Addition of ROCK Inhibitors Alleviates Prostaglandin-Induced Inhibition of Adipogenesis in 3T3L-1 Spheroids

To elucidate the additive effects of the ROCK inhibitors (ROCK-i), ripasudil (Rip) and Y27632 on bimatoprost acid (BIM-A), a prostaglandin analog (PG), on adipose tissue, two- and three-dimensional (2D or 3D) cultures of 3T3-L1 cells, the most well characterized cells in the field of lipid research, were used. The cells were subjected to a variety of analyses including lipid staining, real-time cellular metabolic analysis, the mRNA expressions of genes related to adipogenesis and extracellular matrices (ECMs) as well as the sizes and physical properties of the 3D spheroids by a micro-squeezer. BIM-A induced strong inhibitory effects on most of the adipogenesis-related changes in the 2D and 3D cultured 3T3-L1 cells, including (1) the enlargement and softening of the 3D spheroids, (2) a dramatic enhancement in lipid staining and the expression of adipogenesis-related genes, and (3) a decrease in mitochondrial and glycolytic metabolic function. By adding ROCK-i to the BIM-A, most of these BIM-A-induced effects were cancelled. The collective findings reported herein suggest that ROCK-i eliminated the PG-induced suppression of adipogenesis in the 3T3-L1 cells, accompanied by the formation of enlarged 3D spheroids. Such effects of adding ROCK-i to a PG in preadipocytes on cellular properties appear to be associated with the suppression of PG-induced adverse effects, and provide additional insight into our understanding of lipid-related research.


Introduction
It is well known that adipocytes are involved as master regulators of energy homeostasis by storing large amounts of triacylglycerols or by mobilizing lipids in the cases of an excess of energy or nutritional deficiencies, respectively [1,2]. Adipocyte differentiation, a process in which preadipocytes proliferate and then differentiate into mature adipocytes, is precisely coordinated by various physiological mechanisms, including the gene expression of several adipogenesis-related factors and hormone sensitivity. It is also regulated by several intracellular signaling pathways [3]. Among these factors, prostaglandin derivatives (PGs) secreted from adipocytes both positively and negatively regulate adipogenesis [4]. For example, PGI2 promotes the progression of the differentiation of adipocytes into preadipocytes through PGI2 receptors (IP) by enhancing the expression of important transcription factors that are involved in the activation of the early stage of adipogenesis. These include the CCAAT/enhancer binding protein (C/EBP) β and δ [5,6], resulting in the subsequent activation of the expression of peroxisome proliferator-activated receptor γ (PPARγ), a critical transcription factor that promotes the maturation of adipocytes via adipogenesis. In fact, it was revealed that PGI2 stimulates high-fat diet (HFD)-induced obesity through As shown in these immunolabeled 3D 3T3-L1 spheroids with fibronectin (FN), the main component of the ECMs of the 3D spheroid, their sizes were increased upon adipogenic differentiation (DIF) as compared with those without DIF (CONT). Such DIF-induced effects were inhibited or facilitated by the PG, bimatoprost acid (BIM-A), or the ROCK inhibitor (ROCK-i), ripasudil (Rip), respectively. Although in the 3D spheroids of human orbital fibroblasts (HOFs), Rip suppressed these BIM-A-induced effects, such combined effects by PGs and ROCK-I on lipid metabolisms have not been elucidated yet. Scale bar: 100 μm. MEK: mitogen-activated protein kinase, ERK: extracellular signal-regulated kinase, CREB: the cyclic AMP response element binding protein.

