NANOG Restores Contractility of Mesenchymal Stem Cell-Based Senescent Microtissues

Cultivation of HF-MSCs, cloning, and lentivirus transduction

Human hair follicle MSCs were isolated from a 73-year-old donor ²⁸ and cultured up to passage 14. Based on our previous study, the cells were defined as early passage (EP) before passage 8 and late passage (LP) after passage 14, representing young and senescent cells, respectively. ²⁸ Human dermal fibroblasts were derived from neonatal foreskin as described previously. ²⁹ Human aortic vascular SMCs (VSMCs) were purchased from ATCC (Manassas, VA). Cells were cultured in growth medium comprising Dulbecco's modified Eagle's medium (Gibco, Grand Island, NY) supplemented with 10% FBS (Gibco). Cells were passaged every 4 days using 0.25% trypsin (Gibco). SMC and human fibroblast contraction experiments were performed at passage 9.

The plasmid for reverse tetracycline transactivator (rtTA) was purchased from Addgene (Cambridge, MA). Cloning of NANOG or ZsGreen into the tetracycline controlled (Tet-On) lentiviral vector pNL-EGFP/TREPittdU3 (Addgene), resulting in pTRE-NANOG-IRES-PURO or pTRE-ZsGreen-IRES-PURO as described previously. ³⁰ Recombinant lentiviruses encoding for rtTA, NANOG, or ZsGreen were produced using these plasmids as described elsewhere. ³¹ Cells were cotransduced with NANOG- or ZsGreen-encoding virus and the virus encoding for rtTA. NANOG or ZsGreen expression was induced by DOX (BD, Franklin Lakes, NJ) at a final concentration of 1 μg/mL in the culture medium, unless stated otherwise. Cells that were cultured in the absence of DOX were considered as control (Ctr). Before seeding the microdevice, cells were cultured in the presence of DOX for 5 days unless stated otherwise.

Microdevices to measure contractile properties

To measure contractility, we manufactured a microdevice that contains an array of 13 × 10 microwells, each containing two micropillars with heads, using multilayer microlithography, as described previously. ²⁷ Briefly, SU-8 photoresist was used to create the first layer on silicon wafer. This layer constitutes the base of the micropillar. After exposing the base layer to UV light, the head section (top layer) of the micropillar was spun on top and exposed to UV through the mask. Both head and bottom patterns were exposed to 365 nm UV light at 10 mJ/cm² intensity for 20 and 50 s, respectively. A blocking photoresist layer in between the top and base layers was cast off to prevent unwanted cross-linking of the first layer. Next, the micropillar array pattern was transferred to polydimethylsiloxane (PDMS, Sylgard 184; Dow-Corning, Midland, MI) through replica molding. The final device was casted in a P35 Petri dish using PDMS stamps.

To seed microtissues in the micropillar array, arrays were sterilized in 70% ethanol and then treated with 0.2% Pluronic F127 (BASF, Florham Park, NJ) to form a hydrophobic barrier between the PDMS surface and collagen and prevent collagen from attaching to the PDMS surface. Next, rat tail collagen type I (Corning^®, Riverfront Plaza, NY) was neutralized by NaOH (1 M), mixed with cells (3 × 10⁵ cells per device), and immediately seeded on the device. Microtissues were formed in between the two micropillars and contractile forces were measured after 24 h unless otherwise stated. To this end, images were taken from top and bottom (stationary part) of the micropillars using 10 × objective on an Olympus IX81 microscope (Center Valley, PA). Next, the deflection of micropillar heads from the stationary base was measured using ImageJ software and multiplied by the pillar's spring constant (2.2 ± 0.1 μN/μm) to calculate the force. ²⁷ Contractile force was normalized to the total cell number in each microtissue.

