Temperature-sensitive elastin-mimetic dendrimers: Effect of peptide length and dendrimer generation to temperature sensitivity

Article abstract of DOI:10.1002/bip.22425

Dendrimers are synthetic macromolecules with unique structure, which are a potential scaffold for peptides. Elastin is one of the main components of extracellular matrix and a temperature-sensitive biomacromolecule. Previously, Val-Pro-Gly-Val-Gly peptides have been conjugated to a dendrimer for designing an elastin-mimetic dendrimer. In this study, various elastin-mimetic dendrimers using different length peptides and different dendrimer generations were synthesized to control the temperature dependency. The elastin-mimetic dendrimers formed β-turn structure by heating, which was similar to the elastin-like peptides. The elastin-mimetic dendrimers exhibited an inverse phase transition, largely depending on the peptide length and slightly depending on the dendrimer generation. The elastin-mimetic dendrimers formed aggregates after the phase transition. The endothermal peak was observed in elastin-mimetic dendrimers with long peptides, but not with short ones. The peptide length and the dendrimer generation are important factors to tune the temperature dependency on the elastin-mimetic dendrimer. Copyright

Full text of DOI:10.1002/bip.22425

Temperature-Sensitive Elastin-Mimetic Dendrimers: Effect of Peptide
Length and Dendrimer Generation to Temperature Sensitivity
Chie Kojima,1 Kotaro Irie,2 Tomoko Tada,3 Naoki Tanaka3
1 Nanoscience and Nanotechnology Research Center, Research Organization for the 21st Century, Osaka Prefecture University,
1-2 Gakuen-cho, Naka-ku, Sakai, Osaka 599-8570, Japan
2 Department of Physical Science, Graduate School of Science, Osaka Prefecture University, 1-1 Gakuen-cho, Naka-ku, Sakai,
Osaka 599-8531, Japan
3 Department of Bio-molecular Engineering, Kyoto Institute of Technology, Gosyokaido-cho, Matsugasaki,
Sakyo-ku, Kyoto 606-8585, Japan
Received 22 August 2013; revised 26 September 2013; accepted 8 October 2013
Published online 11 October 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/bip.22425
drimers with long peptides, but not with short ones. The
ABSTRACT:
peptide length and the dendrimer generation are impor-
Dendrimers are synthetic macromolecules with unique
tant factors to tune the temperature dependency on the
structure, which are a potential scaffold for peptides.
C
V
elastin-mimetic dendrimer. 2013 Wiley Periodicals,
Elastin is one of the main components of extracellular
matrix and a temperature-sensitive biomacromolecule.
Previously, Val-Pro-Gly-Val-Gly peptides have been con-
jugated to a dendrimer for designing an elastin-mimetic
dendrimer. In this study, various elastin-mimetic den-
drimers using different length peptides and different den-
drimer generations were synthesized to control the
temperature dependency. The elastin-mimetic dendrimers
formed b-turn structure by heating, which was similar to
the elastin-like peptides. The elastin-mimetic dendrimers
exhibited an inverse phase transition, largely depending
on the peptide length and slightly depending on the den-
drimer generation. The elastin-mimetic dendrimers
formed aggregates after the phase transition. The endo-
thermal peak was observed in elastin-mimetic den-
Inc. Biopolymers 101: 603–612, 2014.
Keywords: elastin; dendrimer; temperature-sensitive;
peptide; phase transition
This article was originally published online as an accepted pre-
print. The “Published Online” date corresponds to the preprint
version. You can request a copy of the preprint by emailing the
Biopolymers editorial office at biopolymers@wiley.com
INTRODUCTION
lastin is one of the main components of extracellular
matrices and has been used as a temperature-sensitive
biomaterial. Elastin is composed of repeat sequences
such as valine-proline-glycine-valine-glycine (Val-
Pro-Gly-Val-Gly, VPGVG).1–3 This kind of peptide is
E
called an elastin-like peptide and is applied to elastin-
mimetic materials. Previous reports showed that elastin-like
polypeptides induced the temperature-dependent structure
change from random coil into b-turn and/or b-spiral, lead-
ing to coacervation because of the hydrophobic interac-
tions.1–3 Most elastin materials have been extracted from
animals or produced by using recombinant protein technol-
ogy.2,3 Since pathogens can become contaminated in these
preparation methods, chemically synthesized elastin materi-
als are desirable. However, it is very difficult to synthesize
Additional Supporting Information may be found in the online version of this
article.
Correspondence to: Chie Kojima; e-mail: c-kojima@21c.osakafu-u.ac.jp
Contract grant sponsor: Tokuyama Science Foundation
Contract grant sponsor: Special Coordination Funds for Promoting Science and
Technology of the Ministry of Education, Culture, Sports, Science and Technology
of Japan (MEXT) (Improvement of Research Environment for Young Researchers
(FY 2008-2012))
C
V
2013 Wiley Periodicals, Inc.
Biopolymers Volume 101 / Number 6
603

604
Kojima et al.
FIGURE 1 Elastin-mimetic dendrimers [ELP1-G4 (A), ELP2-G3 (B), ELP2-G4 (C), and ELP2-G5
(D)].
polypeptides chemically. Short elastin-like peptides, less than elastin-mimetic dendrimer to the temperature dependency. We
four repeats, do not exhibit the temperature-dependent
properties.4 Elastin-mimetic polypeptides are compounds
that link elastin-like peptide in tandem, and the parallel-
linked elastin-mimetic materials exhibit temperature-
dependent properties. Elastin-like peptide-grafted synthetic
polymers with one repeat exhibit the temperature-sensitive
properties.5–7 Since synthetic polymers generally have molec-
ular weight distribution, uniform elastin-mimetics are diffi-
cult to produce using synthetic linear polymers. Gold
nanoparticle was also used as a platform of elastin-like pep-
tides to produce temperature-sensitive nanoparticles.8 How-
ever, it seems difficult to control the binding number of the
peptides on the nanoparticle. Consequently, the phase transi-
tion temperature may be difficultly controlled in this system.
