Preparation of specifically deuterated RNA for NMR studies using a combination of chemical and enzymatic synthesis

Article abstract of DOI:10.1021/ja961274i

The syntheses of ATP, GTP, UTP, and CTP with deuterium labels on the 3', 4', and 5' carbons (2-5) is described. A combination of chemical and enzymatic synthesis is used where D,L-ribose-3,4,5,5'-d₄ (±1) is first produced from glycerol-d_8<

Full text of DOI:10.1021/ja961274i

J. Am. Chem. Soc. 1996, 118, 7929-7940
7929
Preparation of Specifically Deuterated RNA for NMR Studies
Using a Combination of Chemical and Enzymatic Synthesis†
Thomas J. Tolbert and James R. Williamson*
Contribution from the Department of Chemistry, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02139
ReceiVed April 18, 1996X
Abstract: The syntheses of ATP, GTP, UTP, and CTP with deuterium labels on the 3′, 4′, and 5′ carbons (2-5) is
described. A combination of chemical and enzymatic synthesis is used where D,L-ribose-3,4,5,5′-d4 ((1) is first
produced from glycerol-d8 by chemical methods, and then the four 3′,4′,5′,5′-labeled NTPs (2-5) are prepared from
(-1) using enzymes from the purine salvage and pyrimidine biosynthetic metabolic pathways. New procedures
were developed for the large scale preparation of GTP and CTP, and existing procedures were modified for the
preparation of ATP and UTP. A 30-nucleotide RNA derived from the HIV-2 TAR RNA was prepared with unlabeled
NTPs and deuterated NTPs (2-5) to illustrate the dramatic effects of deuteration on the NMR spectra of RNA. The
NOESY spectra of the deuterated RNA exhibits greatly reduced spectral crowding compared to that of the unlabeled
RNA, and assignment of NOEs to the H2′ protons is simplified due to the specific deuteration pattern. Also, the
nonselective T1 and T2 relaxation rates were measured for the deuterated RNA and found to be approximately twice
as long as the T1 and T2 relaxation rates of the unlabeled RNA. The spectral simplification and improved relaxation
properties of the deuterated RNA should prove useful in the study of large RNAs by NMR.
Introduction
significant problems encountered in application of NMR to the
study of larger RNA molecules are spectral crowding in the
ribose region and the large line widths. While the base, imino,
and H1′ protons in RNA have good chemical shift dispersion,
the remaining protons of the ribose ring, comprising over half
of the nonexchangeable protons in RNA, resonate between 4
and 5 ppm. Since there are many biologically interesting RNAs
that are significantly larger than 35 nucleotides, it is important
to develop techniques that would allow the study of these larger
RNAs using NMR. We have developed a combined chemical
and enzymatic synthesis of ribonucleotides deuterated at the 3′,
4′, and 5′ carbons of ribose. These nucleotides will help reduce
the spectral crowding in the ribose region and also reduce the
proton line widths by reducing dipolar relaxation. Specifically
deuterated nucleotides should prove to be a valuable tool in
extending the size of RNAs that can be studied by NMR.
NMR spectroscopy is a powerful tool for studying the
structure of RNA oligonucleotides.1 Short sequences forming
simple structures can be studied using H NMR methods, and
1
recently, heteronuclear methods have been developed that permit
the study of structures up to ≈35 nucleotides.2-16 Two
† Abbreviations: ATP, adenosine 5′-triphosphate; GTP, guanosine 5′-
triphosphate; UTP, uridine 5′-triphosphate; CTP, cytidine 5′-triphosphate;
NTP, nucleoside triphosphate; RNA, ribonucleic acid; nt, nucleotides; NMR,
nuclear magnetic resonance; ppm, parts per million; NOE, nuclear Over-
hauser effect; DIBAL-H, diisobutylaluminum hydride; PRPP, 5-phospho-
D-ribosyl R-1-pyrophosphate; PEP, phosphoenolpyruvate; OMP, orotidine
5′-monophosphate; IR, infrared; UV, ultraviolet; EtOAc, ethyl acetate; s,
singlet; d, doublet; t, triplet; q, quartet; m, multiplet; calcd, calculated;
HRMS, high-resolution mass spectrometry; HRFABMS, high-resolution fast
atom bombardment mass spectrometry; DMSO, dimethyl sulfoxide; Et2O,
diethyl ether; Et, ethyl; Bn, benzyl; THF, tetrahydrofuran; AcOH, acetic
acid; tBuOH, tert-butyl alcohol; NADH, â-nicotinamide adenine dinucle-
otide, reduced form; NAD+, â-nicotinamide adenine dinucleotide; Tris, tris-
(hydroxymethyl)aminomethane; PCR, polymerase chain reaction; IPTG,
isopropyl â-D-thiogalactopyranoside; LB medium, Luria-Bertani medium;
OD, optical density; OPRT, orotate phosphoribosyltransferase; APRT,
adenine phosphoribosyltransferase; XGPRT, xanthine-guanine phospho-
ribosyltransferase; TEABC, triethylammonium bicarbonate; 3PGA, 3-phos-
phoglycerate; R5P, ribose 5-phosphate; PPi, inorganic pyrophosphate; DTT,
dithiothreitol.
Deuteration has been used advantageously in the study of
proteins and nucleic acids with NMR spectroscopy by simpli-
fication of coupling patterns, reduction of line widths, removal
of unessential resonances, increasing NOE intensities, and as a
probe of dynamics.17-31 Although uniform 13C and 15N labeling
of proteins and RNA have been applied to macromolecular
X Abstract published in AdVance ACS Abstracts, August 15, 1996.
(1) Varani, G.; Tinoco, I. J. Q. ReV. Biophys. 1991, 24, 479-532.
(2) Puglisi, J. D.; Tan, R.; Calnan, B. J.; Frankel, A. D.; Williamson, J.
R. Science 1992, 257, 76-80.
(12) Jucker, F. M.; Pardi, A. Biochemistry 1995, 34, 14416-14427.
(13) Hall, K. B. Methods Enzymol. 1995, 261, 542-559.
(14) Pardi, A. Methods Enzymol. 1995, 261, 350-380.
(15) Ye, X.; Kumar, R. A.; Patel, D. J. Chem. Biol. 1995, 2, 827-840.
(16) Allain, F. H.; Gubser, C. C.; Howe, P. W.; Nagai, K.; Neuhaus, D.;
Varani, G. Nature 1996, 380, 646-650.
(17) LeMaster, D. M. Prog. NMR Spectrosc. 1994, 26, 371-419.
(18) Arrowsmith, C. H.; Pachter, R.; Altman, R. B.; Iyer, S. B.; Jardetzky,
O. Biochemistry 1990, 29, 6332-6341.
(19) Hsu, V. L.; Armitage, I. M. Biochemistry 1992, 31, 12778-12784.
(20) Wagner, G. J. Biomol. NMR 1993, 3, 375-385.
(21) Farmer, B. T.; Venters, R. A. J. Am. Chem. Soc. 1995, 117, 4187-
4188.
(22) Yamazaki, T.; Lee, W.; Arrowsmith, C. H.; Muhandiram, D. R.;
Kay, L. E. J. Am. Chem. Soc. 1994, 116, 11655-11666.
(23) Nietlispach, D.; Clowes, R. T.; Broadhurst, R. W.; Ito, Y.; Keeler,
J.; Kelly, M.; Ashurst, J.; Oschkinat, H.; Domaille, P. J.; Laue, E. D. J.
Am. Chem. Soc. 1996, 118, 407-415.
(24) Grzesiek, S.; Anglister, J.; Ren, H.; Bax, A. J. Am. Chem. Soc. 1993,
115, 4369-4370.
(3) Peterson, R. D.; Bartel, D. P.; Szostak, J. W.; Horvath, S. J.; Feigon,
J. Biochemistry 1994, 33, 5357-5366.
(4) Battiste, J. L.; Tan, R.; Frankel, A. D.; Williamson, J. R. Biochemistry
1994, 33, 2741-2747.
(5) Battiste, J. L.; Tan, R.; Frankel, A. D.; Williamson, J. R. J. Biomol.
NMR 1995, 6, 375-389.
(6) Marino, J. P.; Gregorian, R. S.; Csankovszki, G.; Crothers, D. M.
Science 1995, 268, 1448-1454.
(7) Puglisi, J. D.; Chen, L.; Blanchard, S.; Frankel, A. D. Science 1995,
270, 1200-1203.
(8) Allain, F.; Varani, G. J. Mol. Biol. 1995, 250, 333-353.
(9) Aboul-ela, F.; Karn, J.; Varani, G. J. Mol. Biol. 1995, 253, 313-
332.
(10) Greenbaum, N. L.; Radhakrishnan, I.; Hirsh, D.; Patel, D. J. J. Mol.
Biol. 1995, 252, 314-327.
(11) Nowakowski, J.; Tinoco, I. J. Biochemistry 1996, 35, 2577-2585.
S0002-7863(96)01274-7 CCC: $12.00 © 1996 American Chemical Society

