Hippocampal norepinephrine-like voltammetric responses following infusion of corticotropin-releasing factor into the locus coeruleus

Brain Research Bulletin, Vol. 51, No. 4, pp. 319 –326, 2000 Copyright © 2000 Elsevier Science Inc. Printed in the USA. All rights reserved 0361-9230/00/$–see front matter

PII S0361-9230(99)00241-5

Hippocampal norepinephrine-like voltammetric responses following infusion of corticotropin-releasing factor into the locus coeruleus Vitaliy S. Palamarchouk,† Jian-Jun Zhang,‡ Guiying Zhou, Artur H. Swiergiel and Adrian J. Dunn* Department of Pharmacology and Therapeutics, Louisiana State University Medical Center, Shreveport, LA, USA [Received 1 June 1999; Revised 15 September 1999; Accepted 1 October 1999] KEY WORDS: Voltammetry, Norepinephrine, Hippocampus, Corticotropin-releasing factor, Locus coeruleus, Glutamate.

ABSTRACT: Intracerebroventricular (i.c.v.) administration of corticotropin-releasing factor (CRF) increases the activity of noradrenergic neurons in the locus coeruleus (LC) assessed by electrophysiological and neurochemical studies. It has been suggested that this effect of i.c.v. CRF is exerted directly on LC noradrenergic (LC-NE) neurons. Infusion of CRF directly into the LC increases cortical and hippocampal release of norepinephrine (NE) as indicated by in vivo microdialysis studies, but the electrophysiological studies have shown both increases and decreases. The present study used in vivo voltammetry to study changes in the extracellular concentrations of NE in the rat hippocampus in response to infusion of CRF (100 ng) into the LC. When the infusion cannula was located in or very close to the LC, the immediate response to CRF was a small decrease in the NE-like oxidation current, followed by a robust increase after about 6 –7 min. The oxidation current reached a peak around 13 min and returned to baseline by about 30 min after CRF infusion. By contrast with CRF, infusion of glutamate into the LC increased the oxidation current with a delay of around 30 s and a peak within 90 s. The responses to LC infusion of CRF in rats treated with DSP-4 to deplete hippocampal NE were substantially smaller than those in untreated rats, suggesting that the oxidation signals in untreated rats reflected changes in concentrations of NE. The response to glutamate was markedly augmented by pretreatment with the NE reuptake inhibitor, desmethylimipramine, suggesting that the observed responses reflected changes in NE. Infusion of the same dose of CRF into brain structures outside the LC did not elicit consistent changes in oxidation current in the hippocampus. The time course of the responses to CRF is compatible with previously reported electrophysiological responses of LC-NE neurons to CRF and with neurochemical evidence indicating that CRF can affect the activity of LC-NE neurons. The results indicate that CRF may act in or close to the LC to induce release of hippocampal NE, but the delayed response to CRF compared with that to glutamate, suggests that CRF does not directly activate LC-NE neurons. © 2000 Elsevier Science Inc.

INTRODUCTION Numerous studies have indicated that cerebral noradrenergic neurons are activated during stress [9,13,26]. The locus coeruleus (LC) is the primary source for the noradrenergic innervation of the telencephalon [21]. The rate of discharge of the LC noradrenergic (LC-NE) neurons increases during stress [13] and in vivo microdialysis studies have indicated increased release of norepinephrine (NE) in the cerebral cortex and hippocampus (e.g., [1]). Intracerebroventricular (i.c.v.) infusion of corticotropin-releasing factor (CRF) increases the firing rate of LC-NE neurons [30] and has been shown to increase forebrain norepinephrine (NE) release based on measurement of catabolites [8,10,20] and in vivo microdialysis studies [11,16,17,24]. It has been suggested that these effects of CRF may be exerted directly on LC-NE neurons [29]. Recent anatomical studies, including electron microscopy, have identified CRF-containing terminals in contact with both catecholaminergic (immunoreactive for tyrosine hydroxylase) and non-catecholaminergic neurons in or close to the LC [31]. Valentino et al. [30] initially reported that infusion of CRF into the LC had variable results on the firing rate of LC-NE neurons. Direct evidence for an activation was not provided until recently [6]. However, other electrophysiological studies found that even though i.c.v. CRF increased the activity of LC-NE neurons, direct infusion of CRF into the LC decreased rather than increased their activity [3,27]. Studies using in vivo microdialysis have suggested that infusion of CRF into the area of the LC increases NE release in the prefrontal cortex [6,25]. An important question is whether i.c.v. CRF acts directly on the LC after diffusion from the ventricles or whether other neural circuitry is involved. When we studied the cortical and hippocampal release of NE after i.c.v. administration of CRF using in vivo voltammetry, the increase in oxidation

