- Email: [email protected]
Basic fibroblast growth factor-2 and interleukin-1β regulate S100β expression in cultured astrocytes
Pergamon
PII:
Neuroscience Vol. 82, No. 2, pp. 33–41, 1998 Copyright ? 1997 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306–4522/98 $19.00+0.00 S0306-4522(97)00223-6
BASIC FIBROBLAST GROWTH FACTOR-2 AND INTERLEUKIN-1â REGULATE S100â EXPRESSION IN CULTURED ASTROCYTES D. A. HINKLE,* J. P. HARNEY,† A. CAI,† D. C. HILT,‡ P. J. YAROWSKY§ and P. M. WISE†¶ *Department of Physiology, University of Maryland School of Medicine, Baltimore, MD 21202, U.S.A. †Department of Physiology, University of Kentucky College of Medicine, Lexington, KY 40536, U.S.A. ‡AMGEN, 1840 DeHavilland Drive, Thousand Oaks, CA 91320, U.S.A. §Department of Pharmacology and Experimental Therapeutics, University of Maryland School of Medicine, Baltimore, MD 21202, U.S.A. Abstract––Basic fibroblast growth factor and interleukin-1â are known to regulate the expression of other trophic factors and to stimulate reactive gliosis in vivo. S100â is a glial-specific putative neurotrophic factor and has been considered a marker of the reactive status of astrocytes. Therefore, we tested the hypothesis that basic fibroblast growth factor-2 and interleukin-1â achieve their effects by altering S100â gene expression in cultured rat astrocytes using an RNase protection assay. Short-term treatment with basic fibroblast growth factor-2 produced a transient decrease in S100â messenger RNA which was followed by an increase after longer term treatment. In contrast, both short- and long-term treatment with interleukin-1â suppressed S100â messenger RNA. We measured levels of S100â nuclear primary transcript to assess whether alterations in transcriptional rate explain the changes in messenger RNA. Our results indicate that changes in transcription account for changes in steady state levels of messenger RNA since basic fibroblast growth factor-2-induced changes in S100â primary transcript temporally preceded changes in messenger RNA. We further measured intracellular S100â protein levels by enzyme-linked immunosorbent assay to determine whether changes in gene expression were translated into parallel changes in protein. Our results clearly demonstrate that basic fibroblast growth factor-2 and interleukin-1â influence the expression of the S100â gene, that this regulation appears to occur at the level of transcription, and that alterations in messenger RNA are sometimes, but not always, reflected in changes at the level of protein. These observations suggest that basic fibroblast growth factor-2 may amplify its trophic effects, in part, by influencing the expression of another trophic factor. ? 1997 IBRO. Published by Elsevier Science Ltd. Key words: FGF-2, IL-1â, S100â, astrocyte, growth factor, cell culture.
Basic fibroblast growth factor (FGF-2) and interleukin-1â (IL-1â) regulate neurotrophic factor gene expression in cultured astrocytes.33 FGF-2 is an astrocyte mitogen and morphogen in vitro;10 IL-1â is an astrocyte mitogen in vitro.12 In addition, these growth factors appear to play key roles in the glial response to brain lesion.27 Their levels increase early in the response to brain insult, and in vivo infusion of either into the brain can stimulate reactive changes in astrocytes.4,11–13 Thus, these factors may alter the
post-lesion neurotrophic milieu by regulating astrocyte genes and may stimulate reactive gliosis. S100â is a small, acidic, calcium-binding protein that is found predominantly in astrocytes. The homodimeric form of the protein is a well-characterized neuronal survival and neurite-extension factor in vitro, and can protect hippocampal neurons against hypoglycemia-induced cell death.3,20,34 It also modulates astrocyte proliferation and morphology in vitro, and is over-expressed in reactive astrocytes.15,29,35 Therefore, S100â is widely considered to be a marker of reactive astrocytes, and may be important in the mechanism through which these cells respond to brain injury.26 To assess whether increased S100â expression is a component of the glial response to factors that induce reactive gliosis, we measured S100â mRNA levels in cultured astrocytes from the rat cerebral cortex after treatment for various lengths of time
¶To whom correspondence should be addressed. Abbreviations: ABAM, antibiotic-antimycotic; DMEM, Dulbecco’s modified Eagle’s medium; EDTA, ethylenediaminetetra-acetate; ELISA, enzyme-linked immunosorbent assay; FGF-2, basic fibroblast growth factor-2; GFAP, glial fibrillary acidic protein; IL-1â, interleukin1â; PBS, phosphate-buffered saline; PIPES, piperazineN-N*-bis(ethanesulphonic acid); SDS, sodium dodecyl sulphate. 33
34
D. A. Hinkle et al.
with FGF-2 and/or IL-1â. We quantitated S100â primary transcript levels, an indirect measure of the rate of transcription, to determine the mechanism of the treatment-induced changes in mRNA. Finally, we assessed S100â protein levels to determine whether alterations in gene expression were translated into changes in protein levels.
astrocyte monolayers differed from the above in that cells were scraped into solution D with sodium acetate and phenol before sonification, then precipitated, pelleted, and resuspended in water without LiCl treatment. Agaroseformaldehyde gels and the ratio of A260 to A280 were used to determine the concentration and integrity of the purified RNA. Isolation of separate nuclear and cytoplasmic RNA fractions
EXPERIMENTAL PROCEDURES
Cell culture Enriched cerebral cortex astrocyte cultures were prepared from neonatal (P1) rat pups (Sprague–Dawley; Zivic– Miller) using techniques modified from McCarthy and deVellis.25 Briefly, the cerebral hemispheres from a litter of pups were pooled, dissociated into a cell suspension, then plated for three days at 37)C and 5% CO2 in Dulbecco’s modified Eagle’s medium/F12 (DMEM/F12) (1:1, Sigma) with 10% fetal calf serum (BioWhittaker) and 1#ABAM (antibiotic-antimycotic, Gibco). Media were replaced every three to four days until 10–11 days in vitro, at which time the flasks were shaken overnight to remove contaminating microglia and O-2A cells. The astrocyte monolayers were then sub-cultured into flasks in DMEM/F12 with 10% donor calf serum (BioWhittaker) and ABAM. Treatments with human recombinant FGF-2 (bFGF, Gibco) and/or human recombinant IL-1â (Boehringer Mannheim) were administered every other day in fresh 10% calf serumcontaining medium at varying concentrations for varying intervals as detailed in the Results. The sub-culture plating density was designed to produce approximately 80% confluent monolayers consisting of the same number of cells in each treatment group at the time of sample collection (see Results), since S100â levels change upon confluence.22 The cells in the cultures generated by this procedure were consistently composed of 98% type I astrocytes (Glial fibrillary acidic protein (GFAP)+/A2B5"). Immunocytochemistry Astrocytes were sub-cultured on poly--lysine-coated glass chamber slides (LabTek) for at least 24 h before treatment and staining. Slides were incubated with antiGFAP (1:250, DAKO) or anti-S100â (1:50, East Acres) antibodies to label astrocytes, anti-fibronectin (1:100, Sigma) to label fibroblasts, fluorescein isothiocyanateconjugated Griffonia simplicifolia Isolectin-B4 (25 µg/ml, Sigma) to label microglia,31 and/or Hoechst dye (1 µg/ml, Sigma) to label nuclei, then incubated with the appropriate fluorophore-conjugated secondary antibody (1:100, Jackson). Control stains were performed by omitting the primary antibody, where possible, and partially-enriched cultures of meningeal fibroblasts or shake-supernatant microglia were used to test the anti-fibronectin and the isolectin, respectively. In some cultures, anti-A2B5 (1:200, Boehringer Mannheim) immunostains were performed to detect type II astrocytes. Total RNA purification Total cellular RNA was isolated from rat tissues or astrocyte cultures using a modification of the method of Chomczynski.8 In brief, brains or livers were homogenized in solution D (4 M guanidine thiocyanate, 25 mM sodium citrate, 0.5% sarkosyl, 0.1 M â-mercaptoethanol) with phenol and sodium acetate (pH 5), sonified, chloroform extracted, precipitated in isopropanol, the RNA pelleted and resuspended in 4 M LiCl, re-pelleted, and resuspended in TES (10 mM Tris pH 7.5, 1 mM EDTA, 0.5% sodium dodecyl sulphate; SDS). The RNA was again precipitated, pelleted, washed, and resuspended in diethylpyrocarbonatetreated distilled water. Isolation of total RNA from
Nuclear and cytoplasmic RNA fractions were isolated from cultured astrocytes using the method of Roberts et al. with some modifications.17 Monolayers were scraped into phosphate-buffered saline (PBS) and pelleted, then resuspended in lysis buffer (10 mM Tris–HCl pH 7.4, 1.5 mM MgCl2, 0.2 M sucrose, 0.5% Nonidet P-40, 0.25% sodium deoxycholate) and homogenized without disrupting the nuclei. The lysate was layered over cushion buffer (10 mM Tris–HCl pH 7.4, 1.5 mM MgCl2, 0.4 M sucrose), then centrifuged to separate the cytoplasmic fraction and the nuclei. RNA was then isolated from each fraction. Northern analysis Northern analysis was performed using a modification of previously published methods.1 Briefly, total RNA (10– 40 µg) was fractionated through a 1% agarose/0.22 M formaldehyde/1# MOPS gel, transferred to a Nytran membrane, and attached by ultraviolet cross-linking (FisherBiotech FB-UVXL-1000) and vacuum baking. The membrane was rehydrated, pre-hybridized in hybridization buffer (5#standard saline citrate, 5#Denhardt’s solution, 0.1% SDS, 50% formamide, 50 mM phosphate buffer pH 5, and 100 µg/ml yeast RNA), then incubated for 15–17 h at 42)C in fresh hybridization buffer containing 5#105 c.p.m. of a G-50 column-purified, random primer-generated (Prime-a-Gene, Promega) cDNA probe labelled with [á-32P]dCTP. The rat S100â cDNA represents a 302 bp BamHI-HindIII fragment of the 3*-untranslated region (pS24), the genomic clone represents a 0.6 kb BamHIHindIII fragment of the intron 2/exon III boundary (pS1rg5D, sub-clone of pS100-1-5 from Dr Noboru Sueoka, University of Colorado). The membranes were stringently washed, dried, and exposed to film. Solution hybridization/RNase protection assay This assay was performed using a modification of previously published methods.1 Briefly, samples of total or cytoplasmic RNA (5 µg), or the entire nuclear fraction, were dried and resuspended in 30 µl of hybridization solution (80% v/v formamide, 40 mM PIPES, 0.4 M NaCl, 1 mM EDTA). To this was added approximately 1 ng of heatdenatured antisense cRNA probes (S100â cDNA or genomic and 1B15 cDNA representing a 111 bp PstI-BamHI fragment of the cyclophilin open reading frame, from Dr J. L. Roberts, Mt Sinai School of Medicine) which were labelled with [á-32P]UTP by in vitro transcription. A sample of each probe was assessed for percent incorporation, then purified on a G-50 RNA spin column (Boehringer Mannheim) before determination of specific activity. The purified probes and RNA were allowed to hybridize in solution at 45)C for 12–14 h, then incubated with RNase A (20 µg/ml) and RNase T1 (2 µg/ml) for 1 h at room temperature, the digestion terminated with SDS (0.5%) and proteinase K (120 ng/µl), then the RNA hybrids extracted with Tris-buffered phenol/chloroform, precipitated in ethanol, resuspended in denaturing RNA loading buffer (95% v/v formamide, 20 mM EDTA, 0.05% w/v Bromophenol Blue, 0.05% w/v xylene cyanol) and fractionated on a 4% acrylamide/7 M urea gel. The gel was fixed in 10% glacial
