THE JOURNAL OF COMPARATIVE NEUROLOGY 297~106-120(1990)

Cholinergic Manipulation Alters Stimulus-EvokedMetabolic Activity in Cat Somatosensory Cortex SHARON L. JULIANO, WU MA, MARK F. BEAR, AND DON ESLIN Department of Anatomy, USUHS, Bethesda, Maryland 20814 (S.L.J., W.M., D.E.); Center for Neural Science, Brown University, Providence, Rhode Island 02912 (M.F.B.)

ABSTRACT The role of acetylcholine (ACh) in cerebral cortical activity has recently been reevaluated. It now seems clear that this neurotransmitter increases the magnitude of cortical responses. Although substantial information has been gathered regarding the role of ACh in sensory information processing, little is known about the participation of ACh in the organization of maps in the cerebral cortex. T o address this issue, we used 2 methods to manipulate the supply of ACh in the somatosensory cortex of cats: 1) unilateral neurotoxic lesions of the basal forebrain and 2) unilateral topical applications of the cholinergic antagonist, atropine. For each experimental condition, the animal received an injection of 2-deoxyglucose (2DG) while identical somatic stimuli were delivered to the right and left forepaws. In the somatosensory cortex, the 2DG uptake most often occurred in the form of patches that extended from layer I1 to IV. When the patches were reconstructed into %dimensional maps of activity throughout the somatosensory cortex, they formed strips that ran in the rostrocaudal direction. The reconstructed maps revealed that the 2DG patterns in ACh-depleted and the normal cortex were similar in their overall topographic distribution. Depletion or antagonism of ACh, however, caused the stimulus-evoked metabolic label to be reduced in dimension and density. Measurements of background activity levels were obtained by using 1) cytochrome oxidase histochemistry or 2) metabolic activity values in regions of somatosensory cortex that were not specifically stimulated. This analysis indicated that background values in the ACh-depleted hemispheres were not different from those in the normal hemispheres. The absence of ACh therefore appears to reduce the cortical response to stimulation, while background activity values do not change. These observations indicate that ACh plays a significant role in the processing of sensory information and the organization of somatosensory cortical maps. Key words: acetylcholine, basal forebrain, 2-deoxyglucose, cytochrome oxidase, neuronal plasticity, cortical column

Attention has recently focused on the role of acetylcholine (ACh) in cortical processing. ACh has been shown to act as a neuromodulator in the neocortex by facilitating neural activity under specific circumstances. In the visual and somatosensory cortex, ACh appears to enhance neuronal responsivity, especially when paired with appropriate somatic or visual stimulation, without affecting background activity (Donoghue and Carrol, '87; Sato et al., '87a; Sillito and Murphy, '87; Lamour et al., '88; Metherate et al., '88a; Ma et al., '89). In some instances, application of ACh increases neural activity on a long-term basis, outlasting the presence of the drug (Metherate et al., '87; Lamour et al., '88; Metherate et al., '88b; Rasmusson and Dykes, '88). The mechanism of' long-lasting neural enhancement, is not clear, but ACh-mediated excitation appears to occur through a decrease in potassium permeability, leading to a tendency

o 1 9 9 0 WILEY-LISS, INC.

for the cell to fire repetitively (Krnjevic et al., '71; Halliwell and Adams, '82; Brown, '83; McCormick and Prince, '85). ACh has also been identified as a potentially important factor in neuronal plasticity. For example, Bear and Singer ('86) found that depletion of cortical ACh, together with norepinephrine depletion, would interfere with ocular dominance plasticity in the development of kitten area 17. The ability of ACh to produce long-lasting changes in neuronal activity may be further evidence for its participation in "activity-dependent" plasticity in the neocortex. The idea of ACh involvement in neuronal plasticity has led to the suggestion that it may also be implicated in the reorganizaAccepted February 14,1990. Address reprint requests to S.L. Juliano, Department of Anatomy, USIJHS, 4301 Jones Bridge Road, Bethesda, MD 20814.

2DG A N D ACH I N CAT SOMATOSENSORY CORTEX tion of somatotopic cortical maps that occurs following deafferentation or amputation (Metherate et al., '87; Lamour and Dykes, '88). Investigations of neuronal responses after cholinergic manipulation have provided substantial detail regarding the mechanism of cholinergic action and the properties of neurons after ACh application. Little is known, however, about the effect of ACh on cortical maps in either normal or deafferented animals. In this paper we evaluate the role of ACh in the formation of maps in the somatosensory cortex of normal cats. We achieved this by using 2 methods to effectively deplete the somatosensory cortex of ACh: unilateral lesions of the basal forebrain and topical applications of the muscarinic cholinergic receptor antagonist, atropine. Following these manipulations the cats received bilateral somatic stimulation during a 2-deoxyglucose (ZDG) experiment. This procedure allowed us to derive maps of metabolic activity in the normal hemispheres and compare them with those in the ACh depleted hemispheres. A preliminary report of these studies appeared earlier (Juliano e t al., '88).

MATERIALS AND METHODS Experimental design Twelve cats of either sex, weighing 2-3.5 kg were used in this study. The cats were divided into 2 groups: 1) those receiving basal forebrain lesions and 2) those receiving topical applications of atropine. All procedures were performed unilaterally, with the normal, unmanipulated hemisphere acting as an internal control. For the first group, 3 cats received lesions 1 week prior to a 2DG experiment in which they received bilateral tactile stimuli to the forepaw. One of this group received a lesion that did not deplete the somatosensory cortex of ACh. This experiment served as a control to determine whether or not the surgical procedure and/or lesion itself had an effect on the metabolic pattern. The second group of 9 cats received topical applications of atropine. Of this group, 6 received different concentrations of atropine followed by a 2DG experiment. During the 2DG study, the animals received bilateral somatic stimulation. One cat was not stimulated following the 2DG injection to ascertain the effect of topical atropine alone on metabolic uptake. Previous studies have demonstrated that the procedure of opening the skull and applying a control solution t o the brain (e.g., artificial CSF) does not alter the stimulusevoked activity pattern (Juliano et al., '89).

