Biomimetically activated amino acids. Catalysis in the hydrolysis of alanyl ethyl phosphate

Article abstract of DOI:10.1021/ja972406q

Alanyl ethyl phosphate (1) is an activated derivative of alanine that is functionally related to the corresponding aminoacyl adenylate, the initial activated amino acid intermediate in protein biosynthesis. To establish the inherent reactivity of these species, the kinetic parameters for hydrolysis of alanyl ethyl phosphate in water at 25 °C were determined. There is catalysis by acid (k = 4 x 10^-4 M^-1s^-1) and base (k = 1.7 M^-1 s^-1) along with two pH-independent processes (k = 3 x 10^-5 and 1.6 x 10^-3 s^{- 1}) that are connected as a kinetic titration curve of the amino group of alanyl ethyl phosphate (pK(a) = 7.8). The results are consistent with mechanisms proceeding via addition to the carbonyl of water or hydroxide with proton migrations. Reaction with methanol is slower than reactions with water while reaction with 2-propanol leads to complex products. In solutions sufficiently concentrated for ³¹P NMR analysis, alanyl ethyl phosphate also undergoes reactions that produce alanylalanine and other condensation products. Metal ions catalyze the hydrolysis reactions through complex formation. Cupric and zinc ions are most effective (~100-fold larger rate constant than water: association constants > 100 M^-1) with magnesium and calcium forming weaker and less reactive complexes. These results show that aminoacyl alkyl phosphates are sufficiently stable to be used in water and that metal ions can facilitate their reactions. Improved catalysts will be needed to facilitate biomimetic processes such as aminoacylation of t-RNA.

Full text of DOI:10.1021/ja972406q

J. Am. Chem. Soc. 1997, 119, 12089-12094
12089
Biomimetically Activated Amino Acids. Catalysis in the
Hydrolysis of Alanyl Ethyl Phosphate
Ronald Kluger,* Richard W. Loo, and Vince Mazza
Contribution from the Lash Miller Laboratories, Department of Chemistry, UniVersity of Toronto,
Toronto, Ontario, Canada M5S 3H6
ReceiVed July 17, 1997X
Abstract: Alanyl ethyl phosphate (1) is an activated derivative of alanine that is functionally related to the
corresponding aminoacyl adenylate, the initial activated amino acid intermediate in protein biosynthesis. To establish
the inherent reactivity of these species, the kinetic parameters for hydrolysis of alanyl ethyl phosphate in water at 25
°C were determined. There is catalysis by acid (k ) 4 × 10-4 M-1s-1) and base (k ) 1.7 M-1 s-1) along with two
pH-independent processes (k ) 3 × 10-5 and 1.6 × 10-3 s-1) that are connected as a kinetic titration curve of the
amino group of alanyl ethyl phosphate (pKa ) 7.8). The results are consistent with mechanisms proceeding via
addition to the carbonyl of water or hydroxide with proton migrations. Reaction with methanol is slower than reactions
with water while reaction with 2-propanol leads to complex products. In solutions sufficiently concentrated for 31
P
NMR analysis, alanyl ethyl phosphate also undergoes reactions that produce alanylalanine and other condensation
products. Metal ions catalyze the hydrolysis reactions through complex formation. Cupric and zinc ions are most
effective (∼100-fold larger rate constant than water: association constants > 100 M-1) with magnesium and calcium
forming weaker and less reactive complexes. These results show that aminoacyl alkyl phosphates are sufficiently
stable to be used in water and that metal ions can facilitate their reactions. Improved catalysts will be needed to
facilitate biomimetic processes such as aminoacylation of t-RNA.
