Enzymatic synthesis of ribo- and 2′-deoxyribonucleosides from glycofuranosyl phosphates: An approach to facilitate isotopic labeling

Article abstract of DOI:10.1016/j.carres.2017.07.006

Milligram quantities of α-D-ribofuranosyl 1-phosphate (sodium salt) (αR1P) were prepared by the phosphorolysis of inosine, catalyzed by purine nucleoside phosphorylase (PNPase). The αR1P was isolated by chromatography in >95% purity and characterized by ¹H and ¹³C NMR spectroscopy. Aqueous solutions of αR1P were stable at pH 6.4 and 4 °C for several months. The isolated αR1P was N-glycosylated with different nitrogen bases (adenine, guanine and uracil) using PNPase or uridine phosphorylase (UPase) to give the corresponding ribonucleosides in high yield based on the glycosyl phosphate. This methodology is attractive for the preparation of stable isotopically labeled ribo- and 2′-deoxyribonucleosides because of the ease of product purification and convenient use and recycling of nitrogen bases. The approach eliminates the need for separate reactions to prepare individual furanose-labeled ribonucleosides, since only one ribonucleoside (inosine) needs to be labeled, if desired, in the furanose ring, the latter achieved by a high-yield chemical N-glycosylation. 2′-Deoxyribonucleosides were prepared from 2′-deoxyinosine using the same methodology with minor modifications.

Full text of DOI:10.1016/j.carres.2017.07.006

Accepted Manuscript
Enzymatic synthesis of ribo- and 2′-deoxyribonucleosides from glycofuranosyl
phosphates: An approach to facilitate isotopic labeling
Wenhui Zhang, Toby Turney, Ivana Surjancev, Anthony S. Serianni
PII:
S0008-6215(17)30363-4
10.1016/j.carres.2017.07.006
CAR 7420
DOI:
Reference:
To appear in: Carbohydrate Research
Received Date: 17 May 2017
Revised Date: 13 July 2017
Accepted Date: 14 July 2017
Please cite this article as: W. Zhang, T. Turney, I. Surjancev, A.S. Serianni, Enzymatic synthesis of ribo-
and 2′-deoxyribonucleosides from glycofuranosyl phosphates: An approach to facilitate isotopic labeling,
Carbohydrate Research (2017), doi: 10.1016/j.carres.2017.07.006.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Graphical abstract

ACCEPTED MANUSCRIPT
Enzymatic Synthesis of Ribo- and 2’-Deoxyribonucleosides from Glycofuranosyl
Phosphates: An Approach to Facilitate Isotopic Labeling
Wenhui Zhang*, Toby Turney, Ivana Surjancev and Anthony S. Serianni*
Department of Chemistry and Biochemistry, University of Notre Dame,
Notre Dame, IN 46556-5670 USA
Authors for correspondence: wzhangi@nd.edu, aseriann@nd.edu
1

ACCEPTED MANUSCRIPT
Abstract
Milligram quantities of α-D-ribofuranosyl 1-phosphate (sodium salt) (αR1P) were prepared
by the phosphorolysis of inosine, catalyzed by purine nucleoside phosphorylase (PNPase). The
1
13
αR1P was isolated by chromatography in >95% purity and characterized by H and C NMR
o
spectroscopy. Aqueous solutions of αR1P were stable at pH 6.4 and 4 C for several months.
The isolated αR1P was N-glycosylated with different nitrogen bases (adenine, guanine and
uracil) using PNPase or uridine phosphorylase (UPase) to give the corresponding
ribonucleosides in high yield based on the glycosyl phosphate. This methodology is attractive
for the preparation of stable isotopically labeled ribo- and 2’-deoxyribonucleosides because of
the ease of product purification and convenient use and recycling of nitrogen bases. The
approach eliminates the need for separate reactions to prepare individual furanose-labeled
ribonucleosides, since only one ribonucleoside (inosine) needs to be labeled, if desired, in the
furanose ring, the latter achieved by a high-yield chemical N-glycosylation.
2’-
Deoxyribonucleosides were prepared from 2’-deoxyinosine using the same methodology with
minor modifications.
Keyword:
enzymatic
synthesis,
stable
isotopic
labeling,
ribonucleosides,
2’-
deoxyribonucleosides, α-D-ribofuranosyl 1-phosphate
2

ACCEPTED MANUSCRIPT
1. Introduction
The preparation of stable isotopically labeled ribo- and 2’-deoxyribonucleosides requires
1
efficient methods to introduce isotopes into their nitrogen base and furanose constituents, and
2
the assembly of isotopically labeled oligonucleotides requires at least eight different labeled
13
nucleosides. While uniform labeling (for example, with C) can be achieved biologically (e.g.,
13
1b,3
via algal growth on
CO ),
the synthetic challenges to prepare selectively labeled
2
nucleosides in a cost effective manner remain significant, in part because of the large number of
13
2
15
C/ H/ N isotopomers that are possible if both the base and sugar constituents are
1d,1e
considered.
To simplify this labeling, a synthetic strategy was investigated and optimized in
which isotope is initially inserted into the furanosyl ring of inosine, which is prepared chemically
hypoxanthine (HX) + ABR
in high yield. The labeled inosine is then
Vorbrüggen reaction
inosine (I)
Pi
*
catalase (CT)
phosphorolyzed by purine nucleoside
1/2O2 + H2O
H2O2
PNPase
hypoxanthine (HX)
4
uric acid (UA)
phosphorylase (PNPase: E.C. 2.4.2.1) to
xanthine
oxidase (XO)
α-D-ribose 1-phosphate (αR1P)
give α-
salt (αR1P), which is isolated in pure form.
The αR1P is then N-glycosylated
D-ribofuranosyl 1-phosphate sodium
uracil ;
UPase
*
2'-deoxyinosine (dI) can
substitute for I to give
2d-αR1P instead of αR1P.
adenine or guanine;
PNPase
Pi
Pi
uridine (U)
adenosine (A)
or
guanosine (G)
Scheme 1. Preparation of ribonucleosides and 2'-
deoxyribonucleosides from inosine and 2’-deoxyinosine,
respectively.
enzymatically to give a range of labeled
nucleosides (Scheme 1), each of which can
5
2
be converted into a labeled nucleotide for use on DNA/RNA synthesis. Since free nitrogen
bases are used as reactants in the latter N-glycosylation reactions, isotope labeling in both the
sugar and base constituents is easily accomplished without the need for complex purifications.
2’-Deoxyinosine (dI) can used in the place of inosine to prepare 2’-deoxyribonucleosides. This
approach mimics that currently used to prepare monosaccharide isotopomers, wherein, for
13
example, D-[1- C]glucose is prepared chemically and subsequently converted to other labeled
13
13
13
sugars such as
D
-[1- C]mannose,
D
-[2- C]mannose, or D-[1- C]fructose via chemical or
6
enzymatic rearrangements.
3

