What can we learn from the platelet lipidome?

Abstract

Besides their proteome, platelets use, in all responses to the environmental cues, a huge and diverse family of hydrophobic and amphipathic small molecules involved in structural, metabolic and signaling functions; the lipids. Studying how platelet lipidome changes modulate platelet function is an old story constantly renewed through the impressive technical advances allowing the discovery of new lipids, functions and metabolic pathways. Technical progress in analytical lipidomic profiling by top-of-the-line approaches such as nuclear magnetic resonance and gas chromatography or liquid chromatography coupled to mass spectrometry enables either large-scale analysis of lipids or targeted lipidomics. With the support of bioinformatics tools and databases, it is now possible to investigate thousands of lipids over a concentration range of several orders of magnitude. The lipidomic landscape of platelets is considered a treasure trove, not only able to expand our knowledge of platelet biology and pathologies but also to bring diagnostic and therapeutic opportunities. The aim of this commentary article is to summarize the advances in the field and to highlight what lipidomics can tell us about platelet biology and pathophysiology.

Plain Language Summary

What is the context?

• Lipids are a huge and diverse family of molecules strongly involved in biological membranes organization and dynamics, signal transduction, cell metabolism and intercellular communication.

• Earlier seminal works using conventional lipid biochemistry methods have shown the essential role of certain classes of lipids in platelet biology and platelet-related pathologies

What is new?

• The important development of modern lipidomic analyses using mass-spectrometry now provides opportunities to investigate the entire platelet lipidome in different conditions.

• The application of lipidomic approaches to analyze large-scale lipid species allows platelet clinical lipidomics development.

What is the impact?

• Study of the lipidomic landscape of platelets will expand our knowledge of platelet biology and should bring new diagnosis and therapeutic opportunities.

• Evaluating the functional and clinical significance of the data generated by modern platelet lipidomics appears as a vast and exciting challenge.

Keywords:

• Lipidomics
• platelet

Introduction

Platelets are highly reactive to extracellular cues through activation of specific membrane receptors allowing platelet activation, adhesion, secretion, and aggregation. The lipid composition of platelets is crucial for their integrity and adequate responses. Besides forming cell/organelle membranes and modulating their physical properties, lipids act as key intra- and extracellular messengers in signal transduction and cell communication. Following activation, platelets produce and release bioactive lipids with important auto- and paracrine actions. The intracellular functions of lipids include anchoring proteins to membranes, regulating intracellular trafficking and providing energy for cell metabolism. A variety of enzymes such as phospholipases, kinases and phosphatases, synthetases, ligases, oxidases, reductases and transporters tightly control lipid metabolism and functions. Integrating lipidomics into our global view of platelet biology and other omics analysis is of major importance to expand our knowledge and develop precision medicine. Lipidomics is a growing field which has the ambition to describe and quantify all lipid molecular species and to understand their functions. The impressive development of analytical lipidomics methods endorsed by the scientific community and the sharing of international resources such as LIPID MAPS or Lipidomics Standard Initiative (LSI) allow proper structural identification and quantification of lipids [1,2]. Lipids have a large diversity of structures so that a variety of strategic approaches have been developed to analyze them [2,3]. Eight categories of lipids are referenced; fatty acyls, glycerolipids, glycerophospholipids, sphingolipids, sterol lipids, prenol lipids, saccharolipids and polyketides [4]. Several of these lipid categories, especially complex lipids, such as glycerophospholipids and sphingolipids, have a number of molecular species, which further complexify their absolute quantification. More than 10 000 species of glycerophospholipids and 4900 species of sphingolipids are listed in the LIPID MAPS website (https://www.lipidmaps.org/). The lengths, degrees of unsaturation and location of double bound of aliphatic chains of lipids contribute to this variety. They also support the distinct physical and chemical properties of lipids which are major determinants of their function. The different steps of the bioanalysis process of lipids require caution, from their extraction procedure to their storage and analysis. Indeed, unsaturated lipids are prone to oxidation/degradation. Mass spectrometry (MS) is now a gold standard to identify and quantify lipids but the absolute quantification by MS is not always possible, for instance when appropriate standards are not available [3]. In this case, a relative quantification is carried out. The LSI guidelines for lipid analysis and quantification (https://lipidomics-standards-initiative.org/guidelines) are very helpful to standardize lipidomics methodologies and protocols [5]. Nowadays, a variety of strategic lipidomics approaches using nuclear magnetic resonance and gas chromatography (GC) or liquid chromatography (LC) coupled to MS are available and facilitate the analysis of the lipidome, including low-abundance species. Besides in-depth analysis of well-known lipid-based mechanisms of platelet activation such as phospholipase A2 (PLA2) and phospholipase C (PLC) pathways, identification of novel biosynthesis pathways, discovery of new lipids and functions and large-scale analysis of the complete platelet lipidome represent exciting challenges. Moreover, clinical lipidomics applied to platelets may bring important innovations in the field of cardiovascular diseases and other pathologies involving platelets [6].