The formation of proteins requires activation of the carboxyl
group of each amino acid that is added. In most cells, activation
results from the reaction of the amino acid and ATP to form a
mixed anhydride, an aminoacyl adenylate, as an enzyme-bound
intermediate. The aminoacyl adenylate goes on to acylate the
terminal 3′-hydroxyl group of the cognate t-RNA.1,2 The mixed
anhydride is a member of the general class of aminoacyl
phosphate monoesters3 and is one of the principal processes of
molecular activation in biochemistry.4 Although the reaction
patterns of aminoacyl adenylates are significant in catalysis of
this key biological process,5 little is known about these
compounds since they are not readily available.6 How reactive
are they? How can their reactivity be enhanced and perhaps
put to use? The inherent reaction patterns should arise from
the aminoacyl phosphate ester function, independent of the
adenosine moiety. Since we recently developed a convenient
preparation of aminoacyl phosphate esters,3 we have been able
to produce pure compounds in large quantities and begin
systematic studies.
We noted that the compounds are relatively stable in water
but that they can react rapidly with amine nucleophiles.3 In
this paper we focus on the mechanisms and catalysis of
hydrolysis of alanyl ethyl phosphate (1) as a typical member
of the general class. Aminoacyl alkyl phosphates may also be
useful for reactions that mimic those of aminoacyl adenylates
with t-RNAs, transferring the aminoacyl moiety to a terminal
ribose hydroxyl. Thus, we have also assessed the reactivity of
1 with methanol and 2-propanol as preliminary models. We
find that these reactions are slow and complex, establishing the
need for the development of catalysts for aminoacyl transfer to
weak nucleophiles.
Experimental Section
Methods. Commercial reagents were used as received. NMR
1
spectra were recorded at 200, 300, or 400 MHz for H, 100 MHz for
13C, and 121 or 160 for 31P. Kinetic runs were carried out in a jacketed
reaction vessel with a thermoregulated recirculating water bath.
Kinetics. Since alanyl ethyl phosphate is an anhydride of a
carboxylic acid and the monobasic form of ethyl phosphate, a significant
amount of acid is released during hydrolysis in neutral and alkaline
solutions. Therefore, a “pH-stat” was used to measure rates of
appearance of products at pH 7 and higher. The jacketed reaction flask
contained a magnetic stirrer, a glass electrode, and an input tube from
an automatic buret containing 0.01 M sodium hydroxide solution. The
system was continuously flushed with nitrogen (to minimize absorption
of carbon dioxide). A solution of alanyl ethyl phosphate (0.5 mL of
0.005 M) was transferred into the reaction vessel and sufficient 1.0 M
potassium chloride was added to bring the total volume to 5 mL. The
volume of base added from the buret was recorded as a function of
time (through a computer), for approximately 4 half-lives. The data
X Abstract published in AdVance ACS Abstracts, December 1, 1997.
(1) Voet, D.; Voet, J. G. Biochemistry, 2nd ed.; Wiley: New York, 1995;
Section 30.
(2) Schimmel, P. Annu. ReV. Biochem. 1987, 56, 125.
(3) Kluger, R.; Li, X.; Loo, R. W. Can. J. Chem. 1996, 74, 2395-400.
(4) Metzler, D. Biochemistry, the Chemical Reactions of LiVing Cells;
Academic Press: New York, 1977; Chapter 11.
(5) Berg, P. J. Biol. Chem. 1958, 233, 608.
(6) Lewinsohn, R.; Paecht-Horowitz, M.; Katchalsky, A. Biochim.
Biophys. Acta 1967, 140, 24.
S0002-7863(97)02406-2 CCC: $14.00 © 1997 American Chemical Society

12090 J. Am. Chem. Soc., Vol. 119, No. 50, 1997
Kluger et al.
Scheme 1
Scheme 2
were fit to the integrated first-order rate law by nonlinear regression.
The pseudo-first-order kinetic data were used to measure dependence
on concentrations of added species.
For reactions in solutions more acidic than pH 7, the conversion of
reactant to product was followed by periodic recording of 31P NMR
(121 MHz). The integrals of the signal corresponding to alanyl ethyl
phosphate were noted relative to that of a standard solution of disodium
pyrophosphate (0.1 M in deuterium oxide) in a sealed capillary tube
(also serving as a “lock signal” for the NMR). For buffer dependence
studies, 1.0 mL of 0.025 M alanyl ethyl phosphate in sodium acetate
buffer concentrations of 0.2 to 1.0 M (I ) 1.0 with sodium chloride as
needed) was placed in 5-mm NMR tubes. Spectra were obtained with
128 scans and a 2-s delay in the broad band proton-decoupled mode.
