Catecholamine secretory vesicle stimulus-transcription coupling in vivo. Demonstration by a novel transgenic promoter/photoprotein reporter and inhibition of secretion and transcription by the chromogranin A fragment catestatin.

Stimulation of chromaffin cell secretion in vitro triggers not only secretion but also resynthesis of just released catecholamines and chromogranin A, the precursor of the catecholamine release-inhibitory, nicotinic cholinergic antagonist peptide catestatin. Does stimulus-transcription coupling occur in vivo? And does catestatin antagonize secretion and transcription in vivo? To answer these questions, we employed a novel mouse strain harboring a chromogranin A promoter/firefly luciferase reporter transgene. Tissue-specific expression of the reporter was established by both luminescence and reverse transcription-PCR. Secretion and transcription in vivo were triggered by either direct nicotinic stimulation or vesicular transmitter depletion. Nicotinic blockade in vivo was attempted with either the classical antagonist chlorisondamine or the novel antagonist catestatin. Luciferase reporter expression was exquisitely sensitive over a large dynamic range, was specific for the transgenic animals, and paralleled typical neuroendocrine distribution of endogenous chromogranin A. Adrenal ontogeny revealed a rise of embryonic transgene expression until embryonal day 18, with an abrupt postnatal decline. Direct nicotinic stimulation of chromaffin cells caused catecholamine release and transgene transcription, each of which was nearly completely blocked by chlorisondamine. Similar adrenal results were obtained during vesicular catecholamine depletion. Both secretion and transcription were substantially blocked in the adrenal gland by catestatin. In brain and sympathetic nerve, stimulation of transcription was more modest, and reserpine responses were only incompletely blocked by chlorisondamine or catestatin, perhaps because of limited blood-brain barrier penetration by these cationic antagonists. Thus, nicotinic cholinergic stimulus-transcription coupling occurs in vivo and can be provoked either directly or indirectly (by vesicular transmitter depletion). Such coupling triggers the biosynthesis of chromogranin A, the precursor of catestatin. Catestatin itself blocks stimulation of both secretion and transcription in vivo. Thus, chromogranin A and its catestatin fragment may lie at the nexus of nicotinic cholinergic signaling in vivo.


Stimulation of chromaffin cell secretion in vitro triggers not only secretion but also resynthesis of just released catecholamines and chromogranin A, the precursor of the catecholamine release-inhibitory, nicotinic cholinergic antagonist peptide catestatin. Does stimulus-transcription coupling occur in vivo? And does catestatin antagonize secretion and transcription in vivo?
To answer these questions, we employed a novel mouse strain harboring a chromogranin A promoter/firefly luciferase reporter transgene. Tissue-specific expression of the reporter was established by both luminescence and reverse transcription-PCR. Secretion and transcription in vivo were triggered by either direct nicotinic stimulation or vesicular transmitter depletion. Nicotinic blockade in vivo was attempted with either the classical antagonist chlorisondamine or the novel antagonist catestatin. Luciferase reporter expression was exquisitely sensitive over a large dynamic range, was specific for the transgenic animals, and paralleled typical neuroendocrine distribution of endogenous chromogranin A. Adrenal ontogeny revealed a rise of embryonic transgene expression until embryonal day 18, with an abrupt postnatal decline. Direct nicotinic stimulation of chromaffin cells caused catecholamine release and transgene transcription, each of which was nearly completely blocked by chlorisondamine. Similar adrenal results were obtained during vesicular catecholamine depletion. Both secretion and transcription were substantially blocked in the adrenal gland by catestatin. In brain and sympathetic nerve, stimulation of transcription was more modest, and reserpine responses were only incompletely blocked by chlorisondamine or catestatin, perhaps because of limited blood-brain barrier penetration by these cationic antagonists. Thus, nicotinic cholinergic stimulus-transcription coupling occurs in vivo and can be provoked either directly or indirectly (by vesicular transmitter depletion).

