ENERGY AND THE Fee Bie RONALD F, FOXENERGY AND THE EVOLU TION OF LIFE RONALD F. FOX ‘School of Phystes ieorgia Injtttute of Techs W.H, FREEMAN AND COMPANY NEW YORELibracy of Congress Catalaging-in-Publicution Dara Fox, Ronald F. (Ronald Eorcest), 143— ‘Energy and the evolutian af life | Ranald F. Fox Pe cm. nctudes index. 1849-9. ISBN 0-7167-1870-7 (pbk.} 2. Bioenergetics. {, Title. QH37LFG2 1988 87-21958 575 —de19 CIP Copyright © 1988 by W. H. Freemun and Company. No part of this hook may be reproduced by any mechanical, photographic, or electronic process, or in the form of a phonographic recording, nor may it be stored in a retrieval system, transmitted, or otherwise copied for ‘publie o privare use, without the writen permission of the publisher, Printed in the United States of America 1234367890 VB 6543210899 To the memory of Mark Kac and Fritz LipmannIntroduction RL Energy and Matter: Before Life of the Elements during Stellar Evolucion ition of Small Molecules from the Elements onomers—The Buiding Blocks of Living Systems ixed Oligomers—Small Chains of Various Monomers lymers—The Fabric of Living Systems The Uroboros Puzzle 13 21 26 33 35vii | CONTENTS CHAPTER 2 The Primitive Uroboros 2-1, Energy Driven Polymerization 2-2, Energy Metabolism and Polymer Synthesis in Contemporary Organisms 2-3, Initiation of an Uroboros 2.4, Thermodynamic Analysis of the Uroboros CHAPTER 3 Self-Assembly and Control 3-1. Self-Assembly 3-2. Molecular Mospkepeneal and Control Mechanisms 3-3. Phosphate Mechanisms in Organisms 3-4, The Evolution of Energy Metabolism and Storage cuarter 4 Nonlinear Dynamics 4-1, The Physics and Mathematics of Energy-Driven Sysrems 4-2. Reaction-Diffusion Equations for Morphogenesis 4-3, Dimensionality and Instability 4-4, Temporal Dependence in a Driven System: A Turing Model in Time -S. Atcractors in Phase-Space Pictures -6. Closed-Form Solutions and Structural Stability 7. The Transition t Chaos: Experimental Chaos +8, Feigenbaum Universality 9 The J. Maynard Smith Model: Imeger Version 0. Complexity, Cellular Automata, Darwinian Evolution, and the Beain ScEaae 4. CHAPTER 5 Biological Predictabilicy 5-1, What is Evolueion? 3-2, Driven Dissipative Syscems 5-3. The Evolutionary Impact of Simulation 5-4. Nervous Systems Glossary Index 39 40 49 56 74 7 ao 83 8o o4 103 104 LO4 107 WL 114 di? 120 131 135. 142 151 132 134 155 156, 169 v7 os Ma} > Sig = oe T first began thinking about the role of biological evolution roughly thirty years ago when I was a secondary | student. | would visit my father's laboratory, where 1 was allowed to 5 proteinoid microspheres in a test-tube and then watch them under a . I found grear fascination in mixing salts, sugars, and lipids and ying the remarkable self-assembled structures that formed. Most capti- all, however, was their Brownian motion. The microspheres jigeled ly a microsphere divided, and I would watch the twox / PREFACE offspring move away from each other. However, my aptitudes were strongest in the mathematical sciences, so in college 1 studied mathematical physics, with the intention of one day returning to biological questions During my college years (1960-1964), the genetic code was elucidated, and soon thereafter protein biosynthesis was described in great detail. | was nor unaware of these developments. because, even though the study of mathe. matical physics put heavy demands on my time, I kept up wich molecular biology. These events greatly influenced my decision co pursue graduate stud- ies ac the Rockefeller University (1965-1969), where I studied nonequili- brium thermodynamics with George Uhlenbeck and Mark Kac. I also had the good forcune to learn biochemistry and bioenergetics from Fritz Lipmann and Christian de Duve. Ac that time, de Duve had the remarkable foresight to realize that the idea of chemiosmosis deserved a serious hearing and he or- ganized a seminar course on chemiosmosis, which I attended. He invited Peter Mitchell and many of his critics (most of whom are now quite converted) to lecture, and | began to realize that a revolution in bioenergetics, no less sige nificant than the revolution in molecular biology, was about zo take place. By the mid-1970s this revolution was oearly complete, and today, while many details of bioenergetics require more work (alsa true in molecular biology), the essential outline is clear. During the 1970s | pursued an academic career in nonequilibrium statistical mechanics and stochastic processes, I was still looking for chances to connect my studies with biology. While I was a postdoctoral fellow, Uhlenbeck ar- ranged for me to spend a couple of weeks with Max Delbruck to help make this connection. Delbruck was very sympathetic bur tried to convince me that what | should do was simply to become a biologist, as he had done so suc- cessfully. To him, physics and biology were concerned with very different issues and requited very different methods. Ar about the same time, | had. occasion ta visit with Lars Onsager, whom I had met as a student at Rockefeller University when he visited Uhlenbeck. Onsager had retired co Miami, where he came to know my father, and revived his earlier interest in questions of the ongins of life. once asked him whether he thought thar statistical me- chanics would have a significant impact on the study of biological questions. He replied with a typically enigmatic and terse: “No.” Ironically, the man who encouraged me the most, when | visited Kae and Uhlenbeck at Rockefeller University, was Frite Lipmann, After all, he had played the major part in developing our understanding of che role of energy in biology and, while he emphasized the importance of appreciating biochemistry as chemistry, he also understood that energy flow was a physical, rather than a chemical phenom- enon with a strongly thermodynamic flavor. Through Lipmann I learned about the unique biological importance of phosphorous compounds and their sig- nificance with regard to energy flow. The 1970s witnessed yet another scientific revolution, this time in mathe- matical physies, where problems of nonlinear dynamics were being solved and attracting considerable attention. My colleague, Joe Ford, a pioneer in this field during the early 1960s, inviced many scientists—among them Boris Chi- PREFACE / xi ‘and Mitchell Feigenbaum—ro our department to discuss these devel- tents, Atfirse | did not realize what was happening and was mildly amused ear these people talking about clasrical srcifietors, about which 1 thought “everything! As @ postdoctorél fellow at Berkeley, | had expressed my jay at such talk co Karen Ublenbeck, who patiently attempted to tell me rese oscillators weree't harmonic but nonlinear, and that that made all difference. Ir took me nearly a decade for these ideas ca sink in, so char -until the beginning of the 1980s chat I recognized the point of the ntific revolution: that deterministic dynamical systems could display behavior. Even more important for me was the realization chat well- n, The advent of readily available rapid computing coincided with valution, 984, the John Simon Guggenheim Foundation generously awarded me lowship to study the physical basis for biological evolution, This award been stipported by Terrell Hill, Mark Kac, Frit. Lipmansy and James evenson, to whom | am most grateful. The grant was gencrously matched the Georgia Tech Foundation, so that I was able to devore the 1984-1985 lemic year exclusively tc these studies. My objective was to bring together onsequences of the several revolutions mentioned above. | approached project as an experiment in simultaneous writing and research. The ex- 1 also especially grareful to a few close friends who were willing to review eatly version of the manuscript and provide me with their criticisms. They e Larry Gold, Joel Keizer, and Roger Warrell. 1 benefitted greatly ftom the torial efforts of Barbara Brooks, senior editor for W'. H. Freeman and ‘Company, and Martin Silberberg of Manuscripts Associates At saddens me to reflect that Mark Kac (1914-1985) and Prite Lipmann 1900-1986) are no longer able to read my wark. | would have loved to have Feceived their comments about this book. Kac served so well as mentor, guide, friend over the last twenty years, Lipmann encouraged my exploration e origins of life at sevetal crucial steps. It is 10 the memory of Kac and ana that I dedicate this book, dB. FoxENERGY AND THE EVOLUTION OF LIFEAtay book abour the role of energy in be an interdisciplinary work, concerned with biology, chem- s, and physics: The biological and chemical aspect of the book 3, and 5) is descriptive, emphasizing specific substances and reas the mathematical and physical aspect (Chapters 1, 4, and ptual, emphasizing general principles. A glossary of specialized “appears at the end of the book.2) INTRODUCTION Energy Flow in Biological Systems Biological organization and its evolution are consequences of the flow of energy through mater. Frice Lipmann's recognition (19411 of the universal energy currency Of phosphates in biological systems (see Figure I-1) was es. sential to the subsequent revolution in molecular biology, a revolution thar elucidated the processes of the biosynthesis of proteins and of the genes that serve as templates for them. Scientists mapped metabolic pathways (for ex- ample, glycolysis, the tricarboxylic acid cycle, and the electron transport chain) and realized that these pathways are driven by the flow, use, and transduction of energy. This book presents these ideas at an introductory level with em- phasis on energy flow as the central principle. The revolution in molecular biology preceded by only a decade a related revolution in bioenergetics, in which scientists discovered that the process of chemivimesii provides the link berween electron transport energy and phosphate bond energy (Harold, 1986), This discovery enabled biochemisrs to describe reasonably well the mecha- nisms by which cells transduce, use, and store enerey, Mathematical Properties of Driven Dynamic Systems The recognition thar biological evolution is a manifestation of energy-driven organic marrer has focused the attention of biologists on recent revolutionary Adenylic acid Food ane! Fuel Creuine ~P Inorganic P FIGURE Lt The Lipmana cycte. The meeabolic dynamo generates ~P current, This is brushed off by adenylic acid, which likewise functions as the wiring syscem, distributing ehe current Creatine~P, when present served as P accumulator. Camponenis of the metabolic wheel include glycolysis, the: citti¢ ucid cycle, and che electron. transport INTRODUCTION J 3 hematicians aad physicists on understanding che dynamics of onlinear systems. The great mathematician, Jules-Henri Poincaré 2) had gained insights into such systems long before there were However, the systems are best studied by compuer simulation the equations used rarely, possess closed-form solutions. By using irs co apply Poincaré’s ideas, researchers have discovered many new ‘en nonlinea* systems. The concern of this book is with those ime of which are remarkably simple—that are relevant ta the non- aamics of evolving organism: ion of nonlinear dynamics includes an account of the relevance of very simple structures; ir also explores the fact chat the cleterministic “used to describe driven nonlinear systems can give rise to descrip- f chaotic behavior. ‘These behaviors are closely relaced to the nanex- ff closed-form solutions for these equations and to the need for simu- iques to obtain information about their dynamics. These ‘have profound significance for the problem of predictability in - systems, which Chapter 5 relates to the process of biological es on Energy in Biology k presents three perspectives on the role of energy in biology, bue quite cistinee. Chapter 2 gives the first perspective, which sIf-generating structures of biological energy-teansducing mech- ow energy transduction could have happened before any such existed, This is the uroboros puzzle. Chapter 3 considers how ev- sulted from sophistications in energy transduction and is concerned roles of energy regulation (the second perspective) and storage (the tive), especially in certain special energy storage molecules, the Chapters 4 and 5 explore the emergent properties of phos- at biodynamics, in particular the excitable tissues, the muscles heir dependence on phosphagens, and their direct connection of nonlinea: dynamics. The three perspectives of the role of iology cover a range of questions, from che origin of unicellular lution of multicellular organisms and their social structures. The Overviews chat express che conclusions reached from each perspec- author's hypotheses. Throughout the book, putative facts are tiated from explicit hypatheses. tive: The Uroboros Puzzle What is the fundamental problem y? Max Delbruck answers as follows: “Thus there is a clear case for on on Earth from no-life to life. How this happened is a funda- ‘haps the fundariental question of biology.” [Delbruck, 1978, p. Foblem of the initiation of life is beautifully embodied in the ancient 1 urobores. The uroboros, symbolized by a serpent with irs tail in4) INTRODUCTION FIGURE 1-2 ‘The urobaros. Lisually the uroboros is symbolized by ane serperie rather than two. This version has a more direct molecular biology incerpretatian in terms of proteins and polynucleorides. (From Eleazar’s Ural Chymisches Werk, 1260.) The alchemical uses that grew out of the idea of uroboros were related co Greek ideas about its manifestations in the seal world, One of these manifestations was gold, which was equal £0 fire an the scale of Elements: It scemed ro be the only substance that could resist chemical change by fire, and it is yellow like fire and the sun. The sum symbolized gold in ancient literature, and for bor she Greeks and the Egyptians, gald color, the spirit of yold was the ideal of perfection. This spiricual essence was endowed with the power 60 reproduce itself and to transmure all metals into gold. To Plato gold was the ultimate state of inanimare matter, TO his pupil, Aristotle, organic narure, also seeking perfection, progresses through a series of cycles in which the wlimate goal is self-reproduction. This view is the strangest analogue to modern molecular biology: The sun, the symbol of self-begetting gold, is che ultimate source of energy for earth's self-begetting systems. INTRODUCTION (ee Figure 1-2), represents an entity thac is self-xencrating and self- is a self-generating and self-sustaining system that has evolved of being in which its origins are ao longer discernable. Organisms jn energy-transducing structures for obtaining the energy to make reins needed for energy transduction, The uroboros puzzle of the tran- ff noslife co life is: how could this have begun? id Third Perspectives: Evolutionary Trends The uroboros puzzle fundamental biological problem of the transition from no-life co flow supports rie existence of life’s dynamic