Anti-platelet effects of olive oil extract: in vitro functional and proteomic studies
- First Online:
- Cite this article as:
- de Roos, B., Zhang, X., Rodriguez Gutierrez, G. et al. Eur J Nutr (2011) 50: 553. doi:10.1007/s00394-010-0162-3
- 391 Views
Platelets play a key role in haemostasis and wound healing, contributing to formation of vascular plugs. They are also involved in formation of atherosclerosic plaques. Some traditional diets, like the Mediterranean diet, are associated with a lower risk of cardiovascular disease. Components in these diets may have anti-platelet functions contributing to their health benefits.
We studied the effects of alperujo extract, an olive oil production waste product containing the majority of polyphenols found in olive fruits, through measurement of effects on platelet aggregation and activation in isolated human platelets, and through identification of changes in the platelet proteome.
Alperujo extract (40 mg/L) significantly decreased in vitro ADP- (p = 0.002) and TRAP- (p = 0.02) induced platelet activation as measured by the flow cytometry using the antibody for p-selectin (CD62p), but it did not affect the conformation of the fibrinogen receptor as measured by flow cytometry using the antibodies for anti-fibrinogen, CD42a and CD42b. Alperujo extract (100 mg/L) inhibited both collagen- and TRAP-induced platelet aggregation by 5% (p < 0.05), and a combination of hydroxytyrosol and 3,4-dihydroxyphenylglycol were, at least partly, responsible for this effect. Proteomic analysis identified nine proteins that were differentially regulated by the alperujo extract upon ADP-induced platelet aggregation. These proteins represent important mechanisms that may underlie the anti-platelet effects of this extract: regulation of platelet structure and aggregation, coagulation and apoptosis, and signalling by integrin αIIb/β3.
Alperujo extract may protect against platelet activation, platelet adhesion and possibly have anti-inflammatory properties.
KeywordsPlatelet functionMediterranean dietAlperujo extractProteomics
The “Mediterranean Diet”, which is high in olive oil, fruits, vegetables, grains and legumes, is associated with a lower rate of coronary heart disease [1, 2], as well as a reduction in all-cause mortality [3, 4]. Olive oil is the main source of fat in this type of diet [5, 6]. A mechanism by which consumption of the Mediterranean diet might protect against coronary heart disease is by improving the lipoprotein profile  lowering of blood pressure, and improving insulin sensitivity both in healthy and in type 2 diabetic patients . In addition, virgin olive oil (especially the first-pressed “extra virgin” type) contains a wide range of important minor anti-oxidant compounds, such as polyphenols, that contribute to the stability of the oil and may have anti-inflammatory and anti-atherosclerotic properties [9, 10]. Consumption of polyphenols has been associated with prevention from and treatment of atherosclerosis .
Specific components of the phenolic fraction of olive oil may also inhibit platelet function and eicosanoid formation. Supplementation of patients with uncomplicated type 1 diabetes with olive oil waste water providing 25 and 12.5 or 6.6 mg of hydroxytyrosol significantly decreased serum thromboxane B2 (TxB2) production after blood clotting [12, 13]. These in vivo results are in agreement with an in vitro study where pre-incubation of platelet-rich plasma (PRP) with hydroxytyrosol for at least 10 min resulted in maximal inhibition of aggregation induced by ADP or collagen (IC50 were 23 and 67 μM, respectively), and TxB2 production by collagen or thrombin stimulation. In addition, this treatment inhibited the production of TxB2 and 12-hydroxyeicosatetraenoic acid (12-HETE) produced during blood clotting in serum . However, whilst hydroxytyrosol inhibits platelet aggregation in vitro , nothing is known about the effects of hydroxytyrosol on platelet activation, and moreover, no information is available on the effects of other polyphenols from olives and olive oil on both platelet aggregation and activation.
Materials and methods
Ethical approval for the use of human blood samples for the in vitro studies described below was obtained from the North of Scotland Research Ethics Committee.
