Modulation of the peptide backbone conformation by the selenoxo photoswitch

Article abstract of DOI:10.1021/ja1019386

Photocontrol of the backbone conformation is a useful step forward in regulating the bioactivities of peptides and proteins by means of external signals. In the present work, the selenium analogue of a peptide bond was introduced into tetrapeptides to obt

Full text of DOI:10.1021/ja1019386

Published on Web 05/18/2010
Modulation of the Peptide Backbone Conformation by the Selenoxo
Photoswitch
Yun Huang, Gu¨nther Jahreis, Christian Lu¨cke, Dirk Wildemann, and Gunter Fischer*
Max-Planck Research Unit for Enzymology of Protein Folding, Weinbergweg 22, D-06120 Halle/Saale, Germany
Received March 9, 2010; E-mail: fischer@enzyme-halle.mpg.de
The conformation of the polypeptide backbone plays an important
role in determining protein structure, stability, and dynamics.1
External control of conformational changes in a backbone as defined
by photoswitching at the one-bond level would provide a sound
basis for structure-function studies of proteins. Therefore, incor-
poration of a photoresponsive element into the backbone, to control
structure and bioactivity by a soft light switch, is of great interest.
Herein, for the purpose of obtaining more stable peptide bond
conformers characterized by a bathochromic shift of the π-π* band
relative to the oxo or thioxo peptide UV/vis absorption, we
introduced a single selenoxo peptide bond into the peptide
backbone. Principally, the UV/vis spectral properties of amides
allow us to trigger the cis/trans peptide bond photoswitch.2
However, the repetitive character of the peptide bond in proteins
and the low excitation wavelength of ∼200 nm required for
producing excess cis isomer in the photostationary state (PSS) would
considerably hamper the interpretation of photoisomerization
experiments in oligopeptides and proteins. Azobenzene as a
backbone constituent, for instance, was shown to trigger the folding
and unfolding of R-helix and ꢀ-hairpin model peptides on the
femtosecond time scale.3 It should be noted, however, that the
azobenzene moiety interrupts the regularly repeating pattern of a
polypeptide chain, underlining the need for a chemically less
perturbing backbone substitution.
thesis protocols has never been investigated, and selenoxo peptides
have not been described previously. Based on the strategy shown
in Scheme 1, we synthesized the N-protected selenoxo tetrapeptide
methyl esters in acceptable yields, indicating that the selenoxo
peptide moiety is stable under conditions of modified N-Boc
chemistry and mixed anhydride coupling (Table S1). Peptide 1
showed 7% contamination with the CR epimer of the Se-containing
residue, which was probably caused by the selenation procedure
using Woollins’ reagent but did not impede measurements presented
here (Figure S1). Surprisingly, selenoxo peptide 1 proved to be
stable at pH 4.8, 6.5, and 7.6 over a 4-day storage period. Only
5% selenium-oxygen exchange was observed after a 4-day period
of incubation at pH 11, accompanied by partial hydrolysis of the
C-terminal methyl ester group.
Scheme 1. Synthesis of Selenoxo Peptides
Previous studies in our group have shown that irradiation of a
thioxo peptide bond sC(dS)sNRs (R ) H, alkyl) incorporated
in peptides and proteins implements a one-bond cis/trans photo-
switch, but the most effective wavelength of 254 nm is potentially
destructive for sensitive protein constituents.4 In addition, lifetime
limitation in the thermal isomerization (t1/2 ≈ 102 s at 10 °C for
the sC(dS)sNHs moiety) for the photoproduced cis isomer is
less suitable for biochemical investigations. Notably, the process
of photoisomerization itself is completed within a few hundred
picoseconds at high quantum yield and does not exhibit a prolonged
reaction time for longer peptide chains.5 A surprising result of the
first photoswitch application of a thioxo peptide bond in a protein
was that the conformation of the remote Ala4-ψ[CS-NH]-Ala5
moiety, which had previously not been considered important for
catalysis, was implicated in the control of the catalytic machinery
of RNase S. A similar approach gave clear evidence for the isomer
specificity in the functional interaction of the ψ[CS-N]-Pro3 kinin
with the insect kinin receptor.6
The series of synthesized oligopeptides containing the seleno
modification, namely Bz-Val-Gly-Ala-ψ[CSe-NH]Ala-OMe (1),
Bz-Val-Gly-Ala-ψ[CSe-NH]Phe-OMe (2), and Bz-Val-Gly-Ala-
ψ[CSe-N]Pro-OMe (3), comprised secondary selenoxo amides as
well as an imidic selenoxo amide.7 The thioxo tetrapeptide
congeners were incorporated in this work for the sake of comparison.
