the carotid sinus and aortic arch are innervated by arterial baroreceptors that buffer acute changes in blood pressure (BP). One component of this buffering mechanism is the ability of baroreceptors to regulate the release of vasopressin from the neurohypophysis (16, 28). Physiologically, vasopressin is known primarily for its antidiuretic effect (28). The magnocellular neurosecretory cells located in the supraoptic nucleus (SON) and paraventricular nuclei of the hypothalamus release either vasopressin or another peptide hormone, oxytocin, into the peripheral circulation (1, 24, 25). Quantitative RT-PCR studies demonstrate that whereas magnocellular neurons contain mRNA for both vasopressin and oxytocin, neurons in the SON express either vasopressin or oxytocin several orders of magnitude higher than the other peptide (36). Furthermore, the magnocellular neurons of the SON may be characterized as either vasopressin or oxytocin secreting on the basis of their firing patterns and differential responses to afferent input (1, 24, 25). Putative vasopressin neurons are characterized by either phasic or continuous patterns of spontaneous activity, whereas putative oxytocin neurons exhibit continuous patterns of activity (1, 24,25). Moreover, the activation of baroreceptors produces a transient cessation of the spontaneous activity of vasopressinergic magnocellular neurons, unlike oxytocinergic magnocellular neurons that display no sensitivity to baroreceptor activation (24,25). This selective inhibitory effect of baroreceptor activation on vasopressin magnocellular neurons may be used as a tool for distinguishing vasopressin from oxytocin magnocellular neurons in vivo (24, 25).

It has been suggested that catecholaminergic projections from the A1 region of the caudal ventrolateral medulla to the SON may be involved with the baroreceptor-mediated inhibition of vasopressin SON neurons (29). However, previous studies have shown that catecholamine depletions of the SON that completely block the effects of A1 stimulation on SON neurons do not influence the ability of increases in BP to inhibit vasopressin neurons in the SON (8). This suggests that the baroreceptor-mediated inhibition of vasopressin neurons in the SON involves another pathway. A number of studies indicate that the diagonal band of Broca (DBB), located in the forebrain, contains neurons that are activated by stimulating arterial baroreceptors and that when the DBB is electrically stimulated, it selectively inhibited the spontaneous activity of magnocellular vasopressin neurons in the SON (24). Both the DBB- and baroreceptor-induced inhibition of SON vasopressin neurons have been shown to be mediated by GABA_A receptors (24). Importantly, excitotoxin lesions of the DBB significantly reduce the number of SON neurons that are inhibited by baroreceptors (6, 7). Therefore, it has been proposed that the DBB is an integral component of the pathway that relays baroreceptor information to the vasopressin neurons in the SON (24). Additional studies have demonstrated that the DBB does not project directly to the SON, but instead DBB afferents terminate in a region dorsal and lateral to the SON, the so-called perinuclear zone (PNZ) of the SON (14, 34). Because the neurons of the SON are resistant to excitotoxins, it was possible to lesion the neurons in the PNZ without affecting the magnocellular neurons (7, 20). In rats with PNZ lesions, both the DBB- and baroreceptor-induced inhibition of vasopressin SON neurons were significantly reduced (20). These results indicate that both types of inhibition rely on a population of GABAergic interneurons located in the PNZ. Therefore, the DBB neurons that project to the PNZ are not likely to be GABAergic and instead provide an excitatory input to GABA neurons in the PNZ (24). The nature of this projection from the DBB to the PNZ remains to be determined.

In the DBB, electrophysiological studies have identified two primary populations of neurons, one of which is cholinergic (12, 13, 18,19). The other population remains unidentified, although immunohistochemical evidence indicates the DBB contains a significant number of GABA neurons (14, 15, 21). Anatomical studies indicate efferent projections from the DBB to the hippocampal area, medial/lateral preoptic areas, mammillary hypothalamic complex, and the lateral hypothalamus, which includes the PNZ (12, 31, 34,38). Cholinergic neurons in the DBB, typically implicated in behavior control and hippocampal function, do not all project to the hippocampus, suggesting an alternative role for these neurons (38).

