Evidence for Ion-Templation during Macrocyclooligomerization of Depsipeptides

Article abstract of DOI:10.1021/jacs.7b13148

The ion-mediated Mitsunobu macrocyclooligomerization (M-MCO) reaction of hydroxy acid depsipeptides provides small collections of cyclic depsipeptides with good mass recovery. The approach can produce good yields of a single macrocycle or provide rapid access to multiple oligomeric macrocycles in good overall yield. While Lewis acidic alkali metal salts are known to play a role in the outcome of MCO reactions, it is unclear whether their effect is due to an organizational (e.g., templating) mechanism. Isothermal titration calorimetry (ITC) was used to study macrocycle-metal ion binding interactions, and this report correlates these thermodynamic measurements to the (kinetically determined) size distributions of depsipeptides formed during a Mitsunobu-based macrocyclooligomerization (MCO). Key trends have been identified in quantitative metal ion-cyclic depsipeptide binding affinity (K_a), enthalpy of binding (ΔH), and stoichiometry of complexation across discrete series of macrocycles, and they provide the first analytical platform to rationally select a metal-ion template for a targeted size regime of cyclic oligomeric depsipeptides.

Full text of DOI:10.1021/jacs.7b13148

Article
pubs.acs.org/JACS
Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
Evidence for Ion-Templation During Macrocyclooligomerization of
Depsipeptides
Suzanne M. Batiste and Jeffrey N. Johnston*
Department of Chemistry and Vanderbilt Institute of Chemical Biology, Vanderbilt University, Nashville, Tennessee 37235-1822,
United States
S
* Supporting Information
ABSTRACT: The ion-mediated Mitsunobu macrocyclooligomeriza-
tion (M-MCO) reaction of hydroxy acid depsipeptides provides small
collections of cyclic depsipeptides with good mass recovery. The
approach can produce good yields of a single macrocycle or provide
rapid access to multiple oligomeric macrocycles in good overall yield.
While Lewis acidic alkali metal salts are known to play a role in the
outcome of MCO reactions, it is unclear whether their effect is due to
an organizational (e.g., templating) mechanism. Isothermal titration
calorimetry (ITC) was used to study macrocycle−metal ion binding
interactions, and this report correlates these thermodynamic
measurements to the (kinetically determined) size distributions of
depsipeptides formed during a Mitsunobu-based macrocyclooligome-
rization (MCO). Key trends have been identified in quantitative metal
ion-cyclic depsipeptide binding affinity (Ka), enthalpy of binding
(ΔH), and stoichiometry of complexation across discrete series of macrocycles, and they provide the first analytical platform to
rationally select a metal-ion template for a targeted size regime of cyclic oligomeric depsipeptides.
The development of an expedient synthesis19 of collections
of natural product-like cyclodepsipeptides, particularly one that
includes a means to manipulate complex structural elements
(i.e., substituents and configuration), could be a powerful
INTRODUCTION
■
Structurally complex and chemically diverse cyclic peptide and
depsipeptide natural products are attractive scaffolds in the
development of sensors,1 new materials,2 and therapeutics.3
platform to explore chemical diversity for a broad range of
Although synthetic methods using coupling reagents have been
developed to produce these molecules and their derivatives on
specific applications.10,20−23 We reported an initial solution that
includes a hybridization of enantioselective synthesis, umpo-
a commercial scale, the waste generated by their preparation
has long been recognized as a necessary evil.4 Additionally, the
lung amide synthesis (UmAS), and an (apparent) ion-
templated macrocyclooligomerization (MCO). A Mitsunobu
synthesis of macrocyclic peptides and depsipeptides typically
suffers from the disadvantages associated with the preparation
of their linear precursors. To overcome the tedium associated
reaction was adopted as the key transform for both
oligomerization and macrocyclization processes. This stereo-
specific reaction between a carboxylic acid nucleophile and
alcohol electrophile benefits from an irreversible substitution
with a linear approach and solid-phase synthesis, an
oligomerization/macrocyclization approach has been used.5−10
and is therefore inherently kinetically governed. A key feature
Although oligomerizations can greatly decrease step count, they
of this discovery, however, was that ionic salts affect the overall
production efficiency of the MCO, as well as the ring-size
selectivity for arrays of macrocyclic products. We wondered,
however, whether the nature of the salt could be the basis for
predicting how the distribution of ring sizes will change,
presumably as an ion templates the process. The notable
amplification of various ring sizes highlighted in Figure 2A led
to the hypothesis that size selectivity could be correlated to the
binding interactions of each macrocycle with its most suitable
templating alkali metal ion (Figure 2B).
