Simulated Microgravity Disrupts Cytoskeleton Organization and Increases Apoptosis of Rat Neural Crest Stem Cells Via Upregulating CXCR4 Expression and RhoA-ROCK1-p38 MAPK-p53 Signaling

Isolation and culture of rNCSCs

Animals used in this study were purchased from the National Laboratory Animal Center of National Applied Research Laboratories (Taipei, Taiwan), and approval was received from the Institutional Animal Care and Utilization Committee in the Laboratory Animal Center of National Defense Medical Center. The procedures of isolating, culturing, and expanding rNCSCs in vitro were similar to those previously described for mouse NCSCs [90,91]. Briefly, totally three male and three female Sprague–Dawley rats at 6–8 weeks old were sacrificed by carbon dioxide euthanasia, followed by whole-body immersion in a 1:1 mixture of betadine and hydrogen peroxide for about 3 min for disinfection. After squirting the facial region with 75% ethanol, the rats were placed on a dissection microscope in a laminar flow hood, and the whisker pads were dissected and pooled in Hank's balanced salt solution (HBSS; Thermo Scientific HyClone). The whisker follicles were then cut with straight scissors without injuring hair bulbs, and loose adipose and dermal tissues were flushed out with HBSS buffer.

The clean hair follicles were pinned onto a Sylgard-coated glass petri dish and cut longitudinally with a microblade till the appearance of blood, which was then flushed out with HBSS. After making a transverse cut above the level of the cavernous sinus and another cut at the level within the ring sinus, the hair bulges were readily rolled out of the collagen capsule with a bent tungsten needle and, subsequently, washed in a new culture plate in HBSS.

After washing, the isolated hair follicles were individually cultured in the RPMI 1640 medium that was supplemented with L-glutamine (GIBCO^®, Thermo Fisher Scientific Co.), 15% fetal bovine serum (Thermo Scientific HyClone), and 4% AmnioMAX™-II complete medium (GIBCO, Thermo Fisher Scientific) in new 10-cm petri dishes at 37°C with 5% CO₂. The dermal papilla explants of hair follicles outgrew and adhered to the petri dishes. Fifty percent of the old culture medium was replaced with new freshly made medium every day. After 4–5 days, it was detected under the microscope that highly migratory cells emigrated out of the dermal papilla explants and showed stellate morphology and absence of cell–cell contacts. These cells were rNCSCs.

The emigrated rNCSCs were then digested with 0.1% trypsin-EDTA (Invitrogen, Thermo Fisher Scientific) for 2 min at 37°C, centrifuged for 5 min at 300 g, and, finally, passaged to be seeded in new petri dishes at a density of 1 × 10⁶/mL. Subsequently, rNCSCs were passaged every 3–4 days to avoid a high cell density stimulating NCSC differentiation. The rNCSCs of 10–15 passages were used for the following experiments. We found that the cell growth rate, viability, and expression levels and patterns of all protein markers analyzed in this study were indistinguishable among rNCSCs isolated from the six different Sprague–Dawley rats.

Cell counting

Total numbers of rNCSCs were counted by both the hemocytometer and the ADAM-MC automatic cell counter by NanoEnTek, Inc. Both methods obtained identical results. For counting by the hemocytometer, we diluted the culture medium containing rNCSCs for 10-fold and extracted 10 μL of it to mix with an equal volume (10 μL) of trypan blue. As apoptotic and necrotic cells had damaged plasma membranes that were permeable to trypan blue staining and, hence, were deep-blue colored, only the semitransparent and round cells were live and counted.

Cell viability analyses

The cell viability was analyzed by the Vybrant^® MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] cell viability assay kit (Thermo Fisher Scientific). Briefly, 10 μL of 12 mM MTT stock solution was added per 100 μL of freshly replaced medium containing rNCSCs, followed by incubation at 37°C for 2 h, and the reaction was terminated by adding 100 μL of the sodium dodecyl sulfate (SDS)-HCl solution and incubating at 37°C for 4 h. The relative concentration of the solubilized formazan was then determined by reading the optical absorbance at 570 nm. The viability of rNCSCs right after dissociation and centrifugation but before subculturing (into different groups) was set up as the control value (ie, 100% viability).

Encapsulation and decapsulation of rNCSCs

Alginic acid is a type of natural polysaccharides that forms a moisturizing, extracellular matrix and an eggshell-like structure on combination with calcium ions [92]. The hydrogel beads formed by alginic acid provide a three-dimensional growth environment for a high density of the encapsulated cells, and they maintain cell–cell adhesions and cell–matrix interactions mimicking the physiological condition [92 –94]. The procedures of encapsulation and decapsulation of cells both into and out of alginate hydrogel beads have been previously described [93,94]. To encapsulate rNCSCs in hydrogel beads, we mixed 1 mL of 1 × 10⁶/mL of rNCSCs homogeneously with 1 mL of a mixture of 2.2% (w/v) low-viscosity alginic acid (Sigma-Aldrich) and 0.2% (v/v) gelatin (Sigma-Aldrich) in 1× phosphate-buffered saline (PBS) solution at pH 7.4 (Thermo Fisher Scientific).

The mixed cell–gel solution was then pipetted into a 3-mL syringe (Becton, Dickinson and Company) with a 25-G needle (Becton, Dickinson and Company) and injected drop by drop by gently stirring 1 × PBS solution (pH 7.4) containing 100 mM calcium chloride (CaCl₂; Sigma-Aldrich) and 10 mM N-(2-hydroxyethyl) piperazine-N-(2-ethane sulfonic acid) (HEPES) (Sigma-Aldrich). On contact with the CaCl₂ solution, spherical hydrogel beads formed and gradually swelled to ∼2.3–2.5 mm in diameter in 6–8 min of stirring at room temperature. The beads were then washed three times in 1 × PBS and cultured in rNCSC maintenance medium (as described earlier) in the rotary cell culture system (RCCS) or the 2D culturing petri dish.

After 24–48 h of culturing, the hydrogel beads-carrying culture medium was removed from the RCCS or the 2D culturing petri dish, collected into 15-mL conical centrifuge tubes (Becton Dickinson Biosciences), and centrifuged at 300 g for 5 min. For decapsulation, the pellets of rNCSC-harboring hydrogel beads were resuspended in 1 × PBS containing 50 mM trisodium citrate dihydrate (Sigma-Aldrich), 77 mM sodium chloride (Sigma-Aldrich), and 10 mM HEPES (Sigma-Aldrich), and they were gently shaken several times for 5 min at 37°C to depolymerize the hydrogel beads. Decapsulated rNCSCs were collected by centrifugation at 300 g for 5 min and washed in 1 × PBS. rNCSC pellets were centrifuged again at 200 g for 5 min and 2D-cultured in new medium till they adhered to the new petri dish (∼in 12–24 h).

Culture of rNCSCs in the RCCS

The RCCS (Synthecon, Inc.) used in this study includes four small rotating wall vessels (RWVs) that were 0.025 m in radius and 10 mL in size. Through one hole of an RWV, 5 mL of rNCSC maintenance medium without cells was infused before injection of another 5 mL of culture medium containing rNCSC-harboring hydrogel beads, which carried totally 1 × 10⁶ cells. Meanwhile, through the other hole of the same RWV, an empty 5-mL syringe was inserted to exhaust excess air and to avoid the appearance of air bubbles in the medium. The filled 10-mL vessel was then installed onto the RCCS and set up to rotate clockwise at an angular velocity of 12 rpm, that is, 12 revolutions per minute, which equals to the angular displacement in 12 × 2π radians per 60 s, that is, ∼1.2566 rads/s.