Such coupling triggers the biosynthesis of chromogranin A, the precursor of catestatin. Catestatin itself blocks stimu-lation of both secretion and transcription in vivo. Thus, chromogranin A and its catestatin fragment may lie at the nexus of nicotinic cholinergic signaling in vivo.
Stimulation of the splanchnic (efferent, preganglionic sympathetic) nerves innervating the adrenal medulla and sympathetic postganglionic axons releases multiple neurotransmitters, including acetylcholine (acting on neuronal nicotinic cholinergic receptors), ATP, and chromogranin A, precursor of the catecholamine release inhibitory peptide "catestatin" (bovine chromogranin A 344 -364 ) (1).
When secretory stimuli (such as acetylcholine interacting with nicotinic cholinergic receptors) trigger transmitter release from chromaffin cells or sympathetic axons, is the resynthesis of just released transmitters also initiated by the secretory stimulus? We have characterized this process (sometimes called "stimulus-secretion-synthesis coupling" or "stimulustranscription coupling") in chromaffin cells in vitro (7)(8)(9) and established transcriptional activation of chromogranin A by nicotinic cholinergic (physiologic pathway) stimulation. The process occurs at the level of transcript initiation (7), requires particular elements in cis in the proximal promoter (7), and has well defined signal transduction pathways in trans (8,9). However, whether this nicotinic cholinergic stimulation of chromogranin A occurs in vivo is uncertain.
We therefore set out to explore whether chromaffin cell stimulus-transcription coupling (specifically nicotinic cholinergic transcriptional stimulation of chromogranin A) occurs in vivo. To test this possibility, we employed a transgenic strain in which a mouse chromogranin A 4.8-kbp proximal promoter drives the expression of firefly luciferase, an extraordinarily sensitive reporter of gene expression (10 -12). In the in vivo experiments reported here, we used nicotine (mimicking the autonomic ganglionic transmitter acetylcholine) and reserpine (indirect stimulation by vesicular depletion) to evoke cate-cholamine release and expression of the transgene in the adrenal gland, sympathetic nerve, and brain.
At the same time, we wished to test whether the catecholamine release-inhibitory fragment catestatin (1) functions in vivo. Our previous characterization of catestatin in chromaffin cells in vitro (1,(13)(14)(15)(16) documented its mechanism as a nicotinic cholinergic antagonist (17), thereby acting as an autocrine, negative feedback inhibitor of catecholamine release (1). However, it is not yet clear whether catestatin inhibits catecholamine release in vivo (i.e. in an intact organism).
Our in vivo findings confirm in vitro studies of nicotinic signaling to both catecholamine secretion and gene transcription, documenting the role of catestatin in both processes. The studies also provide a novel in vivo photoprobe to investigate stimulus-transcription coupling in experimental cardiovascular disease states.

