Disinhibition of neurons in the region of the dorsomedial hypothalamus (DMH) elicits sympathetically mediated tachycardia in rats through activation of the brain stem raphe pallidus (RP), and this same mechanism appears to be largely responsible for the increases in heart rate (HR) seen in air jet stress in this species. Neurons projecting to the RP from the DMH are said to be concentrated in a specific subregion, the dorsal hypothalamic area (DA). Here, we examined the hypothesis that the location of RP-projecting neurons in the DA correspond to the sites at which microinjection of bicuculline methiodide (BMI) evokes the greatest increases in HR. To determine the distribution of RP-projecting neurons in the DA, cholera toxin B was injected in the RP in four rats. A consistent pattern of retrograde labeling was seen in every rat. In the hypothalamus, RP-projecting neurons were most heavily concentrated midway between the mammillothalamic tract and the dorsal tip of the third ventricle dorsal to the dorsomedial hypothalamic nucleus ∼3.30 mm caudal to bregma. In a second series of experiments, the HR response to microinjections of BMI (2 pmol/5 nl; n = 76) was mapped at sites in the DA and surrounding areas in 22 urethane-anesthetized rats. All injection sites were located from 2.56 to 4.16 mm posterior to bregma, and the microinjections that evoked the largest increase in HR (i.e., >100 beats/min in some instances) were located in a region where RP-projecting neurons were most densely concentrated. Thus RP-projecting neurons in the DA may mediate DMH-induced tachycardia and thus play a role in stress-induced cardiac stimulation.
- dorsomedial hypothalamus
- sympathetic nervous system
- dorsal hypothalamic area
chemical stimulation or disinhibition of neurons in the region of the dorsomedial hypothalamus (DMH) with the GABAA receptor antagonist bicuculline methiodide (BMI) evokes physiological, endocrine, and behavioral changes that mimic the mammalian responses to emotional stress, including marked sympathetically mediated increases in heart rate (HR; see 1, 5, 6, 20, 21). In addition, microinjection of muscimol, a GABAA receptor agonist and neuronal inhibitor, in the region of the DMH before air stress nearly abolishes the increases in HR typically evoked in this paradigm (22, 23). These data support the notion that neurons in the region of the DMH mediate stress-induced tachycardia; however, the identity and exact location of these neurons within this region are currently unknown.
Disinhibition of neurons in the area of the medullary raphe pallidus (RP) elicits a pattern of cardiovascular changes similar to those seen after stimulation of the DMH (14, 19). Recently, the tachycardia evoked by either chemical stimulation of neurons in the region of the DMH or by experimental air stress was shown to be attenuated by microinjection of muscimol in the region of the RP (19, 26). These findings suggested that neurons in the RP play a role in both DMH-induced and stress-induced increases in HR. Retrograde tracing studies from the RP (9–11, 13, 16) showed that neurons that project to this area are highly concentrated in a specific subregion of the DMH known as the dorsal hypothalamic area (DA). These neurons would be likely candidates to mediate the increases in HR resulting from chemical stimulation of the DMH. Thus we hypothesized that the sites most sensitive to the tachycardic effect of microinjection of BMI in the region of the DMH would correspond to the region where these RP-projecting neurons in the DA were concentrated.
To test this hypothesis, we first confirmed the location of the RP-projecting neurons within the DMH by iontophoretic application of the retrograde neuron tracer cholera toxin B (CTb) in the region of the RP previously determined as mediating DMH- and stress-induced tachycardia (19, 26). We then targeted “nanoinjections” of BMI or microinjections of this agent at a greatly reduced dose (2 pmol) and volume (5 nl) to the region of the DMH with the greatest concentration of RP-projecting neurons and also to the surrounding area. The location of each injection site was analyzed for both the magnitude of the increase in HR evoked by the injection and the proximity of the injection site to the region of the DMH containing the densest collection of RP-projecting neurons.
