10 1016j addr 2012 03 009

Prévia do material em texto

Advanced Drug Delivery Reviews xxx (2012) xxx–xxx
ADR-12276; No of Pages 26
Contents lists available at SciVerse ScienceDirect
Advanced Drug Delivery Reviews
j ourna l homepage: www.e lsev ie r .com/ locate /addr
Naturally and synthetic smart composite biomaterials for tissue regeneration☆
Román A. Pérez a, Jong-Eun Won a,b, Jonathan C. Knowles b,c, Hae-Won Kim a,b,d,⁎
a Institute of Tissue Regeneration Engineering (ITREN), Dankook University, Cheonan 330-714, Republic of Korea
b Biomaterials and Tissue Engineering Lab, Department of Nanobiomedical Science & WCU Research Center, Dankook University, Cheonan 330-714, Republic of Korea
c UCL Eastman Dental Institute, University College London, 256 Gray's Inn Road, WC1X8LD London, UK
d Department of Biomaterials Science, School of Dentistry, Dankook University, Cheonan 330-714, Republic of Korea
Abbreviations: BDNF, brain derived neurotrophic fact
phate; CCBD, central cell binding domain; CNT, carbon
DOX, doxorubicin; ECM, extracellular matrix; EGF, epid
granulocyte colony stimulating factor; GF, growth factor
HPMC, hydroxypropylmethylcellulose; HUVEC, human
(3-isocyanatopropyl)triethoxysilane; MSC, mesenchym
phin-3; OCN, osteocalcin; OPN, osteopontin; PA, peptid
methylsiloxane; PEG, poly(ehylene glycol); PEI, polye
PHPMA, poly[n-(2-hydroxypropyl)methacrylamide]; PH
(L-lactic acid); PMMA, poly(methyl methacrylate); pNIP
(vinyl alcohol); RGD, Arg–Gly–Asp (amino acid sequen
VEGF, vascular endothelial growth factor.
☆ This review is part of the Advanced Drug Delivery R
⁎ Corresponding author at: Institute of Tissue Regene
550 3085.
E-mail address: kimhw@dku.edu (H.-W. Kim).
0169-409X/$ – see front matter © 2012 Elsevier B.V. Al
doi:10.1016/j.addr.2012.03.009
Please cite this article as: R.A. Pérez, et al., N
(2012), doi:10.1016/j.addr.2012.03.009
a b s t r a c t
a r t i c l e i n f o
Article history:
Received 15 January 2012
Accepted 7 March 2012
Available online xxxx
Keywords:
Smart biomaterials
Composites
Tissue regeneration
Biomimetic approach
Biofactors delivery
Multifunctional
Stimuli-responsive
The development of smart biomaterials for tissue regeneration has become the focus of intense research in-
terest. More opportunities are available by the composite approach of combining the biomaterials in the form
of biopolymers and/or bioceramics either synthetic or natural. Strategies to provide smart capabilities to the
composite biomaterials primarily seek to achieve matrices that are instructive/inductive to cells, or that
stimulate/trigger target cell responses that are crucial in the tissue regeneration processes. Here, we review
in-depth, recent developments concerning smart composite biomaterials available for delivery systems of
biofactors and cells and scaffolding matrices in tissue engineering. Smart composite designs are possible by
modulating the bulk and surface properties that mimic the native tissues, either in chemical (extracellular
matrix molecules) or in physical properties (e.g. stiffness), or by introducing external therapeutic molecules
(drugs, proteins and genes) within the structure in a way that allows sustainable and controllable delivery,
even time-dependent and sequential delivery of multiple biofactors. Responsiveness to internal or external
stimuli, including pH, temperature, ionic strength, and magnetism, is another promising means to improve
the multifunctionality in smart scaffolds with on-demand delivery potential. These approaches will provide
the next-generation platforms for designing three-dimensional matrices and delivery systems for tissue
regenerative applications.
© 2012 Elsevier B.V. All rights reserved.
Contents
1. What is smart composite biomaterial? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
2. Internal modulation for smart composite biomaterials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
2.1. Tuning physical properties to native tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
2.2. Biomimetic nanotechnology in smart composite chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
2.3. Biomimetic surface tailoring with extracellular matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
3. External modulation for smart composite biomaterials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
3.1. Scaffolds and matrices with delivering potential of biofactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
3.2. Delivery of multiple biofactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
or; bFGF, basic fibroblast growth factor; BMP, bone morphogenetic protein; BSA, bovine serum albumin; CaP, calcium phos-
nanotube; CPC, calcium phosphate cement; DGEA, Asp–Gly–Glu–Ala (amino acid sequence); DNA, deoxyribonucleic acid;
ermal growth factor; FN, fibronectin; GDNF, glial cell line-derived neurotrophic factor; GAG, glycosaminoglycan; G-CSF,
; GFP, green fluorescent protein; HA, hydroxyapatite; HGF, hepatocyte growth factor; HEMA, 2-hydroxyethyl methacrylate;
umbilical vein endothelial cell; IGF, insulin-like growth factor; IKVAV, Ile–Lys–Val–Ala–Val (amino acid sequence); IPTS,
al stem cells; MSN, mesoporous silica nanoparticles; NGF, nerve growth factor; MNP, magnetic nanoparticle; NT3, neutro-
e amphiphile; PAA, poly(acryl amide); PCL, poly(ε-caprolactone); PDGF, platelet derived growth factor; PDMS, polydi-
thylenimine; PEO, poly(ethylene oxide); PET, poly(ethylene terephthalate); PGMMA, poly(glycerolmonomethacrylate);
SRN, Pro–His–Ser–Arg–Asn (amino acid sequence); PLA, poly(lactic acid); PLGA, poly(lactic-co-glycolic acid); PLLA, poly
AAm, poly(N-isopropyl acrylamide); PPG, poly(propylene glycol); PPO, poly(propylene oxide); PTX, paclitaxel; PVA, poly
ce); RNA, ribonucleic acid; siRNA, small interfering RNA; TCP, tricalcium phosphate; TGF, transforming growth factor;
eviews theme issue on "Bionics - nature-inspired smart materials".
ration Engineering (ITREN), Dankook University, Cheonan 330-714, Republic of Korea. Tel.: +82 41 550 3081; fax: +82 41
l rights reserved.
aturally and synthetic smart composite biomaterials for tissue regeneration, Adv. Drug Deliv. Rev.
http://dx.doi.org/10.1016/j.addr.2012.03.009
mailto:kimhw@dku.edu
http://dx.doi.org/10.1016/j.addr.2012.03.009
http://www.sciencedirect.com/science/journal/0169409X
http://dx.doi.org/10.1016/j.addr.2012.03.009
2 R.A. Pérez et al. / Advanced Drug Delivery Reviews xxx (2012) xxx–xxx
4. Stimuli-responsiveness for multifunctional smart composite biomaterials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
4.1. Stimuli responsiveness and drug delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
4.2. Multiple-stimuli responsive smart biomaterials and delivery systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
5. Future perspectives and conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
1. What is smart composite biomaterial?
With respect to biomaterials, the term ‘smart’ refers to the nature
of the interactions between a given biomaterial and the surrounding
cells and tissues. Attributes of smart biomaterials include their in-
structive/inductive or triggering/stimulating effects of the cells and
tissues. These types of biomaterials, in contrast to the surgical appara-
tuses, sensors, and devices, which do not directly contact the host
cells/tissues, canintimately associate with various host constituents
for relatively prolonged periods, providing chemical and physical
cues to the cells. The consequent biological actions can be important
in driving tissue repair and regeneration. Thus, for biomaterials,
‘smart’ encompasses aspects that are relevant to the surrounding
cells and tissues. The instructive and stimulating cues provided by
biomaterials are, in fact, the essential considerations in biomaterial
studies, designed to bolster drug delivering capacity and/or tissue en-
gineering efficacy. Therefore, significant effort that seeks to modulate
the surface and bulk properties or to provide external therapeutic
molecules have been made in biomaterials such as drug delivery sys-
tems and three-dimensional (3D) scaffolds.
Internal modulation mainly involves biomimetic modification of
chemical, physical, and/or biological properties in ways that mimic
the native tissue. If successful, the strategy can provide environmen-
tal cues to the surrounding cells that aid the cells in the recognition
of the biomaterial surface and matrix, enhancing target functions
such as adhesion, proliferation, migration, and tissue differentiation.
Strategies for internal modulation include surface tailoring with bio-
mimetic molecules such as peptides and adhesive proteins, tuning
of the bulk chemistry to produce similarity to native extracellular ma-
trices (ECMs), and adjusting physical properties (such as stiffness) to
natural tissues, are possible strategies. Although this surface modula-
tion can, in a sense, be considered as an external modulation, we
categorized this as an important internal way of engineering bioma-
terials that ultimately have smart actions, fitting the biomimetic
approach by chemistry modulation.
