Polysaccharides present in the glycocalyx and extracellular matrix are highly important for a multitude of functions. Oligo- and polysaccharides-based biomaterials are being developed to mimic the glycocalyx, but the spatial functionalization of these polysaccharides represents a major challenge. In this paper, a series of benzene-1,3,5-tricarboxamide (BTA) based supramolecular monomers is designed and synthesized with mono- (BTA-β-d-glucose; BTA-Glc and BTA-α-d-mannose; BTA-Man) or disaccharides (BTA-β-d-cellobiose; BTA-Cel) at their periphery or a monosaccharide (BTA-OEG4-α-d-mannose; BTA-OEG4-Man) at the end of a tetraethylene glycol linker. These glycosylated BTAs have been used to generate supramolecular assemblies and it is shown that the nature of the carbohydrate appendage is crucial for the supramolecular (co)polymerization behavior. BTA-Glc and BTA-Man are shown to assemble into micrometers long 1D (bundled) fibers with opposite helicities, whereas BTA-Cel and BTA-OEG4-Man formed small spherical micelles. The latter two monomers are used in a copolymerization approach with BTA-Glc, BTA-Man, or ethylene glycol BTA (BTA-OEG4) to give 1D fibers with BTA-Cel or BTA-OEG4-Man incorporated. Consequently, the carbohydrate appendage influences both the assembly behavior and the internal order. Using this approach it is possible to create 1D-fibers with adjustable saccharide densities exhibiting tailored dynamic exchange profiles. Furthermore, hydrogels with tunable mechanical properties can be achieved, opening up possibilities for the development of multicomponent functional biomaterials.
Polysaccharides present in the glycocalyx and extracellular matrix are highly important for a multitude of functions. Oligo- and polysaccharides-based biomaterials are being developed to mimic the glycocalyx, but the spatial functionalization of these polysaccharides represents a major challenge. In this paper, a series of benzene-1,3,5-tricarboxamide (BTA) based supramolecular monomers is designed and synthesized with mono- (BTA-β-d-glucose; BTA-Glc and BTA-α-d-mannose; BTA-Man) or disaccharides (BTA-β-d-cellobiose; BTA-Cel) at their periphery or a monosaccharide (BTA-OEG4-α-d-mannose; BTA-OEG4-Man) at the end of a tetraethylene glycol linker. These glycosylated BTAs have been used to generate supramolecular assemblies and it is shown that the nature of the carbohydrate appendage is crucial for the supramolecular (co)polymerization behavior. BTA-Glc and BTA-Man are shown to assemble into micrometers long 1D (bundled) fibers with opposite helicities, whereas BTA-Cel and BTA-OEG4-Man formed small spherical micelles. The latter two monomers are used in a copolymerization approach with BTA-Glc, BTA-Man, or ethylene glycol BTA (BTA-OEG4) to give 1D fibers with BTA-Cel or BTA-OEG4-Man incorporated. Consequently, the carbohydrate appendage influences both the assembly behavior and the internal order. Using this approach it is possible to create 1D-fibers with adjustable saccharide densities exhibiting tailored dynamic exchange profiles. Furthermore, hydrogels with tunable mechanical properties can be achieved, opening up possibilities for the development of multicomponent functional biomaterials.
Polysaccharides are
abundantly present on the cell surface, i.e.,
the glycocalyx, and in the extracellular matrix, e.g., glycosaminoglycans
(GAGs). They play a number of critical roles in many essential biological
processes, both structurally and functionally,[1,2] such
as water immobilization, growth factor binding, cell attachment, and
signaling.[3−5] The multivalency effect—the cooperative action
of multiple interaction simultaneously—is fundamental in many
of these carbohydrate-receptor interactions.[6] Since specificity and affinity binding are significantly influenced
by the structural and spatial arrangement of the carbohydrate ligands
in polysaccharides, synthetic glycopolymers and glycodendrimers serve
as attractive candidates for biomaterials and they can be used to
investigate interaction mechanisms underlying biological events.[7−9] However, the synthesis of stereo-, composition-, and sequence-defined
glycopolymers or glycodendrimers generally requires multiple steps,
making it challenging and time-consuming. Additionally, it is hard
to achieve adaptive rearrangements in the polymers and therefore,
assembled structures, e.g., supramolecular copolymers, are seen as
attractive alternatives.Inspired by the architectures obtained
through self-organization
of biological macromolecules into complex but highly ordered matter,
supramolecular polymers attained much attention due to their resemblance
in self-assembled properties.[10,11] Dynamic, adaptable,
and responsive supramolecular materials, including supramolecular
glycopolymers, can be formed from monomeric building blocks exploiting
noncovalent interactions in the form of hydrophobic effects, directional
hydrogen bonding, and coordination interactions or π–π
stacking.[12−14] For instance, Stupp and co-workers have studied a
series of glycopeptides that orthogonally self-assemble into nanofibers
driven by hydrogen bonding and hydrophobic effects, showing high potential
application in the stabilization of growth factors and bone regeneration.[15] Brunsveld and co-workers developed a mannose
functionalized discotic molecule that can copolymerize to form supramolecular
polymers with tunable bacterial aggregation efficacy.[16] We have previously synthesized a library of benzene-1,3,5-tricarboxamide(BTA)-based
supramolecular monomers that could be functionalized at their periphery
via azide–alkyne click reactions with monosaccharides and showed
that these could be used to generate fibrous assemblies in water,
however without internal order.[17,18]The modular properties
of supramolecular copolymerization in principle
opens up endless possibilities to create new materials exhibiting
enhanced complexities and functionalities.[18] It is challenging however to achieve the subtle balance between
the complementary recognizing motifs through noncovalent heterointeractions
in supramolecular copolymers. Although some elegant examples have
been reported in organic solvents, e.g., kinetically
controlled living supramolecular copolymers,[19−22] only limited aqueous examples
have been reported so far. Manners and co-workers have published seminal
contributions on crystalline driven supramolecular copolymerization
in aqueous solution, resulting in various morphologies in a living
fashion.[23,24] Oppositely charged polypeptide supramolecular
monomers were demonstrated by Besenius and co-workers to copolymerize
into fibrous structures at neutral pH, whereas they disassembled at
low and high pH.[25]In our continued
efforts toward revealing the underlying mechanisms
and structure-kinetics-function relationship on supramolecular (co)polymerization
in water, we have explored a BTA-based platform that undergoes supramolecular
(co)polymerization to yield different nanostructures.[26−28] The most extensively studied monomer is BTA-OEG4 (Scheme B) bearing a tetraethylene
glycol as periphery, that self-assembles into micrometers long nanofibers
driven by hydrophobic effects and intermolecular hydrogen bonding.
