Effect of intestinal amino acid infusions on hypothalamic single unit activity in the anesthetized cat

Brain Resrcwch B~ii~ii~7, Vol. 10, pp. 15-21, 1983. Printed in the U.S.A.

Effect of Intestinal Amino Acid Infusions on Hypothalamic Single Unit Activity

in the Anesthetized Cat

RI?GINE JEANNINGROS

~~parte~ent de ~~u~op~y~iui~~gie V~g~t~ti~e I.N.P.01 Institut de ~e~~op~~y~i~~~ogi~ et P~y~~~p~ysi~~l~gie du C.N.R.S.

31 Chemin Joseph Aiguier-B.P. 71 13277 Marseille, Cedex 9, Frunw

Received 10 May 1982

JEANNINGROS, R. Effcw c>f intestinrrl nmintr acid injirsions on hypvpotholrrmic single unit trctivir_v in the trwsthetiztd cut. BRAIN RES BULL lo(l) 15-21, 1983.---Single unit discharges in the ventromedial hypothalamic nucleus (VMH) and lateral hy~thalamic area (LH) were recorded extracellularly in anesthetized cats, while amino acid solutions were perfused through the small intestine via implanted cannulae. Test infusions consisted of 5 amino acid mixtures (arginine, leucine, phenylaianine, tryptophan, alanine: 50 mM each), 5.U% casein hydrolysate; arginine 12.5 mM (2.2%); ieucine 125 mM (1.6%). Sixty-six units were recorded in 30 cats. Of 52 tested in the LH, 19 or 37% were affected: 14 increased and 5 decreased in firing rate in response to amino acid intestinal perfusions with a very short latency consistent with activation of intestinal amino acid receptors. Of 14 neurons tested in VMH, only 3 or 21% were activated by amino acid infusions but with a long latency of several minutes which cannot exclude the involvement of a postabsorptive signal. Results are discussed in terms of intestinal afferent control over hypothalamic neuronal activity related to amino acid induced satiety.

Lateral hy~othaIamus Ventromedial hypothalamus Single unit activity Cat

NUMEROUS behavioral studies have demonstrated that the gastrointestinal tract plays an important role in the control of ingestion 14, 5, 6, 12, 13, 18, 21, 221. The presence of nutrients in the duodenum reduces subsequent food intake in var- ious species. Gibbs et al. [6] recently demonstrated that the accumulation of food in the small intestine is necessary to produce normal size meals. Several studies have provided strong evidence that satiety is nutrient specific and postu- lated the chemospecificity of intestinal signals eliciting short term satiety [2,5]. Carbohydrates and amino acids infused in the intestine elicited satiety in sham feeding animals before any significant modification of blood concentration occurred. Novin [181 pointed out that amino acids may also influence feeding through a vagal mechanism. Intestinal vagal receptors responsive to the presence of amino acids in the lumen have been recently identified electrophysiologi- tally in the anestheti~d cat [lo]. They are distinct from intestinal glucoreceptors previously described, since they are in general unresponsive to glucose infusions [ 151.

The adequate stimulus for these preabsorptive receptors was recently identified in our laboratory [ 101. Whether or not the stimulus is effective in modifying hypothalamic activity was an unresolved question. The behavioral experiments of Novin, which were based on the effects of hypothalamic lesions, suggest that amino acid satiating effects are not mediated by the lateral hy~thalamic area (LH) in the rat IlS]. The only single unit recordings following alimentary perfusions varying in osmolality and nutrient properties have been performed in the hypothalamus of anesthetized rat by

Amino acid intestinal infusions Short term satiety

Maddison and Horrell [ 141. They have provided little snpport that nutrient properties of the perfusates exerted some modifying influence on hypothalamic unit discharges.

The purpose of the present experiments was to examine the effects of intestinal amino acid infusions, which are known to activate vagal intestinal amino acid receptors, on the electrical activity of ventromedial (VMH) and lateral hypothalamus (LH). Single unit activity of the LH and VMH was recorded extracellularly in anesthetized cats while amino acid infusions were perfused in the intestinal lumen. Results indicate that intestinal amino acid infusions known to activate the identified chemoreceptors may modulate LH neuronal activity. The effects were specific to chemical stimulation and not related to intestinal distension. There was no evidence that amino acid perfusions exerted any preabsorptive influence on VMH activity.

