- Email: [email protected]
Distribution of choline acetyltransferase-immunoreactive axons in monkey frontal cortex
0306-4522/91 $3.00+ 0.00 PergamonPress plc © 1991IBRO
Neuroscience Vol. 40, No. 2, pp. 363-374, 1991 Printed in Great Britain
DISTRIBUTION OF CHOLINE ACETYLTRANSFERASEIMMUNOREACTIVE AXONS IN MONKEY FRONTAL CORTEX D. A. LEWIS Departments of Psychiatry and Behavioral Neuroscience, University of Pittsburgh, 3811 O'Hara Street, Pittsburgh, PA 15213, U.S.A. Al~tract--Cholinergic neurons of the nucleus basalis of Meynert project to numerous regions of the cerebral cortex. However, little is known about the regional and laminar distributions of cholinergic axons in monkey frontal cortex. In this study, immunohistoehemical techniques were used to identify axons that were immunoreactive for choline acetyltransferase, the enzyme that catalyses the synthesis of acetylcholine, in the frontal cortex of cynomolgus monkeys. Motor cortex contained the greatest density of labeled fibers; the density of labeled fibers was lower in premotor and anterior cingulate cortices and lower still in the association regions of prefrontal cortex. On a laminar basis, choline acetyltransferaseimmunoreactive axons were most dense in layer I to superficiallayer III. Layer V also contained a distinct band of labeled fibers that was particularly prominent in the agranular regions of frontal cortex. The density of labeled fibers was much lower in the deep portion of layer III to layer IV and in layer VI. These findings demonstrate a specific and regionally distinctive cholinergic innervation of monkey frontal cortex that may reveal the anatomical basis for the influence of acetylcholine on the diverse functions of primate frontal cortex.
The frontal lobes in primates contain distinct cortical regions that differ substantially in cytoarchitectonic organization, connectivity and function. The agranular primary motor and premotor regions are located in the most caudal portions of the dorsomedial and lateral surfaces of the frontal lobes. Anterior to these areas, the granular association regions of the prefrontal cortex occupy the lateral, medial and orbital surfaces. In addition, the medial surface of the frontal lobes contains the paralimbic areas of the cingulate cortex. In primates, the cortex of the frontal lobes, especially the prefrontal regions, is proportionally much larger and is substantially more differentiated than in rodents. These changes are associated with marked modifications in the patterns of innervation of frontal cortex by afferent fibers from non-thalamic subcortical nuclei. For example, although dopaminergic axons primarily innervate prefrontal and anterior cingulate cortices in rats, 3'4'9'14'29,44 primary motor cortex contains the greatest density of dopaminergic fibers in monkeys. 5:1'25 In addition, noradrenergic axons have a relatively uniform regional distribution in rodent neocortex, 24,4° whereas in monkey, substantial regional heterogeneity is present in the density of noradrenergic fibers. 27,39 Axons from the cholinergic neurons of the nucleus basalis of Meynert also innervate many, if not all, regions of the primate cerebral cortex, including ACHE, acetylcholinesterase; BSA, bovine serum albumin; CHAT, choline acetyltransferase; DAB, 3,Y-diaminobenzidine; TBS, Tris-buffered saline.
Abbreviations:
those in the frontal lobes. 21,33,36-38 In non-human primates, these projections are topographically organized t9'34'35'41and non-collateralized, at least between frontal and parietal regions.47 Biochemical studies of choline acetyltransferase (CHAT) activity, the enzyme that catalyses the synthesis of acetylcholine, have revealed substantial regional differences in the distribution of this enzyme in monkey cerebral cortex? ~'3s Histochemical investigations of the degradative enzyme acetylcholinesterase (ACHE) produced similar observations in both monkeys and humans, 12"33,36 suggesting that the density of cholinergic innervation was greatest in limbic and paralimbic regions, intermediate in primary motor and primary sensory regions and least in association areas of the cortex. However, interpretation of the histochemical studies has been hindered by questions regarding the specificity of AChE as a marker of cholinergic axons, s,2°,23 The production of specific antibodies directed against ChAT has made it possible to selectively identify cholinergic neurons and axons. For example, Campbell e t al. 6 demonstrated a distinctive laminar organization of the ChAT-immunoreactive, cholinergic afferents to monkey primary auditory cortex. In this study, immunohistochemical techniques were used to determine the regional and laminar distributions of ChAT-positive fibers in the frontal cortex of Old World cynomolgus monkeys. These investigations were conducted in order to determine whether the regional distribution of ChAT-immunoreactive axons in frontal cortex was consistent with the functional organization of the cholinergic system proposed in previous AChE histochemical studies, 36 and
363
364
D.A. LEWIS
if the laminar pattern o f cholinergic innervation was similar or c o m p l e m e n t a r y to that o f projections from other subcortical non-thalamic nuclei. EXPERIMENTAL PROCEDURES
Five male (3.2-4.4 kg) cynomolgus monkeys (Maeaca fascicularis) (Hazleton Research Products Inc., Denver, PA) were used in this study. Animals were deeply anesthetized with ketamine hydrochloride (25 mg/kg, intramuscularly) and pentobarbital sodium (10 mg/kg, intraperitoneally). An endotraclieal tube was inserted and mechanical ventilation with oxygen (2 l/min) was initiated. After the chest was opened, 1.5-2.0ml of 1% aqueous sodium nitrite was injected into the left ventricle of the heart. Animals were then perfused transcardially with cold 1% paraformaldehyde in phosphate buffer (0.15 M) for 15-30 s followed by perfusion with cold 4% paraformaldehyde in phosphate buffer. The latter solution was perfused for approximately 9 min at a flow rate of 300-350 ml/min. Immediately following the perfusion, the brain was removed and sliced into blocks, 3-5 mm thick, and placed in cold fixative for an additional 2 h. Tissue blocks were then washed in a series of cold, graded sucrose solutions and sectioned either coronally or sagittally in a cryostat at 40 #m. Tissue sections were incubated with a rat monoclonal antibody (kindly supplied by B. Wainer) directed against CHAT. This antibody (AB8) recognizes ChAT from a number of mammalian species, 22 and has previously been utilized in immunohistochemical studies in monkey brain) '35 The anti-ChAT antibody was diluted 1:500 in phosphatebuffered saline containing 0.5% Triton X-100, 2% bovine serum albumin (BSA) and 20% normal goat serum. The sections were incubated in the primary antiserum for ap-
proximately 48 h at 4~C and then rinsed in three changes of 0.1-M Tris-buffered saline (TBS) for 30 min. Tissue sections were then incubated for 90min in a 1:50 dilution of goat anti-rat IgG secondary antiserum (Sternberger.-Meyer, Jarrettsville, MD) in TBS containing 0.2% Triton X-100, 2% BSA, 20% normal goat serum and 10% normal cynomolgus monkey serum. Following a 30-rain wash in TBS, the sections were placed for 2 h in a 1:50 dilution of rat peroxidase-antiperoxidase (Sternberger--Meyer) in the same solution as the secondary antiserum. Sections were washed and then incubated again in the secondary and tertiary solutions for 60 and 90 rain, respectively. The tissue was then washed successively in TBS and phosphate buffer and incubated in phosphate buffer containing 0.05% 3,Ydiaminobenzidine (DAB) and 0.015% H20 z. The DAB reaction product was intensified by serial immersions of the mounted sections in aqueous solutions of 0.005% osmium tetroxide and 0.5% thiocarbohydrazide. Some sections were counterstained with Giemsa ~6 which intensifies the DAB reaction product and also produces a Nissl stain of nonimmunoreactive cells. In order to assess the specificity of immunohistochemical labeling, the anti-ChAT antibody was preabsorbed with 50 nmol/ml of human placental ChAT (Sigma, St Louis, MO) for 16 h before incubation with tissue sections. In other experiments, the anti-ChAT antibody was omitted. In both conditions, immunoreactivity in the cortex was abolished compared to controls. Regional and laminar patterns of ChAT immunoreacfivity were determined by comparison of immunohistochemical sections with adjacent Nissl-stained sections and by evaluation of the Giemsa-counterstained immunohistochemical sections. The following cortical regions of the frontal lobe were examined (Fig. 1): primary motor cortex (dorsal area 4), premotor cortex (dorsal area 6), anterior
Fig. 1. Map of the cytoarchitectonic areas of the macaque frontal cortex adapted from Walker: 46 This map depicts the general location of each cytoarchitectonic region; the precise locations of regional boundaries differed slightly across animals. Shaded areas represent those cortical regions examined in the present study. Dots indicate the approximate anatomical location of the indicated photomicrograph. The exact cytoarchitectonic area depicted in each photomicrograph was determined from adjacent Nissl sections.
365
Choline acetyltransferase in monkey frontal cortex cingulate cortex (area 24) and prefrontal cortex (areas 8B, 9-12, 25, 46). These cytoarchiteetonic areas were identified according to published criteria. TM The relative density of labeled fibers was determined with the aid of a computer-assisted image analysis system (Southern Micro Instruments, Atlanta, GA). All labeled fibers were counted in a grid (total area 0.01 mm 2) overlaying a bright-field image at a final magnification of x 400. Results were expressed as the mean (+S.E.) number of fibers in five fields for a given cortical region and layer. Differences among means were assessed with one-way analy-
sis of variance. Tukey's method (ct level = 0.05) was used for between-group comparisons. RESULTS
Morphology of choline acetyltransferase-immunoreactive fibers The vast majority of ChAT-immunoreactive fibers were extremely fine and studded with small round or oval varicosities (Fig. 2A-C). Intervaricose segments
t Ip
F . -4
i
-~
~"
'
:
•
,
"¢
*
~)"-
-
°:
,
,
•,'~ /
-
'
"'4
~,ii.?:,: ,, - ,:, ~ ~
',
~
'
~
17. . . .