* Address for correspondence: Dr. Adrian J. Dunn, Department of Pharmacology, Louisiana State University Medical Center, P.O. Box 33932, Shreveport, LA 71130-3932, USA. Fax: ⫹1-(318) 675 7857; E-mail: [email protected] † Present address: Department of Physical Chemistry, Gubkin’s State Oil and Gas Academy, Orenburg, Russia. ‡ Present address: Department of Pharmacology, Institute of Materia Medica, Chinese Academy of Medical Sciences, 1 Xian Nong Tan Street, Beijing 100050, P.R. China.

319

320

PALAMARCHOUK ET AL.

current occurred with a delay of around 5 min, and a peak around 35 min [32]. Therefore, we have now studied the hippocampal response to CRF infused directly into LC using in vivo voltammetry to obtain a better assessment of the temporal aspects of the response. MATERIALS AND METHODS Animals and Materials Male Sprague–Dawley rats were purchased from Harlan Sprague Dawley (Houston, TX, USA) and weighed between 250 and 350 g at the time of the experiment. For at least 1 week prior to an experiment the rats were housed singly in plastic cages with ad libitum access to Teklad rat chow (Harlan, Madison, WI, USA) and tap water. A 12-h light: 12-h dark cycle with light on at 0700 h was maintained. CRF donated by Dr. Jean Rivier (Peptide Laboratory, The Salk Institute, San Diego, CA, USA) was dissolved in artificial cerebral-spinal fluid (aCSF) prepared as described elsewhere [25]. Glutamate and desmethylimipramine (DMI) were purchased from Sigma Chemical Co. (St. Louis, MO, USA). DSP-4 (N-2-chloroethyl)-N-ethyl-2-bromobenzylamine) was obtained from Research Biochemicals International (RBI, Natick, MA, USA). Stereotaxic surgery and infusion procedure Rats were anesthetized with urethane (dissolved in water, 0.8 – 1.0 g/kg, i.p.). They were placed in a stereotaxic apparatus with the incisor bar set at ⫺10.0 mm. The tips of the voltammetric electrodes were aimed at the hilus of the dentate gyrus of hippocampus using the following coordinates: AP ⫺4.6 mm from bregma; L ⫾ 2.0 mm, and V 3.2 mm below the skull surface. At the same time, a 0.1-mm hole was drilled in the skull above the LC to insert an injector made from fused silica tubing (OD 175 ␮m: Polymicro Technologies, Phoenix, AZ, USA). The injector tip was aimed at the dorsal aspect of the LC using the following coordinates: AP ⫺ 3.3 mm from lambda, L ⫾ 1.3 mm from the midline; and V 6.2 mm below the skull surface. A 0.5-␮l Hamilton syringe driven by an adapted micrometer screw was connected to the injector for subsequent infusions. All procedures involving animals were approved by the Louisiana State University Medical Center Animal Care Committee and conform to National Institutes of Health guidelines.

