Controlled stacking and unstacking of peripheral chlorophyll units drives the spring-like contraction and expansion of a semi-artificial helical polymer

Article abstract of DOI:10.1002/chem.201203569

Developing new strategies for controlling polymer conformations through precise molecular recognition can potentially generate a machine-like motion that is dependent on molecular information - an important process for the preparation of new intelligent nanomaterials (e.g., polymer-based nanomachines) in the field bordering between polymer chemistry and conventional supramolecular sciences. Herein, we propose a strategy to endow a helical polymer chain with dynamic spring-like (contraction/expansion) motion through the one-dimensional self-assembly (aggregation/disaggregation) of peripheral amphiphilic molecules. In this developing system, we employed a semi-artificial helical polysaccharide presenting peripheral amphiphilic chlorophyll units as a power device that undergoes contractive motion in aqueous media, driven by strong π-π interactions of its chlorophyll units or by cooperative molecular recognition of bipyridyl-type ligands through pairs of chlorophyll units, thereby converting molecular information into the regulated motion of a spring. In addition, this system also undergoes expansive motion through coordination of pyridine. We anticipate that this strategy will be applicable (when combined with the established wrapping chemistry of the helical polysaccharide) to the development of, for example, drug carriers (e.g., nano-syringes), actuators (stimuli-responsive films), and directional transporters (nano-railways), thereby extending the frontiers of supramolecular science. Copyright

Full text of DOI:10.1002/chem.201203569

DOI: 10.1002/chem.201203569
Controlled Stacking and Unstacking of Peripheral Chlorophyll Units Drives
the Spring-Like Contraction and Expansion of a Semi-Artificial Helical
Polymer
Munenori Numata,*[a] Daiki Kinoshita,[a] Naoya Hirose,[a] Tomohiro Kozawa,[a]
Hitoshi Tamiaki,[b] Yoshihiro Kikkawa,[c] and Masatoshi Kanesato[c]
1592
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 1592 – 1598

FULL PAPER
Abstract: Developing new strategies
for controlling polymer conformations
through precise molecular recognition
can potentially generate a machine-like
motion that is dependent on molecular
information—an important process for
the preparation of new intelligent
nanomaterials (e.g., polymer-based
nanomachines) in the field bordering
between polymer chemistry and con-
ventional supramolecular sciences.
gregation) of peripheral amphiphilic
molecules. In this developing system,
we employed a semi-artificial helical
polysaccharide presenting peripheral
thereby converting molecular informa-
tion into the regulated motion of a
spring. In addition, this system also un-
dergoes expansive motion through co-
ordination of pyridine. We anticipate
that this strategy will be applicable
(when combined with the established
wrapping chemistry of the helical poly-
saccharide) to the development of, for
example, drug carriers (e.g., nano-sy-
ringes), actuators (stimuli-responsive
films), and directional transporters
(nano-railways), thereby extending the
frontiers of supramolecular science.
amphiphilic chlorophyll units as
a
power device that undergoes contrac-
tive motion in aqueous media, driven
by strong p–p interactions of its chloro-
phyll units or by cooperative molecular
recognition of bipyridyl-type ligands
through pairs of chlorophyll units,
Herein, we propose
a strategy to
endow a helical polymer chain with dy-
namic spring-like (contraction/expan-
sion) motion through the one-dimen-
sional self-assembly (aggregation/disag-
Keywords: chlorophyll
·
helical
polymers · nanomachines · polysac-
charides · supramolecular chemistry
Introduction
larger movement, being relatively rare outside the biological
domain.[2] Unlike most synthetic polymers, enzymes have
the ability to convert chemical energy into continuous me-
chanical motion. The design of artificial polymers exhibiting
controllable movement has remained a difficult challenge.
In this paper, we propose an alternative approach toward
such functional materials through the preparation of semi-
artificial biopolymers that combine the regular conformation
of a helix-forming natural polymer with the dynamic motion
of small molecules through supramolecular aggregation. We
prepared a semi-artificial curdlan (a b-1,3-glucan) featuring
a dendritic amphiphilic Zn-chlorophyll unit attached to each
glucose residue (CUR-Chl); this system acts as an actuator
driven by supramolecular aggregation within the polymer
chain.[3] Because native curdlan adopts a right-handed heli-
cal structure with a helical pitch of approximately 1.8 nm,
with six glucose residues per pitch, the attached chlorophyll
units in CUR-Chl would, therefore, be aligned one-dimen-
sionally along the long axis at 1.8 nm intervals (Figure 1).