Western blot analysis

Cells grown on tissue culture plates were lysed using lysis buffer: 62.5 mM Tris-HCl (pH 6.8 at 25°C), 2% (w/v) sodium dodecyl sulfate (SDS), 10% (v/v) glycerol, 0.1% (w/v) bromophenol blue, supplemented with 41.67 mM dithiothreitol (Cell Signaling, Danver, MA), and protease inhibitor cocktail (Sigma-Aldrich, St. Louis, MO). Protein lysates were denatured at 95°C for 5 min and separated based on their molecular weight by 8% or 10% SDS-polyacrylamide gel electrophoresis. Proteins were transferred to nitrocellulose membranes (Trans-Blot Turbo Transfer System; Bio-Rad, Hercules, CA) and probed for specific proteins using antibodies against NANOG (1:1000 in 5% (w/v) bovine serum albumin; BD) and ACTA2 (1:2000 dilution in 5% (w/v) nonfat dry milk; Sigma-Aldrich). The antibody against glyceraldehyde-3-phosphate dehydrogenase (GAPDH, 1:10,000 dilution in 5% (w/v) nonfat dry milk; Cell Signaling) was used to determine the total amount of protein loaded. Protein bands were detected using horseradish peroxidase-labeled secondary antibodies and a chemiluminescence kit (LumiGLO; Cell Signaling) as per manufacturer's instructions. The protein levels of NANOG and ACTA2 were quantified by measuring the band intensities using ImageJ software and normalized to the intensity of GAPDH. The lysate concentration and exposure time were chosen to ensure that the signal was in the linear range.

Immunocytochemistry

To observe the organization of actin filaments, immunostaining for F-actin using phalloidin stain and ACTA2 using ACTA2 antibody was performed on the 3D microtissues or cells cultured in two dimensions. Briefly, cells or microtissues were fixed in 4% (w/v) paraformaldehyde for 10 min at 25°C, permeabilized by incubating with 0.1% (v/v) Triton X-100/PBS for 10 min at 25°C, and blocked with 5% (v/v) goat serum in 0.01% (w/v) Triton X-100/PBS at 25°C for 1 h. To probe for ACTA2 and MRTF-A, the cells were incubated overnight at 4°C with the following antibodies diluted in blocking buffer: mouse anti-human α-smooth muscle (1:100, A5228; Sigma-Aldrich) and rabbit anti-human MRTF-A (1:200; Santa Cruz, Dallas, TX). Subsequently, the cells were incubated for 1 h at 25°C with Alexa Fluor 488-conjugated goat anti-mouse or goat anti-rabbit antibodies diluted in blocking buffer (1:100; Thermo Fisher Scientific, Waltham, MA) and counterstained with Hoechst 33342 nuclear dye for 5 min (1:400 dilution in PBS; Thermo Fisher Scientific). F-actin staining was performed using Alexa Flour 594 phalloidin (1:50 dilution in PBS +1% (w/v) bovine serum albumin; Thermo Fisher Scientific) as per manufacturer's instructions. Fluorescent images were acquired using the Zeiss Axio Observer Z1 (LSM 510; Zeiss, Oberkochen, Germany) equipped with a digital camera (ORCA-ER C4742-80; Hamamatsu, Bridgewater, NJ). Immunostaining for MRTF-A nuclear translocation was quantified using ImageJ software by measuring the intensity of the MRTF-A signal colocalized with the Hoechst nuclear stain.

RNA isolation, cDNA synthesis, and quantitative real-time polymerase chain reaction

To analyze the mRNA expression profile of ACTA2 after expression and removal of NANOG, the total RNA was isolated from cells plated on tissue culture plates using the RNeasy Mini Kit (QIAGEN, Valencia, CA) as per manufacturer's instructions. Subsequently, cDNA was synthesized from total RNA using the QuantiTect Reverse Transcription Kit (QIAGEN). Quantitative real-time polymerase chain reaction (RT-PCR) was performed using the SYBR Green Kit (Bio-Rad) according to the manufacturer's instructions and specific primers for each gene as follows:

ACTA2 (forward: 5′-GACAGCTACGTGGGTGACGAA-3′; reverse: 5′-GATGCCATGTTCTATCGGGTACT-3′);

MYL9 (forward: 5′-TGACAAGGAGGACCTGCAC-3′; reverse: 5′-CATCATGCCCTCCAGGTATT-3′);

ACTG2 (forward: 5′-CGTGACCTCACGGACTACCT-3′; reverse: 5′-TGTGGCCATCTCATTCTCAA-3′);

TAGLN (forward: 5′-ATGGCCAACAAGGGTCC-3′; reverse: 5′-CTTCAAAGAGGTCAACAG-3′);

MYH11 (forward: 5′- CGCCAAGAGACTCGTCTGG-3′; reverse: 5′-TCTTTCCCAACCGTGACCTTC-3′); and

RPL32 (forward: 5′-AGCGTAACTGGCGGAAAC-3′; reverse: 5′-CGTTGTGGACCAGGAACTTC-3′).

