To study the pharmacological selectivity of tP2YR, we generated a number of clonal 1321N1 human astrocytoma cell lines stably infected with a retrovirus expression vector encoding the tP2YR. Surprisingly, a number of these cell lines exhibited an unusual morphologic and growth alteration not observed with the expression of other purinergic receptors in 1321N1 cells (data not shown). While vector-infected cells were unchanged in appearance from control untransfected 1321N1 cells and formed confluent monolayers of cells, the tP2YR-expressing cells displayed an elongated morphology and showed string-like patterns and cell clusters (Figure 1). Furthermore, the analyses of multiple, independent clonal cell lines indicated that this altered morphology correlated with elevated basal phospholipase C activity (data not shown). This ability to cause morphological transformation of the 1321N1 astrocytoma cell line was unique among the numerous receptors that we have evaluated previously, including the mammalian P2Y₂ receptor (Boyer and Harden, unpublished observation).

The altered morphology of tP2YR-expressing 1321N1 cells suggested that this GPCR may exhibit transforming properties. To investigate this hypothesis further, we determined if the tP2YR could also cause transformation NIH3T3 cells, a nontransformed immortalized mouse fibroblast cell line that we and others have shown to be sensitive to morphologic and growth transformation by Rho GTPase activation 27 . This experimental paradigm has provided a sensitive assay to detect the transforming activity of other GPCR oncoproteins and has been instrumental in dissecting the signal transduction pathways responsible for their transforming activity 21 22 23 24 . Using a focus formation assay, we found that transfection of pZIP-tP2YR in to NIH3T3 cells caused the appearance of foci of transformed cells (Figure 2A). Interestingly, the appearance of these foci was distinct from the foci caused by activated Ras(61L), which are comprised of highly refractile, elongated cells and instead were similar to those caused by activated RhoA(63L), which are comprised of densely-packed, nonrefratile cells. Furthermore, tP2YR-induced foci were similar in appearance to those induced by two other transforming GPCRs that cause transformation by activation of either Rac1-dependent (Mas) or RhoA-dependent (G2A) signaling pathways 21 22 .

To determine if overexpression of the tP2YR reduced the requirement of NIH3T3 cells for serum growth factors, tP2YR-expressing cells were plated at low density in growth medium supplemented with 10% calf serum. After 24 hours, the cultures were switched to growth media supplemented with 10%, 2% or 0.5% calf serum. Untransformed NIH3T3 cells require growth medium supplemented with 10% calf serum for efficient growth in culture. As expected, cells transformed by activated Ras continue to proliferate in 2% or 0.5% serum, whereas vector-transfected control cells exhibited very limited proliferation in growth medium supplemented with 2% or 0.5% serum (Figure 2B). Similar to activated Ras-expressing cells, tP2YR-expressing cells formed dense colonies even in 0.5% serum, indicating that expression of the tP2YR can reduce the growth requirement of NIH3T3 cells for serum growth factors.

To further characterize the ability of tP2YR to cause growth transformation of NIH3T3 cells, we assessed the ability of tP2YR-expressing cells to exhibit anchorage- and serum-independent growth. For these experiments, we isolated NIH3T3 cells stably transfected with expression constructs encoding either H-Ras(61L), RhoA(63L), Mas, tP2YR, or the pZIP-NeoSV(x)1 empty vector control. Following selection in growth medium supplemented with G418, multiple antibiotic-resistant colonies (>100) were pooled together to establish stable cell lines to be used for these analyses. To assay the ability of the tP2YR expressing cells to exhibit anchorage-independent growth, each pooled cell line was suspended in soft agar. As described previously, cells expressing activated Ras or RhoA formed colonies in soft agar 28 and Figure 2C). The tP2YR-expressing cells also readily formed colonies in soft agar indicating that expression of tP2YR can cause anchorage-independent growth. Interestingly, tP2YR-expressing colonies were similar in size and morphology to those formed by Mas-transformed cells. For example, these colonies did not spread laterally in the agar as mutant Ras-expressing colonies, but instead formed clusters of cells that protruded out of the plane of the agar.

The morphological appearance of the tP2YR-induced foci suggested that tP2YR might cause transformation by activation of Rho family proteins. One well-characterized property of transformation by Rho proteins and their activators is their ability to cooperate with activated Raf-1 to cause synergistic transformation of NIH3T3 cells 28 29 30 31 32 . To determine if tP2YR could also cooperate with activated Raf-1, plasmids encoding activated Raf-1(340D) were co-transfected at concentrations that resulted in very low, or no, focus-formation along with either tP2YR, activated Rac1(115I) as a positive control, or the empty vector negative control. As we observed previously, co-expression of activated Rac1(115I) with Raf-1(340D) caused a 100-fold enhancement of focus-forming activity over that seen with expression of Rac1(115I) or Raf alone 28 . Similarly, as shown in Figure 3A, co-transfection of tP2YR with activated Raf caused a 15- to 20-fold synergistic enhancement of focus-forming activity. These results indicate that like activators of Rho proteins, tP2YR can cooperate in a focus formation assay with activated Raf-1.

