UC Santa Barbara
UC Santa Barbara Previously Published Works
Title
Tuning underwater adhesion with cation-π interactions.
Permalink
https://escholarship.org/uc/item/6320820w
Journal
Nature chemistry, 9(5)
ISSN
1755-4330
Authors
Gebbie, Matthew A
Wei, Wei
Schrader, Alex M
et al.
Publication Date
2017-05-01
DOI
10.1038/nchem.2720
Peer reviewed
eScholarship.org Powered by the California Digital Library
University of California
Tuning underwater adhesion with
cation–π interactions
Matthew A. Gebbie
1,2
, Wei Wei
2
, Alex M. Schrader
2,3
,ThomasR.Cristiani
1
, Howard A. Dobbs
4
,
Matthew Idso
4
, Bradley F. Chmelka
4
,J.HerbertWaite
2,3
*
and Jacob N. Israelachvili
1,2,4
*
Cation–π interactions drive the self-assembly and cohesion of many biological molecules, including the adhesion proteins
of several marine organisms. Although the origin of cation–π bonds in isolated pairs has been extensively studied, the
energetics of cation–π-driven self-assembly in molecular films remains uncharted. Here we use nanoscale force
measurements in combination with solid-state NMR spectroscopy to show that the cohesive properties of simple aromatic-
and lysine-rich peptides rival those of the strong reversible intermolecular cohesion exhibited by adhesion proteins of
marine mussel. In particular, we show that peptides incorporating the amino acid phenylalanine, a functional group that is
conspicuously sparing in the sequences of mussel proteins, exhibit reversible adhesion interactions significantly exceeding
that of analogous mussel-mimetic peptides. More broadly, we demonstrate that interfacial confinement fundamentally
alters the energetics of cation–π-mediated assembly: an insight that should prove relevant for diverse areas, which range
from rationalizing biological assembly to engineering peptide-based biomaterials.
N
ature employs a variety of non-covalent interactions to tune
the structures and functions of proteins, peptides and other
complex biological molecules with cation–π interactions
that feature prominently in biological self-assembly
1–4
, molecular
recognition
5–7
and molecular cohesion and adhesion
8–10
. In compo-
site materials, such as protein–solid interfaces, delamination can
occur within a glue (peptide) film, which is called cohesive failure.
Delamination can also occur at a glue– surface (peptide-surface)
interface, which is called adhesive failure (Supplementary Fig. 1).
These two terms are often used interchangeably in the broader
scientific literature, and many adhesives actually fail via cohesive
mechanisms
11–13
.
Cation–π interactions are electrostatic in origin and occur
between cations and electron-rich π orbitals
1,14,15
. Particularly
strong cation–π binding occurs when cations interact with the delo-
calized π orbitals perpendicular to the plane of aromatic rings.
Although cation–π intera ctions are much str onger in the gas phase
than in condensed phases, they still exceed the str ength of hydrogen
bonds, and possibly ev en charge–charge inter actions, in aqueous
solutions
1,16
. As a result, cation–π intera ctions provide an attr a ctiv e
molecular design model to develop molecules that can function as
adhesives in underwa ter environments. Such ma terials could be
used to address a number of substantial engineering challenges,
which range from functioning as biomedical adhesives that can
repla ce damaging screws in surgical applications
17
to pro viding cohe-
sive binding domains that hold together tissue-engineering scaffolds
18
.
Despite this technological promise, the relative binding energetics
of cation– π interactions at interfaces cannot yet be predicted
a priori. Indeed, much of the current understanding of cation–π
binding strengths in condensed phases is either extrapolated from
gas-phase experiments and calculations
1,14–16
or inferred from the
proximity of aromatic and cationic amino acids in protein crystal
structures
2,3,6
. Nevertheless, it remains unclear whether these insights
are directly applicable to rationalizing cation–π energetics a t interfaces.