Externalmodulation is typically germane to smart biomaterials with
drug delivering potential. The external factors that are introduced to
provide triggering and instructive cues to biologics involve chemical
drugs, proteins, and nucleic acids. Typically, those biofactors are incor-
porated in many ways within biomaterials engineered in the form of
porous foam scaffolds, hydrogels, fibers, and particulates. The mecha-
nism of action typically involves membrane receptors or is through in-
tracellular uptake within the cytosol or nucleus. Therefore, design of
biomaterials with drug delivering functionsmust consider the biofactor
of interest, site and type of action, and the location. An essential aspect
of external modulation that is closely related with biomaterial design is
themultiple delivery of biofactors, since this is crucial in scaffold design
for tissue engineering. Furthermore, more controllable actions of the
biofactors can be realized by delivering them in a manner that is re-
sponsive to external or internal stimuli. This responsiveness, in con-
junction with shape memory properties, are being explored in the
design of promising smart biomaterials that have multifunctionality,
and can be categorized in the combinatorial modulation of internal
chemistry and external therapeutic molecules.
In this review, we describe biomaterials with smart properties, par-
ticularly in the form of composites. In this context, composite refers to
materials combined either between different polymers in natural or
Please cite this article as: R.A. Pérez, et al., Naturally and synthetic smart
(2012), doi:10.1016/j.addr.2012.03.009
synthetic and/or those polymers with inorganics, and even sometimes
to those combinatorial biomaterials with biological factors designed
for such smart functions. Themain target tissues in utilizing smart com-
posite biomaterials can range from hard tissues, including bone and
teeth, to soft tissues that require specific composite formulations. This
composite approach is particularly promising and exciting, since the
combined properties between materials largely overcome the limita-
tions that come from a single phasematerial. This helps the biomaterial
meet the requirements for the smart actions, such as physical and
chemical mimetism to native tissues, controlled and sustainable deliv-
ery of biofactors including multiple delivery, stimuli-responsive deliv-
ery, and shape memory multifunctional effects.
2. Internal modulation for smart composite biomaterials
Native tissue ECMs comprise highly specialized composite struc-
ture at the molecular level, organized with macromolecules of differ-
ing chemistry and/or inorganic nanocrystallites. They retain certain
physical and chemical properties required for tissue function. Exploit-
ing biomaterials that are reminiscent of the ECM of the tissue of inter-
est is always challenging. The resulting material will not be identical
to the native ECM; thus, the properties will not exactly mirror those
of the native system, yet, great strides have been made in biomaterial
design, particularly due to developments in nanotechnology. Engi-
neering synthetic biomimetic biomaterials has benefited from com-
bining different types of materials such as natural/natural polymers,
natural/synthetic polymers, and polymers/inorganics. This approach
aims to create biomaterials that possessing physical and chemical
properties equivalent to native tissues, which are hopefully reflected
in the capability to perform relevant biological functions.
Recently, the physical cues of biomaterials, mainly physical stiff-
ness, have gained great attention in dominating cellular responses
and controlling fate. The concept is based on the material mimetism
to native ECM elastic properties. Principally, physical properties such
as elastic modulus and strength intimately reflect the different chem-
ical compositions and their organization at the nano- or micro-scale.
In this regard, the modulation of chemistry and organization is of par-
ticular importance. Awide range of compositional variables is possible
by the composite approach that is tunable to the physical properties of
the native tissues. Tailoring of physical properties is explained based
on polymer–polymer or polymer–inorganic compositions targeting
soft and hard tissue, respectively. Also, nanotechnological advances
in organizing/structuring biomaterial composites have focused on
self-assembled biomaterials with a smart functional moiety in the
nanostructure as an approach to produce instructive smart compos-
ites. Another category in the internal modulation to biomimetism is
surface engineering with ECM molecules, involving the full sequence
recombinant proteins or engineered peptides (short or oligopeptides)
with key domains, or in a fused form that has multiple actions. Strate-
gies for the internal modulation of biomaterials that exploit smart
composites with bulk or surface chemistry and physical traits that
are mimetic to native tissues are schematically illustrated in Fig. 1.
2.1. Tuning physical properties to native tissues
Physical properties including stiffness (elastic modulus) reflect
intimately the chemical composition and the organization of the
composite biomaterials for tissue regeneration, Adv. Drug Deliv. Rev.
http://dx.doi.org/10.1016/j.addr.2012.03.009
Fig. 1. Schematic illustration of internal modulation of biomaterials. The approach exploits smart composites with bulk or surface chemistry, and physical traits that are mimetic to
native tissues. The goal is to provide environmental instructive cues and matrix conditions to stimulate and trigger proteins and cells in their recognition to biomaterials and taking
subsequent biological actions. This can be achieved by (a) tuning elastic properties to trigger cell fate, by a composite approach of using ‘polymer A’ and ‘polymer B’, and further
cross-link of the chemical structure, (b) nanotechnology-assisted self-assembly of peptide amphiphiles into a nanofibrous structure with instructive cues (cell adhesive RGD
motif), and (c) biomimetic surface tailoring with extracellular molecules, including RGD peptide, fibronectin-osteocalcin (FN-OCN), or collagen.
3R.A. Pérez et al. / Advanced DrugDelivery Reviews xxx (2012) xxx–xxx
constituents at the molecular level. Some other physical properties
such as electrical and thermal conductivity, and viscous properties
in the case of visco-elastic polymers, are not covered in this review.
Among the physical properties, recent interest has been given to
the elastic property of substrates and matrices that are stimulating
and instructive to cellular responses, including cell proliferation, mo-
tility, and lineage differentiation, which primarily influences the cell
fate. The stemness of cells, which is vital in the differentiation into
the specific cell type, is dependent on the substrate physical stiffness;
a substrate with physical stiffness engineered to echo that of the na-
tive tissue ECM is important. This indicates that cells favor a substrate
with a physical elastic property that mimics the native ECM. In fact,
this elastic match has long been considered a crucial guideline in de-
signing bone implants. Elastic mismatching in vivo can produce signif-
icant problems in interfacial biomechanical properties, primarily due
to the load concentration onto the high modulus materials, shielding
effective load distribution and resulting in premature failure.
The elastic property of a material can be represented as the pa-
rameter of elastic (Young's) modulus, an index of stiffness of a mate-
rial, and expressed as simple relationship σ=Eε, where σ is stress, ε
is strain and E is elastic modulus [1]. When a biomaterial is labile
and capable of movement in response to forces exerted by cells or ex-
ternally, it is less stiff or flexible, largely correlates with soft tissues,
including nerve and cartilage. On the other hand, if a biomaterial is
rigid and so less apt to deform against such environmental forces, it
is considered to have high stiffness, analogous to hard tissues like
bone and the dentin and cementum of teeth.
Depending on the physical stiffness of the biomaterial substrate,
in vitro cellular responses are modulated. This is directly related
with the term mechanostransduction, which can be defined as the
process by which cells convert their mechanical stimuli into biochem-
ical responses [2]. Anchorage-dependent cells attach to the substrate
Please cite this article as: R.A. Pérez, et al., Naturally and synthetic smart
(2012), doi:10.1016/j.addr.2012.03.009
through the transmembrane proteins (e.g. integrins). Shortly after
this attachment, structural and signaling proteins are recruited at
the intracellular space, forming a focal adhesion complex. Subse-
quently the harmonized action of the proteins provides a signal trans-
duced through the actin–myosin complexes that are found in the
cytoskeleton, which ultimately leads to a certain form and level of
mechanical reaction of the cells [2]. In soft tissues, the substrate is
not strong enough to balance the forces created in the cells and the
formation of stress fibers are not abundant. In contrast, in hard tis-
sues, the substrate is strong enough to enable cells to generate high
forces, which consequently result in the formation of many stress
fibers in the cells [2]. Ultimately, the difference in the cytoskeletal
arrangement (stress fiber distribution) is directly related to further
behavior of a cell, including motility, growth, viability and apoptosis,
as well as the fate of lineage differentiation in the case of stem cells.
Therefore, control over the physical stiffness of the substrate is con-
sidered one of the key strategies to develop instructive smart bioma-
terials that regulate a series of cellular functions.
In a single composition, tuning this elastic property is possible sim-
ply by modulating the density or length of polymer chains, either by
changing the molecular weight of polymers, degree of cross-linking,
or water content in the hydrogels. However, the level of modulation
cannot be broad due to the limited innate physical properties of the
single composition. Moreover, the choice of materials is limited.
