The Influence of Coumestrol on Sphingolipid Signaling Pathway and Insulin Resistance Development in Primary Rat Hepatocytes

Coumestrol is a phytoestrogen widely known for its anti-diabetic, anti-oxidant, and anti-inflammatory properties. Thus, it gets a lot of attention as a potential agent in the nutritional therapy of diseases such as obesity and type 2 diabetes. In our study, we evaluated whether coumestrol affects insulin resistance development via the sphingolipid signaling pathway in primary rat hepatocytes. The cells were isolated from the male Wistar rat’s liver with the use of collagenase perfusion. Next, we incubated the cells with the presence or absence of palmitic acid and/or coumestrol. Additionally, some groups were incubated with insulin. The sphingolipid concentrations were assessed by HPLC whereas the expression of all the proteins was evaluated by Western blot. Coumestrol markedly reduced the accumulation of sphingolipids, namely, ceramide and sphinganine through noticeable inhibition of the ceramide de novo synthesis pathway in insulin-resistant hepatocytes. Moreover, coumestrol augmented the expression of fatty acid transport proteins, especially FATP5 and FAT/CD36, which also were responsible for excessive sphingolipid accumulation. Furthermore, coumestrol altered the sphingolipid salvage pathway, which was observed as the excessive deposition of the sphingosine-1-phosphate and sphingosine. Our study clearly showed that coumestrol ameliorated hepatic insulin resistance in primary rat hepatocytes. Thus, we believe that our study may contribute to the discovery of novel preventive and therapeutic methods for metabolic disorders.


Introduction
The worldwide prevalence of obesity and type 2 diabetes mellitus (T2DM) have spread extensively throughout the last three decades as a consequence of a sedentary lifestyle and unhealthy diet, becoming enormous public health and social burdens [1,2]. There is no doubt that one of the most crucial factors involved in the pathogenesis of these metabolic disorders is insulin resistance. It can be defined as an impaired response to insulin in the skeletal muscles, adipose tissue, and liver. As a result, hyperinsulinemia, hyperglycemia and increased intracellular lipid concentrations in hepatocytes occur, which may finally lead to hepatic steatosis [3,4]. Elevated liver lipid concentrations are the result of (1) insulin signaling impairments, (2) intensified fatty acid (FA) transport overcoming β-oxidation capability, and (3) increased FA esterification into numerous biologically active lipid fractions (including sphingolipids), further augmenting deteriorations in hepatic insulin responsiveness. Numerous studies indicated a clear interplay between the development of insulin resistance and alterations in sphingolipid metabolism. The increased amount of saturated free fatty acids (FFA), like palmitic acid (PA), in the diet results in the promotion of the de novo synthesis of ceramide (CER), one of the most abundant all the incubation periods. Each experiment was always carried out on 10 6 cells per well, in which the percentage of living cells was above 85%.

Experimental Procedures
The treatment with palmitic acid was conducted on serum-starved cells, as previously described [20]. Before the addition of palmitate to the medium, it was conjugated with fatty acid-free bovine serum albumin (BSA, Sigma Aldrich, St. Louis, MO, USA). Briefly, palmitate stock solution was prepared by dissolving PA in a mixture of absolute ethanol and 1 M NaOH, heating to 70 • C, and conjugating with 10% BSA. Next, the stock solution was diluted in serum-free DMEM, containing 10 mM HEPES [6]. The cells were incubated in the presence or absence of palmitic acid at the concentration of 0.5 mM for 18 h. Among these two groups, half of the cells were incubated with 20 or 50 µM of coumestrol and 20 or 50 µM of COM together with 0.5 mM of PA for 18 h. The concentrations of coumestrol used in our experiment were selected based on preliminary studies, which showed that, compared to 20 µM, the 10-µM concentration did not exert statistically significant changes and 60 µM exhibited the same effects as 50 µM (see Supplementary Figure S7). In another set of experiments, the hepatocytes were divided into the same experimental groups as mentioned above and were additionally chased by 100 nM insulin (NovoRapid, Novo Nordisk, ON, Canada) for the last 15 min of 18 h incubation at 37 • C. At the end of each experimental set cell morphology, viability was assessed as well (i.e., Trypan blue staining). Then, all the cells were washed three times with PBS, harvested, and homogenized in ice-cold RIPA (radioimmunoprecipitation assay buffer) lysis buffer containing protease and phosphatase inhibitors (Roche Diagnostics GmbH, Mannheim, Germany).

