Exploring hydrophobicity limits of polyproline helix with oligomeric octahydroindole-2-carboxylic acid

Article abstract of DOI:10.1002/psc.3076

The polyproline-II helix is the most extended naturally occurring helical structure and is widely present in polar, exposed stretches and “unstructured” denatured regions of polypeptides. Can it be hydrophobic? In this study, we address this question usin

Full text of DOI:10.1002/psc.3076

Received: 16 January 2018
DOI: 10.1002/psc.3076
Revised: 24 February 2018
Accepted: 26 February 2018
R E S E A R C H A R T I C L E
Exploring hydrophobicity limits of polyproline helix with
oligomeric octahydroindole‐2‐carboxylic acid
|
Vladimir Kubyshkin
Nediljko Budisa
Institute of Chemistry, Technical University of
Berlin, Müller‐Breslau‐Str. 10, Berlin 10623,
Germany
The polyproline‐II helix is the most extended naturally occurring helical structure and
is widely present in polar, exposed stretches and “unstructured” denatured regions of
polypeptides. Can it be hydrophobic? In this study, we address this question using
oligomeric peptides formed by a hydrophobic proline analogue, (2S,3aS,7aS)‐
octahydroindole‐2‐carboxylic acid (Oic). Previously, we found the molecular principles
underlying the structural stability of the polyproline‐II conformation in these oligo-
mers, whereas the hydrophobicity of the peptide constructs remains to be examined.
Therefore, we investigated the octan‐1‐ol/water partitioning and inclusion in
detergent micelles of the oligo‐Oic peptides. The results showed that the hydropho-
bicity is remarkably enhanced in longer oligomeric sequences, and the oligo‐Oic
peptides with 3 to 4 residues and higher are specific towards hydrophobic environ-
ments. This contrasts significantly to the parent oligoproline peptides, which were
moderately hydrophilic. With these findings, we have demonstrated that the
polyproline‐II structure is compatible with nonpolar media, whereas additional
manipulations of the terminal functionalities feature solubility in extremely nonpolar
solvents such as hexane.
Correspondence
Vladimir Kubyshkin and Nediljko Budisa,
Institute of Chemistry, Technical University of
Berlin, Müller‐Breslau‐Str. 10, Berlin, 10623,
Germany.
Email: kubyshkin@win.tu‐berlin.de; nediljko.
budisa@tu‐berlin.de
Funding information
Deutsche Forschungsgemeinschaft, Grant/
Award Number: FOR 1805
KEYWORDS
hydrophobicity, lipophilicity, micelles, polyproline helix, polyproline‐II, proline
|
1
INTRODUCTION
substructures such as the transmembrane domains of membrane pro-
teins and others relying on hydrophobic motifs.
For these reasons, the capacity of peptides to function as molecu-
lar spacers is frequently problematic in the context of nonpolar media.
|
1.1
Hydrophobic peptides
Proteins utilize various secondary structures for spatial organization
and environmental interaction. The elements of these secondary struc-
tures are stabilized by a set of noncovalent interactions, such as hydro-
gen bonding between the backbones of secondary amides. In general,
polypeptides are optimal for exposure in aqueous media; however,
when present in more hydrophobic media such as membranes, other
stabilizing factors are required. For example, in biological settings, pro-
tein substructures often require chaperones for correct folding and
sophisticated systems for membrane translocation.1,2 Arguably, the
correct folding of hydrophobic segments in proteins is complicated
by the ambivalence of the structure containing both lipophilic side
chains and highly hydrophilic amide backbone. Not surprisingly, in bio-
logical systems, aggregation and misfolding are common for peptide
Nonetheless, such spacers do occur in nature, and some have been
designed in the laboratory. A remarkable example is the WALP and
KALP peptides designed for the biophysical studies of transmembrane
α‐helices.3 These molecular rulers consist of repetitive alanine‐leucine
sequences flanked by tryptophans (WALP) or lysines (KALP) for
correct membrane anchoring of the termini. However, the resulting
constructs are still not free from solubility and aggregation problems.
A potential solution to these issues might be offered by the use of
strongly lipophilic and bulkier side chains. Particularly promising are sil-
icon‐containing amino acids,4 which may additionally stabilize certain
secondary structures.5 In our opinion, the most interesting case
reported to date is the “lipophilic polyproline‐II structure” constructed
by Martin et al.6 In particular, NMR investigations of an oligomeric
J Pep Sci. 2018;e3076.
wileyonlinelibrary.com/journal/psc
Copyright © 2018 European Peptide Society and John Wiley & Sons, Ltd.
1 of 14
https://doi.org/10.1002/psc.3076

2 of 14
KUBYSHKIN AND BUDISA
measurements30-32 and for a variety of other studies based on molec-
ular distance control.33
Polymeric and oligomeric proline sequences serve as good models
for studying features of PII conformation, since the tertiary amide bond
formed by this amino acid opposes formation of regular hydrogen
bond–based structure. Therefore, polymeric proline adopts a relatively
stable all‐trans‐amide PII helix in polar media. Importantly, it is known
that hydrophobic media at times promote alternative structures, such
as polyproline‐I helix (in propan‐1‐ol), which is a more compact helix
formed by an all‐cis‐amide backbone.34 Alternatively, isomerization of
PII helix can proceed into a nonclassical backbone‐side chain hydro-
gen‐bonded β‐sheet‐like assemblies (in trifluoroethanol). The latter
were documented only very recently for peptides formed by specific
proline analogues.35-37
FIGURE 1 Structure of proline analogues forming the lipophilic PII
helix: (2S)‐silaproline (1) and (2S,3aS,7aS)‐octahydroindole‐2‐
carboxylic acid (2)
proline analogue called silaproline (Sip, 1, Figure 1) revealed that the
solution structure is dominated by the polyproline‐II (PII) helix.7 The
latter is a unique natural secondary structure that is not stabilized by
hydrogen bonds.
|
1.3
Hydrophobic polyproline helix
In general, elucidation of potential hydrophobicity limits for the PII
helix could provide a new perspective on the evolution of the amino
acid repertoire for the genetic code and offer an avenue for production
of new‐to‐nature peptide scaffolds and materials. Therefore, research
on this topic covers both fundamental and practical areas of peptide
science. For example, oligomeric silaproline sequences can potentially
be considered very lipophilic PII structures. This amino acid was intro-
duced independently by 2 groups about 2 decades ago.38,39 However,
despite the growing interest in this amino acid, studies are hampered
by their somewhat limited availability.40,41
|
1.2
Polyproline helix
PII helix is the third most abundant secondary structure element (after
α‐helices and β‐sheets) and the most extended helix in natural
proteomes.8 The PII helix is regarded as semiextended, and it features
the same linear increment per 1 amino acid as partially extended
β‐sheet (both 3.0 Å per amino acid). This contrasts to rather compact
α‐helix (1.5 Å per amino acid) and fully extended 2.05‐helix (approxi-
mately 4 Å per amino acid), while the latter is stabilized only in some
specific structural contexts.9
The name “polyproline” is a historical rudiment of the discovery of
this structure for polymeric proline,10 whereas numerical “II” indicates
that this is the second structure emerging upon mutarotation of poly-
meric proline in polar media.11 It is a generic structure allowed for all
coded amino acids but usually favored by polar amino acids,12-14 espe-
cially in the context of the charge repulsion in oligomeric charged
sequences such as oligolysine.15,16 It is also favored in solvent‐exposed
protein parts, fragments lacking a persistent structure, and the dena-
tured state.17,18 Regarding the evolution of the genetic code, recent
considerations suggest that PII helical folding may have been the dom-
inant structure early in the development of biological peptide synthe-
sis, prior to the α‐helix and much earlier than the hydrophobic motif
was recruited by nature.19 This follows the GC‐GCA‐GCAU scheme
of the genetic code evolution as recently summarized by Hartman
and Smith.20,21 Owing to its semiextended nature, the PII protein
domains are predictably highly solvent exposed in aqueous media.22
Although amphipathic PII‐forming sequences do appear in nature,23,24
there is no evidence for the existence of a natural PII helix that would
be purely hydrophobic.
Alternatively, the PII structure can be made hydrophobic by using
oligomers of the proline analogue (2S,3aS,7aS)‐octahydroindole‐2‐car-
boxylic acid (Oic, 2, Figure 1). This amino acid is used in industry as a
structural component of an antihypertensive drug perindopril, thus
making Oic one of the cheapest proline analogues available. Using this
commercially available compound, we previously demonstrated that
oligomeric Oic sequences can induce remarkable stability in the PII
helical fold.29
The stability of the oligo‐Oic PII structure is based on a molecular
interlock mechanism, which propagates as follows: (1) the preceding
amino acid residue prevents the cyclohexane ring from inversion; (2)
the cyclohexane ring becomes fixed in the chair conformation, which
freezes the fused pyrrolidine ring in the exo‐pucker conformation; (3)
the distance and angle between the amide carbonyl groups are optimal
for stabilizing interactions; and (4) the trans‐amide conformation is sta-
bilized (return to no. 1). As a result, the trans‐amide bond and the chair
cyclohexane conformations stabilize (lock) one another (for an illustra-
tion, see also the crystal structure in Figure 4). The stability of the
trans‐amide bond leads to the dominance of the PII helical fold already
at a context of 3 residues, which was observed in our analysis of the 1H
NMR spectra. It is thus possible to construct relatively rigid oligo‐Oic
peptides for further consideration as peptide tags in de novo protein
design or development of peptide materials. However, experimental
data involving the relevant interactions of these scaffolds in hydropho-
bic surroundings are lacking. While it is clear that the chemical formula
indicates a lipophilic structure, it is not clear how lipophilic it is,
whether the PII structure is maintained in nonpolar media; thus what
are the hydrophobic limits of a PII helix? To address these questions,
The most prominent chemical feature of the PII helix is a special
type of fold that does not rely on hydrogen bonding. In contrast to
most conventional peptide structures, the local order in the PII helix
is a product of generic interactions between polar amide groups. The
latter may be quite complex, as these may be described in terms of
n→π* donation,25 dipole‐dipole interaction,26 carbonyl‐carbonyl inter-
action,27 and the cascade effect.28,29 The geometric simplicity of this
secondary structure (1 turn = 3.0 residues ≈ 9.0 Å) is also well appreci-
ated and has been used as a molecular ruler for spectroscopic distance

