Substrate Coupling Strength of Integrin-Binding Ligands Modulates

Sep 21, 2015 - Tian Zhou , Yizhou Zhu , Xia Li , Xiangmei Liu , Kelvin W.K. Yeung , Shuilin Wu , Xianbao Wang , Zhenduo Cui , Xianjin Yang , Paul K. C...
0 downloads 0 Views 2MB Size
Download PDF
Letter pubs.acs.org/NanoLett

Substrate Coupling Strength of Integrin-Binding Ligands Modulates Adhesion, Spreading, and Differentiation of Human Mesenchymal Stem Cells Chun Kit K. Choi,† Yang J. Xu,† Ben Wang,† Meiling Zhu,† Li Zhang,† and Liming Bian*,†,‡ †

Department of Mechanical and Automation Engineering, ‡Shun Hing Institute of Advanced Engineering, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong, China S Supporting Information *

ABSTRACT: Substrate stiffness has been shown to regulate the differentiation fate of human mesenchymal stem cells (hMSCs). hMSCs sense and respond to substrate rigidity by exerting traction forces upon the binding between integrins and integrin-specific ligands present on the substrate surface. However, in previous studies, integrin-specific ligands such as Arg−Gly−Asp (RGD) peptides are always grafted to the substrate by a permanent covalent bond. Whether the coupling strength of integrin-specific ligands on substrate will influence cell behaviors has not been explored. In this work, we have developed a facile platform to investigate the effects of varied coupling strength between the RGD peptide and the glass substrate on stem cell behaviors. Glass coverslips are decorated with positive charges by silanization using (3-aminopropyl) triethoxysilane (APTES) to immobilize negatively charged citrate-capped gold nanoparticles (cit-AuNPs) solely via electrostatic interactions. The monolayer of electrostatically immobilized cit-AuNPs is further conjugated with the thiolated RGD peptides through the sulfur−gold bond. The substrate coupling strength of the RGD peptides, which is dependent on the electrostatic interactions between the APTES-treated glass substrate and the cit-AuNPs, is simply tuned by changing the APTES dosage and, hence, the resultant positive charge density on the surface. A total of 0.5% and 12.5% of APTES are used to fabricate low-coupling-strength surfaces (namely, LCS0.5 and LCS12.5), whereas 25% and 50% of APTES are used to fabricate high-coupling-strength surfaces (namely, HCS25 and HCS50). Fluorescence microscopy shows that hMSCs spread well and form stable actin filamentous structure on HCS surfaces but not on LCS surfaces. Remarkably, hMSCs exhibit enhanced osteogenesis on HCS surfaces as revealed by the immunostaining results of multiple early osteogenic markers. These differential behaviors may be governed by Yes-associated protein (YAP), a mechanosensitive transcriptional regulator of stem cells. Our findings highlight the importance of the substrate coupling strength of integrin-binding ligands on regulating adhesion, spreading, and differentiation of hMSCs. KEYWORDS: cell−substratum interactions, RGD peptide coupling strength, cell adhesion, cell spreading, stem cell differentiation

T

connected to the FAs. This action further mediates various kinds of cell behaviors including cell lineage specification.13−15 Human mesenchymal stem cells (hMSCs) have been under intensive research as a promising cell source for tissue engineering and regenerative medicine applications, owing to the ease of isolation and the multipotency of hMSCs to differentiate into various lineages including adipocytes, osteoblasts, and chondrocytes.16 The ability of directing stem cells to a specific lineage is crucial to the application of hMSCs in stem-cell-based therapies.17,18 Previous studies have shown that stem cells can probe and dramatically respond to various cues from the physical environment presented by the substrate, thereby altering their growth and ability to differentiate along different lineages. Such important cues include the substrate

he adhesion of cells to the culture environment is crucial to their development, including growth and death,1 cell motility,2 and differentiation.3,4 Such cell−substratum interactions are mediated by transmembrane extracellular matrix (ECM) receptors, which also function as mechanoreceptors that transfer mechanical signals to the cytoskeleton, thereby modulating cellular behaviors via the process known as mechanotransduction.5,6 Integrins are membrane-spanning receptors that play a key role in the signal transduction process by mediating the principal links between cytoskeleton and ECM.7−9 Upon binding to specific extracellular ligands, integrins cluster and multiple intracellular adaptor proteins, such as vinculin, are recruited to form focal adhesions (FAs).10 These assemblies create a physical bridge to the actin cytoskeleton, thus enabling cells to sense mechanical cues present on the substrate.11−13 Throughout the formation and maturation of FAs, cells exert forces at adhesions on the substrate by developing a contractile cytoskeleton that is © XXXX American Chemical Society

Received: June 12, 2015 Revised: September 2, 2015

A

DOI: 10.1021/acs.nanolett.5b02323 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Scheme 1. Schematic Illustration of the Fabrication of Arg−Gly−Asp (RGD)-Coupled Surfaces with Varied Substrate Coupling Strength of the Thiolated Integrin-Binding RGD Peptide

Figure 1. Characterization of the synthesized citrate-capped gold nanoparticles (cit-AuNPs). (a) UV−vis spectra of the prepared cit-AuNPs (red trace) and the RGD-coupled AuNPs (blue trace). The red-shift and slight broadening of the absorption peak maximum possibly indicates the successful conjugation of the thiolated-RGD peptides onto the cit-AuNPs via the sulfur−gold bond. (b) Representative TEM image of the prepared cit-AuNPs shows that the particles are uniform and spherical in shape.

