Light-responsive bicyclic peptides

Article abstract of DOI:10.1039/c7ob03178e

In this paper, we describe a method for the synthesis of light-responsive (LR) bicyclic macrocycles from linear peptides composed of 20 natural amino acids. Small molecules, peptide macrocycles, and protein conjugates that reversibly turn their function on and off in response to visible light enabled the fields of photopharmacology and optochemical genetics. Bioactive LR molecules could be produced by grafting azobenzene or other LR-structures onto molecules with known biological functions (e.g., alpha-helical peptides). It is also possible to discover such LR ligands de novo by selecting compounds with a desired function - such as binding to a target - from a library of LR-compounds or a genetically-encoded (GE) library of LR-macrocycles. The bicyclic topology of ligands offers added value such as improved binding and stability when compared to monocyclic peptides, but approaches for the design of bicyclic light-responsive architectures are limited. To address this need, we developed a tridentate C2-symmetric hydroxyl amine and di-chlorobenzene containing azobenzene (HADCAz) LR-linker with two orthogonally reactive functionalities (chlorobenzyl and hydroxylamine) to convert a linear unprotected peptide into a bicyclic peptide in a one-pot, two-step reaction. This linker reversibly isomerizes from the trans to cis form upon irradiation with blue light (365 nm). The resulting bicyclic peptide contains two loops of amino acids, one of which is constrained with an azobenzene moiety that can change the conformation in response to visible light. A scalable synthetic route to the HADCAz linker allowed us to demonstrate its application in multiple synthetic bicyclic peptides with loops that contain 2-5 amino acids.

Full text of DOI:10.1039/c7ob03178e

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
Light-responsive bicyclic peptides†
Mohammad R. Jafari,a Hongtao Yu,a,b Jessica M. Wickware,a Yu-Shan Lin a,b and
Cite this: DOI: 10.1039/c7ob03178e
a
Ratmir Derda
*
In this paper, we describe a method for the synthesis of light-responsive (LR) bicyclic macrocycles from
linear peptides composed of 20 natural amino acids. Small molecules, peptide macrocycles, and protein
conjugates that reversibly turn their function on and off in response to visible light enabled the fields of
photopharmacology and optochemical genetics. Bioactive LR molecules could be produced by grafting
azobenzene or other LR-structures onto molecules with known biological functions (e.g., alpha-helical
peptides). It is also possible to discover such LR ligands de novo by selecting compounds with a desired
function—such as binding to a target—from a library of LR-compounds or a genetically-encoded (GE)
library of LR-macrocycles. The bicyclic topology of ligands offers added value such as improved binding
and stability when compared to monocyclic peptides, but approaches for the design of bicyclic light-
responsive architectures are limited. To address this need, we developed a tridentate C2-symmetric
hydroxyl amine and di-chlorobenzene containing azobenzene (HADCAz) LR-linker with two orthogonally
reactive functionalities (chlorobenzyl and hydroxylamine) to convert a linear unprotected peptide into a
bicyclic peptide in a one-pot, two-step reaction. This linker reversibly isomerizes from the trans to cis
form upon irradiation with blue light (365 nm). The resulting bicyclic peptide contains two loops of amino
acids, one of which is constrained with an azobenzene moiety that can change the conformation in
response to visible light. A scalable synthetic route to the HADCAz linker allowed us to demonstrate its
application in multiple synthetic bicyclic peptides with loops that contain 2–5 amino acids.
Received 25th December 2017,
Accepted 13th March 2018
DOI: 10.1039/c7ob03178e
rsc.li/obc
design would require significant efforts. The discovery of
“static” bioactive ligands from combinatorial chemical
Introduction
Photoswitchable ligands that reversibly change their affinity to libraries is a cornerstone of the modern pharmaceutical indus-
a specific molecular target in response to light are powerful try and chemical biology. It is thus attractive to repurpose the
tools for the investigation of biological systems with spatial power of combinatorial and genetically-encoded screening to
and temporal resolution. The emerging fields of photopharma- develop a versatile suite of discovery strategies for light-respon-
cology1 and optochemical genetics2 are empowered by the dis- sive bioactive molecules.