Intracellular Sphingolipid Analyses
The content of ceramide, sphingosine (SFO), sphinganine (SFA), and sphingosine-1-phosphate (S1P) was determined as previously described in detail [22]. In brief, hepatocytes were homogenized by ultrasonication and lipids were extracted into chloroform. An aliquot of the lipid extract was transferred to a fresh tube with pre-added 40 pmol of N-palmitoyl-D-erythro-sphingosine (C17 base) (Avanti Polar Lipids) as an internal standard and then subjected to alkaline hydrolysis to deacylate ceramide to sphingosine. Free sphinganine, as well as sphingosine released from ceramide, were then converted to their o-phthalaldehyde derivatives and analyzed using an HPLC (high-performance liquid chromatography, PROSTAR, Varian Inc., Palo Alto, CA, USA) system equipped with a fluorescence detector and C18 reversed-phase column (Varian Inc. OmniSpher 5, 4.6 × 150 mm). The isocratic eluent composition of acetonitrile (Merck):water (9:1, v/v) and a flow rate of 1 mL/min was used. Before the sphingolipid analysis, the protein content was measured in all the samples with bovine serum albumin (Sigma-Aldrich, St. Louis, MO, USA) as a standard.

Data Analysis
The data are shown as mean ± SD based on six independent determinations (n = 6). The normality of the data distribution (Shapiro-Wilk test) and homogeneity of the variance (Bartlett's test) were checked. The statistical differences of the data were performed in GraphPad Prism 5 (GraphPad Software, CA, USA) by one-way ANOVA analysis followed by an appropriate post hoc test (pairwise Student's t-test). The results were considered to be statistically significant at p < 0.05.

Effects of Coumestrol on the Expression of Insulin Signaling Pathway Proteins in Primary Rat Hepatocytes
The expression of proteins involved in the insulin signaling pathway was affected during the incubation of primary rat hepatocytes with palmitate (PA) and/or COM in the presence of insulin. We observed a significant decline in pAkt/Akt (phosphorylated protein kinase B/protein kinase B) and pGSK-3β/GSK-3β (phosphorylated glycogen synthase kinase 3β/glycogen synthase kinase 3β) expression ratios after incubation with palmitate and insulin (PA + I, 2-fold lower, p < 0.05, Figure 2A2; 3.6-fold lower, p < 0.05, Figure 2B2; respectively) compared with the C + I group. In turn, considerable elevation in expression ratio of pGSK-3β/GSK-3β was observed after incubation with PA + COM20 + I as well as PA + COM50 + I (3-fold higher, p < 0.05, and 2.9-fold higher, p < 0.05, Figure 2B2, respectively) in comparison with the PA + I group. We did not observe statistically significant results in the expression of pAkt/Akt in the groups incubated simultaneously with PA and COM. However, a trend toward increase was observed in PA + COM20 + I group (p = 0.09, Figure  2A2).

Figure 2.
Representative Western blot membranes of evaluated proteins, i.e., pAkt/Akt-phosphorylated protein kinase B/protein kinase B (A1), pGSK-3β/GSK-3β-phosphorylated glycogen synthase kinase 3β/glycogen synthase kinase 3β (B1), and the corresponding images of total protein staining after membrane transfer used for normalization. The expression of proteins involved in insulin signaling pathway, i.e., pAkt/Akt (A2), pGSK-3β/GSK-3β (B2) in primary rat hepatocytes. The cells were incubated with presence or absence of PA (0.5 mM) or/and COM (20 or 50 μM). Additionally, every group was also incubated with insulin, as it was described in Materials and Methods section. The expression of the proteins was evaluated by the Western blot technique. The data are shown as the mean ± SD and are based on six independent determinations (n = 6); * p < 0.05 significant difference vs. control + insulin group; # p < 0.05 significant difference vs. palmitate + insulin group; C + I, control + insulin; COM20 + I or COM50 + I, coumestrol 20 or 50 μM + insulin; PA + I, palmitic acid + insulin; PA + COM20 + I or PA + COM50 + I, palmitic acid + coumestrol 20 or 50 μM + insulin.

