Connections of the basal forebrain of the weakly electric fish Eigenmannia virescensкод для вставкиСкачать
THE JOURNAL OF COMPARATIVE NEUROLOGY 389:49–64 (1997) Connections of the Basal Forebrain of the Weakly Electric Fish, Eigenmannia virescens CALVIN J.H. WONG* The Neurobiology Unit, Scripps Institution of Oceanography, University of California San Diego, La Jolla, California 92093-0201 ABSTRACT The organization of the ventral nucleus of the ventral telencephalon (Vv) was examined in the weakly electric fish, Eigenmannia virescens. This nucleus, which is considered the teleost homologue to the basal forebrain nuclei of other vertebrates, was subdivided into dorsal and ventral subdivisions, based upon cytoarchitectonic, immunohistochemical, and connectional criteria. Afferent projections were observed from the medial olfactory bulb as well as the terminal nerve ganglion. Telencephalic afferents to the Vv were very restricted, consisting of the supracommissural and the dorsal intermediate nuclei of the ventral telencephalon, the nucleus taenia, and the medial region of the posterior nucleus of the dorsal telencephalon. However, the major afferents to the Vv were diencephalic, particularly those originating from the rostral preoptic area and other hypothalamic nuclei. Additional afferents included the posterior tubercular nucleus, the locus coeruleus, the medial perilemniscal nucleus, and the periventricular nucleus of the posterior tuberculum. Relatively weak projections were observed from the ventral thalamus and the dorsal posterior thalamic nucleus. As described previously, the diencephalic complex of the central posterior thalamic nucleus/prepacemaker nucleus (CP/PPn), which also has cells that innervate the pacemaker circuitry controlling the production of an electric organ discharge, projects to the Vv. Terminal fields of the Vv were observed to be coextensive with afferent cell groups in the preoptic area, lateral and caudal hypothalamic nuclei, and thalamus. An additional efferent target of the Vv was the pretectal nucleus electrosensorius. That many cell groups that are connected with the Vv are also connected with the CP/PPn, particularly the preoptic and hypothalamic nuclei, suggests that the electrocommunicatory system is intimately linked with basal forebrain limbic pathways. J. Comp. Neurol. 389:49–64, 1997. r 1997 Wiley-Liss, Inc. Indexing terms: septum; hypothalamus; reproduction; medial forebrain bundle; teleost The basal forebrain and preoptic nuclei of teleost fishes have been regions of considerable interest due to their role in neuroendocrine control (Peter and Fryer, 1983) and the motivation of reproductive and aggressive behaviors (Demski and Knigge, 1971; see Demski, 1983, for a review; Satou et al., 1984; Fine and Perini, 1994), displaying many functional similarities to homologous regions in other vertebrates. Most anatomical studies on these regions in teleosts have focused on connections with the olfactory bulb (e.g., Bass, 1981b, 1981c; Levine and Dethier, 1985; Sas et al., 1993; Matz, 1995) or projections to the pituitary (Peter and Fryer, 1983; Johnston and Maler, 1992), whereas examination of other connections with these areas has been impaired, in part, due to their deep location and their proximity to the forebrain bundle. Previous experimental studies on the ventral and supracommissural nuclei of the ventral telencephalon (Vv/Vs; Shiga et al., 1985a, 1985b; r 1997 WILEY-LISS, INC. Sloan, 1989) have identified connections with the preoptic area, hypothalamus, and thalamus. However, as injections or lesions were not necessarily confined to a single nucleus, the specific connectivity of individual nuclei remains unknown. Recent tract-tracing studies in the weakly electric fish, E. virescens, identified reciprocal connections between the electrocommunicatory system and the ventral telencephalon, the preoptic area, as well as more caudal regions of the Grant sponsor: NIMH; Grant number: R37 MH26149-18; Grant sponsor: NINCDS; Grant number: RO1 NS22244-08; Grant sponsor: NSF; Grant number: BNS-9106705. *Correspondence to: Dr. Calvin J.H. Wong, Department of Biological Sciences, University of Alberta, Edmonton, Alberta, T6G 2E9 Canada. Received 11 February 1997; Revised 23 June 1997; Accepted 1 July 1997 50 C.J.H. WONG hypothalamus (Keller et al., 1990; Heiligenberg et al., 1991; Wong, 1997a). These findings corroborate previous immunohistochemical studies that identified peptidergic and monoaminergic innervation of diencephalic electrosensory/motor regions (see Zupanc and Maler, 1997, for a review). Gymnotiform fish produce an electric field, in the water around them, called the electric organ discharge (EOD; see Bass, 1986, for a review), of which they can modify the temporal and spectral characteristics for social communication (Hopkins, 1974; Hagedorn and Heiligenberg, 1985; Hopkins, 1988). Two diencephalic regions that act as a sensory-motor interface have been identified to be involved in the detection and production of electrocommunicatory signals. One is the pretectal complex of the nucleus electrosensorius (nE), which receives ascending input from the electrosensory portion of the torus semicircularis (Carr et al., 1981; Keller et al., 1990). Another region is the thalamic complex of the central posterior nucleus/ prepacemaker nucleus (CP/PPn; Zupanc and Zupanc, 1992; Zupanc and Heiligenberg, 1992), which receives input from the nE (Keller et al., 1990; Heiligenberg et al., 1991; Wong, 1997a). The CP/PPn then, in turn, projects to the medullary pacemaker nucleus (Heiligenberg et al., 1981; Kawasaki et al., 1988; Stroh and Zupanc, 1995; Heiligenberg et al., 1996; Wong, 1997a), which sets the timing of the EOD (Dye and Meyer, 1986). Stimulation of the CP/PPn can elicit modulations in the EOD, mimicking those that occur naturally (Kawasaki and Heiligenberg, 1988; Ka- wasaki et al., 1988), suggesting that other regions that project to the CP/PPn might also influence electrocommunicatory behavior. In an effort to expand our understanding of forebrain pathways involved in electrocommunication in gymnotiform fish, the organization and connections of one of the major telencephalic regions connected with the CP/PPn, the Vv, were examined. Some results from these experiments, regarding connections with the CP/ PPn, have been published previously (Wong, 1997a). MATERIALS AND METHODS Adult specimens of both sexes of Eigenmannia virescens were used for experiments (length 10–20 cm). Fish were obtained from a commercial supplier (Bailey’s Tropical Fish, San Diego, CA) and maintained in aquaria containing water of 20 to 30 kOhms/cm resistivity and neutral pH, at a temperature of 26–28°C. Fish were kept on a diurnal cycle of 12 hours light/12 hours dark and fed live blackworms ad libitum. Animal care, anesthesia, surgery, and