| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Systemic Proteomics Research Center and
The Center of Functional Analysis of Human Genome, Korea Research Institute of Bioscience and Biotechnology, Daejeon, Korea;
Department of Biology, Chungnam National University, Daejeon, Korea;
Department of Biological Science, SungKyunKwan University, Suwon, Korea; and
|| Department of Biochemistry, College of Medicine, The Catholic University of Korea, Seoul, Korea
2Correspondence: Jeong Heon Ko, Systemic Proteomics Research Center, Korea Research Institute of Bioscience and Biotechnology, 52 Eoeun-dong, Yuseong-gu, Daejeon 305–333, Korea. E-mail: jhko{at}kribb.re.kr
ABSTRACT
Gangliosides abundant in the nervous system have been implicated in a broad range of biological functions, including the regulation of cell proliferation and death. Glutamate-induced cell death, which is accompanied by an accumulation of reactive oxygen species (ROS), is a major contributor to pathological cell death within the nervous system. However, the mechanism underlying this neuronal cell death has not been fully elucidated. In this study, we report that ganglioside GM3 is involved in neuronal cell death. GM3 was up-regulated in the mouse hippocampal cell line HT22 death caused by glutamate. Increment in GM3 levels by both the exogenous addition of GM3 and the overexpression of the GM3 synthase gene induced neuronal cell death. Overexpression of GM3 synthase by microinjecting mRNA into zebrafish embryos resulted in neuronal cell death in the central nervous system (CNS). Conversely, RNA interference-mediated silencing of GM3 synthase rescued glutamate-induced neuronal death, as evidenced by the inhibition of massive ROS production and intracellular calcium ion influx. 12-lipoxygenase (12-lipoxygenase) (12-LOX) was recruited to glycosphingolipid-enriched microdomains (GEM) in a GM3-dependent manner during oxidative glutamate toxicity. Our findings suggest that GM3 acts as not only a mediator of oxidative HT22 death by glutamate but also a modulator of in vivo neuronal cell death.—Sohn, H., Kim, Y.-S., Kim, H.-T., Kim, C.-H., Cho, E.-W., Kang, H.-y., Kim, N.-S., Kim, C.-H., Ryu, S. E., Lee, J.-H., Ko, J. H. Ganglioside GM3 is involved in neuronal cell death.
Key Words: GM3 • 12-lipoxygenase • glycosphingolipid-enriched microdomains • ROS • neuronal cell death
NEURONAL CELL DEATH underlies the symptoms of many neurological disorders, including ischemic stroke, trauma, PD, Alzheimer’s disease, Huntington’s disease, and amyotrophic lateral sclerosis. A common feature of these neurological disorders is an extensive evidence of oxidative stress, which might be responsible for neuronal cell death, thus contributing to the pathogenesis of disorders. Oxidative stress is the result of the unregulated production of reactive oxygen species (ROS), such as hydrogen peroxide, NO, superoxide, and highly reactive hydroxyl radicals (1
, 2)
.
Glutamate-induced toxicity is implicated in the pathology of neuronal cell death and appears to be mediated by ROS (3)
. In contrast with glutamate-induced excitotoxicity, which depends on the activation of ionotropic glutamate receptors (3
, 4)
, oxidative glutamate toxicity is a nonreceptor-mediated pathway, which involves the glutamate/cystine antiporter system (5)
. An established model system for the study of oxidative glutamate toxicity is the immortalized mouse hippocampal cell line HT22. In this model system, high concentrations of extracellular glutamate interfere with the exchange of glutamate with cystine through the glutamate/cystine antiporter, which normally carries cystine into the cell. It results in depletion of the antioxidant glutathione (GSH), followed by 12-lipoxygenase (12-LOX) activation, massive ROS production, severe oxidative stress, and cyclic GMP-dependent Ca2+ influx, leading to cell death (6
7
8)
. Extensive studies have been made to characterize the cell death cascade for this glutamate-induced neuronal cell death. The synthesis of macromolecules is required for death via oxidative stress (9)
. This death has the morphological features of both apoptosis and necrosis in various types of neuronal cell lines including HT22 cells (10)
. Oxidative toxicity induced by glutamate causes a delayed, sustained activation of extracellular signal-regulated kinase 1/2 (ERK1/2) that is necessary for neuronal cell death (11)
. Nevertheless, the signaling pathways that lead to this form of neuronal cell death are still not fully understood.
