TEMPERATURE DEPENDENCE OF THE CYCLIZATION OF GUANINE AND CYTOSINE MIX HEXANUCLEOTIDES WITH WATER-SOLUBLE CARBODIIMIDE AT 0 - 75 °C

Kunio Kawamura*, Noriyuki Nakahara, Fumitaka Okamoto, and Noriko Okuda
Department of Applied Chemistry, Graduate School of Engineering, Osaka Prefecture University, Sakai, Osaka 599-8531, Japan
Fax: 0722-54-9903; Email: kawamura@chem.osakafu-u.ac.jp

(Received 4 September 2003 Accepted 10 October 2003)

(Abstract)

     The temperature dependence of pseudo-first-order rate constants (kcyc) of the cyclization of hexanucleotides 5'-d(pGCGCG)rC and 5'-d(pGCCCG)rG was investigated in the presence of 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDAC) and imidazole at 0 - 75 °C. Although the association between the elongating oligomer and the activated monomer is necessary for the success of the formation of oligonucleotides on a polynucleotide template or a clay catalyst, the phosphodiester bond formation in the case of cyclization readily occurs without these catalysts. Thus, the cyclization is regarded as a simple model of the formation of phosphodiester bond. The rate constants (kcyc) were compared with those of the cleavage of the ribose phosphodiester bond, in which the cyclization was more than 50 times faster than the cleavage at 75 °C. On the basis of this fact, it is deduced that the prebiotic formations of RNA oligomers such as the template-directed reaction and the mineral catalyzed oligomerization would be possible at high temperatures if the association between elongating oligomer and activated monomer (or activated oligomer) is sufficiently strong.

(Keyword) Cyclization of oligonucleotide, Temperature dependence of the formation and degradation of phosphodiester bond, Hydrothermal origin of life, Oligonucleotide, RNA, Phosphodiester bond formation, Hydrolysis of phosphodiester bond, Cyclicnucleotide, RNA world

Introduction

     The discovery of catalytic functions of ribonucleic acids (RNA) suggests that RNA or RNA-like molecules played central roles for the emergence of genetic information on the primitive earth environments (RNA world hypothesis) [1]. If the RNA world hypothesis is correct then RNA could have been accumulated under the prebiotic earth conditions. The condensation reactions of nucleotides with and without using activated nucleotide monomer have been investigated as primitive polymerase models of oligonucleotides. Examples include the condensation of activated nucleotide monomers on a polynucleotide template [2,3], and the spontaneous oligomerization of activated nucleotide monomer in the presence of several catalysts [4,5]. In addition, the condensation reactions of oligonucleotides using 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDAC) were also investigated [6-9]. In our group, the condensation reaction of hexanucleotides was studied in the presence of EDAC at 0 °C, in which the cyclization of hexanucleotide proceeds efficiently [10].
     On the other hand, it is widely believed that hydrothermal systems such as hydrothermal vents in the deep ocean played important roles for the emergence of life (the hydrothermal origin of life hypothesis) [11]. Thus, the RNA world hypothesis seems to be inconsistent with the hydrothermal origin of life hypothesis since it is imaginable that RNA could not have possessed sufficient stability for preserving information and catalytic ability under the hydrothermal conditions [12]. Although both the rates of the formation and decomposition of RNA should determine the accumulation of RNA molecules under the primitive earth conditions, there have been few investigations on the temperature dependence of the prebiotic formation and decomposition. The comparison of the rates of the formation and degradation of oligonucleotides is essential to estimate the accumulation behavior of RNA. Thus, we have been carried out comparative kinetics of the formation and degradation of RNA from the viewpoint of the hydrothermal origin of life [13,14].
      Our previous study on kinetics of the template-directed reaction (TD reaction) at 40 - 80 °C elucidated that the rate constants of the formation of 3-mer and longer are greater than those of the degradation of oligonucleotides formed by the TD reaction even at 80 °C and higher [13]. However, the actual yield of the TD reaction was very low at 80 °C. The reason is due to that the relative rate of the formation of 2-mer notably decreases at 80 °C and higher. Thus, it was concluded that the TD reaction is not efficient at over 80 °C unless the formation of 2-mer is enhanced. Fortunately, the comparison between the rate constants of the formation of oligoguanylate (oligo(G)) by the TD reaction and the hydrolysis of oligo(G) at elevated temperatures was succeeded in the TD reaction system, since both the formation and degradation of oligo(G) obeyed pseudo-second order processes.
      Recently, we found that the cyclization of hexanucleotides readily occurs in the presence of water-soluble carbodiimide (Fig. 1), which is highly efficient compared with the dimerization of activated nucleotide oligomers in the absence of the polynucleotide template or the clay catalyst [10]. In addition, since the cyclization reaction obeys a pseudo-first-order process, the direct comparison of the rate constants of the formation and cleavage of phosphodiester bond is possible. These facts indicate that the cyclization of oligonucleotides is a suitable model reaction to evaluate the rate constants of the phosphodiester bond formation at high temperatures. In the present study, the scope of the cyclization reactions of hexanucleotides 5'-d(pGCGCG)rC (oligo-6a) and 5'-d(pGCCCG)rG (oligo-6b) was investigated, in which the rate constants (kcyc) were determined at 25 - 75 °C. In addition, the rate constants of the cleavage of the ribose phosphodiester bond of dCrCdGdG (kdeg) were determined in the presence of EDAC at elevated temperatures. The thermodynamic parameters of the cyclization and cleavage of the phosphodiester bond were also determined from the temperature dependence of these rate constants. Possibility of the accumulation of RNA is discussed on the basis of the comparison of kcyc and kdeg.



