Structural basis of glycan interaction in gastroenteric viral pathogens

Abstract

A critical event in the life cycle of a virus is its initial attachment to host cells. This involves recognition by the viruses of specific receptors on the cell surface, including glycans. Viruses typically exhibit strain-dependent variations in recognizing specific glycan receptors, a feature that contributes significantly to cell tropism, host specificity, host adaptation and interspecies transmission. Examples include influenza viruses, noroviruses, rotaviruses, and parvoviruses. Both rotaviruses and noroviruses are well known gastroenteric pathogens that are of significant global health concern. While rotaviruses, in the family Reoviridae, are the major causative agents of life-threatening diarrhea in children, noroviruses, which belong to the Caliciviridae family, cause epidemic and sporadic cases of acute gastroenteritis across all age groups. Both exhibit enormous genotypic and serotypic diversity. Consistent with this diversity each exhibits strain-dependent variations in the types of glycans they recognize for cell attachment. This chapter reviews the current status of the structural biology of such strain-dependent glycan specificities in these two families of viruses.

Figure 1

Rotavirus cryo-EM structure and the crystal structures of cell attachment protein VP8*. (A) The triple layered particle (TLP) is colored with VP4 spikes in red, the VP7 layer in yellow, and the VP6 layer in blue. The cartoon representation of a VP4 spike (PDB ID: 3IYU) is shown with the VP8* domain colored in red and the VP5* domain in orange. (B) Structural overlay of sialidase insensitive P[14] VP8* structure (blue, PDB ID: 4DRV) with VP8* of sialidase-insensitive HR strain Wa (green, PDB ID: 2DWR). The width of the cleft between two twisted β-sheets in the P[14] VP8* is narrower (red arrow) than in the Wa VP8* structure (black arrow). Amino acid 136 is shown in stick and indicated by a black arrow. (C) Interaction between the P[3] VP8* of sialidase sensitive animal strain RRV (PDB ID: 1KQR) and Sia. The P[3] VP8* structure in presented in orange ribbon with the amino acid residues interacting with Sia shown as sticks, and bound Sia is shown as green sticks with oxygen and nitrogen atoms in red and blue, respectively. (D) Interactions between the P[14] VP8* of sialidase insensitive human strain Hal1166 with A-type HBGA. The P[14] VP8* structure in presented in blue ribbon with the amino acid residues interacting with A-HBGA shown as yellow sticks with oxygen and nitrogen atoms colored as in (C). Network of hydrogen bond interactions (dashed lines) are shown.

Figure 2

(A) Sequence alignment of representative VP8*s of different genotypes. The amino acids are colored using Clustal protein color scheme in Jalview [insert ref later]. (B) Classification of VP8* into A–D classes.

Figure 3

(A) X-ray structure of Norwalk virus capsid (PDB ID: 1IHM). The shell domain (S) is shown in blue, the PI and P2 subdomains of the protruding P domain are shown in red and yellow, respectively. (B) Cartoon representation of the P-domain dimer (side view) from the GI.1 HNoV bound to H-type HBGA (PDB ID: 2ZL6). The HBGA binding site lies on the distal P2 subdomain. The P2 subdomain of the individual subunits of the dimer are colored in green and magenta respectively (here and subsequent Figs), and their P1 subdomains are colored in dark and lighter grey. (C) Topology diagram of the P domain highlighting the locations of HBGA binding sites in GI NoVs (yellow box) and GII NoVs (yellow box with lines). The antiparallel β strands (1–6) forming a barrel-like structure in the P2 subdomain are indicated by vertical arrows (magenta) and those in the P1 subdomain (7,8) that contribute to HBGA binding are indicated by grey. The variable loops connecting the β-strands are denoted A–D, P, T and S. (D) Surface representation (top view) of GI P-dimer (PDB ID: 2ZL5) showing distinct HBGA-binding sites (yellow circle) in each of the subunits. Residues that interact with Gal and Fuc moieties of the HBGA (shown in yellow stick representations) are shown in blue and gold colors, respectively. (E) Surface representation (top view) of the GII P domain dimer (PDB ID: 3SLN) showing the HBGA binding site shared between the opposing subunits in the dimer. Residues that interact with the Fuc moiety of the HBGA (shown in yellow stick representations) are colored in yellow.

Figure 4

(A) Gal-centric HBGA interactions in GI NoVs. Shown here as an example is interactions between GI.1 and H-type HBGA. A-type HBGA makes similar interaction with its Gal and N-acetamido groups of N-acetylgalactosamine similar to the Gal and Fuc moieties of the H-type. (B) Fuc-centric HBGA interactions in GII NoVs, shown here as an example is interactions between GII.4 P domain and A-type HBGA (PDB ID: 3SLD). (C) Alterations in length and structure of the P-loop that allows GI.7 bind non-secretor Le^a (PDB I.D. 4P3I), GI.1 with a shorter P–loop cannot make similar interactions (D) Structural alterations in the T-loop that allows 2004 GII.4 variant to interact additionally with di-fucosyl secretor Lewis HBGA (Le^b) (PDB ID: 3SLD), similar interactions with Le^b are not possible in the 1996 GII.4 variant (cyan). The interacting P domain residues are shown as sticks with oxygen and nitrogen atoms in red and blue, respectively.