- Email: [email protected]
A Growth Factor in Bovine and Human Testes Structurally Related to Basic Fibroblast Growth Factor
0022-5347 /88/1402-0422$02.00/0 Vol. 140, August Printed in U.S.A.
THE JOURNAL OF UROLOGY
Copyright© 1988 by The Williams & Wilkins Co.
A GROWTH FACTOR IN BOVINE AND HUMAN TESTES STRUCTURALLY RELATED TO BASIC FIBROBLAST GROWTH FACTOR MICHAEL T. STORY,* JOACHIM SASSE, DANIEL KAKUSKA, STEPHEN C. JACOBS RUSSELL K. LAWSON
AND
From the Medical College of Wisconsin, Departments of Urology and Biochemistry, Milwaukee, Wisconsin and Harvard Medical School, Department of Surgery, Boston, Massachusetts
ABSTRACT
Homogenates of human testes, epididymides and prostate, and calf testes and epididymides are mitogenic for cultured human foreskin fibroblasts. The growth factors appear similar in that they are inactivated by boiling and acid, but not by treatment with reducing agent. The growth factor in human and bovine testes was partially purified from tissue homogenates, prepared in high ionic strength buffer (pH 7.6) containing protease inhibitors, by ammonium sulfate precipitation and two cycles of heparin-Sepharose chromatography. The growth factor in calf testes was also partially purified from tissue extracted in ammonium sulfate without protease inhibitors, acidified to pH 4.5, and precipitated by ammonium sulfate followed by two cycles of heparin-affinity chromatography. A predominant 17,500 molecular weight (MW) growth factor was identified from alkaline homogenates of human and calf testes by its reactivity with antisera prepared against synthetic peptides whose sequences corresponded to residues 1-12 (amino-terminal), 33-43 (internal) and 136-145 (carboxy-terminal) of bovine basic fibroblast growth factor (bFGF). A slightly smaller 16,600 MW peptide from acidic extracts of calf testes also reacted with antisera to the three synthetic peptides. A 15,500 MW peptide, lacking immunoreactivity with antiserum to the amino-terminal synthetic peptide, was also seen. These findings suggest that a growth factor is present in human and calf testes that is structurally related to bFGF. The structure of the growth factors appears to be altered during the isolation procedure. (J. Ural., 140: 422-427, 1988) Jacobs and associates 1 first reported the presence of a growthpromoting factor in extracts of human benign prostatic hyperplasia (BPH) in 1979. The presence of prostatic growth factor (PrGF) in BPH has been confirmed by several investigators.2-5 While comparing the physical chemical and biological properties of PrGF with those of other growth factors, it became apparent that PrGF was similar to a growth factor called seminiferous growth factor (SGF) in the seminiferous epithelium of the calf and several mammalian species. 6 • 7 Recently, we have purified the growth factor from BPH tissue 8 and demonstrated by amino-terminal sequencing that PrGF is structurally related to basic fibroblast growth factor, bFGF. 9 Now we report immunologic data that SGF in calf and human testes is also related to bFGF. MATERIALS AND METHODS
Materials. Human prostate tissue was from surgical specimens obtained at transurethral (TURP) and open prostatectomy for BPH. Human testes and epididymides were from patients, mean 80 years of age, who underwent androgen ablation for prostate cancer. Calf testes and epididymides were obtained fresh from a local meat processing plant. Tris (hydroxymethyl) aminomethane (Tris), diamethylformamide, phenylmethanesulfonyl fluoride, L-1 (tosylamido)-2-phenyl-ethyl chloromethyl ketone, N-ethylmalemide, guanidine-hydrochloride, soybean trypsin inhibitor and keyhole limpet hemocyanin, KLH, were from Sigma (St. Louis, MO). N-hydroxysuccinimidyl maleimidohexanoate ester (MHS) was from Boehringer Mannheim Accepted for publication March 15, 1988. * Requests for reprints: Dept. of Urology, Medical College of Wisconsin, 9200 West Wisconsin Ave., Milwaukee, WI 53226. Supported by NIH grant AM31063. 422
Biochemicals, Indianapolis, IN. Disodium ethylenediaminetetraacetate (EDTA) and glacial acetic acid were from JT Baker Chemical Co. (Phillipsburg, NJ). Ammonium sulfate and sodium chloride were from Mallinckrodt (Paris, KY). All tissue culture medium, antibiotics and serum were from Grand Island Biological Company (Grand Island, NY) and all culture flasks and multiwell plates were from Falcon (Becton Dickinson Labware, Oxnard, CA). Econofluor scintillation cocktail, Protosol tissue solubolizer and (Methyl-3H)-thymidine were from New England Nuclear (Boston, MA). All reagents for SDSPAGE were from LKB (Gaithersburg, MD). Molecular Weight (MW) markers were from Bethesda Research Laboratories (Gaithersburg, MD). Protein was assayed with a kit with BSA as standard (Bio-Rad, Richmond, CA). Cell culture and assay for growth factor activity. The bioassay was based on the ability of a sample to stimulate (3H)-thymidine uptake by quiescent confluent monolayers of human foreskin fibroblasts as described elsewhere. 3 The amount of radioactivity was determined with a model 1219 liquid scintillation counter (LKB) interfaced with an Apple II E for conversion of cpm to dpm. The mean of triplicate samples was determined and a Relative Mitogenic Activity (RMA) was calculated: RMA
= DPM
(Test Sample) - DPM (Buffer) DPM (10% NBS) - DPM (Buffer)
The buffer was the same as in the test sample. One unit of activity was defined as a RMA equal to one, that is, a response equal to that produced by 10% newborn bovine serum (NBS). Tissue homogenization and ammonium sulfate-precipitation. Tissue was transported in saline in an ice bath. The tissue was trimmed free of fat and connective tissue. The epididymidis was separated from the testis, and the tunica albuginea re-
423 ffl0V8d, 'fiS8U88 V\T8r€ Stored frozen ,ncmcc,cco, 0Il8 tO three hurnan testes and enicti,dv:rrn,ctes one to six months; calf testes one to six weeks; calf epididymides two to four months, All steps in preparing tissue homogenates and in isolating growth factor were at OC to 4C. The tissues were thawed, rinsed in saline and minced with scissors. The tissue was weighed and homogenized with a PT 10-35 Polytron equipped with a PT 20ST probe generator (Brinkman Instruments, Inc., Westbury, NY) at a setting of eight for 30 sec at 10 sec intervals with 30 sec cooling between homogenizations. In the "alkaline extraction" procedure, the homogenization buffer (1:2 tissue wet weight: volume), was 50 mM Tris, 1.55 M NaCl pH 7.6 containing EDTA (10 mM), phenylmethanesulfonyl fluoride (1 mM), L-l(tosylamido)-2-phenylethyl chloromethyl ketone (0.3 mM), N-ethylmaleimide (0.05 mM), soybean trypsin inhibitor (10 mg./L). The homogenate was centrifuged at 23,000 g for 40 min. in a Sorvall model RC-2B centrifuge (DuPont Instruments, Newton, CT) and the supernatant was transferred to a clean beaker. An equal volume of buffer was added to the pellet and the homogenization and centrifugation was repeated. The supernatant from the two centrifugations was pooled and filtered through a Metrigard glass fiber membrane (Gelman Sciences, Inc., Ann Arbor, MI) with positive nitrogen pressure. A portion of the filtrate was dialyzed against 50 mM Tris, 0.15 M NaCl pH. 7.6 (TBS) in bags made of Spectra/Por3 tubing, 3,500 MW cutoff (Spectrum Medical Industries, Los Angeles, CA). A precipitate was removed by centrifugation and the supernatant was sterilized by filtration (0.2 µ Acrodisc membranes, Gelman) and assayed for activity and protein. The remaining filtrate was brought to 25 % ammonium sulfate the slow addition of saturated ammonium sulfate, pH 7.6. After mixing two hours, the preparation was centrifuged at 23,000 g for 30 min. The pellet was discarded and the supernatant was brought to 75% ammonium sulfate. After mixing and centrifuging, the pellet was dissolved in TBS and dialyzed against the same buffer. The preparation was centrifuged, Metrigard filtered and a portion was sterilized by filtration and assayed for activity and protein. In the "acidic extraction" procedure, Polytron homogenization was in 0.15 M ammonium sulfate. The remainder of the procedure was from Gospodarowicz, et aL 10 The pH of the homogenate was centrifuged, and the pH of the supernatant was adjusted to 7,0 with one N NaOH. Ammonium sulfate (230 gm.fl.; final concentration, 1.9 M) was added and the precipitate was removed after one hour mixing by centrifugation. Further addition of ammonium sulfate (250 gm.fl. final concentration 3.8 to the supernatant gave a precipitate that was collected after one hour mixing by centrifugation. The itate was dissolved in water and dialyzed overnight against water and then overnight against TBS. _ .,.- .,, •• , studies. Tissue hom,)1,JmE1te:s, 50 mM Tris, 50 mM NaCl, pH were buffer (control and boiled samples), 0,1 M acetic acid, 3 M guanidine-hydrochloride pH 7.6 or 10% 2-mercaptoethanol pH 7.6. After 24 hours at 4C or five min. samples were centrifuged and the supernatant was against Tris buffer, filtered (0.2 µ) and tested for activity, Heparin-Sepharose chromatography. The ammonium sulfate precipitated growth factor was applied to a five cm, X two cm,, 40 ml. bed volume column of heparin-Sepharose (Pharmacia, Piscataway, NJ) equilibrated with TBS. The column was washed with TBS until the absorbance of the eluate at 206 nm. became negligible. The absorbed proteins were then eluted with a gradient (480 ml.) formed from TBS and 50 mM Tris with 3.0 M NaCl. The conductivity of fractions was determined with a conductivity bridge (Yellow Springs Instrument Co., Yellow Spring, OH). A portion of each fraction was diluted in Minimum Essential Medium (Eagle) with 0.5 NBS, sterile filtered and assayed for activity. The growth factor eluting between 1.2 M and 1.8 M NaCl was concentrated by ultrafiltration, YM5 au,nn,HUo
