Screening of FDA-approved Drugs and Identification of Novel Lassa Virus Entry Inhibitors

Lassa virus (LASV) belongs to the Mammarenavirus genus (family Arenaviridae) and causes severe hemorrhagic fever in humans. At present, there are no Food and Drug Administration (FDA)-approved drugs or vaccines specific for LASV. Herein, high-throughput screening of an FDA-approved drug library was performed against LASV entry using a pseudo-type virus enveloping LASV glycoproteins. Two hit drugs, lacidipine and phenothrin, were identified as LASV entry inhibitors in the micromolar range. A mechanistic study revealed that both drugs inhibited LASV entry by blocking low-pH-induced membrane fusion. Moreover, lacidipine irreversibly bound to the LASV glycoprotein complex (GPC), resulting in virucidal activity. Adaptive mutant analyses demonstrated that replacement of T40, located in the ectodomain of the stable-signal peptide (SSP), with lysine (K) conferred LASV resistance to lacidipine without apparent loss of the viral growth profile. Furthermore, lacidipine showed antiviral activity and specificity against both LASV and the Guanarito virus (GTOV), which is also a category A new world arenavirus. Drug-resistant variants indicate that the V36M in ectodomain of SSP mutant and V436A in the transmembrane domain of GP2 mutant conferred GTOV resistance to lacidipine, suggesting that lacidipine might act via a novel mechanism other than calcium inhibition. This study shows that both lacidipine and phenothrin are candidates for LASV therapy, and the membrane-proximal external region of the GPC might provide an entry-targeted platform for inhibitors.


4
(Z), and glycoprotein complex (GPC). The GPC is synthesized as an inactive 66 polypeptide and cleaved twice by the signal peptidase and cellular protease subtilisin 67 kexin isozyme-1/site-1 protease, yielding the retained stable-signal peptide (SSP), the 68 receptor-binding subunit GP1, and the membrane fusion subunit GP2 (7-10). The 69 highly conserved arenavirus SSPs contain 58 amino acids that span the membrane 70 twice, with 8 amino acids in the ectodomain, playing essential roles in GPC 71 maturation and downstream functions (11)(12)(13)(14)(15)(16)(17). LASV utilizes α-dystroglycan (α-DG) 72 as a primary receptor, and successful infections require the receptor switch to 73 lysosome-associated membrane protein 1 (18)(19)(20). 74 To date, no vaccines or specific antiviral agents against LASV are available. Therapy 75 strategies are limited to the administration of ribavirin in the early course of the illness 76 (21). To address this issue, we screened an FDA-approved drug library of 1018 77 compounds. The approved drugs have been intensively investigated for safety, 78 pharmacokinetics, and targets; therefore, screening approved drugs for repurposing 79 will increase the speed of discovery and development for treatment (22,23). Drugs  inhibitors (24). The number of genomic RNA copies of LASVpv was determined to 93 be 1 × 10 7 copies/ml by using a standard curve generated with plasmids carrying the 94 VSVΔG-Rluc. The HTS assay conditions, including the seeding cell density and 95 LASVpv infective dose, were optimized at 1 × 10 4 cells and 1 × 10 2 copies per 96 96-well plate, respectively. Under the optimized conditions, signal-to-basal (S/B) ratio, 97 coefficient of variation (CV), and Z' factor were 41770, 11.9%, and 0.62, respectively, 98 demonstrating that the assay was promising for large-scale screening of inhibitors.

99
The HTS schematic is depicted in Fig. 1A. Inhibitors were defined as primer 100 candidates with inhibition > 50% and no apparent cytotoxicity in duplicate wells at a 101 concentration of 10 μM. Of the 1018 tested compounds, 52 (5.11%) were considered 102 primer candidates. A screening to reconfirm the results was then carried out using 103 these primer candidates over a broader concentration range (3.125 to 50.0 μM). Seven 104 compounds (0.69%) were selected based on their concentration-dependent inhibitory 105 effects and a cell viability > 80%. Subsequently, these 7 compounds were subjected to 106 counter-screening to rule out inhibitors of VSV genome replication and Rluc. Using 107 these criteria, 2 hits, lacidipine and phenothrin, were selected with specific inhibition 108 against LASV GPC activity, while the other 5 compounds were eliminated (Fig. 1B).
Lacidipine is a dihydropyridine voltage-gated Ca 2+ channel antagonist, while 110 phenothrin is a synthetic pyrethroid used for aerosol insecticides. We evaluated the 50% 111 inhibitory concentration (IC50) and 50% cytotoxic concentration (CC50) of both hit 1C and D). The selective index (SI, the ratio of the CC50 to the IC50) for lacidipine 117 was 55.4, while that for phenothrin was > 75.5 (Fig. 1E). The CC50 values for the 2 hit 118 drugs were similar to those previously published for diverse cell systems; however, 119 they were determined using different toxicity assays (25). To validate the antiviral 120 effects, lacidipine and phenothrin were purchased from other commercial sources and 121 tested; the cytotoxic and antiviral effects were similar to the results of our primary 122 screening.

