Substituent effects on energetics of peptide-carboxylate hydrogen bonds as studied by 1H NMR spectroscopy: Implications for enzyme catalysis

Article abstract of DOI:10.1021/jo401762m

Substituent effects in N-H···O hydrogen bonds were estimated by comparing the acidities of two series of model compounds: N-benzoylanthranilic acids (A) and 4-benzoylamidobenzoic acids (B). Intramolecular N-H···O hydrogen bonds were found to be present in the A series of compounds, while B acids were used as control models. The respective pK_a values for A and B acids were determined experimentally in DMSO solution using proton NMR spectroscopy. With X = H, the pK_a for A and B acids were observed to be 7.6 and 11.6, respectively, a difference of 4.0 units (ΔpK_a). However, with X = p-NO ₂, the ΔpK_a value between A and B acids increased to 4.7 units: the pK_a values for A and B acids were determined as 6.7 and 11.4, respectively. The ΔpK_a values between A and B acids as a function of the X substituents were studied in 10 other examples. The effects of X substituents in A acids could be predicted on the basis of the observed linear Hammett correlations, and the sensitivity of each substituent effect was found to be comparable to those observed for the ionization of substituted benzoic acids (ρ = 1.04 for A acids, and ρ = 1.00 for benzoic acids).

Full text of DOI:10.1021/jo401762m

Article
pubs.acs.org/joc
Substituent Effects on Energetics of Peptide-Carboxylate Hydrogen Bonds
as Studied by 1H NMR Spectroscopy: Implications for Enzyme Catalysis
Bright U. Emenike, Albert Tianxiang Liu, Elsy P. Naveo, and John D. Roberts*
Gates and Crellin Laboratories of Chemistry, California Institute of Technology, Pasadena, California 91125
S
* Supporting Information
ABSTRACT: Substituent effects in N−H···O hydrogen bonds
were estimated by comparing the acidities of two series of
model compounds: N-benzoylanthranilic acids (A) and 4-ben-
zoylamidobenzoic acids (B). Intramolecular N−H···O hydro-
gen bonds were found to be present in the A series of compounds,
while B acids were used as control models. The respective pKa
values for A and B acids were determined experimentally in
DMSO solution using proton NMR spectroscopy. With X = H,
the pKa for A and B acids were observed to be 7.6 and 11.6,
respectively, a difference of 4.0 units (ΔpKa). However, with
X = p-NO2, the ΔpKa value between A and B acids increased to
4.7 units: the pKa values for A and B acids were determined as
6.7 and 11.4, respectively. The ΔpKa values between A and B acids as a function of the X substituents were studied in 10 other
examples. The effects of X substituents in A acids could be predicted on the basis of the observed linear Hammett correlations,
and the sensitivity of each substituent effect was found to be comparable to those observed for the ionization of substituted
benzoic acids (ρ = 1.04 for A acids, and ρ = 1.00 for benzoic acids).
very energetic and have joined others in refuting the theory.9
The fact that experimental measurements of single hydrogen-
INTRODUCTION
■
Hydrogen bonding plays a crucial role in enzyme catalysis. The
generalized consensus is that the substantial rate enhancements
associated with enzyme-catalyzed biochemical transformations
partly stem from the myriad of hydrogen bonds provided in the
bond strengths in solution are typically less than 10 kcal/mol,
as previously calculated by Houk et al.,13 underscored the
premise of the LBHB theory. To the best of our knowledge,
reported free energies (ΔG) for the “strongest” O−H···O hydrogen
transition states.1,2 Such stabilization has been estimated to
lower the activation barriers of enzyme-catalyzed reactions by
10 kcal mol−1 or more. On the other hand, non-hydrogen-
bonded factors, such as electrostatic and hydrophobic effects,
also have been reported to provide enzyme transition state
stabilization.3,4 Despite the apparent importance of hydrogen
bonding during biochemical transformation, the exact mecha-
nism at the “reactive site” is still not completely clear. Later,
low-barrier hydrogen bonds (LBHBs) were postulated to be
responsible for enzyme catalysis,5 although such a claim is yet
to be substantiated. LBHBs are unique hydrogen bonds char-
acterized by strong interaction energies (more than 10 kcal/mol)
and downfield 1H NMR signals of the hydrogen-bonded
protons (15−22 ppm). Such LBHBs have been demonstrated
in many enzymes’ active site by X-ray crystallographic6 and
proton NMR studies.7 This evidence led to the suggestion that
LBHBs might be involved in enzyme catalysis.5 For instance,
in the active site of α-chymostrypsin enzyme, Ringe et al.8
reported a proton NMR chemical shift of 18.9 ppm for the low-
barrier N−H···O bond between Asp102 and His57, and the pKa
of His57 is ≥12, which is 6 units more elevated than the pKa of
the imidazole side chain of histidine in water. However, the
suggested roles of LBHBs in enzyme catalysis have been heavily
criticized.9−12 Perrin et al. have pointed out that LBHBs are not
bonds in solution are within 8−11 kcal/mol ranges.14,15
In lieu of the LBHB theory, Shan and Herschlag provided an
alternative and more compelling proposal,16 one that is based
on the stabilization attenuated from multiple hydrogen bonds
instead of single ones. Despite the modest interaction energies
of “normal” single hydrogen bonds (∼5 kcal/mol), recent study
by Kass et al.17 showed that networks of three O−H···O intra-
molecular hydrogen bonds in a simple tetraol model (Figure 1)
increased the acidity of the tertiary alcohol by a factor of 10.21
The pKa of the tetraol was determined to be 16.1 in DMSO,
whereas tert-butyl alcohol (which served as the control model)
has a known pKa value of 32.2 in DMSO.17 The difference in
acidity (ΔpKa = 16.1 units; ΔG = 21.9 kcal/mol) resulting from
the increased stabilization of the charged oxyanion can
potentially explain enzymes’ rate enhancement.
Figure 1. Kass’ tetraol model.