ty-pe men1brane in a stirred cell Amicon (Danvers, and dialyzed agaist 50 mM with 0,25 M NaCl pH 7.6 (calf testes) or 50 mM Tris with 0,5 M NaCl pH 7.6 (human testes). The growth factor recovered from the 40 ml. heparin-affinity column was applied to a 1.6 cm. X 7.5 cm., 15 ml. bed volume column of heparin-Sepharose equilibrated with 50 mM Tris with 0.25 NaCl pH 7.6 (calf testes) or 50 mM Tris with 0.5 M NaCl (human testes). The column was washed with the equilibration buffer until the absorbency of the eluate reached baseline. The absorbed proteins were then eluted with a gradient (280 ml.) formed from 50 mM Tris with 0.25 M NaCl (calf testes) or 50 mM Tris with 0.5 M NaCl (human testes) and 50 mM Tris with 3 M NaCL The conductivity and activity of fractions were measured, Fractions with growth factor were concentrated by ultrafiltration and the salt content was lowered by the addition of 30 volumes of 50 mM Tris with 50 mM NaCl and the sample was reconcentrated. A portion of the sample was analyzed by analytical SDS-PAGE. SDS-PAGE and electroblotting. Analytical SDS-PAGE was performed on a vertical slab gel apparatus (LKB) using 10% to 22.5% linear gradient gels prepared by standard techniques" in the presence of thiol reducing agent, Molecular weight markers were insulin (A and B chains), 2,300 and 3,400; bovine trypsin inhibitor, 6,200; lysozyme, 14,300; /)-lactoglobulin, 18,400; a chymotrypsinogen, 25,700 and ovalbumin, 43,000, Some gels were silver stained. 12 Other gels were electroblotted with a Transphor Unit (LKB) to nitrocellulose paper, 0,2 µ pore size (Schleicher and Schuell, Keene, NH). The transfer was performed in 25 mM Tris-HCl, 0.15 M glycine, 20% methanol pH 8.3 at 100 MA for 15 hours. Preparation of synthetic peptides. Peptide fragments corresponding to sequences located within bovine bFGF were synthesized by solid phase methods 1 "· 14 using an automated Applied Biosystems (Foster City, CA) 430A peptide synthesizer. Peptide one has 12 amino acid residues (pro-ala-leu-pro-gluasp-gly-gly-ser-gly-ala-phe) and corresponds to the amino-terminus 1-12 of FGF, peptide two has 11 residues (arg-ile-hispro-asp-gly-arg-val-asp-gly-val) and is located at position 33-43, and peptide three has 10 amino acid residues (ala-ileleu-phe-leu-pro-met-ser-ala-lys) at position 136-145 of the carboxyl terminal region of FGF. All peptides were synthesized with an additional cysteine residue at the carboxyl-terminus to facilitate conjugation to the carrier protein. Production of polyclonal antisera. 1) Congugation of hapten to carrier protein. The synthetic peptide fragments were conjugated to KLH as a carrier using MH§ as a cross-linking agenL For this purpose, 30 mg. of KLH was dissolved in three mL of PBS pH 7A, and 10 mg. of MHS was dissolved in 0.3 ml. of dimethylformamide 'Nith constant stirring, the disto KLH over a five min. solved MHS was then added The mixture was to react at room temperature 'Nith continued stirring for 45 min. To separate the activated carrier protein from the excess of MHS, the reaction mixture was immediately applied to a 60 cm. x 0.7 cm. gel-filtration column packed with Trisacryl 0.5 (LKB) and equilibrated with PBS/10% DMF, pH 6.0. The flow rate was 50 to 60 ml./hr. and the eluent was monitored at 280 nm. The activated KLH eluting at the void volume was collected. Thirty mg. of peptide was dissolved in three ml. of PBS and the pH adjusted to 7.4 with one N NaOH. The dissolved peptide was combined with the activated KLH and was allowed to react for three hours at room temperature with continuous shaking. The peptide-carrier conjugate was dialyzed against PBS and stored frozen at -20C. 2) Immunization. Eight pound male New Zealand white rabbits were injected at multiple dorsal intradermal sites with 0.5 mg. of KLH-peptide conjugate emulsified in complete Freund's adjuvant. Animals were boostered regularly at four to six week intervals with 0.2 mg. of KLH-peptide conjugate
424
STORY AND ASSOCIATES
emulsified in incomplete Freund's adjuvant. Rabbits were bled from a central ear vein at various time intervals following the initial immunization and the subsequent booster injections. The sera were aliquoted and stored at -20C. The titer was determined by an Enzyme Linked Immunosorbant Assay using unconjugated peptide as antigen and peroxidase-conjugated goat anti-rabbit IgG as secondary antiserum. Immunodetection by anti-bFGF sera. Samples transferred to nitrocellulose paper were stained with antiserum as follows: incubated with 1 % BSA in PBS for one hour to block nonspecific binding sites, washed 3 x 5 min., incubated with antipeptide antiserum at 1:500 dilution overnight, washed 3 x 5 min., incubated with 1:3000 dilution of horseradish peroxidaseconjugated goat anti-rabbit lgG (Bio-Rad) for two hours, washed 3 X 5 min., incubated with enzyme substrate until color developed, washed with distilled water. All antibody dilutions were in 1% BSA/PBS and washing steps were in PBS. The enyzme substrate was from Bio-Rad and prepared as described by the supplier.
TABLE 1. Growth factor activity in alkaline and acidic extracts of various tissues Growth Factor Activity• Units/gm. Wet Weight
Tissue Source
Alkaline Extracts Human prostated open prostatectomy TURP Human testes Human epididymides Calf testes Calf epididymides
2,949 1,761 516 729 1,231 1,798
ND"
(2,758-3,086) (1,198-2,568) (157-1,268) (283-1,307) (304-3,549) (736-2,859)b
• Mean and Range, N :c,: 3 except h where N "Not determined. d From Story, et al. 8
217 (48-375)
ND ND 3,056 (2,520-3,427)
ND
= 2.
TABLE 2. Stability of growth factor activity after heating and treatment with acid, guanidine-hydrochloride and 2-mercaptoethanol % Control Activity
Tissue Source
lOOC (5 min.)
RESULTS
Growth factor activity in extracts of prostate, testes and epididymides. Table 1 compares the growth factor activity in extracts of human prostate, testes, and epididymides and in calf testes and epididymides. All tissues examined had activity. The range of activity in extracts of human bovine testes and epididymides varied by as much as 10-fold. The large variation in preparation of the same type of tissue does not permit assessment of the relative amount of growth factor in these urogenital organs. However, the activity of human prostate homogenates was eight-fold higher in alkaline extracts, than when acidic extraction was performed; whereas, the average activity of calf testes homogenates was higher in acidic extracts. To further characterize the growth factors in tissue homogenates, the stability of biologic activity to heat, acid, dissociating agent and thiol-reducing agent was studied (table 2). The activity in all homogenates reacted similarly. The growth factor was inactivated by boiling and acid (pH 2.9) but was stable to reduction by 2-mercaptoethanol. Treatment with guanidine at pH 7.6 reduced activity by 11% to 33% of control values. Partial purification of the testes-derived growth factor by heparin-affinity chromatography. Since the growth factor in the various tissue homogenates behaved similarly when treated with potentially denaturing agents, the property of the human and calf testes-derived growth factor to bind heparin was investigated. Recent studies 4 ' 8 demonstrated that the human prostate-derived growth factor bound heparin. Therefore, the growth factor in testes homogenates was precipitated by ammonium sulfate and subjected to heparin-affinity chromatography. The predominant activity from both human and calf testes bound to a column of heparin-Sepharose and eluted between 1.2 Mand 1.8 M sodium chloride (figure lA and lC). The growth factor recovered from the heparin-Sepharose column was applied to a second heparin column and the growth factor was eluted as before (figure lB and lD). A summary of the purification and recovery of the growth factor from alkaline extracts of calf testes and human testes is shown in tables 3 and 4, respectively. The growth factor was also partially purified from acidic extracts of calf testes by two cycles of heparinaffinity chromatography (not shown). Immunologic relationship of testes-derived growth factor to bFGF. The partially purified growth factor from alkaline homogenates of human testes, prostate, and alkaline and acidic homogenates of calf testes were subjected to SDS-PAGE and transferred to nitrocellulose paper. Nitrocellulose strips were incubated with rabbit antisera against synthetic peptides with sequence homologies to amino acid residues 1-12 (amino-terminal peptide), 33-43 (internal peptide) and 136-145 (carboxylterminal peptide) of bovine bFGF. The nitrocellulose strips
Acidic Extracts
Human prostate Human testes Human epididymides Calf testes Calf epididymides
Acetic Acid (0.1 M)
Guanidine (3M,PH 7.6)
2-Mercaptoethanol (10%, pH 7.6)
19 10 11 6 3
33 26 11 25 22
118 89 79 91 135
0 13 2 0 0
A
C.