Notably, cation–π binding at interfaces typically involves the
formation of several cation–π binding pairs in close proximity,
in which the electrostatic repulsion between two closely spaced
(positive) pairs can compromise the favourable free energy gained
by forming the two cation–π bonds. The complexation of
anions with cation–π pairs could provide the necessary charge
compensation to eliminate this electrostatic repulsion. Indeed,
researchers have studied the impact of anions on isolated ternary
cation–π–anion binding groups
19–23
, but emphasized that the
three-body interaction term in cation–π–anion complexation is
anti-cooperative and weakens the interaction strength of (destabiliz-
ing) cation–π binding pairs. However, these previous studies did not
account for the electrostatic repulsion between closely spaced
cation–π binding groups, which we hypothesize is an important
general effect at interfaces and in the interiors of folded proteins.
As a result, to gain insight into the impact of electrostatic correlations
between cation–π binding pairs on the energetics of cation–π
interactions is of practical relevance.
In the context of engineered biomaterials, cation- and aromatic-
rich sequences are prevalent in the adhesive proteins of several
marine organisms, including mussels
17
, sandcastle worms
24
and
barnacles
25
. Many researchers have sought to translate these
protein sequences into synthetic, bioinspired adhesives by focusing
predominantly on the role of the catecholic functional group
3,4-dihydroxyphenylalanine (Dopa)
17,26–30
. However, these same
studies
26,28,30
indicate that a reliance on Dopa alone is unrealistic
for engineering an effective wet adhesion in underwater environ-
ments. Further, Dopa is conspicuously sparing or non-existent in
the highly adhesive proteins of some marine organisms, such as
green mussels
8
and barnacles
25
, whereas Dopa is prevalent in
non-adhesive proteins, such as the plaque-coating proteins of
some marine mussels
17
. More recent studies
31,32
identified a possible
synergistic relationship between Dopa and cationic amino acids; yet
none of these studies systematically explored how changes to the
1
Materials Department, University of California, Santa Barbara, California 93106, USA.
2
Materials Research Laboratory, University of California, Santa
Barbara, California 93106, USA.
3
Department of Molecular, Cell and Developmental Biology, University of California, Santa Barbara, California 93106, USA.
4
Department of Chemical Engineering, University of California, Santa Barbara, California 93106, USA.
*
e-mail: jacob@engineering.ucsb.edu; waite@lifesci.ucsb.edu
ARTICLES
PUBLISHED ONLINE: 13 FEBRUARY 2017 | DOI: 10.1038/NCHEM.2720
NATURE CHEMISTRY | VOL 9 | MAY 2017 | www.nature.com/naturechemistry 473
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
aromatic molecular structure impact the strength of cation–π
interactions in adhesive films.
In this work, we test the hypothesis that cation–π interactions
may provide a broader molecular motif that can be used to imbue
peptide-based materials with robust underwater cohesion and
pursue fundamental insights into the energetics of interfacial
cation–π interactions. To test this proposal, we designed a series
of lysine- and aromatic-rich peptides and used nanoscale force
measurements to determine quantitatively the cohesion-interaction
strength present within films composed of each peptide.
Results
Peptide design. All four of the peptides are composed of a sequence
of 36 amino acids, with the numbers and locations of the glycine
(Gly), lysine (Lys) and cysteine (Cys) residues conserved (Fig. 1).
The locations of the aromatic residues and leucine (Leu)
hydrophobic control (X) were also conserved. The four peptides
differ only by progressive hydroxylation of the aromatic residue in
three of the peptides: phenylalanine (Phe), tyrosine (Tyr) and
3,4-dihydroxyphenylalanine (Dopa). The fourth peptide is a Leu
analogue to test whether the strong adhesion forces we measure result
from nonspecific hydrophobic and/or hydrogen-bonding interactions.
