Characterization of acyl adenyl anhydrides: Differences in the hydrolytic rates of fatty acyl-AMP and aminoacyl-AMP derivatives

Article abstract of DOI:10.1021/jo971874f

An improved procedure has been developed to prepare RCOOPO₂-Ado (R = C₆H₁₃ and C₁₅H₃₁), the intermediate in the enzymatic synthesis of acyl-CoA's. The product has been characterized by spectral methods which, in turn, were used to define the hydrolysis rates of RCO-AMP. When R is a fatty acyl group, the hydrolysis rate is 10-fold slower than when R is aminoacyl. In both cases, the hydrolysis rate is enhanced by Mg²⁺. We speculate that the rate acceleration is due to intramolecular participation of the second carbonyl group in the aminoacyl residue.

Full text of DOI:10.1021/jo971874f

J . Org. Chem. 1998, 63, 8661-8667
8661
Ch a r a cter iza tion of Acyl Ad en yl An h yd r id es: Differ en ces in th e
Hyd r olytic Ra tes of F a tty Acyl-AMP a n d Am in oa cyl-AMP
Der iva tives
Otto F. Schall, Iwao Suzuki, Clare L. Murray, J effrey I. Gordon, and George W. Gokel*
Bioorganic Chemistry Program and Department of Molecular Biology & Pharmacology, Washington
University School of Medicine, 660 S. Euclid Avenue, St. Louis, Missouri 63110
Received October 10, 1997 (Revised Manuscript Received August 7, 1998)
An improved procedure has been developed to prepare RCOOPO2-Ado (R ) C6H13 and C15H31), the
intermediate in the enzymatic synthesis of acyl-CoA’s. The product has been characterized by
spectral methods which, in turn, were used to define the hydrolysis rates of RCO-AMP. When R
is a fatty acyl group, the hydrolysis rate is 10-fold slower than when R is aminoacyl. In both cases,
the hydrolysis rate is enhanced by Mg2+
.
We speculate that the rate acceleration is due to
intramolecular participation of the second carbonyl group in the aminoacyl residue.
Sch em e 1
In tr od u ction
The formation of acyl derivatives of coenzyme A is
catalyzed in biological systems by a group of ATP-
dependent acyl-CoA synthetases (EC 6.2.1.3). In recent
years, the genes encoding several of these enzymes have
been isolated from eukaryotes. These enzymes have been
produced using recombinant DNA techniques, setting the
stage for a detailed analysis of the enzyme’s structure-
activity relationships and host-guest interactions, as
well as an assessment of the chemical basis for their
catalysis.1
The overall process of acyl CoA formation is commonly
accepted to occur as follows (eq 1).
In contrast, aminoacyl activation by tRNA synthetases
occurs by use of adenylate (mixed anhydride) rather than
CoA (thioester).3 The single-step activation seems to be
a more primitive activation mechanism than that used
for fatty acids. Therefore, we wondered if there were
intrinsic chemical differences between the fatty acyl and
aminoacyl phosphoryl mixed anhydrides that might
explain why amino acid activation by coenzyme A has
not been observed in biological systems. To address this
question, we improved the standard method used for the
synthesis of these mixed phosphoryl anhydrides, fully
characterized them chemically and spectroscopically, and
then determined their rates of hydrolysis.
RCOOH + ATP + Mg2+ + CoA-SH f
RCOSCoA + AMP + PPi (1)
Berg2 proposed that the activation of the carboxylic
acid functional group by coenzyme A involves two steps
(eqs 2 and 3):
ATP + RCOOH f RCO-AMP + PPi
(2)
RCO-AMP + HS-CoA f RCOS-CoA + AMP (3)
The existence of an acyl-AMP intermediate is attrac-
tive from the chemical perspective because it offers a
solution to an obvious problem: how is RCOOH activated
for attachment of the CoA activating group? Activation
of the fatty acid carbonyl group for nucleophilic substitu-
tion by formation of a mixed phosphoryl carboxyl anhy-
dride obviates the need for direct reaction between
RCOOH and multifunctional CoA (Scheme 1). The mixed
anhydride is presumably required to esterify the thiol
functional group of CoA. In the absence of the mixed
anhydride or some other activating group, carboxylate
acylation of the thiol would likely fail.
Resu lts a n d Discu ssion
Im p r oved Meth od for th e Syn th esis of Acyl Ad e-
n ylp h osp h or yl An h yd r id es. Several methods have
been used historically to synthesize a limited number of
acyl adenylates. Acetyl adenylate was prepared enzy-
matically by Berg from sodium acetate using acetyl CoA
synthetase, ATP, and MgCl2.2b Berg also prepared the
compound nonenzymatically from acetyl chloride and
silver adenylate.2c In the same paper, a preparation is
reported from adenylic acid and acetic anhydride in
aqueous pyridine. A related aqueous pyridine prepara-
tion was used to prepare propanoyl adenylate.4 Talbert
* Corresponding author. Tel. 314/362-9297, Fax 314/362-9298, e-
mail ggokel@pharmsun.wustl.edu.
(1) See, for example, Knoll, L. J .; Schall, O. F., Suzuki, I.; Gokel, G.
W.; Gordon, J . I. J . Biol. Chem. 1995, 270, 20090-20097.
(2) (a) Berg, P. J . Biol. Chem. 1956, 222, 991, 1015, 1025. (b) Berg,
P. J . Biol. Chem. 1956, 222, 1015. (c) Berg, P. J . Biol. Chem. 1956,
222, 1025. (d) Berg, P. Science 1957, 129, 895.
(3) Merrick, W. C. Microbiol. Rev. 1992, 56, 291-315.
(4) Moyed, H. S.; Lipmann, F. J . Bacteriol. 1957, 73, 117.
10.1021/jo971874f CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/30/1998

8662 J . Org. Chem., Vol. 63, No. 24, 1998
Schall et al.
Ta ble 1. In flu en ce of Meta l Ca tion a n d Ba se on th e
F or m a tion of P a lm itoyl-AMP Mixed An h yd r id e (4) fr om
AMP a n d P a lm itic An h yd r id ea
Sch em e 2
exp no.
additive
none
LiOH‚H2O
LiOH‚H2O
NaOH
LiCl
NaCl
MgCl2‚H2O
(i-Pr)2NEt
Et3N
equiv of M+ or base
% yield
1
2
3
4
5
6
7
8
9
60 ( 2
79 ( 3
0b
80
58 ( 1
58
56
84
84 ( 2
0.5
5.0
0.5
0.5
0.5
0.5
0.5
0.5
a
The reactions were carried out according to the optimized
b
procedure (see Experimental Section). No product was isolated
from the reaction mixture.
and Huennekens used dicyclohexylcarbodiimide to couple
adenylic acid and butyric acid in aqueous pyridine.5
Butyroyl adenylate was also prepared by Peng from
butyric acid anhydride and adenylic acid in aqueous
pyridine.6 Hexanoyl adenylate was obtained by White-
house and co-workers7 by using an analogous approach
from hexanoyl anhydride.
