hippocampal afferents to the rat prefrontal cortex: synaptic targets and relation to dopamine...

THE .JOURNAL OF COMPARATIVE NEUROLOGY 369:1-15 (1996)

Hippocampal Afferents to the Rat Prefrontal Cortex: Synaptic Targets and Relation to Dopamine Terminals

DAVID B. CARR AND SUSAN R. SESACK Departments of Neuroscience and Psychiatry, University of Pittsburgh,

Pittsburgh, Pennsylvania 15260

ABSTRACT Meren t s to the prefrontal cortex (PFC) from the hippocampal formation and from

midbrain dopamine (DA) neurons have been implicated in the cognitive and adaptive functions of this cortical region. In the present study, we investigated the ultrastructure and synaptic targets of hippocampal terminals, as well as their relation to DA terminals within the PFC of adult rats. Hippocampal afferents were labeled either by anterograde transport of wheat germ agglutinin-horseradish peroxidase (WGA-HRP) from the ventral hippocampal formation or by anterograde degeneration following fimbria lesion. Hippocampal terminals in the PFC, identified by either method, formed primarily asymmetric axospinous synapses, with a small percentage forming asymmetric axodendritic synapses. Dopamine terminals in the PFC were identified by peroxidase immunocytochemistry for either tyrosine hydroxylase or DA and formed primarily symmetric synapses onto dendritic spines and small caliber dendritic shafts. Spines that received symmetric synaptic contact from DA terminals invariably also received an asymmetric synapse from an unlabeled terminal, forming a triadic complex. Hippocampal and DA terminals in the PFC were not often observed in the same area of the neuropil, and no examples of convergence of hippocampal and DA terminals onto common postsynaptic targets were observed. Further analysis revealed that spines receiving synaptic contact from hippocampal terminals did not receive additional synaptic contact from any other source. However, when localized to the same area of the neuropil, hippocampal and DA terminals were often in direct apposition to one another, without forming axo-axonic synapses. These results suggest that 1) hippocampal terminals primarily form excitatory synapses onto spiny pyramidal neurons, 2) hippocampal afferents are unlikely to be synaptically modulated by DA or non-DA terminals at the level of the dendritic spine, and 3) appositions between hippocampal and DA terminals may facilitate presynaptic interactions between these afferents to the PFC. IXifi WiIey-I,iss, Inc.

Indexing terms: prelirnbic cortex, infralirnbic cortex, hippocampus, subiculurn, ultrastructure

The prefrontal cortex (PFC), one of the most highly evolved structures of the human brain, plays an important role in the organization of behavior. The PFC has been implicated in numerous integrative functions such as learning and memory, cognitive functions, and adaptive processes (Rosenkilde, 1983; Kolb, 1984). One afferent pathway to the PFC that may be crucial for the working memory, cognitive, and autonomic functions of this region originates in the hippocampal formation (Ruit and Neafsey, 1988; Laroche et al., 1990). Both the hippocampal formation and the PFC have been demonstrated to be important for working memory (Mahut et al., 1982; Watanabe and Niki, 1985; Zola-Morgan and Squire, 1986; Friedman and Goldman-Rakic, 1988; Fuster, 19911, and dysfunction in both regions has been implicated in the pathophysiology of schizophrenia (Weinberger, 1987).

The hippocampo-prefrontal pathway in the rat arises in the subiculum and the temporal aspect of area CA1 and courses through the lateral fimbria and precommissural fornix, terminating heavily in the prelimbic and infralimbic cortices (Wyss et al., 1980; Swanson, 1981; Jay et al., 1989; van Groen and Wyss, 1990; Jay and Witter, 1991; Jay et al., 1992). Physiological studies indicate that neurons in the hippocampal formation exert a monosynaptic excitatory influence onto pyramidal neurons in the PFC (Ferino et al., 1987; Laroche et al., 1990), consistent with their use of glutamate as a neurotransmitter (Jay et al., 1992).

Accepted December 15, 1995. Address reprint requests to Dr. Susan R. Sesack, Department of Neurosci-

ence, 446 Crawford Hall, University of Pittsburgh, Pittsburgh. PA 15260. E-mail:sesack~~r bns.pitt.edu

C 1996 WILEY-LISS, INC.

2 D.B. CARR AND S.R. SESACK

Despite extensive investigation of the hippocampal innervation of the PFC by electrophysiological and light microscopic studies, the ultrastructural characteristics of hippocampal terminals, their postsynaptic targets, and their relationships with other afferent systems have not been characterized. In the nucleus accumbens, a limbic structure closely related to the PFC, hippocampal afferents have been demonstrated to terminate predominantly on dendritic spines of medium-sized neurons, with a small proportion of terminals forming synapses on dendritic shafts (Totterdell and Smith, 1989; Meredith and Wouterlood, 1990; Sesack and Pickel, 1990). Because the principal cell type in the PFC, the pyramidal cell, also exhibits numerous dendritic spines (for a review, see DeFelipe and Farifias, 1992) on which it receives the majority of its excitatory synaptic input (Colonnier, 1968; Beaulieu and Colonnier, 1985; McGuire et al., 1991), it appears likely that the hippocampal input to the PFC will also terminate at this dendritic level. To address this question, two tracing techniques, anterograde transport of wheat germ agglutinin-horseradish peroxidase (WGA-HRP) and anterograde degeneration following fimbria transection, were employed in the rat to examine the hippocampal afferents to the PFC by electron microscopy.

Within the nucleus accumbens, dendritic spines that receive synaptic input from hippocampal terminals have also been demonstrated to receive synaptic contact from terminals that use dopamine (DA) as a neurotransmitter (Totterdell and Smith, 1989; Sesack and Pickel, 1990). This arrangement of inputs may serve as an anatomical substrate for the observed modulation of hippocampal input to the nucleus accumbens by DA (Yang and Mogenson, 1984, 1986).

The PFC also receives a prominent innervation from DA fibers originating in the ventral tegmental area and ascend- ing through the medial forebrain bundle (Lindvall and Bjorklund, 1984). The mesoprefrontal DA system has been implicated in the integrative functions of the PFC (see Le Moal and Simon, 1991 for a review). More specifically, lesion studies in rats and primates have demonstrated that this afferent system is vital for the proper functioning of the PFC (Brozoski, 1979; Simon, 1980). In addition, the mesoprefrontal DA system has been implicated in the therapeu- tic actions of antipsychotic drugs (Carlsson, 19881, as well as the reinforcing properties of drugs of abuse (Goeders and Smith, 1980,1986).

Whether hippocampal and DA terminals converge onto common targets in the PFC is not known. However, certain pieces of evidence indicate that such a relationship may exist. First, in rats, hippocampal and DA afferents are concentrated in the deep layers of the PFC, and there is substantial overlap between the two afferents (Descarries et al., 1987; Van Eden et al., 1987; van Groen and Wyss, 1990; Jay and Witter, 1991). Second, in both rats and primates, DA terminals have been demonstrated to form symmetric synaptic contacts onto dendritic spines that also receive asymmetric synaptic contact from non-DA terminals, thus forming a triadic complex (Goldman-Rakic et al., 1989; Verney et al., 1990). The terminals that converge with DA afferents onto common spines display a morphology consistent with that of hippocampal terminals that have been characterized in subcortical regions (Leranth and Frotscher, 1989; Totterdell and Smith, 1989; Meredith and Wouterlood, 1990; Sesack and Pickel, 1990). Based on these observations, it appears likely that hippocampal and

DA afferents to the PFC converge on common dendritic spines. To examine this question, a combined anterograde tract-tracing and immunocytochemical approach was employed at the electron microscopic level. The objective was to determine whether an anatomical substrate exists for a possible functional interaction between these two afferent systems. A preliminary report of these findings has already been presented (Carr et al., 1994).

MATERIALS AND METHODS Two tract tracing methods were used to label the projec-

tions from the hippocampal formation to the PFC. Antero- grade transport of WGA-HRP was used to examine the normal morphology of hippocampal terminals and their postsynaptic targets in the PFC. Anterograde degeneration following ablation of the fimbria was also used to examine postsynaptic targets of hippocampal terminals in the PFC. In addition, degeneration was combined with immunocytochemical localization of DA or tyrosine hydroxylase (TH) to examine the relationship of hippocampal and DA terminals within the PFC.

