Stitched α-helical peptides via bis ring-closing metathesis

Article abstract of DOI:10.1021/ja505141j

Conformationally stabilized α-helical peptides are capable of inhibiting disease-relevant intracellular or extracellular protein-protein interactions in vivo. We have previously reported that the employment of ring-closing metathesis to introduce a single all-hydrocarbon staple along one face of an α-helical peptide greatly increases α-helical content, binding affinity to a target protein, cell penetration through active transport, and resistance to proteolytic degradation. In an effort to improve upon this technology for stabilizing a peptide in a bioactive α-helical conformation, we report the discovery of an efficient and selective bis ring-closing metathesis reaction leading to peptides bearing multiple contiguous staples connected by a central spiro ring junction. Circular dichroism spectroscopy, NMR, and computational analyses have been used to investigate the conformation of these stitched peptides, which are shown to exhibit remarkable thermal stabilities. Likewise, trypsin proteolysis assays confirm the achievement of a structural rigidity unmatched by peptides bearing a single staple. Furthermore, fluorescence-activated cell sorting (FACS) and confocal microscopy assays demonstrate that stitched peptides display superior cell penetrating ability compared to their stapled counterparts, suggesting that this technology may be useful not only in the context of enhancing the drug-like properties of α-helical peptides but also in producing potent agents for the intracellular delivery of proteins and oligonucleotides.

Full text of DOI:10.1021/ja505141j

Article
pubs.acs.org/JACS
Stitched α‑Helical Peptides via Bis Ring-Closing Metathesis
Gerard J. Hilinski,†,∇ Young-Woo Kim,†,∥,∇ Jooyeon Hong,‡ Peter S. Kutchukian,†,#
Charisse M. Crenshaw,†,○ Shaunna S. Berkovitch,†,⊥ Andrew Chang,†,◆ Sihyun Ham,*,‡
and Gregory L. Verdine*,†
†Department of Chemistry and Chemical Biology and Department of Stem Cell and Regenerative Biology, Harvard University,
Cambridge, Massachusetts 02138, United States
‡Department of Chemistry, Sookmyoung Women’s University, Hyochangwon-gil 52, Yongsan-ku, Seoul 140-742, Korea
S
* Supporting Information
ABSTRACT: Conformationally stabilized α-helical peptides
are capable of inhibiting disease-relevant intracellular or
extracellular protein−protein interactions in vivo. We have
previously reported that the employment of ring-closing
metathesis to introduce a single all-hydrocarbon staple along
one face of an α-helical peptide greatly increases α-helical
content, binding affinity to a target protein, cell penetration
through active transport, and resistance to proteolytic
degradation. In an effort to improve upon this technology
for stabilizing a peptide in a bioactive α-helical conformation,
we report the discovery of an efficient and selective bis ring-
closing metathesis reaction leading to peptides bearing multiple contiguous staples connected by a central spiro ring junction.
Circular dichroism spectroscopy, NMR, and computational analyses have been used to investigate the conformation of these
“stitched” peptides, which are shown to exhibit remarkable thermal stabilities. Likewise, trypsin proteolysis assays confirm the
achievement of a structural rigidity unmatched by peptides bearing a single staple. Furthermore, fluorescence-activated cell
sorting (FACS) and confocal microscopy assays demonstrate that stitched peptides display superior cell penetrating ability
compared to their stapled counterparts, suggesting that this technology may be useful not only in the context of enhancing the
drug-like properties of α-helical peptides but also in producing potent agents for the intracellular delivery of proteins and
oligonucleotides.
led to an intense focus on this particular secondary structure.
The α-helical conformation has been stabilized using many
different types of cross-links, including lactam,5−8 disulfide,9−11
thioether,12 azobenzene,13 hydrazone,14 triazole,15 biphenyl,16
bis-triazoylyl,17 oxime,18,19 perfluoroaryl,20 and carbamate.21
Ruthenium-catalyzed ring-closing metathesis (RCM) has
proven useful as a means of introducing helix-constraining
cross-links into peptides, including 310-helices by Grubbs and
co-workers22−24 and α-helices by Arora and co-workers.25 We
have previously reported one such metathesis-enabled α-helix
reinforcement system, the all-hydrocarbon staple, which unlike
all previous systems incorporates both α-methyl groups and a
macrocyclic all-hydrocarbon ring as two distinct helix-stabilizing
elements (Figure 1).26−29 This system affords exceptionally
high levels of α-helix induction and nucleation, and as a direct
result can increase target binding affinity, serum stability, and in
vivo half-life.30−32 Perhaps most significantly, however, it has
been shown in multiple studies employing different stapled
peptides that the presence of the hydrocarbon staple causes
these peptides to be actively transported into mammalian cells
INTRODUCTION
■
Intracellular protein−protein interactions serve pivotal roles in
nearly all aspects of cell function, and their involvement in
disease pathophysiology is well-established. Therapeutic
intervention at these critical interfaces has proven frustrating,
however, because the relatively flat contact surfaces participat-
ing in protein−protein interactions are devoid of hydrophobic
pockets and therefore poorly suited for targeting by small
molecules.1 Though extracellular protein−protein interactions
are amenable to interference with biologics such as monoclonal
antibodies, these large molecules are incapable of penetrating
the cell membrane and accessing the cytoplasm and nucleus.
Thus, there is significant interest in the development of
alternative types of molecules that feature the cell penetrating
ability of small molecules yet can engage protein surfaces with
the efficacy of biologics. Toward this end, we and others have
pursued an approach that involves deconstructing proteins
down to “dominant interfering” peptides comprising a minimal
protein-targeting surface and then stabilizing the bioactive
conformation through the introduction of a synthetic cross-link.
Though numerous synthetic buttressing systems have been
reported to stabilize a variety of folded peptide conformations,
the ubiquity of α-helices in protein−protein interfaces2−4 has
Received: May 22, 2014
Published: August 8, 2014
© 2014 American Chemical Society
12314
dx.doi.org/10.1021/ja505141j | J. Am. Chem. Soc. 2014, 136, 12314−12322

Journal of the American Chemical Society
Article
fusion known to be exceptionally rigidifying. To our knowledge,
this is the first report of α-helical peptides stabilized by multiple
connected staples. We envision this system of installing
contiguous staples along one face of the α-helix as akin to
peptide “stitching”. In this article, we describe highly regio- and
stereospecific RCM reactions on an α-helical scaffold, leading
to efficient, clean formation of stitched α-helical peptides. We
also demonstrate that peptide stitching provides superior helix
induction, thermal stability, proteolytic resistance, and cell
permeability relative to peptide stapling.
