Reversible Folding of a β-Hairpin Peptide by a Metal-Chelating Amino Acid

Article abstract of DOI:10.1002/chem.201700698

5-(1-Hydroxy-pyridin-2(1H)-onyl)-l-alanine (Hop) is a N-hydroxy-1,2-pyridone functionalized α-amino acid with the desired metal-chelating properties of DOPA (3,4-dihydroxy phenylalanine) but without its unwanted redox activity. The Fmoc-protected amino acid Fmoc-l-Hop(tBu)-OH (11) was synthesized from glycine phosphonate followed by enzymatic hydrolysis of the methyl ester yielding the Hop l-isomer in 96 % ee. The amino acid 11 is used in automated peptide synthesis for the assembly of a 14mer β-hairpin peptide with the sequence [dsb^{1, 14}]H-CHXETGKHGHKLVC-OH (X=W, l-Hop). While the 10 π electron containing indole side chain of l-Trp in peptide 14 completes the formation of a hydrophobic cluster and results in 90 % folding, the folded fraction is significantly decreased to approximately 30 % for the 6 π electron l-Hop side chain in peptide 16. Metal chelation of Ga³⁺ reconstitutes the folding of 16 to above 60 % due to the formation of the Ga(16)₃ trimer. The chelation process of 16 is monitored by NMR spectroscopy and the subsequent release of Ga³⁺ by a competitive metal chelator exemplifies the reversible oligomerization of peptide epitopes by metal chelation, bearing the opportunity to synthesize protein-sized aggregates on the basis of reversible chemistry in water.

Full text of DOI:10.1002/chem.201700698

DOI: 10.1002/chem.201700698
Full Paper
&
Amino Acids
Reversible Folding of a b-Hairpin Peptide by a Metal-Chelating
Amino Acid
Jan Reutzel, Timm M. Diogo, and Armin Geyer*[a]
Abstract: 5-(1-Hydroxy-pyridin-2(1H)-onyl)-l-alanine (Hop) is
a N-hydroxy-1,2-pyridone functionalized a-amino acid with
the desired metal-chelating properties of DOPA (3,4-dihy-
droxy phenylalanine) but without its unwanted redox activi-
ty. The Fmoc-protected amino acid Fmoc-l-Hop(tBu)-OH (11)
was synthesized from glycine phosphonate followed by en-
zymatic hydrolysis of the methyl ester yielding the Hop l-
isomer in 96% ee. The amino acid 11 is used in automated
peptide synthesis for the assembly of a 14mer b-hairpin pep-
tide with the sequence [dsb1,14]H-CHXETGKHGHKLVC-OH
(X=W, l-Hop). While the 10 p electron containing indole
side chain of l-Trp in peptide 14 completes the formation of
a hydrophobic cluster and results in 90% folding, the folded
fraction is significantly decreased to approximately 30% for
the 6 p electron l-Hop side chain in peptide 16. Metal chela-
tion of Ga3+ reconstitutes the folding of 16 to above 60%
due to the formation of the Ga(16)3 trimer. The chelation
process of 16 is monitored by NMR spectroscopy and the
subsequent release of Ga3+ by a competitive metal chelator
exemplifies the reversible oligomerization of peptide epi-
topes by metal chelation, bearing the opportunity to synthe-
size protein-sized aggregates on the basis of reversible
chemistry in water.
of Parkinson’s disease.[15] While DOPA unspecifically chelates
metals[15,16] or even silica,[17] HOPOs are known to be excellent
chelators for Fe3+, what makes the amino acid a potential tool
for the treatment of iron overload related diseases such as
cystic fibrosis.[18–22] The uniqueness of the HOPO-functionalized
side chain defines a new class of metal-chelating amino acids
that differ from the often used bipyridine,[23] 8-HQA,[24] or
EDTA/DOTA-like side chain functionalizations[25–27] in metallo-
peptide chemistry. The amino acid was incorporated into
a cyclic b-hairpin peptide of the eif5a hypusination epitope
loop, which has a biological relevance in cancer and HIV re-
search.[28–31] The eif5a protein carries a unique posttranslational
modification (hypusination) that transmits its biological func-
tions such as protein elongation, translation of polyproline
motifs, and cell proliferation. This modification of the K50 side
chain is part of the reverse turn G49KHG52 that is highly con-
served among all eukaryotes.[32–35] The complete b-loop S47-K58
is well resolved in the crystal structure[35] but flexible in solu-
tion.[36] This remarkably long hairpin contains two flexible gly-
cine residues and therefore poses an ambitious challenge for
our recently developed macrocyclization methods for hairpin
epitopes in which the hydrophobic cluster developed by Co-
chran et al.[37–39] serves as a reliable module for the rigidifica-
tion of diverse tetrapeptide loops.[40,41] This method is success-
fully applied here for the b-loop of eif5a to obtain an antiparal-
lel b-hairpin that includes the GKHG b-turn of eif5a. This pep-
tide is then used to develop a general method to trimerize epi-
topes, because eif5a, like many other proteins, forms
oligomers. To this end, we incorporate the new amino acid
Hop into the b-hairpin. We then investigate the influence of
Hop on the folding of the monomeric peptide 16 and finally
Introduction
N-Hydroxy-1,2-pyridones (HOPOs) are versatile structures in
medicinal chemistry with properties ranging from anti-malari-
al,[1] to anti-viral[2] and anti-fungal activities.[2,3] High relaxivity
Gd-HOPO magnetic resonance imaging (MRI) contrast agents
were developed based on their metal-chelating properties.[4–8]
Furthermore, HOPO analogs were used in peptide coupling re-
actions.[9] Here, we present the synthesis of the HOPO-func-
tionalized a-amino acid Fmoc-l-Hop(tBu)-OH (11), starting
from a glycine phosphonate followed by an enzymatic hydroly-
sis with a protease from Bacillus licheniformis to obtain the l-
configured amino acid in 96% ee. Although several synthetic
procedures towards enantioenriched non-natural a-amino
acids are known, such as the asymmetric Strecker syntheses,[10]
the asymmetric Rh-catalyzed hydrogenation of dehydro amino
acids,[11,12] or the stereo induction by Evans-type alkylation or
aziridination[13,14] the described method using enzymatic hy-
drolysis allows rapid access to the enantiopure Fmoc-protected
amino acid 11 for direct usage in automated peptide synthesis.
