ABSTRACT
Peptide nucleic acids (PNAs) are analogs of nucleic acids in which the ribose-phosphate backbone is replaced by a backbone held together by amide bonds. PNAs are interesting as models of alternative genetic systems because they form potentially informational base paired helical structures. Oligocytidylates have been shown to act as templates for formation of longer oligomers of G from PNA G2 dimers. In this paper we show that information can be transferred from DNA to PNA. DNA C4T2C4 is an efficient template for synthesis of PNA G4A2G4 using G2 and A2 units as substrates. The corresponding synthesis of PNA G4C2G4 on DNA C4G2C4 is less efficient. Incorporation of PNA T2 into PNA products on DNA C4A2C4 is the least efficient of the three reactions. These results, obtained using PNA dimers as substrates, parallel those obtained using monomeric activated nucleotides.
It is widely believed that the familiar DNA/RNA/protein world was preceded by an RNA world, i.e. a world without coded polypeptides or DNA. There is, however, no agreement about the origin of the RNA world. The difficulty of envisaging prebiotic syntheses of nucleotides has led some authors to suggest that some simpler informational polymer preceded RNA on the primitive Earth (1 -3 ). This immediately raises the question of transitions between different genetic systems.
PNA is a polymer resembling a nucleic acid in its coding potential, but held together by amide bonds in place of the phosphodiester bonds of DNA and RNA (Fig. 1 ; 4 -6 ). PNA oligomers form double helical complexes with complementary DNA that are more stable than DNA double helices. In an earlier paper we have shown that the PNA sequence C10 is an effective template for synthesis of oligonucleotides with sequence Gn and that DNA C10, reciprocally, facilitates synthesis of PNAs Gn (7 ). In this paper we show that a number of heteropolymeric DNA templates can direct the synthesis of complementary PNA products. In an accompanying paper (8 ) we demonstrate information transfer in the opposite direction, from PNA to RNA. While we do not suggest that PNA was the precursor of RNA on the primitive Earth, we believe that these experiments establish, for the first time, the possibility of information transfer between very different `genetic' polymers.
Unless otherwise stated, all chemicals were reagent grade, purchased from commercial sources and used without further purification. Oligodeoxyribonucleotide templates were synthesized on a 391A DNA synthesizer (Applied Biosystems) in `trityl-on' mode. Oligonucleotides were deprotected in ammonia at 55°C in a Savant Oligoprep lyophilizer, then purified and detrytilated on OPC cartridges (Perkin Elmer). The purity of these oligomers was evaluated by HPLC analysis and in all cases they were found to be satisfactory for use without further purification.
The PNA dimer A2 was synthesized as follows using Boc-protected monomers from Perseptive Biosystems.Boc-aegA(Cbz)-aegA(Cbz)-OEt (1). Ethyl N-((N6-(benzyloxycarbonyl)adenine-9-yl)acetyl)-N-(2-Boc-aminoethyl)glycinate (20 mg, 0.036 mmol) was stirred vigorously in 10 ml neat trifluoroacetic acid (TFA) for 5 min. The TFA was removed in vacuo and the residue washed twice with dry diethyl ether. The compound was resuspended in dry DMF (3 ml) and N-((N6-(benzyloxycarbonyl)adenine-9-yl)acetyl)-N-(2-Boc-aminoethyl)glycine (19 mg), 0.036 mmol), 2-(benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (14 mg, 0.036 mmol) and diisopropylethylamine (13 µl, 0.072 mmol) were added. The reaction mixture was stirred for 1 h at ambient temperature. The title compound was isolated as a white powder upon addition of dry diethyl ether and used directly, without purification, in the next reaction.Boc-aegA(Cbz)-aegA (Cbz)-OH (2). 1 was taken up in THF (5 ml) and cooled to 0°C. 2 N NaOH (5 ml) was added dropwise and the reaction allowed to proceed for 15 min at 0°C and subsequently for 45 min at ambient temperature. The pH was adjusted to pH 3 and the title compound was extracted into dichloromethane. The organic phase was evaporated to dryness in vacuo. The title compound was purified by C18 RP-HPLC using a linear gradient of 0-50% acetonitrile in water (0.1% TFA). Yield 14 mg.NH2-aegA-aegA-OH (A2). 2 was taken up in 300 µl trifluoromethanesulfonic acid:TFA:m-cresol (1:8:1) and left at ambient temperature for 2 h. The title compound was isolated upon precipitation with 1 ml dry diethyl ether. The precipitate was washed five times with diethyl ether and dried in vacuo. The C2, G2 and T2 PNA dimers were made analogously. Mass spectral analyses (FAB+), calculated (found): A2, 569.5 (569.2); C2, 521.3 (521.2); G2, 601.6 (601.2); T2, 551.5 (551.2).