Adipocyte Culture and Adipogenic Differentiation of 3T3-L1 Cells with or without PG and ROCK Inhibitors (ROCK-i)
Two-dimensional (2D) planar and three-dimensional (3D) spheroid cultures of 3T3-L1 preadipocytes (#EC86052701-G0, KAC, Osaka, Japan) and the induction of adipogenic differentiation were performed during 7 days, as described previously [32,33]. To study the drug efficacy of PG and ROCK-i, optimum concentrations of bimatoprost acid (BIM-A, 100 nM) with or without ROCK-is, Ripasudil (Rip, 10 μM) or Y27632 (10 μM), confirmed in our previous studies [32,33], were supplemented. As shown in these immunolabeled 3D 3T3-L1 spheroids with fibronectin (FN), the main component of the ECMs of the 3D spheroid, their sizes were increased upon adipogenic differentiation (DIF) as compared with those without DIF (CONT). Such DIFinduced effects were inhibited or facilitated by the PG, bimatoprost acid (BIM-A), or the ROCK inhibitor (ROCK-i), ripasudil (Rip), respectively. Although in the 3D spheroids of human orbital fibroblasts (HOFs), Rip suppressed these BIM-A-induced effects, such combined effects by PGs and ROCK-I on lipid metabolisms have not been elucidated yet. Scale bar: 100 µm. MEK: mitogen-activated protein kinase, ERK: extracellular signal-regulated kinase, CREB: the cyclic AMP response element binding protein.

Lipid Staining by Oil Red O (2D) or BODIPY (3D)
Lipid staining of the 2D cultured 3T3-L1 cells by an Oil Red O staining assay was performed using a commercial kit (#133102; Abcam, Cambridge, UK). Microscopy images were obtained with a Nikon A1 confocal microscope (Tokyo, Japan) and their quantification was performed by measuring the optical density (O.D.) of the dissolved dye at 500 nm.

Measurement of Real-Time Cellular Metabolic Functions
The oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) of the 2D 3T3-L1 cells were measured using a Seahorse XFe96 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA), as described previously, with minor modifications [34,35]. Briefly, 20 × 10 3 2D 3T3-L1 cells were placed in the wells of a 96-well assay plate in the absence or presence of BIM-A and/or ROCK-i, as above. After replacing the culture medium with a Seahorse XF DMEM assay medium (pH 7.4, #103575-100; Agilent Technologies), supplemented with 5.5 mM glucose, 2.0 mM glutamine and 1.0 mM sodium pyruvate, basal OCR and ECAR values were determined using a Seahorse XFe96 Bioanalyzer and thereafter, the samples were further analyzed after supplementation with 2.0 µM oligomycin, 5.0 µM carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP), 1.0 µM rotenone and antimycin A, and 10 mM 2-deoxyglucose (2-DG). The OCR and ECAR values were normalized to the amount of protein per well.

Physical Properties, Size and Stiffness Analyses of the 3D 3T3-L1 Spheroids
The configuration of the 3D spheroids was observed by phase contrast (PC; Nikon ECLIPSE TS2, Tokyo, Japan) [27]. The resulting photograph of the spheroid was converted into an 8-bit image and the outer circumference of the spheroid was measured using the Image J software version 1.51n (National Institutes of Health, Bethesda, MD, USA), and the area within the circumference was determined as the size of the spheroid. Alternatively, the stiffness (µN/µm), as determined by the force (µN) required to compress the diameter of the 3T3-L1 spheroids by 50% (µm), was determined by a micro-indentation force analysis using a micro-squeezer (CellScale, Waterloo, ON, Canada), as described in a previous study [25].

Other Analytical Methods
Quantitative PCR was performed using predesigned specific primers and statistical analyses, using Graph Pad Prism 8 (GraphPad Software, San Diego, CA, USA), were performed as described previously [32,33]. For an analysis of the difference between groups, a grouped analysis with a two-way analysis of variance (ANOVA) followed by a Tukey's multiple comparison test was performed. Data are presented as the arithmetic means ± the standard error of the mean (SEM).