HRP injection and histochemical visualization The procedure for the injection and visualization of

WGA-HRP has been previously reported (Sesack and Pickel, 1990), and tissue examined in this study was taken from the same animals as described in that investigation. Briefly, WGA-HRP was stereotaxically injected into the ventral hippocampal formation of three adult male Sprague-Dawley rats anesthetized with chloral hydrate (420 mgikg, i.p.1. Approximately 30 nl of WGA-HRP (4% in 0.1 M phosphate buffer, pH 7.4) was pressure injected through glass pipettes (tip size 20-30 Fm) over 30 minutes. Injections were centered in the ventral aspect of the subiculum, but the tracer sometimes spread to the ventral portion of the hippocampus proper and a small portion of the adjacent entorhinal cortex (Fig. 1A).

Following a 2 day survival period, subjects were deeply anesthetized with sodium pentobarbital (100 mgikg i.p.1 and perfused through the aortic arch with 10 ml of 0.9% saline containing 1,000 U/ml of heparin, followed by 50 ml of 3.75% acrolein in 2% paraformaldehyde and 200 ml of 2% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4. The brains were removed, and the forebrain and midbrain were cut into 4-5 mm coronal blocks and stored in the last fixative for 30 minutes.

Vibratome sections (40 Fm) cut from these blocks were processed for HRP histochemistry using either 3,3’ - diaminobenzidine (DAB) as the chromogen for the injection sites (LaVail and LaVail, 1974) or 3,3’,5,5’-tetramethylben- zidine (TMB) to visualize anterogradely-transported HRP (Mesulam, 1978). For light microscopic examination, vibratome sections were mounted on slides, dehydrated through successive alcohols, and coverslipped. Sections through the hippocampal formation were counterstained with thionin. For electron microscopic examination, alternate sections through the PFC were incubated first in TMB, then in DAB to stabilize the reaction product (Lemann et al., 1985). Sections were postfixed for 1 hour in 2% osmium tetroxide in 0.1 M phosphate buffer, dehydrated through successive alcohols and propylene oxide, and embedded in epon (EM bed 812, Electron Microscopy Sciences, Fort Washington, PA) (Leranth and Pickel, 1989). Sections were mounted on epon blocks, and ultrathin sections were cut from the outer

HIPPOCAMPAL AFFERENTS TO THE PFC 3

A

B

Fig. 1. Schematic of coronal sections of the rat brain displaying (A) wheat germ agglutin-horseradish peroxidase (WGA-HRP) injection sites, both the largest (hatched area) and smallest (shaded area), and (B) a representative fimbria lesion (hatched area). 3V, third ventricle; CG, central gray area; CP, caudate putamen; Ent, entorhinal cortex; f, fimbria; ic, internal capsule; ml, medial lemniscus; ot, optic tract; S, subiculum; SNr, substdntia nigra pars reticulata; st, stria terminalis. Adapted from Paxinos and Watson (1986).

surface of the tissue and collected onto copper mesh grids. Sections were counterstained with uranyl acetate and lead citrate and examined with a Zeiss 902 transmission electron microscope.

Fimbria lesion The fimbria was unilaterally ablated by aspiration in

eight adult male Sprague-Dawley rats anesthetized with chloral hydrate (420 mgikg i.p.1 or halothane. One of these animals was originally processed for a previous study (Sesack and Pickel, 1990). In addition to the lateral fimbria, the lesion also included the overlying cortex and corpus callosum. Damage to the most dorsal aspect of the striatum and hippocampus was observed in some animals (Fig. 1B). Two additional animals received sham lesions that included only the cortex and corpus callosum overlying the fimbria, and thus served as controls. Following a 2-3 day survival period, five fimbria-lesioned and the two sham-lesioned animals were deeply anesthetized and perfused with acrolein using the protocol described above. Tissue from these animals was processed for immunocytochemical localization of the catecholamine synthetic enzyme TH. The remain- ing animals were processed for immunocytochemistry by

using an antiserum raised against DA. These animals were perfused with 500 ml of 5% glutaraldehyde in 0.1 M phosphate buffer containing 0.4% sodium metabisulfite, pH 7.4, a fixative compatible with immunocytochemical stain- ing for anti-DA antibodies (Seguela et al., 1988). Vibratome sections through the PFC of animals prepared by either fixative were collected in 0.1 M phosphate buffer and incubated in a solution of 1% sodium borohydride in 0.1 M phosphate buffer for 30 min to improve immunocytochemical labeling (Leranth and Pickel, 1989).

Immunocytochemistry Immunoreactivity for DA or TH was visualized by using

either the peroxidase-antiperoxidase (PAP) method with double bridging (Ordronneau et al., 1981) or the avidin- biotin peroxidase complex (ABC) technique (Hsu et al., 1981). Briefly, sections were incubated for 30 minutes in blocking serum containing 1% bovine serum albumin and 0.04% Triton X-100 in 0.1 M tris-buffered saline (TBS), pH 7.6. For tissue fixed with glutaraldehyde, 0.4% sodium metabisulfite was included in all incubation and rinse solutions to inhibit DA oxidation until the primary antibody was bound. All incubation and washing steps were performed at room temperature under constant gentle agitation. Sections were incubated for 12-15 hour with polyclonal antisera raised in rabbits against TH (1: 1,000, Eugene Tech) or DA (1:5,000, purchased from Dr. Harry Steinbusch, University of Limburg, Maastricht) and diluted in blocking serum.

For tissue processed by the ABC method, sections were rinsed thoroughly in TBS and incubated for 30 minutes in biotinylated goat anti-rabbit antiserum ( 1:400, Vector Labs) diluted in blocking serum. Sections were rinsed again, then incubated for 30 minute in avidin-biotin peroxidase complex (1:200, Vectastain Elite Kit, Vector Labs). For tissue processed by the double-bridge PAP method, sections were incubated for 1 hour in goat anti-rabbit antiserum ( 1 : l O O ) followed by a 1 hour incubation in rabbit PAP (1:100, Sternberger Monoclonals). Sections were then reincubated in the secondary antiserum, followed by the PAP. For tissue processed by either method, the peroxidase reaction product was visualized by placing the sections in a solution containing 0.022% DAB and 0.003% hydrogen peroxide in 0.1 M TBS for 1-6 minutes. The peroxidase reaction was terminated by rinsing the sections in 0.1 M TBS. Sections through the PFC were prepared for light or electron microscopic examination as described above.

Control sections in which normal rabbit serum was substituted for the primary antibody exhibited no immuno- reaction product by light or electron microscopy. Additional tests for specificity, including antigen preadsorption (DA antiserum) or immunoelectrophoresis and immunoprecipi- tation (TH antiserum) were performed by the respective suppliers (Joh et al., 1973; Steinbusch and Tilders, 1987).

Ultrastructural analysis For each animal, ultrathin sections were taken from one

to three vibratome sections through the medial PFC ipsilat- era1 to the fimbria lesion or HRP injection. The area of tissue examined included the deep layers of the prelimbic and infralimbic cortices (Krettek and Price, 1977a), which appear from light microscopic studies to be the area of greatest overlap between hippocampal and DA afferents (Descarries et al., 1987; Van Eden et al., 1987; van Groen and Wyss, 1990; Jay and Witter, 1991). No attempt was


membranes, making it difficult to determine which terminal contained the tracer. For this reason, only structures for which the source of the HRP crystals was evident were included in this study. Terminals labeled with HRP reaction product formed primarily asymmetric synapses on dendritic spines (Fig. 2A-C; Table 1); some of these exhibited discontinuous or “perforated” synaptic densities (Fig. 2A). A small population of labeled terminals formed asymmetric synapses on dendritic shafts (Fig. 2D; Table 1) or symmetric synapses onto dendritic shafts (not shown). Some HRP-labeled terminals were observed to form synapses with multiple postsynaptic targets within a single section (Fig 2A). No examples of HRP-labeled terminals forming axo-somatic or axo-axonic synapses were observed. However, many of the HRP-labeled terminals were in direct apposition to unlabeled terminals without forming axo- axonic specializations (Fig. 2D).

