Bioinformatics exploration of olive oil: molecular targets and properties of major bioactive constituents

This is anOpe Abstract – Olive oil possesses medicinal properties which include antimicrobial, antioxidant and antiinflammatory, anti-diabetes, and anti-cardiovascular diseases. Oleic acid is the most abundant (95%) constituent of olive oil and others include linoleic acid, oleuropein, oleanolic acid, maslinic acid, melatonin, and others. The objective of this study is to predict the molecular targets and properties of key bioactive components of olive oil in human. Bioinformatics methods, which involved pharmacokinetics prediction, target prediction and gene network analyses, were used. The results showed that oleic acid has similar targets with linoleic acid, and showed significant probability of binding to several targets such as fatty acidbinding proteins in the adipose, epidermal, liver and muscle as well as alpha, delta and gamma peroxisome proliferator-activated receptors (PPARs). Carbonic anhydrase showed to be the only significant target of tyrosol, while protein-tyrosine phosphatase 1B, and CD81 antigen were targeted by maslinic acid and oleanolic acid. This study has applauded oleic acid, linoleic acid and tyrosol as olive oil bioactive constituents that have several potential pharmacological effects in humans that modulate several enzymes, receptors and transcription factors. The future work will be to investigate the effects of oleic acid on fatty acid-binding proteins and telomerase reverse transcriptase; melatonin on quinone reductase 2; tyrosol on carbonic anhydrase II; maslinic acid and oleanolic acid on protein-tyrosine phosphatase 1B.


Introduction
Olive (Olea europaea L.) is an ancient plant that belongs to the family Oleaceae, which contain about 600 species within 30 genera (Iaria et al., 2016). The genus Olea L. consists of more than 30 species, which are distributed in Africa, Asia, Europe and Oceania, with only Olea europaea subsp. europaea var. europaea being the cultivated olive (Fogher et al., 2010). Olive oil is produced from the olive plant mainly in the Mediterranean basin which produces 90% of the olive oil consumed worldwide, and it is the principal source of healthy fatty acids, as well as polyphenols and vitamins in minutes (Barbaro et al., 2014;Vasto et al., 2014;Gerber and Hoffman, 2015;Martinez-Gonzalez et al., 2015;Rigacci and Stefani, 2016). Extra virgin olive oil is categorized as a medicinal food because of its nutraceutical benefits and wide range of therapeutic effects such as anti-inflammatory, antioxidant and antimicrobial effect (Cicerale et al., 2012), anti-cardiovascular diseases (Estruch et al., 2006(Estruch et al., , 2018, diabetes (Salas-Salvado et al., 2011, neuronal and geriatrics diseases (Khalatbary, 2013;Rodriguez-Morato et al., 2015).
The difference between extra virgin olive oil (EVOO) and virgin olive oil (VOO) is that EVOO has a maximum acidity of 0.8% and may have no defects. VOO can have an acidity up to 2.0% and has a slight change in taste. Pure olive oils are usually refined olive oils, they are obtained from VOO by refined methods and its free acidity is expressed as less than 0.3% of oleic acid. For example, the concentration of maslinic acid increases as the olive-oil quality decreases, from values of 38 mg.kg À1 for extra virgin to 227 mg.kg À1 for 9.3%-acidity virgin olive, and to 721 mg.kg À1 for crude pomace olive oils (Perez-Camino and Cert, 1999). However, the differences in the concentration of bioactive compounds in EVOO and VOO will affect their bioavailability at the sites of action.
The genome sequence and functional annotation of (Olea europaea L. subsp. europaea var. europaea cv. "Farga") has resulted in 56 349 unique protein coding genes (Cruz et al., 2016), while that of wild olive tree produced over 50 000 protein-coding genes (Unver et al., 2017). Several types of biomarkers such as single nucleotide polymorphisms (SNPs), sequence characterized amplified regions (SCARs) (Fogher et al., 2010), have been used to identify cultivars used in olive oil production (Busconi et al., 2003), and to develop the genetic structure of wild and cultivated olives (Baldoni et al., 2006). Ayed and Rebai (2019) have analyzed 11 Tunisian table olive cultivars based on seven SNP molecular markers (ANTHO3, CALC, FAD2.1, FAD2.3, PAL70, SOD, and SAD.1), to show the possibility of quality authentication and traceability of table olive oil. The functional divergence of oil biosynthesis pathway genes, such as FAD2, SACPD, EAR, and ACPTE, after paralogous event (following duplication), has been responsible for the differential accumulation of oleic and linoleic acids produced in olive when compared with sesame (a closely related oil crop), and the decrease in FAD2 expression and increase in SACPD expression possibly explain the accumulation of exceptionally high levels of oleic acid in olive (Unver et al., 2017).