According to previous reports [47,95], the formula for simulated microgravity (g′) is: g′ = ω ² R/g ₀, where g ₀ = 9.81 m/sec² being the Earth's gravitational acceleration, R = radius from the center of rotation, and ω = angular displacement in radians/time taken (θ/τ). Therefore, the simulated microgravity in our study is calculated as g′ = (1.2566 rads/s)² × 0.025 m/(9.81 m/s²) = 4.024 × 10⁻³.

Real-time quantitative RT-PCR of CXCR4

Total RNA was extracted from decapsulated rNCSCs immediately after culturing (in either the 2D environment or the RCCS) using the RNAqueous^®-Micro total RNA isolation kit (Ambion, Thermo Fisher Scientific) followed by reverse transcription with SuperScript^® III First-Strand Synthesis System (Invitrogen, Thermo Fisher Scientific). The synthesized cDNA was mixed with the CXCR4 primer sets (forward: 5′-CGTCGTGCACAA GTGGATCT-3′, reverse: 5′-GTTCAGGCAACAGTGGAAGA-3′) and the Platinum^® SYBR^® Green qPCR SuperMix-UDG (Invitrogen, Thermo Fisher Scientific), followed by real-time quantitative PCRs using the Applied Biosystems® 7500 Fast Real-time PCR System (Applied Biosystems Group, Thermo Fisher Scientific) located in the Instrument Center of National Defense Medical Center.

The cDNA of 18S rRNA was amplified by the QuantumRNA^™ Classic 18S Internal Standard kit (Ambion, Thermo Fisher Scientific) and used as a normalization control for CXCR4 expression. The cycling program was set up as a 2 min hold at 50°C for UDG incubation followed by 2 min at 95°C and 40 cycles of 95°C for 3 s and of 60°C for 30 s. The expected sizes of the RT-PCR products for CXCR4 and 18S rRNA are 60 and 99 base pairs, respectively.

The results of real-time PCR were analyzed by the DDCt method based on the cycle threshold (Ct) values. The mean Ct values for CXCR4 were normalized against the mean Ct values for 18S rRNA from the same samples, and the quantitative CXCR4 expression was calculated using 2^−ΔΔCt, where ΔΔCt = ΔCt^CXCR4 − ΔCt^{18S rRNA}. The relative levels of CXCR4 expression in different samples were calculated in comparison with the expression level in 2D-cultured rNCSCs, which was presumptively set up as 1.0.

CXCR4 siRNA transfection

The oligonucleotides of CXCR4 siRNA, negative control siRNA, and positive control (GAPDH) siRNA were all obtained from Thermo Fisher Scientific. The sequence of the 21-mer antisense CXCR4 siRNA was 5′-UCGUAAUACGGUAGCCAGCAGdTdT-3′, corresponding to nucleotides 808 − 828 of the NCBI reference sequence NM_022205. We used the Silencer ^® GAPDH siRNA (Thermo Fisher Scientific) for rat as a positive control and the Silencer Negative Control siRNA (Thermo Fisher Scientific) as a negative control. Next, 3 × 10⁵ cells/mL of rNCSCs seeded in six-well culture plates (Nunc, Thermo Scientific) were grown to be 70%−90% confluent, washed, and transferred to serum-free medium, followed by transfection with 2.5 μg of siRNA per well using the Lipofectamine® 2000 Reagent kit (Invitrogen, Thermo Fisher Scientific) in accordance with the manufacturer's instructions.

After 24 h, the siRNA-transfected rNCSCs were washed and transferred to complete maintenance medium, and they were then cultured separately under 2D and 3D (RCCS) conditions at a density of 1 × 10⁶ cells/mL.

Pertussis toxin treatment

Pertussis toxin (PTX) has been shown to inhibit the activity of the α subunit of the heterotrimeric G protein (G_α) via catalyzing ADP-ribosylation and inducing structural changes of G_α [96,97]. We incubated rNCSCs in the maintenance medium at 37°C for 4 h in the presence of 100, 250, 500, or 1,000 ng/mL of PTX (Calbiochem, Merck Millipore), respectively, before culturing in the RCCS. We found that 500 ng/mL of PTX was sufficient to inhibit more than 99% of RhoA GTPase activation and neural differentiation of rNCSCs under simulated microgravity, whereas 1,000 ng/mL of PTX exhibited cell toxicity and greatly decreased the number of rNCSCs. Therefore, we chose to use the concentration of 500 ng/mL of PTX throughout this study.

Immunofluorescence staining

For immunostaining, 1.5 × 10⁵ cells/mL of decapsulated and petri dish-adhered rNCSCs were seeded in 12-well culture plates (Nunc, Thermo Scientific) and cultured for 12 h (till all cells have adhered to the bottom of the plates) in the maintenance medium described earlier. The medium was then aspirated carefully without touching the rNCSCs, which were adhered to the bottom of the plate. The cells were then washed three times in 1 × PBS at pH 7.4, for 5 min each time, followed by fixation in 1 mL of 4% paraformaldehyde (PFA) (Sigma-Aldrich) in PBS for 20 min.

After washing again with 1 × PBS three times, the fixed cells were permeabilized with 0.5% Triton X-100 for 10 min, then blocked with 10% bovine serum albumin (BSA; Sigma-Aldrich) in 1 × PBS for 1 h, to prevent high background signals of immunostaining from non-specific binding of primary and secondary antibodies.

The rNCSCs were then incubated with the following primary antibodies or conjugates that were diluted in 1 × PBS for overnight at 4°C: anti-cleaved caspase-3 (goat polyclonal, 1:100 dilution; Santa Cruz Biotechnology), anti-cortactin (rabbit polyclonal, 1:250 dilution; Abcam), anti-CXCR4 (rabbit polyclonal, 1:100 dilution; Abcam), anti-FoxD3 (rabbit polyclonal, 1:100 dilution; Santa Cruz Biotechnology), anti-microtubule-associated protein 2 (MAP2; mouse monoclonal with the clone number HM-2; 1:500 dilution; Abcam), anti-p21^WAF1/Cip1 (mouse monoclonal with the clone number CP74, 1:100 dilution; Sigma-Aldrich), anti-p38-α MAP Kinase (mouse monoclonal with the clone number p38-3F11, 1:500 dilution; Invitrogen, Thermo Fisher Scientific), anti-phospho-p38 MAPK (pT180/pY182) (rabbit polyclonal, 1:1,000 dilution; Invitrogen, Thermo Fisher Scientific), anti-p53 (mouse monoclonal with the clone number PAb 122, 1:1,000 dilution; Invitrogen, Thermo Fisher Scientific), anti-p75 (rabbit polyclonal, 1:400 dilution; Abcam), anti-active RhoA GTPase (mouse monoclonal with the clone number 26904, 1:100 dilution; NewEast Biosciences), anti-ROCK1 (mouse monoclonal with the clone number B-1, 1:100 dilution; Santa Cruz Biotechnology), anti-phospho-ROCK1 (pS1333) (rabbit polyclonal, 1:100 dilution; GeneTex), anti-Snai1 (mouse monoclonal with the clone number G-7, 1:100 dilution; Santa Cruz Biotechnology), anti-Sox10 (rabbit polyclonal, 1:1,000 dilution; Abcam), anti-Tuj1 (mouse monoclonal with the clone number TU-20, 1:1,000 dilution; Abcam), anti-vimentin (mouse monoclonal with the clone number RV202, 1:500 dilution; Abcam), bovine pancreatic deoxyribonuclease I conjugated to Alexa Fluor^® 594 (1:100 dilution; Molecular Probes, Thermo Fisher Scientific), and phalloidin conjugated to Alexa Fluor^® 488 (1:40 dilution; Molecular Probes, Thermo Fisher Scientific).

