(302c) Injectable, Degradable Thermoresponsive Poly(N-isopropylacrylamide) Hydrogels

Injectable,
Degradable Thermoresponsive poly(N-isopropylacrylamide) Hydrogels

Mathew
Patenaude and Todd Hoare

Department
of Chemical Engineering, McMaster University, Hamilton, Ontario, L8S 4L7

Introduction

            Thermoresponsive hydrogels based on poly(N-isopropylacrylamide)
(PNIPAM) that switch from a water-swollen state at low temperature to a
collapsed state at high temperature have been the subject of extensive
investigation. The volume phase transition temperature of PNIPAM hydrogels, lying
just below normal physiological temperature, has sparked particular interest in
the biomedical applications of these hydrogels as ?smart?, environmentally
sensitive tissue engineering scaffolds, cell growth supports, and drug delivery
vehicles, among other applications. Despite their potential in these biomedical
applications, PNIPAM hydrogels have not achieved clinical acceptance, primarily
due to concerns regarding the ultimate fate of PNIPAM in the body. Though monomeric
NIPAM is toxic, PNIPAM has been shown to be effectively noncytotoxic at
concentrations realistic to many medical applications.¹ However,
possible depolymerization and/or chronic bioaccumulation of PNIPAM represent significant
regulatory barriers to medical use. PNIPAM hydrogels are most often formed
outside of the body since polymerization most often makes use of free radical
chemistry and thermal or UV radiation, which would induce significant cell
toxicity and therefore cannot be performed in deep tissues. While weakly
cross-linked hydrogels, such as those formed through physical interactions, may
be sufficiently viscous to facilitate injection, injection of more highly
elastic hydrogels is impractical. As a result, there is a need for mechanically
robust PNIPAM-based hydrogels that can be introduced into the body through
minimally invasive means and are designed to subsequently degrade into safe
products that can be cleared following use.

We have developed PNIPAM based
hydrogels capable of forming in situ following coinjection of precursor
copolymers into the body via a double-barreled syringe. The chemical
cross-links used to form these hydrogel networks are based on rapidly forming,
yet hydrolytically labile, hydrazone bonds that form between aldehyde and
hydrazide functionalized PNIPAM copolymer precursors. By basing these hydrogels
on hydrazone chemistry, we have developed rapidly forming hydrogels capable of
biodegradation within the body. Further to this, by using PNIPAM precursors of
molecular weights low enough for kidney clearance, these gels may be eliminated
following their biodegradation in vivo.

Materials and Methods

            Hydrogels were fabricated from
hydrazide-functionalized PNIPAM and aldehyde-functionalized PNIPAM. The
hydrazide-functionalized precursor was generated by copolymerizing NIPAM with
acrylic acid and subsequently grafting the resulting copolymer with adipic acid
dihydrazide using aqueous carbodiimide chemistry. Aldehyde-functionalized
PNIPAM was synthesized by copolymerizing NIPAM with 2,3-dihydroxypropyl
methacrylate and subsequently oxidatively cleaving the resulting copolymer
using sodium periodate. The molecular weight of precursor PNIPAM copolymers was
determined using aqueous gel permeation chromatography. The degree of hydrazide
functionality of PNIPAM copolymers was determined through conductometric
titration. Aldehyde functionality was quantified with a carbazate-based
aldehyde detection assay. Hydrazide and aldehyde-functionalized polymers were
mixed via co-injection through a needle. These polymers rapidly form a hydrazone-crosslinked
hydrogel network upon mixing. Reversible thermoresponsiveness was determined
gravimetrically by incubating the gels at alternating temperatures of 25 °C and
37 °C. The degradability of these gels was demonstrated by incubating them in
different concentrations of hydrochloric acid and subsequently determining the
percent of gel loss over a period of time gravimetrically. Cytotoxicity in
response to gel precursors (degradation products) was assayed using an MTT assay.
Finally, both acute and chronic histopathology was determined in BALB/c mice
following sub-cutaneous injection of polymeric precursors via dual-barreled
syringe.

Results

            The
reversible thermal response of the resulting PNIPAM hydrogel is demonstrated in
Figure 1. The initial collapse of the hydrogel network occurs more rapidly than
subsequent cycles, while reswelling of the network takes place over a number of
days before the equilibrium swelling condition is again reached. Although the
hydrogels do not fully recover to the zero strain state following their first
incubation at 37 °C, they do exhibit reversible swelling-deswelling
transitions upon subsequent cycles; this is consistent with previous reports of
conventional, highly-crosslinked PNIPAM hydrogels.² The swelling
response is mirrored by changes in the turbidity of the hydrogel, which
switches from nearly transparent immediately following gel formation to completely
opaque upon thermal collapse to semi-translucent upon rehydration (Figure 1,
side panel). Thus, these injectable hydrogels exhibit the same thermal
deswelling and scattering properties of conventional, nondegradable PNIPAM
hydrogels.

Figure 1. Reversible collapse/swelling of
poly(NIPAM) hydrogel network incubated in 10 mM PBS (pH 7.4) at alternating
temperatures of 25 and37 °C (as noted at top
of graph). Inset: Gel opacity resulting from a transition from the zero-strain
state to 37 °C and reswelling at
25 °C.

