Rat Neuropeptide Y Precursor Gene Expression mRNA STRUCTURE, TISSUE DISTRIBUTION, AND REGULATION BY GLUCOCORTICOIDS, CYCLIC AMP, AND PHORBOL ESTER*

Rat brain neuropeptide Y precursor (prepro-NPY) cDNA clones were isolated and sequenced in order to study regulation of the prepro-NPY gene. Rat prepro-NPY (98 amino acid residues) contains a 36-residue NPY sequence, followed by a proteolysis/amidation site Gly-Lys-Arg, followed by a 30-residue COOH- terminal sequence. The strong evolutionary conservation of rat and human sequences of NPY (100%) and COOH-terminal peptide (93%) suggests that both peptides have important biological functions. In the rat central nervous system, prepro-NPY mRNA (800 bases) is most abundant in the striatum and cortex and moderately abundant in the hippocam- pus, hypothalamus, and spinal cord. The rat adrenal, spleen, heart, and lung have significant levels of pre- pro-NPY mRNA. Regulation of the prepro-NPY mRNA abudance was studied in several rodent neural cell lines. PC12 rat pheochromocytoma and N18TG-2 mouse neuroblastoma cells possess low basal levels of prepro-NPY mRNA, while NGlO8-15 hybrid cells possess high lev- els. or elevation of cAMP by forskolin increased the prepro-NPY mRNA level 2-3-fold or 3-10-fold, respectively. In N18TG-2 cells dexamethasone and forskolin synergistically increased prepro- NPY mRNA 7-fold. Treatment of PC12 cells with the protein kinase C activator phorbol 12-myristate 13- acetate alone elevated prepro-NPY mRNA marginally, but the phorbol ester plus forskolin elicited 20-70-fold increases, which were further enhanced to over 200- fold by dexamethasone and the calcium ionophore A23187. These results indicate that NPY gene expression can be positively regulated by synergistic actions of glucocorticoids, cAMP elevation, and protein kinase C activation.

Rat brain neuropeptide Y precursor (prepro-NPY) cDNA clones were isolated and sequenced in order to study regulation of the prepro-NPY gene. Rat prepro-NPY (98 amino acid residues) contains a 36-residue NPY sequence, followed by a proteolysis/amidation site Gly-Lys-Arg, followed by a 30-residue COOHterminal sequence. The strong evolutionary conservation of rat and human sequences of NPY (100%) and COOH-terminal peptide (93%) suggests that both peptides have important biological functions.
In the rat central nervous system, prepro-NPY mRNA (800 bases) is most abundant in the striatum and cortex and moderately abundant in the hippocampus, hypothalamus, and spinal cord. The rat adrenal, spleen, heart, and lung have significant levels of prepro-NPY mRNA.
Regulation of the prepro-NPY mRNA abudance was studied in several rodent neural cell lines. PC12 rat pheochromocytoma and N18TG-2 mouse neuroblastoma cells possess low basal levels of prepro-NPY mRNA, while NGlO8-15 hybrid cells possess high levels. Treatment of PC12 cells with a glucocorticoid such as dexamethasone or elevation of cAMP by forskolin increased the prepro-NPY mRNA level 2-3-fold or 3-10-fold, respectively. In N18TG-2 cells dexamethasone and forskolin synergistically increased prepro-NPY mRNA 7-fold. Treatment of PC12 cells with the protein kinase C activator phorbol 12-myristate 13acetate alone elevated prepro-NPY mRNA marginally, but the phorbol ester plus forskolin elicited 20-70-fold increases, which were further enhanced to over 200fold by dexamethasone and the calcium ionophore A23187. These results indicate that NPY gene expression can be positively regulated by synergistic actions of glucocorticoids, cAMP elevation, and protein kinase C activation.
Neuropeptide Y (NPY),' a 36-amino acid COOH-terminally amidated peptide, was originally isolated from porcine brain by Tatemoto et al. (1,2) and found to have a structural similarity to peptide YY and pancreatic polypeptide, which are found in the gastrointestinal tract and pancreas. One of the most abundant neuropeptides, NPY is widely distributed in central and peripheral neurons (3)(4)(5) as well as cells of neural crest origin such as adrenal chromaffin cells (6)(7)(8).
NPY is colocalized with catecholamines, y-aminobutyric acid, or other neuropeptides in many neurons (9-12). Evidence is mounting that NPY is involved in the regulation of peripheral and cardiac artery blood pressure (4, 10, 13-15), circadian rhythms (16), release of hypothalamic hormones (17), and feeding behavior (18). NPY also inhibits catecholamine release from the rat vas deferens (19) and cultured bovine chromaffin cells (20). Brief reviews on NPY neuronal systems have appeared recently (21, 22). The nucleotide sequence of human pheochromocytoma prepro-NPY mRNA (NPY mRNA) and the organization of the human NPY gene have been determined (23, 24). Human prepro-NPY (97 amino acid residues) contains a signal peptide sequence, the 36-amino acid NPY sequence, and a 30residue COOH-terminal peptide sequence. The COOH-terminal peptide ("CPON") is reported to be colocalized with NPY in many tissues but as yet has no known biological or pharmacological activity (25,26).
Information about the regulation of the NPY gene by hormones, neurotransmitters, and second-messenger systems is currently lacking. In order to study this regulation in experimental animals and in rodent NPY-containing cell lines (27, 28), we cloned and sequenced rat prepro-NPY cDNA. Using this probe we have examined the tissue-specific expression of the NPY gene in the rat and have delineated regulatory influences of glucocorticoids, CAMP, and phorbol esters on the NPY mRNA abundance in neural cell lines.
Some of these data have been presented in abstract form (29).

