Neuroscience Letters, 145 (1992) 87-92 87 © 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/92/$ 05.00

NSL 08986

Cholecystokinin- and dopamine-containing mesencephalic neurons provide distinct projections to monkey prefrontal cortex

Kristen M. Oeth and David A. Lewis

Departments of Behavioral Neuroscience and Psychiatry, University of Pittsburgh, Pittsburgh, PA 15213 (USA)

(Received 1 May 1992; Revised version received 24 June 1992; Accepted 29 June 1992)

Key words: Neuropeptide; Cynomolgus monkey; Substantia nigra; Ventral mesencephalon; Primate; Retrograde transport; Immunohistochemistry

Retrograde transport and immunohistochemical techniques were utilized to determine if cholecystokinin (CCK)-containing neurons of the primate ventral mesencephalon project to prefrontal cortex, and to examine what relation the CCK innervation of prefrontal cortex bears to the dopaminergic projection to this region. Following injections of Fast blue into monkey prefrontal cortex, retrogradely labeled, CCK-positive neurons were observed predominantly in rostromedial portions of the ventral mesencephalon; these CCK-containing projection neurons were not immunoreactive for tyrosine hydroxylase. Furthermore, dual-labeling studies in the prefrontal cortex revealed that CCK and tyrosine hydroxylase were present in separate populations of axons. These results demonstrate that the CCK innervation of monkey prefrontal cortex arises from both intrinsic and extrinsic sources; in contrast to the rat, the extrinsic CCK innervation of monkey prefrontal cortex is distinct from the dopaminergic mesocortical projection.

In primate neocortex, the neuropeptide cholecystoki- nin (CCK) is present in high concentrations, especially in prefrontal regions [2, 4, 11, 27]. Since CCK is colocalized with GABA in cortical non-pyramidal neurons [8, 23] and these GABA neurons appear to have exclusively in- trinsic axons [10], it has been assumed that all cortical CCK-containing fibers arise from intrinsic CCK-con- taining neurons. However, immunohistochemical studies of the CCK innervation of monkey prefrontal cortex (PFC) have demonstrated that CCK-positive cells are present in greatest density in layers II-superficial III, whereas CCK-containing terminal fields are found in layers II, IV and VI [15]. Since the density of labeled fibers in layer VI far exceed that of CCK-immunoreac- tive (IR) neurons in that layer, it seems unlikely that local non-pyramidal cells are the source of the layer VI termi- nal field. In addition, Golgi studies of monkey PFC have failed to identify any non-pyramidal neurons lo- cated in layers II-superficial III with axons that descend to layer VI [14]. Thus, we hypothesized that an extrinsic source might exist for the CCK-IR layer VI terminal field present in primate PFC.

CCK-containing cells of the ventral mesencephalon

Correspondence: D.A. Lewis, Department of Psychiatry, University of Pittsburgh, W1652 BST, 3811 O'Hara Street, Pittsburgh, PA 15213, USA.

(VMC) were considered to be a potential source of the layer VI CCK-IR terminal field in monkey PFC for sev- eral reasons. First, in the rat, CCK-containing neurons in the VMC project to PFC [22]. Although non-dopa- minergic VMC neurons project to rat PFC [26], the vast majority of the CCK-IR neurons that project to PFC contain tyrosine hydroxylase (TH) [21]. Thus, CCK ap- pears to be colocalized with dopamine (DA) in mesencephalic neurons that project to rodent PFC [21]. Second, immunohistochemical investigations of the DA and CCK innervations of monkey PFC have shown that both systems exhibit a high density of labeled fibers in layer VI [l, 5, 12, 15], suggesting that CCK might be present in DA axons. However, although neurons ex- pressing CCK mRNA are known to be present in mon- key VMC [20], it is not known if the CCK mRNA-posi- tive cells of the VMC project to PFC or if these neurons also contain TH. In this study, we used retrograde trans- port and dual-labeling immunohistochemical techniques in order to determine if CCK-containing neurons of the monkey VMC project to PFC, and if they do, to examine what relation this CCK projection has to the DA inner- vation of PFC.

