Über die Autorin bzw. den Autor

Daniel L. Purich is the editor of Advances in Enzymology and Related Areas of Molecular Biology, Part A: Amino Acid Metabolism, Volume 72, published by Wiley.

Auszug. © Genehmigter Nachdruck. Alle Rechte vorbehalten.

Advances in Enzymology and Related Areas of Molecular Biology, Amino Acid Metabolism, Part a

Interscience Publishers

Chapter One

By STEVEN A. BENNER, SLIM O. SASSI, and ERIC A. GAUCHER, Foundation for Applied Molecular Evolution, 1115 NW 4th Street, Gainesville, FL 32601

CONTENTS

I. Introduction A. Role for History in Molecular Biology B. Evolutionary Analysis and the "Just So" Story C. Biomolecular Resurrections as a Way of Adding to an Evolutionary Narrative

II. Practicing Experimental Paleobiochemistry A. Building a Model for the Evolution of a Protein Family 1. Homology, Alignments, and Matrices 2. Trees and Outgroups 3. Correlating the Molecular and Paleontological Records B. Hierarchy of Models for Modeling Ancestral Protein Sequences 1. Assuming That the Historical Reality Arose from the Minimum Number of Amino Acid Replacements 2. Allowing the Possibility That the History Actually Had More Than the Minimum Number of Changes Required 3. Adding a Third Sequence 4. Relative Merits of Maximum Likelihood Versus Maximum Parsimony Methods for Inferring Ancestral Sequences C. Computational Methods D. How Not to Draw Inferences About Ancestral States III. Ambiguity in the Historical Models A. Sources of Ambiguity in the Reconstructions B. Managing Ambiguity 1. Hierarchical Models of Inference 2. Collecting More Sequences 3. Selecting Sites Considered to Be Important and Ignoring Ambiguity Elsewhere 4. Synthesizing Multiple Candidate Ancestral Proteins That Cover, or Sample, the Ambiguity C. Extent to Which Ambiguity Defeats the Paleogenetic Paradigm

IV. Examples A. Ribonucleases from Mammals: From Ecology to Medicine 1. Resurrecting Ancestral Ribonucleases from Artiodactyls 2. Understanding the Origin of Ruminant Digestion 3. Ribonuclease Homologs Involved in Unexpected Biological Activities 4. Paleobiochemistry with Eosinophil RNase Homologs 5. Paleobiochemistry with Ribonuclease Homologs in Bovine Seminal Fluid 6. Lessons Learned from Ribonuclease Resurrections B. Lysozymes: Testing Neutrality and Parallel Evolution C. Ancestral Transposable Elements 1. Long Interspersed Repetitive Elements of Type 1 2. Sleeping Beauty Transposon 3. Frog Prince 4. Biomedical Applications of Transposons D. Chymase-Angiotensin Converting Enzyme: Understanding Protease Specificity E. Resurrection of Regulatory Systems: The Pax System F. Visual Pigments 1. Rhodopsins from Archaeosaurs: An Ancestor of Modern Alligators and Birds 2. History of Short Wavelength-Sensitive Type 1 Visual Pigments 3. Green Opsin from Fish 4. Blue Opsins 5. Planetary Biology of the Opsins G. At What Temperature Did Early Bacteria Live? 1. Elongation Factors 2. Isopropylmalate and Isocitrate Dehydrogenases 3. Conclusions from "Deep Time" Paleogenetic Studies H. Alcohol Dehydrogenase: Changing Ecosystem in the Cretaceous I. Resurrecting the Ancestral Steroid Receptor and the Origin of Estrogen Signaling J. Ancestral Coral Fluorescent Proteins K. Isocitrate Dehydrogenase V. Global Lessons References

I. INTRODUCTION

A. ROLE FOR HISTORY IN MOLECULAR BIOLOGY

The structures that we find in living systems are the outcomes of random events. These are filtered through processes described by population dynamics and through natural selection to generate macroscopic, microscopic, and molecular physiology. The outcomes are, of course, constrained by physical and chemical law. Further, the outcome is limited by the Darwinian strategy by which natural selection superimposed on random variation searches for solutions to biological problems. The Darwinian strategy need not deliver the best response to an environmental challenge; indeed, it may deliver no response that allows a species to avoid extinction. The outcomes of evolutionary mechanisms therefore reflect history as much as optimization.

