Mapping the structure and function of the E1 and E2 glycoproteins in alphaviruses

Abstract

The 9 A resolution cryo-electron microscopy map of Sindbis virus presented here provides structural information on the polypeptide topology of the E2 protein, on the interactions between the E1 and E2 glycoproteins in the formation of a heterodimer, on the difference in conformation of the two types of trimeric spikes, on the interaction between the transmembrane helices of the E1 and E2 proteins, and on the conformational changes that occur when fusing with a host cell. The positions of various markers on the E2 protein established the approximate topology of the E2 structure. The largest conformational differences between the icosahedral surface spikes at icosahedral 3-fold and quasi-3-fold positions are associated with the monomers closest to the 5-fold axes. The long E2 monomers, containing the cell receptor recognition motif at their extremities, are shown to rotate by about 180 degrees and to move away from the center of the spikes during fusion.

Figure 1. The T = 4 Quasi-Symmetry of SINV

(A) Surface-shaded view of Sindbis virus shown at 9 Å resolution. Left half: The four independent E2 monomers related by T = 4 quasi-symmetry are colored yellow (i3), blue (m4), green (m5), and red (m6). The three yellow monomers are identical in structure and constitute the spikes at the icosahedral 3-fold axes. The blue, green, and red monomers are similar, but not identical, in structure and constitute the spikes at the quasi-3-fold axes. The triangle delineates 1 of the 60 icosahedral asymmetric units. Right half: A cross-section of the virus particle showing the overall organization of the particle. The nucleocapsid is colored red, the lipid bilayer is green, and the glycoproteins are shown in blue. (B) Diagram of the T = 4 quasi-symmetry-related monomers of the E1 protein. The yellow (i3), blue (m4), green (m5), and red (m6), monomers of E1 form heterodimers with the corresponding colored E2 molecules.

Figure 2. Stereodiagrams Showing the Fitted E1 Glycoprotein

Top: The Cα backbone of the E1 monomer fitted into the 9 Å resolution cryo-EM map is shown at the spike around the icosahedral 3-fold axis. E1 is shown in red when fitted as an entire molecule, and it is shown in blue when fitted as two independent rigid bodies (domains I+II and domain III). Center: Domains I and II of the four quasi-symmetry-related E1 monomers shown in yellow, blue, green, and red (as in Figure 1) are superimposed, showing the relative position of domain III. The red monomer, which is closest to the 5-fold axis (see Figure 1), has the greatest deviation from the other monomers. Bottom: The E1 trimer surrounding the icosahedral 3-fold axes (green) superimposed on the E1 trimer surrounding the quasi-3-fold axes (purple) is shown.

Figure 3. Steps in the Maturation and Fusion Processes Involving the Flavivirus E and Alphavirus E1 Proteins

Flavivirus E is shown at the top; alphavirus E1 is shown at the bottem. The maturation process of flaviviruses requires extensive conformational changes of the E protein as the spikes of the immature particle rearrange to become a smooth-surfaced mature particle. Flaviviruses have been postulated to have an intermediate state before fusion and formation of E trimers. Mature alphaviruses have an arrangement of their E1 protein similar to the proposed fusion transition intermediate of flaviviruses. The E1 trimeric postfusion spike is similar to that of flaviviruses.

Figure 4. Mapping the E2 Ectodomain Markers

Left: The E2 difference density for two surface spikes is shown in gray. The Cα backbones of the four quasi-symmetry-related E1 monomers are shown in yellow, blue, green, and red. Center: A top view of the E2 density of one spike looking down the 3-fold axis (shown as the larger purple spot corresponding to the three merged carbohydrate moieties at position 46). Right: A side view of one E2 molecule. Center and Right: The markers on the E2 glycoprotein that correspond to carbohydrate moieties at positions 46, 160, 196, 200, 216, 262, and 318 are shown in purple, red, blue, orange, green, pink, and light blue, respectively. Position 216 (green) was also identified with a cryo-EM map of a Fab-Ross River virus complex. Position 1 (black) was determined by a difference map between SFV (which contains E3) and SINV (which lacks E3).

Figure 5. Identification of Putative E1 Residues that Contact E2 in the Pre- and Postfusion States

(A) The icosahedral 3-fold trimer of E1 molecules (gray) is shown, with residues that contact the E2 density in red. (B) The location of these same residues mapped onto the E1 trimer in the postfusion complex is shown. A majority of these residues are on the external surface of the postfusion trimer. (C) The icosahedral 3-fold trimer of E1 molecules (gray) is shown with residues (green) that make intermonomer contacts in the postfusion state. (D) The postfusion trimer shown with residues (green) that make intermonomer contacts. These observations show that the E1 molecules are rotated about their long axes during the transition from the prefusion to the postfusion state, thereby presumably moving the E2 molecules from the inside to the outside of the E1 trimer while the E1-E2 heterodimer dissociates.

Figure 6. The E1 and E2 Transmembrane Helices

Stereodiagram of E1 residues from 409 to 439 and E2 residues from 363 to 398 fitted into their transmembrane densities. The bend in the E1 density around residue Ala417 is present in all four quasi-symmetry-related positions.

Figure 7. Stereodiagram Showing the Remaining Density after Subtracting the Density Corresponding to the Fitted X-Ray Crystallographic Capsid Protein for Residues 114–265

The residues in the hydrophobic pocket that constitutes the E2 capsid protein binding site are shown in red. An arrow identifies residue 114 at the N terminus of the known capsid structure. The difference density that ends at the arrow probably corresponds to residues 97–113 of the capsid protein.