Five cynomolgus (Macacafascicularis) monkeys (3.0- 5.0 kg) were deeply anesthetized and multiple injections of the fluorescent retrograde tracer Fast blue (Sigma; 5% aqueous) were placed into the PFC of one or both hemi-

12

88

Fig. 1. The distribution of CCK- and TH-containing neurons in the rostral VMC. A: photomicrograph of CCK-IR neurons surrounding the posterior mammillary body. B: section adjacent to (A), labeled with anti-TH antibody. Note the limited number of TH-IR cells located medial and lateral to the posterior mammillary body. C: drawing of Nissl section adjacent to (B). Dashed box indicates the area photographed in (A) and (B). The arrow denotes the midline. D: photomicrograph of CCK-IR neurons 400/~m caudal to area shown in (A C). E: photomicrograph from section adjacent to (D), labeled with anti-TH antibody. Asterisks denote common blood vessels in (D) and (E) and arrowheads point to the same regions of these sections. Note the absence of TH-IR neurons in regions containing abundant CCK-IR cell bodies (arrowheads). F: drawing of Nissl section adjacent to (E); dashed box indicates the area photographed in (D) and (E). Calibration bar (500/lm) applies to A, B, D and E. Cd, caudate nucleus; STN,

subthalamic nucleus; MB, mammillary body; SNpc, substantia nigra, pars compacta.

spheres as previous ly descr ibed [l 5]. In ject ion sites were

loca ted in dorsa l area 9, the banks o f the pr incipal sutcus

(area 46) or the ros t ra l ha l f o f the super ior b ranch o f the

arcuate sulcus (area 8B) [28]. Fo l lowing a survival t ime

o f 15-17 days, 3 monkeys were perfused t ranscard ia l ly

with 4% p a r a f o r m a l d e h y d e as descr ibed previously [15].

The o ther 2 an imals received an inject ion (dorsa l to the

V M C and /o r in t racerebrovent r icu la r ly ) o f colchicine

(Sigma) as descr ibed previous ly [15]; the monkeys were

then mon i to r ed overnight under general anesthes ia and

perfused as above. In addi t ion , 1 cynomolgus m o n k e y

(3.4. kg) and 1 rhesus (Macaca mulatta) m o n k e y (37 days

o f age) were used for cor t ical dua l - labe l ing exper iments ;

these monkeys were anesthet ized and perfused with 4%

pa ra fo rma ldehyde . Af te r perfusion, 3- to 5-mm-thick tis-

sue b locks were taken, c ryopro tec ted and cut on a cry-

os ta t at 40 p m .

Several an t ibodies were used in this s tudy. An t ibod ie s

di rected agains t C C K included a mouse m o n o c l o n a l an-

t ibody which recognizes the te rminal pen tapep t ide o f

C C K (1:1,500-5,000) and a r abb i t an t i se rum di rec ted

agains t the first 27 amino acids o f CCK-33 (1:7,000).

A n t i - T H an t ibodies included a sheep po lyc lona l ant ise-

rum (1:9,000) and 2 mouse monoc lona l an t ibodies

(1: l ,000 a n d 1:20,000), all raised agains t ra t PC12 TH.

The specificity o f all an t ibodies has been previous ly dem-

89

Fig. 2. Fluorescent photomicrographs of a triple-labeled section from rostral VMC. A: retrogradely labeled Fast blue containing neuron. B: the same neuron is immunoreactive for CCK. C: the retrogradely labeled, CCK-positive neuron does not contain TH. Calibration bar equals 50/~m for all

panels.

onstrated [6, 15, 18, 29]. Free-floating adjacent sections were processed for single-labeling experiments according to the avidin-biotin method [9] for TH immunohisto- chemistry and according to the double peroxidase anti- peroxidase method [24] for CCK detection as described previously [12, 15]. Other free-floating sections were processed for double-labeling experiments using an ap- propriate secondary antibody conjugated to fluorescein (1:100-200; Chemicon, Temecula, CA) and a biotinyl- ated secondary antibody (Vector, Burlingame, CA) fol- lowed by a streptavidin-Texas Red conjugate (1:200; Amersham, Arlington Heights, IL) as described previ- ously [7].

In the VMC, the density of CCK-IR cells followed a rostro-caudal gradient such that rostral portions of the VMC contained the highest density of labeled neurons and caudal regions the lowest. At all levels of the VMC, CCK-IR neurons were present predominantly medially. For example, at very rostral levels, CCK-positive neu- rons were observed surrounding the posterior mammil- lary body (Fig. 1A); at a more cuadal level, CCK-IR neurons were present medial to the substantia nigra, pars compacta' (SNpc). The location of CCK-IR neurons was clearly distinct from that of TH-containing neurons in rostral portions of the VMC. For example, a group of CCK-positive neurons was present lateral to the poste- rior mammillary bodies in a region devoid of TH-IR cell bodies (Fig. 1A-C); an additional cluster of CCK-posi- tive neurons was found medial to the rostral SNpc in a region lacking TH-positive somata (Fig. 1 D-F). At more caudal levels of the VMC, CCK-IR cells were much more likely to be located in the same regions as TH- containing neurons. Dual-labeling experiments confirmed that CCK and TH were present in distinct neuronal popula- tions in the rostral VMC; however, TH and CCK were colocalized in neurons of more caudal portions of the VMC.