Most recently, we showed that a dendronized BTA (dBTA), which does
not self-assembles by itself, can be copolymerized with BTA-OEG4 into a more stable nanofibrous structure exhibiting slower
exchange dynamics for both monomers as compared to the individual
assemblies.[27] This copolymerization approach
provides opportunities to build architectures with controlled dynamics
and functional group distributions at the periphery, and the possibility
to generate hydrogels with tunable mechanical properties at higher
concentrations. Today, most of the hydrogels developed contain poly(ethylene
glycol) for water-solubility, which may limit in vivo applications
and lacks the option for functionalization.[30−32] Carbohydrates
can serve as alternatives allowing water solubility and providing
means for tailored cell recognition. In order to control the dynamics,
functionality, and stability of the carbohydrate-based supramolecular
polymers, new avenues have to be discovered to construct complex multicomponent
supramolecular glycopolymers.
Scheme 1
Synthesis Pathways of the BTA-Saccharides
(A) BTA-Glc, BTA-Man, and
BTA-Cel were synthesized by reacting BTA-C12-OH with participating
imidate donors followed by deprotection with NaOMe. (B) BTA-OEG4-Man was synthesized through glycosylation with 2,3,4,6-tetra-O-benzoyl-α-d-mannopyranosyl trichloroacetimidate
to BTA-OEG4 and followed by deprotection with NaOMe.
Synthesis Pathways of the BTA-Saccharides
(A) BTA-Glc, BTA-Man, and
BTA-Cel were synthesized by reacting BTA-C12-OH with participating
imidate donors followed by deprotection with NaOMe. (B) BTA-OEG4-Man was synthesized through glycosylation with 2,3,4,6-tetra-O-benzoyl-α-d-mannopyranosyl trichloroacetimidate
to BTA-OEG4 and followed by deprotection with NaOMe.Herein, by taking advantage of a modular supramolecular
(co)assembly
methodology, a variety of saccharide functionalized supramolecular
polymers and copolymers are generated, incorporating different carbohydrates
of high interest for further biological applications. A family of
saccharide functionalized BTA-based monomers is synthesized via acid
catalyzed glycosylation installing monosaccharides (β-d-glucose, α-d-mannose) or a disaccharide (β-d-cellobiose) on the hydrophobic core (Scheme A). In addition a BTA-tetraethylene glycol
was functionalized with an α-d-mannose to generate
a monomeric building block having the carbohydrate further away from
the BTA core. The saccharides presented at the periphery of the supramolecular
structures are selected as possible antifouling saccharides (glucose
and cellobiose) to reduce aspecific cell interactions, and their antimicrobial
properties (mannose), possibly valuable for future cell experiments.[33] The homopolymerization and copolymerization
behavior of the saccharide based monomeric building blocks has been
elucidated in an aqueous environment, showing the critical role of
the carbohydrate appendage for the self-assembly behavior and internal
order, opening up avenues for biomaterial science.
Results and Discussion
Molecular
Design and Synthesis
A modular synthetic
platform was developed to easily install different saccharides onto
a BTA-based core using a glycosylation reaction. A hydrophobic alkyl
linker of 12 carbons was selected to protect the inner 3-fold hydrogen
bonds from water penetration and to drive assembly by a hydrophobic
collapse. This length was previously proven to be the most optimal
spacing for the ethylene glycol BTA (BTA-OEG4) 1D fibrous
assembly.[34] To this end, a C12 linker was
attached to the benzene core with hydroxyl end groups at the periphery
(BTA-C12-OH, Scheme A). For the glycosylation of d-glucose, d-mannose
and d-cellobiose, reactive trichloroacetimidate donors protected
with participating benzoyl esters were selected to achieve glycosylations
with high 1,2-trans stereoselectivity.[35] The solubility of the starting material in conventional glycosylation
solvents was too low to allow for productive reactions and hexafluoro-iso-propanol was found optimal for the condensation reactions.
