0013-7227/92/1313-1134$03.00/0 Endocrinology Copyright 0 1992 by The Endocrine Society

Vol. 131, No. 3 Printed in U.S.A.

Mitogenic Activity of Epidermal Growth Factor on Newborn Rat Astroglia: Interaction with Insulin-Like Growth Factors* VICTOR K. M. HAN?, ANN SMITH, SHERI BRADSHAWS

WUNNA

MYINT,

KAREN NYGARD,

AND

Medical Research Council Group in Fetal and Neonatal Health and Development (V.K.M.H., A.S., W.M., K.N., S.B.), the Departments of Pediatrics (V.K.M.H.), Anatomy (V.K.M.H.), and Biochemistry (V.K.M.H.. S.B.J. Uniuersitv of Western Ontario, and the Lawson Research Institute, London, Ontario,

ABSTRACT Newborn rat astroglial cells possess epidermal growth factor (EGF) and insulin-like growth factor (IGF) receptors, which suggests that these growth factors regulate their growth and development. To determine the relative roles and interactions between the two growth factors on astroglial growth, primary cultures of astroglial cells from newborn rata (1 day postnatal) were treated with pure peptides, singly or in combination in various concentrations, and the growth response was determined by DNA synthesis ([3H]thymidine incorporation). EGF, IGF-I, and IGF-II, as single peptides, stimulated DNA synthesis, with half-maximal stimulatory concentrations of 0.25 rig/ml for EGF, 2.0 rig/ml for IGF-I, and 25 rig/ml for IGF-II, respectively. These findings indicate that astroglial cells are responsive to these growth factors in physiological concentrations, with the relative sensitivity of EGF > IGF > IGF-II. When EGF and IGF-I were added in combination, the growth stimulatory effect was greater than the additive effects of each growth factor added alone, indicating that the two growth factors act in synergism with each other. In particular, addition of increasing concent&ions of EGF from 0.25-long/ml to a constant concentration of 50 &ml IGF-I resulted in sienificant uotentiation of r3Hlthvmidine incorp&ation of astroglial cells: To de&mine if the syner&stic effect was due to a local synthesis of IGF-I by astroglia, a specific monoclonal

antibody against IGF-I (Sm 1.2) was added to the peptides. Sm 1.2 decreased not only IGF-I-stimulated DNA synthesis, but also EGFstimulated DNA synthesis, suggesting that the effects of EGF were contributed to in part by the local synthesis of IGF-I by astroglial cells. Analysis of conditioned medium from cells treated with EGF revealed a sig&cant increase (-a-fold) in IGF-I (from 4.5 to 8.8 rig/ml), but not IGF-II. To determine if the EGF effect on IGF synthesis was at the level of IGF-I mRNA transcription, stable IGF-I mRNA levels were determined in the astroglial cells before and after stimulation with EGF, using Northern analysis and quantification by densitometry. Astroglia exmessed four IGF-I mRNA transcrints as in the adult and fetal l&er, bit only one (3.6 kilobases) IGF-II mRNA. The stable IGFI and IGF-II mRNA levels were unchanged after EGF stimulation, suggesting that the effect of EGF was not due to an increase in either IGF-I or IGF-II mRNA transcription, but most likely was due to an effect on either translation or p&ttranslational processing or an increase in IGF-I secretion. These findings SUDDOI-~ our hwothesis that EGF and IGFs regulate the growth of &ro&al cells si&y as well as in combination, and that the effect of EGF may be mediated in part by increasing the local synthesis of IGF-I. Interaction between EGF and IGFs may play a role in gliogenesis of neonatal rat brain. (Endocrinology 131: 1134-1142,1992)

D

the developmental process communicate with each other via cell to cell interaction as well as many chemical messengers. This interaction between the two cell types as well as among the same cell type is crucial for normal development of the brain (2). Growth factors constitute one important group of chemical messengers used by the developing brain cells for this communication (1). Nerve growth factor is perhaps the best studied growth factor (see review in Ref. 3). However, the role of nerve growth factor is best demonstrated in the differentiation, survival, and programmed cell death of sympathetic and peripheral nervous systems, and its role in the growth and development of the central nervous system is still not clearly known. Recently characterized related molecules, the brain-derived neurotropic factor and neurotropin3, are more likely candidates in this regard (4). With the recent demonstration that powerful mitogenic and differentiation-promoting insulin-like growth factors (IGFs) and their receptors are expressed in early embryos in rodents (5, 6), it is likely that these growth factors may be involved in the development of the nervous system from the very early stages.

EVELOPMENT of the brain in mammals involves a series of processes by which the primitive neural plate of ectodermal origin grows and matures into a complex organ with multiple cell types (see review in Ref. 1). At the cellular level, the primitive neuroepithelial cells undergo processes of proliferation, differentiation, migration, and programmed cell death (2). Two major cell types develop from the primitive neuroepithelial cells, the neurons and glia, which during

Received December 16, 1991. Address all correspondence and requests for reprints to: Dr. Victor Han, Room 3-425, Lawson Research Institute, 268 Grosvenor Street, London, Ontario, Canada N6A 4V2. *This work was supported by Group Grant MA-10199 from the Medical Research Council of Canada (to V.K.M.H.) and Basil O’Connor Starter Scholar Award 5-679 from the March of Dimes-Birth Defects Foundation (to V.K.M.H.). The experimental animals were treated in accordance with Protocol 90177-7, as auuroved bv the Council on Animal Care, University of Westen-.Onta&. ’ t Scholar and member of the Medical Research Council of Canada Group on Fetal and Neonatal Health and Development. $ Supported by the Graduate Student Bursary of the Department of Pediatrics, University of Western Ontario. 1134

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 18:14 For personal use only. No other uses without permission. . All rights reserved.