At present, there are no data identifying which population of DBB neurons projects to or activates neurons in the PNZ, although preliminary evidence in vitro indicates that the putative cholinergic DBB neurons may be part of this circuit (4). The development of immunotoxins that use neurotoxins coupled to antibodies has made it possible to make highly selective lesions of specific neuronal phenotypes (26, 35). The selective cholinergic toxin 192 IgG-saporin is well documented for its ability to eliminate cholinergic neurons in the basal forebrain, specifically the DBB (9, 26, 35). This toxin combines the ribosomal inactivating protein saporin with a monoclonal antibody for the p75 NGF receptor that is located on cholinergic neurons in the rat (30), allowing injections of 192 IgG-saporin to selectively destroy cholinergic neurons while leaving noncholinergic neurons intact (17, 33). With the use of this immunolesioning approach to deplete the DBB of cholinergic neurons, the present study tested whether or not a cholinergic projection from the DBB participates in the baroreceptor-induced inhibition of vasopressin neurons.

MATERIALS AND METHODS

Animals.

Experiments were conducted on male Sprague-Dawley rats weighing 225–350 g (Harlan; Indianapolis, IN). Animals were maintained in a temperature-controlled environment on a 12:12-h light-dark cycle with food and water available ad libitum. All surgical procedures and experimental protocols were carried out as approved by the Institutional Animal Care and Use Committee at the University of Missouri, Columbia, in accordance with the guidelines of the Public Health Service, American Physiological Society, and Society for Neuroscience.

Drugs.

192 IgG-saporin and unconjugated saporin control (Advanced Targetting Systems, San Diego, CA) were each diluted in artificial cerebrospinal fluid (ACSF) to a final concentration of 0.8 μg/μl. ACSF was prepared using distilled water with the following (in mg/100 ml total volume): 752.0 NaCl, 31.0 KHCO₃, 20.3 MgCl₂-6 H₂O, 180.0 NaHCO₃, 50.0 anhydrous glucose, and 13.0 CaCl₂. The pH of the ACSF was adjusted to 7.4 using 1 N HCl or 1 N NaOH.

Stereotaxic surgery.

Male Sprague-Dawley rats (225–350 g) were anesthetized with pentobarbital sodium (50 mg/kg ip). The rats each received a midline stereotaxic injection directed at the DBB as previously described (5, 6). Each injection was delivered into the DBB over a 5-min period using a 30-gauge steel injector connected to a 5-μl Hamilton syringe. The stereotaxic coordinates were 0.0 mm lateral from bregma, +0.7 mm rostral to bregma, and −7.3 mm ventral from bregma, with minor adjustments of the rostral and ventral coordinates for rats with a body weight in the upper end of the range. After each injection, the injector was left in the tissue for ∼30 s after which it was slowly withdrawn over an additional 5 min. One group of rats (n = 11) each received a single midline stereotaxic injection of 192 IgG-saporin (500 nl, 0.8 μg/μl dissolved in ACSF) into the DBB. A separate group of animals (n = 4) was injected with only the ACSF vehicle using the same volume and coordinates as used for 192 IgG-saporin injections. Likewise, unconjugated saporin (0.8 μg/μl in ACSF) was injected in a third group of rats (n = 6) using the same volume (500 nl) and coordinates. This approach was used because previous studies have demonstrated that these midline injections cover the DBB on both sides of the midline (5, 6). After surgery, the rats were monitored daily to ensure their general health and were allowed at least a 12-day recovery period to maximize cell loss before surgery for electrophysiological experiments.

Electrophysiology protocol.

Adult male Sprague-Dawley rats that were previously injected with either 192 IgG-saporin, ACSF vehicle, or unconjugated saporin were prepared for electrophysiological recording of single-unit activity from the SON via a transpharyngeal approach (5-7,20). Because midline injections into the DBB significantly affect both sides of the DBB (5-7), extracellular recordings were made from either the left or the right SON depending on which side was more suitable for extracellular recording. Rats were anesthetized with pentobarbital sodium (50 mg/kg ip), and catheters were placed in the left femoral artery and vein for BP recording and phenylephrine infusion, respectively. Another catheter was placed in the left jugular vein for supplementary infusion of pentobarbital sodium (2–5 mg/kg as needed) to maintain anesthesia at a level sufficient to suppress withdrawal reflexes. BP was recorded using a pressure transducer connected to a preamplifier (Dalton Electronics Core) and a Pentium computer running Spike 2 data-acquisition software (Cambridge Electronic Design, UK). After catheter instrumentation, the SON and posterior pituitary gland were surgically exposed. A pressure foot was placed on the ventral surface of the hypothalamus to stabilize the area surrounding the SON. A bipolar stimulating electrode was placed in the posterior pituitary to antidromically activate the neurosecretory neurons in the SON. The posterior pituitary was stimulated using single-discharge 1-Hz suprathreshold current pulses (0.1-ms duration, ≤10 mA). The criteria used for antidromic activation included constant latency at threshold, an all-or-none response, and evidence of collision cancellation between spontaneously occurring action potentials and action potentials evoked antidromically from the posterior pituitary.

