Search International and National Patent Collections
Some content of this application is unavailable at the moment.
If this situation persists, please contact us atFeedback&Contact
Note: Text based on automatic Optical Character Recognition processes. Please use the PDF version for legal matters


This application claims priority to the filing date of U.S. Provisional Application Serial No. 60/047,430 filed May 22, 1997.

Field of the Invention
The simple ganglioside GM3, a naturally occurring ligand, has been found to selectively regulate cell proliferation and induce apoptosis of astrocytes. In addition, GM3 has been determined to be useful as a selective chemotherapeutic agent for treating high grade gliomas in human patients.

Background of the Invention
During normal vertebrate development approximately twice as many cells are born, as are finally incorporated into the mature nervous system. This great excess of cells requires that there is a mechanism to eliminate surplus cells, thus preventing an unnecessary nutritional burden to the animal. One mechanism for selective pruning of excess cells is induction of programmed cell death, or apoptosis.
Apoptosis plays an important role in the development of the central nervous system (CNS). However, the factors which indure apoptosis in normal CNS development have not been clearly identified.
Gangliosides, surface components of all mammalian cells, are highly expressed in the CNS compared to many other tissues ( Fishman, P.H. and R.O.

Brady, Biosynthesis and function of gangliosides, Science, 1976, 194: p.906-915; Ledeen, R.W. and R.K. Yu, Gangliosides: structure, isolation and analysis, Methods Enzymol, 1982, 83: p. 1309-191). Specific ganglioside expression is developmentally regulated (Irwin, L.N. and C.C. Irwin, Developmental changes in ganglioside composition of hippocampus, retina and optic tectum, Dev. Neurosci.,

1979, 2: p. 129-138; Irwin, L.N., D.B. Michael, and C.C. Irwin, Ganglioside patterns of fetal rat and mouse brain, J. Neurochem., 1980, 34: p. 1527-1530; Hilbig, R., et al. , Developmental profiles of gangliosides in mouse and rate cerebral cortex, Roux's Arch., 1982, 191: p. 281-284; Rosner, H., Ganglioside changes in the chicken optic lobes as biochemical indicators of brain development and maturation, Brain Res., 1982, 236: p. 49-61). During early development GD3 (Irwin, L.N., D.B. Michael, and C.C. Irwin, Ganglioside patterns of fetal rat and mouse brain, J.

Neurochem., 1980, 34: p. 1527-1530; Hilbig, R., et al., Developmental profiles of gangliosides in mouse and rate cerebral cortex, Roux's Arch., 1982, 191: p. 281-284 ) and GM3 are (Heffer-Lauc, M., et al, Anti-GMS (113Neu5AC-lactosylceramide) ganglioside antibody labels human fetal purkinje neurons during the critical stage of cerebellar development, Neurosci. Lett., 1996, 213: p. 91-94) expressed at high levels in the CNS, which subsequently decrease (Goldman, J.E., et al., GD3 ganglioside is a glycolipid characteristic of immature neuroectodermal cells, J. Neuroimmunol., 1984, 7: p. 179-192). The highest levels of GD3 and GM3 expression correlate with periods of local cell proliferation suggesting that gangliosides might regulate growth and differentiation of CNS neural cells (Bremer,