Effects of Coumestrol on the Expression of Enzymes Involved in Ceramide de Novo Synthesis Pathway in Primary Rat Hepatocytes
Incubation of primary rat hepatocytes with COM50 and PA + COM 20 as well as PA + COM50 resulted in substantial elevation of the SPTLC1 expression compared to the control group (COM50: +118.56%, p < 0.05; PA + COM20: +97.76%, p < 0.05; PA + COM50: +190.1%, p < 0.05, Figure 3A2). Furthermore, we noticed a significant decrease in the expression of SPTLC1 after incubation with PA + COM20 as well as PA + COM50 in comparison with the PA group (−43.7%, p < 0.05; −17.42%, p < 0.05, Figure 3A2, respectively). In turn, the expression of SPTLC2 significantly dropped in COM20 and PA + COM20 groups in comparison with the control group (−56.44%, p < 0.05 and −54.59%, p < 0.05, Figure 3B2, respectively). The expression of SPTLC2 was also decreased in the group incubated with PA + COM20 compared to the PA group (−64.54%, p < 0.05, Figure 3B2).

Figure 2.
Representative Western blot membranes of evaluated proteins, i.e., pAkt/Akt-phosphorylated protein kinase B/protein kinase B (A1), pGSK-3β/GSK-3β-phosphorylated glycogen synthase kinase 3β/glycogen synthase kinase 3β (B1), and the corresponding images of total protein staining after membrane transfer used for normalization. The expression of proteins involved in insulin signaling pathway, i.e., pAkt/Akt (A2), pGSK-3β/GSK-3β (B2) in primary rat hepatocytes. The cells were incubated with presence or absence of PA (0.5 mM) or/and COM (20 or 50 µM). Additionally, every group was also incubated with insulin, as it was described in Materials and Methods section. The expression of the proteins was evaluated by the Western blot technique. The data are shown as the mean ± SD and are based on six independent determinations (n = 6); * p < 0.05 significant difference vs. control + insulin group; # p < 0.05 significant difference vs. palmitate + insulin group; C + I, control + insulin; COM20 + I or COM50 + I, coumestrol 20 or 50 µM + insulin; PA + I, palmitic acid + insulin; PA + COM20 + I or PA + COM50 + I, palmitic acid + coumestrol 20 or 50 µM + insulin.

Effects of Coumestrol on the Expression of Enzymes Involved in Ceramide De Novo Synthesis Pathway in Primary Rat Hepatocytes
Incubation of primary rat hepatocytes with COM50 and PA + COM 20 as well as PA + COM50 resulted in substantial elevation of the SPTLC1 expression compared to the control group (COM50: +118.56%, p < 0.05; PA + COM20: +97.76%, p < 0.05; PA + COM50: +190.1%, p < 0.05, Figure 3A2). Furthermore, we noticed a significant decrease in the expression of SPTLC1 after incubation with PA + COM20 as well as PA + COM50 in comparison with the PA group (−43.7%, p < 0.05; −17.42%, p < 0.05, Figure 3A2, respectively). In turn, the expression of SPTLC2 significantly dropped in COM20 and PA + COM20 groups in comparison with the control group (−56.44%, p < 0.05 and −54.59%, p < 0.05, Figure 3B2, respectively). The expression of SPTLC2 was also decreased in the group incubated with PA + COM20 compared to the PA group (−64.54%, p < 0.05, Figure 3B2).  , and the corre images of total protein staining after membrane transfer used for normalization. The enzymes involved in ceramide de novo synthesis pathway, i.e., SPTLC1 (A2) and SP primary rat hepatocytes. The cells were incubated with presence or absence of PA (0 COM (20 or 50 μM), as it was described in Materials and Methods section. The prote was evaluated by the Western blot technique. The data are shown as the mean ± SD on six independent determinations (n = 6); * p < 0.05 significant difference vs. control 0.05 significant difference vs. palmitate group; C, control; COM20 or COM50, coume μM; PA, palmitic acid; PA + COM20 or PA + COM50, palmitic acid + coumestrol 20 o