The mRNA expression level was quantified using the ΔC_T method and normalized to the expression level of the housekeeping gene, RPL32.

Statistical analysis

Data are shown as mean ± standard deviation (SD). For the kinetics of gene expression, we performed analysis of variance (one-way ANOVA), followed by Fisher's least significant difference as a post hoc analysis. The statistical significance between two mean values was analyzed using two-tailed unpaired Student's t-test with respect to the standard deviations. Probability values less than 0.05 (p < 0.05) were considered statistically significant. Forces were measured from at least 20 microtissues per condition in each experiment and all experiments were repeated three independent times.

NANOG enhances the contractile force generated by 3D tissue constructs

First, we evaluated the force generation capacity of several contractile cell types such as human hair follicles derived MSCs at EP, VSMCs, and fibroblasts. To this end, cells were embedded in collagen gels (3 × 10⁵ cells per device) and allowed to polymerize in microwells, each containing a pair of flexible microposts (Fig. 1A, B), and the force exerted by the cells in each microtissue was calculated by micropost displacement using Hooke's law (Fig. 1C). Our results show that the contractile force of VSMC-based microtissues (1489 ± 278 nN/cell) was significantly higher than the force generated by EP MSCs (200 ± 53 nN/cell), while 5 days of transforming growth factor-beta (TGF-β1) treatment enhanced the contractile force of MSCs to 561 ± 113 nN/cell. The force generated by fibroblasts (422 ± 79 nN/cell) was also higher than MSCs, but still significantly lower than VSMCs and MSC+TGF-β1 (Fig. 1D).

OPEN IN VIEWER

FIG. 1.

Restoring the contractile function of senescent cells by NANOG expression. (A) Schematic representation of the experimental setup (scale bar = 200 μm). (B) Low-magnification picture from the array of microwells. (C) Bottom, top, and cross-sectional views of a microtissue (scale bar = 200 μm). (D) Forces generated by human EP MSCs, VSMCs, and fibroblasts. (E, F) Western blotting and quantification of ACTA2 and NANOG proteins for EP, LP, and LP-NANOG MSCs. (G) Contractile force generated by EP, LP, and LP-NANOG. (H) Immunofluorescence images of EP, LP, and LP-NANOG tissues stained for ACTA2 and F-actin (scale bar = 200 μm). Data are presented as mean ± standard deviation, n = 3 for WB quantification and n = 20 for force measurements. $ denotes p < 0.05 compared with all other samples, * denotes p < 0.05 compared with control. EP, early passage; LP, late passage; MSC, mesenchymal stem cell; VSMC, vascular smooth muscle cell.

Previous work from our laboratory showed that ectopic expression of the pluripotent transcription factor NANOG increased the myogenic differentiation and contractile capacity of SMCs derived from senescent MSCs.^25,30 Therefore, we examined whether NANOG could increase the force generation capacity of 3D senescent tissue constructs. ²⁷ To this end, we developed a tetracycline-regulatable lentiviral system that enabled expression of NANOG upon addition of DOX to the culture medium. Hair follicle (HF)-derived MSCs were transduced at an EP and NANOG was expressed in senescent cells (LP cells) upon addition of DOX in the medium. Senescent cells that were cultured in the absence of DOX were used as Ctr throughout the study.

We found that senescent microtissues, that is, microtissues containing LP cells, expressed significantly lower amount of ACTA2 (Fig. 1E, F) and produced significantly lower contractile force (LP = 77 ± 62 nN/cell) compared with microtissues prepared with EP cells (EP = 179 ± 36 nN/cell) (Fig. 1G). Interestingly, expression of NANOG restored ACTA2 expression and actin filament formation, as well as the contractile function of senescent cells (LP-NANOG = 289 ± 74 nN/cell) (Fig. 1E–G). Cell-mediated contraction of collagen gels occurred rapidly and reached a plateau as soon as 6 h after seeding, but the results were more consistent at 24 h (Supplementary Fig. S1A; Supplementary Data are available online at www.liebertpub.com/tea). Therefore, we chose to perform force measurements at 24 h after cell seeding for the rest of our experiments. We also employed MSCs that were engineered to express ZsGreen (ZsG, a brighter version of GFP) in a DOX-regulatable manner. Upon treatment with DOX, ZsG-expressing microtissues generated similar force as Ctr microtissues (transduced with the NANOG gene, but not treated with DOX) (Supplementary Fig. S1B), suggesting that improved contraction was not affected by DOX, but was solely due to NANOG expression.

