Medical Hypotheses (1998) 51, 511-523

Harcourt Brace & Co. Ltd 1998

A constructionist model predicting the emergence, complementarity and classification of the nucleotide bases
M. MUSSAT, M. E. BE~GIN*, J. P. BUREAU
Laboratoire de Biologie Cellulaire et de Cytog6n#tique Mol#culaire (UPRES-JE 1952), Universite Montpellier I, Facult6 de M#decine, Avenue Kennedy 30900, Nimes, France, *Centre de Recherche Lipergen, 1424, rue Dominion, Sherbrooke, Ou6bec. Canada. Correspondence to: Dr M. Mussat and Professor J.P. Bureau, Laboratoire de Biologie Cellulaire et de Cytog6n#tique Mol6culaire (UPRES-JE 1952), Universit6 de Montpellier I, Facult6 de M6decine, Avenue Kennedy 30900, N~mes, France (Phone: +33 4 66 23 28 06; Fax: +33 4 66 23 07 19)

Abstract - - We are proposing an analytical matrix that models a logical process simulating the emergence of the bases of vital nucleotides. The construction and properties of the matricial model are outlined. The Graph 1 matrix specifies a unique distribution pattern of eight terms coded in binary triplet configurations and obeys specific dynamic laws. The whole set of binary triplet configurations is in dynamic equilibrium. For that reason, it is possible to carry out an analysis by entering by any one of the eight terms on the condition that the operating mode obeys all the internal laws of orientation and symmetries inherent in the matrix. The four chemical elements at the origin of life are distributed following their atomic structures in the order hydrogen (H), carbon (C), nitrogen (N) and oxygen (O) and organized according to the model. The dynamic properties of the model necessitate the running of three successive circular periodic studies per analysis in order to show the emergence of the four bases - adenine, guanine, cytosine and thymine - precisely in that order. The fifth base of the nucleotides - uracil - shows up twice but always in an intermediate position, thus in transition, as it is the case for messenger ribonucleic acid (mRNA). We show also that the model provides for a logical explanation of the law of complementarity of the bases and their chemical classification. It is proposed that subsequent developments of the dynamic laws of this matrix may lead to the study of the logical operations for the formation of protein sequences and of their analysis and to genetic bioprogramming in general. Thus, a strictly logical and dynamic approach to molecular genetics is possible.

*This study was carried out mainly by M. Mussat at the Cell Biology and Molecular Genetics Laboratory (JE: 1952, Professor J. P. Bureau), University of Montpellier I, Ntmes, France. Received 16 May 1997 Accepted 8 September 1997 511

512

MEDICALHYPOTHESES
X!

Introduction
A number of model experiments by various workers on prebiotic chemistry has demonstrated that, given the fight conditions, small starting materials made of H, C, N and O elements under various forms of energy can transform into more complex organic molecules that include amino acids, purines, pyrimidines, nucleotides and possibly monosaccbarides (1,2). These model experiments illustrate that interacting forms of matter and forms of energy can transform their states to generate different complex molecular combinations of biological significance by additions and permutations of their elementary constituents in an organized way. Hence, forms of matter, forms of energy and their states of evolution may be considered as distinct but inter-related forms of relation that could provide a basis for conceptualizing a formal description of the organization of their variations with time. The Big Bang interpretation of cosmological data suggests that our universe had a beginning and has a history. Hence, when related to chronological time, the organization of transforming matter and energy would appear to correspond to an evolution. It follows that the emergence of terrestrial forms of life might have been part of an evolutionary line. The theory of molecular evolution and the theory of biological evolution are both consistent with this idea and recognize that the direction of future chemical and biological transformations along the evolutionary path remains unknown but reflects the action of selective constraints (3-5). These basic observations and considerations suggest that the arrangements of atoms, of molecules and of groups of organic molecules of biological significance may follow a logical sequence of transformations that can be modeled. Of special significance for molecular geneticists is the prebiotic synthesis of purines and pyrimidines, the essential constituents of nucleic acids, and the chemical basis of genotypes of contemporary cells. We have attempted to model such an evolutionary sequence of combinations of the four elements H, C, N, O that would describe qualitatively the emergence of the bases of the nucleotides and predict their chemical and genetic arrangements that are consistent with the data obtained by molecular analysis.

X8 1

111

X2

X~ 101

010

X3

I00

'

~ 000 X~

OO1 X,

X6

Fig. 1 Diagram of Gr~h 1 matrix.

