0361-9230/90 $3.00 + .OO

Brain Research Bulletin, Vol. 24, pp. 577-582. 0 Pergamon Press pk. 1990. Printed in the U.S.A.

Electrophysiological Characterization of Reciprocal Connections Between the Parabrachial Nucleus and the Area Postrema in the Rat S. PAPAS AND A. V. FERGUSON’ Department of Physiology, Queen’s University, Kingston, Ontario, Canada K7L 3N6 Received 28 November

1989

PAPAS, S. AND A. V. PERGUSON. Electrophysiological characterization of reciprocal connections between the parabrachial nucleus and the area postrema in the rat. BRAIN RES BULL 24(4) 577-582, 1990. -Neuroanatomical studies have demonstrated reciprocal connections between the parabrachial nucleus (PBN) and both the area postnzma (AP) and the nucleus tractus solitarius (NTS). To functionally characterize these projections, antidromic identification of AP and NTS neurons projecting to the PBN was attempted. Orthodromic influences on these cells, resulting from PBN stimulation, were also examined. Four percent of AP neurons

tested (n = 74) were antidromically identified as projecting to the PBN [latency (L) = 26 + 4 msec, threshold current (T) = 79 f 11 kA] Parabrachial stimulation orthodromically influenced 24% of AP cells. Equal numbers of these neurons (12%) were excited [L = 25 + 9 msec, duration @) =29 * 14 msec] and inhibited (L=28+8 msec, D= 107 *4O msec). Of 46 NW neurons tested, 11% were antidrnmically identified as projecting to the PBN (L = 12 + 4 msec, T = 612 18 p,A), while orthodromic influences were seen in 41% of these neurons. Initial responses of 30% of the cells were excitatory (L = 34 2 14 msec, D = 63 2 24 msec), PBN stimulation inhibited the remaining 11% of NT’S neurons (L = 30 k 10 msec, D = 108 2 32 msec). These findings suggest that a functional heterogeneity

exists in the PBN efferents to the AP and NTS. However, the small proportion of antidromically identified AP and NTS efferents to the PBN disagrees with neuroanatomical studies suggesting a denser projection. Nucleus tracks

solitarius

Antidromic

Orthodromic

Parabrachial

area postrema (AP) and the adjacent nucleus tractus solitarius (NTS) are regions in the medulla which are strongly associated with central autonomic control. The AP is located on the dorsal surface of the medulla, at the level of the obex of the fourth ventricle. It is the most caudal of the circumventricular organs and characteristically lacks a normal blood-brain barrier. As a result of this modification of the normal blood-brain barrier, the AP provides a route by which systemic substances can influence neural elements within the central nervous system without first

afferents

Parabrachial

efferents

(24) results in a rise in arterial pressure, while AP stimulation elicits a depressor response (8). In addition, both the AP and the NTS contain high concentrations of angiotensin-II receptors (14) and are thought to play a role in mediating this peptide’s central influences on the regulation of body fluid balance (3,4,7, 11, 19). The AP and the medial NTS have also been linked with the lateral PBN with regards to angiotensin-II-induced thirst (18). Although the anatomical relationship of these 3 brainstem structures and their involvement in autonomic control have been well established, studies of the functional nature of the connections between the AP and NTS, and the PBN have not been undertaken. In this investigation, electrophysiological techniques have been used to antidromically identify projections from the AP and NTS to the PBN in anaesthetized rats. At the same time, orthodromic influences of PBN stimulation on neurons in these two areas were examined in order to define the function of these neural connections.

THE

having to cross this barrier to hydrophilic substances. In addition, both the AP and NTS receive afferent visceral information via the vagus (6,12). Afferent information may then be transferred to forebrain autonomic centres either directly or via the pontine parabrachial nucleus (PBN) (10, 17, 21-23, 26, 29). Neuroanatomical studies utilising both anterograde and retrograde tracing techniques have demonstrated reciprocal connections between the PBN and both the AP and NTS (17, 21, 23, 26) in the rat. Individually, each of these 3 regions have been implicated in cardiovascular regulation and in body fluid maintenance. Electrical stimulation of either the rat commissural NTS (8) or the PBN

METHOD

Experiments were performed on male Sprague-Dawley rats

‘Requestsfor reprints should be addressed to Dr. A. V. Ferguson, Department of Physiology, Queen’s University, Kingston, Ontario, Canada K7L 3~6.

