Gb3 glycosphingolipids are the specific receptors for bacterial Shiga toxin. Whereas the trisaccharidic head group of Gb3 defines the specificity of Shiga toxin binding, the lipophilic part composed of sphingosine and different fatty acids is suggested to determine its localization within membranes impacting membrane organisation and protein binding eventually leading to protein internalisation. While most studies use Gb3 extracts, chemical synthesis provides a unique tool to access different tailor-made Gb3 glycosphingolipids. In this review, strategies to synthesize these complex glycosphingolipids are presented. Special emphasis is put on the preparation of Gb3 molecules differing only in their fatty acid part (saturated, unsaturated, α-hydroxylated and both, unsaturated and α-hydroxylated). With these molecules in hand, it became possible to investigate the phase behaviour of liquid ordered/liquid disordered supported membranes doped with the Gb3 species by means of fluorescence and atomic force microscopy. The results clearly highlight the influence of the different fatty acids of the Gb3 sphingolipids on the phase behaviour and the binding properties of Shiga toxin B subunits, even though the membranes were only doped with 5 mol% of the receptor lipid. To obtain fluorescent Gb3 derivatives, either fatty acid labelled Gb3 molecules or head group labelled ones were synthesized. These molecules enabled us to address the question, where the Gb3 sphingolipids are localized prior protein binding by means of fluorescence microscopy on giant unilamellar vesicles. The results again demonstrate that the fatty acid of Gb3 plays a pivotal role for the overall membrane organisation.
Glycosphingolipids (GSLs) are a subclass of glycolipids that are found in cell membranes of various organisms ranging from bacteria to humans. Being typically a minor component of the cell membrane, they are of utmost importance for biological functions that rely on lipid-lipid and lipid-protein interactions (Schnaar and Kinoshita 2015). The chemical structure of 90% of mammalian GSLs is based on glucosyl ceramide, whereas galactosyl ceramide serves as a precursor for the remainder. One can define several major GSL series based on their internal core carbohydrate structure. One can classify them into the ganglio (GalNAc-β-1–4-Gal), globo (Gal-α-1–4-Gal), lacto (Gal-β-1–3-GlcNAc-β-1–3-Gal), and neolacto (Gal-β1-4-GlcNAcβ-1–3-Gal) series. Based on the ganglio GSL series, gangliosides are synthesized, where sialic acids are linked to the glycan structure to produce negatively charged GSLs. In contrast, globosides are neutral lipid molecules.
For the analysis of GSLs, they are extracted from tissues and cells using organic solvents. Extraction procedures are optimized to precipitate and remove proteins and nucleic acids while maximizing solubilization of GSLs along with other lipids. The mixture is then subjected to thin layer chromatography (TLC) to monitor the purity and to separate different GSLs. After separation by TLC, 10–12–10–9 mol quantities of GSL can be chemically detected with orcinol, a reagent for hexoses and with resorcinol-HCl, a reagent for sialic acid. Of note, with these extraction procedures, GSLs are categorized by their glycan structure, whereas the lipid tails are not determined (Fahy et al. 2009).
It is known that GSLs interact with both intracellular as well as exogenous proteins and are critical for membrane organisation, signalling, and recognition events. For example, membrane GSL receptors are known for exogenous microbial virulence factors such as cholera toxin and vero toxins (Johannes and Römer 2010; Lingwood 2011). In this context, major effort has been put in developing synthetic strategies towards gangliosides and the interested reader is referred to an excellent review of Hunter et al. (Hunter et al. 2018). Here, we will solely focus on the less recognized globosides and in particular the globoside Gb3, which is the membrane receptor for vero toxins. Vero toxins, a family of E. coli produced AB5 toxins responsible for the pathology of the haemolytic uremic syndrome (HUS) (Johannes and Billet 2020; Karmali et al. 2010) as well as Shiga toxin are known to specifically bind to Gb3. Okuda et al. (Okuda et al. 2006) have shown that Gb3 synthase knockout mice are completely protected against vero toxin infection demonstrating the high specificity of these toxins towards Gb3.
The globoside Gb3
The globotriaosyl ceramide Gb3, also referred to as PK blood group antigen and CD77 (Johannes and Römer 2010), is the major receptor lipid for a number of lectins. Just recently, Römer and co-workers have shown that Gb3 segregates into different domains at the plasma membrane of cells, which leads to a different binding behaviour of Shiga toxin compared to the lectin LecA from Pseudomonas aeruginosa (Schubert et al. 2020). The receptor lipid is composed of the trisaccharide α-D-galactose-(1 → 4)-β-D-galactose-(1 → 4)-β-D-glucose, bound via an acetal to a sphingosine moiety. One fatty acid is connected via an amide bond to the 2-amine of the sphingosine (Fig. 1).
While a number of studies investigated the specificity of protein binding to Gb3 (Johannes 2017), the variability of the ceramide backbone harbouring different fatty acids has been much less addressed. In various cell types [examples are (primary) human endothelial cell lines, human colon Caco-2, or HCT-8 epithelial cells (Kouzel et al. 2017) and references therein], a conserved repertoire of Gb3 species was found with saturated C16:0, C22:0, or C24:0 fatty acids and the unsaturated C24:1 fatty acid.
The ceramide lipid component of Gb3 has a major effect on Shiga toxin Gb3 binding (Kiarash et al. 1994; Pellizzari et al. 1992). The orientation of the carbohydrate moiety is dependent on both the lipid composition of the glycosphingolipid itself (Nyholm and Pascher 1993) and the membrane in which it is embedded (Lingwood 2011; Mahfoud et al. 2010; Watkins et al. 2014). Moreover, Lingwood and co-workers (Lingwood et al. 2010; Lingwood 2020; Mahfoud et al. 2009) suggested that the pathogenicity of Shiga toxin producing E. coli (STEC) infections is influenced by the different Gb3 species. Despite the importance of the fatty acid of Gb3, the majority of studies has been performed using Gb3 mixtures extracted from different sources, which vary in their fatty acid composition (Bien et al. 2020). Some early studies have determined the binding capacity and affinity of different Gb3 species towards Shiga toxin with the result that Gb3 sphingolipids with long fatty acid chains (C20-C24) have a higher affinity towards Shiga toxin than those with shorter chains (C12, C14) and that unsaturation and α-hydroxylation further enhance binding (Binnington et al. 2002; Kiarash et al. 1994). The finding that mixtures of Gb3 species behave differently than the sole components shows that the interaction with Shiga toxin is complex (Mahfoud et al. 2009). Even though such experiments were partially conducted in the presence of auxiliary lipids, results on membrane-embedded Gb3 species with a defined fatty acid are very rare. To be able to investigate the impact of Gb3 sphingolipids with respect to the chemical nature, synthetic strategies are required to obtain pure compounds. Moreover, to localize these Gb3 molecules in lipid membranes, fluorescent labels attached to Gb3 are highly desirable.
Synthetic strategies to obtain Gb3 glycosphingolipids
Our ultimate goal was to synthesize fluorescent Gb3 glycosphingolipids that do not disturb the natural structure of a lipid bilayer and are still able to allow protein binding. Thus, we followed two directions:
1) Introduction of the fluorophore as part of the fatty acid: Numerous fluorescent fatty acids harboring NBD (Pagano 1990), BODIPY (Kaiser and London 1998), BOIMPY (Patalag et al. 2016), pyrene (Somerharju 2002) and other labels (Patalag and Werz 2016) have been developed in recent decades and have been successfully applied in various fields. However, the membrane organisation suffers either from their steric bulkiness and/or their most often polar properties. Therefore, we sought to integrate a non-bulky, non-polar fluorescent dye into the natural structure of the fatty acid. Building upon the work by Amat-Guerri and co-workers (Souto et al. 1994) and Thiele and co-workers (Kuerschner et al. 2005) we planned to introduce fluorescent conjugated pentaene and hexaene units as an integral part of the acyl chain.
