Conformational studies on nucleotide–amino acid ... - CiteSeerX

Proc. Int. Symp. Biomol. Struct. Interactions, Suppl. J. Biosci., Vol. 8, Nos 3 & 4, August 1985, pp. 823–835. © Printed in India.

Conformational studies on nucleotide–amino acid interactions leading to origin of life R. BALASUBRAMANIAN Department of Crystallography and Biophysics,* University of Madras, Guindy Campus, Madras 600 025, India Abstract. Living processes may be defined as the self-sustained chemical reactions based on the special chemical machinery of nucleic acid-directed protein synthesis. Its genesis may be traced to the molecular interaction between nucleotides and amino acids leading to a primitive adaptor-mediated ordered synthesis of polypeptides. A primitive decoding system is described and its characteristics are shown to imitate, in a primitive manner, the present-day elaborate machinery of protein synthesis. This molecular interaction theory may be rightly considered as the missing link between the Protochemical and Biological Evolution. The origin of chiral specificity observed in living organisms is also traced to this specific molecular interaction in the protobiological milieu. Keywords. Chicken-and-egg problem; primitive decoding apparatus; Proto-tRNA; origin of genetic code; chiral specificity; primitive protein synthesis; link between protochemical and biological evolution; definition of living organisms.

Introduction A study of origin of life necessarily involves the question of what is the underlying principle behind the self-sustained chemical processes that go on in living organisms. These chemical reactions are catalyzed and coordinated by proteins. These chain molecules of amino acid residues are in turn synthesized by a complex machinery involving nucleic acids, and so-produced proteins themselves. Thus the origin of this protein synthesizing machinery poses a chicken-and-egg problem. Moreover, this protein synthesizing mechanism has a peculiar characteristic. The amino acids are not directly and simply condensed into polypeptides through catalysis. An adaptor molecule tRNA, charged with a specific amino acid at its one end, corresponding to a triplet of nucleotides (anticodons) situated at another end, plays a crucial role in a complex process. The anticodon triplet of nucleotide “recognizes” a triplet codon through Watson-Crick base pairing on another long chain nucleic acid namely mRNA. The consecutive triplet codons on mRNA are recognized by appropriate tRNAs having complementary anticodons and the amino acids charged to these tRNAs condense into ordered polypeptide sequences. Thus protein synthesis in biological organisms happens to be a nucleic acid-directed, adaptor mediated, ordered synthesis. It may be noted that this nucleic acid-directed protein synthesis is found in all living organisms and is conspicuous by its absence in all non-living organization of * Contribution No. 665 from this department.

823

824

Balasubramanian

matter. Also, since this forms the basis of all biological functions this may be rightly taken for a definition of life. Even with the minimum number of necessary and sufficient components, viz., mRNA, tRNA and charging enzymes synthetases, this molecular machinery is so complex that it could not have arisen all at once (lock stock and barrel) in the prebiotic milieu. Moreover in any stochastic theory for origin of life, the probability for the establishment of a particular machinery like this is extremely small. Hence one has to look for a simple molecular interaction that could have evolved and culminated into the present-day precision mechanism. Several possible molecular associations have been discussed in the literature (see Hopfield, 1978; Balasubramanian, 1980; Balasubramanian et al., 1980; Hartman, 1984), but none of them seems to give a precise molecular interaction rationale for the origin of this mechanism. Origin of the primitive decoding system We considered a pentanucleotide having uracil at the 5'-end, a purine at the 3'-end flanking any three bases in the middle (Balasubramanian, 1979; Balasubramanian et al., 1980). One of the most favourable conformations of this oligonucleotide is the U-turn conformation in which there is a strong hydrogen bond between the donor N3-H3 of uracil and highly electronegative oxygen of the fourth phosphate, with the second, third and fourth bases stacked into an RNA-11 helical conformation. These latter three residues are termed as primitive anticodons (PAC) for reasons that follow. This conformation has a ‘cleft’ formed between U(l)* and PAC. An amino acid was nestled into this cleft making use of some specific hydrogen bonds. The 3'-end purine of the pentanucleotide was also made use of for this cosy nestling. This hydrogen-bonding molecular interaction between a pentanucleotide and an amino acid at once turned out to be a primitive decoding system, in which the pentanucleotide serves as a primitive tRNA (PIT) of a double-sided template, wherein PAC is capable of base pairing with a codon sequence on another long chain RNA (primitive mRNA, PIM) while PAC’s other side can discriminate amino acid residues, depending upon the base sequence. The specific hydrogen bonds that hold the amino acid in the PIT (figure 1) are: (i) One of the carboxyl oxygen of the amino acid accepts a proton from the 2'hydroxyl of U(l) ⎯AAHB1. (ii) The amino group of amino acid donates one proton to C2–O2 of base U(l)⎯ AAHB2. (iii) This amino group donates another proton to N7 of base A(5)—AAHB3. These three specific hydrogen bonds constrain the amino acid to have a specific configuration in which the R-group of amino acid points towards a surface of PAC, called the backside-surface (another front-side surface is free to interact with codons on PIM, figure 2). The interactive characteristics of this backside surface of PAC changes