-
-'~
" ...................
Fig. 2. Bright-field ( A ~ , G) and diffusion interference contrast (D-F) photomicrographs of ChAT-immunoreactive fibers in monkey frontal cortex• The majority of ChAT-positive fibers in layer I to superficial layer III were extremely thin and varicose (A42), whereas a smaller number of fibers with a thicker and smoother morphology and a radial orientation were present in layers II-VI (D-F). Note the bundle of tangentially oriented fibers in the white matter below the anterior cingulate cortex ((3). Scale bar in B = 3 0 # m and applies to A~2; scale bar in E = 50/~m and applies to D-F; scale bar in G = 400#m.
366
D.A. LEWIS
of the axons were very thin, and at times, barely detectable. Other immunoreactive fibers were thicker in diameter and fairly smooth in appearance, although in some cases they were punctuated at fairly regalar intervals by elongated oval swellings (Fig. 2D). Fibers with this morphological appearance and a radial orientation were commonly observed in layers II-VI (Fig. 2D-F). Some of these fibers turned 90 ° in layers I or V to assume a tangential orientation (Fig. 2E), whereas others bifurcated and formed thinner and more varicose fibers (Fig. 3). Labeled
Fig. 3. Diffusion interference contrast photomicrographs of ChAT-immunoreactive axons in the superficial layers of area 11. Arrows in B and C refer to the same structure indicated by the corresponding arrows in A. Note that both radially and tangentially oriented labeled fibers bifurcate to form axons of smaller diameter. Scale bar in A = 50/~m; in B and C=20#m.
fibers with a similar morphology but with a tangential orientation were also observed in layers l 1I, V and the subcortical white matter (Fig. 2G). No CHATpositive neurons were observed in any areas of cynomolgus monkey frontal cortex.
Regional distribution of choline acetyltransJeraseimmunoreactive fibers The most striking feature of the regional distribution of ChAT-immunoreactive fibers in cynomolgus monkey frontal cortex was a rostral to caudal gradient of increasing fiber density. These regional differences in density of innervation were most apparent in the dorsomedial portions of frontal cortex. In sagittal sections through the medial frontal cortex (Fig. 4), fiber density gradually increased from area l0 in the frontal pole (Fig. 5A) through areas 9 and 8B (Fig. 5B) and then exhibited a stepwise increase between area 8B and the agranular area 6 of premotor cortex (Fig. 5C). Fiber density continued to increase through area 6 and reached a maximum in the caudal portions of area 4 (primary motor cortex). This rostral to caudal gradient of increasing fiber density was also apparent in the dorsal bank of the cingulate sulcus (Fig. 4), but was much less obvious through area 24 (anterior cingulate cortex) on the ventral bank of the cingulate sulcus, or across the regions of the dorsolateral prefrontal cortex. These regional differences in the relative density of ChAT-positive fibers were confirmed by quantitative assessments. For example, the mean (_+ S.E.) density (per 0.01 mm 2) of labeled fibers in layer V was 18.0 (_+0.9) in area 10, 28.8 (_+2.6) in area 8B and 39.2 (_+ 1.2) in area 4 (F2,~2= 35.94; P < 0.001). Pairwise comparisons revealed significant (P <0.05) differences in fiber density among all three areas. Despite this pattern of regional heterogeneity in the density of ChAT-positive axons, within the prefrontal cortex there were relatively few differences in overall fiber density either within or across cytoarchitectonic areas at a given rostral caudal level. This regional similarity in fiber density across regions was most apparent across rostral-prefrontat areas. In more caudal regions, for example in coronal sections taken at the mid-point of the principal sulcus, the density of labeled fibers appeared to be slightly lower on the dorsolaterai surface (area 46; Fig. 6) than on the dorsomedial convexity (area 9; Fig. 7) and orbital (area 11; Fig. 8) surface. This apparent differences in density of labeled fibers was due, in part, to a more prominent band of immunoreactive fibers in layer V of area 9 (Fig. 7) than in area 46 (Fig. 6). These distinctive patterns of axon density across frontal regions were present in each of the five monkeys examined in this study.
Laminar distribution of choline acetyltransferaseimmunoreactive fibers In all regions examined, the greatest density of labeled fibers was present in layer I (Figs 6-8). Axons
Choline acetyltransferase in monkey frontal cortex
367
Fig. 4. Low power dark-field photomicrograph of ChAT immunoreactivity in a sagittal section through medial frontal cortex of a cynomolgus monkey. The frontal pole is to the left and the dorsal surface is at the top. ps, principal sulcus; cs, cingulate sulcus. This section is located approximately 4 mm lateral to the midline so that much of it cuts through the fundus of the cingulate sulcus, revealing both the dorsal and ventral banks of this sulcus. Note the relatively low density of ChAT-immunoreactive fibers in the cortex of the frontal pole and orbital surface. On the dorsal surface, fiber density increases in a rostral to caudal direction; note especially the increasing prominence of labeled fibers in layer V. The rostral to caudal gradient of increasing fiber density in layer V is also evident in the dorsal bank of the cingulate sulcus, a-c, location of the corresponding photomicrographs in Fig. 5. in this layer were predominately tangential in orientation. This plexus of fibers extended into layer II and then diminished in density in the superficial portion of layer III. Some oblique and radial fibers intermixed with tangentially oriented fibers in these layers. Fiber density was substantially lower in the deep portion of layer III, layer IV and the superficial most portions of layer V. In these layers, fibers had both radial and oblique orientations. In the middle to deep portions of layer V, there was an increase in fiber density due to the addition of tangentially oriented fibers to the plexus of oblique and radially oriented fibers. This fiber band was most prominent in motor cortex (area 4), especially in the portion of layer V containing Betz cells (Fig. 9). However, ChAT-positive fibers did not form pericellular baskets or other specializations and thus they did not appear, at least at the level of light microscopy, to be preferentially targeting the somata of Betz or other pyramidal cells. In all regions of frontal cortex, fiber density decreased substantially in layer VI. The deepest portion of this layer contained some larger caliber tangentially oriented fibers that increased in number in the white matter just below the cortex. These labeled fibers were frequently observed to sweep into the cortex where they then became radially arrayed. Within the white matter, ChAT-positive fibers were most dense below the anterior cingulate cortex (Fig. 2G). This laminar pattern of ChAT-immunoreactive axons was, in general, found in all cytoarchitectonic areas of monkey frontal cortex. However, in the more
densely innervated premotor and motor regions, some differences across layers were less apparent. DISCUSSION
Regional distribution patterns
Using an antibody that selectively recognizes CHAT, we found that cholinergic axons innervated every region of monkey frontal cortex examined. However, substantial differences were present across some regions in the density of ChAT-immunoreactive axons. For example, labeled fibers exhibited a rostral to caudal gradient of increasing density that was due predominantly to the greater density of fibers in premotor (area 6) and motor (area 4) regions compared to prefrontal regions. In contrast, within the prefrontal cortex, ChAT-positive fibers had a more uniform regional distribution. Only in the more caudal portions of prefrontal cortex did the medial (area 24) and dorsomedial (areas 9, 8B) regions have a slightly greater density of labeled fibers than the lateral (area 46) regions. These regional patterns of cholinergic innervation parallel those described in biochemical assessments of ChAT activity.21,3sIn these studies, ChAT activity was, in general, higher in the posterior frontal cortex than in prefrontal regions, which contained very uniform levels of ChAT activity. The findings of this study also confirm many observations of previous studies that employed AChE histochemistry. Mesulam et al. 36 reported tow and uniform levels of AChE staining across most prefrontal regions of rhesus monkey cortex, with
368
D.A. LEWIS
greater levels in supracallosal portions of area 24 and in areas 4 and 6. They also described a rostral to caudal increase in AChE staining across areas 6 and 4 similar to the gradient formed by ChAT-positive fibers. Based upon the distribution of AChE staining, Mesulam et al. suggested that cholinergic afferents were most dense in limbic and paralimbic areas (excluding the cingulate gyrus), intermediate in primary sensory and motor regions, and least dense in association cortices. 33'36'38 Our findings, in concert with other investigations of ChAT-immunoreactive fibers in monkey amygdala ~ and primary auditory cortex 6 support this general organizational scheme. For example, some portions of the amygdala, such as the magnocellular subdivision of the basal nucleus, contain a very high density of ChAT-immunoreactive fibers and terminals. ~ In primary auditory cortex, the middle cortical layers also receive a dense plexus of ChAT-positive axons. 6 In addition, in the present study the motor regions of the frontal lobe had a greater density of labeled fibers than did the prefrontal association areas. However, the density of ChAT immunoreactivity differs substantially among the var-
ious nuclei of the monkey amygdala, 1and the density of ChAT-positive axons in the caudal, medial portions of prefrontal cortex may approach that in primary visual cortex (unpublished observations). Thus, these comparisons suggest that although the distribution of ChAT-positive fibers provides evidence consistent with the hypothesis that cholinergic afferents preferentially innervate limbic over motor and sensory regions and the latter regions over association areas, there appear be some exceptions to this general pattern. L a m i n a r distribution p a t t e r n s
Mesulam et al. 36 also reported that AChE-stained fibers were predominately present in layers V and VI in motor and premotor regions of monkey cortex, whereas in prefrontal cortex, lower numbers of labeled fibers were found in layers I and V V I . However, they noted that although some AChE reaction product was clearly in the form of fibers, more frequently it was found 'fin the form of clumps which lacked discernible internal structure". In contrast, ChAT-labeled fibers were very distinctive
Fig. 5. Dark-field photomicrographs of ChAT-immunoreactive fibers in areas 10 (A), 8B (B) and 6 (C) of cynomolgus monkey frontal cortex. Note the regional differences in overall fiber density and in the prominence of the band of labeled fibers in layer V (arrows). Scale bar in C = 400 ~m and applies to A C.
