Microgravity Induces Transient EMT in Human Keratinocytes by Early Down-Regulation of E-Cadherin and Cell-Adhesion Remodeling

Featured Application: Studies on the effects of microgravity on cell–cell and cell–matrix interac-tion may help increase the safety and improve the health of astronauts in space. Abstract: Changes in cell–matrix and cell-to-cell adhesion patterns are dramatically fostered by the microgravity exposure of living cells. The modiﬁcation of adhesion properties could promote the emergence of a migrating and invasive phenotype. We previously demonstrated that short exposure to the simulated microgravity of human keratinocytes (HaCaT) promotes an early epithelial– mesenchymal transition (EMT). Herein, we developed this investigation to verify if the cells maintain the acquired invasive phenotype after an extended period of weightlessness exposure. We also evaluated cells’ capability in recovering epithelial characteristics when seeded again into a normal gravitational ﬁeld after short microgravity exposure. We evaluated the ultra-structural junctional features of HaCaT cells by Transmission Electron Microscopy and the distribution pattern of vinculin and E-cadherin by confocal microscopy, observing a rearrangement in cell–cell and cell–matrix interactions. These results are mirrored by data provided by migration and invasion biological assay. Overall, our studies demonstrate that after extended periods of microgravity, HaCaT cells recover an epithelial phenotype by re-establishing E-cadherin-based junctions and cytoskeleton remodeling, both being instrumental in promoting a mesenchymal–epithelial transition (MET). Those ﬁndings suggest that cytoskeletal changes noticed during the ﬁrst weightlessness period have a transitory character, given that they are later reversed and followed by adaptive modiﬁcations through which cells miss the acquired mesenchymal phenotype. for nuclei and for F-Actin visualization were utilized. Then, the slides were washed with PBS and mounted with 0.1 mM Tris-HCl at pH 9.5: glycerol Negative controls were processed in the same conditions besides primary antibody staining. Finally, immunolocalization was analyzed with a Leica confocal microscope TCS SP2 (Leica Microsystems Heidelberg GmbH, equipped with Ar/ArKr and He/Ne lasers. The laser line was at 488, 543, and 633 nm for FITC, rhodamine, and TO-PRO-3 excitation, respectively. The images were scanned under a 20 × or 40 × oil immersion objective. E-cadherin quantiﬁcation was performed using an optical spatial series with a step size of 1 µ m. Fluorescence intensity (SUM (I), a.u arbitrary units) was evaluated in sized Regions Of Interest (ROI) that included group of cells. For each ROI, the analyzed cell number was determined counting cell nuclei and the SUM (I) of the ROI was normalized versus the cell number. We counted about 500 cells for each sample. The laser intensity and confocal parameters were maintained at the same values to perform 1g and RPM series. The background threshold of ﬂuorescence was determined by the use of negative control in which the primary antibody was omitted. For depth analyses, an optical spatial series with a step size of 5 µ m were recovered. Colocalization analysis was effectuated at the level of periphery of the cell and in an internal compartment at level of cell–cell contact. The ﬂuorescence intensity of vinculin and actin (SUM (I)) at these regions was also quantiﬁed. All analyses were performed using Leica Confocal software. Data were analyzed with Sigma Plot Software and Pearson coefﬁcient was used for correlation analysis. disappearance of main EMT features. A similar pattern can be observed as the same cells are seeded again into a normal gravitational ﬁeld. Deﬁnitely, the emergence of a migratory phenotype should be considered in this cellular model as a transient, adaptive phenomenon enacted by the modiﬁed environmental physical forces. Further studies to elucidate the molecular mechanisms behind the morphological and functional changes we observed are currently undergoing in our laboratory.


Introduction
Microgravity has noticeable effects at the cellular level, producing morphological and functional alterations in a large number of different cell types [1][2][3][4]. It is believed that cellular modifications induced by weightlessness play a fundamental role in the development of parameter that should be carefully considered before conclusive statements could be drawn when performing investigations on living cells cultured in simulated microgravity. Indeed, weightlessness-exposed cells retain the potentiality to reverse to a native phenotype while the epithelial-mesenchymal transition should be viewed as a transitory, adaptive strategy.

Cell Culture
The human keratinocytes cell line HaCaT (kindly provided by Prof. MR Torrisi, Sapienza University, Rome, Italy) was cultured in Dulbecco modified Eagle medium (DMEM) supplemented with 10% Fetal Bovine Serum (FBS) and antibiotics (Penicillin 100 IU/mL, Streptomycin 100 µg/mL, all from Euroclone Ltd., Cramlington, UK). The cells were cultured at 37 • C in an atmosphere of 5% CO 2 in air.

