Tuning gelation time and morphology of injectable hydrogels using Ketone-Hydrazide cross-linking

Article abstract of DOI:10.1021/bm401615d

Injectable, covalently in situ forming hydrogels based on poly(N-isopropylacrylamide) have been designed on the basis of mixing hydrazide-functionalized nucleophilic precursor polymers with electrophilic precursor polymers functionalized with a combination of ketone (slow reacting) and aldehyde (fast reacting) functional groups. By tuning the ratio of aldehyde:ketone functional groups as well as the total number of ketone groups in the electrophilic precursor polymer, largely independent control over hydrogel properties including gelation time (from seconds to hours), degradation kinetics (from hours to months), optical transmission (from 1 to 85%), and mechanics (over nearly 1 order of magnitude) can be achieved. In addition, ketone-functionalized precursor polymers exhibit improved cytocompatibility at even extremely high concentrations relative to polymers functionalized with aldehyde groups, even at 4-fold higher functional group densities. Overall, increasing the ketone content of the precursor copolymers can result in in situ-gellable hydrogels with improved transparency and biocompatibility and equivalent mechanics and stimuli-responsiveness while only modestly sacrificing the speed of gel formation.

Full text of DOI:10.1021/bm401615d

Article
pubs.acs.org/Biomac
Tuning Gelation Time and Morphology of Injectable Hydrogels Using
Ketone−Hydrazide Cross-Linking
Mathew Patenaude, Scott Campbell, Dennis Kinio, and Todd Hoare*
Department of Chemical Engineering, McMaster University, 1280 Main Street West, Hamilton, Ontario, Canada L8S 4L7
*
S Supporting Information
ABSTRACT: Injectable, covalently in situ forming hydrogels
based on poly(N-isopropylacrylamide) have been designed on
the basis of mixing hydrazide-functionalized nucleophilic
precursor polymers with electrophilic precursor polymers
functionalized with a combination of ketone (slow reacting)
and aldehyde (fast reacting) functional groups. By tuning the
ratio of aldehyde:ketone functional groups as well as the total
number of ketone groups in the electrophilic precursor
polymer, largely independent control over hydrogel properties
including gelation time (from seconds to hours), degradation
kinetics (from hours to months), optical transmission (from 1
to 85%), and mechanics (over nearly 1 order of magnitude)
can be achieved. In addition, ketone-functionalized precursor
polymers exhibit improved cytocompatibility at even extremely high concentrations relative to polymers functionalized with
aldehyde groups, even at 4-fold higher functional group densities. Overall, increasing the ketone content of the precursor
copolymers can result in in situ-gellable hydrogels with improved transparency and biocompatibility and equivalent mechanics
and stimuli-responsiveness while only modestly sacrificing the speed of gel formation.
INTRODUCTION
conditions have been developed, including Michael-type
■
addition between a thiol-functionalized precursor and α,β-
Hydrogels have attracted widespread interest as biomaterials
due to their analogous interfacial, mechanical, and phys-
icochemical properties to many soft tissues. By controlling the
1
7−20
unsaturated acid moieties of another precursor,
Diels−
1
Alder “click” chemistry between reactive precursors function-
21−23
alized with furan or maleimide groups,
oxime formation
chemistry of the monomers or linear polymers used to prepare
the hydrogels, the properties of the hydrogels can readily be
engineered to optimize their utility as drug storage and release
2
4
between a carbonyl group and a hydroxylamine, and Schiff
base formation between polymer precursors functionalized with
2
5,26
2
carbonyl and amine functional groups.
Of particular
depots, network scaffolds for tissue engineering applica-
3
−7
8−10
interest, hydrazone bond formation, resulting from the
condensation of carbonyl and hydrazide functional groups,
has been widely reported for use with injectable hydrogels due
to the rapid kinetics of this chemistry in aqueous environments
and the hydrolytic lability of the formed hydrazone group that
tions,
and biosensors.
Hydrogels cross-linked via
covalent bond formation are particularly attractive in that
they generally possess more elastic mechanical properties and
more tunable degradation kinetics than physically cross-linked
1
1
hydrogel formulations, making them more suitable within the
body for long-term use (especially in applications in which the
gels are designed as space-filling or must bear some load).
However, this enhanced elasticity makes most chemically cross-
linked hydrogels unsuitable for introduction within the body
through minimally invasive means such as via injection,
requiring invasive surgical implantation that renders them
27
facilitates both in situ formation and biodegradation. We have
previously reported the use of this chemistry to prepare
injectable environmentally responsive hydrogels based on
poly(N-isopropylacrylamide) (PNIPAM) that can be tailored
to suit a number of biomedical applications, including drug
12−14
delivery and tissue engineering.
12
An ideal hydrogel chemistry for most tissue engineering or
drug delivery applications would facilitate moderate to rapid
rates of gel formation (preventing or minimizing undesirable
diffusion of the precursor polymers away from the injection site
prior to gel formation), form hydrogels with degradable
nonideal for routine clinical use. In response, in recent years,
a great deal of attention has been drawn toward the
development of injectable hydrogels, including physically
cross-linked shear thinning hydrogels and chemically cross-
linked gels that can form in vivo following co-injection of
13
13−16
complementary reactive polymer precursors.
To maintain
the advantages of chemically cross-linked hydrogels in the
context of injectable formulations, a number of cross-linking
chemistries that occur rapidly in water under physiological
Received: November 1, 2013
Revised: January 2, 2014
Published: January 16, 2014
©
2014 American Chemical Society
781
dx.doi.org/10.1021/bm401615d | Biomacromolecules 2014, 15, 781−790

Biomacromolecules
Article
Scheme 1. Synthesis of a Ketal-Protected Ketone Monomer
linkages, and avoid reactions with biomolecules in the gel
environment. In the latter case, ensuring self-gelation instead of
cross-reaction of the gel precursors with biomolecules is critical
to produce hydrogels with sufficiently high cross-link densities
through the gel, and/or degrade the mechanical properties of
the gel. Finding ways to decouple these variables, in particular
gelation time with mechanics and degradation rate, would
greatly improve our ability to engineer specific hydrogels for
particular biomedical applications.
(
i.e., adequate mechanics) as well as prevent or suppress
potential reactions with proteins, which can lead to significant
Recent approaches to addressing this problem have involved
changing the chemical characteristics of the electrophile to alter
hydrogel properties. In particular, McKinnon et al. demon-
strated that using an aryl aldehyde instead of an aliphatic
aldehyde can be used to alter the properties of the hydrogel
28
inflammatory responses. However, most if not all existing
chemistries reported in the literature present limitations with
respect to at least one of these criteria. For example, aldehyde
groups employed in generating hydrazone cross-linked hydro-
gels can induce local toxicity if used at high concentrations or
with sterically hindered polymers which gel more slowly
32
formed. In this work, we seek to address these challenges by
using ketone groups (alone or in combination with aldehyde
groups) as the electrophilic reactive group used to form the
hydrazone cross-link. Ketones are less reactive to hydrazone
(
inducing effective rapid release of aldehyde polymers into the
body) due to their ability to cross-link proteins via their lysine
2
9,30
33
groups.