molecular stare and to evolutionary trends that lead t more complex forms of life, ily multicellular forms that have muscle and nervous tissues. Regulatory isms (perspective two) and storage schemes (perspective chree) involve arures of energy flow. The solution to the uroboros puzzle pinpoines »bond energy as cie special cype of energy used by cells. Phosphate Sl. Phosphate along with calcium, another ‘substance, provide che substrate for bone, a steuctural and procec- | thar has coevolved with excitable tissues, bruck (1977) observed chat it is difficult to reconstruct the “tree cord, Although comparative anatomy, physiology, and biochemisury h more useful chan fossils for such a reconstruction, he concludes char Evolutionary Significance of Excitable Tissues phagens enable organisms to accumulate stores of energy for use during if high energy necd. Early in the history of life, this storage mechanism Simple regulatory device. But it evolved and made the evolution of le tissues: possible. Their rapid use of energy could noc be sustained ut phosphagens. Scle provides an organism with movement, not only of che various parts ody but also of its ent re body through its environment, Such movement y intensifies organism-environment and organism-organism interactions, : usually nonlinear in character. The survival of an individual organism cies depends on che outcome of these interactions. Os tissue provides organisms with the ability to control thetr move- ad'to influence the outcomes of biological interactions. The abi 56 / INTRODUCTION Of nonlinear events is a special cype of nvnliaear dynamic problem. Chapter 4 shows that the optimal approach to this problem is rapid simulation. Thar the nervous system has evolved to bea rapid simulator is the subject of Chapter 5. A list of references follows cach chapter, Three references especially useful for their beautiful illustrations and detailed treatments of molecular and cele lular biology are Lubert Steyer's Biochemistry (1988), Christian de Duve's A Guided Tour of the Living Cell (1984), and Franklin M. Harold's The Vital Force (1986). The reader will be reminded throughout this book to consule these three titles. CHAPTER REFERENCES Berthelot, M., Les originey de falehimie, Georges Steinheil, Paris, 1885 de Duye, C., A Guided Tour of the Living Cell, 2 vols., Scientific American Books, New York, 1984. Delbruck, M., “Mind from Marcer?,” in XU lth Nobel Conference, Gustavus Adolphus. Callene, Oct. (1977), Nawwre uf Life, Edited by William H. Herdcamp, Universit Park Press, Baltimore, 1978. Harold, F. M., The Visad Force: A Study of Bioenerseticn, W. H, Freeman, New York 1986, Hopkins, A. J., Alchemy: Child of Greet Philosophy, AMS Press, New York, 1967, Lipmann, F, “Metabolic Generation and Utilization of Phosphate Bond Energy, Advaueer in Enzymatogy | (1941): 100— 162 % N., Chinete Alchemy: Preliminary Stedier, Harvard University Press, Camb Mass.. 1968. Stryer, L., Brochem#siry. 3d od. W. H. Freeman, New York, 1988. Sivi Energy was present in abundance che verse began. Matter came later, forming from energy as the sratures and pressures of the beginning grew less, This chapter ws of energy and the changes in its form that resulted in con- ‘ate for life. Atoms, molecules, and a suitable surface had co fe could begin. Ps are clear. After stars form chemical ¢lements in a process ucleosynthesis, which involves various energy flaws (Section 1-RGY AND MATTER BE LIFE / 9 8 / CHAPTER I 1), the chemical elements combine into small molecules on the relatively low temperature surface of che earch, with help from narural energy sources (See. tion 1-2), Among these small molecules are the monomers—amino acids, sugars, and es—which when linked together form the larger molecules essential to life (Sections 1-3 and 1-4), Nor so clear is how to form these larger molecules essential for life on the prebiotic carth, since they tend co break up as fast as they form (Section 1. 5), Thus the uroboros puzzle arises: How do you begin to make polymers, when polymers themselves ace needed to make polymers (Section 1-6)? (p) a) denotes the deureron, a nucleus of mass 2 (che proton J in producing @ deuteron, the reaction causes one of the protons beta (plus) decay into a neutron, positron, and neutrino. (Only newly formed star, hot protons surround the deuterons, a con- ds to another rapid nuclear reaction, (p,n) + p— (2p, nm) + ¥. (Reaction 2) 1es a photon, and (2p, n) denotes che nucleus of helium 3 (Hes, at combine with protons. Instead, the protons elastically scatter sach othet) without a net change in their combined energy. Deu- forming *He, are now coo rare to combine appreciably with wently, the amount of *He increases until hor "He collisions take I-l, GRIGIN OF THE ELEMENTS DURING STELLAR EVOLUTION ‘The subject of this book begins somewhar after the beginning of the universe, wheat the scars functioned as factories of chemical clements. The mechanism. by which che chemical elements come intw existence is stellar mucteorynthesis,, The processes involved are an example af how exergy flow produces complex stares of matcer from simple constituents: A combination of gravitational en ergy and nuclear energy converts vast quantities of hydrogen gas, the simplest element, into the nuclei of ocher, more complex, elements. Nucleosynthesi involves nuclear reaction cycles and happens in stages that correlate strong with changes in stellar structure: In the dig bang model of the evolution of the universe, an extremely dense and hor y/em expands and cools, giving rise to stars by collapse of local su percritical masses of mixtures of hydrogen and helium (~25 percent helium). During collapse, half the gravitational porential energy of a supercritical ass is converted inro kinetic energy, causing the gas temperature ro rise; the other half is radiated into space. As the temperature soars, ionization strips electrons from hydrogen atoms, leaving protons. (p.n) + (2p, n)— (2p, 2n) + 2p (Reaction 4) 2n) denotes the nucleus of helium 4 (He, also called « particle by). | rall process, the sum of Reactions (1)—(3), can be written as the clear reaction 6pa + 2p +e” + Dv (¢*) eventually annihilate in collisions with electrons (e~) that usly stripped fram hydrogen atoms, thereby producing light (y). 108 (v) tend to Now our and away from the star into interstellar producing heavier nuclei, these fusion steps release energy. A Or the sun will illustrate, The mass of anc particle is only 99.34 © mass of the four protons from which it is produced. Einstein's ion E = mc* computes the amount of energy produced by nuclear Sehormous and dwarfs the contribution from gravitation potential he stn, 600 million tons of hydtogen fuse every second to form Nn fons of helium; thus 4 million cans of matter turn into energy alent to 800 billion kilowate-hours (kWh) of energy every second. On Gf the sun’s available hydrogen will keep it burning at this rare The Proten-Proton Reaction When the temperature of a forming star reaches 10’ K and its density is 10° gicm? (100) times that of water), the transition to nuclear energy generation begins, with che proton-procon reaction. Below the reaction temperature, the protons (positively charged hydrogen nuclei) repel each other by coulombic forces; but when shis repulsion is overcome by a sufficiently energetic collision berween nwo hot protons, then the protons are close enough together for the short-ranged strong nuclear force co act. This nuclear force is strongly arrractiv€ and creates a new combination of nuclear particles, with the release of energy: in the form of light and/or other particles; that is, illion years; but gravitational energy alone could produce energy for only 50 million years. fon of another nucleon to an @ particle leads to unstable products and *He, which have lifetimes of 10-*! 5, pa pala) eet x (Reaction heous. Ina star's core, where temperature and density are highest, accumulate ac the expense of protons. Astrophysical observations Pp lat particles in the core mix only very slowly wich a star's outer ‘ 4 in which p denotes a proton, n a neurron, e* a positron, and va neutrino.10 / CHAPTER 1 ENERGY AND MATTER) BEFORE LIFE / IL ¢ of this stare were theoretically computed by Hoyle and sub- verified in nuclear experiments, BE nuclei then react with hor a's ro form oxygen nuclei (*%O), and fhia’s to form neon nuclei (*”Ne). The process of adding a's to uclei of carbon, oxygen, and neon yield less energy per gram of es proton fusion into a's, bur it is faster, Consequently, this type ves che red giant sa thar no furcher gravitational collapse occurs pite the less efficienr conversion of energy. envelope of protons. The core of w particles grows in size; but at its present. temperature (10 K), the electrostatic repulsion, which is greater berween q_ Particles than between protons (because a's are doubly charged), inhibits « @ nuclear fusion reactions, Much higher temperatures are required for the ae reaction Nevertheless, ar remperacures of around LO’ K, the nuclei of *He and 4H can combine to make “Be, which can react in the star to form other isotopes of lithium, boron, and beryllium. Such reactions explain how elemengs lighte than carbon come into being, They also produce more a’s 7 The star has now developed a core of hor but nonreactive a's and an ow envelope of protons (ionized hydrogen atoms), electrons, and hydrogen pas, A thin, spherical shell surrounding the a core is che region in which quclear fusion and energy generation continue, As a consequence, this region is the hottest (around 3.x 107 K) and the most luminous, «reactions and from the gravitational collapse of the a core ase the core temperature 10 10° K, which is hot enough to of nuclear reactions, che carban-ritragen (CN) cele, see is cycle is pessible in any star that forms from interstellar gas Origin ot Carton products from supernova explosions of earlier-sencration Gravitational contraction of the core of a's, the next source of energy trans: 0 ture it is dominated by the more powerful proton-proton formation, again illustrates the creative force of energy flow. As the core contracts, it heats up due co canservarion of total energy. The hot core the heats the reactive spherical shell, which expands greatly. Astronomers call sta in this stage of evolution red giants When the core w's reach a temperature of 10* K and densities of 10° cm’, the coulombic repulsions between a’s are overcome, and they can r and fuse. Ar this stage something peculiar happens. The natural product of fusion, the beryllium nucleus (*Be), has a lifetime of only 10~! s before it reverts t0.a's. The fusion reaction, 4 (Reaction 7) da "Be + + (Reaction 4 does not produce "Be fast enough ro dominate the decay of "Be into a’, tl Salpeter suggested (Fowler, 1967) rhac perhaps "Be absorbs another a, formin astable carbon 12 (C) nucleus: "Beta PC 4 y (Reactionta i He, carbon acts as a catalyst for the generation of helium from is, the complex sequence of CN-cycle reactions regenerates °C. dur protons into one « particle, woe’, two v, and three ons from prorons to c's and from e's to '°C are both uncom- driven conversions. The CN cycle shows for the first time the an a trimer, '*C. For approximately every 2000 of these , s native pathway converts '*N into “'N plus an a through “O 8Be + a> VOt + y (Reaction 6) Stmediates. Both the '°O of the CN cycle and the 'O are radio- say rapidly; bee 'O is a stable intermediate thar increases in Fred Hoyle suggested (Fowler, 1967) that che short lifecime of “Be meant Reaction (5) could be fast enough only if a resonant reaction Gecurred produced an excited state of the carbon 12 nucleus (which could then de to the carbon ground state, '7C}; where ‘“C* denotes the carbon nucleus in the excited state, The energy @12} CHAPTER | ENERGY AND MATTER. BEFORE LIFE / 15 If the temperature is high enough (10" K), "C and "“O nuclei react with a's to produce “Ne, *'Mg, 788i, and *S (nuclei of neon, magnesium, silicon and sulfur) as the most stable products, Reactions involving pratons (like q reactions in the CN cycle} lead to the intermediate nuclei of fluorine, so aluminum, and phosphorus. Production of Heavier Nuclei Reactions among larger nuctei require higher temperatures because of thei larger ner charges and their greater coulombic repulsion. However, at 3 10° K the process of photodisintegration begins to contribute to these 1 tions, The absorption of fight by che °*Si nucleus causes it to disintegrate in seven a's, which are sometimes caprured by another “Si co produce a ni nucleus (*°Ni). The °°Ni decays into radioactive cobalt (°°Co), which in tu decays into the stable nucleus of iron (°°Fe). Many other isotopically related nuclei are also produced in similar processes. Appearance of the iron group of elements (Co, Ni, and Fe) in some stars suggests that they have achieved nuclear equilibrium at 4 x 10° K and density of 3. 