A 19F NMR label to substitute polar amino acids in peptides: A CF3-substituted analogue of serine and threonine

Article abstract of DOI:10.1002/anie.201208069

Rigid & polar: The cyclobutane scaffold was used to design the first polar nonperturbing rigid CF₃-substituted amino acid (left in picture) suitable for replacing the serine/threonine residues in peptides. This amino acid imitates the geometry, structure, and function of serine and threonine, but in contrast to those, it can be used in structural studies of membrane-active Ser/Thr-containing peptides by solid-state ¹⁹F NMR spectroscopy. Copyright

Full text of DOI:10.1002/anie.201208069

.
Angewandte
Communications
DOI: 10.1002/anie.201208069
Amino Acids
19
A F NMR Label to Substitute Polar Amino Acids in Peptides:
A CF3-Substituted Analogue of Serine and Threonine**
Anton N. Tkachenko, Pavel K. Mykhailiuk, Sergii Afonin,* Dmytro S. Radchenko,
Vladimir S. Kubyshkin, Anne S. Ulrich,* and Igor V. Komarov*
Structural biology relies heavily on NMR studies of peptides
and proteins that are uniformly or selectively labeled with
NMR-active isotopes (2H, 13C, 15N, 19F). One of the most
informative approaches to analyze membrane-active peptides
in their natural membrane-bound form is solid-state 19F NMR
spectroscopy in oriented lipid bilayers.[1] The main advantages
of this approach are the exquisite sensitivity of fluorine, the
lack of any natural abundance background, the simplicity of
the NMR experiment and data analysis, especially when CF3-
substituted amino acids are used as labels (CF3 labels).
Numerous applications have demonstrated utility of the
designated CF3 labels 1–4,[2] incorporated into peptides in
the place of a bulky hydrophobic amino acid (1,2), a-
aminoisobutyric acid (3), or proline (4).[3] The 19F NMR
parameters from several selectively labeled analogues serve
as orientational constraints to yield the backbone conforma-
tion, alignment, and dynamic behavior of the peptide in the
lipid bilayer. The structure calculation relies entirely on the
fact that the CF3 group must be attached to the peptide
backbone with a fixed and well-defined angle, as seen in
1–4.
All the available CF3 labels 1–4 share the common
characteristic of having hydrophobic side chains. Therefore,
they can only be used as NMR reporters in the place of
similarly hydrophobic residues.[2,3] The substitution of any
polar amino acid with one of the CF3 labels 1–4 might disturb
the conformation and function of a peptide; this is highly
undesirable in structural studies.[1b,d,2d,e] The obvious lack of
appropriate polar and/or charged CF3 labels prompted us to
develop a novel molecular framework for CF3-substituted
amino acids with functionalized side chains. Here, we report
on the design and synthesis of a CF3-substituted serine/
threonine analogue, as well as on the validation of its use in
peptide structural studies. To our knowledge, this is the first
19F label for solid-state NMR analysis that possesses a polar
side chain.
Serine (Ser) and threonine (Thr) polar uncharged residues
are ubiquitous in peptides and proteins.[4] The hydroxy group
in the side chain is often involved in very specific protein
functions, for example, in posttranslational modifications
(phosphorylation, glycosylation), or in catalytic sites of
enzymes.[5] In transmembrane segments or in amphiphilic
helices of membrane proteins, Ser/Thr-rich motifs are known
to form polar binding platforms involved in the recognition of
glycans,[6] other proteins,[7] or to be responsible for the sensing
of membrane curvature.[8] The latter two functions are even
realized in the Ser/Thr-rich segments of stand-alone mem-
brane-active peptides.[7,8] By introducing a suitable CF3 label
to replace a single Ser/Thr residue, new possibilities would
open up to study the structure and function of such peptides in
membranes by using solid-state 19F NMR spectroscopy. This is
especially relevant in cases where the amphiphilic domains
contain only a low number of nonpolar amino acids, which
can make the entire segment inaccessible to the current
labeling scheme based on the known CF3 reporters 1–4.