0.4
0.4
0.3
1.5 0.3
6.0
0.2
1.0 0
4.0~
3.0
2.0
:,.
o.s
0.1
'
0.1
2.0~
1.0
6.0
3.0
B. 0.2
4.0
~
:,. 1.0
2.0 [
I__ 20
m
~
o
o
s
10
2.0~
0 1.0~
J..
w
m
~
FRACTION NUMBER
FIG. 1. Heparin-Sepharose chromatography. A, calf testes (175 gm.) and C, human testes (72 gm.), treated by alkaline extraction procedure described in MATERIALS AND METHODS, were applied to column of heparin-Sepharose (40 ml. bed volume) and washed with equilibration buffer. Salt gradient was started (arrow). Fractions were collected and absorbance (-),conductivity(--) and growth factor activity (hatched area) were measured. Activity eluting from column (bar) was concentrated, dialyzed and applied to 15 ml. bed volume column of heparinSepharose (B, calf testes preparation; D, human testes preparation). The column was washed with equilibration buffer and salt gradient was started (arrow). Fractions were collected and absorbance, conductivity, and growth factor activity were measured. Fractions 24-34 (B) and fractions 26-35 (D) were dialyzed, concentrated and analyzed by SDSPAGE.
were then incubated with peroxidase-conjugated second antibody to detect the bound primary antibody. As seen in figure 2A, lanes 2-4, antiserum to each of the three synthetic peptides of bFGF reacted with human testesderived proteins resolved by SDS-PAGE at MW 17,500. The proteins appeared as a doublet as did the prostate-derived proteins visualized after reaction with antiserum to the aminoterminal peptide (figure 2A, lane 1). A 16,000 MW protein was also visualized by antiserum to the internal peptide (figure 2A, lane 3). When the growth factors from alkaline and acidic extracts of calf testes were electrophoresed side-by-side it could
425
BASIC FGF IN TESTES TABLE 3.
Summary of the isolation of growth factor from calf testes Activity Protein (mg.) Units/mg. Totai
Step 1. Homogenate•
2. Ammonium sulfate precipitate 3. Heparin-Sepharose (40 ml.) 4. Heparin-Sepharose (15 ml.)
3,743 2,296
29.7 56.9
NDb 1.1
111,167 130,642
Fold % Purification Recovery l 1.9
i
2
3
100 118
ND
84,000
ND
76
49,450
54,395
1,665
49
"From 175 gm. tissue. b Not determined. TABLE 4.
Summary of the isolation of growth factor from human testes
Step 1. Homogenate" 2. Ammonium sulfate precipitate 3. Heparin-Sepharose (40ml.) 4. Heparin-Sepharose (15 ml.)
Activity Protein (mg.) Units/mg. Total 1,445 899 NDb
63.2 91,324 206 185,194 ND
0.025 625,000
Fold % Purification Recovery 1 3.3
100 203
24,180
ND
26
15,625
9,889
17
1
2
3
4
5
43
"From 72 gm. tissue. b Not determined.
be seen that the alkaline extracted growth factor was slightly larger than the growth factor from acidic extracts of the same tissue. Whereas, the predominant protein in alkaline extracts of calf testes reacting with antisera to the amino- and carboxylterminal peptides was resolved at MW 17,500 (figure 2B, lanes 1 and 5), a protein of apparent MW 16,600 in acidic extracts reacted with antiserum to all three synthetic peptides (figure 2B, lanes 2-4). In addition, a 15,500 MW protein in acidic extracts reacted with antiserum to the internal and carboxylterminal peptides (figure 2B, lanes 3, 4) but not with antiserum to the amino-terminal peptide (figure 2B, lane 2). DISCUSSION
A growth factor has been identified in human prostate and testes, and in calf testes that is structurally related to bFGF. The evidence for this relationship is 1) that the growth factor activity, like bFGF, is heat and acid labile 15 but not affected thiol-reducing agents, 16 2) that the growth factor binds heparin and requires L4 M to 1.8 M sodium chloride for dissociation, 17 and 3) that antisera prepared against synthetic peptides with sequence homology to amino acid residues 33-43, as well as amino-terminal sequences 1-12 and carboxyl-terminal sequences 136-145 of bovine BFGF react with the prostate grnwth factm and the human and calf testes-derived growth factors (figure 2). Stability studies (table 2) suggest that the same grmvth factor is also present in epididymis. Unfortunately, large variations in the amount of activity in replicate preparations of testes and epididymides (table 1) does not permit assessment of the relative amount of growth factor in various tissues. It is not clear if variations in activity, not seen in prostate, is due to actual differences in the amount of growth factor (or inhibitors) in individual tissues or due to differences in the duration of storage or to uncontrolled differences during homogenization. Possible differences in the amount of growth factor in individual human prostate tissues would not be evidence since homogenates were prepared from pools of 20 or more specimens; whereas, human testis and epididymis preparations were from two to five patients. The availability of the latter tissues also necessitated larger variations in the duration of storage prior to preparing homogenates. However, these factors do not account for the differences among calf tissue preparations. Like prostate, calf tissue homogenates
25.7
2 FIG. 2. Detection of growth factor by antisera to synthetic peptides with sequence homologies to bFGF. Partially purified growth factor preparation (1500 to 3000 units/lane) were electrophoresed by SDSp AGE. Proteins were transferred electrophoretically to nitrocellulose paper and incubated with antisera against synthetic peptides with sequence homologies to bovine bFGF. Bound antibody was visualized by incubation with peroxidase-conjugated goat anti-rabbit IgG. Molecular weight markers were visualized in the original gel by silver staining. A, growth factor from alkaline extracts of human prostate (lane 1) and alkaline extracts of human testes (lanes 2-4). Lanes 1 and 2 were incubated with antiserum to amino-terminal peptide; lime 3, internal peptide; lane 4, carboxyl-terminal peptide. B, growth factor from alkaline extracts of calf testes (lanes 1, 5) and acidic extracts of calf testes (lanes 2-4). Lanes 1 and 2 were incubated with antiserum to amino-terminal peptide; lane 3, internal peptide; lanes 4, 5, carboxyterminal peptide.
were prepared from pools of a large number of animals and the time of storage was short. Regardless of the source, all tissue homogenates show toxicity when tested at high concentrations in the bioassay. Since quantitation of growth factor depends on radionucleotide uptake that is influenced not only by the amount of mitogen present but also by the amount of inhibitor present, it is possible that subtle differences in tissue handling during storage or homogenization could result in different amounts of these agents in homogenates and influence growth factor quantitation. The procedure used to prepare tissue homogenates influenced the amount of activity recovered. Homogenates of human prostate in high ionic strength alkaline buffer containing protease inhibitors had eight-fold higher levels of activity than when tissue was homogenized under acidic conditions in the absence of protease inhibitors. On the other hand, acidic homogenates of calf testes had, on the average, more activity than did
426
STORY AND ASSOCIATES
homogenates prepared by the alkaline extraction procedure (table 1). The difference in behavior of human and bovine tissue under the two extraction conditions may reflect species differences in growth factor structure or differences in tissue specific proteases activated during extraction. Basic FGF has been purified to homogeneity from a number of bovine 10• 17• 24 and human 25- 27 tissues. The sequence of bovine bFGF is known 28 and the sequence of the human growth factor has been predicted from the cloned cDNA. 29 Of the 146 amino acids in the mature proteins only the amino acids at residues 112 and 128 differ. These small changes in the amino acid sequence could influence acid stability or alter susceptibility of the growth factor to tissue specific proteases. The homogenization method also affected the structure of the growth factor isolated (figure 2). Alkaline extraction in the presence of protease inhibitors of human prostate resulted in the isolation of a large form of bFGF. Amino-terminal sequencing9 showed that the growth factor has eight additional amino acids at its amino-terminus not found in bFGF 28 but whose sequence was predicted from the nucleotide sequence by molecular cloning. 29 • 30 Immunodetection of a growth factor of comparable size in alkaline extracts of human and calf testes (figure 2) suggests that the same extended form of the growth factor is present in these tissues, although this has yet to be confirmed by sequence analysis. Antibody reactivity with smaller MW proteins suggests that proteolysis was not completely inhibited in these preparations. Ueno et al. 31 isolated two forms of bFGF from acidic extracts of bovine pituitary when protease inhibitors were included during extraction. One form was identified as (1-146)bFGF and the other was an amino-terminal extended form. When protease inhibitors were omitted during homogenization of calf testes, two forms of the growth factor were detected by immunostaining (figure 2B). One form, apparent MW 16,600, reacted with antisera to all three synthetic bFGF peptides and was likely (1-146)bFGF. A smaller 15,500 MW form reacted with antisera to the internal and carboxyl-terminal peptides, but not with antiserum to the amino-terminal peptides; suggesting that the protein lacked amino-terminal bFGF sequences. A truncated form of bFGF lacking the amino-terminal 15 amino acids has been isolated from bovine adrenal, 24 corpus luteum 10 and kidney22 homogenates prepared without protease inhibitors. It is likely that the 15,500 MW peptide seen in calf testes homogenates is the (16145) bFGF formed by proteolysis of the large precursor form. During preparation of this paper, Ueno et al. 32 reported the purification ofbFGF from bovine testes. The growth factor was purified in the absence of protease inhibitors and found to be the amino-terminal 15-amino acid deleted form. Our previous reports 8 • 9 and the evidence presented here strongly suggest that the growth factor in human prostate, previously called prostatic growth factor, PrGF, 2- 5 and the growth factor in the seminiferous epithelium of the calf and several mammalian species, previously called seminiferous growth factor, SGF, 6 • 7 is bFGF. PrGF has been suggested33 to function in a paracrine manner in the development of the fibrostomal nodule characteristic of BPH. 34 SGF, reported to be present in higher amounts in Sertoli cells, has been suggested to mediate proliferation of spermatogonia and spermatocytes to ensure an orderly, continuous, and abundant production of spermatozoa. 