Choline acetyltransferase in monkey frontal cortex morphologically, and they had precise laminar patterns of distribution. In all regions of monkey frontal cortex, ChAT-immunoreactive fibers were most dense in layer I to superficial layer III; this finding is Consistent with the report that ChAT activity levels are highest in the same layers of human prefrontal cortex. 7 Layer V also contained a band of fibers of increased density, although the prominence of this band exhibited regional specificity that was correlated with certain cytoarchitectonic characteristics. In general, regions with a well-developed layer IV (e.g. areas 10, 46) had only a mild increase in fiber density in layer V, whereas agranular areas, such as motor, premotor and anterior cingulate cortices, had a much more distinctive band (see Fig. 4). These differences were also related to the size of layer V pyramidal
369
neurons. For example, among agranular cortices, the prominence of labeled fibers in layer V increased from anterior cingulate cortex to premotor cortex to primary motor cortex in a manner similar to the increase in size of layer V pyramidal cells across these regions. In addition, in primary motor cortex, the density of ChAT fibers was greatest in the portion of layer V containing Betz cells. However, at the light microscopic level, ChAT-positive axons did not appear to form pericellular baskets or other specializations which might have suggested that they were selectively innervating pyramidal cell somata. These laminar patterns are of interest in comparison to the distribution of ChAT-immunoreactive fibers in monkey primary auditory cortex. 6 In this region, cholinergic axons were very dense in layer I
Fig. 6. Dark-field photomicrograph of ChAT immunoreactivity (A) on the ventral bank of the principal sulcus (area 46) of cynomolgus monkey prefrontal cortex, and a bright-field photomicrograph of an adjacent Cresyl Violet-stained section (B). Roman numerals indicate the cortical layers. WM, white matter. Scale bar in A = 200/~m and applies to A and B.
370
D.A. LEwis
and deep layer III to layer IV, but not in layer V. Although this study employed a different anti-ChAT antiserum, direct comparisons of that antiserum and the one used in the present study have produced similar results in monkey amygdala. 1 Similarly, in monkey primary visual cortex ChAT-positive fibers also preferentially project to layers I and IV (unpublished observations). These comparisons suggest that cholinergic afferents may have both common and
regionally distinctive influences on cortical function. In all cortical regions, the synaptic targets of cholinergic axons may be the distal apical dendrites of pyramidal neurons. Indeed, in rodent cerebral cortex ChAT-positive terminals form synaptic contacts, predominately symmetric, with apical dendrites of pyramidal neurons, 15 and in both rodent and human neocortex apical dendrites have been reported to contain cholinergic receptors. 42'43'45 In addition, the
Fig. 7. Dark-field photomicrograph of ChAT immunoreactivity in the dorsal bank of the cingulate sulcus (area 9) of cynomolgus monkey prefrontal cortex. Note the increased density of labeled fibers in layer I to superficial layer III, and the distinctive band of immunoreactive fibers in the deeper portion of layer V. Roman numerals indicate the cortical layers. WM, white matter. Scale bar = 200 l~m.
Choline acetyltransferase in monkey frontal cortex
371
neuropeptide-containing GABAergic interneurons ences in receptor density across any of the deeper located in the superficial cortical layers ts may also cortical layers. In addition, in prefrontal regions, receive cholinergic input. Finally, in primary sensory there were no laminar differences in the density of areas, cholinergic axons may also selectively innerv- total muscarinic receptors, although the laminar patate neurons in layer IV that receive sensory-specific tern of M2 receptors, especially in area 9, did parallel thalamic input, whereas in motor regions, cholinergic the distribution of ChAT-positive axon terminals. afferents may participate in the regulation of subcor- However, it should also be noted that there is appartically projecting pyramidal cells located in layer V. ~7 ent controversy in the literature regarding the distriComparison of the findings of the present study bution of cholinergic receptors in macaque cortex. with the laminar distribution of muscarinic cholin- For example, Lidow e t al. 28 found that the overall ergic binding sites28 reveals only partial agreement. In density of M2 receptors was similar throughout the primary motor and primary sensory cortices, the cerebral cortex, whereas Mash e t al. 32 found higher density of both M 1 and M2 receptors was greatest in densities of M2 receptors within primary sensory the superficial layers; however, there were no differ- areas. It is possible that knowledge of the laminar
Fig. 8. Dark-field photomicrograph of ChAT immunoreactivity on the orbital surface (area 11) of cynomolgus monkey prefrontal cortex. Roman numerals indicate the cortical layers. WM, white matter. Scale bar = 200 #m.
372
D.A. LEWIS
Fig. 9. Bright-field photomicrograph of ChAT-positive fibers in layer V of primary motor cortex (area 4). Note the high density of immunoreactive fibers in the portion of layer V containing the unlabeled Betz cells (arrows). Scale bar = 100 pro. distribution of nicotinic cholinergic receptors, in concert with that of muscarinic receptors, would provide a more complete picture of the cholinergic innervation of monkey neocortex that would more closely parallel the distribution of ChAT-positive axons. Alternatively, the apparent mismatch between receptor distribution and axon terminals may be related to methodological differences or other factors (see Ref. 13 for review). Comparison to other non-thalamic projections to frontal cortex The distinctive pattern of distribution, and presumably unique functional role, of the cholinergic projections to monkey frontal cortex may be demonstrated by comparison of this afferent system with the organization of other chemically identified afferent systems. In contrast to the relatively uniform regional distribution of cholinergic axons in prefrontal regions, both dopaminergic5'11'26 and noradrenergic27 axons are heterogeneous in density across the same areas. All three systems do show a rostral to caudal gradient of increasing fiber density through the medial frontal cortex, although it is more marked for the dopaminergic and noradrenergic systems than for the cholinergic system. 25'27'39These findings suggest that dopaminergic, noradrenergic and cholinergic axons may all have substantial influence on the activity of primary motor regions of cortex.
On a laminar basis, both dopaminergic and chotinergic axons have their highest densities in the superficial cortical layers, whereas noradrenergic fibers are relatively sparse in these layers. Studies in rodent frontal cortex have suggested that dopaminergic terminals possess muscarinic receptors that can influence dopamine release. 3°'3~ The parallel laminar distribution of dopaminergic and cholinergic axons may provide a basis for a similar interaction of these two systems in the superficial layers of monkey frontal cortex. Noradrenergic axons are most dense in the infragranular layers, especially layer V. Cholinergic and dopaminergic axons also have a relatively high density in this layer, especially in agranular regions. Thus, these three systems may share the same synaptic targets in layer V in at least some areas of frontal cortex. Interestingly, all three afferent systems have a low density of axons in deep layer II! to layer IV, suggesting that they may play a relatively minor role in regulating the activity of the neurons in prefrontal cortex that receive thalamic input) ° Acknowledgements--I would like to thank Dr Bruce Wainer for the generous provision of the anti-ChAT antibody, Sharon Slovenecfor excellenttechnical assistance and Betty Hays for preparing the manuscript. This research was supported by NIMH Research Scientist Development Award MH00519, NIMH grant MH43784 and the Scottish Rite Schizophrenia Research Program.
REFERENCES
I. Amaral D. G. and Bassett J. L. (1989) Cholinergic innervation of the monkey amygdala: an immunohistochemical analysis with antisera to choline acetyltransferase. J. comp. Neurol. 281, 337-361.
