JOURNAL OF CELLULAR PHYSIOLOGY 147:17-26 (1991)
The Basic Fibroblast Growth Factor-Saporin Mitotoxin Acts Through the Basic Fibroblast Growth Factor Receptor DOUGLAS A. LAPPI,* PAMELA A. MAHER, DARLENE MARTINEAU, AND ANDREW BAIRD Department of Molecular and Cellular Growth Biology, The Whittier lnstitute for Diabetes and Endocrinology, 9894 Cenesee Avenue, l a jolla, California 92037 W e have confirmed the hypothesis that a mitotoxin resulting from the conjugation of basic fibroblast growth factor and saporin exerts its cytotoxic effect through specific interaction with the basic fibroblast growth factor (FGF) receptor. Accordingly, the mitotoxin stimulates tyrosine phosphorylation of the 90 kD substrate that characterizes the initial cellular response to basic FGF. Cross-linking experiments show that radio-labeled basic fibroblast growth factor-saporin (FGFSAP) binds to the receptor. Suramin, an inhibitor of growth factor receptor binding, inhibits the cytotoxicity of basic FGF-SAP. In a study of 4 different cell types, there is a decrease in the ED, of the mitotoxin as the receptor number per cell increases. We have verified the cytotoxicity of the mitotoxin in 3 different assay systems. As expected, it is effective in the inhibition of protein synthesis and DNA synthesis, as well as of cell count. Binding of basic FGF-SAP which will result in cytotoxicity occurs very rapidly: 5 minutes of incubation of 10 n M basic FGF-SAP with cells results in 80% inhibition of cell count. The in vitro data indicate that basic FGF-SAP is a receptor specific and potent suicide antagonist of basic FGF. Its potential as an anti-FGF for therapeutic and research uses in vivo is discussed.
Basic fibroblast growth factor (FGF) is a mitogen for both mesodermal and neuroectodermal cells including fibroblasts, vascular and corneal endothelial cells, myoblasts, glial cells, and chondrocytes, among others (for a review, see Baird and Walicke, 1989).It also promotes angiogenesis and neuronal survival. Its primary sequence has been determined and it is part of a family of growth factors that includes acidic FGF, hstlkFGF, int-2, FGF-5, FGF-6 and KGF (Esch et al., 1986; Taira et al., 1987; Delli Bovi et al., 1987; Zhan et al., 1988; Marics et al., 1989; Finch et al., 1989). Several forms of the basic FGF receptor have recently been characterized. Lee et al. (1989) report that the form they have characterized from a chicken cDNA library is a transmembrane protein that contains 3 extracellular immunoglobulin-like domains and an intracellular protein kinase domain. Pasquale !1990), also working with a chicken library, has identified proteins that are homologous with the protein described by Lee et al. that may be receptors for basic FGF-related molecules. Johnson et al. (1990) have described 4 unique human cDNA clones that encode previously undescribed forms of the FGF receptor. These forms have divergence in the regions of the extracellular domain. One clone was isolated from a library of placental cDNA; the other 3 from a library of human umbilical vein endothelial cells in culture. Thus, there appears to be differential expression of the receptor in different systems. Basic FGF has been sug ested to be involved in a number of proliferative iseases and pathological
f
0 1991 WILEY-LISS. INC.
states, including endometriosis (Baird et al., 1986a), diabetic retinopathy (Baird et al., 1985) and tumor growth (Folkman, 1975). Therefore, antagonists for basic FGF are of interest as possible preventative agents for some pathophysiological states. Neutralization of endogenous basic FGF by passive immunization can partially inhibit tumor growth (Baird et al., 1986b1, indicating considerable therapeutic potential for factors capable of preventing basic FGF activity. Furthermore, specific antagonists could be useful for the determination of the physiological role that basic FGF plays in vivo in development (Vlodavsky et al., 19791, as a neurotrophic factor (Walicke et al., 1986) and as a morphogen (Slack et al., 1987; Slack and Isaacs, 1989). We have previously communicated the characterization (Lappi et al., 1989) of a bifunctional protein resulting from the chemical conjugation of basic FGF and saporin (SAP), a powerful ribosome-inactivating protein from the plant Saponaria officinalis (Lappi et al., 1985;Stirpe et al., 1983).This mitotoxin, referred to as basic FGF-SAP, retains the individual activities of both SAP and basic FGF. In a cell-free rabbit reticulocyte protein synthesis system, the mitotoxin inhibits protein synthesis at a level only slightly less than free *To whom reprint requestsicorrespondence should be addressed. Received July 30, 1990; accepted December 13, 1990. Portions of this material were presented at the 11th American Peptide Symposium, San Diego, California, July, 1989.