Vignais and Zabin (V-Z) described the standard
method used for synthesizing palmitoyl adenyl anhy-
dride.8 This involves reaction of adenylic acid with a
carboxylic acid anhydride (Scheme 2). The V-Z synthe-
sis involves LiOH-mediated reaction of AMP (adenyl
phosphate) with palmitoyl anhydride in aqueous solution
to afford the mixed anhydride in approximately 50%
yield.10 This reaction occurs in the opposite sense to the
one postulated for biological systems, i.e., the phosphate
rather than the carboxyl group serves as the nucleophile
(cf. eq 2).
There are three obvious reactions that could diminish
yield in the V-Z procedure. First, the palmitoyl anhy-
dride may be consumed by reaction with the 6-amino
group of AMP. This is an unlikely possibility: we found
that 3′,4′-isopropylideneadenosine gave the amide when
treated with palmitic anhydride in pyridine under strin-
gent conditions (100 °C for 8 h) and then only as a
mixture with the O-acylation product. Second, the use
of Li+ rather than Mg2+ may limit the efficiency of
reaction between phosphate and the anhydride; Mg2+ is
employed in biological systems9 (see eq 1). Furthermore,
Herschlag and J encks10 reported that Mg2+ enhances the
rate of reaction between pyridyl-phosphates and various
oxygen nucleophiles. Hydrolysis of the acyl-AMP deriva-
tive is a third possible contributor to poor yield in the
V-Z reaction. The rate of hydrolysis should be base-
dependent.
When neither cation nor base was present, the maxi-
mum yield obtained was ∼60%. This was accomplished
by conducting the reaction at 23 °C rather than 36 °C,
increasing pyridine concentration (40 f 50%) so that the
reactants were more soluble, and by doing an extractive
workup (pH 3, ether) rather than simply precipitating
the product with acetone. The cation was not the critical
factor in determining yield: substitution of LiCl, NaCl,
or MgCl2 for LiOH (all at [salt] ) 0.5 M) resulted in a
yield of mixed anhydride equivalent to that obtained
when both cation and base were absent (Table 1). The
type of base used was less important than its concentra-
tion: the yield of product was similar (79-84%) if 0.5 M
LiOH, NaOH, (i-Pr)2NEt, or Et3N was used. When 5
equiv of LiOH was employed, no product was obtained
(Table 1).
Ch a r a cter iza tion of th e Vign a is-Za bin Rea ction
P r od u cts. Previously reported characterizations of the
V-Z reaction product, palmitoyl-AMP (4), have been
limited to three qualitative tests: (i) the reaction of its
carbonyl group with hydroxamic acid, (ii) diazotization
of its free 6-amino group by reaction with nitrous acid,
and (iii) periodic acid cleavage of its vicinal, nonacylated
2′,3′-hydroxyl groups.11
A series of spectroscopic studies was performed on 9
compounds to make unequivocal structural assignments
to possible V-Z reaction products (Table 2). Compounds
1 and 2 (adenosine and AMP) were used as reference-
controls. The fully acetylated derivative 3 of adenosine
was used to determine the chemical shift changes induced
by acylation of all hydroxyls and the 6-amino group.
Acetyl was selected as the acyl residue in this case
because its protons are observed as a singlet rather than
as a >30 proton multiplet (see also 7), as would be the
case for palmitoyl. The palmitoyl group was isotopically
labeled in compounds 5 and 6 to facilitate NMR spectral
assignments. Compound 8 was a substrate in the
hydrolytic studies described below. Finally, 9 is a 13C-
labeled version of 7 in which the ribosyl hydroxyls have
been protected so that they could not be acylated.
With these thoughts in mind, we examined the effects
of cation and base on the yield of C15H31COOPO2-Ado (4).
This compound was selected as a model product because
palmitate is among the most abundant fatty acids in
eukaryotic cells and because Vignais and Zabin used it
in their experiments.
1H-NMR analysis revealed that proton chemical shifts
for the 6-amino group were more pH-sensitive than those
for the C2 and C8 protons of adenylate or for the ribosyl
1′-protons. These pH variations in signal position are
illustrated in Table 3 for acetyl-AMP (7) in DMSO-d6.
Note that at pH 5.5 the NH2 signal shifts upfield by 1.5
ppm.
(5) Talbert, P. T.; Huennekens, F. M. J . Am. Chem. Soc. 1956, 78,
4671.
(6) Lee Peng, C. H. Biochim. Biophys. Acta 1956, 22, 42.
(7) Whitehouse, M.; Moeksi, H.; Gurin, S. J . Biol. Chem. 1957, 226,
813.
(8) Vignais, P. V.; Zabin, I. Biochim. Biophys. Acta 1958, 29, 263.
(9) Pecoraro, V. L.; Hermes, J . D.; Cleland, W. W. Biochemistry 1984,
23, 5262-5271.
(10) (a) Herschlag, D.; J encks, W. P. J . Am. Chem. Soc. 1990, 112,
1942-1950. (b) Herschlag, D.; J encks, W. P. Biochemistry 1990, 29,
5172-5179.
(11) J encks, W. P. Methods Enzymol. 1963, 6, 762-766.

Characterization of Acyl Adenylphosphoryl Anhydrides
J . Org. Chem., Vol. 63, No. 24, 1998 8663
Ta ble 2. 1H-NMR Sp ectr a l Sh ifts for 1-9 in DMSO-d 6
compounda
C2-H
C8-H
6-NH2
H1′
amide
CHnCOb
CH3
1 adenosine
2 AMP‚H2O
8.330
8.351
(8.97)
8.691
8.120
8.165
(8.61)
8.265
7.650
7.482
(8.18) (6.80)
5.860
5.911
-
-
-
-
-
-
-
-
9.300
-
3 adenosine, 2′,3′,5′-O-trisacetate, 6-N-acetamide
-
6.206
2.597
2.131
2.088
2.058
2.317
(2.45)
[2.23]
2.317
-
4 AMP-palmitic acid anhydride
8.593
(9.15)
[8.52]
8.571
8.571
8.603
8.589
8.331
(8.62)
[8.07]
8.309
8.309
8.341
8.331
8.152
8.680
(8.20) (6.85)
[7.05] [6.07]
8.651
8.644
8.706
8.683
5.926
-
-
-
0.835
(0.85)
[0.72]
0.834
0.835
-
0.814
-
5 AMP-palmitic acid anhydride (1-13C-palmitate)
6 AMP-palmitic acid anhydride (2,2-d2-palmitate)
7 acetyl-AMP
5.930
5.931
5.948
5.933
6.141
-
-
∼9.1
2.054
2.325
1.954
2.600
8 heptanoyl-AMP
9 adenosine-2′,3′-O-isopropylidene-5′-O-(1-13C-acetate)-6-N-(1-13C-acetamide) 8.679
-
(8.840) (8.795)
-
(6.656) (12.363) (1.899)
(2.701)
-
-
a
Concentrations were ∼20 mM. Resonance positions are given in ppm downfield of residual CD3SOCD2H (δ2.490) or pyridine (δ7.190).