Choline acetyltransferase in monkey frontal cortex
373
2. Barbas H. and Pandya D. N. (1987) Architecture and frontal cortical connections of the premotor cortex (area 6) in the rhesus monkey. J. comp. Neurol. 256, 211-228. 3. Berger B., Tassin J. P., Blanc G., Moyne M. A. and Thierry A. M. (1974) Histochemical confirmation for dopaminergic innervation of the rat cerebral cortex after destruction of the noradrenergic ascending pathways. Brain Res. 81, 332-337. 4. Berger B., Thierry A. M., Tassin J. P. and Moyne M. A. (1976) Dopaminergic innervation of the rat prefrontal cortex: a fluorescence histochemical study. Brain Res. 106, 133-145. 5. Berger B., Trottier S., Verney C., Gaspar P. and Alvarez C. (1988) Regional and laminar distribution of the dopamine and serotonin innervation in the macaque cerebral cortex: a radioautographic study. J. comp. Neurol. 273, 99-119. 6. Campbell M. J., Lewis D. A., Foote S. L. and Morrison J. H. (1987) The distribution of choline acetyltransferase-, serotonin-, dopamine-beta-hydroxylase-, and tyrosine hydroxylase-immunoreactive fibers in monkey primary auditory cortex. J. comp. Neurol. 261, 209-220. 7. DeKosky S. T., Scheff S. W. and Markesbery W. R. (1985) Laminar organization of cholinergic circuits in human frontal cortex in Alzheimer's disease and aging. Neurology 35, 1425-1431. 8. Eckenstein F. and Sofroniew M. V. (1983) Identification of central cholinergic neurons containing both choline acetyltransferase and acetylcholinesterase and of central neurons containing only acetylcholinesterase. J. Neurosci. 3, 2286-2291. 9. Fuxe K., Hokfelt T., Johansson O., Jonsson G., Lidbrink P. and Ljungdahl A. (1974) The origin of the dopamine nerve terminals in limbic and frontal cortex. Evidence for meso-cortico dopamine neurons. Brain Res. 82, 349-355. 10. Giguere M. and Goldman-Rakic P. S. (1988) Mediodorsal nucleus: areal, laminar and tangential distribution of afferents and efferents in the frontal lobe of rhesus monkeys. J. comp. Neurol. 277, 195-213. 11. Goldman-Rakic P. S., Leranth C., Williams S. M., Mons N. and Geffard M. (1989) Dopamine synaptic complex with pyramidal neurons in primate cerebral cortex. Proc. natn. Acad. Sci. U.S.A. 86, 9015-9019. 12. Hedreen J. C., Uhl G. R., Bacon S. J., Fambrough D. M. and Price D. L. (1984) Acetylcholinesterase-immunoreactive axonal network in monkey visual cortex. J. comp. Neurol. 226, 246-254. 13. Herkenham M. (1987) Mismatches between neurotransmitter and receptor Iocalizations in brain: observations and implications. Neuroscience 23, 1-38. 14. Hokfelt T., Ljungdahl A., Fuxe K. and Johansson O. (1974) Dopamine nerve terminals in the rat limbic cortex: aspects of the dopamine hypothesis of schizophrenia. Science 184, 177-179. 15. Houser C. R., Crawford G. D., Salvaterra P. M. and Vaughn J. E. (1985) Immunocytochemical localization of choline acetyltransferase in rat cerebral cortex: a study of cholinergie neurons and synapses. J. comp. Neurol. 234, 17-34. 16. Iniguez C., Gayoso M. J. and Carreres J. (1985) A versatile and simple method for staining nervous tissue using Giemsa dye. J. Neurosci. Meth. 13, 77-86. 17. Jones E. G. (1984) Laminar distribution of cortical efferent cells. In Cerebral Cortex (eds Peters A. and Jones E. G.), Vol. I, pp. 521-553. Plenum Press, New York. 18. Jones E. G., Hendry S. H. C. and DeFelipe J. (1987) GABA-peptide neurons of the primate cerebral cortex: a limited cell class. In Cerebral Cortex (eds Jones E. G. and Peters A.), Vol. 6, pp. 237-266. Plenum Press, New York. 19. Kitt C. A., Mitchell S. J., DeLong M. R., Wainer B. H. and Price D. L. (1987) Fiber pathways of basal forebrain cholinergic neurons in monkeys. Brain Res. 406, 192-206. 20. Lehmann J. and Fibiger H. C. (1979) Acetylcholinesterase and the cholinergic neuron. Life Sci. 25, 1939-1947. 21. Lehmann J., Sturble R. G., Antuono P. G., Coyle J. T., Cork L. and Price D. L. (1984) Regional heterogeneity of choline acetyltransferase activity in primate neocortex. Brain Res. 322, 361-364. 22. Levey A. I., Armstrong D. M., Atweh S. F., Terry R. D. and Wainer B. H. (1983) Monoclonal antibodies to choline acetyltransferase: production, specificity and immunohistochemistry. J. Neurosci. 3, 1-9. 23. Levey A. I., Wainer B. H., Mufson E. J. and Mesulam M.-M. (1983) Co-localization of acetylcholinesterase and choline acetyltransferase in the rat cerebrum. Neuroscience 9, 9-22. 24. Levitt P. and Moore R. Y. (1978) Noradrenaline neuron innervation of the neocortex in the rat. Brain Res. 139, 219-231. 25. Lewis D. A., Campbell M. J., Foote S. L., Goldstein M. and Morrison J, H. (1987) The distribution of tyrosine hydroxylase-immunoreactive fibers in primate neocortex is widespread but regionally specific. J. Neurosci. 7, 279-290. 26. Lewis D. A., Foote S. L., Goldstein M. and Morrison J. H. (1988) The dopaminergic innervation of monkey prefrontal cortex: a tyrosine hydroxylase immunohistochemical study. Brain Res. 449, 225-243. 27. Lewis D. A. and Morrision J. H. (1989) The noradrenergic innervation of monkey prefrontal cortex: a dopamine-flhydroxylase immunohistochemical study. J. comp. Neurol. 282, 317-330. 28. Lidow M. S., Gallager D. W., Rakic P. and Goldman-Rakic P. S. (1989) Regional differences in the distribution of muscarinic cholinergic receptors in macaque cerebral cortex. J. comp. Neurol. 289, 247-259. 29. Lindvall O., Bjorklund A., Moore R. Y. and Stenevi V. (1974) Mesencephalic dopamine neurons projecting to neocortex. Brain Res. 81, 325-331. 30. Liskowsky D. R. and Potter L. T. (1985) Muscarine-binding sites localized to cortical dopamine terminals. Neurosci. Lett. 58, 229-233. 31. Marchi M. and Raiteri M. (1985) On the presence in the cerebral cortex of muscarinic receptor subtypes which differ in neuronal localization, function and pharmacological properties. J. Pharmac. exp. Ther. 235, 230-233. 32. Mash D. C., White W. F, and Mesulam M.-M. (1988) Distribution of muscarinic receptor subtypes within architectonic subregions of the primate cerebral cortex. J. comp. Neurol. 278, 265-274. 33. Mesulam M.-M. and Geula C. (1988) Nucleus basalis (Ch4) and cortical cholinergic innervation in the human brain: observations based on the distribution of acetylcholinesterase and choline acetyltransferase. J. comp. Neurol. 275, 216-240. 34. Mesulam M.-M., Mufson E. J., Levey A. I. and Wainer B. H. (1983) Cholinergic innervation of cortex by the basal forebrain: cytochemistry and cortical connections of the septal area, diagonal band nuclei, nucleus basalis (substantia innominata), and hypothalamus in the rhesus monkey. J. comp. Neurol. 214, 170-197. 35. Mesulam M.-M., Mufson E. J. and Wainer B. H. (1986) Three-dimensional representation and cortical projection topography of the nucleus basalis (Ch4) in the macaque: concurrent demonstration of choline acetyltransferase and retrograde transport with a stabilized tetramethylbenzidine method for horseradish peroxidase. Brain Res. 367, 301-308. 36. Mesulam M.-M., Rosen A. D. and Mufson E. J. (1984) Regional variations in cortical cholinergic innervations: chemoarchitectonics of aeetylcholinesterase-containing fibers in the macaque brain. Brain Res. 311, 245-258.
374
D.A. LEwis
37. Mesulam M.-M. and Van Hoesen G. W. (1976) Acetylcholinesterase-rich projections from the basal forebrain t~t the rhesus monkey to neocortex. Brain Res. 109, 152 157. 38. Mesulam M.-M., Volicer L., Marquis J. K., Mufson E. J. and Green R. C. (1986) Systematic regional differences in the cholinergic innervation of the primate cerebral cortex: distribution of enzyme activities and some behavioral implications. Ann. Neurol. 19, 144-151. 39. Morrison J. H., Foote S. L., O'Connor D. and Bloom F. E. (1982) Laminar, tangential and regional organization of the noradrenergic innervation of monkey cortex: dopamine-beta-hydroxylase immunohistochemistry. Brain Res. Bull. 9, 309 319. 40. Morrison J. H., Grzanna R., Molliver M. E. and Coyle J. T. (1978) The distribution and orientation of noradrenergic fibers in neocortex of the rat: an immunotluorescence study. J. comp. Neurol. 181, 17-39. 41. Pearson R. C. A., Gatter K. C., Brodal P. and Powell T. P. S. (1983) The projection of the basal nucleus of Meynert upon the neocortex in the monkey. Brain Res. 259, 132-136. 42. Schroder H., Zilles K., Luiten P. G., Strosberg A. D. and Aghchi A. (1989) Human cortical neurons contain both nicotinic and muscarinic acetylcholine receptors: an immunocytochemical double-labeling study. Synapse 4, 319-326. 43. Schroder H., Zilles K., Maelicke A. and Hajos F. (1989) Immunohisto- and cytochemical localization of cortical nicotinic cholinoceptors in rat and man. Brain Res. 502, 287-295. 44. Thierry A. M., Blanc G., Sobel A., Stinus L. and Glowinski J. (1973) Dopaminergic terminals in the rat cortex. Science 182, 499-501. 45. Vogt B. A., Townes-Anderson E. and Burns D. L. (1987) Dissociated cingulate cortical neurons: morphology and muscarinic acetylcholine receptor binding properties. J. Neurosci. 7, 959-971. 46. Walker A. E. (1940) A cytoarchitectural study of the prefrontal area of the macaque monkey. J. comp. NeuroL 73, 59-86. 47. Walker L. C., Kitt C. A., DeLong M. R. and Price D. L. (1985) Noncollateral projections of basal forebrain neurons to frontal and parietal neocortex in primates. Brain Res. Bull. 15, 307-314. (Accepted 7 August 1990)
Neuroscience Vol. 40, No. 2, pp. 363-374, 1991 Printed in Great Britain
DISTRIBUTION OF CHOLINE ACETYLTRANSFERASEIMMUNOREACTIVE AXONS IN MONKEY FRONTAL CORTEX D. A. LEWIS Departments of Psychiatry and Behavioral Neuroscience, University of Pittsburgh, 3811 O'Hara Street, Pittsburgh, PA 15213, U.S.A. Al~tract--Cholinergic neurons of the nucleus basalis of Meynert project to numerous regions of the cerebral cortex. However, little is known about the regional and laminar distributions of cholinergic axons in monkey frontal cortex. In this study, immunohistoehemical techniques were used to identify axons that were immunoreactive for choline acetyltransferase, the enzyme that catalyses the synthesis of acetylcholine, in the frontal cortex of cynomolgus monkeys. Motor cortex contained the greatest density of labeled fibers; the density of labeled fibers was lower in premotor and anterior cingulate cortices and lower still in the association regions of prefrontal cortex. On a laminar basis, choline acetyltransferaseimmunoreactive axons were most dense in layer I to superficiallayer III. Layer V also contained a distinct band of labeled fibers that was particularly prominent in the agranular regions of frontal cortex. The density of labeled fibers was much lower in the deep portion of layer III to layer IV and in layer VI. These findings demonstrate a specific and regionally distinctive cholinergic innervation of monkey frontal cortex that may reveal the anatomical basis for the influence of acetylcholine on the diverse functions of primate frontal cortex.
