Polypeptide vesicles with densely packed multilayer membranes

Article abstract of DOI:10.1039/c5sm00820d

Multilamellar membranes are important building blocks for constructing self-assembled structures with improved barrier properties, such as multilamellar lipid vesicles. Polymeric vesicles (polymersomes) have attracted growing interest, but multilamellar polymersomes are much less explored. Here, we report the formation of polypeptide vesicles with unprecedented densely packed multilayer membrane structures with poly(ethylene glycol)-block-poly(γ-(4,5-dimethoxy-2-nitrobenzyl)-l-glutamate) (PEG-b-PL), an amphiphilic diblock rod-coil copolymer containing a short PEG block and a short hydrophobic rod-like polypeptide segment. The polypeptide rods undergo smectic ordering with PEG buried between the hydrophobic polypeptide layers. The size of both blocks and the rigidity of the hydrophobic polypeptide block are critical in determining the membrane structures. Increase of the PEG length in PEG-b-PL results in the formation of bilayer sheets, while using random-coil polypeptide block leads to the formation of large compound micelles. UV treatment causes ester bond cleavage of the polypeptide side chain, which induces helix-to-coil transition, change of copolymer amphiphilicity, and eventual disassembly of vesicles. These polypeptide vesicles with unique membrane structures provide a new insight into self-assembly structure control by precisely tuning the composition and conformation of polymeric amphiphiles.

Full text of DOI:10.1039/c5sm00820d

Soft Matter
View Article Online
View Journal
PAPER
Polypeptide vesicles with densely packed
multilayer membranes†
Cite this:DOI: 10.1039/c5sm00820d
Ziyuan Song,‡a Hojun Kim,‡a Xiaochu Ba,b Ryan Baumgartner,b Jung Seok Lee,a
Haoyu Tang,a Cecilia Leal*a and Jianjun Cheng*ab
Multilamellar membranes are important building blocks for constructing self-assembled structures with
improved barrier properties, such as multilamellar lipid vesicles. Polymeric vesicles (polymersomes)
have attracted growing interest, but multilamellar polymersomes are much less explored. Here, we
report the formation of polypeptide vesicles with unprecedented densely packed multilayer membrane
structures with poly(ethylene glycol)-block-poly(g-(4,5-dimethoxy-2-nitrobenzyl)-L-glutamate) (PEG-b-
PL), an amphiphilic diblock rod-coil copolymer containing a short PEG block and a short hydrophobic
rod-like polypeptide segment. The polypeptide rods undergo smectic ordering with PEG buried between
the hydrophobic polypeptide layers. The size of both blocks and the rigidity of the hydrophobic poly-
peptide block are critical in determining the membrane structures. Increase of the PEG length in PEG-b-PL
results in the formation of bilayer sheets, while using random-coil polypeptide block leads to the
formation of large compound micelles. UV treatment causes ester bond cleavage of the polypeptide
side chain, which induces helix-to-coil transition, change of copolymer amphiphilicity, and eventual
disassembly of vesicles. These polypeptide vesicles with unique membrane structures provide a new insight
into self-assembly structure control by precisely tuning the composition and conformation of polymeric
amphiphiles.
Received 7th April 2015,
Accepted 24th April 2015
DOI: 10.1039/c5sm00820d
www.rsc.org/softmatter
that liposomes have well packed short lipid bilayer structures
whereas polymersomes are based on randomly entangled long
Introduction
Polymeric vesicles (or polymersomes) have found broad applica- hydrophobic polymers. Control over the stability and membrane
tions in encapsulation and drug delivery.1–7 Amphiphilic block permeability of polymersomes rather than the detailed molecular
copolymers with coil–coil structures are often used to form arrangement in the hydrophobic layer has been largely the focus
sandwich-like 2D curved membrane structures with a hydrophobic of study in the past 20 years.1,2,9–11
wall and hydrophilic inner and outer coronas.1,2 A hydrophobic
In contrast to coil–coil block copolymers, rod-coil block
polymer segment with high molecular weight (MW) is essential to copolymers have received increasing attention in solution self-
ensure strong interactions between chains (e.g., chain entangle- assembly due to the special rigid conformation of the rod
ment and/or hydrophobic interaction) to form a kinetically frozen blocks.12–14 The anisotropic alignment of hydrophobic rods
hydrophobic layer with appropriate toughness.8 In addition, favors the formation of 2D membranes with lower curvature
the strong interactions are necessary to impart low membrane rather than the formation of spherical micelles.2,14 Polypeptides,
permeability for the stable encapsulation of cargo.1,9 Although a class of polymeric biomaterials with broad biological and bio-
polymersomes are structurally related to liposomes, as both are medical applications,15–21 can adopt rigid a-helical conformation
vesicles with hollow interiors, one key structural difference is and have been demonstrated to be excellent rod-like building
blocks for vesicle membranes.22–24 For instance, Deming et al.
first reported polypeptide vesicles through the conformation-
a Department of Materials Science and Engineering, University of Illinois at
specific self-assembly of amphiphilic diblock copolypeptides.25,26
Urbana–Champaign, 1304 West Green Street, Urbana, Illinois 61801, USA.
Membrane stability was enhanced by the preferred side-by-side
alignment of rigid rod-like hydrophobic polypeptide helices
along the helical axis and the lower conformational entropy loss
during the assembly process.2,9 The detailed polypeptide vesicle
membrane structures, however, have never been fully studied,
although they are believed to form very thin hydrophobic layers
(several nanometers) with the suggested unilamellar membrane
E-mail: jianjunc@illinois.edu
b Department of Chemistry, University of Illinois at Urbana–Champaign, 505 South
Mathews Avenue, Urbana, Illinois 61801, USA. E-mail: cecilial@illinois.edu
† Electronic supplementary information (ESI) available: Monomer and polymer
characterization details (1H NMR, CD, and ATR-FTIR spectra of synthesized
materials), additional TEM images, and disassembly studies with Nile Red probe.
See DOI: 10.1039/c5sm00820d
‡ Contributed equally.
This journal is ©The Royal Society of Chemistry 2015
Soft Matter

View Article Online
Paper
Soft Matter
Experimental section
Materials
All chemicals were purchased from Sigma-Aldrich (St. Louis,
MO, USA) and used as received unless otherwise specified.