KUBYSHKIN AND BUDISA
3 of 14
we present here a model study of a series of Oic oligomers. Our inten-
tion is also to articulate the principles of the design of a lipophilic PII
structure to allow the use of this conformation for distance control in
extremely hydrophobic environments.
1‐fluorenylmethoxycarbonyl (Fmoc) chloride (8.04 g; 31 mmol;
1.05 equiv.) in dichloromethane (20 mL) was added portion‐wise over
1 minute. The reaction mixture was stirred for 3 hours under cooling.
The solvent was removed under reduced pressure (temperature
approximately 303 K). Five percent aqueous sodium hydrogen carbon-
ate (250 mL), acetone (approximately 100 mL), and water were added
to the residue (net volume approximately 800 mL) under stirring. The
resulting solution was washed with diethyl ether (3 × 150 mL) and then
acidified by the addition of potassium hydrogen sulfate (35 g). The
aqueous phase was extracted with ethyl acetate (5 × 125 mL), and
the organic fractions were dried over sodium sulfate, filtered, and
evaporated to give oilic material, which was then lyophilized from an
acetonitrile‐water mixture to give the final product as a white powder
(8.50 g; 22 mmol; yield 73%). Note that the yield is reduced owing to
the diethyl ether washing step. 1H NMR (CD3OD, 700 MHz), δ: 7.81
(m, 2H), 7.68 to 7.60 (m, 2H), 7.40 (m, 2H), 7.33 (m, 2H), 4.50 (m,
1H), 4.35 (m, 1H), 4.27 (m, 1H), 4.16 (m, 1H), 3.89 and 3.42 (2 m,
1H), 2.38 to 1.20 and 1.01 (series of multiplets, 11H).
|
2
MATERIALS AND METHODS
Synthesis
|
2.1
Methyl esters of the N‐acetylated amino acids were obtained as
described.42
|
2.1.1
Methyl (2S,3aS,7aS)‐1‐
pivaloyloctahydroindole‐2‐carboxylate
Oic (105 mg; 0.62 mmol) was mixed with methanol (2 mL) and
chlorotrimethylsilane (0.25 mL). The mixture was stirred at room tem-
perature for 15 hours. The methanol was then removed under reduced
pressure. The resulting oilic compound was the methyl ester of the
amino acid hydrochloride. 1H NMR (CD3OD, 500 MHz), δ: 4.47 (br
m, 1H, α‐CH), 4.20 (m, 1H, δ‐CH), 2.38 (m, 1H, γ‐CH), 2.16, 2.03,
1.98, 1.81, 1.75, 1.73, 1.64, 1.53, 1.40, 1.27 (10 m, 1H each,
5 × CH2), and 1.38 (s, 9H, (CH3)3C). This compound was dissolved in
dichloromethane (2.5 mL), triethylamine (0.27 mL; 1.9 mmol;
3.1 equiv.), and pivaloyl chloride (0.085 mL; 0.69 mmol; 1.1 equiv.),
and the mixture was stirred at room temperature for 2 hours. It was
then washed with water (1 × 1 mL), 1:2 hydrochloric acid (1 × 1 mL),
and 1M sodium hydrogen carbonate (1 × 1 mL). Next, the solution
was dried over sodium sulfate, filtered, and evaporated. Crude material
was purified on a short silica gel column (approximately 10 g) using
hexane‐ethyl acetate mixture (2:1) as an eluent (Rf = 0.44). The product
was obtained as a clear oil, which later crystallized (142 mg; 0.53 mmol;
yield 85%). 1H NMR (D2O, 700 MHz), δ: 4.40 (t, J = 9.4 Hz, 1H, α‐CH),
4.20 (m, 1H, δ‐CH), 3.68 (s, 3H, CH3O), 2.35 (m, 1H, γ‐CH), 2.06 and
1.95 (2 m, 1H each, CH2), 1.92 and 1.44 (2 m, 1H each, CH2), 1.69
and 1.63 (2 m, 1H each, CH2), 1.62 and 1.16 (2 m, 1H each, CH2),
1.40 and 1.25 (2 m, 1H each, CH2), 1.18 (s, 9H, (CH3)3C). 13C{1H}
NMR (D2O, 176 MHz), δ: 180.0 (CO2Me), 175.7 (N–C═O), 61.3
(α‐CH), 59.2 (δ‐CH2), 52.7 (CH3O), 39.2 (C), 37.9 (γ‐CH), 27.9 (CH2),
27.5 (CH2), 27.3 (3 × CH3), 25.0 (CH2), 23.4 (CH2), 19.4 (CH2). HRMS
(ESI Orbitrap): calcd. for [M + H]+ C15H26NO3 268.1907 Da, found
268.1907 Th. [α]D = −62 (c = 1.5, methanol, 298 K). m.p. 340 K. The
crystals produced after silica gel purification were subjected to X‐ray
diffraction analysis. The crystal structure has been deposited in the
Cambridge Crystallographic Data Centre under the deposition ID
CCDC 1585218.
|
2.1.3
2,2‐Difluoroethyl (2S,3aS,7aS)‐1‐
acetyloctahydroindole‐2‐carboxylate
Oic (149 mg; 0.88 mmol) was mixed with excess of acetic anhydride
(0.3 mL; 3.2 mmol; 3.6 equiv.) in dichloromethane (3 mL). The mixture
becomes clear 15 minutes after the addition of the anhydride, and this
was additionally stirred for the next 2 hours. Solvent was removed
under reduced pressure, water (3 mL) was added in order to quench
residual anhydride, and the mixture was freeze dried. 2,2‐
Difluoroethan‐1‐amine (115 mg; 1.4 mmol; 1.6 equiv.) was mixed with
tert‐butyl nitrite (90%; 190 mg; 1.7 mmol; 1.9 equiv.) in chloroform
(5 mL). This mixture was stirred for 10 minutes during which it was
brought to reflux once. The N‐acetylated amino acid in chloroform
(2 mL) was added to this mixture, and the resulting mixture was stirred
for 12 hours at room temperature. Volatiles were blown off with a
nitrogen gas current, and the resulting crude mixture was purified over
a silica gel column (20 g) using a hexane‐ethyl acetate mixture (1:1) as
an eluent (Rf = 0.37). The product was obtained as a clear oil (217 mg;
0.79 mmol; yield 89%). 1H NMR (CD3OD, 500 MHz), δ, 2 rotamers
(major, s‐trans): 6.14 (tt, JHF = 54 Hz, JHH = 3.3 Hz, minor) and 6.08
(tt, JHF = 55 Hz, JHH = 3.7 Hz, major, 1H, CHF2), 4.70 (t, J = 8.7 Hz,
minor) and 4.43 (dd, J = 9.8 and 8.4 Hz, major, 1H, α‐CH), 4.48 (tt,
JHF = 15 Hz, JHH = 3.2 Hz, minor) and 4.36 (tdd, JHF = 14 Hz, JHH = 3.7
and 2.5 Hz, major, 2H, OCH2), 4.17 (m, minor) and 3.94 (m, major, 1H,
δ‐CH), 2.50 (m, major) and 2.35 (m, minor, 1H, γ‐CH), 2.23, 2.07, 2.01,
1.82, 1.77, 1.76, 1.60, 1.55, 1.38, 1.29 (10 multiplets, all major, 1H
each, 5 × CH2), 2.13 (s, major), and 1.92 (s, minor, 3H, CH3). 13C{1H}
NMR (CD3OD, 126 MHz), δ, 2 rotamers (major, s‐trans): 172.0 (s,
minor CO2), 171.6 (s, major CO2), 170.8 (s, minor N–C═O), 170.2 (s,
major N–C═O), 113.1 (t, JCF = 240 Hz, major CHF2), 113.0 (t,
JCF = 239 Hz, minor CHF2), 62.6 (t, JCF = 27 Hz, minor OCH2), 62.5
(t, JCF = 28 Hz, major OCH2), 59.4 (s, minor α‐CH), 58.9 (s, major
δ‐CH), 58.7 (s, major α‐CH), 57.6 (s, minor δ‐CH), 37.4 (s, major γ‐
CH), 35.7 (s, minor γ‐CH), 32.6, 25.9, 25.0, 23.4, and 19.9 (5 s, minor
CH2), 30.3, 27.4, 25.2, 23.4, and 19.7 (5 s, major CH2), 20.2 (s, minor
CH3), and 19.9 (major CH3). 19F NMR (CD3OD, 471 MHz), δ, 2
rotamers (s‐trans/s‐cis 7.5:1): −127.52 and −128.21 (AB system: 2
|
2.1.2
1‐Fluorenylmethoxycarbonyl‐(2S,3aS,7aS)‐
octahydroindole‐2‐carboxylic acid
Oic (5.01 g; 30 mmol; 1 equiv.) was mixed with diisopropylethylamine
(DIPEA; 18 mL; 103 mmol; 3.4 equiv.) in dichloromethane (130 mL)
under a nitrogen atmosphere. The resulting mixture was cooled down
in an ice bath. Chlorotrimethylsilane (TMS chloride; 8.4 mL; 66 mmol;
2.2 equiv.) was added portion‐wise within 5 minutes. The resulting
clear mixture was stirred for additional 30 minutes. A solution of