In addition, from a fundamental engineering
bond formation than aldehydes, attributable to the presence
perspective, current gelation chemistries are inherently limited
in that the cross-link density, gelation rate, mechanics, and
degradation time of the hydrogel are all intimately coupled;
hydrogels with higher cross-linking potential gel faster and form
of an extra electron-donating carbon at the α position of the
carbonyl group as well as steric hindrance to nucleophilic attack
of the carbonyl group. By adjusting the total number of ketone
groups as well as the aldehyde:ketone ratio in the electrophilic
precursor polymer, we aim to engineer the gelation time to
produce a range of hydrogels with tunable properties. To
demonstrate this phenomenon, we have chosen poly(N-
isopropylacrylamide) (PNIPAM) as the backbone polymer
for gel formation. PNIPAM has attracted significant interest in
the biomedical literature due to its ability to respond to the
temperature of its local environment; PNIPAM hydrogels
undergo a reversible volume phase transition at ∼32 °C that
results in significant and reversible dewelling within the
11
gels that are more elastic and degrade more slowly. Gelation
kinetics and the final morphology of the hydrogel are also often
linked given that rapid cross-linking reactions can induce
gelation faster than the time scale required for diffusional
mixing of precursor polymers. As a result, depending on the
type of mechanical mixing used during the gel formation
process, regions of local heterogeneity may form within the
polymer matrixes of these gels that scatter light (significantly
31
affecting the utility of these gels in ophthalmic applications,
for example), alter the diffusional properties of small molecules
34
hydrogel matrix. This environmental sensitivity allows for
7
82
dx.doi.org/10.1021/bm401615d | Biomacromolecules 2014, 15, 781−790

Biomacromolecules
Article
the design of gels that can both reduce their pore size as well as
switch from highly hydrophilic to partially hydrophobic on
demand to (for example) entrap preloaded drug and slow its
Intermediate A (10 g, 50 mmol) was then added to 180 mL of toluene
along with ethylene glycol (5.85 mL, 100 mmol) and dry para-
tolenesulfonic acid (934 mg, 5 mmol) and refluxed for 15 h. The
reaction mixture was cooled to room temperature, and the ethylene
glycol layer was extracted three times with diethyl ether. The toluene
and ether fractions were combined and washed three times with 5%
3
5
release or adhere and then release cells for regenerative
36
medicine applications. We have previously demonstrated the
fabrication of PNIPAM hydrogels using aldehyde−hydrazide
(w/v) NaOH followed by deionized water. The organic layer was
1
6
chemistry.
While gels with suitable mechanical and
dried over magnesium sulfate, and solvent was removed in a rotary
evaporator. The crude was recrystallized from ethanol to yield pure
intermediate B (Scheme 1B). Intermediate B was subsequently added
to 100 mL of deionized water along with 15 g of NaOH and refluxed
for 2 days, with an additional 60 g of NaOH added slowly over the
course of the reflux. Afterward, the reaction mixture was cooled to
room temperature and extracted three times with 50 mL of
dichloromethane. The organic layers were then combined and dried
over magnesium sulfate, filtered, and concentrated in a rotary
evaporator to yield pure product C (Scheme 1C), a slightly yellow
oil. Finally, the monomer was prepared by adding product C (21.1 mL,
thermoresponsive properties were produced, the gels formed
very quickly (∼10−30 s, problematic in some surgical contexts)
and were extremely opaque. Herein, by using ketone groups to
tune the gelation kinetics, we demonstrate the potential to
create transparent but still mechanically strong and highly
thermoresponsive hydrogels based on PNIPAM precursors.
EXPERIMENTAL SECTION
■
(
Materials. N-Isopropylacrylamide (NIPAM, 99%), acrylic acid
1
80 mmol) to a 20% (w/v) NaOH solution (in water) containing 4-
AA, 99%), adipic acid dihydrazide (ADH, 98%), N′-ethyl-N-(3-
hydroxy TEMPO (10 mg, 0.06 mmol). This reaction mixture was
dimethylaminopropyl)-carbodiimide (EDC, commercial grade), ethyl-
brought to 0 °C in an ice bath, and methacryloyl chloride (16.5 mL,
ene glycol (99.8%), thiolglycolic acid (MAA, ≥98.0%), toluene
1
74 mmol) was added dropwise over 2 h under nitrogen flow. The ice
(
99.8%), acetone (anhydrous, ≥99.9%), chloroacetone (95%),
bath was then allowed to warm to room temperature and the reaction
left to stir overnight in darkness. After this time, stirring was halted and
the product was allowed to collect at the top of the reaction flask. The
pure monomer product (along with inhibitor) (shown in Scheme 1D)
was then isolated using a separatory funnel. The monomer was stored
phthalimide potassium salt (≥98%), methacryloyl chloride
(
≥97.0%), aminoacetaldehyde dimethyl acetal (99%), 4-hydroxy-
TEMPO (97%), and para-toluenesulfonic acid (≥98.5%) were all
purchased from Sigma Aldrich (Oakville, ON). Dimethyl sulfoxide
(
DMSO, reagent grade) was purchased from Caledon Laboratory
1
in the darkness at −20 °C until use. H NMR (600 MHz) in DMSO-
Chemicals (Georgetown, ON). 3T3 Mus musculus mouse cells were
obtained from ATCC: Cederlane Laboratories (Burlington, ON). Cell
proliferation media (which includes Dulbecco’s modified Eagle’s
medium-high glucose (DMEM), fetal bovine serum (FBS), and
penicillin streptomycin (PS)), trypsin-EDTA, and recovery media
were all acquired from Invitrogen (Burlington, ON).
d : RNHCH C(OCH CH O)CH , 1.3 ppm, singlet, 3H;
6
2
2
2
3
CH CCH CONHR′, 2 ppm, singlet, 3H; RNHCH C(OCH CH O)-
2
3
2
2
2
CH , 3.5 ppm, doublet, 2H; RNHCH C(OCH CH O)CH , 4 ppm,
3
2
2
2
3
singlet, 4H; CH CCH CONHR′, 5.35−5.65 ppm, doublet, 2H;
2
3
CH CCH CONHR′, 6 ppm, singlet, 1H. See the Supporting
2
3
1
Information, Figure S1, for the H NMR spectrum.
Synthesis of an Acetal-Protected Aldehyde Monomer (N-
2,2-Dimethoxyethyl)methacrylamide). Aminoacetaldehyde di-
Synthesis of Hydrazide-Functionalized PNIPAM Copolymer.
(
The synthesis of this nucleophilic copolymer was performed as
methyl acetal (50 mL, 46 mmol) was added to a 120 mL stirred
solution of 20% NaOH (w/v). This solution was then cooled to 0 °C
in an ice bath. 4-Hydroxy TEMPO (10 mg, 0.06 mmol) was then
added to this solution as a stabilizer and stirred until it was fully
dissolved in solution. Methacryloyl chloride (47 mL, 48 mmol) was
next added dropwise over the course of 2 h under nitrogen flow, after
which the ice bath was allowed to warm to room temperature and the
reaction was left to stir overnight in darkness. The reaction mixture
was then extracted with 150 mL of petroleum ether. Subsequently, the
aqueous phase was saturated with sodium chloride and was extracted
three times with tert-butyl methyl ether. Additional 4-hydroxy
TEMPO was added to this phase to prevent premature polymerization
of the monomer. The organic phase was then dried over magnesium
sulfate, filtered, and concentrated in a rotary evaporator, yielding the
1
6
previously described (see also the Supporting Information, Scheme
S2). Briefly, N-isopropylacrylamide (4 g, 35 mmol), acrylic acid (1 mL,
14 mmol), and thioglycolic acid (87 μL, 1.25 mmol) were dissolved in
20 mL of absolute ethanol and heated to 56 °C. Following degassing
with nitrogen, dimethyl 2,2′-azobis(2-methylpropionate) (AIBME)
(56 mg, 2.4 μmol) was added, and the solution was allowed to stir
overnight under nitrogen. The solvent was then removed in a rotary
evaporator, and the crude product was redissolved in deionized water
and subjected to exhaustive dialysis followed by lyophilization.