7930 J. Am. Chem. Soc., Vol. 118, No. 34, 1996
Tolbert and Williamson
structure determination with great advantage, introduction of
the isotopic label results in broader line widths for protons due
to the strong dipolar interaction with directly bonded 13C and
15N nuclei.17,21-24 This effect is particularly problematic for
larger molecules. Deuterium labeling offers the alternative
advantage of spectral simplification without causing broader line
widths. One deuteration pattern that would benefit NMR of
RNA is the selective deuteration of the ribose 3′, 4′, and 5′
carbons. Much of the structural information available from
NMR spectra of RNA can be derived from the base protons,
the ribose H1′ and H2′ protons, and the exchangeable protons.
The glycosidic torsion can be estimated from the magnitude of
base to H1′ NOEs, the sugar pucker can be estimated from the
scalar coupling between the H1′ and H2′, and base pairing
information is obtained from the exchangeable protons. Deu-
teration of the 3′, 4′, and 5′ carbons would still permit this
information to be obtained; however, the spectral assignment
would be greatly simplified.
Figure 1. Strategy for synthesis of deuterated nucleotides.
by chemical coupling to form a nucleoside.39 This method
allows a large flexibility in isotopic labeling since the ribose or
the bases can be prepared and labeled separately and then
combined. A variety of isotopic labels have been incorporated
into nucleosides using this general scheme including ribose
deuterated by Raney nickel-2H2O exchange,40 H1-deuterated
ribose,41 H2-deuterated ribose,42 and ribose 13C labeled at the
C1 or C2.43,44 This method also presents problems, however,
since to chemically couple ribose to a base, D-ribose must first
be protected as acetyl 2,3,5-tri-O-benzoyl-â-D-ribofuranoside or
some equivalent, which is a multistep synthetic process.
Furthermore, once nucleosides have been formed chemically,
they must still be converted into NTPs, which requires additional
chemical and enzymatic steps.45 Another method that has been
used, and the one that we have adopted, is isotopic labeling of
the ribose or base separately, followed by enzymatic coupling
to form a nucleotide. Enzymatic coupling allows the direct
conversion of ribose into nucleotides or nucleosides without the
use of protecting groups, which greatly simplifies the synthetic
effort required to produce isotopically labeled NTPs from ribose.
This strategy has been used to produce [1′-14C]AMP and [1′-
3H]AMP starting with isotopically labeled ribose and adenine
using adenine phosphoribosyltransferase.46 Purine nucleoside
phosphorylases have also been used to couple [C8-13C]purines
to ribose-1-phosphate to form [C8-13C]purine nucleosides.45
To synthesize 3′,4′,5′,5′-labeled NTPs (d4-NTPs) we have
developed a synthetic strategy that uses chemical and enzymatic
methods, outlined in Figure 1. Chemical synthesis is used to
convert glycerol-d8 into D-ribose-3,4,5,5-d4 (D-d4-ribose, -1),
and then four enzymatic reactions are used to convert that ribose
into the four d4-NTPs (2-5) with a minimum amount of effort.
The strategy utilizes chemical synthesis to place deuterium at
specific positions, and enzymatic synthesis to perform a series
of complex transformations in a single coupled reaction. These
deuterated ribonucleotides should prove useful for the study of
large RNA molecules, and the preparative methods we have
developed will prove useful for a wide variety of labeling
applications.
Since large RNAs are most readily prepared by transcription
with T7 RNA polymerase, our goal was to prepare deuterated
nucleoside triphosphates (NTPs).32 One method for producing
isotopically labeled NTPs is to harvest total cellular RNA from
bacteria that have been grown on isotopically labeled media,
enzymatically digest the RNA to nucleoside monophosphates
(NMPs), and finally enzymatically convert the NMPs to NTPs.
Uniform 13C, 15N, and 13C/15N labeling of NTPs has been
accomplished by this method by growing Escherichia coli or
Methylophilus methylotrophus on 13C and/or 15N labeled
media.33-36 Unfortunately it is difficult to incorporate specific
isotopic labels with this method due to isotopic scrambling
during biosynthesis. A second method that can be used for
producing deuterated NTPs involves selective oxidation of one
of the hydroxyl groups of ribose, followed by reduction with
deuteride. This method has been used to deuterate both H5′
protons of thymidine, as well as the H3′ of adenosine.37,38
Schemes using selective oxidation/reduction of nucleosides to
incorporate deuterium labels work well for incorporation of
deuterium on a single carbon, but become cumbersome when
applied to deuteration of multiple carbons, and also do not
readily offer the possibility for 13C and 15N incorporation. A
third method for incorporating isotopic labels into nucleotides
is to isotopically label the ribose or base separately, followed
(25) Brush, C. K.; Stone, M. P.; Harris, T. M. Biochemistry 1988, 27,
115-122.
(26) Puglisi, J. D.; Wyatt, J. R.; Tinoco, I. J. Mol. Biol. 1990, 214, 437-
453.
(27) Yamakage, S.-I.; Maltseva, T. V.; Nilson, F. P.; Fo¨ldesi, A.;
Chattopadhyaya, J. Nucleic Acids Res. 1993, 21, 5005-5011.
(28) Agback, P.; Maltseva, T. V.; Yamakage, S.-I.; Nilson, F. P. R.;
Fo¨ldesi, A.; Chattopadhyaya, J. Nucleic Acids Res. 1994, 22, 1404-1412.
(29) Muhandiram, D. R.; Yamazaki, T.; Sykes, B. D.; Kay, L. E. J. Am.
Chem. Soc. 1995, 117, 11536-11544.
(30) Wang, A. C.; Kennedy, M. A.; Reid, B. R.; Drobny, G. P. J. Mag.
Reson. Ser. B 1994, 105, 1-10.
(31) Fo¨ldesi, A.; Yamakage, S. I.; Nilsson, F. P.; Maltseva, T. V.;
Chattopadhyaya, J. Nucleic Acids Res. 1996, 24, 1187-1194.
(32) Wyatt, J. R.; Chastain, M.; Puglisi, J. D. BioTechniques 1991, 11,
764-769.
(33) Batey, R. T.; Inada, M.; Kujawinski, E.; Puglisi, J. D.; Williamson,
J. R. Nucleic Acids Res. 1992, 20, 4515-4523.
(34) Nikonowicz, E. P.; Sirr, A.; Legault, P.; Jucker, F. M.; Baer, L.
M.; Pardi, A. Nucleic Acids Res. 1992, 20, 4507-4513.
(35) Hoffman, D. W.; Holland, J. A. Nucleic Acids Res. 1995, 23, 3361-
3362.
(39) Vorbru¨ggen, H.; Krolikiewicz, K.; Bennua, B. Chem. Ber. 1981,
114, 1234-1255.
(40) Fo¨ldesi, A.; Nilson, F. P. R.; Glemarec, C.; Gioeli, C.; Chatto-
padhyaya, J. Tetrahedron 1992, 48, 9033-9072.
(41) Toyama, A.; Takino, Y.; Takeuchi, H.; Harada, I. J. Am. Chem.
Soc. 1993, 115, 11092-11098.
(42) Cook, G. P.; Greenberg, M. M. J. Org. Chem. 1994, 59, 4704-
4706.
(43) Lancelot, G.; Chanteloup, L.; Beau, J.; Thuong, N. T. J. Am. Chem.
Soc. 1993, 115, 1599-1600.
(36) Batey, R. T.; Battiste, J. L.; Williamson, J. R. Methods Enzymol.
1995, 261, 300-322.
(37) Frank, B. L.; Worth, L. J.; Christner, D. F.; Kozarich, J. W.; Stubbe,
J.; Kappen, L. S.; Goldberg, I. H. J. Am. Chem. Soc. 1991, 113, 2271-
2275.
(38) Robins, M. J.; Samano, V.; Johnson, M. D. J. Org. Chem. 1990,
55, 410-412.
(44) Kline, P. C.; Serianni, A. S. J. Am. Chem. Soc. 1990, 112, 7373-
7381.
(45) SantaLucia, J.; Shen, L. X.; Cai, Z.; Lewis, H.; Tinoco, I. Nucleic
Acids Res. 1995, 23, 4913-4921.
(46) Parkin, D. W.; Leung, H. B.; Schramm, V. L. J. Biol. Chem. 1984,
259, 9411-9417.

Preparation of Specifically Deuterated RNA for NMR Studies
J. Am. Chem. Soc., Vol. 118, No. 34, 1996 7931
Results and Discussion
Chemical Synthesis. For the chemical synthesis, our strategy
was to build upon well-established carbohydrate syntheses to
produce D-d4-ribose (-1) from commercially available, isotopi-
cally labeled starting materials. The ribose synthesis was based
on acyclic carbohydrate syntheses, where 2,3-O-isopropylidene-
D-glyceraldehyde is extended by a Wittig reaction, and the
resulting double bond is then dihydroxylated with osmium
tetroxide, or by Sharpless epoxidation with subsequent ring
opening.47-49 The desired labeling pattern can be obtained if
the protected D-glyceraldehyde is prepared perdeuterated, since
the 3, 4, and 5 carbons of ribose will arise from D-glyceralde-
hyde. Unfortunately, common methods of synthesizing 2,3-O-
isopropylidene-D-glyceraldehyde, such as the periodate oxidation
of 1,2:5,6-di-O-isopropylidene-D-mannitol,50 do not conveniently
allow for deuteration, so a method for obtaining perdeuterated
2,3-O-isopropylidene-D-glyceraldehyde was developed. Initial
attempts involved preparation of perdeuterated D-glyceraldehyde
from perdeuterated allyl alcohols using asymmetric dihydroxy-
lation.51 This method had the advantage of producing a single
enantiomer, but the multistep synthesis of the perdeuterated allyl
alcohol had a low overall yield. A higher yield of the
cyclohexyl ketal of perdeuterated D-glyceraldehyde (7) was
obtained from the racemic mixture of perdeuterated glyceral-
dehydes produced from a simple two-step synthesis starting with
commercially available glycerol-d8. Since the ultimate product
of the chemical synthesis, D,L-ribose, can be subsequently
resolved by the enzymes used to convert D-ribose into NTPs,
the disadvantage of a racemic synthesis was offset by the higher
total yield of D-glyceraldehyde and ease of preparation.
The synthesis of D,L-d4-ribose ((1) from glycerol-d8 is shown
in Figure 2. Protection of glycerol-d8 is followed by Swern
oxidation of ketal (6), resulting in D,L-glyceraldehyde-1,2,3,3-
d4 ketal (7). The triethylamine from the Swern oxidation was
carefully neutralized to prevent deuterium exchange by eno-
lization that can occur under basic protic conditions.48 The
crude aldehyde was unstable to silica gel purification, and was
directly submitted to a Wittig reaction with Ph3PdCHCO2Et
in methanol at 0 °C, which preferentially forms the Z olefin,49
to produce (8) in a 76% yield. Reduction of the ester with
DIBAL-H and protection of the resulting alcohol (9) afforded
the benzyl ether (10). Facial diastereoselectivity in the osmium
tetroxide cis dihydroxylation of the Z olefin47 resulted in the
desired protected ribitol ((11) and protected lyxitol diastere-
omers in a 7:3 ratio. After chromatographic separation of the
diastereomers, protection of the cis-diol, and removal of the
benzyl ether, Swern oxidation of (13 gave the protected ribose
((14). Again, the triethylamine was carefully neutralized by
an aqueous workup, to prevent epimerization under basic
conditions.48 The aldehyde ((14) was unstable to silica gel
purification, and hydrolysis of the crude aldehyde ((14)
afforded crude D,L-d4-ribose ((1). The yield of ribose was
quantitated by enzymatic assay of the crude racemic ribose
mixture. The overall yield of the 10 step synthesis was 12%
D-d4-ribose (-1), (24% (1), as calculated from the starting
material glycerol-d8. The crude D,L-d4-ribose ((1) prepared in
this manner was suitable for direct use in the subsequent
enzymatic reactions.
Figure 2. Synthesis of d4-ribose. Reagents: (a) cyclohexanone, (MeO)3-
CH, H+ (b) oxalyl chloride, DMSO, Et3N (c) Ph3PdCHCO2Et, MeOH
(d) DIBAL-H, CH2Cl2 (e) BnBr, NaH, (nBu)4NI, THF (f) OsO4,
N-methylmorpholine N-oxide, acetone:H2O (8:1) (g) 2-methoxypropene,
H+ (h) Pd on C, H2 (i) oxalyl chloride, DMSO, Et3N (j) H+, THF,
H2O.
Enzymatic Synthesis. Our strategy for the enzymatic
synthesis was to utilize phosphoribosyltransferases for coupling
of 5-phospho-D-ribosyl R-1-pyrophosphate (PRPP) with nitrog-
enous bases to form nucleoside monophosphates. Reactions
catalyzed by phosphoribosyltransferases can be easily linked
to enzymatic PRPP synthesis and NTP formation, allowing
conversion of ribose into a nucleoside triphosphate in a single
coupled enzymatic reaction, as has been shown previously by
Schramm.46 Preparative scale enzymatic reactions for PRPP
synthesis from ribose52 and for NTP formation from NMPs53
have been described previously. Enzymatic procedures for
converting ribose into AMP and UMP have been reported,
offering enzymatic routes to ATP46 and UTP.52 We have
developed enzymatic reactions that produce GTP from ribose
and CTP from UTP, to complete the set of four NTPs required
for RNA synthesis. By combining these methods we were able
to form d4-NTPs (2-5) from D-d4-ribose (-1) in a minimum
number of steps in high yield.
Some of the enzymes necessary for these conversions were
commercially unavailable or only available in costly or low
specific activity forms. These enzymes were purified from
overexpressing strains using a general purification scheme which
readily afforded partially purified enzymes. The enzyme
purification scheme used requires at most one chromatographic
step and results in preparations that are free from undesirable
contaminating activities, such as nonspecific ATPases. Each
enzyme preparation from 1 L of bacterial culture yielded enough
enzyme to catalyze several, from 7 to 180 depending on the
enzyme, reactions of the type described here.
PRPP Synthesis. The PRPP used in NTP formation was
generated in situ from D-ribose in a manner similar to that used
by Gross et al.52 D-d4-ribose (-1) was first phosphorylated by
ribokinase to produce ribose-3′,4′,5′,5′-d4 5-phosphate (d4-R5P),
which was pyrophosphorylated by PRPP synthetase to form
PRPP-3′,4′,5′,5′-d4 (d4-PRPP, 15), as shown in Figure 3a. In
(47) Cha, J. K.; Christ, W. J.; Kishi, Y. Tetrahedron 1984, 40, 2247-
2255.
(48) Ko, Y. S.; Lee, A. W. M.; Masamune, S.; Reed, L. A. I.; Sharpless,
K. B.; Walker, F. J. Tetrahedron 1990, 46, 245-264.
(49) Katsuki, T.; Lee, A. W. M.; Ma, P.; Martin, V. S.; Masamune, S.;
Sharpless, K. B.; Tuddenham, D.; Walker, F. J. J. Org. Chem. 1982, 47,
1373-1378.
(50) Jackson, D. Y. Synth. Commun. 1988, 18, 337-341.
(51) Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.;
Hartung, J.; Jeong, K.; Kwong, H.; Morikawa, K.; Wang, Z.; Xu, D.; Zhang,
X. J. Org. Chem. 1992, 57, 2768-2771.
(52) Gross, A.; Abril, O.; Lewis, J. M.; Geresh, S.; Whitesides, G. M. J.
Am. Chem. Soc. 1983, 105, 7428-7435.
(53) Simon, E. S.; Grabowski, S.; Whitesides, G. M. J. Org. Chem. 1989,
55, 1834-1841.

