Lithium increases melanin-concentrating hormone mRNA stability and inhibits tyrosine hydroxylase gene expression in PC12 cells

Molecular Brain Research 52 Ž1997. 270–283

Research report

Lithium increases melanin-concentrating hormone mRNA stability and inhibits tyrosine hydroxylase gene expression in PC12 cells Franc¸oise Presse, Bruno Cardona, Laetitia Borsu, Jean-Louis Nahon

)

Institut de Pharmacologie Moleculaire et Cellulaire, UPR 411 CNRS, UniÕersite´ de Nice Sophia-Antipolis, 660 route des Lucioles, 06560 Valbonne, ´ France Accepted 12 August 1997

Abstract Melanin-concentrating hormone ŽMCH. is a cyclic peptide involved in the regulation of food-intake behaviour and stress response in mammals. Expression of the MCH gene predominates in hypothalamic neurons. Mechanisms governing the regulation of expression of MCH gene in established cell lines were not explored yet. Here, we analysed the actions of nerve growth factor ŽNGF., dexamethasone, forskolin and lithium on MCH mRNA levels in the PC12 pheochromocytoma cell line. We compared them with those observed on tyrosine hydroxylase ŽTH. mRNA, constitutively expressed in PC12 cells, and neurotensin ŽNT. mRNA, taken as a control. In untreated cells, MCH RNA species of high molecular weight were found. Exposure of cells at a combination of NGF and lithium resulted in decreased expression of these MCH RNAs and in the transient production of mature MCH mRNA. Strikingly, after short exposure of PC12 cells to NGF, lithium per se elicited a marked increase in MCH mRNA levels whilst it exerted a potent inhibitory action on TH mRNA expression. Detailed investigations revealed that lithium enhanced MCH mRNA expression through post-transcriptional mechanisms whereas it regulated TH gene expression mainly at the level of transcription. These results demonstrate that lithium, an agent widely used for treatment of manic depressive illness, can exert an opposite effect on MCH and TH mRNA production in PC12 cells. The MCH gene system in NGF-treated PC12 cells provides a good opportunity for studying the effect of lithium on gene expression at post-transcriptional levels in a neuron-like cellular model. q 1997 Elsevier Science B.V. Keywords: Melanin-concentrating hormone; Tyrosine hydroxylase; Neurotensin; mRNA stability; Gene transcription; Lithium; Glucocorticoids; PC12 cells

1. Introduction Mammalian melanin-concentrating hormone ŽMCH. and the additional pro-MCH derived peptides, namely NEI and NGE, constitute a new family of peptides that are expressed predominantly in the lateral hypothalamus area ŽLHA. and zona incerta ŽZI. w34,40,54x. The MCH axonal network innervates numerous brain areas and the neuronal lobe of the pituitary gland in the rat w4,45x. MCH plays a role as neurotransmitterrneuromodulator in a broad array of neuronal functions Žreviewed in Refs. w3,32x.. Intracerebroventricular ŽICV. injection of MCH in the rat brain modulates secretion of ACTH under resting or stress-conditions possibly through activation of the CRF-synthesizing neurons w22x and depending on circadian control w5x. In

) Corresponding author. Fax: q33 Ž4. 9395-7708; E-mail: [email protected]

0169-328Xr97r$17.00 q 1997 Elsevier Science B.V. All rights reserved. PII S 0 1 6 9 - 3 2 8 X Ž 9 7 . 0 0 2 7 3 - 8

addition, MCH neuronal system is involved in the regulation of food-intake behaviour w38,39,42x. A single MCH-encoding gene has been identified in both rat and mouse genomes w6,46x. Studies regarding the control of MCH biosynthesis have focused essentially on in vivo models w6,36–39x. Cellular models have been established using fetal hypothalamic neurons grown as dissociated primary cells in serum-free medium w10x or using neonatal rat hypothalamic cells w35x. The effect of different agents were examined using these hypothalamic neuronal systems. Indeed, a neurite promoting effect on MCH pericarya was observed after a treatment with nerve growth factor ŽNGF. or in co-cultured models w10x. Furthermore, synthesis and secretion of MCH were enhanced following incubation with dexamethasone or cAMP analogues w35x. However, analysis of MCH gene expression requires a reliable and easily inducible cellular system. In preliminary studies, large screening of immortalized neuronal or endocrine cell types revealed that pheochromocytoma PC12 cells may express MCH mRNA under treat-

F. Presse et al.r Molecular Brain Research 52 (1997) 270–283

ment with combinations of effectors including the NGF, adenylate cyclase activator Žforskolin., dexamethasone and lithium. While not requiring NGF for outright survival, exposure to NGF results in the conversion of the PC12 cells from immature adrenosympathetic precursor-like cells to mature sympathetic-like neurons w18x. PC12 cells can express a wide variety of neuronal transmitters and neuropeptides, under resting andror inducible conditions. In this respect, much attention has been focused on the gene regulation of tyrosine hydroxylase ŽTH., the rate limiting enzyme in catecholamine biosynthesis Žreviewed in Ref. w27x. and neurotensin ŽNT., a neuropeptide co-expressed with dopamine in vivo and in vitro Žreviewed in Refs. w13,14x.. In this paper, we present a detailed analysis of MCH gene expression in PC12 cells treated with multiple combinations of NGF, dexamethasone, forskolin and lithium, and we compare this to the expression of NT and TH genes under the same conditions. We find that lithium markedly potentiates NGF permissive action on MCH mRNA expression mainly through post-transcriptional mechanisms whereas it inhibits TH gene activity at the transcriptional level. Therefore, the PC12 cell system represents a model of value with regards to MCH gene expression in an established neuron-like cell line.

2. Materials and methods 2.1. Cell culture PC12 cells w18x were grown in RPMI 1640 medium supplemented with 10% heat-inactivated horse serum, 5% fetal calf serum and 50 mgrml of gentamicin as described elsewhere w43x. The cells were maintained in a humidified atmosphere at 378C and 5% CO 2 . Cells were plated in 100 mm dishes and used at about 10 7 cellsrdish on the day of experiments. For the complete cocktail treatment, inducers were added to the RPMI medium at the final concentration of 100 ngrml NGF ŽTebu, le Perray-en-Yvelines, France. or 50 ngrml NGF Žlaboratory made preparation., 1 mM dexamethasone, 1 mM forskolin and 20 mM LiCl. For single inducer treatments, cells were treated for various times and concentrations as indicated. At the end of the treatments, cells were washed twice with 1 = PBS and directly used for RNA extraction or frozen at y208C. For the message stability studies, cells were treated with 100 mM 5,6-dichloro-1-b-ribofuranosylbenzimidazole ŽDRB; Calbiochem, La Jolla, CA. for different times. This concentration of DRB was sufficient for blocking transcriptional activity of neuropeptide-encoding genes in PC12 cells w29x. In preliminary experiments, the effect of lithium concentration on MCH and TH mRNA levels was investigated on PC12 cells in presence of 50 ngrml NGF Žnot shown.. A significant rise in MCH mRNA levels was noted at 5mM lithium, maximum increase was observed at

