The low-density lipoprotein receptor promotes infection of multiple encephalitic alphaviruses

Abstract

Members of the low-density lipoprotein receptor (LDLR) family, including LDLRAD3, VLDLR, and ApoER2, were recently described as entry factors for different alphaviruses. However, based on studies with gene edited cells and knockout mice, blockade or abrogation of these receptors does not fully inhibit alphavirus infection, indicating the existence of additional uncharacterized entry factors. Here, we perform a CRISPR-Cas9 genome-wide loss-of-function screen in mouse neuronal cells with a chimeric alphavirus expressing the Eastern equine encephalitis virus (EEEV) structural proteins and identify LDLR as a candidate receptor. Expression of LDLR on the surface of neuronal or non-neuronal cells facilitates binding and infection of EEEV, Western equine encephalitis virus, and Semliki Forest virus. Domain mapping and binding studies reveal a low-affinity interaction with LA domain 3 (LA3) that can be enhanced by concatenation of LA3 repeats. Soluble decoy proteins with multiple LA3 repeats inhibit EEEV infection in cell culture and in mice. Our results establish LDLR as a low-affinity receptor for multiple alphaviruses and highlight a possible path for developing inhibitors that could mitigate infection and disease.

Fig. 1. LDLR promotes infection of EEEV and other alphaviruses in cells.

a Results of CRISPR-Cas9 genome wide loss-of-function SINV-EEEV infection screen in ΔB4galt7 N2a cells. Enriched genes based on robust rank aggregation (RRA) scores in the SINV-EEEV-selected population. b ΔB4galt7 (Control) or ΔB4galt7 ΔLdlr (ΔLdlr) N2a cells were inoculated with SINV-EEEV-GFP (FL93-939), SINV-VEEV-GFP (TrD), SINV-WEEV-GFP (CBA87), SINV-GFP (TR339), SINV-SFV-GFP (SFV4), or SINV-CHIKV-GFP (LR2006) for 6.5 to 24 h (depending on the virus), and infection was assessed by flow cytometry for GFP expression (n = 3 experiments). c Multistep growth curves of SINV-EEEV, SINV-WEEV, SINV-CHIKV, and SINV in ΔB4galt7 (Control), ΔB4galt7 ΔLdlr (ΔLdlr + EV), and LDLR-complemented ΔB4galt7 ΔLdlr (ΔLdlr + Ldlr) N2a cells (n = 3 experiments). d Human THP-1 cells ectopically expressing empty vector (EV), Ldlr, LDLRAD3, and VLDLR were inoculated with SINV-EEEV-GFP, and infection was assessed by flow cytometry (n = 6 experiments). e Multistep growth curves of MADV (Argentina 1936) in ΔB4galt7 (Control), ΔB4galt7 ΔLdlr (ΔLdlr + EV) and complemented ΔB4galt7 ΔLdlr (ΔLdlr + Ldlr) N2a cells (n = 3 experiments). Black asterisks indicate comparisons between ΔB4galt7 and ΔB4galt7 ΔLdlr N2a cells. Blue pound signs indicate comparisons between ΔB4galt7 ΔLdlrad3 and LDLR-complemented ΔB4galt7 ΔLdlrad3 N2a cells. f, g Human THP-1 cells ectopically expressing empty vector (EV), Ldlr, LDLRAD3, or VLDLR were inoculated with SINV-WEEV-GFP (f) or SINV-SFV-GFP (g), and infection was assessed by flow cytometry (n = 3 experiments). Means ± SD are shown. Statistical analysis (P values from left to right): (b) two-tailed unpaired t test: ***P = 0.0009, ***P = 0.0007, *P = 0.0251; (c, e) two-way ANOVA with Dunnett’s post-test: *P = 0.0269, ***P = 0.0002, ***P = 0.0003, **P = 0.0055, **P = 0.0075, ^###P = 0.0007, ^###P = 0.0004, ^####P < 0.0001, ^##P = 0.0036 (c, SINV-EEEV); *P = 0.0467, *P = 0.0178, ****P < 0.0001, **P = 0.0066, **P = 0.0017, ^#P = 0.0168, ^###P = 0.0003, ^##P = 0.0017, ^####P < 0.0001 (c, SINV-WEEV); *P = 0.0461, **P = 0.0053 (c, SINV-CHIKV); *P = 0.0244, ^#P = 0.0186, ^##P = 0.0024 (c, SINV); *P = 0.0280, **P = 0.0038, ***P = 0.0004, ****P < 0.0001, **P = 0.0075, ^#P = 0.0170, ^####P < 0.0001, ^###P = 0.0005, ^####P < 0.0001 (e); (d, f, g) one-way ANOVA with Dunnett’s post-test: **P = 0.0092, **P = 0.0077 (d); ***P = 0.0002 (f); **P = 0.0031, *P = 0.0388, **P = 0.0034 (g); ns, not significant. Source data are provided as a Source Data file.