SwwY Basal forebrain lesion. The lesions were carried out under aseptic conditions and halothane anesthesia 1 week prior to 2DG injections. Two small holes were drilled in the skull over the appropriate coordinates. Injections of Nmethyl-D,L-aspartate (5 % dissolved in Na phosphate buffer a t pH 7.0) were made a t two sites. The coordinates for the first injection (5 pl) were 15.0 mm A, 4.0 mm L, and 7.0 mm D; the coordinates for the second injection (3 pl) were 16 mm A, 7 mm L, and 8.0 mm D (Barstad and Bear, '87, '89; Berman and Jones, '82). Gelfoam was placed in openings in the skull, the skin sutured closed, and the animal allowed to recover. During the recovery period, the animal was monitored carefully. The cats were often found to exhibit turning toward the side of the lesion; this activity disappeared a few hours after the surgery. Other than this, there were no postsurgical complications.

107

Topical applications. The topical application procedure is similar to that reported previously by this laboratory for application of pharmacologic agents (Juliano et al., '89). Briefly, under halothane anesthesia, the animal was intubated with a tracheal cannula and a venous catheter inserted into the long saphenous vein. An opening 10 mm in diameter was made over the somatosensory cortex. A chamber was constructed of dental cement around the opening and the dura removed; artificial CSF was placed in the chamber. A long-acting topical anesthetic ointment was infused in the wounds. At this point, gallamine triethiodide was delivered to achieve neuromuscular blockade. The animal was placed on a respirator and expired CO, and body temperature were monitored and maintained within normal limits. The halothane was removed and replaced with nitrous oxide (70%) and oxygen (30%), in preparation for the 2DG experiment (see below). Preparation of atropine. The atropine (atropine sulfate) was dissolved in CSF to achieve concentrations of 5 to 100 pM; the solutions were placed in the chamber approximately 15 minutes prior to a 2DG injection. The atropine was changed once during the 2DG study. For one animal, 3H atropine (New England Nuclear, N-meth~l-~H-atropine, 87 pCi/mmol) was diluted to a concentration of 50 pM and placed in the chamber for the same period of time that the atropine contacted the brain during a 2DG experiment (about 1hour). This concentration of atropine was shown to be effective in altering the 2DG pattern in earlier experiments. Electrophysiological recordings. For 2 animals multi- and single-unit recordings were conducted to determine the effect of atropine topical applications on neural activity. During this procedure the animal was maintained on 1-1.5% halothane. Pipette electrodes filled with 3 M KC1 and having tip diameters of 5-10 pm were used. Signals from the electrodes were amplified and displayed on an oscilloscope and audio monitor. Receptive fields (RFs) were initially determined by using hand-held probes and brushes. Before and during application of atropine, RFs were carefully mapped by using von Frey hairs. The atropine was infused into the chamber and the neural site assessed for receptive field changes. During this procedure, the investigator did not know the nature of the fluid being infused into the chamber, i.e., a specific concentration of atropine or CSF. 2-Deoxyglucose experiments and histology. The protocol for this portion of the experiments is similar to that described originally by Sokoloff et al. ('77) and published previously by this laboratory (Juliano and Whitsel, '87; Juliano et al., '89). The anesthetic conditions are as described in "Topical applications" above. A somatic stimulus was initiated 5-10 minutes prior to the injection of 2DG; all animals received bilaterally symmetric stimuli to both forepaws. In general, one hemisphere received atropine treatment or an ipsilateral lesion to the basal forebrain, and the other hemisphere served as an internal normal control. Two types of stimulation were used: intermittent vertical displacement (i.e., flutter) and electrocutaneous (previously described in Juliano et al., '89). The flutter stimuli were delivered from a rounded plexiglass probe 0.5 cm in diameter. The probes were attached to a Ling stimulator driven by a wave form generator and power supply. The electrocutaneous stimuli were delivered with ECG electrodes and consisted of repetitive trains of pulses delivered from a Grass

S.L. JULIANO ET AL.

108 stimulator that were 0.9 mA in amplitude, 20 msec in duration, at a rate of 20 per second. The stimuli continued for 45 minutes after the injection of 2-deo~y-D-[l-'~C]glucose (10 pCi/lOO g) a t which time an overdose of pentobarbital Na (35 mg/kg) was administered. The animal was then perfused transcardially with 4% buffered paraformaldehyde, the brain removed, blocked, and quickly frozen in Freon 22. Each brain was stored at -7OoC, and cut at a later date in a cryostat with a temperature of - 16°C at a thickness of 30 pm. The brains were cut in a coronal plane with both hemispheres on a single section, to allow for direct comparison between normal and manipulated sides. Alternate sections were saved for acetylcholinesterase (AChE) and cytochrome oxidase (CO) histochemistry and 2DG autoradiography. Staining for AChE activity used a modification (Jacobowitz and Creed, '83) of the procedure originally described by Koelle ('55).Staining for CO activity used the method described by Wong-Riley ('79). The sections saved for 2DG autoradiography were picked up on #2 coverslips and quickly transferred to a warming tray (60°C) for rapid dehydration. The coverslips were exposed to SB-5 X-ray film (Kodak) along with 14Cstandards (Microscales, Amersham) for 7-10 days. The brain that received 3H-atropine was cut in a similar manner and the sections exposed to LKB Ultrafilm for 10 days. After the development of autoradiographs, selected sections were stained with thionin for identification of cortical cytoarchitecture. Computer analysis. The autoradiographs were analyzed by using a P D P 11/23+ computer interfaced with a video based imaging system (Tommerdahl e t al., '85). The digitized signals were used to determine optical density measurements that were converted to 14C values. We also used the imaging system to generate maps of activity that allowed visualization of the spatial distribution of 14Cdeoxyglucose values across the somatosensory cortex. The reconstruction process involved entering the outer and inner boundaries of selected regions on individual serial sections of cortex along with specific points of reference (e.g., morphological landmarks, cytoarchitectural boundaries). The boundaries corresponded to the interface between layer VI and the white matter (inner boundary) and the location of layer I (outer boundary). The average 14C value of label in the tangential dimension was determined for each autoradiograph by obtaining the average optical density value across layers I1 through V. The bin size for each tangential value is variable, but measures approximately 50 pm. The resulting tangential histogram from each section is displayed as a vertical line of pixels on the video monitor; the bin height, which reflects the 14Cconcentration, is represented by pixel intensity. The lines are then stacked horizontally and aligned with respect to a selected point of reference.