are often low-yielding and size-nonselective. Because cyclic
peptides and depsipeptides tend to be ionophoric in
nature,11−13 metal salt templates have been added to
macrocyclization reactions based on the hypothesis that the
salts can bind their linear precursors and cause a conforma-
tionally enhanced rate of cyclization.14−16 This hypothesis has
been extended to cyclooligomerization reactions in order to
favor different size distributions of cyclic peptide17,18 (Figure 1)
and depsipeptide19 products, but absent from the field is a
broadly effective and rational approach, coupled to a rigorous
analytical platform that might rationalize the product size
distributions. The ultimate goal would be the development of a
tool to predictably select a template for a particular size regime.
Received: December 12, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.7b13148
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 1. Templated cyclooligomerization approaches for peptide synthesis.
Figure 2. (A) Mitsunobu-based MCO highlights and (B) strategy for measurement-based macrocycle synthesis.
The relationship between amplification and template binding
is commonly explored in thermodynamically controlled macro-
cyclooligomerization for the creation of dynamic combinatorial
libraries (DCLs).24−31 In a thermodynamically driven32
templated oligomerization, the product distribution depends
on the relative stabilities of the product−template complexes. A
template that selectively binds one product can shift the
equilibrium toward that species in high yields. However,
optimization of an irreversible (kinetic) templated oligomeriza-
tion reaction is quite different because the product distribution
is dependent on multiple rates, including the template’s role in
transition state stabilization leading to each product.33−35
A
kinetic template could bind early intermediates (lower
oligomers) to slow their cyclization while binding later
intermediates (higher oligomers) to favor a conformation that
undergoes more rapid cyclization. The driving force behind
faster cyclization to a select product in a kinetic templating
reaction is that the template almost invariably binds the product
more strongly than the precursor−template complex, so the
product is also favored thermodynamically.36 Therefore,
B
DOI: 10.1021/jacs.7b13148
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 3. Normalized amplification factors (AFn) induced by ion templates. Significant changes (|AFn| > 0.09) are shown in bold. See Figures 4-6 for
data.
studying the relative binding properties of a range of templates
to the products of a reaction is a practical way to infer their
templating behavior.36,37
solvent mixtures, such as benzene, dichloromethane/acetoni-
trile (3:7), or acetonitrile, led to the observation that the salts
alone and/or salt−macrocycle complex were not sufficiently
soluble. As a result, consistent ITC data could not be obtained.
Methanol is starkly different in polarity to benzene (solvent for
MCO), and it is known that solvent does affect binding
affinity.38,43,44 However, several accounts have shown that,
although binding constants do vary with solvent, the trend of
their relative binding is consistent.45−48 Our goal was to
observe, if possible, trends in binding interactions as a function of
macrocycle and cation size in an attempt to rationalize the
apparent MCO ion-templating effect. The determination of
binding constants in the reaction solvent (benzene) was
therefore deemed less important.
Normalized Amplification Factor (AFn) Analysis. To
correlate trends in binding interactions with changes in
macrocycle size distribution, we needed to establish a term to
determine which template-induced changes are significant.
