- Email: [email protected]
Chapter 10 Electrophysiological analysis of chemosensitive neurons within the area postrema of the rat
W . Hamann and A. lggo (Eds.) Progress in Brain Research, Vol. 14 0 1988 Elsevier Science Publishers B.V. (Biomedical Division)
71
CHAPTER 10
Electrophysiological analysis of chemosensitive neurons within the area postrema of the rat Akira Adachi and Motoi Kobashi Department of Physiology, Okayama University Dental School, Shikata-cho, Okayama 700, Japan
Summary
Introduction
The area postrema (AP) has proved to be a chemoreceptive trigger zone for the emetic response. Evidence indicating other functions of the A P especially in the rat has grown in abundance recently. The present study is an attempt to demonstrate chemosensitive neurons within the A P electrophysiologically. Two different techniques were employed for chemical stimulation of the AP neurons: (1) the floor of the exposed fourth ventricle was superfused with isotonic glucose solution, hypertonic or hypotonic Ringer solution, or LiCl Ringer solution; (2) these solutions were injected into the left carotid artery. To confirm vagal afferent influences on given A P neurons, single shock electrical stimulation was delivered to the cervical vagal trunks of both sides. Three types of chemosensitive neurons were identified: (1) glucose responsive neurons that may participate in the control of satiety, (2) sodium (osmotic pressure) responsive neurons that may contribute to salt preference or control of water balance, and (3) LiCl responsive neurons that may play a role in conditioned taste aversion. Some of these chemosensitive neurons were influenced by the vagal afferent inputs, suggesting a close relationship between chemoreceptor mechanisms of the AP and visceral afferents.
In general the function of the area postrema (AP) has been elucidated as being a chemoreceptive trigger zone for the emetic response (Borison and Wang, 1953). Evidence that indicates other functions of this structure is gradually accumulating (Borison, 1974). Lesions destroying the A P cause changes in salt intake without deficits in gustatory function in the rat (Contreras and Stetson, 1981). Ablation of the A P causes exaggerated consumption of preferred foods in the rat (Edwards and Ritter, 1981). Based on these facts, the authors postulated chemoreceptor functions of the A P not only for the emetic response, but for ingestive behavior that could be important for homeostatic control. Our preliminary study has revealed the existence of glucose responsive neurons as well as sodium responsive ones within the A P in the rat (Adachi and Kobashi, 1985). Glucose responsive neurons induced a marked decrease in the discharge rate in response to topical application of glucose by means of microelectro-osmotic techniques. Two different types of sodium responsive neurons were also identified by means of topical iontophoretic application of sodium ions. The purpose of the present study is to extend our previous observation in order to confirm further the glucose as well as sodium responsive neurons within the
78
AP by use of different techniques for chemical stimulation of the AP. In addition, the AP plays a critical role in the formation of conditioned taste aversions by lithium chloride (Ritter et al., 1980). In this study, effects of chemical stimulation by LiCl on AP neurons were also examined to elucidate LiCl responsive neurons besides the glucose or sodium responsive ones.
A
SUPERFUSION
B
INTRACAROTID INJECTION
Methods A total of 49 male Sprague - Dawley rats (Charles River of Japan), weighing 300 - 400 g were used as subjects. Under urethane - chloralose anesthesia (urethane 0.8 g/kg, chloralose 65 mg/kg, i.p.), a catheter was inserted into the left carotid artery toward the heart as shown in Fig. lB, in order to apply chemical stimuli to the AP through the blood vessel. The right carotid artery was ligated. To examine vagal afferent invasion in the AP, implanted stimulating electrodes were fixed on the cervical vagal trunks of both sides. In one case the celiac branch of the vagus was stimulated. Then the animals were mounted on a stereotaxic apparatus. To place the medulla oblongata in a horizontal position, the incisors were fixed 25 mm below the level of the auditory meatus. The occipital bone was removed with a dental drill, and the caudal part of the medulla was exposed. In this way, the surface of the AP was completely exposed without removing the cerebellum by suction. As illustrated in Fig. lA, the surface of the AP was superfused with standard Ringer solution (Na 147 mEq, K 4 mEq, Ca 4.5 mEq, C1 155.5 mEq). A glass pipette electrode filled with 0.5 M sodium acetate and 2% pontamine sky blue was used for recording unitary discharges. Under inspection through a dissecting microscope, the recording electrode was introduced into the AP by means of a micromanipulator until unit discharges were observed. Then single shock electrical stimulation was delivered to the cervical vagal trunks in order to confirm a vagal afferent supply to the AP. After this examination, superfusion with Ringer solution was switched to isotonic glucose solution (300
Fig. I . Schematic representation of the experimental procedure. Two different techniques employed for chemical stimulation of the AP neurons are illustrated.
mM), hypertonic, hypotonic Ringer solution or LiCl (147 mM) containing Ringer solution respectively. Hypertonicity was achieved by the addition of 147 mM NaCl to the standard Ringer solution. Hypotonicity was achieved by decreasing the NaCl concentration by 74 mM in the standard Ringer solution. These test solutions were also applied through the carotid catheter to elucidate responses of the AP neurons to the chemical stimuli supplied through the fenestrated blood vessel. All records were stored on magnetic tape for later analysis. The number of impulses in each 5 s was plotted by means of a pulse-rate meter. Recording sites were marked by iontophoretic dye injection and examined afterward histologically.