mechanical properties (both stiffness19,20 and viscoelasticity21,22), physical dimension,23 geometry,24,25 ECM composition,26−28 surface topography,29,30 and molecular chirality.31 In essence, the driving forces contributing to the differential cellular responses of stem cells on different substrates mainly rely on the changes in cell adhesion and cytoskeleton structure, which are indeed essential for signal transduction from the extracellular environment into intracellular machineries, as aforementioned. Synthetic integrin-binding ligand Arg−Gly−Asp (RGD)32 is often chemically tethered to the substrate of interest for studying cell−ECM interactions as well as enhancing the celladhesive property of materials.33,34 Extensive studies reveal the pronounced effects of ligand density,35 nanoscale clustering,36,37 and nanospacing of ligands38,39 on adhesion, spreading, focal adhesion dynamics, and motility of mammalian cells. More recently, nanoscale spatial organization of RGD peptides has been shown as a key factor of surface chemistry to direct stem cell fate.28 It is worth pointing out that the coupling strength between the RGD ligand and the substrate, which influences the development of traction forces, may probably

regulate the formation of FAs and, hence, the spreading and contractility of cells grown on the substrate. However, in previous reports, RGD peptides are always grafted to the substrate via a permanent covalent linkage, precluding the tuning of the coupling strength of the RGD on the substrate. As a consequence, developing other means to present the RGD peptide on the substrate with tunable coupling strength will be of great interest in terms of gaining new insights into cell− substrate interactions. In this study, we seek to investigate whether altering the coupling strength of the RGD peptide on the substrate can regulate adhesion and spreading of hMSCs, thus dictating the differentiation pathway of stem cells. To test our hypothesis, we have developed a facile platform to present RGD peptides with varied coupling strength on a glass substrate, as graphically illustrated in Scheme 1. Coverslips are first aminated using different dosages of a common silanization agent called (3-aminopropyl) triethoxysilane (APTES) to generate surfaces with varying positive charge densities. Subsequently, citrate-capped gold nanoparticles (citAuNPs) are immobilized on the silanized glass surfaces, solely via electrostatic interactions between the negative layer of B

DOI: 10.1021/acs.nanolett.5b02323 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 2. Characterization of the RGD-coupled surfaces. (a) Representative scanning electron microscopy (SEM) images of the fabricated surfaces, namely, LCS0.5, LCS12.5, HCS25, and HCS50. As shown in the SEM micrographs, the density of cit-AuNPs electrostatically immobilized on all the fabricated surfaces is apparently constant. In addition, no submonolayer of cit-AuNPs is observed on the surfaces. Scale bar is 1 μm. (b) Representative water contact angle images of the fabricated surfaces. There is no observable difference in the water contact angle among all the surfaces. (c) Representative atomic force microscopy (AFM) images of the fabricated surfaces. Scale bar is 10 μm. Insets show the magnified images of the boxed area being scanned by the AFM cantilever operating at the indicated deflection setpoint voltage (V). Scale bar is 2 μm. Our results show that a higher voltage is required to swipe away the cit-AuNPs immobilized on HCS surfaces than that required for LCS surfaces, indicating that citAuNPs (and thus the RGD peptides) are more stably coupled onto the glass substrate.

AuNPs possess a negative surface charge of −24.7 ± 4.6 mV (Figure S1c, Supporting Information), indicative of the citrate layer.42 This intrinsic property of cit-AuNPs enables their facile deposition on any positively charged surface via electrostatic interactions. Next, we demonstrate the ability of the cit-AuNPs to act as a physical platform to anchor the thiolated RGD peptides (see Materials for the sequence information, Supporting Information) in buffer. We gently mix the citAuNP solution with different concentrations of the thiolated RGD peptide solution for a fixed time and subsequently monitor the conjugation efficiency using UV−vis spectroscopy. After 2 h of incubation with 9 μM of the RGD solution buffered at pH 7.4, we observe a red-shift of ∼1 nm and slight broadening of the λmax of the resultant solution (Figure 1a). These changes possibly reveal the successful coupling of the thiolated RGD peptides onto the cit-AuNPs via the formation of the S−Au bond under these reaction conditions, thereby leading to a change in the local dielectric environment near the nanoparticle surface.43 Silanization is a well-developed method employed to decorate glass substrate with positive charges using modifying agent such as APTES. However, this facile chemistry has rarely been used in stem cell studies. In this work, we attempt to prepare four kinds of RGD-coupled surfaces with different degrees of APTES modification. We first modify glass coverslips with 0.5%, 12.5%, 25%, and 50% of APTES solutions (v/v, in ethanol) in order to produce surfaces with different positive charge densities due to the presence of the primary amino group on each APTES molecule. X-ray photoelectron spectroscopy (XPS) indicates the successful APTES modifica-

