Stereotactic biopsies as a model for studying the interaction between gliomas and normal brain tissue in vitro

In order to study tumor invasiveness, the authors used a model system which allows the direct interaction between glioma cells and normal brain tissue to be studied. The spiral needle biopsy technique provides a tissue core with a completely preserved architecture from brain tumors. These biopsies, once cultured, produce three-dimensional multicellular tumor spheroids; they display morphological features similar to those of the original tumor in vivo. As a target fetal rat brain cell aggregates were used, in order to study the invasive process in the microscope. Not surprisingly, different tumors show different patterns of invasion; this variability is not yet clearly correlated with the clinical course of the patient.

Characteristics of human and rat glioma cells grown in a defined medium

Human (D54Mg, GaMg) and rat (BT5C, BT4Cn) glioma cells cultured in a chemically defined medium showed reduced growth when compared to serum-supplemented medium. The BT5C cells changed from a flat epithelioid morphology to a more glia-like structure. The serum-free medium caused an aggregation of BT5C cells which spontaneously lost anchorage dependence and continued to grow in suspension as multicellular tumor spheroids. The BT4Cn and human cell lines did not show any change in morphology. Flow cytometric DNA measurements showed no change in ploidy for cells grown in serum-free medium. Cell cycle analysis revealed that the same proportion of cells were proliferating (S and G2M phase cells) in serum-free medium as compared to serum-supplemented medium. The reduced growth is probably due to increased cell cycle time.

Whipple's disease. Demonstration of a persisting monocyte and macrophage dysfunction

A patient with Whipple's disease has been followed up for 4 years. Primary involvement was limited to the small intestines, and accumulation of periodic acid-Schiff-positive material, containing typical more or less intact bacillary bodies, was demonstrated within macrophages of affected tissue. After initial oxytetracycline treatment and clinical remission, the patient relapsed, with multiorgan affections. The antibiotic regimen was changed to chloramphenicol, followed by continuous trimethoprim-sulfamethoxazole. Flow cytometric studies showed persisting impairment of monocyte and macrophage intracellular degradation of bacteria during all the 4 years tested. After relapse, reduced activity of several brush border enzymes was demonstrated in distal duodenal biopsy specimens. After 17 months of continuous trimethoprim-sulfamethoxazole therapy complete clinical remission, regression of histopathologic abnormalities, and restoration of duodenal enzyme activities had occurred. The results demonstrate a persisting dysfunction of mononuclear phagocytes from a patient with Whipple's disease, suggesting a primary abnormality of cell-mediated immunity which may promote the susceptibility to the causative bacillus.

DNA content and chromosomes in permanent cultured cell lines derived from malignant human gliomas

DNA content, as determined by flow cytometry (FCM), and ploidy, as determined planimetrically from chromosomes, were measured for 17 established human glioma-derived cell lines. Values obtained by these methods corresponded well for two of two near-diploid lines, two of two hyperdiploid-hypotriploid lines, four of four hypertriploid-hypotetraploid lines, two of six hypertetraploid-hypopentaploid lines, the one hyperpentaploid line and two of two multiclonal lines. For the remaining four lines, ploidy values obtained by karyotyping were more than 10% lower than those obtained by FCM. Since karyotypic ploidy was determined by planimetric measurements of the chromosomes, which corrects for deviations in chromosome size, the lower values could not be explained by large marker chromosomes. Technical problems, such as random chromosomal loss in karyotypic preparations, adherent bits of cytoplasm in nuclear preparations for FCM or differences in nuclear stainability, are possible reasons for the discrepancies. Alternatively, chromosomes in some glioma cell lines may actually contain an increased amount of DNA. FCM and karyotyping provide complementary information in the initial evaluation of cultured cell lines. For sequential studies, such as are used for monitoring cultured lines, the rapidity of FCM makes it the more practical method.