Simulated Microgravity and Cell Exposure
Microgravity conditions were simulated by a Desktop Random Positioning Machine (RPM), which is a particular kind of 3D clinostat [22] manufactured by Dutch Space (Leiden, The Netherlands). The desktop RPM was positioned in a standard incubator. The angular velocity of rotation was set at 90 • /s as maximum and 30 • /s as minimum in a random mode of the machine. HaCaT cells were seeded into cell culture chambers as described below. Chambers containing HaCaT sub-confluent monolayers were fixed onto the RPM as close as possible to the center of the platform. RPM and 1g cultures were kept in the same humidified incubator at 37 • C in an atmosphere of 5% CO 2 in air.
Cells were placed in RPM for 24 h to investigate the early changes in the morphological and functional phenotype. To assess adaptive process to microgravity, HaCaT cells were placed in RPM for 48 and 72 h. For recovery experiments, HaCaT cells were placed in RPM for 24 h and then transferred into a normal gravitational field for further 24 h.

Preparation of Cellular Extracts and Western Blot Analysis
HaCaT cells were seeded into cell chambers in a standard medium. Then, cells were serum-deprived for 4 h and, after this time, they were exposed or not to simulated microgravity for adaptation or recovery experiments. At the end of the experimental time, the cells were washed twice with ice-cold PBS (Phosphate-Buffered Saline, Sigma-Aldrich, Saint Louis, MO, USA) and cell pellets were lysed with RIPA buffer (Sigma-Aldrich). A mix of protease inhibitors (Complete-Mini Protease Inhibitor Cocktail Tablets; Roche, Mannheim, Germany) was added just before use. Then, cellular extracts were centrifuged at 8000× g for 10 min. The protein content of supernatant fractions was determined by using the Bradford assay. For immunoblot analysis, cellular extracts (40 µg of protein for each sample) were separated on SDS-PAGE gels with a concentration of acrylamide specific for the proteins studied. Proteins were blotted onto nitrocellulose membranes (Bio-Rad Laboratories, Hercules, CA, USA) and probed with mouse anti-E-cadherin C-terminal end monoclonal antibody (BD Transduction Laboratories™, San Diego, CA, USA, Cat# 610181); Then, the membranes were incubated with the appropriate HRP-conjugated secondary antibody (sheep anti-mouse IgG ECL antibody, GE Healthcare, Little Chalfont, UK, Cat# NA9310) for 1 h at room temperature. Immunocomplexes were detected with an enhanced chemiluminescence kit (Western Bright ECL HRP Substrate, Advansta Inc., Menlo Park, CA, USA) according to the manufacturer's instructions. All Western blot images were acquired and analyzed through Imaging Fluor S densitometer (Biorad-Hercules). Each experiment was performed three times.

Migration Assay
Cells were incubated at 37 • C with 5% CO 2 under normal gravity conditions (1g) and in microgravity conditions for 24, 48, and 72 h. Migration assay was performed using Cell Culture Insert 12 well 8.0 µm pore size (Falcon, cat. 353182, Lincon Park, NJ, USA) placed in a 12 multiwell (Tanswell Falcon, cat. 351143). After the incubation time, the cells were counted and resuspended in DMEM 0.1% FBS. Then, 5 × 10 5 cells/well were added in the upper chamber of the transwell in DMEM 0.1% FBS, whereas the lower chambers were filled with 800 µL DMEM 10% FBS and incubated for a further 24 h for migration assay.
For recovery experiments, HaCaT cells were placed in RPM for 24 h and then transferred into a normal gravitational field for further 24 h. Then, they were counted and transferred into the upper chamber of the transwell at normal gravitational field for further 24 h for migration assay.
After 24 h of incubation, medium and unmigrated cells in the upper surface of the insert were mechanically removed, and the insert (containing migrated cells in the lower surface) was fixed with paraformaldehyde 4% in PBS (pH 7.4) at 4 • C and staining with Diff Quick solution (DADE, cat. 130832 Network, NJ, USA). Migrated cells were counted under 40× objective using an optical microscope (Axioplan Zeiss Oberköchen, Germany), and the average number ± SD of cells are reported as fold change with respect to control, which is considered as 1. The whole area of each filter was counted. Three independent experiments were performed; each experiment was performed at least in quadruplicate.