Notably, because of the low nucleophilicity of the protic solvent,
minimal hexafluoro-iso-propyl glycosides were formed.
Next, the protecting groups were removed to yield BTA-β-d-glucose (BTA-Glc), BTA-α-d-mannose (BTA-Man),
and BTA-β-d-cellobiose (BTA-Cel) in 31%, 45%, and 12%
overall yield, respectively (Scheme A). Similarly, as shown in Scheme B, BTA-OEG4 was used as precursor
for the synthesis of BTA-OEG4-α-d-mannose
(BTA-OEG4-Man) via direct glycosylation, with an overall
yield of 56%. The integrity and purity of the BTA structures were
confirmed by 1H NMR, 13C NMR, FT-IR spectroscopy,
matrix assisted laser desorption ionization-time-of-flight mass spectrometry
(MALDI-TOF-MS), and liquid chromatography–mass spectrometry
(LC-MS) analysis (Figures S21–26 of the Supporting Information, SI).
Assembly of BTA-Glc, BTA-Man, BTA-Cel, and
BTA-OEG4-Man
We first studied the supramolecular
homopolymerization
of BTA-Cel and BTA-OEG4-Man in MQ water at pH
6.4, using a previously described preparation protocol.[34] To this end BTA-saccharide powder was weighed
in a vial and dissolved in MQ water, after which the samples were
vortexed, heated, vortexed, and allowed to assemble at room temperature,
followed by equilibration overnight prior to measuring. This protocol,
however, was not suitable for BTA-Glc and BTA-Man, since large particles
were unable to reach the (sub)molecular dissolved state in water at
high temperatures,[36,37] and therefore, injection from
a cosolvent was required. Thus, BTA-Glc or BTA-Man was dissolved in
methanol to obtain a concentrated stock solution, which was subsequently
injected into MQ water. After overnight equilibration, dialysis against
MQ water was performed to remove the methanol content. The pH of the
assembled BTA-saccharide samples was about 7 and remained stable over
time.To investigate the resulting supramolecular polymers,
the samples were probed by UV spectrophotometry. Both the BTA-Glc
and BTA-Man showed two absorption maxima at 211 and 225 nm (Figure A) similar to the
BTA-OEG4 previously reported, and indicative of a fibrous
assembly.[28] The UV spectrum of the BTA-Glc
or BTA-Man upon heating remained similar until 60 °C, whereas
at 70 °C, a different spectrum was recorded with a broad maximum
at lower wavelengths (Figure B and 1C) suggesting disassembly at
this temperature. In contrast, BTA-Cel and BTA-OEG4-Man
in water showed an absorption maximum at 196 nm (Figure A), which gradually shifted
to 198 upon heating (Figure S1). A similar
absorbance at around 197 nm was also observed for the previously reported
dBTA, which was shown to form spherical micelles instead of 1D nanofibers.[29] The UV spectra of the BTA-Glc, BTA-Man, BTA-Cel,
and BTA-OEG4-Man thus suggest different assemblies, with
the first two forming 1D assemblies, and the latter two generating
smaller micellar aggregates. In line with these findings, continuous
heating UV experiments and micro-DSC heating experiments showed clear
disassembly temperatures for the BTA-Glc and BTA-Man assemblies, which
were absent in the BTA-Cel and BTA-OEG4-Man samples (Figures S2 and S3).
Figure 1
UV (A–C) and CD
(D–F) spectra of the assembled individual
molecules in water. (A) BTA-Glc and BTA-Man in water show the typical
BTA maxima at 211 and 225 nm, whereas BTA-Cel and BTA-OEG4-Man show maxima at 196 nm. (B) and (C) Stepwise heating of BTA-Glc
and BTA-Man, respectively, from 20 °C to 80 °C showed a
change in aggregation between 60 °C and 70 °C for both of
BTA-Glc and BTA-Man. (D) BTA-Glc and BTA-Man show a biphasic negative
and positive mirror Cotton effect, whereas BTA-Cel and BTA-OEG4-Man are CD silent. (E) and (F) Stepwise heating of BTA-Glc
and BTA-Man, respectively, from 20 to 80 °C showed a drop in
ellipticity between 60 °C and 80 °C. Samples were equilibrated
for 16 min at the designated temperature to allow the formation of
equilibrated aggregates. (cBTA = 50 μM).
UV (A–C) and CD
(D–F) spectra of the assembled individual
molecules in water. (A) BTA-Glc and BTA-Man in water show the typical
BTA maxima at 211 and 225 nm, whereas BTA-Cel and BTA-OEG4-Man show maxima at 196 nm. (B) and (C) Stepwise heating of BTA-Glc
and BTA-Man, respectively, from 20 °C to 80 °C showed a
change in aggregation between 60 °C and 70 °C for both of
BTA-Glc and BTA-Man. (D) BTA-Glc and BTA-Man show a biphasic negative
and positive mirror Cotton effect, whereas BTA-Cel and BTA-OEG4-Man are CD silent. (E) and (F) Stepwise heating of BTA-Glc
and BTA-Man, respectively, from 20 to 80 °C showed a drop in
ellipticity between 60 °C and 80 °C. Samples were equilibrated
for 16 min at the designated temperature to allow the formation of
equilibrated aggregates. (cBTA = 50 μM).Although the peripheral saccharides are situated
far from the hydrogen
bonding core, we observed a preference in helicity upon assembly by
employing circular dichroism (CD) spectroscopy. The CD spectra of
BTA-Cel and BTA-OEG4-Man in water showed a negligible CD
profile (Figure D).
