Abstract

Background

Metabolic syndrome (MetS) is a very common condition that is associated with increased cardiovascular disease (CVD) risk and type 2 diabetes mellitus, itself a risk factor for CVD. Its diverse clinical characteristics illustrate the complexity of the disease, involving several dysregulated metabolic pathways and multiple genetic targets. When caloric intake exceeds caloric expenditure, a positive caloric balance and subsequent storage of fat in adipose tissue appears; this often is cause of adipocyte hypertrophy and obesity. Depending on interaction of genetic and environmental factors, this hypertrophy can be followed by macrophage accumulation within adipose tissue leading to local hypoxia, inflammation, and oxidative stress. These deleterious effects favor adipose functional failure resulting in changes in systemic energy delivery; these also impair glucose consumption and activate self-regulatory mechanisms that extert influence over whole body homeostatic systems and link obesity to numerous health problems associated with MetS [1]. A low-level systemic chronic inflammation, with abnormal cytokine production in adipose tissue, increased acute-phase reactants, and activation of inflammatory signaling pathways, often acting over a period of many years, appears to constitute a potential link to atherosclerosis development in these patients [2]. In addition, changes in postprandial metabolism undergoing every time we have a meal as well as alterations in this state may play an important role in the development of cardiovascular and cardiovascular associated diseases [3]. Moreover, Van Oostromet et al. [4] provides evidence that postprandial triglyceridemia is related to a pro-inflammatory state due to high expression of markers of neutrophilic and monocyte activation.

The anti-atherogenic effects associated with Mediterranean Diet (MD) rich in Virgin Olive Oil (VOO) consumption, could contribute in explaining the low rates of cardiovascular mortality found in Southern European Mediterranean countries, in comparison with other Western populations [5]. It has been demonstrated recently that inhibition of circulating immune cell activation could be a protective mechanism by which MD exerts its healthy effects [6]. Bonani et al. [7] showed that extra VOO consumption reduces inflammatory markers and increase serum antioxidant capacity at postprandial state. Previously, our group has showed how a VOO-rich MD, during postprandial state, reduces inflammatory response of peripheral blood mononuclear cells (PBMCs) mediated by transcription factor NF-κB, when compared to, butter and walnut-enriched diets or Western diets [8,9]. These results have been replicated by others and support the hypothesis that decreasing NF-κB pathway activation is a mechanism involved in the anti-inflammatory effect of a VOO-rich MD [10,11]. Nevertheless, it has been proposed that healthy effects of VOO may be due not only to its oleic acid content but also to the antioxidant and anti-inflammatory capacity of minor components, especially the phenol-rich fraction [12,13]. Our group demonstrated that the VOO phenol fraction also improves endothelial dysfunction and haemostatic profile during postprandial state [14,15]. Other studies showed how these compounds ameliorate lipid profiles and decrease oxidative stress in vivo [12]. However, in animal models, it has been showed that hydroxytyrosol, a phenol found in olive oil, administered after being extracted from its original matrix could be not only non beneficial but indeed harmful for health [16].

It has been speculated that potentially beneficial effects could be due to olive oil modulation on genes involved in proliferative, antioxidant and inflammatory pathways. In order to clarify this question, some human, animal models and in vitro studies were undertaken using gene expression microarray platforms [17]. These studies showed that olive oil is able to modify gene expression coding for proteins participating in cellular mechanisms involved in oxidative stress resistance [18], lipid metabolism [19] and other atherosclerosis-related traits/pathways [20,21]. Nonetheless, it is unresolved whether these changes in gene expression are mediated by oleic acid or executed by olive oil polar minor components, maybe as a consequence of their antioxidant capacity, or by interacting directly with receptors, enzymes or transcription factors. To date, no intervention studies in humans have thoroughly explored the effects of VOO phenol compounds on gene expression. Thus, the main purpose of this study was to identify genes which undergo changes in expression in PBMCs in patients with MetS, after acute intake of breakfast based in phenol-rich virgin olive oil, compared to phenol-poor olive oil breakfast.

Results

Results validation by qRT-PCR

To confirm microarray results using an independent technique, log₂ of the normalized microarray ratio values were compared with log₂ ratio values for four genes obtained from qRT-PCR experiments (JUN, PTGS2, EGR1 and IL1B). Results showed that features of the changes in both procedures were similar with Spearman r coefficients between 0.901 and 0.968 and p-values less than 0.001 (Figure ​(Figure11 and ​and22).