Following ablation of the fimbria, numerous degenerating terminals were observed in the PFC. These terminals displayed characteristics of electron-dense degeneration as described by Mugnaini and Friedrich (1981): shrunken and distorted shape, electron-dense cytoplasm, disrupted vesicles, swollen mitochondria, and often engulfment on non-synaptic sides by astrocytic processes (Fig. 3A,B). Terminals with these morphological characteristics were not observed within the PFC of animals subjected to sham lesions that did not include the fimbria. Degenerating terminals were significantly smaller in area than HRP- labeled terminals. However, there was no significant differ- ence in the length of synaptic specializations between degenerating and HRP-labeled terminals (Table 1).

Like HRP-labeled terminals, degenerating terminals predominantly formed asymmetric synapses on dendritic spines (Fig. 3A,B), with a small percentage forming asymmetric synapses on dendritic shafts (Table 1). No degenerating terminals were found to form symmetric synapses. Some degenerating terminals formed synapses with multiple postsynaptic structures within a single plane of section (Fig. 3A). Most of these terminals contacted multiple dendritic spines; however, one terminal formed an asymmetric synapse on both a dendritic spine and a dendritic shaft. Many degenerating terminals were found in direct apposition to unlabeled terminals, without forming axo-axonic specializations.

DA/TH-labeled terminals in the PFC Within the PFC, terminals immunoreactive for either DA

or TH (DAITH-IR) were observed to predominantly contact dendritic spines and small caliber dendritic shafts (Figs. 3C,D, 4A,B, 5A). When examined in single sections, the majority of DAITH-IR terminals (1481177; 83.6%) were not observed to form synaptic specializations (Fig. 3C, 5B). However, when synaptic specializations were observed, 86.2% were of the symmetric type (Figs. 3D, 4A,B, 5A), while 13.8% were asymmetric.

When examined in serial sections, most dendritic spines that received a symmetric input from a DAITH-IR terminal also received an asymmetric input from an unlabeled terminal, forming a triadic complex (Fig. 3D). In a few cases, DA/TH-IR terminals formed a symmetric synapse on a dendritic spine that was not observed to receive additional synaptic input. However, in each of these cases, technical limitations prevented complete serial examination of these spines. Because observations in other cortical regions indicate that dendritic spines receiving a symmetric input invariably also receive an asymmetric input (Colonnier,

made to compare these two cortical divisions in this study. Within single, non-consecutive thin sections, tissue contained within grid squares (boundaries of the grid mesh = 3,025 pm2) along the epon-tissue interface was examined. All degenerating or HRP-labeled terminals within these areas were photographed a t a magnification of ~ 2 0 , 0 0 0 . These micrographs were used to determine the morphology, area, synaptic length, and postsynaptic targets of hippocampal terminals labeled with HRP or degeneration. A total of 665,500 pm2 of tissue was examined from animals receiving HRP injections (1-2 vibratome sections per animal), and 456,775 pm2 of tissue was examined from fimbria-lesioned animals (1-3 sections per animal). An additional 133,100 pm2 of tissue was examined from animals receiving sham lesions (1-2 sections per animal).

Tissue immunolabeled with either DA or TH antiserum was used to examine the synaptic relationships of degenerating hippocampal and DA terminals. The following steps were taken to reduce false negatives due to inadequate antibody penetration. Only those areas a t the surface of the tissue (eponitissue interface) where antibody penetration is maximal were examined. In addition, only micrographs that contained both degenerating hippocampal and TH- or DA-labeled processes within the same 32.5 pm2 area (area contained within a single micrograph taken at x 12,000 magnification) were analyzed.

Neuronal elements, as described by Peters and colleagues (1991 1, were identified in electron micrographs as follows. Axon terminals were at least 0.2 pm in diameter and contained numerous synaptic vesicles. Dendritic processes contained mitochondria, microtubules, and/or rough endo- plasmic reticulum and exhibited densities postsynaptic to axon terminals. Dendritic spines also contained a postsynaptic density but did not contain the above mentioned organ- elles. Dendritic spines often contained a spine apparatus in single sections; however, the presence of a spine apparatus was not required for a process to be classified as a dendritic spine. As dendritic processes were usually not traced in serial sections, some small-caliber dendritic shafts may have been categorized as dendritic spines.

Asymmetric synapses (Gray’s type 1; Gray, 1959) were identified by thickened postsynaptic densities, while symmetric synapses (Gray’s type 11) had thin postsynaptic densities. Appositions were defined as parallel membrane associations that did not contain synaptic specializations and were not separated by glial processes.

In the case of terminals labeled with DA or TH antisera, synapses were typically thin and lacked an appreciable postsynaptic density. Thus, the widened spacing of parallel apposed plasmalemmal surfaces and the presence of inter- cleft filaments (Peters et al., 1991) were often the only recognized features of symmetric synapses formed by DA- or TH-labeled terminals. This classification has been ap- plied to previous descriptions of catecholaminergic as well as cholinergic synapses in cortex (Seguela et al., 1988, 1990; Smiley and Goldman-Rakic, 1993; Umbriaco et al., 1994).

RESULTS Comparison of HRP-labeled and

degenerating terminals in the PFC Following anterograde transport of WGA-HRP, crystals

of HRP reaction product were found in axon terminals that contained numerous small, clear synaptic vesicles (Fig. 2A-D). These crystals sometimes pierced through terminal


Fig. 2. Electron micrographs showing terminals in the rat prefrontal cortex (PFC) containing crystals of HRP reaction product following anterograde transport from the ventral hippocampal formation. The HRP-labeled terminals (hT) form asymmetric synapses, primarily on dendritic spines t S; curved arrows in A-C), but also on a dendritic shaft tD; curved arrow in D). In A and B, unlabeled terminals (uT) are apposed to, b u t do not synapse on, the spines receiving synaptic input

from HRP-labeled terminals. In B, an unlabeled terminal forms a symmetric synapse (curved arrow) on an adjacent dendrite. In C, both the HRP-labeled terminal and the spine that it synapses upon are surrounded by astrocytic processes (asterisks). In D, an unlabeled terminal forms an asymmetric synapse on the same dendrite as an HRP-labeled terminal. The unlabeled terminal is also in direct apposition (facing arrowheads) to the HRP-labeled terminal. Scale bar, 0.25 pm.


TABLE 1. Comparison of Hippocampal Terminals in the Rat PFC Identified by HRP Transport or Anteroprade Degeneration'

Size Synaptic type and postsynaptic target

Area ikm2) Synaptic length Asymmetric Asymmetric Symmetric (mean 2 S.D.) ibm) (mean 5 S.D.) on spines on dendrites on dendrites

HRP-labeled terminals 0.50 2 0.27 0.31 0.15 67/72 4/72 1/72 ( n = 721 ( n = 77)* 93'> 6% 1 14

In = 117) In = 1311' 98% 2% 0% Degenerating terminals 0.20 t 0.12" 0.29 i 0.12 n.s. 116/119 3/119 0/119

'n.s.. niit significant. HRP, horseradish peroxidase; PFC, prefrnntal cortex *&me terminals form more than one synapse. " P < 0.0005, Student's f test.

1968; Jones and Powell, 1969; Somogyi and Soltesz, 1986; Dehay et al., 19911, these terminals were included in the total number of DA varicosities that formed triadic complexes. In the regions of neuropil examined, it appeared that triadic complexes involving DAITH-IR terminals (as opposed to unlabeled terminals) make up a fairly large proportion of all observed triadic complexes. However, a more thorough quantitative study is required to address this issue in detail.

Relationship between DA/TH-labeled and degenerating terminals in the PFC

In the majority of the tissue analyzed, degenerating hippocampal and DAITH-IR terminals were not found in adjacent areas of the neuropil. In order to quantify this observation, when a degenerating terminal was observed, it was noted whether a DAITH-IR terminal was also located in the same photographic frame (area of frame = 32.5 pm2 at a magnification of ~ 1 2 , 0 0 0 ) . Similarly, when a DA/ TH-IR terminal was observed, it was noted whether a degenerating terminal was also located in the same photographic field. This examination of the neuropil demonstrated that only 32.2% of the degenerating terminals analyzed were observed in the same photographic frame as DAITH-IR terminals, and only 15.6% of the DAITH-IR terminals were observed in the same frame as degenerating terminals. When co-localized in the same field, degenerating terminals and DAITH-IR terminals were often in close proximity. However, they typically contacted separate postsynaptic targets (Fig. 4A,B) or were not both observed to form synaptic contacts within the same plane of section (Fig. 5A,B). In every possible instance, degenerating and DAITH-IR terminals that were observed in close proximity were followed in a limited series of 2-5 adjacent sections. However, in no case were dendritic shafts or spines found to receive both an asymmetric contact from a degenerating terminal and a synaptic or appositional contact from a DAITH-IR terminal (Table 2 ) . Moreover, degenerating hippocampal terminals throughout the neuropil surrounding DAITH-IR varicosities invariably formed asymmetric synapses onto dendritic spines that did not receive additional synaptic contact from any source (Table 2). Further- more, extensive examination of areas of the neuropil that did not contain DAITH-IR terminals revealed only a single example of a degenerating terminal that synapsed on a dendritic spine that received a second synaptic input; in this case an unlabeled terminal was observed to from a second asymmetric synapse on the same spine. Further examination of HRP-labeled tissue also did not reveal any labeled terminals that formed synaptic contact onto dendritic spines that received additional synaptic contact.

