Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
(CANCER RESEARCH 51, 1713-1719. March 15. 1991] Effect of Alkyl-lysophospholipid Brain Tissue in Vitro ' on Glioblastoma Cell Invasion into Fetal Rat Olav Engebraaten,2 Rolf Bjerkvig, and Michael E. Berens1 The (jade Institute, Department of Pathology, University of Bergen, A'-5021 Bergen, Norway ¡O.E., R. B.f, and The Brain Tumor Research Center, University of California, San Francisco, San Francisco, California ¡M.E. B.J ABSTRACT The antitumor effect of alkyl-lysophospholipid (ALP) was studied on a continuous glioma cell line (GaMg) as »ellas on tumor spheroids obtained from three different primary brain tumor biopsies. GaMg monolayer growth was reduced by 50% after treatment with 30 MMALP; cells accumulated in the G2M phase of the cell cycle as determined by flowcytometric analyses. Tumor spheroid growth was reduced by 25 and 44% during treatment with 10 and 30 MMALP, respectively. These drug concentrations also caused a severe destruction of spheroids. No effect on growth or morphology was seen in spheroids treated with 0.1 and 1.0 MMALP. ALP caused a dose-dependent inhibition of invasion by GaMg tumor spheroids into brain aggregates. After 168 h of 1.0 MMALP treatment, the volume of the intact brain aggregate was 90% larger than that in the untreated co-cultures. To further investigate the efficacy of ALP as an anti-invasive drug, co-cultures were performed with specimens obtained from three primary brain tumors: a highly invasive glioblastoma multiforme, an anaplastic astrocytoma, and an astrocytoma. Treatment of spheroids from the most invasive tumor with ALP caused a 7-fold preservation of normal brain tissue relative to control co-cultures. Moreover, the sensitivity of primary glioma spheroids to the anti-invasive effect of ALP seemed to be associ ated with the aggressiveness of the tumor; spheroids from the more malignant specimen (glioblastoma multiforme) were more sensitive than those from the less aggressive tumors. The anti-invasive effect seen with nontoxic concentrations of ALP may prove valuable in the treatment of malignant gliomas. INTRODUCTION Several characteristic changes on the cell surface occur with transformation. For tumors of glial linage, these include alter ations in the expression of growth factor receptors (1-3), as well as changes in the carbohydrate composition on the plasma membrane (4). Gliomas are known for their invasiveness into the normal brain where tumor cells are often found at consid erable distance from the macroscopic border of the primary mass (5). A complete tumor resection is therefore difficult and the prognosis in these patients is generally poor. Drugs that influence the cell membrane or functional components of the cell membrane are therefore of considerable interest in cancer therapy (6). The antineoplastic drug Et-18-O-CH., belongs to a group of ALP,4 which are synthetic analogues of the naturally occurring 2-lysophosphatidylcholine (7). This compound and structural analogues are known to accumulate at the cell surface, where they exert their antineoplastic action (8). These drugs are reReceived9/25/90;accepted1/7/91. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ' O. E. and R. B. were supported by The Norwegian Cancer Society. Grants 88218 and 89002. M. E. B. was supported by a grant from the Preuss Foundation. 2To whom requests for reprints should be addressed, at The Cade Institute. Department of Pathology, University of Bergen. Haukeland Hospital. N-5021 Bergen. Norway. 3Current address: Laboratory of Neuro-Oncology Research, Barrow Neuro logical Institute, 350 West Thomas Road. Phoenix, ÀZ85013-4496. 4 The abbreviations used are: ALP. alkyl-lysophospholipid. DMEM. Dulbecco's modification of minimal Eagle's medium. garded as biological response .modifiers acting both as immunopotentiators and as cytotoxic agents (9). The ALPs have been shown to inhibit growth of several different tumors both in vitro and in vivo (9-11), and exert an anti-invasive effect (12, 13). In this study, the cytotoxic and anti-invasive effects of Et-18O-CHj were studied on a well-characterized human glioma cell line (GaMg) (14) as well as on tumor spheroids generated from 3 different primary brain tumor biopsies (15). For this purpose, an in vitro invasion assay has been developed in which normal fetal rat brain cell aggregates were used as target tissue for human gliomas (16-18). MATERIALS AND METHODS Drug. The ether lipid l-octadecyl-2-methyl-rac-3-glycerophosphocholine was obtained from Medmark Pharma Gmbh, Federal Republic of Germany. The drug was dissolved in 100% ethanol and stored at 4°Cuntil use, at which time serial dilutions in DMEM to the final concentration were made. Cell and Tissue Culture. The human glioma cell line GaMg (14) (passages 25-40) has been characterized before. Briefly, this cell line was obtained from a 42-year-old female and histologically identified as a glioblastoma. The biopsy specimen and primary cultures contained many diploid cells, as analyzed