The design of a CF3 label that is suitable to conservatively
replace Ser/Thr residues in peptides, and in general, the
design of any polar 19F NMR label represents a considerable
challenge.[1d,2d,e] The requirement for a CF3 label not to
perturb the structure and function of the peptides (criterion 1)
is particularly difficult to fulfill. The steric bulk and strong
electron-withdrawing character of fluorine-containing sub-
[*] A. N. Tkachenko, Dr. P. K. Mykhailiuk[+]
Organic Chemistry Department
Taras Shevchenko National University of Kyiv
Volodymyrska 64, 01601 Kyiv (Ukraine)
and
Enamine Ltd.
Matrosova 23, 01103 Kyiv (Ukraine)
Dr. S. Afonin,[+] Prof. A. S. Ulrich
Institute of Biological Interfaces (IBG-2)
Karlsruhe Institute of Technology (KIT)
POB 3640, 76021 Karlsruhe (Germany)
E-mail: sergiy.afonin@kit.edu
anne.ulrich@kit.edu
Homepage: http://www.ibg.kit.edu/nmr/
Dr. D. S. Radchenko, Prof. I. V. Komarov
Institute of High Technologies
Taras Shevchenko National University of Kyiv
Volodymyrska 60, 01601 Kyiv (Ukraine)
E-mail: ik214@yahoo.com
Homepage: http://418lab.chem.univ.kiev.ua
V. S. Kubyshkin, Prof. A. S. Ulrich
Institute of Organic Chemistry and CFN, KIT
Fritz-Haber-Weg 6, 76131 Karlsruhe (Germany)
[+] These authors contributed equally to this work.
[**] We are grateful to Dr. O. V. Shishkin for the X-ray diffractional
analysis and Dr. M. Berditsch for help with antimicrobial assays.
This work was supported by Enamine Ltd. and by DAAD (A.N.T.).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201208069.
1486
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1486 –1489

Angewandte
Chemie
stituents[9] inevitably cause changes in the properties of the
adjacent functional groups. The labeled amino acid must thus
carry the CF3 moiety sufficiently distant from its amino and
carboxylic groups and from any functional groups in the side
chain. In addition, the structural rigidity already mentioned
above (criterion 2), as well as resistance to racemization and
compatibility with standard peptide synthesis procedures
(criterion 3) must also be implemented in the design. One of
the possible logical routes towards designing the new CF3-
substituted Ser/Thr analogue that we report here is illustrated
in Scheme 1.
Scheme 1. Design of CF3 label 6, an analogue of serine or threonine.
Scheme 2. Synthesis of OH-protected cis-13. a) CF3SiMe3, CsF, THF,
258C, 15 h; b) NaH, BnBr, DMF, 258C, 18 h; c) KOH, iPrOH, reflux,
3 h, then HCl; d) (PhO)2P(O)N3, Et3N, toluene, 808C, 5 h, tBuOH,
808C, 15 h; e) 6n HCl, reflux, 6 h; f) KOH, EtOH, reflux, 2 h; g) 2n
HCl in Et2O. Boc=tert-butoxycarbonyl.
To minimize any electronic and structural perturbations of
the polypeptide, the bulky electron-withdrawing CF3 reporter
should be placed at the most distal position to the amino-
carboxylate function, that is, at the Cb atom of the parent Ser/
Thr to construct structure 5. This compound, however, is still
a poor CF3 label. First, molecule 5 is conformationally
flexible, which is undesirable for an NMR label (criterion 2).
Second, the CF3 group is expected to increase the acidity of
the hydroxy function in 5 compared to that of Ser/Thr.
Furthermore, the amino acid (2S,3S)-5[10] substituted for Thr
in an enkephalin-derived hexapeptide had been shown to lead
to a conformational alteration in the main chain of the
peptide owing to the strong electron-withdrawing effect of the
CF3 group.[11]
crystallization and hydrolyzed to cis-12. The molecular
structure of cis-11 was confirmed by X-ray analysis
(Figure 1). Notably, all the protecting groups, including the
A comparison of the reported pKa values for the OH
groups of methanol (15.9),[12] 2,2,2-trifluoroethanol (12.4),[13]
and of Ser/Thr (ca. 13), suggests that the influence of the CF3
group on the acidity of the side chain hydroxy group is almost
equal to that of the aminocarboxylate moiety. Therefore, if
the electron-withdrawing effect of the aminocarboxylate
could be eliminated, then the acidity of the hydroxy group
would be expected to remain comparable to that in Ser/Thr.
