Fig 1.
3.3 inhibits LASV GP-mediated infection and cross-links to the LASV receptor, LAMP1, in cells.
(A) Structure of 3.3 (left) and inhibition of transduction by MLV pseudotyped with the indicated viral glycoproteins (right). Vero cells were incubated with the indicated concentrations of 3.3 for 1h before challenge with MLV encoding GFP and pseudotyped with the indicated viral glycoproteins. Virus transduction is reported as % of GFP-positive cells relative to cells exposed to DMSO vehicle alone. Data are mean ± SD (n = 3). IC50 is the concentration of inhibitor required to reduce transduction by 50%. LUJV: Lujo virus; JUNV: Junin virus; LCMV: Lymphocytic choriomeningitis virus; VSV: Vesicular stomatitis virus; EBOV: Ebola virus. (B) Structure of 1519 (left). 1519 contains a diazirine for target cross-linking and a terminal alkyne for Cu+-catalyzed click-based analysis and inhibits transduction by MLV pseudotyped with LASV GP (right). Vero cells were incubated with the indicated concentrations of 1519 for 1h before challenge with MLV pseudotyped with LASV GP and encoding GFP. Virus transduction is reported as % of GFP-positive cells relative to cells exposed to DMSO vehicle alone. Data are mean ± SD (n = 3). Data was collected in the same experiment as Fig 1A. (C) 1519 co-localizes with LAMP1+ compartments in Vero cells and cross-links in a manner which is competed by the addition of a molar excess of 3.3 but not the inactive 3.3 derivative, 100. Vero cells were co-incubated with 1519 and the indicated competitor compound for 1h at 37°C before UV irradiation followed by click chemistry with TAMRA azide. LAMP1 was visualized by immunofluorescence using an anti-LAMP1 antibody. (D) 1519 co-localizes with LAMP1+ compartments in Huh7.5 cells and cross-links in a manner which is competed by the addition of a molar excess of 3.3 but not the inactive 3.3 derivative, 100. 1519 cross-linking in Huh7.5 LAMP1 knockout (KO) cells is diminished. Huh7.5 pX459 or LAMP1 KO cells were co-incubated with the indicated compounds for 1h at 37°C before UV irradiation followed by click chemistry with TAMRA azide. LAMP1 was visualized by immunofluorescence using an anti-LAMP1 antibody. (E) 1519 cross-links to endogenous LAMP1 in a manner which is competed by the addition of a molar excess of 3.3 but not the inactive 3.3 derivative, 102. Vero cells were incubated with the indicated compounds for 1h at 37°C before UV irradiation. Cells were harvested, lysed and used for click chemistry to attach an Alexa Fluor 488 (AF 488) azide and then subjected to immunoprecipitation using an anti-AF 488 antibody. AF 488 labeled LAMP1 was detected using an anti-LAMP1 antibody. Total AF 488 labeled proteins were detected via in-gel fluorescence. Input protein was detected by Coomassie staining.
Fig 2.
Binding of LASV GP to LAMP1 is dependent on cholesterol and is inhibited by 3.3.
(A) 3.3 blocks the interaction of LASV GP with endogenous LAMP1. 293T LAMP1 KO cells expressing LASV GP-His and 293T pX459 cells were incubated with the indicated concentrations of 3.3 for 1h at 37°C before lysis. Lysates were mixed and subjected to immunoprecipitation with an anti-His antibody. The presence of LASV GP bound LAMP1 was detected by immunoblot with an anti-LAMP1 antibody. (B) LASV GP1-IgG binds to LAMP1 D1 in a cholesterol-dependent manner. Purified LAMP1 D1-His was incubated with the indicated concentrations of cholesterol or epicholesterol for 1h at 37°C before the addition of purified LASV GP1-IgG. Samples were subjected to immunoprecipitation against human IgG and bound LAMP1 D1 was detected by immunoblot with an anti-His antibody. (C) 3.3 inhibits the cholesterol-dependent binding of LASV GP1-IgG to LAMP1 D1. Purified LAMP1 D1-His was co-incubated with 5μM cholesterol and the indicated concentrations of 3.3 or 102 for 1h at 37°C before the addition of purified LASV GP1-IgG. Samples were subjected to immunoprecipitation against human IgG and bound LAMP1 D1 was detected by immunoblot with an anti-His antibody.
Fig 3.
