Ivanovic Lab                   B randeis University, Biochemistry Department

Click here to edit subtitle


Membrane fusion is at the heart of many essential biological processes, including intracellular transport, neurotransmission, gamete union, and cell entry by enveloped viruses. Although our current molecular picture is incomplete for any membrane fusion mechanism, it is now possible to approach such a goal through advanced microscopy techniques that allow the study of biological processes at the single-molecule level. 

Influenza hemagglutinin (HA), cell-recognition and fusion machinery of influenza virus, is the most well studied viral cell entry protein, a key determinant of influenza zoonotic adaptation, but greatly underutilized as drug target. Our goal is to obtain an integrated molecular picture of distinct HA functions, and, in the longer term, also of other viral processes that indirectly influence HA function.

Model of Influenza HA-Mediated Membrane Fusion Deriving from Our Work: (i) Attachment: Influenza HAs densely coved the virion surface (>500 HAs per virion). Individual HA monomer affinities for the receptor are low (mM), but they combine into femtomolar avidities for virions. (ii) Insertion: low pH within endosomes triggers HA conformational changes anywhere at the target membrane interface at random. Productive HAs insert into the target membrane. Otherwise they irreversibly inactivate. (iii) Fusion: 3-5 HA neighbors combine forces to induce membrane merger. 


Predicting adaptation potential of influenza viruses from the sequence or available functional data is a specific case of a ‘grand-challenge’ problem in evolutionary biology – namely, predicting evolutionary paths in a given environment. Solving it would lay the foundation for exploring a fundamentally new approach to inhibitor design that would achieve lasting inhibition by directly targeting viral ability to respond to inhibitor pressure. Uncertainty concerning the dangers of emerging highly pathogenic avian influenza viruses and reported instances of resistance in avian isolates to our last-line anti-influenza drugs is a good example where acquiring this foresight would have major immediate public-health consequences. We are addressing this question both in the context of viral adaptation to cell-entry inhibitors and pandemic adaptation. For these projects, we are moving beyond defining how membrane fusion works; we seek to define the mechanism(s) of its functional change under evolutionary pressure.

Proposed ‘Evolutionary Trap’. Shades of gray represent WT virions. Orange and blue represent less fit genotypes. Mutations that eliminate inhibitor binding (white dots for inhibitor A and white spirals for inhibitor B) generate unfit virions when they arise in a WT background. However, when they arise in an orange or a blue genetic background (for inhibitors A and B, respectively), fitness is restored. If orange virions are more susceptible to inhibitor B than are WT virions, and if blue virions are more susceptible to inhibitor A than are WT virions, a combination of inhibitors then eliminates the context for evolution of resistance.


The general steps of viral membrane fusion were proposed in the mid-nineties, when the crystal structures of the influenza HA in its prefusion and postfusion conformations were solved. Molecular structures of the intermediates remain unknown. We combine single-molecule kinetic analysis of influenza membrane fusion with high-resolution structural characterization of purified HA constructs that trap functionally relevant fusion intermediates.

Negative-Stain Electron Micrographs of Purified Recombinant HA. (Left panel) Neutral-pH HA. (Middle panel) Neutral-pH HA with C-termini bound to 5-nm gold beads forming rosettes. (Right panel) Low-pH HA rosettes of trapped HA refolding intermediates. Gold beads bound to the HA C-termini outline rosette periphery indicative of at most partial HA refolding. Related HA constructs will be used for high-resolution cryo-EM structural analysis.


We are developing single-molecule assays to enable the study of membrane penetration by nonenveloped viruses using reovirus as a model system. Noenveloped viruses lack the cell-derived membrane envelope, and, thus, cannot enter cells with the benefit of membrane fusion. Our appreciation of the molecular mechanism of membrane penetration by nonenveloped viruses lags behind that of enveloped-virus fusion.

Reovirus Particle Envelopment. Negative-stain EM of virus particles getting enveloped by lipid membranes.