(Fig. 1). Graph 1 shows a circular matrix of eight related observables referred to by X 1 t o X 8 and coded in binary triplet terms that configure the continuous dynamic relation interconnecting them over time. The construction of Graph 1 is based on the following considerations: 1. An observable is what can be observed directly. An observable can be described by appropriate analytical variables that refer to three formal forms of relation, i.e. forms of matter, forms of energy and their transitional states. Any variable can be related in space and in time to a set of other measurable variables. 2. Components of each of the three conceptual forms of relation can vary. Being interconnected, any variation in one is related to a variation in the other
two.

Methods
Description of Graph 1: the primary matrix The model represents the continuous, evolutive variation of a dynamic relation between related observables. The primary matrix is illustrated in Graph 1

3. The notion of a continuous and unidirectional time is used as a common reference for the description of the evolutionary relation between the variables. This continuous time can change only in one direction: one can only add upon it; it can never stop or reverse. Changes in time are of two types, either acceleration or non-acceleration. The former is equivalent to a signal and is notified by the symbol 1 whereas the latter, corresponding to a form of uniform relative motion, is equivalent to a non-signal and is notified by the symbol 0. 4. Therefore, relative to time, variations in the three conceptual forms of relation can be referred to by changes in time notified by its two specific modalities of change giving 23 = 8 possible combinations, and 8 only. The eight observable/time positions are termed X1 to X 8 respectively. Using Boolean notation, they can be written in binary symbols assembled in groups of three and arranged

GENETIC BIOPGRAMMING

513

in a sequence of their increasing order as follows: 001 - 010011 - 100101 - 110111. This ordering corresponds to the order of natural numbers 0 to 7 in decimal notation. Each binary triplet configuration may be assigned a numerical value that corresponds to its binary notation. 5. The only valid graphical representation of that ordering of the binary triplet configurations in space is the circle because: (i) only a spherical but not a linear or planar representation can accommodate the integration of all possible observables from all possible spatial locations; and (ii) by convention, the sphere is represented here by its orthogonal projection, the circle. 6. The binary triplet configurations are distributed in correspondence with their initial sequential order (see below) over the eight positions placed at the ends of the four diameters fixed perpendicularly to each other. The symbols X1 to X8 are positioned clockwise in succession on the circle starting with X1 at the top end of the vertical diameter. The binary triplet 111 having the maximal potential is located logically at the highest position at X~ at the top of the graph. The binary triplet 000 having the lowest potential is positioned opposite to the binary triplet 111 at Xs. The binary triplet 001 is situated at position X4 because this is the only position for which the order of the initial sequence of 000 to 111 as well as the symmetry and the orientation of the two families of binary triplet terms are respected. The positioning of the binary triplet 001 at position X6 is not a valid alternative because it would not respect the polarity of the first family (000 to 011) relative to the second family (100 to 111) of binary triplet terms. The positioning of the other binary triplet terms becomes straightforward: 010 at X3, 011 at X2, 100 at X6, 101 at X7 and 110 at X s. The resulting distribution of the binary triplet terms combines space with time and determines an internal continuous spiral trajectory of the evolutive relation which starts at 000 at X5 to end at 111 at Xa after a diametrical transition from X 2 to X6. Note that the first family of four binary triplet terms, from 000 to 011, is oriented counterclockwise, thus is negative or levogyre, whereas the second group of four binary triplet terms, from 100 to 111, is oriented clockwise, thus positive or dextrogyre. The internal negative-positive oscillating movement thus designated by the continuous spiral movement will be referred to as the 'spiral index' or the 'spiral sequence'. This movement plays an important role in the method of analysis to be described below.
000-

co-ordinated terms and presents definite symmetries and properties that specify their relations.