577

578

PAPAS AND F~RGUSO~

(150-300 g) anaesthetized with urethane (1.4 g/kg) or sodium pentobarbital (65 mglkg). Similar electrophysiological data were obtained using both of these anaesthetics. Each animal was fitted with an indwelling femoral arterial catheter (PESO Intramedic) to monitor blood pressure, a jugular venous catheter to administer supplements of sodium pentobarbital when necessary, and an endotrachial tube to facilitate breathing. Throughout the experiment, a feedback controlled heating blanket was used to maintain body temperature at 37 -t 1°C. The animal was placed in a Kopf stereo&c frame and a monopolar or concentric bipolar (Rhodes Medical Ins~ments) stimulating electrode was positioned in the region of the PBN according to the coordinates of Paxinos and Watson (20). A small burr hole was made 8.0 mm posterior to bregma and 1.7 mm lateral to the sagittal midline. The electrode was lowered 6.5 mm ventral to the dura and attached to the skull using a jeweller’s screw and dental cement. The stimulating electrode was connected to an isolated stimulation unit programmed to deliver 0.1 msec current pulses, controlled by a digitimer. The head of the animal was then positioned vertically and midline incision made. The atlanto-occipital membrane was removed to expose the dorsal surface of the medulla, where the surface of the AP is clearly visible. A pressure foot was positioned over the AP to allow for more stable recordings. Single unit recordings were obtained using a NaCl (1 .O M) filled glass recording electrode (15-35 M0, tip diameter < 1 Fm). With the aid of a dissecting microscope, a micromanipulator was used to position the electrode on the dorsal midline surface of the AP and extracellular single unit activity was recorded to a maximal depth of 1000 km. Signals were amplified with a preamplifier, displayed on a digital oscilloscope, and were also fed to a window discriminator and to an on-line microcomputer programmed for spike train analysis in the form of peristimulus histograms. Neurons were antidromically identified as projecting to the PBN through the fulfillment of 3 criteria: an all-or-nothing constant latency response at threshold, collision cancellation of the antidromic spike with a spontaneous spike, and the ability to follow double pulses defivered at high frequency (>I00 Hz) of stimulation (9). O~~romic influences of electrical stimulation in the PBN (40-1~ &A) on the activity of AP or NTS neurons were monitored over a period of 200 stimulations on a peristimulus histogram. In many cases involving monopolar stimulating electrodes currents were often limited to values of 50 p.A or less in order to avoid limb movements induced by higher intensities of stimulation. Therefore, in 7 experiments animals were paralysed with 1.O mgikg of pancuronium bromide (PAVULON, Organon) and artificially ventilated. In this way, higher stimulation currents (up to 1000 (*A) could be used in an attempt to antidromically activate AP and NTS neurons. At the end of each experiment, the animal was fixed by intracardiac perfusion with saline followed by 10% formalin. Brains were stored in formalin for at least 24 hr and, 100 @rn and 50 )*rn coronal sections were taken through the PBN and AP, respectively. Sections were then mounted, stained with cresyl violet, and the anatomical locations of stimulating electrode tips and recording electrode tracts were verified.

A

B

-

O.lmm

FIG. 1. These p~otomicrograp~s illustrate the location of recording electrode tracts in the AP (A) and the NTS (B) as determined histologicaliy. In each case, the arrow indicates the position of the electrode tip.

1000 pm. Cells located in the first 400 pm of each penetration were classified as AP neurons (Fig. lA), in accordance with the anatomical dimensions of the structure. At depths between 600 and 1000 p+rnwe have assumed that all units recorded from were in the NTS (Fig. 1B). Neurons in the 200 pm range between these 2 regions were not included in the present analysis as we did not feel they could be reliably assigned to either the AP or the NT’S group, The locations of stimulating electrode tips were histologically determined to be in the PBN in 34 out of 46 excrements done (Fig. 2). Eighteen of these sites were located in the region of the dorsal PBN. The remaining 12 stimulating sites were situated outside of the PBN. AP Neurons

RESULTS

The maximum dorsal-ventral extent (500 pm) of the rat AP occurs beneath the midpoint of its dorsal surface (2). Thus, recording electrode penetrations were made near the midpoint of the AP’s dorsal surface in order to maximize the recording time spent in the AP. Penetrations were made to a maximum depth of

A total of 74 AP neurons were recorded from and tested with stimulation in the PBN. The spontaneous activities of the AP neurons ranged from less than 1 Hz (46% of neurons), to values greater than 10 Hz in the case of 1 cell. Most AP cells had firing frequencies in the range between 0 and 5 Hz (86%). Neurons were classified as being antidromically activated through the fulfillment

PARABRACHIAL

CONNECTIONS

579

WITH AREA POSTREMA

Rostra1 PBN

c>

Caudal PBN

1)

“.h SCP

.