Received: August 15, 2013
Published: October 15, 2013
© 2013 American Chemical Society
11765
dx.doi.org/10.1021/jo401762m | J. Org. Chem. 2013, 78, 11765−11771

The Journal of Organic Chemistry
Article
Moreover, the stabilizing effect of hydrogen bonding has
been found to be nearly additive. For example, the single intra-
molecular O−H···O hydrogen bond in 2-hydroxybenzoic acid
(salicylic acid) is responsible for its reduced pKa value of 6.6 in
DMSO, which is ∼6 pKa units lower than the acidity of its
isomer (4-hydroxybenzoic acid, pKa = 12.4). On the other hand,
2,6-dihydroxybenzoic acid, which exists with two intramolecular
O−H···O hydrogen bonds has a pKa value of 3.1, which is
∼10 pKa units lower than its intrinsic value of 13.7.16
the hydrogen bond donor (−NH) using a range of electron-
donating and electron-withdrawing substituents. By placing the
substituent groups on the benzamide ring and not on the
benzoic acid ring, the emphasis of the substituents’ electronic
influence is placed on the acidities of each compound’s NH
proton. Because the carboxylic group is farther away from the
substituent than the amide group, changes in the acidities of the
carboxylic acids as a function of substituent groups should
occur primarily as a result of the plausible hydrogen bonding
between the amide and carboxyl functional groups. In order to
tests these hypotheses, we also considered a second set of para-
substituted benzoic acids as control compounds B (13−16).
Note that the differences between the acidities of ortho-
substituted benzoic acids (such as A) and para-substituted
benzoic acids (such as B) previously have been used to provide
quantitative measures of hydrogen bonding in DMSO.16,18
Intramolecular N−H···O Hydrogen Bonds in Solution.
DMSO is a polar aprotic solvent with a smaller dielectric
constant than water (i.e., ε = 46.7 vs 78.4); it has been com-
monly used as a crude model to mimic the interior of enzymes’
hydrophobic active sites. The proton NMR spectroscopic
approach was used largely because of its sensitivity in detecting
N−H···O hydrogen bonding in solution, particularly in aprotic
solvents.22 In general, for the series of model systems A (1−12),
the NH proton appeared to be broadened and centered around
13.5 ppm in DMSO, where 12 exhibited the most down-
field chemical shift (13.9 ppm), not surprisingly by having a
CF3 group directly attached to the amide carbonyl group.
In contrast, 3 was also electron attracting, but its influence was
diminished by having a 4-trifluoromethylphenyl. For the con-
trol model B (13−16), the −NH proton shifts were ∼10.5 ppm,
indicating a chemical shift difference of 3 ppm relative to model
systems A. The differences in the chemical shifts between A and
B are much more profound in the monoanions where these
compounds exist as carboxylates. Thus, for the monoanion of
These elegant examples have prompted examination of the
origin of the hydrogen-bond effects on the pKa of carboxylic acids.
In other words, what relationship can we establish between the
hydrogen-bond strengths and the ionizability of these acids? In this
1
report, H NMR spectroscopy and quantum mechanical DFT
calculations were used to probe the effects of substituent groups
on the stabilization of amide N−H···O hydrogen bonds. The imp-
ortance of substituents in amide hydrogen bonding was demon-
strated using a series of small-molecule models derived from
N-benzoylanthranilic acids (Figure 2). Substituent groups were
Figure 2. Structures of A and B series of compounds, where the
substituent (X) includes a range of electron-donating and electron-
withdrawing substituents.
found to have substantial influences on hydrogen-bond strength
by changes in the acidity of the benzoic acids.
RESULTS AND DISCUSSIONS
■
1
12, the H NMR signal of the NH proton shifted further
In previous studies, salicylic acids have been demonstrated to
be valuable model systems for investigating O−H···O intra-
molecular hydrogen bonds.16,18−20 The replacement of the
hydroxyl group with an amide has also proven beneficial in the
study of N−H···O interaction.21 The improvement of the new
series (1−12, Figure 3) results from electronegatively changing
downfield from 13.9 to 18.4 ppm, whereas that of the 13
monoanion for example, remained essentially the same (i.e.,
from 10.6 to 10.5 ppm). The further downfield chemical shifts
for the NH proton observed in 12-monoanion is typical of a
“low barrier” hydrogen bond in which the bridging proton has
Figure 3. Scheme showing model compounds A (1−12) and B (13−16).
11766
dx.doi.org/10.1021/jo401762m | J. Org. Chem. 2013, 78, 11765−11771

The Journal of Organic Chemistry
Article
become deshielded from the N electrons as a result of N−H
bond elongation (Figure 4). The apparent chemical shift dif-
ference for the NH signals in the carboxylates of 12 and 13
(∼8 ppm i.e., 18.4 vs 10.5 ppm) shows that model systems A
are stabilized by N−H···O hydrogen bond, whereas the B con-
trols are not. The comprehensive NH proton chemical shifts for
the respective monoanions of model system A are listed in Table 1.
Figure 5. Low-energy conformers and the relative free energy for
compound 12 computed at B3LYP/6-311+G(d,p)//B3LYP/6-
31+G(d) level. The selected O···H distances are calculated values in Å.
congruent with the occurrence of the 18.4 ppm downfield
1
chemical shift observed in the H NMR spectrum. In addition,
the calculated H···O distance in the 12-monoanion (1.55 Å)
was substantially shorter than those for the acid (1.82 Å for
12a, and 1.85 Å for 12b). Both the DFT calculations and the
1H NMR results seem to support the idea of strong hydrogen
bonding in model system A.
An X-ray crystal structural analysis was obtained for the
crystalline 1 (shown in Figure 6). The solid state also indicated
1
Figure 4. H NMR signal showing the NH region for the acid 12
(top), and its monoanion (bottom) in DMSO-d6 solvent.