Effects of Coumestrol on the Expression of Enzymes That Share Ceramide De N Synthesis and Salvage Pathways in Primary Rat Hepatocytes
We observed notable changes in the expression of CerS6, CerS4, CerS after incubation with PA + COM20 compared to the PA group (−63.03%, p 4A2; −56.74%, p < 0.05, Figure 4B2; −39.04%, p < 0.05, Figure 4C2; −20.67%, p 4D2, respectively). The expression of CerS4 decreased significantly in grou with COM20 and PA + COM20 in comparison with the control group (−55 −53.1%, p < 0.05, Figure 4B2, respectively). Furthermore, the expression of C creased after incubation with PA + COM50 (−17.03%, p < 0.05, Figure 4D2). Representative Western blot membranes of evaluated proteins, i.e., SPTLC1-serine palmitoyl transferase 1 (A1), SPTLC2-serine palmitoyl transferase 2 (B1), and the corresponding images of total protein staining after membrane transfer used for normalization. The expression of enzymes involved in ceramide de novo synthesis pathway, i.e., SPTLC1 (A2) and SPTLC2 (B2) in primary rat hepatocytes. The cells were incubated with presence or absence of PA (0.5 mM) or/and COM (20 or 50 µM), as it was described in Materials and Methods section. The protein expression was evaluated by the Western blot technique. The data are shown as the mean ± SD and are based on six independent determinations (n = 6); * p < 0.05 significant difference vs. control group; # p < 0.05 significant difference vs. palmitate group; C, control; COM20 or COM50, coumestrol 20 or 50 µM; PA, palmitic acid; PA + COM20 or PA + COM50, palmitic acid + coumestrol 20 or 50 µM.

Effects of Coumestrol on the Expression of Enzymes Involved in the Sphingolipid Salvage Pathway in Primary Rat Hepatocytes
We noticed considerable increase in the expression of SPHK1 after incubation with COM50 and PA + COM20 as well as PA + COM50 groups (+74.82%, p < 0.05; +206.47%, p < 0.05; +71.48%, p < 0.05, Figure 5A2, respectively) compared with the control. Furthermore, the incubation with PA + COM20 resulted in increased expression of the SPHK1 and SPHK2 compared with the PA group (+75.54%, p < 0.05, Figure 5A2; +91.74%, p < 0.05, Figure 5B2; respectively). In contrary, we observed notable decline in the expression of SPHK2 after incubation with PA + COM50 compared to the control (−25.71%, p < 0.05, Figure 5B2). Incubation with PA + COM20 as well as PA + COM50 resulted in significantly increased ASAH1 expression in relation to the PA group (+199.35%, p < 0.05; +204.57, p < 0.05; Figure 5C2, respectively). What is more, significant elevation of ASAH3 expression was detected in PA + COM20 as well as PA + COM50 groups +105.1%, p < 0.05; +137.82%, p < 0.05; Figure 5E2, respectively) compared with the control group, whereas in primary rat hepatocytes treated with PA + COM50 we observed increase in ASAH3 expression compared with the PA group (+43.58%, p < 0.05, Figure 5E2).