Extracellular action potentials from SON neurons were recorded using glass micropipettes filled with 3 M sodium chloride. The signals were amplified (DAGAN 2400 extracellular preamplifier, Minneapolis, MN), filtered, relayed through a window discriminator to a CED 1401 analog-to-digital converter (Cambridge Electronic Design, UK), and sent to a Pentium computer running Spike 2 data-acquisition software.

For the present study, we only analyzed recordings from antidromically activated neurons that had phasic patterns of spontaneous activity, because previous studies using immunocytochemistry indicate that this pattern of activity is a characteristic of vasopressin neurons (3). Continuously active neurons that could be vasopressinergic or oxytocinergic were not included in the current study. This was done because the lesions were intended to eliminate the baroreceptor sensitivity of these cells that would no longer allow us to discriminate continuous vasopressin cells from continuous oxytocin cells. Peripheral baroreceptor stimulation was accomplished using a bolus injection of phenylephrine (10 μg/10 μl iv) sufficient to cause an increase in BP of at least 40 mmHg. Previous studies indicate that a stimulus of this intensity is sufficient to inhibit 90–100% of phasic neurons in the SON (5-7, 20). Phenylephrine-induced baroreceptor stimulation was normally administered either in the 5–15 s immediately after the initiation of a spontaneous phase or during a sustained phasic burst. At least one burst of spontaneous activity was recorded before phenylephrine infusion. A positive response was recorded if the cell shut off during the 20 s immediately after the initiation of a BP increase (5-7, 20). Each phasic neuron was tested for baroreceptor responsiveness no more than five times. Differences between the experimental groups in the number of phasic neurons inhibited by baroreceptor stimulation were tested using chi-square contingency tables.

After the conclusion of the electrophysiological experiments, each rat was overdosed with pentobarbital sodium and perfused transcardially with PBS followed by 4% paraformaldehyde. Brains were subsequently removed and transferred to a 30% sucrose PBS solution to cryoprotect the tissue. Each brain was cut into 40-um sections in a cryostat. Serial sections from the forebrain of each rat were collected in cryoprotectant solution and stored at −20°C until processed for choline acetyltransferase (ChAT) immunocytochemistry, as described below.

PNZ injections.

Normal rats and rats previously injected in the DBB with 192 IgG-saporin (see above) were anesthetized with pentobarbital sodium (50 mg/kg ip) and placed in a stereotaxic frame. Each rat received three unilateral injections of rhodamine-labeled fluorescent latex microspheres (LumaFluor; Naples, FL) into the PNZ. The volume of each injection was 300 nl. The following stereotaxic coordinates were used for the injections with adjustments for body weight: 1.1, 0.9, and 0.7 mm posterior to bregma, 9.0 mm ventral to bregma, and a progression of coordinates 2.4, 2.2, and 2.0 mm lateral to bregma. Each injection was made through 30-gauge steel hypodermic tubing connected to a 5-μl Hamilton microsyringe (Reno, NV) over a 3-min period. The steel injector tubing was left in place for 5 min, after which it was withdrawn over an additional 5-min period to minimize leakage up the cannula tract. After at least 3 days following the PNZ injection to allow retrograde transport of the microspheres, animals were anesthetized with pentobarbital sodium (50 mg/kg ip) and perfused transcardially as described above.

ChAT immunocytochemistry.

Cholinergic neurons in the basal forebrain were identified using ChAT immunocytochemistry (15). Brain sections were rinsed in 0.1 M PBS for 30 min and then incubated in 0.3% hydrogen peroxide in distilled water for 30 min at room temperature. After this incubation, sections were again rinsed for 30 min in PBS and then incubated for 2 h at room temperature in PBS diluent [3% normal horse serum (Sigma; St. Louis, MO) in 0.1 M PBS containing 0.25% Triton X-100]. Sections were subsequently incubated with a mouse monoclonal anti-ChAT antibody (Chemicon International; Temecula, CA), diluted 1:250 in PBS diluent for 48 h at 4°C. After two 30-min rinses in 0.1 M PBS, the tissue was then incubated with an anti-mouse antibody (10 μg/ml PBS diluent) labeled with the fluorophore Oregon green (Molecular Probes; Eugene, OR) for ∼4 h. Sections were rinsed thoroughly with 0.1 M PBS to stop the reaction, mounted on gel-coated slides, and placed under a coverslip with Fluoromount mounting media (Southern Biotechnology Associates; Birmingham, AL).