Dendrimers have highly controllable size, topologies, and
surface properties and are quite different from linear poly-
mers.9–12 Dendrimers can be used as a platform for peptides
because they have multiple functional groups at the periphery.
Multiple antigen peptides, called antigen peptide-conjugated
dendrimers, have been reported.12–14 Dendrimers conjugating
collagen model peptides, such as (Pro-Pro-Gly)n and (Pro-
Hyp-Gly)n, formed collagen-like triple helical structures and
temperature-dependent hydrogels for drug delivery.15–18 One
repeat elastin-like peptide (VPGVG) has been conjugated to
the polyamidoamine (PAMAM) dendrimer and exhibited the
temperature-dependent properties.19 Koga et al reported that
VPGVG and (VPGVG)4 peptides have been conjugated to the
PAMAM dendrimers to produce elastin-mimetics.20 However,
a systematic investigation of this kind of elastin-mimetic den-
drimer remains to be examined. In this study, we investigate
synthesized various elastin-mimetic dendrimers with one and
two repeats of elastin-like peptide [VPGVG (ELP1) and
(VPGVG)2 (ELP2)] and generation 3, 4, and 5 (G3, G4, and
G5): ELP1-G4, ELP2-G3, ELP2-G4, and ELP2-G5 (Figure 1).
The synthesized compounds were characterized by nuclear
magnetic resonance (NMR), circular dichroism (CD) measure-
ments, solution transmittance assay, dynamic light scattering
(DLS), transmission electron microscopy (TEM) analysis, and
differential scanning calorimetry (DSC) to elucidate the phase
transition mechanism of the elastin-mimetic dendrimers and
the influence of peptide lengths and dendrimer generations on
the temperature dependency.
RESULTS AND DISCUSSION
Synthesis of Elastin-mimetic Dendrimers
We previously prepared ELP1-G4.19 ELP2-conjugated den-
drimers (ELP2-G3, ELP2-G4, and ELP2-G5) were synthesized
as shown in Figure 2. The Boc-VPGVG peptide conjugation to
the dendrimers (PAMAM G3-G5) and the following
de-protection were performed repeatedly. Since amine-
terminated elastin-mimetic dendrimers did not exhibit
temperature-dependent properties under the physiological
condition, the N-termini of the peptide-conjugated den-
drimers were acetylated.19 The synthesized compounds were
1
characterized by H NMR (Supporting Information Figure
S1). The spectra of elastin-mimetic dendrimers reveal signals
derived from the PAMAM dendrimer (2.5, 2.7, 2.9, and 3.3
the effect of peptide lengths and dendrimer generations in the ppm), the peptide unit (0.96, 1.9–2.3, and 3.7–4.4 ppm), and
FIGURE 2 Synthetic pathway of elastin-mimetic dendrimers with two repeats of peptide. TFA
and Ac2O are trifluoroacetic acid and acetic anhydride, respectively.
Biopolymers

Temperature-Sensitive Elastin-Mimetic Dendrimers
605
Table I Elastin-mimetic Dendrimers in This Study
PAMAM Dendrimer
(Generation/Terminal
Number)
Peptide (Kind/Binding
Numbera)
Molecular Weightb (Da, ca)
(Peptidec/Dendrimer)
Sample
Transition Temperatured (ꢀC)
ELP1-G4
ELP2-G3
ELP2-G4
ELP2-G5
G4/64
G3/32
G4/64
G5/128
Ac-VPGVG/65
Ac-(VPGVG)2/32
Ac-(VPGVG)2/63
Ac-(VPGVG)2/127
43112 (26207/14151)
34461 (26207/6877)
69320 (52415/14150)
139035 (104829/28697)
4860.9
3963.7
3864.8
3063.3
a Estimated from NMR spectra.
b Given that target peptides and acetyl groups were fully modified.
c Acetyl group is not included.
d Average data estimated from temperature-dependent transmittance measurements using three different lots.
the acetyl unit (2.0 ppm). The integral ratio of the peptide sig- (random coil) were plotted against temperature (Figures 3C
nals to the PAMAM dendrimer indicates that virtually all ter- and 3D). The gradual increase of the 218 nm-cotton effect and
minal amino groups of dendrimer were modified with the the gradual decrease of the 196 nm-cotton effect were observed
peptide twice. From the integral ratios of the acetyl signal to with increasing temperature, indicating that the random coil
the dendrimer signals, essentially all terminal groups were ace- structure was gradually changed into the b-turn structure by
tylated. HPLC analysis of ELP2-modified dendrimers was per- heating. All of the elastin-mimetic dendrimers were similar to
formed (Supporting Information Figure S2). Broad peaks were the elastin-like peptide, thus the elastin-mimetic dendrimers
observed in chromatograms of ELP2-modified dendrimers, are an artificial elastin-mimetic material with temperature-
and there were no contaminations of free peptides. Therefore, dependent conformation change. The negative cotton effect of
we synthesized fully ELP-modified dendrimers. The obtained elastin-mimetic dendrimers at 218 nm was slightly lower than
elastin-mimetic dendrimers in this study are listed in Table I.
that of the corresponding peptides, indicating that conjugation
of the ELPs to the dendrimers slightly interfered with the for-
mation of b-turn structure. However, the dendrimer genera-
tion and the peptide length did not affect the negative cotton
effect at 218 nm very much. This suggests that the peptide
lengths as well as the dendrimer generations did not influence
the formation of the b-turn structure. The negative cotton
effect of elastin-mimetic materials at 196 nm was affected by
peptide lengths, but not by the dendrimer conjugation and the
dendrimer generation. It is likely that longer peptides can fix
the conformation due to the multiple interactions, which pro-
moted the random coil structure formation.