In contrast, a biphasic Cotton effect with minima at 211 and 250 nm
was observed for BTA-Glc and a mirror image biphasic Cotton effect
for BTA-Man. The CD profile of BTA-Glc is similar to that of the CD
of a chiral α-deuteriumBTA-fiber previously reported.[34] As depicted in Figure E and 1F, the CD pattern
remained similar upon heating, with a sudden drop in intensity at
70 °C and became gradually CD silent at 80 °C. This temperature
behavior supports the UV measurements, indicating stable structures
up to 60 °C.Static light scattering (SLS) was performed
to further inform on
potential supramolecular architectures. SLS measures the scattering
intensity as a dependence of the scattering angle, which can provide
information about size, shape, and molar mass of the particles. In
the BTA-Glc and BTA-Man samples, an angle dependent scattering intensity
was observed, indicating elongated structures (Figure S4). In contrast, BTA-Cel and BTA-OEG4-Man
showed low scattering intensities at all angles and an absence of
angular dependency, indicating small spherical-like aggregation.Subsequently, cryogenic transmission electron microscopy (cryo-TEM)
was employed to obtain structural information on the assemblies. Micrometers
long fibers for BTA-Glc and BTA-Man were observed, with a diameter
between 5 and 10 nm (Figure A and 2B), whereas small micelles of
ca. 5 nm in size for BTA-Cel and BTA-OEG4-Man were detected
(Figure C and 2D). Detailed analysis shows in some cases a periodic
feature due to twisting or aggregation; a detail that needs further
studies. BTA-Man formed bundled fibers as shown in Figure B, possibly as a result of
mannose–mannose interactions.[36,37] Self-assembly
and accompanied bundling occurred within seconds, whereas increased
order developed over hours (Figure S5).
Occasionally, a few (bundled) fibers were observed in the BTA-Cel
sample as well, but these were not representative for the content
of the sample. Altogether, these data show that BTA-Glc and BTA-Man
assemble into 1D nanofibers, whereas BTA-Cel and BTA-OEG4-Man do not, indicating that fiber formation is a subtle balance
between hydrophobic to hydrophilic ratio, steric effects of the carbohydrates
and carbohydrate-carbohydrate interactions.[38]
Figure 2
Cryo-TEM
of BTA-Glc (A), BTA-Man (B), BTA-Cel (C), and BTA-OEG4-Man
(D). Micrometers long fibrous structures were observed
for BTA-Glc and BTA-Man while micellar structures for BTA-Cel and
BTA-OEG4-Man. Scale bars indicate 50 nm, cBTA = 250 or 500 μM.
Cryo-TEM
of BTA-Glc (A), BTA-Man (B), BTA-Cel (C), and BTA-OEG4-Man
(D). Micrometers long fibrous structures were observed
for BTA-Glc and BTA-Man while micellar structures for BTA-Cel and
BTA-OEG4-Man. Scale bars indicate 50 nm, cBTA = 250 or 500 μM.Finally, the exchange dynamics of the monomers within the homopolymers
was investigated by employing hydrogen/deuterium exchange mass spectrometry
(HDX-MS), previously utilized in our group.[17,26,30] This label-free method allows the determination
of solvent exchanged monomers, due to an increase in molecular weight
as a result of deuterium exchange of the hydrogen atoms in hydroxyls
and amines. Upon contact with deuterated water, outer hydroxyl groups
readily exchange, whereas the three inner amides exchange simultaneously
upon monomer migration. Concentrated stock solutions of homopolymers
(500 μM or 250 μM in the case of BTA-Man, due to the instability
at higher concentration) were prepared in H2O and subsequently
diluted 100 times in D2O, after which mass spectra through
electrospray ionization mass spectrometry (ESI-MS) were collected
as a function of time. As expected, all hydroxyl groups exchanged
to OD immediately after dilution into D2O, whereas the
well protected inner NH groups in BTA-Man and BTA-Glc exchanged gradually
(Figure S6). Similar to the BTA-OEG4, the data could be fitted with a triexponential fit, indicating
monomers with different mobilities. The fit of BTA-Man followed a
slower decay as compared to BTA-OEG4, demonstrating more
rigid and stable aggregates. In contrast, BTA-Cel and BTA-OEG4-Man were instantaneously fully deuterated indicating a very
weak or absent supramolecular ordered structure, substantiating the
findings described above.
Coassembly of BTA-Glc, BTA-Man, BTA-Cel,
BTA-OEG4-Man and BTA-OEG4
To generate
supramolecular
polymers having the β-d-cellobiose, β-d-glucose or α-d-mannose exposed at the periphery in
different densities, we next turned to the generation of copolymers
to fully explore the scope of the modular BTA-saccharides approach.
