Ultrastructural associations between dopamine terminals and local circuit neurons in the monkey prefrontal cortex: a study of calretinin-immunoreactive cells

ELSEVIER

Neuroscience Letters 200 (1995) 9-12

NEUROSCI[NC[ lETTERS

Ultrastructural associations between dopamine terminals and local circuit neurons in the monkey prefrontal cortex: a study of calretinin-immunoreactive cells Susan R. Sesack a,b,*, Christopher N. Bressler a,b, David A. Lewis .,b aDepartment of Neuroscience, 446 CrawJbrd Hall. University of Pittsburgh, Pittsburgh, PA, 15260, USA bDepartment of Psychiatry, University of Pittsburgh, Pittsburgh, PA, 15260, USA Received 13 June 1995; revised version received 8 September 1995; accepted 13 September 1995

Abstract

Dopamine terminals in the monkey prefrontal cortex (PFC) synaptically target the distal dendrites of both pyramidal cells and GABA intemeurons. We sought to determine whether the latter input includes the innervation of interneurons that utilize calretinin (CalR) as a calcium-binding protein. Sections through prefrontal area 9 of cynomolgus monkeys were processed by immunoperoxidase for tyrosine hydroxylase (TH) to label dopamine varicosities and by pre-embedding immunogold for CalR. Electron microscopic examination of layers 1-3 revealed numerous TH-immunoreactive (TH-ir) terminals, but few were located in the vicinity of CalR-ir dendrites. Although close appositions were sometimes detected between these labeled processes, no synaptic inputs from TH-ir terminals to CalR-ir dendrites were observed. However, in adjacent sections from the same animals, TH-ir terminals were observed to synapse on GABA-ir dendrites. These findings suggest that dopamine afferents to the monkey PFC target the subclasses of GABA interneurons that do not contain CalR. Keywords: Calcium-binding protein; y-Aminobutyric acid (GABA); Tyrosine hydroxylase; Interneuron

Information processing in the neocortex involves extensive interactions between excitatory pyramidal neurons and inhibitory G A B A interneurons [14]. Cortical processing is also regulated by modulatory afferents, such as d o p a m i n e inputs from the ventral midbrain [6,8,15]. The importance of the dopamine innervation of the prefrontal cortex (PFC) is underscored by studies showing that proper functioning of this region depends on an intact d o p a m i n e input [2,10]. Dopamine terminals synapse on the dendritic spines and shafts of pyramidal cells and on the dendrites of presumed local circuit neurons [7,12]. W e have verified and extended these findings [11] by demonstrating that d o p a m i n e terminals synapse on the dendritic shafts of neurons that are immunoreactive for GABA. These findings raise the question of whether dopamine afferents to the PFC provide a uniform synaptic input to all classes of G A B A neurons or whether they selectively * Corresponding author. Tel.: +1 412 6245158; fax: +1 412 6249198; e-maih [email protected].

innervate subpopulations of these cells. Interneuron subclasses can be defined by their morphology, synaptic targets, and differential content of peptides or calciumbinding proteins [4]. For example, in monkey PFC, the calcium-binding proteins calretinin (CalR), calbindin, or parvalbumin are expressed in non-overlapping populations of local circuit neurons [3]. W e sought to determine whether dopamine terminals in the monkey PFC provide direct synaptic input to the dendrites of CalR-containing neurons. Such a relationship seems likely, since CalRpositive cells are prevalent in the superficial layers [3] where dopamine terminal synapses on interneurons are most abundant [ 11,12]. Three adult male cynomolgus (Macaca fascicularis) monkeys were anesthetized and perfused transcardially with 4% paraformaldehyde and 0.2% glutaraldehyde in 0.1 M phosphate buffer at 29°C. Two of these animals were used for the previous study of dopamine terminal synapses on G A B A - i m m u n o r e a c t i v e neurons [11]. A fourth animal was perfused with a cold (4°C) fixative in order to enhance immunostaining; ultrastructural preser-

0304-3940/95/$09.50 © 1995 Elsevier Science Ireland Ltd. All rights reserved SSDI 0304-3940(95) 12076-D

10

S.R. Sesack et al./ Neuroscience Letters 200 (1995) 9-12

Table 1 Relative incidence of associations between TH-ir terminals and dendrites of local circuit neurons immunolabeled for CalR or GABA in the monkey PFC