Matrigel Invasion Assay
In vitro invasion assay was performed using chambers coated with GFR Matrigel (Basement Membrane Matrix Growth Factor Reduced, BD Biosciences, cat.354483). The cells were incubated for 24, 48, and 72 h at 37 • C with 5% CO 2 under normal gravity conditions (1g) and in simulated microgravity condition; then, cells were trypsinized, counted, and resuspended in DMEM 1% FBS. Then, 5.0 × 10 3 cells/well were seeded on the top of the GFR Matrigel in 200 µL of medium 1% FBS, whereas the lower chambers were filled with 800 µL DMEM 10%% FBS. These chambers were incubated for a further 24 h for invasion assay.
For recovery experiments, HaCaT cells were placed in RPM for 24 h and then transferred into a normal gravitational field for further 24 h. Then, they were counted and transferred into the top of GFR Matrigel at normal gravitational field for a further 24 h for invasion assay. At the end of incubation, GFR Matrigel and non-invading cells were mechanically removed with a cotton swab. A polycarbonate filter containing invading cells was fixed with paraformaldehyde 4% in PBS (pH 7.4) at 4 • C. and stained with Diff Quick solution. The filter was analyzed by optical microscope, and four fields/filter were recovered at 10× magnification. Invading cells were counted, and the average number ± SD of cells are reported as fold change with respect to control considered as 1. Three independent experiments were performed; each experiment was performed at least in quadruplicate.
Confocal analysis of GFR Matrigel invasion was done on HaCaT cells dispersed into a single cell suspension; they were labeled using the PKH-26 Red Fluorescent Cell Linker Kit (Sigma-Aldrich, St. Louis, MO, USA), according to manufacturer's instructions.

Immunofluorescence and F-Actin Localization
Fluorescence analysis was performed in cells grown until semi-confluence. HaCaT cells were cultured into 8-well µ-slides (ibidi GmbH, AmKlopferspitz19, D-82152 Martinsried, Germany) and exposed or not to simulated microgravity for adaptation or recovery experiments as described in "Simulated microgravity and cell exposure section". At the end of incubation, the cells were fixed with 4% paraformaldehyde for 10 min at 4 • C and washed twice for 10 min with PBS. The cells were permeabilized for 30 min using PBS, 3% BSA (Bovine Serum Albumin, Santa Cruz Biotechnology), 0.1% Triton X-100 (Sigma-Aldrich), followed by mouse anti-E-cadherin C-terminal end monoclonal antibody (BD Transduction Laboratories™, Cat# 610181), and mouse anti vinculin (Santa Cruz, Cat sc-73614 1:50 dilution).
The cells were washed with PBS and incubated for 1 h at room temperature with the secondary antibody (FITC-AffiniPure donkey anti mouse IgG Cat# 715-095-150, RRID: AB_2340792 purchased from Jackson Immuno Research Labs, 1: 200 dilution). TO-PRO3 iodide fluorescent dye 642/661 (1:5000 in PBS, Invitrogen, cat. T3605, Carlsbad, CA, USA) for nuclei staining and rhodamine phalloidin (Invitrogen Molecular Probes Eugene 1:40 dilution) for F-Actin visualization were utilized. Then, the slides were washed with PBS and mounted with 0.1 mM Tris-HCl at pH 9.5: glycerol (2:3). Negative controls were processed in the same conditions besides primary antibody staining. Finally, immunolocalization was analyzed with a Leica confocal microscope TCS SP2 (Leica Microsystems Heidelberg GmbH, Mannheim, Germany) equipped with Ar/ArKr and He/Ne lasers. The laser line was at 488, 543, and 633 nm for FITC, rhodamine, and TO-PRO-3 excitation, respectively. The images were scanned under a 20× or 40× oil immersion objective. E-cadherin quantification was performed using an optical spatial series with a step size of 1 µm. Fluorescence intensity (SUM (I), a.u arbitrary units) was evaluated in sized Regions Of Interest (ROI) that included group of cells. For each ROI, the analyzed cell number was determined counting cell nuclei and the SUM (I) of the ROI was normalized versus the cell number. We counted about 500 cells for each sample. The laser intensity and confocal parameters were maintained at the same values to perform 1g and RPM series. The background threshold of fluorescence was determined by the use of negative control in which the primary antibody was omitted. For depth analyses, an optical spatial series with a step size of 5 µm were recovered. Colocalization analysis was effectuated at the level of periphery of the cell and in an internal compartment at level of cell-cell contact. The fluorescence intensity of vinculin and actin (SUM (I)) at these regions was also quantified. All analyses were performed using Leica Confocal software. Data were analyzed with Sigma Plot Software and Pearson coefficient was used for correlation analysis.

Transmission Electron Microscopy (TEM) Analysis
Cells cultured as described in the "Simulated microgravity and cell exposure" section were fixed in 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4), postfixed in 1% OsO4 in 0.1 M cacodylate buffer, stained with UAR-EMS stain (Electron Microscopy Sciences), de-hydrated in ethanol, and embedded in epoxy resin. Semithin sections were stained with toluidine blue dye. Ultrathin sections (60 nm) were treated with tannic acid and then contrasted with lead hydroxide. Samples were studied and photographed using a Libra 120 Transmission Electron Microscope (Zeiss) equipped with Wide-angle Dual Speed CCD-Camera sharp eye 2 K (4 Mpx) operated by iTEM software (Soft Image System, Münster, Germany).