271

20 mM LiCl and a 3-fold reduction was found with 60 mM LiCl. A dose dependent decrease in TH mRNA levels, with full inhibition at 60 mM LiCl, was observed. Furthermore, a time course studies revealed that MCH mRNA was detected as early as 3 h after NGF and lithium treatment, reached a maximum at 12–24 h and disappeared after 48 h Žnot shown.. 2.2. Isolation of RNA and northern blot analysis Total RNA was extracted by a modification of the phenol–guanidium isothiocyanate method w9x as described previously w36x. RNA was quantified by absorbance at 260 nm Žthe ratio of absorbance at 260 and 280 nm was close to 2.0 as an indication of RNA purity.. Total RNA was denaturated in 50% formamide, 15% formaldehyde, 5% ficoll solution at 608C for 5 min, then ethidium bromide was added 200 mgrml Žfinal concentration. and the sample was electrophoresed in 1.2% agarose-gel containing 2.2 M formaldehyde. RNA was transferred to hybond ŽN. nylon membrane ŽAmersham; UK. and UV-crosslinked with a Stratagene apparatus. The RNA blot was incubated at 428C for 5 h in prehybridization buffer containing 50% Žvrv. formamide as previously described w36x. Hybridization was performed overnight in the same buffer with 32 P-labelled cDNA Žsee Section 2.3.. Washing was performed successively twice in 5 = SSPE Ž1 = SSPE is 0.18 M NaCl, 0.01 M NaH 2 PO4 ., ŽpH 7.4.r0.1% SDS and 2 = SSPEr0.1% SDS at 50–608C. Filters were visualized by exposing them against X-OMATAR film at y708C with intensifying screens for 3–10 days. The level of expression of specific mRNA was quantified by densitometric scanning of the X-ray film in a Hagfa arcus scanner and using a computerized image analysis screen ŽNIH Image, version 1.51.. The ratio of MCH, NT or TH: b-actin, GAPDH or 18S rRNA hybridization was calculated and we compared the values of the ratio for each treated PC12 cell sample versus the values of the control group taken as 100%. Results are presented as mean " S.E.M. throughout. Statistical significance of data was assessed by using an analysis of variance ŽANOVA.. When the F-value for the effect of the agents tested was significant Ž P - 0.05. the significance of the difference between two means was evaluated by the Fisher’s test. 2.3. DNA probes The rat MCH cDNA probe is pRMCH 11 cDNA w34x. The rat NT cDNA probe is prNT4 w1x. The rat TH cDNA probe is a Pst I fragment of pT51 cDNA w19x subcloned into pBSK vector ŽStratagene, La Jolla, CA.. The b-actin cDNA probe is cDNA pAL41 w2x. The GAPDH cDNA probe is a partial clone ŽpH GAPDH-4. cDNA w11x. The DNA probe identifying the 18S rRNA was kindly provided by Ladoux ŽIPMC, Valbonne, France.. These probes were labelled with w 32 PxdCTP by the random-priming method

272

F. Presse et al.r Molecular Brain Research 52 (1997) 270–283

using a commercial kit ŽPromega, Lyon, France.. The specific activity of the probes was 0.5–2 = 10 9 dpmrmg. 2.4. Determination of the transcription initiation site Primer extension analysis was performed with the RMCH3 primer Ž5X-CCTgTTTgCTgCTCCgTAAAgCCgAAggAg-3X . that was end-labelled with T4 polynucleotide kinase and w g- 32 PxATP. Labelled primer Ž4.10 5 dpm. was annealed to whole cell RNA Ž20 mg. isolated from rat hypothalamus or PC12 cells Žtreated or not with the cocktail of inducers. at 508C for 1 h in 10 mM Tris–HCl ŽpH 7.9., 1 mM EDTA, 125 mM KCl. Transcripts were extended for 1 h at 378C using Moloney murine leukaemia virus reverse transcriptase as described previously w36x. Samples were extracted with phenol–chloroform, ethanol precipitated and finally electrophoresed on a 6% polyacrylamide–8 M urea gel. 2.5. RT–PCR experiments Total RNAs from PC12 cells were reverse-transcribed with oligoŽdT., and the cDNAs were amplified with RPCR3 and RPCR4 primers. The sequence of RPCR3 is ATgCAgAATTTTTgTgAAgTTTAATgCAC. The sequence of RPCR4 is CTCCTTCggCTTTACggAgCAAACAgg. Briefly, 0.5–1 mg RNA was denaturated for 10 min at 808C in presence of 100 ng oligoŽdT., then reverse-transcribed into DNA by using 10 U of avian myeloblastosis virus reverse transcriptase ŽPromega, USA. in 15 ml of a solution containing 50 mM Tris–HCl ŽpH 8.0., 75 mM KCl, 3 mM MgCl 2 , 1 mM each of deoxy NTP, 1 mM dithiothreitol and 10 U of RNasine ŽPromega, Lyon, France.. The mixture was incubated at 378C for 1 h and then treated at 958C for 3 min and quick-chilled on ice. PCR was performed with 2 ml of the reverse transcription reaction mixture in a solution containing single-strength PCR buffer, 200 mM deoxy-NTPs, 1 mM each of 5X- and 3X-primers, and 0.2 U Taq-polymerase ŽAppligene, ` Strasbourg, France. in a total volume of 30 ml. The mixture was overlaid with mineral oil and then amplified with thermal cycles. The PCR profile involved a denaturation at 928C for 2 min, followed by 30 cycles of amplification Ždenaturation 45 s at 928C, annealing 1 min at 528C, extension 1 min 30 s at 728C., and a final extension for 7 min at 728C. Amplification products were electrophoresed onto a 1% agarose gel, transferred to a nylon membrane ŽHybond N, Amersham, Arlington Heights, IL., and hybridized with the 32 P-labelled MCH cDNA probe as previously described w36x. 2.6. Nuclear transcription analysis PC12 cells at about 10 7 per 90 cm diameter plate were treated or not with 100 ngrml NGF and 20 mM lithium for 6 or 24 h. Nuclei were prepared by a method derived

from that of Greenberg and Ziff w17x. Cells were resuspended in lysis buffer containing 10 mM Tris–HCl, pH 7.4, 10 mM NaCl, 3 mM MgCl 2 , vigorously checked and centrifuged 5 min at 1500 rpm at 48C. The nuclear pellet was washed twice in the lysis buffer with 0.3% Nonidet P40 during the last washing, gently resuspended and centrifuged at 48C for a few seconds. The supernatant was mixed with RNA buffer Ž10 mM Tris–HCl, pH 7.5, 7 M urea, 1% SDS, 0.3 M Na acetate, 20 mM EDTA. and RNA were purified following a phenolrchloroform extraction method w17x. After an additional washing in lysis buffer without NP 40, the pellet was resuspended in nucleus storage buffer Ž20 mM Tris–HCl, pH 8.0, 75 mM NaCl, 0.5 mM EDTA, 0.85 mM DTT, 0.125 mM phenylmethylsulfanyl fluoride, 50% glycerol. at a concentration of 1 to 5 = 10 7 nucleirml. The nuclei have been stored at y708C for months without loss of activity. In vitro transcription reactions were carried out in a 200 ml mixture containing 10 mM Tris–HCl, pH 7.9, 140 mM KCl, 2.5 mM MgCl 2 , 0.5 mM MnCl 2 , 1.7 mM spermidine, 14 mM b-mercaptoethanol, 0.1 mM S-adenosyl methionine, 10 mM creatine phosphate, 40 mgrml creatine phosphokinase, 25% glycerol, 1 mM ATP, CTP, GTP, 250 mCi w a- 32 PxUTP Ž400 Cirmmol., 0.5 mM aurintricarboxylic acid and 1–5 10 6 nuclei. Heparin sulphate Ž1 mgrml. or 10% sarkosyl was added to the reaction mixture and incubation was at 258C for 30 min as described previously w33x. PC12 cell nuclei consistently incorporated about 0.2–0.5 dpmrnucleus. Labelled RNA was purified as follows: a solution containing DNase-I, proteinase K Žboth at 0.5 mgrml., and 10 mM CaCl 2 was preincubated for 1 h at 378C. It was then added to the transcription mixture and this combination incubated at 378C for 1 h. Sodium dodecyl sulphate and EDTA were then added Žfinal concentrations of 1% and 15 mM, respectively. and the complete mixture kept at 378C for 10 min. Samples were extracted once with phenol ŽpH 4.0., twice with chloroform–isoamyl alcohol Ž24:1. and the 32 P-labelled RNA precipitated with ethanol at y208C in the presence of 2 M ammonium acetate and 10 mgrml tRNA carrier. The RNA pellets were washed twice with ethanol, dissolved in 100 ml of column buffer Ž10 mM Tris–HCl, pH 7.5, 0.3 M NaCl, 0.1% sodium dodecyl sulphate, 1 mM EDTA. and the remaining free nucleotides removed by filtration through a Sephadex G-50 minicolumn. The purified RNA samples were stored in the column buffer at y208C. MCH genomic fragments Ž0.5 and 2.5 mg., corresponding to the 5X flanking region Žp5.1., the central coding region ŽpG3.4. and the 3X flanking region Žp3.1. Žsee Fig. 4; Presse et al., in preparation. were spotted onto nylon membrane using a slot blot apparatus. The same amounts of TH cDNA w19x, GAPDH cDNA w11x and pBKS plasmid ŽStratagene, USA. were loaded to the same nylon membrane. The filters containing MCH genomic, TH cDNA, and control cDNA sequences were then prehybridized and hybridized to 32 P-labelled RNAs as