Selenoxo amides are well-known intermediates in heterocycle
synthesis8 and were found to be unstable in solution.9 However,
the compatibility between selenoxo substitution and peptide syn-
UV/vis spectroscopy of peptide 1 in phosphate buffer, pH 6.5
(Figure 1) showed a strong absorption band centered at 294 nm
(π-π* transition, ꢀ ) 11280 M-1 cm-1) and a weak band centered
at 366 nm (n-π* transition, ε ) 177 M-1 cm-1), which is
approximately 30 nm red-shifted for each band compared to
1
corresponding thioxo peptides. H NMR spectra of the secondary
selenoxo peptides indicated a cis isomer content of <1% in the
ground state (Figure S2). Excitation of peptide 1 at 286 nm served
to achieve the PSS characterized by a cis content of 11.2%. The
increased cis/trans ratio was accompanied by a small bathochromic
shift of the π-π* transition that completely disappeared during
thermal re-equilibration. Full recovery of the original UV/vis spectra
even after four cycles of irradiation/equilibration indicated no
tendency of the selenoxo peptides to experience photodecomposition
during irradiation. The reconstructed UV spectrum of the pure cis
isomer of peptide 1 showed an absorption maximum at 310 nm (ꢀ
) 12 336 M-1 cm-1), indicating an isomer-specific shift of 16 nm
(Figure S3). This is similar to the magnitude observed for a thioxo
peptide (trans: λmax ) 265 nm, ε265 ) 12 000 M-1 cm-1; cis: λmax
) 281 nm, ε281 ) 15 215 M-1 cm-1).6a Peptide 2 shows a
spectroscopic behavior similar to that of peptide 1, with a first-
order rate constant of (4.4 ( 0.1) × 10-4 s-1 for thermal cis/trans
isomerization. The resulting half time of ∼1600 s showed that
selenoxo substitutions will make a suitable adjustment that satisfies
the conditions for the evaluation of isomer specificities in biochemi-
cal reactions.
9
7578 J. AM. CHEM. SOC. 2010, 132, 7578–7579
10.1021/ja1019386  2010 American Chemical Society

C O M M U N I C A T I O N S
such spectral change was observed for peptide 3 (Figure S8).
Presumably, the sC(dSe)sNHs form is converted to its selenolate
tautomer sC(sSe-)dNs, while peptide 3 is unable to undergo
this reaction. The pH dependent sC(dSe)sNHs absorbances
revealed pKa values of 9.5 and 9.8 for peptides 1 and 2, respectively.
In conclusion, we report the first synthesis of selenoxo peptides.
Our biophysical studies demonstrate that they feature a surprisingly
high stability in aqueous buffer, both in the dark and under the
conditions of photoswitching by UV irradiation. The large Se-
mediated bathochromic shifts of the π-π* transition permitted us
to trigger a .10-fold increase of the cis peptide bond isomer
population in the PSS using irradiation with soft UV light. The
concentration transients of the isomers in the dark after switching
off the light are stable enough to allow for isomer specific
interactions at the one-bond level to be detected. Therefore, the
selenoxo peptide bond may serve as a unique probe to study isomer-
specific contributions in biochemical reactions such as protein
folding, receptor-ligand interactions, and ion channel gating.
Figure 1. UV/vis characterization of photoisomerization for peptide 1 (3.4
× 10-5 mol L-1 peptide in 3.3 × 10-2 mol L-1 phosphate buffer, pH 6.5,
10 °C). Equilibrated peptide (solid line), peptide after 5 min of irradiation
at 286 nm (dashed line), and peptide after four cycles of irradiation/
equilibration (dotted line). Inset: n-π* transition.
Acknowledgment. We are grateful for financial support from
the BMBF (RP ProNet3). We thank Dr. A. Schierhorn and C.