Values in parentheses refer to chemical shifts in pyridine-d5 solutions; those in brackets refer to solutions in D2O:H2O:C5D5N:THF-d8
(0.1:0.03:0.2: 0.67, v/v). For compounds 3 and 7, shifts refer to the acetyl methyl group. Infrared spectral data are included in the
b
Experimental Section.
Ta ble 3. Med iu m p H Dep en d en ce of Ad en in e 1H-NMR
Reson a n ces in AMP -Acetic An h yd r id e, 7
chemical shiftsa for
pHb
H1′
C8-H
C2-H
6-NH2
2.0
3.9
5.5
5.943
5.942
5.897
8.346
8.343
8.126
8.612
8.607
8.424
8.786
8.756
7.260
a
[7])20 mM (DMSO-d6), calibrated against the 2.490 ppm
b
signal of internal CD3SOCD2H. Acidity was adjusted with 5%
aq HCl.
1
The H-NMR spectral assignments were then used to
confirm that the compound obtained from both the
original Vignais-Zabin procedure and from our modified
method was, in fact, palmitoyl-AMP (4, Table 2 and
Experimental Section). 31P- and 13C-NMR spectral analy-
ses revealed no cross signal from the 13C nucleus of 13C-
enriched palmitoyl-AMP (5) to the NH of adenine (see
Experimental Section), thus confirming the absence of
acylation at the 6-NH2 group.
Th e Effects of Ca tion on Hyd r olytic Sen sitivity
of C15H31COOP O2-Ad o. Thin layer chromatographic
analysis indicated that palmitoyl-AMP underwent hy-
drolysis to palmitate and AMP when strirred at ambient
temperature in 5:95 (v/v) H2O:C5H5N for 1-2 h. We
determined the kinetics of hydrolysis at 37 ( 0.1 °C for
both palmitoyl-AMP, 4, and the more soluble heptanoyl
derivative, 8. To do so, we followed changes in their 31P-
NMR spectra because the 31P-resonances of acyl-AMP
and 5′-AMP are separated by nearly 12 ppm (Table 4).
Both 4 and 8 (25 mM) were studied in 20% DMSO-d6/
80% 150 mM N-tris(hydroxymethyl)methyl-3-aminopro-
panesulfonic acid (v/v) at pH ) 9.1. The plot of [RCOO-
F igu r e 1. 31P-NMR-monitored hydrolysis at 37 °C of hep-
tanoyl-AMP, 8. [8] ) 25 mM. b [MgCl2] ) 0 mM; 9 [MgCl2] )
50 mM; 2 [MgCl2] ) 100 mM; and O [MgCl2] ) 150 mM; lines
(s) represent ln 31P-integral intensity values for the 31P
resonances of 8.
PO2-Ado] (as ln(31P integral)) vs time was linear in both
cases over a period of g3 half-lives (e.g., Figure 1). From
the equations of the best exponential fits, [C7AMP] )
90.6e-(-0.001588)t and [C16AMP] ) 90.1e-(-0.001605)t, the hy-
drolysis rate constants were calculated to be similar: 1.59
× 10-3 min-1 for 8 and 1.61 × 10-3 min-1 for 4. This
translates to a half-life (t1/2) of 438 min for heptanoyl-
AMP and 432 min for palmitoyl-AMP under the experi-
mental conditions used. In separate experiments, we
added a cation ([MgCl2] ) 50, 100, or 150 mM). At
[MgCl2] ) 100 mM, the hydrolysis rates for both 4 and 8
increased ∼10-fold to ∼1.7 × 10-2 min-1. As can be seen
in Figure 1, incrementally increasing the concentration

8664 J . Org. Chem., Vol. 63, No. 24, 1998
Ta ble 4. 31P -NMR a n d Hyd r olysis Da ta for F a tty Acyl a n d Am in oa cyl-AMP sa
Schall et al.
compd no.
compdb
additive
Khydrol (min-1
)
t1/2 min δ(31P) ppm before hydrolysis δ(31P) ppm after hydrolysis
8
8
8
8
4
10
10
11
11
C6H13CO-AMP
C6H13CO-AMP
C6H13CO-AMP
C6H13CO-AMP
-
1.6 × 10-3
1.2 × 10-2
1.7 × 10-2
8.2 × 10-2
1.6 × 10-3
1.4 × 10-2
5.0 × 10-2
1.9 × 10-2
7.2 × 10-2
430
58
41
9
430
49
14
37
10
-7.188
-7.13
4.294
3.88
3.885
3.89
50 mM MgCl2
100 mM MgCl2
150 mM MgCl2
-
-7.126
-7.12
C
15H31CO-AMP
-7.280
-7.076
-7.229
-7.110
-7.131
4.24
Cbz-A-AMP
Cbz-A-AMP
Cbz-A-A-AMP
Cbz-A-A-AMP
-
4.240
3.706
4.194
3.886
100 mM MgCl2
-
100 mM MgCl2
31P-derived hydrolysis data and chemical shifts were determined in 20% DMSO-d6:150 mM TAPS (pH ) 9.1) solutions. Concentrations
of the mixed anhydride species were 15 mM. C7 and C16 stand for heptanoyl and hexadecanoyl, respectively. Cbz and “A” stand for
a
b
carboxybenzyloxy and alanine, respectively.
of MgCl2 from 50 mM to 150 mM ([8] ) 25 mM) enhanced
the hydrolysis rate correspondingly. Since 4 equiv of a
divalent salt should have completely occupied all of the
donor sites on 8, the increase in rate may be a medium
effect. Indeed, the largest rate increment is observed
between 0 and 50 mM of added MgCl2.
Hydrolysis of 10 or 11 (25 mM) at 37 ( 0.1 °C was
followed in 20% DMSO-d6 containing 150 mM [3-[[-tris-
(hydroxymethyl)methyl]amino]-1-propanesulfonic acid
(TAPS) (pH ) 9.1) by monitoring the disappearance of
the 7.2 ppm signal and the appearance of a new 31P
resonance at 4.2 ( 0.1 ppm.