F. Presse et al.r Molecular Brain Research 52 (1997) 270–283

described elsewhere w33x. After hybridization the filters were washed twice with 1 = SSC Ž0.15 M NaCl, 0.015 M sodium citrate., 0.1% sodium dodecyl sulphate at 608C for 1 h then once in 2 = SSC solution. Hybridized filters were routinely treated with 10 mgrml RNaseA since significant decrease in the background of the filters was observed after treatment. Filters were then exposed to Kodak XOMAT S film at y708C using two Du Pont intensifying screens ŽCronex Lithting Plus.. Densitometric analysis was performed as described above.

3. Results 3.1. MCH and TH gene expression in PC12 cells Based on preliminary data, we hypothesized that expression of the MCH gene could be under the control of inducers of intracellular secondary pathways in PC12 cells. We first treated PC12 cells with a cocktail of activators, combining NGF Ž100 ngrml., dexamethasone Ž1 mM., forskolin Ž1 mM. and lithium Ž20 mM. and looked for MCH gene transcripts using Northern blotting method. A MCH cDNA probe identified a 0.95 kb MCH related RNA in induced PC12 cells ŽFig. 1A, T24 lanes 2 and 3. but not in control cells Žnot shown.. As expected, the NT probe revealed two bands corresponding to different 3X end-extended NT mRNA in treated PC12 cells w15x. The levels of MCH mRNA were 5 to 10 fold lower than these of NT mRNA by reference to GAPDH mRNA expression in treated PC12 cells, in sharp contrast with their relative levels found in the rat hypothalamus ŽFig. 1A, hyp. lane 1.. Unexpectedly, several RNA species of larger sizes Ž1.4, 3.5 and 4.0 kb. than mature MCH mRNA were clearly identified in untreated PC12 cells ŽFig. 1B, upper panel, C lanes 1 and 2. but not in cells treated for 6 to 12 h with NGF and lithium Žnot shown.. These MCH gene transcripts were also detected in PC12 cells treated for 24 h with NGF and lithium ŽFig. 1B, upper panel, NL 24 lanes 3 and 4. or the four inducers together Žnot shown.. Using intronic probes we revealed that the 3.5 and 4.0 kb RNA forms contained both intron A and B of the rat MCH gene Žnot shown.. Interestingly, the content of TH mRNA was drastically reduced in PC12 cells treated for 24 h with NGF and lithium ŽFig. 1B, mid panel, NL 24 lanes 3 and 4. but not with the four inducers together ŽFig. 2B.. Thus, these data indicate that MCH and NT mRNAs are positively controlled by a combination of NGF, dexamethasone, forskolin and lithium. In addition, NGF and lithium alone activates MCH and inhibits TH gene expression in PC12 cells. To confirm the Northern blot results and to define the transcription initiation site of the MCH gene in PC12 cells, primer extension analysis was carried out using a 30 base-synthetic oligonucleotide probe named RMCH3 which abuts on the translational initiation codon. A major 55

273

bp-extended product was obtained in the rat hypothalamic sample ŽFig. 1C, hyp. lane 1.. A weak band corresponding to the same extended product was identified in PC12 cells treated with combination of all inducers ŽFig. 1C, T24 lane 3. but was absent in PC12 cells under basal condition ŽFig. 1C, C lane 2.. In order to characterize further the MCH gene transcripts found in control or treated PC12 cells, reverse transcription ŽRT. –PCR experiments were performed using RPCR3 and RPCR4 primers that allowed identification of MCH mature transcripts ŽPCR product of 0.7 kb. and intron-containing transcripts including the primary transcript ŽPCR product of 1.3 kb.. As shown in Fig. 1D ŽC lane 3., a single band of 1.3 kb was detected in untreated cells suggesting the presence of intervening sequences within the MCH RNA transcripts of high size identified on Northern blots. Hybridization with specific oligoprobes demonstrated that intron A and B sequences of the MCH gene are encompassed within the 1.3 kb PCR product found in these cells Žnot shown.. A 0.7 kb PCR product was observed in the hypothalamic sample ŽFig. 1D, hyp. lane 1.. In treated cells the 0.7 kb–PCR fragment was found Žblack arrowhead. as well as others including those corresponding to the 1.3 kb RNA species and an alternative splicing RNA Žopened arrowhead. identified previously in rat hypothalamus w49x ŽFig. 1D, T24 lane 5.. 3.2. Regulation of MCH and TH mRNA by combinations of secondary pathway inducers Various combinations of NGF, dexamethasone, forskolin and lithium were tested to determine the effect of individual inducers and to reveal possible synergistic action on MCH and TH gene expression. As summarized in Fig. 2A, NGF and lithium can solely induce MCH mRNA expression, albeit at 10% of the maximum activation. However, activation of MCH gene expression by lithium alone was not observed reliably in separate experiments. In contrast, the increase induced by NGF was found consistently. Lithium acted synergistically with NGF to produce a 5 to 10 fold increase in the MCH mRNA level. Dexamethasone had no effect per se but it enhanced MCH mRNA expression in presence of NGF and lithium. Forskolin reduced the induction of MCH gene activity by NGF and had a highly variable effect in combination with lithium and with all inducers but dexamethasone. In the case of TH gene expression ŽFig. 2B., lithium solely or in combination with either of the other inducers except dexamethasone decreased drastically the TH mRNA levels. Conversely, dexamethasone alone or combined with other inducers except lithium increases TH mRNA content in PC12 cells. Surprisingly, forskolin solely provoked a slight induction of TH gene expression. Forskolin in combination with either NGF or lithium or with both components reduced TH mRNA levels. Unexpectedly, NGF was not required to allow NT

274

F. Presse et al.r Molecular Brain Research 52 (1997) 270–283

mRNA expression as reported by Dobner et al. w15x since treatment with either of the inducers provoked a low but detectable increase in NT mRNA levels Žabout 10 fold less than under maximal induction. ŽFig. 2C.. In addition, a combination of all inducers except lithium elicited a higher response than that obtained when lithium was included. These results indicate that NT gene expression in PC12 cells could be dependent on cellular clone origin or envi-

ronmental influences such as cell density as discussed below. 3.3. Lithium per se regulates MCH mRNA and TH mRNA expressions in NGF-treated PC12 cells Since lithium appeared to enhance MCH mRNA content only when cells were treated with NGF, the effect of

F. Presse et al.r Molecular Brain Research 52 (1997) 270–283

275

Fig. 2. Effect of lithium, forskolin, dexamethasone and NGF on MCH ŽA., TH ŽB. and NT ŽC. mRNA levels in PC12 cells. PC12 cells were treated with either all possible combinations of inducers or no inducer for 24 h, RNA was extracted and analyzed by Northern blotting with specific 32 P-labelled cDNA probes. MCH, TH and NT mRNA levels were quantitated by densitometric scanning of several blots and one representative set of data is depicted. The relative specific mRNA levels are expressed by taking as reference 100 the mean values found in cells treated with NGFrlithium Žfor MCH mRNA., no inducer Žfor TH mRNA. and the combination of all inducers Žfor NT mRNA., respectively. The results are presented as means of replicate Ž n s 4–15. for all combination of inducers determined in three independent experiments. p s 0.001 ŽMCH panel.; p s 0.0001 ŽTH panel.; p s 0.001 ŽNT panel.. ) Indicates significant difference between means according to the Fisher’s test.