The impact of dietary olive oil on cancer development have been studied by Zhang et al. (2019), where RNA-sequencing technology and comprehensive bioinformatics analyses were used to elucidate the molecular processes regulated by dietary fat. Differentially expressed genes (DEGs) were identified and were functionally analyzed by gene ontology (GO), kyoto enrichment of genes and genomes (KEGG). Then, proteinprotein interaction (PPI) network and sub-PPI network analyses were conducted using the STRING database and Cytoscape software. The study suggests that a high olive oil diet aggravates cervical cancer progression in vivo and in vitro (Zhang et al., 2019). Diet-gene interactions are studied by the concept of nutrigenetics and nutrigenomics, which identify gene variants associated with different responses to nutrients and the effect of nutrients on the metabolic pathways and homeostatic regulation, respectively (Muller and Kersten, 2003;Ordovas and Mooser, 2004). The objective of this study is to predict the molecular targets and properties of key bioactive components of olive oil in human. This work corroborates the targets which have been experimentally discovered, and it also predicts the novel targets which have not been clinically explored, which may be of medical importance in treatment of certain diseases such as cancer, atherosclerosis, nephrotoxicity, inflammation and skin disorder.

In Silico target prediction
The structure of several bioactive compounds of olive oil as listed in literature (Fernández-Montesinos et al., 2010;Liu et al., 2010;Segura-Carretero et al., 2010;Bulotta et al., 2014; Tresserra-Rimbau and Lamuela-Raventos 2017), were obtained from the PubChem compound database in canonical SMILES (simplified molecular input line entry specification) format. The SMILES of each of these compounds were used for in silico prediction of target on the SwissTargetPrediction server, where Homo sapiens was selected as target organism (Diana et al., 2019).

In Silico pharmacokinetics
Six active ligands (oleic acid, tyrosol, linoleic acid, oleanolic acid, maslinic acid, and melatonin) were selected based on availability of predicted targets in human with a probability of at least 40% and the SMILES of each of these compounds were used for in silico ADME (absorption, distribution, metabolism, and excretion) screening on Swis-sADME server (Diana et al., 2017). ADME screening was performed at default parameters.

Results and discussion
The predicted targets and pharmacokinetics of six active constituents of olive oil (oleic acid, tyrosol, linoleic acid, oleanolic acid, maslinic acid, and melatonin) are shown in Tables 1 and 2. The choice was based on the fact that these six compounds have predicted targets genes in human with a probability of at least 40% as shown in Table 1 as well as the amount present in the olive oil. Oleic acid which is the main active constituent of olive oil, has six similar targets with linoleic acid, and showed significant probability of binding to several targets such as fatty acid-binding protein in the adipose, epidermal, liver and muscles as well as peroxisome proliferator-activated receptors (alpha, delta and gamma). It has been reported that PPAR-gamma ligands could inhibit tumor necrosis factor (TNF)-alpha, interleukin (IL)-6 and IL-1-beta expression in monocytes, iNOS, matrix metalloprotease-9 (MMP-9) and scavenger receptor-A expression in macrophages among others (Krey et al., 1997). This in silico work currently supports the hypothesis that exogenous fatty acids (FAs) enter cell nuclei by binding to FABPs. On getting inside the cell, dietary FAs are reversibly bound to lipid-binding proteins, such as FABPs and acyl-CoA binding proteins, such as Acyl-CoA desaturase (Esteves et al., 2016). Dietary omega-3 polyunsaturated fatty acids (PUFAs) downregulate fatty acid-binding protein-4 in the adipocytes in a sex-dependent fashion and also modulate stearoyl-CoA desaturase activity in an age and sex-specific manner (Balogun and Cheema, 2016;Feltham et al., 2019).
