Pdgfr, Raf, And Erk Are Phosphorylated And Subsequently Dephosphorylated Upon Stimulation Of Isolated Fibroblasts With Pdgf In Culture

Trip Adler Biological Sciences 54 Final Writing Project 5/6/04

PDGFR, Raf, and ERK Are Phosphorylated and Subsequently Dephosphorylated Upon Stimulation of Isolated Fibroblasts with PDGF in Culture Trip Adler, Mei Rosa Ng, Robin Kachka, Benigno Varela, Danielle Andrews-Lovell, Kedar Gumaste, Kimmie Keefe, Laura Hellett, Laura Perretta, Stephanie Wang

Abstract Extracellular signaling is a process in which a molecule outside of a cell binds to a transmembrane receptor to activate a signaling pathway. One particular example of this is PDGF (platelet-derived growth factor) binding to a RTK (receptor tyrosine kinase) to activate the Ras-ERK kinase pathway in fibroblasts. After choosing an antibody that binds to fibroblasts, using it to purify a culture of this cell type, and then checking the efficiency of the sort using microscopy, we used the pure culture to study the PDGF-activated pathway. We stimulated the culture with PDGF, and took samples of unstimulated cells, cells thirty minutes after stimulation, and cells two hours after stimulation. A western blot was used to observe the relative amounts of PDGFR, Raf, and ERK, in their unphosphorylated and phosphorylated states, at the three different times. It was found that all three proteins are initially unphosphorylated, phosphorylated thirty minutes after stimulation, and less phosphorylated two hours after stimulation. This is consistent with what would be expected in an RTK-controlled Ras-ERK kinase pathway that is activated by PDGF. The protein players are initially phosphorylated in the

2 pathway, but are later dephosphorylated to stop signal transmission.

Introduction Extracellular signaling is one of the fundamental biological processes that allows life on earth to be possible. It is this process that allows cells to respond to their external environments. Understanding the mechanisms of signaling not only allows us to have a better understanding of cells and the way they interact with their surroundings, but many useful applications can result. To increase our understanding of signaling, we studied this process in a pure culture of fibroblasts that we isolated. By stimulating the cells with PDGF and taking samples after different amounts of time, we observed the changing levels of phosphorylated and unphosphorylated signaling proteins. Our results are consistent with what we would expect for the Ras-ERK pathway controlled by RTK (receptor tyrosine kinase) transmembrane protein. To better understand the signaling pathway that takes place when fibroblasts are stimulated by PDGF, we started out by isolating a pure culture of fibroblasts. This first involved selecting the antibody to be used in this process. Immunofluorescence and brightfield microscopy were used to observe how different antibodies stain skin tissue samples. We found out that anti-PDGFR (platelet-derived growth factor receptor) was specific to fibroblasts. We then used this antibody in an antibody-based cell sort, and observed the effectiveness of the sort using phase contrast microscopy. Fluorescence microscopy was then used to further probe the efficiency of the sort. After this we had a pure culture of fibroblasts for use in the final experiment. The fibroblasts were stimulated with PDGF, and samples were taken of unstimulated cells, cells thirty minutes

3 after stimulation, and cells two hours after stimulation. The samples were run on a gel and a western blot was done using antibodies against PDGFR, Raf, and ERK, in their phosphorylated and unphosphorylated forms. It was found that these three proteins tend to be unphosphorylated before stimulation, phosphorylated thirty minutes after stimulation, and less are phosphorylated two hours after stimulation. This fits the current understanding of the RTK-controlled Ras-ERK pathway, in which all of these proteins are initially phosphorylated, but are later dephosphorylated to stop signal transmission.

