The underlying mechanisms of arenaviral entry through matriglycan

Abstract

Matriglycan, a recently characterized linear polysaccharide, is composed of alternating xylose and glucuronic acid subunits bound to the ubiquitously expressed protein α-dystroglycan (α-DG). Pathogenic arenaviruses, like the Lassa virus (LASV), hijack this long linear polysaccharide to gain cellular entry. Until recently, it was unclear through what mechanisms LASV engages its matriglycan receptor to initiate infection. Additionally, how matriglycan is synthesized onto α-DG by the Golgi-resident glycosyltransferase LARGE1 remained enigmatic. Recent structural data for LARGE1 and for the LASV spike complex informs us about the synthesis of matriglycan as well as its usage as an entry receptor by arenaviruses. In this review, we discuss structural insights into the system of matriglycan generation and eventual recognition by pathogenic viruses. We also highlight the unique usage of matriglycan as a high-affinity host receptor compared with other polysaccharides that decorate cells.

Keywords: LARGE1; arenaviruses; lassa virus; matriglycan; receptor recognition; spike complex.

FIGURE 1

Proposed mechanism for matriglycan synthesis. LARGE1 is a homodimeric enzyme that synthesizes matriglycan in a processive manner on α-DG (PDB 7ZVJ). The growing polysaccharide chain alternates between the xylosyltransferase (XylT) domain on one monomer (shown in green), and the glucuronyltransferase (GlcAT) domain on the second monomer (shown in orange). The ‘DXD’ motifs, which comprise part of the catalytic sites, are colored in yellow and pink for the XylT and GlcAT domains, respectively. The matriglycan chain likely wraps around the perimeter of the enzyme, adhering to positively charged patches on the protein’s surface, aiding the enzyme’s processivity. Inset: A schematic diagram of the domain organization of LARGE1, CT: Cytoplasmic, TM: Transmembrane, CC: Coiled Coil. This image was partially created by BioRender.com.

FIGURE 2

Mechanisms of matriglycan recognition by the LASV spike complex and LG4,5-domain-containing proteins. (A) Surface representations of the LG domains 4–5 of laminin α-2 chain, in complex with matriglycan (PDB 5IK5) alongside the Lassa spike complex, bound to matriglycan (PDB 7PVD). (B) The CA²⁺-dependent mechanism of the LG4-5 domain recognition of matriglycan, which chiefly relies on metal coordination. (C) The mechanism of LASV spike complex recognition of matriglycan, which depends on residues of a domain-swapped SKI-1/S1P site. This interaction is dominated by hydrogen bonds and electrostatic interactions. (D) A schematic diagram for the domain organization of the LASV spike complex. SSP: structured signal peptide, GP1: Glycoprotein-1 subunit, GP2: glycoprotein-2 subunit, TM: Transmembrane domain, CT: Cytoplasmic domain. The signal peptidase-1 (SPase) and SKI-1/S1P cleavage sites are indicated. The specific cleavage site by SKI-1/S1P is shown with a red arrow directly after the “RRLL” motif. Note that the use of a single matriglycan chain between the structures, along with only a partial description of key residues used for binding, was carried out for artistic purposes. (E) Stabilization of the LASV spike complex through domain swapping of the terminal GP1 regions (PDB: 7PUY). Monomers (A–C) of the trimeric spike complex are shown in pink, gray, and blue, respectively. The final ∼10 residues of GP1, which include the SKI-1/S1P recognition motif, are shown in stick representation. These segments are interlocked with a neighboring monomer through hydrophobic interactions, while exposing critical residues for matriglycan binding.