Secondary Structure of Elastin-mimetic Dendrimers
CD spectrometry is useful for the estimation of protein sec-
ondary structure. Elastin-mimetics can form b-turn by heat-
ing, which is elucidated when the CD spectrum has a negative
cotton effect around 218 nm.4,19,21–23 The CD spectra of the
elastin-mimetic dendrimers were measured, and the ELP pep-
tides were used as a control. Since ELP1-G4 and the one repeat
peptide were reported in our previous report, ELP2-
conjugated dendrimers (ELP2-G3, ELP2-G4, and ELP2-G5)
were analyzed in this study (Figure 3 and Supporting Informa-
tion Figure S3). The elastin-mimetic dendrimers and the ELP2
exhibited the negative cotton effect that corresponds to b-turn Phase Transition of Elastin-mimetic Dendrimers
structure (218 nm) and random coil (196 nm), similar to The phase transition of aqueous solutions containing the
ELP1-G4 and the ELP1.19 Since conformation of elastin- elastin-mimetic dendrimers (1 mg/mL, pH 7.4, 150 mM NaCl)
mimetics is changed from random coil to b-turn structure by was examined. The solution containing elastin-mimetic den-
heating, the CD spectra were measured at different tempera- drimers became turbid by heating, but the corresponding pep-
tures from 5ꢀC to 65ꢀC. A strong negative cotton effect at 196 tides did not. The change in the transmittance of the solution
nm was observed at low temperature, corresponding to ran- was monitored with increasing temperature. ELP1-G4 was
dom coil structure. The negative cotton effect at 218 nm that examined in the previous report.19 ELP2-conjugated den-
corresponds to the b-turn structure increased at higher tem- drimers were analyzed in this study. Figure 4A shows the
perature in all elastin-mimetics. Since ELP2-conjugated den- temperature-dependent transmittance of ELP2-G3, ELP2-G4,
drimers were turbid after heating above 45ꢀC under our and ELP2-G5 in addition to ELP1-G4. The drastic transmit-
conditions, the data from 5ꢀC to 35ꢀC were obtained and com- tance change at a specific temperature, called a phase transition
pared. The molar ellipticities at 218 nm (b-turn) and 196 nm temperature, was observed in elastin-mimetic dendrimers. The
Biopolymers

606
Kojima et al.
FIGURE 3 Temperature-dependent CD spectra of ELP2-G4 (A) and the ELP2 peptide (B). Correla-
tion of the molar ellipticity that corresponding to random coil (196 nm, C) and b-turn (218 nm, D)
to temperature. Data on ELP1-den and Ac-VPGVG-OH were referred from the reference.19 (C,D)
Open squares and diamonds are ELP1 and ELP2 peptides, respectively. Closed squares, closed trian-
gles, closed diamonds, and closed circle are ELP1-G4, ELP2-G3, ELP2-G4, and ELP2-G5, respectively.
phase transition temperature of ELP2-G3, ELP2-G4, and on the thermosensitivity of the elastin-mimetic dendrimers.
ELP2-G5 were 37ꢀC, 33ꢀC, and 29ꢀC, respectively, which was Figure 4B shows the correlation between salt concentration
estimated from Figure 4A. For the experiment, three independ- and the phase transition temperature. The linear relationship
ent lots of each sample were used. The averaged phase transi- was observed in all elastin-mimetic dendrimers, whose slopes
tion temperatures are listed in Table I. Even though the phase were as follows: 219 for ELP1-G4, 217 for ELP2-G3, 216 for
transition temperature of ELP1-G4 was 48ꢀC, ELP2- ELP2-G4, and 215 for ELP2-G5. The linear decrease of phase
conjugated dendrimers became turbid at less than 40ꢀC. This transition temperature with increasing salt concentration is
suggests that longer peptides promoted the phase transition. consistent with previous observations in other elastin-mim-
Compared with the dendrimer generation, the averaged phase etics.4,19,22,24 It is thought that salt might suppress the hydra-
transition temperatures of G3 and G4 were similar, but that of tion on ELP, which is called the Hofmeister effect.22,24 The
G5 was lower than other dendrimers. This suggests that the slope of the elastin-mimetic dendrimer was similar to ELP
dendrimer generation partly affected the phase transition polypeptides and less steep than for short ELP.4,22,24 These sug-
behavior. Interestingly, the phase transition temperature of gest that the elastin-mimetic dendrimer has similar properties
ELP2-G3 was lower than ELP1-G4, even though the weight of to elastin polypeptides but not to ELP. The peptide length and
peptide moieties of ELP1-G4 and ELP2-G3 were the same the dendrimer generation did not influence on the slope, indi-
(Table I). This suggests that the peptide length in the elastin- cating that the salt effect happened at a similar level in the
mimetic dendrimer dominantly contributed to the decrease of elastin-mimetic dendrimers.
the phase transition temperature. It was reported that elastin-
We investigated the influence of pH on the thermosensitiv-
mimetics indicated the salt-dependent thermosensitive prop- ity of the elastin-mimetic dendrimer (Figure 4C), since ELP1-
erty.4,22,24 We examined the influence of the salt concentration G4 largely responded to pH.19 ELP2-conjugated dendrimers
Biopolymers

Temperature-Sensitive Elastin-Mimetic Dendrimers
607
tiary amine in the dendrimer. Since the dendrimer content in
the ELP2-conjugated dendrimer became less than in the ELP1-
dendrimer, the pH sensitivity may become lower.
