Biomimetic aminoacylation of ribonucleotides and RNA with aminoacyl phosphate esters and lanthanum salts

Article abstract of DOI:10.1021/ja073976l

Aminoacylation of tRNA in cells involves activation of the amino acid as an aminoacyl adenylate, a mixed anhydride with AMP, which reacts with tRNA. We have now established that aminoacyl phosphate esters in the presence of lanthanide ions in water will acylate hydroxyls at the 3′-terminus of RNA or a simple nucleotide. By extension, this will permit synthetically aminoacylated tRNA to be produced in a single-step biomimetic process. The reactions of Boc-4-fluorophenylalanyl ethyl phosphate were followed by HPLC separation, MS, and ¹⁹F NMR analysis. In stoichiometric combination with lanthanum salts in aqueous buffer, Boc-4-fluorophenylalanyl ethyl phosphate rapidly produces 2′- and 3′-monoesters of cytidine and cytidine monophosphate. Reaction of the reagent with RNA in the presence of lanthanum and magnesium salts introduces a specifically detectable signal into the RNA, which is evidence of formation of the aminoacyl ester. When the same RNA is initially oxidized with periodate to convert the 3′-terminal vicinal diol to the cleaved dialdehyde, reaction with the aminoacyl phosphate no longer occurs as evidenced by the lack of a signal in the ¹⁹F NMR spectrum. The results are consistent with a requisite chelation mechanism in which lanthanum serves as a template for both the aminoacyl phosphate and the 3′-terminal diol of RNA and nucleotides. The coordinated diol will then react through specific base-catalyzed intramolecular addition of the alkoxide nucleophile to the acyl group of the aminoacyl phosphate. Assessment of the method with a single tRNA was also achieved using the fluorescent reagent N-dansyl-glycyl ethyl phosphate. Lanthanide-promoted aminoacylation at the 3′-terminus of tRNA^Phe is detected by the introduction of fluorescence (detected directly and by antibody-enhanced emission). This does not occur if the 3′-terminus is converted to the dialdehyde by reaction with periodate.

Full text of DOI:10.1021/ja073976l

Published on Web 12/04/2007
Biomimetic Aminoacylation of Ribonucleotides and RNA with
Aminoacyl Phosphate Esters and Lanthanum Salts
Svetlana Tzvetkova and Ronald Kluger*
Contribution from the DaVenport Chemical Laboratories, Department of Chemistry, UniVersity
of Toronto, Toronto, ON Canada M5S 3H6
Received June 1, 2007; E-mail: rkluger@chem.utoronto.ca
Abstract: Aminoacylation of tRNA in cells involves activation of the amino acid as an aminoacyl adenylate,
a mixed anhydride with AMP, which reacts with tRNA. We have now established that aminoacyl phosphate
esters in the presence of lanthanide ions in water will acylate hydroxyls at the 3′-terminus of RNA or a
simple nucleotide. By extension, this will permit synthetically aminoacylated tRNA to be produced in a
single-step biomimetic process. The reactions of Boc-4-fluorophenylalanyl ethyl phosphate were followed
by HPLC separation, MS, and 19F NMR analysis. In stoichiometric combination with lanthanum salts in
aqueous buffer, Boc-4-fluorophenylalanyl ethyl phosphate rapidly produces 2′- and 3′-monoesters of cytidine
and cytidine monophosphate. Reaction of the reagent with RNA in the presence of lanthanum and
magnesium salts introduces a specifically detectable signal into the RNA, which is evidence of formation
of the aminoacyl ester. When the same RNA is initially oxidized with periodate to convert the 3′-terminal
vicinal diol to the cleaved dialdehyde, reaction with the aminoacyl phosphate no longer occurs as evidenced
by the lack of a signal in the 19F NMR spectrum. The results are consistent with a requisite chelation
mechanism in which lanthanum serves as a template for both the aminoacyl phosphate and the 3′-terminal
diol of RNA and nucleotides. The coordinated diol will then react through specific base-catalyzed
intramolecular addition of the alkoxide nucleophile to the acyl group of the aminoacyl phosphate. Assessment
of the method with a single tRNA was also achieved using the fluorescent reagent N-dansyl-glycyl ethyl
phosphate. Lanthanide-promoted aminoacylation at the 3′-terminus of tRNAPhe is detected by the introduction
of fluorescence (detected directly and by antibody-enhanced emission). This does not occur if the 3′-
terminus is converted to the dialdehyde by reaction with periodate.
