Acid promoted cinnamyl ion mobility within peptide derived macrocycles

Article abstract of DOI:10.1021/ja803299h

We are developing methods that restrict the conformational mobility of peptides and related heteropolymers while simultaneously altering their properties. Our experiments occur as processes wherein a conserved, lipophilic reagent is activated in stages to form composite products with unprotected polyamides in parallel. For each starting oligomer, the goal is to create not one, but rather a collection of products. The intent is for those materials to retain molecular recognition elements of the biopolymer, yet display that functionality as part of stable, cyclic structures having defined shapes and enhanced membrane solubility/permeability. Here we describe reagent 2 and its two-step integration into peptides to afford macrocyclic ethers (e.g., 4 when starting with W-W-Y). When those materials are treated with protic acid in anhydrous solvent, the cinnamyl unit migrates from the oxygen of tyrosine to distribute throughout the structure, forming new products via carbon/carbon bonding. These changes occur concomitantly with acid-promoted rearrangements/cyclizations of the dienyne appendage to generate mixtures containing unique macrocycles such as 15. Similar amalgamations of 2 with more diverse peptides is a means to begin accessing complex peptidomimetics systematically. From a library of screening fractions generated in this way, we have identified a small molecule that selectively promotes hippocampal neurogenesis in the adult mouse brain. Copyright

Full text of DOI:10.1021/ja803299h

Published on Web 09/24/2008
Acid Promoted Cinnamyl Ion Mobility within Peptide Derived Macrocycles
Hongda Zhao,† Lidet Negash,† Qi Wei,† Thomas G. LaCour,† Sandi Jo Estill,† Emanuela Capota,†
Andrew A. Pieper,‡ and Patrick G. Harran*,†
Departments of Biochemistry and Psychiatry, UniVersity of Texas Southwestern Medical Center at Dallas,
5323 Harry Hines BouleVard, Dallas, Texas 75390-9038
Received May 3, 2008; E-mail: harran@chem.ucla.edu
We are developing methods that restrict the conformational
mobility of peptides and related heteropolymers while simulta-
neously altering their properties. Our experiments occur as processes
wherein a conserved, lipophilic reagent (E, Figure 1A) is activated
in stages to form composite products with unprotected polyamides
in parallel. For each starting oligomer, the goal is to create not
one, but rather a collection of products.1 The intent is for those
materials to retain molecular recognition elements of the biopoly-
mer, yet display that functionality as part of stable, cyclic structures
having defined shapes and enhanced membrane solubility/perme-
ability.2
Here we describe compound 2 (Figure 1B) as a specific illus-
tration of E.3 The molecule harbors several electrophilic motifs,
one overtly reactive and the others latent. Aldehyde 2 can be mixed,
independently, with a variety of unprotected peptides to form
N-terminal imines. These react in situ with, for example, aryl-
substituted tosylmethylisocyanides to afford imidazole products of
a net three-component condensation.4 If the original peptide
possesses a phenolic residue, the condensation product can be
treated with catalytic amounts of [allylPdCl]2/Xantphos5 complex
to initiate a macrocycloetherificationswherein the allylic carbonate
precedes a transient metal π-allyl species that captures the pendant
phenol. Synthetic G-W-Y6 is transformed cleanly into macrocycle
4 using these two operations and the methods act similarly on a
range of peptide substrates. Reagent 2 sheds its oxygenation in the
process to become solely unsaturated hydrocarbon in the products.
Figure 1. (A) Generic plan to mold peptides and related nucleophilic
heteropolymers into complex variants having defined shapes and altered
properties [black spheres ) nucleophilic monomers, colored spheres ) latent
and/or overtly reactive electrophiles]. (B) As an example of E, reagent 2
can be incorporated into a range of peptides to afford cyclic composites
having both polar and hydrophobic domains. For instance, mixing 2 with
unprotectedG-W-Yandtreatingtheincipientiminewith4-F-PhCH(NC)SO2tol/
K2CO3 affords an N-terminal imidazole adduct that transforms into cyclic
cinnamyl ether 4 when exposed to 2 mol % [(C3H5)PdCl]2/Xantphos
complex in DMF solution.
Robust ligation and cyclization steps are a start, but the larger
aim seeks more extensive alterations of the peptide brought about
by extended processes (Figure 1A). The design of 27 furthers this
goal. Once intermediates are fashioned into rings, one can initiate
promiscuous long-range rearrangements alongside varying internal
cyclizations. The substance thereby permits one type of complex
molecule (the peptide) to be molded into numerous others.