Data collection for each spectrum took approximately 4 min. Spectra
were taken at varying intervals for 4 to 5 half-lives. Between data
collections, samples were kept at 25 °C in a circulating water bath.
Acidic solutions were prepared from titrated 1 M HCl that was diluted
with 1 M sodium chloride solution. Reactions were followed as
described above.
Data analysis and fitting was done by nonlinear regression on a
computer. For the 31P NMR data, the relative values of the integrals
of the signals corresponding to alanyl ethyl phosphate and disodium
pyrophosphate were fitted to a first-order exponential decay. This gave
an observed first-order rate coefficient for the disappearance of alanyl
ethyl phosphate. For the pH-stat data, the volumes of added sodium
hydroxide solution were fitted to a first-order growth curve to obtain
an observed first-order rate coefficient for the disappearance of alanyl
ethyl phosphate.
of kobs versus pH gives a curve that fits a titration equation,
with the apparent pKa from kinetics being identical with the
value found by equilibrium titration ()7.8 ( 0.1). The kinetic
and thermodynamic titration curves are shown in Figure 1.
The data in Figure 1A were fit to a titration curve equation
to give as limits the rate constants for the uncatalyzed hydrolysis
of the zwitterionic form (below pH 6) and the anionic form
(above pH 9). The lower rate was determined independently
by extrapolation of rates of reactions in buffered solutions (see
below). These values are also determined by fitting the data to
a plot of rate coefficient versus fraction of all material in the
anionic form.7 The rate constant obtained from the plot for
hydrolysis of the zwitterionic form of alanyl ethyl phosphate is
(3.9 ( 0.4) × 10-5 s-1. A plot of observed first-order rate
coefficient versus free base fraction (8 points: 0.2 to 0.8)
extrapolated to unity gives kOH ) (1.6 ( 0.2) × 10-3 s-1
.
Base Catalysis. In more alkaline solutions (reaction mea-
sured for six points between pH 10 and 11), the rate of
hydrolysis is proportional to hydroxide concentration. This is
likely to involve the reaction of hydroxide with the anionic form
of the substrate (A in Scheme 2). The slope of the plot of
observed first-order rate coefficients versus hydroxide concen-
tration gives the second-order rate constant for specific base
catalyzed hydrolysis of alanyl ethyl phosphate: (1.67 ( 0.04)
Synthesis. Tetra-n-butylammonio-N-t-boc-(L)-alanyl ethyl phosphate
was prepared according to the reported procedure3 by using dicyclohexyl
carbodiimide (0.0011 mol) to couple N-(t-boc)-(L)-alanine (0.0010 mol)
and bis(tetra-n-butylammonium)ethyl phosphate (0.0011 mol) in dichlo-
M-1 s-1
.
1
romethane (20 mL). The product is a colorless oil: (76% yield) H
Water Catalysis and Acid Catalysis. There is not a
sufficient release of proton equivalents generated by the
hydrolysis of alanyl ethyl phosphate below pH 7 to permit the
reaction to be followed by a pH-stat in dilute solutions. For
these cases 31P NMR was used to follow the consumption of
reactant (by the integrated signal from alanyl ethyl phosphate
versus an external standard). The first-order rate coefficient
for hydrolysis of alanyl ethyl phosphate in weakly acidic
solutions was obtained for a series of acetate/acetic acid buffers
(0.2, 0.4, 0.6, 0.8 M; see Experimental Section for details)
between pH 4.0 and 5.0. The buffers catalyze the reaction (first
order in buffer concentration). The first-order rate constant for
the reaction in the absence of buffer is obtained by extrapolation
to [buffer] ) 0: pH 4.0, k ) (3.2 ( 0.3) × 10-5 s-1; pH 4.5,
k ) (3.4 ( 0.3) × 10-5 s-1; pH 5.0, k ) (3.4 ( 0.4) × 10-5
s-1. The combination gives a pH-independent rate constant of
(3.3 ( 0.6) × 10-5 s-1 for the spontaneous hydrolysis pathway
of the zwitterionic form.