covery of novel light-responsive molecules that can be actuated
Peptides and their chemical derivatives constitute many of
from the “bioactive” to “inactive” state with visible light. The the biologically active compounds such as hormones, neuro-
rational combination of photochromic compounds such as transmitters, toxins and therapeutic agents. There are more
azobenzenes, spiropyrans and diarylethenes with known bio- than 60 FDA-approved peptide therapeutics on the market and
active small molecules, peptides, and proteins has successfully more than 140 peptide candidates in clinical trials.18 Many
yielded light-responsive ligands for tubulin,3 NMDA recep- therapeutic leads have cyclic and bicyclic topology; 23 out of
tors,4,5 glucagon-like peptide-1 receptor (GLP-1R),6 TRPV1,7 the 60 FDA-approved peptide drugs have a cyclic structure.19
ion channels,8–10 specific DNA sequences,11,12 Bcl-XL,13 The advantages of bicyclic topology include high binding
β-adaptin,14 and histone deacetylases (HDACs)15,16 (for an affinity,20,21 increased stability to enzymatic hydrolysis,20,22
in-depth review see ref. 17). Light-responsive ligands have not and improved permeability through the cell membrane,23,24
been designed yet for many targets, and such a molecular for cyclic peptides when compared to linear peptide precur-
sors. The macrocyclization of a peptide through an inter-
molecular reaction with small molecule linkers or “linchpins”
is now used as a common, versatile strategy for introducing
aDepartment of Chemistry, University of Alberta, Edmonton, AB T6G2G2, Canada.
light-responsiveness into the peptide.25 For example, the
E-mail: ratmir@ualberta.ca
bDepartment of Chemistry, Tufts University, Medford, Massachusetts 02155, USA
chemical modification of α-helical and β-hairpin26,27 peptides
with azobenzene is now used as a general, versatile solution
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
c7ob03178e
for the conversion of bioactive peptides to LR-ligands. The use
This journal is © The Royal Society of Chemistry 2018
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
of peptides composed of unprotected natural amino acids and bidentate C2V symmetric linchpins to C2V linchpins that
translationally-made polypeptide sequences in such modifi- contain LR-moieties, such as Azb, to form LR-cyclic peptide
cations is attractive because the peptides of natural origin macrocycles. An example of such
a linchpin is BSBCA
such as expressed proteins, phage-displayed peptide libraries, (Fig. 1A).12,25 Devising a LR-linchpin with three attachment
or mRNA-displayed libraries, are high-value “starting points is not obvious, because all functionalities that undergo
materials”.
Million-to-billion scale genetically-encoded light-induced
isomerization–such
as
azobenzenes,
libraries of LR-ligands can be produced simply by grafting a diarylethenes,33–35 spiropyranes36,37–have lower C2V symmetry.
light-responsive Azb-linchpin onto two thiol-reactive Wegner, Heinis and co-workers reported a synthesis of a puta-
groups.28,29 The groups of Heinis,28 Ito30 and Derda29 have tive C3V-symmetric linchpin for the modification of peptides
implemented such strategies to modify phage-displayed with three cysteines.38 The authors, however, admitted that the
peptides,31 phage-displayed libraries of peptides and mRNA- use of two tandem Azb functionalities in a linchpin was not
displayed libraries of peptides. These libraries were shown to optimal because these moieties isomerize independently,
be a productive source for the discovery of macrocycles that yielding a poor light-switching performance. The use of a non-
bind streptavidin with binding that can be either turned C3V symmetric linchpin with three identical reactive groups is
“on”28 or “off”29 by irradiation with light.
Bicyclic peptides are high-value molecular scaffolds for the more constitutional isomers.
impractical because its reaction with a peptide yields three or
discovery of potent ligands for diverse protein targets. To date,
A viable topological solution for the synthesis of LR-bicycles
there is no general topological solution for the synthesis of uses a C2V symmetric linchpin with 2 + 1 reactive groups that
light-responsive bicyclic peptides. To form n cycles in a target 2 and 1 orthogonally-reactive functionalities in a
peptide, n + 1 points should be connected through a linchpin peptide. Two types of LR-linchpins (LRLs) can be designed: (i)
with n + 1 reactive functionalities. For example, di- and tri-bro- an LRL with two dynamic loops, both containing an Azb
momethyl benzene (DBMB and TBMB, respectively, Fig. 1A) photo-switch and (ii) an LRL with one dynamic loop and one
are classical static C2V and C3V symmetrical linchpins employed static loop. In this paper we implemented a C2V linchpin that
by Timmerman and coworkers to convert linear unprotected can form the latter bicycle with dynamic and static loops. We
peptide chains with two or three cysteines into mono- and envision that one peptide ring can dynamically change its
bicyclic peptides.32 It is conceptually simple to upgrade static affinity to the receptor of interest, while the second ring can be
reserved to introduce other static properties into the molecule,
such as cell permeability (Fig. 1B and C).