Symmetries and properties of the Graph 1 matrix: symmetrical relations 1. Symmetry relative to the center of the sequence. The initial sequential order of the terms written in their binary triplet configurations defines two families of four binary triplet terms that are distributed symmetrically around the center determined by the binary triplet terms 011 and 100. These two binary triplet terms are the only ones that are opposite and consecutive. They define the first family from 000 to 011 and the second family from 100 to 111. If the binary triplet configurations represent different variables along an evolutionary path, the ordering sequence from 000 to 111 introduces the notion of differences in position, an increase in potential or a hierarchy in polarity, from 000, the minimal potential state, to 111, the maximal potential state. Accordingly, the potential or polarity of the first family is considered lower or negative relative to the potential or polarity of the second family. This ordered sequence can be considered as representing a regular negativepositive oscillation. 2. Symmetry relative to the center of the circle. Each binary triplet configuration faces its opposite diametrically: 111 to 000, 001 to 110, etc. The sum of any two diametrical binary triplet values always totals 111 (or 7 in decimal notation). 3. Specular symmetry. To each binary triplet term, there corresponds a binary triplet term that is its mirror-image. Note that the number in brackets indicates the value of the binary triplet term in decimal notation and that it is independent of the specular geometrical symmetry. For example, the binary triplet 100 (4) has the symmetrical binary triplet 001 (1) for its specular reverse. The binary triplet 110 (6) has the binary triplet 011 (3) as its reverse. The binary triplets 111 (7), 000 (0), 101 (5) and 010 (2) are their own specular symmetrical reverses (palindromes). All these symmetries are reversible. General dynamic properties
In addition to these symmetries, the Graph 1 matrix charts two movements whose ordered trajectories are distinguished graphically on the Graph 1 diagram. The circular clockwise distribution of the observables from X1 to X 8 sequences the first movement; the spiral index traces the second one. The first and second movements are considered external and internal

It is apparent that Graph 1 matrix orders an octet of

514
respectively relative to the center of the matrix. If the binary triplet 111 is signal, its opposite binary triplet 000 is its effector, from which the evolutionary relation linking the observables is effected. In practice, we are concerned mainly with the internal spiral movement. Since the set of related observables is in dynamic equilibrium: (1) any observable can be the effector; (2) any action on any one of the observables will induce a conjugated reaction of the other seven observables; (3) any transitory movement is subjected to the internal laws of symmetry and is ordered along with the spiral index. Thus, the matricial model structures the dynamic and symmetrical relations between specified observables in an operational organized pattern that can be used as an analytical grid for data analysis.

MEDICALHYPOTHESES
This transition is transformative as shown in step 3. Complete the evolution along the spiral sequence of the second cycle as in steps 2, 3 and 4. Repeat the same sequence of steps until the third cycle on the third orbital is completed. The total process of three cycles on the three orbitals reveals 24 propositions. Repeat the entire periodic procedure for each of the remaining binary triplet configurations except 011 at X2. A special treatment is needed for the effector at position 011 at X 2 because it is at the center of transition of the volumetric distribution of the binary triplet configurations. Indeed, the Graph 1 matrix represents not only the linear trajectory on the surface of the sphere but also the volumetric trajectory across its volume. Keeping in mind that Graph 1 is the graphical geometric projection of the distribution of the binary triplets on a sphere, the trajectory of the external movement is imagined first, to penetrate across half of the sphere from the surface towards the center of transition; second, to reach the state of equilibrium at the center and third, to reverse its direction from the center and move across the second half of the sphere towards the surface. Thus, position 011 at X 2 is at the center of transition of the entire external-internal movement as well as the moment of equilibrium and the turning point between two symmetrical but directionally opposed temporal changes. It is evident that the three cycles superimpose on the three sections of the general movement and that to each cycle corresponds a response described by the internal spiral sequence. Therefore, the steps characterizing the trajectory starting at position 011 are: (1) a first spiral sequence on the first cycle oriented counterclockwise; (2) a second spiral sequence on the second cycle oriented reversely; and (3) a third spiral sequence on the third cycle in the counterclockwise orientation. Furthermore, since it is at equilibrium, this configuration is given a value of zero and does not vary. Therefore, there will be no addition on the circular phase of the evolution. However. like the others each transition will be transformative.

Step 6. Step 7.

Step 8. Step 9.

Method of analysis
The particular events of the relation examined are obtained following a set of steps and rules that specify the operations connecting the binary triplet configurations according to the properties of the internal spiral movement described above. An analysis is valid provided that it obeys all the internal laws of orientation and symmetries inherent to the matrix. The sequence of the operations to carry out an analysis is detailed step by step as follows:

Step 1. The induced response starts with an effector,

i.e. the binary triplet configuration (say, 000 at Xs) that is positioned diametrically opposite to its inducing signal (111 at X1). The effector (000 at Xs) evolves additively along the negative circular phase of the spiral sequence from 001 at X 4 to 011 at X2. Then, it transits diametrically from 011 at X2 to 100 at X6. This transition elicits a transformation that implies the substitution of the last addition by the coming one. For example, if the combination ABC is at the position 011 (X2) after the addition of an A to BC and if a B is at position 100 (X6), the transition will mean the loss of A of ABC and the gain of B to yield the newly transformed combination BBC, or B2C at position 100 (X6). The evolution continues along the positive circular phase of the spiral sequence from 100 at X6 to 111 at X1. Thus, the whole spiral sequence is: 000 - 001 - 010 - 011 - transition - 100 - 101 - 110 - 111. The entire sequence defines a cycle. The variable at 111 at X 1 obtained after the completion of the first cycle transits diametrically to the position 000 at X5 of the second cycle to start the second cycle of evolution.