Figure 1. A perspective view of the formation of phosphodiester bond in hexanucleotide with van der Waals radii. Hexanucleotide is oligo-6b, in which the activated terminal is not shown.

Experimental

Oligonucleotides oligo-6a, oligo-6b, dCrCdGdG, d(pC5)rC, d(pC6)rC, and d(pC7)rC were purchased from GENSET (France) as HPLC purified grade. All other reagents used were of analytical grade.
      The condensation reactions were performed in an aqueous solution containing 0.2 M NaCl, 0.075 M MgCl2, 0.1 M imidazole, 0.2 M EDAC, and 0.05 - 0.1 mM oligo-6a, oligo-6b or oligocytidylic acid (oligo(C)) at 25 - 75 °C. The solution containing EDAC was freshly prepared for each reaction. The pH of the sample solutions was adjusted with 0.1 M NaOH or 0.1 M HCl solution. The mixture was allowed to stand for 24 h at 25 °C, 3 h at 50 °C, and 30 min at 75 °C. Aliquot of sample was periodically withdrawn and immediately quenched in liquid nitrogen to stop the reaction. Samples were analyzed by HPLC on a DNA-NPR anion exchange column from TOSOH Co., Tokyo, Japan using a gradient of 0 - 1.2 M NaCl at pH 11 with 0.02 M 2-amino-2-hydroxymethyl-1,3-propanediol.
      The degradation of dCrCdGdG was investigated in a buffer solution, which is the same as used for the condensation reactions of oligo-6a and oligo-6b. Products were analyzed by the reversed-phase HPLC on a ODS-2 column from GL Science Co., Tokyo Japan using a gradient of 0.02 M NaH2PO4, 0.005 M tetrabuthylammonium bromide (TBABr) in water at pH 3.5 mixed with 0.02 M NaH2PO4, 0.005 M TBABr in 60% CH3OH at pH 3.5.
      The melting temperatures (Tm) of d(GCGCGC) and d(GGCCCG) were determined in aqueous solutions containing 0.2 M NaCl, 0.075 M MgCl2, 0.1 M imidazole, and 0.67 mM oligo-6a or oligo-6b at pH = 8.0 using a CSC 6100 Nano II differential scanning calorimeter (DSC) (Calorimetry Sciences Corp., USA).

Results and Discussion

Stability of EDAC
The stability of EDAC was investigated on the basis of our previous study [10]. UV spectra of the solution containing NaCl, MgCl2, imidazole, and EDAC was monitored at 200 - 300 nm at 0 - 75 °C, where the maximum wavelength was 234 nm [10]. The result showed that absorbance at 234 nm decreased to 85 % of the initial amount of EDAC after 7 days at 0 °C, 87 % after 16 h at 25 °C, 77 % after 2 h at 50 °C. The absorbance at 234 nm slightly increased after 30 min at 75 °C. The reaction curves at the beginning of the cyclization reactions were used for the determination of the rate constants of the disappearance of hexanucleotides, at which the influence of the degradation of EDAC was possible to be ignored.

