GLIA 4:378-383 (1991)
FGFs Stimulate IGF Binding Protein Synthesis Without Affecting IGF Synthesis in Rat Astroblasts in Primary Culture CAMILLE LORET,~THJERRYJANET: GERARD LABOURDETTEP HELENE SCHNEID,~AND MICHEL BINOUXl 'Institut National de la Sante et de la Recherche Medicale, Unite de Recherches sur la Regulation de la Croissance, U.142, Hopital Saint Antoine, 75571 Paris Cedex 12 and 'Centre de Neurochimie du CNRS and 31NSERM U.44, 67084 Strasbourg Cedex, France
KEY WORDS
Insulin-like,Astrocyte, Growth, Differentiation
ABSTRACT The production of insulin-like growth factor (1GF)-I and -11 and their binding proteins (BPs) has been studied in new-born rat astroblasts at confluency in primary culture. Under the influence of fibroblast growth factors (FGFs)(acidicand basic),the morphology of the astroblasts was altered, 1251-deoxyuridineincorporation was increased, and glutamine synthetase activity was stimulated. IGF production and IGF mRNA expression remained unchanged. Production of the 32 kDa BP (IGFBP-2),the sole or predominant form under base-line conditions, was enhanced and the 43-39 kDa forms (IGFBP-3) appeared or were increased.Epidermal growth factor (EGF)also stimulated production of these BPs, whereas thrombin and db-CAMP had no effect. Our data suggest that a relationship exists between FGF-induced maturation of astroblasts and the forms of BP they produce. The data also indicate that some factors may act specifically on BP synthesis, without affectingIGF synthesis, and in this way play a role in regulating the bioavailability of the IGFs.
INTRODUCTION
maturation, and differentiation of these cells (Loret et al., 1989b). In this study we set out to investigate the modulation by growth factors, and in particular the FGFs, of the biosynthesis of IGFs and their BPs in astroblasts in primary culture. Parallel studies were done of these factors' effects on cell morphology and growth, as well as on glutamine synthetase activity, which is peculiar to astrocytes in the brain.
Following our initial work on the production of insulin-like growth factors (IGFs)and their binding proteins (BPs) by rat pituitary gland and brain explants in culture (Binoux et al., 19811, numerous studies confirmed the synthesis of IGFs by nervous tissue and their involvement in the control of nerve cell proliferation and maturation (reviews in Daughaday and Rotwein, 1989; Oomura and Plata-Salaman, 1987). We subsequently identified several molecular forms of BP in the culture MATERIALS AND METHODS media of human nervous tissue. The major form in Rat Astroblast Cultures nervous tissue also predominates in cerebrospinal fluid and has a preferential affinity for IGF-I1 (Hossenlopp Primary cultures of new-born rat astroblasts were et al., 1986a). In the brain, astrocytes constitute the major cell type. established according to Booher and Sensenbrenner We have reported that certain growth factors,like acidic and basic fibroblast growth factors (aFGF and bFGF), Received August 15,1990; accepted November 9,1990. provoke morphological changes in rat astroblasts in Address reprint re uests t o C. Loret, INSERM U.142, HBpital Saint Antoine, primary culture and, in addition, stimulate the growth, 184 rue du Faubourgiaint Antoine, 75571 Paris Cedex 12, France. 0 1 9 9 1 Wiley-Liss, Inc.
REGULATION OF IGF BP PRODUCTION IN ASTROBLASTS
(1972), as modified previously (Loret et al., 1989a). After 20 days of culture in Waymouth's MD 705/1 medium (Flow Laboratories) supplemented with 110 mg/ml sodium pyruvate, 50 unitdm1 penicillin, 50 pg/ml streptomycin, and 10% fetal calf serum (Gibco), the confluent astroblasts were switched to a chemically defined medium for performance of the experiments (Weibel et al., 1984). This was the same medium, lacking serum and supplemented with 10 bg/ml IGF-free bovine serum albumin (BSA) (bioMerieux,France) with or without one of the following factors: 5 ng/ml acidic FGF, 5 ng/ml basic FGF (R & D, Minneapolis, MN, USA), 10 ng/ml EGF (Sigma, USA), 2.5 unitdm1 thrombin (gift from J.M. Freyssinet, Centre de Transfusion Sanguine, Strasbourg, France), or 1 mM db-CAMP (Sigma). The culture was then continued for a further 4 days.
379
Competitive binding studies Binding studies were done with BPs extracted from control astroblast culture media. The amounts of BP bindin that were approximately 20-25% 1251-IGF-I and 124-IGF-IIwere determined, after which displacement curves were established using increasing concentrations of IGF-I and IGF-11, each in duplicate. DNA Probes
A human IGF-IA 1,073-bp cDNA insert was isolated from a human liver library (Le Bouc et al., 1986) and subcloned in PGEM 4 (pTG 3906). The 660-bp EcoRIBamHI fragment containing exons 1, 2, 3, and 5 was used. For IGF-11, a human IGF-11663-bp cDNA insert was isolated from a human placental library (Le Bouc et al., 1987) and subcloned in PGEM 4 (pTG 3907). This contains part of exon 4, all of exons 5 and 6, and part of 5-['2511 Iodo-2-Deoxyuridine Incorporation exon 7. cDNA probes were labelled by random priming with After a 16-h culture in the defined medium, 0.5 T P act. 0.5-1 mCi/pg DNA, Kit, AmerpCi/ml 5-[1251]iodo-2-deoxyuridine(1251-dUrd) (spec. C I ~ ~ P - ~ A(spec. act. 5 Wmg, Amersham, England) was added. After a sham). further 24 h, the dishes were rinsed 3 times with 0.9 M NaCl and the cells were scraped off and sedimented. The Northern Blotting radioactivity of the pellets was counted in a gamma counter. Total RNA was isolated according to Chomczynski and Sacchi (1987), using a single extraction with acid guanidium thiocyanate/phenoUchloroform.Total RNA Glutamine Synthetase and Protein Assays samples (12 pg) were submitted to electrophoresis on a After 4 days of culture in the defined medium, cells 1%agarose gel containing 6% formaldehyde and then were rinsed 3 times with 0.9 M NaC1, sedimented, and transferred onto a nylon membrane (Hybond N, Amerfrozen at -20°C. The glutamine synthetase (GS) assay sham). After hybridization, the blots were washed to was performed according to Miller et al. (1978), with final stringency with 0.1 x SSPE (15 mM NaCl, 1 mM slight modifications (Loret et al., 1989a). Proteins were phosphate buffer, pH 6.8,O.l mM EDTA), 0.1% SDS at measured according to Lowry et al. (19511,using BSA as 48°C for 30 min and then exposed to X-ray film (Curix RP1, Agfa) with two intensifying screens (DuPont standard. Cromex Hiplus). IGF Assay and Competitive Binding Studies
Western Ligand Blotting The methods used have been described in detail elseThe method used has been described in detail by where (Binoux et al., 1984, 1986). Samples (7 ml) were collected after 4 days' culture in defined medium, de- Hossenlopp et al. (198610).After SDS-polyacrylamidegel salted, lyophilized, and then submitted to acidic gel electrophoresis (without reducing agent, except in the case of the markers), proteins were electroblotted onto a filtration in order t o separate the IGFs and the BPs. nitrocellulose sheet and the BPs identified using lZ5IIGF-I1 as a probe. IGF assays Pure IGF-I and IGF-I1 from human plasma were a generous gift from R. Humbel (Zurich, Switzerland). IGF-I was determined by radioimmunoassay (RIA) using the antiserum prepared by J.J. Van Wyk (Chapel Hill, NC, USA) and IGF-I1 was measured by protein binding assay using BPs extracted from human cerebrospinal fluid (CSF),which have a selective affinity for IGF-I1(Binoux et al., 1986).