The herein used synthetic procedures can possibly be adapted
towards a large substrate scope.
The amino acid 11 is a redox-stable analogue to 3,4-dihy-
droxy phenylalanine (DOPA), which is used for the treatment
[a] J. Reutzel, Dr. T. M. Diogo, Prof. Dr. A. Geyer
Department of Chemistry, Philipps-University Marburg
Hans-Meerwein-Str. 4, 35032 Marburg (Germany)
E-mail: geyer@staff.uni-marburg.de
Supporting Information and ORCID identification number(s) for the au-
thor(s) of this article can be found under
https://doi.org/10.1002/chem.201700698.
Chem. Eur. J. 2017, 23, 1 – 8
1
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
use its metal-chelating property for epitope oligomerization.
The variation of Hop along a peptide strand allows the trimeri-
zation of further epitopes by metal chelation, bearing the op-
portunity to synthesize aggregates in the size of naturally oc-
curring proteins in a minimal amount of steps based on rever-
sible chemistry in aqueous media.
(12 g of ester were hydrolyzed in a one-step procedure). We
were confident that the proteolytic enzyme would recognize
Cbz-dl-Hop(tBu)-OMe (7) as a substrate and that the enantio-
selective enzymatic hydrolysis of the methyl ester 7 yields the
l-configurated acid 8 and d-configurated ester 9.[48,50] To verify
the stereo specificity of the enzyme, we studied the hydrolysis
of l-, d-, and dl-phenyl alanine methyl ester by NMR-spectros-
copy using a watergate-pulse sequence for water signal sup-
pression. In a set of three experiments a defined amount of
the respective phenyl alanine methyl ester was dissolved in
aqueous phosphate buffer (pH 7.4) and treated with the
Results and Discussion
Synthesis of the protected 1,2-HOPO amino acid
1
enzyme; H-NMR spectra were recorded every 30 min and the
Olefination of the Cbz-protected phosphono glycine trimethy-
lester (1)[42,43] with 5-formyl 2-methoxy pyridine (2)[44,45] yielded
the dehydro amino acid 3. Compound 3 was obtained in an
remaining amount of the methyl ester relative to the carboxyl-
ic acid was calculated from integration of the a protons. Under
these conditions, the d-ester remained nearly unreacted after
24 h, while the l-isomer was completely hydrolyzed after
35 min, providing the opportunity to obtain the enantiopure l-
amino acid by liquid extraction (see the Supporting Informa-
tion for details). The enzymatic hydrolysis of the protected
HOPO amino acid 7 could not be monitored on NMR scale be-
cause of its insolubility in pure phosphate buffer. Various pa-
rameters were investigated such as the N-terminal protecting
group and addition of organic solvents. The conversion was
monitored by RP-HPLC (ACE 3, C18, 3 mm, 150 mm x 3.0 mm
column) and the best combinations were found with Cbz as
the amine protecting group and 1,4-dioxane as a co-solvent
(Figure 1). Changing the N-terminal protecting group to Fmoc
yielded the side-chain-protected amino acid Fmoc-l-Hop(tBu)-
OH 11 with 21% overall yield and in 96% ee over eight steps.
The racemic Fmoc-dl-Hop(tBu)-OH 10 was obtained by saponi-
fication, followed by hydrogenolytic cleavage of the Cbz pro-
tecting group and finalized by Fmoc protection. The l-amino
acid was isolated by liquid extraction and again esterified to in-
1
8:1 mixture of E/Z isomers as determined by H-NMR spectros-
copy. Sodium borohydride reduction in the presence of nickel(-
II)chloride yielded the protected racemic a-amino acid 4
(Scheme 1).
Scheme 1. Syntheses of the racemic 10 and enantiopure amino acid Fmoc-
l-Hop(tBu)-OH (11).