Effects of Coumestrol on the Expression of Enzymes Involved in the Sphingolipid Salv Pathway in Primary Rat Hepatocytes
We noticed considerable increase in the expression of SPHK1 after incubatio COM50 and PA + COM20 as well as PA + COM50 groups (+74.82%, p < 0.05; +206. < 0.05; +71.48%, p < 0.05, Figure 5A2, respectively) compared with the control. Fu more, the incubation with PA + COM20 resulted in increased expression of the S and SPHK2 compared with the PA group (+75.54%, p < 0.05, Figure 5A2; +91.74%, p Figure 5B2; respectively). In contrary, we observed notable decline in the express SPHK2 after incubation with PA + COM50 compared to the control (−25.71%, p Figure 5B2). Incubation with PA + COM20 as well as PA + COM50 resulted in signifi increased ASAH1 expression in relation to the PA group (+199.35%, p < 0.05; +204. 0.05; Figure 5C2, respectively). What is more, significant elevation of ASAH3 expr was detected in PA + COM20 as well as PA + COM50 groups +105.1%, p < 0.05; +13 p < 0.05; Figure 5E2, respectively) compared with the control group, whereas in pr The cells were incubated with presence or absence of PA (0.5 mM) or/and COM (20 or 50 µM), as it was described in Materials and Methods section. The protein expression was evaluated by the Western blot technique. The data are shown as the mean ± SD and are based on six independent determinations (n = 6); * p < 0.05 significant difference vs. control group; # p < 0.05 significant difference vs. palmitate group; C, control; COM20 or COM50, coumestrol 20 or 50 µM; PA, palmitic acid; PA + COM20 or PA + COM50, palmitic acid + coumestrol 20 or 50 µM.

Discussion
Insulin resistance constitutes a crucial factor in the pathogenesis of various metabolic disorders such as obesity, atherosclerosis, hypertension, and type 2 diabetes mellitus. Tak ing into account the rapidly growing number of patients suffering from these diseases worldwide, insulin resistance is considered to be a major health threat in well-developed countries [23]. The body of evidence suggests that phytoestrogens may be efficient in the prevention and supportive treatment of diseases connected with insulin resistance. Thus our study was focused on the assessment of phytoestrogen-coumestrol as a potentia agent in the treatment of hepatic insulin resistance via its influence on sphingolipid metabolism. To show a potential influence of COM on the insulin signaling pathway, we evaluated alterations in phosphorylation of two proteins substantially involved in insulin action, namely, Akt and GSK-3β. They seem to be the most important because phosphor ylation of Akt leads to phosphorylation and inactivation of GSK-3β, the enzyme that plays a pivotal role in the final step of glycogen synthesis as well as in the exacerbation of insulin resistance [24,25]. Ceramide, through stimulation of the different isoforms of the protein kinase C as well as by indirect suppression through induction of the protein phoshphatase 2, leads to the inhibition of the insulin-stimulated Akt phosphorylation [5,26]. Therefore an excess cellular accumulation of ceramide is associated with impaired glucose uptake and, subsequently, an exacerbation of the insulin resistance. After incubation of primary rat hepatocytes with PA and COM in both concentrations (20 μM and 50 μM), we observed a substantial increase of insulin-stimulated pGSK-3β/GSK-3β expression ratio in comparison with the group incubated only with PA and insulin. These data clearly show that COM intensified the phosphorylation of GSK-3β and, subsequently, inactivation o The data are shown as the mean ± SD and are based on six independent determinations (n = 6); * p < 0.05 significant difference vs. control group; # p < 0.05 significant difference vs. palmitate group; C, control; COM20 or COM50, coumestrol 20 or 50 µM; PA, palmitic acid; PA + COM20 or PA + COM50, palmitic acid + coumestrol 20 or 50 µM.