The overall sequence is inspired by a Lys- and Dopa-rich
sequence of 16 amino acids that is present in the mussel foot
protein mefp-5, a strongly adhesive mussel foot protein prominently
featured at the mussel adhesive plaque-solid interface
31,33
. Dopa is
also the dominant aromatic residue in many mussel foot proteins,
and thus we refer to the Dopa peptide as a mussel-mimetic
peptide sequence. Although some mussel proteins also contain an
appreciable Tyr content, Phe is conspicuously deficient in marine
mussel adhesive proteins.
Mica was selected as the substrate material because primary
amines, like Lys, strongly bind to the surface of mica via ion exchange
with the K
+
ions present at the surface of single-crystalline mica
9,34
.
Although individual Lys–mica Coulomb bonds are weaker than
covalent interactions, the peptides form multiple Lys–mica bonds
with an energy of between 3 and 5 k
B
T each. The peptides
irreversibly adsorb to mica under the conditions tested, so the
adhesion forces across the confined peptide films are proportional
to the cohesion interactions between peptide molecules.
Analysis of force–distance profiles. When mica surfaces are
approached and separated in the background buffer solution (no
peptide) of 100 mM acetic acid and 250 mM KNO
3
(hereafter
‘high salt conditions’) using the surface forces apparatus (SFA)
35
(Fig. 2), the forces are reversibly repulsive on both approach and
separation (Fig. 3). The surface separation distance, D,isdefined
with respect to the hard contact of two mica surfaces in an inert
nitrogen atmosphere, where D = 0 nm. The non-monotonic
features measured for surface separation distances, D, of less than
2 nm are consistent with other measurements across highly
concentrated electrolyte solutions and result from the ordering
and/or correlation of ions between the mica surfaces.
Although the focus of this work is on the adhesion forces of pep-
tides in high salt conditions, force–distance curves were also
measured for the Tyr and Phe peptides in 100 mM acetic acid sol-
utions with variable concentrations of KNO
3
. These experiments
are presented in Supplementary Figs 3 and 4 and demonstrate
that cohesion forces systematically decrease as the background sol-
ution salinity is increased. These experiments illustrate that the
deposition salinity may need to be matched to the salinity present
in the application to enable the maximum adhesion performance.
All four of the peptides form diffusive, hydrated surface films that
are approximately 3–5 nm in thickness when solution deposited
onto a single mica surface under high salt conditions (Fig. 3). The
(positive) repulsive forces that are measured when bringing
peptide-coated surfaces together originate from compressing the
diffusive films into tightly packed configurations (Fig. 2). The
ranges of these repulsive forces differ by 3–4 nm among different
experiments. As discussed in Methods, unavoidable changes in
the optical path occur when the surfaces are removed from the
SFA to deposit peptides after calibrating the mica thickness,
which results in up to a 2 nm uncertainty in the peptide-film thick-
nesses. Critically, the magnitudes of the repulsive forces and the dis-
tances over which one can compress the films is characteristic of the
peptide molecular structure and solution salinity, and remains inde-
pendent of variability in the measured film thickness. During
compression, the slope of the force–distance profile is roughly pro-
portional to the compressibility of the peptide films, so we conclude
that the three aromatic peptide films exhibit similar mechanical
properties during compression; the Leu peptide formed
less-compressible films.
HN–R
HN–R HN–R
PEP-X:
CGXKGXKXXGKGKKXXXK
CGXKGXKXXGKGKKXXXK
N-Terminal acetylation C-Terminal amidation
a
b
Leucine (L) Lysine (K)
Phenylalanine (F) Tyrosine (Y) 3,4-dihydroxyphenylalanine
(Dopa)
HN–R
HN–R
R–NH
R–HN
H
3
N
+
O
O
N–R
H
O
N–R
H
N–R
H
O
HO
O
HO
HO
Figure 1 | Sequences and molecular structures of the peptides studied.
a,b, Each of the four peptides included one of the amino acids illustr ated in
b, incorporated in the sequence locations marked by a purple ‘X’ in a.
The Lys residues are conserved in each peptide sequence and are marked
in blue to emphasize the positive charge of Lys at a pH of 2.5.