Following injections of Fast blue into the PFC, retro- gradely labeled neurons were distributed throughout the ventral tegmental area (VTA) and the dorsal tier of the SNpc, consistent with previous reports of the distribu- tion of VMC neurons that project to PFC [17]. Retro- gradely labeled neurons containing TH were present pre- dominantly in the middle and caudal portions of the VMC, particularly in the dorsal tier of the SNpc. These Fast blue/TH-labeled neurons were present throughout the mediolateral extent of the SNpc. in contrast, retro- gradely labeled neurons containing CCK were observed predominantly in the rostral VMC (from the posterior mammillary bodies to the rostral SNpc). These cells were located either near the midline or slightly dorsal to the medial SNpc. Triple-labeling studies confirmed that the retrogradely labeled, CCK-IR cells of the VMC did not contain TH (Fig. 2).

In order to verify the existence of a CCK-positive/TH- negative projection from the VMC to PFC, additional dual-labeling experiments were performed in the PFC. In the cynomolgus monkey, dual-labeling experiments did not detect any fibers containing both CCK and TH. In order to exclude the possibility of a false negative find- ing, experiments were conducted in a 37-day-old rhesus monkey as well. In immature monkeys, the visualization of CCK-containing structures is markedly increased [16] and TH-positive axons are also readily detectable [13]; thus, the utilization of tissue from a young monkey would facilitate the detection of CCK-containing/TH- positive fibers, should they exist. In this animal, both CCK-IR and TH-IR axons were clearly present, but no dual-labeled fibers were observed (Fig. 3), confirming that the CCK and DA projections to PFC arise from separate neuronal populations.

These results demonstrate that the CCK-IR cells pres- ent in the primate VMC exhibit a distinctive pattern of distribution and that they are segregated from the TH-IR

90

J Fig. 3. Fluorescent photomicrographs of dual-label immunohistochemistry in layer VI of prefrontal cortex (37-day-old rhesus monkey). A: TH-IR fibers in layer VI. B: CCK-IR fibers in same field and focal plane as (A). Note the lack of dual-labeled fibers. C: TH-containing fibers in layer VI. D: CCK-IR fibers in same field and focal plane as (C); no dual-labeled fibers are present. Hashmarks denote the layer VI-white matter border.

Calibration bar equals 100/~m for all panels.

neurons of the rostral VTA and SNpc. These findings are consistent with in situ hybridization studies in the mon- key which have shown that C C K mRNA-expressing cells are present medially throughout the VMC [20]. At the level of the posterior mammillary body, CCK m R N A - expressing cells were present in a region in which no TH mRNA-expressing cells were present. In addition, Over- lap in the location of neurons expressing CCK m R N A or T H m R N A was evident more caudally in the VMC.

The findings of the combined retrograde transport and irnmunohistochemical experiments have demonstrated that C C K - I R neurons of the monkey VMC do project to PFC and that these cells do not appear to contain TH. Although it is possible that we failed to detect TH im- munoreactivity in CCK-containing cells of the VMC, or that we were unable to detect C C K immunoreactivity in T H - I R cells of the SNpc, these explanations seem un- likely since our observations were made in colchicine- treated monkeys and they are consistent with studies

using in situ hybridization techniques [20]. Furthermore, this dissociation was also observed in the PFC where C C K - I R fibers were not found to contain TH.

Our findings contrast somewhat with the organization of the rat VMC. In the rat, CCK-positive neurons are observed throughout the VMC; labeled neurons are present in the VTA as well as throughout the medio- lateral extent of the SNpc [22]. Immunohistochemical studies have shown that a high proport ion of these CCK- IR neurons also contain TH [21]. In addition, in situ hybridization experiments in the rat VMC have demon- strated that all CCK mRNA-containing cells also con- tain TH m R N A [19]. In contrast, we observed a much more restricted distribution of CCK- IR cells in the mon- key VMC as well as a prominent dissociation in the pat- terns of organization of CCK- and TH-containing cells. Similar differences between the rat and primate in the chemoanatomy of other VMC projections to cortex have been reported. For example, the neuropeptide neuroten-

91

sin is completely colocalized with TH in neurons of the rat VMC that project to frontal areas [25]. However, dual-labeling studies in human cerebral cortex have dem- onstrated that neurotensin and TH are present in sepa- rate populations of fibers, suggesting that neurotensin and TH are not colocalized in the primate VMC [3].

