Protein Evolution

Andrei Lupas

andrei lupas
  • PhD in Molecular Biology at Princeton University, 1985-91
  • Postdoctoral training at the Gene Center of the University of Munich and at the MPI for Biochemistry, Martinsried, 1993-97
  • Senior Computational Biologist and Assistant Director of Bioinformatics at SmithKline Beecham Pharmaceuticals, 1997-2001
  • Director at the MPI since 2001

Research Interest

Since proteins are essential components of living cells, it is hardly surprising that precursors of most proteins observed today existed at the time of the last common ancestor of all life. But how did proteins evolve? Randomly synthesized polypeptide chains form folded structures in less than one in a billion cases, so it seems impossible that proteins evolved by chance. Our hypothesis is that folded proteins evolved by fusion and accretion from an ancestral set of peptides active as cofactors of RNA-dependent replication and catalysis. Using bioinformatics, we have reconstructed this "vocabulary" of ancient peptides and are now exploring experimentally the processes by which it could have led to the emergence of folded proteins, using methods in biochemistry and structural biology (crystallography and NMR).

We also study how changes in protein structure create new functionality, both by attempting to functionalize newly created proteins in vitro and by exploring structure-function relationships in natural proteins. In the latter, our particular focus is on how type I receptors transduce signals across membranes and AAA ATPases disassemble, unfold and translocate proteins. An essential aspect of our work is the development of new bioinformatic tools, which we deploy in our MPI Bioinformatics Toolkit (

  • ancient vocabulary no labels no red
    click to enlarge

Figure 1: A "vocabulary" of ancient peptides, reconstructed by sequence and structure comparisons of modern proteins. Peptides that interact with nucleic acids are highlighted yellow, with nucleotides blue, and with metals rose. Peptides that form folds by repetition are boxed.

  • cogwheel model
    click to enlarge

Figure 2: Cogwheel model for signal transduction across membranes by axial helix rotation. (A) Structure of a signal transduction domain adjacent to the membrane. (B) Schematic and actual structural transitions in the helices (top view). (C) Model of the transduction process.

Available PhD Projects

Development of software tools for protein sequence analysis and structure prediction.

Selected Reading

1) Alva V, Nam SZ, Söding J, Lupas AN. (2016) The MPI bioinformatics Toolkit as an integrative platform for advanced protein sequence and structure analysis. Nucleic Acids Res. 44(Web Server Issue):W410-5.

2) Alva V, Söding J, Lupas AN. (2015) A vocabulary of ancient peptides at the origin of folded proteins. eLife 4:e09410.

3) Ferris HU, Dunin-Horkawicz S, Hornig N, Hulko M, Martin J, Schultz JE, Zeth K, Lupas AN, Coles M. (2012) Mechanism of regulation of receptor histidine kinases. Structure. 20:56-66.