Fatty acid-related modulations of membrane fluidity in cells: detection and implications

Abstract

Metabolic homeostasis of fatty acids is complex and well-regulated in all organisms. The biosynthesis of saturated fatty acids (SFA) in mammals provides substrates for β-oxidation and ATP production. Monounsaturated fatty acids (MUFA) are products of desaturases that introduce a methylene group in cis geometry in SFA. Polyunsaturated fatty acids (n-6 and n-3 PUFA) are products of elongation and desaturation of the essential linoleic acid and α-linolenic acid, respectively. The liver processes dietary fatty acids and exports them in lipoproteins for distribution and storage in peripheral tissues. The three types of fatty acids are integrated in membrane phospholipids and determine their biophysical properties and functions. This study was aimed at investigating effects of fatty acids on membrane biophysical properties under varying nutritional and pathological conditions, by integrating lipidomic analysis of membrane phospholipids with functional two-photon microscopy (fTPM) of cellular membranes. This approach was applied to two case studies: first, pancreatic beta-cells, to investigate hormetic and detrimental effects of lipids. Second, red blood cells extracted from a genetic mouse model defective in lipoproteins, to understand the role of lipids in hepatic diseases and metabolic syndrome and their effect on circulating cells.

Keywords:

• Membrane fluidity
• functional two photon microscopy
• lipidomic analysis
• saturated fatty acids
• lipoproteins

Introduction

Fatty acids are categorized by their length (number of carbon atoms) and degree of unsaturation (number of double bonds). Saturated linear-chain fatty acids (SFA) are synthesized from acetyl-CoA building blocks, throughout the action of acetyl-CoA carboxylase, fatty acid synthase (FAS), and elongases. Other enzymes, such as Δ9-desaturase (stearoyl-CoA desaturase-1, SCD1) or Δ6-desaturase (D6D), are involved in the transformation of SFA to monounsaturated fatty acids (MUFA). Unlike SFA and MUFA that are available from the diet or through de novo biosynthesis, polyunsaturated fatty acids (PUFA) are formed from the essential fatty acids linoleic acid (18:2n-6) and α-linolenic acid (18:3n-3) by a series of elongation and desaturation reactions mediated by elongases (Elovl2, Elovl4, Elovl5) and desaturases (D5D, D6D) [1–3]. SFA, MUFA, and PUFA are important sources for ATP production via β-oxidation, for energy storage in the form of triglycerides, as precursors for phospholipid biosynthesis and for the generation of numerous ligands for different types of receptors and ligand-activated transcription factors [4–10]. The composition of fatty acids and the nature of the polar head in phospholipids vary greatly among different membranes and even in the same membrane under different metabolic conditions [10–12]. PUFA (such as arachidonic acid, EPA, and DHA), once detached from membrane phospholipids, are transformed enzymatically by cyclooxygenases, lipoxygenases and members of the cytochrome P450 family into numerous biologically active metabolites, such as, prostaglandins, thromboxanes, lipoxins, and eicosanoids. Moreover, these PUFA and their hydroperoxy metabolites can be non-enzymatically converted by radical-induced peroxidation to bioactive mediators such as hydroxyalkenals [4,13].

The analysis of the lipid composition of tissues, cells, and organelles is necessary, but not sufficient, to address functional impairments of living organisms. Aside from the complexity of chemically analyzing numerous molecules [14], the extraction of correlations to meaningful biological interactions is not trivial, because lipids may exhibit diverse functions in different cell types, that greatly enhance inter-individual variability [15]. Biophysical methods can be useful in clarifying the role of lipids in cell physiology and disease processes in a comprehensive manner. These methods address biochemical and biophysical properties of lipids, such as phase behavior and submicrometric physical state of membranes. In cells, membrane lipids can interchange between fluid and gel-like phases, where each lipid molecules contributes to its own spatial arrangement and motional freedom. A broad term commonly adopted to describe the physical state of biological membranes is fluidity, which depends on membrane structure, curvature, microviscosity as well as its phase, lipid structure, packing, and composition. Artificial membrane bilayers composed by a single type of fatty acid can exist in liquid (fluid) or gel-like states depending on the temperature. At low temperatures, fatty acid moieties of phospholipids are laterally ordered and well-packed together in the membrane, in a gel-like phase (L_β). Raising temperatures over a transition value (melting temperature, T_m), which also depends on the kind of fatty acids constituting the membrane, induces the transition to a fluid state (L_α), which is characterized by a larger area occupied by polar heads of phospholipids. Lipid mixtures that mimic the outer leaflet of the plasma membrane (PM) display more complex phase behavior, with multiple phases and a mosaic of gel and fluid domains coexisting in the plane of the bilayer [16].

