- Email: [email protected]
Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling
Journal of Biomechanics ] (]]]]) ]]]–]]]
Contents lists available at SciVerse ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling Igor Titushkin, Shan Sun, Amit Paul, Michael Cho n Department of Bioengineering, University of Illinois, Chicago, IL 60607, United States
a r t i c l e i n f o
abstract
Article history: Accepted 29 September 2012
We have recently shown that altered stem cell biomechanics can regulate the lineage commitment through a family of the membrane–cytoskeleton linker proteins (ERM; ezrin, radixin, moesin). The ERM proteins not only modulate the cell stiffness and actin cytoskeleton organization, but also rearrange focal adhesions and therefore influence the biochemically-directed stem cell differentiation. Combining silencing RNA, atomic force microscopy, and fluorescence microscopy, the role of the ERM proteins involved in the regulation of stem cell biomechanics and adipogenic differentiation was quantitatively determined. Transient ERM knockdown by RNAi caused disassembly of actin stress fibers and focal adhesions and a decrease in the cell stiffness. The silencing RNA treatment not only induced mechanical changes in stem cells but impaired adipogenesis in a time-dependent manner. While siRNA ERM treatment at day 0 substantially interfered with adipogenesis, the same treatment at day 3 of adipogenic differentiation significantly facilitated adipogenesis, as assessed by the expression of adipocyte-specific markers. The intact biomechanics homeostasis appears to be critical for the adipogenic induction. These findings may lead to potential biomechanical intervention techniques and methodologies to control the fate and extent of adipogenesis that would likely be involved in stem cell-based therapeutics for soft tissue repair and regeneration. & 2012 Published by Elsevier Ltd.
Keywords: Adipogenesis ERM proteins Stem cell biomechanics SiRNA AFM
1. Introduction The important role of local biomechanical microenvironment in the stem cell activities and metabolism has long been recognized (Guilak et al., 2009). Physical factors including extracellular matrix (ECM), cell stiffness, cell seeding density and morphology and cytoskeleton tension have all been shown to regulate mesenchymal stem cell commitment to a particular lineage (Engler et al., 2006; McBeath et al., 2004). We have previously demonstrated that human mesenchymal stem cells (hMSCs) exhibit unique biomechanical properties as compared to terminally differentiated specialized cells (Titushkin and Cho, 2006, 2007, 2009, 2011; Titushkin et al., 2010a, 2010b). Several predominant features found in hMSCs include unorganized but thick actin stress fibers and weak membrane–cytoskeleton coupling (see Titushkin and Cho, 2011). While actin stress fibers are largely responsible for the cell stiffness, strong membrane–cytoskeleton coupling adhesion would maintain structural integrity of loadbearing cells (e.g., musculoskeletal cells) subjected to multiple stress cycles (Sheetz et al., 2006). In contrast, a relatively loose n Correspondence to: Department of Bioengineering, University of Illinois, Chicago, 851 S. Morgan St. (M/C 063), Chicago, IL 60607, United States. Tel.: þ1 312 413 9424; fax: þ 1 312 996 5921. E-mail address: [email protected] (M. Cho).
membrane attachment to actin cytoskeleton observed in hMSCs (Titushkin and Cho, 2006) would be expected to facilitate signaling through exo- and endocytosis and cross-membrane trafficking (Jena, 2007). A family of three closely related linker proteins (ezrin, radixin and moesin; ERM) serves as general cross-linkers between the plasma membrane and actin filaments in many mesodermderived cell types (Fie´vet et al., 2007; Louvet-Valle´e, 2000). The amino-terminal half of the ERM proteins are responsible for their interaction with the integral proteins of the plasma membrane, including CD44, ICAM, EBP50 (Mangeat et al., 1999). The carboxyterminal half of the ERM linkers is involved in their interaction with actin filaments. It appears that ERM proteins probably control the cell mechanics both directly by providing the membrane–cytoskeletal linkage (Gautreau et al., 2002; Titushkin et al., 2010a), and indirectly by regulating the actin structure and cell adhesion through Rho-dependent mechanisms (Kotani et al., 1997; Poullet et al., 2001). Our recent results show that, when hMSCs differentiate into osteoblasts, their actins reorganization from sparse thick bundles into a dense thin microfilament network provides more binding sites for ERM linkers, increasing the membrane–cytoskeleton interaction in osteogenically committed hMSCs (Titushkin and Cho, 2011). Significant changes in the cellular mechanics during differentiation suggests that external physical, chemical or genetic manipulation of cellular mechanics
0021-9290/$ - see front matter & 2012 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.jbiomech.2012.09.027
Please cite this article as: Titushkin, I., et al., Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling. Journal of Biomechanics (2012), http://dx.doi.org/10.1016/j.jbiomech.2012.09.027
2
I. Titushkin et al. / Journal of Biomechanics ] (]]]]) ]]]–]]]
0
1
3
7
11
13
17
Day Fig. 1. Images of hMSCs undergoing adipogenesis. F-actins (red), lipids (green) and nuclei (blue) were fluorescently labeled and visualized using rhodamine-phalloidin, LipidTox and DAPI, respectively, at different days of adipogenic differentiation.
could be used to enhance or suppress stem cell induction into the pre-selected phenotype by soluble inductive factors. The ERM linkers could be promising targets for manipulation of the cell mechanics due to their important role in the regulation of both the membrane and the cytoskeleton mechanics. In the present study we applied RNA interference technique to transiently down-regulate of the ERM proteins and modulated the mechanical properties of stem cells undergoing the adipogenic commitment. Since the adipocyte morphology is clearly distinguishable from that of osteoblasts, control of adipogenesis via ERM proteins could shed further insights into the lineage-dependent biomechanical remodeling of stem cells. Indeed, unlike osteogenic differentiation of hMSCs, the extent of adipogenesis was dependent on the time of siRNA transfection. We provide mechanical and biological evidence that the initial biomechanics homeostasis is crucially important for stem cells to respond to differentiation cues such as soluble factors.
2. Materials and methods 2.1. Cell culture and differentiation Human mesenchymal stem cells were obtained from the Tulane Center for Gene Therapy (New Orleans, LA). Based on the flow cytometry results, these cells showed negative staining for CD34, CD36, CD45, and CD117 markers (all less than 2%), and positive staining for CD44, CD90, CD166, CD29, CD49c, and CD105 markers (all more than 95%). The cells were grown in Dulbecco’s modified Eagle’s medium (growth medium, Invitrogen, Carlsbad, CA) with 15% fetal bovine serum, and passages between 3 and 10 were used for all experiments. For adipogenic differentiation, cells were incubated in the adipogenic induction media that contains 1 mM dexamethasone, 10 mg/ml insulin, 200 mM indomethacin and 0.5 mM isobutylmethylxantine.
2.2. siRNA transfection quantitative RT-PCCR siRNA targeting ezrin, radixin and moesin were obtained as a pool of 3 targetspecific siRNA duplexes from Santa Cruz Biotechnology (Santa Cruz, CA). Control siRNA consisted of a scrambled sequence that did not lead to specific degradation of any known cellular mRNA. hMSCs were treated for 5 h with 60 nM ERM siRNA or control siRNA premixed with Transfection reagent (Santa Cruz Biotechnology) in serum-free medium. Cells were then incubated in the adipogenic or normal growth medium. As control scrambled sequence siRNA is not expected to lead to specific degradation of any known cellular mRNA, it did not change the cellular ERM expression levels. In 2 days after siRNA transfection, 490% of ERM siRNAtreated cells were found viable as compared to 96% viability in cells incubated in growth medium without any treatment (Titushkin and Cho, 2011). Treatment of cells with siRNA reduced the ERM proteins’ expression by 1.6-fold. Real-time PCR analysis was performed using 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA). After an initial incubation at 50 1C for 2 min, and denaturation at 95 1C for 10 min, 40 cycles of PCR amplification (95 1C for 15 s, 60 1C for 30 s) were carried out. The data is presented as the fold change in the target gene expression normalized to the endogenous reference gene (GAPDH) and relative to the untreated control. Each reaction was performed in triplicate and the entire RT-PCR analysis was repeated twice for each data point.