Regulation of S100â expression
35
acetic acid/10% methanol, transferred to 3MM paper, then dried and read in a PhosphorImager (Molecular Dynamics). Enzyme-linked immunosorbent assay S100â protein was quantitated in the cell lysates of cortical astrocytes, as detailed in Griffin et al. with minor modifications.14 To obtain cell lysates, monolayers were scraped into PBS, pelleted, and resuspended in PBS. The cells were then sonicated, the cellular debris pelleted, and the supernatant taken for analysis. Bovine brain S100 was used as the standard (East Acres); its S100â concentration was determined using VUSB-1, a well-characterized recombinant form of S100â protein.32 The Bradford protein assay (Bio-Rad) was used to determine the total protein concentration of each sample, which was used to normalize the S100â. Experimental design and statistical analysis Each replicate culture consisted of cells derived from a single litter of rat pups that were treated with FGF-2, IL-1â, both factors (unless otherwise stated), and a corresponding control. The total number of independent replicates (‘‘n’’) for each study is reported in each figure legend. To assess differences in S100â expression between cultures which were treated with FGF-2 and/or IL-1â, the data were analysed by least squares analysis of variance using a General Linear Models Procedure of the Statistical Analysis System (SAS, 1989). Cyclophilin mRNA or total protein were included in the analysis as covariates for the RNase protection and enzyme-linked immunosorbent assay (ELISA) data, respectively, to eliminate variability due to unequal loading of RNA or protein. Significant effects for treatment at any time point were further analysed using independent orthogonal contrasts to detect differences between FGF-2 and control, IL-1â and control, or combination treatment and FGF-2+IL-1â (this is a test for synergy that compares combination treatment to the additive effect of each factor alone at each time point). To evaluate changes in primary transcript, in which only single factor treatments were used, the data were analysed by paired t-tests. The criteria for statistical significance were set at P<0.05. RESULTS
Astrocyte morphology and culture characterization Treatments with FGF-2 (10 ng/ml) induced a robust stellation response which was first detectable after 24 h of treatment and became most highlydifferentiated after five to six days of treatment (data not shown). IL-1â (10 U/ml), in contrast, did not alter the flat, polygonal morphology that was seen in the controls at any time during the seven to eight day treatment period (data not shown). Fig. 1 shows examples of the morphological response of the astrocytes to 48 h of each treatment, a time point at which typical morphological changes were consistently visible. Astrocytes cultured in calf serum-containing media for 48 h or seven days with FGF-2 or IL-1â, or both, were sub-confluent and did not differ in cell number from controls as assessed by cell counts using the Hoechst nuclear dye (data not shown). Under these treatment conditions, the cultures were consistently composed of greater than 98% GFAP or S100â immunoreactive astrocytes, and not more than 0.5% fibroblasts and 0.1% microglia, as determined by counting the proportion of cells immunopositive for
Fig. 1. Comparison of cultured cortical astrocyte morphology by phase-contrast (A,C,E,G) and GFAP immunocytochemistry (B,D,F,H). The cells have been treated for 48 h with control (A,B), IL-1â at 10 U/ml (C,D), FGF-2 at 10 ng/ml (E,F), or both factors (G,H). FGF-2 stimulates morphological stellation, whereas IL-1â does not.
GFAP, S100â, fibronectin, or lectin GSA I-B4, respectively, relative to the total number of Hoechstlabelled cells (data not shown). S100â northern analysis Northern analysis validated the use of pS24 as a probe for S100â mRNA in the RNase protection assay (Fig. 2A). The pS24 (S100â) cDNA recognized a single 1.6 kb transcript in samples of total RNA from rat astrocyte cultures, rat C6 glioma cultures (an immortalized astrocyte cell line), and rat brain, but not rat liver. This transcript is the published size of rat S100â mRNA,21 and was also recognized by the rat S100â genomic clone pS1rg5D (data not shown). The sensitivity of this method was not sufficient to detect heteronuclear S100â RNA. Effects of basic fibroblast growth factor-2 and/or interleukin-1â on the levels of S100â messenger RNA A representative autoradiograph illustrates the changes seen in S100â mRNA after treatment for 24 h, 48 h, or seven days with FGF-2, IL-1â, both factors, or no treatment control (Fig. 2B). The S100â
36
D. A. Hinkle et al.
Fig. 2. A) Northern analysis of rat RNA using S100â cDNA probe (pS24). Total RNA from rat brain (Br, 10 µg), rat cortical astrocyte culture (As, 30 µg), rat C6 glioma cell culture (C6, 30 µg), and rat liver (Lv, 30 µg) was fractionated on an agarose/formaldehyde gel, transferred to a Nytran membrane, and incubated with a 32P-labelled cDNA probe. The transcript recognized is approximately 1.6 kb. B) Representative RNase protection assay autoradiograph from cultured cortical astrocytes treated for the times shown with control (C), FGF-2 at 10 ng/ml (F), IL-1â at 10 U/ml (I), or both factors (IF). 32 P-labelled cRNA probes (P) for S100â (369 nt) and cyclophilin (174 nt) were incubated with 5 µg samples of total RNA or tRNA (t), digested with RNases, then the protected fragments were fractionated on a 4% acrylamide/7 M urea gel. The protected fragments generated were 302 nt and 111 nt in length for S100â and cyclophilin, respectively.
cRNA probe hybridized with a single mRNA species and protected the expected 302 nucleotide fragment. Similarly, the cyclophilin cRNA probe produced the expected 111 nucleotide fragment. The specificity of each probe was further proven by its lack of hybridization with tRNA. In all assays, cyclophilin was used as an internal control since its levels are not regulated by FGF-2 or IL-1â in cultured cortical astrocytes.33 Each treatment paradigm (FGF-2, IL-1â, and FGF-2/IL-1â) had a distinct influence on S100â gene expression (Figs 2B, 3). FGF-2 (10 ng/ml) produced a biphasic effect on S100â gene expression: mRNA levels decreased after 12 h and 24 h of treatment to 86% (P<0.05) and 30% (P<0.001) of the control level, respectively, then were elevated 2.5-fold (P<0.01) after seven days of continuous FGF-2 treatment. IL-1â (10 U/ml) produced a similar decrease in mRNA levels, but with a markedly different pattern. After 48 h of treatment with the cytokine, S100â gene expression was inhibited to 40% of control levels (P<0.02), and continued to be suppressed after seven days of treatment (38% of control, P<0.03). Combination treatment with both FGF-2 and IL-1â produced a rapid, statistically greater suppression of S100â gene expression than the additive effect of each factor alone at 24 h and 48 h. After 24 h of treatment, S100â mRNA levels reached 23% of control (P<0.02, simultaneous treatment with both factors versus the additive effect of each singlefactor treatment), while after 48 h S100â was expressed at only 12% of the control level (P<0.05). After seven days of combination treatment, there was no difference between it and the additive effect of
Fig. 3. S100â gene expression in rat cortical astrocytes in response to treatment with FGF-2 (10 ng/ml) and/or IL-1â (10 U/ml), as measured in total RNA by the RNase protection assay. At each time point and for each treatment, the mean&S.E.M. relative S100â mRNA level (normalized to cyclophilin) is plotted relative to its control (designated as 1.0). Treatment effects are statistically significant by least squares ANOVA with orthogonal contrasts at P<0.05 as follows: FGF-2 or IL-1â versus control (*), combined treatment versus sum of single-factor treatment effects (**), n=3–6.
each factor alone. No changes in S100â mRNA levels were detected under any treatment conditions at 1 or 3 h (data not shown). Similar effects of FGF-2 and IL-1â on the patterns of S100â gene expression were seen in cultures from the hippocampus and hypothalamus (data not shown). Dose–response curves were also generated in cortical astrocyte cultures for each factor at the times
Regulation of S100â expression
37
Fig. 4. Characterization of protected fragments generated by 32P-labelled genomic S100â (S), cyclophilin (C), or both (C+S) cRNA probes in nuclear or cytoplasmic RNA from cultured cortical astrocytes. The probes were incubated with whole nuclear RNA fractions, 5 µg of cytoplasmic RNA, or tRNA (t), digested with RNases, and the protected fragments fractionated on a 4% acrylamide/7 M urea gel. The protected fragments are labelled on the left side of panel A, the undigested S100â (SP) and cyclophilin (CP) probes are shown, and panel B is an over-exposure of the nuclear RNA that better demonstrates the protected primary transcript. Experiments utilized freshly-made probe (less than one-day-old) to decrease the amount of probe breakdown product and to maximize the signal of the primary transcript band, and the protected fragments were fractionated further on the gel than is shown here to prevent signal bands from overlapping.