Preparation of phenolic compounds
All extracts were produced at the Instituto de la Grasa, Seville, Spain. Alperujo, a waste product of the olive oil production process, was obtained by crushing olives to a fine paste, which was subsequently pumped into a two-phase decanter generating a liquid phase (oil) and a semisolid waste (alperujo) . Alperujo is formed from the olive pulp, the pit and the vegetable water, containing a high percentage of moisture (65–70%) and low percentage of fat (2.5–3.5%). It also contains all sugars and other soluble substances from the olive . Alperujo extract was produced by organic solvent extraction from alperujo pre-treated under hydrothermal conditions, as described by us previously . Briefly, alperujo samples were centrifuged, and the liquid fraction was extracted with ethyl acetate and dried under vacuum at 35 °C . Hydroxytyrosol was produced from alperujo extract by a patented industrial system . 3,4-Dihydroxyphenylglycol (DHPG) purification from alperujo extract was performed by a new system that has been patented . The amounts of hydroxytyrosol and DHPG in the alperujo extract were determined by HPLC on a Hewlett-Packard Series 1100 liquid chromatograph system with an UV–visible detector and a Rheodyne injection valve (20 μL-loop). A spherisorb ODS-2 column (250 × 4.6 mm i.d.; particle size 5 μm) (Tecnokroma, Barcelona, Spain) was used at 25 °C. Elution was performed as described by us previously . Briefly, a mobile phase of trifluoracetic acid in water (pH = 2.5) and acetonitrile, with a gradient from 5 to 25% of acetonitrile in 30 min, were used at 1.0 mL/min. Chromatograms were recorded at 280 nm (Fig. 1). The substances were dissolved to the required concentration in PBS and buffered to pH 7.4 before use.
In vitro platelet aggregation
Blood from fasting healthy male and female volunteers with normal platelet function was collected in sodium citrate and centrifuged at 200g for 10 min at room temperature. The 10- to 15-min incubation time was chosen based on previous work with plant bioactive compounds [23–25]. PRP was removed, and the remaining sample was spun at 2,000g for 10 min to isolate platelet-poor plasma (PPP). PPP was used to adjust PRP to a platelet count of 300 × 106 platelets/mL as counted with the Sysmex haematology analyser (KX-21N, Sysmex, Germany). This PRP was used for the measurement of platelet aggregation before and after incubation with different concentrations of alperujo extract, hydroxytyrosol, DHPG or 0.9% of saline as a control for 10 min at 37 °C. Platelet aggregation was then induced by adding collagen (final concentration 2–5 μg/mL) or thrombin receptor analogue peptide (TRAP, final concentration 25 μM) as agonists. The platelet aggregation measurement was performed 90 min after blood sampling using the PACKS 4 platelet aggregation chromogenic kinetic system.
Blood from healthy male and female volunteers with normal platelet function was collected in sodium citrate and diluted 1:10 with HEPES-Mg buffer, pH 7.4 (450 μL) and was incubated with alperujo extract (40 mg/L) or with the same volume of saline for 10 min. Aliquots of these mixtures were then incubated with or without ADP (final concentration, 10 μmol/L), or TRAP (final concentration, 25 μmol/L), or phorbol 12-myristate 13-acetate (PMA; final concentration, 1 μmol/L) for 10 min at room temperature. PMA was used as the positive control marker on the flow cytometer. The peridinin chlorophyll protein (PerCP)-labelled monoclonal anti-CD61 (BD Bioscience), allophycocyanin (APC)-labelled monoclonal antibody anti-CD62P (BD Phamingen) and the fluorescein isothiocyanate (FITC)–labelled monoclonal antibody anti-fibrinogen (DakoCytomation) (all 0.002 μg/μL blood), and FITC-, PerCP- and APC-labelled mouse immunoglobulin G antibodies (BD Biosciences; 0.004 μg/uL) were used as isotype controls. After 20 min of incubation in the dark at room temperature, all samples were quenched with 2 mL of cold PBS and counted using the FACSCalibur (Becton–Dickinson) with CellQuest software (version 2: BD Biosciences). Activated platelets were defined as the percentage of CD61-positive events co-expressing with one of the other two antibodies (CD62P and anti-fibrinogen) binding.