4 of 14
KUBYSHKIN AND BUDISA
ddt, JFF = 295 Hz, JFH = 55 and 15 Hz, major CHF2), −128.4 (dt, JFH = 55
and 15 Hz, minor CHF2). HRMS (ESI Orbitrap): calcd. for [M + H]+
C13H20F2NO3 276.1406 Da, found 276.1399 Th. [α]D = −59 (c = 1.4,
methanol, 298 K).
25 milliseconds for exchange and 3 milliseconds for referencing. The
activation barriers were calculated assuming single transition state
using the Eyring equation.
|
2.3.2
Partitioning
Four to six milligrams of a substrate was mixed with octan‐1‐ol
(1.00 mL) and deionized water (1.00 mL), and this mixture was shaken
for 12 to 24 hours at 298 2 K. In the NMR tubes, 0.30 mL of each
phase was used, and these solutions were diluted by 0.30 mL of aceto-
nitrile‐d3. 1H, 19F, and inverse‐gated decoupled 19F{1H} NMR measure-
ments were performed at 298 K in datasets with identical starting
settings, using calibrated 90° pulses. Only zero‐order phases were
readjusted for reprocessing between equivalent water and octan‐1‐ol
phase spectra. Note that readjustment of the first‐order phase should
be avoided because it may influence the signal intensity, and this may
be indicative of imperfections of the tuning or 90° pulses. After appro-
priate baseline correction, equivalent signals were integrated, and
ratios of the signals were taken as the partitioning constants.
Alternatively, for complicated baseline situations, the spectra were
overlaid, and the intensity was examined by manual readjustments.
|
2.2
Peptide synthesis
Peptide synthesis was performed starting from Fmoc‐Oic preloaded
2‐chlorotrityl resin in dimethylformamide (DMF) as a solvent, as
previously described in detail.29,42 The resin loading was estimated
from the dry weight increment after the preloading procedure as
0.85 mmol g−1. The following coupling reagents were applied: HATU,
1‐[bis(dimethylamino)methylene]‐1H‐1,2,3‐triazolo[4,5‐b]pyridinium
3‐oxide hexafluorophosphate; TBTU, 2‐(1H‐benzotriazole‐1‐yl)‐
1,1,3,3‐tetramethylaminium tetrafluoroborate; HBTU, 2‐(1H‐benzotri-
azole‐1‐yl)‐1,1,3,3‐tetramethylaminium hexafluorophosphate; and
HOBt, 1‐hydroxybenzotriazole. The N‐terminal capping was per-
formed by acetic anhydride/DIPEA for acetylation, pivaloyl chloride/
DIPEA for pivaloylation, and pent‐4‐ynoic acid/TBTU/DIPEA for
pent‐4‐ynoylation. The peptides were cleaved by shaking 15 minutes
with hexafluoroisopropanol/dichloromethane in a 3:1 vol/vol mixture.
The solvents were blown off with a nitrogen current, and the residue
was lyophilized from an acetonitrile‐water mixture. C‐terminally free
crude peptides were purified on a silica gel column using ethyl ace-
tate‐methanol (19:1) as the first eluent and dichloromethane‐methanol
(9:1) as the second eluent. Fractions were analyzed by dry weight after
evaporation. C‐terminal 2,2‐difluoroethyl esters were prepared by
applying 2,2‐difluorodiazoethane as described.42,43 The crude peptides
were purified on a silica gel column using an ethyl acetate‐methanol
(19:1) mixture as an eluent. The peptides were analyzed by thin‐layer
chromatography (iodine chamber detection). For oligomeric peptides
≤5 residues, silica gel purification afforded pure peptide samples. For
the hexameric peptides, additional reversed‐phase high‐performance
liquid chromatography (HPLC) purification was applied (C‐18 column,
water‐acetonitrile gradient elution).
|
2.3.3
Measurements in micelles
Peptides in sodium dodecyl sulfate (SDS) micelles: The peptide amount
was determined gravimetrically, usually 2.0 to 4.0 mg (error 3%‐
10%). This amount was dissolved in acetonitrile and aliquotted into
vials containing varying amounts of 25mM SDS solution. The resulting
mixtures were freeze dried and then dissolved in a mixture of H2O
(545 μL) and D2O (55 μL). A volume of 525 μL of the resulting
solutions was sampled in 5‐mm NMR tubes for further analysis. The
resulting peptide concentration was 1 g L−1 in the oligo‐Oic samples
(equivalent to 1mM for AcO6F2) and 1mM in AcP6F2 and PivP6F2
samples. Consistency of the sample concentrations was checked by
1H NMR recorded with W5 water suppression. 19F{1H} NMR spectra
were recorded with decoupling during acquisition (inverse‐gated
decoupling) at 471 MHz and 298 K. The transmitter frequency was
typically at −127.00 ppm.
Quantitative diffusion measurements were performed using 1H
stimulated echo with bipolar gradients, spoil pulse (500 μs), and a
3‐9‐19 WATERGATE pulse tray for water suppression. The gradient
ramp was 10% to 98% of the maximal z‐gradient with linear incre-
ments giving rise to 128 array experiments. These were processed
using a standard software tool supplied with the NMR machine. The
diffusion time (δ) was 25 milliseconds, and the gradient pulse (Δ/2)
was 1 to 4 milliseconds depending on the expected diffusion values
(typically 1‐2 ms below and 2‐4 ms above the critical micelle concen-
tration [CMC]). Molecular sizes were calculated assuming pure water
dynamic viscosity and 298 K with Equation 1.29
For the nonamer PO9, the peptide sample (20 mg; after silica gel
purification) was dissolved in methanol (0.5 mL). The resulting solution
was kept at 277 K overnight to allow better precipitation. The mixture
was centrifuged, the supernatant was removed, and the residue was
rinsed with methanol (2 × 0.1 mL). The peptide was then dissolved in
a minimal amount of dichloromethane and dried in an open vial, and
the residue was suspended in an acetonitrile‐water mixture and freeze
dried to afford PO9 (11 mg).
|
2.3
Physical chemistry
|
2.3.1
NMR measurements
The temperature unit was calibrated using conventional acidified
methanol sample calibration. For referencing, only the standard sample
lock referencing was applied.
1H exchange spectroscopy NMR analysis was performed for a
20mM sample in deuterium oxide at 280 K and 700‐MHz frequency.
The exchange data were obtained in 1H z‐cross‐relaxation experiments
assisted with z‐gradients, and these spectra were processed as
described.28,29,44 The examined mixing times were 100, 50, 30, and
Â
Ã
1
logD logm2 s−1 ¼ −8:432− logð8:9 þ MW½DaꢀÞ
(1)
3
Diffusion values were calculated for the peptides using molecular
weight (MW). For the micelles, the particle sizes were assumed from
spherical particles45 containing 64 SDS anions.46 At 10mM SDS con-
centration, the mass ratio of the detergent peptide/micelle is 1:2.65;

KUBYSHKIN AND BUDISA
5 of 14
thus, peptide increases the particle size by 38%, and the final size was
assumed to be 24 kDa. The resulting predicted logD value is
−9.89 log m2 s−1
.
Hydrophobic residence (HR) values were calculated using
Equation 2, which assumes averaged molecular weight due to the
equilibration of the peptide between monomeric (MW of the peptide)
and micelle (MW of the micelle approximately 24 kDa). We used an
experimental value of the micelle diffusion logD(M) = −9.92 log m2
−1
s
and experimental peptide diffusion coefficients in the monomeric
peptides at 1 to 2mM SDS as logD(P).
10−3ð logD− logDðPÞÞ−1
10−3ð logDðMÞ− logDðPÞÞ−1
HR ¼
(2)
|
2.3.4
Circular dichroism
Peptide amounts were determined gravimetrically. Dry peptides were
mixed with octan‐1‐ol, 10mM aqueous SDS, or alkanes to afford
1.0mM final amide bond concentration. The spectra were acquired in
a 1‐mm quartz cell at 298 K.
FIGURE 2 LogP values for derivatives of selected amino acids
(octan‐1‐ol/water, 298 K)
|
3.2
C‐terminus
|
2.3.5
Solubility
Both C‐ and N‐terminal parts are important contributors to the lipophi-
licity, because these should be fully exposed to the solvent by the sec-
ondary structure. For example, in a previous study, we used C‐terminal
carboxylates to increase the water solubility of the Oic oligomers.
Indeed, in this case, the water solubility of a hexapeptide was in mM
range.29 In contrast, the use of the peptides in nonpolar media would
require nonionizable C‐terminal groups. Perhaps, the most common
and simple choices are (1) amides introduced during synthesis on
amide resins, (2) methyl esters prepared by post–resin esterification
of peptides, and (3) recently reported C‐terminal 2,2‐difluoroethyl
esters prepared using 2,2‐difluorodiazoethane.43,52 We previously
determined the logP values of the N‐acetyl proline derivatives, which
are shown in Figure 3.42 Notably, these results revealed strong hydro-
philicity for the terminal amide, lower hydrophilicity for both esters,
and the lowest hydrophilicity for the 2,2‐difluoroethyl ester. Thus,
2,2‐difluoroethyl ester was chosen for investigations of the peptide
constructs owing to its highest hydrophobicity. Also beneficial is the
fact that 2,2‐difluoroethyl esters can be subjected to a range of selec-
tive NMR studies based on 19F resonance detection. For example, this
will be applied for the logP measurements in the peptide samples.
Peptide (3‐5 mg) was shaken with hexane (HPLC grade, 0.5‐3.0 mL) for
1 hour. The heterogeneous mixture was centrifuged, the supernatant
portions (100, 200, or 500 μL) were taken, dried, and analyzed by dry
weight.
|
3
RESULTS
|
3.1
Amino acid lipophilicity
Initially, we thought of constructing the hydrophobic peptide oligo-
mers from their individual molecular components. Specifically, the
following factors should be taken into account when designing a
hydrophobic oligopeptide: (1) constructing amino acid residue, (2) pep-
tide chain length, (3) C‐terminal group, and (4) N‐terminal group.
Therefore, we have first characterized the amino acid side chains.
Thus, we determined the experimental octan‐1‐ol/water partitioning
constants for derivatives of several amino acids (Scheme 1). Accurate
partitioning values can be measured by NMR, and few elegant
methods have been developed quite recently.47-50 We measured logP
values for a few amino acid derivatives as shown in Figure 2. The anal-
ysis demonstrated that proline has a relatively polar structure (approx-
imately as polar as alanine), whereas Oic is relatively nonpolar, with a
higher hydrophobicity than the examined canonical hydrophobic
amino acids (Val, Ile, Leu, and Phe). In addition, the difference in the
lipophilicity between the Oic and Pro derivatives (approximately
1.5 logP units) is slightly larger than the previously reported difference
between N‐Fmoc derivatives of Pro and Sip (approximately 1.2 logP
units).51
|
3.3
N‐terminus
Modifications of the N‐terminal part may become more problematic,
because these can interfere with the N‐terminal trans/cis‐amide ratios.
The latter may affect the stability of the PII helix, since the most
FIGURE 3 LogP values for C‐terminal derivatives of N‐acetyl proline
SCHEME 1 Octan‐1‐ol/water partitioning of the model compounds
reported earlier42