2) Attachment of a fluorophore to the carbohydrate head group: To attach fluorophores to a hydrophilic carbohydrate head group has been demonstrated recently by Ando and co-workers and the dynamic behaviour of glycosphingolipids in membranes has been studied (Komura et al. 2016). Commonly the fluorophore is directly attached to the glycan part or via a linker. However, if the glycosphingolipid still needs to be capable of binding a specific protein, an exact knowledge of the protein in complex with the head group of the glycosphingolipid, as obtained by X-ray crystallography data, is required to identify the hydroxy groups of the sugars that are involved in binding and thus cannot be modified. If hydroxy groups are not involved and not completely buried in the binding site, they might be a point of attachment for a linker connected to a fluorophore or any other moiety (Isobe et al. 2007).
For both directions, modular chemical synthesis approaches were pursued that are described in the following chapters. Those readers, who are interested in the biophysical analysis of these synthetic molecules in the context of artificial lipid membranes, are directly referred to the chapters further back of this review.
α-Hydroxylated nervonic acid
The first task to obtain the non-labelled and head group labelled Gb3 sphingolipids was to prepare the required fatty acids, either non-labelled also for the head group labelled Gb3 sphingolipids or fluorescently labelled. We decided on fatty acids with 24 carbon atoms as they are known to be a major part of natural Gb3 sphingolipid mixtures. Whereas saturated (C24:0) and unsaturated (C24:1) fatty acids are commercially available and configurationally defined and α-hydroxylated saturated derivatives had been synthesized in the past (Iwayama et al. 2009; Konen et al. 1975; Patterson et al. 1996), a synthetic route to α-hydroxylated unsaturated derivatives was not known. Oxidizing steps to introduce the α-hydroxy group as they are commonly employed in the synthesis of α-hydroxylated saturated congeners are not possible, because the C–C double bond would suffer.
Thus, we designed a novel synthetic route to α-hydroxylated nervonic acid using a chiral pool approach (Scheme 1). D-Malic acid (1) that contains a defined stereo centre was transformed by a literature-known protocol consisting of protection, reduction and oxidation into aldehyde 2. In the next step, the carbon chain was elongated via a Wittig reaction using a phosphonium salt to furnish 3. Hydrogenation of the olefinic moiety by Pd(OH)2 on charcoal under hydrogen atmosphere and removal of the benzylic protecting group, followed by subsequent oxidation with pyridinium chlorochromate (PCC) afforded aldehyde 4. As the next step a (Z)-selective Wittig reaction was applied. Key is the use of a non-stabilized phosphorus ylide. Traces of the trans-stereoisomer could be easily separated by column chromatography. The isopropylidene protecting group was cleaved under acidic conditions affording target compound 6. The respective fatty acid enantiomer that is found in bacteria was prepared in a completely analogous way starting from L-malic acid (Pawliczek et al. 2014).
Fluorescent oligoene fatty acids
To access oligoene-containing compounds traditional approaches are Pd-catalysed reactions (either as cross-coupling reactions or as carbopalladation cascades of oligoyne units) or Wittig-type reactions. The latter are often superior for capricious compounds, because harsh reactions conditions such as high temperature can be avoided. Therefore, we designed a synthetic route to oligoene fatty acids based on Wittig-type reactions. Originally, Thiele and co-workers investigated this approach (Kuerschner et al. 2005). We modified the route so that it became shorter and the sensitive oligoene was introduced at a later stage by an elimination reaction.
Starting point was commercially available (4E)-decenal (7) that was chlorinated quantitatively in α-position by N-chlorosuccinimide (NCS). The addition of L-proline acting as a proton shuttle was crucial. A Wittig-Horner reaction with phosphonate 9 afforded compound 10 in 65% yield. Problems with low-yielding Wittig-Horner reactions due to the conjugation were diminished by this procedure. After this coupling the elimination triggered by a large excess of DBU took place providing tetraene ester 11. A reduction/oxidation procedure furnished aldehyde 12 as crucial intermediate for the next Wittig reaction with phosphonium salts 13 or 14 of different alkyl chain length. The low reaction temperature guaranteed a predominant formation of the (Z)-isomers 15 and 16 being non-natural counterparts to the common cis-configured unsaturated fatty acids. However, traces of iodine in boiling n-hexane catalysed an isomerization to the all-(E)-pentaene fatty acids 17 and 18 (Scheme 2). A similar synthetic protocol also secured the preparation of the corresponding fatty acids with six conjugated double bonds embedded into the carbon chain.
Photophysical properties of these systems were investigated. Whereas the change of the proximal double bond from (E) to (Z) is with about 2 nm almost negligible, the difference between pentaene and hexaene fatty acids is striking (Fig. 2). Large Stokes shifts of 120 nm in pentaene and 159 nm in hexaene are observed (Patalag and Werz 2012).
The larger π-system shifts the UV-absorption maximum by about 30 nm from 350 to 379 nm to the red. Whereas the molar absorbance for the pentaene fatty acids are in the range of 60,000–70,000 M1 cm−1 (THF), these values increase for the hexaene fatty acids to 70,000–100,000 M1 cm−1 (THF). Although the slim shape and hydrophobic scaffold of the pentaene and hexaene fatty acids was expected to disturb the membrane integrity only mildly, these molecular entities turned out to be quite unstable if they were irradiated with laser light. Moreover, absorption far below 400 nm is not well-suited for fluorescence microscopes with a standard laser equipment. Hence, new fluorescent fatty acids with either a phenyl or a thienothienyl residue in conjugation to an oligoene system were designed with the idea to extend the π-system to red-shift the absorption and emission spectra and simultaneously increase the stability of the fluorescent core structure without increasing the bulkiness of the molecules too much.
The synthesis is similar to fatty acids mentioned above using Wittig-type and reduction/oxidation procedures (Scheme 3). Starting material is benzaldehyde (19); the carbon chain gets successively elongated until compound 24 is reached. Finally, again an isomerization to the all-(E)-phenylpentaene had to take place. Due to the more extended π-system this compound absorbs at 390 nm and provides a higher photo-stability than the previously synthesized oligoene fatty acids. An even larger red shift was observed by embedding the very electron-rich thienothienyl moiety into an oligoene chain. The synthetic access to the thienothienyl-modified fatty acid 31 resembles the previously described ones (Scheme 4). First, thienothiophene-2-carbaldehyde (26) was engaged in a Wittig reaction to furnish isopropylidene-equipped compound 27. A lithiation/carbonylation sequence using DMF as carbonyl source afforded aldehyde 28 which was subjected to a Wittig-Horner reaction using phosphonate. The emerging ester 29 was reduced and subsequently oxidized to aldehyde 30. The latter was transformed by a Wittig reaction/isomerization procedure into the target compound. The photophysical investigations revealed an absorption maximum at λmax = 409 nm. Also, this fatty acid proved to be much more stable than the oligoene fatty acids 17 and 18. However, the drawback of these arene-substituted fatty acids is that (i) the fluorophore is not as slim as in the oligoene fatty acids and (ii) the π-system being stiffer than a saturated hydrocarbon chain is located at the terminus of the fatty acid and not in the middle of the hydrocarbon chain.
Synthesis of non-fluorescent and fluorescent Gb3 glycosphingolipids with defined fatty acids
The power of chemical synthesis towards Gb3 glycosphingolipids is most promising because of the well-defined nature of the resulting material. Three major parts can be differentiated in a glycolipid which lead to the major retrosynthetic cuts: (i) the carbohydrate head group (hydrophilic part) and the ceramide (lipophilic part). The ceramide consists of a (ii) sphingosine such as d-erythro-sphingosine in mammalian glycolipids and (iii) a fatty acid which is linked to the amino group of the sphingosine.
In Gb3 the carbohydrate head group is a triaose which consists of two galactose moieties and one glucose. Whereas the sphingosine stays the same, the fatty acids vary in length, unsaturation and hydroxylation. Here, we concentrated our synthetic efforts on Gb3 molecules with 24 carbon atoms either saturated, mono-unsaturated, or α-hydroxylated. In addition, we also synthesized a Gb3 in which the fatty acid part is α-hydroxylated and unsaturated. The globotriaoside is traced back to three different monosaccharidic building blocks (Scheme 5, 32–34). Glucose derivative 34 was obtained in five steps starting from d-glucal that can be prepared in 100 g scale from D-glucose. The galactosyl building block 32 was synthesized starting from d-galactal (Werz et al. 2007). The third unit, an α-linked galactosyl acceptor, was obtained from 3,6-di-O-benzyl-D-galactal (40) (Scheme 6). A remote-participating group (Bz) was installed at the 4-hydroxyl group to facilitate the formation of the α-galactosidic linkage in the trisaccharide. DMDO-mediated epoxidation of galactal 41 followed by the attack of the emerging acetalic epoxide using p-methoxyphenol (MPOH) led to afforded acetal 42 in 90% yield. Benzyl ether formation and removal of the anomeric protecting group under oxidative conditions led to hemiacetal 43. With trichloroacetonitrile under the basic catalysis the trichloroacetimidate was formed. Because sometimes phosphate donors are superior in glycosylation reactions (Ahadi et al. 2020) the respective trichloroacetimidate was converted into phosphate 32.