that they maximally inhibited S100â gene expression in the above experiments. FGF-2 treatment for 24 h produced maximal suppression of the S100â gene at doses of 1–10 ng/ml. 48 h treatment with IL-1â at 1–10 U/ml produced maximal gene suppression (data not shown). Effects of basic fibroblast growth factor-2 or interleukin-1â on the levels of S100â primary transcript S100â primary transcript levels were quantitated as an indirect measure of the rate of transcription. Approximately 0.15% of the total S100â RNA is in the form of nuclear primary transcript, 1.8% is nuclear mRNA, and the rest is cytoplasmic mRNA (data not shown). The pS1rg5D genomic clone encodes 0.6 kb of S100â sequence at the intron 2/exon III boundary, and produces a protected fragment of this size when it hybridizes with primary transcript (which has intron and exon sequence), or of 272 nucleotides when it hybridizes with mature transcript (mRNA, which has only exon sequence, see Fig. 4). FGF-2 treatment produced a biphasic effect on S100â nuclear primary transcript: it decreased its levels to 72% (P<0.04) and 45% (P<0.04) of controls after 6 h and 12 h of treatment, respectively, then induced an approximately three-fold increase in primary transcript (P<0.04) after 48 h of treatment
(Fig. 5A). Nuclear mRNA decreased to 47% (P<0.02) of the control level after 24 h of treatment. In the cytoplasm, mRNA levels decreased to 79% (P<0.05), 37% (P<0.04), and 69% (P<0.03) of the control level after 12 h, 24 h, and 48 h of treatment, respectively. Thus, the FGF-induced changes in nuclear primary transcript preceded those in nuclear mRNA, which in turn preceded those in cytoplasmic mRNA. In addition, the steep rise in primary transcript at 48 h preceded the elevated expression of mRNA after seven days of treatment reported above. No significant changes were seen after 1 h or 3 h of treatment. Treatment with IL-1â (10 U/ml) for 48 h decreased S100â nuclear primary transcript, nuclear mRNA, and cytoplasmic mRNA levels to 45% (P=0.058), 47% (P=0.056), and 55% (P<0.04) of control levels, respectively (Fig. 5B). The cytoplasmic mRNA decreased to 72% (P<0.05) of control levels after 24 h of IL-1â treatment despite the lack of a significant change in the primary transcript. No significant changes were evident at 3 h, 6 h, or 12 h. Effects of basic fibroblast growth factor-2 and/or interleukin-1â on the levels of S100â protein S100â protein levels from cell lysates were quantitated by an ELISA. The first significant changes were detected after five days and eight days of exposure to FGF-2, at which times S100â protein was elevated
38
D. A. Hinkle et al.
Fig. 6. S100â protein level in cortical astrocytes in response to treatment with FGF-2 (10 ng/ml) and/or IL-1â (10 U/ml). S100â was quantitated by an ELISA, and total protein by a Bradford assay. At each time point and for each treatment, the mean&S.E.M. relative S100â protein level (normalized to total protein) is plotted relative to its control (designated as 1.0). Treatment effects are statistically significant by least squares ANOVA with orthogonal contrasts at P<0.05 for treatment versus control (*), n=4.
Fig. 5. A) S100â nuclear primary transcript, nuclear mRNA, or cytoplasmic mRNA level in cortical astrocytes in response to FGF-2 (10 ng/ml) as measured by the RNase protection assay. At each time point, the mean&S.E.M. relative S100â primary transcript RNA level (normalized to cyclophilin) is plotted relative to its control (designated as 1.0). Treatment effects are statistically significant by paired t-tests at P<0.05 (*), n=3–6. B) S100â nuclear primary transcript, nuclear mRNA, or cytoplasmic mRNA level in cortical astrocytes in response to IL-1â (10 U/ml) as measured by the RNase protection assay. At each time point, the mean&S.E.M. relative S100â primary transcript RNA level (normalized to cyclophilin) is plotted relative to its control (designated as 1.0). Treatment effects are statistically significant by paired t-tests at P<0.05 (*), n=3–6.
3.9-fold (P<0.02) and 10.1-fold (P<0.002), respectively, relative to their controls (Fig. 6). No other treatments produced significant changes in S100â protein concentrations. DISCUSSION
The major goals of this work were to (i) assess whether the trophic effects of FGF-2 and/or IL-1â may, in part, be achieved by influencing the expression of another growth factor, S100â and (ii) deter-
mine whether these trophic factors may act through S100â to initiate changes in astrocytes that are similar to those seen in reactive gliosis. Our data clearly demonstrate that FGF-2 and IL-1â regulate S100â gene expression in cultured rat cortical astrocytes. Further, they support the conclusion that the regulation of this gene occurs, at least in part, at the level of transcription, and that the resulting alterations in mRNA are translated into changes at the level of protein under some circumstances. They suggest that growth factors may reinforce or attenuate their effects by altering the expression of other growth factors. Finally, they suggest that elevated levels of S100â may support or enhance morphological differentiation in vitro, in the long term, that is similar to that seen in reactive gliosis in vivo. However, no clear conclusions can yet be drawn about the involvement of S100â in the mechanisms through which these factors may induce reactive gliosis, since (i) the initiation of morphological changes by FGF-2 occurs in the presence of diminished S100â expression, (ii) IL-1â stimulated no morphological changes, and (iii) it remains unclear how closely our cultures model reactive change in astrocytes in vivo. The appearance of reactive astrocytes after brain lesion is a highly characteristic phenomenon. Although it is known that tissue damage ‘‘stimulates’’ reactive gliosis, the exact mechanisms that underlie its initiation are unknown, as are the consequences of this change to intercellular interactions. Both FGF-2 and IL-1â can increase the number of GFAPimmunoreactive astrocytes after direct infusion into the brain4,11,12,13 and are therefore considered stimulators of reactive gliosis. GFAP and S100â are overexpressed in reactive astrocytes under many lesion conditions. Therefore, it is possible that FGF-2 and/or IL-1â may act through the S100â gene to
Regulation of S100â expression
initiate the reactive phenotype, and/or that the altered expression of this putative neurotrophic factor may influence neuron–glia interactions. To begin to test these ideas, we hypothesized that FGF-2 and/ or IL-1â may influence S100â gene expression in cultured astrocytes. FGF-2 stimulated astrocytes to become stellate, whereas IL-1â did not induce any morphological change. Throughout the study, our major concern was to maintain equal cell density in all treatments at the time of sample collection since S100â expression is regulated by the level of culture confluence.22 We did not detect any effect of either factor on the number of cells after 48 h or seven days of treatment, compared to controls. The reported proliferative effects of these factors are often based on measurements of [3H]thymidine incorporation into astrocyte DNA, and are often assessed in low-serum or serumfree media.2,12 Our data are based solely on counts of cells grown in 10% calf serum, which may have stimulated maximum proliferation and thus obscured any treatment effects. FGF-2 and IL-1â induced different effects on cortical astrocyte S100â mRNA. FGF-2 induced a distinct biphasic effect in gene expression: suppression in response to short-term treatment and stimulation after longer-term treatment. In contrast, the effect of IL-1â was only inhibitory. Combination treatment, using each factor at its maximally effective dose, resulted in a synergistic effect. Together, these findings imply that the regulatory mechanisms employed by each factor in astrocytes are different. Our findings in cortical astrocytes strongly suggest that FGF-induced alterations in S100â mRNA levels were due to changes in the rate of gene transcription. Numerous investigators have demonstrated that changes in primary transcript reflect changes in transcription since the half-life of heteronuclear RNA is extremely short.16,23,24 We therefore quantitated the level of primary transcript in the nucleus as an indirect indicator of gene transcription. S100â nuclear primary transcript was significantly inhibited by FGF-2 at least 12 h before significant decreases were detected in the nuclear and cytoplasmic mRNA, and increased significantly after longer-term treatment. The pattern of the alterations in primary transcript preceded and paralleled those seen in mRNA in response to both short- and longer-term treatment. Further, the magnitudes of the relative decrease and increase in primary transcript were similar to that of cytoplasmic mRNA. These findings clearly implicate transcriptional control as a mechanism through which FGF-2 regulates S100â. Whether or not the IL-1â-induced effects on S100â mRNA levels are explained by regulation of transcription is less clear, since changes in primary transcript did not precede those in mRNA. It is possible that additional mechanisms are involved, such as the regulation of mRNA stability.