The 18.4 ppm for the bridging NH proton is consistent with the
occurrence of LBHBs in solution and agrees with the experimental
value of 18 ppm reported for chymotrypsin and trypsin.23
Figure 6. X-ray crystal structure of 1 crystallized from CH2Cl2
Conformational Analysis for Model System A. To
further understand the structural geometry of the N−H···O
hydrogen bond, conformational search calculations were per-
formed. Molecular mechanics calculation using the Spartan
program (MMFF94) produced two low-energy conformers for
12. These conformers were subsequently reoptimized by den-
sity functional theory (DFT) methods at the B3LYP/6-
31+G(d) level17 to give two stable conformers 12a and 12b
(Figure 5). The relative free energies realized through vibra-
tional frequency calculations at the IEFPCM/B3LYP/6-
311+G(d,p) levels (solvent = DMSO)17 predicted conformer
12a to be more stable by 2.4 kcal/mol. Interestingly, similar
calculations for the monoanion (12-monoanion) revealed a
deshielded N-proton because the N−H bond was 0.06 Å longer
than that of the acid (i.e., 1.08 Å vs 1.02 Å). This finding is
solution; shown distance is in Å.
that N−H···O hydrogen bond is quite possible because the
N−O (2.68 Å) is shorter than the sum of the van der Waals
radii between a nitrogen and an oxygen atom (3.07 Å).
Therefore, the NH proton is located within the van der Waals
distance necessary for N−H···O hydrogen bonding. In addition,
the H···O distance of 2.01 Å is less than the 2.72 Å sum of the
van der Waals radii between hydrogen and oxygen atoms.
Substituent Effects on Hydrogen Bond Strength and
Acidity (pKa). Because model system A indeed appears to form
intramolecular hydrogen bonds in DMSO solutions and also
because the presence of intramolecular hydrogen bond is
expected to perturb the acidity of the benzoic acid, it appears
Table 1. (−NH) Chemical Shifts of the Monoanions, Ionization Constants (Ka), and Hydrogen Bond Interaction Free Energies
for Compounds1−12 in DMSO-d6
a
compd
R
σX
δNH (ppm)
pKa
Kobs
ΔG [kcal/mol]
1
(p-NO2)-Ph
(p-CN)-Ph
(p-CF3)-Ph
(p-Cl)-Ph
Ph
0.78
16.6
16.2
16.1
16.1
15.9
15.7
15.3
16.6
15.6
15.1
14.5
18.4
6.7 0.2
6.9 0.1
7.0 0.1
7.3 0.2
7.6 0.1
7.7 0.1
8.0 0.1
6.9 0.1
7.4 0.2
6.9 0.2
8.2 0.1
6.1 0.2
2.00 × 10−7
1.25 × 10−7
1.00 × 10−7
5.01 × 10−8
2.51 × 10−8
2.00 × 10−8
1.00 × 10−8
1.25 × 10−7
5.01 × 10−8
1.25 × 10−7
6.31 × 10−9
7.94 × 10−7
−6.7
−6.4
−6.3
−5.9
−5.5
−5.3
−4.9
−6.4
−5.9
−6.4
−4.6
−7.5
2
0.66
0.54
0.23
0.00
−0.17
−0.37
0.71
0.12
nab
3
4
5
6
(p-CH3)-Ph
(p−OH)-Ph
(m-NO2)-Ph
(m−OH)-Ph
(o−OH)-Ph
CH3
7
8
9
10
11
12
nab
nab
CF3
a
b
Obtained from ref 27. na: not applicable.
11767
dx.doi.org/10.1021/jo401762m | J. Org. Chem. 2013, 78, 11765−11771

The Journal of Organic Chemistry
Article
Figure 7. Hammett correlation of model system A.
that model system A might be a suitable system to investigate
substituent effects. To determine the pKa of the various acids
prepared, we adopted the NMR method developed by Choi,24
which is a modified version of similar procedure developed by
Perrin et al.25 The subsequently determined pKa values for
1−12 in DMSO solutions are compiled in Table 1.
In DMSO, the stabilization of the conjugate monoanions by
intramolecular hydrogen bonding acts to increase the acidity of
the neutral acids in which the experimentally observed pKa
values (pKobs) are actually lower (i.e., more acidic) than their
corresponding intrinsic values (pKint). The intrinsic pKint refers
to the pKa of the carboxylic acids in the absence of the intra-
molecular hydrogen bonds.18 The difference between pKobs
and pKint provides a quantitative measure of intramolecular
N−H···O hydrogen bond strength according to eq 1.
toward a more positively charged proton. The polarization of
the N−H bond might then lead to a stronger N−H···O hydrogen-
bond interaction. Conversely, when the substituent was an
electron-donating group (such as the OH group 7), a lowered
acidity (i.e., higher pKa of 8.0
0.1 relative to 5) was cal-
culated, which indicates that the hydrogen bond strength in 5 is
0.6 kcal/mol stronger than in 7. On the other hand, com-
pounds 11 and 12 are devoid of benzamide (aromatic) rings;
the effects of the substituent groups in these examples are most
likely transmitted by the inductive effect. Substituting the
methyl group in 11 (pKa = 8.2
0.1) for a trifluoromethyl
group as in 12 (pKa = 7.0 0.2) increased the acidity of 12 by
2.9 kcal/mol per hydrogen bond relative to 11. This represents
the most significant increase in the hydrogen bond free energy
in A series of compounds.
The energetic contributions of substituent groups might
become significant in enzyme catalysis where the transition
states can be stabilized by multiple hydrogen bonds to a single
oxygen atom.28 In this instance, if we assume that the con-
tributions of the three N−H···O hydrogen bonds to the ester
carbonyl are nearly additive, as previously proposed by Shan
and Herschlag,16 then it is possible to stabilize the transition
state by an additional ∼8.7 kcal/mol (i.e., 3 × 2.9 kcal/mol)
simply by tuning the individual hydrogen bonds with a strong
electron-withdrawing group.
One way to trace the origin of substituent effects observed in
model system A is to correlate the various hydrogen-bonding
free energies with the respective Hammett constants.29,30 The
Hammett constant σp is generally recognized as describing
the electron movement through the σ- and the π-frameworks
caused by inductive and resonance effects, respectively.31 In the
original treatment of Hammett constants, ortho-substituted
benzene derivatives were not included because of steric reasons.
However, in model system A, one can assume that, because the
carboxylates and the amide NH are hydrogen bonded, the two
functional groups can be viewed as a composite unit in which
the para substituents on the benzamide ring can electronically
control the hydrogen-bond strength/acid ionization.