EGF AND IGFs ON ASTROGLIAL In most mammalian species, the two major cell types of the developing brain, neurons and glia, undergo proliferation during different stages of embryogenesis. Neurogenesis usually precedes gliogenesis. The timing of these two events is also different among species. In humans, neurogenesis is completed in the early second trimester, and most gliogenesis is finished before birth (7). In contrast in rats, neurogenesis is completed before birth, but gliogenesis predominantly occurs in the fist week of postnatal life (8). It is possible that different mitogenic peptide growth factors may be involved in the regulation of the two processes. In the early postnatal period in rodents, many growth factors are known to be expressed, and it is possible that they may interact with each other in regulating gliogenesis. In this study we examined the role of one of these growth factors, epidermal growth factor (EGF; which is expressed in the kidney and salivary glands of early postnatal rodents) (9) in the growth of glial cells and its possible interactions with IGFs, which are expressed in the developing glial cells (10). Several investigators have shown that IGF-I and IGF-II are synthesized by developing astroglial cells and exert mitogenic action on them (10-12). We and others have previously shown that primary glial cells grown from newborn rats possess IGF receptors (12), have the capability of synthesizing IGFs (lo), synthesize IGF-binding proteins (IGFBPs) (13), and have a growth response to IGFs (12). EGF is also shown to be a potent stimulator of astrocytic proliferation (14-17). EGF, however, is not synthesized by the developing neural cells, but its homolog transforming growth factor-a (TGFol) mRNA is expressed in the brain (18, 19, 20). Glial cells have been shown to possess EGF receptors (17) and, therefore, may respond to its biological actions. During the time when gliogenesis is occurring in the neonatal rat, immunoreactive EGF is detectable in tissues and blood. We hypothesized that EGF and IGFs have interactive as well as independent regulatory actions on the development of glial cells. We have used an in vitro model of developing neonatal rat glial cells to study the biological activity of EGF on the growth of astroglia and its interactions with IGFs in mediating this effect. Materials

and Methods

GROWTH

Island, NY) at room temperature for 30 min with constant shaking. Trypsin activity was inhibited by the addition of twice the volume of basal medium Eagle (BME; Gibco) containing 10% fetal calf serum (Gibco). The cell suspension was then removed and filtered through a 130-pm Nitex monofilament screen (Tetko, Elmsford, NY). The remaining tissues were triturated mechanically and filtered. The filtrate was combined with the initial cell suspension and centrifuged at 2000 x g. The cell pellet was gently vortexed and resuspended in complete medium (BME with 10% fetal calf serum, 5 U/ml penicillin, and 5 pg/ml streptomycin). The number of viable cells was determined by the trypan blue exclusion method in a hemacytometer (>90% viable). A cell suspension of 2.5-3 x lo6 live cells/ml was prepared and plated in 200~ml T-75 tissue culture flasks (75~cm2 growth area; Costar Corp., Cambridge, MA). The cells were cultured at 37 C in 5% CO2 and 95% air with 100% humidity for 12-14 days until confluence. The cells were then shaken at 270 rpm for 18 h at 37 C, and the medium containing suspended cells was removed, discarded, and replaced with new complete medium. The resulting culture consisted of more than 95% flat polygonal astroglial cells, as confirmed by phase contrast microscopy and positive immunostaining with an antibody against glial fibrillary acid protein. The remaining 5% of cells consisted of process-bearing cells (about half of which were astrocytes and half oligodendrocytes), ependymal cells, and fibroblasts. The astroglial cells were used within 1 week of purification and at least one day after the medium change.

PHlThymidine

incorporation studies

The cells were subplated onto 48well plates (Costar) at a density of 1.25 X 10s cells/well and cultured to subconfluence for 48 h. They were incubated in serum-free (SF) BME for another 48 h and exposed to varying concentrations of growth factors, singly or in combination, in SF BME containing BSA (2 mg/ml) or in the latter medium only for controls. For experiments to inhibit the endogenously produced IGF-I, specific monoclonal antibody against human IGF-I, Sm 1.2, was added in various concentrations (expressed as dilution of the stock antibody solution), ranging from I:5000 to l:lOO, together with the peptides. After incubation for 22 h, [methyl-3H]thymidine (0.5 &/well; 1.0 &i/ ml; Amersham Canada Ltd., Oakville, Ontario, Canada) was added to the cultures, and the incubation was continued for another 2 h. The reaction was stopped by aspiration of the incubating solution and washing with cold (4 6) PBS three times. The cult&es were then incubated with 5% trichloroacetic acid (TCA) at 4 C for 20 min. The TCA solution was aspirated, and the cuhures’ were washed again with fresh 5% TCA. After removal of the TCA solution, the cultures were solubilized in 0.5 ml 0.1 N sodium hydroxide. Half of the resulting solution was added to 5 ml scintillation cocktail (ScintiVerse, Fisher Scientific, Napean, Ontario, Canada) and counted in a scintillation counter (Beckman Instruments, Inc., Palo Alto, CA). The remaining half of the solubilized cell solution was neutralized with an equivalent volume of 0.1 M HCI, and the amount of DNA was determined by the diphenylamine method (22), using caif thymus DNA as standard. This ensured that an equivalent number of cells was tested in each well.

Materials Recombinant EGF, IGF-I, and IGF-II were purchased from IMCERA Bioproducts (Terre Haute, IN), and insulin from Sigma Chemical Co. (St. Louis, MO). Monoclonal antibody against human IGF-I (Sm 1.2) was a kind gift from Drs. L. E. Underwood and J. J. Van Wyk, University of North Carolina (Chapel Hill, NC). IGF-I and IGF-II were iodinated by the chloramine-T method, as previously described (21).

Astroglial cultures Relatively pure cultures of polygonal astroglial cells were prepared from cerebral cortices of l-dav-old Snraeue-Dawlev rats (Charles River Breeding Laboratories, Montreal, Q;eb&, Canada), as previously described (12). Briefly, l-day-old rats were anesthetized with an overdose of chloroform and decapitated, and the brains were removed aseptically. Cerebral cortices were dissected, and the meningeal coverings were carefully peeled off. Cortical tissues were cut into small pieces and incubated in 0.1% trypsin and 0.02% EDTA solution (Gibco, Grand

fH]Thymidine

autoradiography

To confirm that the [3H]thymidine that was incorporated into the cells was used for the synthesis of nuclear DNA, astroglia were plated onto 48well tissue culture plates, as described above, until subconfluence, incubated with SF BME for another 48 h, and treated with varvine concentrations of growth factors or insulin for 22 h. The cells were ihe,” incubated with [3H]thymidine (0.5 &/well) for 2 h more. The incubating solution was removed, washed in cold (4 C) PBS three times, and fixed with 4% paraformaldehyde in 70 mr.t phosphate buffer, pH 7.0, at room temperature for 30 min. The cells were then washed with PBS three times and dehydrated through an ascending ethanol series (70%, 90%, and 100%). The bottoms of the tissue culture wells were coated with NTB-3 nuclear track emulsion (Eastman Kodak Co., Rochester, NY), air dried for 2 h, wrapped in foil and kept in a light-proof bag, and exposed at 4 C for 3 weeks. The emulsion was developed in D 19 developer, fixed, washed, and stained with hematoxylin and eosin. The cells were then dehydrated in ascending ethanols and mounted with

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 18:14 For personal use only. No other uses without permission. . All rights reserved.

