Methylamine adenine dinucleotide (MAD): A co-factor to turn alcohol dehydrogenases into aldolases

Article abstract of DOI:10.1139/v02-074

Methylamine adenine dinucleotide (MAD) was prepared in seven steps and 15.3% overall yield from 1-acetoxy-2,3,5-tri-O-benzoyl-D-ribose by chemoenzymatic synthesis. The key step was the phosphate-phosphate coupling between adenosine monophosphate and 2,5-anhydro-1-deoxy-1-phenylacetamide-6-phosphate-D-allitol (3) mediated by carbonyl diimidazole (CDI), followed by the removal of the phenylacetamido group by penicillin G acylase. The MAD co-factor provided a primary amine functionality that was suitably positioned to promote the enamine aldolization of 2-oxopropionamide with aldehydes within the active site of alcohol dehydrogenases, which should lead to the transformation of these stereoselective enzymes into aldolases. Preliminary investigations did not reveal any activity with horse liver alcohol dehydrogenase or yeast alcohol dehydrogenase.

Full text of DOI:10.1139/v02-074

665
Methylamine adenine dinucleotide (MAD):
a co-factor to turn alcohol dehydrogenases
into aldolases
Denis Wahler and Jean-Louis Reymond
Abstract: Methylamine adenine dinucleotide (MAD) was prepared in seven steps and 15.3% overall yield from 1-
acetoxy-2,3,5-tri-O-benzoyl-D-ribose by chemoenzymatic synthesis. The key step was the phosphate–phosphate coupling
between adenosine monophosphate and 2,5-anhydro-1-deoxy-1-phenylacetamide-6-phosphate-D-allitol (3) mediated by
carbonyl diimidazole (CDI), followed by the removal of the phenylacetamido group by penicillin G acylase. The MAD
co-factor provided a primary amine functionality that was suitably positioned to promote the enamine aldolization of 2-
oxopropionamide with aldehydes within the active site of alcohol dehydrogenases, which should lead to the transforma-
tion of these stereoselective enzymes into aldolases. Preliminary investigations did not reveal any activity with horse
liver alcohol dehydrogenase or yeast alcohol dehydrogenase.
Key words: aldolases, alcohol dehydrogenases, co-factor enginering, nucleosides, acylases.
Résumé : On a réalisé une synthèse chimioenzymatique du dinucléotide de méthylamine et d’adénine (« MAD »), en
sept étapes et avec un rendement global de 15,3%, en utilisant le 1-acétoxy-2,3,5-tri-O-benzoyl-D-ribose comme produit
de départ. L’étape clé est le couplage phosphate–phosphate entre le monophosphate d’adénosine et le 2,5-anhydro-1-
désoxy-1-phénylacétamide-6-phosphate-D-allitol (3) catalysée par le carbonyldiimidazole (CDI), suivie de l’élimination
du groupe phénylacétamido par l’acylase de la pénicilline G. Le cofacteur MAD fournit une fonction amine primaire
positionnée de façon appropriée pour favoriser une aldolisation de l’énamine du 2-oxopropanamide avec les aldéhydes
dans le site actif des déshydrogénases de l’alcool, ce qui devrait conduire à une transformation de ces enzymes stéréo-
sélectifs en aldolases. Des études préliminaires indiquent qu’il n’y a pas d’activité avec la déshydrogénase du foie de
cheval ou la déshydrogénase d’alcool de la levure.
Mots clés : aldolases, déshydrogénases d’alcool, cofacteur d’ingénierie, nucléosides, acylases.
[Traduit par la Rédaction] Wahler and Reymond 670
Introduction
stereoselective aldol reactions. The work draws on the struc-
tural analogy between the transition states for hydride trans-
fer and aldol addition to carbonyl compounds.