18
LAPP1 ET AL
SAP. By virtue of the conjugation to basic FGF, however, the mitotoxin possesses a potent cytotoxic activity to cells which possess the basic FGF receptor. Competition studies showed that basic FGF-SAP was able to compete radiolabeled basic FGF with an affinity similar to basic FGF and that an excess of free basic FGF, but not NGF, could inhibit the cytotoxicity of the mitotoxin. We have examined the interaction of the mitotoxin with the basic FGF receptor in an effort to further characterize the activity of the mitotoxin. We confirm here that the mitotoxin acts by binding t o the basic FGF receptor and that through this interaction is cytotoxic to target cells.
EXPERIMENTAL PROCEDURES Reagents L3H1-leucine (specific activity 153 Ciimmol) and [3H]thymidine (specific activity 81.9 Ciimmol) were purchased from Dupont New England Nuclear (Wilmington, DE). L12511-sodium iodide (specific activity 1317 mCiip, ) was purchased from Amersham (Arlington Heights, L). Basic FGF was a kind gift of Dr. Paolo Sarmientos, Farmitalia Carlo Erba (Milan, Italy).
f
Basic FGF-SAP Basic FGF-SAP was synthesized as described in Lappi et al. (1989) with modifications: a 1.8-fold molar excess of SAP, derivatized with N-succinimidyl3(pyridyldithio)proprionate (Pharmacia, Uppsala, Sweden), was used for reaction with basic FGF. After completion of the reaction (2-3 hr), as monitored by an increase in A343nm, the reaction mixture was diluted %fold with water and purified by FPLC (Pharmacia) using a heparin-Sepharose column (0.5 x 5 cm) and salt step gradient. The column was washed with 0.6 M, 1.0 M and 2.0 M NaCl in 10 mM TRIS-HC1, pH 7.2. The 2 M NaCl eluate contained basic FGF-SAP and was dialyzed versus Dulbecco’sPBS. Purity of the conjugate was assured by polyacrylamide gel electrophoresis and, due to the excess of derivatized SAP in the reaction mixture, found to be reproducibly free of free basic FGF. For cross-linking experiments, the conjugate was further purified by Mono S (Pharmacia) chromatography, as described (Lambert et al., 1985).This material showed no free basic FGF in Western blot analysis. Preparation of the lZ5Iderivative of basic FGF-SAP and basic FGF was as described by Feige and Baird (1988). Cell lines Bovine corneal endothelial and adult bovine adrenal endothelial cells were obtained from primary culture according to the method of Gospodarowiczet al. (1986). Baby hamster kidney and 3T3 cells were obtained from the American Type Culture Collection (Rockville,MD). DC215 cell line was a generous gift of Dr. Robert Fox, Department of Immunology, Scripps Clinic and Research Foundation, La Jolla, CA. PFHR9 cells were a kind gift from Dr. Eva Engvall of the La Jolla Cancer Research Foundation, La Jolla, CA. Biological assay of cytotoxic activity Cells were treated with basic FGF-SAP and with control solutions of basic FGF alone, SAP alone and an
equimolar mixture of unconjugated basic FGF and SAP. 1) Cell counting assay. Cell counting assay was performed as described by Lappi et al. (1989) with the following exceptions: PFHR9, 3T3, and the endothelial cells were plated at a concentration of 10,000 per well and incubated for 72 hr, (instead of 48) in the presence of various agents before counting. 2) Protein synthesis inhibition assay. BHK cells in 100 pL of growth media (5% calf serum-HDMEM with F-12, Gibco, Gaithersberg, MD) were added to the wells of a 96-well plate at a concentration of 1000 cells per well. Sixteen hours later, samples were added in triplicate and the plates incubated for 48 hr. 3H-leucine (1 KCi) was then added to the wells for 3 hr. The media were removed from the wells and replaced with trypsinEDTA. When the cells detached, they were transferred to 3-mm filter pa ers. They were precipitated with trichloroacetic aci(Pand the incorporated radioactivity determined by scintillation counting. Control (untreated) counts er min (cpm) mean was 207 ? 