Self-assembly of Elastin-mimetic Dendrimer
As described above, the solutions of elastin-mimetic den-
drimers became turbid. The size of elastin-mimetic particles at
different temperatures was examined by DLS. The intensity-
weighted size distribution and the number-weighted one of
elastin-mimetic dendrimers (ELP1-G4 and ELP2-G4) at differ-
ent temperatures are shown in Supporting Information Figures
S4–S5. The intensity-weighted data contained multiple peaks
at low temperature, indicating that there are several kinds of
particles in the solutions. The diameter of the PAMAM den-
drimer of G3, G4, and G5 are known to be 3.6 nm, 4.5 nm,
and 5.4 nm, respectively.10,25 Thus, small sized particles were a
unimolecular state, and large ones were the assembly. Since the
weighted intensity is calculated from the Rayleigh approxima-
tion (I a d6, where I is the scattering intensity and d is the par-
ticle size), larger particles were dominantly observed in the
intensity-weighted estimation. The number-weighted size dis-
tribution showed a single peak around 6.0 nm and 7.3 nm at
low temperature for ELP1-G4 and ELP2-G4, respectively (Sup-
porting Information Figures S4B and S5B). These suggest that
small portions of elastin-mimetic dendrimers were assembled
at low temperature.
The increase in the size of the elastin-mimetic dendrimers
was also examined after heating. Portions of the large particles
increased with increasing temperature in all samples. To under-
stand the population change of these two types of particles,
average diameters were estimated by cumulant analysis. The
averaged diameters of all elastin-mimetic dendrimer were
changed by increasing temperature (Figure 5). Interestingly,
the averaged diameters of ELP2-conjugated dendrimers drasti-
cally increased above 35ꢀC up to 150 nm (ELP2-G3), 680 nm
(ELP2-G4), and 1010 nm (ELP2-G5). Since the phase transi-
tion temperatures of these dendrimers were 37ꢀC (ELP2-G3),
33ꢀC (ELP2-G4), and 29ꢀC (ELP2-G5), these dendrimers
formed the self-assembly around the phase transition tempera-
ture. Figure 4 shows that the transmittance in the ELP2-G3
starts decreasing at 35ꢀC, even though the phase transition
temperature of ELP2-G3 is 37ꢀC. Thus, the size increase
occurred at 35ꢀC. However, the averaged diameter of ELP1-G4
gradually increased with increasing temperature. Even though
FIGURE 4 Phase transition of elastin-mimetic dendrimers. (A)
Temperature-dependent transmittance for solutions of elastin-
mimetic dendrimers (ELP1-G4, ELP2-G3, ELP2-G4, and ELP2-G5)
in phosphate buffer (pH 7.4 containing NaCl 0.15M). (B) Correla-
tion of the phase transition temperature to NaCl concentration (pH
7.4) in elastin-mimetic dendrimers. The results are shown as an
average from three different lots of elastin-mimetic dendrimer. (C)
Correlation of the phase transition temperature to pH (NaCl
0.15M) in elastin-mimetic dendrimers. The results are shown as an
average from three different lots of elastin-mimetic dendrimer. The
data on ELP1-G4 were referred from our previous report.19
were affected by pH. The higher phase transition temperature ELP1-G4 became turbid around 48ꢀC, the large particles were
was observed at low pH. However, the pH sensitivity in ELP2- formed at relatively large amount below the phase transition
conjugated dendrimers was lower than the ELP1-conjugated temperature. Thus, the self-assembly of ELP1-G4 gradually
dendrimer, and was attenuated with increasing generation. occurred and was independent of the phase transition temper-
The pH sensitivity was caused by the protonation of inner ter- ature. The averaged diameter of elastin-mimetic dendrimers
Biopolymers

608
Kojima et al.
FIGURE 5 Average diameter of elastin-mimetic dendrimers at dif-
ferent temperatures. The same lots shown in Figure 4A were used in
this analysis.
after the phase transition increased with increasing the peptide
length and the dendrimer generation. This suggests that larger
molecules formed larger aggregates after the phase transition.
TEM analysis of ELP2-G4 was performed at room temperature
FIGURE 7 DSC analysis of elastin-mimetic dendrimers, which
was normalized to dendrimer mole (A) and peptide unit mole (B).
The same lots shown in Figure 4A were used in this analysis.
and 50ꢀC (Figure 6). Spherical structures with around 10 nm
in diameter were observed at room temperature, but any
aggregates were not. On the other hand, large aggregates were
observed at 50ꢀC, consistent with our DLS data.
Thermal Process of Elastin-mimetic Dendrimers
Temperature-dependent polymers such as poly(N-isopropyla-
crylamide) (poly(NIPAM)) exhibited the endothermal peak
around the phase transition temperature, at which the dehy-
dration of bound water to the hydrophobic moiety of the poly-
mer occurred.26 It was reported that elastin-mimetics had an
endothermal reaction corresponding to dehydration.22,27 The
thermal behavior of elastin-mimetic dendrimers was examined
by DSC and compared with the previous elastin-mimetics. The
mole heat capacity (Cp) to dendrimer and ELP unit in the
elastin-mimetic dendrimers is shown in Figure 7. ELP2-G3,
ELP-G4, and ELP-G5 exhibited endothermal peaks at 38ꢀC,
36ꢀC, and 32ꢀC, respectively, which corresponds to the phase
transition temperature. This observation was consistent with
the previous elastin-mimetics.22,27 The data, normalized to
dendrimer mole, yield larger endothermal peaks on larger gen-
eration dendrimers (Figure 7A). The larger generation
FIGURE 6 TEM images of ELP2-G4 at room temperature (A) and
50ꢀC (B). Bar. 100 mm. The same lots shown in Figure 4A were used
in this analysis.