7 It has yet to be determined if the growth factor plays a physiologic role in these and other developmental processes. Fibroblast growth factors are known to be mitogenic for cultured endothelial cells 18 and to be angiogenic factors in vivo. 28 In addition, cultured endothelial cells have recently been found to produce the growth factor. 36- 38 The methodology that we have employed in demonstrating bFGF in genitourinary organs does not preclude the possibility that endothelial cells may be the source of the growth factor. If endothelial cells produce bFGF in vivo, this may explain the presence of the
growth factor in a wide variety of tissues. Further studies employing cultured cells derived from tissues and the application of immunohistochemical techniques to localize bFGF are required to identify the cell type responsible for bFGF production in vivo. Acknowledgments. We wish to thank Bonnie Livingston, Susan Swartz and Tom Tebo for expert technical assistance. We are indebted to Nacker Packing Co., Inc., Milwaukee, WI, for their generous supply of calf testes. REFERENCES
1. Jacobs, S. C., Pikna, D. and Lawson, R. K.: Prostatic osteoblastic factor, Invest. Urol., 17: 195, 1979. 2. Tackett, R. E., Heston, W, D. W., Parrish, R, F., Pletscher, L. S, and Fair, W, R.: Mitogenic factors in prostatic tissue and expressed prostatic secretion. J. Urol., 133: 45, 1985. 3. Story, M. T., Jacobs, S, C.and Lawson, R. K.: Partial purification of a prostatic growth factor. J. Urol., 132: 1212, 1984. 4. Nishi, N., Matuo, Y., Muguruma, Y., Yoshitake, Y., Nishikawa, Y. and Wada F.: A human prostatic growth factor (hPGF): partial purification and characterization. Biochem. Biophys. Res. Commun., 132: 1103, 1985. 5. Jinno, H., Ueda, K., Otaguro, K., Kato, T., Ito, J. and Tanaka, R: Prostate growth factor in the extracts of benign prostatic hypertrophy, Eur. Urol., 12: 41, 1986. 6, Feig, L.A., Klabsbrun, M. and Bellve, A. R.: Mitogenic polypeptide of the mammalian seminiferous epithelium: biochemical characterization and partial purification, J. Cell. Biol., 97: 1435, 1983. 7. Bellve, A. R. and Feig, L. A.: Cell proliferation in the mammalian testes: biology of the seminiferous growth factor (SGF). In: Recent Prog. Hormone Res. Edited by R.O. Greep. New York: Academic Press, vol. 40, pp. 534-561, 1984. 8. Story, M. T., Sasse, J,, Jacobs, S. C. and Lawson, R. K.: Prostatic growth factor: purification and structural relationship to basic fibroblast growth factor. Biochemistry, 26: 3843, 1987. 9, Story, M. T., Esch, F., Shimasaki, S., Sasse, J., Jacobs, S. C. and Lawson, R. K.: Amino-terminal sequence of a large form of basic fibroblast growth factor isolated from human benign prostatic hyperplastic tissue. Biochem. Biophys. Res. Commun., 142: 702, 1987. 10. Gospodarowicz, D., Cheng, J., Lui, G, M., Baird, A., Esch, F. and Bohlen, P.: Corpus luteum angiogenic factor is related to fibroblast growth factor. Endocrinology, 117: 2383, 1985. 11. Laemmli, U.K.: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London), 227: 680, 1970. 12. Wray, W., Boulikas, T., Wray, V. P. and Hancock, R.: Silver staining of proteins in polyacrylamide gels. Anal. Biochem., 118: 197, 1981. 13. Merrifield, R. B.: Solid phase peptide synthesis. 1. The synthesis of a tetrapeptide. J. Am. Chem. Soc., 85: 2149, 1963. 14. Sakakibara, S.: The use of hydrogen fluoride in peptide chemistry. In: Chemistry and Biochemistry of Amino Acid, Peptides and Proteins. Edited by B. Weinstein. New York, Dekker, vol. 1, p. 51, 1971. 15. Gospodarowicz, D., Bialecki, H. and Greenburg, G.: Purification of the fibroblast growth factor activity from bovine brain. J. Biol. Chem., 253: 3736, 1978. 16. Maciag, T., Cerundolo, J., Ilsley, S., Kelley, P. R. and Forand, R.: An endothelial cell growth factor from bovine hypothalamus: identification and partial characterization, Proc. Natl. Acad. Sci. USA, 76: 5674, 1979. 17. Lobb, R., Sasse, J., Sullivan, R., Shing, Y., D'Amore, P., Jacobs, J. and Klagsbrun, M.: Purification and characterization ofheparinbinding endothelial cell growth factors. J. Biol. Chem., 261: 1924, 1986. 18. Gospodarowicz, D., Cheng, J., Lui, G-M., Baird, A. and Bohlen, P.: Isolation of brain fibroblast growth factor by heparin-Sepharose affinity chromatography: identity with pituitary fibroblast growth factor. Proc. Natl. Acad. Sci. USA, 81: 6963, 1984. 19. Lobb, R. R. and Fett, J. W.: Purification of two distinct growth factors from bovine neural tissue by heparin affinity chromatography. Biochemistry, 23: 6295, 1984. 20. Klagsbrun, M. and Shing, Y.: Heparin affinity of anionic and
BASIC FGF II~ TESTES
21.
22.
23.
24.
25.
26. 27.
28.
cationic capillary endothelial ceil growth factors. Proc. Natl. Acad. Sci. USA, 82: 805, 1985. Bohlen, P., Baird, A., Esch, F., Ling, N. and Gospodarowicz, D.: Isolation and partial molecular characterization of pituitary fibroblast growth factor. Proc. Natl. Acad. Sci. USA, 81: 5364, 1984. Baird, A., Esch, F., Bohlen, P., Ling, N. and Gospodarowicz, D.: Isolation and partial characterization of an endothelial growth factor. Regul. Pept., 12: 201, 1985. Baird, A., Esch, F., Gospodarowicz, D. and Guillemin, R.: Retinaand eye-derived endothelial cell growth factors: partial molecular characterization and identity with acidic and basic fibroblast growth factor. Biochemistry, 24: 7855, 1985. Gospodarowicz, D., Baird, A., Cheng, J., Lui, G. M., Esch, F. and Bohlen, P.: Isolation of fibroblast growth factor from bovine adrenal gland: physicochemical and biological characterization. Endocrinology, 118: 82, 1986. Bohlen, P., Esch, F., Baird, A., Jones, K. L. and Gospodarowicz, D.: Human brain fibroblast growth factor: isolation and partial chemical characterization. FEBS Letters, 185: 177, 1986. Gospodarowicz, D., Cheng, J., Lui, G-M., Fujii, D. K., Baird, A. and Bohlen, P.: Fibroblast growth factor in the human placenta. Biochem. Biophys. Res. Commun., 128: 554, 1985. Klagsbrun, M., Sasse, J., Sullivan, R and Smith, J. A.: Human tumor cells synthesize an endothelial cell growth factor that is structurally related to basic fibroblast growth factor. Proc. Natl. Acad. Sci. USA, 83: 2448, 1986. Esch, F., Baird, A., Ling, N., Ueno, N., Hill F., Denoroy, L., Klepper, R., Gospodarowicz, D., Bohlen, P. and Guillemin, R.: Primary structure of bovine pituitary basic fibroblast growth factor (FGF) and comparison with the amino-terminal sequence of bovine brain acidic FGF. Proc. Natl. Acad. Sci. USA, 82: 6507, 1985.
427
29. Kmokmva, T., Sasada, R., Iwane, NL and Igarashi, K: Cloning and expression of cDNA encoding human basic fibroblast growth factor. FEBS Letters, 213: 189, 1987. 30. Abraham, J. A., Mergia, A., Whang, J.L., Tumolo, A., Friedman, J., Hjerrild, K. A., Gospodarowicz, D. and Fiddes, J. C.: Nucleotide sequence of a bovine clone encoding the angiogenic protein, basic fibroblast growth factor. Science, 233: 545, 1986. 31. Ueno, N., Baird, A., Esch, F., Ling, N. and Guillemin, R.: Isolation of an amino-terminal extended form of basic fibroblast growth factor. Biochem. Biophys. Res. Commun., 138: 580, 1986. 32. Ueno, N., Baird, A., Esch, F., Ling, N. and Guillemin, R.: Isolation and partial characterization of basic fibroblast growth factor from bovine testes. Mo!. Cell. Endocrinol., 49: 189, 1987. 33. Jacobs, S. C. and Lawson, R. K.: Mitogenic factor in human prostate extracts. Urology, 16: 488, 1980. 34. Moore, R. A.: Benign hypertrophy of the prostate: a morphological study. J. Urol., 50: 680, 1943. 35. Pradhan, B. K. and Chandra, K.: Morphogenesis of nodular hyperplasia prostate. J. Urol., 113: 210, 1975. 36. Schweigerer, L., Neufeld, G., Friedman, J., Abraham, J. A., Fiddes, J. C. and Gospodarowicz, D.: Capillary endothelial cells express basic fibroblast growth factor, a mitogen that promotes their own growth. Nature, 325: 257, 1987. 37. Baird, A. and Ling, N.: Fibroblast growth factors are present in the extracellular matrix produced by endothelial cells in vitro: implications for a role of heparinase-like enzymes in the neovascular response. Biochem. Biophys. Res. Commun., 142: 428, 1987. 38. Vlodavsky, I., Folkman, J., Sullivan, R., Friedman, R., Ishai-Michaeli, R., Sasse, J. and Klagsbrun, M.: Endothelial cell-derived basic fibroblast growth factor: synthesis and deposition into subendothelial extracellular matrix. Proc. Natl. Acad. Sci. USA, 84: 2292, 1987.
