0021-972X/90/7001 -0149$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1990 by The Endocrine Society
Vol. 70, No. 1 Printed in U.S.A.
M. B. ZIMERING, M. L. BRANDI, D. A. D E G R A N G E , S. J. MARX, E. STREETEN, N. KATSUMATA, P. R. MURPHY, Y. SATO, H. G. FRIESEN, AND G. D. AURBACH Metabolic Diseases Branch, National Institutes of Diabetes, Digestive and Kidney Diseases, National Institutes of Health (M.B.Z., M.L.B., D.A.d., S.J.M., E.S., G.D.A.), Bethesda, Maryland 20892; and the Department of Physiology, University of Manitoba Faculty of Medicine (N.K., P.R.M., Y.S., H.G.F.), Winnipeg, Manitoba, Canada R3E 0W3
(FMENl plasma) HS-adsorbed activity eluted with 0.1-0.3 M NaCl, was completely neutralized by specific antibodies against bFGF, and had an apparent mol wt of 110 kD. Active fractions from FMENl plasma prepared by gel filtration in 7 M urea displayed apparent mol wt of about 14-16 kD and showed increased apparent affinity for HS; recovered activity appeared principally in the 3.0-M NaCl eluate. Using a sensitive two-site immunoradiometric assay for bFGF we found 0.4 ng/mL bFGFlike immunoreactivity in the highly purified 3.0-M NaCl eluate from a HS column to which the active components from gel filtration of FMENl plasma in 7 M urea were applied. These results imply that bFGF or closely related factors circulate in FMENl. {J Clin Endocrinol Metab 70: 149,1990)
ABSTRACT. Basic fibroblast growth factor (bFGF) is a potent endothelial cell mitogen found in a variety of normal and tumor tissues. Basic FGF lacks a classical signal sequence, and it is not clear how it is released from cells. bFGF or bFGF-like activity has not been previously demonstrated in plasma. In an earlier study we showed increased mitogenic activity for parathyroidderived epithelial and mesenchymal cells in plasma of subjects with familial multiple endocrine neoplasia type 1 (FMEN1). In the present study we examined the growth-promoting activity of normal and FMENl plasmas [applied to heparin-Sepharose (HS) columns] in parathyroid-derived cloned endothelial cells. FMENl plasma HS-adsorbed activity exceeded normal plasma HS-adsorbed activity in 6 of 8 FMENl plasma samples. Peak
B
ASIC fibroblast growth factor (bFGF) is a potent endothelial cell mitogen found in bovine capillary endothelial cells in vitro (1) and isolated from a variety of normal tissues in vivo, including bovine pituitary, brain, retina, placenta, liver, adrenal, and corpus luteum (2-8). FGFs have also been found in conditioned medium from several tumor cell lines in vitro (9, 10), and a substance resembling bFGF was purified to homogeneity from the extracellular matrix of a rat chondrosarcoma (11). A distinguishing property useful in purification of FGFs is their high affinity for immobilized heparin (11); acidic FGF (aFGF elutes from heparin-Sepharose (HS) columns with 0.8-1.2 M NaCl, and bFGF elutes from HS columns with 1.7-3.0 M NaCl. Despite the presence of abundant cellular stores and extracellular matrix-associated FGFs, the amino acid sequence of bFGF (and aFGF) lacks a conventional signal sequence (12), and bFGF has not been found in normal plasma (13). Familial multiple endocrine neoplasia type 1 (FMENl)
is an autosomal dominant syndrome (14) characterized by hyperplasia and tumors of parathyroid, pancreatic islet, and anterior pituitary glands. We found earlier (15) that a mitogen circulating in FMENl stimulates in vitro growth of mixed cultures of epithelial and mesenchymal cells of parathyroid origin. A clone of parathyroid-derived endothelial cells [bovine parathyroid endothelial (BPE)] was established, characterized,1 and found to be responsive to one or more growth-promoting factors in FMENl plasma (Brandi, M. L., Ornberg, R. L., Sakaguchi, K., Curcio, F., Fattorossi, A., Lelkes, P. I., Matsui, T., Zimering, M. B., and Aurbach, G. D., submitted manuscript). Using HS affinity chromatography and gel filtration we purified an endothelial cell growth factor about 40,000-fold from one FMENl subject's plasma. We report here that a circulating factor(s) in FMENl may be related to bFGF. To our knowledge, this is the first report of bFGF-like bioactivity in human plasma. 1 BPE were developed and characterized as previously described (Brandi, M. L., submitted manuscript). These cells expressed factor VHI-related antigen, took up acetylated low density lipoproteins, formed capillary-like structures in long term culture, expressed bFGF mRNA, displayed diploid chromosome number, grew in strict monolayer, and had mophological features typical of other capillary endothelial cells.