Introduction
In order to make the introduction of synthetic aminoacyl units
directly accessible, we sought a biomimetic process for the
Synthetically aminoacylated tRNAs were envisioned by Hecht
as the basis of a versatile method for ribosomal production of
proteins that contain unnatural amino acids.1 Schultz extended
this to include procedures for in ViVo incorporation of unnatural
amino acids into proteins.2 The reported synthetic procedures
for aminoacyl tRNA require preparation of amino-protected 3′-
aminoacylated diribonucleotides (derived from pCpA), followed
by enzymic ligation to tRNA(-pCpA) and deprotection. In the
cell, aminoacylation of tRNA is promoted by specific cytosolic
enzymes, aminoacyl tRNA synthetases, in a two-part reaction.
The enzymes promote an initial activation of the cognate amino
acid with ATP, producing the corresponding 5′-aminoacyl
adenylate (AA-AMP).3 In a coupled step on the same enzyme,
the aminoacyl moiety is transferred to the 3′-terminal hydroxyl
of the cognate tRNA. Structural and mechanistic analysis
suggests that the enzymic formation of the aminoacyl tRNA is
achieved simply by bringing the components into reactive
proximity while establishing proper alignment for reaction.
synthesis of aminoacyl tRNAs. We developed an aminoacyl
donor that is a functional analogue of the biochemical aminoa-
cyl-AMP intermediate, an aminoacyl alkyl phosphate, as
exemplified by phenylalanyl 5′-adenylate and the simplified
analogue, phenylalanyl ethyl phosphate.
Aminoacyl adenylates have been isolated in complexes with
aminoacyl tRNA synthetases and studied. In some cases, the
activated amino acid undergoes a spontaneous side reaction in
which it forms amides derived from amino groups of the
enzyme.4 In pioneering work on this class of materials Meister
prepared aminoacyl adenylates by condensation of an amino
acid and AMP, establishing that they react with hydroxylamine
and with proteins, as well as with ammonia, amino acids, and
ribonucleic acids.5
In order to direct the nonenzymic reaction specifically to the
3′-terminus of tRNA, we devised a template procedure that takes
advantage of the cis diol functional array of the 2′ and 3′
hydroxyl groups. Chelation of the diol by La3+ in a bis-bidentate
(1) Heckler, T. G.; Chang, L.-H.; Zama, Y.; Naka, T.; Chorghade, M. S.; Hecht,
S. M. Biochemistry 1984, 23, 1468-1473.
(4) Kern, D.; Lorber, B.; Boulanger, Y.; Giege, R. Biochemistry 1985, 24,
(2) Noren, C. J.; Anthony-Cahill, S. J.; Griffith, M. C.; Schultz, P. G. Science
1989, 244, 182-188.
1321-1332.
(5) Moldave, K.; Castelfranco, P.; Meister, A. J. Biol. Chem. 1959, 234, 841-
(3) Mulvey, R. S.; Fersht, A. R. Biochemistry 1978, 17, 5591-5597.
848.
9
15848
J. AM. CHEM. SOC. 2007, 129, 15848-15854
10.1021/ja073976l CCC: $37.00 © 2007 American Chemical Society

Aminoacylation of Ribonucleotides and RNA
A R T I C L E S
temperature for 1 h. The product was extracted with water, freeze-
dried, and used for aminoacylation without further purification.
BocFPheEP, a white hygroscopic solid, was thus prepared in 80% yield.
1H NMR (300 MHz, D2O): δ 7.1 (t, 2H, Ar), 6.9 ((t, 2H, Ar), 4.3
(q, 2H, POCH2CH3), 4.0 (t, 1H, CHCO), 3.85 (m, 2H, ArCH2), 3.1 (q,
8H, N+(CH2CH3)4, 1.2 (s, 9H, tBut), 1.1 (t, 3H, POCH2CH3), 1.0 (t,
12H, N+(CH2CH3)4); 13C NMR (75 MHz, D2O): δ 163.1 (CO-O-
PO), 160.5 (p-F-CAr), 157.3 (CO-O-tBut), 132.2 (p-F-ArC), 130.9
(Ar), 115.0 (Ar), 81.3 (C(CH3)3), 63.6 (P-O-CH2CH3), 55.9 (NH-
CH-CO), 52.1 (N+(CH2CH3)4), 36.2 (Ar-CH2), 27.6 (C(CH3)3), 15.6
(P-O-CH2CH3), 6.7 (N+(CH2CH3)4); 19F NMR (282 MHz, D2O): δ
-117.2; 31P NMR (121 MHz, D2O): δ -6.18; MS ESI (-): found
m/z 391.1203, calculated m/z 391.1196.