E.G., et al, Ganglioside - mediated modulation of cell growth, growth factor binding and receptor phosphorylation, J. Biol. Chem., 1984, 258: p. 6818-6825; Bremer, E.G., J. Schlessinger, and S. Hakomori, Ganglioside-mediated modulation of cell growth, Specific effects of GM3 on tyrosine phosphorylation of the epidermal growth factor receptor, J. Biol. Chem., 1986, 261: p. 2434-2440; Cannella, M.S., et al., Comparison of epi-GM3 with GM3 and GM1 as stimulators of neurite outgrowth, Devel. Brain. Res., 1988, 39: p. 137-143; Durand, M., et al, ed. Evidence for the effects of gangliosides on the development of neurons in primary cultures, Gangliosides and Neuronal Plasticity, ed. R. Tettamanti, et al. 1986, Liviana: Padova, Italy. 295-307; Yim, S.H., et al., Differentiation of oligodendrocytes cultured from developing rat brain is enhanced by exogenous GM3 ganglioside, J. Neurosci. Res., 1994, 38: p. 268-281).
Gangliosides are present in the plasma membrane of most mammalian cells and are enriched in the central nervous system (CNS) (P.H. Fishman, R.O. Brady, Science 194: p. 906-15 (1976); R.W. Ledeen, R.K. Yu, Methods Enzymol 83: p. 139-91 (1982)). The expression of the gangliosides is developmentally regulated in the brain, and the simplest ganglioside GM3 is expressed predominantly in rat brain until embryonic day 16 (El 6), and at lower levels during further development (R.K. Yu, L.J. Macala, T. Taki, H.M. Weinfield, F.S. Yu, J. Neurochem 50: p. 1825-9 (1988)), with low levels of expression persisting throughout adulthood. GM3 is expressed in the ventricular zone of early human brain development (M.

Stojiljkovic, et al, Int. J. Dev. Neurosci 14: p.35-44 (1996)), and in adult rat brain, GM3 is intensely expressed in the white matter and layer VI of cerebrum (M. Kotani, et al., Glycobiology, 4: p. 855-65 (1994)). In vitro, exogenously added gangliosides are rapidly incorporated into the plasma membranes of cells (T.W. Keenan, E. Schmid, W.W. Franke, H. Wiegandt, Exp. Cell Res. 92: p. 259-70 (1975); R.A.

Laine, S. Hakomori, Biochem. Biophys. Res. Commun. 54: p. 1039-45 (1973)) and cause numerous biological effects. One common effect is the modulation of the cell growth. Proliferation of EGF driven 3T3 fibroblast and A431, KB epidermoid carcinoma cell lines (E.G. Bremer, S. Hakomori, P.D. Bowen, E. Raines, R. Ross, J. Biol. Chem. 259: p. 6818-25 (1984); E.G. Bremer, J. Schlessinger, S. Hakomori,

J. Biol. Chem. 261: p. 2434-40 (1986)), PDGF driven 3T3 fibroblast and SH-SY5Y neuroblastoma cell lines (E.G. Bremer, S. Hakomori, P.D. Bowen, E. Raines, R. Ross, J. Biol. Chem. 259: p. 6818-25 (1984); E.G. Bremer, j. Schlessinger, S. Hakomori, J. Biol. Chem. 261: p. 2434-40 (1986); D.L. Hynds, M. Summers, B.J. Van, M.S. O'Dorisio, A.J. Yates, J. Neurochem. 65: p. 2251-8 (1995)), FGF driven

BHK fibroblast cell line (E.G. Bremer, S. Hakomori, Biochem. Biophys. Res. Commun. 106: p. 711-8 (1982)), and insulin driven HL-60 leukemia cell line (N. Nojiri, M. Stroud, S. Hakomori, J. Biol. Chem. 266: p. 4531-7 (1991)), are inhibited by exogenously added GM3. The phosphorylation of the appropriate receptors, EGF-receptor (R), PDGF-R, and Insulin-R, are inhibited by GM3 (E.G.

Bremer, S. Hakomori, P.D. Bowen, E. Raines, R. Ross, J. Biol. Chem. 259: p. 6818-25 (1984); E.G. Bremer, J. Schlessinger, S. Hakomori, J. Biol. Chem. 261: p. 2434-40 (1986); N. Nojiri, M. Stroud, S. Hakomori, J. Biol. Chem. 266: p. 4531-7 (1991); Q. Zhou, S. Hakomori, K. Kitamura, Y. Igarashi, J. Biol. Chem. 269: p. 1959-65 (1994)).

Several studies suggest the another ganglioside, GM1, has supportive effects on CNS neurons (C.L. Schengrund, Brain Res. Bull. 24: p. 131-41 (1990)). Less is known about the supportive effects of GM3. GM3 also inhibits cell growth and modulates cell differentiation of the human leukemia cell lines HL-60 and U937, and induces monocytic differentiation (N. Nojiri, F. Takaku, Y. Terui, Y. Miura, M.