THE JOURNAL OF UROLOGY
Copyright© 1988 by The Williams & Wilkins Co.
A GROWTH FACTOR IN BOVINE AND HUMAN TESTES STRUCTURALLY RELATED TO BASIC FIBROBLAST GROWTH FACTOR MICHAEL T. STORY,* JOACHIM SASSE, DANIEL KAKUSKA, STEPHEN C. JACOBS RUSSELL K. LAWSON
AND
From the Medical College of Wisconsin, Departments of Urology and Biochemistry, Milwaukee, Wisconsin and Harvard Medical School, Department of Surgery, Boston, Massachusetts
ABSTRACT
Homogenates of human testes, epididymides and prostate, and calf testes and epididymides are mitogenic for cultured human foreskin fibroblasts. The growth factors appear similar in that they are inactivated by boiling and acid, but not by treatment with reducing agent. The growth factor in human and bovine testes was partially purified from tissue homogenates, prepared in high ionic strength buffer (pH 7.6) containing protease inhibitors, by ammonium sulfate precipitation and two cycles of heparin-Sepharose chromatography. The growth factor in calf testes was also partially purified from tissue extracted in ammonium sulfate without protease inhibitors, acidified to pH 4.5, and precipitated by ammonium sulfate followed by two cycles of heparin-affinity chromatography. A predominant 17,500 molecular weight (MW) growth factor was identified from alkaline homogenates of human and calf testes by its reactivity with antisera prepared against synthetic peptides whose sequences corresponded to residues 1-12 (amino-terminal), 33-43 (internal) and 136-145 (carboxy-terminal) of bovine basic fibroblast growth factor (bFGF). A slightly smaller 16,600 MW peptide from acidic extracts of calf testes also reacted with antisera to the three synthetic peptides. A 15,500 MW peptide, lacking immunoreactivity with antiserum to the amino-terminal synthetic peptide, was also seen. These findings suggest that a growth factor is present in human and calf testes that is structurally related to bFGF. The structure of the growth factors appears to be altered during the isolation procedure. (J. Ural., 140: 422-427, 1988) Jacobs and associates 1 first reported the presence of a growthpromoting factor in extracts of human benign prostatic hyperplasia (BPH) in 1979. The presence of prostatic growth factor (PrGF) in BPH has been confirmed by several investigators.2-5 While comparing the physical chemical and biological properties of PrGF with those of other growth factors, it became apparent that PrGF was similar to a growth factor called seminiferous growth factor (SGF) in the seminiferous epithelium of the calf and several mammalian species. 6 • 7 Recently, we have purified the growth factor from BPH tissue 8 and demonstrated by amino-terminal sequencing that PrGF is structurally related to basic fibroblast growth factor, bFGF. 9 Now we report immunologic data that SGF in calf and human testes is also related to bFGF. MATERIALS AND METHODS
Materials. Human prostate tissue was from surgical specimens obtained at transurethral (TURP) and open prostatectomy for BPH. Human testes and epididymides were from patients, mean 80 years of age, who underwent androgen ablation for prostate cancer. Calf testes and epididymides were obtained fresh from a local meat processing plant. Tris (hydroxymethyl) aminomethane (Tris), diamethylformamide, phenylmethanesulfonyl fluoride, L-1 (tosylamido)-2-phenyl-ethyl chloromethyl ketone, N-ethylmalemide, guanidine-hydrochloride, soybean trypsin inhibitor and keyhole limpet hemocyanin, KLH, were from Sigma (St. Louis, MO). N-hydroxysuccinimidyl maleimidohexanoate ester (MHS) was from Boehringer Mannheim Accepted for publication March 15, 1988. * Requests for reprints: Dept. of Urology, Medical College of Wisconsin, 9200 West Wisconsin Ave., Milwaukee, WI 53226. Supported by NIH grant AM31063. 422
Biochemicals, Indianapolis, IN. Disodium ethylenediaminetetraacetate (EDTA) and glacial acetic acid were from JT Baker Chemical Co. (Phillipsburg, NJ). Ammonium sulfate and sodium chloride were from Mallinckrodt (Paris, KY). All tissue culture medium, antibiotics and serum were from Grand Island Biological Company (Grand Island, NY) and all culture flasks and multiwell plates were from Falcon (Becton Dickinson Labware, Oxnard, CA). Econofluor scintillation cocktail, Protosol tissue solubolizer and (Methyl-3H)-thymidine were from New England Nuclear (Boston, MA). All reagents for SDSPAGE were from LKB (Gaithersburg, MD). Molecular Weight (MW) markers were from Bethesda Research Laboratories (Gaithersburg, MD). Protein was assayed with a kit with BSA as standard (Bio-Rad, Richmond, CA). Cell culture and assay for growth factor activity. The bioassay was based on the ability of a sample to stimulate (3H)-thymidine uptake by quiescent confluent monolayers of human foreskin fibroblasts as described elsewhere. 3 The amount of radioactivity was determined with a model 1219 liquid scintillation counter (LKB) interfaced with an Apple II E for conversion of cpm to dpm. The mean of triplicate samples was determined and a Relative Mitogenic Activity (RMA) was calculated: RMA
= DPM
(Test Sample) - DPM (Buffer) DPM (10% NBS) - DPM (Buffer)
The buffer was the same as in the test sample. One unit of activity was defined as a RMA equal to one, that is, a response equal to that produced by 10% newborn bovine serum (NBS). Tissue homogenization and ammonium sulfate-precipitation. Tissue was transported in saline in an ice bath. The tissue was trimmed free of fat and connective tissue. The epididymidis was separated from the testis, and the tunica albuginea re-
423 ffl0V8d, 'fiS8U88 V\T8r€ Stored frozen ,ncmcc,cco, 0Il8 tO three hurnan testes and enicti,dv:rrn,ctes one to six months; calf testes one to six weeks; calf epididymides two to four months, All steps in preparing tissue homogenates and in isolating growth factor were at OC to 4C. The tissues were thawed, rinsed in saline and minced with scissors. The tissue was weighed and homogenized with a PT 10-35 Polytron equipped with a PT 20ST probe generator (Brinkman Instruments, Inc., Westbury, NY) at a setting of eight for 30 sec at 10 sec intervals with 30 sec cooling between homogenizations. In the "alkaline extraction" procedure, the homogenization buffer (1:2 tissue wet weight: volume), was 50 mM Tris, 1.55 M NaCl pH 7.6 containing EDTA (10 mM), phenylmethanesulfonyl fluoride (1 mM), L-l(tosylamido)-2-phenylethyl chloromethyl ketone (0.3 mM), N-ethylmaleimide (0.05 mM), soybean trypsin inhibitor (10 mg./L). The homogenate was centrifuged at 23,000 g for 40 min. in a Sorvall model RC-2B centrifuge (DuPont Instruments, Newton, CT) and the supernatant was transferred to a clean beaker. An equal volume of buffer was added to the pellet and the homogenization and centrifugation was repeated. The supernatant from the two centrifugations was pooled and filtered through a Metrigard glass fiber membrane (Gelman Sciences, Inc., Ann Arbor, MI) with positive nitrogen pressure. A portion of the filtrate was dialyzed against 50 mM Tris, 0.15 M NaCl pH. 7.6 (TBS) in bags made of Spectra/Por3 tubing, 3,500 MW cutoff (Spectrum Medical Industries, Los Angeles, CA). A precipitate was removed by centrifugation and the supernatant was sterilized by filtration (0.2 µ Acrodisc membranes, Gelman) and assayed for activity and protein. The remaining filtrate was brought to 25 % ammonium sulfate the slow addition of saturated ammonium sulfate, pH 7.6. After mixing two hours, the preparation was centrifuged at 23,000 g for 30 min. The pellet was discarded and the supernatant was brought to 75% ammonium sulfate. After mixing and centrifuging, the pellet was dissolved in TBS and dialyzed against the same buffer. The preparation was centrifuged, Metrigard filtered and a portion was sterilized by filtration and assayed for activity and protein. In the "acidic extraction" procedure, Polytron homogenization was in 0.15 M ammonium sulfate. The remainder of the procedure was from Gospodarowicz, et aL 10 The pH of the homogenate was centrifuged, and the pH of the supernatant was adjusted to 7,0 with one N NaOH. Ammonium sulfate (230 gm.fl.; final concentration, 1.9 M) was added and the precipitate was removed after one hour mixing by centrifugation. Further addition of ammonium sulfate (250 gm.fl. final concentration 3.8 to the supernatant gave a precipitate that was collected after one hour mixing by centrifugation. The itate was dissolved in water and dialyzed overnight against water and then overnight against TBS. _ .,.- .,, •• , studies. Tissue hom,)1,JmE1te:s, 50 mM Tris, 50 mM NaCl, pH were buffer (control and boiled samples), 0,1 M acetic acid, 3 M guanidine-hydrochloride pH 7.6 or 10% 2-mercaptoethanol pH 7.6. After 24 hours at 4C or five min. samples were centrifuged and the supernatant was against Tris buffer, filtered (0.2 µ) and tested for activity, Heparin-Sepharose chromatography. The ammonium sulfate precipitated growth factor was applied to a five cm, X two cm,, 40 ml. bed volume column of heparin-Sepharose (Pharmacia, Piscataway, NJ) equilibrated with TBS. The column was washed with TBS until the absorbance of the eluate at 206 nm. became negligible. The absorbed proteins were then eluted with a gradient (480 ml.) formed from TBS and 50 mM Tris with 3.0 M NaCl. The conductivity of fractions was determined with a conductivity bridge (Yellow Springs Instrument Co., Yellow Spring, OH). A portion of each fraction was diluted in Minimum Essential Medium (Eagle) with 0.5 NBS, sterile filtered and assayed for activity. The growth factor eluting between 1.2 M and 1.8 M NaCl was concentrated by ultrafiltration, YM5 au,nn,HUo