With the three monosaccharide building blocks the trisaccharidic carbohydrate head was assembled. The sequence starts with the reaction of buildings blocks 33 and 34. TMSOTf was used as Lewis acid to promote the cleavage of the phosphate (Scheme 7). Pivaloate as neighbouring participating group ensured the formation of a β-linkage. Subsequent removal of the Fmoc protecting group followed by the second glycosylation with building block 32 resulted in trisaccharide 44 in 76% yield (over three steps). Remote participation of the benzoate in position 4 of the terminal galactose and low temperature (− 60 °C) secured pure α-selectivity. Previous experiments of a completely benzylated galactosyl phosphate revealed a much lower yield and a worse selectivity. The commonly used procedure to cleave all the benzyl protecting groups by hydrogenolysis at the very end of an oligosaccharide assembly is not a choice, because the double bond of the sphingosine would be destroyed during this hydrogenation. Therefore, all the protecting groups of the globotriaose unit were exchanged by benzoate protecting groups before the lipid part is attached. Such benzoates can be easily cleaved at the end of the synthesis under basic conditions before lipid attachment. To get perbenzoylated material, the three different acyl groups were removed to yield 45, followed by a hydrogenolysis of the benzyl ethers using Pearlman’s catalyst. Perbenzoylation of the naked trisaccharide was performed followed by cleavage of the anomeric benzoate using methyl amine. Hemiacetal 46 was converted into trichloroacetimidate 47 in quantitative yield.
With the protected trisaccharide acceptor 47 in hand, the non-labelled Gb3 sphingolipids can be synthesized by the union with the lipophilic sphingosine alcohol 35a/35b. BF3⋅OEt2 was used as catalyst to trigger this difficult glycosylation reaction; moderate yields were obtained (Scheme 8).
Reduction of the azide moiety by Staudinger reaction and direct coupling with the fatty acids 36–39 under common peptide coupling conditions (e.g., HATU or HOBt/EDCI) afforded four protected glycosphingolipids with different fatty acid chains. Global deprotection was performed under Zemplén conditions resulting in the glycosphingolipids 49–52.This modular procedure readily allowed us to also synthesize Gb3 derivatives with deuterated fatty acids (Bosse et al. 2019). Here, commercially available completely deuterated fatty acids were used for the amidation procedure (Fig. 3).
For the head group labelled Gb3 derivatives, a fluorescent dye was attached to the carbohydrate head group via an oligoethylene glycol (OEG) linker. However, following this strategy includes that the attached fluorophore must not alter the binding behaviour of Shiga toxin B subunits (STxB). Hence, we chose to attach the fluorophore via an OEG spacer to the 2′-OH group of the middle galactose Gb3, known to not participate in STxB binding (Kitov et al. 2003; Kitov and Bundle 2001; Ling et al. 1998). This synthetic route has the advantage that the influence of the fatty acid on the partition of Gb3 in phase-separated bilayers can be readily investigated.
The major goal was to design a modular convergent synthesis for head group labelled Gb3 derivatives with different fatty acids and different OEG linker lengths allowing us to vary the fatty acid and the fluorophore with minimal synthetic effort (Scheme 9). In contrast to semisynthetic methods, this approach ensures that a highly defined material is obtained being prerequisite for quantitative biophysical investigations. The retrosynthetic analysis of the desired structures 56–63 led to four different components. The BODIPY dye 66, which is commercially available was planned to be attached to the carbohydrate head group in the last step of the synthesis via a Huisgen (3 + 2)-cycloaddition (which is often also termed as click-chemistry). As mentioned before, we anticipated that the sphingosine core should be introduced as azido sphingosine 35a in which the azide serves as a masked amine. Amidation with four different fatty acids would lead to the respective glycolipids of a C24 chain length.
The trisaccharide building block is much more difficult to access than the previously shown one without the OEG linker. To attach the OEG at the 2-hydroxyl group of the middle galactose a different building block for this monosaccharide had to be designed. Instead of the permanent pivaloyl ester (in Scheme 7) we used Fmoc as easily cleavable temporary protecting group (Oberli et al. 2008; Werz and Seeberger 2005) in position 2 and PMB in position 4.
Fmoc in 68 ensures as neighbouring participating group the β-selectivity, whereas PMB is easily cleavable in the presence of acids.
Indeed, the assembly showed that already the presence of the Lewis acidic TMSOTf proved to be sufficient to cleave PMB after the formation of the disaccharide (Scheme 10). Glycosylation with the third monosaccharide building block 67 afforded the core trisaccharide. Piperidine as amine base led to a deprotection of the Fmoc and the generation of 70. Trisaccharide 70 was then equipped with a pentenyl chain in the position, where the flexible OEG linker connected to the fluorophore needs to be attached (Scheme 11). In the subsequent step all benzyl protecting groups were removed under Birch conditions. During this procedure the anomeric CH2CH2TMS group and the double bond in the pentenyl handle stayed intact. All emerging hydroxyl groups of the trisaccharide were converted into benzoates. Such esters are advantageous as they can be easily removed at the end of the synthesis without affecting the double bond in the lipid part.
In a further step, the OEG linker is attached to the double bond. To achieve this, a thiol was first introduced as a linking element by a radical approach. The highly nucleophilic thiolate easily reacted with OEG bromides 71 (13 ethylene glycol units) and 72 (3 ethylene glycol units), respectively. To ensure a full protection of all hydroxyl groups, the benzoylation step had to be repeated. Finally, the silylethyl group at the anomeric position was removed with trifluoroacetic acid and the emerging hemiacetals were converted into the corresponding trichloroacetimidates 64 and 65. To build up the glycolipid, the trichloroacetimidates 64/65 were reacted with protected azidosphingosine 35a, which was prepared from the naturally occurring amino acid l-serine. TMSOTf served as Lewis acid for this glycosylation reaction and afforded compounds 73 and 74 in moderate yields (Scheme 12). The azide was transformed in a Staudinger reduction to an amine which was coupled directly to the respective fatty acids; OEG-modified glycosphingolipids were obtained. Global deprotection was achieved using sodium methoxide as base furnishing compounds 75–82, respectively. Click chemistry—or in other words a Cu-catalysed Huisgen (3 + 2)-cycloaddition reaction (Breugst and Reissig 2020)—was used to attach the commercially available BODIPY dye 66 to the terminal triple bond of the linker.
Eight different fluorescently labelled glycosphingolipids 56–63, varying in the OEG linker length and the acyl chain of the fatty acid were obtained (Scheme 13).
Analysis of STxB binding affinities as a function of different Gb3 glycosphingolipids
The functionality of the synthesized Gb3 species can be verified by their ability that STxB binds specifically to membranes doped with the receptor lipids. To quantitatively determine the binding affinity, adsorption isotherms of STxB interacting with Gb3-doped monolayers were measured by means of surface plasmon resonance (SPR) technique and the dissociation constants were determined for the non-fluorescently labelled Gb3 species. According to the procedure described by Nakajima et al. (Nakajima et al. 2001), vesicles composed of DOPC/cholesterol (Chol)/Gb3 were spread on an octanethiol functionalized gold sensor chip. After the monolayer has been established, different STxB concentrations were added and the change in reflectivity was read out as a function of protein concentration. For all synthesized non-labelled Gb3 species, a specific binding of STxB with affinities in the nanomolar regime was measured (Table 1).