MATERIALS AND METHODS
Transgenic Promoter/Reporter Construction-The mouse chromogranin A promoter (Ϫ4.8 kbp to ϩ42 bp) was identified by cDNA hybridization, restriction-mapped, excised from a genomic cosmid clone as an XhoI/HindIII fragment, and subcloned into the polylinker of the firefly luciferase reporter vector pXP1, as described previously (18,19). The promoter's cap (transcription initiation) site was established by primer extension, as described (18,19). Promoter position numbering is according to bases upstream (Ϫ) or downstream (ϩ) of the cap site. The translational start codon (ATG) in mouse chromogranin A begins 258 bases downstream of the cap site (18,19); thus, the promoter fragment used in these experiments is entirely upstream of the amino acidencoding region of the chromogranin A gene, and the first ATG downstream of the cap site encodes the initial Met of firefly luciferase.
Generation of Transgenic Mice-The mouse chromogranin A promoter/firefly luciferase reporter transgenic mouse was created by first digesting the above pXP1 mouse chromogranin A 4.8-kbp promoter/ firefly luciferase reporter construct with PvuI and BamHI, to linearize and excise the promoter/reporter transgene from the vector, and then purifying the linear transgene for microinjection using Sepharose gel separation, CsCl density gradient centrifugation, and dialysis against microinjection buffer (20). The transgene also contains an SV40 polyadenylation signal and SV40 3Ј-untranslated sequence, containing an SV40 small T intron (splice donor and acceptor sites) (Fig. 1A). The male pronucleus of fertilized eggs from the FVB inbred strain (Taconic Farms Inc., Germantown, NY) was used for microinjection of the transgene. The presence of the transgene was evaluated by either Southern blot or polymerase chain reaction. The transgenic animals were then inbred to homozygosity.
First strand cDNA was prepared from 500 ng of total RNA template by reverse transcription (using Omniscript™ and Sensiscript™ reverse transcriptases) at 54°C for 30 min, followed by PCR. The PCR protocol began with a 95°C/15-min step (for simultaneous inactivation of the reverse transcriptases and activation of the HotstarTaq™ DNA polymerase), followed by a three-step amplification profile (94°C denaturing step for 30 s, 55°C annealing step for 30 s, and 72°C extension step for 1 min) for 25 cycles, 72°C for 10 min, and finally holding at 4°C. The reaction mixture was composed of 10 l of 5ϫ RT-PCR buffer, 2 l of dNTP mix (each at 10 mM), 2 l each of forward and reverse primers (each at 10 M), 2 l of RT-PCR enzyme mix, 0.2 l of RNasin® ribonuclease inhibitor (at 40 units/l; Promega), and 5-15 l (depending on the concentration) of the template RNA and RNase-free water to achieve the final desired total volume (50 l/amplification).
After PCR, the products were visualized/photographed on 1.5% agarose gels by ethidium bromide staining/310-nm UV fluorescence. PCR products were then purified with the QIAquick PCR purification kit (Qiagen, Valencia, CA), and amplification of the correct targets was verified by DNA sequencing.
As a negative control, when RNA was pretreated with RNase A (Qiagen), no product in the RT-PCR assay was detected after gel electrophoresis. As a second negative control, no PCR product was obtained when water was included instead of RNA samples in the reaction mixture.
Drug Treatments and Tissue Harvesting-The drugs used were nicotine (2.5 mg/kg intraperitoneally; Sigma), the vesicular monoamine transporter inhibitor reserpine (2.5 mg/kg intraperitoneally; Sigma), the neuronal nicotinic cholinergic antagonist chlorisondamine (5 mg/kg intraperitoneally; Tocris-Cookson), the novel nicotinic antagonist catestatin (20 nmol/25-g mouse intraperitoneally), or vehicle. Catestatin (bovine chromogranin A 344 -358 ; RSMRLSFRARGYGFR) was synthesized and then purified to Ͼ95% homogeneity by RP-HPLC; the identity A n , polyadenylation signal. Distances are not drawn to scale. B and C, sensitivity and dynamic range of the chromogranin A promoter/luciferase reporter transgene assay. Results for brain and adrenal gland homogenates (mean Ϯ S.E.) are shown as log 10 /log 10 plots. The amount of tissue homogenate protein (in typical 100-l assay aliquots) is shown on the horizontal axis. and purity of the final product were verified by re-HPLC and matrixassisted laser desorption ionization mass spectrometry. The catestatin dose (20 nmol/25-g mouse) was calculated so as to achieve a concentration of ϳ4 M in the extracellular space (estimated as 20% of body weight).
In some studies, stimulation of sympathetic outflow was undertaken (directly by nicotine; indirectly by reserpine or vehicle). In dose-and time-dependent studies, reserpine at several doses (1, 5, and 10 mg/kg intraperitoneally) versus mock was injected for 4 and 18 h. In other studies, animals were first pretreated with either a nicotinic cholinergic antagonist (chlorisondamine (5 mg/kg intraperitoneally) or catestatin (20 nmol/25 g intraperitoneally)) or vehicle. 30 min later, the sympathetic stimulant was administered: nicotine (2.5 mg/kg intraperitoneally), reserpine (2.5 mg/kg intraperitoneally), or vehicle.
For studies of the effects of the stimulants on plasma catecholamines, mice were treated with drugs (versus vehicle) as noted above, and 30 min later they were anesthetized intraperitoneally with a rodent anesthesia mixture (ketamine (60 mg/kg of body weight); xylazine (6.4 mg/kg of body weight); acepromazine (1.2 mg/kg of body weight)). After anesthesia was achieved (1-2 min), blood was then collected from the left ventricle into Eppendorf tubes containing acid/citrate/dextrose as anticoagulant. Blood was kept on ice and centrifuged, and the plasma stored at Ϫ70°C until the assay. Plasma catecholamines were assayed by a sensitive radioenzymatic method (21).
In studies of transcriptional activation of the chromogranin A/luciferase transgene, tissues were obtained from control or drug-treated mice after sacrifice by cervical dislocation, typically 16 h after administration of the drug (nicotine, reserpine, or vehicle) perturbing chromogranin A biosynthesis.
Quantitative Assay of the Chromogranin A/Luciferase Transgene Enzymatic Activity-Freshly dissected tissue samples were collected in 500 l of ice-cold lysis buffer (0.1 M potassium phosphate buffer, pH 7.8, 1 mM dithiothreitol, 0.1% Triton X-100) and briefly (15-30 s, depending on the volume of the tissue) homogenized by a Tissuemizer homogenizer with a TR-10 power control device (Tekmar, Cincinnati, OH) set at 50% power output. Typical buffer homogenization volumes were 0.25 ml/ adrenal gland, and 1.0 ml/brain. The homogenate was centrifuged twice at 14,000 rpm for 30 min at 4°C, collecting the supernatant in a fresh microcentrifuge tube after the first spin.
100 l of the clear lysates were then placed in clear plastic polystyrene 12 ϫ 47-mm tubes in an ultrasensitive, low noise luminometer with a 12-watt Peltier-cooled (8°C) photomultiplier tube, a dark count rate of Ͻ100 counts/s (even at 20°C), spectral sensitivity of 390 -620 nm, 20-ns resolution, and quantum efficiency of 24% (AutoLumat LB 953, EG&G Berthold, Bad Wildbad, Germany) with AutoLumat-PC-Control software in a personal computer running DOS on an Intel Celeron RAM chip. 100 l of the luciferase assay buffer (final concentrations: 100 mM Tris acetate, pH 7.8, 10 mM magnesium acetate, 1 mM EDTA, 0.1 mM luciferin substrate, 3 mM ATP cofactor) were injected to each tube, and flash luminescence was recorded for 10 s and saved in a Kaleidagraph spreadsheet in a personal computer. 3-10 l (depending on the tissue type) of the tissue lysates were used for total protein measurement (22), by a Coomassie Brilliant Blue dye binding assay reagent (Bio-Rad). Luciferase activities in the various tissues were normalized to protein concentration and expressed as mean relative light units (RLU) Ϯ S.E. One RLU represents 10 pulses released by light quanta from the photon counter. Assay blanks were obtained by measuring luciferase activity in buffer alone (without tissue homogenate).
Data Analysis and Statistics-Data are reported as the mean value Ϯ one S.E. When only two conditions (e.g. control and experimental) were compared, the data were evaluated by unpaired t tests. When multiple conditions were compared, we used one-way analysis of variance, followed by the Dunnett multiple comparison post hoc test, if appropriate. Statistical significance was concluded at p Ͻ 0.05. Statistics were computed with the programs InStat (GraphPad Software, San Diego, CA), SPSS (Statistical Package for the Social Sciences, Chicago, IL), or Kaleidagraph (Synergy/Abelbeck Software, Reading, PA).