Histology.

Serial sections from the brains of rats injected with either 192 IgG-saporin, unconjugated saporin, or ACSF vehicle were analyzed by light microscopy using an inverted microscope equipped with a mercury lamp for epifluorescence (Olympus IX50, Melville, NY). The sections were analyzed to determine the placement of the injections and to quantify the number of ChAT-positive cells in the DBB. Digital images from each section were captured using a charge-coupled device (CCD) (CCD-72, Dage MTI, IN) connected to a frame-grabber board (LG-3-PCI, Scion, MD) in a Pentium II computer running Scion Image (version 1.4, Scion). Representative sections for three levels of the DBB were selected and equated between animals based on established anatomical coordinates and landmarks. The DBB was subdivided into rostral, medial, and caudal levels corresponding to the following respective coordinates as mapped by Paxinos and Watson (23): 0.7, 0.48, and 0.20 mm anterior to bregma. The rostral DBB was sampled at the level of the small, fused rostral anterior commissure and began just caudal to the fusion of the septum at the level of the genu corpus callosum (23). At the medial DBB level, the anterior commissure was bounded medial laterally by the bed nucleus of the stria terminalis as the basal forebrain fissure became less pronounced. The caudal DBB was sampled at a level just rostral to the optic chiasm and was further delineated by the “comma”-shaped formation of the anterior commissure. The digital images of the ChAT-immunoreactive cells in vertical and horizontal limbs of the DBB from each region were used for cell counts. The number of ChAT-positive cells in each of the three divisions of the DBB was counted and summed together for each animal. Counts were performed by observers blind to the nature of the DBB injection each animal had received.

In the retrograde-labeling study, sections containing the DBB were similarly analyzed. Additional sections containing the PNZ were analyzed to determine the placement of the microsphere injections. For 192 IgG-saporin injections into the DBB, cannula tract was identified at the injection site. In every PNZ-injected rat, the patterns of ChAT-immunoreactive neurons in the DBB were noted, and digital images were collected as described previously. The same sections were then examined, and an additional set of digital images was collected using the filters for the rhodamine fluorescence to identify neurons retrogradely labeled from the PNZ. The pattern and distribution of ChAT-positive cells and neurons retrogradely labeled from the PNZ were compared to determine whether or not cells were doubly labeled.

RESULTS

Histology.

In vehicle-injected control rats, ChAT-positive neurons were distributed throughout the DBB with the highest number in the rostral parts of the nucleus (Table 1). In the rostral DBB, the ChAT-immunoreactive neurons were concentrated primarily on either side of the midline in the more dorsal aspects of the DBB with a few neurons extending ∼1–2 mm laterally in the ventral-basal aspect of the DBB. In the medial DBB, the ChAT-immunoreactive neurons tended to be sparsely distributed along the midline with most of the ChAT-positive cells grouped in two distinct clusters that were located more laterally than those ChAT-immunoreactive neurons observed in the rostral DBB. A second population of neurons was located along the ventral surface of the DBB, stretching laterally from either side of the midline. In the caudal DBB, the ChAT-positive cells were distributed in a pattern similar to the medial DBB, although in the caudal DBB, there were fewer ChAT-immunoreactive cells distributed over a larger area.

Injections of unconjugated saporin into the DBB did not significantly change the number of ChAT-positive neurons (Table 1) or alter the distribution of ChAT-immunoreactive cells in the DBB (Fig.1). DBB injections of 192 IgG-saporin significantly reduced the number of ChAT-immunoreactive neurons in all three levels of the DBB compared with vehicle control and unconjugated saporin injections (P < 0.05, Table 1). In fact, injections of 192 IgG-saporin eliminated virtually all ChAT-positive neurons bilaterally in the rostral and medial levels of the DBB. With the exception of one animal, the only ChAT-positive neurons found in the DBB after 192 IgG-saporin treatment were located 1–2 mm from the injection site in the most caudal aspect of the DBB, particularly in the most ventral and lateral portions.