by flow-cytometric DNA measurement. This changed within one year of subculture, at which time a neartriploid DNA content was found. Karyotyping revealed a mean chro mosome number of 67.1 ±14.3 (SD). The cell lines were grown in DMEM supplemented with 10% heat-inactivated newborn bovine serum, 4 times the prescribed concentration of nonessential amino acids, 2% L-glutamine, penicillin (100 lU/ml), and streptomycin (100 Mg/ml). The cells were kept in a standard tissue culture incubator at 37°C,with 100% relative humidity, 95% air, and 5% CO2. Cells were free of Mycoplasma contamination. Tumor spheroids were initiated using the agar overlay culture method described by Yuhas et ai ( 19). Briefly, spheroids were formed by seeding 5 x IO6cells in 20 ml of growth medium into 80-cttr agar-coated tissue culture flasks (Nunc, Denmark). The spheroids were used between days 7 and 14. Spheroids from Primary Brain Tumor Specimens. Tumor tissue from patients undergoing craniotomy was cut into small fragments and cultured according to the method described by Bjerkvig et al. (15). Briefly, tumor tissue from 3 intracranial tumors, one glioblastoma multiforme, one anaplastic astrocytoma, and one astrocytoma, was dissected out at surgery and transferred to DMEM. Macroscopically representative tissue from the tumor was then selected for tissue culture and prepared for histopathological diagnosis. The specimens were cut into 0.5-mm-diameter fragments and placed into culture in 80-cm2 tissue culture flasks (Nunc) base-coated with 0.75% agar (Difco Agar Noble; Difco Laboratories, Detroit, MI) in DMEM. They were then cultured until they rounded up and formed spheroids. The volume of the overlay suspension was 10 ml of DMEM. The tissue was kept in a standard tissue culture incubator (100% relative humidity, 95% air and 5% CO2). Primary brain tumor spheroids with a diameter of 295 ±5 (SE) Mmwere selected for further use (for further details, see Ref. 15). Assays for Growth. Cell sensitivity to ALP was tested in microtiter culture wells using the MTT assay (20, 21). Briefly, cells in exponential growth phase were collected by trypsinization, and then plated in microtiter wells (1000 cells/well in 200 M')- The following day, ALP was added to the wells to achieve the final concentrations (see "Re- 1713 Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1991 American Association for Cancer Research. ALPS INHIBIT GLIOBLASTOMA suits"). After 5 additional days in culture, 50 p\ of MTT (Sigma Chemical Co., St. Louis, MO) (5 mg/ml in Hanks' balanced salt solution) were added to each well, and the plates were incubated at 37°Cfor 2 h. The liquid contents of the wells were emptied by rapid inversion of the plate, and the wells were dried in a 37°Cdry incubator. Light mineral oil (Sigma) was added to the wells to solubilize the formazan salt, and the absorbance above background (wells containing medium but no cells) was measured at 540 nm in a Titertek MCC340 (Flow Laboratories, McLean, VA). The absorbance from the wells containing untreated cells was used to calculate the fractional absorb ance of the wells with treated cells. Experiments were done in replicates of 6, and the mean absorbance was calculated. The experiment was repeated twice. Monolayer growth curves were obtained by seeding 1.75 x IO5cells in 5 ml into 25-cnr tissue culture flasks (Nunc). After 24 h of culture, the ether lipid was added to achieve final concentrations of 0.1, 1.0, 10, and 30 pM. Three flasks in each treatment group were trypsinatcd and counted electronically, using a Coulter Counter (Coulter Electronics, Luton, Bedfordshire, United Kingdom). Spheroid growth curves were obtained by transferring individual spheroids [diameter. 278 ±3 ¿im(SE)] into 16-mm multiwell dishes (Nunc) base-coated with 0.5 ml of 0.75% DMEM-agar. The overlay suspension contained 1.0 ml of growth medium. The ET-18-O-CH, was added to the wells at 0.1, 1.0, and 10 J/M concentrations. The spheroid size was recorded regularly during 8 days by measuring 2 orthogonal diameters using an inverted phase-contrast microscope with a calibrated reticle in the ocular. Spheroid volume was calculated using the formula for volume of a sphere, and expressed as a fraction relative to the volume at day 0. Ten spheroids were measured in each group and the experiments were done in triplicate. Directional Migration. The area covered by GaMg cells migrating out from the spheroids explanted on a plastic surface was used as an index of cell migration (22). Ten spheroids (diameter, 278 ±3 urn (SE)] in each treatment group were transferred individually into 16-mm multiwell dishes (Nunc). After 2, 5, 8, and 11 days of culture, the diameter of each colony was measured in the presence of different concentrations (0.1, 1.0, 10, and 30 /t\\) of ether lipid. The experiments were done in triplicate. Flow Cytometrj. After days 2, 4, 6, and 8 of exponential monolayer growth, suspensions of the treated and control cells were fixed in icecold 100% ethanol and stored at 40°Cuntil use. The fixed cells were incubated for 15 min with 0.5% pepsin at 37°C,and the free nuclei CELL INVASION l.