The pyridine nitrogen of 4 was successfully oxidized with
meta-chloroperbenzoic acid to give the N-oxide 5. One-pot
acylation of the N-oxide with acetyl chloride and subsequent
methanolysis released the 1,2-hydroxypyridone 6. The tBu pro-
tecting group was introduced as an acid-labile protecting
group for Fmoc-based peptide synthesis to obtain compound
7. The key step was the enzymatic hydrolysis of the methyl
ester with a protease from Bacillus licheniformis (Alcalaseꢁ).[46–50]
The enzymatic hydrolysis of a-amino acid methyl esters with
this protease is known to preferentially occur with substrates
bearing an aromatic substituent in b position.[46,49] Further-
more, it is very cheap and amenable to large scale syntheses
Figure 1. a) Enzymatic hydrolysis of the racemic amino acid ester 7. The l-
configurated carboxylic acid 8 was converted to its methyl ester again to
result in 96% ee determined by chiral HPLC. b) Optimization of the enzymat-
ic hydrolysis of racemic Cbz-dl-Hop(tBu)-OMe 7 as monitored by RP-HPLC.
The half-life term is related to the substrate methyl ester 7, therefore the Bn
protecting group showed slow conversion in comparison to the Cbz pro-
tecting group. The use of DMF as additive decreased the selectivity of the
enzyme. The best results were observed for a-Cbz protecting group and 1,4-
dioxane as co-solvent. The Fmoc protecting group was not compatible be-
cause of its insolubility in these solvent mixtures.
&
&
Chem. Eur. J. 2017, 23, 1 – 8
www.chemeurj.org
2
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
crease its solubility for the normal-phase chiral HPLC analysis,
leading to an enantiomeric excess of 96%.
The ee was confirmed from the dipeptide coupling of the
Fmoc-protected amino acids 10 and 11 with l-Phe-OMe·HCl.
1
The diastereomeric ratio was determined from the H NMR in-
tegral ratio of the amide protons. Assuming that no stereo in-
formation on l-Phe-OMe was lost during the peptide coupling
reaction, the ee was determined directly from the d.r.
(Scheme 2).
Figure 2. NMR titration (300 MHz, 300 K, [D6]DMSO). The upper spectrum
shows the Ga(HOPO)3 complex that was isolated as a solid before. The lower
spectrum shows the HOPO without metal. The expected 3:1 coordination
between 1,2-HOPO and Ga3+ (red) was identified from the increasing educt
signal (blue) after exceeding the 3:1 coordination in the presence of excess
1,2-HOPO. The minor signal set indicates an unknown complex geometry.
Scheme 2. Syntheses of dipeptides from the Fmoc-protected HOPO amino
acids 10 and 11; d.r. and ee were determined by NMR spectroscopy. The in-
tegral ratio of the amidic protons from 12a in comparison to 12b gives the
d.r. The ee was determined under the assumption that no stereo information
of l-Phe is lost during the reaction.
set. Compared to optical spectroscopic methods, NMR spec-
troscopy has the advantage that the 1H signal intensity of
slowly equilibrating complexes yields their relative amounts in
a single experiments. The titration was repeated in H2O/D2O
but the resulting Ga–HOPO complex precipitated from the
NMR solution. This is why the titration in water was performed
directly with peptides, which therefore directly leaded to
metal-chelating experiments with peptides.
Metal-chelating properties
The metal-chelating properties of the 1,2-HOPO motif were
studied by NMR spectroscopy. Ga3+ was used instead of Fe3+
to avoid paramagnetic spectra. The ionic radii of Ga3+ and
Fe3+ are very similar, furthermore, Ga3+ is redox inert and
highly soluble in DMSO and water.[51] A slow exchange rate of
Peptide synthesis
The chiral amino acid Fmoc-l-Hop(tBu)-OH (11) and the race-
mate 10 were used in automated solid-phase peptide synthe-
sis. The peptides are designed by grafting of the eif5a epitope
(T5-H10) onto the b-hairpin motif (C1-E4, K11-C14) developed by
Cochran et al., which stabilizes the b-hairpin by a cysteine
bridge and a hydrophobic cluster of aromatic and aliphatic
side chain contacts (Figure 3).[37–39] Hop was inserted at the po-
sition of Trp into the b-hairpin motif and, like most other
amino acids, is too small or otherwise not qualified to mimic
the known Trp3 side-chain hydrophobic contact with the Leu12
d-methyl groups. The absence of this contact results in
a strong deshielding of the diastereomeric Leu d-CH3 pro-
tons.[41,52] To test this hypothesis and identify metal-chelation
protocols, we synthesized a number of peptides differing by
a single side chain.
1
the metal ion in the H-NMR allows the identification of differ-
ent diastereomeric arrangements of the ligand around the
metal or between other minor components formed during the
titration process. One equivalent of Ga(NO3)3 hydrate was dis-
solved in 450 mL [D6]DMSO and 1, 2, 3, or 4 equivalents of 2-hy-
droxypyridine-1-oxide (1,2-HOPO) were added in 150 mL
1
[D6]DMSO. H NMR spectra were recorded and the downfield
shift of the aromatic signals indicated the complex formation
(Figure 2). At 300 MHz and room temperature, the complex is
in slow equilibrium with the uncomplexed species.
The freshly dissolved solid Ga(HOPO)3 complex 12 (Figure 2,
top) identifies the end point of the titration. The 3:1 ratio of
the isolated compound 12 was proven by elemental analysis.