7932 J. Am. Chem. Soc., Vol. 118, No. 34, 1996
Tolbert and Williamson
(a)
(b)
(c)
Figure 3. (a) Enzymatic synthesis of labeled PRPP from labeled D-ribose. Also shown is ATP regeneration by PEP/PK/MK system driven by
3PGA. Enzymes are (a) ribokinase, (b) PRPP synthetase, (c) myokinase, (d) pyruvate kinase, (e) enolase, and (f) 3-phosphoglycerate mutase. PRPP
synthesis was coupled directly to ATP, GTP and UTP formation shown in parts b and c. (b) Enzymatic synthesis of purines ATP and GTP using
purine salvage. Enzymes are (c) myokinase, (d) pyruvate, kinase, (g) adenine phosphoribosyltransferase, (h) xanthine-guanine phosphoribosyl-
transferase, and (i) guanylate kinase. (c) Enzymatic synthesis of UTP and CTP using pyrimidine biosynthesis. Enzymes are (d) pyruvate kinase, (j)
orotate phosphoribosyltransferase, (k) OMP decarboxylase, (l) nucleoside monophosphate kinase, and (m) CTP synthetase. In the preparative reactions,
CTP synthesis is performed in a reaction separate from UTP formation.
all of the enzymatic reactions, the ATP consumed was regener-
ated by a PEP/pyruvate kinase/myokinase system, where the
phosphoenolpyruvate (PEP) used to drive the reaction was
generated in situ from an excess of 3-phosphoglycerate,53,54 also
shown in Figure 3a. Reactions containing PRPP synthetase were
carried out in potassium phosphate buffer to prevent the
inactivation of PRPP synthetase that occurs at low phosphate
concentration.55 One difficulty encountered in PRPP synthesis
was the inhibition of PRPP synthetase by ADP in the presence
of ribose 5-phosphate, resulting in extremely slow nucleotide
formation in our reactions due to a buildup of ADP and ribose
5-phosphate.56 The buildup of ADP was prevented by reducing
the amount of myokinase and increasing the amount of pyruvate
(54) Hirschbein, B. L.; Mazenod, F. P.; Whiteside, G. M. J. Org. Chem.
1982, 47, 3765-3766.
(55) Switzer, R. L.; Gibson, K. J. Methods in Enzymology 1978, LI, 3-11.

Preparation of Specifically Deuterated RNA for NMR Studies
J. Am. Chem. Soc., Vol. 118, No. 34, 1996 7933
(a)
(b)
Figure 4. HPLC chromatograms of (a) GTP- and (b) CTP-forming reactions.
kinase added to the reactions, to insure that the rate of ADP
formation from AMP was much less than the rate of ATP
formation from ADP and PEP.
GTP synthesis. Only a small guanine peak can be detected in
the HPLC traces, but 2 equiv (per ribose) of solid guanine were
present in the reaction. During the course of the reaction, GMP
formation can be observed at 9 and 22 h, and the conversion of
GMP to GTP is seen in the final HPLC trace. d4-GTP (3) was
produced in 75% yield from D-d4-ribose (-1) by this method.
UTP Synthesis. UTP-3′,4′,5′,5′-d4 (d4-UTP, 4) was synthe-
sized from D-d4-ribose (-1) using two pyrimidine biosynthetic
enzymes, orotate phosphoribosyltransferase and orotidine 5′-
monophosphate decarboxylase (OMP decarboxylase), which
catalyze the formation of UMP from PRPP and orotate. Gross
et al.52 have used these two enzymes to form UMP from PRPP
and orotate, and we have extended this method to produce UTP
from D-ribose and orotate. As in the d4-ATP (2) and d4-GTP
(3) forming reactions, d4-PRPP (15) was first generated in situ
(Figure 3a), and then orotate phosphoribosyltransferase was used
to catalyze the condensation of orotate and PRPP to form d4-
OMP, which was then decarboxylated by OMP decarboxylase
forming d4-UMP, shown in Figure 3c. The sequential action
of nucleoside monophosphate kinase and pyruvate kinase then
produced d4-UTP (4) in 90% isolated yield from (-1) (Figure
3c).
ATP Synthesis. ATP-3′,4′,5′,5′-d4 (d4-ATP, 2) was formed
in a coupled enzymatic reaction from D-d4-ribose (-1) in a
manner similar to that described by Schramm.46 d4-PRPP (15)
was first generated in situ as described above (Figure 3a), and
the purine salvage enzyme adenine phosphoribosyltransferase
(APRT) was used to catalyze the condensation of PRPP and
adenine to form d4-AMP, shown in Figure 3b. d4-ATP (2) was
then formed in the same reaction by the sequential action of
myokinase and pyruvate kinase on d4-AMP (Figure 3b). A
catalytic amount of unlabeled ATP was added to the d4-ATP-
forming reaction to initiate the reaction. In all of the other
reactions, d4-ATP was used as a cofactor to prevent dilution of
the isotopic labels. The initial rate of ATP formation was very
slow, but the reaction rate increases quickly as the concentration
of ATP increases. d4-ATP (2) was produced from (-1) in 78%
isolated yield by this method.
GTP Synthesis. The synthesis of GTP-3′,4′,5′,5′-d4 (d4-GTP,
3) was accomplished by utilizing the E. coli purine salvage
enzyme xanthine-guanine phosphoribosyltransferase (XGPRT).
First, d4-PRPP (15) was generated in situ (Figure 3a) and then
xanthine-guanine phosphoribosyltransferase was used to couple
d4-PRPP (15) and guanine to form d4-GMP, as shown in Figure
3b. Once d4-GMP had been formed, the sequential action of
guanylate kinase and pyruvate kinase in the same reaction
produced d4-GTP (3) (Figure 3b). Although guanine has a very
low solubility in water, GMP formation occurs at an appreciable
rate when solid guanine is added to the reaction in a slurry.
Figure 4a shows HPLC traces during the course of a preparative
CTP Synthesis. CTP-3′,4′,5′,5′-d4 (d4-CTP, 5) was synthe-
sized from d4-UTP (4) by ATP dependent amination of UTP
with glutamine or ammonia using the E. coli pyrimidine
biosynthetic enzyme CTP synthetase, as shown in Figure 3c.
CTP synthetase proved to be a difficult enzyme to work with
since the activity of the enzyme is regulated by the concentration
of substrates and products in solution.57,58 The substrates ATP
and UTP are activators for CTP synthetase, and the product
(57) Long, C. W.; Pardee, A. B. J. Biol. Chem. 1967, 242, 4715-4721.
(58) Anderson, P. M. Biochemistry 1983, 22, 3285-3292.
(56) Switzer, R. L.; Sogin, D. C. J. Biol. Chem. 1973, 248, 1063-1073.