Normalized amplification factor (AFn) is a unitless value
between −1 (complete disappearance of product) and 1
(maximum possible product) to determine significance of
increase in yield of individual products in the presence of a
template, relative to the control and others formed in the
reaction.49−51 This term is generally used to analyze DCL
product mixtures because it accounts for the large size regime
of possible products. To compare amplification factors between
different products, the concentrations of the products are
normalized on the basis of the maximum concentration of
available monomers. Because AFn is based on available
monomer concentration, the term can be applied to ion-
templated MCO when the available monomer concentration is
normalized to represent the isolated yield of macrocycles in the
reaction.30,31 The AFn values for previously reported collections
of macrocycles in the presence of metal ion templates are
shown in Figure 3. To account for experimental variance,
Several methods are commonly used to elucidate the
behavior of a metal complex in solution, such as UV−vis
spectroscopy, NMR spectroscopy, soft-ionization MS spec-
trometry, and circular dichroism (CD) spectroscopy. However,
these analytical methods are not always applicable, especially in
the analysis of complexes that are disorderly, fragile, insoluble,
paramagnetic, and/or have multiple binding modes.38 In
contrast, isothermal titration calorimetry (ITC) is a very
powerful tool for studying the thermodynamics of a binding
event. The endothermic or exothermic heat is directly
measured from the reaction, thus enabling elucidation of an
association constant (Ka) and the enthalpy (ΔH) of the
reaction. Additionally, the ion/macrocycle stoichiometry of
complexation (N) is determined by the inflection point of the
resulting titration curve. Determination of these three
parameters may provide a better thermodynamic character-
ization under optimal macrocycle−metal binding condi-
tions.26,39 Because product formation in the ion-templated
MCO is irreversible, the ability of the different metal cations to
facilitate the efficient synthesis of particular product size
distributions indicates that the transition states of their
templated linear precursors have geometries40−42 that closely
resemble the products. Thus, we speculated that the
thermodynamic macrocycle−metal binding data obtained
from ITC experiments might be correlated to the kinetic
phenomena in the ion-templated MCO.33,37,39
RESULTS AND DISCUSSION
■
Experiment Design. The ion-templated MCO by
Mitsunobu carbon−oxygen bond formation to create the
ester linkage(s) is carried out in benzene, but the ITC
experiments described here were conducted in methanol.
Attempts to prepare ITC runs in other organic solvents and
C
DOI: 10.1021/jacs.7b13148
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 4. Correlation of ITC data with the effect of salts on collections of macrocycles formed using 2. Significantly increased yields are shown in
bold. ITC binding affinities <1 × 103 M−1 are not detected.
Figure 5. Correlation of ITC data with the effect of salts on collections of macrocycles formed using 1. Significantly increased yields are shown in
bold. ITC binding affinities <1 × 103 M−1 are not detected.
template-induced changes are deemed significant when |AFn| >
0.09.49,52,53
cycles formed using tetradepsipeptide 2 and didepsipeptide 1
were determined by ITC (Figures 4 and 5). The data from ITC
can be used to rationalize why different amounts of 24- and 36-
membered rings are produced from tetradepsipeptide mono-
mer 2 and didepsipeptide monomer 1 in the presence of the
same ion templates. Under salt-free conditions, tetradepsipep-
tide 2 inherently produces cyclodimer 4 as the major product,
along with 36-membered ring 6 and 60-membered ring 7 in
Correlation of ITC Measurements with MCO Salt
Effects. In the ion-templated MCO approach to verticilide,
ring-size selectivity largely varied with monomer size, even
though the same D,L-24- and 36-membered rings are formed
from each monomer. To better understand this difference in
behavior, salt−macrocycle binding interactions of the macro-
D
DOI: 10.1021/jacs.7b13148
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 6. Correlation of ITC data with the effect of salts on collections of macrocycles formed from 8. Significantly increased yields are shown in
bold. ITC binding affinities <1 × 103 M−1 are not detected.
minor amounts. However, in the presence of Na+, 4 is formed
exclusively. As expected, 4 strongly and favorably binds Na+ in a
1:1 stoichiometry. However, 6 and 7 also bind Na+, but in 2:1
ion/macrocycle ratios. The selectivity observed in the reaction
despite the measured Na+ binding ability of the other two
potential products can be explained by a positive templating
effect:37,33 Na+ promotes the preorganization14,26 of two
building blocks (2), driven by favorable binding to the linear
precursor of 4 in a conformation that significantly accelerates
an inherently favorable ring closure. Additionally, a combina-
tion of the inherently slow oligomerization of 2 and
preorganization of a two-Na+ ion cyclization template can
explain why oligomerization/cyclization to 6 is much slower
than that to 4. Similar logic can be applied to arrive at a reason
why didepsipeptide 1 did not provide the same results with
Na+: the rate of intramolecular (Mitsunobu) cyclization of 1 is
much faster than that for 2, so although cyclization to 4 is
greater than that without salt conditions, the concentrated
reaction conditions and more reactive monomer (1) render
intermolecular coupling a more competitive process than in the
former example. Additionally, 18-membered ring 3 and 30-
membered ring 5 were found to exist in an equilibrium of 1:2
Na+-to-macrocycle aggregates and 2:1 Na+-to-macrocycle
complexes, averaging 1.5 bound ions.54,55 The observation of
a 1:2 binding pattern with 3 can imply that Na+ promotes the
preorganization of two linear precursors, which then rapidly
undergo cyclodimerization to the 36-membered ring (6) with
the help of a second Na+ ion. Because 5 is substantially larger
than 3, formation of a 1:2 metal−linear peptide complex is
more likely to result in stable aggregates, as opposed to the
precursor of a cyclodimerization reaction. Thus, the ITC
binding pattern observed with 5, which has the strongest 2-Na+
binding constant, can imply that cyclization of the linear
precursor is slowed by the formation of aggregates but is then
accelerated when the complex picks up an additional Na+ ion.