Statistical Analysis
Data were expressed as mean ± standard deviation (SD). Data were analyzed with unpaired two-tailed Student's t-test or analysis of variance (ANOVA) followed by the Bonferroni post-test. Differences were considered significant at the level of p < 0.05. Statistical analysis was performed by using GraphPad Instat software (GraphPad Software, Inc.; San Diego, CA, USA) 3. Results 3.1. E-Cadherin in HaCaT Cells Exposed to Simulated Microgravity 3.1.1. E-Cadherin Expression in HaCaT Exposed to Weightlessness and after Reseeding into a 1g Field E-cadherin expression was evaluated by Western blot analysis, as shown in Figure 1. In HaCaT cells exposed to simulated microgravity for 24 h, we observed a significant decrease in E-cadherin expression as compared to 1g controls. E-cadherin recovers normal values when the cells were placed in RPM for 48 and 72 h or when HaCaT cells were placed in RPM for 24 h and then transferred into a normal gravitational field for further 24 h (recovery). E-cadherin expression was evaluated by Western blot analysis, as shown in Figur In HaCaT cells exposed to simulated microgravity for 24 h, we observed a significant crease in E-cadherin expression as compared to 1g controls. E-cadherin recovers norm values when the cells were placed in RPM for 48 and 72 h or when HaCaT cells w placed in RPM for 24 h and then transferred into a normal gravitational field for furt 24 h (recovery).

E-Cadherin Distribution Pattern in HaCaT Exposed to Weightlessness and after Reseeding into a 1g Field
These results are mirrored by data provided by confocal and electron microsco studies. Indeed, as shown in Figure 2, the fluorescence intensity of E-Cadherin after 2 in RPM-exposed cells decreases significantly, especially nearby areas of cell-to-cell con ( Figure 2B). To confirm these observations, we performed quantitative analyses of fl rescence intensity, and we found a reduction of about 30% in RPM-treated cells with spect to the 1g condition (128.354 a.u ± 12.654 vs. 178.878 a.u ± 10.522 respectively; p < 0 Table 1). Moreover, TEM analysis of HaCaT cells evidences well-organized adherens ju tions in cell growing at 1g ( Figure 2C), whereas a disassembled junctional complex-ch acterized by a reduced electron density-is noticed in RPM-treated cells ( Figure 2D).
For estimating the adaptive process to microgravity, HaCaT cells were placed in R for 48 and 72 h and then fixed and prepared for immunofluorescence experiments lowed by confocal analysis.
After 48 h of simulated microgravity, intense E-cadherin fluorescence is de novo preciable within the cell-cell contact area, with only minimal differences when compa to 1g control ( Figure 3A,B). At 72 h, both 1g and RPM-exposed cells show a sim

E-Cadherin Distribution Pattern in HaCaT Exposed to Weightlessness and after Reseeding into a 1g Field
These results are mirrored by data provided by confocal and electron microscopy studies. Indeed, as shown in Figure 2, the fluorescence intensity of E-Cadherin after 24 h in RPM-exposed cells decreases significantly, especially nearby areas of cell-to-cell contact ( Figure 2B). To confirm these observations, we performed quantitative analyses of fluorescence intensity, and we found a reduction of about 30% in RPM-treated cells with respect to the 1g condition (128.354 a.u ± 12.654 vs. 178.878 a.u ± 10.522 respectively; p < 0.01; Table 1). Moreover, TEM analysis of HaCaT cells evidences well-organized adherens junctions in cell growing at 1g ( Figure 2C), whereas a disassembled junctional complex-characterized by a reduced electron density-is noticed in RPM-treated cells ( Figure 2D). ness. Similarly, cells maintained in RPM for 24 h and then transferred into a normal itational field for a further 24 h recover an E-cadherin distribution pattern that ap undistinguishable from that observed at 1g condition ( Figure 3E,F). Moreover, TEM ysis, performed in the same experimental setting, confirms that after 48 h of microg exposure, adherent junctions are again observable among cells and present ultrastru features overlapping with those noticed in cells never exposed to simulated microg and growing in a 1g field ( Figure 3G,H).   For estimating the adaptive process to microgravity, HaCaT cells were placed in RPM for 48 and 72 h and then fixed and prepared for immunofluorescence experiments followed by confocal analysis.
After 48 h of simulated microgravity, intense E-cadherin fluorescence is de novo appreciable within the cell-cell contact area, with only minimal differences when compared to 1g control ( Figure 3A,B). At 72 h, both 1g and RPM-exposed cells show a similar distribution pattern of E-cadherin ( Figure 3C,D). Overall, these findings suggest that E-cadherin recovers a normal distribution pattern after prolonged exposition to weightlessness. Similarly, cells maintained in RPM for 24 h and then transferred into a normal gravitational field for a further 24 h recover an E-cadherin distribution pattern that appears undistinguishable from that observed at 1g condition ( Figure 3E,F). Moreover, TEM analysis, performed in the same experimental setting, confirms that after 48 h of microgravity exposure, adherent junctions are again observable among cells and present ultrastructural features overlapping with those noticed in cells never exposed to simulated microgravity and growing in a 1g field ( Figure 3G,H). Appl. Sci. 2021, 11, x FOR PEER REVIEW 8 of 20