MATERIALS AND METHODS
All experiments employed adult male Sprague Dawley rats (Harlan, Indianapolis, IN, or Charles River, Wilmington, MA). Animals were maintained on a 12:12-h light-dark cycle and allowed food and water ad libitum. All procedures conformed to the guidelines set forth by the National Institutes of Health and were approved by the Institutional Animal Care and Use Committee.
Experimental protocol: Iontophoretic injection and immunohistochemistry for CTb.
In four rats, the retrograde neuron tracer CTb (List, Campbell, CA) was injected in the RP to confirm the location of RP-projecting neurons within the DMH. The rats weighed between 275 and 350 g and were housed individually. On the day of the microinjection, the rats were anesthetized with pentobarbital sodium (50 mg/kg ip, and supplemented as needed) and placed in a stereotaxic instrument with the incisor bar lowered 3.3 mm below the interaural line. The CTb was reconstituted with 50 μl distilled water to a final concentration of 1.0% and backfilled in a glass micropipette. With the use of sterile technique, the tip of the micropipette (internal tip diameter of 10–15 μm) was then targeted to the RP [stereotaxic coordinates: anterior posterior (AP) −2.8 mm; mediolateral (ML) 0.0 mm; dorsoventral (DV) −1.0 mm with respect to the interaural point]. A constant current of +2.5 μA was applied for 15–20 min for iontophoretic application of CTb. After iontophoresis, the pipette remained in place for 10 min before being withdrawn. After the pipette was removed, the incision was closed, and the animal was returned to its home cage for recovery.
After 2 wk were allowed for transport of the tracer, animals were deeply anesthetized with pentobarbital sodium (100 mg/kg ip) and then transcardially perfused with 120 ml of 0.1 M PBS followed by 180 ml of 4% paraformaldehyde. The brains were removed, postfixed in 4% paraformaldehyde overnight, and then cryoprotected in 30% sucrose. Brains were then blocked, and 40-μm sections were cut on a cryostat. Sections were collected in 0.1 M PBS and stored at 4°C until they were processed as described below.
The protocol used for CTb immunohistochemistry was a modified version of the protocol previously reported by Pan and colleagues (17). Briefly, the sections were washed in 0.05 M PBS followed by 0.3% Triton X-100 in 0.05 M PBS (PBST). The sections were then incubated in goat anti-choleragenoid (List) diluted 1:90,000 in PBST for 16–20 h, biotinylated horse anti-goat secondary antibody (List) diluted 1:200 in PBST for 90 min, and avidin-biotin complex (ABC)-streptavidin complex (Vectastain ABC Elite Kit; Vector Laboratories, Burlingame, CA) for 90 min. Sections were washed in PBST between incubations. The colored reaction product was produced by incubation in a freshly prepared 0.025% 5′,5-diaminobenzidine tetrachloride-0.004% H2O2 solution for 1–20 min, depending on the intensity desired. The reaction was stopped by immersing the sections in 100 ml cold PBS.
Five out of every six sections were then mounted on presubbed slides, allowed to dry overnight, and then covered with a coverslip. The sixth section was mounted on presubbed slides, allowed to dry overnight, and counterstained with 1% Neutral Red dye. Sections were dehydrated in increasing concentrations of ethanol, cleared with xylene, and then covered with a coverslip. The location of retrogradely labeled neurons in the region of the DMH and surrounding areas was determined from unstained tissue sections, whereas the rostrocaudal level of the coronal section with respect to bregma was established by comparison of the Neutral Red-counterstained sections with the atlas of Paxinos and Watson (18). In one experiment, the location of every individual retrogradely labeled neuron on appropriate sections corresponding to those appearing in the atlas of Paxinos and Watson (18) was estimated and plotted on schematics adapted from the atlas.
Experimental protocol: Effect of microinjection of BMI in the DA and surrounding regions on HR in anesthetized rats.
