Activation of brain neurons following central hypervolaemia and hypovolaemia: contribution of baroreceptor and non-baroreceptor inputs

Neuroscience Vol. 95, No. 2, pp. 499–511, 2000 499 Copyright q 1999 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306-4522/00 $20.00+0.00

Effects of blood volume changes on brain Fos expression

Pergamon PII: S0306-4522(99)00426-1 www.elsevier.com/locate/neuroscience

ACTIVATION OF BRAIN NEURONS FOLLOWING CENTRAL HYPERVOLAEMIA AND HYPOVOLAEMIA: CONTRIBUTION OF BARORECEPTOR AND NON-BARORECEPTOR INPUTS P. D. POTTS,* J. LUDBROOK,† T. A. GILLMAN-GASPARI,† J. HORIUCHI* and R. A. L. DAMPNEY*‡ *Department of Physiology and Institute for Biomedical Research, University of Sydney, Sydney NSW 2006, Australia †University of Melbourne Department of Surgery, Royal Melbourne Hospital, Melbourne, Victoria 3050, Australia

Abstract—In the present study we have used the detection of Fos, the protein product of c-fos, to determine the distribution of neurons in the medulla and hypothalamus that are activated by changes in central blood volume. Experiments were conducted in both barointact and barodenervated conscious rabbits, to determine the contribution of arterial baroreceptors to the pattern of Fos expression evoked by changes in central blood volume, induced either by intravenous infusion of an isotonic modified gelatin solution, or by partial occlusion of the vena cava. These procedures resulted in a significant increase and decrease, respectively, in right atrial pressure over a 60 min period. In control experiments, barointact and barodenervated rabbits were subjected to the identical procedures except that no changes in central blood volume were induced. In comparison with the control observations, central hypervolaemia produced a significant increase in the number of Fos-immunoreactive neurons in the nucleus tractus solitarius, area postrema, the caudal, intermediate and rostral parts of the ventrolateral medulla, supraoptic nucleus, paraventricular nucleus, arcuate nucleus, suprachiasmatic nucleus and median preoptic nucleus. The overall pattern of Fos expression induced by central hypervolaemia did not differ significantly between barointact and barodenervated animals. Similarly, the overall pattern of Fos expression induced by central hypovolaemia did not differ significantly between barointact and barodenervated animals, but did differ significantly from that produced by hypervolaemia. In particular, central hypovolaemia produced a significant increase in Fos expression in the same regions as above, but also in the subfornical organ and organum vasculosum lamina terminalis. In addition, compared with central hypervolaemia, hypovolaemia produced a significantly greater degree of Fos expression in the rostral ventrolateral medulla and supraoptic nucleus. Furthermore, double-labelling for tyrosine hydroxylase immunoreactivity demonstrated that neurons in the ventrolateral medulla that expressed Fos following hypovolaemia were predominantly catecholamine cells, whereas following hypervolaemia they were predominantly non-catecholamine cells. Finally, double-labelling for vasopressin immunoreactivity demonstrated that the number of Fos/vasopressin immunoreactive cells in the supraoptic nucleus was approximately 10 times greater following hypovolaemia compared with hypervolaemia, but there were very few such doublelabelled neurons in the paraventricular nucleus in response to either stimulus. The results demonstrate that central hypervolaemia and hypovolaemia each induces reproducible and specific patterns of Fos expression in the medulla and hypothalamus. The degree and pattern of Fos expression was unaffected by arterial baroreceptor denervation, indicating that it is primarily a consequence of inputs from cardiac receptors, together with an increase in the level of circulating hormones such as atrial natriuretic peptide, angiotensin II or vasopressin. Furthermore, the pattern of Fos expression produced by central hypervolaemia and hypovolaemia is distinctly different from that evoked by hypertension and hypotension, respectively [Li and Dampney (1994) Neuroscience 61, 613–634], particularly in hypothalamic regions. These findings therefore indicate that the central pathways activated by changes in blood volume are, at least in part, separate from those activated by changes in arterial pressure. q 1999 IBRO. Published by Elsevier Science Ltd. Key words: blood volume, central cardiovascular pathways, ventrolateral medulla, nucleus tractus solitarius, paraventricular nucleus, immediate-early genes.

Changes in central blood volume reflexly induce powerful compensatory responses. A decrease in blood volume caused by an acute haemorrhage results in a widespread activation of the sympathetic outflow to the heart, adrenal medulla and blood vessels together with an increase in the level of circulating angiotensin and vasopressin. 59 In contrast, an increase in blood volume leads to a reflex inhibition of renal sympathetic outflow, 10,14,47 although the effects on the activity of sympathetic outputs to other vascular beds and to the heart are less clear. 15 Inputs from both arterial baroreceptor and cardiac receptors contribute to these reflex effects, although the former group of receptors appears to have the dominant

role in triggering the response to hypovolaemia, while the latter group is more important in triggering the response to hypervolaemia. 6,36,38 There have been many studies in both anaesthetized and conscious animals examining the central pathways which subserve the responses to a decrease in central blood volume. In particular, studies in rats using the expression of the immediate-early gene c-fos, a marker of neuronal activation, have identified regions in the medulla and hypothalamus that are activated by acute hypovolaemia. 3–5,49,61,64 These studies were all performed on barointact animals, so that little is known of the relative contributions of arterial baroreceptor and non-baroreceptor inputs to the pattern of c-fos expression evoked in these regions by acute hypovolaemia, apart from one study which compared the effect of hypotensive and nonhypotensive haemorrhage on c-fos expression in the paraventricular nucleus and ventrolateral medulla (VLM). 4 Furthermore, although these studies demonstrated that neurons in the VLM and paraventricular nucleus in the

‡To whom correspondence should be addressed. Abbreviations: CI, cardiac index; HR, heart rate; -LI, -like immunoreactive; MAP, mean arterial pressure; NTS, nucleus tractus solitarius; OVLT, organum vasculosum of the lamina terminalis; RAP, right atrial pressure; SFO, subfornical organ; SVCI, systemic vascular conductance index; TH, tyrosine hydroxylase; VLM, ventrolateral medulla. 499

500

P. D. Potts et al.

hypothalamus were activated by hypovolaemia, they did not determine the extent to which catecholamine neurons and vasopressin-secreting neurons in these respective regions were involved. Recent studies have also used the c-fos functional mapping procedure to identify regions in the medulla and hypothalamus of the conscious rabbit 2 or rat 54 that are activated by hypervolaemia, but again the relative contributions of baroreceptor and non-baroreceptor inputs were not determined. In the present study, we have used the detection of Fos, the protein product of c-fos, to determine systematically and quantitatively the distribution of neurons in the medulla and hypothalamus that are activated by central hypervolaemia or central hypovolaemia in conscious rabbits. The arterial pressure, heart rate, right atrial pressure and cardiac output were measured during the periods of central hypervolaemia and hypovolaemia so that the haemodynamic effects associated with the observed pattern of Fos expression in the brain could also be determined. The contribution of arterial baroreceptors to the pattern of Fos expression following each of these stimuli was determined by comparing results obtained in intact animals and in animals with chronically denervated carotid sinus and aortic baroreceptors. Furthermore, we have combined Fos-labelling with immunohistochemical labelling for tyrosine hydroxylase or vasopressin to determine the extent to which catecholamine cells in the pons and medulla, and vasopressin cells in the hypothalamus, are activated by central hypervolaemia or hypovolaemia. The results obtained were compared with those obtained in previous studies from our laboratory in which the effects of hypertension and hypotension on Fos expression in conscious intact and barodenervated rabbits were determined. 31,52 EXPERIMENTAL PROCEDURES