citrate ions surrounding the cit-AuNPs and the positive amino groups decorated on the substrate. By carefully controlling the concentration and the size of the cit-AuNPs, a closely packed monolayer of cit-AuNPs can be formed on the glass surfaces after a specific incubation time. The thiolated RGD peptides are then conjugated to the monolayer of cit-AuNPs through the in situ formation of the sulfur−gold (S−Au) bond prior to cell experiments. Hence, the coupling strength of the peptide is regulated solely by the strength of electrostatic interactions between the nanoparticles and the aminated glass substrates, which is in turn governed by the positive charge density on the surfaces. To the best of our knowledge, it is the first study to investigate the influences of the substrate coupling strength of integrin-binding peptides on cellular behaviors of hMSCs. We first synthesize cit-AuNPs of ∼50 nm in diameter by using a previously reported seed-mediated growth method.40 Inductively coupled plasma optical emission spectrometry (ICP-OES) determines that the atomic Au concentration of the prepared cit-AuNP stock solution is 45 μg/mL (Figure S1a, Supporting Information). UV−vis spectroscopy shows a sharp absorption maximum peak (λmax) at 530 nm of the cit-AuNP solution, thus indicating that the cit-AuNPs are monodispersed and sized close to 50 nm (Figure 1a).41 Typical transmission electron microscopy (TEM) images reveal that the synthesized cit-AuNPs are spherical and uniform. Their physical sizes are measured as 48.1 ± 3.1 nm (Figure 1b). Dynamic light scattering (DLS) measurement shows that the hydrodynamic size of the cit-AuNPs is ∼55 nm (Figure S1b, Supporting Information), which is consistent with the TEM data. More importantly, zeta potential measurement confirms that the citC

DOI: 10.1021/acs.nanolett.5b02323 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

quantitative manner, our AFM results support that the fabricated surfaces show increasing substrate coupling strength of the cit-AuNPs (and thus the RGD peptides) with the increasing APTES modification. To investigate the early stage behaviors of hMSCs cultured on the RGD-coupled surfaces, we allow hMSCs to grow in basal medium for 24 h. Cells are treated equally in all experimental groups in which bare coverslips treated with piranha and coated with BSA only (but without any nanoparticle immobilization) are used as a blank control surface throughout the whole study. Cell nuclei and cytoplasmic actin skeletons are visualized under fluorescence microscopy through DAPI and F-actin staining, respectively. First, the adhesive property of the fabricated surfaces is examined by counting the adhesive cell numbers on over ten images taken under a low magnification (i.e., 4×). With the exception of LCS0.5, all fabricated surfaces show enhanced cell adhesion when compared with the control, revealing the successful coupling of the cell-adhesive RGD peptide (Figure S4, Supporting Information). The reduction of cell adhesion for cells grown on LCS0.5 may be caused by that nonstably presented RGD peptides act as free ligands, thus hindering the attachment of hMSCs onto the substrate, as reported previously.47 Strikingly, we observe that hMSCs respond structurally differently to the RGD-coupled surfaces with varied coupling strength. Representative fluorescence images of hMSCs reveal that hMSCs adopt an elongated and spindlelike shape on both LCS0.5 and LCS12.5 surfaces (Figure 3a,b). In contrast, hMSCs are widely spread on both HCS surfaces, especially for cells grown on LCS50 (Figure 3c,d). The virtually significant difference in cell morphology motivates us to quantify the cell morphological characteristics, specifically the aspect ratio (i.e., major/minor axis) as well as the cell adhesive area (per unit area) using ImageJ. In line with our observations, analysis of the aspect ratio confirms that hMSCs grown on LCS surfaces are highly elongated, whereas hMSCs grown on HCS surfaces appear to be more round in shape (i.e., value closer to 1) (Figure 4a). In addition to the aspect ratio data, hMSCs cultured on HCS surfaces possess a statistically larger adhesive cell area compared with cells on both LCS surfaces (Figure 4b). Notably, cells on HCS50 spread to the greatest extent and exhibit a ∼2.5-fold increase in cell spread area relative to cells on LCS0.5. Such increase in cell spread area with increasing RGD coupling strength is similar to the previously reported increase in cell spread area as the substrate rigidity increased.19 It is worth noting that we observe that hMSCs grown on HCS50 show premature actin stress fibers at this early culture stage. This early formation of stable cytoskeletons within 24 h may affect stem cell behaviors, especially differentiation.48 In short, our results reveal that varying the coupling strength of the RGD peptide on the glass substrate shows a simultaneous modulation of both stem cell morphology and spreading. To further confirm that the observed difference in stem cell behaviors is solely attributable to the effects of varied coupling strength of the RGD peptide on the substrate, we perform a series of control experiments in which hMSCs are grown in the same manner as aforementioned. In one negative control experiment, we seed hMSCs onto coverslips modified with varying APTES dosages only. Upon 24 h of culture, we observe in at least two independent experiments that cells adhere and spread better compared to the blank control surface (without the APTES treatment), regardless of the APTES dosage.

