Mitogenic Activity For Endothelial Cells Research Articles

Thrombospondin 1 as a scavenger for matrix-associated fibroblast growth factor 2

AbstractThe antiangiogenic factor thrombospondin 1 (TSP-1) binds with high affinity to several heparin-binding angiogenic factors, including fibroblast growth factor 2 (FGF-2), vascular endothelial growth factor (VEGF), and hepatocyte growth factor/scatter factor (HGF/SF). The aim of this study was to investigate whether TSP-1 affects FGF-2 association with the extracellular matrix (ECM) and its bioavailability. TSP-1 prevented the binding of free FGF-2 to endothelial cell ECM. It also promoted the mobilization of matrix-bound FGF-2, generating a TSP-1/FGF-2 complex. The region of TSP-1 responsible for these activities was located within the 140-kDa antiangiogenic and FGF-2 binding fragment, whereas the 25-kDa heparin-binding fragment was inactive. Matrix-released FGF-2/TSP-1 complex had a reduced ability to bind to and induce proliferation of endothelial cells. TSP-1 depleted the ECM laid by FGF-2-overproducing tumor cells of its FGF-2-dependent mitogenic activity for endothelial cells. Besides FGF-2, TSP-1 also inhibited VEGF and HGF/SF binding to the ECM and mobilized them from the ECM. Our study shows that TSP-1 acts as a scavenger for matrix-associated angiogenic factors, affecting their location, bioavailability, and function. (Blood. 2003; 102:4399-4406)

Covalently conjugated VEGF–fibrin matrices for endothelialization

Vascular endothelial growth factor (VEGF) is a key factor in endothelial cell biology and blood vessel formation and a candidate therapeutic for the stimulation of angiogenesis-dependent tissue regeneration. The objective of this study was to confer the angiogenic activity of VEGF 121 upon the biomaterial fibrin, a natural substrate for endothelial cell growth and clinically accepted as ‘fibrin glue’. To achieve this, we engineered fibrin-based hydrogels that were covalently modified with VEGF 121. Our laboratory has recently developed novel methodology that allows the covalent incorporation of exogenous bioactive peptides by the transglutaminase activity of factor XIIIa into fibrin during coagulation. Here, this ability of factor XIIIa to crosslink additional proteins within fibrin was employed to covalently incorporate VEGF 121. By recombinant DNA methodology, a mutant VEGF 121 variant, α 2-PI 1–8-VEGF 121, which contains an additional factor XIIIa substrate sequence NQEQVSPL at the aminoterminus, was expressed in E. coli. In soluble form, the mutant protein fully retained its mitogenic activity for endothelial cells. Using 125I-labeled α 2-PI 1–8-VEGF 121, its covalent incorporation and the efficiency of incorporation into fibrin was demonstrated and characterized. The immobilized, fibrin-conjugated VEGF 121 protein remained an active and very efficient mitogen for human endothelial cells grown on two-dimensional VEGF 121-modified fibrin surfaces, and the incorporation of increasing amounts of α 2-PI 1–8-VEGF 121 resulted in dose-dependent enhancement of endothelial cell growth. The VEGF-modified fibrin matrices can be formed as injectable gels in a single-step reaction under physiological conditions in vivo. When used as a ingrowth matrix, such VEGF incorporating materials could be useful in a variety of clinical situations that require an angiogenic response into an ischemic region or inplant.

Identification of Macrophage Migration Inhibitory Factor as a Potent Endothelial Cell Growth-Promoting Agent Released by Ectopic Human Endometrial Cells1

The growth of endometrial cells in ectopic locations (endometriosis) is dependent on the establishment of an adequate blood supply. Neovascularization (angiogenesis) is therefore a vital step toward the progression of this disease. We first revealed the presence of a potent mitogenic activity for human endothelial cells in the culture medium of immortalized human endometriotic cells. The activity, measured by the level of [(3)H]-thymidine incorporation into the DNA of human coronary artery endothelial cells, was then purified by anion exchange high-performance liquid chromatography. Electrophoretic analysis of one of the bioactive fractions that markedly enhanced endothelial cell proliferation showed three distinct bands with apparent molecular masses of 15.8, 12.6, and 6.5 kDa. N-terminal microsequencing of an internal peptide from the 12. 6-kDa protein showed 100% homology with human macrophage migration inhibitory factor (MIF). The protein was positively identified as MIF by Western blot analysis using a specific anti-MIF antibody. Anti-MIF antibody inhibited the bioactivity found in the evaluated fraction and the conditioned medium of primary endometriotic cell cultures, and commercial recombinant human MIF displayed a high mitogenic activity for endothelial cells. Our findings reveal that MIF is released by endometriotic cells and acts as a potent mitogenic factor for human endothelial cells in vitro. This may have a considerable interest, in view of the crucial role of angiogenesis in ectopic endometrial cell growth and activity and in numerous tissues undergoing dynamic physiological changes, such as human endometrium.