EXPERIMENTAL METHODS
■
Figure 1. (A) Olefin-bearing amino acids used in this study. (B)
Schematic structures of helical peptides constrained by an all-
hydrocarbon staple across one turn (i, i + 4) or two turns (i, i + 7)
of an α-helix. The nomenclature Si,i+4S(8) refers to an 8-carbon
metathesized cross-link with (S)-configuration at positions i and i + 4;
Ri,i+7S(11), an 11-carbon cross-link with (R)- and (S)-configuration at
positions i and i + 7, respectively.
Peptide Synthesis. The peptides were prepared using Fmoc
chemistry on Rink Amide MBHA resin (NovaBiochem) with a loading
capacity of 0.59 mmol/g. The dry resin was swelled with CH2Cl2
followed by 1-methyl-2-pyrrolidinone (NMP) before use. The Fmoc
protecting group was removed by treatment with 25% piperidine in
NMP (4 × 5 min). Natural amino acids were coupled for 60 min using
(1-cyano-2-ethoxy-2-oxoethylidenaminooxy)dimethylamino-morpholi-
no-carbenium hexafluorophosphate (COMU) as the activating agent
(4 equiv), 4 equiv of Fmoc-protected amino acid, and 8 equiv of
diisopropyl ethylamine (DIPEA) in NMP. For the coupling of olefin-
bearing amino acids, a reaction time of 2 h was used with Fmoc-
protected amino acid (3 equiv), COMU (3 equiv), and DIPEA (6
equiv). After each coupling or deprotection reaction, the resin was
washed with NMP (5 × 1 min).
through endocytic vesicle trafficking.33−36 Moreover, these
molecules can efficiently escape endosomes and access both
cytoplasmic and nuclear targets, as they are well below the 40
kDa exclusion size of the nuclear pore complex. The
therapeutic potential of stapled peptides has been underscored
by studies demonstrating efficacy in animal models of
disease,33,37−41 and Aileron Therapeutics has initiated clinical
trials of the first stapled peptide therapeutics.
The first-reported all-hydrocarbon staples were incorporated
into a peptide through the introduction of two α-methyl,α-
alkenylglycine units at positions separated by either three (i, i +
4; one helical turn) or six (i, i + 7; two helical turns)
intervening residues. The macrocyclic ring was then closed via
ruthenium-catalyzed RCM to form the Si,i+4S(8) and Ri,i+7S(11)
staple, respectively, with the nomenclature specifying the
configurations of the stapling residues and the final length of
the cross-link (Figure 1B). Recently, we have reported a
hydrocarbon staple that involves the introduction of two α-
methyl, α-alkenylglycine units at positions separated by two
intervening residues (i, i + 3 one helical turn) and subsequent
closure via ruthenium-catalyzed RCM to form the Ri,i+3S(8)
staple.28 These stapling systems have now been employed
successfully by our laboratory and others against a wide variety
of high-value biological targets, as reviewed recently by
Walensky and Bird.32
All-hydrocarbon staples can serve as powerful nucleators of
α-helical structure, such that the incorporation of a single staple
into a 4, 5, or 8 amino acid stretch of a peptide can induce
helical structure throughout a much longer sequence. Addi-
tionally, two or more separate and distinct staples can be
incorporated into a single peptide, conferring increased
proteolytic resistance and enhanced biological activity.37,41
Unsurprisingly, however, the nucleated stretches of stapled
peptides do not show the same extent of protease resistance as
the stapled portion. We therefore sought to extend the
synthetically constrained region over a larger stretch of the
peptide, with the aim of conferring protease resistance even in
harsh environments such as those encountered in the human
intestinal tract.
Metathesis and Purification. Ring-closing metathesis of resin-
bound N-Fmoc, side-chain-protected peptides was performed using 20
mol % of Grubbs I catalyst in 1,2-dichloroethane (DCE) for 2 h at
room temperature. The reactions were monitored by LC/MS after
cleavage of the peptides from a resin aliquot, and additional treatments
with fresh catalyst were used to obtain more complete conversion in
many cases. After the reaction solution was drained, the resin was
washed with DCE (3 × 1 min) and CH2Cl2 (3 × 1 min). After Fmoc
group removal, Fmoc-Ebes-OH (N-Fmoc-2,2′-(ethylenedioxy) bis
(ethylamine) monosuccinamide) or Fmoc-β-alanine-OH was coupled
followed by Fmoc group deprotection. For the acetylated analogs, the
N-terminal amine was treated with 30 equiv of acetic anhydride and 60
equiv of DIPEA in NMP for 45 min. For the fluoresceinated analogs,
the peptide N-termini were exposed overnight to fluorescein
isothiocyanate (7 equiv) in DMF in the presence of DIPEA (10
equiv) at room temperature. The resin was washed with DMF (3 × 3
min), CH2Cl2 (3 × 3 min), and methanol (3 × 3 min) and dried in
vacuo overnight. The peptides were deprotected and cleaved from the
resin by treatment with a mixture of trifluoroacetic acid/
triisopropylsilane/water (95/2.5/2.5) for 2−3 h and precipitated by
addition of a 1:1 mixture of n-pentane and diethyl ether. The
precipitate was collected by centrifugation, dissolved in a 1:1 mixture
of acetonitrile and water, and filtered to remove the resin. The
products were purified by reverse-phase HPLC using a Zorbax C18
column (Agilent, 5 μm, 9.4 × 250 mm) and characterized by liquid
chromatography/mass spectrometry (LC/MS) using a Zorbax C18
column (Agilent, 3.5 μm, 2.1 × 150 mm) and amino acid analysis.
Circular Dichroism (CD). Peptides were dissolved in 20 mM
potassium phosphate buffer (pH 6.5), and the concentrations were
determined by absorbance spectroscopy (extinction coefficient for
tryptophan, λ280 = 5690 cm−1). CD spectra were collected on a Jasco J-
710 spectropolarimeter equipped with a temperature controller using
the following standard measurement parameters: 0.5 nm step
resolution, 20 nm/s speed, 10 accumulations, 1 s response, 1 nm
bandwidth, 0.1 cm path length. All spectra were converted to a
uniform scale of molar ellipticity after background subtraction. Curves
shown are smoothed with standard parameters. Temperature-depend-
ent CD spectra of each peptide (94−100 μM) were recorded at
varying temperatures (every 10 °C from 10 to 90 °C) from 260 to 185
nm. To generate thermal unfolding curves, the ellipticity at 222 nm
was measured every 1 °C from 10 to 90 °C with a temperature slope
of 3 °C/min. To obtain the melting temperatures, we analyzed the
Here we report a next-generation version of the hydro-
carbon-stapling system in which dual hydrocarbon staples
emerge from common attachment points in the peptide,
thereby creating spiro macrocyclic ring junctions, a type of ring
12315
dx.doi.org/10.1021/ja505141j | J. Am. Chem. Soc. 2014, 136, 12314−12322

Journal of the American Chemical Society
Article
thermal unfolding curves using a two-state model with a 95%
confidence interval as previously described.43 Most stitched peptides
did not have a cooperative melting transition point in this temperature
range, and therefore their melting temperatures were indeterminable
by this method.
with Vectashield/DAPI (Vector), and then analyzed using a Zeiss
LSM 510 META microscope at 63× and Zeiss LSM image software.