For isotype negative controls, mouse monoclonal IgG1 (with the clone number NCG01), rabbit polyclonal IgG, and goat polyclonal IgG from Abcam were applied at the same concentration as the respective primary antibodies. On the second day, the cells were washed three times in 1 × PBS and blocked for 20 min with 10% BSA in 1 × PBS, followed by incubation at room temperature for 1 h with the following secondary antibodies conjugated with fluorophores: donkey anti-mouse Alexa Fluor 488, donkey anti-mouse Alexa Fluor^® 555, donkey anti-rabbit Alexa Fluor 488, donkey anti-rabbit Alexa Fluor 555, donkey anti-goat Alexa Fluor 555, and donkey anti-goat Alexa Fluor^® 350 (Molecular Probes, Thermo Fisher Scientific).

After 1 h, the cells were washed in 1 × PBS again, followed by counter-staining with 10 μg/mL of DAPI (1:1,000 dilution; Molecular Probes, Thermo Fisher Scientific) in 1 × PBS for 10 min. Subsequently, the cells were washed and stored in 1 × PBS, followed by photographing under the AXIO observer inverted fluorescence microscope (Zeiss). In line with the fluorescence intensities of different fluorochromes, the exposure time of the AxioCam HRc color CCD camera was set as 300 ms for DAPI, 360 ms for Alexa Fluor 488, and 400 ms for Alexa Fluor 555. Both isotype negative controls and no primary antibody controls showed no specific staining with only very few background signals (Supplementary Fig. S1 and data not shown; Supplementary Data are available online at www.liebertpub.com/scd).

For analysis of cell proliferation, the Click-iT™ EdU Alexa Fluor 555 Imaging Kit (Invitrogen and Molecular Probes, Thermo Fisher Scientific) was used. Briefly, 1.5 × 10⁵ cells/mL of decapsulated and petri dish-adhered rNCSCs were cultured in the 1:1 mixture of the maintenance medium and 20 μM solution of EdU for 12 h in 12-well culture plates (Nunc, Thermo Scientific), followed by fixation in 4% PFA and permeabilization with 0.5% Triton X-100, same as described earlier in the immunostaining procedure. After washing with 1 × PBS containing 3% BSA, 0.5 mL of Click-iT reaction cocktail was added to each well and incubated for 30 min with protection from light. After EdU labeling, the rNCSCs were washed three times in 1 × PBS and blocked for 20 min with 10% BSA in 1 × PBS, followed by immunostaining with primary antibodies.

For analysis of cell apoptosis, the Click-iT^® TUNEL Alexa Fluor 488 Imaging Kit (Invitrogen and Molecular Probes, Thermo Fisher Scientific) was used. Briefly, the decapsulated and petri dish-adhered rNCSCs were fixed and permeabilized as described earlier in the immunostaining procedure, followed by washing twice with deionized and distilled water and incubation with TdT reaction buffer for 10 min at room temperature.

After removing the TdT reaction buffer, the TdT reaction cocktail was added to each well and incubated for 1 h at 37°C, followed by washing twice with 1 × PBS containing 3% BSA. The rNCSCs were then incubated with the Click-iT reaction cocktail for 30 min at room temperature with protection from light, followed by washing in 1 × PBS three times, blocking with 10% BSA in 1 × PBS for 20 min, and finally immunostaining with primary antibodies.

Quantitative and statistical analyses

The image files obtained from the Zeiss fluorescence microscope were input into the Count Nuclei/Cell Sorting Application Module for MetaMorph (MetaMorph Offline vers. 7.0; Universal Imaging Corporation™) for quantitative analyses and comparison among the following groups of rNCSCs: 2D-cultured, RCCS-cultured, RCCS-cultured with PTX treatment, and RCCS-cultured with CXCR4 siRNA transfection. Briefly, for cell counting, the numbers of rNCSCs or nuclei stained with different colors of fluorescence (red or green) and the total numbers of DAPI-stained nuclei were, respectively, counted on the image files of each experimental group.

The statistical histogram in Supplementary Fig. S2 listed the total cell numbers in each group of rNCSCs counted for immunostaining signals of each protein marker in each well of the 12-well culture plates [n = 12, P > 0.05, one-way analysis of variance (ANOVA)]. The average percentages of positive rNCSCs or nuclei in each experimental group were calculated as follows: (the summary of numbers of cells or nuclei stained with red or green fluorescence on all images)/(the summary of total numbers of DAPI⁺ nuclei on all images) × 100%.

For comparison of fluorescent intensities, a specified unit of area (eg, a defined range of the cytoplasm or nucleus) in the 2D-cultured rNCSCs was set up to quantify the fluorescent intensity within the area, which was recorded as a standard value, and the same unit of area was picked up from rNCSCs in the other experimental groups (ie, the RCCS-cultured, PTX-treated, or CXCR4 siRNA-transfected groups), followed by quantification of fluorescent intensity within the area and comparison with the standard value. The relative intensity of a specific marker within an experimental group was calculated by the summation of relative values obtained from all units of areas on all image files within that group, and it was quantified as a fold change of the standard value.

For statistical analyses, we performed one-way ANOVA with post hoc tests of Duncan's multiple-range test and Scheffé's method, which analyzed both the within-group variation and the between-group variation and calculated the relative ratios by use of the SPSS 20.2 software (International Business Machines Corporation). The data of total cell numbers after 24-h culture, relative CXCR4 mRNA expression levels, and positive cell percentages for different markers were all analyzed and compared both within and between groups by one-way ANOVA.

The groups of rNCSCs compared on total cell numbers after 24-h culture include: 2D-cultured without hydrogel beads, 2D-cultured without hydrogel beads and with PTX treatment, 2D-cultured without hydrogel beads and with CXCR4 siRNA transfection, 2D-cultured in hydrogel beads, 2D-cultured in hydrogel beads with PTX treatment, 2D-cultured in hydrogel beads with CXCR4 siRNA transfection, RCCS-cultured (in hydrogel beads), RCCS-cultured (in hydrogel beads) with PTX treatment, and RCCS-cultured (in hydrogel beads) with CXCR4 siRNA transfection (n = 24 in each group). The groups compared on relative CXCR4 mRNA expression levels include: blank control, 2D-cultured, RCCS-cultured, PTX-treated RCCS-cultured, CXCR4 siRNA-transfected RCCS-cultured, and negative siRNA-transfected rNCSCs (n = 12 in each group).