Studies have shown that omega-3 PUFAs or oleic acid could activate the expression of transcription factors such as PPAR (Xu et al., 1999;Brunelleschi et al., 2007) and this may have beneficial effects in human inflammatory bowel disease (Dubuquoy et al., 2006). PPAR has been noted for its indispensable involvement in the neurotrophic effect of oleic acid in neurons (Bento-Abreu et al., 2007). Oleic acid has shown a significantly higher glucosyltransferase (GTF) inhibitory activity and antibacterial activity on Streptococcus mutans (Choi et al., 2010). Study has shown that oleate was the most effective down-regulator of FA biosynthesis and cholesterogenesis of various kind of FAs, where it had reduced mRNA abundance, protein level and the activity of both 3-hydroxy-3-methyl-glutaryl CoA (HMG-CoA) reductase and acetyl-CoA carboxylase (Gnoni et al., 2010).
In this study, carbonic anhydrase showed to be the only significant targets of tyrosol, while protein-tyrosine phosphatase 1B, and CD81 antigen were targeted by maslinic acid and oleanolic acid. Carbonic anhydrase 6, CD209 antigen, and CD44 antigen, have been reported as part of HDL-associated proteins based on the effects of olive oil phenolic compounds (Pedret et al., 2015). Tyrosol is able to inhibit the activation of transcription factors, including NF-kB and STAT-1a, and expression of inducible nitic oxide synthase (iNOS) and cyclooxygenase-2 (COX-2) genes, in cultures of activated macrophages and rat colitis (Moreno, 2003), as well inhibit TNF-a release by LPS-stimulated peripheral blood mononuclear cells isolated from healthy volunteers (Giovannini et al., 2002). Tyrosol also inhibits 5-lipoxygenase, reducing leukotriene B4 and reactive oxygen species (ROS) generation in calcium ionophore-stimulated rat peritoneal leukocytes (De la Puerta et al., 1999). It has been reported that hydroxytyrosol could inhibit human LDL oxidation and platelet aggregation (Morales and Lucas, 2010). Both linoleic and docosahexaenoic acids could cause modulation of gene expression in rat cardiomyocytes .
Melatonin receptors are expressed in the tissues of brain (cerebellum and hippocampus), intestine, kidney, and testis. Natural killer (NK) cells, T-lymphocytes, eosinophils, and mast cells possess melatonin receptors . Melatonin could modulate the biological activity and toxicity of tumor necrosis factor-a (TNF-a), increase of interferon-g production (Fernández-Montesinos et al., 2010). Melatonin administration increases the proliferative response of rat lymphocytes, increases the number of NK cells, stimulates the release of pro-inflammatory cytokines interleukin (IL)-1, enhances phagocytosis and modulates apoptosis .
Maslinic acid (2-a, 3-b-dihydroxyolean-12-en-28-oic acid) is a pentacyclic triterpene abundant in the cuticular lipid layer of olive fruits. Maslinic acid has therapeutic properties related to health and disease, including antiinflammatory, antioxidant, antiviral, antihypertensive, and antitumor activities (Fernández-Navarro et al., 2010). Oleanolic acid has been recognized as an PPAR-a agonist (Huang et al., 2005). Oleanolic acid and maslinic acid could modulate the activity of DNA polymerase beta (POLB) and proteintyrosine phosphatase 1B (PTPN1), aldo-keto reductase family 1 member B10 (AKR1B10), nuclear receptor ROR-gamma ! (RORC), receptor-type tyrosine-protein phosphatase F (LAR) (PTPRF), 11-beta-hydroxysteroid dehydrogenase 1 (HSD11B1) and others as shown in Table 1. POLB is involved in the homeostasis of the number of cells, DNA repair, inflammatory response and aging process. PTPN1 is a non-receptor type tyrosine-specific phosphatase that dephosphorylates the receptor protein tyrosine kinases (such as INSR, EGFR, CSF1R, PDGFR) and dephosphorylates the non-receptor protein tyrosine kinases (such as JAK1, JAK2, JAK3, Src family kinases, STAT1, STAT3 and STAT6) either in the nucleus or the cytoplasm. It negatively regulates numerous signaling pathways and biological processes like hematopoiesis, inflammatory response, cell proliferation and differentiation, and glucose homeostasis. PTPN1 plays a multifaceted and