Neutralizing Antibodies against Lassa Virus Lineage I

Abstract

Lassa virus (LASV) is the causative agent of the deadly Lassa fever (LF). Seven distinct LASV lineages circulate through western Africa, among which lineage I (LI), the first to be identified, is particularly resistant to antibody neutralization. Lineage I LASV evades neutralization by half of known antibodies in the GPC-A antibody competition group and all but one of the antibodies in the GPC-B competition group. Here, we solve two cryo-electron microscopy (cryo-EM) structures of LI GP in complex with a GPC-A and a GPC-B antibody. We used complementary structural and biochemical techniques to identify single-amino-acid substitutions in LI that are responsible for immune evasion by each antibody group. Further, we show that LI infection is more dependent on the endosomal receptor lysosome-associated membrane protein 1 (LAMP1) for viral entry relative to LIV. In the absence of LAMP1, LI requires a more acidic fusion pH to initiate membrane fusion with the host cell relative to LIV. IMPORTANCE No vaccine or therapeutics are approved to prevent LASV infection or treat LF. All vaccine platforms currently under development present only the LIV GP sequence. However, our data suggest that the high genetic diversity of LASV may be problematic for designing both a broadly reactive immunogen and therapeutic. Here, we examine antibodies that are highly potent against LIV yet are ineffective against LI. By pinpointing LI mutations responsible for this decrease in antibody efficacy, we suggest that future vaccine platforms may need to incorporate specific LI-like mutations in order to generate a broadly neutralizing antibody response against all LASV lineages.

Keywords: LAMP1; Lassa fever; Lassa virus; antigenic variation; cryo-EM; hemorrhagic fever virus; neutralizing antibodies; prefusion glycoprotein; structure-based vaccine design; structure-guided immunogen; viral escape.

FIG 1

Structural characterization of the lineage I Lassa virus prefusion glycoprotein. (A) Superimposition of each of the LI-GP trimer structures, with the LIV-GP trimer (PDB accession number 7S8H) shown. GP structures were aligned by a single GP monomer from each structure, resulting in an average RMSD of 1.2 Å between LI and LIV. (B) Side-view of the LI-pfGP-25.10C cryo-EM reconstruction. GP1 and GP2 are colored in light and dark gold, respectively. The light- and heavy-chain domains of the 25.10C Fab are colored light and dark green, respectively. (C) Side-view of the LI-GP-18.5C-M30 cryo-EM reconstruction. GP1 and GP2 are colored in light and dark gold, respectively. The light- and heavy-chain domains of the 18.5C-M30 Fab are colored blue and maroon, respectively. The orientation of the LI-pfGP is rotated 90° from panel B to panel C.

FIG 2

GPC-A antibody-mediated neutralization of LI LASV. (A) Superimposition of the LI-GP-25.10C atomic model (gold and green, respectively) with the LIV-GP-36.1F structure (gray and purple, respectively) (27). (Inset) The position of R95 in LI relative to the LIV-specific 36.1F suggests that this substitution would cause a steric clash with the CDRH3 of the antibody, indicated by a flash symbol. CDRH3 of the pan-LASV 25.10C instead makes a hydrogen bond with the conserved E75 side chain (red dots). (B) Kinetic binding curves for interaction between the GPC-A antibodies 25.10C and 36.1F with the indicated LASV GP monomer. For each panel, the raw data is colored according to GP concentration with the 1:1 fit shown in red. Each experiment was repeated twice, producing similar trends; results from one experiment are shown. (C) Neutralization of wild-type ppVSV-LI-GP and ppVSV-LI-GP bearing an R95M substitution by GPC-A MAbs 25.10C (green) and 36.1F (purple). Each data point is the average of two biological replicates, where each replicate was performed in technical duplicate, with the error bars indicating the standard deviation (SD) from the mean. The data were normalized to infection of Vero cells by ppVSV-LI-GP without MAb.