ΔG = −RT ln (Kobs/Kint)
(1)
The values for the pKint were obtained by assuming that the
measured pKa of the control models (13−16) are equal to
pKint. This assumption has been validated.16,18,21,26 An expected
trend for the controls was that, in the absence of intramolecular
hydrogen bonds, changes in the substituent groups had re-
latively little influence on their acidity because their pKa values
were observed to be within 0.4 unit. The average pKa for 13−
16 was 11.6; this value was used as an approximated pKint value.
By comparing the experimental pKa for 5 (7.6 0.1) to the
averaged values for the control model (11.6), it becomes apparent
that a single amide (N−H···O) intramolecular hydrogen bond
is capable of increasing the acidity of the benzoic acid by 5.5
kcal/mol (ΔpKa = ∼4 units). This energetic value is within the
range of “moderate” hydrogen bonds.27 However, the introduc-
tion of an electron-withdrawing group on the benzamide, such
as the NO2 (1), led to an increase in acidity relative to 5. By
applying eq 1, the pKa of 6.7
0.2 for 1 corresponds to a
6.7 kcal/mol hydrogen bond stabilization of the carboxylate.
Thus, by implication, the NO2 group has increased the strength
of the single N−H···O intramolecular hydrogen bond by an
additional 1.2 kcal/mol relative to 5. As expected, the electron-
withdrawing groups can reduce the π-electron density of the
aromatic ring, which in turn should polarize the N−H bond
The values for the Hammett constants were obtained from
ref 27, and the results are shown in Figure 7. A linear free-energy
11768
dx.doi.org/10.1021/jo401762m | J. Org. Chem. 2013, 78, 11765−11771

The Journal of Organic Chemistry
Article
relationship was obtained for the Hammett plot. The best-fit line
originated from a near-zero intercept (0.04), which makes the
Hammet plot consistent with eq 2. Consequently, the sensitivity
of the substituent effect (ρ) calculated as the slope gave a value
of 1.04. By definition, such a ρ value suggests that substituent
effects in the amide hydrogen bonds are comparable to the sub-
stituent effects observed for the ionization of benzoic acids
derivatives in aqueous conditions. As a result, one can conclude
that as the electron density of the phenyl ring decreases, both the
energetics of N−H···O interaction and the ionizability of the
carboxylic acid group increase.
knowledge, no quantitative measurements have been made for
amide hydrogen bond.
Bearing in mind that 5 (with no substituents) has a pKa of
7.6 0.1, the observed pKa value of 7.4 0.2 for 9 is consistent
with the fact that meta-hydroxyl groups are actually electron-
withdrawing; hence, the Hammett constant has a positive value
(σm = +0.12). On the other hand, hydroxyl groups at the para
position (7) are strongly electron-donating (σp = −0.37),
indicating a resonance effect; therefore, the pKa of 8.0 0.1 is
in accord with the expected trend.
CONCLUSION
log[Ka(X)/Ka(H)] = σρ
■
(2)
The energy of hydrogen bonds in solution has been a central
issue when discussing the roles of hydrogen bonding in
assisting the known effects found in enzyme catalysis. In this
study, we have demonstrated that substituent groups can
bolster the strength of a single amide hydrogen bond up to 2.9
kcal/mol. This finding provides further support for the concept
of the cumulative stabilization power of hydrogen bonds as
proposed by Shan et al18 and Kass et al.17 It could be possible
that substituent groups might provide the missing link between
moderately interacting hydrogen bonds and strong hydrogen
bonds, where strong hydrogen bonds have been related to
enzyme catalysis. Although the mechanism of enzyme catalysis
has been previously linked to the elusive energetic attributes of
the so-called LBHBs, the results of this study provided no
support for “special” hydrogen bond free energy. In fact, in the
As a result of the linear Hammett plot, it appears that, for
model system A, the energetic contribution of any substituent
to N−H···O hydrogen-bond interaction, and by extension to
the acidity of the benzoic acid, can be predicted on the basis of
their Hammett values. However, we caution that such an
assumption might be an oversimplification and potentially
misleading because the free energies of noncovalent inter-
actions, especially those involving aromatic rings, tend to be
sensitive to substituent positioning.32 We note that, in addition
to electronic contribution of substituent groups, these groups
may also lead to significant secondary interactions through field
effects,31,33 or the so-called “local direct interactions”.34 For
instance, when we measured the acidity of 9 and 10 (which are
isomers of 7, shown in Figure 8), pKa values of 7.4 0.2 and
1
monoanion of 12, the appearance of a 18.4 ppm H chemical
shift for the bridging NH proton clearly signals the occurrence
of LBHB, however, the calculated hydrogen-bond free energy
(7.5 kcal/mol) is less than the proposed 10−20 kcal/mol for
LBHB.
Furthermore, because linear free energy relationships were
observed for model system A, it appears that one can make a
reasonable prediction of the substituent effect on amide
hydrogen bonds based on the calculated sensitivity value (ρ =
1.04) and Hammett σ constants. However, one should be
mindful that the relative positions of the substituents are
equally important. This point was demonstrated by comparing
the observed pKa values of 7, 9, and 10, where the formation of
an “outer” hydrogen bond in 10 seems to have strengthened
the N−H···O interaction by an additional 1.5 kcal/mol. The
findings in this work should be of particular importance to
computational enzyme designers because it is crucial to retain
the original geometry of an enzyme’s active site while trying to
find ways of increasing its binding affinity through secondary
substituent effects.
Figure 8. DFT-calculated structures at the B3LYP/6-31+G(d) levels.
The selected chemical shifts were observed in H NMR. The H···O
distance is 1.61 Å in 7, 1.59 Å in 9, and 1.58 Å in 10.