Saito, Proc. Natl. Acad. Sci. USA. 83: p. 782-6 (1986)), while anti-GM3-antibody mediates differentiation of neuro-2a neuroblastoma cells (D. Chatterjee, M. Chakraborty, G.M. Anderson, Brain Res. 583: p. 31-44 (1992); M. Chakraborty, G.M. Anderson, A. Chakraborty, D. Chatterjee, Brain Res. 625: p. 197-202 (1993)). The neuritegenesis of neuro-2a and PC- 12 cells are increased by GM3 (M.S.

Cannella, F.J. Roisen, T. Ogawa, M. Sugimoto, R.W. Ledeen, Brain Res. 467: p. 137-43 1988)), and the number of thickness of processes of cultured oligodendrocytes are also increased (S.H. Yim, E. Yavin, J.A. Hammer, R.H. Quarles, J. Neurochem. 57: p. 2144-7 (1991); S.H. Yim, R.G. Farrer, J.A. Hammer, E. Yavin, R.H. Quarles, /. Neurosci Res. 38: p. 268-81 (1994)).
Although gangliosides are abundant in brain, little is known about their function on CNS cells especially during development.
Gangliosides have been proposed to regulate cellular differentiation in neurons (Cannella, M.S., et al., Comparison of epi-GM3 with GM3 and GM1 as stimulators of neurite outgrowth, Devel. Brain. Res., 1988, 39: p. 137-143;

Dimpfel, W., W. Moller, and U. Mengs, ed. Ganglioside-induced neurite formation in cultures neuroblastoma cells, Gangliosides in Neurological function, ed. M.M. Rapport and A. Gorio. 1981, Raven: New York. 119-124; Ferrari, G., M. Fabris, and A. Gorio, Gangliosides enhance neurite outgrowth in PC12 cells, Dev. Brain Res., 1983, 8: p. 215-221; Katoh-semba, R. S.D. Skaper, and S. Varon, Interaction of GM1 ganglioside with PC12 phenochromocytoma cells: serumand NGF-dependent effects on neurtitic growth and proliferation, J. Neurosci. Res., 1984, 12: p. 299-310; Leskawa, K.C. and E.L. Hogan, Quantitiation of the in vitro neuroblastoma response to exogenous purified gangliosides, J. Neurosci. Res., 1985, 13: p. 539-550; Matta, S.C., G. Yorke, and F.J. Roisen, Neuritic and metabolic effects of individual gangliosides and their interaction with nerve growth factor in cultures of neuroblastoma and phenochromocytoma, Dev. Brain Res., 1986, 27: p. 243-252), while a variety of gangliosides are expressed in normal and transformed glia (Irwin, L.N., D.B. Michael, and C.C. Irwin, Ganglioside patterns of fetal rat and mouse brain, J. Neurochem., 1980, 34: p. 1527-1530; Goldman, J.E., S.S. Geier, and M. Hirano, Differentiation of astrocytes and oligodendrocytes from germinal matrix cells in primary culture, J. Neurosci, 1986, 6(1): p. 52-60; Sung, C.C, et al, Glycolipids and myelin proteins in human oligodendrogliomas, Glycoconjugate J., 1996, 13: p. 433-443; Kim, S.U., G. Moretto, and R.K. Yu, Neuroimmunology of gangliosides in human neurons and glial cells in culture, J. Neurochem., 1986, 15: p. 303-321; Shinoura, N., et al, Ganglioside composition and its relation to clinical data in brain tumors, Neurosurg., 1992, 31: p. 541-549; Stojiljkovic, M., et al, Gangliosides GM1 and GM3 in early human brain development: an immunocyltochemical study, Int. J. Devel. Neurosci., 1996, 14: p. 35-44). The majority of human brain tumors are of glial origin. In children gliomas comprise 67% of CNS tumors, and are the second most frequently diagnosed childhood malignancy (Parkin, D.M., et al, International incidence of childhood cancer, 1988, I ARC Scientific Publication, International Agency for Research on Cancer). In adults brain tumors represent between 40% and 67% of primary CNS tumors (Bohnen, N.I., et al, ed. Descriptive and analytic epidemiology of brain tumors, Cancer of the Nervous System, ed. P.M. Black and J.S. Loeffler, 1997, Blackwell