Experimental Procedure The working electrode was lowered into the hippocampus and the injector was aimed at the LC. At least 10 min was allowed for the voltammetric signal to stabilize before 100 nl of aCSF, CRF (100 ng) or glutamate (0.1 M) was infused over 60 s through the LC injector. When different substances (aCSF or CRF and aCSF or glutamate) were injected into the same structure during the same experiment, the solutions in the injector were separated by a small air bubble. Subsequent infusions (if any) were made only after the voltammetric signal had stabilized. Only one infusion site was studied in each animal, and only one response from an individual animal to a given agent was used for statistical analysis. A total of 89 rats was used for these studies. Fifty-five rats received infusions of CRF and yielded voltammograms that could be analyzed. Seven additional rats were injected with DSP-4 (50 mg/kg, i.p.) 7–10 days before the CRF infusions. Five rats received control infusions of aCSF into LC. Glutamate was infused to the LC of 16 rats. Six of these rats has been pretreated with NE re-uptake blocker, desmethylimipramine (DMI, 20 mg/kg, s.c.) 30 min before glutamate. Histology and Data Analysis Placements of the injector tips in the brain stem and of the electrodes in the hippocampus were determined by examining 40-␮M cresyl violet-stained coronal brain sections and reference to the stereotaxic atlas of Paxinos and Watson [23]. Only data obtained from animals with the working electrode placed in hippocampus and the site of infusion properly identified are included in the data presented. Electrochemical recordings are vulnerable to electromagnetic interference and drift. Therefore, the data from some animals were discarded because of excessive electrical noise, unstable baseline, or premature termination of recordings. The beginning of a voltammetric response was defined by a consistent (at least 10 data points long) rise or decline of oxidation current above the average of 60 baseline data points immediately preceding an infusion of aCSF, CRF or glutamate. In a stable voltammogram this is approximately equal to a response exceeding 2 SD determined for 60 s of baseline. These responses were determined individually for each voltammogram. The peak response was the highest oxidation current recorded for each voltammogram. Two-tailed Student’s t-test was used to determine significance of differences between the groups in times (beginning and maximum) and amplitude of responses. All data are expressed as mean ⫾ SEM.

Chronoamperometric Measurements Electrochemical measurements were performed using a highspeed chronoamperometric apparatus (IVEC-10, Medical Systems Corp., Greenvale, NY, USA). Working electrodes to be inserted in the hippocampus were made from a single 30-␮m carbon fiber with a tip trimmed to a length of 250 ␮m [28]. The electrodes were coated with Nafion to increase their stability and their selectivity against ascorbic acid [14]. Only electrodes exhibiting a NE to ascorbic acid selectivity ratio of at least 250:1 and a linear response to increasing concentrations of NE were used. Silver/silver chloride reference electrodes prepared from silver wire were placed under the scalp. The oxidation potential of ⫹0.55 V vs. Ag/AgCl reference electrode was applied to the working electrode for 100 ms at the rate of 5 Hz. The averages of 5 measurements of oxidation current were recorded at a rate of 1 Hz. Changes in the measured oxidation current reflect changes in concentrations of the oxidizable substrates that, based on in vitro calibration, are converted by the IVEC software and expressed in ␮M.

RESULTS Of 55 rats that received infusions of CRF and yielded voltammograms that could be analyzed, 22 displayed an obvious increase in oxidation current. Among these, 17 rats appeared to have had injector tips placed in the LC or very close to it. The averaged responses of 13 rats infused with CRF with cannulae subsequently shown to be located in or very close to the LC is shown in Fig. 1. The voltammograms of the other four rats were incomplete and thus were not included in this analysis. The initial response to CRF infusion was a small decrease in oxidation current. This decrease was observed in all 13 rats. The nadir was significantly lower than the baseline (t12 ⫽ 2.80, 2p ⬍ 0.05). It was followed by a significant increase in the oxidation current after a delay of 409 ⫾ 56 s. The increase reached a maximum at 789 ⫾ 60 s and had more or less returned to the pre-infusion baseline within 30 min. The response is depicted both as the change in NE concentration and as the ratio of the presumed change in concentration to the maximum

CRF, LC AND HIPPOCAMPAL NOREPINEPHRINE

321

FIG. 1. The hippocampal chronoamperometric response to 100 ng of corticotropin-releasing factor (CRF) infused into the locus coeruleus (LC). The averaged chronoamperometric measurements depict (A) the norepinephrine (NE)-like oxidation current recorded from a single carbon fiber electrode placed in the ipsilateral hippocampus of 13 anesthetized rats in which the tip of the infusion cannula was located in or close to the LC. The ordinate is the mean change in the electrochemical signals converted to concentration by comparison with in vitro standards. (B) The response to 100 nl artificial cerebrospinal fluid (aCSF) infused into the LC of 5 rats. The voltammograms were synchronized to the time at which CRF or aCSF was infused (indicated by the arrow); the bar indicates the (1 min) duration of the infusion. (C) is based on the same data as the top panel, but the ordinate is the mean of the change in concentration (C) at each time divided by the maximal post-infusion change in concentration (Cmax). The peak of the averaged voltammogram does not reach unity because the peaks occurred at different times in different animals.