Together, these observations indicate that fatty-acyl-
AMP’s are sensitive to hydrolysis, that this hydrolysis
does not appear to be affected by varying acyl chain
length from heptanoyl to hexadecanoyl, and that the
presence of a salt containing a Lewis acidic cation
accelerates the hydrolytic cleavage. Herschlag and Jencks4
have also observed a Mg2+-dependent acceleration of the
hydrolysis rate in a chemically related but distinct
system: they thoroughly studied the reactions of RCOO-
with phosphorylated pyridines, picolines, and morpho-
lines and showed how Mg2+ catalyzes the process.
Hyd r olytic Sen sitivity of Am in oa cyl-AMP Der iva -
tives. Aminoacyl adenylates have been prepared previ-
ously by De Moss and co-workers12 who condensed
aminoacyl chlorides with silver adenosine 5′-phosphate.
Berg also obtained aminoacyl adenylates by reaction of
amino acids with ADP in the presence of DCC.13 Mold-
ave, Castelfranco, and Meister14 prepared 15 N-car-
bobenzoxyaminoacyl adenylates by a similar, DCC-
mediated condensation reaction in 40-81% yields. They
noted that the non-nitrogen-protected aminoacyl adeny-
lates “are extremely unstable and rapidly undergo hy-
drolysis and other reactions including conversion to a
compound which appears to be the amino acid ester of
the 2′ (or 3′) hydroxyl group or adenosine 5′-phosphate”.
Kellerman15 studied the preparation and reactions of
benzoyl adenylate and noted that it “is seen to be quite
stable in acid and to be considerably more stable in alkali
than is acetyl adenylate.... Both are more stable than
the aminoacyl adenylates which decompose rapidly at pH
8.” In the case of benzoyl adenylate, the rate of hydroly-
sis was expressed as %-min-1 vs pH. The hydrolysis rate
reported by Kellerman at 37 °C was ∼0 at pH values
between 0 and 8. As the base strength was increased,
reaction rate also increased until 100% hydrolysis was
observed to occur in about 2 min at pH 13.
Remarkably, the hydrolysis of Cbz-Ala-AMP (10) and
Cbz-Ala-Ala-AMP (11) was 10-fold faster than the hy-
drolysis of 4 or 8 under comparable conditions (Table 4).
When MgCl2 was present (100 mM), the hydrolysis rates
for 8 and 11 increased 8-fold and 4-fold, respectively.
Mech a n istic Con sid er a tion s. The ∼10-fold differ-
ence in hydrolysis rates between aminoacyl and fatty-
acyl AMPs could reflect an electron-withdrawing effect
of the amide residue relative to the alkyl chain: R′OC-
ONHCHRCOPO2Ado compared to R′COPO2Ado. One
way to assess this is to evaluate the Hammett parameters
for acetic acid esters. The general formulation is log(k/
k0) ) σF. For the ionization of acetic acid derivatives,16
σI ) -0.03 for the n-hexyl group, +0.26 for NHCOC2H5,
and +0.44 for OCON(CH3)2.17 Since the σI value for alkyl
is essentially 0, and F is constant for the identical reaction
conditions, the equation log(k/k0) ) σF, reduces to log(k/
k0) ) σI. Using this analysis and our data, log(1.4 × 10-2
/
1.6 × 10-3) ) log(8.75) ) 0.94. This value is significantly
higher than expected for either an amide or carbamate
substituent on acetic acid. A simple electronic effect does
not appear to account for this modest but significant
difference.
In our own studies, we compared the hydrolysis rates
of aminoacyl-AMP and fatty-acyl-AMP derivatives. To
do so, we prepared Cbz-Ala-AMP (10) and Cbz-Ala-Ala-
AMP (11) by reaction of the N-protected amino acid or
dipeptide with AMP in the presence of DCC in aqueous
pyridine (see Experimental Section). A single 31P-NMR
signal was observed at 7.2 ( 0.1 ppm for both 10 or 11.
We believe that these observed differences in hydroly-
sis rates are more likely to represent differences in the
(13) Berg, P. J . Biol. Chem. 1958, 233, 608.
(14) Moldave, K.; Castelfranco, P.; Meister, A. J . Biol. Chem. 1959,
234, 841-848.
(15) Kellerman, G. M. J . Biol. Chem. 1958, 231, 427-443.
(16) Charton, M. Prog. Phys. Org. Chem. 1981, 13, 119.
(17) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165-
195.
(12) DeMoss, J . A.; Genuth, S. M.; Novelli, G. D. Proc. Natl. Acad.
Sci. U.S.A. 1956, 42, 325.

Characterization of Acyl Adenylphosphoryl Anhydrides
J . Org. Chem., Vol. 63, No. 24, 1998 8665
Sch em e 3
Sch em e 4
rates of cleavage of an aminoacyl vs fatty acyl group from
AMP rather than differences in the sites of cleavage.
There is a single hydrolytic cleavage site in the fatty acyl
AMP derivative. In the protected aminoacyl AMP de-
rivative, hydrolysis could, in principle, occur at the
urethane of the protecting group, the amide bond of the
dipeptide (if present), or at the mixed anhydride link.
Gisin and Merrifield noted that the free R-amino group
of an anchored dipeptide can undergo the base-catalyzed,
intramolecular reaction shown in Scheme 3.18
This reaction is unlikely to occur for these aminoacyl-
AMP derivatives because there is no free amino group
present that could initiate the cyclization. Alternatively,
an amino group could be formed by hydrolysis of the Cbz
protecting group to afford a dipeptide:
Assuming that the different rates of hydrolysis indicate
that the aminoacyl AMP carbonyl group is more reactive
than the carbonyl group of fatty acyl AMP, we speculate
that the reaction of the two derivatives with the thiol of
CoA to form the corresponding CoA thioester (see eq 2,
above) would not only be equally feasible for both
compounds but would be favored from the chemical
perspective for aminoacyl AMPs. If so, this may have
placed evolutionary pressure on the substrate recognition
site of acyl-CoA synthetases to avoid ligation with amino
acids.
PhCH2OCONHCHMeCONHCHMeCOO-AMP f
H2NCHMeCONHCHMeCOO-AMP
Exp er im en ta l Section
The resulting H-Ala-Ala-AMP derivative could cyclize
to give a diketopiperazine intermediate that, in turn,
could lead to the observed hydrolytic cleavage product.
However, to generate this intermediate, hydrolysis of the
urethane link in the Cbz protecting group would have to
be faster than hydrolysis of the mixed anhydride (acti-
vating group). Even in the unlikely event that urethane
hydrolysis was faster than hydrolysis of the mixed
anhydride, cyclization by this mechanism would not be
possible for 11.