The general microinjection technique employed for reproducible and precise delivery of small volumes has been employed previously (19) and is described in detail here. Glass micropipettes (1.0 mm OD and 0.25 mm ID; A-M Systems) were pulled on a vertical pipette puller to a long tapered tip (David Kopf Instruments) that was beveled at an angle of ∼45° to an outer diameter of 27–33 μm and an inner diameter of 9–15 μm. A glass ocular micrometer graduated in 0.1-mm increments was taped to the back of the micropipette and examined to ensure that the markings of the scale could be visualized and were aligned with the internal column of the micropipette just above the taper. The micropipette was then secured to a modified electrode clamp on the stereotaxic instrument. Plastic tubing was then used to connect the glass micropipette to a picospritzer. After the micropipettewas secured, animals were placed in a stereotaxic instrument (ASI) with the incisor bar raised 5 mm above the interaural line.
The 0.04-μm NeutrAvidin-labeled FluoSpheres (Molecular Probes, Eugene, OR) were added to a stock solution of BMI (4 mM in 0.1 M PBS; Sigma) that was then diluted in 0.1 M PBS to a final concentration of 0.4 mM BMI and 10% microspheres by volume. The fluorescent microspheres were added to the injectate to mark precisely all injection sites. With the use of a binocular dissecting microscope at ×45 magnification, the ocular micrometer was visualized, and the injectate was backfilled in the micropipette until the meniscus reached the top of the ocular micrometer scale. Before the micropipette was lowered in the brain for each injection, brief pulses of the picospritzer (see below) were delivered until the meniscus was positioned exactly at a gradation on the micrometer scale.
After baseline HR and blood pressure had stabilized, each animal received an initial microinjection of solution containing BMI (2 pmol/5 nl) and microspheres (10%) targeted to the DA (stereotaxic coordinates: AP −1.0 mm; ML +2.1 mm; DV −8.1 mm with respect to bregma) or a site in the surrounding region. These injections were performed using the glass micropipette fitted with a reticule eyepiece micrometer and connected to a picospritzer, as described above. The drug solution was delivered by applying multiple brief puffs of compressed nitrogen (8–15 psi) sufficient to produce small advances of the meniscus, and injection volume was estimated by using the binocular dissecting microscope to track the movement of the meniscus of the drug solution. From the inner diameter of the pipet, the volume of injectate delivered by movement of the meniscus by exactly one gradation (0.1 mm) was calculated to be 4.91 nl. Thus, by starting each injection with the meniscus positioned exactly over one marking and viewing the puff-induced movement of the meniscus down the internal column of the micropipette against the background of the ocular micrometer, delivery of injectate volume of 5 ± 0.2 nl could be easily and reliably estimated. Any changes in HR and blood pressure after such an injection were recorded, and the animal was allowed to return to a stable baseline over the next 25–30 min. The micropipette was then repositioned 300–500 μm anterior, posterior, medial, or lateral to the previous injection site, and the response to an identical microinjection of BMI was recorded. This process was then repeated, with no animal receiving more than four microinjections.
After the final microinjection, rats were transcardially perfused, and brains were removed, stored, and sectioned on a cryostat in a similar manner to that described above. Sections were collected in 0.1 M PBS and kept at 4°C until they were mounted on presubbed glass slides. All sections were allowed to dry overnight and were then counterstained with 1% Neutral Red dye, dehydrated in increasing concentrations of ethanol, cleared with xylene, and then covered with a coverslip. The pressure microinjection of the BMI-containing solution also delivered enough of the microspheres to permit visualization of the exact site of injection in mounted brain sections under ultraviolet (UV) light using a broadband rhodamine filter. The injection site was photographed using a model 1.3.0 Spot Digital Camera (Diagnostic Instruments) first under UV light and then under brightfield. The location of each injection site was then plotted on schematic sections adapted from the atlas of Paxinos and Watson (18) by an observer blinded to knowledge of the physiological response evoked at the site.
The baseline HR and blood pressure for each group were analyzed using a one-way ANOVA followed by Tukey's post hoc test. For the studies that attempted to examine the relationship between the magnitude and time course of BMI-evoked changes and the location of the densest accumulation of RP-projecting neurons in the DMH, linear regression analysis was employed for data from all injection sites judged to be in the single coronal plane representing 3.30 mm posterior to bregma (see below).