The structural feature of curdlan is comparable to that of
native DNA. It is well-accepted that the helical pitch of
DNA can be tuned by the physiological conditions; for ex-
ample, the axial expansion of DNA strand can be induced
by complexation with DNA-binding protein.[4] The curdlan
main-chain also has the potential to accommodate such con-
formational changes due to their flexible glycoside linkage.
Actually, we have already found that curdlan can wrap vari-
ous nanomaterials including carbon nanotubes, conductive
polymers, nanoparticles, and change its helical pitch depend-
ing on size and shape of these nanomaterials.[3b] Although
native curdlan has no intrinsic enzyme-like ability to gener-
ate mechanical motion, sequential aggregation and dissocia-
tion of the peripheral chlorophyll units of CUR-Chl driven
by coordination and dissociation of ligands, led to contrac-
tion and expansion (spring-like motion), respectively, of the
helical structure of the main chain.
The dynamic motion of proteins is an essential aspect of
many biological events in cells (e.g., molecular recognition,
storing of molecular information, and the transportation and
chemical conversion of molecules). Several cytoplasmic pro-
teins, including myosins, kinesines, and dynasines, are char-
acterized by their dynamic motion among various conforma-
tions, generally triggered by the hydrolysis of ATP or GTP
as chemical fuel. The precise motion of such proteins means
that they can be regarded as natural machines operating on
the nanometer scale.
Taking inspiration from Natureꢁs molecular machines,
much effort has been exerted to mimic muscular contraction
in artificial systems by taking advantage of elaborately de-
signed motion (e.g., rotations, translations, spring-like
motion) in molecular machinery.[1] Despite significant ad-
vances in this area, most molecular machines have been
based on the motion of a single molecule, with regular con-
formational changes of polymers, potentially leading to
[a] Prof. Dr. M. Numata, D. Kinoshita, N. Hirose, Dr. T. Kozawa
Department of Biomolecular Chemistry
Graduate School of Life and Environmental Sciences
Kyoto Prefectural University, Shimogamo, Sakyo-ku
Kyoto 606-8522 (Japan)
Fax : (+81)75-703-5132
E-mail: numata@kpu.ac.jp
[b] Prof. H. Tamiaki
Graduate School of Life Sciences
Ritsumeikan University
Kusatsu, Shiga 525-8577 (Japan)
[c] Y. Kikkawa, Dr. M. Kanesato
Electoronics and Photonics Research Institute
National Institute of Advanced Industrial Science
and Technology, Tsukuba Center 4, 1-1-1 Higashi
Tsukuba, Ibaraki 305-8562 (Japan)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201203569.
Chem. Eur. J. 2013, 19, 1592 – 1598
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1593

M. Numata et al.
also influence the assembly of the peripheral chlorophyll
units of CUR-Chl in a similar manner. Next, we investigated
whether cooperative aggregation (dissociation) of the pe-
ripheral chlorophyll units could result in contraction (expan-
sion) of the covalently linked curdlan main chain. The UV/
Vis spectrum of CUR-Chl in DMSO (0.046 mgmLÀ1, 2.91ꢂ
10À2 mm/repeating unit) featured sharp absorption bands at
429.1 and 654.7 nm, comparable with those of the monomer-
ic chlorophyll units of Chl-X in this solvent; thereby sug-
gesting that CUR-Chl adopted a single-helix conformation
in DMSO.[6] As DMSO is good solvent for curdlan as well
as for peripheral chlorophyll units, CUR-Chl would adopt
an extended helical conformation, of which the helical pitch
is approximately 1.8 nm. In the present study, we always
started a series of experiments from this form extended in
DMSO. When we added distilled H2O to the DMSO solu-
tion, we observed a significant hypochromic effect of the
chlorophyll units, accompanied by slight redshifts of the
Soret and Q band to 433.4 and 662.0 nm, respectively; the
circular dichroism (CD) intensities corresponding to these
bands decreased accordingly (Figure S3, the Supporting In-
formation). As shown in Figure S2 (the Supporting Informa-
tion), the spectra of CUR-X did not display such variations.