The groups of rNCSCs compared on positive cell percentages for different markers include: 2D-cultured, RCCS-cultured, RCCS-cultured with PTX treatment, and RCCS-cultured with CXCR4 siRNA transfection (n = 12 in each group).

Western blot analyses of cytoplasmic and nuclear protein extracts

The cytoplasmic and nuclear protein fractions were sequentially extracted from decapsulated rNCSCs immediately after culturing (in either the 2D environment or the RCCS) with the NE-PER™ Nuclear and Cytoplasmic Extraction Reagents (Thermo Fisher Scientific), according to the manufacturer's instructions. Briefly, 1 × 10⁶ rNCSCs were trypsinized and centrifuged in a 1.5-mL microcentrifuge tube, lysed with the ice-cold CER I and CER II reagents sequentially, followed by the second centrifugation to obtain supernatants containing the cytoplasmic extracts. The pellet deposition after the second centrifugation was further lysed by the ice-cold NER reagent and centrifuged again to obtain supernatants containing the nuclear extracts.

Equal amounts of cytoplasmic or nuclear protein extracts from each group of rNCSCs were separated using SDS–10% polyacrylamide gel electrophoresis and transblotted onto polyvinylidene difluoride (PVDF) membranes (Millipore). Protein concentration was quantified using the Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific).

Immunoblotting was performed with the same primary antibodies applied for immunofluorescence staining (described earlier). In addition, the anti-β-actin antibody (rabbit polyclonal, 1:100 dilution; Santa Cruz Biotechnology) was applied to detect levels of total actin (including both F-actin and G-actin) in rNCSCs. GAPDH and TATA-box binding protein (TBP), respectively, served as the loading controls of cytoplasmic and nuclear extracts, and they were blotted with the anti-GAPDH (rabbit polyclonal, 1:2,500 dilution; Abcam) and anti-TBP (mouse monoclonal with the clone number mAbcam 51841, 1:1,000 dilution; Abcam) antibodies. For negative controls, the PVDF membranes carrying the cytoplasmic or nuclear protein extracts from different groups of rNCSCs were, respectively, immunoblotted with mouse monoclonal IgG1, rabbit polyclonal IgG, and goat polyclonal IgG from Abcam.

The signals of immunoblots were visualized with the Amersham ECL Western Blotting Detection Kit (GE Healthcare Life Sciences) followed by exposure to X-ray films. The intensities of the immunoblots signals on the X-ray films were then quantified using the ImageQuant^® software with the Molecular Dynamics 375/475 Personal Densitometer SI (GE Healthcare Life Sciences).

Altered morphology of rNCSCs and decreased adherence and survival of rNCSCs with the increase of culturing period under simulated microgravity

We separated hydrogel bead-encapsulated rNCSCs (Supplementary Fig. S3A) into two groups, both of which contained the same cell number at initiation (ie, 1 × 10⁶ cells/mL) and were cultured in the same incubator, with one group cultured under normal gravity in the 2D environment and the other group cultured in the RCCS (Supplementary Fig. S3B) for 12, 24, and 48 h, respectively. The total numbers of rNCSCs cultured in the 2D environment with and without hydrogel encapsulation and rNCSCs cultured in the RCCS were counted after 12, 24, and 48 h (Supplementary Fig. S4A). The doubling time of rNCSCs was estimated to be 21.30 ± 4.58 h.

Interestingly, the rNCSCs cultured without hydrogel encapsulation displayed significantly increased total cell numbers per milliliter at all time points compared with the rNCSCs encapsulated in hydrogel beads (Supplementary Fig. S4A). After decapsulation from the hydrogel beads and subculturing in the new maintenance medium for 2–3 h, all of the 2D-cultured rNCSCs adhered to the new petri dish and appeared in a long spindle shape (Supplementary Fig. S3C); whereas the RCCS-cultured rNCSCs were still suspended and appeared in a round shape (Supplementary Fig. S3D).

After culturing in the RCCS for 12 h and decapsulation, it took at least 12 h for the subcultured rNCSCs to adhere to the new petri dish. After culturing in the RCCS for 24 h and decapsulation, 90% of the rNCSCs adhered to the new petri dish after 24 h of subculturing. On the other hand, after culturing in the RCCS for 48 h and decapsulation, only 47.3% ± 8.6% of the rNCSCs adhered to the new petri dish after 24 h of subculturing. Trypan blue staining of the suspension after 24 h of subculturing indicated that 91.5% ± 7.6% of the suspended rNCSCs were dead.

Cell viability analyses conducted by the MTT assay indicated that the rNCSCs cultured in the 2D environment without hydrogel encapsulation showed a significantly higher viability (93.88%–87.76%) than the rNCSCs encapsulated in hydrogel beads after 12, 24, and 48 h of culturing (n = 24, P < 0.05 after 12 h, P < 0.01 after 24 and 48 h, one-way ANOVA), regardless of normal gravity or simulated microgravity (Supplementary Fig. S4B). On the other hand, the rNCSCs cultured in the RCCS for 48 h without any pretreatment displayed a significantly lower viability (35.75%) than the rNCSCs in the other groups (n = 24, P < 0.01, one-way ANOVA) (Supplementary Fig. S4B). Therefore, culturing in the RCCS for 48 h was harmful to rNCSCs and was, thus, excluded from subsequent analyses.

Decreased expression of neural crest markers and increased expression of neurite markers and active caspase-3 in rNCSCs cultured under simulated microgravity for 24 h

By immunofluorescence staining against four well-characterized neural crest markers, that is, p75 neurotrophin receptor and neural crest-specific transcription factors FoxD3, Snail(Snai1), and Sox10 [98 –103], we confirmed that all rNCSCs cultured under normal gravity (ie, 2D-cultured) expressed all of these neural crest markers (Fig. 1A–D). On the other hand, only 30.9% ± 10.4% of the rNCSCs cultured under simulated microgravity (ie, RCCS-cultured) expressed these neural crest markers (Fig. 1E–H). Interestingly, under the microscope, we found extension of neurite-like structures surrounding the RCCS-cultured rNCSCs that did not express neural crest markers (indicated by arrowheads in Fig. 1E–H).

OPEN IN VIEWER

FIG. 1.

Decreased expression of neural crest markers in the nuclei and appearance of neurite-like structures in the RCCS-cultured rNCSCs. Immunostaining against neural crest markers FoxD3, p75, Snai1, and Sox10 (green) revealed that all 2D-cultured rNCSCs expressed all four neural crest markers analyzed (A–D), whereas more than 50% of the RCCS-cultured rNCSCs did not express neural crest markers and showed the appearance of neurite-like structures (E–H). Note that the neurite-like structures were present in the rNCSCs with negative staining of neural crest markers (indicated by arrowheads in E–H). The scale bar in (A) applies to panels (A–D), and the scale bar in (E) applies to panels (E–H). Both the scale bars in (A) and (E) represent 50 μm. RCCS, rotary cell culture system; rNCSCs, rat neural crest stem cells.