FIG. 2.
Neuroendocrine tissue-specific expression of the chromogranin A promoter/luciferase reporter transgene. A, transgene luciferase expression is specific to mice bearing the construct. Adrenal glands from transgenic and nontransgenic (strain C57BL6) mice were dissected and homogenized in luciferase lysis buffer, and the lysates were assayed in a Berthold luminometer for luciferase activity. Results (mean Ϯ S.E.) are expressed as RLU/g of protein. n ϭ 6 males/group were studied at age 70 -90 days. B, distribution of transgene expression in mouse tissues: endocrine, neuronal, and control. Results are expressed as RLU/g of protein. n ϭ 4 males/group were studied at age 70 -90 days. Ovaries were from n ϭ 4 females.

FIG. 3. Neuroendocrine tissue-specific expression of the chromogranin A promoter: endogenous (chromogranin A) versus transgenic (luciferase) mRNAs.
Total RNA was extracted from freshly dissected tissues, and one-step RT-PCR was done using mouse chromogranin A, luciferase, and cyclophilin gene-specific primers (see "Materials and Methods"). First strand cDNA was prepared from total RNA template by reverse transcription (using Omniscript™ and Sen-siscript™ reverse transcriptases) followed by PCR. PCR products were purified, and amplification of the correct target genes was verified by DNA sequencing. A representative image is shown; bp indicates the size of the specific RT-PCR products for each target.

Neuroendocrine Tissue-specific Expression of the Chromogranin A Promoter/Luciferase Transgene: Luminescence Enzymatic Activity
The chromogranin A promoter/luciferase reporter transgene (Fig. 1A) was an extraordinarily sensitive probe of promoter activity (Fig. 1B); the dynamic range of the system extended over 5 log 10 orders of magnitude, and tissue samples could be diluted over 10 5 -fold (down to nanogram quantities of tissue protein) without reaching the lower limit of detection (Fig. 1B).
Assay of chromogranin A/luciferase in the adrenal gland of transgenic mice showed luciferase expression more than ϳ1000 times background (nonspecific) luminescence in control adrenal glands ( Fig. 2A).
In agreement with the typical cell type-specific pattern of expression of the endogenous chromogranin A gene (23), we found that the chromogranin A-luciferase transgene was expressed in endocrine (adrenal Ͼ Ͼ pituitary Ͼ testis Ͼ ovary Ͼ pancreas) and neuronal tissues (brain stem, hippocampus, frontal cortex, thalamus, olfactory bulb Ͼ striatum, hypothalamus, cerebellum) but not in control tissues (liver, spleen) (Fig. 2B).

Tissue Distribution of Transgene and Endogenous Chromogranin A Transcripts: RT-PCR
To directly compare expression of the two forms of the chromogranin A promoter (endogenous gene versus transgenic), we employed RT-PCR in the same tissues (Fig. 3).
Consistent with the luminescence data ( Fig. 2B), both the endogenous chromogranin A gene (chromogranin A mRNA) and the chromogranin A/luciferase transgene (luciferase mRNA) were expressed in endocrine (adrenal Ͼ gut Ͼ pituitary Ͼ testis, ovary) and neuronal tissues (brain Ͼ vas deferens). The endogenous and transgenic mRNAs displayed similar rank orders of expression in these neuroendocrine sites. There was no detectable expression of either endogenous or transgenic mRNAs in control tissues (skeletal muscle, liver, spleen, heart). Comparability of mRNA load per lane was confirmed by RT-PCR of a "housekeeping" transcript, cyclophilin.

Expression of the Chromogranin A-Luciferase Transgene during Ontogenetic Development: Brain and Adrenal Gland
Previous in situ hybridization studies on the ontogeny (day E16 to P9) of chromogranin A mRNA in rat brain revealed expression in the intermediate cortical layer and dentate gyrus by day E16 (24). In the present study, we detected expression of the chromogranin A/luciferase transgene as early as day E6 (Fig.  4A). A steady increase in expression of the transgene in brain was seen up to day E18, with a subsequent abrupt decline after birth (day P1) and then a further slow decline up to day P30, after which expression stabilized up to day P360 (Fig. 4A). In contrast, FIG. 4. Neuroendocrine transgene expression as a function of age. Ontogenetic and age-dependent expression of the chromogranin A/luciferase transgene. Results (RLU/g of protein) are expressed as mean Ϯ S.E. n ϭ 10 embryonic, postnatal (days P1-P9; both sexes) or n ϭ 6 postnatal (days P30 -P360; all male) animals were studied at each time point. Embryonic mice were dissected from the uteri of females with established time of conception. A, whole embryo/brain. Either the whole embryo (days E6, E8, and E10) or the whole brain (days E12, E14, E16, and E18) was dissected from embryonic mice. The whole brain was dissected from postnatal (days P1-P30) and adult (days P60 -P360) mice. B, adrenal gland. Adrenal glands were dissected from embryonic (day E18), postnatal (days P1-P24), and adult (days P60 -P360) mice.
FIG. 5. Catecholamine release by direct nicotinic cholinergic stimulation and blockade by nicotinic cholinergic antagonists, including catestatin. The sympathoadrenal system was directly activated by the nicotinic cholinergic agonist nicotine (2.5 mg/kg intraperitoneally) versus vehicle (mock). To probe the role of nicotinic cholinergic receptors in secretion after nicotine, animals were pretreated 30 min prior to nicotine with nicotinic cholinergic antagonists, either the classical antagonist chlorisondamine (5 mg/kg intraperitoneally) or the novel antagonist (and chromogranin A fragment) catestatin (20 nmol/25 g intraperitoneally; designed to achieve an extracellular target concentration of ϳ4 M), or vehicle (mock). In each experiment, n ϭ 6 males were studied, at age 60 -70 days. 30 min after nicotine (or vehicle), animals were anesthetized (ketamine, 60 mg/kg of body weight; xylazine, 6.4 mg/kg of body weight; acepromazine, 1.2 mg/kg of body weight), and blood was collected for plasma catecholamine determination (see "Materials and Methods"). Results are shown as mean Ϯ S.E. expression of the transgene in the adrenal gland displayed a substantial increment from day E18 to P1 but was followed thereafter by only ϳ20% fluctuations up to day P360 (Fig. 4B).