Received June 20,1989. Address all correspondence and requests for reprints to Dr. M. B. Zimering, Building 10, Room 9C101, National Institutes of Health, Bethesda, Maryland 20892. * This work was supported in part by the NCI of Canada and the Medical Research Council of Canada.
149
Downloaded from https://academic.oup.com/jcem/article-abstract/70/1/149/2652230 by Rudolph Matas Medical Library user on 17 January 2019
Circulating Fibroblast Growth Factor-Like Substance in Familial Multiple Endocrine Neoplasia Type 1*
ZIMERING ET AL.
150
Subjects and Methods Subjects
was aspirated from wells after 24 h and replaced with medium A without 2% Ultraser-G or 10% Nu-Serum (medium B). Test factors were then added to cells growing in medium B. Growth assays
Informed consent for plasmapheresis was obtained from each subject. After an overnight fast, plasma was collected between 0800-1100 h in citrate by plasmapheresis (Haemonetics V-50, Braintree, MA) at 4,800 rpm and centrifuged at 20,000 X g for 30 min before ammonium sulfate precipitation. FMENl activity was stable for up to 3 yr at -20 C, and fresh or frozen-
After 96-h incubation in the presence of test fractions, cells were washed with Ham's F-12 without phenol red and then incubated at 37 C for 1 h with Hoescht dye 33342 at 10 Mg/mL in Ham's F-12 without phenol red. Medium was discarded, and chromatin-associated fluorescence was determined automatically in a fluorescence microtiter plate reader (Dyna-Tech, Inc., Chantilly, VA) as previously described (16). Fluorescence was a linear function of (BPE) cell number, between 5,000-120,000 cells/well, as shown previously for a variety of other cell types (16). The slope of the dose-response curve was 10 fluorescent units/5000 cells; the midrange of the assay was 40,000 cells =
thawed plasma from the same FMENl subject had similar
80 fluorescent units; the least detectable concentration was
activity.
10,000 cells = 20 fluorescent units. Growth-promoting activity is expressed as a percentage of the control (basal) fluorescence for cells grown in medium without test factors. Simultaneous assays of fractions from HS affinity chromatography of normal and FMENl plasma showed that changes in fluorescence paralleled results of mitogenic activity estimated by [3H]thymidine uptake. The slope of the dose-response curve for plasma samples ranged from 25-45%/log dilution for 6 FMENl plasmas and from 5-15%/log dilution for 6 normal plasmas. The maximal detectable dilution of FMENl plasma from 1 subject that significantly increased growth was 1:1000. Each point represents the mean of sextuplicate determinations, i.e. 1 column of 6 wells/96-well plate. The intraassay coefficient of variation was less than 5%. The interassay coefficient of variation was less than 15%. Standard dose-response curves using synthetic intact bFGF (R&D Systems, St. Paul, MN) showed a linear range of 0.01-10 ng/mL (113 ± 5% to 173 ± 5%), the ED50 was 1 ng/mL (141 ± 5%), the least detectable concentration was 0.01 ng/mL, and the slope was a 30% change in basal fluorescence/log concentration of bFGF.
Plasma samples
Cell culture BPE cells were grown in Coon's modified Ham's F-12 medium containing 1 mM calcium, 0.5 mM magnesium, 2 mM glutamine, 200 mg/L D-galactose, 0.7 mg/L thymidine, 2% Ultraser-G (IBF, Biotechnic, Inc., Cedex, France), and 10% Nu-Serum (Collaborative Research, Inc., Bedford, MA); hereafter referred to as medium A. BPE cells were trypsinized when they reached confluency and plated at a density of 1 X 104 cells/well using a multichannel pipetter in the inner 10 X 6 grid of 96-well plates in medium A. The first and last columns of wells in each grid served as control or basal; test fractions were added to the remaining 8 columns of wells, 1 kind of test fraction/column of wells. For the experiments described in Fig. 1 and Table 1 test fractions were added to cells growing in medium A. For all other experiments reported here medium A
Polyclonal antibodies
FRACTION NUMBER
FIG. 1. HS affinity chromatography of plasma fractions from representative normal (a) and FMENl (b) subjects. The pellet of 30-65% ammonium sulfate fractionations of plasma in 10 mM Na-phosphate buffer, pH 7.4, was applied to HS columns as described in Materials and Methods. One twentieth dilutions of individual fractions were assayed for growth promotion as described in Materials and Methods.