(C/Cmax) observed in that rat to normalize the responses (Fig. 1, bottom panel). The maximum amplitude of the response was 0.412 ⫾ 0.030 ␮M. The response appeared to be associated specifically with CRF rather than to stimuli resulting from the infusions, because infusion of aCSF directly into the LC of 5 rats showed no significant responses (Fig. 1, middle panel). Seven rats were treated with the NE-selective neurotoxin, DSP-4, to deplete

telencephalic NE. Of five DSP-4-treated rats that had correct placements of the electrodes and injection cannulae, only two showed observable responses to CRF. These two rats exhibited markedly smaller responses than those of four untreated ones. The mean response from six rats that received glutamate infusions directly into the LC is shown in Fig. 2. The oxidation current began to increase 34 ⫾ 5 s after infusion and reached a maximum

322

PALAMARCHOUK ET AL.

FIG. 2. Effects of infusion of glutamate on the norepinephrine (NE)-like oxidation current recorded in hippocampus. Glutamate (100 nl 0.1 M) was infused into the locus coeruleus (LC). The lower trace of each panel (n ⫽ 10) shows the effect of glutamate in untreated rats. The upper trace (n ⫽ 6) of each panel shows the effects in rats pretreated with DMI (20 mg/kg, s.c.) 30 min before glutamate. The arrows indicate the time at which glutamate was infused; the bars indicate the duration (1 min) of the infusion. The lower panel is derived from the same data, but is expressed as C/Cmax (see Fig. 1).

at 83 ⫾ 20 s. These responses were relatively small (mean amplitude 0.089 ⫾ 0.012 ␮M) and returned to baseline within 2 min. All animals that received successful infusions of glutamate into the LC displayed increases in oxidation current in the hippocampus. To provide evidence that the responses reflected increases in NE, six rats were pretreated with DMI, a selective uptake inhibitor of NE, which should enhance extracellular concentrations of NE, and thus the voltammetric signal from NE, as has been shown by Mateo et al. [19]. In these rats, infusion of glutamate into LC produced substantially greater responses with a mean delay of 57 ⫾ 15 s from the time of infusion and a maximum at 136 ⫾ 25 s. The amplitude of the response was 0.419 ⫾ 0.035 ␮M. The response returned to baseline within 10 min. There was no initial decrease in oxidation current following glutamate infusion as was observed in CRF-infused rats. With or without DMI pretreatment, the voltammetric responses to glutamate infusions occurred much earlier than with CRF infusions. The temporal data on the responses to CRF and glutamate are summarized in Fig. 3 which indicates the mean time for the start of a peak and its maximum. Clearly, both the onset and the peak responses were much later for CRF than for glutamate. The voltammogram from one rat that was treated sequentially with aCSF, CRF and glutamate is shown in Fig. 4. There was no significant change in oxidation current in response to infusion of aCSF. When CRF was administered, a small decrease occurred.

This was followed, after a delay of several minutes, by a relatively rapid increase with a peak 340 s after infusion, and a slow return of the signal to the preinfusion baseline. The rat was then injected with DMI and 30 min later glutamate was infused into the LC. The signal rose rapidly, reaching a peak at 50 s after infusion and then rapidly returned to baseline. An important aspect of this study was to assess whether any responses were anatomically specific for LC. Therefore, we deliberately varied the coordinates for the infusion cannula around the general area of the LC. Figure 5 indicates the sites for which 55 individual voltammograms were obtained following CRF infusions. The symbols indicate sites at which there were clear-cut responses like those shown in Figs. 1 and 4 (19 cases), as well as three sites at which there were slow responses like those observed to i.c.v. CRF, or no response at all (33 sites). All 17 rats in which infusion of CRF appeared to have been into the LC exhibited responses similar to those of Fig. 4. Among the 38 animals with CRF infusion sites subsequently shown to be located outside LC (lateral to LC, caudal to LC, into the brain stem, or above LC in the cerebellum), an increase in oxidation current was observed in only five rats. In one rat, an increase in oxidation current like that shown in Fig. 1 was observed after infusion of CRF into the second lobe of the cerebellum, but two other animals injected in the same area (squares on Fig. 5, upper panel) produced only a very slow small increase,

