Fusion activation in murine leukemia virus

Abstract

This thesis describes the fusion activation in Murine leukaemia virus (MLV). During entry of retroviruses into uninfected cells, the viral and cellular membranes fuse. The fusion process is mediated by the viral envelope protein composed of three SU and TM heterodimers. The peripheral SU subunit mediates receptor binding and the transmembrane TM subunit the essential fusion function. SU is believed to trap the TM in an inactive folding intermediate, which will be activated by SU dissociation. According to one model receptor binding triggers conformational changes in SU which loosen the SU and TM interaction. Thus subunit dissociation would be the key to control fusion activation. This has been presented as a model for HIV, as well as for MLV retroviruses but not yet proven. Here I present results that support the above model and further suggest a novel activation mechanism for the fusion function in MLV Env.

I found that the SU and TM subunits of MLV Env were linked by a disulphide bond. The disulphide bond has previously been found to be associated with a conserved CWLC peptide sequence in SU. This sequnce resembles an isomerization, motif found in the active site in redox enzymes involved in rearrangement of disulphide bonds in the cell. 1 demonstrate that the CWLC motif was activated by receptor binding to isomerize the intersubunit disulphide bond in Env. This event controlled SU dissociation and TM activation. Specifically, 1 found that the Cys residue of the motif, which was not involved in the inter-subunit disulphide bond, constituted the isomerization-active group. This was shown by thiol mapping of Env subunits before and after SU-TM disulphide bond rearrangement. Receptor binding activated the isomerization reaction. Activation or triggering was also possible by depleting Env of Ca2+, suggesting that the inactive form of Env was stabilized by Ca2+. This in vitro triggering conditions was used to analyze the effect of isomerization on the Env structure. The analysis showed that isomerization caused SU dissociation.

Both the free thiol and the intersubunit disulphide bond were shielded in the native Env structure. However Env triggering, e.g. by receptor binding, resulted in the concomittant exposure of the SU-TM disulphide bond and the CWLC thiol. This was shown by thiol alkylation and DTT reduction of the SUTM disulphide bond. The CWLC thiol alkylation blocked isomerization of the intersubunit disulphide bond and also arrested the fusion function. However DTT reduction of the SU-TM disulphide bond rescued the fusion function and the infectivity of the virus. This formally proved that isomerization, of the SU-TM disulphide bond controlled fusion. These experiments also revealed the alkylated, isomerization, blocked Env as an intermediate structure in Env activation.

Furthermore, I found that the CWLC controlled isomerization reaction could be activated by protein perturbating agents like urea and heat. This indicated that destabilization of the interaction between SU and TM activates isomerization and that TM persists as a metastable protein in the native Env. Finally, biochemical and functional analyzes of the alkylated, isomerization blocked Env intermediate, suggested that MLV fusion involved fusion sites composed of several Env-receptor complexes. Altogether we present here a novel fusion activation pathway for MLV Envs. This involves an internal isomerase motif that triggers an SU-TM disulfide isomerization reaction in the MLV Env leading to fusion.