Nanotechnology in the regulation of stem cell behavior

Abstract

Stem cells are known for their potential to repair damaged tissues. The adhesion, growth and differentiation of stem cells are likely controlled by the surrounding microenvironment which contains both chemical and physical cues. Physical cues in the microenvironment, for example, nanotopography, were shown to play important roles in stem cell fate decisions. Thus, controlling stem cell behavior by nanoscale topography has become an important issue in stem cell biology. Nanotechnology has emerged as a new exciting field and research from this field has greatly advanced. Nanotechnology allows the manipulation of sophisticated surfaces/scaffolds which can mimic the cellular environment for regulating cellular behaviors. Thus, we summarize recent studies on nanotechnology with applications to stem cell biology, including the regulation of stem cell adhesion, growth, differentiation, tracking and imaging. Understanding the interactions of nanomaterials with stem cells may provide the knowledge to apply to cell–scaffold combinations in tissue engineering and regenerative medicine.

• stem cells
• nanotechnology
• nanotopography
• cell adhesion
• differentiation

Introduction

Cells in a physiologic condition are challenged by both soluble and insoluble stimuli from the microenvironment. Most cells either sit on the extracellular matrix (ECM) or are embedded in it, which provides both the chemical and physical properties to support cell migration, proliferation, survival and differentiation [1–4]. Other than the chemical cues that are exerted on cells, physical properties, such as two-dimensional (2D) surface topography, three-dimensional (3D) structures, 3D porosity, matrix stiffness, etc, have become more sophisticated and diverse from bench to medical applications. Among these physical properties, nanoscale features have caught researchers’ attention. These nanoscale structures or features consist of small fibers, pits and grafts on which cells can grow and exert their function in such features that mimic the in vivo microenvironment.

From the physiologic point of view, the cells and tissues are built up in nanoscale dimensions, in which the general size of a cell is around 10 μm in diameter whereas the supporting tissue architecture which forms the intrinsic network is in the range of the nanoscale [5]. This nanoscale architecture promises that nanotechnology will become an important area for tissue engineering and regenerative medicine. Therefore, using microfabrication techniques to generate microstructures may provide inappropriate physical cues to regulate cell behavior at a physiological level. Since the supportive intrinsic architecture is complex and is made of pores, grooves, ridges, etc at the nanoscale, the nanostructures, therefore, provide important physical cues at the cellular levels which regulate cellular behavior similar to physiological conditions.

Nanoscale structures or nanomaterials are defined by the size of the structures or materials (generally at 1–100 nm) that can generally be used in different areas, including tissue engineering, drug release, cell therapy, gene delivery and others [6–8]. The term ‘nanotechnology’ is the technique of designing, constructing, generating and utilizing nanoscale structures or materials, which was first denoted by Professor Norio Taniguchi, when he described semiconductor processes that exhibited characteristics of the order of nanometers [6, 9]. Such manufactured nanoscale devices and systems can provide novel utilizations to significantly improve the physical and biochemical properties of materials for further clinical applications. In the past few decades, nanotechnology has obtained tremendous achievements which improved the performance of nanoscale devices in disease diagnosis, prevention and treatment. This fascinating technology may provide new strategies for regenerative medicine and further lead to the development of interesting and innovative tools to improve and restore tissue functions. Thus, researchers from multidisciplinary fields have incorporated the expertise from materials science, engineering and biology to fabricate biological mimetic nanoscale scaffolds to repair and replace damaged tissues. Given the optimal combination of scaffolds, cell sources and biological signals have become the gold standard of tissue engineering, and the use of micro-nanofabrication techniques to generate scaffolds to guide stem/progenitor cell adhesion, spread, differentiation and migration are now top issues in tissue engineering and regenerative medicine. In this regard, combinations of more-sophisticated nanomaterials with progenitor cells or stem cells (SCs) and proper biological signals may provide further opportunities to improve therapeutic efficacies in regenerative medicine.