CRF, LC AND HIPPOCAMPAL NOREPINEPHRINE

FIG. 3. Summary of the time parameters of the hippocampal norepinephrine (NE) responses to the locus coeruleus (LC) infusion of corticotropinreleasing factor (CRF) and glutamate. Based on the data presented in Figs. 1 and 2. **Significantly different from glutamate-infused rats, p ⬍ 0.001.

unlike those observed after infusion of CRF directly into LC. A fourth rat that had a cannula located in the ventral spinocerebellar tract just above the lateral parabrachial nucleus exhibited a small increase much later than those observed after LC infusions. In a fifth rat, infusion of CRF into the lateral parabrachial nucleus produced a response similar to those observed after LC infusions. Because this site was relatively close to the LC, it is possible that CRF reached the LC, or that some afferents to the LC were affected. DISCUSSION The electrochemical signal recorded from the hippocampus in vivo could have been generated by oxidation of a number of

323 readily oxidized chemical compounds other than NE, such as dopamine (DA), 3,4-dihydroxyphenylacetic acid (DOPAC), 5-hydroxyindoleacetic acid (5-HIAA) or ascorbic acid. Several lines of evidence suggest that in the present experiments the major contributor to the voltammetric signal was NE. (1) There is little DA in the hippocampus, and the electrodes were placed in the hilus of the dentate gyrus, a region where the noradrenergic innervation is particularly dense [18]. (2) The working electrodes were coated with a thin layer of Nafion, which enhances the selectivity of the electrode for DA and NE and minimizes signals from anions such as ascorbic acid, uric acid, and catabolites like DOPAC and 5-HIAA [14]. (3) We observed small or no responses to LC infusion of CRF in DSP-4-treated rats. Because DSP-4 depletes NE by more than 75% in the hippocampus, the results suggest the measured current results from oxidation of catecholamines. Thus it seems most likely that the responses observed were caused by release of NE in the hippocampus. (4) In an earlier in vivo microdialysis study, we observed that CRF administration into the LC was followed by increased extracellular concentrations of NE in the medial prefrontal cortex [25]. The time courses of the microdialysis and voltammetry responses were comparable, bearing in mind the low temporal resolution of the microdialysis study. (5) When the rats were pretreated with DMI, the voltammetric response to glutamate infusion was augmented. DMI is selective for NE over DA transporters [15]. There is general agreement that i.c.v. administration of CRF increases the firing rate of LC-NE neurons [3,27,30]. There is also universal agreement that i.c.v. CRF increases NE release as determined by measurement of catabolites [8,10,20], as well as in vivo microdialysis studies [11,16,17,24], and an in vivo voltammetry study [32]. There is evidence that increases in the firing rate of LC-NE neurons are associated with increased NE release in the cortex [2,6,12] and this is reflected in the agreement between the electrophysiological and neurochemical data. The controversy surrounds the effect of local injections of CRF into the LC. In the pioneering study of Valentino et al. [30], it was reported that responses of LC-NE neurons to the direct application of CRF were quite variable. Nevertheless, it was speculated that the effect of

FIG. 4. The effect of infusion of artificial cerebrospinal fluid (aCSF), corticotropin-releasing factor (CRF) and glutamate (Glu) into locus coeruleus (LC) on the norepinephrine (NE)-like oxidation current recorded in hippocampus. A voltammogram from a single rat. The time (in seconds) following the initial aCSF infusion is indicated on the abscissa. Infusions into the LC are indicated by the arrows (aCSF, 100 nl; CRF, 100 ng in 100 nl; glutamate, 100 nl 0.1 M); the horizontal bars indicate the durations of the infusions. Desmethylimipramine (DMI) was injected 30 min before glutamate.

324

PALAMARCHOUK ET AL.

FIG. 5. Coronal section of the rat brain at the levels of the locus coeruleus (LC) showing the location of the tips of the infusion cannulae. Only the sites for corticotropin-releasing factor (CRF) infusion (100 ng in 100 nl) are shown. The location of the cannula tips in rats in which CRF infusion evoked hippocampal responses like those shown in Figs. 1 and 4 are indicated by closed circles (n ⫽ 19). Those in which infusion did not produce observable responses are depicted by open circles (n ⫽ 33). Sites of infusions that evoked very slow or atypical response are indicated by squares (n ⫽ 3). The location of the cannula tip of the rat used for Fig. 4 is indicated by the arrow.