329; cpm for basic F8F-SAP treated wells at 10 nM was 81.7 t 6.5. 3) Thymidine incorporation inhibition assay. BHK cells were plated and treated with sample as in the protein synthesis inhibition assay. After 48 hr, 3H-thymidine (0.2 WCiiwell) was added to the wells for 3 hr and the plates were frozen overnight. The plates were then thawed and well contents were collected with a Skatron cell harvester onto glass fiber filter sheets and radioactivity incorporation determined by scintillation counting. Control CPM mean was 116,063 t 10,544. CPM for basic FGF-SAP treated wells at 10 nM were 288 k 125. DC215 cells were plated at 2000 cells/ well in wells of a 96-well plate. Sixteen hours and 3 days after, the wells were washed and treatments were added. Ninety-six hours after plating, 200 pCi of 3Hthymidine were added and incubated for 17 hr. The plate was frozen and the next day thawed and harvested by a cell harvester and counted by liquid scintillation techniques. Cross-linkingof basic FGF-SAP to the receptor Cross-linking was performed by Feige and Baird (1988), with minor modifications. BHK cells were plated in 10-cm plates in DMEM:F-12,l:l plus 5% calf serum, gentamycin (Gibco, Grand Island, NY, 50 mg/L) and fungizone (Gibco, 250 pg1L) overnight to reach a subconfluent concentration. Cells were washed 3 times with F-12 medium with 0.2% gelatin and 25 mM HEPES, pH 7.4 (binding buffer) and then incubated with 1 x lo6 cpm of 1251-basi~ FGF in binding buffer (pre-cooled to 4°C) for 2-4 hr on ice. Cells were then washed 3 times with cold binding buffer and then incubated with 0.15 mM disuccinimidyl suberate (Pierce, Rockford, IL) in 10 mM sodium phosphate, 0.15 M NaC1, 1mM EDTA, pH 7.4 (PBS-EDTA) at room temperature for 15 min. The cross-linking reaction was stopped with a wash of 200 mM ethanolamine in PBS-EDTA for 1min. The stop buffer was aspirated off and the plates were washed 2 times with washing buffer (PBS-EDTA with 1mM phenylmethylsulfonylfluoride, leupeptin (1 pg/mL), pepstatin (10 pgimL) and aprotinin (1.7 mUimL) (all protease inhibitors
BASIC FGF-SAPORIN MITOTOXIN
from Sigma, St. Louis, MO) at 4°C. The washing buffer was aspirated and the cells were scraped from the plate with rubber policemen with eppendorf tubes, using a minimum (
The Basic Fibroblast Growth Factor-Saporin Mitotoxin Acts Through the Basic Fibroblast Growth Factor Receptor DOUGLAS A. LAPPI,* PAMELA A. MAHER, DARLENE MARTINEAU, AND ANDREW BAIRD Department of Molecular and Cellular Growth Biology, The Whittier lnstitute for Diabetes and Endocrinology, 9894 Cenesee Avenue, l a jolla, California 92037 W e have confirmed the hypothesis that a mitotoxin resulting from the conjugation of basic fibroblast growth factor and saporin exerts its cytotoxic effect through specific interaction with the basic fibroblast growth factor (FGF) receptor. Accordingly, the mitotoxin stimulates tyrosine phosphorylation of the 90 kD substrate that characterizes the initial cellular response to basic FGF. Cross-linking experiments show that radio-labeled basic fibroblast growth factor-saporin (FGFSAP) binds to the receptor. Suramin, an inhibitor of growth factor receptor binding, inhibits the cytotoxicity of basic FGF-SAP. In a study of 4 different cell types, there is a decrease in the ED, of the mitotoxin as the receptor number per cell increases. We have verified the cytotoxicity of the mitotoxin in 3 different assay systems. As expected, it is effective in the inhibition of protein synthesis and DNA synthesis, as well as of cell count. Binding of basic FGF-SAP which will result in cytotoxicity occurs very rapidly: 5 minutes of incubation of 10 n M basic FGF-SAP with cells results in 80% inhibition of cell count. The in vitro data indicate that basic FGF-SAP is a receptor specific and potent suicide antagonist of basic FGF. Its potential as an anti-FGF for therapeutic and research uses in vivo is discussed.