1. General introduction SCs

The current challenge of regenerative medicine is the shortage of tissues and organs to replace damaged organs in patients who are suffering from chronic diseases. A possible alternative way to circumvent the organ shortage is to orchestrate cell-based therapies with tissue-engineered scaffolds. One potential source of cells, for this purpose, is SCs. SCs derived from either embryonic origins (e.g. embryonic SCs (ESCs)) or adult tissue origins (adult-derived SCs (ADSCs)) for application in cell therapy, tissue engineering and regenerative medicine have drawn researchers’ attention in recent years. These cells are known to possess the potential to differentiate into certain lineages which can ultimately replace cells and restore the vital functions of damaged tissues. SCs can be isolated from different tissues and subsequently maintained and manipulated in the in vitro culture systems. Pluripotent SCs (PSCs), such as ESCs or inducible PSCs (iPSCs), are capable of differentiating into three different germ layers, of the endoderm, mesoderm and ectoderm, but not the placenta [10–12]. The high self-renewal and differentiation capacities of these cells have led to broad applications in tissue engineering and regenerative medicine for treating several diseases, including neurological, hepatic, hematopoietic and diabetic diseases [13–15]. However, one of the major obstacles of using these cells is that they can give rise to teratomas when transplanted into tissues [16, 17] and such malignant tumors can eventually kill patients. These results suggest that both in vitro and in vivo growth control and ESC fate decisions must be carefully managed. In addition, ADSCs (e.g. mesenchymal SCs (MSCs)) isolated from bone marrow, adipose tissue, the placenta, etc, possess limited differentiation ability in that they can only differentiate into specific lineages [18, 19]. These cells can still be induced to differentiate into mesenchymal lineages, such as bone, cartilage, muscle, fat, ligaments, etc [18, 19] and transdifferentiate into other lineages such as neurons and hepatocytes [20, 21]. Recent work demonstrated that MSC fate is dictated by soluble factors, and also physical cues, such as substrate stiffness, ECM topography and mechanical stimulation [22–25]. Understanding environmental cues that regulate SC fate decisions can provide insights into how to maintain and control the differentiation status so that these cells can be prepared for tissue engineering.

Recent studies focused on SC niches and how microenvironmental cues regulate the SC fate. SC niches consist of the ECM, other cells and chemical factors (such as growth factors and cytokines) that regulate SC behaviors [26, 27]. Although traditional biological tools allowed the systematic investigation of the regulation of SC self-renewal and differentiation by soluble chemical factors, epigenetic factors, such as physical signals, have long been neglected. Previous studies revealed mechanical cues in the regulation of SC growth and differentiation [22–25, 28, 29]. In fact, increasing evidence has demonstrated that ‘nanoscale’ physical properties in SC niches may contribute to self-renewal and commitment of SCs into specific lineages [30–34]. These approaches allow the fabrication of SC microenvironments in the nanoscale range, which may be relevant to the study of nanotopography, nanostructures and nanofibrils in controlling primary SC behaviors, such as adhesion, growth and differentiation [35, 36]. In this context, we focus on the application of nanotechnology in regulating SC fate decisions and orchestrating these nanomaterials and SCs in future advances in tissue engineering and regenerative medicine.

2. Nanotopography controls SC behavior

2.1. Cell adhesion to the ECM and its machinery

From tissue engineering aspects, a good-quality scaffold is required and is an essential tool to enable cells to interact with host tissues to repair damaged parts. The best-known scaffold materials are ECM proteins, where these matrix proteins provide structural support and a physical environment for cells residing in tissues. It is known that most ECM proteins form fibrillar networks with sizes ranging from tens to hundreds of nanometers and/or even up to several microns [37, 38]. When cells interact with the ECM, it involves direct contact with the ECM and the subsequent establishment of anchoring points to the ECM. A family of transmembrane proteins on cell membranes that mediate these cell–ECM interactions is known as integrins [39]. Integrins are heterodimerized cell membrane-anchoring proteins which consist of α and β subunits. Combinations of α and β subunits determine their own binding specificity to different ECMs [39]. Cell adhesion and spread against the ECM are mediated by activation and clustering of integrins, which in turn result in the recruitment of proteins forming focal adhesions (FAs) and the organization of actin cytoskeletons [39]. The specific binding motifs on different ECMs can be recognized by different integrins, and integrin activation and clustering are regulated by cell–nanoscale ECM interactions. For example, αvβ3 integrin preferentially recognizes the Arg-Gly-Asp (RGD) peptide sequence [40].