ACCEPTED MANUSCRIPT
Previous applications of nucleoside phosphorylases to prepare labeled (and unlabeled)
13
nucleosides employ mixtures in which a labeled nucleoside (e.g., [1’- C]uridine) is mixed with a
nitrogen base (e.g., adenine), and the mixture is treated with two phosphorylases (e.g., PNPase
13
7
and uridine phosphorylase) to give [1’- C]adenosine (Scheme 2).
This approach gives
equilibrium mixtures containing two nucleosides and two free bases, thus requiring careful and
sometimes time-consuming chromatography to isolate the labeled nucleosides and/or recover
labeled nitrogen base (if used in
Scheme 2. Conventional application of
two nucleoside
the conversion). We reasoned
that the isolation of the αR1P
intermediate implicated in the
conversion shown in Scheme 2
phosphorylases to interconvert isotopically labeled pyrimidine and
purine ribonucleosides. A pyrimidine nucleoside typically serves as
the donor and a purine base as the acceptor, since the reaction
equilibrium favors the purine nucleoside, especially when modified
purine bases are used. Chromatographic purification of the resulting
mixtures is often required to isolate the labeled nucleoside and
recover the labeled precursor.
would simplify the purification of labeled products, improve labeling options in both the sugar
and nitrogen base, and render commercial applications more flexible and reproducible,
assuming that αR1P had a sufficiently long shelf life. In addition, other applications become
available if αR1P can be isolated in pure form. However, to date, αR1P has not been isolated
in pure form for direct use as a nucleoside phosphorylase substrate, it has not been
8
characterized by NMR, and its solution stability remains unknown.
The success of the improved synthetic strategy shown in Scheme 1 depends on (1)
driving the PNPase reaction to completion, and (2) isolating and purifying the αR1P
phosphorolysis product. Inosine (I) was chosen as the common nucleoside precursor because
it, and its 2’-deoxy derivative (2’-deoxyinosine, dI), can be prepared chemically in high yield.
1a,1c,9
Vorbrüggen coupling
of 1-O-acetyl-2,3,5-tri-O-benzoyl-α-D-ribofuranoside (ABR; Scheme
1) to HX to give I is reliable and proceeds in high yield (~80%), and the chemical 2’-
10
deoxygenation of I to give dI also proceeds in reasonable overall yield.
In addition,
phosphorolysis can be driven to completion with I or dI as substrates. Xanthine oxidase (XO;
E.C. 1.1.3.22) converts hypoxanthine (HX) released from the PNPase reaction to uric acid (UA)
4

ACCEPTED MANUSCRIPT
and H O , and catalase (E.C. 1.11.1.6) drives the overall reaction to completion by converting
2
2
11
H O to O and H O (Scheme 1).
This coupling strategy has been used previously in an
2
2
2
2
12
enzyme assay of inosine, and was applied here to enable large-scale preparations of αR1P.
2. Results and Discussion
2.1. Preparation and Isolation of
α-D-Ribofuranosyl 1-Phosphate (Sodium Salt) (αR1P)
from Inosine. PNPase reaction conditions were
optimized using a 55 µM solution of I (ε251 = 1.09
4
-1
-1 13
×10 M cm ) in 10 mM sodium phosphate (NaP )
i
o
buffer at pH 7.4 and 37 C containing PNPase (0.2
units), and the appearance of HX was monitored
4
-
spectrophotometrically at 248 nm (ε248 = 1.08×10 M
1
-1 13
Figure 1. Formation of HX and UA in a 55 µM
cm )
(Figure 1).
The same reaction was
solution of I treated with only PNPase (νo
=
10.3 ± 3.7 µM product/min) or with a mixture
of PNPase and xanthine oxidase (νo = 19.5 ±
0.6 µM product/min). Product concentrations
were determined spectrophotometrically at
248 nm (for HX) and 290 nm (for UA).
repeated with the addition of 2 units of XO and
catalase (1000 units), and the production of UA was
4
-1
-1 14
monitored at 290 nm (ε290 = 1.32×10 M cm ) .
The reaction without XO reached equilibrium in ~3 min with only ~50% of the inosine
phosphorolyzed, but complete phosphorolysis was
observed after 11 min when XO and catalase were
present. The initial velocity of the reaction also
increased when XO and catalase were present.
13
Selective C-labeling of I was then used to
13
directly observe the fate of the labeled carbon by
C
15
Scheme 3. Phosphorolysis of [1'-13C]inosine
to give α-[1-13C]R1P.
NMR during phosphorolysis.
The production of
αR1P was confirmed in a reaction involving a 500 µM
13
solution of [1’- C]inosine in 50 mM NaP buffer at pH 7.4 containing PNPase (1 units), XO (2
i
5

ACCEPTED MANUSCRIPT
13
1
units) and catalase (2000 units) (Scheme 3). The C{ H} NMR spectrum of the reaction
mixture obtained before the addition of PNPase contained a prominent singlet at 90.9 ppm
arising from the labeled C1’ of I (Figure 2A). After the
o
addition of PNPase and 3 h of reaction at 37 C,
spectrophotometric assay indicated that the conversion
was complete. This result was confirmed by the
13
1
C{ H} NMR spectrum of the reaction mixture, which
lacked the signal at 90.9 ppm and contained a new
13
1
Figure 2. C{ H} NMR spectra (150 MHz)
13
prominent doublet at 100.3 ppm (Figure 2B). The latter
of the PNPase reaction with [1’- C]inosine
as the substrate. (A) Reaction mixture
without PNPase showing only the labeled
C1’ signal of [1’- C]inosine. (B) Reaction
13
doublet was assigned to C1 of α-
D-[1- C]R1P, with
13
13 31
mixture after the addition of PNPase and
the signal splitting caused by geminal C- P spin-
o
incubation for 3 h at 37 C, showing the
13
labeled C1 signal of α-D-[1- C]R1P. The
2
coupling ( J
= 5.2 Hz). These results confirmed
C1,O,P
signal splitting (inset) is caused by
2
J
.
C1,O,P
that I was completely phosphorolyzed to αR1P and HX
when XO and catalase were present in the reaction mixture.
The reaction scale was then increased to 250 µmol of I (67 mg) to facilitate the isolation
and purification of αR1P. When the reaction was complete, the reaction mixture was applied to
o
a Bio-Gel P-2 size exclusion column at 22 C and eluted with DI water. The αR1P eluted as a
o
single peak, and fractions were pooled, concentrated and stored at pH 6.4 and 4 C. The
concentration of αR1P in the latter solution was
1
12 mM, as determined by H NMR (see Figures
16
S1–S3), giving an isolated yield of 74% based
on I. Little if any degradation occurred in dilute
aqueous solutions of αR1P after storage for three
o
months at pH 6.4 and 4 C.
1
2
The 1D H NMR spectrum of αR1P in H O
2
1
Figure3. Expanded regions of the
2
H NMR
2
spectrum (600 MHz) of αR1P in H O at 22 °C
2
at p H 6.4 (Figure 3) (see Figure S4 for the full
and pH 6.4, externally referenced to DSS,
showing signal assignments and J-couplings.
spectrum) is essentially first-order at 600 MHz
6

ACCEPTED MANUSCRIPT
1
(Tables 1 and 2). A comparison of the H chemical shifts of αR1P with those of methyl α-
D
-
20
ribofuranoside (MeαR)
shows that only H1 of αR1P is significantly affected by O1
1
13
Table 1. H and C Chemical Shift Assignments
for αR1P and Methyl α- -Ribofuranoside (MeαR)
phosphorylation, shifting downfield by 0.5
D
1
ppm. The remaining H chemical shifts are
essentially identical in both compounds.
3
J
values in αR1P (4.1 Hz) and MeαR
H1,H2
(4.2 Hz) are virtually identical, confirming the
α-configuration in the phosphorylated
a 1
13
H and C chemical shifts in ppm relative to
2
external DSS; measured at 22 °C in H O (80
2
3
product. The remaining J values in αR1P
HH
b
mM). Values taken from ref. 17 and adjusted
using the formula δDSS = δTMS + 2.0 ppm.
18
are very similar to corresponding values in
MeαR, suggesting similar ring and exocyclic hydroxymethyl conformations in both
20
compounds.
13
1
The C{ H} NMR spectrum of αR1P (Figure S5) showed the C1 and C2 signals split by
2
3
J
and J
, respectively (Tables 1 and 2). Phosphorylation at O1 affects δ significantly,
C2,P C1
C1,P
31
1
shifting this signal upfield by ~6 ppm relative to that for MeαR. The P{ H} NMR spectrum of
αR1P is shown in Figure S6 (Supporting Information).
2.2. N-Glycosylation Reactions to Produce Ribonucleosides from
αR1P. The isolated
αR1P was used to prepare ribonucleosides by enzyme-catalyzed N-glycosylation (Scheme 1).
A reaction mixture containing αR1P,
1
1
13 31
1
31
Table 2. H- H, C- P, H- P Coupling Constants in
αR1P and Methyl α-
D
-Ribofuranoside (MeαR)
adenine and PNPase in DI water
2
containing 10% v/v H O was incubated
2
o
at 37 C (Scheme 4), and the reaction
1
was monitored by H NMR. About 60%
a
2
of the αR1P was converted to adenosine
In Hz ± 0.1 Hz at 22 °C in H O; br denotes broaden
2
b
signal (< 0.5 Hz). Values were taken from ref. 19.
after 4 h (Figure S7), but signals from
o
both compounds disappeared after incubation at 37 C overnight (Figure S7), presumably the
consequence of microbial degradation. The reaction was rerun with a drop of CHCl added to
3
7

