A Rare Olive Compound Oleacein Improves Lipid and Glucose Metabolism, and Inflammatory Functions: A Comprehensive Whole-Genome Transcriptomics Analysis in Adipocytes Differentiated from Healthy and Diabetic Adipose Stem Cells

Oleacein (OLE), a rare natural compound found in unfiltered extra virgin olive oil, has been shown to have anti-inflammatory and anti-obesity properties. However, little is known regarding the mechanisms by which OLE influences metabolic processes linked to disease targets, particularly in the context of lipid metabolism. In the present study, we conducted whole-genome DNA microarray analyses in adipocytes differentiated from human adipose-derived stem cells (hASCs) and diabetic hASCs (d-hASCs) to examine the effects of OLE on modulating metabolic pathways. We found that OLE significantly inhibited lipid formation in adipocytes differentiated from both sources. In addition, microarray analysis demonstrated that OLE treatment could significantly downregulate lipid-metabolism-related genes and modulate glucose metabolism in both adipocyte groups. Transcription factor enrichment and protein–protein interaction (PPI) analyses identified potential regulatory gene targets. We also found that OLE treatment enhanced the anti-inflammatory properties in adipocytes. Our study findings suggest that OLE exhibits potential benefits in improving lipid and glucose metabolism, thus holding promise for its application in the management of metabolic disorders.


Introduction
Adipocytes, also known as lipocytes and fat cells, originate from mesenchymal stem cells and serve the purpose of storing energy in the form of fat [1]. These cells play a crucial role in responding to the overall energy state of an organism by reacting to anabolic hormones such as insulin and catabolic hormones such as glucagon, which in turn affects the processes of lipogenesis and lipolysis [2]. Apart from its role in the systemic energy state, adipocytes also release multiple adipokines that possess crucial regulatory functions in various physiological processes, such as feeding behavior, energy expenditure, and glucose homeostasis, among others [3,4]. The maintenance of metabolic activity in healthy adipocytes is dependent on processes such as angiogenesis and mitochondrial biogenesis, as described in previous literature [5,6]. Conversely, adipocyte dysfunction can occur as a result of chronic hypoxia, inflammation, and mitochondrial dysfunction, leading to a range of conditions, including insulin resistance, hypertension, and atherosclerosis [7,8]. This imbalance is thought to be a major contributor to the development of type 2 diabetes [9]. on the underlying regulatory mechanism. Hence, our study aims to evaluate the potential effects of OLE on lipid accumulation in adipocytes. We investigated the biological effects of OLE by conducting whole-genome DNA microarray analysis on adipocytes differentiated from healthy and diabetic hASCs. Furthermore, we explored the potential preventive impact of OLE on adipocytes differentiated from diabetic donors' adipose stem cells.

Effect of OLE on hASCs Viability and Proliferation
First, we examined if hASCs retain their capacity to proliferate in the presence of OLE. hASCs were incubated with 0, 1,5,10,20,40,80, and 100 µM of OLE, and an MTT assay was performed after 24 h. The viability of the treated cells relative to the vehicle control was assessed as a percentage of optical density readings. We found that OLE did not show any cytotoxicity [26]. There was no significant difference in OLE-treated hASCs compared to nontreated control cells at concentrations ranging from 0 to 40 µM, with only minimal changes in % activity. However, OLE was found to significantly stimulate cell proliferation at the concentration of 80 µM and more (Supplementary Figure S1).

Effect of OLE on Lipid Accumulation during Adipocytic Differentiation of hASCs
OLE treatments were added at the concentration of 1, 10, 20, and 40 µM. Then, we used relative fluorescence unit (RFU) to quantify the lipid accumulation rate in hASCs ( Figure 1A). Compared to undifferentiated cells, we observed significant lipid accumulation in all adipogenic-induced groups, both OLE-treated and nontreated. However, OLE-treated cells had significantly lower RFU compared to nontreated cells at concentrations of 20 µM and 40 µM. Induced hASCs treated with 20 and 40 µM OLE exhibited almost identical intensity values. Therefore, we decided on the OLE concentration of 20 µM for subsequent gene expression analysis.

Effects of OLE on Gene Expression Profiling during hASC Adipogenic Differentiation
We performed whole-genome DNA microarray analysis to investigate the effects of OLE on gene expression profiling during hASC adipocytegenic differentiation. Genes with a 2-fold change (FC) and a significance level of p ≤ 0.05 were identified as differentially expressed genes (DEGs). The volcano plots display the FC values of all gene probe sets in the adipocyte-differentiated and undifferentiated group (diff vs. undiff), as well as the OLEtreated and adipocyte-differentiated and undifferentiated groups (diff + OLE vs. undiff) (Figure 2A,B). A total of 19,668 genes were identified in both comparison groups. In the hASCs adipocyte differentiation group, 2883 genes were differentially expressed compared to the undifferentiated group, of which 1365 DEGs were upregulated, and 1518 DEGs were downregulated. In the OLE-treated group, of 2782 DEGs, 1373 were upregulated, and 1409 were downregulated. Bar graphs show the distribution of FC values of up and downregulated DEGs ( Figure 2C).
Moreover, we assessed the degree of overlap in the number of all DEGs in diff vs. un-diff and diff + OLE vs. undiff ( Figure 2D). We identified 508 and 407 unique genes in diff vs. undiff and diff + OLE vs. undiff, respectively, and 2375 genes commonly regulated in both conditions. We checked the significantly enriched gene ontology biological processes (GOBP) by the common genes. We have identified some particular genes that were significant and display basic regulation, such as the regulation of proteolysis (GO:0030162), protein phosphorylation (GO:0006468), DNA metabolic process (GO:0006259), and the regulation of kinase activity (GO:0043549) in metabolic processes. Significantly enriched cellular processes included microtubule cytoskeleton organization (GO:0000226), cell population proliferation (GO:0008283), chromosome organization (GO:0051276), and mitotic cell cycle (GO:0000278) (Supplementary Figure S2). Nile red fluorescence intensity quantification. Lipid content was quantified by relative fluorescence unit (RFU) of Nile red staining. Data were shown as mean ± SEM (n = 3/dose). ns, non-significant; * p < 0.05 statistical difference was calculated using one-way ANOVA followed by Tukey's post hoc test.