B)

. : SrN n

-0.5

mm

-0.5

-0.5

mm

-

mm

diddle PBN

0.4 mm

FIG. 2. The schematic drawings (A-C) were taken from coronal sections of the rat PBN, at the rostra1 (A), middle (B), and caudal (C) regions of its rostro-caudal extent. Electrodes were located within the parabrachial region (circles) in 34 of the 46 experiments done. A photograph of a histological section showing a stimulating electrode site (arrow) in the dorsal rostra1 PBN is also shown (D). The squares (A-C) represent stimulating sites located outside of the PBN. DPBN-dorsal parabrachial nucleus, 4V-fourth ventricle, LC-locus coemleus, MTN-motor trigeminal nucleus, SCPsuperior cerebellar peduncle, STN-principal sensory trigeminal nucleus, VPBN-ventral parabrachial nucleus.

of 3 criteria. Examples of these criteria are illustrated in Fig. 3. Four percent (n = 3) of the tested AP cells were antidromically identified as projecting to the PBN. The average latency and threshold of the antidromic responses of these cells was 26 5 4 msec and 79 +- 11 pA, respectively (Table 1 -all values reported as mean 2 SEM). Assuming a distance of approximately 4.0 mm between the PBN and the AP, the conduction velocities of these cells would average 0.2 m/set. A larger proportion of the AP neurons (24%) were orthodromically influenced by parabrachial stimulation (Fig. 4). The initial effect of parabrachial stimulation on these neurons was either an excitation (12% of cells) or an inhibition (12% of cells) of activity (Fig. 5). The excitatory effects had an average latency of 25 2 9 msec and duration of 29 k 14 msec (Table 1). Inhibitory influences had an average latency of 28 + 8 msec and duration of 107 f 40 msec (Table 1). The activity of the remaining 72% of AP neurons tested with PBN stimulation was not affected. Twenty-one AP cells were tested with stimulation of the regions surrounding the PBN (Fig. 2). None of these AP neurons were antidromically activated by stimulation. The majority of these neurons (95%) were also unaffected by stimulation in PBN

at currents of up to 1000 p.A. However, 1 AP cell showed an enhanced level of activity with stimulation. In this animal histology showed the position of the stimulating electrode to be approximately 300 pm caudal to the PBN. It is thus possible that this orthodromic effect may have resulted from the activation of fibres of passage, projecting caudally from the PBN to the AP.

NTS Neurons The range of spontaneous activities of NTS cells was similar to that seen in the AP. Firing frequencies ranged from less than 1 Hz (41% of neurons) to greater than 10 Hz (3% of neurons), with most values (85%) falling in the 0 to 5 Hz range. Of the 46 NT’S cells tested with parabrachial stimulation, 11% (n = 5) were antidromitally identified as projecting to the PBN (Fig. 3). The antidromic responses of these cells had an average latency of 12 k 4 msec and threshold current of 612 18 pA (Table 1). The mean conduction velocity of these neurons was calculated to be approximately 0.3 m/set. As in the AP, a larger proportion (41%) of neurons tested were orthodromically influenced by stimulation of the PBN. An initial

580

PAPAS AND FERGlJSON

Antidromic Criteria

*

2. kL

t t _-i

0.5 mV

20ms FIG. 3. This figure is an exampb of the criteria which must be fulfilled in order to antidromically identify a neuron. These records are taken from an NT’Sneuron antidromically identified as projecting to the PBN. Each of the illustrated criteria is made up of 4 superimposed oscilloscope sweeps and

in each case the stimulus artifact is represented by an arrow. The antidromic criteria are: 1) an all-or-nothing constant latency response (evoked spike) to a threshold stimulus, 2) collision cancellation of the evoked spike with a sinuous spike(s) (the star represents where the ~ti~omic spike would normally occur), and 3) the ability of the evoked spike to follow high frequencies (1100 Hz) of stimulation.