METHOD

Experiments were performed in 30 male and female adult cats weighing between 2.5 and 3.0 kg, anesthetized with chloralose injected intravenously (75 mg/kg, IV) after halo- thane induction. Animals were curarized and ventilated arti- ficially. Rectal body temperature was maintained at 382 1°C by means of a warming blanket.

The cats were fixed in a stereotaxic apparatus and a lo- calized hemicraniotomy was achieved on the left side. The

15

16 JEANNINGRCXS

Al

i I

A2

Amino acids

FIG. 1. Typical bursts activation of LH neurons by an intestinal amino acid infusion occurring at a short latency cabou~ 10 SW). Solid black bar indicates period of infusion,

holes were located over predeterm~ed recording brain sites according to a cat brain atlas [S]. Electrode placements were at the following coordinates, in mm: VMH 10.5 to I 1.5 posterior; 1.5 lateral: 5.5 to 6.5 deep; LH 10.5 and 12.0 posterior; 3.0 lateral; 2.0 to 4.5 deep.

Microelectrodes were driven by a micromanipulator to the predete~ined neural sites. Extracell~ar action poten- tials were recorded through tungsten microelectrodes having a DC resistance of 5 Mf2 at 1000 Hz, lo-” A. Action poten- tials were amplified by a high input impedance preamplifier, displayed on a conventional Tektronix oscilloscope, recorded on magnetic tape and optical UV recorder and then analyzed by an impulse frequency meter on the basis of am- plitude and frequency.

Recording sites were marked by passing 20 PA of anodal direct current through the electrode tip for a 20 set period. The head of the animal was perfused with a 1% form01 solution at the end of the experiment. The brain was fixed in the same solution for 24 hr and sectioned sagittally by means of a freezing microtome.

Duodentrl Perfusiotz

A segment of the proximal part of the small intestine in- cluding the duodenum, except the first 2-3 cm corresponding to the intestinal bulb, and the jejunum was selected and can- nulated both proximally and distally. This technique has been utilized successfully in previous studies in which ex- tracellular recordings were made from nodose g~giion neurons under similar conditions [lO,lSl. A clamp prevented the in~s~ons from reaching the stomach. This segment was perfused with 25 ml of test or control soiutions maintained at a constant temperature (38°C) in a water bath, and delivered over a 10 second period by means of a syringe. Amino acid solutions were dissolved in a Krebs Henseieit buffer pH 7.4. Perfusion stimuli consisted of the following solutions: 5.0% casein hydrolysate; a mixture of 5 amino acids (arginine, leucine, phenylalanine, tryptophan, afafline at a concentration of 0.05 molesil each); a&nine 0.125 moles/l (2.2%): leucine 0.125 moles/l (1.6%).

Thirty to 60 seconds later the remaining perfusate was flushed out by means of air infusion, then the intestinal lumen was rinsed with 25 ml of bicarbonate saline pH 7.4. At least 20 minutes were allowed to elapse before presentation of the next stimuius.

Spikes 13s

Ii ! ! 6

contra: / 0

-15 0 15 30 45 60

Time (S)

FIG. 2. Spikes per second presented as a function of time in seconds for an LH cell. A: chemosoecific effect of amino acid intestinal

1

infusion. B: no effect of the control in&ion., Arrows iixlicate when the intestinal infusion started and stoped during the experiment.

Control infusions were carried out using bicarbonate saline alone, before and after the amino acid infusion. In addition, rapid injection of air was used to test for possible mechanical effects related to intestinal wall distension and to satisfy the criteria of chemospecifcity for amino acid re- SpWtSeS.

Electrid Shdurion of‘ Vugtcb !I’cr\7~

The left cervical vagus was dissected away from the carotid artery. The whole vagus nerve in the mid-cervical

INTESTINAL AMINO ACIDS ON HYPOTHALAMUS 17

Control

A2

__-- .-

I I I j j 1)” I I I l”lll

- Leucine

TT- I

FIG. 3. ~~~rnos~eci~c &ect of one single amino acid (leucine 125 mM) infused in the intestine on LH neuronal activity, in contrast to a mechanical effect of saline infusion. A: control infusion (bicarbonate saline) has no effect on the large unit, whereas a mechanical inhibition of the little unit occurs. B: activation by intestinal infusion of leucine of another large LH unit, which occurred at a latency of 20 set (indicating a chemical effect). The recordings 1 and 2 are continuous in time. C: pluriunitary response following electrical stimulation of vagus nerve at the same recording site. Solid black bars indicate periods of infusion.