The frontal lobes in primates contain distinct cortical regions that differ substantially in cytoarchitectonic organization, connectivity and function. The agranular primary motor and premotor regions are located in the most caudal portions of the dorsomedial and lateral surfaces of the frontal lobes. Anterior to these areas, the granular association regions of the prefrontal cortex occupy the lateral, medial and orbital surfaces. In addition, the medial surface of the frontal lobes contains the paralimbic areas of the cingulate cortex. In primates, the cortex of the frontal lobes, especially the prefrontal regions, is proportionally much larger and is substantially more differentiated than in rodents. These changes are associated with marked modifications in the patterns of innervation of frontal cortex by afferent fibers from non-thalamic subcortical nuclei. For example, although dopaminergic axons primarily innervate prefrontal and anterior cingulate cortices in rats, 3'4'9'14'29,44 primary motor cortex contains the greatest density of dopaminergic fibers in monkeys. 5:1'25 In addition, noradrenergic axons have a relatively uniform regional distribution in rodent neocortex, 24,4° whereas in monkey, substantial regional heterogeneity is present in the density of noradrenergic fibers. 27,39 Axons from the cholinergic neurons of the nucleus basalis of Meynert also innervate many, if not all, regions of the primate cerebral cortex, including ACHE, acetylcholinesterase; BSA, bovine serum albumin; CHAT, choline acetyltransferase; DAB, 3,Y-diaminobenzidine; TBS, Tris-buffered saline.
Abbreviations:
those in the frontal lobes. 21,33,36-38 In non-human primates, these projections are topographically organized t9'34'35'41and non-collateralized, at least between frontal and parietal regions.47 Biochemical studies of choline acetyltransferase (CHAT) activity, the enzyme that catalyses the synthesis of acetylcholine, have revealed substantial regional differences in the distribution of this enzyme in monkey cerebral cortex? ~'3s Histochemical investigations of the degradative enzyme acetylcholinesterase (ACHE) produced similar observations in both monkeys and humans, 12"33,36 suggesting that the density of cholinergic innervation was greatest in limbic and paralimbic regions, intermediate in primary motor and primary sensory regions and least in association areas of the cortex. However, interpretation of the histochemical studies has been hindered by questions regarding the specificity of AChE as a marker of cholinergic axons, s,2°,23 The production of specific antibodies directed against ChAT has made it possible to selectively identify cholinergic neurons and axons. For example, Campbell e t al. 6 demonstrated a distinctive laminar organization of the ChAT-immunoreactive, cholinergic afferents to monkey primary auditory cortex. In this study, immunohistochemical techniques were used to determine the regional and laminar distributions of ChAT-positive fibers in the frontal cortex of Old World cynomolgus monkeys. These investigations were conducted in order to determine whether the regional distribution of ChAT-immunoreactive axons in frontal cortex was consistent with the functional organization of the cholinergic system proposed in previous AChE histochemical studies, 36 and
363
364
D.A. LEWIS
if the laminar pattern o f cholinergic innervation was similar or c o m p l e m e n t a r y to that o f projections from other subcortical non-thalamic nuclei. EXPERIMENTAL PROCEDURES
Five male (3.2-4.4 kg) cynomolgus monkeys (Maeaca fascicularis) (Hazleton Research Products Inc., Denver, PA) were used in this study. Animals were deeply anesthetized with ketamine hydrochloride (25 mg/kg, intramuscularly) and pentobarbital sodium (10 mg/kg, intraperitoneally). An endotraclieal tube was inserted and mechanical ventilation with oxygen (2 l/min) was initiated. After the chest was opened, 1.5-2.0ml of 1% aqueous sodium nitrite was injected into the left ventricle of the heart. Animals were then perfused transcardially with cold 1% paraformaldehyde in phosphate buffer (0.15 M) for 15-30 s followed by perfusion with cold 4% paraformaldehyde in phosphate buffer. The latter solution was perfused for approximately 9 min at a flow rate of 300-350 ml/min. Immediately following the perfusion, the brain was removed and sliced into blocks, 3-5 mm thick, and placed in cold fixative for an additional 2 h. Tissue blocks were then washed in a series of cold, graded sucrose solutions and sectioned either coronally or sagittally in a cryostat at 40 #m. Tissue sections were incubated with a rat monoclonal antibody (kindly supplied by B. Wainer) directed against CHAT. This antibody (AB8) recognizes ChAT from a number of mammalian species, 22 and has previously been utilized in immunohistochemical studies in monkey brain) '35 The anti-ChAT antibody was diluted 1:500 in phosphatebuffered saline containing 0.5% Triton X-100, 2% bovine serum albumin (BSA) and 20% normal goat serum. The sections were incubated in the primary antiserum for ap-
proximately 48 h at 4~C and then rinsed in three changes of 0.1-M Tris-buffered saline (TBS) for 30 min. Tissue sections were then incubated for 90min in a 1:50 dilution of goat anti-rat IgG secondary antiserum (Sternberger.-Meyer, Jarrettsville, MD) in TBS containing 0.2% Triton X-100, 2% BSA, 20% normal goat serum and 10% normal cynomolgus monkey serum. Following a 30-rain wash in TBS, the sections were placed for 2 h in a 1:50 dilution of rat peroxidase-antiperoxidase (Sternberger--Meyer) in the same solution as the secondary antiserum. Sections were washed and then incubated again in the secondary and tertiary solutions for 60 and 90 rain, respectively. The tissue was then washed successively in TBS and phosphate buffer and incubated in phosphate buffer containing 0.05% 3,Ydiaminobenzidine (DAB) and 0.015% H20 z. The DAB reaction product was intensified by serial immersions of the mounted sections in aqueous solutions of 0.005% osmium tetroxide and 0.5% thiocarbohydrazide. Some sections were counterstained with Giemsa ~6 which intensifies the DAB reaction product and also produces a Nissl stain of nonimmunoreactive cells. In order to assess the specificity of immunohistochemical labeling, the anti-ChAT antibody was preabsorbed with 50 nmol/ml of human placental ChAT (Sigma, St Louis, MO) for 16 h before incubation with tissue sections. In other experiments, the anti-ChAT antibody was omitted. In both conditions, immunoreactivity in the cortex was abolished compared to controls. Regional and laminar patterns of ChAT immunoreacfivity were determined by comparison of immunohistochemical sections with adjacent Nissl-stained sections and by evaluation of the Giemsa-counterstained immunohistochemical sections. The following cortical regions of the frontal lobe were examined (Fig. 1): primary motor cortex (dorsal area 4), premotor cortex (dorsal area 6), anterior
Fig. 1. Map of the cytoarchitectonic areas of the macaque frontal cortex adapted from Walker: 46 This map depicts the general location of each cytoarchitectonic region; the precise locations of regional boundaries differed slightly across animals. Shaded areas represent those cortical regions examined in the present study. Dots indicate the approximate anatomical location of the indicated photomicrograph. The exact cytoarchitectonic area depicted in each photomicrograph was determined from adjacent Nissl sections.
365
Choline acetyltransferase in monkey frontal cortex cingulate cortex (area 24) and prefrontal cortex (areas 8B, 9-12, 25, 46). These cytoarchiteetonic areas were identified according to published criteria. TM The relative density of labeled fibers was determined with the aid of a computer-assisted image analysis system (Southern Micro Instruments, Atlanta, GA). All labeled fibers were counted in a grid (total area 0.01 mm 2) overlaying a bright-field image at a final magnification of x 400. Results were expressed as the mean (+S.E.) number of fibers in five fields for a given cortical region and layer. Differences among means were assessed with one-way analy-
sis of variance. Tukey's method (ct level = 0.05) was used for between-group comparisons. RESULTS
Morphology of choline acetyltransferase-immunoreactive fibers The vast majority of ChAT-immunoreactive fibers were extremely fine and studded with small round or oval varicosities (Fig. 2A-C). Intervaricose segments
t Ip
F . -4
i
-~
~"
'
:
•
,
"¢
*
~)"-
-
°:
,
,
•,'~ /
-
'
"'4
~,ii.?:,: ,, - ,:, ~ ~
',
~
'
~
17. . . .
