taste neurons in the cortex of the alert cynomolgus monkey

4
Brain ResearchBulletin, Vol. 28, pp. 333-336. 0 Pergamon Press pk. 1992. Printed in the U.S.A 036179230/92 $5.00 + .OO RAPID COMMUNICATION Taste Neurons in the Cortex of the Alert Cynomolgus Monkey CARLOS R. PLATA-SALAMAN AND THOMAS R. SCOTT*’ School of Life and Health Sciences and *Department of Psychology, University of Delaware, Newark, DE 19716 Received 21 October 1991 PLATA-SALAMAN, C. R. AND T. R. SCOTT. Taste neurons in the cortex of the alert cynomolgus monkey. BRAIN RBS BULL 28(Z) 333-336, 1992.-The activity of single neurons in the gustatory cortex of alert cynomolgus monkeys was analyzed. Taste-evoked activity in response to the four prototypical taste stimuli was recorded from a cortical gustatory area comprising the frontal operculum and adjoining anterior insula. Spontaneous activity for 364 gustatory neurons was 3.9 24.9 (mean * SD) spikes/s. Mean net (gross minus spontaneous) discharge rates for all gustatory neurons were: 1.0 M glucose=4.9? 11.6, 0.3 M NaCl= 3.227.1, 0.001 M quinine HCI=2.6?5.8, and 0.01 M HCI= 1.7 +4.6. The results from intensity-response functions imply that the perception of each basic taste quality in the nonhuman primate is based on the activity of the appropriate neural subgroup rather than on the mean activity of all taste cells. Therefore a more meaningful index of the effectiveness of a stimulus may be the discharge rate it evokes from the subset of gustatory neurons for which it is the best stimulus. Glucose was the best stimulus for 142 cells (including ties), from which it elicited a mean net response of 10.3 spikes/s; NaCl was best for 107 neurons which gave a mean 8.7 spikes/s; quinine HCl evoked 6.2 spikes/s from the 74 cells that responded best to it; HCl elicited 5.9 spikes/s from the 49 neurons for which it served as best stimulus. The response characteristics of cortical taste cells indicate heterogeneous features, and significantly different patterns from those reported in other nonchemical sensory systems. Taste Cerebral cortex Nervous system Monkey Electrophysiology ANATOMICAL, neurophysiological and clinical evidence indi- cates that the primary gustatory area in the cerebral cortex of the monkey is located in the frontal operculum and adjoining anterior insula (1, 3, 7, 10). In this communication we report on the characteristics of taste-responsive cells in the gustatory cortex of the alert cyno- molgus monkey. METHOD The subjects were seven male cynomolgus monkeys (Macaca fascicularis) weighing 3.8-7.0 kg during the course of data col- lection. The complete surgical procedure and additional methodology have been reported elsewhere (7). Briefly, each monkey was se- dated with ketamine hydrochloride (10 mg/kg, IM) and anesthe- tized to a surgical level with sodium pentobarbital (3040 mg/ kg, IM). Under sterile conditions, a stainless steel ring, to which a microdrive could be fitted during recording sessions, was ste- reotaxically implanted near the midline of the skull at the an- teroposterior level of the insular-opercular cortex. Antibiotics and topical antiseptics were used for several days for postopera- tive prophylaxis. After recovery of about 10 days, recording sessions were initiated. The alert monkey normally adopted a relaxed sitting position and his comfort was continuously at- tended to. Glass-insulated tungsten electrodes with tip sizes of about 2 x 5 pm were used to isolate individual cells from the insular-opercular cortex. The sterile electrode was advanced with a hydraulic microdrive and a chronic adaptor system. Conven- tional electrophysiological recording techniques were used. A variety of taste stimuli were used to identify gustatory neurons. In this report we summarize the findings with the four prototypical stimuli: 1 .O M glucose, 0.3 M NaCl, 0.001 M qui- nine HCl, and 0.01 M HCl. These concentrations were based on previous electrophysiological results in the primary gustatory cortex (10) and on human psychophysical studies. Stimuli were delivered through a hand-held 3 .O-ml syringe in quantities of 1 .O ml. Each stimulus was followed by a 2.0 to 5.04 deionized water rinse and a rest period of 15-180 s. An IBM-PC/AT com- puter was used to count action potentials for 3 s before and 5 s after stimulus application. Stimulus onset time was marked with precision through a circuit that was completed as the delivered stimulus contacted the tongue. The position of each recording site was determined by X-ray photography after each track and by histological techniques after the experiment (7). RESULTS Neurons responsive to chemical stimulation of the tongue ‘Requests for reprints should be addressed to T. R. Scott, Ph.D., Department of Psychology, University of Delaware, Newark, DE 19716. 333