First, the supramolecular copolymerization of BTA-Cel or BTA-OEG4-Man was explored with the monomers that are capable of generating
stable fibers, BTA-Glc, BTA-Man or BTA-OEG4. To this end
coassembled structures were prepared by mixing the individual assembled
stocks in a 2:1, 1:1, or 1:2 ratio, following the assembly preparation
protocol described above. Co-assembling BTA-Cel and BTA-Glc resulted
in the typical absorption maxima at 211 and 225 nm (Figure A), indicating the successful
copolymerization into fibrous structures. The corresponding CD-spectra
showed a decrease in ellipticity with increasing BTA-Cel content,
suggesting less well-ordered conformations in the supramolecular copolymer
(Figure D). BTA-Cel
copolymerized with BTA-OEG4 also provided similar UV spectra
(Figure C), whereas
the corresponding CD-spectra showed a reversed biphasic Cotton effect
as compared to BTA-Glc, which changed only moderately with changing
monomer ratio. Of note, the chirality of this helix is opposite to
that of the polymer, even though the chirality of the appendages,
consisting both of β-d-glucose-type moieties, is similar.
The stable CD effect of the BTA-Cel/BTA-OEG4 copolymers
suggests that stable fiber can be formed as a result of the proper
lateral alignment between the cellobiose moieties, which cannot be
attained in structures composed of solely BTA-Cel monomers.[39−42] The coassembly of BTA-Glc with BTA-OEG4 again showed
a similar UV-absorption profile (Figure B), with a significant loss of chirality
(Figure E). Assemblies
formed by BTA-OEG4 showed a large CD and LD (linear dichroism)
effect, probably due to a macroscopic orientation rather than due
to chirality,[36,37] therefore, the CD spectra of
the BTA-OEG4 were not used for analyses. The LD signal
in all other samples was small and therefore had a minimal influence
on the CD (Figure S7).
Figure 3
UV (top) and CD (bottom)
spectra of coassembled BTAs in water at
20 °C. (A) and (D) BTA-Glc coassembled with BTA-Cel. (B) and
(E) BTA-Glc coassembled with BTA-OEG4. (C) and (F) coassembly
of BTA-Cel with BTA-OEG4. CD spectra of BTA-OEG4 were discarded due to a large LD effect. Note that the 1:2 Glc:OEG4 mixture displayed a small negative LD effect. (cBTA, total = 50 μM).
UV (top) and CD (bottom)
spectra of coassembled BTAs in water at
20 °C. (A) and (D) BTA-Glc coassembled with BTA-Cel. (B) and
(E) BTA-Glc coassembled with BTA-OEG4. (C) and (F) coassembly
of BTA-Cel with BTA-OEG4. CD spectra of BTA-OEG4 were discarded due to a large LD effect. Note that the 1:2 Glc:OEG4 mixture displayed a small negative LD effect. (cBTA, total = 50 μM).In order to further investigate the coassembled aggregates, the
existence of a hydrophobic pocket was investigated using a nile red
fluorescence assay.[38] In water, nile red
shows very low fluorescence at around 665 nm (Figure S8A–C), while in contact with a hydrophobic
environment an increase in fluorescence with a blue shift of the maximum
is detected. At a total BTA concentration of 50 μM, all measured
samples showed a blue shift to around 615 nm, with an increase in
fluorescence intensity as compared to nile red in water, thus indicative
of a hydrophobic pocket. No significant differences between the homo-
and copolymers could be deduced form the assay.Next, SLS measurements
of the mixtures were performed to investigate
the aggregation size in more detail. All mixtures showed similar aggregation
size as compared to BTA-OEG4, further supporting the formation
of fibrous assemblies (Figure S8D–F). In addition, small-angle X-ray scattering (SAXS) was employed
for the BTA-Cel assemblies with and without BTA-OEG4. The
SAXS profile of the BTA-Cel, incapable of forming fibers by itself,
displayed a plateau at lower q values, indicating
small spherical-like particles, whereas the BTA-Cel:BTA-OEG4 coassemblies showed a slope at lower q, indicating
long fibrous structures (Figure S9). The
dimensions obtained from the fits of the SAXS data corroborate well
with the dimensions measured in cryo-TEM (Figure ) and TEM of the 1:1 BTA-Cel:BTA-OEG4 (Figure C).
Cryo-TEM of the other 1:1 mixtures (BTA-Glc:BTA-Cel and BTA-Glc:BTA-OEG4) also showed micrometers long fibrous structures and in some
cases a periodic feature was observed similar as that of the homopolymers.
These data show that BTA-Cel monomers are able to coassemble with
BTA-OEG4 and BTA-Glc monomers to form 1D supramolecular
fibrous structures instead of self-sorting into individual aggregates.
Figure 4
Cryo-TEM
of 1:1 mixtures. (A) Glc:Cel, (B) Glc:OEG4,
(C) Cel:OEG4, (D) OEG4-Man:OEG4,
(E) Man:OEG4, and (F) Man:OEG4-Man. (C) TEM
without staining. Fibrous structures and micelles were observed. cBTA = 500 μM.