Figure 2
demonstrates the degradability of the synthesized PNIPAM hydrogels in response
to acid-catalyzed hydrolysis at constant [H+] concentrations. The increase in
degradation rate with increasing acid concentration demonstrates that bulk loss
of hydrogel occurs by proton-catalyzed hydrolysis of the hydrazone crosslinks. 
GPC further confirmed the regeneration of the initial reactive copolymers upon
degradation (i.e. the carbon-carbon polymer backbone is not
hydrolytically susceptible). By degrading the hydrogels at defined chemical
crosslinking points, we expect that the toxic effects observed with monomeric
byproduct may be avoided without compromising the potential clearance of the
polymer degradation products through the kidney.

Figure 2. Degradation of PNIPAM hydrogels on
incubation at 37 °C in 0.1 M, 0.5 M,
and 1 M HCl at constant ionic strength.

In vitro and in
vivo toxicity studies of the injectable PNIPAM hydrogels and the hydrazide and
aldehyde-reactive precursor copolymers are shown in Figure 3. In vitro
cell viability assays were performed using an MTT assay with 3T3 mouse
fibroblasts and retinal pigment epithelial (RPE) cells. PNIPAM-co-ADH induces a
slight cytotoxic response at concentrations greater than 400 μg/mL in 3T3
cells (Figure 3A) and no significant toxicity at any tested concentration in
RPE cells (Figure 3B). Poly(NIPAM-cooxoethyl methacrylate) induces only mild
cytotoxicity to both cell types at low-to-moderate  concentrations only becoming
cytotoxic at very high concentrations (i.e., 2000 μg/mL, Figures 3A,B). Overall,
any observed cytotoxicity occurs at polymer concentrations well above any local
concentrations expected due to the slow hydrolytic degradation of hydrogel
under physiological conditions.

        Acute (48 h post-injection) in vivo toxicity assays were performed via
subcutaneous injection of BALB/c mice with 6 wt % solutions of PNIPAM polymer
precursors (Figures 3-C and D), and a 6 wt % in situ-formed hydrogel  (Figure
3-E). Hematoxylin-eosin (H&E) staining indicates that the reactive
copolymer solutions are largely cleared from the injection site 48 h post-injection
and induced only a very mild acute inflammatory response. The in situ-gelled
PNIPAM hydrogel remains at the injection site after 48 h and induces only a
mild inflammatory response at the tissue-hydrogel interface consisting
predominantly of macrophages with a few neutrophils. Following a five-month chronic
incubation of the gel in vivo (Figure 3F) the gel remains at the site of
introduction, although the quantity of residual gel was somewhat lower than the
initial injected volume. However, minimal to no fibrous capsule formation was
observed at the gel-tissue interface and no sign of a chronic
inflammatory response was noted either within the residual gel or within the
tissue adjacent to the gel.  Together, these results suggest that the gel is
well-tolerated within the subcutaneous space and may be amenable for practical
use in in vivo applications.

Figure 3. Cytotoxicity and biocompatibility
of injectable hydrogels and precursor/degradation product polymers: (A) MTT
viability assays of 3T3 mouse fibroblasts in the presence of both precursor
copolymers at various concentrations; (B) MTT viability assays of RPE retinal
pigment epithelial cells in the presence of both precursor copolymers at
various concentrations; (C-F) H&E stained
sections of mouse subcutaneous tissue: (C) 6 wt % poly(NIPAM-co-ADH) in PBS,
after 48 h; (D) 6 wt % poly(NIPAM-co-oxoethyl methacrylate) in PBS, after 48 h;
(E) PNIPAM in situ-formed hydrogel from 6 wt % polymer precursor solutions in
PBS, after 48 h; (F) PNIPAM in situ-formed hydrogel from 6 wt % polymer
precursor solutions in PBS, after five months. Tissue labels on (F) are
pertinent to all histological samples.

Conclusions

            We have developed a
novel method to synthesize covalently cross-linked and in situ-gellable thermoresponsive
hydrogels based on poly(N-isopropylacrylamide) using reversible and rapid
hydrazide-aldehyde chemistry to link functionalized PNIPAM oligomers. The
hydrogels exhibit the same thermal swelling-deswelling responses as
conventional PNIPAM hydrogels but can be degraded back into the reactive polymer
gel precursors via an acid-catalyzed hydrolysis process. Furthermore, the
combination of rapid gelation with low toxicity observed herein suggests that
this hydrazide-aldehyde oligomer cross-linking approach may be translatable to
the design of injectable, degradable covalently cross-linked hydrogels based on
a range of biocompatible synthetic polymers.

References

            1 ? Patenaude, M.; Hoare, T. ACS Macro Letters.
2012, 1,409-413.

            2 ? Garbern, J. C.; Stayton, P. S. ACS Biomacromolecules 2010,
11, 1833-1839.

Acknowledgments

Funding from the Natural Sciences
and Engineering Research Council of Canada (NSERC) and the Ontario Ministry of
Research and Innovation (Early Researcher Award program) is gratefully
acknowledged.