Additive Effects of Rock-Is on PGs on 2D Cultured 3T3-L1 Cells
In our preceding study, we found that the clinically used ROCK-i, Rip, significantly suppressed the effects of BIM-A on 3D human orbital fibroblasts (HOFs) spheroids. The fact that BIM-A induces the most prominent effect on human orbital fatty tissues [36,37] among the cur- rently used PGs, suggests a possible crosslink between ROCK-i and PGs within adipocytes [26]. In the current study, to obtain currently unidentified insights into the crosslinking effect of ROCK-i and PGs toward fatty tissues in the case of the addition of ROCK-i to PG, we examined the additive effects of ROCK-i (Rip or Y27632) to BIM-A on the adipogenesis of 2D and 3D cultured 3T3-L1 cells, the most frequently used adipocyte in adipocyte-related research [38]. Oil Red O lipid staining and the gene expression of the DIF+-related genes in the 2D cultured 3T3-L1 cells were significantly enhanced upon adipogenesis, and markedly suppressed by BIM-A, except for the expressions of Ap2, as described in our previous studies [32,33,39]. Upon the addition of ROCK-i, the staining intensities for Oil Red O again increased significantly. In contrast, however, the expressions of Pparγ, Ap2 and Leptin were further markedly decreased or slightly increased by the addition of Rip or Y27632, respectively.   (C1-C3)). All experiments were performed in triplicate using fresh preparations, each consisting of 5 specimens. Data are presented as the arithmetic means ± the standard error of the mean (SEM). ** p < 0.01, *** p < 0.005, **** p < 0.001 (ANOVA followed by a Tukey's multiple comparison test). In terms of the mRNA expression of ECMs ( Figure 3), a marked downregulation or upregulation was observed in Col1 and Fn, or Col4 and Col6, respectively, upon DIF+ addition. In the presence of BIM-A, although the DIF+-induced changes in these ECMs were not significantly altered, all four ECM molecules were substantially downregulated by the addition of ROCK-i to BIM-A. Real-time cellular metabolic measurements ( Figure 4) demonstrated that although both OCR and ECAR were significantly reduced by a monotreatment with either BIM-A and ROCK-i, the values were, to the contrary, enhanced in the case of BIM-A and Y27632. Therefore, these collective results suggest that some unknown synergy, and not simply the additive effects of ROCK-i and BIM-A, may be at play here. demonstrated that although both OCR and ECAR were significantly reduced by a monotreatment with either BIM-A and ROCK-i, the values were, to the contrary, enhanced in the case of BIM-A and Y27632. Therefore, these collective results suggest that some unknown synergy, and not simply the additive effects of ROCK-i and BIM-A, may be at play here.   ) and extracellular acidification rate (ECAR, panel (B)) were measured, and thereafter, they were further measured after subsequent supplementation with oligomycin (a complex V inhibitor), FCCP (a protonophore), and rotenone/antimycin A (complex I/III inhibitors) and 2DG (a hexokinase inhibitor). The main parameters of the cellular metabolic analysis are shown in panels (C) and (D), respectively. Basal OCR was calculated by subtracting the OCR with rotenone/antimycin A from the OCR at baseline. ATP-linked respiration was analyzed by subtracting (a hexokinase inhibitor). The main parameters of the cellular metabolic analysis are shown in panels (C,D), respectively. Basal OCR was calculated by subtracting the OCR with rotenone/antimycin A from the OCR at baseline. ATP-linked respiration was analyzed by subtracting the OCR with oligomycin from the OCR at baseline. Maximal respiration was calculated by subtracting the OCR with rotenone/antimycin A from the OCR with FCCP. The basal ECAR was calculated by subtracting the ECAR with 2DG from the ECAR at baseline. Glycolytic capacity was calculated by subtracting the ECAR with 2DG from the ECAR with oligomycin. The glycolytic reserve was calculated by subtracting the ECAR at baseline from the ECAR with oligomycin. All experiments were performed in triplicate using fresh preparations (n = 3). Data are presented as the mean ± the standard error of the mean (SEM). * p < 0.05 (ANOVA followed by a Tukey's multiple comparison test).