Neuropilin-1 Is Expressed by Endothelial and Tumor Cells as an Isoform-Specific Receptor for Vascular Endothelial Growth Factor

Vascular endothelial growth factor (VEGF), a major regulator of angiogenesis, binds to two receptor tyrosine kinases, KDR/Flk-1 and Flt-1. We now describe the purification and the expression cloning from tumor cells of a third VEGF receptor, one that binds VEGF 165 but not VEGF 121. This isoform-specific VEGF receptor (VEGF 165R) is identical to human neuropilin-1, a receptor for the collapsin/semaphorin family that mediates neuronal cell guidance. When coexpressed in cells with KDR, neuropilin-1 enhances the binding of VEGF 165 to KDR and VEGF 165-mediated chemotaxis. Conversely, inhibition of VEGF 165 binding to neuropilin-1 inhibits its binding to KDR and its mitogenic activity for endothelial cells. We propose that neuropilin-1 is a novel VEGF receptor that modulates VEGF binding to KDR and subsequent bioactivity and therefore may regulate VEGF-induced angiogenesis.

Production of heparin-binding endothelial mitogens by bovine uterine fibroblastic and epithelial cells

This study was conducted to determine whether early-passage cultures of bovine endometrial fibroblastic (Fb, n = 7 uteri) and epithelial (Ep, n = 3 uteri) cells produce endothelial mitogens in vitro and to begin characterization of these mitogens. Confluent cultures of Fb and Ep were incubated for 72 h in serum-free media, and the resulting conditioned media (CM) were evaluated for effects on proliferation of bovine aortic endothelial cells. CM from these Fb cultures (n = 8) and Ep cultures (n = 4) stimulated (147 +/- 10% (mean +/- SEM), p < 0.01, and 124 +/- 8%, p < 0.10, respectively) proliferation of endothelial cells compared with control (unconditioned) media. Most of the mitogenic activity of a sample of Fb CM and a sample of Ep CM from one individual uterus bound to heparin-agarose, and each exhibited two major peaks of activity that eluted at 0.9-1.0 and 1.7-1.8 M NaCl; the Fb CM also exhibited an additional heparin-binding peak eluting at 0-0.1 M NaCl. Pooled Fb CM (n = 8 cultures from 7 animals) also contained mitogenic activity for endothelial cells that bound to heparin-agarose, but exhibited three major peaks, eluting at 0.6, 1.1, and 1.8 M NaCl. Pooled Ep CM (n = 4 cultures from 3 animals) showed only one peak of mitogenic activity, which eluted at 0.9 M NaCl. Further characterization indicated that heat treatment reduced the activity of all heparin-binding Fb CM and Ep CM peaks, except the Fb CM peak eluting at 1.7 M NaCl. Trypsin reduced the activity of all peaks except one. Protein-A-purified antibody against fibroblast growth factor 1 (FGF-1) had no or only a slight effect on the mitogenic activity of the peaks. Mitogenic activity of the Fb CM peak eluting at 0.6 M NaCl was reduced by antibody against FGF-2. Activity of the Fb CM and Ep CM peaks that eluted at 1.7-1.8 M NaCl also was immunoneutralized by antibody to FGF-2. These data demonstrate that early passage cultures of endometrial Fb and Ep cells produce heparin-binding endothelial mitogens that appear to be immunologically related to FGF-2. These heparin-binding endothelial mitogens may influence endometrial vascular growth.

Production of heparin-binding endothelial mitogens by bovine uterine fibroblastic and epithelial cells