Number of animals Total area examined (/,tm2) Area examined containing both immunomarkers in the same 32.5/~m 2 field ~ m 2 ) a Total number of TH-labeled terminals analyzed Number of TH-labeled terminals forming associations with dendrites b Number (%) of TH-labeled terminals forming associations with immunolabeled dendrites b Number (%) of TH-labeled terminals forming synapses on immunolabeled dendrites

Local circuit neurons immunolabeled for CalR

Local circuit neurons immunolabeled for GABA

4 363000 3023 118 50 6 (12) 0 (0)

2 129000 2373 88 62 16(26) c 8(13) d

aArea encompassed by an electron micrograph at 12 0 0 0 x magnification. blncludes synapses and close appositions not separated by astrocytic processes. cp = 0.09, Fisher's exact test dp = 0.01, Fisher's exact test

vation of this tissue was compromised relative to the other animals. Coronal sections through Walker's [13] prefrontal area 9 were cut at 50/~m on a vibratome and treated for 30 min with 1% sodium borohydride. Sections were processed for dual immunoperoxidase and pre-embedding immunogold-silver staining as described previously [11]. A mouse monoclonal antibody against tyrosine hydroxylase (TH) was used as a marker for dopamine terminals [11]. Previous studies have demonstrated that this antibody preferentially labels dopamine as opposed to noradrenergic varicosities in cortex (for discussion, see [1,11]). The sections were incubated for 12-15 h in a solution containing 1:8000 mouse anti-TH antibody and 1:1000 rabbit anti-CalR polyclonal antiserum (SWant). TH fibers were visualized by the avidinbiotin-peroxidase method (Vector Labs). CalR-immunoreactive processes were subsequently visualized by incubating sections in secondary IgG conjugated to 1 nm gold particles (Amersham) and enhancing the bound gold by silver solution (Amersham). Control sections were processed by substitution of normal serum for the primary antibody species; no immunoreactivity was detected in these sections by light or electron microscopy. Additional tests for specificity of primary antibodies, including preadsorption controls, immunoprecipitation, and Western blot analyses, have been performed in previous investigations [3,9,16]. As described previously [11], tissue sections were postfixed in 2% osmium tetroxide, dehydrated, and embedded in plastic. Ultrathin sections were cut from layers 1-3 of area 9, collected on copper grids, and counterstained with heavy metals. The tissue surface was scanned randomly in the electron microscope for regions where TH-immunoreactive (TH-ir) and CalR-immunoreactive (CalR-ir) processes were in close proximity. To avoid false negatives due to inadequate penetration of immunoreagents, only micrographs that included both specific immunoperoxidase and immunogold-silver staining

were included for relative quantitative comparisons (see Table 1). At least two vibratome sections and at least 78 6 5 0 ~ m 2 of the PFC were examined for each of the four animals. In the superficial layers of the monkey PFC, peroxidase immunoreactivity for TH was localized to axons and terminal varicosities that contained synaptic vesicles. In many instances, TH-ir terminals either failed to contact dendritic processes in single sections (Fig. 1A), or were found in close apposition to dendrites without exhibiting evidence of synaptic specializations (Fig. 1B,C). Other TH-ir processes formed synapses on unlabeled dendritic spines or shafts. Most of the synapses were of the symmetric type, although a few exhibited asymmetric synaptic densities (Fig. 2). Gold-silver immunoreactivity for CalR was distributed to perikarya, dendrites, and axon terminals. CalR-ir dendrites typically exhibited a varicose morphology, rarely gave rise to spines, and received extensive synaptic input from primarily unlabeled (Fig. 1), but occasionally CalRir terminals. CalR-ir axon terminals formed primarily symmetric synapses on unlabeled dendrites. TH-ir and CalR-ir processes were infrequently detected in the same area of neuropil. When they were found in close proximity, they were typically separated by intervening unlabeled processes (Fig. 1A). Occasionally, TH-ir terminals were closely apposed to CalR-ir dendrites but did not form obvious synaptic specializations at these sites (Fig. 1B,C; Table 1). Extensive examination of tissue from three monkeys revealed no evidence of synaptic input from TH-ir terminals to CalR-ir dendrites. In tissue from an additional monkey perfused with cold fixative to enhance immunostaining, TH-ir terminals were detected more frequently, but they also failed to show synaptic associations with CalR-ir dendrites. TH-ir terminals in the vicinity of CalR-ir dendrites did synapse on unlabeled dendrites that exhibited the varicose morphology and abundant synaptic input characteristic of GABA interneu-