NMR (200 MHz, chloroform-d) δ 1.19 (t, 3H), 1.30 (t, 12H), 1.34 (d,
3H), 1.38 (s, 9H), 3.36 (q, 8H), 3.97 (m, 2H), 4.21 (m, 1H), 5.15 (d,
1H); 31P NMR (chloroform-d) δ -6.8.
(L)-Alanyl ethyl phosphate was prepared by treatment of tetraethyl-
ammonium-N-t-boc-(L)-alanyl ethyl phosphate with trifluoroacetic acid.
Overall yield: 40% from N-(tert-butoxycarbonyl)-(L)-alanine): mp 120
°C dec; IR (KBr) 1764, 1624, 1574, 1251, 1230, 1075, 1047 cm-1; 1H
NMR (200 MHz, deuterium oxide) δ 1.19 (t, 3H), 1.50 (d, 3H), 3.95
(m, 2H), 4.16 (q, 1H); 13C NMR (100 MHz, deuterium oxide) δ 15.63
(s), 16.22 (d, JC-P ) 6.6 Hz), 49.95 (d, JC-P ) 7.3 Hz), 64.78 (d, JC-P
) 6.6 Hz), 167.92 (d, JC-P ) 8.8 Hz); 31P NMR (121 MHz, deuterium
oxide) δ -6.58. FAB MS calculated 197, found (m/z) 196 (M - H).
All further discussion of this compound omits the stereochemical
descriptor.
Results
The hydrolysis of alanyl ethyl phosphate produces alanine
and ethyl phosphate (Scheme 1).
The hydrolysis of alanyl ethyl phosphate is acid catalyzed:
(0.1 M HCl, kobs ) 4.8 × 10-5 s-1; 0.5 M HCl, kobs ) 2.3 ×
10-4 s-1; 1.0 M HCl, kobs ) 3.8 × 10-4 s-1). The second-
order rate constant for specific acid catalysis is (3.8 ( 0.5) ×
10-4 M-1 s-1 from the linear correlation of the observed first-
order rate coefficients and acid concentrations.
Both the zwitterionic (HA) and anionic forms (A) of alanyl
ethyl phosphate should undergo hydrolysis (Scheme 2) as given
by the rate law in eq 1. ST is the concentration of alanyl ethyl
phosphate in all forms; fHA and fA are the fractions in the
zwitterionic form and anionic forms, respectively.
kobs[ST] ) (kfHA + k′fA)[ST]
(1)
Overall Rate Law. A rate law from four mechanisms that
predominate in different conditions is summarized in eq 2. This
A series of observed first-order rate coefficients was measured
by pH-stat in solutions whose pH brackets the pKa of the
zwitterionic form of alanyl ethyl phosphate (see below). A plot
(7) Jencks, W. P. Catalysis in Chemistry and Enzymology; McGraw-
Hill: New York, 1969; Chapter 11.

Catalysis in the Hydrolysis of Alanyl Ethyl Phosphate
J. Am. Chem. Soc., Vol. 119, No. 50, 1997 12091
Figure 1. (A) pH vs rate coefficient (s-1) for the hydrolysis of alanyl ethyl phosphate. (B) Titration of alanyl ethyl phosphate with sodium hydroxide
(relative volume). Both fit the standard titration equation with the pKa for the amino group of alanyl ethyl phosphate as 7.8 ( 0.1.
Reaction of Alanyl Ethyl Phosphate with Alcohols. As a
preliminary to the use of these materials for aminoacylation of
t-RNA, we examined the inherent reactivity of alanyl ethyl
phosphate with alcohols. Alanyl ethyl phosphate reacts with
methanol to give alanine methyl ester and ethylphosphoric acid.