6 of 14
KUBYSHKIN AND BUDISA
common alternative structure (polyproline‐I helix) requires isomeriza-
tion of the whole backbone into cis‐amides. Hartree‐Fock simula-
tions53 and ion‐mobility mass spectrometry54 results have suggested
that the polyproline isomerization mainly proceeds in a defined
manner from one terminus to another onward. Thus, the N‐terminal
functionality appears to be an important contributor to the peptide
conformation.
For example, the use of N‐terminal pivaloyl group has been previ-
ously introduced with the aim to lock the N‐terminal amide in the
trans‐amide conformation.55 We found that, for proline derivatives,
increasing the steric bulk of the N‐acyl substituent (from N‐acetyl to
N‐pivaloyl) stabilizes the trans‐amide bond by 5.5 kJ mol−1 owing to
the differences in the steric requirements of the α‐ and δ‐segments
of the pyrrolidine ring (Table 1, entries 1‐2).29,44 However, for Oic,
which bears a substituent at the δ‐carbon atom, we expected that an
increased bulk of the N‐acyl moiety in the Oic case may destabilize
the trans‐amide bond or even reverse the trans/cis equilibrium around
this residue following steric considerations.56-59
FIGURE 4 The crystal structure of the methyl ester of N‐pivaloyl Oic
reveals the trans‐amide bond conformation, which stabilizes the chair
conformation of the pyrrolidine ring and fixes the proline ring in the
exo‐C4‐pucker conformation. The strength of the intercarbonyl n→π *
interaction is enhanced owing to the decreased distance between the
interacting atoms. The ellipsoids are at 30% displacement probability.
Grey indicates carbon; blue, nitrogen; red, oxygen
|
3.4
Peptide length: lipophilicity
In order to characterize the influence of the peptide length on hydro-
phobicity, we prepared a series of oligopeptides up to 6 residues using
the procedure shown in Scheme 2. Fmoc‐Oic‐OH was prepared in
good yields using Fmoc chloride with TMS chloride assistance in
dichloromethane. The Fmoc‐Oic‐OH was then preloaded onto a 2‐
chlorotrityl‐polystyrene resin.60 The growth of the peptide chain was
achieved by applying HATU or TBTU/HOBt as the coupling reagents.
The synthesized peptides were cleaved from the resin using mild
hexafluoroisopropanol‐dichloromethane cleavage conditions61 and
purified on a short silica gel column. The peptides with free C‐termini
were converted to the esters using 2,2‐difluorodiazoethane, as
described previously.42,43 Purification of the esterified peptides was
performed by silica gel chromatography. This procedure yielded pep-
tides in good purity, except of the longest hexapeptide AcO6F2, which
contained some amount of the shorter peptide AcO5F2. The final puri-
fication of this peptide was afforded by standard reversed‐phase
HPLC. Indeed, we noticed that conventional silica gel chromatography
was unable to sufficiently separate Oic oligomers, but under the HPLC
We thus compared trans/cis ratios for 2 model compounds with
N‐acetyl (not bulky) and N‐pivaloyl (bulky) moieties. To our surprise,
the N‐pivaloyl Oic derivative exhibited an increased trans‐amide dom-
inance with the increased bulk of the N‐terminal group (Table 1,
entries 3‐4). Thus, the trans‐amide becomes even more stabilized with
bulky N‐terminal groups, contrary to our expectation. This outcome
might be due to the limited interference from the N‐acyl moiety on
the cyclohexane ring in Oic when fixed in the chair conformation. This
conclusion is supported by the crystal structure of this derivative, as
shown in Figure 4, which illustrates that the tert‐butyl moiety can suf-
ficiently fit in the allowed space without altering the conformation. It
additionally visualizes the interlock of the cyclohexane and the trans‐
amide bond conformations. These results suggest that the N‐pivaloyl
moiety is sufficient to lock the N‐terminal acyl group in the trans‐amide
state for both Pro and Oic and may not decrease the stability of the PII
fold. The lipophilicity gain due to the acyl‐to‐pivaloyl group exchange
was approximately +1.3 logP units.
TABLE 1 Physicochemical properties of the N‐terminally modified amino acid models as determined by 1H NMR
Amide Properties in Deuterium Oxide Solution
Thermodynamics
Kinetics
k, s−1
T, K trans→cis
Ea, kJ mol−1
Amino Acid
ΔGtrans/cis,
Entry Derivative
LogP
T, K Ktrans/cis
kJ mol−1
cis→trans
trans→cis cis→trans Reference
1
−0.44 0.05 298 4.95 0.05
−4.0 0.1 310 0.0070 0.0005 0.033 0.002 88.8 0.2 84.8 0.2 Kubyshkin
and
Budisa29
2
+0.81 0.13 298
46
2
−9.5 0.1 298
0.622 0.043
36.5 2.0
74.2 0.1 64.1 0.1 Kubyshkin
and
Budisa44
3
4
+1.03 0.05 298 9.38 0.40
+2.27 0.11 280 118 17
−5.5 0.1 310
−11.1 0.4 280
0.008 0.001
0.256 0.019
0.074 0.013 88.5 0.3 82.7 0.5 Kubyshkin
and
Budisa29
33.6 8.0
71.6 0.2 60.2 0.7 This work

KUBYSHKIN AND BUDISA
7 of 14
measured by using 19F NMR detection. Results demonstrate that the
oligoprolines have a better preference towards aqueous medium,
with a steady decrease of the logP at higher oligomeric numbers.
Interestingly, the lipophilicity difference between the N‐pivaloylated
peptide PivP6F2 and the N‐acetylated one AcP6F2 was identical as
in the small molecule models (entry 2 vs 1 in Table 1). The oligo‐Oic
peptides were remarkably lipophilic, having logP values increasing with
the oligomeric number, and this value reached the method detection
limits (this is 3 < logP)48 already at the oligomeric number 3 and higher.
|
3.5
Peptide length: inclusion into detergent
micelles
The octan‐1‐ol/water partitioning constants usually offer an accurate
measure of the hydrophobic tendencies in a series of analogous com-
pounds. However, logP values are poor indicators of the biologically
relevant environments, in which the hydrophobic phase is small rela-
tive to the aqueous phase. Therefore, we next aimed to characterize
distribution of the peptides between water and detergent micelles
formed by SDS (Scheme 3).
SDS has a CMC of 8.2mM at 298 K, and therefore, SDS micelles
feature small volume of the hydrophobic phase versus water. For
example, at 10mM, micelles will be formed, despite the presence of
only 2.9 mg SDS per 1 mL of water; thus, the alkane phase is about
1/600 of the net weight under these conditions. Inclusion into deter-
gent micelles can be seen from the 19F NMR spectra in samples
containing peptides and SDS below and above the CMC of the deter-
gent, which indicated a characteristic resonance shift for each oligomer
from the oligo‐Oic series (see Supporting Information). The affinity to
the micelles was followed by observation of the diffusion coefficients,
as measured by diffusion‐ordered NMR spectroscopy. 1H detection
was applied for simultaneous detection of the peptide and the
detergent. We measured diffusion parameters for the oligo‐Oic
peptide series, and hexaproline peptides (AcP6F2 and PivP6F2) were
used as reference.
SCHEME 2 Preparation of the oligomeric peptides (oligo‐Oic)
conditions, these hydrophobic peptides were easily separated,
producing relatively sharp elution peaks for each oligomer (see also
Figure 8B). N‐Pivaloylated peptide PivO6F2 was obtained analogously.
Oligoproline reference peptides were prepared starting from proline
preloaded 2‐chlorotrityl resin with TBTU reagent and without the
purification steps. The resulting peptide sequences are summarized in
Table 2.
With these peptides, we then measured octan‐1‐ol/water
partitioning values. In this case, concentration in the phases was
As seen from the diffusion (Figure 5A) of the detergent, the
micelle formation was overall not perturbed by the presence of the
peptides except for the samples close to the CMC (5 and 10mM),
where the most hydrophobic peptides slowed down the detergent dif-
fusion. The peptide diffusion coefficients logD are indicative of the
individual particle sizes, and these can be relatively well calculated
using previously derived Equation 129 assuming water viscosity, since
presence of SDS changes water viscosity only slightly.62 The mono-
meric peptide diffusion was calculated using peptide molecular
TABLE 2 Oligomeric peptides and their lipophilicity values (octan‐1‐
ol/water, 298 K)
Peptide ID
Peptide Sequence
LogP
Oligo‐Oic
AcO1F2
AcO2F2
AcO3F2
AcO4F2
AcO5F2
AcO6F2
PivO6F2
Ac‐(Oic)1‐OCH2CHF2
Ac‐(Oic)2‐OCH2CHF2
Ac‐(Oic)3‐OCH2CHF2
Ac‐(Oic)4‐OCH2CHF2
Ac‐(Oic)5‐OCH2CHF2
Ac‐(Oic)6‐OCH2CHF2
Piv‐(Oic)6‐OCH2CHF2
Oligoproline Reference
Ac‐(Pro)1‐OCH2CHF2
Ac‐(Pro)2‐OCH2CHF2
Ac‐(Pro)3‐OCH2CHF2
Ac‐(Pro)4‐OCH2CHF2
Ac‐(Pro)5‐OCH2CHF2
Ac‐(Pro)6‐OCH2CHF2
Piv‐(Pro)6‐OCH2CHF2
+1.39 0.05
+2.52 0.08
+3.2 0.2
>3
>3
>3
>3
AcP1F2
AcP2F2
AcP3F2
AcP4F2
AcP5F2
AcP6F2
PivP6F2
+0.02 0.06
−0.32 0.01
−0.49 0.03
−0.76 0.03
−0.99 0.05
−1.15 0.07
+0.16 0.03
SCHEME 3 Partitioning of the peptides between water and micelle
fractions. effMW = effective molecular weight; this appears as
weighted average of the 2 peptide states: in solution (monomeric) and
in micelles