* Throughout the text, numbers within parenthesis (like U(l)) refer to the residue number in the pentanucleotide.

Nucleotide-amino acid interactions

825

Figure 1. Conformation of Primitive tRNA(PIT) holding an amino acid through three hydrogen bonds AAHB1, AAHB2, AAHB3. The U-turn hydrogen bond N3–H3 – – – OL is seen in the middle of the diagram. There is another hydrogen bond O2'– –H2 of G(3) – – – O1 of C(4) which constrain the bases G(3) and C(4) to have a standard RNA-11 conformation. Such a hydrogen bond is absent between sugars (2) and (3). The residues in the pentanucleotide are numbered from the 5'-uracil and these residue numbers are shown within parenthesis.

with the sequences of bases in PAC and we show below specific correlations between the PAC sequences and the R-groups of amino acids for favourable interactions; and this correlation corresponds to genetic code. Thus our molecular interaction theory at once gives a rationale for the origin of a decoding system and also for the origin of the genetic code (see also Balasubramanian, 1981, 1982). The genetic code Let us consider a few key amino acids from the codon table and the corresponding anticodon sequences (Balasubramanian and Raghunathan, 1984). Figure 3 shows Threonine nestled in a PIT having PAC sequence 5'-GGU-3'. It may be seen that OG2H of Thr gets locked up into a good hydrogen bond with O4 of U(4). This hydrogen bond uniquely fixes the orientations of the other groups attached to CB, viz., HB and the methyl group CG1 (H3). The methyl group has interactive contact with the base

826

Balasubramanian

Figure 2. The “inside” or the backside and the “outside” or front-side of the PAC are shown schematically over a suitable projection of the molecular moiety. The frontside of PAC is in such a configuration as to readily base-pair with a codon-sequence while the backside molecular surface can specifically interact with side-chain of an amino acid residue. The C L E F T formed by the U-turn conformation and the amino acid nestled therein are schematically shown for a clear comprehension of the molecular mechanism of the primitive decoding system.

G(3) providing a close packed conformation for molecular association of this PIT and Thr. Energy calculations were carried out by varying the side chain dihedral angles χ1, χ21, χ22 of Thr. Since many of the conformations were eliminated by short contacts we could start at a reasonable conformation, vary one dihedral angle at a time and cyclize this procedure to arrive at the minimum energy conformation of χ1= χ21= – 40° and χ22 = – 90°. The parameters of the hydrogen bond OG2-HG2– – – O4 (of U(4)) are: Η – – – Ο = 1·8 Å and angle Ο—Η – – – Ο = 155°. The β-carbon of Thr is asymmetric and on interchanging any two groups attached to CB we get allothreonine. When we tried to nestle allothreonine in this PIT, the interaction became unfavourable due to loss of the hydrogen bond and due to short contacts between its R-group and PAC bases, particularly G(3). Thus our molecular interaction theory presents a new angle of view and interesting possibilities for the elimination of non-proteinous amino acids in the very early stages of the development

Nucleotide-amino acid interactions

827

Figure 3. A suitable projection of the complex between PIT 5'-UGGUA-3' and LThreonine. The key atoms of interactions are marked. For further discussions see text.

of decoding system (for other explanations see Rohlfing and Saunders, 1978; Weber and Miller, 1981). When Gln is nestled in PIT having PAC sequence 5'-UUG-3', two hydrogen bonding interactions are possible, between R-group of Gln and U(3) and G(4) of PAC. NE1—Hl (Gln) – – – O4 of U(3) NE1—H2(Gln) – – – O6 of G(4) On replacement of bases U(3) and G(4) by other bases either short contacts develop or hydrogen bonds are lost. For the sister amino acid Asn in association with PAC, 5'-GUU-3', the following two hydrogen bonds are formed ND1—H1 (Asn) – – –O4 of U(3) ND1—H2 (Asn) – – – O4 of U(4).