killing were carried out in compliance with guidelines set out by the Animal Subjects Committee for the University of California, San Diego. Normal anatomy The normal anatomy of the telencephalon was examined from the reference series of E. virescens, from the library of the Heiligenberg laboratory. These series were transverse Abbreviations AC ATh cG CE CP/PPn CR-l-ir Dc Dd DFl Dl Dld Dlp Dlv Dm1 Dm2 Dm2v Dp Dpm DPn Ec EOD Er TS FB G Ha Hc Hd Hl Hv ICL JAR LCe LFB LL MFB MgT MOB nAPv anterior commissure anterior thalamic nucleus commissure of Goldstein central nucleus of the inferior lobe complex of the central posterior nucleus (thalamus) and the prepacemaker nucleus calretinin-like immunoreactivity central division of the dorsal telencephalon dorsal division of the dorsal telencephalon lateral diffuse nucleus of the inferior lobe dorsolateral telencephalon dorsolateral telencephalon, dorsal subdivision dorsolateral telencephalon, posterior subdivision dorsolateral telencephalon, ventral subdivision dorsomedial telencephalon, subdivision 1 dorsomedial telencephalon, subdivision 2 dorsomedial telencephalon, subdivision 2, ventral dorsoposterior telencephalon dorsoposterior telencephalon, medial subdivision dorsal posterior nucleus (thalamus) caudal entopeduncular nucleus electric organ discharge rostral entopeduncular nucleuse torus semicircularis efferents forebrain bundle glomerular nucleus anterior hypothalamic nucleus caudal hypothalamic nucleus dorsal hypothalamic nucleus lateral hypothalamic nucleus ventral hypothalamic nucleus internal cellular layer jamming avoidance response locus coeruleus lateral forebrain bundle lateral lemniscus medial forebrain bundle magnocellular tegmental nucleus medial olfactory bulb anterior periventricular nucleus nE nE< nLTa nPPv nRLl nT OB PC PeG PGl PGm PGr PLm POC PPa PPn PPp R Rd SPPn TA tBH TeO TP TPP TSd TSv Vc Vd Vi Vid Vir Vl VMTh Vp Vs Vv Vv-d Vv-v * nucleus electrosensorius nucleus electrosensorius, ‘down’subdivision anterior region of the lateral tuberal nucleus posterior periventricular nucleus lateral nucleus of the lateral recess nucleus taenia olfactory bulb posterior commissure periglomerular nucleus preglomerular complex, lateral subdivision preglomerular complex, medial subdivision preglomerular complex, rostral subdivision medial perilemniscal nucleus postoptic commissure anterior periventricular preoptic nucleus prepacemaker nucleus posterior periventricular preoptic nucleus red nucleus dorsal raphé nucleus sublemniscal prepacemaker nucleus anterior tuberal nucleus basal hypothalamic tract optic tectum posterior tuberal nucleus periventricular nucleus of the posterior tuberculum torus semicircularis, dorsal subdivision torus semicircularis, ventral subdivision ventral telencephalon, central nucleus ventral telencephalon, dorsal nucleus ventral telencephalon, intermediate nucleus ventral telencephalon, dorsointermediate nucleus ventral telencephalon, rostrointermediate nucleus ventral telencephalon, lateral nucleus ventromedial thalamic nucleus ventral telencephalon, posterior nucleus ventral telencephalon, supracommissural nucleus ventral telencephalon, ventral nucleus ventral telencephalon, ventral nucleus, dorsal subdivision ventral telencephalon, ventral nucleus, ventral subdivision nucleus of the zona limitans (thalami) BASAL FOREBRAIN CONNECTIONS IN GYMNOTIFORM FISH and sagittal sections (15-µm paraffin) of brains, either stained with the Nissl stain, cresyl violet, or with KlüverBerrera to reveal fiber tracts and a cresyl violet counterstain to reveal cell bodies. Because all experimental material consisted of 50-µm sections cut with a Vibratome and, therefore, had considerably different tissue shrinkage from paraffin material, the anatomical organization of the forebrain was further examined by using numerous cases of material (50-µm Vibratome sections) from other experiments that did not involve label of basal forebrain nuclei. These brains were counterstained with the Nissl stain, neutral red (Sigma Chemical Co., St. Louis, MO). The nomenclature used in the present study is that presented in the whole brain atlas of the related gymnotiform Apteronotus leptorhynchus (Maler et al., 1991). To confirm that the corresponding cell groups in both taxa were being assigned the same name, transverse sections of A. leptorhynchus were also studied. Immunohistochemistry To clarify further the anatomical organization of the basal forebrain region, an antibody to the calcium binding protein, calretinin, was used. Calretinin immunoreactivity has been shown to be differentially distributed among different populations of neurons (Baimbridge et al., 1992), particularly in the forebrain of other vertebrates (e.g., Jacobowitz and Winsky, 1991) and was, therefore, used as an additional tool for distinguishing cell groups of the ventral telencephalon. Recent biochemical studies in mormyrid and gymnarchid fishes identified an epitope recognized by this antibody, that is of the same molecular weight as mammalian calretinin and not calbindin (Friedman and Kawasaki, 1997). Although it is likely that this protein is calretinin, no functional role, in terms of calcium binding, or implication of homology with calretinin-like immunoreactive (CRl-ir) structures in other vertebrates will be considered in its present usage. It is stressed that the usage of this antibody is as a differential marker, like a Nissl stain or many histochemical stains. Three adult E. virescens were used for immunohistochemical studies. Animals were deeply anesthetized in tricaine methane sulfonate (MS-222; Sigma) and perfused transcardially with saline, followed by 30 ml of cold (4°C) 4% paraformaldehyde in 0.1 M phosphate buffer. Brains were removed and post-fixed at 4°C. Free-floating Vibratome sections (50 µm) were cut into 0.02 M phosphate buffered saline (PBS, pH 7.6) and rinsed three times, for 10 minutes each, in PBS. After rinsing, sections were placed in a blocking solution of 5% normal goat serum (NGS; Vector Laboratories, Burlingame, CA) in PBS with 0.3% Triton X-100 (PBST) for 1 hour. After blocking, sections were transferred to a 1% NGS solution in PBST containing the primary antibody (rabbit anti-calretinin antibody; SWant, Bellinzona, Switzerland) at 1:12,000 dilution. Sections were incubated for 2 days at 4°C in the primary antibody solution. After incubation, sections were rinsed three times, for 10 minutes each, in PBS, and then transferred to the secondary antibody solution. This consisted of 1% NGS in PBST with a biotinylated goat anti-rabbit IgG (Vector) at 1:300 dilution. After 2 hours of incubation in secondary antibody, sections were rinsed three times for 10 minutes in PBS, then reacted by the avidin-biotin/peroxidase protocol de- 51 scribed below. Alternate sections were used for immunohistochemistry and normal