Engineered Human Cathelicidin Antimicrobial Peptides Inhibit Ebola Virus Infection

Abstract

The 2014-2016 West Africa Ebola virus (EBOV) outbreak coupled with the most recent outbreaks in Central Africa underscore the need to develop effective treatment strategies against EBOV. Although several therapeutic options have shown great potential, developing a wider breadth of countermeasures would increase our efforts to combat the highly lethal EBOV. Here we show that human cathelicidin antimicrobial peptide (AMP) LL-37 and engineered LL-37 AMPs inhibit the infection of recombinant virus pseudotyped with EBOV glycoprotein (GP) and the wild-type EBOV. These AMPs target EBOV infection at the endosomal cell-entry step by impairing cathepsin B-mediated processing of EBOV GP. Furthermore, two engineered AMPs containing D-amino acids are particularly potent in blocking EBOV infection in comparison with other AMPs, most likely owing to their resistance to intracellular enzymatic degradation. Our results identify AMPs as a novel class of anti-EBOV therapeutics and demonstrate the feasibility of engineering AMPs for improved therapeutic efficacy.

Keywords: Drugs; Molecular Biology; Viral Microbiology.

Graphical abstract

Figure 1

Amino Acid Sequences of Engineered AMPs All the engineered AMPs are C-terminally amidated and derived from the sequence of human native LL-37 (hLL-37). 17BI is a short name for 17BIPHE2 as previously published.

Figure 2

Engineered AMPs Inhibit the Infection of Pseudo-EBOV Virion in Cell Lines and Human Primary Macrophages (A) AMPs inhibited pseudo-EBOV (VSV-eGP) infection in HeLa cells. VSV-eGP viruses were preincubated with an EBOV-neutralizing monoclonal antibody 13C6 (at 10 μg/mL, as a positive control) or individual AMPs (at 5 μM) for 30 min at 37°C in PBS buffer (pH 7.4) and then added to HeLa cells. After 20 h of culture, HeLa cells were harvested for flow cytometry analysis of GFP expression (percentages of GFP-positive cells represent percentages of cells infected with VSV-eGP). (B) AMPs inhibited pseudo-EBOV (VSV-eGP) infection in HeLa cells in a dose-dependent manner. VSV-eGP viruses were added to HeLa cells together with monoclonal antibody 13C6 (at 5, 10, or 20 μg/mL) or individual AMPs (at 2.5, 5, or 10 μM). After 20 h of culture, HeLa cells were harvested for flow cytometry analysis to measure viral infection. ∗∗∗, p < 0.001 by two-way ANOVA test. (C) AMPs inhibit pseudo-EBOV (VSV-eGP) infection in human primary macrophages. Human monocytes were freshly purified from blood and differentiated to macrophages by culturing the cells in media containing M-CSF. Macrophages were then infected with VSV-eGP in the presence of mAb 13C6 (at 5, 10, or 50 μg/mL) or AMPs (at 2.5, 5, or 10 μM). Cells were analyzed by flow cytometry at 20 h post infection. ∗, p < 0.05; ∗∗, p < 0.01 by two-way ANOVA test. (D) A summary of the efficacy and toxicity of the AMPs in HeLa cells and human primary macrophages. Cells were infected with VSV-eGP together with AMPs at different concentrations ranging from 0.1 to 50 μM and analyzed by flow cytometry at 20 h post infection to evaluate the IC₅₀ (half maximal inhibitory concentration). To measure cytotoxicity, cells were treated alone with the AMPs at different concentrations for 20 h. Cell viability was measured by MTT cell proliferation assay to determine the TC₅₀ (half maximal toxicity concentration).

Figure 3

Engineered AMPs Inhibit Wild-Type EBOV Infection HeLa cells were preincubated with individual AMPs at different concentrations (16 serial dilutions from 50 μM) for 2 h at 37°C and then infected with EBOV (Zaire-Kiwit) at MOI of 4. At 24 h post infection, cells were fixed, permeabilized, and immuno-stained with anti-EBOV GP antibodies followed by fluorescence-conjugated secondary antibody to identify the infected cells. (A) Representative immunostaining images (red: nuclear staining; green: anti-EBOV GP staining). (B) A summary of IC₅₀ of the AMPs.