To implement this idea in our design of the label, we decided
to place a cyclobutane ring between the aminocarboxylic
moiety and the OH group. This should also diminish further
the electronic and steric influence of the CF3 group on the
aminocarboxylate moiety, and provide the necessary rigidity
(criterion 2) to the CF3 label 6 thus designed.
Figure 1. Molecular structure of cis-11.[20] For clarity, individual con-
formers are selected for the iPr and benzyloxy substituents (see the
Supporting Information for details). F green, O red, N blue.
benzyl group, were readily removable (yielding cis-6·HCl)
under reflux of cis-11 in 6n HCl. Later on, we advantageously
used this acid lability of the OBn protection in the final stages
of solid-phase peptide synthesis (SPPS). Deprotection of cis-
12 under mild conditions gave the OBn derivative, cis-13.
Installment of the N-Fmoc protecting group onto cis-13,
to prepare the building block for use in SPPS, required some
experimentation. The standard methods,[17] which employ
Fmoc-Cl or N-(9-fluorenylmethoxycarbonyloxy)succinimide
(Fmoc-OSu) under Schotten–Baumann conditions, gave very
poor yields of the protected compound cis-14. The two-step
approach,[18] without isolation of the putative bis-silylated
intermediate cis-15 resulted in acceptable 54% yield of cis-14
(Scheme 3).
The synthesis of 6 started from the ketodicarboxylate 7[14]
(Scheme 2), which was transformed into the trifluoromethyl
alcohol 8 by using the Ruppert–Prakash reagent[15,16] in
a reasonable yield. As a convenient OH protecting group
needed for further transformations we chose the benzyl
group, considering its ease of introduction and orthogonality
to the further planned N-Fmoc protection. The OBn deriv-
ative 9 was subjected to monohydrolysis followed by a Curtius
rearrangement and N-Boc protection. This set of reactions led
to a mixture of diastereomers, in which cis-11 prevailed (cis/
trans ca. 2:1); it was isolated in 39% yield after fractional
To evaluate the compatibility of cis-14 with the SPPS
(criterion 3), we first synthesized a model tripeptide Ac-Gly-
(cis-6)-Gly-NH2 (16). The OBn protecting group was
removed by prolonged (ca. 7 days) treatment with the
Angew. Chem. Int. Ed. 2013, 52, 1486 –1489
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1487

.
Angewandte
Communications
Scheme 3. Synthesis of OBn-N-Fmoc-protected cis-14. a) SiMe3Cl,
DIEA, CH2Cl2, 08C, 30 min; b) Fmoc-Cl, 258C, 3 h. DIEA=diisopropyl-
ethylamine.
cleavage cocktail during the obligatory cleavage off the resin.
The synthesis of 16 also served an additional purpose, namely,
to provide a polypeptide framework for studying the acid–
base properties of the cis-6 side chain. The pH-dependent
ionization was studied by liquid-state NMR spectroscopy. The
19F NMR chemical shift of the CF3 group was sensitive to the
protonation state of the OH group, thus allowing us to
determine the titration curve for 16, giving a pKa value of
11.7 Æ 0.1 (23 8C). The value is close to the corresponding pKa
value in both Ser and Thr (ca. 13), thus justifying our design.
Hence, from a chemical perspective amino acid cis-6 could
serve as a replacement of these natural amino acids. It
resembles them not only by size, polar character of the side
chain, and spatial position of the hydroxy group, but also by
the acid–base properties of this functional group.
Figure 2. Evaluation of cis-6 as a new, polar CF3 label. A) Circular
dichroism analysis (0.05 mgmLÀ1 peptide concentration, 258C) of
TA-17-wt (top), TA-18 (middle), and TA-19 (bottom). B) Solid-state
19F NMR spectra of TA-18 (top) and TA-19 (bottom) in oriented
dimyristoyl phosphatidylcholine bilayers, with varying peptide content
(in mol%). PB=phosphate buffer, TFE=2,2,2-trifluoroethanol,
SDS=sodium dodecyl sulfate.