This study was conducted to determine whether early-passage cultures of bovine endometrial fibroblastic (Fb, n = 7 uteri) and epithelial (Ep, n = 3 uteri) cells produce endothelial mitogens in vitro and to begin characterization of these mitogens. Confluent cultures of Fb and Ep were incubated for 72 h in serum-free media, and the resulting conditioned media (CM) were evaluated for effects on proliferation of bovine aortic endothelial cells. CM from these Fb cultures (n = 8) and Ep cultures (n = 4) stimulated (147 +/- 10% (mean +/- SEM), p < 0.01, and 124 +/- 8%, p < 0.10, respectively) proliferation of endothelial cells compared with control (unconditioned) media. Most of the mitogenic activity of a sample of Fb CM and a sample of Ep CM from one individual uterus bound to heparin-agarose, and each exhibited two major peaks of activity that eluted at 0.9-1.0 and 1.7-1.8 M NaCl; the Fb CM also exhibited an additional heparin-binding peak eluting at 0-0.1 M NaCl. Pooled Fb CM (n = 8 cultures from 7 animals) also contained mitogenic activity for endothelial cells that bound to heparin-agarose, but exhibited three major peaks, eluting at 0.6, 1.1, and 1.8 M NaCl. Pooled Ep CM (n = 4 cultures from 3 animals) showed only one peak of mitogenic activity, which eluted at 0.9 M NaCl. Further characterization indicated that heat treatment reduced the activity of all heparin-binding Fb CM and Ep CM peaks, except the Fb CM peak eluting at 1.7 M NaCl. Trypsin reduced the activity of all peaks except one. Protein-A-purified antibody against fibroblast growth factor 1 (FGF-1) had no or only a slight effect on the mitogenic activity of the peaks. Mitogenic activity of the Fb CM peak eluting at 0.6 M NaCl was reduced by antibody against FGF-2. Activity of the Fb CM and Ep CM peaks that eluted at 1.7-1.8 M NaCl also was immunoneutralized by antibody to FGF-2. These data demonstrate that early passage cultures of endometrial Fb and Ep cells produce heparin-binding endothelial mitogens that appear to be immunologically related to FGF-2. These heparin-binding endothelial mitogens may influence endometrial vascular growth.

Effect of heparin on bovine epithelial lens cell proliferation induced by heparin affin regulatory peptide.

HARP (heparin affin regulatory peptide) is an 18 kDa heparin binding protein, also known as HB-GAM or pleiotrophin (PTN) which has been primarily isolated from brain and uterus, and displays neurite outgrowth, angiogenic and mitogenic activities. Previously, we have expressed the human cDNA encoding human HARP in NIH 3T3 cells. Purified recombinant HARP displayed mitogenic activity for endothelial cells. Its NH2-terminal sequence indicates that the HARP molecule possesses a three amino acid extension from the signal peptide more than the NH2-terminal described. For HB-GAM or PTN, these three amino acids may be essential for the stability and the mitogenic activity of this growth factor. In an attempt to further study the mode of action of this growth factor, we have investigated the mitogenic effect of HARP on various cell types. In contrast to FGF-2, HARP failed to induce stimulation of DNA synthesis on a CCL39 cell line. However, we found that in quiescent bovine epithelial lens (BEL) cells, the stimulation of DNA synthesis induced by HARP is dose-dependent (EC50: 2.5 ng/ml) and maximal stimulation is as potent as that induced by FGF-2 (EC50: 25 pg/ml). Interestingly, when BEL cells were allowed to quiesce in the presence of serum, the stimulation induced by HARP is considerably less potent. In this highly responsive cell system, heparin could potentiate the mitogenic activity of HARP at very low doses (0.1-1 microgram/ml) and inhibit this activity at concentrations of 10 micrograms/ml. In contrast to its protective effect on FGF-1 and -2, heparin was unable to preserve HARP from tryptic and chymotryptic degradations.

Initial characterization of mitogenic activity of ovine corpora lutea from early pregnancy.