RESULTS Basal forebrain lesion experiments Lesion identification. Three animals received lesions of the basal forebrain. In 2 of these animals the lesion led to depletion of ACh, as demonstrated by AChE histochemistry (Fig, 1). An example of the site of an effective lesion can be seen in Figure 2. It includes regions of the ventrolateral globus pallidus, the substantia innominata, and the diagonal band of Broca; these basal forebrain structures have been identified as the primary source of neurons that project to the somatosensory cortex of cats (Barstad and Bear, '87, '89). The lesions were identified using CO histochemistry,

because with this stain, the lesion site was observed as a clearly distinguished pale area (Fig. 3); this location was verified in the Nissl stain. The extent of the cortical ACh depletion was quantified by counting the AChE positive fibers in the normal and depleted hemispheres. The fibers were counted in layers 11-IV (i.e., the region that contained patches of evoked 2DG uptake); 10 sections from each hemisphere of the somatosensory cortex were included. A grid of 100 pm2 was positioned over 3 sites in a vertical array within layers 11-IV and the number of fibers tallied. These calculations revealed that uf the 2 lesions that successfully depleted the somatosensory cortex of ACh, one led to an 81% reduction of AChE positive fibers and the second led to a 73'h reduction. In all of these experiments, the lesions were well delineated and did not extend posterior to anterior portions of the globus pallidus, or into the thalamus. Nevertheless, in order to rule out potential retrograde degeneration in the thalamus due to trauma from the surgery or insertion of the injection cannula, we examined the ventrobasal thalami of each animal. Nissl stains of this region indicate that the thalami on the side of the lesion appear similar to those in the opposite normal side. The distribution, size, and density of neurons in the ventrobasal complex were the same in both hemispheres. 2-Deoxyglucose label. The 2DG uptake evoked by somatosensory stimulation was primarily patch-like, as in previous studies (Juliano and Whitsel, '87; Juliano et al., '89). In the normal hemisphere, the patches ranged from 550 to 1,000 pm in tangential dimension. A patch corresponded to metabolic activity that was a t least 55% above white matter values, and the surrounding cortical regions were 20-4090 above white matter values. In most experiments, as in earlier studies, the patches extended from layers I to IV; the label occasionally reached into upper portions of layer V. Measurements of patches were made through layer IV, which usually corresponded to the widest portion of the patch. The patches found in the hemisphere ipsilateral to the basal forebrain lesion were reduced in dimension and intensity from those in the normal hemispheres (Figs. 4, 5). In these experiments, the right and left hemispheres were processed together on a single autoradiograph, thus it was possible to directly compare the control and manipulated hemispheres on a single section. The parameters for image analysis were held constant for measuring values in right and left hemispheres. Although reduced in intensity, the label in the depleted hemispheres occurred in the same layers as that in the normal side. The distribution of a patch of 2DG uptake through the cortical laminae can be seen in Figure 5. Two-dimensional maps of the metabolic uptake also indicate that the distribution of evoked activity on the side of the lesion is shrunken compared to that on the normal side. Figure 6 demonstrates a map of 2DG uptake distributed through the normal and lesioned hemispheres. On the normal side (on the left), the 2DG pattern occurs as a strip that extends for 3 mm. On the ACh depleted side, the strip is not as extensive (extending for 2 mm), nor is i t as dense. When the basal forebrain lesion was not effective in depleting the neocortex of ACh, there was no difference in the dimension or intensity of the label from the normal hemisphere. Figures 7 and 8 summarize the results of these experiments. Background activity. CO activity was evaluated in the somatosensory cortex of the animals with basal fore-

2DG AND ACH IN CAT SOMATOSENSORY CORTEX

109

I

II

111

IV

V

Vl

-'

Fig. 1. Acetylcholinesterase histochemistry through the right and left somatosensory cortex of an animal that sustained a unilateral lesion of the basal forebrain. On the left is the normal staining pattern; on the

i

right is the pattern observed after cholinergic depletion with a basal forebrain lesion. Cortical layers are indicated with Roman numerals. WM = white matter. Scale bar = 100 pm.

S.L. JULIANO ET AL.

110

A 18.5

A 17.5

A 16.5

A 15.5

A 14.5

Fig. 2. Line drawings indicating the location of a typical lesion in coronal sections. T h e lesion is shown with shading. T h e drawings are arranged with the most anterior section toward the top. Ca, caudate;

Put, putamen; GP, globus pallidus; AC, anterior commissure; SI, substantia innorninata; DBV, diagonal band of Broca, ventral limb; DBH, diagonal band of Broca, horizontal limb. Scale bar = 11.2 rnrn.

brain lesions. No differences were observed between normal and ACh-depleted hemispheres. The 2 hemispheres appeared qualitatively the same, with the CO patt.ern staining heavily in layers 11-IV, slightly reduced in layer I. The staining was weakest in layer V and became slightly increased in layer VI (Fig. 9). Measurements of 2DG uptake in regions of the somatosensory cortex that were depleted of ACh but not activated by the somatic stimulus (e.g., foot region) also indicated that the background cortical activity values were not different in the normal versus lesioned hemispheres (Fig. 8).