Experiments were performed on male New Zealand White rabbits (Walter and Eliza Hall Institute for Medical Research breeding colony, body weight ˆ 2.64 ^ 0.11 kg). All experimental procedures were carried out in accordance with the Guidelines for Animal Experimentation of the National Health and Medical Research Council of Australia. Surgical procedures The surgical procedures have been described in detail previously. 36,38 They were done under halothane anaesthesia after induction with thiopental sodium (25 mg/kg i.v.) and endotracheal intubation. During intrathoracic procedures the rabbits were artificially ventilated. Post-operatively, a long-lasting analgesic buprenorphine (Temgesic, Reckitt and Coleman, 60 mg/kg) was administered subcutaneously. First, an inflatable cuff was placed around the thoracic vena cava (caval cuff) via a right thoracotomy. Two weeks later, a transit-time Doppler flow probe (Transonics Systems, Ithaca, Type 6S) was placed extrapericardially around the ascending aorta, via a left thoracotomy. In 11 rabbits, sinoaortic barodenervation was performed two weeks later. The aortic nerves were identified and divided, and the carotid bifurcations stripped and painted with 2% aqueous phenol. The experiments were performed seven to 10 days (barodenervated group) or 21– 24 days (barointact group) after the last surgical operation, at which time all rabbits were in good health and were gaining weight. On the day of the experiment, the following procedures were done under local anaesthesia (0.5% lignocaine). The tube leading to the caval cuff and the plug for the flow probe were retrieved from their subcutaneous positions. A catheter was inserted into each central ear artery, and advanced to the root of the ear, to measure arterial pressure and for blood sampling. A catheter was inserted into a marginal ear vein, for drug injection. Then the rabbit was lightly anaesthetized with thiopental sodium (15–25 mg/kg i.v.) supplemented by local

anaesthesia with 0.5% lignocaine. A double-lumen catheter was inserted via the right external jugular vein, so that its tip lay either in the right ventricle or in the abdominal vena cava. When the rabbit had fully recovered from the anaesthesia (within 15 min after insertion of the catheter) it was placed in its box and the catheter was then withdrawn into the right atrium. This location was identified by one of the following criteria: (i) the conversion of a right ventricular pressure tracing into a right atrial pressure tracing; or (ii) inflation of the caval cuff caused a fall, rather than a rise, in pressure. One lumen of the catheter was used to measure right atrial pressure, while the other was used for blood sampling. The experimental procedures commenced 4 h after completion of all preliminary procedures. Cardiorespiratory measurements Cardiac output (ascending aortic blood flow) was measured by connecting the flow probe to a meter (Transonic Systems, Ithaca, Model T206). Each probe had been calibrated in vitro by means of a roller pump and cellophane dialysis tubing, and the electronic zero was checked in vivo on an oscilloscope to ensure that end-diastolic flow was zero. 38 Arterial and right atrial pressures were measured by Viggo Spectramed P23XL strain gauges zeroed at heart level. Heart rate was measured by a tachometer triggered by the aortic flow pulse. The haematocrit was measured in duplicate using blood samples of 0.2 ml, taken from an ear artery. In all experimental protocols, haematocrit measurements were made 10 min before the experimental period began and again immediately after it ended. In the hypervolaemia protocol, the haematocrit was also measured at 15 min intervals during the experimental period. The haematocrit values were then used to estimate the percentage changes in blood volume, relative to the pre-experimental control period, at the end of or, in the case of hypervolaemia, during the experimental period. The arterial pO2 and pCO2 were measured from arterial blood samples of 0.6 ml (Radiometer ABL3 Gas Analyser, Copenhagen, Denmark). The samples were taken 10 min before and immediately after the end of the experimental period. Test for the effectiveness of sinoaortic denervation The baroreceptor–heart rate reflex was tested during the resting period before the start of the experimental protocol. This was done by inflating the caval cuff so that the arterial pressure decreased at a rate of 1–2 mmHg/s until the heart rate stabilized (usually within 20 s). In the barointact rabbits, this resulted in an increase in heart rate of 99 ^ 14 beats/min, whereas in the barodenervated rabbits the change in heart rate was 21 ^ 3 beats/min. Experimental protocols For a period of 4 h before the experiment started, and during the experimental period, the rabbits sat within a box with a wire mesh lid, under quiet environmental conditions at an ambient temperature of 21–238C. The haemodynamic variables were recorded on a Grass Model 7 Polygraph and the signals sent to an IBM-PC for A-D processing and computation of 60 s means. The primary variables were cardiac output, mean arterial pressure (MAP, mmHg), right atrial pressure (RAP, mmHg) and heart rate (HR, beats/min). The derived variables were cardiac index (CI, cardiac output/body weight, ml/min/kg) and the systemic vascular conductance index (SVCI, 100 × CI/MAP, expressed as units). Normovolaemic control. In this group of experiments the rabbits (two barointact and three barodenervated) remained undisturbed in their box during a 60 min period. Acute central hypervolaemia. This was induced in barointact (n ˆ 4) and barodenervated (n ˆ 4) rabbits by infusion of an isotonic modified gelatin solution (Haemaccel, Behring), maintained at 378C, via one lumen of the right atrial catheter. The total volume delivered over the 60 min experimental period was 48.6 ^ 1.7 ml/kg. It was given as a loading dose of 1.37 ml/kg/min over 12 min, followed by infusion at 0.67 ml/kg/min for the remainder of the 60 min period. Based on previous observations, 15 it was estimated this would result in a sustained increase in blood volume of 30–35% and in RAP of 3–5 mmHg.

501

Effects of blood volume changes on brain Fos expression

Acute central hypovolaemia. This was induced in barointact (n ˆ 4) and barodenervated (n ˆ 4) rabbits by inflation of the caval cuff by a saline-filled micrometer-driven syringe. 37 This was done in a graded fashion for 4 min so that CI fell by 8.5% of its baseline value per minute. This has been shown to produce a central hypovolaemia approximately equivalent to that produced by withdrawal of 24% of blood volume. 37 The CI was maintained at 66% of the baseline level for a further 56 min, after which the caval cuff was deflated and the CI returned to its baseline level. Following a waiting period of 45 min after the end of the period of central hypervolaemia or hypovolaemia, the rabbits were deeply anaesthetized with thiopental sodium (50–75 mg/kg, i.v.), intubated and artificially ventilated. The chest wall was opened and the flow probe removed from the aorta. A solution of heparin (10,000 IU) was injected intravenously, and the animal was perfused transcardially with 1 l of saline containing 0.5% sodium nitrite, followed by 2 l of 4% paraformaldehyde in 0.1 M phosphate-buffered saline (pH 7.2–7.4). The brain was removed and immersed in 30% sucrose in 0.1 M phosphatebuffered saline for 24–72 h. Five series of coronal sections (40 mm) were cut on a freezing microtome and stored in 1% azide in 0.1 M Trisbuffered saline at 48C. Staining procedure for Fos, tyrosine hydroxylase and vasopressin immunoreactivity The series of free-floating sections from each brain was processed for Fos immunoreactivity using an avidin–biotin–peroxidase procedure, as we have described in detail previously. 50 Sections from the medulla and pons were also stained for tyrosine hydroxylase (TH) immunoreactivity, and those from the hypothalamus for vasopressin immunoreactivity, after being stained for Fos immunoreactivity, according to the double-labelling procedure described in detail previously. 31,50 Microscopic analysis and quantitation The procedures for determining the distribution and number of Foslike immunoreactive (Fos-LI) neurons and double-labelled neurons within each brain region were the same as described in detail previously. 23,50 These procedures were performed by observers who were unaware of the experimental group pertaining to the brain sections that were examined. The distributions of single and doublelabelled neurons were mapped and quantified using the Magellan Image Analysis Program and a 486DX-33 computer. In regions where Fos-LI neurons extended over considerable rostrocaudal distances, such as the nucleus tractus solitarius (NTS) and VLM, several sections at different levels (approximately 0.4 mm apart) were mapped. Anatomical structures were identified by reference to the atlas of Meessen and Olszewski. 43 In the case of the VLM, this was divided into three parts—rostral (defined as the part extending from the level 1.8 mm to 3.0 mm rostral to the obex), intermediate (from the obex to the level 1.8 mm more rostral) and caudal (the part caudal to the obex). In each animal, the mean number per section of labelled neurons of each type (e.g., Fos-LI, TH or vasopressin immunoreactive, or doublelabelled neurons) was calculated for each region. This was done by counting bilaterally, for each region, the number of labelled neurons of each type in the sections that had been mapped. Except in the case of the area postrema, subfornical organ (SFO), organum vasculosum lamina terminalis (OVLT) and median preoptic nucleus, which are all midline structures, these numbers were then divided by two so that the final values presented in the results represent the means per section for one side. Statistical analysis Data are summarized as between-rabbit means ^ 1 S.E.M. For the purpose of analysing the cardiovascular variables, means over the 60 s immediately preceding the times 20, 40 and 60 min of the experimental period were used. The baseline values at time 0 min (also measured as the means of these values over the preceding 60 s) were analysed by two-factor ANOVA, the factors being baroreceptor status (intact or denervated) and protocol (normovolaemia control, hypervolaemia, hypovolaemia). The same procedure was used to analyse the blood gas results. The profiles of the haemodynamic variables over time (see Figs 1, 2) were compared by the interaction between baroreceptor

status and time within repeated measures ANOVA, using the Greenhouse–Geisser procedure to correct for serial autocorrelation. Preliminary examination of the counts of Fos-LI cells or doublelabelled cells in different brain regions showed that the group variances increased pari passu with the means. Therefore, the values were logtransformed before further analysis. The latter took the form of twoway ANOVA, the factors being baroreceptor status and protocol, baroreceptor status and brain region, or brain region and protocol. Subsequently, pairwise contrasts among the three protocols were made, using the Tukey–Kramer procedure to control the familywise type 1 error rate. The statistical analyses were performed using the software package SYSTAT 7.01 (SPSS, Chicago). A P-value ˆ 0.05 was taken to indicate a statistically significant effect. RESULTS