Consequently, biomembranes are characterized by several phases that overall reflect changes in the lipidome as imposed by dietary, environmental, and cellular conditions. Moreover, membrane lipids (i.e. phospholipids, cholesterol, phosphosphingolipid, sphingolipids, gangliosides [16]) compartmentalize and coalesce in organelle membranes, bilayer leaflets and in physically segregated domains within a leaflet. The maintenance of physiological cell membrane fluidity is a prerequisite for proper membrane function, and associated with cell viability and normal cell growth and division. For example, membrane fluidity and phase separation have a pivotal role in the potency of ligand binding to membrane receptors, on direct cell–cell interaction and on the modulation of the activity of membrane enzymes, receptors, channels, and transporters [16]. Membrane fluidity alterations have been indeed described in various pathologies, such as thrombocythemia, hyperlipidemia, hypercholesterolemia, hypertension, diabetes mellitus, obesity, sepsis, allergies, and burn injuries, in cases of alcohol abuse, in Alzheimer's disease and in schizophrenia [17–19]. In addition to the effects of lipids as signaling molecules (e.g. PiP3, sphingosine-1-phosphate), or modulators of gene expression (e.g. via SRBEPs), altered lipid phase behavior in cell membranes may cause functional impairments and onset of several pathologies [16]. Figure 1 summarizes the principal factors affecting membrane physical state.

Figure 1. Factors inducing membrane fluidity changes. Nutrients and environmental factors affect membrane fluidity by altering: (1) temperature and/or pressure, (2) lipid and protein composition, and by inducing (3) protein and lipid modifications. Regulation and homeostasis of membrane fluidity are obtained mainly by varying lipid composition through enzymatic action.

Last generation fluorescence imaging-based detection methods, such as functional two-photon microscopy (fTPM), can detect impairments in subcellular organelles with several advantages that include high specificity, high resolution, fine subcellular localization and temporal detection of changes. Combined with other independent assays, the real-time detection of alterations in lipid composition and modified cell signaling has become feasible. The fluorescence spectra of the probe Laurdan, which incorporates into the lipid phase in membrane, is correlated to its physical state (Figure 2) [20]. Laurdan's excited-state relaxation, independent of the head-group type in phospholipids, is highly sensitive to the presence and mobility of water molecules within the membrane bilayer [21], yielding information on membrane fluidity by a shift in its emission spectrum depending on the surrounding lipid phase state (i.e. bluish in ordered, gel phases and greenish in disordered, liquid-crystalline phases). Two-photon infrared excitation techniques have been successfully applied to detect Laurdan emission [22]. By using this probe, coexisting lipid domains were characterized on the basis of their distinctive fluorescence spectra and dual-wavelength ratio measurements [23–25], which map changes in the structure of PM [19,22,26–28].

Figure 2. Laurdan probe monitors membrane fluidity in cells by a shift in emission spectrum. Laurdan intensity images of INS-E1 islets were recorded simultaneously with emission in the range of 400–460 nm (I_G) and 470–530 nm (I_R). The Laurdan emission spectrum is reported in the graph (diamonds). Membrane fluidity can be measured in terms of ratio of emission intensities by using Generalized Polarization (GP) value, defined as GP = (I_G−I_R)/(I_G+I_R). The GP goes from −1 (very fluid) to 1 (very gel-like) (2) and it is calculated for each pixel to obtain the fluidity map. Scale bar is 10 μm.

In this work, we examined fatty acid-related membrane remodeling capacity by using the Laurdan probe and fTPM in two cases: first, pancreatic beta-cells exposed to increasing glucose and/or fatty acid concentrations. Second, red blood cells (RBC) isolated from a genetic mouse model defective in lipoproteins and exhibiting hepatic dysfunction and features of the metabolic syndrome.