Platelet lipidomics for the discovery of new lipids and lipid-based mechanisms

The essential implication of lipids in platelet biology has been highlighted by earlier seminal works and include PLC, phospholipase D (PLD) and phosphoinositide 3-kinases pathways, PLA2 activation and subsequent production of eicosanoids, generation of lysophospholipids, activation of the sphingolipid metabolism, organization of lipid rafts or externalization of phosphatidylserine (PS) (Figure 1) [7–9]. These lipids have either intracellular signaling functions that contribute to platelet secretion, adhesion and aggregation or act as extracellular mediators further activating platelets and influencing hemostasis as well as inflammation, wound healing and other biological processes involving platelets. Most of our current knowledge of platelet lipids is coming from experiments using conventional techniques including thin layer chromatography (TLC), gas chromatography (GC), high-pressure liquid chromatography (HPLC) coupled or not to radiolabelling approaches. These specific lipid biochemistry approaches brought us fundamental understanding of the role of lipids in vascular biology. A striking example is the discovery of prostaglandins and other eicosanoids derived from arachidonic acid, which was awarded the Nobel Prize to S.K. Bergström, B.I. Samuelsson and J.R. Vane in 1982. However, modern lipidomics is heavily relying on technologies for the analysis of the structure and abundance of lipids. For about 20 years, the development of highly sensitive mass spectrometers coupled to LC or GC together with appropriate bioinformatics assistance has changed our approaches of lipid analysis in general and may have a strong impact on our knowledge of platelet biology in the future.

Figure 1. Lipid-dependent pathways in the regulation of platelet functions. The figure highlights the major lipid-mediated mechanisms described so far in blood platelets. The different colors of the boxes point out different functions or lipid families. Thromboxane A2 is a potent proaggregatory and vasoconstrictor, platelet-activating factor can stimulate platelet aggregation and activate immune cells, lysophosphatidic acid can stimulate platelet aggregation and induce cell proliferation, several eicosanoids are potent regulators of inflammation, sphingosine-1-phosphate is a pleiotropic lipid mediator involved in a large array of processes including inflammation, immunity, endothelial barrier integrity or tumorigenesis, phosphoinositides are key actors in intracellular signal transduction leading to platelet secretion, adhesion and aggregation, phosphatidic acid is involved in intracellular signal transduction [7–9]. PS, phosphatidylserine; IP3, inositol trisphosphate; DAG, diacylglycerol; T×A2, thromboxane A2; PI(3,4,5)P₃, phosphatidylinositol 1,4,5 trisphosphate; PI3P, phosphatidylinositol 3 monophosphate.

Nowadays, several techniques are available, including different mass spectrometer detectors in ultra-high resolution (FTICR Fourier-transform ion cyclotron resonance), high resolution (Orbitrap and Time of Flight QTOF) and low resolution (Quadrupole and ion trap). Hybrid instruments combining these different detectors offer multiple ionization and fragmentation modes. These systems can be used by injecting a lipid extract directly (Shotgun) or after chromatographic separation with GC, HPLC or supercritical fluid (SFC). These systems are now miniaturized and reach a very high sensitivity. They can be complemented by ion mobility which can be placed before the ionization, or in the mass detector at different positions. This fourth dimension allows the separation of certain isomers of lipids. Finally, these detectors can be used for the spatial localization of lipids on tissue slices by means of matrix-assisted laser desorption (MALDI), electrospray desolvation (DESI), or with a focused primary ion beam (SIMS).