Results It is noticed that spontaneous discharges have been rarely observed within the AP in previous studies, due presumably to a sparse neuronal distribution in this structure. Therefore, responses have been infrequently obtained even when exploring within the AP by means of the microelectrode technique. Figs. 2 and 3 show two different types of AP neurons in response to glucose. When superfusion of the exposed surface of the fourth ventricle with standard Ringer solution was switched to isotonic glucose solution (indicated by G ) , a marked
19
decrease in discharge rate is seen (Fig. 2Aa). The decreased neuronal activity gradually returned to the control discharge rate when glucose solution was changed back to the Ringer solution. This response is characterized by a negative correlation between the discharge rate and glucose concentration. The superfusion of hypertonic Ringer solution containing 294 mM NaCl (indicated by S ) induced no response in this neuron. Fig. 2Ab shows responses of this glucose responsive neuron to single shock electrical stimulation (intensity 3 V , duration 3 ms) of the celiac branch of the vagus. The latency from the onset of single shock electrical stirnulation until the unit discharge was evoked was approximately 160 ms. Because preceding spontaneous discharges did not cancel the evoked discharges by collision (see the second and third traces from the top), this neuron was activated by afferent invasion via the celiac branch of the vagus. It implies that some neurons within the AP are responsive to changes in glucose concentration of the cerebrospinal fluid (CSF) and also receive afferent signals from the abdominal organs. Fig. 2Ba presents a response of this type of
Fig. 2. Responses of two different glucose responsive AP neurons. Superfusion of the fourth ventricle with isotonic glucose solution (G) induced a marked decrease in discharge rate of an AP neuron (Aa). Note that hypertonic Ringer solution (S) is ineffective. (Ab) Responses to single shock electrical stimulation of the celiac branch of the vagus. Each oscillograph sweep is triggered by a stimulus pulse. Evoked unit discharge is seen in each trace. (B,) Effect of intracarotid infusion of glucose on the other AP neuron. Note clear suppression by glucose (G) and no response to hypertonic Ringer solution (S). (Bb) Oscillograph record of discharges. Ten sweeps were superimposed. This neuron elicits no response t o electrical stimulation of the celiac branch of the vagus.
Fig. 3. (A) Effect of the superfusion with isotonic glucose solution ( G ) ,hypertonic (S) or hypotonic (S') Ringer solution. Only glucose elicits marked increases in discharge rate. The intracarotid infusion of glucose indicated by G(ca) induces the same response. (B) Evoked response to single shock electrical stimulation of the right cervical vagal trunk. (C) Oscillograph records of unit discharges: (1) spontaneous discharges recorded from the time indicated by arrow 1 in A; (2) discharges responding to glucose recorded from the time indicated by arrow 2 in A. Note a marked difference in the discharge rate between CI and C2.
glucose responsive neuron to the intracarotid infusion of isotonic glucose solution. This neuron rapidly responded to the glucose infusion (indicated by G ) , displaying a marked decrease in discharge rate followed by a return to the control discharge rate after cessation of the infusion. The same infusion with hypertonic Ringer solution (indicated by s) did not elicit any change in discharge rate. The response to the glucose infusion is similar to that of the superfusion with glucose solution. This unit did not respond to single shock electrical stimulation delivered to the celiac branch of the vagus as shown in Fig. 2Bb. Opposite responses to glucose were observed in the other AP neuron. As presented in Fig. 3A, superfusion with glucose (indicated by G ) markedly increased the discharge rate. Intracarotid infusion of the same solution also produced an increase in discharge rate as indicated by G(ca). However, no responses were observed when the superfusion with Ringer solution was switched to that with either hyper- or hypotonic Ringer solution (indicated by S and S ' respectively). This neuron responded to single shock electrical stimulation of the right cervical vagal trunk (same intensity and duration as in Fig. 2) but not to that of the left one
80
as shown in Fig. 3B. Oscillographic records of the discharges are illustrated in Fig. 3C, and C,. Comparison of C, with C, also clearly. indicates the increase in discharge rate in response to the superfusion with glucose. Typical responses to hyper- and hypotonicity of
Fig. 4. (A,) Response to the superfusion with hypertonic Ringer solution (S). Note a marked increase in discharge rate during a horizontal bar indicated by S followed by a decrease in discharge rate (after-effect). Hypotonic Ringer solution (S ’ ) causes a decrease in discharge rate but glucose ( G ) does not. (A,) Responses to electrical stimulation of the left and right vagus. Ten sweeps were superimposed. No response is elicited. (B,) Response to the superfusion with hypotonic Ringer solution (S‘). No response to superfusion with isotonic glucose ( G ) or hypertonic Ringer solution (S) is elicited. (Bb)The same oscillograph records as in A,. No response is elicited.