Cellular Uptake of Oligoarginine Analogs. HeLa cells were
cultured in DMEM with 10% fetal bovine serum, seeded at 20,000
cells per chamber in eight-chamber slides (Nalge Nunc), and allowed
to adhere for 48 h at 37 °C with 5% CO2. Cells were washed with PBS,
incubated in serum-free medium at 37 °C for 1 h, and then treated
with 5 μM fluorescein-labeled peptide (originally at 1 mM in DMSO)
in serum-free DMEM at 37 °C for 4 h. After incubation, the cells were
washed twice with PBS and with DMEM without serum and then
directly analyzed using a Zeiss LSM 510 META microscope at 25× or
63× and Zeiss LSM image software.
Peptide Digestion Assay. 2.5 μL of trypsin solution (5 μM,
Sigma) in digestion buffer (0.1 M NH4HCO3 buffer, pH 8.0) was
added to 250 μL of a fluoresceinated peptide solution (40 μM) in the
digestion buffer (enzyme/substrate = 1/800), and the resulting
mixture was incubated with rapid shaking (500 rpm) at room
temperature. Thirty μL of digestion mixture was taken at 0, 5, 10, 30,
60, 120, minutes and then quenched with 2 μL of trifluoroacetic acid.
The residual substrate and hydrolyzed products were quantified by
LC/MS-based peak detection at 494 nm. Each experiment was
performed in duplicate. The half-life, t1/2, was determined by linear
regression analysis using KaleidaGraph (Synergy Software) from a plot
of ln S (S is % residual substrate) versus time (min) (t1/2 = −ln 2/
slope).
Molecular Modeling. A Monte Carlo conformational search was
performed to locate all low-energy conformations of each linker in the
helical state. To generate starting conformations for the MC search, a
15-residue polyalanine peptide was built with a right-handed helical
conformation using MacroModel’s Maestro GUI.44 Hydrocarbon
cross-links were manually added and then fully minimized, while all
non-cross-linker atoms were held frozen. For each isomer, at least two
distinct 5−10,000 step Monte Carlo conformational searches were
run. For all calculations, energies were evaluated using the OPLS2005
force field as implemented in Macromodel. For all minimizations the
Polak−Ribiere conjugate gradient (PRCG) method was employed,
and the convergence criterion for the minimization of gradient norm
was set to <0.05 kJ/mol·Å. We employed the GB/SA solvation
treatment,45 modeling the solvent as chloroform as all metathesis
reactions were carried out in 1,2-DCE. Bond dipole cutoffs were
employed to truncate the electrostatic and GB terms. Nonbonded
cutoffs were as follows: 8 Å in van der Waals, 99999.0 Å in charge−
charge (effectively infinite), 203/2 Å (89.4 Å) in charge−dipole, and 20
Å in dipole−dipole. Harmonic constraints (100 kJ/mol) were placed
on each backbone dihedral angle to maintain the helical conformation
throughout the search. At each step of the Monte Carlo search, 2−5
cross-link dihedrals were randomly selected, and their values were
adjusted by 0−180°. The C-terminal C−C bond adjacent to each
olefin was temporarily broken during each step, allowing for dihedral
perturbations along the cross-link, and then reattached after dihedral
modification. After each step, up to 500 steps of minimization were
performed, if convergence was not achieved in fewer steps, and
conformations within 50 kJ of the global minimum were saved. After
the search, all remaining structures were fully minimized, and all
conformations within 15 kJ of the global minimum were kept, while
redundant structures (RMSD < 0.25 Å) were removed. Runs were
carried out until the number of new structures obtained after pooling
the conformations obtained from the current run with those obtained
from all previous runs was <10% of the initial number of structures,
suggesting that conformational space had been fully explored (Tables
S1 and S2).
Flow Cytometry. Jurkat cells (ATCC #TIB-152) were grown in
RPMI-1640 supplemented with 10% fetal bovine serum and penicillin/
streptomycin (500 units/mL and 0.5 mg/mL, respectively). For
analysis, cells were washed with PBS and resuspended in serum-free
RPMI-1640 at a density of 106 cells/mL. Cells were incubated in
serum-free media at either 37 or 4 °C for 1 h. Peptides (100 μM in
DMSO) were added to 150,000 cells at a final concentration of 1 μM
and incubated for 1 h. Samples were resuspended in 500 μL PBS and
analyzed by flow cytometry on a LSRII flow cytometer (Becton
Dickinson). Experiments were performed independently at least three
times. Typically, total Jurkat population size was roughly 1,000. Cells
were gated for histogram analysis based on forward and side scatter
values indicative of single, live cells: FlowJo’s automatic gating tool was
used to conservatively isolate the population of single, live cells from
the population of cells with low forward and high side scatter. The
gated population accounted for 60−80% of the total cell population.