The isopropyl ester of N-Boc-4-fluorophenylalanine was synthesized
according to the method of Hassner and Alexanian.10 N-Boc-4-
fluorophenylalanine (1.0 mmol) was dissolved in 15 mL of dichlo-
romethane. DCC (1.1 mmol), isopropanol (1.1 mmol), and 4-pyrroli-
dinopyridine (0.1 mmol) were added. The solution was stirred at room
temperature for 2 h. Dicyclohexylurea was removed by filtration, and
the filtrate was washed with water, acetic acid, and again with water.
The isopropyl ester was dried over magnesium chloride, and the solid
was dried in vacuum.
1H NMR (300 MHz, CD3OD): δ 7.2 (t, 2H, Ar), 6.99 ((t, 2H, Ar),
4.97 (quintet, 1H, CH(CH3)2), 4.26 (t, 1H, CHNH), 2.92 (m, 2H,
ArCH2), 1.38 (s, 9H, tBut), 1.19 (dd, 6H, CH(CH3)2); 19F NMR (282
MHz, CD3OD): δ -118.9; MS ESI (+): found m/z 348.2 (325.17+
Na), calculated m/z 325.17 (+23.5).
Reactions of the aminoacyl phosphates were performed at room
temperature with stirring. All reacting components were used in
equimolar amounts. Lanthanide salts (triflate or trichloride) were the
last component added in order to minimize the competing lanthanum-
promoted hydrolysis of BocFPheEP prior to acylation.
Products were monitored by reversed phase HPLC at 263 nm and
then 19F NMR at 367 or 282 MHz in D2O, with shifts recorded relative
to CFCl3. When 19F NMR was used to follow the reactions, LaCl3 was
used as a catalyst since the triflate group contains fluorine atoms with
signals that interfere with those of the reactants and products. In cases
where the ester products were separated by HPLC, the residual TFA
from the mobile phase was removed by solid-phase extraction (PL-
HCO3 MP SPE Tubes, Polymer Labs) or acetic acid was used in its
place.
complex6 along with the aminoacyl alkyl phosphate sets up the
system for specific-base-catalyzed acylation (Scheme 1). The
potential for this type of reaction selectivity is consistent with
models in which esters form from the reaction of a single
hydroxyl in a 1,2-diol with benzoyl methyl phosphate and
lanthanide ions in water.7,8
We now report promising results on aminoacylation reactions
of nucleosides and nucleotides with Boc-4-fluorophenylalanyl
ethyl phosphate (BocFPhePEP), a fluorinated amino acid
derivative whose reaction we were able to follow by 19F NMR
spectroscopy. The results provide strong evidence that a direct,
selective reaction at the 2′ and 3′ hydroxyls is readily achieved.
We also report the successful direct extension of the aminoa-
cylation reaction to RNA.
Materials and Methods
Commercial reagents were used without further purification. RNA
type VI from Torula yeast and tRNAPhe from yeast were purchased
from Sigma. High-resolution mass spectrometry was performed at the
QStar Chemistry Mass Spectral Facility, University of Toronto. HPLC
analysis utilized a C18 reversed phase preparative column (7.8 mm ×
300 mm), eluting with 40% acetonitrile (HPLC grade) with 0.1%
trifluoroacetic acid in deionized water. The flow rate was 3.0 mL/min,
at room temperature. Eluting species were detected at 263 nm.
Boc-4-fluorophenylalanyl ethyl phosphate (BocFPheEP) was pre-
pared from the amino acid according to the general method previously
reported by Loo and Kluger.9 Ethyl dichlorophosphate (25 mmol) was
converted to the free acid by addition to 4.5 mL water over 10 min in
an ice-cooled round-bottom flask and then stirred for 1 h. Hydrogen
chloride that formed as a byproduct was removed by rotary evaporation
with vacuum. The resulting ethyl phosphoric acid was neutralized with
2 equiv of tetraethylammonium hydroxide. The neutral solution was
freeze-dried. Boc-4-fluorophenylalanine (BocFPhe, 1.62 mmol) was
activated with dicyclohexylcarbodiimide (DCC, 1.12 mmol) in dichlo-
romethane for 3 min. Tetraethylammonium ethyl phosphate (1.12 mmol)
in dichloromethane was added, and the mixture was stirred at room
Scale-up of the reaction with RNA: The reactants were dissolved
in 0.5 mM pH 8 EPPS. A 60 µL aliquot of RNA solution (20 mg/mL
buffer) was first mixed with 100 µL of 100 mM MgCl2 solution.