Reaction behavior of cyclization at 0 - 75 °C
The condensation reaction of oligo-6a and that of oligo-6b were monitored by HPLC and two main peaks appeared (peak-1 and peak-2), where peak-1 was assigned to oligo-6a or oligo-6b [10]. The characterization of peak-2 was carried out in our previous study using specific enzymatic hydrolyses; the peak-2 fraction was assigned to cyclic hexanucleotide (cyclic-oligo-6) [10]. It was confirmed that peak-2 disappeared by the treatment with ribonuclease T2 (RNaseT2) and was not changed by the alkaline phosphatase treatment. The reaction curves at 25 - 75 °C are shown in Fig. 2. A relatively large amount of the dimerized product (12-mer) of oligo-6b was detected in the condensation of oligo-6b at 0 °C, and a very small amount of the dimerized product of oligo-6a was barely detected. The yield of the dimerized product of oligo-6b became negligible at 25 °C and higher temperatures. Since imidazole was necessary for the cyclization reactions of oligo-6a and oligo-6b, the formation of imidazolide onto the 5'-terminal phosphate of hexanucleotide is assumed in the presence of EDAC. The formation of imidazolide is supported by several studies [6-8,10]. Cyclic hexanucleotides were rapidly formed from oligo-6a and oligo-6b, but the decomposition of cyclic-oligo-6 was hardly observed (Fig. 2). Thus, the cyclization is obviously faster than the degradation of cyclic-oligo-6. The reaction for the case of oligo-6a is expressed in equations (1) and (2).

     5'-d(pGCGCG)rC → cyclic-oligo-6a     (1)
     cyclic-oligo-6a → 5'-d(GCGCG)rC>p    (2)




Figure 2. Reaction curves for the cyclization of oligo-6a. [NaCl] = 0.2 M, [MgCl2] = 0.075 M, [imidazole] = 0.1 M, [EDAC] = 0.2 M, [oligo-6a] = 1 x 10-4 M, pH = 8.0. ○: oligo-6a, ●: cyclic-oligo-6a, □: unknown.


Pseudo-first-order rate plots of the disappearance of oligo-6a and oligo-6b showed that the reactions obey pseudo-first-order kinetics at the beginning of the reactions, at which the reactions proceed about 50 % (Fig. 2). The reaction was gradually slowed from the pseudo-first-order process. This is mainly caused by the degradation of EDAC and possibly by the degradation of cyclic-oligo-6, so that the rate constants were determined from the linear part of the pseudo-first-order rate plots (kcyc) (Fig. 3, Table 1). The magnitudes of kcyc are not much different between oligo-6a and oligo-6b.




Figure 3. Pseudo-first-order rate plots for the disappearance of oligo-6a. Reaction conditions are the same as shown in Figure 2. Temperature (°C), ○: 25, ●: 50, □: 75.

Table 1. The rate constants for the formation of cyclic-hexanucleotides.
T / °C kcyc /s-1


oligo-6aoligo-6b
0a(1.6 ± 0.1) x 10-6(6.9 ± 0.1) x 10-7
25(2.4 ± 0.1) x 10-5(1.6 ± 0.1) x 10-5
50(2.3 ± 0.1) x 10-4(2.2 ± 0.1) x 10-4
75(1.6 ± 0.1) x 10-3(1.6 ± 0.1) x 10-3
Reaction conditions, [NaCl] = 0.2 M, [MgCl2] = 0.075 M, [imidazole] = 0.1 M, [EDAC] = 0.2 M,
[hexanucleotide] = 1 x 10-4 M, pH = 8.0. Hexanucleotides, oligo-6a: 5'-d(pGCGCG)rC, oligo-6b: 5'-d(pGCCCG)rG. a Data from [10].

     On the other hand, kinetics of the cleavage of phosphodiester bond at elevated temperatures have been investigated in the presence of EDAC; the rate constants of the cleavage of phosphodiester bond is much smaller than kcyc [14]. Naturally, the reaction curves in Fig. 2 indicate that the rate of the degradation of cyclic-oligo-6 ismuch smaller than that of the formation. In this study, the rate constants of the cleavage of ribose phosphodiester bond of dCrCdGdG (kdeg) were determined in the presence of EDAC at 50 - 85 °C (Table 2) to compare the rate constants of the cleavage and cyclization. HPLC charts for the degradation are shown in Fig. 4. The cleavage of the ribose phosphodiester bond of dCrCdGdG is consistent with pseudo-first-order kinetics, where dCrC>p and dGdG are formed. The -rCdG- sequence in dCrCdGdG is identical to that of cyclic-oligo-6a. The fact that the magnitudes of kdeg are similar to those obtained without EDAC [14] indicates that EDAC does not affect the cleavage of phosphodiester bond. Incidentally, the cyclic-oligo-6b forms the phosphodiester bond of the -rGdG- sequence, and the degradation rate of the phosphodiester bond of linear dCrCdGdG may be different from that of cyclic-oligo-6. These differences may cause different stability of the phosphodiester bond. However, the sequence and location of the phosphodiester bond within a hairpin loop or a single strand does not affect the rate of the cleavage of phosphodiester bond [15].