2.3. Immunostaining For immunofluorescence studies, the samples were fixed in 3.7% formaldehyde and permeabilized in cold ( 20 1C) acetone. Nonspecific binding sites were blocked using a 3% bovine serum albumin. F-actins were stained with 5 mM rhodaminephalloidin (Invitrogen). ERM proteins were labeled with rabbit ERM antibody (Cell Signaling Technology, Danvers, MA), and secondary antibody conjugated with AlexaFluor488 dye (Invitrogen). Focal adhesions were labeled using mouse antibody to vinculin (Chemicon, Temecula, CA), and secondary fluorescent antibody conjugated with FITC. Samples were imaged by a laser scanning confocal system (Radiance 2100MP, Bio-Rad, Hercules, CA) using a 60 objective lens (NA 1.4). 2.4. Cell stiffness measurements The live cell elasticity was measured with a Novascan atomic force microscope (AFM, Novascan Technologies, Ames, IA) as described earlier (Titushkin and Cho, 2007). To minimize the effect of glass substrate on the cell elasticity measurements, we used an indentation depth up to 500 nm ( 10–15% of the average cell height) for data analysis. A total of 70–100 cells of each condition were measured and analyzed according to the Hertz model (Radmacher, 2002). 2.5. Quantification of lipid droplets Two different techniques were applied to assess formation of lipid droplets. Cells were washed, fixed in 3.7% formaldehyde, permeabilized with 0.1% saponin, and lipids were fluorescently labeled using LipidTox (1:200 , Invitrogen) for 30 min at room temperature. For quantitative measurements, cells were stained with 0.3% Oil Red O dye (Sigma-Aldrich), and Oil Red O was extracted from the cells with isopropanol. Absorbance was measured at 500 nm.
3. Results 3.1. Biomechanical remodeling during adipogenesis Using fluorescence microscopy, changes in the stem cell cytoskeleton and morphology were monitored. Representative images of actin remodeling and appearance of lipid droplets were recorded as a function of time (Fig. 1). As expected, hMSCs undergo rapid morphological changes to assume round shape and reduction in the cell size, as demonstrated by rhodamine-labeled actins. More strikingly, rounded cellular morphology appeared clear at day 1 of adipogenesis along with signals from the fluorescently labeled lipids. By day 7 the intracellular space was filled with lipid droplets. Implication could be that the cell rounding preceded the formation of lipid droplets, and that biomechanics remodeling may not be simply a consequence but necessary cellular event for adipogenesis. This postulate was quantitatively supported, in part, by measuring the hMSCs’ elastic moduli using AFM microindentation (Fig. 2). The cell stiffness decreased exponentially (t1/2 ¼3.16 days), and no further significant decrease was observed after 7 days of adipogenic differentiation. However, while the lack of changes in the hMSCs’ biomechanical properties after this time point was evident, lipid droplets continued to accumulate inside the cells.
Please cite this article as: Titushkin, I., et al., Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling. Journal of Biomechanics (2012), http://dx.doi.org/10.1016/j.jbiomech.2012.09.027
I. Titushkin et al. / Journal of Biomechanics ] (]]]]) ]]]–]]]
Elastic Modulus (kPa)
5 *
4
3
2
1
0 0
1
2
3 7 Time (Day)
13
15
Fig. 2. Time-dependent changes of cell stiffness. AFM indentation was performed at selected days of adipodifferentiation and the cell stiffness was quantified by measuring the elastic modulus using the Hertz model (Titushkin and Cho, 2007). The elastic moduli were fitted to a single exponential function (solid line drawn) with the characteristic half time of 3.16 days. Each value represents mean 7 SEM from 70 to 100 microindentation experiments using 20–30 cells. The cell stiffness measurements were statistically different from the cells at day 0 (po 0.05), except those marked with an asterisk (*).
3.2. Role of ERM proteins The total level of ERM proteins, i.e. the 80 kDa band (corresponding to ezrin and radixin) and the 75 kDa band (corresponding to moesin) was obtained by densiometric measurements. Control siRNA consisted of a scrambled sequence that did not lead to specific degradation of any known cellular mRNA. Confirmed by Western blot analysis, the siRNA transfection reduced the ERM protein expression by 1.6 fold (Titushkin and Cho, 2011). A similar level of RNAi-mediated ERM downregulation was also observed in cells incubated both in the growth or adipogenic media, suggesting that the soluble factors did not interfere with inhibition of the ERM gene expression. Adipogenic commitment of hMSCs by the soluble factors changed both the actin cytoskeleton organization and redistributed the ERM proteins (Fig. 3). Actin, focal adhesion (identified by vinculin) and ERM proteins were immunostained and fluorescently visualized for hMSCs undergoing adipogenesis for 7 days. The thick actin stress fibers were replaced with shorter actin filaments (Fig. 3A). Besides, large focal adhesions typically found in normal stem cells (Chen et al., 2007; Titushkin and Cho, 2009) were substituted with much smaller adhesion complexes. The ERM proteins were observed to correlate with morphological remodeling and concentrated near the membrane (Fig. 3B). Cell treatment with siRNA ERM prior to adipogenic induction (day 0) interfered with adipogenically-mediated actin reorganization and still retained short but thick bundles of actin filaments (Fig. 3C), and stained ERM proteins distributed along the cell periphery (Fig. 3D). In contrast, the same siRNA treatment applied at day 3 of the induction caused even higher density of short actin filaments (Fig. 3E) than in cells undergoing biochemical adipogenic differentiation but not treated with siRNA, and the ERM distribution (Fig. 3F) also resembled that found in normally differentiating hMSCs. These images suggest that the dynamics of the biomechanical remodeling may be dependent on the time of siRNA ERM transfection. As expected, the siRNA ERM transfection modulated the dynamics of adipogenically-induced mechanical changes in hMSCs. For example, RNAi treatment either on day 0 or on day 3 caused an immediate decrease in the elastic modulus. However, after 12 days of induction in the adipogenic medium, no statistically significant differences in the cell elasticity
3
were observed between siRNA-treated and untreated but otherwise differentiating to adipocytes (Fig. 4). To validate whether biochemically-induced adipogenesis depended on the initial biomechanics homeostasis, triglyceriderich vesicles of lipid droplets were visualized and quantified using fluorescence imaging and Oil Red O cytological staining (Fig. 5). After 12 days of induction by the adipogenic soluble factors, lipid droplets were observed in the cytoplasm of at least 70% of cells that were committed to the adipogenic lineage. Cells incubated in the growth media without the adipogenic soluble factors expressed a baseline expression for lipid droplets (Fig. 5A) and, in response to the soluble factors, it was significant increased (Fig. 5B). However, the siRNA ERM treatment before adipogenic induction (day 0) resulted in far fewer lipid droplets (Fig. 5C), whereas the same treatment on day 3 of adipogenic induction increased the formation of lipid droplets (Fig. 5D) that was even greater than the positive control (Fig. 5B). These results were quantitatively confirmed by measuring the absorbance of Oil Red O dye extracted from the cells. The early (i.e., day 0) ERM knockdown in adipo-differentiating hMSCs decreased the Oil Red O absorbance by approximately 2-fold in comparison with non-treated adipogenic cells. In contrast, the late (i.e., day 3) ERM knockdown significantly increased Oil Red O staining in adipogenically committed hMSCs. The Oil Red O absorbance for both siRNA-treated and untreated adipo-differentiating cells was higher than for negative control cells in the growth medium. Finally, using quantitative RT-PCR amplification, adipocytespecific gene expression was measured. The mRNA levels of PPARg, LPL and aP2 were significantly increased by a factor of 6.0, 3.8, and 2.8, respectively, in adipo-differentiating cells as compared to hMSCs in the growth medium (Fig. 6). The siRNA ERM treatment on day 0 suppressed the three genes’ expressions. However, consistent with fluorescence images and Oil Red O staining, the ERM proteins knockdown at the later stage of adipodifferentiation (day 3) further increased the expression of these genes by day 12 of adipogenic differentiation.