excitatory response occurred in 30% of NTS neurons tested (Fig. 6) with an average latency and d~ation of 34114 msec and 63 Z!Z 24 msec, respectively (Table 1). The rem~ning 1I % of orthodromically influenced NTS cells showed an inhibition in activity after stimulation (Fig. 6). This decrease in activity had a latency of 30 + 10 msec on average and a duration of 108 + 32 msec (Table 1). Forty-eight percent of the NTS neurons tested with PBN stimulation were not affected. The effects of s~ulation in regions outside the PBN was examined in 17 NTS cells (Fig. 2). None of these cells could be ~ti~omic~ly activated by stimulation in these sites outside PBN. Twelve percent (n=2) of cells did show an increase in activity following stimulation. Histological verifications of the stimulating electrode sites being immediately caudal to the PBN in these cases again suggest the possibility of the activation of fibres of passage projecting to the AP as an expl~ation of these findings. DISCUSSION

The data presented in this investigation offer electrophysiological support for the existence of reciprocal connections between the PBN and both the AP and NTS. Neuroanatomical studies have suggested that predominant projections in this network travel from the AP (2326) and NTS (17,21) to the lateral PBN. Visceral info~ation then may travel via PBN efferents to several forebrain regions, which include the amygdala, the bed nucleus of the stria terminalis (16, 22, 29), and the hypothalamic and preoptic subnuclei (1, 16, 22, 25).

i 1:.‘-. -I

0.1 mV

20 ms

FIG. 4. These 4 digital oscilloscope sweeps were taken from an AP neuron orthodromicahy excited by parabrachiaf stimulation. The arrow in the first sweep represents the time of the stimulus artifact.

After considering the findings of these neuroanatomical studies, it was surprising to find that electrophysiological analysis revealed only a small number of AP (4%) and NTS (11%) neurons which could be antidromically identified as projecting to the PBN. These neurons were not localised to any specific region of either the AP or NTS. In many studies, attempts are made to limit the electrically induced activation of efferents, in this case from the AP and NTS, to a highly localised region around the stim~a~g electrode tip, thus lhniting the number of neurons intmenced. Taking into account the apparent conduction velocities of the few antkiromitally activated cells encountered in this study, the neurons in

581

PABABBACHIAL CONNECTIONS WITH AREA POSTRBMA TABLE 1 SUMMARYOF MEANLATENCIES AND

DURATIONS OF ANTIDROMIC (AD), EXCITATORY @XC) AND JNHBITORY (INH) ORTHODROMIC RESPONSES IN THE AP AND NTS (NE = NO EFFECT)

AP Neurons AD (n=3) EXC (n = 9) INB (n=9)

100

200

Time

300

400

600

bns)

PIG. 5. The peristimulus histograms shown in this figure were collected from AP neurons whose activities were excited (A) and inhibited (B) by parabrachial stimulation. Each histogram was collected over a period of ZOOstimulations. Tire filled bin in each case represents the time of the stimulus artifact.

cSEM

4 9 8

29 rt 14 107 rt 40 -

-

NTS Neurons AD (n=5) EXC (n = 14) INH (n=5) NE (n = 22)

0

Duration (msec)

c SEM

26k 25t 28k

NE (n=52)

B)

Latency (msec)

122 4 34 f 14 30 5 10 -

63 & 24 108 zk 32 -

of the AP (23) and over the entire NTS, although those carrying visceral information to the PBN are located in the caudal half (6, 1’7,21). Thus, the recording electrodes were in a position to record from these cells. It is possible, however, that the ~pulation of cells projecting to the PBN are not being recorded by our methods because of their small size (5, 13, 15). Neuroanatomical studies have determined PBN efferent connections with the rat AP (23) and the NTS (22) to be of a relatively