2.2. Nano-engineered ECM substrate regulates integrin-mediated SC adhesion to ECM

To test whether interactions of cell–nanoscale are mediated by integrins/FAs, a previous study using block-polymer micelle nanolithography [41] deposited RGD peptide-conjugated gold nanoparticles (NPS) onto a glass surface and tested the ligand spacing in cell spread and FA dynamics. They found that cells plated with a 108 nm spaced pattern showed delayed spread and rapid FA turnover compared to cells plated with a 58 nm spaced pattern. Cells plated with a 108 nm spaced pattern also exhibited rapid and repeated protrusion–retraction cycles, suggesting instability of the FAs and higher motile behavior of these cells [42]. Results from that study indicate that a proper ligand density and spacing are crucial to dictate the maturation and stabilization of integrin-mediated adhesion, which in turn, regulates cell spread and FA maturation. In addition, to generate a pit-shaped nanoscale topography (14 − , 29- and 45 nm deep pits) using polymer demixing to fabricate poly(l-lactic acid) (PLLA) and polystyrene (PS), Lim et al [43] discovered that selected integrin-mediated cell attachment and spread, and FA formation were increased in human fetal osteoblasts plated on shallower-depth nanopits. Park et al [44] created self-assembling layers by anodizing titanium sheets to generate vertically oriented TiO₂ nanotubes with diameters ranging 15–100 nm to investigate human (h)MSC responses upon plating on the substrate. They found that MSCs preferentially adhered to thinner nanotubes (< 50 nm) and optimal cellular responses (e.g. cell proliferation and migration) occurred with nanotubes with a lateral spacing of 15–30 nm. Clustering of β1 integrin and paxillin in FAs with a well-spread morphology was found in cells plated on thinner nanotubes (15 nm), while cells plated on 100 nm nanotubes displayed only a few FAs with an un-spread morphology and cells eventually died due to anoikis. The integrin-focal adhesion kinase-extracellular-signal regulated kinase signaling cascade might be responsible for relaying nanotopographical effects (figure 1) [44]. These results indicate that a lateral spacing of <50 nm is crucial for integrin clustering, leading to subsequent cell adhesion and spread against the ECM substrate [44]. To further determine the spatial distance of adhesion sites for integrin clustering-mediated cell adhesion, Huang et al used micelle nanolithography to produce ordered and disordered gold nanopatterns on glass to quantitatively generate nanoscale patterns and test integrin-dependent cell adhesion. The cyclic RGD ligand was covalently linked to an Au nanopattern using a self-assembled monolayer method [45] and biotinylated poly(ethylene glycol) was used to block nonspecific adhesion (figure 2). The results revealed that the nanopattern-ordered substrate with local inter-ligand distances of <70 nm was critical for integrin clustering and subsequent FA formation, actin cytoskeletal rearrangement and cell adhesion in mouse pre-osteoblasts. It appears that although the global inter-ligand spacing was > 70 nm, the disordered close proximity of the spots resulted in more integrin clustering than with the ordered nanopattern. This finding appears to be similar to that of a previous study conducted by Kühn, in which he proposed that the integrin-binding motif (Gly-Phe-Hyp-Gly-Glu-Arg, or GFOGER) at the collagen fibril surface demonstrated a repetitive distance of 68 nm [46]. These results suggest that a critical involvement of inter-ligand spacing controls the clustering and activation of integrins.

Figure 1. (A) SEM images of self-assembling layers of vertically oriented TiO₂ nanotubes generated by anodizing titanium sheets. (B) Fluorescence images of FA (red, paxillin) and actin (green) staining of MSCs grown on 15 nm TiO₂ nanotubes (a, c) or 100 nm nanotubes for 1 (a, b) and 3 days (c, d). Blue, nuclear staining (DAPI). (C), (D) Schematic illustration of a hypothetical model of nanoscale spacing in directing SC fate. (C) A spacing of 15 nm is suitable for integrin clustering and subsequent FA assembly and actin cytoskeletal organization, while a nanotube spacing of > 70 nm suppresses FA assembly. The images are reprinted with permission from Park et al 2007 Nano Lett. 7 1686–91 [44]. Copyright 2007 American Chemical Society.

Figure 2. (A), (B) Schematic illustration of the preparation of a nanopatterned surface for cell adhesion. (A) Cartoons indicate the micelle nanolithographic technique to make (1) ordered and (2) disordered gold nanopatterns on the glass. PS homopolymers were used to make the ordering-interference reagent. (B) Fabrication of nanopatterned ligands on a layer of M-PEG-Si(OMet)₃ to prevent cell adhesion. The protruding gold NPs were biofunctionalized with c(-RGDfK-)-thiol ligands, and cell adhesion on the resulting Arg-Gly-Asp (RGD)- nanopatterned surface was examined. (C), (D) Schematic illustration of the regulation of integrin clustering and subsequently formed FA on ordered/disordered nanopatterned surfaces. Lateral spacing between two neighboring RGD ligands of < 70 nm led to integrin clustering and further FA complex formation (C), while at a spacing of > 70 nm, neither integrin clustering nor FA formation occurred in this context. Images are reprinted with permission from Huang et al 2009 Nano Lett. 9 1111–6 [45]. Copyright 2009 American Chemical Society.

2.3. Nano-helical/fibril structure regulates SC adhesion

It is also known that ECM proteins assemble a fibril network with defined periodicity [47]. The nanohelical shape and periodicity of ECM proteins are important because mutations that occur in structural genes, such as procollagen-1, result in changes in the periodicity of collagen fibrils, leading to several diseases, including osteogenesis imperfecta [48, 49]. To test whether the nanohelical structure and periodicity are important to SC adhesion, Das et al [50] examined 1 mM gels of 16-2-16 l-tartrate which form a twisted nanohelical structure in the presence of chiral tartrate counterions. The handedness of the ribbons depends on the d or l helical configuration of tartrate, and the twist pitch continuously decreases upon increasing the enantiomeric excess [51] (figure 3). These helical organic nanoribbons are based on the self-assembling organic nanostructures of Gemini-type amphiphiles to access chiral silica nanoribbons with two different shapes and periodicities (twisted ribbons and helical ribbons). Using chemical modification, the fluorescence of the RGD peptide can be covalently immobilized on helical nanoribbons. They found that a specific periodicity of 63 nm (± 5 nm) with the helical ribbon shape induced formation of fibrillar FAs and led to osteogenic induction in MSCs, while 100 nm (± 15 nm) twisted nanoribbons did not lead to an osteogenic commitment [50]. Interestingly, helical nanoribbon-induced MSC osteogenic differentiation did not require chemical induction factors. This study suggested that MSCs can resolve the helical and nonhelical nanostructure of the microenvironment and indicated that a proper 3D nanotopography may be crucial for MSCs making an osteogenic commitment. However, detailed mechanisms of how these cells sense the nanohelical structure of the fibrillar network and turn it into signals to regulate MSC adhesion and differentiation remain to be elucidated.