Step 2. Step 3.

Step 4.

Step 5.

The main rules that characterize the program of the workings of the matrix can be extracted from the above and be formulated as below:

Rule 1. Any binary triplet configuration can be used

as the effector.

GENETIC BIOPGRAMMING

515

Rule 2. Any analysis starts with the effector. Rule 3. The position of the effector determines the position of the start of the evolution along the spiral sequence. Rule 4. The two circular evolutions (positions X 5 to X 2 and X 6 to X1) of the spiral sequence are additive. Rule 5. The two diametrical transitions in the spiral sequence (X~ to X 5 and X 2 to X6) are transformative. Rule 6. The sequence continues recursively for three complete cycles. The transitional change to the successive cycle is at the X ~to X5 transition. Rule 7. The movement starting at position 011 at X 2 requires a reversal of its direction on the second cycle and the value of the binary triplet configuration is neutral.

X~

O X1 111

O X2

1o /OlO
100 ~ 0 0 Xs H 1

x~
H

ooo

x4
C

Fig. 2 Distribution of the elements H, C, N and O on Graph 1 matrix.

In conclusion, these operating rules reveal an algorithm by which diverse vital phenomena can be described. To illustrate the power of this algorithm by an example relevant to molecular genetics, it will suffice to write the four basic elements H, C, N and O sequentially according to the algorithm operated by the Graph 1 matrix and to observe the logical appearance of chemicals with formulae that include those of the nucleotidic bases.
Results Generation of the bases of the nucleotides

triplet O 111 (XI). One sees immediately that combinations of letters made of H, C, N and O are obtained which evoke chemical formulas. An interpretation is allowed once the third cycle is fully completed. Significant combinations should appear on the counterclockwise side since the four fundamental bases are themselves levogyre material structures. To avoid unnecessary repetitions, we will expose here in detail only the first analysis using the binary triplet 000 as effector (Fig. 3). Then, we will give only the results obtained with the other analyses successively. 1st cycle. The movement starts at H 000 (Xs) and proceeds counterclockwise. We obtain the additive sequence H CH CHN CHNO at 011 X2. At the transition phase, an O is loss and an H is gained yielding CH2N at 100 X 6. Then, the sequence continues clockwise to give C2H2N C2H2N 2 C2H2N20 at 111 Xs. 2rid cycle. By losing an O and gaining an H during the transition phase, C2H2N20 at 111 X 8 is transformed into C2H3N 2. Then, we proceed on the second cycle as above. We
o b t a i n C3H3N 2 C3H3N 3 C3H3N30 -

Chemical analyses have shown that the terrestrial living beings are composed of approximately 96% H, C, N and 0, and that the material support of heredity is an agglomerate of nucleotides where four main bases are adenine, guanine, cytosine and thymine. To begin with, the elements are distributed in the ordered sequence according to the Graph 1 matrix organization following their increasing atomic number: H (1), C (12), N (14) and O (16). The coincidences with the matrix positions as illustrated in Figure 2 are as follows: H 000 (Xs), C 001 (X4), N 010 (X3), O 011 (X2) -transition - H 100 (X6) , C 101 (XT), N 110 (X8) and O 111 (X1). In conformity with the properties of Graph 1 matrix, the operations will consist in an addition for each evolutionary phase in clockwise and counterclockwise directions and of a transformation for each transition phase. From any starting effector, we proceed in the same manner and engage in three successive cycles changing orbit at the X1 to X5 transition. The first analysis will begin logically with the binary triplet H 000 (Xs). When the three cycles are completed, we pass to the next analysis, using the binary triplet C 001 (X4), and so forth, until the binary

transition - C3H4N 3 C4H4N3 C4H4N4

C4H4N4O at 111 18.

3rdcycle. Proceeding the same way, we obtain C4H4N40 - transition - C4HsN 4 CsH5N 4 CsHsN5 CsHsNsO - transition CsH6N 5 C6HrN5 C6H6N6 CrH6N60. At this stage, two families of chemicals with formulae corresponding to adenine (CsHsNs) and to guanine (CsH5NsO) have appeared, as predicted on the counterclockwise phase of the movement.