Preparation of treated platelets
Blood from healthy male and female volunteers with normal platelet function was collected in sodium citrate and centrifuged at 200g for 15 min at 37 °C. The top 2/3rd of the PRP was removed, and PGI2 was added (final concentration of ~50 ng/mL). Four millilitres of PRP was layered onto an albumin cushion of ~0.5 mL of 50% bovine serum albumin (fatty acid free, pH adjusted to 7.4) and warmed to 37 °C. Samples were spun at 1,600g for 15 min at room temperature. The platelet layers were taken off and re-suspended in warmed Tyrodes calcium-free buffer (containing 50 ng/mL PGI2) to obtain a platelet count of ~300 × 106 platelets/mL. Upon full degradation of the PGI2, half of the platelet suspensions (n = 5) were incubated with alperujo extract (40 mg/L), and the other half of the platelet suspensions (n = 5) were incubated with saline at room temperature for 15 min. Then, ADP was added (final concentration of 10 μmol/L) for 10 min at 37 °C in order to measure the effect of alperujo extract on the ADP-activated platelet proteome. ADP was used as a mild agonist to enable comparison with the outcomes of the platelet activation assays and to prevent platelets from clotting during the incubation time. Samples were then centrifuged at 8,700g for 10 min at 4 °C, and supernatants were removed. The remaining platelet pellet was washed with cold Tyrodes buffer, and samples were again centrifuged at 8,700g for 10 min at 4 °C, after which the Tyrodes buffer was removed.
Protein isolation from platelets
An extraction buffer (50 mM Tris, 100 mM KCL, 20% glycerol pH 7.1, add 0.1% Triton) was added to the platelet pellets, and samples were sonicated for 2 × 10 s on ice. Samples were then centrifuged at 8,700g for 10 min at 4 °C, after which the supernatant was carefully removed for protein analysis. Cytosolic platelet protein concentrations were measured by the bicinchoninic acid (BCA) protein assay (Sigma).
Two-dimensional gel electrophoresis
Two-dimensional gel electrophoresis was performed as described by us previously [26–30]. Briefly, 1 gel was run per incubation (total of n = 10 gels) using BioRad immobilized pH gradient (IPG) 24 cm strips (pH 4–7) for separation of the proteins in the first dimension. Gels were stained with Flamingo stain (BioRad) according to the manufacturer’s instructions. 2D-electrophoresis gels were analysed using PDQuest software (BioRad). Only protein spots that were present in all replicates were used for subsequent analysis. Spots with densities that significantly differed between treatment groups were excised from the SDS–PAGE gels using the BioRad ExQuest spot cutter, alkylated and trypsinized using a standard protocol on a MassPrep Station (Micromass), and analysed by electrospray LC mass spectrometric methods as described [26–30]. The total ion current (TIC) data were submitted for database searching using the MASCOT search engine (Matrix Science) using the MSDB database with the following search criteria: allowance of 0 or 1 missed cleavages; peptide mass tolerance of ±1 Da; fragment mass tolerance of ±0.8 Da, trypsin as digestion enzyme; carbamidomethyl modification of cysteine; methionine oxidation as partial modification; and charged state as MH+.
Data are presented as means ± SD. A two-sided t-test of two independent samples was carried out on protein spot density data (Genstat/E10.0, VSN International Ltd) with a p-value of p < 0.05 considered significant.