-
-'~
" ...................
Fig. 2. Bright-field ( A ~ , G) and diffusion interference contrast (D-F) photomicrographs of ChAT-immunoreactive fibers in monkey frontal cortex• The majority of ChAT-positive fibers in layer I to superficial layer III were extremely thin and varicose (A42), whereas a smaller number of fibers with a thicker and smoother morphology and a radial orientation were present in layers II-VI (D-F). Note the bundle of tangentially oriented fibers in the white matter below the anterior cingulate cortex ((3). Scale bar in B = 3 0 # m and applies to A~2; scale bar in E = 50/~m and applies to D-F; scale bar in G = 400#m.
366
D.A. LEWIS
of the axons were very thin, and at times, barely detectable. Other immunoreactive fibers were thicker in diameter and fairly smooth in appearance, although in some cases they were punctuated at fairly regalar intervals by elongated oval swellings (Fig. 2D). Fibers with this morphological appearance and a radial orientation were commonly observed in layers II-VI (Fig. 2D-F). Some of these fibers turned 90 ° in layers I or V to assume a tangential orientation (Fig. 2E), whereas others bifurcated and formed thinner and more varicose fibers (Fig. 3). Labeled
Fig. 3. Diffusion interference contrast photomicrographs of ChAT-immunoreactive axons in the superficial layers of area 11. Arrows in B and C refer to the same structure indicated by the corresponding arrows in A. Note that both radially and tangentially oriented labeled fibers bifurcate to form axons of smaller diameter. Scale bar in A = 50/~m; in B and C=20#m.
fibers with a similar morphology but with a tangential orientation were also observed in layers l 1I, V and the subcortical white matter (Fig. 2G). No CHATpositive neurons were observed in any areas of cynomolgus monkey frontal cortex.
Regional distribution of choline acetyltransJeraseimmunoreactive fibers The most striking feature of the regional distribution of ChAT-immunoreactive fibers in cynomolgus monkey frontal cortex was a rostral to caudal gradient of increasing fiber density. These regional differences in density of innervation were most apparent in the dorsomedial portions of frontal cortex. In sagittal sections through the medial frontal cortex (Fig. 4), fiber density gradually increased from area l0 in the frontal pole (Fig. 5A) through areas 9 and 8B (Fig. 5B) and then exhibited a stepwise increase between area 8B and the agranular area 6 of premotor cortex (Fig. 5C). Fiber density continued to increase through area 6 and reached a maximum in the caudal portions of area 4 (primary motor cortex). This rostral to caudal gradient of increasing fiber density was also apparent in the dorsal bank of the cingulate sulcus (Fig. 4), but was much less obvious through area 24 (anterior cingulate cortex) on the ventral bank of the cingulate sulcus, or across the regions of the dorsolateral prefrontal cortex. These regional differences in the relative density of ChAT-positive fibers were confirmed by quantitative assessments. For example, the mean (_+ S.E.) density (per 0.01 mm 2) of labeled fibers in layer V was 18.0 (_+0.9) in area 10, 28.8 (_+2.6) in area 8B and 39.2 (_+ 1.2) in area 4 (F2,~2= 35.94; P < 0.001). Pairwise comparisons revealed significant (P <0.05) differences in fiber density among all three areas. Despite this pattern of regional heterogeneity in the density of ChAT-positive axons, within the prefrontal cortex there were relatively few differences in overall fiber density either within or across cytoarchitectonic areas at a given rostral caudal level. This regional similarity in fiber density across regions was most apparent across rostral-prefrontat areas. In more caudal regions, for example in coronal sections taken at the mid-point of the principal sulcus, the density of labeled fibers appeared to be slightly lower on the dorsolaterai surface (area 46; Fig. 6) than on the dorsomedial convexity (area 9; Fig. 7) and orbital (area 11; Fig. 8) surface. This apparent differences in density of labeled fibers was due, in part, to a more prominent band of immunoreactive fibers in layer V of area 9 (Fig. 7) than in area 46 (Fig. 6). These distinctive patterns of axon density across frontal regions were present in each of the five monkeys examined in this study.
Laminar distribution of choline acetyltransferaseimmunoreactive fibers In all regions examined, the greatest density of labeled fibers was present in layer I (Figs 6-8). Axons
Choline acetyltransferase in monkey frontal cortex
367
Fig. 4. Low power dark-field photomicrograph of ChAT immunoreactivity in a sagittal section through medial frontal cortex of a cynomolgus monkey. The frontal pole is to the left and the dorsal surface is at the top. ps, principal sulcus; cs, cingulate sulcus. This section is located approximately 4 mm lateral to the midline so that much of it cuts through the fundus of the cingulate sulcus, revealing both the dorsal and ventral banks of this sulcus. Note the relatively low density of ChAT-immunoreactive fibers in the cortex of the frontal pole and orbital surface. On the dorsal surface, fiber density increases in a rostral to caudal direction; note especially the increasing prominence of labeled fibers in layer V. The rostral to caudal gradient of increasing fiber density in layer V is also evident in the dorsal bank of the cingulate sulcus, a-c, location of the corresponding photomicrographs in Fig. 5. in this layer were predominately tangential in orientation. This plexus of fibers extended into layer II and then diminished in density in the superficial portion of layer III. Some oblique and radial fibers intermixed with tangentially oriented fibers in these layers. Fiber density was substantially lower in the deep portion of layer III, layer IV and the superficial most portions of layer V. In these layers, fibers had both radial and oblique orientations. In the middle to deep portions of layer V, there was an increase in fiber density due to the addition of tangentially oriented fibers to the plexus of oblique and radially oriented fibers. This fiber band was most prominent in motor cortex (area 4), especially in the portion of layer V containing Betz cells (Fig. 9). However, ChAT-positive fibers did not form pericellular baskets or other specializations and thus they did not appear, at least at the level of light microscopy, to be preferentially targeting the somata of Betz or other pyramidal cells. In all regions of frontal cortex, fiber density decreased substantially in layer VI. The deepest portion of this layer contained some larger caliber tangentially oriented fibers that increased in number in the white matter just below the cortex. These labeled fibers were frequently observed to sweep into the cortex where they then became radially arrayed. Within the white matter, ChAT-positive fibers were most dense below the anterior cingulate cortex (Fig. 2G). This laminar pattern of ChAT-immunoreactive axons was, in general, found in all cytoarchitectonic areas of monkey frontal cortex. However, in the more
densely innervated premotor and motor regions, some differences across layers were less apparent. DISCUSSION
Regional distribution patterns
Using an antibody that selectively recognizes CHAT, we found that cholinergic axons innervated every region of monkey frontal cortex examined. However, substantial differences were present across some regions in the density of ChAT-immunoreactive axons. For example, labeled fibers exhibited a rostral to caudal gradient of increasing density that was due predominantly to the greater density of fibers in premotor (area 6) and motor (area 4) regions compared to prefrontal regions. In contrast, within the prefrontal cortex, ChAT-positive fibers had a more uniform regional distribution. Only in the more caudal portions of prefrontal cortex did the medial (area 24) and dorsomedial (areas 9, 8B) regions have a slightly greater density of labeled fibers than the lateral (area 46) regions. These regional patterns of cholinergic innervation parallel those described in biochemical assessments of ChAT activity.21,3sIn these studies, ChAT activity was, in general, higher in the posterior frontal cortex than in prefrontal regions, which contained very uniform levels of ChAT activity. The findings of this study also confirm many observations of previous studies that employed AChE histochemistry. Mesulam et al. 36 reported tow and uniform levels of AChE staining across most prefrontal regions of rhesus monkey cortex, with
368
D.A. LEWIS
greater levels in supracallosal portions of area 24 and in areas 4 and 6. They also described a rostral to caudal increase in AChE staining across areas 6 and 4 similar to the gradient formed by ChAT-positive fibers. Based upon the distribution of AChE staining, Mesulam et al. suggested that cholinergic afferents were most dense in limbic and paralimbic areas (excluding the cingulate gyrus), intermediate in primary sensory and motor regions, and least dense in association cortices. 33'36'38 Our findings, in concert with other investigations of ChAT-immunoreactive fibers in monkey amygdala ~ and primary auditory cortex 6 support this general organizational scheme. For example, some portions of the amygdala, such as the magnocellular subdivision of the basal nucleus, contain a very high density of ChAT-immunoreactive fibers and terminals. ~ In primary auditory cortex, the middle cortical layers also receive a dense plexus of ChAT-positive axons. 6 In addition, in the present study the motor regions of the frontal lobe had a greater density of labeled fibers than did the prefrontal association areas. However, the density of ChAT immunoreactivity differs substantially among the var-
ious nuclei of the monkey amygdala, 1and the density of ChAT-positive axons in the caudal, medial portions of prefrontal cortex may approach that in primary visual cortex (unpublished observations). Thus, these comparisons suggest that although the distribution of ChAT-positive fibers provides evidence consistent with the hypothesis that cholinergic afferents preferentially innervate limbic over motor and sensory regions and the latter regions over association areas, there appear be some exceptions to this general pattern. L a m i n a r distribution p a t t e r n s
Mesulam et al. 36 also reported that AChE-stained fibers were predominately present in layers V and VI in motor and premotor regions of monkey cortex, whereas in prefrontal cortex, lower numbers of labeled fibers were found in layers I and V V I . However, they noted that although some AChE reaction product was clearly in the form of fibers, more frequently it was found 'fin the form of clumps which lacked discernible internal structure". In contrast, ChAT-labeled fibers were very distinctive
Fig. 5. Dark-field photomicrographs of ChAT-immunoreactive fibers in areas 10 (A), 8B (B) and 6 (C) of cynomolgus monkey frontal cortex. Note the regional differences in overall fiber density and in the prominence of the band of labeled fibers in layer V (arrows). Scale bar in C = 400 ~m and applies to A C.