By adding Ga3+ to the 1,2-HOPO solution, the signals of the
uncomplexed HOPO completely disappear. The new signal set
in the 1:1 ratio from metal to ligand is converted to another
new signal set that arises in the 2:1 spectra and reaches its
maximum in the 3:1 spectra, before the amount of unbound
1,2-HOPO increases again. The ratio of complexed to uncom-
plexed HOPO decreases significantly from the 3:1 spectra, indi-
cating a 3:1 coordination from 1,2-HOPO to Ga3+. The same ti-
tration was performed with the amino acid 6 and showed
a similar trend but without a detectable second minor signal
Our hypothesis was that the aromatic side chain Hop is too
small to stabilize the monomeric b-hairpin but that the cooper-
ative binding of three Hop-containing peptides around one
Ga3+ ion should restructure the b-hairpin shape. First, the
native l-Trp3 was exchanged for Hop and several other aromat-
ic amino acids with the aim of studying the influence of the ar-
omatic side chain on the hydrophobic interactions and the sta-
bility of the b-hairpin. Figure 3 shows the side chains and
Table 1 shows the linear peptide sequences.
Chem. Eur. J. 2017, 23, 1 – 8
www.chemeurj.org
3
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
The linear precursors of peptides 15–19 were oxidized in
aqueous buffer for disulfide formation, purified by RP-HPLC
and analyzed by NMR spectroscopy in aqueous phosphate
buffer/D2O (9:1, pH 3.0). We used the already published meth-
ods to characterize and quantify the amount of folded b-hair-
pins containing the hydrophobic cluster developed by Cochran
et al.[37–39,41,40] Qualitatively, the dispersion of the amide NH pro-
tons is an obvious indicator of an intact hydrophobic cluster
(Figure 4). Second, both Gly residues in positions 6 and 9 next
to the turn appear as a dd instead of a t, if they are in a pre-
ferred conformation. Third, the d-CH3 groups of Leu12 receive
an upfield shift towards negative ppm regions from the ring
current of the aromatic side chain in position 3 of the b-hair-
pin. Figure 4 sums up the results of the peptide synthesis. The
l-Hop-containing peptide was obtained from the use of the
enantiopure amino acid 11 during peptide synthesis, while the
d-Hop peptide was isolated by incorporation of the racemic
amino acid 10, followed by the separation of the diastereo-
meric peptides by RP-HPLC. The highest dispersion of the
amide region was observed for the peptide containing 1-l-
Naphthylalanine (1-l-Nal) together with the Leu12-d-CH3 upfield
shift and the diastereotopic splitting of the two Leu12 methyl
groups. The dispersion of Hop peptides is slightly higher for
the l-isomer, while the d-isomer behaves very similar to l-di-
phenylalanine (Dpa), which has no stabilizing effect. Compar-
ing the oxidized and reduced form of the l-Trp peptide 14
and 15, it is evident that the hydrophobic contacts of the di-
sulfide are necessary to complete the hydrophobic cluster of
the b-hairpin. Another important structural property is the bi-
cyclic aromatic 10 p electron ring system at position 3 of the
b-hairpin. All substitutions, except the conservative Nal peptide
19, significantly destabilize the folding, because the hydropho-
bic cluster is incomplete without a large side chain. We esti-
mated the destabilizing effects with a two-site model in fast
chemical exchange, which allows the calculation of a “folded
Figure 3. The hydrophobic cluster that was developed by Cochran and cow-
orkers comprises two times four amino acids (CHWE and KLVC) on opposite
sides of a b-hairpin. The complete 14mer b-hairpin peptides 15–19 of this
study (C1HXE4T5GKHGH10K11LVC14) are extended by the b-turn-region
T48GKHGH62 (blue) from the eif5a-hypusin loop structure.[30, 31,37–39,53] The aro-
matic amino acid Xaa3 within the antiparallel b-sheet was mutated. The Hop
amino acid was inserted into the peptide as enantiopure amino acid 11 to
give the l-Hop peptide 16, as well as the racemic 10. Diastereomers were
obtained from the synthesis with the racemic amino acid, which were sepa-
rated on preparative HPLC to give the d-Hop peptide 17.
Table 1. Sequences of the synthesized peptides (Dpa=l-b,b-diphenyl
alanine, Nal=l-1-naphthyl alanine, dsb=disulfide bridged).
Peptide
Sequence
Folded fraction[a]
[%]
14
15
16
17
18
19
H-CHTrpETGKHGHKLVC-OH
9
88
31
18
17
100[b]
[dsb1,14]H-CHTrpETGKHGHKLVC-OH
[dsb1,14]H-CHHopETGKHGHKLVC-OH
[dsb1,14]H-CHhopETGKHGHKLVC-OH
[dsb1,14]H-CHDpaETGKHGHKLVC-OH
[dsb1,14]H-CHNalETGKHGHKLVC-OH
[a] The percentages of the folded fraction are calculated using the Leu
d-CH3 random coil 1H NMR shift as unfolded reference[54] and the l-1-Nal
peptide 19 as completely folded reference. [b] Reference value.
Figure 4. NMR analysis of the synthetic peptides. The best dispersion is obtained with 1-l-Nal, the aromatic side chain of which makes a hydrophobic contact
with Leu12, thus stabilizing the secondary structure of the peptide. The reduced form of the hydrophobic cluster shows no dispersion in the amide and ali-
phatic region (see the Supporting Information).