Moving to a slightly larger cation, K+, with tetradepsipeptide
2, the 36-membered ring is now significantly increased relative
to the control, while the 24-membered ring is decreased. The
36-membered ring has the greatest enthalpically favorable 1:1
binding to K+ but is not produced selectively, unlike the 24-
membered ring with Na+. This can be explained by the overall
efficacy of K+ as a cyclization template. An effective template
must have the capacity to bind early intermediates in the
oligomerization but also prevent them from competitively
cyclizing.33,37 In this case, the untemplated cyclodimerization
reaction is very fast, and addition of K+ slows it, but not enough
to selectively favor formation of the cyclotrimer (6). However,
addition of K+ to the MCO with 1 produces the 36-membered
ring in nearly double its original amount. In conjunction with
the inherent fast intermolecular oligomerization of monomer 1,
and strong binding to K+, the rate of 36-membered ring
cyclization can be selectively increased. ITC indicates that the
other ring sizes either weakly bind (Ka < 1 × 104 M−1)51 or do
not bind K+ at detectable levels, so their cyclization rates and
respective yields are minimally affected.
The largest cation (Cs+) with tetradepsipeptide 2 signifi-
cantly amplified both of the larger ring sizes relative to the
control but did not favor them substantially as observed with 4
and Na+. ITC data shows that 6 and 7 have similar, but weak,
binding affinities to Cs+ (Figure 4). However, using Cs+ with
didepsipeptide monomer 1 does not result in significant
amplification of macrocycles relative to the salt-free control,
aside from a modest increase in 36-membered ring (6). In these
two examples, the modest change in size distribution can largely
be attributed to the lack of a driving force from formation of a
strong Cs+-binding product.
E
DOI: 10.1021/jacs.7b13148
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 7. Application of structural and thermodynamic trends to templated macrocycle synthesis by MCO.
Correlation of ITC Measurements With Salt Effects
and D,D-Macrocycles. The use of enantioselective catalysis to
prepare the depsipeptide monomer provided a synthesis
platform to investigate a broad scope of precursors that
might exhibit a more diverse range of behaviors in the MCO
reaction. It was more tractable, therefore, to question whether
the trends observed with monomers 1 and 2 might also apply
to monomers with different relative configurations (Figure 6).
Unlike its L,L-analogue (1), L,D-monomer (8) inherently
favored direct cyclization to the 6-membered ring (9), in
addition to producing small amounts of 18-membered (10) and
24-membered rings (11) in salt-free conditions. However,
addition of Na+ doubled the amounts of both 10 and 11, while
the 6-membered ring (9) was significantly decreased. Addi-
tionally, a small amount of a 30-membered ring (12), which
was not observed in salt-free conditions, was formed.
Macrocycles 10, 11, and 12 strongly bind Na+, while 9 does
not at detectable levels, which can explain why the rate of direct
cyclization to 9 was so greatly decreased. Additionally, the
binding affinities of 10 and 11 to Na+ are strong enough to
drive preorganization of their linear precursors and subsequent
cyclization at a rate much faster than intermolecular
oligomerization and cyclization to 12, despite its slightly
stronger binding affinity to Na+. Additionally, 12 exhibits a
positive enthalpy of binding, suggesting that the binding is due
to an unfavorable conformational change26,56,57 as opposed to
hydrogen bonding. Because binding interactions of cyclized
products are known to reflect their respective linear precursors,
it can be inferred that the transition state leading to 12 binds
Na+ in a conformation that does not promote cyclization
(negative templating effect),33 which can explain why the rate of
cyclization to 12 is not more competitive with 10 and 11.