Confocal Microscopy. Temperature-Dependent Cellular Uptake
of RNase A Analogs. U2OS cells (kindly provided by the Schreiber
laboratory at the Broad Institute of Harvard and MIT) were cultured
in DMEM with 10% fetal bovine serum, seeded at 100,000 cells per
chamber in four-chamber slides (Nalge Nunc), and allowed to adhere
overnight at 37 °C with 5% CO2. Thirty minutes prior to peptide
addition, slides were either preincubated at 4 °C or left at 37 °C. Cells
were washed with PBS and then incubated with 2 μM fluorescein-
labeled peptide (originally at 0.1 mM in DMSO) in serum-free
DMEM at either 4 or 37 °C for 4 h. After incubation, the cells were
washed twice with PBS and then fixed for 15 min in 4% formaldehyde
(Sigma-Aldrich) in PBS. Cells were washed twice with PBS, mounted
RESULTS AND DISCUSSION
■
The critical synthetic challenge in producing stitched peptides
is to control the regiochemistry and stereochemistry of RCM
reactions on peptides having multiple chemically equivalent
olefins. Peptide 1, a typical substrate for stitching, contains four
olefins, one each borne on (S)-α-methyl,α-pentenylglycine (S5)
and (S)-α-methyl,α-octenylglycine (S8), and two borne on bis-
pentenylglycine (B5, 2-(((9H-Fluoren-9-yl)methoxy)-
carbonylamino)-2-(pent-4-enyl)hept-6-enoic acid) (Figure
2A). RCM on 1 could therefore in principle generate three
possible bis-metathesized products, 2−4 (Figure 2A). Of
particular concern was the possibility that the two olefins in
B5 might preferentially react with each other during RCM
(reaction a). To experimentally investigate all the possible
RCM reaction pathways shown in Figure 2A, we turned to
model studies examining each single-RCM process in isolation
(Table 1). Significantly, we found the intraresidue RCM of B5
(reaction a) to be highly disfavored, undergoing only 23%
conversion to the cyclononenylglycine product with the model
peptide under our typical reaction conditions (20 mol %
Grubbs I catalyst, 3 × 2h) (Table 1, entry I). This RCM
reaction was found to be disfavored even with the monomeric
Fmoc-protected ethyl ester of B5, in which case the RCM failed
completely (Figure S1). These results are in agreement with
previous observations that RCM to form a 9-membered
carbocyclic ring or larger is typically ineffective.46,47 Just as
undesired reaction a proved unfavorable, so also did undesired
reaction b (entry II). Taken together, these results suggested
that the a+b bis-RCM pathway leading to product 2 would be
highly unlikely.
To test the likelihood of formation of the undesired product
3, we tested reaction c and found no evidence of product
formation (entry III). This particular reaction manifold would
give an Si,i+4R(8) staple, which is known from both
experimental and computational studies to be highly disfavored.
In data not shown, the peptide designed to test reaction d gave
high conversion only after three treatments with catalyst (entry
IV), whereas the peptide designed to test reaction f gave high
conversion after only two treatments with catalyst (entry VI).
12316
dx.doi.org/10.1021/ja505141j | J. Am. Chem. Soc. 2014, 136, 12314−12322

Journal of the American Chemical Society
Article
Of the six possible mono-RCM reactions, the two most
efficient ones were e and f. Should this preferential reactivity be
retained in a peptide containing the four olefinic tethers
required to introduce a stitch, then the e+f pathway would be
expected to provide product 4 selectively. Fully protected,
resin-bound substrate 1 was subjected to RCM under the same
conditions as used in the single-RCM reactions described
above. LC/MS analysis of the crude reaction mixture revealed
84% conversion to bis-RCM products following three 2-h
metathesis reactions (Table 1, entry XII). A single major
product accounted for ∼80% of the total product distribution
having the correct molecular mass. A number of minor
products having the correct molecular mass and together
comprising ∼20% of the total product distribution were also
observed, but none was formed in sufficient amounts to warrant
further characterization. As stapling studies have demonstrated
that small amounts of minor double bond isomers (trans for i,i
+4 and cis for i,i+7) are formed during peptide stapling RCM
reactions, such isomers of the major product are likely to
account for the bulk of the minor products observed.
Based on the results of the single-RCM reactions, 4 is the
most likely structure for the major bis-RCM product. To
further validate this assignment of the stitched product, we
performed a Monte Carlo conformational search and identified
all low energy conformations of each linker in the helical state.
The molecular modeling study indicates that the most stable
olefin isomer of 4 is thermodynamically more stable than that
of 3 by ∼15 kcal/mol. This is in part due to three syn-pentane
interactions that arise only in product 3: two involve the spiro
junction while the other involves the N-terminal attachment of
the i,i+4 staple (Figure 2C). Regarding olefin geometry in
product 4, we observed a preference of ∼2.5 kcal/mol for a cis
double bond in the i,i+4 staple. Though there is no significant
energetic preference for either double bond orientation in the
i,i+7 staple, the absence of any significant ring constraint most
likely results in a trans double bond that reflects the intrinsic
preference of the catalyst. We have thus far been unable to
confirm these predictions by NMR, as the signals for the
olefinic protons overlap and show uninterpretable coupling
patterns.
Encouraged by the efficiency of the RCM leading to the i + 4
+ 7 stitched product 4, we explored the possibility of forming
stitched peptides having other stitching configurations. Toward
this end, various peptide substrates bearing the olefin-
containing amino acid residues at different positions were
designed based on the established half-site rules (Table 1,
entries XIII-XVII). All of the peptide substrates designed to
incorporate a stitch composed of two favorable half-site staples
underwent a smooth RCM reaction to give the corresponding
double-RCM products. It should be noted that the peptide
bearing S5 at both positions i and i+8 along with B5 at i+4
(entry XV) demonstrated only modest conversion to a complex
mixture of four stitched products, with no single major product
detected. This particular peptide was designed to further
demonstrate the stereoselectivity of RCM reactions on an α-
helix scaffold; according to computational and experimental
single-RCM studies, the RCM of the pro-(R) side-chain of B5
with any of three olefin partners in this substrate would be
unfavorable (entries I, III, and X). On the contrary, the pro-(S)
side-chain of B5 would favorably react with S5 at either position
i or i + 8 to form an Si,i+4S(8) staple. The examples provided in
entries XIII-XV clearly demonstrate that the RCM reactions
Figure 2. (A) The α-helical tetra-olefinic peptide designed to undergo
bis-RCM. Three regioisomeric bis-RCM pathways are possible (a+b, c
+d, and e+f); these would yield products 2, 3, and 4, respectively. (B)
Schematic structure of the main product, the stitched peptide 4. The
stereochemical configuration of the spiro carbon (red dot) was
established by modeling. (C) Molecular models of the global minima
of all-Ala i + 4 + 7 stitched conformers. Product 3 is less stable than
product 4 by ∼15 kcal/mol due in part to the three syn-pentane
interactions highlighted by red arrows.
This result suggests a kinetic preference to form the staple
provided by reaction f, the product of which, Ri,i+7S(11), is also
predicted by modeling studies to be thermodynamically more
stable than Si,i+7S(11) by 6.5 kcal/mol. Taken together, these
results indicate that the c+d bis-RCM pathway leading to
undesired product 3 is unfavorable. By contrast, reactions e and
f, that lead respectively to the well-established Si,i+4S(8) and
R
i,i+7S(11) staples, proceeded efficiently to provide high
conversions to the corresponding stapled products (entries V
and VI).