To validate the use of cis-6 as a 19F NMR label in
structural studies, we chose the natural membrane-active
antimicrobial peptide Temporin A (TA)[19] as a model com-
pound.
We
synthesized
the
wild-type
sequence
(FLPLIGRVLSGIL-NH2, TA-17-wt), and two CF3-labeled
analogues, by using either the novel cis-6 (FLPLIGRVL-cis-6-
GIL-NH2, TA-18), or the known hydrophobic CF3 label 2
(FLPLIGRVL-2-GIL-NH2, TA-19). In each case the native
Ser at position 10 was replaced. In accordance with our
published approach,[1] the peptides were compared in terms of
their functional activity, structure, and behavior in the solid-
state 19F NMR experiment.
In a standard antimicrobial assay (two-fold dilution series
to obtain the minimal inhibitory concentration, MIC), all
three peptides were active against Gram-positive S. aureus
and showed low activity against Gram-negative E. coli, so the
modified peptides TA-18 and TA-19 displayed the same
selectivity as the wild-type TA. The MIC values, however,
varied significantly depending on the mutation at the Ser
position. While for both TA-17-wt and TA-18 the MIC values
were comparable (8 mgmLÀ1), in the case of TA-19 the
activity decreased to a MIC of 32 mgmLÀ1. Furthermore, the
circular dichroism spectra (Figure 2A) showed that both
TA-17-wt and TA-18 adopt a random coil conformation in
aqueous solutions but fold as a helix in membrane-mimetic
environments. Peptide TA-19, in contrast, was helical only in
the presence of TFE but not SDS, and exhibited a significant
tendency to aggregate in aqueous buffer and in the presence
of detergent micelles. These data demonstrate the superiority
of cis-6 over 2 as a CF3 label for Ser.
spectra, data not shown). Under these conditions, TA-18 and
TA-19 showed
a synchronous concentration-dependent
change of the 19F NMR signal (Figure 2B), with a threshold
at approximately 2 mol% peptide. This change in splitting is
indicative of a realignment of the helical peptide in the lipid
membrane, as previously observed for several other peptides
as well.[1,3b,d–h] Remarkably, the dipolar splitting of the CF3
group in the peptide containing 2 was large when the cis-6-
labeled peptide showed a minimal value (ca. 0 kHz), and vice
versa. Thus, the CF3 groups of the two labels appear to be
aligned orthogonally with respect to each other. This relation-
ship is useful, as one could obtain two independent NMR
constrains from the same site of the peptide (as in refer-
ence [3h]), provided that it can be nonperturbingly substi-
tuted with 2 and cis-6. The full structural analysis of the
membrane-bound Temporin A peptide will require several
such NMR constraints[1] and is currently ongoing.
In conclusion, we have designed, synthesized, and con-
firmed the compatibility of cis-1-amino-3-hydroxy-(trifluoro-
methyl)cyclobutanecarboxylic acid (cis-6) as the first polar
CF3 label for solid-state 19F NMR structure analysis of
membrane-active peptides.
Received: October 7, 2012
Published online: December 12, 2012
Both the labeled peptides, TA-18 and TA-19, were readily
reconstituted in oriented lipid bilayers despite the solubility
differences, and remained nonaggregated and uniformly
aligned. The bilayer integrity was preserved in both samples
and did not change with time (as judged from 31P NMR
Keywords: amino acids · fluorine · serine · NMR spectroscopy ·
threonine
.
1488
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1486 –1489

Angewandte
Chemie
[4] M. B. Swindells, M. W. MacArthur, J. M. Thornton, Nat. Struct.
[1] a) A. S. Ulrich, Prog. Nucl. Magn. Reson. Spectrosc. 2005, 46, 1 –
21; b) A. S. Ulrich, P. Wadhwani, U. H. N. Dꢀrr, S. Afonin, R. W.
Glaser, E. Strandberg, P. Tremouilhac, C. Sachse, M. Berditch-
evskaia, S. L. Grage in NMR spectroscopy of biological solids
(Ed.: A. Ramamoorthy), CRC Press, Boca Raton, FL, 2006,
p. 215 – 236; c) S. L. Grage, S. Afonin, A. S. Ulrich, Methods
Mol. Biol. 2010, 618, 183 – 207; d) V. S. Kubyshkin, I. V.