Congeners 7 and 8 harbor this same macrocycle and contain the
dienyne of 4 intact or its formal Cope rearranged variant,
respectively. Products 9 likely derive from 6 wherein a secondary
cyclization installs a conjugated dihydronapthalene motif. Lastly,
structures 10 result from a third mode of cinnamyl unit migration,
in this case to engage the trisubstituted olefin of the dienyne
appendage to generate regioisomers of the unsaturated meta-
cyclophane shown.
In the case of 4, exposing the molecule to protic acid in an-
hydrous solution initiates a number of striking isomerizations.8
Treatment with excess MeSO3H for 3 h in dry MeNO2 at 0 °C
affords a crude mixture from which we isolate 5-10 via preparative,
mass-guided HPLC (Figure 2).9 Structure assignments derive from
1
mass spectra and combined H, 13C, gCOSY, gHMBC, gHMQC,
and NOESY NMR analyses.
Use of reagent 2 translates simple protonations into varying
products and these track changes in the structure of polyamide
starting material. For example, if instead of G-W-Y, our three-step
process [(1) ligation with 2, (2) cycloetherification,11 (3) acid
treatment] is repeated beginning with synthetic W-W-Y, acid
induced mobility of the cinnamyl unit again transforms a 23-
membered cyclic ether into products having 21, 19, and 15-
membered macrocycles (11-14, Figure 2). However, in this case
we also isolate materials 15 and 16. The ansa bridge in 15 likely
reflects a side-chain propargylic cation, rather than losing a proton
Phenol 5 is a product of noncanonical aryl Claisen rearrange-
ment10 of the cinnamyl ether in 4 concomitant with cyclization of
its dienyne appendage to an alkynyl cyclohexene (3:1 mixture of
olefin regioisomers). Structure 6 derives similarly except wherein
the putative cinnamyl ion pair formed by heterolysis of the allylic
ether in 4 has undergone migration to the adjacent tryptophan
residuesforming a new C-C bond at the indole 5-position.
† Department of Biochemistry.
‡ Department of Psychiatry.
9
13864 J. AM. CHEM. SOC. 2008, 130, 13864–13866
10.1021/ja803299h CCC: $40.75
2008 American Chemical Society

C O M M U N I C A T I O N S
processed in parallel, with product mixtures from each being
fractionated and arrayed for biochemical evaluations.
As a pilot study, we processed GWY, WWY and GGY using
reagent 2 and the three steps described. Product mixtures from each
were divided into 10-12 fractions by preparative HPLC. These
fractions were lyophilized to powders and redissolved in DMSO
to provide ∼10 mM stock solutions. With this small screening set,
we chose an in vivo assay where a functional response could be
tracked reliably. The assay scored for hippocampal neurogenesis
in the adult mouse brain. Data has established a correlation between
postnatal neural stem cell proliferation in the subgranular zone of
the dentate gyrus and forms of human neuropsychiatric disease
characterized by cognitive deficits, such as schizophrenia.13,14 There
is interest in agents that stimulate neuronal precursor cell prolifera-
tion in the hippocampus as potential pharmacotherapy in this area.
Isolated HPLC fractions were diluted (103-fold) in artificial
cerebrospinal fluid and infused intracerebroventricularly at a
constant rate into the left lateral ventricle of adult C57BL/J6 mice
by means of an implanted Alzet osmotic minipump. Mice were
awake and unconstrained during the infusion and administered daily
intraperitoneal injections of bromo deoxyuridine (BrdU, 50 mg/
kg) as a marker of cell divisions. After infusion was complete (7
days), mice were transcardially perfused, and brain tissue was
stained with antibodies against BrdU. Hippocampal neurogenesis
was evaluated by light microscopy contralateral to the side of pump
implantation (Figure 3A) to avoid artifacts from tissue damage.
Every fifth section through the rostral-caudal extent of the hippo-
campus was analyzed, and the total number of BrdU+ cells was
normalized against the volume of the dentate gyrus. Similarly
infused recombinant fibroblast growth factor (FGF, 0.04 mg/ Kg/
day) served as a positive control (Figure 3B).
Select pools derived from each of the processed peptides
displayed activity in this format. We pursued fraction five from
G-G-Y because it elicited the most robust and selective15 response
(Figure 3C). This fraction contained several products (Figure 3D),
and these were separated into an additional three pools and
rescreened in vivo as described. The pool found active now
contained four main products (Figure 3E), and these were chro-
matographed a third time to afford four components, one of which
(Figure 3F) retained the majority of neurogenesis activity. The
structure of the active constituent was subsequently assigned as
meta cyclophane 17 on the basis of mass spectra and multidimen-
sional NMR analyses. Pure 17 reproducibly elicits marked (Figure
3G) and selective neural cell proliferation in the hippocampal
dentate gyrus and its activity is dose dependent in a dilution series
(Figure 3H).
Figure 2. Products derived from processing G-W-Y and W-W-Y indepen-
dently with reagent 2, using the chemistry depicted in Figure 1B followed by
treatment with MsOH in anhydrous MeNO2. Ar ) p-F-phenyl. ∆38,39 (R )
Me) and ∆39,41 (R ) CH2) isomers of cyclophane 10 characterized separately.
14 observed in ∆38,39 form only. Where present, the alkynyl cyclohexene
motif (a) is isolated as a mixture of ∆34,35 and ∆34,36 olefin isomers.
to afford 11, engaging the proximal tryptophan residue in a
transannular electrophilic aromatic substitution. Structure 16 com-
pletes a set. In 11 the cinnamyl unit has shifted from O to C at
tyrosine, in 12/13 it has migrated to the adjacent tryptophan residue,
and in 16 it has traversed the length of the tripeptide to engage the
distal tryptophan, in this instance bonding to the indole C-2 position
via a branched linkage.12
The mechanism by which 17 acts is not yet known. While its
properties and performance are being examined further, results here
show that, even from a very small array, we can track and purify
an active constituent from mixtures. Iterative rescreening of
increasingly homogeneous product fractions minimizes the prob-
ability of isolating pure false positives. Time invested in structure
determinations is then of value.
Structures 5-16 harbor six new macrocycle types. They were
prepared from just two small peptides using one reagent (2) and
three steps. The materials generally show well-dispersed 1H NMR
spectra and, relative to the peptides from which they derive,
improved solubility in organic solvents and detergent containing
buffer solutions. Precise de novo syntheses of similar molecules
would be considerably more involved; and therein lies the potential.
We have characterized the current examples in detail for demon-
stration. However, this is time-consuming. In fact, each isolate
requires analyses comparable to those used in a natural product
structure determination. Going forward, up-front analyses are not
necessary. Rather, oligomers will be assembled systematically and
As chemistry in this venue is refined, processes will be extended
to include oxidation/oxygenations and screening sets will expand
considerably. New methods will be established to amalgamate next
generation reagents into diverse synthetic polyamides and variants.
Unmodified, these polymers generally have poor properties for
pharmacological research. Our goal is to purge those limiting
characteristics while retaining the value of functionally complex,
stereochemically rich raw materials that can be made systematically
by machine. Doing so would provide broad potential to discover
9
J. AM. CHEM. SOC. VOL. 130, NO. 42, 2008 13865

C O M M U N I C A T I O N S
Figure 3. (A) Normally proliferating cells (anti BrdU, black dots) in the subgranular zone of the mouse hippocampal dentate gyrus (DG, lower left). (B)
Magnified DG from a mouse infused with FGF (0.04 mg/Kg/d for 7 d) showing increased BrdU incorporation [see Supporting Information for detailed
protocols]. (C) In vivo screening of products derived from processing G-G-Y with 2 identified a chromatography fraction (#5) that elicits BrdU incorporation
comparably to FGF. HPLC analysis of this fraction (D) showed it contained several products. These were divided into three pools by HPLC and rescreened
in vivo. The active fraction (E) was resolved into components, one of which (F) retained the majority of neurogenesis activity and was assigned structure
17 (∼1:1 mixture of isomers, stereochemistry at C9 and C38 unassigned). Isolated 17 elicits marked BrdU incorporation in the DG when infused at 0.1
mg/kg/d for 7d (G) and its activity is dose dependent (H).
Wildeman, J.; Oldenziel, O. H. J. Org. Chem. 1977, 42, 1153–1159.
(5) Klingensmith, L. M.; Strieter, E. R.; Barder, T. E.; Buchwald, S. L.
Organometallics 2006, 25, 82–91.
(6) Peptides assembled using Fmoc-based solid phase synthesis and isolated
as C-terminal n-butylamides. Residue abbreviations: G ) glycine, Y )
tyrosine, W ) tryptophan.
(7) Aldehyde 2 is prepared on gram scales beginning with commercial
isophthalaldehyde and obtained in 60% ee (unoptimized) upon asymmetric
conjugate reduction of the corresponding enal. See Supporting Information.
(8) Structure and unimolecular reactions of the 1-phenylallyl cation: Olah, G. A.;
Asensio, G.; Mayr, H. J. Org. Chem. 1978, 43, 1518–1520.
(9) The combined mass recovery of isomeric products approaches 80% (goal
efficiency) in these experiments. Yield of individual products range from
∼3% to 20% (see Supporting Information) and can vary with acid exposure
time.
(10) (a) Nonoshita, K.; Banno, H.; Maruoka, K.; Yamamoto, H. J. Am. Chem.
Soc. 1990, 112, 316–322. (b) Wipf, P.; Rodr´ıguez, S. AdV. Synth. Catal.
2002, 344, 434–440.
(11) This affords a macrocyclic ether directly analogous to 4 in 79% yield.
(12) In the W-W-Y series, we have not isolated products derived from internal
tryptophan/tryptophan cross-linking nor did DDQ oxidation of the crude
acidolysis mixture generate isolable materials that fluoresce upon excitation
at 254 nm. See: Stachel, S. J.; Habeeb, R. L.; van Vranken, D. L. J. Am.
Chem. Soc. 1996, 118, 1225–1226.
complex peptidomimetics useful as probes in areas of biology where
recombinant biologics and/or small drug-like heterocycles can be
limited.16
Acknowledgment. Financial support provided by the Robert
A. Welch Foundation, the Chilton Family Endowment, the Hartwell
Foundation, the Morton H. Meyerson Family Tzedakah Fund, and
Merck Laboratories is gratefully acknowledged. We thank Dr.
Carlos Amezcua and Professor John MacMillan (UTSW) for
invaluable advice and Steven L. McKnight for tireless support of
our programs.
Supporting Information Available: Experimental procedures and
characterization data for new compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) For a dissimilar design but related concept see: Ostresh, J. M.; Husar, G. M.;
Blondelle, S. E.; Dorner, B.; Weber, P. A.; Houghten, R. A. Proc. Natl.
Acad. Sci. U.S.A. 1994, 91, 11138–11142. Confer also: (b) Spiegel, D. A.;
Schroeder, F. C.; Duvall, J. R.; Schreiber, S. L. J. Am. Chem. Soc. 2006,
128, 14766–14767.
(2) Jackson, S.; DeGrado, W.; Dwivedi, A.; Parthasarathy, A.; Higley, A.;
Krywko, J.; Rockwell, A.; Markwalder, J.; Wells, G.; Wexler, R.; Mousa,
S.; Harlow, R. J. Am. Chem. Soc. 1994, 116, 3220–3230.
(13) Pieper, A. A.; Wu, X.; Han, T.; Estill, S. J.; Dang, Q.; Wu, L. C.; Reece-
Fincanon, S.; Dudley, C. A.; Richardson, J. A.; Brat, D. J.; McKnight,
S. L. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 14052–14057.
(14) Reif, A.; Fritzen, S.; Finger, M.; Strobel, A.; Lauer, M.; Schmitt, A.; Lesch,
K. P. Mol. Psych. 2006, 11, 514–522.
(15) Fractions 4 and 8-10 (omitted in Figure 3C) caused BrdU staining in areas
outside the dentate gyrus, indicative of inflammation. They were discarded.
(16) Li, L.; Thomas, R. M.; Suzuki, H.; De Brabander, J. K.; Wang, X.; Harran,
P. G. Science 2004, 305, 1471–1474.
(3) Wei, Q.; Harran, S.; Harran, P. G. Tetrahedron 2003, 59, 8947–8954.
(4) (a) Sisko, J.; Kassick, A. J.; Mellinger, M.; Filan, J. J.; Allen, A.; Olsen,
M. A. J. Org. Chem. 2000, 65, 1516–1524. (b) van Leusen, A. M.;
JA803299H
9
13866 J. AM. CHEM. SOC. VOL. 130, NO. 42, 2008