region was stimulated by means of a bipolar silver stimulat- ing electrode which was tied in place around the nerve. Stimulus parameters used were 45 V, 0.5 msec so that the C fibers which innervate the intestine were also stimulated. These values were selected on the basis of results obtained in a previous study [9] and where C fibers were specifically stimulated by the intravenous administration of phenyl- diguanide [ 10,151. Recordings were made from ipsilateral neurons in VMH and LH, before amino acid stimulations, in order to verify the existence of vagal projections at the recording site.

RESULTS

A total of 66 cells were tested in 30 cats, 52 were recorded from the LH (26 experiments) and 14 from the VMH (13 experiments). Units which were responsive to mechanical stimulations, such as intestinal distention, pinching or digital compressions of the wall, were not considered in this study. On the basis of their latencies, the hypothalamic responses were classified in two groups: (1) cells exhibiting changes in firing rate with a latency less than or equal to 20 set; (2) cells responding with a latency greater than one minute. The latency of the first group unequivocally indicates a preabsorptive effect of chemical infusions, since within this time, no sign&ant changes in the blood amino acid concentration had appeared [7]. Therefore, under these conditions, effects are obviously due to a preabsorptive effect of amino acid

intestinal infusions. With the long latency of the second group, we cannot be confident that the response of hypothalamic neurons was due to a preabsorptive effect of the intestinal infusion. For that reason, only the cells belonging to the first group were considered in this study. In addition, the effect of vagus nerve stimulation was tested at the recording site on some unitary responses to amino acids.

Of the 52 units tested which were generally silent, 19 or 37% responded in a chemospecific manner to amino acid intestinal infusions, 14 exhibited increases in firing rate and 5 exhibited decreases in firing rate. Figure 1 shows a typical activation of a LH unit by amino acid intestinal infusion. All the LH units responsive to amino acids belonged to type I, since changes in their discharge frequency occurred at a latency of 2-20 sec. Figure 2 illustrates a chemospecific activation of a LH unit occurring some seconds after amino acid infusion. No effect was produced with infusion of buffer alone. All the units taken into consideration responded chemospecifically to amino acid infusions, since neither con- troi infusions nor mechanical stimulations were effective in producing changes in their discharge frequency.

The mixture of 5 amino acids or one amino acid alone were also tested and were able to produce a change in the LH neuronal activity. Figure 3 illustrates the chemospecific activation of leucine infusion (125 mM) occurring at a latency of 20 sec. A control infusion with bicarbonate saline alone

18 JEANNINGROS

RINSING

-qq-Yzi ST. V;GUS

200

FIG. 4. Chemospecific effect of both intestinal amino acid infusion and rinsing. A: activation of amino acid of a LH neuron occurred at a latency of 4 sec. B: the emptying of intestine has no significant effect. C: the mechanical distention of intestine 6th a rapid injection ofair has no effect on this unit. D: the rinsing with buffer produced an important rebound of the chemosensitive unit, occurring at the same latency as the amino acid effect produced in A. Recordings 1 and 2 are continuous in time. At this recording site. the electrical stimulation of vagus nerve produced a pl~ri~~i~ry response of LH cells. Solid black bars indicate periods of infusion.

was not able to activate this leucine sensitive unit but did inhibit immediately another spontaneous unit. When amino acids were flushed out after a test infusion and then the lumen immediately rinsed with bicarbonate saline, sometimes the effect produced by amino acids was enhanced. This special effect of the rinsing had always a latency similar to the test latency and was not due to a mechanical stimuia- tion. Figure 4 illustrates the rapid activation, occurring at a latency of 4 set, of a LH unit by amino acid infusion (A 1, AZ). The emptying of intestine (B) decreased the firing rate which, then, was strongly reactivated at a short latency (4 see) by the rinsing of intestine (Dl, D2). A sudden injection of air was ineffective in changing the discharge frequency of this unit, proving that the effects of both amino acid and rinsing are related to chemical stim~ation (C). Figure 5 shows a LH cell which is activated by both amino acid and glucose intestinal infusions, at the same latency of 9 set which is not consistent with the latency of a mechanical effect. On the contrary, some cells in the VMH were inhib- ited by the same amino acid infusions at a very short latency (less than 0.5 set), demonstrating that the effect of intestinal perfusion is related to a mechanical stimulation. Figure 6 illustrates a non-specific effect of amino acid infusion. The activation of this LH unit is immediate (less than 0.5 set) and is reproduced by bicarbonate saline alone. This is typical of the effect of mechanical stimulation of the intestinaf wall.