Antiserum to bFGF-(1-24) generated as outlined previously (17) was supplied by Dr. Andrew Baird (Whittier Institute, La Jolla, CA). The antiserum showed less than 1% cross-reactivity with up to 1 Mg/mL each of platelet-derived growth factor, insulin-like growth factor-I, epidermal growth factor, transforming growth factor-/3 (TGF/3), or aFGF in a liquid phase RIA. Intact bFGF gave a competition curve parallel to that bFGF-(1-24) in the RIA. The immunoglobulin G (IgG) fraction obtained after 3-fold concentration of the antiserum by ammonium sulfate precipitation and protein-A affinity chromatography, anti-bFGF-(l-24), was stored at -70 C and used in the experiments described here. Control antibodies were rabbit antihuman IgG obtained from Kierkegaard and Perry, Inc. (Gaithersburg, MD). Two-site IRMA of bFGF Antiserum to bFGF-(l-lO) was generated as previously described (10). Assay materials and procedures were previously described (18, 19).
Downloaded from https://academic.oup.com/jcem/article-abstract/70/1/149/2652230 by Rudolph Matas Medical Library user on 17 January 2019
FMEN1 subjects (three males and five females) were randomly selected and ranged in age from 25-56 yr (mean age, 38 yr). Normal subjects (four males and two females) ranged in age from 32-56 yr (mean age, 40 yr).
JCE & M • 1990 Vol70«Nol
CIRCULATING bFGF-LIKE SUBSTANCE IN FMENl
151
TABLE 1. Growth-promoting activity in heparin eluates of plasma from FMENl and normal subjects Average peak activity
Average flow-through activity
Average HS -adsorbed activity
141 ± 7d 116 ± 9
HS affinity chromatography of plasma fractions The pellet from 30-65% ammonium sulfate fractionations of 30 mL plasma in 10 mM Na-phosphate buffer, pH 7.4, was applied to a HS column (bed volume, 30 mL) equilibrated with 10 mM Na-phosphate buffer, pH 7.4. The column was washed with the same buffer and eluted with a gradient of the same buffer containing equal volumes of 0.0 or 2.0 M NaCl. The flow rate was 30 mL/h. Individual fractions and pools of flow through fractions (no. 3-28) or HS-adsorbed fractions (no. 75100) were dialyzed in Spectrapor tubing [mol wt (MW),
Vol. 70, No. 1 Printed in U.S.A.
M. B. ZIMERING, M. L. BRANDI, D. A. D E G R A N G E , S. J. MARX, E. STREETEN, N. KATSUMATA, P. R. MURPHY, Y. SATO, H. G. FRIESEN, AND G. D. AURBACH Metabolic Diseases Branch, National Institutes of Diabetes, Digestive and Kidney Diseases, National Institutes of Health (M.B.Z., M.L.B., D.A.d., S.J.M., E.S., G.D.A.), Bethesda, Maryland 20892; and the Department of Physiology, University of Manitoba Faculty of Medicine (N.K., P.R.M., Y.S., H.G.F.), Winnipeg, Manitoba, Canada R3E 0W3
(FMENl plasma) HS-adsorbed activity eluted with 0.1-0.3 M NaCl, was completely neutralized by specific antibodies against bFGF, and had an apparent mol wt of 110 kD. Active fractions from FMENl plasma prepared by gel filtration in 7 M urea displayed apparent mol wt of about 14-16 kD and showed increased apparent affinity for HS; recovered activity appeared principally in the 3.0-M NaCl eluate. Using a sensitive two-site immunoradiometric assay for bFGF we found 0.4 ng/mL bFGFlike immunoreactivity in the highly purified 3.0-M NaCl eluate from a HS column to which the active components from gel filtration of FMENl plasma in 7 M urea were applied. These results imply that bFGF or closely related factors circulate in FMENl. {J Clin Endocrinol Metab 70: 149,1990)
ABSTRACT. Basic fibroblast growth factor (bFGF) is a potent endothelial cell mitogen found in a variety of normal and tumor tissues. Basic FGF lacks a classical signal sequence, and it is not clear how it is released from cells. bFGF or bFGF-like activity has not been previously demonstrated in plasma. In an earlier study we showed increased mitogenic activity for parathyroidderived epithelial and mesenchymal cells in plasma of subjects with familial multiple endocrine neoplasia type 1 (FMEN1). In the present study we examined the growth-promoting activity of normal and FMENl plasmas [applied to heparin-Sepharose (HS) columns] in parathyroid-derived cloned endothelial cells. FMENl plasma HS-adsorbed activity exceeded normal plasma HS-adsorbed activity in 6 of 8 FMENl plasma samples. Peak
B