Direct Nicotinic Cholinergic Stimulation of the Sympathoadrenal System: Acute Transmitter Release, Subsequent Adrenal Stimulus-Transcription Coupling, and Role of Catestatin in Vivo
Acute Transmitter Release-Direct activation of nicotinic cholinergic receptors by nicotine caused acute (30 min) ϳ2.7fold release of catecholamines (norepinephrine and epinephrine) from storage vesicles into the bloodstream (Fig. 5). When animals were pretreated with the classical neuronal nicotinic antagonist chlorisondamine, the catecholamine increment was blunted by Ͼ80%, confirming specific mediation of the response by nicotinic receptors.
Stimulus-Transcription Coupling-Chromogranin A (25) and its catestatin fragment (14) are co-released by exocytosis with catecholamines. In cultured chromaffin cells in vitro (7-9), exocytotic stimuli also program the resynthesis of just released catecholamine storage vesicle contents, a process known as "stimulus-secretion-synthesis coupling" or "stimulus-transcription coupling." Does this phenomenon occur in vivo? Treatment with nicotine resulted in a ϳ2-fold increment in expression of the chromogranin A/luciferase transgene, confirming the phenomenon in vivo, and the increment was Ͼ80% inhibited by the classical nicotinic cholinergic antagonist chlorisondamine (Fig. 6A).
Catestatin in Vivo-The chromogranin A fragment catestatin (1), a novel nicotinic cholinergic antagonist stored and released with catecholamines by exocytosis (14), also blocked the catecholamine secretory response to nicotine by Ͼ80% (Fig. 5), indicating that this peptide functions as a nicotinic antagonist upon secretion in vivo, extending its role as an autocrine, negative feedback inhibitor of catecholamine release (1). Furthermore, catestatin also blocked the transcriptional response to nicotine by Ͼ90% (Fig. 6A), establishing an entirely new role for the peptide on gene expression in vivo.

Stimulus-Transcription Coupling in Brain after Direct Nicotinic Cholinergic Stimulation
Besides stimulation of neurotransmitter release, nicotine triggers expression of genes encoding enzymes involved in neurotransmitter synthesis such as tyrosine hydroxylase (26), neurotransmitter transporters such as the vesicular acetylcholine transporter (27), neuropeptides such as neuropeptide Y (28), and transcription factors such as c-Fos and the cAMP-response element-binding protein (29).

FIG. 6. Stimulus-transcription coupling after direct nicotinic cholinergic stimulation and blockade by nicotinic cholinergic
antagonists, including catestatin. The sympathoadrenal system was directly activated by the nicotinic cholinergic agonist nicotine (2.5 mg/kg intraperitoneally) versus vehicle (mock). To probe the role of nicotinic cholinergic receptors in transcription after nicotine, animals were pretreated 30 min prior to nicotine with nicotinic cholinergic antagonists, either the classical antagonist chlorisondamine (5 mg/kg intraperitoneally) or the novel antagonist (an chromogranin A fragment) catestatin (20 nmol/25 g intraperitoneally; designed to achieve an extracellular target concentration of ϳ4 M), or vehicle (mock). In each experiment, n ϭ 6 males were studied, at age 60 -70 days. 16 h after nicotine (or vehicle), animals were sacrificed to harvest adrenal glands and brains for transgene (chromogranin A promoter/luciferase reporter) and protein assays. The results (mean Ϯ S.E.) obtained in the adrenal glands (A) and brains (B) were expressed as RLU/g of protein.
FIG. 7. Indirect stimulation of the sympathoadrenal system by vesicular transmitter depletion: acute transmitter release and blockade by nicotinic cholinergic antagonists, including catestatin. The sympathoadrenal system was indirectly activated by the VMAT inhibitor reserpine (2.5 mg/kg intraperitoneally) versus vehicle (mock). To probe the role of nicotinic cholinergic receptors in secretion after reserpine, animals were pretreated 30 min prior to reserpine with nicotinic cholinergic antagonists, either the classical antagonist chlorisondamine (5 mg/kg intraperitoneally) or the novel antagonist (and chromogranin A fragment) catestatin (20 nmol/25 g intraperitoneally; designed to achieve an extracellular target concentration of ϳ4 M), or vehicle (mock). In each experiment, n ϭ 6 males were studied, at age 70 -80 days. 30 min after reserpine (or vehicle), animals were anesthetized (ketamine, 60 mg/kg of body weight; xylazine, 6.4 mg/kg of body weight; acepromazine, 1.2 mg/kg of body weight), and blood was collected for plasma catecholamine determination (see "Materials and Methods"). Results are shown as mean Ϯ S.E.
Nicotinic stimulation elevated brain expression of the chromogranin A transgene by ϳ26% (Fig. 6B). This increment was blocked ϳ70% by the classical nicotinic antagonist chlorisondamine and ϳ80% by the novel nicotinic antagonist catestatin.