tions on all of the surfaces, even for the lowest APTES dosage (i.e., 0.5%) (Figure S2, Supporting Information). Next, the APTES-modified surfaces are electrostatically coated with a monolayer of cit-AuNPs that serves as a platform for the subsequent conjugation of the thiolated-RGD peptides on the substrate. In order to minimize the nonspecific adhesion of cells, we finally immerse the RGD-coupled glass coverslips into the buffered solution of heat-inactivated bovine serum albumin (BSA), a protein that does not support cell adhesion, for 1 h before cell experiments.36 We hypothesize that the coupling strength between the cit-AuNPs (and thus the RGD peptides) and the glass substrate is regulated solely by the resultant positive charge density on the surface. Therefore, we define the surfaces modified with 0.5% and 12.5% of APTES as lowcoupling strength (LCS) surfaces, whereas the surfaces modified with 25% and 50% of APTES are described as highcoupling strength (HCS) surfaces. Although it is difficult to quantitatively determine the exact amount of positive charges on the glass substrates, our cell experiments adequately show significant difference in cellular responses of hMSCs cultured on different RGD-coupled surfaces, as described later in the text. By literature precedent, the lateral spacing of RGD peptide can regulate the stem cell behaviors and direct the differentiation fate.44,45 In order to diminish the potential effects of this cell-regulatory factor in our system, it is essential to keep the density of deposited cit-AuNPs nearly constant for all the RGD-coupled surfaces. Through trial and error, we identify the suitable concentration of the cit-AuNP solution and the appropriate incubation time so that glass surfaces with nearly constant nanoparticle density (per unit area) can be produced, independent of the degree of APTES modification. Representative SEM images clearly show not only that the cit-AuNP density is well controlled but also that a monolayer of citAuNPs is formed on all the fabricated surfaces without the formation of a submonolayer structure (Figure 2a). This ensures that any cellular responses from hMSCs observed in our subsequent experiments are originated from only the effects of varied substrate coupling strength of the RGD peptide but not due to the nanospacing effects. We next characterize the hydrophilic nature of the RGD-coupled surfaces by performing contact angle measurements. Representative water contact angle images show that there is no observable difference in the hydrophilicity among all the fabricated surfaces (Figure 2b). More importantly, we attempt to examine the relative coupling strength between the cit-AuNPs (and thus the RGD peptides) and the different APTES-treated glass substrates by conducting atomic force microscopy (AFM) operating in contact mode.46 In our study, we use the same type of AFM cantilever (with the force spring constant of 0.6 N/m) to scan the nanoparticles immobilized on each surface. We adjust the applied force on each surface by systematically increasing the deflection setpoint voltage (V) from the lowest 1.000 V until a value is reached to clear almost 100% of nanoparticles. This final voltage applied to remove all the nanoparticles, which correlates to the minimal force required to sweep away the particles, is recorded for comparison. As expected, a relatively larger force is needed to remove the cit-AuNPs on HCS surfaces when compared to that on LCS surfaces (Figure 2c), indicating that cit-AuNPs on HCS surfaces are more stably immobilized on the surfaces than that on LCS surfaces. More specifically, an 8-fold voltage increase is required for removing all the particles on HCS50 than that on LCS0.5 (Figure S3, Supporting Information). In a semiD

DOI: 10.1021/acs.nanolett.5b02323 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Furthermore, numbers of adherent cells and degree of cell spreading become higher with the increasing degree of APTES modification (Figure S5, Supporting Information), revealing that increasing positive charge densities on substrates do promote the stem cell adhesion.49 In a follow-up negative control study, we culture hMSCs on glass substrate treated with varying APTES dosages and immobilized with cit-AuNPs but without the conjugation of the RGD peptides. We observe no statistical difference in both aspect ratio and adhesive cell area among all the negative control surfaces, regardless of the amount of surface positive charges. Fluorescence images show similar morphology of hMSCs (as spindle-like) when grown on these control surfaces in the absence of RGD (Figure S6, Supporting Information). Our results indicate that the monolayer of cit-AuNPs not only acts as an anchor platform for the subsequent coupling of the thiolated RGD peptides but also significantly masks the charge effect of APTES-treated surfaces. In the positive control experiment, we decorate glass substrates with different dosages of (3-mercaptopropyl) trimethoxysilane (MPTMS), which is another silanization agent with the structure similar to APTES except that it bears a thiol group (but not an amino moiety). Cit-AuNPs are then covalently linked onto the surfaces via the formation of S− Au bond instead of electrostatic interactions. The thiolated RGD peptides are subsequently conjugated onto the surface of cit-AuNPs via the formation of S−Au bond as well. We hypothesize that hMSCs should show similar responses to the MPTMS-treated surfaces because cit-AuNPs (and thus the RGD peptides) are now covalently attached to the glass substrate. Therefore, no difference in coupling strength should exist. As expected, we observe no significant difference in either adherent cell numbers or cell morphology when hMSCs are seeded onto the MPTMS-treated surfaces (Figure S7, Supporting Information). Cells spread well with the formation of mature actin cytoskeletons on all positive control surfaces, regardless of the MPTMS dosage. Collectively, results obtained from these extensive control experiments support that the distinct responses observed in the hMSCs grown on the RGDcoupled surfaces most likely are originated from the varied coupling strength of the RGD peptide on the glass substrate. Based on our observations, we postulate that hMSCs grown on the surfaces with low-coupling strength of the RGD peptide (i.e., LCS0.5 and LCS12.5) tend to extend their cell bodies to search for other RGD-anchoring points in order to support their adhesion and growth, thereby developing an elongated morphology as previously shown (Figure 3a,b). This is

Figure 3. Representative fluorescence micrographs of the hMSCs cultured on (a) LCS0.5, (b) LCS12.5, (c) HCS25, (d) HCS50, and (e) blank control surface for 24 h. Coverslips treated with piranha only is used as the control for all the cell experiments. In the left panel, white arrows indicate widely spread cells, whereas yellow arrows indicate elongated cells. Scale bar is 200 μm. In the right panel, highmagnification images are taken to show the morphological feature of cells grown on the fabricated surfaces. Scale bar is 100 μm.