When degenerating hippocampal and DAITH-IR terminals were co-localized within the same field, they were often

in direct apposition to one another without forming axo- axonic specializations (Fig. 6, Table 2 ) . Degenerating terminals were also frequently apposed to unlabeled terminals (Table 2 ) . Apposed degenerating and DAITH-IR terminals were sometimes surrounded on non-apposed and non- synaptic sides by astrocytic processes (Fig. 6A,B) In many cases, the degenerating terminal of the apposed pair was observed to form an asymmetric synapse on an adjacent dendritic spine.

Examination of the nucleus accumbens shell To confirm that the lack of observed convergence be-

tween degenerating and DAITH-IR terminals in the PFC was not due to technical limitations, tissue from the shell region of the nucleus accumbens, an area where convergence between these inputs has been reported (Sesack and Pickel, 1990), was analyzed from the same fimbria-lesioned animals. Examination of this tissue revealed numerous examples of degenerating hippocampal terminals forming asymmetric synapses on dendritic spines that also received a synaptic or appositional contact from DAITH-IR terminals (Fig. 7). No attempt was made to quantify the fre- quency of this occurrence in this study.

DISCUSSION The present results, which are shown schematically in

Figure 8, suggest that hippocampal and DA afferents to the PFC target either separate dendritic compartments of common pyramidal neurons or separate neuronal populations. In addition, these inputs may interact via non- synaptic mechanisms a t pre- or postsynaptic sites.

Ultrastructural characteristics of hippocampal afferents to the PFC

In the present study, two different anterograde tract- tracing techniques were employed to maximize the advan- tages inherent in each technique. Anterograde transport of WGA-HRP allows for the examination of the normal morphology of axon terminals containing the transported lec- tin, but has been demonstrated to label a smaller population of hippocampal efferents than anterograde degeneration following fimbria transection (Sesack and Pickel, 1990). Conversely, anterograde degeneration following fimbria lesion labels a greater population of hippocampal efferents, but the normal morphology of degenerating terminals is typically compromised (Mugniani and Friedrich, 1981 ). However, a major advantage of the anterograde degeneration method is that it can be combined with single label immunocytochemistry to examine the relationships between multiple afferent pathways. Thus, the use of anterograde degeneration and peroxidase immunocytochemistry is the most sensitive combination of markers currently


Fig. 3. Electron micrographs demonstrating the morphology and synaptic contacts of terminals in the PFC labeled by anterograde degeneration following fimbria lesion (A and B) or by peroxidase immunoreactivity for tyrosine hydroxylase (TH; C and D). In A and B, degenerating terminals (dT) exhibit a distorted shape, electron dense cytoplasm, disrupted vesicles (v), and swollen mitochondria (m). In B, the degenerating terminals form asymmetric synapses (curved arrows) on dendritic spines. The degenerating terminal and its postsynaptic

target are enveloped by an astrocytic process (asterisks). In C and D, TH-labeled terminals (thT) form close appositions (straight arrows in C) or symmetric synapses (curved arrow in D! on unlabeled spines or dendritic shafts (D!. The spines also receive asymmetric synapses (curved arrows) from unlabeled terminals (LIT). In D, the TH-labeled terminal is closely apposed (facing arrowheads! to an unlabeled terminal. Scale bar, 0.25 pm.


Fig. 4. Electron micrographs demonstrating the typical spatial relationship between degenerating and TH-labeled terminals located in adjacent areas of the neuropil in the PFC. In A and B, degenerating terminals (dT) form asymmetric synapses (curved arrows) on unlabeled spines. In both cases, the degenerating terminals are partially surrounded by astrocytic processes (asterisks). The degenerating terminals

and their postsynaptic targets are spatially distant from TH-labeled terminals (thT) in the surrounding neuropil. The TH-labeled terminals form symmetric synapses (curved arrows) on unlabeled dendrites in both A and B. The symmetric synapse formed by the TH-labeled terminal in B is also evident in an adjacent section (inset). Scale bar, 0.5 pm.

available for labeling hippocampal and dopamine afferents, respectively (Sesack and Pickel, 1990).

A number of observations suggest that anterograde transport of WGA-HRP from the ventral hippocampal formation and anterograde degeneration following fimbria lesion labeled a similar population of terminals in the medial PFC. Specifically, HRP-labeled and degenerating

terminals did not differ significantly in their types of synaptic contacts, postsynaptic targets, or synaptic lengths. The observed differences in the morphology and cross- sectional area of degenerating and HRP-labeled terminals were most likely the result of shrinkage and glial engulfment due to the process of degeneration (Mugniani and Friedrich, 1981 1, Other subtle differences in the population


Fig. 5. Electron micrographs demonstrating the typical spatial In B, a degenerating terminal forms an asymmetric synapse (curved arrow) on a dendritic spine. Astrocytic processes (asterisks) surrounding the degenerating terminal separate it and its postsynaptic spine from a nearby TH-labeled process (th). Scale bar, 0.25 pm.

relationship of degenerating and TH-labeled profiles when co-localized in the neuropil of the PFC. In A, a TH-labeled terminal (thT) forms a symmetric synapse (curved arrow) on an unlabeled dendrite (D) that is separated from a degenerating terminal idT) by an intervening process.

TABLE 2. Associations Between Degenerating Hippocampal Terminals and DA/TH-Labeled or Unlabeled Terminals in the Rat PFC'

an excitatory influence (Peters et al,, 1991). This conc~usion is consistent both with physiological data demonstrating that hippocampal stimulation evokes primarily excitatory responses in PFC neurons (Laroche et al., 1990; Jay et al., 1992) and with anatomical and biochemical evidence that hippocampofugal fibers utilize an excitatory amino acid transmitter, most likely glutamate (Walaas and Fonnum,

Total area of tissue examined that contained both degenerating and

Number of degenerating terminals Number of DA TH terminals

Number of dexrnerating terminals synapsing on the same dendrite as

DA 'I'H-labeled terminals within the same 32.5 &m'area* 1,105 kmz 50 53

iO<L, Number of degenerating termmals synapsing on the same dendrite as

DA TH trrminals .~

unlabeled t&rnmals 0 10% i Number of degenerating terminals apposed to DK'I'H terminals 11 i225? I

12 124'3) Number ofdegewrating termmals apposed to unlabeled terminals

IDA. dopaminr . TH. tyrosine hydroxylas? *Area of i t field photographed at x 12,000

of terminals labeled by the two techniques may be explained by the minor involvement of other afferent pathways to the PFC, such as those arising from the amygdala or entorhinal cortex (Krettek and Price, 1977b; Sarter and Markowitsch, 1983; Swanson and Kohler, 1986). However, the similarity between terminals labeled by WGA-HRP transport and anterograde degeneration indicates either that these afferents did not significantly contribute to the population of degenerating and HRP-labeled terminals, or that their pattern of termination is similar to hippocampal afferents.

The ultrastructural characteristics of HRP-labeled and degenerating terminals observed in this study agree well with morphological descriptions of hippocampal terminals in subcortical structures (Leranth and Frotscher, 1989; Totterdell and Smith, 1989; Meredith and Wouterlood, 1990; Sesack and Pickel, 1990). The observation that hippocampal terminals formed almost exclusively asymmetric contacts indicates that this pathway is likely to mediate

1979; Zaczek et al., f9?9; Walaas and Fonnum, 1980; Christie et al., 1987; Fuller et al., 1987; Jay et al., 1992).