Discussion
Insulin resistance constitutes a crucial factor in the pathogenesis of various metabolic disorders such as obesity, atherosclerosis, hypertension, and type 2 diabetes mellitus. Taking into account the rapidly growing number of patients suffering from these diseases worldwide, insulin resistance is considered to be a major health threat in well-developed countries [23]. The body of evidence suggests that phytoestrogens may be efficient in the prevention and supportive treatment of diseases connected with insulin resistance. Thus, our study was focused on the assessment of phytoestrogen-coumestrol as a potential agent in the treatment of hepatic insulin resistance via its influence on sphingolipid metabolism. To show a potential influence of COM on the insulin signaling pathway, we evaluated alterations in phosphorylation of two proteins substantially involved in insulin action, namely, Akt and GSK-3β. They seem to be the most important because phosphorylation of Akt leads to phosphorylation and inactivation of GSK-3β, the enzyme that plays a pivotal role in the final step of glycogen synthesis as well as in the exacerbation of insulin resistance [24,25]. Ceramide, through stimulation of the different isoforms of the protein kinase C as well as by indirect suppression through induction of the protein phoshphatase 2, leads to the inhibition of the insulin-stimulated Akt phosphorylation [5,26]. Therefore, an excess cellular accumulation of ceramide is associated with impaired glucose uptake and, subsequently, an exacerbation of the insulin resistance. After incubation of primary rat hepatocytes with PA and COM in both concentrations (20 µM and 50 µM), we observed a substantial increase of insulin-stimulated pGSK-3β/GSK-3β expression ratio in comparison with the group incubated only with PA and insulin. These data clearly show that COM intensified the phosphorylation of GSK-3β and, subsequently, inactivation of this enzyme, which may be correlated with the beneficial anti-diabetic effect of used phytoestrogen [27]. We did not observe any significant changes in the expression ratio of pAkt/Akt in the groups treated with COM. It may be suspected that a longer incubation time with COM would be more effective in Akt phosphorylation because of the trend toward increase observed in the PA + COM20 + I group. Interestingly, Lin et al. revealed that GSK-3β activation, via its dephosphorylation, was induced by CER in the study conducted on mouse T hybridoma cell line 10I [28]. The incubation of primary rat hepatocytes with PA and COM in two different concentrations resulted in notably decreased cellular accumulation of CER and SFA. These results are in agreement with a study conducted by Vinayavekhin et al., who showed that treatment with another dietary phytoestrogen from plant C. comosa, namely, diarylheptanoid, decreased concentration of CER and, subsequently, reduced insulin resistance development in ovariectomized rats [29]. On the contrary, Charytoniuk et al. showed that the treatment with enterolactone, which also belongs to phytoestrogens, and PA simultaneously resulted in elevated concentration of CER in malignant hepatocytes (HepG2). However, this discrepancy might be related to the proapoptotic influence of enterolactone on the malignant HepG2 cells used in that experiment. Noteworthy, in the group incubated with COM20 alone, we reported a substantially increased concentration of CER, whereas in the study conducted by Charytoniuk et al. an increased content of SFA, not CER, after incubation with enterolactone alone was reported. This suggests that coumestrol, similarly to enterolactone, may exert various effects on the sphingolipid accumulation, depending on the presence or absence of palmitate excess and cell type [9]. As observed in our study, increased accumulation of CER may be a result of increased expression of proteins involved in FFA uptake including FABPpm, FATP, and FAT/CD36. Fatty acid transport proteins' expression highly correlates with augmented delivery of palmitate and other FFA, which constitutes an essential substrate to the de novo CER synthesis pathway. Bonen et al. indicated that reduced FAT/CD36 expression in CD36 knockout mice and, subsequently, diminished fatty acid transport was correlated with decreased CER accumulation and ameliorated insulin resistance occurrence in muscle cells [30]. The simultaneous incubation of primary rat hepatocytes with PA and COM in concentrations of 20 µM and 50 µM revealed notably decreased expression of both FATP5 and FAT/CD36. This might contribute to the reduced concentrations of CER observed in these groups, as a consequence of diminished delivery of the crucial substrates for the de novo synthesis of this compound. Furthermore, we noticed significantly decreased expression of the enzymes involved in the CER de novo synthesis pathway, namely, SPTLC1, SPTLC2, and CerS2/4/5/6 after treatment with combined PA and COM20. These results clearly showed that the attenuated accumulation of CER may occur through inhibition of the de novo synthesis pathway in lower COM concentration and