In addition to increasing the number of tP2YR-induced foci, co-expression with activated Raf altered the morphology of the transformed foci (Figure 3B). As described previously, co-expression of activated Raf-1 with either RhoA or G2A causes the appearance of transformed foci that are similar to those caused by activated Ras and contain elongated, highly refractile cells 21 . In contrast, co-transfection of activated Rac1 or Mas with activated Raf-1 caused the appearance of foci that were distinct, and contained rounded cells with a non-refractile appearance 21 . The transformed foci caused by co-expression of Raf-1 and tP2YR were indistinguishable in appearance from those caused by expression of activated Raf-1 and either activated Rac1 or Mas, suggesting that tP2YR may activate Rac1.

The appearance of tP2YR foci suggests that it causes transformation by activation of Rac1 or a related protein. To address this possibility, we assessed the ability of tP2YR to activate Rac1 and RhoA in pull-down assays. These assays take advantage of the increased affinity of activated Rac1 or RhoA for the isolated Rac/Rho binding domains (RBDs) of Pak and Rhotekin, effectors of Rac and RhoA, respectively. For these experiments, NIH3T3 cells were transfected with either pcDNA3 or tP2YR along with either Tiam1-C1199 (positive control for Rac1 activation) or Ect2-DH/PH/C (positive control for RhoA activation). Cell lysates were incubated with GST-Pak-RBD to assess Rac1 activation or GST-Rhotekin-RBD to assess RhoA activation. As shown in Figure 4B and 4C, expression of tP2YR increased activation of both Rac1 and RhoA. These data indicate that tP2YR can cause activation of Rho family proteins.

We next investigated which Gα subunit(s) may link tP2YR to Rho protein activation. P2Y-purinergic receptors can couple to either Gα_q, activating PLC or to Gα_i inhibiting adenylyl cyclase activity. Typically purinergic receptors do not couple to both of these G-proteins. However, our previous studies indicated that this tP2YR has the unique ability to robustly activate both Gα_q and Gα_i 25 26 . Therefore, we sought to determine if activation of either or both of these Gα subunits may contribute to tP2YR transforming activity by examining the ability of constitutively activated mutants of Gα_i and Gα_q either, alone, or in combination, to cause focus-formation in NIH3T3 cells. When transfected alone, neither activated Gα_i, nor activated Gα_q caused the appearance of transformed foci in our NIH3T3 cells (data not shown). However, co-expression of activated Gα_i and Gα_q caused the appearance of transformed foci that were identical in appearance to those induced by the tP2YR or Mas (Figure 5A). These results suggested that the coordinate activation of both Gα_i and Gα_q by the tP2YR maybe responsible for its transforming activity.

To examine the requirement for Gα_i activation for focus formation by these GPCR oncogenes, NIH3T3 cells were transfected with expression constructs encoding either Mas, tP2YR, G2A, or Ras(61L) and the transfected cultures were maintained in the presence or absence of PTX to inactivate Gα_i. Treatment of cells with PTX caused a 90% inhibition of Mas- and tP2YR-induced focus-formation (Figure 5B). In contrast, PTX treatment did not cause any reduction in oncogenic Ras-induced focus formation, indicating that the reduction seen with tP2YR was not due to nonspecific inhibition by PTX. Interestingly, PTX also caused a 60% reduction in G2A-induced focus-formation. We previously showed that G2A causes transformation by activation of Gα13, which in turn causes activation of RhoA by stimulation of the Lsc/p115 RhoGEF 22 23 . These data suggest that GPCR oncogenes such as Mas, tP2YR and to a lesser extent G2A rely on Gα_i activity for induction of a transformed phenotype.

While our in vitro Rac1 and RhoA activation assays indicate that tP2YR can activate both Rho family proteins, based on the appearance of transformed foci we favor the interpretation that activation of Rac1 (or a related protein) maybe more important for the transformed phenotype induced by tP2YR. For example, the appearance of tP2YR-induced foci was identical to those induced by expression of the Mas oncogene, which causes transformation by activation of Rac1 rather than RhoA 21 . In contrast, G-proteins and GPCRs such as Gα_12/13 and G2A cause the appearance of Rho-like foci and require RhoA protein function for transformation 22 23 24 33 34 . Additionally, tP2YR cooperated with activated Raf in a focus formation assay causing the appearance of foci that resemble those by activated Raf plus Rac1 or Rac1 activators such as Mas 21 . This is in contrast to the appearance of foci induced by activated Raf plus RhoA or RhoA activators such as G2A 22 . These data suggest that while tP2YR is capable of activating both Rac1 and RhoA in an in vitro biochemistry assay; the morphologically predominant phenotype induced by expression of tP2YR suggests that Rac1 activation maybe more important for the transforming activity of this oncogene.