LiCl alone after a short exposure of PC12 cells to NGF was studied in more detail. The cells were treated either with NGF Ž50 ngrml. and 20 mM LiCl for different times Žcondition I; Fig. 3A., with 50 ngrml NGF for 6 h before 20 mM lithium was added to the media for various times Žcondition II, Fig. 3A., or with 50 ngrml NGF for 6 h before cells were washed with phosphate-buffered saline prior to addition of 20 mM lithium and grown for up to 24 h Žcondition III, Fig. 3A.. As shown in Fig. 3B, combined treatment of NGF and lithium Žcondition I. resulted in an increase of the levels of MCH mRNA, reaching a maximum after 12 h and fol-

lowed by a four-fold decrease during the next 12 h. It is worth mentioning that maximum induction was found between 12 and 24 h, depending on the initial culture conditions Žnot shown.. NGF alone induced a slight timedependent increase in MCH mRNA levels Žnot shown.. Under conditions II and III, NGF stimulated MCH mRNA accumulation to 7–10% of the reference level Ž24 h-grown cells. after 6 h of treatment in PC12 cells Ž0 h, conditions II and III.. Addition of lithium to cells maintained in presence of NGF elicited a biphasic response, i.e., an initial response that tended to a plateau at 6 h Ž40% of the levels found after 24 h of culture. followed by a two-fold

Fig. 1. MCH gene expression in PC12 cells. ŽA. Northern blot analysis of PC12 cells treated for 24 h ŽT24. with a combination of NGF Ž100 ngrml., lithium Ž20 mM., forskolin Ž1 mM. and dexamethasone Ž1 mM.. 20 mg of RNA extracted from rat hypothalamus Žlane 1., or two plates ŽA, B. of treated PC12 cells Žlanes 2 and 3. were separated in a 1.2% agarose gel containing formaldehyde, transferred to nylon membrane and probed successively with 32 P-labelled MCH cDNA Žtop panel., NT cDNA Žmiddle panel. and GAPDH cDNA Žbottom panel.. The length of the RNA species is indicated on the right. ŽB. Northern blot analysis of control PC12 cells Žlanes 1 and 2. or cells treated for 24 h with NGF Ž100 ngrml. and lithium Ž20 mM. ŽNL 24 ; lanes 3 and 4.. 20 mg of RNA extracted from PC12 cells was analyzed as described above. The filter was probed successively with 32 P-labelled MCH cDNA Žtop panel., TH cDNA Žmiddle panel. and GAPDH cDNA Žbottom panel.. ŽC. Mapping of the transcription initiation site of rat MCH gene. Primer extension was performed with RMCH3 oligo probe using RNA extracted from hypothalamus Žhyp., lane 1., control PC12 cells ŽC, lane 2. and PC12 cells treated with a combination of the four inducers ŽT24, lane 3.. The labelled fragments were resolved on a denaturing polyacrylamide gel and sized by comparison with marker V ŽBoehringer Mannheim, RFG.. The arrowhead indicates the position of the extended fragment found in the hypothalamus and treated PC12 cells. Additional bands in lanes 1 and 2 correspond to primer extension artefacts. ŽD. RT–PCR analysis of MCH gene expression. RT–PCR experiments were performed with RNA extracted from the hypothalamus Žhyp., lane 1., untreated PC12 cells ŽC, lanes 2,3. and PC12 cells treated with combination of all inducers ŽT24, lanes 4,5. as described in A.. The PCR products were run on a 0.8% agarose gel and Southern blot analysis with 32 P-labelled MCH cDNA probe was carried out. Sizes of the PCR products corresponding to the primary transcript Ž1.3 kb., mature MCH mRNA Ž0.7 kb, black arrowhead. and MGOP mRNA Ž0.5 kb, open arrowhead. are indicated w49x. NRT: not reverse-transcribed, RT: reverse transcribed.

276

F. Presse et al.r Molecular Brain Research 52 (1997) 270–283

Fig. 3. Time course of inductionrinhibition of MCH and TH mRNA expression by lithium per se. ŽA. Schematic representation of the experimental procedures. PC12 cells were grown in the presence of NGF and lithium Žtreatment I., in the presence of NGF for 6 hours before lithium was added directly Žtreatment II. or after PBS washing Žtreatment III.. RNAs were isolated at the indicated times Žarrows. and Northern blot analysis was performed as described in Fig. 1. ŽB. The relative MCH mRNA levels were determined by taking as reference 100 the values found at 24 h for cells under treatment I. ŽC. The relative TH mRNA levels were determined by taking as reference 100 the values found at 0 h for cells under treatment I. Statistical significance was determined as in Fig. 2.

F. Presse et al.r Molecular Brain Research 52 (1997) 270–283

277

Fig. 4. Effect of NGFrlithium treatment on MCH and TH gene transcription in PC12 cells. ŽA. MCH gene organization. The black boxes indicate the exons, thick lines correspond to the introns, and the flanking regions are noted as thin lines. The MCH genomic clones derived from the lMCH-G1 clone ŽPresse et al., in preparation. are depicted as well as the location of some oligonucleotides used for primer extension ŽRMCH3. and RT–PCR ŽRPCR3-RPCR4. experiments. ŽB. Nuclei were isolated from control cells ŽC. at 6 ŽNL6. and 24 h ŽNL 24 . after NGFrlithium treatment. ‘Run on’ transcription experiments were performed as described in Section 2. The two sets of autoradiograms correspond to two independent experiments. pKS: bluescript plasmid; p5.1, pG3.4 and p3.1 are depicted in A; pTH: TH cDNA; GAPDH: GAPDH cDNA. ŽC. Densitometric analysis of ‘run on’ transcription and Northern blot data. The signal intensities corresponding to TH gene expression assayed by Northern blotting of cytoplasmic RNA Žnot shown. and ‘run on’ transcription were compared by reference with GAPDH gene expression for untreated cells Ž0 h., and lithiumrNGF-treated cells for 6 and 24 h. A.U.: arbitrary units.

278

F. Presse et al.r Molecular Brain Research 52 (1997) 270–283

elevation of MCH mRNA levels at 12 h and a small decrease at 24 h Žcondition II, Fig. 3B.. Importantly, these results indicated that the same lag of about 12 h is necessary to obtain the maximum elevation in MCH mRNA levels after addition of lithium even though PC12 cells were pretreated with NGF. Under condition III, lithium added to PC12 cells kept in absence of NGF induced a sharp peak of MCH mRNA level at 30 min followed by an asymptotic increase of MCH mRNA expression during the first 6 h of induction and resulted in constant production of MCH mRNA between 6 and 24 h. This plateau represents about 60% of the reference values of MCH mRNA expression found at 24 h under condition I. Under control conditions Žcondition I, Fig. 3C., lithium inhibited TH mRNA expression in a time-dependent manner while a similar decrease was noted between 1 and 12 h of treatment. Under condition II, lithium added directly to NGF-differentiated cells induced a transitory increase of TH mRNA levels after 12 h of culture. However, this was

not observed reliably in other experiments suggesting that this NGF-induced antagonism on lithium action was highly dependent on the cell culture conditions. Under condition III, a sharp and transitory increase was noted, then lithium provoked a decrease faster than that observed when lithium and NGF were initially combined Žcondition I, Fig. 3C.. This suggests a bimodal effect of lithium on TH gene expression in NGF-treated PC12 cells; a short-term transitory activation of TH mRNA synthesis followed by a long-term inhibitory action. Alternatively, we should also consider a general activatory effect when cells were washed before the addition of lithium since MCH gene expression exhibited a similar transient peak of activation under condition III ŽFig. 3B.. 3.4. Nuclear transcription and mRNA stability studies The changes in MCH and TH mRNA levels could be due to a transcriptional effect or to post-transcriptional