Table 2. The rate constants for the cleavage of tetranucleotide and the degradation of water-soluble carbodiimide.
T / °Ckdeg / s-1kEDAC / s-1
50(1.8 ± 0.1) x 10-6(2.3 ± 0.1) x 10-6
75(2.4 ± 0.1) x 10-5(1.9 ± 0.1) x 10-5
85(6.6 ± 0.2) x 10-5(3.6 ± 0.2) x 10-5
Reaction conditions, [NaCl] = 0.2 M, [MgCl2] = 0.075 M, [imidazole] = 0.1 M, [EDAC] = 0.2 M, [dCrCdCdG] = 5 x 10-5 M, pH = 8.0.



Figure 4. HPLC charts for the degradation of dCrCdGdG. [NaCl] = 0.2 M, [MgCl2] = 0.075 M, [imidazole] = 0.1 M, [EDAC] = 0.2 M, [dCrCdGdG] = 5 x 10 -5 M, pH = 8.0. Temperature: 75 °C. *The peak does not correspond to any reaction educts and products since it appears by the injection of water.

Thus, the fact that the magnitude of kcyc is more than 50 times greater than that of kdeg straightforwardly means that the degradation of cyclic-oligo-6a and cyclic-oligo-6b is much slower than the cyclization of oligo-6a and oligo-6b. From the temperature dependence of kcyc and kdeg, the apparent activation energy (Ea,app), the apparent enthalpy change (ΔH≠), and the apparent entropy change (ΔS≠) of the reactions were calculated (Table 3). The Arrhenius plots visualize the difference between kcyc and kdeg (Fig. 5). The magnitudes of ΔH≠ and ΔS≠ for oligo-6a are somewhat smaller than those for oligo-6b. The fact that the value of ΔS≠ for oligo-6b is somewhat greater than that for oligo-6a may reflect the difference between the association of -rC…*pdG- in oligo-6a and that of -rG…*pdG- in oligo-6b.
      By the way, on the course of the study on the degradation of dCrCdGdG, the degradation of EDAC was possible to be observed on the reversed-phase HPLC (Fig. 4) and the rate constants of the degradation of EDAC were determined (Table 2). The results show that the rate of the degradation of EDAC is much smaller than those of the cyclization reactions of oligo-6a and oligo-6b. Thus, the concentration of EDAC is regarded to remain constant during the cyclization reactions of oligo-6a and oligo-6b.

Table 3. Activation parameters for the formation and cleavage of phosphodiester bond.
Reactions
Ea,app / kJ mol-1ΔH≠ / kJ mol-1ΔS≠ / J mol-1
present system
cyclization

oligo-6a73 ± 171 ± 1-96 ± 1

oligo-6b82 ± 279 ± 2-70 ± 6
cleavage

dCrCdGdG99 ± 296 ± 2-58 ± 5
TD reaction a
formation

2-mer37 ± 534 ± 5-193 ± 17

3-mer74 ± 1471 ± 14-41 ± 44

4-mer85 ± 282 ± 2-3 ± 7
ImpG hydrolysis
82 ± 579 ± 5-40 ± 16
cleavage of oligo(G)
without poly(C)112 ± 11109 ± 1142 ± 32
with poly(C)109 ± 16106 ± 1617 ± 44
a Determined from the data shown in [13].


Figure 5. Temperature dependence of the cyclization of oligo-6 and the cleavage of ribose phosphodiester bond of dCrCdGdG. ○: oligo-6a, ●: oligo-6b, □: cleavage of dCrCdGdG in the presence of EDAC,■: that in the absence of EDAC [14].



      On the basis of previous mechanistic studies on the prebiotic formation of oligonucleotides with and without a poly(C) template [16,17,18], the reaction model of the phosphodiester bond formation from activated nucleotide and terminal ribose is generally expressed in following equations,

     -NOH + *pN- ⇆ -NOH…*pN-     (3)
     -NOH…*pN- → -NpN-           (4)

where -NOH and *pN- are the terminal group of elongating oligomer and the activated terminal phosphate group, and -NOH_*pN- indicates the association between -NOH and *pN-. The equilibrium constant of the outersphere association (3) is Kass and the rate constant of the phosphodiester bond formation from the associate is kphos. Thus, the aver all the reaction rate constant kform is expressed in equation (5) [18].

     kform = kphos Kass                (5)