Figure 3. (A) SEM images of twisted nanoribbons (left panel) and helical nanoribbons (right panel). (B) Fluorescence images of SCs plated on twisted nanoribbons (left panels) or on helical nanoribbons (right panels) and stained with vinculin (red), F-actin (green) and nuclei (blue). (C) Quantified results of relative areas of FAS on control glass, control glass grafted with Arg-Gly-Asp (RGD), twisted nanoribbon-RGD and helical nanoribbon-RGD. The images are reprinted with permission from Das et al 2013 ACS Nano 7 3351–61 [50]. Copyright 2013 American Chemical Society.

In addition to these helical structures, the other natural structure of collagen is a unique triple helix which requires the (Gly–X–)n repeating sequence to form triple-helix collagen [52]. Collagen-regulated cellular functions are known to be mediated by interactions of the specific epitope, GFOGER, to receptors, such as α2β1 integrin [53]. To elucidate the detailed mechanisms of how this triple helix regulates integrin-dependent cell adhesion, Yamazaki et al [54] developed a self-assembled peptide and they used step-wise regional-selective cysteine chemistry to generate disulfide bonds between the three strands in a head-to-tail trend. The staggered trimers were then elongated by self-assembly. Using this technique, they were able to develop collagen-mimetic triple-helix peptides and tested collagen receptor-mediated cell adhesion using these collagen-mimetic triple-helix peptides. They found that in the presence of the integrin-binding sequence, GFOGER, in the mimetic peptide, cell adhesion, cell spread and FA formation were comparable to those of native collagen and were solely mediated by α2β1 integrin [54]. This study suggested a possible application of the synthetic mimetic triple-helix peptide in tissue engineering. Taken together, these studies demonstrated the importance of nanotopography in integrin-mediated cell adhesion and highlighted the potential application of these nanotopographical techniques of fabricating scaffolds for regenerative medicine.

2.4. Nanotopographic cues regulate SC growth and differentiation

As described previously, topographical cues may differ in size and shape in vivo. Different sizes and shapes are crucial for cell adhesion and may be important for SCs differentiating into specific lineages. By using different nanofabrication methods or a three-step process of electron beam lithography, nickel die fabrication and hot embossing, Dalby et al [55, 56] generated nanopatterns with different arrangements on a polymethylmethacrylate (PMMA) substrate with pit diameter at about 120 nm, 100 nm in depth and 300 nm center-to-center spacing [55, 56]. hMSCs plated on these nanotopographies with controlled disorder showed a higher spreading area, with more FAs, a well-organized cytoskeleton and increased expression of osteogenic markers compared to those highly ordered or randomly dispersed nanotopographies. Similar results were obtained by Khang et al [57]. Furthermore, by combining nanotechnology, fluorescence in situ hybridization and DNA microarray, Tsimbouri et al [58] (from Dalby's group) were able to elucidate direct signaling from nanotopology-regulated nuclear organization and gene expression profiles in hMSCs. In this study, they were able to determine the up- and down-regulation of genes on individual chromosomes and gene activity upon plating cells on different nanotopographies. These nanotopography-induced deregulated genes accompanied by changes in the nuclear positioning of chromosomes and topographical-related changes occurred toward the end of the chromosomes where osteogenic-related genes are clustered. These results again demonstrate that a proper nanotopography can direct SC commitment through changes in chromosome re-positioning in the nucleus which may subsequently result in gene de-regulation.

In addition, Wang et al also preformed chemically modified nanografted surfaces to investigate whether topographical or chemical cues regulate myoblast differentiation. Results revealed that C2C12 myoblasts grown on RGD-conjugated grooved PS displayed better adhesion, proliferation and myogenic differentiation. In addition, the anisotropic topography directed the alignment of myoblasts and myotube formation, which mimicked the in vivo organization of muscle cell alignment [59]. These results suggest that a proper topography and surface chemistry are sufficient to provide local environmental cues to guide hMSC commitment into specific lineages.