Science: Cambridge Mass. 3-24). CNS tumors account for approximately 2% of all adult malignancies (Giles, G.G., B.K. Armstrong, and L.N. Smith, Cancer in Australia, 1987), but are responsible for disproportionately high number of years of life lost (Hoffman, R.M., ed. Fertile seed and rich soil; The development of clinically relevant models of human cancer by surgical orthotopic implantation of intact tissue,

Anticancer drug development guide: Preclinical screening, clinical trials and approval., ed. B.A. Teicher, 1997, Humana Press: Totowa, N.J. 127-144). Of primary CNS malignancies, glioblastoma multiforme (GBM) has the poorest prognosis, with median survival between 29 to 36 weeks (Ammirati, M., et al, Effects of the extent of surgical resection on survival and quality of life in patients with supratentorial glioblastomas and anaplastic astrocytomas, Neurosurgery, 1987, 21: p. 201-206; Harsch, G.R., et al, Reoperating for recurrent glioblastoma and anaplastic astrocytoma, Neurosurgery., 1987, 21: p.615-621). Even with aggressive surgical resection followed by adjuvent therapy (radiotherapy, immunotherapy, chemotherapy), the vast majority of GBM patients succumb to their disease (Wen, P.Y. and D. Schiff, ed. Clinical evalustion of patients with astrocytomas,

Contemporary issues in Neurological surgery - Astrocytomas: Diagnosis, Treatment and Biology., ed. P.M. Black, W.C. Schoene, and L.A. Lampson, 1993, Blackwell Scientific: Boston, Mass. 26-36). Treatment of CNS malignancies is complicated due to the sensitivity of collateral brain tissue and the profound consequences of damage to cells adjacent to the tumor.

Summary of the Invention
The present invention relates to the ability of the ganglioside, GM3 to inhibit proliferation and induce apoptosis in proliferating CNS cells. The present invention further demonstrates the ability for GM3 to reduce cell numbers in primary cultures of rapidly proliferating human glial tumors and the 9L rat gliosarcoma cell line. In addition, GM3 is shown to have no effect on quiescent cultures of normal human CNS cells. A single injection of GM3 three days after intracranial implantation of tumor cells in a murine xenograft model system resulted in a significant increase in the symptom-free survival period of host animals. Therefore,

GM3 is useful as a chemotherapeutic agent for human high grade gliomas.
As such, a first aspect of the invention includes a novel chemotherapeutic agent for inducing apoptosis in proliferating neural cells comprising GM3.
A second aspect of the invention relates to selective induction of apoptosis in proliferating neural cells using a GM3 chemotherapeutic agent.
A further aspect of the invention relates to treatment of a patient having a brain tumor to inhibit proliferation and induce apoptosis of proliferating neural cells in said patient.
An additional aspect of the invention relates to a chemotherapeutic regimen to treat patients in need thereof.

Therefore, the present invention has the advantage of being able to selectively treat brain tumors including high grade gliomas in patients thereby enhancing the prognosis of the patients by effectively increasing the median survival time for such patients.
Another advantage relates to the use of a chemical agent normally produced in the body in a chemotherapeutically effective composition thereby reducing the risk of adverse side effects typically related to such treatments.
Still other advantages and benefits of the invention will become apparent to those skilled in the art upon reading and understanding the following detailed description of the preferred embodiments.