Cytoskeleton Modifications of HaCaT Exposed to Simulated Microgravity
To assess the cytoskeleton reorganization occurring in keratinocytes exposed to simulated microgravity for 24, 48, and 72 h, we investigated changes involving the distribution pattern of F-actin and Vinculin, which is a molecule that is highly expressed in focal adhesion, and it is also recruited at cadherin containing adherents junctions [18].

F-Actin and Vinculin Organization after 24 h of Simulated Microgravity
As we described in a previous paper [18], cells cultured in 1g for 24 h showed F-actin distributed in peripheral cortical bundles, while in simulated microgravity exposed cells, a relevant increase of lamellipodia, filopodia, and ruffles were frequently observed. Overall, those findings suggest that cells exposed to weightlessness acquire a motile/invasive phenotype.
To obtain further morphological insights, we performed confocal analysis of the cell to recognize how vinculin colocalizes and correlates with actin filament in shaping FAs, which plays a pivotal role in modulating cell adhesion to the substrate and, henceforth, cell migration. The colocalization analysis and fluorescence intensity of vinculin and actin (SUM (I)) at the free front of the cell membrane, and in an internal compartment at the level of cell-cell contact was performed as described in the Materials and Methods section. As shown in Figure 4A, in control condition, vinculin is predominantly distributed nearby the cell-cell contacts. Colocalization and correlation analyses showed a few or no colocalization of vinculin with actin at the leading edge of the cell culture ( Figure 4C) and no significant positive relationship. (r = 0.174 p > 0.05). In addition, in cell contact, there is no significant relationship between F-actin and vinculin (r = 0.29 p > 0.05), although these two proteins are tightly associated (at a few points, the cellular contacts colocalize), as indicated in the green and red picks in the figure ( Figure 4E). After 24 h of exposure to weightlessness, vinculin was mostly displaced at the cell periphery, and it showed an increased density in the newly formed ruffles and filopodia ( Figure 4B). Colocalization and correlation analyses demonstrate that vinculin colocalizes with actin filament at the level of membrane protrusions ( Figure 4D), and the correlation coefficient revealed a positive relationship between these two markers (r = 0.97; p < 0.001), indicating the presence of FA. By contrast, both colocalization ( Figure 4F) and the association between F-actin and vinculin decrease at the site of cell adhesion contacts, and there is no significant positive relationship (r = 0.29 p > 0.05) between the two markers. These results confirm the formation of new FAs necessary to the acquisition of the motile/invasive phenotype by the cells exposed to simulated microgravity.

F-Actin and Vinculin Organization at 48-72 h of Simulated Microgravity
To verify if the acquired phenotype at 24 h is stably maintained by cells exposed to microgravity, we extended the exposure times to 48 and 72 h. Surprisingly, after 48 h of simulated microgravity, lamellipodia and ruffles markedly decreased when compared to 24 h RPM cultured cells. After 72 h of microgravity exposition, the cytoskeleton architecture observed in the microgravity condition recovered almost completely the cytoskeleton organization usually observed in 1g control.
Specifically, by confocal analysis, we observed that after 48 h of microgravity, the exposition of vinculin at the cell periphery was appreciably diminished when compared to cells exposed for 24 h in weightlessness. Colocalization analyses demonstrated little or no colocalization (Figure 5C), although a good correlation with actin (r = 0.92; p > 0.05) can be recorded in few points. At cell-to-cell contact, colocalization with F-actin was observed ( Figure 5E) and, as described in 1g treated samples, there is no significant positive relationship between the two markers (r = 0.21 p > 0.05).    After 72 h of microgravity exposition, the distribution of vinculin is almost concentrated within the areas of cell-to-cell contact, in which it shows a striking colocalization with F-actin bundles ( Figure 5D,F), while little or no colocalization was observed in cell membranes at the leading edge of culture. Correlation analyses of vinculin with actin at the leading edge of the cell culture and in cell-cell contact showed no significant positive relationships (r = 0.173 p > 0.05; r = 0.57 p > 0.05 respectively), as we observed in the control condition.
Overall, these findings witness that after longer periods of microgravity exposure, cells recover their native phenotype, losing their migrating/invasive properties.

