• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Sep 2, 2003; 100(18): 10219–10224.
Published online Aug 15, 2003. doi:  10.1073/pnas.1737042100
PMCID: PMC193542

Glutamate-binding affinity of Drosophila metabotropic glutamate receptor is modulated by association with lipid rafts


Metabotropic glutamate receptors (mGluRs) are responsible for the effects of glutamate in slow synaptic transmission, and are implicated in the regulation of many processes in the CNS. Recently, we have reported the expression and purification of a mGluR from Drosophila melanogaster (DmGluRA), a homologue of mammalian group II mGluRs. We have shown that ligand binding to reconstituted DmGluRA requires the presence of ergosterol in the liposomes [Eroglu, C., Cronet, P., Panneels, V., Beaufils, P. & Sinning, I. (2002) EMBO Rep. 3, 491-496]. Here we demonstrate that the receptor exists in different affinity states for glutamate, depending on the membrane composition. The receptor is in a high-affinity state when associated with sterol-rich lipid microdomains (rafts), and in a low-affinity state out of rafts. Enrichment of the membranes with cholesterol shifts the receptor into the high-affinity state, and induces its association with rafts. The receptor was crosslinked to photocholesterol. Our data suggest that sterol-rich lipid rafts act as positive allosteric regulators of DmGluRA.

Keywords: cholesterol, metabotropic glutamate receptor, nano-electrospray ionization tandem MS, regulation of G protein-coupled receptors

Metabotropic glutamate receptors (mGluRs) are implicated in the regulation of many physiological and pathological processes in the CNS, including synaptic plasticity, learning and memory, motor coordination, pain transmission, and neurodegeneration (reviewed in refs 1 and 2). Glutamate also functions as a neurotransmitter in the invertebrate CNS (3), where it performs both excitatory and inhibitory actions (4, 5). The first Drosophila mGluR cloned, (DmGluRA), shares the structural features characteristic for mammalian mGluRs (6). Recently, we have reported the expression of DmGluRA in insect cells by recombinant baculovirus infection (7), and in transgenic Drosophila melanogaster (8), where the receptor is directed to the photoreceptor cells in the fruit fly eye, and is colocalized with rhodopsin in specialized membrane stacks, called rhabdomeres.

Unlike many other G protein-coupled receptors (GPCRs), which possess a ligand-binding pocket in their heptahelical transmembrane domain (7TMD), mGluRs have a very large extracellular domain responsible for ligand binding (1). The glutamate-binding domains of mGluRs have been expressed as soluble secreted receptors (9-11). The crystal structure of the glutamate-binding domain of mGluR1 shows the presence of two globular domains separated by a hinge region (12). The receptor exists in two different conformations, open (resting) and closed (active), in the ligand-free form. Ligand (glutamate) binding stabilizes the closed (active) conformation rather than inducing it (12, 13). It is still not known for the full-length receptor how the ligand-binding domain (LBD) is functionally coupled to the 7TMD. Recently, a model for the crosstalk between these two domains has been presented (14). This model postulates the existence of modulators that act on the 7TMD, which in turn influences ligand binding. Interestingly, we have previously (8) shown for reconstituted DmGluRA, that ligand binding depends on the presence of ergosterol in the liposomes. This finding suggests that a sterol-7TMD interaction is involved in the regulation of ligand binding.

Sterols are known to affect the function of various membrane proteins (reviewed in ref. 15), including GPCRs (16, 17). Sterols might mediate their action on membrane proteins, either indirectly by modulating the biophysical properties of the lipid bilayer, or by a direct interaction with the receptor acting as a cofactor (15, 17). Sterols are known to promote phase separation in membranes, thus playing an important role in the formation of lipid microdomains, which are also known as lipid rafts (reviewed in refs. 18 and 19). Detergent-resistant sterolsphingolipid-enriched membranes (DRMs), which can be isolated as a light (floating) fraction by using density gradients after detergent extraction at low temperatures, are derived from these lipid rafts (18, 20). Partitioning in and out of lipid rafts is thought to be involved in the sorting and regulation of membrane proteins (21). In addition, lipid rafts are implicated in the establishment of polarity in certain cell types (18, 22), including neurons (23). In this article, we show that sterols and rafts are involved in the regulation of DmGluRA, a key molecule for synaptic transmission.