To characterize angiogenic factors produced by ovine corpora lutea (CL) during early pregnancy, two experiments were performed. In Experiment 1, luteal explants from days 12, 18, 24, and 30 (n=4 ewes/day) after mating were incubated in serum-free medium for 6 h. Luteal-conditioned media (LCM) were evaluated for their ability to stimulate proliferation of endothelial and 3T3 cells, as well as migration of endothelial cells. Pools of the LCM (one pool/day) then were characterized biochemically. In Experiment 2, two pools of LCM from days 24 of pregnancy were evaluated for their effects on endothelial cell, 3T3 cell, and ovine luteal cell proliferation. These pools of LCM then were concentrated by ultrafiltration and subjected to heparin-agarose affinity chromatography with salt gradient (0-4 M NaCl in buffer) elution, and fractions were evaluated for mitogenic activity for endothelial and 3T3 cells. The resulting five peaks of mitogenic activity from heparin-agarose chromatography were characterized biochemically. The five peaks of mitogenic activity were further purified by using chromatography, then were concentrated and subjected to SDS-PAGE and Western analysis for FGF-2. Ovine CL from each day of early pregnancy secreted mitogens (P<0.05) for endothelial (285 +/- 8% of unconditioned media controls) and 3T3 (142 +/- 7%) cells as well as factors which stimulated migration of endothelial cell (153 +/- 8% of controls). LCM pool from day 24 of pregnancy also stimulated (P<0.05) proliferation of ovine luteal cells in a dose-dependent manner. In Experiment 1, mitogenic activity for endothelial cells was greater than 100 kDa, heat-labile, trypsin-sensitive and bound to DEAE-Sephacel and heparin-agarose columns, but not to a CM-Sepharose column. Antibody against FGF-1 did not affect mitogenic activity of LCM for endothelial and 3T3 cells, whereas treatment with FGF-2 antibody decreased (P<0.05) mitogenic activity of LCM for both endothelial and 3T3 cells. In Experiment 2, heparin-agarose affinity chromatography resolved five peaks of mitogenic activity: a non-heparin-binding peak that was specific for 3T3 cells, three heparin-binding peaks that were specific for endothelial cells, and one heparin-binding peak that was specific for 3T3 cells. In Experiment 2, heparin-, heat-, or trypsin-treatment and immunoneutralization with FGF-1 or FGF-2 antibodies influenced mitogenic activity of all of the peaks. Whereas SDS-PAGE demonstrated several bands of protein within each peak, Western analysis was unable to detect the presence of FGF-2 in any of the heparin-binding peaks. These data demonstrate that ovine CL from early pregnancy produce mitogenic factors that can be resolved into 5 separate peaks of activity with differing affinities for heparin. These data also indicate that the endothelial mitogens produced by CL of early pregnancy are immunologically related to, but biochemically distinct from FGF-2. Mitogens for endothelial and other cells likely play a role in regulation of luteal function during early pregnancy in sheep.

Tissue factor controls the balance of angiogenic and antiangiogenic properties of tumor cells in mice.

Meth-A sarcoma cells were stable transfected to overexpress (sense construct) or underexpress (antisense construct) tissue factor. In vitro, there was no difference in plating efficiency or growth between these cell lines. In vivo, tumor cells transfected to overexpress tissue factor grew more rapidly, and established larger and more vascularized tumors than control transfectants. Antisense transfectants grew the slowest and were the least vascularized. Anticoagulation of mice with warfarin did not alter the difference between these tumor lines. Tumor cells over-expressing tissue factor released more (compared with control transfectants) mitogenic activity for endothelial cells in parallel with enhanced transcription of vascular permeability factor/vascular endothelial cell growth factor (VEGF/VPF), and diminished transcription of thrombospondin (TSP2), a molecule with anti-angiogenic properties. Antisense tissue factor transfectants, while releasing the lowest amount of mitogenic activity, had increased thrombospondin and decreased VEGF/VPF transcription compared with control transfectants or wild-type cells. Experiments with these sense, antisense, truncated sense, or vector tumor lines gave comparable results in complete medium, serum free medium or in the presence of hirudin, indicating that the activation of the coagulation mechanism was not likely to be responsible for changes in tumor cell properties. These results suggest that tissue factor regulates angiogenic properties of tumor cells by altering the production of growth regulatory molecules of endothelium by a mechanism distinct from tissue factor activation of the coagulation mechanism.

Identification of basic fibroblast growth factor in papillary carcinoma of the thyroid.

Basic fibroblast growth factor (bFGF) was identified in the papillary carcinoma of the human thyroid. Immunohistochemically, it was found that the reactivity for bFGF was localized in the cytoplasm of the neoplastic cells of the five papillary carcinomas. However the extract of the papillary carcinomas contained the mitogenic activity for endothelial cells. This bioactive molecule was determined as bFGF by using the heparin-Sepharose affinity chromatography and western blot analysis. The bFGF derived from human thyroid papillary carcinoma and the recombinant human bFGF stimulated the bromodeoxyuridine incorporation by the cultured human thyroid papillary carcinoma cells. These cells also showed positive staining for thyroglobulin and cytokeratin. These results indicate that bFGF, probably produced by the neoplastic cells, plays an important role in the development of papillary carcinoma of the thyroid with stimulation of angiogenesis as well as proliferation of the parenchymal cells.

Production of mitogenic factor(s) by ovine corpora lutea throughout the estrous cycle.