In all four animals receiving iontophoretic application of the retrograde neuron tracer CTb in the RP, distribution of labeled neurons was similar. Furthermore, this pattern of labeling confirmed the results of the similar studies of Hosoya and colleagues (11) in which wheat germ agglutinin-horseradish peroxidase had been used as a retrograde tracer. Because the results were nearly identical in all four rats tested, the exact pattern of labeling from a single representative experiment was used for Figs. 1–3 (see Figs. 1 and 2).
Hypothalamic neurons labeled with CTb were noted at each rostrocaudal level from the paraventricular nucleus (PVN) to the posterior hypothalamus according to the atlas of Paxinos and Watson (18). However, the distinguishing characteristic between labeling at these different levels was the increased density of labeled neurons in the region of the DA just dorsal to the dorsomedial hypothalamic nucleus (DMN) in sections where a distinct zona compacta was evident (see Figs. 1 and 2). The majority of these cells was in a cluster roughly 600 μm in diameter located from ∼3.14 to 3.60 mm caudal to bregma midway between the tip of the third ventricle and the mammillothalamic tract. Anterior to this area, in a region between the DMN and the PVN, CTb-positive neurons were more diffusely distributed, with no particular hypothalamic region containing a markedly greater concentration of cells. Few neurons were located in the ventral region of the hypothalamus or dorsal to the mammillothalamic tract. Moving posterior from the DA, the density of labeled cells decreased, and retrogradely labeled cells were located primarily more ventrally and laterally compared with the distribution seen in more anterior regions.
In a series of 22 animals, the increases in HR after microinjection of BMI (2 pmol/5 nl) in the DA and surrounding regions were recorded. A total of 76 injections (no more than 4 in a single rat) met the criterion for inclusion in this study (i.e., a stable baseline HR <400 beats/min). The average baseline HR before each microinjection was 360 ± 2 beats/min, and the corresponding average baseline blood pressure was 90 ± 1 mmHg (see Table 1). Injection sites were classified into three groups on the basis of the magnitude of the tachycardia produced by BMI, >50 beats/min (n = 21), 25–49 beats/min (n = 30), or <25 beats/min (n = 25). Regardless of the magnitude of the increase in HR, the chronotropic effect of microinjecting BMI in the DMH began within 30 s in all animals and often in as little as 5 s. Each animal returned to stable baseline HR within 30 min of the onset of the chronotropic effect. In one-half of the animals (11/22), the first injection evoked the largest increase in HR; in the remainder of the experiments, the greatest increases in HR were elicited by the second (5 experiments), third (2 experiments), or fourth (4 experiments) injections. There were no significant differences in baseline HR between any of the three groups, although the baseline blood pressure before injections that evoked increases in HR >50 beats/min was significantly higher than that in the other two groups (ANOVA followed by Tukey's post hoc test; see Table 1).
In general, injection sites at which microinjections increased HR by >50 beats/min (n = 21) were located in a region where RP-projecting neurons were most heavily concentrated in retrograde tracing experiments, midway between the mammillothalamic tract and the dorsal tip of the third ventricle dorsal to the DMN ∼3.30 mm caudal to bregma. Increases in HR of 25–49 beats/min were evoked from sites (n = 30) adjacent to this region, and, with the exception of several injections at sites just posterior, little or no response (<25 beats/min) was elicited at more distant sites (n = 25), including sites in the DMN itself.
To determine whether the increases in HR were related in magnitude to the increases in blood pressure, and the time-to-peak change in HR, linear regression analyses were performed. There was no significant interaction between the maximal change in HR and the time-to-peak increase in HR (R = 0.16, F = 2.04, P > 0.05). However, there was a weak but significant correlation between the changes in blood pressure from baseline and the change in HR from baseline (R = 0.38, F = 12.81, P < 0.05), indicating that greater increases in blood pressure tended to be associated with the largest increases in HR.