123
Lacidipine and Phenothrin Inhibited GPC-mediated Membrane Fusion. 124 Arenavirus GPCs have a unique structure in which the cleaved SSP is retained and 125 non-covalently associates with GP2; many arenavirus entry inhibitors have been 126 shown to bind with and stabilize the prefusion forms of GPC to prevent membrane 127 fusion (26-28); therefore, we asked whether the 2 hit drugs act via a similar 128 mechanism. To address this, 293T cells transfected with GPC were incubated with 129 either drug and subjected to a low-pH pulse to promote fusion. As shown in Fig. 2A, a 130 low pH in the GPC-transfected cells induced obvious membrane fusion, whereas a 131 7 neutral pH had no effect. Both drugs inhibited syncytium formation at all tested 132 concentrations, suggesting that both drugs inhibit GPC conformational changes 133 induced by an acidic environment.

134
To further quantitatively evaluate the inhibitory activities, fusion efficacy was 135 determined using a dual-luciferase assay. As shown in Fig. 2B, the maximum 136 inhibitory rates for lacidipine and phenothrin were 42.4% and 81.2%, respectively, at 137 the range of concentrations tested. Notably, phenothrin exhibited great activity (~80%) 138 against GPC-mediated membrane fusion even at the lowest concentration tested (12.5 139 μM). Together, these results show that both drugs, especially phenothrin, prominently 140 inhibit GPC-mediated membrane fusion.

141
Lacidipine Irreversibly Binds to GPC. Lacidipine and phenothrin inhibited 142 GPC-mediated membrane fusion; therefore, we asked whether the drugs irreversibly 143 bind to GPC and prevent conformational changes. To test this, we conducted a 144 virucidal assay to investigate the binding ability of the hit drugs to native GPC.

145
LASVpv was mixed with drugs for 1 h; the mixture was then diluted 200-fold and 146 added to the cells for 1 h. As shown in Fig. 3A, luciferase activity was not suppressed 147 in the phenothrin group, indicating that phenothrin did not irreversibly bind to the 148 neutral-pH forms of GPC (29). However, it is possible that phenothrin binds to a 149 structural GPC intermediate resulting from the low pH. A reduction > 70% was 150 observed in the lacidipine group, suggesting that lacidipine irreversibly interacts with 151 the pre-fusion conformation of GPC and inhibits LASV entry. 152 We next investigated the inhibitory effects of the drugs on binding, which is  When viruses were treated with 10 μM lacidipine, the viral titer after 12 rounds of 171 passaging was about 100-fold higher than that in the wild type (WT) virus (Fig. 4A).

172
The lacidipine-resistant virus isolate was plaque purified and sequenced for the entire 173 glycoprotein precursor GPC region. An amino acid substitution was observed in the 174 isolated clone, but markedly absent in the DMSO-treated virus, resulting in a revealed that the T40K plaques were similar to the WT plaques, whereas T40R and 195 T40D plaques were smaller, and T40A plaques were mid-range. These results

196
suggested that the infectivity of SSP T40R and T40D mutant viruses is milder than 197 10 that of the WT virus, and position 40 in SSP was tolerable for K. 198 We next investigated sensitivity of the four mutant viruses to lacidipine. As shown in  (Fig. 6A). To investigate the functional significance of these residues, a 240 pseudo-type GTOV containing the mutants was used to evaluate lacidipine sensitivity. 241 12 GTOV was much less sensitive when either the V36M or V436A mutant was 242 generated, and lacidipine sensitivity was further reduced when both sites were 243 changed (Fig. 6B). As V36 is similarly conserved in NW pathogenic viruses, we 244 reasoned that residues other than these selective mutations contributed to the 245 sensitivity of LASV and GTOV to lacidipine. have proven to be essential in GPC maturation, fusion, and infectivity (12, 41). In the 276 current study, we demonstrate that replacement of T40 with a charged amino acid (K, 277 R, or D) confers LASV resistance to lacidipine, and position 40 is tolerable for K 278 without detectable changes in LASVrv growth kinetics. These results suggest that T40 279 together with lacidipine might interplay with the residues located in the 280 membrane-proximal ectodomain of GP2, thus stabilizing the native structure of GPC.

281
Although position 40 is tolerable for K, the stabilization interaction between T40 and 282 lacidipine is collapsed by the replacement of K.  Lacidipine-sensitivity depends on accessibility of the interface to the drug as well as 296 the stability of the drug-GPC complex. 297 We also identified phenothrin via HTS; phenothrin is a pyrethroid usually used in 298 pesticide products and is effective at inhibiting LASV entry. We demonstrated that 299 phenothrin has activity against MOPV, GTOV, and CHPV, with IC50 values lower 300 than 10 μM. The structure of phenothrin is similar to that of the LASV specific entry