Brief Description of the Drawings
FIGURE 1 shows TUNEL labeling showing DNA fragmentation for proliferating astrocyte cells treated with 50μM GM3.
FIGURE 2A and 2B show proliferative astrocyte morphology without

GM3 treatment and with GM3 treatment.
FIGURE 3A and 3B show the effect of GM3 on non-proliferative astrocytes.
FIGURE 4 shows TUNEL labeling showing DNA fragmentation for proliferating neurons and glial precursor cells treated with GM3.
FIGURE 5 shows the effect of GM3 treatment on human glial tumor growth and on rat 9L cell line in vitro.
FIGURE 6A-6E show morphological changes in control cell cultures (6A, 6C) and GM3 treated cell cultures (6B, 6D, 6E).
FIGURE 7 shows the effect of increasing the concentration of GM3 on human glioblastoma multiforms (GBM's) in vitro.
FIGURE 8 shows the effects of GM3 on normal human central nervous system (CNS) tissue cultures.
FIGURE 9 shows the effect of GM3 treatment on survival time of animals implanted with 9L rat tumor cells.

Detailed Description of the Preferred Embodiment
Exposure of proliferating rat CNS (central nervous system) neural cells to the ganglioside GM3 inhibits cell proliferation and induces apoptosis. The present invention demonstrates that GM3 treatment of human CNS tumor cells inhibits their growth in vitro and results in prolonged symptom-free post-implant intervals in a xenograft brain tumor model.
The mechanism by which GM3 inhibits the expansion of human GBM in vitro is unclear. Without intending to be bound by any particular theory, it seems likely that GM3 may act to both inhibit tumor cells proliferation, as well as to induce apoptosis in actively proliferating cells. Consistent with this hypothesis, GM3 treatment resulted in a significant inhibition of proliferation of rat neural cell precursors and a rapid induction of apoptosis which is correlated with an up regulation of p27kιp"' expression and a reduction of pRb (hyperphosphorylated retinoblastoma protein) expression. Two lines of evidence suggest that human tumor cells also undergo apoptosis in response to GM3 exposure. First, the number of cells in GM3 treated cultures was reduced compared to both the starting population and to control cultures suggesting cell death. Second, the rounded morphology and nuclear fragmentation seen following GM3 exposure is characteristic of apoptosis.
The degree of response of cultures to GM3 exposure appears to be cell or tumor specific. While all tumors studied in the present invention were characterized as glioblastoma multiforme (GBM) and where grown at similar densities under identical conditions, their response to GM3 varied. In some cultures the reduction in cell number was greater than 80%, while in parallel experiments using cultures of a different tumor, the reduction in cell number was only 35%. This difference in response to GM3 treatment is unlikely to be due to culture passage number since the seven primary cultures were derived from the original plating, or early passages of the resected tumor.
One factor which might regulate responsiveness of cells from different tumors to GM3 treatment is the rate of proliferation. Non-proliferative rodent CNS cells do not undergo apoptosis in response to GM3 treatment in vitro, but rather become increasingly differentiated. Similarly, intraventricular injection of GM3 into developing rats results in cell death only in ventricular and subventricular proliferating cell populations. As seen in the present invention, GM3 did not reduce cell number in quiescent cultures of normal human brain cells and there were no changes in neurological behavior associated with injection of GM3 in rats. Thus, cells in cultures of different tumors may have different cell cycle times and the response to GM3 may be directly correlated with cell cycle time (i.e. faster cycle times result in increased cell death).
The present invention demonstrates that GM3 treatment significantly decreased the growth of primary cultures of human and rodent tumor cells, while not significantly altering cell number in quiescent cultures of normal human brain. In addition, a single treatment with GM3 significantly extended the symptom-free, post implant period in nude mice with intracranially implanted rat 9L brain tumor cells. Taken together, these data demonstrate that GM3 provides an effective therapeutic treatment for human high grade gliomas.
Below are several examples showing both in vitro and in vivo results upon application of GM3 to both tumorous and normal cell lines. The GM3 used in the examples was obtained from Sigma and was 98% pure from bovine brain. Normal saline (sterile) was used as a carrier.