It has been elucidated that the enhancement of the association (3) is important in both the TD reaction and the oligonucleotide formation on the clay catalyst [16,17]. In addition, the phosphodiester bond formation (4) can be promoted on the poly(C) template or the clay surface. The enhancement of the processes (3) and (4) in the present cyclization reaction does not occur since neither the template nor the clay catalyst is applied.
      Here, it is known that the formation of dimers and trimers from nucleoside 5'-phosphorimidazolide monomers (ImpN) in the absence of the template or the clay catalyst is much slower than that in the presence; the rate constants are in the range 7.3 x 10-7 - 2.9 x 10-6 s-1 M-1 for dimers and 8.9 x 10-6 - 2.0 x 10-5 s-1 M-1 for trimers at 25 °C [17]. Since the unit of these rate constants (s-1 M-1) is different from that for the cyclization (s-1), the direct comparison of these values is meaningless. Nevertheless, the cyclization is obviously more efficient than the formation of 2-mer and 3-mer from ImpN in the absence of the clay catalyst or the template. While the reaction yield of the cyclization reactions is reached to 50 %, that of the formation of dimer and trimer is normally less than 5 %. This is probably due to that the cyclization is an intramolecular reaction and the formation of 2-mer and 3-mer is an intermolecular reaction in the absence of the clay catalyst or the template. In addition, the existence of surrounding sequences of -NOH…*pN- may enhance the association in the cyclization reaction compared with the formation of 2-mer and 3-mer from ImpN on a poly(C) template since the association of 2 - 3 monomer molecules with elongating oligomer on a poly(C) template is important for the TD reaction [16]. A molecular model visualizes that oligo-6a and oligo-6b form easily this type of associate when the activated phosphate terminal approaches the terminal ribose (Fig. 1). This may result the strong association between both the terminal residues.
      In our previous study, oligo-6a and oligo-6b were chosen since it was expected that these oligomers could yield the dimers of oligo-6a and oligo-6b other than cyclic-oligo-6; oligo-6a and oligo-6b are possible to form self-complementary duplex. The Tm values under the same condition used for the cyclization were 64.2 °C for d(pGCGCGC) and 32.5 °C for d(pGGCCCG). However, the yield of dimers was very low and the main products were cyclic-hexanucleotides. The values of Tm may reflect the difference of the rate constants of kcyc between oligo-6a and oligo-6b at 25 - 50 °C, but the difference disappeared at 75 °C. In addition, the extent of the dimerized products of oligo-6a and oligo-6b decreased with increasing temperature, where the dimers of oligo-6a and oligo-6b were observed at 0 °C, but the dimers were not possible to be identified at over 25 °C. Here, the rate constants of the cyclization of oligo(C) were evaluated in the same buffer solution used for the cyclization reactions of oligo-6a and oligo-6b. Although oligo(C)s do not form the self-complementary duplex, the values of kcyc of oligo(C)s are not much different from those of oligo-6a and oligo-6b (Table 4). Conclusively, the influence of the formation of self-complementary duplex is very weak on the cyclization of the oligomers. A possible mechanism including 2 processes is proposed for the cyclization; the terminal phosphate of oligonucleotide is converted to phosphorimidazolide in the presence of imidazole and EDAC as the first step, and then the cyclization occurs. The second step would be much faster than the first step since the imidazolide of oligonucleotides was not detected during the formation of cyclic-oligonucleotides. During the organic synthetic preparation of imidazolide of some oligonucleotides by the method of Joyce et al. [19], the cyclic-oligonucleotides was formed efficiently and instantly [20]. The reactivity to form the cyclic-oligonucleotides in organic solvent may reflect the fast cyclization of oligonucleotides in aqueous solution.
Table 4. The rate constants for the cyclization of oligo(C).
oligo(C)kcyc /s-1
p(C)5rC(1.2 ± 0.1) x 10-5
p(C)6rC(1.2 ± 0.1) x 10-5
p(C)7rC(2.0 ± 0.1) x 10-5
Reaction conditions, [NaCl] = 0.2 M, [MgCl2] = 0.075 M, [imidazole] = 0.1 M, [EDAC] = 0.2 M, [oligo(C)] = 1 x 10-4 M, pH = 8.0, 25 °C.