Cryo-TEM
of 1:1 mixtures. (A) Glc:Cel, (B) Glc:OEG4,
(C) Cel:OEG4, (D) OEG4-Man:OEG4,
(E) Man:OEG4, and (F) Man:OEG4-Man. (C) TEM
without staining. Fibrous structures and micelles were observed. cBTA = 500 μM.The influence of aging on the self-assembled structures was investigated
since this may enhance the helical order as a result of improved packing
over time. Most of the samples investigated did not show a difference
in UV and CD upon aging the samples for 2 months, except for the mixtures
BTA-Cel:BTA-OEG4 (1:1) and BTA-Glc:BTA-Cel (1:1) (Figure S10). In the case of BTA-Cel:BTA-OEG4, the chirality was slightly enhanced (Figure S10B). In contrast, BTA-Glc:BTA-Cel went from a CD
silent profile at day 1, to a CD profile displaying a Cotton effect
after 2 months. LD did not affect this behavior (Figure S10C). This indicates slow readjustments into a more
stable supramolecular conformation exhibiting a strong CD effect highlighting
the need for careful alignment to induce helical order.Overall,
the data clearly show that all studied copolymerizations
led to 1D assemblies. The highest induction of optical activity, hence
indicating the highest order in the helical aggregate, is observed
for the BTA-Glc homopolymers, which decreased when coassembled with
BTA-Cel and which almost vanished when copolymerized with BTA-OEG4. In contrast, the coassembly of BTA-Cel with BTA-OEG4 showed an induction of optical activity with a similar CD
intensity for all the mixtures measured. The reduction in CD intensity
might be the result of induced steric hindrance when the more bulky
BTA-Cel is incorporated in BTA-Glc, thereby disordering the fiber
backbone.Similarly, the supramolecular copolymerization of
BTA-Man, BTA-OEG4-Man, and BTA-OEG4, as well
as BTA-Glc with BTA-Man
at different mixing ratios of 2:1, 1:1, and 1:2 was investigated with
UV, CD, SLS (Figures S11–14) and
cryo-TEM (Figure D–F).
Combined, these techniques proved the formation of 1D micrometers
long fibers with diameters of ca. 8 nm. The exchange dynamics of the
copolymers (1:1 ratio) were monitored by HDX-MS, revealed different
monomer exchange profiles for all mixtures measured (Figures and S15). This substantiates the successful copolymerization, since self-sorting
would result in similar exchange profiles as the homopolymers (Figure , 50% compared to
their 100% counterpart). Upon copolymerizing BTA-OEG4-Man
with BTA-OEG4 a slower monomer exchange as compared to
their 100% counterpart is observed (Figure A), suggesting a stabilization effect upon
coassembly. This stabilization behavior was previously observed when
dBTA was copolymerized with BTA-OEG4.[27] In contrast, an increased monomer exchange was observed
when BTA-Man was copolymerized with either BTA-OEG4 or
BTA-OEG4-Man (Figure B and 5C), suggesting a destabilizing
effect. In other words by copolymerization of different comonomers,
the exchange dynamics of BTA-Man can be tuned.
Figure 5
HDX-MS curves of homopolymers
and copolymers (1:1 ratio) after
100 times dilution into D2O. The graphs highlight the amount
of remaining unexchanged monomers (BTA3NH) as a function of time.
(A) BTA-OEG4-Man coassembled with BTA-OEG4,
(B) BTA-Man with BTA-OEG4, and (C) BTA-OEG4-Man
and BTA-Man. The data was fitted with a triexponential fit. cBTA = 5 μM.
HDX-MS curves of homopolymers
and copolymers (1:1 ratio) after
100 times dilution into D2O. The graphs highlight the amount
of remaining unexchanged monomers (BTA3NH) as a function of time.
(A) BTA-OEG4-Man coassembled with BTA-OEG4,
(B) BTA-Man with BTA-OEG4, and (C) BTA-OEG4-Man
and BTA-Man. The data was fitted with a triexponential fit. cBTA = 5 μM.
The Influence of Sample Preparation and Purity on Self-Assembly
During these studies we were confronted with subtleties that are
worth publishing. The sample preparation method is critical in the
assembly to avoid undesirable kinetic traps and is key to reproducibility.
As mentioned in the paragraph on homopolymerization, BTA-Glc and BTA-Man
require a cosolvent to allow assembly to occur since the carbohydrate–carbohydrate
interactions in the solid are too strong to be disrupted by solely
water addition and thermal energy. Interestingly, when BTA-Glc was
mixed with 15% of the 2-armed BTA-Glc—unforeseen encountered
with a less pure sample—water solubility was readily obtained
without the need for a cosolvent. Not only the water solubility has
changed, also the CD pattern was different (Figure E) as compared to pure BTA-Glc (Figure D). Contrary, the
assemblies were shown to be more stable at higher temperatures, since
the transition temperature for disassembly was shifted from 60 °C
to 70–80 °C (Figures A, 6B, and S16). Allowing the 2-armed BTA-Glc to self-assemble in water
revealed an even higher thermal stability reaching beyond 80 °C,
although precipitation was observed by scattering at higher wavelengths
(Figure C). Co-assembling
the mixture of BTA-Glc (85% BTA-3Glc, 15% BTA-2Glc) with BTA-Cel or
BTA-OEG4 (Figure S17) showed
subtle differences in UV and CD patterns as well as compared to copolymerized
pure BTA-Glc (Figure A and 3B). Moreover, solubility and self-assembly
in buffer was investigated as well (Figure S18). BTA-Glc and BTA-Man in phosphate buffered saline (PBS) precipitated,
whereas BTA-Cel and BTA-OEG4-Man readily dissolved, although
the latter did not self-assemble into 1D fibers. In contrast, all
copolymers did dissolve and form elongated assemblies in buffer similar
to their counterparts in pure water. These results highlight that
the composition of the samples have not only a profound effect on
the water and buffer solubility but also on the nanoscopic ordering
of the structure, hence studying copolymers is highly important in
understanding complex assembly processes.