RESULTS Effects of Growth Factors and db-CAMP on Cell Morphology, Growth, and GS Activity Only the FGFs, epidermal growth factor (EGF), and db-CAMP induced significant morphological changes after 4 days in the defined medium. These effects, involving reduced cell body size and the appearance of
380
LORET ET AL.
processes, have been reported previously (Loret et al., 1988,1989a). In three experiments, mitogenic activity, as reflected by 1251-deoxyuridineincorporation measured after 40-h culture, was stimulated by a factor of approximately 6 by the FGFs and thrombin and by a factor of 3 4 by EGF, whereas db-CAMP had an inhibitory effect on the residual proliferation of the cells (Fig. 1A). GS activity, by contrast, was most strongly stimulated by EGF and db-CAMP (by factors of about 7 and 6, respectively). aFGF and bFGF provoked 3-4-fold stimulation, and thrombin only 1.5-2 (Fig. 1B). Expression of IGFs and Their mRNAs Under control conditions the rat astroblasts secreted both IGF-I and IGF-I1into the culture medium. None of the factors of hormones tested had any effect on IGF production. Fig. 2A shows the IGF concentrations detected in the culture media of control and aFGF- or bFGF-treated cells. Despite wide variations in the amounts of IGF secreted from one experiment to another in both treated and untreated cultures, the ratio of IGF-IYIGF-I concentrations remained close to 2 (Fig. 2B). mRNA expression was determined by Northern blotting. IGF-I1 mRNAs were more abundant than IGF-I mRNAs (Fig. 3). For the IGF-I1 gene, four lengths of mRNA were detected, of 4.9,2.1,1.5, and 1kb, the 4.9 kb messenger being the major form. For IGF-I, two sizes of mRNA of approximately 4.1 and 1.7 kb were found. None of the factors tested had any effect on their expression (Fig. 3). Modulation of BP Secretion The BPs secreted into the culture medium by the astroblasts were analyzed by Western ligand blotting and compared with those found in serum, which reflects liver production in vivo. In serum there are six molecular forms. Three of these, of 43,41, and 39 kDa, represent different states of glycolysationof the same protein, as determined in experiments using N-glycanase (data not shown). The other three, of 32,29, and 24 kDa, are distinct species. The migration profiles of BPs secreted by control astroblasts were similar to those of BPs in new-born rat serum, with a preponderant 32-kDa form appearing as a doublet. The 43-39-kDa triplet and the 24-kDa BP were absent or barely detectable. Treatment with the FGFs increased the 32-kDa BP and resulted in the appearance or increase of the 43-39kDa form. Among the other growth factors, only EGF induced the appearance of the high molecular weight species. The remainder had no visible effect (Fig. 4). The binding characteristics of the major BP secreted were investigated using the culture medium of un-
Y
Control a F G F
bFGF
Thr
E G F dbcAMP
Fig. 1. Effects of acidic fibroblast growth factor (aFGF), basic fibroblast growth factor (bFGF), e idermal growth factor {EGF), thrombin (Thr) and db-cAMP on [12511-~eoxyundine 1251dURdincorporation (A) and glutamine srthetase (GS) actifity (B) in rat astroblast cultures. Factors were ad ed at day 20 as described in Materials and Methods.A typical experiment is shown. Error for triplicate dishes was below 10% for incorporation and below 15%for the lutamine synthetase assay. Similar results were obtained in two furtier experiments.
treated astroblasts (in which only the 32-kDa BP was present) after acidic extraction. The BP proved to bind IGF-I1 preferentially, whichever tracer was used (Fig. 5). The affinity constants for IGF-I and IGF-11, calculated from Scatchard analysis, were 0.4 x loloM-l and 0.8 x 10" M-l (mean of two experiments). Finally, it is worth mentioning here that other growth factors and hormones tested [platelet-derived growth factor (PDGF), transforming growth factorp (TGFP), triiodotyronine, hydrocortisone] had no effect on either IGF production, mRNA expression, or BP secretion (results not shown).