Only small trends concerning the binding affinities were found. While the highest affinity was determined for the Gb3 mixture from porcine erythrocytes with a KD of about 4 nM identical to that determined for Gb3C24:0 and very similar to that for the hydroxylated species, the monolayers containing the unsaturated Gb3 species show slightly higher KD values. In previous reports, binding studies were only performed with Gb3 species harbouring fatty acids of different length (Binnington et al. 2002; Kiarash et al. 1994; Mahfoud et al. 2009). As the fatty acid chain length in this study is identical for all Gb3 species, we can conclude that the unsaturation is responsible for the different binding affinity probably owing to an altered packing property of the Gb3 molecules in the membrane. Similarly, the binding affinity is slightly diminished if Chol is added to the mixture. To illuminate the aspect of membrane organisation in more detail, we performed fluorescence and atomic force microscopy imaging on lipid bilayers resembling the outer leaflet of the plasma membrane of eukaryotic cells (van Meer and Kroon 2011), doped with the different Gb3 sphingolipids.
Phase behaviour of liquid ordered/liquid disordered coexisting supported membranes doped with different Gb3 sphingolipids
The outer leaflet of the plasma membrane of eukaryotic animal cells contains sphingomyelin (SM) and Chol. A general mixture frequently used in literature, which is also known as the “raft mixture” is composed of DOPC/SM/Chol (40:40:20). At room temperature, this mixture phase separates into a liquid ordered (lo) and a liquid disordered (ld) phase. We used mixtures composed of DOPC/SMporc/Chol/Gb3 (40:25:20:5), where 5 mol% of SM from porcine brain (SMporc) was replaced by Gb3 (Schütte et al. 2014, 2015). As expected, all lipid mixtures show lo/ld coexistence. The two different phases can be distinguished by the lipid fluorophore that preferentially partitions into the ld phase (Sezgin et al. 2012) and the height difference of about 0.6–0.8 nm between lo and ld phase as obtained by atomic force micrographs (Azouz et al. 2019; Sullan et al. 2010) (Figs. 5, 7). Of note, in case of Gb3C24:1, three different phases could be distinguished in the fluorescence micrographs, which we assigned to the lo phase (black), an intermediate phase (li phase, dark green) and the ld phase (green) (Fig. 5c). The three different phases were also reflected in different heights of the phases in the AFM topographs with a larger height difference of 1.6 ± 0.2 nm between the lowest and the highest phase compared to the other cases (Fig. 7). The images clearly show that the fatty acid attached to Gb3 significantly influences the phase behaviour of the DOPC/SMporc/Chol lipid mixtures even though it only constitutes 5 mol% of the total lipids. The result suggests that Gb3 strongly interacts with other lipids and probably partitions between the lo and ld phase differently, dependent on the fatty acid composition. Indeed, this is what we found using fluorescently labelled Gb3 sphingolipids. This notion is also reflected in the differences in the lo fractions. Even though the overall lipid composition is the same, the lo fraction in case of the hydroxylated species (50) is significantly lower (56 ± 5%) compared to the saturated non-hydroxylated one (49, 74 ± 7%) (Fig. 7).
The question next arises how does binding of STxB alters the phase behaviour of the coexisting supported lipid bilayers. It has been shown previously, that binding of STxB to Gb3 alters the lipid membrane organization (Römer et al. 2007; Safouane et al. 2010). The protein can bind up to 15 receptor molecules suggesting a mutual influence of the membrane organisation on the bound protein and vice versa. Hence, it is well conceivable that the difference in partitioning of the Gb3 sphingolipids is also reflected in a different reorganisation behaviour of the membranes upon STxB binding. To investigate this aspect in more detail, fluorescently labelled STxB (60 nM STxB-Cy3) was added to the phase-separated solid supported bilayers and fluorescence and atomic force microscopy images were taken (Fig. 6). From previous studies, it is known that STxB binds exclusively to the lo phase implying that Gb3 is localised in the lo phase after protein binding (Windschiegl et al. 2009). Indeed, also in case of the chemically well-defined Gb3 molecules, STxB binds solely to the lo phase. Even in the case of the unsaturated Gb3 (51), STxB binds only to the lo phase (dark), while the li phase (dark green) is protein free. However, the mode of protein binding appears to be different. Given by the observed height differences in the atomic force micrographs, we distinguish between STxB clusters (high STxB density) indicated by a height difference of Δh between 1.5–2.5 nm (refenced to the ld phase) and less densely packed STxB with only a height difference of Δh = 0.7–1.3 nm as observed by AFM topographs (Fig. 7).
STxB binding to the different Gb3 species, except for the unsaturated one (51), results in the formation of high-density protein clusters and low-density protein areas. Such clusters in combination with Gb3 molecules harbouring a saturated fatty acid might inhibit membrane bending (Römer et al. 2007) due to the rigid structure (Pezeshkian et al. 2015; Watkins et al. 2014). The unsaturated Gb3 species 51 appears to be the most important one for the first step of Shiga toxin internalisation. STxB binding leads only to protein clusters concomitant with a large area demand of the lipids (Gracià et al. 2010; Kollmitzer et al. 2013). This is supposed to induce negative curvature eventually leading to the observed Shiga toxin induced invaginations. Another aspect is the reorganisation of the lipids between lo and ld phase. Here, membranes containing Gb3 molecules with an α-hydroxylation show an increase in the lo fraction after protein binding suggesting a large redistribution of lo phase lipids.
These results let us conclude that the Gb3 distribution between lo and ld phase might be different dependent on the fatty acid. However, while we can safely assume that the majority of Gb3 molecules are localised, where STxB has been bound (Schütte et al. 2014, 2015), it remains unclear how the Gb3 species are distributed in lo/ld coexisting membranes.
Localizing fluorescently labelled Gb3 glycosphingolipids in lipid membranes of GUVs
To be able to address the question, where Gb3 is localized prior to STxB binding, we pursued an approach of fluorescently labelling Gb3. In our first strategy, we synthesized Gb3 molecules with labelled fatty acids (Fig. 4). However, as it turned out that the binding behaviour of STxB was altered in the presence of these molecules, a second labelling strategy was followed, in which a fluorophore was attached to the head group of Gb3 (Scheme 13).
Two different fatty acids with either a phenylpentaene moiety (Scheme 3) or a thienothienyl-moiety (Scheme 4) (Patalag et al. 2017) were attached to obtain the Gb3 species 54 and 55. To analyse the partition of these Gb3 sphingolipids in phase-separated membranes, we made use of giant unilamellar vesicles (GUVs). GUVs can be doped with lipid-like dyes that partition specifically either in the lo or ld phase (Baumgart et al. 2007; Sezgin et al. 2012). GUVs were prepared by the electroformation technique and were composed of DOPC/SM/Chol/Gb3/Dy731-DOPE (39.5/35/20/5/0.5). Importantly, we have chosen a ld phase marker (Dy731-DOPE) with an absorption spectrum shifted to long wavelengths (λex = 732 nm) to not overlap with the emission spectra of the Gb3 molecules, which would result in an undesired fluorescence resonance energy transfer. Dy731-DOPE was chemically synthesized and partitions into the ld phase (Patalag et al. 2017). To determine the fraction of Gb3 species partitioning into the lo phase, we read out the fluorescence intensities of the lo phase and the ld phase from the intensity line profiles of confocal images (Fig. 8), namely, Ilo and Ild. According to Eq. 1 an apparent partitioning coefficient (%lo) was calculated:
For compound 54, %lo was determined to be 45%, whereas %lo = 24% was calculated for compound 55 (Fig. 9). However, even though we could show that the Gb3 species are capable of binding STxB using fluorescently labelled Cy3-STxB, the results also demonstrated that STxB binds to the ld phase in contrast to what is known from naturally occurring Gb3-containing membranes.
This finding clearly demonstrates that small changes in the packing density in the lo phase induced by bulky groups such as the phenyl ring and the thiophene ring can largely alter the phase behaviour of the glycosphingolipids. It has been reported previously that sphingolipids containing BODIPY-FL or NBD in the fatty acid chain do not partition in the lo phase owing to the bulkiness inserted into the hydrophobic part of the membrane (Baumgart et al. 2007; Komura et al. 2016; Sengupta et al. 2008; Sezgin et al. 2012). Similarly, it was observed that GM1 being the natural receptor lipid of cholera toxin, partitions and binds cholera toxin exclusively in the ld phase, if the acyl chain is labelled with BODIPY-FL, even though it is known that cholera toxin binds to the lo phase in case of unlabelled GM1 (Bacia et al. 2005; Komura et al. 2016; Sezgin et al. 2012).