Indirect Stimulation of the Sympathoadrenal System by Vesicular Depletion: Initial Transmitter Release, Subsequent Adrenal Stimulus-Transcription Coupling, and Role of Catestatin in Vivo
Initial Transmitter Release and Catestatin-Vesicular monoamine transporter (VMAT) inhibition by reserpine causes an initial release of stored catecholamines in vitro (30) and in vivo (31), followed by a prolonged inhibition of catecholamine uptake into storage granules in the adrenal medulla (32) and noradrenergic nerves (33,34). Treatment with reserpine ( Fig.  7) caused acute catecholamine secretion, to plasma levels ϳ30% greater than basal. Pretreatment with the classical nicotinic cholinergic antagonist chlorisondamine completely blocked the acute secretory response (Fig. 7), suggesting that the acute secretion might be, at least in part, mediated by reflex increments in efferent preganglionic sympathetic (splanchnic) nerve traffic (35,36).
Catestatin in Vivo-Catestatin not only completely blocked depletion-induced increments of plasma catecholamines (Fig.  7) but also diminished by Ͼ60% the chromogranin A transgene activation after reserpine (Fig. 8B). Thus, catestatin in vivo seems to exert nicotinic cholinergic antagonist activity on both secretory and transcriptional processes in the sympathoadrenal system.
In the present study, transmitter depletion caused time-and dose-dependent increments of transgene expression in the brain, with ϳ3.3-fold stimulation at 5 mg of reserpine/kg (Fig.  8C). The increment was blocked ϳ30% by the classical nicotinic antagonist chlorisondamine and ϳ35% by the novel nicotinic antagonist catestatin (Fig. 8D).

Stimulus-Transcription Coupling in Peripheral Postganglionic Sympathetic Axons (vas Deferens)
The male vas deferens is a rich source of postganglionic sympathetic nerve terminals (25,31). Contraction of smooth muscle in the vas deferens is elicited by sympathetic nerves co-releasing the neurotransmitters ATP and norepinephrine from large dense core vesicles (42,43). Reserpine depletes catecholamines from these vesicles (44). Since nicotinic cholinergic receptors are expressed on vas deferens nerve terminals (45), nicotinic cholinergic stimulation should evoke catecholamine and chromogranin A co-release from these same vesicles.
In these studies of the vas deferens (Fig. 9), vesicular depletion by reserpine increased chromogranin A transgene expression by ϳ1.75-fold, whereas direct nicotinic cholinergic stimulation activated the transgene by ϳ1.5-fold.

DISCUSSION
Neuroendocrine Expression-Specific expression of the chromogranin A promoter/luciferase reporter was verified by its restriction only to mice bearing the transgenic construct ( Fig.  2A). Consistent with the reported neuroendocrine distribution of chromogranin A (2, 23, 46, 47), we found expression of the chromogranin A/luciferase transgene in endocrine and neuronal tissues but not in control tissues (Fig. 2B).
The 4.8-kbp mouse proximal chromogranin A promoter directs correct neuroendocrine expression of the gene in transfected neuroendocrine cells in vitro (19,47). Does this promoter region also direct appropriate expression in vivo? The rank orders of endogenous chromogranin A mRNA and transgenic luciferase mRNA expression were across cell types (Fig. 3), indicating that the 4.8-kbp mouse chromogranin A proximal promoter fragment contains information sufficient to direct correct neuroendocrine specific expression in vivo.
Neuroendocrine Ontogeny of Transgene Expression-Ontogenetic expression of chromogranin A has been explored previously in birds and mammals. In the chick, chromogranin A was detected in the carotid body (day E9) (54) and gizzard (day E12) (55). Chromogranin A mRNA was detected in rat brain at day E16 (24) and in rat enterochromaffin-like cells at day E18 (56). In the fetal pig, chromogranin A immunoreactivity was detected at days F24 -F27 in sympathetic ganglia, days F37-F42 in chromaffinoblasts, and days F54 -F56 in adrenal medullary cells, declining after day F76 and postnatally (57). In the human adrenal gland, chromogranin A was detected in chromaffin cell precursors by 9 weeks of gestation (58), whereas in the fetal lung chromogranin A was detected as early as 12-14 weeks (59). In the present study, detection of chromogranin A/luciferase transgene expression at day E6 may indicate an especially early role in embryonic development.

