Basement membrane invasion of glioma cells mediated by integrin receptors

Role of extracellular matrix in tumor invasion: migration of glioma cells along fibronectin-positive mesenchymal cell processes

OBJECTIVE

The major morbidity of glioma lies in its infiltrative growth. One of the major patterns of this invasive growth is the formation of Scherer's secondary structures associated with the blood vessels and the leptomeninges. To better understand the role of extracellular matrix (ECM) in glioma invasion, we investigated in vitro the interaction between glioma cells and the meningeal mesenchymal tissue from the brain. As an aid to this study, ECM in glioma cell line spheroids was compared with that in primary fetal brain aggregates.

METHODS

To study the expression of ECM, four glioma cell lines (U-87 MG, U-251 MG, AN1/lac-z, and HF-66) and primary cells from fetal rat brain were grown as spheroids and monolayers. To sudy the role of ECM in glioma invasion, spheroids from the glioma cell lines were grown over established cultures of fetal meningeal and mesenchymal tissue. Expression of fibronectin, laminin, tenascin, collagen VI, and chondroitin sulfate proteoglycan was studied by immunofluorescence.

RESULTS

Expression of ECM by the spheroids was variable. U-87 MG expressed most of the ECM components robustly, whereas AN1/lac-z expressed them all weakly. Fetal rat brain aggregates produced minimal ECM. In cocultures of glioma spheroids and fetal meningeal mesenchymal tissue, individual cells from the glioma spheroids that expressed least fibronectin (AN1/lac-z and U-251 MG) migrated along the fibronectin-positive mesenchymal cells in the culture dish. Cells from the other two lines (U-87 MG and HF-66) that expressed fibronectin strongly did not demonstrate such behavior. None of the other ECM components showed a similar association; mesenchymal cells did not express laminin as strongly as fibronectin, and glioma cells were not observed to align with the laminin-positive structures.

CONCLUSION

This study suggests that fibronectin may play a key role in intracerebral invasion of glioma cells.

Suppression of matrix metalloproteinase-2-mediated cell invasion in U87MG, human glioma cells by anti-microtubule agent: in vitro study

Because microtubules are important components of cell motility and intracellular transport, it is reasonable to propose that the depolymerizing effect of an antimicrotubule agent, estramustine, on glioma microtubules would modulate cell invasiveness. To determine whether matrix metalloproteinases, key factors in cell invasion, are affected by exposure to estramustine, a cell proliferation assay, a zymogram, a collagenolysis assay and a haptoinvasion assay were used in this study. The zymogram revealed that an activated (62 kDa) form of matrix metalloproteinase-2 diminished with increasing estramustine concentrations. The collagenolysis assay demonstrated approximately 2.5- to 21-fold lower rates of enzymatic activity suppressed by estramustine in a dose-dependent manner at estramustine concentrations of 1, 5, and 10 microM, compared with the control group. On the haptoinvasion assay, no statistically significant difference was seen in the 0.5 microM estramustine group, whereas 1-10 microM estramustine groups revealed significant suppression of invasion from 6 to 24 h in a dose-dependent manner. The results suggest that estramustine suppresses the invasion of U87MG cells in vitro using the decreasing available matrix metalloproteinase-2, an effect caused by the disassembly of microtubules. Suppression of the infiltrative capacity of malignant glioma cells could be of significant value in the treatment of this disease.

Translational research models in neuro-oncology

Tumor development and progression in the nervous system are poorly understood. Consequently, even though there seems to be little possibility of major advances in existing clinical modalities used to treat malignant brain tumors, no targeted molecular therapies have risen to take their place. The variability and plasticity of brain neoplasms make them an illusive target for study and therapeutic intervention. Further complicated by infiltration of vital nervous tissue, clinical studies have serious practical limitations and the ability to assess tumor progression in vivo is still a developing technology. Evaluation of potential new therapies for brain tumors is heavily dependent on the development of more informative and cognate experimental models. To develop and validate new models, it is particularly important to integrate clinical, pathological, and cell biological characterizations of malignant brain tumors. This discussion provides an overview of developments in tumor cell culture and the impact of animal models on brain tumor research. Studies of malignant glial neoplasms associated with a dismal prognosis in patients receive particular attention.