Effects of OLE on Gene Expression Profiling during hASC Adipogenic Differentiation
We performed whole-genome DNA microarray analysis to investigate the effects of OLE on gene expression profiling during hASC adipocytegenic differentiation. Genes with a 2-fold change (FC) and a significance level of p ≤ 0.05 were identified as differentially expressed genes (DEGs). The volcano plots display the FC values of all gene probe sets in the adipocyte-differentiated and undifferentiated group (diff vs. undiff), as well as the OLE-treated and adipocyte-differentiated and undifferentiated groups (diff + OLE vs. undiff) (Figure 2A,B). A total of 19,668 genes were identified in both comparison groups. In the hASCs adipocyte differentiation group, 2883 genes were differentially expressed compared to the undifferentiated group, of which 1365 DEGs were upregulated, and 1518 Figure 1. Effect of OLE treatment on cellular lipid accumulation in hASCs induced to adipocytic differentiation. Cells were cultured in an adipocytic differentiation medium for 7 days and then incubated with 20 µM of OLE for 14 days. Intracellular lipid levels were evaluated quantitatively and qualitatively using Nile red staining. (A) Representative fluorescence image. The scale bar indicates 200 µm. (B) Nile red fluorescence intensity quantification. Lipid content was quantified by relative fluorescence unit (RFU) of Nile red staining. Data were shown as mean ± SEM (n = 3/dose). ns, non-significant; * p < 0.05 statistical difference was calculated using one-way ANOVA followed by Tukey's post hoc test.
Next, we checked the enriched GOBP terms with the 407 unique DEGs in diff + OLE vs. undiff to identify the OLE exclusive function. The parent terms, including immune system process (GO:0002376), metabolic process (GO:0008152), and signaling pathway (GO:0023052), were significantly enriched. For example, the positive regulation of immune response (GO:0050778) and the regulation of the T cell receptor signaling pathway (GO:0050856) were the significantly enriched immune system processes. The regulation of apoptotic signaling pathway (GO:2001233), protein kinase B signaling (GO:0043491), MAPK cascade (GO:0000165), and the regulation of Notch signaling pathway (GO:0008593) were the significantly enriched signaling pathways.
In addition, several enriched terms related to metabolic processes suggest that bioenergetics were more activated by the OLE treatment. These enriched terms were specifically associated with protein glycation processes, including glycoprotein metabolic processes (GO:0009100) and protein glycosylation (GO:0006486), as well as lipidation processes such as lipoprotein metabolic processes (GO:0042157) and lipid metabolism processes such as fatty acid metabolic processes (GO:0006631).

Gene Ontology Analysis of OLE-Treated hASCs Induced to Adipogenic Differentiation
To clarify the molecular changes in adipocytes after adding OLE, we compared adipocytes with or without treatment (diff + OLE vs. diff). A total of 19,665 distinctive probe sets were recognized. The volcano plot is based on FC value with the criteria of 1.1-fold change ( Figure 3A), and the bar graph illustrates the distribution of FC ( Figure 3B). The top 10 up and downregulated DEGs were listed in detail (Tables 1 and 2). A total of 848 genes were differentially expressed, with 381 upregulated DEGs and 467 downregulated DEGs. We identified the main GOBP terms with more than three qualifying genes. The GO terms for immune system processes (GO:0002376) and metabolic processes (GO:0008152) were examined.   . The x-axis displays the gene radio, and the y-axis represents the enrichment BPs, the size of bubbles pertains to the count/number of genes within the pathway, and the colors refer to the significance −log10 (p-value). Among these terms, the downregulated DEGs showed significant enrichment in terms related to the regulation of B cell proliferation and lymphocyte migration and activation. Additionally, the fatty-acyl-CoA biosynthetic process exhibited the highest enrichment score, followed by the triglyceride metabolic process and polysaccharide metabolic process. Moreover, terms related to the fatty acid metabolic process, cellular lipid catabolic process, and ATP metabolic process were all downregulated ( Figure 3C). The carbohydrate biosynthetic process presented the most significant value, with the largest number of genes also enriched in the downregulated biological process (BP) results. In DEGs that were upregulated and enriched in BPs, several terms were observed, including T cell proliferation, megakaryocyte, and myeloid differentiation. Additionally, the biological processes of mitochondrial electron transport and glycolipid catabolic metabolism were identified. Notably, carbohydrate derivative formation again exhibited the highest significance and the largest number of genes, suggesting that OLE may impact carbohydrate metabolic processes.