Co-factor engineering is a purely chemical approach to
enzyme modification, which offers the possibility to turn a
given enzyme into a completely different biocatalyst by
changing the nature of the reaction being catalyzed (1). Im-
pressive examples of this approach include the transforma-
tion of phytases into peroxidases by combining them with
vanadate (2), and the alkylation of the active-site cysteine
residue in cysteine proteases such as papain with various co-
factors to create novel enzyme activities (3). Herein we de-
scribe the synthesis and preliminary evaluation of
methylamine adenine dinucleotide (MAD) as a co-factor to
turn alcohol dehydrogenases into aldolases. This type of co-
factor engineering should enable one to take advantage of
the substrate versatility and high and predictable
stereoselectivity of alcohol dehydrogenases, as has been ex-
plored in depth by Jones and co-workers (4), to achieve
Results and discussion
The aldol reaction efficiently combines carbon–carbon
bond formation with the creation of asymmetric centers.
Aldol products are useful synthons appearing either as syn-
thetic intermediates or in the final structures of many natural
products. Natural aldolase enzymes provide biocatalytic
aldolization routes to polyhydroxylated compounds, in es-
sence carbohydrates and close derivatives (5). By contrast,
less oxygenated aldols, such as polypropionates, remain the
domain of pure synthetic chemistry (6). One notable excep-
tion is the aldolase catalytic antibodies developed by Wagner
et al. (7), which allow for the preparation of optically and
stereochemically pure non-carbohydrate aldols through
Received 15 January 2002. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 28 May 2002.
This paper is dedicated to J. Bryan Jones on the occasion of his 65th birthday in recognition of his contributions to biocatalysis.
D. Wahler and J.-L. Reymond.1 Department of Chemistry & Biochemistry, University of Bern, Freiestrasse 3, 3012 Bern,
Switzerland.
1Corresponding author (e-mail: jean-louis.reymond@ioc.unibe.ch).
Can. J. Chem. 80: 665–670 (2002)
DOI: 10.1139/V02-074
© 2002 NRC Canada

666
Can. J. Chem. Vol. 80, 2002
Scheme 1. Analogy between the targeted enamine-mediated
aldolization with 1 and the hydride transfer process with NAD.
Scheme 2. Retrosynthetic analysis.
H
N
ADPO
ADPO
NH2
R
O
O
O
O
H
OH OH
OH OH
H
O
R
H
2
MAD (1)
H
NH2
O
H
N
ADPO
NH
O
O
NH2
N
H
NH2
N
OH OH
H
N
ADPO
HO
P
HO
HO
O
O
N
O
P
N
O
O
O
O
HO
O
Me
CONH2
+
OH OH
OH OH
AMP
OH OH
O
3
ADPO
NH2
Hydride transfer with
Nicotinamide Adenine Dinucleotide
(NAD)
O
BzO
OAc
OH OH
O
+
CN
+
2 H
OBz OBz
Enamine mediated aldolization with
Methylamine Adenine Dinucleotide
(MAD) 1
4
kinetic resolution of aldols by enantio- and stereoselective
retro-aldolization (8). These catalytic antibodies demonstrate
the versatility of the enamine mechanism for the aldol reac-
tion. We have shown earlier that enamine-mediated
aldolization by catalytic antibodies is possible using a co-
factor approach by combining antibodies raised against a
quaternary benzyl ammonium ion with the corresponding
benzylamine (9).
nicotinamide against modified nicotinamides (12). We ap-
proached the synthesis from a completely different and
purely synthetic angle. MAD (1) could be obtained from
deprotection of the protected MAD precursor (2), which
would itself be the product of phosphate–phosphate coupling
between adenosine monophosphate and a suitably protected
synthetic monophosphate (3) (Scheme 2). Similar phos-
phate–phosphate coupling reactions have been described for
the preparation of UDP-sugars (13). The phenacetyl group
on the primary amine was planned for technical reasons as a
hydrophobic and UV-active group to facilitate the isolation,
purification, and identification of the coupling product 1.
Furthermore, the phenacetyl group could be removed enzy-
matically under mild aqueous conditions, which do not af-
fect the labile diphosphate and glycosidic linkages present in
the target molecule (14).