Conductometric titration (ManTech, Inc.) indicated that the
copolymer contained 15.2 ± 0.4 mol % acrylic acid residues per
chain. The acrylic acid residues were then converted to hydrazide
groups by dissolving the polymer in deionized water along with a 10-
fold molar excess of ADH and adjusting the solution pH to 4.75 using
0.1 M HCl. A 2-fold molar excess of EDC (predissolved in 10 mL of
deionized water) was then added to the solution, and the reaction was
allowed to continue (maintaining a constant pH of 4.75 throughout
the reaction via addition of 0.1 M HCl) until no change in pH was
observed, typically on the order of ∼4 h. At this point, the solution was
neutralized using 0.1 M NaOH and subjected to exhaustive dialysis
and subsequent lyophilization. Conductometric titration indicated that
97 ± 3% of acrylic acid groups had been converted to hydrazide
groups, resulting in ∼14.7 mol % hydrazide groups per PNIPAM
copolymer chain. DMF gel permeation chromatography was
conducted using a Waters 590 HPLC pump along with three Waters
Ultrastyragel Linear columns operating in series and a Waters 410
refractive index detector. GPC indicated a number average molecular
weight of 26.5 kDa (polydispersity 1.55) for this hydrazide-
functionalized copolymer based on narrow molecular weight poly-
ethylene glycol standards (Waters).
acetal-protected aldehyde monomer. The monomer was stored in the
1
darkness at −20 °C until use. H NMR (600 MHz) in DMSO-d :
6
CH CCH COONHR′, 1.75 ppm, singlet, 3H; RNHCH CH(OCH ) ,
2
3
2
3 2
3
.25 ppm, triplet, 1H; RNHCH CH(OCH ) , 3.35 ppm, singlet, 6H;
2 3 2
R N H C H C H ( O C H ) , 4 . 4 1 p p m , d o u b l e t , 2 H ;
2
3
2
CH CCH COONHR′, 5.35−5.65 ppm, doublet, 2H; RNHR′, 8
2
3
ppm, singlet, 1H (see the Supporting Information, Scheme S1).
Synthesis of a Ketal-Protected Ketone Monomer (N-((2-
Methyl-1,3-dioxolan-2-yl)methyl)methacrylamide). Synthesis of
the protected ketone monomer was conducted on the basis of a
37
modification of a previously reported protocol. Chloroacetone (10
mL, 12.5 mmol) and the potassium salt of phthalimide (25.5 g, 13.8
mmol) were added to 150 mL of stirred dry acetone. The solution was
then heated to 80 °C for 20 h, after which it was cooled to room
temperature and the acetone was removed in a rotary evaporator. The
resulting solid was then redissolved in methylene chloride and washed
repeatedly with water. The methylene chloride layer was dried over
magnesium sulfate, filtered, and removed using a rotary evaporator.
The resulting yellow crude solid was washed with diethyl ether several
times until the solid became white; this solid was subsequently dried in
a vacuum oven to yield purified intermediate A (Scheme 1A).
Synthesis of Aldehyde- and/or Ketone-Functionalized
PNIPAM Copolymers. N-Isopropylacrylamide (4 g, 35 mmol) was
copolymerized with various ratios of protected aldehyde to protected
ketone monomer (Scheme 2), keeping the total mole percent of these
7
83
dx.doi.org/10.1021/bm401615d | Biomacromolecules 2014, 15, 781−790

Biomacromolecules
Article
Scheme 2. Synthesis of Protected Ketone and/or Aldehyde Monomers with NIPAM and Subsequent Deprotection to Reveal the
Electrophilic Ketone and/or Aldehyde Moieties
a
Table 1. Compositions and Properties of Electrophilic PNIPAM Copolymer Chains
gel identification
aldehyde content (Xna)
ketone content (Xnk)
total electrophile content (Xn)
total NIPAM content (Xnip
)
Mw (kDa)
1
7
5
2
0
0
00Ald/0Ket
5Ald/25Ket
0Ald/50Ket
5Ald/75Ket
Ald/100Ket
Ald/100Ket*
1
0
0.12 ± 0.03
0.11 ± 0.02
0.11 ± 0.04
0.13 ± 0.03
0.12 ± 0.05
0.42 ± 0.05
0.88 ± 0.03
0.89 ± 0.02
0.89 ± 0.04
0.87 ± 0.03
0.88 ± 0.05
0.58 ± 0.05
21.2 ± 3.2
20.4 ± 2.3
21.8 ± 3.3
22.5 ± 2.1
20.6 ± 1.9
19.5 ± 2.6
0.72 ± 0.05
0.45 ± 0.07
0.23 ± 0.03
0
0
0.28 ± 0.05
0.55 ± 0.07
0.77 ± 0.03
1
1
a
The gel identification codes represent the anticipated mole fractions of the aldehyde:ketone monomers based on the recipe, while the actual mole
1
fractions of aldehyde and ketone functional groups within the polymer chains (as determined by H NMR) are given as Xna and X , respectively.
nk
The mole fractions of monomer types, either electrophilic monomer or NIPAM, are given as X and X , respectively. GPC results are reported
n
nip
relative to a polyethylene glycol standard curve.
protected functional groups (and thus electrophilic reactive groups for
covalent gelation) constant. The recipes used to prepare these ketone-
and/or aldehyde-functionalized copolymers are given in Table 1. The
polymerizations were carried out as described for the hydrazide-
containing copolymer. Following removal of ethanol with a rotary
evaporator, the protecting acetal and ketal groups were removed by
redissolving the obtained crude polymer in 3 M HCl and allowing the
resulting solution to stir at room temperature for 24 h. This solution
was then exhaustively dialyzed and lyophilized to isolate the purified
copolymers. Aldehyde and ketone group content was quantified via
NMR by measuring the relative peak intensity at 9.5 ppm (RCOH,
of a needle. The time required to form a stable gel was assessed using a
standard vial inversion technique following co-extrusion of the reactive
2
1,38,39
polymer precursors into a 20 mL glass scintillation vial.
The
time required to form a gel (i.e., the time required such that the
polymer did not flow within 5 s following vial inversion) was noted for
each gel type. In order to assess the potential formation of
microdomains and/or phase separation, gels were formed in
polystyrene cuvettes (1 cm path length) and the transparency of the
resulting gels was measured at 550 nm using a DU 800 UV/visible
spectrophotometer (Beckman Coulter) following 24 h of incubation at
room temperature. The phosphate buffer maintained the pH at 7.4 in
all gelation experiments. A total of n = 4 replicates were performed per
sample in each assay, with the reported error bars representing the
standard deviation of the measurements.
aldehyde) and 2.1 ppm (RCOCH , ketone) relative to a
3
tetramethylsilane internal standard. GPC was performed as previously
described for the hydrazide-functionalized copolymer.