Anhydrous N,N-dimethylformamide (DMF) was dried by a column
packed with 4 Å molecular sieves and stored in a glovebox.
Anhydrous tetrahydrofuran (THF) and hexane were dried by a
column packed with alumina. Methoxy poly(ethylene glycol)
amines (mPEG-NH2, 1 kDa, 2 kDa and 5 kDa) were purchased
from Laysan Bio (Arab, AL, USA). Spectra/Por RC dialysis tubings
with molecular weight cut off (MWCO) of 1 kDa were purchased
from Spectrum Laboratories (Rancho Dominguez, CA, USA).
Carbon film and holey carbon film supported copper grids
(200 mesh) were purchased from Electron Microscopy Sciences
(Hatfield, PA, USA).
Instrumentation
1H nuclear magnetic resonance (NMR) spectra were recorded
on a Varian U500 MHz or a VXR-500 MHz spectrometer.
Chemical shifts were reported in ppm and referenced to the
solvent proton impurities. Gel permeation chromatography
(GPC) experiments were performed on a system equipped with
an isocratic pump (Model 1100, Agilent Technology, Santa
Clara, CA, USA), a DAWN HELEOS multi-angle laser light
scattering (MALLS) detector (Wyatt Technology, Santa Barbara,
CA, USA), and an Optilab rEX refractive index detector (Wyatt
Fig. 1 Schematic illustration of polypeptide vesicle formation with
densely packed multilayer membrane and UV-triggered disassembly pro-
cess. The polypeptide length l is estimated by assuming ideal a-helical
conformation, the layer thickness d and the membrane thickness (domain
size) L are estimated from SAXS results.
structure.25,26 Given the rigidity of polypeptide rods compared Technology, Santa Barbara, CA, USA). The detection wavelength
to lipids, polypeptide vesicles with such nanometer-thin stiff of HELEOS was set at 658 nm. Separations were performed
membrane may have compromised vesicle stability and poorly using serially connected size exclusion columns (102 Å, 103 Å,
controlled membrane permeability.
104 Å, 105 Å, and 106 Å Phenogel columns, 5 mm, 300 ꢀ 7.8 mm,
In this article, we use poly(ethylene glycol)-block-poly(g-(4,5- Phenomenex, Torrance, CA, USA) at 60 1C using DMF contain-
dimethoxy-2-nitrobenzyl)-L-glutamate) (PEG-b-PL), a rod-coil ing 0.1 mol Lꢁ1 LiBr as the mobile phase. The MALLS detector
diblock polymeric amphiphile with a hydrophobic polypeptide was calibrated using pure toluene and can be used for the
segment, to show how the membrane structures of polypeptide determination of the absolute molecular weights. The MWs of
vesicles can be manipulated. With cryogenic TEM and X-ray polymers were determined based on the dn/dc value of each
scattering techniques, we show for the first time clear evidence polymer sample calculated offline by using the internal calibra-
of densely packed multilayer membrane structures of poly- tion system processed by the ASTRA 6 software (version 6.1.1.17,
peptide vesicles. These structures result presumably from Wyatt Technology, Santa Barbara, CA, USA). Circular dichroism
further assembly of unstable unilamellar structures, and closely (CD) measurements were carried out on a JASCO J-815 CD
resemble liposomal multilamellar vesicle (MLV) membrane spectrometer (JASCO, Easton, MD, USA). The polymer samples
structures with the absence of water spacing between layers were prepared at a concentration of 0.40 mg mLꢁ1 in aqueous
(Fig. 1). Random-coil PEG segments play an important role in solution at pH = 7 and the solution was placed in a quartz cell
segregating the PL macromolecular rods into a smectic-like with a pathlength of 0.10 cm. The mean residue molar ellipticity
ordering, while being short enough to be collapsed and buried of each polypeptide was calculated on the basis of the measured
between the layers of polypeptide helices. Although a-helical apparent ellipticity by following the literature-reported formulas:
polypeptides such as poly(g-benzyl-L-glutamate) (PBLG) often ellipticity ([y] in deg cm2 dmolꢁ1) = (millidegrees ꢀ mean residue
pack nematically with long-range directional order but no weight)/(path length in millimetres ꢀ concentration of poly-
positional order,27–30 we are able to obtain smectic ordering peptide in mg mLꢁ1).33,34 Transmission electron microscopy
of rod-like polypeptides through the use of flexible PEG domains (TEM) and cryogenic TEM (cryo-TEM) images were collected
whereas previously this packing could only be obtained through using JEOL 2100 cryo transmission electron microscope. Ultra-
the use of genetically engineered monodisperse polypeptides.31,32 violet (UV) light was generated from an OmniCure S1000 UV
The incorporation of photo-responsive PL segments also lamp (Lumen Dynamics Group, Mississauga, Ontario, Canada).
enables the study of the triggered disassembly of such vesicles Infrared spectra were recorded on a Perkin Elmer 100 serial FTIR
by changing the side-chain charge state and disrupting the spectrophotometer calibrated with polystyrene film (PerkinElmer,
amphiphilic structure (Fig. 1).
Santa Clara, CA, USA). Fluorescent spectra were recorded on a
Soft Matter
This journal is ©The Royal Society of Chemistry 2015

View Article Online
Soft Matter
Paper
Perkin Elmer LS 55 fluorescence spectrometer (PerkinElmer, (s, 1H, ArH), 6.50 (s, 1H, NH), 5.48 (q, 2H, ArCH2–), 4.46 (t, 1H,
Santa Clara, CA, USA).
a-H), 4.00 (s, 3H, CH3O–), 3.96 (s, 3H, CH3O–), 2.65 (t, 2H,
–CH2CH2CO–), 2.24 (m, 2H, –CH2CH2CO–). 13C NMR (CDCl3):
d 172.2, 169.6, 153.7, 151.9, 148.9, 140.5, 125.9, 111.6, 108.6,
64.4, 57.0, 56.8, 56.7, 29.8, 27.1.