Co-factor design
Drawing on the mechanistic analogy between the aldol re-
action and the reduction of aldehydes, where in both cases a
nucleophile is added to an electrophilic carbonyl group, we
envisioned that co-factor engineering might convert alcohol
dehydrogenases into aldolases. Thus MAD is expected to
promote an enamine-mediated cross-aldolization between
oxo-propionamide and an aldehyde within the binding
pocket of an alcohol dehydrogenase (Scheme 1). The en-
zyme binding pocket should provide the correct relative po-
sitioning of the reactants and induce an electrophilic
activation of the aldehyde carbonyl group, thus catalyzing
the carbon–carbon bond formation using the primary amine
functional group of the co-factor as a chemical catalyst.
Aldol reactions readily occur under amine catalysis in aque-
ous environments (10). It should be noted that a catalytic ef-
fect on the rate-limiting carbon–carbon bond forming step
should be sufficient to induce an observable catalysis of the
overall process. Indeed, catalysis of the enamine formation
step, which is not rate-limiting in enamine-mediated
aldolization catalysis (11), is not required, but it might be-
come necessary to optimized the catalyst.
Synthesis
Intermediate 3 was first prepared from the commercially
available 1-acetoxy-2,3,5-tri-O-benzoyl-D-ribose (4) (Scheme 3).
Thus, reaction with trimethylsilyl cyanide and catalytic
SnCl4 gave cyanide 5 (15), which was selectively reduced
using sodium tris(trifluoroacetoxy)borohydride in tetrahy-
drofuran to give primary amine 6 as described by Sauer and
Schneller (16). Amine 6 was then directly acylated with
phenacetyl chloride to give amide 7. The benzoyl groups
were removed under basic conditions to afford triol 8. Direct
phosphorylation of the primary alcohol in 8 with phophorus
oxychloride in triethylphosphate solvent, a common method
in nucleoside chemistry (17), gave intermediate 3 in 69%
yield. Activation to the imidazolide and coupling with the
tributylammonium salt of adenosine monophosphate yielded
protected dinucleotide 2. Finally, treatment with penicillin G
acylase gave free-methylamine adenine dinucleotide 1,
Retrosynthetic analysis
Various nicotinamide adenine dinucleotide (NAD) analogs
have been prepared by enzyme-catalyzed exchange of
© 2002 NRC Canada

Wahler and Reymond
667
Scheme 3. Synthesis of MAD.
Scheme 4. Fluorogenic assays used to detect alcohol dehydrogenase
and aldolase activities.
1. TMSCN, SnCl4 (quant.)
O
O
BzO
BzO
BzO
BzO
ADH, NAD+
X
OAc
OBz
O
OH
OBz
Me
OCoum
Me
OCoum
5 (X = CN)
4
9
10
2. NaBH(OTFA)3
6 (X = CH2NH2)
BSA,
pH > 7
3. BnCOCl, Et3N
ADH, NAD+
OH
7 (X = CH2NHCOBn)
(65 % from 5)
OCoum
Me
5. POCl3, PO(OEt)3
(69 %)
4. K2CO3, MeOH
(74 %)
O
BSA, pH > 7
O
11
HO
HO
OCoum
NHCOBn
Me
OH
12
8
O
OH
HO
O
O
Me
OCoum
= Coum
umbelliferone (13)
6. i) Bu3N, H2O; ii) CDI, DMF; iii)
AMP, DMF, 25°C, 2 d; iv) alkaline
phosphatase, 25°C, 19h (64 % to 2)
Me
14
Aldolase
O
HO
P
O
O
HO
7. Penicilllin G acylase, pH 8, 25°C,
3 h (100 %)
NHCOBn
HO
OH
Table 1. Inhibition constants for 1 and 2 with alcohol dehydrogenases.
3
O
O
P
NH2
Inhibitor
ADH
Apparent Ki ( M)
HO
O
850
1400
3.0
OH
O
1
1
2
2
Horse liver
Yeast
Horse liver
Yeast
O
HO
OH
P
O
N
N
O
2000
H2N
N
OH
N
Note: In competition with NAD+ for the oxidation of substrate 9.