BocFPheEP and lanthanum trichloride were added to give a concentra-
tion of 20 mM. The final buffer concentration was 75 mM. Distilled
water was added to bring the reaction volume to 0.4 mL. Sephadex
G-25 spin columns were used for isolation of the RNA-esters.
Cleavage of the terminal diol of RNA was accomplished with sodium
periodate solution. RNA (9.3 mg) was dissolved in 0.5 mL of 100 mM
magnesium chloride solution and 0.5 mL of 0.5 M EPPS buffer. Sodium
periodate (5 mL of 0.1 M solution) was added, and the reaction was
kept in the dark at room temperature for 20 min.11 Excess sodium
periodate was precipitated by addition of about 1.0 mg of potassium
chloride at 0 °C.12 The supernatant was removed, and RNA was
precipitated with cold ethanol, collected, and dried.
Fluorescence Studies. Commercially supplied N-dansyl-glycine was
converted to the corresponding ethyl phosphates anhydrides according
to the procedure described above for BocFPheEP. Commercially
available purified yeast tRNAPhe was used as the reaction substrate.
Nuclease-free water (Fermentas) was used in all experiments with
tRNA.
(6) Clarke, P. A.; Arnold, P. L.; Smith, M. A.; Natrajan, L. S.; Wilson, C.;
Chan, C. Chem. Commun. 2003, 2588-2589.
(10) Hassner, A.; Alexanian, V. Tetrahedron Lett. 1978, 4475-4478.
(11) Proudnikov, D.; Mirzabekov, A. Nucleic Acid Res. 1996, 24 (22), 4535-
4542.
(12) Chernetskii, V. P.; Ponomareva, E. A.; Stavitskii, V. V. Khim. Geterotsikl.
Soedin. 1970, 6, 987-988.
(7) Kluger, R.; Cameron, L. J. Am. Chem. Soc. 2002, 124, 3303-3308.
(8) Cameron, L.; Wang, S.; Kluger, R. J. Am. Chem. Soc. 2004, 126, 10721-
10726.
(9) Kluger, R.; Li, X.; Loo, R. W. Can. J. Chem. 1996, 74, 2395-2400.
9
J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007 15849

A R T I C L E S
Tzvetkova and Kluger
Scheme 1. La3+ Chelation and Aminoacylation
Scheme 2. Reaction of 1 with Cytidine and Lanthanum Triflate Produces 2′- and 3′-Esters
Polyclonal rabbit anti-dansyl antibody (anti-DNS Ab) was obtained
from Molecular Probes. Aminoacylation of tRNA was carried out for
1 h at room temperature, as described for nucleotides and bulk RNA
in the preceding sections. After completion of the reaction, the products
were isolated and purified by passage through a G-25 column. tRNA
concentrations were determined spectrophotometrically based on extinc-
tion coefficient obtained from literature.
The dansyl (DNS) group has an excitation maximum at 330 nm and
an emission maximum at 550 nm. The excitation maximum of the DNS-
aminoacyl phosphates was determined to be at 327 nm, and the emission
maximum, at 540 nm. The instrumental settings were as follows:
increment, 1 nm; excitation slit, 1 nm; emission slit, 1 nm; integration
time, 1 s. Spectra were acquired from 400 to 650 nm. All measurements
were performed at room temperature in nuclease-free water solutions.
Antibodies were used to test further the introduction of the
fluorophore onto tRNA at a more sensitive level. An anti-DNS antibody
was used in the fluorometric studies to determine the presence of a
DNS-group on the tRNA after aminoacylation. It was supplied as a
solution (1 mg/mL in phosphate-buffered saline pH 7.2, containing 5
mM sodium azide). Based on information supplied with the reagents,
the maximum fluorescence enhancement for DNS is about 10-fold over
the fluorescence of the free dye. Steric interference with the covalently
bound fluorophore will lower the enhancement produced by antibody
binding to the fluorophore.
Oxidation of tRNAPhe with NaIO4 was performed as described for
bulk RNA by the method of Proudnikov and Mirzabekov.11 Fresh
solutions of 0.1 M NaIO4 were prepared immediately before the
reaction. A 5 µL aliquot of this solution was used to oxidize 0.1 mL of
5.5 × 10-5 M tRNA. The reaction was carried out in the dark for 20
min at room temperature. Excess NaIO4 was precipitated by addition
of KCl at 0 °C.12 The supernatant was removed and transferred to a
vial. The oxidized tRNA was then isolated by precipitation with cold
ethanol and centrifugation for 30 min at 13 000 rpm. The resulting
white pellets were dried and used for the aminoacylation test reaction.