Biological and biochemical characterization of cell lines derived from initiation-promotion transformed C3H/10T1/2 cells

A two-stage transformation protocol was used to chemically transform the mouse embryo fibroblasts, C3H/10T1/2 Cl 8. To initiate the cells 0.37 microM 20-methylcholanthrene was used and 0.17 microM of the tumor promoter 12-O-tetradecanoyl-phorbol-13-acetate was employed to complete the transformation process. Six weeks later transformed foci were identified and isolated by the ring-cloning technique. Altogether eight different foci were trypsinized resulting in a total of 12 morphologically transformed subclones. Three of these clones, designated TPA 41, TPA 42 and TPA 482, have been characterized in detail. Their growth morphologies were different. The TPA 482 cells grew in a criss-cross pattern with piled up foci, thus showing a characteristic type III morphology. The TPA 482 clone did not show cell-density growth inhibition and grew in soft agar. The TPA 41 and TPA 42 clones exhibited cell-density growth inhibition, grew as monolayers and formed only few colonies in soft agar. Late passages of the TPA 42 clone acquired growth characteristics similar to TPA 482. The C3H/10T1/2 Cl 8 and the TPA 41 cells were not tumorigenic when transplanted into syngeneic mice. TPA 482 cells were strongly tumorigenic, producing tumors in 6/6 mice in 21 days. The TPA 42 cells were also tumorigenic, the first tumors appearing after 4 weeks; all animals injected with TPA 42 cells had tumors after 8 weeks. All tumors observed appeared to be fibrosarcomas. Flow cytometric analysis indicated differences in DNA distributions between tumor cells grown in vitro and the tumors in vivo. Two-dimensional gel analysis of the total cellular and the nuclear proteins showed an increase in the TPA 42 and TPA 482 cells of an acidic 48,000 and a basic 83,000 mol. wt polypeptides, and a decrease of a neutral polypeptide of mol. wt 46,000, located in the nucleus of TPA 482 cells.

Phagocytic capacity of normal and malignant rat glial cells in culture

The phagocytic capacity of 4 continuous rat glioma cell lines (BT2C, BT4Cn, BT5c, and 9L) and normal BD IX fetal rat glial cells in culture has been studied. This was done by flow cytometric measurements of single cells from monolayer cultures having ingested fluorescent bacteria, zymosan particles, red blood cells, or fragments of normal glial cells. In addition, phagocytosis was studied in a three-dimensional culture system. The BT4Cn, BT5C, and 9L cell lines were tumorigenic and invasive both in vivo and in organ culture in vitro. In contrast, BT2C has shown variable tumorigenicity and does not seem to be invasive. The phagocytic capacity of the cell lines was compared to their destructive properties during invasion. Depending on the particle type, 30-40% of the normal glial cells were phagocytic. The fractions of phagocytic glioma cells were dependent on the particle type and the prey load. Of the invasive cell lines, BT5C showed high phagocytic activity both in monolayer and three-dimensional cultures. Two of the invasive cell lines (BT5C and 9L) had about the same fraction of phagocytic cells as normal glial cells. These 2 cell lines showed highly destructive growth during invasion. In contrast, the third invasive cell line (BT4Cn) had almost no phagocytic cells. The BT4Cn cells showed single-cell invasion with little destruction of target tissue. The noninvasive cell line (BT2C) showed low phagocytic activity, and almost no destruction was observed in the border zone between tumor cells and normal tissue. Phagocytosis seems to be an inherent property of both normal and malignant glial cells, although the fraction of phagocytic cells varies from one cell line to another. In organ culture high phagocytic capacity of invasive glioma cells seems to be related to destructive activity on the normal brain tissue during invasion.

Transmission and scanning electron microscopy of N2 microbubble-activated human platelets in vitro

When N2 microbubbles are stirred in platelet-rich plasma, they cause a fall in the number of free platelets. Changes in the platelets ultrastructure during this interaction between gas bubbles and platelets have been studied by transmission and scanning electron microscopy. Platelets, and aggregates of platelets, adhere to the surface of the N2 microbubbles. This adhesion induces ultrastructural changes (shape change, pseudopod formation, granule centralization, fusion, and disappearance) that are similar to "classical" agonists like ADP, collagen, and thrombin. These ultrastructural studies further strengthen our previous contention that N2 microbubbles activate platelets in a way similar to these physiologic agonists, and show that the previously reported fall in the number of free platelets is due to platelet aggregation. Platelet aggregates are also present in the interstices between gas bubbles. Fixed N2 microbubbles have been demonstrated, and these can be broken like cracked eggs by means of the electron beam in the electron microscope. Possible mechanisms for activation of the platelets by the gas bubbles are: through ADP released from some few platelets; by diffusion of gas after bubble-platelet interaction; and through certain fractions of the plasma proteins and lipids in the bubble surface that may act as binding sites for the platelets and promote adhesion and spreading of the platelets over the surface.