The finding of a CCK-positive projection from the VMC to the PFC, in concert with previous studies [15], demonstrates that the CCK innervation of primate PFC is a combination of both intrinsic and extrinsic systems. This suggests that the influence of CCK on prefrontal cortical function is more complex than previously thought, particularly given that CCK is expressed within the PFC by neurons with different patterns of axonal arborization [14] and thus, different roles in prefrontal cortical circuitry. In addition, these findings suggest the possibility of unique functional interactions between the CCK-containing and DA systems in primate PFC that may have relevance for the study of human disorders such as schizophrenia.

We thank Drs. J. Walsh and J. Reeve for the generous provision of the anti-CCK antibodies, Drs. J. Haycock, A. Acheson and G. Kapatos for the generous provision of the anti-TH antibodies and M. Brady for excellent photographic assistance. This work was supported by NIMH Research Scientist Development Award MH00519 and NIMH Grant MH45156.

1 Berger, B., Trottier, S., Verney, C., Gaspar, P. and Alvarez, C., Regional and laminar distribution of the dopamine and serotonin innervation in the macaque cerebral cortex: a radioautographic study, J. Comp. Neurol., 273 (1988) 99-119.

2 Emson, EC., Rehfeld, J.F. and Rossor, M.N., Distribution ofchol- ecystokinin-like peptides in the human brain, J. Neurochem., 38 (1982) 1177-1179.

3 Gaspar, E, Berger, B. and Febvret, A., Neurotensin innervation of the human cerebral cortex: lack of colocalization with catecholam- ines, Brain Res., 530 (1990) 181-195.

4 Geola, F.L., Hershman, J.M., Warwick, R., Reeve, J.R., Walsh, J.H. and Tourtellotte, W.W., Regional distribution of cholecystoki- nin-like immunoreactivity in the human brain, J. Clin. Endocrinol. Metab., 53 (1981) 270-275.

5 Goldman-Rakic, ES., Leranth, C., Williams, S.M., Mons, N. and Geffard, M., Dopamine synaptic complex with pyramidal neurons in primate cerebral cortex, Proc. Natl. Acad. Sci. USA, 86 (1989) 9015 9019.

6 Haycock, J.W. and Waymire, J.C., Activating antibodies to tyro- sine hydroxylase, J. Biol. Chem., 257 (1982) 9416-9423.

7 Hayes, T.L. and Lewis, D.A., Non-phosphorylated neurofilament protein and calbindin immunoreactivity in layer III pyramidal neu- rons of human neocortex, Cerebr. Cortex, 2 (1992) 56-67.

8 Hendry, S.H.C., Jones, E.G., DeFelipe, J., Schmechel, D., Bran- don, C. and Emson, EC., Neuropeptide-containing neurons of the cerebral cortex are also GABAergic, Proc. Natl. Acad. Sci. USA, 8 l (1984) 6526-6530.

9 Hsu, S.-M., Raine, L. and Fanger, H., Use of avidin-biotin-peroxi- dase complex (ABC) in immunoperoxidase techniques: a compari- son between ABC and unlabeled antibody (PAP) procedures, J. Histochem. Cytochem., 29 (1981) 577-580.

10 Jones, E.G., Hendry, S.H.C. and DeFelipe, J., GABA-peptide neu- rons of the primate cerebral cortex: a limited cell class. In A. Peters and E.G. Jones (Eds.), Cerebral Cortex, Vol. 6, Plenum, New York, 1987, pp. 237 266.

11 Larsson, L.-I. and Rehfeld, J.F., Localization and molecular heter- ogeneity of cholecystokinin in the central and peripheral nervous system, Brain Res., 165 (1979) 201-218.

12 Lewis, D.A., Foote, S.L., Goldstein, M. and Morrison, J.H., The dopaminergic innervation of monkey prefrontal cortex: a tyrosine hydroxylase immunohistochemical study, Brain Res., 449 (1988) 225 243.