Local spread of malignant neuroepithelial tumors

A short review of invasiveness of primary malignant neoplasms in the nervous system is given. Invasiveness implies progressive spread and destruction locally, which eventually leads to a fatal outcome in the patient. In particular, the malignant cells are able to rapidly migrate over large parts of the brain. This process includes the capacity to adhere to a substratum, usually constituted by the various components of the extracellular matrix, followed by detachment and migration. Anatomical structures and local regulatory factors in the brain influence the direction and extent of this migration. Several model systems are now available for monitoring the aggressiveness of such tumours both in vivo and in vitro, and different phenotypic properties characteristic of invasive cells have been elucidated. Although still in its infancy, and currently as an experimental approach, anti-invasive therapy may in the future be an interesting alternative to conventional chemotherapy of brain tumours.

Stimulation of glioma-cell migration by laminin and inhibition by anti-alpha3 and anti-beta1 integrin antibodies

An induction of laminin in the confrontation zone between tumor cells and normal brain tissue has been observed in our model systems in vivo and in vitro. In order to study the effects of ECM components on glioma-cell migration and invasion, we have used 2 lacZ-transfected glioma cell lines, AN1/lacZ and U-251 /lacZ. Cell migration from multicellular spheroids was studied using different types of media: DMEM with 10% serum, Ultra Culture medium, and filtrated DMEM with serum in which the protein fraction > 100 kDa had been removed by ultrafiltration. Laminin, fibronectin and collagen type-IV were individually added to the different media, and cell migration from the spheroids was studied. The results show that cell migration in both cell lines, was stimulated by laminin and fibronectin. Collagen type-IV stimulated only cell migration of U-251/lacZ cells. Scanning electron microscopy revealed an extensive change in cell shape as a result of laminin stimulation. Flowcytometric studies showed that both AN1/lacZ and U-251/lacZ strongly express the alpha3 beta1 integrin receptor, which can bind to several ECM components (laminin, fibronectin, collagen). Immunofluorescence microscopy demonstrated that the same integrin sub-units were expressed in multicellular spheroids. When monoclonal antibodies to alpha3 and beta1 were added to the laminin-stimulated cultures, cell migration was significantly reduced. This indicates that the alpha3 beta1 integrin receptor plays an important role during glioma-cell migration.

In vitro inhibition of cell proliferation, viability, and invasiveness in U87MG human glioblastoma cells by estramustine phosphate

OBJECTIVE

Several determinants of cell motility are highly dependent on the cytoskeleton, in particular, microtubules. To our knowledge, there have been no previous reports regarding the anti-invasive ability by an antimicro-tubule agent, estramustine phosphate (EMP), on glioblastoma cell lines. We investigated the modulated cell proliferation and invasiveness by EMP in vitro.

METHODS

We determined the relative survival rate by cell proliferation assay and the percent survival fraction by monotetrazolium assay. Furthermore, an invasion index was used to quantify the migrating and invasive potential of the human glioblastoma cell line, U87MG, in Boiden's chamber with reconstituted basement membrane (Matrigel; Collaborative Research, Lexington, MA).

RESULTS

We found that 0.5 mumol/L EMP had no effect in any of the assays. Concentrations of 1, 5, and 10 mumol/L demonstrated a concentration- and time-dependent depression in all of the assays. A range of drug concentration of EMP, 1 to 10 mumol/L, in which cell invasiveness was successfully inhibited, was comparable with antiproliferative capacity.