Because it was demonstrated that culturing in alternative conditions might cause stress-induced formation of neurite-like structures on adult MSCs [104], we performed triple immunostaining for the following markers: the neurite marker neuron-specific class III β-tubulin (Tuj1) or MAP2 [105 –107], and the neural crest marker p75 or Sox10, as well as cleaved caspase-3 subunits, including the immature p20 peptide and the mature p17 subunit [108,109]. In the 2D-cultured rNCSCs, none expressed Tuj1 or MAP2, and 35.2% ± 10.7% expressed cleaved caspase-3 (Figs. 2A, E and Fig. 3).

OPEN IN VIEWER

FIG. 2.

Decreased expression of neural crest markers and increased expression of neurite markers and cleaved caspase-3 in the RCCS-cultured rNCSCs. Triple immunostaining against a neural crest marker (p75 or Sox10; green), a neurite marker (Tuj1 or MAP2; blue), and cleaved caspase-3 (red) revealed significantly decreased numbers of p75⁺ and Sox10⁺ cells whereas dramatically increased numbers of Tuj1⁺, MAP2⁺, and cleaved caspase-3⁺ cells in the untreated RCCS-cultured rNCSCs (B, F), in comparison with the 2D-cultured rNCSCs (A, E), the PTX-pretreated RCCS-cultured rNCSCs (C, G), and the CXCR4 siRNA-pretransfected RCCS-cultured rNCSCs (D, H). In the untreated RCCS-cultured group, not all p75⁻ or Sox10⁻ rNCSCs expressed Tuj1 or MAP2, and not all Tuj1⁺ or MAP2⁺ rNCSCs expressed cleaved caspase-3; whereas all cleaved caspase-3⁺ cells expressed Tuj1 or MAP2 and exhibited neurite-like structures (indicated by arrows in B, F). The scale bar in (A) applies to panels (A–D), and the scale bar in (E) applies to panels (E–H). Both the scale bars in (A) and (E) represent 50 μm. CXCR4, C-X-C chemokine receptor 4; MAP2, microtubule-associated protein 2; PTX, pertussis toxin.

OPEN IN VIEWER

FIG. 3.

Statistical histogram of average percentages of positive cells for different protein markers. The numbers of total DAPI-stained nuclei and the positive cells immunostained for different protein markers or labeled with EdU or TUNEL in rNCSCs cultured under different conditions were individually counted by the Count Nuclei/Cell Sorting Application Module for MetaMorph Offline vers. 7.0. The average percentages of positive cells for each marker were calculated as “(the summary of numbers of cells or nuclei stained with red or green fluorescence on all images)/(the summary of total numbers of DAPI⁺ nuclei on all images) × 100%” for each group of rNCSCs. One-way ANOVA with n = 12 was applied for each marker analyzed. *P < 0.05 for p53⁺ and cleaved caspase-3⁺ cells; **P < 0.01 for MAP2⁺, Tuj1⁺, CXCR4⁺, active RhoA⁺, and phospho-p38⁺ cells. ANOVA, analysis of variance.

Interestingly, 68.9% ± 9.5% of the RCCS-cultured rNCSCs co-expressed Tuj1 and cleaved caspase-3, and 72.5% ± 12.7% of the RCCS-cultured rNCSCs co-expressed MAP2 and cleaved caspase-3, indicating that the cells with neurite-like structures were undergoing apoptosis (Figs. 2B, F and 3). The percentage of active caspase-3-positive nuclei in the RCCS-cultured rNCSCs was significantly higher than that in the 2D-cultured rNCSCs (Fig. 3; n = 12, P < 0.01, one-way ANOVA). Western blot analyses further confirmed the marked increase of MAP2 and Tuj1 expression and the dramatic decrease of FoxD3, p75, Snai1, and Sox10 expression in the RCCS-cultured rNCSCs compared with the 2D-cultured rNCSCs (Supplementary Figs. S5 and S6).

Plasma membrane blebbing and cytoskeleton disorganization in the RCCS-cultured rNCSCs

As microgravity environment has been reported to affect cytoskeleton organization and/or actin polymerization in a variety of cells, including breast cancer cells, endothelial cells, epithelial cells, fibroblasts, glial tumor cells, lymphocytes, MSCs, osteoblasts, utricular hair cells, etc. [4,27,42,44 –48], it was of interest to analyze cytoskeleton organization in the RCCS-cultured rNCSCs. Phalloidin staining demonstrated an organized and a nearly parallel arrangement of F-actin in the 2D-cultured rNCSCs under normal gravity (Fig. 4A); whereas in the RCCS-cultured rNCSCs under simulated microgravity for 24 h, the actin filaments were generally shorter and disorganized with membrane ruffling (indicated by arrows in Fig. 4B), though the overall immunostaining intensities of F-actin were comparable between the 2D-cultured and RCCS-cultured rNCSCs (Fig. 4A, B).

OPEN IN VIEWER

FIG. 4.

Disorganized F-actin, membrane blebbing, and an increased level of G-actin in the RCCS-cultured rNCSCs. Phalloidin staining revealed that F-actin was organized in an extended and a nearly parallel pattern under normal gravity in the 2D-cultured rNCSCs (A), and under simulated microgravity in the PTX-pretreated (C) or CXCR4 siRNA-pretransfected (D) rNCSCs. However, in the untreated RCCS-cultured rNCSCs (B), F-actin appeared much shorter and disorganized, and membrane blebbing or ruffling was observed (indicated by arrows). Deoxyribonuclease staining revealed a dramatically increased level of G-actin in the RCCS-cultured rNCSCs (B) compared with the 2D-cultured (A), PTX-pretreated (C), or CXCR4 siRNA-pretransfected (D) rNCSCs. The scale bar in (A) represents 50 μm and applies to panels (A–D). F-actin, filamentous actin; G-actin, globular actin.

On the other hand, deoxyribonuclease staining revealed a greatly increased level of G-actin in the RCCS-cultured rNCSCs compared with that in the 2D-cultured rNCSCs (compare Fig. 4A, B), which was consistent with the previously reported increase of the G-actin level in microgravity-stimulated human MSCs [44,48]. Interestingly, western blot analyses revealed that the cytoplasmic level of total β-actin in the RCCS-cultured rNCSCs was approximately twofold of that in the 2D-cultured rNCSCs (Supplementary Fig. S5), indicating that the total actin level was greatly increased after culturing under simulated microgravity.

Among a variety of actin-binding proteins in the cytoskeleton, cortactin has been characterized to be an important molecular scaffold and a mediator for actin polymerization and organization [110,111], and vimentin, which is an intermediate filament protein required for neuritogenesis [112 –114], was shown to be down-regulated in adult rat MSCs during stress-induced formation of neurite-like structures [104,115]. Therefore, we also analyzed the expression of cortactin and vimentin in the 2D-cultured and RCCS-cultured rNCSCs.

Interestingly, both cortactin and vimentin showed decreased expression and perturbed organization in the RCCS-cultured rNCSCs (Fig. 5B), in contrast to the strong expression and well-organized pattern observed in the 2D-cultured rNCSCs (Fig. 5A). Our results of western blotting also indicated dramatically reduced protein levels of cortactin and vimentin in the RCCS-cultured rNCSCs (Supplementary Fig. S5).

OPEN IN VIEWER

FIG. 5.

