Advanced biomaterials for repairing the nervous system: what can hydrogels do for the brain?
Transcript of Advanced biomaterials for repairing the nervous system: what can hydrogels do for the brain?
RESEARCH:Review
Materials Today � Volume 00, Number 00 � June 2014 RESEARCH
Advanced biomaterials for repairing thenervous system: what can hydrogels dofor the brain?Zin Z. Khaing1,*, Richelle C. Thomas2, Sydney A. Geissler3 and Christine E. Schmidt1,*
1 J. Crayton Pruitt Family Department of Biomedical Engineering, The University of Florida, United States2McKetta Department of Chemical Engineering, The University of Texas at Austin, United States3Department of Biomedical Engineering, The University of Texas at Austin, United States
Newly developed hydrogels are likely to play significant roles in future therapeutic strategies for the
nervous system. In this review, unique features of the central nervous system (i.e., the brain and spinal
cord) that are important to consider in developing engineered biomaterials for therapeutic applications
are discussed. This review focuses on recent findings in hydrogels as biomaterials for use as (1) drug
delivery devices, specifically focusing on how the material can change the delivery rate of small
molecules, (2) scaffolds that can modify the post-injury environment, including preformed and
injectable scaffolds, (3) cell delivery vehicles, discussing cellular response to natural and synthetic
polymers as well as structured and amorphous materials, and (4) scaffolds for tissue regeneration,
describing micro- and macro-architectural constructs that have been designed for neural applications. In
addition, key features in each category that are likely to contribute to the translational success of these
biomaterials are highlighted.
IntroductionRecent advances both in our understanding of the nervous system
and the availability of sophisticated biomaterials have significant-
ly changed the landscape of potential strategies for repairing the
nervous system. New imaging and staining techniques along with
the development of novel materials have opened the possibility to
directly design biomaterials tailored for a particular application. In
this short review, recent advances in the use of hydrogels made
from both natural and synthetic polymers are discussed (see Fig. 1
as reference) and their evaluations using in vitro and in vivo pre-
clinical models are presented where applicable. This review is by
no means a comprehensive review of biomaterials for nervous
system repair: the readers are referred to a number of other excel-
lent review articles for further studies [1–11]. This review intro-
duces important obstacles to central nervous system (CNS)
regeneration focusing on the unique characteristics of the CNS.
Additionally, the injury environment and scarring are unique to
the CNS in that a glial scar introduces a physical and chemical
Please cite this article in press as: Z.Z. Khaing, et al., Mater. Today (2014), http://dx.doi.org/
*Corresponding authors:. Khaing, Z.Z. ([email protected]),
Schmidt, C.E. ([email protected])
1369-7021/� 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.mattod.2014.05.011
barrier to regeneration. Understanding the specific requirements
and taking them into consideration in designing therapeutic plat-
forms are likely to result in the successful development of next
generation biomaterials. Briefly, the blood barriers, the endoge-
nous immune system within the CNS, and the mechanical prop-
erties of CNS tissues are discussed. The focus of this review is on the
use of hydrogels as scaffolds to aid regeneration within the central
nervous system (CNS) (i.e., the brain and spinal cord). In particu-
lar, the following systems and applications are highlighted: (1)
hydrogels as drug delivery platforms, (2) hydrogels to modify the
post-injury environment, (3) hydrogels for cell delivery and (4)
hydrogels for tissue regeneration.
Features of the central nervous system to consider forbiomaterial developmentThere are a number of challenges that are unique to the CNS that
must be considered for development of therapeutic biomaterials.
First, the CNS is segregated from the circulating blood by the blood
brain barrier (BBB) and the blood spinal cord barrier (BSCB). These
barriers are made up of tight junctions between extracellular mem-
branes of endothelial cells and astrocytes. In normal physiological
10.1016/j.mattod.2014.05.011
1
RESEARCH Materials Today �Volume 00, Number 00 � June 2014
MATTOD-336; No of Pages 9
FIGURE 1
Synthetic and natural materials can be used to deliver small molecules or cells to the nervous system. Micro and macro architecture can also be introduced
to these hydrogels to alter injury environments and direct cell behavior. Chemical structures of some common polymers are presented here. Abbreviations:
PLGA – poly(lactic-co-glycolic acid), PEG – poly(ethylene glycol), PCL – polycaprolactone.
RESEARCH:Review
conditions, few molecules can passively move from the circulating
blood to the CNS extracellular fluid. This makes it difficult, if not
impossible, to deliver drugs and small molecule therapeutics to the
CNS intravenously. It has been suggested that hydrogels made from
both natural and synthetic polymers that can be intrathecally
placed into localized areas show great promise to deliver therapeu-
tics into the brain and spinal cord.
Under normal conditions, the BBB and BSCB also isolate the
CNS tissue from the circulating immune cells. Thus, the CNS is
considered an ‘immune privileged’ organ. Major immune cells
within the CNS include (1) microglia, which are resident immune
cells that are distributed throughout the brain and the spinal cord,
and (2) perivascular macrophages that are located in the capillar-
ies. After injury or in response to disease, microglia, astroctyes,
macrophages, oligodendroctyes and even neurons to an extent,
can all respond and release inflammatory cytokines [12]. There-
fore, for biomaterials to modify the injury environment, the
materials must interact with the resident cells’ immune response
in a positive manner [13].
The brain and spinal cord are some of the softest tissues in the
body with compressive moduli around 2000 Pa [14–16]. Matching
the mechanical properties of an implanted biomaterial to that of
the host tissue can significantly affect the success of the implanted
material in vivo [2,17]. In vitro studies of neural progenitor cell
(NPC) differentiation showed that hydrogels that best matched
the stiffness of the brain provided the most optimal results for
neuronal differentiation [16,18]. Hence, determining the appro-
priate mechanical property is a key factor to consider when using
biomaterials in the CNS.
Injury environment of the nervous systemA major difference between the peripheral nervous system (PNS)
and CNS is the capacity for peripheral nerves to regenerate; CNS
axons do not regenerate appreciably in their native environment.
Please cite this article in press as: Z.Z. Khaing, et al., Mater. Today (2014), http://dx.doi.org/
2
After injury in the CNS, macrophages infiltrate the site of injury
much more slowly than they do in the PNS. This delays the
removal of inhibitory myelin associated proteins from the injury
site. Macrophage recruitment is also limited because cell adhesion
molecules in the distal end of the injured spinal cord are not
appreciably up-regulated. Additionally, astrocytes in the CNS
become ‘reactive’, and produce glial scar tissue.
