Amantadine has potential for the treatment of COVID-19 because it inhibits known and novel ion channels encoded by SARS-CoV-2

Abstract

The dire need for COVID-19 treatments has inspired strategies of repurposing approved drugs. Amantadine has been suggested as a candidate, and cellular as well as clinical studies have indicated beneficial effects of this drug. We demonstrate that amantadine and hexamethylene-amiloride (HMA), but not rimantadine, block the ion channel activity of Protein E from SARS-CoV-2, a conserved viroporin among coronaviruses. These findings agree with their binding to Protein E as evaluated by solution NMR and molecular dynamics simulations. Moreover, we identify two novel viroporins of SARS-CoV-2; ORF7b and ORF10, by showing ion channel activity in a X. laevis oocyte expression system. Notably, amantadine also blocks the ion channel activity of ORF10, thereby providing two ion channel targets in SARS-CoV-2 for amantadine treatment in COVID-19 patients. A screen of known viroporin inhibitors on Protein E, ORF7b, ORF10 and Protein 3a from SARS-CoV-2 revealed inhibition of Protein E and ORF7b by emodin and xanthene, the latter also blocking Protein 3a. This illustrates a general potential of well-known ion channel blockers against SARS-CoV-2 and specifically a dual molecular basis for the promising effects of amantadine in COVID-19 treatment. We therefore propose amantadine as a novel, cheap, readily available and effective way to treat COVID-19.

Fig. 1. Amantadine among other inhibitors blocks the ion channel function of SARS-CoV-2 Protein E.

a Summarized and averaged I/V relations in SARS-CoV-2 Protein E-expressing oocytes revealed significantly different current activity compared to control (uninjected) oocytes. The current activity at −85 mV in SARS-CoV-2 Protein E-expressing oocytes normalized to that of uninjected oocytes. b–i Eight compounds (10 μM) were tested on SARS-CoV-2 Protein E activity. Amantadine (b), HMA (e), emodin (f), and xanthene (g) blocked the activity, whereas no inhibition of SARS-CoV-2 Protein E by rimantadine (c), adamantane (d), pyronin B (h), or pyronin Y (i) was observed. b–i current activity at −85 mV with the treatment of the specific drugs normalized to the current activity obtained without treatment. Statistical significance was determined with unpaired Student t test, **P < 0.01; ***P < 0.001, of n = 4 biologically independent experiments.

Fig. 2. CSPs from ¹H-¹⁵N HSQC spectra of full-length Protein E from SARS-CoV-2 in the apo form and after adding amantadine (red) or rimantadine (black).

a The backbone amide region of the spectrum, with notable sequence-specific resonance assignments indicated. b Expansion of the spectrum showing shift perturbations for the residues G10, S16, V17, and L27. Measurements were recorded at 50 °C using a Bruker Prodigy probe on a 600 MHz instrument.

Fig. 3. MD simulations models of Protein E (aa 8–65) in complex with amantadine and HMA.

a Complex with amantadine, with inset in (b) zooming in at the binding site. c Complex with HMA, with inset in (d) zooming in at the binding site. The models depict an outward orientation of the drugs with Protein E in an open state (ligand carbons and Protein E in gray) from restrained 100 ns-MD simulations with OPLS2005 force field and a force constant of 2 kcal mol Å⁻² to the Ca atoms of Protein Ε. Ligand and amino–acid residues (shown in only one of the five α-helices for clarity) are shown as sticks. The PDB ID 5X29 was used as starting structure (aa 8–65) for the Protein E-amantadine complex.

Fig. 4. Overview of four viroporins encoded by SARS-CoV-2: Protein E, Protein 3a, ORF7b, and ORF10.

a Overall structure and ion channel function of known (Proteins E and 3a) and novel viroporins identified in the current study (ORF7b and ORF10). The membrane topology was predicted by TMHMM2 and displayed using Protter. b Ion channel activity mediated by Protein E, Protein 3a, ORF7b, and ORF10 electrophysiologically monitored in X. laevis oocytes, here shown as summarized and averaged I/V relations in oocytes expressing the viroporins compared to control (uninjected) oocytes of n = 3 biologically independent experiments.

Fig. 5. All four viroporins of SARS-CoV-2 can be blocked by drugs.

a Protein 3a-, b ORF7b-, and c ORF10-mediated current activity was monitored in the absence and presence of selected drugs (10 µM). Amantadine was ineffective on Protein 3a (a) and ORF7b (b), but blocked the activity of ORF10 (c). Xanthene blocked the activity of both Protein 3a (a) and ORF7b (b), while emodin blocked ORF7b (c). a–c Current activity at −85 mV with the treatment of the specific drugs normalized to the current activity obtained without treatment. Statistical significance was determined with unpaired Student t test, **P < 0.01; ***P < 0.001, of n = 3 biologically independent experiments.