Connexin 43 enhances the adhesivity and mediates the invasion of malignant glioma cells

Abstract

A hallmark of astrocytic tumors is their infiltrative nature. Although their aggressive and typically widespread dispersal in the adult brain differs fundamentally from that of other brain tumors, little is known about their cellular basis. Astrocytic tumors express the gap junction protein connexin 43 (Cx43), and we show here that Cx43 expression induced the morphological transformation of glioma cells into an epithelial phenotype. In a short-term aggregation assay, Cx43 expression was associated with a several-fold increase in the competence of glioma cells to aggregate. Antibodies directed against the extracellular domain of Cx43 restored the connexin-deficient phenotype, as manifested by a dose-dependent reduction in aggregation. Apart from their role in gap junction formation, connexins may therefore be considered a distinct class of membrane proteins with adhesive properties. Moreover, implanted Cx43-expressing glioma cells established functional gap junction channels with host astrocytes and dispersed through a substantially greater volume of brain parenchyma than mock- and mutant Cx43-transfected sister cells. Cx43 expression therefore may modulate not only the adhesion of astrocytes to one another, but the spread of glial tumor cells throughout astrocytic syncytia. These observations widen our concept of the potential interactions between tumor cells and their surroundings and suggest that both connexin proteins and their derived gap junctions are critical determinants of the invasiveness of central gliomas.

Fig. 1.

Forced expression of connexins is associated with epithelialization and adhesion. Forced expression of Cx-proteins yields both a phenotypic transformation and an increase in adhesiveness of C6 glioma cells. A–C, Phase micrographs of C6-mock transfected cells (A), C6-Cx43 cells (B), and C6-Cx32 cells (C).D–F, Short-term aggregation of C6-mock transfected cells (D), C6-Cx43 cells (E), and C6-Cx32 cells (F). Forced expression of either Cx43 or Cx32 was associated with increased aggregation. Mock-transfected or wild-type C6 cells failed to aggregate. Insets, Immunolabeling of Cx43 or Cx32 in matched cultures. Scale bar: A–C, 30 μm; D–F, 25 μm;insets, 60 μm.

Fig. 2.

Antibodies directed against Cx43 diminish astrocytic dye coupling. All connexins share a similar topology consisting of four membrane-spanning segments with the N and C termini both on the cytoplasmic side, flanking two extracellular loops (E1 and E2). E1 and E2 are highly conserved regions, each containing three cysteine residues that dock hemichannels in neighboring cells to form a functional gap junction.A, Predicted topology of Cx43, with four transmembrane segments, two external loops (E1 and E2), and the N and C termini located at the cytoplasmic side of the membrane. Polyclonal antibodies were raised against segments that included amino acids 46–76 and 186–206 of E1 and E2, respectively (indicated by red lines). B,C, Immunolabeling of confluent cultures of cortical astrocytes with antibodies directed against E1 (B) and E2 (C). Cx43 plaques were located in areas of cell-to-cell contact.D, E, Inhibition of dye transfer in cortical astrocytes after incubation in anti-E1 IgG Fab (60 μg/ml) (E), compared with vehicle-treated control cultures (D). Gap junctional coupling was quantified by transfer of CDCF (green) from DiIC₁₈ labeled cells (red) to unlabeled identical cells. Donor cells appear yellow because of to the merge of red and green labeling. The coupling index of 8.4 ± 0.4 in control astrocytic cultures was reduced to 2.7 ± 0.7 in plates exposed to anti-Cx43E1 and 3.6 ± 0.5 in cultures exposed to anti-Cx43E2 (p < 0.0001 by two-tailed, one-way ANOVA). Scale bar (shown in E): B,C, 30 μm; D, E, 45 μm.

Fig. 3.

Dose-dependent inhibition of aggregation by anti-Cx43. A, Anti-Cx43 (60 μg/ml) almost completely blocked aggregation of C6-Cx43 cells, whereas 10 μg/ml anti-Cx43 (B) attenuated but did not entirely block aggregation. In contrast, >90% of vehicle-treated control cells formed aggregates (C). Scale bar, 20 μm.D, Histogram summarizes the net proportion of aggregated cells in anti-Cx43-treated cultures (10, 30, and 60 μg/ml Fab, and 60 μg/ml Fab preabsorbed with respective blocking peptide). Controls included Fabs prepared from normal rabbit serum (NRS), anti-NgCAM (α8D9), and anti-rat L1 (αL1) (Goldman et al., 1996), each at 60 μg/ml (*p < 0.01; ANOVA and post hocBonferroni t test).