Example 1
In this example, GM3 is shown to induce apoptosis of proliferating astrocytes. Astrocyte cells were prepared from cerebral cortices of newborn (P0) Sprague-Dawley rats as described in Smith et al., Dev.Biol. 138, 377-390 (1990), and maintained in Delbecco's modified Eagles medium (DMEM) supplemented with 10% fetal bovine serum (FBS).
To examine the mechanisms of GM3 induced astrocyte cell death, the cells were double labeled with anti-GFAP antibody and the terminal deoxynucleotidyl transferase-mediated dUTP-digoxigenin nick-end labeling (TUNEL) method. Morphological changes of astrocytes were apparent after 18 hours of incubation with 50 μM of GM3. By 72 hours, most of the cells were highly shrunk and their nuclei condensed, compatible with the induction of apoptosis. Consistent with these morphological changes, the proportion of TUNEL positive cells increased substantially. By 72 hours of incubation, some of the dead cells detached from cover slips, so the proportion of TUNEL positive cells in these cultures may be an understatement of the extent of cell death. In all the cultures, a few GFAP negative cells, presumably meningeal or endothelial cells, appeared normal even after 72 hours of incubation with GM3.
Since any random DNA cleavage caused by necrosis can potentially be labeled by TUNEL, apoptotic internucleosomal DNA fragmentation was assayed (Figure 1). Characteristic DNA fragmentation was first seen after 48 hours and was clearly apparent after 72 hours of incubation with GM3, consistent with the TUNEL staining data. Ultrastructurally, characteristic features of apoptotic cell death such as cell shrinkage, condensation of chromatin, loss of the integrity of the nuclear and plasma membranes were observed (Figure 2).

Example 2
To determine if the response to GM3 was correlated with the proliferative state of astrocytes, GM3 was added to contact-inhibited non-proliferating astrocyte cultures. GM3 did not induce changes in cell morphology or cell death in these contact-inhibited non-proliferating astrocytes (Figure 3).

Upon analyzing the results from Examples 1 and 2, it is seen that GM3 selectively induces apoptosis in immature, proliferating astrocytes. Further, the results also demonstrate that active proliferation of neural cells also contributes to the susceptibility of those cells to GM3 treatment.

Example 3
In this example, GM3 is shown to induce apoptosis of neurons and glial precursors. Mixed cell cultures (post-mitotic neurons, neuronal and glial precursors) were prepared from El 5 brain according to Smith et al., Dev. Biol, 138, 377-390 (1990), and maintained in DMEM + 10% FBS or in DMEM + 1% FBS +

1% N2 supplement.

To evaluate the effect of GM3 on neurons and glial precursors, the mixed cultures were treated with GM3, and cell death was assessed by double labeling with anti-β-tubulin-positive and A2B5-positive cells were TUNEL-positive and the characteristic DNA fragmentation became apparent (Figure 4). By 72 hours of incubation, approximately 80 % of the cells were dead. The remaining cells appeared intact and the majority of these cells were β-tubulin-positive neurons, consistent with the process-outgrowth promoting effect of GM3 on mature neurons. When these remaining cells were incubated with BrdU during the last 24 hours of GM3 treatment and double labeled with anti-β-tubulin and anti-BrdU antibodies, the majority (>99%) of remaining viable cells were β-tubulin+/BrdU-, suggesting that post-mitotic neurons, like non-proliferative astrocytes were less sensitive to GM3 treatment. Together, these observations indicate that exposure to GM3 induces apoptosis in all classes of proliferating but not non-proliferating neural cells.
In the following Examples, all tumor specimens were diagnosed as glioblastoma multiforme (GBM), corresponding to W.H.O. grade 1.1.3 and/or St.

Anne-Mayo grade IV. Non-neoplastic cells were obtained from a 3 yr. old male that underwent a left hemispherectomy for Rasmussen's disorder.
Cultures were generated using standard protocols. Briefly, tumor samples were dissociated enzymatically, plated at densities between 1X105"6 on 12mm poly-L-lysine coated cover slips and maintained at 37°C with 5% CO2.