The majority of hippocampal terminals labeled by either anterograde tracing method formed synaptic contacts with dendritic spines. As the majority of dendritic spines in the cerebral cortex arise from pyramidal neurons (DeFelipe and Farinas, 1992), this finding indicates that these neurons are the primary cell type contacted by hippocampal terminals. Although a small population of hippocampal terminals formed synaptic contacts on dendritic shafts, these dendrites were not examined in serial sections, making it difficult to state whether their morphology was consistent with pyramidal or non-pyramidal cells. However, it appears that any hippocampal innervation of non-pyramidal neurons in the PFC represents, at best, a minor projection of this pathway.

Ultrastructural characteristics of DA terminals in the PFC

The ultrastructural characteristics of DA/TH-IR terminals observed in this study are consistent with previous investigations of the PFC that used either DA or TH antiserum or both (Van Eden et al., 1987; Seguela et al., 1988; Goldman-Rakic et al., 1989; Verney et al., 1990).


Fig. 6 . Serial electron micrographs of the PFC demonstrating degenerating terminals and TH-labeled processes in direct apposition to one another. In the series A-B and the series C-D, degenerating terminals (dT) form asymmetric synapses (curved arrows) on unlabeled

spines and are directly apposed (facing arrowheads) to TH-labeled processes (th). In the series A-B, the degenerating terminal, its postsynaptic spine, and the apposed TH-labeled process are contacted by astrocytic processes (asterisks). Scale bar, 0.25 pm.

Thus, these terminals probably represent dopaminergic afferents from the VTA. Although some labeling of noradrenergic (NE) terminals by the TH antiserum cannot be disproved, it would appear that such labeling did not significantly influence the results for the following reasons. First, our light microscopic examination revealed that fibers labeled by either DA or TH antiserum exhibited similar morphological features and were concentrated in the deep layers of the PFC; these fibers did not demonstrate the morphology or laminar pattern of innervation characteristic of NE fibers (Lindvall and Bjorklund, 1984). Second, the comparable ultrastructural features of terminals labeled with either DA or TH antiserum suggests that a similar population of fibers is being recognized at the electron microscopic level, consistent with previous studies directly comparing these markers (Goldman-Rakic et aI., 1989; Verney et al., 1990). Finally, it has been demonstrated that in the primate cortex, TH antisera do not significantly label NE axons and terminals (for a thorough discussion see Noack and Lewis, 1989; Akil and Lewis, 1993). This observation has also been reported in subcortical sites in the rat (Asan, 1993). For these reasons, it is concluded that NE terminals either did not significantly contribute to the population of TH-IR terminals in this study, or that labeled NE terminals, similar to DAITH-IR terminals, did not converge with hippocampal terminals on common targets.

The majority of DAITH-IR terminals did not form a recognizable synaptic contact within a single plane of

section, consistent with the observations of Descarries and colleagues that serial section analysis is often required to detect synapses formed by monoaminergic and cholinergic terminals in cortex (Seguela et al., 1988, 1989, 1990; Umbriaco et al., 1994). When synaptic junctions were observed, the results were consistent with previous reports that DA terminals form primarily symmetric synapses on dendritic spines and distal dendrites (Seguela et al., 1988; Goldman-Rakic et al., 1989; Verney et al., 1990). The targets of DA terminals in the PFC are known to include both spiny pyramidal neurons and non-spiny GABAergic interneurons (Goldman-Rakic et al., 1989; Smiley and Goldman-Rakic, 1993; Sesack et al., 1995a).

The observation that dendritic spines receiving a symmetric input from a DAITH-IR terminal invariably also received an asymmetric input from an unlabeled terminal matches previous descriptions in the rat and monkey PFC (Goldman-Rakic et al., 1989; Verney et al., 1990). In theoretical models, this triadic arrangement of inputs onto a dendritic spine is believed to provide a substrate by which one input can modulate the efficacy of an excitatory input, without itself significantly altering the electrical activity of the parent dendrite (Koch and Poggio, 1983b; Pongracz, 1985; Shepherd and Brayton, 1987; Rall and Segev, 1988; Segev and Rall, 1988; Qian and Sejnowski, 1990). Although the percentage of all dendritic spines that receive conver- gent symmetric and asymmetric inputs was not calculated in this study, previous descriptions of other cortical areas indicate that 5 3 0 % of dendritic spines receive such dual


Fig. 7. Serial electron micrographs of the nucleus accumbens shell displaying convergence of degenerating and TH-labeled terminals. In the series A-B, a degenerating terminal (dT) forms asymmetric synapses (curved arrows) on unlabeled spines. These spines also receive a

symmetric synapse (curved arrow in A) or a close apposition (straight arrow in B) from a TH-labeled terminal (thT). The degenerating terminal is surrounded on non-synaptic sides by an astrocytic process (asterisks). Scale bar, 0.25 km.

inputs (Jones and Powell, 1969; Koch and Poggio, 1983a; Beaulieu and Colonnier, 1985). This suggests that only a select number of excitatory inputs onto pyramidal cells are synaptically modulated by any neuroactive substance at the level of the dendritic spine. In the PFC, our preliminary observations that a large proportion of symmetric synapses on dendritic spines are formed by DA as opposed to non-DA terminals indicates that the mesoprefrontal DA system is a major component of the modulatory inputs onto these spines. The identity of the non-DA terminals that also formed symmetric synaptic contacts on dendritic spines receiving asymmetric input is not known. However, it has been demonstrated that GABAergic, cholinergic, and noradrenergic terminals occasionally participate in this conver- gent synaptic complex within the cortex (Houser et al., 1984; Papadopoulos et al., 1989; Seguela et al., 1990; Verney et al., 1990; Umbriaco et al., 1994).

Relationship between hippocampal and DA terminals in the PFC

The lack of observed postsynaptic convergence between hippocampal and DA terminals on common spines in the PFC is contrary to our original hypothesis. Although examples of convergence may have been missed due to limitations inherent in the methods, several observations suggest that a significant convergence of these afferents was not overlooked. 1) Efforts were made to minimize false negatives due to limited antibody penetration of tissue exposed to low concentrations of detergent. Precautions

included the use of both TH and DA antibodies, which produced similar results, and confining the analysis to the epon/tissue interface where maximal antibody penetration occurs. 2) Because anterograde degeneration of hippocampal terminals occurs throughout the section and is not limited to the tissue surface, it is significant that even in areas where DAITH-IR terminals were detected, hippocampal terminals were not often found in the same field. This observation suggests that the apparent overlap between these two afferents observed at the light microscopic level does not necessarily extend to the ultrastructural level. 3) Examination of tissue labeled by either HRP transport or anterograde degeneration indicates that the majority of dendritic spines receiving synaptic input from hippocampal terminals do not receive any additional synaptic contact, at least in the single sections examined. This finding suggests that hippocampal input to pyramidal cells of the PFC may not be subject to modulation at the level of the dendritic spine by any other synaptic input. 4) Examination of the shell regon of the nucleus accumbens in the same animals used for this study revealed numerous examples of convergence of hippocampal and DA afferents. This finding illus- trates that the convergence of these inputs can be readily observed in a region where it is reported to occur (Totterdell and Smith, 1989; Sesack and Pickel, 1990).

The use of a single section, rather than a complete serial section analysis, is one factor that could explain the inabil- ity to detect synaptic convergence between hippocampal and DA afferents to the PFC. However, in instances where


Hippocampal Terminal

Dopamine Terminal

Unlabeled Terminal

Fig. 8. Schematic diagram of the observed and potential relationships between hippocampal and DA terminals in the PFC. 1) Hippocam- pal terminals form primarily asymmetric synapses on dendritic spines of pyramidal cells, 2) with a small proportion forming asymmetric synapses on dendritic shafts. 3 ) Hippocampal terminals are often in direct apposition to unlabeled terminals. 4) DA terminals form primarily symmetric synapses on dendritic spines and shafts. 5) DA terminals

DAITH-IR terminals and degenerating terminals were in close proximity, a limited serial section analysis (2-5 contigu- ous sections) revealed no cases where these two inputs converged onto common targets. The possibility exists that DA terminals contact dendrites that also receive hippocampal input at sites more distant than can be detected in a limited serial section analysis. The functional implications of such a distant convergence are unclear, since DA may not be able to selectively modulate hippocampal input without affecting other afferents to the same dendrite. Alterna- tively, DA and hippocampal afferents may contact separate populations of neurons in the PFC. A more thorough serial reconstruction study will be required to examine these questions.