4. Discussion We have recently proposed a model to characterize the biomechanical state of the cell (Titushkin et al., 2010a, 2010b). While this model is not limited to stem cells, we were able to elucidate unique biomechanical properties of bone marrow-derived hMSCs. As expected, the cell stiffness is predominantly dictated by the actin organization and regulated by the ERM proteins. By suppressing the ERM proteins’ expressions by siRNA treatment, the cell stiffness can be modulated. However, when the intended differentiation is completed, the cell elasticity in the siRNA-treated or untreated cells did not differ significantly. In contrast, the tissue-specific protein and gene expression levels were different and much more depended on the siRNA treatment. These findings suggest the following. First, the cellular mechanical parameters alone may be insufficient to accurately characterize the stem cell differentiation and supports the notion that the stem cell would have to detect, process, and integrate multiple endogenous and exogenous mechanical and physicochemical signals to determine its response and lineage specification (Guilak et al., 2009). Second, the dynamics of biomechanical remodeling is important for the stem cell commitment and differentiation. In adipogenesis, the cell stiffness is essentially the same by day 12 of adipogenesis, independent of the siRNA treatment. However, interference with the stem cell biomechanics significantly prevented or facilitated adipogenesis, indicating that preserving the initial cellular mechanical homeostasis is required for proper stem cell differentiation. It is evident that biomechanics of stem cells is involved in the regulation of stem cell responses to the soluble factors, potentially
Please cite this article as: Titushkin, I., et al., Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling. Journal of Biomechanics (2012), http://dx.doi.org/10.1016/j.jbiomech.2012.09.027
4
I. Titushkin et al. / Journal of Biomechanics ] (]]]]) ]]]–]]]
Elastic Modulus (kPa)
Fig. 3. Effect of the ERM proteins knockdown on actin cytoskeleton structure, focal adhesion, and ERM distribution on day 7 of adipogenesis. (A) In cells not treated with siRNA ERM, the phallodin-labeled F-actins (red) are no longer bundled and vinculin staining using FITC-conjugated antibody (green) shows disappearance of large focal adhesions, and the ERM proteins distribution, visualized with Alexa488-conjugated antibody, is consistent with cell rounding (B); when treated with siRNA ERM at day 0, thick bundles of stress actin fibers remain intact (C), and concentrated ERM proteins are observed along the cell periphery (D); when treated with siRNA at day 3 however, the actin organization; (E) is similar to that found in (A), and the ERM proteins distribution appears not well organized and more diffusive (F). All images are 100 mm 100 mm in size.
3
2
growth medium adipogenic RNAi on day 3 RNAi on day 0
1 day 0
day 3
day 6
day 9
day 12
Fig. 4. Dynamics of the cell stiffness changes of hMSCs undergoing adipogenesis. As expected, hMSCs cultured in the regular growth media showed no changes in the cell stiffness. The ERM siRNA transfection on day 0 or day 3 of the adiopgenic induction caused a significant reduction in the cell elastic modulus and continued to decrease. However, the dynamic of cellular biomechanical remodeling was most evident when siRNA treatment was applied at day 3.
opening up a new research field correlating biomechanics and soluble factors which are considered orthogonal cues. An important consequence of transient ERM knockdown is the temporal modulation of the cellular biomechanics that recovers to its normal state determined by the local microenvironment and the cell differentiation status. The stem cell commitment by modulation of its mechanical properties through transient ERM genes silencing appears feasible. Temporary ERM knockdown may facilitate or inhibit the rearrangement of actins, membrane– cytoskeleton association (Titushkin and Cho, 2011) and the overall cell mechanical characteristics driven by soluble inductive factors into the targeted phenotype pattern. Generally, the ERM knockdownmediated mechanical changes are more relevant to adipogenesis
(e.g., low cell stiffness, weak membrane–cytoskeleton interaction) than osteogenesis (higher Young’s modulus). Besides, the dynamics of siRNA mediated ERM down-regulation ( 1 week) is more comparable to the time course of adipogenic differentiation ( 2 weeks) rather osteogenic differentiation (44 weeks). Comparison of the final cell morphology of adipocytes and osteoblasts suggests that rapid cytoskeletal rearrangement required for adipocytes should be better facilitated by modulation of the ERM linker proteins. These observations may explain, in part, more pronounced effects of RNAi-mediated ERM knockdown on adipogenesis (current study) than osteogenesis (Titushkin and Cho, 2011). During induction by the soluble factors, the stem cell mechanical properties are adjusted to those of fully differentiated cells.
Please cite this article as: Titushkin, I., et al., Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling. Journal of Biomechanics (2012), http://dx.doi.org/10.1016/j.jbiomech.2012.09.027
I. Titushkin et al. / Journal of Biomechanics ] (]]]]) ]]]–]]]
5
0.8
OD (500 nm)
0.6
0.4
0.2
0 normal
adipogenic
adipogenic, RNAi day 0
adipogenic, RNAi day 3
Fig. 5. Modulation of hMSCs adipogenic commitment by ERM protein knockdown. At day 12 lipid droplets were visualized in cells incubated in the growth media (A) and in the adipogenic media (B). (C,D) ERM siRNA treatments on day 0 or day 3 suppresses or facilitates the formation of lipid droplets. All images are 500 mm 500 mm in size. (E) Quantitative measurements of adipogenic differentiation by the Oil Red O absorbance.
Relative mRNA expression
8
6
4
2
0
normal
adipogenic PPAR
adipogenic, RNAi day 0 LPL aP2
adipogenic, RNAi day 3
Fig. 6. qRT-PCR analysis of adipose tissue-specific biomarker expression in hMSCs after 12 days of adipogenic differentiation. Cells treated with silencing RNA on day 0 show a reduction in the PPAR, LPL and aP2 expression. In contrast, the expressions of the three genes are up-regulated with siRNA treatment on day 3. The PPAR expression is notably enhanced.
However, the dynamics of mechanical changes appears to be lineage-dependent. The adipogenic induction factors produce a much faster decrease in the cell elastic modulus than the osteogenic factors.