Therefore, a more promising approach is to combine more than two
different compositions, producing composites either between poly-
mers or between polymer and inorganics. Particularly in the latter
case, securing good bonding properties between the components is
important, otherwise the inorganic phase will be a weakening factor
in the whole composite system. Biomaterials such as poly(acryl
amide) (PAA), poly(ethylene glycol) (PEG), collagen, and glycosami-
noglycans (GAGs) have recently been engineered to tailor the physical
composite biomaterials for tissue regeneration, Adv. Drug Deliv. Rev.
http://dx.doi.org/10.1016/j.addr.2012.03.009
4 R.A. Pérez et al. / Advanced Drug Delivery Reviews xxx (2012) xxx–xxx
properties [3–20]. However, this area is still in its infancy; proof-of-
concept has been made only with a simple model biomaterial and/or
under two-dimensional (2D) conditions. When considering the rapid
development of a variety of composite biomaterials and matrices for
tissue engineering, this concept should be realized in more relevant
systems that are useful for target tissues. In this sense, some of the im-
proved biological functions of cells in vitro or tissues in vivo on the
composite biomaterials should be considered, not only from the
chemical composition standpoint, but also in light of the physical
functionality such as the matched elastic properties.
PAA gels were the first to be devised and, thus, have been widely
studied in an effort to obtain different physical elasticity by changing
the amount of acryl amide or bisacryl amide. Depending on the stiffness
of the modulated gels, in a variety of cells many different behaviors
were regulated, including kidney cell locomotion and focal adhesions,
neural cell growth and interactions, rat annular cell morphology, cyto-
skeletal structure, apoptosis and the ECM regulatory gene expression,
andmesenchymal stem cell lineage differentiation [3–10]. More recent-
ly, PEG has also been used, since itsmechanical stiffness can be easily al-
tered [11–15]. Many studies elucidating the close relationship between
substrate rigidity and cells have also been carried out using different cell
types, such as osteoblast cell line MC3T3, neuronal cell line PC12, fibro-
blast cell line NIH 3T3, and primary adult human dermal fibroblasts
[11–15]. Depending on cell type, the force balance between cells and
the substrate should be different because each type of cells possesses
its own contractile force that senses the substrate stiffness differently
(i.e., different level). Moreover, in response to the substrate, the cellular
status can also change, undergoing a certain level of differentiation or
maturation, which ultimately alters the force balance in relation to the
underlying substrate; in that case, cells might also alter the physical
microenvironment by secreting extracellular matrices that are more
favorable for their behavior and relevant to tissues to which they are
conforming. Moreover, as cells proliferate to a certain level, the cell–
cell communication may also become dominant over the cell–matrix
interaction.
Experiments with PAA gels with stiffness modulated above 2 kPa
by changing the monomer concentration showed that fibroblasts
and endothelial cells developed a spreading morphology and their
actin stress fibers were highly dependent on the stiffness, which
was not readily observed in the cells after confluence. This suggests
that signaling from cadherins in cell–cell interactions should override
the cell–matrix interaction [3]. Neutrophils were observed to be in-
sensitive to the different surface stiffness varying from 2 to 55 kPa,
in terms of cell morphology, cytoskeletal structure, and adhesion
[3]. While neural cells have generally been shown to favor a much
softer substrate compared to other tissue types, including cartilage,
muscle, and bone, it has also been shown that PC12 cells on the
very soft substrates of polyacrylamide gel tailored with stiffness
below 0.1 kPa sharply decrease neuriteoutgrowth and branching,
demonstrating there is some threshold of substrate rigidity that is fa-
vorable for this type of neural cell behavior [4]. On the other hand,
cortical neuron outgrowth was observed to be insensitive to substrate
stiffness modulated over the range of 0.26–13 kPa [5]. When neural
stem/progenitor cells derived from the forebrain of adult female Wis-
tar rats were cultured on composite chitosan-PAA gels, the optimal
stiffness values for differentiation into neuron (b1 kPa) and prolifer-
ation (3.5 kPa) were different. This suggests that the substrate that
the cells require is highly dependent on the status of the cells [6]. In
the case of bone-associated cells, the biomaterial substrates with
stiffness over 100 kPa, which is close to the stiffness of natural bone,
usually favorably dictate cellular responses, including initial cell
adhesion and osteogenesis [7,12]. When preosteoblast MC3T3 cells
were cultured on the PEG-based composite gels, made with PEG-
diacrylate mixed with different amounts of nonacrylated PEG, the
osteogenic differentiation was promoted more highly on the stiffer
matrices [11]. If the starting cell types had more stemness character,
Please cite this article as: R.A. Pérez, et al., Naturally and synthetic smart
(2012), doi:10.1016/j.addr.2012.03.009
like mesenchymal stem cells (MSCs), the final lineage was shown to
be more subtly regulated by the substrate rigidity and the differenti-
ation lineage was potentially predictable by the underlying substrate.
MSCs cultured on PAA gels with different stiffness can undergo differ-
ent lineage differentiation: neurogenesis on soft tissues, myogenesis
on intermediate stiffness, and osteogenesis on stiff matrices [8]. Sim-
ilar results were observed for neural stem cells, with an optimum
value of stiffness (e.g. 0.5 kPa) leading to increased production of
neuronal markers [9]. Cellular apoptosis has been shown to be highly
dependent on substrate physical stiffness. Rat annulus fibrosus cells
were mechanosensitive when cultured on different PAA matrices
with different stiffness ranging from 1 to 63 kPa. They showed round-
ed morphology and increased apoptosis on soft substrates, while
higher metalloproteinase levels were evident as early as 24 h on
stiff substrates (Fig. 2) [10].
While stiffness can be modulated by changing the synthetic poly-
mer composition, the incorporation of ECM molecules within the
structure of synthetic polymers also affects stiffness. Adhesive ligands
incorporated within PEG gels influence the physical properties of the
gels [16]. When fibrinogen was combined with PEG, the gel stiffness
was affected, leading to phenotypic plasticity of smooth muscle cells
[17,18]. Incorporation of natural polymers within synthetic polymeric
gels to make composites has also been tried in an effort to modulate
cellular behaviors by the alteration of physical properties of the gels.
Collagen addition to PEG during the cross-linking process also signifi-
cantly altered the mechanical stiffness of the gels, leading to positive
effects on neural cell responses [13,19]. Gelatin has also been intro-
duced within a new type of synthetic polymer gel, hydroxyphenyl
propionic acid, via enzyme-mediated cross-linking. Depending on
the composite composition, the stiffness was effectively controlled,
which led to stiffness-dependent behaviors of human MSCs; highly
proliferative on stiffer gels with better spreading, more organized cy-
toskeletons and stable focal adhesion and migration rate, exhibiting
neurogenic behaviors on gels with stiffness over 0.6–2.5 kPa, while
myogenic on the gels with stiffness higher than 8 kPa [21]. A combina-
tion of poly(L-lysine) and hyaluronic acid was used to control the stiff-
ness of substrates; stiff substrates promoted the formation of focal
adhesions and enhanced proliferation, whereas soft matrices were
not favorable for anchoring or proliferating skeletal muscle cells [22].
Efforts to modulate the stiffness have also beenmade between natural
polymers, by combining different compositions and changing the de-
gree of cross-linking. Collagen and GAGs cross-linked with different
types of agents were observed to have different stiffness while main-
taining nominally the same chemical composition. Softer matrices
showed higher cellmediated contraction, presenting higher osteogen-
ic maturation, whereas stiffer ones presented less differentiation but
higher cell number [10,20]. Silk-elastin composite biomaterials were
also developed to have a range of elasticity that influencing the
myogenic/osteogenic stimulation of C2C12 and hMSCs cells [23].
When Arg–Gly–Asp (RGD) adhesive ligand was combined with algi-
nate gel via photocrosslinking, the matrix stiffness increased, leading
to increased angiogenic potential of adipose progenitor cells while
inhibiting adipose differentiation [24].
As demonstrated above, the internal composition changes poten-
tially modulated the physical elastic property, which in turn pro-
foundly influenced the initial spreading, growth, and migration
behavior of cells, dictating the cell differentiation into tissue specific
lineages, and even determining cell death. This is considered a power-
ful tool in developing biomaterials that mimic the ECM physical con-
ditions and, thus, provide instructive and smart matrices for cells in
their regenerative processes. While the modulation of physical stiff-
ness has been possible, mainly in the polymeric gels like PAA and
PEG, by altering the network structure of the matrices like cross-
linking density and polymerization, more recent studies have devel-
oped new composites incorporating natural proteins and ECM mole-
cules that have more appropriate range of stiffness values and
composite biomaterials for tissue regeneration, Adv. Drug Deliv. Rev.
http://dx.doi.org/10.1016/j.addr.2012.03.009
Fig. 2. Effects of substrate stiffness on cell morphology. Phase-contrast photomicrographs of rat annulus fibrosus cells cultured on soft (a), intermediate (b), or rigid (c) substrates
for 24 h. Cells on soft (d), intermediate (e), or rigid (f) substrates for 1, 3, 6, 12, and 24 h were fixed with 4% paraformaldehyde and their F-actin was stained with phalloidin-
fluorescein isothiocyanate. The isolated cells appeared to have no stress fibers (d). Apparent stress fibers formed when cells were cultured on rigid substrates for 24 h (f).