Astrocytic response after injury is characterized by cellular
hypertrophy, astrocyte proliferation, process extension, and in-
creased production of the intermediate filament proteins glial
fibrillary acidic protein (GFAP), vimentin, and nestin [19]. This
response of astrocytes to injury is known as reactive gliosis. When
an injury does not involve penetrating the dura mater, the scar
tissue formed is mainly composed of astrocytes; however, when
the dura is broken there is infiltration of meningeal fibroblasts in
addition to reactive astrocytes [20]. Most glial scar tissue is thought
to include, in addition to the reactive astrocytes, NG2+ oligoden-
drocyte precursor cells, meningeal cells, infiltrating macrophages,
and activated microglia. Schwann cells from adjacent dorsal roots
have also been found within the CNS scar tissue in experimental
injuries where the dura was broken. The mature glial scar includes
proteoglycans such as neurocan, phosphocan, versican, and bre-
vican along with secreted proteins including Semaphorin 3A and
3D [21]. One notable component of glial scar tissue is chondroitin
sulfate proteoglycans (CSPGs). CSPGs have been shown to inhibit
axonal outgrowth in many neuronal systems [22]. In sites distant
from traumatic injury, astrocytes can also become larger in size
and transform into a more pronounced stellate shape. These
‘activated astrocytes’ have been known to produce soluble trophic
factors that enhance the survival of neurons and glial cells in the
vicinity of the astrocytes [23]. Therefore, the effects of activated
astrocytes after injury on axonal regeneration and neuronal plas-
ticity is unclear. Moreover, myelin-related glycoproteins have long
been implicated in creating a non-hospitable environment for the
10.1016/j.mattod.2014.05.011
Materials Today � Volume 00, Number 00 � June 2014 RESEARCH
MATTOD-336; No of Pages 9
FIGURE 2
A schematic representation of the three fundamental ways in which
hydrogels are being developed for therapies to repair the nervous system
is shown here. (a) Hydrogels alone or as composite materials withmicroparticles have been used to create highly tunable delivery platforms
for small molecule therapeutics into the CNS. (b) Hydrogels can also be
used to deliver cells into the CNS. (c) Hydrogels have been utilized, withand without internal architecture, as scaffolds to deliver cells and repair
injured CNS tissue.
RESEARCH:Review
severed axons of the CNS [24–27]. Combined, these factors con-
stitute a formidable environment for regenerating axons in the
CNS.
Although the idea of simply inhibiting the formation of scar
after injury may be an attractive option, it is important to note that
formation of scar tissue after injury to the CNS encourages tissue
homeostasis. Contusion spinal cord injury (SCI) studies completed
in transgenic mice that targeted the depletion of mitotic astrocytes
immediately surrounding the injury site support this idea. The
results show an increased infiltration of inflammatory cells, cellu-
lar degeneration and increased lesioned area compared to control
mice with a similar injury [28]. Therefore it appears that there are
specific beneficial aspects of the glial cell response. Specifically the
production of cytokines, growth factors, and other extracellular
matrix (ECM) components that participate in re-establishing the
BBB after injury have been shown advantageous to the post-injury
environment. Simply inhibiting the formation of scar is not a
viable option. Instead, strategies that will specifically antagonize
the receptors for the inhibitors may be a better way to interfere
with the growth inhibitory nature of the injured CNS micro-
environment.
Hydrogels as drug delivery platformsAs stated above, delivering molecules to the CNS is a challenging
problem due to the presence of BBB and BSCB. Direct intrathecal
injections into the extracellular fluid have been successful in some
clinical [29–32] and experimental/pre-clinical settings [33,34].
Current research in this field involves examining biocompatible
and bio-inert polymers as platforms for controlled release of
bioactive molecules within the CNS for sustained drug delivery
(Fig. 2a). It is important to note that, after injury to the brain and
spinal cord, there is a temporary leak in the BBB and BSCB that
allows more molecules and cells to enter the CNS from the
circulating blood compared to normal physiological conditions
[35]. However, the extent to which this ‘leakiness’ can be taken
advantage of for cell and drug delivery is unclear. Hydrogels with
matching mechanical properties to that of the nervous tissue
(compressive modulus �2000 Pa) can be placed in the intrathecal
space to maintain homogeneous mechanical landscape with the
surrounding soft CNS tissue [16]. In light of this feature, hydrogels
made from a number of natural and synthetic polymers have been
examined for their ability to deliver therapeutics directly into the
brain and spinal cord.
Naturally derived polymers including fibrin [36–39], hyaluro-
nan–methylcellulose (HAMC) blend [40–45], hyaluronic acid [46–
48], agarose [49] and chitosan [50] have been used to successfully
deliver molecules to the CNS. For example, an injectable hydrogel
system developed by the Shoichet group shows promise for intra-
thecal delivery of growth factors into the injured spinal cord. This
group developed HAMC hydrogels that are safe, effective [51], and
able to deliver trophic factors [52]; modified versions of these
HAMC hydrogels can also be used as vehicles to deliver cells into
the injured spinal cord [53]. In one case, the delivery of erythro-
poietin (EPO) into the intrathecal space using HAMC hydrogels
after SCI in a rodent model resulted in reduction in lesion cavity
size and had a higher neuroprotective effect compared to direct
injection of EPO alone into the intrathecal space [52]. In other
cases, HAMC has been successfully combined with nanoparticles
Please cite this article in press as: Z.Z. Khaing, et al., Mater. Today (2014), http://dx.doi.org/
resulting in composite biomaterials with highly tunable delivery
profiles [42,54]. Fibrin-based hydrogels have also been successfully
used for delivery of growth factors [39,36,55,56], scar inhibiting
enzyme chondroitinase ABC [57], and progenitor/stem cells in
combination with growth factors [58–60] in rodent models of SCI.