CONCLUSION

The data add to the findings that EMP not only offers selective antiproliferative activity against glioblastoma but also reduces invasiveness, consistent with its main mechanism of action. Such findings form the basis for the development of agents that use non-DNA targets for the treatment of glioblastomas and may improve control over tumor proliferation and invasion.

Glioma invasion in the central nervous system

Invading glioma cells seem to follow distinct anatomic structures within the central nervous system. Tumor cell dissemination may occur along structures, such as the basement membranes of blood vessels or the glial limitans externa, that contain extracellular matrix (ECM) proteins. Frequently, invasive glioma cells are also found to migrate along myelinated fiber tracts of white matter. This behavior is most likely a consequence of using constitutive extracellular ligands expressed along the pathways of preferred dissemination. The extracellular space in anatomic structures, such as blood vessel basement membranes or between myelinated axons, is profoundly different, thus suggesting that glioma cells may be able to use a multiplicity of matrix ligands, possibly activating separate mechanisms for invasion. In addition, enzymatic modification of the extracellular space or deposition of ECM by the tumor cells may also create a more permissive environment for tumor spread into the adjacent brain. Tumor cell invasion is defined as translocation of neoplastic cells through host cellular and ECM barriers. This process has been studied in other cancers, in which a cascade of events has been described that involves receptor-mediated matrix adhesion, degradation of matrix by tumor-secreted metalloproteinases, and, subsequently, active cell locomotion into the newly created space. Although some of these mechanisms may play an important role in glioma invasion, there are some significant differences that are mainly the result of the profoundly different composition of the extracellular environment within the brain. This review focuses on the composition of central nervous system ECM and the recent evidence for the use by glioma cells of multiple invasion mechanisms in response to this unique environment.

Vascular and perivascular GD3 expression in human glioma

A major hallmark of gliomas is their intense neovascularisation. Ganglioside GD3, is one of the major gangliosides which has been implicated in tumour angiogenesis. Recently we reported that GD3 was a potent stimulator of vascular endothelial growth factor release in human glioma cell lines. In the present study we were able to detect GD3-immunoreactivity in 10 out of 10 cases of glioblastoma multiforme and 7 out of 10 cases of anaplastic astrocytoma while low grade tumours were negative. Interestingly, GD3 was intensively expressed in hypervascularised areas of high grade gliomas. These data support the involvement of this ganglioside in brain tumour angiogenesis.

Modulation of matrix metalloprotease-2 and invasion in human glioma cells by alpha 3 beta 1 integrin

We have investigated the effect of integrin antibodies to a well-characterized alpha 5 beta 1 (fibronectin receptor) and to a multi-specific alpha 3 beta 1 (laminin, collagen, and fibronectin receptor), on the expression of matrix metalloproteases and the invasion ability of two human glioblastoma cell lines, SNB19 and U251. Cell adhesion assays indicated that both cell lines adhere to fibronectin, type IV collagen and laminin. Adhesion of cells to fibronectin was inhibited by a RGD peptide. Cells treated with anti-alpha 3 beta 1 or anti-alpha 5 beta 1 antibodies expressed increased levels of MMP-2. An in vitro matrigel assay also showed that the alpha 3 beta 1 antibody-treated cells had greater invasive ability than the controls. Immunofluorescence data showed that glioma cells treated with either anti-alpha 3 beta 1 or anti-alpha 5 beta 1 antibodies expressed diminished alpha 3 beta-1 and alpha 5 beta 1 integrins relative to the controls. The data show that treatment of cells with alpha 3 beta 1 antibody diminishes the integrin expression on the cell surface and increases the MMP-2 activity and invasiveness.