Why should there be any difference in the hydrolysis
rates of fatty-acyl and aminoacyl-AMPs at the carboxyl
group? The carbonyl group of an amino acid might be
more hindered than the corresponding functional group
of the fatty acid, and therefore hydrolysis would be slower
for the former. Our experimental results indicate that
the opposite is true. A plausible explanation for the rate
difference is that there is anchimeric assistance to
hydrolysis of the aminoacyl-AMP derivatives that is not
possible for the fatty acyl derivatives. This point is
illustrated in Scheme 4. As shown, the second carbonyl
group in aminoacyl-AMP can serve as an intramolecular
nucleophile to give the oxazolidinone. This is not possible
for fatty-acyl-AMPs which lack the second carbonyl
group. Thus, the reactivities of the aminoacyl and fatty
acyl carboxyl groups would differ. The cyclization shown
in the scheme is well-known: it is analogous to that
observed when the terminal amino acid of a peptide is
cleaved during the Edman degradation. Such a reaction
would occur, presumably with the same rate, whether
the AMP derivative was attached to a protected amino
acid or to a protected dipeptide as in 11.
1H-NMR spectra were recorded at either 300 or 500 MHz.
13C- and 31P-NMR spectra were recorded at corresponding
frequencies. Proton chemical shifts are reported in ppm (δ)
downfield from Me4Si or using residual, nondeuterated, solvent
(CHCl3, δ 7.240; CD3SOCD2H, δ 2.490; or pyridine-d4-H, δ
7.190) as a reference signal. Signals are reported in the
following order: chemical shift, peak multiplicity (br ) broad;
s ) singlet; d ) doublet; t ) triplet; m ) multiplet), integration,
and assignment. Phosphorus and carbon chemical shifts are
reported relative to 85% H3PO4 (0.00 ppm) and pyridine-d5
(triplet at 123.44 ppm) as references, respectively. Infrared
spectra were recorded on an FT-IR instrument and were
calibrated against the 1601 cm-1 band of polystyrene. Melting
points were determined in open capillaries and are uncor-
rected. High resolution mass spectrometric data were obtained
at the Washington University Resource for Biomedical and
Bioorganic Mass Spectrometry. TLC analyses were performed
using silica gel 60 F-254. Silica gel was used for flash
chromatography as originally described by Still et al.19 All
reactions were conducted under dry N2 or protected from
moisture with Drierite. THF and Et2O were dried and distilled
from sodium benzophenone ketyl. Anhydrous solvents were
stored over activated molecular sieves (3 Å) and were dis-
pensed by syringe under dry N2. 1-13C-hexadecanoic acid, 2,2-
d2-hexadecanoic acid, Cbz-Ala-Ala-OH, Cbz-Ala-OH, and deu-
terated NMR solvents were obtained commercially. All other
solvents and reagents were of the best grade available com-
mercially and were used without further purification unless
otherwise stated.
Ad en in e, a d en osin e (1), a d en osin e-5′-m on op h osp h a te
(2), a n d 2′, 3′-O-isop r op ylid en ea d en osin e were obtained
from Aldrich Chemical Co. and were used as received. In some
cases, commercial AMP‚H2O was dehydrated by storing over
P2O5 in vacuo (e0.05 Torr) overnight.
1-13C-Hexa d eca n oic An h yd r id e. 1-13C-Hexadecanoic acid
was dissolved in anhydrous Et2O (30 mL) and stirred while a
solution of DCC (865 mg, 4.2 mmol) in Et2O (20 mL) was added
dropwise during 20 min. The resulting white suspension was
stirred (12 h) and filtered, and the filtrate was evaporated in
vacuo. The residue was crystallized (hexanes) to yield the
(18) (a) Gisin, B. F.; Merrifield, R. B. J . Am. Chem. Soc. 1972, 94,
3102-3106. (b) Barany, G.; Merrifield, R. B. Solid-Phase Peptide
Synthesis. In The Peptides; Gross, E., Meienhofer, J . Eds.; Academic
Press: New York, 1979; Vol. 2, pp 1-284. (c) Pedroso, E.; Grandas,
A.; de las Heras, X.; Eritja, R.; Giralt, E. Tetrahedron Lett. 1986, 743-
746. (d) Grant, G. A., Ed.; Synthetic Peptides; W. H. Freeman & Co.:
New York, 1992; p 143.
(19) Still, W. C.; Kahn, M.; Mitra, A. J . Org. Chem. 1978, 43, 2923.

8666 J . Org. Chem., Vol. 63, No. 24, 1998
Schall et al.
anhydride (850 mg, 90%) as white needles, mp 62-63 °C (lit.20
mp for the nonlabeled compound: 64 °C). 1H-NMR: 0.856 (t,
6H, CH3), 1.230 (broad s, 48H, (CH2)12), 1.623 (m, 4H, CH2-
CH213CdO), 2.410 (q, 4H, CH213CdO, J HCC ) 7.2 Hz).
2,2-d 2-Hexa d eca n oic An h yd r id e. 2,2-d2-Hexadecanoic
acid was added to Ac2O (10 mL, 105 mmol) and stirred and
heated to reflux for 1.5 h. The solution was cooled, evaporated
in vacuo to a light yellow syrup, dissolved in toluene (10 mL),
evaporated, and then held under high vacuum (<0.5 Torr)
overnight. The resulting waxy solid was crystallized from
hexanes to afford the anhydride (840 mg, 88%) as a white solid,
mp 63-64 °C. 1H-NMR: 0.855 (t, 6H, CH3), 1.233 (broad s,
48H, (CH2)12), 1.611 (m, 4H, CH2CD2CdO).
water aspirator. This procedure was repeated four additional
times. Distilled water (3 mL) was added to the resulting
whitish gel, and the pH of this suspension was brought to ∼3
with 5% aq HCl. The gel turned into a white suspension and
was extracted with Et2O (3 × 30 mL) by stirring vigorously
and separating the top organic layer by suction as before. The
white solid was transferred to a 100-200 Å fritted funnel and
the remaining aqueous solution filtered. Successive washings
with 5% aq HCl (5 mL), acetone (3 × 5 mL), and Et2O (2 × 5
mL), followed by overnight drying in vacuo (ambient temper-
ature, over P2O5), afforded a white powder in 82% yield with
properties identical to those of 4.
Ad en osin e-5′-p h osp h or ic-(1-13C-h exa d eca n oic) An h y-
d r id e, 5. AMP (70 mg, 0.202 mmol) was dissolved in 40%
H2O:pyridine (v/v, 7 mL) containing LiOH‚H2O (8.5 mg, 0.20
mmol). The resulting clear solution was warmed (37 °C), and
1-13C-hexadecanoic anhydride (200 mg, 0.40 mmol) in THF (12
mL) was added at once. The initial white suspension turned
clear in 1-2 min; the final solution temperature was 36 °C.