Likewise, injectable, in situ-gelling forms of agarose have been
used in combination with lipid microtubules loaded with bioac-
tive molecules after SCI [49,61]. Bioactive molecules have been
delivered to the CNS with chitosan-based hydrogels [62–67],
microparticles [68–70] or nanoparticles [71–80]. Chitosan-based
biomaterials in particular have largely been explored for delivery
through the nasal route. This delivery paradigm is attractive since
the nasal passage has a large surface area, porous endothelial
membrane, high total blood flow and is readily accessible. Indeed,
10.1016/j.mattod.2014.05.011
3
RESEARCH Materials Today �Volume 00, Number 00 � June 2014
MATTOD-336; No of Pages 9
RESEARCH:Review
molecules that enter through the nasal mucosa can have direct
access to the brain and spinal cord, and can bypass the blood–brain
barrier via the olfactory and trigeminal associated pathways
[81,82]. The mucoadhesive property of chitosan increases the
dwell time and concentration of the deliverables at the site which
results in a higher local concentration and an increase of absorp-
tion. A variety of bioactive molecules including antibiotics [70],
insulin [65,81], anti-viral [62,63,80] and antidepressants [78,79]
have been successfully delivered into the CNS using chitosan-
based biomaterials. For example, Thoren et al. successfully used
chitosan-based materials to deliver insulin-like growth factor 1 as a
possible treatment of Alzheimer’s disease or stroke [81]. Others
have used chitosan to deliver vaccines for cell-mediated immune
response [62,63,67]. In summary, there is ample evidence that
natural-based polymers are well suited for use in nasal, intrathecal
and direct drug delivery applications into the CNS.
Although naturally derived biomaterials are referenced more
often for their innate bioactive properties, biocompatible synthet-
ic polymers are also commonly used. Synthetic biomaterials allow
more precise control of bulk properties and degradation rates
because synthesis parameters can be easily modified. To this
end, aligned and electrospun fibers [83] and freeze-dried macro-
porous scaffolds [84] have been made from poly-L-lactic acid (PLA)
to deliver bioactive molecules into both brain and spinal cord.
Moreover, synthetic polymers such as poly(lactic-co-glycolic acid)
(PLGA) have been widely used, in experimental settings, to pro-
duce microparticles [85–88], and nanoparticles [42,54,85,89,90].
Both micro- and nano-size PLGA particles loaded with bioactive
molecules have been used on their own [75,91], but these particles
have also been used in combination with hydrogels to localize the
particles (as mentioned above with HAMC hydrogels) and to
further tune the delivery profiles of the bioactive molecules
[54,85,89].
Synthetic polymers can also be used as modifiers for controlled
release of small molecules into the CNS. Poly(ethylene glycol)
(PEG) is one of the most commonly used synthetic biomaterials in
biomedical applications and PEG hydrogels have been developed
to deliver growth factors [1,18] and steroids [92] into the CNS in
pre-clinical rodent models. In some studies, PEG has been used to
modify polyethylenimine (PEI) as part of a gene delivery system.
PEGylation, the addition of PEG onto polymers, can increase
solubility and decrease protein binding in vivo. For example,
researchers compared the use of either PEI alone or PEI conjugated
with PEG for gene delivery to the spinal cord. They found that the
addition of one linear chain of PEG onto each PEI monomer was
sufficient to significantly increase the efficiency of transgene
expression in a rodent spinal cord [93]. PEG has also been used
to modify the growth factor for obtaining longer bioactivity by
increasing the half-life of the growth factor [94,95]. Moreover, a
combination of synthetic polymers (i.e., polyethylene glycol–
polycaprolactone (PEG–PCL) was used to deliver silencing RNAs
via a nasal delivery system in a pre-clinical rodent model [96].
As mentioned previously, bulk properties and degradation rates
are more easily modified in synthetic polymers. This can be highly
advantageous in developing a platform that can then be modified
for tissue or disease specific applications. However, the degrada-
tion products of both synthetic and natural polymers can be
bioactive and, to the best of our knowledge, currently there is
Please cite this article in press as: Z.Z. Khaing, et al., Mater. Today (2014), http://dx.doi.org/
4
limited information on possible biological effects of degradation
products from synthetic polymers [97,98]. For example, lactic acid
and glycolic acid are known byproducts of PLGA. Recently, an
intriguing article by Rinholm et al. suggests that exogenous appli-
cation of lactic acid to cultured slices of developing mouse brain
cortex can support oligodendrocyte development and myelina-
tion [99]. Future studies examining the presence and possible
effects of degradation products from commonly used polymers,
such as PLGA, are warranted when applied in vivo.
Hydrogels to modify the post-injury environmentHydrogels can be particularly useful in the CNS post injury envi-
ronment by potentially serving as a substrate to re-establish tissue
continuity. Prefabricated scaffolds have been implanted [100], but
can cause secondary injury to the surrounding tissue during
delivery [101]. Minimally invasive injectable hydrogels are partic-
ularly well suited for these types of applications. The liquid poly-
mer can fill irregularly shaped void spaces and injections via
needles to deliver the liquid polymer circumvent the BBB and
BSCB. The physical properties of hydrogels that are relevant to
consider in this post-injury environment are the porosity, chemi-
cal composition and mechanical properties. Hydrogel porosity is
essential to stabilize the post-injury environment by permitting
nutrient flow into and out of the scaffold. Porosity also affects cell
infiltration [2], cell distribution, as well as cell growth, prolifera-
tion, vascularization and local angiogenesis. Hydrogel interaction
with host cells is also largely a result of the hydrogels’ chemical and
mechanical compatibility. Mechanically, injectable hydrogels
with storage moduli similar to surrounding tissue have been the
most successful because they are able to more closely mimic the
native environment [102–105].
In one study, the implantation of injectable fibrin hydrogels
enhanced neural fiber sprouting and dural resealing following SCI
[36]. Another injectable hydrogel, HAMC, was placed into the
spinal cord after injury and the results showed that there was
reduced inflammation in the animals with HAMC injection com-
pared to controls [44]. Preformed HA hydrogels implanted into the
transected spinal cord showed that animals with HA implants had
lowered macrophage/microglia cell infiltration and astrocytic re-
sponse compared to control animals [100]. In some cases, PEG
administered following SCI has been shown to help repair dam-
aged membranes and prevent paralysis in pre-clinical models
[106–108]. In the clinic, there are a number of FDA-approved dura
replacement membrane products made from synthetic (Dura-
SealTM from PEG hydrogel) and natural-based polymers (Dur-
GenTM from collagen, SynthecelTM from cellulose) with varying
success. Synthetic polymers do not provide natural cues and are
more difficult for the host’s cells to remodel [109] compared to
hydrogels made from native ECM molecules, and therefore may be
more suitable as barriers but are less suitable for modifying the
cellular response in the post-injury environment.