On the Chemical Evolution of RNA
      The cyclization of hexanucleotide is regarded as a simple model of the phosphodiester bond formation. The reaction was possible at 75 °C and would be possible at higher temperatures, where the cyclization is much faster than the cleavage of the ribose phosphodiester bond. Besides, our previous study on the TD reaction displayed that the rate constant of the formation of 3-mer and longer oligo(G)s is much greater than that of the degradation of oligo(G); the low efficiency of the TD reaction at 80 °C and higher is mainly due to the relatively small rate constant of the 2-mer formation from two monomers. Thus, the fact that the cyclization of oligonucleotides is possible at elevated temperatures is not surprising. In the TD reaction, the rate constant of the formation of 4-mer and longer at 60 °C is 59 times greater than that of the hydrolysis of single-strand oligo(G) and 688 times greater than that of double-strand oligo(G) with poly(C) [13]. The difference of the rates between the cyclization and degradation in the present system is comparable to that between the formation and degradation of oligo(G) in the TD reaction.
      The comparison of the activation parameters of the cyclization and the TD reaction is also useful to find similarity between the present system and the TD reaction system. The activation parameters were shown in Table 3. Since the activation parameters were not determined in our previous study on the TD reaction [13], the values of ΔH≠ and ΔS≠ were calculated from the rate constants and briefly discussed in the present paper (Table 3). The enthalpy change for the formation of phosphodiester bond in the TD reaction is of greater advantage than the degradation of oligo(G). A trend that ΔS≠ for the formation of oligo(G) increases in the order 2mer < 3-mer < 4-mer indicates that the formation of 4-mer is easier than the 2-mer formation from the viewpoint of the entropy change. This is in agreement with the conventional TD reaction model given in equations (3) and (4), in which the association between the elongating 3-mer and the activated monomer on a poly(C) is easier than that between two activated monomers. Besides, the comparison of the cyclization of oligo-6a or oligo-6b with the TD reaction indicates that the magnitude of ΔH≠ of the cyclization is similar to that of the formation of 3-mer and 4-mer in the TD reaction. The magnitude of Ea,app of the hydrolysis of dCrCdGdG is greater than those of the cyclization of oligo-6a and oligo-6b. The same trend is observed in the relationship between the values of Ea,app for the formation and degradation in the TD reaction system. Besides, the magnitude of ΔS≠ of the cleavage of dCrCdGdG is much different from that of oligo(G). This is probably due to the presence of Zn2+ in the TD reaction [13], in which the cleavage of oligo(G) obeys pseudo-second-order kinetics. Conclusively, the thermodynamic parameters displayed similarity between the present system and the TD reaction.
      The present study showed that the cyclization of hexanucleotides is highly efficient. This is consistent with the fact that cyclic-2-mer and longer cyclic-oligomers readily form from ImpN in the presence of clay catalysis or in the presence of metal ion catalyst, in which the yield of cyclic-3-mer from uridine 5'-phosphorimidazolide or inosine 5'-phosphorimidazolide on montmorillonite clay catalyst is notably high [17,21]. From the viewpoint of the formation of RNA under the primitive earth conditions, the cyclization of oligonucleotides is regarded as the termination pathway. Thus, it should be noted that the cyclization could have been disadvantageous for the formation of oligonucleotide under the primitive earth conditions. From this view, the elongation using the activated mononucleotide monomers and non-activated primers should have been superior to that using the activated oligonucleotide monomers under the primitive earth conditions.
      The extrapolation of the temperature dependence of the rate constants of the cyclization vs. degradation in the present study and the oligonucleotide formation vs. degradation of the TD reaction [13] displayed that the rate of the formation of oligonucleotides would be much faster than that of the degradation at over 100 °C although the difference between the formation and degradation rates gradually decreases with temperature. These facts imply that the phosphodiester bond formation from the associate between the activated terminal residues shown in equation (4) would be possible for the case of unimolecular cyclization at high temperatures. Ideally, the magnitude of kphos would not be different between the cyclization and dimerization of oligonucleotides. Besides, the low efficiency of the dimerization of oligonucleotides is probably due to the small association constant (Kass) between the two oligonucleotide molecules. Thus, if the association (Kass) between the terminal residues for the dimerization is selectively enhanced, the relative rate of the dimerization of oligonucleotide could be increased in the present system. Furthermore, the consecutive elongation of oligonucleotides with activated mononucleotide monomers in the presence of a polynucleotide template or a clay catalyst would be possible at high temperatures if the association between the elongating oligonucleotide and the activated monomer is effective at high temperatures. Naturally, it is reasonable that the association between nucleotide monomer and oligomers become weak without any additives at high temperatures. However, this problem of the weak association at high temperatures is possible to be solved indeed in hyperthermophiles. In these organisms, enzymes or other molecules should be adapted to enhance the association between the elongating oligonucleotide and the activated monomer during the phosphodiester bond formation. Thus, it will be important to search possible prebiotic enzymes and/or reaction environments to promote the association under primitive earth conditions.

Conclusions

In this study, the temperature dependence of the rate constants of the cyclization of hexanucleotides was studied. The cyclization of hexanucleotides is more than 50 times faster than the degradation of phosphodiester bond at 75 °C. The magnitude of ΔH≠ of the cyclization is consistent with that of the formation of 3-mer and longer oligomers in the TD reaction. The difference between kcyc and kdeg supports that the prebiotic RNA formation such as the template-directed reaction and the clay catalyzed formation of oligonucleotides could have been sufficiently fast at high temperatures if the association between the elongating oligomer and the activated nucleotide monomer was facilitated by prebiotic enzymes only for the case of oligomerization.