Figure 4

Engineered AMPs Inhibit EBOV Cell Entry but Do Not Affect EBOV Replication in a Mini-Genome System (A) Engineered AMPs targeted pseudo-EBOV (VSV-eGP) at the early stage of infection. Human monocyte-derived primary macrophages were treated with 5 μM of individual AMPs at 4 h before VSV-eGP infection (-4 h), at the same time with VSV-eGP (0 h), or at 4 or 8 h after VSV-eGP infection (+4 h or +8 h, respectively). Macrophages were harvested for flow cytometry analysis of GFP levels at 20 h post infection. ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001 by two-way ANOVA test. (B) Engineered AMPs inhibit the cell entry of Ebola virus-like particles (VLPs). Ebola VLPs were produced by co-expressing the EBOV matrix protein, VP40 (fused to β-lactamase), and the EBOV GP. Vero cells were pre-treated with AMPs (at 2.5, 5, and 10 μM) or cathepsin B inhibitor CA-074 methyl ester (at 50 μM as a positive control) for 1 h at 37°C and then infected with Ebola VLPs. At 4 h post infection, a membrane-permeable β-lactamase substrate (CCF-2AM) was added to the culture and incubated for 1 h at room temperature. The release of β-lactamase by VLPs into the cytoplasm of cells (which is the result of successful processing and cell entry of VLPs from the endosomes) was measured by fluorescence emission of CCF-2AM substrate within 4 h of infection. ∗, p < 0.05; ∗∗, p < 0.01 by two-way ANOVA test. (C) Engineered AMPs do not affect EBOV replication in a mini-genome system. The EBOV mini-genome plasmids (encoding a firefly luciferase reporter gene flanked by the leader and trailer sequences from EBOV genome) were co-transfected into HeLa cells with four supporting plasmids encoding EBOV replication complex components NP, L, VP35, and VP30 and a plasmid encoding Renilla luciferase (for normalization of transfection efficiency). The ratio of firefly to Renilla luciferase represented the relative levels of EBOV mini-genome replication. ∗∗∗, p < 0.001 by one-way ANOVA test.

Figure 5

LL-37 and Engineered AMPs Block the Cleavage of EBOV GP by CatB, but Not by CatL (A) LL-37 and engineered AMPs blocked CatB-dependent cleavage of EBOV GP. CatB was pre-incubated with the CatB inhibitor (CA-074 methyl ester at 50 μM, as a positive control) or AMPs (at 10 μM) in 100 mM sodium acetate buffer pH 5.0 at 37°C for 30 min. Ebola GP protein was added to the reaction and incubated at 37°C for 1 h. The reaction mixtures were boiled in SDS-PAGE loading buffer and applied to denatured SDS PAGE followed by western blot with anti-Ebola GP mAb 13C6 (to detect the 110 kD full-length GP) and rabbit anti-Ebola GP polyclonal antibodies (to detect the 19 kD cleaved GP product). (B) None of the tested AMPs blocked CatL-dependent cleavage of EBOV GP. The same reaction was performed as in (A) except using CatL instead of CatB.

Figure 6

AMPs Composed of L-Amino Acids Are Susceptible to Degradation by CatS and Blockage of CatS Enhances AMP Inhibition of EBOV Infection (A) AMPs composed of L-amino acids were susceptible to degradation by CatS. AMPs (5 μM) were incubated with or without CatS (8 μg/mL) in a sodium acetate buffer (pH 5.0) for 60 min at 37°C. The reaction mixture was boiled in SDS sample buffer and loaded onto 4%–20% gradient gel, followed by staining with Coomassie blue. (B) Blockage of CatS enhanced AMP inhibition of pseudo-EBOV infection. Vero cells were infected with VSV-eGP in the absence or presence of CatB inhibitor (50 μM) as a positive control, CatS inhibitor (75 μM), AMPs (5 μM), or the combination of CatS inhibitor (75 μM) and AMPs (5 μM). After 20–24 h of culture, cells were harvested and applied to flow cytometry analysis.