Materials and Methods

Preparation of Membrane Fractions. Enriched plasma membrane fractions were prepared from transgenic Drosophila melanogaster heads expressing DmGluRA in the photoreceptor cells (DrDGR membranes), and insect cells (Sf9) infected by a recombinant baculovirus expressing DmGluRA (Sf9DGR membranes) (8). Crude lysate from transgenic Drosophila heads or postnuclear supernatant from insect cells were mixed 1:2 with OptiPrep (Nycomed, Oslo) resulting in 40% solution; the 30% and 5% OptiPrep solutions were laid on top. Plasma membrane enriched fraction formed a dense band at 30-5% interface after 3 h of centrifugation at 100,000 × g.

Immunodetection of DmGluRA. DmGluRA was detected by a polyclonal antibody against a peptide in the extracellular domain of the receptor (described in ref. 8). Horseradish peroxidase-conjugated secondary antibodies (Sigma) were used for detection and the blots were developed by an enhanced chemiluminescence kit (Amersham Pharmacia). Chemiluminescence was detected on BioMax MR1 films (Kodak).

Reconstitution of DmGluRA. DmGluRA expressed in Sf9 suspension cells was purified by an immunoaffinity column as described in refs. 7 and 8. Lipids dioleoyl-phosphatidylcholine (DOPC), dipalmitoyl-phosphatidylcholine (DPPC), and cholesterol were purchased from Avanti Polar Lipids. Reconstitution was performed as described in ref. 8.

Glutamate Binding Assay. Homologous competitive binding experiments were performed on Sf9DGR and DrDGR membranes, as described (7, 8), in the presence of varying concentrations of cold l-glutamate (10-3 to 10-8 M) to determine the affinity of the receptor to glutamate. Membranes (20-40 μg) were incubated with 1.5 μM tritiated glutamate (Amersham Pharmacia Biotech) in 20 mM Tris, (pH 7.4) buffer containing 0.2% BSA for 2 h at 25°C. The binding mixtures were filtered through nitrocellulose membrane filters (pore size 0.45 μm, 25-mm diameter; Sigma), and the filters were washed four times with 2 ml of 50 mM Tris (pH 7.4) 250 mM NaCl/0.1% BSA. The radioactivity on the filters was counted. Reconstituted receptor (1 μg) was tested for binding in the same way, but with different washing stringencies.

The total binding and IC50 values were calculated by a GraphPad (San Diego) program, after entering the binding data (dpm) versus the logarithm of concentration of cold l-glutamate. The y axis was then converted to percent total binding to simplify comparison.

Enrichment of Membrane Sterol Content. Methyl-β-cyclodextrin (MβCD; Sigma)-cholesterol complexes were formed as described in ref. 24. Isolated plasma membrane fraction was treated with different concentrations of MβCD-cholesterol complexes as described in ref. 16. After treatment, membranes corresponding to 2 μg of membrane protein were analyzed with a cholesterol oxidase assay kit (Molecular Probes) to check sterol incorporation, following the manufacturer's protocol. The maximum cholesterol incorporation (≈8-fold of the original sterol content) was reached when membranes were treated with 240 μM of the complex.

Isolation of Light-Density DRMs. The protocol was adapted from ref. 25. Membrane protein (800 μg) was solubilized with 1% Triton X-100 (TX-100, Serva, Germany) in, and the extraction mixture was mixed 1:2 with OptiPrep; the 30% and 0% OptiPrep solutions were laid on the top (1.2 and 0.2 ml, respectively). After 2 h of spinning at 55,000 rpm (Beckman TLS55 rotor) at 4°C, 350-μl fractions were taken from top to bottom. Ten-microliter samples from each fraction were analyzed by Western blotting.

Crosslinking Cholesterol to DmGluRA. Photoactivatable analogue of cholesterol, photocholesterol, radioactively labeled with tritium, and in complex with MβCD (26), was a kind gift from C. Thiele (Max Planck Institute of Cell Biology and Genetics, Dresden, Germany). Isolated plasma membranes from fruit fly heads (1 mg of total membrane protein) or insect cells (3 mg of total membrane protein) were incubated with MβCD-[3H]photocholesterol complex (0.3 mCi, 2 mg; 1 Ci = 37 GBq). The photocholesterol was crosslinked by irradiation by using the filtered (λ > 310 nm) beam of a high-pressure mercury lamp. One-tenth of the membranes were analyzed as total membrane protein, and the remainder was immunoprecipitated (pulled down) by using 50 μl of 7G11-Sepharose (described in ref. 8). The total membrane protein fractions and the pull-down fractions were analyzed by SDS/7.5% PAGE. The labeled proteins were detected by autoradiography on x-ray films (Hyperfilm, Amersham Pharmacia).

Lipid Analysis. Membrane pellets were extracted, according to the method of Bligh and Dyer (27). After solvent evaporation, samples were resuspended in methanol, and were further processed for nano-electrospray ionization tandem MS (nano-ESI-MS/MS) as described (28). Cholesterol and ergosterol analysis of the sulfated sterols was performed as described (29).