Gangliosides, sialic acid-containing glycosphingolipids, are ubiquitously present in the plasma membrane of essentially all mammalian cells and are particularly abundant in the nervous system (12
13
14)
. Although their biological functions have not been completely clarified, recent studies indicate that they are implicated in a broad range of biological functions, including cell–cell interaction, cell activation, and signal transduction (15
16
17)
. In addition, other studies point to the involvement of gangliosides in the regulation of cell proliferation and cell death (18
, 19)
. The simplest ganglioside GM3, a major ganglioside in many mammalian cells, is expressed predominantly in the early embryonic brain and at lower levels in adulthood (20
, 21)
. Exogenously added GM3 is incorporated into the plasma membrane and modulates cell growth and differentiation in the CNS, acting as an antitumor agent against human brain tumors (22
23
24)
. In fact, exogenously added GM3 inhibits cell proliferation and induces apoptosis in gliomas as well as immature proliferating glial and neuronal cells (23
, 25)
. In addition, GM3 overexpression induces apoptosis and reduces malignant potential in murine bladder cancer (26)
. GD3 interacts with mitochondria and recruits it to apoptotic pathways, contributing to the mitochondria-dependent apoptosome activation and subsequent apoptosis triggered by death ligands (18
, 19
, 27
, 28)
.
On the basis of these observations, we investigated the involvement of gangliosides in neuronal cell death. In this study, we show that an increase in intracellular GM3 levels is prerequisite for the massive ROS production in glutamate-induced neuronal cell death. In addition, overexpression of GM3 synthase in the zebrafish animal model resulted in a specific neuronal cell death in the CNS.
MATERIALS AND METHODS
Cell culture, zebrafish maintenance, and reagents
HT22 cells, a mouse hippocampal cell line, were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Invitogen, Carlsbad, CA) supplemented with 10% FBS (Hyclone, Logan, UT). Adult fish were maintained at 28.5°C with a 14 h light/10 h dark cycle. Embryos were washed and cultured at 28.5°C in Ringer’s solution. The embryonic stage was determined by hours postfertilization (hpf) and by microscopic observation, as referenced in The Zebrafish Book (29)
. Some of the embryos were treated with phenylthiocarbamide (1-phenyl-2-thiourea, PTU; Sigma, St. Louis, MO) to suppress melanin synthesis. D-threo-1-phenyl-2-decanoylamino-3-morpholino-1-propanol (PDMP) was from Calbiochem. L-glutamic acid, carbonyl cyanide p-trifluoromethoxyphenyl-hydrazone (FCCP), cobalt chloride, and gangliosides (GM3, GD3, and GT1b) were purchased from Sigma.
Ganglioside extraction and high-performance thin-layer chromatography (HPTLC)
Gangliosides were extracted as described previously (30)
with minor modification. In a typical experiment, cells were extracted twice with chloroform:methanol (1:1, v/v) and partitioned between diisopropyl ether and 1-butanol by addition of 0.1% NaCl, resulting in a final diisopropyl ether:1-butanol:0.1% NaCl ratio of 6:4:5. The lower phase, including the polar glycosphingolipids, was purified from salts and low MW contaminants using a Sep-Pak C18 cartridge (Waters Corp., Milford, PA). The eluted gangliosides were applied to HPTLC plates, which were developed in chloroform:methanol:0.25% KCl (60:35:8, v/v/v). Gangliosides were visualized by spraying resorcinol reagent followed by heating for 30 min at 120°C. The concentration of gangliosides was determined by quantifying the sialic acid in the eluates.
GM3 immunofluorescence (flow cytometry/confocal microscopy)
GM3 levels on plasma membranes were determined by indirect membrane immunofluorescence using flow cytometry or confocal microscopy. Cells were trypsinized, washed twice, and stained with mouse monoclonal anti-GM3 (Seikagaku Corp., Tokyo, Japan), which had been diluted in PBS (1:100) containing 0.1% BSA and 0.05% sodium azide, for 1 h at 4°C. After washing twice, cells were incubated with a secondary goat antimouse immunoglobulins conjugated with FITC (DAKO, Glostrup, Denmark) diluted in PBS (1:10,000) containing 0.1% BSA for 30 min at 4°C, washed twice, and resuspended in PBS, and then cell suspensions were used for confocal microscopy (Zeiss LSM 510) or flow cytometry using FACSCaliburTM (BD Biosciences, San Jose, CA).
