in-depth analysis of the likely hundreds of CADs that are FDA-approved drugs or drug candidates in order to understand their likely potential to cause phospholipidosis in vitro and or in vivo, as well as identify any correlations between physicochemical properties such as logP, pKa and antiviral mechanisms. CADs may actually provide useful structure activity relationship insights that can help us find and design broader spectrum antivirals to fight the next pandemic.

Figure 1.

Finding the needle in the haystack. Recently described molecules with in vitro and in vivo activity against SARS-CoV-2 which are likely CADs.