Expeditious, potentially primordial, aminoacylation of nucleotides

Article abstract of DOI:10.1002/anie.200501591

(Chemical Equation Presented) In the beginning: Mixed carboxylic phosphoric anhydrides 3, formed from 3′-nucleotides 1 and amino acid N-carboxyanhydrides 2, undergo competing rearrangement to 2′-aminoacyl esters 4 and cyclization to 2′,3′-cyclic phosphates 5. The intramolecular aminoacyl transfer is faster than the cyclization despite the ease with which 2′,3′-cyclic phosphates are formed through any other form of phosphate activation.

Full text of DOI:10.1002/anie.200501591

Angewandte
Chemie
Nucleotides
generate 5’-aminoacyladenylates (5’-aa-AMP) 1 (A = ade-
nine) through an initial nucleophilic attack of an amino acid
carboxylate on ATP. The aa-RS then catalyzes the intermo-
lecular aminoacyl transfer from 1 to the 2’/3’-terminus of a
cognate tRNA to give adenosine-5’-monophosphate (5’-
AMP) (2) and an equilibrating mixture of 3’- and 2’-aa-
tRNAs 3 and 4 (Scheme 1).
DOI: 10.1002/anie.200501591
Expeditious, Potentially Primordial,
Aminoacylation of Nucleotides**
Jean-Philippe Biron, Alastair L. Parkes, Robert Pascal,*
and John D. Sutherland*
In contemporary biochemistry, enzymatically synthesized,
activated aminoacyl-tRNA esters (aa-tRNA) serve as sub-
strates for coded peptide synthesis by ribosomes,[1] and a
major goal is to understand the evolutionary path through
which this process arose. As a first step towards this goal, we
have been looking to find a prebiotically plausible means for
the aminoacylation of ribonucleotide 2’-/3’-hydroxy groups. A
number of simple, potentially prebiotic, activated amino acid
derivatives have been reported, most notably N-carboxyan-
hydrides (NCAs).[2] The case for NCAs in prebiotic evolution
has recently been strengthened by the finding that they can be
produced from amino acids through the action of the simple
volcanic gas carbonyl sulfide.[3]
We therefore decided to investigate the aminoacylation of
nucleotides by NCAs. Initially we were concerned that the
nucleobase amino groups of adenine and cytosine would
prove more nucleophilic than the 2’-/3’-hydroxy groups;
however, recent findings show that NCAs react with inorganic
phosphate to give aminoacyl phosphates.[4] This suggests that
it might be possible to generate nucleotide aminoacyl esters
by initial aminoacylation of the phosphate monoester,
followed by intramolecular aminoacyl transfer. Nucleotide
aminoacylation by way of intermediate carboxylic phosphoric
anhydrides would be analogous to the chemistry of amino-
acyl-tRNA synthetase (aa-RS) enzymes. These enzymes
Scheme 1. The biochemistry and chemistry of 5’-aa-AMP 1: biochemi-
cally (upper), 1 is used by aa-RS enzymes for the aminoacylation of
cognate tRNA. In the absence of enzymes (lower), 1 can undergo
intramolecular aminoacyl transfer and hydrolysis reactions.
When separated from the protective environment of an
aa-RS, 5’-aa-AMPs are highly unstable and undergo hydrol-
ysis and isomerization in aqueous solution.[5,6] The isomer-
ization involves a slow initial intramolecular aminoacyl
transfer from the 5’-phosphate to the 3’-hydroxy group via
an eight-membered transition state to give the 3’-aminoacyl
ester 5, followed by rapid equilibration of the latter with a 2’-
aminoacyl ester. If it were possible to generate 5’-aa-AMP 1
from 5’-AMP 2 by reaction with an NCA, then this intra-
molecular transfer would hopefully result in the sought-after
formation of aminoacyl esters.
[*] J.-P. Biron, Dr. R. Pascal
Dynamique des Syst›mes BiomolØculaires Complexes UMR 5073-
CC 017
DØpartement de Chimie, UniversitØ Montpellier 2
Place Eug›ne Bataillon, 34095 Montpellier (France)
Fax: (+33)46-763-1046
E-mail: rpascal@univ-montp2.fr
Dr. A. L. Parkes, Prof. Dr. J. D. Sutherland
Schoolof Chemistry, The University of Manchester
Oxford Road, Manchester M139PL (UK)
Fax: (+44)161–275–4939
E-mail: john.sutherland@manchester.ac.uk
We first investigated whether the aminoacylation of a
simple model phosphate monoester with an NCA was
possible. We chose methyl phosphate 6 as the model
phosphate monoester and the valyl derivative 7 to allow
comparison with the previous work on inorganic phosphate[4]
and because 7 can be easily prepared and stored (Scheme 2).[7]
In a general sense, the reaction of a phosphate monoester
with any electrophile is faster at pH values above 7 when the
phosphate is in its dianionic state, (pKa for the monoanionic
state ꢀ 6–7 (Supporting Information)); however, NCA
hydrolysis/polymerization is also favored at high pH values.
[**] This work was carried out within working group D27/001/02 of the
EU COSTaction “Prebiotic Chemistry and Early Evolution”, and was
funded by COST and the Engineering and PhysicalSciences
Research Council. We thank Dr Jo Kirkpatrick (Oxford University) for
the LC/ESI MS analysis and Jean-Pascal Gimeno for experimental
assistance in the chemistry of the mixed anhydride 8. We gratefully
acknowledge Peter Strazewski (UniversitØ Claude Bernard, Lyon 1)
for his helpful comments on the manuscript and Auguste
Commeyras (UniversitØ Montpellier 2) for facilitating the initiation
of this collaborative project.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2005, 44, 6731 –6734
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6731

