Growth Factors Outside of the Platelet-derived Growth Factor (PDGF) Family Employ Reactive Oxygen Species/Src Family Kinases to Activate PDGF Receptor α and Thereby Promote Proliferation and Survival of Cells*

The vitreous contains a plethora of growth factors that are strongly implicated in the formation of fibroproliferative diseases such as proliferative vitreoretinopathy. Although platelet-derived growth factors (PDGFs) are present in the vitreous, vitreal growth factors outside of the PDGF family activated the PDGF α receptor (PDGFRα) and promoted disease progression in a rabbit model of proliferative vitreoretinopathy (H. Lei, G. Velez, P. Hovland, T. Hirose, D. Gilbertson, and A. Kazlauskas (2008) submitted for publication.) In this report we investigated the mechanism by which non-PDGFs activated PDGFRα. We found that non-PDGFs increased the cellular level of reactive oxygen species (ROS) and that this event was necessary and sufficient for phosphorylation of PDGFRα. We speculated that the underlying mechanism was ROS-mediated inhibition of phosphotyrosine phosphatases, which antagonize receptor auto-phosphorylation. However, this did not appear to be the case. Non-PDGFs promoted tyrosine phosphorylation of catalytically inactive PDGFRα, and thereby indicated that at least one additional tyrosine kinase was involved. Indeed, preventing expression or blocking the kinase activity of Src family kinases suppressed non-PDGF-dependent tyrosine phosphorylation of PDGFRα. Thus non-PDGFs increased the level of ROS, which activated Src family kinases and resulted in phosphorylation of PDGFRα. Finally, although non-PDGFs induced only modest phosphorylation of PDGFRα, proliferation and survival of cells in response to non-PDGFs was significantly enhanced by expression of PDGFRα. These studies reveal a novel mechanism for activation of PDGFRα that appears capable of enhancing the responsiveness of cells to growth factors outside of the PDGF family.

The platelet-derived growth factor (PDGF) 2 family consists of five ligands, which assemble dimeric receptors consisting of homo-or hetero-combinations of the two PDGF receptor (PDGFR) subunits (1)(2)(3). Analysis of cultured cells and/or mice lacking PDGFs or PDGFRs has provided vast and convincing evidence that PDGF/PDGFRs regulate many cellular responses and thereby contribute to a wide spectrum of physiological processes (3).
In a rabbit model of PVR, PDGFRs are essential for disease formation (4,28,29), and the vitreous contains high levels of PDGFs (15). These observations suggest that vitreal PDGFs activate PDGFRs and thereby facilitate the development of PVR. Surprisingly, neutralizing vitreal PDGFs only modestly attenuate experimental PVR, 3 which suggests that PDGFRs are undergoing activation via a non-traditional route. Indeed, vitreal growth factors outside of the PDGF family promote modest tyrosine phosphorylation of PDGFR␣. 3 Other investigators have also reported increased tyrosine phosphorylation of PDGFRs by agents that are not PDGF. Certain agonists of G protein-coupled receptors, autoantibodies in the blood of scleroderma patients and agents that are within the bone marrow (but are probably not PDGFs) promoted tyrosine phosphorylation of PDGFR (31)(32)(33)(34)(35)(36)(37)(38). Thus there is a growing appreciation that the direct, PDGF-based mechanism is not the only route to induce tyrosine phosphorylation of PDGFRs. The goal of this study was to identify the intracellular events and mediators by which non-PDGFs induced tyrosine phosphorylation of PDGFR␣ and to assess whether indirect activation of PDGFR␣ contributed to cellular responses induced by non-PDGFs.

EXPERIMENTAL PROCEDURES
Cell Culture-F, F␣, and F␣⌬X cells were previously described (4). 3 Briefly, they are mouse embryo fibroblasts derived from mice null for both pdgfr genes. They were immortalized with SV40 T antigen. F␣ and F␣⌬X cells are F cells in which we re-expressed the human PDGFR␣ that was the fulllength or the mutant lacking the majority of the extracellular domain, respectively.
