БИОЛОГИЧЕСКИЕ МЕМБРАНЫ, 2009, том 26, № 3, с. 234-239
УДК 577.352.
GANGLIOSIDE GM1 INCREASES LINE TENSION AT RAFT BOUNDARY
IN MODEL MEMBRANES
© 2009 r. S. A. Akimov, E. A. Hlaponin, P. V. Bashkirov, I. A. Boldyrev*, I. I. Mikhalyov*,
W. G. Telford**, I. M. Molotkovskaya*
Frumkin Institute of Physical Chemistry and Electrochemistry, Russian Academy of Sciences, Leninsky pr.,
31/5, Moscow, 119991 Russia Tel./fax: (+7-495)-952-55-82, e-mail megakrot@mail.ru *Shemyakin-Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences, ul. Miklukho-Maklaya,
16/10, Moscow, 117997 Russia **Experimental Transplantation and Immunology Branch, National Cancer Institute, National Institutes of Health,
10 Center Drive, Bethesda, Maryland 20892, USA Received February 12, 2009
Gangliosides are significant participants in suppression of immune system during tumor processes. It was shown that they can induce apoptosis of T-lymphocytes in a raft-dependent manner. Fluorescence confocal microscopy was used to study distribution and influence of ganglioside GM1 on raft properties in giant unilamellar vesicles. Both raft and non-raft phase markers were utilized. No visible phase separation was observed without GM1 unless lateral tension was applied to the membrane. At 2 mol % of GM1 large domains appeared indicating macroscopic phase separation. Increase of GM1 content to 5 mol % resulted in shape transformation of the domains consistent with growth of line tension at the domain boundary. At 10 mol % of GM1 almost all domains were pinched out from vesicles, forming their own homogeneous liposomes. Estimations showed that the change of the GM1 content from 2 to 5-10 mol % resulted in a several-fold increase of line tension. This finding provides a possible mechanism of apoptosis induction by GM1. Incorporation of GM1 into a membrane leads to an increase of the line tension. This results in a growth of the average size of rafts due to coalescence or merger of small domains. Thus, necessary proteins can find themselves in one common raft and start the corresponding cascade of reactions.
Key words: raft, ganglioside, apoptosis, line tension
Cancer cells tend to be very resistant to apoptosis, both due to their ability to evade the host immune response and due to the inherent resistance to apoptotic stimuli [1]. Sphingolipids, particularly gangliosides, can suppress the antitumor immune response by blocking cytolysis mediated by T-lymphocytes and natural killer cells. Tumor-derived gangliosides can also influence other cells important for antitumor activity of the immune system. While known for some time, the mechanism of ganglioside-mediated immunosuppres-sion is not well understood.
Certain gangliosides expressed in the plasma membranes of malignant tumor cells rarely or never occur in the normal cellular membranes. GM2 and GD3 are well-characterized examples and are considered to be tumor markers [2]. Ganglioside shedding by tumor cells (neuroblastoma, melanoma, lymphoma, mammary carcinoma, hepatoma, etc.) is a well-known process [3, 4]. This shedding results in a high level of tumor-specific gangliosides in serum and ascites. Ganglio-sides shed by rapidly growing tumors may be major contributors to the generalized suppression of the immune system [4].
Previously extrinsic or receptor-mediated apoptosis activated by FasL (ligand for Fas/CD95 receptor) or TNFa (tumor necrosis factor) has been shown to be a lipid raft-dependent process [5]. Ligand interaction with corresponding death receptors leads to receptor trimerization. This in turn leads to the acid sphingomy-elinase (aSMase) activation with a release of ceramide (Cer) from sphingomyelin [6] and liberated ceramide probably leads to the raft formation. This suggestion was supported by the ability of Cer to induce apoptosis of cells defective on aSMase [7]. Later, natural ceram-ide was shown to induce raft formation in artificial lipid membranes [8]. These findings led to the hypothesis that raft assembly, which follows the death receptors recruitment, is important for signal molecule docking during apoptosis.
The idea that lipid domains with different physical properties could exist in membranes has been known since the 1970s. Numerous studies using different approaches showed that plasma membrane could contain liquid-disordered (Ld) domains coexisting with liquid-ordered domains (Lo). Natural lipids having saturated acyl chains (most sphingolipids) were generally found
to form lipid bilayer with high gel-to-liquid melting temperature (Tm). Lipids having unsaturated acyl chains (like most natural phospholipids) were generally found to have low Tm. The simplest lipid mixture that could mimic eukaryotic plasma membrane behavior is a ternary mixture, which includes lipids with high Tm (e.g., sphingolipids), low Tm lipids (e.g., natural phosphatidylcholine), and cholesterol [9-13].
Gangliosides have been previously shown to induce apoptosis in cytotoxic T- lymphocytes. Moreover, the GMl-induced apoptotic pathway is dependent on the apoptosis-associated signaling/effector enzyme caspase-8 [14]. Since induced apoptosis in cytotoxic T-lymphocytes has also been found to be dependent on membrane reorganization, lipid raft formation may also play a direct signaling role in this process, as it does in Fas-mediated apoptosis signaling.