Peptide-
coated
mica
Force, F
R
Piezo
Buffer
D
Mica
Surfaces separated
a
SFA overview
Mica
1–4 nm
Mica
Surfaces compressed
Contact region between surfaces
b
Mica
Figure 2 | Schematic of the SFA set-up and illustration of the surface–
peptide–surface interface studied. a, SFA experimental set-up. The peptide
was solution deposited onto mica surfaces before placing it into the SFA,
and all the experiments were performed in a pH 2.5 buffer solution of 100 mM
acetic acid and 250 mM KNO
3
, without additional dissolved peptide.
b, When the two surfaces are compressed into hard contact, a multilay er
peptide film is confined between the two surfaces. When the surfaces are
separated, failure occurs within the peptide film, which means that the
measured work of adhesion is proportional to the intermolecular interactions
between peptide molecules.
ARTICLES
NATURE CHEMISTRY DOI: 10.1038/NCHEM.2720
NATURE CHEMISTRY |VOL9|MAY2017|www.nature.com/naturechemistry474
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
In contrast, the adhesion forces that are measured when separ-
ating the mica surfaces exhibit a significant dependence on the
peptide molecular structure (Fig. 3). For the Leu peptide (hydro-
phobic control) at high salt conditions, the work of adhesion was
measured as 1.3 ± 0.4 mJ m
–2
and the attractive force extended
over a distance of 1–2 nm. The range and magnitude of this force
is consistent with the pulling apart of weak, nonspecific cohesion
interactions that probably result from hydrophobic interactions
and intermolecular hydrogen bonding along the peptide backbones.
All three aromatic peptides exhibit dramatically increased
adhesion relative to the Leu control (Fig. 3). As a pronounced
strong adhesion is measured only when the peptides contain both cat-
ionic (Lys) and aromatic moieties, and this adhesion is strongly
impacted by changing the solution salinity, we conclude that the
most-probable source of the strong adhesion measured in the three
aromatic peptides is intermolecular cation–π interaction. Solid-state
NMR spectroscopy corroborates this conclusion, as discussed below.
The work of adhesion measured for the mussel-mimetic Dopa
peptide, 3.6 ± 0.4 mJ m
–2
, and for the Tyr peptide, 4.0 ± 0.6 mJ m
–2
,
are similar, with the Tyr peptide yielding a work of adhesion that
slightly exceeds that of the mussel-mimetic Dopa peptide. This
observation indicates that the interfacial Lys–Tyr and Lys–Dopa
cation–π complexation energies are similar. This result also
implies that the cation–π interaction between Dopa and Lys is the
dominant mechanism that mediates molecular cohesion in
Dopa- and Lys-containing proteins, peptides and synthetic molecules.
Unexpectedly, the measured work of adhesion for the Phe
peptide is 10 ± 3 mJ m
–2
, which is more than double that measured
for the Tyr and Dopa peptides. From this result, we conclude that
the Phe–Lys cation–π complexation energy is surprisingly strong
compared with the Tyr–Lys and Dopa–Lys complexation energies.
Solid-state NMR spectroscopy. Solid-state NMR spectroscopy
complements the SFA measurements by establishing the
molecular proximities and orientations of lysine and aromatic
residues in solid (non-crystalline) peptides as a model of the
intermolecular interactions that occur in peptide-rich confined
films. Two-dimensional (2D)
13
C{
1
H} heteronuclear correlation
(HETCOR) experiments use through-space dipolar couplings to
correlate the isotropic chemical shifts of nearby (<1 nm)
1
H and
13
C nuclei. The solid-state 2D
13
C{
1
H} HETCOR spectrum of Tyr
in Fig. 4 shows many well-resolved correlations that arise from
dipolar-coupled
13
C and
1
H nuclei of the Tyr peptide. Most
intensity correlations in this 2D spectrum originate from directly
bound intraresidue
1
H and
13
C nuclei, which allows their
assignments to specific
1
H and
13
C moieties of the lysine,
tyrosine, cysteine and glycine residues of the Tyr peptide. These
resonance assignments are corroborated by the solid- and
solution-state NMR spectra of neat peptides and polypeptides
reported in the literature
36–40
.