Linear regression analysis was also employed to examine the relationship between the intensity of the tachycardic response to microinjection of BMI and the relative location of the injection site with respect to the region in the DA where RP-projecting neurons were concentrated in retrograde-tracing experiments. Neurons retrogradely labeled from the RP were most dense in a rostrocaudal plane ∼3.30 mm posterior to bregma, and within this plane the densest concentration of labeled cells could be approximated as being one-half the distance from the tip of the third ventricle to the mammillothalamic tract. Therefore, for all injection sites judged to be in this single rostrocaudal plane (n = 18), the distance from the injection site to the point midway between the third ventricle and the mammilothalamic tract was estimated. This distance was found to be inversely correlated with the magnitude of the increase in HR evoked at that site (R = 0.60, F = 9.66, P < 0.05, n = 18). Thus the largest increases in HR evoked by microinjection of BMI tended to be associated with injection sites closest to the location of the densest collection of neurons in the DA that project to the RP.
The results of a representative experiment shown in Fig. 3 support this conclusion and also demonstrate the degree of resolution attainable with the microinjection technique employed. Here, the increase in HR evoked by injection of BMI in three sites varied widely but systematically. Injection at the first site, a site that corresponded closely to the point where, according to retrograde tracing experiments, RP-projecting neurons were most densely concentrated, resulted in an increase in HR of >100 beats/min. However, identical injection at a site 300 μm lateral to this increased HR by less than one-half this amount, and a third injection at a site still more lateral by 300–600 μm lateral to the first injection site had little discernible effect on HR. In contrast, a fourth identical microinjection 300 μm anterior to the first site resulted in an increase in HR nearly identical to that seen after the initial injection. Thus the progressive decrement of the HR response to injections at successively more lateral locations was not the result of loss of responsiveness of the preparation but was more likely a consequence of the relative density of RP-projecting neurons at the injection site.
The results of this study confirm that RP-projecting neurons in the region of the DMH are highly concentrated in a specific area of the dorsal hypothalamus (9, 11, 16) and demonstrate for the first time that sites at which microinjection of BMI produces the largest increases in HR correlate closely with this same region. The specific location of the densest accumulation of these RP-projecting neurons in the DA does not correspond to any of the anatomically defined nuclear regions of the hypothalamus as defined by Paxinos and Watson (18) but is adjacent to several hypothalamic areas of interest with regard to sympathetic nervous regulation, including the hypothalamic paraventricular nucleus (PVN), located <1 mm anteriorly. However, because of the degree of resolution demonstrated here (Fig. 3), the present results clearly define the most active region with respect to changes in HR as closely coincident with that at which RP-projecting neurons in the DA are concentrated.
To achieve this necessary degree of anatomical resolution, two refinements in our previous microinjection techniques were necessary. First, the location was marked by simultaneous delivery of BMI with fluorescent latex microspheres that, unlike microinjected drugs, are known to diffuse little if at all from the site of injection in the brain (15). Thus the exact site of injection (although not necessarily the sphere of influence affected by the more diffusible microinjected BMI) could be positively visualized using fluorescent microscopy (see Fig. 3) without the potential for error imposed by applying the markers with a second microinjection at the end of the study. Second, the dose of BMI and volume of injectate in which it was delivered (2 pmol/5 nl) represented reductions of 60–80 and 90–95%, respectively, over those used in our previous microinjection studies in this region (4, 19, 25). This reduction in both dose and volume allowed the comparison of increases in HR evoked by individual nanoinjections within the region of the DA and surrounding areas, whereas the degree of resolution achievable in previous studies had only been sufficient to differentiate responses evoked from much larger and more distinct regions (i.e., the DMH vs. the PVN, Bailey and DiMicco, 2001; DeNovellis, et al., 1995). Thus the magnitude of the increase in HR evoked by chemical stimulation of sites in the region of the DA correlated with the relative density of neurons retrogradely labeled from the RP at this site (see Figs. 1–3).