Choline acetyltransferase in monkey frontal cortex morphologically, and they had precise laminar patterns of distribution. In all regions of monkey frontal cortex, ChAT-immunoreactive fibers were most dense in layer I to superficial layer III; this finding is Consistent with the report that ChAT activity levels are highest in the same layers of human prefrontal cortex. 7 Layer V also contained a band of fibers of increased density, although the prominence of this band exhibited regional specificity that was correlated with certain cytoarchitectonic characteristics. In general, regions with a well-developed layer IV (e.g. areas 10, 46) had only a mild increase in fiber density in layer V, whereas agranular areas, such as motor, premotor and anterior cingulate cortices, had a much more distinctive band (see Fig. 4). These differences were also related to the size of layer V pyramidal
369
neurons. For example, among agranular cortices, the prominence of labeled fibers in layer V increased from anterior cingulate cortex to premotor cortex to primary motor cortex in a manner similar to the increase in size of layer V pyramidal cells across these regions. In addition, in primary motor cortex, the density of ChAT fibers was greatest in the portion of layer V containing Betz cells. However, at the light microscopic level, ChAT-positive axons did not appear to form pericellular baskets or other specializations which might have suggested that they were selectively innervating pyramidal cell somata. These laminar patterns are of interest in comparison to the distribution of ChAT-immunoreactive fibers in monkey primary auditory cortex. 6 In this region, cholinergic axons were very dense in layer I
Fig. 6. Dark-field photomicrograph of ChAT immunoreactivity (A) on the ventral bank of the principal sulcus (area 46) of cynomolgus monkey prefrontal cortex, and a bright-field photomicrograph of an adjacent Cresyl Violet-stained section (B). Roman numerals indicate the cortical layers. WM, white matter. Scale bar in A = 200/~m and applies to A and B.
370
D.A. LEwis
and deep layer III to layer IV, but not in layer V. Although this study employed a different anti-ChAT antiserum, direct comparisons of that antiserum and the one used in the present study have produced similar results in monkey amygdala. 1 Similarly, in monkey primary visual cortex ChAT-positive fibers also preferentially project to layers I and IV (unpublished observations). These comparisons suggest that cholinergic afferents may have both common and
regionally distinctive influences on cortical function. In all cortical regions, the synaptic targets of cholinergic axons may be the distal apical dendrites of pyramidal neurons. Indeed, in rodent cerebral cortex ChAT-positive terminals form synaptic contacts, predominately symmetric, with apical dendrites of pyramidal neurons, 15 and in both rodent and human neocortex apical dendrites have been reported to contain cholinergic receptors. 42'43'45 In addition, the
Fig. 7. Dark-field photomicrograph of ChAT immunoreactivity in the dorsal bank of the cingulate sulcus (area 9) of cynomolgus monkey prefrontal cortex. Note the increased density of labeled fibers in layer I to superficial layer III, and the distinctive band of immunoreactive fibers in the deeper portion of layer V. Roman numerals indicate the cortical layers. WM, white matter. Scale bar = 200 l~m.
Choline acetyltransferase in monkey frontal cortex
371
neuropeptide-containing GABAergic interneurons ences in receptor density across any of the deeper located in the superficial cortical layers ts may also cortical layers. In addition, in prefrontal regions, receive cholinergic input. Finally, in primary sensory there were no laminar differences in the density of areas, cholinergic axons may also selectively innerv- total muscarinic receptors, although the laminar patate neurons in layer IV that receive sensory-specific tern of M2 receptors, especially in area 9, did parallel thalamic input, whereas in motor regions, cholinergic the distribution of ChAT-positive axon terminals. afferents may participate in the regulation of subcor- However, it should also be noted that there is appartically projecting pyramidal cells located in layer V. ~7 ent controversy in the literature regarding the distriComparison of the findings of the present study bution of cholinergic receptors in macaque cortex. with the laminar distribution of muscarinic cholin- For example, Lidow e t al. 28 found that the overall ergic binding sites28 reveals only partial agreement. In density of M2 receptors was similar throughout the primary motor and primary sensory cortices, the cerebral cortex, whereas Mash e t al. 32 found higher density of both M 1 and M2 receptors was greatest in densities of M2 receptors within primary sensory the superficial layers; however, there were no differ- areas. It is possible that knowledge of the laminar
Fig. 8. Dark-field photomicrograph of ChAT immunoreactivity on the orbital surface (area 11) of cynomolgus monkey prefrontal cortex. Roman numerals indicate the cortical layers. WM, white matter. Scale bar = 200 #m.
372
D.A. LEWIS
Fig. 9. Bright-field photomicrograph of ChAT-positive fibers in layer V of primary motor cortex (area 4). Note the high density of immunoreactive fibers in the portion of layer V containing the unlabeled Betz cells (arrows). Scale bar = 100 pro. distribution of nicotinic cholinergic receptors, in concert with that of muscarinic receptors, would provide a more complete picture of the cholinergic innervation of monkey neocortex that would more closely parallel the distribution of ChAT-positive axons. Alternatively, the apparent mismatch between receptor distribution and axon terminals may be related to methodological differences or other factors (see Ref. 13 for review). Comparison to other non-thalamic projections to frontal cortex The distinctive pattern of distribution, and presumably unique functional role, of the cholinergic projections to monkey frontal cortex may be demonstrated by comparison of this afferent system with the organization of other chemically identified afferent systems. In contrast to the relatively uniform regional distribution of cholinergic axons in prefrontal regions, both dopaminergic5'11'26 and noradrenergic27 axons are heterogeneous in density across the same areas. All three systems do show a rostral to caudal gradient of increasing fiber density through the medial frontal cortex, although it is more marked for the dopaminergic and noradrenergic systems than for the cholinergic system. 25'27'39These findings suggest that dopaminergic, noradrenergic and cholinergic axons may all have substantial influence on the activity of primary motor regions of cortex.
On a laminar basis, both dopaminergic and chotinergic axons have their highest densities in the superficial cortical layers, whereas noradrenergic fibers are relatively sparse in these layers. Studies in rodent frontal cortex have suggested that dopaminergic terminals possess muscarinic receptors that can influence dopamine release. 3°'3~ The parallel laminar distribution of dopaminergic and cholinergic axons may provide a basis for a similar interaction of these two systems in the superficial layers of monkey frontal cortex. Noradrenergic axons are most dense in the infragranular layers, especially layer V. Cholinergic and dopaminergic axons also have a relatively high density in this layer, especially in agranular regions. Thus, these three systems may share the same synaptic targets in layer V in at least some areas of frontal cortex. Interestingly, all three afferent systems have a low density of axons in deep layer II! to layer IV, suggesting that they may play a relatively minor role in regulating the activity of the neurons in prefrontal cortex that receive thalamic input) ° Acknowledgements--I would like to thank Dr Bruce Wainer for the generous provision of the anti-ChAT antibody, Sharon Slovenecfor excellenttechnical assistance and Betty Hays for preparing the manuscript. This research was supported by NIMH Research Scientist Development Award MH00519, NIMH grant MH43784 and the Scottish Rite Schizophrenia Research Program.
REFERENCES
I. Amaral D. G. and Bassett J. L. (1989) Cholinergic innervation of the monkey amygdala: an immunohistochemical analysis with antisera to choline acetyltransferase. J. comp. Neurol. 281, 337-361.