Synthesis of c-(4,5-dimethoxy-2-nitrobenzyl)-L-glutamate
(DMNB-L-Glu) and c-(4,5-dimethoxy-2-nitrobenzyl)-D-glutamate
(DMNB-D-Glu)
DMNB-L-Glu was synthesized following the reported procedure.19 Glu. The final product was obtained as light yellow crystal (80%
In a 250 mL flat bottom flask, N,N,N0,N0-tetramethylguanidine yield). H NMR (CDCl3): d 7.71 (s, 1H, ArH), 6.97 (s, 1H, ArH),
DMNB-D-Glu-NCA was synthesized similarly using DMNB-D-
1
(1.1 mL, 8.77 mmol) was added dropwise to a stirred mixture 6.20 (s, 1H, NH), 5.50 (q, 2H, ArCH2–), 4.44 (t, 1H, a-H), 4.01
of L-glutamic acid (0.65 g, 4.45 mmol) and L-glutamic acid (s, 3H, CH3O–), 3.97 (s, 3H, CH3O–), 2.65 (t, 2H, –CH2CH2CO–),
copper(II) complex (1.05 g, 2.15 mmol) in DMF (4.0 mL) and 2.26 (m, 2H, –CH2CH2CO–).
distilled (DI) water (0.6 mL). The solution was stirred at room
temperature for 2 h until all solids were dissolved and then
more DMF (3.0 mL) was added. 4,5-Dimethoxy-2-nitrobenzyl
bromide (2.5 g, 9.06 mmol) was added to the above solution in
one portion. The reaction mixture was stirred at 40 1C for 24 h.
Acetone (100 mL) was added to the mixture and stirred for 2 h
Synthesis of poly(ethylene glycol)-block-poly(c-(4,5-dimethoxy-
2-nitrobenzyl)-L-glutamate) (PEG-b-PL), poly(ethylene glycol)-
block-poly(c-(4,5-dimethoxy-2-nitrobenzyl)-D-glutamate)
(PEG-b-PD) and poly(ethylene glycol)-block-poly(c-(4,5-
dimethoxy-2-nitrobenzyl)-DL-glutamate) (PEG-b-PDL)
until a fine precipitate was obtained. The violet solid was In a glovebox, DMNB-L-Glu-NCA (60 mg, 0.16 mmol) was dis-
collected by filtration, followed by mixing with freshly prepared solved in DMF (1.50 mL), followed by adding the DMF solution
ethylenediaminetetraacetic acid (1.89 g)/sodium bicarbonate of mPEG-NH2 (0.02 mol Lꢁ1, 407 mL, 0.008 mmol, M/I = 20). The
(1.08 g) aqueous solution (15 mL) to remove excessive copper polymerization mixture was stirred at room temperature. FTIR
salts. The mixture was stirred for another 24 h. The crude product was used to monitor the polymerization until the conversion
was collected by filtration and washed with DI water. The solid was above 99%. The polymer was then precipitated by cold
was further purified by recrystallization from isopropanol/DI hexane/ether (1: 1, v/v) and collected by centrifugation. The final
water (1 : 1, v/v). Isopropanol was then removed under vacuum polymer PEG-b-PL was obtained as viscous yellow solid after
and DI water was removed via lyophilization to yield the final removing the solvent residue under vacuum (80–87% yield). The
product DMNB-L-Glu as light yellow powder (1.74 g, 77% yield). glassware was wrapped with aluminum foil to avoid light
The glassware was wrapped with aluminum foil to avoid exposure during the whole process. 1H NMR (CDCl3/TFA-d,
light exposure during the whole process. 1H NMR (DMSO-d6/ 85: 15, v/v): d 7.66 (s, 1H, ArH), 6.97 (s, 1H, ArH), 5.40 (s, 2H,
D2O–DCl (35 wt%), 9 : 1, v/v): d 7.59 (s, 1H, ArH), 7.11 (s, 1H, ArCH2–), 4.67 (s, 1H, a-H), 3.94 (s, 3H, CH3O–), 3.92 (s, 3H,
ArH), 5.30 (s, 2H, ArCH2–), 3.86 (t, 1H, a-H), 3.83 (s, 3H, CH3O–), CH3O–), 3.78 (s, 4H, –OCH2CH2–), 2.59 (s, 2H, –COCH2CH2–),
3.78 (s, 3H, CH3O–), 2.60 (m, 2H, –CH2CH2CO–), 2.04 (m, 2H, 2.14 (d, 2H, –COCH2CH2–).
–CH2CH2CO–).
PEG-b-PD was synthesized similarly using DMNB-D-Glu-NCA
DMNB-D-Glu was synthesized similarly using D-glutamic acid as the monomer. The final product was obtained as a viscous
and D-glutamic acid copper(II) complex. The final product yellow solid (82–85% yield). 1H NMR (CDCl3/TFA-d, 85 : 15, v/v):
was obtained as light yellow powder (78% yield). 1H NMR d 7.65 (s, 1H, ArH), 6.96 (s, 1H, ArH), 5.39 (s, 2H, ArCH2–), 4.64
(DMSO-d6/D2O–DCl (35 wt%), 9 : 1, v/v): d 7.58 (s, 1H, ArH), (s, 1H, a-H), 3.93 (s, 3H, CH3O–), 3.92 (s, 3H, CH3O–), 3.77
7.10 (s, 1H, ArH), 5.29 (s, 2H, ArCH2–), 3.85 (t, 1H, a-H), 3.82 (s, 4H, –OCH2CH2–), 2.59 (s, 2H, –COCH2CH2–), 2.13 (d, 2H,
(s, 3H, CH3O–), 3.77 (s, 3H, CH3O–), 2.58 (m, 2H, –CH2CH2CO–), –COCH2CH2–).
2.03 (m, 2H, –COCH2CH2–).
PEG-b-PDL was synthesized similarly by mixing DMNB-L-
Glu-NCA and DMNB-D-Glu-NCA with a 1 : 1 ratio as the mono-
mers. The final product was obtained as yellow solid (78–82%
yield). H NMR (CDCl3/TFA-d, 85 : 15, v/v): d 7.63 (s, 1H, ArH),
6.97 (s, 1H, ArH), 5.38 (s, 2H, ArCH2–), 4.64 (s, 1H, a-H), 3.93 (s, 6H,
Synthesis of c-(4,5-dimethoxy-2-nitrobenzyl)-L-glutamate
N-carboxyanhydride (DMNB-L-Glu-NCA) and c-(4,5-dimethoxy-2-
nitrobenzyl)-D-glutamate N-carboxyanhydride (DMNB-D-Glu-NCA)
1
In a dried 250 mL two-neck round bottom flask, DMNB-L-Glu CH3O–), 3.76 (s, 4H, –OCH2CH2–), 2.58 (s, 2H, –COCH2CH2–),
(0.70 g, 2.04 mmol) was added and dried under vacuum for 2 h. 2.14 (d, 2H, –COCH2CH2–).