Experiments involving topical applications of atropine Nine cats received topical applications of atropine. Control experiments. One cat received a topical application of 'H atropine to observe the extent that the atropine was absorbed into the brain. The 'H atropine was placed on the brain in the same manner, for the same length of time as in the 2DG experiments. Figure 10 shows that a 50 KM concentration of atropine is absorbed through layer V. Optical density values taken through the distribution of

2DG A N D ACH I N CAT SOMATOSENSORY CORTEX

111

Fig. 3. Cytochrome oxidase histochemistry taken through a coronal section demonstrating a lesion of the basal forebrain. The lesion is unilateral and indicated with arrows. The cell poor area appears pale

with this stain. CA, caudate; IC, internal capsule; Put, putamen; GP, globus pallidus; SI, substantia innominata; dbh, diagonal band of broca, horizontal limb. Scale bar = 2.5 mm.

%-atropine on the autoradiograph indicate that 50% of the total amount of atropine is present in lower layer V (Fig. 10); therefore the concentration of atropine was reduced by a factor of 0.5 in its layer V location. Although the atropine is not present through the full thickness of the cortex, the 2DG uptake in the normal hemispheres concentrates in the cortical layers where the atropine was present, i.e., the central and upper layers. Another cat received a topical application of 100 pM atropine but no somatic stimuli were delivered during the 2I)G study. The intent of this experiment was to determine

if atropine itself had any effect on metabolic uptake, in the absence of stimulation. No unusual alterations in 2DG uptake were observed; maps of metabolic activity show no differences in the normal versus atropine treated hemisphere (Fig. 11). An analysis of RF properties was carried out on 2 cats. The intention of obtaining neurophysiological data was to determine if application of atropine to the cerebral cortex produced dramatic changes in neural properties. In these animals, neuronal responses were determined before and after application of various doses of atropine. RF and thres-

Fig. 4. Autoradiographs taken in coronal section through the somatossnsory cortex of an animal that received a basal forebrain lesion; the normal hemisphere is on the left. Arrows point to patches of stimulusevoked 2DG uptake. The metabolic activity on the lesioned side (right) is very faint compared with that found in the normal hemisphere. The

animal was stimulated bilaterally with a flutter stimulus to the ventral surface of digit 2. The autoradiographs from this animal are reconstructed and shown in Figure 6. The images were photographed directly from the autoradiographic film. Scale bar = 1.35 mm.

S.L. JULIANO ET AL.

112

80%

b LAYERS

LAYERS

Fig. 5. On the top are digitized autoradiographs taken through the somatosensory cortex of an animal that received a basal forebrain lesion. The normal hemisphere is on the left. Patches of label are boxed in. The regions boxed in are reconstructed on the bottom into 3-dimensional histograms demonstrating optical density values within a patch. The tangential (a)and vertical axes (b)are indicated with letters on the

autoradiograph and on the histogram. Cortical layers are shown on the bottom. Peak activity values are indicated on the histogram, and are normalized as a percent of white matter, set to 0 for purposes of the histogram reconstruction. The evoked activity on the lesioned side is reduced in size and density value. Scale bar = 500 pm.

hold of response was determined using von Frey hairs. Our recording procedures did not determine alterations in the magnitude of response, but concentrated on receptive field changes. Analysis of 10 neurons after application of either 10 pM, 50 pM, or 100 p M concentrations of atropine demonstrated that no obvious changes were observed. Although this was a crude measure of neuronal responsivity, and carried out under different anesthetic conditions than the 2DG experiments, our purpose was to observe if topical applications of atropine caused gross disturbances in the cortical response, which might be reflected in global metabolic activity levels. Effect of different concentrations of atropine. The concentrations of atropine were chosen as those known to be effective in causing conductance changes in neurons recorded intracellularly in slice preparations, but not a t a level known to cause an anesthetic cellular effect (Williamson and Sarvey, ’85; Boyer and Sarvey, personal communication). The lowest dose of 5 pM did not elicit a metabolic pattern that differed from the normal side. A dose of 10 pM atropine caused a slight reduction in the dimension of the stimulusevoked patches, but did not cause changes in the patch density (Fig. 7). Doses of 10,40,50, or 100 pM atropine all generated changes in the metabolic pattern. Doses of 40,50, and

100 pM caused alterations that were similar in quality and quantity, whereas the lowest dose of 5 p M produced no change in the pattern. The observation that a 5 pM dose did not affect the 2DG label reinforces the notion that the metabolic alterations found with higher doses are not due to the procedure of topical application per se. It also appears that a dose of 100 pM did not induce local anesthetic effects, at least as measured by 2DG, since for the animal that received no somatic stimulation, metabolic activity levels in the region contacted by the atropine were the same as those on the opposite normal side (Fig. 8). This was also true for 2DG uptake levels in regions of cortex contacted by atropine but not specifically activated by somatic stimulation (see below and Fig. 8). Effects of atropine on 2DG uptake. The application of atropine produced a pattern of label that was similar to the distribution found after the basal forebrain lesion. On single autoradiographs, the individual patches of 2DG ‘uptake were reduced in dimension and density (Figs. 12, 113). The distribution of the label through the cortical layers did not change. Therefore the decreased activity was present in the same layers as on the normal side, only reduced in dimension and intensity. Figure 13 represents the distri bution of a patch of label through the cortical layers.

2DG A N D ACH I N CAT SOMATOSENSORY CORTEX

113

Fig. 6. Two-dimensional, digitized reconstruction of metabolic activity evoked in the somatosensory cortex by bilateral stimulation to the ventral surface of digit 2. Individual sections were unfolded and aligned or( the fundus of the coronal sulcus (CS). See text for details about the reconstruction procedure. The scale is expanded in the anteroposterior direction to reveal details of the metabolic pattern. On the left is the pattern evoked in the normal hemisphere; on the right is the pattern

evoked in the hemisphere depleted of acetylcholine by a unilateral basal forebrain lesion. The darkest metabolic activity represented in these reconstructions (i.e., black) corresponds to values 70% or more above white matter values; the lightest grey activity corresponds to 45-50% above white matter values. Other hues of grey represent metabolic activity levels in between these values. Med, medial; Ant, anterior.