Results
In our design, the LR-linchpin should meet three criteria: (i) it
should include functionalities that can selectively react with
two orthogonally reactive handles present in or derived from
natural amino acid side-chains. Herein, we focus on the thiol
functionalities of cysteine and the glyoxal functionality pro-
duced from N-terminal serine by mild oxidation with sodium
periodate.39 (ii) The linker should have a half-life of thermal
relaxation of more than 10 min after irradiation to provide
sufficient time for potential biological studies, such as ligand–
receptor interactions. (iii) The LRL should switch in response
to visible light (λmax of >360 nm), because most biological
systems are sensitive to light with λ below 350 nm.40
We synthesized four different azobenzene cores (Fig. 2A,
1a–1d) which carried the suitable functionalities (–OH, –NH2,
and Ar–OH) for further installation of cysteine and glyoxal-
reactive groups. We did not observe any changes in the spectra
after the irradiation of compounds 1a and 1b with 380 nm and
470 nm LEDs, respectively, which is most likely due to the fast
Fig. 1 Strategy and linchpins for the synthesis of cyclic and bicyclic thermal relaxation of azobenzene (Fig. S1A and B†). This is
consistent with other published studies,40 that found the
single or double alkylation of one or both aromatic amines at
peptides with dynamic and/or static surfaces. (A) Thiol-reactive linchpins
used for the generation of monocyclic (DBMB), bicyclic (TBMB), and
monocyclic (BSBCA) LR-peptides. (B) Scheme of the synthesis of bicyclic
LR-peptides from an unprotected linear peptide made of 20 natural
either end of Azb to result in fast switching. Compound 1c
exhibited the desired properties: its trans to cis isomerization
amino acids. (C) A bicyclic LR-peptide may contain a static and a
dynamic surface, each targeting a specific binding function.
was induced by irradiation with 380 nm light and the thermal
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Organic & Biomolecular Chemistry
Paper
relaxation of the cis form back to the trans form occurred with
a half-life of 43 min (Fig. 2C).
The presence of an amide linkage in 1c was crucial for an
optimal light-response because core 1d, in which the amide
was replaced by an ester bond, showed the maximum absorp-
tion at 320 nm which is in the bio-incompatible UV region
(Fig. S1D†). Based on the structure–activity relationship of the
aforementioned azobenzenes, we believed that the arrange-
ment of functional groups in azobenzene 1c yielded near
optimal photochemical properties (Fig. 2C and D). Installing
the desired functional groups onto the Azb core 1c (Fig. 2B)
yielded hydroxyl amine, di-chlorobenzene containing azo-
benzene (HADCAz) in seven steps from dimethyl-5-(amino)iso-
phthalate in 18.5% overall yield (Scheme 1). HADCAz exhibited
maximum absorption at 350 nm, which is close to its parent
Azb core 1c. HADCAz was successfully switched to 71% cis-
isomer in its photostationary state after irradiation with
365 nm light, and relaxed thermally to the trans-isomer with a
half-life of ∼60 min (Fig. 3).
To explore the utility of HADCAz for the bicyclization of
peptides, we investigated the reaction of HADCAz with the
peptide sequences SXnCRRRRC where Xn was W, SW, KSW, or
DKSW (Fig. 4A), which included the major reactive functional-
ities present in natural amino acids (amine, hydroxyl, car-
boxylic acid, and indole). The N-terminal serine can be con-
verted to glyoxal, whereas cysteines flank a tetra-arginine
sequence that has been used by Pei and co-workers to enhance
the cell permeability of bicyclic peptides.41 We found that
forming the first cycle by the reaction of Cys with chloroben-
zyls was not a suitable strategy, because the resulting sulfides
are significantly more susceptible to periodate oxidation than
the disulfides (Fig. 4C).42 Adducts with +16 Da and +32 Da in
the mass spectrum confirmed the oxidation of sulfides
Fig. 2 Synthesis of Azb cores and LR-linchpin HADCAz. (A) Potential
Azb precursors to bicyclic macrocycles and their light-induced isomeri-
zation properties. (B) Minimalistic design of the light-responsive linker
HADCAz for forming bicyclic macrocycles (see Scheme 1 for a complete
synthetic route to HADCAz). (C) Light-induced isomerization properties
of the precursor molecules shown in (A). (D) Absorption spectra of
HADCAz before and after irradiation with 380 nm light.