This is not surprising since adipogenic commitment is generally expected to proceed at a higher rate than osteogenic commitment (Pittenger et al., 1999). Modulation of the cell mechanical
Please cite this article as: Titushkin, I., et al., Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling. Journal of Biomechanics (2012), http://dx.doi.org/10.1016/j.jbiomech.2012.09.027
6
I. Titushkin et al. / Journal of Biomechanics ] (]]]]) ]]]–]]]
parameters during cell differentiation is likely mediated by highly dynamical actin cytoskeleton remodeling. As reported earlier, polymeric actin is the major determinant of the cytoskeleton elasticity rather than microtubules or intermediate filaments in many cell types (Takai et al., 2005; Trickey et al., 2004; Titushkin et al., 2010a). The actin cytoskeleton participates in the stem cell fate regulation by controlling multiple intracellular signaling pathways (Guilak et al., 2009). First, it has been shown recently that ROCK-dependent cytoskeletal tension and the contractile forces play important roles in the stem cell fate regulation (McBeath et al., 2004). The ability of Rho-ROCK signaling to direct osteogenic versus adipogenic hMSCs differentiation relies on nonmuscle myosin II activity, which emphasizes the role of cytoskeleton in mechanotransduction (Engler et al., 2006). Second, polymeric actin in the cell is closely associated with the focal adhesions providing a mechanical link to the extracellular environment, and therefore playing a key role in mediating cellular responses to external physical cues (Docheva et al., 2007). Third, ERM proteins-mediated actin interaction with the plasma membrane determines the membrane mechanical parameters (e.g., surface tension), and regulates the transport of active agents by endocytosis and exocytosis (Morris and Homann, 2001). The ERM proteins in this strategic position might regulate the whole cell mechanics not only by providing physical membrane– cytoskeleton coupling, but also by participating in signaling mechanisms affecting the cytoskeleton structure and cell adhesion (Mangeat et al., 1999; Poullet et al., 2001). Indeed, the ERM involvement in Rho transduction pathway was demonstrated in several studies (Mackay et al., 1997; Mangeat et al., 1999). On one hand, Rho-dependent phosphorylation and conformational activation of the ERM proteins is required for association of the ERM family and vinculin with the membrane (Kotani et al., 1997). On the other hand, the role of ERM as a downstream Rho effector in the reorganization of actin cytoskeleton has been established (Louvet-Valle´e, 2000). Rho enzyme, in turn, has been shown to regulate formation of the stress fibers and the focal adhesions in many types of cultured cells (Albiges-Rizo et al., 2009). Moreover, the ERM family proteins have been implicated in the assembly of focal adhesions (Poullet et al., 2001). For example, ezrin has been shown to induce FAK (focal adhesion kinase) activation independently of the cell–matrix adhesion. The dual role of ERM proteins as both mechanical linkers and regulators of the intracellular signaling cascades make them potent candidates for modulation of the stem cell biomechanics and cell commitment to a particular phenotype. For adipogenesis, time-dependent suppression of the ERM proteins is critical for facilitating or inhibiting the intended stem cell differentiation. Possible differential role of the individual ERM proteins involved in the regulation of stem cell biomechanics is currently under investigation. Based on our results from the current study, potential development of biomechanical therapeutic approaches for soft tissue regeneration appears feasible.
Conflict of interest statement None.
Acknowledgments This work was supported in part by the NIH grants (CA113975 and HL083298).
References Albiges-Rizo, C., Destaing, O., Fourcade, B., Planus, E., Block, M.R., 2009. Actin machinery and mechanosensitivity in invadopodia, podosomes and focal adhesions. Journal of Cell Science 122, 3037–3049. Chen, H., Titushkin, I., Stroscio, M., Cho, M., 2007. Altered membrane dynamics of quantum dot-conjugated integrins during osteogenic differentiation of human bone marrow derived progenitor cells. Biophysical Journal 92, 1399–1408. Docheva, D., Popov, C., Mutschler, W., Schieker, M., 2007. Human mesenchymal stem cells in contact with their environment: surface characteristics and the integrin system. Journal of Cellular and Molecular Medicine 11, 21–38. Engler, A.J., Sen, S., Sweeney, H.L., Discher, D.E., 2006. Matrix elasticity directs stem cell lineage specification. Cell 126, 677–689. Fie´vet, B., Louvard, D., Arpin, M., 2007. ERM proteins in epithelial cell organization and functions. Biochimica et Biophysica Acta 1773, 653–660. Gautreau, A., Louvard, D., Arpin, M., 2002. ERM proteins and NF2 tumor suppressor: the Yin and Yang of cortical actin organization and cell growth signaling. Current Opinion in Cell Biology 14, 104–109. Guilak, F., Cohen, D.M., Estes, B.T., Gimble, J.M., Liedtke, W., Chen, C.S., 2009. Control of stem cell fate by physical interactions with the extracellular matrix. Cell Stem Cell 5, 17–26. Jena, B.P., 2007. Secretion machinery at the cell plasma membrane. Current Opinion in Structural Biology 17, 437–443. Kotani, H., Takaishi, K., Sasaki, T., Takai, Y., 1997. Rho regulates association of both the ERM family and vinculin with the plasma membrane in MDCK cells. Oncogene 14, 1705–1713. Louvet-Valle´e, S., 2000. ERM proteins: from cellular architecture to cell signaling. Biology of the Cell 92, 305–316. Mackay, D.J., Esch, F., Furthmayr, H., Hall, A., 1997. Rho- and rac-dependent assembly of focal adhesion complexes and actin filaments in permeabilized fibroblasts: an essential role for ezrin/radixin/moesin proteins. Journal of Cell Biology 138, 927–938. Mangeat, P., Roy, C., Martin, M., 1999. ERM proteins in cell adhesion and membrane dynamics. Trends in Cell Biology 9, 187–192. McBeath, R., Pirone, D.M., Nelson, C.M., Bhadriraju, K., Chen, C.S., 2004. Cell shape, cytoskeletal tension, and RhoA regulate stem cell lineage commitment. Developmental Cell 6, 483–495. Morris, C.E., Homann, U., 2001. Cell surface area regulation and membrane tension. Journal of Membrane Biology 179, 79–102. 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., 1999. Multilineage potential of adult human mesenchymal stem cells. Science 284, 143–147. Poullet, P., Gautreau, A., Kadare´, G., Girault, J.A., Louvard, D., Arpin, M., 2001. Ezrin interacts with focal adhesion kinase and induces its activation independently of cell–matrix adhesion. Journal of Biological Chemistry 276, 37686–37691. Radmacher, M., 2002. Measuring the elastic properties of living cells by the atomic force microscope. Methods in Cell Biology 68, 67–90. ¨ Sheetz, M.P., Sable, J.E., Dobereiner, H.G., 2006. Continuous membrane– cytoskeleton adhesion requires continuous accommodation to lipid and cytoskeleton dynamics. Annual Review of Biophysical and Biomolecular Structures 35, 417–434. Takai, E., Costa, K.D., Shaheen, A., Hung, C.T., Guo, X.E., 2005. Osteoblast elastic modulus measured by atomic force microscopy is substrate dependent. Annals of Biomedical Engineering 33, 963–971. Titushkin, I., Cho, M., 2006. Distinct membrane mechanical properties of human mesenchymal stem cells determined using laser optical tweezers. Biophysical Journal 90, 2582–2591. Titushkin, I., Cho, M., 2007. Modulation of cellular mechanics during osteogenic differentiation of human mesenchymal stem cells. Biophysical Journal 93, 3693–3702. Titushkin, I., Cho, M., 2009. Regulation of cell cytoskeleton and membrane mechanics by electric field: role of linker proteins. Biophysical Journal 96, 717–728. Titushkin, I.A., Sun, S., Shn, J., Cho, M., 2010a. Physicochemical control of adult stem cell differentiation: shedding light on potential molecular mechanisms. Journal of Biomedicine and Biotechnology , http://dx.doi.org/10.1155/2010/ 743476. Titushkin, I., Shin, J., Cho, M., 2010b. A new perspective for stem cell mechanobiology: biomechanical control of stem cell behavior and fate. Critical Reviews in Biomedical Engineering 38, 393–433. Titushkin, I., Cho, M., 2011. Altered osteogenic commitment of human mesenchymal stem cells by ERM protein-dependent modulation of cellular biomechanics. Journal of Biomechanics 44, 2692–2698. Trickey, W.R., Vail, T.P., Guilak, F., 2004. The role of the cytoskeleton in the viscoelastic properties of human articular chondrocytes. Journal of Orthopedic Research 22, 131–139.