Acknowledgement

Mr. Takayuki Tsuji helped the measurement of the rate of the cyclization of oligocytidylic acids. We thank Professor Harumi Fukada for the measurement of DSC in College of Agriculture and Professor Taketoshi Nakahara for the use of HPLC in Department of Applied Chemistry in Osaka Prefecture University. This research was supported by a Grant-in-Aid for Scientific Research(C) (1550150) from Japan Society for the Promotion of Science (JSPS) and the Sumitomo Foundation 1998.

References

[1] Gilbert, W. The RNA World, Nature 319, 618 (1986); Gesteland, R. F. and Atkins J. F. Ed., The RNA World, Cold Spring Harbor Laboratory Press, New York, 1993.
[2] Lohrmann, R. and Orgel, L. E. Reactions of adenosine 5'-phosphorimidazolide with adenosine analogs on a polyuridylic acid template, J. Mol. Biol. 113, 193-198 (1977); Lohrmann, R. and Orgel, L. E. Efficient catalysis of polycytidylic acid-directed oligoguanylate formation by Pb2+, J. Mol. Biol. 142, 555-567 (1980); Inoue, T. and Orgel, L. E. Oligomerization of (guanosine 5'-phosphor)-2-methylimidazolide on poly(C), An RNA polymerase model, J. Mol. Biol. 162, 201-217 (1982); Inoue, T. and Orgel, L. E. A nonenzymatic RNA polymerase model, Science 219, 859-862 (1983).
[3] Sawai, H., Totsuka, S., Yamamoto, K. and Osaki, H. Non-enzymatic template-directed ligation of 2'-5' oligoribonucleotides. Joining of a template and a ligator strand, Nucleic Acids Res. 26, 2995-3000 (1998).
[4] Sawai, H., Kuroda, K. and Hojo, H. Uranyl ion as a highly effective catalyst for internucleotide bond formation, Bull. Chem. Soc. Jpn. 62, 2018-2023 (1989); Sawai, H., Higa, K. and Kuroda, K. Synthesis of cyclic and acyclic oligocytidylates by uranyl ion catalyst in aqueous solution, J. Chem. Soc., Perkin Trans. 1992, 505-508.
[5] Ferris, J. P. and Ertem, G. Oligomerization of ribonucleotides on montmorillonite: Reaction of the 5'-phosphorimidazolide of adenosine, Science 257, 1387-1389 (1992); Ertem, G. and Ferris, J. P. Synthesis of RNA oligomers on heterogeneous templates, Nature 379, 238-240 (1996).
[6] Terfort, A. and von Kiedrowski, G. Self-replication by condensation of 3-amino-benzamidines and 2-formylphenoxyacetic acids, Angew. Chem. Int. Ed. Engl. 31, 654-656 (1992); von Kiedrowski, G. A Self-replicating hexadexynucleotide, Angew. Chem. Int. Ed. Engl. 25, 932-935 (1986); Sievers, D. and von Kiedrowski, G. Self-replication of complementary nucleotide-based oligomers, Nature 369, 221-224 (1994).
[7] Kanaya, E. and Yanagawa, H. Template-directed polymerization of oligoadenylates using cyanogen bromide, Biochemistry 25, 7423-7430 (1986).
[8] Sawai, H. and Wada, M. Nonenzymatic template-directed condensation of short-chained oligouridylates on a poly(A) template, Origins Life Evol. Biosphere 30, 503-511 (2000).
[9] Streltsov, S. A., Khorlin, A. A., Victorova, L. S., Kochetkova, S. V., Tsilevich, T. L. and Florentiev, V. L. Trivaline 'catalyzes' 5'-pdGTT oligomerization in solution, FEBS Lett. 298, 57-60 (1992).
[10] Kawamura, K. and Okamoto, F. Condensation reaction of hexanucleotides containing guanine and cytosine with water soluble carbodiimide, Nucleic Acids. Symp. Ser. 44, 217-218 (2000); Kawamura, K. and Okamoto, F. Cyclization and dmierization of hexanucleoides containing guanine and cytosine with water-soluble carbodiimide, Viva Origino 29, 162- 167 (2001).
[11] White, R. H. Hydrolytic stability of biomolecules at high temperatures and its implication for life at 250 °C, Nature 310, 430-432 (1984); Yanagawa, H. and Kojima, K. Thermophilic microspheres of peptide-like polymer and silicates formed at 250 °C. J. Biochem. 97, 1521-1524 (1985); Baross, J. A. and Hoffman, S. E. Submarine hydrothermal vents and associated gradient environments as sites for the origin and evolution of life, Origins Life 15, 327-345 (1985); Miller, S. L. and Bada, J. L. Submarine hot springs and the origin of life, Nature 334, 609-611(1988); Holm, N. G. ED., Special Issue-Marine Hydrothermal Systems and the Origin of Life; Origins Life Evol. Biosphere 22, 5-242 (1992) and therein.