Biopolymers

Temperature-Sensitive Elastin-Mimetic Dendrimers
609
above the threshold (Step 2), which led to the aggregation
(Step 3); and consequently, the turbidity rapidly increased
around the phase transition temperature (Step 4).
The effects of peptide length and dendrimer generation to
the phase transition behavior are discussed. First, ELP1-G4
and ELP2-G4 are compared to understand the effects of pep-
tide length. The step 1 conformation change was not signifi-
cantly affected by the peptide length, but the following steps
were largely affected. Our DSC experiments detected endother-
mal signals in ELP2-G4, but not in ELP1-G4, which suggests
that the drastic dehydration occurred in ELP2-G4, but not in
ELP1-G4 (step 2). This result may be because the amount of
hydrated water is small in ELP1-G4 owing to the limited length
of the ELP. Accordingly, the average diameter of the ELP1-G4
gradually increased with increasing temperature. Our DLS data
suggest that ELP1-G4 assembled each other even at low tem-
perature and the equilibrium was changed from unimolecular
form to assembled one by heating (Step 3). When the
assembled form reached the threshold, the turbidity changed
drastically. The phase transition temperature significantly
decreased with elongated peptide length (Step 4). The balance
of the hydrophobicity is a key factor of the temperature-
dependent polymers.11,26 It was reported that molecular weight
of elastin-mimetics affected the phase transition behaviors.2,3
Thus, the decreased phase transition temperature came from
the increased hydrophobic ratio and the increased molecular
weight. The pH sensitivity was also different: ELP1-G4 was
more sensitive to pH than ELP2-G4. The pH dependency
came from the protonation of the inner dendrimer. The den-
drimer moiety in ELP1-G4 is larger than that in ELP2-G4.
FIGURE 8 Possible mechanism of phase transition in elastin-
mimetic dendrimers.
dendrimer contains large amounts of peptide, which is a cause
of the large endothermal signal. However, normalized to pep-
tide unit (VPGVG) mole, the endothermal peaks of all ELP2-
conjugated dendrimers look similar, suggesting that the pep-
tide induced the dehydration. However, ELP1-G4 did not
show any significant peaks under our condition, suggesting
that the dehydration did not occur around the phase transition
temperature. It was reported that dendrimers conjugating
NIPAM units at the termini did not show any significant endo-
thermal signals.26 It is likely that short peptide units at the sur-
face of the dendrimer may provide only small amounts of
bound water.
Current Model of Phase Transition Mechanisms in
Elastin-mimetic Dendrimers
The possible mechanism on the temperature dependency of Since the peptide components in the material increased by the
the elastin-mimetic dendrimers is shown in Figure 8. Our CD longer peptide, ELP2-G4 did not affect the protonation in the
results indicated that the conformation of the ELP at the den- dendrimer very much.
drimer surface was changed from random coil to b-turn by
ELP2-G3, ELP2-G4, and ELP2-G5 are compared to under-
heating. Since it was reported that the b-turn structure induced stand the effects of dendrimer generation. The Step 1 confor-
the assembly,2,3,21 the phase transition occurred by increasing mation change was not affected by the dendrimer generation
temperature. In other words, the conformation change was a very much. The Step 2 dehydration was essentially unchanged,
cue for the phase transition. Interestingly, the conformation normalized to the peptide unit mole. These results indicate
change was gradual by heating, but the turbidity was changed that the conformation change and dehydration were contrib-
drastically. This suggests that the phase transition occurred uted to the ELP unit. The Step 3 self-assembly was affected by
when the hydrophobicity of the dendrimer with b-turn struc- the dendrimer generation. The temperature to exhibit the dras-
ture was beyond the threshold. The size of the elastin-mimetic tic change in diameter decreased with increasing the dendrimer
dendrimers with the long peptides aggregates increased drasti- generation, which corresponds to the phase transition temper-
cally around the phase transition temperature. The phase tran- ature. The diameter after the phase transition became larger in
sition corresponded to the dehydration of the ELP, except in the elastin-mimetic dendrimer with larger dendrimer genera-
the ELP1-conjugated dendrimer. These results indicate that the tion. Even though the hydrophobicity balance is unchanged in
dehydrated ELP induced the self-assembly of the elastin- the different generation dendrimers, the molecular weights are
mimetic dendrimers. In summary, conformation of ELP at the increased. The large molecular weight induced the larger aggre-
termini of elastin-mimetic dendrimers changed from random gates. The phase transition temperature decreased with enlarg-
coil to b-turn by heating (Step 1); the dehydration occurred ing the dendrimer generation (Step 4). Thus, the decreased
Biopolymers

610
Kojima et al.
phase transition temperature came from the increased molecu-
lar weight. It was reported that the occupied surface areas per
peptide at the dendrimer periphery (G3, G4, and G5) can be
calculated as 1.3 nm2, 1.0 nm2 and 0.8 nm2, respectively, given
that the dendrimers are spherical.28 Thus, the terminal crowd-
ing effect may occur at the dendrimer periphery to induce the
cooperative behavior of the ELPs. However, the difference of
the phase transition temperature is limited, and thus, the phase
transition temperature may largely depend on the hydropho-
bicity balance.