Additive Effects of ROCK-Is to PGs on 2D Cultured 3T3-L1 Cells
Since the nature of adipogenesis of the 3T3-L1 cells was significantly different between the 2D planar cultures and 3D spheroid culture conditions [31], we also studied the additive effect of ROCK-i to BIM-A on DIF+ in 3D 3T3-L1 spheroids that were prepared as describe in a previous study [27,33] The following results were consistent with the results reported in a previous study: (1) DIF-induced 3D 3T3-L1 spheroids became smaller during the 7-day culture; (2) upon adipogenesis (DIF+), their mean area sizes became significantly larger; and (3) such DIF+induced effects were significantly inhibited in the presence of 100 nM BIM-A. The addition of 10 µM ROCK-i to 100 nM BIM-A induced a substantial increase in the sizes of 3D spheroids until Day 7 ( Figure 5). Concerning the micro-squeezer analysis, the 3D 3T3-L1 spheroids became physically softened upon adipogenesis (DIF+), and this stiffness was substantially increased or decreased by the treatment of BIM-A or BIM-A/and ROCK-I ( Figure 6).
Bioengineering 2022, 9, x FOR PEER REVIEW 10 of 21 the OCR with oligomycin from the OCR at baseline. Maximal respiration was calculated by subtracting the OCR with rotenone/antimycin A from the OCR with FCCP. The basal ECAR was calculated by subtracting the ECAR with 2DG from the ECAR at baseline. Glycolytic capacity was calculated by subtracting the ECAR with 2DG from the ECAR with oligomycin. The glycolytic reserve was calculated by subtracting the ECAR at baseline from the ECAR with oligomycin. All experiments were performed in triplicate using fresh preparations (n = 3). Data are presented as the mean ± the standard error of the mean (SEM). * p < 0.05 (ANOVA followed by a Tukey's multiple comparison test).

Additive Effects of ROCK-Is to PGs on 2D Cultured 3T3-L1 Cells
Since the nature of adipogenesis of the 3T3-L1 cells was significantly different between the 2D planar cultures and 3D spheroid culture conditions [31], we also studied the additive effect of ROCK-i to BIM-A on DIF+ in 3D 3T3-L1 spheroids that were prepared as describe in a previous study [27,33] The following results were consistent with the results reported in a previous study: (1) DIF-induced 3D 3T3-L1 spheroids became smaller during the 7-day culture; (2) upon adipogenesis (DIF+), their mean area sizes became significantly larger; and (3) such DIF+induced effects were significantly inhibited in the presence of 100 nM BIM-A. The addition of 10 μM ROCK-i to 100 nM BIM-A induced a substantial increase in the sizes of 3D spheroids until Day 7 ( Figure 5). Concerning the micro-squeezer analysis, the 3D 3T3-L1 spheroids became physically softened upon adipogenesis (DIF+), and this stiffness was substantially increased or decreased by the treatment of BIM-A or BIM-A/ and ROCK-I ( Figure 6).  (CONT) or their adipogenic differentiation (DIF) with or without the combination of 100 nM BIM-A (BIM-A) and ROCK-i (10 µM Ripasudil (Rip) or 10 µM Y27632). The mean area sizes (µm 2 ) of the spheroids were measured and compared among experimental groups on Day 7. All experiments were performed in triplicate using fresh preparations, each consisting of 16 spheroids. Data are presented as the arithmetic means ± the standard error of the mean (SEM). ** p < 0.01, **** p < 0.001 (ANOVA followed by a Tukey's multiple comparison test). eering 2022, 9, x FOR PEER REVIEW experiments were performed in triplicate using fresh preparation Data are presented as the arithmetic means ± the standard error o p < 0.001 (ANOVA followed by a Tukey's multiple comparison te As shown in Figure 7A,B, the DIF+-induced enhance intensities was significantly inhibited in the presence of BIM served upon the addition of ROCK-i to BIM-A. The mRNA sis-related genes was also significantly increased with DIF The specimens collected on Day 7 were subjected to a physical solidity analysis. Among the above experimental conditions, the requiring force (µN) was measured and force/displacement (µN/µm) was potted (right panel). ** p < 0.01, *** p < 0.005, **** p < 0.001 (ANOVA followed by a Tukey's multiple comparison test).