Global high throughput methods can be obtained by direct introduction of the lipid extract in high-resolution mass spectrometer system (shotgun Lipidomics), or after chromatographic separation. These methods allow profiling of series of complex mixtures in a relatively short time. This untargeted mode of lipidomics has several advantages when analyzing populations of patients for instance, because it allows identification of unanticipated changes in lipid molecular species. Such an unbiased approach can also contribute to the identification of new lipids. Besides these great advantages, the global high throughput methods are generally not quantitative and can miss low abundance lipids. In 2016, a first global study identified more than 5000 putative lipid species in resting human platelets with changes in about 900 lipids following thrombin stimulation [10,11]. Interestingly, this study pointed to an unanticipated lipidomic circuit providing, through cPLA2, the substrates for mitochondrial energy generation during platelet activation.

Besides global methods, targeted lipidomics by LC/MS/MS is generally highly sensitive and can allow accurate quantification and in-depth study of a given platelet family of lipids, such as phosphoinositides, eicosanoids or sphingolipids. Specific platelet lipid families such as phosphoinositides have been recently analyzed by LC/MS/MS in resting and activated mice and human platelets [12]. Phosphoinositides are important lipids highly involved in signal transduction and in intracellular membrane trafficking in all eukaryotic cells [13], including platelets [14,15]. Compared to conventional methods (TLC and HPLC), there are several advantages to analyze phosphoinositides by LC/MS/MS including the non-requirement of radioisotopes (³²Pi or H³-inositol) and the opportunity to detect the molecular species of the different phosphoinositides. In our study [12], several molecular species (fatty acyl chain composition) of phosphatidylinositol (PI), phosphatidylinositol monophosphate (PIP), phosphatidylinositol bisphosphate (PIP₂) and phosphatidylinositol trisphosphate (PIP₃) were identified. An important increase of low-abundance molecular species of PIP₂ was observed following stimulation by thrombin or collagen-related peptide, suggesting a possible function for these lipids [12]. This phosphoinositide profiling also revealed similarities and differences between mouse and human platelets. For instance, following thrombin stimulation, some molecular specie of PIP₂ dramatically increased in mouse platelets and only weakly in human platelets. Moreover, thrombin was much more efficient than collagen-related peptide to stimulate PIP₃ production in human platelets while it was the opposite in mouse platelets. Very recent LC/MS/MS analytical developments through chiral separation [16,17], allow even more in-depth analysis by separating the different regioisomers of PIP (PI3P, PI4P, PI5P) and PIP₂ (PI(3,4)P₂, PI(3,5)P₂, PI(4,5)P₂) which was not possible until now (Figure 2). These new methods, sensitive enough to allow the analysis of low-abundance molecular species of all phosphoinositides, open exciting perspectives in the field.

Figure 2. Detection of the molecular species of phosphatidylinositol monophosphate isomers by a LC/MS/MS method. The figure illustrates the separation of phosphatidylinositol monophosphate regioisomers (zoom of the total ionic current of an equimolar mixture of PI3P, PI4P and PI5P), using a chiral column as described recently [14,15]. The detection of the molecular species of the different phosphatidylinositol monophosphates (fatty acyl composition: C38:4>C38:3>other species) can be achieved by LC/MS/MS as schematically shown in the figure. A similar analysis can be performed for phosphatidylinositol bisphosphate regioisomers (PI(3,4)P₂, PI(3,5)P₂ and PI(4,5)P₂).

The combination of technologies can improve the identification and quantification of lipids. For instance, a recent study integrating shotgun analysis and targeted lipidomics performed on human and mouse platelets has provided a first quantitative analysis of mouse platelet lipidome [18,19]. This study identified about 400 lipid species, 15 being the most abundant ones and less than 20% had an alteration of their abundance following thrombin stimulation. Interestingly, using a mouse model of sphingomyelin phosphodiesterase deficiency, the authors observed a correlation between an increase in lysosphingomyelin concentration and platelet response inhibition. Although the mechanisms involved remain unclear this study sheds light on a novel mechanism regulating platelet function. In platelets, the sphingolipids metabolism is highly active. For instance, sphingosine-1-phosphate, generated by sphingosine kinase, is released by activated platelets and acts as an extracellular mediator stimulating specific G-protein coupled receptors [20].