Hydrogel Formation, Gelation Time, and Transparency
Assays. Hydrazide (nucleophilic) and aldehyde/ketone (electro-
philic) functionalized polymer precursors were dissolved separately in
PBS at 25 °C at a concentration of 6% (m/v) for each precursor
polymer and added to separate barrels of a double barrel syringe
Hydrogel Cross-Linking Efficiency. Polymer precursors 100Ald/
0Ket, 50Ald/50Ket, 0Ald/100Ket, and 0Ald/100Ket* were dissolved
in D O at 6% (m/v) and were added to one barrel of different double
2
barrel syringes. A 6% (w/v) solution of hydrazide-functionalized
PNIPAM in D O was then added to the second barrel of each of these
2
(Scheme 3). Depressing the plunger of the syringe caused the polymer
syringes, and hydrogels were made by co-extruding each polymer pair
precursors to mix along a baffled channel and co-extrude out of the tip
into different NMR tubes. Spectral measurements were taken
7
84
dx.doi.org/10.1021/bm401615d | Biomacromolecules 2014, 15, 781−790

Biomacromolecules
Article
type was determined using parallel plate geometry with a plate
diameter of 7 mm and a plate spacing of 1 mm. An initial strain sweep
was first conducted between 1 and 100% at 1 Hz in order to identify
the linear viscoelastic range of these gels. A strain within this linear
viscoelastic range was then selected and maintained constant during a
frequency sweep from 1 to 100 rad/s to measure G′. Measurements
were performed on gels initially incubated at room temperature and at
Scheme 3. Co-Injection of Hydrazide and Aldehyde/Ketone
Complementary PNIPAM Precursors into a Silicone Mold
Using a Double Barrel Syringe to Form a Hydrazone Cross-
Linked Hydrogel
3
7 °C. A total of n = 4 replicates are performed on each sample, with
the reported error bars representing the standard deviation of the
measurements.
Cytocompatibility Assays. The cytocompatibility of the hydrogel
1
6
precursors (which are also the degradation products) was assessed
using an in vitro MTT assay with 3T3 Mus musculus mouse cells similar
1
4
to that described by Campbell et al. Briefly, 24-well polystyrene
plates were cultured with 10 000 3T3 cells within 1 mL of proliferation
media. After 24 h, the cells were exposed to various concentrations of
hydrogel precursor polymers (0.1, 0.4, 0.8, 1.2, 1.6, and 2 mg/mL for
the lower concentration sets and 2, 4, 8, 12, 16, and 20 mg/mL for the
higher concentration set, n = 4 per concentration) and incubated for
an additional 24 h. The media and polymer solutions were replaced
with 1 mL of a 0.4 mg/mL MTT solution in proliferation media for
the last 3 h of incubation. After this, the MTT solution was aspirated
off and the formazan produced by the cells was dissolved in 400 μL of
DMSO. The absorbance of the formazan solution was read using a
Biorad microplate reader (model 550) at 570 nm against a 630 nm
baseline (accounting for media absorbance) and compared to the
absorbances measured in cell-only wells in which no material (i.e., only
media) was added to calculate the relative cell viabilities according to
eq 2.
immediately following co-extrusion into the tubes for each gel type as
well as 48 h after co-extrusion (allowing cross-linking to occur).
Changes in peak intensity corresponding to aldehyde protons (COH,
∼
9.5 ppm) and terminal methyl groups of ketone moieties (CH3,
∼2.15 ppm), both of which decrease following hydrazone bond
cell viability (%)
formation, were tracked relative to spectra of polymer precursors alone
at a total concentration of 3% (matching the concentration of each
(
absorbance
− absorbancepolymer−exposed,630nm)
polymer−exposed,570nm
=
precursor in each final gel) in D O. Sixteen scans were taken for each
(absorbanceblank,570nm − absorbanceblank,630nm)
2
sample type at a relaxation delay of 4 s per scan. The ratio of peak
integrations relative to the isopropyl group intensity of NIPAM
moieties was measured for both precursor alone and co-extruded
polymer solutions and compared at different time points in the
experiment (initially and at 48 h). For 50Ald/50Ket gels, measure-
ments were taken for both aldehyde and ketone functional groups
independently of each other, allowing for the calculation of the total
residual electrophilic group content in the gel. Quantification of the
number of electrophilic groups (of a given type) consumed through
cross-linking was made according to eq 1
(
2)
Error bars for each polymer tested represent the standard deviation of
the cell viability percentages measured for the four replicates
performed. A Student’s t test (at 95% confidence) is performed to
assess the statistical significance of differences between cell viability
results.
RESULTS
■
Reactive Precursor Polymer Synthesis. As per Table 1,
an approximately stoichiometric incorporation of monomers at
their desired mole fractions within each polymer chain was
obtained for both the ketal- and acetal-protected monomers,
resulting in electrophilic polymer chains with statistically equal
functional group contents (and therefore cross-linking
potential) in all hydrogels studied. For an equal mass mixture
of the polymers (each at 6%), the hydrazide polymers contain
functional groups in approximately a 1.5-fold excess over
electrophilic groups on complementary precursor polymer
chains, advantageous biologically given that hydrazides are
functionally bio-orthogonal with biomolecule functional groups,
whereas aldehydes and, to a much lesser extent, ketones are
not. The ∼42 mol % 0Ald/100Ket* electrophilic polymer is the
exception to this relationship and is designed to offer enhanced
gel formation kinetics and cross-link density relative to the
other precursor polymers prepared. The molecular weight of
each electrophilic precursor polymer was maintained at ∼20
kDa (independent of functional group incorporation, Table 1),
while the nucleophilic chain had a molecular weight of 26.5
kDa. Considering that hydrazone bond hydrolysis regenerates
the precursor polymers (i.e., the precursor polymers are the
degradation products), each tested polymer has a molecular
weight below the lower molecular weight cutoff of the
glomerulus basement membrane of the renal excretory system
⎛
⎝
Ie,g/Ii,g
⎞
%
electrophilic groups consumed = ⎜1 −
⎟ × 100%
⎜
⎟
I /I
e,s i,s ⎠
(1)
where I , I , I , and I are the integrations corresponding to the
e,g i,g e,s
i,s
electrophilic (ketone or aldehyde) peaks of the gel after 48 h, the
isopropyl groups of the gel after 48 h, the electrophilic peaks of the
polymer precursor solution initially (t = 0), and the isopropyl groups
of the polymer precursor solution initially (t = 0), respectively.
Hydrogel Degradation. Hydrogel precursor solutions were co-
extruded into 0.44 mL silicone molds, as shown in Scheme 3. The
solution-containing molds were then covered with silicone slabs and
left for 48 h at room temperature in order to ensure equilibrium cross-
linking of the polymer networks. Hydrogels were then placed in cell
culture inserts (12-well format, Falcon) and were subsequently
submerged in PBS at pH 7.4. The rates of degradation for each gel
type were determined gravimetrically at predetermined time intervals
after drawing off excess (unbound) water from the surface of the gels
using a Kimwipe. A total of n = 4 replicates are performed on each
sample, with the reported error bars representing the standard
deviation of the measurements.