The strong hypochromic effect evident in the UV/Vis spec-
tra suggested that columnar stacking among the peripheral
chlorophyll units occurred along the main chain, in which
À
the direct Zn O coordination interactions, that is, the for-
mation of the J-aggregate, is suppressed as a result of their
immobilization on the curdlan chain. Based on our previous
studies of related CD spectra, we could estimate whether
the semi-artificial curdlan derivative presenting peripheral
dye units adopted a single- or triple-helix structure; upon
forming a triple-helix, the CD intensity of the dye units
would increase dramatically as a result of strong p–p stack-
ing interactions among them.[7] In this present case, howev-
er, we did not observe such a dramatic change in CD inten-
sity, even at higher compositions of H2O, suggesting that the
steric bulk of the peripheral dendritic units caused CUR-
Chl to retain a single-helix conformation. Therefore, the p–
p stacking interactions suggested in the UV/Vis and CD
spectra presumably occurred in an intrapolymer manner,
leading to a conformational change of the curdlan main-
chain.
We used atomic force microscopy (AFM) to gain some in-
sight into the morphological changes of CUR-Chl that oc-
curred upon addition of H2O. Sharp peaks in the UV/Vis
spectra and the weak CD intensity suggested that CUR-Chl
adopted an extended fibrous structure in DMSO.[8] Fig-
ure 2a reveals, however, that the addition of H2O to the
DMSO solution caused CUR-Chl to undergo a dramatic
morphological change to form rod-like structures (contract-
ed fibers) having a diameter (height) of approximately
1.5 nm and a length of 30–50 nm. Assuming an ideal helical
structure, we would expect the length of CUR-Chl to be ap-
proximately 200 nm;[9] the considerably shorter length ob-
served for the rod-like structures is consistent with contrac-
tion of the main chain of CUR-Chl through stacking of its
Figure 1. a) Chemical structure of CUR-Chl and b) potential ligands for
coordination with its chlorophyll units. c) Schematic representation of the
controllable motion (contraction/expansion) of CUR-Chl through molec-
ular recognition of its peripheral moieties.
Results and Discussion
We synthesized CUR-Chl from the terminal alkyne-func-
tionalized chlorophyll derivative Chl-X and the azido-func-
tionalized curdlan derivative CUR-N3 according to our pre-
viously reported procedures (Scheme S1 and Scheme S2, the
Supporting Information). We also employed Chl-X to inves-
tigate the fundamental self-assembly behavior of the periph-
eral chlorophyll units of CUR-Chl. The UV/Vis spectra of
solutions of Chl-X in DMSO/H2O at different compositions
(Figure S2, the Supporting Information) revealed that it ex-
isted as a monomer in DMSO, whereas its Soret and Q
bands both were shifted to longer wavelengths (from 429.0
to 431.1 nm and from 655.2 to 729.0 nm, respectively) upon
the addition of H2O, accompanied by a significant hypochro-
mic effect. These shifts in absorption peaks are indicative of
À
the formation of J-aggregates, stabilized through Zn O co-
ordination and p–p stacking interactions.[5] Accordingly, we
suspected that variations in the DMSO/H2O ratio would
1594
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 1592 – 1598

Semi-Artificial Helical Polymer
FULL PAPER
with the helical pitch (1.8 nm), whereas DABCO is consider-
ably shorter. Firstly, we mixed DPE with the DMSO solu-
tion containing CUR-Chl and H2O was added to the resul-
tant DMSO solution, adjusting DMSO/H2O=5:95 (v/v), ex-
pecting to suppress the stacking of peripheral chlorophyll
units. Figure 3a and b reveal the changes in the UV/Vis and
CD spectra. Upon increasing the [DPE]/[chlorophyll] ratio
from 0 to 100, we observed hyperchromic effects for both
the Soret and Q bands, with accompanying slight blueshifts
(ca. 14 nm). In addition, a new signal, assignable to DPE,
appeared near 300 nm. These variations in the UV/Vis spec-
tra of CUR-Chl are comparable with those observed after
the addition of pyridine, which is indicative of the suppres-
sion of the stacking of the chlorophyll units. The CD spectra
provided further evidence for the chlorophyll units interact-
ing with the DPE ligands; in Figure 3b, we observe a more-
intense split-type CD signal in the Q band region (positive
Figure 2. AFM images of CUR-Chl samples was prepared from aqueous
solutions in the a) absence and b) presence of pyridine ([pyridine]/[chlor-
ophyll]=80); mica surface, dynamic mode.
peripheral chlorophyll units. In some cases, the stacking oc-
curred in a random manner with accompanying intrapoly-
mer globule aggregation of the main chain; the globule ag-
gregate structures can be also confirmed in the same AFM
image in addition to the rod-like structures.