anatomy. By using a Vectastain avidin-biotin preoxidase complex kit (ABC; Vector), an ABC solution was prepared (4 drops A 1 4 drops B in 12 ml PBST). After 30 minutes of prebinding of the ABC solution, sections were incubated in the ABC solution in small, covered dishes at 4oC for 1 hour. Sections were then washed three times for 10 minutes in PBS and once for 10 minutes in Tris buffer (TB; 0.1 M, pH 7.2), then processed by the peroxidase/3,3(-diaminobenzidine (DAB) procedure. Sections were presoaked for 15 minutes in a solution of TB, and 0.04% DAB (Sigma). After presoak, H2O2 was added to a final concentration of 0.0018%. The reaction was allowed to proceed for 5 to 15 minutes, depending on the intensity of the background label. The reaction was stopped by washes in TB. Sections were washed at least three times for 10 minutes in TB and mounted on chrom-alum gelatin-coated slides. Tract tracing A total of 18 animals was used for tract tracing studies. Injections were made into the ventral telencephalon (6 cases), the preoptic area (2 cases), the hypothalamic nuclei (4 cases), and the nucleus electrosensorius (nE; 6 cases). Over 40 additional cases of injections into other structures (dorsal telencephalon, thalamus, torus, tectum, midbrain tegmentum) were analyzed as controls for label attributable to spread. Technical considerations of tract-tracing with Neurobiotin in E. virescens, particularly with regard to transneuronal transport, have been described in detail in a previous study (Wong, 1997a). Processing of tissue for Neurobiotin tract tracing was from a protocol modified from previous procedures (Kita and Armstrong, 1991; Lapper and Bolam, 1991; Huang et al., 1992). Anesthesia consisted of placing the fish in a 1:15,000 solution of MS-222 until somatic reactions stopped. This procedure was followed by immobilization with an intramuscular injection (2 µl of 2 mg/ml in saline) of the cholinergic blocker gallamine triethiodide (Sigma). The fish was then placed in a fish holder with aerated water perfusing the gills. The tail was placed in a plastic tube, and a wire inserted in the tube so that the EOD could be amplified and monitored. The surgical site was prepared with a topical application of 2 mg/ml lidocaine (Veterinary Companies of America, Tempe, AZ), and the skull was opened to expose the rostral pole of the brain. Injections were placed in brain nuclei, based on stereotaxic reference to the CP/PPn, which was initially localized by iontophoresing the excitatory amino acid, L-glutamate, and monitoring changes in the EOD (Kawasaki et al., 1988; Keller et al., 1990; Heiligenberg et al., 1991; Heiligenberg et al., 1996). As EOD modulations could be elicited by iontophoresis of L-glutamate into the nE, this nucleus was identified by its direct stimulation (Keller and Heiligenberg, 1989). The single-barrel electrode was replaced with a triple-barrel electrode consisting of two barrels filled with glutamate and one with the tracer, Neurobiotin (Vector). The Neurobiotin was made up at 2% in 0.2 µm filtered 1 M KCl. Tip resistances for the Neurobiotin electrodes were approximately 5 MOhms. Upon reidentification of the site, tracer was iontophoresed with positive DC current (ca. 12 µA, 10 seconds on, 15 seconds off) for 20 minutes up to 1 hour. After iontophoresis of Neurobiotin, the electrode was left in place for 5 to 10 minutes, and then slowly withdrawn with backing current 52 C.J.H. WONG to avoid leakage. The hole in the skull was patched with Gelfoam (Upjohn Company, Kalamazoo, MI), and the site was sealed with Vetbond (3M, St. Paul, MN). Survival times were 8 to 30 hours at 27°C for retrograde and anterograde transport of tracer. Fish were killed in MS-222 and perfused transcardially with 0.9% NaCl followed by 30 ml of fixative. The fixative consisted of 4% paraformaldehyde and 0.25% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4). Brains were then removed and post-fixed overnight. Free-floating Vibratome sections were processed by using the ABC method described above, with the following differences in the protocol: (1) after sectioning, sections were prebleached in 0.5% H2O2 in PBS for 10 minutes to inhibit endogenous peroxidases in the tissue; (2) sections were incubated in ABC solution overnight at 4°C; (3) additional intensification with 0.064% nickel ammonium sulfate was used; (4) every section was reacted, and subsequently counterstained, with neutral red. Chartings of representative cases were made by using a camera lucida. Brightfield photomicrographs were taken by using an Olympus BH-2 photomicroscope, with TMAX100 black and white film (Kodak, Rochester, NY). Contrast enhancement filters used for photomicrography were a Wratten 44 filter to enhance the counterstain and neutral density filters to adjust exposure. Prints were made on Kodak multigrade resin-coated (RC) paper or Ilford multigrade RC paper (Ilford Photo, Paramus, NJ). RESULTS Cytoarchitectural organization The midline nuclei of the basal forebrain of E. virescens are a complex series of nuclei, bound laterally by the forebrain bundle. The basal forebrain consists of several distinctive subdivisions. The nomenclature used for the telencephalon of ray-finned fishes is that of Nieuwenhuys (1963), as adapted to a related gymnotiform fish, A. leptorhynchus (Maler et al., 1991). Under this nomenclature, the telencephalon is parceled into a dorsal telencephalon (D) and a ventral telencephalon (V), with individual nuclei given positional descriptors. Generally, the dorsal telencephalon is recognized as comprising the pallium, whereas the ventral telencephalon is the subpallium. In the present study, I will address only the ventral nucleus of the ventral telencephalon (Vv) and its immediate surround. Precommissural regions. The Vv is a large, complex nucleus situated at the base of the telencephalon. At rostral levels (Fig. 1A1,B1), along the midline, the Vv can be observed as a wing-shaped nucleus extending dorsolaterally from the ependymal zone. The Vv extends from the dorsal, caudal aspect of the olfactory bulb to the level of the anterior commissure. It is capped by the dorsal nucleus of the ventral telencephalon (Vd), which can be observed as a series of cell laminae, appearing as arches parallel to the ventricular wall. The lateral extent of the Vv is bound by the forebrain bundle (FB). Dorsolateral to the Vv, a cup-shaped group of darkly staining cells, termed the central nucleus of the ventral telencephalon (Vc) can be recognized. At the lateral edge of the ventral telencephalon, a diffuse sheet of cells can be observed as the lateral nucleus of the ventral telencephalon (Vl). The Vv can be divided into a ventral subnucleus (Vv-v) and a dorsal subnucleus (Vv-d). The dorsal subnucleus is observed as a cigar-shaped stream of relatively small (5–8 µm), lightly