Please cite this article as: Titushkin, I., et al., Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling. Journal of Biomechanics (2012), http://dx.doi.org/10.1016/j.jbiomech.2012.09.027
Contents lists available at SciVerse ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling Igor Titushkin, Shan Sun, Amit Paul, Michael Cho n Department of Bioengineering, University of Illinois, Chicago, IL 60607, United States
a r t i c l e i n f o
abstract
Article history: Accepted 29 September 2012
We have recently shown that altered stem cell biomechanics can regulate the lineage commitment through a family of the membrane–cytoskeleton linker proteins (ERM; ezrin, radixin, moesin). The ERM proteins not only modulate the cell stiffness and actin cytoskeleton organization, but also rearrange focal adhesions and therefore influence the biochemically-directed stem cell differentiation. Combining silencing RNA, atomic force microscopy, and fluorescence microscopy, the role of the ERM proteins involved in the regulation of stem cell biomechanics and adipogenic differentiation was quantitatively determined. Transient ERM knockdown by RNAi caused disassembly of actin stress fibers and focal adhesions and a decrease in the cell stiffness. The silencing RNA treatment not only induced mechanical changes in stem cells but impaired adipogenesis in a time-dependent manner. While siRNA ERM treatment at day 0 substantially interfered with adipogenesis, the same treatment at day 3 of adipogenic differentiation significantly facilitated adipogenesis, as assessed by the expression of adipocyte-specific markers. The intact biomechanics homeostasis appears to be critical for the adipogenic induction. These findings may lead to potential biomechanical intervention techniques and methodologies to control the fate and extent of adipogenesis that would likely be involved in stem cell-based therapeutics for soft tissue repair and regeneration. & 2012 Published by Elsevier Ltd.
Keywords: Adipogenesis ERM proteins Stem cell biomechanics SiRNA AFM
1. Introduction The important role of local biomechanical microenvironment in the stem cell activities and metabolism has long been recognized (Guilak et al., 2009). Physical factors including extracellular matrix (ECM), cell stiffness, cell seeding density and morphology and cytoskeleton tension have all been shown to regulate mesenchymal stem cell commitment to a particular lineage (Engler et al., 2006; McBeath et al., 2004). We have previously demonstrated that human mesenchymal stem cells (hMSCs) exhibit unique biomechanical properties as compared to terminally differentiated specialized cells (Titushkin and Cho, 2006, 2007, 2009, 2011; Titushkin et al., 2010a, 2010b). Several predominant features found in hMSCs include unorganized but thick actin stress fibers and weak membrane–cytoskeleton coupling (see Titushkin and Cho, 2011). While actin stress fibers are largely responsible for the cell stiffness, strong membrane–cytoskeleton coupling adhesion would maintain structural integrity of loadbearing cells (e.g., musculoskeletal cells) subjected to multiple stress cycles (Sheetz et al., 2006). In contrast, a relatively loose n Correspondence to: Department of Bioengineering, University of Illinois, Chicago, 851 S. Morgan St. (M/C 063), Chicago, IL 60607, United States. Tel.: þ1 312 413 9424; fax: þ 1 312 996 5921. E-mail address: [email protected] (M. Cho).
membrane attachment to actin cytoskeleton observed in hMSCs (Titushkin and Cho, 2006) would be expected to facilitate signaling through exo- and endocytosis and cross-membrane trafficking (Jena, 2007). A family of three closely related linker proteins (ezrin, radixin and moesin; ERM) serves as general cross-linkers between the plasma membrane and actin filaments in many mesodermderived cell types (Fie´vet et al., 2007; Louvet-Valle´e, 2000). The amino-terminal half of the ERM proteins are responsible for their interaction with the integral proteins of the plasma membrane, including CD44, ICAM, EBP50 (Mangeat et al., 1999). The carboxyterminal half of the ERM linkers is involved in their interaction with actin filaments. It appears that ERM proteins probably control the cell mechanics both directly by providing the membrane–cytoskeletal linkage (Gautreau et al., 2002; Titushkin et al., 2010a), and indirectly by regulating the actin structure and cell adhesion through Rho-dependent mechanisms (Kotani et al., 1997; Poullet et al., 2001). Our recent results show that, when hMSCs differentiate into osteoblasts, their actins reorganization from sparse thick bundles into a dense thin microfilament network provides more binding sites for ERM linkers, increasing the membrane–cytoskeleton interaction in osteogenically committed hMSCs (Titushkin and Cho, 2011). Significant changes in the cellular mechanics during differentiation suggests that external physical, chemical or genetic manipulation of cellular mechanics
0021-9290/$ - see front matter & 2012 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.jbiomech.2012.09.027
Please cite this article as: Titushkin, I., et al., Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling. Journal of Biomechanics (2012), http://dx.doi.org/10.1016/j.jbiomech.2012.09.027
2
I. Titushkin et al. / Journal of Biomechanics ] (]]]]) ]]]–]]]
0
1
3
7
11
13
17
Day Fig. 1. Images of hMSCs undergoing adipogenesis. F-actins (red), lipids (green) and nuclei (blue) were fluorescently labeled and visualized using rhodamine-phalloidin, LipidTox and DAPI, respectively, at different days of adipogenic differentiation.
could be used to enhance or suppress stem cell induction into the pre-selected phenotype by soluble inductive factors. The ERM linkers could be promising targets for manipulation of the cell mechanics due to their important role in the regulation of both the membrane and the cytoskeleton mechanics. In the present study we applied RNA interference technique to transiently down-regulate of the ERM proteins and modulated the mechanical properties of stem cells undergoing the adipogenic commitment. Since the adipocyte morphology is clearly distinguishable from that of osteoblasts, control of adipogenesis via ERM proteins could shed further insights into the lineage-dependent biomechanical remodeling of stem cells. Indeed, unlike osteogenic differentiation of hMSCs, the extent of adipogenesis was dependent on the time of siRNA transfection. We provide mechanical and biological evidence that the initial biomechanics homeostasis is crucially important for stem cells to respond to differentiation cues such as soluble factors.
2. Materials and methods 2.1. Cell culture and differentiation Human mesenchymal stem cells were obtained from the Tulane Center for Gene Therapy (New Orleans, LA). Based on the flow cytometry results, these cells showed negative staining for CD34, CD36, CD45, and CD117 markers (all less than 2%), and positive staining for CD44, CD90, CD166, CD29, CD49c, and CD105 markers (all more than 95%). The cells were grown in Dulbecco’s modified Eagle’s medium (growth medium, Invitrogen, Carlsbad, CA) with 15% fetal bovine serum, and passages between 3 and 10 were used for all experiments. For adipogenic differentiation, cells were incubated in the adipogenic induction media that contains 1 mM dexamethasone, 10 mg/ml insulin, 200 mM indomethacin and 0.5 mM isobutylmethylxantine.
2.2. siRNA transfection quantitative RT-PCCR siRNA targeting ezrin, radixin and moesin were obtained as a pool of 3 targetspecific siRNA duplexes from Santa Cruz Biotechnology (Santa Cruz, CA). Control siRNA consisted of a scrambled sequence that did not lead to specific degradation of any known cellular mRNA. hMSCs were treated for 5 h with 60 nM ERM siRNA or control siRNA premixed with Transfection reagent (Santa Cruz Biotechnology) in serum-free medium. Cells were then incubated in the adipogenic or normal growth medium. As control scrambled sequence siRNA is not expected to lead to specific degradation of any known cellular mRNA, it did not change the cellular ERM expression levels. In 2 days after siRNA transfection, 490% of ERM siRNAtreated cells were found viable as compared to 96% viability in cells incubated in growth medium without any treatment (Titushkin and Cho, 2011). Treatment of cells with siRNA reduced the ERM proteins’ expression by 1.6-fold. Real-time PCR analysis was performed using 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA). After an initial incubation at 50 1C for 2 min, and denaturation at 95 1C for 10 min, 40 cycles of PCR amplification (95 1C for 15 s, 60 1C for 30 s) were carried out. The data is presented as the fold change in the target gene expression normalized to the endogenous reference gene (GAPDH) and relative to the untreated control. Each reaction was performed in triplicate and the entire RT-PCR analysis was repeated twice for each data point.