Representative images of cells spread on different substrates showed that cells on soft substrates (d) spread very little compared with the extent of spreading on stiff substrates
(f). Original magnification ×160. Adapted and reprinted with permission from Ref. [10].
5R.A. Pérez et al. / Advanced Drug Delivery Reviews xxx (2012) xxx–xxx
better biocompatibility. This concept of physical elasticity as a deter-
minant of cellular behavior should be borne in mind when developing
smart biomaterials for regenerative purposes and in the understand-
ing of the cellular phenomena occurring on the biomaterials, aside
from chemical or biochemical cues.
2.2. Biomimetic nanotechnology in smart composite chemistry
A more elegant tailoring of the internal chemistry of biomaterials
has become possible by means of nanotechnology. Nanotechnological
advances have spurred the creation of composite biomaterials with
better properties and functions, enabling control over the processes
that produce a composite that echoes native ECM organization and
structure. Self-assembled biomaterials with cell-instructive cues are
a promising avenue of smart composite biomaterials.
Self-assembly is defined as the spontaneous free energy driven
association of several individual entities under thermodynamic equi-
librium, to form a well-organized and well-defined structure to max-
imize the benefit of the individual without external instruction [25]. It
is based on the interaction of the different individual molecules
through weak bonds, such as electrostatic forces, van der Waals inter-
actions, or hydrogen bonds [25]. The resultof the interaction is a self-
associated hierarchical structure. The key point in nanotechnology
driven self-assembly is the design of peptides with specific structure
that allows the formation of these hierarchical structures when im-
mersed in a liquid by the aforementioned weak bonds. Many types
of self-assembled structures can be obtained, which include nanofi-
bers, nanotubes, and nanovesicles. These forms are mainly dependent
on the chemistry of building units/blocks that are composed of one or
Please cite this article as: R.A. Pérez, et al., Naturally and synthetic smart
(2012), doi:10.1016/j.addr.2012.03.009
more different peptides [25,26]. In particular, two types have been
utilized as biomaterials: one is self-assembled from peptides and
the other is from synthetic hybrid block polymers.
Peptide amphiphile (PA) is an attractive building unit that is com-
prised of peptides that can self-assemble into 3D nanofibers. PA basi-
cally combines a hydrophobic tail, a beta sheet-forming segment
(although it can also be alpha or random coil), a charged group re-
gion, and finally the bioactive epitope that can be varied to target a
specific function [27]. While the hydrophobic segment is oriented to-
wards the inside, the hydrophilic region is towards the outside, where
the bioactive target functional motifs can also be put together, which
consequently assemble to form a long fibrous structure. Therefore,
the design of bioactive functional motifs is of special importance in
the regulation of cellular functions. The most common epitope used
is the RGD sequence to allow favorable cell adhesion [27,28]. A
laminin-derived epitope Ile–Lys–Val–Ala–Val (IKVAV) sequence was
also introduced for improving neural cell functions [27,29]. While
this type of PA self-assembled nanofiber is, in itself, an effective scaf-
fold in the form of hydrogels that contain cells inside, providing 3D
substrates for them to adhere to and proliferate, the introduced bio-
active functional epitopes within the structure are primarily the key
aspect of the smart and cell-triggering nature of the PA biomaterials.
Thus, many potential opportunities are open to utilize the PA-based
biomaterials in the repair and regeneration of various types of tissues
and other applications. Among else, injectable hydrogels have shown
significant value as a tissue regenerative material, in applications
such as bone and nerve regeneration and angiogenesis [27]. When
the PA containing epitope IKVAV derived from laminin was investi-
gated for neural differentiation, the epitope peptide effectively
composite biomaterials for tissue regeneration, Adv. Drug Deliv. Rev.
http://dx.doi.org/10.1016/j.addr.2012.03.009
6 R.A. Pérez et al. / Advanced Drug Delivery Reviews xxx (2012) xxx–xxx
stimulated longer neurite outgrowths in a 2D culture system, largely
suppressing the formation of astrocytes [30]. In vivo results in a
mouse spinal cord injury model showed the use of PA reduced cell
death at the injury site and astrogliosis, accompanied by enhanced
limb functionality [31]. Angiogenesis was stimulated by incorporating
PA that was designed to have a heparin binding domain and, conse-
quently, have an affinity to growth factors like fibroblast growth
factor-2 (FGF-2), bone morphogenetic protein-2 (BMP-2), and vas-
cular endothelial growth factor (VEGF), demonstrating significant
stimulation of in vivoblood vessel formation [32]. To stimulate hydroxy-
apatite (HA) bonemineral formation targeting hard tissues, the epitope
was designed to incorporate a specific amino acid, phosphoserine,
which is able to start biomineralization due to the presence of acidic
moieties. Formation of the HAmineral phase on the PAwhen immersed
in calcium and phosphate solutions was reported, and the alignment of
the HA crystal along c-axis was preferentially parallel to the PA fiber
long axis [33]. When phosphorylated PA was used, mineralization of
HA crystals was also greatly enhanced. The HA composition and struc-
ture was similar to those of native bone with controlled and spatial
selective deposition of HA in a 3D environment (Fig. 3(a, b)) [34]. It is
considered that the incorporation of amino acids or functional groups
that are highly negatively charged (carboxylated, sulfated, or phosphor-
ylated) within the epitopes should favor the mineralization of HA
Fig. 3. Some representative nanotechnological approaches to self-assemble into smart biom
bone mineralization process that may have potential applications as scaffolds for bone rep
self-assembly process into a nanofiber with approximately 5–7 nm in diameter (a). A phosp
alization of HA (based on TEM analyses, in panel b). Adapted and reproduced with permissi
was genetically-engineered to contain specific functional groups (c). Representative examp
oriented nucleation of the HA within the nanofibrous phage in a manner with the c-axis par
with permission from Refs. [51] and [55].
Please cite this article as: R.A. Pérez, et al., Naturally and synthetic smart
(2012), doi:10.1016/j.addr.2012.03.009
crystals, providing potential uses such as biofunctional smart bone
matrices. Apart from being used as scaffolds or matrices, when the PA
was used to tailor the surface of biomaterials, such as coating for
shape-memory alloys in stents and joints repair, its biological functions
such as improvement of cell adhesion was also well demonstrated if
these PA presented the RGD sequence in the epitope [35–38].
Peptides designed to have specific biological functions are often in-
troduced within the synthetic polymeric compositions to produce self-
assembled peptide/synthetic polymer composite biomaterials. In one
case, biological molecules are primarily combined with the synthetic
polymers through covalent bonds. Block copolymers are often designed
to self-assemble into an ordered structure. Within the composition,
peptide sequences are combined to form hybrid block copolymers.
Many of them combine β-sheet forming peptides such as β-amyloid
mimics, silk mimics, fibrin mimics, and elastin mimics [39–41]. In the
case of silk, the hybridization of silk-like β-sheet polypeptides with
PEG to form triblock copolymers increases the silk-mediated beneficial
effects on mechanical properties [40]. PEG, due to its structure and hy-
drophilicity, is widely used to combine with coil forming structures to
form the self-assembled structures [42], demonstrating the ability to
bind drugs and genes for tissue specific delivery [43,44]. Furthermore,
certain genetically engineered protein domains have been incorporated
into the polymer structures to produce genetically modified smart
aterials. Self-assembly building units are designed to favor tissue cells and mimic the
air and regeneration. (a,b) Peptide amphipile (PA) self-assembly nanotechnology: PA
horylated group introduced within the epitope of PA was effective in inducing miner-
on from Ref. 34. (c, d) Bacteria phage nanotechnology: A stable building block of phage
le showing a self-assembly of the building units into β-sheet bundles, which allows the
allel to the bundles (based on TEM analyses, as in panel (d)). Adapted and reproduced
composite biomaterials for tissue regeneration, Adv. Drug Deliv. Rev.
http://dx.doi.org/10.1016/j.addr.2012.03.009
7R.A. Pérez et al. / Advanced Drug Delivery Reviews xxx (2012) xxx–xxx
biomaterials for drug delivery and tissue engineering [45,46]. A recent
report, based on smart hybrid copolymers, described an example
work targeting bone tissue. Poly[n-(2-hydroxypropyl)methacrylamide]
(PHPMA) and β-sheet peptide were graft copolymerized to self-
assemble into hydrogels [47]. Complementary β-sheet of the peptides
(TTRFFWTFTTT and TTEFTWTFETT) were shown to play a role in nucle-
ation of theHAmineral phase, demonstrating thepotential of thehybrid
composite hydrogel as useful bone tissue engineering scaffolds. While
the nucleation on specific sites was due to the presence of certain pep-
tides, the anisotropic pore morphology in the scaffold was interestingly
dominanton the mineralization morphology of HA. As demonstrated
above, although the synthetic copolymers have a self-assembling char-
acter to form 3D structures they may have weaknesses in the biological
functions in the stimulating and instructive activities to target cells and
tissues. Therefore, the hybridization of smart biofunctional peptides
specifically designed for target functions within the polymer structure
is a promising strategy to exploit smart composite 3D scaffolds and
matrices for tissue regeneration.