S.R. Sesack et al. /Neuroscience Letters 200 (1995) 9-12

Fig. 1. Electron micrographs of the monkey PFC depicting peroxidase immunoreactivity for TH in axon terminals (TH-T) and immunogoldsilver labeling for CalR in dendrites (CalR-D). (A) TH is localized to an axon with terminal varicosities that form no obvious synaptic contacts in this single section. The TH-T lies close to a proximal CalR-D that receives symmetric synaptic input (straight arrows) from two unlabeled terminals (uT). N, nucleus. (B,C) A TH-T shown in adjacent serial sections is closely apposed (straight arrow) to a CalR-D but forms no obvious synaptic specialization. Two uTs form asymmetric synapses (curved arrows) on the same dendrite. Scale bars, 0.5/tm for (A), 0.25/~m for (B,C). rons (Fig. 2) [5,1 1,12]. Furthermore, in adjacent sections, TH-positive terminals were observed to synapse on G A B A - i r dendrites [ 11 ]. Semi-quantitative analyses confirmed that, compared to TH-ir and CalR-ir processes, TH-ir and G A B A - i r processes were more frequently detected in the same area of neuropil, as evidenced by the smaller total area of tissue needed to yield an equivalent number of micrographs acceptable for analysis (Table 1). TH-ir terminals and G A B A - i r dendrites were also more frequently in close apposition to each other than were TH-ir and CalR-ir processes. Finally, 50% of the associations between TH-ir terminals and G A B A - i r dendrites exhibited synaptic specializations, whereas none of the close appositions between TH-ir and CalR-ir processes were synaptic (Table 1). The combined results of this study and our previous investigation [1 1] indicate that dopamine afferents to the

11

monkey PFC may synaptically target only a subset of G A B A neurons. However, the negative findings with respect to CalR-positive interneurons need to be interpreted cautiously, as limitations in the sensitivity of the method could result in the failure to detect a minor synaptic input to this cell population [11]. Nevertheless, several observations suggest that CalR-containing neurons are rarely, if ever, among the local circuit neurons synaptically targeted by dopamine terminals in the primate PFC. First, the larger area of tissue that had to be scanned in order to detect TH-ir terminals in close proximity to CalR-ir versus G A B A - i r dendrites suggests that CalRcontaining cells are not the preferred interneuron target. Second, the failure to detect synaptic contacts between TH-ir terminals and CalR-ir dendrites even in tissue fixed to promote greater immunostaining suggests that this connection was not missed because of efforts to preserve ultrastructure. Third, the synapses detected between TH-ir terminals and G A B A - i r dendrites in adjacent sections of the same animals indicate that one or more of the nonCalR cell classes may be preferentially innervated by dopamine afferents. The latter conclusion is supported by observations that TH-ir terminals synapse on dendrites that bear morphological features of interneurons [5] and are adjacent to CalR-ir dendrites, but are not labeled. These CalRnegative interneurons are likely to contain one of the other major calcium-binding proteins, calbindin or parvalbumin [3]. Since the synaptic connections of these local circuit neurons (e.g. basket and chandelier cells) include the soma and axon initial segments of pyramidal neurons [14], their regulation by dopamine would provide a powerful, albeit indirect, means for modulating the excitability of pyramidal cells. However, the synaptic relationships between calbindin- or parvalbumin-containing

Fig. 2. Electron micrograph depicting the synaptic relationship of a THT to an unlabeled dendrite (uD) in the vicinity of a CalR-D. The TH-T forms an asymmetric synapse (curved arrow) on an uD that has a varicose shape and receives additional asymmetric (curved arrows) or symmetric (straight arrow) synaptic input from several uTs. The CalRD in the adjacent neuropil receives no synaptic input in this single section. Scale bar, 0.5/,tm.