ASIC fibroblast growth factor (bFGF) is a potent endothelial cell mitogen found in bovine capillary endothelial cells in vitro (1) and isolated from a variety of normal tissues in vivo, including bovine pituitary, brain, retina, placenta, liver, adrenal, and corpus luteum (2-8). FGFs have also been found in conditioned medium from several tumor cell lines in vitro (9, 10), and a substance resembling bFGF was purified to homogeneity from the extracellular matrix of a rat chondrosarcoma (11). A distinguishing property useful in purification of FGFs is their high affinity for immobilized heparin (11); acidic FGF (aFGF elutes from heparin-Sepharose (HS) columns with 0.8-1.2 M NaCl, and bFGF elutes from HS columns with 1.7-3.0 M NaCl. Despite the presence of abundant cellular stores and extracellular matrix-associated FGFs, the amino acid sequence of bFGF (and aFGF) lacks a conventional signal sequence (12), and bFGF has not been found in normal plasma (13). Familial multiple endocrine neoplasia type 1 (FMENl)
is an autosomal dominant syndrome (14) characterized by hyperplasia and tumors of parathyroid, pancreatic islet, and anterior pituitary glands. We found earlier (15) that a mitogen circulating in FMENl stimulates in vitro growth of mixed cultures of epithelial and mesenchymal cells of parathyroid origin. A clone of parathyroid-derived endothelial cells [bovine parathyroid endothelial (BPE)] was established, characterized,1 and found to be responsive to one or more growth-promoting factors in FMENl plasma (Brandi, M. L., Ornberg, R. L., Sakaguchi, K., Curcio, F., Fattorossi, A., Lelkes, P. I., Matsui, T., Zimering, M. B., and Aurbach, G. D., submitted manuscript). Using HS affinity chromatography and gel filtration we purified an endothelial cell growth factor about 40,000-fold from one FMENl subject's plasma. We report here that a circulating factor(s) in FMENl may be related to bFGF. To our knowledge, this is the first report of bFGF-like bioactivity in human plasma. 1 BPE were developed and characterized as previously described (Brandi, M. L., submitted manuscript). These cells expressed factor VHI-related antigen, took up acetylated low density lipoproteins, formed capillary-like structures in long term culture, expressed bFGF mRNA, displayed diploid chromosome number, grew in strict monolayer, and had mophological features typical of other capillary endothelial cells.
Received June 20,1989. Address all correspondence and requests for reprints to Dr. M. B. Zimering, Building 10, Room 9C101, National Institutes of Health, Bethesda, Maryland 20892. * This work was supported in part by the NCI of Canada and the Medical Research Council of Canada.
149
Downloaded from https://academic.oup.com/jcem/article-abstract/70/1/149/2652230 by Rudolph Matas Medical Library user on 17 January 2019
Circulating Fibroblast Growth Factor-Like Substance in Familial Multiple Endocrine Neoplasia Type 1*
ZIMERING ET AL.
150
Subjects and Methods Subjects
was aspirated from wells after 24 h and replaced with medium A without 2% Ultraser-G or 10% Nu-Serum (medium B). Test factors were then added to cells growing in medium B. Growth assays
Informed consent for plasmapheresis was obtained from each subject. After an overnight fast, plasma was collected between 0800-1100 h in citrate by plasmapheresis (Haemonetics V-50, Braintree, MA) at 4,800 rpm and centrifuged at 20,000 X g for 30 min before ammonium sulfate precipitation. FMENl activity was stable for up to 3 yr at -20 C, and fresh or frozen-
After 96-h incubation in the presence of test fractions, cells were washed with Ham's F-12 without phenol red and then incubated at 37 C for 1 h with Hoescht dye 33342 at 10 Mg/mL in Ham's F-12 without phenol red. Medium was discarded, and chromatin-associated fluorescence was determined automatically in a fluorescence microtiter plate reader (Dyna-Tech, Inc., Chantilly, VA) as previously described (16). Fluorescence was a linear function of (BPE) cell number, between 5,000-120,000 cells/well, as shown previously for a variety of other cell types (16). The slope of the dose-response curve was 10 fluorescent units/5000 cells; the midrange of the assay was 40,000 cells =
thawed plasma from the same FMENl subject had similar
80 fluorescent units; the least detectable concentration was
activity.