ACCEPTED MANUSCRIPT
the reaction mixture, and with excess adenine to drive the conversion to completion. Under
these modified conditions, >90% conversion of αR1P to adenosine was observed after 20 h at
o
21
o
37 C (Figure 4). No degradation of adenosine was observed after 2 days at 37 C.
Only ~2% of αR1P was converted to
guanosine after 21 h using the same reaction
conditions to prepare adenosine (Figure S8),
presumably because of the low solubility of
Scheme 4. N-Glycosylation of αR1P with adenine
to give adenosine, catalyzed by PNPase.
2
guanine. When a reaction solution containing
80:10:10 v/v/v H O: H O:DMSO-d
6
was
2
2
22
o
used, an ~40% conversion to guanosine was observed after incubation at 37 C for 9 days
21b
(Figure S9).
PNPase remained active under these reaction conditions, and unreacted αR1P
was detected at the end of the reaction, confirming its stability under these long reaction times.
Two unidentified by-products were also observed (Scheme S1), one decreasing and the other
23
increasing in concentration as reaction time increased.
Enzyme-mediated N-glycosylation of αR1P was also used to prepare uridine. Reaction
mixtures containing αR1P, uracil, and uridine phosphorylase (UPase; E.C. 2.4.2.3) in 90:10 v/v
1
2
o
H O: H O were incubated at 37 C (Scheme S2). About 75% of the αR1P was converted to
2
2
21b
uridine after 2.5 h (Figure S10),
and this
conversion increased to 85% after
additional 18 h and with additional uracil
supplied to the reaction mixture. Cytidine,
however, was not detected in a similar
reaction with cytosine after 20 h of reaction
1
Figure 4. H NMR spectrum (600 MHz) of a reaction
mixture containing αR1P, adenine and PNPase in
90%:10% v/v H O: H O after 20 h at 37 C.
1
2
o
2
2
Signals arising from the β-ribofuranosyl constituent
of adenosine appear between 3–6 ppm
(assignments shown in red). Weak signals are also
observed from unreacted αR1P. Signals H2 and H8
arise from the base constituent of adenosine, and
signals B1 and B2 arise from unreacted adenine.
(Figure S11), indicating that cytosine is not
24
a UPase substrate.
However, since
uridine can be converted to cytidine
8

ACCEPTED MANUSCRIPT
25
chemically in high yield, access to all four ribonucleosides is possible from either an unlabeled
or labeled inosine precursor.
13
13
The practicality of preparing
C-labeled ribonucleosides via N-glycosylation of
C-
13
labeled αR1P was investigated by phosphorolyzing [1’- C]inosine (250 µmol, 67 mg) with
13
PNPase and isolating the resulting α-
D
-[1- C]R1P by Bio-Gel P-2 gel filtration chromatography
13
as described above. This reaction gave ~185 µmol of α-D-[1- C]R1P in an isolated yield of 74%
13
(Figure S12). A reaction mixture containing α-
D
-[1- C]R1P (107 µmol, 25 mg), adenine (1.48
mmol, 200 mg) and PNPase (4.0
U) in DI water containing one drop
o
of CHCl was incubated at 37 C
3
(Scheme 4). A >95% conversion
13
of
α-
D
-[1- C]R1P
to
[1’-
13
C]adenosine was observed after
o
21
3 days at 37 C (Figure S13).
1
13
Figure 5. H NMR spectra (600 MHz) of [1’- C]adenosine
Excess adenine was removed by
centrifugation of the reaction
mixture, the supernatant was
concentrated and applied to a Bio-
that was purified on a Bio-Gel P2 column, showing signal
assignments. The H1 signal is split by the large
(165.6 Hz).
(1.3 Hz) were also obtained from the H2, H3 and
H4 signals, respectively.
1
J
C1’,H1’
2
3
J
(3.1 Hz),
J
(5.1 Hz) and
C1’,H2’
C1’,H3’
3+4
J
C1’,H4’
o
Gel P-2 column (2.5 cm x 100 cm) at 22 C, and the column was eluted with DI water. The
13
elution volumes of [1’- C]adenosine (~650 mL) and adenine (~800 mL) were greater than
13
those of α-
D
-[1- C]R1P (~200 mL) and inorganic phosphate salts (~270 mL), presumably due
26
to attraction between the nitrogen base and the Bio-Gel P-2 gel matrix.
This chromatography
proved very effective in removing unreacted αR1P and salts from the ribonucleoside product,
13
13
and in separating [1’- C]adenosine from unreacted adenine. [1’- C]Adenosine was obtained
in high isolated yield (94%) and in high purity (Figure 5). Since unreacted adenine is
conveniently recovered by centrifugation and/or subsequent chromatography, the use of
9

ACCEPTED MANUSCRIPT
isotopically labeled nitrogen bases in N-glycosylation reactions with either unlabeled or
isotopically labeled αR1P is practical from an economic standpoint.
13
Small-scale N-glycosylation reactions with α-D-[1- C]R1P were also conducted to
13
13
prepare [1’- C]guanosine (data not shown) and [1’- C]uridine (Figure S14). No attempts were
13
made to purify the product C-labeled ribonucleosides, but purification by chromatography on
13
Bio-Gel P-2 should be achievable based on findings from the purification of [1’- C]adenosine.
2.3. Preparation and Isolation of 2-Deoxy-
α-D-ribofuranosyl 1-Phosphate (Sodium Salt)
(2d-
αR1P) from 2’-Deoxyinosine (dI). A small-scale phosphorolysis reaction of 2’-deoxyinosine
(dI) (20 µmol, 5 mg) was initially conducted in 50 mM sodium phosphate buffer at pH 7.4 and 37
o
C to prepare 2-deoxy-α-
D
-erythro-pentofuranosyl 1-phosphate (sodium salt) (2-deoxy-α-
D
-
O
ribofuranosyl 1P; 2d-αR1P).
2’-
N
N
HO
NaPi
NH
HO
O
O
2 Na+
HO
N
OH
PNPase
NaPi
O
Deoxyinosine was completely
2ꢀ
OPO3
OH
2ꢀdeoxyꢀ ꢀribose
OH
O
D
2dꢀ R1P
OH
2'ꢀdeoxyinosine
consumed after 2 days, but the 2d-
N
NH
PNPase
HO
αR1P product was unstable,
N
H
N
N
O
N
N
catalase
NH
XO
O
H2O2
1/2O2 + H2O
NaPi
with >80% degrading to 2-deoxy-D-
O
H
N
erythro-pentose (2-deoxy-D-ribose)
NH
OH
2'ꢀdeoxyꢀN7ꢀinosine
O
N
H
O
N
27
H
(Scheme 5).
Thus, a larger-scale
Scheme 5. Phosphorolysis of 2'-deoxyinosine, catalyzed by
PNPase, to give 2d-αR1P.
reaction with dI (0.25 mmol, 63 mg)
o
was then conducted at 22 C in 50
mM sodium phosphate buffer at pH 7.4. After 20 h, ~80% of the dI had converted to 2d-αR1P
28
with little degradation.
The product 2d-αR1P was purified on a Bio-Gel P-2 column (2.5 cm x
o
100 cm) developed at 4 C, giving an isolated ~20% yield of 2d-αR1P based on dI (Figure S15).
Some 2d-αR1P degradation occurred during column purification.
PNPase-mediated phosphorolysis of dI (40 µmol, 10 mg) was also investigated in 10 mM
o
sodium phosphate buffer (Scheme 5) at 22 C. Under these conditions, the yield of 2d-αR1P
1
was ~49% after 6 h. The H NMR spectrum of the product mixture showed the presence of 2’-
7
deoxy-N -inosine (~29%), which presumably formed by reversal of the PNPase reaction at
10