Fig. 5 . (A) Response of an AP neuron to the superfusion with LiCl containing Ringer solution (Li). Note the lack of response to equiosmotic hypertonic Ringer solution (S).Glucose is also ineffective. (B) Oscillograph records of discharges elicited by electrical stimulation of the vagus of both sides.
some A P neurons are presented in Fig. 4A and B. When the superfusion with standard Ringer solution was switched to that with hypertonic Ringer solution (indicated by S), a marked increase in discharge rate was observed (Fig. 4Aa). Immediately after switching back to standard Ringer solution, a suppression of neural response was recognized. This neuron shows an opposite response to hypotonic Ringer solution (indicated by S ’ ) , so a marked decrease in discharge rate is seen during a horizontal bar indicated by S ’ . No response is elicited by switching the superfusion to isotonic glucose solution (indicated by G). Single shock electrical stimulation of the cervical vagal trunk of both sides elicited no response in this neuron as shown in Fig. 4.4,. Fig. 4Ba presents a response of the other A P neuron to hypotonicity. This neuron responded only to the hypotonic superfusion (indicated by S ‘ ) with a marked increase in discharge rate. Both isotonic glucose solution (G) and hypertonic Ringer solution (S) were ineffective. The same electrical stimulation of the vagus elicited no response as shown in Fig. 4Bb. These neurons were not supplied by the vagal afferent inputs. A specimen record from a lithium chloride responsive neuron within the A P is shown in Fig. 5 . When the superfusion with standard Ringer solution was switched to that with LiCl containing Ringer solution (indicated by Li), an increase in discharge rate was observed. However, hypertonic Ringer solution that was equiosmotic to LiCl Ringer solution elicited no response (Fig. 5A). It indicates that the LiCl responsive neuron is insensitive to osmotic pressure as well as glucose. This neuron responded to electrical stimulation of the cervical vagal trunks of both sides. Fig. 5B presents evoked discharges by electrical stimulation of the left and right vagus (same intensity and duration as in Fig. 2). Latencies are approximately 46 ms. A total of 46 neurons within the area postrema were examined. Among them, 25 neurons responded to electrical stimulation of the cervical vagal trunk. Four neurons decreased and one neuron increased the discharge rates in response to glucose.
81
Two neurons increased and two other neurons decreased the discharge rates in response t o hypertonic Ringer solution. One neuron increased and one neuron decreased in response to hypotonic Ringer solution. Two LiCl responsive neurons were also observed. Some of these neurons responded only to chemicd stimuli, others responded to both chemical and single shock electrical stimulation of the vagus. Thus, no correlation between the vagal afferent supplies and chemosensitivities of these AP neurons was recognized.
Discussion The AP is a circumventricular organ of the fourth ventricle and lacks a blood - brain barrier. The existence of neural elements within this organ has been confirmed histologically (Leslie, 1986). These facts suggest various chemoreceptor mechanisms to detect blood- and CSF-borne chemical information not only for triggering the emetic response but also for homeostatic control. Many ablation experiments of the A P indicated other functions than the emetic response, for instance, control of satiety (Edwards and Ritter, 1981, 1986; Ritter and Edwards, 1984), control of sodium and water balance (Wise and Ganong, 1960; Contreras and Stetson, 1981; Edwards and Ritter, 1982), formation of conditioned taste aversion (Berger et al., 1973; Ritter et al., 1980) and cardiovascular control (Barnes et al., 1984). A postulate of AP chemoreceptors is common to these authors. The present study provides direct evidence for the existence of such chemoreceptor mechanisms which participate in these controls. The glucose responsive neurons within the A P may play a role in the control of food intake. Oomura et al. (1969) were the first to reveal glucose responsive neurons in the hypothalamus by electrophysiological methods. Glucose responsive neurons were also found in the nucleus tractus solitarii (NTS) (Adachi et al., 1984; Mizuno and Oomura, 1984). Neural connections between the glucose responsive neurons in the central nervous system and a peripheral hepatic glucoreceptor (Nii-
jima, 1969, 1982) was elucidated (Shimizu et al., 1983; Adachi et al., 1984). These neural networks suggest a hierarchical organization of glucose information in the body fluid processing by central and peripheral nervous systems (Oomura, 1983). The AP responsive neurons could be involved in this organization. The possible function of the A P sodium (or osmo-) responsive neurons favors a similar conception to that of the glucose responsive ones in association with the control of sodium and water balance of the body fluid. Sodium (or osmo-) responsive neurons are located in the hypothalamus (Oomura et al., 1969) and the NTS as well (Kobashi and Adachi, 1985). Peripheral osmoreceptors exist in the liver (Adachi et al., 1976) and their afferent signals project to the caudal portion of the NTS (Adachi, 1981, 1984) and in turn to the hypothalamus via the parabrachial nucleus (PB) (Kobashi and Adachi, 1986). Convergence of the hepatic osmoreceptive signals onto the NTS sodium responsive neurons was also confirmed (Kobashi and Adachi, 1985). A survey of these previous reports strongly suggests involvement of the AP sodium responsive neurons in the neural control system for isosmosis and isovolemia of the body fluid. Carpenter and Briggs (1986) recorded responses of neurons of the canine A P to iontophoretic application of insulin. They concluded that these responses provide further support for a critical role of the AP in triggering the emetic reflex. The existence of LiCl responsive neurons within the rat AP indicates that this brain region is involved in neural mechanisms inducing nausea, even though the emetic response is not easily elicited in the rat. In turn, such chemosensitive neurons may play a part in the formation of conditioned taste aversion. As many other chemicals such as scopolamine, amphetamine, hydroxytryptophan and so on induce conditioned taste aversion, it may be possible that the LiCl responsive neuron also responds to these chemicals. A postulated scheme is presented in Fig. 6. Because some of the chemosensitive AP neurons
82
References
PB NTS
X
Fig. 6. Schematic diagram illustrating a possible neural connection of AP neurons. CSF, cerebrospinal fluid; R, chemoreceptor neuron; C, capillary; FE, fenestration; AP, area postrema; PB, parabrachial nucleus; NTS, nucleus tractus solitarii; X, vagus.