To not touch the fatty acid, another strategy was pursued based on the modification of the head group of Gb3 (Scheme 13). As it is well conceivable that even a fluorophore attached to the head group still influences the phase behaviour of lo/ld phase-separated bilayers, we synthesized two Gb3 series, one with three OEG units, the other one with 13 OEG units. Indeed, the length of the spacer influences the partition of the Gb3 species. The Gb3 sphingolipids with the longer linker (13 OEG units) partitions by 15–33% more in the lo phase than those with only 3 OEG units (Fig. 9). Previous studies reported on similar observations. A lipid that is known to prefer the lo phase was found in the ld phase, if the fluorophore was directly coupled to the head group (Klymchenko and Kreder 2014). If a linker was placed between the fluorophore and the head group, the lipid re-partitions to the lo phase (Honigmann et al. 2013). Apparently, a long linker helps separating the fluorophore from the membrane interface so that it does not disturb the lipid packing. Lipids with longer linkers partition more likely in the lo phase (Bordovsky et al. 2016; Momin et al. 2015).
As we assume that the fluorescently labelled Gb3 with the long linker reflects the natural behaviour of the Gb3 sphingolipids in a better way, we focused our study on these species and analysed the influence of the fatty acid on the partition behaviour. As expected, Gb3 with a saturated fatty acid partitions predominantly in the lo phase (Fig. 9) (74% for 56 and 71% for 57). If the saturated fatty acid is replaced by a mono-unsaturated fatty acid, an almost equal partitioning of Gb3 between lo and ld phase is found (47% for 60 and 50% for 61). The influence of an unsaturated fatty acid on the phase behaviour of coexisting lipid membranes has been also described by Maté et al. (Maté et al. 2014). They reported that if the fatty acid of SM is changed from a C24:0 fatty acid to a C24:1 fatty acid in a DOPC/SM/Chol membrane, the phase separation disappears. These findings support our notion that the packing of the unsaturated Gb3 species disfavours its partition in the lo phase, similarly to what has been reported by Legros et al. for in vivo conditions (Legros et al. 2017). For primary human blood brain barrier endothelial cells, they found that Gb3 with a C24:1 fatty acid resides more strongly in non-detergent-resistant membranes compared to Gb3 with a C24:0 fatty acid. Our results have also implications for the proposed membrane reorganisation upon STxB binding. As more Gb3 is localized in the ld phase prior protein binding, a larger reorganisation will take place upon STxB binding assuming that the majority of Gb3 is in the lo phase after STxB has been bound.
Besides the saturation of the fatty acid, we also investigated the influence of the α-hydroxylation on the Gb3 partition behaviour. Our results show that the OH-group does not influence the Gb3 phase distribution. Others also found that the introduction of an OH-group does not alter the order parameter of the lipids (Morrow et al. 1995a, b). However, we have seen that in case of supported membranes, those lipid bilayers with α-hydroxylated Gb3 sphingolipids exhibit a reduced lo phase fraction (Schütte et al. 2014). Slotte (2013) showed that the 2-OH group increases the hydration in the membrane interface. Moreover, it decreases the affinity of a sphingolipid for sterols, which means that less Chol is recruited into the lo phase in case of hydroxylated Gb3 compared to the non-hydroxylated species. This results in a smaller lo fraction, even though the amount of Gb3 in the lo fraction is the same.
We not only investigated the influence of the fatty acid of the Gb3 molecules, but also looked into more detail into the impact of the fatty acid of SM. In general, long chain sphingolipids are known to be able to interdigitate. For SM with fatty acids of more than 20 carbon atoms, it was reported that they even interdigitate in the liquid-crystalline phase (Maulik et al. 1986; Niemelä et al. 2006; Takahashi et al. 2007). Glycosphingolipids with a fatty acid of 24 carbon atoms can also interdigitate (Niemelä et al. 2006; Róg et al. 2016). Thus, we hypothesized that the SM species influence the partitioning of the Gb3 molecules by a specific and probably similar packing behaviour. In a series of experiments with SM with different fatty acids ranging from C16:0 to C24:0, we saw a clear trend that the partition of Gb3 into the lo phase was favoured for SM harbouring fatty acids with 18 or more carbon atoms, which might indicate that the packing into the lo phase is favoured if the chain lengths match with the possibility to interdigitate (Fig. 9).
The ultimate question remained whether STxB is capable of binding to the head group fluorescently labelled Gb3 molecules. Indeed, we could demonstrate that STxB binds to 56 and that it is localized in the lo phase as expected from the natural Gb3 molecules. However, to readily visualize STxB binding, 5 mol% of Gb3 were required. This concentration already leads to partial self-quenching of the fluorescence which prevented us from quantifying the Gb3 distribution after protein binding.
Packing behaviour of Gb3 glycosphingolipids analysed by NMR spectroscopy
As the packing behaviour of Gb3 appears to be an important aspect of its partitioning, we illuminated this aspect using 2H solid-state NMR spectroscopy (Bosse et al. 2019). To perform these experiments, perdeuterated Gb3 lipid acyl chains were synthesized (Fig. 4). With these compounds in hand, we were able to study the influence of Gb3 lipids with a C24 fatty acid [49, 49(D)] on the properties of phospholipid membranes composed of a lo/ld phase-separated lipid membranes composed of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC)/N-palmitoyl-D-erythro-sphingosylphosphorylcholine (SMC16:0)/Chol/Gb3 (40/35/20/5, n/n) by means of 2H solid-state NMR spectroscopy. From the solid-state 2H NMR spectra, a phase separation into a POPC-rich ld and a SMC16:0/Chol-rich lo phase was concluded. In the mixture with 49, SMC16:0 showed higher order parameters as it preferentially co-localizes with Chol in the Chol-rich lo phase, while POPC showed a lower order parameter being in the ld phase which is depleted of Chol. This result is in agreement with previous 2H NMR data on raft-like mixtures (Bartels et al. 2008; Bunge et al. 2008; Veatch et al. 2007).
49(D) showed an unusual order parameter profile of the acyl chain, which flattens out for the last ~ 6 methylene segments. For the upper chain segments (C2–C16) the order parameter profile of 49(D) shows similar order parameters as known for phospholipids in the liquid crystalline state. Such an odd chain conformation could be an indication for a partial chain interdigitation (Lewis et al. 1994; Morrow et al. 1993) and/or a very fluid midplane region of the membrane as a result of a strong mismatch of the fatty acid chain length of Gb3 and SM. Indeed, if the fatty acids are similar, i.e., 53 (Gb3C16:0) is used, the Gb3 is well associated with the SMC16:0/Chol-rich lo phase.
In the presence of STxB, Gb3 preferentially partitioned into the lo membrane phase as described before (Windschiegl et al. 2009). Especially the short acyl chain 53(D) showed very similar chain order parameters as SMC16:0. Surprisingly, we observed isotropic contributions in the 2H NMR powder spectra for all lipids, which were most pronounced for the Gb3 molecules in the presence of STxB. Such isotropic contributions are a result of highly curved membrane structures with a curvature radius of < 10 nm. Curvature induction has been proposed to be part of the mechanism of Shiga toxin internalisation.
With the aim to elucidate the impact of individual Gb3 species on the phase behaviour of lipid membranes and their ability to bind STxB, we designed synthetic strategies to chemically access pure Gb3 sphingolipids. We developed a modular and very universal chemical approach applicable also to other glycosphingolipids, which allowed us to produce a number of different Gb3 species. These Gb3 sphingolipids were reconstituted into phase-separated lipid membranes.