\ tITRO and allowed lo "mature" for 18 days. At this time, the fetal cells had formed stable aggregates and developed into morphologically mature brain cells. Mature aggregates, with a diameter of 281 ±4 /tm (SE), were transferred individually into agar-coated 96-well tissue culture dishes (Nunc). Tumor spheroids with the same size as the aggregates were also transferred to the wells, and DMEM (containing the actual con centrations of ALP) was added, giving a total volume of 300 ¿Jin each well. The normal brain aggregates and the tumor spheroids were then brought into proximity by use of a sterile syringe. The orthogonal diameters of the brain aggregates and tumors were then measured. The co-cultures were fixed after 72 or 168 h for light-microscopic exami nation and morphometric analyses. In the co-cultures involving the GaMg glioma cell line, there were 7 observations in each treatment group at the different fixation points, and the experiments were done in triplicate. In the invasion experiments using tissue from intracranial tumors, there were 5 co-cultures in each treatment group. Light Microscopy. The co-cultures were fixed for 24 h in 2% glutaraldehyde in sodium-cacodylate buffer (300 ±25 mOsm). The speci mens were then washed in the same buffer, without glutaraldehyde, and postfixed for l h in 1% OsO4 before serial dehydration in increasing gradients of ethanol to 100%. Embedding in Epon 812 was performed using graded mixtures of epon-propylenoxide. The specimens were polymerized for 48 h at 60°C. Morphometry. In order to determine the amount of remaining brain tissue after 72 and 168 h of tumor cell invasion. Chavalieri's principle for direct estimation of volume from systematically sampled sections was used (27, 28). Serial semithin sections (1.5 pm) were prepared from the specimens using a Reichert-Jung Microtome 2040 (Reichert-Jung, Austria). Every 15th section was sampled starting with one of the 10 first sections that was randomized. The sections were then stained with toluidine blue for light-microscopic examination. The brain tissue and the tumor tissue were easily discriminated by the difference in mor phology. The areas of the tumor tissue and remaining brain tissue were then measured on each section by morphometry, using a Kontron Image Analysis system (Kontron, Federal Republic of Germany). The area of tumor spheroid and brain aggregate measured on each slide, multiplied by the distance between every sampled section (15 x 1.5 ^m), resulted in an estimate of the volume for each slice. The total volume of the tumor tissue as well as the remaining brain tissue was calculated by summation of all the individual volumes. The coefficient of error in this method for estimating volumes of sections is accepted to be less than 5% (27). Statistics. The Student's t test was used to determine the level of significance of the difference between individual groups. The doseresponse relationship in the growth curve experiments, and in the invasion assay, was assessed statistically by regression analysis. were subsequently washed in 0.9% saline and treated for 5 min with 1 mg/ml ribonuclease dissolved in 0.9% saline (Sigma). Staining of DNA was performed with a mixture of ethidium bromide and mithramycin (for details, see Ref. 23). The relative cellular DNA content was measured with a Phywe ICP-22 pulse cytophotometer (Ortho Instru ments, Westwood, MA). Normal human lymphocytes were used for instrument adjustment and calibration. Cell cycle analysis was per RESULTS formed using a planimetrie method originally described by Baish et al. (24); 15,000 cells were analyzed in each of 3 replicate observations in MTT—Assay and Tumor Growth. The GaMg cells showed a the different treatment groups (coefficient of variation <5%). dose-dependent inhibition of growth by ALP after 6 days in the Assay for Tumor Cell Invasiveness. A standardized in vitro invasion MTT assay; the 90% inhibitory concentration was 50 MM(Fig. assay was used to study the anti-invasive effects of Et-18-O-CH., (17, 1). Cells in the growth curve assay were in exponential phase 18). Fetal rat brain cell aggregates were used as a normal target tissue for 4 days, showing 13% (P < 0.025) and 35% (P < 0.005) for the tumor spheroids. Such brain aggregates have a well-developed neurophil with mature neurons, astrocytes, and oligodendrocytes. My- growth inhibition by 10 and 30 MMALP, respectively (Fig. 2). No effect from 0.1 or 1.0 MMtreatment was detected. elinated axons and synaptic complexes are frequently observed in the Dose-dependent inhibition of spheroid growth was detected aggregates thus showing many similarities to the normal adult brain (16, 25). The procedure for obtaining fetal rat brain aggregates has using 10 and 30 MMALP (Fig. 3). These concentrations of ALP been described in detail elsewhere (16). Briefly, 18-day-old fetuses from also elicited severe damage to the integrity of the spheroids pregnant BD IX