are excited by the vagal afferent but others are not, there can be a possibility that receptor-like neurons (R) are located in the AP. The former neurons themselves can also be chemosensitive, because convergence of the vagal afferent signals to chemosensitive neurons in the NTS has been ascertained (Adachi et al., 1984; Kobashi and Adachi, 1985). These neurons are responsive to either blood- or CSF-borne substances, because application of the chemical stimuli to both the fourth ventricle and the carotid artery elicited the neuronal responses. Close connections between the A P and the PB as well as the NTS have been clarified (Shapiro and Miselis, 1982). It is easily delineated that the internal chemical information detected in the AP may be transmitted to the other nuclei in the central nervous system, so that various reflexes including emesis can be induced. Certainly further studies should be necessary to elucidate such neural mechanisms. Acknowledgements
This work was supported by Grant-in-Aid for General Scientific Research 62480378 from the Ministry of Education, Science and Culture in Japan and a Grant from the Society for Research on Umami Taste.
Adachi, A. (1981) Electrophysiological study of hepatic vagal projection to the medulla. Neurosci. Lerr., 24: 19-23. Adachi, A. (1984) Projection of the hepatic vagal nerve in the medulla oblongata. J. Auton. Nerv. Syst., 10: 287-293. Adachi, A. and Kobashi, M. (1985) Chemosensitive neurons within the area postrema of the rat. Neurosci. Lerr., 5 5 : 137- 140. Adachi, A., Niijima, A. and Jacobs, H.L. (1976) An hepatic osmoreceptor mechanism in the rat: electrophysiological and behavioral studies. Am. J. Physiol., 231: 1043 - 1049. Adachi, A,, Shimizu, N., Oomura, Y. and Kobashi, M. (1984) Convergence of hepatoportal glucose-sensitive afferent signals to glucose-sensitive units within the nucleus of the solitary tract. Neurosci. Lett., 46: 215 -218. Barnes, K.L., Ferrario, C.M., Chernicky, C.L. and Borosnihan, K.B. (1984) Participation of the area postrema in cardiovascular control in the dog. Fed. Proc., 43: 2959 - 2962. Berger, B.D., Wise, C.D. and Stein, L. (1973) Area postrema damage and bait shyness. J. Comp. Physiol. Psycho/., 83: 475 - 479. Borison, H.L. (1974) Area postrema: chemoreceptor trigger zone for vomiting - is that all? Life Sci., 14: 1807- 1817. Borison, H.L. and Wang, S.C. (1953) Physiology and pharmacology of vomiting. Pharmacol. Rev., 5 : 193 - 230. Carpenter, D.O. and Briggs, D.B. (1986) Insulin excites neurons of the area postrema and causes emesis. Neurosci. Lerr., 68: 85 - 89. Contreras, R.J. and Stetson, P.W. (1981) Changes in salt intake after lesions of the area postrema and the nucleus of the solitary tract in rats. Brain Rex, 211: 355-366. Edwards, G.L. and Ritter, R.C. (1981) Ablation of the area postrema causes exaggerated consumption of preferred foods in the rat. Brain Res., 216: 276. Edwards, G.L and Ritter, R.C. (1982) Area postrema lesions increase drinking to angiotensin and extracellular dehydration. Physiol. Behav., 29: 943 - 947, Edwards, G.L. and Ritter, R.C. (1986) Area postrema lesions: cause of overingestion is not altered visceral nerve function. A m . J . Physiol., 251: R575-R581. Kobashi, M. and Adachi, A. (1985) Convergence of hepatic osmoreceptive inputs on sodium-responsive units within the nucleus of the solitary tract of the rat. J . Neurophysiol., 54: 212-219. Kobashi, M. and Adachi, A. (1986) Projection of nucleus tractus solitarius units influenced by hepatoportal afferent signal to parabrachial nucleus. J. Auton. Nerv. Sysf., 16: 153 - 158. Leslie, R.A. (1986) Comparative aspects of the area postrema: fine-structural considerations help to determine its function. Cell. Mot. Neurobiol., 6: 95 - 120. Mizuno, Y. and Oomura, Y. (1984) Glucose responding neurons in the nucleus tractus solitarius of the rat: in vitro study. Brain Res., 307: 109- 116.
83 Niijima, A. (1969) Afferent impulse discharges from glucoreceptors in the liver of the guinea pig. Ann. N. Y. Acad. Sci., 157: 690-700. Niijima, A. (1982) Glucose-sensitive afferent nerve fibers in the hepatic branch of the vagus nerve in the guinea-pig. J. Physiol. (London), 332: 315 -323. Oomura, Y. (1983) Glucose as a regulator of neuronal activity. In: A.J. Szabo (Ed.), Advances in Metabolic Disorders, Vol. 7, Academic Press, New York, pp. 31 -65. Oomura, Y . , Ono, T., Ooyama, H. and Wayner, M.J. (1969) Glucose and osmosensitive neurons of the rat hypothalamus. Nature (London), 222: 282 - 284.
Ritter, R.C. and Edwards, G.L. (1984) Area postrema lesion cause overconsumption of palatable foods but not calories. Physiol. Behav.. 32: 923 - 927. Ritter, S., McGlone, J.J. and Kelley, K.W. (1980) Absence of lithium-induced taste aversion after area postrema lesion. Brain Res., 201: 501 -506. Shapiro, R.E. and Miselis, R.R. (1982) An efferent projection from the area postrema and the caudal medial nucleus of the solitary tract to the parabrachial nucleus in rat. Proc. SOC. Neurosci., 8: 269. Wise, B.L. and Ganong, W.F. (1960) Effect of brain-stem stimulation of renal function. A m . J . Physiol., 198: 1291 1295.