Another nanotechnological development of smart and instructive
self-assembled biomaterials is bacteria phage technology, which has
recently provided an interesting and promising nanostructured plat-
form for tissue regeneration. It is based on the use of viruses as tem-
plates (building blocks or units) to produce materials, since they are
able to self-assemble into an ordered structure [48]. Moreover, they
display functional peptides on the materials that have been genetical-
ly engineered to have a biological activity. Only a few studies have yet
explored phage technology for biomaterial and tissue engineering
applications. The M13 phage has been most commonly used, mixed
with polyvinypyrrolidone (PVP) and electrospun to produce micro
and nanofibers, demonstrating the exploitation of genetically engi-
neered functionality into mechanically robust virus fibers [49,50]. For
the specific targeting of neural regeneration, phages were genetically
engineered to display RGD and IKAV peptides in the building unit, facil-
itating both cell adhesion and neurite outgrowth. The phages were
mixed with neuronal progenitor cells within liquid agarose, which
was allowed to become a gel to generate neuro-functional nanostruc-
tured biomaterials [51]. In another exemplar study, phages with RGD-
engineered sequences in the building block were patterned on a cover
slip via silane treatment, which allowed directional growth and encap-
sulation of fibroblasts [52]. Engineering phage structure has also been
targeted for mineralization of bone crystal HA. Different peptides that
are responsible for the nucleation of HA were introduced in the phage,
which was then self-assembled into a nanofiber. The phages were
able to self-assemble into β-structure bundles between the peptides
displayed on the side walls. The structure allowed the oriented nucle-
ation of the HA within the nanofibrous phage in a manner with the
c-axis parallel to the bundles (Fig. 3(c, d) [53–55]. This new type of
self-assembled nanostructured biomaterials may be useful as scaffolds
for the regeneration of hard tissue including bone and teeth. Because
of the robust structural stability of bacteria, utilizing them as the build-
ing units with diverse peptide functionality should facilitate full use of a
natural design to exploit biomimetic smart biomaterials.
The possibility of designing peptides that comprise the building
units, in such a way that they are biologically functional and instruc-
tive for directing cellular responses in repair and regenerative
processes, makes self-assembled nanotechnology a promising tool
to the internal modulation of chemistry for smart composites. As in
the form of either amphiphile or bacteria phage as the building blocks
that contain a combination of functional peptides, self-assembly into
a 3D material form needs more research to find promising scaffolding
platforms for tissue regeneration.
2.3. Biomimetic surface tailoring with extracellular matrices
While biomimetism is possible by engineering the entire structure
of biomaterials to compositionally mimic native ECMs, it is sometimes
Please cite this article as: R.A. Pérez, et al., Naturally and synthetic smart
(2012), doi:10.1016/j.addr.2012.03.009
better to tailor the surfaces that directly interface with the biological
environments. This region—the biointerface—is the key place in
whichmost of the initial series of events occurs between the biomate-
rial and the host and involves dissociation of surface molecules/ions,
protein interactions, and cell anchorage and adhesion. Therefore,
biointerface control with native tissue ECMs is by far one of the most
effective ways of providing scaffolds and implants with smart cell in-
structive functions. Because biomaterial surface engineering with
ECM-mimicking proteins has been intensively studied over the last
decade [56–60], here we review the most recent work on the surface
engineering of biomaterials and scaffolds with protein molecules
and designer equivalents, such as peptides that mimic the ECMs of
target tissues and/or confer benefits to their biological functions, ulti-
mately providing biomimetic smart surface conditions to cells.
The first and most widely used surface-tethering moieties are the
proteins that comprise the ECM components of native tissues. Among
these, adhesive ligands have proven to be one of the most attractive
proteins to tailor the surface of synthetic biomaterials, controlling
the first essential step of cellular recognition of the biomaterials.
Because most synthetic polymers including poly(α-hydroxyl acids)
are highly hydrophobic, which hampers biological events of adhesive
protein binding, the presence of adhesive ligands should significantly
increase the rate of cellular interactions. The surface of synthetic bioma-
terial scaffolds with various forms, including porous foams, nanofibers,
and microspheres, have been tethered with adhesive proteins, includ-
ing collagen, fibronectin, and laminin. Due to the many attractive fea-
tures, electrospun nanofibers of synthetic biopolymers including poly
(lactic acid) (PLA) and poly(ε-caprolactone) (PCL) have recently been
modified with collagen or fibronectin. Significant improvement of
hydrophility and cell attachment and proliferation has been reported
[61,62]. Mainly for neural cells, laminin is an important ECM protein
that mimics the nerve basement membrane. Laminin-tethered PCL
nanofiber scaffolds enhance peripheral nerve regeneration [63]. The
tethering strategy of those proteins onto the nanofibrous surface is pri-
marily through a covalent-linkage by the activation of the biopolymer
surface. This is achieved by treatment using alkaline solutions, plasma,
or radiation to reveal active groups that facilitate the immobilization
of the adhesive molecules [64–67]. Moreover, many detailed processes
that have been confirmed on biopolymer films were applied to the
nanostructured (nanofibrous) or complex-shaped 3D scaffolds [62,63,
68,69]. While those adhesive proteins have mainly been applied to the
surface of synthetic biopolymers, natural polymers like chitosan and al-
ginate, which have poor cell binding affinity and are largely devoid of
adhesive ligands, benefit also from this surface tailoring. For natural
polymers, the covalent linkage with adhesive proteins is much easier
and direct as they have innate chemical functional groups that allow
covalent bonds [70]. Immobilization of collagen or fibronectin onto
chitosan scaffolds significantly improves the adhesion process of
many types of cells including chondrocytes and MSCs [71–73].
Apart from the adhesive proteins, some other compositions of
ECM molecules, including GAGs, have also been used to improve the
surface properties of biomaterials. This tends to be more effective
for synthetic biopolymers. Poly(dimethylsiloxane) (PDMS) surface
treated with hyaluronic acid and collagen, which was used in neural
interfacing areas, displayed increased hydrophilicity, cell growth,
and neural differentiation [74]. The surface of PCL scaffolds functiona-
lized with chondroitin sulfate in conjunction with collagen also
showed improved chondrocyte adhesion and proliferation [75]. Hep-
arin is often used to tailor biomaterial surfaces to allow biological
functionality, including the capturing of growth factors like FGFs
[76], as these bind specifically to heparin for a sustained period.
Growth factors are mainly considered to be incorporated within the
internal structure of biomaterials,allowing a more effective and con-
trolled release than when those are tethered on the surface, even
though some studies have also reported the surface functionalization
of biomaterials using growth factors, such as BMPs, NGFs, VEGFs, and
composite biomaterials for tissue regeneration, Adv. Drug Deliv. Rev.
http://dx.doi.org/10.1016/j.addr.2012.03.009
8 R.A. Pérez et al. / Advanced Drug Delivery Reviews xxx (2012) xxx–xxx
FGFs [77–79]. As shown with heparin, GAG functionalization of
biomaterial surfaces has additional benefits of holding or capturing
growth factors as well as binding tightly with ECM proteins like colla-
gen. A chondroitin sulfate functionalized polypyrrole conduction
polymer was shown to be effective in subsequent immobilization of
collagen molecules, which consequently enhanced neurite outgrowth
in neural cells [80].
Inorganic biomaterials and the polymer/inorganic composites
have also been shown to achieve improved biological functionality
through the immobilization of ECM molecules [81,82]. The surface of
silane-based inorganic biomaterials was functionalized with fibronec-
tin to improve the initial adhesion of bone cells [83,84]. Composites of
synthetic biopolymers with bone mineral HA nanoparticles were also
functionalized with ECM molecules, such as collagen, aiming to im-
prove cell adhesion process, which is beneficial for securing a large
population of cells and further rapid differentiation and mineraliza-
tion [3]. Compared to the polymer surface, the HA mineral phase
present on the surface is more favorable for the adsorption of proteins
such as fibronectin and bone-related proteins [85–87]. Therefore, pro-
tein tethering on such surfaces is mainly achieved by affinity binding
such as ionic bonds utilizing the charged amino acid sequences of
the protein and the calcium and/or phosphate groups in HA [85–87].