10° gem’. The subsequent evolution of these stars can tak one of several different directions thar will produce still heavier clement ‘One outcome is a supernova event, in which the star explodes and distrib the products of its reactions into interstellar space. Later-gencration stars w accrete this enriched matcer as they farm. For che purposes of this book, however, discussion of the evolution of the elements will end with °Fe: the hot star, nuclear reactions occur in a plasma of charged (ionized) nu and the electrons that have been stripped from chem. Ar much lower tea peratures (10°10 K}, planet formation occurs concomitantly with ato element and molecule formation from these electrons and charged nuclei, _ a elements needed for biology are indicated in gray. Some of the AE in our treaement, such us aluminum (Al) and silican (Si) since cearh's crust. us amounts of energy—much more than from changes in tial energy, This nuclear cnergy is created by the conversion gy. The CN cycle converts protons into a's, with carbon Biological Elements, the Primordial Dozen - In addition, the a's can react direcely wich oxygen nuclei The dynamic properties of molecular systems that we recognize as biolog are fully represented by chemical reactions among the initial 26 ele. (Figure 1-1), Furthermore, for initial inquiries into che transition from 104 10 life, the primordial dozen (H, C, N, O, Na, Mg, P, S, Cl, K, Ca, and Fe) tial energy, heat, nuclear energy, and light energy. subser of the 26 elements, suffices as the basic sec from which all biomolee ‘TION OF SMALL MOLECULES FROM THE (Cu), and zinc (Zn)—are necessary for higher-level nervous tissue functions: ELEMENTS Story now changes from the interiors of stars to the surface earth) that formed around a particular star (the sun) shortly system was born. On the earth, energy flow at relatively low Swill convert the primordial dozen elements into more complex 1¢ biologically relevant small molecules. Summary of Stellar Nucleosynthesis Gravitational collapse in stars converts gravitational potential energy it heat and lightand drives the fusion of protons into a particles. Further coll! and concomitant heating fuses a's into carbon nuclei. These fusion proc4 / CHAPTER I Evolution of the Earth Prevailing models of the evolution of the earth envisage an early stag aceretion about 4.5 billion years ago, Around 4 hillion years ago the p was so hot that the iron-group elements (Co, Ni, and Fe) melted and, be of their great density, passed through che lighter, silicate rock co form q core of the earth. This event is called the row catastrophe. The surface began to cool and develop a solid crust, which was very thin relative to earth's radius. The iron-nickel core solidified under great pressure, Cug opinion is that the solid iron-nickel core is surrounded by a region of fi iron and nickel; together these regions account for roughly half the volume, Floating on this liquid iron and nickel is a molten mantle, pri of silicates, perhaps 2000 km thick. The solid siliceous surface crust ig ¢ about LOO km thick, about as thick as an ege shell compared with the diam of the exe. Outgassing through the crust built an atmosphere rich in H.0, No, and CO. By 3.6 billion years ago, the oceans bad begun to form: and crustal plates, solidified out of che lighter material and floating on the molt interior, were being shaped and moved on the planet's surface by an on-goiny Process called plate tectonics. On this enrly earth with ics nascent oceans, tem peratures were about 7 x 10’ K and densities were becween | and 10 gf Because light gases such as hydrogen and helium had escaped fro! atmosphere, the environment was composed of hydrogen in molecules me complex than H and of heavier elements. Table 1-1 shows estimaces of abundances of the elements in various environments including those organisms. Almost all the hydrogen on the earth and in organisms in water; oxygen is abundant in warer, in COs, and in silicates in the cf Note for later discussion that the amoune of phosphorus in organisms is 1 gtearer than in any other environment except the carth’s lithosphere. Free Energy of Formation, Predictor of Stability Consideration of the free energy of formation (AGP) of various molet Bives some idea of what substances to expect an the earth's surface (lithosp + hydrosphere + atmosphere). By convention, the free energy of form: is zero for the most stable form of every element ar 298 K and 1 atmosphe of pressure (chat is, AG? = 0). Thus carbon as crystalline graphire has @ energy of formation of 0.0 and is stable; but as diamond it has a positive: energy of formation (AG? = 0.685 keal/mol) and so is slightly unstable. 1-2 presents che empirical formulas, molar weights, and standard free ene! of formation for a variety of elements, ions, and biologically imporcaat 1 ecules. Because H2O was abundant on the primitive earth, the properti many of the ions are given for aqueous solutions (aq), which are the biologi relevant forms, (The units kilocalories per mol can be converted to kilojoul! per mole by multiplication by 4.186.) ‘The more negative the free energy of formation the more stable a subs TL 10.0 0.02 22 10.67 0.22, 27.0 “He 1.67 0.18 83.94 0.01 73.53 13.3 0.0L 0,02 85.79 0.19 0.02 47.33, <.01 <01 3h a a 4 53.3 <01 eo a6 2,46 = ENERGY AND MATTER OG 2.07 BEFORE LIFE / 26s és 2.46 3,47 45 15 7.8516 | CHAPTER 1 TABLE 1-2 Free Energies of Formation tion (continued) Molar Empirical weight Enopirécal Name fornsule tain) format Acetaldehyde GH,O 44 CH.O; Acetic acid C2H,O, 60 cHOS Acetate (ag) GHSOF 39 CoH) 205 Acetyl CoA, Ca HyoOyNoPsS 825 CoHsOoP Acetyl phosphate CoH50,P Ho CoHisOuF2 ei-Aconitate CoH,O2- vt C\H.O, Alanine C3H;O,N Bo CoH 1206 L-Arginine €oHj02Ny 175 CoH 205 L-Asparagine C,HsO.N, 132 hate CoHisOoP L-Aspartate C.H.O,N~ 132 CyHsO4N ‘Ammonium ion (aq) NH is CsH\pO5N; Hydrogen carbonate tag) CHO; 61 CyHpOy Carbon dioxide ¢g) co; dd CyHoOQeP Carbon dioxide (ag) cop 44 } €3H;0,N Carbon monoxide (g) co 28 ‘-phosphate CsH,O,P Citrate CgH,OF- 189 HO” Creatine C\HLO.N, 131 Me Creatinine C\HFON, 113 ERE Cysteine C\H-O,NS 121 HO; Carbont ts, geaphitey Gs 12a +s Chlorinet om 1 a Chlorine ion (ag) ar 45.5 ie Caleiumt (e) Ca, 40m a Caleium (agt Cat* 40 Fe Calcium hydrogen phosphate (¢) CaHPO, 138 C,H,0}- Dihydroxyacetone phosphare CyH-OgP 170 CoHi302N Erythrase 4-phosphate CsHoO,P 200 CHF Ethanol C,H,O 46 C\HyOF Formaldehyde CH;0 30 CHO Mater wetght (glial 4G AS 180 260 340 115 180 180 260 146 146 92 iva 75. 170 7 2 Wy a4 M4 27 35.80 55.8 55.8 189 BL 144 89 332 ENERGY AND MATTER: BEFORE LIFE! 17 ~ AG Sou tkealimoli* 85,1 B38 218.7 420.0 6213 Mad 220.6 219.1 420.5 166.5 125.4 16: 319.2 90.0 307.1 37.6 0.0 36.7 327 65 —28,7 0.0 20.3 2518 ( CHAPTER 1 TABLE 1-2 Free Energies of Formation (continued) ENERGY AND MATTER: BEFORE LIPE / 19 mation (ioutinised) Male Molar “ rica weight =AG os Enipirieal weight Emptric ; : Nave fornvala taleaol) foveasla (girth (sated) t-Leucine CoH ON ial % 177.3 Methane (g) CH 16 80 118.8 Methanol CH.O 32 Si 32n 00 i-Methionine 3H, O2NS 19 Sie 28H 0.0 Magnesiumt (¢) Mg, Man Na, Bn 0.0 Magnesium jon (ag) My"? 24.3 Na* 23 62.6 Nitrite ion (agi Nor 46 Nach 58.5 O18 Nitrate ion (ay) NOS 62 C\H.O,N 105 122.1 Nitrogen? (2) Ny 28 CiH,O,N 119 122.9 Oxalate C.0% 8s CHiO.N Isl 92.5 Oxaloacerate C,H,0f> 130 CuHLO.N. 204 20.9 ol 48.7 Oxygent (2) Oy 32 CHyON,» GO 8. Phosphoric acid HsPA, oR CoH) ,0:N 115 86.0 Ng 33 Dihydrogen phosphate ion (ag) HsPOy 97 C.H,02N, 135 3 2 CHO: M45 178.7 Hydrogen phosphate ion (aq) HPO} 06 Ais . HO 18 36.7 Phosphate ion (ag) PO} 95 Phosphorus? («, whited Py Bln Pyruvate C\HOs 87 aphoerlafie He as €§ equilibrium. However, the rate at which equilibrium is Howsshumts(p} Ky, 39:19 tnany orders of magnitude for different substances, Con- Potassium tom (ait Kt 30.1 rium penederarons, are only sugseste es Settee: “3 4S. prominent product, given its free energy of formation Mibese Sephosphere Ss OaP 22 and quart (S102), a major crustal component, has a free Ribuloge 5-phosphate CyH i OgP 230 of —192.4 kcal/mol. Calcium phosphate (CaHPO,) is aL : be Quartz («) teilicon dioxide) SiO, 60 Re uilibrium,as are the aqueous ions of magnesium {Mg’ ~ } seBheayletanine GHnON, 185 ay, expressed by RT where R is the gas constant (1.99 call Sedoheptulose T-phosphate CoH QyoP 290 dard against which to compare these values. At T = 298 Riecénale C,H,02- 6 tatistical mechanics shows that at equilibrium the relative a ste i reschice of a molecule with a free energy of formation Suceinyl CoA CosHaOiN:Ps5 882 the Boltemann formula exp(—AG%oy/RT). Almast every Be P os Sucrose (CEH 342 .€xcept the elements themselves, is thermally stable at T20) CHAPTER 1 ENERGY AND MATTER: BEFORE LIFE / 21 = 298 K. (25 °C). Note that molecules of biological significance are particularly amino acids, such as alanine and glycine, and simple carbohyde such as fructose and glucose. Therefore the basic building blocks of matter are the relatively stable and therefore abundant small molecules spontaneously form from the elements, a fact thar is consistent with the, servation that sueh molecules are also found in interstellar space, in meteorig and on the surface of the moon, molecules, such as CO2, Ny, and H20, into combinations amino acids, which become the building blocks for life. ONOMERS: THE BUILDING BLOCKS OF LIVING SYSTEMS all molecules {MW = 300) that naturally arise from the effeces the primordial dozen elements, the monomers are of special nclude amin acids, monosaccharides (sugars), and the purine Two views of the atmosphere are shown in Table 1-1, one of which is laly idine bases. What distinguishes the monomers and makes them spe- primitive. This particular primitive atmosphere contains primarily COs, Ny They can link together into larger molecules, the mixed oligomers, and H»Q—nor the highly reducing conditions postulared in the Oparin- ones, the polymers. scenario. Instexd of a highly reducing primitive atmosphere, many im ‘ocess that forms these linkages, found almost everywhere gators have now proposed a less reducing, or even nonreducing, p dehydration condensation because formation of the linkage atmosphere such as that shown: in the table. The poinc here is not to lecule. Tae following reaction shows the linkage of two which conditions really existed on the primitive earth, but to point out rs, in which the H2O arises from portions of the carboxy! a number of possible atmospheres might have given rise to conditions f amino acids 1 and 2, respectively. The product (a di- able to life. The Oparin-Urey scenario for a primitive atmosphere Possible Primitive Atmospheres the important peptide linkage of proteins: methane (CH4), ammonia (NH), and water (HO), as opposed to oO dioxide (CO,), No, and HO, or other choices, When one of a varie I energy fluxes passes through one of these various mixtures of gases, the cn come is @ vast variety of molecules with negative free energies of fort many of which are listed in Table 1-2, The type of energy flux—for ex: sparks, ultraviolet (UV) radiation, and heat—and the initial composi the atmosphere affect the composition of the product mixture, (Dera garding these experiments can be found in the references.) OF course, | initial mixture docs not have a nitrogen-containing component, n0 nil containing products can be formed. Many of the molecules that can be made by adding energy to gas like the above are biologically relevant because life arose under similar cumstances. The experimental setting arranged by scientists and the setting arranged by stellar‘and geophysical processes automatically p the molecules that are most likely to form from the primordial dozen eles Moreover, a variety of settings in ehe atmosphere, hydrosphere, and [thas may be relevant simultaneously. Products from one setting may enter a and thereby enrich it. Exactly what happened 4 billion years ago on the! may remain unknown, but an understanding of the possibilities is within | H i P R O Bits iy * --€ | H [Oo UH H oO Anne acid 1 Amino acid 2 Pepride linkage R, O Ra O lll | d Ni C—C—N—C— + H:0 \ H Dipepride Water molecule of dehydration condensation is the linkage of two phosphate Summary of the Formation of Small Molecules duce pyrophasphate= Elements formed during stellar nucleosynthesis combine into variet small molecules at the relatively low temperatures on a planetary crust: nd if later sec:ions, stricradherence to the correct chatge states of molecules fluxes, such as lightning, ultraviolet radiation, and volcanic heat, cont wed unless i is a crucial consideration.22) CHAPTER [ ENERGY AND MATTER: BEFORE LIFE | 23 ° I avail | i HO H+ HO pea HO po + 10 o- ‘ o oO Phosphate Phosphate Pyrophosphate Water Amino Acids, the Monomers of Proteins ly as closed s in solution, such as B-D-glucose: When researchers mix CH), H)O, and NH; and make them react by axist largely as closed rings i p-D-2 plying an cnergy flux of electric discharge (Miller, 1953) or of beat ( and Fox, 1964), the elements H, C, N, and © combine. A: CH,OH the resulting molecular species. Their general structure, in aqueous milie H Q OH « H H—C—COs OH H I H NHe HO HOH ups (—OH) in monosaccharides take part in the dehydracion id the moromers into polysaccharides, Table 1-3 lists the s that are important for this discussion, Their structures ate iochemiscry text {Stryer, 1988]. xe structure (CHOH)},—or, equivalently, (CH:O},—has a mation per CH; unit that depends on w. Formaldehyde, s a free energy of formation of ~ 31 kcal/mol, compared with mol for CO, and H2O (think of CO, + H»O—> CH20 + O2). n = 6, has —36.5 kcal/mol per CH20 unit, In all cases, the organisms have more than 100 different R groups, of which they use only. to make their proteins. The structures of these special 20 are in any. st biochemistry ext (Server, 1988], The dehydration linkage berween the acids that make up protein chains occurs between the carboxyl and’a groups and does not involve the R groups, The simplest amino acid, glycine, for which R is only one atom of hydrog has a free energy of formation of —90 keal/mol in HyO, whereas its constituents (CO), NH4, and CH) together have a free energy of form * TABLE 1-; of ~ 123 keal’mol. Consequently, although glyeine is enezgy-rich rela Simple ees its constituent small molecules, it has a negative free energy of formato! es = relative to its constirwent elements. ‘Thus glycine is of intermediate th Monasaccheride ” Linear Ring dynamic stability, as are the other amino acids. Amino acids existed Glyceraldehyde 3 x primitive earth because the geochemical conditions on its surface we equilibrium conditions but invalved a variety of energy fluxes, which gene! Erythrose 4 molecules of intermediate stability. Ribose. 5 x x Ribulose 5 x Sugars, the Monomers of Polysaccharides Xylulose 5 x Melvin Calvin (1950) reacted H. and CO: together using an energy Glucose 6 x x radioactivity to produce monosaccharides (sugar monomers), all of ae Fructose 6 x x the general structure (CHOH),, where m= 3, 4,... , 7. For examp! | Sedoheptclose 7 ao coraldahwde che ccicee (hres barbae acca, hag the:follderne eric Lite rprelose24) CHAPTER | BNERGY AND MATTER BEFORE LIFE 25 CH NEL, Hyco feo-llo bey 4 b i g NH, yen phosphate ion Pantothenic acid 4 ‘Adenine (a purine} Guanine ¢a pusine " tae! . ee a “1 u Ht q ine of‘ aa Mt ™ N—-H ry, rc oes TN yg ey th i eR 6.) -Dimethylisoalloxazine (oa SF ALOALIA Aiteenin BD sriatatee: red with the elements themselves. This fact shows thac like the amino acids, are of intermediate stability from an folecules, the Monomers of Mixed Oligomers wii A ‘species are also regarded as monomers because chey can pdebydration ta praduce mixed oligomers. These include many importance. Figure 1-2 displays a rep- he purines and pyrimidines, for example, are important cam- ribonucleic acid (DNA) and ribonucleic acid (RNA) and > coenzymes, the catalytic centers of many enzymes. Noligomers ace of contemporary biological significance, and all Products of sbislegical geophysical processes (Ore and Kimball, mple, CH, NH}, and H2O can react in an energy flux of 8 lenine ot ina flux of electric discharges to produce porphyrin; ible constituent of primitive planetary settings. However, ¢ does not imply that biology is present; it is only a pre- ogy can become a feature of our picture only after these mon- Cyeosine (a pysimidined ‘Thymine (a pyrimidine) FIGURE 1-2 Monomers: purines, pyrimidines, and coenzymes.26 / CHAPTER | omers link into larger structures, The next section explores this possi and its requiremems for energy. 