REGULATION OF IGF BP PRODUCTION IN ASTROBLASTS
A
11
1.1
381
B
r = 0.74
0
p < 0.001
1
./
1
a /
0
0.9
--
0.7
-
. 0
0.7
/
v
./ /
0
\
0.6
C
-
0
E
0.6
/
W
Y
LL (3
/
0.8
h
E
/
/
0.9
0.8
/ /
I
0.5
0.5
(3 0.4
0.4
0.3
0.3
0.2
0.2 ’
01
0.1’
0
0 Control
a FGF
b FGF
n=13
n = 9
n=lO
I
I
1
1
I
1
0.1
0.2
0.3
0.4
0.5
0.6
IGF
f
0.7
- I(ng/ml)
Fig. 2. A: Effects of acidic fibroblast owth factor (aFGF) and basic fibroblast growth factor (bFGF) on cumulative insulin-like growth factor (gF1-1(hatched columns) and IGF-I1production (0 en columns)by rat astroblasts after 4 days’ culture (mean 2 SD). B: Relationships between IGF-I1and &F-I concentrations after 4 days’culture in defined medium without (controls)and with aFGF, bFGF, EGF, thrombin, or db-CAMP.
DISCUSSION Under our experimental conditions, only the FGFs and EGF were capable of inducing changes in cell 4.9 + 4.1 + morphology, increased proliferation, and enhanced GS activity (which reflects cell maturation). The other reg2.1 + 1.7+ 1.5 + ulatory factors tested had partial effects: thrombin 1.0 * failed to affect cell morphology, and db-CAMP, which is the second messenger for norepinephrine (Narumi kb kb et al., 1978), had no effect on proliferation. These findings confirm and expand our previously published data IGF I IGF II (Loret et al., l988,1989a, b). Fig. 3. Northern blot analysis of insulin-like growth factor (1GF)-I Astroblasts in primary culture secreted twice as much and IGF-I1mRNAs isolated from rat astroblast cultures after 4 days in IGF-I1 into the culture medium as IGF-I. Also, IGF-I1 defined medium without (controls) and with acidic fibroblast growth factor (aFGF) or basic fibroblast growth factor (bFGF).Exposure time mRNAs were more abundant than IGF-I mRNAs. These was 8 days for IGF-I mRNAs and 4 days for IGF-I1mRNAs. findings differ from those of Ballotti et al. (1987) who
382
LORET ET AL.
serum
1.0 -
Culture medium
l Z 5 1 - 1 I~ ~
0.8 -
0
0.6 -
F4
BPs
0.1
(kDa) 43 + 41 + 3 9 .)
24
-
0.2 -
+
1.0-
0.8 -
Fig. 4. Western ligand blot analysis of binding proteins (BPs) in rat astroblast culture media. Profiles in serum are shown for purposes of comparison. For abbreviations, see Fig. 1legend.
0
F4 F?
0.6-
0.4 -
observed no IGF-I1mRNA in rat astrocytes in secondary culture. None of the regulatory factors tested had any effect on either IGF production or expression of their mRNAs. Under all the experimental conditions used, production of the two IGFs mirrored expression of their mRNAs. IGF production was accompanied by production of some of the BPs. In the control astroblasts, it was mainly the 32-kDa BP that was secreted. This corresponds to the BP purified from the culture media of rat liver BRL 3A cells, whose cDNA has recently been cloned (Ballard et al., 1989; Brown et al., 1989; Margot et al., 1989)and which is now known as rat IGFBP-2. It has a preferential affinity for IGF-11,which, however, is less marked than that of the correspondinghuman form predominating in CSF and the central nervous system (Hossenloppet al., 1986a). Other authors have reported BP secretion by rat astrocytes in culture: Han et al. (1988) used affinity cross-linking to analyze the BPs they detected and their data can therefore not be compared with ours; Ocrant et al. (1989) used Western ligand blotting and their profiles largely resemble ours. The new findings emerging from this study concern stimulation of BP secretion by the growth factors that influence the proliferation and maturation of astroglial cells. Among the factors tested, only aFGF, bFGF, and EGF provoked increased production of the 32-kDa BP and the appearance or increase of the 43-39-kDa triplet. This triplet corre-
0.2 -
0.0 , 0.1
B I
0.5
1.0
2.0
ng of IGF added per tube Fig. 5. Com etitive inhibition b insulin-like owth factor (1GF)-I and -11 of the finding of 1251-IGF-P(A) and IGF-ff (B) to IGF binding protein (IGF BP)-2extracted from the culture medium ofuntreated rat astroblasts (controls).The amount used er tube was that giving 25% bindin of tracer in the absence of coldPGF (B,,). This corres onded to 19 pf eqnuivalent of culture medium for the bindin of '251-IGP-Iand 16 p1 fort e binding of 1251-IGF-II.The competitivefinding curves for the two tracers were obtained from the same experiment and similar results were obtained from a second experiment.
sponds to the preponderant form in adult rat serum, known as IGFBP-3 (Ballard et al., 1989; Hossenlopp et al., 1987). In serum it appears to be GH-dependent, unlike IGFBP-2, which predominates in fetal and newborn serum (Hossenloppet al., 1987).The similar effects of the FGFs and EGF on BP secretion may be accounted for by the fact that their receptors both have tyrosine kinase activity (Ushiro and Cohen, 1980; Rozengurt, 1986). Also, in astroblasts cultured for 20 days, they exhibit comparable modulation of the synthesis of the
REGULATION OF IGF BP PRODUCTION IN ASTROBLASTS
383
major proteins seen in two-dimensional LDS-polyization and evidence of biologic function. J. Neuroscience, 8:3135-3143. acrylamide gel electrophoresis (Loret et al., 1989a). Hossenlop P., Seurin, D., Segovia-Quinson, B., and Binoux, M. Our observations suggest that some relationship ex(1986a) hentification of an insulin-like growth factor bindin tein in human cerebrospinalfluid with a selective affinity for If$? ists between the forms of BP secreted by astroblasts and FEBS Lett., 208:439-444. the stage of cell maturation regulated by aFGF, bFGF, Hossenlopp, P., Seurin, D., Segovia-Quinson, B., Hardouin, S., and or EGF. They also show that certain factors are capable Binoux, M. (198613) Anal sis of serum insulin-like growth factor binding proteins using Jestern blotting: Use of the method of of acting specifically on BP synthesis without affecting titration of the binding proteins and competitive binding studies. IGF synthesis and are therefore involved in regulating Anal. Biochem., 154:138-143. Hossenlo P., Seurin, D., Se ovia, B., Portolan, G., and Binoux, M. the bioavailability of the IGFs.
ACKNOWLEDGMENTS We are grateful to C. Lassarre, B. de Galle, and M.F. Knoetgen for their technical assistance. C. Loret is a recipient of a grant from the Fondation pour la Recherche Medicale.