Hydrogel Mechanical Properties. The mechanical properties of
hydrogels formed by co-extruding hydrazide-functionalized PNIPAM
and PNIPAM functionalized with different ratios of ketones to
aldehydes were determined using oscillatory rheology with an ARES
rheometer (Texas Instruments). The storage modulus (G′) of each gel
7
85
dx.doi.org/10.1021/bm401615d | Biomacromolecules 2014, 15, 781−790

Biomacromolecules
Article
(
∼40 kDa) and thus has the potential to be cleared following
electrophile content (∼12 mol %), consistent with the higher
reactivity of aldehyde groups relative to ketone groups that
would shift the equilibrium of hydrazone bond formation/
breaking toward the product (cross-linked) side. However, the
cross-link density of a gel prepared with an aldehyde-
functionalized precursor polymer can be matched using a
ketone-functionalized polymer precursor functionalized with a
higher total mole fraction of ketone functionality (Table 1).
The percentage of available ketone groups cross-linked is
approximately equal between the 0Ald/100Ket (∼12 mol %
ketone) and 0Ald/100Ket* (∼42 mol % ketone) precursors,
consistent with an equilibrium driving hydrazone bond
formation; however, the 4-fold increase in the total electrophilic
functional group content of 0Ald/100Ket* relative to 0Ald/
degradation.
Hydrogel Formation Kinetics. Hydrogels prepared with
any combination of aldehyde/ketone functionalization could be
easily co-extruded from a double-barrel syringe, and all
combinations tested successfully gelled within the silicone
molds to form elastic hydrogels within the 24 h gelation period
used for screening. Results of the vial inversion assay, shown in
Table 2, indicate that an increase in ketone content of the
Table 2. Time Required to Form a Hydrogel for the Various
Gel Types Assayed According to the Vial Inversion Test
gel ID
gel formation time (min)
100Ald/0Ket
75Ald/25Ket
50Ald/50Ket
25Ald/75Ket
0Ald/100Ket
0Ald/100Ket*
0.09 ± 0.02
0.4 ± 0.1
24 ± 2
45 ± 6
65 ± 13
2.0 ± 0.5
1
00Ket gels results in formation of ∼4−5-fold more cross-links
in the hydrogel. While it is possible that the absolute extent of
cross-linking in (unbuffered) D O will differ from that in
2
(
buffered) H O, it is clear from this data that the cross-link
2
density of the hydrogel can be scaled directly by the number of
reactive functional groups in the precursor polymers.
Hydrogel Mechanics. The results of oscillatory rheometry
conducted on a range of hydrogels prepared with precursor
polymers containing varying ratios of aldehyde and ketone
groups are shown in Figure 1. A clear relationship exists
electrophilic polymer leads to an increase in the time required
to form a gel, with the gelation time increasing dramatically
when the ketone concentration is increased to >50 mol %.
This increased gelation time associated with ketone-rich
polymers can be attributed to the lower electrophilicity of the
ketone group relative to the aldehyde group, making ketones
less susceptible to nucleophilic attack by the hydrazide-
functionalized PNIPAM chains. When the number of ketone
groups in the precursor polymers is quadrupled to ∼42 mol %
(
0Ald/100Ket*), gelation occurs 30-fold faster than when a
12 mol % ketone-functionalized precursor polymer is used
0Ald/100Ket). Interestingly, the gelation time observed using
∼
(
the ∼42 mol % ketone-functionalized precursor is still ∼20-fold
slower than that observed when using an aldehyde-function-
alized polymer with ∼12 mol % functional groups to form gels.
Hydrogel Cross-Linking Efficiency. By comparing the
aldehyde and ketone group NMR spectral intensities
immediately following co-extrusion and 48 h postextrusion,
an estimate of the cross-link density of the respective hydrogels
can be made. Changes in peak intensities were measured
Figure 1. Elastic moduli versus frequency for PNIPAM hydrogels
prepared with precursor polymers containing different aldehyde:ke-
tone functional group ratios and different ketone functional group
mole fractions (0Ald/100Ket* = ∼42 mol % ketones; all other
hydrogels ∼12 mol % electrophile).
relative to 3% (w/v) solutions of precursor only in D O (see
2
representative raw data spectra in the Supporting Information,
Figures S2 and S3). Table 3 shows the percentage of ketone
and/or aldehyde functional groups consumed in each gel as a
function of precursor chemistry and the total cross-link density
of each hydrogel following 48 h of gel formation.
Hydrogels prepared with aldehyde-rich precursor polymers
contain significantly more cross-links relative to those prepared
with ketone-rich precursor polymers at the same total
between the elastic modulus (G′) and the ratio of aldehyde to
ketone groups of the electrophilic precursor polymers used to
synthesize the hydrogels, with an increasing aldehyde to ketone
ratio yielding gels with an increased G′. This suggests that
aldehydes are significantly more effective at forming cross-links
than ketone groups following co-extrusion, consistent with the
NMR result (Table 3) and the lower reactivity of ketone
functional groups. Interestingly, the 0Ald/100Ket* hydrogels
(
prepared with ∼42 mol % ketone precursor polymers as
Table 3. Electrophilic Functional Group Consumption and
Cross-Link Density for Hydrogels Prepared with Different
Aldehyde:Ketone Ratios and Total Ketone Contents as
opposed to ∼12 mol % total electrophilic comonomer
precursor polymers used to prepare the other hydrogels)
exhibited analogous mechanical properties to the 100Ald/0Ket
hydrogel, with the additional available ketone functional groups
facilitating an increase in cross-link formation; again, this result
matches that of the NMR experiment (Table 3). Thus, while
the use of ketone-functionalized polymers results in a
significant decrease in the elastic modulus of the resulting
hydrogels at equal functional group contents, modulus
matching can still be performed independent of the
aldehyde:ketone ratio of the reactive copolymers by changing
1
Determined by H NMR
%
electrophilic
no. of electrophilic
functional groups
cross-link
density
functional groups
consumed
1
8
(×10 cm−3)
18
hydrogel ID
consumed (×10 )
1
00Ald/0Ket
0Ald/50Ket
Ald/100Ket
Ald/100Ket*
74 ± 6
42 ± 5
14 ± 3
17 ± 5
22.4 ± 1.8
12.3 ± 1.4
3.9 ± 0.5
19.3 ± 1.4
11.2 ± 0.9
6.1 ± 0.7
1.9 ± 0.2
9.6 ± 0.7
5
0
0
7
86
dx.doi.org/10.1021/bm401615d | Biomacromolecules 2014, 15, 781−790

Biomacromolecules
Article
the total ketone content in the electrophilic hydrogel precursor
polymer.