1
6.9 0.2 were determined, respectively. This implies that 10 is
about 1.5 kcal/mol more acidic than 7. The higher acidity of 10
might result from the cumulative effect of the two conjugated
hydrogen bonds. Positioning the hydroxyl group at the ortho
carbon in 10 creates the possibility of forming a second “outer”
hydrogen bond with the amide (as shown in Figure 8),
which will compliment the “inner” N−H···O hydrogen bond in
stabilizing the carboxylate. DFT calculations and proton NMR
spectroscopy provided further support for this arrangement. In
the DFT-optimized structures, the inner H···O distance in 10
was calculated to have the shortest distance, which suggests a
stronger N−H···O hydrogen bond. Furthermore, as expected,
the downfield chemical shifts of the bridging NH proton for 7,
9, and 10 were almost identical, (i.e., 15.3, 15.6, and 15.1 ppm,
respectively). However, the corresponding OH chemical shift
of 10 was 2.1 and 2.6 ppm more downfield than in 7 and 9,
respectively. With these results, it is therefore not completely
unreasonable to infer that the higher acidity of 10 (∼ 1.5 kcal/mol)
relative to 7 is due to the additional ortho hydroxyl hydrogen
bond. In fact, Kass et al. found similar effects in alcohols using
a heptaol model compound.17 However, to the best of our
EXPERIMENTAL SECTION
■
General Procedure for the Preparation of Model Com-
pounds.35 To a solution of anthranilic acid (1.1 mmol, 151 mg) in
anhydrous THF (4 mL) was 1.0 mmol of the appropriate benzoyl
chloride added at room temperature. After cooling the solution using
ice−water bath, 1.5 mmol of triethylamine was added dropwise and
reaction was stirred at room temperature for additional 4−12 h. The
mixture was poured into a 20−30 mL cold solution of 1.0 M HCl, and
precipitates were collected by filtration. Recrystallization from THF−
hexane solution afforded the desired compound in quantitative yields.
2-(4-Nitrobenzamido)benzoic acid36 (1): H NMR (600 MHz,
1
dmso) δ 13.67 (s, 1H), 12.20 (s, 1H), 8.58 (d, J = 8.3 Hz, 1H), 8.39
(d, J = 8.0 Hz, 2H), 8.14 (d, J = 8.5 Hz, 2H), 8.02 (d, J = 7.9 Hz, 1H),
7.65 (t, J = 7.8 Hz, 1H), 7.22 (t, J = 7.6 Hz, 1H). 13C NMR (151 MHz,
dmso) δ 170.3, 163.5, 149.9, 140.8, 140.5, 134.7, 131.7, 129.0, 124.6,
11769
dx.doi.org/10.1021/jo401762m | J. Org. Chem. 2013, 78, 11765−11771

The Journal of Organic Chemistry
Article
124.1, 120.8, 117.8. HRMS (EI-TOF) m/z calcd for C14H10O5N2
[M+] 286.0590, found 286.0590
(d, J = 1.3 Hz, 1H), 7.26−6.99 (m, 1H), 2.12 (s, 3H). 13C NMR (126
MHz, dmso) δ 169.9, 168.9, 141.3, 134.4, 131.5, 123.0, 120.4, 116.9,
25.4.
2-(4-Cyanobenzamido)benzoic acid37 (2): 1H NMR (600
MHz, dmso) δ 12.16 (s, 1H), 8.58 (d, J = 8.3 Hz, 1H), 8.22−7.86
(m, 4H), 7.62 (t, J = 7.8 Hz, 1H), 7.19 (t, J = 7.5 Hz, 1H). 13C NMR
(151 MHz, dmso) δ 170.3, 163.7, 140.9, 138.9, 134.7, 133.4, 131.7,
128.3, 123.9, 120.7, 118.6, 117.6, 114.8. HRMS (EI-TOF) m/z calcd
for C15H11N2O3 [M + H]+ 267.0770, found 267.0777
2-(Trifluoroacetamido)benzoic acid40 (12): 1H NMR (600
MHz, dmso) δ 13.67 (s, 1H), 12.43 (s, 1H), 8.24 (d, J = 8.1 Hz, 2H),
8.01 (d, J = 7.6 Hz, 2H), 7.66 (t, J = 7.5 Hz, 2H), 7.31 (t, J = 7.4 Hz,
2H). 13C NMR (126 MHz, dmso) δ 169.8, 154.7 (m, J = 36.6 Hz),
138.0, 134.7, 131.7, 125.8, 121.6, 119.5, 116.0 (m, J = 288.9 Hz).
4-(4-Nitrobenzamido)benzoic acid43 (13): 1H NMR (500
MHz, dmso) δ 12.79 (s, 1H), 10.83 (s, 1H), 8.38 (d, J = 8.9 Hz,
1H), 8.20 (d, J = 8.9 Hz, 1H), 7.94 (dd, J = 23.7, 8.9 Hz, 2H). 13C
NMR (126 MHz, dmso) δ 167.3, 164.8, 151.0−149.8, 143.2, 140.7,
130.8, 129.8, 126.5, 124.0, 120.1.
2-(4-(Trifluormethyl)benzamido)benzoic acid38 (3): H NMR
1
(500 MHz, dmso) δ 12.24 (s, 1H), 8.66 (d, J = 7.7 Hz, 1H), 8.15 (d,
J = 7.8 Hz, 2H), 8.07 (dd, J = 7.9, 1.5 Hz, 1H), 7.99 (d, J = 6.9 Hz,
2H), 7.69 (t, J = 7.9 Hz, 1H), 7.25 (t, J = 7.6 Hz, 1H). 13C NMR (126
MHz, dmso) δ 170.4, 164.0, 141.1, 138.8, 134.8, 132.4, 132.2, 131.7,
128.5, 126.5, 123.9, 120.7, 117.6. HRMS (EI-TOF) m/z calcd for
C15H10F3NO3 [M + H]+ 310.0691, found 310.0701
4-Benzamidobenzoic acid44 (14): 1H NMR (500 MHz, dmso) δ
10.53 (s, 1H), 7.99−7.95 (m, 1H), 7.93 (d, J = 2.9 Hz, 1H), 7.64−7.60
(m, 1H), 7.57−7.52 (m, 1H). 13C NMR (126 MHz, dmso) δ 167.4,
166.4, 143.8, 135.8−135.2(m), 132.3, 130.7, 128.9, 128.2, 125.9, 119.9.