The reaction mixture for the aminoacylation of yeast tRNAPhe
consisted of tRNAPhe, DNS-glycyl ethyl phosphate (DNSGlyEP), and
La(OTf)3 in equimolar concentrations (2.47 × 10-5 M), excess MgCl2
(10-3 M), buffered with pH 8, 3 × 10-5 M EPPS. The reaction was
performed in the dark at room temperature for 1 h. The resulting tRNA/
oxtRNA and DNS-aminoacyl-tRNA were isolated using the G-25 spin-
column. This treatment removes unreacted DNS-aminoacyl phosphate
as well as the product from its hydrolysis, the DNS-amino acid. This
treatment removes Mg2+ and La3+ as well, thus avoiding any possible
metal-catalyzed hydrolysis of the aminoacyl-tRNA.
Results
One of the challenges of assessing the success of the
aminoacylation reaction of a nucleotide (or RNA) is detection
of the product and establishment of the site of aminoacylation.
We obtain this information by first introducing an NMR-
detectable derivative that carries a label not normally found in
RNA. Reaction of BocFPheEP with RNA (Scheme 1) is
expected to give an ester that is detectable by its 19F NMR
signal. We assessed the specific site of the aminoacylation
reaction by comparing the reaction after periodiate oxidation
cleavage of the 3′-terminal diol of RNA. We find that the
resulting dialdehyde does not react with the aminoacyl phos-
phate, confirming the necessity of the intact vicinal diol for the
aminoacylation process. Because of the quantities of material
necessary for NMR analysis, we used bulk RNA for this part
of the study.
Aminoacylation of Nucleosides. The reaction of cytidine
with BocFPheEP (1) in the presence of 1 equiv of lanthanum
triflate occurs rapidly, producing a mixture of the 2′- and 3′-
BocFPhe esters of cytidine (2 and 3) along with BocFPhe from
the competing hydrolysis reaction of BocFPheEP (Scheme 2,
Figure 1). Analysis by HPLC indicates that the products form
within 30 s and remain intact thereafter (Figure 2).
The products were isolated and characterized as monoesters
of cytidine (ES MSI (+) m/z calculated 508, found 509.2).
Independent studies in our laboratory have revealed that the
2′-ester elutes ahead of the 3′-ester and that the concentration
of the 3′-ester is about twice that of the 2′-ester. The ratio 2′-
ester/3′-ester/BocFPhe was 1:2:0.5. There is no 5′-ester detected.
The reaction was also studied by 19F NMR (Figure 3) using
lanthanum chloride in place of lanthanum triflate. The reaction
was stopped with EDTA after 10 s. The signal corresponding
to BocFPheEP appears at δ -117.2. In addition, a new peak at
9
15850 J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007

Aminoacylation of Ribonucleotides and RNA
A R T I C L E S
Figure 1. 19F NMR (282 MHz) spectrum of BocFPheEP (δ -117.2) and
its hydrolysis product, BocFPhe (δ -117.8).
Figure 4. Aminoacylation of 5′CMP at room temperature. HPLC chro-
matograms at 60 min of reaction in 100 mM EPPS pH 8.
Figure 2. Aminoacylation of cytidine with BocFPheEP: Reversed phase
HPLC of product after 30 s reaction. Peak contents were analyzed by MS
and NMR. The initial peak is cytidine. The two subsequent peaks contain
the 2′ and 3′ BocFPhe esters of cytidine (see Scheme 2). The last peak is
BocFPhe.
Figure 5. Aminoacylation of 5′CMP at room temperature. HPLC chro-
matograms at 60 min of reaction in 10 mM EPPS, initially pH 8, then
lowered to pH 6.5.
HPLC and analyzed by MS ESI (-); m/z calculated ) 588.15,
found ) 587.2, indicating that only monoesters of 5′CMP are
1
formed. H NMR analysis revealed that the products are the
2′- and 3′-O-monoesters of 5′CMP. There is a shift in the signal
of the 1′ proton of the furanose ring that varies with the location
of the ester (2′ vs 3′). The 2′-ester signal appears as a doublet
at δ 6.14 while the 3′-ester signal appears as a doublet at δ
5.80. The esters rapidly equilibrate, with peaks from both esters
1
observed in the H NMR spectrum from an initial sample of
either ester. The 19F NMR spectrum of the reaction after 1 h
indicates a new peak at δ -116.98 (Figure 6), which is distinct
from the products of the hydrolysis reaction (Figure 7). 19F NMR
of the esters alone in the absence of La3+ and buffer (after HPLC
purification) gives a signal at δ -117.15. Further analysis of
the same product by ESI MS and 1H NMR identifies the material
as a mixture of the 2′- and 3′-esters of 5′CMP.