Because pyridine has the potential to suppress the stack-
À
ing of chlorophyll units through Zn N coordination, we
added it to the aqueous solution containing the contracted
CUR-Chl, expecting to reform the original helical structure.
Indeed, upon addition of pyridine ([pyridine]/[chlorophyll]=
80), we observed a hyperchromic effect in both the Soret-
and Q band regions, suggesting that the stacking of chloro-
phyll moieties dissociated into “monomeric” states in aque-
ous solution (Figure S3, the Supporting Information).[10] We
suspect that if CUR-Chl could adopt a triple-helix, such a
hyperchromic effect of the chlorophyll units would not
occur, even in the presence of excess pyridine, due to the
strong stacking of the chlorophyll moieties. Therefore, we
believe that CUR-Chl adopted a single-stranded structure in
the presence of pyridine. The AFM image of CUR-Chl after
addition of pyridine reveals many extended fibrous struc-
tures (Figure 2b).[11] The line profile of this image provided
diameters and lengths of approximately 1.0 and 150–180 nm,
respectively, implying that the curdlan main chain existed in
its regular helical structure (Figure S6, the Supporting Infor-
mation). Accordingly, contraction/expansion motion of
CUR-Chl can be achieved through the addition of H2O/pyri-
dine; CUR-Chl behaves as a nanometer-sized spring that
can be contracted and expanded through supramolecular
stacking and unstacking of its peripheral chlorophyll units.
On the basis of the fundamental motion of contraction
and expansion of CUR-Chl in DMSO upon the addition of
H2O and pyridine, respectively, we wondered whether diaza-
type ligands could control the degree of contraction through
cooperative interactions between each ligand and two chlor-
ophyll units. We tested the effects of four ligands: 1,4-
diazabicycloACHTUNGTRENNUNG[2.2.2]octane (DABCO), 4,4ꢁ-bipyridyl (Bipy),
Figure 3. a,c) UV/Vis and b,d) CD spectra of CUR-Chl at various
1,2-di(4-pyridyl)ethylene (DPE), and 4,4’-(p-phenylenebisvi-
nylene)bispyridine (PVP), all with different molecular
lengths (Figure 1), expecting a tunable contraction of the
structures based on molecular information. Among these li-
gands, the molecular lengths of PVP are almost comparable
[ligand]/[chlorophyll] ratios ([CUR-Chl]=0.046 mgmLÀ1
; DMSO/H2O
5:95; 0.5 cm cell, RT). Ligand: a,b) DPE, c,d) PVP; black dotted lines in-
dicate the spectra of CUR-Chl without ligands. e) Absorbance at 660 nm
in (c), plotted with respect to the [PVP]/[chlorophyll] ratio. f) CD intensi-
ty at 375 nm in (d), plotted with respect to the [PVP]/[chlorophyll] ratio.
Chem. Eur. J. 2013, 19, 1592 – 1598
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1595

M. Numata et al.
at 671.5 nm, negative at 653.5 nm) upon increasing the
[DPE]/[chlorophyll] ratio, which is indicative of direct p–p
stacking of adjacent chlorophyll units. This type of CD
signal did not appear after the addition of pyridine, suggest-
ing that cooperative coordination between DPE and two
chlorophyll units was occurring on the surface of CUR-Chl.
Because the molecular length of DPE (0.9 nm) is shorter
than the helical pitch of CUR-Chl (1.8 nm), cooperative co-
ordination would lead to slight distortion of the main chain.
We suspect that the split-type CD signal arose from strong
p–p stacking between the pairs of chlorophyll p-surfaces
that did not participate in coordination with DPE; the con-
traction of the main chain through cooperative coordination
would promote such direct p–p stacking interactions. Nota-
bly, the spectral changes in both the UV/Vis and CD spectra
were almost saturated at a [DPE]/[chlorophyll] ratio of 80,
strongly supporting the view that the interactions occurred
in a nonspecific manner and that thermodynamically stable
complexes were formed through precise molecular recogni-
tion.
stacking interactions that occurred by slight contraction of
the main chain in the presence of DPE or Bipy. Figure 3e
and f present plots of the UV/Vis absorbance and CD inten-
sities at 660 and 375 nm, respectively, with respect to the
[PVP]/[chlorophyll] ratio, revealing that the spectral changes
were saturated at a value of [PVP]/[chlorophyll] of 0.5.