FIG 3

Arginine insertions increase neutralization potency of GPC-B MAb 18.5C. (A) Neutralization of ppVSV-LI by wild-type 18.5C and enhanced 18.5C-M30. Each data point is the average of two biological replicates, where each replicate was performed in technical duplicate, with the error bars indicating the SD from the mean. The data were normalized to infection of Vero cells by ppVSV-LI-GP without MAb. (B) Depictions of the novel salt-bridges (red dots) formed by engineered arginine residues at positions 54 and 101 of the 18.5C-M30 heavy chain with LI (gold) residues D400 and D407, respectively.

FIG 4

Single residue substitutions in LASV-LI decrease GPC-B antibody neutralization activity. (A) Cryo-EM structure of LI (gold) in complex with 18.5C-M30 (maroon) and crystal structure of LIV (gray) in complex with 18.5C (blue) (PDB accession number 6P91) shown as cartoons. In LIV Y62 hydrogen bonds with 18.5C heavy chain K58 (black dots) to stabilize the N-terminal β-strand of LIV. In LI, which lacks Y62, this region is disordered. (B) Neutralization of wild-type and mutant ppVSV-LI-GP bearing a Y62ins by 18.5C (left) and 25.6A or 37.7H (right). (C) Superimposition of LASV-LI GP (gold) in complex with 18.5C-M30 (light blue) and LASV-LIV GP (gray) shown as a cartoon. The 196 to 207 loop is colored magenta with residue S199 shown as sticks. The 196 to 207 loop is disordered in LI and likely adopts a different orientation relative to LIV, due to the S198R substitution (modeled here in cyan) that would sterically clash with surrounding GP1 elements. The FA2B glycan at position N390 (48) is modeled and oriented to the location of the single ordered NAG residue in LI GP. (D) Neutralization of wild-type and mutant ppVSV-LI-GP with either R198S or with genetic removal of glycan N389 via an N389D mutation by 18.5C (left) and 25.6A or 37.7H (right). (E) Cartoons of LI-GP and LIV in complex with 18.5C-M30 or parental 18.5C, respectively, are shown. H398 in LIV GP forms a pi-cation stacking interaction with Y33 of CRDL1 and hydrogen bonds with the mainchain oxygen of Y92 in CDRL3 (black dots). Q397 of LI can form a hydrogen bond only with the mainchain oxygen of Y92 (red dots). (F) Neutralization of wild-type and mutant ppVSV-LI-GP bearing a Q397H mutation by 18.5C (left) and 25.6A or 37.7H (right). In panels B, D, and F, each data point represents the average of two biological replicates, each performed in technical duplicate, with error bars indicating the SD from the mean. Data were normalized to infection by ppVSV-LI-GP without MAb. The same wild-type ppVSV-LI neutralization data are shown in each panel for comparison.

FIG 5

Greater dependency on LAMP1 for cellular entry of LI-GP relative to LIV-GP. (A) Superimposition of LI with 18.5C-M30, LI with 25.10C, and LIV with 18.5C (PDB accession number 6P91) at the histidine triad. Residues R95, Y93, and H229 exhibit greater mobility in LI-GP than LIV-GP. (B) Infectivity assays measuring relative infection of Vero cells and HAP1/LAMP1⁻ cells by pseudovirus as a function of virus concentration. ppVSV-LI-GP and ppVSV-LIV-GP are shown in gold and gray, respectively. Each data point represents the mean percentage of cells infected with pseudovirus. Error bars indicate the SD from the mean of two biological replicates that each have two technical duplicates. (C) Fusogenic profile of ppVSV-LASV in Vero cells as a function of pH. Each data point is the average of two biological experiments, each with four technical replicates. Error bars indicate the SD from the mean. Data were normalized to the maximum infectivity of ppVSV-LI-GP as 100% of the control in which pseudovirions naturally infected cells in the absence of treatment with acid or lysosomotropic agents (NH₄Cl).