staining, round cells, extending dorsolaterally from the ventricular wall. The Vv-d can be distinguished from the Vv-v, which generally appears as a more heterogeneous aggregate. Postcommissural regions. Further caudally (Fig. 1C1), the Vv is flanked along its ventral aspect by the supracommissural nucleus (Vs) that appears as a bed nucleus of the anterior commissure (AC). Because the supracommissural nucleus is intermingled with fibers of the anterior commissure and is bound rostrally by the bifurcation of a prominent blood vessel, the transition between the posterior Vv and anterior Vs can be difficult to discern. The Vs can be distinguished from the Vv in E. virescens, as consisting of slightly larger (10 µm) and diffusely scattered cells that stain relatively poorly for Nissl substance. In the present description, the Vs is considered to consist of both the region immediately dorsal to the commissure and the loosely scattered elements, situated around the horn of the commissure. At commissural and postcommissural levels, the posterior nucleus of the ventral telencephalon (Vp) can be observed dorsal to the Vd (Fig. 1C1,D1). The Vp can be identified as a ventrolaterally extending sheet of darkly staining cells. A loosely laminated, triangular region of smaller cells is observed ventromedial to the Vp and has been traditionally considered as part of the Vp (Maler et al., 1991). The Vv and Vs are replaced by the intermediate nucleus of the ventral telencephalon (Vi), characterized as scattered, diffuse elements extending dorsolaterally from the ventricular wall (Fig. 1D1). Ventral to the anterior commissure, the anterior periventricular preoptic nucleus (PPa) can be observed as a densely staining, tightly clustered, peri- and paraventricular cell group, bound by a lateral neuropil that is continuous with the medial forebrain bundle (Fig. 1C1). The caudal aspect of the PPa is displaced laterally by the laminar sheets of cells of the posterior periventricular preoptic nucleus (PPp) which is capped by the Vi (Fig. 1D1). Caudal to the PPp, the anterior hypothalamic nucleus (Ha) can be observed to eventually replace the dorsal aspect of the PPp (not shown). Calretinin immunohistochemistry The basal forebrain nuclei are richly endowed with calretinin-like immunoreactivity (CR-l-ir). In particular, the Vv, Vl, and Vc display deeply staining CR-l-ir perikarya (Fig. 1A2,B2). In contrast, the Vd displays a diffuse, terminal field of CR-l-ir with few immunoreactive cell bodies. The subregions of Vv can be recognized further by using CR-l-ir. The Vv-d appears as a dense cluster of immunoreactive perikarya, in contrast to the Vv-v, which has only a few scattered cells among a rich plexus of CR-l-ir fibers and terminals. At the level of the anterior commissure, CR-l-ir cells in the Vv are bound ventrally by the relatively CR-l-ir cell-poor supracommissural nucleus (Fig. 1C2). These are eventually replaced by the Vi, which, in general, is lacking in CR-l-ir, except for an occasional cell (Fig. 1D2). However, the Vp, can be recognized as having a dense meshwork of fibers, similar to the Vd. The PPa consists of a sheet of heavily immunoreactive cells that appear rostrally near the ventricle and are displaced caudolaterally by the PPp (Fig. 1C2, D2). The most rostral aspect of the BASAL FOREBRAIN CONNECTIONS IN GYMNOTIFORM FISH 53 Fig. 1. Cytoarchitecture and calretinin-like immunoreactivity of the basal forebrain of Eigenmannia virescens. A1,B1,C1,D1: Nisslstained transverse sections through the ventral telencephalon and preoptic area (50-µm thick, 200-µm intervals). A2,B2,C2,D2: Calreti- nin-like immunoreactivity in the section immediately caudal to the section presented in the left hand column. For abbreviations, see list. Scale bar 5 200 µm in D2 (applies to A1–D2). PPp appears to have few CR-l-ir cells, whereas the transitional zone with the Ha, appears to contain lightly immunoreactive cell bodies (not shown). the ventromedial aspect of the bulb constituting the terminal nerve ganglion (not shown). Telencephalon. In the dorsal telencephalon, retrogradely labeled cells were consistently observed in the posterior nucleus (Dp). These bipolar and multipolar cells were generally confined to the medial aspect of the Dp (Fig. 2F). In the ventral telencephalon, retrogradely labeled somata were observed ipsilateral to the injection site within the Vs (Figs. 2D, 3B). More caudally, a band of cells could be observed extending throughout the most caudal aspect of the dorsal intermediate nucleus (Vid; Figs. 2F, 4C). Labeled somata were not observed in the entopeduncular nuclei or other subnuclei of the ventral telencephalon with injections confined to the Vv. A bilateral projection was also observed from a sheet of cells of the nucleus taenia (nT), situated along the telencephalo-diencephalic junction (Figs. 2F, 4C). Rostrally, the nT appears immediately medial to the caudal entopeduncular nucleus. At more caudal levels, the labeled cells within nT cap the caudal entopeduncular nucleus and eventually appear subadjacent to the Dp. Afferents to the Vv Injections were placed in the rostral Vv. Of the six injections placed into the Vv, only three injection sites were discretely confined to the Vv without significant encroachment on adjacent areas. Figure 2 documents a global injection centered on and mostly confined to Vv. Although, in the following description, I present the extent of labeled cells and fibers from this global injection (Fig. 3A), cell groups considered afferents or efferents (Table 1) were assessed by smaller, confined injections and control injections. Most connections to and from the Vv were primarily ipsilateral with a weaker contralateral component. Olfactory bulb. The most rostral extent of labeled cells were mitral cells within the medial aspect of the olfactory bulb (Figs. 2A, 4A). In contrast, labeled cells were not observed in the lateral portion of the bulb after injections that were strictly confined to Vv. Additionally, relatively large cell bodies were observed bilaterally along Fig. 2. A–I: Camera lucida drawings of representative sections of label after a global injection of Neurobiotin centered on the ventral nucleus of the ventral telencephalon (Vv). Large dots indicate retrogradely labeled cell bodies; small dots indicate terminal fields. Broken lines indicate fibers. Asterisks in G, H indicate the zona limitans. Inset: side view of the brain of E. virescens. Numbers indicate the corresponding level for this figure and Figure 1. Fb, forebrain; Cb, cerebellum; ELL, electrosensory lateral line lobe. For abbreviations, see list. Scale bar 5 200 µm in I (applies to A–I), 1 mm in inset. BASAL FOREBRAIN CONNECTIONS IN GYMNOTIFORM FISH 55 Fig. 3. A: Photomicrograph of a Neurobiotin injection site into the Vv charted in Figure 2C. B: Labeled cells