7934 J. Am. Chem. Soc., Vol. 118, No. 34, 1996
Tolbert and Williamson
CTP can act as an inhibitor. Using conditions similar to those
used for synthesis of the other nucleotides, no CTP production
was observed with CTP synthetase. After trying several
different reaction conditions, it was found that by diluting UTP
concentration from approximately 5.5 to 1 mM and increasing
the ATP concentration from 0.17 to 0.22 mM, it was possible
to convert UTP into CTP. Under these conditions CTP
concentration is low enough to prevent feedback inhibition, and
the ATP concentration is high enough to keep the enzyme active
at low UTP concentrations. Because of the problems encoun-
tered with CTP synthetase, CTP was produced from UTP rather
than the large coupled enzymatic reactions where ATP, GTP,
and UTP were produced directly from ribose. In addition, a
small amount of GTP was added as a positive affector of
glutamine dependent amine transfer.57 The conversion of UTP
into CTP catalyzed by CTP synthetase was monitored by HPLC
as shown in Figure 4b. The amount of ATP relative to the
other nucleotides is much greater in the CTP-forming reaction
than in the GTP-forming reaction (Figure 4a), and additional
ATP was added after 3.5 h in order to increase the rate of the
reaction. During the course of the conversion of UTP into CTP,
a number of NDP species become visible, but these are
eventually converted to NTPs in the final HPLC trace. After
careful optimizing, this deceptively simple single enzyme
reaction coupled with ATP regeneration produced d4-CTP (5)
in 87% yield from d4-UTP (4).
a
b
NMR of d4-RNA. To illustrate the effects of the 3′,4′,5′,5′-
deuteration pattern, we prepared a 30-nucleotide RNA derived
from the HIV-2 TAR bulged loop by in vitro transcription with
both unlabeled and deuterated NTPs. The 1H NMR spectra of
both in D2O are shown in Figure 5. The spectral crowding in
the ribose region of the deuterated RNA is greatly reduced, and
individual H2′ resonances are clearly observed between 4 and
5 ppm. The advantages of this deuteration pattern are dramati-
cally illustrated in a comparison of the base to ribose region of
a NOESY spectrum of TAR RNA acquired on unlabeled RNA
and d4-RNA, shown in Figure 6, parts a and b, respectively.
The base to H2′ NOEs are unambiguously identified in Figure
6b, and the spectral overlap in this region is greatly reduced. A
similarly dramatic effect of the ribonucleotides-d4 is observed
in the ribose region of the NOESY spectrum, shown in Figure
7. The H1′ to H2′ cross peaks can be unambiguously identified
in Figure 7b. The diagonal for the region between 4 and 5
ppm is completely devoid of cross peaks, except for one weak
peak that results from an unusual H2′-H2′ NOE from the loop
region of TAR. This peak was not previously identified in 3D-
or 4D-13C resolved NOESY spectra using uniformly 13C-labeled
TAR RNA, even though all of the ribose protons were
assigned.59
Figure 5. 500 MHz 1H NMR spectra of HIV-2 TAR RNA (a)
unlabeled (inset is the secondary structure of HIV-2 TAR RNA) (b)
3′,4′,5′,5′-2H4-labeled.
of relaxation efficiency could significantly improve the spectra
of RNAs that are much larger.
Conclusion
This paper describes a synthesis of d4-NTPs (2-5) that
utilizes both chemical and enzymatic techniques. A chemical
synthesis of D,L-d4-ribose ((1) that uses perdeuterated glycerol
as a starting material and produces (1 in 24% overall yield is
reported. Three multienzyme reactions are described that mimic
purine salvage and pyrimidine biosynthetic pathways to convert
D-d4-ribose (-1) into d4-ATP (2), d4-GTP (3), and d4-UTP (4)
in 78%, 75%, and 90% yields, respectively. Another enzymatic
reaction is reported where d4-UTP (4) is converted into d4-CTP
(5) in 87% yield. These enzymatic reactions represent an
efficient method for incorporating isotopically labeled ribose
or bases into the four NTPs required for RNA transcription.
The d4-ribonucleotides described here should prove extremely
valuable in the structural characterization of large RNAs using
NMR spectroscopy. The method preparation of these nucle-
otides is sufficiently flexible that a number of different labeling
patterns with 15N/13C/2H can be readily prepared. In this way,
ribonucleotides can be tailor-made to extract particular informa-
tion of interest from simplified and heteronuclear spectra.
Specific deuteration, in particular, offers the advantages of
reduced line width and spectral simplification, without the
sacrifice of sensitivity inherent in random fractional deuteration.
Although the cross peaks in the NOESY spectra of d4-RNA
in Figures 6b and 7b appear marginally narrower than the
unlabeled RNA in Figures 6a and 7a, relaxation rates were
determined to provide a more quantitative measure of the effects
of deuteration on dipolar relaxation. Nonselective T1 and T2
measurements were performed on unlabeled TAR and d4-TAR
RNAs, and the relaxation rates were determined for a number
of resolved peaks, summarized in Table 1. The effective T1
and T2 relaxation times for the d4-RNA were increased by
approximately a factor of 2 compared to unlabeled RNA. These
results are comparable to the effects of deuteration observed in
proteins60 and DNA.28 Although for the 30-nucleotide TAR
RNA studied here the deuteration does not qualitatively improve
the NOESY spectrum by a relaxation effect, a 50% reduction
(59) Brodsky, A. S.; Williamson, J. R., personal communication 1996.
(60) Markus, M. A.; Dayie, K. T.; Matsudaira, P.; Wagner, G. J. Magn.
Reson. Ser. B 1994, 105, 192-195.

Preparation of Specifically Deuterated RNA for NMR Studies
J. Am. Chem. Soc., Vol. 118, No. 34, 1996 7935
Figure 6. Expansion of the base to ribose region of a NOESY spectrum of (a) TAR RNA and (b) d4-TAR RNA.
Indicator Strips pH 0-14 made by EM Science. NMR spectra were
Experimental Section
recorded on a Varian XL-300 or a Varian VXR-500 spectrometer,
infrared spectra were recorded on a Perkin Elmer 1600 series FT-IR,
and ultraviolet spectra were recorded on a Hitachi U-2000 UV/Vis
spectrophotometer. Mass spectra were obtained with a Hewlett Packard
5890/5971 GC/mass spectrometer for low resolution and a Finnigan
MAT 8200 for high resolution.
Materials and Methods/General Procedures. Reactions were
conducted under an argon atmosphere unless otherwise noted, and
solvents were dried and distilled.61 Chemicals were purchased from
Aldrich and Sigma. Glycerol-d8, D2O, and CDCl3 were purchased from
Cambridge Isotopes Laboratories. Phosphoglycerate mutase from rabbit
muscle and nucleoside monophosphate kinase from beef liver were
purchased from Boehringer Mannheim. Enolase from baker’s yeast,
myokinase from chicken muscle, pyruvate kinase from rabbit muscle,
and guanylate kinase from porcine brain were purchased from Sigma.
One unit of enzyme corresponds to 1 µmol/min of activity under assay
conditions. The pH of aqueous solutions was determined by ColorpHast
D,L-2,3-O-Cyclohexylideneglycerol-1,1′,2,3,3′-d5 (6). Cyclohex-
anone (3.49 g, 35.6 mmol) and trimethylorthoformate (3.77 g, 35.6
mmol) were dissolved in CH2Cl2 (150 mL). Glycerol-d8 (98%) (3.56
g, 35.6 mmol) and a catalytic amount of p-toluenesulfonic acid (10
mg, 50 µmol) were added, and the heterogeneous mixture was stirred
for 9 h at room temperature, after which the solution was homogeneous
and the reaction complete. The reaction mixture was concentrated and
the product purified by flash chromatography (5-40% EtOAc in
(61) Perrin, D. D.; Armarego, W. L. Purification of Laboratory
Chemicals, 3rd ed.; Pergamon Press Inc.: New York, 1988.

7936 J. Am. Chem. Soc., Vol. 118, No. 34, 1996
Tolbert and Williamson
Figure 7. Expansion of the ribose to ribose region of a NOESY spectrum of (a) TAR RNA and (b) d4-TAR RNA.
hexane) to yield 6.07 g (34.1 mmol, 95%) of 6 as a colorless oil. 1H
NMR (300 MHz, CDCl3) δ 2.86 (s br, 1H), 1.85-1.28 (m, 10 H); 13
g, 17.2 mmol) was added to CH2Cl2 (200 mL) in a flask cooled to
-78 °C, and DMSO (2.68 g, 34.4 mmol) was added slowly via syringe.
The resulting mixture was stirred for 5 min before 6 (2.78 g, 15.6 mmol)
was added. The mixture was stirred for 30 min at -78 °C, and then
triethylamine (6.32 g, 62.5 mmol) was added slowly. After stirring
for 15 min at -78 °C, the dry ice/acetone bath was removed. After
30 min, the reaction was quenched with 75 mL of 0.55M HCl. The
organic layer was separated and washed with saturated NaHCO3 (75
mL) and saturated NaCl (75 mL). The organic layer was dried over
MgSO4 and then concentrated. The crude aldehyde 3 was carried on
directly to the next reaction without purification.
C
NMR (75 MHz, CDCl3) δ 109.7, 75.0 (t, J ) 23 Hz), 64.6 (quintet, J
) 23 Hz), 62.2 (quintet, J ) 21 Hz), 36.2, 34.6, 24.9, 23.8, 23.6; IR
(neat) 3442, 2944, 2860, 2215, 2106, 1449, 1367, 1333, 1288, 1257,
1231, 1169, 1105, 1063, 1011, 974, 941, 907; HRMS m/z 177.1409
2
(177.1408 calcd for C9H11O3 H5, M+).
D,L-2,3-O-Cyclohexylideneglyceraldehyde-1,2,3,3′-d4 (7). Alcohol
6 was dissolved in CH3OD and evaporated two times to exchange the
1
hydroxyl proton, remove protons, and prevent H exchange that can
occur with the aldehyde under basic conditions. Oxalyl chloride (2.18