decreased transport of PA into hepatocytes regardless of the concentration that subsequently increased insulin sensitivity. The salvage pathway, besides the de novo pathway and sphingomyelin hydrolysis, constitutes another important route of CER synthesis in the cell. Indeed, it is estimated that the sphingolipid salvage pathway may participate even in 90% of total sphingolipid biosynthesis [7]. Thus, to examine the possible involvement of this pathway, we assessed the cellular concentration of SFO and sphingosine-1-phosphate (S1P) as well as the expression of SPHK1 and SPHK2. Incubation of the cells in the presence of PA and COM20 led to notably increased expression of both sphingosine kinases (SPHK1 and SPHK2) and, as a consequence, increased accumulation of S1P in this group. Our findings are in agreement with the studies showing that lowered activity of sphingosine kinases resulted in decreased S1P concentration and impaired insulin signaling as well as the development of diabetes in [31,32]. In turn, the increased activity of these enzymes and increased S1P levels significantly ameliorated insulin sensitivity [32]. The concentration of SFO was increased in the cells incubated with PA and COM but only in a lower concentration. These changes were reflected in the increased expression of ceramidases, which transform CER into SFO. Indeed, the expression of acid ceramidase and alkaline ceramidase was significantly elevated in primary rat hepatocytes incubated with combined PA and COM simultaneously. Thus, we suspect that COM intensified S1P synthesis from SFO, a part of the salvage pathway that was intensified after treatment of primary rat hepatocytes with COM. S1P is widely known for its anti-inflammatory effect and plays a pivotal role in the function, regeneration, and survival of the cell [33]. COM may augment the synthesis of S1P in the hepatocytes, which may constitute the protective mechanism against inflammation and progression of the hepatic insulin resistance that may occur at the beginning of the steatosis development. However, some studies indicated that increased accumulation of S1P in the liver constitutes a factor of hepatic steatosis progression and development to fibrosis [34,35]. Thus, the role of S1P in the pathogenesis of hepatic insulin resistance, diabetes, and obesity in humans is still unclear and should be fulfilled by further studies. The de novo pathway and salvage pathway share the same ceramide synthases, which are crucial for generation of either dihydroceramide or ceramide. This may suggest that two, potentially very different, pathways remain linked. Thus, the cellular response on the stimulus may depend on which pathway will be promoted in the overall ceramide biosynthesis [7]. Therefore, as described above, changes in the expression of the CerS2/4/5/6 after simultaneous treatment with PA and COM may not only diminish the ceramide synthesis through de novo pathway but also by regulation of the salvage pathway. These results suggest that coumestrol led to increased synthesis of the beneficial S1P by elevating the expression of the SPHK1 and SPHK2. Moreover, it also significantly reduced the ceramide formation by decreasing the expression of CerS2/4/5/6 as these enzymes are mainly involved in the formation of ceramide from palmitate. Last but not least, Raichur et al. showed that the increased concentration of ceramides was markedly associated with impaired lipid β-oxidation and subsequently caused hepatic steatosis and insulin resistance development in CerS 2 −/+ mice [36]. In our study, we assessed the possible influence of COM on β-oxidation by the incubation of primary rat hepatocytes with combined PA and COM50, which revealed that the expression of β-HADH was notably increased compared to the PA group, indicating that β-oxidation was enhanced. These results suggest that COM via intensified β-oxidation of FA diminishes FA availability to the de novo CER synthesis pathway. Thus, it may constitute a putative mechanism of the protection of the cell against the enhanced accumulation of CER and, in consequence, ameliorated insulin sensitivity.

Conclusions
In conclusion, our study demonstrated that the treatment of primary rat hepatocytes with coumestrol during increased availability of PA leads to the amelioration of hepatic insulin resistance. The possible mechanisms of the outcomes seem to be complex and involve alterations in the sphingolipid metabolism pathway components, including inhibiting the ceramide de novo synthesis pathway as well as augmenting the sphingolipid salvage pathway. What is more, COM significantly decreased the expression of proteins responsible for FFA uptake and enhanced the mitochondrial FA β-oxidation, which also affected the rate of sphingolipid deposition in hepatocytes. There is no doubt that the popularity of plant-based diets is increasing worldwide, being a strong alternative to Western eating patterns. Thus, we believe that our results may contribute, at least partially, to the discovery of novel preventive and therapeutic methods for metabolic disorders, especially obesity and T2DM.