REFERENCES Ballard, J., Baxter, R., Binoux, M., Clemmons, D., Drop, S., Hall, K., Hintz, R., Rechler, M., Rutanen, E., and Schwander, J. (1989) On the nomenclature of the IGF binding proteins. Acta Endocrinol. (Copenh.),121:751-752. Ballotti, R., Nielsen, F.C., Pringle, N., Kowalski, A,, Richardson,W.D., Van Obberghen, E., and Gammeltoft, S. (1987) Insulin-like growth factor I in cultured rat astrocytes: Expression of the gene, and receptor tyrosine kinase. EMBO J., 6:3633-3639. Binoux, M., Hossenlopp, P., Lassarre, C., and Hardouin, N. (1981) Production of insulin-like growth factors and their carrier by rat pituitary gland and brain explants in culture. FEBS Lett., 124:178-183. Binoux, M., Seurin, D., Lassarre, C., and Gourmelen,M. (1984) Preferential measurement of insulin-like growth factor (IGF) I-related peptides in serum with the aid of IGF binding proteins (IGF BPs) produced by rat liver in culture. Estimation of serum IGF BP levels. J. Clin. Endocrinol.Metab., 59:453-462. Binoux, M., Lassarre, C., and Gourmelen, M. (1986) Specific assay for insulin-like growth factor (IGF) I1 usin the IGF binding roteins extracted from human cerebrospinal fiuid. J. Clin. Enjocrinol. Metab., 63:1151-1155. Booher, J. and Sensenbrenner, M. (1972) Growth and cultivation of dissociated neurons and glial cells from embryonic chick, rat and human brain in flask cultures. Neurobiology,2:97-105. Brown, A.L., Chiariotti, L., Orlowski, C.C., Mehlman, T., Burgess, W.H.,Ackerman,E.J., Bruni, C.B., andRechler, M.M. (1989) Nucleotide sequence and expression of a cDNA clone encoding a fetal rat binding protein for insulin-like growth factors. J . Biol. Chem., 2645148-5154. Chomczynski, P. and Sacchi, N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroformextraction. Anal. Biochem., 162:15&159. Dauehadav. W.H. and Rotwein. P. (1989) Insulin-like mowth factors I an% 11. Pe tide, messenger kbonucleic acid and gene structures, serum, anttissue concentrations.Endocrinol. Rev., 10:6&91. Han, V.K.M., Lauder, J.M., and DErcole, J. (1988) Rat astroglial somatomedidinsulin like growthfactor binding proteins: Character-
(1987) Fiterogeneity of insdin-like growth factor binding proteins and relationships between structure and affinity. 2. Forms released b human and rat liver in culture. Eur. J. Biochem., 170:133-142. Le Bouc, Y., Dreyer, D., Jaeger, F., Binoux, M., and Sonderme er, P (1986) Complete charaderization of the human IGF-I nuccotide se uence isolated from a newly constructed adult liver cDNA library. F&SLett., 196:108-112. Le Bouc, Y., Noguiez, P., Sondermeyer,P., Dreyer, D., Girard, F., and Binoux, M. (1987) A new 5’ non-coding region for human lacental insulin-like growth factor I1 mRNA expression. FEBS Lett., 222:181-185. Loret, C., Sensenbrenner, M., and Labourdette, G. (1988) Maturationrelated gene expression of rat astroblasts in vitro studied b two dimensional polyacrylamide gel electrophoresis. Cell 5iffer.i 2k37-46. Loret, C., Sensenbrenner, M., and Labourdette, G. (1989a)Differential enotypic expression induced in cultured rat astroblasts by acidic ibroblast growth factor, epidermal growth factor and thrombin. J. BWZ.Chem., 264:8319-8327. Loret, C., Laeng, P., Sensenbrenner, M., and Labourdette, G. (1989b) Acidic and basic fibroblast growth factors similarly regulate the rate of bios thesis of rat astroblast proteins. FEBS Lett., 257:324-328. Lowry, Rosebrough, N.J., Farr, A.L., and Randall, R.J. (1951) Protein measurement with the Folin phenol reagent. J . Biol. Chem., 193:265-275. Mar ot, J.B., Binkert, C., Mary, J.L., Landwehr, J., Heinrich, G., and Scfwander, J. (1989) A low molecular weight insulin-like y w t h factor binding protem from rat: cDNAcloningand tissue distn ution of its messenger RNA. Mol. Endocrinol., 3:1053-1060. Miller, R.E., Mackenberg, R., and Gershman, H. (1978) Regulation of dutamine svnthetase in cultured 3T3-Ll cellsbv insulin. hvdrocortilone and dibutyryl cyclic AMP. Proc. Nati. Acad.’Sci. USA, 75:1418-1422. Narumi, S.. Kimelberg, H.K., and Bourke, R.S. (1978) Effects of norepinephrine on the morphology and some enzyme activities of primary monolayer cultures from rat brain. J. Neurochem., 31:147%1490. Ocrant, I., Pham, H., Oh, Y., and Rosenfeld, R.G. (1989) Characterization of insulin-like growth factor binding proteins of cultured rat astroglial and neuronal cells. Biochem. Biophys. Res. Commun., 159:1316-1322. Oomura, Y. and Plata-Salaman, C.R. (1987) Somatomedins (insulinlike growth factors)and the nervous system. In: Insulin, Insulin-Like GrowthFactors, and Their Rece tors in the CentralNervous System. M.K. Raizada, I.M. Phillips, anfD. LeRoith, eds. Plenum Press, New York, pp. 215-244. Rozengurt, E. (1986) Early signals in the mitogenic response. Science,
P
&.,
224.161-166 -- - .- -- - - - .
Ushiro, H. and Cohen, S. (1980) Identification ofphosphotyrosine as a product of epidermal growth factor-activatedprotein kinase in A-431 cell membranes. J. Biol. Chem.. 255:8363-8365. Weibel, M., Pettmann, B., Daune, G., Labourdette, G., and Sensenbrenner, M. (1984) Chemically defined medium for rat astroglial cells in primary culture. Int. J . Dev. Neurosci., 2:355-366.