Gersching for MS measurements and M. Seidel for technical
assistance with NMR data collection.
Theoretical calculations have indicated that both the electron
delocalization from N to chalcogen and the planarity of the
chalcogen amide bonds increase in the order O < S < Se.11 This
translates into the heights of the rotational barriers, with a value of
20.5 kcal mol-1 for the selenoxo peptide 1 followed in decreasing
order by the corresponding thioxo (19.4 kcal mol-1) and oxo
peptides (17.1 kcal mol-1, Table S2). The decelerating influence
of the selenium substitution on the cis/trans isomerization is driven
entirely by a large unfavorable activation entropy term (Table S2).
Supporting Information Available: Full experimental details,
including peptide synthesis and analysis, UV spectroscopic character-
izations, and NMR data. This material is available free of charge via
the Internet at http://pubs.acs.org.
References
(1) (a) Baldwin, R. L. Biophys. Chem. 2002, 101, 203–210. (b) Rose, G. D.;
Fleming, P. J.; Banavar, J. R.; Maritan, A. Proc. Natl. Acad. Sci. U.S.A.
2006, 103, 16623–15633. (c) Yang, X. Y.; Wang, M.; Fitzgerald, M. C.
Bioorg. Chem. 2004, 32, 438–449.
Table 1. Characterization of the Peptide Bond Cis/Trans
Isomerization of Thioxo and Selenoxo Peptides
(2) (a) Li, Y.; Garrell, R. L.; Houk, K. N. J. Am. Chem. Soc. 1991, 113, 5895–
5896. (b) Holtz, J. S. W.; Li, P. S.; Asher, S. A. J. Am. Chem. Soc. 1999,
121, 3762–3766. (c) Wang, Y.; Purello, R.; Georgiou, S.; Spiro, T. G. J. Am.
Chem. Soc. 1991, 113, 6368–6377.
(3) (a) Renner, C.; Kusebauch, U.; Loweneck, M.; Milbradt, A. G.; Moroder,
L. J. Pept. Res. 2005, 65, 4–14. (b) Bredenbeck, J.; Helbing, J.; Kumita,
J. R.; Woolley, G. A.; Hamm, P. Proc. Natl. Acad. Sci. U.S.A. 2005, 102,
2379–2384. (c) Schrader, T. E.; Schreier, W. J.; Cordes, T.; Koller, F. O.;
Babitzki, G.; Denschlag, R.; Renner, C.; Loweneckt, M.; Dong, S. L.;
Moroder, L.; Tavan, P.; Zinth, W. Proc. Natl. Acad. Sci. U.S.A. 2007, 104,
15729–15734. (d) Jurt, S.; Aemissegger, A.; Gu¨ntert, P.; Zerbe, O.; Hilvert,
D. Angew. Chem., Int. Ed. 2005, 45, 6297–6300.
(4) (a) Frank, R.; Jakob, M.; Thunecke, F.; Fischer, G.; Schutkowski, M. Angew.
Chem., Int. Ed. 2000, 39, 1120–1122. (b) Zhao, J.; Wildemann, D.; Jakob,
M.; Vargas, C.; Schiene-Fischer, C. Chem. Commun. 2003, 2810–2811.
Zhao, J. Z.; Micheau, J. C.; Vargas, C.; Schiene-Fischer, C. Chem.sEur.
J. 2004, 10, 6093–6101.
(5) (a) Satzger, H.; Root, C.; Gilch, P.; Zinth, W.; Wildemann, D.; Fischer, G.
J. Phys. Chem. B 2005, 109, 4770–4475. (b) Helbing, J.; Bregy, H.;
Bredenbeck, J.; Pfister, R.; Hamm, P.; Huber, R.; Wachtveitl, J.; De Vico,
L.; Olivucci, M. J. Am. Chem. Soc. 2004, 126, 8823–8834.
(6) (a) Wildemann, D.; Schiene-Fischer, C.; Aumu¨ller, T.; Bachmann, A.;
Kiefhaber, T.; Lu¨cke, C.; Fischer, G. J. Am. Chem. Soc. 2007, 129, 4910–
4918. (b) Huang, Y.; Cong, Z. Y.; Yang, L. F.; Dong, S. L. J. Pept. Sci.