2.3. Immunostaining For immunofluorescence studies, the samples were fixed in 3.7% formaldehyde and permeabilized in cold ( 20 1C) acetone. Nonspecific binding sites were blocked using a 3% bovine serum albumin. F-actins were stained with 5 mM rhodaminephalloidin (Invitrogen). ERM proteins were labeled with rabbit ERM antibody (Cell Signaling Technology, Danvers, MA), and secondary antibody conjugated with AlexaFluor488 dye (Invitrogen). Focal adhesions were labeled using mouse antibody to vinculin (Chemicon, Temecula, CA), and secondary fluorescent antibody conjugated with FITC. Samples were imaged by a laser scanning confocal system (Radiance 2100MP, Bio-Rad, Hercules, CA) using a 60 objective lens (NA 1.4). 2.4. Cell stiffness measurements The live cell elasticity was measured with a Novascan atomic force microscope (AFM, Novascan Technologies, Ames, IA) as described earlier (Titushkin and Cho, 2007). To minimize the effect of glass substrate on the cell elasticity measurements, we used an indentation depth up to 500 nm ( 10–15% of the average cell height) for data analysis. A total of 70–100 cells of each condition were measured and analyzed according to the Hertz model (Radmacher, 2002). 2.5. Quantification of lipid droplets Two different techniques were applied to assess formation of lipid droplets. Cells were washed, fixed in 3.7% formaldehyde, permeabilized with 0.1% saponin, and lipids were fluorescently labeled using LipidTox (1:200 , Invitrogen) for 30 min at room temperature. For quantitative measurements, cells were stained with 0.3% Oil Red O dye (Sigma-Aldrich), and Oil Red O was extracted from the cells with isopropanol. Absorbance was measured at 500 nm.
3. Results 3.1. Biomechanical remodeling during adipogenesis Using fluorescence microscopy, changes in the stem cell cytoskeleton and morphology were monitored. Representative images of actin remodeling and appearance of lipid droplets were recorded as a function of time (Fig. 1). As expected, hMSCs undergo rapid morphological changes to assume round shape and reduction in the cell size, as demonstrated by rhodamine-labeled actins. More strikingly, rounded cellular morphology appeared clear at day 1 of adipogenesis along with signals from the fluorescently labeled lipids. By day 7 the intracellular space was filled with lipid droplets. Implication could be that the cell rounding preceded the formation of lipid droplets, and that biomechanics remodeling may not be simply a consequence but necessary cellular event for adipogenesis. This postulate was quantitatively supported, in part, by measuring the hMSCs’ elastic moduli using AFM microindentation (Fig. 2). The cell stiffness decreased exponentially (t1/2 ¼3.16 days), and no further significant decrease was observed after 7 days of adipogenic differentiation. However, while the lack of changes in the hMSCs’ biomechanical properties after this time point was evident, lipid droplets continued to accumulate inside the cells.
Please cite this article as: Titushkin, I., et al., Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling. Journal of Biomechanics (2012), http://dx.doi.org/10.1016/j.jbiomech.2012.09.027
I. Titushkin et al. / Journal of Biomechanics ] (]]]]) ]]]–]]]
Elastic Modulus (kPa)
5 *
4
3
2
1
0 0
1
2
3 7 Time (Day)
13
15
Fig. 2. Time-dependent changes of cell stiffness. AFM indentation was performed at selected days of adipodifferentiation and the cell stiffness was quantified by measuring the elastic modulus using the Hertz model (Titushkin and Cho, 2007). The elastic moduli were fitted to a single exponential function (solid line drawn) with the characteristic half time of 3.16 days. Each value represents mean 7 SEM from 70 to 100 microindentation experiments using 20–30 cells. The cell stiffness measurements were statistically different from the cells at day 0 (po 0.05), except those marked with an asterisk (*).
3.2. Role of ERM proteins The total level of ERM proteins, i.e. the 80 kDa band (corresponding to ezrin and radixin) and the 75 kDa band (corresponding to moesin) was obtained by densiometric measurements. Control siRNA consisted of a scrambled sequence that did not lead to specific degradation of any known cellular mRNA. Confirmed by Western blot analysis, the siRNA transfection reduced the ERM protein expression by 1.6 fold (Titushkin and Cho, 2011). A similar level of RNAi-mediated ERM downregulation was also observed in cells incubated both in the growth or adipogenic media, suggesting that the soluble factors did not interfere with inhibition of the ERM gene expression. Adipogenic commitment of hMSCs by the soluble factors changed both the actin cytoskeleton organization and redistributed the ERM proteins (Fig. 3). Actin, focal adhesion (identified by vinculin) and ERM proteins were immunostained and fluorescently visualized for hMSCs undergoing adipogenesis for 7 days. The thick actin stress fibers were replaced with shorter actin filaments (Fig. 3A). Besides, large focal adhesions typically found in normal stem cells (Chen et al., 2007; Titushkin and Cho, 2009) were substituted with much smaller adhesion complexes. The ERM proteins were observed to correlate with morphological remodeling and concentrated near the membrane (Fig. 3B). Cell treatment with siRNA ERM prior to adipogenic induction (day 0) interfered with adipogenically-mediated actin reorganization and still retained short but thick bundles of actin filaments (Fig. 3C), and stained ERM proteins distributed along the cell periphery (Fig. 3D). In contrast, the same siRNA treatment applied at day 3 of the induction caused even higher density of short actin filaments (Fig. 3E) than in cells undergoing biochemical adipogenic differentiation but not treated with siRNA, and the ERM distribution (Fig. 3F) also resembled that found in normally differentiating hMSCs. These images suggest that the dynamics of the biomechanical remodeling may be dependent on the time of siRNA ERM transfection. As expected, the siRNA ERM transfection modulated the dynamics of adipogenically-induced mechanical changes in hMSCs. For example, RNAi treatment either on day 0 or on day 3 caused an immediate decrease in the elastic modulus. However, after 12 days of induction in the adipogenic medium, no statistically significant differences in the cell elasticity
3
were observed between siRNA-treated and untreated but otherwise differentiating to adipocytes (Fig. 4). To validate whether biochemically-induced adipogenesis depended on the initial biomechanics homeostasis, triglyceriderich vesicles of lipid droplets were visualized and quantified using fluorescence imaging and Oil Red O cytological staining (Fig. 5). After 12 days of induction by the adipogenic soluble factors, lipid droplets were observed in the cytoplasm of at least 70% of cells that were committed to the adipogenic lineage. Cells incubated in the growth media without the adipogenic soluble factors expressed a baseline expression for lipid droplets (Fig. 5A) and, in response to the soluble factors, it was significant increased (Fig. 5B). However, the siRNA ERM treatment before adipogenic induction (day 0) resulted in far fewer lipid droplets (Fig. 5C), whereas the same treatment on day 3 of adipogenic induction increased the formation of lipid droplets (Fig. 5D) that was even greater than the positive control (Fig. 5B). These results were quantitatively confirmed by measuring the absorbance of Oil Red O dye extracted from the cells. The early (i.e., day 0) ERM knockdown in adipo-differentiating hMSCs decreased the Oil Red O absorbance by approximately 2-fold in comparison with non-treated adipogenic cells. In contrast, the late (i.e., day 3) ERM knockdown significantly increased Oil Red O staining in adipogenically committed hMSCs. The Oil Red O absorbance for both siRNA-treated and untreated adipo-differentiating cells was higher than for negative control cells in the growth medium. Finally, using quantitative RT-PCR amplification, adipocytespecific gene expression was measured. The mRNA levels of PPARg, LPL and aP2 were significantly increased by a factor of 6.0, 3.8, and 2.8, respectively, in adipo-differentiating cells as compared to hMSCs in the growth medium (Fig. 6). The siRNA ERM treatment on day 0 suppressed the three genes’ expressions. However, consistent with fluorescence images and Oil Red O staining, the ERM proteins knockdown at the later stage of adipodifferentiation (day 3) further increased the expression of these genes by day 12 of adipogenic differentiation.