FGFs Stimulate IGF Binding Protein Synthesis Without Affecting IGF Synthesis in Rat Astroblasts in Primary Culture CAMILLE LORET,~THJERRYJANET: GERARD LABOURDETTEP HELENE SCHNEID,~AND MICHEL BINOUXl 'Institut National de la Sante et de la Recherche Medicale, Unite de Recherches sur la Regulation de la Croissance, U.142, Hopital Saint Antoine, 75571 Paris Cedex 12 and 'Centre de Neurochimie du CNRS and 31NSERM U.44, 67084 Strasbourg Cedex, France
KEY WORDS
Insulin-like,Astrocyte, Growth, Differentiation
ABSTRACT The production of insulin-like growth factor (1GF)-I and -11 and their binding proteins (BPs) has been studied in new-born rat astroblasts at confluency in primary culture. Under the influence of fibroblast growth factors (FGFs)(acidicand basic),the morphology of the astroblasts was altered, 1251-deoxyuridineincorporation was increased, and glutamine synthetase activity was stimulated. IGF production and IGF mRNA expression remained unchanged. Production of the 32 kDa BP (IGFBP-2),the sole or predominant form under base-line conditions, was enhanced and the 43-39 kDa forms (IGFBP-3) appeared or were increased.Epidermal growth factor (EGF)also stimulated production of these BPs, whereas thrombin and db-CAMP had no effect. Our data suggest that a relationship exists between FGF-induced maturation of astroblasts and the forms of BP they produce. The data also indicate that some factors may act specifically on BP synthesis, without affectingIGF synthesis, and in this way play a role in regulating the bioavailability of the IGFs.
INTRODUCTION
maturation, and differentiation of these cells (Loret et al., 1989b). In this study we set out to investigate the modulation by growth factors, and in particular the FGFs, of the biosynthesis of IGFs and their BPs in astroblasts in primary culture. Parallel studies were done of these factors' effects on cell morphology and growth, as well as on glutamine synthetase activity, which is peculiar to astrocytes in the brain.
Following our initial work on the production of insulin-like growth factors (IGFs)and their binding proteins (BPs) by rat pituitary gland and brain explants in culture (Binoux et al., 19811, numerous studies confirmed the synthesis of IGFs by nervous tissue and their involvement in the control of nerve cell proliferation and maturation (reviews in Daughaday and Rotwein, 1989; Oomura and Plata-Salaman, 1987). We subsequently identified several molecular forms of BP in the culture MATERIALS AND METHODS media of human nervous tissue. The major form in Rat Astroblast Cultures nervous tissue also predominates in cerebrospinal fluid and has a preferential affinity for IGF-I1 (Hossenlopp Primary cultures of new-born rat astroblasts were et al., 1986a). In the brain, astrocytes constitute the major cell type. established according to Booher and Sensenbrenner We have reported that certain growth factors,like acidic and basic fibroblast growth factors (aFGF and bFGF), Received August 15,1990; accepted November 9,1990. provoke morphological changes in rat astroblasts in Address reprint re uests t o C. Loret, INSERM U.142, HBpital Saint Antoine, primary culture and, in addition, stimulate the growth, 184 rue du Faubourgiaint Antoine, 75571 Paris Cedex 12, France. 0 1 9 9 1 Wiley-Liss, Inc.
REGULATION OF IGF BP PRODUCTION IN ASTROBLASTS
(1972), as modified previously (Loret et al., 1989a). After 20 days of culture in Waymouth's MD 705/1 medium (Flow Laboratories) supplemented with 110 mg/ml sodium pyruvate, 50 unitdm1 penicillin, 50 pg/ml streptomycin, and 10% fetal calf serum (Gibco), the confluent astroblasts were switched to a chemically defined medium for performance of the experiments (Weibel et al., 1984). This was the same medium, lacking serum and supplemented with 10 bg/ml IGF-free bovine serum albumin (BSA) (bioMerieux,France) with or without one of the following factors: 5 ng/ml acidic FGF, 5 ng/ml basic FGF (R & D, Minneapolis, MN, USA), 10 ng/ml EGF (Sigma, USA), 2.5 unitdm1 thrombin (gift from J.M. Freyssinet, Centre de Transfusion Sanguine, Strasbourg, France), or 1 mM db-CAMP (Sigma). The culture was then continued for a further 4 days.
379
Competitive binding studies Binding studies were done with BPs extracted from control astroblast culture media. The amounts of BP bindin that were approximately 20-25% 1251-IGF-I and 124-IGF-IIwere determined, after which displacement curves were established using increasing concentrations of IGF-I and IGF-11, each in duplicate. DNA Probes
A human IGF-IA 1,073-bp cDNA insert was isolated from a human liver library (Le Bouc et al., 1986) and subcloned in PGEM 4 (pTG 3906). The 660-bp EcoRIBamHI fragment containing exons 1, 2, 3, and 5 was used. For IGF-11, a human IGF-11663-bp cDNA insert was isolated from a human placental library (Le Bouc et al., 1987) and subcloned in PGEM 4 (pTG 3907). This contains part of exon 4, all of exons 5 and 6, and part of 5-['2511 Iodo-2-Deoxyuridine Incorporation exon 7. cDNA probes were labelled by random priming with After a 16-h culture in the defined medium, 0.5 T P act. 0.5-1 mCi/pg DNA, Kit, AmerpCi/ml 5-[1251]iodo-2-deoxyuridine(1251-dUrd) (spec. C I ~ ~ P - ~ A(spec. act. 5 Wmg, Amersham, England) was added. After a sham). further 24 h, the dishes were rinsed 3 times with 0.9 M NaCl and the cells were scraped off and sedimented. The Northern Blotting radioactivity of the pellets was counted in a gamma counter. Total RNA was isolated according to Chomczynski and Sacchi (1987), using a single extraction with acid guanidium thiocyanate/phenoUchloroform.Total RNA Glutamine Synthetase and Protein Assays samples (12 pg) were submitted to electrophoresis on a After 4 days of culture in the defined medium, cells 1%agarose gel containing 6% formaldehyde and then were rinsed 3 times with 0.9 M NaC1, sedimented, and transferred onto a nylon membrane (Hybond N, Amerfrozen at -20°C. The glutamine synthetase (GS) assay sham). After hybridization, the blots were washed to was performed according to Miller et al. (1978), with final stringency with 0.1 x SSPE (15 mM NaCl, 1 mM slight modifications (Loret et al., 1989a). Proteins were phosphate buffer, pH 6.8,O.l mM EDTA), 0.1% SDS at measured according to Lowry et al. (19511,using BSA as 48°C for 30 min and then exposed to X-ray film (Curix RP1, Agfa) with two intensifying screens (DuPont standard. Cromex Hiplus). IGF Assay and Competitive Binding Studies
Western Ligand Blotting The methods used have been described in detail elseThe method used has been described in detail by where (Binoux et al., 1984, 1986). Samples (7 ml) were collected after 4 days' culture in defined medium, de- Hossenlopp et al. (198610).After SDS-polyacrylamidegel salted, lyophilized, and then submitted to acidic gel electrophoresis (without reducing agent, except in the case of the markers), proteins were electroblotted onto a filtration in order t o separate the IGFs and the BPs. nitrocellulose sheet and the BPs identified using lZ5IIGF-I1 as a probe. IGF assays Pure IGF-I and IGF-I1 from human plasma were a generous gift from R. Humbel (Zurich, Switzerland). IGF-I was determined by radioimmunoassay (RIA) using the antiserum prepared by J.J. Van Wyk (Chapel Hill, NC, USA) and IGF-I1 was measured by protein binding assay using BPs extracted from human cerebrospinal fluid (CSF),which have a selective affinity for IGF-I1(Binoux et al., 1986).