rats (26) were removed by cesarean section. The brains when examined by phase-contrast microscopy. The spheroids of the fetuses were carefully dissected out, minced, washed 3 times in had a ruffled surface appearance and pronounced cell shedding phosphate-buffered saline, and serially trypsinized. The resulting singleinto the medium. Spheroids treated with ALP at 0.1 and 1.0 cell suspension was then centrifuged at 1500 rpm for 5 min and MM,or the control wells maintained smooth external surfaces resuspended in DMEM; IO7cells were then seeded into 24-well dishes and did not show any signs of cytotoxicity. (Nunc) previously coated with 0.5 ml of DMEM-agar. The cells were Cell Cycle Effects. Treatment with 30 //M Et-18-O-CH, allowed to reaggregate for 2 days. Thereafter, the aggregates were transferred to a larger number of wells (5 to 7 aggregates in each well) caused an accumulation of cells in the G2M compartment of 1714 Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1991 American Association for Cancer Research. ALPS INHIBIT GLIOBLASTOMA 1.00 0.10 0.03 Dose (uM) Fig. 1. Dose-response relationship analyzed in the MTT assav. of treated GaMg monolayer culture as 3.0Conlrol 2.5 - 0.1 l ^M I 2.0- lOuM 30 uM CELL INVASION 1\ VITRO aggregates consisted of a loosened neuropil. In co-cultures treated with 0.1 and 1.0 ^M ALP, the brain aggregate was intact, lacking only the peripheral fibrous layer. The center of the aggregate was preserved, with no change in morphology from control aggregates (Fig. IB); 10 n\i ALP had serious toxic effects on both the normal and the tumor tissue (Fig. 1C). Necrosis developed after 72 h and increased in magnitude within 168 h. All the primary brain tumor biopsies progressively invaded and destroyed the brain tissue. In the glioblastoma multiforme, extensive destruction of the target tissue was evident within 72 h, with an almost total destruction of brain tissue after 168 h in the control co-cultures (Fig. 8). The lower grade brain tumor spheroids invaded and destroyed the brain tissue to a lesser extent, as reported previously (17) (Fig. 8). The invasion of the glioblastoma multiforme was severelyinhibited by 1.0 ¡IM ALP. The treatment preserved the remain ing brain volume by 31% after 72 h (P < 0.05). After 168 h, such treatment caused a 7-fold preservation of normal brain tissue (P < 0.025) as compared with untreated co-cultures (see Figs. 8 and 9). The reduction in invasiveness was less pronounced for the less invasive, lower grade tumors. In treated co-cultures of the 4.0-1 o 3.5- > § 3.0- Time (Days) Fig. 2. Growth curves of GaMg cells during continuous exposure to ALP. Bars. SEM from triplicate measurements. No reduction in exponential cell growth was observed at ALP doses lower than 10 »JM. •a JS o OÃ- the cell cycle (21.9% as compared with 15.3% for control cultures: P < 0.005) and a concomitant significant reduction of cells in the G, phase of the cell cycle (from 68 to 59%, P < 0.005) (Fig. 4). This effect was evident by the second day of drug exposure and persisted through the 8 days of measurement. Directional Migration. Treatment with ALP at low concen trations (0.1 and 1.0 /UM)did not cause any effect on cell outgrowth from the spheroids (Fig. 5); 10 uM ALP caused a 27% reduction in outgrowth area after 11 days of drug exposure (P < 0.005), whereas 30 ^M drug reduced the colony area to 40% of controls (P < 0.0005). Anti-Invasive Kffcct. In co-cultures between GaMg spheroids and brain aggregates, progressive destruction and invasion of the brain aggregate by the glioma cells occurred. After 168 h of 1.0 MMALP treatment, the volume of the intact brain aggregate was 90% larger than that in control co-cultures. Dose-depend ent preservation of brain aggregate was evident after 72 h, which reached statistical significance (P < 0.005) at 168 h (Fig. 6A). There was no significant difference in glioma spheroid volumes after 72 or 168 h with the concentrations used in these experiments (Fig. 6Ä). In control co-cultures, brain aggregates were severely de stroyed by invading glioma cells within 168 h (Fig. 7.4). The structure of the brain aggregate was disrupted, the outer fibrous layer of glial-like cells was lost, and the remaining core of the 12345678 Time (days) Fig. 3. Growth of GaMg multicellular tumor spheroids treated with ALP. Spheroid growth is recorded relative to the calculated volume at day 0, using the formula for the volume of a sphere. Bars, SEM from 10 replicate spheroids at each treatment dose. 30 uM 1.0 uM G,s c;2M UNA content Control Fig. 4. Cell cycle distribution of GaMg monolayer cultures as analyzed by flow cytometry after 2 days of treatment with ALP. showing an accumulation of cells in the G;M phase, and concomitant reduction of cells in (he G, phase. Analysis at days 4. 