Fig. 5. Determination of the stability of MCH and TH mRNA. ŽA. and ŽB. PC12 cells were treated with NGFrlithium for different times in the presence ŽqDRB. or absence of 100 mM DRB. The amounts of MCH ŽA. and TH ŽB. mRNA were determined by scanning densitometry and normalized to 18S rRNA expression. ŽC. and ŽD. PC12 cells were treated for 24 h with NGFrlithium before 100 mM DRB was added ŽqDRB. or not to cultures and the amounts of MCH ŽC. and TH ŽB. mRNA were quantitated as described in Fig. 2. t s 0 in ŽC. and ŽD., time at which the DRB was added.

F. Presse et al.r Molecular Brain Research 52 (1997) 270–283

regulatory mechanisms. The relative transcriptional activity of the MCH and TH genes was evaluated by performing a nuclear ‘run-on’ assay in nuclei of PC12 cells grown for 6 or 24 h in serum-containing medium supplemented or not with NGF Ž100 ngrml. and lithium Ž20 mM.. For this purpose, 32 P-labelled RNAs corresponding to nascent RNA chains were isolated from nuclei and hybridized to fragments of DNA representative of the 5X flanking region of the MCH gene Žp5.1., the central portion ŽpG3.4. and the 3X flanking region Žp3.1. ŽFig. 4A.. Fragments of TH cDNA and GAPDH cDNA were spotted on the same membrane and hybridized to the same labelled RNA samples. As a control, RNAs were isolated from cytoplasm of the same cells used for nucleus isolation and a Northern blot was performed. As expected, treatment of PC12 cells with NGF and lithium produced a marked stimulation in MCH mRNA levels and down-regulated TH mRNA expression Žnot shown.. Hybridization of MCH gene fragments with in vitro elongated RNAs generated similar very low but reliable hybridization signals in both untreated and treated PC12 cells ŽFig. 4B, top panel.. In addition, hybridization was clearly detected in the 5X-flanking as well as the 3X-flanking sequences of the MCH gene ŽFig. 4B, bottom panel. in control cells ŽC. and at low levels in treated cells ŽNL 24. suggesting that the transcripts which were elongated in vitro include sequences both upstream and downstream from the coding region of the MCH gene. Taken together these results suggest that transcriptional activation of MCH gene may not be directly involved in mediating NGFrlithium response. By contrast, comparison of the relative transcription of THrGAPDH genes ŽFig. 4B and C. with the THrGAPDH mRNA ratios reveals a similar decrease after 6 or 24 h-treatment indicating that the regulation of the TH gene by lithium operates mainly at the transcriptional level. To further assess the relative contribution of post-transcriptional and transcriptional mechanisms in the regulation of MCH and TH genes under lithium treatment, mRNA stability studies were performed using PC12 cells in presence of the transcriptional inhibitor DRB. DRB inhibits transcription by inducing premature termination of RNA polymerase II-mediated RNA synthesis w8,28x. To begin to characterize the early induction of MCH mRNA and inhibition of TH mRNA, we examined the effect of DRB when added to NGF and lithium for a short-term treatment Ž0–12 h. and the steady-state levels of MCH mRNA and TH mRNA were determined by Northern blots. As shown in Fig. 5A, MCH mRNA expression in the presence of DRB was transiently decreased at 6 h and remained unchanged at 12 h in comparison with untreated cells. This indicates that arrest of transcription with DRB has either a slight negative effect on MCH mRNA synthesis or provokes a delay in the onset of reduced degradation. Most importantly, DRB treatment did not prevent the rise of MCH mRNA levels induced by

279

lithium and NGF at 12 h. TH mRNA was stabilized under DRB addition ŽFig. 5B. suggesting either that newly initiated transcription of a gene product was necessary to elicit TH mRNA inhibition or that DRB treatment modified the turn-over of the TH mRNA. To analyse the decay of the steady-state levels of MCH and TH mRNA after long-term exposure to NGF and lithium, PC12 cells were treated for 24 h with NGF and lithium, then DRB was added to fresh medium and the amount of MCH mRNA ŽFig. 5C. and TH mRNA ŽFig. 5D. measured by Northern blot. In the presence of DRB an increase in MCH mRNA levels was observed during the first 3 h of treatment followed by a rapid decay at 6 h. At 3 h, an average 4-fold increase in MCH mRNA stability was noted in comparison with the cells grown in absence of DRB ŽFig. 5C.. Treatment of PC12 cells with DRB after long-term treatment with lithium and NGF induced slight TH mRNA stabilization during the first 3 h but the levels reached the same values as those found in control cells during the last 6 h of treatment ŽFig. 5D.. These findings indicate that in long-term NGFrlithium-stimulated PC12 cells, DRB is producing a similar suppressing effect on MCH and TH mRNA degradation although it was less pronounced in the case of TH mRNA. The half-life of MCH mRNA and TH mRNA in absence or presence of DRB following a long-term treatment with NGF and lithium was estimated by regression analysis of the decline in mRNA levels by reference to the 100% value found at 0 h Žsee Fig. 5C and D and not shown.. The half-life of MCH mRNA was 3.0 " 0.5 h in absence of DRB and estimated to 9.0 " 2.0 h in DRB-treated cells. The increase in half-life for MCH mRNA in PC12 cells under DRB treatment was difficult to measure because an increase in MCH mRNA content was found during the first 3 h of DRB treatment by comparison to the level of reference at 0 h. The average half-life was determined therefore between the 6 and 12 h of treatment. Similar analysis revealed that the half-life of TH mRNA in DRB-treated and -untreated cells was similar, close to 10.5 h.

4. Discussion In this report, we examined the mechanisms by which known inducers of intracellular secondary pathways, involved in the control of NT gene expression w15x, regulate the synthesis of MCH and TH mRNA in PC12 cells. In particular, the effect of lithium on NGF-differentiated PC12 cells was investigated in detail because the lithium enhancement of MCH mRNA expression seems dependent of post-transcriptional regulations. This represents a novel mechanism of lithium action in this particular neuron-like cell line. Obviously, in vivo studies using animal models must be performed to ascertain whether modulation of the stability of MCH mRNA represents a potential target of lithium’s therapeutic action w23x.