Thus, two chlorophyll units cooperatively captured one PVP
ligand to form a stable CUR-Chl/PVP complex, without
causing distortional strain of the original helical structure.
The higher affinity of CUR-Chl to PVP, compared with
other tested ligands, is due to its comparable molecular
length with the original helical pitch of CUR-Chl. The calcu-
lated structure of the CUR-Chl/PVP complex also support
this view; that is, PVP can be incorporated into the p-space
constructed by two chlorophyll units without any changing
of the original structure of CUR-Chl (see the Supporting In-
formation, Figure S11).
Figure 4 displays AFM images of CUR-Chl in the pres-
ence of various ligands at different [ligand]/[chlorophyll]
ratios (for height-profile analyses, see Figures S6–S9, the
Next, when we employed Bipy (length: 0.7 nm) as the
ligand, which is shorter than DPE, and we observed similar
split-type CD signals (Figure 3b) in the Q band region, sup-
porting our view that coordination with a somewhat shorter
ligand promoted direct p–p stacking on the other p surfaces
(Figure S4, the Supporting Information). On the other hand,
the addition of DABCO, which is almost the same size as
pyridine itself, resulted in only recovery of the CD intensity,
with no split-type CD signal, upon increasing the [DABCO]/
[chlorophyll] ratio from 0 to 100 (Figure S5, the Supporting
Information). Even in this case, however, the hyperchromic
effect of the chlorophyll units was evident in the UV/Vis
spectra, indicative of the dissociation of the stacked chloro-
phyll units, which would allow CUR-Chl to adopt expanded
fibrous structures.
The ligand PVP features two pyridyl termini and has a
length of approximately 1.6 nm, which is comparable with
the helical pitch of native curdlan. In addition, PVP features
an absorption near 350 nm that is distinct from that of the
chlorophyll units, thereby providing additional information
concerning any specific interactions. Figure 3c and d display
UV/Vis and CD spectra, respectively, of CUR-Chl upon in-
creasing the [PVP]/[chlorophyll] ratio from 0.1 to 1.0. In Fig-
ure 3c the intensities of both the Soret and Q bands in-
creased upon addition of PVP, which is consistent with PVP
coordinating to the chlorophyll units. In addition, the inten-
sity of the absorption of PVP at 375 nm increased upon in-
creasing the [PVP]/[chlorophyll] ratio. Notably, induced-CD
(i-CD) from PVP appeared at 320 and 375 nm (Figure 3d),
the intensity of which also increased upon increasing the
[PVP]/[chlorophyll] ratio, implying that PVP coordinated
tightly with the chlorophyll units.[12] On the other hand, we
did not observe a split-type CD band in the Q band region,
in contrast to the CD spectra of the CUR-Chl/DPE and
CUR-Chl/Bipy complexes. This discrepancy suggests that
the main chain of CUR-Chl maintained the original helical
structure after incorporation of PVP, with none of the p–p
Figure 4. AFM images of CUR-Chl in the presence of various ligands:
a) [DABCO]/[chlorophyll]=5; b) [DABCO]/[chlorophyll]=30; c) [DPE]/
[chlorophyll]=5; d) [DPE]/[chlorophyll]=30; e) [PVP]/[chlorophyll]=
0.5; f) [PVP]/[chlorophyll]=30. In image (e), random coil structures are
also present in addition to straight fibers. The dot-like structures in (f)
represent excess PVP ligand that did not interact with CUR-Chl.