SYF and SYF ϩ Src cells were previously described (39). Briefly, they are mouse embryo fibroblasts derived from src, yes, and fyn knockout mice. SYF␣ and SYF␣ ϩ Src cells were generated by expressing human PDGFR␣ in SYF and SYF ϩ Src cells, respectively (40). The SYF␣ and SYF␣ ϩ Src cells were used in these studies and are designated as "Ϫ/Ϫ" and "Src" in Fig. 3.
Ph␣ and R627 cells were previously described (41). Briefly, they are spontaneously immortalized mouse embryo fibroblasts derived from mice harboring the Ph deletion that includes the pdgfra gene. The wild-type or kinase-inactive human pdgfra cDNA was expressed in parental Ph cells to generate Ph␣ and R627, respectively. ARPE19 (RPE19) cells are a human retinal pigment epithelial cell line that was purchased from American Type Culture Collection. RPE19␣ cells were generated by expressing PDGFR␣ in the parental RPE19 cells (15), which naturally express a very low level of PDGFR␣.
Immunoprecipitation and Western Blot-Cells were grown to 90% confluence and then incubated for 24 h in DMEM or DMEM/F-12 without serum. The cells were exposed to the desired agents as follows: NAC, Tiron: 10 mM for 10 min; Catalase: 144 nM for 10 min; SU6656: 1 M for 1 h; unless indicated otherwise growth factors were added for 10 min at a concentration of 50 ng/ml for PDGF and 100 ng/ml for bFGF. The cells were washed twice with ice-cold phosphate-buffered saline, and then lysed in extraction buffer (10 mM Tris-HCl, pH 7.4, 5 mM EDTA, 50 mM NaCl, 50 mM NaF, 1% Triton X-100, 20 g/ml aprotinin, 2 mM Na 3 VO 4 , 1 mM phenylmethylsulfonyl fluoride). Lysates were centrifuged for 15 min at 13,000 ϫ g, 4°C, and PDGFR␣ was immunoprecipitated from clarified lysates as previously described (15). The immunoprecipitating antibody was a crude rabbit polyclonal (27P). The blotting antibody was a 1:1 mixture of anti-phosphotyrosine antibodies (4G10/PY20). The primary blot membrane was stripped and reprobed with the PDGFR␣ antibody, 27P, or number 3164 from Cell Signaling. At least three independent experiments were performed. Signal intensity was determined by densitometry using Quantity One (Bio-Rad) and normalized for the amount of PDGFR␣ in each sample (15).
DCF Assay-Intracellular H 2 O 2 generation was measured indirectly using the fluorescent dye, 2Ј,7Ј-dichlorofluorescein diacetate (DCFH-DA), as described by several other groups (44 -46). Briefly, serum-deprived RPE19␣, rabbit conjunctiva fibroblast, and F␣ were stimulated for 10 min with bFGF (100 ng/ml) in DMEM without phenol red or PVR rabbit vitreous. Subsequently the cells were rinsed twice with Krebs-Ringer solution and incubated in Krebs-Ringer solution containing DCFH-DA (5 M). DCFH-DA is non-polar and readily diffuses into cells, where it is hydrolyzed to the non-fluorescent polar derivative DCFH and thereby trapped within the cells. In the presence of H 2 O 2 , DCFH is oxidized to the highly fluorescent DCF. Culture dishes were sealed with paraffin film and placed in a CO 2 incubator at 37°C for 5 min, after which DCF fluorescence was photographed using a Zeiss Axiovert 135 inverted immunofluorescence microscope equipped with a digital camera. To avoid photooxidation of DCF, we collected the fluorescent images at a 9.0-s exposure under green light and identical contrast and brightness parameters for all samples. Cells that were cultured in 96-well plates and treated in the same way were read with a Bio-Tek fluorescence plate reader at excitation and emission wavelengths of 485 and 528 nm, respectively.