Plasmid construction and transient transfection
The cDNA of mouse GM3 synthase were amplified from total RNA of HT22 cells by using the following primers: Forward primer, 5'-CCGGATCCCAGCACAAGATGA-3'; Reverse primer, 5'-GTGGATCCGAGTTCGCTGTGGATG-3' based on the GenBankTM sequence information (AB018048). Purified PCR products were subcloned into the pcDNA3.1 (Invitrogen) vector. The correct sequence and orientation of the constructs were confirmed by DNA sequencing. Vector ST3GalV-pcDNA3.1 containing mouse GM3 synthase or the pcDNA3.1 empty vector was transfected into HT22 cells using LipofectamineTM (Invitrogen) as the transfection agent, according to the manufacturer’s instructions. One day after transfection, RT-PCR, phase-contrast microscopy, and a cell viability assay were performed.
RT-PCR
Total RNA was prepared from the transfected HT22 cells using the RNeasy RNA isolation kit (Qiagen, Hilden, Germany). Total RNA (2 µg) was reverse-transcribed in a 20 µl reaction system by using a Monkey Murine Leukemia Virus Reverse Transcriptase (Invitrogen) under conditions described by the supplier. We amplified 1/100 reverse-transcribed cDNA by 30 cycles under standard conditions in a 20 µl PCR system with two primers for mouse GM3 synthase [Forward primer, 5(prime)-CAGCACAAGATGAGAAGACCC-3(prime), and Reverse primer, 5(prime)-GCGTAGTATTCAACGTCCGAC-3(prime)]. The gene expression levels of mouse GM3 synthase were normalized against ß-actin.
Cell viability assay
Cell viability was assessed with a Cell Counting Kit-8 (Dojindo Laboratories, Kumamoto, Japan). Cells were seeded and grown for 24 h in 96-well plates prior to any treatment. After treating with various reagents according to the experimental design, 10 µl of kit reagent was added and the solution was incubated for another 1.5 h. Cell viability was determined by scanning with a microplate reader at 450 nM. The results were expressed relative to the control values specified in each experiment and were subjected to statistical analysis.
Stable cell line expressing GM3 synthase siRNA by RNA interference
Mouse GM3 synthase expression was silenced by the small interfering RNA (siRNA) expression system (31)
. Three sequences, (a) TTACACCGCTGAAGAATGT, (b) CCAGTTCGATGTGGTAATA, and (c) TACTACGCCAATGATTTGT, were chosen and cloned in the pSilencerTM 1.0-U6 expression vector (Ambion, Austin, TX). The specificity of the sequence was verified by a BLAST search of the public database. The control vector or vector containing mouse GM3 synthase siRNA was cotransfected with pcDNA3.1 hygro vector (Invitrogen) into HT22 cells using LipofectamineTM Reagent (Invitrogen). Transfected cells were selected by culturing in hygromycin (250 µg/ml), and pure populations were used in the experiments. The inhibitory expression efficiency of mouse GM3 synthase siRNA was determined by RT-PCR for mouse GM3 synthase. Sequence c, above, showing the highest efficacy was selected for experiments.
ROS measurement
ROS production was evaluated by staining cells with dichlorodihydrofluorescein diacetate (H2DCFDA; Molecular Probes, Carlsbad, CA). Cells were washed twice with DMEM containing 10% FBS, incubated in 10 µM H2DCFDA diluted with DMEM for 20 min at 37°C, washed with PBS, and trypsinized. Trypsinized cells were washed twice with ice-cold PBS, resuspended in PBS, and analyzed by flow cytometry using FACSCaliburTM (BD Biosciences).
Ca2+ measurement
Intracellular Ca2+ was measured using Fluo-4 acetoxymethyl (AM) esters (Molecular Probes) and flow cytometry. Cells were washed, trypsinized, and incubated in 2.5 µM Fluo-4 diluted with DMEM containing 10% FBS and 1 mM EDTA (pH 7.4) for 30 min at 37°C. Incubated cells were washed twice with DMEM containing 10% FBS, incubated for a further 30 min to allow complete de-esterification of intracellular AM esters, and analyzed by flow cytometry using FACSCaliburTM (BD Biosciences).