and fibers observed at the level of the anterior commissure. Note the heavy fiber label both within the anterior periventricular preoptic area (PPa) and the supracommissural nucleus (Vs). For abbreviations, see list. Scale bars 5 100 µm. Less consistently observed were labeled cells in the rostral-most region of the central nucleus of the dorsal telencephalon (Dc; Fig. 2B,C). Due to the proximity of the labeled cells in Dc to the rostral injection sites and that this cell group was not observed to be labeled with very discrete injections, it is likely that the labeled somata in Dc constitute spread of tracer from the injection site. Diencephalon and mesencephalon. Caudally, labeled cells were observed throughout the medial aspect of the preoptic area (Figs. 2D,E, 3B). Retrogradely labeled somata were observed primarily throughout the anterior periventricular preoptic nucleus (PPa) and the posterior periventricular preoptic nucleus (PPp). In contrast, more caudal aspects of the preoptic region, including the anterior hypothalamic nucleus (Ha) and anterior periventricular nucleus (nAPv), contained few retrogradely labeled somata (Fig. 2F). The lateral (Hl), ventral (Hv), and caudal (Hc) hypothalamic nuclei, and the anterior portion of lateral tuberal nucleus (nLTa) contained numerous labeled somata (Figs. 2G,H,I, 4D). Occasionally, a few cells were observed within the lateral nucleus of the lateral recess (nRLl). Other major afferent cell groups observed include the central posterior thalamic nucleus/prepacemaker nucleus (CP/PPn), the posterior tubercular nucleus (TP), and the locus coeruleus (LCe; not shown). Occasionally, cells were observed in the dorsal posterior thalamic nucleus (DPn), the ventral thalamus, the periventricular nucleus of the posterior tuberculum (TPP) and an isthmic TABLE 1. Connections of the Ventral Nucleus of the Ventral Telencephalon1 Eigenmannia virescens (this study) Afferents to the Vv Olfactory bulb Telencephalon Dorsal Telencephalon Ventral Telencephalon Diencephalon Hypothalamus Thalamus Posterior tuberculum Mesencephalon Olfactory bulb Telencephalon Diencephalon Hypothalamus Thalamus Pretectum 1For Carassius auratus (Sloan, 1989) Afferents to the Vv/Vs Terminal nerve, medial olfactory bulb Terminal nerve, olfactory bulb Dpm Vs, Vid, nT Dc, Dl Vc, Vd, Vp, Vi-nT PPa, PPp Ha, nAPv (minor) Hl, Hv, Hc nLTa, nRLl (minor) CP/PPn DPn (minor) VMTh (minor) TPP, TP LCe PLm Efferents from the Vv2 Medial olfactory bulb Vs (Vi, minor) PPa PPa, PPp (Ha, Hv, Hl, nRLl, Hc) DPn DP/PPn nE< (rest of nE, minor) Hd, Hc CP DPn TPP, TP LCe Raphe Efferents from the Vv/Vs Vc Dc, Dl PPa Habenula abbreviations, see list. 2Efferents indicate observed terminal fields. Due to label of axon collaterals and difficulty in mapping the precise topography with the hypothalamic nuclei, efferents that were not clearly identified are indicated with brackets. 56 C.J.H. WONG Fig. 4. Photomicrographs of afferents to the Vv. A: Retrogradely labeled mitral cells in the medial olfactory bulb (differential interference contrast optics). B: Retrogradely labeled somata in the dorsal intermediate nucleus of the ventral telencephalon (Vid). C: Labeled somata in nucleus taenia (nT). Note the heavy fiber label in the medial forebrain bundle (MFB; arrowheads) in contrast to the lateral forebrain bundle. D: Labeled cells (arrowheads) and fibers in the ventral hypothalamic nucleus (Hv) and lateral hypothalamic nucleus (Hl) in contrast to the anterior tuberal nucleus (TA). For abbreviations, see list. Scale bars 5 50 µm in A–C, 200 µm in D. nucleus, the medial perilemniscal nucleus (PLm; Maler et al., 1991). could be traced ventrally and caudally to innervate the preoptic area and a few fibers were observed to extend dorsolaterally toward the rostral aspect of the ventrolateral nucleus of the dorsal telencephalon (Dlv). However, this projection to the Dlv was very weak and not observed with confined injections into the Vv. A few fibers were also observed within the rostral entopeduncular nucleus (Fig. 3B). A dense plexus of terminals and fibers was observed in both the medial cellular region of the rostral preoptic area (PPa, PPp) and the lateral neuropil, including the medial forebrain bundle, which can be observed medial to the entopeduncular nucleus (Figs. 2D,E,3B). Further caudally, fibers could be traced within the medial forebrain bundle (MFB) to the transverse level of the anterior hypothalamus (Fig. 4C). At this level, a dorsal band of fibers (Fig. 2F) was observed to continue caudally to innervate the thalamus, in particular the dorsal neuropil between the CP/PPn and the DPn (Fig. 2H). A second band of labeled fibers appeared to course ventrally and interweave with the commissural fibers of the postoptic commissure (Fig. 2F). Two separate components of this band could be traced. A Efferents from the Vv After injections into Vv, anterogradely labeled fibers were observed, coextensive with many cell groups that contained retrogradely labeled somata. However, as Neurobiotin is transported in both retrograde and anterograde directions, it is unclear as to what degree labeled fibers constitute efferents from the Vv or anterograde label of collaterals of afferents to the Vv. Therefore, the description of labeled fibers will be presented followed by descriptions of reciprocally placed injections. The extent by which labeled fibers can be attributed to labeled efferent cells of the Vv, and not labeled collaterals of afferents to the Vv, are derived from these injections. Rostrally, labeled fibers were observed to exit the injection site toward the olfactory bulb. Caudally, labeled fibers were observed to exit from the injection site to the level of the commissure of Goldstein, where many of the fibers appeared to decussate, innervating homotopic structures of the contralateral hemisphere. However, some fibers BASAL FOREBRAIN CONNECTIONS IN GYMNOTIFORM FISH 57 Fig. 5. Anterogradely labeled fibers and terminals in the pretectal nucleus electrosensorius, down subdivision (nE<). Scale bar 5 50 µm. lateral component appeared to course dorsal to the preglomerular complex and innervate the pretectum (Figs. 2G, 5). A medial component could be traced along the lateral edge of the anterior tuberal nucleus to a prominent terminal field in the lateral and caudal hypothalamic nuclei (Figs. 2G,H,I, 4D). Some fibers continue caudally through the basal hypothalamic tract (tBH) toward the pituitary (not shown). A band of fibers could be observed ventral to the thalamus, near the zona limitans (Fig. 2G,H). This band of very fine labeled fibers could be traced along the dorsal aspect of the lateral recess to a terminal field in the lateral nucleus of the lateral recess (nRLl). Injections into the PPa/PPp Because heavily labeled cells and terminals were observed in the ipsilateral PPa/PPp, after