For more information, see Supporting Materials and Methods, which is published as supporting information on the PNAS web site, www.pnas.org.


DmGluRA Has Different Affinities to Glutamate, Depending on the Membrane Environment. Previously, we have described the overexpression of DmGluRA in Drosophila photoreceptor cells and in Sf9 cells (7, 8). We have purified and reconstituted DmGluRA, and showed that the ligand binding depends on the presence of sterols in the liposomes (8). To try to understand how sterols affect the activity of DmGluRA, we decided to further characterize the receptor. To achieve this end, we determined the affinity of the receptor to glutamate. Competitive binding assays were performed on both DrDGR membranes and Sf9DGR membranes.

Interestingly, DmGluRA expressed in Sf9DGR membranes had >10-fold higher affinity to glutamate than did the same receptor expressed in the fruit fly eye in DrDGR membranes, (IC50 = 4.17 versus 54.5 μM; Fig. 1).

Fig. 1.
DmGluRAs have different affinity to glutamate when expressed in Sf9 cells or in fruit fly heads. Competitive glutamate binding assays on Sf9DGR membranes (A) and DrDGR membranes (B) show that the receptor expressed in Sf9DGR membranes has 10-fold higher ...

Because our previous study pointed out the importance of the lipid environment in ligand binding to the receptor (8), we tested whether the cause of this dramatic difference in affinity was due to differences in the lipid environment. To make this determination, we analyzed the lipid composition of the membranes from both sources by nano-ESI-MS/MS (30). The fatty acyl chains of phospholipids in DrDGR membranes were shorter and less saturated, whereas in Sf9DGR membranes, phospholipids contained longer and more saturated acyl chains (Table 1). The main sterol species from both membranes were different as shown (31-32). DrDGR membranes contain ergosterol, whereas Sf9DGR membranes contain cholesterol (see Fig. 7A, which is published as supporting information on the PNAS web site). The amounts of sterol per mg of membrane protein were similar in both membranes (see Table 3, which is published as supporting information on the PNAS web site). Sf9DGR membranes contain sphingomyelin and other sphingolipids (Table 2 and Fig. 7C).

Table 1.
Main PC and PE species of DrDGR and Sf9DGR membranes
Table 2.
Main sphingolipid species of Sf9DGR and DrDGR membranes

In contrast, DrDGR membranes lack sphingomyelin, but contain other sphingolipid species [glucosylceramides (GlcCer), 14:1;18:0 and 14:1;20:0 and phosphoethanolamine Cer (PECer), 14:1;20:0)], including hydroxylated GlcCer species: GlcCer 20:0-OH (73.6%) and GlcCer 22:0-OH (26.5%; Table 2 and Fig. 7C).

The High-Affinity State of DmGluRA Correlates with Its Association with DRMs. Sterols associate with the long and saturated fatty acyl chains of sphingolipids and phospholipids, and form tightly packed lipid microdomains, which are also known as lipid rafts. Lipid rafts are thought to be the source of DRMs. Taking into account the results of our lipid analysis, we asked whether the differences in affinity we observed could be the result of a biophysical state of the membrane, rather than the sterol content per se. To confirm our findings, we tested whether DmGluRA associates with DRMs. Sf9DGR and DrDGR membranes were treated with TX-100, and subjected to density gradient centrifugation (see Materials and Methods). Western Blot analysis of fractions from the density gradient showed that DmGluRA was mostly DRM-associated (floating to the light density fraction) in Sf9DGR membranes (Fig. 2A), whereas it was mostly not DRM-associated (in the bottom fractions) in DrDGR membranes (Fig. 2B).

Fig. 2.
DRM association of DmGluRA. (A) In Sf9DGR membranes, DmGluRA floats to the light density fractions (lane 1) after TX-100 extraction on ice, indicating that the receptor is associated with DRMs. (B) In DrDGR membranes, the receptor is mostly soluble ...

These results, together with the affinity data from DrDGR and Sf9DGR membranes, suggest that raft association is involved in the modulation of the affinity of the receptor to glutamate.