Cultures were grown for at least 24 hours prior to addition of 100 μM GM3 suspended in media to experimental cultures. Parallel control cultures received media alone at the commencement of the experiment. The majority of assays were performed from cells derived from the original plating or first passage of a tumor sample. Cultures were grown in DMEM media with 10% FBS and N2 supplement containing insulin, transferrin, selenium, progesterone, putrescine, 3, 3 ',5 triiodo-L-thyronine, thyroxine and fraction V BSA.
Following exposure to GM3 for seven days, cultures were fixed in 4% paraformaldehyde, stained with the DNA stain Dapi, (1 : 10,000; Molecular Probes) for 5 min. and examined on a Zeiss Axiophot microscope. To quantify remaining cells, the number of cells in 10 consecutive 40X randomly selected fields was counted and the mean determined. For each tumor at least 3 independent cultures were assayed and the data pooled. To determine the effect of increasing GM3 concentrations, these studies were repeated at concentrations of 100 μM, and 400 μM in parallel.

Example 4
In the following example, GM3 is applied to human high grade gliomas in vitro to determine whether GM3 treatment inhibited expansion of primary human tumor cells. Tumor samples used in this example were high grade lesions from a variety of regions of the tumor. Seven independent tumor samples were assayed.
In each tumor sample, 100 μM GM3 was applied in a single application. This resulted in a significant reduction in the number of cells in these experimental cultures over a period of seven days (Figure 5).
The effect of GM3 treatment on cell number in the tumor cultures, relative to matching untreated control cultures, was tumor specific. For example, while cell number was reduced by approximately 57% in cultures of GBM-1, cell number was reduced by greater than 85% in cultures of GBM-7. The effect of GM3 was not limited to primary tumor cultures. Cell number in cultures of rat 9L cell line, an established brain tumor model for in vivo studies, was reduced by approximately 65%, close to the average reduction in the human primary cultures (Figure 5).
The morphology of cells in GM3 treated cultures was different from the control cultures (Figure 6A-6E). In control cultures, cell density increased during 7 days in vitro and the majority of cells were flattened and well adhered. By contrast in GM3 treated cultures, cell density decreased over the same period and the residual cells were rounded and highly refractile. Many cells in GM3 treated cultures, but not in control cultures, had chromosomal fragmentation revealed by the DNA stain Dapi (Figures 6A-6E) and GM3 cultures contained a significant amount of cellular debris. The effects of GM 3 treatment on cell number were increased at higher concentrations. In parallel cultures of tumor GBM-5 exposed to 100 μM or 400 μM GM3, there were significantly fewer cells in cultures treated with 400 μM GM3 than in those treated with 100 μM GM3 (Figure 7).

Example 5
To determine whether GM3 treatment reduced cell number in quiescent cultures of normal human CNS, exogenous GM3 (100 μM) was added to cultures derived from non-neoplastic CNS tissue. The density of cells was not significantly altered curing 7 days in vitro. Furthermore, there was no significant difference in the number of cells in GM3 treated cultures compared to controls (Figure 8), suggesting GM3 was not toxic to normal human CNS cells, consistent with all previous studies.

Example 6
To assess the effects of GM3 on tumor growth in vivo a 9L rat gliosarcoma cells were transplanted intracranially into mice. Between 1-1.5X10° cells were stereotactically injected into the brain parenchyma of 18 adult Swiss nude mice, in two separate experiments. In each experiment, the animals were randomly assigned to either control or experimental groups immediately following the implantation procedure. Seventy two hours after cell implantation control animals received an intracranial injection of 5 ml of 0.9% sodium chloride while experimental animals received 0.3 mg GM3 in 5 mol of 0.9% sodium chloride. All animals were evaluated daily for signs of neurological impairment (seizures, hemiparesis, ataxia, dyskinesia, gait difficulties, etc.) and, when detected, the animal was sacrificed. Survival rates from both experiments were pooled, the means +/-S.D. [standard deviation] determined and the statistical significance assessed by a 2 tailed