Reversibility of the Cytoskeleton Modifications
After 24 h of microgravity exposure, cells were replaced into a normal gravitational field, and cytoskeleton architecture was investigated after a further 24 h of 1g conditioning. Under these conditions, both F-actin filament organization and vinculin distribution recovered almost completely their normal distribution pattern. As shown in Figure 6, most of the cells reacquire a cortical F-actin organization as well as cell-to-cell contact of vinculin distribution, overlapping the picture observed in cells growing in 1g ( Figure 6). Noticeably, F-actin and vinculin both colocalize mostly in proximity of cell-to-cell contact areas ( Figure 6C). Correlation analyses of vinculin with F-actin at the leading edge of cell cultures and in cell-cell contact regions showed no significant positive relationship (r = 0.218, p > 0.05; r = 0.465, p > 0.05 respectively). These data demonstrate that the migratory phenotype acquired in the first 24 h in simulated microgravity conditions is almost completely reversed when cells are replaced into a normal gravity field.

Migration and Invasive Capability of HaCaT Exposed to Simulated Microgravity
To get insights into the migrating/invasive properties acquired by HaCaT cells exposed to microgravity for short periods, we planned specific experiments as previously described.

Cell Migration
Exposure to microgravity in RPM for 24 h strongly increases cell migration with respect to 1g control (3.44 ± 0.38 vs. 1.00 ± 0.17; p < 0.01), as observed in a preliminary study [11]. However, after 48 h of simulated microgravity, cell migration resulted significantly decreased when compared to 24-h RPM treated cells (1.95 ± 0.31 vs. 3.44 ± 0.3; p < 0.05). Yet, after 72 h of microgravity exposition, cell migration was further decreased, approaching values observed in 1g cultured cells (0.94 ± 0.14 vs. 1.00 ± 0.17; p = n.s.) ( Figure 7A). Similar findings were obtained when HaCaT were replaced in 1g field for 24 h, after 24 h of weightlessness exposure. Indeed, the recovery experiment shows that cell migration was significantly reduced when compared to 24 h RPM-treated cells (1.28 ± 0.26 vs. 3.44 ± 0.38, § § p < 0.01), whereas there was no significant difference with respect to 1g controls at 48 h (1.28 ± 0.26 vs. 1.00 ± 0.18 p = n.s.). In addition, no significant difference was observed with respect to cells maintained for 48 h in RPM condition (Figure 7 panel A  and table). vinculin distribution, overlapping the picture observed in cells growing in 1g ( Figure 6). Noticeably, F-actin and vinculin both colocalize mostly in proximity of cell-to-cell contact areas ( Figure 6C). Correlation analyses of vinculin with F-actin at the leading edge of cell cultures and in cell-cell contact regions showed no significant positive relationship (r = 0.218, p > 0.05; r = 0.465, p > 0.05 respectively). These data demonstrate that the migratory phenotype acquired in the first 24 h in simulated microgravity conditions is almost completely reversed when cells are replaced into a normal gravity field.   Data expressed as fold increase of control value considered as 1 are means of three independent experiments, each performed in quadruplicate, with S.D. represented by vertical bars (* p < 0.05; ** p < 0.01 versus 1g control; $ p < 0.05; $$ p < 0.01 versus the cells exposed to RPM for 24 h by ANOVA followed by Bonferroni post-test). Tables illustrate the number of migrating and invading cells/field, respectively.

Cell Invasiveness
In vitro invasion assay was performed using chambers coated with GFR Matrigel. As shown in Figure 7B, exposure to microgravity for 24 h strongly increases cell invasion with respect to 1g conditions (4.26 ± 0.66 vs. 1.00 ± 0.14; p < 0.01). To further confirm such behavior, we estimated by confocal microscopy the depth of Matrigel invasion reached by HaCaT cells exposed to microgravity for 24 h. As reported in Figure 8, HaCaT cells maintained at normal gravity form ovoidal structures and invade the matrix up to a depth of 49.71 ± 3.91 µm ( Figure 8A-C), while cells growing in microgravity appear scattered and able to invade deeper, reaching a depth of 119.99 ± 5.67 µm (p < 0.001; Figure 8D-F). Overall, such findings confirm that cells exposed to microgravity for short periods dramatically increase their invasiveness capability. Appl. Sci. 2021, 11, x FOR PEER REVIEW 16 of 20