516

MEDICAL HYPOTHESES

C~I~N.O

C~H~N,

Fig. 3

Analysis of the evolution of H from position 000 X 5.

To find the other bases, we proceed along the spiral trajectory from the binary triplet C 001 (X4). The resulting sequence will be 001 (effector) 09 010 011 111 - transition - 000 100 101 110. The chemical formulae obtained after three successive cycles are illustrated in Figure 4. One notes the singular appearance of chemicals with formulae corresponding to uracil (C4H4N202) in an intermediate position and corresponding to the binary triplet 101 (XT). It is noticeable that uracil appeared in a transitory position which is in line with its presence in mRNA, which is in a transition situation and not in DNA. Chemicals with formulae corresponding to pyrazinoic (CsHaN202) and parabanic (C3H2N203) acids as well as hydantoine (C3H4N202) have appeared. No significant event is identified following the analysis starting with the binary triplet N 010 (X3) (Fig. 5). The binary triplet O 011 (X2) is positioned at a turning point of the whole relation and demands a particular treatment. For the reasons discussed above, we emphasize again: (1) that the direction of the second cycle is reversed relative to the other two; and (2) that the value of the binary triplet is neutral. It follows that the movement specifying the sequence

will start at the binary triplet O 011 (X2) as the effector and will follow the spiral trajectory on the first cycle in the usual direction, then will reverse its direction on the second cycle and will continue its trajectory on the third cycle in the usual counterclockwise direction. Notice that the value of X 2 being neutral, the values of X2 and X1 are equal. Figure 6 shows the chemical formulae obtained after these operations. We find that chemicals with formulae corresponding to cytosine (C4HsN30) appeared on the third cycle at position 001 (X4) in the counterclockwise direction. Chemicals with formulaes corresponding to azomycine (C3H3N302) and cyanocetic acid (CaHsN30) have appeared also. It is important to note that no interpretable chemical substance is obtained when the sequence is built following the three normal revolutions, i.e. if we had not taken into account the reversal of the direction of the second cycle and the absence of addition at X 2. This result confirms that the particular treatment used for the binary triplet configuration at the turning point X2 is the only one valid logically because only this movement implies all the laws of organization of the Graph 1 matrix. All the other sequences resulting from the re-

GENETIC BIOPGRAMMING

517

CsH'N30*

CH N 0 s ~ 3~

Fig. 4

Analysis of the evolution of C from position 001

X 4.

C~H.N~O, C~H4N40~ CH,N20, ~.N~Os

C~HsN306
Fig. 5 Analysis of the evolution of N from position 010 X 3.

518

MEDICAL HYPOTHESES

C~H,N40=

C,H~N,O

Azomyclne

C~H~N~O2

C~I-hN~O~
C3H~N~O2

0
NO

C 2 N0 C NO H~

CHNO ~3=

CH3N=O

Cytosine

CH ~ 3s O N Cyanocetic

Fig. 6 Analysisof the evolutionof 0 from position011 X 2,

maining binary triplets to be examined are obtained following the normal algorithm. Hence, the sequence resulting from the binary triplet 100 (X6) generates chemicals with formulae corresponding to thymine on the third cycle at position 001 (X4) in the counterclockwise orientation (Fig. 7). The analyses of the binary triplets 101, 110 and 111 do not produce any significant events (Figs 8, 9 & 10).

mental laws of molecular genetics is immediately recognized: the continuous pairing of A - T and G - C and inversely.

Verification of the two classes of nucleotidic bases

Secondly, the bases of the nucleotides can be identified according to their starting and finishing binary triplet configurations following their respective cycles. For example, A and G have appeared in the same cyclic analysis of the starting binary triplet 000. A ended up at the binary triplet 010 (X3) and G at the binary triplet 011 (X2). Accordingly, one can write each of the four bases followed by their finishing binary triplet configuration over their starting binary triplet configuration as below: A: 011 G: - 000 000 010 and C: 001 001 T: - 011 100

Verification of the law of complementarity

The representation of the generation of the bases of the nucleotides based on the Graph 1 matrix also reveals two other relevant observations concerning the nucleotidic bases. Firstly, the four bases composing DNA have not appeared randomly; on the contrary, they have appeared following a rigorous sequence: A - G - C - T. If the four bases are distributed superimposed on the same reference binary triplet configurations as shown in Figure 11, the following sequence is obtained: A 000 G 001 C 010 T 011 transition - A 100 G 101 C 110 T 111. Knowing that all the components of the Graph 1 matrix are diametrically opposed two by two and that they constitute a perfect symmetry, one of the funda-

One sees immediately that this disposition reveals the classes of the bases of the nucleotides: A and G have the same denominator and correspond to the purines, whereas C and T have the same numerator and correspond to the pyrimidines.