Communications
Scheme 2. Aminoacylation of model phosphate monoester by Val-NCA
7.
To slow such degradative reactions while still allowing
reaction of the NCA with the phosphate group of 6, we
conducted our experiments under very slightly acidic con-
ditions. In this case, although 6 was predominantly mono-
anionic, some of the reactive dianionic form still existed.
Using 1H NMR spectroscopy for analysis, we observed a low-
level conversion of 6 and 7 into l-valyl methyl phosphate 8
(not isolated, according to 1H NMR integration ꢀ 23% based
on 7, see Supporting Information). The conversion was only
transient, however, and 8 was gradually converted back into
the starting methyl phosphate 6 along with valine and
oligovaline.
Scheme 3. Aminoacylation/cyclization of 3’- and 2’-nucleotides by Val-
NCA 7.
These observations imply that the reaction of 6 and 7 to
give 8 is an equilibrium that is rapidly established and lies in
favor of 6 and 7. The equilibrium is eventually dissipated
through the hydrolysis of 7 and 8 to valine, and the subsequent
reaction of the valine with further 7 and 8 giving oligovaline.
The existence of the proposed equilibrium was proven when
we found that 8, as prepared by conventional synthesis, was
almost completely converted into 6 and 7 in the presence of
bicarbonate. Once again, hydrolysis/polymerization of the
activated valine derivatives subsequently yielded the free
amino acid and oligomers.
Undeterred by this concern, we initially studied the
reaction of cytidine-3’-monophosphate (3’-CMP) 11 (base
(B) = cytosine (C)) with Val-NCA 7, again under very slightly
acidic conditions. After 55 min, no new signals consistent with
9 (B = C) were observed in the 1H NMR spectrum. However,
new signals were observed which were tentatively assigned to
the aminoacyl-transfer product 13 (B = C) (Figure 1a)). In
We next turned our attention to the reaction of nucleo-
tides with Val-NCA 7. 5’-AMP 2 showed similar behavior to
methyl phosphate 6, and we observed an equilibrium
production of 5’-Val-AMP 1 (R = iPr). However, the equilib-
rium was dissipated by NCA hydrolysis/polymerisation
before the rearrangement of 1 into 5 occurred (see Supporting
Information).[5,8] We next studied the reaction of other
regioisomeric nucleotides with Val-NCA 7. We recognized
that if a nucleoside 3’/2’-valyl phosphate 9/10 could be
transiently formed from the corresponding 3’/2’-nucleotide
11/12 and 7, then intramolecular aminoacyl transfer to the
2’/3’-OH group, via a seven-membered transition state[9]
might now be faster than hydrolysis/polymerization of the
NCA or the intermediate mixed anhydrides (Scheme 3). We
reasoned, therefore, that even an unfavorable equilibrium
situation might lead to subsequent intramolecular aminoacyl
transfer, and that this transfer should then displace the
equilibrium according to Le Chatelierꢀs principle. There was a
major concern with the 3’/2’-nucleotides, however, as we
foresaw cyclization of the intermediate 3’/2’-aminoacyl phos-
phates 9/10 to give the 2’,3’-cyclic phosphate 15 as a likely
competing reaction (Scheme 3). Indeed, 2’/3’-phosphates can
be cyclized with great ease, to 2’,3’-cyclic phosphates via five-
membered transition states, by almost any form of phosphate
activation.[10]
Figure 1. 1H NMR spectra (500 MHz, D2O) of the products of the
reaction of 3’-CMP 11 (B=C) with Val-NCA 7. a) Spectrum acquired
after 1 h. b) Spectrum acquired after spiking the sample with authentic
15 (B=C) (0.5 mg).
particular, a diagnostic triplet between d = 5.4 and 5.5 ppm
was shown by COSYanalysis to correspond to 2’-H of the new
species, and the downfield shift of this signal was consistent
with a 2’-ester. Although 13 was not isolated,[11] integration of
the 1H NMR spectrum of the reaction mixture suggested that
it had been formed in ꢀ 9% overall yield (Table 1).
6732
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 6731 –6734