10,000 cells = 20 fluorescent units. Growth-promoting activity is expressed as a percentage of the control (basal) fluorescence for cells grown in medium without test factors. Simultaneous assays of fractions from HS affinity chromatography of normal and FMENl plasma showed that changes in fluorescence paralleled results of mitogenic activity estimated by [3H]thymidine uptake. The slope of the dose-response curve for plasma samples ranged from 25-45%/log dilution for 6 FMENl plasmas and from 5-15%/log dilution for 6 normal plasmas. The maximal detectable dilution of FMENl plasma from 1 subject that significantly increased growth was 1:1000. Each point represents the mean of sextuplicate determinations, i.e. 1 column of 6 wells/96-well plate. The intraassay coefficient of variation was less than 5%. The interassay coefficient of variation was less than 15%. Standard dose-response curves using synthetic intact bFGF (R&D Systems, St. Paul, MN) showed a linear range of 0.01-10 ng/mL (113 ± 5% to 173 ± 5%), the ED50 was 1 ng/mL (141 ± 5%), the least detectable concentration was 0.01 ng/mL, and the slope was a 30% change in basal fluorescence/log concentration of bFGF.
Plasma samples
Cell culture BPE cells were grown in Coon's modified Ham's F-12 medium containing 1 mM calcium, 0.5 mM magnesium, 2 mM glutamine, 200 mg/L D-galactose, 0.7 mg/L thymidine, 2% Ultraser-G (IBF, Biotechnic, Inc., Cedex, France), and 10% Nu-Serum (Collaborative Research, Inc., Bedford, MA); hereafter referred to as medium A. BPE cells were trypsinized when they reached confluency and plated at a density of 1 X 104 cells/well using a multichannel pipetter in the inner 10 X 6 grid of 96-well plates in medium A. The first and last columns of wells in each grid served as control or basal; test fractions were added to the remaining 8 columns of wells, 1 kind of test fraction/column of wells. For the experiments described in Fig. 1 and Table 1 test fractions were added to cells growing in medium A. For all other experiments reported here medium A
Polyclonal antibodies
FRACTION NUMBER
FIG. 1. HS affinity chromatography of plasma fractions from representative normal (a) and FMENl (b) subjects. The pellet of 30-65% ammonium sulfate fractionations of plasma in 10 mM Na-phosphate buffer, pH 7.4, was applied to HS columns as described in Materials and Methods. One twentieth dilutions of individual fractions were assayed for growth promotion as described in Materials and Methods.
Antiserum to bFGF-(1-24) generated as outlined previously (17) was supplied by Dr. Andrew Baird (Whittier Institute, La Jolla, CA). The antiserum showed less than 1% cross-reactivity with up to 1 Mg/mL each of platelet-derived growth factor, insulin-like growth factor-I, epidermal growth factor, transforming growth factor-/3 (TGF/3), or aFGF in a liquid phase RIA. Intact bFGF gave a competition curve parallel to that bFGF-(1-24) in the RIA. The immunoglobulin G (IgG) fraction obtained after 3-fold concentration of the antiserum by ammonium sulfate precipitation and protein-A affinity chromatography, anti-bFGF-(l-24), was stored at -70 C and used in the experiments described here. Control antibodies were rabbit antihuman IgG obtained from Kierkegaard and Perry, Inc. (Gaithersburg, MD). Two-site IRMA of bFGF Antiserum to bFGF-(l-lO) was generated as previously described (10). Assay materials and procedures were previously described (18, 19).
Downloaded from https://academic.oup.com/jcem/article-abstract/70/1/149/2652230 by Rudolph Matas Medical Library user on 17 January 2019
FMEN1 subjects (three males and five females) were randomly selected and ranged in age from 25-56 yr (mean age, 38 yr). Normal subjects (four males and two females) ranged in age from 32-56 yr (mean age, 40 yr).
JCE & M • 1990 Vol70«Nol
CIRCULATING bFGF-LIKE SUBSTANCE IN FMENl
151
TABLE 1. Growth-promoting activity in heparin eluates of plasma from FMENl and normal subjects Average peak activity
Average flow-through activity
Average HS -adsorbed activity
141 ± 7d 116 ± 9
HS affinity chromatography of plasma fractions The pellet from 30-65% ammonium sulfate fractionations of 30 mL plasma in 10 mM Na-phosphate buffer, pH 7.4, was applied to a HS column (bed volume, 30 mL) equilibrated with 10 mM Na-phosphate buffer, pH 7.4. The column was washed with the same buffer and eluted with a gradient of the same buffer containing equal volumes of 0.0 or 2.0 M NaCl. The flow rate was 30 mL/h. Individual fractions and pools of flow through fractions (no. 3-28) or HS-adsorbed fractions (no. 75100) were dialyzed in Spectrapor tubing [mol wt (MW),