To further test this conclusion, we decided to increase the sterol content of DrDGR membranes. This end can be achieved by treatment of membranes with preformed water-soluble inclusion complexes of MβCD with cholesterol (33). Ergosterol could not be used because it failed to form stable complexes with MβCD. Increasing concentrations of MβCD-cholesterol complex treatment resulted in increase of membrane 3-hydroxyl sterol content up to 8-fold in DrDGR membranes (see Materials and Methods). Competitive glutamate binding assays were performed on these membranes after treatment. The analysis of the binding curve showed an ≈50-fold increase in affinity to glutamate (IC50 = 1.2 μM; Fig. 3A). Interestingly, the total binding was also increased (up to 2.5-fold) after sterol treatment. This increase in the total binding could be due to either the activation of previously inactive receptor present in the membranes by the addition of sterols, or to the increase in affinity as the binding assay is done with [3H]glutamate concentrations (1.5 μM) close to the IC50 values. Next, DRM association of DmGluRA was examined in these membranes. On treatment with MβCD-cholesterol, a significant portion of the receptor was incorporated into DRMs (Fig. 3B). Together, these results suggest a role of raft association in the regulation of the affinity of the receptor to glutamate.

Fig. 3.
Affinity of DmGluRA to glutamate depends on its DRM association. (A) Competitive binding assays on DrDGR membranes treated with 240 μM MβCD-cholesterol complex shows that increase in membrane sterol content shifts DmGluRA to a high-affinity ...

DmGluRA Directly Interacts with Cholesterol. We have shown that the sterol enrichment of DrDGR membranes increases the affinity of DmGluRA to glutamate, and induces its partition into DRMs. Next, we asked whether DmGluRA directly interacts with sterols. To confirm this possibility, we used a photoactivatable analogue of cholesterol, known as photocholesterol, that has previously been used for labeling proteins with strong affinity to cholesterol (26). DrDGR and Sf9DGR membranes were treated with MβCD-[3H]photocholesterol complex, and is crosslinked by light activation (see Materials and Methods). After labeling, total membrane proteins from these membranes were analyzed by SDS/PAGE, and the labeled proteins were detected by autoradiography (Fig. 4A, lanes 1-3). A strong band corresponding to the monomer size of the receptor was visualized after SDS/PAGE in DrDGR membranes labeled with [3H]photocholesterol. A protein band of identical apparent molecular mass was also present in the fraction pulled down by a DmGluRA-specific monoclonal antibody (7G11) conjugated to Sepharose (described in ref. 8) from these membranes (Fig. 4A, lane 4), demonstrating that this band corresponds to DmGluRA. In Sf9 membranes, labeling of DmGluRA in Sf9DGR membranes was harder to distinguish from the background (Fig. 4A, lanes 2 and 3). However, the labeled receptor was detected in the pull-down fraction from Sf9DGR membranes (Fig. 4A, lane 5), whereas no such band was present in the pull-down fraction from uninfected Sf9 cell membranes (Fig. 4A, lane 6), thus demonstrating that the band that was seen having the correct apparent molecular mass belongs to DmGluRA. Interestingly, Western blot analysis of the same fractions (Fig. 4B) showed that on labeling with photocholesterol, the receptor ran mainly as a monomer in SDS/PAGE, and that there was a slight difference in migration between the receptor from DrDGR and Sf9DGR membranes. The pattern of DmGluRA recognition in Western blot analysis matches exactly with the autoradiography signals, thus confirming that the suspected band in autoradiography represents DmGluRA. DmGluRA is, to our knowledge, the first GPCR known to crosslink to cholesterol. This result suggests a direct interaction between the receptor and cholesterol.

Fig. 4.
Photoactivatable analogue of cholesterol (photocholesterol) crosslinks to DmGluRA. DrDGR, Sf9DGR, and uninfected Sf9 cell membranes were treated with MβCD-photocholesterol complex. (A) Total membrane and 7G11 pull-down fractions were analyzed ...

Partition of DmGluRA Into Lipid Rafts Modulates Its Binding Affinity.