2.5. Nanofibers regulate ESC stemness

One of the big challenges in tissue engineering is to fabricate a biomimetic microenvironment that can mimic SC niches to control SC self-renewal and differentiation. Recent studies used electrospun nanofibers to mimic the in vivo SC microenvironment. Using either natural or synthetic polymers or their combination, researchers were able to generate a porous 3D architecture of electrospun nanofibers with excellent interconnectivity [60, 61]. As mentioned above, SC self-renewal and lineage commitment respond to both chemical and physical stimuli, and increased evidence has demonstrated that physical cues mediate SC fates. Early studies using a conventional method maintained human (h)ESCs in an undifferentiated status including the application of feeder cells (growth-incompetent mouse embryonic fibroblasts (MEFs)) and the leukemia inhibitory factor [10]. However, from a cell therapy point of view, this method might not be able to generate massive amounts of cells for tissue engineering purposes and contamination by cells from other species may result in serious side effects. To overcome these difficulties, a previous study applied Matrigel or laminin-coated plates to replace the MEF feeder layer. hESCs can remain undifferentiated for at least 130 population doublings without losing hESC stemness [62]. This model can help researchers eliminate contamination by cells derived from other species as well as unknown factors generated from co-culture systems. To further establish an SC niche-mimicking microenvironment, Liu et al used polymethylglutarimide (PMGI) in tetrahydrofuran and cyclopentanone to generate electrospun nanofibers. In the absence of mouse MEF feeder cells, PMGI nanofibers served as a cellular scaffold to maintain the self-renewal of mouse ESCs without losing pluripotency [63]. In addition, both density and topological effect of the PMGI nanofibers were important to maintain ESC self-renewal and undifferentiation status. Similar results were obtained using a commercially available synthetic polyamine scaffold (Ultra-Web) to maintain the mouse ESC self-renewal ability [64]. Unfortunately, this scaffold/feeder cell-free system was unable to replace the conventional method to maintain hESCs. Instead, several studies reported that nanofibers promote both mouse and human ESC differentiation into specific lineages [65–68]. To improve this system, Gauthaman et al used electrospun polycarprolactone (PCL)/collagen and PCL/gelatin nanofibers with diameters of around 280 nm in the presence of mouse MEFs to maintain hESC stemness. Their results revealed that the growth rates of hESCs on both nanoscaffolds in the presence of mouse MEFs significantly increased compared to the conventional control. The sizes of hESC colonies were also larger on both scaffolds than the control [69]. Although this novel culture system might not be able to reduce contamination by xenogenic cells and unexpected secreted factors generated from co-culture systems, this improved system may have the potential to scale up production for further study and treatment.

2.6. Nanofibers regulate adult SC fates

2.6.1. Surface-modified nanofibers enhance hematopoietic SC homing and adhesion.

In addition to regulating ESC fate decisions, recent reports also demonstrated the application of nanofibrils in adult SC lineage commitment. Chua et al produced surface-functionalized poly(acrylic acid (PAA))-grafted polyethersulfone (PES) nanofiber meshes (with diameters of around 500 nm) to investigate the adhesion and expansion of human umbilical cord blood hematopoietic stem/progenitor cells. Results demonstrated that the aminated nanofiber mesh and film facilitated a higher degree of adhesion and increased expansion of CD34⁺CD45⁺ cells [70]. To mimic the local microenvironment to attract more bone marrow-derived hematopoietic SCS (BM-HSCs), Ma et al generated E-selectin-coated electrospun nanofibers with poly(dl-lactide-co-glycolide (PLGA)) blended with type I collagen. Results showed that in the presence of an E-selectin coating, the capture efficiency of BM-HSCs was significantly enhanced and was more rapid than the uncoated control [71]. These studies indicate that proper modification of nanofibers may result in a mimetic microenvironment which provides sufficient topographical and chemical cues to facilitate HSC homing.

2.6.2. Nanofibers enhance SC-mediated neurogenic differentiation and neurite guidance.

On the control of neurogenic differentiation, several studies indicated that nanoscale features can guide SCs toward a neurogenic lineage. Kijeńska et al [72] demonstrated that nanofibers made of poly(l-lactic acid)-co-poly (ε-caprolactone) (poly(LLA-CL) with collagen I/III showed increased neural SC (NSC) proliferation, cells aligned on the scaffold and expression of neuron-specific neurofilaments. Yang et al also performed electrospinning of nano/micro scale PLLA aligned fibers on coverslips (with diameters ranging from 150 nm to 3 μm) in the regulation of neural SC (NSC) differentiation. The results demonstrated that the rate of NSC differentiation to neuronal cells was higher when NSCs were plated on PLLA nanofibers than that of micro fibers, regardless of the fiber alignment [73], suggesting the importance of nanotopography rather than microtopography. Similar results were obtained from Qu et al [74] using electrospun silk fibroin that demonstrated the beneficial effect of the use of nanofiber (with diameters ranging up to 400 nm) rather than microfiber (1.2 μm). Given these topographical features, including fiber dimension and pattern which may be important to induce neurogenic differentiation, He et al [75] used either aligned PLLA scaffold (with diameters ranging from 307 to 917 nm) or random fibers (with diameters ranging from 327 to 1150 nm) to study NSC differentiation. The results showed that with a similar diameter range, longer neurites could be obtained on aligned fibers than random fibers, while cell proliferation and viability controls were regulated differently by medium-size fibers or randomly aligned fibers, respectively. These results suggest that the surface topographies of fibrous scaffolds, including the fiber pattern, dimension and mesh size, play roles in regulating NSC behavior.