Discussion
HaCaT cells exposed to simulated microgravity show a biphasic response, displaying increased motility and invasiveness during the early period (24 h), which was followed by the recovering of a noninvasive morphological phenotype that is partly observed after 48 h and fully acquired at 72 h. Similar results were obtained by seeding again into a normal gravitational field those invasive HaCaT cells previously exposed to simulated microgravity for 24 h. Overall, these findings suggest that changes occurring at the morphological and functional level in cells exposed to microgravity for short periods have a transitory character and then are followed by adaptive modifications through which cells strive in recovering their native phenotype. Both E-cadherin and vinculin distribution patterns are likely to play a key role during those changes.
The actin ability to generate/transduce forces through cytoskeleton remodeling is a pivotal tool that enables cells in mechanosensing the environmental imbalance of physical forces. The transduction of mechanical stresses involves the clustering of transmembrane receptors that are instrumental in the successive assembly of focal contacts and adhesions that physically connect the extracellular matrix to the cell cytoskeleton [23]. Binding enhances the bundling of F-actin at the cell-substratum adhesion sites with the concomitant maturation of both the focal adhesions and the actin stress fibers [24], which represent mandatory steps to trigger the required tension for ensuring cell adherence and motility capability. It has been reported that exposure to microgravity reduces the number and total area of focal adhesions per cell [25], thus significantly reducing adherence to the substratum and migratory capability [26], albeit some contradictory results have been reported [27]. These changes are mandatory for the emergence of floating populations in cell cultures exposed to weightlessness, i.e., an organoid-like structure in which cells aggregate themselves after detaching from the substrate [28]. Therefore, our results are apparently at odds with those reported for other kind of cells. Differences in the experimental model, namely those involving different times of observation, can probably help in explaining this conundrum. Indeed, we observed increased motility only within the limited window of the first 24 h of microgravity exposure, given that after that time, As noticed for cell migration, invasiveness decreases after 48 h of RPM with respect to 24-h RPM treated cells (2.91 ± 0.42 vs. 4.26 ± 0.66; p < 0.05). Moreover, after 72 h of microgravity exposition, cell invasion was significantly suppressed when compared to the 24 h RPM treated cells, (1.38 ± 0.08 vs. 4.26 ± 0.66; p < 0.01); HaCaT cells almost completely lose their invasive property, reaching values superimposable to those measured in 1g control cells (1.38 ± 0.08 vs. 1.00 ± 0.14 p = n.s.). Again, similar results were obtained when cells exposed to microgravity for 24 h were reseeded into a 1g field for 24 h: cells displayed significantly reduced invasiveness values in respect to the microgravity condition at 24 h (1.88 ± 0.31 vs. 4.26 ± 0.66 p < 0.05), the migrating phenotype "disappears", and cells recover values similar to those displayed by 1g controls at 48 h (1.88 ± 0.31 vs. 1.00 ± 0.19; p = n.s.). In addition, no significant difference was observed with respect to cells maintained for 48 h in RPM condition (Figure 7 panel B and table).