71
CHAPTER 10
Electrophysiological analysis of chemosensitive neurons within the area postrema of the rat Akira Adachi and Motoi Kobashi Department of Physiology, Okayama University Dental School, Shikata-cho, Okayama 700, Japan
Summary
Introduction
The area postrema (AP) has proved to be a chemoreceptive trigger zone for the emetic response. Evidence indicating other functions of the A P especially in the rat has grown in abundance recently. The present study is an attempt to demonstrate chemosensitive neurons within the A P electrophysiologically. Two different techniques were employed for chemical stimulation of the AP neurons: (1) the floor of the exposed fourth ventricle was superfused with isotonic glucose solution, hypertonic or hypotonic Ringer solution, or LiCl Ringer solution; (2) these solutions were injected into the left carotid artery. To confirm vagal afferent influences on given A P neurons, single shock electrical stimulation was delivered to the cervical vagal trunks of both sides. Three types of chemosensitive neurons were identified: (1) glucose responsive neurons that may participate in the control of satiety, (2) sodium (osmotic pressure) responsive neurons that may contribute to salt preference or control of water balance, and (3) LiCl responsive neurons that may play a role in conditioned taste aversion. Some of these chemosensitive neurons were influenced by the vagal afferent inputs, suggesting a close relationship between chemoreceptor mechanisms of the AP and visceral afferents.
In general the function of the area postrema (AP) has been elucidated as being a chemoreceptive trigger zone for the emetic response (Borison and Wang, 1953). Evidence that indicates other functions of this structure is gradually accumulating (Borison, 1974). Lesions destroying the A P cause changes in salt intake without deficits in gustatory function in the rat (Contreras and Stetson, 1981). Ablation of the A P causes exaggerated consumption of preferred foods in the rat (Edwards and Ritter, 1981). Based on these facts, the authors postulated chemoreceptor functions of the A P not only for the emetic response, but for ingestive behavior that could be important for homeostatic control. Our preliminary study has revealed the existence of glucose responsive neurons as well as sodium responsive ones within the A P in the rat (Adachi and Kobashi, 1985). Glucose responsive neurons induced a marked decrease in the discharge rate in response to topical application of glucose by means of microelectro-osmotic techniques. Two different types of sodium responsive neurons were also identified by means of topical iontophoretic application of sodium ions. The purpose of the present study is to extend our previous observation in order to confirm further the glucose as well as sodium responsive neurons within the
78
AP by use of different techniques for chemical stimulation of the AP. In addition, the AP plays a critical role in the formation of conditioned taste aversions by lithium chloride (Ritter et al., 1980). In this study, effects of chemical stimulation by LiCl on AP neurons were also examined to elucidate LiCl responsive neurons besides the glucose or sodium responsive ones.
A
SUPERFUSION
B
INTRACAROTID INJECTION
Methods A total of 49 male Sprague - Dawley rats (Charles River of Japan), weighing 300 - 400 g were used as subjects. Under urethane - chloralose anesthesia (urethane 0.8 g/kg, chloralose 65 mg/kg, i.p.), a catheter was inserted into the left carotid artery toward the heart as shown in Fig. lB, in order to apply chemical stimuli to the AP through the blood vessel. The right carotid artery was ligated. To examine vagal afferent invasion in the AP, implanted stimulating electrodes were fixed on the cervical vagal trunks of both sides. In one case the celiac branch of the vagus was stimulated. Then the animals were mounted on a stereotaxic apparatus. To place the medulla oblongata in a horizontal position, the incisors were fixed 25 mm below the level of the auditory meatus. The occipital bone was removed with a dental drill, and the caudal part of the medulla was exposed. In this way, the surface of the AP was completely exposed without removing the cerebellum by suction. As illustrated in Fig. lA, the surface of the AP was superfused with standard Ringer solution (Na 147 mEq, K 4 mEq, Ca 4.5 mEq, C1 155.5 mEq). A glass pipette electrode filled with 0.5 M sodium acetate and 2% pontamine sky blue was used for recording unitary discharges. Under inspection through a dissecting microscope, the recording electrode was introduced into the AP by means of a micromanipulator until unit discharges were observed. Then single shock electrical stimulation was delivered to the cervical vagal trunks in order to confirm a vagal afferent supply to the AP. After this examination, superfusion with Ringer solution was switched to isotonic glucose solution (300
Fig. I . Schematic representation of the experimental procedure. Two different techniques employed for chemical stimulation of the AP neurons are illustrated.
mM), hypertonic, hypotonic Ringer solution or LiCl (147 mM) containing Ringer solution respectively. Hypertonicity was achieved by the addition of 147 mM NaCl to the standard Ringer solution. Hypotonicity was achieved by decreasing the NaCl concentration by 74 mM in the standard Ringer solution. These test solutions were also applied through the carotid catheter to elucidate responses of the AP neurons to the chemical stimuli supplied through the fenestrated blood vessel. All records were stored on magnetic tape for later analysis. The number of impulses in each 5 s was plotted by means of a pulse-rate meter. Recording sites were marked by iontophoretic dye injection and examined afterward histologically.