Open in a separate window

Figure 1

Microarrays results validation by qRT-PCR. Mean ( ± S.E.M.) of relative expression values after the consumption of low and high phenol content olive oil. One way ANOVA p value is showed. Gene expression values were log transformed before statistical analyses.

Open in a separate window

Figure 2

Correlation between microarrays and qRT-PCR results. Correlation analysis between log₂ ratio of the gene expression values from microarrays and qRT-PCR experiments. Ratio of the gene expression values correspond to gene expression after consumption of high phenol content olive oil divided by gene expression after the consumption of low phenol content olive oil. Spearman r values are shown in the graphs.

Pathway analysis

In order to investigate functional relationships in the set of differentially expressed genes, we used the Ingenuity Pathway Analysis Software [22] (Ingenuity Systems, Redwood City, CA USA) which employs a predefined knowledge base containing over 10.000 curated human genes. Of the 98 differentially expressed genes found in patients with MetS during the postprandial period after intake of olive oil phenols, two transcripts (LOC284454, AF351612) showed no entries within the knowledge base and only 81 genes were eligible for network analysis. Inflammatory disorder was the most highly represented disorder (39 genes, p = 2.11E-19). From these, 35 genes (PTGS2, IL1B, IL6, OSM, CCL3, CXCL1, CXCL2, CXCL3, CXCR4, NAMPT, DUSP1, DUSP2, EGR1, EGR2, EGR3, EREG, FOSB, G0S2, JUN, JUNB, NFKBIA, NFKBIZ, NR4A1, NR4A2, PER1, SOCS3, SOD2, TAGAP, TNFAIP3, ZFP36, AREG, CA2, CD69, CD83, CDKN2A) were under-expressed and after intake of virgin olive oil with high content in phenolic compounds, 4 genes were over-expressed (CCR2, CA1, CPVL, FN1). Cellular functions most strongly associated with the differentially expressed genes were cell death (41 genes, p = 7.53E-11), cell migration (24 genes, p = 3.88E-8), cell division (23 genes, p = 6.29E-8), cell proliferation (32 genes, p = 9.90E-7) and transcription (25 genes, p = 3.55E-05). Network analysis yielded 14 different sub-networks, from which, one was identified as preferentially enriched by genes found differentially expressed after high-phenol VOO intake. This top-scoring sub-network included 26 genes whose related top-function was 'inflammatory diseases' with a probability value of 10^-54 (Fisher's exact test) of gene interrelationships being by chance. All sub-networks were merged to obtain the overall network shown in Figure ​Figure3.3. Interacting proteins were added using Ingenuity Pathways Knowledge Base database. Finally non-connected genes or those connected by two or more edges were removed, except for those genes that were found to be differentially expressed in our microarrays analysis.

Open in a separate window

Figure 3

Ingenuity Pathway Analysis Network. Network of phenol-rich VOO modulated genes. Gray symbols denote that the gene was found over-expressed or under-expressed by phenols in microarrays analysis.

Discussion

In the present study, we have observed that the phenol fraction of VOO in vivo is able to repress the expression of several genes related to inflammation pathways in patients with MetS during postprandial period (Additional File 1). This finding draws interest since pro-inflammatory state remains as one component of MetS [23] and low-grade inflammation is often associated with endothelial dysfunction [24], which by itself is associated to the development of atherosclerosis [25].

The PTGS2 gene encodes prostaglandin-endoperoxide synthase 2 (COX-2), an inducible isozyme involved in prostaglandin biosynthesis using arachidonic acid as substrate. In macrophages, and other cells, COX-2 activity is rapidly increased by various stimuli, such as pro-inflammatory cytokine IL1β. Substantial evidence indicates that up regulated PTGS2 expression and prostaglandin synthesis indeed influence chronic inflammatory conditions [26]. De la Puerta et al. observed that murine macrophages showed significantly reduced IL1β production and COX-2 activity after olive oil-enriched diet [27]. Recently, it has been demonstrated that hydroxytyrosol, one of the most important phenol compounds found in virgin olive oil, attenuates in vitro LPS-induced transcription of PTGS2 [28]. Our study showed that in vivo intake of phenol-rich virgin olive oil in humans is associated with a decreased expression of both IL1B and PTGS2, as compared to low-phenol olive oil intake. Those effects could contribute to reduced inflammation during postprandial state in agreement with anti-inflammatory effects observed after VOO-rich MD consumption [8,9].