CRF, LC AND HIPPOCAMPAL NOREPINEPHRINE i.c.v. CRF was directly exerted on LC-NE neurons. However, a careful study of the direct effects of CRF on LC-NE neurons revealed only inhibitory responses [3]. Similar results were obtained in a preliminary report from another laboratory [27]. A recent study from Valentino’s group reported that direct application of CRF into the LC increased the firing of LC neurons, and that CRF in the LC was about 20-fold more effective than i.c.v. CRF [6]. This complemented earlier data from the same laboratory suggesting that administration of CRF antagonists into the LC attenuated or prevented the electrophysiological responses to nitroprusside [5,22]. The in vivo microdialysis studies indicate that CRF injected into LC increased NE release in the prefrontal cortex [6,25]. This effect appeared to be anatomically selective for the LC [25]. In the Smagin et al. study [25], the NE responses peaked within the first two dialysis samples (22 min each). In the Curtis study, the increase occurred only in the first 20-min sample and appeared to parallel the electrophysiological activation of LC-NE neurons [6]. The finer temporal resolution of the voltammetric recording in the present study suggests that the response of hippocampal NE was not immediate, but appeared after a mean delay of around 6 –7 min and did not reach a peak until around 13 min. Interestingly, the increase was preceded by a decrease, which was observed in all 13 rats and is consistent with the inhibitory electrophysiological responses reported by Stowe et al. [27] and Borsody and Weiss [3]. In the studies from the Valentino laboratory, the time course of the electrophysiological response observed to CRF infusion into the LC is not easy to discern from the published reports. In a recording presented from one rat, the increased firing appeared to start within 2 min after infusion, and reached a maximum around 7 min [6]. But data from another rat in the same report, suggests a maximum around 13–14 min, consistent with our voltammetric observations of apparent NE release, but the start of the peak cannot be determined. Infusion of glutamate into the LC induced a response that was considerably faster than that to CRF. Thus the possibility must be considered that the effect of CRF was not exerted directly on CRF-NE neurons. It could be argued that because CRF is a much larger molecule than glutamate, it would diffuse more slowly and thus the response would be expected to be slower than to glutamate. But it is questionable whether CRF would take 6 –7 min, when glutamate takes less than 30 s. It is relevant that mice lacking the CRF gene exhibit normal cerebral noradrenergic responses to footshock, indicating that CRF is not essential for the noradrenergic responses in stress [7]. It is useful to compare the present voltammetric results with those obtained following i.c.v. administration of CRF [32]. Those studies revealed increases in hippocampal NE release with a delay of around 5 min, and a peak at around 35 min. Thus the delay in the response following intra-LC infusion of CRF was similar but the peak response occurred somewhat earlier than with i.c.v. infusions, suggesting that if similar mechanisms are involved, the LC infusions were closer to the active site. It is important to note that the i.c.v. dose of CRF (1 ␮g) was 10 times higher than the LC dose of CRF. Other data suggest that there is some peripheral contribution to the LC-NE response to i.c.v. CRF. Both the electrophysiological response [3] and the voltammetric response [32] were inhibited by pretreatment with the ganglionic blocker, chlorisondamine. Moreover, ␤-adrenergic receptor antagonists known to penetrate the brain poorly similarly attenuated the electrophysiological [3] and voltammetric responses [32]. These results suggest that peripheral sympathetic mechanisms contribute to the response, consistent with the known activation of the autonomic nervous system by i.c.v. CRF [4].

325 The major findings of this study are that CRF injected directly into the LC increased the apparent release of NE in the hippocampus. Such an effect was not observed consistently when CRF was injected into a variety of surrounding regions. However, the hippocampal response to CRF was slow starting around 6 –7 min, reaching a peak at 13 min, and dissipating by 30 min. In contrast, infusion of glutamate into the LC produced a much more rapid response, hippocampal NE release increased within 30 s, and reached a peak within 90 s. Thus although CRF might act directly on LC-NE neurons to alter their activity, it is more likely that the effect is indirect. ACKNOWLEDGEMENTS

This work was supported by grants from the National Institutes of Health (MH50947) and the Air Force Office of Scientific Research (F49620-93-1-0125DEF). The technical assistance of Jonathan Bennett is greatly appreciated.