HPLC analysis of the reaction samples was performed on a RPC-5 column (4.6 × 250 mm) as previously described (17 ,18 ). Reaction products were eluted with a linear gradient of NaClO4 (pH 12, 0-0.06 M over 60 min) and monitored by UV absorption at 254 nm. Peak areas were integrated manually and in estimating yields allowance was made for differences in the extinction coefficients of the bases at pH 12. Product yields are given as the sum of yields of oligomers relative to the template if we do not explicitlystate otherwise. Our HPLC system does not allow us to detect oligomers with fewer than four G, T or U residues (18 ). Yields of oligomers in experiments involving PNA C2 or PNA A2, therefore, are lower limits that do not take account of heterotetrameric products. Peaks of G4, G6 and G8 in the figures were assigned by co-injection with authentic synthetic PNAs Gn of appropriate lengths. The assignment of peaks to oligomers GnA2 and GnC2 in Figures 3 and 4 was indirect. First we assumed that peaks that appeared only in reactions that included G2 or A2 or G2 and C2 but not in reactions with G2 alone were heteropolymers. Next, we argued that oligomers GnX2 should be retained less well on an anion exchange column than Gn + 2 oligomers, since the former carry n negative charges and the latter n + 2 (18 ). Our assignments follow from these assumptions. The assignment of peaks to GnT2 oligomers depends on our observation that GnT2 and Gn+2 oligomers which have the same charges have very similar retention times (18 ). Our HPLC analysis does not discriminate between GnX2 and X2Gn sequences. Base composition rather than the sequence order is intended when a formula is in parentheses.
Stock solutions of PNA dimers were prepared in deionized water at the following concentrations: 5 mM PNA G2; 5 mM PNA G2 + 5 mM PNA X2; 5 mM PNA G2 + 10 mM PNA X2; 2.5 mM PNA G4 + 5 mM PNA X2 (X = A, T or C). The solutions were used immediately after preparation. The stock solutions (5 µl) of PNA substrates, with or without 2.5 µmol template, were evaporated to dryness. Reaction was initiated by addition of 5 µl freshly prepared 0.2 M N-(3-dimethylaminopropyl)-N'-ethylcarbodiimide hydrochloride (EDC) solution in 0.4 M 2-methylimidazole buffer (pH 7.5 at room temperature). The reaction mixture was vortexed and left at room temperature for 12 h. Reactions were stopped by shock freezing and stored at -80°C until analyzed. Prior to HPLC analysis the samples were dissolved in 100 µl running buffer (2 mM Tris-HClO4, pH 12) (17 ). An aliquot of 20 µl of this solution was added to 1 ml running buffer (pH 12) and heated to 95°C for 15 min to hydrolyze side products formed by attack of EDC on G and T residues (7 ). Samples for identification of Gn products by co-injection were prepared by mixing an aliquot of the reaction mixture with an aliquot of a reference solution of PNA Gn prior to heating.
1 Scheme 1. The intramolecular rearrangement of PNA. The attack of the terminal amine on the neighboring sidechain leads to a product with a base transferred to the N-terminus. The reaction is pH dependent and proceeds faster under alkaline conditions.The HPLC chromatograms of the products from reactions on a DNA C10 template are shown in Figure 2 . The products from PNA G2 on this template confirmed previous published results (7 ) and yielded 86% of PNA Gn (Fig. 2 a). The addition of 1 equiv either PNA A2 (Fig. 2 b), PNA C2 (Fig. 2 c) or PNA T2 (Fig. 2 d) to PNA G2 had little effect on product distribution. In no case could we see evidence for misincorporation in excess of 3%. The total yield of longer oligomers from mixtures of PNA G2 and PNA C2 was only 75% of that observed in the other reactions. This inhibition may have been caused by association of PNA G2 and PNA C2 in solution.