Hydrogel Transparency. UV/vis absorbance measure-
ments performed at 550 nm were used to assess the
transparency of the hydrogels prepared with the different
precursor polymers. A lack of transparency at this wavelength
indicates the presence of microdomains of a size similar to the
wavelength of the incident light. Figure 2 indicates that
Figure 3. Temperature-dependent thermal collapse following gel
incubation at 37 °C for PNIPAM hydrogels prepared with precursor
polymers containing different aldehyde:ketone functional group ratios
and different ketone functional group mole fractions (0Ald/100Ket*
∼
42 mol % ketones; all other hydrogels ∼12 mol % electrophile). The
dashed line represents the initial mass of each hydrogel tested.
restricts gel deswelling. Interestingly, the highly functionalized
0
Ald/100Ket* gels retain their ability to thermally deswell at 37
°
C to roughly the same degree as hydrogels prepared with ∼12
mol % electrophilic group precursors despite the ∼42 mol %
ketone monomer functionalization of this precursor polymer
(
i.e., only ∼58 mol % of the total monomer units are NIPAM).
The slightly lower phase transition temperature of ketone-
functionalized polymers relative to aldehyde polymers likely
offsets the lower NIPAM content in this case. As such, based on
the basis of Figure 3, the thermal phase transition can be
maintained independent of the cross-link density and opacity of
the hydrogel by varying both the ratio of aldehyde:ketone
groups as well as the mole percentage of ketone groups in the
electrophilic precursor polymer.
Figure 2. Transmittance of 550 nm light through hydrogels prepared
with precursor polymers containing aldehyde:ketone functional group
ratios and different ketone functional group mole fractions (0Ald/
100Ket* ∼ 42 mol % ketones; all other hydrogels = 12 mol %
electrophile). Pictures of each hydrogel are provided above each
transparency result for reference.
Hydrogel Degradation. Hydrogels prepared by hydra-
zide−ketone/aldehyde chemistry can degrade via the hydrolysis
of the hydrazone cross-linking moieties, as previously
increasing the aldehyde content of the precursor polymers
results in significantly less transparent hydrogels, indicating the
presence of microdomains and regions of heterogeneity within
the bulk gel. Conversely, gels prepared with precursors
containing a high fraction of ketone groups remain transparent,
with the higher transparency largely preserved as the number of
ketone functional groups is increased (comparing 0Ald/100Ket
with ∼12 mol % ketones and 0Ald/100Ket* with ∼42 mol %
ketones); this result suggests that the ketone group itself plays a
role in maintaining the gel transparency even independent of
the gelation time, since the 0Ald/100Ket* gel forms 10-fold
faster than the 50Ald/50Ket gel but maintains significantly
higher transparency.
14−16
demonstrated.
Figure 4 indicates that the rate of hydrogel
degradation (as measured in PBS at 37 °C) can be tuned on the
basis of the aldehyde to ketone ratio used to prepare the
electrophilic precursor polymers. 0Ald/100Ket gels (containing
Hydrogel Phase Transition. Incubation of the hydrogels
in PBS at 37 °C induced deswelling in all gels, consistent with
the phase transition behavior of PNIPAM (Figure 3). Of note,
while the precursor polymers themselves are significantly more
hydrophilic than PNIPAM (with LCST values of 78 °C for
PNIPAM−hydrazide, 45 °C for PNIPAM−aldehyde, and 37 °C
for PNIPAM−ketone, with both electrophilic chains possessing
∼
12 mol % functionalization), the phase transition of the
hydrogel occurs at a temperature analogous to that of PNIPAM
prepared via standard free radical processes. Similar degrees and
kinetics of deswelling are observed across each gel type;
however, an increase in ketone content leads to a somewhat
higher degree of thermal collapse of the polymer network. This
discrepancy between different gel types can be attributed
primarily to differences in cross-link density, by which the
higher degree of cross-linking in aldehyde-rich gels elastically
Figure 4. Hydrogel degradation profiles (expressed in terms of the gel
mass at different time points normalized to the initial mass of the gel
following co-extrusion) following gel incubation at 37 °C for PNIPAM
hydrogels prepared with precursor polymers containing different
aldehyde:ketone functional group ratios and different ketone func-
tional group mole fractions (0Ald/100Ket* ∼42 mol % ketones; all
other hydrogels ∼12 mol % electrophile). The dashed line represents
the initial mass of each hydrogel tested.
7
87
dx.doi.org/10.1021/bm401615d | Biomacromolecules 2014, 15, 781−790

Biomacromolecules
12 mol % ketone groups and no aldehydes) degrade rapidly
Article
∼
(∼18 days) after gel formation; in contrast, 100Ald/0Ket gels
containing ∼12 mol % aldehyde groups and no ketones)
(
degrade over a much longer time period (∼93 days). This
difference between these two hydrogel types is likely
attributable to the large observed variation in cross-linking
between the two gels (Table 3), with the aldehyde-function-
alized precursor polymers producing hydrogels with signifi-
cantly higher cross-link densities. Of note, the 0Ald/100Ket*
hydrogel degrades more slowly than the 100Ald/0Ket hydrogel
with the same cross-link density (as per G′ measurements,
Figure 2, and NMR, Table 3), as the extra α carbon makes the
hydrazone group less reactive to water hydrolysis. As such, by
varying both the aldehyde:ketone ratio and the total number of
ketone functional groups, the gel degradation rate can be tuned
independently of the cross-link density, typically not possible
with other reported injectable hydrogel formulations. Interest-
ingly, while the 0Ald/100Ket gels undergo a nearly linear
degradation profile following the onset of degradation, 100Ald/
Figure 5. Comparison of cell viability (relative to a cell-only control)
of 3T3 mouse fibroblast cells in the presence of high concentrations
(
0
>2 mg/mL) of 0Ald/100Ket* (∼42 mol % ketones) and 100Ald/
Ket (∼12 mol % aldehydes). The asterisk indicates that the difference
in the viabilities of the two polymers at a given concentration is
statistically significant (p < 0.05).
0
Ket gels exhibit an additional secondary degradation profile,
with a very slow degradation rate initiated at ∼78 days
following >90% degradation) until complete gel degradation is
(∼42 mol % ketones) precursor polymers at concentrations as
high as 20 mg/mL (2%), a concentration typically far in excess
of those screened in an in vitro cell culture assay. At these
higher concentrations, and particularly in the range 4−12 mg/
mL, the PNIPAM−ketone polymer exhibited significantly less
cytotoxicity than the PNIPAM−aldehyde polymer (p < 0.05).
The PNIPAM−aldehyde polymer started to exhibit significant
cytotoxicity (<80% relative viability) at concentrations above 4
mg/mL, whereas the PNIPAM−ketone polymer only exhibited
equivalent cytotoxicity at a concentration of 20 mg/mL despite
containing quadruple the mole fraction of electrophilic reactive
functional groups. This result suggests that the use of ketones
in partial or complete place of aldehydes significantly reduces
the potential cytotoxicity of the precursor polymer as well as
the degradation products.
(
achieved. These hydrogels remained as small spherical opaque
samples during this secondary degradation step. This
phenomenon may be attributable to the presence of
inhomogeneities in these highly opaque hydrogels (see Figure
2), in which regions of higher and lower cross-link densities
may be present; secondary hydrolysis may be attributable to the
slower degradation of the more highly cross-linked pockets in
the hydrogel. A similar secondary degradation rate was
observed for gels possessing intermediate aldehyde to ketone
ratios, with the effect decreasing as the ketone content is
increased (and the transparency of the gel improves, see Figure
2
). Furthermore, the 0Ald/100Ket* gel that has a matched
cross-link density to 100Ald/0Ket but significantly higher
transparency exhibits no secondary linear region in its
degradation profile, again supporting the hypothesis that an
increase in ketone content in the precursor polymers leads to a
greater degree of gel homogeneity and therefore a more
uniform gel degradation profile.
