N. R. Rivera et al. / Tetrahedron Letters 57 (2016) 1090–1092
1091
O
P
NH2
NH2
N
NH2
BH3
1) R = H
2) R =
3) R =
N
N
N
N
N
N
N
O-
N
N
KRED
O
N
S
P
OH
O
P
OH
OH
H
N
4
) R =
OR
5
OH
1
OH
NADP
NADPH
O
OH
H
Figure 1. (R)-9-(2-Hydroxypropyl)adenine 1 and (R)-9-(2-hydroxypropyl)adenine
containing compounds.
Scheme 2. Enzymatic asymmetric reduction with KRED and NADP with a cofactor
recycling system using isopropanol.
ketoreductase (KRED) enzymes have been widely used in our lab-
oratories for the preparation of chiral alcohols.10 Furthermore,
KRED enzymes have been used to prepare deuterium labeled alco-
hols from ketones.11 We envisioned that a short and direct route to
SIL-[D6]-1 and [3H]-1 would be through a biocatalytic enantiose-
lective reduction of a ketone intermediate (Scheme 1).12 For SIL
synthesis (Path A), the acidic protons of ketone 5 would allow for
facile H/D exchange followed by an enzymatic reduction with an
appropriate deuterium source to give the desired SIL-[D6]-1. A sim-
ilar approach (Path B) would provide tritium labeled alcohol [3H]-1
via a direct enzymatic reduction of ketone 5 with an appropriate
tritium source.
Table 1
Representative KRED reduction enzyme screen results
Entry
Enzyme
% conversiona
% eeb
1
2
3
4
5
6
7
8
9
P1B02
P1B10
P1H10
P2C02
97.3
30
12.2
99.7
44.6
2.3
23.3
19.2
11.9
>99 (R)
97.7 (R)
>99 (R)
97.5 (S)
97.5 (R)
68.4 (R)
40.6 (R)
81.6 (R)
92.8 (S)
P2D11
KRED 101
KRED 119
KRED 130
KRED NADH 101
With this synthetic approach in mind, we first sought to iden-
tify conditions to reduce ketone 513 enzymatically. A number of
KRED enzymes14 were screened using NADP as a cofactor and
iPrOH to regenerate the cofactor (Scheme 2). The representative
results summarized in Table 1 show that either enantiomer can
be accessed in very high enantioselectivities and conversions
(entries 1 and 4). With the optimal enzyme identified (entry 1,
P1B02 enzyme), the reaction conditions were optimized with
respect to enzyme and cofactor loading, reaction concentration,
and amount of iPrOH required. Typically with these reactions, in
order to drive the reaction equilibrium in favor of the desired pro-
duct, an excess amount of iPrOH is used. Furthermore, the acetone
byproduct can be vented to drive the product formation.
In practice SIL compounds to be used as internal standards gen-
erally require a molecular weight increase of at least 4 atomic mass
units (amu). When the difference is less than 3 amu the isotope
peaks mass signal of the analyte may interfere with the signal of
the internal standard.15 In addition, as a general guideline, we typ-
ically require that the unlabeled content (D0) cannot be above 0.1%.
For this particular substrate, we targeted a mass increase of 6 amu.
With these specifications in mind, we began the synthesis of SIL-
[D6]-1 with hydrogen–deuterium exchange of the acidic protons
of ketone 5. After some optimization with regard to deuterated sol-
vents used, pH and temperature, we found that the exchange can
be accomplished by heating ketone 5 in a mixture of D2O/MeOD/
ACN at 50 °C to give ketone [D5]-5 in very high isotopic incorpora-
tion with [D5] as the most abundant isotope as shown in Table 2.16
a
Conversion, as area % of alcohol 1 relative to ketone 5, was determined by HPLC
analysis.
b
ee was determined by chiral HPLC.
Table 2
Isotopic distribution of ketone [D5]-5
m/z
Isotopic species
% distribution
192
193
194
195
196
197
198
D0
D1
D2
D3
D4
D5
D6
0.01
0.02
0.17
2.38
19.10
67.87
10.47
Some minor H/D exchange was observed at the C-8 position in the
adenine core resulting in [D6] isotope species with m/z = 198.
The final deuterium was introduced enzymatically using
[2-2H]-iPrOH as the deuterium source. We initially established that
a single deuterium transfer experiment using P1B02 enzyme with
unlabeled ketone 5 and [2-2H]-iPrOH gave the corresponding [D1]-
labeled alcohol with 100% deuterium incorporation as measured
by MS analysis. Thus, taking ketone [D5]-5 and subjecting it to
KRED reduction conditions gave the desired isotopically labeled
alcohol SIL-[D6]-1 in high enantiopurity (>99% ee) and yield
(90%). More importantly, the isotopic distribution is such that D0
is less than 0.1% as shown in Table 3, with [D6] as the most abun-
dant isotope. It is important to note that due to the propensity of
the (acidic) deuterium labels of ketone [D5]-5 to back exchange
to hydrogen, the enzymatic reduction was carried out in
NH2
NH2
N
N
N
enzyme
N
N
N
deuterium
source
N
N
Path A
D
D
OH
CD3
SIL-[D6]-1
D
O
NH2
D
D
CD3
N
N
N
H/D exchange
Table 3
[D5]-5
Isotopic distribution of alcohol SIL-[D6]-1
N
O
m/z
Isotopic species
% distribution
5
Path B
NH2
N
194
195
196
197
198
199
200
201
D0
D1
D2
D3
D4
D5
D6
D7
0.01
0.03
0.04
0.29
2.25
16.79
58.80
21.79
N
N
N
enzyme
tritium source
T
OH
[3H]-1
Scheme 1. Enzymatic synthesis of isotopically labeled (R)-9-(2-hydroxypropyl)
adenine.