Figure 3. 19F NMR (282 MHz, D2O) spectrum of the reaction mixture of
10 mM cytidine, 10 mM BocFPheEP, 10 mM LaCl3, and 100 mM EPPS
pH 8, quenched with EDTA at 10 s. Peaks were identified as the 2′- and
3′-BocFPhe-cytidine esters (δ -116.9), BocFPheEP (δ -117.2), and
BocFPhe (δ -117.8).
In order to estimate the likely 19F NMR chemical shift of the
ribose ester products, we prepared the isopropyl ester of
BocFPhe. The chemical shift of the peak for the ester is at δ
-118.9, which is close to those signals assigned to the esters
from the nucleic acid derivatives.
δ -116.9 appears, while the known peak of the free acid,
BocFPhe, is at δ -117.8. The shoulders on the peaks of
BocFPheEP and BocFPhe (Figure 1) are mainly due to
aggregation during spectra acquisition or incorporation of
deuterium in the compound.
Aminoacylation of Nucleotides. Reactions of 5′CMP were
also evaluated. Reactions of 5′CMP in 100 mM pH 8 EPPS
were complete, as determined by the complete consumption of
BocFPheEP, in 40 min, and only one ester resulted. In contrast,
reactions conducted at pH 6.5 were not complete after 60 min,
consistent with the reaction being specific base catalyzed
(Figures 4 and 5). The products were isolated by preparative
Aminoacylation of RNA. Reactions with yeast tRNAPhe were
conducted using conditions as described for nucleotides. The
materials could not be analyzed by NMR due to the small
amounts (concentrations in the nano- to micromolar range). In
our experience, product analysis by 19F NMR requires at least
millimolar concentrations. In order to work at such concentra-
tions for product analysis, we used a readily available mixture
9
J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007 15851

A R T I C L E S
Tzvetkova and Kluger
Figure 9. (Left) 19F NMR spectrum of the solution from the reaction of
periodate-oxidized RNA with BocFPheEP in the presence of lanthanum
chloride after 1 h, prior to isolation of RNA. (Right) the 19F NMR spectrum
of the same amount of material after isolation of all RNA species via size-
exclusion chromatography. Trace peaks result from slightly exceeding the
column’s capacity.
Figure 6. 19F NMR (376 MHz, D2O) of the aminoacylation reaction
solution (30 mM LaCl3, 5′CMP, and BocFPheEP, 34 mM buffer pH 8)
after 1 h.
purification with the size exclusion chromatography at similar
intensities. Thus, we conclude that these are the signals of
products resulting from aminoacylation of RNA with BocF-
PheEP.
The chemical shifts of the NMR signals corresponding to
the products are similar to those of the BocFPhe-derived esters
of nucleosides and nucleotides. In order to test the requirement
for the terminal diol for aminoacylation, the subject RNA was
treated with excess NaIO4 to oxidize and cleave the 3′-terminal
diol. We assume that the reaction goes close to completion. The
oxidized RNA was combined with BocFPheEP and lanthanum
chloride in buffer containing excess magnesium chloride. The
hydrolysis product was removed as a precipitate by filtration.
The 19F NMR spectrum of the products after 1 h of reaction
time (Figure 9) was obtained by addition of 0.3 mL of D2O to
0.2 mL of the mixture that had been purified through a Sephadex
G-25 spin column.
Figure 7. 19F NMR (376 MHz, D2O) of the hydrolysis reaction of
BocFPheEP under the same conditions at 5 min of reaction as in Figure 6.
The initial spectrum consists of peaks that correspond to
residual BocFPheEP and its hydrolysis product, BocFPhe. The
material analyzed after purification by size exclusion chroma-
torgraphy has a 19F NMR spectrum that consists of only two
very small peaks with chemical shifts of BocFPheEP and
BocFPhe, indicating that no ester formation occurs (Figure 9).
Thus, oxidation converts RNA to a species that is predictably
unreactive under conditions where ester formation occurs readily
with material that has not been oxidized. These results strongly
support a conclusion that reaction occurs at the terminal diol in
the native RNA sample.
