Gliomas are a type of primary brain tumor that display an extensive invasive behavior, but do not metastasize. The growth and invasiveness into the surrounding normal brain tissue makes gliomas a major challenge for clinical intervention 123. Therefore, the elucidation of the mechanisms involved in tumor growth, invasion and angiogenesis could aid in the treatment of these tumors.

The invasive cellular behavior of malignant gliomas is determined by receptor mediated cell-substratum contacts and cell-cell interaction, as well as cellular locomotion 4. Clues to the invasion process have been ascertained through clarification of the key roles played by the extracellular matrix (ECM), cell-adhesion molecules and matrix degrading proteases 5.

Among other elements, vascular endothelial growth factor (VEGF) is one of the most prominent angiogenic growth factors. In fact, it has been shown that VEGF is secreted by the C6 rat glioma cell line 6. This growth factor is produced by almost all solid tumors and its receptors are highly expressed on vascular endothelial cells and, predominantly, in vessels in the proximity of the tumor 7. Moreover, some other systems may be involved in the growth and invasion of the gliomas. Glutamate secreting glioma cells demonstrate a growth advantage and glioma cells release excitotoxic concentrations of glutamate, thus killing neurons close to the tumor border and opening space for tumor growth 8910.

Another important signalling molecule that may be involved in glioma development is ATP 1112. In glioma C6 cells, ATP can bind and activate P2Y2, whereas its degradation product, ADP, activate P2Y1 and P2Y12 receptors 13141516.

In previous studies, we showed the involvement of the purinergic system in glioma proliferation in different cell types 17 and that glioma cells present a clear resistance to death induced by cytotoxic concentrations of ATP when compared with normal brain tissue 18. Taken together with its inability to metabolize extracellular ATP 19, the purinergic system may form part of a novel and important mechanism associated with the malignance of these tumors.

ATP may be liberated into the extracellular space by the excitotoxic death of the normal host cells and by the injury caused by tumor resection, which is the mainstay of initial therapy for gliomas 20. We hypothesize that ATP liberated under these conditions may, besides inducing proliferation of the glioma itself, induce the cell death of normal tissue and, consequently, increase the interstitial space around the malignant glioma.

To test this hypothesis, we examined the effect of co-injection of apyrase, an ATP depleting enzyme, in a C6 rat glioma experimental model, which has been extensively used to test antitumoral interventions.

In the present study we tested the effect of the apyrase enzyme on the growth of glioma cells in vivo. We have previously shown that C6 gliomas, a cell line that was originated from N-nitrosomethylurea-treated rats, exhibited almost undetectable ATP and ADP-degrading activity; however, our results did not allow the identification of which NTPDase family members are expressed by these cells 19. To investigate which of the ecto-NTPDases are present in the C6 glioma cell line, total RNA was isolated from cultured glioma cells and the cDNA was analyzed by Real Time-PCR with primers for the NTPDase family. Figure 1 shows that C6 cells effectively have a low expression of all NTPDases investigated, in comparison with normal astrocytes. We, therefore, chose the purified apyrase, which hydrolyzes ATP and ADP at a similar rate to test our hypothesis of the ATP participation in the process of glioma invasion in host brain.

The implanted tumors have characteristics that are closer to those of human glioblastomas with the C6 cells growing in the intracerebral, intraventricular and intraparenchymal spaces (data not shown). Haematoxylin and eosin examination showed that the implanted tumor has a high mitotic index, nuclear pleomorphism, foci of tumor necrosis, intratumoral hemorrhage and parenchymal invasion (Table 1). In some cases, pathological analysis identified pallisading cells delineating the foci of necrosis and lymphocytic infiltration, with the occasional formation of edema fluid and neovascularization, which are characteristics of glioblastoma multiforme in humans (Table 1) (Fig. 3a and 3c).

In order to analyze whether the tumors implanted into host brain were responsive to drugs that are clinically used to treat gliomas and to validate our model, the rats were treated with the alkylating agent, temozolomide 25. In fact, tumor size analysis showed that the rats treated with temozolomide presented a significant reduction in tumor size when compared with the control groups (Fig. 2a). In addition, rats treated with temozolomide survive significantly longer than their respective controls, demonstrating a relative benefit of 100% in the rats' survival, that means, at the time that 100% of control animals died all animals from the treated group were still alive (data not shown).

Confirming our purinergic hypothesis of glioma growth, the results presented in Figure 2b show that the rats with implanted glioma co-injected with apyrase had a significant reduction in the tumor size (p < 0.05) in comparison with the rats injected only with gliomas or with gliomas plus inactivated apyrase, indicating that it is not the presence of the polypeptide, but the enzymatic activity that is responsible for the effect of apyrase. The effect of apyrase (2 U apyrase/10⁶ cells) on C6 cell proliferation in culture showed that the enzyme was not toxic to the cells but instead, induced an increase by around 40% of the cell number after 24 h of treatment. Additionally, the C6 cells were detached from the surface handled exactly as for the injection (with or without apyrase) and were subsequently re-plated in DMEM/FCS for 24 and 48 h. Cell viability was analyzed by trypan blue exclusion and MTT assays immediately after the cell preparation for injection and after 24 and 48 h of culture. The cells presented more than 90% viability when compared to the control cells in all conditions tested, indicating that the effect of apyrase is not probably on the initial viability of the injected cells but rather on the subsequent tumor implantation and growth.