8 of 14
KUBYSHKIN AND BUDISA
At the SDS concentrations 1 to 2mM (below the CMC) all the pep-
tides were found in the monomeric form in the aqueous solution
(Figure 5B). Increase of the detergent concentration did not affect
the peptide diffusion in the case of the hexaproline references, show-
ing that the peptides are residing in the aqueous phase. In order to pro-
vide a numerical measure, here, we introduce the HR value, which
represents the fraction of time during which the peptide solute is in
the hydrophobic (micellar) fraction as shown in Scheme 3. For example,
at 10mM SDS, the HR value is about 0% for AcP6F2 and PivP6F2
(peptide is in water), whereas in the case of the oligo‐Oic peptides, this
was 0.9% for AcO1F2, 23% for AcO2F2, and 26% for AcO3F2, and for
the highest oligomers, the HR approached 100% (peptide is in
micelles). Thus, the oligo‐Oic peptides demonstrate an exceptional
selectivity towards the hydrophobic phase, and this increased with
the peptide length. The selectivity is remarkable, considering a very
small fraction of the hydrophobic environment. From this, it follows
that oligo‐Oic peptides can potentially be applied as hydrophobic tags
for anchoring on hydrophobic surfaces and in biomembranes.
|
3.6
Solubility in hexane
We also compared the solubility of the Oic hexapeptides in aqueous
and alkane media. Whereas AcO6F2 is soluble in water at 0.3mM,
PivO6F2 dissolved only at 0.05mM. At the same time, when we
determined the solubility in hexane, this was found to be 0.7 and
7mM, respectively. Thus, the oligo‐Oic peptides can be considerably
more soluble in hexane than in water. These results also illustrate
importance of the N‐terminal group's contribution to the peptide
lipophilicity.
FIGURE 5 Inclusion of the peptides into the SDS detergent micelles.
Diffusion of the detergent (A) and the peptide (B) as found in 1H DOSY
spectra in the peptide/SDS aqueous samples at 298 K. The samples
contained 1, 2, 5, 10, 15, and 25mM of SDS, and the peptide
concentration was 1 g L−1, which corresponds to 1mM concentration
for AcO6F2, solvent H2O/D2O 9:1. CMC indicates critical micelle
concentration; DOSY, diffusion‐ordered NMR spectroscopy; HR,
hydrophobic residence; SDS, sodium dodecyl sulfate
|
3.7
Conformational analysis
In the previous section, we demonstrated the selectivity of oligo‐Oic
sequences towards hydrophobic phases. However, it is not yet clear
whether the PII conformation was maintained under the examined con-
ditions. To this end, we analyzed circular dichroism (CD) spectra of the
oligo‐Oic peptides in octan‐1‐ol and SDS micelles as shown in Figure 6.
It is worth to note that the CD shape observed for Oic oligomers
may differ from that of other PII‐forming peptide sequences.
Differences can result from the fact that Oic bears two (α‐CH‐N and
weights, whereas the whole micelle diffusion was calculated assuming
SDS micelles composed of 64 anions and complete inclusion of the
peptide (see Section 2.3.3). Predicted diffusion parameter
logD − 9.89 log m2 s−1 agrees with the one found experimentally for
the most hydrophobic peptides −9.92 log m2 s−1
.
FIGURE 6 Mean residue circular dichroism of the oligomeric peptides in octan‐1‐ol (A) and aqueous SDS micelles (10mM detergent) (B) at 1mM
amide, 298 K. SDS indicates sodium dodecyl sulfate

KUBYSHKIN AND BUDISA
9 of 14
δ‐CH‐N) chiral centers with cis‐relative configuration around the amide
chromophore, and this increases the symmetry of the residue to some
extent. As a result, the absolute intensity of the negative band at
212 nm (about −5 L mol−1 cm−1) was approximately 2‐fold lower than
that of the 204‐nm band in oligoproline peptides or the 196‐nm band
in short oligolysine.15,42 However, at this point, we limited our interpre-
tation to the change in the spectral shape within the peptide series. The
results suggest that the molecular conformation of the oligomeric Oic
peptides appears for the oligomers with 3 residues and persists in lon-
ger oligomers. It should not be ruled out that the shape of the CD curves
could be permissive for a contribution of β‐conformation, as this is also
the closest region on the Ramachandran plot. The β‐structure can be
speculated from the position of the main negative band, which is unusu-
ally red shifted (212 nm). Nonetheless, the observed red shift can be a
result of other effects. For example, it is that the negative band position
is highly dependent on the absorption properties of the amide bond,
whereas hydrophobic environments produce so‐called secondary
effects due to the polarization properties of the media.63
potentially enable chemical ligation using Cu‐catalyzed azide‐alkyne
cycloaddition. The free C‐terminus was chosen in order to improve
the solubility of the C‐terminally charged peptides in biologically rele-
vant buffers.
Three peptides, PO3, PO6, and PO9, were prepared from Fmoc‐
Oic preloaded 2‐chlorotrityl resin using Fmoc‐Oic‐OH/TBTU/HOBt/
DIPEA in a 2.5/2.4/2.5/5 molar ratio to the initial loading of the resin,
whereas the whole synthesis was performed in DMF. The resulting
peptides were cleaved from the resin and subsequently purified using
conventional silica gel column chromatography (Scheme 4). Peptide
PO3 was pure after this purification. Peptide PO6 contained a small
amount of a peptide that is 1 amino acid shorter, and the longest pep-
tide, PO9, was a heterogeneous mixture of oligopeptides with 5 to 9
Oic residues in the sequence, according to liquid chromatography‐
mass spectrometry analysis.
We further found that PO9 peptide sample produced a white pre-
cipitate when dissolved in methanol. Precipitation occurred at a certain
concentration threshold and increased with higher concentrations and
longer incubation times. This trend is depicted in Figure 8A, where the
dissolved peptide fraction was monitored by 1H NMR. The integral
intensity of the signal increased linearly with the concentration of
the dissolved peptide in fresh samples; after incubation, the peptide
intensity decreased in the most concentrated samples owing to precip-
itation. At the same time, the shorter oligomers PO3 and PO6 did not
We then examined CD spectra for the most lipophilic peptide from
the oligomeric series PivO6F2. Results shown in Figure 7 indicate
remarkable dependence on the negative band position from the
presence of the polar groups in the solvent, as this shifts from 211 to
214 nm when going from octan‐1‐ol to alkanes, thus indicating a
notable secondary effect. At the same time, spectrum in the SDS
micelles illustrates presence of the peptide in the hydrophobic core
of the micelle. Indeed, the linear peptide length of a PII‐forming
hexapeptide should be about 18 Å, well below the hydrophobic core
diameter, which is about 30 Å.46 Therefore, the peptide may be
completely immersed in the hydrophobic core of the micelle under
the examined conditions.
show any precipitation when dissolved in methanol up to 20 g L−1
.
The aggregation of polyproline structures has only been sporadi-
cally reported in the literature64; however, in the case of oligo‐Oic pep-
tides, this process is apparently driven by the hydrophobicity of the
cyclohexane side chain decoration. The precipitate formed by the
crude PO9 peptide was soluble in dichloromethane, similar to all of
the other oligo‐Oic peptides studied here. Further analysis showed
that this precipitate was a relatively pure sample of the target product
PO9, whereas the soluble fraction was enriched with the shorter olig-
omeric products (Figure 8B). Thus, precipitation favors longer peptide
sequences, and this feature can be exploited for the purification of
the target peptide. We also assume that the complication in the pep-
tide synthesis occurs owing to the low solubility of the longer peptide
sequences in DMF. For example, we noticed that during the synthesis
of higher oligomers, the resin did not produce a significant increase in
the volume, as typically occurs with other peptide sequences.
|
3.8
Synthesis of hydrophobic tags
Next, we decided to prepare a series of oligo‐Oic peptides that can be
applied to model hydrophobic peptide tags with a defined linear length
resulting from the assumed rigid PII skeleton. First, we synthesized 3
oligopeptides capped with a pent‐4‐ynoyl moiety at the N‐terminus
and free at the C‐terminus. The N‐terminal modification should
FIGURE 7 Mean residue circular dichroism of PivO6F2 in a series of
solvents at 1mM amide concentration and 298 K. SDS micelles are at
10mM concentration of the detergent. SDS indicates sodium dodecyl
sulfate
SCHEME 4 Synthesis of the hydrophobic peptide tags PO3, PO6,
and PO9 with N‐terminal pent‐4‐ynoyl modification