Importantly, the 2D
13
C{
1
H} HETCOR spectrum (Fig. 4) of Tyr
also includes intensity correlations that result from inter-residue
interactions, specifically among the lysine and tyrosine side
chains. In particular,
1
H signals at ∼6.6 ppm of the aromatic
1
H
moieties of the tyrosine residues are correlated with
13
C signals
between 20 and 30 ppm (Fig. 4, red arrows) assigned to the alkyl l
and m moieties of the lysine residues. These correlations unambigu-
ously establish the close proximities of the alkyl groups of lysine
with the aromatic tyrosine moieties.
Furthermore, the
13
C signals at ∼40 ppm (Fig. 4, red shaded
region) from the alkyl j
+
moieties are correlated with
1
H signals
at 6.9 ppm of the ε
+1
H moieties of the protonated lysine amide
groups. By comparison, a 2D
13
C{
1
H} HETCOR spectrum
(Supplementary Fig. 5) collected from the Leu peptide without aro-
matic residues under otherwise identical conditions shows that the
intensity correlation from the same j
+13
C and ε
+1
H moieties occurs
at
∼7.5 ppm in the
1
H dimension. The large displacement (0.6 ppm)
of this correlated intensity to a lower frequency in the spectrum of
the Tyr peptide (Fig. 4) indicates that the ε
+1
H groups experience
ring-current effects that are associated with a substantial fraction
of these
1
H moieties positioned near the centres of the aromatic
rings of the tyrosine side chains, as shown schematically in Fig. 4.
Such a configuration of the tyrosine and lysine side chains is consist-
ent with cation–π interactions among the protonated lysine and
tyrosine side chains and corroborates the analyses of the SFA data.
Discussion
The adhesion forces that we measured for all three aromatic- and
Lys-containing peptides are consistent with the work of adhesion
measured previously for various Dopa-containing mussel adhesive
proteins between mica surfaces, under similar conditions of salinity
and pH (ref. 17). Surprisingly, the Phe peptide exhibits an adhesive
performance between mica surfaces that exceeds the performance of
the mussel-mimetic Dopa peptide and even rivals that of the
ab
Force/radius, F/R (mN m
–1
)
Phe
Buffer
Leu
Tyr
Dopa
Distance, D (nm)
0123456
10
1
Compression
Force/radius, F/R (mN m
–1
)
10
0
–10
–20
–30
–40
–50
Tyr:
4.0 ± 0.6 mJ m
–2
Phe:
10 ± 3 mJ m
–2
Out
Buffer
Dopa: 3.6 ± 0.4
mJ m
–2
Leu: 1.3 ± 0.4
mJ m
–2
Separation
Work of adhesion (mJ m
–2
)
0
–2
–4
–8
–6
2
–10
Distance, D (nm)
0123456
Figure 3 | Representative force–distance data measured for peptides between mica surfaces. a, Representative force–distance profilesmeasuredwhentwo
mica surfaces are brought together in 100 mM acetic acid and 250 mM KNO
3
. Measurements were also performed under variable solution salinities and are
presented in Supplementary Figs 4 and 5. Positive forces are repulsive and negative forces are attractive. The black data were measured in the absence of
adsorbed peptide, and each of the coloured curves corresponds to an experiment in which the peptide was adsorbed onto a single mica surface.
b, Representative profiles measured when mica surfaces are separated in 100 mM acetic acid and 250 mM KNO
3
. The black force–distance profile measured
in the absence of peptide exhibits repulsive behaviour, which supports that the adhesion measured in the presence of peptide films results from peptide
intermolecular cohesion. The aver age work of adhesion and associated uncertainty quoted on the plot was obtained from at least ten different force–distance
profiles for each peptide. Small variations in peptide-film thicknesses were measured between different experiments, and these variations exhibited no
dependence on the peptide molecular structure. Importantly, the work of adhesion does not exhibit a systematic dependence on the film thickness, which
implies that the cohesion interactions are independent of minor variations in the film thickness.