Angewandte
Chemie
Table 1: Yields of the products observed in the reactions of 3’/2’-
nucleotides 11/12 with Val-NCA 7.
supported by experiments with 2’-AMP 12 (B = A) and Val-
NCA 7 in which we found that both the 3’-ester 14 (B = A)
and the cyclic phosphate 15 (B = A) were formed in only trace
amounts. This marked contrast in aminoacylation behavior
between 3’- and 2’-nucleotides was not anticipated. It now
seems likely that the efficiency of the aminoacylation reaction
depends on the pKa value of the phosphate and 2’/3’-OH
groups, and possibly on conformational effects such as the
furanose ring pucker. The pKa value of the phosphate
presumably influences the efficiency of the intramolecular
aminoacyl-transfer step through a correlation with the leav-
ing-group ability. Although the absolute pKa values reported
in the literature for the individual nucleotides differ (Sup-
porting Information),[12–14] there is a consensus that the 3’-
phosphates are more acidic than the 2’-phosphates by
approximately 0.2 pKa units and this is consistent with the
more efficient conversion of 11 into 13. Within the 3’-
phosphate series, the greater yield of 13 (B = A) relative to 13
(B = C) is possibly related to the fact that aden-9-yl is a better
stabilizer for the 2’-oxyanion than any other aglycons.[15]
The results described herein indicate that uncoded amino-
acylation of 3’-nucleotides by NCAs is a prebiotically
plausible reaction. The chemical mechanism of this process
suggests that it should also be possible for 3’-phosphoryl
oligoribonucleotides to undergo aminoacylation by NCAs in
the same way, but with potential assistance by the ribonucleic
acid chain through stereocomplementary binding. It is there-
fore possible that ribozymes, capable of coded aminoacyla-
tion by NCAs might have evolved and could have been key
intermediates in the emergence of translation.[16] The use of
5’-aa-AMPs as intermediates in coded aminoacylation of
RNA is most likely a later evolutionary development.
Nucleotide
t
Starting
material
[%]
Valyl ester
(13 or 14)
[%]
Divalyl
ester
[%]
15
[%]
[min]
11 (B=C)
11 (B=A)
55
30
90
45
90
90.6
82.2
79.5
94.9
99.0
(13) 8.7
(13) 13.8
(13) 13.4
(14) 3.0
(14) 0.5
trace
3.9
5.6
0
0.7
trace
1.5
2.1
0.5
12 (B=C)
12 (B=A)
0
As our experiment was initiated at a pD value close to the
pKa of the monoanionic form of 11, the yield of 13, based
upon the reactive dianionic form of 11, is probably consid-
erably higher. This suggests that isomerization of the inter-
mediate 9 into 13 is highly efficient relative to the fast
backwards reaction of 9 with CO2. After further time had
passed (> 1 h), 1H NMR spectroscopic analysis suggested that
13 underwent very slow reversion to 11, presumably by a
combination of hydrolysis and nucleophilic attack by the
amino group of valine and valyl derivatives, including 13. The
1
minor signals in the H NMR spectrum (accompanying the
signals of 13) in the reaction of 11 and 7 were shown, by
spiking the reaction with an authentic sample, to be due to the
2’,3’-cyclic phosphate 15 (B = C) (Figure 1b)). However, this
anticipated by-product was formed in much lower amounts
(13/15 > 10:1), and we have not been able to prove the
structure of the second minor product definitively, but MS
and HMBC data strongly suggest that it is the 2’-divalyl
analogue of 13.
We next investigated the effect of the nucleobase on this
remarkable conversion and examined the reaction of 3’-AMP
11 (B = A) with Val-NCA 7. Again, we could not detect the