The identity of alanine methyl ester was confirmed by com-
parison of 1H NMR and thin-layer chromatographic patterns to
those for genuine alanine methyl ester. The rate of methanolysis
of alanyl ethyl phosphate in methanol, by 31P NMR at 25 °C,
is very slow, with an observed first-order rate constant of (6.8
( 0.2) × 10-5 s-1 (t1/2 ) 170 min).
The reaction of alanyl ethyl phosphate in 2-propanol (a model
for the 2° alcohol of the ribosyl ring of t-RNA) is even slower.
The cloudy suspension of alanyl ethyl phosphate undergoes only
a partial reaction over a period of days. After 1 week at room
Figure 2. pH vs rate coefficient profile for hydrolysis of alanyl ethyl
temperature, the 31P NMR indicates 86% of the alanyl ethyl
phosphate (25 °C) as described in the text: (b) data from 31P NMR;
phosphate remained intact. 1H NMR analysis of the crude
mixture does not indicate a signal for a 2-propyl group although
alanine residues other than alanyl ethyl phosphate are evident.
(O) data from a pH-stat.
Table 1. Rate Constants for the Hydrolysis of Alanyl Ethyl
Phosphate (25 °C) Derived from Fitting Experimental Data to Eq 1
To increase the rate of the reaction between alanyl ethyl
term
value
phosphate and 2-propanol, heat and base were applied. (Alanyl
ethyl phosphate is soluble in refluxing 2-propanol.) After 24
h, 31P NMR indicated that all the starting material was
consumed. Both chromatography and 1H NMR revealed a
mixture of products. Similarly, a number of products appeared
when the reaction was repeated at room temperature with an
equivalent of triethylamine or imidazole. These products were
not identified (this is the subject of ongoing studies).
kH
(3.8 ( 0.5) × 10-4 M-1 s-1
(3.3 ( 0.6) × 10-5 s-1
(1.6 ( 0.2) × 10-3 s-1
(1.67 ( 0.04) M-1 s-1
kHA
kA
kOH
was used to generate a pH-rate profile (Figure 2) with the rate
constants in Table 1.
Metal Ion Catalysis. Initially, the effects of the divalent
ions of copper, zinc, magnesium, and calcium on the rate of
hydrolysis of alanyl ethyl phosphate [0.0005 M] were screened
with metal ion concentrations of 0.005 M at pH 7. Cupric and
zinc ions cause a significant increase at these low levels, while
magnesium and calcium do not. The detailed concentration
dependence was then measured.
V ) kobs[1] ) (kH[H+] + kHAfHA + kAfA + kOH[OH-])[1]
(2)
Complications. Since alanyl ethyl phosphate and alanine
(from hydrolysis) contain an amino group, they can also react
as nucleophiles with the electrophilic carbonyl center of alanyl
ethyl phosphate. At higher concentrations of substrate, this
competes with hydrolysis since the free amino group is a much
better nucleophile than water toward alanyl ethyl phosphate and
it is present in reasonable quantities at pH > pKa of the
The effect of variation in ion concentration on the rate of the
reaction was found for each metal. The addition of each
increased the observed first-order rate coefficient up to satura-
tion. The curves fit the equation for saturating binding. As in
other catalytic systems, the fit provides an apparent association
constant for the metal and substrate as well as the rate constant
for the decomposition of the metal complex. The curve for
catalysis by cupric ion is shown in Figure 3.
1
zwitterionic form. Thin-layer chromatography and H NMR
indicated the formation of such products. Since these result
from reactions that are second order in aminoacyl compound,
they are promoted by the high concentrations used for NMR-
based kinetic studies. The solutions used for pH-stat kinetics
do not lead to significant second-order processes since the
samples are dilute. A further complication of the NMR method
is the necessary use of buffers. They catalyze the reaction,
requiring the rate due to buffer to be separated.