OH
Conditions: 20 mM aq. borate, pH 8.8, 2 mg mL–1 BSA, 100
M
1
substrate 9, 0.1 mM and 1 mM NAD+.
which was isolated as the trifluoroacetate salt by reversed-
phase HPLC.
ously between amines and aldehydes with a pH-independent
equilibrium constant of approximately 10 mM (19), the ex-
periment showed that the binding pocket of the alcohol
dehydrogenases did not strongly bind to the imines formed
between these carbonyls and our primary amine co-factor.
Interactions between MAD and alcohol dehydrogenases
Horse liver alcohol dehydrogenase and yeast alcohol
dehydrogenase, both of which are commercially available
and represent a broad spectrum of alcohol dehydrogenases,
were selected as the target enzymes. We first tested whether
MAD could inhibit the redox activity of the enzymes, and
thus could bind to the active site. The alcohol dehydrogenase
activity was measured using the fluorogenic oxidation of the
chiral alcohols (S)-9 and (S)-11 (Scheme 4) (18). The oxida-
tion products 10 and 12 undergo a rapid b-elimination in the
presence of bovine serum albumin (BSA) to give strongly
fluorescent umbelliferone 13, which is readily detected at
micromolar concentrations in microtiter plates.
Inhibition experiments of alcohol oxidation showed that
our primary amine co-factor MAD (1) bound to the alcohol
dehydrogenases, albeit weakly (Table 1). Only protected pre-
cursor 2 showed a good affinity for horse liver alcohol
dehydrogenase. The inhibition of the ADH oxidation activity
by 1 (1 mM) was unaffected by the presence of carbonyl
compounds such as phenacetaldehyde, benzaldehyde, anis-
aldehyde, 1,2,3,6-tetrahydrobenzaldehyde, acetaldehyde, acetyl-
acetone, crotonaldehyde, cinnamaldehyde, or 2-oxopropiona-
mide (1 mM each). Since imine formation occurs spontane-
Evaluation of MAD as an aldolization co-factor
Aldolization activity was investigated using a series of
fluorogenic aldols such as 14 and close derivatives under the
same conditions as those used for the ADH-activity assay
(Table 2). Under these experimental conditions, these sub-
strates undergo clean retro-aldolization in the presence of
aldolase catalysts such as antibody 38C2 (20). Although
these aldols do not posess the oxoamide carbonyl that would
make them optimally suited to alcohol dehydrogenases ac-
cording to our design, the approximation was judged close
enough considering that alcohol dehydrogenases also easily
accept various redox co-factors where the carboxamide
group has been changed to other groups. There was no ob-
servable retro-aldolization catalysis with MAD and the
dehydrogenases for these substrates, despite the fact that
corresponding alcohol 11 is a good oxidation substrate for
these enzymes. Direct aldolization was also tested for the
cross-aldolization between 2-oxopropionamide and acetalde-
hyde, as well as for the direct dimerization of acetaldehyde
© 2002 NRC Canada

668
Can. J. Chem. Vol. 80, 2002
Table 2. Assays for aldolization activity of MAD (1).
Entry
Substrate
Conditions
MAD (1)
Detection
Reaction
1
2
Fluorogenic aldol (14)
0.1 mM, 2 mg mL–1 BSA
1 mM
None
Fluorescencea
n.r.
n.r.
3
4
2-Oxopropionamide, acetaldehyde
Acetaldehyde
10 mM each
10 mM
1 mM
None
RP-HPLC, 214 nmb
RP-HPLC, 230 nmb
n.r.
n.r.
5
6
1 mM
None
No catalysis
Spontaneous
Note: All assays in 20 mM aq. borate pH 8.8, 25°C, were followed over 48 h.
aꢀex = 360 nm, ꢀem = 460 nm
bAnalytical column: Vydac 218TP-54 (C18, pore size 300 Å), 0.45 × 22 cm, detection by UV at 214 or 230 nm, isocratic elution at 1.5 mL min–1, A =
0.1 % TFA in H2O, B = 50:50 CH3CN–H2O). tR (2-oxopropionamide) = 3.9 min at 5% B, tR (crotonaldehyde) = 5.2 min at 15% B.