The elution profiles of products from control reactions in the absence of a template are shown in Figure 3 a for PNA G2 (i), PNA G2/A2 in a 1:1 ratio (ii) and PNA G2/A2 with a 2-fold excess of A2 (iii). The predominant product peak in every case corresponds to G4; hexamers are less abundant and only traces of octamer are detectable. In the presence of PNA A2 we saw less heteromeric products containing A2 than would be expected if the bases were co-oligomerized with equal efficiency. We believe that PNA G2 dimers and longer G oligomers must associate in a configuration that favors homocondensation.
The HPLC traces of mismatch and read-through products of PNA G2 in the absence of PNA A2 on C4T2C4 (Fig. 3 b) show products up to the decamer in 21% yield. Incorporation of PNA A2 in the presence of PNA G2 on a DNA C4T2C4 template occurs efficiently and with high fidelity (Fig. 3 c). The yield of PNA A2-containing products is >68% when equal amounts of PNA G2 and A2 are used. The read-through of PNA G2 was only a minor side reaction, as can be seen by the ratio of the yield of Gn oligomers to that of A-containing oligomers. The read-through products formed <5% of the total yield. The area of the peak corresponding to the full-length product G4A2G4 relative to that for (G4A2) is noteworthy. Once PNA (G4A2) is formed it is rapidly extended to the full-length product by consecutive addition of two PNAs G2. The yield of G4A2G4 was 20%. When 2 equiv PNA A2 were used (Fig. 3 d) the proportion of read-through products decreased to <2%. The total yield of oligomers increased to >71%, but the amount of full-length product did not increase beyond 22%. In the presence of PNA A2 we see products longer than the decamer, which are presumably formed in a non-template reaction, perhaps due to dangling-base association of the PNA dimers with the full-length double helical product (20 ).
The elution profiles of the products from non-template reactions of a mixture of PNA G2 and PNA C2 (data not shown) were very similar to those illustrated in Figure 3 a. The predominant product peak is the tetramer, some hexamer is formed and we detected traces of octamer. The reaction produced mainly homo-oligomers of G.
Oligomerization of PNA G2 in the absence of PNA C2 but in the presence of template (Fig. 4 a) gave a total yield of oligomers of 22%. This yield was comparable with the yield we obtained with PNA G2 on C4T2C4 (Fig. 3 a). Substantial amounts of Gn oligomers up to the decamer were obtained along with only traces of the 12mer.
When an equimolar mixture of PNA C2 and PNA G2 was activated in the presence of DNA C4G2C4 template (Fig. 4 b) the total yield of oligomers was >29%. The distribution of products indicates that once PNA C2 is incorporated into the growing strand elongation proceeds rapidly to the full-length complementary transcript. PNA G2 mismatch incorporation to give G6 accounted for 35% of the hexameric products, but G8 was not formed in >2% yield. Thus correct PNA (G4C2) transcripts were elongated much more efficiently than G6. When 2 equiv PNA C2 were used in the template-directed reaction the yield of products increased to ~35%. The selectivity and product distribution were only marginally influenced by the increase in the amount of PNA C2.
The reaction of a mixture of PNA C2 and PNA G2 on a C4G2C4 template gave 18% of G8. The yield of all C2-containing products, including the dominant full-length transcript G4C2G4, was 12%. This was higher than the yield of PNA C2-containing products obtained in the corresponding reactions with PNA G2 and PNA C2.
Non-template reactions of mixtures of PNA G2 and T2 (data not shown) gave similar product distributions to those previously described for reactions of PNA G2 with PNA A2 or PNA C2. The tetramer was the major product, accompanied by some hexamer and a trace of octamer. Reaction of PNA G2 in the absence of PNA T2 on C4A2C4 gave a yield of 22% Gn oligomers.
The results of reactions obtained with PNA G2 and 1 equiv PNA T2 on the C4A2C4 template are difficult to interpret. T and G bases in the PNA products are both deprotonated at pH 12. Consequently, PNA Gn and PNA (Gn-2T2) have very similar retention times on RPC-5 (18 ). Shoulders on the Gn peaks and sometimes splitting of the Gn peaks may indicate the presence of products containing PNA T2, but this is not certain.