Lipid
Staining with BODIPY and the Quantitative PCR of Adipogenesis (DIF+)-Related Genes, including Pparγ, Ap2 and Leptin ( Figure 6) As shown in Figure 7A,B, the DIF+-induced enhancement of the BODIPY staining intensities was significantly inhibited in the presence of BIM-A, while no change was observed upon the addition of ROCK-i to BIM-A. The mRNA expression of the adipogenesis-related genes was also significantly increased with DIF+. However, inconsistent with the BODIPY staining results, the addition of ROCK-i to BIM-A caused an enhancement in the mRNA expression of these adipogenesis-related genes ( Figure 7C1-C3).   All experiments were performed in duplicate using fresh preparations, each consisting of 16 spheroids. Data are presented as the arithmetic means ± the standard error of the mean (SEM). * p < 0.05, ** p < 0.01, *** p < 0.005, **** p < 0.001 (ANOVA followed by a Tukey's multiple comparison test).

The Expression of ECMs including Collagen (Col) 1, 4 and 6, and Fibronectin (Fn)
Analyzed by qPCR ( Figure 8) and Immunocytochemistry (Figure 9) Concerning the mRNA expression of ECMs (Figure 8), BIM-A caused significant changes in the DIF+-induced downregulation of Col1 and Fn, and the upregulation of Col4 and Col6, similar to the results of the 2D cell culture experiment described above. However, the additive effects were also different between the 2D and 3D spheroids. That is, the addition of Rip to BIM-A caused a significant increase in Col1, Col4, and Col6 and a relative increase in Fn, but Y27632 caused a significant increase only in Col4. In terms of this discrepancy for the ECM expressions between the 2D and 3D cell cultures, we previously reported that the nature of 2D and 3D cultures was significantly different based upon their trypsin sensitivity [33] as well as the efficacy of adipogenesis [32].

The Expression of ECMs Including Collagen (Col) 1, 4 and 6, and Fibronectin (Fn)
Analyzed by qPCR ( Figure 8) and Immunocytochemistry (Figure 9) Concerning the mRNA expression of ECMs (Figure 8), BIM-A caused significant changes in the DIF+-induced downregulation of Col1 and Fn, and the upregulation of Col4 and Col6, similar to the results of the 2D cell culture experiment described above. However, the additive effects were also different between the 2D and 3D spheroids. That is, the addition of Rip to BIM-A caused a significant increase in Col1, Col4, and Col6 and a relative increase in Fn, but Y27632 caused a significant increase only in Col4. In terms of this discrepancy for the ECM expressions between the 2D and 3D cell cultures, we previously reported that the nature of 2D and 3D cultures was significantly different based upon their trypsin sensitivity [33] as well as the efficacy of adipogenesis [32].  In contrast to the mRNA expression data, immunolabeling indicated that: (1) the levels of COL1 or COL4 significantly decreased upon the DIF+ treatment; (2) the addition of BIM-A to DIF+ induced a significant increase in COL1; (3) the addition of Rip to BIM-A caused a decrease in COL1 and an increase in COL6 levels; and (4) the addition of Y27632 to BIM-A induced an increase in all ECMs, except for COL1 (Figure 9). These discrepancies between gene expression and immunostaining can be attributed to the spatial distribution of the individual ECM molecules that are expressed within the 3D conformation of the spheroid, as was suggested in a previous study [33].
Bioengineering 2022, 9, x FOR PEER REVIEW 14 of 21 In contrast to the mRNA expression data, immunolabeling indicated that: (1) the levels of COL1 or COL4 significantly decreased upon the DIF+ treatment; (2) the addition of BIM-A to DIF+ induced a significant increase in COL1; (3) the addition of Rip to BIM-A caused a decrease in COL1 and an increase in COL6 levels; and 4) the addition of Y27632 to BIM-A induced an increase in all ECMs, except for COL1 (Figure 9). These discrepancies between gene expression and immunostaining can be attributed to the spatial distribution of the individual ECM molecules that are expressed within the 3D conformation of the spheroid, as was suggested in a previous study [33].  The staining intensities of the ECMs of the spheroids that were stained as above are plotted. All experiments were performed in duplicate using fresh preparations consisting of 5 spheroids each. Data are presented as the arithmetic mean ± the standard error of the mean (SEM). * p < 0.05, ** p < 0.01, *** p < 0.005, **** p < 0.001 (ANOVA followed by a Tukey's multiple comparison test).
These collective results indicate that although the BIM-A-treated 3D spheroids were small and densely packed with lipids, the addition of ROCK-i to BIM-A induced 3D 3T3-L1 spheroids with greatly increased sizes, presumably due to the rich Cols framework as compared to 3D spheroids without ROCK-I and comparable amounts of lipids within both 3D spheroids.