Upload: thomas-r

Post on 25-Dec-2016

214 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Taste neurons in the cortex of the alert cynomolgus monkey

Brain Research Bulletin, Vol. 28, pp. 333-336. 0 Pergamon Press pk. 1992. Printed in the U.S.A 036179230/92 $5.00 + .OO

RAPID COMMUNICATION

Taste Neurons in the Cortex of the Alert Cynomolgus Monkey

CARLOS R. PLATA-SALAMAN AND THOMAS R. SCOTT*’

School of Life and Health Sciences and *Department of Psychology, University of Delaware, Newark, DE 19716

Received 21 October 1991

PLATA-SALAMAN, C. R. AND T. R. SCOTT. Taste neurons in the cortex of the alert cynomolgus monkey. BRAIN RBS BULL 28(Z) 333-336, 1992.-The activity of single neurons in the gustatory cortex of alert cynomolgus monkeys was analyzed. Taste-evoked activity in response to the four prototypical taste stimuli was recorded from a cortical gustatory area comprising the frontal operculum and adjoining anterior insula. Spontaneous activity for 364 gustatory neurons was 3.9 24.9 (mean * SD) spikes/s. Mean net (gross minus spontaneous) discharge rates for all gustatory neurons were: 1.0 M glucose=4.9? 11.6, 0.3 M NaCl= 3.227.1, 0.001 M quinine HCI=2.6?5.8, and 0.01 M HCI= 1.7 +4.6. The results from intensity-response functions imply that the perception of each basic taste quality in the nonhuman primate is based on the activity of the appropriate neural subgroup rather than on the mean activity of all taste cells. Therefore a more meaningful index of the effectiveness of a stimulus may be the discharge rate it evokes from the subset of gustatory neurons for which it is the best stimulus. Glucose was the best stimulus for 142 cells (including ties), from which it elicited a mean net response of 10.3 spikes/s; NaCl was best for 107 neurons which gave a mean 8.7 spikes/s; quinine HCl evoked 6.2 spikes/s from the 74 cells that responded best to it; HCl elicited 5.9 spikes/s from the 49 neurons for which it served as best stimulus. The response characteristics of cortical taste cells indicate heterogeneous features, and significantly different patterns from those reported in other nonchemical sensory systems.

Taste Cerebral cortex Nervous system Monkey Electrophysiology

ANATOMICAL, neurophysiological and clinical evidence indi- cates that the primary gustatory area in the cerebral cortex of the monkey is located in the frontal operculum and adjoining anterior insula (1, 3, 7, 10).

In this communication we report on the characteristics of taste-responsive cells in the gustatory cortex of the alert cyno- molgus monkey.

METHOD

The subjects were seven male cynomolgus monkeys (Macaca fascicularis) weighing 3.8-7.0 kg during the course of data col- lection.

The complete surgical procedure and additional methodology have been reported elsewhere (7). Briefly, each monkey was se- dated with ketamine hydrochloride (10 mg/kg, IM) and anesthe- tized to a surgical level with sodium pentobarbital (3040 mg/ kg, IM). Under sterile conditions, a stainless steel ring, to which a microdrive could be fitted during recording sessions, was ste- reotaxically implanted near the midline of the skull at the an- teroposterior level of the insular-opercular cortex. Antibiotics and topical antiseptics were used for several days for postopera- tive prophylaxis. After recovery of about 10 days, recording sessions were initiated. The alert monkey normally adopted a relaxed sitting position and his comfort was continuously at-

tended to. Glass-insulated tungsten electrodes with tip sizes of about 2 x 5 pm were used to isolate individual cells from the insular-opercular cortex. The sterile electrode was advanced with a hydraulic microdrive and a chronic adaptor system. Conven- tional electrophysiological recording techniques were used.

A variety of taste stimuli were used to identify gustatory neurons. In this report we summarize the findings with the four prototypical stimuli: 1 .O M glucose, 0.3 M NaCl, 0.001 M qui- nine HCl, and 0.01 M HCl. These concentrations were based on previous electrophysiological results in the primary gustatory cortex (10) and on human psychophysical studies. Stimuli were delivered through a hand-held 3 .O-ml syringe in quantities of 1 .O ml. Each stimulus was followed by a 2.0 to 5.04 deionized water rinse and a rest period of 15-180 s. An IBM-PC/AT com- puter was used to count action potentials for 3 s before and 5 s after stimulus application. Stimulus onset time was marked with precision through a circuit that was completed as the delivered stimulus contacted the tongue.