GSH measurement
For GSH measurements, cells were trypsinized, collected, washed, resuspended in 0.01 N HCl, including 5% 5-sulfosalicylic acid for the removal of proteins from sample solution and lysed by freezing and thawing. The supernatant was assayed for intracellular GSH content using a GSH quantification kit from Dojindo Laboratories according to the manufacturer’s recommendations.
12-LOX activity assay
12-LOX activity was measured by quantifying 12-hydroxyeicosatetraenoic acid [12(S)-HETE], the stereospecific hydroxyl product of 12(S)-hydroperoxyeicosatetraenoic acid, and the 12-LOX metabolite of arachidonic acid (AA) using a 12(S)-HETE immunoassay kit (R&D Systems Inc., Minneapolis, MN). For sample preparation, cells were washed twice, harvested, and resuspended in PBS (pH 7.4). And then, the cell suspension was sonicated, acidified to pH 3.5 by adding 2 N HCl, and extracted by ethyl acetate (1:2, v/v) three times. The supernatant was dried by nitrogen gassing and resuspended in ethanol and the levels of 12(S)-HETE were analyzed according to the manufacturer’s recommendations.
Sucrose gradient ultracentrifugation
HT22 cells were treated with 5 mM glutamate or PBS buffer as control for 7.5 h and equal numbers of cells (2x107) was sonically lysed in presence of 1% TX-100 for 10 min. The cell lysates were subjected to 5–43% sucrose gradient ultracentrifugation at 200,000 g for 18 h in a SW55Ti rotor (Beckman, Beckman, CA) as described elsewhere (32)
. After centrifugation, solutions were fractionated into 0.5 ml of aliquots, each of which was used for the quantification of GM3 and 12-LOX. Following a partial purification of each fraction according to a previous report (33)
, GM3 was quantified by HPTLC, as described above. Immunoblot analysis of 12-LOX was performed using an anti-12-LOX polyclonal antibody pAb (Cayman Chemical, Ann Arbor, MI).
Whole-mount in situ hybridization
To prepare the antisense RNA probe for zebrafish GM3 synthase, we amplified a cDNA fragment by RT-PCR based on GenBankTM sequence information (NM_199516). The primers used were forward 5'-AGTGAAGATCTCTTATTTTGAGCA-3' and reverse 5'-GCCATTGGTTTCAACAGAAGCTGC-3'. The 1,164 bp cDNA fragment product was subcloned into the pGEM-T Easy vector, linearized with NcoI, and antisense RNA for GM3 synthase was in vitro transcribed using a SP6 RNA polymerase and digoxigenin-labeled UTP. Whole-mount in situ hybridization was performed as described previously (34)
.
Microinjection
The GM3 synthase cDNAs for the mouse and zebrafish were subcloned into the pCS2+ vector and sense RNAs encoding full-length zebrafish GM3 synthase were transcribed in vitro using the SP6 Message Machine (Ambion). To obtain one- to two-cell embryos, mating pairs were set up in a transparent tank. Eggs were collected and immediately washed in egg water at 28.5°C. Embryos at the one- to two-cell stage were arranged in agar troughs in 100 mM-diameter dishes. Microinjections were carried out with a WPI microinjector and picopump controller.
Acridine orange staining
Acridine orange (Sigma) was diluted in fish water to a final concentration of 1 µg/ml. Dechorionated embryos were placed in this solution for 30 min, washed briefly in fish water for 1 h, and immediately observed with GFP optics on a Leica DM5000B fluorescence microscope.
Data analysis
Flow cytometric data were plotted as histograms using the data analysis program CELLQuest Pro (BD Biosciences). The median fluorescence of each peak was obtained and multiplied by the number of events in that peak. Data are presented as mean ± SD. Statistical comparisons were performed using the Student’s t test. A value of P < 0.05 was considered to be significant.