Preparation of Specifically Deuterated RNA for NMR Studies
J. Am. Chem. Soc., Vol. 118, No. 34, 1996 7937
Table 1
a
c
T1
T2
TAR
d4-TAR
ratiob
TAR
d4-TAR
ratiod
H2′/(H3′,H4′,H5′a,H5′b)e
H1′/H5
H8/H2/H6
4.3 ( 0.7
5.6 ( 0.7
5.4 ( 1.2
9.3 ( 1.0
10.3 ( 1.3
10.2 ( 1.6
2.2 ( 0.2
1.8 ( 0.1
1.9 ( 0.2
0.04 ( 0.01
0.10 ( 0.02
0.09 ( 0.03
0.12 ( 0.02
0.17 ( 0.03
0.17 ( 0.06
2.6 ( 0.3
1.8 ( 0.3
2.0 ( 0.3
a Average T1 relaxation time b Ratio of d4-TAR and TAR T1 relaxation times. c Average T2 relaxation time. d Ratio of d4-TAR and TAR T2
relaxation times. e The global average of relaxation times were calculated for these protons in the unlabeled sample due to spectral overlap in this
region.
Ethyl (Z)-D,L-4,5-O-Cyclohexylidene-2-pentenoate-3,4,5,5-d4 (8).
The crude aldehyde 7 was dissolved in 100 mL of methanol and cooled
to 4 °C. Ph3PdCHCO2Et, (carbethoxymethylene)triphenylphosphorane,
was added, and the reaction was stirred for 24 h at 4 °C. After
concentration in vacuo the residue was purified by flash chromatography
(5% Et2O in petroleum ether) to yield 2.91 g (11.9 mmol, 76%) of a
light yellow oil. 1H NMR (300 MHz, CDCl3) δ 5.83 (s, 1 H), 4.17 (q,
concentrated. Separation of diastereomers was accomplished by careful
flash chromatography (15% EtOAc in hexane). The slower running
diastereomer corresponds to the ribitol configuration ((11), 1.21 g (3.7
mmol, 63%) of the yellow oil (11 was separated from the faster eluting
diastereomer. 1H NMR (300 MHz, CDCl3) δ 7.41-7.22 (m, 5 H),
4.56 (d, 1 H, J ) 11.9 Hz), 4.51 (d, 1 H, J ) 11.9 Hz), 3.85 (t, 1 H,
J ) 4.9 Hz), 3.68 (d, 2 H, J ) 4.9 Hz), 3.31 (s br, 1 H), 3.10 (s br, 1
H), 1.70-1.30 (m, 10 H); 13C NMR (75 MHz, CDCl3) δ 137.5, 128.3
(2 C), 127.8, 127.7 (2 C), 109.7, 75.2 (t, J ) 23 Hz), 73.5, 72.3 (t, J
) 22 Hz), 71.5, 70.9, 65.2 (quintet, J ) 23 Hz), 36.1, 34.6, 25.0, 23.9,
23.6; IR (neat) 3442, 2934, 2861, 2153, 1452, 1367, 1286, 1106, 1008,
2 H, J ) 7.1 Hz), 1.75-1.33 (m, 10 H), 1.29 (t, 3 H, J ) 7.1 Hz); 13
C
NMR (75 MHz, CDCl3) δ 165.5, 149.0 (t, J ) 23 Hz), 120.4, 110.1,
72.5 (t, J ) 24 Hz), 68.2 (quintet, J ) 23 Hz), 60.2, 36.1, 34.8, 25.0,
23.8, 23.7, 14.0; IR (neat) 2981, 2935, 2862, 2229, 2174, 2111, 1716,
1634, 1448, 1372, 1349, 1272, 1198, 1164, 1117, 1036; HRMS m/z
2
909, 738, 699; HRMS m/z 326.2028 (326.2027 calcd for C18H22O5 H4,
2
244.1610 (244.1608 calcd for C13H16O4 H4, M+).
M+).
(Z)-D,L-4,5-O-Cyclohexylidene-2-penten-3,4,5,5-d4-1-ol (9). Z ester
8 (4.21 g, 17.2 mmol) was added to 200 mL of CH2Cl2 and was cooled
to 0 °C. DIBAL-H (1 M in hexanes, 43.1 mmol) was slowly added
with a syringe over a period of 10 min. The mixture was stirred at
0-4 °C for 30 min, then the reaction was quenched with 100 mL of 1
M NaOH. The sodium hydroxide solution was added dropwise to
prevent explosive evolution of hydrogen. The quenched reaction was
stirred for 8 h at room temperature, then the aqueous and organic layers
were separated. The aqueous layer was extracted with 3 × 100 mL of
EtOAc. The organic layers were combined and dried over MgSO4,
concentrated, and purified by flash chromatography (15-30% EtOAc
in hexane), affording 3.24 g (16.0 mmol, 93%) of a light yellow oil
(9). 1H NMR (300 MHz, CDCl3) δ 5.81 (t, 1 H, J ) 6.5 Hz), 4.28
(dd, 1 H, J ) 13.2, 6.9 Hz), 4.15 (dd, 1 H, J ) 13.2, 6.1 Hz), 2.95 (s
br, 1 H), 1.75-1.30 (m, 10 H); 13C NMR (75 MHz, CDCl3) δ 132.9,
129.0 (t, J ) 24 Hz), 109.9, 70.8 (t, J ) 22 Hz), 68.2 (quintet, J ) 24
Hz), 58.2, 36.2, 35.3, 24.9, 23.8, 23.7; IR (neat) 3416, 2934, 2861,
2226, 2175, 2108, 1448, 1366, 1338, 1285, 1234, 1166, 1116, 1081,
D,L-1-(Benzyloxy)-4,5-O-cyclohexylidene-2,3-O-isopropyliden-
eribitol-3,4,5,5-d4 ((12). Diol (11 (2.03 g, 6.2 mmol), 2-methoxy-
propene (0.54 g, 7.5 mmol), and CH2Cl2 (100 mL) were combined,
and a single crystal of p-toluenesulfonic acid was added. The reaction
was stirred for 1 h and then concentrated. Purification by flash
chromatography (10-20% Et2O in petroleum ether) yielded 2.11 g of
the yellow oil (12 (5.8 mmol, 93%). 1H NMR (300 MHz, CDCl3) δ
7.46-7.22 (m, 5 H), 4.67 (d, 1 H, J ) 12.1 Hz), 4.55 (d, 1 H, J )
12.1 Hz), 4.41 (dd, 1 H, J ) 7.9, 3.1 Hz), 3.83 (dd, 1 H, J ) 10.6, 3.1
Hz), 3.63 (dd, 1 H, J ) 10.6, 7.9 Hz), 1.42 (s, 3 H), 1.34 (s, 3 H),
1.67-1.24 (m, 10 H); 13C NMR (75 MHz, CDCl3) δ 137.9, 128.1 (2
C), 127.7 (2 C), 127.4, 109.9, 108.7, 77.2 (t, J ) 22 Hz), 76.7, 73.3,
72.2 (t, J ) 23 Hz), 68.5, 66.7 (quintet, J ) 23 Hz), 36.3, 34.7, 27.7,
25.2, 24.9, 23.8, 23.7; IR (neat) 2934, 2861, 2117, 1452, 1380, 1370,
1271, 1243, 1219, 1182, 1109, 1070, 1052, 736, 698; HRMS m/z
2
366.2339 (366.2340 calcd for C21H26O5 H4, M+).
D,L-4,5-O-Cyclohexylidene-2,3-O-isopropylideneribitol-
3,4,5,5-d4 ((13). The benzyl ribitol (12 (2.34 g, 6.4 mmol) was
dissolved in 75 mL of THF and hydrogenated over 5% palladium on
carbon at 1 atm of hydrogen for 22 h. The reaction was filtered through
Celite and the filtrate was concentrated. Purification by flash chro-
matography (20-40% Et2O in petroleum ether) afforded 1.62 g (5.9
mmol, 92%) of alcohol (13. 1H NMR (300 MHz, CDCl3) δ 4.36 (dd,
1 H, J ) 7.3, 5.7 Hz), 3.90 (dd, 1 H, J ) 11.9, 7.3 Hz), 3.83 (dd, 1 H,
J ) 11.9, 5.7 Hz), 2.99 (s, 1 H), 1.72-1.25 (m, 10 H), 1.40 (s, 3 H),
1.35 (s, 3 H); 13C NMR (75 MHz, CDCl3) δ 110.4, 108.5, 77.6 (t, J )
24 Hz), 77.4, 72.2 (t, J ) 23 Hz), 66.9 (quintet, J ) 22 Hz), 60.6,
36.2, 34.6, 27.6, 25.0, 24.8, 23.9, 23.7; IR (neat) 3512, 2986, 2935,
2862, 2175, 2118, 1449, 1371, 1219, 1180, 1112, 1062, 861; HRMS
2
1035, 978, 928; HRMS m/z 202.1503 (202.1503 calcd for C11H14O3 H4,
M+).
(Z)-D,L-1-(Benzyloxy)-4,5-O-cyclohexylidene-2-pentene-
3,4,5,5-d4 (10). Sodium hydride, 60% dispersion in mineral oil (2.56
g, 64.1 mmol), was washed with hexane to remove the mineral oil.
THF (100 mL) was added to the sodium hydride. Allylic deuterated
alcohol 9 (3.24 g, 16.0 mmol), benzyl bromide (3.01 g, 17.6 mmol),
and tetrabutylammonium iodide (100 mg, 0.3 mmol) were added to
this mixture. The reaction was stirred for 24 h at room temperature
and quenched by addition of saturated NaCl (75 mL). Water was added
to dissolve precipitated salts, and the aqueous layer was separated. The
aqueous layer was extracted with three 50 mL portions of EtOAc, then
the combined organic layers were dried over MgSO4 and concentrated.
Purification by flash chromatography (5-15% Et2O in petroleum ether)
yielded 4.49 g of the yellow oil (6) (15.4 mmol, 96%). 1H NMR (300
MHz, CDCl3) δ 7.41-7.20 (m, 5 H), 5.80 (t, 1 H, J ) 6.4 Hz), 4.53
(d, 1 H, J ) 11.8 Hz), 4.48 (d, 1 H, J ) 11.8 Hz), 4.11 (d, 2 H, J )
6.4 Hz), 1.75-1.33 (m, 10 H); 13C NMR (75 MHz, CDCl3) δ 137.9,
130.5 (t, J ) 24 Hz), 129.8, 128.1 (2 C), 127.5 (2 C), 127.4, 109.6,
71.9, 70.9 (t, J ) 23 Hz), 68.1 (quintet, J ) 22 Hz), 65.4, 36.1, 35.2,
24.9, 23.7, 23.6; IR (neat) 2934, 2860, 2226, 2174, 2107, 1451, 1365,
1166, 1086, 1028, 1011, 738, 698; HRMS m/z 292.1970 (292.1972
2
m/z 276.1870 (276.1871 calcd for C14H20O5 H4, M+).
D,L-4,5-O-Cyclohexylidene-2,3-O-isopropylideneribose-
3,4,5,5-d4 ((14). Oxalyl chloride (53 mg, 0.42 mmol) was added to
CH2Cl2 (200 mL) and cooled to -78 °C. DMSO (66 mg, 0.84 mmol)
was added slowly via syringe, and the mixture was stirred for 5 min
before the deuterated alcohol (13 (104 mg, 0.38 mmol) was added to
the flask as a solution in CH2Cl2 (30 mL). After the reaction was stirred
for 30 min at -78 °C, triethylamine (154 mg, 1.52 mmol) was added
slowly to the mixture. After stirring for 15 min at -78 °C, the dry
ice/acetone bath was removed. Thirty minutes later, the reaction was
quenched with an aqueous HCl solution (15 mL, 0.12 M). The organic
and aqueous layers were immediately separated, and the organic layer
was washed with saturated aqueous NaHCO3 solution (20 mL) and
saturated aqueous NaCl solution (20 mL). The organic layer was dried
over MgSO4 and then concentrated. Purification by flash chromatog-
raphy (20-40% Et2O in petroleum ether) yielded 94 mg (0.34 mmol,
90%) of protected ribose ((14). 1H NMR (300 MHz, CDCl3) δ 9.75
(d, 1 H, J ) 1.8 Hz), 4.63 (d, 1 H, J ) 1.8 Hz), 1.73-1.28 (m, 10 H),
1.54 (s, 3 H), 1.38 (s, 3 H); 13C NMR (75 MHz, CDCl3) δ 197.2, 111.2,
110.7, 81.8, 78.4 (t, J ) 23 Hz), 72.6 (t, J ) 23 Hz), 66.5 (quintet, J
2
calcd for C18H20O3 H4, M+).
D,L-1-(Benzyloxy)-4,5-O-cyclohexylideneribitol-3,4,5,5-d4 ((11).
Benzyl deuterated allyl alcohol 10 (1.71 g, 5.85 mmol), N-methylmor-
pholine N-oxide (1.37 g, 11.7 mmol), and osmium tetroxide 2.5% in
tBuOH (3.05 g, 0.3 mmol OsO4) were combined in a solution of 1:8
D2O/acetone (72 mL) and stirred for 12 h at room temperature. The
reaction was quenched by addition of 40 mL of 30% Na2SO3. After
stirring for 1 h, the solution was extracted with four 75 mL portions of
EtOAc. The combined organic layers were dried over Na2SO4 and