Haemodynamic effects of induced central hypervolaemia and hypovolaemia The baseline levels of MAP, RAP, CI and SVCI, measured at time 0 min, did not differ significantly according to baroreceptor status or protocol (P . 0.12 in all cases). For all rabbits, the baseline MAP, RAP, CI and SVCI were 91 ^ 2 mmHg, 21.2 ^ 0.4 mmHg, 143 ^ 5 ml/min/kg and 158 ^ 7 units, respectively. However, the baseline HR was significantly higher in the barodenervated than in the barointact rabbits (252 ^ 8 vs 217 ^ 17 beats/min, P , 0.05). Furthermore, the baseline pO2 and pCO2 levels in the arterial blood of the barodenervated rabbits (79 ^ 4 and 39.5 ^ 1.5 mmHg, respectively) were also significantly different from the levels in the barointact rabbits (90 ^ 4 and 33.2 ^ 1.0 mmHg, P , 0.005 in both cases). However, the pO2 and pCO2 levels in both barointact and barodenervated rabbits did not change significantly as a consequence of either induced central hypervolaemia or hypovolaemia (P . 0.8 in all cases). Normovolaemic control group. In the normovolaemic control rabbits (n ˆ 5), there was no consistent change in any of the haemodynamic variables over the 60 min period of observation in either the barointact or barodenervated rabbits (P . 0.14 in all cases). Acute central hypervolaemia. Infusion of Haemaccel over a period of 60 min resulted in an increase in blood volume (as calculated from the change in haematocrit) of 32 ^ 3% and 33 ^ 4% in the barointact and barodenervated rabbits, respectively. Over the 60 min experimental period, the time-course of change in the haemodynamic variables (Fig. 1) was not consistently different in the barointact and barodenervated groups (P . 0.25 in all cases). In both groups, there was an asymptotic rise in RAP, CI and SVCI, while there was a continuous increase in HR. The MAP showed no consistent change. Acute central hypovolaemia. Over the 60 min experimental period, the time-course of change in RAP, CI, MAP and HR (Fig. 2) was not consistently different in the barointact and barodenervated groups (P . 0.11 in all cases). There was a sustained fall in RAP, CI and MAP, and a modest rise in HR in both groups of animals. However, there was a significant difference in the time-course of change of SVCI in the barointact and barodenervated groups (P , 0.05). In the barointact rabbits, SVCI fell abruptly to and remained at a lower level during the period of central hypovolaemia, but in the

502

P. D. Potts et al.

Fig. 1. Graphs showing the changes in mean arterial pressure (MAP), cardiac index (CI), right atrial pressure (RAP), heart rate (HR) and systemic vascular conductance index (SVCI) during a 60 min period of hypervolaemia induced by infusing Haemaccel solution (beginning at time 0 min) in barointact and barodenervated rabbits. For all cardiovascular variables there was no significant difference between the barointact and barodenervated groups in the timecourse of change in response to hypervolaemia, as indicated by the P-values (.0.25 in all cases) shown on the graphs.

Fig. 2. Graphs showing the changes in mean arterial pressure (MAP), cardiac index (CI), right atrial pressure (RAP), heart rate (HR) and systemic vascular conductance index (SVCI) during a 60 min period of central hypovolaemia induced by partial occlusion of the vena cava (beginning at time 0 min) in barointact and barodenervated rabbits. As indicated by the P-values shown on the graphs, there was a statistically significant difference between the barointact and barodenervated groups in the time-course of change in SVCI in response to hypervolaemia, but no significant difference in the case of all other cardiovascular variables.

503

Effects of blood volume changes on brain Fos expression

Table 1. Number of Fos-like immunoreactive neurons per section in different brain regions in control rabbits and following a period of hypervolaemia or hypovolaemia in barointact and barodenervated rabbits Hypervolaemia Control (n ˆ 5)

Hypovolaemia

Intact (n ˆ 4)

Denervated (n ˆ 4)

Intact (n ˆ 4)

Denervated (n ˆ 4)

19 ^ 4 18 ^ 9 9^1 18 ^ 1 19 ^ 2

19 ^ 5 7^1 13 ^ 4 20 ^ 5 19 ^ 4

16 ^ 3 28 ^ 11 12 ^ 2 22 ^ 4 27 ^ 5

12 ^ 3 15 ^ 3 10 ^ 1 25 ^ 3 28 ^ 3

hypo ˆ hyper . . . control hypo . hyper . . . control hypo ˆ hyper . . . control hypo ˆ hyper . . . control hypo . . . hyper . . . control

6^1 11 ^ 1 16 ^ 2 8^1

23 ^ 7 60 ^ 11 68 ^ 9 30 ^ 5

24 ^ 17 74 ^ 16 63 ^ 10 20 ^ 3

128 ^ 17 70 ^ 18 90 ^ 14 27 ^ 7

117 ^ 21 55 ^ 14 104 ^ 6 31 ^ 4

hypo . . . hyper . . . control hypo ˆ hyper . . . control hypo . .hyper . . . control hypo ˆ hyper . . . control

4 ^ 0.3 9 ^ 0.2 19 ^ 1

12 ^ 7 23 ^ 6 43 ^ 15

6^1 29 ^ 3 34 ^ 4

26 ^ 8 49 ^ 5 138 ^ 24

19 ^ 3 39 ^ 8 122 ^ 20

hypo . . . hyper ˆ control hypo . hyper . . . control hypo . . . hyper ˆ control

Medulla NTS AP CVLM IVLM RVLM Hypothalamus SON PVN AN ScH Lamina terminalis SFO MnPO OVLT

4^1 2 ^ 0.2 5^1 6^1 6^1

Pairwise contrasts

Values are mean ^ S.E.M. Pairwise contrasts were made among the three protocols, using the Tukey–Kramer procedure. . . . corresponds to P , 0.001, . . to P , 0.01, . to P , 0.05, and ˆ to P . 0.05. AN, arcuate nucleus; AP, area postrema; CVLM, caudal ventrolateral medulla; IVLM, intermediate ventrolateral medulla; MnPO, median preoptic nucleus; NTS, nucleus tractus solitarius; OVLT, organum vasculosum lamina terminalis; PVN, paraventricular nucleus; RVLM, rostral ventrolateral medulla; ScH, suprachiasmatic nucleus; SFO, subfornical organ; SON, supraoptic nucleus.

Fig. 3. Drawings of sections through the medulla at the level approximately 1 mm rostral to the obex, showing the distribution of Fos-LI cells following central hypervolaemia in representative experiments in a barointact and a barodenervated rabbit. Each small filled circle represents one Fos-LI cell in a 40 mm-thick section. DmV, dorsal motor nucleus of the vagus; IO, inferior olive; NA, nucleus ambiguus; TS, tractus solitarius; Vsp, spinal trigeminal nucleus; XII, hypoglossal nucleus.

barodenervated rabbits, there was only a small initial fall followed by a rise to above-baseline levels. Fos expression following central hypervolaemia and hypovolaemia Normovolaemic control group. In the control group of animals, a low level of Fos expression was observed in all brain regions examined. The pattern and degree of Fos expression were similar in both the barointact and barodenervated control animals, in confirmation of previous studies in which a low baseline level of Fos expression was observed in both barointact and barodenervated rabbits. 31,52 The data for the barointact and barodenervated control animals in this

study were therefore grouped for the purpose of comparison with the rabbits in which central hypervolaemia or hypovolaemia was produced (Table 1). Acute central hypervolaemia. Central hypervolaemia resulted in a significant increase, compared with the control group of animals, in the degree of Fos expression in a number of regions in the medulla and hypothalamus in both barointact and barodenervated animals (Table 1). However, ANOVA showed that there was no significant effect of baroreceptor status on the pattern of Fos expression in the different brain regions examined (P . 0.7). In particular, the distribution and density of Fos-LI cells in the NTS and VLM were very similar whether the baroreceptors were intact or denervated, as

