Our demonstration that specific downregulation of endogenous FIP200 protein can increase Pyk2 activity and induce apoptosis adds to the body of evidence that FIP200 can regulate Pyk2 and underscores the importance of FIP200 in the regulation of cell survival. Collectively, our results indicate that FIP200 regulation of Pyk2 can play an important role in the survival of glioblastoma tumor cells and suggests the intriguing possibility that this mechanism may contribute to sustained tumor growth by promoting not only the survival of tumor cells but also the intra-tumor glial cells and the tumor-associated endothelial cells.
These results underscore the importance of FIP200 in the regulation of cell survival in that they demonstrate that downregulation of FIP200 can induce apoptosis. The potential role of pPyk2 in promoting the apoptosis induced by FIP200 downregulation was indicated by the ability of a TAT-Pyk2 fusion protein containing the autophosphorylation site to inhibit the pro-apoptotic effect, whereas a TAT-Pyk2 fusion protein containing the potential Grb2 binding site (Y881) did not. This is consistent with the prior report that the N-terminus and the kinase domains of Pyk2, but not the C-terminus, are necessary for the pro-apoptotic effect of Pyk2 
. Several other mechanisms have been implicated in FIP200 regulation of apoptosis in other cell types. In the FIP200-null mouse embryos, no changes in Pyk2 activity were detected in the hepatocytes or cardiomyocytes and the enhanced apoptosis was attributed to a decrease in S6 kinase activity 
. The TSC1&2 complex is thought to negatively regulate cell proliferation through S6 kinase, mTOR, and 4E-BP1 
. On analysis of the p70 S6 kinase activity in the glioblastoma cells we did find that there was a slight reduction (an estimated 30%) in this activity on downregulation of FIP200. This small decrease in S6 kinase activity could contribute to the inhibition of proliferation we observed on FIP200 downregulation but the current results suggest that the mechanism by which FIP200 regulates apoptosis in glioblastoma tumor cells and brain MvECs differs from that in the embryonic murine cardiocytes and hepatocytes in that it requires Pyk2 activity. It also has been reported that FIP200 upregulates the expression of the retinoblastoma (Rb
) gene 
, thus the downregulation of endogenous FIP200 protein in our cells could potentially inhibit cell proliferation by decreasing the Rb
gene expression. Although the association of FAK with the death domain kinase receptor interacting protein (RIP) is thought to prevent the pro-apoptotic signal of RIP 
, we could not detect an association of FAK with RIP in the U-87MG cells by co-immunoprecipitation (D. Wang and C.L. Gladson unpublished observation).
Although both Pyk2 and FAK were expressed in the glioblastoma cells and human brain MvEC, we found that FIP200 preferentially co-immunoprecipitated with Pyk2 in the absence of FIP200 downregulation in the unmanipulated U-87MG glioblastoma and brain MvEC cells propagated in complete media. The association of the endogenous FIP200 with Pyk2 in the lysates of the unmanipulated glioblastoma cells and human brain MvEC propagated in complete media suggests that the FIP200 regulation of Pyk2 activity in these cells may be a direct effect. The interaction of FIP200 with Pyk2 has been shown to inhibit Pyk2 activity in other cell types 
and Pyk2 activation has been shown to promote its dissociation from FIP200. Thus, it is possible that downregulation of the endogenous FIP200 may lead to exposure of the autophosphorylation site permitting autophosphorylation of Pyk2 
The activities of Pyk2 and FAK and their contribution to specific cell functions, including apoptosis, are regulated in large part by their shuttling between different compartments of the cell. The largely cytoplasmic and diffuse pattern of FIP200 that we observed on immunofluorescence analysis of both the U-87MG cells and the primary human brain MvECs is consistent with the pattern reported for fibroblasts 
. The distribution of Pyk2 also was predominantly cytoplasmic and diffuse, which is consistent with the preferential co-precipitation of FIP200 with Pyk2 we observed, whereas FAK was largely localized to the focal adhesions and FIP200 was not detectable at these sites. The regulation of localization of FAK and Pyk2 to focal adhesions can be independently influenced by the presence or absence of specific binding partners, for example, gelsolin has been reported to bind Pyk2 but not FAK 
as well as differences in the affinity of FAK and Pyk2 for shared binding partners within the focal adhesions, such as paxillin 
. FIP200 localization to the nucleus has been reported in other cell types 
, emphasizing a potential role for FIP200 in regulation of transcription through, for example, its potential interaction with Rb, which is a regulator of transcription. We did not observe a clear-cut association of FIP200 with the nuclei of the U-87MG or primary human MvECs, but we did not utilize biochemical assays to explore this possibility and cannot rule out a partial nuclear localization. However, the pattern of expression of FIP200 in the tumor cells in the glioblastoma biopsies was heterogeneous. Although the pattern was predominantly cytoplasmic, nuclear staining was observed and partial FIP200 localization to the nucleus was observed in tumor cells in 10 of 36 biopsy samples. In contrast, in the tumor-associated endothelial cells of glioblastoma biopsies, the distribution of FIP200 was largely cytoplasmic with only rare nuclear staining.
We observed a reduced association of Fyn with FAK on FIP200 downregulation in both U-87MG cells and the brain MvEC although the levels of activity of FAK and its expression were unaffected by the downregulation. This supports the concept that activated Pyk2 competes with FAK for binding partners that affect downstream signaling pathways, as shown in other cell types by Zhao and colleagues 
. Activated Pyk2 can signal the activation of p38MAP kinase 
, a MAP kinase that can promote apoptosis 
, and Pyk2 induction of apoptosis in cardiomyocytes requires Src activation of p38MAP kinase 
. FAK can promote cell proliferation and cell survival through the activation of other MAP kinase(s) in a cell-type and experimental context-dependent manner reviewed in 
It should be emphasized that there was heterogeneity in the levels of expression of FIP200 in the tumor cells and that FIP200 was not detectable in all of the tumor cells within a single biopsy specimen. In addition, we did not detect FIP200 protein in tumor cells in 8 of 36 of the glioblastoma biopsies, although FIP200 was detected in scattered glial cells and the endothelial cells in the tumor-associated vessels in 3 of these 8 tumors. Currently, it is not known whether the failure to detect FIP200 protein in tumor cells in eight of the glioblastoma biopsy samples is due to a gene deletion in these cells or an environmentally triggered reduction in the levels of FIP200 mRNA or protein.
The location of the FIP200
gene on chromosome 8q11 
has led other investigators to formulate the hypothesis that FIP200 could be an “anti-tumor” or “tumor suppressor” gene. Recently, however, Wei and colleagues 
created a conditional knockout of FIP200 in the MMTV-Cre model of breast cancer and showed no promotion of tumorigenesis, suggesting that FIP200 inactivation alone does not promote tumor formation. In addition, a prior report demonstrated that overexpression of FIP200 in breast cancer cells resulted in an inhibition of both cell cycle progression and clonogenic cell survival due to FIP200 promotion of p21 expression 
. It is possible that the level of FIP200 protein in cells, and the cell type, influences its function. For example, the kinetics of FIP200 association with some of its binding partners, such as Pyk2, FAK or the p21 or Rb promoter, could be altered when FIP200 is overexpressed and this could be affected by the experimental context and the specific cell-type. Taken together with the expression of FIP200 in the tumor-associated endothelial cells in the glioblastoma tumors but not the normal brain, these data underscore the importance of elucidation of the factors that regulate the expression of FIP200.