discrete injections into Vv, two injections were made into the rostral preoptic area to further examine the organization of efferents from Vv. These injections were not exclusively confined to the PPa or PPp and hence will be described briefly. Injections centered into PPa/PPp retrogradely labeled many cell groups, similar to what was observed after injections into Vv. In considering these regions in which labeled cells were observed from both injections into the Vv and the PPa/ PPp, it is unclear as to what degree these cell groups project to both the Vv and the PPa/PPp, or the Vv only. It is possible that subsequent spread of tracer into the forebrain bundle from injections into the PPa/PPp might explain overlapping patterns in labeled afferents. Of these caudal regions that were observed to be retrogradely labeled from both injections into Vv and the PPa/PPp, only connections between the PPa/PPp with the CP/PPn were fully corroborated (Wong, 1997a). After injections of Neurobiotin centered in the PPa/PPp (Fig. 6), retrogradely labeled somata were observed both within the Vv-v and the Vv-d. Anterogradely labeled fibers were also observed in both Vv-v and Vv-d; however, labeled terminals appeared to be more prominent in the Vv-v (Fig. 7A). Two cell groups were identified that had retrogradely labeled cell bodies uniquely from injections into the PPa/ PPp and not injections confined to the Vv. One afferent cell group to the PPa/PPp was the dorsal hypothalamic nucleus (Hd; Fig. 7B). A second afferent cell group that was Fig. 6. Injection site into the preoptic area centered at the interface of the anterior and posterior periventricular preoptic nuclei (PPa/PPp). White arrow indicates electrode tracking. For abbreviations, see list. Scale bar 5 100 µm. observed was located in the midbrain tegmentum in the vicinity of the SPPn (Fig. 7D). These multipolar tegmental neurons were similar in morphology to those that were typically labeled after injections into the pacemaker; however, they were situated ventrolateral to the SPPn (indicated by asterisks in Fig. 7D). This finding suggests that these cells comprise a distinct subregion of the midbrain tegmentum and are not SPPn neurons proper. Similar to injections into the Vv, numerous anterogradely labeled terminals were observed in the CP/PPn coextensive with labeled cell bodies (Fig. 7C). Injections into the hypothalamic nuclei Injections were made in the hypothalamic nuclei to test hypotheses of connectivity with the preoptic area and the Vv. However, because none of the injections placed were specifically confined to a single hypothalamic nucleus and that fibers originating from hypophysiotrophic cell groups in the Vv and PPa/PPp course through the basal hypothalamic tract (Johnston and Maler, 1992), the specific organization of forebrain afferents to these various hypothalamic nuclei remains unknown. Briefly, after injections centered into the lateral hypothalamic nucleus encroaching on the anterior tuberal nucleus and nucleus of the lateral recess, both labeled cell bodies and fibers were observed in both subdivisions of the Vv and throughout the preoptic area. 58 C.J.H. WONG Fig. 8. A: An injection of Neurobiotin, centered on, but not confined to, the nE<. B: Observed retrogradely labeled cells in the Vv-v. Inset: High-powered photomicrograph of cells in Vv-v. Scale bars 5 100 µm in A,B, 25 µm in the inset. Injections into the nucleus electrosensorius As described above, a minor efferent projection was observed from the Vv to the pretectal nucleus electrosensorius. Although an occasional fiber was observed in most subdivisions (not shown), the most prominent pretectal target of the Vv was the ‘down’subdivision of the nE (nE<), and this projection was minor in comparison to fibers distributed through the thalamus, preoptic area, and hypothalamic nuclei. Because the nE< could be located by glutamate stimulation (Keller and Heiligenberg, 1989), precise injections were placed in this nucleus, and the presence of this connection was tested (Fig. 8A). Retrograde transport of Neurobiotin confirmed this minor projection and revealed that the cell bodies of origin were situated in the Vv-v (Fig. 8B). Control injections Fig. 7. Labeled cells and fibers after an injection in the rostral preoptic area. A: Labeled cells and fibers in Vv. Note the heavier innervation of the ventral subdivision of Vv (Vv-v). B: Retrogradely labeled somata in the dorsal hypothalamic nucleus. C: Retrogradely labeled somata and anterogradely labeled fibers in the dorsal aspect of the medial CP/PPn. Note terminal swellings within the CP/PPn (arrows). D: Retrogradely labeled somata (arrows) ventrolateral to the sublemniscal prepacemaker nucleus. Asterisks indicate boundaries of the SPPn. For abbreviations, see list. Scale bars 5 100 µm in A, 25 µm in B, 50 µm in C, 100 µm in D. Further work will be necessary to clarify the precise topographic organization of connections that the Vv and PPa/PPp share with individual hypothalamic nuclei. Due to the depth of the Vv and PPa/PPp, one major concern is that injections into these structures might have encroached onto overlying nuclei. Only three injections into the Vv were small enough so as to be specifically confined to the Vv, without any significant encroachment on the rest of the telencephalon. Therefore, subtractive analysis was used to further consider hypotheses of connections for larger injections. Control injections were placed into the telencephalon such that Vv or the PPa were specifically excluded from the injection site. The observed pattern of labeling is briefly described below. One control injection consisted of an injection centered on the overlying pallium, in the second subdivision of the dorsomedial telencephalon (Dm2; Fig. 9A). In this case, there were no labeled cells observed in the CP/PPn, nor fibers observed in the nE or medial preoptic area. Anterogradely labeled fibers were observed within the most lateral aspect of the lateral hypothalamic nucleus. However, the ventral hypothalamic nucleus as well as more medial portions of the lateral hypothalamic nucleus in which labeled fibers and cells were more typically observed after discrete injections into Vv or the PPa (e.g., Fig. 4D), were not observed to be labeled with an injection centered on the Dm2. A second control injection consisted of a large injection of Neurobiotin, centered on, although not confined to, the Vd with only minor encroachment on the Vv (Fig. 9B). In this BASAL FOREBRAIN CONNECTIONS IN GYMNOTIFORM FISH 59 Fig. 9. Camera lucida drawings of the center of injection sites for control injections. See text for details. A: Injection of Neurobiotin centered on the second subdivision of the dorsomedial telencephalon (Dm2). B: Injection of Neurobiotin centered on, although not confined to, the dorsal