Fig. 4.

Separation and distinction of connexin-mediated adhesion and channel formation. Plaque formation, but not functional gap junction channels, was required for Cx-mediated adhesion.A, Diffuse Cx43 immunoreactivity in the cytosol of C6 cells transfected with mutant C61S-Cx43, which harbors a cysteine-to-serine point mutation at position 61. The C61S-Cx43 protein did not form gap junction plaques and correspondingly did not increase the adhesive capability of sister cells (B).C, C6 cells transfected with a chimeric construct Cx40*43C3 displayed immunoreactive plaques at cell–cell interfaces (arrows) and an increase in cell aggregation (D). Scale bar: A,C, 25 μm; B, D, 30 μm.E, Histograms summarizing both the incidence of dye transfer and the percentage of aggregating cells among the clones studied (*p < 0.01; ANOVA and post hoc Bonferroni t test).

Fig. 5.

Implanted glioma cells establish functional gap junctions with host astrocytes. Intercellular transfer of Lucifer yellow from CMTMR-labeled glioma cells (C6-Cx43) to GFAP-positive astrocytes. A, Diagram illustrating preparation and intrastriate implantation of glioma cells labeled with the cell tracker, CMTMR. Three days later, vibratome sections were prepared, and isolated CMTMR-labeled glioma cells that had migrated a minimum of 80 μm from the injection site were patch clamped with a pipette solution containing 4% of the gap junction permeable dye Lucifer yellow.B, Confocal image of the live slice obtained during the experiment illustrating the CMTMR-labeled glioma cells (red) and Lucifer yellow (white). The impaled CMTMR-labeled glioma cell is marked by a red arrowhead.C, Confocal imaging of the same field after the slice was fixed and immunostained for GFAP (green). Lucifer yellow transferred from the CMTMR-labeled glioma cell (red arrowhead) to two GFAP-positive astrocytes (yellow arrowhead) and six GFAP-negative (small white arrow) neighboring cells. Scale bars: B, 100 μm; C, 20 μm.

Fig. 6.

Cx43 expression directs aggressive parenchymal invasion. Analysis of parenchymal versus adluminal invasion of CMTMR-labeled (A) astrocytes, C6-Cx43 (B), C6-mock clone 1 (C), and C6-Cx40*43C3 (D). The cells were implanted 4 d before rats were killed. The vasculature was outlined by fluorescein-dextran (green; 2 × 10⁶ Da, administered intravenously). Vibratomesections (100 μm) were optically sectioned using confocal imaging with a step size of 0.5 μm and reconstructed in three dimensions. All cell types studied invaded brain along the vessels with few or no cells embedded in the brain parenchyma (C). The exceptions to this rule were astrocytes (A) and C6-Cx43 cells (B). Astrocytes and C6-Cx43 cells were frequently positioned in tissue without direct contact to the vasculature, indicating that these two cell types were capable of invading the brain tissue without vessel guidance (white arrows). Cx40*43 hybrid mutants, which aggregated as well as wild-type Cx43 in the short-term in vitro assay (Fig.3D,E) but failed to establish functional gap junctions with host astrocytes in vivo, were relatively scarce in the parenchyma (D, two white arrowheads). Because of intracellular compartmentalization of CMTMR, only part of the invading cells (especially astrocytes) is visible inA–D. Scale bar, 100 μm.E, Comparison of the invasive pattern of astrocytes and C6 clones. The number of CMTMR-labeled cells was counted as a function of their distance from the tumor border. The number of cells was quantified in a rectangle with the dimensions 387 × 580 × 100 μm = 22 × 10⁻¹²m³. The positions of CMTMR-labeled cells in relation to the vessels were scored as either adluminal (directly contacting vessel; C, D) or parenchymal (without contacting vessel; A, B). Data represent mean ± SEM from a total of six implanted tumors (except C6-C61S-Cx43, which include 4 tumors). *p < 0.01 (ANOVA and post hoc Bonferroni ttest).