Another important pathway in platelets involves PLA2. Among the fatty acyls generated by PLA2, arachidonic acid is a major precursor for oxidative transformation to eicosanoids by lipoxygenase and cyclooxygenase pathways. Following stimulation, platelets produce a number of eicosanoids, including thromboxane A2 (TXA2), a proaggregatory lipid. T×A2 released by activated platelets acts as an auto- and paracrine molecule through the G-protein coupled receptor TP-receptor [7,8]. The calcium-sensitive cytosolic PLA2α isoform (cPLA2α) has a major role in human platelets as demonstrated in a rare inherited human deficiency of this enzyme leading to impaired platelet eicosanoid generation associated with platelet dysfunction [21]. A platelet lipidomic profiling study has highlighted the importance of PLA2α (cPLA2α) in platelet activation and clarified the substrates of this PLA2 isoform [22]. Moreover, the lysophospholipids generated by cPLA2 were identified but their role in platelet remains poorly understood. This study also suggested the involvement of cPLA2 in the production of a unique eicosanoid downstream of PAR4 and GPVI triggering.

Besides eicosanoids generation through oxidation of arachidonic acid, lipidomic studies using tandem MS mode have provided new data on the oxidized phospholipids (i.e. phospholipid-esterified eicosanoids) generated by platelets during activation. Interestingly, these lipids contribute to the platelet procoagulant activity by helping clotting factors in plasma to work more efficiently [7,23,24]. Thus, specific lipidomic approaches point to oxidized phospholipids generated by activated platelets as important players in hemostasis and thrombosis.

These examples show how investigation of the platelet lipidome by appropriate top of the line approaches and rigorous methodologies can bring new layers of information and unexpected discoveries to increment our understanding of platelet biology. Lipidomics can also be applied to platelet organelles after isolation as well as to platelet-generated extracellular vesicles (microparticles and exosomes) which may provide important information.

Platelet clinical lipidomics

The application of lipidomic approaches to measure large-scale lipid species is now extending into the clinical field, as illustrated by the term “clinical lipidomics” which aims to quantify the lipidome of cells, biopsies or body fluids from patients and link it to clinical data, genomics and proteomics [25]. This discipline is expected to allow identification of diagnostic/prognostic biomarkers and to help the monitoring of therapy. The clinical importance of platelets is well-recognized in cardiovascular diseases and in bleeding diathesis. It is also strongly emerging in immune, inflammatory and likely malignant diseases. Clinical lipidomics of platelets is therefore expected to provide new insights in various pathologies. However, it is important to keep in mind that there is a certain degree of heterogeneity of the platelet lipidome in healthy individuals [10,11]. The platelet lipid profile is in part dynamic and variables like age, gender and diet may contribute to this interindividual heterogeneity. Nevertheless, the platelet lipidome has been reported to vary in different diseases such as liver disease [26], hyperlipoproteinemia and dyslipoproteinemias [27,28], arterial hypertension [29], cancer [30], and following dietary supplement of n-3 polyunsaturated fat [31].

However, whether specific platelet lipidomic profiles may become diseases signature remains to be established. Moreover, how altered platelet lipidome can be involved in platelet-dependent pathologies is still poorly understood. Yet, some inherited pathologies are directly linked to platelet lipid modifications. This includes the loss of PS exposure due to TMEM16F scramblase mutations impairing the procoagulant function of platelets in the bleeding Scott syndrome [32], the rare inherited cPLA2 deficiency leading to platelet dysfunction [21], the decreased expression of PLCβ2 linked to hypo-responsive platelets [33], or the loss of function of the PI(4,5)P₂ 5-phosphatase affecting platelet responses in the LOWE syndrome [34]. Some mouse models have also pointed to new important lipid-related pathways that should stimulate to prioritize a lipidomic analysis in patients. For instance, as mentioned above, mice lacking sphingomyelin phosphodiesterase 1 have platelet dysfunction suggesting that human platelets deficient in this enzyme, such as platelets from Niemann-Pick patients, may be affected. A mouse model of sitosterolemia caused by a mutation in the ABCG5 or ABCG8 transporter genes has linked the accumulation of free plant sterols in platelet membranes to dysregulation of platelet functions leading to macrothrombocytopenia and bleeding [35]. Lipidomic studies on platelets from sitosterolemia patients should bring interesting information as well as potential diagnostic and disease monitoring markers.

Moreover, it could be of interest to investigate whether other inherited platelet pathologies, not directly linked to lipid metabolism, induce platelet lipidome changes as this may unravel new regulatory pathways.

Acquired platelet dysfunction can also be linked to lipidome changes. The cyclooxygenase inhibitor aspirin, a widely used antithrombotic agent, is well known to profoundly alters the production of platelet eicosanoids [7]. Moreover, treatment with the antagonist of the P2Y12 ADP receptor, ticagrelor, has been shown to modify the lipid composition of the platelet plasma membrane and the lipid rafts [36]. Infectious diseases can also change the platelet lipidome. For instance, COVID-19 has been shown to alter the platelet lipidome of patients correlated with changes in platelet reactivity, suggesting that other infectious diseases may also impact the platelet lipidome [37].