ACCEPTED MANUSCRIPT
27
relatively low P concentrations (Figure S16-A).
When the reaction time was increased,
i
significant hydrolysis of 2d-αR1P was observed, giving 2-deoxy-D-ribose as the major product
(Figure S16-C).
o
Phosphorolysis reactions conducted with dI at 4 C for 28 days, however, showed good
7
conversion to 2d-αR1P (~60% yield based in dI) and 2’-deoxy-N -inosine (~40% yield), with
29
little unreacted dI detected and minimal hydrolysis to 2-deoxy-
D
-ribose (Figure S16-B).
At 4
o
30
C, XO activity was reduced significantly, HX accumulated in the reaction solution, and back
7
31
reactions were more favored, the latter responsible for the formation of 2’-deoxy-N -inosine.
o
The conversion of HX to UA was slow at 4 C, but the continuous removal of HX promotes the
7
formation of 2d-αR1P, as indicated by the lower percentage of 2’-deoxy-N -inosine (~25%) and
higher percentage of 2d-αR1P (~71%) found in solutions after 42 days of reaction. Very little 2-
deoxy-
D-ribose (~4%) was detected after 42 days, showing that 2d-αR1P is relatively stable in
o
10 mM P buffer at pH 7.4 and 4 C. Low temperature and 10 mM P buffer at pH 7.4 favor the
i
i
7
production of 2’-deoxy-N -inosine from dI, and thus not the optimal condition to prepare 2d-
αR1P.
2.4. N-Glycosylation Reactions to Produce 2’-Deoxyribonucleosides from 2d-
αR1P.
Purified 2d-αR1P was converted into 2’-deoxyadenosine (dA) and 2’-deoxyuridine (dU) in 90:10
1
2
v/v H O/ H O solvent at pH 6.9 containing either PNPase or UPase, respectively. These
2
2
o
o
reactions were conducted for 18 h at 22 C, followed by 4 h at 37 C. Yields of dA and dU were
21,32
~60% and ~52%, respectively (Figures S17–S18).
As found for the ribo series, 2’-
deoxycytidine (dC) could not be prepared because cytosine is not a UPase substrate (Figure
25
S19), but dU can be converted to dC in good yield by chemical methods. 2’-Deoxyguanosine
(dG) was obtained in low yield (<5%) using PNPase even when DMSO was added to the
32
reaction mixture (Figure S20). Use of thymidine phosphorylase (TPase, E.C. 2.4.2.4) in N-
21b,33
glycosylation of 2d-αR1P with thymine gave thymidine in a yield of ~60% (Figure S21).
In
N-glycosylation reactions to prepare 2’-deoxyribonucleosides from 2d-αR1P, significant
hydrolysis of 2d-αR1P to give 2-deoxy-
D
-ribose was observed, in contrast to behavior found for
11

ACCEPTED MANUSCRIPT
the N-glycosylation of αR1P. No attempt was made to isolate 2’-deoxyribonucleosides from
these reactions in this work, although the nucleoside products should be separable from 2-
deoxy-D-ribose on a Bio-Gel P-2 column, based on results obtained for the purification of [1’-
13
C]adenosine.
3. Conclusions
Milligram quantities of α-D-ribofuranosyl 1-phosphate (sodium salt) (αR1P) have been
prepared and purified from the enzyme-catalyzed phosphorolysis of inosine. αR1P is stable in
o
dilute aqueous solutions at pH 6.4 and 4 C for at least three months. Purified αR1P was N-
glycosylated to give adenosine and uridine in >85% yields and guanosine in ~40% yield (Table
3). Extension of the method to prepare 2d-αR1P from 2’-deoxyinosine showed the former was
unstable
in
Table 3. Optimized Reaction Conditions to Prepare αR1P and 2d-αR1P
by Nucleoside Phosphorolysis, and to Prepare Ribonucleosides and 2’-
Deoxyribonucleosides by Glycosyl Phosphate N-Glycosylation
phosphorolysis reaction
mixtures at pH 7.4 and
o
22 C, and that 2’-deoxy-
7
N -inosine formed in
reactions run at relatively
low concentrations of
phosphate buffer. Under
optimal phosphorolysis
conditions,
2d-αR1P
a
b
[αR1P] = 4 mM; pH 6.9; 37 °C; base in >8 fold molar e xcess. [2d-
αR1P] = 4 mM; pH 6.9; 22 °C for 18 h and then at 37 °C for 4 h; base in
could be isolated in pure
form in ~20 % yield.
c
1
>8 fold molar excess. Reaction yield determined by H NMR.
Although not detailed in this report, follow-up work showed that 2’-deoxyinosine phosphorolysis
o
at pH 8.5 and 22 C for reaction times up to 18 h gave very high yields of 2d-αR1P (> 95%) with
very little degradation to 2-deoxy-D-ribose during the course of the reaction, demonstrating that
PNPase, XO and catalase possess sufficient catalytic activity at this elevated pH to effect
12

ACCEPTED MANUSCRIPT
phosphorolysis. We also found that degradation (hydrolysis) of 2d-αR1P on Bio-Gel P-2
chromatographic columns was very low if the column was eluted with distilled water adjusted to
a pH of 8–8.5. Isolated 2d-αR1P was N-glycosylated to give 2’-deoxyadenosine, 2’-
deoxyuridine and thymidine in 50–60% yields (Table 3).
Neither cytidine nor 2’-deoxycytidine could be prepared by N-glycosylation of αR1P or 2d-
αR1P, respectively, but uridine and dU obtained by N-glycosylation can be converted to cytidine
25
and 2’-deoxycytidine, respectively, via chemical methods.
13
13
Using optimized reaction conditions, α-
D
-[1- C]R1P prepared from [1’- C]inosine was
13
13
converted to [1’- C]adenosine on a 27 mg scale. [1’- C]Adenosine was obtained in 94% yield
and in >99% purity after chromatography on a Bio-Gel P-2 column, demonstrating the feasibility
and practicality of the method for the synthesis of isotopically labeled nucleosides.
One-pot multi-enzyme syntheses of ribonucleosides and 2’-deoxyribonucleosides have
7,34
been reported previously,
but these methods give nucleoside mixtures that require
purification, rendering the overall process time-consuming. By isolating αR1P or 2d-αR1P as
described herein, nucleoside product purification after N-glycosylation is simplified significantly.
The method described herein
allows the incorporation of stable
isotopes into the furanose
constituent of nucleosides using I
or dI, both of which can be
Scheme 6. Current synthetic routes used to prepare NMPs containing
selective stable isotopic enrichment. (A) Williamson route.7d,36 (B)
Truncated Williamson route.1d (C) Route from this study. For routes
A–C, the total number of enzymes required for the process, and the
number that are commercially available, are indicated. Reactions
marked by an asterisk are ATP-requiring, and that marked by two
asterisks is a NAPDH-requiring reaction (dehydrogenase).
prepared in good yield by
chemical
glycosylation
of
appropriately
labeled
35
pentoses.
When only base
labeling is desired, unlabeled αR1P or 2d-αR1P, generated by phosphorolysis of unlabeled I or
dI, respectively, is N-glycosylated with an unprotected labeled base. Isotopic labeling in both
the base and furanose constituents is achieved using furanose-labeled I or dI and labeled
13