7938 J. Am. Chem. Soc., Vol. 118, No. 34, 1996
Tolbert and Williamson
) 23 Hz), 36.5, 34.5, 27.3, 25.4, 25.0, 23.9, 23.7; HRMS m/z 274.1715
on a setting of 7, with a 2.5 min interval between bursts. Cellular
debris was removed by centrifugation at 31000g for 30 min. A 0.1
volume of a 20% streptomycin sulfate solution was then added to the
protein supernatant. After stirring for 15 min, the resulting precipitate
was removed by centrifugation at 31000g for 30 min. The supernatant
was then fractionated by ammonium sulfate precipitation. The am-
monium sulfate pellets were dissolved in the appropriate buffer and
dialyzed against 2 L of buffer for 8 h. The resulting solution was then
loaded onto a 2.5 × 14 cm DEAE column (DEAE-650M Toyopearl
resin from Supelco) that had been washed thoroughly with 3 M KCl
and then equilibrated with H2O. After loading the protein solution,
the column was washed with 100 mL of H2O, and the enzymes were
eluted with a salt gradient, the fractions being collected every 8 mL.
The column fractions were assayed for protein and then pooled and
concentrated by ammonium sulfate precipitation. Enzymes were stored
at -20 °C in 40-50% glycerol buffer solutions.
Ribokinase. The partial purification of ribokinase was accomplished
by a simplified form of the method of Hope et al.65 Inducible strain
MRi240/pJGK10 was grown on A+B minimal media64 supplemented
with glucose (4 g/L), ribose (4 g/L), and thiamine (7 mg/L). One liter
of media was inoculated with a 5 mL overnight culture of MRi240/
pJGK10. The cells were grown to an OD of 0.6, then 1 mL of a
solution of 3-â-indoleacrylic acid dissolved in ethanol (25 mg/mL) was
added to the culture to induce ribokinase production. Cells were
harvested 14 h after the addition of 3-â-indoleacrylic acid, and the cell
pellet was resuspended in 40 mL of 50 mM potassium phosphate buffer,
pH 7.5, and 2 mM 2-mercaptoethanol (buffer A). Cell disruption,
removal of cell debris, and streptomycin sulfate precipitation were
carried out as described in the general procedure. The 0-80%
saturation ammonium sulfate pellet was collected and dissolved in buffer
A for dialysis. DEAE chromatography was performed with a 500 mL
linear gradient of 0 to 300 mM KCl in buffer A. Column fractions
were assayed for ribokinase activity using the assay of Gross et al.52
The final ammonium sulfate pellet was dissolved in a minimum amount
of buffer A containing 40% glycerol. A total of 700 units of ribokinase
were obtained from 1 L of bacterial culture.
2
(274.1714 calcd for C14H18O5 H4, M+).
D,L-Ribose-3,4,5,5-d4 ((1). The protected deuterated ribose (14
(22 mg, 78 µmol) was dissolved in 7 mL of H2O:THF:AcOH (3:3:1).
The flask was attached to a reflux condenser and heated to 60 °C for
24 h. The mixture was evaporated under vacuum to afford crude D,L-
d4-ribose ((1). The amount of D-d4-ribose (-1) produced was
determined by enzymatic assay to be 31 µmol, corresponding to 78%
yield from starting protected D-d4-ribose -14 and 39% yield from
((14). NMR revealed the presence of the expected complex mixture
of R- and â-anomers of the furanose and pyranose forms of ribose.
2
HRFABMS (glycerol) m/z 137.0747 (137.0748 calcd for C5H5O4 H4,
[M - OH]+).
Enzymatic D-Ribose Assay. The amount of D-d4-ribose present in
the crude racemic mixture produced by the chemical synthesis was
quantitated by coupling D-ribose phosphorylation by ribokinase to
oxidation of NADH. This assay is similar to the ribokinase assay of
Gross et al.,52 except that the amount of D-ribose is limiting and the
absolute amount of NADH consumed is measured. Since there is a
1:1 relationship between NADH oxidation and D-ribose phosphoryla-
tion, the concentration of D-ribose can then be calculated by determining
the amount of NADH consumed. The assay solution (1 mL) contained
0.2 mM NADH, 1 mM PEP, 10 mM MgCl2, 3 mM ATP, 50 mM Tris-
HCl buffer pH 7.8, 2 units of lactate dehydrogenase, 2 units of pyruvate
kinase, and 1 unit of ribokinase. The assay solution was equilibrated
after the addition of the enzymes, and then the assay was started by
addition of a small aliquot of the unknown D-ribose solution. The
absorbance of NADH at 340 nm was monitored until no further change
was observed, and the total change in absorbance was used to calculate
the amount of NADH consumed, and therefore the amount of D-ribose
present in the aliquot, using an extinction coefficient of 6220 cm-1
mol-1
.
Cloning the Gene for Adenine Phosphoribosyltransferase. The
gene encoding E. coli adenine phosphoribosyltransferase was cloned
from the E. coli strain JM109 genome based on the reported gene
sequence62 using PCR with the oligonucleotides dCCG CGC GAA TTC
ATG ACC GCG ACT GCA CAG CAG CTT and dCCG GCG CTG
CAG TTA ATG GCC CGG GAA CGG GAC AAG as primers.62 The
PCR product was digested with EcoRI and PstI and ligated into
expression plasmid pKK223-3, that had been prepared by digestion
with EcoRI and PstI and dephosphorylation with calf intestinal alkaline
phosphatase. Transformation of this construct into E. coli strain JM109
produced an IPTG inducible adenine phosphoribosyltransferase over-
producing strain JM109/pTTA6.
Cloning the Gene for Xanthine-Guanine Phosphoribosyltrans-
ferase. The gene encoding E. coli xanthine-guanine phosphoribo-
syltransferase was cloned from the E. coli strain JM109 genome based
on the reported gene sequence63 using PCR with the oligonucleotides
dCCG CGC GAA TTC ATG AGC GAA AAA TAC ATC GTC ACC
and dCCG GCG CTG CAG TTA GCG ACC GGA GAT TGG CGG
GAC as primers.63 The PCR product was digested with EcoRI and
PstI and ligated into expression plasmid pKK223-3, that had been
prepared by digestion with EcoRI and PstI and dephosphorylation with
calf intestinal alkaline phosphatase. Transformation of this construct
into E. coli strain JM109 produced an IPTG inducible xanthine-guanine
phosphoribosyltransferase overproducing strain JM109/pTTG2.
General Enzyme Purification Scheme. All seven of the enzymes
not obtained commercially were overexpressed and purified using a
similar purification procedure. Cells were disrupted by sonication,
followed by streptomycin sulfate precipitation, ammonium sulfate
precipitation, and DEAE anion exchange chromatography. The general
outline of this purification is given below, with specific details and
exceptions for each enzyme following.
5-Phosphoribosyl-r-1-pyrophosphate Synthetase. The partial
purification of 5-Phosphoribosyl-R-1-pyrophosphate synthetase (PRPP
synthetase) was accomplished by a modified form of the procedure of
Switzer et al.55 A 5 mL overnight culture of strain HO561/pHO11,66
which constitutively expresses PRPP synthetase, was used to inoculate
a 1 L culture in LB media which was then grown for 12 h. After cell
harvest the cell pellet was resuspended in 40 mL of buffer A. After
sonication and removal of cellular debris, a 0.1 volume of 20%
streptomycin sulfate solution was added to the supernatant and the
mixture was stirred at 0 °C for 15 min. Before centrifuging, the
streptomycin sulfate mixture was heat treated in a water bath at 55 °C
for 5 min and then immediately cooled to 0 °C in an ice bath for at
least 4 min. The resulting precipitate was removed by centrifugation
at 31000 g for 30 min. The supernatant was brought to 35% saturation
of ammonium sulfate and stirred for 15 min. The ammonium sulfate
pellet was collected and dissolved in 25 mL of buffer A. The pH of
this solution was adjusted to 4.6 by addition of 1 M acetic acid at 0
°C, followed by immediate centrifugation for 10 min at 31000g. The
acid precipitate was collected and dissolved in buffer A. No DEAE
chromatography was used in the purification of PRPP synthetase. The
resulting solution was concentrated by ammonium sulfate precipitation
and stored in buffer A containing 50% glycerol. PRPP synthetase
activity was assayed using the method of Gross et al.52 From 1 L of
bacterial culture, 47 units of PRPP synthetase were obtained.
Orotate Phosphoribosyltransferase. The partial purification of
orotate phosphoribosyltransferase (OPRT) was accomplished by a
modified form of the procedure of Poulsen et al.67 Strain JM109/
pJTA43,68 which constitutively expresses OPRT, was grown on LB
media. One liter of medium was inoculated with a 25 mL overnight
culture and shaken for 14 h at 37 °C. The cells were harvested and
Cells were grown at 37 °C in either LB media or A+B minimal
media of Clark and Maaløe64 containing 50 µg/mL of ampicillin and
supplemented as indicated. All steps after cell growth were carried
out in a 4 °C cold room or on ice, unless otherwise specified. Cells
were harvested by centrifugation at 6000g for 15 min. The cell pellets
were resuspended in the appropriate buffer and disrupted with 30 30-s
sonication bursts using a Fisher Scientific 550 Sonic Dismembrator
(65) Hope, J. N.; Bell, A. W.; Hermodson, M. A.; Groarke, J. M. J.
Biol. Chem. 1986, 261, 7663-7668.
(66) Bjarne, H.-J. Mol. Gen. Genet. 1985, 201, 269-276.
(67) Poulsen, P.; Jensen, K. F.; Valentin-Hansen, P.; Carlsson, P.;
Lundberg, L. G. Eur. J. Biochem. 1983, 135, 223-229.
(68) Jensen, K. F.; Andersen, J. T.; Poulsen, P. J. Biol. Chem. 1992,
267, 17147-17152.
(62) Hershey, H. V.; Taylor, M. W. Gene 1986, 43, 287-293.
(63) Pratt, D.; Subramani, S. Nucleic Acids Res. 1983, 11, 8817-8823.
(64) Clark, D. J.; Maaløe, O. J. Mol. Biol. 1967, 23, 99-112.