39
The FGF-induced elevation in S100â mRNA was translated into an increase in intracellular S100â protein. Longer-term FGF-2 treatment significantly elevated S100â protein levels, relative to controls, which may have been due to the increased mRNA levels seen at this time, and/or to a decreased rate of protein turnover. The biological repercussions of this elevation of S100â within the astrocytes are unknown, but may include the stimulation of morphological differentiation, glial filament turnover, protein phosphorylation, and/or calcium signal transduction.5,6,19 S100â is a positive regulator of morphological stellation in glial cells, and we observed that the relative elevation in protein occurred at approximately the same time that our cultures reached their greatest level of morphological differentiation.35 Thus, elevated levels of S100â induced by longerterm treatment with FGF-2 appear to have supported, or even enhanced, the morphological differentiation that we observed in the astrocyte cultures. Whether this resulted from the action of S100â as an intracellular or extracellular modulator, which has been described in other systems, was not evaluated.28,30 These data further link S100â with morphological changes in cultured astrocytes which may be similar to changes seen in reactive astrocytes in vivo. However, we did not detect an inhibition of S100â protein in response to short-term FGF-2 treatment despite the marked suppression of mRNA. This lack of a change in protein may be due to the transient nature of the decrease in mRNA in the face of an abundant and stable protein pool. Van Eldik et al. have demonstrated in cultured glial cells that there is a delay of 24–48 h before changes in S100â mRNA are reflected in changes in protein.29 These data suggest that mechanisms other than S100â are important in the initiation of morphological differentiation by FGF-2, since decreased gene expression and unchanged protein levels did not appear to hinder the development of morphological stellation. Therefore, it remains difficult to conclude what role S100â plays in the early stages of the response of astrocytes to FGF-2, and no clear hypothesis can yet be generated relevant to its potential role in FGF-induced reactive gliosis. IL-1â treatment produced a trend towards a decrease in S100â protein that did not reach statistical significance, but was able to suppress the stimulatory FGF-2 effect since combination treatment produced an increasing trend in S100â protein that was not statistically significant. The inability of the IL-1âinduced changes to reach statistical significance suggests that S100â protein does not decrease in response to this treatment, but may reflect the limits of the sensitivity of the ELISA, particularly when compared to the RNase protection assay, to detect decreases in S100â below a low baseline. It is also possible that more complex mechanisms exist for the regulation of S100â protein at the level of translation or turnover. These data, combined with the lack of
40
D. A. Hinkle et al.
morphological changes in response to IL-1â, make it difficult to hypothesize what role, if any, S100â plays in IL-1â-induced reactive gliosis in vivo. CONCLUSIONS
In summary, we have demonstrated that both FGF-2 and IL-1â regulate S100â expression in cultured rat astrocytes. Our findings lend further credence to the concept that trophic factors often act by enhancing or suppressing other trophic factors.7,9,18 They also emphasize that elevated S100â, at least in
the long term, may contribute to changes in cultured astrocytes that may be similar to those seen in reactive gliosis. However, our data do not yet allow conclusions to be drawn relevant to the role of S100â in the mechanisms of FGF-2- or IL-1â-induced reactive change in astrocytes. Acknowledgements—We wish to thank Katherine Rosewell and Suzanne Steman for their superb technical assistance. This work was supported by NIH AG02224 to PMW. David Hinkle was a predoctoral trainee of NIH Training Grant HD07170.
REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.
Ausubel F. M., Brent R., Kingston R. E., Moore D. D., Seidman J. G., Smith J. A. and Struhl K. (1994) Current Protocols in Molecular Biology. John Wiley. Avola R., Ragusa N., Reale S., Costa A., Insirello L. and Stella A. M. G. (1993) Effect of growth factors on macromolecular synthesis in primary rat astroglial cell cultures. Annals NY Acad. Sci. 692, 192–200. Barger S. W., VanEldik L. J. and Mattson M. P. (1995) S100â protects hippocampal neurons from damage induced by glucose deprivation. Brain Res. 677, 167–170. Barotte C., Eclancher F., Ebel A., Labourdette G., Sensenbrenner M. and Will B. (1989) Effects of basic fibroblast growth factor (bFGF) on choline acetyltransferase activity and astroglial reaction in adult rats after partial fimbria transection. Neurosci. Lett. 101, 197–202. Baudier J., Delphin C., Grunwald D., Khochbin S. and Lawrence J. J. (1992) Characterization of the tumor suppressor protein p53 as a protein kinase C substrate and a S100b-binding protein. Proc. natn. Acad. Sci. U.S.A. 89, 11627–11631. Bianchi R., Giambanco I. and Donato R. (1993) S-100 protein, but not calmodulin, binds to the glial fibrillary acidic protein and inhibits its polymerization in a Ca2+-dependent manner. J. biol. Chem. 268, 12669–12674. Carroll P., Sendtner M., Meyer M. and Thoenen H. (1993) Rat ciliary neurotrophic factor (CNTF): gene structure and regulation of mRNA levels in glial cell cultures. Glia 9, 176–187. Chomczynski P. and Sacchi N. (1987) Single-step methods of RNA isolation by acid guanidinium thiocyanate-phenolchloroform extraction. Analyt. Biochem. 162, 156–159. DeLeon M., Van Eldik L. J. and Shooter E. M. (1991) Differential regulation of S100B and mRNAs coding for S100-like proteins (42A and 42C) during development and after lesion of rat sciatic nerve. J. Neurosci. Res. 29, 155–162. Engele J. and Bohn M. C. (1991) The neurotrophic effects of fibroblast growth factors on dopaminergic neurons in vitro are mediated by mesencephalic glia. J. Neurosci. 11, 3070–3078. Finklestein S. P., Apostolides P. J., Caday C. G., Prosser J., Philips M. F. and Klagsbrun M. (1988) Increased basic fibroblast growth factor (bFGF) immunoreactivity at the site of focal brain wounds. Brain Res. 460, 253–259. Giulian D. and Lachman L. B. (1985) Interleukin-1 stimulation of astroglial proliferation after brain injury. Science 228, 497–499. Giulian D., Woodward J., Young D. G., Krebs J. F. and Lachman L. B. (1988) Interleukin-1 injected into mammalian brain stimulates astrogliosis and neovascularization. J. Neurosci. 8, 2485–2490. Griffin W. S. P., Stanley L. C., Yeralan O., Ronvnaghi C. R. and Marshak D. R. (1993) Methods for the study of cytokines in human neurodegenerative disease. Meth. Neurosci. 17, 268–287. Griffin W. S. T., Stanley L. C., Ling C., White L., MacLeod V., Perrot L. J., White C. L. III and Araoz C. (1989) Brain interleukin 1 and S-100 immunoreactivity are elevated in Down syndrome and Alzheimer disease. Proc. natn. Acad. Sci. U.S.A. 86, 7611–7615. Herman J. P., Schafer M. K.-H., Watson S. J. and Sherman T. G. (1991) In situ hybridization analysis of arginine vasopressin gene transcription using intron-specific probes. Molec. Endocr. 5, 1447–1456. Jakubowski M. and Roberts J. L. (1991) Multiplex solution hybridization-ribonuclease protection assay for quantitation of different ribonucleic acid transcripts from snap-frozen neuroendocrine tissues of individual animals. J. Neuroendocr. 4, 79–89. Johnson-Green P. C., Dow K. E. and Riopelle R. J. (1992) Neurite growth modulation associated with astrocyte proteoglycans: influence of activators of inflammation. Glia 5, 33–42. Kligman D. and Hilt D. C. (1988) The S100 protein family. Trends biochem. Sci. 13, 437–442. Kligman D. and Marshak D. R. (1985) Purification and characterization of a neurite extension factor from bovine brain. Proc. natn. Acad. Sci. U.S.A. 82, 7136–7139. Kuwano R., Usui H., Maeda T., Fukui T., Yamanari N., Ohtsuka E., Ikehara M. and Takahashi Y. (1984) Molecular cloning and the complete nucleotide sequence of cDNA to mRNA for S-100 protein of rat brain. Nucl. Acids Res. 12, 7455–7465. Labourdette G. and Marks A. (1975) Synthesis of S-100 protein in monolayer cultures of rat-glial cells. Eur. J. Biochem. 58, 73–79. Lorang D., Lundblad J. R., Blum M. and Roberts J. L. (1991) Hormonal regulation of proopiomelanocortin transcription and RNA processing in AtT20 cells. Endocrinology 128, 509–540. Maurer R. A. (1982) Estradiol regulates the transcription of the prolactin gene. J. biol. Chem. 257, 2133–2136. McCarthy K. D. and DeVellis J. (1980) Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue. J. Cell Biol. 85, 890–902.
Regulation of S100â expression 26. 27. 28. 29. 30. 31. 32. 33. 34. 35.