Discussion
Since orbital fatty tissue grows within a three-dimensional (3D) conic space, our group recently established a more suitable and representative in vitro model using a three-dimensional (3D) drop culture technique to replicate this adipocyte-spreading environment [25], using human orbital fibroblasts (HOFs) [27,28] as well as 3T3-L1 cells [33], which are the most extensively characterized cells that are used in adipogenesis-related research [40]. Härmä et al. recently reported that the use of automated image analysis for the evaluation of phenotypic or morphometric aspects facilitated cell-based 3D assays in basic research as well as in drug discovery and target validation [41]. In an independent study, we also demonstrated that the characterization of the physical properties, sizes and stiffness of the 3D spheroids could also be used for the evaluation of several drug-induced effects [26][27][28]32,33]. For example, most recently, we found that ROCK-i significantly inhibited the downsizing and hardening effects of 3D HOFs spheroids that were induced by a PG, namely, bimatoprost acid (BIM-A) [26]. However, and quite interestingly, we The staining intensities of the ECMs of the spheroids that were stained as above are plotted. All experiments were performed in duplicate using fresh preparations consisting of 5 spheroids each. Data are presented as the arithmetic mean ± the standard error of the mean (SEM). * p < 0.05, ** p < 0.01, *** p < 0.005, **** p < 0.001 (ANOVA followed by a Tukey's multiple comparison test).
These collective results indicate that although the BIM-A-treated 3D spheroids were small and densely packed with lipids, the addition of ROCK-i to BIM-A induced 3D 3T3-L1 spheroids with greatly increased sizes, presumably due to the rich Cols framework as compared to 3D spheroids without ROCK-I and comparable amounts of lipids within both 3D spheroids.

