Selforganization of Matter and the
Evolution of Biological Macromolecules
Manfred Eigen
Die Naturwissenschaften
(The Science of Nature)
1971
Manfred Eigen
• Born: 1927 in Bochum, Germany
• PhD: University of Göttingen, 1951
• Dissertation on the specific heat of heavy
water and aqueous electrolyte solutions
• Max-Planck-Institut für Physikalische Chemie,
Göttingen
• Nobel Prize: Chemistry, 1967
• for his studies of extremely fast chemical
reactions, effected by disturbing the
equilibrium by means of very short pulses of
energy
• His works range from the thermodynamic properties
of water and aqueous solutions, and the theory of
electrolytes, through thermal conductivity and
sound absorption, to fast ionic reactions
I: Introduction
• The question of the origin of life appears as a question of “cause and effect”
• In molecular biology: Which came first, the protein or the nucleic acid?
• Protein = “function”
• Nucleic acid = “information”
• “Function” cannot occur in an organized manner unless “information” is present
• “Information” only acquires its meaning via the “function” for which it is coding
• It is a closed loop system
• Although the line from which the loop is formed must have originated somewhere, the starting
point will have lost all its importance as soon as the circle is closed
• Theory of selforganization is required to solve such a problem of interplay between cause and
effect
• Molecular selforganization includes many random events without instructed functional
significance
• What really matters is how certain such random events are able to feed back to their origin and
thus themselves become the cause of some amplified action
• This may build up to a macroscopic functional organization, which includes selfreplication,
selection, and evolution to a level of sophistication where the system can escape the
prerequisites of its origin and change the environment to its own advantage
The cell is a self-replicating entity
Nucleic acids and
proteins are intimately
linked together in their
reproduction cycle
How did the origin of life start?
• At the “beginning” there was molecular chaos; no functional organization among the immense
variety of chemical species
• Origin of life must have started from random events
• Information theory: instruction requires information
• Information content:
• Selecting a situation with Z1 out of Z0 possible
outcomes
• γ types of digits to sequences of v
• Biological macromolecules contain an enormous
information capacity
• Nucleotides: γ =4
• Amino Acids: γ =20
How big is this information capacity?
Comparison: age of the Earth
= 1017 seconds
Conclusion: The information
capacity of DNA/RNA/Protein
is incredibly large
Introduction of a value parameter
• Information alone is insufficient, there needs to be a “value” parameter
to characterize the level of evolution
• “Valuation” characterizes the degree of selforganization of a functional
order and defines a gradient for evolution
• Information gains value by selection
• Example:
• Protein of 100AA residues has 10^130 different choices of sequences
• Nature is sophisticated: utilizes parameters such as structural
advantages to affect the mechanism of valuation
What properties of matter are required to start self-organization?
Phases of evolution:
1) A prebiotic chemical phase
2) The phase of self-organization to replicating individuals nonliving living
3) The evolution of individual species
• “All authors agree on the conclusion that the essential building stones of biological
macromolecules – amino acids, energy-rich nucleoside phosphates such as ATP and its base
homologs, as well as many other biochemical compounds – could form, where required,
and polymerize under prebiotic conditions”
• Catalytic function in combination with various feedback mechanisms causing certain self-
enhancing growth properties of the system are one of the decisive prerequisites for self-
organization
• Environmental conditions are also required for self-organization, as autocatalytic growth
cannot occur in completely or nearly equilibrated systems
• (Equilibrium is a state of maximum entropy)
Thermodynamic aspect of evolution
Chemical thermodynamics: the study of the interrelation of heat and work with chemical reactions
• The energy of the universe is constant.