Materials and Methods Immunohistochemistry and Brightfield Microscopy: Five skin samples from pigs were sectioned, fixed, and treated with a peroxide block. While one of the samples was not treated with a primary antibody, the other four were treated with anti-EGFR (epidermal growth factor receptor), anti-PDGFR (platelet-derived growth factor receptor), anti-Cytokeratin, or anti-Collagen antibodies. The samples were then incubated with a secondary antibody, which was linked to biotin. This in turn was bound to a fusion protein of streptavidin linked to HRP (horse radish peroxidase) enzyme. After this the HRP substrate was added, which results in a red precipitate, and the nuclei were stained blue with hematoxylin. Lastly, mounting media was added directly to the sample area. Brightfield microscopy was used to observe these prepared slides at 40X magnification. Cell Sorting and Phase Contrast Microscopy: Intact skin tissue was treated with trypsin and then homogenized to create a cell suspension. This was inserted into an opticell with growth media. Trypsin/EDTA solution was used to remove cells from the

4 membrane. The cells were centrifuged at 2,500 rpm for ten minutes. The pellet was resuspended in DMEM/PBS solution. To isolate fibroblasts, anti-PDGFR antibodies were added as a primary antibody. This suspension was then injected into a new opticell, before adding secondary antibody bound to magnetic beads. A magnet was used to hold these beads and the cells bound to them in place, while the unattached cells were removed. Growth media was then added so the cells could continue to grow. This entire process starting after the initial addition of trypsin and the homogenization was done twice. Before and after the sort, a phase-contrast image was taken of the cell culture. Cell Counting: Before the second cell sort, 500 µl of suspension was removed for a cell count. First, 100 µl of trypan blue was added to the suspension. 10 µl of this was transferred to one chamber of the hemocytometer, where the numbers of white and blue cells were counted. Each 1 mm2 section of the hemocytometer represented a total volume of 1x10-4 ml. By counting the average number of white cells per 1 mm2 section, an estimate of the number of living cells per ml could be calculated. Immunofluorescense and Fluorescence Microscopy: The opticell was cut open and the cells were fixed with methanol. After blocking with PBS/1%BSA and staining the nuclei blue, a mixture of rabbit anti-PDGFR and goat anti-EGFR primary antibodies was added. The secondary antibody mixture was then added, which consisted of antirabbit antibodies conjugated to Fluor488, which appears green in fluorescence microscopy, and anti-goat conjugated to Fluor546, which appears red. These cells were then observed by fluorescence microscopy at 40X magnification. Gel Electrophoresis: Using the pure cultures of fibroblasts, some were stimulated by PDGF, and of these some were collected, washed and stored as cell pellets

5 after thirty minutes, while others were collected after two hours. Some cells were also left unstimulated. These three different samples were resuspended and added to SDSPAGE (sodium dodecyl sulfate-polyacrylamide gel electrophoresis) loading dye, and this was loaded into a 12% acrylamide gel. There were four groups of four lanes, and each group consisted of Precision Plus Protein standards marker in lane 1, unstimulated extract in lane 2, extract taken thirty minutes after stimulation in lane 3, and extract taken two hours after stimulation in lane 4. The gel was run for fifty minutes at 150 V. Western Blotting: When the gel was done running, it was placed in CAPS electroblotting buffer. A sandwich was prepared and the gel box was filled with CAPS buffer. The transfer was run at .200 amps for forty-five minutes, and then the membrane was placed in a blocking solution of BSA(2%)/TBS-Tween. There were four primary antibody solutions used: 1) rabbit anti-PDGFRβ, 2) rabbit anti-p-PDGFRβ (antiphosphorylated-PDGFRβ), 3) rabbit anti-Raf-1 and mouse anti-p-Raf-1, and 4) rabbit anti-ERK1/2, and mouse anti-p-ERK1/2. These antibodies were diluted to 1/1000 in BSA(2%)/TBS-Tween solution and added to the membranes. The secondary antibodies were anti-rabbit, which were conjugated to fluorophores that appeared green when scanned, and anti-mouse, which were conjugated to fluorophores that appeared red. A total of four western blot membranes were scanned.

Results We first determined that anti-PDGFR (platelet-derived growth factor receptor) antibody is good for isolating fibroblasts. To do this, five skin samples from research pigs were observed using brightfield microscopy (Figure 1). The nuclei in the samples