Secretion of matrix metalloproteinase-2 (72 kD gelatinase/type IV collagenase = gelatinase A) by malignant human glioma cell lines: implications for the growth and cellular invasion of the extracellular matrix

Human glioma cells (T98G and A172 cell lines) were cultured on various extracellular matrix (ECM) components including type I, IV and V collagens, fibronectin, laminin, and reconstituted basement membrane (Matrigel), and the role of matrix metalloproteinases (MMPs) in their growth and invasion was examined. T98G glioma cells grew well on these ECM components and invaded the reconstituted basement membrane. In contrast, A172 glioma cells showed growth inhibition on collagen types IV and V and Matrigel without invasion of the Matrigel. Gelatin zymography and enzyme immunoassays demonstrated that T98G glioma cells, but not A172 cells, secrete a large amount of matrix metalloproteinase-2 (MMP-2, 72 kD gelatinase/type IV collagenase = gelatinase A), and this was confirmed by immunoblotting and immunohistochemistry. Of the two different tissue inhibitors of metalloproteinases (TIMP-1 and TIMP-2), T98G cells produced only TIMP-1 during culture on Matrigel, whereas A172 cells secreted both. Although both human recombinant TIMP-1 and TIMP-2 stimulated T98G cell growth slightly on Matrigel, the in vitro invasiveness was significantly reduced by only recombinant TIMP-2. These results suggest that MMP-2 plays an important role in the ECM invasion of T98G human glioma cells in vitro.

Heterogeneic modulation of malignant behavior in human glioma cells in defined and serum-containing media

Malignant features in three glioma cell lines were studied in four defined media of various complexity. The cell lines D37MG, D54MG, and GaMG were able to grow in monolayer culture in all media examined, and as multicellular tumor spheroids in the two most nutrient-rich media. In the defined media, none of the cell lines were able to migrate in a migration assay on poly-D-lysine-coated plastic surfaces. Flow cytometric analysis of the GaMG cell line demonstrated no medium-dependent selection of subclones of glioma cells in spheroids cultured for 30 d. Morphological diversity of spheroids varied according to the supplementation of the media. The capacity of glioma cells to invade cellular rat brain aggregates was intact in the media examined. However, glioma migration was severely inhibited by the lack of specific serum components. This study demonstrates that glioma growth and invasion was heterogeneously preserved in the defined media used. Depending on the assay to be used in the study of glioma cell behavior, the degree of medium supplementation has to be considered.

CD44/chondroitin sulfate proteoglycan and alpha 2 beta 1 integrin mediate human melanoma cell migration on type IV collagen and invasion of basement membranes

Tumor cell invasion of basement membranes (BM) represents one of the critical steps in the metastatic process. Tumor cell recognition of individual BM matrix components may involve individual cell adhesion receptors, such as integrins or cell surface proteoglycans, or may involve a coordinate action of both types of receptors. In this study, we have focused on the identification of a cell surface CD44/chondroitin sulfate proteoglycan (CSPG) and alpha 2 beta 1 integrin on human melanoma cells that are both directly involved in the in vitro invasion of reconstituted BM via a type IV collagen-dependent mechanism. Interfering with cell surface expression of human melanoma CSPG with either p-nitro-phenyl-beta-D-xylopyranoside treatment or anti-CD44 monoclonal antibody (mAb) preincubation (mAb) preincubation inhibits melanoma cell invasion through reconstituted BM. These treatments also strongly inhibit melanoma cell migration on type IV collagen, however, they are ineffective at inhibiting cell adhesion to type IV collagen. Purified melanoma cell surface CD44/CSPG, or purified chondroitin sulfate, bind to type IV collagen affinity columns, consistent with a role for CD44/CSPG-type IV collagen interactions in mediating tumor cell invasion. In contrast, melanoma cell migration on laminin (LM) does not involve CD44/CSPG, nor does CD44/CSPG bind to LM, suggesting that CD44/CSPG-type IV collagen interactions are specific in nature. Additionally, anti-alpha 2 and anti-beta 1 integrin mAbs are capable of blocking melanoma cell invasion of reconstituted BM. Both of these anti-integrin mAbs inhibit melanoma cell adhesion and migration on type IV collagen, whereas only anti-beta 1 mAb inhibits cell adhesion to LM. Collectively, these results indicate that melanoma cell adhesion to type IV collagen is an important consideration in invasion of reconstituted BM in vitro, and suggest that CD44/CSPG and alpha 2 beta 1 integrin may collaborate to promote human melanoma cell adhesion, migration, and invasion in vivo.