RESULTS
GM3 is up-regulated in glutamate-induced HT22 cell death
An increase in GM3 levels is required for the induction of terminal differentiation, culminating in death by apoptosis, of human colonic carcinoma cells in vitro (35)
, as has been observed during cell death by metastasis-suppressing gene products (36)
, and induced by treatment with tumor necrosis factor-
(TNF-) (37)
. GD3 synthesis is induced during Fas (APO-1/CD95)-mediated apoptosis and is then thought to mediate the apoptotic effect (18)
. To determine whether changes in GM3 or GD3 levels could be monitored during glutamate-induced neuronal cell death, we analyzed the ganglioside pattern in glutamate-treated HT22 cells. An HPTLC analysis of the total cellular gangliosides revealed that GM3 was the major ganglioside in HT22 cells and its cellular concentration increased in a time-dependent manner after glutamate treatment, although GD3 was not detected (Fig. 1
A). GM3 appeared as double bands in HPTLC, probably due to the heterogeneity of the fatty acid composition, as described previously (35
, 38)
. To examine the relationship between an increase in GM3 levels and typical features associated with glutamate-induced HT22 death such as a massive ROS production and Ca2+ influx (6
7
8)
, we also examined GM3 levels on the cell surface by flow cytometric analysis using an anti-GM3 antibody (GMR6) after treatment with glutamate alone, glutamate plus FCCP, a mitochondrial uncoupler, or glutamate plus cobalt chloride, a calcium ion channel blocker. Although these inhibitors prevented glutamate-induced cell death (8, data not shown), they had no measurable effect on GM3 levels (Fig. 1B
). Similar results were observed in the confocal microscopic study (Fig. 1C
), suggesting that an increment in GM3 levels is either an independent event on or upstream from ROS production and Ca2+ influx.
|
Both the exogenous addition and endogenous overexpression of GM3 induce neuronal cell death
We investigated the issue of whether GM3 is directly involved in the cell death process. Exogenously added GM3, but not GD3 and GT1b, induced cell death in HT22 cells in a dose-dependent manner, as evidenced by a cell viability assay. The effect of GM3 on HT22 cell death was observed at more than 20 µM GM3, whereas equal concentrations of the other gangliosides were ineffective (Fig. 2
A). These results are similar to those reported for rat glioma cells (39)
. To confirm that endogenous GM3 accumulation is also sufficient to trigger cell death, we subsequently transiently overexpressed the gene coding for mouse GM3 synthase (ST3GalV; lactosylceramide 2,3-N-acetyl sialic acid transferase) in HT22 cells, which is in charge of GM3 synthesis by adding a sialic acid to lactosylceramide. An RT-PCR analysis showed that the ST3GalV-transfected cells contained high mRNA levels of GM3 synthase (Fig. 2B
). Cell proliferation was inhibited significantly and massive cell death was observed in the ST3GalV-transfectants compared with mock-transfectants (Fig. 2C, D
). These results suggest that an increase in GM3 levels induces HT22 cell death.
|
Pharmacological inhibition of GM3 synthesis prevents glutamate-induced neuronal cell death
We then asked whether glutamate-induced cell death could be interrupted by blocking GM3 synthesis. To interfere with intracellular ganglioside neosynthesis, we used an inhibitor of glucosylceramide synthase, PDMP, a well known reagent used in functional studies of glycosphingolipids including gangliosides (18)
. GM3 immunofluorescence by flow cytometric analysis showed that pretreatment with PDMP lowered the GM3 levels in HT22 cells and blocked the increase in GM3 levels induced by glutamate (Fig. 3
A). The viability of PDMP-pretreated HT22 cells prior to glutamate treatment was similar to that of the glutamate-untreated control cells, indicating that GM3 depletion by PDMP pretreatment was able to substantially block glutamate-induced cell death (Fig. 3B, C
).