41
Miyake T., Hattori T., Fukuda M. and Kitamura T. (1989) Reactions of S-100-positive glia after injury of mouse cerebral cortex. Brain Res. 489, 31–40. Neary J. T., Baker L., Jorgensen S. L. and Norenberg M. D. (1994) Extracellular ATP induces stellation and increases glial fibrillary acidic protein content and DNA synthesis in primary astrocyte cultures. Acta neuropath. 87, 8–13. Selinfreund R. H., Barger S. W., Pledger W. J. and Van Eldik L. J. (1991) Neurotrophic protein S100beta stimulates glial cell proliferation. Proc. natn. Acad. Sci. U.S.A. 88, 3554–3558. Selinfreund R. H., Barger S. W., Welsh M. J. and Van Eldik L. J. (1990) Antisense inhibition of glial S100beta production results in alterations in cell morphology, cytoskeletal organization, and cell proliferation. J. Cell Biol. 111, 2021–2028. Shashoua V. E., Hesse G. W. and Moore B. W. (1984) Proteins of the brain extracellular fluid: evidence for release of S-100 protein. J. Neurochem. 42, 1536–1541. Streit W. J. (1990) An improved staining method for rat microglial cells using the lectin from Griffonia simplicifolia (GSA I-B4). J. Histochem. Cytochem. 38, 1683–1686. Van Eldik L. J., Staecker J. L. and Winningham-Major F. (1988) Synthesis and expression of a gene coding for the calcium-modulated protein S100â and designed for cassette-based, site-directed mutagenesis. J. biol. Chem. 263, 7830–7837. Vige X., Costa E. and Wise B. C. (1991) Mechanism of nerve growth factor mRNA regulation by interleukin-1 and basic fibroblast growth factor in primary cultures of rat astrocytes. Molec. Pharmac. 40, 186–192. Winningham-Major F., Staecker J. L., Barger S. W., Coats S. and Van Eldik L. J. (1989) Neurite extension and neuronal survival activities of recombinant S100beta proteins that differ in the content and position of cysteine residues. J. Cell Biol. 109, 3063–3071. Zimmer D. B. and Van Eldik L. J. (1989) Analysis of the calcium-modulated proteins, S100 and calmodulin, and their target proteins during C6 glioma cell differentiation. J. Cell Biol. 108, 141–151. (Accepted 24 April 1997)
PII:
Neuroscience Vol. 82, No. 2, pp. 33–41, 1998 Copyright ? 1997 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306–4522/98 $19.00+0.00 S0306-4522(97)00223-6
BASIC FIBROBLAST GROWTH FACTOR-2 AND INTERLEUKIN-1â REGULATE S100â EXPRESSION IN CULTURED ASTROCYTES D. A. HINKLE,* J. P. HARNEY,† A. CAI,† D. C. HILT,‡ P. J. YAROWSKY§ and P. M. WISE†¶ *Department of Physiology, University of Maryland School of Medicine, Baltimore, MD 21202, U.S.A. †Department of Physiology, University of Kentucky College of Medicine, Lexington, KY 40536, U.S.A. ‡AMGEN, 1840 DeHavilland Drive, Thousand Oaks, CA 91320, U.S.A. §Department of Pharmacology and Experimental Therapeutics, University of Maryland School of Medicine, Baltimore, MD 21202, U.S.A. Abstract––Basic fibroblast growth factor and interleukin-1â are known to regulate the expression of other trophic factors and to stimulate reactive gliosis in vivo. S100â is a glial-specific putative neurotrophic factor and has been considered a marker of the reactive status of astrocytes. Therefore, we tested the hypothesis that basic fibroblast growth factor-2 and interleukin-1â achieve their effects by altering S100â gene expression in cultured rat astrocytes using an RNase protection assay. Short-term treatment with basic fibroblast growth factor-2 produced a transient decrease in S100â messenger RNA which was followed by an increase after longer term treatment. In contrast, both short- and long-term treatment with interleukin-1â suppressed S100â messenger RNA. We measured levels of S100â nuclear primary transcript to assess whether alterations in transcriptional rate explain the changes in messenger RNA. Our results indicate that changes in transcription account for changes in steady state levels of messenger RNA since basic fibroblast growth factor-2-induced changes in S100â primary transcript temporally preceded changes in messenger RNA. We further measured intracellular S100â protein levels by enzyme-linked immunosorbent assay to determine whether changes in gene expression were translated into parallel changes in protein. Our results clearly demonstrate that basic fibroblast growth factor-2 and interleukin-1â influence the expression of the S100â gene, that this regulation appears to occur at the level of transcription, and that alterations in messenger RNA are sometimes, but not always, reflected in changes at the level of protein. These observations suggest that basic fibroblast growth factor-2 may amplify its trophic effects, in part, by influencing the expression of another trophic factor. ? 1997 IBRO. Published by Elsevier Science Ltd. Key words: FGF-2, IL-1â, S100â, astrocyte, growth factor, cell culture.
Basic fibroblast growth factor (FGF-2) and interleukin-1â (IL-1â) regulate neurotrophic factor gene expression in cultured astrocytes.33 FGF-2 is an astrocyte mitogen and morphogen in vitro;10 IL-1â is an astrocyte mitogen in vitro.12 In addition, these growth factors appear to play key roles in the glial response to brain lesion.27 Their levels increase early in the response to brain insult, and in vivo infusion of either into the brain can stimulate reactive changes in astrocytes.4,11–13 Thus, these factors may alter the
post-lesion neurotrophic milieu by regulating astrocyte genes and may stimulate reactive gliosis. S100â is a small, acidic, calcium-binding protein that is found predominantly in astrocytes. The homodimeric form of the protein is a well-characterized neuronal survival and neurite-extension factor in vitro, and can protect hippocampal neurons against hypoglycemia-induced cell death.3,20,34 It also modulates astrocyte proliferation and morphology in vitro, and is over-expressed in reactive astrocytes.15,29,35 Therefore, S100â is widely considered to be a marker of reactive astrocytes, and may be important in the mechanism through which these cells respond to brain injury.26 To assess whether increased S100â expression is a component of the glial response to factors that induce reactive gliosis, we measured S100â mRNA levels in cultured astrocytes from the rat cerebral cortex after treatment for various lengths of time
¶To whom correspondence should be addressed. Abbreviations: ABAM, antibiotic-antimycotic; DMEM, Dulbecco’s modified Eagle’s medium; EDTA, ethylenediaminetetra-acetate; ELISA, enzyme-linked immunosorbent assay; FGF-2, basic fibroblast growth factor-2; GFAP, glial fibrillary acidic protein; IL-1â, interleukin1â; PBS, phosphate-buffered saline; PIPES, piperazineN-N*-bis(ethanesulphonic acid); SDS, sodium dodecyl sulphate. 33
34
D. A. Hinkle et al.
with FGF-2 and/or IL-1â. We quantitated S100â primary transcript levels, an indirect measure of the rate of transcription, to determine the mechanism of the treatment-induced changes in mRNA. Finally, we assessed S100â protein levels to determine whether alterations in gene expression were translated into changes in protein levels.
astrocyte monolayers differed from the above in that cells were scraped into solution D with sodium acetate and phenol before sonification, then precipitated, pelleted, and resuspended in water without LiCl treatment. Agaroseformaldehyde gels and the ratio of A260 to A280 were used to determine the concentration and integrity of the purified RNA. Isolation of separate nuclear and cytoplasmic RNA fractions
EXPERIMENTAL PROCEDURES
Cell culture Enriched cerebral cortex astrocyte cultures were prepared from neonatal (P1) rat pups (Sprague–Dawley; Zivic– Miller) using techniques modified from McCarthy and deVellis.25 Briefly, the cerebral hemispheres from a litter of pups were pooled, dissociated into a cell suspension, then plated for three days at 37)C and 5% CO2 in Dulbecco’s modified Eagle’s medium/F12 (DMEM/F12) (1:1, Sigma) with 10% fetal calf serum (BioWhittaker) and 1#ABAM (antibiotic-antimycotic, Gibco). Media were replaced every three to four days until 10–11 days in vitro, at which time the flasks were shaken overnight to remove contaminating microglia and O-2A cells. The astrocyte monolayers were then sub-cultured into flasks in DMEM/F12 with 10% donor calf serum (BioWhittaker) and ABAM. Treatments with human recombinant FGF-2 (bFGF, Gibco) and/or human recombinant IL-1â (Boehringer Mannheim) were administered every other day in fresh 10% calf serumcontaining medium at varying concentrations for varying intervals as detailed in the Results. The sub-culture plating density was designed to produce approximately 80% confluent monolayers consisting of the same number of cells in each treatment group at the time of sample collection (see Results), since S100â levels change upon confluence.22 The cells in the cultures generated by this procedure were consistently composed of 98% type I astrocytes (Glial fibrillary acidic protein (GFAP)+/A2B5"). Immunocytochemistry Astrocytes were sub-cultured on poly--lysine-coated glass chamber slides (LabTek) for at least 24 h before treatment and staining. Slides were incubated with antiGFAP (1:250, DAKO) or anti-S100â (1:50, East Acres) antibodies to label astrocytes, anti-fibronectin (1:100, Sigma) to label fibroblasts, fluorescein isothiocyanateconjugated Griffonia simplicifolia Isolectin-B4 (25 µg/ml, Sigma) to label microglia,31 and/or Hoechst dye (1 µg/ml, Sigma) to label nuclei, then incubated with the appropriate fluorophore-conjugated secondary antibody (1:100, Jackson). Control stains were performed by omitting the primary antibody, where possible, and partially-enriched cultures of meningeal fibroblasts or shake-supernatant microglia were used to test the anti-fibronectin and the isolectin, respectively. In some cultures, anti-A2B5 (1:200, Boehringer Mannheim) immunostains were performed to detect type II astrocytes. Total RNA purification Total cellular RNA was isolated from rat tissues or astrocyte cultures using a modification of the method of Chomczynski.8 In brief, brains or livers were homogenized in solution D (4 M guanidine thiocyanate, 25 mM sodium citrate, 0.5% sarkosyl, 0.1 M â-mercaptoethanol) with phenol and sodium acetate (pH 5), sonified, chloroform extracted, precipitated in isopropanol, the RNA pelleted and resuspended in 4 M LiCl, re-pelleted, and resuspended in TES (10 mM Tris pH 7.5, 1 mM EDTA, 0.5% sodium dodecyl sulphate; SDS). The RNA was again precipitated, pelleted, washed, and resuspended in diethylpyrocarbonatetreated distilled water. Isolation of total RNA from