NANOG expression and microtissue contraction increased with increasing DOX concentration

Next, we determined the effect of NANOG concentration on force generation. To this end, cells were treated with the indicated concentrations of DOX for 5 days before embedding them into the collagen hydrogels to form microtissues. We found that increasing the concentration of DOX resulted in increased expression of NANOG and led to increased expression of the prototypical myogenic marker ACTA2 (Fig. 2A, B). At 1 μg/mL of DOX, NANOG expression reached a plateau and ACTA2 expression increased ninefold compared with Ctr (0 μg/mL DOX). In addition to increased expression, confocal immunofluorescence microscopy showed that NANOG expression induced ACTA2 and F-actin polymerization and filamentous organization, indicative of contractile SMC phenotype (Fig. 2C). Finally, the magnitude of the contractile force increased in a DOX concentration-dependent manner from 72 ± 41 nN/cell at 0 μg/mL to 321 ± 63 nN/cell at 1 μg/mL of DOX (Fig. 2D).

OPEN IN VIEWER

FIG. 2.

NANOG increased the contractile force of senescent microtissues. MSCs were treated with the indicated DOX concentrations and the contractile force was measured using micropost devices. (A, B) Western blotting analysis and quantification of NANOG and ACTA2 protein levels as a function of DOX concentration. (C) Confocal immunofluorescence microscopy of microtissues stained for ACTA2 (green) protein and F-actin (red) at the indicated DOX concentrations (scale bar = 200 μm). (D) Contractile force as a function of DOX concentration. Data are presented as mean ± standard deviation, n = 3 for WB quantification and n = 20 for force measurements. $ denotes p < 0.05 from all other samples, * denotes p < 0.05 compared with 0 μg/m, and # denotes p < 0.05 compared with 0.01 μg/mL DOX. DOX, doxycycline.

Effect of temporal expression of NANOG on the contractile force of microtissues

Next, we examined whether the time period of NANOG expression affected the level of force generated by microtissues. To this end, MSCs were cultured in the presence of DOX (1 μg/mL) for different times as indicated (between 1 and 5 days) and then embedded in collagen gels at the same time to form microtissues. As shown in Figure 3A, all cells were cultured for 5 days and all microtissues were cultured for 24 h. Interestingly, NANOG protein level reached saturation after 1-day treatment with DOX, while the ACTA2 protein level gradually increased by 3-, 7-, and 11-fold after 2, 4, and 5 days, respectively (Fig. 3B, C). Similarly, filamentous organization of actin also increased with time of DOX treatment, reaching maximum on day 4 (Fig. 3D). Accordingly, the force generated by the tissues also increased with time of DOX treatment, from 72 ± 36 nN/cell (no DOX) to 156 ± 42 nN/cell after 1 day to 237 ± 63 nN/cell after 3 days of DOX treatment and remained at that level thereafter (Fig. 3E).

OPEN IN VIEWER

FIG. 3.

Effect of time of NANOG expression on the contractile force of MSCs. (A) Schematic illustration of the experimental design. MSCs were cultured in the presence of 1 μg/mL DOX for 1 to 5 days and subsequently seeded in the microdevices for 24 h. (B, C) Western blot and quantification of NANOG and ACTA2 protein levels. (D) Immunostaining of microtissues for F-actin (red) and ACTA2 (green) (scale bar = 200 μm). (E) Contractile force as a function of the time of DOX treatment. * denotes p < 0.05 compared with control; # denotes p < 0.05 compared with 1 and 2 days; and $ denotes p < 0.05 compared with 4 days of NANOG expression. Data are presented as mean ± standard deviation, n = 3 for WB quantifications and n = 20 for force measurement.

Additive effects of NANOG and TGF-β1 on contractile function of senescent microtissues

Previous studies have shown that TGF-β1 induced formation of actin filaments and contractility in MSCs.^32–34 Therefore, we examined whether NANOG and TGF-β1 might increase contractility in an additive manner. To this end, we cultured MSCs in the presence of 1 μg/mL DOX and 2 ng/mL TGF-β1 for 5 days before measuring contractile force exerted by the microtissues.