Addition of K+ to monomer 8 decreased production of the
monocyclized product, but only modestly increased the 30-
membered ring (12) relative to the control and the other
macrocyclic products. ITC data show that product 11 weakly
binds K+ and that products 9, 10, and 12 do not bind at
detectable levels. In this example, the modest change in size
distribution can be attributed to the lack of a driving force from
formation of a strong K+-binding product. The most notable
results of this series were observed when Cs+ was used as an
additive: the 30-membered ring (12) was isolated as the major
product, when it was not originally observed in salt-free
conditions. The 30-membered ring (12) is the strongest 1:1
Cs+ binder, followed by the 24-membered ring, while 9 and 10
do not have detectable binding. Here, Cs+ effectively accelerates
the rate of cyclization to the two Cs+-binding macrocycles,
which resulted in their increased production.
Prediction of Salt Effects Using Structural and
Thermodynamic Trends. On the basis of the trends observed
in Figures 4−6, the distribution of macrocycles using salt
additives as a variable suggests that inherent structure-based
reactivity can be inferred from ITC analysis of macrocycle−salt
binding. To demonstrate this principle, a new monomer (13)
was synthesized,58 and its MCO products from the
untemplated reaction were analyzed by ITC before running
the MCOs with salts (Figure 8). Without salt, didepsipeptide
13 provided 18-, 24-, and 36-membered rings. The overall yield
of the untemplated MCO using 13 was quite low (47%), likely
due to the steric effect of A1,3-strain (from the p-chloromandelic
acid residue) on the Mitsunobu reaction. The low initial
efficiency provided a workable starting point to study the effects
of salts, especially those that might induce more reactive
conformations by templation. The thermodynamic interactions
of 14, 15, and 16 with various salts were measured by ITC, and
predictions for how the size distribution might change in the
presence of salts were made based on the structural and
thermodynamic trends (Figure 7) observed in previous
examples. With Na+, the 24-membered ring (15) has an
order of magnitude stronger 1:1 binding affinity than the other
two macrocycles, and thus, it should be amplified the most.
Macrocycles 14 and 16 also have strong binding affinities to
Na+ but are not expected to be greatly increased. Preorganiza-
tion to form 15 will likely be faster than cyclization to 14, and
16 requires a double-ion template, which will form more slowly
than the single-ion template required for 15. In general, 14, 15,
and 16 have fairly weak binding affinities to the larger metal
cations (K+ and Cs+), compared to the previously described
examples. Macrocycles 14 and 16 both have enthalpically
favorable, weak 1:1 binding affinities for K+, and 15 has a
strong, enthalpically disfavored 2:1 binding affinity. The
thermodynamically disfavored double-ion binding of the 24-
membered ring may result in a diminished rate of cyclization
and potentially promote the formation of a 30-membered ring,
which is not formed in the untemplated oligomerization. Lastly,
15 and 16 weakly bind Cs+, while 14 does not, which is
hypothesized to result in modest amplification of both of the
former.
F
DOI: 10.1021/jacs.7b13148
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 8. Use of ITC to predict changes in product size distribution from 13. Significantly increased yields are shown in bold. ITC binding affinities
<1 × 103 M−1 are not detected.
Comparison of Predicted Salt Effects with Exper-
imental Outcomes. Next, these predictions were compared
to the outcomes when 13 was subjected to MCO conditions
using the various ion templates. As shown in Figure 8, the size
distributions of macrocycles formed in the presence of salts are
largely aligned with the predictions made from ITC measure-
ments. Out of the three salt additives, Na+ and Cs+, provided
size distributions that most closely resembled the predictions.
The addition of Na+ substantially amplified the strongest-
binding macrocycle (15) and modestly increased 16, while Cs+
modestly increased 15 and 16, as predicted. The template effect
of K+ was more difficult to predict because a 30-membered ring
was not isolated from the salt-free conditions; hence, its binding
affinity to K+ is unknown, and the binding affinities of all three
isolated macrocycles were weak. The modest amplification of
15, concomitant with little to no amplification of the other two
ring sizes relative to the control, may be due to K+ causing the
reaction to stall at the 24-membered linear precursor. The 24-
membered ring (15) has an enthalpically unfavorable double-
ion binding to K+, which could mean that its linear precursor’s
rate of inter- and intramolecular coupling are both slowed,
resulting in the reaction stalling at this intermediate. Because
ion binding is a reversible process, the amplification of 15 could
be due to untemplated cyclization of some of its accumulated
linear precursor.