Scheme 1 Synthetic pathway for HADCAz.
This journal is © The Royal Society of Chemistry 2018
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
oxime ligation between the purified peptide glyoxal and
HADCAz in water or water–organic mixtures such as 50%
aqueous DMF (pH 4.5) was unsuccessful and yielded no appar-
ent conversion (Fig. S3D–F†). The aniline catalyst (100 mM,
pH 4.5)43 accelerated these reactions but, unfortunately, it also
gave rise to the nucleophilic substitution of the benzyl chlor-
ides of HADCAz by aniline (Fig. 4D and Fig. S3G–H†). The
most optimal oxime bond formation conditions involved
adding the lyophilized peptide glyoxal to a DMF solution of
HADCAz (1.2 eq.). The reaction was complete in 2 h at 55 °C
(Fig. 4A) or after 24 h of incubation at room temperature
(Fig. S3A–C†). Once the oxime bond was formed, the sub-
sequent bicyclization of 7 was straightforward. Introducing
three equivalents of TCEP into the reaction mixture resulted in
the double intramolecular nucleophilic substitution of both
chlorine atoms with the thiols yielding the quantitative “peak-
to-peak conversion” of monocycle 7 to bicycle 8 (Fig. 4B).
HRMS confirmed the identity of the bicycle.
We note that there exists an alternative cyclization pathway
that starts with an alkylation of thiols in intermediate 6 by
HADCAz (Fig. 4E). We observed that subsequent cyclization via
the intramolecular formation of the oxime bond was not repro-
ducible as it often yielded a significant amount of insoluble
byproducts. It is tempting to suggest that the difference in the
efficiency in bicyclization via 7→8 vs. 9→8 pathways originates
from different geometric constraints of nucleophilic substi-
tution and addition reactions. It is possible that the exo-tet
Fig. 3 Thermal relaxation of HADCAz. (A) LCMS traces of the thermal
relaxation of HADCAz after irradiation with 365 nm light. (B) Scheme of
the isomerization of trans-HADCAz to the cis form. (C) Fitting the
thermal relaxation data to the first order kinetics model determined the
cis-ratio to be 71% in its photostationary state after irradiation, and the
thermal relaxation half-life to be 59 min.
(Fig. S4†). In contrast, the oxidation of a disulfide peptide cyclization of 7 to yield 8 is geometrically relaxed because it
(Fig. 4A) with sodium periodate at pH 7.0 yielded a peptide involves an acceptor on the freely-rotating benzylic bonds. In
glyoxal with an intact disulfide bond. Our attempt to perform contrast, the formation of 8 from precursor 9 requires con-
Fig. 4 Synthetic pathways for the generation of LR-bicyclic peptides. (A) Successful route to bicyclization starts from the oxidative conversion of
N-terminal serine to glyoxal followed by the reaction with HADCAz to form an oxime bond; subsequent reduction of the disulfide bond with TCEP
triggers the intramolecular formation of two thioether bonds to close both cycles. (B) Conversion of monocycle 7 to bicycle 8 occurs as quantitative
“peak-to-peak” conversion when monitored by HPLC. Mass-spectrometry confirmed the identity of the bicycle. (C–E) Plausible but unsuccessful
routes to the same product. (C) Formation of the first cycle through thioether bonds is not a suitable first step because subsequent periodate oxi-
dation results in the oxidation of the thioethers. (D) Oxime formation in the presence of aniline results in the undesired substitution of either one or
both chlorines by aniline. (E) Although it is possible to convert intermediate 6 to monocyclic precursor 9, its cyclization to the desired product 8 was
inconsistent, and yielded a significant amount of uncharacterized insoluble byproducts (possibly oligomeric analogs of 8).