Only 3 units or 21% out of the 14 tested in the VMH responded in a chemospecific manner to amino acid infusions by increasing their discharge frequency. All the units correspond to type II defined in the-first paragraph of result%, since the fatencies of responses were very high (several minutes). All these responses could not be attributed to mechanical effects. Figure 7 (A and A’) shows a strong active- tion of several VMH units 3 minutes after an ammo acid infusion, whereas a glucose infusion (10%) produced an op- posite and more rapid effect occurring in some seconds (Fig. 7C and C’). Both amino acid and glucose infusions produced a specific effect related to chemical stimulation, since mechanical stimulations such as rapid intestinal~ injections of water elicited an immediate inhibition of these units (Fig. 7B). The same mechanical inhibition occurred ifglucose was infused rapidly.

While recording in VMH and LH, an electrical stimulus known to activate also the non-medullated fibers (C type) was sometimes tested in regard to the effect on units responsive to amino acid infusion. In addition to a multiun~t evoked response (Figs. 3 and 4). the vagal stimulation produced changes in discharge frequency of the units responsive to amino acids in the same way as the effects of the intestinal infusions.


VMH

Bl Ii

AA5 5~UVl 28

FIG. 5. Simultaneous recording in VMH and LH. A: mechanical effect of amino acid intestinal infusion on VMH cells occurring at a very short latency (less than 0.5 set). B: chemospecific effect of amino acids on a LH unit occurring at a latency of 9 set, and persisting after 1 min. C: chemospecific effect of glucose infusion on the same LH unit responsive to amino acid, occurring at the same latency of 9 sec. Recordings I and 2 are continuous in time. Solid black bars indicate periods of infusion.

Amino acid 1

I - Control

FIG. 6. Typical mechanical effect of amino acid infusion (A) which is characterized by both a very short latency (less than 0.5 set) and an identical effect of the control infusion (B) with bicarbonate saline alone, occurring with the same very short latency. Solid black bars indicate periods of infusion.

Microscopic examination of the brain sections confirmed that recording sites were within the LH or the VMH as designated in the atlas of Jasper and Ajmone-Marsan IS].

DISCUSSION

These data provide evidence that amino acids infused in the intestine result in the modulation of LH neuronal activity: 37% of the neurons tested in LH were affected in the anesthetized cat. This effect is chemospecific, since 19 LH cells responsive to amino acids were affected neither by control infusions using the buffer alone, nor by mech~ical stimulations. The changes in frequency following amino acid infusions occur at a short latency of 2-20 sec. This is quite consistent with the latency of activation of vagai amino acid receptors at the intestinal level HO], which depends upon the

time necessary for the chemical substance to reach the nervous endings in the lamina propria. Recently, studies performed in the rat intestine in vivo provided evidence that, within this short time and until 30 see, amino acid infusions result in no significant changes in amino acid blood concentration [7]. So, it can be logically assumed that the modulation observed in the lateral hypothalamic activity is supplied by preabsorptive signals arising from the intestine. The vagal amino acid intestinal receptors, recently identified eIec- trophysioIogica1, are in good position to convey rapid information about the amino acids before they reach the blood. Our previous studies in the same animal have evidenced large projections of vagus nerve at the hy~thaIamic level [91.