RESULTS Effects of Growth Factors and db-CAMP on Cell Morphology, Growth, and GS Activity Only the FGFs, epidermal growth factor (EGF), and db-CAMP induced significant morphological changes after 4 days in the defined medium. These effects, involving reduced cell body size and the appearance of
380
LORET ET AL.
processes, have been reported previously (Loret et al., 1988,1989a). In three experiments, mitogenic activity, as reflected by 1251-deoxyuridineincorporation measured after 40-h culture, was stimulated by a factor of approximately 6 by the FGFs and thrombin and by a factor of 3 4 by EGF, whereas db-CAMP had an inhibitory effect on the residual proliferation of the cells (Fig. 1A). GS activity, by contrast, was most strongly stimulated by EGF and db-CAMP (by factors of about 7 and 6, respectively). aFGF and bFGF provoked 3-4-fold stimulation, and thrombin only 1.5-2 (Fig. 1B). Expression of IGFs and Their mRNAs Under control conditions the rat astroblasts secreted both IGF-I and IGF-I1into the culture medium. None of the factors of hormones tested had any effect on IGF production. Fig. 2A shows the IGF concentrations detected in the culture media of control and aFGF- or bFGF-treated cells. Despite wide variations in the amounts of IGF secreted from one experiment to another in both treated and untreated cultures, the ratio of IGF-IYIGF-I concentrations remained close to 2 (Fig. 2B). mRNA expression was determined by Northern blotting. IGF-I1 mRNAs were more abundant than IGF-I mRNAs (Fig. 3). For the IGF-I1 gene, four lengths of mRNA were detected, of 4.9,2.1,1.5, and 1kb, the 4.9 kb messenger being the major form. For IGF-I, two sizes of mRNA of approximately 4.1 and 1.7 kb were found. None of the factors tested had any effect on their expression (Fig. 3). Modulation of BP Secretion The BPs secreted into the culture medium by the astroblasts were analyzed by Western ligand blotting and compared with those found in serum, which reflects liver production in vivo. In serum there are six molecular forms. Three of these, of 43,41, and 39 kDa, represent different states of glycolysationof the same protein, as determined in experiments using N-glycanase (data not shown). The other three, of 32,29, and 24 kDa, are distinct species. The migration profiles of BPs secreted by control astroblasts were similar to those of BPs in new-born rat serum, with a preponderant 32-kDa form appearing as a doublet. The 43-39-kDa triplet and the 24-kDa BP were absent or barely detectable. Treatment with the FGFs increased the 32-kDa BP and resulted in the appearance or increase of the 43-39kDa form. Among the other growth factors, only EGF induced the appearance of the high molecular weight species. The remainder had no visible effect (Fig. 4). The binding characteristics of the major BP secreted were investigated using the culture medium of un-
Y
Control a F G F
bFGF
Thr
E G F dbcAMP
Fig. 1. Effects of acidic fibroblast growth factor (aFGF), basic fibroblast growth factor (bFGF), e idermal growth factor {EGF), thrombin (Thr) and db-cAMP on [12511-~eoxyundine 1251dURdincorporation (A) and glutamine srthetase (GS) actifity (B) in rat astroblast cultures. Factors were ad ed at day 20 as described in Materials and Methods.A typical experiment is shown. Error for triplicate dishes was below 10% for incorporation and below 15%for the lutamine synthetase assay. Similar results were obtained in two furtier experiments.
treated astroblasts (in which only the 32-kDa BP was present) after acidic extraction. The BP proved to bind IGF-I1 preferentially, whichever tracer was used (Fig. 5). The affinity constants for IGF-I and IGF-11, calculated from Scatchard analysis, were 0.4 x loloM-l and 0.8 x 10" M-l (mean of two experiments). Finally, it is worth mentioning here that other growth factors and hormones tested [platelet-derived growth factor (PDGF), transforming growth factorp (TGFP), triiodotyronine, hydrocortisone] had no effect on either IGF production, mRNA expression, or BP secretion (results not shown).
REGULATION OF IGF BP PRODUCTION IN ASTROBLASTS
A
11
1.1
381
B
r = 0.74
0
p < 0.001
1
./
1
a /
0
0.9
--
0.7
-
. 0
0.7
/
v
./ /
0
\
0.6
C
-
0
E
0.6
/
W
Y
LL (3
/
0.8
h
E
/
/
0.9
0.8
/ /
I
0.5
0.5
(3 0.4
0.4
0.3
0.3
0.2
0.2 ’
01
0.1’
0
0 Control
a FGF
b FGF
n=13
n = 9
n=lO
I
I
1
1
I
1
0.1
0.2
0.3
0.4
0.5
0.6
IGF
f
0.7
- I(ng/ml)
Fig. 2. A: Effects of acidic fibroblast owth factor (aFGF) and basic fibroblast growth factor (bFGF) on cumulative insulin-like growth factor (gF1-1(hatched columns) and IGF-I1production (0 en columns)by rat astroblasts after 4 days’ culture (mean 2 SD). B: Relationships between IGF-I1and &F-I concentrations after 4 days’culture in defined medium without (controls)and with aFGF, bFGF, EGF, thrombin, or db-CAMP.