ACCEPTED MANUSCRIPT
nitrogen bases. While isotopically labeled bases are used in molar excess during N-
glycosylation, excess base can easily be recovered at the end of the reaction and recycled.
A comparison of the labeling methodology described herein with competing methods
7d,36
described previously is shown in Scheme 6. The Williamson method
isotopically labeled
coenzymes to prepare labeled ribo NMPs. The route is based on the pentose phosphate
, shown in A, uses
D
-glucoses as the labeled precursor, and requires 14 enzymes and several
pathway in which
D
-glucose 6P (G6P) is oxidized to 6-phospho-D-gluconate (6PG) by G6P
+
dehydrogenase, with concomitant reduction of NADP to NADPH. The NADPH is recycled
through the reduction of α-ketoglutarate to
L
-glutamate by glutamate dehydrogenase. Since
-ribose 5P (R5P) to α- -phosphoribosyl
ATP is required to convert -glucose to G6P and
D
D
D
pyrophosphate (PRPP), ATP regeneration is accomplished with the enzymes, creatine kinase
37
and myokinase, and with 2’-deoxyATP serving as an ATP surrogate and creatine phosphate
as the primary phosphate donor. Since C1 of the
D-glucose precursor is lost during its
conversion to
D-ribulose 5P (Ru5P), only
D-glucose labeled at C2–C6 are viable substrates,
38,6
despite the fact that labeling at C1 of
D-glucose is the easiest and least costly to perform.
The enzymes needed to convert PRPP to ribo NMPs (adenine phosphoribosyltransferase, uracil
phosphoribosyltransferase; xanthine-guanine phosphoribosyltransferase) are also not
commercially available. CTP synthetase, needed to convert UTP to CTP since the latter cannot
be prepared directly from PRPP and cytidine, is also not commercially available. Route A is
13
2
attractive for uniform labeling in the furanose ring of NMPs, since uniformly C- and/or H-
labeled -glucoses are commercially available. Applying route A to prepare selectively labeled
D
ribo NMPs is less attractive, since selectively labeled
D
-riboses can be prepared almost as
39
easily as selectively labeled D-glucoses.
1d
The relatively complex route shown in A (Scheme 6) has been truncated recently by
recognizing that
D-ribose is a more practical and cost-effective labeled precursor than D-glucose
for the selective labeling of NMPs (route B, Scheme 6). While it has been argued that labeled
D-riboses have only recently become available commercially for this purpose, routes to prepare
14

ACCEPTED MANUSCRIPT
13
39
virtually all C isotopomers of
D-ribose were described >20 years ago, and many ribose
40
isotopomers have been commercially available since that time.
Like route A, an ATP
-ribose to -R5P, and -R5P to
regeneration system is needed because two conversions (
D
D
D
37
PRPP) require ATP, and the same dATP system used in route A has been applied. Route B
requires six enzymes and two cofactors (dATP and creatine phosphate).
Five of the six
enzymes (ribokinase, adenine phosphoribosyltransferase, uracil phosphoribosyltransferase,
xanthine-guanine phosphoribosyltransferase, and CTP synthetase) are not commercially
available (only phosphoribosylpyrophosphate synthetase is commercially available). Thus,
while route (B) is a significant improvement over route A, most of the enzymes required to
implement it are not commercially available, thus reducing its attractiveness.
The routes shown in C (Scheme 6) are described in this report. An important distinction
between routes C and routes A–B is that both ribo- and 2’-deoxyribonucleosides can be
prepared by the former, whereas the latter apply only to ribonucleotide synthesis. Focusing on
the ribo route in C, four enzymes (PNPase, uridine phosphorylase, xanthine oxidase, catalase)
are required to prepare αR1P and to convert it into three ribonucleosides (A, G and U), and all
are commercially available. Cytidine can be prepared chemically in good yield from uridine
25
(analogous to the use of CMP synthetase in routes A–B). Since route C gives nucleosides as
end-products, the latter can be converted directly to phosphoramidite derivatives for use in the
chemical synthesis of RNA. Alternatively, if NMPs are needed, reliable chemical routes to
41
prepare them from the four ribonucleosides are available, although nucleoside kinases could
be also used for this purpose; to our knowledge, however, these kinases are not commercially
1d,36,42
available. The NMPs can then be converted enzymatically to NTPs
for use in the
1d,36,42,43
enzymatic synthesis of RNA.
While route B starts with a simpler labeled precursor
(ribose) than route C (inosine), this advantage is offset by need to express multiple enzymes
and by the relative ease of chemical synthesis of labeled inosine from labeled -ribose. Finally,
D
route B was developed as a one-pot synthesis to give NMPs directly, whereas route C allows
15

ACCEPTED MANUSCRIPT
the isolation of a key intermediate, αR1P, which can be stored for future nucleoside syntheses
or used for other purposes.
The above analysis demonstrates that routes B and C are more attractive than route A
for the preparation of selectively labeled ribo NMPs, owing to their greater simplicity. The
relative merits of routes B and C are likely to change over time as more enzymes become
commercially available (assuming their costs and shelf-lives are reasonable); the main current
limitation of route B is the fact that five of the six required enzymes cannot be obtained
conveniently. This problem is overcome in route C by investing in a more complex labeled
starting material, prepared by chemical methods, and by employing chemical methods to
convert
the
product
to
ribonucleosides
ribonucleoside
Scheme 7. Anticipated modifications to route C shown in Scheme 6. In
which αR1P is produced directly from -ribose via the use of ribokinase
D
(RK) and phosphopentomutase (PPM). Conversion of αR1P to
ribonucleosides involves the same enzymes in Scheme 6, and
conversion of ribonucleosides to NMPs is achieved either enzymatically
(nucleoside kinases) or chemically (e.g., POCl3) (see text). The same
route can also be used to prepare 2'-deoxyribonucleosides, since RK
5’-
monophosphates, although over
time enzymes may become
available commercially to effect
and PPM accept 2-deoxy-
respectively, as substrates.
D-ribose and 2'-deoxy-D-ribose 5P,
the latter conversions. From a commercial viewpoint, being able to isolate, purify and store the
key intermediate, αR1P, adds additional appeal to route C by allowing a given αR1P isotopomer
to be combined with different nitrogen base isotopomers as needs dictate. Unlike route B, route
C can also be used to prepare labeled 2’-deoxyribonucleosides, adding to its versatility.
Future modifications of route C are also anticipated in which inosine is eliminated as the
source of αR1P (Scheme 7). In this scenario,
D-ribose is converted to
D-ribose 5P with
44
ribokinase, and the latter isomerized to αR1P by phosphopentomutase.
While the mutase
reaction is not likely to favor the glycosyl phosphate, the reaction would be driven to completion
by coupling it to the nucleoside phosphorylase reaction. In practice, the overall conversion of
D
-
ribose to ribonucleosides would be achieved in a one-pot reaction (e.g., the αR1P would not be
45
isolated). Since ribokinase can phosphorylate 2-deoxy-D-ribose to give 2-deoxy-D-R5P, and
16