[12] Kawamura, K. Inspection of the RNA world hypothesis from the viewpoint of the hydrothermal origin of life: Reconstruction of the scenario on the origin of life, Viva Origino 28, 129-138 (2000).
[13] Kawamura, K. and Umehara, M. Kinetic analysis of the temperature dependence of the template-directed formation of oligoguanylate from the 5'-phosphorimidazolide of guanosine on a poly(C) template with Zn2+, Bull. Chem. Soc. Jpn. 74, 927-935 (2001).
[14] Kawamura, K. Kinetics and activation parameter analyses of hydrolysis and interconversion of 2',5'- and 3',5'-linked dinucleotide monophosphate at extremely high temperatures, Biochim. Biphys. Acta 1620, 199-210 (2003); Kawamura, K. Kinetic analysis of the cleavage of the ribose phosphodiester bond within guanine and cytosine-rich oligonucleotides and dinucleotides at 65 - 200 °C and its implications concerning the chemical evolution of RNA, Bull. Chem. Soc. Jpn., 76, 153-162 (2003); Kawamura, K. Hydrolytic stability of ribose phosphodiester bonds within several oligonucleotides at high temperatures using a real-time monitoring method for hydrothermal reactions, Chem. Lett. 1120-1121 (2001); Kawamura, K., Kameyama, N. and Matumoto, O. Kinetics of hydrolysis of ribonucleotide polymers in aqueous solution at elevated temperatures: Implications of chemical evolution of RNA and primitive ribonuclease, Viva Origino 27, 107-118 (1999).
[15] Zagorowska, I., Mikkola, S. and Lonberg, H. Hydrolysis of phosphodiester bonds within RNA hairpin loops in buffer solutions: the effect of secondary structure on the inherent reactivity of RNA phosphodiester bonds, Helv. Chim. Acta 82, 2105-2111 (1999).
[16] Kanavarioti, A., Bernasconi, C. F., Alberas, D. J., and Baird, E. E. Kinetic dissection of individual steps in poly(C)-directed oligoguanylate synthesis from guanosine 5'-monophosphate 2-methylimidazolide, J. Am. Chem. Soc. 115, 8537-8546 (1993).
[17] Kawamura, K. and Ferris, J. P. Kinetics and mechanistic analysis of dinucleotide and oligonucleotide formation from the 5'-phosphorimidazolide of adenosine on Na+-montmorillonite, J. Am. Chem. Soc. 116, 7564-7572 (1994); Ding, P. Z., Kawamura, K. and Ferris, J. P. Oligomerization of uridine phosphorimidazolides on montmorillonite: A model for the prebiotic synthesis of RNA on minerals, Origins Life Evol. Biosphere 26, 151-171 (1996); Kawamura K. and Ferris, J. P. Clay catalysis of oligonucletide formation: Kinetics of the reaction of the 5'-phosphorimidazolides of nucleotides with the non-basic heterocycles uracil and hypoxanthine, Origin Life Evol. Biosphere 29, 563-591(1999). [18] Kawamura, K., Kuranoue, K. and Umehara, M. Chemical evolution of RNA monomers and RNA polymers: Implications from search for the prebiotic pathway of formation of RNA from adenosine 5'-triphosphate in the presence of thermal condensation products of amino acids as primitive enzymes, Viva Origino 30, 123-134 (2002).
[19] Joyce, G. F., Inoue, T. and Orgel, L. E. Non-enzymatic template-directed synthesis on RNA random copolymers, Poly(C,U) templates, J. Mol. Evol. 176, 278-306 (1984).
[20] Kawamura, K. unpublished data.
[21] Prabahar, K. J. Cole, T. D. and Ferris, J. P. Effect of phosphate activating group on oligonucleotide formation on montmorillonite: The regioselective formation of 3',5'-linked oligoadenylate, J. Am. Chem. Soc., 116, 10914-10920 (1994); Prabahar, K. J. and Ferris, J. P. Adenine derivatives as phosphate-activating groups for the regioselective formation of 3',5'-linked oligoadenylates on montmorillonite: Possible phosphate-activating groups for the prebiotic synthesis of RNA, J. Am. Chem. Soc., 119, 4330-4337 (1997).