6 were stained blue with hematoxylin, and in each sample a different cell type was stained red using different primary antibodies and secondary antibodies conjugated to the HRP (horse radish peroxidase) enzyme. In each image the dermis, epidermis, and components of these two tissue types are visible. Figure 1a is the negative control, because the primary antibody did not specifically bind to any proteins in the skin. This is apparent because only one color of staining is present and it is staining the nuclei. The staining in the second image was done with anti-EGFR (epidermal growth factor receptor) antibody (Figure 1b). This is obvious because the red staining is localized to the epidermis, and is concentrated in the basal layer, where actual growth takes place. Figure 1c was stained with anti-Cytokeratin antibody, and it is the dark red staining in the outer layer of the epidermis is what explains this. In Figure 1d the dermis was stained with anti-Collagen antibody. Because collagen makes up most of the dermis, it makes sense that the red color is spread out throughout this tissue layer. Finally, Figure 1e was stained with anti-PDGFR antibodies. Because this stains fibroblasts, it can be seen concentrated near the fibroblasts in the dermis and near the inner layer of the epidermis. Therefore, we determined that anti-PDGFR antibody is the antibody best suited for use in the isolation of fibroblasts from intact skin tissue.

(a)

(b)

7

(c)

(d)

(e) Figure 1. Brightfield Microscopy of Skin Cells. The distinction between the dermis, epidermis, and components of these tissues is clear. In each image, cell nuclei are stained in blue and a different protein is stained in red using a different primary antibody. (a) Negative control. The antibody used did not recognize any proteins in the skin. (b) Staining of epidermis using anti-EGFR antibodies. (c) Staining of Keratinocytes with anti-Cytokeratin antibodies. (d) Staining of dermis with anti-Collagen antibodies. (e) Staining of fibroblasts with anti-PDGFR antibodies. Once it was determined that anti-PDGFR antibodies bind to the cell surface of fibroblasts, these antibodies were used to isolate fibroblasts from skin tissue. After degrading the connective tissue of the dermis, the resulting culture of cells was photographed using phase contrast microscopy (Figure 2a). In this mixture, the fibroblasts were the smaller cells with pointy edges, and the other skin cells could easily be distinguished by their different sizes and shapes. Using anti-PDGFR as a primary

8 antibody specific to fibroblasts and a secondary antibody bound to a magnetic bead, the fibroblasts were separated from the other cells by magnetic sorting. The resulting culture of pure fibroblasts is shown in Figure 2b. A hemocytometer was used to count that there were about 3.50x106 healthy cells per milliliter of solution. The healthy cells were distinguished from the unhealthy ones using trypan blue. Roughly 87% of the counted cells were white, which indicates that they were healthy. To further probe the efficiency of the cell sort, immunofluorescence was used. The nuclei were stained blue with DAPI, while a mixture of rabbit anti-PDGFR and goat

(a) (b) Figure 2. Phase Contrast Images of Skin Cell Culture Before and After Cell Sort. (a) Before a cell sort that selects for fibroblasts, it can be seen that there are several different types of cells. The smaller ones with pointy edges are fibroblasts, while the cells of other sizes and shapes are other skin cells. (b) After the cell sort, the only type of cells present is fibroblasts. anti-EGFR antibodies, which are specific for fibroblasts and epithelial cells, respectively, were added. Secondary antibodies conjugated to Fluor488 and Fluor546 stained the fibroblast membranes green and the epithelial membranes red, respectively. As a negative control, this process was first done to a culture of pure epithelial cells. Figure 3a shows this result, with the nuclei blue and all the membranes red. In contrast, Figure 3b shows the immunofluorescence image after the cell sort. In this case, there are no epithelial cells present. Instead, all the nuclei are surrounded by green membranes, which

9 indicates that these are all fibroblasts. In addition, the membranes have a shape that is characteristic of fibroblasts, in which there are spiky points. Therefore, the lack of epithelial cells suggests that this is a fairly pure culture of fibroblast cells that is ready for use in another experiment.

(a) (b) Figure 3. Immunofluorescence Images of Epithelial Cells and Skin Cell Culture After Cell Sort. The cell nuclei were stained blue with DAPI. The membranes were stained using a mixture of anti-PDGFR and anti-EGFR antibodies, and secondary antibodies conjugated to Fluor488 and Fluor546 made fibroblasts appear green and epithelial cells appear red, respectively. (a) As a negative control, a pure culture of epithelial cells was used in immunofluorescence microscopy. It can be seen that there are only cells with red membranes present, which are epithelial cells. (b) After the cell sort, only fibroblasts, with membranes that are stained green, are present.