Electrophysiology.

Data were obtained from a total of 77 SON neurons antidromically activated from the posterior pituitary, displaying a phasic discharge pattern characteristic of vasopressin neurons (seematerials and methods). As many as seven phasically active neurons were recorded from any one animal. Phasic SON neurons were classified on the basis of their sensitivity to phenylephrine-induced changes in BP activity.

Extracellular recordings were obtained from 21 phasic vasopressin SON neurons in rats injected in the DBB with ACSF. Nineteen of these cells were completely inhibited by phenylephrine-induced elevations in BP (Table 2, Fig.2). In experiments using rats injected in the DBB with unconjugated saporin, 18 of 20 phasic neurons were inhibited by comparable increases in BP (Table 2, Fig. 2). A total of 36 phasic vasopressin neurons were recorded from animals with 192 IgG-saporin injections placed in the DBB. Thirty-three of these cells were completely inhibited by phenylephrine-induced BP elevations (Table2, Fig. 2). Analysis of BP sensitivity of phasic vasopressin neurons revealed no statistically significant changes in the number of phasic SON neurons displaying baroreceptor-induced inhibition after cholinergic DBB lesions (Table 2, P > 0.05).

The electrophysiological characteristics of the phasic neurons from each of the three experimental groups are summarized in Table3 along with the BP data. Because none of the DBB injections significantly affected the number of phasic neurons that were inhibited by baroreceptor stimulation, we did not separate the neurons responsive and unresponsive to baroreceptor stimulation for this analysis. Discharge properties of antidromically activated phasic SON neurons were not significantly different among the three groups. There was a significant difference for the peak BP produced by the phenylephrine injections (Table 3). The BP peak that was observed for the rats injected with the 192 IgG-saporin was significantly lower that for than the other two groups. There were no other differences in BP among the groups.

DBB neurons: retrograde labeling from the PNZ and ChAT immunoreactivity in normal and 192 IgG-saporin-injected rats.

Five rats had injections in the PNZ that did not include the SON (normal rats, n = 3; 192 IgG-saporin, n= 2). In each of these rats, the center of the injections was located in the lateral hypothalamus ∼1 mm posterior to bregma (23), dorsal and lateral to the SON (Fig.3). The injections covered from ∼0.5 mm posterior to 1.3 mm posterior to bregma. In normal rats and those injected with 192 IgG-saporin, the pattern of retrograde labeling was comparable. In each rat, the DBB contained labeled cells located in the vertical and horizontal limbs of the DBB, ipsilateral to the injection, throughout the rostral caudal length of the nucleus, as previously described (12). The neurons that were retrogradely labeled from the PNZ were separate and distinct from the ChAT immunofluorescence in the DBB (Figs. 4and 5). Sections taken from normal rats containing the vertical limb of the medial DBB displayed neurons retrogradely labeled from the PNZ that were primarily located lateral to ChAT-immunoreactive neurons (description noted above). Furthermore, none of the sections examined contained neurons that were both retrogradely labeled and ChAT positive (Figs. 4 and 5). The rostral portion of the DBB contained the lowest amount of retrograde labeling from the PNZ, and these retrogradely labeled cells were confined to the dorsal aspect of the rostral DBB and medial septum. ChAT-positive neurons in this part of the DBB were located below the area containing retrogradely labeled cells, and there were no doubly labeled neurons in this region. Retrograde labeling in the caudal DBB was primarily in regions bordering the lateral extents of the horizontal limb in the region of the medial forebrain bundle and ventral pallidum. In the caudal portion of the DBB, there were no retrogradely labeled neurons located within the DBB, and none of the ChAT-positive neurons were retrogradely labeled from the PNZ.

Sections containing the DBB from the animals previously injected with 192 IgG-saporin displayed similar patterns of retrograde labeling from the PNZ despite the elimination of ChAT-positive neurons (Fig. 5). In the rostral, medial, and caudal parts of the DBB from 192 IgG-saporin-injected rats, the pattern and number of retrogradely labeled neurons was comparable to the DBB neurons retrogradely labeled from the PNZ in sections taken from normal rats (Fig. 5).