Cell Proliferation Assay-Cells (F, F␣, and F␣⌬X) were seeded into 6-well plates at a density of 60,000 cells/well in DMEM plus 10% fetal bovine serum. After 8 h the cells had attached; the medium was aspirated, the cells were rinsed twice with phosphate-buffered saline, and the cells were cultured in serum-free DMEM and treated with certain reagents as follows: 1) DMEM only; 2) bFGF (1, 4, 20, 100, and 200 ng/ml); 3) bFGF (20 ng/ml) plus NAC (10 mM); 4) PDGF-A (50 ng/ml); and 5) PDGF-A plus NAC (10 mM). The media were replaced every day. The cells were counted in a hemocytometer on days 1, 2, 3, and 4, and at least three independent experiments were performed.
Apoptosis Assay-Cells were seeded into 10 cm-dishes at a density of 2 ϫ 10 5 cells per dish in DMEM plus 10% fetal bovine serum. After the cells had attached the dishes, they were treated as described above under "Cell Proliferation Assay." On Day 3, the cells were harvested and stained with fluorescein isothiocyanate-conjugated Annexin V and propidium iodide according to the instructions provided with the apoptosis kit (BD Biosciences). The cells were analyzed by flow cytometry in Coulter Beckman XL (47). At least three independent experiments were performed. The raw data from a representative experiment are shown in supplemental Fig. S1.
Statistics-All data were analyzed using the unpaired t test. p values of Ͻ0.05 were considered statistically significant.

Non-PDGF-triggered Tyrosine Phosphorylation of PDGFR␣ was Dependent on ROS-It has been demonstrated that
increasing the level of reactive oxygen species (ROS) can lead to tyrosine phosphorylation of many proteins, including the PDGF ␤ receptor (PDGFR␤) (48). Consequently, we tested if non-PDGFs increased ROS, and whether this event was likely to be involved with tyrosine phosphorylation of the closely related PDGFR␣. In these experiments we used vitreous from normal rabbits (instead of rabbits with PVR) to mimic the conditions at the start of the disease. Unlike the vitreous from rab-bits with PVR, vitreous from normal rabbits had undetectable amounts of PDGFs, although it did have a variety of growth factors outside of the PDGF family (15). Fig. 1 (A and B) shows that either rabbit vitreous, or bFGF (a representative vitreal non-PDGF), increased the level of ROS in three cell lines that are highly relevant to experimental and/or clinical PVR. Furthermore, exogenously added hydrogen peroxide strongly stimulated phosphorylation of PDGFR␣ (Fig. 1C), indicating that a rise in ROS was sufficient to increase the phosphotyrosine content of PDGFR␣. Pretreating cells with anti-oxidants NAC or Tiron (10 mM for 10 min), or an enzyme that scavenges H 2 O 2 (catalase, 144 nM, for 10 min) prevented non-PDGFs and H 2 O 2 from phosphorylating PDGFR␣ (Fig. 1C). In contrast, direct activation of PDGFR␣ and FGFR1 was unaffected by antioxidants ( Fig. 1, C and D). These results indicate that ROS was produced in response to non-PDGFs and was required for them to induce tyrosine phosphorylation of PDGFR␣.
One of the mechanisms by which ROS could increase tyrosine phosphorylation of PDGFR␣ is by inactivating phosphotyrosine phosphatases that dephosphorylate autophosphorylated PDGFR␣ (49,50). To test this ROS/phosphotyrosine phosphatase hypothesis we determined if non-PDGFs induced tyrosine phosphorylation of kinase-inactive PDGFR␣, which is incapable of autophosphorylating (51). Using a previously characterized pair of cell lines that expressed either the wild-type or kinase-inactive (R627) PDGFR␣ (51), we found that there was no difference in the extent of tyrosine phosphorylation of the two receptors in response to any of the three agents that promoted receptor phosphorylation in an ROS-dependent manner ( Fig. 2A). As expected, direct activation of PDGFR␣ via PDGF increased tyrosine phosphorylation of the wild-type, but not R627 PDGFR␣ ( Fig. 2A). Western blot analysis of total cell lysates prepared from the two cell lines indicated that the wildtype and R627 PDGFR␣ were expressed to similar levels (Fig.  2B). These findings indicate that the ROS/phosphotyrosine phosphatase hypothesis was either incorrect or incomplete. Furthermore, the mechanism by which non-PDGFs increased phosphorylation of PDGFR␣ involved a kinase that was not PDGFR␣ itself.