OLE Exerted Anti-Inflammatory Effect and Regulated Lipid Metabolism in Adipocyte
Next, we conducted pathway enrichment analysis to obtain a comprehensive and detailed pathway term list by using the Kyoto Encyclopedia of Genes and Genomes (KEGG) and BioPlanet_2019 database of diff + OLE vs. diff. The correlation among the enrichment score of various pathways in the KEGG database is shown in Figure 4A. It was observed that upregulated DEGs have a significant impact on nonalcoholic fatty liver disease and transcriptional misregulation in cancer pathways. The pathway of glycosphingolipid biosynthesis exhibited the highest enrichment score in the metabolism function. Furthermore, the upregulated DEGs showed enhancement of the phosphatidylinositol signaling system and P13K-Akt signaling pathway. The downregulated DEGs had a significant impact on various pathways, including T2D, which exhibited the highest enrichment score, followed by chemical carcinogenesis, lipid metabolism, and atherosclerosis in human diseases. In the metabolic function, the downregulated DEGs had a noticeable impact on pathways such as mucin-type O-glycan biosynthesis, fatty acid, pyruvate, cholesterol metabolism, and oxidative phosphorylation. Additionally, several signaling pathways, including adipocytokine signaling, PPAR signaling, insulin signaling, sphingolipid signaling, and the JAK-STAT signaling pathway, were regulated by the downregulated DEGs. The inhibition of certain pathways related to lipid metabolism by OLE indicates its potential beneficial effects in the treatment of type 2 diabetes (T2D).
To identify functionally clustered pathways, we used Uniform Manifold Approximation and Projection (UMAP) dimension reduction technique from the BioPlanet2019 database ( Figure 5B). In the given analysis, the terms with similar gene sets are placed closer together. It was observed that the pattern of DEGs in both upregulated and downregulated clusters was the same but exhibited different significant terms. The downregulated DEGs exhibited the highest significant functions, which included chylomicron-mediated lipid transport and fatty acid biosynthesis represented by the color 'red'. The downregulated DEGs showed highly enhanced pathways related to the proteasome complex, apoptosis regulation, and the activation of NF-kB in B cells, but these terms were not enriched in the upregulated DEGs. Additionally, the terms that were more enriched in the 'orange' cluster in the downregulated DEGs included interleukin-9 regulation, cytokine-cytokine receptor interaction, inflammasomes, NOD-LIKE receptor proteins, and apoptosis. However, in the upregulated DEGs clusters, the SMAD2/3-SMAD4 transcriptional activity showed significant enrichment, and the 'gray' points represent ERBB4 signaling events, interleukin-6 signaling, and CBL signaling regulation, which was also enriched in upregulated DEGs. The downregulation of DEGs resulting from the regulation of lipid and NF-κB transcription factor activity suggests that OLE not only regulates lipids but also exhibits anti-inflammatory effects on hASCs. Along with the upregulation of the CBL signaling pathway, the evidence indicates that OLE may also downregulate insulin resistance [27]. For each gene set term, TF-IDF values were calculated and then reduced in dimensionality using the Uniform Manifold Approximation and Projection (UMAP) technique. The resulting terms were plotted using the first two UMAP dimensions on the scatter graph. The Leiden algorithm was applied to the TF-IDF values, leading to the identification of clusters. Each term was then assigned a color that corresponds to its respective cluster. Most significant terms in selected clusters are shown in the figure.

Effects of OLE on Gene Expression Profiling in Adipocytes Differentiated from Diabetic-hASCs
OLE treatment at 1 µM concentration for 14 days on d-hASCs adipocyte differentiation showed significantly smaller lipid droplets ( Figure 5A). To further investigate the potential impact of OLE on adipogenesis in d-hASCs, we performed GO analysis to identify enriched BPs. Probes that exhibited an absolute value FC of 1.18 or higher and were sorted based on the top 500 DEGs in the OLE-treated group compared to the adipocyte differentiation group were selected. The distribution of all DEGs is illustrated in Figure 5B.
In terms of enriched BP identified through the GO analysis, the downregulated DEGs were notably enriched in immune system-related terms ( Figure 5C). These included the inflammatory response to an antigenic stimulus, B cell receptor signaling pathway, lymphocyte migration, leukocyte activation and migration, immune effector process, and T cell activation, among others. Additionally, in terms of metabolic processes, the downreg-