Given that the guidance of neurites to their specific location was also important for neural regeneration, several studies used nanofibers to investigate the guidance control of neurite. The results found that nerve cell/NSC elongation and neurite outgrowth on aligned nanofibers was parallel to the direction of orientation of nanofibers [76, 77]. To promote neurogenic differentiation, the guidance of neurites from pre-synaptic neurons toward post-synaptic neurons or muscle was also found to be important. This strategy is important for spinal cord injuries given that proper guidance from biomaterial scaffolds can help bridge injury gaps. Studies conducted by several groups established nerve guidance features which can lead to nerve regeneration in combination with SC therapies. Results from Neal et al [78] showed that the alignment and composition of nanofibers were important for peripheral nerve regeneration [78]. They found that the content of laminin in laminin-PCL nanofibers was important for neurite guidance in an in vitro assay. More importantly, when using a rat tibial nerve defect model, they discovered that the retrograde nerve conduction speed in those animals receiving aligned nanofiber conduits was significantly faster than that of animals receiving random ones. Taken together, these studies demonstrate the feasibility of using these nanofibers for peripheral nerve regeneration and highlight a potential application of the combination of nanofibers with soluble signals or SC therapy for nerve regeneration.

2.6.3. Nanofibers enhance SC-mediated adult tissue differentiation.

The application of hMSCs for regenerating the skeletal system has been consistently reported. The most common strategy to induce MSC differentiation into an osteogenic lineage is to use dexamethasone-driven osteogenic induction or an endogenous soluble factor such as bone morphogenetic proteins [18, 22, 23]. Increasing evidence was also reported of nanofibers facilitating MSC differentiation into an osteogenesis lineage. Gandhimathi et al [79] applied biocompatible PCL/poly(α, β)-dl-aspartic acid/collagen/nanohydroxyapatite (PCL/PAA/Col/n-HA) nanofibers (with diameters from 150 to 250 nm) with a calcium phosphate coating. hMSCs plated onto these blended scaffolds showed increased proliferation and osteogenic differentiation. In addition, using a novel electrospinning method, Rampichová et al were able to fabricate 3D elastic PCL nanofibril scaffolds. The adhesion, spread, proliferation and survival of hMSCs were better when cells were plated on a 3D matrix than on a 2D scaffold. Expression of osteogenic-related genes was higher in cells grown on the 3D matrix [80]. Similar results were obtained using hyaluronic acid (HA)-containing composite nanofibers, PCL nanofibers, HA-chitosan nanofibers and others to increase osteogenic differentiation both in vitro and in vivo [81–82].

In addition to osteogenic induction, combining proper nanofibers with MSCs was also reported to be applied to cardiomyogenic induction for regenerating the heart. Using a fibrin-based multiscale electrospun PLGA composite scaffold, Sreerekha et al investigated MSC differentiation toward a myocardial lineage and the potential utilization of composite scaffold-MSCs for myocardial regeneration. In this multiscale scaffold, the range of diameters of fibrin was 50–300 nm, while PLGA fibers ranged from 2 to 4 μm. This design was to mimic the structural hierarchy of native myocardial tissues. hMSCs grown on the fibrin-PLGA composite scaffold for 14 days showed increased cardiomyocyte markers, such as cardiac calponin, α-sarcomeric actinin and tropomyosin. More importantly, the fibrin scaffold degraded and the newly synthesized collagen scaffold replaced it [83]. In addition, the strategy of using PCs, soluble signals and nanofiber scaffolds was applied to treat a myocardial infarction (MI) model. Padin-Iruegas et al coupled biotinylated self-assembling peptide nanofibers of tethering biotinylated insulin-like growth factor-1 (b-IGF-1) with a biotin sandwich method to generate nanofiber-IGF-1 (NF-IGF-1) [84]. With this design, they were able to mix cardiac PCs (CPCs) with NF-IGF-1 to treat MI rats. The combination therapy of CPCs plus NF-IGF-1 resulted in greater increases in the cardiac cell number, size and functions and a decrease in the infract size. This combined treatment also increased the length density of newly formed coronary arterioles and activated residential CPCs through paracrine effects [85]. Together, these studies provide evidence showing that synthetic biomimetic nanoscaffolds can provide a niche-like environment for SC/PC fate decisions and application of the tissue engineering triad (cells, signals and scaffold) is able to accelerate the regeneration of damaged tissues.