After 10 min, the solution was poured into acetone (80 mL,
-20 °C). After 2 h in a freezer at -20 °C, the suspension was
filtered, and the last portion of supernatant and solids were
combined and centrifuged; the remaining solvent was de-
canted. The gellike residue was suspended in H2O (20 mL),
the pH was adjusted to ∼2 (5% aq HCl), Et2O (20 mL) was
added, and the two-phase system was stirred vigorously and
then centrifuged. The solid was filtered (100-200 Å fritted
funnel) and washed with acetone (-20 °C, 2 × 2 mL) and then
with Et2O (3 × 2 mL). After being dried in vacuo (0.2 Torr,
ambient temp, P2O5), 5 was obtained (86 mg, 72%) as a white
solid, mp ∼175 °C (darken) 179-180 °C (dec). IR (KBr): 3321,
3276, 2920, 2851, 1713, 1607, 1557, 1504, 1470, 1412, 1242,
Ad en osin e-2′,3′,5′-O-tr isa ceta te-6-N-a ceta m id e, 3. To a
solution of adenosine (200 mg, 0.75 mmol) in pyridine (20 mL)
was added a solution of Ac2O (1.52 g, 1.4 mL, 15 mmol) in
pyridine (5 mL) during 5 min. The clear solution was heated
under reflux for 8 h, cooled, and quenched with EtOH (5 mL).
After evaporation of the solvent, the thick, yellow, oily residue
was dissolved in minimum CHCl3 and chromatographed (flash,
SiO2, 10% MeOH:CHCl3) to afford 3 (180 mg, 55%) as a yellow
gum. IR (CCl4): 3246-2900, 1752, 1703, 1609, 1589, 1542,
1522, 1470, 1374, 1283, 1225, 1035 cm-1 1H-NMR: 2.058 (s,
.
3H, CH3CdO(O)), 2.088 (s, 3H, CH3CdO(O)), 2.131 (s, 3H,
CH3CdO(O)), 2.597 (s, 3H, CH3CdO(NH)), 4.320-4.450 (m,
3H, 5′-Hs + 4′-H), 5.637 (t, 1H, 3′-H), 5.927 (t, 1H, 2′-H), 6.223
(d, 1H, 1′-H), 8.265 (s, 1H, 8-H), 8.691 (s, 1H, 2-H), ∼9.3 (broad,
∼1H, amide-H, exchanged in D2O). HRMS (FAB, [M + H]+)
calcd for C18H21N5O8: 436.1468; found: 436.1455.
Ad en osin e-5′-p h osp h or ic H exa d eca n oic An h yd r id e
(p a lm itoyl-AMP ), 4. AMP (200 mg, 0.58 mmol) was dissolved
in 35 mL of 40% H2O:pyridine (v/v) containing 0.1% (18 mg,
0.43 mmol) of LiOH‚H2O. The resulting clear solution was
warmed to ∼40 °C, and a solution of palmitic anhydride (600
mg, 1.21 mmol) in THF (70 mL) was added at once. The initial
white suspension turned clear in 1-2 min; the final temper-
ature of the solution was 36 °C. After 10 min, the solution
was diluted with acetone (250 mL, -20 °C), and the solution
temperature was held at -20 °C. After 15 min, the mixture
was filtered, and the last portion of supernatant and solids
was combined and centrifuged; the remaining solvent was then
decanted. The gellike solid was suspended in H2O (25 mL)
and the pH adjusted to ∼2 with 5% aq HCl. Et2O (30 mL)
was added and the two-phase system stirred vigorously and
then centrifuged. The solid was separated, washed succes-
sively with ice-chilled acetone (2 × 15 mL) and Et2O (2 × 15
mL), and then crystallized from H2O:EtOH (1:15 v/v) to afford,
after drying over P2O5 at 60 °C and 0.1 Torr overnight, 4 as a
white, amorphous powder (185 mg, 55%) mp: ∼175 °C
(shrink), 180-182 °C (dec). IR (KBr): 3292, 3127, 2919, 2850,
1758, 1702, 1608, 1560, 1508, 1470, 1420, 1242, 1080, 1030,
1140, 1119, 1099, 923, 817 cm-1 1H-NMR (DMSO-d6): 0.834
.
(t, 3H, CH3), 1.208 (broad s, 24H, (CH2)12), 1.441 (m, 2H, CH2-
CH213CdO), 2.317 (q, 2H, CH213CdO, J HCC ) 7.2 Hz), 4.085
(broad s, 3H, 4′ and 5′-Hs), 4.169 (t, 1H, 3′-H), 4.563 (t, 1H,
2′-H), 5.930 (d, 1H, 1′-H), 8.309 (s, 1H, 8-H), 8.571 (s, 1H, 2-H),
8.651 (broad s, ∼2H, -NH2, exchanged in D2O). HRMS (FAB,
[M + H]+) calcd for 13C12C25H44N5O8P: 587.3039; found:
587.3022.
Ad en osin e-5′-p h osp h or ic (2,2′-d 2-h exa d eca n oic) An -
h yd r id e, 6. AMP (70 mg, 0.202 mmol) was dissolved in 40%
H2O:pyridine (v/v, 7 mL) containing LiOH‚H2O (8.5 mg, 0.20
mmol). The resulting clear solution was warmed to ∼37 °C,
and a solution of 2,2′-d2-hexadecanoic (palmitic) anhydride
(200 mg, 0.40 mmol) in THF (12 mL) was added at once. A
white suspension immediately resulted. This initial suspen-
sion became clear during 2-3 min. The final temperature of
the solution was 36 °C. After 10 min, the solution was poured
into acetone (80 mL, -20 °C), and the temperature was
adjusted to -20 °C. After 1 h in a -20 °C freezer, the
suspension was filtered. The resulting gellike material was
dissolved in H2O (2.5 mL), the pH was adjusted to ∼2 (5% aq
HCl), Et2O (20 mL) was added, and the two-phase system was
stirred vigorously and then centrifuged. The solid was sepa-
rated and washed with acetone (0 °C, 2 × 5 mL) and then with
Et2O (2 × 5 mL). After being dried in vacuo (0.2 Torr, ambient
temp, P2O5), 6 (96 mg, 80%) was obtained as a white solid,
mp 176-181 °C (darken, shrink). IR (KBr): 3292, 133, 2921,
2851, 2773, 2360, 2344, 1756, 1654, 1649, 1609, 1558, 1508,
817 cm-1
.
1H-NMR (DMSO-d6): 0.835 (t, 3H, CH3), 1.207
(broad s, 24H, (CH2)12), 1.441 (broad s, 2H, CH2CH2CdO),
2.317 (t, 2H, CH2CdO), 4.085 (broad s, 3H, 4′ and 5′-Hs), 4.166
(t, 1H, 3′-H), 4.563 (t, 1H, 2′-H), 5.926 (d, 1H, 1′-H), 8.331 (s,
1H, 8-H), 8.593 (s, 1H, 2-H), 8.680 (broad s, ∼2H, NH2,
exchanged in D2O). HRMS (FAB, [M + H]+) calcd for
C
26H44N5O8P: 586.3006; found: 586.2989.