GENETIC BIOPGRAMMING

519

Fig. 7

Analysis of the evolution of H from position 100 X 6.

Fig. 8

Analysis of the evolution of C from position 101 X 7.

520

MEDICAL HYPOTHESES

Fig. 9

Analysis of the evolution of N from position 110 X 8.

oloo
Fig. 10

5oyo

CsH~ CN,~80~ HO , , CNe 66 HO

Analysis of the evolution of O from position 111 X a.

GENETIC BIOPGRAMMING

5 21

C x~

T x, 111

T x~

100~001
x~ .4
Fig. 11 matrix.

ooo x~ A

x, G

Distribution of the bases A, G, C and T on Graph 1

Discussion
We have reported here only the main arguments that have led to the construction of the definitive representation of our model. A more complete and detailed description of the logical combinations, dead ends and explorations surrounding the discovery of the Graph 1 matrix in the sequence of its unfolding has been published previously in a different form (6). The solution arrived at for the distribution of the eight binary triplet terms configured on the circular model is the only one acceptable because it is the only one that is logically in conformity with all of the initial premises and internal symmetries. Each member of the triplet is of symbolic character and we have suggested that they represent conceptual forms of matter, forms of energy and forms of interactions describing transitional states of evolution. These terms designate forms of relation by which our perception of information exchanges can be described rationally and efficiently. The three concepts are inseparable and their dynamic relationship is organized as proposed by the Graph 1 matrix. The correspondence of two similar observables with two positions showing different binary triplet configurations illustrates the ranking function of the latter: they are used to rank the observables being studied at the right positions on the ordered trajectory of the spiral index. Hence, two similar observables can be used at two different positions on the sequence of the evolutionary dynamic relation without causing any ambiguity in the reasoning. An advantage of using the binary notation is to allow data analysis by computerized means. The precise reason for three cycles is presently unknown, but it seems consistent with the concept of the triplet nature of the arrangement. We recognize, with other scientists, that we cannot at present describe the primary origin

of energy or of matter although the contributions of many eminent contemporary physicists and philosophers have greatly advanced thinking on their relationship. This conjecture may benefit from further explorations. According to recent reviews on prebiotic chemistry (1,2,7), it may be concluded that (1): with respect to the kinds of reaction mixture used, the elemental composition is of more significance than the kinds of molecule used; (2) HCN is likely to have played a pivotal role as a starting compound in the synthetic sequences of the bases of the nucleic acids; (3) none of the compounds synthesized in the model experiments has been reported to be optically active; (4) the processes that occurred spontaneously on the earth were operationally simple at the same time as being mechanistically complex; (5) the evolutionary complexification of chemicals of biological significance appears as the result of selective chemical reactions and; (6) the geobiological relevance of most theoretical kinetic models of possible reactions and hypothetical model experiments appears arbitrary and remains uncertain. Clearly, prebiotic chemistry would benefit from opening up to other possibilities. There are reasons to believe that our model could be used as a guide for the undertaking of new, and perhaps more appropriate, model experiments. As with most models of purine and pyrimidine syntheses, it includes HCN as a primary reactant. It predicts the emergence of families of chemicals that include the five nucleotidic bases through an ordered cyclic process unfolding in conformity with the reactions depicted in Graph 1. Although not suggesting possible mechanisms, the model shows a selective pathway leading to the synthesis of chemicals with base-like formulae with the correct optical activity, a characteristic that has not been observed in other prebiotic models. Not intended to tell us the mechanisms of the chemical reactions inolved, but rather to demonstrate the ordered sequence of their synthesis and to predict their arrangement as seen in contemporary cells, our model introduces the possibility that the chemistry of life and part of its genetic aspects may be interpreted as the result of a logical evolutionary informative process. It is apparent that the logic used for the construction of our model may be exploited to explore the emergence and arrangements of other chemicals of biological significance including amino acids, nucleotides, lipids and saccharides. Such attempts are in progress. Recent attempts have been made to structuralize biology (8-12). Living systems are by their nature defined in a dynamic sense. Hence, one should study life-forms not only from their elements of matter, but also in conjunction with their energetic and evolutionary processes. The question of the logical nature