Synthesis of Boc-(VPGVG)2-conjugated Dendrimers
Boc-VPGVG conjugation to VPGVG-dendrimer was performed
according to the same procedure described in Synthesis of Boc-
VPGVG-conjugated dendrimers except the quantities of the materials
and the reaction times. For Boc-(VPGVG)2-G3, VPGVG-G3 (2.0
lmol), Boc-VPGVG (0.078 mmol), HOAt (0.047 mmol), HATU
(0.095 mmol), and TEA (0.38 mmol) in the DMSO/DMF/chloroform
mixture (344 lL) were reacted for 5 days. Yield 77%. For Boc-
(VPGVG)2-G4, VPGVG-G4 (0.75 lmol), Boc-VPGVG (0.065 mmol),
HOAt (0.047 mmol), HATU (0.094 mmol), and TEA (0.40 mmol) in
the DMSO/DMF/chloroform mixture (344 lL) were reacted for 5
days. Yield 73%. For Boc-(VPGVG)2-G5, VPGVG-G5 (0.34 lmol),
Boc-VPGVG (0.067 mmol), HOAt (0.079 mmol), HATU (0.079
mmol), and TEA (0.32 mmol) in the DMSO/DMF/chloroform mix-
ture (344 lL) were reacted for 6 days. Yield 97%.
MATERIALS AND METHODS
Materials
Acetylation of Elastin-mimetic Dendrimers
Amino-terminated ethylenediamine core PAMAM dendrimers of G3,
G4, and G5 were purchased from Sigma-Aldrich (St. Louis, MO). o-
(7-Benzotriazol-1-yl)–1,1,3,3-tetramethyluronium hexafluorophos-
phate (HBTU), o-(7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluro-
nium hexafluorophosphate (HATU), 1-hydroxy-7-azabenzotriazole
(HOAt), and Ac-VPGVG were obtained from Watanabe Chemical
Industries (Hiroshima, Japan). Triethylamine (TEA), trifluoroacetic
acid (TFA), acetic anhydride, and celite (545RVS) were obtained from
Nacalai Tesque (Kyoto, Japan).
Boc-(VPGVG)2-dendrimers (G3 1.5 lmol; G4 0.51 lmol; G5 0.31
lmol) were dissolved in TFA (3 mL) and incubated at 4ꢀC for 2 h.
The reaction mixture was evaporated after the addition of water,
which was repeated four times. Freeze-dried (VPGVG)2-dendrimers
were dissolved in acetic anhydride (5 mL) and stirred at 40ꢀC for 2 h.
The reaction mixture was evaporated. The mixture was dialyzed
(pore: MW1000) in aqueous solution (pH 8) for a day. The final
product was obtained by freeze-drying. Yield 74% for ELP2-G3, 92%
1
for ELP2-G4, and 91% for ELP2-G5. H NMR (400 MHz, D2O): d
0.96 (m, Hc for Val), 1.92–2.31 (m, Hb, and Hc for Pro and Hb for
Val), 2.03 (s, Ac), 2.45 (br, ANCH2CH2CONHCH2CH2 for den-
drimer), 2.69 (br, ANCH2CH2CONHCH2CH2NA for dendrimer),
2.87 (br, ANCH2CH2CONHCH2CH2 for dendrimer), 3.32 (br,
Synthesis of Elastin-mimetic Dendrimers
Boc-VPGVG and ELP-G4 were produced according to our previous
report.19 Other elastin-mimetic dendrimers (ELP2-G3, ELP2-G4, and
ELP2-G5) were synthesized as shown in Figure 2. Three lots of each
elastin-mimetic dendrimer were synthesized. The typical synthetic
methods are described below.
ANCH2CH2CONHCH2CH2
and
ANCH2CH2CONHCH2CH2
NHCO-peptide for dendrimer), 3.71–4.44 (m, Hd for Pro, Ha for Val,
Ha for Gly and Ha for Pro).
Synthesis of Boc-VPGVG-conjugated Dendrimers
Boc-VPGVG conjugation to dendrimer was performed according to
our previous report.19 Boc-VPGVG-conjugated G3 and G5 den-
drimers were synthesized as follows. PAMAM dendrimer of G3 (2.4
lmol), Boc-VPGVG (0.098 mmol), and HOAt (0.061 mmol) were
dissolved in dimethylsulfoxide (DMSO)/N,N-dimethylformamide
(DMF)/chloroform at the ratio of 3/3/2 (344 lL). HATU (0.12 mmol)
was then added under nitrogen. TEA (0.53 mmol) was added, fol-
lowed by stirring at room temperature for 3 days. After the addition
of deionized water to the mixture, it was purified using a Sephadex
LH-20 column with methanol as the eluent. The final product was
obtained by freeze-drying. Yield 81%. The same procedure was used
for the PAMAM dendrimer of G5 (0.56 lmol), Boc-VPGVG (0.11
mmol), HOAt (0.068 mmol), HATU (0.13 mmol), and TEA (0.53
mmol) in DMSO/DMF/chloroform mixture (344 lL). Yield 63%.
Synthesis of ELP2 (Ac-(VPGVG)2)
Boc-VPGVG-OBzl was synthesized according to procedures in our
previous report.19 Boc-VPGVG-OBzl (1.65 mmol) was dissolved and
incubated in TFA (3 mL) for 3 h at 4ꢀC. The reaction mixture was
evaporated. After the addition of water, the subsequent evaporation
was performed, which was repeated four times. Yield (VPGVG-OBzl)
1
ꢁ 100%. H NMR (400 MHz, D2O): d 0.88–1.12 (m, Hc for Val),
1.97-2.34 (m, Hb, and Hc for Pro and Hb for Val), 3.66- 4.19 (m, Hd
for Pro, Ha for Val and Ha for Gly), 4.51 (m, Ha for Pro), 5.22 (s, Bzl),
7.44 (s, phenyl). Boc-VPGVG (0.29 mmol) and VPGVG-OBzl (0.28
mmol) were dissolved in distilled acetonitrile (10 mL). HBTU (0.284
mmol) was added to the mixture and stirred for 4 days. After the
addition of deionized water, the mixture was purified using a Sepha-
dex LH-20 column with methanol as the eluent. The final product
was obtained by freeze-drying. Yield 55%. 1H NMR (400 MHz,
DMSO): d 0.86 (m, Hc for Val), 1.36 (s, Boc), 1.83-2.08 (m, Hb, and
Boc Removal From Boc-VPGVG-conjugated Dendrimers
Boc removal was performed according to our previous report, and Hc for Pro and Hb for Val), 3.57-4.17 (m, Hd for Pro, Ha for Val and
VPGVG-conjugated G4 dendrimer was reported.19 VPGVG- Ha for Gly), 4.30 (m, Ha for Pro), 5.11 (s, Bzl), 7.34(s, phenyl).