Materials and methods

Materials

Normal- and glucose-free RPMI 1640 medium, fetal calf serum (FCS) and l-glutamine were from Biological Industries (Kibbutz Beit-Haemek, Israel). Sigma-Aldrich (Sigma, St. Louis, MO and Rehovot, Israel) supplied fatty acid free-BSA (bovine serum albumin, fraction V) and dimethyl sulfoxide (DMSO). Laurdan was purchased from Molecular Probes, Inc. (Eugene, OR).

Cell cultures

INS-1E cells were transfected with IAPP-mCherry protein targeted to insulin granules (IG) (courtesy of Dr. Patrick E. MacDonald, University of Alberta, Edmonton, Canada), as described [29]. Cells were plated at a density of 10⁵ cells/cm² in small IBIDI plates, in RPMI 1640 medium containing 11 mM glucose. After one day, the medium was changed to media supplemented with 5, 11, or 25 mM glucose. Cells were then loaded with Laurdan and emission images were acquired at zero time (when the cells were fed with the different glucose concentrations), 24- and 48 h. Other plates that were incubated for 32 h at the same glucose concentrations (5, 11, and 25 mM glucose) received 0, 200, 300, and 500 μM, The incubation medium with palmitic acid (PA) was serum free supplemented with 0.5% (w/v) fatty acid-free BSA PA during the last 16 h of incubation. In addition, other cells incubated for 48 h at 5, 11, or 25 mM glucose for 48 h received the different PA concentrations for 3 h only.

Animal model

Lcat−/− mice (courtesy of Prof. Silvia Santamarina-Fojo) and Apoa1−/− and C57BL/6 mice (both from Jackson Laboratory, Bar Harbor, ME) were bred in the animal facility of the University of Patras Medical School. Male mice of 10–16 weeks of age were used for experiments. Mice were allowed unrestricted access to food and water under a 12 h light/dark cycle. Animal studies were governed by the EU guidelines of the Protocol for the Protection and Welfare of Animals and conducted in accordance with the Declaration of Helsinki and authorized by the appropriate committee of the Laboratory Animal Center of the University of Patras Medical School and the Veterinary Authority of the Prefecture of Western Greece.

For the isolation of RBC, blood samples were collected into EDTA collection tubes. RBCs were isolated following blood centrifugation at 4000 rpm for 10 min.

Laurdan two-photon microscopy

A 1 mM stock of Laurdan was prepared in DMSO in dimethyl sulfoxide. Cells (pancreatic beta cells or RBC) were incubated with Laurdan in a 5% CO₂ incubator for 30 min in the dark. Afterwards, the plates were placed on the inverted confocal microscope (DMIRE2, Leica Microsystems, Wetzlar, Germany) and Laurdan intensity images were obtained using a 60× objective (NA 1.4) under excitation at 800 nm with a mode-locked Titanium-Sapphire laser (Chameleon, Coherent, Santa Clara, CA). Internal photon multiplier tubes collected images in an eight bit, unsigned images at a 400 Hz scan speed.

Laurdan intensity images were recorded simultaneously with emission in the range of 400–460 nm and 470–530 nm and imaging was performed at room temperature. For each specimen, 4–6 field images were analyzed with the GPicture software [24].

This program measures membrane fluidity in terms of ratio of emission intensities by using Generalized Polarization (GP) value, defined as (1)

GP value was calculated for each pixel using the two Laurdan intensity images [I(400–460) and I(470–530)]. The calibration factor G was obtained from the GP values of solutions of Laurdan in DMSO. GP images (as eight-bit unsigned images) were pseudocolored in Image-J. GP histograms values were determined within multiple Regions of Interest (PM of single cells) for each sample, and their mean ± SD (n = 15–20) determined and utilized for further statistical analysis (two-tailed Student’s t-test). Line profiles and analysis of acquired images were performed with image J.

Results

Beta cells: effects of nutrient overload

Altered lipid metabolism, transport, and storage are considered risk factors in the etiology of the metabolic syndrome. Thus, amelioration of dyslipidemia is considered a major therapeutic target for the reduction of risk of developing insulin resistance, type 2 diabetes, and related cardiovascular complications [30–32]. Numerous studies have shown that the combination of hyperlipidemia and hyperglycemia is highly toxic to cells. The term “glucolipotoxicity” was coined to describe how nutrient overload deteriorates the function and viability of pancreatic beta cells, induces insulin resistance in skeletal and cardiac muscles, alters hepatic functions and aggravates diabetes-associated complications [33–36]. In this context, we used as a model system the pancreatic beta INS-1E cell line, which was derived from rat insulinoma. This cell line has been widely investigated due to glucose-stimulated insulin secretion that resembles normal beta cells [37].