The cytokine-cytokine interaction pathway contains a network of proteins (chemokines and their receptors) involved in the recruitment and activation of leukocytes during inflammatory response. Expression of genes such as CCL3, CXCL1, CXCL2, CXCL3, CXCR4, IL1B, IL6, and OSM is described under-expressed after acute intake of phenol-rich olive oil in our intervention study on diet. CCL3 gene, which codes for macrophage inflammatory protein-1 (MIP-1), has been implicated in monocyte infiltration of adipose tissue, an action that could significantly influence a pro-inflammatory pattern within endothelial cells [29]. CXCL1, CXCL2 (MIP2A), and further, CXCL3 (MIP2B) are genes for small and structurally related chemokines that regulate cell trafficking of various types of leukocytes through interactions with a subset of G protein-coupled receptors. Elgazar-Carmon et al. have demonstrated in mice that early neutrophil infiltration of adipose tissue may be mediated by CXCL1, a process that would precede macrophage infiltration after long-term consumption of a high-fat diet [30]. IL6 encodes a cytokine which is secreted to serum and induces a transcriptional response involved in a wide variety of inflammation-associated conditions, including MetS and type 2 diabetes mellitus (T2DM) [31]. On the other hand, it has been proposed that IL6 and OSM, which encodes oncostatin M, a growth regulator and member of the IL6 group of cytokines, can contribute to the increased cardiovascular risk in obese patients by up regulating PAI -1 in adipose tissue [32].

Activation of NF-κB and MAPK pathway, a cascade of phosphorilation events that result in the activation of transcription factors like CREB and AP-1, synergize for expression of inflammatory genes through coordinate bindings of transcription factors to κB and AP-1 sites which have been found together in the promoters of e.g. IL6 and TNF-α, and many other inflammatory genes [33]. Chemokine repression found in our study could be consequence of phenols interaction with this inflammation signaling system, since expression of some genes involved in NF-κB/MAPK/AP-1 signaling pathways was also modulated after phenol-rich olive oil based breakfast. NF-κB is a transcription factor activated by pro-inflammatory cytokines [34] and oxidative stress mediators [35]. Recently Pierce et al. have demonstrated that NF-κB activation is important in mediating vascular endothelial dysfunction in obese humans [36]. The product of SGK1 gene, encoding a serum/glucocorticoid regulated kinase with a role in stress response and by itself being a downstream target for PI₃K signaling, enhances nuclear NF-κB activity by phosphorylating an inhibitory kinase IKKα [37]; so repression on expression of SGK1 gene by olive oil phenols would decrease the NF-κB activation. In addition, NFKBIA gene, which encodes to IκBα, a member of an inhibitory IκB family that retains NFκB into the cytoplasm, remained under-expressed after acute intake of phenol-rich olive oil. It has been reported that NF-κB binds to the IκBα promoter in order to activate its transcription [38]. Thus, this negative feedback mechanism results in rapid cycles of inhibition and stimulation of NF-κB where a decrease on NF-κB activation is accompanied by a reduction on NFKBIA gene expression, as observed in our results. The hypothesis that NF-κB activation is decreased by olive oil phenols is also supported by two in vivo studies which showed reduced NF-κB activation after olive oil consumption [8,9]. Additionally, in vitro studies showing attenuated NF-κB activation by resveratrol support the hypothesis that this pathway employs a shared mechanism by which polyphenols reduce expression of genes encoding inflammatory cytokines and adhesion molecules [39].