EXPERIMENTAL PROCEDURES AND RESULTS~
Structure of Rat Brain NPY mRNA--Rat brain NPY cDNA clones were isolated (Fig. 1, Miniprint). The sequences of rat NPY mRNA and its gene product are compared in Fig. 2 with the previously determined human sequences (23). An uncommon AUGAUG sequence begins the 294-nucleotide open reading frame in rat NPY mRNA, in contrast to human NPY mRNA, in which a single AUG opens the reading frame. The

FIG. 2.
Nucleotide sequence of rat brain NPY cDNA, deduced amino acid sequence of rat prepro-NPY, and comparison of rat sequences with reported human sequences (23). Nucleotide sequence alignment was performed with the NUCALN program (62) on the DEC-10 computer of the NIH Computer Center. Identical bases in the mRNA sequences are indicated with dots. Nonidentical amino acids are indicated with bold type in both rat and human sequences. The signal peptide, NPY (lacking the terminal amide), and the COOHterminal peptide sequences are shown in the labeled boxes. The initiating codon of the rat mRNA is presumed to be the first AUG. Triple stars indicate chain termination. The putative AAUAAA polyadenylation signals are underlined.
sequence surrounding the first AUG in the rat mRNA corresponds better than the sequence surrounding the second AUG to the optimal context for polypeptide chain initiation suggested by Kozak (30). Therefore, initiation at the first AUG is assumed in the numbering of amino acid residues of rat prepro-NPY in Fig. 2. The deduced sequence of rat brain prepro-NPY (98 amino acid residues, M, 11,032) includes a sequence (residues 30-65) indentical to human NPY. A cleavage by signal peptidase between GlyZ9 and Tyr30 (analogous to Alam-TyP in human prepro-NPY) would produce pro-NPY (69 residues, M, 8,059). The mature active NPY peptide (M, 4,271) and COOH-terminal peptide (MI 3,463) then would be generated from pro-NPY by proteolysis within the region Gl~~-Lys~~-Arg68, followed by an enzymatically catalyzed amidation (31,32) in which Tyr@-Glyee is converted to Tyr-NH,.
The overall similarity of rat and human NPY mRNA sequences is high (81 and 94% for nucleotides and amino acids, respectively), with extensive conservation in the coding region and near the polyadenylation signal. The amounts of amino acid identity in the three coding domains are 86,100, and 93% for the signal, NPY, and COOH-terminal peptides, respectively. The amounts of nucleotide identity are 82, 94, and 90%, respectively.
Tissue Distribution of NPY mRNA-NPY mRNA from various rat brain regions electrophoresed as a broad band of mean length 800 bases (Fig. 3A). A similar estimate of length was obtained with purified poly(A)+ RNA from rat striatum, cortex, hypothalamus, and spinal cord (not shown). NPY mRNA from peripheral tissues electrophoresed in narrower bands of smaller average size (720 bases) (Fig. 3B). The abundances of NPY mRNA and NPY immunoreactivity in brain regions and other tissues were determined (Table I, see also Fig. 4 in Miniprint). The highest abundance in brain was found in striatum, with cortex slightly lower. The spinal cord, hypothalamus, and hippocampus have intermediate levels.
The levels of NPY immunoreactivity in rat brain regions are