• In any spontaneous process, there is always an increase in entropy of the universe
• The entropy of a perfect crystal at 0 Kelvin is zero
State functions:
• Internal energy (U): refers to energy contained within the system, while excluding the kinetic energy
of motion of the system as a whole and the potential energy of the system as a whole due to
external force fields. It keeps account of the gains and losses of energy of the system
• Enthalpy (H): a defined thermodynamic potential, that consists of the internal energy of the system
(U) plus the product of pressure (p) and volume (V) of the system
• Entropy (S): is a measure of the number of specific ways in which a thermodynamic system may be
arranged, commonly understood as a measure of disorder
• Gibbs free energy (G): a thermodynamic potential that measures the "usefulness" or process-
initiating work obtainable from a thermodynamic system
II. Phenomenological Theory of Selection, II.1 the concept of
“information”
Depiction of the “information box”
Upper box:
• Energy rich and energy deficient monomeric
digits can flow in and out
• Polymeric sequences representing
“information” are assembled and
decomposed
• Both may be enzyme catalyzed processes
• Steady state is maintained via a control of
fluxes/concentrations
Lower box:
• Equilibrium is represented, involving
fluctuations +/- n
• In thermodynamic equilibrium there are no
net macroscopic flows of matter or of energy,
either within a system or between systems
• A system will spontaneously evolve toward
thermodynamic equilibrium, the configuration
with maximum entropy
Relating the information box to evolution
• It is theorized through a series of phenomenological equations, that each mutation leading to an
increase in the selective value corresponds to a negative fluctuation of entropy production,
indicating instability of the existing steady state
• Evolution at constant flows corresponds to a sequence of such instabilities, in which the
dominant species im die out in favor of the new species im+1 according to a finite selective
advantage
• (WFm+1-WFm) where WF is derived as a selective value
• This can be compared to mean productivity (E), where W > E **** (many additional equations
and parameters are used to derive this comparison)
• Whenever a mutant with selective advantage (W > E) occurs, it will inevitably outgrow the
former distribution
• This gives a physical definition to the Darwinian term “fittest”
• Darwin states “This preservation of favorable individual differences and variations, and the
destruction of those which are injurious, I have called Natural Selection, or survival of the
Fittest”
• He notes a physical origin: “It is no valid objection that science as yet throws no light on
the far higher problem of the essence or origin of life”
• Conclusion: Fitness is dependent upon constant fluxes of entropy, leading to instability and
selective advantages, leading to evolution
Stochastic (random) approach to selection
So far, selection has been treated as a deterministic process
Limitations:
1) The elementary process leading to a specific mutant is non-deterministic
2)The growth process itself is subject to statistical fluctuations
What does selection look like at steady state?
Probabilities in the steady state model (modified Pascal
triangle)
• K = redundancy in each copy
• Each step represents a change of the population by one
with an equal chance for addition or removal
• Chart represents dissymmetry brought about by
extinction
What it means?
• The individual information content narrows down to only a few, or even
one, highly redundant information carriers before the total information
content is extinguished
• Such selection behavior, caused by the autocatalytic nature of the
formation process, represents a typical case of survival of the survivors
• There is no other criterion for selection than the outcome (either survive or
don’t)
IV: Self-organization Based on Complementary Recognition:
Nucleic Acids
• Complementary instruction is based
on exclusive pair formation between
A and U or C and G
• Biochemist reasoning: “specific
enzymes”
• Physical chemist reasoning: “specific
forces”
• Specific complimentary paired
structures proposed by Watson and
Crick
Why do the bases pair this way?
Other Possibilities
• Hydrogen bonding alone is
insufficient
• There is a difference in geometry of
the different pairs
• Important so that the isomorphic
structure of the two pairs will be
advantageous
What are the relative stabilities?
Experiments were preformed to assess stability of
base pairing combinations
• Experiments done by:
A. Rich, J. Derkosh, T. Funck, R.
Hopman, and F. Eggers
• Relative stability measured by adding
nucleosides to nonpolar solvents
• Conclusion: The complementary pairs
AU and GC are the strongest when
compared to the alternatives
• AU is at least ten times more stable
than UU or AA
• GC is more stable than CC or GG
• GC is more stable than AU
This preferential complementary base pairing is an important prerequisite of code formation
Codons with three digits are preferred for pairing with anti-codon
Codons with less than three would be very
unstable (at least for A and U)
Codons with more than three become too
“sticky” (at least for G and C)
These stability parameters are important
because the lifetime of a codon-anticodon pair
should not exceed milliseconds so that enzymes
with corresponding turnover numbers can adapt
optimally
Conclusion: The reason the triplet codon is in
place is not only for the coding of at least 20
amino acids
Additional Theorization
Can nucleic acids organize a self-replicating and further evolving unit
without catalytic help?