3. Application of NPs in SC isolation, tracking and imaging

The application of nanotechnology is mainly focused on developing NPs for entrapment and delivery of genetic materials, biomolecules and soluble factors to reinforce the effects of scaffolds in regenerative medicine [86]. Another potential application of these NPs is to isolate cells from mixed populations. As mentioned above, control of ESC growth and differentiation must be carefully managed. As ESCs are induced to differentiate into specific lineages, a very small population of ESCs might not receive the differentiating signals and retain their original ESC characteristics. Upon transplantation of the entire population of cells, even this very small population of undifferentiated cells may have devastating effects. Therefore, isolation of differentiated cells from undifferentiated ESCs has become an important issue. Using a phage display system [87], researchers discovered a series of peptide sequences which can specifically bind to NSCs [88]. By combining the novel peptide (HGEVPRFHAVHL, HGE) with quantum dots, Zhao et al [89] discovered that this HE-quantum dot combination specifically recognized 48/34 kDa proteins on membranes of NSCs derived from monkey ESCs but not hESCs. This study suggests that the quantum dot-conjugated specific peptides could be used to study lineages committed of ESCs and are a potential system for isolating differentiated cells from the ESC-differentiated cell populations.

One of the important issues of using SCs for cell therapy is whether these cells can be homed to target tissues to regenerate damaged tissues. In early studies, cells could only be visualized at the end of the experiment when the animals were sacrificed. Therefore, understanding SC homing to designated tissues using novel labeling techniques/materials without influencing SC proliferation, differentiation or migration is important for tissue engineering and regenerative medicine. Proper labeling allows for practical detection of implanted cells and tracing cells at the defect area to assess the participation of these cells in tissue regeneration. For example, to monitor whether SCs are involved in repair processes, hMSCs were labeled with diverse NPs, including quantum dots, fluorescence-labeled silica NPs, gold NPs and superparamagnetic iron oxide NPs [86] in order to track these cells during live imaging. For example, Ruan et al [90] used polyamidoamine dendrimer-modified magnetic nanoparticles (dMNPs) as a delivery system to introduce the pluripotent transcription factors Oct4, Sox2, Lin 28 and Nanog to generate lentiviruses to induce iPSCs. The resultant lentivirus was ten-fold more potent than viruses produced using the liposome method. After forming iPSCs, these cells were labeled with fluorescent magnetic NPs. The fluorescence signals were visualized by fluorescence microscopy and magnetic NPs could be detected by magnetic resonance imaging. Successful cellular uptake and long-term retention of these NPs in cells are beneficial for the tracking and labeling of these cells after being implanted [91]. Although these different NPs can effectively and efficiently enter and label cells, several concerns were also raised about the application of NPs, i.e. cytotoxicity. The cytotoxic effect is associated with the material size, morphology, composition, surface charge and hydrophobicity [92]. These NP characteristics result in an increase of cytosolic reactive oxygen species, which cause chromosomal aberrations and eventual cell death [93]. Thus, in the context of biocompatibility, enhancing cellular uptake of these NPs for tracking and labeling but with lower cytotoxicity and interference with cellular differentiation need to be carefully addressed. Future research on NP applications should focus on improvements in order to have wider applications for cell labeling, imaging and tracking in different tissues.

4. Bringing nanotechnology from 2D to 3D

As cells reside in the microenvironment, cell-to-cell, cell-to-ECM and even cell-to-chemical stimuli are all regulated spatiotemporally. Thus, in vitro culture systems bring the complex microenvironment from 3D to 2D, which allows us to easily design tools for studying cellular signaling, disease progression and drug testing. When bringing the culture system from 2D to 3D, responses of cells to this spatial change significantly differ. For example, growing Madin–Darby canine kidney (MDCK) cells on regular tissue culture dishes in the presence of hepatocyte growth factor (HGF) resulted in a scattering effect in which cells tended to migrate out of the islet-like structure [94]. On the other hand, if MDCK cells were grown in a 3D collagen gel in the presence of HGF, these cells developed branching tubules [95]. Similar responses can be observed by growing mammary epithelial cells on 2D or 3D collagen gels and the increased collagen stiffness turned mammary epithelial cells from a normal to a cancer-like phenotype [96]. These studies are examples depicting the importance of the 3D microenvironment in controlling cellular behaviors in vivo. Given the inherent complexity of the tissue ECM, bringing the traditional 2D culture into 3D and with nanotopographical designs is by no means a simple task. One possible approach to generate tissue mimetic 3D nanoscaffolds is to use cell-recognizable ECM peptide sequences (such as RGD, GFOGER, etc) and create self-assembling 3D networks. Using this strategy, Silva et al [97] fabricated a self-assembling 3D nanoscaffold with a diameter of 5–8 nm which contained isoleucine–lysine–valine–alanine–valine (IKVAV) peptide sequences derived from laminin. When encapsulating NSCs into the nanoscaffold, the artificial nanoscaffold induced very rapid differentiation of cells into neurons, while suppressing astrocyte development. In addition, Zhang et al [98] further used self-assembling ECM-mimetic bioceramic nanoreservoirs to deliver recombination fibronectin/cadherin (rFN/CDH) chimera proteins to modulate MSC differentiation to bone. The MSCs interacted with the rFN/CDH nanoscaffold through α5β1/αvβ3 integrin, and this nanoscaffold-regulated MSC adhesion, proliferation and osteogenic differentiation were mediated by α5β1/αvβ3 integrin. These 3D nanoscaffolds demonstrated excellent bio-physicochemical and biocompatibility properties, and further served as nanoreservoirs to regulate cell behaviors and functions and promote SC differentiation into designed lineages through ECM–integrin interactions.