In our proposed model, GM1 clusters, normally associated with small inactive rafts, assemble to form a large lipid platform, where functionally related proteins can interact. Moreover, gangliosides may initiate assembly of rafts functionally different from ones formed by natural ceramide. To test this hypothesis from a mechanistic standpoint, giant liposomes were assembled with fluorophore-conjugated lipids that could identify phase state of Ld and Lo membrane regions. When natural gangliosides GM1 were simultaneously incorporated into this system, assembly of small membrane domains into larger ordered ones was indeed observed by confocal microscopy. These results suggest that gangliosides can really promote large raft formation important for apoptotic signaling.
EXPERIMENTAL
Materials. 1,2-Dioleoyl-sn-glycero-3-phosphocho-line (DOPC), cholesterol (Chol), egg sphingomyelin (eSM), 1,2-dioleoyl-sn-glycero-3-phosphoethanola-mine-N-(Lissamine Rhodamine B Sulfonyl) (Rho-DOPE) were purchased from Avanti Polar Lipids Inc. (USA) and GM1 was purchased from Sigma Chemical Co. and used without further purification. The synthesis and properties of the novel fluorescent probe Me4-BODIPY-8-yl-C5-SM (Bodipy-SM) are described in [15]. Ganglioside fluorescent probe N-BODIPY-FL-C5-GM1 (BODIPY-GM1) was synthesized as described in [16]. Unless otherwise indicated, all other chemicals were from Sigma-Aldrich (USA).
GUV preparation. Giant unilamelliar vesicles (GUVs) were prepared by gentle hydration as described in [17]. Briefly, lipid probes were added to an aliquot of the lipid solution DOPC : SM:Ch (1 : 1 : 1) containing fixed amounts of liquid-disordered (Ld) and liquid-ordered (Lo) fluorescent phase markers and varying amounts of GM1 in chloroform. Lipid solution was rotary evaporated in a round-bottomed flask and kept in vacuum (20 Pa) for 60 min. The resulting dried lipid
film was incubated in a solution of 300 mM sucrose and 10 mM HEPES, pH 7.2, for 48 h at 40°C. The characteristic size of GUVs was 7-10 ^m.
We used Rho-DOPE as a Ld phase marker [18-21]. For the Lo phase visualization, fluorescent lipid probes BODIPY-SM or BODIPY-GM1 were used [15, 2024]. The probes were added in pairs during the above preparation, Rho-DOPE + BODIPY-SM or Rho-DOPE + + BODIPY-GM1, so that two probes were present in vesicles in each experiment. Concentration of BODIPY-SM was 0.2 mol %, that of BODIPY-GM1 and Rho-DOPE, 0.5 mol %. Unlabelled ganglioside GM1 was added simultaneously into the DOPC:eSM:Chol lipid solution at concentrations of 0, 2, 5, or 10 mol %.
Confocal fluorescence imaging. Confocal fluorescence images of GUVs were acquired with an inverted laser confocal microscope (Nikon Eclipse C1 Plus, Nikon Instruments, Japan) at a 100x magnification with oil immersion. A 488-nm Ar+ laser and a 543-nm HeNe laser were simultaneously used to excite BODIPY-SM (or BODIPY-GM1) and Rho-DOPE, respectively.
Immediately before the confocal microscopy experiment, the GUV suspension was diluted with equal volume of mixture of 150 mM KCl and 10 mM HEPES. This resulted in a decrease of density of the solution surrounding GUVs leading to sedimentation of the vesicles on the bottom of the experimental chamber. The isotonicity of the solutions inside and outside GUVs was controlled using an osmometer Osmomat 030 (Gonotec, Germany). In some cases, we applied osmotic pressure by adding pure water to external solution in order to induce GUV swelling. The pressure was generated by the concentration difference equivalent to 30 mM. The temperature of the experimental chamber was maintained at 20°C. BODIPY and Rhodamine fluorescence was detected using narrow bandpass filters, 510-530 nm for BODIPY and 580-600 nm for Rhodamine. Images were collected and stored on the computer hard disk for further analysis.
RESULTS AND DISCUSSION
GUVs in the absence of unlabelled GM1. Typical images of a GUV containing lipid probes Rho-DOPE and BODIPY-SM in the absence of GM1 are presented in Fig. 1. The vesicles are large (about ten micrometers in diameter or larger) and well formed, easily distinguishable by confocal microscopy. The distribution of fluorophore is easily observable. For the case when a different pair of probes was used (Rho-DOPE and BO-DIPY-GM1) the picture was qualitatively the same (not shown). Both probes were uniformly distributed throughout the vesicle in Fig. 1, without visible heterogeneity. This indicated that either vesicle phase separation did not take place under such conditions or non-
236
AKIMOV h «p.
uniform domains were too small to be resolved optically.
Для дальнейшего прочтения статьи необходимо приобрести полный текст. Статьи высылаются в формате PDF на указанную при оплате почту. Время доставки составляет менее 10 минут. Стоимость одной статьи — 150 рублей.