Overall, these single-stapling studies demonstrated a high
degree of regio- and stereochemical control over the RCM
reactions on these peptide substrates; the exquisite selectivity of
RCM in these peptides is clearly evident from comparison of
entries III and V in Table 1, in which inversion of a single
stereogenic center causes a nearly quantitative reaction to fail.
We believe that the exceptional selectivity of these reactions
with chemically equivalent olefins results from the fact that the
peptide substrates are highly helical when undergoing RCM,
owing to the presence of multiple helix-promoting α,α-
disubstituted residues,48 the hydrophobic environment of the
cross-linking reaction provided by the nonpolar solvent 1,2-
DCE,49 and the protected side-chains.50 Stated otherwise, the
reactions are controlled by α-helix templating.
12317
dx.doi.org/10.1021/ja505141j | J. Am. Chem. Soc. 2014, 136, 12314−12322

Journal of the American Chemical Society
Article
Table 1. Sequences of Peptide Substrates and Percent Conversions for Metathesis Reaction
a
Percent conversion [desired RCM products/(desired RCM products + intermediates + starting material)] to stitched or stapled products after
b
three 2 h RCM reactions (unless otherwise noted) as determined by reversed-phase HPLC. Ratio of the major stitched (or stapled) final product to
c
the sum of all minor stitched (or stapled) final products as determined by reverse-phase HPLC. Percent helicity was calculated from mean-residue
ellipticity at 222 nm ([θ]222) using −31,500(1 − 2.5/n) and 0 deg cm2 dmol−1 as the values for 100 and 0% helicity, respectively; n is the number of
amino acid residues in the peptide.42 No product detected after two 2 h RCM reactions. Four stitched products were detected with relative
d
e
f
abundances of 3:3:2:1. CD studies were not conducted due to the complexity of the reaction. Percent helicity of two major stitched products after
isolation. Percent helicity of mixture of four stitched products. U represents α-aminoisobutyric acid (Aib), which mimics the helix-stabilizing effect
of the α,α-disubstituted S5. Four stitched products were detected with relative abundances of 4:2:2:1 A single stapled product was detected.
g
h
i
j
among multiple olefin side-chains in α-helical templates
proceed in a highly stereoselective manner.
S2B and S2C) using a combination of 2D NOESY and TOCSY
spectra (Figures S3 and S4).
With a series of stitched peptides in hand, we next
investigated the effects of stitching on the conformational
preference and structural stability of the peptides. In far
ultraviolet CD analyses, the stitched peptides displayed CD
spectra characteristic of a typical right-handed α-helix (Figure
3A), with a maximum near 195 nm and minima at 208 and 222
nm.42 Based on the mean residue ellipticity at 222 nm, all of the
stitched analogs displayed enhanced α-helical character
compared to the unmodified parent sequence 9 (Table 1).
The i + 4 + 7 stitched peptide 4 displays the highest helical
content of all the peptides studied here, 86%, which is far
higher than that of its i, i + 7 single-stapled counterpart 10
(60%) (Figure 3B). Notably, the CD spectrum of 4 is not
concentration dependent, suggesting that aggregation is not
responsible for the high α-helicity of this stitched peptide
(Figure S2A). The lower helical content of the i + 7 + 4
stitched peptide 7 (45/28%) when compared to 10 (60%) or 4
(86%) may result from a syn-pentane interaction that occurs
between the two staples of 7. The α-helical structure of the
stitched peptides was further confirmed by an NMR spectros-
copy study of an i + 4 + 7 stitched analog 13 possessing
enhanced water-solubility (Table 1, entry XXII and Figures
To examine the effect of stitching on conformational
stability, we compared stitched peptide 4 with its stapled and
unmodified counterparts in terms of resistance to thermal and
chemical denaturation. In thermal unfolding experiments,
stapled peptide 10 underwent a cooperative melting transition
with an approximate Tm of 55 °C, which is far higher than that
of unmodified peptide 9 (Tm = 19 °C) (Figure 3C).
Remarkably, the i + 4 + 7 stitched peptide 4 showed additional
thermal stability, as it retained more than 60% of its α-helicity
even at 90 °C. However, 4 did not have a cooperative melting
transition point in this temperature range, precluding a Tm
determination by this method. In an unfolding experiment
using guanidinium hydrochloride (GnHCl), 4 resists conforma-
tional disruption in response to increasing guanidinium salt up
to 6 M, whereas 10 showed a cooperative denaturing behavior
reaching a saturation point at ∼4 M GnHCl (Figure 3D).
One of the most significant features of stapled peptides is
their enhanced cellular uptake by endocytic vesicle trafficking.
To test whether stitched peptides retain this important
property, we performed fluorescence-activated cell sorting
(FACS) analyses on Jurkat cells exposed to the peptides at
37 °C. As expected, both fluorescein-labeled 4 and 10 caused
12318
dx.doi.org/10.1021/ja505141j | J. Am. Chem. Soc. 2014, 136, 12314−12322

Journal of the American Chemical Society
Article
observed upon treatment with the stapled analog 10 was
significantly decreased in comparison to that observed at 37 °C
(Figure 3F), which was expected from previous studies.33,34
Interestingly, however, the cellular uptake of the stitched analog
4 appears to be much less affected by a low temperature than its
stapled counterpart as confirmed by FACS (Figure 3F) as well
as confocal microscopy (Figure 3G,H). These results imply that
passive membrane diffusion may contribute to the cellular
uptake of the stitched peptide.
Importantly, it is unclear from the current data whether the
observed increase in cell penetration afforded by all-hydro-
carbon stitches relative to all-hydrocarbon staples is due to
increased helicity, greater hydrocarbon content in the cross-
link, or a combination of these factors. It is possible that the
difference in cell penetration is the result of a change in the
dominant mechanism of cell entry, and future studies will be
performed to further investigate the reasons for these
observations.
Encouraged by the markedly enhanced cell penetration of
stitched peptide 4, we investigated the effect of stitching on the
cell penetration of oligoarginine, a prototypical cell penetrating
peptide (CPP).51,52 Oligoarginines are intrinsically unstruc-
tured in solution, mainly due to charge repulsion among the
positively charged side-chain guanidinium groups. Therefore,
they represent one of the most challenging cases for
stabilization in an α-helical structure. In addition, oligoarginines
are among the most potent CPPs and have been frequently
used as cargo-delivery vectors, as reviewed by Brock51 and
Copolovici and colleagues.52 Limiting the use of oligo-L-
arginines in in vivo applications, however, is their high
susceptibility to enzymatic degradation. We were therefore
interested to determine the extent to which stitching could
improve the proteolytic stability of oligoarginine.