1596
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 1592 – 1598

Semi-Artificial Helical Polymer
FULL PAPER
Supporting Information), revealing that the degree of con-
traction of CUR-Chl was strongly dependent on the length
of the ligand as well as its molar ratio. With DABCO, the
shortest ligand, a morphological transition from dot-like
structures to worm-like fibers occurred upon increasing the
[DABCO]/[chlorophyll] ratio from 5 to 30 (Figure 4a,b); at
a [DABCO]/[chlorophyll] ratio of 80, the fibers became fur-
ther entangled (Figure S9, the Supporting Information),
with an average length of approximately 100 nm—shorter
than the native helical structure of CUR-Chl obtained after
the addition of pyridine. Unlike the effect of added pyridine,
the coordination of DABCO to the chlorophyll units would
suppress the free expansion of CUR-Chl, implying that the
degree of the spring-like contraction of CUR-Chl can be
controlled by a diaza-type ligand. Furthermore, the UV/Vis
and CD spectra suggested that not all of the chlorophyll
units participated in the coordination interactions, even in
the presence of excess DABCO; therefore, the partial stack-
ing of the chlorophyll units in the CUR-Chl/DABCO com-
plex presumably led to worm-like morphologies. The effect
of DPE was similar, but, in this case, p–p stacking was facili-
tated after coordination of DPE, as revealed in the UV/Vis
and CD spectra. Indeed, at a [DPE]/[chlorophyll] ratio of
30, only globule structures were present (Figure 4d), with
no evidence for any transformation toward fiber-like struc-
tures, due to strong p–p stacking between adjacent chloro-
phyll units, which would appear randomly along the helix,
resulting in intrapolymer globular aggregation. At a [DPE]/
[chlorophyll] ratio of 80, however, the CUR-Chl/DPE com-
plexes had transformed from globules to extended structures
(Figure S7, the Supporting Information). In contrast, addi-
tion of PVP to the aqueous solution of CUR-Chl resulted in
the formation of straight fibers having an average length of
approximately 200 nm—the largest degree of extension of
spring among our tested ligands. Figure 4e and f reveal that
straight fiber structures were present at [PVP]/[chlorophyll]
ratios of 0.5 and 30, respectively;[13] the average diameter,
determined from the height profiles of the extended fibers
(Figure S8, the Supporting Information), was approximately
1.8 nm—considerably thicker than the fiber structures ob-
tained from the other ligands. Through effective coordina-
tion to the PVP ligands, the chlorophyll units were aligned
in a columnar manner, resulting in a larger cross-section di-
ameter in the AFM image. Among our diaza-type ligands
tested, the perfect transformation to straight and stiff fiber
structures occurred only in the presence of PVP; this mor-
phology reflected the rigid structure of PVP and its quanti-
tative interaction with the chlorophyll units. Overall, we
could control the degree of the contraction of CUR-Chl
springs through precise molecular recognition at its periph-
eral chlorophyll units, with the molecular information from
each ligand being amplified along the helical structure, lead-
ing to the creation of unique ligand-dependent morpholo-
gies. In the present system, p–p stacking of adjacent chloro-
phyll units, triggered by the addition of H2O, was the main
driving force behind the compression of the CUR-Chl
springs, with coordination of pyridine acting as a trigger to
release the strained energy of the main chain. The degree of
the contraction of the spring can be precisely tuned by the
diaza-type ligands depending on their molecular length. This
type of molecular motion, through contraction and expan-
sion of a helical polymer, differs from that found in biologi-
cal systems or other artificial constructions.
Next, we investigated whether the extended CUR-Chl
could be directly transformed to its contracted form through
removal of the ligands from the peripheral p-space. We
tested this idea through the addition of aqueous HCl
(24 equiv with respect to PVP) to a solution of PVP-extend-
ed CUR-Chl ([PVP]/[chlorophyll] ratio=0.5). The absorp-
tion band in the UV/Vis spectrum at 375 nm, assignable to
PVP, shifted to longer wavelength (ca. 10 nm), which is con-
sistent with the protonation of PVP (Figure 5a); in addition,
the Soret and Q bands of the chlorophyll units exhibited hy-
pochromic effects, suggesting that p–p stacking of the chlor-
ophyll units occurred along the helical structure. Further-
more, the addition of aqueous HCl caused the CD intensi-
ties at both the Soret and Q bands to decrease and, notably,
the i-CD signal from PVP to disappear (Figure 5b). These
features strongly support the view that protonation of PVP
led to its removal from the chiral groove constructed by the
chlorophyll units, thereby transforming CUR-Chl to its con-
tracted form. AFM images provided direct evidence for the
contraction process; Figure 5d reveals that all of the extend-
ed CUR-Chl fibers had transformed into globular aggre-
gates (for a larger-area image, see Figure S10, the Support-
ing Information). In a control experiment, when we added
Figure 5. a) UV/Vis and b) CD spectra of CUR-Chl/PVP (extended
form) recorded before (solid blue line) and after (red solid line) addition
of aqueous HCl ([CUR-Chl]=0.023 mgmLÀ1; DMSO/H2O 5:95; 1.0 cm
cell; RT); dotted lines: spectra of CUR-Chl only, before and after addi-
tion of HCl. c,d) AFM images of CUR-Chl/PVP; c) before (extended
form) and d) after (contracted form) addition of aqueous HCl.