828

Balasubramanian

Comparing the hydrogen bonding interactions pertaining to these two sister amino acids, it is seen that the longer side-chain of glutamine is accommodated by the farther carbonyl O6 of G(3) while the shorter side chain of Asn is interacting with the nearer carbonyl O4 of U(3). Thus the last two bases of PAC seem to be uniquely suited for these sister amino acids Gln and Asn though the bases in the first (wobble) positions of ΡACs do not seem to play any specific role. Actually this is the case with most of the amino acids as the wobble anticodon is nonspecific for amino acids. In the codon table one can readily appreciate the fact that though a doublet code takes care of most of the amino acids, a triplet frame for a comma-free reading seems to be necessary. In our postulate there is a structural explanation for such a situation. In the PIT, U-turn hydrogen bonded conformation is an important feature for forming a cleft to nestle an amino acid. In this conformation U(1) is locked up by a strong hydrogen bond with the fourth phosphate allowing a trinucleotide (PAC) to remain in a regular helical conformation for possible anticodonic function. Thus triplet frame of reading the code is a natural consequence of this type of molecular association. The wobble behaviour of the first anticodon also seems to be a natural consequence of the U-turn formation. This conformation actually precludes another hydrogen bond between the first two sugars of the anticodon triplet, allowing the first base to wobble, while it facilitates a hydrogen bond between the second and third sugars of PAC, constraining these bases for a standard base-pairing with the codon. This gives a structural explanation for the degeneracy in the code and explains why an RNA happens to be the adaptor for protein synthesis. (Balasubramanian and Seetharamulu, 1980, see also 1983). Cystein is a sulphur containing amino acid having the anticodon sequence 5'-GCA3'. Our calculations show that the side chain-PAC interaction is characterised by two hydrogen bonds N6—H6 of A(4) – – – SG of Cys SG—Η of Cys – – – N7 of A(4). The second and third position of PAC seem to be uniquely fixed by considerations of hydrogen bonding and steric possibilities. When we considered A in the second and C in the third position of PAC, two hydrogen bonds were feasible, but the amino acid conformation has an intra-amino acid short contact since sulphur has to be in cis position with respect to carboxyl oxygen. Methionine is an amino acid which has a non-degenerate single code and interestingly its lengthy and bulky (sulphur at δ position) side chain reaches out specifically to interact with the first position of PAC and there are two possible hydrogen bonds N4—H4 of A(3) – – – SD of Met N6—H6 of C(2) – – – SD of Met. Interaction energies of all the other amino acids and their cognate anticodons have also been studied critically and they would be reported elsewhere.

Nucleotide-amino acid interactions

829

Primitive protein synthesis Another interesting and important aspect of our theory is that when two PITs holding their cognate amino acids, happen to base-pair with consecutive triplets on PIM, there is a possibility of dynamic interaction by which, the amino acid held by PIT1 is brought to proper juxtaposition of the amino acid held by PIT2, facilitating the formation of peptide bond. Sequential happening of such events can lead to a primitive but ordered synthesis of proteins in accordance with the sequence of bases (hence codons) on PIM. Thus this molecular interaction neatly mimicks the present-day machinery of nucleic acid-directed adaptor-mediated ordered synthesis of proteins, though in a miniature form. This is probably the ‘logic’ behind the later evolution of present-day sophisticated machinery. This is perhaps the rationale for the origin of this peculiar coupling process between nucleic acids and proteins forming the basis of self-sustained biochemical processes found in all living organisms. A schematic diagram of this primitive decoding apparatus is shown in figures 4 and 5. As a first step in substantiating the possibility of such molecular events, the essential features for their stereochemical feasibility were simulated using a computer (Balasubramanian and Seetharamulu, 1984). A projection diagram of the ‘final’ conformation in which the CO group of amino acid1 of PIT1 is brought to

Figure 4. Schematic diagram of PIT1 holding amino acid AA1 while PAC coupling with its complementary codon in PIM. The adjacent codon is coupled to PIT2 carrying AA2. A dynamic interaction is triggered and AA1 is brought to juxtaposition to AA2 as shown in figure 5.