Effects of olive oil phenolics on ex vivo platelet aggregation
Effects of alperujo extract on ex vivo platelet activation
Effect of alperujo extract (40 mg/L) on ADP- and TRAP-induced platelet activation, measured as the percentage of activated platelets expressing CD62p or fibrinogen receptor
Activated platelets (%)
36.7 ± 14.1
44.6 ± 12.7
52.0 ± 14.6
59.2 ± 10.2
70.6 ± 13.3
76.9 ± 9.8
44.3 ± 14.1
48.6 ± 18.5
Effects of alperujo extract on the platelet proteome
Measures of confidence for protein identification and characterization by MS/MS analysis of platelet proteins which were significantly increased or decreased upon incubation with alperujo and ADP in activated platelets
Actin, cytoplasmic 2
EITALAPSTMK + Ox (M)
Rho GDP-dissociation inhibitor 1
TDYMVGSYGPR + Oxidation (M)
MLVVLLQANR + Ox (M)
NFATSLYSMIK + Ox (M)
MLVVLLQANR + Ox (M)
NFATSLYSMIK + Ox (M)
Platelet membrane glycoprotein IIb
Fibrinogen gamma chain precursor
Protein disulphide isomerase-related protein 5 (ERp5)
Serine/threonine protein phosphatase
Alperujo extract from olives is a new by-product from the modern two-phase processing technique used in olive oil production. Upon olive oil extraction, only a low percentage of the total phenolic compounds present in the olive fruits end up in the virgin olive oil; the majority of phenolic compounds (about 98%) remain in the liquid and solid olive-mill waste. Therefore, alperujo represents an extremely rich source of phenolic compounds , which potentially could have anti-platelet effects, comparable with those observed in the literature [32, 33]. We found that alperujo extract inhibited both human platelet activation and aggregation. At a concentration of 40 mg/L, alperujo extract significantly inhibited both ADP- and TRAP-induced platelet activation, measured as the percentage of activated platelets expressing P-selectin, which is released to the platelet surface during activation to mediate the rolling of the blood cells on the surface of the endothelium and the initial recruitment of leucocytes to the site of injury during inflammation . At this concentration, the inhibition of ADP- and TRAP-induced platelet activation measured as the percentage of platelets expressing the activated fibrinogen receptor by alperujo extract did not reach statistical significance (Table 1). At a concentration of 100 mg/L, alperujo extract significantly inhibited both collagen- and TRAP-induced platelet aggregation (Fig. 2). Inhibition of collagen- and TRAP-induced platelet aggregation was even more pronounced when exposing the platelets to alperujo extract at a concentration of 500 mg/L (28 and 16%, respectively). Our HPLC results suggested that 500 mg/L alperjuo extract contained approximately 40 mg/L hydroxytyrosol and 5 mg/L of DHPG, which on their own did not inhibit collagen- or TRAP-induced platelet aggregation, apart from DHPG in the highest concentration (Fig. 2). However, a mixture of 40 mg/L of hydroxytyrosol and 5 mg/L of DHPG did significantly inhibit collagen-induced platelet aggregation by 12% and TRAP-induced platelet aggregation by 16%. This suggests that an interaction between hydroxytyrosol and DHPG could be, for a large part, responsible for the anti-platelet effects of alperujo. In addition, other compounds contained in the alperujo extract could also be implicated in the effects observed. For example, 1.1% of the dry weight of alperujo extract was verbascoside, a conjugated glucoside of hydroxytyrosol and caffeic acid (Fig. 1). However, verbascoside is easily hydrolysed into hydroxytyrosol and a caffeoyl group, with hydroxytyrosol believed to be responsible for the activities of verbascoside . HPLC of the alperujo extract showed the presence of a range of more than 15 compounds that, considering their elution time and absorbance characteristics, are likely to have phenolic structures (Fig. 1). The strong effects observed on collagen-mediated aggregation in other studies support this suggestion, since phenolics can affect collagen-induced platelet TxA2 formation [32, 33]. Proteomics of platelets incubated with alperujo extract, compared with control, elucidated three important mechanisms that may underlie the anti-platelet effects of this extract: regulation of platelet structure and aggregation, coagulation and apoptosis, and signalling by integrin αIIb/β3.
Platelet structure and platelet aggregation
Alperujo extract caused a significant 1.5-fold increase in platelet protein γ-actin levels (Table 2). The actin-based cytoskeleton maintains the discoid shape of the human platelet using a cytoplasmic scaffolding composed of rigidly cross-linked actin filaments. During activation of platelets, however, the cytoskeleton also can get quickly taken apart and reassembled in a process influenced by platelet proteins and signalling pathways, which initiate cell shape changes mediated by the actin filament system of cells . Increased protein levels of γ-actin upon treatment with alperujo extract may promote cytoskeletal stability of the resting platelet and reduce the incidence of shape-change in response to external stimuli. Alperujo extract also increased protein levels of Rho GDP-dissociation inhibitor 1, a negative regulator of Rho family GTPases . Rho family GTPases regulate the formation of actin cytoskeletal structures and play important roles in cytoskeletal reorganization . Rho is, however, also involved in the regulation of thrombin-induced platelet aggregation, since treatment of platelets with Rho GDP-dissociation inhibitor 1 blocked the aggregation . Increased protein levels of Rho GDP-dissociation inhibitor 1 may therefore, at least partly, explain the effects seen on inhibition of platelet aggregation by alperujo extract in this study.