DISCUSSION Under our experimental conditions, only the FGFs and EGF were capable of inducing changes in cell 4.9 + 4.1 + morphology, increased proliferation, and enhanced GS activity (which reflects cell maturation). The other reg2.1 + 1.7+ 1.5 + ulatory factors tested had partial effects: thrombin 1.0 * failed to affect cell morphology, and db-CAMP, which is the second messenger for norepinephrine (Narumi kb kb et al., 1978), had no effect on proliferation. These findings confirm and expand our previously published data IGF I IGF II (Loret et al., l988,1989a, b). Fig. 3. Northern blot analysis of insulin-like growth factor (1GF)-I Astroblasts in primary culture secreted twice as much and IGF-I1mRNAs isolated from rat astroblast cultures after 4 days in IGF-I1 into the culture medium as IGF-I. Also, IGF-I1 defined medium without (controls) and with acidic fibroblast growth factor (aFGF) or basic fibroblast growth factor (bFGF).Exposure time mRNAs were more abundant than IGF-I mRNAs. These was 8 days for IGF-I mRNAs and 4 days for IGF-I1mRNAs. findings differ from those of Ballotti et al. (1987) who
382
LORET ET AL.
serum
1.0 -
Culture medium
l Z 5 1 - 1 I~ ~
0.8 -
0
0.6 -
F4
BPs
0.1
(kDa) 43 + 41 + 3 9 .)
24
-
0.2 -
+
1.0-
0.8 -
Fig. 4. Western ligand blot analysis of binding proteins (BPs) in rat astroblast culture media. Profiles in serum are shown for purposes of comparison. For abbreviations, see Fig. 1legend.
0
F4 F?
0.6-
0.4 -
observed no IGF-I1mRNA in rat astrocytes in secondary culture. None of the regulatory factors tested had any effect on either IGF production or expression of their mRNAs. Under all the experimental conditions used, production of the two IGFs mirrored expression of their mRNAs. IGF production was accompanied by production of some of the BPs. In the control astroblasts, it was mainly the 32-kDa BP that was secreted. This corresponds to the BP purified from the culture media of rat liver BRL 3A cells, whose cDNA has recently been cloned (Ballard et al., 1989; Brown et al., 1989; Margot et al., 1989)and which is now known as rat IGFBP-2. It has a preferential affinity for IGF-11,which, however, is less marked than that of the correspondinghuman form predominating in CSF and the central nervous system (Hossenloppet al., 1986a). Other authors have reported BP secretion by rat astrocytes in culture: Han et al. (1988) used affinity cross-linking to analyze the BPs they detected and their data can therefore not be compared with ours; Ocrant et al. (1989) used Western ligand blotting and their profiles largely resemble ours. The new findings emerging from this study concern stimulation of BP secretion by the growth factors that influence the proliferation and maturation of astroglial cells. Among the factors tested, only aFGF, bFGF, and EGF provoked increased production of the 32-kDa BP and the appearance or increase of the 43-39-kDa triplet. This triplet corre-
0.2 -
0.0 , 0.1
B I
0.5
1.0
2.0
ng of IGF added per tube Fig. 5. Com etitive inhibition b insulin-like owth factor (1GF)-I and -11 of the finding of 1251-IGF-P(A) and IGF-ff (B) to IGF binding protein (IGF BP)-2extracted from the culture medium ofuntreated rat astroblasts (controls).The amount used er tube was that giving 25% bindin of tracer in the absence of coldPGF (B,,). This corres onded to 19 pf eqnuivalent of culture medium for the bindin of '251-IGP-Iand 16 p1 fort e binding of 1251-IGF-II.The competitivefinding curves for the two tracers were obtained from the same experiment and similar results were obtained from a second experiment.
sponds to the preponderant form in adult rat serum, known as IGFBP-3 (Ballard et al., 1989; Hossenlopp et al., 1987). In serum it appears to be GH-dependent, unlike IGFBP-2, which predominates in fetal and newborn serum (Hossenloppet al., 1987).The similar effects of the FGFs and EGF on BP secretion may be accounted for by the fact that their receptors both have tyrosine kinase activity (Ushiro and Cohen, 1980; Rozengurt, 1986). Also, in astroblasts cultured for 20 days, they exhibit comparable modulation of the synthesis of the
REGULATION OF IGF BP PRODUCTION IN ASTROBLASTS
383
major proteins seen in two-dimensional LDS-polyization and evidence of biologic function. J. Neuroscience, 8:3135-3143. acrylamide gel electrophoresis (Loret et al., 1989a). Hossenlop P., Seurin, D., Segovia-Quinson, B., and Binoux, M. Our observations suggest that some relationship ex(1986a) hentification of an insulin-like growth factor bindin tein in human cerebrospinalfluid with a selective affinity for If$? ists between the forms of BP secreted by astroblasts and FEBS Lett., 208:439-444. the stage of cell maturation regulated by aFGF, bFGF, Hossenlopp, P., Seurin, D., Segovia-Quinson, B., Hardouin, S., and or EGF. They also show that certain factors are capable Binoux, M. (198613) Anal sis of serum insulin-like growth factor binding proteins using Jestern blotting: Use of the method of of acting specifically on BP synthesis without affecting titration of the binding proteins and competitive binding studies. IGF synthesis and are therefore involved in regulating Anal. Biochem., 154:138-143. Hossenlo P., Seurin, D., Se ovia, B., Portolan, G., and Binoux, M. the bioavailability of the IGFs.
ACKNOWLEDGMENTS We are grateful to C. Lassarre, B. de Galle, and M.F. Knoetgen for their technical assistance. C. Loret is a recipient of a grant from the Fondation pour la Recherche Medicale.