I t is important to note that this investigation was con- fined to the deep layers of the PFC and the possibility of convergence between hippocampal and DA terminals in the superficial layers of the PFC cannot be excluded. Although future ultrastructural examination will be required to test this possibility, the low density of DA terminals within the superficial layers suggests that convergence between these afferents is unlikely.

Although hippocampal and DA terminals did not converge onto common dendritic spines, they were frequently in direct apposition to one another when co-localized in the same area of neuropil. These terminal associations did not

often form triadic complexes in which the spine postsynaptic to a DA terminal also receives asymmetric synaptic input from an unlabeled terminal. 6) DA and hippocampal terminals are often in direct apposition to one another, when localized to the same area of neuropil. 7) DA terminals may synapse on the same dendrites that receive input from hippocampal terminals, but at sites too distant to be observed in single sections.

display the ultrastructural characteristics of axo-axonic synapses, consistent with the observation that synapses involving two axon terminals, either immunoreactive for DA or unlabeled, are encountered only rarely, if at all in the PFC (Van Eden et al., 1987; Seguela et al., 1988; Smiley and Goldman-Rakic, 1993). Whether these terminal appositions represent functional sites of interaction is not known. However, it may be significant that degenerating hippocampal and DAITH-IR terminals remained in contact with one another despite glial envelopment of other non-synaptic portions of degenerating terminals.

Close appositions between DAITH-IR terminals and unlabeled terminals forming asymmetric synapses were also commonly observed in the PFC and resemble terminal associations between cortical and DA terminals in subcortical structures (Bouyer et a]., 1984; Sesack and Pickel, 1990, 1992). These findings suggest that presynaptic associations may be a characteristic feature of DA terminals in forebrain regions, although their functional significance is unclear. The recent findings that immunoreactivity for D1 (Smiley et al., 1994) or Dz (Sesack et al., 199513, DA receptors is sometimes localized to terminals forming asymmetric synapses within the cortex are consistent with the hypothesis that terminal appositions represent functional sites of non-synaptic interaction. At the least, terminal appositions may facilitate presynaptic interactions by limiting the

HIPPOCAMPAL AFFERENTS TO THE PFC

distance over which DA must diffuse in order to reach its target site. Whether such interactions occur between DA and hippocampal terminals in the PFC is not known, and future physiological examination will be required to address this question.

Functional significance The observation that hippocampal terminals in single

sections did not converge onto common dendritic spines with any other synaptic input suggests that hippocampal input onto PFC neurons is not likely to be synaptically modulated at the level of the dendritic spine by DA or any other neuroactive substance. Although this input may be modulated by inputs onto the parent dendrite, such an arrangement, according to theoretical studies, lacks the selectivity of modulation found in the triadic complex. Thus, hippocampal afferents to the PFC may mediate a strong excitatory influence onto pyramidal neurons, despite their relatively distal site of termination.

The observation that hippocampal input to the PFC does not appear to be synaptically modulated at the level of the dendritic spine has important implications for understand- ing information processing in the cortex. Thus, although 5-30% of all dendritic spines have been estimated to receive modulatory synaptic input (Jones and Powell, 1969; Koch and Poggio, 1983a; Beaulieu and Colonnier, 1985; Dehay et al., 1991), our findings suggest that only select excitatory inputs are synaptically modulated in this manner. A similar finding has been reported in the visual cortex (Dehay et al., 199 1). Future ultrastructural examination will be required to determine which excitatory cortical afferents are subject to such modulation in the PFC.

Our findings, taken together with previous observations (Goldman-Rakic et al., 1989; Verney et al., 1990), suggest that DA afferents to the PFC play an important role in the modulation of excitatory input at the level of the dendritic spine. Such a mechanism may contribute to the reported cognitive functions of DA within the PFC (Brozoski, 1979; Simon, 1980; Sawaguchi and Goldman-Rakic, 1991, 1994). The present findings suggest that hippocampal afferents are not among those excitatory connections that are subject to synaptic modulation by DA. The afferent systems to the PFC that do converge with DA terminals on common dendritic spines are not known. Possible candidates for this triadic association include the amygdala, contralateral PFC, thalamic nuclei, perirhinal cortex, and entorhinal cortex (Divac et al., 1978; Sarter and Markowitsch, 1983; Swanson and KohIer, 1986; McDonald, 1987; Audinat et al., 1988; Sesack et al., 1989; McDonald, 1991; Van Eden et al., 1992). Future anatomical and physiological investigations are required to determine which afferent system or systems account for the observed convergence with DA afferents on common spines.

13

ACKNOWLEDGMENTS This work is supported by USPHS grant MH50314 (SRS)

and an NSF Neural Processes in Cognition Fellowship (DBC). The authors thank Thomas Harper for skilled technical assistance, Steven King and Beth Ann Basola for assisting in the data analysis, and Dr. J. Patrick Card and Beth Fisher for thoughtful comments on the manuscript. The authors gratefully acknowledge Dr. Theresa Milner and Claudia Farb for assistance with the WGA-HRP transport.

LITERATURE CITED Akil, M., and D.A. Lewis (1993) The dopaminergic innervation of monkey

entorhinal cortex. Cerebral Cortex 3:533-550. Asan, E. (1993) Comparative single and double immunolabelling with

antisera against catecholamine biosynthetic enzymes: Criteria for the identification of dopaminergic, noradrenergic and adrenergic structures in selected rat brain areas. Histochemistry 99:427-442.

Audinat, E., F. Conde, and F. Crepe1 (1988) Cortico-cortical connections of the limbic cortex of the rat. Exp. Brain Res. 69r439-443.

Beaulieu, C., and M. Colonnier (19851 A laminar analysis of the number of round-assymmetrical and flat-symmetrical synapses on spines, dendritic trunks, andcell bodiesinarea 17ofthecat. J . Comp. Neurol.231180-189.

Bouyer, J .J . , D.H. Park, T.H. Joh, and V.M. Pickel (1984) Chemical and structural analysis of the relation between cortical inputs and tyrosine hydroxylase-containing terminals in rat neostriatum. Brain Res. 302.267- 275.

Brozoski, T.J. , R.M. Brown, H.E. Rosvold, and P.S. Goldman (19791 Cognitive deficit caused by regional depletion of dopamine in prefrontal cortex of rhesus monkey. Science 205929-932.

Carlsson, A. (1988) The current status of the dopamine hypothesis of schizophrenia. Neuropsychopharmacology I: 179-186.

Carr, D.B., H.C. Wolf, and S.R. Sesack (1994) Hippocampal afferents to the rat prefrontal cortex: synaptic targets and relation to dopamine terminals. SOC. Neurosci. Abstr. 20.530.

Christie, M.J., R.J. Summers, J.A. Stephenson, C.J. Cook, and P.M. Beart (1987) Excitatory amino acid projections to the nucleus accumbens septi in the rat: A retrograde transport study utihzing D["Hlaspartate and PJH IGABA. Neuroscience 22.425-439.

Colonnier, M. ( I 968) Synaptic patterns on different cell types in the different laminae of the cat visual cortex. An electron microscope study. Brain Res. 9268-287.

DeFelipe, J . , and I. Farinas (1992) The pyramidal neuron of the cerebral cortex: morpholog~cal and chemical characteristics of the synaptic inputs. Prog. Neurobiol. 39:563-607.

Dehay, C., R.J. Douglas, K.A.C. Martin, and C. Nelson !I9911 Excitation by genicuIocortical synapses is not 'vetoed' a t the level of dendritic spines in cat visual cortex. J. Physiol. 440r723-734.

Descarries, L., B. Lemay, G. Doucet, and B. Berger (1987) Regonal and laminar density of the dopamine innervation in adult rat cerebral cortex. Neuroscience 21:807-824.

Divac, I., A. Kosmal, A. Bjorklund, and 0. Lindvall (1978) Subcortical projections to the prefrontal cortex in the rat as revealed by the horseradish peroxidase technique. Neuroscience 3.785-796.