10 of 14
KUBYSHKIN AND BUDISA
FIGURE
9
Space‐filled models of oligopeptides in PII helical
conformation. Yellow indicates hydrogen; grey, carbon; blue,
nitrogen; red, oxygen
sequence, the backbone NH groups are indeed exposed to the
medium. When NH bonds are eliminated in oligoproline structures,
the peptide may become more tolerant to exposure in nonpolar envi-
ronment; thus, one can expect oligoproline to become hydrophobic.
Analysis of the experimental octan‐1‐ol/water logP values for the
reference oligoproline series (Table 1) indicates an ambivalent charac-
ter of the PII helix formed by proline, with a slight preference towards
aqueous medium. Moreover, it is also documented that the nonpolar
media can promote isomerization of oligoprolines into more compact
polyproline‐I helical arrangement,34 which is otherwise disfavored by
the amide‐amide interactions in the backbone. Thus, the oligoproline‐
based PII helix is poorly compatible with nonpolar media, owing to
the backbone amide accessibility to the solvent. Finally, for biologically
relevant systems, where water is highly abundant, oligoproline
sequences may only prefer the aqueous environment. This conclusion
follows from our observation of the diffusion properties of hexaproline
peptides in the presence of SDS, which showed the diffusion of a
monomeric peptide that was not bound to micelles (Figure 5).
FIGURE 8 A, Precipitation of the PO9 peptide from its mixture with
shorter oligomers followed by 1H NMR. Freshly lyophilized peptide
was first dissolved in methanol‐d4 and kept at 298 K for 24 hours. The
fraction observed by the 1H NMR in solution is a monomeric peptide
as seen from the diffusion properties of the resonances. B, Analytical
HPLC chromatograms of the precipitate (top) and the supernatant
(bottom) produced by this process. HPLC indicates high‐performance
liquid chromatography
Therefore, the resin fails to swell, which hinders contact of the cou-
pling reagents with the peptide. Nonetheless, it should be noted that
the synthesis was performed on a regular fashion, and we have not
tried to further solve the synthesis complications at this time (eg, by
applying low‐loading resin).
|
4
DISCUSSION
Therefore, in order to design a hydrophobic PII helix, proline struc-
ture should be further decorated by more lipophilic elements. The first
example was provided by Martin et al, who reported on a PII helix
formed by a hydrophobic proline analogue, silaproline.6,7 Indeed,
silaproline is 1.2 logP units more lipophilic than proline (in N‐Fmoc
derivatives),51 and this may create the driving force for the preferences
towards nonpolar media. In our examination of model compounds, we
found that Oic is 1.5 logP units more lipophilic than proline (Figure 2).
Peptides are essentially polar constructs, as follows from the high
polarity of the amide bond. Therefore, it may represent a fundamental
challenge to design a stable PII helix for nonpolar media, as this
semiextended conformation allows high backbone exposure to the
environment. To illustrate this argument, we modeled a 9‐residue PII
helix formed by oligoalanine, oligoproline, and oligo‐Oic peptides using
previously reported crystal structure of a hexa‐Oic peptide as a tem-
plate (Figure 9).29 It can be seen clearly that in the polyalanine

KUBYSHKIN AND BUDISA
11 of 14
It thus may be a good starting point for construction of a PII helix affine
to purely hydrophobic environments based on oligo‐Oic structure.
In our experiments of octan‐1‐ol/water and SDS micelle/water
(Figure 5) partitioning, we demonstrated high selectivity of the oligo‐
Oic sequences towards hydrophobic environments. For example, logP
went above the common detection limit level (logP > 3) at the oligo-
meric number 3 (Table 2). Diffusion experiments demonstrated that
the peptides are solely present in the micellar fraction in 10mM SDS
starting from oligomeric number 4 and in higher oligomers (Figure 5).
Furthermore, synthesis of the peptides on the solid phase indicated
complications in growing sequences longer than 5 Oic residues owing
to the low solubility of the hydrophobic peptides in DMF. Finally, we
discovered that the nonameric peptide can be precipitated from meth-
anol, and this feature can be further exploited towards purification of
the longest oligomeric peptides (Figure 8). All these observations illus-
trate an increasing hydrophobicity of the oligo‐Oic peptides with
increasing peptide chain length.
These effects are usually quite weak for α‐helical structures
(<0.5 nm),63 but in a semiextended PII helix with exposed backbone
carbonyl groups, these may produce larger shifts.
As already mentioned above, at this point, we could not
completely exclude the β‐conformation contribution. To address this
problem, we modeled a possible isomerization of the PII helix into a
β‐like structure by readjustment of the ψ angles from 145 (PII helix)
to 120 (likewise in a β‐strand) in a model (Oic)9 structure, and this is
shown in Figure 10. As seen from the figure, the PII helical structure
arranges the outer cyclohexane rings in a relatively rigid fashion, creat-
ing a cavity between the cyclohexane rings of residues N and N + 3. At
this point, we speculated that this may produce a significant increase in
the surface energy, which can in parts be compensated by intercalation
of the solvent molecules. Contrary to this situation, potential isomeri-
zation into the β‐like conformation significantly increases the distance
between the stacked cyclohexane rings, and this eliminates the cavity
feature. Following this scenario, intercalation of more or less bulky sol-
vent may influence the putative PII ↔ β‐conformation equilibrium. For
this reason, we examined CD spectra of the peptide PivO6F2 in
alkanes with different spatial structure: linear (n‐hexane), cyclic (cyclo-
hexane), and branched (isooctane). Careful inspection of the results
shown in Figure 7 show minimal dependence on the CD band positions
from the solvent branching, thus indicating that the speculated equilib-
rium might not be present. Thus, CD curves might indicate a relatively
pure PII helix.
Remarkably, the choice of end groups could play a crucial role in
peptide architecture, and this can significantly alter both hydrophobic-
ity and solubility (Figure 3, Table 1). For example, by taking relatively
lipophilic N‐terminal N‐pivaloyl and C‐terminal 2,2‐difluoroethyl ester
groups, we showed that a hexa‐Oic peptide can be dissolved in hexane
up to 7mM concentration.
We found that the trans‐amide bond formed by an Oic residue
permits bulky N‐terminal acyl‐groups (Table 1); therefore, N‐terminal
modifications should not compromise the PII helix formation. Further
analysis of the CD spectra demonstrated that the peptides reached
regular secondary structure with the oligomeric number 3 and higher
(Figure 6). Interestingly, the spectra showed an extraordinary red‐
shifted position of the negative band, when measured in octan‐1‐ol
(211 nm) or SDS micelles (212 nm). The position of this band is influ-
enced by the absorption properties of the amide bond, and this would
be 196 nm in oligolysine with secondary amide bond and 204 nm in
oligomeric proline with tertiary amide bond.15,42 This can be further
affected by the nature of the amino acid, as the band shifts to
207 nm for hydrophobic (Sip)11 sequence (in trifluoroethanol).6
Encapsulation of the PII backbone in a permanent hydrophobic shell
in oligo‐Oic may produce a further red shift, whereas hydrophobic sol-
vents can enhance this shift owing to the so‐called secondary effects.
Considering the stabilizing forces, it is known that transmembrane
α‐helices are remarkably uniformed in the structure when placed in the
hydrophobic environment.65 This fact is attributed to increasing
strength of the stabilizing hydrogen bonds in nonpolar media.66
Conversely, the PII helix is not stabilized by macrocyclic hydrogen
bonds (13 atoms forming a ring in an α‐helix), but it results from
short‐distant interactions between the neighboring residues (5 atoms
in a ring; see Figure 4 for illustration).25,28 The latter may be weak,
and the previously found estimates for the interaction energies are
around 3 to 3.5 kJ mol−1 per residue for proline28 and ≤5 kJ mol−1
per residue for Oic with locked exo‐pucker, with ≤2 kJ mol−1 per resi-
due stability enhancement via the cascade effect.29 At the same time,
this interaction is less dependent on environment conditions such as
temperature, or presence of denaturing additives due to the
FIGURE 10 Space‐filled models for oligo‐
Oic peptides in PII helix and its isomerization
into a β‐like conformation. Yellow indicates
hydrogen; grey, carbon; blue, nitrogen; red,
oxygen