6. and 8 rc\ caled identical effects of ALP. 1715 Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1991 American Association for Cancer Research. Al.l'S IMIIBII (,l IODI ASIOMA anaplastic astrocytoma, the brain tissue was 53% (P < 0.05) larger after 168 h (Fig. 9), whereas treatment of the astrocytoma preserved the amount of target tissue by 34% (not significant; see Figs. 8 and 9). CELI. INVASION /.Y IÕTRO V DISCUSSION Malignant gliomas are characterized by a highly invasive behavior in the brain (5, 29). The lack of effective treatment of such tumors implies a poor prognosis for the patients, and new therapeutic strategies are therefore needed. ALP is known to cause cytostatic and cytotoxic effects in both rat and human gliomas (10, 11). The present study confirms these results showing reduced cell growth after ALP treatment. In addition, it is shown that ALP is an anti-invasive drug not only for glioma cell lines but also for biopsies obtained from primary human brain tumors. Treatment with ALP led to a dose-dependent growth inhibi tion of GaMg human glioma cells in monolayer culture. The MTT assay showed the 90% inhibitory concentration to be approximately 50 /¿M, and in the growth curve studies, 30 /IM ALP resulted in 35% growth reduction after 4 days of drug exposure during exponential growth. Prior studies against var ious tumor cell lines have shown that concentrations above 20 /¿M are required for growth inhibition in monolayer culture (11, 30). In addition, short-term treatment (<2 h) with ALP does not exert an antiproliferative effect (10). Tumor spheroid growth curves showed that 10 /¿MALP inhibited growth after 6 days in tissue culture and had a dele terious effect on the spheroid morphology. The increased drug * ItfSWR ' *t«V?WV%ÄVl B 40 -i 2 4 6 Time (days) 8 10 12 Fig. 5. Monolayer outgrowth of GaMg spheroids after addition of ALP. Bars, SEM from 6 replicate spheroids at each treatment dose. Fig. 7. Photomicrographs of co-culture experiments between GaMg spheroids and rat brain aggregates. A, control co-cultures showing extensive invasion and degradation of the brain tissue (B'). B, co-culture treated with 1.0 ^M ALP. The brain aggregate (B') is well preserved. C. co-cultures treated with toxic doses of ALP (10 tiM) B'. remnants of the brain aggregate: bars. 50 ¿im. (I.(K)K -j Fig. 6. Effect of ALP treatment on volumes of rat brain aggregates (A) and GaMg tumor spheroids in co-culture (/?). Volumes represent values calculated from reconstructed image analysis of serial sections. Bars, SEM from 7 replicate co-culture experiments in each group. Time (hours) Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1991 American Association for Cancer Research. ALPS INHIBIT GLIOBLASTOMA CELL INVASION /.V IITRO * B _,>v»" ^^ ' ; • •:•*. •'* Fig. 8. Photomicrographs of co-culture experiments involving the spheroids from primary brain tumor specimens. A. spheroid from glioblastoma multiforme confronted with brain aggregate (B') after 72 h of incubation. B, co-culture between glioblastoma multiforme and brain tissue (B'}. treated with 1.0 JJ.MALP for 72 h. C, control co-culture between a spheroid originating from an anaplastic astrocytoma and a fetal rat brain aggregate (B') after 168 h of incubation, D, co-culture between anaplastic astrocytoma spheroid and brain aggregate (B') after 16X h of incubation, treated with 1.0 /IM ALP (bars. 5 sensitivity of GaMg in spheroid culture compared with monolayer culture may be related to the increased cell-to-cell inter actions brought into play by 3-dimensional culture. Since the mechanism of action of ALP is dependent on integration of the drug in the plasma membrane of the target cells (9), it is reasonable to speculate that monolayer cultures provide a less susceptible target for the drug. This is not a novel finding, as differential sensitivity of tumor cells based on assay methods has been reported previously (31-33). In an in vivo study, treatment by ALP caused an accumulation of the drug in the tumor tissue, as compared with normal tissue (34). If this is the case, in 3-dimensional tissue culture is presently not known. A concentration of 30 n\\ ALP caused an accumulation of cells in the G2M phase of the cell cycle after 2 days' exposure 0.11(17 : Fig. 9. Measured volumes from serially sectioned co-cultures of a glioblastoma multiforme, an anaplas tic astrocytoma. and an astrocytoma. Bars, SEM from 5 replicate co-cultures in each group. Ghoblastoma multiforme 72 hours I68hours 1717 Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1991 American Association for Cancer Research. ALPS INHIBIT (¡I.IOBIASTOMA CELL INVASION l.