Chem. Eur. J. 2013, 19, 1592 – 1598
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1597

M. Numata et al.
ences cited therein; b) B. M. Rosen, C. J. Wilson, D. A. Wilson, M.
Peterca, M. R. Imam, V. Percec, Chem. Rev. 2009, 109, 6275–6540;
c) E. Ohta, H. Sato, S. Ando, A. Kosaka, T. Fukushima, D. Hashi-
zume, M. Yamasaki, K. Hasegawa, A. Muraoka, H. Ushiyama, K.
Yamashita, T. Aida, Nat. Chem. 2011, 3, 68–73.
aqueous HCl to CUR-Chl in the absence of PVP, we did not
observe any changes in either the UV/Vis or CD spectra,
nor any morphological transitions in AFM images.
[3] Selected review: a) M. Numata, S. Shinka, Chem. Commun. 2011,
47, 1961–1975; b) M. Numata, M. Asai, K. Kaneko, A.-H. Bae, T.
Hasegawa, K. Sakurai, S. Shinkai, J. Am. Chem. Soc. 2005, 127,
5875–5884.
[4] a) T. Nishinaka, Y. Ito, S. Yokoyama, T. Shibata, Proc. Natl. Acad.
Sci. USA 1997, 94, 6623–6628; b) T. Nishinaka, A. Shinohara, Y.
Ito, S. Yokoyama, T. Shibata, Proc. Natl. Acad. Sci. USA 1998, 95,
11071–11076.
[5] a) H. Tamiaki, A. R. Holzwarth, K. Schaffner, J. Photochem. Photo-
biol. B 1992, 15, 355–360; b) V. Huber, M. Katterle, M. Lysetska, F.
Wꢃrthner, Angew. Chem. 2005, 117, 3208–3212; Angew. Chem. Int.
Ed. 2005, 44, 3147–3151; c) R. F. Kelley, M. J. Tauber, M. R. Wasie-
lewski, Angew. Chem. 2006, 118, 8147–8150; Angew. Chem. Int. Ed.
2006, 45, 7979–7982; d) V. Huber, M. Lysetska, F. Wꢃrthner, Small
2007, 3, 1007–1014; e) H. Tamiaki, M. Kunieda, Handbook of Por-
phyrin Science 2011, 11, 223–290; f) M. Numata, D. Kinoshita, N.
Taniguchi, H. Tamiaki, A. Ohta, Angew. Chem. 2012, 124, 1880–
1884; Angew. Chem. Int. Ed. 2012, 51, 1844–1848; g) S. Shoji, T. Ha-
shishin, H. Tamiaki, Chem. Eur. J. 2012, 18, 13331–13341.
[6] Curdlan adopts triple helix in water, but it dissociates into single
chain in DMSO. When H2O is added to this DMSO solution, cur-
dlan can retrieve its original triple helix. This unique self-assembling
nature of curdlan has been investigated by us and other groups (see
Ref. [3a], and references cited therein).
Conclusion
We have demonstrated that spring-like contraction/expan-
sion motion of a polymer can be achieved for a semi-artifi-
cial helical polysaccharide presenting a regular 61 helical ar-
rangement of chlorophyll units at its periphery. The aggrega-
tion and dissociation of the peripheral chlorophyll units,
regulated by precise molecular recognition events, endowed
the helical polysaccharide with dynamic spring-like motion.
The degree of elongation was tunable through variation of
the length of the added ligand. In a previous study, we dem-
onstrated the potential wrapping ability of the helical poly-
saccharide; the present dynamic motion might also be ach-
ievable after incorporation of guest molecules or polymers,
potentially leading to directional movement on the polymer
chain.[3a] We anticipate that this present system will open up
new opportunities for polymer-based nanomachines (e.g.,
drug carriers, molecular transporters, actuators) that are ap-
plicable to nano- and biomaterials.
[7] When curdlan adopts a triple-helix structure, the CD intensity of the
peripheral chromophore units increases as a result of their regular
arrangement on the helical curdlan structure; see: M. Ikeda, S. Har-
aguchi, M. Numata, S. Shinkai, Chem. Asian J. 2007, 2, 1290–1298.
[8] We could not directly confirm the morphology of CUR-Chl by
AFM because of its nonvolatility of DMSO. Because DMSO is a
good solvent for both curdlan and the peripheral chlorophyll units,
CUR-Chl adopted an extended fibrous structure in this solvent, as
was evidenced by its UV/Vis and CD spectra.