To evaluate secretion of mitogenic factors by ovine corpora lutea (CL) at several stages of development, luteal explants from days 5 (n = 12 ewes), 10 (n = 6 ewes), and 15 (n = 6 ewes) of the estrous cycle were incubated in serum-free medium for 24 h. Luteal-conditioned media (LCM) were evaluated for their ability to stimulate proliferation of endothelial, BALB/3T3, and ovarian granulosa cells. After mitogenic activity of LCM from individual animals was evaluated, pools of LCM from each day of the estrous cycle were subjected to anion-exchange, cation-exchange, and heparin-affinity chromatography to characterize mitogenic activity. Pools of LCM also were utilized for ultrafiltration, heat-treatment, trypsin-treatment, and immunoneutralization studies. Results demonstrated that ovine CL secrete mitogenic activity that stimulates (P less than 0.01) proliferation of endothelial (135.7 +/- 5.3% of control) and granulosa cells (188.9 +/- 2.9%) but not 3T3 (103.2 +/- 2.5%) cells. Differences between stages of luteal development were not observed. The mitogenic activity bound to diethylaminoethyl-Sephacel and heparin-agarose, but not to carboxymethyl-Sepharose, indicating that ovine luteal mitogenic factor(s) is anionic and may belong to the heparin-binding growth factor (HBGF) family. In addition, the mitogenic activity was heat-labile, trypsin-sensitive, and appeared to have a M(r) greater than 100,000. Mitogenic activity for endothelial cells was partially neutralized with a specific antibody against HBGF-1 and was completely abolished with a specific antibody against HBGF-2. Moreover, HBGF-1 and HBGF-2 were immunolocalized in histological sections of CL from days 5 (n = 5 ewes), 10 (n = 5 ewes), and 15 (n = 5 ewes) after estrus. These findings are the first report of a major mitogenic factor(s) produced by cyclic ovine CL and indicate this factor is an HBGF-2-like molecule.

Production of mitogenic factor(s) by ovine corpora lutea throughout the estrous cycle

To evaluate secretion of mitogenic factors by ovine corpora lutea (CL) at several stages of development, luteal explants from days 5 (n = 12 ewes), 10 (n = 6 ewes), and 15 (n = 6 ewes) of the estrous cycle were incubated in serum-free medium for 24 h. Luteal-conditioned media (LCM) were evaluated for their ability to stimulate proliferation of endothelial, BALB/3T3, and ovarian granulosa cells. After mitogenic activity of LCM from individual animals was evaluated, pools of LCM from each day of the estrous cycle were subjected to anion-exchange, cation-exchange, and heparin-affinity chromatography to characterize mitogenic activity. Pools of LCM also were utilized for ultrafiltration, heat-treatment, trypsin-treatment, and immunoneutralization studies. Results demonstrated that ovine CL secrete mitogenic activity that stimulates (P less than 0.01) proliferation of endothelial (135.7 +/- 5.3% of control) and granulosa cells (188.9 +/- 2.9%) but not 3T3 (103.2 +/- 2.5%) cells. Differences between stages of luteal development were not observed. The mitogenic activity bound to diethylaminoethyl-Sephacel and heparin-agarose, but not to carboxymethyl-Sepharose, indicating that ovine luteal mitogenic factor(s) is anionic and may belong to the heparin-binding growth factor (HBGF) family. In addition, the mitogenic activity was heat-labile, trypsin-sensitive, and appeared to have a M(r) greater than 100,000. Mitogenic activity for endothelial cells was partially neutralized with a specific antibody against HBGF-1 and was completely abolished with a specific antibody against HBGF-2. Moreover, HBGF-1 and HBGF-2 were immunolocalized in histological sections of CL from days 5 (n = 5 ewes), 10 (n = 5 ewes), and 15 (n = 5 ewes) after estrus. These findings are the first report of a major mitogenic factor(s) produced by cyclic ovine CL and indicate this factor is an HBGF-2-like molecule.

Endothelial cell chemotactic factor derived from human glioma cell lines.

The authors report a study of the human umbilical vein endothelial cell chemotactic factor derived from human malignant glioma cell lines. The endothelial cell chemotactic activity of serum-free conditioned medium from cultures of U-373MG, U-251MG, or U-105MG cell lines was measured using a 48-well microchemotaxis chamber. The best response was from U-373MG, which was selected for further study. Chemotactic activity was contained in materials unadsorbed and adsorbed to the heparin-affinity column. Because the higher activity was seen in the unadsorbed material, it was used for characterization and partial isolation. The chemotactic activity was decreased under the condition of tumor protein synthesis inhibition. Heating, exposure to acid, and trypsin digestion also decreased the activity. The factor was found to be a protein with a relative molecular weight of greater than 200 kD; it has no mitogenic activity for endothelial cells in vitro and, partially purified, it was not identical to any other known endothelial cell chemotactic or mitogenic factor. Fibronectin was not detected, and anti-fibronectin antibody failed to inhibit the activity of the factor. These results suggest that malignant glioma cells produce a yet unknown endothelial cell chemotactic factor.