HPLC-grade acetonitrile and MilliQ de-ionized water using
a Waters prepak cartridge 500 g (RP-C18 20 m, 300 Å
Table 3. Analytical RP-HPLC conditions for MAD and intermediates.
pore size) installed on a Waters Prep LC 4000 system from
Millipore, flow rate 100 mL min–1, gradient +0.5% min–1
CH3CN, detection by UV at 214 nm. TLC were performed
with fluorescent F254 glass plates. MS and HRMS (high
resolution mass spectra) were provided by Dr. Thomas
Schneeberger (University of Bern). Fluorescence measure-
ments were carried out using a Cytofluor II Plate-Reader
from Perseptive Biosystems.
Compound
% A
% B
tR (min)
3.0
5.2
10.0
2
8
3
2
1
70
90
95
98
30
10
5
2
Note: Isocratic elution at 1.5 mL min–1, detection by UV at 254 nm, A =
0.1% TFA in H2O, B = 50:50 CH3CN–H2O, analytical column: Vydac
218TP-54 (C18, pore size 300 Å), 0.45 × 22 cm.
2,5-Anhydro-3,4,6-tri-O-benzoyl-1-deoxy-1-phenyl-
acetamide-D-allitol (7)
to crotonaldehyde. In both cases there was no observable
product formation.
As described in literature (15),
a
solution of
NaBH3(O2CCF3) (1.0 g, 6.6 mmol) was prepared by adding
under nitrogen pure trifluoroacetic acid (0.51 mL, 6.7 mmol)
to an ice-bath-cooled, stirred suspension of NaBH4 (0.26 g,
6.8 mmol) in dry THF (20 mL). A solution of cyanide 5
(2.2 g, 4.7 mmol) in dry THF (13 mL) was added at 0°C,
and the mixture was stirred at 25°C for 18 h. The reaction
was quenched with de-ionized water (5 mL) and concen-
trated to dryness. The residue was partitioned between
EtOAc (20 mL) and sat. aq. NaHCO3 (7 mL), and cooled in
an ice bath. BnCOCl (2 mL, 14.7 mmol) was added, and the
mixture was vigorously stirred at 0°C for 3 h. The layers
were separated, the organic phase was washed once with wa-
ter, dried over Na2SO4, and concentrated to dryness. The res-
idue was purified by flash chromatography on silica gel
(Hex–EtOAc, 50:50) to give 7 (1.8 g, 3 mmol, 65 % over the
two steps) as a yellow syrup. IR (neat) (cm–1): 3013, 1727,
1654, 1601, 1584, 1522, 1496, 1452, 1316, 1269, 1216,
Conclusion
MAD (1) has been prepared in seven steps by an efficient
chemoenzymatic synthesis from protected D-ribose derivative
4. Penicillin G acylase provided the key to remove the
phenacetyl-protecting group of the primary amine function,
which was used as a hydrophobic, UV-visible tag to facili-
tate the purification of intermediates. The structural and
mechanistic analogy between the putative aldolization chem-
istry of MAD and the redox chemistry of the natural NAD
co-factor suggests that our synthetic co-factor should be able
to transform alcohol dehydrogenases into stereoselective
aldolases. Our preliminary investigation using the alcohol
dehydrogenases from horse liver and from yeast, however,
do not seem to indicate any activity of this type. Although
the enzyme preparations used might not contain a high con-
centration of active protein, we must conclude that aldoliza-
tion catalysis as designed does not occur within the active
site of these enzymes at a level that is detectable under our
conditions. Further experiments with other alcohol dehydro-
genases and other aldolization substrates may, however, af-
ford different results. MAD also offers an attractive
opportunity for enzyme evolution, which should be obtain-
able using fluorogenic substrates such as 14 as screening
tools.