Free fatty acids (FFA), which are provided to pancreatic beta-cells from the diet and de novo synthesis, are not only used for the synthesis of phospholipids and triglycerides, but also serve as important regulators of insulin secretion. Normally, increased levels of FFA augment insulin secretion from beta cells. The latter interacts with hepatic insulin receptors to augment triglyceride biosynthesis and incorporation into lipoproteins. In the adipose tissue insulin facilitates fatty acid storage in triglycerides [31,32]. These physiological responses are compromised when beta cells are chronically exposed to high levels of FFA and glucose that induce insulin resistance in peripheral tissue and beta cell dysfunction [6,33–35]. These abnormalities are accompanied with alterations in membrane fluidity, palmitoylation of proteins, generation of ceramides, mitochondrial dysfunction, ER-stress, autophagy, and apoptosis [36].

Many studies have shown that nutrient overload affects the structure, function, and viability of beta cells [6,38–41]. We mimicked nutrient overload conditions in these cells in vitro by exposing them to increasing concentrations of glucose and PA. We have reported before that non-toxic levels of these two nutrients augment glucose-stimulated insulin secretion up to a threshold when each, or the combination of both, becomes detrimental to the cells [10]. INS-1E cells expressing the red fluorescent m-cherry protein in IG were similarly treated in the present study and effects of high levels of glucose and PA on the fluidity states of PM and membrane of IG were determined with Laurdan using fTPM, as described above. The cells were cultured at increasing glucose levels (5, 11, and 25 mM glucose) for 48 h without or with 500 μM of PA during the last 16 h of incubation. For this purpose, we first labeled IG in the cells with the red fluorescent m-cherry protein, targeted to IG, as described above. Following the incubation with glucose and/or PA, the cells were loaded with Laurdan and the experimental confocal acquisition conditions were set for simultaneous two-photon excitations of the two probes. This allowed for the measurement of membrane fluidity in single IG and determining the biophysical properties of PM and IG membranes. Figure 2 depicts Laurdan intensity images of insulin secreting INS-1E cells in culture that were excited with a two-photon laser and recorded simultaneously at emission ranges of 400–460 (I_G) and 470–530 nm (I_R). The corresponding Laurdan emission spectrum is reported in the graph (blue diamonds). Membrane fluidity is expressed in terms of ratio of emission intensities by using GP value, as described above. The GP ranges from −1 (very fluid-purple) to +1 (very gel-like-orange) (2) and it is calculated for each pixel to obtain a fluidity map (Figure 2). The calculated GP of the PM of INS-1E cells shown in Figure 2 was 0.38. Figure 3(A) depicts high-resolution fluorescence images of Laurdan emission for fluidity investigation along with m-cherry emission images in cells exposed to 11 mM glucose without or with 500 μM PA. Fluidity maps of a typical INS-1E cell are shown in Figure 3(A) and are characterized by a very gel-like PM (yellow), a slightly less gel-like circular region in the central part of the cell (violet) surrounded by a very fluid region (purple). M-cherry labels IG have spherical shapes of about 0.5–1 μm diameter. The m-cherry fluorescence (red LUT) co-localizes in a greater extent with the intermediate fluidity central region (yellow domains).

Figure 3. Effects of nutrient overload on beta cells. (A) High-resolution fluorescence images of Laurdan emission for fluidity investigation along with m-cherry emission images. Membrane fluidity is represented in terms of ratio of emission intensities by using the Generalized Polarization (GP) value. The GP ranges from −1 (very fluid-purple (black in print version)) to +1 (very gel-like-orange (white in print version)). M-cherry labels insulin granules, which have spherical shapes of about 0.5–1 μm diameter. (B) GP values of plasma membrane (PM) in cells incubated with increasing glucose concentration for 32 h. Scale bar is 10 μm.