3 8 2 Summary of the Origin of Monomers : + HO Using a variety of energy fluxes, simple precursors combine to be ‘ H HOA HOH monomers. In che labararory in simulated, primitive earth environment . include CHy, NHs, CHO, and HO, scientists bave produced amino ie em sugars, and the genetic bases. / een the G-1 carbon of glucose and the C-2 carbon of fructose | of dehydration condensation called & wfyeoride linketxe, l-4, MIXED OLIGOMERS: SHORT CHAINS OF VARIOUS MONOMERS This section discusses mixed oligomers (so called because they consist o (0 ten monomers that are not all che same kind), which are constituents enzymes, genes, vitamins, and such important molecules as lactose and 5 Subsections treat oligosaccharides, nucleosides, nucleotides, and coenzy Figure 1-3 shows several biologically prominent mixed oligomers-a constituent moieties, each of which combines by the dehydration Many af these species are coeazymes chat are indispensible to energy olism in contemporary organisms: Notice that nicotinamide is a dehydi linkage berween nicotinic acid and ammonia. ( Nov all mixed oligomers form in the abiological context of the p earth. Section 1-5 addresses their actual occurrence, along with that polymers. nomer), One molecule of the furanose rife and a molecule imine combine by dehydration condensation to form adenosine: NH» Oligosaccharides Oligosaccharides are comprised of two, three, four, five, six, or mor monomers. Sometimes chey form as Aydralyits products of molecules s maltose (which has owo sugar monomers) or cellobiase (which has cv hydrolysis, water inserts itself into a molecule, the reverse of dehy condensation.) The sugars lactase and sucrose also are oligasaccharid sucrose forms as the dehydration condensate of n-glucose and D-f H Adenosine H ie A;~P + P (Activation of amine Ajp~P + WA, + P + heat (Hydroly, Peptides Assisted activation is the key to producing ag macromolecules, which can have so-called emergent ‘ivated opertiet of systems can be predicted only with great of meta ee ie properties of a system's components. ‘Thus amino acid) molecules are difficult to predict from the behaviors : is that some macromol- Ay~P + A; AWA, +P (Peptide bond forts mers. Che emergent property is thai A,;~ P + Ay~ P— AA, ~ P + P (Activated peptide fort act as biological catalysts. A modest catalytic activity ydrolysis rate of one peptide bond per month. A turn- cond is 100 times faster than hydrolysis though it is 10" the best contemporary enzyme. Carbonic anhydrase holds where A denotes an amino acid, ¢ an integer index, P a phosphate he is ighest rate of turnover of substrate molecules: 6 x 10 a water molecule, the symbol ~ an energy-rich phosphare bond, and bol — a typical covalent chemical bond with less free energy cont ‘The equations show that these reactions are not aided by any caté In addition, they are virtually irreversibile.* The peptide bond and A, bas a AG (free energy content relative to hydrolysis) of * RTs, kcal/mol. This means that A-—A, will hydrolyze and bur the bond energy is otherwise not useful: thetical and incomplete constructs cannot initiate a mo- they demonstrate the problems involved AA, + WA; + A; + heat (Peptide hydrolysis) watalysis (Hypothetical) Assume that three particular ces Of five amino acids) exist: is is only an approximarion, albeit a good one when the reaction involves a free Aa Ay Ag —Au—Aa, large compared with the scale set by che thermal energy: RTax = 0:59 keall Fich phosphate bonds in phosphoanhydride and amina acyl phosphate groups have Ab Ay. — Ay, —Ap—Ap, of hydrolysis differences of about 10 RT 2a. This makes hydrolysis virtually Six 10") Aci—Ag—Ac—Ae—Acs44 / CHAPTER 2 THE PRIMITIVE UROBOROS ! 45 which are abbreviated ,, F,,, and E., respectively. Further, Suppose pentapeptides exhibic catalytic propertivs: E, catalyzes the forma P from a hypothetical energy flux, Es, catalyzes activation of aming P~P, and E, catalyzes polymerization of A, ~ P and A; ~ Pp: ‘Abundances of Amino Acids in Various Meteorites Hot terrestrial tana energy flux + 2P ay P-P Glycine Glycine : Alanine Alanine ween Y soe Glutamic acid Glutamic acid ‘Aspartic acid Aspartic seid. Valine Serine Proline Threonine Assume chat these three catalytic functions proceed fast enough so: aa rates prevail aver the hydrolysis rate. That is, although these cat tapeptides are also subject to hydrolysis, they possess long half. were essential components, they would appear, regenerated, amor ucts of the polymerizations. This is pare of the uroboros requirem Isoleucine Leucine Laboratory Simulations Protein Syuthesis (Incomplete Caustruct) Mixcures of amino ack Redising alnosphere and electeic Hieting to dryness, yield short chains of amino acids called thermal proteins, discharge tefuotds. Experiments (Fox and Dose, 1977) have shown that the le the amino acid sequences of proteinoids are not random but ared largely by the composition of che amino-acid mixture and other r ditions, and chat the yield is aided by the addition of phosphat shows the constituents of amino acid mixtures in various envi About che time life began, this process could have been active Glycine Alanine Glutamic acid Aspartic acid crust. Pools and lakes of liquid witer containing dissolved small a Setine including amino acids, had accurnulated in shallow depressions Valine When the water evaporated or splashed onto the surrounding rock, oi h coline in the rock would have reacted the amino acids to form proteino refilled che pools or tides washed the proteinoids back inta the DN yasdanaiitea me If some of the proteinoids possessed catalytic functions such Pjiaier i 76) cucecfal syashests of 3 vobuarcial .and E., then polymer spathesis wauld begin. Such a sel-bege j Perr ar telaciiE abundince of abparic and would alse be self-maintaining: the primitive uroboros. 5; otherwise, heating amino acid mixtures yields E i = ss this incon rar. Compositional analyses, of hundceds of proteins ad SR ERORe Bare: Era ' show relative abundances of these ewo amino steuct of protein synthesis. fi anmals, glueamic acid is in greatest abundance Polysucleotide Synthesis (Incomplete Construct) An cnerey- sition from nucleotide monomers to polynucleotide polymers: attractive features that are not possible with proteins.” The seructl oligonucleotide molecules supports the existence of base-patring ™ ‘Aucleotide monomers on an oligonucleotide template tains a sequence homologous to the sequence used as the double-stranded form is literally replicated by im of polymerization. This mechanism reproduces : 7 y oF 4 biochemistry textbable such a amie : i Sa eee re coals he loriesy ona lnehominty oestb ne ie Sequence: For the:single-stranded form, the first46) CHAPTER 2 THE PRIMITIVE URQBOROS activated nucleotides do nor spontaneously polymerize c. Bur where could a template have come from ini- roblem exists: Even in experiments with polynucleo- oer nucleotides, polymerization is difficult co achieve momers tend to hydrolyze faster than they polymerize. the slow reaction rate of nucleotides is overcome é ymerases; and the hydrolysis of activated monomers is ofrreactions, also catalyzed by enzymes, that continuously An addition to these problems, the coupling of any ix to ~P ~ P production, which is required for paly- also requires proteins. How could an uroboros have been Camlpric feedback effects Food ar sunlight Ener “a Ps Yap free PoP ne uP x 4) Monomers Polymers yee nH Monomers anisms use to regenerate the critical monomers help soblems of polynucleotide synthesis and uroboros ini for generate purine mononucleorides simply by linking phate ion, and a purine base; they begin with ribose- kage thar is not energyrich). Through a succession of ; pieces of aspartic acid and glutamine (amino acids), bamyl phosphate—organisms synthesize the purine in ‘sugar. These reactions require energy. To synthesize es no fewer than five high-energy phosphate bonds tivate intermediates in the pachway. Enzymes catalyze servations imply that a procein-generating system must fore polynucleotides became abundant, They also em- bond energy was plentiful, chen an amine acid-to- we been driven in the direction of these later phos- cesses. It is my view that modern polynucleotides are a ontinuing stage of energy transduction development in sphate bond energy dominated, and still dominates, the atter. This view is elaborated later with respect to met- ent mechanisms, and major changes in phylogeny. N, 4 2P—P+N,- Pi on RGY METABOLISM AND POLYMER [5 IN CONTEMPORARY ORGANISMS ms used by contemporiry organisms to generate phos- jour the ive uroboros must have co be recognized as living. sins'in many ways, and most of the wypes of proteins that r catalyzing many reactions, including those in energy synthesis of monomers, protuins, and polynucleotides. sequences of tees sequence fidelity during polynucleotide repli- product.on. y Metabolism lism pathway is a series of reactions that makes energy ‘principal pathways used by organisms are g/yealysis, che | way (or hexose monophosphate shunt), and the electran of o 6 nx tofo # sie May: i a oo in two forms, one for photosynthetic cells and In aerobic cells, carbohydrate oxidation and electron d by the éricarboxylic acted cycle (Keel cle) which alycolysis and places them in intermed to the electron cransport chain. § 2-2, 2-3, and 2-4 serve to emphasize the cencral fey Processing. These pathways are complex, have many and enzymes, Qa iate matecules, I + HzN—C—O- _ (Pyrophospha' (Carbumate) Chain in each of the preceding sequences, formation Is captures same carbohydrate bond energy that the elec- €ventually uses co make ATP. The electron transport ; Feactions that involve iron-sulfur proteins, quinone, and the electrons and cwo protons reduce molecular oxygen Transition from Heterotrophic to Autotrophic Polymeri# In the most primitive stages of the development of polymeriza omers, polymerization required ready-made molecules and thu trophic, The monomers were available as by-products of geophTHE PRIMITIVE UROBOROS / 51 Glucose ADF Glucase- phosphate Fructose-. The elvetrans carried away by 3NADH and FADH retain mucly of the OFig energy af the acerite. and another porrion of the free energy i in the phosphute-d Potential of GTP, Onaloacetate reforms, and with the addition of more aceey! Continates by extracting electrons from NADH, transferring them to FAD, FADH), which then transfers chem t one of these iron-sulfur pI cytachromes contain heme iron (iron complexed to nitrogen-ato phyrin ring) which is linked to the protein pare of the mol bridges (bonds). The several cytochromes are distinguishable by components. The energy metabolism pathways just described depend on la of enzymes whose synthesis is driven by the energy gener pathways," ‘ cyt, cytochrome; FeS, iron-sulfur protein, and F) and ATP synthesizing eoxyme. ms that specify the amino acid sequences of all che ‘Organism reside in its genes. The process of protein 8 the beginning of life appears in this section; the reader iochemistry or molecular biology for grearer detail 1986; Stryer, 1981; and Watson, 1975). ive portions of enormous double-stranded molecules yibonucleic acid (DNA), known as chromosomes. | Blectron transport, which depends an oxygen, is more complestand more ree 4m acceria, if occurs in the ¢ell membrane, correlaced wich an intricate spatial membrane proteins, whereas elycalysis occurs within the gellike matrix af tHe bur oxygen or special arrangement of enzymes, Mast of the cell's total catbol harvested by electron transpart (aout 90 percent) cacher than by glycalys!34 / CHAPTER 2 Each strand consists of a sugar and phosphate backhone, to which the nucleic acid bases, Connecting the wo. Spiraling strands of Dy merous hydrogen bonds, which form between the pairs of bases: and thymine (T), guanine (G) and cytosine (C), When sex cells. enzymes replicate the strands of DNA in such a way that every inherits a full complement of parental genes. Before enzymes must separare the entwined strands and [ onucleotides along the separated strand Tand G-C), and phosphodiester bonds must form beeween adjacen nucleotides, If these events could vecur Spontaneously, DNA wa be self-replicating, provided the nctivation energy were available, Fy replication requires the aid of numerous enzymes. For instan bacteria, initiation of DNA replication requires the so-called uny aymes and RNA polymerase (a large ine up energy-act cording to base-pairi CAA Glo 3 Protein), After synthesis ofa CAG Gln of ribonucleic acid (RNA), DNA polymerase III (another large pro —— the picture. Finally, DNA polymerase | and DNA ligase finish the | AAU Asn has used 12 t0 20 enzymes, ‘ AAC Asn Constructing’ protein from the code in-a gene is one of the mi processes in cell biology. After enzymes unwind and separate th Thr AAA Lys of DNA, the enzyme RNA polyme rasé transcribes the gene into RNA (mRNA), synthesizing a new chain of nucleotides using template; so the mRNA sequence is always complementary to the) sequence. An earlier and separate transeription has made three RNAs (FRNA), which organize into-a cellular body called ribo : GAA Glu of protein syathesis. The ribosome binds the strand of mRNA, Th R ; GAG Glu (tRNA), also teanseribed earlier from special genes, contain a - three bases called am-anticodon, which by base-pairing, can "re uf three bases tcadons}on che mRNA. The their order constitute the code structed. Ac least one specific nds of bases on for one amino acid of the protei tRNA exists for each amino acid. forms a complex with mRNA and tRNA to read the mRNA and formation of a peptide bond berween the last amino acid on tht and each new amino acyl (RNA chat is brought to it. The energy for protein synthesis is derived from ATP; an € z ERNA synthetase, binds its CRNA, its amino acid, and ATP. Fits bonds to the amino acid, activating it into an amino acyl adenyl synthetase (which is specific to the RNA) bonds this activated af the tRNA. The resulting amino-acyl-tRNA is still sufficiently enet form a spontaneous peptide bend, In other words, as a ribosome the m-RNA strand, everytime it reads a codon it “calls in” the tRNA, which positions its amino acid an the growing chain, Wi bond anchors i With a tripler code and only four bases, there are at most 64: amino acids. Thus the genetic code is redundant, as is shown This redundancy is accommodated by 30 to 40: tRNAs and at I synthetases. THE Ofeins, and the synthe: PRIMITIVE UGU UGC UGA UGG T Cys Cys End Trp cGU cGc CGA CGG AGU AGC AGA AGG GGU GGc GGA GGG Arg Arg Arg Arg Ser Arg Arg Gly Gly Gly Gly UROBOROS | 55 Amino Number of acid Ala Arg Ash Asp — Cys Gla Glu Gly His Ue Lew Lys Mee Phe Pro Ser The Trp Tyr Val End RH ROUN BRN Ne & codons tain about 50 proteins, bue che ability of the sibaearte slation of m-RNA into protein requires the action o| alled factors. Some factors function only if they bind to of GTP, che energy-rich analogue of ATP. The energy of the formation of the peptide bond berween amino acids ‘smooth functioning of the complex of ribosome, mRNA, tch of protein synthesis emphasizes, the replication of of proteins depends on genes, ns. Both processes require activation energy, which is ATP is generaced by the energy metabolism pathways, es. This is the uroboros puzzle, How could ic have it must have been simpler. ; ¥ Organism, both energy metabolism and biopolymer