Preparation of Specifically Deuterated RNA for NMR Studies
J. Am. Chem. Soc., Vol. 118, No. 34, 1996 7939
resuspended in 40 mL of 100 mM Tris buffer, pH 7.6, 2 mM EDTA,
and 2 mM 2-mercaptoethanol (buffer B). Sonication, removal of
cellular debris, and streptomycin sulfate precipitation were carried out
as described in the general procedure. The 0-75% saturation am-
monium sulfate pellet was collected from the supernatant of the
streptomycin sulfate precipitation and then dialyzed against buffer B.
The dialysate was subjected to DEAE chromatography, by eluting with
a 500 mL linear gradient from 50-300 mM KCl in buffer B. Column
fractions were assayed for OPRT activity according to the method of
Schawartz and Neuhard.69 Fractions containing OPRT were pooled
and ammonium sulfate precipitated. The ammonium sulfate pellet was
dissolved in a minimum amount of buffer B containing 40% glycerol.
From 1 L of culture, 31 units of orotate phosphoribosyltransferase were
obtained.
streptomycin precipitation and then dialyzed against buffer C. The
dialysate was subjected to DEAE chromatography with a 500 mL
gradient of 50-300 mM KCl in buffer C. Column fractions containing
XGPRT were identified by SDS PAGE gel electrophoresis, pooled,
concentrated by ammonium sulfate precipitation, and stored in buffer
C containing 50% glycerol. Xanthine-guanine phosphoribosyltrans-
ferase activity was estimated by the method described below, and from
1 L of culture, 28 units of XGPRT were obtained.
Xanthine-Guanine Phosphoribosyltransferase Assay. Xanthine-
guanine phosphoribosyltransferase activity was estimated by observing
GMP formation in a small GMP-forming reaction by injecting aliquots
onto an HPLC column as described in the preparative nucleotide
synthesis section below. The assay conditions were as follows for a 2
mL reaction: 1 mM PRPP, 10 mM MgCl2, 50 mM potassium phosphate
buffer, pH 7.5, 1 mg of solid guanine, and 1 unit of inorganic
pyrophosphatase. The reaction was stirred at room temperature and
initiated by adding an aliquot of XGPRT solution. The time elapsed
before the reaction was complete, determined by GMP formation
observed by HPLC, was used to calculate the activity of the XGPRT
solution. This method provides a rough estimate of the activity of the
XGPRT solution. However, the assay is very similar to our GTP-
forming reactions and thus suitable for empirical determination of the
amount of XGPRT required for preparative reactions.
CTP Synthetase. CTP synthetase was purified by a modified form
of the procedure of Anderson.58 One liter of LB medium was inoculated
with a 5 mL overnight culture of strain JM109/pMW5,71 which
constitutively overproduces CTP synthetase. The cells were grown for
12 h and then harvested. Sonication was performed as described in
the general procedure in buffer C containing 20 mM glutamine.
Removal of cellular debris and streptomycin sulfate precipitation were
conducted as in the general procedure. From the resulting supernatant,
the 0-65% ammonium sulfate pellet was extracted. This was dissolved
in buffer C containing 4 mM glutamine and was dialyzed against the
same buffer. The dialysate was subjected to DEAE chromatography
with a 500 mL gradient of 50-300 mM potassium phosphate buffer,
pH 7.5, containing 70 mM 2-mercaptoethanol and 4 mM glutamine.
Column fractions were assayed by the method of Anderson,58 and
fractions containing CTP synthetase were pooled, ammonium sulfate
precipitated, and dissolved in a minimum amount of buffer B containing
50% glycerol. From 1 L of culture, 38 units of CTP synthetase activity
were obtained.
Preparative Nucleotide Synthesis. All enzymatic reactions were
carried out under argon. Reactions were monitored by HPLC on a 25
× 4.6 mm Vydac 303NT405 nucleotide column, using a linear gradient
from 100% buffer A (0.045 M NH4CO2 brought to pH 4.6 with
phosphoric acid) to 100% buffer B (0.5 M NaH2PO4 brought to pH
2.7 with formic acid) in 10 min at a flow rate of 1 mL/min, with
detection at 254 nm. Potassium phosphate buffer (50mM) was used
in reactions containing PRPP synthetase because the enzyme is
inactivated in solutions with low phosphate concentration.55 The pH
of the reactions was monitored using pH paper and adjusted periodically
with 1 M NaOH or HCl to maintain the pH between 7.0 and 7.9. NTPs
were purified by using boronate affinity chromatography on Affigel
601 to remove the majority of salts and proteins in order to prepare
the nucleotides for transcription. Reaction mixtures were concentrated
under vacuum and then dissolved in a minimum amount of 1 M
triethylammonium bicarbonate (TEABC) solution, pH 9.5, prepared
by bubbling CO2 into an aqueous solution of triethylamine. In cases
where a residue could not be dissolved, the solutions were vacuum
filtered, and the precipitate was washed thoroughly with TEABC. These
solutions were loaded onto boronate affinity columns and purified as
described elsewhere.33
Orotidine-5′-monophosphate Decarboxylase. The partial purifica-
tion of OMP decarboxylase was accomplished by a modified version
of the procedure of Turnbough et al.70 Strain pDK26/N100,70 which
constitutively overproduces orotidine-5′-monophosphate decarboxylase
(OMP decarboxylase), was grown on A+B minimal media64 supple-
mented with 8 g of glucose, 0.5 g of casamino acids, 1 mL of 240 mM
UMP, and thiamine (7 mg/L). One liter of medium was inoculated
with a 5 mL overnight culture and then shaken at 37 °C for 12 h. After
cell harvest, the cells were resuspended in 40 mL of 64 mM Tris buffer
(pH 7.8) with 5 mM 2-mercaptoethanol (buffer C). Sonication, removal
of cellular debris, and streptomycin sulfate precipitation were carried
out as described in the general procedure. The 45-75% ammonium
sulfate pellet was collected from the supernatant of the streptomycin
sulfate precipitation and dissolved in buffer C for dialysis. The resulting
dialysate was subjected to DEAE chromatography, eluting with a 500
mL linear gradient of 0-300 mM NaCl in buffer C. Column fractions
containing OMP decarboxylase were detected by the method of
Turnbough et al.70 Active fractions were pooled, ammonium sulfate
precipitated, and stored in buffer C containing 40% glycerol. From 1
L of culture, 400 units of OMP decarboxylase were obtained.
Adenine Phosphoribosyltransferase. IPTG inducible, adenine
phosphoribosyltransferase (APRT) overproducing strain JM109/pTTA6
was grown on A+B minimal medium64 supplemented with 22 mL
glycerol, 2 g of glucose, and thiamine (7 mg/L). The medium was
inoculated with a 5 mL overnight culture, grown for 12 h, and then
induced with IPTG (0.234 g/L) for another 6 h. Cell harvest and
streptomycin sulfate precipitation were carried out as described in the
general procedure using buffer C. The 0-80% ammonium sulfate
precipitate was collected from the supernatant of the streptomycin
sulfate precipitation. The ammonium sulfate pellet was dialyzed and
then subjected to DEAE chromatography with a 500 mL linear gradient
of 0-300 mM KCl in buffer C. Column fractions containing APRT
were detected by the nonradioactive APRT assay described below. The
fractions containing APRT were pooled, ammonium sulfate precipitated,
and finally dissolved in a minimum amount of buffer C containing
50% glycerol and stored at -20 °C. From 1 L of culture, 360 units of
APRT were obtained.
Adenine Phosphoribosyltransferase Assay. Adenine phosphori-
bosyltransferase activity was measured by coupling the APRT reaction
to oxidation of NADH in a manner similar to the PRPP synthetase
assay.52 Conversion of NADH to NAD+ was monitored at 340 nm.
The assay solution (1 mL) contained 0.2 mM NADH, 1 mM PEP, 10
mM MgCl2, 1.5 mM PRPP, 1.5 mM adenine hydrochloride, 3 mM
ATP, 50 mM Tris-HCl buffer, pH 7.8, 2 units of lactate dehydrogenase,
2 units of pyruvate kinase, and 2 units of myokinase. The assay solution
was equilibrated, a 20 µL aliquot of APRT solution was added to start
the assay, and the absorbance change at 340 nm was monitored as a
function of time. The activity was obtained using an extinction
coefficient of 6220 cm-1 mol-1 for NADH.
Xanthine-Guanine Phosphoribosyltransferase. IPTG inducible,
xanthine-guanine phosphoribosyltransferase (XGPRT) overproducing
strain JM109/pTTG2 was grown on 1 L of LB media inoculated with
a 5 mL overnight culture for 12 h and then induced with IPTG (0.234
g/L) for another 6 h. Cells were harvested and streptomycin sulfate
precipitated as in the general procedure using buffer C. The 0-75%
ammonium sulfate fraction was collected from the supernatant of the
Adenosine-3′,4′,5′,5′-d4 5′-Triphosphate (2). Crude D,L-d4-ribose
((1) containing 42.3 mg (275 µmol) of D-d4-ribose as determined by
enzymatic ribose assay, was placed in a three-necked flask containing
adenine hydrochloride (87.5 mg, 510 µmol) and sodium 3-phospho-
glycerate (2.55 mmol). This was dissolved in 50 mL of a solution
containing 10 mM MgCl2, 10 mM dithiothreitol, and 50 mM potassium
phosphate buffer, pH 7.5. Argon was bubbled through this solution
for 5 min to remove oxygen. To this solution was added 125 units of
phosphoglycerate mutase, 50 units of enolase, 200 units of pyruvate
(69) Schwartz, M.; Neuhard, J. J. Bacteriol. 1975, 121, 814-822.
(70) Turnbough, C. L.; Kerr, K. H.; Funderburg, W. R.; Donahue, J. P.;
Powell, F. E. J. Biol. Chem. 1987, 262, 10239-10245.
(71) Weng, M.; Makaroff, C. A.; Zalkin, H. J. Biol. Chem. 1986, 261,
5568-5574.