Figure 4. Statistical results of the structural responses of the hMSCs cultured on the RGD-coupled surfaces for 24 h. (a) Quantification of aspect ratio of the hMSCs. (b) Quantification of adhesive cell area. Over 100 cells are counted on multiple fluorescence images to obtain the data, which are presented as mean ± sd. *, significant statistical difference compared to other experimental groups (p < 0.01); †, significant statistical difference compared to the blank control surface (p < 0.05); ‡, significant statistical difference compared to the negative control (p < 0.01). E

DOI: 10.1021/acs.nanolett.5b02323 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

that we observe a global effect on the differential responses in stem cell morphology but not a localized effect (Figure S8, Supporting Information). More importantly, we observe appreciable difference in the cytoskeletal assembly between the hMSCs cultured on LCS surfaces and those on HCS surfaces as shown in high-magnification images (Figure S9, Supporting Information). For hMSCs grown on LCS surfaces, no well-defined actin fiber is observed in the cytoplasmic area of the cells. This lack of stress fibers might be a consequence of that the cell adhesion provided by LCS surfaces is too weak to support the development of extensive stress fibers. In sharp contrast, hMSCs grown on HCS surfaces express prominent, ordered, and continuous actin filaments that span over the whole cytoplasm, particularly for cells on HCS50. As previously reported, increasing cell spreading on the substrate promotes the maturation of actin stress fibers.15,24 In line with these findings, our observations strongly correlate a high degree of cell spreading with enhanced assembly of the cytoskeleton of hMSCs. To better understand the difference in the structural responses of the hMSCs, we again quantify cell morphology and spreading. Statistical results reveal the same trend of both the structural parameters as previously obtained from hMSCs cultured for 24 h (Figure 6). Particularly, there is a decreasing trend of the aspect ratio when cells are grown on the surfaces with increasing substrate coupling strength of the RGD peptides, whereas the cell spread area is monotonically increasing with the coupling strength (Figure S10, Supporting Information). These results support our assertion that the varied coupling strength of the RGD peptide on the substrate exhibits significant effects on hMSC behaviors for at least 3 days of culture. Next, we anticipate investigating how the coupling strength of RGD peptide on the substrate is related to cell adhesion and how the effects are responsible for the observed difference in hMSC behaviors. As focal adhesions (FAs) are the centers responsible for cell anchorage and the organization of the actin cytoskeleton, we perform immunofluorescence staining to visualize vinculin, a key anchor protein of FAs. As observed in at least two independent experiments, hMSCs grown on LCS surfaces display very weak vinculin staining (Figure 5a,b, right panel), whereas hMSCs cultured on HCS surfaces show much stronger fluorescence signals that are located near the outer edge of cell membrane as well as the ventral part of cells (Figure 5c,d, right panel). Most importantly, we can observe punctate vinculin clusters in the cells grown on HCS50 both in the perinuclear region and at the cell periphery. Accordingly, we suggest that the varied coupling strength of the RGD peptide on the substrate may have contributed to the observed difference in the vinculin distribution of the cell samples in which mature FAs can be formed only when the RGD peptides are stably coupled to the substrate. Extensive studies show that cell morphology as well as cell size dictate the commitment of stem cell differentiation.15,23,51,52 Our results indicate that hMSCs grown on the fabricated surfaces with varied coupling strength of the RGD peptide on the substrate display pronounce difference in cell morphology, spreading, and actin assembly. Therefore, it is highly possible that hMSCs may show altered differentiation potential when cultured on LCS surfaces versus HCS surfaces. As a proof-of-concept work, we choose osteogenic differentiation as a model to study because osteogenesis of stem cells is sensitive to morphological changes and cell spreading. We

consistent with a recent study on stem cells using titanium oxide (TiO2) nanotubes as the substrate which shows that hMSCs grown on TiO2 nanotubes of a bigger diameter (i.e., 70 and 100 nm) elongate and extend their filopodia to search for protein agglomerates which facilitate cell adhesion.50 In our case, we also observe that hMSCs cultured on LCS surfaces start to produce thin and long filopodia, analogous with the results of the previous report. In order to determine whether the observed cellular responses of stem cells toward the RGD-coupled surfaces will last for a longer time of culture, we allow hMSCs to grow on the fabricated surfaces for 3 days and subsequently stain the cells as before. Remarkably, hMSCs still exhibit different responses toward our fabricated surfaces. As recorded by a fluorescence microscope, hMSCs grown on LCS0.5 further elongate like a spindle while cells on LCS12.5 start to spread sparsely (Figure 5a,b). Substantially, hMSCs grown on both HCS surfaces spread profusely (Figure 5c,d). It is worth noting

Figure 5. High-magnification fluorescence micrographs of the hMSCs cultured on (a) LCS0.5, (b) LCS12.5, (c) HCS25, (d) HCS50, and (e) blank control surface for 3 days. Left panel shows the overlaid images of the hMSCs stained for DAPI (blue), F-actin (red), and vinculin (green). Right panel shows the corresponding vinculin channel of the hMSCs. White arrows indicate punctate vinculin clusters of the cells grown on HCS surfaces. Scale bar is 50 μm. F

DOI: 10.1021/acs.nanolett.5b02323 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 6. Statistical results of the structural responses of the hMSCs cultured on the RGD-coupled surfaces for 3 days. (a), (b) Quantification of shape factor and adhesive cell area of hMSCs grown on the fabricated surfaces, respectively. Over 100 cells are counted on multiple fluorescence images to obtain the data, which are presented as mean ± sd. Symbols are used identically as previously shown in Figure 4.