12 of 14
KUBYSHKIN AND BUDISA
sequence. Org Lett. 2017;19:2937‐2940. https://doi.org/10.1021/acs.
orglett.7b01172
favorability of the intramolecular 5‐membered ring formation.
Recently, we42 and others67 have provided the data, indicating that
the polar PII‐stabilizing interaction energy may also increase in nonpo-
lar environments, although with a small factor (increase by ≤2 kJ mol
−1). Since β‐strand structure is the least stabilized owing to the car-
bonyl‐carbonyl interactions,68 the β‐contribution in the absence of
an alternative stabilizing force seems very unlikely. For example, isom-
erization of some model oligoproline‐based structure into the β‐like
antiparallel sheets was reported in nonpolar media, when this was
assisted by backbone‐side chain hydrogen bond formation,35-37 but
not otherwise.35 Thus, the CD curves in the oligo‐Oic sequences can
be attributed to dominating PII helix, which is encapsulated in a hydro-
phobic shell and is still sensitive to the presence of the polar groups in
the solvent via the secondary effects.
6. Martin C, Lebrun A, Martinez J, Cavelier F. Synthesis of
homopolypeptides with PPII structure. J Polym Sci Part A: Polym Chem.
2013;51:3103‐3109. https://doi.org/10.1002/pola.26705
7. Martin C, Legrand B, Lebrun A, Berthomieu D, Martinez J, Cavelier F.
Silaproline helical mimetics selectively form an all‐trans PPII helix.
Chem
A Eur J. 2014;20:14240‐14244. https://doi.org/10.1002/
chem.201404820
8. Adzhubei AA, Sternberg MJE, Makarov AA. Polyproline‐II helix in
proteins: structure and function. J Mol Biol. 2013;425:2100‐2132.
https://doi.org/10.1016/j.jmb.2013.03.018
9. Peggion C, Moretto A, Formaggio F, Crisma M, Toniolo C. Multiple,
consecutive, fully‐extended 2.05‐helix peptide conformation. Peptide
Sci. 2013;100:621‐636. https://doi.org/10.1002/bip.22267
10. Cowan PM, McGavin S. Structure of poly‐L‐proline. Nature.
1955;176:501‐503. https://doi.org/10.1038/176501a0
In summary, we were able to demonstrate that the use of a hydro-
phobic proline analogue allows the construction of a relatively stable
PII helix, which is yet remarkably selective towards nonpolar media
and can be soluble in hydrocarbons. As a result, oligomeric Oic
sequences provide a tool for molecular spacing in extremely hydropho-
bic environments such as organic solvents or biomembranes. Further
studies of this novel hydrophobic polyproline motif are possible using
the synthesis and purification of the oligo‐Oic sequences described
herein. Notably, in most cases, HPLC purification can be avoided, thus
enabling very easy handling of the material. In addition, the ready
commercial availability of the enantiomerically pure Oic monomer with
the (2S,3aS,7aS) configuration will enable further extensive exploration
of oligo‐Oic sequences for applications in peptide, materials, and life
sciences.
11. Kurtz J, Berger A, Katchalski E. Mutarotation of poly‐L‐proline. Nature.
1956;178:1066‐1067. https://doi.org/10.1038/1781066a0
12. Mansiaux Y, Joseph AP, Gelly J‐C, de Brevern AG. Assignment of
polyproline II conformation and analysis of sequence‐structure rela-
tionship. PLoS One. 2011;6:e18401. https://doi.org/10.1371/journal.
pone.0018401
13. Shi Z, Chen K, Liu Z, Ng A, Bracken WC, Kallenbach NR. Polyproline II
propensities from GGXGG peptides reveal an anticorrelation with
β‐sheet scales. Proc Natl Acad Sci U S A. 2005;102:17964‐17968.
https://doi.org/10.1073/pnas.0507124102
14. Brown AM, Zondlo NJ. A propensity scale for type II polyproline helices
(PPII): aromatic amino acids in proline‐rich sequences strongly disfavor
PPII due to proline‐aromatic interactions. Biochemistry. 2012;51:5041‐
5051. https://doi.org/10.1021/bi3002924
15. Rucker AL, Creamer TP. Polyproline II helical structure in protein
unfolded states: lysine peptides revisited. Protein Sci. 2002;11:980‐
985. https://doi.org/10.1110/ps.4550102
ACKNOWLEDGEMENT
16. Johannessen C, Kapitán J, Collet H, Commeyras H, Hecht L, Barron LD.
Poly(L‐proline) II helix propensities in poly(L‐lysine) dendrigraft
generations from vibrational Raman optical activity. Biomacromolecules.
2009;10:1662‐1664. https://doi.org/10.1021/bm9002249
V.K. acknowledges Deutsche Forschungsgemeinschaft (DFG) research
group FOR 1805 for a postdoctoral position.
17. Shi Z, Chen K, Liu Z, Kallenbach NR. Conformation of the backbone in
unfolded proteins. Chem Rev. 2006;106:1877‐1897. https://doi.org/
10.1021/cr040433a
CONFLICT OF INTEREST
There is no conflict of interest to declare.
18. Rath A, Davidson AR, Deber CM. The structure of “unstructured”
regions in peptides and proteins: role of the polyproline II helix in pro-
tein folding and recognition. Biopolymers. 2005;80:179‐185. https://
doi.org/10.1002/bip.20227
ORCID
Vladimir Kubyshkin http://orcid.org/0000-0002-4467-4205
Nediljko Budisa http://orcid.org/0000-0001-8437-7304
19. Kubyshkin V, Budisa N. Synthetic alienation of microbial organisms by
using genetic code engineering: why and how? Biotechnol J. 2017;12.
1600097. https://doi.org/10.1002/biot.201600097
REFERENCES
20. Hartman H, Smith TF. The evolution of the ribosome and the genetic
1. Nyathi Υ, Wilkinson ΒΜ, Pool ΜΡ. Co‐translational targeting and
translocation of proteins to the endoplasmic reticulum. Biochim Biophys
Acta–Mol Cell Res. 2013;1833:2392‐2402. https://doi.org/10.1016/j.
bbamcr.2013.02.021
code. Life. 2014;4:227‐249. https://doi.org/10.3390/life4020227
21. SmithTF, Hartman H. The evolution of class II aminoacyl‐tRNA synthe-
tases and the first code. FEBS Lett. 2015;589:3499‐3507. https://doi.
org/10.1016/j.febslet.2015.10.006
2. Gloge F, Becker AH, Kramer G, Bukau B. Co‐translational mechanisms
of protein maturation. Curr Opin Struct Biol. 2014;24:24‐33. https://
doi.org/10.1016/j.sbi.2013.11.004
22. Stapley BJ, Creamer TP. A survey of left‐handed polyproline II helices.
Protein Sci. 1999;8:587‐595. https://doi.org/10.1110/ps.8.3.587
23. Bicudo TC, Bicudo RC, Forato LA, et al. γ‐Zein secondary structure in
solution by circular dichroism. Biopolymers. 2008;89:175‐178. https://
doi.org/10.1002/bip.20884
3. Killian JA. Synthetic peptides as models for intrinsic membrane
proteins. FEBS Lett. 2003;555:134‐138. https://doi.org/10.1016/
S0014‐5793(03)01154‐2
24. Pujals S, Fernández‐Carneado J, Kogan MJ, Martinez J, Cavelier F,
Giralt E. Replacement of a proline with silaproline causes a 20‐fold
increase in the cellular uptake of a pro‐rich peptide. J Am Chem Soc.
2006;128:8479‐8483. https://doi.org/10.1021/ja060036c
4. Rémond E, Martin C, Martinez J, Cavelier F. Silicon‐containing amino
acids: synthetic aspects, conformational studies, and applications to
bioactive peptides. Chem Rev. 2016;116:11654‐11684. https://doi.
org/10.1021/acs.chemrev.6b00122
5. Fanelli R, Berthomieu D, Didierjean C, et al. Hydrophobic α,α‐disubsti-
tuted disilylated TESDpg induces incipient 310‐helix in short tripeptide
25. Newberry RW, Raines RT. The n→π* interaction. Acc Chem Res.
2017;50:1838‐1846. https://doi.org/10.1021/acs.accounts.7b00121