NATURE CHEMISTRY DOI: 10.1038/NCHEM.2720
ARTICLES
NATURE CHEMISTRY | VOL 9 | MAY 2017 | www.nature.com/naturechemistry 475
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
most-adhesive Dopa-containing mussel foot protein tested to date,
mefp-5 (ref. 33)
Nevertheless, Dopa is a biologically important functional group
that exhibits diverse chemical reactivity
17
. For example, Dopa can
chelate multivalent ions
41,42
and exhibits a propensity to autoxidize
and irreversibly crosslink at neutral-to-basic pH conditions
43,44
.
Certainly, this reactivity can be used to advantage. Unfortunately,
in many circumstances the oxidation state of Dopa cannot be con-
trolled easily
43–45
, which leaves the adhesive properties of Dopa-
containing molecules compromised by premature autoxidation.
Our results suggest that molecules incorporating Lys and a
balance of both Dopa and chemically stable Phe could provide an
attractive alternative for developing underwater adhesives, hydrogel
binding groups and other applications that involve molecular
cohesion in harsh oxidizing environments.
Our results also provide additional evidence to support the
importance of cation–π interactions in marine bioadhesion. Many
mussel proteins comprise sequences that are rich in both Lys and
Dopa residues
17
. Recently, the adhesion of synthetic biomimetic
small-molecule monolayers to mica surfaces was shown to depend
critically on the synergy between Dopa and Lys functional
groups
32
, with the conclusion that the primary role of Lys is to
eject hydrated cations from mica surfaces to enable Dopa-surface
bidentate hydrogen bonding. This prior study
32
proposed that an
analogous effect occurs in the larger adhesion proteins utilized by
marine mussels. This interpretation predicts that surface-binding
interactions between Dopa-containing peptides and mica surfaces
should significantly exceed those in Tyr- or Phe-based peptides,
whereas we observed similar adhesion forces for the Dopa and
Tyr peptides, and increased adhesion in the Phe peptide.
We rationalize this observation by noting that Maier et al.
32
studied monolayers of a small molecule, for which adhesive
failure necessarily occurs at the molecule–mica interface. Here,
the failure plane is shifted out into the peptide film, which results
in cohesive failure, as is the case for most practical adhesives com-
posed of larger molecules, such as peptides and proteins, that
assemble to encapsulate particles and/or cover surface heterogene-
ities
11–13,46
. For large molecule multilayers that strongly bind to sur-
faces through multiple parallel covalent bonds, hydrogen bonds
and/or strong ionic bonds, molecule–molecule interactions are
often weaker (and/or more transient) than molecule–surface inter-
actions. This shifts failure planes away from surfaces and into the
films
13
, which renderers the overall adhesive performance
critically dependent on intermolecular cohesion.
Thus, the molecule–surface binding force and/or energy of the
mussel-mimetic Dopa peptide may exceed the surface binding of
the Phe peptide; this does not contradict the observation that the
cation–π-mediated cohesion in Phe significantly exceeds that of
Dopa. Furthermore, these results can be explained without invoking
bidentate Dopa hydrogen bonding because a combination of Lys
electrostatic interactions and/or peptide-backbone hydrogen
bonds appear to be sufficient for strong peptide –surface binding.
Nevertheless, prior evidence
31,32,44,47
indicates that bidentate hydro-
gen bonding should be important whenever the film-failure plane is
located at the molecule–surface interface, especially in the absence
of Lys residues.