&
&
Chem. Eur. J. 2017, 23, 1 – 8
www.chemeurj.org
4
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
fraction”. The chemical shift of the shielded Leu12 d-CH3 was
used as the calculating parameter. A random coil chemical
shift built the reference for the unfolded state, while the shift
of Leu d-CH3 of peptide 19 was set as the completely folded
reference value (Table 1).[54] The l-Hop peptide 16 evidently be-
longs to the unfolded peptides with a folded fraction of only
31%. This destabilization of monomer 16 increases the possi-
bility for trimerization, only if the trimer is folded around the
metal ion.
signal nor line broadening was observed. When 1,2-HOPO is
added to this solution of 16, the original spectrum of the pep-
tide is obtained again together with a white precipitate. The
trimeric 1,2-HOPO complex of Ga3+ is insoluble in water and
precipitates from the NMR solution. This proves that the ob-
served NMR shifts come from metal complexation and it also
clarifies that the metal complexation of the peptides is a rever-
sible process.
Conclusion
Metal-chelating peptides
We describe the synthesis of 5-(1-Hydroxy-pyridin-2(1H)-onyl-l-
alanine (Hop) 11 and its use as an Fmoc-protected amino acid
in peptide synthesis for the assembly of novel metallopeptides.
The herein applied enzymatic hydrolysis of a racemic b-aromat-
ic a-amino acid methyl ester to yield the enantiopure l-config-
ured carboxylic acid is a fast and efficient way to obtain
unique non-natural a-amino acids ready for peptide synthesis
and can possibly be adapted to a broad substrate scope, open-
ing doors to novel functionalized amino acids in metallopep-
tide chemistry. The resulting HOPO-functionalized amino acid
11 is a redox stable analogue of the clinically relevant DOPA
and was successfully introduced into a b-hairpin epitope motif,
in which it destabilized the secondary structure of the peptide
16. Upon treatment with Ga3+, the folded fraction of the
HOPO peptide 16 was significantly increased and by the addi-
tion of a competitive metal chelator the reversibility of this
chelation process was proven. The trimeric b-hairpins in the
peptide complex are in intermediate exchange on the NMR
time scale and show significant line-broadening from the dif-
ferent diastereomeric arrangements of partially folded b-hair-
pins around the metal center. The incorporation of the metal-
binding amino acid 11, without significant destabilization of
the peptide shape, is promoted here as an alternative to estab-
lished methods of epitope oligomerization.[55] The herein used
replacement of Fe3+ with the diamagnetic Ga3+ can further be
used as a model system to study the role of Fe3+ in various
bioinorganic processes on NMR scale, providing new insights
in biomedical research. For further applications, peptides con-
taining the HOPO amino acid 11 could be used for the devel-
opment of peptide-based MRI/PET contrast agents or in pep-
tide catalysis.
The l-Hop-containing peptide 16 was treated with one equiva-
lent of Ga(NO3)3 in aqueous phosphate buffer/D2O. Similar to
the HOPO titration experiment in [D6]DMSO (Figure 2), the aro-
matic signals of the pyridone are significantly downfield shift-
ed. The very evident upfield shift of the aliphatic region that is
visible in Figure 5 is a clear indication of the refolding of the b-
hairpin in the trimeric complex. The additional line broadening
is explained by the different diastereomeric arrangements of
the three b-hairpins around the metal center. The chemical
shift value of 0.04 ppm for Leu12-d-CH3 yields a folded fraction
of 64% for the trimer. To exclude secondary effects from other
side chain functionalities in the peptide, for example, metal
complexation by His side chains, the l-Trp3 peptide 15 was ti-
trated in the same manner and neither significant shifts of any
Experimental Section
General information
If not mentioned otherwise, solvents were distilled by rotary evap-
oration. For reactions under inert gas atmosphere, solvents were
dried under argon atmosphere as follows: DCM was distilled from
anhydrous CaH2, THF was dried with KOH and distilled from
sodium/benzophenone, and toluene was distilled from sodium.
The reaction procedure was monitored by thin layer chromatogra-
phy (TLC) using Merck Silica Gel 60 F245 plates and visualized by UV
light or staining with ninhydrin (200 mg ninhydrin, 0.5 mL acetic
acid, 4.5 mL water, 100 mL n-butanol) or vanillin (6.00 g vanillin,
1.5 mL conc. sulfuric acid, 95 mL EtOH). Flash chromatography was
performed using Merck 60 M (0.040–0.063 mm) silica gel as station-
Figure 5. Ga3+ complexation with the l-Hop peptide 16. As a third equiva-
lent of metal salt is added, a clear chemical shift of the aliphatic signals is
visible, due to the formation of the 3:1 complex. Line broadening occurs be-
cause of the formation of several diastereomers around the metal center.
When an excess of a better chelator (in this case 1,2-HOPO) is added, the
original spectra (without Ga3+ addition) is obtained and a white solid precip-
itates from the NMR solution. The white precipitate was analyzed and clearly
characterized as the Ga(HOPO)3 complex 13.