tolecular mechanisms, where energy requirements for net by energy metabolism pathways that need numerous ist5are the products of polymer synthesis. To ensure id ‘appropriare quantities of catalysts, the cell's genetic36 / CHAPTER 2 Seale for Evolution (comsinued) Beginuing apparatus directs protein synthesis, which requires the symthesiy of (years Bed ypes of polynucleotides (DNA and RNA). And polynuclearid requires energy and protein catalysts. Together, the processes of cote tabolism, gene-directed protein synthesis, and gene replication cane molecular uroboros, but one of onside ha Period Permian 290% 108 table complexity. The problem g {0 initiate a molecular uroboras can be solved only if the peimitics Beotniet NN Possesses a much simpler structure and also has the capacity ta oth contemporary complexity, The next sect ssi A ississiPP zeae molecular uroboros. soni prseniss migclel (aaa Mississippian 36 Devonian 410% 10° 2-3. INITIATION OF AN UROBOROS Throughout the discussions that follow and lead 10 the proposed m the primitive uroborus, the reader should carry a feeling for geologi Initiation of the relatively simple, primiti iri 430% 108 happen quick! n after i i iv — “ happen quickly, even after conditions on che primitive earch ‘were fay ies hfe. A look at che history of biological change on the geological tim lps One appreciate the time scale of che earlis i rae scale of the earliest stages of evolutio Ordovielan 500 10° Though fossils of microorganisms are scarce, atid fossils of macrom are nonexistent, several clues to early life do exist, About 3.5 billio ago, bacteria growing in extensive clumps lay down the substrates fo : i macolice fossils (Ballard, 1983, p. 44), the oldest of which are in AW oe oe These bacteria were probably heterotrophs, which did not use phorosye TABLE 2.3 a0 10 PALE 2 700 10" Geological Time Seale for Evolution Beginning a Eris Povicad (years wr.) Stgmaficant CENOZOIC Quarternary 2x Loe Development of m pestesnanic ee Tertiary, 63x 10% Mammals dominate; ieee a a plants spread and { 2100x108 MESOZOIC Cretaceous Tan x 10° Extinction of dinoss vonnainee 2500 x 10° floweri jants ap jowering plants 2700% 10° Jurassic 200% 10% Primitive mammals, reptiles; birds appt oe Archean 3500 x 10" conifers abundant 3800 x 10° 240% 10° Repriles dominate; appear; cycads appt 4500 x LO Significant evenis Repiiles advance; huge plants decline Cozl forms from undecayed layers of wegeration Insects abundant; amphibians advance; first reptiles: tree ferns; conifers First trees and forests: amphibians appear: primitive insects, searpions, and spiders on land Fishes advance; ozone blocks UV, planes spread on land; coral reefs extensive Fishes (primitive vertebrates): explosion in the vatiecy of sea lite (invertebrates), deposits of limestone Masses; diversification of invertebrates such as brachiopods, sails, trilobites, sponges First organisms visible to unaided eye; sponges Sexual reproductian; lime- secreting algae First aerobes Stromasolites widespread Early red sandstones (Q; in atmosphete} Stromarolires increase Photosynthesis by blue-green algae Bacterial sccomarolite fossils Iron catastrophe; origin of oceans; origin of Accretion of the planer Earth58 ( CHAPTER 2 THE PRIMITIVE UROBOROS | 59 or oxygen for enna: later I shall propose how such eatly organisms may p fig eomplicaced interactions of ribosomes, DNA. RNAs, synthetases, been energized. At this time, the generic apparatus for protein sy es, and protein factors. The system muse have been simple (and the the rudiments of energy metabolism were evolving. Stromatolites y ust have been simple} and must also have had the capability for in their ascendancy 2.5 billion years ago, even though photosynthe sy into contemporary complexity. evalved (2.7 to 3.0 billions years ago), By 2.1 billion years ago the fj sandstanes formed—red because the OQ» content of the atmosphere creased enough (as a result of photosynthesis) co rust the iron in # However, only about 1.6 billion years ago did atmospheric oxygen fy plenutul enough to support aerobic fife. Sexual reproduction began 900) lion years azo, and the first metazoans large enough to be visible atose 700 million years ago, All these events occurred in the first 3 billion yi life, in the geological era known as the Pre-Cambrian, i All advanced forms of life have evolved during the last half bi ‘Energy Transduction pling of oxidatior-reduction energy 10 phosphate bond enerky “been the first event on the way ro life” (Lipmann, 1971). Even had cooled enough for liquid water to accumulate, much thermal till present, caus.ng some areas of the earth's crust co be wer and tot and dry. The oxidation state of iron in the crust and in other combinations was either Fe(II) or Fei {II} depending on the envir Every new fossil find changes some detail of the picture, such as which @ [Gp fron is one pocestial source of readily available osidation-reduc- ganism first arrived on dry land. Bue whichever one did, a sufficient Another is sulfur, which occurs in different oxidation states, such of atmospheric O2 occurred first, forming an ozone layer in the Upp sulfate, each oF which has a large negative free energy of for mosphere. This ozone layer is essential because it filters out UV tad -2), The transduction of this oxidation-reduction energy into from the sun, which would have been deadly to terrestrial life, Th versatile and usable form for organic synthesis requires its conversion layer initially occurred during the Ordovician period, around 400 milli 1 © bond energy. This conversion, in my hypothesis, could have ago, and saturated during the Silurian, by geophysical conditions that may have been especially scrong in catastrophe. contemporary systems suggests two models of che origin sduction: oxidative phosphorylation and chemiosmotic phas- ‘The mechanism of substratc-level oxidative phosphorylation, in glycolysis, may have come earlicr than chemiasmotic phos- ich occurs aerobically and photosynthecically. The only other of substrate-level phosphorylation occurs in Thiohecillus bacteria, lizing organism. Both of these oxidative phasphorylation mech- lve sulfur. A closer examination might reveal early mechanistic The Character of Primordial Life The energy metabolism of the earliest life on earth was probably a phorasynthetic nor aerobic, but was similar co glycolysis or to some: or analogue of it, As already mentioned, the glycolitic pathway is indep ‘of photosynthesis and the electron cransport chain. Ic is the least sophi of the energy pathways, struceurally, in that its enzymes function in and are warer soluble, whereas che enzymes of the other two. pathway highly organized within lipid membranes. Nevertheless, glycolysis sis, the steps of interest are those that metabolize glyceraldchyde- isphosphoglycerate, which in turn phosphorylates ADP co ATP the phosphate ir the 3-position is nor involved in energy coupling, 1€ enzyme works through a sulfur atom on a cysteine residue at site co form a thiohemiacetal enzyme intermediate. Step (2: NAD~ A-carbon, forming a thioester, that is, an aldehyde oxidizes to a Step (3); Inorganic phosphate phosphorolyzes the thioescer, form- sphoglycerare and recycling the free enzyme, Step (4): ADP is d to form ATP. Reforming the oxidizing agent NAD* js the lation-reduction energy that | must outline here. My hypothesis contemporary pathways. The links with primordial mecha tinct, and their macromolecular substructures were nor fossilized. Th is the most pr ‘ontemporary energy pathway, i be called primordial because of its dependence on enzymes. Similarl contemporary gene-directed protein biosynthesis machinery does not primordial relics. And those links with the primordial also became during the first balf-billion years of life. There are no relics, but there may be clues. Some key enzymes in lh and other pathways have sulfur-containing amino-acid side chains in § tive sites; other enzymes use nonheme iren, Pyrophosphate doubles in some microorganisms (Kulaev and Vagabov, 1983). Perhaps amet for coupling oxidation-reduction energy to phosphate bond energy coul been a system that used sulfur and iron and did not require modern en - The system had to promote polymerizations of amino acids and nucle SH some of this section is hypotherical, the experimental existence of60 | CHAPTER 2 THE PRIMITIVE UROBOROS / 61 cone microme:er (um) in diameter, called mérrespheres (Figure ESSH Solfo canstie pheres enclose chains of amino acids and other solutes in the hich they form. Although the proteinoids contain mo lipids, ie hydrophobic amine acid residues causes the boundaries af the to act as membranes (some are even double layered). The bound- to keep amino acid chains inside the microspheres; but sometimes cals and protons can cross a boundary (though some of che more icrospheres are even impervious (0 pratons}. ' joids can ace as hormones as has been shown experimentally by d formed from six amino acids—glutamic acid, glycine, histidine, phenylalanine, and tryptophan, In a specific sequence, these amino jee the active portion of che melanocyte-stimulating hormone; this ipplied to a frog's skin causes black spots to appear. Black spots also OH, nel ~o- fori H oO H Lethe 5.0) — Tea pe HC-DH HE-OH i Heo NAD? H oO e a NADH + H® 0 on Ho=P—0 =a a © HE-OH 44, c= | HO-P-o-e—H fF 5 He —OH, }oou 8 a? TE ng—o~l-on Suluretyme FIGURE 2-6 Glyeeruldehyde-3-# — 1, 3-bisphosphoglycerate + ATP, Phosphorylacion involves intermediarion of 0 sulfisr atom on a cysteine residue of the 4 the proteinoids is nor: Dry heating certain combinations of the 20 bioley amino acids automatically results in amino acid chains called thermal OF proteinoids, The lengehs-and compositions of these proreinoids the composition of che amino acid mixture, the temperature, and of ditions of the experiment. The construct went only this far: splashin solutions of amino acids onto hot igneous rock and then washing the Proteinoids back into the water. otomicrograph. Nore the cemuriable uniformity in size (-2 micromecers notenaid microspheres. Rach microsphere is self-assembled from Proteinoids in the Laborwiory Biotlemists have discovered that wht y 10! molecules of proteinaid, One gram of proteinoid yields 10° ra 10” teinoids dissolve in water, they spontaneously assemble (selassembTHE PRIMITIVE UROBOROS ! 65 62.) CHAPTER 2 appeared when the experimenters applied the proceinoid to a frop’s cause the proteinoid contained an abnormally large amount of the specific sequence. Microspheres on the Prebiotic Earth On the prebiotic earth, p microspheres could have formed a metastable environment in which processes could occur relatively protected: from harsh geophysical pro A proteinoid microsphere containing sullur-proteinoids would provide vironment for the oxidation-to-phosphorylation process described ei this section. (If the sulfur acom is the key to the enzyme activity in this p then the contemporary enzyme is simply a more efficient catalyse th sulfur-proteinoid.) NAD” is a mixed oligomer (a dehydration conde that would have been available on a marginal basis in the primordial envi ment: It could have been present in some microspheres. If it acted cyeli thar is, if i¢ was reoxidized—cthen small amounts would have been sut initially. The oxidatien of NADH could have occurred by the action of of the previously mentioned sources of oxidation-reduction energy; mo the membranelike activity of the proteineid microsphere could have facilitar ¥ by quiaone and Fo(lll), in proteinpid microspheres, The quinonelike this coupling, A clue to how to accomplish this rcoxidation comes : (Q, i one of the products of small molecule abiowenesis: it acts here #3 temporary organisms. 7 ie aero ahurrle. The electron transport chains in contemporary cells, which are res] for most of the energy harvested after glycolysis, use a quinone de: carry out a key step in the oxidation-reduction sequence (see Figure quinone simply moves between electron donor and electron ace Protein appears to be required for its function. Figure 2-8 depicts idation of NADH, mediated by a quinonelike substance (Q), whi oxidized by the oxidation potential of ferric iron, Fe(III), which i to have been abundant in the primitive environment and, therefore, ne microspheres. (The iron could be a residual of the iron catastrophe: result from UV irradiation of organic iron compounds, which pro stances such as ferricyanide.) This mechanism also suggests that proror from inside to outside the microspheres, thereby generating a mes electrochemical potential. Protons could partially neutralize inorg phate on the outside, thereby allowing ic to penetrate che proteinoid more easily (see Figure 2-9), The final step is the production of pyrophospl my hypothesized source of primordial phosphate bond energy. The Thiebaciltas mechanism for oxidizing SO3~ 10 SO} is perk simpler. The mechanism used by this organism is, of course, cat enzymes (Figure 2-10}; however, the acceptor of electrons from SO} [step (1) in the figure], could have been iron(II) in the byp scenario, In a primordial scheme based on this mechanism, | shall nee the formation of pyrophosphate P ~ P, instead of ATP, as in Step requirement yields the simple model in Figure 2-11, Once some mechanism, perhaps nor unlike che one just propose! transduction, in proteingid microspheres. G denotes glyceraldehyde (or environmentally generated oxidation-reduction energy ro the production eee eos He pr deh sulle pioieinald cailysefoe the Wee ei raid he 2s ca ores Ngoc any | iaiee ar a one saci Blyceraldchyde-3-phosphare. Dotted li 2Fet 11 ar mw Microsphene 2e (1) a 2Fe (LY 13-dieG64 / CHAPTER 2 = THE PRIMITIVE UROBDOROS / 65 am sactjons ‘The nex result of energy-couplings mechanisms depends iE the various competing processes. For example, thus far the iles readily available pyrophosphate bond energy thar either could heat through hydrolysis or could activate amino acids by linking, Fo form energy-rich carboxyl phosphates, Which outcome domi- nds on rare, not cquilibrium. Equilibrium favors hydrolysis of the hares. Or, hydralysis of the carboxyl phosphaces could follow the activation of amino acids, which would produce the same net result cohace hydrolysis. Rates ate che determining factor. The hydrolysis ion bond in the absence of an enzyme may be as slow as 10 second. Even if pyrophosphaces hydrolyze as often ag 1077 per vation and polymerization could dominate tat 10-' eo 10° per jour approaching the rates rypical of highly evolyed morlern en- f «0 10!" per second). Activated amino acids could have polymerized been favorable fur the activation and polymerization of monomers i ihey hydrolyzed, even though hydrolysis was the ultimate equilibrium gomers and polymers. Perhaps one of the earliest consequences of abi Fundamentally, which rate dominares is an experimental question pyrophosphate would have been the generation of mixed oligomers tha yy important details, such as ion concentrations and pH are crit- the catalycic functions of coenzymes, } Wever, the face thar Lie exists in its present form suggests thar on the When considering energy metabolism, especially such sequences Off h phosphate energy drove the polymerization of amino acids tions as those in Figure 2-3, in which five coenzymes participate, [have thar a stage of metabolism that was coenzyme dependent but virtuall free (excepr for the proceinoids) evolved before anything like com polymer synthesis arose. Although réaction rares in such systems wor much slower than in contemporary, enzyme-catalyzed organisms, che easily be grearer than the hydrolysis rates for small mixed oligamers (¢6 with the discussion of peptide hydrolysis in Section 2-1), With P ~ Be ADP APS s03- re) Suliate S08 Sulfite FIGURE 2-10 Sulfur oxidation in Thisbacillas, APS is adenosine-3 fi Sut cates Tate APS 13 adenosinc-3 “phasphosulfate, alsa called active Genetics in the Microsphere f proteinoid can yield 100 million to a billion microspheres; ion with the prebiotic environment were myriads of natural ex- Tam proposing thar in some of these experiments the production driving this relatively simple system of reactions, a set of coenzym te bond energy was sufficient to polymerize amino acids, perhaps have evolved that would possess enough functional diversity to pron he proteinoid involved was catalytic for peptide bond formation (Sec- connected sysrem of reactions that also could regenerate the coenzyn 21), If some proteinoid could promote nucleotide polymerization with would be uroboroslike behavior, but withour genetics 3" phosphodiester linkage, then small oligoribanucleaeides (short chains evolutionary, Veould have formed. Speculative detail about these steps is precluded e, bur now the stage is set far the development of molecular ge- ‘2 primordial level: Pyrophosphate-cnergized microspheres could id amino acids and produced short oligomers of RNA.