Figure 3(B) shows that GP values of PM, recovered from fluidity maps, remained unchanged (GP = 0.38) in cells that were incubated with increasing glucose concentrations (5, 11, and 25 mM) for 32 h. The GP values of IG displayed a more fluid state (GP = 0.26) than the PM at all glucose concentrations. Yet, after the addition of 500 μM PA for 16 h the PM of INS-1E exhibited a significant decrease of fluidity in a glucose concentration dependent manner. Generally, under these conditions the fluidity of the PM decreased by ∼15% and ∼27% at 5 mM and 25 mM, respectively. Of interest is the accompanying ∼40% decrease in GP values of IG membranes. Indeed, we have recently reported that the content of PA in total cell phospholipids increased in the presence of exogenously added 500 μM PA from 45.6 ± 0.3 to 81.6 ± 1.0, 48.2 ± 0.3 to 91.7 and 53.4 ± 2.5 to 100.0 ± 1.3 ng/10⁶ in cells incubated for 24 h with culture medium containing 5, 11, or 25 mM glucose, respectively [11]. Concomitantly, total PUFA content decreased from 74.6 ± 3.2 to 52.9 ± 1.0, from 53.1 ± 0.8 to 36.6 ± 0.5, and from to 48.1 ± 0.9 to 32.7 ± 0.7 ng/10⁶ (mean ± SEM, n = 4), respectively. Noteworthy, total fatty acid content remained unchanged under these experimental conditions [11]. The total membrane lipidome analysis, indicating the relative enrichment in PA in membrane phospholipids and concomitant reduction in PUFA abundance, correlates well with the significant decrease of fluidity of the IG membranes as depicted in Figure 3(A).

It is worth noting that by increasing glucose concentration the release of PUFA from membrane phospolipids, in particular omega-6 linoleic and arachidonic acids, occurs together with their increased peroxidation to 4-HNE, indicating an oxidative stress response in the cells [10,11]. In fact, we found that 4-HNE activates PPARδ in the INS-1E cells and enhances insulin secretion, suggesting that a mild oxidative reactivity is indeed a physiological stimulus that takes part in the whole homeostatic control mechanism, provided that the lipid membrane composition is also maintained in an adequate balanced [10,12,13].

Impact of Lcat−/− and Apoa1−/− on the fluidity of mouse RBCs

In this second paradigm, we investigated the potential role of altered cell membrane fluidity on lipoprotein induced cellular changes studying RBC from Lcat−/− and Apoa1−/− mice, which exhibit impaired high density lipoprotein (HDL) metabolic pathway and HDL cholesterol (HDL-C) homeostasis [42,43].

The analysis of membrane fluidity of RBC isolated from these mice was carried out by Laurdan two-photon microscopy revealing that in both animal models PM fluidity is altered at baseline in comparison with RBC from WT mice. Image processing and analysis showed that Lcat−/− PM displayed a lower GP index than WT mice (0.692 ± 0.003 vs. 0.705 ± 0.006, p < .001), which is associated with a more fluid state. On the contrary, Apoa1−/− PM demonstrated an increased GP, which correlates with a more gel-like state of their RBC membrane (0.711 ± 0.002 vs. 0.705 ± 0.006, p < .05) (Figure 4).

Figure 4. Analysis of membrane fluidity of peritoneal macrophages isolated from APOA1 or LCAT deficient mice by Laurdan two-photon microscopy. High-resolution fluorescence images of Laurdan emission for Lcat−/− and Apoa1−/− mice show a drastically altered RBC membrane fluidity at baseline than WT mice. Scale bar is 20 μm.