504

P. D. Potts et al.

Table 2. Number of Fos-like immunoreactive neurons immunoreactive for tyrosine hydroxylase or vasopressin in different brain regions in control experiments and following a period of hypervolaemia or hypovolaemia Control (n ˆ 3)

Hypervolaemia (n ˆ 8)

Medulla (Fos/tyrosine hydroxylase neurons) NTS 1 ^ 0.3 5 ^ 1* CVLM 2^1 4^1 IVLM 3 ^ 0.1 4^1 RVLM 2^1 6 ^ 1* Hypothalamus (Fos/vasopressin neurons) PVN 0.4 ^ 0.1 SON 12 ^ 5

Hypovolaemia (n ˆ 8) 5 ^ 1* 6 ^ 1* 14 ^ 2*† 13 ^ 1*† 2^1 120 ^ 7†

Values are mean ^ S.E.M. *P , 0.05 compared with corresponding values in the control group; †P , 0.05 compared with corresponding values in the hypervolaemia group. CVLM, caudal ventrolateral medulla; IVLM, intermediate ventrolateral medulla; NTS, nucleus tractus solitarius; PVN, paraventricular nucleus; RVLM, rostral ventrolateral medulla; SFO, subfornical organ; SON, supraoptic nucleus.

shown for example in Fig. 3. However, there was a highly significant difference in the overall pattern of Fos expression when the group of rabbits subjected to hypervolaemia was compared with the control group (P , 0.001). As shown in Table 1, in all brain regions except the SFO and OVLT the number of Fos-LI cells following hypervolaemia was significantly greater than in the control group (P , 0.001 in all cases). In medullary and hypothalamic regions outside those listed in Table 1 there was little or no Fos expression following central hypervolaemia, as was also the case in the control group of animals. Acute central hypovolaemia. Similarly, a period of induced central hypovolaemia also resulted in a significant increase, compared with the control group of animals, in the degree of Fos expression in a number of regions (Table 1). As with the hypervolaemic stimulus, there was no significant effect of baroreceptor status on the pattern of Fos expression in the different brain regions examined (P . 0.3), nor was there any significant degree of Fos expression in medullary and hypothalamic regions outside those listed in Table 1. However, there was a highly significant difference in the overall pattern of Fos expression when the hypovolaemia group was compared with the control group and also when compared with the hypervolaemia group (P , 0.001 in both cases). Medulla oblongata. Following either hypervolaemia or hypovolaemia, there was increased Fos expression, compared with the control group, in the NTS, area postrema, and in the rostral, caudal and intermediate parts of the VLM (Table 1). The major differences between the hypervolaemia and hypovolaemia groups were in the rostral VLM and area postrema, where the numbers of Fos-LI cells were significantly greater following hypovolaemia than following hypervolaemia (Table 1). The distribution of Fos-LI neurons was similar following either hypervolaemia or hypovolaemia. In the NTS, Fos-LI neurons were found mainly at or close to (within 1.5 mm) the level of the obex, while in the VLM they were located mainly ventrolateral to the nucleus ambiguus. Neurons that were immunoreactive for both Fos and TH were also found in the NTS and in all parts of the VLM following either central hypervolaemia or hypovolaemia.

However, the number of double-labelled Fos/TH neurons in the caudal, intermediate and rostral parts of the VLM, expressed as a percentage of the number of Fos-LI cells in each of these regions, was much lower following hypervolaemia compared with hypovolaemia (caudal VLM: 38 ^ 3 vs 56 ^ 3%; intermediate VLM: 35 ^ 4 vs 61 ^ 3%; rostral VLM: 31 ^ 3 vs 48 ^ 3%). These differences were highly significant (P , 0.001 in all cases). Overall, there were approximately twice as many double-labelled Fos/TH neurons in the rostral, intermediate and caudal parts of the VLM following hypovolaemia compared with hypervolaemia (Table 2). The distribution of Fos/TH neurons in the rostral and intermediate VLM in a typical experiment in which central hypovolaemia was induced, together with a photomicrograph of double-labelled neurons, is shown in Fig. 4. Hypothalamus. There were increased numbers of Fos-LI cells in the paraventricular, supraoptic, arcuate and suprachiasmatic nuclei in the hypothalamus following both hypervolaemia and hypovolaemia (Table 1, Fig. 5). A comparison between the hypovolaemia and hypervolaemia groups revealed no difference in the numbers of Fos-LI cells in the paraventricular and suprachiasmatic nuclei, but a much greater (approximately five times) number of Fos-LI cells in the supraoptic nucleus following hypovolaemia (Table 1; Fig. 5C, D). There was also a significantly greater number of Fos-LI cells in the arcuate nucleus following hypovolaemia compared with hypervolaemia (Table 1). Double labelling for vasopressin immunoreactivity showed, in confirmation of previous studies (e.g., see Ref. 31), many neurons in the paraventricular and supraoptic nuclei that were immunoreactive for vasopressin. In the paraventricular nucleus, however, very few Fos-LI neurons (average one to two per section) were double labelled for vasopressin following either hypervolaemia or hypovolaemia. In the supraoptic nucleus there was a very large number of double-labelled Fos/vasopressin neurons following hypovolaemia but a much smaller number following hypervolaemia (Fig. 5, Table 2). Furthermore, whereas the doublelabelled Fos/vasopressin neurons represented approximately 90% of all Fos-LI cells in the supraoptic nucleus following hypovolaemia, this proportion was much lower (approximately 40%) following hypervolaemia. Lamina terminalis. In the circumventricular organs (SFO and OVLT), there was no significant increase, compared with the controls, in Fos expression following hypervolaemia. In contrast, there was a very marked increase in the number of Fos-LI cells in both of these regions following hypovolaemia (Table 1, Fig. 5F). Fos expression in the median preoptic nucleus increased significantly following both hypervolaemia and hypovolaemia, although the increase was greater following hypovolaemia (Table 1). DISCUSSION

The results of this study show for the first time that the degree and pattern of Fos expression in the brainstem and hypothalamus in response to both central hypervolaemia and hypovolaemia are not significantly affected by inputs from arterial baroreceptors. However, there are clear-cut differences in the pattern of Fos expression induced by central hypovolaemia compared with central hypervolaemia.

Effects of blood volume changes on brain Fos expression

505

Fig. 4. Drawings of sections through the ventrolateral medulla at the intermediate (A) and rostral (B) levels, showing the distribution of Fos-LI neurons (indicated by small filled circles) and double-labelled Fos/TH neurons (indicated by small filled circles inside larger open circles) following central hypovolaemia in a representative experiment. The bright-field photomicrograph in (C) shows examples of a Fos-LI neuron (indicated by a short double arrow) and Fos/TH neurons (indicated by a longer single arrow). IO, inferior olive; NA, nucleus ambiguus; P, pyramid; Vsp, spinal trigeminal nucleus.

Furthermore, there are also clear differences in the patterns of Fos expression resulting from central hypervolaemia and hypovolaemia, and those resulting from hypertension and hypotension, respectively, as determined in previous studies from our laboratory. 31,50,52 The nature of these differences, and their implications in elucidating the central mechanisms that subserve the reflex response to changes in central blood volume, are discussed in detail below. Methodological considerations Experiments in this study were performed on conscious animals, and thus avoided the major problem that anaesthetic agents have an important effect on the function of central neurons, particularly in forebrain regions which were a major focus of this study. 28 We have previously discussed in detail the limitations of the method of Fos expression as a marker of neuronal activation, 31,52 so these will be only briefly summarized here. The control group of experiments,

in which rabbits were subjected to the same surgical procedures as in the hypervolaemia and hypovolaemia groups, demonstrated that the baseline level of Fos expression was very low under the experimental conditions of this study. With regard to the sensitivity of Fos expression, it has been shown that in comparison to other immediate-early geneencoded proteins such as Fos B, Jun, Jun B, Jun D and Krox-24, Fos appears to be the most effective marker of neuronal activation under a variety of experimental conditions. 29 Nevertheless, not all activated neurons express Fos. For example, Fos expression in somatic motoneurons in the spinal cord is either absent or considerably delayed following stimulation. 21 Furthermore, we have previously shown that Fos expression in central neurons (e.g., those subserving the baroreceptor reflex) will only occur if the stimulus is of a substantial magnitude and duration. 31 Thus, a small perturbation in arterial pressure, although sufficient to evoke a reflex sympathetic response, may not necessarily