Stimulus-Transcription Coupling in Vivo and the Role of Nicotinic Cholinergic
Receptors-Nicotine is a powerful stimulant of the sympathoadrenal system, causing release of catecholamines from postganglionic sympathetic neurons and adrenal medulla (60). Nicotine also acts centrally to stimulate dopamine release from nigrostriatal and mesocorticolimbic neurons and norepinephrine release from hippocampus, cerebellum, and locus coeruleus neurons (61)(62)(63). In agreement with these findings as well as our previous in vitro studies (1,7), here we documented acute nicotinic stimulation of catecholamine release as reflected by ϳ2.7-fold increments in plasma catecholamine levels (Fig. 5).
Enkephalins (64) and catecholamine biosynthetic enzymes (tyrosine hydroxylase, dopamine ␤-hydroxylase, and phenylethanolamine-N-methyltransferase) are up-regulated in response to nicotinic cholinergic stimulation (7,29,66,67). Nicotine can also function centrally to activate tyrosine hydroxylase in the adrenal medulla, both transcriptionally and posttranslationally (67). Consistent with our in vitro studies of the transfected chromogranin A promoter in chromaffin cells (7-9), we documented ϳ2-fold direct nicotinic stimulation of chromogranin A/luciferase transgene expression in the adrenal gland (Fig. 6A), and the increment was blocked Ͼ80% by chlorisondamine. Three conclusions can be drawn: 1) the stimulustranscription mechanism for repletion of vesicular contents postexocytosis also functions in vivo; 2) the stimulus-transcription process is specifically triggered by the nicotinic cholinergic receptor; 3) a 4.8-kbp proximal chromogranin A promoter is sufficient to confer the nicotinic transcriptional response in vivo.
Reserpine was also used to explore the stimulus-transcrip- FIG. 9. Stimulus-transcription coupling in peripheral postganglionic sympathetic axons (vas deferens nerve termini). n ϭ 6 male mice/group, age 60 -80 days, were subjected to direct (nicotine, 2.5 mg/kg intraperitoneally) or indirect (vesicular depletion by reserpine, 2.5 mg/kg intraperitoneally) activation of the sympathoadrenal system. Control animals received vehicle (mock). 16 h later, the animals were sacrificed for vas deferens dissection, homogenization, and luciferase assay. Results (RLU/g of protein) are expressed as mean Ϯ S.E. tion process (Figs. 7-9). The mechanism of action of reserpine involves catecholamine store depletion from sympathetic nerve endings and chromaffin cells (33,34) through reversible competitive inhibition of VMATs (69). Previous reports revealed that splanchnic denervation diminished reserpine-induced adrenal medullary catecholamine secretion and depletion (70 -72), indicating efferent sympathetic neural mediation of reserpine action on chromaffin cell catecholamine release. In agreement with these reports, we found complete blockade of reserpine-stimulated catecholamine secretion by the classical nicotinic cholinergic antagonist chlorisondamine (Fig. 7). Ͼ4fold transcriptional activation of the chromogranin A/luciferase transgene was also inhibited Ͼ70% by chlorisondamine, documenting the participation of splanchnic nerve traffic in a reflex response (Fig. 8, A and B).
Thus, both nicotine and reserpine activate not only acute catecholamine release (Figs. 5 and 7) but also chronic stimulustranscription coupling (Figs. 6 and 8), albeit by rather different initial mechanisms. Nonetheless, each process involves an ob-ligate role for nicotinic cholinergic receptors: in the case of nicotine itself, a direct role (Figs. 5 and 6), and an indirect, reflex role in splanchnic neurotransmission in the case of vesicular depletion by reserpine (Figs. 7 and 8).
Stimulus-Transcription Coupling in Neurons and Brain-Both nicotine (Figs. 6 and 9) and reserpine (Figs. 8 and 9) provoked increments in chromogranin A/luciferase transgene expression, in both peripheral postganglionic sympathetic axons ( Fig. 9) and brain ( Figs. 6B and 8, C and D). Reserpine was more powerful than nicotine in stimulating the transgene, whether in the adrenal gland (Fig. 8, A and B), peripheral noradrenergic nerve terminals (Fig. 9), or brain (Fig. 8, C  and D).
In neurons, the relative degree of transgene activation by either nicotine (ϳ1.26-to ϳ1.5-fold) (Figs. 6 and 9) or reserpine (ϳ1.75-to ϳ3.3-fold) (Figs. 8, C and D, and 9) was quantitatively similar in brain (Figs. 6B and 8, C and D) and peripheral sympathetic nerves (Fig. 9). However, the degree of luciferase stimulation in neurons was substantially less than the corre- The nicotinic cholinergic agonist (acetylcholine surrogate) nicotine binds to the nicotinic cholinergic receptor, inducing extracellular Na ϩ influx, depolarizing the cell membrane, and causing influx of calcium through voltage-gated calcium channels. Influx of calcium induces both catecholamine release by exocytosis (all-or-none secretion) and chromogranin A gene transcription through a pathway (62-64) involving activation of protein kinase C (PKC) and mitogen activated protein kinase (MAPK). Catestatin is formed within and secreted from chromaffin granules and inhibits subsequent catecholamine secretion and chromogranin A gene transcription by acting as an endogenous nicotinic cholinergic antagonist. The VMAT inhibitor reserpine impairs (Ϫ) catecholamine storage by inhibiting uptake of dopamine or catecholamines into the chromaffin granules, resulting in acute nonexocytotic catecholamine release. By contrast, chronic reflex splanchnic sympathetic activation after catecholamine depletion by reserpine (ϩ) causes the release of acetylcholine, which in turn acts at neuronal/ganglionic nicotinic cholinergic receptors to induce both catecholamine secretion and chromogranin A gene transcription. NE, norepinephrine; ERK, extracellular signal-regulated kinase; CRE, cAMP-response element; CREB, cAMP-response element-binding protein.