Phosgene (15 wt% in toluene, 2.0 mL, 2.80 mmol) was added
Preparation of copolymer self-assemblies in aqueous solution
along with anhydrous THF (30 mL), the mixture was stirred
at 50 1C for 2 h under the protection of drying tube. Solvent Dried PEG-b-PL diblock copolymer powder (5 mg, or with
THF was then removed under vacuum to obtain a yellow solid. 0.1 wt% Nile Red) was dissolved in DMF (1.0 mL) in a small vial
The crude product was purified by recrystallization from THF– charged with a magnetic stir bar, followed by dropwise addition
hexane (1 : 5, v/v) three times in a glovebox to obtain light yellow of DI water (4.0 mL) via syringe pump (KD Scientific, Holliston,
crystal (0.63 g, 84% yield). The resulting DMNB-L-Glu-NCA MA, USA. Addition speed: 0.1 mL minꢁ1). The suspension was
monomer was stored at ꢁ30 1C in the glovebox. The glassware stirred at room temperature for 2 h, and then transferred to a
was wrapped with aluminum foil to avoid light exposure during dialysis bag (MWCO = 1 kDa). The assemblies were dialyzed
1
the whole process. H NMR (CDCl3): d 7.70 (s, 1H, ArH), 6.97 against DI water for 4 h to remove DMF (water changed
This journal is ©The Royal Society of Chemistry 2015
Soft Matter

View Article Online
Paper
Soft Matter
every hour). The resulting suspension was used for subsequent their bilayer structure,25,26,37–41 we first evaluated whether high
studies.
MW polypeptides could be used to prepare vesicles with thick
membranes and, thus, potentially improved stability. We synthe-
sized PEG1k-b-PL100 and used the co-solvent method42 to drive the
formation of self-assembled structures. After water was dropwise
added into the DMF solution of PEG1k-b-PL100, macroscopic
precipitation was observed instead of self-assembly, suggesting
uncontrolled polypeptide chain interaction. The relatively short
hydrophilic PEG block may not be able to prevent the random
packing of rod-like polypeptides.25
Transmission electron microscopy (TEM)
TEM samples were prepared on carbon film supported copper
grids (200 mesh). One drop (B10 mL) of diluted copolymer
assembly aqueous suspension (0.25–0.5 mg mLꢁ1) was placed
on the grid and allowed to interact with the surface for 10 min.
Filter paper was then used to remove the residual polymers and
liquid. The sample on the grid was imaged using JEOL 2100 cryo
TEM at 80 kV.
Copolymers with reduced polypeptide DP were then synthe-
sized in order to better control helix–helix packing for stable
colloidal suspensions. Self-assemblies from three amphiphilic
Cryogenic transmission electron microscopy (Cryo-TEM)
copolymers with shorter polypeptide lengths, PEG1k-b-PL10
,
Cryo-TEM samples were prepared on holey carbon film supported
copper grids (200 mesh) using Vitrobot (FEI, Hillsboro, OR,
USA). One drop (B10 mL) of diluted copolymer assembly
aqueous suspension (0.25–0.5 mg mLꢁ1) was placed on the
grid, and the drop was blotted with blotting paper. The solution
residue was then vitrified by rapidly immersing it into liquid
ethane. The vitrified sample was transferred to a JEOL 2100
cryo TEM microscope for imaging using a cryo-holder at 200 kV.
The temperature of the sample was kept below ꢁ180 1C during
the course of sample preparation and imaging.
PEG1k-b-PL20, and PEG1k-b-PL40, were prepared. The a-helical
conformation of the synthesized PL blocks in aqueous environ-
ment was confirmed by circular dichroism (CD) and ATR-FTIR
spectroscopy (Fig. S6 and S7†). Based on the TEM analysis, we
found that PEG1k-b-PL20 could largely form hollow vesicular
structures with a diameter around 400 nm through closure of
the lamellar membrane (Fig. 2B). Such PEG/polypeptide ratio
was found to be crucial to form stable vesicular morphology.
PEG1k-b-PL10 copolymers showed irregular membrane structures
presumably because of the weak interaction of the ultra-short
polypeptide (PL 10-mer) (Fig. 2A).26 PEG1k-b-PL40 mainly formed
large broken pieces of membrane likely due to the stiff nature of
the PL 40-mer, which makes it difficult for the membrane to curve
into vesicles (Fig. 2C).25 In addition, we noticed all three copolymer
assemblies showed uncommonly high contrast on carbon film
grids without staining. This observation was quite different from
other amphiphiles with similar lengths (e.g., lipids), indicating
a thick membrane structure (Fig. S8†).
In order to understand how copolymer composition and
polypeptide conformation influence the assembly structure,25,26,38
we synthesized four more polypeptides with variable PEG lengths
and polypeptide conformations. PEG2k-b-PL20 and PEG2k-b-PL40,
with increased hydrophilic PEG size and bulkiness, showed the
sheet-like typical bilayer membrane structures with low contrast
in the absence of stain (Fig. 2D, E and Fig. S8†). The self-assembly
of PEG1k-b-PD20 and PEG1k-b-PDL20 were also studied to elucidate
the effects of polypeptide conformation on the self-assembled
structures. CD analysis indicated PD segments adopted left-
handed a-helical conformation, while the racemic PDL block
showed a random-coil conformation with no Cotton effect
(Fig. S6†). PEG1k-b-PD20 assemblies exhibited similar hollow
vesicular morphology with thick membranes, substantiating
that the helical sense has no effect on the assembly behavior
Small/wide-angle X-ray scattering (SAXS/WAXS)
Copolymer assembly samples were prepared in 1.5 mm quartz
X-ray capillaries (Hilgenberg Glas, Germany) and the SAXS/
WAXS experiments were conducted in a home built (Forvis
Technologies, Santa Barbara, CA, USA) equipment composed of
a Xenocs GeniX3D CuKa Ultra Low Divergence X-ray source
(1.54 Å/8 keV), with a divergence of B1.3 mrad. The 2D diffrac-
tion data were radially averaged upon acquisition on a Pilatus
300 K 20 Hz hybrid pixel Detector (Dectris) and integrated using
FIT2D software (http://www.esrf.eu/computing/scientific/FIT2D)
from ESRF.35,36
UV irradiation studies
The copolymer aqueous suspension (1.0 mL) was transferred
to a small vial charged with a magnetic stir bar, the vial was
then placed under UV lamp for UV irradiation (l = 365 nm,
I = 0–50 mW cmꢁ2, t = 0–30 min). The UV intensity and
irradiation time was controlled through the UV lamp to study
the disassembly process.