SFKs Were Required for Non-PDGF-dependent Tyrosine Phosphorylation of PDGFR␣-We considered whether this kinase was an Src family kinase (SFK), because SFKs can be activated by hydrogen peroxide (52,53) and have many substrates, including PDGFRs (54). All of the agents that promoted phosphorylation of PDGFR␣ activated c-Src (Fig. 3A). Furthermore, phosphorylation of PDGFR␣ induced by bFGF, rabbit vitreous (RV) or hydrogen peroxide was either undetectable or severely attenuated in mouse embryo fibroblasts that lacked SFKs (SYF cells, "Ϫ/Ϫ") (Fig. 3B). Re-expression of c-Src restored phosphorylation of PDGFR␣ by all three agents without altering the expression level of PDGFR␣ (Fig.  3, B and C). In contrast (and consistent with previous observations (39)), direct activation of PDGFR␣ did not require expression of SFKs (Fig. 3B). Similarly, pharmacologically inhibiting SFKs had no effect on PDGF-dependent phosphorylation of PDGFR␣, yet it blocked non-PDGF-dependent phosphorylation of PDGFR␣ (Fig. 4). These results identify SFKs as an intracellular mediator of non-PDGFdependent tyrosine phosphorylation of PDGFR␣.

Indirect Activation of PDGFR␣ Promoted Proliferation and Survival of Cells in Response to
Non-PDGFs-Although direct activation of PDGFRs robustly induced cellular responses such as proliferation and survival, whether the modest increase in the phosphotyrosine content of PDGFR␣ induced by non-PDGFs was also capable of doing so was an open question. Previous investigators found that non-PDGF-triggered signaling events related to cell proliferation and survival (activation of Akt) were enhanced by expression of PDGFR␣ (55). The increase in such signaling events may be sufficient to enhance cellular responses. To test this idea we considered if indirect activation of PDGFR␣ promoted cellular responses by comparing bFGF-induced cell proliferation and survival in cells that did or did not express PDGFR␣. A dose response experiment indicated that both responses plateaued at 20 ng/ml bFGF ( Fig. 5A and 6A). bFGFinduced proliferation was enhanced by expression of fulllength PDGFR␣ (Fig. 5B and Table 1). When we assessed apoptosis in this panel of cell lines we observed enhanced survival (reduced apoptosis) under both basal and bFGF-stimulated conditions in the PDGFR␣-expressing cells as compared with the non-expressing cells (Fig. 6B and Table 1). The enhanced survival in the absence of exogenously added growth factors may by due to indirect activation of PDGFR␣ by endog-FIGURE 2. PDGFR␣ kinase activity is dispensable for its phosphorylation in response to non-PDGFs. A, mouse embryo fibroblasts expressing either the full length (␣) or kinase inactive (R627) PDGFR␣ were treated as described in the legend of Fig. 1. PDGFR␣ was immunoprecipitated and subjected to an anti-phosphotyrosine Western blot. In three independent experiments bFGF, RV, and H 2 O 2 induced a 3.0 Ϯ 0.5-, 6.1 Ϯ 0.6-, and 34.5 Ϯ 3.1-fold increase, respectively, of the wild-type receptor; whereas the -fold increase for the kinase-dead receptor was 3.2 Ϯ 0.4-, 6.2 Ϯ 0.5-, and 36.5 Ϯ 3.6-fold, respectively. The results show that phosphorylation of PDGFR␣ by non-PDGFs was independent of the receptor's kinase activity. B, cleared lysates from the indicated cell lines were subjected to Western blotting using an anti-PDGFR␣ or anti-RasGAP antibody. The resulting data are representative of at least two