The dependence of rate on zinc, magnesium, and calcium
ions was also found. All enhance the reaction, but the effects
of magnesium and calcium ions are smaller. These results are
shown in Figures 4, 5, and 6.

12092 J. Am. Chem. Soc., Vol. 119, No. 50, 1997
Kluger et al.
Figure 6. Effect of calcium ion concentration on the observed first-
order rate coefficient of hydrolysis of alanyl ethyl phosphate.
Figure 3. Dependence of observed first-order rate coefficient for
hydrolysis alanyl ethyl phosphate.
Scheme 3
Table 2. Rate and Association Constants from Fitting Kinetic
Data (Scheme 3 and Figures 3-6) for Metal-Catalyzed Hydrolysis
of Alanyl Ethyl Phosphate at pH 7.0, 25 °Ca
metal
k2, s-1
k2/kobs
K1, M-1
Cu(II)
Zn(II)
Mg(II)
Ca(II)
2.6 × 10-2
2.7 × 10-2
1.0 × 10-2
0.9 × 10-2
100
135
50
250
141
28
45
15
a The “k2/kobs” entry refers to the acceleration observed for the metal
complex at pH 7.
Discussion
Figure 4. Effect of zinc ion concentration on the observed rate
coefficient for hydrolysis of alanyl ethyl phosphate.
The hydrolysis of alanyl ethyl phosphate in water follows
the general reaction patterns of acyl phosphate monoesters,8-12
complicated by the presence of the R-amino group. Since the
amino group exists in protonated and neutral forms, it alters
the reactivity at the acyl center. Scheme 4 summarizes a kinetic
scheme that can be used to connect the data to reaction
mechanisms in neutral and alkaline solutions. The mechanisms
are consistent with the data and are distinguished only by
protonation states.
Alanyl ethyl phosphate is shown in the zwitterionic (SH) and
anionic (S) forms that are the major species in neutral solutions.
The acidity constant Ka connects S (plus a proton) and SH. A
set of tetrahedral intermediates can form by the addition of water
or hydroxide ion to S and SH.
These intermediates in Scheme 4 are tautomeric forms that
can revert directly to reactants; they are also likely to lose a
proton since the least basic sites are protonated. Loss of the
proton from the ether oxygen leads to the same intermediates
as formed by addition of hydroxide (I and L). A shift of the
same proton gives the somewhat more stable intermediates, I
and K. These can expel ethyl phosphate as the monoanion
rather than as the more basic dianion. (Some of the intermedi-
ates differ by the position of a proton and are therefore difficult
to distinguish. The addition of hydroxide to SH leads to
transition states with the same net charge, as does addition of
water to S.) Addition of hydroxide may be rate determining
since its reverse is likely to be slower than the competing loss
Figure 5. Effect of magnesium ion concentration on the observed first-
order rate coefficient for hydrolysis of alanyl ethyl phosphate.
The simplest scheme that can be used to fit the data in each
saturation plot is shown in Scheme 3.
The association constant, K1, and the maximal rate constant,
k2, are determined from the fit of the data to the equation for
each plot, where S is alanyl ethyl phosphate and M represents
divalent metal ions.
(8) Disabato, G.; Jencks, W. P. J. Am. Chem. Soc. 1961, 83, 4400.
(9) Disabato, G.; Jencks, W. P. J. Am. Chem. Soc. 1961, 83, 4393.
(10) Kluger, R.; Tsui, W.-C. J. Org. Chem. 1980, 45, 2723.
(11) Kluger, R.; Grant, A. S.; Bearne, S. L.; Trachsel, M. R. J. Org.
Chem. 1990, 55, 2864-8.
kobs ) K1k2[M]/(1 + K1[M])
(3)
The rate constants and association constants obtained from
fitting to give these plots are given in Table 2.
(12) Wodzinska, J. M., Ph.D. Thesis, University of Toronto, 1994.