EGF AND IGFs ON ASTROGLIAL

1136

Crystal/Mount (Biomeda Corp., Foster City, CA). The radiolabeled nuclei were viewed with an inverted microscope and manually counted. The labeling index was determined by counting total and radiolabeled nuclei in 10 random fields/well; 3 wells were used for each experimental condition. Therefore, the mean and SEM were calculated from a total of 30 fields for each condition. IGF- I and IGF- II RIA IGF-I and IGF-II concentrations were determined in conditioned medium samples by a modification of a previously described RIA method (21, 23). Briefly, IGF-I and IGF-II in the conditioned medium samples (500 ~1 each) were incubated for 1 h on ice with an equal volume of 1 M acetic acid, separated from IGFBPs by gel filtration chromatography on a 0.9 X 28-cm column of Sephadex G-50 fine (360 X 12 mm; flow rate, 25 ml/h), and eluted with 1 M acetic acid at room temperature. Fractions of 1 ml were neutralized with 2 M NaOH, dried on a vacuum centrifuge, and resuspended in 250 rl phosphate buffer before RIA. Initially, all fractions from gel filtration were subjected to RIA to establish the elution profile of IGF-I and IGF-II. Consequently, IGF-containing fractions were pooled for each conditioned medium sample. The recovery of IGF-I and IGF-II was determined by 1) addition of 10,000 cpm [1251]IGF-Ior [‘251]IGF-II or 2) addition of 100 ng IGF-I or 300 ng IGF-II to 200 ~1 serum before acidification and gel filtration. Recoveries of radiolabeled or nonlabeled IGF-I or IGF-II were in each case greater than 90%. The possible presence of IGFBP contamination in IGFcontaining fractions was assessed by their ability to displace [‘251]IGF-II from activated charcoal, as previously described in detail (24). For the RL4 of IGF-I, the primary antiserum used was rabbit antihuman IGF-I, kindly provided by Drs. J. J. Van Wyk and L. E. Underwood, Department of Pediatrics, University of North Carolina, through the National Pituitary Agency, NIH (final assay dilution, 1:12,000), and for the RIA of IGF-II, monoclonal antibody against human IGF-II was obtained from Amano Biochemicals (Troy, VA; final assay dilution, 0.13 nM). After extractions, conditioned medium samples were assayed in duplicate at 50%, 25%, 12.5%, and 6.25% dilutions. Human IGF-I (IMCERA Bioproducts) and human IGF-II (Bachem, Inc., Torrance, CA) were used for the appropriated standard curves, as previously described (19). The cross-&a&vity of human IGF-I in the RIA for IGF-II was less than l%, and that of IGF-II in the RIA for IGF-I was less than 3%. These assavs have previously been used to measure IGF-I and IGF-II in conditioned medium from fetal chondrocytes (23). Northern

blot hybridization

analysis

Total RNA was prepared from astroglia (two T-75 flasks per experimental condition) -using the guanidine thiocyanate-cesium chloride method (25). Before Northern analvsis, the inteeritv of total RNA preparations was checked by subjecdng the RNA samples to agaroseurea gel electrophoresis and staining with ethidium bromide (26). The degraded samples were discarded. Total RNA (20 pg) was denatured and electrophoresed in 1% agarose gels containing 6% formaldehyde. The RNAs were then transferred to a Zeta-Probe blotting membrane (Bio-Rad Laboratories, Richmond, CA) by capillary transfer technique (27). After transfer, the blots were baked at 70 C for 1 h. Blots were then hybridized with 32P-labeled cDNA probes [l-2 X lo6 cpm/ml buffer containing 5 X SSPE (0.75 M NaCl, 44 mu Na2HPOI.2H20, 5 mu EDTA, pH 7.4), 7% sodium dodecyl sulfate (SDS), and 5 @g/ml denatured salmon sperm DNA] at 42 C overnight. The specific cDNA encoding rat IGF-I was a gift from Dr. L. Murphy (University of Manitoba, Winnipeg, Manitoba, Canada), and that for mouse IGF-II was a gift from Dr. Greame Bell (University of Chicago, Chicago, IL). The cDNA inserts were labeled with [3’P]deoxy-CTP to specific activities of 1-2 X IO9 cpm/pg by the random priming technique, using the oligo-labeling kit (Pharmaaa-Canada, Inc.; Baie d;Urfe, duebet. Canada). The blots were washed twice (30 min each) in 1 x SSC (0.15 M NaCl, 15 mM Na citrate, pH 7.0)-0.1% SDS at 42’C and twice (30 min each) in 0.1 X SSC-0.1% SDS at 42 C. They were air dried and subjected to autoradiography using intensifying screens at -70 C. The blots were stripped after hybridization with specific cDNA probes by washing twice for 30 min at 90 C in 0.01 x SSC and 0.5% SDS.

GROWTH

Endo. 1992 Voll31. No 3

Consistency in the relative amounts of total RNAs being loaded to each lane of the gel was checked by probing the blots with a radiolabeled cDNA for 18s ribosomal RNA (a gift from Dr. David Denhardt, Rutgers University, Piscataway, NJ). The autoradiograms were quantified using laser densitometry ((Ultrascan XL, LKB, Bromma, Sweden). The relative densities of the bands were expressed as arbitrary absorbance units (au) per mm. To correct for minor differences in loading of total RNA in Northern blots, a ratio of the relative density of each specific band or bands to the relative density of the 18s ribosomal RNA band was calculated before comparisons were made. Calculations

and statistical

analysis

In each experiment a minimum of six wells of astroglia were tested for each condition. The results were calculated from three experiments and expressed as the mean f 1 SEM unless otherwise indicated. Significant differences between a treatment condition and the control were calculated using Student’s t test or, where multiple treatment conditions were compared, an analysis of variance. For nonparametric measurements, the Mann-Whitney rank sum test (for comparison of two experimental groups) was used (28). To determine if the addition of IGF-I and EGF in combination gave a synergistic or additive effect, an analysis of variance technique with quadratic and cubic terms was used in a SAS statistical analysis system (SAS Institute, Inc., Cary, NC). A synergistic effect was determined to be present if a significant interaction (P < 0.05) was present among values for different treatment conditions. To equalize for the minor loading differences among RNA samples in Northern blot hybridization, the area under the curve densitometric readings of the different hybridizing mRNA bands were normalized for the signal intensity of the 18s ribosomal RNA by calculating the signal ratio between the two. For IGF-I and IGF-II in which multiple mRNA species were present, the total densitometric reading of all of the hybridizing bands was taken as the total mRNA level of either IGF-I or IGF-II.