522 and systematization of the process of energetic/ massive observable events in evolution has been discussed recently in a biological context (13-15). In line with this concept, we propose that the Graph 1 matrix represents graphically the geometric projection of a solution of the complex holonomic equation of transformation of observables definable in terms of energy, matter and evolution by which biology can be structuralized. As will be illustrated in future publications, this type of matrix permits a deeper understanding of biological systems by integrating diffuse information into clear patterns of knowledge. It is a way to express complex ideas of biology visually, thereby making them more accessible to analysis and verification. The Graph 1 matrix is just one step in that direction. The analysis showed that the generation and appearance of chemicals with formulae that include the bases of the nucleotides could result from a strictly oriented process based on logical operations sequenced in an orderly way in concordance with the matrix. One could raise the objection that the algorithm found is arbitrary and constitutes in itself nothing more than an isolated and fertile but otherwise meaningless coincidence. It is not so, for the following three reasons: firstly, the emergence at the right place of the matrix of chemicals with formulae specifying the bases, their precise order of appearance as well as the confirmation of the law of complementarity of the bases, validate the workings of the matrix and its congruence with observed analytical data. Secondly, the heuristic power of Graph 1 matrix extends to many examples taken from other areas of biology including biochemistry (6). The feasibility of general applications argues against a mere anecdotal curiosity. Thirdly, it can be developed into other matrices that predict the ordering of some genetic events (submitted for publication). It may also be argued that the combination of the four elements used is but one of 8! = 40 320 possible combinations. The point is that this particular ranking combination of the four elements follows precisely the order of their increasing atomic structure and that when this precise sequence is organized and treated according to the Graph 1 matrix, it generated four families of chemicals that included the four bases of DNA successfully and in an orderly way and accounted by the same operation for their grouping into two classes and their complementarity. The key assumption is that one of the many compounds in the families of the chemicals generated does correspond to the corresponding base. Of course, the alternative assumption is to assume that such a base is not included. If the latest case were true, it would also mean that the synthesis of the corresponding base would be selectively and specifi-

MEDICAL HYPOTHESES

cally restricted. Since nothing justifies such a restrictive constraint and since molecular data show that the nucleotidic bases exist, the assumption that the corresponding nucleotidic base is included in the families of chemicals generated appears more acceptable than its negative alternative. As far as we are aware, given that the bases are included in the families of the chemicals generated, this model is unique in predicting at the same time the emergence, the law of complementarity and the two classes of the bases of the nucleotides. Moreover, it suggests that an order of efficiency of biological significance can be extracted from the apparent multiplicity of reactive chemicals. The model illustrates how dynamic and symmetrical relations between related specified observables can be structured in an operational organized pattern that can be used as an analytical grid for data analysis. The operating rules of the matrix reveal an algorithm by which diverse vital phenomena, including those relevant to genetics, can be described in an organized way. Corresponding to a logical reality, we conclude that our model is not arbitrary and that it is valid for the cases where it applies. By providing a solution that compels by its logic, coherence and simplicity, we suggest that this model deserves proper consideration. This work is a prolegomenon to investigations into the structuring of data obtained by molecular genetics in order to integrate them into a general and coherent logical operational model that may lead to a better definition of the complex ordering system. Molecular geneticists have concentrated mostly on describing the mechanics of genetic information transfer; however, no machine can be understood merely by a description of its hardware: the theoretical aspects must be considered as well. Through the years, one of us (MM) has designed software that offers a solution for deciphering the logic ordering the molecular mechanisms of genetic information transfer. An analysis of the data obtained by molecular biology using models of the type described here could reveal the existence of hidden logical mechanisms implicated in the ordering of genetic events. Such an analysis would shed a new light on the apparent arbitrariness of the genetic code system (16-18) in particular and of the genetic organization in general. We have presented here an analysis of the emergence of chemicals with formulae corresponding to the nucleotidic bases using the Graph 1 matrix, but it appears also possible to exploit the use of other appropriate analytical matrix models for the study of, for example, the organization of DNA bases into genes and its relationship with the organization of amino acid sequences into proteins. Our proposed models may complement favorably other models on genetic algorithms that have appeared recently (19-21). Such studies are in