7940 J. Am. Chem. Soc., Vol. 118, No. 34, 1996
Tolbert and Williamson
kinase, 25 units of myokinase, 5 units of ribokinase, 2 units of PRPP
synthetase, and 2 units of adenine phosphoribosyltransferase. The
reaction was initiated by adding a catalytic amount of ATP (0.5 mg,
0.9 µmol) to this solution. The reaction was stirred at 25 °C and
monitored by HPLC. The pH was maintained between 7.0 and 7.9
with 1 M HCl. An additional 0.5 mmol of 3-phosphoglycerate was
added every 5 h. Formation of ATP was initially slow but increased
as the reaction proceeded due to the increase in ATP concentration.
The reaction was stopped after 29 h by concentration under vacuum
and storage at -20 °C. Purification of the d4-ATP (2) was ac-
complished by boronate affinity chromatography. Compound 2 was
quantitated by absorption at 259 nm (ꢀ259 ) 15 400 cm-1 mol-1). This
reaction yielded 213 µmol of 2 (78% yield). 1H NMR (500 MHz, D2O)
δ 8.55 (s, 1H), 8.26 (s, 1H), 6.14 (d, 1H, J ) 6.0 Hz), 4.80 (d, 1H, J
) 6.0 Hz); MS m/z 510.1 (510.07 calcd for C10H112H4N5O13P3).
Guanosine-3′,4′,5′,5′-d4 5′-Triphosphate (3). Crude D,L-d4-ribose
((1) containing 42.3 mg (275 µmol) of D-d4-ribose as determined by
enzymatic ribose assay, was placed into a three-necked flask containing
guanine (77.1 mg, 510 µmol) and sodium 3-phosphoglycerate (2.55
mmol). This was mixed with 50 mL of a solution containing 10 mM
MgCl2, 10 mM dithiothreitol, and 50 mM potassium phosphate buffer,
pH 7.5. Argon was bubbled through this inhomogeneous solution for
5 min to remove oxygen. To this mixture was added 250 units of
phosphoglycerate mutase, 100 units of enolase, 200 units of pyruvate
kinase, 38 units of myokinase, 5 units of ribokinase, 2 units of PRPP
synthetase, and 0.2 units of xanthine-guanine phosphoribosyltrans-
ferase. The reaction was initiated by adding 4.4 µmol of d4-ATP (2).
The reaction was stirred at 25 °C and monitored by HPLC. The pH
was maintained between 7.0 and 7.9 with 1 M HCl. Additional
3-phosphoglycerate was added to the reaction at a rate of 0.24 mmol
every 5 h. After 4 h 2 units of guanylate kinase was added, and after
21 h, an additional 4.4 µmol of 2 was added. The reaction was stopped
after 72 h by concentration under vacuum and storage at -20 °C.
Purification of the d4-GTP (3) was accomplished by boronate affinity
chromatography. Compound 3 was quantitated by the absorbance at
253 nm (GTP ꢀ253 ) 13 700 cm-1 mol-1). This reaction yielded 206
µmol of 3 (75% yield). 1H NMR (500 MHz, D2O) δ 8.14 (s, 1H),
5.93 (d, 1H, J ) 6.0 Hz), 4.80 (d, 1H, J ) 6.0 Hz); MS m/z 526.2
(526.07 calcd for C10H112H4N5O14P3).
Uridine-3′,4′,5′,5′-d4 5′-Triphosphate (4). Crude D,L-d4-ribose ((1)
containing 84.7 mg (550 µmol) of D-d4-ribose as determined by
enzymatic ribose assay, was placed into a three-necked flask containing
potassium orotate (198 mg, 1.02 mmol) and sodium 3-phosphoglycerate
(10.2 mmol). This was dissolved in 100 mL of a solution that contained
10 mM MgCl2, 10 mM dithiothreitol, and 50 mM potassium phosphate
buffer, pH 7.5. Argon was bubbled through this solution for 5 min to
remove oxygen. To this solution was added 375 units of phospho-
glycerate mutase, 150 units of enolase, 300 units of pyruvate kinase,
38 units of myokinase, 5 units of ribokinase, 2 units of PRPP synthetase,
1.5 units of orotate phosphoribosyltransferase, and 10 units of orotidine-
5′-monophosphate decarboxylase. The reaction was initiated by adding
6.5 µmol of d4-ATP (2) and then stirred at 25 °C. The reaction was
monitored by HPLC, and the pH was maintained between 7.0 and 7.9
with 1 M HCl. An additional 0.35 mmol of 3-phosphoglycerate was
added every 5 h. After 4 h 1 unit of nucleoside monophosphate kinase
was added to catalyze the formation of UDP from UMP. After 21 h,
10.9 µmol of 2 was added to stimulate the rate of the reaction. The
reaction was stopped after 72 h by concentration under vacuum and
storage at -20 °C. Purification of the d4-UTP (4) was accomplished
by boronate affinity chromatography. Compound 4 was quantitated
by the absorbance at 260 nm (ꢀ260 ) 10 000 cm-1 mol-1). This reaction
yielded 495 µmol of 4 (90% yield). 1H NMR (500 MHz, D2O) δ 7.96
(d, 1H, J ) 8.1 Hz), 5.99 (d, 1H, J ) 5.4 Hz), 5.97 (d, 1H, J ) 8.4
Hz), 4.38 (d, 1H, J ) 5.4 Hz); MS m/z 487.1 (487.05 calcd for
C9H102H4N2O15P3).
Cytidine-3′,4′,5′,5′-d4 5′-Triphosphate (5). d4-UTP (4) (247 µmol),
glutamine (369 mg, 2.52 mmol), and sodium 3-phosphoglycerate (505
µmol) were placed into a three-necked flask. This was dissolved in
250 mL of a solution containing 10 mM MgCl2 and 50 mM Tris-HCl
buffer, pH 7.8. Argon was bubbled through this solution for 5 min to
remove oxygen. To this solution was added 250 units of phospho-
glycerate mutase, 100 units of enolase, 200 units of pyruvate kinase,
and 5.5 units of cytidine-5′-triphosphate synthetase. To initiate the
reaction 26.9 µmol of d4-ATP (2) and 2.5 µmol of d4-GTP (3) were
added. The reaction was stirred at 25 °C and monitored by HPLC.
After 3.5 h, 26.9 µmol of d4-ATP, 2.5 µmol of d4-GTP, and 0.86 mmol
of 3-phosphoglycerate were added to stimulate the rate of the reaction.
The reaction was stopped after 8 h by concentration under vacuum
and storage at -20 °C. Purification of the d4-CTP (5) was ac-
complished by DEAE chromatography. The 5 was eluted from the
DEAE resin with a TEABC salt gradient. Boronate affinity chroma-
tography was not possible due to the use of tris buffer, which binds to
the boronate affinity resin. Compound 5 was quantitated by absorbance
at 259 and 280 nm, using the extinction coefficients ATP ꢀ259 ) 15 400,
ATP ꢀ280 ) 1911, CTP ꢀ259 ) 7204, and CTP ꢀ280 ) 6905. This
reaction yielded 215 µmol of 5 (87% yield). 1H NMR (500 MHz, D2O)
δ 8.07 (d, 1H, J ) 7.7 Hz), 6.21 (d, 1H, J ) 7.7 Hz), 5.99 (d, 1H, J
) 4.4 Hz), 4.34 (d, 1H, J ) 4.4 Hz); MS m/z 486.1 (486.06 calcd for
C9H112H4N3O14P3).
RNA Synthesis. TAR RNA (5′GGCCAGAUUGAGCCUGG-
GAGCUCUCUGGCC3′) was synthesized by in Vitro transcription with
T7 RNA polymerase using unlabeled NTPs from Sigma and d4-NTPs
(2-5).72,32 The RNA was purified by 20% PAGE, electroeluted, and
dialyzed against 50 mM NaCl, 0.1 mM EDTA, and 10 mM sodium
phosphate, pH 6.4.
NMR Experiments. NOESY spectra of TAR and d4-TAR were
recorded on a Varian VXR-500 spectrometer with a spectral width of
5500 Hz, acquiring 4096 complex points in t2 and 644 in t1, 64 or 32
scans per FID for TAR and d4-TAR, respectively, a relaxation delay
of 1.6 s, and a mixing time of 0.200 s. Sample conditions were 50
mM NaCl, 0.1 mM EDTA, 10 mM sodium phosphate, pH 6.4, 25 °C.
RNA concentrations were 0.83 mM for TAR and 1.12 mM for d4-
TAR. Twice the number of scans were acquired per FID for the
unlabeled TAR sample to compensate for the difference in concentra-
tions between the two RNA samples.
Nonselective T1 relaxation rates for individual resonances were
measured using inversion recovery, and nonselective T2 relaxation rates
were measured using a CPMG experiment. For both T1 and T2
measurements a relaxation delay of 75 s between each of four scans
was used.
Acknowledgment. We would like to thank Dr. JoAnne
Stubbe, Dr. Kaj Frank Jensen, Dr. Charles L. Turnbough Jr.,
and Dr. Howard Zalkin for generously providing plasmids and
overproducing strains. We thank Dr. Pete Wishnok for nucleo-
side triphosphate mass spectra, Alex Brodsky and John Battiste
for help with the 500 MHz NMR, Robert Batey for help with
cloning, Dr. Jean-Luc Montchamp for helpful discussions on
synthesis, and Dr. Patrick Zarrinkar for critical reading of the
manuscript. This work was supported by the Searle Scholars
Program of the Chicago Community Trust, and grants from the
National Institutes of Health (GM-53320 and GM-53757).
JA961274I
(72) Milligan, J. F.; Uhlenbeck, O. C. Methods Enzymol. 1989, 180, 51-
62.