Even though these phase-separated membranes contained only 5 mol% of the non-labelled Gb3 sphingolipids, they significantly influenced the overall membrane organisation. This is of particular interest in light of lipid reorganisation and protein domain formation as these aspects are discussed to be important during the first step of the internalisation process of Shiga toxin and Shiga-like toxins giving rise to the infection of cells. Our results obtained from artificial membrane systems clearly suggest that the unsaturated Gb3 species are the most important ones for the first step of Shiga toxin internalisation. Binding of STxB results in highly dense protein clusters connected with a large area demand of the lipids This leads to an asymmetric reduction in membrane area, which is proposed to be prerequisite for membrane invaginations induced by Shiga toxin. This process is probably even facilitated as the unsaturated species have a lower bending modulus compared to the saturated ones. To address where Gb3 is localized before STxB binding, fluorescently labelled Gb3 species were synthesized. It has been turned out that great care has to be taken to ensure that the fluorescent label does not influence the phase behaviour of the Gb3 molecules as well as their binding behaviour to STxB. Head group labelled glycosphingolipids eventually enabled us to address the question how the fatty acid of a glycosphingolipid influences its distribution in lo/ld phase-separated membranes. Our results clearly demonstrate that the fatty acid (un)saturation significantly shifts the Gb3 molecules from the lo phase (C24:0) to the ld phase (C24:1). This is another important aspect with respect to lipid reorganisation as STxB exclusively binds to Gb3 in the lo phase. Upon STxB binding the amount of redistributed Gb3 and probably also other lo phase lipids then depend on the fatty acid of Gb3. The length match of the fatty acids of SM and Gb3 appears to play another decisive role in determining, where the Gb3 glycosphingolipids are preferentially localized. Overall, it appears that the combination of the attached fatty acids of SM and Gb3 impacts the distribution of the Gb3 glycosphingolipids. Hence, a picture emerges in which the overall lipid distribution between the different domains in a lipid membrane, which is influenced by the fatty acid composition of Gb3 and the Shiga toxin induced membrane reorganisation partially determines the invagination of the protein into the host cell.
Ahadi S, Awan SI, Werz DB (2020) Total synthesis of tri-, hexa- and heptasaccharidic substructures of the O-polysaccharide of Providencia rustigianii O34. Chem Eur J 26:6264–6270. https://doi.org/10.1002/chem.202000496
Azouz M, Cullin C, Lecomte S, Lafleur M (2019) Membrane domain modulation of Aβ1-42 oligomer interactions with supported lipid bilayers: an atomic force microscopy investigation. Nanoscale 11:20857–20867. https://doi.org/10.1039/C9NR06361G
Bacia K, Schwille P, Kurzchalia T (2005) Sterol structure determines the separation of phases and the curvature of the liquid-ordered phase in model membranes. Proc Natl Acad Sci USA 102:3272–3277. https://doi.org/10.1073/pnas.0408215102
Bartels T, Lankalapalli RS, Bittman R, Beyer K, Brown MF (2008) Raftlike mixtures of sphingomyelin and cholesterol investigated by solid-state 2H NMR spectroscopy. J Am Chem Soc 130:14521–14532. https://doi.org/10.1021/ja801789t
Baumgart T, Hunt G, Farkas ER, Webb WW, Feigenson GW (2007) Fluorescence probe partitioning between Lo/Ld phases in lipid membranes. Biochim Biophys Acta 1768:2182–2194. https://doi.org/10.1016/j.bbamem.2007.05.012
Bien T, Perl M, Machmüller AC, Nitsche U, Conrad A, Johannes L, Müthing J, Soltwisch J, Janßen K-P, Dreisewerd K (2020) MALDI-2 mass spectrometry and immunohistochemistry imaging of Gb3Cer, Gb4Cer, and further glycosphingolipids in human colorectal cancer tissue. Anal Chem 92:7096-7105. https://doi.org/10.1021/acs.analchem.0c00480
Binnington B, Lingwood D, Nutikka A, Lingwood CA (2002) Effect of globotriaosyl ceramide fatty acid α-hydroxylation on the binding by verotoxin 1 and verotoxin 2. Neurochem Res 27:807–813. https://doi.org/10.1023/A:1020261125008
Bordovsky SS, Wong CS, Bachand GD, Stachowiak JC, Sasaki DY (2016) Engineering lipid structure for recognition of the liquid ordered membrane phase. Langmuir 32:12527–12533. https://doi.org/10.1021/acs.langmuir.6b02636
Bosse M, Sibold J, Scheidt HA, Patalag LJ, Kettelhoit K, Ries A, Werz DB, Steinem C, Huster D (2019) Shiga toxin binding alters lipid packing and the domain structure of Gb3-containing membranes: a solid-state NMR study. Phys Chem Chem Phys 21:15630–15638. https://doi.org/10.1039/c9cp02501d
Breugst M, Reissig H-U (2020) The Huisgen reaction: milestones of the 1,3-dipolar cycloaddition. Angew Chem Int Ed Engl 59:12293–12307. https://doi.org/10.1002/anie.202003115
Bunge A, Muller P, Stockl M, Herrmann A, Huster D (2008) Characterization of the ternary mixture of sphingomyelin, POPC, and cholesterol: Support for an inhomogeneous lipid distribution at high temperatures. Biophys J 94:2680–2690. https://doi.org/10.1529/biophysj.107.112904
Fahy E, Subramaniam S, Murphy RC, Nishijima M, Raetz CRH, Shimizu T, Spener F, van Meer G, Wakelam MJO, Dennis EA (2009) Update of the LIPID MAPS comprehensive classification system for lipids. J Lipid Res 50(Suppl):S9–14. https://doi.org/10.1194/jlr.R800095-JLR200
Gracià RS, Bezlyepkina N, Knorr RL, Lipowsky R, Dimova R (2010) Effect of cholesterol on the rigidity of saturated and unsaturated membranes: fluctuation and electrodeformation analysis of giant vesicles. Soft Matter 6:1472–1482. https://doi.org/10.1039/b920629a
Honigmann A, Mueller V, Hell SW, Eggeling C (2013) STED microscopy detects and quantifies liquid phase separation in lipid membranes using a new far-red emitting fluorescent phosphoglycerolipid analogue. Faraday Discuss 161:77–89. https://doi.org/10.1039/C2FD20107K
Hunter CD, Guo T, Daskhan G, Richards MR, Cairo CW (2018) Synthetic strategies for modified glycosphingolipids and their design as probes. Chem Rev 118:8188–8241. https://doi.org/10.1021/acs.chemrev.8b00070
Isobe H, Cho K, Solin N, Werz DB, Seeberger PH, Nakamura E (2007) Synthesis of fullerene glycoconjugates via a copper-catalyzed Huisgen cycloaddition reaction. Org Lett 9:4611–4614. https://doi.org/10.1021/ol702128z
Iwayama Y, Ando H, Ishida H, Kiso M (2009) A first total synthesis of ganglioside HLG-2. Chemistry 15:4637–4648. https://doi.org/10.1002/chem.200802706
Johannes L (2017) Shiga Toxin-a model for glycolipid-dependent and lectin-driven endocytosis. Toxins 9:340. https://doi.org/10.3390/toxins9110340
Johannes L, Billet A (2020) Glycosylation and raft endocytosis in cancer. Cancer Metastasis Rev 39:375–396. https://doi.org/10.1007/s10555-020-09880-z
Johannes L, Römer W (2010) Shiga toxins-from cell biology to biomedical applications. Nat Rev Microbiol 8:105–116. https://doi.org/10.1038/nrmicro2279
Kaiser RD, London E (1998) Determination of the depth of BODIPY probes in model membranes by parallax analysis of fluorescence quenching. Biochim Biophys Acta Biomembr 1375:13–22. https://doi.org/10.1016/S0005-2736(98)00127-8