Our findings thus suggest that RP-projecting neurons in the DA are likely to mediate increases in HR resulting from chemical stimulation of the larger region of the DMH. Further support for this notion emerges from a reevaluation of our findings regarding the effect of identical bilateral microinjections of muscimol, a GABAA receptor agonist and neuronal inhibitor, in various regions of the hypothalamus between the PVN and the posterior hypothalamus (including the DMH) on air stress-induced tachycardia (23). Microinjection of muscimol in the region of the DMH but not the PVN resulted in marked suppression of the increases in HR evoked by 20 min of air stress. This finding suggested that activation of the same neurons in the region of the DMH is responsible for the increases in HR seen after microinjection of BMI and in response to air jet stress. However, within the DMH, the sites at which injections of muscimol provided the greatest degree of suppression were located specifically in the region of the DA shown in the present study to be the location of the greatest density of RP-projecting cells (see Fig. 4 in Ref. 23). Thus a reevaluation of Stotz-Potter and colleagues' study not only supports the present results but also points to a previously unrecognized role for RP-projecting neurons in the DA in mediating stress-induced tachycardia.
Our suggestion that DMH-induced tachycardia is mediated through a direct projection to the RP is based almost entirely on the following two key observations: 1) the present results showing that the specific sites where BMI acts in the DMH to produce this effect are coincident with the region in the DA where RP-projecting neurons are most densely concentrated and 2) our previous finding that microinjection of muscimol in the RP markedly reduces the tachycardia elicited from the DMH (19). However, an alternative interpretation is suggested by a report that microinjection of muscimol in the midbrain periaqueductal gray (PAG) nearly abolished the increases in HR elicited by microinjection of BMI in the DMH (3). Based upon their results, da Silva and coworkers (3) suggested that DMH-induced cardiac stimulation was mediated through projections to the PAG, an area known to receive afferent inputs from the DMH. Furthermore, anterograde tracing studies suggest that neurons in the DA project to the PAG in the rat (12), although the precise location of these PAG-projecting neurons has not been described in this species. If their distribution closely approximates that of RP-projecting neurons, then an alternative explanation for our findings is that disinhibition of the DMH elicits activation of the RP and tachycardia not through a direct projection but by means of a relay through neurons in the PAG. Nonetheless, in the absence of this knowledge, the close correspondence between functionally active sites in the DMH and the location of RP-projecting neurons seen in the present study seems unlikely to be coincidental and points to some other explanation for the data of da Silva and colleagues. One possibility stems from the fact that the region of the PAG studied may be a source of ascending glutamate-mediated excitatory drive to neurons in the DA. The region of the DMH receives afferent input from the PAG in rats (7, 24), and ascending input from the PAG appears to facilitate defensive rage reactions evoked at the level of the medial hypothalamus in cats (2), suggesting that this input is excitatory. We have shown previously that blockade of ionotropic glutamate receptors in the region of the DMH prevents or reverses the increases in HR that are seen when the area is disinhibited by BMI (21). If afferents from the PAG represent the source of this necessary tonic excitatory drive, then inhibiting the PAG with muscimol should have an effect similar to that reported by da Silva and colleagues.
Before this study, the location of the neurons responsible for evoking tachycardia after chemical stimulation of medial hypothalamic sites had been localized only to the general region of the DMH (4). Previous studies identified the RP as the principal intermediate relay along this pathway (19) and provided an opportunity to identify the location of these cells using well-established retrograde tracing techniques. Our study confirms the presence of a dense population of RP-projecting neurons located within the DA just dorsal to the dorsomedial nucleus at the level where a clear zona compacta is evident and provides the first evidence that activation of these neurons is largely responsible for the tachycardia resulting from chemical stimulation of the medial hypothalamus. Further study is needed to assess the potential roll of RP-projecting neurons in the DA in stress-induced tachycardia.
This work was supported by National Institutes of Health Grants NS-19883 and MH-65697.
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.
- Copyright © 2004 the American Physiological Society