Chem. Eur. J. 2017, 23, 1 – 8
www.chemeurj.org
5
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
ary phase. NMR analysis of organic molecules was performed on
a Bruker Avance 300, Bruker Avance DRX 500, and Bruker Avance
500 spectrometer at 300 K. Chemical shifts are reported in ppm
with the solvents residue as internal standard. Peptides were ana-
lyzed on a Bruker Avance AVII 600 MHz spectrometer using Water-
gate-pulse sequence. NOESY spectra were obtained with a mixing
time of 300 ms. 50 mm potassium phosphate buffer pH 3.0/D2O
(9:1) was used as solvent with a trace of 3-(trimethylsilyl)propionic-
2,2,3,3-d4 acid sodium salt as internal standard for 1H (d=
0.00 ppm). ESI mass spec analyses were performed on a Thermo
Fisher Scientific LTQ-FT spectrometer by the department of mass
analysis at Philipps-University Marburg. The resolution was set to
100.000.
moved under reduced pressure and the peptide was precipitated
from cold diethyl ether, washed, and lyophilized. Crude peptides
were dissolved in (NH4)2CO3 buffer (50 mL, pH 8.9) and the reaction
progress was monitored by RP-HPLC. If necessary, the pH of the so-
lution was adjusted to 8.9 by addition of solid (NH4)2CO3. After
complete conversion, the solution was lyophilized and the peptide
was purified by RP-HPLC.
Acknowledgements
We thank the Jꢂrgen Manchot Stiftung, Dꢂsseldorf, for a PhD
fellowship.
Enzymatic hydrolysis of 7
Conflict of interest
Cbz-dl-Hop(OtBu)-OMe (7, 0.80 g, 2.00 mmol, 1.0 equiv) was dis-
solved in 1,4-dioxane (8.0 mL) and phosphate buffer (42 mL 0.4m,
pH 7) and a solution of protease from bacillus licheniformis (1.4 mL,
min. 2.4 UgÀ1, EC 3.4.21.62, Sigma–Aldrich) were added. The solu-
tion was stirred at room temperature for 2 h (consumption of the
educt was monitored by RP-HPLC). The mixture was extracted with
diethyl ether (200 mL) and the layers were separated. The aqueous
layer was acidified to pH 1 with HCl (1m) and extracted with ethyl
acetate (200 mL) three times. The organic layers were combined,
dried over anhydrous sodium sulfate, and the solvent was removed
under reduced pressure. 380 mg (0.97 mmol, 97% relative to l-
isomer) of the enantiopure Cbz-l-Hop(tBu)-OH 8 were obtained as
a white solid.
1H-NMR (300 MHz, [D6]DMSO): d=7.45 (d, 1H, 4J=1.7 Hz, H-6),
7.39–7.19 (m, 6H, Ph+H-4), 6.37 (d, 1H, 4J=8.2 Hz, NH), 6.37 (d,
1H, 3J=8.6 Hz, H-3), 4.96 (m, 2H, Ph-CH2), 3.95 (ddd, 1H, 3J=
9.1 Hz. 3J=8.2 Hz, 3J=4.3 Hz, a-H), 2.88 (dd, 2J=13.9 Hz, 3J=
4.3 Hz, b-H), 2.60 (dd, 2J=14.0 Hz, 3J=9.1 Hz, b-H), 1.27 ppm (s,
9H, C(CH3)3); 13C-NMR (75 MHz, [D6]DMSO): d=173.1 (CO2H), 158.6
(C-2), 156.0 (Cbz-Cq), 140.3 (C-4), 138.7 (C-6), 136.9 (Ph-Cq), 129.7
(Ph), 128.2 (Ph), 127.5 (Ph), 120.8 (C-3), 114.6 (C-5), 86.1 (C(CH3)3),
67.4 (Ph-CH2), 55.8 (a-C), 34.2 (b-C), 27.8 ppm (C(CH3)3); HRMS
(ESI+, DCM): m/z calcd for C20H23N2O6: 387.1562 [M+H+]; found:
387.1567.
The authors declare no conflict of interest.
Keywords: hydroxypyridones
spectroscopy · oligomerization · peptides
·
metal chelation
·
NMR
[1] M. Isaka, M. Tanticharoen, P. Kongsaeree, Y. Thebtaranonth, J. Org.
Chem. 2001, 66, 4803–4808.
[2] P. M. J. Clement, H. M. Hanauske-Abel, E. C. Wolff, H. K. Kleinman, M. H.
Park, Int. J. Cancer 2002, 100, 491–498.
[3] B. B. Abrams, H. Hꢃnel, T. Hoehler, Clin. Dermatol. 1991, 9, 471–477.
[4] C. De Cola, G. Fiorillo, A. Meli, S. Aime, E. Gianolio, I. Izzo, F. De Riccardis,
Org. Biomol. Chem. 2014, 12, 424–431.
[5] K. N. Raymond, V. C. Pierre, Bioconjugate Chem. 2005, 16, 3–8.
[6] A. Datta, K. N. Raymond, Acc. Chem. Res. 2009, 42, 938–947.
[7] J. J. J. Yang, L. Wei, O. Zurkiya, W. Yang, S. Li, J. Zou, Y. Zhou, A. L. W.
Maniccia, H. Mao, F. Zhao, J. Am. Chem. Soc. 2008, 130, 9260–9267.
[8] P. D. Garimella, A. Datta, D. W. Romanini, K. N. Raymond, M. B. Francis, J.
Am. Chem. Soc. 2011, 133, 14704–14709.
[9] S. Y. Han, Y. A. Kim, Tetrahedron 2004, 60, 2447–2467.
[10] S. J. Zuend, M. P. Coughlin, M. P. Lalonde, E. N. Jacobsen, Nature 2009,
461, 968–970.
[11] P. Etayo, A. Vidal-Ferran, Chem. Soc. Rev. 2013, 42, 728–754.
[12] C. Nꢄjera, J. M. Sansano, Chem. Rev. 2007, 107, 4584–4671.