Results and discussion
We prepared PEG-polypeptide amphiphilic diblock copolymers (Fig. 2F). Racemic PEG1k-b-PDL20 only formed large compound
through ring-opening polymerization (ROP) of g-(4,5-dimethoxy-2- micelles instead of 2D sheets (Fig. 2G),42,43 similar as what was
nitrobenzyl)-L-glutamate N-carboxyanhydride (DMNB-L-Glu-NCA) previously reported for polypeptide vesicles with random-coil
(Scheme S1 and Fig. S1–S5†).19 The resulting diblock copolymers hydrophobic polypeptide blocks.25,26 In the absence of side-by-side
are named as PEGm-b-PXn, where ‘‘m’’ is the MW of PEG, ‘‘n’’ is the ordering of helical polypeptide rods, coil–coil diblock copolymers
degree of polymerization (DP) of polypeptides, and ‘‘X’’ refers to with pure hydrophobic effects cannot form stable bilayer membrane
the stereochemistry of the amino acid residues of the polypeptide at relatively low hydrophobic-to-hydrophilic ratio.2,25,26,39
block (L, D or DL). Since previously reported polypeptide vesicles
From the self-assembly behavior of the seven copolymers
only have B20 hydrophobic polypeptide repeating units in shown in Table 1, we have clearly shown that both morphology
Soft Matter
This journal is ©The Royal Society of Chemistry 2015

View Article Online
Soft Matter
Paper
Fig. 2 The self-assembly morphology of amphiphilic diblock copolymers PEG-b-polypeptide. (A–E) Composition-dependent self-assembly morphol-
ogy: (A) PEG1k-b-PL10, (B) PEG1k-b-PL20, (C) PEG1k-b-PL40, (D) PEG2k-b-PL20 and (E) PEG2k-b-PL40. (F and G) Conformation-dependent self-assembly
morphology: (F) PEG1k-b-PD20 and (G) PEG1k-b-PDL20. (H) Cryogenic TEM images of PEG1k-b-PL20 vesicle.
and membrane structures of PEG-b-PL assemblies are governed Two interactions may play a role in the assembly of these
not only by the block composition of amphiphiles but also multilayer structures. First, previous work on polypeptide based
the conformation of hydrophobic segments. The membrane rod-coil amphiphiles demonstrates that the interactions between
structures of assemblies were also studied by cryogenic TEM a-helical rigid rod-like hydrophobic segments first drive the
(cryo-TEM) to confirm the difference between PEG1k-b-PL and formation of 2D bilayer sheets through anisotropic side-by-side
PEG2k-b-PL assemblies. PEG1k-b-PL20 showed uniform, hollow ordering.25,26 Since conformational entropy loss during the
vesicular structures with high contrast (Fig. 2H and Fig. S9†). segregation process of stiff helical polypeptides is insignificant,
The membrane thickness was estimated to be 40 nm, which the minimization of interfacial energy dominates the assembly
is much thicker than typical bilayer structure (B6 nm for two process of PEG-b-PL forming bilayer structures.2 Second, the
polypeptide blocks with DP = 20, assuming ideal a-helix). On interaction between the formed bilayer and the solvent is
the other hand, PEG2k-b-PL40 showed spherical sheets with relatively high in energy, as the PEG is not long and bulky
much lower contrast (Fig. S9†).
enough to fully solvate and stabilize the formed bilayer in the
We propose that the formation of the unusually thick aqueous environment. This leads to higher interfacial energy at
membrane of PEG1k-b-PL is due to the further assembly of
‘‘unstable’’ bilayer sheets into a multilayer structure (Fig. 3).
Table 1 The compositions of the PEG-b-polypeptide copolymers and
their self-assembled structures
Copolymer
Compositiona
Morphologyb
Membranec
PEG1k-b-PL10
PEG1k-b-PL20
PEG1k-b-PL40
PEG2k-b-PL20
PEG2k-b-PL40
PEG1k-b-PD20
PEG1k-b-PDL20
PEG22-b-PL10
PEG22-b-PL22
PEG22-b-PL44
PEG44-b-PL20
PEG44-b-PL38
PEG22-b-PD22
PEG22-b-PDL22
V, I
V
I
S
S
V
LCM
Multilayer
Multilayer
Multilayer
Bilayer
Bilayer
Multilayer
—
a
b
Obtained copolymer composition determined by 1H NMR. Morphol-
ogy determined visually from regular TEM images. V = vesicles;
I = irregular aggregates, S = membrane sheets, LCM = large compound
c
Fig. 3 Proposed self-assembly process of PEG1k-b-PL to form multilayer
membrane and PEG2k-b-PL to form bilayer membrane.
micelle. Membrane structure determined by cryo-TEM images and
SAXS results.
This journal is ©The Royal Society of Chemistry 2015
Soft Matter

View Article Online
Paper
Soft Matter
the hydrophobic–hydrophilic interface, resulting in unstable
bilayers subject to further assembly for minimized interfacial
energy. The bilayers assemble and form multilayer structures
with PEG segments collapsed and buried between polypeptide
helical domains and water molecules excluded between
bilayers (Fig. 3). To our best knowledge, this is the first report
of further assembly of unstable bilayers into densely packed
multilayer membrane structures in polymersomes.
The use of PEG1k as the hydrophilic block is important in
this polypeptide-based assembly system. PEG1k block is not
only an effective separator to segregate the polypeptides, but also
acts as a short and flexible unit connecting the helical layers.