|
RNA interference-mediated silencing of GM3 synthase rescues glutamate-induced neuronal cell death, preventing massive ROS production, 12-LOX activation, and intracellular Ca2+ increase
To further investigate the role of GM3 in glutamate-induced cell death, we directly targeted mouse GM3 synthase by RNA interference using plasmid-based siRNA constructs (pSilencerTM 1.0-U6 from Ambion) and established a stable cell line expressing siRNA for GM3 synthase. mRNA expression of the GM3 synthase gene was examined by RT-PCR. HT22 cells stably expressing siRNA for GM3 synthase (siST3GalV), but not control cells expressing the empty vector (mock), exhibited a reduction in the basal expression concentration of GM3 synthase, as well as an increment in GM3 levels during the cell death induced by glutamate (Fig. 4
A, B). We also monitored glutamate-induced cell death in these cells by a cell viability assay. As expected, treatment of siST3GalV cells with glutamate did not result in cell death, whereas, under the same conditions, it induced cell death in mock cells (Fig. 4C, D
). These results suggest that the up-regulation of GM3 is required for glutamate-induced neuronal cell death. We next explored the possible mechanism by which the up-regulation of GM3 mediates cell death in response to glutamate. Previous studies reported that ROS plays a central role in glutamate-induced neuronal cell death and its production in massive quantities is required in death (3
, 6
, 8)
. Therefore, we compared the production of ROS in siST3GalV cells to that in mock cells by flow cytometric analysis using H2DCFDA. As seen in Fig. 4E
, the glutamate-induced increase in ROS was blocked in siST3GalV cells but not in mock cells, indicating that an increase in GM3 levels is required for the massive ROS production. Since an increase in intracellular Ca2+ is another common feature of glutamate-induced neuronal cell death (6
7
8
9)
, we investigated the effect of GM3 depletion on Ca2+ influx in glutamate-treated cells. As shown in Fig. 4F
, we observed that a reduction in GM3 synthase resulted in the inhibition of an increase in the intracellular Ca2+. These data indicate that the massive ROS production and intracellular Ca2+ increase are dependent on GM3 levels in glutamate-induced neuronal cell death. GSH depletion induced by glutamate causes ROS production via 12-LOX activation in neuronal cell death (6)
. To gain further insights into the link between GM3 and ROS, we investigated GSH depletion and 12-LOX activities in siST3GalV cells after glutamate treatment. 12-LOX activation was significantly blocked in siST3GalV cells compared to mock control cells after glutamate treatment (Fig. 4H
). However, intracellular GSH depletion was not affected by a decrease in endogenous GM3 levels (Fig. 4G
).
|
12-LOX is activated by being recruited to GEM
A deeper insight of how the up-regulation of GM3 is related to the oxidative death of HT22 cells was gained by monitoring the trafficking of 12-LOX along with GM3. Cells were lysed in 1% Triton X-100 (TX-100) and the lysates were fractionated in 5–43% sucrose gradient centrifugation, resulting in fractions F1 to F9. Proteins solubilized by TX-100 were found in the F7–F9 fractions, as assessed by the position of ß-actin (Fig. 5
C). However, TX-100-insoluble components formed a turbid layer near the interface of 5 and 30% sucrose solutions, corresponding to near the F3 fractions. A small portion of ß-actin was found in the F3 and F4 fractions, which may be due to the possible interaction of ß-actin with the TX-100-insoluble components (40)
. GM3 was found exclusively in F2-F4 (Fig. 5A
), indicating that GM3 is localized in glycosphingolipid-enriched microdomains (GEM). Consistent with Fig. 1A
, an increase in GM3 content in GEM was observed on treatment with glutamate. Although the content of GM3 in GEM was significantly lower in siST3GalV cells, the glutamate-induced increment of GM3 content was also observed in the cells. Note that a significant portion of 12-LOX was colocalized with GM3 in the GEM regions and that treatment with glutamate caused higher amounts of 12-LOX to be tethered in the GEM regions (Fig. 5B
). When the content of GM3 was low, 12-LOX was undetectable in those regions.
|
Considering the fact that the content of arachidonic acid (AA), a substrate for 12-LOX, is universally described as being low and our data in which AA is nearly equally distributed between the cytosolic and membrane fractions (data not shown), it is likely that the rate of production of 12-HETE is very low in the cytosol. A majority of AA is instead metabolized to 12-HETE in the GEM fraction, in which the local concentrations of AA and 12-LOX are relatively high. This finding is supported by reports that the activated 12-LOX is translocated to the membrane (41)
and our data (Fig. 4H
). Although it is unclear at present how an increase in the concentration of GM3 is connected to a concomitant increase in 12-LOX in the GEM regions, these results imply that glutamate may bring about an oxidative stress, at least in part, by inducing the mass-production of 12-HETE through the recruitment of 12-LOX in the GEM.