Figure 6
Influence of purities
on BTA-Glc assembly. A–C indicates
UV spectra and D–F CD spectra. (A) and (D) pure BTA-Glc showing
a transition between 60 and 70 °C. (B) and (E) 85% BTA-3Glc
mixed with 15% BTA-2Glc showing disassembly at 80 °C. (C) and
(F) BTA-2Glc revealing a stability beyond 80 °C. Samples were
equilibrated for 16 min at the designated temperature to allow the
formation of equilibrated aggregates. (cBTA = 50 μM).
Influence of purities
on BTA-Glc assembly. A–C indicates
UV spectra and D–F CD spectra. (A) and (D) pure BTA-Glc showing
a transition between 60 and 70 °C. (B) and (E) 85% BTA-3Glc
mixed with 15% BTA-2Glc showing disassembly at 80 °C. (C) and
(F) BTA-2Glc revealing a stability beyond 80 °C. Samples were
equilibrated for 16 min at the designated temperature to allow the
formation of equilibrated aggregates. (cBTA = 50 μM).
Hydrogel Formation
The formation of hydrogels was investigated
with selected copolymers capitalizing on the fiber entanglement, most
strongly observed for the BTA-Glc fibers. Hydrogel formation of BTA-Man
with BTA-OEG4 (1:2, HG1), BTA-OEG4-Man with BTA-OEG4 (1:2, HG2) and BTA-Glc
with BTA-OEG4 (1:2, HG3) was optimized starting
from the sample preparation protocol described above. Fast cooling
in an ice bath after heating and vortexing proved to be necessary
to form a more transparent hydrogel (5 wt %) as compared to cooling
at room temperature. The mechanical properties of the hydrogels were
investigated by rheological measurements. Strain dependent oscillatory
rheology of the hydrogels displayed a linear region with a larger
storage modulus (G′) as compared to the loss
modulus (G′′) indicating a viscoelastic
material (Figure ).
Moreover, a gel-to-sol transition was observed upon 400% strain (Figure A–C). Viscoelastic
hydrogels were formed in all cases with low storage moduli resulting
in extremely soft gels (G′ ≈ 42 Pa, G′ ≈ 20 Pa and G′ ≈ 47 Pa). The storage modulus of HG1 and HG3 are similar when compared to hydrogels formed from BTA-OEG4 (Figure S19), whereas HG2 shows lower mechanical properties. The anticipated self-healing
properties of the hydrogels were investigated in step–strain
measurements revealing fast recovery (Figure S20).
Figure 7
Rheology measurements of 5 wt % hydrogels formed by BTA-Man: BTA-OEG4 (1:2, HG1, A,D), BTA-OEG4-Man: BTA-OEG4 (1:2, HG2, B,E), and BTA-Glc: BTA-OEG4 (1:2 HG3, C,F) at 37 °C showing the storage and
loss moduli (G′, G′′).
(A–C) Strain dependent oscillatory rheology (fixed angular
frequency of 1 rad/s) with a photograph of the inverted vial containing
the corresponding hydrogel in the inset. (D–F) Frequency sweep
measurements (fixed applied strain of 1%).
Rheology measurements of 5 wt % hydrogels formed by BTA-Man: BTA-OEG4 (1:2, HG1, A,D), BTA-OEG4-Man: BTA-OEG4 (1:2, HG2, B,E), and BTA-Glc: BTA-OEG4 (1:2 HG3, C,F) at 37 °C showing the storage and
loss moduli (G′, G′′).
(A–C) Strain dependent oscillatory rheology (fixed angular
frequency of 1 rad/s) with a photograph of the inverted vial containing
the corresponding hydrogel in the inset. (D–F) Frequency sweep
measurements (fixed applied strain of 1%).These initial proof-of-concept experiments show that by carefully
selecting monomers, hydrogels can be fabricated with specific functionalities
due to supramolecular copolymerization. Although the prepared hydrogels
are extremely weak, it is anticipated that the mechanical properties
can be further enhanced by increasing the wt % or including cross-links.
Together with their self-healing properties, their modular approach
and the option to be modified with specific carbohydrates at the periphery,
these supramolecular hydrogels are very attractive candidates for
biomedical applications, e.g., the culture of soft tissues.
Conclusions
Inspired by the various functions of polysaccharides present in
the glycocalyx and the extracellular matrix, biomaterials based on
poly- and oligosaccharides have been previously developed. Here, BTA-based
supramolecular polymers presenting mono- (β-d-glucose
and α-d-mannose) and disaccharides (β-d-cellobiose) at the periphery were developed as ethylene glycol substitutes
previously described and interrogated on their self-assembly properties
using several techniques. A combination of UV, CD, light scattering
and cryo-TEM proved to be powerful in elucidating the aggregation
behavior into great detail. BTA-Glc and BTA-Man self-assembled into
chiral 1D helical supramolecular polymers. In contrast, BTA-Cel and
BTA-OEG4-Man formed small spherical-like assemblies instead.