allow hMSCs to differentiate on the RGD-coupled surfaces in osteogenic induction medium for 7 days following 1 day of adhesion. Afterward, we fix and then immunostain the cells against runt-related transcription factor 2 (RUNX2), which is the key early stage osteogenic marker.53 Strikingly, hMSCs cultured on HCS surfaces display much higher fluorescence signals of nuclear RUNX2 compared to cells cultured on LCS surfaces (Figure 7, left panel), thus indicating enhanced differentiation ability toward osteogenesis of hMSCs grown on HCS surfaces. To support our observations, we carry out ratiometric analysis of the nuclear and cytoplasmic fluorescence of RUNX2 as well as calculate the percentage of hMSCs expressing RUNX2 over 500 cells. Consistently, we obtain a monotonically increasing trend for both the relative fluorescence intensity of nuclear RUNX2 and the percentage of positively stained hMSCs when cells are cultured on the RGDcoupled surfaces with increasing coupling strength (Figure S11, Supporting Information). To provide additional evidence of the distinct osteogenic differentiation of the hMSCs cultured on the fabricated substrates, we further stain the cell samples against alkaline phosphatase (ALP), which is another early marker for osteogenesis.53 Representative bright-field micrographs reveal that hMSCs differentiated on HCS surfaces show significant positive staining of ALP, whereas cells on LCS surfaces show limited and weak cell staining (Figure 7, right panel). This apparent trend is consistent with the immunofluorescence staining results of RUNX2. Particularly, hMSCs grown on HCS50 exhibit the strongest staining for both RUNX2 and ALP. Taken together, our staining results support that the higher coupling strength of the RGD peptide on the substrate promotes osteogenesis of hMSCs. These findings align well with the significant difference observed in the morphology and cytoskeleton assembly of hMSCs grown on different RGDcoupled surfaces, considering that osteogenic cells are typically more widely spread15 and have extensive stress fibers associated with actomyosin contractility.48 Accordingly, we suggest that a high coupling strength of integrin-binding ligands on substrate favors osteogenic differentiation of hMSCs. Yes-associated protein (YAP) has newly emerged as a sensor and mediator of mechanical cues presented by the cellular microenvironment.54 The activity of YAP has known to be regulated by substrate stiffness, cell spreading, and cytoskeletal tension. Furthermore, osteogenic differentiation induced in stem cells has been shown to be positively correlated with YAP activity. In light of these findings, we anticipate that YAP may be a key regulator that is responsible for the observed difference in stem cell behaviors when grown on the RGD-coupled

Figure 7. Examination of the early stage osteogenic markers in the hMSCs cultured on the RGD-coupled surfaces upon 7 days of osteogenic induction. Left panel shows the representative fluorescence micrographs of the RUNX2-stained hMSCs differentiated on (a) LCS0.5, (b) LCS12.5, (c) HCS25, (d) HCS50, and (e) blank control surface. Right panel shows the corresponding bright-field micrograph of the ALP staining for the same cell sample. Scale bar is 200 μm.

surfaces. To elucidate the role of YAP, we again induce osteogenesis of the hMSCs for 7 days following 1 day of adhesion on the RGD-coupled surfaces and subsequently G

DOI: 10.1021/acs.nanolett.5b02323 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

In this work, we have developed a facile platform to study several effects of the substrate coupling strength of integrinspecific peptides on growth and differentiation of hMSCs. A well-known integrin-binding ligand RGD is used in this proofof-concept study and its coupling strength on the glass substrate is controlled solely by electrostatic interactions between the positively charged aminated glass surface and the negatively charged cit-AuNPs. Notably, hMSCs show differential structural responses in terms of cell morphology and spreading when grown on the RGD-coupled surfaces with varied substrate coupling strength. HCS surfaces promote adhesion as well as spreading of the seeded hMSCs in both short- and long-term culture (i.e., 24 h and 3 d, respectively). In the presence of induction factors, staining of two osteogenesisrelated markers collectively indicate that hMSCs grown on HCS surfaces exhibit enhanced differentiation ability toward osteogenesis. The distinct stem cell responses may be in part governed by YAP, a newly emerged mechanosensitive transcriptional regulator of stem cells. Our results show, for the first time, that varying the coupling strength of the RGD peptide on the substrate modulates adhesion, spreading, and differentiation of hMSCs. This work not only highlights a new physical parameter that has to be carefully considered when designing new biomaterials for stem-cell-based therapies but also opens up an avenue for generating new classes of cell-interactive materials with tunable coupling strength of cell-adhesive ligands.

immunostain the cells against YAP. It is worth noting that hMSCs seem to maintain their distinct morphology on different RGD-coupled surfaces even after 7 days of differentiation (Figure 8, left panel). Notably, the hMSCs differentiated on

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.5b02323. Materials and methods, characterization of the synthesized cit-AuNPs, X-ray photoelectron spectra of the RGD-coupled surfaces, additional statistical results of cell adhesion and structural responses, examination of the substrate coupling strength of immobilized cit-AuNPs, representative fluorescence micrographs of the hMSCs cultured on the control surfaces, low- and highmagnification fluorescence micrographs of the hMSCs cultured on the fabricated surfaces for 3 days, and quantification of RUNX2 and YAP staining of the hMSCs cultured on the fabricated surfaces for 7 days. (PDF)

Figure 8. Examination of the YAP expression in the hMSCs cultured on the RGD-coupled surfaces upon 7 days of osteogenic induction. Left panel shows the representative fluorescence micrographs of the YAP-stained hMSCs differentiated on (a) LCS0.5, (b) LCS12.5, (c) HCS25, (d) HCS50, and (e) blank control surface. Right panel shows the corresponding YAP channel of the hMSCs. Scale bar is 100 μm.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: +852 3943 8342. Address: Room 213 William Mong Engineering Building, The Chinese University of Hong Kong, Shatin, Hong Kong, China.

