Emmanuel David Tannenbaum

 

How to Contact Me?

Short Biography

I came to Harvard in September, 1998 after completing my undergraduate studies at the University of Minnesota , where I majored in Chemical Engineering and in Mathematics. What attracted me to Harvard was its Chemical Physics program. I quickly joined Eric Heller's group , which sits in both the Physics and Chemistry departments.

The focus of my graduate research was in semiclassical quantum mechanics . I also pursued an experimental collaboration with William Klemperer and his group, modeling the vibrational predissociation of the van der Waals complex ArHF. I defended my Ph.D. thesis in May, 2002, and, after a brief stint in the Israeli army , I returned to Harvard in March, 2003 to start a postdoc in mathematical biology with Eugene Shakhnovich.

Feel free to download a copy of my CV .

RESEARCH OVERVIEW

Recent Work

My postdoctoral research has been focused on a class of evolutionary dynamics models known as the quasispecies equations . These were introduced by Manfred Eigen in 1971 as a way to model the evolutionary dynamics of single-stranded RNA molecules in in vitro evolution experiments. The quasispecies equations have emerged as a powerful formalism for modeling evolutionary dynamics in a variety of contexts. While the major focus of quasispecies research has been on viral evolution, the model has also been a useful tool for understanding bacterial evolution, the emergence of cancer, and immune response.

The central result of the quasispecies model is a localization to delocalization transition over the genome space termed the error catastrophe . Briefly, at low mutation rates, the replicative selection of viable genomes is sufficiently strong so that the organisms' genotypes cluster about one or a few viable genomes. Thus, the population does not consist of a single genotype corresponding to the fastest replicator in the population, but is rather a "clan" of related individuals (i.e. a "quasispecies."). Beyond some critical mutation rate, replicative selection is no longer sufficiently strong to localize the population about the viable genotypes, leading to delocalization over the entire genome space. In this regime, the fraction of viable organisms becomes negligible, and no discernible quasispecies is present. In short, natural selection can no longer overcome the high mutation rates and produce a viable population of organisms.

The error catastrophe has been observed experimentally, and has been shown to be the basis for a number of anti-viral treatments (Ribavirin against Hepatitis-C, and 5-hydroxydeoxycytidine against HIV).

My work in quasispecies theory has included the following:

(1) Incorporation of Repair : This research was motivated by a desire to understand the role that mismatch-repair deficient strains, or mutators , play in the emergence of antibiotic drug resistance and cancer in multicellular organisms. We developed a model within the quasispecies framework that incorporated genetic repair, and computed the equilibrium fraction of mutators as a function of mutation rate and repair efficiency. We showed that, in addition to the error catastrophe, there exists a second localization to delocalization transition, the "repair catastrophe", over the repair genome subspace.

(2) Extension to Semiconservative Replication : We developed a formulation of the quasispecies model appropriate for the semiconservative nature of DNA replication. We regard this work as a necessary first step toward making the quasispecies model a useful tool for analyzing the evolutionary dynamics of DNA-based genomes. In contrast to the conservative mode of replication assumed in previous quasispecies models, the error threshold has a finite upper bound for semiconservative replication. This implies that semiconservatively replicating genomes are considerably less robust to mutations than conservatively replicating genomes. In particular, it is not possible to "out-replicate" the error catastrophe with semiconservative replication.

(3) Solution of the Model for Genomes Consisting of Arbitrary Numbers of Genes : This paper considered the quasispecies dynamics over fitness landscapes more complicated than the Single Fitness Peak landscape. Specifically, we looked at genomes with arbitrary numbers of genes, whose fitness was determined by which genes in the genome were functional, and which were not. The central result of this paper was that, instead of a single error catastrophe, it is possible to have a series of localization to delocalization transitions, which we called an "error cascade." We don't assume any particular form for the fitness function, so that our model can capture fairly nontrivial behavior (such as relocalization) which arises due to correlations amongst the genes. The delocalization pattern can therefore be used to extract correlations amongst different genes in a genome, and determine the contributions that the various genes make to the fitness of the organism.

(4) Imperfect Lesion Repair : Following recent work by Yisroel Brumer and Eugene I. Shakhnovich (Y. Brumer and E.I. Shakhnovich, "The Importance of DNA Repair in Tumor Suppression," Phys. Rev. E 70, 061912 (2004)), we developed an ordered strand pair formulation of the quasispecies model appropriate for semiconservative replication with imperfect lesion repair. In our original derivation of the semiconservative quasispecies equations, we considered a dynamics over the space of complementary strand pairs. This requires that any mismatches remaining after cell division (which were either not repaired or erroneously repaired by mechanisms such as mismatch repair) are repaired. However, because after cell division is complete parent/daughter strand discrimination is lost, a post-replication mismatch is correctly repaired with a 50% probability. Imperfect lesion repair is an extension of the semiconservative model which does not assume that all post-replication mismatches are repaired.