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 1 Isolated yield, thermal relaxation half-life, and the ratio of the
cis-isomer in the photostationary state after irradiation of peptides
3a–3d
in the trans-state and cis-state that shed some light on the
potential changes in the conformation of the peptides and the
azobenzene core in response to illumination. Fig. S14 and
S15† contain selected structures and Derda_PDB.zip has PDB
of all structures. The most remarkable observation was that
the 198/200 MD simulations of “dark” 3a (ground state) con-
verged on structures that contained the cis-configuration for
the aromatic amide bond nearest to the azobenzene. In “light”
3a structures, all amide bonds were of trans-configuration
(Fig. 5). The same amide bond was of cis-configuration with a
73/200 conformation of simulated “dark” 3b structures
(Fig. S16†). “Dark” 3c and 3d contained cis amide bonds in
Relaxation
half-time
(min)
cis-Isomer in the
photostationary
state (%)
Isolated
yield (%)
Sequence
3a
3b
3c
3d
SWCRRRRC
11
74
63
68
144.3 18.7
243.1 33.4
117.2 21.0
29.5 1.3
86.2 2.0
91.0 1.9
81.9 1.5
74.9 3.3
SSWCRRRRC
SKSWCRRRRC
SDKSWCRRRRC
strained endo-trig cyclization with the acceptor orbital being only 4/200 conformations (Fig. S16†). UV spectroscopy corro-
part of an oxaloyl functionality that contains no rotating borated that the NvN bond in the “dark” state 3a is indeed
bonds (due to extended conjugation).
trans (Fig. S5†). ROESY NMR spectroscopy confirmed that the
Optimized bicyclization conditions (Fig. 4A) occurred repro- aromatic amide bond is indeed uniquely cis in “dark” 3a (Fig. 5C
ducibly for peptides with the structure of SXnCR4C, where n > 1. and S17†), whereas it is trans in “dark” 3b–3d (Fig. S17 and
For Xn = SW, KSW or DSKW, the isolated yields were 74%, 63% page S81 of the ESI†). In 1D NMR spectroscopy, the proton of
and 68%, respectively (Table 1). HRMS and NMR spectroscopy this amide bond exhibited a change in the chemical shift from
confirmed the structures of all products (Fig. S2A–C†). The iso- 9.4 ppm in 3a to 9.7 ppm in 3b, 3c and 3d (see the ESI,
lated yield for the peptide sequence SWCRRRRC (n = 1) was page S77†). Combined evidence from the MD simulation and
lower (11%) and the reaction contained a mixture of un- spectroscopy, thus, suggest that the smallest bicycle 3a cannot
identified byproducts (Fig. S2D†). The low yield of the reaction simultaneously contain a trans NvN bond and all amide
for this sequence highlights the lower limit of the bicycliza- bonds in the trans-configuration. The aromatic amide bond,
tion. The distance between the cysteine and glyoxal functional- which has the lowest rotational barrier, has to adopt the
ities has to be at least two amino acids. To investigate the cis-conformation in the “dark” trans (NvN) state. As 3a is
upper limit, we foresee studying the bicyclization with several photoexcited to the cis (NvN) state, the backbone geometry is
peptide sequences SXnCR4C where Xn is 10–20 amino acids.
relaxed and the aromatic amide bond can adopt the trans-
Bicyclic products 3a–3d successfully switched to their cis- conformation. The relaxation of the amide bond predicted by
isomers after irradiation with 365 nm light, as indicated by a MD simulation to the trans state upon light irradiation of 3a
decrease in their absorption at 350 nm (Fig. S5†). We deter- could be further confirmed in the future by NMR spectroscopy
mined the ratio of the cis-isomer to trans-isomer for each of the irradiated bicycle 3a.
peptide by LCMS in time intervals after irradiation (Fig. S6A†).