Hydrogels for cell deliverySource and cell type have been the focus of cell transplantation
research in the nervous system, with attempts to implant Schwann
cells [110,111], NPCs [112], bone marrow stromal cells [113,114],
adult stem cells [115], induced pluripotent stem cells [116] and
embryonic stem cells [117], among others (Fig. 2b). Recently,
10.1016/j.mattod.2014.05.011
Materials Today � Volume 00, Number 00 � June 2014 RESEARCH
MATTOD-336; No of Pages 9
RESEARCH:Review
research has also focused on mode of cell delivery. Improving
widespread problems with low transplanted cell viability after SCI
[118], and uncontrolled cell differentiation (transplanted and
endogenous cells) in an ischemic stroke model [119] have come
to the forefront of today’s research. Biomaterial scaffolds can
mitigate a number of these issues by acting as delivery vehicles
for cells into injured areas of the nervous system, such as after SCI
or traumatic brain injury (TBI) [120,121]. Scaffolds that include
natural ECM proteins can direct cell behavior by providing cues to
cells during migration [122], differentiation [15], and regeneration
after TBI, stroke, and in other injury models [2,123]. These attri-
butes allow hydrogels to provide benefit to the endogenous cells
surrounding the implant site as well as to the transplanted cells.
Cells are often implanted directly into the lesioned cavity to
repair the injured CNS tissue, therefore maintaining cell viability
at the site of interest is important [124]. Xiong et al. placed
collagen I scaffolds with human marrow stromal cells into a TBI
model and observed a significant increase in cell viability com-
pared to cells transplanted in culture medium only [125]. In a
separate investigation, thiol crosslinked hyaluronan–heparin–col-
lagen–polyethylene glycol diacrylate hydrogels were used as NPC
carriers to the necrotic stroke cavity. Again, significant cell survival
of transplanted cells was observed compared to cells transplanted
in medium alone (Fig. 3a) [98]. Notably, when the cells were
transplanted in combination with gels, there was a significant
reduction in microglia/macrophage infiltration. This is an inter-
esting observation because it suggests that the presence of hydro-
gels either modified the post-injury inflammatory environment
(as discussed in the previous section) and/or reduced the inflam-
mation in response to the presence of transplanted cells (Fig. 3b)
[98]. In a separate study using a complete transection model of
thoracic SCI, Lu et al. demonstrated success at increasing trans-
planted cell viability as well as functional recovery after SCI with
the delivery of NPCs in a fibrin matrix compared to cells in
medium (Fig. 3c) [120]. It is important to note that in the Lu
study, the researchers included a cocktail of growth factors in the
pre-hydrogel solution. Therefore, combinations of supportive scaf-
folds, growth factors and appropriate cells are needed for effec-
tively repairing the injured tissue.
Transplanted cells can also serve as a consistent source of growth
factors when they are delivered to the CNS [126]. To achieve this,
these cells must be protected from the immune system for long-
term viability. Some researchers are using microspheres with cells
encapsulated inside or adhered to the outside to protect the cells
from the immune system upon delivery in vivo [69,127]. Specifi-
cally, alginate capsules can shield brain derived neurotrophic
factor (BDNF)-producing fibroblasts from the immune system
when implanted in a SCI model [128,129]. This is an intriguing
option to deliver cells to the injured CNS without immune sup-
pression; this option could provide a viable treatment without
additional medication. In using biomaterials to isolate trans-
planted cells from the immune system, synthetic materials may
provide more immune protection than natural polymers because
of the absence of bioactive components.
Neural stem and progenitor cells have also been transplanted
with the intent that these cells will differentiate into specific cell
types that can participate in the recovery process. In this case,
utilizing biomaterials with natural ECM molecules, matching
Please cite this article in press as: Z.Z. Khaing, et al., Mater. Today (2014), http://dx.doi.org/
mechanical properties, and introducing tissue relevant factors
and topographical cues are ways to direct cell differentiation.
Hyaluronic acid-based hydrogels with mechanical properties sim-
ilar to that of brain and spinal cord tissue have been shown to
direct NPC differentiation [16]. Moreover, NPCs can also respond
to structural cues and align to follow controlled microstructure
created within hyaluronic acid hydrogels [130]. In 2009, Christo-
pherson et al. examined the effects of fiber diameter using elec-
trospun polyethersulfone nanofibers coated with laminin on NPC
differentiation [131]. In this study, researchers found that under
differentiation conditions, adult NPCs exhibited increased oligo-
dendrocyte differentiation on 283 nm fibers and increased neuro-
nal differentiation on 749 nm fibers compared to cells grown on
tissue culture plates [131]. Therefore, topographical cues and
mechanical cues within hydrogels are additional tools to direct
the fate of NPCs. In a different investigation, Lim et al. observed
that aligned electrospun polycaprolactone (PCL) fibers significant-
ly increased NPC differentiation toward neurons and alignment
when compared to NPCs grown on tissue culture treated plastic or
randomly oriented fibers [132]. Effects of fiber diameter on neu-
rons and glia were examined. The authors found that topographi-
cal features such as large fiber diameter and aligned fibers can
induce apoptosis of glial cells, increasing the percentage of neuro-
nal cells. This is an exciting finding that introduces additional
features for creating more defined regions for specific cell types
within a hdyrogel; for example, fiber diameter could control the
cell phenotype boundary of white matter and gray matter within
the spinal cord [132]. It seems clear that currently available bio-
materials can increase cell viability, modify the inflammatory
response and encourage differentiation of implanted cells. Next
generation hydrogels for CNS repair will likely include native ECM
components and similar topography to native tissue.
Hydrogels for tissue regenerationThe brain and spinal cord consist of highly specific regions and
intricate architecture (e.g., the hippocampus and subventricular
zone in the brain, white and gray matter of the spinal cord) [133].