Results Effects of EGF and IGFs on astroglial

DNA synthesis

EGF stimulated DNA synthesis of astroglia in a dosedependent manner (Fig. 1). In each of the three experiments performed (results of only two experiments are shown in Fig. l), EGF stimulated astroglial DNA synthesis at concentrations as low as 0.25 rig/ml. The maximum stimulatory effect was usually achieved with 10 rig/ml, and the use of higher concentrations was associated with less than the maximum effect (Fig. 2). To compare the relative mitogenic potencies of EGF and IGFs, a dose-response stimulation test was performed with the three growth factors (Fig. 2). EGF was the most potent, with half-maximum stimulation (EDSO)being achieved with 0.25 rig/ml, followed by IGF-I, with an EDS0of 2.0 rig/ml, and then by IGF-II, with an EDS0of 25 rig/ml. The maximum stimulation was achieved with 10 rig/ml EGF, 50 rig/ml IGFI, and 100 rig/ml IGF-II. These results indicate that EGF is the most potent mitogen of the three growth factors tested. To demonstrate that the [3H]thymidine that was incorporated into the cells was used for the synthesis of nuclear DNA, [3H]thymidine labeling autoradiography was performed, and the labeling index was calculated. Figure 3 shows that in astroglial cells, basal DNA synthesis (A) was increased by the addition of 10 rig/ml EGF (B), 50 rig/ml IGF-I (C), or 100 rig/ml IGF-II (D). The labeling indices were 3.96 f 0.36% for controls (basal condition), 17.83 + 0.82%

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 18:14 For personal use only. No other uses without permission. . All rights reserved.

EGF

Control

AND

1

0 1

EGF

IGFs

10

ON

ASTROGLIAL

GROWTH

1137

100

(rig/mlx log)

1. Tritiated thymidine incorporation dose response of primary rat astroglial cells to increasing concentrations of EGF (0.1-10 rig/ml). A total of three experiments were performed, which resulted in similar dose responses. The results from two experiments [Exp 1 (0) and Exp 2 (Cl)] are shown. Values are the mean + SD of six wells. The control represents [3H]thymidine incorporation of astroglia when the cells were incubated only with SF medium. FIG.

100 7

SO-

&J-

40 -

20 -

02 0

01

1

1

Growth Factor Concentration

10

100

1000

(nglml x log)

FIG. 2. Tritiated thymidine incorporation dose response of primary astroglial cells to increasing concentrations of EGF (W), IGF-I (O), and IGF-II (0). Values are the mean f SEM of three experiments. EGF is the most potent of the three growth factors, with a half-maximum stimulation dose of 0.25 rig/ml. The half-maximum stimulation dose for IGF-I is 2.0 rig/ml, and that for IGF-II is 25 rig/ml.

for EGF (10 rig/ml), 12.8 + 0.48% for IGF-I (50 rig/ml), and 9.15 f 0.7% for IGF-II (100 rig/ml; all treatment conditions were significantly different from the control, P < 0.05). Effects of EGF and IGF-I

on astroglial

DNA

synthesis

To test if EGF and IGF-I were acting on the astroglial cells in an additive or synergistic manner, astroglia were treated with EGF or IGF-I alone or with two different EGF and IGFI combinations: constant concentrations of IGF-I (10 or 50 rig/ml) with increasing concentrations of EGF (0.1-10 ng/ ml; Fig. 4) or constant concentrations of EGF (2.5 or 10 ng/ ml) with increasing concentrations of IGF-I (l-50 rig/ml; Fig. 5).

FIG. 3. Tritiated thymidine incorporation autoradiography of primary astroglial cells under control (basal) conditions (A) and after stimulation with 10 rig/ml EGF (B), 50 rig/ml IGF-I (C), or 100 rig/ml IGF-II (D). The labeling indices are 3.96 + 0.36% for controls, 17.83 + 0.82% for 10 rig/ml EGF, 12.8 + 0.48% for 50 rig/ml IGF-I, and 9.15 f 0.7% for 100 rig/ml IGF-II. Astroglial cells are not only the most sensitive to EGF, but also have the highest tritiated thymidine incorporation response, as evidenced by the labeling index.

When increasing concentrations of EGF were added to a low constant concentration of IGF-I (10 rig/ml), only an additive effect was observed. When EGF was added to a high constant concentration of IGF-I (50 rig/ml), synergistic effects were observed with EGF concentrations of 0.25 ng/ ml and higher (P < 0.001; Fig. 4). When increasing concentrations of IGF-I were added to a low constant concentration of EGF (2.5 rig/ml), only an additive effect was observed. When IGF-I was added to a high constant concentration of EGF (10 rig/ml), synergistic effects were observed with IGFI concentrations of l-25 rig/ml (P < 0.001) and additive effects with 50 rig/ml IGF-I. The actions of IGF-II were similar to those of IGF-I (data not shown). These findings suggest that EGF and IGF-I potentiate the actions of each other in stimulating astroglial DNA synthesis. EGF and local synthesis

of IGF-I

To test whether the synergistic effect between EGF and IGF-I was due to stimulation of IGF-I synthesis by EGF, we added increasing concentrations of specific bioinhibitory monoclonal antibody against IGF-I (Sm 1.2) to EGF (10 ng/ ml). To demonstrate the specificity of this antibody in inhibiting IGF-I, the effect on EGF action was compared with its effect on IGF-I (50 rig/ml) and insulin (1 pg/ml; Fig. 6). Sm 1.2 had little or no effect on basal DNA synthesis. As expected, Sm 1.2 significantly decreased IGF-I action from an antibody dilution of 1:5000 and higher, whereas it had

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 18:14 For personal use only. No other uses without permission. . All rights reserved.

EGF AND IGFs ON ASTROGLIAL

1138

GROWTH

Endo. 1992 Vol131. No 3

70 60 -

.-56

50 -

30000

E g, $

40 -

E .-2::v 0 E za 2

30 -

20000

10000

i 0

20 -

0

10 -

0

1.5000

1.2500

11000

1500

1:lOO

Sm 1.2 Concentration

02

EGF (rig/ml x log) 4. Tritiated thymidine incorporation response of primary astroglial cells to varying concentrations of EGF alone (0) or in combination with 10 rig/ml IGF-I (w) or 50 rig/ml IGF-I (0). Values represent the mean + SD. When increasing concentrations of EGF (0.1-10 rig/ml) were added to a constant low concentration of IGF-I (10 rig/ml), only an additive response was observed. When increasing concentrations of EGF (0.1-10 rig/ml) were added to a constant high concentration of IGF-I (50 rig/ml), a synergistic response was observed with EGF concentrations of 0.25 rig/ml and higher (*, P < 0.001). FIG.