\ \ÕTRO in monolayer culture. Cytotoxic effects have been associated with increased lipid fluidity (35), the formation of blebs and holes in the plasma membrane (8), and modulation of the protein kinase C system (36, 37). ALP does not seem to inhibit microtubule assembly or reassembly in MO4 mouse fibrosar coma cells (12). An induction of multinucleated cells by ALP treatment in sensitive cell lines has been found previously (38). The accumulation of cells in G:M phase of the cell cycle may be due to an interference of ALP with the plasma membrane with a retarded cell division as the result. The most profound and novel observation from our studies is the effect of nontoxic concentrations of ALP on the invasive behavior of glioma cells into an in vitro model of brain tissue. Since complete surgical resection of astrocytomas and glioblastomas is not feasible, and since these tumors are highly refrac tory to chemotherapy and radiation, agents that block other phenotypic traits of these malignant cells may have exciting potential as new treatment modalities against these neoplasms. The invasion of tumor spheroids into fetal rat brain aggre gates was inhibited after 72 h of co-culture, but the magnitude of inhibition did not reach statistical significance until 168 h. This may be due to a slow integration of drug molecules into the plasma membrane (9), which would lead to a delayed action of ALP. The retarded anti-invasive effect may also arise from a technical limitation of the assay, in which more extensive invasion by GaMg control co-cultures is necessary for differ ences between control and treated dishes to become significant. Other authors have described a total prevention of tumor cell invasion into chick heart fragments after incubation with ALP (12, 13). The discrepancy between the demonstration of early and complete inhibition of invasion in their studies and the delayed efficacy in our report may be attributable to differences in the 2 model systems used. The mechanisms of glioma cell invasion into fetal rat brain aggregates, which represents a tissue of neuroectodermal origin, may be expected to be different from those required for the invasion into precultured chick heart fragments. Furthermore, by virtue of the natural invasiveness of glioma cells into brain and the extraordinarily low incidence of glioma metastasis extracranially (29), brain aggregate may be a more relevant model for the investigation of processes operative in the clinical setting of gliomas. Invasion may be due to specific interactions between cells and the interstitial matrix of the invaded tissue; the composition of the matrix would be markedly different in chick heart versus rat brain. Finally, different tumors may show various responses to anti-invasive therapy (13). Additional impetus to the potential utility of ALP as an antiinvasive therapy against human gliomas is gained from our results using primary glioma spheroid cultures in the invasion assay. The profound inhibition of the glioblastoma cell invasion at a concentration of 1.0 MMsuggests that ALPs induce pertur bations of the cell membrane, thus interfering with the invasive phenotype of glial cell tumors. Because of the relationship between the invasiveness of the glial tumor, and the ability of ALP to inhibit the invasion for the tumors studied, we suggest that this property is directly interfered with by ALP. The biochemical origin of this anti-invasive effect of ALP should be further exploited for the development of therapy and to better understand glioma progression. ALP at a concentration of 1.0 MMdid not inhibit cell migra tion (spreading of spheroids onto monolayer), but did inhibit invasion after 168 h of co-culture. It is likely, therefore, that ALP exerts direct anti-invasive effects on glioma cells inde pendent from influences on motility (migration) and prolifera tion. The mechanism(s) responsible for the anti-invasive effect are not known, but alterations in the cell surface carbohydrates (4, 39, 40) could play a major role, as carbohydrates are found in relatively high concentrations on the cell surface in the form of glycopeptides, glycosylated proteins (cell recognition mole cules), and glycolipids (gangliosides). All these molecules are thought to be important for cellular interactions and recogni tion (41, 42). It is known that malignant cells have a carbohydrate com position different from that of nonmalignant cells, and that ALP treatment changes the carbohydrate moieties on the cell membrane (4, 43). If a change in carbohydrate composition is associated with invasiveness, this may explain the transition of noninvasive, transformed cells to invasive cells by ALP treat ment, as have been documented in a previous study (4). Changes in carbohydrate composition in embryonal cells treated by ALP may also explain the limited therapeutic index seen in our study, in which 10 MMof this drug destroyed the fetal rat brain aggregates. In contrast to this, ALP has been shown not to interact with cellular DNA and has no mutagenic effects (44). In addition, limited toxic effects have been observed from ALP during Phase I clinical trials (9, 45). ALP has been shown to penetrate the blood-brain barrier more readily than other cytostatics (46). Furthermore, there is evidence for an enrichment of this compound in tumors in vivo, as compared with normal tissue (34). This makes the drug interesting as a potential agent in brain tumor therapy. The relatively large inhibition of invasion achieved by ALP at nontoxic doses suggests a potential use for ALP, or its derivatives in the treatment of malignant gliomas. ACKNOWLEDGMENTS The authors are greatly indebted to Dr. O. D. Laerum for valuable advice and criticism. REFERENCES 1. Libermann. T. A.. Nusbaum. H. R., Razón. N.. Kris. R.. Lax. !.. Whittle. N., Waterfield. M. D.. Ullrich. A., and Schlessinger, J. Amplification, enhanced expression and possible rearrangement of EG F receptor gene in primary 1985. " human brain tumors of glial origin. Nature (Lond.). 