Effects of OLE on Gene Expression Profiling in Adipocytes Differentiated from Diabetic-hASCs
OLE treatment at 1 µM concentration for 14 days on d-hASCs adipocyte differentiation showed significantly smaller lipid droplets ( Figure 5A). To further investigate the potential impact of OLE on adipogenesis in d-hASCs, we performed GO analysis to identify enriched BPs. Probes that exhibited an absolute value FC of 1.18 or higher and were sorted based on the top 500 DEGs in the OLE-treated group compared to the adipocyte differentiation group were selected. The distribution of all DEGs is illustrated in Figure 5B.
In terms of enriched BP identified through the GO analysis, the downregulated DEGs were notably enriched in immune system-related terms ( Figure 5C). These included the inflammatory response to an antigenic stimulus, B cell receptor signaling pathway, lymphocyte migration, leukocyte activation and migration, immune effector process, and T cell activation, among others. Additionally, in terms of metabolic processes, the downregulated DEGs were involved in o-glycan processing, the regulation of lipid metabolic process, and carbohydrate metabolic process. The upregulated DEGs were found to be associated with several enriched BPs, including somatic diversification of immune receptors, immune system development, T cell differentiation, as well as leukocyte and lymphocyte differentiation. Moreover, in terms of metabolic processes, the upregulated DEGs were related to protein demannosylation, mRNA and DNA processing, and carbohydrate derivative biosynthetic processes.
The results of the study indicate that the regulation of lipid metabolism and antiinflammatory effects observed in healthy adipocytes were also present in adipocytes obtained from hASCs. The use of OLE resulted in improved immune function-related terms, as evidenced by the downregulation of DEGs in adipocytes from d-hASCs. These findings suggest that OLE may have a potential regulatory role in these metabolic pathways.

Differential Effect of OLE on Adipocytes Differentiated from Healthy and Diabetic hASCs
To find the target anti-inflammatory and metabolic functions of OLE, we performed a comparative analysis of adipocytes from healthy and diabetic conditions types. We overlapped the DEGs in adipocytes with OLE compared to nontreatment with OLE in hASCs and d-hASCs ( Figure 6A). The adipocytes differentiated from d-hASCs exhibited 1038 unique genes, while the normal adipose cells showed 813 unique genes, with an overlap of 36 genes between the two conditions. These genes exhibited distinct and clearly defined biological states, displaying consistent expression patterns in adipogenesis, other types of differentiation (COAGULATION, MYC_TARGETS_V1), cell cycle regulation (E2F_TARGETS, G2M_CHECKPOINT, MITOTIC_SPINDLE), innate immune system function (COMPLEMENT), and glycolysis. Subsequently, we performed a transcription factor (TF) enrichment analysis using the TRRUST database ( Figure 6B). In the top 20 TFs, half of the TFs differed, while four TFs, SP1(Specificity Protein 1), STAT3 (Signal Transducer and Activator of Transcription 3), NFκB1 (Nuclear Factor Kappa-Light-Chain-Enhancer of Activated B Cells 1) and RELA (v-Rel Avian Reticuloendotheliosis Viral Oncogene Homolog A) were same expressed with different significance in both groups. This implies that the four TFs may be the key to regulating genes related to health and diabetes adipocytes, which are related to anti-inflammation or lipogenesis.
Shared BPs, three from the immune system and six from metabolic processes, were identified ( Figure 6C). The results indicate that OLE has a higher significance value in adipocytes differentiated from d-hASCs in the immune system, particularly in the innate immune response, acute inflammatory response, and regulation of immune response when compared to the hASCs group. In terms of metabolic processes, the highest value associated with carbohydrate metabolic processes was observed in the hASC group, while a smaller value was observed in the d-hASC group. This trend was also observed in protein glycosylation, O-glycan processing, and RNA biosynthetic process. Glycoprotein metabolic process and histone modification showed similar values in both groups. The confirmation of these points was supported by the intensity value ( Figure 6D). In the adipocyte-induced in d-hASCs, a larger number of genes associated with inflammation were downregulated in comparison with hASCs. This shows a stronger anti-inflammatory function of OLE under disease conditions. It is noteworthy that the terms associated with metabolic processes in adipocytes from hASCs were more prominent than those in the d-hASCs group, but there were more upregulated genes in d-hASCs than in hASCs. This indicates that OLE is more effective in upregulating the metabolic processes in adipocytes from d-hASCs, particularly those related to carbohydrate and glycoprotein metabolism.

PPI Network Analysis of Adipocytes Differentiated from hASCs and d-hASCs
Next, the protein-protein interaction (PPI) network was curated to elucidate the links between expressed genes to find the connections in DEGs with potential proteins. The PPI analysis revealed 295 seeds with 1646 interacting nodes in adipocyte-hASC ( Figure 7A) and 429 seeds with 923 nodes in adipocytes in d-hASCs groups ( Figure 7B). The detailed top 20 seeds in each adipocyte with degree and expression value were listed ( Figure 7C). There were five seeds shown in two adipocyte groups. The protein: ubiquitin C (UBC) was the top node with the highest degree in the two groups (degree = 208 in hASC, degree = 325 in d-hASC). We also provided the top 10 seeds list that appeared in the two adipocyte networks (Tables 3 and 4), along with detailed information about each protein, in-

