Peptides by Activation of Amino Acids with CO on (Ni,Fe)S Surfaces: Implications for the Origin of Life

See allHide authors and affiliations

Science  31 Jul 1998:
Vol. 281, Issue 5377, pp. 670-672
DOI: 10.1126/science.281.5377.670


In experiments modeling volcanic or hydrothermal settings amino acids were converted into their peptides by use of coprecipitated (Ni,Fe)S and CO in conjunction with H2S (or CH3SH) as a catalyst and condensation agent at 100°C and pH 7 to 10 under anaerobic, aqueous conditions. These results demonstrate that amino acids can be activated under geochemically relevant conditions. They support a thermophilic origin of life and an early appearance of peptides in the evolution of a primordial metabolism.

The activation of amino acids and the formation of peptides under primordial conditions is one of the great riddles of the origin of life. We have now found that under the hot, anaerobic, aqueous conditions of a setting with magmatic exhalations, amino acids are converted into peptides. Under these conditions we previously demonstrated the conversion of carbon monoxide into activated acetic acid in an aqueous slurry of coprecipitated (Ni,Fe)S at 100°C (1).

Peptides were formed from phenylalanine (F), tyrosine (Y), and glycine (G). In each run 500 μmol of the amino acid were reacted in a slurry of 1 mmol of FeS and 1 mmol of NiS in 10 ml of water with 4 mmol of CO gas (1 bar) in the presence of 500 μmol of hydrogen sulfide (H2S) or methanethiol (CH3SH) at 100°C and pH 7 to 10. In some of the runs 500 μmol of Na2HPO4 were added. After 1, 2, or 4 days, we determined the yield of the peptides and the pH in the water phase (2) (Table 1). No peptides were detectable, if under otherwise identical conditions CO was replaced by Ar, or if neither H2S nor CH3SH was added, or if both NiS and FeS were absent. In runs 13 and 14 and 19 to 22, about 3 nmol of tripeptides (Y-Y-Y) were detected after 1 and 4 days.

Table 1

Formation of dipeptides froml-phenylalanine (LF), l-tyrosine (LY),d,l-tyrosine (DY,LY) and glycine (G) in the presence of CO, H2S (or CH3SH), and 1 mmol of NiS + 1 mmol of FeS. The amount of G-G in run 25 is an average of four reactions with a standard deviation of ±1.5.

View this table:

In separate experiments it was determined that under these same conditions dipeptides hydrolyzed rapidly. For example, 100 μmol of the dipeptide G-G generated 124 or 48 μmol of G in 1 day at pH 8.4 under otherwise the same conditions as in runs 25 or 27, respectively. This means that the amounts of dipeptides given in Table 1 constitute a balance between reactions of condensation and reactions of hydrolysis. Such a counter-operation of synthetic (anabolic) and hydrolytic (catabolic) reactions with corresponding steady-state concentrations of the products is typical for all extant metabolisms.

The experiments with l-phenylalanine andl-tyrosine produced both epimeric dipeptides as a result of racemization. In the case of l-tyrosine, racemization was extensive after 4 days. These results mean that in an origin of life on (Fe,Ni)S at elevated temperatures, amino acids would be racemic. In a chemo-autotrophic origin of life (3) with a catalytic feedback of amino acids or short oligopeptides as ligands for catalytic metal centers homochirality of the amino acids or of their peptides is not essential. Homochirality becomes increasingly important with increasing chain lengths of the peptides. It is of interest that the oligopeptides of cell walls have both d– andl–amino acids (4).

The results of Table 1 show that the formation of peptides is strongly dependent on the pH of the reaction medium. In the absence of phosphate, significant peptide yields were obtained at pH of about 8 to 9.5. In the presence of phosphate, the productive pH range was broader. In our reaction system the pH of the aqueous reaction medium decreased with time. Therefore, we followed for three runs the development of pH and peptide concentration (Fig. 1). The results show that the peptide concentration rose and fell as the pH moved into and out of the optimum range. These results suggest that the control of pH and of the pH-dependence of the pathways were among the first problems to be solved by the early organisms. The lowering of the pH can be explained by the formation of acids from CO in our reaction system. We demonstrated this effect in an experiment with 1 mmol of FeS, 1 mmol of NiS, 4 mmol of CO, and 0.5 μmol of CH3SH in 10 ml of water at 100°C. After 17 days we detected 67 μmol of CH3COOH, 600 μmol of HCOOH, and 1.3 mmol of CO2.

Figure 1

Plot of development of pH and dipeptide yields over time. The numerals in circles denote the yields of LF-LF from LF; the numerals in rectangles, the yields of LY-LY from LY; and the numerals in hexagons, the yields of G-G. The standard deviation of the yields of G-G is up to ±11% (four repeats).

The reaction mechanism has both a thermodynamic and a kinetic aspect. Thermodynamically, the formation of peptides under our hot, dilute, aqueous conditions is endergonic. Therefore the mechanism must explain energy coupling with an exergonic reaction. Mechanisms based on three exergonic reactions may be considered: (i) the oxidative conversion of CO to CO2 via COS; (ii) the hydrolytic conversion of CO to HCOOH; and (iii) the formation of acetic acid from CH3SH and CO via activated acetic acid. These possibilities cannot be assumed to be jointly exhaustive or mutually exclusive.