We prepared an oligo-L-arginine 12-mer (R12), one of the
most potent oligoarginine CPPs known,53 along with its i, i + 7
stapled (R12-Stp), i + 4 + 7 stitched (R12-Stc), and i + 4 + 7
unstitched (R12-Unstc) analogs (Table 1, entries XXIII−
XXVI). CD analyses indicated that unmodified R12 was
essentially devoid of secondary structure. Introduction of a
staple improved the helicity to 23% at room temperature (R12-
Stp), and stitching provided a significant additional increase to
59% (R12-Stc) (Figure 4A). On the contrary, R12-Unstc
displayed only 15% helicity, indicating that global rigidification
by macrocyclization plays a more profound role in helix-
stabilization than the localized conformational restriction
imparted by the introduction of α,α-disubstituted amino acids
(data not shown).
To further investigate the effect of stitching on the structural
dynamics of an oligo-L-arginine 12-mer, we performed fully
atomistic molecular dynamics (MD) simulations with explicit
water. Whereas the unmodified R12 exhibits no significant
secondary structure during the simulations, the computed
helicities for R12-Stp and R12-Stc are 21.3 and 56.4%,
respectively, in excellent agreement with our experimental
data (Figures 4B−D). Based on residue-specific helicity
calculations, R12-Stc displays helical character throughout the
entire peptide, whereas the helicity of R12-Stp is localized to
residues 5−8.
Figure 3. (A) CD spectra of stitched peptides at 20 °C. (B) CD
spectra of stitched peptide 4 (green) with its unmodified (blue) and
stapled (red) counterparts. (C and D) Conformational stability of
stitched peptide 4 against thermal (C) and guanidinium hydrochloride
denaturation (D) in comparison with its unmodified (blue) and
stapled (red) counterparts. Curves were not normalized to
demonstrate the different values of each peptide at the highest
temperature or guanidinium salt concentration. (E and F) Flow
cytometry profile of Jurkat cells treated with stitched peptide 4 in
comparison to its unmodified (blue) and stapled (red) analogs at 37
°C (E) and 4 °C (F). (G and H) Confocal microscope images of
U2OS cells treated at 4 °C with 2 μM of fluorescein (green)-labeled
10 (G) or 4 (H). The nuclei are stained by 4′6′-diamino-2-
phenylindole (DAPI, blue). Scale bars, 20 μm.
an increase in cell fluorescence, indicative of cellular uptake,
whereas the unmodified analog 9 did not show any detectable
cell permeability (Figure 3E).
Indeed, stitched peptide 4 displayed a higher degree of cell
permeability than its stapled counterpart 10, as evidenced by an
increase in mean and median cellular FITC signals. When the
experiment was performed at 4 °C, the cellular fluorescence
To further understand the molecular origin for the enhanced
helix nucleation of all-hydrocarbon-stitching compared to
stapling, we performed a solvation analysis to calculate the
number of water molecules around each residue of R12, R12-
Stp, and R12-Stc in their equilibrium structures. The average
12319
dx.doi.org/10.1021/ja505141j | J. Am. Chem. Soc. 2014, 136, 12314−12322

Journal of the American Chemical Society
Article
Figure 5. (A) Confocal microscope images of HeLa cells treated with
5 μM fluorescein (green)-labeled analogs at 37 °C for 1 h in the
presence of DAPI (blue). Scale bars, 20 μm. (B) Quantitative
comparison of cellular uptake of fluorescein-labeled peptides using
epifluorescence. Error bars represent the mean standard deviation of
two independent experiments. (C) Stability toward trypsin digestion.
R12-Stc displays a 20-fold longer half-life (t1/2) than R12-Stp.
Figure 4. (A) CD spectra of dodeca-L-arginine (R12) and its stapled
(R12-Stp) and stitched (R12-Stc) analogs at 20 °C. (B−D) The most
distributed conformations and the Ramachandran plots of the
ensemble structures of each analog. R12 adopts no defined secondary
structure, and R12-Stc displays a more extended α-helical region than
R12-Stp.
possibility that the higher cytosolic accumulation of R12-Stc
may be attributable to enhanced metabolic stability provided by
the presence of the stitch, which we assessed by examining the
ability of stitching to protect against proteolytic degradation.
Oligoarginines are extremely vulnerable to trypsin proteolysis,
as trypsin predominantly cleaves at the carboxyl side of
positively charged amino acids such as lysine and arginine. As
expected, after treatment with trypsin for 5 min at 25 °C at an
enzyme/substrate ratio of 1/800, the parent oligoarginine R12
was completely digested. However, under the same conditions,
41% of the stapled version and 94% of the stitched version
remained intact. Indeed, greater than 50% of the stitched
analog remained intact even after 60 min of trypsin exposure,
with a half-life 20-fold longer (t1/2 = 64 min) than the
corresponding stapled counterpart (t1/2 = 3.1 min) (Figure
5C). These results imply that greater helix stability is
accompanied by enhanced resistance to digestion by proteases.
The additional macrocylic bridge in the stitched analog most
likely physically blocks the access of the protease to the
cleavable residues and also prevents the peptide from adopting
the extended conformation required for insertion into the
enzyme active site.55
number of water molecules for R12 is 3.4 per residue, and
those water molecules are evenly distributed throughout the
peptide. For R12-Stp, the average number of water molecules is
decreased to 2.5 per residue, and we observed a modest
localized dehydration of the peptide backbone spanning
residues 4−8. R12-Stc displays extensive dehydration of the
backbone of residues 2−11, with an average of 1.8 water
molecules per residue. Because helix unfolding in an aqueous
environment is triggered primarily by the access of water to the
backbone hydrogen bonds,49,50,54 these results indicate that the
all-hydrocarbon staples and stitches shield the backbone
hydrogen bonds from competition by solvent water molecules.
Our experimental and computational analyses, therefore,
indicate that the enhanced α-helix stabilization conferred by
all-hydrocarbon stitching is due to improved conformational
rigidification as well as the creation of a more extended
hydrophobic microenvironment around the backbone hydro-
gen bonds.50
To examine the effect of peptide stitching on cell
penetration, we treated HeLa cells with fluorescein-labeled
analogs (5 μM) at 37 °C for 1 h and analyzed their cellular
uptake using confocal microscopy, which revealed that R12-Stc
fully retained the cell penetrating ability of R12 (Figure 5A).