Catalysis in the Hydrolysis of Alanyl Ethyl Phosphate
J. Am. Chem. Soc., Vol. 119, No. 50, 1997 12093
Scheme 4
Scheme 5
of ethyl phosphate. Where the reaction involves addition of
water, its reverse should be faster than loss of ethyl phosphate
(in any ionization state), since neutral water is a better leaving
group. Proton transfers among the reactive intermediates may
also be rate determining since the concentrations of these species
are low, even if they are associated with large rate constants.
The presence of a hydroxide-catalyzed reaction of alanyl ethyl
phosphate begins to appear in the high pH range of Figure 2.
The most likely route is addition of hydroxide to S, forming L.
Although S is somewhat less electrophilic than SH because of
the relative electron withdrawing ability of the amino and
ammonio substituents, the increased hydroxide concentration
favors the route via L at high pH.
The pH-rate profile in Figure 2 shows a pH-independent
rate in the region that corresponds to the alanyl ethyl phosphate
being present as SH (Scheme 4). The rate is consistent with
the addition of water to SH, tautomerization to form I, and
elimination (k1). In solutions with pH > pKa for loss of the
proton of the ammonium ion of SH, the reaction has a larger
water rate. This can involve a route via J or K. In acyl
phosphate esters the rate of reaction has a dependence of the
basicity of the attacking group (ânuc) of 0.9,12 indicating that
the transition state for addition resembles the intermediate. The
elimination step, as the microscopic reverse, should have a
similar transition state. Both the alkyl phosphate monoanion
and dianion are accessible in neutral solutions, so neither must
be avoided. The lesser extent of protonation of the amino group
is exactly compensated by the increase in hydroxide concentra-
tion as pH increases. Therefore, the route via J should be
favored over that via the less prevalent tautomer, K. Thus, the
rate constant for the water reaction in the upper plateau could
also give the second-order rate constant for the hydroxide
dependent hydrolysis ((3.2 ( 0.3) × 103 M-1 s-1).
There is also an acid-catalyzed route involving SH as
demanded by the data in Figure 2 that show an increase in acidic
solutions. Scheme 5 suggests the likely pathway: through
reasonable proton transfers, the intermediate can expel neutral
ethyl phosphoric acid, an excellent leaving group.
The electron-withdrawing effect of a protonated amino group
R to the carbonyl of an acyl phosphate should accelerate its
rate of hydrolysis (σi ) 0.60)13 compared to that of unsubstituted
acyl phosphates. On the other hand, the unprotonated amino
group does not enhance the reaction electronically (σi ) 0.10).13
Information in Table 3 shows where the effects of the amino
group are most significant. The rate constant for the acid-
catalyzed hydrolysis of alanyl ethyl phosphate at 25 °C is
somewhat greater than that of benzoyl methyl phosphate at 58
°C.12 For a typical activation barrier this corresponds to a factor
of about 10. The “water reaction” in acidic solution has a larger
difference (∼30-fold or ∼300 at the same temperature). The
(13) Charton, M. J. Org. Chem. 1964, 29, 1222.

12094 J. Am. Chem. Soc., Vol. 119, No. 50, 1997
Kluger et al.
Table 3. Rate Constants for the Hydrolysis Alanyl Ethyl Phosphate and Benzoyl Methyl Phosphate
acyl phosphate
conditions
kH (M-1 s-1
)
kH O (s-1
)
kOH (M-1 s-1
)
2
benzoyl methyl phosphate
alanyl ethyl phosphate
58 °C, I ) 1.0
25 °C, I ) 1.0
1.9 × 10-4
6.5 × 10-7
2.4
1.67
3.9 × 10-4
2.0 × 10-5, 1.6 × 10-3
Scheme 6
water reaction at pH > pKa of the amino group is about 2500-
fold faster at the lower temperature (or ca. 25 000 times faster
at the same temperature) for alanyl ethyl phosphate. This
suggests that the N-protonated species reacts with hydroxide.
Finally, the hydroxide dependent rate of hydrolysis of benzoyl
methyl phosphate at 58 °C exceeds that of alanyl ethyl phosphate
at 25 °C by the smallest ratio, consistent with a minimal effect
of the neutral amino group on the adjacent anion.