[13] D. A. Evans, T. C. Britton, J. A. Ellman, R. L. Dorow, J. Am. Chem. Soc.
1990, 112, 4011–4030.
Peptide synthesis
[14] R. Dharanipragada, E. Nicolas, G. Toth, V. J. Hruby, Tetrahedron Lett.
1989, 30, 6841–6844.
2-CTC resin (Iris Biotech, 1.63 mmolgÀ1) was loaded with Fmoc-l-
Cys(Trt)-OH (2.0 equiv) and DIPEA (6.0 equiv) in DMF for 4 h. The
resin was washed with DMF twice, further treated with a mixture
of DCM/MeOH/DIPEA (80:15:5) for 30 min twice, and afterwards
washed several times with DMF, DCM, and MeOH. The loaded resin
was dried under vacuum. The loading was estimated to
0.50 mmolgÀ1 through monitoring the UV/Vis absorption of fluo-
rene at 289 nm and 300 nm after cleavage of the Fmoc protecting
group. The peptides were synthesized on a microwave-equipped
CEM Liberty Blue automated peptide synthesizer by Fmoc strategy
under standard conditions (5 equiv Fmoc-AA-OH, 5 equiv DIC,
5 equiv Oxyma) at 508C for 10 min. Fmoc deprotection was per-
formed with 20% piperidine in DMF (0.1% HOBt) for 2.5 min at
508C. The peptides were cleaved from the resin with a solution
containing TFA/phenol/TIPS/H2O 88:5:5:2 for 3 h. Peptides were
precipitated from 40 mL of cold diethyl ether, washed two times
with ether, and lyophilized in water with a minimum amount of
acetonitrile. Crude NMR analysis indicated incomplete cleavage of
the tBu protecting group from the pyridone when the peptide was
cleaved from the resin. Therefore, the crude precipitated peptide
was dissolved in 95% TFA and stirred overnight. The acid was re-
[15] W. H. Chiu, C. Depboylu, G. Hermanns, L. Maurer, A. Windolph, W. H.
Oertel, V. Ries, G. U. Hçglinger, Neuropharmacology 2015, 95, 367–376.
[16] S. Mainer, J. M. Davis, Bioinorg. Chem. 1978, 9, 187–203.
[17] X. Kꢃstele, P. Klꢂfers, F. Kopp, J. Schuhmacher, M. Vogt, Chem. Eur. J.
2005, 11, 6326–6346.
[18] R. A. Yokel, A. M. Fredenburg, P. W. Durbin, J. Xu, M. K. Rayens, K. N. Ray-
mond, J. Pharm. Sci. 2000, 89, 545–555.
[19] D. Kalinowski, D. Richardson, Pharmacol. Rev. 2005, 57, 547–583.
[20] S. Chillappagari, C. Mꢂller, P. Mahavadi, A. Guenther, L. Nꢃhrlich, J. Rose-
nblum, B. K. Rubin, M. O. Henke, J. Cystic Fibrosis 2016, 15, 325–331.
[21] J. I. Lachowicz, V. M. Nurchi, G. Crisponi, M. G. Jaraquemada-Pelaez, M.
Arca, A. Pintus, M. A. Santos, C. Quintanova, L. Gano, Z. Szewczuk,
Dalton Trans. 2016, 45, 6517–6528.
[22] W. W. Grody, R. J. Desnick, Genet. Med. 2001, 3, 87–90.
[23] B. Imperiali, T. J. Prins, S. L. Fisher, J. Org. Chem. 1993, 58, 1613–1616.
[24] H. S. Lee, G. Spraggon, P. G. Schultz, F. Wang, J. Am. Chem. Soc. 2009,
131, 2481–2483.
[25] B. Cuenoud, A. Y. Schepartz, Tetrahedron 1991, 14/15, 2535–2542.
[26] L. M. De Leꢅn-Rodriguez, Z. Kovacs, G. R. Dieckmann, A. D. Sherry,
Chem. Eur. J. 2004, 10, 1149–1155.
[27] M. S. Cooper, E. Sabbah, S. J. Mather, Nat. Protoc. 2006, 1, 314–317.
[28] M. Hoque, H. M. Hanauske-Abel, P. Palumbo, D. Saxena, D. D’Alliessi
Gandolfi, M. H. Park, T. Pe’ery, M. B. Mathews, Retrovirology 2009, 6, 90.
&
&
Chem. Eur. J. 2017, 23, 1 – 8
www.chemeurj.org
6
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
[29] D. Saxena, M. Spino, F. Tricta, J. Connelly, B. M. Cracchiolo, A.-R. Ha-
nauske, D. D’Alliessi Gandolfi, M. B. Mathews, J. Karn, B. Holland, PLoS
One 2016, 11, e0154842.
[44] P. J. Coleman, K. M. Brashear, B. C. Askew, J. H. Hutchinson, C. A.
McVean, L. T. Duong, B. P. Feuston, C. Fernandez-Metzler, M. A. Gentile,
G. D. Hartman, J. Med. Chem. 2004, 47, 4829–4837.
[30] H. M. Hanauske-Abel, D. Saxena, P. E. Palumbo, A. R. Hanauske, A. D. Lu-
chessi, T. D. Cambiaghi, M. Hoque, M. Spino, D. D. A. Gandolfi, D. S.