Specific binding between the group of γ-carboxy glutamic acid
sequences of osteocalcin (OCN), a key noncollageneous bone ECM
protein, and a group of calcium ions present in the HA crystal lattice
structure is one representative example [88]. This protein recognition
to themineral surface is very beneficial for utilizing in biomaterials for
bone regeneration. Not only the HA-based scaffolds, but also the
biopolymer scaffolds treated with HA mineral, tethering with OCN is
possible with highly specific and avid binding. It also preserves the
biological activity of the protein in a manner that is safer and more
effective compared to the general covalent bindings, which can deter
the conformation change in native proteins. Future studies are
expected to focus on the protein tailoring of HA mineral surfaces for
bone regeneration.
While the entire native structure of the components present in
ECM is favored in terms of preserving the biological functions, their
utilization with biomaterials requires special consideration. This is
due to the structural conformation and unfolding that are associated
with the processing conditions to be undertaken with biomaterials,
including solvents, temperature and pH, which differ from the biolog-
ical conditions. Since the domains of native proteins that are relevant
to biological effectiveness are well-recognized, the engineering of
those key functional domains in the form of short peptides or oligo-
peptides is an attractive strategy in the biomimetic tailoring of the
surface of biomaterials. Furthermore, the use of short peptides avoids
the inherent disadvantage of the proteins that have to be purified,
which may involve an immune response and/or infection risk. During
the tethering processes of the protein on the biomaterials surface,
only some proteins present the proper orientation for cell adhesion
to take place [89].
As the key domain of adhesive proteins, the RGD sequence has
shown considerable in vitro cellular functionality. But, similar benefits
are not always obtained in vivo. This is likely to be because of exten-
sive remodeling following implantation, since adsorptive proteins
can alter cell reactivity with the RGD [90]. The beneficial response is
more likely to be achieved with biomaterials that have limited biolog-
ical activity but which possess architecture and mechanical proper-
ties that favor the repair of tissues or, even better, if the tethered
RGD sequences are combined with other functional domains [90].
An extensive review on the RGD sequence and its biomaterial-
associated benefits is available [90].
Recent advances have also extended the use of RGDon the surface of
novel developed biomaterials, including polymer/inorganic composites,
alginatemicrospheres, cardiac tissue engineering alginate scaffolds, and
thermosensitive polymers [91–97]. Several notable strategies include
Please cite this article as: R.A. Pérez, et al., Naturally and synthetic smart
(2012), doi:10.1016/j.addr.2012.03.009
the utilization of the RGD sequence in conjunction with natural poly-
mers, such as collagen and hyaluronic acid and the application of gradi-
ent immobilized RGD peptide on thermally sensitive polymers to
regulate cellular attachment as well as detachment [98]. Aside from
RGD sequences, the P-15 oligopeptide has been studied extensively
with respect to bone regeneration [99,100]. P-15 is longer, relative to
RGD,whichmay be the basis of its superiority for in vivo bone formation
when tethered to surfaces [101]. Other peptides designed from the na-
tive proteins, including Asp–Gly–Glu–Ala (DGEA) from collagen,
fibronectinIII10 or Pro–His–Ser–Arg–Asn (PHSRN) from fibronectin,
and Ac-CGGASIKVAVS from laminin, exert specific biological functions
[102–107]. In particular, DGEA, which is a bioactive collagen peptide,
can be specifically linked to the surface of HA in a reaction mediated
by heptaglutamate-E7; the linkage stimulated adhesion and differenti-
ation of MSCs into osteoblasts in vitro, and increases new bone forma-
tion and bone contact around the biomaterials surface in an in vivo rat
model [107]. New peptide sequences designed to incorporate key sec-
tions of the cell binding domain of FN, including FNIII10 or PHSRN,
have been used to tailor biopolymer surfaces, resulting in improved
cell spreading and adhesion [102,103]. Polypeptides that mimic the
elastic ECMprotein, elastin, have also been engineered and immobilized
on the surface of polycarbonate urethane polymer via cross-linking. The
modified surface promoted smooth muscle cell adhesion, spreading,
and retention during culture [104]. To stimulate neural cells, a super-
porous Ac-CGGASIKVAVS peptide was designed with the laminin key
functional domain, which was cross-linked in a gradient to collagen
scaffolds; the design substantially improved neurite outgrowth of
human fetal neural precursor cells [106].
Engineering of multifunctional proteins that possess more than
two functional domains of proteins originated from different proteins
has been done with the aim of controlling multiple or sequential cel-
lular processes. This fusion protein technology has potential in the
modulation of synthetic biomaterials that more closely mimic native
ECM compositions, where a number of multiple-functioning proteins
are integrated to regulate multiple or sequential cellular processes
[60,108].
One exemplar study was to utilize the domains of two proteins
that take part in initial cellular adhesion and further bone specific dif-
ferentiation. FN with a central cell binding domain (CCBD) was fused
into the bone matrix proteins, osteopontin (OPN) or OCN, two major
nanocollageneous bone ECM proteins involved in many essential
steps in osteogenic differentiation and mineralization. A multifunc-
tional protein FN-OCN was designed to utilize the surface of HA
[109]. The presence of the OCN sequence was shown to drive the
HA-specific affinity-binding between calcium ions present in the
crystal lattice of HA and the highly negatively charged γ-glutamic
acid sequence of OCN [109]. The binding affinity to HA was superior
in the fusion protein compared to FN, demonstratingthe role of
OCN in binding to HA. The tethered FN-OCN protein was shown to
regulate initial cell adhesion, which was benefited from FN, as well
as stimulate osteoblastic differentiation at a later stage, which
resulted from the function of OCN (Fig. 4) [109]. In a similar approach,
the cell adhesion domain of FN was combined with fibroblast growth
factor, promoting cellular adhesion, proliferation, and differentiation
of osteoblastic cells [110].
Another recent fusion protein technology in bone regeneration
area has combined the excellent mechanical properties of silk with
the key bone ECM protein, bone sialoprotein, in an effort to stimulate
cell attachment and differentiation, as well as accelerating the depo-
sition of calcium phosphate. The fusion protein was shown to en-
hance the mineralization process and stimulate MSC differentiation
towards the osteogenic lineage [111]. Even the fusion of different
types of BMPs (e.g., BMP4 and BMP7) can increase differentiation of
bone marrow stem cells compared to the single type of BMPs [112].
Targeting tissues other than bone, including nerve, recent studies
have exploited to engineer multifunctional proteins and peptides.
composite biomaterials for tissue regeneration, Adv. Drug Deliv. Rev.
http://dx.doi.org/10.1016/j.addr.2012.03.009
Fig. 4. Fusion protein FN-OCN tethered on the HA biomaterial surface accelerates initial
cell adhesion and further osteoblastic functionality. (a) FN-OCN protein adhesion to HA
is highly favored by the specific recognition of OCN to the Ca ions in the HA crystal
lattice (vs. BSA or FN). (b) Consequent cell adhesion and (c) ALP osteoblastic activity
are greatly stimulated by the synergistic action of fusion protein (vs. w/o protein or
FN). MC3T3-E1 cells were used to assess the cell adhesion (at 1 h) and alkaline
phosphatase (ALP) activity (at 14 days). Illustrations referred to Ref. [109].
9R.A. Pérez et al. / Advanced Drug Delivery Reviews xxx (2012) xxx–xxx
For neural tissue regeneration, a bifunctional peptide composed of
collagen-like repetitive sequence and laminin-derived sequence,
forming AG73-G3-(PPG)5, was adsorbed onto PLA polymer surfaces
through hydrophobic interaction mediated by the PPG5 region in
the peptide [113]. The fusion protein formed an ECM-like layer
composed of the collagen structural portion and laminin-derived
signaling sequence. Enhanced neurite outgrowth of PC12 cells was
evident when grown on the fusion protein-adsorbed PLA substrate.
A similar study also introduced elastic structural domain VGVPG
and RGD cell binding domain (VGRGD) and combined them into
one fusion protein [114]. Fibroblasts and neuroblasts cultured on sur-
faces functionalized with this fusion protein showed cell adhesion
similar to that obtained on fibronectin. Furthermore, the fibroblasts
Please cite this article as: R.A. Pérez, et al., Naturally and synthetic smart
(2012), doi:10.1016/j.addr.2012.03.009
exhibited a flattened polygonal morphology, whereas the neuroblasts
synthesized new DNA and actively proliferated. One recent approach
that has garnered great interest is to combine the elastin-like pro-
teins with different types of proteins or amino acid sequences with
multifunctionality—recombinamers [115,116]. These can also be
used as matrices or scaffolds as a whole, rather than as surface tailor-
ing moieties. This approach is beyond the scope of this review. The
reader is referred to other, more relevant, reviews [116].
3. External modulation for smart composite biomaterials
Incorporating the external factors that have therapeutic functions
within the internal structure that are subsequently to be released is
an essential part of the strategy to produce smart biomaterials. This
has primarily been considered a common and facile tool to generate
3D scaffolds and matrices with therapeutic actions to stimulate and
induce cells for tissue engineering applications, and thus, is a major
part of designing contemporary drug delivering scaffolds. In this sec-
tion, we review the recent advances on scaffold and matrix systems
that have the potential of enabling delivery of biofactors that include
chemical drugs, proteins, and genes, with particular emphasis on the
strategic tools to deliver multiple biofactors.