Discussion

A combined approach based on lipidomic analysis of membrane phospholipids, performed before and the current fTPM-based analysis of membrane physical state of cells, was employed to investigate the effects of fatty acids under varying nutritional and pathological conditions in two case studies: pancreatic beta-cells under fatty acid/glucose feeding and RBC from a genetic mouse models defective in lipoproteins. The first was investigated to elucidate how glucose- and fatty acid-induced phospholipid remodeling affects the biophysical properties of membranes, and the second was useful to understand the role of lipids in hepatic diseases and metabolic syndrome and their effect on circulating cells. The relationship between lipidome variation and membrane biophysical state is not straightforward: the lipid composition of cellular membranes is continuously adjusted to modify their fluidity in response to changes in their physiochemical environment. The length and degree of unsaturation of natural fatty acids determines their impact on the membrane biophysical properties: SFA with short chains form less viscous membranes, unsaturated fatty acids form more fluid membranes than SFA [44]. Cholesterol regulates fluidity of membranes by acting as an “anti-freeze” agent: below the melting temperature T_m (gel-like state) it increases membrane fluidity, while above T_m (fluid state) the rigidity of cholesterol ring reduces the freedom of motion of adjacent acyl chains, and thereby decreases membrane fluidity [45]. Sphingolipids are also incorporated into membranes and affect their structure and function [16]. Membrane and transmembrane proteins can alter lipid density, packing and interactions, and selective alterations in membrane lipid composition may in turn influence the homo- and/or heterodimerization of cell-surface proteins, receptor interaction with effector proteins and subsequent signaling pathways, protein aggregation, as well as transmembrane transport of small molecules [46–48]. The enzymatic activity of desaturases and elongases can directly affect membrane fluidity by increasing the abundance of MUFA and PUFA in membrane phospholipids [45]. Moreover, non-enzymatic peroxidation transforms unsaturated fatty acids, mainly PUFA, into oxidized metabolites, altering membrane composition and properties [17,19,49,50]. Concomitantly, PUFA metabolites activate several signaling pathways, either in protective or detrimental manners [6,8,19,51–53]. Similarly, sphingolipids are an abundant source for signaling molecules, such as sphingosine-1-phosphate that also regulate various cell functions [16]. One aspect of the adaptive mechanism cells employ is the rapid remodeling of membrane phospholipids, that rapidly varies at any given stimulus through the process of membrane remodeling (e.g. Land’s cycle [54]). Lipid remodeling rapidly alters membrane composition, either by modifications like lipid peroxidation or by exchanges among different types of fatty acids. The latter process is determined by the availability and abundance of fatty acids to cells and the organism and by the function of enzymes that remodel phospholipids. Therefore, the homeostatic control of membranes requires a balanced pool of fatty acids that allows for their fast adaptive recovery following metabolic, oxidative, or radical stress.

Pancreatic beta cells exposed to increasing concentrations of PA complete the incorporation of PA and the release PUFA in less than 2 h [11]. This entails rapid activation of phospholipase A2 (PLA2) for PUFA hydrolysis and of other enzymes to mediate PA acylation into phospholipids [10,11]. The peroxidation process that accompanies the release of PUFA in the cell interior, demonstrated by 4-HNE formation in beta cells under glucose administration, can be regarded as a critical event for augmented insulin secretion [10,11,13]. The cellular mechanisms, taking part in the rapid adaptive response of the cells, involve activation of cell surface receptors for fatty acids (e.g. the Gq-coupled Free Fatty Acid Receptor 1, also known as GPR40), generation of various SFA, MUFA, and PUFA by elongation and desaturation reactions, enzymatic transformation of PUFA to numerous mediators (such as prostaglandins, thromboxanes, leukotrienes, lipoxins, and eicosanoids) and non-enzymatic peroxidation of PUFA to generate other active mediators, such as hydroxyalkenals [4,6,8,9,51,55–57].

In addition, SFA, like PA, can modify cell functions by serving as a precursor molecule for ceramide and sphingolipids biosynthesis or for direct palmitoylation of proteins [39,58–61]. Overall effects of these lipidome variations induced by nutrient overload resulted in differential effects in PM and IG membrane fluidity. Laurdan-based fTPM analysis clearly showed that the exposure of INS-1E cells to different glucose concentrations over time and/or exposure to PA diminished the basal difference between the fluidity of the PM and the IG membranes. Glucolipotoxic conditions could therefore have an impact on the integrity, fluidity, and function of beta cells and on the mechanism of IG recruitment to the PM and insulin secretion. Collectively, these interactions and metabolites are paramount in the adaptive and compensatory mechanisms cells employ to cope with nutrient overload such as hyperglycemia and hyperlipidemia, which are associated with type 2 diabetes and obesity.

In the second paradigm, RBC membranes were used as a reporter of the status of mice with genetic defects in lipoproteins. Liver is the central organ that regulates lipid turnover. First pass metabolism of dietary lipids (e.g. cholesterol, fatty acids) in the liver directs them to storage in extrahepatic tissues, mainly in adipocytes. This requires biochemical transformations of fatty acids, triglyceride biosynthesis, packing in lipoproteins and export to the systemic circulation. Complex interactions among the different species of lipoproteins and peripheral organs and tissues throughout hormonal, neural and metabolic regulators determine the fate of the cargo lipids and their distribution among peripheral tissues [62].