conjugated G3 and G5 dendrimers were synthesized as follows. Boc-
Boc-(VPGVG)2-OBzl (0.16 mmol) was dissolved in ethanol
VPGVG-dendrimers (G3 1.9 lmol; G5 0.35 lmol) were dissolved in (5 mL). Pd-C (0.060 mmol) was added to the mixture and reduced
TFA (3 mL) and incubated at 4ꢀC for 2 h. The reaction mixture was with H2 gas for 1 day. The mixture was filtered with celite (545RVS)
evaporated after the addition of water, which was repeated four times. and evaporated. The final product was obtained after vacuum drying.
Yield ꢁ 100%.
Yield 98%. 1H NMR (400 MHz, DMSO): d 0.86 (m, Hc for Val), 1.37
Biopolymers

Temperature-Sensitive Elastin-Mimetic Dendrimers
611
(s, Boc), 1.82–2.08 (m, Hb, and Hc for Pro and Hb for Val), 3.57–4.16
(m, Hd for Pro, Ha for Val and Ha for Gly), 4.30 (m, Ha for Pro).
Boc-(VPGVG)2-OH (0.032 mmol) was dissolved and incubated in
TFA (3 mL) for 2 h at 4ꢀC. The reaction mixture was evaporated.
After the addition of water, the subsequent evaporation was per-
formed, which was repeated four times. The product was dissolved in
2 mL of deionized water to adjust to pH 9. NaOH (4N, 0.28 mL)
solution and acetic anhydride (0.14 mL) were added in a drop-wise
manner and stirred for 12 h at 4ꢀC. The solution pH was adjusted to
7 and the mixture was purified using a Sephadex LH-20 column with
ing average method. The molar ellipticity at 196 nm and 218
nm was estimated from the spectra, and the averaged data of
three different lots of dendrimer were reported.
MEASUREMENT OF PHASE TRANSITION
Various solutions for elastin-mimetic dendrimers [1 mg/mL,
10 mM buffer (pH 4, 5, 6, and 7.4) containing 0–1M NaCl]
were prepared by using 100 mM phosphate buffer (pH 7.4) or
citrate buffer (pH 4, 5, and 6), 4M NaCl solution, and 10 mg/
mL dendrimer solution. The turbidity was measured at
500 nm using a Jasco Model V-630 spectrophotometer
equipped with a Peltier-type thermostatic cell holder coupled
with an ETC-717 controller. The heating rate of the sample cell
was maintained at 1.0ꢀC min21. The phase transition tempera-
ture was taken as temperature at a transmittance of 50%. The
averaged phase transition temperature of three different lots of
dendrimer was reported.
1
methanol as the eluent. Yield 62%. H NMR (400 MHz, D2O): d
0.92–1.19 (m, Hc for Val), 1.98–2.34 (m, Hb, and Hc for Pro and Hb
for Val), 2.04 (s, Ac), 3.72–4.23 (m, Hd for Pro, Ha for Val and Ha for
Gly), 4.41 (m, Ha for Pro). HPLC: Rt 16.6 min, purity 94%.
CHARACTERIZATION
The obtained products were characterized by 1H-NMR (JEOL,
400 MHz).
The HPLC system was equipped with a Cosmosil 5C18-
MS-II column (Nacalai Tesque) and a UV detector (220 nm;
UV-2075Plus, Jasco, Tokyo, Japan). Samples (5 lL) were
injected with an autosampler (AS-2057Plus, Jasco) and eluted
with methanol/20 mM phosphate buffer 5 10/90 at 1.0 mL
min21. Acetonitrile was increased to 60% over 30 min.
DLS analysis was performed, as follows. Solutions for elastin-
mimetic dendrimers (1 mg/mL, 10 mM phosphate buffer (pH
7.4) containing 0.15M NaCl) were prepared and filtered (0.45
lm), followed by DLS analysis using ELSZ-DN2 (Otsuka Elec-
tronics, Osaka, Japan). The temperature was adjusted to 15ꢀC,
25ꢀC, 35ꢀC, 45ꢀC, and 55ꢀC. After 15 min incubation at each
temperature, the data were measured three times.
CONCLUSION
We synthesized and characterized the fully ELPs-conjugated
dendrimers with different peptide lengths and dendrimer gen-
erations. The conformation of elastin-mimetic dendrimers
changed from random coil into b-turn by heating, essentially
independent of the dendrimer structure. Elastin-mimetic den-
drimers underwent a temperature-dependent phase transition,
which was affected by salt concentration and pH. The phase
transition temperature was largely influenced by the peptide
length and slightly influenced by the dendrimer generation.
The diameter of the elastin-mimetic dendrimer enlarged
around the phase transition temperature, at which the dehy-
dration occurred in long peptide-conjugated dendrimers. The
enlarged diameter and the endothermal peak normalized to
dendrimer mole became greater with increasing peptide length
and dendrimer generation. Thus, the peptide length and den-
drimer generation are important factors to tune the
temperature dependency on the elastin-mimetic dendrimers.