13 Lewis, D.A. and Harris, H.W., Differential laminar distribution of tyrosine hydroxylase-immunoreactive axons in infant and adult monkey prefrontal cortex, Neurosci. Lett., 125 ( 1991) 151 - 154.

14 Lund, J.S. and Lewis, D.A., Local circuit neurons of macaque pre- frontal cortex: Golgi and immunocytochemical characteristics, J. Comp. Neurol., submitted.

15 Oeth, K.M. and Lewis, D.A., Cholecystokinin innervation of mon- key prefrontal cortex: an immunohistochemical study, J. Comp. Neurol., 301 (1990) 123 137.

16 Oeth, K.M. and Lewis, D,A., Postnatal development and sources of cholecystokinin (CCK) immunoreactivity in monkey prefrontal cortex, Soc. Neurosci. Abstr., 16 (1990) 1094.

17 Porrino, L.J. and Goldman-Rakic, P.S., Brainstem innervation of prefrontal and anterior cingulate cortex in the rhesus monkey re- vealed by retrograde transport of HRP, J. Comp. Neurol., 205 (1982) 63-76.

18 Rohrer, H., Acheson, A.L., Thibault, J. and Thoenen, H., Develop- mental potential of quail dorsal root ganglion cells analyzed in vitro and in vivo, J. Neurosci., 6 (1986) 2616-2624.

19 Savasta, M., Ruberte, E., Palacios, J.M. and Mengod, G., The colo- calization of cholecystokinin and tyrosine hydroxylase mRNAs in mesencephalic dopaminergic neurons in the rat brain examined by in situ hybridization, Neuroscience, 29 (1989) 363 369.

20 Schalling, M., Friberg, K., Seroogy, K., Riederer, P., Bird, E., Schiffman, S.N., Mailleux, P., Vanderhaegen, J.-J., Kuga, S., Gold- stein, M., Kitahama, K., Luppi, P.H., Jouvet, M. and Hrkfelt, T., Analysis of expression of cholecystokinin in dopamine ceils in the ventral mesencephalon of several species and in humans with schiz- ophrenia, Proc. Natl. Acad. Sci. USA, 87 (1990) 8427 8431.

21 Seroogy, K.B., Dangaran, K., Lim, S., Haycock, J.W. and Fallon, J.H., Ventral mesencephalic neurons containing both cholecystoki- nin- and tyrosine hydroxylase-like immunoreactivities project to forebrain regions, J. Comp. Neurol., 279 (1989) 397414.

22 Seroogy, K.B. and Fallon, J.H., Forebrain projections from chole- cystokinin-like immunoreactive neurons in the rat midbrain, J. Comp. Neurol., 279 (1989) 415~J,34.

23 Somogyi, P., Hodgson, A.J., Smith, A.D., Nunzi, M.G., Gorio, A. and Wu, J.Y., Different populations of GABAergic neurons in the visual cortex and hippocampus of cat contain somatostatin- or cholecystokinin-immunoreactive material, J. Neurosci., 4 (1984) 2590-2603.

24 Sternberger, L.A., Immunocytochemistry, 3rd edn., Churchill Liv- ingstone, New York, 1986.

25 Studler, J.M., Kitabgi, P., Tramu, G., Herve, D., Glowinski, J. and Tassin, J.P., Extensive colocalization of neurotensin with dopamine in rat meso-cortico-frontal dopaminergic neurons, Neuropeptides, 11 (1988) 95 100.

92

26 Swanson, L.W., The projections of the ventral tegmental area and adjacent regions: a combined fluorescent retrograde tracer and im- munofluorescence study in the rat, Brain Res. Bull., 9 (1982) 321 353.

27 Taquet, H., Javoy-Agid, F., Mauborgne, A., Benoliel, J.J., Agid, Y., Legrand, L.C., Tramu, G., Cesselin, F. and Hamon, M., Bio- chemical mapping of cholecystokinin-, substance P-, [met]enkeph- alin-, [leu]enkephalin-, and dynorphin A (l-8)-like immunoreactivi- ties in the human cerebral cortex, Neuroscience, 27 (1988) 871 883.

28 Walker, A.E., A cytoarchitectural study of the prefrontal area of the macaque monkey, J. Comp. Neurol., 73 (1940) 59-86.

29 Wolf, M.E., LeWitt, P.A., Bannon, M.J., Dragovic, L.J. and Kapa- tos, G., Effect of aging on tyrosine hydroxylase protein content and the number of dopamine nerve terminals in human caudate, J. Neu- rochem., 56 (1991) 1191 2000.