The position of each recording site was determined by X-ray photography after each track and by histological techniques after the experiment (7).

RESULTS

Neurons responsive to chemical stimulation of the tongue

‘Requests for reprints should be addressed to T. R. Scott, Ph.D., Department of Psychology, University of Delaware, Newark, DE 19716.

333

Page 2: Taste neurons in the cortex of the alert cynomolgus monkey

334 PLATA-SALAMAN AND SCOTT

TABLE I

CHARACTERISTICS OF TASTE-RESPONSIVE CELLS IN THE MONKEY GUSTATORY CORTEX

S.A. BOT

1 .O M Glucose

E.A. for all neurons (n=364): 4.9 i 11.6 Excitation (n=2.50, 68.7%): 7.8 t 12.8

Inhibition (n=96, 26.4%): -2.0 -+ 2.2

No response (n= 18, 4.9%): 0.0

E.A. for G.B. neurons (N= 142 or 38.2%): 10.3 + 16. I Excitation(n=129or90.8%): 11.7 2 16.2 (range=0.6 to 115.1)

Inhibition (n= 13 or 9.2%): - 3.5 t 3.0 (range= -0.1 to -9.0)

0.3 M NaCl

3.9 + 4.1t 1.6 't 3.8

0.67 I 0.23

0.67 r 0.23 0.71 -t 0.23

E.A. for all neurons (n=364): 3.2 i- 7.1

Excitation (n=244, 67.0 %): 5.7 ? 7.3

Inhibition (n=98, 26.9%): -- 2.2 2 I .9

No response (n = 22, 6.0%): 0.0

E.A. for N.B. neurons (N= 107 or 28.8%): 8.7 i 9.6

Excitation (n= 100 or 93.5%): 9.6 + 9.3 (range=0.4 to 46.0)

Inhibition (n=7 or 6.5%): --3.S ? I .9 (range= - 1.4 to -6.2)

0.001 M QHCl

3.1 k 3.61 2.4 k 1.9

0.66 i 0.24

0.65 ~fl 0.25 0.70 f 0.21

E.A. for all neurons (n=364): 2.6 t S.X

Excitation (n=233, 64.0%): 4.8 + 6.2

Inhibition (n = 102, 28.0%): - 1.6 & I .8 No response (n = 29, 8.0%): 0.0

E.A. for Q.B. neurons (N=74 or 19.9%): 6.2 -t 8.0

Excitation (n=63 or 85.1%): 8.0 2 7.2 (range=0.4 to 37.4)

Inhibition (n=ll or 14.9%): -4.0 + 3.5 (range= -0.1 to -- 12.0)

0.78 f 0.17 3.4 f 3.63: 0.78 i- 0.14

4.1 f 3.5 0.76 2 0.28

0.01 M HCI

E.A. for all neurons (n =364): I .7 2 4.6

Excitation (n=214, 58.8%): 3.8 + 4.9

Inhibition (n= 116. 31.9%): - I.6 ? 1.7

No response (n=34. 9.3%): 0.0

E.A. for H.B. neurons (N=49 or 13.2%): 5.9 i 6.6 0.79 2 0.14

Excitation (n=45 or 91.8%): 6.8 ? 6.1 (range=0.6 to 27.1) 4.1 I4.1 0.78 & 0.14

Inhibition (n=4 or 8.2%): --4.0 I 3.2 (range= -0.7 to 7.9) 3.0 i 4.7 0.87 t 0.12

Data: spikes/s (mean 2 SD). Abbreviations: BOT, breadth-of-tuning; E.A.. evoked activity; G.B., glucose best-responsive neurons; N.B., NaCl best-responsive neurons; H.B., HCI best-responsive neurons; Q.B., quinine HCI best-responsive neurons: S.A., spontaneous activity.

*B.A. for 362 neurons (excludmg two cells with over 40 spikes/s) was 3.7 ~4.0. j-Excludes one cell with >40 spikes/s. fEx- eludes one cell with >20 spikes/s.