Formation of Amides. The free amino group of alanyl ethyl
phosphate or alanine can add to another alanyl ethyl phosphate
to form a dipeptidyl molecule. Oligomeric products can result
from further reactions of alanyl ethyl phosphate and the
dipeptide. We observe no products that retain the acyl
phosphate functionality although the amino group is available
in alanyl ethyl phosphate. The pKa of the amino group of alanyl
ethyl phosphate is 2 units lower than that of alanine (9.9).
Hence, there should be much more alanyl ethyl phosphate than
alanine in its nucleophilic amine form, but it should be a weaker
nucleophile. An alternative explanation for the absence of any
dipeptidyl phosphate would be an undetected intramolecular
reaction that forms an oxazolone14 or diketopiperazine.15 Oli-
gomerization is a second-order process and minimized in dilute
solutions, as reported for phenylalanyl adenylate.16 We have
investigated the general use of N-protected aminoacyl alkyl
phosphates as reagents for peptide synthesis in water. They
give reasonable yields, and no loss of chiral integrity can be
detected at the stereocenters.3,17 This is presumably the result
of the anionic character of the material suppressing enolate
formation.17
Reactions with Alcohols. Since alanyl ethyl phosphate reacts
slowly with water, it is not surprising that it undergoes solvolysis
slowly in methanol. The alcohol can react by a similar
mechanism with the transfer of the hydroxyl proton in the
intermediate. The less polar solvent should discourage the more
reactive ionic forms. The reaction of alanyl ethyl phosphate
with 2-propanol was not observed to proceed at all toward the
ester. This is consistent with its being even less polar and more
hindered.
to the effective rate. The phosphate anion provides a localized
site for binding. This position is not particularly effective for
promoting hydrolysis since stabilizing the carbonyl tetrahedral
intermediate and its associated transition states is most critical
(Scheme 6). If the binding were completely localized at the
phosphate anion, catalysis would likely be due to a decreased
electrostatic barrier.
The amino group may serve as a template to direct binding
in this direction. Thus, catalysis can be accomplished by a small
portion of the complex being bidentate-coordinated (to the
phosphate and oxyanion of the intermediate and related transi-
tion states).18 The use of specific ligands on the metal may
further facilitate such coordination.
Conclusions
Acid, base, and metal ions enhance the reaction of water with
alanyl ethyl phosphate. The kinetic patterns are consistent with
reactions involving intermediates that undergo proton transfers.
Reactions with alcohols are quite slow and will require the
design of catalysts in order to be useful. One potential
application for such reactions could be in producing synthetically
aminoacylated t-RNAs to make proteins with unnatural amino
acid derivatives in their backbone.19
Acknowledgment. We are grateful for the support for this
research that was provided by an operating grant from the
Natural Sciences and Engineering Research Council of Canada.
Richard Loo received a postgraduate fellowship from the same
organization.
Metal Ion Catalysis. The results in Table 2 show that one
of the major differences between metals is their affinities for
alanyl ethyl phosphate. The rate constants for the complexes
(at saturation) are not related to the affinity, but both contribute
(14) Benoiton, N. L.; Chen, F. M. F. Can. J. Chem. 1981, 59, 384.
(15) Purdie, J. E. B.; Benoiton, N. L. J. Chem. Soc., Perkin Trans. 1973,
13, 1845.
(16) Lacey, J. C., Jr.; Senaratne, N.; Mullins, D. W., Jr. Origins Life
1984, 15, 45.
JA972406Q
(18) Kluger, R.; Wong, M. K.; Dodds, A. K. J. Am. Chem. Soc. 1984,
106, 1113.
(19) Noren, C. J.; Anthony-Cahill, S. J.; Griffith, M. C.; Schultz, P. G.
Science (Washington, D.C.) 1989, 244, 182.
(17) Loo, R. W. Ph.D. Thesis, University of Toronto, 1995.