Discussion
HaCaT cells exposed to simulated microgravity show a biphasic response, displaying increased motility and invasiveness during the early period (24 h), which was followed by the recovering of a noninvasive morphological phenotype that is partly observed after 48 h and fully acquired at 72 h. Similar results were obtained by seeding again into a normal gravitational field those invasive HaCaT cells previously exposed to simulated microgravity for 24 h. Overall, these findings suggest that changes occurring at the morphological and functional level in cells exposed to microgravity for short periods have a transitory character and then are followed by adaptive modifications through which cells strive in recovering their native phenotype. Both E-cadherin and vinculin distribution patterns are likely to play a key role during those changes.
The actin ability to generate/transduce forces through cytoskeleton remodeling is a pivotal tool that enables cells in mechanosensing the environmental imbalance of physical forces. The transduction of mechanical stresses involves the clustering of transmembrane receptors that are instrumental in the successive assembly of focal contacts and adhesions that physically connect the extracellular matrix to the cell cytoskeleton [23]. Binding en-hances the bundling of F-actin at the cell-substratum adhesion sites with the concomitant maturation of both the focal adhesions and the actin stress fibers [24], which represent mandatory steps to trigger the required tension for ensuring cell adherence and motility capability. It has been reported that exposure to microgravity reduces the number and total area of focal adhesions per cell [25], thus significantly reducing adherence to the substratum and migratory capability [26], albeit some contradictory results have been reported [27]. These changes are mandatory for the emergence of floating populations in cell cultures exposed to weightlessness, i.e., an organoid-like structure in which cells aggregate themselves after detaching from the substrate [28]. Therefore, our results are apparently at odds with those reported for other kind of cells. Differences in the experimental model, namely those involving different times of observation, can probably help in explaining this conundrum. Indeed, we observed increased motility only within the limited window of the first 24 h of microgravity exposure, given that after that time, HaCaT cells progressively recovered the native phenotype, losing the invasive/motile phenotype. Moreover, effects of weightlessness on F-actin cytoskeleton are still a matter of debate. While a number of studies have reported that microgravity decreases F-actin expression, leading to disorganization of the cytoskeleton [29], other reports have reported opposite findings, with increased F-actin and stress fiber formation with the emergence of lamellipodia protrusions [30]. It is well known that vinculin has several effects on actin polymerization in newly formed adhesions. Vinculin can cross-link and bundle actin filaments, and it can recruit actin modifiers such as the Arp2/3 complex, which is a potent nucleator of actin polymerization. By contrast, vinculin and F-actin polymerization at cell-cell adhesions is less understood [18]. In our study, through analyzing the fluorescence intensity (SUM (I)) of actin and vinculin in cells exposed to simulated microgravity, we obtained a positive correlation between the two markers only at the level of newly formed FAs in cells cultured in the microgravity condition for 24 h. This correlation persists but decreases in membrane protrusions observed at 48 h of simulated microgravity, while at 72 h, it disappears and reaches values overlapping those observed in cells cultured in 1g. On the basis of these results, we can hypothesize that in the FAs that we observed and analyzed, vinculin is able to recruit actin filament within the FAs, while this phenomenon does not occur in cell-cell contact or in membrane protrusions of cells deprived of their migratory capability. Furthermore, in HaCaT cells cultured in RPM for 24 h and then transferred into a normal gravitational field for a further 24 h (recovery conditions), both F-actin filament and vinculin recovered almost completely the distribution and correlation pattern that we observed at 1g. All these data demonstrated that the emergence of FAs and consequently the F-actin remodeling occurring during the first 24 h should be considered a transient and adaptive phenomenon. Noticeably, these data outline that a critical factor in the emergence of the migrating/invasive phenotype is represented by the concomitant down-regulation of E-cadherin in cells exposed to microgravity for limited periods. Indeed, reduction in the expression pattern of E-cadherin during weightlessness indicated a transitory commitment toward an invasive phenotype, which is a critical feature that is also recorded in thyroid cancer cells exposed to short-term microgravity [30]. It is worth noting that the concomitant reduction in vinculin expression, which is witnessed by confocal studies during the first 24 h of microgravity exposure, can ultimately contribute in decreasing E-cadherin values, given that vinculin exerts a positive regulatory influence upon E-cadherin levels [31]. We hypothesize that the loosening of cell-to-cell adhesion, which is ascribed to reduced E-cadherin levels, plays a key role in favoring the emergence of a migratory/invasive phenotype despite the fact that the concomitant changes occurring in the cytoskeleton-i.e., the disorganization of F-Actin network and weakening of FA-would impair the migrating capability of HaCaT cells, at least in principle. This conclusion deserves to be emphasized, as it pinpoints E-cadherin as the pivotal factor that can switch cells toward a motile, invasive phenotype. Given that the metastatic potential of tumors is highly dependent on such capabilities, it is worth noting that the critical target of the treatment should be the restoration of proper E-cadherin values, as already underlined [32].
Loss of cell-to-cell adhesions is instrumental in enabling the disassembly of the cell cluster and in promoting their migratory capability. When and if such a condition persists, then organoid-like structures-in which cells are detached from the substratum and are recognizable as floating aggregates in the supernatant-appear [33]. In many microgravitybased models, the emergence of these organoid bodies is associated with the persistence of adherent cells, so it can be assumed that microgravity enacts the partitioning of living cells into two different populations, which is characterized by different morphological and functional properties [34]. However, by using the HaCaT model, we did not observe the formation of any organoid-like structure. On the contrary, these cells were shown to be able to counteract microgravity effects on both E-cadherin and F-actin, thus reestablishing a native, non-motile phenotype within 48-72 h of adaptation in simulated microgravity. These results suggest that caution should be kept when considering data from weightlessness-based experiments, given that different exposure times can generate different biological behaviors. Overall, living cells seem to successfully cope with the microgravity challenge by enacting an "adaptive strategy", which mostly relies on the intrinsic plasticity of cells.

Conclusions
HaCaT cells exposed to simulated microgravity show a biphasic response, displaying increased motility and invasiveness during the early period (24 h), followed by the recovering of a noninvasive morphological phenotype at 48 and 72 h. Although changes in the cytoskeleton and F-actin organization would in principle inhibit the migrating capability of HaCaT cells by reducing cell adhesiveness to the substratum, the concomitant reduction in cell-to-cell adhesion fostered by the significant decrease in the E-cadherin expression pattern promoted the appearance of a migrating/invasive phenotype. However, those changes are transitory, and the native phenotype is re-established within 48-72 h in microgravity cultures. It is worthy of interest that the morphological recovery is paired by the concomitant disappearance of main EMT features. A similar pattern can be observed as the same cells are seeded again into a normal gravitational field. Definitely, the emergence of a migratory phenotype should be considered in this cellular model as a transient, adaptive phenomenon enacted by the modified environmental physical forces.
Further studies to elucidate the molecular mechanisms behind the morphological and functional changes we observed are currently undergoing in our laboratory.