830

Balasubramanian

Figure 5. In the final conformation of the dynamic interaction, AA1 and AA2 are shown schematically to be in proper juxtaposition for formation of peptide bond.

juxtaposition to NH group of amino acid2 of PIT2 is shown in figure 6. The relevant atoms Ο and Η are brought to hydrogen bonding distance of 2 A. Auto-condensation of amino acids (without enzymes) is possible if amino acid esters are brought to juxtaposition as elucidated by the experimental works of Weber and Orgel (1978), Fukuda et al. (1981), Folda et al. (1982), and Mullins et al. (1984). In order to see whether adenylated esters of amino acids could be accommodated into our proposed mechanism, we did generate the molecular moiety (figure 6) and found that there is no steric hindrance due to the additional molecular moiety. Weber and Orgel (1978) envisage that for primitive protein synthesis, the amine group of one amino acid ester might have been constrained to lie in just the correct position to attack the ester group of another, or they could be attached to tRNA-like sequences which, through their specific tertiary structure held the amino acids together in the correct orientation, and this is precisely what emerges from our molecular interaction theory. Moreover in this interaction, the amino acids are brought to juxtaposition in an orderly and sequential way through adaptor molecules adapting themselves to consecutive triplet codons on long RNA (PIM). The link between protochemical and biological evolution Thus at one stage of prebiotic chemical evolution when nucleic acids, and amino acid esters were available in a primitive milieu, the above type of molecular associations among them would naturally lead to a primitive nucleic acid-directed adaptormediated ordered synthesis of polypeptides and when such ordered polypeptide

Nucleotide-amino acid interactions

831

Figure 6. The final conformation shown schematically in figure 5 is shown here as an actual molecular projection generated in a computer. Adenylate esters of amino acids are shown projected, since such esters are known to undergo autocondensation into peptides without the aid of enzymes.

sequences happen to be those corresponding to crude enzymes like synthetases, polymerases and replicases, biochemical processes would become, viable and selfsustaining, to evolve into living organisms. Thus our proposition of molecular interaction theory bridges the missing link between the protochemical and biological evolution. Origin of chiral specificity Another interesting feature of our proposition is that it gives a novel approach in understanding the possible origin of chiral specificity (Balasubramanian, 1983). For other explanations see proceedings of the 2nd International Symposium, 1981. The specific nature of the hydrogen-bonding interaction between penta-ribonucleotides and amino acids are such that L-amino acids interact more favourably (of the order of 5 kcal/mol) with β-D-ribonucleotides than D-amino acids. Thus there is a chiral specificity between β-D-ribonucleotides and L-amino acids (Balasubramanian and Seetharamulu, 1984). Thus in the possible origin of this genetic decoding apparatus, the co-existence or conjunction of L-peptides and β-D-ribonucleotides becomes a necessity because of their cooperative template fitting interaction; of course, the mirrored combination of optical antipodes could also trigger the ordered synthesis of D-peptides (figure 7).

832

Balasubramanian

Figure 7. Schematic diagram of primitive tRNAs of opposite chiralities synthesising L- and D-peptides.

In this context an interesting possibility suggests itself. Two protein synthesizing machineries, one manufacturing L-peptides (say, L-system) and another D-peptides (say, D-system) might have developed autonomously. The two systems are capable of evolving independently. Natural mutations on the sequences of PIMs and the errors in the primitive translation would bring in new sequences of polypeptides and the Darwinian process of survival of the fittest would lead to more and more sophisticated decoding systems. But L-system and D-system would take entirely different pathways since each of them is dependent respectively on its own enentiomorphic PIM-sequences

Nucleotide-amino acid interactions

833

and the chance-mutations on them. At one stage, let us suppose that L-system develops an enzyme that is capable of efficiently cleaving all D-peptides. Thus L-system attains an evolutionary supremacy over D-system which would not be able to survive because of this “killer enzyme” (figure 8). D-system could also develop a “killer enzyme” to Lsystem. But since the two systems are independently evolving, the probability of both the systems developing a killer enzyme at the same time is extremely small and whichever system happened to bring forth a killer enzyme for the other system, survived to outlast the other in the evolutionary race.

Figure 8. Schematic diagram showing the dependence of polypeptide sequences on the PIM sequence of the corresponding chirality. Sequences of L-enzyme and those of D-enzyme are completely different. One of them could have first developed a killer enzyme for the other and hence its evolutionary survival while the other could have been eliminated from our biosphere even at the early stages of evolution.

834

Balasubramanian

Since this hypothesis is based on a mechanism of active and effective destruction of the system of opposite chirality this turns out to be a specific and forceful rationale for the establishment of biological organisms of a given chiral specificity in a biosphere. It is also testable in the sense that one could look for such an enzyme in the primitive or archaic organisms. As such, the existence of antibiotics for the presently existing Lsystem gives credence to the hypothesis and it should be noted that peptide antibiotics indeed contain D-amino acids that are inevitable for their function (Davies, 1977).