ACCEPTED MANUSCRIPT
25,44
PPM can isomerize 2-deoxy-
D
-R5P to give 2d-αR1P,
the route in Scheme 7 could also be
used to prepare 2’-deoxyribonucleosides and their phosphates.
4. Experimental
1
13
1
4.1. NMR Spectroscopy. High-resolution H and C{ H} NMR spectra were obtained
o
at 22 C using 5-mm NMR tubes on a 600-MHz FT-NMR spectrometer equipped with a 5-mm
1
19 15 31
1
H- F/ N- P dual broadband probe.
H NMR spectra were collected with a ~6000 Hz
spectral window and ~4 s recycle time, and FIDs were zero-filled to give final digital resolutions
13
1
of <0.025 Hz/pt. C{ H} NMR spectra were collected with a ~25k Hz spectral window and ~3.6
s recycle time, and FIDs were zero-filled to give final digital resolutions of <0.1 Hz/pt. FIDs were
processed with resolution enhancement (Gaussian or sine-bell functions) to improve resolution
and facilitate the measurement of small J-couplings (≥ 0.5 Hz), and reported J-couplings are
accurate to 0.1 Hz, unless otherwise stated.
±
4.2. Spectrophotometry. A double-beam spectrophotometer was used to monitor the
o
phosphorolysis reactions at 22 C. Pure water or sodium phosphate buffer at same
concentration as that in the reaction solution was used as the blank. The UV absorption was
measured at 248 nm for hypoxanthine, at 254 nm for adenosine, and at 290 nm for uric acid.
4.3. Mass Spectrometry. Mass spectral data were obtained on a micrOTOF II mass
spectrometer (Bruker) with direct infusion in CH CN:H O (9:1). An electrospray ionization
3
2
probe (ESI) was used with a spray voltage of 3.8 kV and sheath gas flow rate of 4.0 L/min (0.8
o
Bar). The capillary temperature was set at 180 C and the ion polarity was set as negative. Full
MS scans were collected over a range of 50–1650 m/z.
4.4. Test Reaction to Convert Inosine to
α-D-Ribofuranosyl 1-Phosphate Sodium Salt
(
α
R1P). Inosine (1.5 mg, 5.5 µmol) and PNPase (0.2 units) were added to 100 mL of 10 mM
o
sodium phosphate buffer at pH 7.4. The reaction mixture was incubated at 37 C and the
4
-1
-1 13
production of hypoxanthine (ε248 = 1.08×10 M cm ) was monitored spectrophotometrically
by measuring the UV absorption of three samples at each time point. The reaction reached
17

ACCEPTED MANUSCRIPT
equilibrium after ~3 min. To another 100 mL of 10 mM sodium phosphate buffer at pH 7.4 were
added inosine (1.5 mg, 5.5 µmol), PNPase (0.2 units), xanthine oxidase (2 units) and catalase
o
(1000 units). The reaction mixture was incubated at 37 C and the production of uric acid (ε290
=
4
-1
-1 14
1.32×10 M cm ) was monitored spectrophotometrically by measuring the UV absorption of
three samples at each time point. The reaction was complete after 11 min.
4.5. Scale-up Reaction to Isolate
α-D-Ribofuranosyl 1-Phosphate Sodium Salt (αR1P)
from Inosine Phosphorolysis. Inosine (67.0 mg, 250 µmol), PNPase (8 units), XO (0.4 units),
catalase (2,000 units) and a few drops of CHCl (to inhibit bacterial growth) were added to
3
sodium phosphate buffer (50 mM, 500 mL) at pH 7.4, and the reaction mixture was incubated at
o
o
37 C. After two days, the reaction mixture was concentrated to ~10 mL at 30 C in vacuo and
o
most of the sodium phosphate was removed by crystallization at 4 C. The sample (~3 mL) was
loaded on a Bio-Gel P-2 column (2.5 cm × 100 cm) and eluted at ~0.8 mL/min with DI water,
1
and fractions (~15 mL) were analyzed by H NMR. Fractions containing αR1P (fraction 13–21)
were combined and reloaded on a second Bio-Gel P-2 column of similar dimensions, and
elution gave a pure sample of αR1P for further characterization. An overall isolated yield of ~74%
1
1
13
was determined by H NMR. See Tables 1 and 2 for H NMR (600 MHz) and C NMR (150
-
MHz) data. HRMS: [M - H] calcd for C H O P 229.0119, found 229.0122.
5
10 8
4.6. Conversion of
αR1P to Adenosine. Adenine (10 mg, 74 µmol) was added to a
reaction mixture (1.0 mL) containing αR1P (4 mM) and PNPase (0.2 units) dissolved in 90:10
1
2
v/v H O: H O at pH 6.9. Some of the adenine remained undissolved in the resulting solution.
2
2
One drop of CHCl (~40 ꢁL) was added to the mixture to inhibit bacterial growth, and reaction
3
o
1
21
mixture was incubated at 37 C for 20 h and assayed by H NMR (Figure 4).
4.7. Conversion of R1P to Guanosine. PNPase (0.4 units) and guanine (10 mg, 66
µmol) were added to 1.0 mL of a 80:10:10 v/v/v H O: H O:DMSO-d solution containing αR1P
α
1
2
2
2
6
(4 mM) at pH 6.8. One drop of CHCl (~40
ꢁL) was added, and the reaction mixture was
3
1
incubated at 37 °C for 16 h. Guanosine (~10% yield ) was observed by H NMR. After 9 days of
21b
reaction, ~40 % of the αR1P was converted to guanosine (Figure S9).
18

ACCEPTED MANUSCRIPT
4.8. Conversion of
α
R1P to Uridine. Uracil (1.8 mg, 16 µmol) and one drop of CHCl
3
1
2
(~40 ꢁL) were added to 1.0 mL of a 90:10 v/v H O: H O solution containing αR1P (4 mM) and
2 2
o
UPase (5.0 ꢁg) at pH 6.9. The reaction mixture was incubated at 37 C for 2.5 h, after which
1
time H NMR indicated that ~75% of the αR1P was converted into uridine (Figure S10A).
o
Additional uracil (1.8 mg, 16 µmol) was added and the reaction mixture was incubated at 37 C
1
for an additional 18 h, after which time H NMR showed that 85% of the αR1P was converted
21b
into uridine (Figure S10B).
4.9. Conversion of
αR1P to Cytidine. Cytosine (1.8 mg, 16 µmol) and one drop of
1
2
CHCl (~40 L) were added to 1.0 mL of a 90:10 v/v H O: H O solution containing αR1P (4
ꢁ
2
2
3
o
mM) and UPase (15 ꢁg) at pH 6.9. The reaction mixture was incubated at 37 C for 20 h and
1
assayed by H NMR (Figure S11). No cytidine was detected in the reaction mixture.
13
4.10. Scaled-up Reaction to Prepare and Isolate [1’- C]Adenosine. Adenine (200 mg,
13
1.48 mmol) was added to a reaction mixture (30 mL) containing α-D-[1- C]R1P (107 µmol) and
PNPase (4 units) dissolved in DI water at pH 7.7. Some of the adenine remained undissolved in
the resulting reaction mixture. One drop of CHCl (~40 L) was added to the solution to inhibit
ꢁ
3
o
1
bacterial growth. The reaction mixture was incubated at 37 C and monitored by H NMR.
13
13
After three days of reaction, >95% of α-D-[1- C]R1P was converted to [1’- C]adenosine. The
remaining undissolved adenine was removed by centrifugation, and the supernatant was
o
concentrated to ~4 mL at 30 C in vacuo. Additional solid adenine formed during concentration,
and was separated from the aqueous solution by centrifugation; the solid was washed twice with
water (1.5 mL × 2). The aqueous solution and washes were combined and loaded on a Bio-Gel
P-2 column (2.5 cm × 100 cm), which was eluted at ~0.8 mL/min with DI water. Fractions (~15
mL) were analyzed at 254 nm using a double-beam spectrophotometer. Fractions containing
13
o
[1’- C]adenosine (fractions 43–49) were combined and concentrated at 30 C in vacuo, giving
1
the product as a white solid (27 mg, 101 µmol, 94%). H NMR showed the product 1’-
13
C]adenosine to be > 99% pure (Figure 5). Unreacted adenine eluted after adenosine
1
21b
(fractions 52–57), whose structure was confirmed by H NMR.
19