3.3 protects LAMP1, but not LAMP2, from thermal denaturation in cells.
(A) 3.3 but not the inactive derivatives, 100, 102 and 103, raises the Tm of endogenous LAMP1. 293T cells were incubated with 10μM of the indicated compounds or DMSO for 1h at 37°C. Cells were harvested and heated to the indicated temperatures prior to lysis. Lysed samples were centrifuged and the supernatants were analyzed by immunoblot with an anti-LAMP1 antibody. (B) 3.3 does not alter the thermal denaturation profile of endogenous LAMP2. 293T cells were incubated with 10μM 3.3 or DMSO for 1h at 37°C. Cells were harvested and heated to the indicated temperatures prior to lysis. Lysed samples were centrifuged and the supernatants were analyzed by immunoblot with an anti-LAMP2 antibody. (C) 3.3 dose-dependently protects endogenous LAMP1 from thermal denaturation at 80°C. 293T cells were incubated with the indicated concentrations of 3.3 or DMSO for 1h at 37°C. Cells were harvested and heated to 80°C prior to lysis. Lysed samples were centrifuged and the supernatants were analyzed by immunoblot with an anti-LAMP1 antibody.
Fig 4.
Cross-linking of 1519 to LAMP1 D1 is inhibited by cholesterol and vice versa.
(A) A photoreactive cholesterol analog (photoclick cholesterol) cross-linking to purified LAMP1 D1 is competed by a molar excess of cholesterol or 3.3 but not by a molar excess of epicholesterol or 102. Purified LAMP1 D1-His was incubated with the indicated compounds for 1h at 37°C prior to UV irradiation and click chemistry with an AF 488 azide. Labeled protein was detected by immunoblot with an anti-AF 488 antibody. Input LAMP1 D1 was detected using an anti-His antibody. (B) 1519 cross-links to purified LAMP1 D1 in a manner which is competed by a molar excess of 3.3 but not the inactive derivative, 102. Purified LAMP1 D1-His was incubated with the indicated compounds for 1h at 37°C prior to UV irradiation and click chemistry with an AF 488 azide. Labeled protein was detected by immunoblot with an anti-AF 488 antibody. Input LAMP1 D1 was detected using an anti-His antibody. (C) 1519 cross-links to purified LAMP1 D1 in a manner which is competed by a molar excess of 3.3 and cholesterol but not epicholesterol. Purified LAMP1 D1-His was incubated with the indicated concentrations of 1519 and 3.3, cholesterol or epicholesterol for 1h at 37°C prior to UV irradiation and click chemistry with an AF 488 azide. Labeled protein was detected by immunoblot with an anti-AF 488 antibody and input LAMP1 D1 was detected using an anti-His antibody.
Fig 5.
3.3 binding to the hydrophobic central pocket of LAMP1 D1.
(A) Docking of 3.3 (yellow) in the predicted structure of LAMP1 D1. Residues within the hydrophobic pocket predicted to contact 3.3 are colored orange. Residues on the surface at the edge of the pocket predicted to contact the diphenyl group of 3.3 are colored red. (B) Mutational analysis of predicted 3.3 contacts with LAMP1 D1 on cross-linking by photoclick cholesterol and by 1519. Purified LAMP1 D1 containing the indicated mutations were incubated with photoclick cholesterol (top) or 1519 (bottom) for 1h at 37°C prior to UV irradiation and click chemistry with AF 488 azide. Labeled protein was detected by immunoblot with an anti-AF 488 antibody. Input LAMP1 D1 was detected using an anti-His antibody. (C) Mutational analysis of predicted 3.3 contacts with LAMP1 D1 on transduction by MLV LASV GP and 3.3 inhibition. 293T LAMP1 KO cells transfected with the indicated LAMP1 mutants were incubated with the indicated concentrations of 3.3 for 1h before challenge with MLV pseudotyped with LASV GP and encoding GFP. Virus transduction is reported as % of GFP-positive cells relative to cells exposed to DMSO vehicle alone. Inset shows the MLV LASV GP transduction efficiency relative to that on cells transfected with wild-type (WT) LAMP1. Data are mean ± SD (n = 3). The IC50 of the WT curve differs significantly from all other curves (p < 0.001). Inset: ns (not significant); *** (p < 0.001). Statistical analyses were performed using one-way ANOVA with Tukey post-hoc analysis.