Protein plasticity and evolution
Previous and Current Research
In order for the cell to function properly, proteins must be robust to both changes in their environment and errors made during their synthesis. At the same time, proteins also need to be able to evolve novel functions to survive on long evolutionary timescales. The very same processes, i.e. genetic and phenotypic mutations, generate the diversity that leads to functional innovations and broken proteins, ultimately resulting in novel organisms, diseases and in some cases extinction.
In short, proteins exhibit evolutionary plasticity. But how do proteins remain robust and facilitate innovations at the same time? How can we distinguish variations, genetic and phenotypic, that are good or bad?
In my previous research I explored genetic variations, with particular focus on epistasis between mutations. In my new lab, we focus on phenotypic variations that are orders of magnitude more frequent than genotypic mutations. These variations are generated by the stochastic noise inherent to biological systems, such as transcriptional errors, ribosomal slippage, conformational flexibility and noisy expression.
We aim to define and quantify the phenotypic plasticity of a protein, and to identify the compensatory mechanisms that buffer otherwise deleterious mutations. We wish to reveal the evolutionary potential of latent phenotypes to create novel functions and to influence gene-disease associations.
We mainly perform computational experiments using the vast amount of experimental data available (genome sequencing, RNA-Seq, mass-spectrometry proteomics), but we also test our hypotheses experimentally, generate our own data, and work closely with experimental collaborators and clinicians.
Future Projects and Goals
We study phenotypic plasticity of proteins, its mechanism, origin, evolutionary history, potential for innovations, and its role in diseases. We complement the increasing amount of high-throughput data with rigorous computational studies, and provide hypotheses for experimental validation.
Predicting phenotypic mutations
Transcriptional and translational errors generate diverse sequences, most dysfunctional, some harboring novel functions. We are developing an algorithm for predicting transcriptional and ribosomal slippage sites by training on known sites in transcriptomics and proteomics data and using evolutionary information. We aim at discovering novel frameshift and STOP codon read-through variants that we will validate by shotgun proteomics in collaboration with MPI-CBG Mass-spectrometry unit. We will explore the evolutionary potential of slippage sites and test whether they could lead to promiscuous functions.
Phenotypic variability: how do new functions emerge?
Evolution needs raw material for selection, for instance promiscuous function that is generated by phenotypic variability. We will continue exploring how new protein functions evolve by: i) functional characterization of the novel sequences generated by phenotypic mutations (predicted in the previous section); ii) studying disordered proteins, which can have different functions originating from different structural states. We previously found that co-evolutionary information can reveal the potential for functional structural states of disordered proteins that take on multiple conformations or fold only upon binding. We are further exploring this experimentally challenging class of proteins and predict their functional diversification and evolutionary origins using co-evolutionary and phylogenetic methods.
Quantifying protein plasticity
We plan to unify our knowledge about genetic, phenotypic and environmental noise and quantify phenotypic plasticity from a systems perspective. We explore associations and epistatic relationships between the different types of phenotypic mutations.
Predicting the fitness effects of mutations is a long-standing challenge. We will incorporate the phenotypic plasticity score to create an algorithm to predict the effects of mutations. We will provide an accessible database and tool for the biomedical community as well as seek collaboration with genetic diagnostics researchers. We hope to transform disease gene prioritization for Mendelian gene discovery, as more than 4,000 Mendelian phenotypes still lack an associated gene despite worldwide efforts.
Methodological and Technical Expertise
- Bioinformatics, algorithm development
- Computational experiments
- Next-generation sequencing data analysis
- Basic molecular biology: PCR, cloning, E. coli as a model system
Toth-Petroczy A*, Palmedo P*, John Ingraham JI, Thomas A. Hopf TA, Berger B, Sander C, Marks DS
Structured states of disordered proteins from genomic sequences.
Cell 2016, 167(1):158-70
Rockah-Shmuel L, Toth-Petroczy A, Sela A, Wurtzel O, Sorek R and Tawfik DS
The Occurrence and Bypass of Frame-shifting Insertion-Deletions (InDels) to give Functional Proteins are correlated.
PloS Genetics 2013 (10) e1003882
Dellus-Gur E*, Toth-Petroczy A*, Elias M, Tawfik DS
What makes a protein fold amenable to functional innovations? Fold polarity and stability trade-offs.
J Mol Biol 2013; 425(14):2609-21
Toth-Petroczy A and Tawfik DS
Protein insertions and deletions enabled by neutral roaming in sequence space.
Mol Biol Evol 2013; 30(4):761-71
Toth-Petroczy A and Tawfik DS
Slow protein evolutionary rates are dictated by surface-core association.
PNAS 2011, 108(27):11151-6
* equal contribution