1
1178, 1124, 1070, 1026. H NMR (CDCl3) d: 8.09–7.22 (m,
20H), 6.03 (NH), 5.56 (dd, J = 4.4, 5.9 Hz, 1H), 5.32 (dd,
J = 6.3 Hz, 1H), 4.45 (m, 4H), 3.60 (dd, J = 5.2, 5.5 Hz,
2H), 3.51 (s, 2H). 13C NMR (CDCl3) d: 172.1, 169.9, 167.0,
166.1, 135.4, 134.1, 134.0, 130.4, 130.0, 129.6, 129.3,
129.1, 128.0, 80.8, 80.5, 73.3, 73.2, 64.8, 44.2, 41.2.
MS m/z: 594 (M+).
2,5-Anhydro-1-deoxy-1-phenylacetamide-D-allitol (8)
To a solution of K2CO3 (0.24 g, 1.68 mmol) in methanol
(20 mL) and water (4 mL) was added a solution of com-
pound 7 (1.0 g, 1.68 mmol) in methanol (20 mL). The mix-
ture was stirred vigorously overnight at 25°C, and the
solvent was removed in vacuo. The residue was dissolved in
dichloromethane and water, and extracted several times with
Experimental
All reagents and enzymes were purchased from Aldrich or
Fluka. Chromatographies (flash) were performed with
Merck Silicagel 60 (0.040–0.063 mm). CH2Cl2 was distilled
from P2O5 prior to use. Preparative HPLC was done with
© 2002 NRC Canada

Wahler and Reymond
669
1
water. The aqueous phase was then purified by preparative
D-allitol (1) as a colorless solid; mp 128°C. H NMR (D2O)
d: 8.60 (s, 1H), 8.35 (s, 1H), 6.07 (br d, 1H), 4.60–4.00 (m,
11H), 2.96 (m, 2H). 13C NMR (D2O–CD3OD) d: 151.5, 149.1,
146.3, 143.3, 119.5, 89.0, 85.3, 84.0, 79.3, 75.7, 73.2,
72.3, 71.4, 66.1, 41.8. 31P NMR (D2O) d: –10.5. MS m/z:
632 ([M – 2H + Na + K]+), 616 ([M – 2H + 2Na]+),
610 ([M – H + K]+), 594 ([M – H + Na]+), 572 (M+).
HPLC to afford 8 (0.35 g, 1.24 mmol, 74 %) as a light yel-
1
low syrup (Table 3). H NMR (CD3OD) d: 7.28 (m, 5H),
3.85 (m, 4H), 3.69 (dd, J = 2.9, 11.8 Hz, 1H), 3.53 (dd, J =
2.9, 11.8 Hz, 1H), 3.51 (s, 2H), 3.40 (m, 2H). 13C NMR
(CD3OD) d: 173.9, 136.2, 129.7, 129.3, 127.6, 85.1, 82.9,
73.2, 72.0, 62.6, 43.7, 42.1.
2,5-Anhydro-1-deoxy-1-phenylacetamide-6-phosphate-D-
allitol (3)
Acknowledgments
To a solution of 8 (0.430 g, 1.53 mmol) in (EtO)3PO
(15 mL) was added under N2 a solution of POCl3 (0.8 mL,
8.7 mmol) and H2O (0.080 mL) in (EtO)3PO (10 mL), in a
bath cooled with acetone–ice. The mixture was stirred at
0°C for 3 h, and then POCl3 (0.8 mL) and H2O (0.070 mL)
in (EtO)3PO (10 mL) were added at 0°C. The reaction mix-
ture was then stored at 4°C for the night, and NaHCO3
(4.8 g) in H2O (40 mL) was added at 0°C. After one night at
4°C, the solution was extracted several times with CH2Cl2 to
remove (EtO)3PO. The aqueous phase was purified by pre-
parative RP-HPLC to afford 3 (0.38 g, 1.05 mmol, 69 %) as
This work was supported by the University of Bern, the
Swiss National Science Foundation, the European COST
program (Action D12) and the Swiss Office Fédéral de
l’Education et de la Science.
References
1. (a) J.A. Duine. Trends Biotechnol. 9, 343 (1991); (b) E.T. Kai-
ser, D.S. Lawrence, and S.E. Rokita. Ann. Rev. Biochem. 54,
565 (1985); (c) E.T. Kaiser. Angew. Chem. 100, 945 (1988);
Angew. Chem. Int. Ed. Engl. 27, 913 (1988).