In vivo functional study of GM3 synthase in the zebrafish animal model
To explore the in vivo role of GM3, we used the zebrafish as an animal model system. The zebrafish is an ideal animal model for the study of the development and disease processes in the nervous system because of the easy access to and analysis of transparent external developing embryos (42)
. We first identified the presence of a zebrafish homologue of GM3 synthase in the zebrafish EST database (GenBankTM accession no. NM_199516) and cloned its full-coding cDNA by RT-PCR. The zebrafish GM3 synthase gene encodes a protein of 364 amino acid residues, which has a high similarity (61.4%) to mouse GM3 synthase (Fig. 6
A). To examine the expression of the zebrafish GM3 synthase gene at various developmental stages, we performed whole-mount in situ hybridization by using a DIG-labeled antisense RNA probe. Zebrafish GM3 synthase transcripts exhibited a roughly ubiquitous expression, with a high concentration of expression notably in the CNS (Fig. 6B
). The expression pattern of GM3 synthase in zebrafish is similar to that observed in embryonic rat brain (21)
. Although there are several reports on GM3-related knockout mice (43
, 44)
, the in vivo role of GM3 synthase is largely unknown. We utilized zebrafish for gene overexpression system by microinjecting in vitro-transcribed mRNAs into zebrafish embryos. To our knowledge, it is the first report in which gain-of-function study of GM3 synthase is performed in vertebrates. Zebrafish embryos injected with mouse GM3 synthase and zebrafish GM3 synthase transcripts (up to 300 pg) show relatively normal phenotypes in their overall development: early cell division, gastrulation, somitogenesis, and early organogenesis. However, when embryos were stained with acridine orange, which is widely used to detect dying cells during embryonic development in zebrafish and other animals (45)
, there was a dramatic change in the extent of neuronal cell death in GM3 synthase-overexpressing embryos, especially in the CNS (mouse GM3 synthase: 66%, n=36; zebrafish GM3 synthase: 59%, n=27). The death was apparent especially in the forebrain, midbrain, and midhindbrain boundary, indicating that tissues in the brain region are sensitive to GM3 overexpression (Fig. 6C
, panels d–f). At this early stage, the tissues were found to be composed mainly of neuronal and neuronal precursor cells; a majority of cells in the brain region expressed huC, a pan-neuronal and -neuronal precursor marker (46)
, and some portion of the cells expressed -tubulin, which functions in detecting fully differentiated neurons in the neuronal network. However, glial cells were nearly undetectable even at 72 hpf, as assessed by the expression of myelin basic protein, a glial cell marker (Fig. 6C
, panels g–i) (47)
. Taken together, these results strongly suggest that GM3 overexpression causes neuronal cell death in the in vivo zebrafish animal model system.
|
DISCUSSION
Many lines of evidence have revealed that gangliosides are involved in the death of various types of cells, among which the roles of GM3 and GD3 have been extensively investigated. GD3 is expressed at increased levels in a variety of tumors (48)
. In the CNS, GD3 is a minor ganglioside in normal adult brains (49)
, but it is expressed in activated microglia (50)
and in reactive astrocytes (51)
. GM3 is also involved in the apoptotic death of human carcinoma cells as well as glial cells (23
, 39)
. However, the functional role of GM3 and GD3 in neuronal cell death is poorly understood. The findings herein suggest that GM3 is increased in immortalized mouse hippocampal HT22 cells that are exposed to glutamate treatment and mediates the oxidative toxicity induced by glutamate. The death response to glutamate was significantly alleviated by lowering endogenous GM3 levels through RNA interference of GM3 synthase. GD3 and another ganglioside GT1b were present in much lower levels compared to GM3 (Fig. 1A
). Nonetheless, equal amounts of GD3 or GT1b, which would be an excessive amount for HT22 cells, did not affect cell viability, indicating that HT22 neuronal cell death is specifically sensitive to GM3. GD3 contributes to cell death by disrupting the mitochondrial transmembrane potential leading to the release of cytochrome c and the activation of caspase 3 (18
, 28)
. However, general apoptotic parameters such as the activation of Bax and caspase 3, mitochondrial membrane depolarization, DNA fragmentation, and cytochrome c release are not required to execute the cell death program initiated by glutamate or other forms of oxidative stress in neuronal cells (7
, 10
, 52)
. This finding suggests that GM3 is a specific mediator of glutamate-induced oxidative neuronal death.