506

P. D. Potts et al.

Fig. 5. Drawings of sections through the hypothalamus at the level of the paraventricular nucleus (A and B), supraoptic nucleus (C and D) and organum vasculosum lamina terminalis (E and F), showing the distribution of Fos-LI neurons (indicated by small filled circles) and double-labelled Fos/vasopressin neurons (indicated by small filled circles inside larger open circles) following central hypervolaemia (A, C, E) or central hypovolaemia (B, D, F) in representative experiments. BNST, bed nucleus of the stria terminalis; IC, internal capsule; OT, optic tract; OVLT, organum vasculosum lamina terminalis; PVN, paraventricular nucleus; ScH, suprachiasmatic nucleus; SON, supraoptic nucleus.

result in Fos expression in the central pathways mediating the reflex. These factors therefore need to be considered when interpreting the results of the present study. The procedure used to denervate the carotid sinus and aortic baroreceptors was the same as in previous studies from our laboratory. 51,52 The effectiveness of the denervation was indicated by the fact that there was no reflex tachycardia in response to an induced decrease in arterial pressure, although it was very marked in the barointact animals. Nature of the stimuli Central hypervolaemia. The method of infusing Haemaccel as a means of producing hypervolaemia was chosen because it has been shown in a previous study in conscious rabbits to produce similar increases in blood volume and right atrial pressure as those resulting from infusion of whole blood at the same rate. 15 The increases in blood volume, as estimated from the changes in haematocrit, were very similar in barointact and barodenervated rabbits (32 and 33%, respectively). It is most unlikely that inputs from arterial baroreceptors or chemoreceptors contributed significantly to the Fos expression evoked by central hypervolaemia, because the pattern of haemodynamic changes as well as that of Fos expression

in the brain were very similar in both intact and sinoaortic denervated rabbits (in which both arterial baroreceptors and chemoreceptors were denervated). Furthermore, there was no significant change in either arterial pressure or blood pO2 during the period of hypervolaemia in the intact group of animals. It therefore seems most likely that the main input that was responsible for Fos expression in central neurons following the period of hypervolaemia originated from cardiac receptors, particularly atrial receptors. This is consistent with previous studies showing that cardiac receptors appear to have the dominant role in triggering the reflex response to hypervolaemia. 6,36,38 In addition, volume loading causes the release of atrial natriuretic peptide 2,27 which may also influence brain neurons via an action on neurons within the circumventricular organs. 24 Central hypovolaemia. The method used to produce central hypovolaemia (graded inflation of a caval cuff) has been shown to be a highly reliable method of reducing central blood volume in unanaesthetized rabbits. 37 It has been shown previously, 36,38 and confirmed in the present study, that there is a compensatory decrease in systemic vascular conductance during the period of central hypovolaemia mediated by the arterial baroreceptor reflex which has the

Effects of blood volume changes on brain Fos expression

effect of minimizing the fall in arterial pressure. However, the fact that the degree and pattern of Fos expression following hypovolaemia were very similar in intact and barodenervated rabbits indicates that unloading of arterial baroreceptors was not the primary stimulus to the evoked Fos expression. This is further supported by the observation, as discussed in more detail below, that there were marked differences between the pattern of Fos expression in response to central hypovolaemia compared with that produced by hypotension, which is mainly mediated by arterial baroreceptors. 52 The lack of contribution of baroreceptor inputs to the Fos expression could be explained by the fact that the decrease in mean arterial pressure in the barointact group during the period of central hypovolaemia was modest (approximately 10– 15 mmHg) and may not have been sufficient, by itself, to evoke a significant level of Fos expression in central neurons. Consistent with this, Li and Dampney 31 have shown previously that a sustained change in mean arterial pressure of 10 mmHg in barointact conscious rabbits results in relatively little Fos expression in the brain. At the same time, it could be argued that baroreceptor inputs do make a small contribution to the pattern of Fos expression evoked by central hypovolaemia, but that this was not revealed by the statistical comparison between the barointact and barodenervated groups because of the small number of animals in each group. However, the P-value that resulted from comparison of the overall pattern of Fos expression between the two groups (using two-way ANOVA) was .0.3, and thus not close to the level of statistical significance. Thus, although it is possible that there are small differences between barointact and barodenervated animals in the patterns of brain Fos expression evoked by central hypovolaemia, it would appear that much larger sample sizes would be required to reveal such differences. It is unlikely that arterial chemoreceptors contributed significantly to the Fos expression evoked by central hypovolaemia, because this did not result in a significant change in arterial blood pO2. However, the central hypovolaemia would cause an unloading of cardiac receptors, as measured by the decrease in right atrial pressure, and this is therefore a likely factor responsible for the Fos expression. In addition, it has been shown that central hypovolaemia produced in conscious rabbits by the same procedure as in the present study results in a marked activation of the renin–angiotensin system. 37 Thus, another likely factor contributing to the Fos expression is an increase in the level of circulating angiotensin II, acting on neurons in circumventricular organs of the lamina terminalis (SFO and OVLT). Consistent with this, a high level of Fos expression was observed in these nuclei following central hypovolaemia. Comparison of the effects of hypervolaemia hypovolaemia on activation of brain neurons

and

Medulla oblongata. In the medulla, both central hypervolaemia and hypovolaemia evoked an increase in Fos expression (compared with control animals) in the NTS, area postrema, and the caudal, intermediate and rostral VLM, although there were clear differences in the degree or pattern of labelling in all of these regions, except for the NTS. In the case of the NTS, the degree of Fos expression, and the distribution of Fos-LI neurons within the nucleus, were similar following either central hypervolaemia or hypovolaemia.

507

This is consistent with a previous electrophysiological study showing that some NTS neurons are excited, whereas others are inhibited, by an increase in central blood volume, 22 and is also very similar to the results of a previous study showing that hypertension and hypotension each results in a similar degree and pattern of Fos expression in the NTS. 31 A major difference, however, between the effect of changes in arterial pressure on Fos expression in the NTS and the effect of changes in blood volume is that the former is dependent upon arterial baroreceptor inputs, 52 whereas the latter, as discussed above, is likely to be due to inputs from cardiac receptors. A second difference is that the degree of Fos expression evoked by a large increase in blood volume (of approximately 30%) results in a much smaller number of FosLI cells in the NTS than that observed following a moderate rise in arterial pressure of 20 to 30 mmHg. 31 This suggests that inputs from cardiac receptors either have a less potent effect on NTS neurons (below the threshold required to produce Fos expression), or else may affect a much smaller population of neurons within this nucleus. Hypervolaemia and hypovolaemia both resulted in increased Fos expression in the area postrema. With regard to hypervolaemia, our findings confirm similar findings in the rat. 54 The area postrema receives direct afferent inputs from vagal receptors 26 and also has receptors for atrial natriuretic peptide. 24 Thus, the activation of area postrema neurons in response to hypervolaemia could be a consequence of excitatory inputs from cardiac receptors or increased levels of atrial natriuretic peptide, or both. With regard to hypovolaemia, the activation of area postrema neurons is unlikely to be due to an increased level of circulating angiotensin II, because we have previously found that direct intravenous infusion of angiotensin II in the conscious barodenervated rabbit results in no significant increase in Fos expression in the area postrema, 51 consistent with the fact that the density of angiotensin II receptors in this region is very low in the rabbit. 44 However, area postrema neurons can be excited by circulating vasopressin, 20 the level of which is likely to have increased, as indicated by the activation of supraoptic vasopressin neurons in response to central hypovolaemia. In the caudal and intermediate VLM, the numbers of FosLI cells following either hypervolaemia or hypovolaemia were very similar. There were clear differences, however, in the proportion of Fos-LI cells in these regions that were double labelled for TH following hypervolaemia or hypovolaemia. Following hypervolaemia, the Fos-LI cells in these regions were predominantly non-catecholamine cells, whereas following hypovolaemia they were predominantly catecholamine cells. The non-catecholamine neurons in the caudal/intermediate VLM that were activated following hypervolaemia may be interneurons within the central reflex pathway that inhibit the sympathetic vasomotor outflow to the kidney or other vascular beds, analogous to the baroreceptor inhibitory interneurons that have been identified within this region that project to the rostral VLM. 1,17,62 Such neurons have been shown to be non-catecholaminergic. 7 The finding that following hypovolaemia the Fos-LI cells in the caudal/intermediate VLM were predominantly catecholamine cells confirms the previous study of Chan and Sawchenko 8 in the rat, who showed that hypotensive haemorrhage elicited Fos expression predominantly in catecholamine cells in that region. It was not clear in their study, however, whether this effect was dependent upon

