Past Research: Proteins, from Sequence Function Annotation to Molecular Modeling
Although I cloned a gene as an undergraduate, my research has been focused mostly on proteins. I first worked on membrane proteins with Dr. Ludger Johannes at the Curie Intitute (Paris), where I characterised the localization of four membrane proteins in lipid rafts.After this biochemical project, I moved towards computational work. During my stay with Dr. Christos Ouzounis and Dr. Benjamin Audit, at EBI (Cambridge, UK), I developed a novel method for protein function prediction based on sequence similarity clues [1,2].
Following this work, where proteins were considered at the sequence level only, I was motivated to get a better understanding of their behaviour, in molecular terms. For this purpose, the Kinesin was a challenging subject because of its exceptional mechanical properties: this video [3] shows a kinesin (green) walking along a microtubule (Red), and this one [4], gives an idea of the protein conformational changes taking place during this process. I used two main approaches to investigave the Kinesin's mechanical properties: (i) a detailed approach using molecular dynamics, and (ii) a coarse-grained model. The latter, called GNM (Gaussian Network Model), gave interesting results that will be published within the next months.
PhD Research: Protein Complexes - Classification, Evolution, and Assembly
It is becoming increasingly clear that most proteins do not function alone within cells, but rather form complexes [5], thus, the protein complex point of view is presumably the state to consider in order to best capture the molecular details of the organization of life. Surprisingly however, most studies on protein structure consider single proteins, or at best interacting domain pairs, while protein complexes can be made of multiple proteins, and each protein can be made of multiple domains.Since there was obviously scope for research on this topic, I decided to focus on protein complexes of known structure during my PhD. My first task has been to build a classification of complexes. Indeed, to quantify the diversity and understand the relationships between large sets of objects, a natural way to proceed is to classify them. Carl von Linné, during the 18th century, developed the classification of species now widely used in biology. Similarly, I created the 3D Complex classification, which provides the first framework for the investigation of protein complexes of known atomic structure [6,7].
I have been using the classification to ask specific questions about protein complexes, e.g.:
- What are the mechanisms driving their emergence and evolution?
- What are the mechanisms involved in their assembly?
Current Research: Evolvability, Evolution and Dynamics of Protein-Protein interactions
It has been known for a while that it is quite easy to disrupt a protein-protein interaction: all one needs is to knock-out a "Hot-Spot" residue. But does this work the other way around? In other words could a single point mutation often trigger the formation of new protein-protein interactions?From a simple "amino-acid" composition perspective the answer is yes [9]! Importantly, this suggests that protein-protein interactions might be more evolvable than we imagine. The downside though, is that it means many PPIs probably exist but have not evolved to achieve a particular function (those can be called non-functional or promiscous PPIs) [10,11,12].
Related to this, questions I am now addressing at the bench are:
- What is the impact of (i) abundance, (ii) localization, and (iii) surface composition of a protein on its promiscuous interactions?
- Can the dynamic character of interactions (e.g. the fact that they exist in one condition but not in another) help distinguish functional ones from promiscuous ones?
- Can we use the promiscuous character of proteins to our advantage?
References
[1] Levy ED, Ouzounis CA, Gilks WR, Audit B. Probabilistic annotation of protein sequences based on functional classifications. BMC Bioinformatics. 2005 Dec 14;6:302.[2] Audit B, Levy ED, Gilks WR, Goldovsky L, Ouzounis CA. CORRIE: enzyme sequence annotation with confidence estimates. BMC Bioinformatics (In press)
[3] Asenjo AB, Krohn N, Sosa H. Configuration of the two kinesin motor domains during ATP hydrolysis. Nat Struct Biol. 2003 Oct;10(10):836-42. Epub 2003 Sep 14.
[4] Vale RD, Milligan RA. The way things move: looking under the hood of molecular motor proteins. Science. 2000 Apr 7;288(5463):88-95. Review.
[5] Gavin AC, Aloy P & al. Proteome survey reveals modularity of the yeast cell machinery. Nature. 2006 Mar 30;440(7084):631-6. Epub 2006 Jan 22.
[6] Pereira-Leal JB, Levy ED, Teichmann SA. The origins and evolution of functional modules: lessons from protein complexes. Philos Trans R Soc Lond B Biol Sci. 2006 Mar 29;361(1467):507-17. Review.
[7] Levy ED, Pereira-Leal JB, Chothia C, Teichmann SA. 3D complex: a structural classification of protein complexes. PLoS Comput Biol. 2006 Nov 17;2(11):e155
[8] Levy ED, Boeri-Erba E, Robinson CV, Teichmann SA. Assembly reflects evolution of protein complexes.Nature, Jun 26;453(7199):1262-5
[9] Levy ED A simple definition of structural regions in proteins and its use in analyzing interface evolution. J Mol Biol. 2010 Nov 5;403(4):660-70. Epub 2010 Sep 22.
[10]Levy ED*, Landry CR*, Michnick SW. Signaling through cooperation. Science. 2010 May 21;328(5981):983-4.
[11] Landry CR*, Levy ED*, Michnick SW. Weak functional constraints on phosphoproteomes. Trends Genet. 2009 Apr 4.
[12] Levy ED*, Landry CR*, Michnick SW. How perfect can protein interactomes be? Sci Signal. 2009 Mar 3;2(60):pe11.
[13] Tarassov K, Messier V, Landry CR, Radinovic S, Serna Molina MM, Shames I, Malitskaya Y, Vogel J, Bussey H, Michnick SW. An in vivo map of the yeast protein interactome. Science. 2008 Jun 13;320(5882):1465-70. Epub 2008 May 8.