(5) Stem Cell Division and Tissue Renewal : We applied the ideas from our work on imperfect lesion repair to develop a point-mutation model describing the evolutionary dynamics of a population of adult stem cells. Because of their connection to tissue aging and the emergence of cancer, adult stem cells are currently a topic of intense investigation. The model we developed incorporated a long-standing, and recently confirmed hypothesis regarding the nature of adult stem cell division. This immortal strand hypothesis states that when an adult stem cell divides to produce a stem cell and a differentiating tissue cell, the stem cell retains the chromosomes containing the oldest DNA strands of the genome. Presumably, the oldest DNA strands are the most accurate templates for daughter strand synthesis, and therefore their retention in the stem cells greatly slows the rate of accumulation of mutations in the stem cell population. Our theoretical model established that lesion repair should be completely suppressed in order to optimally preserve the genetic integrity of the stem cell population over the course of a human lifetime. This result is in qualitative, and possibly quantitative, agreement with experiment.

Future Directions

There is evidence which suggests that selection is an underlying principle driving the self-organization of a variety of systems. Examples include pathway selection in the brain, the structure of RNA biochemical networks, and of course, the self-organization of humans into networked societies. This implies that quasispecies theory, hypercycles, and other tools from evolutionary dynamics could be useful for understanding a variety of phenomena, such as the emergence of addiction, learning, and biological regulatory networks.

Thus, while my current research plans are focused on quasispecies theory, for future work I would like to move into other areas for which selection models will be relevant. In this vein, I have recently become interested in agent-based modeling methods. I also recently submitted a paper, titled "An RNA-centered view of eukaryotic cells," in which I speculate that eukaryotic cells should be regarded in many ways as RNA-"communities." In the paper, I describe in detail what I mean by such terminology.

Publications

  1. E. Tannenbaum, "An RNA-centered view of eukaryotic cells," q-bio.MN/0412027 , submitted to Proc. Roy. Soc. London B ( PDF ).
  2. E. Tannenbaum, J.L. Sherley, and E.I. Shakhnovich, "Evolutionary dynamics of adult stem cells: Comparison of random and immortal strand segregation mechanisms," Physical Review E 71 , 041914 (2005) ( PDF ).
  3. E. Tannenbaum, J.L. Sherley, and E.I. Shakhnovich, "Imperfect DNA Lesion Repair in the Semiconservative Quasispecies Model: Derivation of the Hamming Class Equations and Solution of the Single-Fitness Peak Landscape," Physical Review E 70 , 061915 (2004) ( PDF ).
  4. S. Kallush, E. Tannenbaum, and B. Segev, "Local Group Velocity and Path-Delay: Semi-Classical Propagators for the Time Evolution of Wigner Functions in Deep Tunneling and in Dispersive Media," Chemical Physics Letters 396 , 261 (2004) ( PDF ).
  5. E. Tannenbaum and E.I. Shakhnovich, ``Solution of the Quasispecies Model
    for an Arbitrary Gene Network,'' Physical Review E 70 , 021903 (2004) ( PDF ).
  6. E. Tannenbaum, E.J. Deeds, and E.I. Shakhnovich, ``Semiconservative
    Replication in the Quasispecies Model,'' Physical Review E 69 , 061916 (2004) ( PDF ).
  7. E. Tannenbaum and E.I. Shakhnovich, ``Error and Repair Catastrophes:
    A Two-Dimensional Phase Diagram in the Quasispecies Model,'' Physical Review E 69 , 011902 (2004) ( PDF ).
  8. E. Hershkovitz, E. Tannenbaum, S.B. Howerton, A. Sheth, A. Tannenbaum,
    and L.D. Williams, ``Automated Identification of RNA Conformational Motifs:
    Theory and Application to the HM LSU 23S rRNA,'' Nucleic Acids Research 31 , 6249 (2004) ( PDF ).
  9. E. Tannenbaum, E.J. Deeds, and E.I. Shakhnovich, ``Equilibrium
    Distribution of Mutators in the Single Fitness Peak Model,''
    Physical Review Letters 91 , 138105 (2003) ( PDF ).
  10. E. Tannenbaum and E.J. Heller, ``Determination of Bound-Free
    Dissociative Couplings Via Classical Fourier Coefficients,''
    The Journal of Chemical Physics 117 , 9574 (2002) ( PDF ).
  11. E. Tannenbaum, K.J. Higgins, W. Klemperer, B. Segev, and E.J. Heller,
    ``A Perturbative Approach to Vibrational Predissociation Rates: Application
    to ArHF,'' The Journal of Physical Chemistry B 106 , 8100 (2002) ( PDF ).
  12. E. Tannenbaum, ``Partial-Differential-Equation-Based Approach to
    Classical Phase-Space Deformations,'' Physical Review E 65 , 066613
    (2002) ( PDF ).
  13. E. Tannenbaum and E.J. Heller, ``Semiclassical Quantization Using
    Invariant Tori: A Gradient-Descent Approach,'' The Journal of Physical
    Chemistry A     105 , 2803 (2001) ( PDF ).
  14. Y. Kannai and E. Tannenbaum, ``Paths Leading to the Nash Set for
    Non-Smooth Games,'' The International Journal of Game Theory 27 ,
    393 (1998) ( PDF ).