Probably due to the bulky and highly hydrophilic character of BTA-Cel
and BTA-OEG4-Man, spherical micelles are favored over 1D
fibers. In contrast, coassembling BTA-Cel or BTA-OEG4-Man
with BTA-Glc, BTA-Man, or BTA-OEG4 resulted in 1D fibrous
supramolecular copolymers with BTA-Cel or BTA-OEG4-Man incorporated
in the supramolecular copolymers. Interestingly, chirality was induced
upon mixing BTA-Cel with BTA-OEG4, in contrast to a reduction
or even suppression of chirality when BTA-Glc was copolymerized with
BTA-Cel or BTA-OEG4. This highlights that the chiral order
of the supramolecular aggregates is the result of careful packing
of the peripheral chiral saccharides and not due directly to the cooperative
effect of the core. The supramolecular structures generated and investigated
here provide a deeper understanding of supramolecular copolymerization—an
area with many unknowns to be discovered—and taught us that
subtle changes in monomer structure, preparation, and aging can have
a profound effect on the packing of the monomers as well as their
stability in time. In future experiments the mechanical properties
in the gel state can be further tuned by controlling the weight percentage
and ratio of different BTA-based monomers, and more functionality
can be incorporated by mixing in, e.g., peptide functionalized BTAs.
Carefully selecting monomers in supramolecular copolymerization opens
avenues to fabricate endless variants of multicomponent functional
polymeric materials, having functionalities that are not accessible
in supramolecular homopolymers.
Authors: Sietze Reitsma; Dick W Slaaf; Hans Vink; Marc A M J van Zandvoort; Mirjam G A oude Egbrink Journal: Pflugers Arch Date: 2007-01-26 Impact factor: 3.657
Authors: Tom F A De Greef; Maarten M J Smulders; Martin Wolffs; Albert P H J Schenning; Rint P Sijbesma; E W Meijer Journal: Chem Rev Date: 2009-11 Impact factor: 60.622
Authors: Jonathan K Armstrong; Georg Hempel; Susanne Koling; Linda S Chan; Timothy Fisher; Herbert J Meiselman; George Garratty Journal: Cancer Date: 2007-07-01 Impact factor: 6.860
Authors: Pierre Carçabal; Rebecca A Jockusch; Isabel Hünig; Lavina C Snoek; Romano T Kroemer; Benjamin G Davis; David P Gamblin; Isabelle Compagnon; Jos Oomens; John P Simons Journal: J Am Chem Soc Date: 2005-08-17 Impact factor: 15.419
Authors: Giulia Morgese; Bas F M de Waal; Silvia Varela-Aramburu; Anja R A Palmans; Lorenzo Albertazzi; E W Meijer Journal: Angew Chem Int Ed Engl Date: 2020-08-18 Impact factor: 15.336
Authors: Marle E J Vleugels; Silvia Varela-Aramburu; Bas F M de Waal; Sandra M C Schoenmakers; Beatriz Maestro; Anja R A Palmans; Jesús M Sanz; E W Meijer Journal: Biomacromolecules Date: 2021-11-30 Impact factor: 6.988
Authors: Sandra M C Schoenmakers; Bart W L van den Bersselaar; Shikha Dhiman; Lu Su; Anja R A Palmans Journal: Org Biomol Chem Date: 2021-10-06 Impact factor: 3.876
Authors: Shahzad Hafeez; Huey Wen Ooi; Dennis Suylen; Hans Duimel; Tilman M Hackeng; Clemens van Blitterswijk; Matthew B Baker Journal: J Am Chem Soc Date: 2022-02-23 Impact factor: 15.419
Authors: Ciqing Tong; Joeri A J Wondergem; Marijn van den Brink; Markus C Kwakernaak; Ying Chen; Marco M R M Hendrix; Ilja K Voets; Erik H J Danen; Sylvia Le Dévédec; Doris Heinrich; Roxanne E Kieltyka Journal: ACS Appl Mater Interfaces Date: 2022-04-10 Impact factor: 10.383
Authors: Shahzad Hafeez; Fiona R Passanha; Antonio J Feliciano; Floor A A Ruiter; Afonso Malheiro; René P M Lafleur; Nicholas M Matsumoto; Clemens van Blitterswijk; Lorenzo Moroni; Paul Wieringa; Vanessa L S LaPointe; Matthew B Baker Journal: Biomater Sci Date: 2022-08-24 Impact factor: 7.590
Authors: René P M Lafleur; Svenja Herziger; Sandra M C Schoenmakers; Arthur D A Keizer; Jahaziel Jahzerah; Bala N S Thota; Lu Su; Paul H H Bomans; Nico A J M Sommerdijk; Anja R A Palmans; Rainer Haag; Heiner Friedrich; Christoph Böttcher; E W Meijer Journal: J Am Chem Soc Date: 2020-09-29 Impact factor: 15.419
Authors: Silvia Varela-Aramburu; Giulia Morgese; Lu Su; Sandra M C Schoenmakers; Mattia Perrone; Luigi Leanza; Claudio Perego; Giovanni M Pavan; Anja R A Palmans; E W Meijer Journal: Biomacromolecules Date: 2020-09-29 Impact factor: 6.988