Nuclear and cytoplasmic RNA fractions were isolated from cultured astrocytes using the method of Roberts et al. with some modifications.17 Monolayers were scraped into phosphate-buffered saline (PBS) and pelleted, then resuspended in lysis buffer (10 mM Tris–HCl pH 7.4, 1.5 mM MgCl2, 0.2 M sucrose, 0.5% Nonidet P-40, 0.25% sodium deoxycholate) and homogenized without disrupting the nuclei. The lysate was layered over cushion buffer (10 mM Tris–HCl pH 7.4, 1.5 mM MgCl2, 0.4 M sucrose), then centrifuged to separate the cytoplasmic fraction and the nuclei. RNA was then isolated from each fraction. Northern analysis Northern analysis was performed using a modification of previously published methods.1 Briefly, total RNA (10– 40 µg) was fractionated through a 1% agarose/0.22 M formaldehyde/1# MOPS gel, transferred to a Nytran membrane, and attached by ultraviolet cross-linking (FisherBiotech FB-UVXL-1000) and vacuum baking. The membrane was rehydrated, pre-hybridized in hybridization buffer (5#standard saline citrate, 5#Denhardt’s solution, 0.1% SDS, 50% formamide, 50 mM phosphate buffer pH 5, and 100 µg/ml yeast RNA), then incubated for 15–17 h at 42)C in fresh hybridization buffer containing 5#105 c.p.m. of a G-50 column-purified, random primer-generated (Prime-a-Gene, Promega) cDNA probe labelled with [á-32P]dCTP. The rat S100â cDNA represents a 302 bp BamHI-HindIII fragment of the 3*-untranslated region (pS24), the genomic clone represents a 0.6 kb BamHIHindIII fragment of the intron 2/exon III boundary (pS1rg5D, sub-clone of pS100-1-5 from Dr Noboru Sueoka, University of Colorado). The membranes were stringently washed, dried, and exposed to film. Solution hybridization/RNase protection assay This assay was performed using a modification of previously published methods.1 Briefly, samples of total or cytoplasmic RNA (5 µg), or the entire nuclear fraction, were dried and resuspended in 30 µl of hybridization solution (80% v/v formamide, 40 mM PIPES, 0.4 M NaCl, 1 mM EDTA). To this was added approximately 1 ng of heatdenatured antisense cRNA probes (S100â cDNA or genomic and 1B15 cDNA representing a 111 bp PstI-BamHI fragment of the cyclophilin open reading frame, from Dr J. L. Roberts, Mt Sinai School of Medicine) which were labelled with [á-32P]UTP by in vitro transcription. A sample of each probe was assessed for percent incorporation, then purified on a G-50 RNA spin column (Boehringer Mannheim) before determination of specific activity. The purified probes and RNA were allowed to hybridize in solution at 45)C for 12–14 h, then incubated with RNase A (20 µg/ml) and RNase T1 (2 µg/ml) for 1 h at room temperature, the digestion terminated with SDS (0.5%) and proteinase K (120 ng/µl), then the RNA hybrids extracted with Tris-buffered phenol/chloroform, precipitated in ethanol, resuspended in denaturing RNA loading buffer (95% v/v formamide, 20 mM EDTA, 0.05% w/v Bromophenol Blue, 0.05% w/v xylene cyanol) and fractionated on a 4% acrylamide/7 M urea gel. The gel was fixed in 10% glacial
Regulation of S100â expression
35
acetic acid/10% methanol, transferred to 3MM paper, then dried and read in a PhosphorImager (Molecular Dynamics). Enzyme-linked immunosorbent assay S100â protein was quantitated in the cell lysates of cortical astrocytes, as detailed in Griffin et al. with minor modifications.14 To obtain cell lysates, monolayers were scraped into PBS, pelleted, and resuspended in PBS. The cells were then sonicated, the cellular debris pelleted, and the supernatant taken for analysis. Bovine brain S100 was used as the standard (East Acres); its S100â concentration was determined using VUSB-1, a well-characterized recombinant form of S100â protein.32 The Bradford protein assay (Bio-Rad) was used to determine the total protein concentration of each sample, which was used to normalize the S100â. Experimental design and statistical analysis Each replicate culture consisted of cells derived from a single litter of rat pups that were treated with FGF-2, IL-1â, both factors (unless otherwise stated), and a corresponding control. The total number of independent replicates (‘‘n’’) for each study is reported in each figure legend. To assess differences in S100â expression between cultures which were treated with FGF-2 and/or IL-1â, the data were analysed by least squares analysis of variance using a General Linear Models Procedure of the Statistical Analysis System (SAS, 1989). Cyclophilin mRNA or total protein were included in the analysis as covariates for the RNase protection and enzyme-linked immunosorbent assay (ELISA) data, respectively, to eliminate variability due to unequal loading of RNA or protein. Significant effects for treatment at any time point were further analysed using independent orthogonal contrasts to detect differences between FGF-2 and control, IL-1â and control, or combination treatment and FGF-2+IL-1â (this is a test for synergy that compares combination treatment to the additive effect of each factor alone at each time point). To evaluate changes in primary transcript, in which only single factor treatments were used, the data were analysed by paired t-tests. The criteria for statistical significance were set at P<0.05. RESULTS
Astrocyte morphology and culture characterization Treatments with FGF-2 (10 ng/ml) induced a robust stellation response which was first detectable after 24 h of treatment and became most highlydifferentiated after five to six days of treatment (data not shown). IL-1â (10 U/ml), in contrast, did not alter the flat, polygonal morphology that was seen in the controls at any time during the seven to eight day treatment period (data not shown). Fig. 1 shows examples of the morphological response of the astrocytes to 48 h of each treatment, a time point at which typical morphological changes were consistently visible. Astrocytes cultured in calf serum-containing media for 48 h or seven days with FGF-2 or IL-1â, or both, were sub-confluent and did not differ in cell number from controls as assessed by cell counts using the Hoechst nuclear dye (data not shown). Under these treatment conditions, the cultures were consistently composed of greater than 98% GFAP or S100â immunoreactive astrocytes, and not more than 0.5% fibroblasts and 0.1% microglia, as determined by counting the proportion of cells immunopositive for
Fig. 1. Comparison of cultured cortical astrocyte morphology by phase-contrast (A,C,E,G) and GFAP immunocytochemistry (B,D,F,H). The cells have been treated for 48 h with control (A,B), IL-1â at 10 U/ml (C,D), FGF-2 at 10 ng/ml (E,F), or both factors (G,H). FGF-2 stimulates morphological stellation, whereas IL-1â does not.
GFAP, S100â, fibronectin, or lectin GSA I-B4, respectively, relative to the total number of Hoechstlabelled cells (data not shown). S100â northern analysis Northern analysis validated the use of pS24 as a probe for S100â mRNA in the RNase protection assay (Fig. 2A). The pS24 (S100â) cDNA recognized a single 1.6 kb transcript in samples of total RNA from rat astrocyte cultures, rat C6 glioma cultures (an immortalized astrocyte cell line), and rat brain, but not rat liver. This transcript is the published size of rat S100â mRNA,21 and was also recognized by the rat S100â genomic clone pS1rg5D (data not shown). The sensitivity of this method was not sufficient to detect heteronuclear S100â RNA. Effects of basic fibroblast growth factor-2 and/or interleukin-1â on the levels of S100â messenger RNA A representative autoradiograph illustrates the changes seen in S100â mRNA after treatment for 24 h, 48 h, or seven days with FGF-2, IL-1â, both factors, or no treatment control (Fig. 2B). The S100â
36
D. A. Hinkle et al.
Fig. 2. A) Northern analysis of rat RNA using S100â cDNA probe (pS24). Total RNA from rat brain (Br, 10 µg), rat cortical astrocyte culture (As, 30 µg), rat C6 glioma cell culture (C6, 30 µg), and rat liver (Lv, 30 µg) was fractionated on an agarose/formaldehyde gel, transferred to a Nytran membrane, and incubated with a 32P-labelled cDNA probe. The transcript recognized is approximately 1.6 kb. B) Representative RNase protection assay autoradiograph from cultured cortical astrocytes treated for the times shown with control (C), FGF-2 at 10 ng/ml (F), IL-1â at 10 U/ml (I), or both factors (IF). 32 P-labelled cRNA probes (P) for S100â (369 nt) and cyclophilin (174 nt) were incubated with 5 µg samples of total RNA or tRNA (t), digested with RNases, then the protected fragments were fractionated on a 4% acrylamide/7 M urea gel. The protected fragments generated were 302 nt and 111 nt in length for S100â and cyclophilin, respectively.
cRNA probe hybridized with a single mRNA species and protected the expected 302 nucleotide fragment. Similarly, the cyclophilin cRNA probe produced the expected 111 nucleotide fragment. The specificity of each probe was further proven by its lack of hybridization with tRNA. In all assays, cyclophilin was used as an internal control since its levels are not regulated by FGF-2 or IL-1â in cultured cortical astrocytes.33 Each treatment paradigm (FGF-2, IL-1â, and FGF-2/IL-1â) had a distinct influence on S100â gene expression (Figs 2B, 3). FGF-2 (10 ng/ml) produced a biphasic effect on S100â gene expression: mRNA levels decreased after 12 h and 24 h of treatment to 86% (P<0.05) and 30% (P<0.001) of the control level, respectively, then were elevated 2.5-fold (P<0.01) after seven days of continuous FGF-2 treatment. IL-1â (10 U/ml) produced a similar decrease in mRNA levels, but with a markedly different pattern. After 48 h of treatment with the cytokine, S100â gene expression was inhibited to 40% of control levels (P<0.02), and continued to be suppressed after seven days of treatment (38% of control, P<0.03). Combination treatment with both FGF-2 and IL-1â produced a rapid, statistically greater suppression of S100â gene expression than the additive effect of each factor alone at 24 h and 48 h. After 24 h of treatment, S100â mRNA levels reached 23% of control (P<0.02, simultaneous treatment with both factors versus the additive effect of each singlefactor treatment), while after 48 h S100â was expressed at only 12% of the control level (P<0.05). After seven days of combination treatment, there was no difference between it and the additive effect of
Fig. 3. S100â gene expression in rat cortical astrocytes in response to treatment with FGF-2 (10 ng/ml) and/or IL-1â (10 U/ml), as measured in total RNA by the RNase protection assay. At each time point and for each treatment, the mean&S.E.M. relative S100â mRNA level (normalized to cyclophilin) is plotted relative to its control (designated as 1.0). Treatment effects are statistically significant by least squares ANOVA with orthogonal contrasts at P<0.05 as follows: FGF-2 or IL-1â versus control (*), combined treatment versus sum of single-factor treatment effects (**), n=3–6.
each factor alone. No changes in S100â mRNA levels were detected under any treatment conditions at 1 or 3 h (data not shown). Similar effects of FGF-2 and IL-1â on the patterns of S100â gene expression were seen in cultures from the hippocampus and hypothalamus (data not shown). Dose–response curves were also generated in cortical astrocyte cultures for each factor at the times
Regulation of S100â expression
37
Fig. 4. Characterization of protected fragments generated by 32P-labelled genomic S100â (S), cyclophilin (C), or both (C+S) cRNA probes in nuclear or cytoplasmic RNA from cultured cortical astrocytes. The probes were incubated with whole nuclear RNA fractions, 5 µg of cytoplasmic RNA, or tRNA (t), digested with RNases, and the protected fragments fractionated on a 4% acrylamide/7 M urea gel. The protected fragments are labelled on the left side of panel A, the undigested S100â (SP) and cyclophilin (CP) probes are shown, and panel B is an over-exposure of the nuclear RNA that better demonstrates the protected primary transcript. Experiments utilized freshly-made probe (less than one-day-old) to decrease the amount of probe breakdown product and to maximize the signal of the primary transcript band, and the protected fragments were fractionated further on the gel than is shown here to prevent signal bands from overlapping.