Another aspect of this study was the effect of the specific
lanthanide ion on the acylation reaction. Thus, we evaluated
La3+, Pr3+, Nd3+, and Yb3+ as well as Sc3+ and Mg2+ as
catalysts. Table 1 represents the percent conversion of 5′CMP
with each of these metal cations. While Mg2+ has high affinity
for phosphates, it does not coordinate effectively to the diol
leading to no formation of esters. Some Lewis acid properties
of Sc3+ compare to those of lanthanides; however, it provides
only a trace of ester from 5′CMP. In contrast, all lanthanides
effectively promoted formation of the ester with the efficiency
of the conversion decreasing with increasing atomic number
(Table 1). Specifically, the early lanthanides, exemplified by
La and Pr (atomic numbers 57 and 59), are better catalysts for
acylation reactions than Yb (atomic number 70). It is the earlier
lanthanides that have larger ionic radii (La 103 pm, Pr 99 pm,
Figure 8. Aminoacylation of bulk RNA with BocFPheEP in the presence
of La3+ 19F NMR spectrum of the reaction mixture after 1 h (left) and
:
after purification and dilution (right).
of RNA that contains RNAs of different sizes and types,
including tRNAs. As all types of RNA provide a terminal 3′-
diol moiety that is required to test the aminoacylation reaction,
this provides a test of the process. 19F NMR spectra were
recorded after reactions had proceeded for 1 h (Figure 8).
Samples were passed through a size-exclusion column (Sepha-
dex G-25) to remove salt, buffer, and starting material (Figure
8). Most of the free amino acid from hydrolysis was removed
by filtration before the spectra were recorded. The spectra after
purification indicate two large peaks at δ -117.19 and δ
-117.51 (Figure 8) in the region where the signals from the
model esters were observed, which are also present prior to
9
15852 J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007

Aminoacylation of Ribonucleotides and RNA
A R T I C L E S
Figure 10. Emission spectrum of the product of reaction of yeast tRNAPhe
with DNSGlyEP and lanthanum triflate at pH 8 in absence (lower) and
presence (upper) of a DNS-specific antibody.
Figure 11. Emission spectra of the products from reaction of periodate-
oxidized yeast tRNAPhe with DNSGlyEP and lanthanum triflate in the
absence and presence of anti-DNS antibody.
Table 1. Relative Conversion of 5′CMP to Esters of BocFPhe as
Promoted by Different Metal Ions; the Amounts Are Based on
HPLC Peak Integration
lanthanum. Using the same reaction conditions, the lack of
antibody enhanced fluorescence (Figure 11) in comparison with
the untreated sample shows that the principal site of aminoa-
relative
conversion
cylation is the 3′-terminus of tRNAPhe
.
ion
atomic number
of 5
′
CMP, %
Mg2+
Sc3+
La3+
Pr3+
12
21
57
59
60
70
0
0.2
31
30
29
21
Discussion
These results show that lanthanum salts effectively direct
aminoacylation by aminoacyl phosphate esters to the 3′-terminal
of tRNA in direct analogy to the biochemical process that is
accomplished by enzymic binding. Since the diol and the acyl
phosphate functional groups are the coordinating entities, it
appears that the method will extend to a large variety of
aminoacyl phosphates and RNA. (We also find that amino group
protection is not necessary for the reaction of the acyl phosphate,
although the protected material will last for a considerably longer
time in storage. These results are currently the subject of
extended studies.)
Nd3+
Yb3+
Nd 98.3 pm). Established aspects of the coordination chemistry
of lanthanides suggest that larger ions would coordinate more
readily with the diol due to the size of the span between the
hydroxyls. Later lanthanides have smaller radii (Yb 86.8 pm),
and greater positive charges will make them more selective for
phosphates (and not diols). Thus, where the goal is the
hydrolysis of the phosphodiester, the later lanthanides are
preferred as catalysts.13
Earlier reports that lanthanum ions in water will preferentially
promote transfer of the acyl group of an acyl phosphate ester
to one hydroxyl of a diol14,15 implied that an aminoacyl
phosphate would be able to react selectively with the only
common diol in RNA, which is at the 3′-terminus. By exploiting
this unique feature and the reagent’s specificity, we have now
shown that the reaction can serve as a basis for 3′-terminal
specific aminoacylation of RNA.
Specific Reaction with tRNA. While the reaction patterns
and analysis with 19F NMR indicate that the reaction has
appropriate selectivity with respect to the 3′-terminus of RNA,
it is also important to conduct a reaction with a single
homogeneous tRNA. The NMR method is not sufficiently
sensitive to work with the quantities available for such a test.
Therefore, we sought an alternative, more sensitive method.
Since fluorescence detection is much more sensitive than NMR,
we prepared a fluorescent derivative of an amino acyl phosphate.