Furthermore, epifluorescence microscopy analysis using a lower
concentration (1 μM) with a longer incubation time (8 h)
indicated that the cellular uptake of R12-Stp and R12-Stc was
increased by 21% and 75%, respectively, compared to R12
(Figure 5B). It is noteworthy that the stitched analog has the
smallest net positive charge, indicating that the number of
positively charged residues is not the only factor affecting cell
penetration of polycationic CPPs. We also considered the
CONCLUSIONS
■
This work demonstrates that ruthenium-mediated RCM
reactions on an α-helical peptide scaffold can be remarkably
selective for chemically equivalent but stereo- and regiochemi-
cally distinct olefins. The stitched α-helical peptides thus
generated feature exceptional thermal stabilities, and they also
exhibit dramatically enhanced stability against chemical
denaturation and proteolytic digestion. In addition, the stitched
peptides show greatly improved cell penetrating ability
compared to stapled peptides. Therefore, we believe stitched
peptides warrant further investigation as ligands for chemical
genetics applications, next-generation therapeutic agents, and
tools for macromolecule cargo delivery.
12320
dx.doi.org/10.1021/ja505141j | J. Am. Chem. Soc. 2014, 136, 12314−12322

Journal of the American Chemical Society
Article
(14) Cabezas, E.; Satterthwait, A. C. J. Am. Chem. Soc. 1999, 121,
3862.
(15) Cantel, S.; Isaad, A. L. C.; Scrima, M.; Levy, J. J.; DiMarchi, R.
D.; Rovero, P.; Halperin, J. A.; D’Ursi, A. M.; Papini, A. M.; Chorev,
M. J. Org. Chem. 2008, 73, 5663.
(16) Muppidi, A.; Doi, K.; Edwardraja, S.; Drake, E. J.; Gulick, A. M.;
Wang, H. G.; Lin, Q. J. Am. Chem. Soc. 2012, 134, 14734.
(17) Lau, Y. H.; de Andrade, P.; Quah, S.-T.; Rossmann, M.; Laraia,
ASSOCIATED CONTENT
* Supporting Information
General information, synthesis of the B5 amino acid, CD data,
mass spectrometry data, NMR methods and spectra, discussion
of NMR data, discussion of molecular modeling data, and MD
methods. This material is available free of charge via the
Internet at http://pubs.acs.org.
■
S
L.; Skold, N.; Sum, T. J.; Rowling, P. J. E.; Joseph, T. L.; Verma, C.;
̈
Hyvonen, M.; Itzhaki, L. S.; Venkitaraman, A. R.; Brown, C. J.; Lane,
̈
AUTHOR INFORMATION
■
D. P.; Spring, D. R. Chem. Sci. 2014, 5, 1804.
(18) Frost, J. R.; Vitali, F.; Jacob, N. T.; Brown, M. D.; Fasan, R.
ChemBioChem 2013, 14, 147.
(19) Haney, C. M.; Horne, W. S. J. Pept. Res. 2014, 20, 108.
(20) Zou, Y.; Spokoyny, A. M.; Zhang, C.; Simon, M. D.; Yu, H.; Lin,
Y. S.; Pentelute, B. L. Org. Biomol. Chem. 2014, 12, 566.
(21) Mazzier, D.; Peggion, C.; Toniolo, C.; Moretto, A. Biopolymers
2014, 102, 115.
Corresponding Authors
gregory_verdine@harvard.edu
sihyun@sookmyung.ac.kr
Present Addresses
∥College of Pharmacy, Dongguk University, 32 Dongguk-Ro,
Ilsandong-Gu, Goyang-Si, Gyeonggi-Do 420−820, Korea.
#Cheminformatics, Merck Research Laboratories, Boston, MA
02115, United States.
(22) Blackwell, H. E.; Grubbs, R. H. Angew. Chem., Int. Ed. 1998, 37,
3281.
○Chemical Biology and Proteomics Laboratory, The Salk
Institute for Biological Studies, La Jolla, CA 92037, United
States.
(23) Blackwell, H. E.; Sadowsky, J. D.; Howard, R. J.; Sampson, J. N.;
Chao, J. A.; Steinmetz, W. E.; O’Leary, D. J.; Grubbs, R. H. J. Org.
Chem. 2001, 66, 5291.
(24) Boal, A. K.; Guryanov, I.; Moretto, A.; Crisma, M.; Lanni, E. L.;
Toniolo, C.; Grubbs, R. H.; O’Leary, D. J. J. Am. Chem. Soc. 2007, 129,
6986.
⊥Center for Human Genetics Research, Massachusetts General
Hospital, Boston, MA 02114, United States.
◆Medical Scientist Training Program, Stony Brook University,
Stony Brook, NY 11794, United States.
(25) Chapman, R. N.; Dimartino, G.; Arora, P. S. J. Am. Chem. Soc.
2004, 126, 12252.
(26) Schafmeister, C. E.; Po, J.; Verdine, G. L. J. Am. Chem. Soc.
2000, 122, 5891.
(27) Kim, Y. W.; Verdine, G. L. Bioorg. Med. Chem. Lett. 2009, 19,
2533.
Author Contributions
∇These authors contributed equally.
Notes
The authors declare no competing financial interest.
(28) Kim, Y. W.; Kutchukian, P. S.; Verdine, G. L. Org. Lett. 2010, 12,
3046.
(29) Shim, S. Y.; Kim, Y. W.; Verdine, G. L. Chem. Biol. Drug Des.
ACKNOWLEDGMENTS
■
We thank Andriy Didovyk, Qian Chu, and Eric Smith for
assistance with the thermal melting data analysis, epifluor-
escence assay, and the editorial work, respectively. P.S.K. and
C.M.C. were supported by NSF graduate research fellowships.
A.C. was funded by the Harvard College Research Program.
This research was supported by the Harvard and Dana-Farber
Program in Cancer Chemical Biology and by a Korean
Research Foundation Grant funded by the Korean Government
(MOEFRD, Basic Research Promotion Fund)
(2012R1A2A01004687).
2013, 82, 635.
(30) Verdine, G. L.; Hilinski, G. J. Methods Enzymol. 2012, 503, 3.
(31) Verdine, G. L.; Hilinski, G. J. Drug Discovery Today: Technol.
2012, 9, e41.
(32) Walensky, L. D.; Bird, G. H. J. Med. Chem. [Online early access].
DOI: 10.1021/jm4011675 (accessed March 6, 2014).
(33) Walensky, L. D.; Kung, A. L.; Escher, I.; Malia, T. J.; Barbuto, S.;
Wright, R. D.; Wagner, G.; Verdine, G. L.; Korsmeyer, S. J. Science
2004, 305, 1466.
(34) Bernal, F.; Tyler, A. F.; Korsmeyer, S. J.; Walensky, L. D.;
Verdine, G. L. J. Am. Chem. Soc. 2007, 129, 2456.