It is therefore not surprising that biology finds its roots in natural history. The classical fields in this classical discipline include systematic zoology, botany, paleontology, and planetary science. Here, seemingly trivial details (such as the physiology of the panda's thumb) have proven enlightening to naturalists as they attempt to understand the interplay of chance and necessity in determining the outcome of evolution (Gould, 1980; Glenner et al., 2004). These roots in natural history are not felt as strongly in modern molecular biology, however. Molecular biology emerged in the twentieth century as an alliance between biology and chemistry. The alliance has been enormously productive, but largely without reference to systematics, history, or evolution. Today, we have the chemical structures of millions of biomolecules and their complexes: as small as glucose and as large as the human genome (Venter et al., 2001). X-ray crystallography and nuclear magnetic resonance spectroscopy locate atoms within biomacromolecules with precisions of tenths of nanometers. Biophysical methods measure the time course of biological events on a microsecond scale (Buck and Rosen, 2001). These and other molecular characterizations, written in the language of chemistry, have supported industries such as drug design and foodstuff manufacture, all without any apparent need to make reference to the history of their molecular components or the evolutionary processes that generated them.

The success of this reductionist approach has caused many molecular biologists to place a lower priority on historical biology. Indeed, the archetypal molecular biologist has never studied systematics, paleontology, or Earth science. The combination of chemistry and biology has generated so much excitement that history seems no longer to be relevant, and certainly not necessary, to the practice of life science or the training of life scientists.

Nearly overlooked in the excitement, however, has been the failure of molecular characterization, even the most detailed, to generate something that might be called "understanding." The human genome provides an example of this. The genome is itself nothing more (and nothing less) than a collection of natural product structures. Each structure indicates how carbon, hydrogen, oxygen, nitrogen, and phosphorus atoms are bonded within a molecule that is special only in that it is directly inherited. It has long been known to natural product chemists that such biomolecular structures need not make statements about the function of the biomolecule described, either in its host organism or as the host organism interacts with its environment to survive and reproduce. This has poven to be true for genomic structures as well.

Genomic sequences do offer certain opportunities better than other natural product structures when it comes to understanding their function. Comparisons of the structures of genes and proteins can offer models for their histories better than comparisons of the structures of other natural products (Hesse, 2002). As was recognized nearly a half century ago by Pauling and Zuckerkandl (1963), a degree of similarity between two gene or protein sequences indicates, to a degree of certainty, that the two proteins share a common ancestor. Two homologous gene sequences may be aligned to indicate where a nucleotide in one gene shares common ancestry with a nucleotide in the other, both descending from a single nucleotide in an ancestral gene. An evolutionary tree can be built from an alignment of many sequences to show their familial relationships. The sequences of ancestral genes represented by points throughout the trees can be inferred, to a degree of certainty, from the sequences of the descendent sequences at the leaves of the tree.

The history that gene and protein sequences convey can then be used to understand their function. In its most general form, the strategy exploits the truism that any system, natural or human-made, from the QWERTY keyboard to the Federal Reserve banking system, can be better understood if one understands both its structure and its history.

Much understanding can come first by analyzing the sets of homologous sequences themselves. Thus, credible models for the folded structure of a protein can be predicted from a detailed analysis of the patterns of variation and conservation of amino acids within an evolutionary family (Benner et al., 1997a; Gerloff et al., 1999; Rost, 2001), if these are set within a model of the history of the family (Thornton and DeSalle, 2000). The quality of these predictions has been demonstrated through their application to protein structure prediction contests as well as through the use of predicted structures to detect distant protein homologs (Benner and Gerloff, 1991; Gerloff et al., 1997; Tauer and Benner, 1997; Dietmann and Holm, 2001). More recently, analysis of patterns of variation and conservation in genes is used to determine whether the gross function of a protein is changing and which amino acids are involved in the change (Gaucher et al., 2001, 2002; Bielawski and Yang, 2004).

Computer analysis of protein sequences from an evolutionary perspective has emerged as a major activity in the past decade. Here, sets of protein sequences are studied computationally within the context of an evolutionary model in an effort to better connect evolving sequences with changing function. Our purpose is to review strategies that go beyond simple computational manipulation of gene and protein sequences. In this review we explore experiment as a way to exploit the history captured within the chemical structures of DNA and protein molecules.

Our focus will be the emerging field known variously as experimental paleogenetics, paleobiochemistry, paleomolecular biology, and paleosystems biology. Practitioners of the field resurrect ancient biomolecular systems from now-extinct organisms for study in the laboratory. The field was started 20 years ago (Nambiar et al., 1984; Presnell and Benner, 1988; Stackhouse et al., 1990) for the specific purpose of joining information from natural history, itself undergoing a surge of activity, to the chemical characterization of biomolecules, with the multiple intents of helping molecular biologists select interesting research problems, generating hypotheses and models to understand the molecular features of biomolecular systems, and providing a way of experimentally testing historical models.

