Activation of the spike proteins of alpha- and retroviruses

Abstract

Enveloped viruses like human immunodeficiency virus type 1 (HIV-1) enter cells by fusing their membrane with that of the cell plasma membrane, while others, like influenzavirus and alphavirus enter via uptake into endosomes and fusion with the endosomal membrane. For this they carry spike proteins. The spikes are composed of three copies of a subunit pair. One subunit has membrane fusion activity, whereas the other one binds to a structure (receptor) present on the cell surface. It also controls (chaperones) the activation of the fusion active subunit so that fusion does not occur prematurely. The fusion subunit is anchored in the virus membrane at one end and carries a fusion peptide (fp) at the other one. While hidden in the native spike the fp becomes exposed after spike activation (trigger). The fusion subunit interacts via the fp with the target membrane in an extended conformation and then forces the viral and the cell membranes together for fusion by a backfolding reaction. This model is mostly based on biochemical and structural studies using isolated subunits. However, a full understanding of the spike activation mechanism can only be obtained by studying the complete spike, i.e the trimer of the subunit pair. This was the aim of my thesis work.

I used low dose electron cryomicroscopy (cryo-EM) to capture images of my objects that had been frozen in liquid ethane in their native and partly activated states. This procedure facilitates the analysis of the virus and the spikes in their intact form, free from e.g. staining artifacts. By computer aided processing of the particle images, the three-dimensional structures were obtained. For my studies I used an alphavirus, Semliki Forest virus (SFV), and two retroviruses, Moloney mouse leukemia virus (Mo-MLV) and HIV-1. The SFV spike is triggered by low pH in the cell endosome and the retrovirus spikes by receptor binding.

One dilemma for SFV is that its spike during biosynthesis has to pass acidic compartments. How can it avoid activation? The chaperone and the fusion subunits are called E2 and E1. E2 is made as a precursor p62, which is cleaved by cellular furin into E2 and E3. The p62-E1 complex is acid resistant. I analyzed an SFV mutant with uncleaved p62-E1 spikes by cryo-EM and found that the E3 portion of p62 forms an extra contact with E1, which can explain the acid resistance of the precursor spike. A subsequent question was how low pH alleviated the E2 chaperoning of E1. In the native spike antibody mapping showed that the E1 fp is protected by E2 in the tip region. Partial triggering by low pH resulted in subunit dissociation in the membrane proximal but not in the distal fp containing part of the spike. This suggests a tight regulation of the steps in the fusion process.

While the SFV spike was studied in the context of the whole virus the retrovirus spikes were solubilized and isolated as trimers. Cryo-EM showed that the retrovirus spikes were hollow cage-like structures where the three subunit pairs formed a common roof and floor and separated lobes on the sides. Analyses of a partially activated form of the Mo-MLV spike showed an outward rotation of the top domain of the chaperone subunit. This opened up the cage from above, most likely to allow the fusion subunit to reach out to the cell membrane. I conclude that cryo-EM offers a powerful approach to study virus activation.