Fitting the data to a first order kinetics model determined the
relative amount of the cis isomer in the photostationary state
(Fig. S6B†). Interestingly, we found that the ratio of the cis-
Discussion
isomer in the photostationary state for bicyclic peptides 3a–3c In conclusion, we developed a C2V symmetric, tridentate
is higher than that for HADCAz alone (Table 1). The thermal LR-linchpin that can react with peptides of the structure
relaxation half-life for 3a, 3b and 3c was 144, 243, and SXnCXmC to form LR-bicyclic macrocycles. The reaction occurs
117 min, respectively, which was higher than the half-life of site-specifically on unprotected peptides consisting of only
HADCAz (59 min). The ratio of the cis-isomer in the photosta- natural amino acids and cyclization results in satisfactory
tionary state for 3d was 75%, which was close to the cis-ratio of yields for peptides that contain a medium sized ring (n > 1).
HADCAz; however, its thermal relaxation half-life dropped to The dynamic loop of the bicyclic products (SXnC) contained an
29 min. The higher cis-ratio and relaxation half-life of the pro- azobenzene moiety which was successfully switched to its cis-
ducts could be because of greater ring strain in 3a–3c. We isomer in response to 365 nm light for n = 1–4. The number of
suspect that this modulation of photoswitching could also be amino acids in the loop modulates the switching properties of
sequence-specific, but the evaluation of different sequences the product. The static ring of the molecule harbors a tetra-
extends beyond the scope of this paper.
arginine sequence which has been shown to increase the per-
To understand the changes in the conformation of peptides meability of bicyclic peptides into mammalian cells. We were
upon irradiation, we employed 100 ps molecular dynamics not able to perform oxime ligation under aqueous conditions,
(MD) simulation in explicit TIP4P water for peptides 3a–3d and the current HADCAz bicyclization reaction is incompatible
with the NvN bond constrained as the cis-state (“dark state”) with potentially attractive applications such as the modifi-
or trans-state (“light state”) (details of the simulation are avail- cation of proteins or phage-displayed peptides.29,31 Further
able in the ESI†). Although bicycles 3a–3d do not exhibit a investigation will be required to identify suitable aqueous reac-
well-defined structure to be detected by NMR, MD simulations tion conditions, or to integrate another existing N-terminal
of bicycles 3a–3d predict an ensemble of structures with Azb ligation method into our reaction.44
This journal is © The Royal Society of Chemistry 2018
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 5 Molecular dynamics (MD) analysis of the photoisomerization of the smallest bicycle 3a. (A) Conformational preferences of dihedral angles
adjacent to the core NvN functionality in the HADCAz crosslinker in azo-trans-3a (“dark 3a”) (B) after 100 MD simulations in TIP4P water. The struc-
tures of both HADCAz and the peptide backbone diverged in 100 simulations (see Fig. S14,† and PDB files in Derda_PDB.zip in the ESI†). Interestingly,
every structure enforced the cis-configuration for the aromatic amide adjacent to HADCAz. The same MD simulation predicted the trans-amide bond
in dark 3b, 3c and 3d (Fig. S15 and S16†). (C) ROESY NMR spectroscopy confirmed the cis-configuration of the amide bond in dark 3a: (i) we observed
NOE interactions between (H1,H2) and (H4,H5), which is possible only in the cis-amide configuration; (ii) we observed no (H1,H2)–(H3) interactions
present in 3b, 3c and 3d that contain a trans-amide bond (Fig. S16, S17 and NOE summary on page S81†). (D) MD simulation of azo-cis-3a (“light 3a”)
simulations in TIP4P water converged on well-defined clusters that lock azobenzene in two possible configurations. (E) Representative structures of
cis-3a predicted by MD analysis (see Movie S1† for all structures). Every structure has a trans-configuration for the aromatic amide bond.
The requirement for the cis amide bond geometry in the
ground state of 3a detected by MD simulation and NMR spec-
Acknowledgements
troscopy explains the dramatic decrease in the yield of the bicy- This work was supported by research grants from the National
clization reaction (11%, Table 1 and Fig. S13†). The relaxation Science and Engineering Research Council of Canada (NSERC,
of the geometric constraints in the photoexcited state suggests #492943), NSERC Accelerator Supplement, and the Alberta
that the yield of this bicyclization with the HADCAz linker in Glycomics Centre. Infrastructure support was provided by the
small peptides could be increased if the reaction is performed Canadian Foundation for Innovation (CFI) New Leaders
under light irradiation, forcing the HADCAz linker to adopt Opportunity. We thank Mark Miskolzie and Ryan McKay for
the cis-conformation. The platform we described in this paper assistance in NMR spectra acquisition and interpretation and
could be employed in the future for the development of Dr. Nicholas J. Bennett for assistance with the alignment and
ligands that dynamically regulate cell functions via binding to visualization of PDB structures. JMW was supported by an
intracellular signalling components. Such bicyclic LR-inhibi- Alberta Innovates—Summer Studentship.
tors will expand the growing toolbox of optochemical genetics2
and photo-pharmacology,1 and permit the investigation of cell
References
signaling pathways that are not currently possible to study
using conventional tools.