These geometrically complex regions have been explored as design
criteria for developing biomaterials for the CNS. Many fundamen-
tal studies using cultured cells have shown that creating a complex
topographical landscape [130] with submicron and nanoscale
features [130] can direct neural cell behavior. As mentioned in
the previous section, injectable hydrogels have the ability to
conform to the shape of irregular cavities, but creating architec-
tures and altering pore sizes within hydrogel is difficult [101]. In
vitro studies of electrospun nanofibers showed embryonic stem cell
differentiation into neurons, astrocytes and oligodendrocytes
[134]. Moreover, the presence of parallel nanofibers enhanced
the neurite extension and orientation [134]. Seidlits et al. reported
increased neurite outgrowth of hippocampal cells on 3D micro-
structures created using 2-photon lithography. The authors creat-
ed sub-micron structures of bovine serum albumin coated with
laminin-derived peptides within HA scaffolds that directed neurite
extension and orientation [135]. In another investigation,
researchers compared hippocampal NPC affinity to polydimethyl-
siloxane microsctructure to chemical cues (laminin or nerve
growth factor) in vitro. NPCs more preferentially extended neurites
onto the microstructures than onto non-structured chemically
10.1016/j.mattod.2014.05.011
5
RESEARCH Materials Today �Volume 00, Number 00 � June 2014
MATTOD-336; No of Pages 9
Please cite this article in press as: Z.Z. Khaing, et al., Mater. Today (2014), http://dx.doi.org/10.1016/j.mattod.2014.05.011
FIGURE 3
Hydrogels can be used to deliver cells to an injury site. This has been shown to increase cell viability, decrease immune response, have an effect on cell
differentiation, and increase functional recovery. (a) Zhong et al. transplanted neural progenitor cells (NPCs) into the stroke cavity 7 days after focal ischemiastroke induction with or without hyaluronan-heparin-collagen hydrogels [98]. Two weeks after transplantation, transplanted cells were immunostained with
anti-green fluorescent protein (GFP) and quantified by stereological optical fractionator procedures. Cell viability significantly increased when cells were
transplanted with a hydrogel compared to cells in saline (p = 0.035). The increase in cell viability is attributed to the delivery with a hydrogel scaffold? (b)Zhong et al. also quantified the immune response after NPC transplantation by immunostaining transplanted NPCs with anti-GFP and active macrophages
and microglia with anti-Iba1. Activated macrophages and microglia were quantified around the injury site. There was a significantly decreased quantity of
infiltrating activated macrophages and microglia in the injury site with composite hydrogel and cell implantation over cell implantation alone (p = 0.004),
quantified by stereological optical fractionator procedures. This shows the positive effect of the hydrogel on the injury environment or on the implantedcells ability to alter the injury environment. (c) Lu et al. transplanted NPCs to fill a thoracic level complete transection site 2 weeks following injury [114].
7 weeks post transplantation, transplanted NPCs (GFP) were observed to be evenly distributed throughout the injury area (1), to differentiate into mature
neurons (2), to robustly extend axons beyond the lesion into the caudal spinal cord in gray matter and white matter (3–5), and rats were seen to have
significant functional improvement over the animals with saline alone injections (6, p < 0.01). These results support the idea that the presence of hydrogelscaffold during transplantation can distribute cells throughout the transection area as well as provide support for neuronal differentiation of transplanted
NPCs.
6
RESEARCH:Review
Materials Today � Volume 00, Number 00 � June 2014 RESEARCH
MATTOD-336; No of Pages 9
RESEARCH:Review
coated substrates [136], suggesting that microstructure plays a
critical role in cell behavior. In 2010, Baiguera et al. made electro-
spun PCL into fibrous membranes with micron-scale and submi-
cron-scale features. Hippocampal astroctyes and endothelial cells
cultured on micron-scale features allowed cell infiltration whereas
submicron-scale features did not [137]. Therefore, determining the
correct feature size for a specific application is an important
parameter to consider. These fundamental findings from in vitro
work have been incorporated in hydrogel-based biomaterials and
the resulting hydrogels have been used in vivo with some success
[138,141].
Architecture within hydrogels can also be accomplished by
introducing a porous network (Fig. 2c). Micro- and macrostructure
in hydrogel scaffolds have been created using a number of meth-
ods including porogen leaching [138], freeze drying [97,101],
crystal templating [139], multiphoton lithography [135] and
molding [140]. Porogen leaching and gas foaming create pore
sizes relatively consistent with the size of the porogen or gas used
to create the pores [102,138]. It is not entirely clear if having a
consistent pore size is a critical requirement for successful bioma-
terial integration but optimum pore sizes have been determined
for a number of applications [102]. Martinez-Ramos et al. used
porogen leaching to synthesize poly(ethyl acrylate)–poly (hydrox-
yl ethyl acrylate) copolymer blend matrices with two different
geometries for implantation near the subventricular zone (SVZ) of
the brain [138]. Pre-clinical in vivo implants of hydrogels with
aligned 40 mm diameter pores and with interconnected orthogo-
nal 80 mm pores showed local angiogenesis and neurite extension
within each of the scaffolds in the SVZ [138]. In in vitro experi-
ments, NPCs readily infiltrated the scaffold and differentiated into
neurons and astrocytes regardless of the scaffold geometry [138]. It
appears from this study that specific scaffold geometry does not
play a significant role in scaffold integration or cell penetration in
the brain; it is likely that the overall porosity plays the largest role.
In another study, PCL scaffolds were created with two kinds of
internal architectures using a molding method: (1) longitudinal
channels with microgrooves within a cylinder, or (2) orthogonally
intersecting channels and axial microgrooves within a cylinder
[141]. When implanted into adult rat cortex, the group with
orthogonal channel implants that more closely mimic the native
ECM showed the most tissue in-growth compared to the longitu-
dinal channel implant or control scaffold implant groups. It is
important to note that in this study the pore sizes were hundreds
of microns. Therefore, when the pores are larger than 100 mm the
internal architecture may play a more significant role in tissue in-
growth from the host. In 2005, Tian et al. implanted freeze-dried
hyaluronic acid-poly-D-lysine composite gels with mean pore dia-
meters ranging from 90 to 230 mm in a TBI model in adult rats.
These porous scaffolds with internal architecture (fiber-like struts
and sheet-like features) showed successful integration with host
tissue, endothelialization and new ECM formation by six weeks
[97].