508

P. D. Potts et al.

inputs from baroreceptor or cardiac receptors, since both groups of receptors would be unloaded by hypotensive haemorrhage. Furthermore, Badoer et al. 4 found that hypotensive but not non-hypotensive haemorrhage evoked Fos expression in the VLM, suggesting that central hypovolaemia alone is not sufficient to activate neurons in this region. Our study, however, has shown clearly that a significant degree of Fos expression is produced following central hypovolaemia in the absence of arterial baroreceptor inputs. At the same time, the degree of Fos expression in these regions resulting from central hypovolaemia of a magnitude sufficient to reduce the cardiac output by 30–35% was only about half that evoked by a moderate hypotension of 20–30 mmHg, 31 which is entirely dependent upon baroreceptor inputs. 52 Thus, unloading of baroreceptors appears to be a more potent stimulus for activation of caudal/intermediate VLM neurons than unloading of cardiac receptors. An interesting and unexpected finding of the present study was that central hypervolaemia induced a significant increase in Fos expression in the rostral VLM. Given that hypervolaemia causes a reflex decrease in renal sympathetic activity 10,14,47 and an overall increase in systemic vascular conductance, it might be expected that sympathoexcitatory neurons within the rostral VLM would be predominantly inhibited rather than excited by this stimulus. However, the heart rate increased during hypervolaemia, in confirmation of a previous study in rats 35 indicating that cardiac sympathetic activity is increased. Thus, hypervolaemia may result in differentiated changes in sympathetic output, so that certain subgroups of presympathetic neurons within the rostral VLM (such as the cardiac presympathetic neurons) may be activated rather than inhibited. In contrast, hypertension has very little effect on Fos expression in the rostral VLM, 31 consistent with the fact that this stimulus reflexly evokes inhibition of the vast majority of vasomotor neurons in this region. 32,63 Hypovolaemia also resulted in increased Fos expression in the rostral VLM, to a greater extent than that evoked by hypervolaemia. This can be explained as a consequence of disinhibition arising from unloading of vagally innervated cardiac receptors. Consistent with this interpretation, it has been shown that rostral VLM neurons are inhibited by vagal inputs. 63 However, the increase in Fos expression induced by hypovolaemia was much less than that induced by hypotension, 31 indicating that unloading of baroreceptors also appears to be a more potent stimulus for activation of rostral VLM neurons than unloading of cardiac receptors. An alternative explanation for the fact that hypervolaemia and hypovolaemia both evoked an increase in Fos expression in the VLM is that in this region the activation of neurons is dependent on the change in central blood volume from baseline levels, rather than the direction of the change (i.e. increase or decrease). If this is true, however, it implies that the processing of signals to VLM neurons activated by changes in central blood volume is quite different from the processing of signals to other brain regions (e.g., the supraoptic nucleus) where hypervolaemia and hypovolaemia resulted in markedly different degrees of Fos expression. Hypothalamus. There is much evidence that a number of hypothalamic nuclei plays key roles in the compensatory responses to changes in blood volume. In particular, it has been shown previously that hypovolaemia, even in the

absence of hypotension, results in activation of neurons in the paraventricular and supraoptic nuclei. 4,5,61,64 The present study confirms these findings, and demonstrates that the degree of activation in these nuclei is unaffected by baroreceptor denervation. A very interesting new finding, however, was that virtually none of the neurons in the paraventricular nucleus that expressed Fos in response to hypovolaemia was immunoreactive for vasopressin, although 90% of the Fos-LI neurons in the supraoptic nucleus were double labelled for vasopressin. In a previous study, hypotension in conscious rabbits also induced a high level of Fos expression in the paraventricular and supraoptic nuclei, but in contrast to the effects of hypovolaemia, approximately 40% of the paraventricular neurons activated by hypotension were immunoreactive for vasopressin. 31 Therefore, the results of the present study, taken together with this previous study, indicate that inputs to the paraventricular nucleus stimulated by hypovolaemia are directed almost exclusively to non-vasopressin neurons within the nucleus, whereas those stimulated by hypotension are directed to both vasopressin and non-vasopressin paraventricular neurons. This in turn implies that the inputs to hypothalamic nuclei activated by unloading of cardiac receptors and baroreceptors are conveyed via central pathways which are at least partly separate. These pathways are believed to include synapses within the A1 neurons in the caudal/intermediate VLM, 13 which as discussed above were also activated in response to central hypovolaemia. Hypervolaemia also resulted in increased Fos expression in the paraventricular and supraoptic nuclei, although in the case of the latter nucleus the increase was small, much less than that evoked by hypovolaemia. With regard to the paraventricular nucleus, a number of recent studies has demonstrated that this nucleus plays a major role in the compensatory autonomic response to a volume load (for review see Ref. 11). For example, electrophysiological studies have identified spinally projecting neurons in the paraventricular nucleus that are activated by a volume load 33 or by systemic infusion of atrial natriuretic peptide, 34 which is known to be released as a consequence of volume loading. 27 Both of these effects are mediated by vagal afferent inputs. 33,34 In addition, stimulation of neurons in the paraventricular nucleus can produce inhibition of renal sympathetic activity which is similar to that produced by a volume load. 16 Finally, lesions of the parvocellular portion of the paraventricular nucleus have been shown to attenuate the renal sympathoinhibition and vasodilation reflexly evoked by a volume load. 19,35 Thus, it has been suggested that a volume load results in stimulation of vagally innervated cardiac receptors, leading to activation of paraventricular neurons which in turn causes inhibition of renal sympathetic activity. 11 Our results demonstrating activation of paraventricular neurons in response to hypervolaemia are therefore entirely consistent with this hypothesis. It is also interesting to note that very few (approximately 1%) of the paraventricular neurons activated by hypervolaemia were immunoreactive for vasopressin, in view of recent evidence that spinally projecting paraventricular neurons that release vasopressin as a neurotransmitter have an excitatory action on the renal sympathetic outflow. 39 Thus, our results suggest that the paraventricular neurons activated by hypervolaemia may be predominantly sympathoinhibitory. Recent anatomical and electrophysiological studies indicate that this could be

509

Effects of blood volume changes on brain Fos expression

mediated either by a direct paraventriculospinal connection or via inhibitory inputs to rostral VLM sympathoexcitatory neurons. 53,60,66 A recent study in the barointact conscious rabbit also reported that a volume load produced by Haemaccel infusion resulted in a high level of Fos expression in the paraventricular nucleus, but attributed this to activation by inputs arising from neurons in the OVLT that are excited by increased plasma osmolality associated with the Haemaccel infusion. 2 This is unlikely to explain the activation of paraventricular neurons in response to Haemaccel infusion in our study, however, because there was no significant increase in Fos expression in the OVLT in response to this stimulus. Moreover, osmotic stimuli are known to elicit activation of vasopressin neurons within the paraventricular nucleus, 40,41 whereas in our study very few vasopressin neurons in this nucleus were activated by the hypervolaemic stimulus. Our results therefore support the finding of Randolph et al. 54 in barointact rats that an isotonic volume load activates neurons in the paraventricular nucleus, and extend it by demonstrating that these effects are independent of arterial baroreceptor inputs and are not mediated by vasopressin neurons. There was also a moderate but statistically significant increase, compared with the control group, in the number of Fos-LI cells in the supraoptic nucleus following hypervolaemia. In contrast to hypovolaemia, only a minority (approximately 40%) of the supraoptic neurons that expressed Fos in response to hypervolaemia was immunoreactive for vasopressin. It is thus possible that the non-vasopressin neurons in the supraoptic nucleus that were activated by hypervolaemia were oxytocin-secreting neurons, which have been shown to be activated by volume expansion in the rat. 54 Oxytocin is believed to play a role in fluid regulation, because its plasma level is increased by volume expansion 18 and it acts on the kidney to produce natriuresis. 65 Increased Fos expression also occurred following either hypervolaemia or hypovolaemia in the hypothalamic arcuate and suprachiasmatic nuclei. Although the arcuate nucleus has long been known to be involved in neuroendocrine regulation, 9 recent studies have shown that it also plays a role in body fluid homeostasis. 30,57 Anatomical studies have also demonstrated that the arcuate nucleus receives afferent inputs from both the NTS 25 and the paraventricular nucleus, 55 which are therefore possible routes by which arcuate neurons are activated in response to hypervolaemia or hypovolaemia. The suprachiasmatic nucleus plays a key role in the diurnal regulation of various physiological processes, 46 and there is now evidence that this includes the regulation of vasopressin and oxytocin secretion. 12 The afferent pathways that mediate activation of suprachiasmatic neurons in response to hypervolaemia or hypovolaemia are not clear, but may include an input from the SFO. 45 The precise role of the suprachiasmatic