HCS surfaces show brighter nuclear YAP fluorescence than the cells on LCS surfaces as well as those on the blank control surface (Figure 8, right panel). Ratiometric analysis of the nuclear and cytoplasmic fluorescence of YAP consistently show an increment in nuclear fluorescence signals from cells cultured on LCS surfaces to that on HCS surfaces as quantified by using ImageJ (Figure S12, Supporting Information). Our results are in good agreement with the findings that stem cells with a higher degree of spreading and more prominent stress fibers exhibit increased YAP level.54 These data indicate that the effects of varied substrate coupling strength between the RGD peptide and the glass substrate on stem cell behaviors may be governed in part by YAP as a mechanosensitive transcriptional regulator.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We are grateful for support from the Research Grants Council of Hong Kong (Project No. 439913) through an Early Career Scheme grant, the National Natural Science Foundation of China for the project 31300796, the Shun Hing Institute of Advanced Engineering (The Chinese University of Hong Kong) for the project BME-8115043, and the Chow Yuk Ho Technology Centre for Innovative Medicine (The Chinese H

DOI: 10.1021/acs.nanolett.5b02323 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

(33) Pizzo, A. M.; Kokini, K.; Vaughn, L. C.; Waisner, B. Z.; VoytikHarbin, S. L. J. Appl. Physiol. 2005, 98, 1909−1921. (34) Massia, S. P.; Hubbell, J. A. Ann. N. Y. Acad. Sci. 1990, 589, 261−70. (35) Deeg, J. A.; Louban, I.; Aydin, D.; Selhuber-Unkel, C.; Kessler, H.; Spatz, J. P. Nano Lett. 2011, 11, 1469−76. (36) Koo, L. Y.; Irvine, D. J.; Mayes, A. M.; Lauffenburger, D. A.; Griffith, L. G. J. Cell Sci. 2002, 115, 1423−33. (37) Huang, J.; Grater, S. V.; Corbellini, F.; Rinck, S.; Bock, E.; Kemkemer, R.; Kessler, H.; Ding, J.; Spatz, J. P. Nano Lett. 2009, 9, 1111−6. (38) Cavalcanti-Adam, E. A.; Volberg, T.; Micoulet, A.; Kessler, H.; Geiger, B.; Spatz, J. P. Biophys. J. 2007, 92, 2964−74. (39) Wang, X.; Yan, C.; Ye, K.; He, Y.; Li, Z.; Ding, J. Biomaterials 2013, 34, 2865−74. (40) Grabar, K. C.; Brown, K. R.; Keating, C. D.; Stranick, S. J.; Tang, S. L.; Natan, M. J. Anal. Chem. 1997, 69, 471−7. (41) Bastus, N. G.; Comenge, J.; Puntes, V. Langmuir 2011, 27, 11098−105. (42) Brewer, S. H.; Glomm, W. R.; Johnson, M. C.; Knag, M. K.; Franzen, S. Langmuir 2005, 21, 9303−7. (43) Levy, R.; Thanh, N. T.; Doty, R. C.; Hussain, I.; Nichols, R. J.; Schiffrin, D. J.; Brust, M.; Fernig, D. G. J. Am. Chem. Soc. 2004, 126, 10076−84. (44) Frith, J. E.; Mills, R. J.; Cooper-White, J. J. J. Cell Sci. 2012, 125, 317−27. (45) Wang, X.; Li, S.; Yan, C.; Liu, P.; Ding, J. Nano Lett. 2015, 15, 1457−67. (46) Eppler, A. S.; Rupprechter, G.; Anderson, E. A.; Somorjai, G. A. J. Phys. Chem. B 2000, 104, 7286−7292. (47) Zhang, Z.; Ni, J.; Chen, L.; Yu, L.; Xu, J.; Ding, J. J. Biomed. Mater. Res., Part B 2012, 100B, 1599−1609. (48) Rodriguez, J. P.; Gonzalez, M.; Rios, S.; Cambiazo, V. J. Cell. Biochem. 2004, 93, 721−731. (49) Lai, Y.; Xie, C.; Zhang, Z.; Lu, W.; Ding, J. Biomaterials 2010, 31, 4809−17. (50) Oh, S.; Brammer, K. S.; Li, Y. S.; Teng, D.; Engler, A. J.; Chien, S.; Jin, S. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 2130−5. (51) Yao, X.; Peng, R.; Ding, J. Biomaterials 2013, 34, 930−9. (52) Peng, R.; Yao, X.; Cao, B.; Tang, J.; Ding, J. Biomaterials 2012, 33, 6008−19. (53) Komori, T. J. Cell. Biochem. 2006, 99, 1233−9. (54) Dupont, S.; Morsut, L.; Aragona, M.; Enzo, E.; Giulitti, S.; Cordenonsi, M.; Zanconato, F.; Le Digabel, J.; Forcato, M.; Bicciato, S.; Elvassore, N.; Piccolo, S. Nature 2011, 474, 179−83.

University of Hong Kong). This work is also supported by the Jiangsu Provincial Special Program of Medical Science (BL2012004).