Fig. 2. LDLR expression promotes SINV-EEEV attachment and internalization.

a SINV-EEEV (FL93-939) was incubated with THP-1 cells expressing empty vector (EV), LDLR, or VLDLR at 4°C for 90 min. After extensive washing, bound virions were quantified as the ratio of viral RNA (vRNA) to GAPDH gene mRNA level via RT-qPCR and then normalized to the control EV cells. b After removal of unbound virus, the temperature was increased to 37°C for 1 h to allow internalization. After proteinase K digestion to remove surface-associated virus, intracellular viral RNA was measured and presented as in (a). Mean ± SD (n = 4 experiments in quadruplicate, all data points shown). Statistical analysis (P values from left to right) on mean values of four experiments: one-way ANOVA with Dunnett’s post-test: (a) **P = 0.0024, **P = 0.0039; (b) **P = 0.0013, **P = 0.0043. Source data are provided as a Source Data file.

Fig. 3. LDLR directly binds to EEEV virions.

a Schematic of domain structure of LDLR (left) and LA3-OL mini receptor (right). b–d Biolayer interferometry (BLI) graphs showing EEEV VLPs binding to immobilized full length LDLR-Fc (LDLR-Fc) (b), EEEV VLPs binding to immobilized single LDLR LA domains (LA-Fc) (c), and LDLR proteins binding to immobilized EEEV VLPs (d) (one of two experiments is shown). e Confocal microscopy images of LDLR puncta stained on the plasma membrane surface of N2a cells. LDLR (green), nuclei (DAPI, blue), and cell membrane (wheat germ agglutinin, WGA, magenta). WT denotes ΔB4galt7 N2a cells, and KO denotes ΔB4galt7 ΔLdlr N2a cells. Scale bar, 10 μm for low magnification images and 2 μm for zoomed in images (n = 2 experiments, with representative fields shown). f LA3 mini-receptor LA3-OL promotes infection of SINV-EEEV-GFP in transduced THP-1 cells (n = 3 experiments). g Mapping of the LDLR LA3 domain binding site for EEEV. THP-1 cells ectopically expressing indicated mutations were inoculated with SINV-EEEV-GFP. Cell surface expression data of mutants is shown in Supplementary Fig 3e. Red columns indicate mutations that significantly impair the ability of the LA3 domain to support SINV-EEEV infection (n = 4 experiments). Statistical analysis (P values from left to right): one-way ANOVA with Dunnett’s post-test: (f) **P = 0.0028, ****P < 0.0001; (g) one-way ANOVA with Welch and Brown-Forsythe post-test: *P = 0.0185, *P = 0.0252, ****P < 0.0001, *P = 0.0336, ****P < 0.0001, **P = 0.0022, ***P = 0.0003. Column heights indicate mean values, and error bars denote SD. Source data are provided as a Source Data file.

Fig. 4. Neutralization and protection of alphavirus infection by LA3 tandem repeat domain proteins.

a, b Effects of preincubation of LDLR-Fc (a), LBD-Fc (a), or LA3-Fc (b) on VSV and SINV-EEEV infection of THP-1 cells ectopically expressing LDLR (n = 3 experiments). hE16 is an anti-West Nile virus mAb (negative control). c Binding of soluble LA domain-Fc molecules with repeat domains to immobilized EEEV as measured by ELISA (n = 3 experiments). d Dose-dependent effect of preincubation of LA3-Fc, LA3(3)-Fc, and LA3(5)-Fc on SINV-EEEV infection (n = 3 experiments). e, Schematic of a pentamer of LA3(5). f–h Effect of pre-incubation of LA3(5)-COMP (or LDLRAD3 D1-COMP, negative control) protein on infection by SINV-EEEV (f), SINV-WEEV (g), or SINV-SFV (h) in THP-1 cells ectopically expressing LDLR (n = 3 experiments). i–l Four-week-old C57BL/6J mice were inoculated subcutaneously with 2 x 10² FFU of MADV (Argentina 1936) pre-mixed with 50 μg of LA3(5)-COMP or LDLRAD3-D1-COMP. A second 100 μg dose of LA3(5)-COMP or LDLRAD3-D1-COMP was administered by intraperitoneal injection at day +1. i Weights were measured daily through the duration of the experiment. j Survival was monitored through 14 days (i, j: n = 9, three experiments). k, l At 3 or 5 days post-infection, the indicated tissues were collected (n = 9 two experiments), and MADV RNA levels were quantified by RT-qPCR (k) and normalized to a standard curve from a stock of known titer. Infectious virus was quantified by focus-forming assay (l). Limits of detection (LoD) are indicated by dashes lines. Means ± SD are shown. Statistical analysis (P values from left to right): f, g one-way ANOVA with Dunnett’s post-test: ****P < 0.0001 (f). ****P < 0.0001, **P = 0.0087(g); j Log-rank (Mantel-Cox) test, ***P = 0.0003; k, l two-tailed Mann–Whitney test: **P = 0.0026 (spleen, 3 dpi), **P = 0.0090 (brain, 3 dpi), *P = 0.0267 (kidney, 5 dpi), **P = 0.0019 (brain, 5 dpi) (k); l **P = 0.0019; ns, not significant. Source data are provided as a Source Data file.