Ferino, F., A.-M. Thierry, and J . Glowinski (1987) Anatomical and electrophysiological evidence for a direct projection from Ammon's horn to the medial prefrontal cortex in the rat. Exp. Brain Res. 65:421-426.

Friedman, H.R., and P.S. Goldman-Rakic (1988) Activation of the hippocampus and dentate gyrus by working-memory: a 2-deoxyglucose study of behaving rhesus monkeys. J. Neurosci. 8:4693-4706.

Fuller, T.A., F.T. Russchen, and J .L. Price (1987) Sources of presumptive glutamatergic/asparterg~c afferents to the rat ventral striatopallidal regon. J. Comp. Neurol. 2.58.317-338.

Fuster, J.M. (1991) The prefrontal cortex and its relation to behavior. Prog. Brain Res. 87:201-211.

Goeders, N.E., and J.E. Smith (1986) Reinforcingproperties ofcocaine in the medial prefrontal cortex: Primary action on presynaptic dopaminergx terminals. Pharmac. Biochem. Behav. 25191-199.

Goeders, N.E., and J .E. Smith (1980) Cortical dopaminergc involvement in cocaine reinforcement. Science 221:773-775.

Goldman-Rakic, P.S., C. Leranth, S.M. Williams, N. Mons, and M. Geffard (1989) Dopamine synaptic complex with pyramidal neurons in primate cerebral cortex. Proc. Natl. Acad. Sci. 86.9015-9019.

Gray, E.G. (1959) Axo-somatic and axo-dendritic synapses of the cerebral cortex: An electron microscope study. J. Anat. 9 6 4 2 0 4 3 3 .

Houser, C.R., J .E. Vaughn, S.H.C. Hendry, E.G. Jones, and A. Peters ! 1984) GABA neurons in the cerebral cortex. In E. Jones and A. Peters !edsl: Cerebral Cortex. New York: Plenium, pp. 63-89.

Hsu, S.-M., L. Raine. and H . Fanger (1981) Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques. J. Histochem. Cyto- chem. 29:577-580.

Jay, T. M., A.-M. Thierry, L. Wiklund, and J. Glowinski (19921 Excitatory amino acid pathway from the hippocampus to the prefrontal cortex. Contribution of AMPA receptors in hippocampo-prefrontal cortex transmission. Eur. J. Neurosci. 4: 1285-1295.


Jay, T.M., J. Glowinski, and A.-M. Thierry (1989) Selectivity of the hippocampal projection to the prelimbic area of the prefrontal cortex in the rat. Brain Res. 505337-340.

Jay, T.M.. and M.P. Witter 11991) Distribution of hippocampal CA1 and subicular efferents in the prefrontal cortex of the rat studied by means of anterograde transport of Phaseolus vulgaris-Leucoagglutinin. J . Comp. Neural. 313:574-586.

Joh, I., C. Gegham, and D. Reis (1973) Immunocytochemical demonstration of increased tyrosine hydroxylase protein in sympathetic ganglia and adrenal medulla elicited by reserpine. Proc. Natl. Acad. Sci. U.S.A. 70: 2 76 7-2 77 1.

Jones, E.G., and T.P.S. Powell (1969) Morphological variations in the dendritic spines of the neocortex. J . Cell Sci. 5:509-529.

Koch. C., and T. Poggio (1983a) Electrical properties of dendritic spines. Trends Neurosci. 6:80-83.

Koch, C., and T. Poggio (198313) A theoretical analysis ofelectrical properties ofspines. Proc. R. Soc. Lond. 218:455-477.

Kolb, B. (19841 Functions of the frontal cortex of the rat: A comparative review. Brain Res. Rev. 8.65-98.

Krettek, J .E., and J.L. Price (1977a) The cortical projections of the mediodorsal nucleus and adjacent thalamic nuclei in the rat. J . Comp. Neurol. 171:157-192.

Krettek, J.E., and J.L. Price (197713) Projections from the amygdaloid complex to the cerebral cortex and thalamus in the rat and cat. J . Comp. Neurol. 172:687-722.

Laroche, S., T.M. Jay, and A.-M. Thierry (1990) Long-term potentiation in the prefrontal cortex following stimulation of the hippocampal CAli subicular region. Neurosci. Lett. 114:184-190.

LaVail. J .H., and M.M. LaVail 11974) The retrograde intra-axonal transport of horseradish peroxidase in the chick visual system. J . Comp. Neurol. 157:303-357.

Le Moal, M., and H. Simon (1991) Mesocorticolimbic dopaminergic network: functional and regulatory roles. Physiol. Rev. 71:155-234.

Lemann, W., C.B. Saper, D.B. Rye, and B.H. Wainer (1985) Stabilization of TMB reaction product for electron microscopic retrograde and anterograde fiber tracing. Brain Res. Bull. 14:277-281.

Leranth, C., and M. Frotscher (1989) Organization of the septa1 region in the rat brain: Cholinergic-GABAergic interconnections and the termination of hippocampo-septa1 fibers. J . Comp. Neurol. 289:304-314.

Leranth, C., and V.M. Pickel (1989) Electron microscopic preembedding double-immunostaining methods. In L. Heimer and L. Zaborszky (eds): Neuroanatomical Tract Tracing Methods 2: Recent Progress. New York: Plenum, pp. 129-172.

Lindvall, O., and A. Bjorklund (1984) General organization of cortical monoamine systems. In L. Descarries, T. Reader, and H. Jasper (eds): Monoamine Innervation of the Cerebral Cortex. New York: Alan R. Liss, Inc.. pp. 9-40.

Mahut. H., S. Zola-Morgan, and M. Moss (1982) Hippocampal resections impair associative learning and recognition memory in the monkey. J . Neurosci. 21214-1229.

McDonald, A.J. (1987) Organization of amygdaloid projections to the mediodorsal thalamus and prefrontal cortex: A fluorescence retrograde transport study in the rat. J . Comp. Neurol. 26246-58.

McDonald, A.J. (1991) Organization of amygdaloid projections to the prefrontal cortex and associated striatum in the rat. Neuroscience 44:l-14.

McGuire. B.A., C.D. Gilbert, P.K. Rivlin, and T.N. Wiesel (1991) Targets of horizontal connections in macaque primary visual cortex. J. Comp. Neurol. 305:370-392.

Meredith, G.E., and F.G. Wouterlood (1990) Hippocampal and midline thalamic fibers and terminals in relation to the choline acetyltransferase- immunoreactive neurons in nucleus accumbens of the rat: a light and electron microscopic study. J . Comp. Neurol. 296.904-221.

Mesulam, M.-M. (1978) Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: A non-carcinogenic blue reaction-product with superior sensitivity for visualizing neural afferents and efferents. J . Histochem. Cytochem. 26:106-117.

Mugniani, E., and V.L. Friedrich (1981) Electron microscopy: Identification and study of normal and degenerating neural elements by electron microscopy. In L. Heimer and M.J. RoBards (eds): Neuroanatomical Tract-Tracing Methods. New York: Plenum, pp. 377-406.

Noack, H.J., and D.A. Lewis (1989) Antibodies directed against tyrosine hydroxylase differentially recognize noradrenergic axons in monkey neocortex. Brain Res. 500:313-324.

Ordronneau, P. , P.B.-M. Lindstrom, and P. Petrusz (1981) Four unlabeled antibody bridge techniques: A comparison. J. Histochem. Cytochem. 29: 1397-1404.

Papadopoulos, G.C., J.G. Parnavelas, and R.M. Bujis (1989) Light and electron microscopic immunocytochemical analysis of the noradrenaline innervation of the rat visual cortex. J. Neurocytol. 18: 1-10.

Paxinos, G., and C. Watson i 1986) The Rat Brain in Stereotaxic Coordinates. New York: Academic Press.

Peters, A,, S.L. Palay, and H. Webster (1991) The Fine Structure of the Nervous System: Neurons and Their Supporting Cells. New York: Oxford.

Pongracz, F. (1985) The function of dendritic spines: A theoretical study. Neuroscience 15933-946.

Qian, N., and T.J. Sejnowski (1990) When is an inhibitory synapse effective? Proc. Natl. Acad. Sci. 87:8145-8149.

Rall, W., and I. Segev 11988) Synaptic integration and excitable dendritic spine clusters: Structureifunction. In R. Lasek and M. Black (eds): Intrinsic Determinants of Neuronal Form and Function. New York: Liss, pp. 263-282.