that they maximally inhibited S100â gene expression in the above experiments. FGF-2 treatment for 24 h produced maximal suppression of the S100â gene at doses of 1–10 ng/ml. 48 h treatment with IL-1â at 1–10 U/ml produced maximal gene suppression (data not shown). Effects of basic fibroblast growth factor-2 or interleukin-1â on the levels of S100â primary transcript S100â primary transcript levels were quantitated as an indirect measure of the rate of transcription. Approximately 0.15% of the total S100â RNA is in the form of nuclear primary transcript, 1.8% is nuclear mRNA, and the rest is cytoplasmic mRNA (data not shown). The pS1rg5D genomic clone encodes 0.6 kb of S100â sequence at the intron 2/exon III boundary, and produces a protected fragment of this size when it hybridizes with primary transcript (which has intron and exon sequence), or of 272 nucleotides when it hybridizes with mature transcript (mRNA, which has only exon sequence, see Fig. 4). FGF-2 treatment produced a biphasic effect on S100â nuclear primary transcript: it decreased its levels to 72% (P<0.04) and 45% (P<0.04) of controls after 6 h and 12 h of treatment, respectively, then induced an approximately three-fold increase in primary transcript (P<0.04) after 48 h of treatment
(Fig. 5A). Nuclear mRNA decreased to 47% (P<0.02) of the control level after 24 h of treatment. In the cytoplasm, mRNA levels decreased to 79% (P<0.05), 37% (P<0.04), and 69% (P<0.03) of the control level after 12 h, 24 h, and 48 h of treatment, respectively. Thus, the FGF-induced changes in nuclear primary transcript preceded those in nuclear mRNA, which in turn preceded those in cytoplasmic mRNA. In addition, the steep rise in primary transcript at 48 h preceded the elevated expression of mRNA after seven days of treatment reported above. No significant changes were seen after 1 h or 3 h of treatment. Treatment with IL-1â (10 U/ml) for 48 h decreased S100â nuclear primary transcript, nuclear mRNA, and cytoplasmic mRNA levels to 45% (P=0.058), 47% (P=0.056), and 55% (P<0.04) of control levels, respectively (Fig. 5B). The cytoplasmic mRNA decreased to 72% (P<0.05) of control levels after 24 h of IL-1â treatment despite the lack of a significant change in the primary transcript. No significant changes were evident at 3 h, 6 h, or 12 h. Effects of basic fibroblast growth factor-2 and/or interleukin-1â on the levels of S100â protein S100â protein levels from cell lysates were quantitated by an ELISA. The first significant changes were detected after five days and eight days of exposure to FGF-2, at which times S100â protein was elevated
38
D. A. Hinkle et al.
Fig. 6. S100â protein level in cortical astrocytes in response to treatment with FGF-2 (10 ng/ml) and/or IL-1â (10 U/ml). S100â was quantitated by an ELISA, and total protein by a Bradford assay. At each time point and for each treatment, the mean&S.E.M. relative S100â protein level (normalized to total protein) is plotted relative to its control (designated as 1.0). Treatment effects are statistically significant by least squares ANOVA with orthogonal contrasts at P<0.05 for treatment versus control (*), n=4.
Fig. 5. A) S100â nuclear primary transcript, nuclear mRNA, or cytoplasmic mRNA level in cortical astrocytes in response to FGF-2 (10 ng/ml) as measured by the RNase protection assay. At each time point, the mean&S.E.M. relative S100â primary transcript RNA level (normalized to cyclophilin) is plotted relative to its control (designated as 1.0). Treatment effects are statistically significant by paired t-tests at P<0.05 (*), n=3–6. B) S100â nuclear primary transcript, nuclear mRNA, or cytoplasmic mRNA level in cortical astrocytes in response to IL-1â (10 U/ml) as measured by the RNase protection assay. At each time point, the mean&S.E.M. relative S100â primary transcript RNA level (normalized to cyclophilin) is plotted relative to its control (designated as 1.0). Treatment effects are statistically significant by paired t-tests at P<0.05 (*), n=3–6.
3.9-fold (P<0.02) and 10.1-fold (P<0.002), respectively, relative to their controls (Fig. 6). No other treatments produced significant changes in S100â protein concentrations. DISCUSSION
The major goals of this work were to (i) assess whether the trophic effects of FGF-2 and/or IL-1â may, in part, be achieved by influencing the expression of another growth factor, S100â and (ii) deter-
mine whether these trophic factors may act through S100â to initiate changes in astrocytes that are similar to those seen in reactive gliosis. Our data clearly demonstrate that FGF-2 and IL-1â regulate S100â gene expression in cultured rat cortical astrocytes. Further, they support the conclusion that the regulation of this gene occurs, at least in part, at the level of transcription, and that the resulting alterations in mRNA are translated into changes at the level of protein under some circumstances. They suggest that growth factors may reinforce or attenuate their effects by altering the expression of other growth factors. Finally, they suggest that elevated levels of S100â may support or enhance morphological differentiation in vitro, in the long term, that is similar to that seen in reactive gliosis in vivo. However, no clear conclusions can yet be drawn about the involvement of S100â in the mechanisms through which these factors may induce reactive gliosis, since (i) the initiation of morphological changes by FGF-2 occurs in the presence of diminished S100â expression, (ii) IL-1â stimulated no morphological changes, and (iii) it remains unclear how closely our cultures model reactive change in astrocytes in vivo. The appearance of reactive astrocytes after brain lesion is a highly characteristic phenomenon. Although it is known that tissue damage ‘‘stimulates’’ reactive gliosis, the exact mechanisms that underlie its initiation are unknown, as are the consequences of this change to intercellular interactions. Both FGF-2 and IL-1â can increase the number of GFAPimmunoreactive astrocytes after direct infusion into the brain4,11,12,13 and are therefore considered stimulators of reactive gliosis. GFAP and S100â are overexpressed in reactive astrocytes under many lesion conditions. Therefore, it is possible that FGF-2 and/or IL-1â may act through the S100â gene to
Regulation of S100â expression
initiate the reactive phenotype, and/or that the altered expression of this putative neurotrophic factor may influence neuron–glia interactions. To begin to test these ideas, we hypothesized that FGF-2 and/ or IL-1â may influence S100â gene expression in cultured astrocytes. FGF-2 stimulated astrocytes to become stellate, whereas IL-1â did not induce any morphological change. Throughout the study, our major concern was to maintain equal cell density in all treatments at the time of sample collection since S100â expression is regulated by the level of culture confluence.22 We did not detect any effect of either factor on the number of cells after 48 h or seven days of treatment, compared to controls. The reported proliferative effects of these factors are often based on measurements of [3H]thymidine incorporation into astrocyte DNA, and are often assessed in low-serum or serumfree media.2,12 Our data are based solely on counts of cells grown in 10% calf serum, which may have stimulated maximum proliferation and thus obscured any treatment effects. FGF-2 and IL-1â induced different effects on cortical astrocyte S100â mRNA. FGF-2 induced a distinct biphasic effect in gene expression: suppression in response to short-term treatment and stimulation after longer-term treatment. In contrast, the effect of IL-1â was only inhibitory. Combination treatment, using each factor at its maximally effective dose, resulted in a synergistic effect. Together, these findings imply that the regulatory mechanisms employed by each factor in astrocytes are different. Our findings in cortical astrocytes strongly suggest that FGF-induced alterations in S100â mRNA levels were due to changes in the rate of gene transcription. Numerous investigators have demonstrated that changes in primary transcript reflect changes in transcription since the half-life of heteronuclear RNA is extremely short.16,23,24 We therefore quantitated the level of primary transcript in the nucleus as an indirect indicator of gene transcription. S100â nuclear primary transcript was significantly inhibited by FGF-2 at least 12 h before significant decreases were detected in the nuclear and cytoplasmic mRNA, and increased significantly after longer-term treatment. The pattern of the alterations in primary transcript preceded and paralleled those seen in mRNA in response to both short- and longer-term treatment. Further, the magnitudes of the relative decrease and increase in primary transcript were similar to that of cytoplasmic mRNA. These findings clearly implicate transcriptional control as a mechanism through which FGF-2 regulates S100â. Whether or not the IL-1â-induced effects on S100â mRNA levels are explained by regulation of transcription is less clear, since changes in primary transcript did not precede those in mRNA. It is possible that additional mechanisms are involved, such as the regulation of mRNA stability.