4-Acetamidobenzoic acid:39 (15): 1H NMR (500 MHz, dmso) δ
10.23 (s, 1H), 7.88 (d, J = 8.8 Hz, 2H), 7.69 (d, J = 8.8 Hz, 2H), 2.08
(s, 3H). 13C NMR (126 MHz, dmso) δ 169.3, 167.4, 143.8, 130.8,
125.3, 118.6, 24.6.
2-(4-Cholorobenzamido)benzoic acid36 (4): 1H NMR (300
MHz, dmso) δ 13.65 (s, 1H), 12.14 (s, 2H), 8.64 (dd, J = 8.4, 1.0 Hz,
2H), 8.03 (dd, J = 7.9, 1.6 Hz, 2H), 7.93 (d, J = 8.7 Hz, 5H), 7.69−
7.58 (m, 7H), 7.19 (td, J = 8.0, 1.2 Hz, 2H). 13C NMR (126 MHz,
dmso) δ 170.5, 164.1, 141.3, 137.5, 134.7, 133.7, 131.7, 129.5, 129.4,
123.6, 120.4, 117.2. HRMS (EI-TOF) m/z calcd for C14H10O3NCl
[M]+ 275.0349, found 275.0348.
4-(2,2,2-Trifluoroacetamido)benzoic acid45 (16): 1H NMR
(300 MHz, dmso) δ 13.20−12.80 (br s, 1H), 11.51 (s, 1H), 7.88 (dd,
J = 50.8, 8.2 Hz, 6H).13C NMR (126 MHz, dmso) δ 167.1, 155.2 (m),
140.8, 130.8, 128.02, 121.4−121.0 (m), 116.1 (m).
2-Benzamidobenzoic acid36 (5): H NMR (600 MHz, dmso) δ
1
13.75 (s, 1H), 12.15 (s, 1H), 8.68 (dd, J = 8.4, 1.1 Hz, 1H), 8.03 (dd,
J = 7.9, 1.6 Hz, 1H), 7.92 (dd, J = 7.0, 1.5 Hz, 2H), 7.67−7.59 (m,
2H), 7.56 (ddd, J = 6.7, 4.5, 1.3 Hz, 2H), 7.22−7.14 (m, 1H). 13C
NMR (151 MHz, dmso) δ 170.4, 165.1, 141.6, 135.0, 134.8, 132.6,
131.7, 129.7, 129.4, 129.0, 127.4, 123.4, 120.3, 116.9.
Measurement of Ionization Constants in DMSO. In a typical
running, equimolar amounts of a test acid AH (in this case 1−11 with
unknown pKa) and a conjugate acid B (a standard acid BH with
known pKa) are combined in solution in an NMR tube. The resulting
changes in the chemical shifts of the reporter nucleus (1H or 13C) were
used to judge the position of equilibrium between the two acids.
Because this method is a relative measure of the acids’ strengths
between AH and BH, the chemical shifts of the two acids (δAH and
δBH) and their corresponding monoanions (δA and δB) must be
known. Equation 3 describes the relationship between the chemical
shifts and the ionization constants of the two acids used. The
unknowns in eq 3 are the equilibrium chemical shifts for when AH and
B− or A− and BH are mixed together, i.e., δa and δb, respectively. The
advantage of this technique over others are that (1) only a single
measurement, which can be repeated many times if desired, is required
in order to determine the acidity (Ka) of the unknown compound, and
(2) the standard measurement approach ensures that the ionization
equilibrium is restricted only to the first deprotonation step because
the model systems used in this study have more than one readily
ionizable proton. The best results for using eq 3 are obtained when
AH and BH are separated by ∼2 pKa units. p-Bromobenzoic acid
(pKa = 10.5), p-nitrobenzoic acid (pKa = 9.1), and salicylic acid (pKa =
6.6) were used as the standard BH acids in DMSO solution.
2-(4-Methylbenzamido)benzoic acid39 (6): 1H NMR (500
MHz, dmso) δ 12.15 (s, 1H), 8.73 (d, J = 8.1 Hz, 1H), 8.06 (d, J = 7.9
Hz, 1H), 7.85 (d, J = 8.1 Hz, 2H), 7.65 (t, J = 7.9 Hz, 1H), 7.38 (d, J =
7.9 Hz, 2H), 7.19 (t, J = 7.6 Hz, 1H), 2.39 (s, 4H). 13C NMR (126
MHz, dmso) δ 170.5, 165.0, 142.7, 141.8, 134.8, 132.2, 131.7, 129.9,
127.5, 123.2, 120.2, 116.7, 21.5.
2-(4-Hydroxybenzamido)benzoic acid36 (7): 1H NMR (500
MHz, dmso) δ 13.67 (s, 1H), 12.06 (s, 2H), 10.25 (s, 2H), 8.72 (d, J =
8.4 Hz, 2H), 8.05 (dd, J = 7.9, 1.5 Hz, 1H), 7.82 (d, J = 8.7 Hz, 2H),
7.72−7.61 (m, 1H), 7.23−7.10 (m, 1H), 6.92 (d, J = 8.7 Hz, 2H).13C
NMR (126 MHz, dmso) δ 170.6, 164.9, 161.5, 142.0, 134.7, 131.7,
129.6, 125.5, 122.9, 120.1, 116.6, 116.0. HRMS (EI-TOF) m/z calcd
for C14H12NO4 [M + H]+ 258.0766, found 258.0761
2-(3-Nitrobenzamido)benzoic acid40 (8): H NMR (500 MHz,
1
dmso) δ 12.33 (s, 9H), 8.84−8.81 (m, 1H), 8.72 (t, J = 1.9 Hz, 5H),
8.61 (dd, J = 8.4, 1.0 Hz, 6H), 8.54 (ddd, J = 7.8, 1.6, 1.0 Hz, 1H),
8.49−8.45 (m, 7H), 8.35 (ddd, J = 7.8, 1.7, 1.0 Hz, 6H), 8.17−8.14
(m, 1H), 8.04 (dd, J = 8.0, 1.5 Hz, 6H), 7.96 (ddd, J = 8.1, 7.3, 1.5 Hz,
1H), 7.91−7.85 (m, 7H), 7.67 (ddd, J = 8.5, 7.4, 1.7 Hz, 7H), 7.24
(ddd, J = 7.9, 7.4, 1.2 Hz, 6H). 13C NMR (126 MHz, dmso) δ 170.5,
162.9, 158.9, 154.9, 148.5 (d, J = 10.3 Hz), 146.2, 140.9, 136.3, 133.8,
131.7, 131.2, 127.1, 124.0, 122.7−122.2 (m), 120.7, 117.7. HRMS (EI-
TOF) m/z calcd for C14H10O5N2 [M + H]+ 286.06680, found
286.0671.