12

S.R. Sesack et al. / Neuroscience Letters 200 (1995) 9-12

interneurons and d o p a m i n e afferents to the primate PFC r e m a i n to be established. This investigation represents the first d e m o n s t r a t i o n of a specific class o f neurons in the primate P F C that appears not to be targeted by d o p a m i n e afferents. T h e findings suggest that d o p a m i n e does not exert a diffuse and ubiquitous e f f e c t on all cortical cell classes, but rather selectively innervates populations o f neurons that are distinct in their n e u r o c h e m i s t r y , c o n n e c t i v i t y , and function. Further studies to identify w h i c h p y r a m i d a l and interneuron cell classes r e c e i v e d o p a m i n e input will h a v e important implications for u n d e r s t a n d i n g cortical processing and the role o f d o p a m i n e m o d u l a t i o n in c o g n i t i v e function. T h e authors thank Christopher S n y d e r and D a r l e n e M e l c h i t z k y for technical assistance. This w o r k was supported by U S P H S grants M H 5 0 3 1 4 (S.R.S.), M H 4 3 7 8 4 (D.A.L.), Research Scientist Development Award M H 0 0 5 1 9 (D.A.L.), and an N I M H Center for the N e u r o science of Mental Disorders MH45156, [1] Akil, M. and Lewis, D.A., The dopaminergic innervation of monkey entorhinal cortex, Cereb. Cortex, 3 (1993) 533-550. [2] Brozoski, T.J., Brown, R.M., Rosvold, H.E. and Goldman, P.S., Cognitive deficit caused by regional depletion of dopamine in prefrontal cortex of rhesus monkey, Science, 205 (1979) 929-932. [3] Cond6, F., Lund, J.S., Jacobowitz, D.M., Baimbridge, K.G. and Lewis, D.A., Local circuit neurons immunoreactive for calretinin, calbindin D-28k or parvalbumin in monkey prefrontal cortex: distribution and morphology, J. Comp. Neurol., 341 (1994) 95116. [4] DeFelipe, J., Neocortical neuronal diversity: chemical heterogeneity revealed by colocalization studies of classic neurotransmitters, neuropeptides, calcium-binding proteins, and cell surface molecules, Cereb. Cortex, 3 (1993) 273-289. [5] Freund, T.F., Magloczky, Z., Soltesz, I. and Somogyi, P., Synap-

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13] [14]

[15]

[16]

tic connections, axonal and dendritic patterns of neurons immunoreactive for cholecystokinin in the visual cortex of the cat, Neuroscience, 19 (1986) 1133-1159. Gaspar, P., Stepniewska, I. and Kaas, J.H., Topography and collateralization of the dopaminergic projections to motor and lateral prefrontal cortex in owl monkeys, J. Comp. Neurol., 325 (1992) 1-21. Goldman-Rakic, P.S., Leranth, C., Williams, S.M., Mons, N. and Geffard, M., Dopamine synaptie complex with pyramidal neurons in primate cerebral cortex, Proc. Natl. Acad. Sci. USA, 86 (1989) 9015-9019. Lewis, D.A., Foote, S.L., Goldstein, M. and Morrison, J.H., The dopaminergic innervation of the monkey prefrontal cortex: a tyrosine hydroxylase immunohistochemical study, Brain Res., 449 (1988) 225-243. Lewis, D.A., Melchitzky, D.S. and Haycock, J.W., Expression and distribution of two isoforms of tyrosine hydroxylase in monkey brain, Brain Res., 656 (1994) 1-13. Sawaguchi, T. and Goldman-Rakic, P.S., D 1 dopamine receptors in prefrontal cortex: involvement in working memory, Science, 251 (1991) 947-950. Sesack, S.R., Snyder, C.L. and Lewis, D.A., Axon terminals immunolabeled for dopamine or tyrosine hydroxylase synapse on GABA-immunoreactive dendrites in rat and monkey cortex, J. Comp. Neurol., (1995) in press. Smiley, J.F. and Goldman-Rakic, P.S., Heterogeneous targets of dopamine synapses in monkey prefrontal cortex demonstrated by serial section electron microscopy: a laminar analysis using the silver-enhanced diaminobenzidine sulfide (SEDS) immunolabeling technique, Cereb. Cortex, 3 (1993) 223-238. Walker, A.E., A cytoarchitectural study of the prefrontal area of the macaque monkey, J. Comp. Neurol., 73 (1940) 59-86. White, E.L., Cortical Circuits: Synaptic Organization of the Cerebral Cortex - Structure, Function, and Theory, Birkhliuser, Boston, 1989. Williams, S.M. and Goldman-Rakic, P.S., Characterization of the dopaminergic innervation of the primate frontal cortex using a dopamine-specific antibody, Cereb. Cortex, 3 (1993) 199-222. Wolf, M.E., LeWitt, P.A., Bannon, M.J., Dragovic, L.J. and Kapatos, G., Effect of aging on tyrosine hydroxylase protein content and the relative number of dopamine nerve terminals in human caudate, J. Neurochem., 56 (1991) 1191-1200.