2. (a) F. van de Velde, L. Könemann, F. van Rantwijk, and R.A.
Sheldon. Chem. Commun. 1891 (1998); (b) F. van de Velde,
I.W.C.E. Arends, and R.A. Sheldon. J. Inorg. Biochem. 80, 81
(2000); (c) F. van de Velde, L. Könemann, F. van Rantwijk,
and R.A. Sheldon. Biotechnol. Bioeng. 67, 87 (2000); (d) F.
van de Velde, I.W.C.E. Arends, and R.A. Sheldon. Top. Catal.
13, 259 (2000).
1
a colorless syrup. H NMR (CD3OD) d: 7.25 (m, 5H), 4.00
(m, 6H), 3.54 (s, 2H), 3.41 (d, J = 4.8 Hz, 2H). 13C NMR
(CD3OD) d: 174.4, 137.0, 130.2, 129.6, 127.9, 83.7, 83.3,
73.7, 72.8, 67.5, 43.8, 42.4.
6-Adenosine-diphosphate-2,5-anhydro-1-deoxy-1-phenyl-
acetamide-D-allitol (2)
3. (a) H.L. Levine, Y. Nakagawa, and E.T. Kaiser. Biochem.
Biophys. Res. Commun. 76, 64 (1977); (b) H.L. Levine and
E.T. Kaiser. J. Am. Chem. Soc. 100, 7670 (1978); (c) J.T.
Slama, S.R. Oruganti, and E.T. Kaiser. J. Am. Chem. Soc. 103,
6211 (1981); (d) J.T. Slama, C. Radziejewski, S.R. Oruganti,
and E.T. Kaiser. J. Am. Chem. Soc. 106, 6778 (1984).
4. (a) J.B. Jones and I.J. Jakovac. Can. J. Chem. 60, 19 (1982);
(b) A.R. Krawczyk and J.B. Jones. J. Org. Chem. 54, 1795
(1989).
5. T.D. Machajewski and C.H. Wong. Angew. Chem. 112, 1406
(2000); Angew. Chem. Int. Ed. Engl. 39, 1352 (2000).
6. (a) T. Mukaiyama, K. Banno, and K. Narasaka. J. Am. Chem.
Soc. 96, 7503 (1974); (b) D.A. Evans, J.V. Nelson, and T.R.
Taber. Top. Stereochem. 13, 1 (1982); (c) I. Paterson. Pure
Appl. Chem. 64, 1821 (1992); (d) E.M. Carreira, R.A. Singer,
and W. Lee. J. Am. Chem. Soc. 116, 8837 (1994).
Tri-n-butylammonium-AMP (0.165 g, 0.3 mmol) was dis-
solved in dry DMF (0.5 mL) at 25°C under N2. A solution of
carbonyl diimidazole (CDI, 0.102 g, 0.63 mmol) in dry
DMF (1 mL) was added, and the mixture was stirred at 25°C
for 2 h. The excess CDI was destroyed by adding MeOH
(0.026 mL), and the excess MeOH was removed under high
vacuum. A solution of compound 8 (0.100 g, 0.28 mmol) in
dry DMF (2.5 mL) was added, and the reaction was stirred
at 25°C for 2 days. The solvent was removed in vacuo, and
the residue taken in aq. 50 mM Tris buffer (10 mL,
pH 7.45). The mixture was treated with ~150 U (0.067 mL)
of calf intestinal alkaline phosphatase at 25°C overnight. Af-
ter concentration in vacuo, the crude product was purified by
preparative HPLC to give 2 (0.122 g, 0.18 mmol, 64 %) as a
1
colorless solid; mp 132–133°C. H NMR (D2O–CD3OD) d:
7. J. Wagner, R.A. Lerner, and C.F. Barbas. Science, 270, 1797
(1995).