independent experiments and show that the level of PDGFR␣ was similar in both cell lines. FIGURE 3. SFKs were required for phosphorylation of PDGFR␣ by non-PDGFs. A, RPE19␣ cells were serumstarved for 24 h and exposed to PDGF-A (50 ng/ml), bFGF (100 ng/ml), RV, or H 2 O 2 (5 mM) for 10 min at 37°C. Cleared lysates were subjected to Western blot using a phospho-Src (Tyr-416) antibody. The stripped membrane was reprobed with a pan-Src antibody. The ratio of the resulting signals is shown at the bottom. In three independent experiments the average Ϯ S.D. was 3.0 Ϯ 0.3, 2.5 Ϯ 0.2, 2.7 Ϯ 0.3, and 36.2 Ϯ 3.5 for PDGF-A, bFGF, RV, and H 2 O 2 , respectively. These data show that all agents activated Src. Although the magnitude of the increase in phosphorylation was modest, it is consistent with what other groups report using this approach (30,64) and what we have found using a different approach to monitor activation of SFKs (43). B, mouse embryo fibroblasts that lacked all SFKs (39) were reconstituted with either an empty expression vector (Ϫ/Ϫ), or wild type c-Src (Src) (40) and stimulated with the indicated agents as described in panel A. In three independent experiments the PDGF-A, bFGF, RV, and H 2 O 2 -dependent -fold increases in phosphorylation of PDGFR␣ in Ϫ/Ϫ and Src-expressing cells were 28.5 Ϯ 2.5 and 29.2 Ϯ 3.2, 0.99 Ϯ 0.1 and 4.8 Ϯ 0.5, 0.98 Ϯ 0.2 and 6.2 Ϯ 0.6, and 2.2 Ϯ 0.3 and 38.5 Ϯ 2.3, respectively. The results indicate that, although c-Src greatly increased the ability of non-PDGFs to increase the phosphotyrosine content of PDGFR␣, PDGF-dependent phosphorylation of PDGFR␣ was independent of expression of SFKs. C, cleared lysates from the indicated cell lines were subjected to Western blotting using an anti-PDGFR␣ or anti-RasGAP antibody. The resulting data are representative of at least two independent experiments and show that the level of PDGFR␣ was similar in both cell lines. enously produced factors. Cultured cells produce growth factors (56), and PDGFR␣ can be indirectly activated by a variety of such factors. 3 We concluded that expression of PDGFR␣ enabled enhanced proliferation and survival in response to non-PDGFs such as bFGF.
Another way to consider these data is in the reverse order. The non-PDGF was more effective at driving proliferation and survival of PDGFR␣-expressing cells as compared with the cells that did not express PDGFR␣ (Figs. 5 and 6 and Table 1). Because FGFR1 was expressed to comparable levels in both cell lines (Fig. 6C), the difference in the magnitude of the responses was not simply due to a lower level of FGFR1 in the less-responsive cells.
The cells used in these studies secrete both PDGF-C and the proteases necessary to process it (15), which raised the possibility that the enhanced responsiveness of the PDGFR␣-expressing cells was the result of direct activation of PDGFR␣ via endogenously produced PDGF-C. To address this possibility we repeated the experiments with cells the expressed ␣⌬X PDGFR␣. This truncated receptor cannot be activated by PDGF (because it lacks the majority of the extracellular domain), yet is fully transactivatable. 3 As shown in Figs. 5 and 6 and Table 1, the ␣⌬X PDGFR␣ potentiated non-PDGFdriven cellular responses as well as the full-length receptors. These findings rule out the possibility that a PDGF-C/PDGFR␣ autocrine loop is responsible for the greater responses in PDGFR␣-expressing cells.