Glioma cell interactions with fetal rat brain aggregates in vitro and with brain tissue in vivo

Two malignant rat neurogenic cell lines have been confronted with reaggregation cultures of fetal rat brain in a stationary culture system. In this organ culture system several morphological aspects of the developing brain are imitated. The malignant cell lines showed invasive and progressive growth into the brain tissue, ending with a completely destroyed aggregate within 10 days of coculture. One cell line (BT5C) showed solid invasion with groups of cells replacing different layers of the brain tissue. A considerable degree of normal cell lysis was seen, both at the edge of invasion and at distant parts of the aggregate. This cell line seemed to activate macrophages in the aggregates. In addition, conditioned medium from these cultures caused degeneration of the brain aggregates. The other cell line (BT4Cn) showed invasion by single cells where the replacement and destruction of the tissue was not accompanied by distant lysis. Both cell lines showed the same pattern of invasion in vitro as was seen in the brains of isogeneic animals in vivo, showing that the individual tumor characteristics were also reflected during invasive growth in organ culture.

Reaggregation of fetal rat brain cells in a stationary culture system. I: Methodology and cell identification

A stationary tissue culture system for reaggregation cultures of rat brain cells is described. Aggregates were formed by placing cells at high concentration in liquid overlay cultures on a nonadherent nutrient agar surface. No physical stress in the form of rotation or shaking was applied to the aggregating cell population. Transmission electron microscopy and immunohistochemistry showed that the cells developed from homogeneously dispersed, immature cells in Day 4 aggregates, to mature astrocytes, oligodendrocytes, and neurons in Day 20 aggregates. Twenty days older aggregates had a tightly packed neuropil which was most prominent in a cell-sparse outer layer of the aggregates. When the aggregates were allowed to adhere to a substrate, both glial fibrillary acidic protein (GFAP) positive and negative cells were observed migrating out from the aggregates. Cells giving a positive reaction for neuron specific enolase (NSE) were also present. This reaggregation procedure, with transfer of selected brain cell aggregates into agar-coated multiwells is an alternative three-dimensional culture system which can be potentially useful in the study of morphogenesis and cell interactions in the nervous system.

Reaggregation of fetal rat brain cells in a stationary culture system. II: Ultrastructural characterization

Ultrastructural characteristics of fetal rat brain cell aggregates in a three-dimensional stationary culture system are described. Transmission electron microscopy showed immature cells which developed into mature astrocytes, oligodendrocytes, and neurons during 20 d in culture. This was accompanied by the development of a neuropil where myelinated axons and synaptic complexes were observed. In addition to confirming earlier ultrastructural investigations on fetal rat brain cell aggregates, the stationary culture system also showed the presence of histiotypic regions within the aggregates. These regions consisted of ependymal cells where cilia were observed on the cell surfaces. Structures resembling subependymal basement membrane labyrinths were also observed. Macrophages seemed to be more numerous in the stationary cultures as compared to other culture systems. The stationary culture system may provide aggregates that are ultrastructurally more complex than those obtained by rotation mediated systems.

Growth arrest and polyploidization induced by metahalone microtubule inhibitors on rat glioma cells in culture

Two pyrimidine analogs (metahalones) NY 3163, NY 3170 have been tested for their effects on the growth of rat glioma cells in monolayer and in spheroid culture. Both substances have earlier been found to inhibit the microtubule system in malignant cells. In glioma cells, arrest of mitosis was accompanied by repeated cycles of DNA synthesis, leading to different levels of polyploidization up to 16 and 32 ploid cells. The effect was not reversible through a culture period of 4 days. Reduced ability of directional migration of cells on a plastic surface was seen for 16 days after exposure. The reaction to microtubule inhibitors seems to differ depending on the type of tumour cells, where some malignant cells have earlier been reported to escape mitotic arrest and proceed to the next G1 phase, in contrast to the present glioma cells which undergo polyploidization.