280

F. Presse et al.r Molecular Brain Research 52 (1997) 270–283

The NT gene has been the subject of the most detailed studies on the molecular mechanisms by which lithium acts in PC12 cells w14x. NGF and lithium can substitute with each other to enhance dexamethasone and forskolin effects but act on NT gene transcription through distinct intracellular pathways w7,24x. We reproduce here the synergistic increase in the levels of NT mRNA with several of these inducers but the relative contributions of NGF, lithium, forskolin and dexamethasone in this process were clearly different from those reported previously w15x. In particular, under our conditions and in contrast to previous works w15,25,47x, lithium was not required to induce maximum NT mRNA accumulation. Cellular contact or the production of diffusible factors which depend on cell density might have an influence on the levels of NT mRNA in PC12 cells. In addition, one must consider the difference in the PC12 cellular clone origin. Our results dealing with the action of glucocorticoids, NGF and forskolin on TH mRNA are in general agreement with previous works Žw30x, reviewed in Ref. w27x.. In particular, dexamethasone alone or in combination with NGF andror forskolin positively regulates TH gene activity. In contrast, lithium alone or in combination with NGF andror forskolin exerts a strong inhibitory effect on TH mRNA expression. This down-regulation operates mainly at the transcriptional level while secondary post-transcriptional mechanisms might also contribute to the decrease of steady state levels of the TH mRNA as suggested by our studies using the DRB transcription inhibitor. Our data provide the first evidence that lithium could act directly on TH gene activity in a neuron-like cell model. Multiple signal transduction systems and their cognate transcription factors may participate in the regulation of TH mRNA synthesis Žw30,51x, reviewed in Ref. w27x.. The presence of a functional AP-1 site in the promoter region of the TH gene w53x strongly suggests that this represents the ultimate target integrating the lithium-induced intracellular responses as found in the model of NT gene regulation w24x. In this context, the lithium-provoked TH mRNA decline was almost completely prevented by adding DRB suggesting that a transcriptionally lithium-inducible gene product mediated TH gene repression. NGF and lithium rapidly activate the expression of AP1 genes such as c-fos, c-jun and fra-1 as well as AP-1 complex binding activity in PC12 cells w7x. Therefore, it is tempting to speculate that the inhibition of lithium effects by DRB reflects the repression of AP-1 genes. However, other regulatory proteins, such as these recognizing the E box dyed motifs necessary for tissue-specific transcription of the TH gene w53x or the Oct 2 transcription factor which binds to an octamer motif involved in repression of TH gene activity w12x, may also be involved. Our most striking result is that MCH mRNA content may be strongly up-regulated by lithium per se in NGF-induced PC12 cells. NGF appears to play a permissive role for the increase in MCH mRNA levels detectable as early

as 3 h after induction. Lithium greatly enhanced MCH mRNA expression after 12–24 h treatment in combination with NGF or following a short exposure Ž6 h. of PC12 cells to NGF Ž5 to 20-fold by reference with the induction noted with NGF only.. However, the kinetics of induction by lithium were clearly different depending on the presence of NGF in the culture media Žsee Fig. 3B., suggesting that these two agents cooperate through distinct permissive intracellular pathways as reported for NT gene regulation in PC12 cells w14x. In addition, long exposure to lithium Ž) 24 h. in combination with NGF alone or with the whole mixture of inducers resulted in the decrease of MCH mRNA content indicating that activation of MCH mRNA expression by lithium was transitory in the PC12 cell model. The cAMP analog 8-Br-cAMP promotes both the synthesis and release of MCH in cultured rat hypothalamic cells w35x suggesting that cAMP-activated intracellular cascade might function in MCH pericarya. In PC12 cells, the adenylate cyclase activator Žforskolin. had no effect per se. It had contrasting effect on the regulation of MCH gene exerted by NGF and lithium. In particular, forskolin treatment prevents NGF-induced MCH mRNA increase while it allows induction of MCH mRNA levels in presence of lithium Žsee Fig. 2A.. This suggests that differential crosstalks may be exerted between the PKA-activated pathway and lithium or NGF-activated second messenger systems. At present, a sequence that might confer responsiveness to cyclic AMP ŽCRE; w16x. has not been found within the available sequence of the rat MCH gene ŽPresse et al., in preparation. preventing us from concluding whether cAMP activated pathway could control directly or indirectly the MCH gene transcription. Glucocorticoids activate MCH mRNA expression in the rat hypothalamus w36x and increase both the synthesis and secretion of MCH in cultured rat hypothalamic cells w35x. While dexamethasone had no effect per se on MCH mRNA expression in PC12 cells, it potentiated the positive action of lithium and NGF on MCH mRNA synthesis Žsee Fig. 2A.. A full glucocorticoid response element ŽGRE. and a partial GRE are located in the 5X flanking region of the rat MCH gene ŽPresse et al., in preparation.. Our results suggest that glucocorticoids might act cooperatively with lithium andror NGF to increase MCH gene activity and that either of these GRE might be involved in such a regulation. The mechanisms regulating MCH mRNA production by lithium in PC12 cells are of particular interest given the unexpected finding that the regulation of MCH mRNA stability represents the main determinant of its expression. Transcription blockade with DRB during the initial phase of MCH mRNA induction by the NGF and lithium resulted in a transitory down-regulation of MCH mRNA content but did not prevent the strong increase of MCH mRNA levels after 12 h of culture. In addition, DRB-induced stabilization of MCH mRNA within the first 3 h

F. Presse et al.r Molecular Brain Research 52 (1997) 270–283

following long-term NGFrlithium treatment might be attributed to the inhibition of a labile gene product required for MCH mRNA degradation. Similar stabilization of mRNA after DRB treatment has been observed in the cases of the c-fos w44x, the NPY w29x and the VIP w48x transcripts and likely resulted from the loss of one or a number of proteins involved in degradation of these RNAs. Thus we provide clear evidence that the majority of the steady state variations of MCH mRNA in PC12 cells does not result from transcriptional mechanisms but that the MCH mRNA degradative pathway is somehow altered by lithium treatment. To our knowledge, this represents the first example of lithium action on neuropeptide-encoding RNA production at a post transcriptional level. Do the changes observed at the MCH mRNA level may be parallel with similar modifications in peptide synthesis and release? Using a specific RIA to assess MCH levels we failed to reveal the presence of MCH precursor or peptide into PC12 cells or in the media due to cross-reactivities of MCH antisera with unknown compounds of PC12 peptide extracts Žour unpublished data.. Other antisera should be available soon and will allow determination of the pro-MCH and derived peptide levels. The MCH gene transcript of 0.95 kb identified in the treated cells may correspond to the reported MCH mRNA with a short poly A tail found in extra hypothalamic brain areas w37x and in peripheral tissues w20x of the rat. Indeed, enzymatic removal of the polyŽA. tract with RNaseH indicated that the MCH mRNA found in PC12 cells had a shorter polyŽA. tail than this isolated from hypothalami ŽBorsu et al., in preparation.. Unexpectedly, we identified several MCH RNA species of higher size than the mature mRNA in untreated cells. These transcripts drastically decreased following short-term lithium treatment and were detected again in 24 h-treated cells, albeit at low levels Žunpublished data.. The nature of these MCH RNA species remains poorly defined. Northern blot and RT–PCR experiments indicated the synthesis of transcripts containing the intronic sequences of the MCH gene in untreated cells. Therefore, the large MCH RNA species could represent large precursors of the MCH mRNA. In this context, the run-on transcription assay suggests that transcripts containing both 5X-flanking and 3X-flanking MCH gene regions might be produced in untreated and NGFrlithium treated PC12 cells Žsee Fig. 4B.. It is now of an obvious interest to elucidate the structure of these MCH RNA-related species and to reveal their possible involvement in the post-transcriptional regulation of MCH gene activity. The observation that lithium up-regulated MCH mRNA and down-regulated TH mRNA in the same cells might be relevant in a physiological context. Both MCH and catecholamine neuronal systems can be considered as nonspecific regulators of brain activity, involved in the controls of appropriate behavioral response according to the environmental context w32,41x. Experimental stress provoked marked down-regulation of MCH mRNA expres-