We have shown that sterols are required for the activity of DmGluRA, and that the high-affinity state of the receptor correlates with its association with lipid rafts. To distinguish whether the effect of sterols on the binding properties of DmGluRA depends only on the sterol concentration in the membrane or in the ability of sterols to induce the formation of lipid rafts, we have reconstituted purified DmGluRA into liposomes of two different compositions. Both types of liposomes had the same cholesterol concentration (40% molar). The first type (non-raft; NR) had 60% unsaturated DOPC. The second type (raft-like; R) contained 30% saturated DPPC, as well as 30% DOPC. The absence of saturated acyl chains in NR liposomes impairs formation of sterol-rich lipid microdomains. The saturated acyl chains of DPPC form a complex with cholesterol, thus promoting the formation of lipid rafts (18, 34, 35). DmGluRA was reconstituted into NR and R liposomes with similar efficiency (data not shown). Glutamate binding experiments were performed on these proteoliposomes (see Materials and Methods). Because the amount of purified reconstituted receptor was limiting, we could not perform competitive binding assays as before. Instead, we measured the specific glutamate binding under low- and high-stringency conditions. Specific binding of glutamate was detected for both types of liposomes when the assay was performed with low stringency (two washes; Fig. 5). The specific binding to DmGluRA reconstituted in R liposomes was 1.3-fold higher than that of the receptor reconstituted in NR liposomes (Fig. 5). Interestingly, the specific binding in NR liposomes was dramatically reduced when the stringency of the binding assay was increased by adding a third washing step. The third wash did not diminish the specific binding in R liposomes (Fig. 5, ≈10-fold more specific binding in R liposomes compared with NR liposomes). This result suggests that the receptor in NR liposomes had lower affinity than the receptor in R liposomes, because the radioactive ligand could be displaced more easily from the receptor in NR liposomes. If the affinity would depend only on the membrane sterol concentration, there should be no difference. Therefore, we conclude that the affinity of the receptor is not regulated by the sterol content alone, but rather by the partitioning of the receptor into lipid rafts.

Fig. 5.
Radioactive glutamate binding to proteoliposomes [60% DOPC: 40% cholesterol non-raft (NR), and 30% DOPC: 30% DPPC: 40% cholesterol raft-like (R)] were measured by rapid filtration assays (see Materials and Methods). Excess glutamate (1 mM) was added ...


Affinity of DmGluRA to Glutamate Is Regulated by DRM Association. We have previously shown (8) that glutamate binding to the purified DmGluRA strictly depends on the presence of ergosterol. In this article, we show that the receptor has different affinity states, depending on the lipid environment, and we link the high-affinity state of DmGluRA to its association with sterol-rich membrane microdomains. We show that the high-affinity state of DmGluRA is induced by its association with lipid rafts. In addition, we show a direct interaction of cholesterol with DmGluRA by crosslinking. This interaction is most likely the major driving force for its DRM association.

The oxytocin receptor is the only other GPCR for which it is known that ligand binding also depends on the presence of sterols (36-38). This receptor belongs to the rhodopsin-like GPCR family, which possesses a ligand-binding pocket in its transmembrane domain. Two distinct affinity states of the receptor have been reported, and the high-affinity state is facilitated by increasing the cholesterol content (16, 33, 37, 39). However, the oxytocin receptor does not associate with DRMs (38), and no direct interaction with sterols has been reported. Therefore, sterols might modulate the function of GPCRs in different ways. Our results indicate that sterols are required for ligand binding of DmGluRA; however, the affinity for the ligand is modulated, not by the sterol content itself, but rather, by the ability of sterols to induce raft formation (Fig. 5).

DRM Association of DmGluRA. Raft formation depends on the relative ratios of sterols, sphingolipids, and the degree of fatty acyl saturation (18). Lipid analysis of Sf9DGR and DrDGR membranes showed significant differences in composition. Whereas Sf9DGR membranes contain sphingomyelin and saturated phospholipid species, DrDGR membranes lack sphingomyelin, and are enriched in glycerophospholipids with short and unsaturated acyl chains (Table 1, Table 3, and Fig. 7 B and C). Consequently, DmGluRA was associated with DRMs in Sf9DGR, but not in DrDGR membranes (Fig. 2). Even though the sterol species are different in these two membranes, biophysical studies on lipid bilayers have shown that ergosterol and cholesterol affect the lipid bilayer properties in a similar fashion (18, 35), and ergosterol has been shown to have even higher propensity to form rafts (35). However, the shorter and more unsaturated fatty acyl chains in DrDGR membranes (Table 1) would reduce the driving force of ergosterol to induce phase separation. In addition, lack of sphingomyelin might affect the extent of raft formation. However, certain ceramide species that are enriched in fruit fly head DRMs might partially replace sphingomyelin.

The fact that DmGluRA expressed in DrDGR membranes is not associated with DRMs cannot be explained simply by the lack of DRMs in these membranes, as lipid analysis showed their presence (Table 3 and Fig. 7 B and C). DrDGR membranes are not only derived from photoreceptor cells, but they are also derived from other cell types, such as nerve cells and the brain. In fact, rhabdomeres from fruit fly photoreceptor cells where DmGluRA is expressed (8) must be optimized for rhodopsin function, which is facilitated by the presence of unsaturated phospholipids, and is impaired by higher concentrations of cholesterol (40, 41). Studies on rod disk membranes (mammalian rhabdomeres) showed that they are poor in cholesterol and sphingolipids, and contain high amounts of polyunsaturated phospholipids (42), and therefore do not support raft formation (17). However phase separation can be induced by cholesterol enrichment of these membranes by using MβCD-cholesterol complexes (17). Our data are in line with these observations. We could induce the association of DmGluRA with DRMs in DrDGR membranes by adding cholesterol (Fig. 3B). In addition, the higher crosslinking efficiency of photocholesterol to DmGluRA in these membranes suggests that rhabdomeres, like rod disk membranes, contain a low level of sterols (Fig. 4A). However, in insect cell membranes, the crosslinking efficiency is lower as DmGluRA is in rafts, and therefore, photocholesterol must compete with endogenous sterols.