Histopathology, classification, and grading of gliomas

Neoplastic transformation occurs in all glial cell types of the human nervous system, producing a wide variety of clinico-pathological entities and morphological variants. Astrocytomas are most common and span an unusually wide spectrum, ranging from the slowly growing juvenile pilocytic astrocytoma to the highly malignant glioblastoma multiforme. Diffusely infiltrating astrocytomas of the cerebral hemispheres show an inherent tendency for progression towards a more malignant phenotype. This change is morphologically categorized in histologic grading schemes (e.g., WHO Grade II to IV) and is associated with the sequential acquisition of genetic alterations, including mutations in the p53 and homozygous deletions of the p16 tumour suppressor genes. Loss of heterozygosity on chromosomes 10 and 19q as well as amplification of the EGF receptor are largely restricted to malignant gliomas and thus considered late events in astrocytoma progression. Gliomas often show phenotypic expression of different glial cell lineages (e.g., oligoastrocytoma). Recent studies suggest that the occurrence of mixed gliomas is not indicative of a polyclonal origin but rather reflects altered gene expression, leading to a change in the balance of growth factors influencing glioma differentiation.

Matrix metalloproteinases and tissue inhibitors of metalloproteinases in human gliomas

The gene expression of five matrix metalloproteinases (MMPs) and two tissue inhibitors of metalloproteinases (TIMPs) was studied in human gliomas in vivo and in vitro to evaluate their roles in glioma invasion. Simultaneous expression of one to four MMP genes and two TIMP genes was found in 17 surgical glioma specimens, and one MMP (gelatinase A) gene and two TIMP genes were simultaneously expressed in tissue of three brains. The concomitant overexpression of gelatinase A, gelatinase B, and occasional matrilysin genes was associated with the malignancy of gliomas and accompanied by overexpression of the TIMP-1 gene. In five human glioma cell lines, gelatinase A, TIMP-1, and TIMP-2 genes were constitutively expressed in alll cell lines: the matrilysin gene in three cell lines; the stromelysin gene in two cell lines; and the interstitial collagenase gene in one cell line. There was a clear difference in the expression of gelatinase B and stromelysin genes between surgical glioma specimens and glioma cell lines: the gelatinase B gene was not expressed constitutively in vitro but was overexpressed in vivo, whereas the stromelysin gene was not expressed in vivo but was expressed in some cell lines. To find the cause of that difference in vivo and in vitro, the transcriptional regulations of MMP and TIMP genes by tumor promoter, growth factors, or cytokines were studied in vitro. Interstitial collagenase, gelatinase B, stromelysin, and TIMP-1 genes were upregulated in many cell lines by phorbol-12-myristate-13-acetate (PMA) and in some cell lines by epidermal growth factor, tumor necrosis factor-alpha, or interleukin-1 beta. Transforming growth factor-beta 1 (TGF beta 1) upregulated gelatinase A and matrilysin genes in some cell lines, and there were no clear responses from any MMP and TIMP genes to interleukin-6. Thus, the transcriptional modulation of MMP genes by these growth factors and cytokines seemed insufficient to explain the difference in gelatinase B and stromelysin gene expressions in vivo and in vitro and was suggestive of the genetic alteration of glioma cells in vitro, the heterogeneous cell population in glioma tissues, or both. Furthermore, the in vitro invasion of glioma cells through Matrigel in response to PMA, TGF beta 1, or TIMP-1 was assessed by chemoinvasion assay. In most cell lines, invasion was significantly stimulated by PMA or TGF beta 1 but suppressed by TIMP-1.(ABSTRACT TRUNCATED AT 400 WORDS)