8.64 (s, 1H), 8.34 (s, 1H), 7.23 (m, 5H), 6.09 (s, 1H),
4.03–3.81 (m, 9H), 3.55–3.39 (m, 6H). 13C NMR (D2O–CD3OD)
d: 172.2, 151.0, 146.0, 136.1, 130.0, 129.9, 129.6, 127.9,
89.0, 85.3, 83.1, 82.5, 75.8, 73.1, 72.2, 71.3, 67.7, 66.2,
43.3, 41.8. 31P NMR (D2O) d: –10.5, –14.8. MS m/z: 713
([M + Na]+), 691 ([M + H]+). HRMS m/z: calcd. for
C24H33N6O14P2: 691.1530; found: 691.1547.
8. G. Zhong, R.A. Lerner, and C.F. Barbas. Angew. Chem. 111,
3957 (1999); Angew. Chem. Int. Ed. Engl. 38, 3738 (1999).
9. (a) J.-L. Reymond and Y. Chen. Tetrahedron Lett. 36, 2575
(1995); (b) J.-L. Reymond. Angew. Chem. 107, 2471 (1995);
Angew. Chem. Int. Ed. Engl. 34, 2285 (1995).
10. For aldol condensations catalyzed by primary and secondary
amines in water, see: (a) F.G. Fischer and A. Marschall. Chem.
Ber. 64, 2825 (1931); (b) W. Langenbeck and G. Borth. Chem.
Ber. 75, 951 (1942); (c) T.A. Spencer, H.S. Neel, T.W.
Flechtner, and R.A. Zayle. Tetrahedron Lett. 6, 3889 (1965);
(d) C.D. Gutsche, D. Redmore, R.S. Buriks, K. Nowotny, H.
Grassner, and C.W. Armbruster. J. Am. Chem. Soc. 89, 1235
(1967). For Schiff base mechanism of aldolases, see: (e) E.
Grazi, P.T. Rowley, T. Cheng, O. Tchola, and B.L. Horecker.
Biochem. Biophys. Res. Commun. 9, 38 (1962).
6-Adenosine-diphosphate-1-amino-2,5-anhydro-1-deoxy-
D-allitol (1)
Compound 2 (0.060 g, 0.087 mmol) is dissolved in MeOH
(1 mL), water (2.5 mL), and NaOH (0.1 M, 2.7 mL, pH 8) at
25°C. Penicillin G acylase immobilized on Eupergit C14 (15 U,
0.140 g) was added and the suspension was stirred at
25°C for one day. After filtration, the solution was
lyophilized, and the residue purified by preparative HPLC
(A–B, 1.5 mL min–1) to afford 0.050 g (0.087 mmol, quant.)
of 6-adenosine-diphosphate-1-amino-2,5-anhydro-1-deoxy-
11. J.-L. Reymond and Y. Chen. J. Org. Chem. 60, 6970 (1995).
12. S. Tono-Oka. Bull. Chem. Soc. Jpn. 55, 1531 (1982).
© 2002 NRC Canada

670
Can. J. Chem. Vol. 80, 2002
13. G. Baisch and R. Öhrlein. Bioorg. Med. Chem. 5, 383 (1997).
14. H. Waldmann. Tetrahedron Lett. 29, 1131 (1988).
15. P.D. Cook and D.J. McNamara. J. Heterocyclic Chem. 23, 155
(1986).
E. Huberman, and M. Grifantini. J. Med. Chem. 41, 1702
(1998).
18. G. Klein and J.-L. Reymond. Helv. Chim. Acta, 82, 400
(1999).
16. D.R. Sauer and S.W. Schneller. Synthesis, 9, 747 (1991).
17. (a) M. Yoshikawa, T. Kato, and T. Takenishi. Bull. Chem. Soc.
Jpn. 42, 3505 (1969); (b) P. Franchetti, L. Cappellacci, P.
Perlini, H.N. Jayaram, A. Butler, B.P. Schneider, F.R. Collart,
19. J. Hine, B.C. Menon, J.H. Jensen, and J. Mulders. J. Am.
Chem. Soc. 88, 3367 (1966).
20. R. Pérez Carlòn, N. Jourdain, and J.-L. Reymond. Chem. Eur.
J. 6, 4154 (2000).
© 2002 NRC Canada