Positive Allosteric Modulation of DmGluRA by Raft Association. According to the current model of ligand binding to mGluRs (12-14, 43), the LBD oscillates between closed (active) and open (resting) conformations, and the binding of glutamate stabilizes the closed conformation. Similarly, the 7TMD also oscillates between different states (active and resting), which are proposed to depend on the conformation of the LBD. The reverse should also apply, suggesting that an allosteric interaction exists between these two domains (14). Therefore, the ligand-binding affinity can be altered by modulating the frequency of oscillation between the active and resting states of the 7TMD. We show that the ligand-binding properties of DmGluRA critically depend on sterols and lipid rafts. We suggest that both modulate ligand binding by acting on the 7TMD. In the absence of sterols, the receptor is trapped in the resting state (no binding of glutamate), and therefore is inactive. The association of the receptor with sterol-rich or -poor domains alters the equilibrium between the active and resting states (Fig. 6). For the raft-associated receptor, the active form is favored, and out of rafts the resting form is favored. Therefore, the affinity of the receptor to glutamate is high when in rafts, and is low when out of rafts. In conclusion, we propose that the association with sterol-rich lipid microdomains acts as a positive allosteric regulator of DmGluRA.

Fig. 6.
Model for the regulation of ligand binding to DmGluRA by sterols. The receptor oscillates between an active and a resting state. Ligand binding shifts the equilibrium to the active state by stabilizing the LBD (12, 14). Our data suggest that sterol-rich ...

Physiological Role of Raft Association of mGluRs. DmGluRA is homologous to group II mammalian mGluRs (6, 44). Group II mGluRs (2/3 presynaptic), like DmGluRA, are associated with rafts in a sterol-dependent manner (see Fig. 8, which is published as supporting information on the PNAS web site). In addition, different affinity states of mGluR2 have been reported when expressed in different cell types (1), which is analogous to our observations on DmGluRA. This finding implies that ligand binding to group II mGluRs might be regulated by sterols in a similar way.

A recent study (45) suggests that mGluR1α (group I, postsynaptic) is not in rafts, whereas evidence is provided on the raft association of metabotropic GABA receptors (GABAB), which are localized presynaptically similar to group II mGluRs. Therefore, it is possible that raft association (in a cholesterol-dependent manner) is a characteristic of presynaptic metabotropic neurotransmitter receptors, and is involved in their sorting. Indeed, there is substantial evidence that rafts play a role in axonal polarization of nerve cells (46, 47). In addition, lipid rafts have been recently shown to facilitate the long-range axonal transport of synaptic vesicle precursors by inducing dimerization of Unc104 motor protein (48).

Raft association of DmGluRA must be a dynamic process to have a regulatory function. One of the possibilities of regulation would be the modification of the membrane composition at the presynaptic region. Even though there is very limited data on the dynamics of lipid composition at the synapse, there is evidence that lipids are playing important roles in the formation and regulation of endocytosis and exocytosis at these sites (49). In addition, cholesterol is an important constituent of the synaptic vesicles (26).

Interestingly, cholesterol has been identified as a factor released by glial cells that increases synapse number and efficacy (50-52). This result suggests cholesterol-dependent mechanisms in the regulation of synaptic transmission, and plasticity in mammalian CNS (53). mGluRs are primarily involved in these processes. Therefore, by regulating the cholesterol level of presynaptic membranes, glial cells might modulate the ligand-binding characteristics of presynaptic mGluRs (group II), and facilitate synaptic transmission and plasticity.

Supplementary Material

Supporting Information:


We thank C. Thiele (Max Planck Institute, Dresden, Germany) for photocholesterol, and for helpful discussions concerning the crosslinking experiment; V. Panneels, M. Bagnat, and P. Osten for help, advice, and discussions; and A. Kuhn, B. Segnitz, and I. Leibrecht for excellent technical assistance. This work has been supported by a European Union QLK-3CT-grant (to I.S.), and a German Research Foundation grant (Deutsche Forschungsgemeinschaft) (to B.B. and F.W.). Ç.E. is supported by the European Molecular Biology Laboratory Ph.D. program.