Ad en osin e-5′-m on op h osp h or ic Hexa d eca n oic Mixed
An h yd r id e (p a lm itoyl-AMP ), 4. Im p r oved P r oced u r e. 5′-
AMP‚H2O (200 mg, 0.548 mmol) was dissolved by magnetic
stirring in a solution of 8 mL of 50% H2O:pyridine (v/v)
containing 23 mg (0.57 mmol) of NaOH. A solution of palmitic
anhydride (570 mg, 1.152 mmol) in 8 mL of THF was added
by pipet in three portions. A white suspension resulted
immediately, and stirring was continued at room temperature.
After 15 min, the suspension had slowly thinned to an almost
clear solution. Et2O (25 mL) was then added and the two-
phase system stirred vigorously for 2 min.21 After 1-2 min
had elapsed, the ether phase was removed by suction using a
1471, 1420, 1323, 1240 1100, 1067, 1034 cm-1
.
1H-NMR
(DMSO-d6): 0.835 (t, 3H, CH3), 1.207 (broad s, 24H, (CH2)12),
1.428 (m, 2H, CH2CD2CdO), 4.083 (broad s, 3H, 4′ and 5′-Hs),
4.169 (t, 1H, 3′-H), 4.563 (t, 1H, 2′-H), 5.931 (d, 1H, 1′-H), 8.309
(s, 1H, 8-H), 8.571 (s, 1H, 2-H), 8.644 (broad s, ∼2H, NH2,
exchanged in D2O). HRMS (FAB, [M + H]+) calculated for
C
26(2H)2(1H)42N5O8P: 588.3131; found: 588.3124.
Aden osin e-5′-m on oph osph or ic Acetic An h ydr ide (acet-
yl-AMP ), 7. AMP (300 mg, 0.86 mmol) was dissolved in 40%
H2O:pyridine (v/v, 27 mL) containing LiOH‚H2O (36 mg, 0.86
mmol). The resulting solution was warmed to 40 °C, and a
solution of acetic anhydride (214 mg, 200 µL, 2.1 mmol) in THF
(50 mL) was added all at once. The resulting white suspension
quickly turned clear (t ) 27 °C). After 10 min, the solution
was poured into acetone (-20 °C, 300 mL). After 15 min in a
(20) CRC Handbook of Chemistry and Physics; Lide, D. R., Ed.; CRC
Press: Boca Raton, FL, 1993-1994; pp 3-359.
(21) This procedure removes most of the pyridine which may
otherwise foster hydrolysis of the carboxylic phosphoric anhydride.

Characterization of Acyl Adenylphosphoryl Anhydrides
J . Org. Chem., Vol. 63, No. 24, 1998 8667
freezer at -20 °C, the suspension was filtered, and the residual
white gel was washed with acetone (10 °C, 20 mL) and Et2O
(15 mL). The resulting gellike solid was dissolved in H2O (2.5
mL) and adjusted to pH ) 2 (5% aq HCl). Absolute EtOH (-20
°C, 200 mL) was added, and the suspension was concentrated
in vacuo to ∼100 mL (water bath temperature ∼35 °C) and
then allowed to stand (rt) overnight. Filtration yielded 7 (180
mg, 54%) as an amorphous, white powder (Rf ) 0.43 in
n-BuOH:acetone:HOAc: 5% NH4OH:H2O [9:5:3:3:1]; mp ∼183
°C (darken), 205 °C (foam, dec). IR (KBr): 3439-3119, 1945,
1757, 1698, 1608, 1561, 1508, 1476, 1421, 1372, 1331, 1243,
Ad en osin e-5′-m on op h osp h or ic N-(Ca r boxyben zyloxy)-
a la n yl Mixed An h yd r id e (Cbz-a la n yl-AMP ), 10. By a
procedure identical to that for the preparation of 11 (see
below), AMP (340 mg, 0.98 mmol), N-Cbz-Ala-OH (225 mg,
1.02 mmol), water (1.2 mL), pyridine (5.2 mL), concentrated
HCl (90 µL, 1.1 mmol), and DCC (5.15 g, 25 mmol dissolved
in 6 mL of pyridine) yielded mixed anhydride 11 as a white
powder in 55%. 1H- and 31P-NMR integrals indicate that the
powder consists of a mixture of AMP and Cbz-Ala-AMP in an
approximately 3.5:6.5 ratio. 1H-NMR (DMSO-d6): 1.097 (d,
3H, -CH3, J HH ) 6.6 Hz), 4.069 (m, 3H, 4′- and 5′-H’s), 4.166
(t, 1H, 3′-H), 4.217 (m, 1H, OdC-CH-CH3), 4.579 (t, 1H, 2′-
H), 5.005 (s, 2H, PhCH2), 5.933 (d, 1H, 1′-H), 7.322 (m, 5H,
phenyl), 7.844 (broad s, 2H, NH2, exchanged with D2O), 8.220
(d, 1H, amide-NH, exchange with D2O), 8.412 (s, 1H, 8-H),
8.532 (s, 1H, 2-H). 31P-NMR: -7.076 (s).
1088, 1015, 925, 824 cm-1 1H-NMR (DMSO-d6): 2.054 (s, 3H,
.
CH3CdO), 4.100 (m, 3H, 4′ and 5′-H’s), 4.185 (m, 1H, 3′-H),
4.571 (t, 1H, 2′-H), 5.948 (d, 1H, 1′-H), 8.341 (s, 1H, 8-H), 8.603
(s, 1H, 2-H), 8.706 (broad s, ∼2H, -NH2, exchanged in D2O).
HRMS (FAB, [M + H]+) calcd for C12H16N5O8P: 390.0815;
found: 390.0803.