intermediate aminoacyl phosphate 9 (B = A) but observed
the aminoacyl-transfer product 13 (B = A) in a maximal
overall yield of ꢀ 14% after 30 min (Table 1). In this
experiment the putative 2’-divalyl analogue of 13 was also
formed in a greater quantity and, after 90 min, had accumu-
lated to the level of ꢀ 6%. On the basis of the amount of the
reactive dianionic starting material, the combined synthesis
efficiency of 13 (B = A) and the divalyl analogue is again,
probably considerably higher. Close examination of the
1H NMR spectra revealed the presence of the 2’,3’-cyclic
phosphate 15 (B = A) in low yield (13/15 > 8:1). It thus
appeared that the nucleobase had a relatively small effect on
the aminoacylation of 3’-nucleotides with Val-NCA, and as
such, we switched our attention to the isomeric 2’-nucleotides
with the expectation of equally efficient aminoacylation.
Through the use of the same conditions employed for the
3’-isomer, we found that cytidine-2’-monophosphate (2’-
CMP) 12 (B = C), in the presence of Val-NCA 7, gave the
3’-ester 14 (B = C) in only ꢀ 3% yield after 45 min (Table 1).
Not only was the aminoacyl-transfer product formed in low
yield, but it was accompanied by a comparable ( ꢀ 2%)
amount of the 2’,3’-cyclic phosphate 15 (B = C) (14/15 < 2:1).
The lower yield of the ester product and the higher yield of 15
suggest that aminoacyl transfer from 10 is less efficient than it
is from the 3’-phosphoryl isomer 9. This conclusion was
Received: May 10, 2005
Revised: June 27, 2005
Published online: September 27, 2005
Keywords: amino acids · cyclization · N-carboxyanhydrides ·
.
ribonucleotides · RNA
[1] P. Schimmel, Annu. Rev. Biochem. 1987, 56, 125.
[2] K. Wen, L. E. Orgel, Origins Life Evol. Biosphere 2001, 31, 241.
[3] L. Leman, L. Orgel, M. R. Ghadiri, Science 2004 306, 283.
[4] J.-P. Biron, R. Pascal, J. Am. Chem. Soc. 2004, 126, 9198.
[5] N. S. M. D. Wickramasinghe, M. P. Staves, J. C. Lacey, Biochem-
istry 1991, 30, 2768.
[6] N. S. M. D. Wickramasinghe, J. C. Lacey, Origins Life Evol.
Biosphere 1992, 22, 361.
[7] In preliminary work, we found that other NCAs (e.g. Ile-NCA)
behave in a manner similar to 7 and we are now extending the
study to NCAs that are most conveniently generated in situ in
water (e.g. Gly-NCA).
[8] J. C. Lacey, N. Senaratne, D. W. Mullins, Origins Life 1984, 15,
45.
[9] D. H. Rammler, Y. Lapidot, H. G. Khorana, J. Am. Chem. Soc.
1963, 85, 1989.
[10] L. E. Orgel, R. Lohrmann, Acc. Chem. Res. 1974, 7, 368.
[11] Our early attempts to isolate 13 were thwarted by its instability
in the presence of (oligo)valine at the pH value of the reaction
and over the time period required for HPLC. The successful
characterization of 13 by LC–MS (see the Supporting Informa-
Angew. Chem. Int. Ed. 2005, 44, 6731 –6734
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6733

Communications
tion) involved an acidic mobile phase, and we are now
attempting preparative separations under similar conditions.
[12] D. D. Perrin, Dissociation Constants of Organic Bases in
Aqueous Solution, Butterworths, London, 1965.
[13] S. C. Gupta, N. B. Islam, D. L. Whalen, H. Yagi, D. M. Jerina, J.
Org. Chem. 1987, 52, 3812.
[14] L. F. Cavalieri, J. Am. Chem. Soc. 1953, 75, 5268.
[15] I. Velikyan, S. Acharya, A. Trifonova, A. Fꢁldesi, J. Chattopad-
hyaya, J. Am. Chem. Soc. 2001, 123, 2893.
[16] V. Borsenberger, M. A. Crowe, J. Lehbauer, J. Raftery, M.
Helliwell, K. Bhutia, T. Cox, J. D. Sutherland, Chem. Biodivers-
ity 2004, 1, 203.
6734
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 6731 –6734