Spontaneous activity for all neurons (n= 364): 3.9 -t 4.9 (range =O.O to 44.7)*. Breadth-of-tuning (entropy) coefficient (n= 364): 0.70 + 0.22 (range =O.O to I .O)

were found in a cortical area whose rostra1 boundary corre- sponded to the orbitofrontal cortex/frontal operculum junction. The area extended about 4 mm posteriorly to the frontal opercu- lum and adjoining anterior insula, about 7 mm in the dorsoven- tral plane, and about 3.0 mm mediolaterally

The mean spontaneous rate of taste-responsive neurons in this report was 3.9 ? 4.9 spikes/s (n = 364). Taste-responsive cells were characterized by small action potentials (usually 200- 500 p.V).

The criterion we selected for an evoked response was sponta- neous rate k2.33 SD QKO.01) maintained for a 5-s period af- ter stimulus application. Evoked responses to application of the four prototypical stimuli are shown in Table 1. The mean evoked responses to each of the prototypical stimuli across all neurons were: glucose = 4.9 spikes/s, NaCl= 3.2, quinine HCl= 2.6, and HCl = 1.7. When considering only the mean evoked activity

corresponding to the neural subgroup most responsive to each prototypical stimulus, discharge rates were: glucose = 10.3 spikes/s (n= 142), NaCl=8.7 (n= 107), quinine HCl=6.2 (n=74), and HCl=5.9 (n=49). The stability of our recordings was con- firmed by the test-retest reliability coefficient of >0.90 obtained by comparing response profiles generated across neurons by the first versus the second application of a stimulus while recording from a given cell.

The breadth of a neuron’s response across the four prototypi- cal stimuli was quantified by calculating the cell’s coefficient of entropy, a value that may range from 0.0 (total specificity to one chemical) to 1.0 (equal responsiveness to all four chemi- cals) (11). The mean coefficient for the 364 neurons in the monkey cortex was 0.70 (Table 1). This indicates greater speci- ficity than the mean value of 0.87 derived from the monkey nu- cleus tractus solitarius (9), but greater breadth than the mean

Page 3: Taste neurons in the cortex of the alert cynomolgus monkey

4STE IN MONKEY CORTEX

Fruit juice

=IlIm .&_-.

0.01 M HCI

1.0 M Glucose

0.001 M Quinine

-II I 0.001 M NaCl

I _+___ II .- -------. t I -II

0.03 M NaCl

0.3 M NaCl

I

0.1 M NaCl

1.0 M NaCl

FIG. 1. Responses of a representative neuron in primary taste cortex to fruit juice (probe stimulus for identifying gustatory neurons), prototypi- cal stimuli, and to an intensity series of NaCI. The cell responded with a fixed latency of 420 ms. Stimulus onset is marked on the bottom trace. Full sweep=2 s. From Scott et al. (7) with permission.

value of 0.39 obtained from orbitofrontal cortex neurons (6). A comparison among the breadth-of-tuning coefficients correspond- ing to the neural subgroup most responsive to a particular stimu- lus demonstrated that cells tuned to glucose or NaCl were significantly @<O.Ol) more narrowly tuned than those most re- sponsive to quinine HCl or HCl. As shown in Table 1, there was no clear evidence that a neural subgroup’s breadth of sensi- tivity was a function of evoked activity levels. The responses of a single gustatory neuron to an array of taste qualities and to a concentration series of NaCl are shown in Fig. 1.

335

Plots of the location of over 200 neurons as a function of its best stimulus revealed that there was no clear indication of chemotopic organization in the gustatory cortex (not shown).

DISCUSSION

These results show that taste-responsive cells are character- ized by slow spontaneous activity with over 45% of the cells displaying a mean of O-2 spikes/s. Derived data from mean counts across all taste cells also show low evoked activity due to the many cells that were unresponsive to a particular proto- typical stimulus (Table 1). Categorization of each neuron into four subsets according to the stimulus that evoked its greatest response resulted in a considerable higher evoked activity. In- tensity-response functions determined across 62 taste-responsive cells showed that mean discharge rate was a direct function of stimulus concentration for glucose ( 10P3-1 .O M), NaCl (lo- 3- 1 .O M), and quinine HCl (10-5-3 X lop3 M) (7). The slopes of these functions, when calculated from the activity of neurons that responded best to each stimulus, matched the slopes of in- tensity-response functions from human psychophysical experi- ments. The implication of these results is that taste perception of the four prototypical stimuli is conveyed by the activity in subgroups of neurons most sensitive to a particular quality rather than by the responses of all taste cells.