Ad en osin e-5′-m on op h osp h or ic (N-ca r boxyben zyloxy-
a la n in e)-N-a la n in e Mixed An h yd r id e (Cbz-a la n yl-a la n yl-
AMP ), 11. AMP (340 mg, 0.98 mmol) and N-Cbz-Ala-Ala-OH
(300 mg, 1.02 mmol) were suspended in water (1.2 mL) and
dissolved by addition of 5.2 mL of pyridine. Concentrated HCl
(90 µL, 1.1 mmol) was added with a microsyringe causing some
fumes to form. The clear solution was cooled to -10 °C in an
acetone:ice bath, and then a solution of DCC (5.15 g, 25 mmol)
in 6 mL of pyridine was added in one portion and stirred
vigorously. After 1 h the cold bath was removed, and the
reaction was allowed to warm to room temperature. The light
white suspension thickened considerably with time. After 2.5
h (rt), the thick yellow suspension was filtered through a
fritted funnel and the filtrate added to 150 mL of acetone held
at -50 °C (dry ice:acetone bath). The resulting precipitate was
quickly centrifuged in 30-mL batches while keeping the
remaining suspension cold (-50 °C). The collected solids were
transferred onto a 15-20 µm fritted funnel, washed with
acetone (3 × 10 mL) and Et2O (2 × 10 mL), dried for 5-10
min, and placed in vacuo (e0.05 Torr) overnight to afford a
white powder (305 mg, 51%). 1H- and 31P-NMR integrals
indicate that the powder consists of a mixture of AMP and
Cbz-Ala-Ala-AMP in an approximately 3:7 ratio. 1H-NMR
(DMSO-d6): 1.025 (d, 3H, CH3, J HH ) 6.6 Hz), 1.193 (d, 3H,
CH3, J HH ) 6.6 Hz), 4.072 (m, 4H, 4′- and 5′-H’s, and
OdCCHCH3), 4.176 (t, 1H, 3′-H), 4.267 (m, 1H, OdCCHCH3),
4.584 (t, 1H, 2′-H), 5.007 (s, 2H, PhCH2), 5.935 (d, 1H, 1′-H),
7.331 (m, 5H, phenyl), 7.883 (broad s, 2H, NH2, exchanged with
D2O), 8.235 (m, 2H, amide-NH’s, exchange with D2O), 8.402
(s, 1H, 8-H), 8.515 (s, 1H, 2-H). 31P-NMR: -7.110 (s).
Ad en osin e-5′-m on op h osp h or ic Hep ta n oic An h yd r id e
(h ep ta n oyl-AMP ), 8. AMP (140 mg, 0.404 mmol) was dis-
solved in a solution of 14 mL of 40% H2O:pyridine (v/v)
containing LiOH‚H2O (20 mg, 0.480 mmol). The resulting
clear solution was warmed to ∼40 °C, and a solution of
heptanoic anhydride (198 mg, 0.8 mmol, 2 equiv) in 24 mL of
distilled THF was added in one portion. A very light white
suspension quickly formed and disappeared almost immedi-
ately. After 10 min at 40 °C, the solution was poured into
200 mL of acetone held at -20 °C, resulting in precipitation
of a gel. After 1 h in a freezer at -20 °C, the suspension was
filtered. The resulting gellike material was dissolved in 1 mL
of distilled water and the pH adjusted to ∼2 with 5% aq HCl.
The white suspension was poured into a centrifuge tube,
mixed, and washed with Et2O (3 × 4 mL, centrifuged and
decanted each time). After the third washing, the aqueous
phase was also decanted, and the residual solid was placed in
a fritted funnel (100-200 Å) and further washed with acetone
(3 × 5 mL) and Et2O (2 × 5 mL) and air-dried to yield a white
powder. After being dried overnight in vacuo (<0.2 Torr, rt)
over P2O5, 142 mg (71%) of the product was collected (mp
turned light brown ∼168 °C, then shrunk at 177-178 °C and
darkened. Rf ) 0.68 in in n-BuOH: acetone:H2O:5% NH4OH:
AcOH [9:5:3:2:2 (v/v)] on SiO2. 1H-NMR (DMSO-d6): 0.814 (t,
3H, CH3), 1.201 (broad s, 6H, (CH2)3), 1.444 (m, 2H, CH2-
CH2CdO), 2.325 (t, 2H, CH2CdO), 4.097 (broad m, 3H, 4′ and
5′-H’s), 4.172 (t, 1H, 3′-H), 4.564 (t, 1H, 2′-H), 5.933 (d, 1H,
1′-H), 8.331 (s, 1H, 8-H), 8.589 (s, 1H, 2-H), 8.683 (broad s,
∼2H, NH2, exchanged in D2O). HRMS (FAB, [M + H]+) calcd
for C17H26N5O8P: 460.1597; found: 460.1586.
Hyd r olysis Kin etics. Hydrolysis reactions were conducted
in 5 mm NMR tubes containing the mixed anhydride species
at a concentration of 25 mM. The solvent was a mixture of
20% DMSO-d6 and 80% 150 mM aqueous TAPS buffer (v/v)
at pH ) 9.1. The temperature was held at 37.0 ( 0.1 °C. In
those experiments in which MgCl2 was present, its concentra-
tion in the final solution was 50 mM, 100 mM, or 150 mM.
The 31P signal, relative to external (sealed glass capillary) 85%
H3PO4 was monitored. Samples were obtained as frequently
as could be managed experimentally, typically 10-15 min
between data points. A minimum of 15 data points was
collected for each run except at [MgCl2] ) 150 mM which
occurred too rapidly for this to be practical.
Ad en osin e-2′,3′-O-isop r op ylid en e-5′-O-(1-13C-a ceta te)-
6-N-(1-13C-a ceta m id e), 9. 2′,3′-O-Isopropylideneadenosine
(150 mg, 0.48 mmol) was dissolved in pyridine (8 mL)and 1-13C-
acetic anhydride (300 mg, 276 µL, 2.88 mmol) was added. The
mixture was stirred at 90 °C for 2.5 h, and the solution was
allowed to cool overnight while stirring. EtOH (5 mL) was
added, and the mixture was evaporated to afford a nearly
colorless syrup. EtOH (15 mL) was added, the mixture was
evaporated, and the resulting gum was dissolved in minimal
CHCl3 and then purified by Chromatotron chromatography
(Al2O3 rotor, 4 mm thickness, 5:95 (v/v) MeOH:CHCl3). Prod-
uct 9 (Rf ) 0.23) was isolated (73 mg, 33%) from late fractions
as a white foam. IR (CCl4): 3229, 3119, 2992, 2938, 2361,
2343, 1707, 1674, 1660, 1609, 1587, 1559, 1542, 1470, 1372,
1258, 1212, 1106, 1078 cm-1
.
1H-NMR: 1.379 (s, 3H, i-Pr-
Ack n ow led gm en t. This work was supported by
NIH grants to J .I.G. and G.W.G. (GM 36262 and AI
30188). We also acknowledge the Washington Univer-
sity Resource for Biomedical and Bioorganic Mass
Spectrometry. We are grateful toDrs. D. Andre´ D’Avignon
and J eff Kao for valuable assistance.
CH3), 1.601 (s, 3H, i-Pr-CH3), 1.954 (d, 3H, CH313CdO(O), J HCC
) 6.9 Hz), 2.600 (d, 3H, CH313CdO(NH), J HCC ) 6.6 Hz), 4.286
(m, 2H, 5′-Hs), 4.493 (m, 1H, 4′-H), 5.036 (dd, 1H, 3′-H), 5.455
(dd, 1H, 2′-H), 6.141 (d, 1H, 1′-H), 8.152 (s, 1H, H-8), 8.679 (s,
1H, H-2), 9.1 (broad s, ∼1H, amide-H, exchanged in D2O).
HRMS (FAB, [M + H]+) calculated for (13C)2(12C)15H21N5O6:
394.1637; found: 394.1622.
J O971874F