ty-pe men1brane in a stirred cell Amicon (Danvers, and dialyzed agaist 50 mM with 0,25 M NaCl pH 7.6 (calf testes) or 50 mM Tris with 0,5 M NaCl pH 7.6 (human testes). The growth factor recovered from the 40 ml. heparin-affinity column was applied to a 1.6 cm. X 7.5 cm., 15 ml. bed volume column of heparin-Sepharose equilibrated with 50 mM Tris with 0.25 NaCl pH 7.6 (calf testes) or 50 mM Tris with 0.5 M NaCl (human testes). The column was washed with the equilibration buffer until the absorbency of the eluate reached baseline. The absorbed proteins were then eluted with a gradient (280 ml.) formed from 50 mM Tris with 0.25 M NaCl (calf testes) or 50 mM Tris with 0.5 M NaCl (human testes) and 50 mM Tris with 3 M NaCL The conductivity and activity of fractions were measured, Fractions with growth factor were concentrated by ultrafiltration and the salt content was lowered by the addition of 30 volumes of 50 mM Tris with 50 mM NaCl and the sample was reconcentrated. A portion of the sample was analyzed by analytical SDS-PAGE. SDS-PAGE and electroblotting. Analytical SDS-PAGE was performed on a vertical slab gel apparatus (LKB) using 10% to 22.5% linear gradient gels prepared by standard techniques" in the presence of thiol reducing agent, Molecular weight markers were insulin (A and B chains), 2,300 and 3,400; bovine trypsin inhibitor, 6,200; lysozyme, 14,300; /)-lactoglobulin, 18,400; a chymotrypsinogen, 25,700 and ovalbumin, 43,000, Some gels were silver stained. 12 Other gels were electroblotted with a Transphor Unit (LKB) to nitrocellulose paper, 0,2 µ pore size (Schleicher and Schuell, Keene, NH). The transfer was performed in 25 mM Tris-HCl, 0.15 M glycine, 20% methanol pH 8.3 at 100 MA for 15 hours. Preparation of synthetic peptides. Peptide fragments corresponding to sequences located within bovine bFGF were synthesized by solid phase methods 1 "· 14 using an automated Applied Biosystems (Foster City, CA) 430A peptide synthesizer. Peptide one has 12 amino acid residues (pro-ala-leu-pro-gluasp-gly-gly-ser-gly-ala-phe) and corresponds to the amino-terminus 1-12 of FGF, peptide two has 11 residues (arg-ile-hispro-asp-gly-arg-val-asp-gly-val) and is located at position 33-43, and peptide three has 10 amino acid residues (ala-ileleu-phe-leu-pro-met-ser-ala-lys) at position 136-145 of the carboxyl terminal region of FGF. All peptides were synthesized with an additional cysteine residue at the carboxyl-terminus to facilitate conjugation to the carrier protein. Production of polyclonal antisera. 1) Congugation of hapten to carrier protein. The synthetic peptide fragments were conjugated to KLH as a carrier using MH§ as a cross-linking agenL For this purpose, 30 mg. of KLH was dissolved in three mL of PBS pH 7A, and 10 mg. of MHS was dissolved in 0.3 ml. of dimethylformamide 'Nith constant stirring, the disto KLH over a five min. solved MHS was then added The mixture was to react at room temperature 'Nith continued stirring for 45 min. To separate the activated carrier protein from the excess of MHS, the reaction mixture was immediately applied to a 60 cm. x 0.7 cm. gel-filtration column packed with Trisacryl 0.5 (LKB) and equilibrated with PBS/10% DMF, pH 6.0. The flow rate was 50 to 60 ml./hr. and the eluent was monitored at 280 nm. The activated KLH eluting at the void volume was collected. Thirty mg. of peptide was dissolved in three ml. of PBS and the pH adjusted to 7.4 with one N NaOH. The dissolved peptide was combined with the activated KLH and was allowed to react for three hours at room temperature with continuous shaking. The peptide-carrier conjugate was dialyzed against PBS and stored frozen at -20C. 2) Immunization. Eight pound male New Zealand white rabbits were injected at multiple dorsal intradermal sites with 0.5 mg. of KLH-peptide conjugate emulsified in complete Freund's adjuvant. Animals were boostered regularly at four to six week intervals with 0.2 mg. of KLH-peptide conjugate
424
STORY AND ASSOCIATES
emulsified in incomplete Freund's adjuvant. Rabbits were bled from a central ear vein at various time intervals following the initial immunization and the subsequent booster injections. The sera were aliquoted and stored at -20C. The titer was determined by an Enzyme Linked Immunosorbant Assay using unconjugated peptide as antigen and peroxidase-conjugated goat anti-rabbit IgG as secondary antiserum. Immunodetection by anti-bFGF sera. Samples transferred to nitrocellulose paper were stained with antiserum as follows: incubated with 1 % BSA in PBS for one hour to block nonspecific binding sites, washed 3 x 5 min., incubated with antipeptide antiserum at 1:500 dilution overnight, washed 3 x 5 min., incubated with 1:3000 dilution of horseradish peroxidaseconjugated goat anti-rabbit lgG (Bio-Rad) for two hours, washed 3 X 5 min., incubated with enzyme substrate until color developed, washed with distilled water. All antibody dilutions were in 1% BSA/PBS and washing steps were in PBS. The enyzme substrate was from Bio-Rad and prepared as described by the supplier.
TABLE 1. Growth factor activity in alkaline and acidic extracts of various tissues Growth Factor Activity• Units/gm. Wet Weight
Tissue Source
Alkaline Extracts Human prostated open prostatectomy TURP Human testes Human epididymides Calf testes Calf epididymides
2,949 1,761 516 729 1,231 1,798
ND"
(2,758-3,086) (1,198-2,568) (157-1,268) (283-1,307) (304-3,549) (736-2,859)b
• Mean and Range, N :c,: 3 except h where N "Not determined. d From Story, et al. 8
217 (48-375)
ND ND 3,056 (2,520-3,427)
ND
= 2.
TABLE 2. Stability of growth factor activity after heating and treatment with acid, guanidine-hydrochloride and 2-mercaptoethanol % Control Activity
Tissue Source
lOOC (5 min.)
RESULTS
Growth factor activity in extracts of prostate, testes and epididymides. Table 1 compares the growth factor activity in extracts of human prostate, testes, and epididymides and in calf testes and epididymides. All tissues examined had activity. The range of activity in extracts of human bovine testes and epididymides varied by as much as 10-fold. The large variation in preparation of the same type of tissue does not permit assessment of the relative amount of growth factor in these urogenital organs. However, the activity of human prostate homogenates was eight-fold higher in alkaline extracts, than when acidic extraction was performed; whereas, the average activity of calf testes homogenates was higher in acidic extracts. To further characterize the growth factors in tissue homogenates, the stability of biologic activity to heat, acid, dissociating agent and thiol-reducing agent was studied (table 2). The activity in all homogenates reacted similarly. The growth factor was inactivated by boiling and acid (pH 2.9) but was stable to reduction by 2-mercaptoethanol. Treatment with guanidine at pH 7.6 reduced activity by 11% to 33% of control values. Partial purification of the testes-derived growth factor by heparin-affinity chromatography. Since the growth factor in the various tissue homogenates behaved similarly when treated with potentially denaturing agents, the property of the human and calf testes-derived growth factor to bind heparin was investigated. Recent studies 4 ' 8 demonstrated that the human prostate-derived growth factor bound heparin. Therefore, the growth factor in testes homogenates was precipitated by ammonium sulfate and subjected to heparin-affinity chromatography. The predominant activity from both human and calf testes bound to a column of heparin-Sepharose and eluted between 1.2 Mand 1.8 M sodium chloride (figure lA and lC). The growth factor recovered from the heparin-Sepharose column was applied to a second heparin column and the growth factor was eluted as before (figure lB and lD). A summary of the purification and recovery of the growth factor from alkaline extracts of calf testes and human testes is shown in tables 3 and 4, respectively. The growth factor was also partially purified from acidic extracts of calf testes by two cycles of heparinaffinity chromatography (not shown). Immunologic relationship of testes-derived growth factor to bFGF. The partially purified growth factor from alkaline homogenates of human testes, prostate, and alkaline and acidic homogenates of calf testes were subjected to SDS-PAGE and transferred to nitrocellulose paper. Nitrocellulose strips were incubated with rabbit antisera against synthetic peptides with sequence homologies to amino acid residues 1-12 (amino-terminal peptide), 33-43 (internal peptide) and 136-145 (carboxylterminal peptide) of bovine bFGF. The nitrocellulose strips
Acidic Extracts
Human prostate Human testes Human epididymides Calf testes Calf epididymides
Acetic Acid (0.1 M)
Guanidine (3M,PH 7.6)
2-Mercaptoethanol (10%, pH 7.6)
19 10 11 6 3
33 26 11 25 22
118 89 79 91 135
0 13 2 0 0
A
C.