281

sion in the hypothalamus w36x and induced TH gene expression and enzyme activity in the brain Žreviewed in Refs. w31,52x.. In addition, while monoamines activate the stress response, MCH prevents ACTH rise following mild stress w5x. The opposite effects and regulation of MCH and catecholamines might reflect functional antagonism in the central nervous system. The close location of MCH neurons and monoamine fibres within the medial forebrain bundle crossing the lateral hypothalamus w4,26x and the possible coexistence of MCH and catecholamines in the multipolar neurons of the ZI w50x provide further arguments for functional cross-talk. In addition, colocalization of pro-MCH derived peptides and monoamines in resident macrophages of the gastrointestinal tract has been recently documented w21x, suggesting possible interactions between these two transmitter systems in the regulation of intestinal functions. In summary, we provided here a detailed analysis of MCH, TH and neurotensin gene expression in PC12 cells treated with NGF, dexamethasone, forskolin and lithium, four compounds activating distinct but somewhat cooperative signal transduction pathways. In focused studies we demonstrated that combined lithium and NGF treatment result in a marked elevation of MCH mRNA and a suppression of TH mRNA production. Strikingly, the lithium up-regulates MCH gene likely by stabilising its mRNA, while it decreases transcription of tyrosine hydroxylase. It is now of an obvious interest to examine the effects of lithium at therapeutic concentrations on MCH and TH gene regulation in the rat brain and to compare them with those reported in the present studies. Acknowledgements We are grateful to Drs. Patrick Kitabgi and Stephen McSorley ŽI.P.M.C., Valbonne, France. for critical reading of the manuscript. We thank Drs. Michel Darmon ŽCentre de Biochimie, Nice, France., Paul R. Dobner ŽUniversity of Massachusetts Medical Center, Worcester, USA. and Nicole Faucon-Biguet ŽUMR 9923 CNRS, Paris, France. very much for the generous gift of the GAPDHrb-actin cDNA, NT cDNA and TH cDNA, respectively. We are indebted to Dr. Carole Rovere ` and Pierre Barbero ŽI.P.M.C., Valbonne, France. for providing the PC12 cell line and kind assistance. We thank J. Kervella for typing the manuscript and F. Aguila for preparation of the artworks. This work was supported by the ‘‘Association pour la Recherche sur le Cancer’’ ŽARC 6987. and the ‘‘Ministere ` de l’Education Nationale’’ ŽGrant 95G0099.. Laetitia Borsu is a recipient of a ‘allocation MENESR’. References w1x M.J. Alexander, M.A. Miller, D.M. Dorsa, B.P. Bullock, R.H. M elloni, P.R. Dobner, S.E. Leeman, Distribution of neurotensinrneuromedin N mRNA in rat forebrain: unexpected

282

w2x

w3x w4x

w5x

w6x

w7x

w8x

w9x

w10x

w11x

w12x

w13x

w14x

w15x

w16x w17x w18x

w19x

w20x

w21x

F. Presse et al.r Molecular Brain Research 52 (1997) 270–283 abundance in hippocampus and subiculum, Proc. Natl. Acad. Sci. USA 86 Ž1989. 5202–5206. S. Alonso, A. Minty, Y. Bourlet, M. Buckingham, Comparison of three actin-coding sequences in the mouse; evolutionary relationship between the actin genes of warm-blooded vertebrates, J. Mol. Evol. 23 Ž1986. 11–22. B.B. Baker, Melanin-concentrating hormone updated. Functional considerations, Trends Endocrinol. Metab. 5 Ž1994. 120–126. J.C. Bittencourt, F. Presse, C. Arias, C. Peto, J. Vaughan, J.L. Nahon, W. Vale, P.E. Sawchenko, The melanin-concentrating hormone system of the rat brain: an immuno- and hybridization histochemical characterization, J. Comp. Neurol. 319 Ž1992. 218–245. M.T. Bluet-Pajot, F. Presse, Z. Voko, ¨ C. Hoeger, F. Mounier, J. Epelbaum, J.L. Nahon, Neuropeptide-E-I antagonizes the action of melanin-concentrating hormone on stress-induced release of adrenocorticotropin in the rat, J. Neuroendocrinol. 7 Ž1995. 297–303. C. Breton, F. Presse, G. Hervieu, J.L. Nahon, Structure and regulation of the mouse melanin-concentrating hormone mRNA and gene, Mol. Cell. Neurosci. 4 Ž1993. 271–284. B.P. Bullock, G.P. McNeil, P.R. Dobner, Synergistic induction of neurotensin gene transcription in PC12 cells parallels changes in AP-1 activity, Mol. Brain Res. 27 Ž1994. 232–242. L.A. Chodosh, A. Fire, M. Samuels, P.A. Sharp, 5,6-Dichloro-1-bD-ribofuranosylbenzimidazole inhibits transcription elongation by RNA polymarase II in vitro, J. Biol. Chem. 264 Ž1989. 2250–2257. P. Chomczynski, N. Sacchi, Single-step method of RNA isolation by acid guanidinium thiocyanate–phenol–chloroform extraction, Anal. Biochem. 162 Ž1987. 156–159. N. Compagnone, D. Fellmann, C. Bugnon, C. Jaquemard, Diffusible factors from rat arcuate nucleus and Broca’s diagonal band nucleus increase size an neurite outgrowth, respectively, of cultured melanin-concentrating hormone containing neurons, Brain Res. 628 Ž1993. 137–144. C. Dani, J.M. Blanchard, M. Piechaczyk, S.E. Sabouty, L. Marty, P. Jeanteur, Extreme instability of myc mRNA in normal and transformed human cells, Proc. Natl. Acad. Sci. USA 81 Ž1984. 7046– 7050. S.J. Dawson, S.O. Yoon, D.M. Chikaraishi, K.A. Lillycrop, D.S. Latchman, The Oct-2 transcription factor represses tyrosine hydroxylase expression via a heptamer TAATGARAT-like motif in the gene promoter, Nucleic Acids Res. 22 Ž1994. 1023–1028. A.Y. Deutch, D.S. Zahm, The current status of neurotensin-dopamine interactions. Issues and speculations, Ann. NY Acad. Sci. 668 Ž1992. 232–252. P.R. Dobner, E. Kislauskis, B.P. Bullock, Cooperative regulation of neurotensinrneuromedin N gene expression in PC12 cells involves AP-1 transcription factors, Ann. NY Acad. Sci. 668 Ž1992. 17–29. P.R. Dobner, A.S. Tischler, Y.C. Lee, S.R. Bloom, S.R. Donahue, Lithium dramatically potentiates neurotensinrneuromedin N gene expression, J. Biol. Chem. 263 Ž1988. 13983–13986. R.H. Goodman, Regulation of neuropeptide gene expression, Annu. Rev. Neurosci. 13 Ž1990. 111–127. M.E. Greenberg, E.B. Ziff, Stimulation of 3T3 cells induces transcription of the c-fos proto-oncogene, Nature 311 Ž1984. 433–438. L.A. Greene, A.S. Tischler, Establishment of a noradrenergic clonal line of rat adrenal pheochromocytoma cells which respond to nerve growth factor, Proc. Natl. Acad. Sci. USA 73 Ž1976. 2424–2428. B. Grima, A. Lamouroux, F. Blanot, N. Faucon Biguet, J. Mallet, Complete coding sequence of rat tyrosine hydroxylase mRNA, Proc. Natl. Acad. Sci. USA 82 Ž1985. 617–621. G. Hervieu, J.L. Nahon, Pro-melanin concentrating hormone messenger ribonucleic acid and peptides expression in peripheral tissues of the rat, Neuroendocrinology 61 Ž1995. 348–364. G. Hervieu, K. Volant, O. Grishina, M. Descroix-Vagne, J.L. Nahon, Similarities in cellular expression and functions of melaninconcentrating hormone and atrial natriuretic factor in the rat digestive tract, Endocrinology 137 Ž1996. 561–571.