4. Discussion We have recently proposed a model to characterize the biomechanical state of the cell (Titushkin et al., 2010a, 2010b). While this model is not limited to stem cells, we were able to elucidate unique biomechanical properties of bone marrow-derived hMSCs. As expected, the cell stiffness is predominantly dictated by the actin organization and regulated by the ERM proteins. By suppressing the ERM proteins’ expressions by siRNA treatment, the cell stiffness can be modulated. However, when the intended differentiation is completed, the cell elasticity in the siRNA-treated or untreated cells did not differ significantly. In contrast, the tissue-specific protein and gene expression levels were different and much more depended on the siRNA treatment. These findings suggest the following. First, the cellular mechanical parameters alone may be insufficient to accurately characterize the stem cell differentiation and supports the notion that the stem cell would have to detect, process, and integrate multiple endogenous and exogenous mechanical and physicochemical signals to determine its response and lineage specification (Guilak et al., 2009). Second, the dynamics of biomechanical remodeling is important for the stem cell commitment and differentiation. In adipogenesis, the cell stiffness is essentially the same by day 12 of adipogenesis, independent of the siRNA treatment. However, interference with the stem cell biomechanics significantly prevented or facilitated adipogenesis, indicating that preserving the initial cellular mechanical homeostasis is required for proper stem cell differentiation. It is evident that biomechanics of stem cells is involved in the regulation of stem cell responses to the soluble factors, potentially
Please cite this article as: Titushkin, I., et al., Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling. Journal of Biomechanics (2012), http://dx.doi.org/10.1016/j.jbiomech.2012.09.027
4
I. Titushkin et al. / Journal of Biomechanics ] (]]]]) ]]]–]]]
Elastic Modulus (kPa)
Fig. 3. Effect of the ERM proteins knockdown on actin cytoskeleton structure, focal adhesion, and ERM distribution on day 7 of adipogenesis. (A) In cells not treated with siRNA ERM, the phallodin-labeled F-actins (red) are no longer bundled and vinculin staining using FITC-conjugated antibody (green) shows disappearance of large focal adhesions, and the ERM proteins distribution, visualized with Alexa488-conjugated antibody, is consistent with cell rounding (B); when treated with siRNA ERM at day 0, thick bundles of stress actin fibers remain intact (C), and concentrated ERM proteins are observed along the cell periphery (D); when treated with siRNA at day 3 however, the actin organization; (E) is similar to that found in (A), and the ERM proteins distribution appears not well organized and more diffusive (F). All images are 100 mm 100 mm in size.
3
2
growth medium adipogenic RNAi on day 3 RNAi on day 0
1 day 0
day 3
day 6
day 9
day 12
Fig. 4. Dynamics of the cell stiffness changes of hMSCs undergoing adipogenesis. As expected, hMSCs cultured in the regular growth media showed no changes in the cell stiffness. The ERM siRNA transfection on day 0 or day 3 of the adiopgenic induction caused a significant reduction in the cell elastic modulus and continued to decrease. However, the dynamic of cellular biomechanical remodeling was most evident when siRNA treatment was applied at day 3.
opening up a new research field correlating biomechanics and soluble factors which are considered orthogonal cues. An important consequence of transient ERM knockdown is the temporal modulation of the cellular biomechanics that recovers to its normal state determined by the local microenvironment and the cell differentiation status. The stem cell commitment by modulation of its mechanical properties through transient ERM genes silencing appears feasible. Temporary ERM knockdown may facilitate or inhibit the rearrangement of actins, membrane– cytoskeleton association (Titushkin and Cho, 2011) and the overall cell mechanical characteristics driven by soluble inductive factors into the targeted phenotype pattern. Generally, the ERM knockdownmediated mechanical changes are more relevant to adipogenesis
(e.g., low cell stiffness, weak membrane–cytoskeleton interaction) than osteogenesis (higher Young’s modulus). Besides, the dynamics of siRNA mediated ERM down-regulation ( 1 week) is more comparable to the time course of adipogenic differentiation ( 2 weeks) rather osteogenic differentiation (44 weeks). Comparison of the final cell morphology of adipocytes and osteoblasts suggests that rapid cytoskeletal rearrangement required for adipocytes should be better facilitated by modulation of the ERM linker proteins. These observations may explain, in part, more pronounced effects of RNAi-mediated ERM knockdown on adipogenesis (current study) than osteogenesis (Titushkin and Cho, 2011). During induction by the soluble factors, the stem cell mechanical properties are adjusted to those of fully differentiated cells.
Please cite this article as: Titushkin, I., et al., Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling. Journal of Biomechanics (2012), http://dx.doi.org/10.1016/j.jbiomech.2012.09.027
I. Titushkin et al. / Journal of Biomechanics ] (]]]]) ]]]–]]]
5
0.8
OD (500 nm)
0.6
0.4
0.2
0 normal
adipogenic
adipogenic, RNAi day 0
adipogenic, RNAi day 3
Fig. 5. Modulation of hMSCs adipogenic commitment by ERM protein knockdown. At day 12 lipid droplets were visualized in cells incubated in the growth media (A) and in the adipogenic media (B). (C,D) ERM siRNA treatments on day 0 or day 3 suppresses or facilitates the formation of lipid droplets. All images are 500 mm 500 mm in size. (E) Quantitative measurements of adipogenic differentiation by the Oil Red O absorbance.
Relative mRNA expression
8
6
4
2
0
normal
adipogenic PPAR
adipogenic, RNAi day 0 LPL aP2
adipogenic, RNAi day 3
Fig. 6. qRT-PCR analysis of adipose tissue-specific biomarker expression in hMSCs after 12 days of adipogenic differentiation. Cells treated with silencing RNA on day 0 show a reduction in the PPAR, LPL and aP2 expression. In contrast, the expressions of the three genes are up-regulated with siRNA treatment on day 3. The PPAR expression is notably enhanced.
However, the dynamics of mechanical changes appears to be lineage-dependent. The adipogenic induction factors produce a much faster decrease in the cell elastic modulus than the osteogenic factors.