3.1. Scaffolds and matrices with delivering potential of biofactors
While there have been many scaffolds and matrices with different
forms and compositions developed to load and deliver biofactors, the
delivery strategy should be established based on the type of mole-
cules to deliver. Biofactors can generally be grouped as chemical
drugs, proteins, or nucleic acids (genes) [117–119]. Depending on
the type, the biological action can differ; therapeutic activity is
achieved either in direct contact with cells through the cell mem-
brane receptors, or after cellular uptake requiring cytoplasmic ac-
tions, or even after penetration into nucleus (Fig. 5).
Those biofactors are primarily incorporated within the internal
structure of biomaterials during the processing routes, or are otherwise
bonded or adsorbed on the surfaces of the preformed biomaterials,
depending on the actions of the therapeutics and the target cells/
tissues.While the former ismore relevant to gain long-term therapeutic
effects in a more sustainable and time-dependent manner, the latter
mainly targets direct actions with the contact cells. The delivering bio-
materials are developed in various forms, including porous foam scaf-
folds, hydrogels, nano/microfibers, and nano/microparticulates, and
their combinations [119], and thematerials compositions are also either
biopolymers of natural or synthetic origin, bioactive inorganics, or their
composites [120–125]. Therefore, a variety of designs are useful to de-
velop a system that can effectively load and deliver target biomolecules
to the site of injury. Among all the possibilities, herewe review themost
recent advances in the delivery systems of scaffolds that are utilized as
cell supportingmatrices, and thus are engineered to incorporate biofac-
tors within the internal structure and release them in a controllable and
sustainablemanner, and even to have smart actions to trigger and stim-
ulate appropriate cellular functions.
The most common scaffolds used to incorporate biofactors are poly-
mers, and the biofactors have been incorporated either during the pro-
cess of the scaffolds or after the fabrication. Because most biofactors
require water-based solutions, natural biopolymers are preferred over
synthetic ones. Many natural polymers including collagen, gelatin, chit-
osan and GAGs, have charged functional groups and present a more or
less ionic affinity to therapeutic biomolecules such as growth factors.
Scaffolds that load the growth factors and release them over certain pe-
riods have been extensively studied [122,123,126–128]. Here we show
some of the essential examples of the recent advances of those scaffold
systems for growth factor delivery. While those composites of natural
biopolymers containing growth factors are easily implemented, growth
factors are freely released from the system through water diffusion
composite biomaterials for tissue regeneration, Adv. Drug Deliv. Rev.
http://dx.doi.org/10.1016/j.addr.2012.03.009
Fig. 5. Therapeutic action schemes of scaffolding delivery systems depending on biofactors to deliver, where scaffolds incorporating growth factor, gene-loaded nanoparticles, or
chemical drugs are releasing target biofactors which are delivered to cells. Intereaction involves via either (i) receptor-ligand bindings in the case of growth factors, or intracellular
uptake of biofactors (ii) in naked form such as chemical drugs or (iii) with the help of gene loaded nanocarriers via endocytosis. (iv) certain genes (such as siRNA) are designed to
temporarily express within the cytosol, and (v) sometimes the gene loaded nanocarriers are targeted into the nucleus through the penetration of nuclear pores.
10 R.A. Pérez et al. / Advanced Drug Delivery Reviews xxx(2012) xxx–xxx
because of the hydrogel characteristic of the polymers. Therefore, the
affinity binding approach between growth factors and biopolymer net-
works has been demonstrated. Several studies reported that heparin
functionalized to collagen allows more stable incorporation of bFGF
and NGF with higher affinity binding and facilitates their sustainable
release from the matrix, as the growth factors have a heparin binding
domain [128,129]. Likewise, chondroitin sulfate, which has a similar
network of highly negative-charged groups and is categorized as
GAGs like heparin, can improve the binding ability to BMP-2 and subse-
quent prolonged release when combined into collagen scaffolds [123].
The effectiveness of heparin in delivering growth factors was also con-
firmed in composite scaffolds made of calcium phosphate and collagen
[130]. Alongwith heparin, fibrin has high affinity to growth factors and,
thus, has been used as a delivery system. A fibrin gel incorporating
transforming growth factor beta-1 (TGFβ-1) displayed a slow release
profile and was consequently effective in chondrogenic differentiation
while suppressing osteogenic differentiation [131]. For skin regenera-
tion, epidermal growth factor (EGF) fused with the fibrin-binding
domain of fibronectin has reportedly shown higher affinity than the
EGF alone to the fibrin matrix, with the EGF-loaded fibrin promoting
the growth of fibroblasts and keratinocytes, and wound repair [132].
To prolong the release profile of the biofactors from the matrices,
the candidate molecules are often encapsulated first within micro-
spheres that release more slowly, which is then embedded within
the scaffolds or hydrogels. A system combining protein-loading poly
(lactic-co-glycolic acid) (PLGA) microspheres within collagen and
hyaluronic acid gel-like scaffolds was developed to permit tunable
and sustainable protein release kinetics [133]. For the support of neu-
ral stem cell maintenance and proliferation, a composite systemmade
of hyaluronic acid hydrogel that incorporates PLGAmicrosphere load-
ed with brain-derived neurotrophic factor (BDNF) and VEGF was de-
veloped. The composite appears to be a promising scaffold that
provides an ECM mimicking niche for stem cells and creates a
Please cite this article as: R.A. Pérez, et al., Naturally and synthetic smart
(2012), doi:10.1016/j.addr.2012.03.009
permissive microenvironment for angiogenesis and neural
regeneration [121]. Another system presented a hydrogel consisting
of 2-hydroxyethyl methacrylate (HEMA) intowhich PCLmicrocarriers
with active molecule levonorgestrel were encapsulated [134]. Due to
the composite structure, the system showed almost zero-order release
kinetics of the drug over 4 months [134]. A similar approach has also
been taken in bioceramics and biopolymer/bioceramic composite
scaffolds [124,125,135,136]. Porous HA scaffolds were ionically com-
bined with biodegradable PLGA microspheres loading dexametha-
sone, and the composite system demonstrated about 4 weeks of
drug release and corresponding new bone formation in vivo [125].
Moreover, HA/polyurethane scaffolds incorporating antibiotic drug-
loaded ethyl cellulose microspheres were developed and proved to
be effective drug delivering scaffolds for bone regeneration [135]. In
fact, this composite system of scaffolds or hydrogels incorporating
biofactor-loaded microspheres has attracted much attention as a
means to deliver multiple biofactors (more than two biofactors) by
means of loading additional biofactors directly within the scaffold.
This is detailed in the following section.
Along with the porous scaffolds and water-containing hydrogels,
spherical forms of biomaterials (microspheres and microcapsules)
are also considered to provide effective 3D substrate conditions for
cellular growth and delivery and further tissue engineering as inject-
able devices [137–141]. When the microspherical cell carriers possess
the delivering ability of biofactors, the cell instructive therapeutic po-
tential can be greatly improved [137]. PLA/PLGA-based polymeric
nanocomposite microspheres were suggested as 3D scaffolds for
stem cell therapy with sustainable drug release properties [142]. Pro-
teins loaded within the microspheres of block copolymers PLGA-PEG-
PLGA with varying compositions were effectively and continuously
released, without an initial burst, indicating the potential of the
approach as a new cell delivery and therapy tool for injured tissues
[143]. A recent advance involves a porous structure within the
composite biomaterials for tissue regeneration, Adv. Drug Deliv. Rev.
http://dx.doi.org/10.1016/j.addr.2012.03.009
11R.A. Pérez et al. / Advanced Drug Delivery Reviews xxx (2012) xxx–xxx
microspheres to hold a large cell population. This approach demon-
strated a better delivery potential than the dense microspheres. Com-
bining the delivery potential of therapeutic molecules with those
porous cell-carriers warrants further study.
In terms of cell-delivering or carrying biomaterials, microencapsu-
lation systems have long shown great promise. Alginate-based hydro-
gel microspheres were developed a long time ago to encapsulate
tissue cells; this approach allows good viability and tissue-specific
cell functions [144–147]. Some recent advances on this microencap-
sulation technology have highlighted composite systems that incor-
porate biofactors or biofactors-loading particles. A composite
delivery system made of alginate-poly(L-lysine)-alginate microen-
capsulated myoblasts incorporating dexamethasone-loaded PLGAmi-
crospheres has proven to be an effective composite release system.