Results It is noticed that spontaneous discharges have been rarely observed within the AP in previous studies, due presumably to a sparse neuronal distribution in this structure. Therefore, responses have been infrequently obtained even when exploring within the AP by means of the microelectrode technique. Figs. 2 and 3 show two different types of AP neurons in response to glucose. When superfusion of the exposed surface of the fourth ventricle with standard Ringer solution was switched to isotonic glucose solution (indicated by G ) , a marked
19
decrease in discharge rate is seen (Fig. 2Aa). The decreased neuronal activity gradually returned to the control discharge rate when glucose solution was changed back to the Ringer solution. This response is characterized by a negative correlation between the discharge rate and glucose concentration. The superfusion of hypertonic Ringer solution containing 294 mM NaCl (indicated by S ) induced no response in this neuron. Fig. 2Ab shows responses of this glucose responsive neuron to single shock electrical stimulation (intensity 3 V , duration 3 ms) of the celiac branch of the vagus. The latency from the onset of single shock electrical stirnulation until the unit discharge was evoked was approximately 160 ms. Because preceding spontaneous discharges did not cancel the evoked discharges by collision (see the second and third traces from the top), this neuron was activated by afferent invasion via the celiac branch of the vagus. It implies that some neurons within the AP are responsive to changes in glucose concentration of the cerebrospinal fluid (CSF) and also receive afferent signals from the abdominal organs. Fig. 2Ba presents a response of this type of
Fig. 2. Responses of two different glucose responsive AP neurons. Superfusion of the fourth ventricle with isotonic glucose solution (G) induced a marked decrease in discharge rate of an AP neuron (Aa). Note that hypertonic Ringer solution (S) is ineffective. (Ab) Responses to single shock electrical stimulation of the celiac branch of the vagus. Each oscillograph sweep is triggered by a stimulus pulse. Evoked unit discharge is seen in each trace. (B,) Effect of intracarotid infusion of glucose on the other AP neuron. Note clear suppression by glucose (G) and no response to hypertonic Ringer solution (S). (Bb) Oscillograph record of discharges. Ten sweeps were superimposed. This neuron elicits no response t o electrical stimulation of the celiac branch of the vagus.
Fig. 3. (A) Effect of the superfusion with isotonic glucose solution ( G ) ,hypertonic (S) or hypotonic (S') Ringer solution. Only glucose elicits marked increases in discharge rate. The intracarotid infusion of glucose indicated by G(ca) induces the same response. (B) Evoked response to single shock electrical stimulation of the right cervical vagal trunk. (C) Oscillograph records of unit discharges: (1) spontaneous discharges recorded from the time indicated by arrow 1 in A; (2) discharges responding to glucose recorded from the time indicated by arrow 2 in A. Note a marked difference in the discharge rate between CI and C2.
glucose responsive neuron to the intracarotid infusion of isotonic glucose solution. This neuron rapidly responded to the glucose infusion (indicated by G ) , displaying a marked decrease in discharge rate followed by a return to the control discharge rate after cessation of the infusion. The same infusion with hypertonic Ringer solution (indicated by s) did not elicit any change in discharge rate. The response to the glucose infusion is similar to that of the superfusion with glucose solution. This unit did not respond to single shock electrical stimulation delivered to the celiac branch of the vagus as shown in Fig. 2Bb. Opposite responses to glucose were observed in the other AP neuron. As presented in Fig. 3A, superfusion with glucose (indicated by G ) markedly increased the discharge rate. Intracarotid infusion of the same solution also produced an increase in discharge rate as indicated by G(ca). However, no responses were observed when the superfusion with Ringer solution was switched to that with either hyper- or hypotonic Ringer solution (indicated by S and S ' respectively). This neuron responded to single shock electrical stimulation of the right cervical vagal trunk (same intensity and duration as in Fig. 2) but not to that of the left one
80
as shown in Fig. 3B. Oscillographic records of the discharges are illustrated in Fig. 3C, and C,. Comparison of C, with C, also clearly. indicates the increase in discharge rate in response to the superfusion with glucose. Typical responses to hyper- and hypotonicity of
Fig. 4. (A,) Response to the superfusion with hypertonic Ringer solution (S). Note a marked increase in discharge rate during a horizontal bar indicated by S followed by a decrease in discharge rate (after-effect). Hypotonic Ringer solution (S ’ ) causes a decrease in discharge rate but glucose ( G ) does not. (A,) Responses to electrical stimulation of the left and right vagus. Ten sweeps were superimposed. No response is elicited. (B,) Response to the superfusion with hypotonic Ringer solution (S‘). No response to superfusion with isotonic glucose ( G ) or hypertonic Ringer solution (S) is elicited. (Bb)The same oscillograph records as in A,. No response is elicited.
Fig. 5 . (A) Response of an AP neuron to the superfusion with LiCl containing Ringer solution (Li). Note the lack of response to equiosmotic hypertonic Ringer solution (S).Glucose is also ineffective. (B) Oscillograph records of discharges elicited by electrical stimulation of the vagus of both sides.