This paper was submitted directly (Track II) to the PNAS office.

Abbreviations: mGluR, metabotropic glutamate receptor; DmGluRA, Drosophila melanogaster GluR; GPCR, G protein-coupled receptor; 7TMD, heptahelical transmembrane domain; DRM, detergent-resistant membrane; DrDGR membranes, fruit fly head membranes from fruit flies expressing DmGluRA in the photoreceptor cells; Sf9DGR membranes, Sf9 cell membranes expressing the receptor by recombinant baculovirus infection; DOPC, dioleoylphosphatidylcholine; MβCD; methyl-β-cyclodextrin; R, raft-like; NR, non-raft; LBD, ligand-binding domain.


1. Conn, P. J. & Pin, J. P. (1997) Annu. Rev. Pharmacol. Toxicol. 37, 205-237. [PubMed]
2. De Blasi, A., Conn, P. J., Pin, J. & Nicoletti, F. (2001) Trends Pharmacol. Sci. 22, 114-120. [PubMed]
3. Bicker, G., Schafer, S., Ottersen, O. P. & Storm-Mathisen, J. (1988) J. Neurosci. 8, 2108-2122. [PubMed]
4. Bolshakov, V., Gapon, S. A. & Magazanik, L. G. (1991) J. Physiol. 439, 15-35. [PMC free article] [PubMed]
5. Kehoe, J. (1994) Neuron 13, 691-702. [PubMed]
6. Parmentier, M. L., Pin, J. P., Bockaert, J. & Grau, Y. (1996) J. Neurosci. 16, 6687-6694. [PubMed]
7. Panneels, V., Eroglu, C., Cronet, P. & Sinning, I. (2003) Protein Expression Purif. 30, 275-282. [PubMed]
8. Eroglu, C., Cronet, P., Panneels, V., Beaufils, P. & Sinning, I. (2002) EMBO Rep. 3, 491-496. [PMC free article] [PubMed]
9. Okamoto, T., Sekiyama, N., Otsu, M., Shimada, Y., Sato, A., Nakanishi, S. & Jingami, H. (1998) J. Biol. Chem. 273, 13089-13096. [PubMed]
10. Peltekova, V., Han, G., Soleymanlou, N. & Hampson, D. R. (2000) Brain Res. Mol. Brain Res. 76, 180-190. [PubMed]
11. Han, G. & Hampson, D. R. (1999) J. Biol. Chem. 274, 10008-10013. [PubMed]
12. Kunishima, N., Shimada, Y., Tsuji, Y., Sato, T., Yamamoto, M., Kumasaka, T., Nakanishi, S., Jingami, H. & Morikawa, K. (2000) Nature 407, 971-977. [PubMed]
13. Tsuchiya, D., Kunishima, N., Kamiya, N., Jingami, H. & Morikawa, K. (2002) Proc. Natl. Acad. Sci. USA 99, 2660-2665. [PMC free article] [PubMed]
14. Parmentier, M. L., Prezeau, L., Bockaert, J. & Pin, J. P. (2002) Trends Pharmacol. Sci. 23, 268-274. [PubMed]
15. Burger, K., Gimpl, G. & Fahrenholz, F. (2000) Cell Mol. Life Sci. 57, 1577-1592. [PubMed]
16. Gimpl, G., Burger, K. & Fahrenholz, F. (1997) Biochemistry 36, 10959-10974. [PubMed]
17. Niu, S. L., Mitchell, D. C. & Litman, B. J. (2002) J. Biol. Chem. 277, 20139-20145. [PubMed]
18. Brown, D. A. & London, E. (1998) J. Membr. Biol. 164, 103-114. [PubMed]
19. Simons, K. & Ikonen, E. (1997) Nature 387, 569-572. [PubMed]
20. van Meer, G. (2002) Science 296, 855-857. [PubMed]
21. Ikonen, E. (2001) Curr. Opin. Cell Biol. 13, 470-477. [PubMed]
22. Simons, K. & Toomre, D. (2000) Nat. Rev. Mol. Cell Biol. 1, 31-39. [PubMed]
23. Tsui-Pierchala, B. A., Encinas, M., Milbrandt, J. & Johnson, E. M. (2002) Trends Neurosci. 25, 412-417. [PubMed]
24. Keller, P. & Simons, K. (1998) J. Cell Biol. 140, 1357-1367. [PMC free article] [PubMed]
25. Bagnat, M., Keranen, S., Shevchenko, A. & Simons, K. (2000) Proc. Natl. Acad. Sci. USA 97, 3254-3259. [PMC free article] [PubMed]