FIG. 6. Tritiated thymidine incorporation response of primary astroglial cells to varying concentrations of monoclonal antibody against IGF-I (Sm 1.2) in control conditions and in the presence of 50 rig/ml IGF-I, 1 pg/ml insulin (Ins), and 10 rig/ml EGF. Sm 1.2 inhibited the mitogenic activity of IGF-I, but not that of insulin on astroglia, in a dose-dependent manner. Sm 1.2 also inhibited the mitogenic activity of EGF in a dose-dependent manner, suggesting that EGF activity may be due in part to the local synthesis of IGF-I by astroglia. TABLE IGF-II treated

1. Radioimmunoassayable in the conditioned medium with varying concentrations EGF (w/ml) 0 (Control) 1 10 25

concentrations of IGF-I and of primary rat astroglial cells of EGF IGF-I hdmu

4.5 4.5 8.8 8.2

+ + f *

IGF-II bdml)

0.25 0.25 0.42" 0.22"

7.2 6.8 9.4 8.6

t ?I + t

1.10 0.24 0.90 0.84

Cells were grown in T-75 flasks to confluence in 10% FCS containing BME, purified by the differential adhesion technique, incubated in SF BME for 2 days, then incubated with SF BME containing 2 mg/ml BSA with or without varying concentrations of EGF for 24 h. The conditioned medium was collected and subjected to RIA as described in Mater& and Methods. The cells were subjected to RNA extraction and Northern blot hybridization analysis. The data are the mean + SD of three independent experiments run in triplicate. "P < 0.05.

I

0

I, r,

I

IO

1

100

IGFl(ng/mlxlog) 5. Tritiated thymidine incorporation response of primary astroglial cells to varying concentrations of IGF-I alone (0) or in combination with 2.5 rig/ml EGF (m) or 10 rig/ml EGF (0). Values represent the mean + SD. When increasing concentrations of IGF-I (l-50 rig/ml) were added to a constant low concentration of EGF (2.5 rig/ml), only an additive response was observed. When increasing concentrations of IGF-I (l-50 rig/ml) were added to a constant high concentration of EGF (10 rig/ml), a synergistic response was observed (*, P < 0.001). A similar, although not identical, response was obtained when IGF-I was substituted with IGF-II (data not shown).

FIG.

no effect on the action of insulin. Sm 1.2 decreased the biological action of EGF when antibody dilutions of 1:lOOO

or higher were used. Since Sm 1.2 does not recognize EGF in a RIA or in Western blots (Underwood, L. E., personal communication), it is unlikely that it directly inhibits EGF; more likely, it inhibits the locally synthesized IGF-I. In addition, since it is known that Sm 1.2 recognizes IGF-II (29), it is possible that EGF may have action on local synthesis of IGF-II. Effect of EGF on astroglial

IGF-I

and IGF-II

synthesis

Under basal conditions, astroglia synthesize both IGF-I and IGF-II into the conditioned medium. The effect of EGF on astroglial IGF-I and IGF-II synthesis was tested by measuring the changes in IGF-I and IGF-II concentrations in the conditioned medium in response to varying doses of EGF (1, 10, and 25 rig/ml; Table 1). The IGF-I concentration in the conditioned medium increased nearly 2-fold, from 4.5 f 0.25 to 8.8 + 0.42 rig/ml (P < 0.05), when treated with 10 rig/ml

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 18:14 For personal use only. No other uses without permission. . All rights reserved.

EGF AND IGFs ON ASTROGLIAL EGF, and no further increase was observed with 25 rig/ml EGF. IGF-II concentrations did not change with either of the EGF treatments. These findings suggest that EGF stimulates astroglial IGF-I synthesis, but not that of IGF-II. Effect of EGF on astroglial expression of IGF-I and IGF-II genes To determine whether the effect of EGF on astroglial IGFI synthesis was due to its effect on IGF-I gene expression, stable IGF-I mRNA levels were measured by Northern blotting on total RNAs extracted from astroglia under basal (control) conditions or after treatment with 10 rig/ml EGF. For comparison, cells were treated with 50 rig/ml IGF-I and 1 rg/ml insulin and analyzed in the same manner (Fig. 6). As positive controls, total RNA from livers of 20-day gestation rat fetuses and adults were analyzed in the same Northern blot. Adult liver expressed greater abundance of IGF-I mRNA than fetal liver. At least four IGF-I transcripts [1.2-7.4 kilobases (kb)] were observed in both adult and fetal liver. The smallest (1.2 kb) and the largest (7.4 kb) mRNAs were more abundant in the adult. Astroglia also expressed four IGF-I transcripts, and the abundance was equivalent to that in fetal liver (Fig. 7A). The abundance of all four transcripts appeared to be unchanged from control values when astroglia were treated with EGF. This finding was confirmed by densitometry when the IGF mRNA levels were normalized for 18s ribosomal RNA. For comparison, IGF-I levels also did not change when astroglia were treated with 50 rig/ml IGF-I or 1 pg/ml insulin. To determine whether EGF treatment altered IGF-II mRNA expression, the Northern blot was stripped and hybridized with 32P-labeled mouse IGF-II cDNA probe (Fig. 7B). Fetal liver showed high abundance of six IGF-II mRNA transcripts, ranging from 1.0-5.0 kb. No IGF-II transcripts were observed in the adult rat liver. Astroglia expressed only the 3.6-kb transcript, the abundance of which did not change with EGF treatment. In comparison, neither IGF-I nor insulin treatment altered its abundance. The same treatments, however, have been shown to alter the abundance of IGFBP-2 mRNA abundance (30), indicating that the method is adequate to detect any significant changes that might occur. Discussion Development of the brain occurs in different stages, with the neurons proliferating first (neurogenesis), followed by glial cells (gliogenesis) (7). In the rat, gliogenesis occurs in the first week of postnatal life (8), and growth factors may play an important role in the regulation of this process (1). Many growth factors are expressed in the developing early postnatal rat; those that are expressed locally in the developing brain (e.g. IGFs) may act in either an autocrine or paracrine mode, and those that are expressed in other organs and tissues (e.g. EGF) may act with an endocrine mechanism (1). Several investigators have previously shown that EGF stimulates the proliferation of astrocytes and other brain celIs