PPI Network Analysis of Adipocytes Differentiated from hASCs and d-hASCs
Next, the protein-protein interaction (PPI) network was curated to elucidate the links between expressed genes to find the connections in DEGs with potential proteins. The PPI analysis revealed 295 seeds with 1646 interacting nodes in adipocyte-hASC ( Figure 7A) and 429 seeds with 923 nodes in adipocytes in d-hASCs groups ( Figure 7B). The detailed top 20 seeds in each adipocyte with degree and expression value were listed ( Figure 7C). There were five seeds shown in two adipocyte groups. The protein: ubiquitin C (UBC) was the top node with the highest degree in the two groups (degree = 208 in hASC, degree = 325 in d-hASC). We also provided the top 10 seeds list that appeared in the two adipocyte networks (Tables 3 and 4), along with detailed information about each protein, including an enriched functional network. In both groups, UBC appears to be closely associated with regulating the activity of the NF-κB TF, protein phosphorylation, and the MAPK cascade. Additionally, hepatocyte nuclear factor 4 alpha (HFH4A), which is associated with lipid homeostasis and response to carbohydrate stimulus, was upregulated in the d-hASCs group. On the other hand, Amyloid Beta Precursor Protein (APP) was found to be linked to immune response. The majority of the hub genes identified were involved in RNA and DNA processing. This suggests that these genes may have a common role in the regulation of adipocyte function and metabolism, regardless of the presence of diabetes after OLE treatment. Additionally, this indicates that these genes could have important implications in the treatment of adipogenesis. Activation of JUN kinase activity, Superoxide metabolic process, regulation of cytokine secretion, regulation of T cell proliferation, regulation of cell-cell adhesion, Activation of MAPK activity, Lipid transport, Fatty acid metabolic process, Activation of immune response ciated with lipid homeostasis and response to carbohydrate stimulus, was upregulated in the d-hASCs group. On the other hand, Amyloid Beta Precursor Protein (APP) was found to be linked to immune response. The majority of the hub genes identified were involved in RNA and DNA processing. This suggests that these genes may have a common role in the regulation of adipocyte function and metabolism, regardless of the presence of diabetes after OLE treatment. Additionally, this indicates that these genes could have important implications in the treatment of adipogenesis.