Ka(δ − δA)(δBH − δb) = Kb(δb − δB)(δAH − δ )
(3)
a
a
Preparation of Samples in DMSO. The monoanion salts were
prepared by mixing a CH3OH solution of the appropriate acid with
1 equiv of tetrabutylammonium hydroxide solution (1.0 M in
CH3OH) respectively. Both H2O and CH3OH were subsequently
removed by evaporation under reduced pressure, followed by heating
the sample (∼70 °C) for 1−2 h under high vacuum to yield solids. In
cases where an oily product persists after heated drying, the product is
dissolved in minimum THF, and the solid product is precipitated with
hexanes. Dried samples were then stored in a positive pressure glove-
box for subsequent use. The solute concentration used for NMR
analysis in 99.9% DMSO-d6 was about 0.01 M or less for all samples.
Spectra Measurements. NMR spectra were recorded with a
300 MHz NMR spectrometer using the default pulse sequence in the
software. Typical spectra parameters for 1H spectra: 16 scans, spectral
width 9600 Hz, relaxation delay of 1 s, and acquisition time of 4 s.
Unless otherwise stated, all measurements were at a regulated tempera-
ture of 25 °C.
2-(3-Hydroxybenzamido)benzoic acid41 (9): 1H NMR (500
MHz, dmso) δ 12.15 (s, 1H), 9.98 (s, 1H), 8.71 (dd, J = 8.4, 0.9 Hz,
1H), 8.05 (dd, J = 7.9, 1.6 Hz, 1H), 7.65 (ddd, J = 8.6, 7.4, 1.6 Hz,
1H), 7.43−7.34 (m, 3H), 7.24−7.15 (m, 1H), 7.08−6.99 (m, 1H). 13C
NMR (126 MHz, dmso) δ 170.4, 165.2, 158.3, 141.6, 136.4, 134.7,
131.7, 130.5, 123.3, 120.2, 119.7, 117.7, 117.0−116.4 (m), 114.5.
HRMS (EI-TOF) m/z calcd for C14H12NO4 [M + H]+ 258.0766,
found 258.0764.
2-(2-Hydroxybenzamido)benzoic acid42 (10): H NMR (500
1
MHz, cd3od) δ 8.75 (dd, J = 8.5, 1.0 Hz, 1H), 8.14 (dd, J = 8.0, 1.6 Hz,
1H), 7.84 (dd, J = 8.4, 1.6 Hz, 1H), 7.61 (ddd, J = 8.6, 7.4, 1.7 Hz,
1H), 7.45 (ddd, J = 8.3, 7.4, 1.6 Hz, 1H), 7.19 (ddd, J = 8.1, 7.3, 1.2
Hz, 1H), 6.99−6.92 (m, 2H). 13C NMR (126 MHz, cd3od) δ 170.3,
168.5, 161.1, 140.7, 134.1, 133.9, 131.3, 126.8, 122.9, 120.4, 118.9,
117.7, 116.5, 115.6. HRMS (EI-TOF) m/z calcd for C14H12NO4 [M +
H]+ 258.0766, found 258.0765.
Theoretical Calculations. Conformational distributions were
performed using the SPARTAN ‘08 program. The selected pre-
optimized conformers were then fully optimized by DFT methods
2-Acetamidobenzoic acid40 (11): 1H NMR (500 MHz, dmso) δ
11.05 (s, 1H), 8.55−8.38 (m, 1H), 7.96 (dd, J = 7.9, 1.6 Hz, 1H), 7.56
11770
dx.doi.org/10.1021/jo401762m | J. Org. Chem. 2013, 78, 11765−11771

The Journal of Organic Chemistry
Article
using the Gaussian09 package.46 All energy minima were verified to
have 0 imaginary frequencies by vibrational frequency analysis.
(20) Nagy, P. I.; Dunn, W. J.; Alagona, G.; Ghio, C. J. Phys. Chem.
1993, 97, 4628.
(21) Shi, Z. S.; Krantz, B. A.; Kallenbach, N.; Sosnick, T. R.
Biochemistry 2002, 41, 2120.
ASSOCIATED CONTENT
■
(22) Rudner, M. S.; Jeremic, S.; Petterson, K. A.; Kent, D. R.; Brown,
K. A.; Drake, M. D.; Goddard, W. A.; Roberts, J. D. J. Phys. Chem. A
2005, 109, 9076.
(23) Markley, J. L. Biochemistry 1978, 17, 4648.
(24) Choi, P. J.; Petterson, K. A.; Roberts, J. D. J. Phys. Org. Chem.
2002, 15, 278.
(25) Perrin, C. L.; Fabian, M. A. Anal. Chem. 1997, 69, 108.
(26) Pinto, S. S.; Diogo, H. P.; Guedes, R. C.; Cabral, B. J. C.; da
Piedade, M. E. M.; Simoes, J. A. M. J. Phys. Chem. A 2005, 109, 9700.
(27) Mariam, Y. H.; Musin, R. N. J. Phys. Chem. A 2008, 112, 134.
(28) Zhang, Y. K.; Kua, J.; McCammon, J. A. J. Am. Chem. Soc. 2002,
124, 10572.
(29) Hammett, L. P. J. Am. Chem. Soc. 1937, 59, 96.
(30) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165.
(31) Roberts, J. D.; Moreland, W. T. J. Am. Chem. Soc. 1953, 75,
2167.