Basic fibroblast growth factor (FGF) is a mitogen for both mesodermal and neuroectodermal cells including fibroblasts, vascular and corneal endothelial cells, myoblasts, glial cells, and chondrocytes, among others (for a review, see Baird and Walicke, 1989).It also promotes angiogenesis and neuronal survival. Its primary sequence has been determined and it is part of a family of growth factors that includes acidic FGF, hstlkFGF, int-2, FGF-5, FGF-6 and KGF (Esch et al., 1986; Taira et al., 1987; Delli Bovi et al., 1987; Zhan et al., 1988; Marics et al., 1989; Finch et al., 1989). Several forms of the basic FGF receptor have recently been characterized. Lee et al. (1989) report that the form they have characterized from a chicken cDNA library is a transmembrane protein that contains 3 extracellular immunoglobulin-like domains and an intracellular protein kinase domain. Pasquale !1990), also working with a chicken library, has identified proteins that are homologous with the protein described by Lee et al. that may be receptors for basic FGF-related molecules. Johnson et al. (1990) have described 4 unique human cDNA clones that encode previously undescribed forms of the FGF receptor. These forms have divergence in the regions of the extracellular domain. One clone was isolated from a library of placental cDNA; the other 3 from a library of human umbilical vein endothelial cells in culture. Thus, there appears to be differential expression of the receptor in different systems. Basic FGF has been sug ested to be involved in a number of proliferative iseases and pathological
f
0 1991 WILEY-LISS. INC.
states, including endometriosis (Baird et al., 1986a), diabetic retinopathy (Baird et al., 1985) and tumor growth (Folkman, 1975). Therefore, antagonists for basic FGF are of interest as possible preventative agents for some pathophysiological states. Neutralization of endogenous basic FGF by passive immunization can partially inhibit tumor growth (Baird et al., 1986b1, indicating considerable therapeutic potential for factors capable of preventing basic FGF activity. Furthermore, specific antagonists could be useful for the determination of the physiological role that basic FGF plays in vivo in development (Vlodavsky et al., 19791, as a neurotrophic factor (Walicke et al., 1986) and as a morphogen (Slack et al., 1987; Slack and Isaacs, 1989). We have previously communicated the characterization (Lappi et al., 1989) of a bifunctional protein resulting from the chemical conjugation of basic FGF and saporin (SAP), a powerful ribosome-inactivating protein from the plant Saponaria officinalis (Lappi et al., 1985;Stirpe et al., 1983).This mitotoxin, referred to as basic FGF-SAP, retains the individual activities of both SAP and basic FGF. In a cell-free rabbit reticulocyte protein synthesis system, the mitotoxin inhibits protein synthesis at a level only slightly less than free *To whom reprint requestsicorrespondence should be addressed. Received July 30, 1990; accepted December 13, 1990. Portions of this material were presented at the 11th American Peptide Symposium, San Diego, California, July, 1989.
18
LAPP1 ET AL
SAP. By virtue of the conjugation to basic FGF, however, the mitotoxin possesses a potent cytotoxic activity to cells which possess the basic FGF receptor. Competition studies showed that basic FGF-SAP was able to compete radiolabeled basic FGF with an affinity similar to basic FGF and that an excess of free basic FGF, but not NGF, could inhibit the cytotoxicity of the mitotoxin. We have examined the interaction of the mitotoxin with the basic FGF receptor in an effort to further characterize the activity of the mitotoxin. We confirm here that the mitotoxin acts by binding t o the basic FGF receptor and that through this interaction is cytotoxic to target cells.
EXPERIMENTAL PROCEDURES Reagents L3H1-leucine (specific activity 153 Ciimmol) and [3H]thymidine (specific activity 81.9 Ciimmol) were purchased from Dupont New England Nuclear (Wilmington, DE). L12511-sodium iodide (specific activity 1317 mCiip, ) was purchased from Amersham (Arlington Heights, L). Basic FGF was a kind gift of Dr. Paolo Sarmientos, Farmitalia Carlo Erba (Milan, Italy).