1 W. A. Velema, W. Szymanski and B. L. Feringa, J. Am. Chem.
Soc., 2014, 136, 2178.
2 T. Fehrentz, M. Schonberger and D. Trauner, Angew. Chem.,
Int. Ed., 2011, 50, 12156.
Conflicts of interest
The corresponding author of this publication serves as the
founder and CEO of 48Hour Discovery Inc.
3 M. Borowiak, W. Nahaboo, M. Reynders, K. Nekolla,
P. Jalinot, J. Hasserodt, M. Rehberg, M. Delattre, S. Zahler,
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Organic & Biomolecular Chemistry
Paper
A. Vollmar, D. Trauner and O. Thorn-Seshold, Cell, 2015, 23 E. A. Villar, D. Beglov, S. Chennamadhavuni, J. A. Porco Jr.,
162, 403.
D. Kozakov, S. Vajda and A. Whitty, Nat. Chem. Biol., 2014,
10, 723.
4 S. Berlin, S. Szobota, A. Reiner, E. C. Carroll,
M. A. Kienzler, A. Guyon, T. Xiao, D. Trauner and 24 Q. Chu, R. E. Moellering, G. J. Hilinski, Y.-W. Kim,
E. Y. Isacoff, eLife, 2016, 5, DOI: 10.7554/eLife.12040.
T. N. Grossmann, J. T. H. Yeh and G. L. Verdine,
5 L. Laprell, E. Repak, V. Franckevicius, F. Hartrampf,
MedChemComm, 2015, 6, 111.
J. Terhag, M. Hollmann, M. Sumser, N. Rebola, 25 D. C. Burns, F. Zhang and G. A. Woolley, Nat. Protoc., 2007,
D. A. DiGregorio and D. Trauner, Nat. Commun., 2015, 6, 8076. 2, 251.
6 J. Broichhagen, T. Podewin, H. Meyer-Berg, Y. von Ohlen, 26 A. Aemissegger, V. Krautler, W. F. van Gunsteren and
N. R. Johnston, B. J. Jones, S. R. Bloom, G. A. Rutter, D. Hilvert, J. Am. Chem. Soc., 2005, 127, 2929.
A. Hoffmann-Roder, D. J. Hodson and D. Trauner, 27 S. L. Dong, M. Loweneck, T. E. Schrader, W. J. Schreier,
Angew. Chem., Int. Ed., 2015, 54, 15565.
W. Zinth, L. Moroder and C. Renner, Chem. – Eur. J., 2006,
12, 1114.
28 S. Bellotto, S. Chen, I. Rentero Rebollo, H. A. Wegner and
C. Heinis, J. Am. Chem. Soc., 2014, 136, 5880.
7 J. A. Frank, M. Moroni, R. Moshourab, M. Sumser,
G. R. Lewin and D. Trauner, Nat. Commun., 2015, 6, 7118.
8 M. Volgraf, P. Gorostiza, S. Szobota, M. R. Helix,
E. Y. Isacoff and D. Trauner, J. Am. Chem. Soc., 2007, 129, 29 M. R. Jafari, L. Deng, P. I. Kitov, S. Ng, W. L. Matochko,
260.
K. F. Tjhung, A. Zeberoff, A. Elias, J. S. Klassen and
R. Derda, ACS Chem. Biol., 2013, 9, 443.
30 M. Z. Liu, S. Tada, M. Ito, H. Abe and Y. Ito, Chem.
Commun., 2012, 48, 11871.