Decreased expression of cortactin and vimentin in the RCCS-cultured rNCSCs. Double immunostaining revealed that the expression levels of both cortactin and vimentin were significantly decreased in the untreated RCCS-cultured rNCSCs (B) compared with the 2D-cultured rNCSCs (A), and PTX-pretreated (C) or CXCR4 siRNA-pretransfected (D) RCCS-cultured rNCSCs. In addition to the decreased expression levels, both cortactin and vimentin showed disorganized patterns in the untreated RCCS-cultured rNCSCs (B). The scale bar in (A) represents 50 μm and applies to panels (A–D).

Increased levels of CXCR4, active RhoA GTPase, and phosphorylated ROCK1 and p38 MAPK in the RCCS-cultured rNCSCs

Previous studies have demonstrated that the intracellular trafficking and signal transduction of the CXCR4, which regulates migration of neural crest cells [116 –118], are mediated by RhoA GTPase [119 –121], which regulates actin polymerization and is down-regulated by modeled microgravity in human MSCs [44,48,86,122,123]. Therefore, we analyzed protein levels and distributions of CXCR4 and active RhoA GTPase in the 2D-cultured and RCCS-cultured rNCSCs by immunostaining.

In the 2D-cultured rNCSCs, CXCR4 was mainly distributed in the perinuclear region in the cytoplasm, whereas active RhoA was detected merely at a background level (Fig. 6A). In the RCCS-cultured rNCSCs, the immunostaining signals of both CXCR4 and active RhoA were dramatically increased and colocalized both in the cytoplasm and on the plasma membranes (indicated by arrows in Fig. 6B). Western blot analyses indicated that the relative protein levels of CXCR4 and active RhoA in the RCCS-cultured rNCSCs were, respectively, 3.7- and 7.7-fold of the levels in the 2D-cultured rNCSCs.

OPEN IN VIEWER

FIG. 6.

Increased levels of CXCR4 and active RhoA GTPase in the RCCS-cultured rNCSCs. Double immunostaining revealed only background levels of CXCR4 and active RhoA GTPase in the 2D-cultured rNCSCs (A) and CXCR4 siRNA-pretransfected RCCS-cultured rNCSCs (D). In more than 50% of the untreated RCCS-cultured rNCSCs (B), however, the levels of both CXCR4 and active RhoA GTPase were dramatically increased and colocalized on the plasma membranes and in the cytoplasm (indicated by arrows; n = 12, P < 0.01, one-way ANOVA; see also Fig. 3). In PTX-pretreated RCCS-cultured rNCSCs (C), the expression level of CXCR4 was comparable with that in the untreated RCCS-cultured rNCSCs (B) (indicated by arrows); whereas the level of active RhoA was significantly lower than that in the untreated RCCS-cultured rNCSCs (B) and more comparable with that in the 2D-cultured rNCSCs (A) and CXCR4 siRNA-pretransfected RCCS-cultured rNCSCs (D), indicating that PTX pretreatment inhibited RhoA activation but not CXCR4 expression. The scale bar in (A) represents 50 μm and applies to panels (A–D).

We then analyzed mRNA levels of CXCR4 in 2D-cultured and RCCS-cultured rNCSCs by real-time quantitative RT-PCR, with the vehicle blank control, siRNA negative control, and internal standard control of 18S rRNA extracted from each group of rNCSCs (Fig. 7A). When the average CXCR4 mRNA level in the 2D-cultured rNCSCs was presumptively set up as 1.0, the relative CXCR4 mRNA level in the rNCSCs cultured in the RCCS for 24 h was 3.46 ± 0.87, which was significantly increased (n = 18, P < 0.01, one-way ANOVA) (Fig. 7B). Although the relative CXCR4 mRNA level in the rNCSCs cultured in the RCCS for 12 h was 1.57 ± 0.64-fold of that in the 2D-cultured rNCSCs, the difference was not statistically significant (n = 18, P > 0.05, one-way ANOVA) (Fig. 7B).

OPEN IN VIEWER

FIG. 7.

Increased expression of CXCR4 mRNA in the RCCS-cultured rNCSCs. (A) A representative gel electrophoresis result of the RT-PCR products of 18S rRNA and CXCR4 cDNA revealed a significantly increased mRNA level of CXCR4 relative to 18S rRNA (the internal standard) in the untreated and PTX-pretreated rNCSCs after culturing in the RCCS for 24 h (respectively labeled with “RCCS 24h” and “RCCS 24h + PTX” on the top), in comparison with the blank control, 2D-cultured rNCSCs (labeled with “2D culture” on the top), rNCSCs cultured in the RCCS for 12 h (labeled with “RCCS 12h” on the top), CXCR4 siRNA-pretransfected RCCS-cultured rNCSCs (labeled with “RCCS 24h + CXCR4 siRNA” on the top), and the siRNA negative control. The relative CXCR4 mRNA levels were quantitated and shown in the histogram in (B), with the mRNA level in the 2D-cultured rNCSCs presumptively set up as 1.0. It is noteworthy that, although the relative CXCR4 mRNA level in the rNCSCs cultured in the RCCS for 12 h was 1.57 ± 0.64-fold of that in the 2D-cultured rNCSCs, the difference was not statistically significant (ie, P > 0.05). One-way ANOVA with n = 18 was applied, and double asterisks (**) indicated P < 0.01.

These results indicated that CXCR4 was not significantly upregulated at the transcriptional level in the rNCSCs until culturing under simulated microgravity for 24 h.

Because active RhoA GTPase binds to and activates the Rho-associated protein kinases ROCK1 and ROCK2 [124 –126], and ROCK1 rather than ROCK2 has been shown to play an essential role in stress-induced actin destabilization [127,128], we then examined the levels of activated ROCK1 in the 2D-cultured and RCCS-cultured rNCSCs. We analyzed the total and activated (Ser1333-phosphorylated) ROCK1 proteins by immunostaining [129]. In the 2D-cultured rNCSCs under normal gravity, the total ROCK1 was distributed nearly homogenously in the cytoplasm, whereas the Ser1333-phosphorylated ROCK1 was detected predominantly in the perinuclear region (Fig. 8A).

OPEN IN VIEWER

FIG. 8.

Increased phosphorylation of ROCK1 in the RCCS-cultured rNCSCs. Double immunostaining revealed that, in the 2D-cultured rNCSCs (A), and PTX-pretreated (C) or CXCR4 siRNA-pretransfected (D) RCCS-cultured rNCSCs, total ROCK1 was detected nearly homogenously at low levels in the cytoplasm; whereas phosphorylated ROCK1 was mainly detected in the perinuclear region. Interestingly, the levels of both total and phosphorylated ROCK1 were dramatically increased on the plasma membranes in more than 50% of the untreated RCCS-cultured rNCSCs (indicated by arrows in B). The scale bar in (A) represents 50 μm and applies to panels (A–D).