Komarov, S. Afonin, P. K. Mykhailiuk, S. L. Grage, A. S.
Ulrich in Fluorine in Pharmaceutical and Medicinal Chemistry:
From Biophysical Aspects to Clinical Applications (Eds.: V.
Gouverneur, K. Mꢀller), Imperial College Press, London, 2012,
pp. 91 – 139; e) K. Koch, S. Afonin, M. Ieronimo, M. Berditsch,
A. S. Urich, Top. Curr. Chem. 2012, 306, 89 – 118.
[2] a) S. Afonin, R. W. Glaser, M. Berditchevskaia, P. Wadhwani,
K. H. Gꢀhrs, U. Mçllmann, A. Perner, A. S. Ulrich, ChemBio-
Chem 2003, 4, 1151 – 1163; b) R. W. Glaser, C. Sachse, U. H. N.
Dꢀrr, P. Wadhwani, A. S. Ulrich, J. Magn. Reson. 2004, 168, 153 –
163; c) E. Strandberg, P. Wadhwani, P. Tremouilhac, U. H. N.
Dꢀrr, A. S. Ulrich, Biophys. J. 2006, 90, 1676 – 1686; d) P. K.
Mikhailiuk, S. Afonin, A. N. Chernega, E. B. Rusanov, M. O.
Platonov, G. G. Dubinina, M. Berditsch, A. S. Ulrich, I. V.
Komarov, Angew. Chem. 2006, 118, 5787 – 5789; Angew. Chem.
Int. Ed. 2006, 45, 5659 – 5661; e) P. K. Mykhailiuk, S. Afonin,
G. V. Palamarchuk, O. V. Shishkin, A. S. Ulrich, I. V. Komarov,
Angew. Chem. 2008, 120, 5849 – 5851; Angew. Chem. Int. Ed.
2008, 47, 5765 – 5767.
[3] a) S. Afonin, U. H. N. Dꢀrr, R. W. Glaser, A. S. Ulrich, Magn.
Reson. Chem. 2004, 42, 195 – 203; b) R. W. Glaser, C. Sachse,
U. H. N. Dꢀrr, P. Wadhwani, S. Afonin, E. Strandberg, A. S.
Ulrich, Biophys. J. 2005, 88, 3392 – 3397; c) S. L. Grage, A. V.
Suleymanova, S. Afonin, P. Wadhwani, A. S. Ulrich, J. Magn.
Reson. 2006, 183, 77 – 86; d) S. Afonin, P. K. Mykhailiuk, I. V.
Komarov, A. S. Ulrich, J. Pept. Sci. 2007, 13, 614 – 623; e) P.
Wadhwani, J. Bꢀrck, E. Strandberg, C. Mink, S. Afonin, A. S.
Ulrich, J. Am. Chem. Soc. 2008, 130, 16515 – 16517; f) S. Afonin,
S. L. Grage, M. Ieronimo, P. Wadhwani, A. S. Ulrich, J. Am.
Chem. Soc. 2008, 130, 16512 – 16514; g) E. Strandberg, N.
Kanithasen, D. Tiltak, J. Bꢀrck, P. Wadhwani, O. Zwernemann,
A. S. Ulrich, Biochemistry 2008, 47, 2601 – 2616; h) D. Maisch, P.
Wadhwani, S. Afonin, C. Bçttcher, B. Koksch, A. S. Ulrich, J.
Am. Chem. Soc. 2009, 131, 15596 – 15597; i) M. Salwiczek, P. K.
Mikhailiuk, S. Afonin, I. V. Komarov, A. S. Ulrich, B. Koksch,
Amino Acids 2010, 39, 1589 – 1593; j) M. Ieronimo, S. Afonin, K.
Koch, M. Berditsch, P. Wadhwani, A. S. Ulrich, J. Am. Chem.
Soc. 2010, 132, 8822 – 8824; k) D. Grasnick, U. Sternberg, E.