Discussion
In the present study, we have performed an integrated transcriptome analysis of OLEtreated adipocytes differentiated from hASCs and d-hASCs. We explored the prospects of OLE in anti-inflammatory functions and its effect on glucose and lipid metabolic functions.
The abnormal expansion and accumulation of adipocytes, resulting from an excessive differentiation of adipose stem cells towards an adipogenic lineage, have been linked to various health conditions such as obesity, insulin resistance, inflammation, metabolic disorders, and cardiovascular diseases [28,29]. Searching for compounds that are more efficient and effective in preventing adipocyte differentiation and promoting new adipocyte differentiation without depleting the reserve of stem cells has been an ongoing endeavor [30]. In this study, we explored adipocytes differentiated from both healthy and diabetic adipose-derived stem cells to investigate the impact of OLE on metabolic processes.
In hASCs, OLE exhibited normal proliferative ability in adipocytes, while the trend of lipid accumulation capacity decreased in adipocytes. Upon analyzing global microarray data under undifferentiated healthy adipose stem cell conditions, the unique genes showed significant enrichment in terms related to protein modification, such as protein phosphory-lation, glycosylation, and lipidation. These modifications suggest that post-translational modifications occur in adipocytes upon treatment with OLE and regulate important cellular processes such as glucose uptake, lipid metabolism, adipocyte differentiation, insulin signaling, adipogenesis, lipolysis, and adipokine secretion [31,32]. We also found that the OLE downregulated fatty-acyl-CoA biosynthetic, triglyceride metabolic, and cellular lipid catabolic processes in adipocytes in BPs. Moreover, the findings from KEGG analysis indicated a downregulation of fatty acid metabolism and PPAR signaling pathway; meanwhile, the BioPlanet database highlighted the most significant cluster related to lipid transport. These results were supported by the research conducted by the Lepore team, who reported that OLE exhibited a protective effect against high-fat diet-induced adiposity in mice by reducing the levels of peroxisome proliferator-activated receptor gamma (PPARγ) and fatty acid synthase proteins [24]. All these results suggested that PPAR signaling may be a particular interest pathway, as the related genes directly impact lipid homeostasis.
Interestingly, T2D was the most highly enriched pathway in human diseases by downregulated DEGs, whereas nonalcoholic fatty liver disease was also prominently enriched, as indicated by the upregulated DEGs. We also observed a decrease in the expression of the adipokine signaling pathway in adipocytes differentiated from hASCs, which could indicate the presence of insulin resistance [33]. This case was also supported by an upregulation of the P13K-AKT signaling pathway in the KEGG result. Previous research has shown that the P13K-AKT pathway can downregulate insulin resistance, leading to improved glucose uptake, control of gluconeogenesis, and enhanced glycogen storage in a db/db mice model [34]. Despite this, even in the presence of insulin resistance, OLE treatment has the potential to preserve normal glucose metabolism in adipocytes through alternative pathways or transcription factors.
We also investigated the effects of OLE on adipocytes differentiated from d-hASCs. Treatment with a low concentration of OLE led to a reduction in adipocyte size and downregulation of lipid metabolic processes. These findings suggested that OLE has the potential to modify cellular processes related to adipocyte differentiation, which could have implications for diabetes treatment. This modulation of lipid metabolism is further supported by prior research carried out by our team employing adipocytes from d-hASCs on compounds with squalene derivatives [14]. Apart from these, more terms related to immune response were shown in the d-hASCs group. This is because, after adipocyte differentiation, adipose tissue inflammation attracts immune cells, which triggers the activation of resident macrophages and exacerbates insulin resistance and systemic inflammation [35]. Adipocytes are often inflamed due to pro-inflammatory cytokines in obese individuals, which can impact adipocyte function [36]. T cell proliferation was found to be positively regulated, while the migration and proliferation of other immune cells, such as B cells, lymphocytes, and leukocytes, were found to be downregulated, suggesting the presence of an immunosuppressive environment that could help prevent excessive inflammation in adipose tissue via OLE administration.
This phenomenon was also shown in the healthy adipocyte group. Not only was there a downregulation of T cells, B cells, lymphocytes, and leukocytes, but also the "NF-B transcription factor activity" term was downregulated in adipocytes differentiated from hASCs. Moreover, the DEGs regulated by NFκB1 and RELA were shown in two adipocytes that belong to the NF-κB family and play an important role in inflammation and metabolic disease [37]. Adipose tissue is known to be associated with chronic low-grade inflammation [38]. Previous research on human cells demonstrated that OLE amplifies antiinflammatory activity in human macrophages via the expression of the CD163 receptor [39]. Additionally, in adipocytes with Simpson-Golabi-Behmel syndrome, OLE mitigated the expression of inflammation-related genes, resulting in decreased activation of NF-κB, which is consistent with our study [22,40]. OLE's ability to rapidly activate signaling molecules through pharmacological principles may regulate the nutritional modulation of inflammatory diseases [41]. All these findings revealed that certain inflammation-related functions could be regulated in adipocytes by OLE.
For finding the potential target of the OLE effect in adipocytes from healthy and diabetic stem cells, DEGs related to TFs and their' PPI networks were checked. Various genes involved in the transcriptional regulation of adipogenesis differentiation were induced by OLE treatment. Our TF analysis revealed that SP1 was the most abundant regulator of DEGs, displaying the highest significance in both groups. SP1 is known to activate the transcription of the glucose transporter Glut-1, which plays a crucial role in glucose uptake in various cell types, including adipocytes [42]. Additionally, emerging evidence suggests that both SP1 and STAT3 can suppress the expression of pivotal genes implicated in adipocyte differentiation, including PPARγ, C/EBPα, and FABP4, thereby potentially restoring insulin sensitivity and mitigating obesity [43,44]. Additionally, SUMOylation (SUMO) was identified as a regulator of PPARγ activity in our PPI analysis, offering a potential strategy for developing safer drugs for treating type 2 diabetes by inhibiting insulin-sensitizing activity without affecting adiposity in mice [45]. ELAVL1 binds to a cyclic nucleotidedependent stabilizing domain in the 3 untranslated region of Na(+)/glucose cotransporter mRNA. The drugs inferred for the ELAVL1 gene are reported to affect glucocorticoids, anti-inflammatory agents, small molecules, and anti-inflammatory glucocorticoids [46]. Based on these findings, it is possible to hypothesize that although the upregulation of the P13K-AKT pathway in hASCs may lead to insulin resistance, the identified targets have the potential to restore normal glucose metabolism through alternative pathways. This suggests that these targets could help mitigate insulin resistance.
Furthermore, while the effects of OLE on lipid and glucose metabolism, as well as immune response, remained consistent, our results revealed variations between the two sources of differentiated adipocytes. STAT3 participates in various cellular processes and is involved in the regulation of carbohydrate and lipid metabolism [47]. Despite the fact that STAT3 is regulated in both groups, its expression level is higher in the d-hASC group due to the activation of STAT3 by pro-inflammatory cytokines, increasing its levels, as has been observed in significantly upregulated T2D models [48]. Additionally, we found that PPARγ was only shown to control DEGs in the d-hASCs group, which are concerned with T2D risk through the involvement in adipocyte differentiation and energy homeostasis in a previous study [49]. Moreover, the exclusive presence of HNF4A expression in adipocytes derived from d-hASCs, but not in the hASCs group, highlights its strong association with the "lipid metabolism" in one human research related to diabetes [50]. This gene also holds potential as a candidate gene for type 2 diabetes in individuals with severe obesity [51]. The unique expression patterns observed suggest that these genes may serve as potential targets for regulating lipid metabolism in adipocytes derived from diabetic adipose stem cells, requiring further investigation.
According to structure characteristics among secoiridoid derivatives in EVOO, OLE, derived from oleuropein, possesses a secoiridoid aglycone structure similar to that of oleuropein [52]. Previous studies reported that oleuropein aglycone exhibited diverse pharmacological effects, including anti-hyperglycemic and lipid-lowering properties [53], as well as the ability to enhance UCP1 expression in brown adipose tissue, thereby activating β-adrenergic signaling in obese rats induced by a high-fat diet [54]. In our study, we identified alterations in terms that belong to lipid and glucose metabolism, as well as ATP biosynthetic processes, in adipocytes derived from healthy adipose stem cells. These findings suggest a potential involvement of the aglycone structure of OLE in the regulation metabolism on fat accumulation and glucose. Additionally, it is noteworthy that oleuropein aglycone exhibited notable stability under acidic conditions simulating gastric digestion, with limited hydrolysis observed [55,56]. Similarly, OLE demonstrated favorable stability in a gastric digestion model [56,57]. Here, we utilized OLE by employing a one-step synthesis method from oleuropein, which demonstrated higher efficiency compared with the more complex over 10-step synthesis method [58] and the use of H-mont did not result in any decrease in the yield of OLE [20]. This suggests that the stable and highly pure structure of OLE may have a significant role in regulating adipocyte fat and glucose metabolism.
It is important to note that our findings are derived from whole-genome transcriptomelevel microarray data, offering preliminary insights for future investigations, particularly concerning adipocytes derived from diverse sources. Most findings are consistent with previous studies regarding adipocyte metabolic homeostasis and immune response. Further in vitro studies are necessary to investigate the potential targets and preventive effects of OLE, as well as to elucidate the structural characteristics of OLE in relation to its functional properties. Meanwhile, in vivo studies focusing on diabetic model treatment are required to confirm the potential functionalities of OLE observed in the present study.
Based on the latest research, the link to metabolic processes in OLE, such as glucose and lipid metabolism, as well as chroma-tin-modifying enzymatic activities, through computational de-orphanization (a structure-based tool) has been experimentally validated [59]. In our study, it appears that OLE promotes adipocyte metabolism in both hASCs and d-hASCs. Additionally, OLE has been shown to regulate lipid and glucose metabolic functions and enhance anti-inflammatory properties during adipocyte differentiation. This is achieved by balancing glucose metabolism through glycosides and glycolipids (Figure 8). OLE has the potential to aid in the remedy of insulin resistance and prevent the onset of early T2D. Overall, our findings provide a promising foundation for the development of a novel class of therapeutics derived from natural compounds.  Figure 8. Summary of biological events regulated by OLE in adipocytes differentiated from hASCs.