Cell and tissue culture of the central nervous system: recent developments and current applications

A survey of methods for cell and tissue culture of the central nervous system (CNS) is given. This includes a brief historical outline and description of methods in current use. Recent methodological improvements are emphasized, and it is shown how these are applied in modern neurobiological research. Both monolayer cell cultures and three-dimensional organ culture systems are widely used, each having advantages and limitations. In recent years, there has been considerable improvement of culture for prolonged periods in chemically defined media. Brain tissue from a wide spectrum of species have been used, including different types of human brain cells which can be propagated for several months. At present, these culture systems are employed for dynamic studies of the developing, the adult and ageing brain. It is possible to select neurons and the different classes of glial cells for culture purposes. Cell culture of the CNS has given new insights into the biology of brain tumours. Culture systems for experimental tumour therapy in vitro are also available. Recently, it has been shown that organ cultures of brain tissue can be used as targets for invasive glioma cells, enabling a direct study of the interactions between tumour cells and normal tissue to take place.

DNA content and chromosomal composition of malignant human gliomas

A short review is given on DNA aberrations and chromosomal composition of malignant human gliomas. By flow cytometric DNA analysis, a wide range of different ploidies has been reported in biopsied gliomas, from diploid to strongly aneuploid nuclear DNA. However, with the preparation and analysis methods used so far, no clear relationship between the type of ploidy and histology or prognosis has been established. A high proportion of glioblastomas is near-diploid, indicating a high degree of biologic malignancy is not necessarily connected to aberration of the nuclear DNA content. It is possible that improved methods giving a higher degree of resolution will allow separation of the near-diploid populations of malignant human gliomas from normal diploid cells and permit the detection of subpopulations with small differences from the dominant DNA mode. Chromosomal studies of malignant gliomas have confirmed that the majority of them have near-diploid stemlines. These populations are seldom normal diploid, however, as both numerical and structural abnormalities are usually present. In addition, chromosomal analyses have shown that when gliomas are bimodal, the polyploid populations are usually doubled versions of the near-diploid ones. In contrast to the near-diploid populations that characterize biopsied malignant gliomas, both FCM studies and karyotyping have demonstrated that permanent cultured cell lines derived from malignant gliomas are usually near-triploid or near-tetraploid. Sequential karyotypic studies of these tumors from biopsy through establishment in vitro have shown an evolutionary pattern consisting of doubling of the original stemline, followed by gains or losses of individual chromosomes with new marker formation in late culture. Evaluation of biopsied malignant gliomas by karyotyping has also demonstrated that subgroups of them are characterized by specific numerical and structural deviations. These groupings may prove useful in predicting prognosis or responsiveness to specific therapeutic regimens. The specific chromosomal abnormalities observed in malignant human glioma may provide clues as to the genes important in glial transformation. As chromosomal loci for production of structural proteins and enzymes involved in glial metabolism are mapped and patterns of oncogene activation and amplification are determined for human gliomas, the meaning of the nonrandom chromosomal changes seen in these tumors may become clear.

Interaction between rat glioma cells and normal rat brain tissue in organ culture

A system for coculture between normal rat brain fragments and multicellular spheroids of rat glioma cells is described. The tumor cells were derived from fetal BD IX rats treated transplacentally with the carcinogen N-ethyl-N-nitrosourea (CAS: 759-73-9). Brain fragments were obtained from fetal BD IX rats and precultured for 20 days before confrontation with multicellular tumor spheroids. The cocultures were grown in nonadherent stationary organ culture for 30 days. Due to morphologic similarities between normal brain cells and tumor cells, the tumor cells were labeled with tritiated thymidine, which made them easily recognizable in autoradiographs. The two structures adhered to each other, and glioma cells progressively invaded and replaced the normal brain tissue. Invasion and replacement are characteristic features of brain tumors in vivo. Therefore, this organotypic and syngeneic model may be useful for investigation of these phenomena outside the body.