ACCEPTED MANUSCRIPT
4.11. Isolation and Purification of 2-Deoxy-
α
-
D-ribofuranosyl 1-Phosphate Sodium Salt
(2d-
α
R1P) from 2’-Deoxyinosine Phosphorolysis.
2’-Deoxyinosine (63 mg, 0.25 mmol),
PNPase (20 units), XO (1.0 unit), catalase (2,000 units) and one drop of CHCl were added to
3
sodium phosphate buffer (50 mM, pH 7.4, 500 mL), and the reaction mixture was incubated at
o
4
-1
-1
22 C.
The production of uric acid (ε290 = 1.32×10 M cm ) was monitored
o
spectrophotometrically. After incubation at 22 C for 20 h, the reaction solution was
o
concentrated at 30 C in vacuo and most of the sodium phosphate was removed by
1
o
26
crystallization at 4 C. The presence of 2d-αR1P was confirmed by H NMR.
The reaction
solution (~5 mL) was loaded on a Bio-Gel P-2 column (2.5 cm × 100 cm) and eluted at ~0.8
o
o
mL/min with DI water at 4 C. Fractions (~15 mL) were collected, concentrated at 30 C and
1
analyzed by H NMR to afford 2d-αR1P in ~85% purity (Figure S15). An overall isolated yield of
1
~20% was determined by H NMR. Additional column fractions contained 2d-αR1P and 2-
deoxy-
yield.
D-ribose, as well as P ; these fractions were not processed further to improve the isolated
i
In a second test reaction, 2’-deoxyinosine (10.0 mg, 0.04 mmol), PNPase (4 units), XO
(0.5 units), catalase (1,000 units) and one drop of CHCl were added to sodium phosphate
3
o
buffer (10 mM, pH 7.4, 80 mL) and the mixture was incubated at 22 C for 6 h. A portion of
o
reaction mixture (40 mL) was held at 22 C for an additional 28 d, and another portion (40 mL)
o
1
was held at 4 C for an additional 42 d. During these incubations, H NMR spectra were
collected periodically to monitor the reaction progress. In addition to 2d-αR1P and 2-deoxy-
D
-
7
1
29
ribose, 2’-deoxy-N -inosine was detected in these reaction mixtures by H NMR.
4.12. Conversion of 2d- R1P to 2’-Deoxyribonucleosides. Nucleobases (adenine,
guanine, cytosine, uracil, thymine) were added in molar excess to 1.0 mL of 90:10 v/v
α
1
2
H O: H O solutions containing 2d-αR1P (4 mM) and 0.2 units of PNPase (for adenine and
2
2
guanine), UPase (for cytosine and uracil) or TPase (for thymine), one drop of CHCl (~40
ꢁL)
3
was added to each reaction mixture, and the reaction mixtures at pH 6.9 were incubated at 22
o
o
C for 18 h followed by 37 C for 4 h. Adenine (10 mg, 74 µmol) and guanine (10 mg, 66 µmol)
20

ACCEPTED MANUSCRIPT
remained largely as solids in these reaction mixtures. The guanine reaction mixture also
contained 10% DMSO to increase the concentration of the nitrogen base in solution. Uracil (3.6
mg, 32 µmol), cytosine (3.6 mg, 32 µmol) and thymine (3.9 mg, 32 µmol) were soluble under
1
these reaction conditions. The reactions were monitored by H NMR (Figure S17–21) and
1
reaction products were identified by comparison of the observed H chemical shifts to those of
21,32
authentic 2’-deoxyribonucleosides.
Acknowledgements
This work was supported by a grant from the National Science Foundation to A.S. (CHE
13
1402744). We thank Omicron Biochemicals, Inc. for a gift of [1’- C]inosine.
Supplementary Information
1
13
1
Description of the method used to determine αR1P concentrations in solution; H, C{ H} and
31
1
P{ H} NMR spectra of reaction mixtures and purified products; structures of intermediates and
by-products; summary schematic of enzymatic conversions of αR1P to adenosine, guanosine
and uridine.
Notes
A.S. is the president and CEO of Omicron Biochemicals, Inc..
References
1. (a) Kline, P. C.; Serianni, A. S. J. Am. Chem. Soc., 1990, 112, 7373–7381. (b) Batey, R. T.;
Inada, M.; Kujawinski, E.; Puglisi, J. D.; Williamson, J. R. Nucleic Acids Res., 1992, 20,
4515–4523. (c) Bandyopadhyay, T.; Wu, J.; Serianni, A. S. J. Org. Chem., 1993, 58, 5513–
5517. (d) Longhini, A. P.; LeBlanc, R. M.; Becette, O.; Salguero, C.; Wunderlich, C. H.;
Johnson, B. A.; D'Souza, V. M.; Kreutz C.; Dayie T. K. Nucleic Acids Res., 2016, 44, e52.
21

ACCEPTED MANUSCRIPT
(e) Alvarado, L. J.; Longhini, A. P.; LeBlanc, R. M.; Chen, B.; Kreutz C.; Dayie, T. K.
Methods Enzymol., 2014, 549, 133-162
2. (a) Kellenbach, E.; Remerowski, M.; Eib, D.; Boelens, R.; Van der Marel, G.; Van den Elst,
H.; Van Boom, J.; Kaptein, R. Nucleic Acids Res., 1992, 20, 653–657. (b) Zimmer, D. P.;
Crothers, D. M. Proc. Natl. Acad. Sci., 1995, 92, 3091–3095. (c) Louis, J. M.; Martin, R. G.;
Clore, G. M.; Gronenborn, A. M. J. Biol. Chem., 1998, 273, 2374–2378.
3. (a) Behrens, P. W.; Sicotte, V. J.; Delente, J. J. Appl. Phycol., 1994, 6, 113–121. (b)
Nikonowicz, E. P.; Sirr, A.; Legault, P.; Jucker, F. M.; Baer, L. M.; Pardi, A. Nucleic Acids
Res., 1992, 20, 4507–4513. (c) Batey, R. T.; Battiste, J. L.; Williamson, J. R. Methods
Enzymol., 1995, 261, 300–322.
4. (a) Shirae, H.; Yokozeki, K.; Kubota, K. Agric. Biol. Chem., 1988, 52, 1499–1504. (b)
Hennen, W. J.; Wong, C. H. J. Org. Chem., 1989, 54, 4692–4695. (c) Krenitsky, T. A.;
Koszalka, G. W; Tuttle, J. V. Biochemistry, 1981, 20, 3615–3621. (d) Kline, P. C.; Schramm,
V. L. Biochemistry 1993, 32, 13212–13219. (e) Parks, R. E.; Agarwal, R. P. The Enzymes,
1972, 7, 483–514. (f) Serra, I.; Ubiali, D.;Piškur, J.; Christoffersen, S.; Lewkowicz, E. S.;
Iribarren, A. M.; Albertini, A. M.; Terreni, M. ChemPlusChem, 2013, 78, 157–65.
5. (a) Van Rompay, A. R.; Johansson, M.; Karlsson, A. Pharmacol. Ther., 2000, 87, 189–198.
(b) Yoshikawa, M.; Kato, T.; Takenishi, T. Tetrahedron lett., 1967, 8, 5065–5068.
6. Zhang, W.; Zhao, S.; Serianni, A. S. Methods Enzymol., 2015, 565, 423–458.
7. (a) Li, Y.; Ding, Q.; Ou, L.; Qian, Y.; Zhang, J. Biotechnol. Bioprocess Eng., 2015, 20, 37–43.
(b) J. Fernández-Lucas, Appl. Microbiol. Biotechnol., 2015, 99, 4615–4627. (c) Arevalo, R.
M.; Deroncelé, V. M.; Gómez, C. L.; Gilabert, M. P.; Company, C. E.; Boliart, J. C. US. Pat.,
20120264175A1, 2012. (d) Scott, L. G.; Tolbert, T. J.; Williamson, J. R. Methods Enzymol.,
2000, 317, 18-38.
8. Tozzi, M. G.; Camici, M.; Mascia, L.; Sgarrella, F.; Ipata, P. L. FEBS J., 2006, 273, 1089–
1101.
9. Vorbrüggen, H.; Ruh-Pohlenz, C. Synthesis of nucleosides; Wiley Online Library, 2000.
22