Since the ELP2-conjugated dendrimers exhibited the phase
transition around the body temperature, they are potent
temperature-dependent biomaterials.11,19
The same solutions were used for the DSC. DSC measure-
ments over the range 20–60ꢀC were carried out with a Nano-
DSC 2 (Calorimetry Science Corporation (CSC), Salt lake City,
UT). The heating rate was 1ꢀC/min.
TEM analysis was performed, as follows. The solution for
ELP2-G4 (1 mg/mL, 10 mM phosphate buffer (pH 7.4) con-
taining 0.15M NaCl) was prepared and incubated at 4ꢀC and
50ꢀC. The sample solutions were dropped on carbon-coated
copper grids at room temperature and 50ꢀC, negatively stain-
ing with phosphotungstic acid TEM images were obtained
using JEM-2000FMX (JEOL, Tokyo, Japan).
CD MEASUREMENTS
The authors are grateful to Dr. Akihiro Nomoto and Prof. Akiya
Ogawa (Osaka Prefecture University) for their generous supply of
H2 gas in peptide synthesis.
CD spectra were measured by using a J-820 spectropolarimeter
(Jacso) from 5ꢀC to 65ꢀC. Before the measurements, the sam-
ple solutions [0.05 mg/mL, 10 mM phosphate buffer
(pH 5 7.4)] were incubated at each temperature for 10 min.
The CD spectra were obtained using a 0.1 cm path length cell, REFERENCES
by signal integrating 10 scans from 190 to 260 nm at a scan
speed of 50 nm/min. Data were processed by the simple mov-
1. Tatham, A. S.; Shewry, P. R. Trends Biochem Sci 2000, 25, 567.
2. Urry, D. W. Angew Chem Int Ed 1993, 32, 819.
Biopolymers

612
Kojima et al.
3. Urry, D. W.; Urry, K. D.; Szaflarski, W.; Nowicki, M. Adv Drug 17. Kojima, C.; Suehiro, T.; Tada, T.; Sakamoto, Y.; Waku, T.;
Deliv Rev 2010, 62, 1404.
Tanaka, N. Soft Matter 2011, 7, 8991.
18. Kojima, C.; Suehiro, T. Chem Lett 2011, 40, 1249.
19. Kojima, C.; Irie, K. Biopolym Pept Sci, in press.
20. Koga, T.; Iimura, M.; Higashi, N. Macromol Sci 2012, 12, 1043.
21. Nicolini, C.; Ravindra, R.; Ludolph, B; Winter, R. Biophys
J 2004, 86, 1385.
4. Nuhn, H.; Klok, H. A. Biomacromolecules 2008, 9, 2755.
5. Ayres, L.; Vos, M. R. J.; Adams, P. J. H. M.; Shklyarevskiy, I. O.;
van Hest J. C. M. Macromolecules 2003, 36, 5967.
6. Roberts, S. K.; Chilkoti, A.; Setton, L. A. Biomacromolecules
2007, 8, 2618.
22. Reguera, J.; Urry, D. W.; Parker, T. M.; McPherson, D. T.;
ꢀ
7. Conrad, R. M.; Grubbs, R. H. Angew Chem Int Ed 2009, 48,
8328.
Rodrıguez-Cabello J. C. Biomacromolecules 2007, 8, 354.
23. Reiersen, H.; Clarke, A. R.; Rees, A. R. J Mol Biol 1998, 283,
255.
8. Lemieux, V.; Adams, P. H. H. M.; van Hest, J. C. M. Chem
Commun 2010, 46, 3071.
24. Cho, Y.; Zhang, Y.; Christensen, T.; Sagle, L. B.; Chilkoti, A.;
Cremer, P. S. J Phys Chem B 2008, 112, 13765.
25. Uppuluri, S.; Keinath, S. E.; Tomalia, D. A.; Dvornic, P. R. Mac-
romolecules 1998, 31, 4498.
9. Lee, C. C.; MacKay, J. A.; Frechet, J. M. J.; Szoka, F. C. Nat Bio-
tech 2005, 12, 1517.
10. Svenson, S.; Tomalia, D. A. Adv Drug Deliv Rev 2005, 57, 2106.
11. Kojima, C. Exp Opin Drug Deliv 2010, 7, 307.
12. Astruc, D.; Boisselier, E.; Ornelas, C. Chem Rev 2010, 110, 1857.
13. Crespo, L.; Sanclimens, G.; Pons, M.; Giralt, E.; Royo, M.;
Albericio, F. Chem Rev 2005, 105, 1663.
26. Haba, Y.; Kojima, C.; Harada, A.; Kono, K. Angew Chem Int Ed
2007, 46, 234.
27. Yamaoka, T.; Tamura, T.; Seto, Y.; Tada, T.; Kunugi, S.; Tirrell,
D. A. Biomacromolecules 2003, 4, 1680.
14. Sadler, K.; Tam, J. P. J Biotechnol 2002, 90, 195.
15. Kojima, C.; Tsumura, S.; Harada, A.; Kono, K. J Am Chem Soc
2009, 131, 6052.
28. Suehiro, T.; Kojima. C.; Tsumura, S.; Harada, A.; Kono, K. Bio-
polymers 2010, 93, 640.
16. Suehiro, T.; Tada, T.; Waku, T.; Tanaka, N.; Hongo, C.; Yamamoto,
S.; Nakahira, A.; Kojima. C. Biopolymers 2011, 95, 270.
Reviewing Editor: Eric J. Toone
Biopolymers