(32) Gung, B. W.; Emenike, B. U.; Alverez, C. N.; Rakovan, J.;
Kirschbaum, K.; Jain, N. Tetrahedron Lett. 2010, 51, 1648.
(33) Roberts, J. D.; Carboni, R. A. J. Am. Chem. Soc. 1955, 77, 5554.
(34) Wheeler, S. E.; Houk, K. N. J. Am. Chem. Soc. 2008, 130, 10854.
(35) Wiklund, P.; Bergman, J. Tetrahedron Lett. 2004, 45, 969.
(36) Prashanth, M. K.; Revanasiddappa, H. D. Med. Chem. Res. 2013,
22, 2665.
(37) Shultz, P.; Bouchez, L. PCT Int. Appl. WO/2012/129562 A2
20120927, 2012.
(38) Hamprecht, G.; Rohr, W.; Varwig, J.; Wuerzer, B. Ger. Offen
DE 3308239 A1 19840913, 1984
S
* Supporting Information
Cartesian coordinates of the optimized structures used in the
theoretical conformational analysis and the crystallographic
data (cif) for compound 1. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*robertsj@caltech.edu
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Acknowledgement is made to the National Science Foundation
under Grants CHE-0841550, TG-CHE1000106, and to the
donors of the Petroleum Research Fund administered by the
American Chemical Society for support for this research. Other
important support came from the Summer Undergraduate
Research Fellowship Program (SURF) at the California
Institute of Technology, the Senior Scientist Mentor Program
of the Camille and Henry Dreyfus Foundation, and the
NORAC Grant to Caltech by Dr. and Mrs, Chester M.
McCloskey. We are indebted to Merck and Company, Dr.
David J. Mathre, and Edith M. Roberts for their helpful
financial assistance. The facilities of the MCS used in these
studies were established with grants from DURIP-ONR and
DURIP-ARO, with additional support from ONR, ARO, NSF,
NIH, DOE, Chevron, Nissan, Dow Corning, Intel, Pfizer,
Boehringer-Ingelheim, and Sanofi-Aventis.
(39) Zentmyer, D. T.; Wagner, E. C. J. Org. Chem. 1949, 14, 967.
(40) Errede, L. A.; Oien, H. T.; Yarian, D. R. J. Org. Chem. 1977, 42,
12.
(41) Sinreih, M.; Sosic, I.; Beranic, N.; Turk, S.; Adeniji, A. O.;
Penning, T. M.; Rizner, T. L.; Gobec, S. Bioorg. Med. Chem. Lett. 2012,
22, 5948.
(42) Hauteville, M.; Ponchet, M.; Ricci, P.; Favrebonvin, J. J.
Heterocycl. Chem. 1988, 25, 715.
(43) Abbel, R.; Frey, H.; Schollmeyer, D.; Kilbinger, A. F. M.
Chem.Eur. J. 2005, 11, 2170.
REFERENCES
■
(1) Jencks, W. P. Adv. Enzymol. Relat. Areas Mol. Biol. 1975, 43, 219.
(2) Jencks, W. P. Acc. Chem. Res. 1976, 9, 425.
(3) Reed, G. H.; Poyner, R. R.; Larsen, T. M.; Wedekind, J. E.;
Rayment, I. Curr. Opin. Struct. Biol. 1996, 6, 736.
(4) Richard, J. P. Biochemistry 2012, 51, 2652.
(5) Gerlt, J. A.; Kreevoy, M. M.; Cleland, W. W.; Frey, P. A. Chem.
Biol. 1997, 4, 259.
(44) Zheng, C. H.; Yang, H.; Zhang, M.; Lu, S. H.; Shi, D.; Wang, J.;
Chen, X. H.; Ren, X. H.; Liu, J.; Lv, J. G.; Zhu, J.; Zhou, Y. J. Bioorg.
Med. Chem. Lett. 2012, 22, 39.
(45) Ban, H.; Gavrilyuk, J.; Barbas, C. F. J. Am. Chem. Soc. 2010, 132,
1523.
(46) Frisch, M. J.; et al. Gaussian09; Gaussian, Inc.: Wallingford CT,
2009.
(6) Usher, K. C.; Remington, S. J.; Martin, D. P.; Drueckhammer, D.
G. Biochemistry 1994, 33, 7753.
(7) Gu, Z. T.; Drueckhammer, D. G.; Kurz, L.; Liu, K.; Martin, D. P.;
McDermott, A. Biochemistry 1999, 38, 8022.
(8) Cassidy, C. S.; Lin, J.; Frey, P. A. Biochemistry 1997, 36, 4576.
(9) Perrin, C. L. Acc. Chem. Res. 2010, 43, 1550.
(10) Guthrie, J. P. Chem. Biol. 1996, 3, 163.
(11) Warshel, A.; Sharma, P. K.; Kato, M.; Xiang, Y.; Liu, H. B.;
Olsson, M. H. M. Chem. Rev. 2006, 106, 3210.
(12) Schutz, C. N.; Warshel, A. Proteins: Struct., Funct., Bioinf. 2004,
55, 711.
(13) Chen, J. G.; McAllister, M. A.; Lee, J. K.; Houk, K. N. J. Org.
Chem. 1998, 63, 4611.
(14) Kolonko, K. J.; Reich, H. J. J. Am. Chem. Soc. 2008, 130, 9668.
(15) Kato, Y.; Toledo, L. M.; Rebek, J. J. Am. Chem. Soc. 1996, 118,
8575.
(16) Shan, S. O.; Herschlag, D. J. Am. Chem. Soc. 1996, 118, 5515.
(17) Shokri, A.; Abedin, A.; Fattahi, A.; Kass, S. R. J. Am. Chem. Soc.
2012, 134, 10646.
(18) Shan, S. O.; Loh, S.; Herschlag, D. Science 1996, 272, 97.
(19) Mori, N.; Asano, Y.; Irie, T.; Tsuzuki, Y. Bull. Chem. Soc. Jpn.
1969, 42, 482.
11771
dx.doi.org/10.1021/jo401762m | J. Org. Chem. 2013, 78, 11765−11771