Regulation of Rat Neuropeptide Y Precursor Gene Expression
in good agreement with NPY mRNA levels except in the hypothalamus, which has much more NPY peptide relative to its NPY mRNA content. NPY mRNA was found to be abundant also in the adrenal gland, spleen, heart, and lung. Regulation of NPY Gene Expression by Glucocorticoids and CAMP in Cultured Cells"PC12 rat pheochromocytoma cells, which are studied as a model of neural crest cell differentiation 4400-A 2400- 5, thyroid; 6, liver; 7, duodenum +jejunum; 8, spleen; and 9, striatum of brain.

300-
(33, 34), possess a low level of NPY mRNA (0.16 & 0.02 pg/ pg total RNA ( n = 39)) that is consistent with its low NPY content, previously reported (28). Treatment of cells with the potent glucocorticoid dexamethasone (1 p~) elicited a gradual elevation to 2.1-2.6 times the control between 24 and 96 h (Fig. 5). The half-maximal and maximal effects were achieved with 1.7 X lo-' M and M dexamethasone, respectively (not shown). As shown in Fig. 5, treatment of cells with 10 p~ forskolin, a diterpene activator of adenylate cyclase, elicited a rapid but transient increase in the NPY mRNA level, which was maximally 2.7 times the control at 8 h. The response to forskolin was apparently desensitized by 72 h.
The response to dexamethasone + forskolin at early times equaled that by forskolin alone, but at 48 and 72 h was greater than the effects of each drug alone. The results suggest that the rat NPY mRNA abundance can be positively regulated by cAMP and glucocorticoids, acting rapidly ( 5 4 h) and slowly (>12 h), respectively.
Untreated N18TG-2 mouse neuroblastoma cells (35) also possess a low level of NPY mRNA (0.13 & 0.02 pg/pg RNA ( n = 5)). As shown in Fig. 6, this level was only slightly increased by dexamethasone or forskolin alone during 6-48 h of treatment, but the combination of both drugs produced marked elevations of up to 6.7 times the control. These results demonstrate that NPY mRNA in mouse cells is positively regulated by both glucocorticoids and cAMP elevation, with greater synergism than with PC12 cells under the conditions employed.

Regulation of NPY Gene Expression by Combinations of
Phorbol Ester, Calcium Zonophore, Glucocorticoid, and CAMP-To elucidate the possible effects of calcium-depend-

TABLE I
The distribution of NPY mRNA and NPY immunoreactivity in rat tissues A, the NPY mRNA contents of total RNA preparations from various brain regions or tissues were determined by dot-blot and Northern blot hybridization analyses as described under "Experimental Procedures." The data are expressed as means of a t least two determinations. Standard errors indicate variation in means obtained with two RNA preparations. B, the absolute amounts of NPY immunoreactivity (first column) were determined by radioimmunoassay as described under "Experimental Procedures." The data are means f S.E. obtained from the number of samples shown in parentheses. Each sample was an extract from tissue obtained from a single rat. NT, not tested.  h. At the indicated times, cells were harvested and RNA was purified. Abundances of NPY mRNA (-800 bases) were determined by Northern blot analysis of total RNA (40 pg) and are expressed as percent of the apparent absolute abundance at zero time, which was estimated to be 0.21 pg/pg RNA. The unexpected decline at 96 h in response to dexamethasone + forskolin was found to be reproducible. forskolin, and dexamethasone were tested (Table 11, experi- 14 ment 1). PMA or A23187 alone had little or no effect on the NPY mRNA abundance, while together they elicited moderate stimulation in the presence of dexamethasone. More strikingly, the combination of forskolin + PMA elicited large elevations (20-70-fold) that were further potentiated (up to approximately 200-fold) by A23187 and/or dexamethasone. The half-maximally and maximally effective concentrations of PMA were 6 and 40 nM, respectively, in the presence of forskolin (not shown). 40-Phorbol 12,13-didecanoate, which is inactive in protein kinase C activation in contrast to 4pphorbol esters such as PMA, was inactive in increasing the NPY mRNA abundance in the presence of forskolin (Table  11, experiment 2).
(The former derivative was less active than the latter or forskolin in both this effect and the production of short processes extended from the cell bodies.) The combination of either cAMP derivative with PMA elicited much higher elevations than the sum of effects by single compounds. Addition of the calcium ionophore further potentiated the effects of cAMP derivatives + PMA. These results are consistent with those obtained with forskolin (experiments 1 and 2). 8-Bromo-cGMP was less effective than 8-bromo-CAMP and did not significantly synergize with PMA.
The time course (Fig. 7) of the effects of various combinations of inducers indicates that the apparent half-maximal effects of forskolin + PMA f A23187 occurred at approximately 6 h of treatment. (The maximal stimulation was less than that in other experiments, probably because of a higher basal level of NPY mRNA, but qualitatively the effects of the regulators are identical.) The results of Table I1 and Fig. 7 indicate a strong and synergistic positive regulation of NPY gene expression by concomitant activations of protein kinase C and CAMP-dependent protein kinase. Glucocorticoid appears to exert a weaker, probably permissive, action. The calcium ionophore A23187 also exerts a modest potentiating action, probably by facilitating the action of PMA. Qualitatively similar results were also obtained with NPY mRNA levels in SK-N-MC human neuroblastoma cells: suggesting that these conclusions are applicable to human as well as rat NPY gene expression.