39
The FGF-induced elevation in S100â mRNA was translated into an increase in intracellular S100â protein. Longer-term FGF-2 treatment significantly elevated S100â protein levels, relative to controls, which may have been due to the increased mRNA levels seen at this time, and/or to a decreased rate of protein turnover. The biological repercussions of this elevation of S100â within the astrocytes are unknown, but may include the stimulation of morphological differentiation, glial filament turnover, protein phosphorylation, and/or calcium signal transduction.5,6,19 S100â is a positive regulator of morphological stellation in glial cells, and we observed that the relative elevation in protein occurred at approximately the same time that our cultures reached their greatest level of morphological differentiation.35 Thus, elevated levels of S100â induced by longerterm treatment with FGF-2 appear to have supported, or even enhanced, the morphological differentiation that we observed in the astrocyte cultures. Whether this resulted from the action of S100â as an intracellular or extracellular modulator, which has been described in other systems, was not evaluated.28,30 These data further link S100â with morphological changes in cultured astrocytes which may be similar to changes seen in reactive astrocytes in vivo. However, we did not detect an inhibition of S100â protein in response to short-term FGF-2 treatment despite the marked suppression of mRNA. This lack of a change in protein may be due to the transient nature of the decrease in mRNA in the face of an abundant and stable protein pool. Van Eldik et al. have demonstrated in cultured glial cells that there is a delay of 24–48 h before changes in S100â mRNA are reflected in changes in protein.29 These data suggest that mechanisms other than S100â are important in the initiation of morphological differentiation by FGF-2, since decreased gene expression and unchanged protein levels did not appear to hinder the development of morphological stellation. Therefore, it remains difficult to conclude what role S100â plays in the early stages of the response of astrocytes to FGF-2, and no clear hypothesis can yet be generated relevant to its potential role in FGF-induced reactive gliosis. IL-1â treatment produced a trend towards a decrease in S100â protein that did not reach statistical significance, but was able to suppress the stimulatory FGF-2 effect since combination treatment produced an increasing trend in S100â protein that was not statistically significant. The inability of the IL-1âinduced changes to reach statistical significance suggests that S100â protein does not decrease in response to this treatment, but may reflect the limits of the sensitivity of the ELISA, particularly when compared to the RNase protection assay, to detect decreases in S100â below a low baseline. It is also possible that more complex mechanisms exist for the regulation of S100â protein at the level of translation or turnover. These data, combined with the lack of
40
D. A. Hinkle et al.
morphological changes in response to IL-1â, make it difficult to hypothesize what role, if any, S100â plays in IL-1â-induced reactive gliosis in vivo. CONCLUSIONS
In summary, we have demonstrated that both FGF-2 and IL-1â regulate S100â expression in cultured rat astrocytes. Our findings lend further credence to the concept that trophic factors often act by enhancing or suppressing other trophic factors.7,9,18 They also emphasize that elevated S100â, at least in
the long term, may contribute to changes in cultured astrocytes that may be similar to those seen in reactive gliosis. However, our data do not yet allow conclusions to be drawn relevant to the role of S100â in the mechanisms of FGF-2- or IL-1â-induced reactive change in astrocytes. Acknowledgements—We wish to thank Katherine Rosewell and Suzanne Steman for their superb technical assistance. This work was supported by NIH AG02224 to PMW. David Hinkle was a predoctoral trainee of NIH Training Grant HD07170.
REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.
Ausubel F. M., Brent R., Kingston R. E., Moore D. D., Seidman J. G., Smith J. A. and Struhl K. (1994) Current Protocols in Molecular Biology. John Wiley. Avola R., Ragusa N., Reale S., Costa A., Insirello L. and Stella A. M. G. (1993) Effect of growth factors on macromolecular synthesis in primary rat astroglial cell cultures. Annals NY Acad. Sci. 692, 192–200. Barger S. W., VanEldik L. J. and Mattson M. P. (1995) S100â protects hippocampal neurons from damage induced by glucose deprivation. Brain Res. 677, 167–170. Barotte C., Eclancher F., Ebel A., Labourdette G., Sensenbrenner M. and Will B. (1989) Effects of basic fibroblast growth factor (bFGF) on choline acetyltransferase activity and astroglial reaction in adult rats after partial fimbria transection. Neurosci. Lett. 101, 197–202. Baudier J., Delphin C., Grunwald D., Khochbin S. and Lawrence J. J. (1992) Characterization of the tumor suppressor protein p53 as a protein kinase C substrate and a S100b-binding protein. Proc. natn. Acad. Sci. U.S.A. 89, 11627–11631. Bianchi R., Giambanco I. and Donato R. (1993) S-100 protein, but not calmodulin, binds to the glial fibrillary acidic protein and inhibits its polymerization in a Ca2+-dependent manner. J. biol. Chem. 268, 12669–12674. Carroll P., Sendtner M., Meyer M. and Thoenen H. (1993) Rat ciliary neurotrophic factor (CNTF): gene structure and regulation of mRNA levels in glial cell cultures. Glia 9, 176–187. Chomczynski P. and Sacchi N. (1987) Single-step methods of RNA isolation by acid guanidinium thiocyanate-phenolchloroform extraction. Analyt. Biochem. 162, 156–159. DeLeon M., Van Eldik L. J. and Shooter E. M. (1991) Differential regulation of S100B and mRNAs coding for S100-like proteins (42A and 42C) during development and after lesion of rat sciatic nerve. J. Neurosci. Res. 29, 155–162. Engele J. and Bohn M. C. (1991) The neurotrophic effects of fibroblast growth factors on dopaminergic neurons in vitro are mediated by mesencephalic glia. J. Neurosci. 11, 3070–3078. Finklestein S. P., Apostolides P. J., Caday C. G., Prosser J., Philips M. F. and Klagsbrun M. (1988) Increased basic fibroblast growth factor (bFGF) immunoreactivity at the site of focal brain wounds. Brain Res. 460, 253–259. Giulian D. and Lachman L. B. (1985) Interleukin-1 stimulation of astroglial proliferation after brain injury. Science 228, 497–499. Giulian D., Woodward J., Young D. G., Krebs J. F. and Lachman L. B. (1988) Interleukin-1 injected into mammalian brain stimulates astrogliosis and neovascularization. J. Neurosci. 8, 2485–2490. Griffin W. S. P., Stanley L. C., Yeralan O., Ronvnaghi C. R. and Marshak D. R. (1993) Methods for the study of cytokines in human neurodegenerative disease. Meth. Neurosci. 17, 268–287. Griffin W. S. T., Stanley L. C., Ling C., White L., MacLeod V., Perrot L. J., White C. L. III and Araoz C. (1989) Brain interleukin 1 and S-100 immunoreactivity are elevated in Down syndrome and Alzheimer disease. Proc. natn. Acad. Sci. U.S.A. 86, 7611–7615. Herman J. P., Schafer M. K.-H., Watson S. J. and Sherman T. G. (1991) In situ hybridization analysis of arginine vasopressin gene transcription using intron-specific probes. Molec. Endocr. 5, 1447–1456. Jakubowski M. and Roberts J. L. (1991) Multiplex solution hybridization-ribonuclease protection assay for quantitation of different ribonucleic acid transcripts from snap-frozen neuroendocrine tissues of individual animals. J. Neuroendocr. 4, 79–89. Johnson-Green P. C., Dow K. E. and Riopelle R. J. (1992) Neurite growth modulation associated with astrocyte proteoglycans: influence of activators of inflammation. Glia 5, 33–42. Kligman D. and Hilt D. C. (1988) The S100 protein family. Trends biochem. Sci. 13, 437–442. Kligman D. and Marshak D. R. (1985) Purification and characterization of a neurite extension factor from bovine brain. Proc. natn. Acad. Sci. U.S.A. 82, 7136–7139. Kuwano R., Usui H., Maeda T., Fukui T., Yamanari N., Ohtsuka E., Ikehara M. and Takahashi Y. (1984) Molecular cloning and the complete nucleotide sequence of cDNA to mRNA for S-100 protein of rat brain. Nucl. Acids Res. 12, 7455–7465. Labourdette G. and Marks A. (1975) Synthesis of S-100 protein in monolayer cultures of rat-glial cells. Eur. J. Biochem. 58, 73–79. Lorang D., Lundblad J. R., Blum M. and Roberts J. L. (1991) Hormonal regulation of proopiomelanocortin transcription and RNA processing in AtT20 cells. Endocrinology 128, 509–540. Maurer R. A. (1982) Estradiol regulates the transcription of the prolactin gene. J. biol. Chem. 257, 2133–2136. McCarthy K. D. and DeVellis J. (1980) Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue. J. Cell Biol. 85, 890–902.
Regulation of S100â expression 26. 27. 28. 29. 30. 31. 32. 33. 34. 35.
41
Miyake T., Hattori T., Fukuda M. and Kitamura T. (1989) Reactions of S-100-positive glia after injury of mouse cerebral cortex. Brain Res. 489, 31–40. Neary J. T., Baker L., Jorgensen S. L. and Norenberg M. D. (1994) Extracellular ATP induces stellation and increases glial fibrillary acidic protein content and DNA synthesis in primary astrocyte cultures. Acta neuropath. 87, 8–13. Selinfreund R. H., Barger S. W., Pledger W. J. and Van Eldik L. J. (1991) Neurotrophic protein S100beta stimulates glial cell proliferation. Proc. natn. Acad. Sci. U.S.A. 88, 3554–3558. Selinfreund R. H., Barger S. W., Welsh M. J. and Van Eldik L. J. (1990) Antisense inhibition of glial S100beta production results in alterations in cell morphology, cytoskeletal organization, and cell proliferation. J. Cell Biol. 111, 2021–2028. Shashoua V. E., Hesse G. W. and Moore B. W. (1984) Proteins of the brain extracellular fluid: evidence for release of S-100 protein. J. Neurochem. 42, 1536–1541. Streit W. J. (1990) An improved staining method for rat microglial cells using the lectin from Griffonia simplicifolia (GSA I-B4). J. Histochem. Cytochem. 38, 1683–1686. Van Eldik L. J., Staecker J. L. and Winningham-Major F. (1988) Synthesis and expression of a gene coding for the calcium-modulated protein S100â and designed for cassette-based, site-directed mutagenesis. J. biol. Chem. 263, 7830–7837. Vige X., Costa E. and Wise B. C. (1991) Mechanism of nerve growth factor mRNA regulation by interleukin-1 and basic fibroblast growth factor in primary cultures of rat astrocytes. Molec. Pharmac. 40, 186–192. Winningham-Major F., Staecker J. L., Barger S. W., Coats S. and Van Eldik L. J. (1989) Neurite extension and neuronal survival activities of recombinant S100beta proteins that differ in the content and position of cysteine residues. J. Cell Biol. 109, 3063–3071. Zimmer D. B. and Van Eldik L. J. (1989) Analysis of the calcium-modulated proteins, S100 and calmodulin, and their target proteins during C6 glioma cell differentiation. J. Cell Biol. 108, 141–151. (Accepted 24 April 1997)