High density lipoprotein affects membrane cholesterol composition and homeostasis mainly through its ability to promote the exchange of cholesterol among tissues, in particular via reverse cholesterol transport (RCT), to the liver. An RCT cycle begins with efflux of free cholesterol and phospholipids from the PM to the lipid poor apolipoprotein A1 (APOA1) and is mediated by ATP-binding cassette A1 (ABCA1) transporter. This results in the formation of nascent, discoidal particles of HDL composed of two antiparallel molecules of APOA1 and a bilayer of phospholipids and free cholesterol arranged in a horse-shoe like structure [63–65]. Discoidal HDL particles are then converted into spherical structures through the action of the plasma enzyme lecithin–cholesterol acyltransferase (LCAT), which esterifies lipoprotein free cholesterol using the fatty acyl group on the position C-2 of lecithin as the acyl donor [66]. Eventually these modifications lead to the creation of mature spherical HDL particles, capable of delivering their cholesterol cargo to the liver for excretion by the scavenger receptor class B type I (SR-BI). Fatty acid-based lipidomics represent a powerful tool to achieve such information [59,60], thus providing RBC membrane lipidome profiles that trigger research on the mechanisms that underlie alteration of membrane properties. As matter of facts, membrane fatty acid is a comprehensive health biomarker whose use in diagnostics is matter of intense work, such as recently appeared in obesity [61]. On the other hand, lipid modifications in HDL have a profound impact on membrane fluidity and functionality. Based upon the present data on RBC from ApoA or LCAT deficient mice, additional studies are planned to identify precisely how HDL regulates the physical state and the functionality of PM in RBC and other cells, and in relation to animal and human physiology.

In conclusion, combining the imaging data with molecular information, such as lipidomics and signaling cascades, a multidisciplinary and integrated approach emerges, with invaluable contribution for investigating and analyzing membrane organization and remodeling mechanisms. The high spatial and temporal resolutions of functional microscopy techniques make them ideal to monitor in real-time and on live cells the behavior of biological membranes in a variety of nutritional and stress conditions. Our observation also suggests that this integrated approach may open a new perspective to nutritional and molecular effects on cells, that may later be applied to respective nutritional research and the development of nutritional and pharmaceutical intervention strategies that may prolong the adaptive phase and allow for stable metabolic homeostasis.

Acknowledgements

GM would like to acknowledge Mario Amici and Daniela Samengo for their excellent technical assistance and the support by Fondi di Ateneo, UCSC Rome, Italy (Linea D1 to GM, GP and MDS). The confocal analysis has been performed at Labcemi, UCSC, Rome. SS is Adolf D. and Horty Storch Chair in Pharmaceutical Sciences, at the Faculty of Medicine, The Hebrew University of Jerusalem, Israel. He is affiliated with the David R. Bloom Center for Pharmacy and the Dr. Adolf and Klara Brettler Center for Research of Molecular Pharmacology and Therapeutics in the Hebrew University. KK is the chairman of the Pharmacology Department of the University of Patras Medical School and the director of the University of Patras research network “MetSNet”. The authors acknowledge the contribution of Lipinutragen srl in providing membrane lipidomic methodologies and information from its human lipidome database.

Disclosure statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.

Funding

All authors gratefully acknowledge the support and networking opportunities they received and enjoyed from COST (European Cooperation in Science and Technology) Action CM1201 on “Biomimetic Radical Chemistry”. This COST Action also provided Short-Term Scientific Mission (STSM) fellowships to BD and PP for training in confocal microscopy at GM Laboratory in Rome and for GM training in beta cell research at the laboratory in Jerusalem. SS would like to acknowledge support from Legacy Heritage Biomedical Science Partnership of the Israel Science Foundation of the Israel Academy of Sciences and Humanities (1429/13), the Vigevani Foundation, the Brettler Center for Research of Molecular Pharmacology and Therapeutics in the Hebrew University. GM would like to acknowledge Università Cattolica del Sacro Cuore, 10.13039/501100005743 [D1 2015]. KK would like to acknowledge the support of RioPharm Pharmaceuticals (Patras, Greece) the action “Supporting post-doctoral researchers” of the Operational Program “Education and Lifelong Learning” of the Greek General Secretariat of Research and Technology (grant LS4-858) and the action Siemens Excellence award with the acronym “EC3HDLglucose” financed by the Hellenic State Scholarships Foundation (IKY).