Behavioral reports indicated that the satiating effect of chemical intestinal loads are vagally mediated in the rat fl8]. In several cases where amino acid infusions resulted in a

20 JEANN tNCROS

A

FIG. 7. Effect of intestinal infusions on VMH multiunitary activity. A: an amino acid infusion performed in A produces a strong multiunitary activation which occurs only after 3 mm (A’). B: a rapid injection of distilled water (E.D.) produces an immediate multiunitary inhibition related to a mechanical effect. An injection of air produced the same effect. C: a slow infusion of glucose produce\ a multiunitar~ inhibition with a short latency of 25 see related with a chemospecific effect of glucose which is different of amino acid effect. Recordings C, C’ :md C” are continuous in time. Solid black bars indicate periods of infusion

significant change in frequency, stimulation of vagus nerve also resulted in a significant change in frequency in the same direction. The rebound activity observed after the rinsing of intestinal content has also been reported by Maddison and Horrell in the rat following alimentary intestinal perfusions [14]. They suggested that there were no relationships between hypothalamic unit activity and the flushing perfusion of distilled water. In fact, the same enhancing effect of rinsing has also been observed at the vagai protoneurone level using vagal activity recording in the nodose ganglia in the cat [lo]. This effect on LH units is thus not surprising, since it quite reflects the activity modulation of intestinal chemoreceptors. It is noteworthy that the distilled water or bicarbonate saline alone elicited no changes in firing rate of LH units responsive to amino acids. It is likely that this rebound activity nonspecifically occurred as the nutrient intestinal absorption was enhanced, so that the stimulation of nervous endings was facilitated. It was also observed when the intestinal motility was increased (Gonella, personal communi~tion). The observation that the rebound activity produced by the rinsing could be higher than the activity

resulting from amino acid infusion is difficult to explain. The possible activation of the same LH unit by both glucose and amino acid intestinal infusions, as illustrated in Fig. 5. indicated that lateral hypothalamic neurons do not have a very strict specificity regarding the infused nutrient. However, all the LH responses considered in this study were discrimimr- tive of chemical and mechanical stimuli. Since the chemoreceptors located in the intestine are relatively specific in regard to the type of nutrient 1 IO], it can be suggesterf that different messages arising from the intestine are inte- grated at the hypothalamic levei. Several examples of inte- grative properties of the lateral hypothalamic neurons have already been reported [ 17,Zlj.

In contrast to LH, few VMH cells were affected by chemical intestinal stimulation. The present results suggest that amino acid intestinal infusions do not rapidly affect the spontaneous firing rates of VMH cells. In three cats, significant changes in discharge frequency were observed in the VMH after several minutes. These effects cannot be related unequivocally to a preabsorptive signal, and possible effects of changes in amino acid blood concentration on postabsorp-


tive receptors located at liver or hypoth~amic level must be kept in mind 120,211. These findings are consistent with No- vin’s results exhibiting that amino acid intestinal infusions did not change the food intake in VMH lesioned or vagoto- mized rats 181. But more extensive experimentation is re- quired in order to determine the possible modulation of VMH by intestinal chemical information. Nevertheless, the chemospecific effect of glucose intestinal infusion on VMH cells which occur in some seconds (Fig. 7) is quite consistent with the activation of intestinal glucoreceptors [15]. Moreover, two types of nutrients have produced two oppo- site effects with different latencies indicating the involvement of different mechanisms.

Previous studies have demonstrated that mechanical or nutrient properties of gastrointestinal contents influence hypothalamic single unit activity, in different species [l, 3, 141. Results recently obtained in our laboratory show large projections of mechanoreceptors located at different sites along

the gastrointestin~ tract (manuscript in preparation). But until the present study, there was little support that amino acid intestinal infusions affect LH neuronal activity [ 141. The results of these experiments suggest a distinct modulation of VMH and LH by amino acids infused in the intestine of the cat. In addition, these findings support the hypothesis of a vagal influence of LH through the activation of amino acid intestinal receptors, while a preabsorptive modulation of VMH by amino acids seems to be absent. This provides additional support for the integration of chemical signals arising from intestine in short term satiety mechanisms con- trolled by LH [16,18] and could be interpreted as further evidence of no involvement of VMH by such preabso~tive amino acid-induced satiety signals [ 113. Finally, these electrophysiological findings obtained in the cat are consistent with the previous behavioral and electrophysiological data obtained in the rat.

REFERENCES

1. Anand, B, K. and R. V. Pillai. Activity of single neurons in the hypothalamic feeding centers: Effect of gastric distension. J. Physiol. 192: 63-67, 1%7.

2. Anderson, G. H. Control of protein and energy intake: Role of plasma amino acids and brain nemotransmitters. Can. J. ~hysiol. Phrrrmuc,. 57: 1043-1057, 1979.