nucleus, like that of the arcuate nucleus, in the regulation of body fluid balance remains to be determined. Lamina terminalis. There was a large increase in Fos expression in the OVLT and SFO of the lamina terminalis in response to hypovolaemia, but not hypervolaemia. This is presumably due to a direct action of circulating angiotensin II on circumventricular neurons in these regions, 48 which are known to contain a high density of angiotensin II receptors. 44 An increase in circulating angiotensin II could result from an increase in renin release either reflexly evoked by unloading of arterial baroreceptors and/or cardiac receptors, or in the case of the barodenervated rabbits, as a consequence of the marked hypotension associated with the central hypovolaemia. 56 The activation of neurons in the OVLT and SFO in turn would explain the activation of neurons in the median preoptic nucleus in response to hypovolaemia, since this nucleus receives excitatory inputs from the OVLT and SFO. 42 In addition, the median preoptic nucleus also receives inputs from the NTS, 42 which may also have contributed to the activation of this nucleus in response to hypovolaemia. Furthermore, the activation of neurons within the OVLT and SFO may also contribute to activation of the supraoptic and paraventricular nuclei in response to central hypovolaemia, because the latter nuclei receive inputs from angiotensin II-sensitive neurons in the OVLT and SFO, both directly and indirectly via the median preoptic nucleus. 42,49,58 CONCLUSIONS

The results of this study show that both central hypervolaemia and hypovolaemia evoke Fos expression in distinct populations of neurons in the medulla and hypothalamus in the unanaesthetized rabbit. The overall patterns of Fos expression in response to the two stimuli were different, although in several nuclei Fos expression occurred in response to either stimulus. In contrast to the effects of induced changes in arterial pressure, 31,52 medullary and hypothalamic Fos expression evoked by changes in central blood volume were not dependent upon inputs from arterial baroreceptors. Taken together with the results of previous studies, the results also indicate that the activation of hypothalamic neurons in response to signals from cardiac receptors and arterial baroreceptors is conveyed, at least in part, by separate central pathways.

Acknowledgements—The work was supported by the National Health and Medical Research Council of Australia, the National Heart Foundation of Australia and the Ramaciotti Foundations.

REFERENCES

1. 2. 3. 4. 5.

Agarwal S. K. and Calaresu F. R. (1991) Monosynaptic connection from caudal to rostral ventrolateral medulla in the baroreceptor reflex pathway. Brain Res. 555, 70–74. Badoer E., McKinlay D., Trigg L. and McGrath B. P. (1997) Distribution of activated neurons in the rabbit brain following a volume load. Neuroscience 81, 1065–1077. Badoer E., McKinley M. J., Oldfield B. J. and McAllen R. M. (1992) Distribution of hypothalamic, medullary and lamina terminalis neurons expressing Fos after hemorrhage in conscious rats. Brain Res. 582, 323–328. Badoer E., McKinley M. J., Oldfield B. J. and McAllen R. M. (1993) A comparison of hypotensive and non-hypotensive hemorrhage on fos expression in spinally projecting neurons of the paraventricular nucleus and rostral ventrolateral medulla. Brain Res. 610, 216–223. Badoer E. and Merolli J. (1998) Neurons in the hypothalamic paraventricular nucleus that project to the rostral ventrolateral medulla are activated by haemorrhage. Brain Res. 791, 317–320.

510 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47.

P. D. Potts et al. Badoer E., Moguilevski V. and McGrath B. P. (1998) Cardiac afferents play the dominant role in renal nerve inhibition elicited by volume expansion in the rabbit. Am. J. Physiol. 274, R383–R388. Blessing W. W., Hedger S. C., Joh T. H. and Willoughby J. O. (1987) Neurons in the area postrema are the only catecholamine-synthesizing cells in the medulla or pons with projections to the rostral ventrolateral medulla (C1 area) in the rabbit. Brain Res. 419, 336–340. Chan R. K. W. and Sawchenko P. E. (1994) Spatially and temporally differentiated patterns of c-fos expression in brainstem catecholaminergic cell groups induced by cardiovascular challenges in the rat. J. comp. Neurol. 348, 433–460. Chronwall B. M. (1985) Anatomy and physiology of the neuroendocrine arcuate nucleus. Peptides 6, Suppl. 2, 1–11. Clement D. L., Pelletier C. L. and Shepherd J. T. (1972) Role of vagal afferents in the control of renal sympathetic nerve activity in the rabbit. Circulation Res. 31, 824–830. Coote J. H. (1995) Cardiovascular function of the paraventricular nucleus of the hypothalamus. Biol. Signals 4, 142–149. Cui L. N., Saeb-Parsy K. and Dyball R. E. J. (1997) Neurones in the supraoptic nucleus of the rat are regulated by a projection from the suprachiasmatic nucleus. J. Physiol. 502, 149–159. Day T. A. (1989) Control of neurosecretory vasopressin cells by noradrenergic projections of the caudal ventrolateral medulla. In Central Neural Organization of Cardiovascular Control. Progress in Brain Research (eds Ciriello J., Caverson M. M. and Polosa C.), Vol. 81, pp. 301–317. Elsevier, Amsterdam. DiBona G. F. and Sawin L. L. (1985) Renal nerve activity in conscious rats during volume expansion and depletion. Am. J. Physiol. 248, F15–F23. Faris I. B., Iannos J., Jamieson G. G. and Ludbrook J. (1981) The circulatory effects of acute hypervolemia and hemodilution in conscious rabbits. Circulation Res. 48, 825–834. Gardner J., Al-Ani M., Lovick T. A. and Coote J. H. (1995) Differential pattern of sympathetic nerve activity elicited by chemical stimulation in the paraventricular nucleus of the rabbit. J. Physiol. 483, 100P. Gieroba Z. J., Li Y.-W. and Blessing W. W. (1992) Characteristics of caudal ventrolateral medullary neurons antidromically activated from rostral ventrolateral medulla in the rabbit. Brain Res. 582, 196–207. Haanwinckel M. A., Elias L. K., Favaretto A. L. V., Gutkowska J., McCann S. M. and Antunes-Rodrigues J. (1995) Oxytocin mediates atrial natriuretic peptide release and natriuresis after volume expansion in the rat. Proc. natn. Acad. Sci. U.S.A. 92, 7902–7906. Haselton J. R., Goering J. and Patel K. P. (1994) Parvocellular neurons of the paraventricular nucleus are involved in the reduction in renal nerve discharge during isotonic volume expansion. J. auton. nerv. Syst. 50, 1–11. Hasser E. M. and Bishop V. S. (1981) Reflex effects of vasopressin after blockade of VI receptors in the area postrema. Circulation Res. 67, 265–271. Herdegen T., Kovary K., Leah J. and Bravo R. (1991) Specific temporal and spatial distribution of Jun, Fos and KROX-24 proteins in spinal neurons following noxious transynaptic stimulation. J. comp. Neurol. 313, 178–191. Hines T., Toney G. M. and Mifflin S. W. (1994) Responses of neurons in the nucleus tractus solitarius to stimulation of heart and lung receptors in the rat. Circulation Res. 74, 1188–1196. Hirooka Y., Polson J. W., Potts P. D. and Dampney R. A. L. (1997) Hypoxia-induced Fos expression in neurons projecting to the pressor region in the rostral ventrolateral medulla. Neuroscience 80, 1209–1224. Imura H., Nakao K. and Itoh H. (1992) The natriuretic peptide system in the brain: implications in the central control of cardiovascular and neuroendocrine functions. Front. Neuroendocr. 13, 217–249. Ito H. and Seki M. (1998) Ascending projections from the area postrema and the nucleus of the solitary tract of Suncus murinus: anterograde tracing study using Phaseolus vulgaris leucoagglutinin. Okaj. Folia anat. Jap. 75, 9–31. Kalia M. and Sullivan J. M. (1982) Brainstem projections of sensory and motor components of the vagus nerve in the rat. J. comp. Neurol. 211, 248–265. Kohno M., Clegg K. B. and Sambhi M. P. (1987) Effects of volume change in circulating immunoreactive atrial natriuretic factor in rats. Hypertension 10, 171–175. Korner P. I. (1980) Central nervous control of autonomic cardiovascular function. In Handbook of Physiology—Cardiovascular System 1 (ed. Field J. W.), Chap. 20, pp. 691–739. American Physiological Society, Washington. Lante´ri-Minet M., Weil-Fugazza J., de Pommery J. and Mene´trey D. (1994) Hindbrain structures involved in pain processing as revealed by the expression of c-Fos and other immediate early gene proteins. Neuroscience 58, 287–298. Lepetit P., Grange E., Gay N. and Bobillier P. (1992) Comparison of the effects of chronic water deprivation and hypertonic saline ingestion on cerebral protein synthesis in rats. Brain Res. 586, 181–187. Li Y.-W. and Dampney R. A. L. (1994) Expression of Fos-like protein in brain following sustained hypertension and hypotension in conscious rabbits. Neuroscience 61, 613–634. Li Y.-W., Gieroba Z. J., McAllen R. M. and Blessing W. W. (1991) Neurons in rabbit caudal ventrolateral medulla inhibit bulbospinal barosensitive neurons in rostral medulla. Am. J. Physiol. 261, R44–R51. Lovick T. A. and Coote J. H. (1988) Effects of volume loading on paraventriculo-spinal neurones in the rat. J. auton. nerv. Syst. 25, 135–140. Lovick T. A. and Coote J. H. (1989) Circulating natriuretic factor activates vagal afferent inputs to paraventriculo-spinal neurones in the rat. J. auton. nerv. Syst. 26, 129–134. Lovick T. A., Malpas S. and Mahony M. T. (1993) Renal vasodilatation in response to acute volume load is attenuated following lesions of parvocellular neurons in the paraventricular nucleus in rats. J. auton. nerv. Syst. 43, 247–256. Ludbrook J. and Graham W. F. (1984) The role of cardiac receptor and arterial baroreceptor reflexes in control of the circulation during acute change of blood volume in the conscious rabbit. Circulation Res. 54, 424–435. Ludbrook J., Potocnik S. J. and Woods R. L. (1988) Simulation of acute haemorrhage in unanaesthetized rabbits. Clin. exp. Pharmac. Physiol. 15, 575–584. Ludbrook J. and Ventura S. (1996) Roles of carotid baroreceptor and cardiac afferents in hemodynamic responses to acute central hypovolemia. Am. J. Physiol. 270, F680–F687. Malpas S. C. and Coote J. H. (1994) The role of vasopressin in the sympathetic response to paraventricular nucleus stimulation in the anaesthetised rat. Am. J. Physiol. 266, R228–R236. McDonald T. J., Li C., Nijland M. J., Caston-Balderrama A. and Ross M. G. (1998) Fos response of fetal sheep anterior circumventricular organs to osmotic challenge in late gestation. Am. J. Physiol. 275, H609–H614. McKinley M. J., Bicknell R. J., Hards D., McAllen R. M., Vivas L., Weisinger R. S. and Oldfield B. J. (1992) Efferent neural pathways of the lamina terminalis subserving osmoregulation. Prog. Brain Res. 91, 395–402. McKinley M. J., Pennington G. L. and Oldfield B. J. (1996) Anteroventral wall of the third ventricle and dorsal lamina terminalis—headquarters for control of body fluid homeostasis. Clin. exp. Pharmac. Physiol. 23, 271–281. Meessen H. and Olszewski J. (1949) A Cytoarchitectonic Atlas of the Rhombencephalon of the Rabbit. S. Karger, Basel. Mendelsohn F. A. O., Allen A. M., Clevers A. J., Denton D. A., Tarjan E. and McKinley M. J. (1998) Localization of angiotensin II receptor binding in rabbit brain by in vitro autoradiography. J. comp. Neurol. 270, 372–384. Moga M. M. and Moore R. Y. (1997) Organization of neural inputs to the suprachiasmatic nucleus in the rat. J. comp. Neurol. 389, 508–534. Moore R. Y. (1995) Organization of the mammalian circadian system. In Circadian Clocks and their Adjustment (ed. Waterhouse J. M.), pp. 88–99. John Wiley, Chichester. Morita H. and Vatner S. F. (1985) Effects of volume expansion on renal nerve activity, renal blood flow, and sodium and water excretion in conscious dogs. Am. J. Physiol. 249, H1538–H1548.