After verifying by immunofluorescence that the culture of cells was only composed of fibroblasts, this pure culture was used in further experimentation. First, cell extracts were taken before the cells were treated in any way. Then the fibroblasts were stimulated with PDGF, and cells were collected thirty minutes and two hours after stimulation. Four different SDS-PAGE gels were loaded with marker in lane 1, extract before stimulation in lane 2, extract thirty minutes after stimulation in lane 3, and extract two hours after stimulation in lane 4 (Figure 4). Western blots were done such that each

10 set of extracts was stained with a different primary antibody. Figure 4a shows staining with anti-PDGFR antibody. It can be seen that before stimulation, there is a strong green band, which indicates the presence of unphosphorylated PDGFR. However, after thirty minutes, this band gets weaker, indicating the disappearance of this protein, and then after two hours, the band gets stronger again. Meanwhile, in Figure 4b, which shows staining with anti-p-PDGFR (phosphorylated-PDGFR) antibody, the band is weak at first and gets stronger after thirty minutes, indicating that more p-PDGFR is accumulating, but then it weakens again after two hours. Figure 4c shows a similar pattern when staining is with anti-Raf antibody. It illustrates the presence of Raf at first, its initial disappearance after stimulation with PDGF, and then reappearance a little later. When Raf disappears, Figure 4d shows that pRaf begins to appear, and when Raf appears, p-Raf disappears. The merge shows that both Raf and p-Raf are present after the initial stimulation, but more Raf returns at the two-hour mark. Figure 4f shows that ERK is present the whole time but especially before stimulation and two hours after stimulation. After stimulation, the amount p-ERK initially increases (Figure 4g). This is clarified in Figure 4h, which shows the merge. 1 2 3 4

1 2 3 4

(a) 1 2 3 4

(b) 1 2 3 4

1 2 3 4

(c) 1 2 3 4

1 2 3 4

(d)

1 2 3 4

(e)

11

P (f) (g) (h) Figure 4. Western Blot Analysis of Fibroblast Cells Treated with PDGF. In each western blot image, lane 1 contains the marker, lane 2 contains fibroblast extract of unstimulated cells, lane 3 contains extract thirty minutes after the cells were stimulated with PDGF, and lane 4 contains extract two hours after stimulation. (a) Staining with anti-PDGFR antibodies. (b) Staining with anti-p-PDGFR antibodies. (c) Staining with anti-Raf antibodies. (d) Staining with anti-p-Raf antibodies. (e) Merge of images (d) and (e). (f) Staining with anti-ERK antibodies. (g) Staining with anti-p-ERK antibodies. (h) Merge of images (f) and (g). The band is the most yellow, as opposed to green, thirty minutes after stimulation, which indicates that this in when the most p-ERK is present. Therefore, with all three proteins, they tend to be present in the unphosphorylated form before stimulation, the phosphorylated form thirty minutes after stimulation, and again in the unphosphorylated form after two hours.

Discussion In our first experiment, we used brightfield microscopy to determine that antiPDGFR (platelet-derived growth factor receptor) antibody is a good antibody for use in isolating fibroblasts. We then did a cell count and an antibody-based cell-sort to isolate fibroblasts. We confirmed that the sort was successful using phase contrast microscopy. To further probe the efficiency of the sort, we used fluorescence microscopy to observe the lack of epithelial cells. In the final experiment, we stimulated our pure culture of fibroblasts with PDGF and took samples after different amounts of time. By western

12 blotting, we observed the relative amounts of three different proteins in their phosphorylated and unphosphorylated forms. A similar pattern emerged for all three proteins. For PDGFR, Raf, and ERK, all three were present mainly in their unphosphorylated forms before stimulation. Thirty minutes after stimulation, all three of these tended to exist more in their phosphorylated forms, while two hours after stimulation, they were again unphosphorylated. It should be pointed out that in the western blot that shows the relative amounts of phosphorylated PDGFR, it is a little difficult to see the change in the strength of the band at the three different times. This is because the first band, from the unstimulated cells, is relatively strong. It is therefore more difficult to detect a change in band strength after thirty minutes and two hours. However, our interpretation was that the band gets stronger after thirty minutes and weaker after two hours. Our results showing the relative amounts of the three phosphorylated and unphosphorylated proteins are consistent with what is currently known about RTK (receptor tyrosine kinase) pathways that are controlled by PDGF. Figure 5 shows a typical RTK pathway that involves the proteins Ras and ERK. When cells in culture are stimulated with a growth factor, such as PDGF, this molecule binds to its receptors, such as PDGFR in this case. The PDGFR monomers in the cell membrane then dimerize and