DISCUSSION
In this study we have attempted to lay the groundwork for investigations of the regulation of NPY gene expression in a variety of experimental systems. Furthermore, we show that NPY gene regulation appears to be controlled more strongly by the synergistic actions of multiple regulators (CAMP, protein kinase C activation, and glucocorticoids) than by single regulators.
The rat prepro-NPY cDNA sequence reported here is in agreement with reports of Allen et al. (37) and Larhammar et al. (38) published after the initial submission of this paper. While the high conservation of NPY sequences in human and rat species was expected, it is especially interesting that the S. L. Sabol, preliminary results.
sequences of the 30-residue COOH-terminal peptide, which has no reported biological or pharmacological activity, are highly conserved (93%, two divergent residues at positions 87 and 96). This degree of conservation between human and rat sequences is typical of that for polypeptide hormones such as the A and B chains of insulin (90-95%) or corticotropin (95%) and is higher than that of such sequences as insulin C-peptide (71-74%) and signal peptides (70-85%), which are apparently important mainly for precursor biosynthesis and processing ( Ref. 39 and references therein). Therefore, we propose that the prepro-NPY COOH-terminal sequence or a fragment thereof has an important biological function.
The abundances of NPY mRNA in brain regions (Table I) are generally parallel to those of NPY peptide, even though some NPY neurons possess cell bodies (containing mRNA) and terminals (containing stored NPY) in different brain regions. As expected, NPY mRNA is highly abundant in the striatum and cerebral cortex, where numerous intrinsic NPYcontaining neurons are found (12,21,22). The hypothalamus is the only region having a major discrepancy between relative NPY mRNA and NPY peptide levels, suggesting that much of the stored NPY in hypothalamus is from extrinsic rather than intrinsic neurons and/or that NPY turnover is relatively low in hypothalamus. The significant amounts of NPY mRNA in rat heart and lung may reflect the innervation of these organs by intrinsic neurons containing NPY (40, 41). Glucocorticoids positively regulate several, perhaps many, genes governing neurotransmitter and neuropeptide biosynthesis such as tyrosine hydroxylase (42) and proenkephalin genes (43, 44). During rat development glucocorticoids can direct the differentiation of bipotential neural crest cells toward the chromaffin cell phenotype (45). In the present study we demonstrate positive regulation of NPY mRNA levels by glucocorticoids in three rodent cell lines. The steroid specificity of this response reflects a glucocorticoid receptormediated response (Fig. 8). These results are consistent with a previous report showing a 2-fold increase in NPY peptide content in NG108-15 cells treated with dexamethasone (27).
Cyclic AMP elevation stimulates the transcription of several neuropeptide genes (46-49) and tyrosine hydroxylase (42). The actions of cAMP and glucocorticoids are sometimes synergistic, as in the case of proenkephalin gene transcription (49). In the present study of effects on the NPY mRNA abundance, we found a variable synergism between cAMP and glucocorticoids in PC12 and N18TG-2 cells (Figs. 5 and 6). However, we found a much more powerful synergism between cAMP and PMA (Table 11, Fig. 7). A similar synergism between these two compounds, although smaller in magnitude, was previous reported for the regulation of vasoactive intestinal polypeptide precursor mRNA levels in human neuroblastoma cells (50). Another possible example of such synergism is a cAMP regulatory element of the human proenkephalin gene that responds to phorbol ester only in the presence of elevated cyclic nucleotide levels (46). The elevations of the NPY mRNA abundance elicited by forskolin + PMA are of sufficient magnitude to involve, most likely, transcriptional activation.
Regulatory elements required for either CAMPor phorbol ester-stimulated transcription have been identified for several genes, and consensus sequences for upstream regulatory element cores have recently been proposed for cAMP (46-48,51, 52) and phorbol ester (53, 54) regulation. In examining the 5"flanking region of the recently reported sequence of the rat NPY gene (38) for the presence of these motifs, we noted a 28-base imperfect palindrome (bases -88 to -65, GGGAGT-CACCCGGGCGTGACTGCC), which is well conserved in the