No, such a system would not be able to organize itself into any type of
correlated function
What does it need to do so?
V: Self-organization Via Cyclic Catalysis: Proteins
Can proteins alone form self-instructive systems? – No.
• Unlike nucleic acids, there is no property of self- or complementary instruction
inherent to amino acids
• Certain amino acids such as collagen may have a complimentary arrangement
of certain amino acids, but it is not an inherent property of the digits
• Proteins lack the very important property of mutagenicity
• When an error occurs, they would not be able to reproduce their error copy
• Specific spatial folding hinders any straightforward copying process
X-ray structure of chymotrypsin provides excellent example for the
dependence of function on a sophisticated structure
• Groups of distant sequential location are brought
together into a precisely fixed special arrangement
• Diversity of specific recognition sites is demonstrated by
large variety of antibodies
• Active center of chymotrypsin indicates recognition by
proteins is a unique result of special folding and not any
inherent property of the digits
Catalytic networks can be made through a series of enzyme
catalyzing functions
• A branched catalytic network results from the
activation of one catalyst by another
Examples:
• 1 ) Activation of chymotrypsin from
chymotrypsinogen via enzymatic cleavage of a
peptide bond close to one end of the chain
• 2) Certain proteins have the ability to catalyze the
condensation of a limited number of amino acids
into chains of specified sequences (various
pentapeptides)
• Some enzymes have polyfunctional branches to
make the network self productive
• If a loop is large enough, all auxiliary functions
can easily be located in the branches
A number of different, interacting cycles could exist
Three cycles are shown:
• Two individual cycles and loop involving
both cycles
• Due to the feedback, the cycles do not
compete for selection, rather stabilizing
each other
• The matrix of the rate coefficients of the
total system is a reflection of reaction
behavior
• Columns=species
• Rows=reactions
• Circled (+) represents coupling points of
the loops
Additional Theorization
• Two properties of proteins make them more suitable than nucleic acids for starting
self-organization:
• The higher precision of recognition of certain substrates gained with the help
of their tertiary structure
• The higher information content involved in a multistep cycle (with branches),
as opposed to a single chain of restricted length
• The major disadvantage is that proteins which catalyze their own reproduction
through specific cycles will not automatically reproduce their mutants resulting
from error copying, even if these were to offer advantages
V: Selfordering by encoded catalytic function: The requirement
of cooperation between nucleic acids and proteins
• Nucleic acids provide complementary instruction as the basis of selective
selfreproduction and code formation using a quaternary digit system
• Cooperitivity of digit interaction allows for selective and adequately
processed codon-anticodon recognition
• Characteristic single strand structures can be targets for enzymatic
recognition
• Proteins provide an enormous functional and recognitive diversity and specificity
• They can link up via catalytic couplings, increasing information capacity
• Structure modifications may provide controlling properties
• Nucleic acids provide selective advantages (beneficial mutations) for protein
catalytic functions
A Self-reproductive Hyper-Cycle
The model consists of:
Ii: nucleotide sequences of limited chain length
(complementary single strands of RNA)
Ei: catalytically active polypeptide chains
It is important that this is a closed cycle so that
different Ik do not compete and select against each
other
10 Properties of the model:
1) Each cycle has autocatalytic growth properties
A Self-reproductive Hyper-Cycle
The model consists of:
Ii: nucleotide sequences of limited chain length
(complementary single strands of RNA)
Ei: catalytically active polypeptide chains
It is important that this is a closed cycle so that
different Ik do not compete and select against each
other
10 Properties of the model:
2) Independent cycles compete for selection
A Self-reproductive Hyper-Cycle
The model consists of:
Ii: nucleotide sequences of limited chain length
(complementary single strands of RNA)
Ei: catalytically active polypeptide chains
It is important that this is a closed cycle so that
different Ik do not compete and select against each
other
10 Properties of the model:
3) Selection will be very sharp, possibly resembling
“all or none” behavior
A Self-reproductive Hyper-Cycle
The model consists of:
Ii: nucleotide sequences of limited chain length
(complementary single strands of RNA)
Ei: catalytically active polypeptide chains