13 PDGF membrane GDP PDGFR Monomers

PDGFR Dimer

Phosphorylated PDGFR Dimer

GDP Inactive     Disassociated MEK         Active Raf

GRB2    Sos      Inactive Ras

GTP Active Raf

GTP Active Ras            Inactive Raf

Active MEK      Inactive ERK               Active ERK

Figure 5. A Typical RTK Pathway. In a typical RTK pathway, such as one that involves PDGFR (platelet-derived growth factor receptor), PDGFR initially exists as transmembrane monomers. Upon binding of PDGF, the monomers dimerize and phosphorylate each other. This phosphorylated dimer can then bind GRB2, which binds Sos. This localizes Sos near the membrane so that it can bind to inactive Ras-GDP. This promotes dissociation of GDP and then binding of GTP, which produces Ras in its membrane-bound active form. Ras-GTP binds an inactive Raf and phosphorylates it. GTP hydrolysis leads to dissociation of Raf from Ras, and disassociated active Raf activates MEK by phosphorylating it. Active MEK in turn phosphorylates ERK so that it is in its active form. This can dimerize to enter the nucleus and activate many transcription factors. phosphorylate each other. This binds GRB2, which in turn binds Sos, which when recruited to the membrane converts Ras-GDP to its active form, Ras-GTP. This protein then can active a pathway in which Raf gets phosphorylated. In this active form, it activates MEK by phosphorylating it, which in turn activates ERK by phosphorylating it. ERK can then dimerize to enter the nucleus and activate many transcription factors. We

14 started our experiment in this same way, because we added PDGF to our cells in culture. It would make sense that this would bind to receptors and activate an RTK pathway. Our results are consistent with the Ras-ERK kinase pathway, because we started with relatively high levels of unphosphorylated PDGFR, Raf, and MEK, and low levels of these proteins in the phosphorylated state. It was only after the fibroblasts were stimulated with PDGF that the initial levels of these proteins in the phosphorylated state increased and the levels in the unphosphorylated state decreased. This fits in with what is known about RTK pathways, because all of these proteins get phosphorylated after being initially stimulated with PDGF. Our results also showed that after two hours, these three proteins were present less in the phosphorylated state and more in the unphosphorylated state. This is also consistent with what is known about RTK-controlled Ras-ERK kinase pathways. In living cells, proteins that get activated by being phosphorylated in a pathway must eventually be deactivated so that the signal is not transmitted forever. Therefore, proteins tend to eventually get dephosphorylated. It would make sense that this is what is happening to PDGFR, Raf, and ERK, and this is why the balance shifts from the phosphorylated to the unphosphorylated state. It is likely that PDGF activated the Ras-ERK pathway in the culture of fibroblasts that we studied. This explains the initial phosphorylation and then subsequent dephosphorylation of PDGFR, Raf, and ERK. Further experiments can be done to see which other proteins are involved in this pathway. A similar pattern of phosphorylation and dephosphorylation of the proteins being tested would indicate that the proteins are involved in the pathway. The experiment can also be done using more time points, which might provide information about the order in which the proteins get phosphorylated and

15 dephosphorylated. More information about this pathway can lead to many useful applications. One is that it can lead to cures for cancer, because PDGF can be a powerful mitogen. Research is also useful because PDGF is a very important factor in development, new blood vessel formation, and the healing of wounds. Addition experiments can also be done to study this pathway in types of cells other than fibroblasts, and other types of growth factors. The possibilities are endless and all information about these cellular processes can lead to countless applications.

Literature Cited Heldin, C. and A. Östman and L. Rönnstrand. Biochimica et Biophysica Acta 1378 (pp. F79-F113). Signal Transduction Via Platelet-derived Growth Factor Receptors. Elsevier Science, 1998. Young, B. and J. W. Heath. Skin, Chpt 9 (pp. 157-159, 162-164). Wheater’s Functional Histology. Harcourt Publishers, 2000.