Macro-architecture that mimics the natural tissue is thought to
encourage regeneration. Repairing the spinal cord presents an
interesting challenge to biomaterials scientists and engineers in
that the white and gray matter architectures are dramatically
different. Depending on the injury type and location, different
architectures may be more relevant. The white matter contains
Please cite this article in press as: Z.Z. Khaing, et al., Mater. Today (2014), http://dx.doi.org/
well-aligned axon tracts, whereas the gray matter houses local
motor neurons and interneuron pools and would likely require less
directed growth. Friedman et al. theorized that creating architec-
ture with specific features to mimic spinal cord tracts would
encourage regeneration [142], however, the suggested designs
have not been fully developed. Hydrogel scaffolds with macro-
architecture have been explored recently. Stokols et al. designed
templated agarose scaffolds for use after dorsal column lesions.
The extrusion method created aligned linear cylindrical structure
for white matter regeneration. The authors observed host integra-
tion including Schwann cell infiltration and vascularization in
implants. Further, the researchers noted a minimal increase in
inflammation near the tissue-injury interface compared to un-
treated animals. Axon extension was observed only when Matri-
gel1 or fibrin matrices were included within the implants [143],
suggesting that natural ECM cues are required for axonal growth
within the graft. Hydrogel scaffolds with slightly more complex
architecture have been explored recently to repair the spinal cord.
Wong et al. cast five PCL scaffolds with different architectures and
implanted them into a complete transection thoracic spinal cord.
The authors compared three designs with non-specific architec-
ture and two designs with specific architecture that more closely
mimics white and gray matter of the spinal cord. At three months,
non-specific architecture implants had higher inflammatory re-
sponse (increased macrophages, fibroblasts, astrocytes) as well as
lesion expansion compared to the structure-mimicking implants.
Scaffolds with the structure-mimicking designs also encouraged
more axonal infiltration, astrocyte migration and myelinated
axon growth into the material compared to the other designs
[140]. In this study, meaningful functional recovery was not
assessed, but it provided a strong argument in support of anatom-
ically relevant macro-architectural cues for enhancing regenera-
tion.
In summary, when it comes to incorporating structure into
hydrogel scaffolds, researchers must determine the most relevant
biological features, and size scale and determine how to create
similar architectures within biomaterials. Although, there are new
scaffolds being tested in pre-clinical models in the laboratory,
currently there are none available in the clinic. Future work in
this area needs to focus on fine-tuning the optimal morphological
layout, length scales, and feature sizes. Molding and multiphoton
lithography show the most promise as methods to create robust,
relevant geometry within hydrogels because they provide precise
control over macro- and microarchitecture. It is clear from funda-
mental studies that topography can play a significant role in
directing cell behavior. Therefore, biomaterials that better mimic
the native architecture will continue to be an important area of
study for CNS regeneration.
Concluding remarksLooking ahead, advances being made in the field of biomaterial
science will significantly impact the next generation of therapies
for repair in the CNS. Hydrogels in particular are strong candidates
for clinical relevance in the future. Indeed, as mentioned in this
short review, there are a number of hydrogel genres that have been
used successfully in the CNS in the laboratory. However, before
any biomaterial can be used in the clinic, we must consider
thoroughly how materials interact with surrounding cells and
10.1016/j.mattod.2014.05.011
7
RESEARCH Materials Today �Volume 00, Number 00 � June 2014
MATTOD-336; No of Pages 9
RESEARCH:Review
host tissue in vivo. Key considerations and material features that we
feel are important include: (1) a better understanding of the
degradation products of parent polymers used for drug delivery
applications, (2) equal consideration of the possible routes of
delivery including nasal, intrathecal and direct implantation,
(3) careful examination of possible biological effects of the hydro-
gels and their degrading products, (4) inclusion of native ECM
components within the scaffolds during cell delivery for better cell
survival and integration with the host tissue and (5) addition of
biologically relevant topography and architectural features within
scaffolds. Significant advances have been made to improve both
functional recovery and cell viability following SCI. Additionally,
progenitor cell differentiation and the reduction in inflammation
as a result of hydrogel implantation are notable achievements in
this area of research. Future efforts should focus on a more com-
prehensive approach to developing and studying biomaterials
which include relevant in vivo models.
AcknowledgmentsThe authors wish to acknowledge related funding from the
Mission Connect (#012-113 Z.Z.K. and C.E.S.), The Craig Neilsen
Foundation (#222456, C.E.S.), NSF-CBET (#1159774 C.E.S.), NSF-
DMR (#0805298 C.E.S.), NIH R21 (#R21 NS074162 C.E.S.), NSF-
GFRP (#2011112479, S.A.G.), and the National GEM Consortium
(R.C.T.).
References
[1] D.R. Nisbet, et al. J. Biomater. Appl. 24 (1) (2009) 7.
[2] J.T.S. Pettikiriarachchi, et al. Aust. J. Chem. 63 (8) (2010) 1143.
[3] D. Macaya, M. Spector, Biomed. Mater. 7 (1) (2012) 012001.
[4] D.A. McCreedy, S.E. Sakiyama-Elbert, Neurosci. Lett. 519 (2) (2012) 115.
[5] K. Straley, S.C. Heilshorn, Conf. Proc. IEEE Eng. Med. Biol. Soc. 2009 (2009) 2101.
[6] Y. Zhong, R.V. Bellamkonda, J. R. Soc. Interface 5 (26) (2008) 957.
[7] Z.Z. Khaing, C.E. Schmidt, Neurosci. Lett. 519 (2) (2012) 103.
[8] H. Kim, et al. Trends Biotechnol. 30 (1) (2012) 55.
[9] G. Orive, et al. Nat. Rev. Neurosci. 10 (9) (2009) 682.
[10] H. Cao, et al. Adv. Drug Deliv. Rev. 61 (12) (2009) 1055.
[11] R. Uibo, et al. Biochim. Biophys. Acta 1793 (5) (2009) 924.
[12] O.K. Bitzer-Quintero, I. Gonzalez-Burgos, Neural Plast. 2012 (2012) 348642.
[13] R.J. McKeon, et al. J. Neurosci. 11 (11) (1991) 3398.
[14] P.C. Georges, et al. Biophys. J. 90 (8) (2006) 3012.
[15] A.J. Engler, et al. Cell 126 (4) (2006) 677.
[16] S.K. Seidlits, et al. Biomaterials 31 (14) (2010) 3930.
[17] H. Liao, et al. Acta Biomater. 4 (5) (2008) 1161.
[18] J.A. Burdick, et al. Biomaterials 27 (3) (2006) 452.