The field has now explored approximately a dozen biomolecular systems (Table 1). These include digestive proteins (ribonucleases, proteases, and lysozymes) in ruminants to illustrate how digestive function arose from nondigestive function in response to a changing global ecosystem, fermentive enzymes from fungi to illustrate how molecular adaptation supported mammals as they displaced dinosaurs as the dominant large land animals, pigments in the visual system adapting to function optimally in different environments, steroid hormone receptors adapting to changing function in steroid-based regulation of metazoans, and proteins from very ancient bacteria, helping to define environments where the earliest forms of bacterial life lived.

To date, approximately 20 narratives have emerged where specific molecular systems from extinct organisms been resurrected for study in the laboratory. In general, understanding delivered by experimental paleogenetics was not accessible in other ways. After a brief introduction of the strategies and problems in experimental paleoscience, we review each of these narratives.

Our goal here is also to strengthen awareness among molecular and biomedical scientists of the ability of experimental paleogenetics to place meaning on biological data. We believe that current efforts falling under the rubric of "systems biology" can be complemented and strengthened when combined with a historical prespective. Understanding will not, we suspect, arise from still more, and still more quantitative, molecular, chemical, and geometric characterization of cellular, organ, and organism-defined systems. At the very least, the analysis must go further, to include the organism, the ecosystem, and the physical environment, which extends from the local habitat to the planet and the cosmos (Feder and Mitchell-Olds, 2003). Without an understanding of the history, we expect that efforts in reductive systems biology are likely to fall short of their promise to deliver understanding.

The same applies to the broader scientific community. Reviewing the first paleomolecular resurrections (Stackhouse et al., 1990; Jermann et al., 1995) a decade ago, Nicholas Wade, writing in the New York Times Magazine (Wade, 1995), expressed displeasure. "The stirring of ancient artiodactyl ribonucleases," he wrote, "is a foretaste of biology's demiurgic powers." He then suggested that biomolecular "resurrection [remain] an unroutine event." Given the absence of hazard presented by paleomolecular resurrections (pace Jurassic Park), it seems unwise to forgo understanding that comes from bringing into the laboratory for study biomolecules from the past. As the size of the genome sequence database grows, and the gap between chemical data compilations and biological understanding increases, we suspect that experimental paleogenetics will be the key tool to bridge the gap.

B. EVOLUTIONARY ANALYSIS AND THE "JUST SO" STORY

After a half century of reductionist biology, evolutionary analysis must struggle to enter the mainstream of discussion within the molecular sciences. As Adey noted a decade ago, many scientists view evolutionary hypotheses as being inherently resistant to experimental test and therefore fundamentally nonscientific (Adey et al., 1994). Certainly, neither the sequence nor the behavior of a protein from an organism that went extinct a billion years ago can be known with the same precision as the sequence or behavior of a descendent living today. But much in science is useful even though it is not known with certainty. Our job in paleogenetics, as with all science, is to manage whatever uncertainty we encounter.

The arcane nature of the most prominent debate in molecular evolution has also led many molecular biologists to shun evolutionary discussions. Many biomedical researchers do not understand how their view of biology might be different if it turns out that dog diverged from humans and rodents before or after humans and rodents themselves diverged. Similarly, it has always been unproductive in chemistry to ask whether alterations in chemical structure generally have any specific impact on behavior; productive discussions in chemistry focus on specific chemical structures and specific changes. Yet the neutralist-selectionist debate that consumed molecular evolution for nearly a decade asked precisely such questions (Hey, 1999).

Further, the accidents that shape molecular biology cannot be reproduced in the laboratory; they may, in fact, never be known in detail. Complex biological systems are chaotic. Small differences in input can have large impacts on the output. The notion that biology would be rather different had a fly flapped his wings differently in the Triassic is a compelling reason to marginalize historical narratives (Lorenz, 1963, 1969).

Further difficulties are encountered with historical narratives, even when they do explain a biological fact. Here, they can easily be viewed as "just so" stories. This epithet is pejorative; it indicates that the narrative is constructed ad hoc to explain a specific fact (how the zebra got his stripes), makes no reference to facts verifiable outside the fact being explained (we have no way to verify that an ancestral zebra took a nap under a ladder), and could easily be replaced by a different story, just as compelling, explaining a different observation (if the modern zebra had spots, the story might be that the ancestral zebra took a nap beneath a philodendron such as Monstera friedrichsthalii).

Continues...

Excerpted from Advances in Enzymology and Related Areas of Molecular Biology, Amino Acid Metabolism, Part aby Daniel L. Purich Copyright © 1998 by Daniel L. Purich. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.