Reactions of PNA T2 and PNA G4 (Fig. 5 d) yielded G8 as predominant product. This system is of particular interest because the anticipated T-containing products, (G4T2) and G4T2G4, were well resolved from G4 and G8 respectively, so that incorporation of PNA T2 can be examined directly. About 4% full-length product G4T2G4 was formed, along with ~5% (G4T2).
The oligomerization of PNA G2 on a DNA C10 template has been shown to be an efficient reaction (7 ). The specific synthesis of homo-oligomers of G when a 1:1 mixture of PNA G2 and a second PNA dimer is activated in the presence of DNA C10 now shows that this reaction proceeds with high fidelity. Transcription of a DNA C4X2C4 template into its complement by reacting a mixture of PNA G2 and PNA (X')2, where X' is the complement of X, is efficient and has high fidelity only when PNA A2 is incorporated opposite template T2. Incorporation of PNA C2 opposite G2 is accompanied by extensive misincorporation of PNA G2. We do not have direct quantitative data for incorporation of PNA T2 opposite A2. Thus the order of efficiency of incorporation of PNA dimers on complementary DNA C4X2C4 templates is G2 > A2 > C2 > T2. This order is somewhat different from that observed in experiments with 5'-phosphorimidazolides of ribonucleotides, where the order was shown to be G > A [approx] C > T (21 ). The major difference is that A and C are incorporated with comparable efficiency using RNA substrates, whereas PNA A2 is incorporated much more efficiently than PNA C2. Incorporation of PNA C2 in the presence of PNA G2 is probably inhibited by hybridization of the two substrates in solution.
When DNA templates C4X2C4 (X = A, T or G) are incubated with G2 alone product oligomers G6, G8 and G10 are always formed. Elongation past the mismatch could plausibly occur in two ways. Either the template could loop out or the mismatched pair could be stabilized by non-Watson-Crick interstrand base pairing or intrastrand stacking in the product strand. Looping out of the product would lead to formation of oligomers up to the 8mer, while stacking of the mismatched base pairs would lead to oligomers up to the 10mer. Since the 10mer is always formed in significant amounts, it seems that stacking of mismatched base pairs is always important. However, looping out is not precluded and our data are not sufficient to allow us to estimate the relative importance of `mispairing' and looping out. Whatever the mechanism of misincorporation, it is clear that it competes effectively with faithful copying except in the case of a C4T2C4 template. When PNA G4 is substituted for PNA G2 the major product, irrespective of the nature of X, is G8. The absence of significant amounts of G12 in these experiments argues strongly in favor of the looping out mechanism for the G4 substrate.
All of our experiments were carried out at room temperature. It is possible, but we think unlikely, that higher fidelities could be achieved by working at lower temperatures. Unfortunately, this is not practical on account of the limited solubility of PNA. It is also possible that poly(A) would act as a template for oligomerization of PNA T2 at sufficiently low temperatures, although it is ineffective at room temperature (data not shown).
Information can clearly be transferred efficiently and with high fidelity from DNA to PNAs in the case of polypyrimidine templates. Information transfer from purine-containing DNA templates is less efficient and has lower fidelity, particularly for incorporation of T. These latter results parallel those obtained when ribonucleotides are polymerized on DNA or RNA templates (21 ).
This work was supported by NSCORT/EXOBIOLOGY grant no. NAGW-2881 from the National Aeronautics and Space Administration and a grant from The Danish National Research Foundation.We thank Aubrey R.Hill Jr for technical assistance and Sylvia Bailey for manuscript preparation.
This article has been cited by other articles:
REFERENCES
CiteULike 
Connotea 
Del.icio.us What's this?
U. J. Meierhenrich, G. M. Munoz Caro, J. H. Bredehoft, E. K. Jessberger, and W. H.-P. Thiemann
Identification of diamino acids in the Murchison meteorite
PNAS,
June 22, 2004;
101(25):
9182 - 9186.
[Abstract]
[Full Text]
[PDF]
This Article 
Abstract
Print PDF (85K)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (36)
Request Permissions
Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Articles by Schmidt, J. G.
Articles by Orgel, L. E.
Search for Related Content
PubMed
PubMed Citation
Articles by Schmidt, J. G.
Articles by Orgel, L. E.
Social Bookmarking
What's this?