The 2-dimensional maps of evoked activity also indicate that the distributed pattern in the atropine treated side is not as extensive as that in the normal hemisphere. This can be seen in Figures 14 and 15, which demonstrate maps of 2DG uptake following application of 100 pM and 50 pM, respectively. Background activity. Background values of 2DG uptake were taken from regions of somatosensory cortex that were not selectively activated by the somatic stimulus. These measurements also indicate that application of the atropine itself did not cause changes in metabolic activity, since there were no appreciable differences in the treated versus normal hemispheres (Figs. 8 , l l ) .

application of atropine, this suggests that the metabolic pattern reductions are due to decreased action of ACh at muscarinic receptors. The supposition is that in the normal cortex, ACh facilitates evoked activity. It is interesting to note that the decreased metabolic activity after AChdepletion was demonstrated in nitrous-oxide anesthetized animals. Although we did not study the effects of anesthesia systematically, we would expect the consequences of AChdepletion to be even more dramatic in unanesthetized animals. Conversely, it is possible that deep barbiturate anesthesia would eliminate the interhemispheric differences after unilateral manipulation of cortical ACh. In this regard, the fact that our electrophysiological experiments were carried out under halothane anesthesia may help to explain our lack of effect on receptive field properties after topical applications of atropine. A number of earlier studies indicated that neural or metabolic responses were not diminished after basal forebrain lesions leading to cholinergic depletion (Lamour et al., '82; Lamarca and Fibiger, '84; London et al., '84). These experiments differed from ours in that they examined either spontaneous neural activity or background metabolic activity, in brains that were not specifically stimulated. Such results are thus consistent with the outcome of the present experiments, since we found that resting metabolic levels, or metabolic activity levels as visualized with CO, do not change with cholinergic depletion. I t appears that regions of cortex specifically responding to stimulation are the sites altered in the absence of ACh. This conclusion is supported by the work of Sat0 et al. ('87b) in visual cortex. These researchers studied the visual cortex of cats that received basal forebrain lesions and found that the responses of

DISCUSSION What does ACh do in the somatosensory cortex? These findings demonstrate that ACh is important for normal processing of somatosensory information a t the level of the cerebral cortex. Our study suggests that the reduction of ACh reduces metabolic responsivity. Cholinergic inactivity, either by basal forebrain lesion or local cortical pharmacologic antagonism reduces the dimension and intensity of evoked metabolic activity, while leaving the background activity unaffected. The observation that background activity levels appear unaltered, as seen in background 2DG measurements and in CO staining, indicates that the response to stimulation is selectively modified while global activity levels are not changed. Because the results found after basal forebrain lesions were duplicated by topical

S.L. JULIANO ET AL.

114

a

Patch Width Normal 8 treated hemispheres

1

1200

800

rn* I

40( 20( 0

5uM

lOuM

50uM lOOuM

40uM

Normal *patch

BF+

BF+

BF-

Treated

width in rnlcrons

b rn e t a b 0

I

Patch Density Normal 8 treated hemispheres 100

80

I C

a C

60

t

I V

I

40

t Y V

20

a I U

e 8

0 5uM

lOuM

40uM

50uM lOOuM

Normal Fig. 7. a: Bar graph indicating the widths of stimulus-evoked patches of metabolic activity elicited by somatic stimulation. T h e graph compares the dimension of patches in normal hemispheres (solid) with hemispheres on the opposite sides t h a t were treated (hatched) either with topical applications of atropine or a n ipsilateral basal forebrain lesion. The doses of atropine are indicated on the abscissa. BF+ refers to

BF+

BF+

BF-

Treated

lesions that successfully depleted the somatosensory cortex of ACh. BF-refers to a lesion t h a t did not deplete the somatosensory cortex of ACh. b: Graph of density values in stimulus-evoked patches for the same conditions a s in a. Values on the ordinate indicate metabolic activity levels t h a t are expressed a s percent above the density value of white matter.

2DG A N D ACH I N CAT SOMATOSENSORY CORTEX

m e t a b 0

I I

115

100

80 -

C

a

60 -

C

t I V

I

40

V

20

t Y

a I U

e 8

0

5uM

8, stimulation ns, no stimulation

lOuM 40uM 50uM 1 0 p l O C p Normal

BF+

BF+

BF-

Treated

Fig. 8. Bar graph demonstrating metabolic activity values for background levels in the right and left hemispheres of animals that were treated with topical applications of atropine or unilateral basal forebrain lesions. Metabolic activity levels are expressed as percent above white matter values. T h e background values were taken from regions of

the somatosensory cortex that were not specifically stimulated. “s” refers to a n animal that received a topical application of 100 pM atropine as well as somatic stimulation; “ns” refers to an animal that was treated with a topical application of atropine, hut received no stimulation during the 2DG experiment.

neurons in area 17 were sluggish and depressed in response to visual stimulation. Earlier studies indicate that in sensory cortical regions, ACh is generally facilitatory. Although there are a number of reports indicating inhibitory effects, it is likely that inhibitory responses in the cerebral cortex are due to excitation of inhibitory interneurons (McCormick and Prince, ’85).In visual cortex, Sillito and Kemp (’83) demon-

strated that iontophoretic application of ACh increases the neuronal response to specific stimuli without enhancing overall responsivity. In the somatosensory cortex, many cells studied by Metherate et al. (’88a) were facilitated by application of ACh when paired with the appropriate somatic stimulation, while few neurons changed their responsivity by iontophoresis of ACh alone. A long-term enhancement of evoked

Fig. 9. Cytochrome oxidase (CO) histochemistry taken in t h e coronal plane through the somatosensory cortex of a n animal t h a t received a basal forebrain lesion. T h e normal hemisphere is on the left; the hemisphere ipsilateral to the lesion is on the right. T h e CO pattern appears similar on both sides. Scale bar = 1 mm.