5. Potential limitations of using nanomaterials in SC biology

As we described previously, nanotechnology has achieved outstanding progress in tissue engineering and regenerative medicine in a very short period of time. Incorporating cell therapies and chemical signals into nanotechnology has become an emerging interdisciplinary field. However, mechanisms of interactions between nanomaterials and SCs are still not well-clarified, given the interactions of nanomaterials with SCs, the effects of these nanomaterials on SC behavior and the metabolic pathways of NPs that are taken up. One of the big challenges using these nanomaterials is cytotoxicity. As discussed earlier, the fabricated nanomaterial size, shape, surface properties and properties of the material itself contribute to the induction of cytotoxicity. Several studies indicated that in vitro and in vivo mechanisms of toxicity were due to the type or size of NPs [99, 93]. To circumvent these cytotoxicity issues, several groups used different strategies to improve the material effects on SC death [100–102]. In addition, the use of degradable synthetic materials to generate nanomaterials may result in local increases in degraded metabolites. For example, although monomers of PLLA, PGA and PLGA were previously designed using chemicals that exist in metabolic process of the body, the degradation of these polymers results in the local accumulation of acid metabolites which can change the local environment to acidic and subsequently influence cell behaviors. Another possible challenge is metal-containing materials in NPs. These metals include aluminum, iron, silver and other heavy metals. Although several studies evaluated the cytotoxic effects of superparamagnetic iron oxide particles in different cell types, these results showed that internalization of these NPs by SCs did not influence cell growth, viability or differentiation [103, 104]. The accumulation of these heavy metals, for example, using Al₂O₃ NPs to NSCs for neural regeneration, may result in the accumulation of aluminum in nerve tissues which was proved to induce Alzheimer's disease [105, 106]. In conclusion, despite the application of nanotechnological impacts on tissue engineering and regenerative medicine and the very exciting outcomes of such applications, it is still necessary to understand the detailed mechanisms of how these nanomaterials interact with cells to regulate cell fate.

6. Conclusions

Cellular adhesion to scaffolds is the first and most important step in interacting with materials. The feature that dictates cell adhesion to scaffolds often acts at micro- to nanoscale levels due to the involvement of nanoscale topography and integrins. In the current review, the aforementioned studies demonstrate the importance of nanoscale features and materials to the regulation of SC adhesion, growth and differentiation into specific lineages. Although quite a few research teams are working on fabricating biomimetic nanoscaffolds for tissue engineering, the detailed mechanisms of how these nanoscaffolds and nanotopographical features influence SC behaviors remain to be elucidated. In addition, combinations of SC therapy and nanotechnology in tissue engineering and regenerative medicine have achieved significant advances. These combinations allow nanotechnology to engineer scaffolds with various features to control SC fate decisions. However, determining how to design and fabricate functionalized scaffolds with micro- and nanotechnology for each specific application and improving cellular responses have become important trends in tissue engineering and regenerative medicine. This includes the fabrication of 3D biomimetic nanointerfaces onto which SCs can adhere and spread, forming a niche-like microenvironment which can guide SCs to proceed with either self-renewal or differentiation and even provides adequate signals to attract endogenous progenitor cells to heal damaged tissues. Although it showed significant improvements in some in vitro and in vivo studies, determining how to scale it up to industrial levels requires further investigations, and potential obstacles as to whether it can mimic endogenous ECM 3D hierarchical structures are still not known. Continued studies require the state-of-the-art design of reliable scaffolds with low toxicity, controlled 2D surfaces for cell adhesion and assembly in a 3D fashion for tissue engineering and regenerative medicine.

Acknowledgments

We apologize for being unable to cite all relevant papers in the manuscript due to space constraints. We thank Ms Ya-Ting Hsieh for her excellent administrative assistance. This work was supported in part by a grant from the National Science Council, Taiwan (NSC101-2320-B-006-016) and a Taipei Medical University startup grant (TMU101-AE1-B04) to YKW; a Taipei Medical University Excellent Young Investigator Grant (TMU101-AE3-Y03) to CLT; and a grant from E-Da hospital, I-Shou University to KCW (EDAHP-101018).