Finally, we tested if the enhanced responsiveness in PDGFRexpressing cells was sensitive to anti-oxidants. When assayed in the presence of NAC, the basal proliferation of cells decreased and apoptosis increased somewhat (Figs. 5B and 6B and Table  1). More importantly, while bFGF promoted proliferation and survival in NAC-treated cells, the responses were not greater in PDGFR-expressing cells as compared with the F cells. These observations indicated that ROS is essential for non-PDGFs to promote cellular response in PDGFR␣-expressing cells.

DISCUSSION
The findings described herein begin to elucidate intracellular events and identify effectors that are required for growth factors outside of the PDGF family to induce tyrosine phosphorylation of PDGFR␣ (Fig. 7). Non-PDGFs bind and activate their own receptors and thereby increase the cellular level of ROS. This leads to a rise in SFK activity, which either directly or indirectly phosphorylates PDGFR␣. This series of events is associated with increased proliferation and survival of cells exposed to non-PDGFs. The indicated cells were seeded at a density of 60,000 cells per well of 6-well plates. The cells were permitted to attach, and after two rinses with phosphate-buffered saline, serum-free medium was added that was (ϩ) or was not (Ϫ) supplemented with bFGF. The media were replaced every day, and the cell number was counted on day 3. A, a dose-response curve showing that 20 ng/ml bFGF induced the maximal response. B, the cells were treated with 20 ng/ml bFGF, 50 ng/ml PDGF-A, 10 mM NAC as indicated. The bars indicate the raw data (cell number), whereas the -fold change is presented in Table 1. The mean Ϯ S.D. of three independent experiments is shown; *, p Ͻ 0.05 using an unpaired t test. Although bFGF promoted proliferation of all cell lines, the response was enhanced in cells expressing PDGFR␣, and this phenomenon was blocked by NAC. C, cleared lysates from the indicated cell lines were subjected to Western blotting using an anti-PDGFR␣ or anti-RasGAP antibody. The resulting data are representative of at least two independent experiments and show that the level of PDGFR␣ was similar in both cell lines.

TABLE 1 Expression of PDGFR␣ promoted non-PDGF-driven cell proliferation and survival
The indicated cellular responses were assayed as described in the legend of Figs. 5 and 6. The mean Ϯ S.D. of the bFGF-induced-fold change from three independent experiments is shown.  MARCH 6, 2009 • VOLUME 284 • NUMBER 10

JOURNAL OF BIOLOGICAL CHEMISTRY 6333
Tyrosine phosphorylation of receptor tyrosine kinases by ligands other than their cognate receptors has been previously reported. For instance, activation of G protein-coupled receptors increases the activity of metalloproteases that release bound heparin-binding-EGF, which activates the EGF receptor (57). This mechanism seems relevant to PDGFR␣ phosphorylation, because the cell lines under investigation secrete PDGF-C and the proteases necessary to activate it (15). Thus non-PDGFs could promote either secretion, or processing of PDGF-C, and thereby drive tyrosine phosphorylation of PDGFR␣. However, this mechanism seems unlikely in light of the fact that removing the extracellular domain of PDGFR␣ did not prevent its tyrosine phosphorylation by non-PDGFs. 3 Profoundly different mechanisms appear responsible for direct (PDGFdriven) and indirect (via non-PDGFs) phosphorylation of PDGFR␣. Direct activation involves the extracellular ligand-binding domain, depends on activating the receptor's intrinsic kinase activity and proceeds without any apparent input from SFKs. In contrast, indirect activation appears to involve coordinating events that result in activation of SFKs that either directly or indirectly phosphorylate the resting receptor. These findings reveal that PDGFR␣ can be phosphorylated by SFKs under certain circumstances and in this regard is similar to other receptor tyrosine kinases. For instance, c-Src phosphorylates EGFR and potentiates EGF-dependent cellular responses (58,59). Moreover, SFKs directly or indirectly promote tyrosine phosphorylation of the hepatocyte growth factor receptor (Met) in colon cancer cell lines (60).