The dexamethosone released from the PLGA generates a potential
immune-privileged local environment to the cells that are microen-
capsuled and ensheathed [148]. A composite microencapsulation gel
system made of thiolated heparin with acrylated PEG was also devel-
oped to carry hepatocytes inside. Moreover, hepatocyte growth factor
(HGF) was incorporated within the gel via affinity-binding with hep-
arin. The system allows the slow release of HGF (only 40% release
after 30 days), and promotes hepatic functions [149]. Cell delivering
microencapsulation systems have recently been developed using
temperature-reversible polymers, which are reversible in the sol–
gel transition depending on temperature, simplifying cell manipula-
tion and allowing injectable delivery and subsequent gel formation.
These are considered intelligent cell delivery systems for tissue engi-
neering. This issue will be detailed in Chapter 4.
While biopolymers are versatile in incorporating biofactors, bioac-
tive inorganics such as calcium phosphates and glasses have signifi-
cant limitations in delivering biofactors because they primarily
require high thermal processes in the shape formulation. In this man-
ner, the bioactive inorganics are generally made into composites with
biopolymers mainly those in natural origin to allow shape formability.
However, some of the valuable physicochemical properties of bioinor-
ganic nanoparticles (mainly calcium phosphates), such as high elec-
trostatic charge, surface area and roughness, improve the interaction
with and affinity to biofactors, allowing suitable matrices for drug de-
livering scaffolds [150]. Among the bioactive inorganics, some groups
have self-setting property. Calcium phosphate cements (CPCs) are
among the most attractive group of inorganic biomaterials for use in
biofactor delivery. Some recent studies developed CPC-based compos-
ite biomaterials for this purpose [141,151–154]. α-tricalcium
phosphate-based CPC can self-harden and be formulated into micro-
spheres with the help of collagen to deliver biomolecules. Bovine
serum albumin (BSA), used as a model protein, was safely loaded
within the microspheres and then released sustainably over a month
[141]. In order to stimulate osteoinduction, BMP-2 was incorporated
within tetracalcium phosphate/dicalcium phosphate anhydrous-
based CPC composite with chitosan, whichshowed significant im-
provement of osteoblastic cell functions [155]. The addition of alginate
into CPC based on calcium carbonate/monocalcium phosphate mono-
hydrate prolonged the release of gentamicin, providing a reservoir
system for antibiotic delivery with bone regeneration capability
[156]. Incorporation of polymer microspheres as loading biofactors
within CPC has also been pursued to achieve a system with a sustain-
able release profile [153]. In the delivery systems of biofactors partic-
ularly targeting hard tissues, the composite approach of bioactive
inorganics like CPCs with biopolymers is very promising, in that the
biopolymers facilitates shape formation and mechanical flexibility,
while the bioactive inorganics provide active and bone cell stimulating
matrix conditions with biologically favorable and relevant ionic
sources within the composition [151,157–159]. Calcium phosphate
(CaP) mineral particles have also been used to load biological proteins
and chemical drugs such as alendronate and antibiotics [160,161].
PLGAmicrospheres surface-mineralized with biomimetic CaP showed
Please cite this article as: R.A. Pérez, et al., Naturally and synthetic smart
(2012), doi:10.1016/j.addr.2012.03.009
great potential to bind proteins and their sustainable release. Cyto-
chromeC and BSA, used asmodel proteins, were effectively and tightly
bound to the CaP mineral phase, and then subsequently released al-
most linearly for over a month [162]. When the proteins were encap-
sulated within the inner PLGA part the release rate was further
reduced. A delivery system of drugs within the CaP microspheres
was also developed, where alendronate was in-situ loaded [163]. A
sustainable release pattern of drug over 40 days was evident, and
the release rate was controllable by modulation of the proportion of
amorphous phase and the consequent degradation rate. The drug-
loaded CaP microspheres demonstrated biological activity, effectively
inhibiting osteoclast differentiation. Incorporation of the CaP nano-
particle with pre-loaded drugs into biopolymers has been sought to
develop slow drug-releasing scaffold systems. The CaP nanospheres
combined with poly(L-lactic acid) (PLLA)-PEG hybrid polymer can
sustain the release of ibuprofen over 150 h [163]. Coating of CaP nano-
particles with PLGA was also effective in producing a sustainable de-
livery system. Tigecycline loaded into the composite system showed
a sustained release over 20 days [164].
Currently, one interesting and attractive form of biomaterials scaf-
folds is the nanofiber, which is mainly produced by an electrospin-
ning process. A number of target tissues including skin, nerve,
muscle, cartilage, blood vessel and bone, have utilized the nanofi-
brous matrices in support of cells, for treatment of damage or disease,
or for implementing tissue-engineered constructs [165–170]. There-
fore, utilizing the nanofibrous scaffolds as drug delivery systems has
become attractive. Although synthetic biopolymers have shown bet-
ter mechanical properties than natural ones when formulated into a
nanofibrous structure, the organic solvents used to dissolve the syn-
thetic polymers are not readily available for the use of biofactors.
Even so, the hydrophilic biofactors are segregated and not homoge-
nously distributed within the synthetic matrices. For the loading of
growth factors, some common biological proteins such as BSA were
used to hold and stabilize the growth factors like NGF, which was
subsequently dispersed in the co-solvent of the synthetic copolymer
of ε-caprolactone-ethyl ethylene phosphate and then electrospun
into nanofibers [171]. The use of BSA significantly stabilized the
growth factors, showing a sustainable release profile over 90 days. In-
stead of using BSA, collagen was used with PCL synthetic polymer and
also showed similar effects on epidermal growth factor (EGF) release
from the electrospun nanofibers [62]. Heparin has also been highly
effective in stabilizing growth factors like EGF and bFGF within the
PLA nanofibers [172]. However, those mixture systems are considered
rather case-specific, not being applicable to general systems, and have
limitations in controlling the drug release profiles.
A more elegant and general strategy to gain sustainable and con-
trolled release pattern of biofactors from the electrospun nanofibers is
the core-shell (or dual-concentric) design [173–175]. The core-shell
structure is possible by a specific design of nozzle, a part of the electro-
spinning apparatus, to produce concentrically aligned inner and sheath-
ing outer nozzles. Two different solutions are fed into each nozzle part
and then electrospun into a core-shell structured nanofiber [176]. The
core part is primarily composed of water-soluble biopolymers contain-
ing target biomolecules, while the shell area is made of biopolymers
that are effective in sheathing and protection of the inner part, thus
allowing sustainable release of the drugs. As the shell part is in direct
contact with the biological milieu, the composition should also be bio-
compatible and is preferred to trigger cell adhesion and, possibly, to fa-
cilitate degradation. As such, the release of the encapsulated biofactors
is profiled in a highly dependent manner both on the core composition
and on the outer shell properties; namely, the release pattern will not
only depend on the interactions between the core biomaterials and bio-
factors, but also on the degradation rate of the shell composition and
diffusivity of water molecules through the shell. Therefore, a more sus-
tainable release profile may be facilitated by choosing an inner compo-
sition with stability and high affinity with biofactors and an outer
composite biomaterials for tissue regeneration, Adv. Drug Deliv. Rev.
http://dx.doi.org/10.1016/j.addr.2012.03.009
12 R.A. Pérez et al. / Advanced Drug Delivery Reviews xxx (2012) xxx–xxx
material with low solubility and thicker layer. Some recent studies have
highlighted the effectiveness of this core-shell design for prolonged de-
livery of growth factors. Silkfibroin/PCL core-shell nanofiberswere pro-
posed as a potential tissue engineering and drug release system [174].
Water soluble synthetic polymers, PEG or PEO, were also often used as
the core material sheathed to deliver growth factors. PEG/PCL core-
shell nanofiber incorporating BMP-2 was shown to have slow release
of rhBMP-2 and biological functions in vitro and in vivo for bone tissue
[173]. For neural tissue regeneration, a core-shell nanofiber made of
outer PLLA-PCL and an inner BSA/NGF portion was developed. In vivo
results of the drug-loading nanofiber implanted in the 10 mm gap of
long sciatic nerves showed excellent nerve reconstruction with similar
performance to the autograft [175].
Eventually, the delivery systems of protein like growth factors are
designed to release themmainly outside of cells, while simultaneously
allowing interactions with cellular membrane receptors to transmit
instructive signals to the intracellular compartments. Therefore, sub-
stantial attention has been directed to the development of those scaf-
folds and matrices to load the proteins safely and deliver them at a
sustained rate. On the other hand, the delivery strategy of genes, in-
cluding plasmid DNAs, mitochondrial RNA, and small interfering
RNA (siRNA), has been sought in a different waymainly in the interac-
tion with cells, as the release of the genes should occur in the intracel-
lular compartments. Thus, the delivering biomaterials are developed
mainly as nanomaterials in the form of nanoparticulates and nanocap-
sules that can contain a large amount of genetic material inside the
particle space as well as penetrating effectively through the cell mem-
brane and even into the nucleus, in order to provide signals and cues
in genetic modification. Studies of the development of non-viral
gene delivery vehicles from polymeric materials including liposomes,
poly(ethylenimine) (PEI), chitosan nanoparticles, dendrimers,