DISCUSSION

The DBB has been implicated as an important region in the central pathway regulating the baroreceptor-induced inhibition of vasopressin release (5-7, 24). The baroreceptor-induced inhibition of magnocellular neurosecretory vasopressin cells in the SON may be mediated by DBB activation of the PNZ, a GABAergic interneuronal population surrounding the SON (14, 20, 24, 34). The DBB contains a variety of neurochemical cell types (2, 13, 15,21), some of which have distinct electrophysiological characteristics (12, 13, 18, 19). It is not known which of these cell types projects to the PNZ and thus participates in the baroreflex inhibition of vasopressinergic neurons in the SON, although a preliminary study suggested that cholinergic DBB neurons might be involved (4).

This experiment directly tested the role of cholinergic neurons of the DBB in the baroreceptor-mediated inhibition of SON vasopressin neurons. This was accomplished by lesioning these cells with an immunotoxin that is specific for cholinergic neurons (26, 33, 35, 38) and then examining the baroreceptor sensitivity and basal firing characteristics of phasic SON neurons. We verified the depletion of cholinergic neurons in the DBB using ChAT immunofluorescence. Our results demonstrate that DBB injections of 192 IgG-saporin nearly eliminated all of the ChAT-positive neurons in the rostral and medial portions of the DBB and reduced the number of ChAT-positive cells in the caudal DBB by ∼95%. Nevertheless, the elimination of cholinergic neurons in the DBB did not significantly alter the inhibitory effects of baroreceptor activation on phasic vasopressin SON neurons. Ninety-two percent of the phasic putative vasopressin neurons recorded from the SON of rats injected with 192 IgG-saporin were inhibited by baroreceptor activation. This is comparable to the 91% and 90% of phasic neurons inhibited by baroreceptor activation observed in the control groups. These data demonstrate that the cholinergic neurons in the DBB are not a necessary component of the pathway that mediates the baroreceptor-induced inhibition of vasopressin neurons in the SON.

This finding is supported by the data from our other study in which we combined retrograde tract tracing from the PNZ with ChAT immunofluorescence. In normal rats, rhodamine-labeled latex microspheres injected into the PNZ produced significant retrograde labeling in the DBB as has been previously described (14,34). When the retrograde labeling from the PNZ was combined with ChAT immunofluorescence, it revealed two largely separate populations of neurons within the DBB, cholinergic neurons, and retrogradely labeled neurons. In fact, we failed to observe any neurons in the DBB that were both retrogradely labeled from the PNZ and ChAT positive. This observation indicates that the population of DBB neurons that project to the PNZ is noncholinergic.

We also looked at the retrograde labeling of DBB neurons from the PNZ in rats previously injected in the DBB with 192 IgG-saporin. In these rats, the 192 IgG-saporin all but eliminated the ChAT-positive neurons in the DBB. However, the pattern of retrograde labeling from the PNZ was unaffected by lesioning the cholinergic neurons in the DBB. These data indicate again that cholinergic neurons in the DBB do not project to the PNZ and therefore may not be part of the pathway that ultimately brings baroreceptor information to the SON. In addition, these data demonstrate that the lesions produced by the 192 IgG-saporin were specific in that the noncholinergic DBB neurons that project to the PNZ were not affected by the immunotoxin. This is consistent with the results from other studies that have used 192 IgG-saporin to produce specific lesions of cholinergic neurons (22, 26, 33, 35,38). Electrophysiological studies of the DBB indicate that excitotoxic lesions of the vertical limb of the DBB largely eliminate the baroreflex inhibition of phasic SON neurons (6). If the 192 IgG-saporin had produced a nonspecific lesion, we would have expected to see more of an effect on the baroreflex-mediated inhibition of the vasopressin neurons in the SON. Although it could be argued that these observations themselves might be sufficient to demonstrate that the cholinergic neurons in the DBB are not part of the pathway from the DBB to the PNZ, it has been noted that many of the cholinergic cells in the DBB function as interneurons (38). Therefore, it is possible that the cholinergic cells could still be involved even if they do not directly project to the PNZ. This necessitates the functional study that we conducted.