GROWTH

1139

in vitro (I4-19,31,32). Wang et al. (17) further characterized the EGF receptors in primary glial and neuronal cells cultured from l-day-old rats and demonstrated that the functional EGF receptors are predominantly localized in the glial cells. Addition of EGF to glial cells produced a dose-dependent stimulation of thymidine incorporation as well as multiplication of cells over a 6-day period. In this study we have demonstrated that astroglia are very sensitive to the mitogenie actions of EGF in concentrations that may be physiologically present in the circulation at this stage (33). Astroglial cells in our study appeared to be more sensitive to EGF action than those in the study by Wang et al. (17). Half-maximal stimulation was achieved by 0.25 rig/ml, maximum stimulation was achieved by 10 rig/ml, and the maximum mitogenie response was higher (550% VS.300% of control). The differences between the two cell types could explain the minor differences between the two studies. In our study astroglia were cultured from cerebral cortices only, a relatively pure population of flat astroglial cells was purified from a mixed population of astroglia by differential adhesion technique, and the cells were tested within 2 weeks of plating. The differences were minimal, and the mitogenic responses of astroglial cells to EGF in this study are supportive of findings in the previous study of Wang et al. (17). Polpliker et al. (9) have previously shown that EGF is detectable in the circulation of the rat at this stage, but EGF mRNA is not detectable in the kidney until the latter part of the first week of postnatal age. Another member of the EGF family of peptides, TGFLu, may be the alternative ligand of biological importance. TGFa is detectable in the rat embryo from early gestation (34-36) and in the maternal decidua immediately after implantation (26). Recently, we have demonstrated the presence of TGFcv mRNA in the rat brain (18). Other investigators have also demonstrated the presence of TGFa mRNA and/or precursor in the growing or adult rat brain under normal (19) or injury repair conditions (20). TGFa is found to be equally or slightly less potent than EGF in this system (our unpublished observations) and, therefore, could be a more relevant ligand. Whether the biologically active ligand is EGF or TGFa, newborn rat astrocytes possess high affinity EGF receptors (17) that may bind to either of them. Judging from the comparative dose-response study, astroglia are more sensitive to EGF than IGF-I or IGF-II, suggesting an important biological role for either EGF or TGFa in the proliferation of astroglia. The synergistic actions between EGF and IGF-I in stimulating DNA synthesis could be due to changes in the affinity or number of IGF receptors in the presence of EGF or vice versa. We have tested the former by determining specific [‘251]IGF-I binding to astroglia in the presence of different concentrations of EGF and have not demonstrated any changes in either the receptor affinity or number (data not shown). Similarly, specific [lz51]EGF binding to astroglia was not altered in the presence of IGF-I. Specific IGF-I or EGF receptor-blocking antibodies are not available in the rat, and therefore, receptor-blocking studies could not be performed to confirm the results directly. These studies indicate that the synergistic actions of IGF-I and EGF are unlikely to be due

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 18:14 For personal use only. No other uses without permission. . All rights reserved.

1140

EGF AND IGFs ON ASTROGLIAL

GROWTH

Endo. 1992 Vol131.

FIG. 7. Northern blot hybridization of total RNA from primary astroglia for mRNAs encoding IGF-I (A) and IGF-II (B). Total RNA from adult and fetal (20 days gestation) rat liver (20 pg each lane) are loaded in lanes 1 and 2, respectively, to compare the relative abundance as well as the different mRNA sizes expressed by astroglia with those in liver. Total RNA (20 rg each lane) was loaded from controls (lane 3) and astroglia treated with 10 rig/ml EGF (lane 4). Astroglia treated with 50 rig/ml IGF-I (lane 5) and 1 pg/ml insulin (lane 6) are shown in comparison. The autoradiogram in A was exposed for 2 days at -70 C. The autoradiogram in B was exposed for 1 day at -70 C to avoid overexposure of IGF-II mRNAs in fetal rat liver. To demonstrate that equal quantities of total RNA were loaded into each lane, the Northern blot was stripped and probed with radiolabeled 18s ribosomal cDNA. Densitometric analysis of the hybridizing IGF-I and IGF-II mRNAs did not show any difference between control and EGF treatment (three experiments were performed).

A

LIVER

B

ASTROGLIA

--

LIVER

No 3

ASTROGLIA

--

7.4 kb +

+

3.6 kb

2.5 kb 2.0 kb -

1.2 kb +

18s +

1

2

to alterations in the affinity or number of receptors for either EGF or IGF-I. We went on to test the other possibility that EGF is altering the local synthesis of IGF-I or IGF-II, or vice ~ersp. Since EGF or TGFcv peptide was not detectable in the astroglial cell-conditioned medium nor were the mRNAs encoding these peptides present in these cells (our unpublished observations), it is unlikely that IGF-I or IGF-II was altering the astroglial synthesis of EGF or TGFa. The availability of the bioinhibitory antibody to IGF-I (Sm 1.2) allowed us to test whether EGF was altering the local synthesis of IGF-I. Sm 1.2 has been reported to be inhibitory to IGF-I action in a wide variety of cells in many species (37). However, it has been shown to recognize IGF-II with equivalent affinity (29) (Underwood, L. E., personal communication), and therefore, it is possible that the effects of EGF could also be due to alterations in the local synthesis of IGF-II. Analysis of radioimmunoassayable concentrations of IGFI and IGF-II in the conditioned medium of astroglia treated with EGF indicates that EGF stimulates astroglial IGF-I synthesis, but not IGF-II synthesis. Similar findings have been reported in primary astroglia and C6 glioma cell lines in a preliminary report by Chemausek et al. (38). Since we have recently shown that EGF increases IGFBP-2 production by astroglial cells (30), it is also possible that the increased level of IGFBP-2 in the medium could be interfering with the IGF RIA, particularly for IGF-II. However, we have validated the efficiency of our method of stripping IGFs from their binding proteins in conditioned medium from a wide variety of cell types (23), including astroglial cell-conditioned medium, and have found it to be over 95% efficient. It is, therefore, unlikely that the increase in IGFBP levels in the astroglial cell-conditioned medium influenced the detection of changes

3

4

5

6

1

2

3

4

5

6

in IGF concentrations, Previous investigators have shown that the developing brain expresses both IGF-I and IGF-II mRNAs (10). However, in situ hybridization studies in adult rats have demonstrated that strong hybridization was observed in the choroid plexus, with little or no hybridization signal observed in other regions of the brain (39, 40). Similar findings were obtained in the rat embryo (41). These studies, however, do not exclude the possibility that IGF-I or IGF-II mRNAs are expressed in the brain in low abundance, but only indicate that the choroid plexus expresses a greater abundance of IGF-II mRNA than any other brain region. Using a sensitive RNAse protection assay, Rotwein et al. (10) demonstrated the presence of IGFI mRNA in developing and adult brains and cultured astroglia. In addition, studies using in situ hybridization in developing rodents did not investigate the newborn in the first l2 weeks, when gliogenesis is most active, and therefore, it is possible that failure to detect IGFs in the developing brain may be due in part to the different developmental stages at which the studies were performed. In this study primary astroglia were cultured from l-day-old rats and were investigated after approximately 2 weeks in culture. It is, however, also possible that IGF mRNAs expressed in these cells may be induced by tissue culture conditions. Recent studies in brain injury models, which demonstrated that astroglia at the site of injury express IGF-I and/or IGF-II (42), also suggest that astroglia in culture may represent an injury response. This study clearly demonstrates that cultured astroglia express both IGF-I and IGF-II genes, and that specific mRNAs are detectable using Northern blot analysis on total RNA samples. Comparison with adult and fetal liver in the

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 18:14 For personal use only. No other uses without permission. . All rights reserved.