PEG2k, when used in the amphiphilic copolymer, provides
sufficient solubilizing ability and bulkiness to hydrate and
Fig. 4 X-ray scattering scans obtained for five different assemblies of
PEG-b-PL. (A) SAXS scans (logy) reveal a Bragg reflection (indicated by
solid arrow) at the characteristic repeat distances of d = 2p/q = 12.2, 13.4
stabilize the bilayer membrane in aqueous solution, which and 18.5 nm for PEG1k-b-PL10, PEG1k-b-PL20, and PEG1k-b-PL40, respec-
tively. A weak second order Bragg peak (indicated by the dash arrow) is
prefers to form traditional bilayer membrane with low interfacial
energy (Fig. 3). The increased size and hydrophilicity of PEG2k
segments also make it difficult to be buried between layers of
also observed for the PEG1k-b-PL40 sample (at repeat distance d = 9.0 nm).
The peak positions were determined by a Gaussian fit to the SAXS lines
after a polynomial background reduction. The subtracted SAXS data in the
helices; therefore, PEG2k-b-PL assembled membranes resemble
q interval of the first peak is shown on the inset. (B) WAXS scans show a
the previous reported polypeptide vesicle systems,25,26,37–40 prominent Bragg peak (indicated by solid arrow) at characteristic distance
d (nm) observed for all assemblies. PEG1k-b-PL10 (d = 0.66), PEG1k-b-PL20
which used sterically hindered hydrophilic polymer blocks
including oligo(ethylene glycol) based a-helical polypeptides,25
(d = 0.71), PEG1k-b-PL40 (d = 0.77), PEG2k-b-PL20 (d = 0.75), and PEG2k-b-
PL40 (d = 0.77). Another weak Bragg peak (indicated by the dash arrow) is
glycopolypeptides,38,40 Y-shaped branched PEG,39 or charged
also observed for PEG1k-b-PL20 suspension (at repeat distance d = 1.4 nm).
polypeptides that prevented close-packing of bilayers.26,37
The peak positions were determined by a Gaussian fit to the WAXS lines
Similar as PEG2k-b-PL, the further assembly of bilayers in these after a polynomial background reduction shown on the inset.
systems is therefore not favored considering the bulkiness or
the charge repulsion of these hydrophilic segments.
To confirm the multilayered structures and further elucidate is also their size, relative to the polypeptide segment, affects the
how PEG1k-b-PL amphiphiles were arranged in the membrane, degree of ordering. For example, the PEG1k-b-PL10 membrane
small-angle X-ray scattering (SAXS) and wide-angle X-ray scat- gives rise to the strongest Bragg peak intensity compared to other
tering (WAXS) were used to analyze five assemblies in aqueous copolymers, indicating a more ordered smectic phase. The hydro-
suspensions (Fig. 4A and B). In the SAXS regime, a Bragg peak philic PEG domain is an effective separator to segregate the PL
was observed for the PEG1k-b-PL assembly at the characteristic block, and also function as a ‘spring’ to compensate the geo-
distances of d = 2p/q (nm): PEG1k-b-PL10 (d = 12.2), PEG1k-b-PL20 metrical mismatch of polypeptide layer. Increase of the poly-
(d = 13.4), and PEG1k-b-PL40 (d = 18.5). A weak second order peptide length, with reduced fraction of PEG1k, results in
Bragg peak was also observed for the PEG1k-b-PL40 sample decreased segregation effects between helical domains and
(at d = 9.0 nm) (Fig. 4A). These results indicate the existence decreased geometrical compensation effects, eventually leading
of a multilayer arrangement in all PEG1k-b-PL membranes, to uncontrolled packing (as PEG1k-b-PL100 mentioned above).
where each layer comprises about 8, 4, and 3 copolymer units Indeed, for the PEG2k-b-PL systems where the membrane can
of PEG1k-b-PL10, PEG1k-b-PL20, and PEG1k-b-PL40, respectively barely be observed or appears to be very thin in TEM, no Bragg
(assuming ideal a-helix, 0.15 nm per polypeptide repeating reflections were detected by SAXS.
unit). The membrane thickness of self-assemblies, or in other
Fig. 4B shows the WAXS data obtained for all assemblies. A
words, the domain size L which relates to the number of layers prominent reflection at the characteristic spacings of d = 2p/q
in each membrane estimated by the full width at half max- (nm) is observed for all the samples: PEG1k-b-PL10 (d = 0.66),
imum of the primary Bragg peaks, is determined to be L = 32, PEG1k-b-PL20 (d = 0.71), PEG1k-b-PL40 (d = 0.77), PEG2k-b-PL20
42, 49 nm for PEG1k-b-PL10, PEG1k-b-PL20, and PEG1k-b-PL40
,
(d = 0.75), and PEG2k-b-PL40 (d = 0.77). Regardless of segment
respectively (Fig. 1). The results correlate well with the length, these Bragg reflections result from the rise per turn as
membrane thickness observed by cryo-TEM of PEG1k-b-PL20 previously reported.44 In addition, we also observed another
(40 nm). One should note that the Bragg reflections in the low intensity Bragg peak at low q (q = 4.5 nmꢁ1) before the most
SAXS regime are rather broad. While there are many possible prominent peak at ca. q = 9 nmꢁ1 for PEG1k-b-PL20. The corre-
reasons for peak broadening, we should expect to obtain a local sponding distance 1.4 nm matches the diameter of a-helical
distribution of d spacings for PEG-b-PL assemblies due to the polypeptides.45 This WAXS reflection arises from a lateral short
fact that the PEG moiety is flexible and may act as a ‘spring’ range order of the polypeptides within the bilayer membrane as
between hydrophobic PL units. In addition, the fact that there similarly observed in other a-helix systems.31 Through Gaussian
is no water within the multilayers affects the contrast of the peak fitting, the short range order domain size of PEG1k-b-PL20
X-ray signal. While the PEG domains make the peaks broader, it vesicles in lateral direction is calculated as approximately 2.4 nm,
Soft Matter
This journal is ©The Royal Society of Chemistry 2015

View Article Online
Soft Matter
Paper
PEG-b-PL diblock copolymers exhibited similar turbidity and
color change, indicating the same disassembly mechanism as
PEG1k-b-PL20 vesicles. Nile Red was used as a hydrophobic indicator
to evaluate the UV-responsiveness of PEG1k-b-PL20 vesicles.48 The
disassembly process was studied with different UV intensities and
irradiation times, taking advantage of the tunability and precise
control of the light trigger.49–51 By changing the UV intensity
from 0.5 to 50 mW cmꢁ2, we were able to successfully control
the disassembly kinetics. Lower UV intensity led to a slower
disassembly process, while longer irradiation time (10 min)
completely disrupted the vesicle structure (Fig. S10†).