human NPY gene (24), that contains two sequences (-87 to -80 in the upper strand and -66 to -73 in the lower strand) that resemble the phorbol ester motif (T(G/T)AGTCA(G/C)) and to a lesser extent the cAMP motif ((T/G)ACGTCAG). Phorbol ester regulatory elements can bind the transcription factor AP-1 (53, 54), while the cAMP regulatory element binds an apparently similar or identical factor (52).
The interaction of the cAMP and protein kinase C regulatory systems reported here is an example of "monodirectional control" discussed by Nishizuka (36). However, the exact point(s) of interaction of the systems with respect to NPY gene expression are unknown. Several mechanisms of crosstalk between the two systems are known (reviewed in Refs. 36 and 55) and may be relevant to this study. However, a more interesting possible mechanism would be activation of one or more nuclear transcription factorb), such as AP-1 or factor(s) interacting with AP-1, by phosphorylation by both types of protein kinases at different sites on the same factor or on different factors. Cooperation of this nature among multiple regulators may be a feature of many genes.
Our results indicate that the regulation of the NPY gene is complex and interesting. Unraveling the mechanisms involved should shed light on general aspects of gene regulation. ethanol. Cells exposed to the stated concentratma of these compounds appeared morphaloglcally deqatured in formaldehyde-farmamide, diluted svith 0.9 M NaCI-0.09 M sodium citrate. and filtered slowly onto a Gene %em Plus membrane (New England Nuclear) within a dotblat manifold. For standardization, at leaat 5 quantities (2.5-100 pg) ofpBL-NPYl twnscnpts were similarly filtered with eamgr rat liver V A (3 bg). The membranes were hybridized with nick-trsnslated rNPY2 cDNA. Aytorsdiographic cignalri were quantltated by densitometry and corrected for the difference in size between the natural mRNA (800 h a w ) and pBLNPYl transcripts (544 bases) and for pBL-NPYl transmipt purity. The NFY mRNA abundance8 in other rqions of rat brain. and N18TG-2 and PC12 cells were determined by Northern-blot analyois and comparison of autoradiograpluc si&e with those of a welectrophoresed diquat of a standardized rat etristum RNA preparation.
Dot.blot analysis WBB found tu be insumnently sensitive for guantitption of NPY mRNA in untrpstedPC12 cells.
b a y Of NPY. Brains of adult male Sprague-Dawley rats (200 g), which we= dissected by the method of Glowinski and Iverqen (59), and peripheral organs were extracted as described previqqsly (8). The NPY content was deternuned by radioimmunoassay, as described previously (8). The antiserum, which recognizes either internal or ammo-terminal epltape(s) of NPY, peptide YY and <0.004% erges-reactivity with human or avian pancreatic polypeptide. Phe-Met-exhibited the same sensitivity to human and porcine NPY and had 8% moas-reactivity with porcine Arg-Phe-NHa, a-melanocyte stimulatmg hormone, OF substance P peptides.