This is not surprising since adipogenic commitment is generally expected to proceed at a higher rate than osteogenic commitment (Pittenger et al., 1999). Modulation of the cell mechanical
Please cite this article as: Titushkin, I., et al., Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling. Journal of Biomechanics (2012), http://dx.doi.org/10.1016/j.jbiomech.2012.09.027
6
I. Titushkin et al. / Journal of Biomechanics ] (]]]]) ]]]–]]]
parameters during cell differentiation is likely mediated by highly dynamical actin cytoskeleton remodeling. As reported earlier, polymeric actin is the major determinant of the cytoskeleton elasticity rather than microtubules or intermediate filaments in many cell types (Takai et al., 2005; Trickey et al., 2004; Titushkin et al., 2010a). The actin cytoskeleton participates in the stem cell fate regulation by controlling multiple intracellular signaling pathways (Guilak et al., 2009). First, it has been shown recently that ROCK-dependent cytoskeletal tension and the contractile forces play important roles in the stem cell fate regulation (McBeath et al., 2004). The ability of Rho-ROCK signaling to direct osteogenic versus adipogenic hMSCs differentiation relies on nonmuscle myosin II activity, which emphasizes the role of cytoskeleton in mechanotransduction (Engler et al., 2006). Second, polymeric actin in the cell is closely associated with the focal adhesions providing a mechanical link to the extracellular environment, and therefore playing a key role in mediating cellular responses to external physical cues (Docheva et al., 2007). Third, ERM proteins-mediated actin interaction with the plasma membrane determines the membrane mechanical parameters (e.g., surface tension), and regulates the transport of active agents by endocytosis and exocytosis (Morris and Homann, 2001). The ERM proteins in this strategic position might regulate the whole cell mechanics not only by providing physical membrane– cytoskeleton coupling, but also by participating in signaling mechanisms affecting the cytoskeleton structure and cell adhesion (Mangeat et al., 1999; Poullet et al., 2001). Indeed, the ERM involvement in Rho transduction pathway was demonstrated in several studies (Mackay et al., 1997; Mangeat et al., 1999). On one hand, Rho-dependent phosphorylation and conformational activation of the ERM proteins is required for association of the ERM family and vinculin with the membrane (Kotani et al., 1997). On the other hand, the role of ERM as a downstream Rho effector in the reorganization of actin cytoskeleton has been established (Louvet-Valle´e, 2000). Rho enzyme, in turn, has been shown to regulate formation of the stress fibers and the focal adhesions in many types of cultured cells (Albiges-Rizo et al., 2009). Moreover, the ERM family proteins have been implicated in the assembly of focal adhesions (Poullet et al., 2001). For example, ezrin has been shown to induce FAK (focal adhesion kinase) activation independently of the cell–matrix adhesion. The dual role of ERM proteins as both mechanical linkers and regulators of the intracellular signaling cascades make them potent candidates for modulation of the stem cell biomechanics and cell commitment to a particular phenotype. For adipogenesis, time-dependent suppression of the ERM proteins is critical for facilitating or inhibiting the intended stem cell differentiation. Possible differential role of the individual ERM proteins involved in the regulation of stem cell biomechanics is currently under investigation. Based on our results from the current study, potential development of biomechanical therapeutic approaches for soft tissue regeneration appears feasible.
Conflict of interest statement None.
Acknowledgments This work was supported in part by the NIH grants (CA113975 and HL083298).
References Albiges-Rizo, C., Destaing, O., Fourcade, B., Planus, E., Block, M.R., 2009. Actin machinery and mechanosensitivity in invadopodia, podosomes and focal adhesions. Journal of Cell Science 122, 3037–3049. Chen, H., Titushkin, I., Stroscio, M., Cho, M., 2007. Altered membrane dynamics of quantum dot-conjugated integrins during osteogenic differentiation of human bone marrow derived progenitor cells. Biophysical Journal 92, 1399–1408. Docheva, D., Popov, C., Mutschler, W., Schieker, M., 2007. Human mesenchymal stem cells in contact with their environment: surface characteristics and the integrin system. Journal of Cellular and Molecular Medicine 11, 21–38. Engler, A.J., Sen, S., Sweeney, H.L., Discher, D.E., 2006. Matrix elasticity directs stem cell lineage specification. Cell 126, 677–689. Fie´vet, B., Louvard, D., Arpin, M., 2007. ERM proteins in epithelial cell organization and functions. Biochimica et Biophysica Acta 1773, 653–660. Gautreau, A., Louvard, D., Arpin, M., 2002. ERM proteins and NF2 tumor suppressor: the Yin and Yang of cortical actin organization and cell growth signaling. Current Opinion in Cell Biology 14, 104–109. Guilak, F., Cohen, D.M., Estes, B.T., Gimble, J.M., Liedtke, W., Chen, C.S., 2009. Control of stem cell fate by physical interactions with the extracellular matrix. Cell Stem Cell 5, 17–26. Jena, B.P., 2007. Secretion machinery at the cell plasma membrane. Current Opinion in Structural Biology 17, 437–443. Kotani, H., Takaishi, K., Sasaki, T., Takai, Y., 1997. Rho regulates association of both the ERM family and vinculin with the plasma membrane in MDCK cells. Oncogene 14, 1705–1713. Louvet-Valle´e, S., 2000. ERM proteins: from cellular architecture to cell signaling. Biology of the Cell 92, 305–316. Mackay, D.J., Esch, F., Furthmayr, H., Hall, A., 1997. Rho- and rac-dependent assembly of focal adhesion complexes and actin filaments in permeabilized fibroblasts: an essential role for ezrin/radixin/moesin proteins. Journal of Cell Biology 138, 927–938. Mangeat, P., Roy, C., Martin, M., 1999. ERM proteins in cell adhesion and membrane dynamics. Trends in Cell Biology 9, 187–192. McBeath, R., Pirone, D.M., Nelson, C.M., Bhadriraju, K., Chen, C.S., 2004. Cell shape, cytoskeletal tension, and RhoA regulate stem cell lineage commitment. Developmental Cell 6, 483–495. Morris, C.E., Homann, U., 2001. Cell surface area regulation and membrane tension. Journal of Membrane Biology 179, 79–102. 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., 1999. Multilineage potential of adult human mesenchymal stem cells. Science 284, 143–147. Poullet, P., Gautreau, A., Kadare´, G., Girault, J.A., Louvard, D., Arpin, M., 2001. Ezrin interacts with focal adhesion kinase and induces its activation independently of cell–matrix adhesion. Journal of Biological Chemistry 276, 37686–37691. Radmacher, M., 2002. Measuring the elastic properties of living cells by the atomic force microscope. Methods in Cell Biology 68, 67–90. ¨ Sheetz, M.P., Sable, J.E., Dobereiner, H.G., 2006. Continuous membrane– cytoskeleton adhesion requires continuous accommodation to lipid and cytoskeleton dynamics. Annual Review of Biophysical and Biomolecular Structures 35, 417–434. Takai, E., Costa, K.D., Shaheen, A., Hung, C.T., Guo, X.E., 2005. Osteoblast elastic modulus measured by atomic force microscopy is substrate dependent. Annals of Biomedical Engineering 33, 963–971. Titushkin, I., Cho, M., 2006. Distinct membrane mechanical properties of human mesenchymal stem cells determined using laser optical tweezers. Biophysical Journal 90, 2582–2591. Titushkin, I., Cho, M., 2007. Modulation of cellular mechanics during osteogenic differentiation of human mesenchymal stem cells. Biophysical Journal 93, 3693–3702. Titushkin, I., Cho, M., 2009. Regulation of cell cytoskeleton and membrane mechanics by electric field: role of linker proteins. Biophysical Journal 96, 717–728. Titushkin, I.A., Sun, S., Shn, J., Cho, M., 2010a. Physicochemical control of adult stem cell differentiation: shedding light on potential molecular mechanisms. Journal of Biomedicine and Biotechnology , http://dx.doi.org/10.1155/2010/ 743476. Titushkin, I., Shin, J., Cho, M., 2010b. A new perspective for stem cell mechanobiology: biomechanical control of stem cell behavior and fate. Critical Reviews in Biomedical Engineering 38, 393–433. Titushkin, I., Cho, M., 2011. Altered osteogenic commitment of human mesenchymal stem cells by ERM protein-dependent modulation of cellular biomechanics. Journal of Biomechanics 44, 2692–2698. Trickey, W.R., Vail, T.P., Guilak, F., 2004. The role of the cytoskeleton in the viscoelastic properties of human articular chondrocytes. Journal of Orthopedic Research 22, 131–139.
Please cite this article as: Titushkin, I., et al., Control of adipogenesis by ezrin, radixin and moesin-dependent biomechanics remodeling. Journal of Biomechanics (2012), http://dx.doi.org/10.1016/j.jbiomech.2012.09.027