nucleus of the ventral telencephalon (Vd). For abbreviations, see list. Scale bar 5 200 µm. case, there were only a few labeled somata observed in the preoptic area, the CP/PPn, and the nT, which contrasted with injections that were centered more ventrally. There were numerous labeled cells in the intermediate nucleus of the ventral telencephalon; however, these were accompanied by dense labeling of fibers, in contrast to injections centered in the Vv in which there was very sparse anterograde labeling of fibers in the Vi. Additionally there were no retrogradely labeled somata observed in the olfactory bulb, the Vs, or the hypothalamic nuclei. Retrogradely labeled cells were observed in the central nucleus of the dorsal telencephalon, after injections of tracer into the torus semicircularis and optic tectum. Therefore, it is unclear as to whether observed labeled cells in Dc actually project to Vv or whether interruption of fibers of passage may account for labeled cells in Dc. DISCUSSION The results of the present study are summarized in Figures 10, 11, and in Table 1. The ventral nucleus of the ventral telencephalon of E. virescens (Vv) is heavily and reciprocally connected with the rostral nuclei of the preoptic area (PPa, PPp), the hypothalamic nuclei (nLTa, Hl, Hv, Hc, nLTa, nRLl), and as reported previously (Wong, 1997a), the thalamus (CP/PPn). Other inputs originate from the olfactory bulb and terminal nerve, caudal telencephalon (Dpm, Vid, nT), posterior tuberculum (TPP, TP), the locus coeruleus (LCe), and the medial perilemniscal nucleus (PLm). Moreover, subdivisions of the Vv can be recognized based upon cytoarchitectonic, chemoarchitectonic, and connectional criteria. Subdivisions of the ventral nucleus of the ventral telencephalon The present study recognizes all the major subdivisions of the ventral telencephalon and rostral preoptic area that have been identified in other teleosts (see Northcutt and Braford, 1980; Northcutt and Davis, 1983, for a review). In E. virescens, the Vv consists of at least two recognizable subdivisions. Only one other study, that of Bass (1981a) on the channel catfish, Ictalurus punctatus specifically addresses subdivisions of Vv and the present interpretations differ somewhat from that of Bass (1981a). The two subdivisions of Vv identified by Bass (1981a) cannot be Fig. 10. Organization of connections that are topographically organized with respect to the dorsal (Vv-d) and ventral subdivisions of the Vv (Vv-v). A: Afferent projections to the Vv. Relative strength of projections is indicated by thickness of the lines. B: Efferent projections of the Vv. Same presentation as in A. Other projections for which the topographic organization is unknown are listed in Table 1. See text for details. For abbreviations, see list. recognized in E. virescens, and therefore the Vv-v in E. virescens likely corresponds to both subdivisions of Vv recognized by Bass (1981a). Based on the position of olfactory terminal fields in the related gymnotiform A. leptorhynchus (see Sas et al., 1993, for details), it is likely that the dorsal subnucleus of the Vv (Vv-d) in the present study corresponds to the ventral subnucleus of Vd, recognized in the channel catfish (Vd-v; Bass, 1981a). It would also appear to correspond to the Vs of Levine and Dethier (1985), in the common goldfish, Carassius auratus. Although the secondary olfactory projections of E. virescens were not examined in the present study, based on cytoarchitecture, similar subdivisions can be recognized to be present in A. leptorhynchus. In addition to cytoarchitectonic criteria, chemoarchitecture distinguishes subdivisions of the Vv. The Vv-d can be 60 C.J.H. WONG recognized as being distinctive from both the Vv-v and the Vd in having tightly clustered CR-l-ir perikarya. This contrasts with the Vv-v, which has a high density of periventricular immunoreactive somata and only scattered CR-l-ir somata laterally. Moreover, it also contrasts with the Vd, which contains primarily meshwork of fine, immunoreactive fibers and terminals. In A. leptorhynchus, the Vv-d is extremely rich in substance-P immunoreactive fibers and terminals in contrast to the Vv-v and the Vd (Weld and Maler, 1992). Moreover, the Vv-d appears to be more heavily innervated by enkephalinergic fibers (Richards and Maler, 1996), further corroborating the heterogeneous nature of the Vv. Connectional evidence distinguishes the two subdivisions of the Vv. The Vv-d and to a lesser degree, the ventral aspect of the Vd, share prominent reciprocal connections with the thalamic CP/PPn (see Fig. 3a of Wong, 1997a) and appear to overlap with the medial olfactory terminal field (Sas et al., 1993). Moreover, the Vv-d provides a source of afferents to the olfactory bulb (Sas et al., 1993). In contrast, the Vv-v receives heavy input from the PPa/PPp, whereas the Vv-d receives a somewhat weaker projection. The Vv-v appears to be only weakly connected with the thalamus (Wong, 1997a). Consideration of the transitional zone between the Vv and the Vs is more problematic. The original cytoarchitectonic description of Vs was merely as the supracommissural extension of the Vd (Nieuwenhuys, 1963). Therefore, the Vs would be situated dorsal and caudal to the Vv (Northcutt and Braford, 1980; Northcutt and Davis, 1983). However, reexamination of the application of the term Vs to teleosts (Nieuwenhuys, 1963) reveals that this designation grouped together several nuclei that had been recognized as distinctive by previous authors. The present study considers the Vs as a bed nucleus of lightly staining cells situated ventral to the Vv (see Results). The location is in agreement with the atlas of the related gymnotiform A. leptorhynchus (Maler et al., 1991) as well as the atlas of the goldfish (Peter and Gill, 1975). It would appear to correspond to the bed nucleus of the commissura hippocampi of Schnitzlein (1962). Technical considerations to tract-tracing experiments The Vv receives input from the olfactory bulb (Bass, 1981b; Levine and Dethier, 1985; Sas et al., 1993; Matz, 1995) and has cells that project to the pituitary (Johnston and Maler, 1992). Due to the proximity of the injection sites to the forebrain bundle, one major concern is that damaged fibers of passage might have taken up tracer. Therefore, I discuss the following considerations of alternative hypotheses regarding observed patterns of labeled cell groups. Some cell groups, that were observed to be retrogradely labeled after injections into the Vv, also project to the bulb. Therefore, labeled cells might have resulted from interruption of bulbopetal fibers (e.g., Vs, Dpm, nT). However, many cell groups that are afferent to the bulb were not observed to be retrogradely labeled, after discrete injections of tracer into Vv. These include the Dc shell, the rostral aspect of