Tumor cell spheroids induce a mitogenic response in endothelial cells.

Conditioned media from spheroids of human colorectal tumor cell lines HT29, SW620, SW480 and CaCo-2 were tested for their mitogenic activity on human umbilical vein endothelial (HUVE) and bovine capillary endothelial (BCE) cells. All spheroid-conditioned media were capable of increasing the labeling indices in both endothelial cell populations, with the mitogenic activity decreasing in the following order: SW620, HT29/SW480 and CaCo-2. The endothelial cell mitogenic activity of the spheroids was not dependent on the presence of necrosis nor on their in vivo cytokinetic characteristics. We conclude that tumor spheroids release mitogenic activity for endothelial cells, but this stimulation depends on the age of the tumor spheroid as well as on the type of tumor used.

Production of platelet-derived endothelial cell growth factor by normal and transformed human cells in culture.

Platelet-derived endothelial cell growth factor (PD-ECGF) is a 45-kDa endothelial cell mitogen which has angiogenic properties in vivo. We report here that human foreskin fibroblasts, a human squamous cell carcinoma cell line, and 2 out of the 3 human thyroid carcinoma cell lines investigated produce PD-ECGF, whereas 21 other cell lines examined do not. The positive cell lines contained a 1.8-kilobase PD-ECGF mRNA, and a 45-kDa protein could be demonstrated in lysates of the cell lines by immunoblotting and immunoprecipitation using a specific antiserum against PD-ECGF. Furthermore, the cell lysates contained mitogenic activity for endothelial cells that was neutralized by the PD-ECGF antiserum. PD-ECGF was found to be secreted only slowly from the producer cells, consistent with the previous finding that the primary translation product lacks a signal sequence. The restricted expression and intracellular sequestration of PD-ECGF imply a strictly controlled function in endothelial cell proliferation and angiogenesis. Aberrant production of PD-ECGF may play a role in tumor angiogenesis.

Release of an endothelial cell growth factor from cultured porcine thyroid follicles.

It has been proposed from in vivo studies that thyroid angiogenesis during thyroid enlargement may be due to paracrine mitogenic factors released by epithelial thyroid cells. To study this paracrine growth regulating communication between thyroid cells and endothelial cells in vitro, culture medium from isolated porcine thyroid follicles was investigated for a growth promoting effect on porcine aortal endothelial cells. Serum-free conditioned medium (CM) from thyroid follicles in suspension culture contains a dose-related mitogenic activity which stimulates endothelial cell growth up to 197%. Stimulation of the thyroid follicles with TSH (1 mU/ml) significantly reduced the mitogenic activity for endothelial cells in CM to 131%. Thyroid hormones had no influence on mitogenic activity in CM. When follicles were treated with iodide (20 microM) during CM production, no proliferation of endothelial cells was observed by this CM. In contrast, CM from epidermal growth factor-treated thyroid follicles significantly enhanced the mitogenic activity for endothelial cells up to 235%. The mitogenic activity was precipitable by saturated ammonium sulfate, showed high affinity to heparin by chromatography on heparin-sepharose, and was abolished after treatment of CM with trypsin. On gel electrophoresis the heparin-binding fraction showed a double band with a mol wt of 15 and 15.5 k. These data show a paracrine mitogenic activity on endothelial cells released by thyroid follicles which is regulated by TSH, epidermal growth factor, and iodide in parallel with the direct effect of these substances on thyroid cell growth. The data suggest that the mitogenic factor is a polypeptide, which belongs to the heparin-binding growth factors.

Pituitary follicular cells produce basic fibroblast growth factor.

Cultured monolayers of bovine pituitary follicular cells, which transport ions, contain high amounts of mitogenic activity for endothelial cells which, on the basis of gene expression analysis, heparin-Sepharose elution profile, bioassay, immunoblotting, radioimmunoassay, and radioreceptor assay, has been identified as basic fibroblast growth factor (bFGF). These data indicate that follicular cells may be a major source of bFGF in the pituitary gland. Considering that bFGF has been proposed to play a role in paracrine regulation of pituitary hormone secretion, the data also suggest that these cells may exert important local regulatory functions.