REFERENCES Ballard, J., Baxter, R., Binoux, M., Clemmons, D., Drop, S., Hall, K., Hintz, R., Rechler, M., Rutanen, E., and Schwander, J. (1989) On the nomenclature of the IGF binding proteins. Acta Endocrinol. (Copenh.),121:751-752. Ballotti, R., Nielsen, F.C., Pringle, N., Kowalski, A,, Richardson,W.D., Van Obberghen, E., and Gammeltoft, S. (1987) Insulin-like growth factor I in cultured rat astrocytes: Expression of the gene, and receptor tyrosine kinase. EMBO J., 6:3633-3639. Binoux, M., Hossenlopp, P., Lassarre, C., and Hardouin, N. (1981) Production of insulin-like growth factors and their carrier by rat pituitary gland and brain explants in culture. FEBS Lett., 124:178-183. Binoux, M., Seurin, D., Lassarre, C., and Gourmelen,M. (1984) Preferential measurement of insulin-like growth factor (IGF) I-related peptides in serum with the aid of IGF binding proteins (IGF BPs) produced by rat liver in culture. Estimation of serum IGF BP levels. J. Clin. Endocrinol.Metab., 59:453-462. Binoux, M., Lassarre, C., and Gourmelen, M. (1986) Specific assay for insulin-like growth factor (IGF) I1 usin the IGF binding roteins extracted from human cerebrospinal fiuid. J. Clin. Enjocrinol. Metab., 63:1151-1155. Booher, J. and Sensenbrenner, M. (1972) Growth and cultivation of dissociated neurons and glial cells from embryonic chick, rat and human brain in flask cultures. Neurobiology,2:97-105. Brown, A.L., Chiariotti, L., Orlowski, C.C., Mehlman, T., Burgess, W.H.,Ackerman,E.J., Bruni, C.B., andRechler, M.M. (1989) Nucleotide sequence and expression of a cDNA clone encoding a fetal rat binding protein for insulin-like growth factors. J . Biol. Chem., 2645148-5154. Chomczynski, P. and Sacchi, N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroformextraction. Anal. Biochem., 162:15&159. Dauehadav. W.H. and Rotwein. P. (1989) Insulin-like mowth factors I an% 11. Pe tide, messenger kbonucleic acid and gene structures, serum, anttissue concentrations.Endocrinol. Rev., 10:6&91. Han, V.K.M., Lauder, J.M., and DErcole, J. (1988) Rat astroglial somatomedidinsulin like growthfactor binding proteins: Character-
(1987) Fiterogeneity of insdin-like growth factor binding proteins and relationships between structure and affinity. 2. Forms released b human and rat liver in culture. Eur. J. Biochem., 170:133-142. Le Bouc, Y., Dreyer, D., Jaeger, F., Binoux, M., and Sonderme er, P (1986) Complete charaderization of the human IGF-I nuccotide se uence isolated from a newly constructed adult liver cDNA library. F&SLett., 196:108-112. Le Bouc, Y., Noguiez, P., Sondermeyer,P., Dreyer, D., Girard, F., and Binoux, M. (1987) A new 5’ non-coding region for human lacental insulin-like growth factor I1 mRNA expression. FEBS Lett., 222:181-185. Loret, C., Sensenbrenner, M., and Labourdette, G. (1988) Maturationrelated gene expression of rat astroblasts in vitro studied b two dimensional polyacrylamide gel electrophoresis. Cell 5iffer.i 2k37-46. Loret, C., Sensenbrenner, M., and Labourdette, G. (1989a)Differential enotypic expression induced in cultured rat astroblasts by acidic ibroblast growth factor, epidermal growth factor and thrombin. J. BWZ.Chem., 264:8319-8327. Loret, C., Laeng, P., Sensenbrenner, M., and Labourdette, G. (1989b) Acidic and basic fibroblast growth factors similarly regulate the rate of bios thesis of rat astroblast proteins. FEBS Lett., 257:324-328. Lowry, Rosebrough, N.J., Farr, A.L., and Randall, R.J. (1951) Protein measurement with the Folin phenol reagent. J . Biol. Chem., 193:265-275. Mar ot, J.B., Binkert, C., Mary, J.L., Landwehr, J., Heinrich, G., and Scfwander, J. (1989) A low molecular weight insulin-like y w t h factor binding protem from rat: cDNAcloningand tissue distn ution of its messenger RNA. Mol. Endocrinol., 3:1053-1060. Miller, R.E., Mackenberg, R., and Gershman, H. (1978) Regulation of dutamine svnthetase in cultured 3T3-Ll cellsbv insulin. hvdrocortilone and dibutyryl cyclic AMP. Proc. Nati. Acad.’Sci. USA, 75:1418-1422. Narumi, S.. Kimelberg, H.K., and Bourke, R.S. (1978) Effects of norepinephrine on the morphology and some enzyme activities of primary monolayer cultures from rat brain. J. Neurochem., 31:147%1490. Ocrant, I., Pham, H., Oh, Y., and Rosenfeld, R.G. (1989) Characterization of insulin-like growth factor binding proteins of cultured rat astroglial and neuronal cells. Biochem. Biophys. Res. Commun., 159:1316-1322. Oomura, Y. and Plata-Salaman, C.R. (1987) Somatomedins (insulinlike growth factors)and the nervous system. In: Insulin, Insulin-Like GrowthFactors, and Their Rece tors in the CentralNervous System. M.K. Raizada, I.M. Phillips, anfD. LeRoith, eds. Plenum Press, New York, pp. 215-244. Rozengurt, E. (1986) Early signals in the mitogenic response. Science,
P
&.,
224.161-166 -- - .- -- - - - .
Ushiro, H. and Cohen, S. (1980) Identification ofphosphotyrosine as a product of epidermal growth factor-activatedprotein kinase in A-431 cell membranes. J. Biol. Chem.. 255:8363-8365. Weibel, M., Pettmann, B., Daune, G., Labourdette, G., and Sensenbrenner, M. (1984) Chemically defined medium for rat astroglial cells in primary culture. Int. J . Dev. Neurosci., 2:355-366.