Strandberg, P. Wadhwani, A. S. Ulrich, Eur. Biophys. J. 2011, 40,
529 – 543; l) P. Wadhwani, E. Strandberg, N. Heidenreich, J.
Bꢀrck, S. Fanghꢁnel, A. S. Ulrich, J. Am. Chem. Soc. 2012, 134,
6512 – 6515.
Biol. 1995, 2, 596 – 603.
[5] a) G. M. Simon, B. F. Cravatt, J. Biol. Chem. 2010, 285, 11051 –
11055; b) Review: S. Urban, W. Dickey, Genome Biol. 2011, 12,
231 – 241.
[6] A. Boisramꢂ, A. Cornu, G. Da Costa, M. L. Richard, Eukaryotic
Cell 2011, 10, 1317 – 1330.
[7] a) J. P. Dawson, J. S. Weinger, D. M. Engelman, J. Mol. Biol.
2002, 316, 799 – 805; b) B. North, L. Cristian, X. Fu Stowell, J. D.
Lear, J. G. Saven, W. F. Degrado, J. Mol. Biol. 2006, 359, 930 –
939; c) J. M. Rauceo, R. De Armond, H. Otoo, P. C. Kahn, S. A.
Klotz, N. K. Gaur, P. N. Lipke, Eukaryotic Cell 2006, 5, 1664 –
1673.
[8] a) B. Antonny, Annu. Rev. Biochem. 2011, 80, 101 – 123; b) G.
Drin, B. Antonny, FEBS Lett. 2010, 584, 1840 – 1847.
[9] a) B. E. Smart, Properties of Fluorinated Organic Compounds.
ACS Monograph 187, Vol.1, ACS, Washington DC, 1995;
b) B. E. Smart, J. Fluorine Chem. 2001, 109, 3 – 11; c) R. Smits,
C. D. Cadicamo, K. Burger, B. Koksch, Chem. Soc. Rev. 2008, 37,
1727 – 1739; d) V. A. Soloshonok, I. I. Gerus, Y. L. Yagupolskii,
V. P. Kukhar, Zh. Org. Khim. 1987, 23, 2308 – 2313 [Chem. Abstr.
109:55185].
[10] D. Seebach, E. Juaristi, D. D. Miller, C. Schickli, T. Weber, Helv.
Chim. Acta 1987, 70, 237 – 261.
[11] T. Kitamoto, S. Marubayashi, T. Yamazaki, Tetrahedron 2008, 64,
1888 – 1894.
[12] P. Ballinger, F. A. Long, J. Am. Chem. Soc. 1960, 82, 795 – 798.
[13] P. Ballinger, F. A. Long, J. Am. Chem. Soc. 1959, 81, 1050 – 1053.
[14] D. S. Radchenko, P. K. Mykhailiuk, I. V. Komarov, A. V. Bez-
dudny, Synlett 2009, 1827 – 1829.
[15] I. Ruppert, K. Schlich, W. Volbach, Tetrahedron Lett. 1984, 25,
2195 – 2198.
[16] G. K. S. Prakash, A. K. Yudin, Chem. Rev. 1997, 97, 757 – 786.
[17] a) C. D. Chang, M. Waki, M. Ahmad, J. Meienhofer, E. O.
Lundell, J. D. Haug, Int. J. Peptide Prot. Res. 1980, 15, 59 – 66;
b) R. C.deL. Milton, E. Becker, S. C. F. Milton, J. E. J. Baxter,
J. F. Elsworth, Int. J. Peptide Prot. Res. 1987, 30, 431 – 432;
c) D. A. Wellings, E. Atherton, Methods Enzymol. 1997, 289,
44 – 67.
[18] D. R. Bolin, I. I. Sytwu, F. Humiec, J. Meienhofer, Int. J. Peptide
Prot. Res. 1989, 33, 353 – 359.
[19] A. Giacometti, O. Cirioni, W. Kamysz, G. D’Amato, C. Silvestri,
M. S. Del Prete, A. Licci, J. Lukasiak, G. Scalise, J. Antimicrob.
Chemother. 2005, 55, 272 – 274.
[20] CCDC-890358 (1) contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif
Angew. Chem. Int. Ed. 2013, 52, 1486 –1489
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1489