A)

question are most likely small in size and unmyelinated. Thus, one

would expect that higher currents would be required in order to activate such axons. Unfortunately, stimulus-induced movement is associated with the application of a higher stimulus intensity. In an attempt to overcome this diffkulty, paralysed and ventilated animaIs were used in some experiments, allowing us to employ currents up to 1 mA in an attempt to antidromically activate neurons. In all of our experiments, stimulation seemed to be well locahied as indicated by the observations that stimulation outside of the PBN (in 12 of the 46 experiments) produced no antidromic effects, and o~~o~c influences were seen only when the stim~ating electrode was in the vicinity of fibres of passage. Findings in the studies employing higher currents in the PBN did not differ from those in which lower currents were used, suggesting that the currents used in most of the experiments were sufficient to antidromically activate the AP and NTS efferents to the PBN. However, possible explanations for the few antidromitally identi~ed AP and NTS neurons do exist. Although tracer studies suggest that these efferent projections to the PBN are prominent, the density of the axons may have been exaggerated, for example by the spread of the injection site into adjacent areas. In fact, some neuroanatomical studies of AP projections have not mentioned the PBN as an efferent projection site (13, 27, 28). However, this explanation seems unlikely as retrograde studies in the PBN have demons~~d that the majority of AP neurons project to the PBN (26). The most likely explanation for the small number of antidromically identified neurons found appears to be that these cells were not being preferentially recorded. The population of neurons involved in the PBN projection are found in the caudal two-thirds

1

b-

II

0 100

0

200

I 300

5

4-

3-

2-

II

o0

I 100

IIIIII 200

Time

300

tmsf

PIG. 6. These peristimuhrs histograms were collected from NTS cells

which showed an excitatory (A) and an inhibitory (B) response to parabrachiaf stimulation. Each histogram was collected over a period of 200 stimulations. The filled bin in each histogram represents the time of the stimulus artifact.

582

PAPAS AND FERGUSON

light density when compared to its afferent projections from these 2 regions. Although the endings of our electrophysiologic~ investigation did not verify the existence of a large afferent projection to the PBN from the AP. our observations were consistent with the reported density of PBN efferents to the AP in tracer studies. PBN projections to the AP appear to be relatively light since only 24% of the AP neurons tested were orthodromitally influenced by PBN stimulation. However, o~hodromic irdluences were seen on a relatively large propo~ion (41%) of the NTS neurons recorded from, again supporting the anatomical demonstration of PBN efferents projecting to this region. The methods used in this investigation did not allow us to determine whether these effects of PBN stimulation were due to activation of mono- or polysynaptic pathways. The initial response of neurons in the AP and NTS to PBN stimulation was either excitatory or inhibitory in nature suggesting a functional heterogeneity in the projection. Although there was some variation in the latencies and durations of these responses, the average values of these parameters did not differ between the AP and the NTS @>0.05, Table 1). In addition, there did not

appear to be any organization of the neurons with respect to their response to PBN stimulation and their location in the AP or the NTS. Electrophysiological evidence has now been provided for the reciprocal connections between the PBN and both the AP and NTS of the rat. In addition, a step towards the understanding of the functional properties of these projections has been made. This network of connections provides a route by which visceral information may be relayed to forebrain regions involved in autonomic control, and perhaps undergo modification in the process. Electrophysiological techniques may now be employed to determine the relationship of the neuronal projections in this network, to the physiological role of these neurons. This may be of particular interest in the case of the AP, whose cells receive both neural and blood-borne info~ation (19). ACKNOWLEDGEMENTS

Thanks to Pauline Smith for her excellent technical assistance. This work was supported by the Heart and Stroke Foundation of Ontario and the Medical Research Council of Canada.

REFERENCES 1. Berk, M. L.; Finkelstein, J. A. Affetent projections to the preoptic