Adhesion of human glioma cell lines to fibronectin, laminin, vitronectin and collagen I is modulated by gangliosides in vitro

Adhesion of eight cell lines, derived from human gliomas of different histological types, to fibronectin, collagen I, vitronectin, and laminin was investigated in vitro. The glioma cell lines were found to attach to these substrates to different extents. Interestingly, all cell lines strongly attached to laminin. In addition, glioma cell adhesion was found to be dose dependent. Moreover, adhesion of three cell lines to fibronectin and collagen I was partially inhibited and to vitronectin completely prevented by GRGDTP peptide, indicating the involvement of integrin receptors in glioma cell adhesion. We have demonstrated, recently, that gangliosides play an important role in promoting glioma cell invasion of the reconstituted basement membrane, Matrigel, in vitro. In order to study the mechanism of action of gangliosides in this process, the role of six gangliosides (GM1, GM3, GD3, GD1a, GD1b, and GT1b) in cell adhesion to the four proteins was investigated in three cell lines. Although all gangliosides, with the exception of GM3, were found to enhance cell adhesion to these proteins to different extents, GD3 proved to be the most effective adhesion-promoting ganglioside in all three cell lines. GM3 was found to inhibit cell adhesion to the four proteins in one cell line but enhanced cell adhesion in two other cell lines. The three cell lines were found to express both GD3 and gangliosides recognised by the A2B5 antibody. Furthermore, adhesion of the three cell lines to fibronectin, vitronectin, laminin, and collagen I was inhibited by incubation with A2B5, demonstrating the involvement of intrinsic cell membrane gangliosides in adhesion of glioma cells to these proteins. Taken together with the observation that gangliosides modulate integrin receptor function, these data suggest that gangliosides may play a central role in the control of the adhesive and invasive properties of human glioma cells.

Expression of type VI collagen during glioblastoma cell invasion in brain tissue cultures

Human glioblastoma cells, U-87 MG, were utilized in two separate rat brain tissue culture systems. In both cases, the glioblastoma cells deeply penetrated and formed tumor masses inside the brain tissues. Immunofluorescence technique, utilizing anti-type VI collagen antibodies demonstrated strong immunoreactivity of type VI collagen in the tumor masses, invading cells, and cell groups. We suggest that type VI collagen may be involved in tumor cells infiltration and invasion of healthy rat brain tissues. Furthermore, the brain tissue culture method may provide a rapid in vitro model with which cellular and extracellular determinants of invasiveness may be studied.

Interactions of glioma cells and extracellular matrix

The communication between tumor cells and extracellular matrix (ECM) is responsible for clinically important features of malignant gliomas, such as cerebral invasion and leptomeningeal spread. The synthesis of ECM components, ECM-degrading activities and ECM receptors as well as the interaction between ECM components and their receptors represents the molecular basis for these processes. Recent studies have shown that proteases and integrins, the major group of ECM receptors, may be over-expressed by astrocytic tumor cells. Furthermore, integrins and the hyaluronate receptor CD44 have been found to be involved in adhesion and basement membrane invasion of glioma cells. Critical issues which are poorly understood so far include the ECM composition of the normal human brain and of brain tumors, the function of individual ECM components and receptors in a neuro-oncological context, and the molecular processes mediating the diffuse invasion of glioma cells into the brain.

Cell surface gangliosides are involved in the control of human glioma cell invasion in vitro

The involvement of cell surface gangliosides in glioma cell invasion in vitro was examined using the ganglioside-specific antibody A2B5 and Matrigel-coated 8-microns porosity polycarbonate filters. Invasion of six cell lines derived from glial tumours of different histological grades was found to be markedly inhibited by A2B5 (50-96% inhibition) in a dose-dependent manner. Furthermore, exogenous gangliosides were found to prevent cell invasion when they were incubated with cells during the invasion assay. These results suggest that cell surface gangliosides are involved in glioma cell invasion in vitro, probably because of their adhesion-promoting action to basement membrane components.