sponding increment in the adrenal gland after either nicotine (ϳ2-fold) (Fig. 6A) or reserpine (ϳ4-fold) (Fig. 8, A and B). Impaired passage of nicotine or reserpine across the bloodbrain barrier (73) cannot easily explain the adrenal versus brain/neuron disparity in transgene activation, since vas deferens nerve termini are certainly outside the brain, and both nicotine and reserpine display at least some brain penetration (73), especially over a 16 -18-h time course. Reasons for these adrenal versus neuronal discrepancies in transgene response are not immediately apparent but might include the well known differences in the subunit composition of nicotinic receptors in the two sites (17) or the marked quantitative differences in catecholamine storage vesicle size, composition, and abundance in chromaffin cells versus noradrenergic nerves (25,31,74).
The ability of the classical nicotinic cholinergic antagonist chlorisondamine to block stimulus-transcription coupling (Fig.  6) also differed somewhat by site (adrenal versus brain) and by stimulus (nicotine versus reserpine). In the adrenal gland, both stimuli to transcription (nicotine (Fig. 6A) and reserpine (Fig.  8B) were blocked Ͼ70% by chlorisondamine. By contrast, in brain only nicotine-stimulated transcription (Fig. 6B) was substantially (Ͼ70%) blocked by chlorisondamine, whereas reserpine-stimulated transcription (Fig. 8D) was reduced only ϳ30% by chlorisondamine. Here we must consider the potential role of the blood-brain barrier in hindering transport of chlorisondamine into the brain; chlorisondamine's two obligate positively charged quaternary amines (i.e. bis-quaternary amine structure), coupled with the relatively brief (30-min) period between chlorisondamine pretreatment and subsequent stimulus (nicotine or reserpine) administration, would diminish the probability of effective and timely brain penetration (73) for central nicotinic cholinergic blockade. Indeed, other investigators (68,75) have found that effective central nervous system nicotinic cholinergic blockade by parenteral chlorisondamine may require larger doses or longer periods of time (as long as 12-21 days of pretreatment).
Role of Catestatin in Vivo-The chromogranin A biologically active fragment catestatin (e.g. bovine chromogranin A 344 -364 ) is proteolytically cleaved from chromogranin A in catecholamine storage vesicles in vivo (14), whereupon it can be co-released by exocytosis along with catecholamines (14). It then acts as a nicotinic cholinergic antagonist (17) to block catecholamine release in response to the physiological trigger, acetylcholine (1). Whereas catestatin's nicotinic cholinergic blocking actions have been extensively characterized in cultured chromaffin cells in vitro (1,(13)(14)(15)(16), its potential actions in vivo have not been extensively explored.
Here we found that catestatin in vivo effectively blocked not only the catecholamine secretory responses (nicotine (Fig. 5) and reserpine (Fig. 7)) but also the adrenal transcriptional responses (nicotine (Fig. 6) and reserpine (Fig. 8)) to nicotinic cholinergic-mediated stimuli. In each case (secretion or transcription), blockade by catestatin was comparable in magnitude with blockade achieved by the classical nicotinic cholinergic antagonist chlorisondamine. In one setting (reserpineinduced brain transcription) (Fig. 8D), catestatin and chlorisondamine were only partially (ϳ30 -35%) effective in blocking stimulus-induced transcription. In this setting, it is noteworthy that both chlorisondamine and catestatin are predicted to exhibit poor penetration of the blood-brain barrier (73) (chlorisondamine because of its two obligate quaternary amine moieties (i.e. bis-quaternary amine structure) and catestatin (as used here: bovine chromogranin A 344 -358 ; RSMRLSFRAR-GYGFR) because of its highly cationic nature (calculated pI of ϳ12.7, with 33% Arg content) and relatively high molecular weight (1861 g/mol) for a drug). Finally, catestatin is a somewhat less potent inhibitor of catecholamine release from neurites than chromaffin cells (1).
Conclusion and Perspectives-Thus, chromogranin A (and its catestatin fragment) seem to be at the virtual nexus of nicotinic cholinergic signaling to both secretion and transcription in vivo. The catecholamine release-inhibitory peptide catestatin seems clearly to function in vivo as a nicotinic cholinergic antagonist, blocking agonist-induced increments in both secretion (Figs. 5 and 7) and transcription (Figs. 6 and 8, B and D) in sympathochromaffin cells. These actions of exogenous catestatin clarify and put into functional perspective recent observations in humans on the inferred actions of endogenous catestatin in vivo (65), in which we observed that a diminution of plasma catestatin was associated with augmented risk of developing hypertension and increased pressor responses to environmental stressors.
These in vivo findings extend the significance of our previous in vitro studies of nicotinic signaling to catecholamine secretion and gene transcription in chromaffin cells (1,(13)(14)(15)(16), establish a fundamental role for chromogranin A and its catestatin fragment at the nexus of nicotinic cholinergic signaling (Fig. 10), and provide a sensitive, novel in vivo photoprobe to investigate stimulus-transcription coupling in experimental cardiovascular disease states. Fig. 10 presents a diagram summarizing our principal findings and conclusions. Although simply a model, it does present a framework for integrating the results and formulating new hypotheses for future testing.