REFERENCES 1. Abercrombie, E. D.; Keller, R. W.; Zigmond, M. J. Characterization of hippocampal norepinephrine release as measured by microdialysis perfusion: Pharmacological and behavioral studies. Neuroscience 27: 897–904; 1988. 2. Berridge, C. W.; Abercrombie, E. D. Relationship between locus coeruleus discharge rates and rates of norepinephrine release within neocortex as assessed by in vivo microdialysis. Neuroscience 93: 1263–1270; 1999. 3. Borsody, M. K.; Weiss, J. M. Influence of corticotropin-releasing hormone on electrophysiological activity of locus coeruleus neurons. Brain Res. 724:149 –168; 1996. 4. Brown, M. R.; Fisher, L. A.; Spiess, J.; Rivier, C.; Rivier, J.; Vale, W. Corticotropin-releasing factor: Actions on the sympathetic nervous system and metabolism. Endocrinology 111:928 –931; 1982. 5. Curtis, A. L.; Grigoriadis, D. E.; Page, M. E.; Rivier, J.; Valentino, R. J. Pharmacological comparison of two corticotropin-releasing factor antagonists: In vivo and in vitro studies. J. Pharmacol. Exp. Ther. 268:359 –365; 1994. 6. Curtis, A. L.; Lechner, S. M.; Pavcovich, L. A.; Valentino, R. J. Activation of the locus coeruleus noradrenergic system by intracoerulear microinfusion of corticotropin-releasing factor: Effects on discharge rate, cortical norepinephrine levels and cortical electroencephalographic activity. J. Pharmacol. Exp. Ther. 281:163–172; 1997. 7. Dunn, A. J. Footshock-induced changes in brain catecholamines and indoleamines are not mediated by CRF or ACTH. Neurochem. Int. In press. 8. Dunn, A. J.; Berridge, C. W. Corticotropin-releasing factor administration elicits a stress-like activation of cerebral catecholaminergic systems. Pharmacol. Biochem. Behav. 27:685– 691; 1987. 9. Dunn, A. J.; Kramarcy, N. R. Neurochemical responses in stress: Relationships between the hypothalamic-pituitary-adrenal and catecholamine systems. In: Iversen, L. L.; Iversen, S. D.; Snyder, S. H., eds. Handbook of psychopharmacology. New York: Plenum Press; 1984:455–515. 10. Emoto, H.; Tanaka, M.; Koga, C.; Yokoo, H.; Tsuda, A.; Yoshida, M. Corticotropin-releasing factor activates the noradrenergic neuron system in the rat brain. Pharmacol. Biochem. Behav. 45:419 – 422; 1993. 11. Emoto, H.; Yokoo, H.; Yoshida, M.; Tanaka, M. Corticotropin-releasing factor enhances noradrenaline release in the rat hypothalamus assessed by intracerebral microdialysis. Brain Res. 601:286 –288; 1993. 12. Florin-Lechner, S. M.; Druhan, J. P.; Aston-Jones, G.; Valentino, R. J. Enhanced norepinephrine release in prefrontal cortex with burst stimulation of the locus coeruleus. Brain Res. 742:89 –97; 1996. 13. Foote, S. L.; Aston-Jones, G.; Bloom, F. E. Impulse activity of locus coeruleus neurons in awake rats and monkeys is a function of sensory stimulation and arousal. Proc. Natl. Acad. Sci. USA 77:3033–3037; 1980. 14. Gerhardt, G. Rapid chronocoulometric measurements of norepineph-

326

15.

16.

17. 18. 19.

20.

21. 22. 23.