Chemicals
OLE was chemically synthesized from oleuropein using proton-exchanged montmorillonite as a catalyst and used after column separation and freeze-drying [20]. Briefly, oleuropein powder (1.53 g, purity = 88.0%) was treated in the presence of proton-exchanged montmorillonite (3.06 g) in DMSO (10.0 mL) containing 7.16 mmol of H 2 O at 150 • C for 3 h without stirring. After removing the catalyst by filtration, the crude products were extracted with AcOEt/H 2 O. OLE was isolated via column chromatography at a yield of 75% (0.598 g).

Cells Culture and Techniques
hASCs were purchased from ATCC ® (SCRC-4000TM, Manassas, VA, USA). These cells were isolated from the adipose tissue of a Caucasian female. Cell culture was performed using a modified process based on the previous description [60,61]. Briefly, the cells were cultured in a DMEM/F12 (11320033, Thermo Fisher Scientific, Waltham, MA, USA) based medium. This medium consisted of 10% (v/v) fetal bovine serum (FBS), 1% (v/v) penicillin/streptomycin, 1% GlutaMax (35050061, Thermo Fisher Scientific), and 1‰ (v/v) 2-mercaptoethanol (21985023, Thermo Fisher Scientific, USA), under conditions at 37 • C in a 5% CO 2 environment. The cells were seeded at a density of 5000 viable cells per cm 2 and passaged after reaching 80% confluence as recommended by the provider. d-hASCs were purchased from Lonza (PT-5008, Walkersville, MD, USA). Detailed culture conditions have been explained in our previous study [14]. Briefly, the cells were cultured at 37 • C in a 5% CO 2 humidified incubator using ASC growth medium sourced from the Lonza company. The ASC growth medium comprised ASC basal medium supplemented with 10% FBS, 1% L-glutamine, and gentamicin-amphotericin B (GA-1000), following the manufacturer's instructions.

Adipocyte Differentiation of hASCs and d-hASCs
White adipocyte differentiation of hASCs was performed using the DMEM/F12-based medium as previously described according to the regularly used protocol with slight modifications to ensure successful adipocyte induction [62]. Adipogenic differentiation was initiated on the second day after cell confluence, followed by a 7-day induction period and a 14-day maintenance period. The induction cocktail contained 100 nM insulin (093-06351, Wako, Japan), 1 µM dexamethasone (10008980, Cayman Chemical, Ann Arbor, MI, USA), 0.5 mM IBMX (10008978, Cayman Chemical, USA), and 1 µM rosiglitazone. For the maintenance medium, 100 nM insulin and 1 µM dexamethasone were prepared with a DMEM-F12-based medium. To generate white adipocytes, hASCs were seeded at a density of 5 × 10 4 cells per cm 2 and grown to confluence. Both the induction and maintenance medium was subsequently changed every 4 days for a total of 21 days.
For induced adipocytes from d-hASCs, the cells were seeded in the preadipocyte growth medium (PGM-2 TM Bulletkit, Lonza, Walkersville, MD, USA). When the cells reach 80% confluency, change the growth medium to adipogenic differentiation medium to induce adipogenicity. Rhinsulin, dexamethasone, IBMX, and indomethacin were added to the adipogenic differentiation medium. Refresh the medium every 10 to 12 days [14].

Cell Viability Assay
The cell viability of the hASCs was analyzed using the 3-(4,5-Dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) assay. Succinate dehydrogenase in the mitochondria of living cells can reduce exogenous MTT to water-insoluble formazan and deposit it in cells, whereas dead cells have no such function. In brief, cells at the density of 5 × 10 4 cells/mL were seeded in a 96-well plate. After 16 h, the cells were treated with oleacein concentrations of 0, 1, 5, 10, 20, 40, 80, and 100 µM for 24 h and then incubated with MTT (0.5 mg/mL) for an additional 4 h. The forming formazan was dissolved using 10% SDS. Using a microplate reader, the absorbance was read at 570 nm. We conducted three independent experiments.
Data were calculated as cell proliferation (%) = (mean OD value of treated well − mean OD value of the blank)/(mean OD value of untreated well − mean OD value of the blank) × 100%.