I

I

II

111

--

I

IV

I

V

VI

DISTANCE Fig. 10. Demonstration of the absorption of "H-atropine in the somatosensory cortex. On the right is an autoradiograph of the distribution of 'H-atropine. On the left is a Nissl stain of the same section. Cortical laminae are indicated with Roman numerals. On the bottom is a curve indicating the optical density values taken in a vertical slice

through the autoradiograph. The curve demonstrates that 50% of a given concentration of topical atropine should he present in the center of layer V. Arrows point to the same site in both images. Scale bar = 250 pm.

Fig. 11. Two-dimensional, digitized reconstruction of metabolic activity in an animal that received a topical application of 100 p M atropine, but was not stimulated. The treated hemisphere is on the right. No evoked activity is present in either hemisphere, and both the treated and the normal side appear similar in overall density. The

darker grey represented in these reconstructions corresponds to values 25-30% above white matter values; the lighter grey corresponds to values 20-25% above background. Other conventions are as for Figure 6. CS, coronal sulcus. Med, medial; Ant, anterior.

Fig. 12. Autoradiographs taken through the somatosensory cortex of an animal t h a t received a topical application of atropine (50 pM). T h e animal received electrocutaneous stimuli to the central and pisiform pads bilaterally. Arrows point to patches of stimulus evoked activity.

b

DENSITY

T h e evoked activity on the treated side (on the right) is reduced. The normal hemisphere is on the left. Autoradiographs were photographed directly from the autoradiographic film and are not digitized. Scale bar = 1.5pm.

DENSITY

a

c

LAYERS Fig. 13. Digitized autoradiographs taken from the right and left hemispheres of a n animal t h a t received a 50 pM topical application of atropine. On t h e bottom are 3-dimensional reconstructions of the

LAYERS patches of stimulus-evoked label outlined with boxes. See Figure 5 for details. Scale bar = 1 mm.

S.L. JULIANO ET A],.

118

Fig. 14. Two-dimensional reconstruction of stimulus-evoked activity of an animal that was treated with a topical application of 100 pM atropine. The point of alignment for these reconstructions was the fundus of the ansate sulcus (ANS). The normal hemisphere is on the left; the atropine treated hemisphere is on the right. The darkest activity

values represented here (i.e., black) correspond to densities 65-70%) above white matter values. The lightest grey in the reconstruction represents values 50-55% above white matter densities. Other conventions are as for Figure 6. Med, medial; Ant, anterior.

potentials was found in cat somatosensory cortex when the basal forebrain was stimulated concurrently with peripheral stimulation (Rasmusson and Dykes, '88). Topical applications of atropine in the same experiment also appeared to block this effect. In the auditory cortex, ACh is also neuromodulatory, as evidenced by the observation that ACh or muscarinic agonists modulated spontaneous and evoked activity, but were more likely to affect evoked than spontaneous activity (McKenna et al.. '88). Metherate and Weinberger ('89) also found that frequency-specific alterations in auditory receptive fields occurred during ACh application and were blocked by atropine. Taken together, the balance of available evidence indicates that ACh released in sensory cortex by basal forebrain afferents facilitates responses evoked by input from the thalamus.

interpretation because in the 2DG experiments, the middle and upper layers of the somatosensory cortex were selectively activated by stimulation. It was rare for the metabolic activity to extend deeply into layer V. This pattern of stimulus-evoked metabolic activity is similar to that seen previously in cat somatosensory cortex (Juliano et al., '89). Recent reports of cat visual and somatosensory cortex suggest that neurons in all cortical layers are cholinoceptive. Metherate et al. ('88a) describe that somatosensory neurons responding to iontophoretic application of ACh are distributed similarly to all neurons encountered with receptrve fields. Sato et al. ('87a) find that neurons encountered in layers I1 through VI of striate cortex respond to ACh application. Sillito and Kemp ('83) describe neurons facilitated by ACh in all layers of visual cortex. Studies detailing the distribution of AChE or ChAT in various regions of cerebral cortex indicate that cholinergic fibers are present in all layers, but exhibit dense Dlexuses in specific sites, which vary-with cortical region (Mesulam et al., '84; Bear et al., '85; Stichel and Singer, '87). In our own evaluation of AChE positive fibers, all cortical layers are

Laminar distribution In the present experiments, we found that atropine applied topically began decreasing its concentration in layer Va. We do not believe, however, that this confounds our

2DG A N D ACH I N CAT SOMATOSENSORY CORTEX

119

Fig. 15. Two-dimensional reconstruction of stimulus-evoked activity that resulted from an animal unilaterally treated with a 50 p M topical application of atropine. T h e sections were aligned on the fundus 01 the coronal sulcus (CS). T h e treated hemisphere is on the right; the

normal hemisphere on the left. T h e darkest metabolic activity levels represented here (i.e., black) are 75-80% above white matter values; the lightest grey levels are 50-55% above white matter values. Other conventions are as for Figure 6.

richly reactive, with the densest distribution in layers I and V (Fig. 1).Although the distribution of AChE positive fibers may not consistently parallel the distribution of fibers reactive for choline acetyl transferase (ChAT) (Lysakowski el, al., '89), studies comparing the relationship between AChE and ChAT distributions in cat visual cortex suggest that the laminar arrangements are similar (Stichel and Singer, '87). The distribution of cholinergic fibers, however, does not always correspond to the arrangement of receptors sensitive to ACh. Although ACh activates both muscarinic and nicotinic receptors, it is very likely that the cholinergic action producing excitation of neurons in the neocortex is muscarinic in nature (Krnjevic and Phillis, '63; Krnjevic e t al., '71; McCormick and Prince, '85). In many species, including cat, muscarinic receptors are distributed most densely in layers 1-111 of the somatosensory cortex (Shaw et al., '86; Mash et al., '88; Sampson et al., '88). Thus it appears that the upper layers of the cortex are likely to respond to ACh application, given the distribution of AChE positive fibers and muscarinic receptors, and that these layers also respond to the kind of somatic stimulation delivered in these experiments, given the finding that the highest density of 2DG uptake was found in layers 11-IV. All this taken together implies that our topical applications of atropine contacted the appropriate regions of cortex to modulate somatosensory information.