Effects of blood volume changes on brain Fos expression

511

48. Oldfield B. J., Badoer E., Hards D. K. and McKinley M. J. (1994) Fos production in retrogradely labelled neurons of the lamina terminalis following intravenous infusion of either hypertonic saline or angiotensin II. Neuroscience 60, 255–262. 49. Petrov T., Harris K. H., MacTavish D., Krukoff T. L. and Jhamandas J. H. (1995) Hypotension induces Fos immunoreactivity in NADPH-diaphorase positive neurons in the paraventricular and supraoptic hypothalamic nuclei of the rat. Neuropharmacology 34, 509–514. 50. Polson J. W., Potts P. D., Li Y.-W. and Dampney R. A. L. (1995) Fos expression in neurons projecting to the pressor region in the rostral ventrolateral medulla after sustained hypertension in conscious rabbits. Neuroscience 67, 107–123. 51. Potts P. D., Hirooka Y. and Dampney R. A. L. (1999) Activation of brain neurons by circulating angiotensin II: direct effects and baroreceptor-mediated secondary effects. Neuroscience 90, 581–594. 52. Potts P. D., Polson J. W., Hirooka Y. and Dampney R. A. L. (1997) Effects of sinoaortic denervation on Fos expression evoked by hypertension and hypotension in conscious rabbits. Neuroscience 77, 503–520. 53. Pyner S. and Coote J. H. (1999) Identification of an efferent projection from the paraventricular nucleus of the hypothalamus terminating close to spinally projecting rostral ventrolateral medullary neurons. Neuroscience 88, 949–957. 54. Randolph R. R., Li Q., Curtis K. S., Sullivan M. J. and Cunningham J. T. (1998) Fos expression following isotonic volume expansion of the unanesthetized male rat. Am. J. Physiol. 274, R1345–R1352. 55. Ranson R. N., Motawei K., Pyner S. and Coote J. H. (1998) The paraventricular nucleus of the hypothalamus sends efferents to the spinal cord of the rat that closely appose sympathetic preganglionic neurones projecting to the stellate ganglion. Expl Brain Res. 120, 164–172. 56. Reid I. A., Morris B. J. and Ganong W. F. (1978) The renin–angiotensin system. A. Rev. Physiol. 40, 377–410. 57. Rosas-Arellano M. P., Solano-Flores L. P. and Ciriello J. (1996) Arcuate nucleus inputs onto subfornical organ neurons that respond to plasma hypernatremia and angiotensin II. Brain Res. 707, 308–313. 58. Saper C. B. and Levisohn D. (1983) Afferent connections of the median preoptic nucleus in the rat: anatomical evidence for a cardiovascular integrative mechanism in the anteroventral third ventricle (AV3V) region. Brain Res. 288, 21–31. 59. Schadt J. C. and Ludbrook J. (1991) Hemodynamic and neurohumoral responses to acute hypovolemia in conscious mammals. Am. J. Physiol. 260, H305–H318. 60. Shafton A. D., Ryan A. and Badoer E. (1998) Neurons in the hypothalamic paraventricular nucleus send collaterals to the spinal cord and to the rostral ventrolateral medulla in the rat. Brain Res. 801, 239–243. 61. Shen E., Dun S. L., Ren C. and Dun N. J. (1992) Hypovolemia induces Fos-like immunoreactivity in neurons of the rat supraoptic and paraventricular nuclei. J. auton. nerv. Syst. 37, 227–230. 62. Terui N., Masuda N., Saeki Y. and Kumada M. (1990) Activity of barosensitive neurons in the caudal ventrolateral medulla that send axonal projections to the rostral ventrolateral medulla in rabbits. Neurosci. Lett. 118, 211–214. 63. Terui N., Saeki Y. and Kumada M. (1986) Barosensory neurons in the ventrolateral medulla in rabbits and their responses to various afferent inputs from peripheral and central sources. Jpn. J. Physiol. 36, 1141–1164. 64. Ueta Y., Levy A., Lightman S. L., Hara Y., Serino R., Nomura M., Shibuya I., Hattori Y. and Yamashita H. (1998) Hypovolemia upregulates the expression of neuronal nitric oxide synthase gene in the paraventricular and supraoptic nuclei of rats. Brain Res. 790, 25–32. 65. Verbalis J. G., Mangione M. P. and Stricker E. M. (1991) Oxytocin produces natriuresis at physiological plasma concentrations. Endocrinology 128, 1317–1322. 66. Yang Z. and Coote J. H. (1998) Influence of the hypothalamic paraventricular nucleus on cardiovascular neurones in the rostral ventrolateral medulla of the rat. J. Physiol. 513, 521–530. (Accepted 23 August 1999)