Students t test. Of the animals in group 1, one of the animals demonstrated symptoms of a rapidly progressing infection and had to be sacrificed prior to exhibiting any symptoms of tumor growth. This animal was excluded from the data set.
Injection of GM3 significantly prolonged the symptom-free period in host animals (Figure 9). Among control animals (N=9) the mean post implant survival period was 18 days (+/-3). At 13 days approximately 20% (2/9) animals demonstrated neurological impairment and no animals survived without noticeable neurological impairment beyond 23 days post implantation. By contrast, among experimental animals that received a single injection of GM3 (N=8), the mean post implant survival period was 23 days (P-value compared to control of 0.03).

Although one animal developed neurological impairment at day 17, two animals were symptom-free until 30 days post implant; that is 7 days after all control animals were sacrificed.
The above Example demonstrates that a single treatment of GM3 increased the mean symptom free survival time of host animals following implantation of 9L cells. Based on the in vitro assays, this increased survival time most likely reflects reduction in the rate of growth of implanted tumor cells following GM3 treatment. Since the median survival for a patient diagnosed with a GBM is estimated between 29-36 weeks, any increase in median survival would be beneficial, provided there was little inherent treatment toxicity. GM3 is a naturally occurring molecule that enhances maturation rather than induce cell death in non-proliferative neural cells. Thus GM3 treatment is unlikely to severely compromise the quality of remaining life.
Several additional approaches may enhance the symptom-free survival period of GM3 treated host animals. First, in vitro studies suggest the efficacy of

GM3 induced cell death is enhanced at higher concentrations. While any effect on non-proliferative neural cells may be more pronounced at higher concentrations, increasing GM3 concentrations may result in extended survival periods. Second, the current studies used a single injection of GM3, 3 days after tumor implantation. Multiple injections of GM3 should result in a greater limitation of tumor growth and concomitant extension of symptom-free survival periods. Recent in vitro studies indicate that sustained application of GM3 over a period of 1 week resulted in greater cell loss in the majority of tumor cultures approaching 100%.
Taken together these studies identify GM3 as a treatment agent for patients diagnosed with GBM. Treatment of proliferating CNS cells with GM3 inhibits cell proliferation and induces apoptosis. Likewise, growth of human primary GBM tumor cells is inhibited by GM3 treatment. By contrast, normal, non-proliferative CNS cells are not induced to undergo apoptosis following GM3 treatment. This strong proliferation dependence of GM3 responsiveness indicates that GM3 is a powerful agent in the treatment of highly proliferative human CNS malignancies such as GBM.
Induction of apoptosis of various types of neural cells may be accomplished in practicing the present invention. Glioblastoma multiformes (GBMs), astrocytomas (astrogliomas), ohgodendrocytomas, ependymomas, glial tumors and other various mixed gliomas are within the scope of cells to be treated according to the present invention.
Various treatment regimens may be utilized in accordance with the present invention. For example, the administration of GM3 may be in a single application, multiple applications, or as a slow release suspension polymer formulation. The amount to be administered is dictated by the amount needed to inhibit tumor growth. Typical ranges of effective amounts of GM3 to be administered to a patient range from about 5μM to 50μM (or about lmg to 16g.). The treatment regimen should last at least until tumor growth has been completely inhibited. Generally, this may be from about one day to about six months (or longer, if necessary).
Various conventional means for delivering chemotherapeutic agents may be used in accordance with practicing the present invention. Such means include direct application of the GM3 to the tumor (either alone or in combination with a slow release polymer), intraarterial injection, and stereotactic injection intratumorally.
In addition, other adjuvant therapies and chemotherapeutic treatment methods may be used in conjunction with the GM3 treatment. For example, pre- or post- administration of radiation therapy is an effective adjuvant therapy which may be used in conjunction with the GM3 treatment.
The invention has been described with reference to the preferred embodiments. Obviously, modifications and alterations will occur to others upon reading and understanding the preceding detailed description. It is intended that the invention be construed as including all such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.