0.4
0.4
0.3
1.5 0.3
6.0
0.2
1.0 0
4.0~
3.0
2.0
:,.
o.s
0.1
'
0.1
2.0~
1.0
6.0
3.0
B. 0.2
4.0
~
:,. 1.0
2.0 [
I__ 20
m
~
o
o
s
10
2.0~
0 1.0~
J..
w
m
~
FRACTION NUMBER
FIG. 1. Heparin-Sepharose chromatography. A, calf testes (175 gm.) and C, human testes (72 gm.), treated by alkaline extraction procedure described in MATERIALS AND METHODS, were applied to column of heparin-Sepharose (40 ml. bed volume) and washed with equilibration buffer. Salt gradient was started (arrow). Fractions were collected and absorbance (-),conductivity(--) and growth factor activity (hatched area) were measured. Activity eluting from column (bar) was concentrated, dialyzed and applied to 15 ml. bed volume column of heparinSepharose (B, calf testes preparation; D, human testes preparation). The column was washed with equilibration buffer and salt gradient was started (arrow). Fractions were collected and absorbance, conductivity, and growth factor activity were measured. Fractions 24-34 (B) and fractions 26-35 (D) were dialyzed, concentrated and analyzed by SDSPAGE.
were then incubated with peroxidase-conjugated second antibody to detect the bound primary antibody. As seen in figure 2A, lanes 2-4, antiserum to each of the three synthetic peptides of bFGF reacted with human testesderived proteins resolved by SDS-PAGE at MW 17,500. The proteins appeared as a doublet as did the prostate-derived proteins visualized after reaction with antiserum to the aminoterminal peptide (figure 2A, lane 1). A 16,000 MW protein was also visualized by antiserum to the internal peptide (figure 2A, lane 3). When the growth factors from alkaline and acidic extracts of calf testes were electrophoresed side-by-side it could
425
BASIC FGF IN TESTES TABLE 3.
Summary of the isolation of growth factor from calf testes Activity Protein (mg.) Units/mg. Totai
Step 1. Homogenate•
2. Ammonium sulfate precipitate 3. Heparin-Sepharose (40 ml.) 4. Heparin-Sepharose (15 ml.)
3,743 2,296
29.7 56.9
NDb 1.1
111,167 130,642
Fold % Purification Recovery l 1.9
i
2
3
100 118
ND
84,000
ND
76
49,450
54,395
1,665
49
"From 175 gm. tissue. b Not determined. TABLE 4.
Summary of the isolation of growth factor from human testes
Step 1. Homogenate" 2. Ammonium sulfate precipitate 3. Heparin-Sepharose (40ml.) 4. Heparin-Sepharose (15 ml.)
Activity Protein (mg.) Units/mg. Total 1,445 899 NDb
63.2 91,324 206 185,194 ND
0.025 625,000
Fold % Purification Recovery 1 3.3
100 203
24,180
ND
26
15,625
9,889
17
1
2
3
4
5
43
"From 72 gm. tissue. b Not determined.
be seen that the alkaline extracted growth factor was slightly larger than the growth factor from acidic extracts of the same tissue. Whereas, the predominant protein in alkaline extracts of calf testes reacting with antisera to the amino- and carboxylterminal peptides was resolved at MW 17,500 (figure 2B, lanes 1 and 5), a protein of apparent MW 16,600 in acidic extracts reacted with antiserum to all three synthetic peptides (figure 2B, lanes 2-4). In addition, a 15,500 MW protein in acidic extracts reacted with antiserum to the internal and carboxylterminal peptides (figure 2B, lanes 3, 4) but not with antiserum to the amino-terminal peptide (figure 2B, lane 2). DISCUSSION
A growth factor has been identified in human prostate and testes, and in calf testes that is structurally related to bFGF. The evidence for this relationship is 1) that the growth factor activity, like bFGF, is heat and acid labile 15 but not affected thiol-reducing agents, 16 2) that the growth factor binds heparin and requires L4 M to 1.8 M sodium chloride for dissociation, 17 and 3) that antisera prepared against synthetic peptides with sequence homology to amino acid residues 33-43, as well as amino-terminal sequences 1-12 and carboxyl-terminal sequences 136-145 of bovine BFGF react with the prostate grnwth factm and the human and calf testes-derived growth factors (figure 2). Stability studies (table 2) suggest that the same grmvth factor is also present in epididymis. Unfortunately, large variations in the amount of activity in replicate preparations of testes and epididymides (table 1) does not permit assessment of the relative amount of growth factor in various tissues. It is not clear if variations in activity, not seen in prostate, is due to actual differences in the amount of growth factor (or inhibitors) in individual tissues or due to differences in the duration of storage or to uncontrolled differences during homogenization. Possible differences in the amount of growth factor in individual human prostate tissues would not be evidence since homogenates were prepared from pools of 20 or more specimens; whereas, human testis and epididymis preparations were from two to five patients. The availability of the latter tissues also necessitated larger variations in the duration of storage prior to preparing homogenates. However, these factors do not account for the differences among calf tissue preparations. Like prostate, calf tissue homogenates
25.7
2 FIG. 2. Detection of growth factor by antisera to synthetic peptides with sequence homologies to bFGF. Partially purified growth factor preparation (1500 to 3000 units/lane) were electrophoresed by SDSp AGE. Proteins were transferred electrophoretically to nitrocellulose paper and incubated with antisera against synthetic peptides with sequence homologies to bovine bFGF. Bound antibody was visualized by incubation with peroxidase-conjugated goat anti-rabbit IgG. Molecular weight markers were visualized in the original gel by silver staining. A, growth factor from alkaline extracts of human prostate (lane 1) and alkaline extracts of human testes (lanes 2-4). Lanes 1 and 2 were incubated with antiserum to amino-terminal peptide; lime 3, internal peptide; lane 4, carboxyl-terminal peptide. B, growth factor from alkaline extracts of calf testes (lanes 1, 5) and acidic extracts of calf testes (lanes 2-4). Lanes 1 and 2 were incubated with antiserum to amino-terminal peptide; lane 3, internal peptide; lanes 4, 5, carboxyterminal peptide.
were prepared from pools of a large number of animals and the time of storage was short. Regardless of the source, all tissue homogenates show toxicity when tested at high concentrations in the bioassay. Since quantitation of growth factor depends on radionucleotide uptake that is influenced not only by the amount of mitogen present but also by the amount of inhibitor present, it is possible that subtle differences in tissue handling during storage or homogenization could result in different amounts of these agents in homogenates and influence growth factor quantitation. The procedure used to prepare tissue homogenates influenced the amount of activity recovered. Homogenates of human prostate in high ionic strength alkaline buffer containing protease inhibitors had eight-fold higher levels of activity than when tissue was homogenized under acidic conditions in the absence of protease inhibitors. On the other hand, acidic homogenates of calf testes had, on the average, more activity than did
426
STORY AND ASSOCIATES
homogenates prepared by the alkaline extraction procedure (table 1). The difference in behavior of human and bovine tissue under the two extraction conditions may reflect species differences in growth factor structure or differences in tissue specific proteases activated during extraction. Basic FGF has been purified to homogeneity from a number of bovine 10• 17• 24 and human 25- 27 tissues. The sequence of bovine bFGF is known 28 and the sequence of the human growth factor has been predicted from the cloned cDNA. 29 Of the 146 amino acids in the mature proteins only the amino acids at residues 112 and 128 differ. These small changes in the amino acid sequence could influence acid stability or alter susceptibility of the growth factor to tissue specific proteases. The homogenization method also affected the structure of the growth factor isolated (figure 2). Alkaline extraction in the presence of protease inhibitors of human prostate resulted in the isolation of a large form of bFGF. Amino-terminal sequencing9 showed that the growth factor has eight additional amino acids at its amino-terminus not found in bFGF 28 but whose sequence was predicted from the nucleotide sequence by molecular cloning. 29 • 30 Immunodetection of a growth factor of comparable size in alkaline extracts of human and calf testes (figure 2) suggests that the same extended form of the growth factor is present in these tissues, although this has yet to be confirmed by sequence analysis. Antibody reactivity with smaller MW proteins suggests that proteolysis was not completely inhibited in these preparations. Ueno et al. 31 isolated two forms of bFGF from acidic extracts of bovine pituitary when protease inhibitors were included during extraction. One form was identified as (1-146)bFGF and the other was an amino-terminal extended form. When protease inhibitors were omitted during homogenization of calf testes, two forms of the growth factor were detected by immunostaining (figure 2B). One form, apparent MW 16,600, reacted with antisera to all three synthetic bFGF peptides and was likely (1-146)bFGF. A smaller 15,500 MW form reacted with antisera to the internal and carboxyl-terminal peptides, but not with antiserum to the amino-terminal peptides; suggesting that the protein lacked amino-terminal bFGF sequences. A truncated form of bFGF lacking the amino-terminal 15 amino acids has been isolated from bovine adrenal, 24 corpus luteum 10 and kidney22 homogenates prepared without protease inhibitors. It is likely that the 15,500 MW peptide seen in calf testes homogenates is the (16145) bFGF formed by proteolysis of the large precursor form. During preparation of this paper, Ueno et al. 32 reported the purification ofbFGF from bovine testes. The growth factor was purified in the absence of protease inhibitors and found to be the amino-terminal 15-amino acid deleted form. Our previous reports 8 • 9 and the evidence presented here strongly suggest that the growth factor in human prostate, previously called prostatic growth factor, PrGF, 2- 5 and the growth factor in the seminiferous epithelium of the calf and several mammalian species, previously called seminiferous growth factor, SGF, 6 • 7 is bFGF. PrGF has been suggested33 to function in a paracrine manner in the development of the fibrostomal nodule characteristic of BPH. 34 SGF, reported to be present in higher amounts in Sertoli cells, has been suggested to mediate proliferation of spermatogonia and spermatocytes to ensure an orderly, continuous, and abundant production of spermatozoa. 