some A P neurons are presented in Fig. 4A and B. When the superfusion with standard Ringer solution was switched to that with hypertonic Ringer solution (indicated by S), a marked increase in discharge rate was observed (Fig. 4Aa). Immediately after switching back to standard Ringer solution, a suppression of neural response was recognized. This neuron shows an opposite response to hypotonic Ringer solution (indicated by S ’ ) , so a marked decrease in discharge rate is seen during a horizontal bar indicated by S ’ . No response is elicited by switching the superfusion to isotonic glucose solution (indicated by G). Single shock electrical stimulation of the cervical vagal trunk of both sides elicited no response in this neuron as shown in Fig. 4.4,. Fig. 4Ba presents a response of the other A P neuron to hypotonicity. This neuron responded only to the hypotonic superfusion (indicated by S ‘ ) with a marked increase in discharge rate. Both isotonic glucose solution (G) and hypertonic Ringer solution (S) were ineffective. The same electrical stimulation of the vagus elicited no response as shown in Fig. 4Bb. These neurons were not supplied by the vagal afferent inputs. A specimen record from a lithium chloride responsive neuron within the A P is shown in Fig. 5 . When the superfusion with standard Ringer solution was switched to that with LiCl containing Ringer solution (indicated by Li), an increase in discharge rate was observed. However, hypertonic Ringer solution that was equiosmotic to LiCl Ringer solution elicited no response (Fig. 5A). It indicates that the LiCl responsive neuron is insensitive to osmotic pressure as well as glucose. This neuron responded to electrical stimulation of the cervical vagal trunks of both sides. Fig. 5B presents evoked discharges by electrical stimulation of the left and right vagus (same intensity and duration as in Fig. 2). Latencies are approximately 46 ms. A total of 46 neurons within the area postrema were examined. Among them, 25 neurons responded to electrical stimulation of the cervical vagal trunk. Four neurons decreased and one neuron increased the discharge rates in response to glucose.
81
Two neurons increased and two other neurons decreased the discharge rates in response t o hypertonic Ringer solution. One neuron increased and one neuron decreased in response to hypotonic Ringer solution. Two LiCl responsive neurons were also observed. Some of these neurons responded only to chemicd stimuli, others responded to both chemical and single shock electrical stimulation of the vagus. Thus, no correlation between the vagal afferent supplies and chemosensitivities of these AP neurons was recognized.
Discussion The AP is a circumventricular organ of the fourth ventricle and lacks a blood - brain barrier. The existence of neural elements within this organ has been confirmed histologically (Leslie, 1986). These facts suggest various chemoreceptor mechanisms to detect blood- and CSF-borne chemical information not only for triggering the emetic response but also for homeostatic control. Many ablation experiments of the A P indicated other functions than the emetic response, for instance, control of satiety (Edwards and Ritter, 1981, 1986; Ritter and Edwards, 1984), control of sodium and water balance (Wise and Ganong, 1960; Contreras and Stetson, 1981; Edwards and Ritter, 1982), formation of conditioned taste aversion (Berger et al., 1973; Ritter et al., 1980) and cardiovascular control (Barnes et al., 1984). A postulate of AP chemoreceptors is common to these authors. The present study provides direct evidence for the existence of such chemoreceptor mechanisms which participate in these controls. The glucose responsive neurons within the A P may play a role in the control of food intake. Oomura et al. (1969) were the first to reveal glucose responsive neurons in the hypothalamus by electrophysiological methods. Glucose responsive neurons were also found in the nucleus tractus solitarii (NTS) (Adachi et al., 1984; Mizuno and Oomura, 1984). Neural connections between the glucose responsive neurons in the central nervous system and a peripheral hepatic glucoreceptor (Nii-
jima, 1969, 1982) was elucidated (Shimizu et al., 1983; Adachi et al., 1984). These neural networks suggest a hierarchical organization of glucose information in the body fluid processing by central and peripheral nervous systems (Oomura, 1983). The AP responsive neurons could be involved in this organization. The possible function of the A P sodium (or osmo-) responsive neurons favors a similar conception to that of the glucose responsive ones in association with the control of sodium and water balance of the body fluid. Sodium (or osmo-) responsive neurons are located in the hypothalamus (Oomura et al., 1969) and the NTS as well (Kobashi and Adachi, 1985). Peripheral osmoreceptors exist in the liver (Adachi et al., 1976) and their afferent signals project to the caudal portion of the NTS (Adachi, 1981, 1984) and in turn to the hypothalamus via the parabrachial nucleus (PB) (Kobashi and Adachi, 1986). Convergence of the hepatic osmoreceptive signals onto the NTS sodium responsive neurons was also confirmed (Kobashi and Adachi, 1985). A survey of these previous reports strongly suggests involvement of the AP sodium responsive neurons in the neural control system for isosmosis and isovolemia of the body fluid. Carpenter and Briggs (1986) recorded responses of neurons of the canine A P to iontophoretic application of insulin. They concluded that these responses provide further support for a critical role of the AP in triggering the emetic reflex. The existence of LiCl responsive neurons within the rat AP indicates that this brain region is involved in neural mechanisms inducing nausea, even though the emetic response is not easily elicited in the rat. In turn, such chemosensitive neurons may play a part in the formation of conditioned taste aversion. As many other chemicals such as scopolamine, amphetamine, hydroxytryptophan and so on induce conditioned taste aversion, it may be possible that the LiCl responsive neuron also responds to these chemicals. A postulated scheme is presented in Fig. 6. Because some of the chemosensitive AP neurons
82
References
PB NTS
X
Fig. 6. Schematic diagram illustrating a possible neural connection of AP neurons. CSF, cerebrospinal fluid; R, chemoreceptor neuron; C, capillary; FE, fenestration; AP, area postrema; PB, parabrachial nucleus; NTS, nucleus tractus solitarii; X, vagus.
are excited by the vagal afferent but others are not, there can be a possibility that receptor-like neurons (R) are located in the AP. The former neurons themselves can also be chemosensitive, because convergence of the vagal afferent signals to chemosensitive neurons in the NTS has been ascertained (Adachi et al., 1984; Kobashi and Adachi, 1985). These neurons are responsive to either blood- or CSF-borne substances, because application of the chemical stimuli to both the fourth ventricle and the carotid artery elicited the neuronal responses. Close connections between the A P and the PB as well as the NTS have been clarified (Shapiro and Miselis, 1982). It is easily delineated that the internal chemical information detected in the AP may be transmitted to the other nuclei in the central nervous system, so that various reflexes including emesis can be induced. Certainly further studies should be necessary to elucidate such neural mechanisms. Acknowledgements
This work was supported by Grant-in-Aid for General Scientific Research 62480378 from the Ministry of Education, Science and Culture in Japan and a Grant from the Society for Research on Umami Taste.