Cell Measurement and Lipid Accumulation Assay on hASCs
The hASC cells were subjected to an induction medium to induce adipocyte differentiation, which was followed by a maintenance medium in both 37 • C and 5% concentrations of CO 2 environments. Lipid droplet formation was observed using the Leica TCS SP8 confocal microscope, and the cellular lipid accumulation was determined using Nile red (144-08811, FUJIFILM, Tokyo, Japan). The Nile red working solution concentration was 25 µg/mL by dissolving in DMSO. To quantify lipid accumulation, cells were seeded in a black 96-well plate and differentiated into adipocytes for 21 days. After washing with PBS, 5 µL of the working solution was added to each well containing PBS and incubated for 30 min at 37 • C. Fluorescence was measured by using a microplate reader (Varioskan LUX, Thermo Fisher Scientific, Waltham, MA, USA) with excitation at 485 nm and emission at 535 nm, according to the manufacturer's introduction. For each concentration, we conducted three independent experiments. RFU was used to quantify the fluorescence intensity emitted by the Nile Red. The data were calculated as RFU = [(mean RFU value of treated well − mean RFU value of blank) × 100%]/(mean RFU value of untreated well − mean RFU value of blank).

RNA Extraction and Quantification
The cells were cultured in a 6-well plate with total volumes of 3 mL during induction and maintenance medium conditions. An Isogen kit (Nippon Gene, Tokyo, Japan) was used to extract the total RNA according to the manufacturer's instructions. RNA was quantified using the Nanodrop 2000 spectrophotometer (ThermoScientific, Waltham, MA, USA).

Microarray Experiment Processing
The GeneChip 3 IVT PLUS Reagent Kit (Affymetrix Inc., Santa Clara, CA, USA) and Affymetrix ® 3 IVT Array Strips for Affymetrix's GeneAtlas ® System were utilized for the gene expression profiling of d-hASC, which was carried out following the instructions in the manufacturer's user guide. As previously described, the total RNA was extracted, and then 250 ng per sample was used for strand synthesis. GeneAtlas ® Hybridization station, GeneAtlas ® Fluidics Station, and GeneAtlas ® Imaging Station were used for hybridization, washing, staining, and scanning for the human genome array strips after the samples were fragmented and labeled. For the hASC group, we used Clariom S Assay Human with the SST-RMA algorithm, and for the d-hASC group, we used Human GENO-U219 with the RMA algorithm.

Gene Expression Analysis
The probe intensity for various comparisons was conducted using Transcriptome Analysis Console software (version 4.0.2, Thermo Fisher Scientific, Waltham, MA, USA) A threshold value of FC > 2 (in linear space) and a p-value of < 0.05 (one-way between-subject analysis of variance ANOVA) were set to identify differentially expressed genes (DEGs) in analyzing the comparisons to normal hASCs (undiff). After adipocyte differentiation, DEGs were filtered using a linear fold change (FC) greater than 1.1 and a p-value less than 0.05 for comparing the OLE + diff and diff groups in hASCs. For DEGs in d-hASCs, the FC values of the top 500 genes (FC = 1.18) were extracted and sorted in descending order for clustering and more focused enrichment analysis.
Metascape (http://metascape.org, accessed on 1 March 2023), a web-based tool, was utilized to conduct gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses [63]. The BioPlanet_2019 database was used to cluster genes function under the Enrichr online tool (https://maayanlab.cloud/Enrichr/, accessed on 9 March 2023) [64][65][66]. The transcription factor (TF) and DEG connections were examined using TRRUST (transcriptional regulatory relationships unraveled by sentencebased text-mining)) [67] and PPI networks, using the PPI networks constructed via the NetworkAnalyst tool (https://www.networkanalyst.ca/NetworkAnalyst/home.xhtml, accessed on 27 March 2023) [68], which utilizes the IMEx Interactome database [69]. The initial set of seeds and nodes was constructed based on the first-order generic PPI. We set the degree filter value as 3.0 and the tissue filter as adipose tissue and manually minimized the network more specifically. A bioinformatics online tool (https://www.bioinformatics. com.cn/, accessed on 15 March 2023) was used to generate the balloon plot and bubble plot, while the lollipop illustration was visualized using Sangerbox 3.0 together with the circle plot under the condition of FDR based on Benjamini-Hochberg (http://sangerbox.com/, accessed on 27 March 2023) [70].

Statistical Analysis
One-way ANOVA was performed using GraphPad Prism 9 software (San Diego, CA, USA) to determine statistical significance. A p-value of 0.05 was considered significant. The results are reported as mean ± SEM. The results represent the average of three parallel experiments.

Conclusions
In summary, our findings suggest that OLE has the potential to confer health benefits, particularly with regard to reducing inflammation and maintaining metabolic homeostasis in adipocytes differentiated from both healthy and diabetic adipose stem cells. Moreover, we have identified potential targets of OLE that regulate lipid and glucose metabolism. While our study is under the whole-genome transcriptomics analysis, it provides a basis for future investigations into the therapeutic potential of OLE in the fields of chemical biology and drug discovery.