26. Thiele, C., Hannah, M. J., Fahrenholz, F. & Huttner, W. B. (2000) Nat. Cell Biol. 2, 42-49. [PubMed]
27. Bligh, E. G. & Dyer, W. J. (1959) Can. J. Biochem. Physiol. 37, 911-917. [PubMed]
28. Bruegger, B., Sandhoff, R., Wegehingel, S., Gorgas, K., Malsam, J., Helms, J. B., Lehmann, W. D., Nickel, W. & Wieland, F. T. (2000) J. Cell Biol. 151, 507-518. [PMC free article] [PubMed]
29. Sandhoff, R., Bruegger, B., Jeckel, D., Lehmann, W. D. & Wieland, F. T. (1999) J. Lipid Res. 40, 126-132. [PubMed]
30. Bruegger, B., Erben, G., Sandhoff, R., Wieland, F. T. & Lehmann, W. D. (1997) Proc. Natl. Acad. Sci. USA 94, 2339-2344. [PMC free article] [PubMed]
31. Rietveld, A., Neutz, S., Simons, K. & Eaton, S. (1999) J. Biol. Chem. 274, 12049-12054. [PubMed]
32. Marheineke, K., Grunewald, S., Christie, W. & Reilander, H. (1998) FEBS Lett. 441, 49-52. [PubMed]
33. Gimpl, G., Klein, U., Reilander, H. & Fahrenholz, F. (1995) Biochemistry 34, 13794-13801. [PubMed]
34. Ahmed, S. N., Brown, D. A. & London, E. (1997) Biochemistry 36, 10944-10953. [PubMed]
35. Xu, X., Bittman, R., Duportail, G., Heissler, D., Vilcheze, C. & London, E. (2001) J. Biol. Chem. 276, 33540-33546. [PubMed]
36. Klein, U. & Fahrenholz, F. (1994) Eur. J. Biochem. 220, 559-567. [PubMed]
37. Klein, U., Gimpl, G. & Fahrenholz, F. (1995) Biochemistry 34, 13784-13793. [PubMed]
38. Gimpl, G. & Fahrenholz, F. (2000) Eur. J. Biochem. 267, 2483-2497. [PubMed]
39. Gimpl, G., Burger, K., Politowska, E., Ciarkowski, J. & Fahrenholz, F. (2000) Exp. Physiol. 85, 41S-49S. [PubMed]
40. Mitchell, D. C., Niu, S. L. & Litman, B. J. (2001) J. Biol. Chem. 276, 42801-42806. [PubMed]
41. Niu, S. L., Mitchell, D. C. & Litman, B. J. (2001) J. Biol. Chem. 276, 42807-42811. [PubMed]
42. Stinson, A. M., Wiegand, R. D. & Anderson, R. E. (1991) Exp. Eye Res. 52, 213-218. [PubMed]
43. Jensen, A. A., Greenwood, J. R. & Brauner-Osborne, H. (2002) Trends Pharmacol. Sci. 23, 491-493. [PubMed]
44. Parmentier, M. L., Joly, C., Restituito, S., Bockaert, J., Grau, Y. & Pin, J. P. (1998) Mol. Pharmacol. 53, 778-786. [PubMed]
45. Becher, A., White, J. H. & McIlhinney, R. A. (2001) J. Neurochem. 79, 787-795. [PubMed]
46. Ledesma, M. D., Simons, K. & Dotti, C. G. (1998) Proc. Natl. Acad. Sci. USA 95, 3966-3971. [PMC free article] [PubMed]
47. Ledesma, M. D., Brügger, B., Bunning, C., Wieland, F. T. & Dotti, C. G. (1999) EMBO J. 18, 1761-1771. [PMC free article] [PubMed]
48. Klopfenstein, D. R., Tomishige, M., Stuurman, N. & Vale, R. D. (2002) Cell 109, 347-358. [PMC free article] [PubMed]
49. Cremona, O. & De Camilli, P. (2001) J. Cell Sci. 114, 1041-1052. [PubMed]
50. Mauch, D. H., Nagler, K., Schumacher, S., Goritz, C., Muller, E. C., Otto, A. & Pfrieger, F. W. (2001) Science 294, 1354-1357. [PubMed]
51. Barres, B. A. & Smith, S. J. (2001) Science 294, 1296-1297. [PubMed]
52. Pfrieger, F. (2002) Curr. Opin. Neurobiol. 12, 486-490. [PubMed]
53. Koudinov, A. R. & Koudinova, N. V. (2001) FASEB J. 15, 1858-1860. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...