CONCLUSION
■
In summary, we have implemented ITC as a tool to study
macrocycle-template binding interactions, and we show that
these measurements can be correlated to size distributions of
depsipeptides formed during Mitsunobu-based MCOs. These
studies have identified key trends in quantitative metal ion−
cyclic depsipeptide binding interactions across discrete
collections of macrocycles. Importantly, these behaviors can
be elucidated across collections of macrocycles formed from
monomers of different sizes, relative configurations, and side-
chains. In general, the ITC data indicates that there is a trend
where macrocycles with a strong template binding affinity (Ka >
1 × 104) will be formed in substantially greater amounts in the
presence of that template. Two particularly notable instances of
this phenomenon were observed with 15, which was doubled in
the presence of Na+ (Ka = 1.21 × 105 M−1), and 12, which was
formed as the major product in the presence of Cs+ (Ka = 1.20
× 104 M−1) but not formed at all without a salt. The
stoichiometry and enthalpy parameters from the ITC data, in
combination with relative binding affinities, provide us with a
G
DOI: 10.1021/jacs.7b13148
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(6) Bertram, A.; Hannam, J. S.; Jolliffe, K. A.; Gonzal
Turiso, F.; Pattenden, G. Synlett 1999, 1999, 1723.
́ ́
ez-Lopez de
full picture on how the ion template will affect the size
distribution of products in a collection of macrocycles. In
general, a 1:1 metal ion/macrocycle complex observed by ITC
indicates a faster rate of formation than what would be
observed with a 2:1 complex, simply because preorganization of
a 1-ion cyclization complex is faster than one that requires 2
ions. ITC allowed 1:2 complexes to be observed, and these
were found to correlate with inhibited monocyclization of
smaller rings and increased formation of their corresponding
cyclodimer in the presence of the salt additive. However, when
observed with larger rings, 1:2 complexes correlate with overall
slowed cyclization. The enthalpy parameter served as a good
indicator for whether or not cyclization to a particular ring size
would be favorable in the presence of a given template. This
parameter is especially important as a predictive measure if
multiple macrocycles in a series have strong binding to a
particular ion. In general, negative binding enthalpy with strong
ion affinity resulted in an increased rate of cyclization, and
positive binding enthalpy with strong ion affinity resulted in a
slowed rate of cyclization. The thermodynamic trends
described here provide the first analytical platform to rationally
select a template for a targeted size regime of cyclo-
depsipeptides. Our current efforts are directed at broadening
the synthetic impact of this method by continuing to expand
our scope of cyclodepsipeptides, while uncovering more
structure-related trends with ITC. Future studies will explore
this tool further, perhaps as a measure to correlate antibiotic
activity of cyclodepsipeptides with ion-binding selectivity.59
(7) Beyer, R. L.; Singh, Y.; Fairlie, D. P. Org. Lett. 2008, 10, 3481.
(8) Wipf, P.; Miller, C. P.; Grant, C. M. Tetrahedron 2000, 56, 9143.
(9) Rothe, M.; Kreiss, W. Angew. Chem., Int. Ed. Engl. 1977, 16, 113.
(10) Beeler, A. B.; Acquilano, D. E.; Su, Q.; Yan, F.; Roth, B. L.;
Panek, J. S.; Porco, J. A., Jr. J. Comb. Chem. 2005, 7, 673.
(11) Pitchayawasin, S.; Kuse, M.; Koga, K.; Isobe, M.; Agata, N.;
Ohta, M. Bioorg. Med. Chem. Lett. 2003, 13, 3507.
(12) Heitz, F.; Kaddari, F.; Heitz, A.; Raniriseheno, H.; Lazaro, R. Int.
J. Pept. Protein Res. 1989, 34, 387.
(13) Grell, E.; Funck, T.; Sautee, H. Eur. J. Biochem. 1973, 34, 415.
(14) Marti-Centelles, V.; Pandey, M. D.; Burguete, M. I.; Luis, S. V.
Chem. Rev. 2015, 115, 8736.
(15) Burke, S. D.; McDermott, T. S.; O’Donnell, C. J. J. Org. Chem.
1998, 63, 2715.
(16) Liu, M.; Tang, Y. C.; Fan, K. Q.; Jiang, X.; Lai, L. H.; Ye, Y. H. J.
Pept. Res. 2005, 65, 55.
(17) Marti-Centelles, V.; Burguete, M. I.; Luis, S. V. J. Org. Chem.
2016, 81, 2143.
(18) Blake, A. J.; Hannam, J. S.; Jolliffe, K. A.; Pattenden, G. Synlett
2000, 2000, 1515.