Rosenkilde, C.E. 11983) Functions of the prefrontal cortex: Behavioral investigations using ablation and electrophysiological techniques in rats, cats, dogs, and monkeys. Acta Physiol. Scand. Suppl514:4-58.

Ruit, K.G., and E.J. Neafsey (1988) Cardiovascular and respiratory responses to electrical and chemical stimulation of the hippocampus in anesthetized and awake rats. Brain Res. 457:310-321.

Sarter, M., and H.J. Markowitsch (1983) Convergence of basolateral amygdaloid and mediodorsal thalamic projections in different areas of the frontal cortex of the rat . Brain Res. Bull. 10:607-622.

Sawaguchi, T., and P.S. Goldman-Rakic (1991) D1 dopamine receptors in prefrontal cortex: Involvement in working memory. Science251:947-950.

Sawaguchi, T., and P.S. Goldman-Rakic (1994) The role of D1-dopamine receptor in working memory: Local injections of dopamine antagonists into the prefrontal cortex of rhesus monkeys performing and oculomotor delayed-response task. J . Neurophysiol. 71 :515-528.

Segev, I., and W. Rall 11988) Computational study of an excitable dendritic spine. J. Neurophysiol. 60:499-523.

Seguela, P., K.C. Watkins, and L. Descarries (1988) Ultrastructural features of dopamine axon terminals in the anteromedial and the suprarhinal cortex ofadult rat. Brain Res. 44211-22.

Seguela, P., K.C. Watkins, and L. Descarries (1989) Ultrastructural relationships of serotonin axon terminals in the cerebral cortex of the adult rat. J . Comp. Neurol. 289:129-142.

Seguela, P., K.C. Watkins, M. Geffard, and L. Descarries (1990) Noradrena- line axon terminals in adult rat neocortex: An immunocytochemical analysis in serial thin sections. Neuroscience 35249-264.

Sesack, S.R., A.Y. Deutch, R.H. Roth, and B.S. Bunney 119891 Topographical organization of the efferent projections of the medial prefrontal cortex in the rat: An anterograde tract-tracing study with Phaseolus uulgaris- leucoagglutinin. J . Comp. Neurol. 290:213-242.

Sesack, S.R., and V.M. Pickel (1990) In the rat medial nucleus accumbens. hippocampal and catecholaminergic terminals converge on spiny neurons and are in apposition to each other. Brain Res. 527:266-279.

Sesack, S.R., and V.M. Pickel (1992) Prefrontal cortical efferents in the rat synapse on unlabeled neuronal targets ofcatecholamine terminals in the nucleus accumbens septi and on dopamine neurons in the ventral tegmental area. J . Comp. Neural. 320:145-160.

Sesack, S.R., C.1,. Snyder, and D.A. Lewis 11995a) Axon terminals immunolabeled for dopamine or tyrosine hydroxylase synapse on GABA- immunoreactive dendrites in ra t and monkey cortex. J. Comp. Neurol. 363.3644280.

Sesack, S.R., S.W. King, C.N. Bressler, S.J. Watson, and D.A. Lewis (1995bl Electron microscopic visualization of dopamine D2 receptors in the forebrain: Cellular, regional, and species comparisons. Soc. Neurosci. Abstracts 21:365.

Shepherd, G.H., and R.K. Brayton (1987) Loac operations are properties of computer-simulated interactions between excitable dendritic spines. Neuroscience 21: 151-165.

Simon, H., B. Scatton, and M. Le Moal (1980) Dopaminergic A10 neurones are involved in cognitive functions. Nature 286:150-151.

Smiley, J .F. , and P.S. Goldman-Rakic 11993) 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) immunolabelling technique. Cerebral Cortex 3323-238.

Smiley, J .F. , A.I. Levey, B.J. Ciliax, and P.S. Goldman-Rakic 11994) D1 dopamine receptor immunoreactivity in human and monkey cerebral cortex: Predominant and extrasynaptic localization in dendritic spines. Proc. Natl. Acad. Sci. U.S.A. 91:5720-5724.

Somogyi, P., and I. Soltesz (1986) Immunogold demonstration of GABA in synaptic terminals of intracellularly recorded, horseradish peroxidase-


tilled basket cells and clutch cells in the cat’s visual cortex. Neuroscience 19:1051-1065.

Steinbusch, H.W.M., and F.J.H. Tilders 11987) Immunohistochemical techniques for light-microscopical localization of dopamine, noradrenaline, serotonin and histamine in the central nervous system. In H.W.M. Steinbusch led): Monoaminergic Neurons: Light Microscopy and Ultra- structure. New York: John Wiley & Sons, pp. 125-166.

Swanson. L.W., and C. Kohler (19861 Anatomical evidence for direct projections from the entorhinal area to the entire cortical mantle in the rat. J . Neurosci. 6.3010-3023.

Swanson, L.W. 11981) A direct projection from Ammon’s horn to prefrontal cortex in the rat. Brain Res. 21 7:150-154.

Totterdell. S., and A.D. Smith (1989) Convergence of hippocampal and dopaminergic input onto identified neurons in the nucleus accumbens of the rat. J. Chem. Neuroanat. 2:285-298.

Umbriaco, D., K.C. Watkins, L. Descarries, C. Cozzari, and B.K. Hartman ( 19941 Ultrastructural and morphometric features of the acetylcholine innervation in adult rat parietal cortex: An electron microscopic study in serial sections. J. Comp. Neurol. 348:351-373.

Van Eden, C.G., E.M.D. Hoorneman, R.M. Buijs, M.A.H. Matthijssen, M. Geffard, and H.B.M. Uylings 11987) Immunocytochemical localization of dopamine in the prefrontal cortex of the rat a t the light and electron microscopical level. Neuroscience 22349-862.

Van Eden, C.G., V.A.F. Lamme, and H.B.M. Uylings (1992) Heterotopic cortical afferents to the medial prefrontal cortex in the rat: A combined retrograde and anterograde tracer study. Eur. J. Neurosci. 4.77-97.

van Groen. T., and J.M. Wyss (1990) Extrinsic projections from area CA, of the rat hippocampus: Olfactory, cortical, subcortical, and bilateral hippocampal formation projections. J. Comp. Neurol. 302515-528.

Verney, C., C. Alvarez, M. Geffard, and B. Berger (1990) Ultrastructural double-labelling study of dopamine terminals and GABA-containing neurons in rat anteromedial cerebral cortex. Eur. J. Neurosci. 2.960-972.

Walaas, I., and F. Fonnum (1979) The effects of surgical and chemical lesions on neurotransmitter candidates in the nucleus accumbens of the rat. Neuroscience 4r209-216.

Walaas, I., and F. Fonnum (1980) Biochemical evidence for glutamate as a transmitter in hippocampal efferents to the basal forebrain and hypothala- mus in the rat brain. Neuroscience 5:1691-1698.

Watanabe, T., and H. Niki (1985) Hippocampal unit activity and delayed response in the monkey. Brain Res. 325241-254.

Weinberger, D.R. (1987) Implications of normal brain development for the pathogenesis of schizophrenia. Arch. Gen. Psychiatr. 44:660-669.

Wyss, J.M., L.W. Swanson. and W.M. Cowan 11980) The organization of the fimbria, dorsal fornix, and ventral hippocampal commissure in the rat. Anat. Embryol. 158:303-316.

Yang, C.R., and G.J. Mogenson (1984) Electrophysiological responses of neurones in the nucleus accumbens to hippocampal stimulation and the attenuation of the excitatory responses by the mesolimbic dopaminergic system. Brain Res. 324:69-84

Yang, C.R., and G.J. Mogenson i 1986) Dopamineenhances terminal excitabil- ity of hippocampal-accumbens neurons via D2 receptor: Role of dopamine in presynaptic inhibition. J. Neurosci. 62470-2478.

Zaczek, R., J.C. Hedreen, and J.T. Coyle (1979) Evidence for a hippocampal- septa1 glutamatergic pathway in the rat . Exp. Neurol. 65:145-156.

Zola-Morgan, S., and L.R. Squire (1986) Memory impairment in monkeys following lesions limited to the hippocampus. Behav. Neurosci. 100:155- 160.

hippocampal afferents to the rat prefrontal cortex: synaptic targets and relation to dopamine...

Documents