DISCUSSION
■
By changing the ratio of ketone to aldehyde groups in the
electrophilic polymer involved in the formation of hydrazone
cross-linked injectable hydrogels, significant control can be
exerted over several key engineering properties of these
hydrogels. The slower reactivity of ketones relative to aldehydes
significantly slows the rate of gelation, increasing the time
required for gel formation from several seconds in the case of
aldehyde-functionalized polymers to over 1 h in the case of only
ketone-functionalized polymers (Table 2). We anticipate that
copolymers containing between 25 and 50 mol % ketones
(yielding gelation times on the order of several minutes) may
be ideal for many biomedical applications in this respect,
allowing surgeons sufficient time to administer the gel at the
site(s) desired while avoiding premature gelation inside the
syringe. Ketone-functionalized precursor polymers also resulted
in significantly more transparent hydrogels (Figure 2), a
phenomenon also attributable to the slower gelation rate with
these copolymers. Slower gelation allows more time for
diffusional mixing of the (miscible) precursor polymers before
the matrix is fixed by cross-link formation, decreasing the
number of microdomain inhomogeneities within the matrix and
thus leading to more transparent gels. The uniformity of
hydrogels has two significant advantages in biomedical
applications. First, transparency is required in some applications
of hydrogels, particularly ophthalmic applications (e.g., eyedrop
formulations, intraocular lenses, or vitreal substitutes) that are
Polymer Cytocompatibility. The in vitro cytotoxicity of
the hydrogel precursor polymers, which are also the
16
degradation products of the hydrogels, was screened using
an MTT assay with 3T3 mouse fibroblasts. Numerous previous
studies from our group have shown that the PNIPAM−
hydrazide component exhibits no significant cytotoxicity at the
typical concentrations used in cell culture assays as well as in
14−16
vivo in subcutaneous mouse models.
Therefore, the in vitro
study herein was focused on a comparison of the cytotoxic
effects of the PNIPAM−aldehyde and PNIPAM−ketone
polymers, along with polymers containing a combination of
aldehyde and ketone components. MTT assay results for
polymers prepared with different aldehyde:ketone ratios
indicated cell viabilities of at least 80% for all polymer ratios
tested at concentrations up to 2 mg/mL (see the Supporting
Information, Figure S4), viabilities that have been associated
1
with little to no practical cytotoxicity. Notably, the maximum
concentration tested (2 mg/mL) is expected to be greater than
the concentration of free polymer that the body would be
exposed to given the relatively rapid gelation process of the
hydrogels along with their slow degradation.
Figure 5 shows the results of an MTT assay performed on
the 100Ald/0Ket (∼12 mol % aldehydes) and 0Ald/100Ket*
7
88
dx.doi.org/10.1021/bm401615d | Biomacromolecules 2014, 15, 781−790

Biomacromolecules
Article
particularly attractive for injectable hydrogel formulations.
Second, gel uniformity reduces confounding intramatrix
morphologies that may serve to complicate drug diffusion
throughout the gel and simplifies the development of
mathematical expressions governing drug release, improving
the predictability of hydrogel performance in a drug delivery
application. Similarly, if the inhomogeneities are on the same
length scale as cells, local differences in cell responses to the
hydrogel may result, leading to typically undesirable nonuni-
form cell distributions or cell behaviors within a tissue
engineering matrix.
Hydrogel properties can be tuned by varying both the ratio
of aldehyde:ketone groups as well as the total number of either
functional group, with varying the amount of ketone groups of
particular interest. Hydrogels prepared with ∼42 mol % ketone-
functionalized precursor polymers (0Ald/100Ket*) maintain
relatively high transparency (Figure 2) even though they gel
much more quickly than even some of the more opaque
hydrogels prepared using aldehyde−ketone mix precursor
polymers (Table 2) and exhibit both analogous mechanical
CONCLUSIONS
■
Injectable hydrogels prepared by mixing hydrazide-function-
alized PNIPAM with electrophilic PNIPAM precursor poly-
mers containing mixtures of aldehyde and ketone groups can be
tuned to exhibit desired gelation rates, degradation rates,
opacities, and mechanical properties. Specifically, by changing
the aldehyde:ketone ratio in the electrophilic polymer used for
gelation and/or the mole percentage of ketones incorporated in
the polymer, largely independent control over the gelation rate,
opacity, degradation rate, and cross-link density can be
achieved, unlike in currently reported in situ gelling chemistries.
The hydrogels maintain their volume phase transition proper-
ties irrespective of the composition of the electrophile
precursor polymer; furthermore, the presence of higher ketone
fractions significantly improves the cytocompatibility of the
precursor polymers (and, by extension, the degradation
products) at even extremely high polymer concentrations,
facilitating the use of such hydrogels in switchable biomedical
applications. This chemistry approach could be applied to any
synthetic or natural polymer backbone to produce highly
tunable hydrogel compositions by simple mixing of well-
defined precursor polymers applicable to a variety of different
biomedical challenges.
(
Figure 1) and thermoresponsive properties (Figure 3) as well
as slower degradation rates (Figure 4) relative to hydrogels
prepared with ∼12 mol % aldehyde-functionalized precursor
polymers. In this way, manipulating the total number of fast
(
aldehyde) and slow (ketone) gelling functional groups can
ASSOCIATED CONTENT
■
result in largely independent control over gelation time, gel
degradation kinetics, gel transparency, hydrogel mechanics, and
phase transition behavior to a degree not currently possible
with existing injectable hydrogel approaches. Specifically, for all
currently reported injectable hydrogel chemistries, gelation
times and hydrogel degradation rates are both directly linked to
the ultimate cross-link density of the hydrogel; the use of
mixtures of ketones and aldehydes for hydrogel preparation
decouples these parameters to provide largely independent
control over multiple key hydrogel properties. This degree of
control significantly improves the potential translatability of
these materials in a surgical context, since hydrogels with
desired mechanical properties (i.e., for matching elastic moduli
of native tissues) can be delivered with a gelation time chosen
by the surgeon administering the hydrogel that is most suitable
to the procedure to be conducted.
Maximizing the ketone content of the precursor polymers
also has demonstrated advantages in terms of maintaining high
cell tolerance to the precursor polymers. Aldehyde groups can
react with amine-containing cellular components via Schiff base
formation, potentially disrupting regular cellular processes
necessary for cell survival and proliferation (particularly at
higher polymer concentrations); in comparison, ketones are
inherently less reactive due to the presence of two adjacent
alpha carbons instead of one, which reduces the electrophilicity
of the carbonyl group and renders ketones significantly more
*
S
Supporting Information
Schemes related to the chemistry of aldehyde monomer
preparation and hydrazide polymer preparation, raw NMR
data for the estimation of cross-link density in the hydrogels,
and MTT assay results for electrophilic precursor polymers
with different aldehyde:ketone ratios. This material is available
free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
*
Corresponding Author
E-mail: hoaretr@mcmaster.ca.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Funding from the 20/20: NSERC Ophthalmic Materials
Research Network, the Natural Sciences and Engineering
Research Council of Canada (Discovery Grants Program), and
the Ontario Graduate Scholarship Program (MP) is gratefully
acknowledged.