2008, 14, 1062–1068.
a
Peptides
kc/t (s-1
)
% cisb
pKac
d
Bz-VGAψ[CS-NH]A-OMe
Bz-VGAψ[CS-NH]F-OMe
(6.7 ( 0.7) × 10-3
-
11-13e
11-13e
9.5
9.8
-
(2.6 ( 0.2) × 10-3
13.5
11.2
20.5
37.4
Bz-VGAψ[CSe-NH]A-OMe 1 (9.9 ( 0.1) × 10-4
Bz-VGAψ[CSe-NH]F-OMe 2 (4.4 ( 0.1) × 10-4
f
Bz-VGAψ[CSe-N]P-OMe 3
1.3 × 10-5
a Determined at 10 °C. b Extrapolated to PSS. c pKa refers to NH
dissociation. d Not observable by NMR. e Reference 12. f Extrapolated
from the Eyring plot.
As peptidyl prolyl bonds play an important role in protein folding
and signaling,13 we characterized the photoisomerization of the
imidic selenoxo peptide bond in peptide 3. The UV/vis absorption
bands at 302 nm (ε ) 11 024 M-1cm-1) and 373 nm (ε ) 272
M-1cm-1) were attributed to the π-π* and n-π* transitions,
respectively (Figure S5). In the ground state, peptide 3 contained
∼8.5% cis isomer. After irradiation at 296 nm for 10 min, the cis
content increased to 37.4% (Table 1). However, unlike the case
for peptides 1 and 2, there was no obvious band shift for peptide
3 after photoswitching. It may result from the similar intramolecular
environment (alkyl groups) of CdSe in both the trans and cis
isomer, whereas the NH proton, in the case of secondary selenoxo
peptides, makes an isomer-specific difference. It is noteworthy that
the cis isomer was especially stable, with a half time of isomer-
ization of 70 min at 40 °C. The linear Eyring plot extrapolated to
(7) To stay consistent with the name “thioxo peptides” and to distinguish the
common “selenopeptide”, which contains selenocysteines, we refer to these
compounds as ‘selenoxo peptides’.
(8) Murai, T. In Chalcogenocarboxylic Acid DeriVatiVes; Kato, S., Ed.; Topics
in Current Chemistry, Vol. 251; Springer-Verlag Berlin: Berlin, 2005; pp
247-272.
(9) Olszewska, T.; Pyszno, A.; Milewska, M. J.; Gdaniec, M.; Polonski, T.
Tetrahedron: Asymmetry 2005, 16, 3711–3717.
(10) Baxter, I.; Hill, A. F.; Malget, J. M.; White, A. J. P.; Williams, D. J. Chem.
Commun. 1988, 1190–1191.
(11) (a) Glendening, E. D.; Hrabal, J. A. J. Am. Chem. Soc. 1997, 119, 12940–
12946. (b) Niu, S. Q.; Li, G. M.; Zingaro, R. A.; Reibenspies, J. H.; Ichiye,
T. Heteroatom Chem. 2002, 13, 380–386.
10 °C indicated a rotational barrier of 22.9 kcal mol-1
.
(12) Jensen, K. A. Arch. Pharm. Chem. Sci. Ed. 1981, 9, 93–116.
(13) (a) Lu, K. P.; Finn, G.; Lee, T. H.; Nicholson, L. K. Nat. Chem. Biol.
2007, 3, 619–629. (b) Fischer, G.; Aumu¨ller, T. ReV. Physiol. Biochem.
Pharmacol. 2004, 148, 105–150. (c) Eckert, B.; Martin, A.; Balbach, J.;
Schmid, F. X. Nat. Struct. Mol. Biol. 2005, 12, 619–623.
An additional feature peculiar to secondary amide selenoxo
peptides was a decrease of the 294 nm absorbance at pH 11.3 while
a new absorption band appeared near 260 nm, exhibiting an
isosbestic point at 275 nm (Figures S6 and S7). Moreover, the
original spectrum recovered when titrated from pH 11.3 to 6.3. No
JA1019386
9
J. AM. CHEM. SOC. VOL. 132, NO. 22, 2010 7579