Temozolomide modifies caveolin-1 expression in experimental malignant gliomas in vitro and in vivo

Abstract

Background: Caveolin-1 is a protein that displays promotive versus preventive roles in cancer progression according to circumstances. Temozolomide (TMZ) is the standard chemotherapeutic to treat glioma patients. The present work aims to characterizeTMZ-induced effects on caveolin-1 expression in glioma cells.

Methods: Human astroglioma (U373 and T98G) and oligodendroglioma (Hs683) cell lines were used in vitro as well as in vivo orthotopic xenografts (Hs683 and U373) into the brains of immunocompromisedmice. In vitro TMZ-induced effects on protein expression and cellular localization were determined by Western blot analysis and on the actin cytoskeleton organization by means of immunofluorescence approaches. In vivo TMZ-induced effects in caveolin-1 expression in human glioma xenografts were monitored by means of immunohistochemistry.

Results: TMZ modified caveolin-1 expression and localization in vitro and in vivo after an administration schedule that slightly, if at all, impaired cell growth characteristics in vitro. Caveolin-1 by itself (at a 100-ng/ml concentration) was able to significantly reduce invasiveness (Boyden chambers) of the three human glioma cell lines. The TMZ-inducedmodification in caveolin-1 expression in flotation/raft compartments was paralleled by altered Cyr61 and β(1) integrin expression, two elements that have already been reported to collaborate with caveolin-1 in regulating glioma cell biology, and all these features led to profound reorganization of the actin cytoskeleton. An experimental Src kinase inhibitor, AZD0530, almost completely antagonized the TMZ-induced modulation in caveolin-1 expression.

Conclusion: TMZ modifies caveolin-1 expression in vitro and in vivo in glioma cells, a feature that directly affects glioma cell migration properties.

Figure 1

Temozolomide modulates raft resident caveolin-1 expression in vitro and increases caveolin-1 expression in vivo in experimental GBMs. (A) Western blot analysis of the expression of raft resident caveolin-1 in the different flotation fractions (F4–F12) obtained from T98G GBM cells left untreated (CT) or treated with 100 µM TMZ for 7 h/d for five consecutive days. Raft resident caveolin-1 expression was measured on the third day (day 3) after the end of the TMZ treatment. (B) Caveolin-1 immunoreactivity in Hs683 GBM tumors implanted into brains of immunodeficient mice and treated (Bb and Bc) or not treated (Ba) with TMZ (40 mg/kg per os, three times per week for three consecutive weeks, with immunohistochemical analyses performed 1 week after the end of the treatment). Caveolin-1 expression was absent from the Hs683 tumor (T) left untreated (Ba), in which only the vessels (Vs) were immunoreactive for caveolin-1. Caveolin-1 expression was marked in Hs683 tumors (T) after TMZ treatment (Bb). Caveolin-1 expression in the tumor bulk (T) was associated with accumulation of caveolin-1 in the extracellular matrix around the invasive part of the tumor (black arrows in Bc). The normal brain (NB) expressed weak levels of caveolin-1 (Bc).

Figure 2

Temozolomide decreases β₁ integrin and modulates Cyr61 and actin cytoskeleton organization in GBM cell lines. (A) Western blot analysis of the expression of β₁ integrin in U373, T98G, and Hs683 GBM cells left untreated (CT) or treated with 100 µM TMZ for 7 h/d for five consecutive days. The β₁ integrin expression was measured on the third (day 3), fifth (day 5), and seventh (day 7) days after the end of the TMZ treatment. (B) Western blot analyses of the expression of Cyr61 in U373, T98G, and Hs683 GBM cell lines left untreated (CT) or treated with 100 µM TMZ using a treatment schedule identical to that described above for β₁ integrin. (C) Fluorescence microscopic visualization of the actin cytoskeleton (red fluorescence shows globular (nonpolymerized) actin and green fluorescence shows fibrillary (polymerized) actin) under control conditions (CT) and 3 (TMZ D3) and 5 (TMZ D5) days after termination of TMZ (100 µM) treatment in U373 (Ca) and Hs683 (Cb) glioma cells according to a treatment schedule identical to that described above for β₁ integrin.

Figure 3

Extracellular caveolin-1 decreases GBM cell line invasion. Invasion was evaluated using Matrigel Boyden chambers. The number of invasive cells was counted in 10 fields per chamber (the chamber surface includes 30 fields), and the experiment was conducted in triplicate. The mean of the 10 values was used to calculate the percentage of invasive cells in the population seeded, and the control invasion rate shown on each graph was reported to be 100%. The effects of 100 ng/ml recombinant caveolin-1 were examined both when the protein was added in the upper (“Top”; promigratory or antimigratory factor) or the lower (“Bottom”; chemoattractant vs chemorepulsive factor) compartments of the chamber. Results are expressed as mean of the triplicate ± SE.

Figure 4

The TMZ-induced caveolin-1 modulation is Src-dependent in Hs683 GBM cells. (A) Western blot analysis of the expression of soluble caveolin-1 in U373, T98G, and Hs683 GBM cells left untreated (CT) or treated with 100 µM TMZ for 7 h/d for five consecutive days. Soluble caveolin-1 expression was measured on the third (day 3) and fifth (day 5) days after the end of the TMZ treatment. (B) Western blot analyses of soluble caveolin-1 expression in Hs683 glioma cells treated with TMZ (100 µM) four times per week (day 1–4) for 7 h/d, the EGFR inhibitor (10 µM) (erlotinib; day 1), the Src inhibitor AZD0530 (10 µM) (day 1), and combination of the inhibitors and TMZ (+TMZ) compared with control untreated cells (Ct). Soluble caveolin-1 expression was measured on day 5.