It is important that this is a closed cycle so that
different Ik do not compete and select against each
other
10 Properties of the model:
4) The system will be able to utilize very small
selective advantages and evolve quickly
A Self-reproductive Hyper-Cycle
The model consists of:
Ii: nucleotide sequences of limited chain length
(complementary single strands of RNA)
Ei: catalytically active polypeptide chains
It is important that this is a closed cycle so that
different Ik do not compete and select against each
other
10 Properties of the model:
5) The cyclic coupling will provide an information
capacity which is adapted to the requirements of the
system
A Self-reproductive Hyper-Cycle
The model consists of:
Ii: nucleotide sequences of limited chain length
(complementary single strands of RNA)
Ei: catalytically active polypeptide chains
It is important that this is a closed cycle so that
different Ik do not compete and select against each
other
10 Properties of the model:
6) Genotypic mutations can be used to evolve the
system
A Self-reproductive Hyper-Cycle
The model consists of:
Ii: nucleotide sequences of limited chain length
(complementary single strands of RNA)
Ei: catalytically active polypeptide chains
It is important that this is a closed cycle so that
different Ik do not compete and select against each
other
10 Properties of the model:
7) The system selects against parasitic branches if
they have selective values smaller than that of the
members of the cycle
A Self-reproductive Hyper-Cycle
The model consists of:
Ii: nucleotide sequences of limited chain length
(complementary single strands of RNA)
Ei: catalytically active polypeptide chains
It is important that this is a closed cycle so that
different Ik do not compete and select against each
other
10 Properties of the model:
8) The only type of branch that can co-exist with the
cycle is a branch whose selective value exactly
matches that of the cycle
A Self-reproductive Hyper-Cycle
The model consists of:
Ii: nucleotide sequences of limited chain length
(complementary single strands of RNA)
Ei: catalytically active polypeptide chains
It is important that this is a closed cycle so that
different Ik do not compete and select against each
other
10 Properties of the model:
9) Only compartmentalized systems can utilize
functional branches (brought about by mutations)
exclusively to their own advantage (and thus also
allow evolution of the branches)
A Self-reproductive Hyper-Cycle
The model consists of:
Ii: nucleotide sequences of limited chain length
(complementary single strands of RNA)
Ei: catalytically active polypeptide chains
It is important that this is a closed cycle so that
different Ik do not compete and select against each
other
10 Properties of the model:
10) A system enclosed in a compartment may
individualize by linking its code units into a stable
chain
VII: Evolution Experiments
“A theoretical model is only as much as its capacity for experimental testing”
Video
Spiegelman experiments
• QB replicase system: phage QB utilizes a specific replicating enzyme which
recognizes exclusively QB RNA
• Infectious viral RNA was reproduced in cell-free media
• Cell-free solution was subject to serial transfer
• The final product contained less than one in 1015 of the initial phage
template, yet the sample was as infectious as the original one
VIII. Conclusion: Entropy time diagram for a selection process
at constant overall flows of digits
• The occurrence of a mutant
exhibiting a selective advantage
corresponds to a negative
fluctuation of entropy production
• This causes an instability/break
down of the steady state
• The former master copy dies out
and the mutant grows to a
dominant level
Conclusion
• The theory explains the general principle of selection and evolution at the
molecular level, based on a stability criterion of the thermodynamic theory of
steady states
• Evolution appears to be an inevitable event, given the presence of certain matter
with specified autocatalytic properties and a finite amount of free energy flow
necessary to compensate for the steady state production of entropy
• The theory may explain how to construct simple molecular models representing
possible precursors of living cells
• “The fact that ‘selection’ and ‘evolution’ - in a certain analogy to equilibrium in
thermodynamics - can be characterized by extremum principles allows a physical
foundation and a quantitative formulation of Darwin’s principle”
Moving Forward
“Evolution at the molecular level may be considered a game in which the intelligence
of the player is replaced by a selective “instinct” for advantage among randomly
occurring events. Therefore game theory, as introduced by John von Neumann,
which in recent years has been developed to a high level of sophistication, is the key
to any further generalization of evolution theory”
Laws of the Game (Eigen and Winkler)