REFERENCES
■
(
1) Geckil, H.; Xu, F.; Zhang, X.; Moon, S.; Demirci, U.
Nanomedicine 2010, 5, 469−484.
(
(
2) Hoare, T. R.; Kohane, D. S. Polymer 2008, 49, 1993−2007.
3) Lee, K. Y.; Mooney, D. J. Chem. Rev. 2001, 101, 1869−1880.
40
bio-orthogonal to typical functional groups found in vivo.
Thus, while care must be taken not to slow the gelation time to
the point where the precursor polymers diffuse away prior to
gelling once injected in vivo, the ability to precisely tune the
gelation time using ketone−aldehyde copolymers of defined
compositions offers control over multiple hydrogel properties
of interest. Note that all hydrogels studied in this paper are
prepared with the same total mass concentration of precursor
polymers (6%), a parameter that if varied would provide an
additional mechanism to control the properties of the final
hydrogels.
(4) Galperin, A.; Long, T. J.; Ratner, B. D. Biomacromolecules 2010,
2583−2592.
(5) Van Vlierberghe, S.; Dubruel, P.; Schacht, E. Biomacromolecules
2
(
(
011, 12, 1387−1408.
6) Tan, H.; Marra, K. G. Materials 2010, 3, 1746−1767.
7) Nicodemus, G. D.; Bryant, S. J. Tissue Eng., Part B 2008, 14, 149−
1
65.
(
8) Yang, W.; Xue, H.; Carr, L. R.; Wang, J.; Jiang, S. Biosens.
Bioelectron. 2011, 26, 2454−2459.
(9) Toma, M.; Jonas, U.; Mateescu, A.; Knoll, W.; Dostalek, J. J. Phys.
Chem. C 2013, 117, 11705−11712.
7
89
dx.doi.org/10.1021/bm401615d | Biomacromolecules 2014, 15, 781−790

Biomacromolecules
Article
(
10) Khimji, I.; Kelly, E. Y.; Helwa, Y.; Hoang, M.; Liu, J. Methods
013, 64, 292−298.
11) Hiemstra, C.; van der Aa, L. J.; Zhong, Z.; Dijkstra, P. J.; Feijen,
J. Biomacromolecules 2007, 8, 1548−1556.
12) Azab, A. K.; Doviner, V.; Orkin, B.; Kleinstern, J.; Srebnik, M.;
Nissan, A.; Rubinstein, A. J. Biomed. Res., Part A 2007, 2, 414−422.
13) Guvendiren, M.; Lu, H. D.; Burdick, J. A. Soft Matter 2012, 8,
60.
14) Campbell, S. B.; Patenaude, M.; Hoare, T. Biomacromolecules
2
(
(
(
2
(
2
(
(
(
013, 14, 644−653.
15) Patenaude, M.; Hoare, T. Biomacromolecules 2012, 13, 369−378.
16) Patenaude, M.; Hoare, T. ACS Macro Lett. 2012, 1, 409−413.
17) Pritchard, C. D.; O’Shea, T. M.; Siegwart, D. J.; Calo, E.;
Anderson, D. G.; Reynolds, F. M.; Thomas, J.; Slotkin, J. R.; Woodard,
E. J.; Langer, R. Biomaterials 2011, 32, 587−597.
(
18) Yu, Y.; Deng, C.; Meng, F.; Shi, Q.; Feijen, J.; Zhong, Z. J.
Biomed. Mater. Res., Part A 2011, 99, 316−326.
19) Sekine, Y.; Moritani, Y.; Ikeda-Fukazawa, T.; Sasaki, Y.;
Akiyoshi, K. Adv. Healthcare Mater. 2012, 1, 722−728.
20) Censi, R.; Fieten, P. J.; Di Martino, P.; Hennink, W. E.;
Vermonden, T. J. Controlled Release 2010, 148, 28−29.
21) Wei, H. L.; Yang, Z.; Zheng, L. M.; Shen, Y. M. Polymer 2009,
0, 2836−2840.
22) Nimmo, C. M.; Owen, S. C.; Shoichet, M. S. Biomacromolecules
011, 12, 824−830.
23) Tan, H.; Rubin, J. P.; Marra, K. G. Macromol. Rapid Commun.
011, 32, 905−911.
24) Grover, G. N.; Lam, J.; Nguyen, T. H.; Segura, T.; Maynard, H.
D. Biomacromolecules 2012, 13, 3013−3017.
25) Vetrík, M.; Pra ny, M.; Hruby, M.; Michal
Stab. 2011, 96, 756−759.
26) Sivakumaran, D.; Maitland, D.; Hoare, T. Biomacromolecules
011, 12, 4112−4120.
27) Ito, T.; Yeo, Y.; Highley, C. B.; Bellas, E.; Kohane, D. S.
Biomaterials 2007, 28, 3418−3426.
28) Beauchamp, R. O.; Clair, M. B., St; Fennell, T. R.; Clarke, D. O.;
Morgan, K. T.; Kari, F. W. Crit. Rev. Toxicol. 1992, 22, 143−174.
29) Gough, J. E.; Scotchford, C.; Downes, S. J. Biomed. Mater. Res.
002, 61, 121−130.
30) O’Brien, P. J.; Siraki, A. G.; Shangari, N. Crit. Rev. Toxicol. 2005,
5, 609−662.
31) Meid, J.; Friedrich, T.; Tieke, B.; Lindner, P.; Richtering, W.
Phys. Chem. Chem. Phys. 2011, 13, 3039−3047.
32) McKinnon, D. D.; Domaille, D. W.; Cha, J. N.; Anseth, K. S.
Adv. Mater. 2013, DOI: 10.1002/adma.201303680.
(
(
(
5
(
2
(
2
(
(
̌
d
́
́
́
́
ek, J. Polym. Degrad.
(
2
(
(
(
2
(
3
(
(
(33) Uzu, S.; Kanda, S.; Imai, K. Analyst 1990, 115, 1477−1482.
(34) Cai, S.; Suo, Z. J. Mech. Phys. Solids 2011, 59, 2259−2278.
(35) Kang Derwent, J. J.; Mieler, W. F. Trans. Am. Ophthalmol. Soc.
2
(
008, 106, 206−213.
36) Hou, Y.; Matthews, A. R.; Smitherman, A. M.; Bulick, A. S.;
Hahn, M. S.; Hou, H.; Han, A.; Grunlan, M. Biomaterials 2008, 29,
3
(
(
175−3184.
37) Adachi, J.; Sato, N. J. Org. Chem. 1972, 37, 221−225.
38) Motokawa, R.; Morishita, K.; Koizumi, S.; Nakahira, T.; Annaka,
M. Macromolecules 2005, 38, 5748−5760.
39) Dong, Y.; Hassan, W.; Zheng, Y.; Saeed, A. O.; Cao, H.; Tai, H.;
Pandit, A.; Wang, W. J. Mater. Sci.: Mater. Med. 2012, 23, 25−35.
40) Sletten, E. M.; Bertozzi, C. R. Angew. Chem., Int. Ed. 2009, 38,
974−6998.
(
(
6
7
90
dx.doi.org/10.1021/bm401615d | Biomacromolecules 2014, 15, 781−790