RESULTS
human NPY eDNA probe nelded SLX posltwe clones, depxted ~n Rg. 1 Four of these (colleet~vely Cloning of rat NPY cDNA The screerung of two rat cDNA llbranes I" the Wll vector w t h B termed rNPYl), from a b r a n eDNA hbrary, appeared ~dent~eal by restnctlon analym and DNA sequenctng and had an unexpectedly large insert of 1,078 bp. Another elone (rNPY2) f~o m the same llbrary had an tnsert of 511 bp. An add~t~onal Independent clone (rNPY3) from a rat hypothalamic cDNA library had an Insert of 149 bp. The three independent clones were found to contain resons of dentl leal sequence 8s shown, although rNPYl eontamed an addttional 5' flsnkmg 583 bp. Because the rNPYl Insert hybndized w t h two rat brain mRNA epeeles havmg d i l k e n t tmsue dlstnbutlons. we concluded that most of the sddltional 5 sequence of the rNPYl xnsert (shown 85 a lagged h e m Fig. 1) is eDNA from an unrelated gene Itgated to NPY cDNA dunng the eonstructmn of the hbrary. Therefore the NPY eDNA sequence reported ~n th13 study IS hrmted to that of the rNPY2 ~nsert. No dlscrepannes were bund ~n nucleotide sequences common to the three clones . " ~ bran resons, as shown for rtnatum m Fig. 4, domonstrsted that NPY-immunoreactmty eluted ~n HPLC analysis of NPY immunoreactivity in r a t brain. HPLC analyses of extracts of each rat a angle peak having a retentLon ttme ~dent~cal to that ofsynthetic human NPY but slightly shorter than that of poreme NPY, whleh has a Leu ~n place ofMet46. This result 1s consmtent ulth the findmg (Rg. 2) that rat NPY 1s sdent~eal to human rather than to porem NPY. The HPLC profiles also show that the data comerrung NPY rmmunoreactmty ~n b r a n (Table I)  HI'LC analvsm gf NPY-~mrnunorwctw&y ~n rat stnaturn An aeetlc scld ethanol extract of strratum t m u o was lyophlllzed. red~ssolved zn water. and passed through a n octsdecyls~lyl shca cartndge iSep-Pak C18, Waters Assoaates) ~n water Peptides were eluted wtth 4 ml 60% aeetamtnle eontatmng 0 1% tnfluoroacetx a c d The samples were Iyophhed, dmolved I" 0 2 ml water and eentnfuged at 15.000 x g for 10 mm. The supernatant fraetlon (0.1 ml eantamrng 1 9 pmolos NPY Immunoreaetwlty) was applled to a reverse-phase BIO-SII ODS 10 column (4 x 250 mm). The column was developed Hnth a h e a r gradtent of 20.60% aeetomtnlo (AcCNI in 0 1% trifluoroaeetx scld over a period of 60 mm at a flow rate of 1 rnllmin. One ml fractmns were collected, lyophrliaed, reeonstltuted ~n water and radmmmunoassayed to mdleate elutmn positions ofsynthetx human and porcine NPY and porcine pepttde YY (PYY).
determmed NPY content. The recovery of standard NPY on the HPLC column WBB 91%. Arrows pattern of regulation by cAMP was obtained wlth NG108-15 mouse neuroblastoma x rat glloma tiBects of glncocorticoids and CAMP on Npy -A levels inNGl0e-15 hybrid cells A &ffelent hybnd cells. compared to PC12 and N18TG-2 cells. The hybnd cells. whxh have been extenswely studled mth respect to the mduetlon ofneuronal properties by prolonged cAMP elevatm (611, were previously shown to Contam NPY peptide (27). We found that untreated NG108-15 cells contam blot analysis of NG108-15 DNA revealed m a n l y the m o u e rather than the rat NPY gene (not remarkably high amounts of 800-nueleotlde NPY mRNA (11.4 k 1 5 pglpg RNA ( " 5 ) ) . Southern shown). The parental N18TG 2 neuroblastoma and CGRU-1 gltoma cells contam NPY mRNA I" low (FIE 6) and neghpble amounts. respectwely. In NGlOR-15 hybrld cells. the response to glueomrtimids resembles that in PC12 cells: however. forakolin treatment rapidly and persistently depressed the NPY mRNA abundance by 80.85% (Fig. 8). The half-maximal reductmn due to forskalin occurred a t approximately 12 h (not shown). NPY mRNA was also reduced by 1 mM 8. bromo-cAMP or 10 pM prostaglandin E l (which activates adenylate cyclase). while 1 mM 8-bmnm cOMP had no effect (not shown). Other treatment. that differentista NG108-15 cella without primarily elevating CAMP levels. such a8 28 dimethyladfoxids and e m deprivation, also reduce the NPY mRNA abundlmncs. This mggestd that CAMP may be acting negatively in thew cella by II m e c b s m different h m that mediating positive regulation in PC12 and N18TG2 cells.  114eo~ymrtim~terone. E, beta-es!mdiol. P. progestemne. and T. WtmtemM.