* Phong hat . P ‘ext four subsections, | present speculations on how these micra- aw ? ? re could have developed a primitive genetics having the capabilicy of gel Aves tO a more complex system. Moreover, my scheme will suggest the ch many features of che contemporary mechanism could have their oO oO iS pri ee ape ‘. : o T feel SQi 2FeUM) 2Fetily H primordial one. 1 present these speculations partly because I feel Sule Acaive sulle Prrophosphite FIGURE 2:11 P : f V flcto- explanation of this circumstuice has the same sort of impact chat the Coupled sulfite oxidation and formation of high-energy P ~ P, in proteingid microsph 4 Excited stare of the °C nucleus had for stellar alc ahetis tsee Section 1-1) Step (1), oxidative phosphorylation of sulfire c6. fo ive sulfare. Step. hysieal H oe : em active sulfate. Step (2), phos BE Chemical law . The niversal laws with the 7 of phosphate to farm P ~ P, wich the release of sulfare. Standaed axidation-reduction Be acct; rere x te banttciog of univers ich the particulars porcntias for thete reactions at either acid or bas i : i : avaiable iaigonsing BOD, cid or basic pH confirm thar sufficient ened Ais Section 2-1) does not apply ar this level beenusé’ Ie tequires's special, very66! CHAPTER 2 genetic apparatus, though complex, is not so complex that plausible for it cannot even be conceived. A current view exists that this pro origin is insurmountable and chat therefore life muse have originated where else where conditions were different (Crick, 1981). [reject such as defeatist and unimaginative. Primitive Polypeptide Formation Once again, che aext step in forn this model derives from the contemporary mechanism: In all known gene-directed protein biosynthesis today, after ATP activates: amin tRNA synthetase immediately places them on their cognate tRN, 2*—OH carhoxyl-ribose ester linkage so formed is depicted in Figu This ester bond is energy-rich to the same degree thar the precursot acyl adenylate (the activated amino acid) was energy-rich. Thus the subs formation of a pepride bond becween two activated amino acids dynamically favored, In this model of primordial gene-directed protein biosynthesis, a that RNA served both aru primordial gene and as a printordial messeng : No DNA was involved, nor were ribosomes, tRNAS, or (RNA syni which evolved later. First, pyrophosphate would activate the amino proteinoid microsphere to become amino acyl carboxyl phosphates. would reace with the 2'—OH groups of the RNA, forming amino acyl | esters (carboxyl-ribose esters), as in Figure 2-13, step (1). Unlike d case, in which these esters form only at the 3! end, in chis madel th form on any of the 2’—OH groups in the RNA. A modern RNA mt usually has cations (positively charged ions), such as Mz", bound to ie phate groups, which are negatively charged; in this model, when an ester formed, the positively charged amino groups moved close to , ot rc i a FIGURE 2-12 a Carboxy!-ribose ester of RNA, contempotaty The 3' end of a -RNA is shown, wi * al 2'—GH amino acyl ester at che top. Be z i a Carbonyl phovphate: ANA ‘Aina 3671 RNA 3! end FIGURE 2-15 ion. E : ackl-RNA esters for pe} a < z z 5 : é 3 & 3 2 3 ge ae ae 5g ef at we Le fe ze aS 20GH) CHAPTER 2 THE PRIMITIVE UROBOROS J 69 th its third-base redundancy, called wobble). A list of the firse two ‘all codons of modem RNA codes for nearly all of the 20 amino acids ‘ology appears in Table 2-4. When tesearchers are able to produce k i physicochemical conditions for this model in the laborarory, chen Efe cothesis of the origin of che code, particularly the specific code in Figure an be tested experimentally. _ / “hypothesized coding mechanism offers some specificity of interaction cn activated amino acids and the base sequences of RNA. This specificity or possess che fidelity af che contemporary coding mechanism because re among the myriad microspheres, which sel-assemble so easily, & not the biological sense. It need only favor one amino acid slightly more y than another. For example, lysine might bond between A and A only cent more often than histidine, and visa versa between C and A. "This luce a mechanism with a relatively high mutation rate (mutation is | would creare the possibility among the billions of microspheres for lution. phosphates, and displaced the meral cations, When enough esters formed. 4 conformation change occurred, as in step (2). The esterified RNA would fon a right-handed helix of three nucleotides per turn, with the phosphates ont inside and che bases-and amino acids on the outside, When amino acids. terified to every third ribose of RNA, they would be adjacent to each other. in this helical conformation. Moreover, they would be adjacent in the sei that the amino group of one amino acid was adjacent co the ester carboxyl o| the next amino acid located three riboses away in the 3' direction, In thé uncoiled conformation of RNA, amino acids that are three riboses apart roughly 20 angstroms (A) apart, and even amino acids on adjacent ribose: about 7 A apart; these distances are too great far any chemical bonds to form, Bur in the model, the amino and carboxy! groups of some amino acids would be couching in the helical conformation and could! then form the energetica favorable peptide bonds. benween them. This pracess would continue until amino acids, three riboses apart, were bonded into a polypeptide, albeit a sh one. To test the plausibility of these steps, | have buile CPK and skelet molecular models that accurately reflect che distances involved, Wher the polypeptide was complete and left the RNA, the RNA wou revert ta its nonhelical conformation and could serve as a template for firth protein synthesis—or for ies own complementary copying (to be descril jtion from No-Life ta Life Once microspheres existed thar were ‘of making RNA ard then small proteins on this RNA, a particular below), The separation of the growing polypeptide from the RNA might ha TABLE 2-4 involved nothing more complex than spontaneous hydrolysis of the connect pierre Gove carboxy! ester linkage.” Doubles codon Aine cecil The base spacing in the model (like-ribose spacing) would be three ba “* os WD per amino acid, identical to the contemporary coding system, excepe tha cA ysine the model ir would be a consequence of che special helical conformarion. AC Threonine most plausible argument (Eigen, [97 1) given for the three-base cade AG Serine of arginine contemporary mechanism is that two hases are not enough ca code for ‘ - aminoacids (4 x 4 = 16 codons), and a two-base hydrogen bonding of (RN AL Isoleucine to mRNA is not strong enough te function smoothly or maintain specifiel CA Histidine On the other hand, a four-base code, although more than rich enough’ ce Proline account for the 20 amino acid alphabet (4 x 4 x 4 X 4 = 256 codons), : co cao strong a bond berween © and mRNA and thus ce too slow am ce Arginine anism. This leaves the three-base code as the best form. cae cu Leucine But where is the coding in the primordial model? How could a spe “ ino acid bond (by esterification) a a specific base sequence on the R GA Aspartate or glutamate U hypothesize thar the esterification of an amino acid beeween ewo ad! Ge Alanine hases (for example, B, and Bz in Figure 2-13) is selective, depending 011 we Glycine affinities of the owo bases and the amino acid residue R, This would const Gu Valine UA End ("punctuation") 0, whereas for a thermally (As stated above, P AV is ignorable.) Therefore, AU + PAV (and constant?) system the second law stares that AG <0. Bialogic: @ case; instead, AU + PAV ~ \). are thermally buttered, Experimentally, AG < 0 is always true for + PAV ~ and AS, <0; s0 AG = AU + PAV - TAS assembly of a complex from its dispersed subunits. Nevertheless, it will be greaver than zero (—AS, > 0 if AS, < 0), But this ‘fra yet include AS, in AS, which is the key cerm because during dispersed subunits (initial state) to ageregation (final state) appears to! a the displaced wazer molecules gain translational entropy, making the direction of greater organization, or increased order. As noted at d lécatise several warer molecules are usually bound to every subunic, of Chapter 2, increased order means an apparent decrease in eatropy AS, |, The overall effect is that AS > 0; s0 AG < 0 because of AS <0} but this is nor necessarily in conflict with the second law b translaconal entropy of displaced water molecules, Thus the biological systems are not isolated. Mereaver, AG < 0 and AS < 0: peess is “entropy driven,” that is, driven by the entropy increase patible provided that Free Energy and Entropy Considerations for Self-Assembly Let the initial stare for self-assembly be dispersed subunits, and lee: state be the aggregated complex. The free energy change AG is defin AG = Grint — Grnicial |au+Pav[>ras OLECULAR MORPHOGENESIS AND CONTROL and MECHANISMS : aly does not require added energy, Once the subunits are synthe- lermal motions bring them together, whereupon they sponta- Y means of multiple specific weak bonds. The structures they AU + PAV as compauridiiAd the ehaccnic snleeuiia OF Oe elements decane Boro Phosphate, sos to ridiitest repeooable ATHADE te of their properties, including che suitability of phosphorus for energy Bp cecil the available-fod source. This. prestimahly, tt wet olism. However, given che primordial dozen elements, many of the abse characteristics of the living stare cannot be predicted from first principle current description of the living state is largely ex post facto and is hig by emergent properties. lular eukaryotes, the purpose of the phosphageas is clear: 1d Cr—P are energy storage molecules thac are quickly available for shorolyzing ADF into ATP. etic Distribution af Phasphagens Researchers have found large of polyphosphates in protozoa in the form of granules thar are com- ted with DNA fibrils and cellular zones rich in ribosomes. In ir eukaryotes, such as yeast and fungi, polyphosphates may constitute nc of the dry weight. In both prokaryotes and unicellular cukaryores, S Serve severa. functions, but evidence exists in both cases f 3-4. THE EVOLUTION OF ENERGY METABOLIK AND STORAGE Major transitions in form and behavior have occurred im the cours evolution of life on earth: transitions from unicellular to multicellar life Primitive heterocraph to autocroph co modern hererocroph, from anaerove © Of as teady sources of high-energy phosphate—that is, as phos- photoaurotroph to aerobe, from sessile organism co motile, and from a9 BS. Only micoorganisms use polyphosphates as phosphagens, whereas ofganism to terrestrial, (See the geological time scale, Table 2-3.) These animals use Arg—P and Cr—P, The difference berween Arg—P sitions are so well delineated chat this section will consider them in term (CrP users is illuminating. Superficially ic appears that invertebrates, metamorphosis at che phylogenetic level, rather than in che usual bio Paand vertebrates use Cr—P, but closer inspection af the evidence context of ontogeny. [n every instance, the metamorphosis is directly. atper distinction: Prechordata use Arg—P, and Chordata use Cr—P. i energy metabolism, regulation, and storage: | lata include all vertebrates and the so-called Protochordata. Pro-{BLY AND CONTROL | 96.) CHAPTER 3 ‘ ' lin tochordata comprise three classes: Tunicata (sea-squirts), Enteropneuista¢ ‘Chordate Hine worms), and Cephalochorda (Jancelets or amphioxus), Acorn worms, the Primitive invertebrate chordates, have a nervous system that is che mog tive of any group of animals having an organ-system level af constey Three anatomical structures distinguish Cr—P users: (1) a notochord tilagenous rad extending the length of the body and supporting the sof (2) a dorsal tubular nerve cord; and (3) pharyngeal gill slits, Embry, Protochordata and Chordaca posses these three structures. The Precha which do not possess these structures, include all other multicellular tebrares, such ay Coelenterata, Plaryhelminthes, Nemertea, Mollugea, thropada, and Echinedermara, ‘ FSA triminve Biochemical analyses of many phyla and species seem to indiea mesoderm cell Arg—P, asa source of quick energy, supports muscle activity and rud nervous activity, whereas Cr—P, which also supports muscle activity, i ¢ sential for higher level nervous tissue activiey. The notion that thes phagens are prerequisites to muscle and nerve activity follows from thei tions: The cnergy demand of muscle is so greac chat glycolysis and transport cannot regenerate ATP fast enough to supply it, Mammalian can work.ar a rate that demands LO”? moles of ATP per gram of mi minute. The ATP available per gram of resting muscle is 5 x 10-& is sufficient for only 0.3 5 of activity, The phosphagen stores, however, a rapidly regenerated source of ATP chat can last for several minutes, The evolutionary perspective is thar early multicellular species, pr worms, were able to use phosphagens (originally stored merely to ATP/ADP ratios while accumulating abundant energy from food sot energize muscle tissue for rapid motility. An evolucionary relationship: to exist among the muscle proteins (actin and myosin) and relate thar confer a cytoskeleton and motility on unicellular forms, Muse natural development in multicellular forms, but its use is possible only if PRO MG calls Giese HAE phagens are available. The sequence hypothesized here is that abundantebt eegee and vascular system: ecto Se eee cas Gerais) in the environment leads to energy storage in phosphagens, which eh Bee uct pe dietn one encelb a ee eae to cytoskeletons and simple, cell movement and to che appearance © aris oe one derive mosdy from ncural cresticells, which beaia tissue, both of which conferring evolutionary advantages. Beye Bi rlgreshascfaimed teomectadermcell velopmental evens requite cell motility within the developing gas- this requires phosphagen. Why Cr—P leads to one owrcame and nother is unknown. Nevertheless, with Cr—P, sophisticated ner- Protected by bone, evolved. Blastopore: FIGURE 4-8 Arthroped and chordawe mesoderm cells, odermate, which are invertebrates, might appear to be ia the .