7 It has yet to be determined if the growth factor plays a physiologic role in these and other developmental processes. Fibroblast growth factors are known to be mitogenic for cultured endothelial cells 18 and to be angiogenic factors in vivo. 28 In addition, cultured endothelial cells have recently been found to produce the growth factor. 36- 38 The methodology that we have employed in demonstrating bFGF in genitourinary organs does not preclude the possibility that endothelial cells may be the source of the growth factor. If endothelial cells produce bFGF in vivo, this may explain the presence of the
growth factor in a wide variety of tissues. Further studies employing cultured cells derived from tissues and the application of immunohistochemical techniques to localize bFGF are required to identify the cell type responsible for bFGF production in vivo. Acknowledgments. We wish to thank Bonnie Livingston, Susan Swartz and Tom Tebo for expert technical assistance. We are indebted to Nacker Packing Co., Inc., Milwaukee, WI, for their generous supply of calf testes. REFERENCES
1. Jacobs, S. C., Pikna, D. and Lawson, R. K.: Prostatic osteoblastic factor, Invest. Urol., 17: 195, 1979. 2. Tackett, R. E., Heston, W, D. W., Parrish, R, F., Pletscher, L. S, and Fair, W, R.: Mitogenic factors in prostatic tissue and expressed prostatic secretion. J. Urol., 133: 45, 1985. 3. Story, M. T., Jacobs, S, C.and Lawson, R. K.: Partial purification of a prostatic growth factor. J. Urol., 132: 1212, 1984. 4. Nishi, N., Matuo, Y., Muguruma, Y., Yoshitake, Y., Nishikawa, Y. and Wada F.: A human prostatic growth factor (hPGF): partial purification and characterization. Biochem. Biophys. Res. Commun., 132: 1103, 1985. 5. Jinno, H., Ueda, K., Otaguro, K., Kato, T., Ito, J. and Tanaka, R: Prostate growth factor in the extracts of benign prostatic hypertrophy, Eur. Urol., 12: 41, 1986. 6, Feig, L.A., Klabsbrun, M. and Bellve, A. R.: Mitogenic polypeptide of the mammalian seminiferous epithelium: biochemical characterization and partial purification, J. Cell. Biol., 97: 1435, 1983. 7. Bellve, A. R. and Feig, L. A.: Cell proliferation in the mammalian testes: biology of the seminiferous growth factor (SGF). In: Recent Prog. Hormone Res. Edited by R.O. Greep. New York: Academic Press, vol. 40, pp. 534-561, 1984. 8. Story, M. T., Sasse, J,, Jacobs, S. C. and Lawson, R. K.: Prostatic growth factor: purification and structural relationship to basic fibroblast growth factor. Biochemistry, 26: 3843, 1987. 9, Story, M. T., Esch, F., Shimasaki, S., Sasse, J., Jacobs, S. C. and Lawson, R. K.: Amino-terminal sequence of a large form of basic fibroblast growth factor isolated from human benign prostatic hyperplastic tissue. Biochem. Biophys. Res. Commun., 142: 702, 1987. 10. Gospodarowicz, D., Cheng, J., Lui, G, M., Baird, A., Esch, F. and Bohlen, P.: Corpus luteum angiogenic factor is related to fibroblast growth factor. Endocrinology, 117: 2383, 1985. 11. Laemmli, U.K.: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London), 227: 680, 1970. 12. Wray, W., Boulikas, T., Wray, V. P. and Hancock, R.: Silver staining of proteins in polyacrylamide gels. Anal. Biochem., 118: 197, 1981. 13. Merrifield, R. B.: Solid phase peptide synthesis. 1. The synthesis of a tetrapeptide. J. Am. Chem. Soc., 85: 2149, 1963. 14. Sakakibara, S.: The use of hydrogen fluoride in peptide chemistry. In: Chemistry and Biochemistry of Amino Acid, Peptides and Proteins. Edited by B. Weinstein. New York, Dekker, vol. 1, p. 51, 1971. 15. Gospodarowicz, D., Bialecki, H. and Greenburg, G.: Purification of the fibroblast growth factor activity from bovine brain. J. Biol. Chem., 253: 3736, 1978. 16. Maciag, T., Cerundolo, J., Ilsley, S., Kelley, P. R. and Forand, R.: An endothelial cell growth factor from bovine hypothalamus: identification and partial characterization, Proc. Natl. Acad. Sci. USA, 76: 5674, 1979. 17. Lobb, R., Sasse, J., Sullivan, R., Shing, Y., D'Amore, P., Jacobs, J. and Klagsbrun, M.: Purification and characterization ofheparinbinding endothelial cell growth factors. J. Biol. Chem., 261: 1924, 1986. 18. Gospodarowicz, D., Cheng, J., Lui, G-M., Baird, A. and Bohlen, P.: Isolation of brain fibroblast growth factor by heparin-Sepharose affinity chromatography: identity with pituitary fibroblast growth factor. Proc. Natl. Acad. Sci. USA, 81: 6963, 1984. 19. Lobb, R. R. and Fett, J. W.: Purification of two distinct growth factors from bovine neural tissue by heparin affinity chromatography. Biochemistry, 23: 6295, 1984. 20. Klagsbrun, M. and Shing, Y.: Heparin affinity of anionic and
BASIC FGF II~ TESTES
21.
22.
23.
24.
25.
26. 27.
28.
cationic capillary endothelial ceil growth factors. Proc. Natl. Acad. Sci. USA, 82: 805, 1985. Bohlen, P., Baird, A., Esch, F., Ling, N. and Gospodarowicz, D.: Isolation and partial molecular characterization of pituitary fibroblast growth factor. Proc. Natl. Acad. Sci. USA, 81: 5364, 1984. Baird, A., Esch, F., Bohlen, P., Ling, N. and Gospodarowicz, D.: Isolation and partial characterization of an endothelial growth factor. Regul. Pept., 12: 201, 1985. Baird, A., Esch, F., Gospodarowicz, D. and Guillemin, R.: Retinaand eye-derived endothelial cell growth factors: partial molecular characterization and identity with acidic and basic fibroblast growth factor. Biochemistry, 24: 7855, 1985. Gospodarowicz, D., Baird, A., Cheng, J., Lui, G. M., Esch, F. and Bohlen, P.: Isolation of fibroblast growth factor from bovine adrenal gland: physicochemical and biological characterization. Endocrinology, 118: 82, 1986. Bohlen, P., Esch, F., Baird, A., Jones, K. L. and Gospodarowicz, D.: Human brain fibroblast growth factor: isolation and partial chemical characterization. FEBS Letters, 185: 177, 1986. Gospodarowicz, D., Cheng, J., Lui, G-M., Fujii, D. K., Baird, A. and Bohlen, P.: Fibroblast growth factor in the human placenta. Biochem. Biophys. Res. Commun., 128: 554, 1985. Klagsbrun, M., Sasse, J., Sullivan, R and Smith, J. A.: Human tumor cells synthesize an endothelial cell growth factor that is structurally related to basic fibroblast growth factor. Proc. Natl. Acad. Sci. USA, 83: 2448, 1986. Esch, F., Baird, A., Ling, N., Ueno, N., Hill F., Denoroy, L., Klepper, R., Gospodarowicz, D., Bohlen, P. and Guillemin, R.: Primary structure of bovine pituitary basic fibroblast growth factor (FGF) and comparison with the amino-terminal sequence of bovine brain acidic FGF. Proc. Natl. Acad. Sci. USA, 82: 6507, 1985.
427
29. Kmokmva, T., Sasada, R., Iwane, NL and Igarashi, K: Cloning and expression of cDNA encoding human basic fibroblast growth factor. FEBS Letters, 213: 189, 1987. 30. Abraham, J. A., Mergia, A., Whang, J.L., Tumolo, A., Friedman, J., Hjerrild, K. A., Gospodarowicz, D. and Fiddes, J. C.: Nucleotide sequence of a bovine clone encoding the angiogenic protein, basic fibroblast growth factor. Science, 233: 545, 1986. 31. Ueno, N., Baird, A., Esch, F., Ling, N. and Guillemin, R.: Isolation of an amino-terminal extended form of basic fibroblast growth factor. Biochem. Biophys. Res. Commun., 138: 580, 1986. 32. Ueno, N., Baird, A., Esch, F., Ling, N. and Guillemin, R.: Isolation and partial characterization of basic fibroblast growth factor from bovine testes. Mo!. Cell. Endocrinol., 49: 189, 1987. 33. Jacobs, S. C. and Lawson, R. K.: Mitogenic factor in human prostate extracts. Urology, 16: 488, 1980. 34. Moore, R. A.: Benign hypertrophy of the prostate: a morphological study. J. Urol., 50: 680, 1943. 35. Pradhan, B. K. and Chandra, K.: Morphogenesis of nodular hyperplasia prostate. J. Urol., 113: 210, 1975. 36. Schweigerer, L., Neufeld, G., Friedman, J., Abraham, J. A., Fiddes, J. C. and Gospodarowicz, D.: Capillary endothelial cells express basic fibroblast growth factor, a mitogen that promotes their own growth. Nature, 325: 257, 1987. 37. Baird, A. and Ling, N.: Fibroblast growth factors are present in the extracellular matrix produced by endothelial cells in vitro: implications for a role of heparinase-like enzymes in the neovascular response. Biochem. Biophys. Res. Commun., 142: 428, 1987. 38. Vlodavsky, I., Folkman, J., Sullivan, R., Friedman, R., Ishai-Michaeli, R., Sasse, J. and Klagsbrun, M.: Endothelial cell-derived basic fibroblast growth factor: synthesis and deposition into subendothelial extracellular matrix. Proc. Natl. Acad. Sci. USA, 84: 2292, 1987.