(19) (a) Batiste, S. M.; Johnston, J. N. Proc. Natl. Acad. Sci. U. S. A.
2016, 113, 14893. (b) Leighty, M. W.; Shen, B.; Johnston, J. N. J. Am.
Chem. Soc. 2012, 134, 15233.
(20) Liu, S.; Gu, W.; Lo, D.; Ding, X. Z.; Ujiki, M.; Adrian, T. E.;
Soff, G. A.; Silverman, R. B. J. Med. Chem. 2005, 48, 3630.
(21) Chen, Y.; Bilban, M.; Foster, C. A.; Boger, D. L. J. Am. Chem.
Soc. 2002, 124, 5431.
(22) Garcia-Martin, F.; Cruz, L. J.; Rodriguez-Mias, R. A.; Giralt, E.;
Albericio, F. J. Med. Chem. 2008, 51, 3194.
(23) Hewitt, W. M.; Leung, S. S.; Pye, C. R.; Ponkey, A. R.;
Bednarek, M.; Jacobson, M. P.; Lokey, R. S. J. Am. Chem. Soc. 2015,
137, 715.
(24) Corbett, P. T.; Leclaire, J.; Vial, L.; West, K. R.; Wietor, J. L.;
Sanders, J. K.; Otto, S. Chem. Rev. 2006, 106, 3652.
(25) Ludlow, R. F.; Liu, J.; Li, H.; Roberts, S. L.; Sanders, J. K.; Otto,
S. Angew. Chem., Int. Ed. 2007, 46, 5762.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.7b13148.
■
S
Procedures, complete experimental details, and analytical
data for all new compounds (PDF)
(26) Roberts, S. L.; Furlan, R. L.; Otto, S.; Sanders, J. K. Org. Biomol.
Chem. 2003, 1, 1625.
1H NMR and 13C NMR results (PDF)
(27) Corbett, P. T.; Sanders, J. K.; Otto, S. J. Am. Chem. Soc. 2005,
127, 9390.
(28) Corbett, P. T.; Sanders, J. K.; Otto, S. Chem. - Eur. J. 2008, 14,
AUTHOR INFORMATION
Corresponding Author
*jeffrey.n.johnston@vanderbilt.edu
ORCID
Jeffrey N. Johnston: 0000-0002-0885-636X
Notes
The authors declare no competing financial interest.
■
2153.
(29) Furlan, R. L.; Ng, Y. F.; Otto, S.; Sanders, J. K. J. Am. Chem. Soc.
2001, 123, 8876.
(30) Di Stefano, S. J. Phys. Org. Chem. 2010, 23, 797.
(31) Berrocal, J. A.; Cacciapaglia, R.; Stefano, S. D.; Mandolini, L.
New J. Chem. 2012, 36, 40.
(32) Furlan, R. L.; Otto, S.; Sanders, J. K. Proc. Natl. Acad. Sci. U. S. A.
2002, 99, 4801.
(33) Anderson, S.; Anderson, H. L.; Sanders, J. K. M. Acc. Chem. Res.
1993, 26, 469.
ACKNOWLEDGMENTS
■
Research reported in this publication was supported by the
National Institute of General Medical Sciences (NIH GM
063557).
(34) Fahrenbach, A. C. Pure Appl. Chem. 2015, 87, 205.
(35) Lucas, D.; Minami, T.; Iannuzzi, G.; Cao, L.; Wittenberg, J. B.;
Anzenbacher, P., Jr.; Isaacs, L. J. Am. Chem. Soc. 2011, 133, 17966.
(36) Anderson, H. L.; Sanders, J. K. M. J. Chem. Soc., Chem. Commun.
1989, 1714.
(37) Anderson, H. L.; Sanders, J. K. M. Angew. Chem., Int. Ed. Engl.
1990, 29, 1400.
(38) Kano, K.; Kondo, M.; Inoue, H.; Kitagishi, H.; Colasson, B.;
Reinaud, O. Inorg. Chem. 2011, 50, 6353.
(39) Lee, Y. J.; Ho, T. H.; Lai, C. C.; Chiu, S. H. Org. Biomol. Chem.
REFERENCES
■
(1) Ma, Y.; Marszalek, T.; Yuan, Z.; Stangenberg, R.; Pisula, W.;
Chen, L.; Mullen, K. Chem. - Asian J. 2015, 10, 139.