(35) Moellering, R. E.; Cornejo, M.; Davis, T. N.; Del Bianco, C.;
Aster, J. C.; Blacklow, S. C.; Kung, A. L.; Gilliland, D. G.; Verdine, G.
L.; Bradner, J. E. Nature 2009, 462, 182.
(36) Grossmann, T. N.; Yeh, J. T. H.; Bowman, B. R.; Chu, Q.;
Moellering, R. E.; Verdine, G. L. Proc. Natl. Acad. Sci. U.S.A. 2012, 109,
17942.
(37) Bird, G. H.; Madani, N.; Perry, A. F.; Princiotto, A. M.; Supko, J.
G.; He, X.; Gavathiotis, E.; Sodroski, J. G.; Walensky, L. D. Proc. Natl.
Acad. Sci. U.S.A. 2010, 107, 14093.
(38) Chang, Y. S.; Graves, B.; Guerlavais, V.; Tovar, C.; Packman, K.;
To, K. H.; Olson, K. A.; Kesavan, K.; Gangurde, P.; Mukherjee, A.;
Baker, T.; Darlak, K.; Elkin, C.; Filipovic, Z.; Qureshi, F. Z.; Cai, H.;
Berry, P.; Feyfant, E.; Shi, X. E.; Horstick, J.; Annis, D. A.; Manning, A.
M.; Fotouhi, N.; Nash, H.; Vassilev, L. T.; Sawyer, T. K. Proc. Natl.
Acad. Sci. U.S.A. 2013, 110, E3445.
REFERENCES
■
(1) Hopkins, A. L.; Groom, C. R. Nat. Rev. Drug Discovery 2002, 1,
727.
(2) Guharoy, M.; Chakrabarti, P. Bioinformatics 2007, 23, 1909.
(3) Jochim, A. L.; Arora, P. S. Mol. BioSyst. 2009, 5, 924.
(4) Jochim, A. L.; Arora, P. S. ACS Chem. Biol. 2010, 5, 919.
(5) Osapay, G.; Taylor, J. W. J. Am. Chem. Soc. 1990, 112, 6046.
(6) Yu, C. X.; Taylor, J. W. Tetrahedron Lett. 1996, 37, 1731.
(7) Phelan, J. C.; Skelton, N. J.; Braisted, A. C.; McDowell, R. S. J.
Am. Chem. Soc. 1997, 119, 455.
(8) Taylor, J. W. Biopolymers 2002, 66, 49.
(9) Pease, J. H. B.; Storrs, R. W.; Wemmer, D. E. Proc. Natl. Acad. Sci.
U.S.A. 1990, 87, 5643.
(10) Jackson, D. Y.; King, D. S.; Chmielewski, J.; Singh, S.; Schultz, P.
G. J. Am. Chem. Soc. 1991, 113, 9391.
(39) Green, B. R.; Klein, B. D.; Lee, H. K.; Smith, M. D.; Steve
White, H.; Bulaj, G. Bioorganic & medicinal chemistry 2013, 21, 303.
(40) Hao, Y. J.; Wang, C.; Cao, B.; Hirsch, B. M.; Song, J.;
Markowitz, S. D.; Ewing, R. M.; Sedwick, D.; Liu, L. L.; Zheng, W. P.;
Wang, Z. H. Cancer Cell 2013, 23, 583.
(11) Leduc, A. M.; Trent, J. O.; Wittliff, J. L.; Bramlett, K. S.; Briggs,
S. L.; Chirgadze, N. Y.; Wang, Y.; Burris, T. P.; Spatola, A. F. Proc.
Natl. Acad. Sci. U.S.A. 2003, 100, 11273.
(12) Galande, A. K.; Bramlett, K. S.; Burris, T. P.; Wittliff, J. L.;
Spatola, A. F. J. Pept. Res. 2004, 63, 297.
(41) Bird, G. H.; Boyapalle, S.; Wong, T.; Opoku-Nsiah, K.; Bedi, R.;
Crannell, W. C.; Perry, A. F.; Nguyen, H.; Sampayo, V.; Devareddy,
(13) Kumita, J. R.; Smart, O. S.; Woolley, G. A. Proc. Natl. Acad. Sci.
U.S.A. 2000, 97, 3803.
12321
dx.doi.org/10.1021/ja505141j | J. Am. Chem. Soc. 2014, 136, 12314−12322

Journal of the American Chemical Society
Article
A.; Mohapatra, S.; Mohapatra, S. S.; Walensky, L. D. J. Clin. Invest.
2014, 124, 2113.
(42) Chen, Y. H.; Yang, J. T.; Martinez, H. M. Biochemistry 1972, 11,
4120.
(43) Favrin, G.; Irback, A.; Samuelsson, B.; Wallin, S. Biophys. J.
2003, 85, 1457.
(44) MacroModel, 9.1 ed.; Schrodinger LLC: New York, NY, 2005.
(45) Still, W. C.; Tempczyk, A.; Hawley, R. C.; Hendrickson, T. J.
Am. Chem. Soc. 1990, 112, 6127.
(46) Clarke, P. A.; Grist, M.; Ebden, M.; Wilson, C.; Blake, A. J.
Tetrahedron 2005, 61, 353.
(47) Clark, J. S.; Marlin, F.; Nay, B.; Wilson, C. Org. Lett. 2003, 5, 89.
(48) Toniolo, C.; Crisma, M.; Formaggio, F.; Peggion, C. Biopolymers
2001, 60, 396.
(49) Vila, J. A.; Ripoll, D. R.; Scheraga, H. A. Proc. Natl. Acad. Sci.
U.S.A. 2000, 97, 13075.
(50) Gao, J. M.; Bosco, D. A.; Powers, E. T.; Kelly, J. W. Nat. Struct.
Mol. Biol. 2009, 16, 684.
(51) Brock, R. Bioconjugate Chem. 2014, 25, 863.
(52) Copolovici, D. M.; Langel, K.; Eriste, E.; Langel, U. ACS Nano
2014, 8, 1972.
(53) Kosuge, M.; Takeuchi, T.; Nakase, I.; Jones, A. T.; Futaki, S.
Bioconjugate Chem. 2008, 19, 656.
(54) Avbelj, F.; Luo, P.; Baldwin, R. L. Proc. Natl. Acad. Sci. U.S.A.
2000, 97, 10786.
(55) Tyndall, J. D.; Nall, T.; Fairlie, D. P. Chem. Rev. 2005, 105, 973.
12322
dx.doi.org/10.1021/ja505141j | J. Am. Chem. Soc. 2014, 136, 12314−12322