Experimental Section
General procedure: Briefly, copper(II) bromide (0.51 mg), ascorbic acid
(2.6 mg), and propylamine (0.27 mL) were added to a solution of 6-
azido-6-deoxycurdlan (CUR-N3, 5.1 mg) in DMSO (5 mL). Chl-X
(76 mg) was added and then the mixture was stirred for 3 days at room
temperature. The FTIR spectrum confirmed that the click reaction be-
tween CUR-N3 and Chl-X was quantitative (disappearance of the signal
for the azide units). The reaction product was subjected to dialysis, lyo-
philization, and washing with ether to give Chl-Chl as a powder (96%
yield). Detailed procedures for the chemical modification of native cur-
dlan are provided in the Supporting Information.
[9] The molecular weight of CUR-N3, the precursor of CUR-Chl, was
approximately 980000 gmolÀ1. Accordingly, the average length of
CUR-Chl was approximately 200 nm.
[10] This reformation process proceeds very slowly, taking more than
12 h. Therefore, we observed morphological and spectral changes
after leaving the resultant solution over more than 12 h. As an alter-
native procedure, we also mixed pyridine with the DMSO solution
containing CUR-Chl, and then H2O was added to the resultant
DMSO solution, adjusting DMSO/H2O 5:95, expecting to suppress
the stacking of chlorophyll units more quickly. In this case, we could
also confirm the same morphological and spectral changes. Based
on these fundamental experiments utilizing the pyridine ligand, we
first prepared a DMSO solution containing both CUR-Chl and
ligand, and then added H2O to the resultant DMSO solution to
form CUR-Chl/ligand complexes.
Acknowledgements
This study was partially supported by Grants-in-Aid for Scientific Re-
search (B) (22310069), Exploratory Research (23651118).
[11] Figure S6 in the Supporting Information reveals the effect of the
[pyridine]/[chlorophyll] ratio on the morphology of CUR-Chl.
[12] The contracted CUR-Chl samples were prepared through a globule
aggregation process from its extended form. Therefore, there is
slight discrepancy in CD spectra obtained in the absence of ligand.
[13] The UV/Vis and CD spectral changes suggested that CUR-Chl in-
teracted with PVP quantitatively. However, at a [PVP]/[chlorophyll]
ratio of 0.5 AFM images showed random coil CUR-Chl in addition
to straight fiber structures, presumably because of the dissociation
of PVP from CUR-Chl during sample preparation. Accordingly, to
ensure quantitative conversion from coils to straight fibers, an
excess of PVP ([PVP]/[chlorophyll]=30) was added prior to record-
ing the AFM image.
[1] a) G. S. Kottas, L. I. Clarke, D. Horinek, J. Michl, Chem. Rev. 2005,
105, 1281–1376 and references cited therein; b) K. Kinbara, T.
Aida, Chem. Rev. 2005, 105, 1377–1400 and references cited there-
in; c) V. Berl, I. Huc, R. Khoury, M. J. Krische, J.-M. Lehn, Nature
2000, 407, 720–723; d) M. Barboiu, J.-M. Lehn, Proc. Natl. Acad.
Sci. USA 2002, 99, 5201–5206; e) K. Maeda, H. Mochizuki, M. Wa-
tanabe, E. Yashima, J. Am. Chem. Soc. 2006, 128, 7639–7650; f) H.-
J. Kim, E. Lee, H. Park, M. Lee, J. Am. Chem. Soc. 2007, 129,
10994–10995; g) V. Percec, J. G. Rudick, M. Peterca, P. A. Heiney, J.
Am. Chem. Soc. 2008, 130, 7503–7508; h) K. Miwa, Y. Furusho, E.
Yashima, Nat. Chem. 2010, 2, 444–449; i) Y. Ferrand, Q. Gan, B.
Kauffmann, H. Jiang, I. Huc, Angew. Chem. 2011, 123, 7714–7717;
Angew. Chem. Int. Ed. 2011, 50, 7572–7575.
[2] a) M. M. Caruso, D. A. Davis, Q. Shen, S. A. Odom, N. R. Sottos,
S. R. White, J. S. Moore, Chem. Rev. 2009, 109, 5755–5798 and refer-
Received: October 7, 2012
Published online: January 10, 2013
1598
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 1592 – 1598