[19] H.Y. Yang, et al. Mol. Chem. Neuropathol. 21 (2–3) (1994) 155.
[20] M.C. Shearer, J.W. Fawcett, Cell Tissue Res. 305 (2) (2001) 267.
[21] R.J. Pasterkamp, et al. Mol. Cell. Neurosci. 13 (2) (1999) 143.
[22] J. Silver, J.H. Miller, Nat. Rev. Neurosci. 5 (2) (2004) 146.
[23] C.M. Liberto, et al. J. Neurochem. 89 (5) (2004) 1092.
[24] P.A. Harvey, et al. J. Neurosci. 29 (19) (2009) 6285.
[25] W.B. Cafferty, S.M. Strittmatter, J. Neurosci. 26 (47) (2006) 12242.
[26] X. Wang, et al. J. Neurosci. 22 (13) (2002) 5505.
[27] T. GrandPre, et al. Nature 417 (6888) (2002) 547.
[28] J.R. Faulkner, et al. J. Neurosci. 24 (9) (2004) 2143.
[29] B. Gardner, et al. Paraplegia 33 (10) (1995) 551.
[30] T.M. Miller, et al. Lancet Neurol. 12 (5) (2013) 435.
[31] G. Lind, et al. Acta Neurochir. Suppl. 97 (Pt 1) (2007) 57.
[32] S. Erdine, J. De Andres, Pain Pract. 6 (1) (2006) 51.
[33] B.R. Felice, et al. Toxicol. Pathol. 39 (5) (2011) 879.
[34] C.E. Kang, et al. J. Control. Release 144 (1) (2010) 25.
[35] A.S. Lossinsky, R.R. Shivers, Histol. Histopathol. 19 (2) (2004) 535.
[36] P.J. Johnson, et al. Biotechnol. Bioeng. 104 (6) (2009) 1207.
[37] P.J. Johnson, et al. Soft Matter 6 (20) (2010) 5127.
[38] S.E. Sakiyama-Elbert, J.A. Hubbell, J. Control. Release 69 (1) (2000) 149.
[39] S.J. Taylor, et al. J. Control. Release 113 (3) (2006) 226.
Please cite this article in press as: Z.Z. Khaing, et al., Mater. Today (2014), http://dx.doi.org/
8
[40] J.W. Austin, et al. Biomaterials 33 (18) (2012) 4555.
[41] M.D. Baumann, et al. J. Control. Release 138 (3) (2009) 205.
[42] M.D. Baumann, et al. Biomaterials 31 (30) (2010) 7631.
[43] M.J. Caicco, et al. J. Biomed. Mater. Res. A 101 (5) (2013) 1472.
[44] D. Gupta, et al. Biomaterials 27 (11) (2006) 2370.
[45] M.J. Cooke, et al. Biomaterials 32 (24) (2011) 5688.
[46] W.M. Tian, et al. J. Control. Release 102 (1) (2005) 13.
[47] R.K. Mittapalli, et al. Mol. Cancer Ther. 12 (11) (2013) 2389.
[48] Y. Wang, et al. Pharm. Res. 28 (6) (2011) 1406.
[49] A. Jain, et al. PLoS ONE 6 (1) (2011) e16135.
[50] Z. Yang, et al. Biomaterials 31 (18) (2010) 4846.
[51] M.S. Shoichet, et al. Prog. Brain Res. 161 (2007) 385.
[52] C.E. Kang, et al. Tissue Eng. Part A 15 (3) (2009) 595.
[53] A.J. Mothe, et al. Biomaterials 34 (15) (2013) 3775.
[54] C.E. Kang, et al. Cells Tissues Organs 197 (1) (2013) 55.
[55] M.D. Wood, et al. J. Biomed. Mater. Res. A 89 (4) (2009) 909.
[56] S.J. Taylor, S.E. Sakiyama-Elbert, J. Control. Release 116 (2) (2006) 204.
[57] A.J. Hyatt, et al. J. Control. Release 147 (1) (2010) 24.
[58] P.J. Johnson, et al. Cell Transplant. 19 (1) (2010) 89.
[59] S.M. Willerth, et al. Stem Cells 25 (9) (2007) 2235.
[60] S.M. Willerth, et al. Stem Cell Res. 1 (3) (2008) 205.
[61] A. Jain, et al. Biomaterials 27 (3) (2006) 497.
[62] Y. Wu, et al. Eur. J. Pharm. Biopharm. 81 (3) (2012) 486.
[63] Y. Wu, et al. Biomaterials 33 (7) (2012) 2351.
[64] H. Nazar, et al. Eur. J. Pharm. Biopharm. 77 (2) (2011) 225.
[65] A.K. Agrawal, et al. Pharmazie 65 (3) (2010) 188.
[66] I.A. Alsarra, et al. Drug Dev. Ind. Pharm. 35 (3) (2009) 352.
[67] J. Wu, et al. Biomaterials 28 (13) (2007) 2220.
[68] E. Gavini, et al. J. Pharm. Sci. 100 (4) (2011) 1488.
[69] N.B. Skop, et al. Acta Biomater. 9 (6) (2013) 6834.
[70] E. Gavini, et al. J. Pharm. Sci. 100 (2011) 1488–1502.
[71] J. Malmo, et al. PLoS ONE 8 (1) (2013) e54182.
[72] M. Yemisci, et al. Methods Enzymol. 508 (2012) 253.
[73] A. Mahapatro, D.K. Singh, J. Nanobiotechnol. 9 (2011) 55.
[74] A. Hafner, et al. J. Microencapsul. 28 (8) (2011) 807.
[75] Y. Cho, R.B. Borgens, Exp. Neurol. 233 (1) (2012) 126.
[76] A. Mistry, et al. J. Drug Target. 17 (7) (2009) 543.
[77] K. Nagpal, et al. Drug Deliv. 20 (3–4) (2013) 112.
[78] K. Nagpal, et al. Int. J. Biol. Macromol. 59 (2013) 20.
[79] K. Nagpal, et al. Drug Deliv. 19 (8) (2012) 378.
[80] K. Nagpal, et al. Chem. Pharm. Bull. (Tokyo) 58 (11) (2010) 1423.
[81] R.G. Thorne, et al. Neuroscience 127 (2) (2004) 481.
[82] U. Westin, et al. Eur. J. Pharm. Sci. 24 (5) (2005) 565.