Conclusion In conclusion we would like to bring out the fact that this postulate of molecular interaction between pentanucleotides and amino acids possess almost all the salient and unique features of the contemporary protein synthesizing machinery. They can be summarized thus: (i)

(ii) (iii) (iv)

(v)

(vi) (vii)

(viii)

(ix) (x)

(xi)

The rationale behind the origin of the uniquely peculiar adaptor strategy for protein synthesis in biological processes is straightaway explained as the direct consequence of this type of molecular interaction, The triplet coding is a necessary outcome of the U-turn conformation of PIT as elaborated earlier. Again the wobble behaviour of the first base of PAC has been shown to be a necessary consequence of the formation of U-turn hydrogen bond. The glaring feature of the genetic code, viz, the middle base of codon playing a key role, the first less so, and the third being almost non-specific is reflected in primitive decoding system as the amino acid side chain neatly positioned near the middle base of anticodon, is farther from the third base and farthest from the first. In our postulate, the possibility of dynamic interaction is such that the adaptor mediated synthesis would proceed from 5'- to 3'-end of the mRNA, which is what is happening in contemporary protein synthesis. Similarly PIT’s hold on amino acid is such that the synthesis would proceed from amino to carboxyl end for the polypeptide. As discussed earlier PIT’s cleft is such that the non-proteinous amino acids might have been eliminated from this system of polypeptide synthesis in the early stages of evolution. The remarkable features of stereospecificity between β-D-ribonucleotides and L-amino acids and the possible emergence of chiral uniqueness in living process has already been elaborated. In PIT, uracil residue preceeding PAC plays a crucial role. In contemporary tRNAs U is an invariant residue preceeding anticodon triplets. In PIT, a purine succeeding PAC plays an important role in holding and positioning an amino acid. And in present-day tRNAs the base that follows anticodons is always a purine. Last but not the least, crystal structures of tRNAs reveal the existance of Uturn conformation in the anticodon loop and it is perhaps the ‘fossil’ evidence

Nucleotide-amino acid interactions

835

for our molecular interaction theory for origin of life. It seems as though the bits of a jig-saw puzzle fall into their right places when we consider this particular molecular interaction between pentanucleotides and amino acids. Acknowledgements The author thanks Drs. P. Seetharamulu and G. Raghunathan for some of the calculations carried out by them. A part of the work is supported by a grant from the Department of Science and Technology, New Delhi, for “Studies on Origin and Evolution of Life”. References Balasubramanian, R. (1979) Indian J. Biochem. Biophys., 16, 442. Balasubramanian, R. (1980) in Proceedings of the Workshop on Origin and Evolution of Life and Intelligence in the Universe held in Bombay (Jan. 14–16, 1980) ISRO, India. Balasubramanian, R. (1981) Trends Biochem. Sci., 7, 9. Balasubramanian, R. (1982) BioSystems, 15, 99. Balasubramanian, R. (1983) Origins Life, 13, 109. Balasubramanian, R. and Seetharamulu, P. (1980) J. Madras Univ., B43, 54. Balasubramanian, R. and Seetharamulu, P. (1983) J. Theoret. Biol., 101, 77. Balasubramanian, R. and Seetharamulu, P. (1985) J. Theoret. Biol., 113, 15. Balasubramanian, R. and Seetharamulu, P. and Raghunathan, G. (1980) Origins Life, 10, 1. Balasubramanian, R. and Raghunathan, G. (1985) BioSystems, (in press). Davies, J. S. (1977) in Chemistry and Biochemistry of Amino Acids, Peptides and Proteins, (ed. Β. Weinsten) (New York and Basel: Marcel Dekker Inc.) Vol. 4, p. 1. Folda, Th., Gros, L. and Ringsdorf, H. (1982) Macromol. Chem. Rapid Commun., 3, 167. Fukuda, K., Shibasaki, Y. and Nakahara, H. (1981) J. Macromol. Mol. Sci. Chem. A15, 999. Hartman, H. (1984) Origins Life, 14, 643. Hopfield, J. J. (1978) Proc. Natl. Acad. Sci. USA, 75, 4334. Mullins, Jr., D. W., Senasatne, N. and Lacey Jr. J. C. (1984) Origins Life, 14, 597. Proceedings of the 2nd International Symposium, University of Bremen, July 16–18,1980, Origins Life, 11, No. 2, 1981. Rohlfing, D. L. and Saunders, M. A. (1978) J. Theoret. Biol., 71, 787. Weber, A. L. and Miller, S. L. (1981) J. Mol Evol., 17, 273. Weber, A. L. and Orgel, L. E. (1978) J. Mol. Evol., 11, 189.