■

REFERENCES

(1) Chen, C. S.; Mrksich, M.; Huang, S.; Whitesides, G. M.; Ingber, D. E. Science 1997, 276, 1425−1428. (2) Palecek, S. P.; Loftus, J. C.; Ginsberg, M. H.; Lauffenburger, D. A.; Horwitz, A. F. Nature 1997, 385, 537−40. (3) Dike, L. E.; Chen, C. S.; Mrksich, M.; Tien, J.; Whitesides, G. M.; Ingber, D. E. In Vitro Cell. Dev. Biol.: Anim. 1999, 35, 441−8. (4) Yao, X.; Peng, R.; Ding, J. Adv. Mater. 2013, 25, 5257−86. (5) Wang, N.; Butler, J. P.; Ingber, D. E. Science 1993, 260, 1124−7. (6) Wang, N.; Naruse, K.; Stamenovic, D.; Fredberg, J. J.; Mijailovich, S. M.; Tolic-Norrelykke, I. M.; Polte, T.; Mannix, R.; Ingber, D. E. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 7765−70. (7) Hynes, R. O. Cell 1987, 48, 549−54. (8) Hynes, R. O. Cell 1992, 69, 11−25. (9) Humphries, J. D.; Byron, A.; Humphries, M. J. J. Cell Sci. 2006, 119, 3901−3903. (10) Critchley, D. R. Curr. Opin. Cell Biol. 2000, 12, 133−9. (11) Geiger, B.; Bershadsky, A.; Pankov, R.; Yamada, K. M. Nat. Rev. Mol. Cell Biol. 2001, 2, 793−805. (12) Geiger, B.; Spatz, J. P.; Bershadsky, A. D. Nat. Rev. Mol. Cell Biol. 2009, 10, 21−33. (13) Balaban, N. Q.; Schwarz, U. S.; Riveline, D.; Goichberg, P.; Tzur, G.; Sabanay, I.; Mahalu, D.; Safran, S.; Bershadsky, A.; Addadi, L.; Geiger, B. Nat. Cell Biol. 2001, 3, 466−72. (14) Riveline, D.; Zamir, E.; Balaban, N. Q.; Schwarz, U. S.; Ishizaki, T.; Narumiya, S.; Kam, Z.; Geiger, B.; Bershadsky, A. D. J. Cell Biol. 2001, 153, 1175−86. (15) McBeath, R.; Pirone, D. M.; Nelson, C. M.; Bhadriraju, K.; Chen, C. S. Dev. Cell 2004, 6, 483−95. (16) Pittenger, M. F.; Mackay, A. M.; Beck, S. C.; Jaiswal, R. K.; Douglas, R.; Mosca, J. D.; Moorman, M. A.; Simonetti, D. W.; Craig, S.; Marshak, D. R. Science 1999, 284, 143−147. (17) Caplan, A. I. J. Cell. Physiol. 2007, 213, 341−7. (18) Abdallah, B. M.; Kassem, M. Gene Ther. 2008, 15, 109−16. (19) Engler, A. J.; Sen, S.; Sweeney, H. L.; Discher, D. E. Cell 2006, 126, 677−89. (20) Saha, K.; Keung, A. J.; Irwin, E. F.; Li, Y.; Little, L.; Schaffer, D. V.; Healy, K. E. Biophys. J. 2008, 95, 4426−38. (21) Cameron, A. R.; Frith, J. E.; Cooper-White, J. J. Biomaterials 2011, 32, 5979−93. (22) Cameron, A. R.; Frith, J. E.; Gomez, G. A.; Yap, A. S.; CooperWhite, J. J. Biomaterials 2014, 35, 1857−68. (23) Connelly, J. T.; Gautrot, J. E.; Trappmann, B.; Tan, D. W.; Donati, G.; Huck, W. T.; Watt, F. M. Nat. Cell Biol. 2010, 12, 711−8. (24) Kilian, K. A.; Bugarija, B.; Lahn, B. T.; Mrksich, M. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 4872−7. (25) Peng, R.; Yao, X.; Ding, J. Biomaterials 2011, 32, 8048−57. (26) Trappmann, B.; Gautrot, J. E.; Connelly, J. T.; Strange, D. G.; Li, Y.; Oyen, M. L.; Cohen Stuart, M. A.; Boehm, H.; Li, B.; Vogel, V.; Spatz, J. P.; Watt, F. M.; Huck, W. T. Nat. Mater. 2012, 11, 642−9. (27) Wen, J. H.; Vincent, L. G.; Fuhrmann, A.; Choi, Y. S.; Hribar, K. C.; Taylor-Weiner, H.; Chen, S.; Engler, A. J. Nat. Mater. 2014, 13, 979−87. (28) Ye, K.; Wang, X.; Cao, L.; Li, S.; Li, Z.; Yu, L.; Ding, J. Nano Lett. 2015, 15, 4720−9. (29) Kantawong, F.; Burgess, K. E.; Jayawardena, K.; Hart, A.; Burchmore, R. J.; Gadegaard, N.; Oreffo, R. O.; Dalby, M. J. Biomaterials 2009, 30, 4723−31. (30) Yim, E. K.; Darling, E. M.; Kulangara, K.; Guilak, F.; Leong, K. W. Biomaterials 2010, 31, 1299−306. (31) Yao, X.; Hu, Y.; Cao, B.; Peng, R.; Ding, J. Biomaterials 2013, 34, 9001−9. (32) Ruoslahti, E. Annu. Rev. Cell Dev. Biol. 1996, 12, 697−715. I

DOI: 10.1021/acs.nanolett.5b02323 Nano Lett. XXXX, XXX, XXX−XXX