ACCEPTED MANUSCRIPT
10. Kawasaki, A. M.; Wotring, L. L.; Townsend, L. B. J. Med. Chem., 1990, 33, 3170–3176.
11. Pal, S.; Nair, V. Biocatal. Biotransform., 1997, 15, 147–158.
12. Gardiner, D. G. Anal. Biochem., 1979, 95, 377–382.
13. Jones, N. R. Analyst, 1960, 85, 111–115.
14. Shintani, H. Pharm. Anal. Acta., 2013, S7:004.
15. (a) Zhang, W.; Carmichael, I.; Serianni, A. S. J. Org. Chem., 2011, 76, 8151–8158. (b)
Zhang, W.; Serianni, A. S. J. Am. Chem. Soc., 2012, 134, 11511–11524.
16. Malz, F.; Jancke, H. J. Pharm. Biomed. Anal., 2005, 38, 813–823.
17. Gorin, P. A.; Mazurek, M. Carbohydr. Res., 1976, 48, 171–186.
18. Harris, R. K.; Becker, E. D.; Cabral de Menezes, S. M.; Goodfellow, R.; Granger, P. Pure
Appl. Chem., 2001, 73,1795–1818.
19. Serianni, A. S.; Barker, R. Can. J. Chem., 1979, 57, 3160–3167.
20. (a) Wu, G. D.; Serianni, A. S.; Barker, R. J. Org. Chem., 1983, 48, 1750–1757. (b) Wu, J.;
Bondo, P. B.; Vuorinen, T.; Serianni, A. S. J. Am. Chem. Soc., 1992, 114, 3499–3505.
21. (a) Ciuffreda, P.; Casati, S.; Manzocchi, A. Magn. Reson. Chem., 2007, 45, 781–784. (b)
Biological Magnetic Resonance Data Bank, http://bmrb.wisc.edu/.
22. Calleri, E.; Cattaneo, G.; Rabuffetti, M.; Serra, I.; Bavaro, T.; Massolini, G.; Ubiali, D. Adv.
Synth. Catal., 2015, 357, 2520–2528.
23. Robins, M. J.; Zou, R.; Guo, Z.; Wnuk, S. F. J. Org. Chem., 1996, 61, 9207–9212.
24. (a) Saunders, P. P.; Wilson, B. A.; Saunders, G. F. J. Biol. Chem., 1969, 244, 3691–3697.
(b) Ling, F.; Inoue, Y.; Kimura, A. Appl. Environ. Microb., 1990, 56, 3830–3834.
25. Ouwerkerk, N.; Steenweg, M.; De Ruijter, M.; Brouwer, J.; van Boom, J. H.; Lugtenburg, J.;
Raap, J. J. Org. Chem., 2002, 67, 1480–1489.
26. (a) Khym, J. X. Ana. Biochem., 1974, 58, 638–641. (b) Uziel, M.; Cohn, W. E. Nucleic
Acids Protein Syn., 1965, 103, 539–541.
23

ACCEPTED MANUSCRIPT
27. (a) Church, T. J.; Carmichael, I.; Serianni, A.S. J. Am. Chem. Soc., 1997,119, 8946–8964.
(b) Bandyopadhyay, T.; Wu, J.; Stripe, W. A.; Carmichael I.; Serianni A. S. J. Am. Chem.
Soc., 1997,119, 1737–1744.
28. (a) Komatsu, H.; Awano, H. J. Org. Chem., 2002, 67, 5419–5421. (b) Hatano, A.; Harano,
A.; Takigawa, Y.; Naramoto, Y.; Toda, K.; Nakagomi, Y.; Yamada, H. Bioorg. Med. Chem.,
2008, 16, 3866–3870.
29. (a) Seela, F.; Leonard, P. HeIv. Chim. Acta, 1998, 81, 2244–2263. (b) Marfurt, J.; Parel, S.
P.; Leumann, C. J. Nucleic Acids Res., 1997, 25, 1875–1882. (c) Robins, M. J.; Zou, R.;
Guo, Z.; Wnuk, S. F. J. Org. Chem., 1996, 61, 9207–9212.
30. Mondal, M. S.; Mitra, S. Biochemistry, 1994, 33, 10305–10312.
31. (a) Ropp, P. A.; Traut, T. W. J. Biol. Chem., 1991, 266, 7682–7687. (b) Zhou, X.; Szeker,
K.; Janocha, B.; Böhme, T.; Albrecht, D.; Mikhailopulo, I. A.; Neubauer, P. FEBS J., 2013,
280, 1475–1490.
32. Kline, P. C.; Serianni, A. S. Magn. Reson. Chem., 1990, 28, 324–330.
33. Lemieux, R. U. Can. J. Chem., 1961, 39, 116–120.
34. (a) Lazrek, H. B.; Ouzebla, D.; Rochdi, A.; Redwane, N.; Vasseur, J. J. Lett. Org. Chem.,
2006, 3, 313–314. (b) Dudycz, L. W.; Wright, G. E. Nucleos. Nucleot., 1984, 3, 33–44.
35. (a) Recondo, E. F.; Rinderknecht, G. L. Helv. Chim. Acta, 1959, 42, 1171–1173. (b)
Vorbrüggen, H.; Krolikiewicz, K.; Bennua, B. Chem. Ber., 1981, 114, 1234–1255.
36. Tolbert, T. J.; Williamson, J. R. J. Am. Chem. Soc. 1997, 119, 12100–12108.
37. Wennefors, C. K.; Dobrikov, M. I.; Xu, Z.; Li, P.; Shaw, B. R. Bioorg. Chem. 2008, 36, 169–
177.
38. Serianni, A. S.; Vuorinen, T.; Bondo, P. B. J. Carbohydr. Chem. 1990, 9, 513–541.
39. Serianni, A. S.; Bondo, P. B. J. Biomol. Struct. Dyn. 1994, 11, 1133–1148.
40. Omicron Biochemicals, Inc., http://www.omicronbio.com/.
24

ACCEPTED MANUSCRIPT
41. (a) Yoshikawa, M.; Kato, T.; Takenishi, T.; Tetrahedron Lett. 1967, 8, 5065–5068. (b)
Waehneldt, T. V.; Fox, S. W. Biochim. Biophys. Acta, Nucleic Acids Protein Synth. 1967,
134, 1–8. (c) Taktakishvili, M.; Nair, V. Tetrahedron Lett. 2000, 41, 7173–7176.
42. Lu, K.; Miyazaki, Y.; Summers, M. F. J. Biomol. NMR. 2010, 46, 113–125.
43. Alvarado, L. J.; LeBlanc, R. M.; Longhini, A. P.; Keane, S. C.; Jain, N.; Yildiz, Z. F.; Tolbert,
B. S.; D'Souza, V. M.; Summers, M. F.; Kreutz, C.; Dayie, T. K. ChemBioChem. 2014, 15,
1573–1577.
44. Moustafa, H. M.; Zaghloul, T. I.; Zhang, Y. P. Anal. Chem. 2016, 501, 75–81.
45. Chuvikovsky, D. V.; Esipov, R. S.; Skoblov, Y. S.; Chupova, L. A.; Muravyova, T. I.;
Miroshnikov, A. I.; Lapinjoki, S.; Mikhailopulo, I. A. Bioorg. Med. Chem. 2006, 14, 6327–
6332.
25

ACCEPTED MANUSCRIPT
Highlights
•
•
•
The α-
D
-ribofuranosyl 1-phosphate was prepared by enzymatic method, isolated
by chromatography and characterized by NMR.
The method to prepare isotopically labeled ribo- and 2’-deoxyribonucleosides
was significantly improved through isolating glycosyl phosphates.
[1’-13C]adenosine was prepared with high yield and high purity.