9 M. Volgraf, P. Gorostiza, R. Numano, R. H. Kramer,
E. Y. Isacoff and D. Trauner, Nat. Chem. Biol., 2006, 2, 47.
10 D. L. Fortin, M. R. Banghart, T. W. Dunn, K. Borges,
D. A. Wagenaar, Q. Gaudry, M. H. Karakossian, T. S. Otis, 31 M. R. Jafari, J. Lakusta, R. J. Lundgren and R. Derda,
W. B. Kristan, D. Trauner and R. H. Kramer, Nat. Methods,
2008, 5, 331.
11 A. M. Caamano, M. E. Vazquez, J. Martinez-Costas,
Bioconjugate Chem., 2016, 27, 509.
32 P. Timmerman, J. Beld, W. C. Puijk and R. H. Meloen,
ChemBioChem, 2005, 6, 821.
L. Castedo and J. L. Mascarenas, Angew. Chem., Int. Ed., 33 T. Fukaminato, T. Hirose, T. Doi, M. Hazama, K. Matsuda
2000, 39, 3104. and M. Irie, J. Am. Chem. Soc., 2014, 136, 17145.
12 L. Guerrero, O. S. Smart, G. A. Woolley and R. K. Allemann, 34 O. Babii, S. Afonin, M. Berditsch, S. Reisser,
J. Am. Chem. Soc., 2005, 127, 15624.
13 S. Kneissl, E. J. Loveridge, C. Williams, M. P. Crump and
R. K. Allemann, ChemBioChem, 2008, 9, 3046.
P. K. Mykhailiuk, V. S. Kubyshkin, T. Steinbrecher,
A. S. Ulrich and I. V. Komarov, Angew. Chem., Int. Ed., 2014,
53, 3392.
14 A. Martin-Quiros, L. Nevola, K. Eckelt, S. Madurga, 35 C. C. Warford, V. Lemieux and N. R. Branda, in Molecular
P. Gorostiza and E. Giralt, Chem. Biol., 2015, 22, 31. Switches, 2nd edn, 2011, vol. 1, p. 1.
15 S. A. Reis, B. Ghosh, J. A. Hendricks, D. M. Szantai-Kis, 36 L. Chen, J. Wu, C. Schmuck and H. Tian, Chem. Commun.,
L. Tork, K. N. Ross, J. Lamb, W. Read-Button, B. X. Zheng, 2014, 50, 6443.
H. T. Wang, C. Salthouse, S. J. Haggarty and R. Mazitschek, 37 R. Klajn, Chem. Soc. Rev., 2014, 43, 148.
Nat. Chem. Biol., 2016, 12, 317.
38 S. Bellotto, R. Reuter, C. Heinis and H. A. Wegner, J. Org.
16 W. Szymanski, M. E. Ourailidou, W. A. Velema, F. J. Dekker
and B. L. Feringa, Chem. – Eur. J., 2015, 21, 16517.
17 R. J. Mart and R. K. Allemann, Chem. Commun., 2016, 52,
12262.
Chem., 2011, 76, 9826.
39 K. F. Geoghegan and J. G. Stroh, Bioconjugate Chem., 1992,
3, 138.
40 A. A. Beharry and G. A. Woolley, Chem. Soc. Rev., 2011, 40,
18 K. Fosgerau and T. Hoffmann, Drug Discovery Today, 2015,
20, 122.
4422.
41 W. Lian, B. Jiang, Z. Qian and D. Pei, J. Am. Chem. Soc.,
19 P. Vlieghe, V. Lisowski, J. Martinez and M. Khrestchatisky,
Drug Discovery Today, 2010, 15, 40.
2014, 136, 9830.
42 B. J. Evans, J. T. Doi and W. K. Musker, J. Org. Chem., 1990,
20 C. Heinis, T. Rutherford, S. Freund and G. Winter, Nat.
Chem. Biol., 2009, 5, 502.
55, 2580.
43 A. Dirksen, T. M. Hackeng and P. E. Dawson, Angew. Chem.,
21 J. Löfblom, J. Feldwisch, V. Tolmachev, J. Carlsson, S. Ståhl
and F. Y. Frejd, FEBS Lett., 2010, 584, 2670.
22 Á. Roxin and G. Zheng, Future Med. Chem., 2012, 4, 1601.
Int. Ed., 2006, 45, 7581.
44 C. B. Rosen and M. B. Francis, Nat. Chem. Biol., 2017, 13,
697.
This journal is © The Royal Society of Chemistry 2018
Org. Biomol. Chem.