Heller, PLoS One 2013, 8, e74414.
[31] A. Abbruzzese, H. M. Hanauske-Abel, M. H. Park, S. Henke, J. E. Folk, Bio-
chim. Biophys. Acta Protein Struct. Mol. Enzymol. 1991, 1077, 159–166.
[32] P. Saini, D. E. Eyler, R. Green, T. E. Dever, Nature 2009, 459, 118–121.
[33] E. Gutierrez, B. S. Shin, C. J. Woolstenhulme, J. R. Kim, P. Saini, A. R. Bus-
kirk, T. E. Dever, Mol. Cell 2013, 51, 35–45.
[45] L. Struk, J. G. Sosnicki, Synthesis 2012, 735–746.
[46] K. Yazawa, K. Numata, Molecules 2014, 19, 13755–13774.
[47] M. R. Blomberg, T. Borowski, F. Himo, R.-Z. Liao, P. E. M. Siegbahn, Chem.
Rev. 2014, 114, 3601–3658.
[48] T. Nuijens, C. Cusan, J. A. W. Kruijtzer, D. T. S. Rijkers, R. M. J. Liskamp,
P. J. L. M. Quaedflieg, Synthesis 2009, 809–814.
[49] P. Vossenberg, R. Beeftink, M. C. Stuart, H. Tramper, Biotechnol. Prog.
2013, 29, 870–875.
[34] Y. Seko, T. Fujimura, T. Yao, H. Taka, R. Mineki, K. Okumura, K. Murayama,
Sci. Rep. 2015, 5, 13737.
[50] P. Vossenberg, H. H. Beeftink, T. Nuijens, M. A. Cohen Stuart, J. Tramper,
J. Mol. Catal. B 2012, 75, 43–49.
[35] C. M. Costa-Neto, L. T. Parreiras-e-Silva, R. Ruller, E. B. Oliveira, A. Miran-
da, L. Oliveira, R. J. Ward, Biochem. Biophys. Res. Commun. 2006, 347,
634–640.
[36] A. Malyutin, S. Musalgaonkar, S. Patchett, J. Frank, A. W. Johnson, The
EMBO Journal 2017, 36, 854–868.
[37] T. Blandl, A. G. Cochran, N. J. Skelton, Protein Sci. 2003, 12, 237–247.
[38] S. J. Russell, A. G. Cochran, J. Am. Chem. Soc. 2000, 122, 12600–12601.
[39] S. J. Russell, T. Blandl, N. J. Skelton, A. G. Cochran, J. Am. Chem. Soc.
2003, 125, 388–395.
[40] A. Schrimpf, A. Geyer, ChemBioChem 2016, 17, 2129–2132.
[41] S. Fischer, A. Geyer, Angew. Chem. Int. Ed. 2014, 53, 3849–3853; Angew.
Chem. 2014, 126, 3928–3933.
[51] K. Drꢃger, Z. Naturforsch. A 1988, 43, 262–270.
[52] S. Fischer, M. Lamping, M. Gold, Y. Rçttger, D. Brçdje, R. Dodel, R.
Frantz, M. A. Mraheil, T. Chakraborty, A. Geyer, Bioorg. Med. Chem. 2016,
25, 603–608.
[53] Y. Bin Teng, X. X. Ma, Y. X. He, Y. L. Jiang, J. Du, C. Xiang, Y. Chen, C. Z.
Zhou, Proteins Struct. Funct. Bioinf. 2009, 77, 736–740.
[54] D. S. Wishart, C. G. Bigam, A. Holm, R. S. Hodges, B. D. Sykes, J. Biomol.
NMR 1995, 5, 67–81.
[55] A. M. Roeder, Y. Roettger, A. Stꢂndel, R. Dodel, A. Geyer, J. Biol. Chem.
2013, 288, 27638–27645.
[42] B. M. Schiavi, D. J. Richard, M. M. Joulliꢆ, J. Org. Chem. 2002, 67, 620–
624.
[43] M. Berwe, W. Jçntgen, J. Krꢂger, Y. Cancho-Grande, T. Lampe, M. Mi-
chels, H. Paulsen, S. Raddatz, S. Weigand, Org. Process Res. Dev. 2011,
15, 1348–1357.
Manuscript received: February 14, 2017
Accepted Article published: April 3, 2017
Final Article published: && &&, 0000
Chem. Eur. J. 2017, 23, 1 – 8
www.chemeurj.org
7
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
Reversible peptide folding: 5-(1-Hy-
droxy-pyridin-2(1H)-onyl)-l-alanine
(Hop), a redox stable N-hydroxy-1,2-pyri-
done-functionalized a-amino acid with
the desired metal-chelating properties
of DOPA (3,4-dihydroxy phenylalanine),
was synthesized in eight steps with
96% ee and used as Fmoc-protected
amino acid for the assembly of a 14mer
b-hairpin peptide. The secondary struc-
ture of the peptide, which was destabi-
lized by the introduction l-Hop, could
be reversibly reconstituted by subse-
quent additions of Ga3+ and a competi-
tive metal chelator.
&
Amino Acids
J. Reutzel, T. M. Diogo, A. Geyer*
&& – &&
Reversible Folding of a b-Hairpin
Peptide by a Metal-Chelating Amino
Acid
&
&
Chem. Eur. J. 2017, 23, 1 – 8
www.chemeurj.org
8
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!