Adachi, A. (1981) Electrophysiological study of hepatic vagal projection to the medulla. Neurosci. Lerr., 24: 19-23. Adachi, A. (1984) Projection of the hepatic vagal nerve in the medulla oblongata. J. Auton. Nerv. Syst., 10: 287-293. Adachi, A. and Kobashi, M. (1985) Chemosensitive neurons within the area postrema of the rat. Neurosci. Lerr., 5 5 : 137- 140. Adachi, A., Niijima, A. and Jacobs, H.L. (1976) An hepatic osmoreceptor mechanism in the rat: electrophysiological and behavioral studies. Am. J. Physiol., 231: 1043 - 1049. Adachi, A,, Shimizu, N., Oomura, Y. and Kobashi, M. (1984) Convergence of hepatoportal glucose-sensitive afferent signals to glucose-sensitive units within the nucleus of the solitary tract. Neurosci. Lett., 46: 215 -218. Barnes, K.L., Ferrario, C.M., Chernicky, C.L. and Borosnihan, K.B. (1984) Participation of the area postrema in cardiovascular control in the dog. Fed. Proc., 43: 2959 - 2962. Berger, B.D., Wise, C.D. and Stein, L. (1973) Area postrema damage and bait shyness. J. Comp. Physiol. Psycho/., 83: 475 - 479. Borison, H.L. (1974) Area postrema: chemoreceptor trigger zone for vomiting - is that all? Life Sci., 14: 1807- 1817. Borison, H.L. and Wang, S.C. (1953) Physiology and pharmacology of vomiting. Pharmacol. Rev., 5 : 193 - 230. Carpenter, D.O. and Briggs, D.B. (1986) Insulin excites neurons of the area postrema and causes emesis. Neurosci. Lerr., 68: 85 - 89. Contreras, R.J. and Stetson, P.W. (1981) Changes in salt intake after lesions of the area postrema and the nucleus of the solitary tract in rats. Brain Rex, 211: 355-366. Edwards, G.L. and Ritter, R.C. (1981) Ablation of the area postrema causes exaggerated consumption of preferred foods in the rat. Brain Res., 216: 276. Edwards, G.L and Ritter, R.C. (1982) Area postrema lesions increase drinking to angiotensin and extracellular dehydration. Physiol. Behav., 29: 943 - 947, Edwards, G.L. and Ritter, R.C. (1986) Area postrema lesions: cause of overingestion is not altered visceral nerve function. A m . J . Physiol., 251: R575-R581. Kobashi, M. and Adachi, A. (1985) Convergence of hepatic osmoreceptive inputs on sodium-responsive units within the nucleus of the solitary tract of the rat. J . Neurophysiol., 54: 212-219. Kobashi, M. and Adachi, A. (1986) Projection of nucleus tractus solitarius units influenced by hepatoportal afferent signal to parabrachial nucleus. J. Auton. Nerv. Sysf., 16: 153 - 158. Leslie, R.A. (1986) Comparative aspects of the area postrema: fine-structural considerations help to determine its function. Cell. Mot. Neurobiol., 6: 95 - 120. Mizuno, Y. and Oomura, Y. (1984) Glucose responding neurons in the nucleus tractus solitarius of the rat: in vitro study. Brain Res., 307: 109- 116.
83 Niijima, A. (1969) Afferent impulse discharges from glucoreceptors in the liver of the guinea pig. Ann. N. Y. Acad. Sci., 157: 690-700. Niijima, A. (1982) Glucose-sensitive afferent nerve fibers in the hepatic branch of the vagus nerve in the guinea-pig. J. Physiol. (London), 332: 315 -323. Oomura, Y. (1983) Glucose as a regulator of neuronal activity. In: A.J. Szabo (Ed.), Advances in Metabolic Disorders, Vol. 7, Academic Press, New York, pp. 31 -65. Oomura, Y . , Ono, T., Ooyama, H. and Wayner, M.J. (1969) Glucose and osmosensitive neurons of the rat hypothalamus. Nature (London), 222: 282 - 284.
Ritter, R.C. and Edwards, G.L. (1984) Area postrema lesion cause overconsumption of palatable foods but not calories. Physiol. Behav.. 32: 923 - 927. Ritter, S., McGlone, J.J. and Kelley, K.W. (1980) Absence of lithium-induced taste aversion after area postrema lesion. Brain Res., 201: 501 -506. Shapiro, R.E. and Miselis, R.R. (1982) An efferent projection from the area postrema and the caudal medial nucleus of the solitary tract to the parabrachial nucleus in rat. Proc. SOC. Neurosci., 8: 269. Wise, B.L. and Ganong, W.F. (1960) Effect of brain-stem stimulation of renal function. A m . J . Physiol., 198: 1291 1295.