3. Barone. F. C.. M. .I. Wavner. C. S. Weiss and C. R. Almli. Effects of intragastric water infusion and gastric distension on hypothalamic neuronal activity. Bruin Rcs. Bull. 4: 267-282, 1979.

4. Campbell, C. S. and J. 0. Davis. Peripheral control of food intake: Interaction between test diet and postingestive chemoreception. Physiol. B&or. 12: 377-384, 1974.

5. Geary, N. Food intake and behavioral caloric compensation after protein depletion in the rat. Physial. Behnv. 23: 108%1098. 1979.

6. Gibbs, J., S. P. Maddison and E. T. Rolls. Satiety role of the small intestine examined in sham-feeding Rhesus monkeys. J. cctmp. physiol. Psycho!. 95: 1003-1015, 1981.

7. Hajjar, J. J. and H. P. Schedl. Amino acid influx across the mucosal border of the rat intestine in vivo. Biochim. biophys. Actrr 649: 759-768, 1981.

8. Jasper, H. H. and C. Ajmone-Marsan. A Stertwtuxic Atlas of the Di~ncephulon of the Cut. Ottawa: The National Research Council of Canada, 1954.

9. Jeanningros, R. and N. Mei. Vagal and splanchnic effects at the level of the ventro-median nucleus of the hypothalamus in the cat. Brain Rcs. 185: 239-251, 1980.

10. Jeanningros, R. Vagal unitary responses to intestinal amino acid infusions in the anesthetized cat: A putative signal for protein- induced satiety. Physiol. Behnv. 28: 9-21, 1982.

11. Le Magnen, J. The metabolic basis of dual periodicity of feeding in rats. Brhtrv. Bruin Sci. 4: 561-607, 1981.

12. Lepkovsky, S., M. K. Dimick, F. Furuta, S. E. Feloman and R. Park. Stomach and upper intestine in the rat in the regulation of food intake. J. Nutr. IDS: 1491-1499, 1975.

13. Liebling, D. S., J. D. Eisner, J. Gibbs and G. P. Smith. Intesti- nal satiety in rats../. camp. physiof. Psychof. 89: 955-965, 1975.

14. Maddison, S. and R. I. Horrell. Hypothalamic unit responses to alimentary perfusions in the anesthetized rat. Bruin Rrs. Bull. 4: 259-266, 1979.

15. Mei, N. Vagal glucoreceptors in the small intestine of the cat. J. Physiol. 282: 485-506, 1978.

16. Myers, R. D. and M. L. MC Caleb. Feeding: satiety signal from intestine triggers brain’s noradrenergic mechanism. Science 209: 1035-1037, 1980.

17. Nicolaidis, S. Lateral hypothalamic control of metabolic factors related to feeding. Djffb~t~~~~~g~~ 20: 426-434, 1981.

18. Novin, D., J. Sanderson and M. Gonzalez. Feeding after nutrient infusions: Effects of hy~thalami~ lesions and vagotomy. Physiol. Behav. 22: 107-l 13, 1979.

19. Rolls, E. T. Neurophysiology of feeding. In: Appetite and Food Intake, edited bv T. Silverstone. Dahlem: Verlaa Chemie. 1976. pp. 21-42. -

20. Russek, M. Hepatic receptors and the neurophysiological mechanisms controlling feeding behavior. In: Neuroscience Re- senvch, vol. 4, edited by S. Ehrenpreis and 0. C. Solinksky. New York: Academic Press, 1971, pp. 213-283.

21. Smith, G. P. and J. Gibbs. Postprandial satiety. Prog. Psychnbiol. physiol. Psychol. 8: 179-242, 1979.

22. Vanderweele, D. A,, D. Novin, M. Rezek and J. D. Sanderson. Duodenal or hepatic-portal glucose perfusion: Evidence for duodenaliy-based satiety. Physic>/. Behav. IZ: 467-473, 1974.

23. Wayner, 111. J.. T. Ono, A. De Young and F. C. Barone. Effects of essential amino acids on central neurons. Phu~~~i~~. Biwkrm. Behtrv. 3: Suppl. 1, 85-90, 1975.

Documents

Effect of intestinal amino acid infusions on hypothalamic single unit activity in the anesthetized cat