¢, bur evolutionists view them as the link between other in- d the Chordata. Embryologically, they are so similar to acora early specialist on echinoderms described the Larva of an acorn chinoderm larva. However, the acorn worm develops into a true whereas the echinoderm larva ends up as a starfish or sea ur- ind Cr—P as phosphagens, thus they are transitional organisms hagen viewpoint as well. Embryolagical Considerations Embryology strengthens the pret views. Motility, which began in unicellular forms, plays a crucial’ rol development of a multicellular organism, especially during the early s embryogenesis when blastul and gastrula form, Only ewo lines of em ogical development exist in evolution: the arthropod line and the chi line, These ewo lines differ in the way that mesoderm cells form in the: stage (Figure 3-8), In the arthropod line (including flarworms, neme! mollusks, annelids, and arthropods), the primitive mesoderm cells bud | primitive endoderm cells of the blastula. In the chordate line (inclt Chordata and the Echinodermata), the mesoderm cells’ bud off as €6# acs from the endoderm, Refinements and Chemistry of Phosphagens About the ime lordate and the arthropod lines separated from a common ancestor, like organisms made up the arthropod line: true arthropods and later from ancestral worms. The trigger for this separation into ines may have been the emergence in chordaces of C-—P in place P- At this time, evolution may have tried other phosphagens as well.98 } CHAPTER 3 Prechordares: Arginine phosphate CH.—CH;—CH—CH—COuH NH» Chordares: __Crewsine phosphate —CH—COH Pélychetes: oglycocyamine wnnnnvnsnnnteng NH—CH.—co—u Phospholombrigine o | NEC CH EG -0-F 0-08 OH NH, Phosphotaurocyamine Octopus: FIGURE 4.9 Phosphagens using the energy-rich —N~P bond. SELEASSEMBLY AND CONTROL / 99 1, researchers hiaye found several other phosphagens in the annelid fed polychetes. Even the octopus, a mollusk, has its own unique — Chemically, all these phosphagens are similar to Arg—P and ure 3-9), and cack involves the energy tich —N~P bond. Such are effective donors of energy-rich phosphate for conversion of | Struccuce Correlates cleton of cells is based on proteins thatare relared, or in some tical, to the muscle proteins actin and myosin, Their energy comes and cyclic-AMP and calcium trigger the kinases that regulate their early multicellular life forms were worms (flat worms), which rageous motility with their muscles because they stared energy Icium is a component of the worm's muscle regulatory processes. hropod lineage, the earliest mollusks that evolved were lictle more with protective stells of calcium carbonate, true mollusks evolved sk shells have formed deposits of limestone nearly everywhere on calcium carbonate appears to be a natural product of a calcium- ergetic process coupled ro a metabolism that produces the waste ms ate aerobes which produce CO, as an ‘of energy metabolism te lineage uses Cr—P, which may have provided an enbanced share utilization. Not only did a more sophisticated nervous system ently correlated with Cr—P, but nervous tissue used more cal- Mineral crystalline ‘form Organic mtr Cell walls Calcite (CaCO,) Cellulose pectins, lignins Exoskelecor Silica (SiO) Pectins Exoskeletor Calcite, aragonite Pistein (€aCO,) (conchiolin) Exoskeletor Calcite (CaCO) Chitin, proteins Endoskeleton, Hydroxyapatite Collagen bone, (Cajq( PO (OH) 2) cartilage, tooth100} CHAPTER 3 SELF-ASSEMBLY AND CONTROL | 101 cium and phosphate, both a8 control substances (Ca?* and Phosphate} V.,and ¥, Robin, “Eiseribution of Phosphagens in Erranc and Sedentary active ion fluxes {Ca * ). Muscle contains substrate-level amounts of boy H in Steatier a Een peetie Dietbon ho edited by K. A, Munday, Per- and phosphate because the protein filaments of actin and myog vss, New York, 196+ and pheeehs ate sites about fey 400 A along theit lengths. Thos Be Sister and E. L. Smich, . Priapl of Biochesetstry, Sed edition, MeGraw- of calcium phosphate as bone seems ro be a natural conseque York, 1964, pp. 749-751 metabolism. Calcium phosphate is a stronger material than calcium Table 3-1 summars $ the occurrences of skeletal structures in orga REFERENCES Section 3-1 Fox, R. F,, Binlogtca! Energy Transduction: The Uroboros, John Wiley, New Yor Nomura, M., "Assembly of Bacterial Ribosomes,” Scfewce 179 (1973; 864, Schrodinger, E., Whar w Life, Cambridge University Press, Cambridge, 1944. Section 3-2 Riadt, K. P., "Expression of Isozymes and Their Function in Differentiat Differentiation, edited by L. Nover, M. Luckner, ind B. Parthice, Spri Berlin, L982, Section 3-3 Adelscein, R. 8, and E. Eisenberg, Annnal Reviews of Biocbemisery 49 (1980) deDuve, €., A Guided Tour of the Living Cell, WH. Erceman, New Ys Chapter Eisenberg, E., and ‘TL, Hill, “Muscle Contraction and Free Energy Tra sd Biological Systems,” Setence 227, (1985): 999-1006. This paper cont d account of hew the muscle proteins, actin and myosin, are arranged in how ATP causes contraction of the muscle. Harold, F. M,, The Viswl Force: A Study of Bioenergetics, W. H. Freeman, 1986, Chapter 11. Kemer, J,, “Variational Principles in Nonequilibrium Thermodynamics, (1977): 219-226. Klee, C. B., T. H, Crouch, and P. G. Richman, Annual Reviews of Bio (1981): 489, e Robin, C, §., and O. M, Rosen, “Protein Phosphorylacian,” Annual Re chemistry 44 (1975); 831, Schramm, M., and Z. Selinger, Science 225 (1984): 1350. Section 3-4 Baldwin, E., Dymarnre Aspects of Bioebemistry, 4th edition, Cambridge Unis Cambridge, England, 1965. Buchsbaum, R., Animal without Backboner, Universicy of Chicago Pres 1948.CHAPTER The mathematical character of driven, ipative systems is the main copic of this chapter. The important inear dynamics fer chis discussion is that the equations used ro systems do not usually possess closed-form solucions and must Pmodeled. Computer simulation demonstraces chat even very sim- this type can show extremely complex and sometimes chaotic ehaviors of such systems are reminiscent of the behaviors of Bur Shue: of such aystens provide iavtalicinto life's evakiGon.ie ‘ CS i 105 lad) CHAPTER 4 NONLINEAR DYNAM! that the phenomenon is explained by a system of reaction-dif The chapter begins wich aq example: Turing’s mathematical a cai i n ord fons; which describe the space and time variations of chemical description of the embryological formation and differentiation : tion 4-23. This is followed by discussions af mathematical Jimena they move through space by diffusion and engage Hishenees - 4-3), numerical solutions for a simple Turing model (Section ed), aay Beiaring’s iden wes chee reaction diffusion equations govern the sbice fries oF chaotic attractors (Section 4-5), Harmonic ostllacor and dj ne movement of morphagews, chemical species in a growing the Pendulum systems demonstrave how the behavior of dilferential equate cific developmental events and thereby control its morphology, change with dissipative perturbations (Section 4-6), Section dct ak amples of emergent behavior and dynamic chaos, and Sections 4. discuss models thar demonstrate nonlinear dynamics, especially in the of predictability, Chapter 4-10 describes the usefulness to organisms simulation and concludes that che evolutionary sigaificance of the beg it has become @ rapid simulator of nonlinear processes Mathematical termiaology and mechanisms abound in Sections 42, 49, Readers who feel uncomfortable with the mathematics can use, sections for later reference and sc on co Section 4-10, However, any has a rudimentary programming ability (or a friend who does) could: work through Section 4-9, ; Difficulties with the Mathematics reaction-diffusicn equations are coupled partial differential equa- J concentrations of the substances involved, each concentration ng a function of chree spatial coordinares and one temporal co- ithough the diffusion terms are linear in these variables, the re- ‘e generally, at least bilinearly, nonlinear. Therefore, the: task closed-form solutions of such systems of coupled partial differ- tions is formidable at best and usually is impossible. Mathematicians omputer simulations for such problems, but these simulations dy limiced by the size of problems thar computers can accommodate er simulation must use discrete coordinates to represent the four pace and time coordinates, and as few as one hundred discrete of the four coordinates results in 10* discrete coordinate sites, nae “ai known for every one of the Energy flow is a necessary bur not» sufficient condition for the living $4 oe Hee ic heme Ue Urabecihes by codes of matter since a Living tae also evatves, The conditions of energy f ee sogrosinaces ibe ayuem wf comnugur rel capacity to evolve are both satislied by phosphate eneray uransducde DPA cnc roge sophoticaced ccoderh comibieces easier acting with proteins and polynucleotides; thus the living state as Be aiibles conveniently. Nevertheless, progress in this ap- consequence of special substances and their emergenc properties a en rd waeahet processing and new computer consequence of energy flow, The preceding chapters have empha ae rove’ P substances and their emergent properties; this chapter emphasizes $01 eval principles thar have biological significance. From the viewpoint fF and marhematics, the study of energy flaw is an active research general ideas still evolving. Much of the work on energy-driven § volves dynamic descriptions given by aonlinear differential equatiot not be solved in closed form in terms of standard functions. Serio! such equations became possible only with the advent of powerful, puting machines. However, any reader who has access to a micro can explore some of the behavior of driven nonlinear systems using programs given later in this chaprer, 4-1. THE PHYSICS AND MATHEMATICS OF ENERGY-DRIVEN SYSTEMS led the difficuity of analyzing coupled partial differential equa- king acellular view; thac is, he replaced the continuum by a tissue laced partial differencial equations in x, », 2, and ? by ordinary uations in f alone. His motive was the cellular structure af tis- necessities of computers although his approach anticipates such irably. azen concentrations X and ¥', the partial differential equations 4-2, REACTION-DIFFUSION EQUATIONS FOR MORPHOGENESIS a 2 ae ME A) = FOX, YC) + DV? XK, Fi A mathematical approach to understanding embryological morphoge mation and differentiation of tissue) was attempted in 1951 by Alan in his classic paper “The Chemical Basis of Morphogenesis” (Turing: a £) = e(X(r, V(r.) + DV? Yor, 4)106) CHAPTER 4 NONL SEAR DYNAMICS / 107 where D and D’ are diffusion constants, and f and ¢ are arbitrary, even nonlinear, functions of the morphogea concentrations. Fora ring assembly of cells, using Turing's cell index r = 1, 2...., Ny the sy: ordinary differencial equations equivalent co Equation (1) is aginal disk of Drosophila melanogaster using such a mechanism, In the investigators obtained che closed-form solution of coupled par- tial equations. With appropriate choices for the values of rhe dif: fants and reaction rates, the homogencows sparial sate is unstable ‘perturbations of appropriace wavelength. As a result, spontancous d fev B . ‘of spatially patterned states occurs, This mechanism prediccs. che Bir den TA Biker Be He tera of development found for the wing disk when the disk is Dies | Db a growing ellipse containiag eso morphogens X and Y. The actual Girt Xa Vd + ey — eK fepuities of X and Y’ are nor required tor known) for this analysis. # ideas led to the concept af sel/-organization in a spatial system. If J and instabilities of the type just wiscussed occur, thermal wheter = 1,2)... anid /is the center-to-cenrer distance berwaent peur ene re ons favor spatially parcerned states aver homogeneity. That is, par- mstances can produce a general resule. He fully appreciated chat f instability occurred in driven systems and that ene of che mor- (say, X) had to be aurocatalycic and one (say, ¥) had co inhibir the 1¢ aurocatalytic morphogen X reacted by converting chemical input cells. Turing’s primary goal was to study the anset of instabilities in the eq an approach characteristic of many studies that followed. He investigate ple geometric assemblies of cells such as rings, sheets, and spherical He was able co identify six ways for instability to arise in rings. instability was a transition from homogeneity and symmetry to 9 broken symmetry. (The cerm instability is perhaps too strong, both in work and in later work by others.) Whar happens is that when the investi inafluid. Would an elliptical sheet of cells ever develop a swimming that is, a pattern in space and time thac resembles che swimming pattern of some sort. The old state becomes unstable, whereas the can be very stable unless the parameter is altered again. Further al ‘disks. The model of the flaeworm might be an elastic elliptical Je cells containing Turing’s two morphogens, XN and ¥, which ‘ : the twitching of contracting muscle cells. The sheet would be immersed stability of oe behavior in Space and rime. Sequences of sushisty ad would behave according to hydrodynamic equations and to the ly occur as a parameter changes, Turing claimed co understan Riba cqustiond of the morphogens, Morphogen X triggers cans a by using a spherical shell of cells co representa blastula, which ‘muscle cells when NX is sufliciently concentrated, The input con- spherical symmetry. A breakdown of this symmetry occurred und could be a food substance distributed in the medium outside the , could conditions 2¢ such thar thermal Auctuations would trigget ation in which the elliptical sheet “swims” through its food? This f problem Turing's approach addresses, and che next generation processing in digital computers may make numerical simulation metry. Turing concluded his analysis by suggesting thar digital would be useful, ‘Turing realized thac che