(2) Gokel, G. W.; Leevy, W. M.; Weber, M. E. Chem. Rev. 2004, 104,
2723.
(3) White, A. M.; Craik, D. J. Expert Opin. Drug Discovery 2016, 11,
2016, 14, 1153.
1151.
(40) Thakkar, A.; Trinh, T. B.; Pei, D. ACS Comb. Sci. 2013, 15, 120.
(41) Yep, Y.-h.; Gao, X.-m.; Liu, M.; Tang, Y.-c.; Tian, G.-l. Int. J.
Pept. Res. Ther. 2003, 10, 571.
(42) Liu, M.; Tang, Y. C.; Fan, K. Q.; Jiang, X.; Lai, L. H.; Ye, Y. H. J.
Pept. Res. 2005, 65, 55.
(4) Constable, D. J. C.; Dunn, P. J.; Hayler, J. D.; Humphrey, G. R.;
Leazer, J. L., Jr.; Linderman, R. J.; Lorenz, K.; Manley, J.; Pearlman, B.
A.; Wells, A.; Zaks, A.; Zhang, T. Y. Green Chem. 2007, 9, 411.
(5) Sokolenko, N.; Abbenante, G.; Scanlon, M. J.; Jones, A.; Gahan,
L. R.; Hanson, G. R.; Fairlie, D. P. J. Am. Chem. Soc. 1999, 121, 2603.
H
DOI: 10.1021/jacs.7b13148
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(43) Cougnon, F. B.; Au-Yeung, H. Y.; Pantos, G. D.; Sanders, J. K. J.
Am. Chem. Soc. 2011, 133, 3198.
(44) Sessler, J. L.; Gross, D. E.; Cho, W. S.; Lynch, V. M.;
Schmidtchen, F. P.; Bates, G. W.; Light, M. E.; Gale, P. A. J. Am. Chem.
Soc. 2006, 128, 12281.
(45) Amini, M. K.; Shamsipur, M. J. Phys. Chem. 1991, 95, 9601.
(46) Shamsipur, M.; Irandoust, M. J. Solution Chem. 2008, 37, 657.
(47) Izatt, R. M.; Pawlak, K.; Bradshaw, J. S.; Bruening, R. L. Chem.
Rev. 1991, 91, 1721.
(48) Danil de Namor, A. F.; Matsufuji-Yasuda, T. T.; Zegarra-
Fernandez, K.; Webb, O. A.; El Gamouz, A. Croat. Chem. Acta 2013,
86, 1.
(49) Hamieh, S.; Saggiomo, V.; Nowak, P.; Mattia, E.; Ludlow, R. F.;
Otto, S. Angew. Chem., Int. Ed. 2013, 52, 12368.
(50) Ulatowski, F.; Sadowska-Kuziola, A.; Jurczak, J. J. Org. Chem.
2014, 79, 9762.
(51) Ulatowski, F.; Dąbrowa, K.; Jurczak, J. Tetrahedron Lett. 2016,
57, 1820.
(52) Variance within control MCO experiments and within
templated experiments was insignificant and fell within the range of
reported DCL literature examples, so AFn > 0.09 can represent
significant templated amplification. See Supporting Information for
details.
(53) For ITC experiments, salts were chosen for their solubility in
MeOH, to ensure accurate concentrations. Prior work suggests cation-
dependent product size distribution (ref 19a). Thus, counteranions are
not included in the figures for simplicity. However, specific salts used
in ITC or MCO experiments are detailed in the Supporting
Information.
(54) Both macrocycles gave atypical sigmoidal ITC thermograms
with inflection points at 0.5 and 1.5. See Supporting Information for
details.
(55) Darby, S. J.; Platts, L.; Daniel, M. S.; Cowieson, A. J.; Falconer,
R. J. J. Therm. Anal. Calorim. 2017, 127, 1201.
(56) Kandeel, M.; Al-Taher, A.; Nakashima, R.; Sakaguchi, T.;
Kandeel, A.; Nagaya, Y.; Kitamura, Y.; Kitade, Y. PLoS One 2014, 9,
e94538.
(57) Otto, S.; Engberts, J. B. Org. Biomol. Chem. 2003, 1, 2809.
(58) See Supporting Information for monomer synthesis.
(59) Ovchinnikov, Y. A. FEBS Lett. 1974, 44, 1.
I
DOI: 10.1021/jacs.7b13148
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX