Proteomorphous objects from abiotic backbones

Article abstract of DOI:10.1002/anie.200603390

(Figure Presented) Big is beautiful: A folded synthetic molecule with a conformation that compares in size with the tertiary folds of a small protein and yet only consist of non-natural units is described. By not controlling the helical handedness allows

Full text of DOI:10.1002/anie.200603390

Communications
DOI: 10.1002/anie.200603390
Helical Structures
Proteomorphous Objects from Abiotic Backbones**
Nicolas Delsuc, Jean-Michel LØger, StØphane Massip, and Ivan Huc*
The functions of proteins rest on the ability of peptidic
sequences to fold into well-defined conformations. Previously,
molecular folding was believed to belong mainly to biological
macromolecules but, over the last decade, chemists have
shown that numerous families of synthetic oligomers (fol-
damers) also adopt folded structures,[1–4] some of which
display biological activities.[3–6] Foldamers have opened the
path to mimic not only protein structure but also protein
function in artificial systems. However, they have thus far
been limited in size and mostly consist of simple analogues of
secondary structural elements as, for example, isolated helical
or linear strands. Herein, we present a large foldamer
(> 8 kDa) that compares to a modest-sized protein tertiary
fold in terms of its dimension and structural complexity. Our
results show that design and stepwise synthetic strategies can
be devised to access abiotic, very large, yet conformationally
defined architectures that resemble proteins somewhat more
closely than previously described sizeable non-peptidic struc-
tures, such as discrete self-assemblies[7,8] or dendrimers.[9,10]
These proteomorphous objects represent novel platforms to
envisage molecular recognition and enzymelike catalysis on a
Scheme 1. Structures and helical folding of oligoamide sequences.
Compound 1 possesses a branched architecture made of two helices
each comprising 17 units (8+1+8) separated by an ethylene glycol
spacer. Oligomer 2 consists of a single helix of nine units (4+1+4).
The formula of helical tetrameric and octameric precursors[12] 3 and 4,
respectively, show hydrogen bonds (dashed lines) and electrostatic
repulsions (double headed arrows) that, along with intramolecular
aromatic stacking, stabilize the helically folded conformers. The helical
conformations have been characterized both in the solid state and in
solution.[11–15] R=isobutoxy residues ((CH3)2CHCH2O-) that protrude
from the helices toward the solvent.
different scale.
In designing a large artificial folded structure, one must
take into account the feasibility of the synthesis. Considering
the difficulty of the stepwise preparation of long peptides in
significant quantities, we determined that linear sequences of
monomers may not be the best target for practical purposes.
Alternatively, large branched architectures have been pre-
pared by using efficient divergent or convergent synthetic
schemes.[9,10] Branched structure 1 (8.2 kDa, C464H464N70O74)
was thus chosen as our first design (Scheme 1). It consists of
two helices connected by a short ethylene glycol spacer that
imposes proximity, and thus interactions, between their
peripheral residues, as in protein tertiary motifs. Each helix
comprises two identical octameric quinoline oligoamides
linked to a bis(aminomethyl)pyridine connector. Quinoline
oligoamides derived from 8-amino-2-quinolinecarboxylic acid
adopt particularly robust helical conformations in a wide
range of nonprotic (CDCl3, DMSO)[11–14] and protic (MeOH,
H2O)[15] solvents and thus represent versatile “building
blocks” to construct proteinlike objects in a modular fashion.
The multistep synthesis of 1 includes two key reactions
(Scheme 2): the divergent attachment of two pyridine units to
the central ethylene glycol spacer to give intermediate 8 and
the subsequent simultaneous ligation of four quinoline
octameric amides to give 1 (see the Supporting Information).
The synthesis has been completed on a 50-mg scale but could
easily be scaled up to produce larger amounts.
As detailed below, an essential feature of the design of 1,
which greatly facilitated the study of its structure, is that it
contains no chiral group to favor a right (P) or left (M)
handedness of the helices.[14] The proposal that 1 consists of
two distinct helices relies on the premise that each bis-
(aminomethyl)pyridine connector that links a pair of quino-
line oligomers inserts itself in the quinoline helical motif. This
was validated by a conformational study of the shorter
[*] N. Delsuc, Dr. I. Huc
Institut EuropØen de Chimie et Biologie
2 rue Robert Escarpit, 33607 Pessac (France)
Fax: (+33)540-002-215
E-mail: i.huc@iecb.u-bordeaux.fr
Prof. J.-M. LØger, Dr. S. Massip
Laboratoire de Pharmacochimie
146 rue LØo Saignat, 33076 Bordeaux (France)
[**] This work was supported by an ANR grant (project no. NT05-
3_44880), by the CNRS, the University of Bordeaux I and the
University Victor Segalen Bordeaux II, the Ministry of research
(predoctoral fellowship to N.D.), and by the Conseil RØgional
d’Aquitaine.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
214
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 214 –217

Angewandte
Chemie
relative quinoline helical segments so that they have the same
handedness. In contrast, an analogue of 2, in which the
pyridine ring is replaced by a phenyl group, shows little
preference for the meso or the racemic species, and undergoes
faster helix inversion than 2: the temperature of coalescence
1
of the H NMR signals of diasterotopic motifs of isobutoxy
side chains protons in CDCl3 is above 558C for 2 (see the
Supporting Information) compared with 258C for its C
analogue.[16]
The same reasoning applies to the structure of 1: each of
its two large helices may have P or M helicity and thus 1 may
exist as a mixture of P-P/M-M and P-M species. Importantly,
the two helices of 1 are connected side by side and not end to
end anymore as in the two helical segments of 2. Thus, any
preference for the racemic or meso species in 1 should reveal
interactions between the side chains of the helices, which
amount to tertiary interactions in proteins. Taking into
account the size of 1, its NMR spectrum in CDCl3 is strikingly
simple (Figure 1b). The symmetrical branched architecture is
conserved in the solution conformation, giving rise to a
degenerate spectroscopic signature. Essentially, one set of
signals can be observed in the spectrum, which implies a
strong preference for either the meso or the racemic
conformer and thus significant helix–helix interactions.
Small signals amounting to about 7% can also be observed
and were eventually assigned to the other conformer (see
below).
The unambiguous assignment of the preferred conforma-
tion of 1 in CDCl3 to the meso P-M species was made possible
by a crystal structure (see the Supporting Information) that
showed that the two helices have opposite handedness
(Figure 2b–d) and by a NMR spectrum of the same freshly
dissolved crystals, which showed that the crystals correspond
to the major species in CDCl3. The structure validates the
design strategy and provides a clear illustration of the size of
1, which compares with that of a 75-residue protein. This
structure is significant in that it represents a rare case of
genuine tertiary, albeit primarily steric, interactions in an
abiotic system. The very fact that 1 easily crystallizes (over-
night) suggests a well-defined three-dimensional conforma-
tion. By comparison, only very small dendrimers can crystal-
lize.[18] The structure reveals a perpendicular orientation of
the two helices. This conformation is apparently imposed by
steric repulsions between the isobutoxy groups in position 4
of each helical segment (Figure 2c), which would otherwise
clash if the helices were oriented parallel. Presumably,
residues in position 6 and 7 of each octameric sequence may
also contribute to contacts between the two helices and
participate in the stabilization of the P-M species at the
expense of the P-P/M-M form.
Scheme 2. Synthesis of folded oligomeric sequences 1 and 2. a) BH3,
THF, quant.; b) BocOBoc, THF, 75%; c) TFA, CH2Cl2, quant.; d) 3, HBTU,
HOBT, EtiPr2N, DMF, 65%; e) H2, Pd black, MeOH/EtOAc, 73%; f) HO-
(CH2)2OH, DIAD, PPh3, THF, 89%; g) TFA, CH2Cl2, quant.; h) 4, HBTU,
HOBT, EtiPr2N, DMF, 44%. Boc=tert-butoxycarbonyl, TFA=trifluoroacetic
acid, HBTU=O-(benzotriazol-l-yl)-N,N,N’,N’-tetramethyluronium hexa-
fluorophosphate, HOBT=hydroxybenzotriazole, DMF=N,N-dimethyl-
formamide, DIAD= diisopropylazodicarboxylate, Bn=benzyl.
oligomer 2 (Scheme 1). The two quinoline helical segments of
2 may, in principle, have P or M helicity. Thus, 2 may exist as a
mixture of a racemic P-P/M-M and a meso P-M species that
interconvert through the handedness inversion of one helical
segment. This process is slow on the NMR time scale and
should give rise to two sets of signals on the 1H NMR
spectrum of 2 if the two species coexist.[15,16] However, the
spectrum of 2 shows that it consists of a single set of signals
(Figure 1a). A crystal structure of 2 confirmed that this
species corresponds to the racemic P-P/M-M mixture for
which the two helical segments of each molecule always
possess the same handedness (Figure 2a). This preference
presumably arises from favorable interactions between the
endocyclic nitrogen of the pyridine unit and the two
neighboring amide protons that set the orientation of the
1
Figure 1. Selected region of the 400 Mhz H NMR spectra of 1 and 2
The characterization of the meso helix underlines the
usefulness of not controlling handedness. If identical chiral
residues had been introduced in the helices, they would have
favored one of the two homochiral species (P-P or M-M),[14]
therefore playing against the natural tendency of the system.
Induction of helix handedness by a neighboring helix has
rarely been observed.[19,20] As shown herein, it represents a
novel, reliable, and quantitative probe of helix–helix inter-
actions. In contrast, previous studies of synthetic peptidic
at 258C showing amide resonances. a) Compound 2 at equilibrium in
CDCl3. b) Branched structure 1 at equilibrium in CDCl3. c) Compound
1 freshly dissolved in CDCl3 after one week of equilibration in toluene.
d) Compound 1 at equilibrium in [D8]toluene. The stars and circles
indicate signals assigned individually to the P-M and P-P/M-M con-
formers of 1, respectively. Other signals of the two species overlap.
Full spectra and experiments that led to the assignment of the signals
to P-P/M-M and P-M conformers based on the monitoring of
chemical-shift variations in solvent mixtures are reported in the
Supporting Information.
Angew. Chem. Int. Ed. 2007, 46, 214 –217
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
215

Communications
Figure 2. Crystal structures of helix 2 and of the branched structure 1. a) Side view and topview of the right-handed helix of 2 in the solid state
(quinoline segments in grey and pyridine unit in red). Alkoxy residues have been omitted for clarity. b) View of the structure of 1 as van der Waals
spheres (carbon in gray, hydrogen in white, oxygen in red, nitrogen in blue). c) Same view of 1, however, the structure is represented as tubes
except for the residues that are responsible for intramolecular helix–helix interactions that are shown as van der Waals spheres (ethylene glycol
spacer in gray, isobutoxy chains in position 4, 6, and 7 of each four subhelix in blue, green, and red, respectively; numbering begins at each of the
four N termini). Residues in position 4 clearly lie at the helix–helix interface. d) Two views of the folded structure of 1 in which residues have been
omitted to show the perpendicular arrangement of the right-handed helix (gray) and of the left-handed helix (red). e) For the purpose of size
comparison, the crystal structure of the scorpion toxin protein is shown at the same scale as all other structures in this figure.[17] The scorpion
toxin protein contains 65 amino acids (7.1 kDa, C311H423N86O95S8) and is 15% smaller than 1.
helix bundles showed that handedness is determined by chiral
centers on each amino acid and interactions between helices
are probed by the onset of helicity.[21]
of a dichloroethane molecule in a cavity between the two
helices in the crystal structure of 1 suggests that solvent is
intimately associated with helix–helix interactions (see the
Supporting Information).
This study demonstrates that is possible to design,
synthesize, and characterize folded molecules that, by their
size (> 8 kDa) and structural complexity, mimic the tertiary
fold of a small protein yet only consist of non-natural units.
Additionally, it shows that purposely not controlling helical
handedness within secondary motifs allows the observation of
tertiary interactions between helical modules through induc-
tion of helix–helix side-by-side handedness.
Although the abiotic proteomorphous foldamer pre-
sented herein is the first of its kind, it has a large potential
for development. Its structure is amenable to specific changes
to promote attractive interactions between the helices. For
example, tuning the spacer length and level of branching, the
size of each helical module, and the nature of the side chains
may allow us to define binding sites and position catalytic
Furthermore, we observed that helix–helix interactions
and the preference for the meso species of 1 much depends on
the solvent. A complete shift of equilibrium occurs in
aromatic solvents in which the racemic P-P/M-M form
dominates (70% compared with 7% in CDCl3) at the
expense of the meso P-M form (30% compared with 93%
in CDCl3). Upon dissolving crystals grown in chlorinated
solvents into deuterated toluene or benzene, or crystals grown
in an aromatic solvent into CDCl3, the slow interconversion of
each species into the other can be monitored and be fitted to a
single exponential decay curve. Interestingly, helix inversion
is faster in toluene (characteristic time of 3.5 h) than in
chloroform (characteristic time of 10 h) at 258C. The exact
mechanism by which the solvent influences the proportions
between the two species and the rates at which they
interconvert is not well established. However, the presence
216
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 214 –217

Angewandte
Chemie
[9] G. R. Newkome, C. N. Moorefield, F. Vꢀgtle, Dendrimers and
groups within the folded structure or onto its exterior, which
may be decorated so as to interact over large surfaces with
specific targets.
Dendrons: Concepts, Syntheses, Applications, Wiley-VCH,
Weinheim, 2001.
[10] Dendrimers and other Dendritic Polymers (Eds: J. M. J. Frꢁchet,
D. A. Tomalia), Wiley, Chichester, 2001.
[11] H. Jiang, J.-M. Lꢁger, I. Huc, J. Am. Chem. Soc. 2003, 125, 3448.
[12] H. Jiang, J.-M. Lꢁger, C. Dolain, P. Guionneau, I. Huc,
Tetrahedron 2003, 59, 8365.
[13] C. Dolain, A. Grꢁlard, M. Laguerre, H. Jiang, V. Maurizot, I.
Huc, Chem. Eur. J. 2005, 11, 6135.
[14] C. Dolain, H. Jiang, J.-M. Lꢁger, P. Guionneau, I. Huc, J. Am.
Chem. Soc. 2005, 127, 12943.
[15] E. Gillies, C. Dolain, J.-M. Lꢁger, I. Huc, I. , J. Org. Chem. 2006,
71, 7931.
[16] C. Dolain, J.-M. Lꢁger, N. Delsuc, H. Gornitzka, I. Huc, Proc.
Natl. Acad. Sci. USA 2005, 102, 16146.
[17] G. D. Smith, R. H. Blessing, S. E. Ealick, J. C. Fontecilla-Camps,
H. Hauptman, D. H. Housset, D. A. Langs, R. Miller, Acta
Crystallogr. Sect. D 1997, 53, 551.
[18] B. Huang, J. R. Parquette, Org. Lett. 2000, 2, 239.
[19] E. J. Spek, Y. Gong, N. R. Kallenbach, J. Am. Chem. Soc. 1995,
117, 10773.
[20] K. Maeda, M. Ishikawa, E. J. Yashima, J. Am. Chem. Soc. 2004,
126, 15161.
[21] M. Mutter, G. G. Tuchscherer, C. Miller, K.-H. Altmann, R. I.
Carey, D. F. Wyss, A. M. Labhardt, J. E. Rivier, J. Am. Chem.
Soc. 1992, 114, 1463; M. R. Ghadiri, C. Soares, C. Choi, J. Am.
Chem. Soc. 1992, 114, 825; K. Suzuki, H. Hiroaki, D. Kohda, H.
Nakamura, T. Tanaka, J. Am. Chem. Soc. 1998, 120, 13008.
Received: August 18, 2006
Published online: October 19, 2006
Keywords: chirality · helical structures · interactions ·
.
supramolecular chemistry · X-ray diffraction
[1] D. J. Hill, M. J. Mio, R. B. Prince, T. S. Hughes, J. S. Moore,
Chem. Rev. 2001, 101, 3893.
[2] I. Huc, Eur. J. Org. Chem. 2004, 17.
[3] R. P. Cheng, S. H. Gellman, W. F. DeGrado, Chem. Rev. 2001,
101, 3219.
[4] D. Seebach, A. K. Beck, D. J. Bierbaum, Chem. Biodiversity
2004, 1, 1111.
[5] E. A. Porter, X. Wang, H.-S. Lee, B. Weisblum, S. H. Gellman,
Nature 2000, 404, 565.
[6] “Versatile oligo(N-subtitued) glycines: the many roles of
peptoids in drug discovery”: J. A. Patch, K. Kirshenbaum, S. L.
Seurynck, R. N. Zuckermann, A. E. Barron in Pseudo-Peptides
in Drug Discovery (Ed.: P. E. Nielsen), Wiley-VCH, Weinheim,
2004, p. 1.
[7] M. Tominaga, K. Suzuki, M. Kawano, T. Kusukawa, T. Ozeki, S.
Sakamoto, K. Yamaguchi, M. Fujita, Angew. Chem. 2004, 116,
5739; Angew. Chem. Int. Ed. 2004, 43, 5621.
[8] L. R. MacGillivray, J. L. Atwood, Nature 1997, 389, 469.
Angew. Chem. Int. Ed. 2007, 46, 214 –217
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
217