313: 144-147. 2. Bigner. S. H.. Burger. P. C.. Wong. A. J.. Werner. M. H.. Hamilton. S. R.. Muhlbayer. L. H.. Vogelstein. B., and Bigner. D. D. Gene amplification in malignant human gliomas: clinical and histopathologic aspects. J. Neuropathol. Exp. Neurol.. 47: 191-205. 1988. 3. Nister. M., Libermann. T. A.. Betsholtz. C.. Pctterson. M.. Clacsson-Welsh. L., Schlessinger. J.. and \Vestermark. B. Expression of messenger RNAs for platelet derived growth factor and transforming growth factor <>and their receptors in human malignant glioma cell lines. Cancer Res.. 4Ãœ: 3910-3918. 1988. 4. Bolschcr. J. G. M.. Schallier. D. C. C.. Smets. L. A., Rooy. H. V.. Collard. J. G.. Bruyneel. E. A., and Marcel. M. M. K. Effect of cancer-related and drug-induced alterations in surface carbohydrates on the invasive capacity of mouse and rat cells. Cancer Res.. 46:4080-4086. 1986. 5. Halperin. C., and Int. Burger. P. C. Oncol. The fallacy the locali/cd supratentorial malignant E. gliomas. J. Radial. Bio'l.ofBiophys.. IS: 505-509. 1988. 6. Spector. A. A., and Burns. C. P. Biological and therapeutic potential of membrane lipid modification in tumors. Cancer Res.. 47: 4529-4537. 1987. 7. Ishaq. K. S.. Capobianco. M., Piantadosi. C'.. Noseda. A.. Daniel. L. W., and Modest. E. J. Synthesis and biological evaluation of ether linked derivatives of phosphatidylinositol. Pharm. Res.. 6: 216-224. 1989. 8. Noseda, A., White. J. G., Godwin. P. L.. Jerome. W. G., and Modest, E. J. Membrane damage in leukemic cells induced by ether and ester lipids: an electron microscopic study. Exp. Mol. Pathol.. 50: 69-83. 1989. 9. Berdel. W. E.. Andreesen. R.. and Munder. P. G. Synthetic alkyl-phospholipid analogs: a new class of antitumor agents. In: J. F. Kuo (ed.). Phospholipids and Cellular Regulation. Vol. 2. pp. 42-73. Boca Raton. Ft.: CRC Press. 1985. 10. Berdel. W. E.. Greiner. E.. Fink, l .. Stavrou. D.. Reichert. A.. Rastetter. J.. 1718 Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1991 American Association for Cancer Research. ALPS INHIBIT GLIOBI.ASTOMA CELL INVASION l.\ UTRO Hoffman. D. R.. and Snyder. F. Cytotoxicity of alkyl-lysophospholipid derivatives and low-alkyl-cleavage enzyme activities in rat brain tumor cells. Cancer Res.. 43: 541-545. 1983. Berdel. \V. E.. Greiner. E.. Fink, li., /anker. K. S., Stavrou. D.. Trappe. A.. Fahlbusch. R.. Reichert. A., and Rastetter. J. Cytotoxic effects of alkylIvsophospholipids in human brain tumor cells. Oncology. 41: 140-145, 1984. Storme. G. A.. Berdel. \V. E.. Slitters»ijk. W. J. V.. Bruyneel. E. A.. De Bruyne. G.. and Marcel. M. M. K. Anti-invasive effect of racemic l-ooctadecyl-2-o-niethylglycero-3-phosphocholine on MO4 mouse fibrosarcoma cells in vitro. Cancer Res.. 45: 351-357. 1985. Storme. G. A.. Bruyneel. E. A.. Schallier. D. C. C.. Bolscher. J. G. M.. Berdel. \V. E., and Marcel. M. M. K. Effect of lipid derivatives on invasion //; vitro and on surface glycoproteins of three rodent cell types. Lipids. 22: 847-850. 1987. Akslen. L. A., Andersen. K-J.. and Bjerkvig. R. Characteristics of human and rat glioma cells grown in a defined medium. Anticancer Res., ÕÕ: 797804. 1988. Bjerkv¡g.R., Tonnesen. A.. Laerum. O. D.. and Backlund. E. O. Multicellular tumor spheroids from human gliomas maintained in organ culture. J. Neurosurg.. 72: 463-475. 1990. Bjerkvig. R.. Steinsvag. S. K., and Laerum. O. D. Reaggregation of fetal rat brain cells in a22: stationary I: methodology and cell identifica tion. In Vitro. 180-192.culture 1986.system ' in stereology and its prediction. J. Microscopy. 147: 229-263. 1987. 28. Gundersen. H. J. G.. Bendtsen. T. f.. Korbo. L.. Marcussen, N.. M0ller. A., Nielsen, K.. Nyengaard. J. R.. Pakkenberg. B.. Sorensen. F. B.. Vesterby. A., and West. M. J. Some new. simple and efficient stereological methods and their use in pathological research and diagnosis. APMIS. 96: 379-394. 1988. 29. Sherbet. G. V. The metastatic spread of cancer: reflections upon its mecha nisms based on the malignant glioma model. London. FL: Macmillan Press. 1987. 30. Runge. M. H„Andreesen. R.. Pfleiderer, A., and Munder. P. G. Destruction of human solid tumors by alkyl lysophospholipids. J. Nati. Cancer Inst., 64: 1301-1306. 1980. 31. Berens. M. E., Giblin, J. R.. Dougherty. D. V.. Hoifodt. H. K.. Tveit. K.. and Rosenblum. M. L. Comparison of in vitro cloning assays for drug sensitivity of human brain tumors. Br. J. Neurosurg.. 2: 227-234, 1988. 32. Yohen. K. H.. Slyman. D. J.. Bregman. M. D.. and Meyskensjr, F. L. Radiation survival of murine and human melanoma cells utilizing two assay systems: monolayer and soft agar. Br. J. Cancer.. 57: 64-69. 1988. 