EGF AND IGFs ON ASTROGLIAL same Northern blot shows that astroglia express the same size IGF-I transcripts. The abundance is similar to that in the 20-day gestation fetal liver, which is at the level astroglia may be expected to express. IGF-I mRNA levels, as quantified by densitometry, were not different in the EGF-treated cells and untreated control cells. IGF-I- and insulin-treated cells also did not show any alteration in IGF-I mRNA abundance. The increase in radioimmunoassayable levels of IGF-I in the conditioned medium of EGF-treated astroglia, with no change in stable IGF-I mRNA levels, suggests that the effect of EGF on astroglial IGF-I production is unlikely to be at the level of gene transcription. The regulation may be at any level of posttranscriptional processing of IGF-I, be it at the level of translation, posttranslational modification of IGF-I precursor, or secretion. It is also possible that Northern blot analysis may not be sensitive enough to detect the changes. Using a more sensitive solution hybridization (RNAse protection) assay, we were not able to demonstrate any alteration in IGF-I mRNA levels (data not shown), and therefore, it is unlikely that EGF has a regulatory action on IGF-I gene expression. It was interesting to note that astroglia expressed only the 3.5-kb IGF-II transcript, in contrast to the fetal liver, which expressed six IGF-II mRNAs with sizes ranging from 1.2-4.6 kb. The 3.5-kb mRNA is the most abundant of all transcripts, possibly originates at the 3’-promoter, and is transcribed by the use of exons 3, 4,5, and 6 (43). The presence of only the 3.5-kb mRNA in the astroglia indicates that transcription of the IGF-II gene in the astroglia is significantly less complex than that in any other tissue (43) and offers an. excellent system to study the activities and regulation of the 3’promoter. The basal (control) level of expression of IGF-II mRNA by the astroglia is not influenced by EGF, indicating that the potentiating action of EGF and IGF-I and the reduction in EGF mitogenic activity by anti-IGF antibody Sm 1.2 are not due to the effect of EGF on astroglial IGF-II synthesis. In summary, we have demonstrated that both EGF and IGFs have mitogenic activity on the developing astroglia of the neonatal rat cerebral cortex. EGF appears to be more potent than the IGFs in stimulating DNA synthesis. EGF potentiates the actions of IGF by stimulating the astroglial synthesis of IGF-I and not IGF-II. This action of EGF is not at the level of gene transcription, but could be at any level of posttranscriptional processing of IGF-I mRNA. Interactions between different growth factors, EGF and IGF-I in this case, may play an important role in the growth and development of the central nervous system. Acknowledgments We would like to thank Dr. David Hill and Mrs. D. De Souza for their help in the RIA of IGF-I and IGF-II. We are also grateful to the following colleagues for their generous gifts: Drs. L. E. Underwood and J. J. Van Wyk of University of North Carolina (Chapel Hill, NC), for the monoclonal antibody against IGF-I (Sm 1.2), Dr. Liam Murphy of University of Manitoba (Winnipeg, Manitoba, Canada) for the rat IGF-I cDNA, Dr. Greame Bell of University of Chicago (Chicago, IL) for the mouse IGF-II cDNA, and Dr. David Denhardt of Rutgers University for the 18s ribosomal cDNA probe. We also wish to thank Dr. L. Stitt of the Department of Epidemiology and Biostatistics, University of Western

GROWTH

Ontario, for advice tarial assistance.

1141

on statistical

methods,

and Ms.

K. Ayers

for secre-

References 1. Gurney ME 1990 Peptide growth factors and the nervous system. In: Spdrn MB, Roberts AB reds) Peptide Growth Factors and Their Receutors II. Surineer-Verlaa, New York, vol 2:345-369 2. Shodter EM 1991 Neurotroihic factors in the brain: where do IGFs fit?. In: Spencer EM (ed) Modern Concepts of Insulin-Like Growth Factors. Elsevier, New York, pp 297-307 3. Chao MV 1990 Nerve growth factor. In: Sporn MB, Roberts AB (eds) Peptide Growth Factors and Their Receptors II. SpringerVerlag, New York, vol 2:135-166 4. Ernfors P, Ibanez CF, Ebendal T, Olson L, Persson H 1990 Molecular cloning and neurotrophic activities of a protein with structural similarities to nerve growth factor: developmental and topographical expression in the brain. Proc Nat1 Acad Sci USA 87:5454-5458 5. Rappolee DA, Werb Z 1991 Endogenous insulin-like growth factor II mediates growth in preimplantagon mouse embryos. In: Spencer EM (ed) Modern Concepts of Insulin-Like Growth Factors. Elsevier, Ned York, pp 3-8 * 6. Bondy CA, Werner H, Roberts CT, LeRoith D 1990 Cellular pattern of insulin-like growth factor I (IGF-I), type 1 IGF receptor gene expression in early organogenesis: comparison with IGF-II gene expression. Mol Endocrinol4:1386-1398 7. Dobbing J, Sands J 1979 Comparative aspects of the brain growth spurt. Early Hum Dev 3:79-83 8. Skoff RI’ 1980 Neuroglia: a reevaluation of their origin and development. Path01 Res Pratt 168:279-300 9. Popliker M, Shatz A, Avivi A, Ullrich A, Schlessinger J, Webb CG 1987 Onset of endogenous synthesis of epidermal growth factor in neonatal mice. Dev Biol 119:38-44 10. Rotwein P, Burgess S, Milbrandt JD, Krause JE 1988 Differential expression of insulin-like growth factor genes in rat central nervous system. Proc Nat1 Acad Sci USA 85:265-269 11. Renoir D, Honegger P 1983 Insulin-like growth factor I (IGF I) stimulates DNA svnthesis in fetal rat brain cell cultures. Dev Brain Res 7:205-213 ’ 12. Han VKM, Lauder JM, D’Ercole AJ 1987 Characterization of somatomedin/insulin-like growth factor receptors and correlation with biologic actions in cultured neonatal rat astroglial cells. J Neurosci 7:501-511 13. Han VKM, Lauder JM, D’Ercole AJ 1988 Rat astroglial somatomedin/insulin-like growth factor binding proteins: characterization and evidence of biologic function. J Neurosci 8:3135-3143 14. Leutz A, Schachner M 1981 Epidermal growth factor stimulates DNA synthesis of astrocytes in primary cerebellar cultures. Cell Tissue Res 220:393-404 15 Guentert-Lauber 8, Honegger P 1985 Responsiveness of astrocytes in serum-free aggregate cultures to epidermal growth factor: dependence on cell cycle and the epidermal growth factor concentration. Dev Neurosci 7:286-295 R, Condorelli DR, Surrentino S, Turpeenoja L, Costa A, 16 Avola Giuffrida-Stella AM 1988 Effect of epidermal growth factor and insulin on DNA, RNA and cytoskeletal protein labeling in primary rat astroelial cell cultures. I Neurosci Res 19:230-238 17. Wang Si, Shiverick KT, bgilvie S, Dunn WA, Raizada MK 1989 Characterization of epidermal growth factor receptors in astrocytic glial and neuronal cells in primary culture. Endocrinology 124:240247 18. Seroogyi KB, Han VKM, Lee DC 1991 Regional expression of transforming growth factor-a mRNA in the rat central nervous system. Neurosci Lett 125:241-245 JH, Annis CM, Gentry LE, Twardzik DR, Loughlin SE 19. Fallon 1990 Localization of cells containing transforming growth factoralpha precursor immunoreactivity in the basal ganglia of the adult rat brain. Growth Factors 2:245-250 20. Junier Ml’, Ying JM, Costa ME, Hoffman G, Hill DF, Ojeda SR 1991 Transforming growth factor o( contributes to the mechanism