Invasiveness of primary brain tumors

Primary malignant neoplasms of the nervous system differ from other types of malignancy in several ways. Clinical progression is due to local invasive growth, while metastases outside the skull are rare. The tumors show no sharp delimitation from the surrounding normal tissue. At the edge, an ill-defined area of invasive tumor cells, reacting glial cells and inflammatory cells is present. At the same time the primary brain tumors are biologically heterogeneous. In this review, a short survey of markers for malignancy in primary brain tumors is given, and some properties of importance for invasive behavior, are listed. These include different cellular enzymes, phagocytotic property, locomotive and proliferative characteristics. Studies of primary brain tumors in situ show invasive growth into the surrounding brain tissue, often followed by hemorrhage and necrosis. In addition spread of tumor cells takes place along preexisting intracranial structures. Recently, several systems for the study of brain tumor invasiveness in culture have been elaborated. Both experimental and human gliomas have been tested. The target tissues include organ culture of embryonic chick heart muscle, chorioallantoic membrane, fetal rat brain tissue and reconstructed vessel walls. It has been shown that glioma cells are able to split junctions between normal cells. They destroy and phagocytose the normal cells and penetrate the normal tissue. The use of brain tissue and reaggregated brain cell cultures as target for glioma cells in culture opens the possibility for an elucidation of invasiveness as one of the most important properties of malignancy in the nervous system.

The relative resistance of non-cycling cells in 9L multicellular spheroids to spirohydantoin mustard

Surviving fractions of 9L spheroid cells treated with 1.5, 3.0 and 6.0 micrograms/ml of spirohydantoin mustard (SHM) decreased when assayed 6 hr after treatment but increased thereafter. Flow cytometric analysis showed that exponentially growing 9L monolayer cells treated with SHM accumulated at the G2/M border within 24 hr. Cells dissociated from spheroids treated with 3 and 6.0 micrograms/ml of SHM accumulated at the G2/M border during the first 24 hr after treatment and remained there for the next 12 hr. However, 50% of the cells remained at the 2C DNA peak. Spheroid cells with 2C DNA content 24 hr after treatment were assumed to be non-cycling cells at the time of treatment, and cells that accumulated at the G2/M peak appeared to be cycling cells at the time of treatment; approximately 50% of cells in untreated 9L spheroids are in the noncycling pool. G1 and/or early S phase 9L cells in exponential growth elutriated immediately after treatment with SHM had significantly lower (P greater than 0.001) plating efficiencies than 9L cells in S and G2/M phases. When spheroids were dissociated, elutriated and plated for colony-forming efficiency 24 hr after treatment with 3 micrograms/ml of SHM, fractions enriched in 2C DNA content had significantly higher (P greater than 0.001) plating efficiencies than elutriated cells enriched in 4C DNA. These results indicate that SHM is less effective against non-cycling 9L spheroid cells with 2C DNA content than against cycling 9L spheroids cells.

Cell cycle age response of 9L cells to 1,3-bis(2-chloroethyl)-1-nitrosourea and modification by alpha-difluoromethylornithine

We have determined the cell cycle age response of 9L rat brain tumor cells to 1,3-bis(2-chloroethyl)-1-nitrosourea using centrifugal elutriation to obtain populations of cells enriched in G1, S, and G2-M phases. While cells in all phases of the cell cycle were killed by 20 or 40 microM 1,3-bis(2-chloroethyl)-1-nitrosourea, cells in G1 and G2-M phases were more sensitive than cells in S phase. The differential sensitivity was more pronounced at the higher dose, which will markedly alter the distribution of cells through the cell cycle. In a clinical setting, this factor could affect the efficacy of either fractionated or multimodality protocols. Treatment with alpha-difluoromethylornithine, a polyamine biosynthesis inhibitor, potentiated the cytotoxic effects of 20 microM 1,3-bis(2-chloroethyl)-1-nitrosourea against G1- and G2-M- but not against S-phase cells; however, at a higher dose of 1,3-bis(2-chloroethyl)-1-nitrosourea (40 microM), the cytotoxicity was potentiated for cells in all phases of the cell cycle. In alpha-difluoromethylornithine-treated cells, the phenomenon could be reversed by adding 1 mM putrescine 24 hr before treatment with 1,3-bis(2-chloroethyl)-1-nitrosourea. Therefore, the potentiation of 1,3-bis(2-chloroethyl)-1-nitrosourea cytotoxicity appears to be related to polyamine depletion.