33. West. C. M. L., and Sutherland, R. M. A radiobiological comparison of human tumor soft agar clonogenic assays. Int. J. Cancer. 37:897-903. 1986. 34. Arnold. B.. Reuther. R.. and Weltzien, H. U. Distribution and metabolism of synthetic alkyl analogs of lysophosphatidylcholine in mice. Biochim. Biophys. Acta. 530: 47-55. 1978. 35. N'oseda. A.. Godwin. P. L.. and Modest. E. J. Effects of antineoplastic ether 17. Bjerkvig, R.. Laerum. O. D.. and Mella. O. Glioma cell interactions with fetal rat brain aggregates in vitra, and with brain tissue in viro. Cancer Res.. 46: 4071-4079. 1986. 18. Engebraatcn. ().. Bjerkvig. R.. l.und-Johansen. M.. Pedersen. P-H.. Wester. K.. Backlund. E-O.. and Laerum. O. D. Interaction between fetal rat brain tissue and brain tumor biopsies in vilro. Acta Neuropathol.. XI: 130-140. 1990. 19. Yuhas. J. M.. Li. A. P.. Martinez. A. O.. and Ladman. A. J. A simplified method for production and growth of multicellular spheroids. Cancer Res.. 57:3639-3643.1977. 20. Mosmann. T. Rapid colorimetrie assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J. Immunol. Methods. 65:55-63. 1983. 21. Carmichael. J.. Mitchell, J. B.. DcGraf. W. G.. Gamson, J.. Gazdar, A. F.. Johnson. B. li.. Glatstein. E.. and Minna, J. D. Chemosensitivity testing of human lung cancer cell lines using the MTT assay. Br. J. Cancer. 57: 540547. 1985. 22. Nederman. T. Effects of vinblastinc and 5-fluorouracil on human glioma and thyroid cancer cell monolayers and spheroids. Cancer Res.. 44: 254-258. 1984. 23. Barlogie. B.. /ante, J.. Drewinco. B.. Schumann. J.. Buchner. T., Gohde. W., Hart, J. S.. and Johnson. D. A. New preparation and staining procedures for specific and rapid analysis of DNA distributions. In: W. Gohde. J. Schumann, and T. Buchner (eds.). Pulse Cytophotometry II. pp. 125-136. Ghent, FL: European Press. 1976. 24. Baish. H.. Gohde. W.. and Linden. \V. Analysis of PCP-data to determine the fraction of cells in the various phases of the cell cycle. Rad. Environ. Biophys.. 12: 31-39. 1975. 25. Trapp. B. D.. llonegger. P.. and Richelson, E. Morphologic differentiation of mechanically dissociated fetal rat brain aggregating cell cultures. Brain Res.. 160: 235-252. 1981. 26. Druckrey. H. Genotypes and phenotypes of ten inbred strains of BD rats. Arzneimittelforschung. 21: 1274-1278. 1971. 27. Gundersen. H. J. G.. and Jensen. E. B. The efficiency of systematic sampling lipids on model and biological membranes. Biochim. Biophys. Acta. 945: 92-100. 1988. 36. Hclfman. D. M.. Barnes, K. C.. Kinkade. J. M., Vogler. W.. Shoji. R.. ;..id Kuo. J. F. Phospholipid-sensitive Ca-dependent protein phosphorylation system in various types of leukemic cells from human patients and in human Icukemic cell lines HL60 and K562. and its inhibition by alkyl-lysophospho lipid. Cancer Res.. 43: 2955-2961. 1983. 37. Nishizuka. V. Studies and perspectives of protein kinase C. Science (Wash ington DC). 233: 305-312. 1986. 38. Roos. G.. and Berdel. \V. E. Sensitivity of human hematopoietic cell lines to an alkyl-lysophospholipid-derivativc. Leuk. Res.. 10: 195-202. 1986. 39. Steck. P. A.. Moser. R. P.. Bruner, J. M.. Liang. L.. Friedman. A. N.. Hwang. T-L., and Yung. W. K. A. Altered expression and distribution of heparin sulfate proteoglycans in human gliomas. Cancer Res.. 49: 2096-2103. 1989. 40. Bolscher. J. G. M.. Bruyneel. E. A., vanRoy, H.. Schallier. D. C. C.. Marcel. M. M. K.. and Smets. L. A. Decreased fucose incorporation in cell surface carbohydrates is associated with inhibition of invasion. Clin. Exp. Metastasis. 7:557-569. 1989. 41. Niedieck. B.. and Lohler. J. Expression of A-fucosyl-.V-acetyllactosamine on glia cells and its putative role in cell adhesion. Acta Neuropathol.. 75: 173184. 1987. 42. lozzo, R. V. Cell surface heparin sulfate proteoglycan and the neoplastic phenotype. J. Cell. Biochem.. 37: 61-78. 1988. 43. Dennis. J. W.. and Laferte. S. Tumor cell surface carbohydrate and the metastatic phenotype. Cancer Metastasis Rev.. 5: 185-204, 1987. 44. King. M. T.. Eckhardt. K.. Gocke. E.. Wild. D.. Berdel. W. E., and Munder. P. G. Failure to detect mutagenic effects of antitumor alkyl-lysophospholipids. Cancer Lett.. 12: 217. 1981. 45. Berdel. W. E., Bausert. W. R. E.. Fink. U.. Rastetter. J.. and Munder. P. G. Anti-tumor action of alkyl-lysophospholipids. Anticancer Res., I: 345-352. 1981. 46. Unger. C.. Eibl. H.. von Heyden. H-W., Krisch. B., and Nagel. G. A. Bluthirnschranke und penetration von zytostatika. Klin. Wochenschr.. 63: 565571. 1985. 11. 12. 13. 14. 15. 16. 1719 Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1991 American Association for Cancer Research. Effect of Alkyl-lysophospholipid on Glioblastoma Cell Invasion into Fetal Rat Brain Tissue in Vitro Olav Engebraaten, Rolf Bjerkvig and Michael E. Berens Cancer Res 1991;51:1713-1719. Updated version E-mail alerts Reprints and Subscriptions Permissions Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/51/6/1713 Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected]. Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1991 American Association for Cancer Research.