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 18:14 For personal use only. No other uses without permission. . All rights reserved.

1142

21. 22.

23.

24.

25.

26.

EGF

AND

IGFs

ON ASTROGLIAL

by which hypothalamic injury induces precocious puberty. Proc Nat1 Acad Sci USA 889743-9747 Hill DJ 1990 Relative abundance and molecular size of immunoreactive insulin-like growth factors 1 and II in human fetal tissues. Early Hum Dev 21:49-58 Richards GM 1974 Modifications of the diphenylamine reaction giving increased sensitivity and simplicity in the estimation of DNA. Anal Biochem 57369-376 Hill DJ, De Sousa D 1990 Insulin is a mitogen for isolated epiphyseal growth plate chondrocytes from the fetal lamb. Endocrinology 126:2661-2670 Romanus JA, Terre1 JE, Yang YW, Nissley SP, Rechler MM 1986 Insulin-like growth factor carrier proteins in neonatal and adult rat serum are immunologically different: demonstration using a new radioimmunoassay for the carrier protein from BRL-3A rat liver cells. Endocrinology 118:1743-1748 Han VKM, Lund PK, Lee DC, D’Ercole AJ 1988 Expression of somatomedin/insulin-like growth factor messenger ribonucleic acids in the human fetus: identification, characterization and tissue distribution. J Clin Endocrinol Metab 66:422-429

Han VKM, Hunter III ES, Pratt RM, Zendegui

for astrocytes. 32.

33.

34.

35.

36.

37.

38.

JG, Lee DC 1987

Expression of rat transforming growth factor alpha during development occurs predominantly in the maternal decidua. Mol Cell Biol7:2335-2343 27. Southern EM 1975 Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol98:503-517 28. Snedecor GW, Cochran WG 1980 Statistical Methods, ed 7. Iowa State University Press, Ames, pp 144-145 29. Han VKM, Hill DJ, Strain AJ, Towle AC, Lauder JM, Underwood LE, D’Ercole AJ 1987 Identification of somatomedin/insulin-like growth factor immunoreactive cells in the human fetus. Pediatr Res 22:245-251 insulin-like growth 30. Bradahaw S, Smith A, Han VKM, Astroglial factor binding proteins: synthesis and regulation of gene expression by hormones and growth factors. 2nd International Symposium on Insulin-Like Growth Factors/Somatomedins, San Francisco CA, 1991, p 268 (Abstract) M, Broderick JT 1989 A soluble brain molecule 31. Nieto-Sampedro related to epidermal growth factor receptor is a mitogen inhibitor

GROWTH

39.

40.

41.

42.

43.

J Neurosci

Endo. 1992 Vo1131.No3

Res 22:28-35

Reynolds BA, Weiss S 1992 Generation

of neurons and astrocytes from isolated cells of the adult mammalian central nervous system. Science 255:1707-1710 Oka Y, Ghishan FK, Greene HL, Orth DN 1983 Effect of mouse epidermal growth factor/urogastrone on the functional maturation of rat intestine. Endocrinology 112:940-944 Twardzik DR 1985 Differential expression of transforming growth factor-alpha during prenatal development of the mouse. Cancer Res 45:5413-5416 Lee DC, Rochford R, Todaro GJ, Villarreal LP 1985 Developmental expression of rat transforming growth factor-alpha mRNA. Mol Cell Biol5:3644-3646 Wilcox JN, Derynk R 1988 Developmental expression of transforming growth factor alpha and beta in mouse fetus. Mol Cell Biol 8:3415-3422 Russell WE, Van Wyk JJ, Pledger WJ 1984 Inhibition of the mitogenic effects of plasma by a monclonal antibody to somatomedin-C. I’roc Nat1 Acad Scl USA 81:2389-2392 Chernausek SD, Mitchell V, Hengsen HT, Production of insulinlike erowth factor I (IGF I). IGF bindine oroteins bv cells of neural origii. 71st Annual Meeting of The En&&ine Society, Seattle WA, 1990, p 354 (Abstract) Stylianopoulou F, Efstratiadis A, Herbert J, Pintar J 1988 Pattern of the insulin-like growth factor II gene expression during rat embryogenesis. Development 103:497-506 Hynes MA, Brooks PJ, Van Wyk JJ, Lund PK 1988 Insulin-like growth factor II messenger ribonucleic acids are synthesized in the choroid plexus of the rat brain. Mol Endocrinol 1:47-54 Beck F, Samani NJ, Penschow JD, Thorley B, Tregear GW, CoghIan JP 1987 Histochemical localization of IGF-I and -11 mRNA in the developing rat embryo. Development 101:175-184 Logan A, Hill DJ, Phillips ID, Gonzalez A-M, Insulin-like growth factor peptide and IGF II mRNA localization in the normal and injured rat brain. 2nd International Symposium on Insulin-Like Growth Factors/Somatomedins, San FranciscoCA, 1991, p 153 (Abstract) Daughaday WH, Rotwein P 1989 Insulin-like growth factors I and II. Peptide, messenger ribonucleic acid and gene structures, serum and tissue concentrations. Endocr Rev 10:68-91

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 18:14 For personal use only. No other uses without permission. . All rights reserved.