Coagulation and apoptosis
Signalling via integrin αIIb/β3 (fibrinogen receptor)
Protein levels of integrin αIIb were lower upon treatment with alperujo extract (Table 2, Fig. 3), and this could impair the formation of the full fibrinogen receptor, integrin αIIb/β3. During platelet activation, conversion of integrin αIIb/β3 to a high affinity conformation initiates platelet–platelet interaction through binding of fibrinogen and von Willibrand factor. This step leads to rapid platelet aggregation that physically plugs ruptured endothelial cell surface. This is an essential and common step in full platelet aggregation in response to all stimuli . In our study, we observed lower protein levels of integrin αIIb despite concomitant increased protein levels of the fibrinogen γ precursor upon treatment with alperujo extract. Fibrinogen is a soluble glycoprotein, consisting of a hexamer containing two sets of three different chains (α, β and γ), linked to each other by disulphide bonds. The main γ-chain form is γA, but a fraction of fibrinogen contains a differentially spliced γ chain called γ′, which presents itself in plasma mainly as a heterodimer with the common γA chain as γA/γ′ fibrinogen. Platelets can take up fibrinogen from plasma, but are also capable of synthesizing this compound through a translation process. Interestingly, fibrinogen stored in the platelet α-granules completely consists of γA/γA fibrinogen. The carboxy-terminal end of the γA chain is involved in binding to platelet integrin αIIb-β3, but the γ′ chain lacks the binding site required for platelet adhesion and therefore reduces thrombin-induced platelet aggregation . Our LC–MS/MS analyses did not enable us to distinguish between the γA or γ′ chain. Increased protein levels of the latter variant could imply an anti-platelet effect that would underpin the effects of alperujo extract on platelet aggregation and activation.
Protein levels of endoplasmic reticulum protein 5 (or ERp5) were also significantly increased by alperujo extract. ERp5, a thiolesterase that is present mainly on platelet intracellular membranes, is rapidly recruited to the platelet cell surface and associates with the β3 subunit of integrin αIIb/β3 upon platelet activation. Indeed, inhibition of ‘cell-surface’ ERp5 using inhibitory antibodies decreased integrin activation and granular secretion, albeit not sufficient to completely inhibit platelet aggregatory responses . It is currently unclear how and whether increased protein levels of ERp5 could contribute to anti-platelet effects of alperujo since it is not known whether our proteomics approach could differentiate between intracellular or cell-surface protein levels.
In conclusion, the phenolic compounds in alperujo extract may protect against platelet activation, platelet adhesion and possibly have anti-inflammatory properties, as evidenced by the lower percentage of activated platelets expressing P-selectin. However, these effects were instigated by relative high concentrations of phenolic compounds that will be difficult to obtain in vivo from olive oil consumption only. The beneficial effects could be accomplished by regulation of pathways involved in organization of platelet structure and aggregation, coagulation and apoptosis, and signalling by integrin αIIb/β3. This study provides additional evidence that specific phenolic compounds from a Mediterranean diet may improve platelet function. The extraction of these compounds, perhaps for future human consumption purposes, can be accomplished in a sustainable manner through effective use of a by-product from olive oil manufacturing process.
The laboratory of BdR is funded by Scottish Government Rural and Environment Research and Analysis Directorate (RERAD). We thank Dr. Juan Fernandez-Bolanos, Instituto de la Grasa, Seville, Spain, for providing us with the alperujo extract, hydroxytyrosol and 3,4-dihydroxyphenylglycol.