Treatment of cells with ATP can influence cell growth and proliferation in opposing ways depending on the cell type and the combination of purinoceptors expressed. In rare instances, stimulation of P2Y receptors can promote cellular proliferation and oncogenic transformation 5 6 12 13 . Since the ability of tP2YR to cause tumorigenic transformation is unique among the receptors that we have characterized, the possibility remains that this avian receptor is constitutively activated. One characteristic of the tP2YR that sets it apart from other purinoceptors is its ability to simultaneously activate both Gα_i and Gα_q 25 26 . This observation prompted us to test the hypothesis that transformation by tP2YR maybe due to the coordinated activation of both Gα_i and Gα_q. We found that while neither Gα_i nor Gα_q when expressed alone in NIH3T3 cells induced focus formation, when co-expressed, these proteins induced the appearance of transformed foci. Furthermore, the morphology of Gα_i and Gα_q induced foci was identical to those induced by expression of either tP2YR or Mas indicating that similar pathways may be activated to promote transformation. While previous studies have demonstrated only a mild focus formation activity of Gα_q in NIH3T3 cells 35 , others indicate that Gα_q expression induces apoptotic cell death 36 37 , we have been able to establish stable cell lines that express Gα_q using a relatively weak cytomegalovirus promoter (pcDNA3). Furthermore, focus formation was not observed in these stable cell lines indicating that Gα_q expression alone is not transforming in the NIH3T3 strain used in these studies. Nor was Gα_q transforming when co-expressed with activated Raf (data not shown). Finally, focus formation by both Mas and tP2YR requires Gα_i activity supporting the hypothesis that it is the coordinated activation of these two Gα subunits, which leads to growth transformation by these GPCR oncoproteins.

How coordinated activation of Gα_i and Gα_q by tP2YR leads to activation of Rho proteins remains unknown. The mammalian P2Y₂ receptor can activate both RhoA and Rac1 through interaction with integrins mediated by an RGD sequence motif located in the first extracellular loop of the receptor 13 . The interaction of the P2Y₂ receptor with integrins is necessary for coupling to Gα_o leading to activation of Rac1 and Gα₁₂ and to activation of RhoA but is not required for activation of Gα_q 14 . Since, the tP2YR does not contain an RGD sequence, it is likely that activation of Rho proteins are mediated by this alternative pathway. Previous studies have found that the activation of Gα_i and Gα_q by expression of multiple P2Y receptors on the same cell can cause a synergistic rise in second messenger production 13 . It is possible that this synergistic activation of second messengers can lead to Rho activation.

The tP2YR mammalian retrovirus expression construct pLXSN-tP2YR has been described previously 25 . pZIP-NeoSV(x)1 retrovirus expression vectors encoding wild type Mas, and the constitutively activated and transforming Rac1(61L), Rac1(115I), H-Ras(61L), RhoA(61L), or Raf-1(340D) mutant proteins and pCDNA3 expression vectors encoding GTPase-deficient, constitutively activated mutants of Gα_q(QL), Gα_i(QL), Rac1(12V), and RhoA(12V) have also been described previously 38 39 .

1321N1 human astrocytoma cells were grown in Dulbecco's minimum Eagle's medium (DMEM) supplemented with 5% fetal calf serum. Infection of 1321N1 astrocytoma cells was done as described 25 . NIH3T3 mouse fibroblasts were grown in DMEM supplemented with 10% calf serum and transfected as described previously 40 . For focus formation assays, 60 mm dishes were transfected and maintained in growth medium for three weeks. Individual foci were either photographed and/or stained with 0.4% crystal violet before quantitation. To evaluate the ability of pertussis toxin (PTX) treatment to inhibit transformation, focus formation assays were performed as above in growth medium supplemented with 100 ng/ml of PTX (Sigma). Growth medium and PTX were replaced every other day for 14-16 days. For generation of stable cell lines, transfected cultures were maintained in growth medium supplemented with 400 μg/ml G418 (Geneticin; GIBCO-BRL) and following selection, multiple G418-resistant colonies (>100) were pooled. Assays to examine growth in soft agar and low serum have been described previously 41 .

Assays to determine cellular Rac1-GTP and RhoA-GTP levels were performed as described previously 42 43 . Briefly, NIH3T3 cells were transiently transfected with either empty pcDNA3 expression vector or those encoding tP2YR, the C-terminal 1199 amino acids of Tiam-1, C1199 44 or the DH/PH/C domain of Ect2 45 , using Lipofectamine Plus Reagent (Invitrogen Inc.). Cells were grown for 24 h in DMEM supplemented with 0.5% fetal calf serum. Cell lysates were incubated with GTP-dependent binding domains for Pak1 (GST-PAK-RBD) or Rhotekin (GST-Rhotekin-RBD) as we have described previously 42 43 .