w22x D. Jezova, V. Bartanusz, I. Westergren, B.B. Johansson, J. Rivier, W. Vale, C. Rivier, Rat melanin-concentrating hormone stimulates adrenocorticotropin secretion: evidence for a site of action in brain regions protected by the blood brain barrier, Endocrinology 130 Ž1992. 1024–1029. w23x R.S. Jope, M.B. Williams, Lithium and brain signal transduction systems, Biochem. Pharmacol. 47 Ž1994. 429–441. w24x E. Kislauskis, P.R. Dobner, Mutually dependent response elements in the Cis-regulatory region of the neurotensinrneuromedin N gene integrate environmental stimuli in PC12 cells, Neuron 4 Ž1990. 783–795. w25x E. Kislauskis, P.R. Dobner, Identification of Cis-regulatory sequences required for induction of neurotensinrneuromedin N gene expression in PC12 cells, Methods Neurosci. 9 Ž1992. 347–361. w26x K.M. Knigge, D. Baxter-Grillo, J. Speciale, J. Wagner, Melanotropic peptides in the mammalian brain: the melanin-concentrating hormone, Peptides 17 Ž1996. 1063–1073. w27x S.C. Kumer, K.E. Vrana, Intricate regulation of tyrosine hydroxylase activity and gene expression, J. Neurochem. 67 Ž1996. 443–462. w28x O. Laub, E.B. Jakobovits, Y. Aloni, 5,6-Dichloro-1-b-ribofuranosylbenzimidazole enhances premature termination of late transcription of simian virus 40 DNA, Proc. Natl. Acad. Sci. USA 77 Ž1980. 3297–3301. w29x R.A. Lerchen, D.-Y. Yum, R. Krajcik, C.A. Minth-Worby, Transcriptional vs. Posttranscriptional control of neuropeptide Y gene expression, Endocrinology 136 Ž1995. 833–841. w30x E.J. Lewis, C.A. Harrington, D.M. Chikaraishi, Transcriptional regulation of the tyrosine hydroxylase gene by glucocorticoid and cyclic AMP, Proc. Natl. Acad. Sci. USA 84 Ž1987. 3550–3554. w31x T. Nagatsu, Genes of human catecholamine synthesizing enzymes and their regulation during stress, In: R. Kveinansky, R. McCarty, J. Axelrod ŽEds.., Stress: Neuroendocrine and Molecular Approaches, Vol. 1, Gordon and Breach, New York, 1992, pp. 287–295. w32x J.L. Nahon, The melanin-concentrating hormone: from the peptide to the gene, Crit. Rev. Neurobiol. 8 Ž1994. 221–262. w33x J.L. Nahon, J.L. Danan, M. Poiret, I. Tratner, M. Jose-Estanyol, J.M. Sala-Trepat, The rat alpha-fetoprotein and albumin transcription units: transcriptional control and comparison of the sequence organization and promoter region, J. Biol. Chem. 26 Ž1987. 12479–12487. w34x J.L. Nahon, F. Presse, J.C. Bittencourt, P. Sawchenko, W. Vale, The rat melanin-concentrating hormone mRNA encodes multiple putative neuropeptides coexpressed in the dorsolateral hypothalamus, Endocrinology 125 Ž1989. 2056–2065. w35x D.G. Parkes, W. Vale, Secretion of melanin-concentrating hormone and neuropeptide-EI from cultured rat hypothalamic cells, Endocrinology 131 Ž1992. 1826–1831. w36x F. Presse, G. Hervieu, T. Imaki, P.E. Sawchenko, W. Vale, J.L. Nahon, Rat melanin-concentrating hormone messenger ribonucleic acid expression: marked changes during development and after stress and glucocorticoid stimuli, Endocrinology 131 Ž1992. 1241– 1250. w37x F. Presse, J.L. Nahon, Differential regulation of Melanin-concentrating hormone gene expression in distinct hypothalamic areas under osmotic stimulation in rat, Neuroscience 55 Ž1993. 709–720. w38x F. Presse, I. Sorokovsky, J.P. Max, S. Nicolaidis, J.L. Nahon, Melanin-concentrating hormone is a potent anorectic peptide regulated by food-deprivation and glucopenia in the rat, Neuroscience 71 Ž1996. 735–745. w39x D. Qu, D.S. Ludwig, S. Gammeltoft, M. Piper, M.A. Pelleymounter, M.J. Cullen, W.F. Mathes, J. Przupek, R. Kanarek, E. Maratos-Flier, A role for melanin-concentrating hormone in the central regulation of feeding behaviour, Nature 380 Ž1996. 243–247. w40x P.Y. Risold, D. Fellmann, J. Rivier, W. Vale, C. Bugnon, Immunoreactivities for antisera to three putative neuropeptides of the rat melanin-concentrating hormone precursor are coexpressed in neurons of the rat lateral dorsal hypothalamus, Neurosci. Lett. 136 Ž1992. 145–149.

F. Presse et al.r Molecular Brain Research 52 (1997) 270–283 w41x J.P. Rivot, M. Peschanski, B. Calvino, Les systemes non specifiques., ` ´ Med. Sci., Lexique Neurobiol, Ž1992. 81–82. w42x M. Rossi, S.J. Choi, D. O’Shea, T. Miyoshi, M.A. Ghatei, S.R. Bloom, Melanin-concentrating hormone acutely stimulates feeding, but chronic administration has no effect on body weight, Endocrinology 138 Ž1997. 351–355. w43x C. Rovere, ` F. De Nadai, J.-N. Bidard, J.-C. Cuber, P. Kitabgi, PC12 cells can be induced to produce but do not process, the neurotensinrneuromedin N precursor, Peptides 14 Ž1993. 983–989. w44x A.-B. Shyu, M.E. Greenberg, J.G. Belasco, The c-fos transcript is targeted for rapid decay by two distinct mRNA degradation pathways, Gene Develop. 3 Ž1989. 60–72. w45x G. Skofitsch, D.M. Jacobowitz, N. Zamir, Immunohistochemical localization of a melanin concentrating hormone-like peptide in the rat brain, Brain Res. Bull. 15 Ž1985. 635–649. w46x R.C. Thompson, S.J. Watson, Nucleotide sequence and tissuespecific expression of the rat melanin concentrating hormone gene, DNA Cell Biol. 9 Ž1990. 637–645. w47x A.S. Tischler, L.A. Ruzicka, P.R. Dobner, A protein kinase inhibitor, staurosporine, mimics nerve growth factor induction of neurotensinrneuromedin N gene expression, J. Biol. Chem. 266 Ž1991. 1141–1146. w48x P.J. Tolentino, L. Villa-Komaroff, Regulation of vasoactive intestinal polypeptide and galanin mRNA stabilities, Mol. Brain Res. 39 Ž1996. 89–98.

283

w49x G. Toumaniantz, J.C. Bittencourt, J.L. Nahon, The rat melanin-concentrating hormone gene encodes an additional putative protein in a different reading frame, Endocrinology 137 Ž1996. 4518–4521. w50x M.V. Ugrumov, J. Taxi, A. Tixier-Vidal, J. Thibault, M.S. Mitskevich, Ontogenesis of tyrosine hydroxylase-immunopositive structures in the rat hypothalamus. An atlas of neuronal cell bodies, Neuroscience 29 Ž1989. 135–156. w51x S. Vyas, N. Faucon Biguet, J. Mallet, Transcriptional and post-transcriptional regulation of tyrosine hydroxylase gene by protein kinase C, EMBO J. 9 Ž1990. 3707–3712. w52x T.C. Wessel, K.-T. Kim, K.S. Kim, J.M. Carroll, H. Baker, T.H. Joh, Gene expression and regulation of monoamine enzymes during stress. In: R. Kvetnansky, R. McCarty, J. Axelrod ŽEds.., Stress: Neuroendocrine and molecular approaches., Vol. 1, Gordon and Breach, New York, 1992, pp. 297–312. w53x S.O. Yoon, D.M. Chikaraishi, Tissue-specific transcription of the rat tyrosine hydroxylase gene requires synergy between an AP-1 motif and an overlapping E box-containing dyad, Neuron 9 Ž1992. 55–67. w54x N. Zamir, G. Skofitsch, M.J. Bannon, D.M. Jacobowitz, Melaninconcentrating hormone: unique peptide neuronal system in the rat brain and pituitary gland, Proc. Natl. Acad. Sci. ŽUSA. 83 Ž1986. 1528–1531.