area and hypothalamic regions in the rat brain. Neuroscience 6: 1601-1624; 1981. 2. Brooks, M. J.; Hubbard, J. I.; Sirett, N. E. Extracellular recording in rat area postrema in vitro and the effects of cholinergic drugs. serotonin and angiotensin II. Brain Res. 261:85-90; 1983. 3. Casto, R.; Phillips, M. 1. Mechanism of pressor effects by ~giotensin in the nucleus tractus solitarius of rats. Am. J. Physiol. 247:R75-R8 1; 1984. 4. Casto, R.; Phillips, M. 1. Cardiovascular actions of microinjections of angiotensin II in the brain stem of rats. Am. J. Physiol. 246: R81 l-R816 1984. 5, Chemicky, C. L.; Barnes, K. L.; Conomy, J. P.; Ferrario. C. M. A mo~hoIogic~ cha~cte~~tion of the canine area postrema. Neurosci. Lett. Xh37-43; 1980. 6. Contreras, R. J.; Beckstead, R. M.; Norgren, R. The central projections of trigeminal, facial, glossopharyngeal and vagus nerves: an autoradiographic study in the rat. J. Auton. Nerv. Syst. 6:303-322; 1982. 7. Diz, D. I.; Barnes, K. L.; Ferrario, C. M. Hypotensive actions of microinj~tions of angio~nsin II into the dorsal motor nucleus of the vagus. J. Hypertens. 2:53-56; 1984. 8. Ferguson, A. V.; Marcus, P. Area postrema stimulation induced cardiovascular changes in the rat. Am. J. Physiol. 255:R8554860; 1988. tech9 Ferguson, A. V.; Renaud, L. P. In vivo electrophysiological niques in the study of peptidergic neurons and actions. In: Boulton, A. A.; Baker, G. B.; Pittman, Q. J., eds. Neuromethods. Clifton, NJ: The WumanaPress; 1987:379-408. of the efferent 10. Fulwiier, C. E.; Saper, C. B. Subnuclear org~ization connections of the parabrachial nucleus in the rat. Brain Res. Rev. 7:229-259; 1984. 11. Joy, M. D.; Lowe, R. D. Evidence that the ama postrema mediates the central cardiovascular response to angiotensin II. Nature 228: 13031304; 1970 12. Kalia, M.; Sullivan, J. M. Brainstem projections of sensory and motor com~nents of the vagus nerve in the rat. J. Camp. Neurol. 211: 248-264; 1982. 13. King, G. W. Topology of ascending brainstem projections to nucleus parabrachialis in the cat. J. Comp. Neural. 191:615-638; 1980. 14. Mendelsohn, F. A. 0.: Quirion, R.; Saavedra, J. M.; Aguilera G.

15. 16. 17. 18.

19.

20. 21.

22. 23. 24. 25

26.

27.

28.

29.

Autoradiographic localization of angiotensin II receptors in rat brain. Proc. Natl. Acad. Sci. USA 81:1575-1579; 1984. Morest, D. K. A study of the structure of the area postrema with Golgi methods. Am. J. Anat. 107:291-303; 1960. Norgren, R. Taste pathways to hypothalamus and amygdala. J. Comp. Neural. 16617-30; 1976. Norgren, R. Projections from the nucleus of the solitary tract in the rat. Neuroscience 3:207-218; 1978. Ohman, L. E.; Johnson, A. K. Brain stem mechanisms and the inhibition of angiotensin-induced drinking. Am. J. Physiol. 256: R264-R269; 1989. Papas, S.; Smith, P.; Ferguson, A. V. Electrophysiological evidence that systemic angiotensin influences rat area postrema neurons. Am. J. Physiol. 258:R70-R76; 1990. Paxinos, G.; Watson, C. The rat brain in stereotaxic coordinates. New York: Academic Press; 1982. Ricardo, J. A.; Koh, E. T. Anatomical evidence of direct projections from the nucleus of the solitary tract to the hypothalamus, amygdala, and other forebrain structures in the rat. Brain Res. 153:1-26; 1978. Saper, C. B.; Loewy, A. D. Efferent connections of the parabrachial nucleus in the rat. Brain Res. 197:291-317; 1980. Shapiro, R. E.; Miselis, R. R. The central neural connections of the area nostrema of the rat. J. Camp. Neuroi. 234:3&l--364, 1985. Sved: A. F. Pontine pressor sites which release vasopressin. Brain Res. 369:143-150; 1986. Tribollet, E.; Dreifuss, J. J. Localization of neurones projecting to the hypothalamic paraventricular nucleus area of the rat: a horseradish peroxidase study. Neuroscience 61315-1328; 1981. van der Kooy, D.; Koda, L. Y. ~g~ization of the projections of a ci~umven~cu1~ organ: the area postrema in the rat. 3. Camp. Neural. 219:328-338; 1983. Vigier, D.; Portalier, P. Efferent projections of the area postrema demonstrated by autoradiography. Arch. Ital. Biol. 117:308-324; 1979. Vigier, D.; Rouviere, A. Afferent and efferent connections of the area postrema demonstrated by the horseradish peroxidase method. Arch. Ital. Biol, 117:325-339; 1979. Voshart, K.; van der Kooy, D. The organization of the efferent projections of the parabrachial nucleus to the forebrain in the rat: a retrograde fluorescent double-labeling study. Brain Res. 212:271286; 1981.