Choline acetyltransferase in monkey frontal cortex
373
2. Barbas H. and Pandya D. N. (1987) Architecture and frontal cortical connections of the premotor cortex (area 6) in the rhesus monkey. J. comp. Neurol. 256, 211-228. 3. Berger B., Tassin J. P., Blanc G., Moyne M. A. and Thierry A. M. (1974) Histochemical confirmation for dopaminergic innervation of the rat cerebral cortex after destruction of the noradrenergic ascending pathways. Brain Res. 81, 332-337. 4. Berger B., Thierry A. M., Tassin J. P. and Moyne M. A. (1976) Dopaminergic innervation of the rat prefrontal cortex: a fluorescence histochemical study. Brain Res. 106, 133-145. 5. Berger B., Trottier S., Verney C., Gaspar P. and Alvarez C. (1988) Regional and laminar distribution of the dopamine and serotonin innervation in the macaque cerebral cortex: a radioautographic study. J. comp. Neurol. 273, 99-119. 6. Campbell M. J., Lewis D. A., Foote S. L. and Morrison J. H. (1987) The distribution of choline acetyltransferase-, serotonin-, dopamine-beta-hydroxylase-, and tyrosine hydroxylase-immunoreactive fibers in monkey primary auditory cortex. J. comp. Neurol. 261, 209-220. 7. DeKosky S. T., Scheff S. W. and Markesbery W. R. (1985) Laminar organization of cholinergic circuits in human frontal cortex in Alzheimer's disease and aging. Neurology 35, 1425-1431. 8. Eckenstein F. and Sofroniew M. V. (1983) Identification of central cholinergic neurons containing both choline acetyltransferase and acetylcholinesterase and of central neurons containing only acetylcholinesterase. J. Neurosci. 3, 2286-2291. 9. Fuxe K., Hokfelt T., Johansson O., Jonsson G., Lidbrink P. and Ljungdahl A. (1974) The origin of the dopamine nerve terminals in limbic and frontal cortex. Evidence for meso-cortico dopamine neurons. Brain Res. 82, 349-355. 10. Giguere M. and Goldman-Rakic P. S. (1988) Mediodorsal nucleus: areal, laminar and tangential distribution of afferents and efferents in the frontal lobe of rhesus monkeys. J. comp. Neurol. 277, 195-213. 11. Goldman-Rakic P. S., Leranth C., Williams S. M., Mons N. and Geffard M. (1989) Dopamine synaptic complex with pyramidal neurons in primate cerebral cortex. Proc. natn. Acad. Sci. U.S.A. 86, 9015-9019. 12. Hedreen J. C., Uhl G. R., Bacon S. J., Fambrough D. M. and Price D. L. (1984) Acetylcholinesterase-immunoreactive axonal network in monkey visual cortex. J. comp. Neurol. 226, 246-254. 13. Herkenham M. (1987) Mismatches between neurotransmitter and receptor Iocalizations in brain: observations and implications. Neuroscience 23, 1-38. 14. Hokfelt T., Ljungdahl A., Fuxe K. and Johansson O. (1974) Dopamine nerve terminals in the rat limbic cortex: aspects of the dopamine hypothesis of schizophrenia. Science 184, 177-179. 15. Houser C. R., Crawford G. D., Salvaterra P. M. and Vaughn J. E. (1985) Immunocytochemical localization of choline acetyltransferase in rat cerebral cortex: a study of cholinergie neurons and synapses. J. comp. Neurol. 234, 17-34. 16. Iniguez C., Gayoso M. J. and Carreres J. (1985) A versatile and simple method for staining nervous tissue using Giemsa dye. J. Neurosci. Meth. 13, 77-86. 17. Jones E. G. (1984) Laminar distribution of cortical efferent cells. In Cerebral Cortex (eds Peters A. and Jones E. G.), Vol. I, pp. 521-553. Plenum Press, New York. 18. Jones E. G., Hendry S. H. C. and DeFelipe J. (1987) GABA-peptide neurons of the primate cerebral cortex: a limited cell class. In Cerebral Cortex (eds Jones E. G. and Peters A.), Vol. 6, pp. 237-266. Plenum Press, New York. 19. Kitt C. A., Mitchell S. J., DeLong M. R., Wainer B. H. and Price D. L. (1987) Fiber pathways of basal forebrain cholinergic neurons in monkeys. Brain Res. 406, 192-206. 20. Lehmann J. and Fibiger H. C. (1979) Acetylcholinesterase and the cholinergic neuron. Life Sci. 25, 1939-1947. 21. Lehmann J., Sturble R. G., Antuono P. G., Coyle J. T., Cork L. and Price D. L. (1984) Regional heterogeneity of choline acetyltransferase activity in primate neocortex. Brain Res. 322, 361-364. 22. Levey A. I., Armstrong D. M., Atweh S. F., Terry R. D. and Wainer B. H. (1983) Monoclonal antibodies to choline acetyltransferase: production, specificity and immunohistochemistry. J. Neurosci. 3, 1-9. 23. Levey A. I., Wainer B. H., Mufson E. J. and Mesulam M.-M. (1983) Co-localization of acetylcholinesterase and choline acetyltransferase in the rat cerebrum. Neuroscience 9, 9-22. 24. Levitt P. and Moore R. Y. (1978) Noradrenaline neuron innervation of the neocortex in the rat. Brain Res. 139, 219-231. 25. Lewis D. A., Campbell M. J., Foote S. L., Goldstein M. and Morrison J, H. (1987) The distribution of tyrosine hydroxylase-immunoreactive fibers in primate neocortex is widespread but regionally specific. J. Neurosci. 7, 279-290. 26. Lewis D. A., Foote S. L., Goldstein M. and Morrison J. H. (1988) The dopaminergic innervation of monkey prefrontal cortex: a tyrosine hydroxylase immunohistochemical study. Brain Res. 449, 225-243. 27. Lewis D. A. and Morrision J. H. (1989) The noradrenergic innervation of monkey prefrontal cortex: a dopamine-flhydroxylase immunohistochemical study. J. comp. Neurol. 282, 317-330. 28. Lidow M. S., Gallager D. W., Rakic P. and Goldman-Rakic P. S. (1989) Regional differences in the distribution of muscarinic cholinergic receptors in macaque cerebral cortex. J. comp. Neurol. 289, 247-259. 29. Lindvall O., Bjorklund A., Moore R. Y. and Stenevi V. (1974) Mesencephalic dopamine neurons projecting to neocortex. Brain Res. 81, 325-331. 30. Liskowsky D. R. and Potter L. T. (1985) Muscarine-binding sites localized to cortical dopamine terminals. Neurosci. Lett. 58, 229-233. 31. Marchi M. and Raiteri M. (1985) On the presence in the cerebral cortex of muscarinic receptor subtypes which differ in neuronal localization, function and pharmacological properties. J. Pharmac. exp. Ther. 235, 230-233. 32. Mash D. C., White W. F, and Mesulam M.-M. (1988) Distribution of muscarinic receptor subtypes within architectonic subregions of the primate cerebral cortex. J. comp. Neurol. 278, 265-274. 33. Mesulam M.-M. and Geula C. (1988) Nucleus basalis (Ch4) and cortical cholinergic innervation in the human brain: observations based on the distribution of acetylcholinesterase and choline acetyltransferase. J. comp. Neurol. 275, 216-240. 34. Mesulam M.-M., Mufson E. J., Levey A. I. and Wainer B. H. (1983) Cholinergic innervation of cortex by the basal forebrain: cytochemistry and cortical connections of the septal area, diagonal band nuclei, nucleus basalis (substantia innominata), and hypothalamus in the rhesus monkey. J. comp. Neurol. 214, 170-197. 35. Mesulam M.-M., Mufson E. J. and Wainer B. H. (1986) Three-dimensional representation and cortical projection topography of the nucleus basalis (Ch4) in the macaque: concurrent demonstration of choline acetyltransferase and retrograde transport with a stabilized tetramethylbenzidine method for horseradish peroxidase. Brain Res. 367, 301-308. 36. Mesulam M.-M., Rosen A. D. and Mufson E. J. (1984) Regional variations in cortical cholinergic innervations: chemoarchitectonics of aeetylcholinesterase-containing fibers in the macaque brain. Brain Res. 311, 245-258.
374
D.A. LEwis
37. Mesulam M.-M. and Van Hoesen G. W. (1976) Acetylcholinesterase-rich projections from the basal forebrain t~t the rhesus monkey to neocortex. Brain Res. 109, 152 157. 38. Mesulam M.-M., Volicer L., Marquis J. K., Mufson E. J. and Green R. C. (1986) Systematic regional differences in the cholinergic innervation of the primate cerebral cortex: distribution of enzyme activities and some behavioral implications. Ann. Neurol. 19, 144-151. 39. Morrison J. H., Foote S. L., O'Connor D. and Bloom F. E. (1982) Laminar, tangential and regional organization of the noradrenergic innervation of monkey cortex: dopamine-beta-hydroxylase immunohistochemistry. Brain Res. Bull. 9, 309 319. 40. Morrison J. H., Grzanna R., Molliver M. E. and Coyle J. T. (1978) The distribution and orientation of noradrenergic fibers in neocortex of the rat: an immunotluorescence study. J. comp. Neurol. 181, 17-39. 41. Pearson R. C. A., Gatter K. C., Brodal P. and Powell T. P. S. (1983) The projection of the basal nucleus of Meynert upon the neocortex in the monkey. Brain Res. 259, 132-136. 42. Schroder H., Zilles K., Luiten P. G., Strosberg A. D. and Aghchi A. (1989) Human cortical neurons contain both nicotinic and muscarinic acetylcholine receptors: an immunocytochemical double-labeling study. Synapse 4, 319-326. 43. Schroder H., Zilles K., Maelicke A. and Hajos F. (1989) Immunohisto- and cytochemical localization of cortical nicotinic cholinoceptors in rat and man. Brain Res. 502, 287-295. 44. Thierry A. M., Blanc G., Sobel A., Stinus L. and Glowinski J. (1973) Dopaminergic terminals in the rat cortex. Science 182, 499-501. 45. Vogt B. A., Townes-Anderson E. and Burns D. L. (1987) Dissociated cingulate cortical neurons: morphology and muscarinic acetylcholine receptor binding properties. J. Neurosci. 7, 959-971. 46. Walker A. E. (1940) A cytoarchitectural study of the prefrontal area of the macaque monkey. J. comp. NeuroL 73, 59-86. 47. Walker L. C., Kitt C. A., DeLong M. R. and Price D. L. (1985) Noncollateral projections of basal forebrain neurons to frontal and parietal neocortex in primates. Brain Res. Bull. 15, 307-314. (Accepted 7 August 1990)