Importantly, a targeted lipidomic study has shown an increase in ceramides, di- and triacylglycerol, sphingomyelin, lysophosphatidylcholine, acylcarnitines and non-enzymatic oxidized lipids in platelets from patients with symptomatic coronary artery disease [38,39]. This study shows that patients platelets can uptake oxidized LDL increasing the oxidation of other platelet lipids. More recently, significant changes in the platelet lipidome were observed between acute coronary syndrome and chronic coronary syndrome patients [40]. How this specific platelet lipidome signature impacts on platelet function and on the pathological state remains an open question. One possibility is that platelets engage transcellular lipid changes, either by releasing lipids or through generation of extracellular vesicles which level is frequently elevated in cardiovascular diseases, and may contribute to modulate the pathological states [41].

Recently, we had the opportunity to evaluate the platelet lipidomic fingerprint in obesity. A targeted lipidomic analysis was used to compare resting platelets from twelve non-diabetic severely obese patients (BMI>40 kg/m²) and their age- and gender-matched lean controls [42]. A significant 10% decrease in free cholesterol, phosphatidylcholine (PC) and phosphatidylethanolamine (PE) as well as a trend in PI and SM decrease was observed in platelets from the obese group compared to the control group. This decrease may be due to the release of these lipids from platelet membranes as extracellular vesicles which have been shown to display a distinctive proteome profile in obese individuals [43].

Another interesting aspect is the generation of still poorly known platelet lipids that may control hemostasis and thrombosis. For instance, oxidized phospholipids including PE, PI and PC-esterified 12S-hydroxyeicosatetraenoic acid (12S-HETE) or the anionic phospholipid, phosphatidylthreonine, closely related to PS, are produced by activated platelets [23,24]. Although present in relatively small amounts, these lipids appear to contribute to the platelet procoagulant activity. Their presence in extracellular vesicles may amplify coagulation and contribute to intercellular communications.

The storage of platelets and other blood products is an important aspect in transfusion medicine. Bioactive lipid generated during the storage of platelets, and other blood products, may have functional implications once these products are transfused. Platelets modify their lipidome during storage and release bioactive lipids as well as extracellular vesicles containing these lipids [44–47]. Mapping the platelet lipidome during storage has the potential to bring important information to further improve the quality and security of blood products.

Finally, other exciting developments include lipidomics and MALDI-MS imaging (or spatial lipidomics) [48] of thrombus recovered by endovascular thrombectomy from stroke patients. Lipid signatures may represent unique reporters of platelet or leukocyte involvement and activation. Correlating clinical data to the composition of the clots analyzed by histochemistry and lipidomics may provide unprecedented insight into the origin of the clot and help clinicians to establish personalized therapies.

Despite these promises, to fulfill expectations, the platelet clinical lipidomics field will have to face important challenges, such as providing reference values, identifying and validating diagnostic/prognostic biomarkers and lipid signatures in well-characterized patient cohorts and demonstrating the clinical utility of lipidomic profiling [6]. The new advances in technology and bioinformatics tools should help to reach these goals in the near future.

Conclusion

With the development of high resolution, high sensitivity and rapid scanning instruments, the lipidomics field offers now a variety of approaches to investigate the many, yet unexplored, facets of the platelet lipidome. Scientific hypothesis derived from studies on lipid-related pathways, platelet diseases linked to genetic defects on lipid metabolizing enzymes or transporters and cardiovascular diseases such as coronary artery disease should help to prioritize platelet lipidomics. No doubts that large-scale analysis of the complete platelet lipidome and targeted lipidomics in different situations will continue to shed light on novel lipids and mechanisms important for platelet function. The platelet lipidome holds great potential to provide biomarkers that may have major implications in diagnostic/prognostic, and follow-up and treatment of diseases. Emerging approaches using stable isotope labeling strategies will allow analysis of dynamic changes in lipid metabolism and will open perspectives to discover novel platelet metabolic pathways. Evaluation of the functional and clinical significance of the information generated by modern platelet lipidomics will be a vast and exciting challenge in the near future.

Acknowledgments

The authors thank all members of the B.P. team for helpful discussions.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study was supported by research funding from Inserm and Fondation pour la Recherche Médicale (DEQ202203014742).