Discussion
Since orbital fatty tissue grows within a three-dimensional (3D) conic space, our group recently established a more suitable and representative in vitro model using a threedimensional (3D) drop culture technique to replicate this adipocyte-spreading environment [25], using human orbital fibroblasts (HOFs) [27,28] as well as 3T3-L1 cells [33], which are the most extensively characterized cells that are used in adipogenesis-related research [40]. Härmä et al. recently reported that the use of automated image analysis for the evaluation of phenotypic or morphometric aspects facilitated cell-based 3D assays in basic research as well as in drug discovery and target validation [41]. In an independent study, we also demonstrated that the characterization of the physical properties, sizes and stiffness of the 3D spheroids could also be used for the evaluation of several drug-induced effects [26][27][28]32,33]. For example, most recently, we found that ROCK-i significantly in-hibited the downsizing and hardening effects of 3D HOFs spheroids that were induced by a PG, namely, bimatoprost acid (BIM-A) [26]. However, and quite interestingly, we also found that the biological features of the 3D spheroids were different between HOFs and 3T3-L1 cells; that is, DIF+ induced an increase or decrease in the stiffness of HOFs [26][27][28] and 3T3-L1 cells [32,33], respectively. These facts rationally suggest that the additive effects of ROCK-i on PGs may also be different between HOFs and 3T3-L1 cells.
Adipogenic differentiation from precursor stem cells to adipocytes is thought to be accomplished by two steps: determination and terminal differentiation [42]. Among these, the terminal differentiation step is relatively well identified as compared to the determination step. It has also been revealed that such adipogenesis processes are critically regulated by several key genes; that is, essential regulators of adipogenesis, the peroxisome proliferatoractivated receptor γ (PPARγ) and CCAAT enhancer-binding protein α (C/EBPα) induce the expression of metabolic genes, including glucose transporter 4 (GLUT4), fatty acid binding protein 4 (FABP4) and leptin [42]. Throughout these gene functions, lipid droplets begin to appear within the cytoplasm and grow into mature adipocytes [42]. This adipogenesis process is known to also be modulated by several extracellular signaling proteins, including insulin/IGF-1, TGFβ, WNT10b, FGF, BMPs, PGs and others. Among these, PG-induced suppression of adipogenesis has begun to draw a lot of attention as a possible DUES etiology in the ophthalmology field. In addition, a significant contribution of ROCK signaling toward lipid adipogenesis has also been suggested [16]. That is, it was shown that the Rho/ROCK signaling pathway negatively regulates adipocyte differentiation, in which several underlying mechanisms, including a mechanism mediated by WNT genes, the regulation of actinomyosin formation, which is a key determinant of adipogenesis, the inhibition of the insulin requirement, and others have been suggested. However, such studies regarding ROCK signaling as it relates to adipogenesis have not been accomplished. Furthermore, except for our previous study using HOFs [26], no additional studies have been reported suggesting the existence of a possible crosslink between ROCK and PG signaling within lipid adipogenesis. In the current study, using a universal adipocyte cell line, 3T3-L1 cells, we confirmed that a crosslink exists in both signaling within lipid adipogenesis in general based upon the observations that the addition of ROCK-i cancelled the BIM-induced suppression of the several DIF+ structural and functional properties in both 2D and 3D cultured 3T3-L1 cells.
In terms of the anti-adipogenic effect of PGF2α, the FP-receptor-activated mitogenactivated protein kinase/extracellular signal-regulated kinase, the (MEK)/extracellular signal-regulated kinase (ERK) cascade and the binding of the cyclic AMP response element (CRE) binding protein (CREB) to the COX-2 promoter are known to be involved. Interestingly, in addition to this pathway, it is also known that the activation of the FP receptor by PGF2α is linked to the Gα12-ROCK signaling pathway. This observation suggests that the underlying mechanism may cause a crosslinkage between ROCK-i and PGs on adipogenesis that was reported in the current study. To support this possibility, Gα12 signaling appears to be involved in the regulation of the growth of skeletal muscle by modulating the metabolism of ECMs, similar to the current observation that pan-ROCK-is, Rip or Y27632 caused significant changes in the physical properties of the 3T3-L1 spheroids by modulating the expression of ECM molecules in addition to adipogenesis [32] (Table 1). Interestingly, in the case in which combinations of BIM-A and ROCK-i were used, in addition to some of the changes being able to be rationally explained by their independent effects, other