According to the pathological analysis, the malignant gliomas induced by C6 injection and co-injected with apyrase presented a significant reduction in the mitotic index and other histological characteristics that indicate a less invasive/proliferative tumor (Table 1). Less lymphocytic infiltration and peritumoral edema were observed as well as the absence of pseudopalisading necrosis (Fig. 4b and 4d). Reduction of proliferation induced by apyrase co-injection was confirmed by counting the percentage of Ki67 positive glioma cell nuclei (Table 2) (Fig. 4a and 4b).

Since tumor growth is dependent on the ability to induce angiogenesis, we performed immunohistochemical experiments with CD31. According to counts with CD31, vessel density and neoformation was higher in the C6 group 20 days after implantation, in comparison with the group co-injected with apyrase (Table 2). Confirming this observation, the implanted gliomas with apyrase presented less VEGF staining in comparison to the control group (glioma alone) (Fig. 4c and 4d).

In the present study, we investigated the effect of the co-injection of the enzyme apyrase with implanted C6 gliomas cells. It is well known that in vitro assays for growth and invasion only represent isolated aspects of the multi-cascade process of the in vivo tumor growth 4. For this reason, we decided to use an in vivo glioma model that, despite its limitations, is useful for the study of growth, angiogenesis and invasion of gliomas 3.

Extracellular ATP could be hydrolyzed by the action of a cascade of enzymes, which includes the family of E-NTPDases: NTPDase1 (CD39), NTPDase2 (CD39L1), NTPDase3 (CD39L3), NTPDase5 (CD39L4) and NTPDase6 (CD39L2) 26 and the more recently described NTPDase7 and NTPDase8 262728. Some of these enzymes have been related with different tumors 293031 and recent studies demonstrated that NTPDase5 (CD39L4) is a proto-oncogene PCPH, which is involved in tumor growth 32. According to our results, the C6 cells present a very low expression of the NTPDase1 (CD39), NTPDase2 (CD39L1) and NTPDase3, the three main enzymes involved in the cell surface ATP degradation when compared to astrocytes. Previous studies showed that nucleotides and nucleosides induce proliferation in glioma cell lines 17 while ATP can mediate death in dissociated primary cerebellar granules or striatal neurons and in hippocampal organotypic cultures 33. The indirect autocrine nature of this process should be noted. The tumor induces, through the death of normal cells, the release of more ATP that stimulates its own multiplication and, as a consequence, the cell death of other cells opening space to be occupied by tumor cells and liberating more ATP to continue the invasive process. Thus we hypothesize that the re-establishment of NTPDase activity may reduce glioma growth.

Purified apyrase is an enzyme widely used to deplete ATP and hydrolyzes ATP and ADP at the same rate as NTPDase1 34. Therefore, the apyrase may regulate the extracellular nucleotide concentration and, thus, modulate their biological effects when activated by P2Y and P2X receptors.

Our experiments demonstrate that the co-injection of apyrase significantly diminished the growth of implanted gliomas in rats after twenty days of tumor induction. Considering that the injection of apyrase was done only at the moment of implantation, the effect of ATP depletion is probably at the implantation and initial growth of the glioma. This could be of therapeutic interest, since the application of apyrase in the surgical resection cavity could be helpful in reducing the initial growth of invaded tumor.

In addition to the reduction in the glioma size, pathological analysis demonstrated the lack of some important malignant characteristics, typical of the glioblastomas, in the tumors co-injected with apyrase (Table 1). Among these, of note is the reduction of the mitotic index in the rats treated with apyrase, indicating that ATP has an important role in glioma proliferation in vivo. The participation of ADP in this process should not be excluded since this nucleotide can bind to P2 receptors and is potentially hydrolyzed by the injected apyrase.

As mentioned above, angiogenesis is an extremely important process for sustained tumor growth. Studies have demonstrated that C6 and human glioma cell lines secrete VEGF 63536. Since invasive properties are largely dependent on the presence of new vessels, CD31 was studied and demonstrated a greater staining in the samples from rats implanted with the C6 cells in relation to those co-injected with apyrase (Table 2). This result was confirmed by the stronger VEGF staining in rats implanted with glioma, when compared with the apyrase group (Fig. 4c and 4d). These results clearly indicate that microvascular proliferation is decreased in tumors treated with apyrase in relation to the control group, with a consequent decrease in the blood flow in tumor areas with high cell proliferation. The insufficient vascular supply may become the limiting factor for tumor growth 7.

In conclusion, the data presented herein clearly show that the presence of an ATP-depleting enzyme at the moment of glioma implantation causes a significant decrease in the growth, angiogenesis and proliferation index. The exact mechanisms to explain these observations remain under evaluation, however, the participation of ATP and the ecto-nucleotidases may be associated with the development of this type of brain tumor in an in vivo glioma model.

CNS, central nervous system; DAB, 3,3'diaminobenzidine tetrahydrochloride; DMEM, dulbecco's modified eagle's medium; DMSO, dimethyl sulfoxide; ECM, extracellular matrix; FCS, fetal calf serum; H&E, hematoxylin and eosin; VEGF, vascular endothelial growth factor.

The author(s) declare that they have no competing interests.

FBM performed cell growth, glioma injection, drug treatments, conducted H&E and immunohistochemical analysis of gliomas, was responsible for data analysis and drafting of the manuscript. DLO participated in the glioma injection. PWG performed drug treatments and tumor size measurements. JS performed cell growth. SW participated in the glioma injection and glutamate uptake assay. MRW carried out Real Time PCR studies. LM and MIAE conducted H&E and immunohistochemical analysis of gliomas. GL participated in the study design and drafting of the manuscript. AMOB conceived of the study, participated in its design, coordination and in the draft of the manuscript. All authors read and approved the final manuscript.