On the other hand, in more than 50% of the RCCS-cultured rNCSCs, the levels of both the total and phosphorylated ROCK1 were significantly increased and distributed predominantly on the plasma membranes (indicated by arrows in Fig. 8B). Translocation of ROCK1 to the plasma membranes and increased ROCK1 phosphorylation in RCCS-cultured rNCSCs may be concomitant with increased RhoA activity (Fig. 6B), as previous studies have demonstrated that plasma membrane translocation is a prerequisite of ROCK1 activation by RhoA GTPase [130,131]. Western blot analyses indicated that the relative protein levels of total ROCK1 and phosphorylated ROCK1 in the RCCS-cultured rNCSCs were, respectively, 1.9- and 5.2-fold of the levels in the 2D-cultured rNCSCs.

Because both RhoA and ROCK1 were activated in the rNCSCs cultured under simulated microgravity, we were interested in analyzing the total and phosphorylation levels of p38 MAPK, which was a major downstream effector of ROCK1 [132,133], in the 2D-cultured and RCCS-cultured rNCSCs. The total p38 MAPK was detected in either the cytoplasm or nuclei of rNCSCs, whereas the Thr180/Tyr182-phosphorylated and activated p38 was restricted to the nuclei (Fig. 9), as demonstrated by previous studies [134,135]. The summations of immunostaining intensities of the total p38 MAPK were comparable between the 2D-cultured and RCCS-cultured rNCSCs (compare Fig. 9A, B).

OPEN IN VIEWER

FIG. 9.

Increased phosphorylation and nuclear localization of p38 MAPK in the RCCS-cultured rNCSCs. Double immunostaining against total and phosphorylated p38 MAPK revealed that phosphorylated p38 MAPK was located only in the nucleus, with p38⁺ nuclei coinciding with phospho-p38⁺ nuclei; whereas cytoplasmic p38 MAPK was unphosphorylated. The percentages of both p38⁺ and phospho-p38⁺ nuclei were significantly increased in the RCCS-cultured rNCSCs (B) in comparison with those in the 2D-cultured rNCSCs (A), and PTX-pretreated (C) or CXCR4 siRNA-pretransfected (D) RCCS-cultured rNCSCs (n = 12, P < 0.01, one-way ANOVA; see also Fig. 3). The scale bar in (A) represents 50 μm and applies to panels (A–D). MAPK, mitogen-activated protein kinase.

On the other hand, both the intensity and the percentage of nuclei positive for Thr180/Tyr182-phosphorylated p38 were significantly higher in the RCCS-cultured rNCSCs than in the 2D-cultured rNCSCs (Figs. 3 and 9; n = 12, P < 0.01, one-way ANOVA). Western blot analyses indicated that the relative protein levels of total cytoplasmic p38 MAPK and phosphorylated nuclear p38 MAPK in the RCCS-cultured rNCSCs were, respectively, 0.5- and 4.9-fold of the levels in the 2D-cultured rNCSCs. These results indicated increased p38 MAPK activation in rNCSCs cultured under simulated microgravity for 24 h.

PTX treatment or CXCR4 siRNA transfection restored cell viability, cell morphology, cytoskeleton organization, and RhoA-ROCK1-p38 signaling in the RCCS-cultured rNCSCs

Given that both the RhoA activity and CXCR4 expression were significantly increased in the RCCS-cultured rNCSCs (Figs. 6B and 7), we were interested in analyzing the effects of inhibiting RhoA activation or CXCR4 expression on cytoskeleton disorganization and signaling perturbation in the RCCS-cultured rNCSCs.

For RhoA inhibition, before culturing in the RCCS, we incubated rNCSCs in the maintenance medium containing, respectively, 100, 250, 500, or 1,000 ng/mL of PTX, which is an ADP-ribosylating factor and inhibitor of the Rho family of GTPases [136,137]. We found that 500 ng/mL of PTX was sufficient to inhibit more than 99% of neurite marker expression (Figs. 2C, G and 3) and RhoA GTPase activation (Figs. 3 and 6C) in the RCCS-cultured rNCSCs, whereas 1,000 ng/mL of PTX exhibited cell toxicity and greatly decreased the total number of rNCSCs (data not shown). Therefore, PTX at the concentration of 500 ng/mL was applied for all analyses in this study. We found that PTX pretreatment significantly increased the total numbers and cell viability of rNCSCs after culturing in the RCCS for 24 and 48 h to comparable levels as in the 2D-cultured rNCSCs (Supplementary Fig. S4).

PTX pretreatment also increased the percentages of p75⁺ and Sox10⁺ rNCSCs but decreased the percentages of Tuj1⁺, MAP2⁺, and active caspase-3⁺ rNCSCs after RCCS culturing to comparable levels as in the 2D-cultured rNCSCs (Figs. 2C, G and 3). Western blotting results also showed that, in the cytoplasm of the PTX-pretreated and RCCS-cultured rNCSCs, FoxD3, p75, Snai1, and Sox10 expression was increased; whereas MAP2 and Tuj1 expression was decreased to comparable levels as in the cytoplasm of the 2D-cultured rNCSCs (Supplementary Figs. S5 and S6).

In addition, PTX pretreatment suppressed the G-actin and total β-actin levels as well as plasma membrane blebbing (Fig. 4C and Supplementary Fig. S5), increased the cortactin and vimentin levels (Fig. 5C and Supplementary Fig. S5), decreased RhoA GTPase activation (Figs. 3, 6C, and Supplementary Fig. S5), and decreased ROCK1 and p38 MAPK phosphorylation (Figs. 8C, 9C, Supplementary Figs. S5 and S6) in the RCCS-cultured rNCSCs. Nonetheless, PTX pretreatment did not affect the increased CXCR4 expression after RCCS culturing (indicated by arrows in Fig. 6C). The inhibitory effect of PTX on the simulated microgravity-induced cytoskeleton disorganization and apoptosis-activating signaling persisted for at least 48 h during culturing in the RCCS (Supplementary Fig. S4 and data not shown).

Similar to the effects of PTX pretreatment, transfection of CXCR4 siRNA into rNCSCs before culturing in the RCCS significantly increased total numbers and cell viability of rNCSCS; increased expression of neural crest markers FoxD3, p75, Snai1, and Sox10 (Figs. 2D, H, 3, and Supplementary Fig. S6), and expression of cytoskeleton components cortactin and vimentin (Fig. 5D and Supplementary Fig. S5); as well as decreased activation of RhoA GTPase (Figs. 3, 6D, and Supplementary Fig. S5), ROCK1 (Fig. 8D and Supplementary Fig. S5), and p38 MAPK (Figs. 3, 9D, and Supplementary Fig. S6).

On the other hand, CXCR4 siRNA pretransfection decreased expression of G-actin and total β-actin (Fig. 4D and Supplementary Fig. S5), as well as plasma membrane blebbing (Fig. 4D), and dramatically decreased the levels of both CXCR4 protein and mRNA in the RCCS-cultured rNCSCs (Figs. 6D, 7, and Supplementary Fig. S5). The mRNA level in the RCCS-cultured rNCSCs was reduced by CXCR4 siRNA transfection from 3.46 ± 0.87-fold to 0.92 ± 0.38-fold of that in the 2D-cultured rNCSCs (Fig. 7B; n = 24, P < 0.01, one-way ANOVA). The inhibitory effect of CXCR4 siRNA on the simulated microgravity-induced cytoskeleton disorganization and apoptosis-activating signaling persisted for at least 48 h during culturing in the RCCS (Supplementary Fig. S4 and data not shown).