currently unexplained and unidentified changes (designated as N.I.) were also recognized ( Table 1). More interestingly, such combined effects by BIM-A and ROCK-i were also different between the 3T3-L1 cells and HOFs [26] (Table 1). Therefore, given the above collective findings, these findings strongly support the existence of relationships between ROCK-i and PGs on adipogenesis. Thus, the mechanisms responsible for such unidentified relationships will need to be investigated in a future project. Table 1. Summary of the effects of mono-or combined treatments of PGs and ROCK inhibitors on the physical properties of the 3D 3T3-L1 spheroids, and adipogenesis and gene expression in the 2D planar and 3D spheroid cultured 3T3-L1 cells and human orbital fibroblasts (HOFs).
In conclusion, in the current study, we report on new observations to suggest that the addition of ROCK-i cancelled the PG-induced suppression of adipogenesis in 3T3-L1 cells, resulting in the formation of greatly enlarged 3D spheroids. This points to additional insights into our understanding of lipid-related research in general. However, the current study also had the following limitations: (1) It is known that there are two functionally diverse isoforms of ROCKs, ROCK 1 and ROCK 2, in a variety of tissues, including adipocytes [43][44][45][46][47][48]. Among these, ROCK2, but not ROCK1, is thought to be responsible for the anti-adipogenic activity of Rho in 3T3-L1 and mouse embryonic fibroblasts (MEFs) [49,50], in which ROCK1 and/or ROCK2 are responsible for the unidentified crosslinking with PGs. In fact, the ROCK2-i, KD025, induced quite different effects with PGs in 3D HOFs spheroids [26]. However, as shown above, since the drug-induced effects by ROCK-is and/or BIM-A also are substantially different between 3T3-L1 and HOFs (Table 1), additional studies using siRNA or isotypespecific inhibitors for ROCK 1 or 2, to determine which isoform among ROCK1 and ROCK2 plays a key role, will need to be carried out. (2) The mechanism responsible for causing such large and soft 3D spheroids upon the addition of ROCK-i to BIM-A has not been identified yet. As a possible mechanism, we speculate that the additive effects of ROCK-i to BIM-A induced: (1) a significant upregulation of COL6, which mainly makes up the structural frame of the 3D spheroid, and (2) relatively insufficient amounts of lipid and FN content within the 3D spheroid may contribute to the formation of such large-sized and soft 3D spheroids, as shown in Figure 10. In support of this speculation, it has been reported that ROCK contributes to the actin cytoskeleton and fibronectin matrix assembly [51][52][53], and that ROCK-i alters cytoskeletal arrangement and cell shape [54,55]. In addition, we also found that ROCK-i only caused the formation of large and soft 3D 3T3-L1 spheroids as was observed in our previous study [32], the effects of which were similar, but less intense, compared to the combination of ROCK-i and PG [32]. However, real-time cellular metabolic measurements by a Seahorse Bioanalyzer identified some quite new and interesting results, suggesting the existence of some unknown type of synergy, and that not simply the additive effects of ROCK-i and BIM-A are involved in this process. Therefore, to elucidate the underlying mechanisms responsible for this, we also plan to perform RNA sequence analyses to identify the types of upstream and downstream regulations that are involved in this process. BIM-A: bimatoprost acid, Rip: ripasudil, 2D: two-dimensional planar culture, 3D: three-dimensional spheroid culture, Pparγ: peroxisome proliferator-activated receptor γ, Ap2: adipocyte protein 2, Col: collagen, Fn: fibronectin, ↑: significant increase (p < 0.05), ↑↑: significant increase (p < 0.01), ↑↑↑: significant increase (p < 0.005), ↓: significant decrease (p < 0.05), ↓↓: significant decrease (p < 0.01), ↓↓↓: significant decrease (p < 0.005), * results for Rip and Y27632 from 3T3-L1 cells [17] and ** HOFrelated data [10] were taken from our previous studies. Parentheses indicate the effects of combinations of BIM-A (B), and Rip (R) or Y27632 (Y), and can be explained by their independent actions; for example, B < R refers to the finding that the effect of B was higher than for R, or cannot be explained because these were affected non-independently together (N.I.).  As shown in the immunolabeled 3D spheroids, a significant upregulation of collagen6 (COL6) as the main component of the structural frame of the 3D spheroid, and relatively insufficient amounts of lipid and fibronectin (FN) content within the 3D spheroid, represent a possible mechanism for the formation of the huge-sized and soft 3D spheroids upon the administration of ROCK-i to BIM-A. Scale bar: 100 µm. CONT: 3T3-L1 preadipocyte, Rip+BIM-A: adipogenic differentiated 3T3-L1 in the presence of R-p and BIM-A, BODIPY: lipid staining by BODIPY.

Conflicts of Interest:
The authors declare no conflict of interest.