(McCormick, '89, for review). This hypothesis has led to the suggestion that enduring changes are permitted to occur only while cells actively discharge in response to stimulation (Bear and Singer, '86; Metherate et al., '87). Our studies involving basal forebrain lesions suggest that absence of ACh leads to enduring changes in the somatic map activated by digit stimulation. If map plasticity requires a threshold amount of driven activity, ACh may participate by facilitating normally subthreshold inputs. The stimulus-evoked maps we find in the absence of cholinergic activity appear to represent an image of the cortical response without the normal influence of ACh that enhances neural responses in the presence of somatic input. Dykes and Lamour ('88) suggest that a substantial population of neurons without receptive fields exists in the somatosensory cortex. One factor that can induce a population of inactive neurons to become active is ACh. Our data reinforces the idea that borders in somatotopic maps are labile, and that ACh is an important variable in determining map borders. Deficits in sensory discriminative ability are known to occur in Alzheimer's disease (Freedman and Oscar-Berman, '87). Recently, Wozniak et al. ('89a,b) demonstrated that rats with cholinergic depletion in the neocortex are impaired in sensory discrimination tasks. These deficits appear to be due to impairments in sensory information processing and not a result of memory limitations (Wozniak, '89b). The reductions we find in the metabolic map in the absence of ACh may be akin to reductions in sensory discriminative ability such as those that are found in Alzheimer's disease.

Is ACh involved in cortical plasticity? In the somatosensory and visual cortex, ACh causes long-lasting changes in neural responsivity (Metherate et al., '87; Sillito and Murphy, '87; Metherate et al., '88b; Rasmusson and Dykes, '88). The mechanism that produces thiese changes may be due to alterations in potassium currents that are sensitive to increases in neural activity

ACKNOWLEDGMENTS The authors thank Dr. John Sarvey for helpful comments on an earlier version of the manuscript. This work was

S.L. JULIANO ET AL.

120 supported by PHS NS-24014 (S.L.J.) and DO1)-RO-7064

(S.L.J.).

LITERATURE CITED Barstad, K.E., and M.F. Bear (1987) The cholinergic innervation of somatic Neurosci. Abs. 13r249. sensory cortex in the cat. SOC. Barstad, K.E., and M.F. Bear (1990) The cholinergic innervation of somatic sensory cortex in the cat. J. Neurophysiol, in press. Bear, M.F., K.M. Carnes. and F.F. Ebner (1985) An investigation of cholinergic circuitry in cat striate cortex using acetylcholinesterase histochemistry. J. Comp. Neurol. 234:411-430. Bear, M.F., and W. Singer (1986) Modulation of visual cortical plasticity by acetylcholine and noradrenaline. Nature 320:172-176. Berman, A.L., and E.G. Jones (1982) “The Thalamic and Basal Telencephalon of the Cat: A Cytoarchitectonic Atlas With Stereotaxic Coordinates. Madison, WI: U. of Wisconsin Press. Brown, D.A. (1983) Slow cholinergic excitation-a mechanism for increasing neuronal excitability. T.I.N.S. 6:302-307. Donoghue, J.P., and K.L. Carrol (1987) Cholinergic modulation of sensory responses in rat primary somatic sensory cortex. Brain Res. 408:367-371. Dykes, R., and Y. Lamour (1988) Neurons without demonstrable receptive fields outnumber neurons having receptive fields in samples from the somatosensory cortex of anesthetized or paralyzed cats and rats. Brain Res. 440:133-143. Freedman, M., and Oscar-Berman, M. (1987) Tactile discrimination learning in Alzheimer’s and Parkinson’s diseases. Arch. Neurol. 44r394-398. Halliwell, J.V., and P.R. Adams (1982) Voltage clamp analysis of muscarinic excitation in hippocampal neurons. Brain Res. 250r71-92. Jacobowitz, D.M., and G.J. Creed (1983) Cholinergic projection sites of the nucleus of tractus diagonalis. Brain Res. Bull. 10:365-371. Juliano, S.L., D.E. Eslin, and W. Ma (1988) Manipulation of cortical cholinergic innervation alters stimulus-evoked metabolic activity in cat Neurosci. Abs. 14:844. somatosensory cortex. SOC. Juliano, S.L., and B.L. Whitsel (1987) A combined 2-deoxyglucose and neurophysiological study of primate somatosensory cortex. J. Comp. Neurol. 263514-525. Juliano, S.L., B.L. Whitsel, M. Tommerdahl, and S.S. Cheema (1989) Determinants of patchy metabolic labeling in the somatosensory cortex of cats: A possible role for intrinsic inhibitory circuitry. J. Neurosci. 9:l-12. Koelle, G.B. (1955) The histochemical identification of acetylcholinesterase in cholinergic adrenergic and sensory neurons. J. Pharmacol. Exp. Ther. 114:167-184. Krjevic, K., and J.W. Phillis (1963) Pharmacological properties of acetylcholine-sensitive cells in the cerebral cortex. J. Physiol. (Lond.) 166:328-350. Krnjevic, K., R. Pumain, and L. Renaud (1971) The mechanism of excitation by acetylcholine in the cerebral cortex. J. Physiol. (Lond.) 215:247-268. Lamarca, M.V., and H.C. Fibiger (1984) Deoxyglucose uptake and choline acetyltransferase activity in cerebral cortex following lesions of the nucleus basalis magnocellularis. Brain Res. 307:366-369. Lamour, Y., P. Dutar, and A. Jobert (1982) Spread of acetylcholine sensitivity in the neocortex following lesion of the nucleus basalis. Brain Res. 252:377-381. Lamour, Y., P. Dutar, A. Jobert, and R.W. Dykes (1988) An iontophoretic study of single somatosensory neurons in rat granular cortex serving the limbs: A laminar analysis of glutamate and acetylcholine effects on receptive-field properties. J. Neurophysiol. 60:725-750. Lamour, Y., and R.W. Dykes (1988) Somatosensory neurons in partially deafferented rat hindlimb granular cortex subsequent to transection of the sciatic nerve: Effects of glutamate and acetylcholine. Brain Res. 449:1%33. London, E.D., M. McKinney, M. Dam, A. Ellis, and J.T. Coyle (1984) Decreased cortical glucose utilization after ibotenate lesion of the rat ventromedial globus pallidus. J. Cereb. Blood Flow Metab. 4:381-390. Lysakowski, A,, B.H. Wainer, G. Bruce, and L.B. Hersh (1989) An atlas of the regional and laminar distribution of choline acetyltransferase immunoreactivity in rat cerebral cortex. Neuroscience 28:291-336. Ma, W., C.F. Hohmann, J.T. Coyle, and S.L. Juliano (1989) Lesions of the basal forebrain alter stimulus-evoked metabolic activity in mouse somatosensory cortex. J. Comp. Neurol. 288:414427.