GENETICBIOPGRAMMING progress and their application m a y lead to the possibility o f a strictly logical and d y n a m i c approach to h u m a n m o l e c u l a r genetics and m a y allow us not only to verify or m o d i f y the results and interpretations obtained by m o l e c u l a r biology, but clarify a n u m b e r of ambiguities and reveal singularities that will be the objects o f later discussions. T h e m a i n a i m o f this article is to introduce the first m a t r i x o f a group o f three that m o d e l s the logical s y s t e m g o v e r n i n g genetic operations at the m o l e c u l a r level. In this particular article, the m o d e l is used to s h o w that the e m e r g e n c e o f the nucleotidic bases o f D N A , their g r o u p i n g into two classes (purines and pyrimidines) and the law of base-pairing m a y o b e y a logical rule. Our d e m o n s t r a t i o n is intentionally logical rather than c h e m i c a l or m a t h e m a t i c a l to e m p h a s i z e the logical nature underlying the c h e m i c a l operations. T h e r e is no shortage o f e x c e l l e n t descriptions of m o l e c u l a r b i o l o g y , with all its discoveries and the insight it has gained into structures and reaction m e c h a n i s m s in biology. It appears that the only e l e m e n t lacking in the n e w and recent k n o w l e d g e is its integration into a general understanding of Nature. R e l e v a n t m o d e l s of such integration h a v e b e e n published recently (22-24). A d d i n g up to this theme, our m o d e l with its extension, to be d e c r i b e d in detail elsewhere, is intended to be used as a tool for the study o f the ordering and integration system organ i z i n g genetic m o l e c u l a r events and is an attempt to a c h i e v e a better u n d e r s t a n d i n g o f its nature.

523 2. Fox S W, Dose K. Molecular evolution and the origin of fife. New York: Marcel Dekker 1977: 370. 3. Gitlepsie J. The causes of molecular evolution. New York: Oxford University Press, 1991: 250. 4. Li W H, Graur D. Fundamentals of molecular evolution. Sunderland, MA: Sinaner, 1991: 140. 5. Sharp P M. In search of molecular darwinism. Nature 1997; 385: 111-112. 6. Mussat M. Biodynamique (4 fasc.). Paris: Le Francois, 1974: 619. 7. Norden B. The asymmetry of life. J Mol Evol 1978; 11: 313-332. 8. Sibatani A. How to stmcmralize biology? In: Goodwin B, Sibatani A, Webster G, eds. Dynamic Structures in Biology. Edinburgh: University Press, 1989: 16-30. 9. Ji S, ed. Biocybemetics: a machine theory of biology. In: Molecular Theories of Cell Life and Death. New Brunswick: Rutgers University Press, 1991: 2-237. 10. Elsasser W M. Principles of a new biological theory: a summary. J Theor Biol 1981; 89: 131-150. 11. Elsasser W M. Reflections on a theory of organisms. Quebec: Orbis Publishing, 1987: 160. 12. Smith H A, Welchy G R. Cytosociology: a field-theoretic view of cell metabolism. In: Ji S, ed. Molecular Theories of Cell Life and Death. New Brunswick: Rutgers University Press, 1991: 282-322. 13. Lupasco S. Le principe d'antagonisme et la logique de l'6nergie. Paris: Hermann 1951: 137. 14. Atlan H. L'organisation biologique et la throrie de l'information. Paris: Hermann, 1972: 300. 15. Jaffelin J. Pour une thdorie de 1'information g~nrrale. Paris: Jaffelin 1993: 376. 16. Cullmann G, Labouygues J M. The mathematical logic of life. Origins Life 1984; 14: 747-755. 17. Cullmann G, Labouygues J M. Le code gdn~tique, code instantan6 absolument optimal. C R Acad Sc 1985: 301: 157-160. 18. Gatlin L L. Information Theory and the Living System. New York: Columbia University Press, 1972: 210. 19. Forrest S. Genetic algorithms: principle of natural selection applied to computation. Scienee 1993; 261: 872-878. 20. Stefanini S, Camussi A. APLOGEN: an object-oriented genetic algorithm performing Monte Carlo optimization. Comput Appl Biosci 1993; 9: 695-700. 21. Von Heijne G. Computer analysis of DNA and protein sequences. Eur J Biochem 1991; 199: 253-256. 22. Monod J. Le hasard et la ndcessitr. Paris: Le Seuil, 1970: 197. 23. Lima de Faria A. Evolution without selection. Amsterdam: Elsevier Science, 1988: 372. 24. Eigen M. Steps towards life. [Translated by Paul Wooley] New York: Oxford University Press, 1992: 173.

Acknowledgements
We thank Rodolphe Hunzinger for technical assistance, Francois Gu6rin for the design of the figures, and Laurent Henry and C61ine Ginestier-Veme for their with the help typescript.

References
1. Brack A, Raulin F. L'6volution chimique et les origines de la vie. Paris: Masson, 1991: 181.