[83] N.J. Schaub, R.J. Gilbert, J. Neural Eng. 8 (4) (2011) 046026.
[84] C.M. Patist, et al. Biomaterials 25 (9) (2004) 1569.
[85] S.A. Chvatal, et al. Biomaterials 29 (12) (2008) 1967.
[86] A. Goraltchouk, et al. J. Control. Release 110 (2) (2006) 400.
[87] G.E. Rooney, et al. Tissue Eng. Part A 17 (9–10) (2011) 1287.
[88] E.Y. Tan, et al. Pharm. Res. 24 (12) (2007) 2297.
[89] J.C. Stanwick, et al. J. Control. Release 160 (3) (2012) 666.
[90] Y.C. Wang, et al. Biomaterials 29 (34) (2008) 4546.
[91] Y.T. Kim, et al. Biomaterials 30 (13) (2009) 2582.
[92] P.J. Gaillard, et al. J. Control. Release 164 (3) (2012) 364.
[93] G.P. Tang, et al. Biomaterials 24 (13) (2003) 2351.
[94] D.P. Ankeny, et al. Exp. Neurol. 170 (1) (2001) 85.
[95] R.G. Soderquist, et al. J. Biomed. Mater. Res. A 91 (3) (2009) 719.
[96] T. Kanazawa, et al. Biomaterials 34 (36) (2013) 9220.
[97] W.M. Tian, et al. Tissue Eng. 11 (3–4) (2005) 513.
[98] J. Zhong, et al. Neurorehabil. Neural Repair 24 (7) (2010) 636.
[99] J.E. Rinholm, et al. J. Neurosci. 31 (2) (2011) 538.
[100] Z.Z. Khaing, et al. J. Neural Eng. 8 (4) (2011) 046033.
[101] P.Z. Elias, M. Spector, J. Tissue Eng. Regen. Med. (2012), http://dx.doi.org/
10.1002/term.1621.
[102] N. Annabi, et al. Tissue Eng. Part B: Rev. 16 (4) (2010) 371.
[103] K. Whang, et al. Tissue Eng. 5 (1) (1999) 35.
[104] S. Woerly, Biomaterials 14 (14) (1993) 1056.
[105] R.J. Wade, J.A. Burdick, Mater. Today 15 (10) (2012) 454.
[106] S. Kouhzaei, et al. Neurol. Res. 35 (4) (2013) 415.
[107] A. Nehrt, et al. J. Neurotrauma 27 (1) (2010) 151.
[108] R.B. Borgens, et al. J. Exp. Biol. 205 (Pt 1) (2002) 1.
[109] J. Patterson, et al. Mater. Today 13 (1–2) (2010) 14.
[110] T. Hadlock, et al. Tissue Eng. 6 (2) (2000) 119.
[111] H.E. Olson, et al. Tissue Eng. Part A 15 (7) (2009) 1797.
10.1016/j.mattod.2014.05.011
Materials Today � Volume 00, Number 00 � June 2014 RESEARCH
MATTOD-336; No of Pages 9
ARCH:Review
[112] J.D. Kocsis, et al. J. Neurotrauma 21 (4) (2004) 441.
[113] H. Itosaka, et al. Neuropathology 29 (3) (2009) 248.
[114] D. Lu, et al. Neurosurgery 61 (3) (2007) 596.
[115] M. Tohill, et al. Neurosci. Lett. 362 (3) (2004) 200.
[116] A. Wang, et al. Biomaterials 32 (22) (2011) 5023.
[117] L. Cui, et al. Stem Cells 26 (5) (2008) 1356.
[118] A.M. Parr, et al. Neuroscience 155 (3) (2008) 760.
[119] S. Kelly, et al. Proc. Natl. Acad. Sci. U.S.A. 101 (32) (2004) 11839.
[120] P. Lu, et al. Cell 150 (6) (2012) 1264.
[121] Y.J. Ren, et al. J. Bioact. Compat. Polym. 24 (1) (2009) 56.
[122] E. Tziampazis, et al. Biomaterials 21 (5) (2000) 511.
[123] W. Daly, et al. J. R. Soc. Interface 9 (67) (2012) 202.
[124] M.J. Cooke, et al. Soft Matter 6 (20) (2010) 4988.
[125] Y. Xiong, et al. Brain Res. (2009) 1263–2183.
[126] B. Rihova, Adv. Drug Deliv. Rev. 42 (1–2) (2000) 65.
[127] R.M. Hernandez, et al. Adv. Drug Deliv. Rev. 62 (7–8) (2010) 711.
Please cite this article in press as: Z.Z. Khaing, et al., Mater. Today (2014), http://dx.doi.org/
[128] C.A. Tobias, et al. J. Neurotrauma 22 (1) (2005) 138.
[129] C.A. Tobias, et al. J. Neurotrauma 18 (3) (2001) 287.
[130] S.K. Seidlits, et al. Nanomedicine (Lond.) 3 (2) (2008) 183.
[131] G.T. Christopherson, et al. Biomaterials 30 (4) (2009) 556.
[132] S.H. Lim, et al. Biomaterials 31 (34) (2010) 9031.
[133] S.C. Joshi, et al. Int. J. Pattern Recognit. Artif. Intell. 11 (08) (1997) 1317.
[134] J. Xie, et al. Nanoscale 2 (1) (2010) 35.
[135] S.K. Seidlits, et al. Adv. Funct. Mater. 20 (9) (2010) 1368.
[136] N. Gomez, et al. J. R. Soc. Interface 4 (13) (2007) 223.
[137] S. Baiguera, et al. J. Mater. Sci. Mater. Med. 21 (4) (2010) 1353.
[138] C. Martinez-Ramos, et al. J. Biomed. Mater. Res. A 100 (12) (2012) 3276.
[139] S.A. Zawko, C.E. Schmidt, Acta Biomater. 6 (7) (2010) 2415.
[140] D.Y. Wong, et al. J. Neurotrauma 25 (8) (2008) 1027.
[141] D.Y. Wong, et al. J. Neurosurg. 109 (4) (2008) 715.
[142] J.A. Friedman, et al. Neurosurgery 51 (3) (2002) 742.
[143] S. Stokols, et al. Tissue Eng. 12 (10) (2006) 2777.
10.1016/j.mattod.2014.05.011
9
RESE