f
Basic FGF-SAP Basic FGF-SAP was synthesized as described in Lappi et al. (1989) with modifications: a 1.8-fold molar excess of SAP, derivatized with N-succinimidyl3(pyridyldithio)proprionate (Pharmacia, Uppsala, Sweden), was used for reaction with basic FGF. After completion of the reaction (2-3 hr), as monitored by an increase in A343nm, the reaction mixture was diluted %fold with water and purified by FPLC (Pharmacia) using a heparin-Sepharose column (0.5 x 5 cm) and salt step gradient. The column was washed with 0.6 M, 1.0 M and 2.0 M NaCl in 10 mM TRIS-HC1, pH 7.2. The 2 M NaCl eluate contained basic FGF-SAP and was dialyzed versus Dulbecco’sPBS. Purity of the conjugate was assured by polyacrylamide gel electrophoresis and, due to the excess of derivatized SAP in the reaction mixture, found to be reproducibly free of free basic FGF. For cross-linking experiments, the conjugate was further purified by Mono S (Pharmacia) chromatography, as described (Lambert et al., 1985).This material showed no free basic FGF in Western blot analysis. Preparation of the lZ5Iderivative of basic FGF-SAP and basic FGF was as described by Feige and Baird (1988). Cell lines Bovine corneal endothelial and adult bovine adrenal endothelial cells were obtained from primary culture according to the method of Gospodarowiczet al. (1986). Baby hamster kidney and 3T3 cells were obtained from the American Type Culture Collection (Rockville,MD). DC215 cell line was a generous gift of Dr. Robert Fox, Department of Immunology, Scripps Clinic and Research Foundation, La Jolla, CA. PFHR9 cells were a kind gift from Dr. Eva Engvall of the La Jolla Cancer Research Foundation, La Jolla, CA. Biological assay of cytotoxic activity Cells were treated with basic FGF-SAP and with control solutions of basic FGF alone, SAP alone and an
equimolar mixture of unconjugated basic FGF and SAP. 1) Cell counting assay. Cell counting assay was performed as described by Lappi et al. (1989) with the following exceptions: PFHR9, 3T3, and the endothelial cells were plated at a concentration of 10,000 per well and incubated for 72 hr, (instead of 48) in the presence of various agents before counting. 2) Protein synthesis inhibition assay. BHK cells in 100 pL of growth media (5% calf serum-HDMEM with F-12, Gibco, Gaithersberg, MD) were added to the wells of a 96-well plate at a concentration of 1000 cells per well. Sixteen hours later, samples were added in triplicate and the plates incubated for 48 hr. 3H-leucine (1 KCi) was then added to the wells for 3 hr. The media were removed from the wells and replaced with trypsinEDTA. When the cells detached, they were transferred to 3-mm filter pa ers. They were precipitated with trichloroacetic aci(Pand the incorporated radioactivity determined by scintillation counting. Control (untreated) counts er min (cpm) mean was 207 ? 329; cpm for basic F8F-SAP treated wells at 10 nM was 81.7 t 6.5. 3) Thymidine incorporation inhibition assay. BHK cells were plated and treated with sample as in the protein synthesis inhibition assay. After 48 hr, 3H-thymidine (0.2 WCiiwell) was added to the wells for 3 hr and the plates were frozen overnight. The plates were then thawed and well contents were collected with a Skatron cell harvester onto glass fiber filter sheets and radioactivity incorporation determined by scintillation counting. Control CPM mean was 116,063 t 10,544. CPM for basic FGF-SAP treated wells at 10 nM were 288 k 125. DC215 cells were plated at 2000 cells/ well in wells of a 96-well plate. Sixteen hours and 3 days after, the wells were washed and treatments were added. Ninety-six hours after plating, 200 pCi of 3Hthymidine were added and incubated for 17 hr. The plate was frozen and the next day thawed and harvested by a cell harvester and counted by liquid scintillation techniques. Cross-linkingof basic FGF-SAP to the receptor Cross-linking was performed by Feige and Baird (1988), with minor modifications. BHK cells were plated in 10-cm plates in DMEM:F-12,l:l plus 5% calf serum, gentamycin (Gibco, Grand Island, NY, 50 mg/L) and fungizone (Gibco, 250 pg1L) overnight to reach a subconfluent concentration. Cells were washed 3 times with F-12 medium with 0.2% gelatin and 25 mM HEPES, pH 7.4 (binding buffer) and then incubated with 1 x lo6 cpm of 1251-basi~ FGF in binding buffer (pre-cooled to 4°C) for 2-4 hr on ice. Cells were then washed 3 times with cold binding buffer and then incubated with 0.15 mM disuccinimidyl suberate (Pierce, Rockford, IL) in 10 mM sodium phosphate, 0.15 M NaC1, 1mM EDTA, pH 7.4 (PBS-EDTA) at room temperature for 15 min. The cross-linking reaction was stopped with a wash of 200 mM ethanolamine in PBS-EDTA for 1min. The stop buffer was aspirated off and the plates were washed 2 times with washing buffer (PBS-EDTA with 1mM phenylmethylsulfonylfluoride, leupeptin (1 pg/mL), pepstatin (10 pgimL) and aprotinin (1.7 mUimL) (all protease inhibitors
BASIC FGF-SAPORIN MITOTOXIN
from Sigma, St. Louis, MO) at 4°C. The washing buffer was aspirated and the cells were scraped from the plate with rubber policemen with eppendorf tubes, using a minimum (