Increased levels of nuclear p53 and apoptosis, but unchanged rates of cell death and proliferation in the RCCS-cultured rNCSCs

Given that the RCCS-cultured rNCSCs showed an increased level of active caspase-3 (Figs. 2B, F and 3) as well as increased phosphorylation and nuclear translocation of p38 MAPK (Figs. 3 and 9B), we were interested in analyzing the expression of p53, which has been reported to be phosphorylated and activated by p38 MAPK [138 –140], in the 2D-cultured and RCCS-cultured rNCSCs. As p53 is well known for its critical role in inducing cell apoptosis [141,142], we performed double immunostaining for p53 and TUNEL or p53 and active caspase-3 to assess the correlation between p53 expression and the death of rNCSCs. Our immunostaining results revealed that the percentage of p53⁺ nuclei in the RCCS-cultured rNCSCs was significantly increased to ∼1.8-fold of that in the 2D-cultured rNCSCs (Fig. 3; compare Figs. 10A, B and 11A, B).

OPEN IN VIEWER

FIG. 10.

Increased percentage of p53⁺ nuclei in spite of unchanged rate of cell death in the RCCS-cultured rNCSCs. Double staining for p53 (red) and TUNEL (green) revealed a significantly increased number and percentage of p53⁺ nuclei in the untreated RCCS-cultured rNCSCs (B) in comparison with those in the 2D-cultured rNCSCs (A), and PTX-pretreated (C) or CXCR4 siRNA-pretransfected (D) RCCS-cultured rNCSCs (n = 12, P < 0.01, one-way ANOVA; see also Fig. 3). On the other hand, the numbers and percentages of TUNEL⁺ nuclei were comparable among the 2D-cultured (A), untreated RCCS-cultured (B), PTX-pretreated RCCS-cultured (C), and CXCR4 siRNA-pretransfected RCCS-cultured (D) rNCSCs (see also Fig. 3). The percentage of p53⁺/TUNEL⁺ double-positive nuclei (indicated by arrows) in the untreated RCCS-cultured rNCSCs (B) was significantly higher than the percentages in the 2D-cultured (A), PTX-pretreated RCCS-cultured (C), and CXCR4 siRNA-transfected RCCS-cultured (D) rNCSCs, indicating that p53 was more positively correlated with the death of untreated rNCSCs after culturing under simulated microgravity. The scale bar in (A) represents 50 μm and applies to panels (A–D).

OPEN IN VIEWER

FIG. 11.

Colocolization of p53 and cleaved (active) caspase-3 and increased rate of apoptosis in the RCCS-cultured rNCSCs. Double immunostaining revealed that p53 was colocalized with cleaved (active) caspase-3 in the nuclei of rNCSCs under both normal gravity and simulated microgravity, indicating that nuclear p53 expression was associated with apoptosis of rNCSCs. The percentages of p53⁺ and cleaved caspase-3⁺ nuclei in the untreated RCCS-cultured rNCSCs (B) were significantly increased to averagely 1.8-fold of those in the 2D-cultured rNCSCs (A), PTX-pretreated RCCS-cultured (C), and CXCR4 siRNA-transfected RCCS-cultured (D) rNCSCs (see also Fig. 3). The scale bar in (A) represents 50 μm and applies to panels (A–D).

In addition, western blot analyses indicated that the nuclear protein level of p53 in the RCCS-cultured rNCSCs was significantly increased to averagely 4.2-fold of that in the 2D-cultured rNCSCs (Supplementary Fig. S6). Both PTX treatment and CXCR4 siRNA transfection before RCCS culturing led to a decrease of the percentage of p53⁺ nuclei and nuclear expression of p53 in rNCSCs to comparable levels as in the 2D-cultured rNCSCs (Figs. 3, 10C, D, 11C, D, and Supplementary Fig. S6). The significantly increased percentage of active caspase-3⁺ nuclei in the RCCS-cultured rNCSCs was also decreased by PTX pretreatment (Figs. 3 and 11C; n = 12, P < 0.01, one-way ANOVA) and CXCR4 siRNA pretransfection (Figs. 3 and 11D; n = 12, P < 0.01, one-way ANOVA).

Western blot analyses also revealed that the nuclear level of cleaved caspase-3 in the untreated RCCS-cultured rNCSCs was averagely 2.6-fold of the levels in the 2D-cultured, PTX-pretreated RCCS-cultured, and CXCR4 siRNA pretransfected RCCS-cultured rNCSCs. On the other hand, the percentages of TUNEL⁺ nuclei in the 2D-cultured and PTX-pretreated or CXCR4 siRNA-pretransfected RCCS-cultured rNCSCs (Figs. 3, 10C, D) were comparable with those in the untreated RCCS-cultured rNCSCs (Figs. 3 and 10B).

To evaluate the correlation between TUNEL⁺ and active caspase-3⁺ signals in rNCSCs cultured under normal or simulated microgravity, we performed double immunostaining for TUNEL and active caspase-3 in different groups of rNCSCs (Supplementary Fig. S7A–D). We found that the percentages of TUNEL⁺/active caspase-3⁺ double-positive nuclei (indicated by arrows in Supplementary Fig. S7A−D) in total active caspase-3⁺ nuclei were comparable among the 2D-cultured, untreated RCCS-cultured, PTX-pretreated RCCS-cultured, and CXCR4 siRNA-pretransfected RCCS-cultured rNCSCs (Supplementary Fig. S7E), indicating that the proportions of cells at the late phase of apoptosis in total apoptotic cells were comparable among these four groups of rNCSCs.

On the other hand, in the untreated RCCS-cultured rNCSCs, the percentage of TUNEL⁺/active caspase-3⁺ double-positive nuclei in total TUNEL⁺ nuclei was significantly increased to averagely 1.5-fold of the percentages in the 2D-cultured, PTX-pretreated RCCS-cultured, and CXCR4 siRNA-pretransfected RCCS-cultured rNCSCs (Supplementary Fig. S7E), indicating that the proportion of apoptotic cells in total dying cells in the untreated RCCS-cultured group was significantly higher than the proportions in the other three experimental groups (Supplementary Fig. S7E; n = 12, P < 0.01, one-way ANOVA).

We next analyzed cell proliferation and expression of the cyclin-dependent kinase inhibitor p21^Waf1/Cip1, a major downstream effector of p53 [143 –145], in 2D-cultured and RCCS-cultured rNCSCs. Interestingly, both the proliferation rates, which were calculated as the percentages of EdU⁺ cells, and the percentages of p21⁺ nuclei were comparable among the 2D-cultured, RCCS-cultured, PTX-pretreated, and CXCR4 siRNA-pretransfected rNCSCs (Fig. 3 and Supplementary Fig. S8). It is noteworthy that p21⁺ nuclei were stained with a low level of EdU (indicated by arrows in Supplementary Fig. S8) or no EdU at all, consistent with the role of p21 in suppressing cell proliferation [146 –148]. Our western blotting results also indicated that the nuclear protein levels of p21 were comparable among the 2D-cultured, untreated RCCS-cultured, PTX-pretreated RCCS-cultured, and CXCR4 siRNA pretransfected RCCS-cultured rNCSCs (Supplementary Fig. S6).