Mash, D.C., W.F. White, and M.-M. Mesulam (1988) Distribution of muscarinic receptor subtypes within architectonic subregions of the primate cerebral cortex. J. Comp. Neurol. 278:265-274. McCormick, D.A. (1989) Cholinergic and noradrenergic modulation of thalamocortical processing. T.I.N.S. 12215-221 McCormick, D.A., and D.A. Prince (1985) Two types of muscarinic response to acetylcholine in mammalian cortical neurons. Proc. Natl. Acad. Sci. U.S.A. 82:6344-6348. McKenna, T.M., J.H. Ashe, G.K. Hui, and N.M. Weinberger (1988) Muscarinic agonists modulate spontaneous and evoked unit discharge in a u d tory cortex of cat. Synapse 2:54-68. Mesulam, M.-M., A.D. Rosen, and E.J. Mufson (1984) Regional variations in cortical cholinergic innervation: Chemoarchitectonics of acetylcholinesterase-containing fibers in the Macaque brain. Brain Res. 311:245-258. Metherate, R., N. Tremblay, and R.W. Dykes (1987) Acetylcholine permits long-term responsiveness in cat primary somatosensory cortex. Neuroscience 2273-81. Metherate, R., N. Tremblay, and R.W. Dykes (1988a) The effects of acetylcholine on response properties of cat somatosensory cortical neurons. J. Neurophysiol. 59:1231-1252. Metherate, R., N. Tremblay, and R.W. Dykes (1988b) Transient and prolonged effects of acetylcholine on responsiveness of cat somatosensory cortical neurons. J. Neurophysiol. 59.1253-1276. Metherate, R., and N.M. Weinberger (1989) Acetvlcholine uroduces stimulusspecific receptive field alterations in cat auditory Eortex. Brain Res. 480:372-377. Rasmusson, D.D., and R.W. Dykes (1988) Long-term enhancement of evoked potentials in cat somatosensory cortex produced by co-activation of the basal forebrain and cutaneous receptors. Exp. Brain Res. 70:276-286. Sampson, S.M., C. Shaw, M. Wilkinson, and D.D. Rasmusson (1988)Sensory deafferentation fails to modify muscarinic receptor binding in raccoon somatosensory cortex. Brain Res. Bull. 20r597-601. Sato, H., Y. Hata, K. Hagihara, and T. Tsumoto (198713) Effects of cholinergic depletion on neuron activities in the cat visual cortex. J. Neurophysiol. 58:781-794. Sato, H., Y . Hata, H. Masui, and T. Tsumoto (1987a) A functional role of cholinergic innervation to neurons in the cat visual cortex. J. Neurophysiol. 58:765-780. Shaw, C., M. Wilkinson, M. Cynader, M.C. Needler, C. Aoki, and S.E. Hall (1986) The laminar distributions and postnatal development of neurotransmitter and neuromodulator receptors in cat visual cortex. Brain Res. Bull. 16:661-671. Sillito, A.M., and J.A. Kemp (1983) Cholinergic modulation of the functional organisation of the cat visual cortex. Brain Res. 289r143-155. Sillito, A.M., and P.C. Murphy (1987) The cholinergic modulation of cortncal function. In E.G. Jones and A. Peters (eds): Cerebral Cortex, v. 6. pp 161-185, Plenum, NY. Sokoloff, L., M. Reivich, C. Kennedy, M.H. Des Rosiers, C.S. Patlak, K.D. Pettigrew, 0. Sakurada, and M. Shinohara (1977) (“C) Deoxyglucoss method for the measurement of local cerebral glucose utilization: Theory, procedure, and normal values in the conscious and anesthetized albino rat. J. Neurochem. 28:897-916. Stichel, C.C., and W. Singer (1987) Quantitative analysis of the choline acetyltransferase-immunoreactive axonal network in the cat primary visual cortex: I. Adult cats. J. Comp. Neurol. 258:91-98. Tommerdahl, M., R. Baker, B.L. Whitsel, and S.L. Juliano (1985) A method for reconstructing patterns of somatosensory cerebral cortical activity. Biomed. Sci. Instrum. 21:93-98. Williamson, A.M., and J.M. Sarvey (1985) Effects of cholinesterase inhibitors on evoked responses in field CAI of the rat hippocampus. J. Pharmacol. Exp. Ther. 235448-454. Wong-Riley, M.T.T. (1979) Changes in the visual system of monocularly sutured or enucleated cats demonst,rable with cytochrome oxidase histochemistry. Brain Res. 171:11-28. Wozniak, D.F., G.R. Stewart, S. Finger, J.W. Olney, and C. Cozzari (1989a) Basal forebrain lesions impair tactile discrimination and working memory. Neurobiol. Aging. 10r173-179. Wozniak, D.F., G.R. Stewart, S. Finger, and J.W. Olney (1989b) Comparrson of behavioral effects of nucleus basalis magnocellular lesions and somatosensory cortex ablation in the rat. Neuroscience, 32685-700.