A variety of biological effects of gangliosides are exerted by a physical and functional association of gangliosides with signaling molecules in glycosphingolipid-enriched microdomains (GEM). GEM acts as a "glycosignaling domain" either by recruiting molecules that are involved in signaling pathways, allowing their molecular interaction, or by modulating signaling function (53)
. Activated 12-LOX was reported to be translocated to the plasma membrane (41)
and to functionally associate with phospholipase A2 isoforms (54)
. It has not been reported that 12-LOX exists, at least temporarily, in GEM. Our functional study revealed that a subset of 12-LOX pool was colocalized with GM3 in GEM and the concentration of 12-LOX in GEM was dependent on that of GM3. Cellular levels of AA are very low and the production of 12-HETE by soluble 12-LOX would be very limited. It is assumed that GM3 acts as a death-inducing molecule by tethering 12-LOX in GEM, in the vicinity of which the local concentration of AA is thought to be relatively high (55)
. This assumption is supported by the overproduction of 12-HETE and the massive production of ROS in glutamate-treated HT22 cells, in which GM3 was accumulated in GEM together with 12-LOX (Figs. 4
and 5)
. Likewise, a reduction in endogenous GM3 levels significantly blocked both massive ROS production and intracellular Ca2+ influx induced by glutamate. 12-HETE has dual functions: It is, in itself, required for the generation of ROS and activates soluble guanylyl cyclase (sGC) to produce cyclic GMP (cGMP), which in turn stimulates channels required for Ca2+ influx late in the cell death pathway (52)
. Ca2+ influx into cells is also a common feature of cell death by oxidative glutamate toxicity in many forms of neuronal cell death (6
7
8
9)
and massive ROS production and Ca2+ influx appear to be tightly coupled (8)
. Another aspect of glutamate-induced oxidative stress is GSH depletion. GSH depletion not only leads to a loss of defensive power against oxidative stresses but also triggers the activation of 12-LOX, peroxide production and Ca2+ influx (6)
. Considering our findings that GM3 levels did not affect GSH depletion (Fig. 4G
), GSH depletion appears to occur in an independent pathway from glutamate-induced increase in GM3 and to cooperate with the increased GM3 in activating 12-LOX. The proposed mechanism underlying HT22 neuronal cell death by oxidative glutamate toxicity is summarized in Fig. 7
.
|
A gain-of-function study of GM3 synthase revealed that GM3 induced neuronal cell death in an in vivo zebrafish system, which was also observed in glutamate-induced HT22 cell death. Notwithstanding the similarity, they show somewhat different aspects of cell death; oxidative HT22 cell death by glutamate toxicity, though a form of programmed cell death, lacks DNA fragmentation and nuclear condensation and, thus, is distinct from classical apoptosis (52)
. In contrast, GM3-induced neuronal death in zebrafish embryo was accompanied by DNA breakage, as accessed by acridine orange staining. However, the regions of the brain that were stained by acridine orange in our study consisted mainly of nonhippocampal neuronal cells and it is probable that GM3 causes an apoptotic death of such neuronal cells by evoking the related molecules. For example, 12-HETE induces the c-Jun-dependent apoptosis of cortical neurons and the blockade of 12-LOX expression protects neurons from apoptotic death (56)
. In addition, there is another report indicating that AA induces the apoptotic death of nonhippocampal neuronal cells by generating ROS through 12-LOX and cytochrome P450 (57)
. Taken together, it is tempting to speculate that, while GM3 is involved in the equipment of oxidative machinery, the consequences can be necrotic or apoptotic, depending on the type of neuronal cell.
ACKNOWLEDGMENTS
The HT22 cells were obtained from David Schubert at the Salk Institute. This research was supported by grants from KRIBB Research Initiative program, Glycomics Program, and 21C Frontier Program from the Korea Ministry of Science and Technology. H.-T. Kim, C.-H. Kim, and N.-S. Kim were supported by a grant (R01–2004-000–10095–0) from the Basic Research Program of the Korea Science and Engineering Foundation.
FOOTNOTES
1 These authors contributed equally to this work. 
3 These authors share senior authorship.
Received for publication August 31, 2005. Accepted for publication January 19, 2006.
REFERENCES
. J. Cell Biol. 152,997-1006This article has been cited by other articles:
|
|
 |
|
  S. Khanna, S. Roy, H.-A Park, and C. K. Sen Regulation of c-Src Activity in Glutamate-induced Neurodegeneration J. Biol. Chem., August 10, 2007; 282(32): 23482 - 23490. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | S |