Although direct activation of PDGFR is thought to proceed by the same mechanism, regardless of which PDGF isoform is used, there may be multiple mechanisms by which agents outside of the PDGF family promote phosphorylation of PDGFRs. Tyrosine phosphorylation of PDGFR␣ by angiotensin (an agonist of a G protein-coupled receptor) is independent of SFKs (33). In contrast, there was a strict requirement for SFKs in order for polypeptide growth factors to promote tyrosine phosphorylation of PDGFR␣. Thus there appear to be at least two different routes by which resting PDGFRs can be engaged in signaling events triggered by agents outside of the PDGF family.
In light of the promiscuity of ROS/SFKs, one may expect that all receptor tyrosine kinases would be phosphorylated when these intracellular agents are increased/activated. Yet this turned out not to be the case. PDGF promoted a rise in ROS and activated SFKs, yet there was no perceptible increase in the phosphotyrosine content of FGFR1. 3 How the ROS/SFK-di-FIGURE 6. Expression of PDGFR␣ promoted non-PDGF-driven survival. The indicated cells were seeded at a density of 200,000 cells/10-cm dish. The cells were permitted to attach, and after two rinses with phosphatebuffered saline, serum-free medium was added that was (ϩ) or was not (Ϫ) supplemented with bFGF. The media were replaced every day. On day 3 the cells were harvested, stained with fluorescein isothiocyanateconjugated Annexin V and propidium iodide, and analyzed by flow cytometry. Cells in early apoptosis were defined as the population that was positive for Annexin V, but negative for propidium iodide. A, a doseresponse curve showing that 20 ng/ml bFGF induced the maximal response. B, the cells were treated with 20 ng/ml bFGF, 50 ng/ml PDGF-A, 10 mM NAC as indicated. The bars indicate the % of apoptotic cells (raw data from a representative experiment is in supplemental Figs. S1 and S2), whereas the -fold change is presented in Table 1. The mean Ϯ S.D. of three independent experiments is shown; *, p Ͻ 0.05 using an unpaired t test.
Although bFGF protected all cell lines from apoptosis, it was more effective for cells that expressed PDGFR␣, and this phenomenon was blocked by NAC. C, cleared lysates prepared from F and F␣ cells were subjected to Western blot analysis using the indicated antibodies. The data show that the level of FGFR1 was not altered by expression of PDGFR␣. Although we organized our findings to emphasize the differences between the direct and indirect mechanisms by which PDGFR␣ was activated, there may be some overlap. As noted above, direct activation of PDGFR␣ via PDGF increased the level of ROS, which was not required for tyrosine phosphorylation of the receptor. However, increasing the level of ROS appeared to be required for cell survival even when the PDGFR␣ was activated directly. PDGF protected F␣ cells from apoptosis much more efficiently in the absence of NAC than it its presence ( Fig. 6 and Table 1). In contrast, NAC treatment had little impact on PDGF-dependent proliferation of these cells. These findings suggest that ROS was required for some of the responses, even when the receptor was activated directly.
Our data re-enforce the emerging concept that PDGFRs can be activated by more than just PDGFs, and that this can boost the responsiveness of cells to non-PDGFs. How do these tissue culture findings extend to physiology/pathology? Because knocking out either PDGFR␤ or PDGF-B gene in mice resulted in very similar phenotypes (61,62), it appears that PDGF-B is the primary activator of PDGFR␤ during development. In contrast, activation of PDGFR␣ by non-PDGFs appears to play an important role in the manifestation of fibroproliferative diseases such as PVR. 3 Furthermore, there is a breach of traditional signaling specificity in the context of breast cancer. A subpopulation of lung cancer patients treated with EGFR inhibitors developed lung cancers that overexpressed the hepatocyte growth factor receptor, which acted via the kinase-independent ErbB3 to activate phosphatidylinositol 3-kinase (63). Thus nontraditional activation of receptor tyrosine kinases appears to be more relevant to pathology than development. Identifying the mediators of these pathways and elucidating the mechanisms by which they drive cellular responses is likely to provide the conceptual foundation to confront endemic diseases.