Conclusions
We reported the formation of the unprecedented, densely packed
multilayer membranes using PEG1k-b-PL block copolymers. In
such membrane, polypeptide rods exhibit smectic ordering, with
PEG1k segments buried between the layers of helices. The PEG1k
moiety behaves as an effective separator to enable polypeptide
segregation, and is short enough to allow the interaction of the
neighboring polypeptide layers for further assembly to eventually
form the multilayered structures. Tuning copolymer composition
Fig. 5 Disassembly of polypeptide vesicles upon UV irradiation (l = 365 nm).
(A) CD spectra of PEG5k-b-PL20 upon UV irradiation at 10 mW cmꢁ2 for 0, 5,
10 and 30 min. PEG5k was used to solubilize PL segments in aqueous solution and polypeptide conformation lead to the change of self-
for CD tests; (B, C) TEM images of PEG1k-b-PL20 before and after UV irradiation
assembly morphology and membrane structures. When diblock
copolymer composition is properly adjusted (PEG1k-b-PL20),
a hollow vesicular morphology is obtained. UV irradiation
(10 min, 10 mW cmꢁ2). The insets show photography of PEG1k-b-PL20 aqueous
suspension before and after UV irradiation.
alters polypeptide conformation and copolymer amphiphilicity,
resulting in the complete disassembly of self-assembled structures.
which corresponds to about two units. With more understand-
The combination of thick membrane structures and trigger-
ings on copolymer arrangement in the multilayer membranes
responsive moieties in this new polypeptide vesicle may open
from the X-ray scattering studies, it is clear that the multilayer
up new opportunities for the design of membranes of encap-
membrane from PEG-b-PL assembly is structurally different
from the liposomal MLV and other multilamellar polymersome
sulation and delivery systems with improved barrier properties.
Meanwhile, the unique vesicle membrane structures may also
as no water is found in the densely packed multilayer membrane
provide new insights into the self-assembly behavior of rod-
structure (Fig. 1).7,46,47
coil block copolymers, showing how precise control of block
When stimuli responsive functionalities are introduced in
copolymer composition and secondary structure influence the
vesicle-forming polymeric amphiphiles, it is possible to control
molecular arrangement and eventually control the morphology
the morphological transition or membrane permeability, thus
and membrane structures of assemblies.
potentially broaden the application of polymersomes in encap-
sulation and delivery.10,11,14,22 As the copolymer has built-in
photo-responsive PL domain, we next studied the trigger-induced Acknowledgements
disassembly of PEG1k-b-PL20 vesicles. Under UV irradiation, the
This work was supported by The Dow Chemical Company via the
Dow-University of Illinois sponsored research program (J.C.),
cleavage of the 4,5-dimethoxy-2-nitrobenzyl (DMNB) ester bond
results in the transition from the hydrophobic, a-helical PL to the
and NSF (CHE-1308485, J.C.), and the University of Illinois
anionic, random-coil poly(L-glutamic acid) (Fig. 1). The elimina-
(C.L.). We thank Dr Keith Harris, Dr Liang Hong, and Dr Joshua
tion of the PL helical structure was verified by CD spectroscopy
S. Katz for helpful discussions, and Dr Wacek Swiech for the
(Fig. 5A). This conformation change, together with the change of
amphiphilicity of the copolymer, induced the disassembly of
assistance on TEM and cryo-TEM studies.
vesicles. After 10 min UV irradiation, the turbid vesicle suspension
became yellowish and clear, indicating the release of the DMNB
group and the complete disassembly of the self-assembled struc-
References
ture. The resulting post-irradiation samples were further analyzed
by TEM, which showed the disappearance of regular hollow
spherical structures previously observed and the formation of
irregular aggregates from PEG-b-poly(glutamic acid) dried on
the TEM grid (Fig. 5B and C). All other self-assemblies from
1 D. E. Discher and A. Eisenberg, Science, 2002, 297, 967–973.
2 M. Antonietti and S. Forster, Adv. Mater., 2003, 15, 1323–1333.
3 D. E. Discher, V. Ortiz, G. Srinivas, M. L. Klein, Y. Kim,
C. A. David, S. S. Cai, P. Photos and F. Ahmed, Prog. Polym.
Sci., 2007, 32, 838–857.
This journal is ©The Royal Society of Chemistry 2015
Soft Matter

View Article Online
Paper
Soft Matter
4 D. H. Levine, P. P. Ghoroghchian, J. Freudenberg, G. Zhang, 27 C. Robinson and J. C. Ward, Nature, 1957, 180, 1183–1184.
M. J. Therien, M. I. Greene, D. A. Hammer and R. Murali, 28 C. Robinson, Tetrahedron, 1961, 13, 219–234.
Methods, 2008, 46, 25–32.
29 W. G. Miller and E. L. Wee, J. Phys. Chem., 1971, 75,
5 D. A. Christian, S. Cai, D. M. Bowen, Y. Kim, J. D. Pajerowski
1446–1452.
and D. E. Discher, Eur. J. Pharm. Biopharm., 2009, 71, 463–474. 30 J. C. Horton, A. M. Donald and A. Hill, Nature, 1990, 346,
6 J. S. Lee and J. Feijen, J. Controlled Release, 2012, 161,
473–483.
7 S. Rangelov and A. Pispas, Polymer and Polymer-Hybrid
Nanoparticles: From Synthesis to Biomedical Applications,
CRC Press, Boca Raton, FL, 2013.
44–45.
31 S. M. Yu, V. P. Conticello, G. Zhang, C. Kayser,
M. J. Fournier, T. L. Mason and D. A. Tirrell, Nature, 1997,
389, 167–170.