The levels of spontaneous and evoked activity of cortical (frontal operculum and anterior insula) taste cells are in contrast with activity levels in other sensory systems. Discharge rates are low in taste cells-even when considering only the neural sub- groups most responsive to a particular prototypical taste. Neu- rons in the primary visual cortex of the alert monkey show a mean spontaneous rate of about 7 spikes/s, and their mean evoked activity increases to about 100 spikes/s in response to an optimal stimulus (12). Similarly, neurons in the auditory cortex of the alert monkey generate over 100 spikes/s in response to tone stimulation (2). The limited range of discharge rates over which prototypical taste stimuli are coded in the monkey cortex may indicate that: 1) the optimal gustatory stimulus has yet to be employed, although this possibility is unlikely since scores of other tastants have been used with similar results, 2) gustatory neurons possess intrinsic properties that preclude high response rates, although there are no ultrastructural features to suggest that separate cortical areas exhibit different metabolic activity (5), or 3) gustatory neurons may have multiple excitatory and inhibitory inputs, and only a subset of the excitatory inputs may be activated during chemical stimulation of the tongue. This no- tion is supported by the presence of a variety of nongustatory cells within the primary taste cortex that modify their discharge in response to mouth movements, tactile stimulation of the oral cavity, approach or anticipation of stimulus application, or tongue extension (7). The slow but extended responses of the taste sys- tem may also serve the needs of the gustatory system, which responds not only to momentary stimulation of taste receptors, but also to the slowly changing state of the viscera (8).

Taste-responsive cells showed greatest sensitivity to glucose among the four prototypical stimuli (Table l), followed by NaCl, quinine HCl, and finally HCI. This is relevant to the dis- criminative capacity of humans for whom the perceptions of sweetness and saltiness are nearly always unequivocal while those of sourness and bitterness are often confused (4).

ACKNOWLEDGEMENTS

This work was supported by grants from the National Science Foun- dation and from the Campbell Institute for Research and Technology. Ms. Virginia L. Smith-Swintosky was involved in several aspects of these studies.

Page 4: Taste neurons in the cortex of the alert cynomolgus monkey

336 PLATA-SALAh%hLN AND SCOTT

REFERENCES

I.

2.

3.

4.

5.

6.

Benjamin, R. M.; Burton, H. Projection of taste nerve afferents to anterior opercular-insular cortex in squirrel monkey (Suimiri sci- ureus). Brain Res. 7:221-231; 1968. Brugge, J. F.; Merzenich, M. M. Responses of neurons in auditory cortex of the macaque monkey to monaural and binaural stimula- tion J. Neurophysiol. 36:1138-1158; 1973. Bomstein, W. S. Cortical representation of taste in man and mon- key. II. The localization of the cortical taste area in man, a method of measuring impairment of taste in man. Yale J. Biol. Med. 13: 133-156; 1940. Meiselman, H. L.; Dzendolet, E. Variability in gustatory quality identification. Percept. Psychophys. 2:496-498; 1967. Peters, A.; Kaiser-Abramof, I. R. The small pyramidal neuron of the rat cerebral cortex: the synapses upon dendritic spines. Zell- forsch. Mikrosk. Anat. 100:487-506; 1969. Rolls, E. T. Information processing in the taste system of primates. J. Exp. Biol. 146:141-164; 1989.

7.

8.

9.

10.

11.

12.

Scott, T. R.; Plata-Salaman, C. R.; Smith, V. L.; Giza, B. K. Gustatory neural coding in the monkey cortex: stimulus intensity. J. Neurophysiol. 657686; 1991, Scott, T. R.; Yaxley, S. Interaction of taste and ingestion. In: Ca- gan, R., ed. Neural mechanisms in taste. Boca Raton: CRC Press; 1989:147-177. Scott, T. R.; Yaxley, S.; Sienkiewicz, Z. J.; Rolls, E. T. Gustatory responses in the nucleus tractus solitarius of the alert cynomolgus monkey. J. Neurophysiol. 55:182-200; 1986. Scott, T. R.; Yaxley, S.; Sienkiewicz, Z. J.; Rolls, E. T. Gustatory responses in the frontal opercular cortex of the alert cynomolgus monkey. J. Neurophysiol. 56:876-890; 1986. Smith, D. V.; Travers, J. B. A metric for the breadth of tuning of gustatory neurons. Chem. Sens. 4:215-229; 1979. Trotter, Y.; Thorpe, S. J.; Celebrini, S.; Pouget, A.; Imbert, M. Processing of orientation in Vl of the awake monkey. Sot. Neuro- sci. Abstr. 15:1056; 1989.