With this in mind, we address the sometimes contradictory con-
clusions as to the importance of Dopa bidentate hydrogen bonding
for promoting underwater adhesion. Specifically, several groups
(including ours)
31–33
previously concluded that Dopa-mediated
bidentate hydrogen bonding is critical to enable mussel proteins
to achieve a strong underwater adhesion. Many of these studies con-
trolled for the role of hydrogen bonding by chemically oxidizing
Dopa to dopaquinone, which demonstrated a corresponding
decrease in adhesion. However, dopaquinone is a reactive functional
group that can induce a wide range of chemical and/or confor-
mational changes within protein and peptides
17
. Thus, oxidizing
Dopa to dopaquinone does much more than remove the opportu-
nity for bidentate hydrogen bonding.
Recently, there have been efforts
10,31
to compare the adhesion of
mussel-mimetic peptides and recombinant proteins that incorpor-
ate Tyr in peptides analogous to the Dopa peptides to test the
impact of bidentate hydrogen bonding. Peptide adhesion was
observed to be similar for the Dopa and Tyr functionalities in pep-
tides that contained signifi cant numbers of cationic residues
10
,in
agreement with the current study. In contrast, underwater adhesion
was seen to depend on the presence of bidentate Dopa hydrogen
binding in peptides that lacked cationic residues
31
. Notably, the
adhesion of non-cationic sequences was observed to be significantly
lower than that measured for positive (cationic) peptides that are
rich in aromatic groups
31
.
Further, many prior studies did not focus on establishing
whether films fail through adhesive or cohesive mechanisms,
which leaves open the possibility that film-failure mechanisms
may also play a large role in determining the importance of biden-
tate hydrogen bonding. For example, bidentate hydrogen bonding
200 180 160 140 120 100 80 60 40 20 ppm
11
10
9
8
7
6
5
4
3
2
1
ppm
m,k,l
b,c,d,e
γ
n
α
13
C
1
H
1
H single pulse
13
C{
1
H} CP MAS
m
k
l
j
+
g
n
α
a
γ a
b
d
c
e
f
g
e,d,f
b,c
i
h
g,j
+
/j
neut.
j
neut.
j
+
/j
neut.
ε
+
/ε
neut.
I
k
m
6.9 ppm
ε
+
ε
neut.
NH
R
O
NH
R
SR
N
H
HN
O
HO
R
R
NH
R
O
H
NH
R
NH
R
O
HN R
H
3
N
+
HO
H
3
N
H
H
H
H
H
H
+
Figure 4 | Solid-state 2D
13
C{
1
H} HETCOR MAS NMR spectrum acquired
from bulk Tyr peptide with a 1D
13
C{
1
H} CP MAS NMR spectrum along
the top horizontal axis and a single-pulse
1
HMASNMRspectrumalong
the left vertical axis. Red arro ws indicate the intensity correla tions that
result from the close proximities (<1 nm) of the aromatic b–e
1
Hmoietiesof
the tyrosine residues (purple letters) and alkyl l and m
13
Cmoietiesofthe
lysine side chains (blue letters). The superscript ‘+’ denotes chemical shifts
associat ed with protona ted (posi tive) lysine residues, whereas the superscript
‘neut.’ denotes chemical shifts associated with neutral lysine residues.
The intersection of the shaded red bands indicates a correlated intensity that
arises from the proximate alkyl j
13
C moieties and protonated amide ε
+1
H
moieties of lysine residues, which resonate approxima tely 0.6 ppm to a
lower frequency in the
1
H dimension compared with a Leu sample measured
under otherwise identical conditions (Supplementary Fig. 5). Such a
displacement is consistent with ring-current effects tha t would result from a
configuration of the lysine and tyrosine side chains shown schematically in
the inset, associated with inter-residue contact through cation–π electron
interactions. All the NMR measurements were conducted at 11.74 T under
10 kHz MAS conditions at 0 °C.
ARTICLES
NATURE CHEMISTRY DOI: 10.1038/NCHEM.2720
NATURE CHEMISTRY |VOL9|MAY2017|www.nature.com/naturechemistry476
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