32 S. M. Yu, C. M. Soto and D. A. Tirrell, J. Am. Chem. Soc., 2000,
122, 6552–6559.
8 B. M. Discher, Y.-Y. Won, D. S. Ege, J. C.-M. Lee, F. S. Bates,
D. E. Discher and D. A. Hammer, Science, 1999, 284, 33 A. J. Adler, N. J. Greenfield and G. D. Fasman, in Methods
1143–1146.
Enzymol., ed. S. N. T. C. H. W. Hirs, Academic Press, 1973,
vol. 27, pp. 675–735.
9 J. F. Le Meins, O. Sandre and S. Lecommandoux, Eur. Phys.
J. E: Soft Matter Biol. Phys., 2011, 34, 1–17.
10 M.-H. Li and P. Keller, Soft Matter, 2009, 5, 927–937.
11 F. Meng, Z. Zhong and J. Feijen, Biomacromolecules, 2009,
10, 197–209.
12 H. A. Klok and S. Lecommandoux, Adv. Mater., 2001, 13,
1217–1229.
13 Y.-b. Lim, K.-S. Moon and M. Lee, J. Mater. Chem., 2008, 18,
2909–2918.
14 J. Zhang, X.-F. Chen, H.-B. Wei and X.-H. Wan, Chem. Soc.
Rev., 2013, 42, 9127–9154.
15 H. Lu, J. Wang, Y. Bai, J. W. Lang, S. Liu, Y. Lin and
J. Cheng, Nat. Commun., 2011, 2, 206.
34 N. J. Greenfield, Nat. Protoc., 2006, 1, 2876–2890.
35 A. P. Hammersley, S. O. Svensson, M. Hanfland, A. N. Fitch
and D. Hausermann, High Pressure Res., 1996, 14, 235–248.
36 A. P. Hammersley, ESRF Internal Report, 1998, ESRF98-
HA01T, FIT2D V9.129 Reference Manual V123.121.
37 J. Rodriguez-Hernandez and S. Lecommandoux, J. Am.
Chem. Soc., 2005, 127, 2026–2027.
38 J. Huang, C. Bonduelle, J. Thevenot, S. Lecommandoux and
A. Heise, J. Am. Chem. Soc., 2012, 134, 119–122.
39 K. Osada, H. Cabral, Y. Mochida, S. Lee, K. Nagata, T. Matsuura,
M. Yamamoto, Y. Anraku, A. Kishimura, N. Nishiyama and
K. Kataoka, J. Am. Chem. Soc., 2012, 134, 13172–13175.
16 N. P. Gabrielson, H. Lu, L. Yin, D. Li, F. Wang and J. Cheng, 40 J. R. Kramer, A. R. Rodriguez, U.-J. Choe, D. T. Kamei and
Angew. Chem., Int. Ed., 2012, 51, 1143–1147. T. J. Deming, Soft Matter, 2013, 9, 3389–3395.
17 L. Yin, Z. Song, K. H. Kim, N. Zheng, N. P. Gabrielson and 41 A. R. Rodriguez, J. R. Kramer and T. J. Deming, Biomacro-
J. Cheng, Adv. Mater., 2013, 25, 3063–3070. molecules, 2013, 14, 3610–3614.
18 L. Yin, Z. Song, Q. Qu, K. H. Kim, N. Zheng, C. Yao, 42 L. Zhang and A. Eisenberg, Science, 1995, 268, 1728–1731.
I. Chaudhury, H. Tang, N. P. Gabrielson, F. M. Uckun and 43 Y. Mai and A. Eisenberg, Chem. Soc. Rev., 2012, 41,
J. Cheng, Angew. Chem., Int. Ed., 2013, 52, 5757–5761.
19 L. Yin, H. Tang, K. H. Kim, N. Zheng, Z. Song, N. P. Gabrielson, 44 Y. Wang, F. L. Filho, P. Geil and G. W. Padua, Macromol.
H. Lu and J. Cheng, Angew. Chem., Int. Ed., 2013, 52, 9182–9186. Biosci., 2005, 5, 1200–1208.
20 H. Tang, L. Yin, K. H. Kim and J. Cheng, Chem. Sci., 2013, 4, 45 A. Niehoff, A. Mantion, R. McAloney, A. Huber, J. Falkenhagen,
5969–5985.
3839–3844.
C. Goh, A. Thu¨nemann, M. Winnik and H. Menzel, Colloid
Polym. Sci., 2013, 291, 1353–1363.
46 A. Jesorka and O. Orwar, Annu. Rev. Anal. Chem., 2008, 1,
801–832.
47 A. Napoli, M. Valentini, N. Tirelli, M. Muller and J. A. Hubbell,
Nat. Mater., 2004, 3, 183–189.
48 A. P. Goodwin, J. L. Mynar, Y. Ma, G. R. Fleming and
21 Z. Song, N. Zheng, X. Ba, L. Yin, R. Zhang, L. Ma and
J. Cheng, Biomacromolecules, 2014, 15, 1491–1497.
22 A. Carlsen and S. Lecommandoux, Curr. Opin. Colloid Interface
Sci., 2009, 14, 329–339.
23 C. Deng, J. Wu, R. Cheng, F. Meng, H.-A. Klok and Z. Zhong,
Prog. Polym. Sci., 2014, 39, 330–364.
´
24 H. Lu, J. Wang, Z. Song, L. Yin, Y. Zhang, H. Tang, C. Tu,
Y. Lin and J. Cheng, Chem. Commun., 2014, 50, 139–155.
25 E. G. Bellomo, M. D. Wyrsta, L. Pakstis, D. J. Pochan and
T. J. Deming, Nat. Mater., 2004, 3, 244–248.
26 E. P. Holowka, D. J. Pochan and T. J. Deming, J. Am. Chem.
Soc., 2005, 127, 12423–12428.
J. M. J. Frechet, J. Am. Chem. Soc., 2005, 127, 9952–9953.
49 N. Fomina, J. Sankaranarayanan and A. Almutairi, Adv. Drug
Delivery Rev., 2012, 64, 1005–1020.
50 G. Liu and C.-M. Dong, Biomacromolecules, 2012, 13,
1573–1583.
51 Y. Zhao, Macromolecules, 2012, 45, 3647–3657.
Soft Matter
This journal is ©The Royal Society of Chemistry 2015