After intake of virgin olive oil with high content in phenolic compounds we found a decreased postprandial expression of DUSP1 and DUSP2. Those genes encode dual serine-threonine phosphatases, which down regulate members of p38, MAPK/ERK and SAPK/JNK, the three final effectors of the MAPK pathway [33]. In addition, TRIB1, another gene repressed by phenol-rich olive oil, is involved in MAPK signaling, participating in the activation of ERK proteins [40] and being up regulated in human atherosclerotic plaques [41]. Thus, reduction of TRIB1 expression by olive oil phenols could promote decreased ERK activation. This observation agrees with in vitro studies demonstrating that phenol compounds of green tea down regulate PTGS2 expression by decreasing ERK and p38 MAPK activation [42]. Our results allow us to hypothesize that olive oil phenols influence activation of AP-1, which consists of a variety of heterodimers of Jun, Fos and activating transcription factor families [43], by means of two different mechanisms: a) one direct, involving repression of JUN, JUNB and FOSB as observed after phenol-rich olive oil intake; and/or b) another one indirect, through MAPK pathway, where the relative intensity and duration of activation determine the type of response.

Lastly, biomedical literature and text mining tasks were performed to identify interactions of differentially expressed genes in PBMCs as response to phenol-rich VOO with conditions clustered around MetS such as obesity, hypertension, dyslipemia, hyperglycemia, or T2DM. Recently, Chen et al. have described a macrophage-enriched gene network (MEGN) of ~1237 genes referred as having causal relationship with complex-disease traits associated with MetS [44]. Thirteen genes share our set of differentially expressed genes in PBMCs after acute intake of phenol-rich olive oil and MEGN: JUN, RGS1, CXCL2, ANXA3, RASGEF1B, CD83, CA2, EGR2, DIAPH3, CCL3, and TLR7, PSAP and IFIT3. De Mello et al. assessed individuals with both impaired glucose metabolism and MetS on how long-term weight loss affects expression of cytokines in PBMCs. Weight reduction resulted in a decrease in of IL1B expression [45]. Kaiser et al. showed by microarrays analysis in PBMCs a set of 22 over-expressed genes in T1DM and T2DM compared to healthy subjects [46]. Interestingly, 8 of the identified 22 over-expressed genes in T2DM were repressed by olive oil phenols, according to our intervention study (IL1B, EGR2, EGR3, PTGS2, FOSB, CXCL1, SGK, and TRIB1). In addition, PBEF1, another gene involved in the pathogenesis of T2DM [47], was also found repressed after consumption of phenols-rich olive oil. Taken together, this finding could lead to potential therapeutic implications in T2DM.

Conclusions

Our study shows that intake of breakfast based in virgin olive oil being rich in phenol compounds is able to repress expression of several pro-inflammatory genes in vivo, thereby switching activity of PBMCs to a less deleterious inflammatory profile. These results provide at least a partial molecular basis for risk reduction of cardiovascular disease observed in Mediterranean countries, where VOO represents a main source of dietary fat. Admittedly, other lifestyle factors are also likely to contribute to lowered risk of CVD in this region. Nonetheless, our data suggest that mechanisms by which these micronutrients in phenol-rich olive oil would exert their anti-inflammatory effect could involve pathways related to NF-κB/AP-1, cytokine-cytokine receptor interaction, arachidonic acid metabolism, and MAPK. These findings strengthen the relationship between inflammation, obesity and diet and provide evidence at transcription level of control of healthy effects derived from VOO consumption in humans. However, it would be interesting to evaluate whether these beneficial effects are maintained after prolonged feeding and if these effects are carried out by one or several olive oil phenolic compounds, or if they are consequence of a synergic effect of the total phenolic fraction.

Methods

Abbreviations used

ppm: part per million; MetS: metabolic syndrome; CVD: cardiovascular disease; VOO: virgin olive oil; MD: Mediterranean Diet; PBMC: peripheral blood; mononuclear cells; MEGN: macrophage-enriched gene network.

Authors' contributions

AC and JR contributed equally to this work. AC, JR, JL-M, and FP-J: designed research; AC, JMF, and MU: provision of study materials and subjects; AC, JR, JMF, AJ, and MS-G: performed research; LDP and MU: contributed new reagents/analytic tools; AC, AJ, and MS-G: collection and assembly of data; AC, JR, JMF, LDP, CM, PP-M, JL-M, and FP-J: analyzed data; CM, PP-M, JL-M, and FP-J: provided statistical expertise; AC: wrote the paper; JR, LDP, MU, JL-M and FP-J: critical review of the manuscript; FP-J: obtained funding. None of the authors had any personal or financial conflicts of interest.

Supplementary Material

References