Karmali MA, Gannon V, Sargeant JM (2010) Verocytotoxin-producing escherichia coli (VTEC). Vet Microbiol 140:360–370. https://doi.org/10.1016/j.vetmic.2009.04.011
Kiarash A, Boyd B, Lingwood CA (1994) Glycosphingolipid receptor function is modified by fatty acid content. Verotoxin 1 and verotoxin 2c preferentially recognize different globotriaosyl ceramide fatty acid homologues. J Biol Chem 269:11138–11146
Kitov PI, Bundle DR (2001) Synthesis and structure–activity relationships of di- and trisaccharide inhibitors for Shiga-like toxin Type 1. J Chem Soc Perkin Trans 1:838–853. https://doi.org/10.1039/b009685g
Kitov PI, Shimizu H, Homans SW, Bundle DR (2003) Optimization of tether length in nonglycosidically linked bivalent ligands that target sites 2 and 1 of a Shiga-like toxin. J Am Chem Soc 125:3284–3294. https://doi.org/10.1021/ja0258529
Klymchenko AS, Kreder R (2014) Fluorescent probes for lipid rafts: from model membranes to living cells. Chem Biol 21:97–113. https://doi.org/10.1016/j.chembiol.2013.11.009
Kollmitzer B, Heftberger P, Rappolt M, Pabst G (2013) Monolayer spontaneous curvature of raft-forming membrane lipids. Soft Matter 9:10877–10884. https://doi.org/10.1039/C3SM51829A
Komura N, Suzuki KGN, Ando H, Konishi M, Koikeda M, Imamura A, Chadda R, Fujiwara TK, Tsuboi H, Sheng R, Cho W, Furukawa K, Furukawa K, Yamauchi Y, Ishida H, Kusumi A, Kiso M (2016) Raft-based interactions of gangliosides with a GPI-anchored receptor. Nat Chem Biol 12:402–410. https://doi.org/10.1038/nchembio.2059
Konen DA, Silbert LS, Pfeffer PE (1975) Alpha-anions. VII. Direct oxidation of enolate anions to 2-hydroperoxy- and 2-hydroxycarboxylic acids and esters. J Org Chem 40:3253–3258. https://doi.org/10.1021/jo00910a020
Kouzel IU, Pohlentz G, Schmitz JS, Steil D, Humpf H-U, Karch H, Müthing J (2017) Shiga toxin glycosphingolipid receptors in human Caco-2 and HCT-8 colon epithelial cell lines. Toxins 9:338. https://doi.org/10.3390/toxins9110338
Kuerschner L, Ejsing CS, Ekroos K, Shevchenko A, Anderson KI, Thiele C (2005) Polyene-lipids: a new tool to image lipids. Nat Methods 2:39–45. https://doi.org/10.1038/nmeth728
Legros N, Dusny S, Humpf H-U, Pohlentz G, Karch H, Müthing J (2017) Shiga toxin glycosphingolipid receptors and their lipid membrane ensemble in primary human blood-brain barrier endothelial cells. Glycobiology 27:99–109. https://doi.org/10.1093/glycob/cww090
Lewis RN, McElhaney RN, Monck MA, Cullis PR (1994) Studies of highly asymmetric mixed-chain diacyl phosphatidylcholines that form mixed-interdigitated gel phases: Fourier transform infrared and 2H NMR spectroscopic studies of hydrocarbon chain conformation and orientational order in the liquid-crystalline state. Biophys J 67:197–207. https://doi.org/10.1016/S0006-3495(94)80470-9
Ling H, Boodhoo A, Hazes B, Cummings MD, Armstrong GD, Brunton JL, Read RJ (1998) Structure of the shiga-like toxin I B-pentamer complexed with an analogue of its receptor Gb3. Biochemistry 37:1777–1788. https://doi.org/10.1021/bi971806n
Lingwood CA (2011) Glycosphingolipid functions. Cold Spring Harb Perspect Biol 3:a004788. https://doi.org/10.1101/cshperspect.a004788
Lingwood C (2020) Verotoxin receptor-based pathology and therapies. Front Cell Infect Microbiol 10:123. https://doi.org/10.3389/fcimb.2020.00123
Lingwood CA, Binnington B, Manis A, Branch DR (2010) Globotriaosyl ceramide receptor function—where membrane structure and pathology intersect. FEBS Lett 584:1879–1886. https://doi.org/10.1016/j.febslet.2009.11.089
Mahfoud R, Manis A, Lingwood CA (2009) Fatty acid-dependent globotriaosyl ceramide receptor function in detergent resistant model membranes. J Lipid Res 50:1744–1755. https://doi.org/10.1194/jlr.M800385-JLR200
Mahfoud R, Manis A, Binnington B, Ackerley C, Lingwood CA (2010) A major fraction of glycosphingolipids in model and cellular cholesterol-containing membranes is undetectable by their binding proteins. J Biol Chem 285:36049–36059. https://doi.org/10.1074/jbc.M110.110189
Maté S, Busto JV, García-Arribas AB, Sot J, Vazquez R, Herlax V, Wolf C, Bakás L, Goñi FM (2014) N-nervonoylsphingomyelin (C24:1) prevents lateral heterogeneity in cholesterol-containing membranes. Biophys J 106:2606–2616. https://doi.org/10.1016/j.bpj.2014.04.054
Maulik PR, Atkinson D, Shipley GG (1986) X-ray scattering of vesicles of N-acyl sphingomyelins. Determination of bilayer thickness. Biophys J 50:1071–1077. https://doi.org/10.1016/S0006-3495(86)83551-2
Momin N, Lee S, Gadok AK, Busch DJ, Bachand GD, Hayden CC, Stachowiak JC, Sasaki DY (2015) Designing lipids for selective partitioning into liquid ordered membrane domains. Soft Matter 11:3241–3250. https://doi.org/10.1039/c4sm02856b
Morrow MR, Singh D, Lu D, Grant CW (1993) Glycosphingolipid acyl chain orientational order in unsaturated phosphatidylcholine bilayers. Biophys J 64:654–664. https://doi.org/10.1016/S0006-3495(93)81424-3
Morrow MR, Singh D, Grant CWM (1995a) Glycosphingolipid acyl chain order profiles: Substituent effects. Biochim Biophys Acta Biomembr 1235:239–248. https://doi.org/10.1016/0005-2736(95)80010-D
Morrow MR, Singh DM, Grant CW (1995b) Glycosphingolipid headgroup orientation in fluid phospholipid/cholesterol membranes: similarity for a range of glycolipid fatty acids. Biophys J 69:955–964. https://doi.org/10.1016/S0006-3495(95)79969-6
Nakajima H, Kiyokawa N, Katagiri YU, Taguchi T, Suzuki T, Sekino T, Mimori K, Ebata T, Saito M, Nakao H, Takeda T, Fujimoto J (2001) Kinetic analysis of binding between Shiga toxin and receptor glycolipid Gb3Cer by surface plasmon resonance. J Biol Chem 276:42915–42922. https://doi.org/10.1074/jbc.M106015200
Niemelä PS, Hyvönen MT, Vattulainen I (2006) Influence of chain length and unsaturation on sphingomyelin bilayers. Biophys J 90:851–863. https://doi.org/10.1529/biophysj.105.067371
Nyholm PG, Pascher I (1993) Orientation of the saccharide chains of glycolipids at the membrane surface: Conformational analysis of the glucose-ceramide and the glucose-glyceride linkages using molecular mechanics (MM3). Biochemistry 32:1225–1234. https://doi.org/10.1021/bi00056a005
Oberli MA, Bindschädler P, Werz DB, Seeberger PH (2008) Synthesis of a hexasaccharide repeating unit from Bacillus anthracis vegetative cell walls. Org Lett 10:905–908. https://doi.org/10.1021/ol7030262
Okuda T, Tokuda N, Numata S-i, Ito M, Ohta M, Kawamura K, Wiels J, Urano T, Tajima O, Furukawa K, Furukawa K (2006) Targeted disruption of Gb3/CD77 synthase gene resulted in the complete deletion of globo-series glycosphingolipids and loss of sensitivity to verotoxins. J Biol Chem 281:10230–10235. https://doi.org/10.1074/jbc.M600057200
Pagano RE (1990) The Golgi apparatus: insights from lipid biochemistry. Biochem Soc Trans 18:361–366. https://doi.org/10.1042/bst0180361
Patalag LJ, Werz DB (2012) Fluorescent penta- and hexaene fatty acids by a Wittig-Horner/elimination strategy. J Org Chem 77:5297–5304. https://doi.org/10.1021/jo300624h
Patalag LJ, Werz DB (2016) Benzothiadiazole oligoene fatty acids: fluorescent dyes with large Stokes shifts. Beilstein J Org Chem 12:2739–2747. https://doi.org/10.3762/bjoc.12.270