Cholinergic neurons in the basal forebrain, which includes the DBB, are typically described as playing a functional role in behavioral processes, memory, and hippocampal function (22, 26, 33, 35,38). Functional studies employing cholinergic saporin lesions indicate that cholinergic DBB neurons participate in spatial working memory (22) but not avoidance behavior (38). The degree of cholinergic depletion associated with the impairment in spatial working memory (22) was comparable to the loss of cholinergic neurons produced in our study. The present study separated the DBB into three rostral-caudal extents to more accurately quantify the extent of cholinergic cell loss, and the lesion appeared nearly complete in all regions. The small number of neurons in the most caudal extent of the DBB in the lesioned rats was largely attributable to one animal and was in the most ventral and lateral extents of the nucleus, nearly at the level of the median preoptic nucleus. Therefore, the failure of the cholinergic lesions of the DBB to influence the activity of SON neurons was not likely due to an insufficient lesion. This contention is also supported by the results from our retrograde tract tracing studies that showed that PNZ-projecting DBB neurons were not lesioned by the 192 IgG-saporin.

Thus our results demonstrate that cholinergic neurons of the DBB do not project to the PNZ and are not a critical part of the central pathway that mediates the baroreceptor-induced inhibition of SON vasopressinergic neurons. Therefore, the neurochemical phenotype of the DBB neurons that do relay baroreceptor information to the PNZ remains undefined. To date, the most compelling evidence of cholinergic DBB innervation of the PNZ is the identification of putative cholinergic neurons in the DBB that were antidromically activated from the SON-PNZ region in a hypothalamic explant preparation (4). However, the antidromic activation of DBB neurons from the PNZ in the previous study did not discriminate between fibers of passage and axon terminals. This is further complicated by the PNZ's overlap with the medial forebrain bundle (23). This anatomical overlap could result in activation of DBB neurons projecting to more caudal regions such as the tuberomammillary nuclei (37) and not the PNZ.

The DBB contains a number of neuronal phenotypes that may be responsible for or alternatively influence the baroreflex inhibition of phasic SON neurons. Although the majority of DBB neurons are GABAergic, it is highly unlikely that the baroreceptor-related pathway from the DBB to the SON involves a GABA projection from the DBB to the PNZ. The current experimental evidence indicates that the DBB is responsible for activating a population of GABAergic neurons in the PNZ that, in turn, inhibit the activity of vasopressin neurons in the SON (20,24). Thus the baroreceptor-mediated activation of a population of GABA neurons in the DBB that project to GABAergic PNZ neurons is not consistent with the current experimental evidence.

In addition to substantial populations of GABAergic and cholinergic neurons, the DBB contains a number of other cell types, including neurons, that produce lutenizing hormone-releasing hormone, gonadotropin-releasing hormone, calretinin, vasopressin, substance P, neurotensin, and other neuropeptides (2, 10, 17, 15). A recent study also indicates that some DBB neurons may be glutaminergic (11). Evidence from experiments using hypothalamic brain-slice preparations indicate that local inhibitory interneurons to hypothalamic magnocellular neurons may be activated by glutamate (27, 32). Therefore, it is possible that glutamate is released by DBB neurons onto neurons in the PNZ, resulting in the inhibition of SON neurons. The excitotoxins employed in both DBB- and PNZ-lesion studies (6, 7, 20) that produced lesions that significantly reduce the number vasopressin neurons that are inhibited by baroreceptor activation require somatic glutamate receptors to produce cell death. This indicates that the DBB and PNZ neurons that participate in the baroreflex inhibition of vasopressin neurons in the SON have functional glutamate receptors. This is consistent with the notion that glutamate could serve as the neurotransmitter from the DBB to the PNZ. However, it has been demonstrated that gonadotropin-releasing hormone (GnRH) neurons in the DBB are resistant to glutamate excitotoxicity (10). Therefore, these excitotoxin-resistant GnRH neurons in the DBB may not be involved in the baroreceptor-mediated inhibition of vasopressin neurons in the SON. Determining which population of DBB neurons participates in the baroreceptor regulation of vasopressin-releasing neurons in the SON will be the focus of future investigations.

FOOTNOTES

• This work was done during the tenure of a predoctoral fellowship award to R. J. Grindstaff from the American Heart Association (AHA) Missouri affiliate and was supported by National Heart, Lung, and Blood Institute Grants T32-HL-07094 (R. R. Grindstaff), R29-HL-55692 (J. T. Cunningham), and K02-HL-03620 (J. T. Cunningham) and an AHA Missouri affiliate grant in aid (J. T. Cunningham).

• Address for reprint requests and other correspondence: J. T. Cunningham, Dalton Cardiovascular Research Center, Univ. of Missouri, Research Park Dr., Columbia, MO 65211.

• The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.