While NANOG or TGF-β1 increased the contractile force significantly, the effect of NANOG was more pronounced than that of TGF-β1 (NANOG: 279 ± 52; TGF-β1: 195 ± 62; Ctr: 92 ± 51 nN/cell). Interestingly, the combination of NANOG and TGF-β1 increased the level of force even further to 529 ± 78 nN/cell (Fig. 4A). In agreement, the combination of TGF-β1 and NANOG also increased the expression level and filamentous organization of actin over either NANOG or TGF-β1 alone (Fig. 4B–D).

OPEN IN VIEWER

FIG. 4.

Additive effects of NANOG and TGF-β1 on restoring the contractile function of senescent microtissues. (A) Effects of NANOG, TGF-β1, or both on the contractile force of microtissues. (B, C) Protein level of ACTA2 in the presence of NANOG, TGF-β1, or both as measured by western blot. (D) Confocal immunofluorescence microscopy of microtissues for ACTA2 (green) protein and F-actin (red) (scale bar = 200 μm). Data are presented as mean ± standard deviation, n = 3 for WB and n = 20 for force measurement. $ denotes p < 0.05 compared with all other samples. TGF-β1, transforming growth factor-beta1.

The effects of NANOG persist for several days after DOX removal

Next, we examined whether the effects of NANOG are transient or whether they persist when NANOG expression is terminated. To this end, cells were treated with DOX (1 μg/mL) to induce NANOG expression for a 5-day window and then DOX was removed for the subsequent 0, 5, 10, or 20 days (Fig. 5A). All microtissues were prepared on day 20 and cultured for 24 h before they were tested. Interestingly, the level of NANOG decreased rapidly upon removal of DOX (half-life ∼2.5 ± 0.5 h) and was barely detectable by 10 h (Fig. 5B). Despite the short half-life of NANOG, the ACTA2 protein level was sustained up to 5 days and decreased significantly (40% decrease, n = 3, p < 0.05) by day 10 and even further by day 20 (60% decrease, n = 3, p < 0.05) compared with the level on day 0 (Fig. 5C). The contractile force was sustained up to 10 days following DOX withdrawal and decreased to levels of Ctr tissues after 20 days (Fig. 5D).

OPEN IN VIEWER

FIG. 5.

The effects of NANOG persist even after DOX removal. (A) Schematic of the experimental design. (B) Western blot for NANOG and quantification of NANOG protein level as a function of time after DOX removal. (C) Western Blot for ACTA2 and quantification as a function of time after DOX removal. (D) Contractile force generated by microtissues as a function of time after DOX removal. Data are presented as mean ± standard deviation, n = 3 for WBs and n = 20 for the force measurements. * denotes p < 0.05 compared with control; # denotes p < 0.05 compared with 0 days.

It is known that upon actin polymerization, MRTF-A is released from G-actin and translocates to nucleus, where the MRTF-A/SRF complex affects expression of cytoskeletal genes. ³⁵ To begin to understand how the force generation capacity of microtissues could persist in the absence of NANOG, we examined actin polymerization by following the nuclear localization of MRTF-A. Briefly, the cells were treated with DOX for 5 days, then serum starved for 24 h, and stimulated with media containing 10% serum for 10 min. As shown in Figure 6A, B, MRTF-A nuclear translocation and actin polymerization remained significantly higher than the control up to 5 days after removal of DOX, suggesting that actin polymerization continued during that time. On the other hand, quantitative real-time RT-PCR also showed that the mRNA levels of ACTA2, ACTG2, TAGLN, MYH11, and MYL9 increased significantly upon NANOG expression, but decreased rapidly and reached control levels by 48 h after DOX removal (Fig. 6C–G). These data show that the transcription of myogenic genes decreased quickly upon termination of NANOG expression, but actin polymerization and therefore contractile function were sustained for several days.

OPEN IN VIEWER

FIG. 6.

MRTF-A nuclear localization, actin filament organization, and myogenic gene expression following DOX withdrawal. (A) Immunostaining for MRTF-A (green) and F-actin (red) as a function of time after DOX removal (scale bar = 100 μm). (B) Quantification of nuclear MRTFA. Intensity values are normalized to the Ctr level. (C–G) Quantitative RT-PCR for ACTA2, ACTG2, TAGLN, MYH11, and MYL9 as a function of time after DOX removal. Data are presented as mean ± standard deviation, n = 3. * denotes p < 0.05 compared with control and # denotes p < 0.05 compared with 0 days. Ctr, control.