KUBYSHKIN AND BUDISA
13 of 14
26. Kamer KJ, Choudhary A, Raines RT. Intimate interactions with carbonyl
groups: dipole‐dipole or n→π*? J Org Chem. 2013;78:2099‐2103.
https://doi.org/10.1021/jo302265k
44. Kubyshkin V, Budisa N. Amide rotation trajectories probed by symme-
try. Org Biomol Chem. 2017;15:6764‐6772. https://doi.org/10.1039/
C7OB01421J
27. Rahim A, Saha P, Jha KK, Sukumar N, Sarma BK. Reciprocal carbonyl‐
carbonyl interactions in small molecules and proteins. Nat Commun.
2017;8:78. https://doi.org/10.1038/s41467‐017‐00081‐x
45. Hayashi S, Ikeda S. Micelle size and shape of sodium dodecyl sulfate in
concentrated sodium chloride solutions. J Phys Chem. 1980;84:744‐
751. https://doi.org/10.1021/j100444a011
28. Kubyshkin V, Durkin P, Budisa N. Energetic contribution to both acidity
46. Duplâtre G, Ferreira Marques MF, da Graça Miguel M. Size of sodium
dodecyl sulfate micelles in aqueous solutions as studied by positron
annihilation lifetime spectroscopy. J Phys Chem. 1996;100:16608‐
16612. https://doi.org/10.1021/jp960644m
and conformational stability in peptide models. New
2016;40:5209‐5220. https://doi.org/10.1039/C5NJ03611A
J Chem.
29. Kubyshkin V, Budisa N. Construction of
a polyproline structure
with hydrophobic exterior using octahydroindole‐2‐carboxylic acid.
Org Biomol Chem. 2017;15:619‐627. https://doi.org/10.1039/
C6OB02306A
47. Mo H, Balko KM, Colby DA. A practical deuterium‐free NMR method
for the rapid determination of 1‐octanol/water partition coefficients
of pharmaceutical agents. Bioorg Med Chem Lett. 2010;20:6712‐6715.
https://doi.org/10.1016/j.bmcl.2010.08.145
30. Schuler B, Lipman EA, Steinbach PJ, Kumke M, Eaton WA. Polyproline
and the “spectroscopic ruler” revisited with single‐molecule fluores-
cence. Proc Natl Acad Sci U S A. 2005;102:2754‐2759. https://doi.
org/10.1073/pnas.0408164102
48. Linclau B, Wang Z, Compain G, et al. Investigating the influence of
(deoxy)fluorination on the lipophilicity of non‐UV‐active fluorinated
alkanols and carbohydrates by a new logP determination method.
Angew Chem Int Ed. 2016;55:674‐678. https://doi.org/10.1002/
anie.201509460
31. Garbuio L, Lewandowski B, Wilhelm P, et al. Shape persistence of
polyproline II helical oligoprolines. Chem A Eur J. 2015;21:10747‐
10753. https://doi.org/10.1002/chem.201501190
49. Soulsby D, Chica JAM. Determination of partition coefficients using 1
H
NMR spectroscopy and time domain complete reduction to amplitude‐
frequency table (CRAFT) analysis. Magn Reson Chem. 2017;55:724‐
729. https://doi.org/10.1002/mrc.4582
32. Sobakinskaya E, Schmidt M, Regner T. Theory of FRET “spectroscopic
ruler” for short distances: application to polyproline. J Phys Chem B.
2018;122:54‐67. https://doi.org/10.1021/acs.jpcb.7b09535
50. Stéen EJL, Nyberg N, Lehel S, et al. Development of
a simple
33. Dobitz S, Aronoff MR, Wennemers H. Oligoprolines as molecular
entities for controlling distance in biological and material sciences.
Acc Chem Res. 2017;50:2420‐2428. https://doi.org/10.1021/acs.
accounts.7b00340
proton nuclear magnetic resonance‐based procedure to estimate the
approximate distribution coefficient at physiological pH (logD7.4):
evaluation and comparison to existing practices. Bioorg Med Chem Lett.
2017;27:319‐322. https://doi.org/10.1016/j.bmcl.2016.11.048
34. Zanna N, Milli L, Del Secco B, Tomasini C. Factors affecting the stabili-
zation of polyproline II helices in a hydrophobic environment. Org Lett.
2016;18:1662‐1665. https://doi.org/10.1021/acs.orglett.6b00532
51. Cavelier F, Vivet B, Martinez J, et al. Influence of silaproline on peptide
conformation and bioactivity. J Am Chem Soc. 2002;124:2917‐2933.
https://doi.org/10.1021/ja017440q
35. Sonar MV, Ganesh KN. Water‐induced switching of β‐structure to
polyproline II conformation in the 4S‐aminoproline polypeptide via
H‐bond rearrangement. Org Lett. 2010;12:5390‐5393. https://doi.
org/10.1021/ol1021993
52. Peng S‐Q, Zhang X‐W, Zhang L, Hu X‐G. Esterification of carboxylic
acids with difluoromethyl diazomethane and interrupted esterification
with trifluoromethyl diazomethane:
a fluorine effect. Org Lett.
2017;19:5689‐5692. https://doi.org/10.1021/acs.orglett.7b02866
36. Bansode ND, Sonar MV, Ganesh KN. A nanofiber assembly directed by
the non‐classical antiparallel β‐structure from 4S‐(OH) proline polypep-
tide. Chem Commun. 2016;52:4884‐4887. https://doi.org/10.1039/
C6CC00838K
53. Kang YK, Jhon JS, Park S. Conformational preferences of proline
oligopeptides. J Phys Chem B. 2006;110:17645‐17655. https://doi.
org/10.1021/jp0629792
54. Shi L, Holliday AE, Shi H, et al. Characterizing intermediates along the
transition from polyproline I to polyproline II using ion mobility
spectrometry‐mass spectrometry. J Am Chem Soc. 2014;136:12702‐
12711. https://doi.org/10.1021/ja505899g
37. Bansode ND, Madhanagopal B, Sonar MV, Ganesh KN.
Stereodependent and solvent‐specific formation of unusual β‐structure
through side chain‐backbone H‐bonding in C4(S)‐(NH2/OH/NHCHO)‐
L‐prolyl polypeptides. Biopolymers. 2017;108:e22981. https://doi.org/
10.1002/bip.22981
55. Boussard G, Marraud M, Aubry A. Experimental investigations on the
backbone folding of proline‐containing model tripeptides. Biopolymers.
1979;18:1297‐1331. https://doi.org/10.1002/bip.1979.360180602
38. Vivet B, Cavelier F, Martinez J. Synthesis of silaproline, a new proline
surrogate. Eur J Org Chem. 2000;807‐811. https://doi.org/10.1002/
(SICI)1099-0690(200003)2000:5%3C807::AID-EJOC807%3E3.0.
CO;2-E
56. Beausoleil E, Lubell WD. Steric effects on the amide isomer
equilibrium of prolyl peptides. Synthesis and conformational analysis
of N‐acetyl‐5‐tert‐butylproline N'‐methylamides.
1996;118:12902‐12908. https://doi.org/10.1021/ja962013b
J Am Chem Soc.
39. Tacke R, Merget M, Bertermann R, Bernd M, Beckers T, Reissmann T.
Syntheses and properties of silicon‐ and germanium‐containing α‐
amino acids and peptides: a study on C/Si/Ge bioisosterism. Organome-
tallics. 2000;19:3486‐3497. https://doi.org/10.1021/om000169l
57. Halab L, Lubell WD. Use of steric interactions to control peptide
turn geometry. Synthesis of type VI β‐turn mimics with 5‐tert‐
butylproline.
10.1021/jo990294a
J Org Chem. 1999;64:3312‐3321. https://doi.org/
40. Martin C, Vanthuyne N, Miramon H, Martinez J, Cavelier F. Resolution
of protected silaproline for a gram scale preparation. Amino Acids.
2012;43:649‐655. https://doi.org/10.1007/s00726‐011‐1112‐y
58. Halab L, Lubell WD. Influence of N‐terminal residue stereochemistry on
the prolyl amide geometry and the conformation of 5‐tert‐butylproline
type VI β‐turn mimics. J Pept Sci. 2001;7:92‐104. https://doi.org/
10.1002/psc.297
41. Chung JYL, Shevlin M, Klapars A, Journet M. Asymmetric synthesis of
N‐Boc‐(R)‐silaproline via Rh‐catalyzed intramolecular hydrosilylation
of dehydroalanine and continuous flow N‐alkylation. Org Lett.
2016;18:1812‐1815. https://doi.org/10.1021/acs.orglett.6b00548
59. Chaume G, Simon J, Lensen N, Pytkowicz J, Brigaud T, Miclet E.
Homochiral versus heterochiral trifluoromethylated pseudoproline
containing dipeptides: a powerful tool to switch the prolyl‐amide bond
conformation. J Org Chem. 2017;82:13602‐13608. https://doi.org/
10.1021/acs.joc.7b01944
42. Kubyshkin V, Budisa N. Hydrolysis, polarity, and conformational impact
of C‐terminal partially fluorinated ethyl esters in peptide models.
Beilstein J Org Chem. 2017;13:2442‐2457. https://doi.org/10.3762/
bjoc.13.241
43. Mykhailiuk PK, Kishko I, Kubyshkin V, Budisa N, Cossy J. Selective
19F‐labeling of functionalized carboxylic acids with difluoromethyl
diazomethane (CF2HCHN2). Chem A Eur J. 2017;23:13279‐13283.
https://doi.org/10.1002/chem.201703446
60. Barlos K, Gatos D, Kallitsis J, et al. Darstellung geschützter
peptid‐fragmente unter einsatz substituierter triphenylmethyl‐harze.
Tetrahedron Lett. 1989;30:3943‐3946. https://doi.org/10.1016/
S0040‐4039(00)99290‐6

14 of 14
KUBYSHKIN AND BUDISA
61. Bollhagen R, Schmiedberger M, Barlos K, Grell E. A new reagent for the
cleavage of fully protected peptides synthesised on 2‐chlorotrityl chlo-
ride resin. J Chem Soc Chem Commun. 1994;2559‐2560. https://doi.
org/10.1039/C39940002559
67. Siebler C, Maryasin B, Kuemin M, et al. Importance of dipole moments
and ambient polarity for the conformation of Xaa‐Pro moieties—a com-
bined experimental and theoretical study. Chem Sci. 2015;6:6725‐
6730. https://doi.org/10.1039/C5SC02211H
62. Kushner LM, Duncan BC, Hoffman JI. A viscometric study of the
micelles of sodium dodecyl sulfate in dilute solutions. J Res Natl Stand.
1952;49:85‐90. https://doi.org/10.6028/jres.049.011
68. Bartlett GJ, Choudhary A, Raines RT, Woolfson DN. n→π* interactions
in proteins. Nat Chem Biol. 2010;6:615‐620. https://doi.org/10.1038/
nchembio.406
63. Chen Y, Wallace BA. Secondary solvent effects on the circular dichro-
ism spectra of polypeptides in non‐aqueous environments: influence of
polarisation effects on the far ultraviolet spectra of alamethicin.
Biophys Chem. 1997;65:65‐74. https://doi.org/10.1016/S0301‐
4622(96)02225‐9
SUPPORTING INFORMATION
Additional Supporting Information may be found online in the
supporting information tab for this article.
64. Tooke L, Duitch L, Measey TJ, Schweitzer‐Steiner R. Kinetics of the
self‐aggregation and film formation of poly‐L‐proline at high tempera-
tures explored by circular dichroism spectroscopy. Biopolymers.
2010;93:451‐457. https://doi.org/10.1002/bip.21361
How to cite this article: Kubyshkin V, Budisa N. Exploring
65. Kim S, CrossTA. Uniformity, ideality, and hydrogen bonds in transmem-
brane α‐helices. Biophys J. 2002;83:2084‐2095. https://doi.org/
10.1016/S0006‐3495(02)73969‐6
hydrophobicity limits of polyproline helix with oligomeric
octahydroindole‐2‐carboxylic acid.
J Pep Sci. 2018;e3076.
https://doi.org/10.1002/psc.3076
66. Sheu S‐Y, Yang D‐Y, Selzle HL, Schlag EW. Energetics of hydrogen
bonds in peptides. Proc Natl Acad Sci U S A. 2003;100:12683‐12687.
https://doi.org/10.1073/pnas.2133366100