PALAMARCHOUK ET AL. rine overflow and clearance in CNS tissues. In: Boulton, A. A.; Baker, G. B.; Adams, R. N., eds. Voltammetric methods in brain systems. Totowa, NJ: Humana Press, Inc.; 1995:117–151. Horn, A. S.; Coyle, J. T.; Snyder, S. H. Catecholamine uptake by synaptosomes from rat brain. Structure-activity relationships of drugs with differential effects on dopamine and norepinephrine neurons. Mol. Pharmacol. 7:66 – 80; 1971. Lavicky, J.; Dunn, A. J. Corticotropin-releasing factor stimulates catecholamine release in hypothalamus and prefrontal cortex in freely moving rats as assessed by microdialysis. J. Neurochem. 60:602– 612; 1993. Lee, E. H. Y.; Chang, S. Y.; Chen, A. Y. J. CRF facilitates NE release from the hippocampus: A microdialysis study. Neurosci. Res. 19:327– 330; 1994. Loy, R.; Koziell, D. A.; Lindsey, J. D.; Moore, R. Y. Noradrenergic innervation of the adult rat hippocampal formation. J. Comp. Neurol. 189:699 –710; 1980. Mateo, Y.; Pineda, J.; Meana, J. J. Somatodendritic ␣2-adrenoceptors in the locus coeruleus are involved in the in vivo modulation of cortical noradrenaline release by the antidepressant desipramine. J. Neurochem. 71:790 –798; 1998. Matsuzaki, I.; Takamatsu, Y.; Moroji, T. The effects of intracerebroventricularly injected corticotropin-releasing factor (CRF) on the central nervous system: Behavioral and biochemical studies. Neuropeptides 13:147–155; 1989. Moore, R. Y.; Bloom, F. E. Cerebral catecholamine neuron systems: Anatomy and physiology of the norepinephrine and epinephrine systems. Annu. Rev. Neurosci. 2:113–168; 1979. Page, M. E.; Berridge, C. W.; Foote, S. L.; Valentino, R. J. Corticotropin-releasing factor in the locus coeruleus mediates EEG activation associated with hypotensive stress. Neurosci. Lett. 164:81– 84; 1993. Paxinos, G.; Watson, C. The rat brain in stereotaxic coordinates. San Diego, CA: Academic Press; 1997.

24. Shimizu, H.; Bray, G. A. Modulation by corticotropin-releasing factor of monoamine metabolism in the lateral hypothalamus. Neurosci. Lett. 103:74 – 80; 1989. 25. Smagin, G. N.; Swiergiel, A. H.; Dunn, A. J. Corticotropin-releasing factor administered into the locus coeruleus, but not the parabrachial nucleus, stimulates norepinephrine release in the prefrontal cortex. Brain Res. Bull. 36:71–76; 1994. 26. Stone, E. A. Stress and catecholamines. In: Friedhoff, A. J., ed. Catecholamines and behavior. Vol. 2: Neuropsychopharmacology. New York: Plenum Press; 1975:31–72. 27. Stowe, Z. N.; Landry, J. C.; Owens, M. J.; Plotsky, P. M.; Nemeroff, C. B. The effects of directly applied CRF on the firing rates of individual locus ceruleus neurons. Am. Coll. Neuropsychopharmacol. 34th Ann. Mtg. 152; 1995. 28. Swiergiel, A. H.; Palamarchouk, V. S.; Dunn, A. J. A new design of carbon fiber microelectrode for in vivo voltammetry using fused silica. J. Neurosci. Meth. 73:29 –33; 1997. 29. Valentino, R. J. Corticotropin-releasing factor: Putative neurotransmitter in the noradrenergic locus coeruleus. Psychopharmacol. Bull. 25: 306 –311; 1989. 30. Valentino, R. J.; Foote, S. L.; Aston-Jones, G. Corticotropin-releasing factor activates noradrenergic neurons of the locus coeruleus. Brain Res. 270:363–367; 1983. 31. Van Bockstaele, E. J.; Colago, E. E.; Valentino, R. J. Corticotropinreleasing factor-containing axon terminals synapse onto catecholamine dendrites and may presynaptically modulate other afferents in the rostral pole of the nucleus locus coeruleus in the rat brain. J. Comp. Neurol. 364:523–534; 1996. 32. Zhang, J.-J.; Swiergiel, A. H.; Palamarchouk, V. S.; Dunn, A. J. Intracerebroventricular infusion of CRF increases extracellular concentrations of norepinephrine in the hippocampus and cortex as determined by in vivo voltammetry. Brain Res. Bull. 47:227–284; 1998.