important role in the development of the immune system. AKR1B10 is highly expressed in the small intestine, colon and adrenal gland, and plays a critical role in detoxifying dietary and lipid-derived unsaturated carbonyls, and their glutathione-conjugates carbonyls. This protein is involved in the retinol metabolism pathway. Thus, maslinic acid and Oleanolic acid could interfere with retinoid metabolic process. Study has shown that maslinic acid and oleanolic acid significantly reduce hyperlipidemia induced by a highcholesterol diet and lower the expression of the acyl-CoA cholesterol acyltransferase (ACAT) gene (Liu et al., 2010). As shown in Table 2, oleic acid is moderately soluble, with a high gastrointestinal absorption, serves as an inhibitor for CYP1A2 and CYP2C9, and has a high bioavailability score. This could justify why oleic acid possesses striking therapeutic effects in the intestine, liver and adipose tissues. Although tyrosol has features closely similar ADME properties to oleic acid, it could permeate the blood-brain barrier (BBB) with no action of the cytochromes and not affected by P-glycoprotein. Tyrosol is bioavailable in humans, even from moderate doses of olive oil consumption with substantial variance among women and men (Covàs et al., 2003). The half-life of tyrosol is estimated to be 2-4 h in humans (Covàs et al., 2003). Among the six compounds investigated in this study, tyrosol has the highest skin permeability rate, followed by melatonin. Linoleic acid and melatonin could permeate the BBB and have a high gastrointestinal absorption.
Histone deacetylase 3 (HDAC3), HDAC5, LSD1 (a histone demethylase), Atrophin1, and BCL11A have been reported to interact with TLX to regulate the expression of target genes, and has been validated in human Y79 retinoblastoma cells (Wang and Xiong, 2016). Study has shown that oleic acid triggers hippocampal neurogenesis by binding to TLX/NR2E1 and changes it to a transcriptional activator from a transcriptional repressor of cell cycle and neurogenesis genes (Kandel et al., 2020). Thus, through TLX (orphan nuclear hormone receptor), oleic acid could indirectly activate Wnt7a expression, suppresses the expression of p21 via a p53-dependent mechanism, promotes EGFR signaling in brain cells, and modulates the mitogen-activated protein kinase (MAPK) pathways (Wang and Xiong, 2016) and enhances STAT1 function (Beiting et al., 2015). A bioinformatics study conducted to predict peroxisome proliferatoractivated receptors (PPARs) gene targets on a genome-wide basis, has shown that PPARs could directly regulate some genes such as chromatin remodeling, DNA damage response, Wnt, and mitogen-activated protein kinase (MAPK) (Lemay and Hwang, 2006). FAs and eicosanoids can regulate gene transcription through PPARs (Kliewer et al., 1997;Kersten et al., 2000).

Conclusion
Knowledge of bioinformatics as it is being applied in lipidomics of a given organism, has not been fully applied to the study of lipids as therapeutics, in order to predict the  acclaimed medicinal properties. This study has applauded oleic acid, linoleic acid and tyrosol as olive oil bioactive constituents that have several potential pharmacological effects in human by modulating several enzymes, receptors and transcription factors. Moreover, these molecular effects of olive oil indicate its medicinal importance in the treatment of oxidative stress, inflammation, cardiovascular diseases, obesity, diabetes, and age-related diseases. Furthermore, chemical biology and in silico simulation of pharmacological potential of oleic acid (such as molecular docking and dynamics, drugdrug interaction) will yield significant insights to the presently unexplored molecular mechanisms of action to explain the therapeutic effect of olive oil. The future work will be to investigate the effects of oleic acid on fatty acid-binding proteins and telomerase reverse transcriptase; melatonin on quinone reductase 2; tyrosol on carbonic anhydrase II; maslinic acid and oleanolic acid on protein-tyrosine phosphatase 1B.
Conflicts of interest. The authors declare no conflicts of interest.