wave patterns he saw were the result of "a supply of free energy.” Calling A the substance with greatest free em B the substance with least, he wrote that “Energy for the whole pi obtained by the degradation of A into 5." Substances G are cataly schemes and govern che rates at which the morphogens X and ) chemically. Turing’s models involved a single mechanism. For appropriate diffusion constants, reaction rates, and initial substance concen! stable hamogencous spatial state can be a solution to the equation situations, For example, perturbations of wavelengths outside some: range around a characteristic wavelength fy decay away, returning (07 tially homogeneous states or perturbation of wavelengchs within th range around /y grow, yielding a spatially parterned final state. Stuart Ae man and others (1978) achieved a derailed account of the embrye, deas extend the consequences of energy flow from the macramo- vel described in Chapters 2 and 3 to the tissue level of multicellular 4-3, DIMENSIONALITY AND INSTABILITY fatical description of molecular evolution must take into account a ‘Aumber of variables (chat is, it has many dimensions); but in some dynamics is governed by oaly afew (L, 2, or 3) dominant variables as few dimensions). These cases of small dimension are the best1OB ¢ CHAPTER 4 NONLINEAR DYNAMICS { 109 understood mathematically. Although the dimension of a simple mo that of w realistic model are vastly different, some features appear ta be, For example, self-organization of a morphogen partern is conceptually. to the self-assembly of proceins. Both exbibic a tendency away fon geneity coward asymmetric patterns, and both require an energy inpup, assembly, the energy inpue results in protein synthesis; in self-organiza chemical input results in an autocatalytic reaction step +) begin with three dimensions: xX = f(X,¥, 2) X,Y, 2) o Z = AX, Y, Z) " of behavior have biological importance chat is discussed in Chapter on general themes that dynamic systems exhibir. Turing used two morph because the generic reaction-diffusion behavior of a single morph centration is stable uniformity (that is, the homogencous, symme stare (equilibrium) is asymptotically stable), To produce an instability, tical Chaos at least twa morphogen concentrations. This difference in behavior of dimension (number of morphogen concentrations) js a provabl matical theorem. q Dynamic systems of ordinary differential equarions exhibit the dimension explicitly: The one-dimensional case is city of trajectory orbits is characterized by numbers called Lia- . Given an initial phase-space point (X., 5, Z,), the rate at boring phase-space point (X,Y, Z!) separates from the initial two trajectories evolve in time, yields the Liapunov exponents h dimension. The separation along the ith axis grows like ame exponent). Ifo, < 0, then nearby poiats approach each other jin time; if o, > 0, then nearby points separate exponentially. terized as the occurrence of at least one Liapuney exponent in zero, Section 4-7 gives detailed examples of chaos. ponents also provide a measure of the rate of change of a volume © points as the volume moves through phase space according ta of equations, The rate of volume change is denoted by Ag. If pace volume grows and becomes asymptotically unbounded. n the evolution is a""contraction,” which is typical of dissipative volume grows Lke e**', tic states of interest are chose occurring in dissipative systems. teast one /, ¢; > 0, and Ag < 0, The conventional numbering Liapunoy exponents is ro start with the largest, that is, x FLX) is, they asymptotically approach a fixed limit withouc os derivative with respect co time never changes sign). The proof of the preceding claim follows as a special case of a gene inn dimensions. In dimensions, a phase-space point has » coordin: pled first-order (having only first derivatives with respect to time) differential equations yield trajectories that never self-intersect. I tory did intersece itself, then two different trajectories could use section point as the initial value, which contradicts the existence of solution. In one dimension, chis implies monoronicity; that is, the cannot reverse its motion in one dimension fora first-order equi dition, many second-order systems (having second derivatives with ume) can be recast as higher-dimensional first-order systems. issipative systems require ¢, > 0 and Ay < 0. In cwo dimensions che equations are m1 >ox> oy > (6) X= fx,y) =aX,¥) Ay = 3 oo; a ms OE eg Xe GM XG os el, Using this identity to see an These equations peri Consequence for dimension requires introducing one more tech- nificance of two variables in Turine’s theory. Suill other cypes of dynamic behavior are possible in higher dimensle! of w couplec, first-order ordinary differential equations can dasaw — 1 dimensional Poincaré map. Bounded motions inNONLINEAR DYNAMICS ¢ LIL MO 4 CHAPTER 4 ‘TEMPORAL DEPENDENCE IN A DRIVEN SYSTEM: A TURING MODEL IN TIME n dimensions are of interest. A dissipative system, withouc inp; decays to equilibrium. With inputs, ic may have stable steady stares, have stable time-dependenc states instead of stable homogeneous § motions are to remain bounded, then some-varialles must be oscilla erally nonharmonically) abour some fixed value. The needed analy nique is as follows: Observe the system, choose an oscillating vai choose a reference value for that variable in the range of its ascillarioy every time the value of the oscillating variable passes through the value, record the values of all the other variables. For example, lee the: be Xa, Na,--., Ng. Choose X, for observation, and choose the: value XY. Whenever X(t) = NY, record Ny(t), X44), - 02, Ky 0, dX (Oat was either positive or negative. The record above con: for which both signs of dX fd occur, Restricting the sign co that al t = 0 results in a mapping of ( . &,,) thar is the Poinearé jon of simple two-dimensional dynamics for a system that remains ogeneous focuses attention on the time dependence. The system by the equations x= me xt = pe 18) xy — by des an autocatalyric step, the x7 term, and a driving term a. Both ¢ positive. Setting % = 7 = 0 and solving for x and y gives che ‘The prohibition on crajectory crossing discussed earlier does noe. xhab os a common aéymptotic equilibrium stare. In fact, dissipative sys a ath not driven do end up at a single-point equilibrium. All trajectories OS SE except at their final, asymptotic limit, the common equilibrium point, volume of n-dimensional phase space vanishes when all of it reaches ing the linearized dynamics at (x, 9) gives the stability matrix for librium state. In the driven case, rhe final state may be time de| may retain a finite “volume” (of lower dimension than #), a/zero: dimension » (more about this point later). ‘The Poincaré map also can be characterized by Liapunov exponent a volume (2 — 1 dimensional) contraction factor Ay. For the differential equation system, the » — 1 dimensional Poincaré map # — | Liapunow exponents, again denoted by a, > 02 > > Oyo these are distince from the o's for che #-dimensional differential flow (10) (@ —alé -b ab+b 0 Baa (e- 9) 21(-2 —w + 2) ay pair of complex conjugate values. > @, 2 perturbation near the equilibrium state will return to the ate by spiraling into ic (Figure 4-1). The spiral morion is a com- -hegative real perts of A= and the imaginary parts, which give Oscillations that look like spirals. For 6 > at, che real parts tive, and the equilibrium is nor stable. Instead, che dynamics iptatic time-dependent state. Although, in this case, the li- Ons grow exponentially without limit, the true solutions co the Hons do noc. When a/é > 4, the spirals are counterclockwise and-end up at che equilibrium point. When 4 > a/b, the $0 spirals away from (4, (a + 47)/) in a counterclockwise sense, sstOry never settles down on a single point; it continues its coun- Gunded circuie forever. Syaamics approaches an equilibrium point in a stable way, the ‘stable fixed porns for the dynamics. When, instead, the dynamics =i Ay = So A dissipative, chaotic Poincaré map must have Ag < 0 and gi > 0. step of the argument is thar for a chree-cimensional differential eq tem, che rwo-dimensional Poincaré map must yield a1 > 0 and 6 0, which is a possibility; whereas for a cwo-dimensional differenti system, the one-dimensional Poinearé map cannot yield both oy = 08a <0 I Therefore, a system that is characterized by only two degrees 0} can exhibit patterned temporal behavior without the potential whereas aif higher-dimensional cases have che potential for chat states, Biological control mechanisms may have evolved to FesP mathematical differences in dimension: Restriction to two det grees of freedom would produce contre] without risk of chaot112 / CHAPTER 4 NONLINEAR DYNAMICS / 113 aver - FIGURE 4-1 Fixed poine and limir eyele, forever produces an asymptotic time-dependent, closed krajectory, jectory is called a frm? cycle, Figure 4-1 shows these cases for th example. The limit cycle can be approached from inside or outsid upon the initial values for x and y. J Said horizontal arrows the sign of % up is + and down is ~ | Much can be learned about this type of dynamic system even thi aba. and left is —. are no closed-form solutions.* ‘The lack of closed-form solutions is no! of insufficient cleverness but of their nonexistence. Only numerit are passible, except in very special cases. In the preceding exampl form solutions exist, However, graphical techniques and numetic tations On computers make the behavior of the system entirely good qualitative understanding ¢an be achieved with graphical mel whereas quanticarive understanding is aided by computers. For é particular system given in Equation (3) can be understood qualit plotting the vector field (%, 3) on the xy plane. Inspection of | shows that the «-y7 plane can be divided into four regions in and x = y are deduced from Equation (3). The figure shows that around the fixed point (4, (# + 67)/4) is counterclockwise, Another important feature of an unstable equilibrium state in: tem is its behavior under thermal perturbations. Assume that th in the state (4, (2 + 47/6) with 6 > ald, chen any thermal flu changes (x, y) from (4, (2 + &*)!6) will cause the state of the sys away from (4, (¢ + 67)/4) and toward the limir cycle. The limit cy with respecr to thermal fluctuations although it is somewhat sm thermal noise. Simulating rhe solution to Equation (3) on a microcomputer is xts) and y(r) againse ¢ produces spiking for sufficiently large values ee Tin * This is an analogue of the nonexistence of closed-form expressions far the coots Af PS 7 FIGURE 43 of degree 4 or higher, whereas such expressions do exist for degrees 2, 3, and. Spiking signals.1d) CHAPTER 4 NONLINEAR DYNAMICS / 115 4-3), The time-dependent spiking pattern is a self-generated, eme erty of the nonlinear dynamics, even when all inputs are constants Biological spiking patterns, such as in neurons, could also arise f fundamental mathematical mechanisms, 4-5. ATTRACTORS IN PHASE-SPACE PICTU) The geometric (graphical) approach co the theary of differential (especially ordinary differential equations) has led to the concept: fractor; When a dynamic system has a stable fixed point, the attractors [io » 0 0 fixed point; and when che system has a stable limit cycle, the attra limit cycl Tone- and two-dimensional sysrems of ordinary first. ferential ¢quations, this exhausts the possibilities; but for higher richer ateractors exist. In three dimensions, the so-called strange, or chaotic, attractor apy simple example of its behavior arises in the Rossler model (wh physical ar biological significance buc is mathematically minimal), Frequency (Ue) FIGURE £4 Bifurcstions in the Rossler madel en this nonlinearity can appear in the épectrum as the presence pmonics of the base frequency. In the Rossler model, < depends term 2x, whereas x is linearly dependent on z. Therefore, if ency w, then so does x(t), whereas the product xz contributes =y ny does x r were c follows from simple trigonometric idencitics such as x + ay Hcos(2ar) + 1). Because 2(f) contains the harmonic 2u, a os [xz contains 3w and dw also. These higher-order harmonics ate b+ ax oO compared with the base frequency in the power spectrum, All for’ = 2.6 and ‘s shown in the figure. pnew happens when ¢ = 3.G orc = 4.1: Subharmonic frequencies power spectrum for 2(r). These correspond to the bifurcated the xy plane. When c = 2,6, the trajectories are simple, and Facing one cycle is T = 2m/w, where w is the base Frequency. the trajectory has bifurcated, and the time for tracing the entire aT. ‘This corresponds to a subharmonic frequency of w!2. For ¢ ithe entire doub y bifurcated trajectory, takes 47, which impl w/4, However, nonlinearity, the 2 term, will generate: super- these subharmoaics, just as before, for c = 2.6. These super- Up as frequencies such as 3/4, 3u/2, and soon. The cor er spectra for 2(t) are shown in Bigure 4-4, along with the in which 4, 4, and © are positive constants. This extensively stud shows interesting behavior fora = 6 = 0.2 and ¢ variable. Beca three-dimensional system, it is easiest to visualize its behavior © mensional plane; the x-y plane is a natural choice, When ¢ = 2.6, the projected onto the xy plane is a simple limit cycle. However, wl the projection is a double limit cycle. This doubling is called b1 ¢ = 4.1, the projected trajectory has doubled again. The trajectories in Figure 4-4, which also shows the asymptotic limit-cycle behavi the reansient approach to it for different initial conditions, which wol the figure). The self-crossing of the trajectory fore = 3.6ande =4 in conilict with the non-self-crossing cequitement of first-order, oft ferential equations because these are projected trajectories thar dong in three-dimensional space While these trajectories are doubling in the x+y plane, the behav third variable = is also changing. These changes are nicely descr poteer spectriam of z(t), which is found by taking the temporal Fourier of the correlation function for 2(1). The correlation function is (2 this means to muleiply zt) by 2( + s) and then to average over of fora given s, after transients (rapid initial changes) die out 4n jectory is on the limir cycle. The Fourier transform is then perforn respect tos. If z(t) simply oscillares, then its power spectrum sho® peak at the frequency of the oscillation. If, in addition, che equatio OM sequence continues as ¢ increases beyond 4.1, In fact, bi- ‘More-and mare frequently for smaller and smaller increases 80 many bifurcations have occurred that the x-7 plot looks it bwo-dimensional region of the x-y plane. The corresponding for 2(1) shows noisy contributions from an apparent contin- ities (Figure 4-5), icles in figure 4-4 look like one-dimensional curves in a cwo- in. In figure 4-5, the arcractor seems almost two dimensional. crease in attractor dimensionality is characteristic of actraccors