The first possibility of an energy coupling with the oxidative conversion of CO to CO2 via COS (Fig. 2A) involves a thiazolidinedione or an oxazolidinedione (Leuchs anhydride) or its 2-thio derivative as the species suffering nucleophilic attack; this reaction has the advantage of being nonanionic. This mechanism is supported by the formation of COS in this reaction system (1), and by the observation that dipeptides were formed (data not shown), if under otherwise identical conditions we replaced CO and H2S (or CH3SH) by COS, but not in the absence of NiS and FeS.

Figure 2

Notional representation of alternative ligand sphere reaction mechanisms. The symbol “aa” represents an amino acid of the formula RCHNH2-COOH, “aa-aa” represents its dipeptide and “?” an unknown intermediate, perhaps a 2-mercapto-exazolinone. In the cyclic intermediate of mechanism A positions 1 and 2 may be occupied alternatively by O and CO or by O and CS, respectively.

The second possibility of an energy coupling with the conversion of CO to HCOOH is supported by the facile formation of formic acid in our experiments at rates that seem to correspond to the rates of peptide formation. The detection of small amounts ofN-formyl-amino acids indicates that a reaction channel from CO to HCOOH proceeds through an activated formic acid. However, a mixed anhydride (H2N-CHR-CO-O-CO-H) as the species suffering nucleophilic attack by an amino acid may be ruled out, because it would react intramolecularly rather than intermolecularly. This inference leads us to a hypothetical mechanism in which oxazolinone acts as the nonionic species for the nucleophilic attack by the amino acid, as shown in Fig. 2B.

The third possibility of an energy coupling with the formation of acetic acid is supported by the detection of acetic acid under our reaction conditions in the presence of CH3SH, by the previously demonstrated (2) formation of thioacetic acid (CH3-COSH) or its methyl ester (CH3-CO-SCH3), and by the formation of small amounts of the N-acetyl-amino acids in our system. A mechanism to the one shown in Fig. 2B may proceed through a 2-methyl-oxazolinone. Such a mechanism could make only a small contribution, because the formation of dipeptides proceeds similarly well in the presence and absence of CH3SH and because the rate of peptide formation is higher than the rate of acetic acid formation. We did not detect any dipeptides, if under otherwise identical conditions CO was replaced by 500 μmol of CH3COSH or CH3COSCH3. This result means that neither of these two activated compounds is situated in a reaction channel to the peptides.

The problem of the endergonic nature of peptide formation in dilute aqueous solutions, calculated for glycine to have a maximum around 100°C (5), has previously been approached by changing the thermodynamic conditions with heating in dry condition (6), drying/wetting cycles (7, 8), or high salt concentrations (9). Under the dilute aqueous conditions most relevant for the origin of life, activation of the amino acids by coupling with hydrolysis reactions (10) notably of inorganic polyphosphates (11) has been suggested. It is, however, not clear how under hot aqueous conditions such hydrolytically sensitive coupling compounds, if geochemically available at all, could resist rapid equilibration. In our system the chemical potential is maintained for a long time, and under natural conditions it may be readily replenished by volcanic exhalations containing CO.

Kinetically the (Fe,Ni)S catalyst in our system promotes the productive reaction channel to peptides as compared to other nonproductive channels. It means that the reaction occurs in the ligand sphere of the sulfide mineral. Previously, copper ions have been used as catalyst for peptide formation in the presence of high salt concentrations (9). However, under anaerobic conditions with even a small sulfide activity, copper ions cannot exist.

Most prior attempts to produce peptides under primordial conditions have been beset by the formation of large amounts of unreactive diketopiperazines (7, 12). For example, in drying-wetting cycle experiments with glycine on montmorillonite, the molar ratio of diketopiperazine to diglycine was more than 4:1 (13). The formation of diketopiperazines was so far only suppressed, if the amino acid was activated in the form of a Leuchs anhydride, which required organic activation agents such as carbodiimides (14). In our system the diketopiperazines form minor byproducts. For example, in run 25, the amount of the diketopiperazine was 3.5 ± 0.5 μmol; while in run 13 the amount of the diketopiperazine of tyrosine is 5 μmol after 1 day and 8 μmol after 4 days (15). Both mechanisms shown in Fig. 2 would disfavor the formation of the diketopiperazine. They would also explain the racemization by resonance-stabilized enolization.

Our result supports the theory of a thermophilic origin of life with a primordial surface metabolism on transition metal sulfide minerals. It means that a continuously recycling library of peptides was generated on the surfaces of a library of (Fe,Ni)S structures. It raises the possibility that CO and Ni had a much greater role in the primordial metabolism than in any of the known extant metabolisms. All known extant organisms are found in habitats with low activities of CO and Ni. This could explain why they resorted to the formation of CO from CO2 and to the elimination of nickel from many enzymes (16).


View Abstract

Navigate This Article