We reacted dansyl-glycyl ethyl phosphate (DNSGlyEP) with
tRNAPhe under the conditions used for the NMR studies but at
lower concentration levels and isolated the modified tRNA. We
also oxidized the tRNA with periodate and ran the aminoacy-
lation conditions as a control. We isolated both products and
tested for the specific effect of presence of the dansylated tRNA
with a dansyl-specific antibody, which serves as a fluorescence
enhancer (details in the Materials and Methods). With tRNAPhe
we observed the expected enhancement (Figure 10), whereas
with tRNAPhe that has first been oxidized we see no significant
enhancement. Since the dansylamide is stable under the reaction
conditions, the source of the fluorescence is necessarily the
aminoacylated tRNA. Using the same control as that in the NMR
experiments, specific oxidation of the 3′-terminus with periodate
converts the 1,2-diol to a dialdehyde that should not chelate
All reactions were performed in aqueous solution, and
products from hydrolysis of the aminoacylating reagent as well
as esters were formed. In the presence of a lanthanum ion, the
hydrolysis is very fast if no other ligand/nucleophile other than
water is present. However, when nucleosides or nucleotides were
introduced, mainly ester products were obtained. In the absence
of the metal ion, hydrolysis is slow and acylation does not occur.
We observe that aminoacylation of nucleosides with ami-
noacyl alkyl phosphates and a lanthanum ion occurs rapidly.
In all cases in our study, two monoacylated products were
obtained in the reaction with nucleosides. Control studies in
our laboratory with ribose, purines, and pyrimidines confirm
that there is neither formation of an ester derived from the 5′
hydroxyl nor an amide formed from reaction of an amino group
of the heterocyclic base. This is consistent with coordination
of lanthanum for reaction exclusively with the terminal 2′ and
(14) Evans, C. H. Biochemistry of Lanthanides; Plenum Press: New York, 1990.
(15) Chargaff, E., Davidson, J. N., Eds. The Nucleic Acids; Academic Press:
New York, 1955; Vol. 1.
(13) Komiyama, M.; Takeda, N.; Shigekawa, H. Chem. Commun. 1999, 1443-
1451.
9
J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007 15853

A R T I C L E S
Tzvetkova and Kluger
3′ hydroxyl groups. The ratio of 2′- to 3′-ester (1:2) is
comparable to the equilibrium ratios reported in the literature.16
It is expected that complexes of lanthanum with uncharged
sugars will not be as strong as those with phosphates.14 Thus,
the reaction occurs rapidly within the ionized bis-bidentate
complex because the overall association of lanthanum with the
diol is weak enough to permit the hydroxyl to alter its
coordination and attack the acyl group of the coordinated acyl
phosphate.
A large part of the effect of coordination to lanthanum comes
from the high effective concentration of ionized hydroxyl groups
that is achieved. The pKa of the OH groups on the ribose is
approximately 12.5.15 Water bound to lanthanum has a first pKa
of about 7.17 As well, coordination directs nucleophilic reaction
of the alkoxide to the carbonyl carbon of the acyl phosphate
(see Scheme 1). Consistent with the apparent importance of
ionization is our observation that at pH 8 and 6.5 the regiospeci-
ficity (terminal 2′ vs terminal 3′) and the extent of ester
formation from a nucleotide (5′CMP) are different. At higher
pH, a greater proportion of the lanthanum-coordinated water is
deprotonated so it can act as an intramolecular base catalyst.
This promotes esterification as well as the hydrolysis of the
aminoacyl phosphate. At lower pH, more ester products
(compared to hydrolysis products) are produced.
Based on these very promising results, we envision the work
extending in several directions that take advantage of the RNA
terminal ester specificity of these reactions. For example, in
order to be able to work at even lower concentrations, the
fluorescently labelled aminoacyl phosphates will assist in
detecting the incorporation of acyl phosphate derivatives of
amino acids at N-termini of proteins. We also expect that there
will be other applications that utilize our general method for
preparing 3′-terminal esters of RNA.
Acknowledgment. We thank Steven Rathgeber for his
assistance and discussions. The Natural Sciences and Engineer-
ing Research Council of Canada provided support through the
Discovery Grants program.
(16) Griffin, B.; Jarman, M.; Reese, C.; Sulston, J.; Trentham, D. Biochemistry
1966, 5, 3638-3649.
(17) Burgess, J. Metals in Solution; Ellis Horwood Limited: New York, 1978;
pp 259-289.
JA073976L
9
15854 J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007