Patalag LJ, Jones PG, Werz DB (2016) BOIMPYs: Rapid access to a family of red-emissive fluorophores and NIR dyes. Angew Chem Int Ed Engl 55:13340–13344. https://doi.org/10.1002/anie.201606883
Patalag LJ, Sibold J, Schütte OM, Steinem C, Werz DB (2017) Gb3 glycosphingolipids with fluorescent oligoene fatty acids: Synthesis and phase behavior in model membranes. ChemBioChem 18:2171–2178. https://doi.org/10.1002/cbic.201700414
Patterson JE, Ollmann IR, Cravatt BF, Boger DL, Wong C-H, Lerner RA (1996) Inhibition of oleamide hydrolase catalyzed hydrolysis of the endogenous sleep-inducing lipid cis -9-octadecenamide. J Am Chem Soc 118:5938–5945. https://doi.org/10.1021/ja954064z
Pawliczek M, Wallbaum J, Werz D (2014) Synthesis of enantiomerically pure α-hydroxylated nervonic acid—a chiral pool approach to α-hydroxylated unsaturated fatty acids. Synlett 25:1435–1437. https://doi.org/10.1055/s-0033-1341276
Pellizzari A, Pang H, Lingwood CA (1992) Binding of verocytotoxin 1 to its receptor is influenced by differences in receptor fatty acid content. Biochemistry 31:1363–1370. https://doi.org/10.1021/bi00120a011
Pezeshkian W, Chaban VV, Johannes L, Shillcock J, Ipsen JH, Khandelia H (2015) The effects of globotriaosylceramide tail saturation level on bilayer phases. Soft Matter 11:1352–1361. https://doi.org/10.1039/c4sm02456g
Róg T, Orłowski A, Llorente A, Skotland T, Sylvänne T, Kauhanen D, Ekroos K, Sandvig K, Vattulainen I (2016) Interdigitation of long-chain sphingomyelin induces coupling of membrane leaflets in a cholesterol dependent manner. Biochim Biophys Acta 1858:281–288. https://doi.org/10.1016/j.bbamem.2015.12.003
Römer W, Berland L, Chambon V, Gaus K, Windschiegl B, Tenza D, Aly MRE, Fraisier V, Florent J-C, Perrais D, Lamaze C, Raposo G, Steinem C, Sens P, Bassereau P, Johannes L (2007) Shiga toxin induces tubular membrane invaginations for its uptake into cells. Nature 450:670–675. https://doi.org/10.1038/nature05996
Safouane M, Berland L, Callan-Jones A, Sorre B, Römer W, Johannes L, Toombes GES, Bassereau P (2010) Lipid cosorting mediated by shiga toxin induced tubulation. Traffic 11:1519–1529. https://doi.org/10.1111/j.1600-0854.2010.01116.x
Schnaar RL, Kinoshita T (2015) Essentials of glycobiology: glycosphingolipids, vol 3. Cold Spring Harbor, NY
Schubert T, Sych T, Madl J, Xu M, Omidvar R, Patalag LJ, Ries A, Kettelhoit K, Brandel A, Mely Y, Steinem C, Werz DB, Thuenauer R, Römer W (2020) Differential recognition of lipid domains by two Gb3-binding lectins. Sci Rep 10:9752. https://doi.org/10.1038/s41598-020-66522-8
Schütte OM, Ries A, Orth A, Patalag LJ, Römer W, Steinem C, Werz DB (2014) Influence of Gb3 glycosphingolipids differing in their fatty acid chain on the phase behaviour of solid supported membranes: Chemical syntheses and impact of Shiga toxin binding. Chem Sci 5:3104–3114. https://doi.org/10.1039/c4sc01290a
Schütte OM, Patalag LJ, Weber LMC, Ries A, Römer W, Werz DB, Steinem C (2015) 2-Hydroxy fatty acid enantiomers of Gb3 impact Shiga Toxin binding and membrane organization. Biophys J 108:2775–2778. https://doi.org/10.1016/j.bpj.2015.05.009
Sengupta P, Hammond A, Holowka D, Baird B (2008) Structural determinants for partitioning of lipids and proteins between coexisting fluid phases in giant plasma membrane vesicles. Biochim Biophys Acta 1778:20–32. https://doi.org/10.1016/j.bbamem.2007.08.028
Sezgin E, Levental I, Grzybek M, Schwarzmann G, Mueller V, Honigmann A, Belov VN, Eggeling C, Coskun U, Simons K, Schwille P (2012) Partitioning, diffusion, and ligand binding of raft lipid analogs in model and cellular plasma membranes. Biochim Biophys Acta 1818:1777–1784. https://doi.org/10.1016/j.bbamem.2012.03.007
Sibold J, Kettelhoit K, Vuong L, Liu F, Werz DB, Steinem C (2019) Synthesis of Gb3 glycosphingolipids with labeled head groups: distribution in phase-separated giant unilamellar vesicles. Angew Chem Int Ed Engl 58:17805–17813. https://doi.org/10.1002/anie.201910148
Slotte JP (2013) Molecular properties of various structurally defined sphingomyelins—correlation of structure with function. Prog Lipid Res 52:206–219. https://doi.org/10.1016/j.plipres.2012.12.001
Somerharju P (2002) Pyrene-labeled lipids as tools in membrane biophysics and cell biology. Chem Phys Lipids 116:57–74. https://doi.org/10.1016/S0009-3084(02)00020-8
Souto AA, Acuña AU, Amat-Guerri F (1994) A general and practical synthesis of linear conjugated pentaenoic acids. Tetrahedron Lett 35:5907–5910. https://doi.org/10.1016/S0040-4039(00)78216-5
Sullan RMA, Li JK, Hao C, Walker GC, Zou S (2010) Cholesterol-dependent nanomechanical stability of phase-segregated multicomponent lipid bilayers. Biophys J 99:507–516. https://doi.org/10.1016/j.bpj.2010.04.044
Takahashi H, Hayakawa T, Kawasaki Y, Ito K, Fujisawa T, Kodama M, Kobayashi T (2007) Structural characterization of N -lignoceroyl (C24:0) sphingomyelin bilayer membranes: a re-evaluation. J Appl Crystallogr 40:s312–s317. https://doi.org/10.1107/S002188980700859X
van Meer G, de Kroon AIPM (2011) Lipid map of the mammalian cell. J Cell Sci 124:5–8. https://doi.org/10.1242/jcs.071233
Veatch SL, Soubias O, Keller SL, Gawrisch K (2007) Critical fluctuations in domain-forming lipid mixtures. Proc Natl Acad Sci USA 104:17650–17655. https://doi.org/10.1073/pnas.0703513104
Watkins EB, Gao H, Dennison AJC, Chopin N, Struth B, Arnold T, Florent J-C, Johannes L (2014) Carbohydrate conformation and lipid condensation in monolayers containing glycosphingolipid Gb3: Influence of acyl chain structure. Biophys J 107:1146–1155. https://doi.org/10.1016/j.bpj.2014.07.023
Werz DB, Seeberger PH (2005) Total synthesis of antigen bacillus anthracis tetrasaccharide–creation of an anthrax vaccine candidate. Angew Chem Int Ed Engl 44:6315–6318. https://doi.org/10.1002/anie.200502615
Werz DB, Castagner B, Seeberger PH (2007) Automated synthesis of the tumor-associated carbohydrate antigens Gb-3 and Globo-H: incorporation of alpha-galactosidic linkages. J Am Chem Soc 129:2770–2771. https://doi.org/10.1021/ja069218x
Windschiegl B, Orth A, Römer W, Berland L, Stechmann B, Bassereau P, Johannes L, Steinem C (2009) Lipid reorganization induced by Shiga toxin clustering on planar membranes. PLoS ONE 4:e6238. https://doi.org/10.1371/journal.pone.0006238
The authors are grateful to the DFG (SFB 803, project A05) for financial support. C.S. thanks Ole M. Schütte, Loan Vuong, Fangyuan Li, and Jutta Gerber Nolte for their experimental work. D. B. W. thanks Annika Ries, Lukas J. Patalag and Katharina Kettelhoit for their synthetic efforts. Open Access funding provided by Projekt DEAL.
Conflict of interest
The authors declare no conflict of interest.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Special Issue: Multicomponent lipid membranes.
About this article
Cite this article
Sibold, J., Ahadi, S., Werz, D.B. et al. Chemically synthesized Gb3 glycosphingolipids: tools to access their function in lipid membranes. Eur Biophys J 50, 109–126 (2021). https://doi.org/10.1007/s00249-020-01461-w
- Fluorescence microscopy
- Model membranes
- Organic synthesis
- Shiga toxin