the Dp, and the dorsomedial telencephalon (Sas et al., 1993). These data suggest that the contribution of labeled cells, attributable to interruption of bulbopetal fibers, was probably minor. Two major efferent targets of the Vv were the Vs and the anterior preoptic area (PPa, PPp). The centromedial subdivision of the medial olfactory tract, that carries bulbofugal fibers to the ventral telencephalon and possibly the preoptic area (Bass, 1981b; Levine and Dethier, 1985; although see Sas et al., 1993; Matz, 1995), courses near the Vv. Thus the bulbofugal fibers to these regions may have been interrupted by the present injections. Evidence against such an interpretation is as follows: (1) both the habenular commissure and the anterior commissure carry commissural fibers of bulbofugal cells, however, generally lacked any labeled fibers from injections confined to the Vv; (2) many of the cell groups that receive bulbofugal input from the medial olfactory tract, contained only very sparse label of fibers or terminals (e.g., Vir, Vc); and (3) at supracommissural levels, the medial olfactory terminal field was observed to be relatively lightly labeled, possibly from labeling of collaterals, in contrast to the heavy terminal labeling in the Vs and the PPa (e.g., Figs. 2D, 3B). The presence of heavy anterogradely labeled fibers in the Vs, PPa, and PPp, in contrast to what was observed in other bulbofugal cell groups, after injections into Vv, suggests that the Vv does project to the Vs, PPa, and PPp. A last confound is that labeled cells and fibers observed within the Vv, after injections into the PPa/PPp, may be due to interruption of afferents from and efferents to Vv. There are two pieces of evidence against such an interpretation. First, after injections into the PPa/PPp, most anterogradely labeled fibers were observed within the Vv-v, in contrast to the Vv-d. The Vv-d receives a heavy thalamofugal projection (Wong, 1997a), and it is clear that labeled fibers in the Vv-d were few in comparison, after relatively confined injections into the PPa (Fig. 7A). A second argument against such an interpretation is the high density of cellular, terminal, and fiber labeling in the PPa/PPp after injections into the Vv. This finding suggests that the Vv and the PPa are heavily interconnected. Because there was considerable overlap in cell groups that were retrogradely labeled from injections either into PPa or Vv (Dpm, nT, hypothalamic nuclei), it remains unclear as to whether these nuclei project to Vv exclusively or also project to the PPa/PPp in addition to the Vv. Injections of anterograde tracer into these regions will be necessary to delineate the full extent of overlap in afferent cell groups to the basal forebrain and preoptic area. Anterograde confirmation of thalamic afferents to both the PPa/PPp and basal forebrain have been demonstrated previously (Wong, 1997a). It is also possible that individual cells may project to both the Vv and PPa/PPp, further coordinating these two areas. This hypothesis will need to be tested by double-label retrograde studies. Other concerns of spread were further tested by control injections centered into overlying structures. These injections revealed very different patterns of labeled cells and fibers in comparison to what was observed after injections centered in the Vv (see Results). Specifically, labeled cells were not observed in the preoptic area, ventral, lateral, or caudal hypothalamic nuclei, nE, or CP/PPn. Labeled cells were observed in the Dc after injections into the torus and tectum, as has been reported in other teleost fishes (Ito and Kishida, 1977; Murakami et al., 1983; Echteler, 1984). Therefore, it is unclear whether Dc actually projects to the Vv or whether interruption of fibers of passage may account for the observed labeled cells. BASAL FOREBRAIN CONNECTIONS IN GYMNOTIFORM FISH Afferents to the ventral telencephalon The connections identified in the present study are in good agreement with those identified previously in the goldfish (Sloan, 1989). Some connections have also been identified in the land-locked salmon, although identification of correspondence between specific cell groups is limited by differences in cytoarchitectonic delineation (Oncorhynchus nerka; Shiga et al., 1985a, 1985b). These results are summarized in Table 1. It is possible that the differences in connectivity may be due to species-specific differences or methodological differences such as the sensitivity of the tracer used, and the size, placement, and interpretation of injection sites. In E. virescens, after injections of tracer confined to Vv, retrogradely labeled cells were observed within the medial aspect of the olfactory bulb. Only after injections that were clearly not confined to the Vv were retrogradely labeled cells observed in the lateral portion of the bulb. This finding reinforces the notion that the olfactory bulb is a heterogeneous structure consisting of at least a medial and lateral subdivision based upon connectional (Bass, 1981b; Levine and Dethier, 1985; Sas et al., 1993; Matz, 1995) and immunohistochemical (Weld and Maler, 1992; Sas et al., 1993) criteria. In the present study, retrogradely labeled cells were observed bilaterally in the olfactory bulb, confirming the bilateral nature of bulbar efferents to the ventral telencephalon (Bass, 1981b; Sas et al., 1993). In contrast, in the common goldfish (Sloan, 1989), retrogradely labeled somata were observed in only the ipsilateral bulb after injections into the ventral telencephalon. However, it is possible that the differences may be due to the increased sensitivity of tracer detection for Neurobiotin versus the horseradish-peroxidase (HRP) tracing technique used previously (Sloan, 1989). Indeed, previous investigators identified a bilateral bulbofugal projection to Vv in the goldfish after injections into the olfactory bulb (Levine and Dethier, 1985). Several large round somata were retrogradely labeled along the ventromedial aspect of the bulb. Although the axons from these cells were not traced to the retina, it is likely that these larger cells are ganglion cells of the terminal nerve (Von Bartheld and Meyer, 1986). Telencephalic afferents to the Vv in E. virescens are very restricted (Vid, Vs, Dpm, and nT), although additional afferents have been reported in the goldfish (Sloan, 1989; Vd, Vc, Vp, Dl, Dc). However, as connections reported previously include afferents to Vs (Sloan, 1989), it is possible that the differences may be due to inclusion of the Vs in injection sites in studies on the goldfish (Sloan, 1989). The present study does suggest that the Vv and Vs are intimately related as has been described in other fishes (Shiga et al., 1985a, 1985b; Sloan, 1989). Additionally, in the present study, as labeled cells were not observed in Dc after discrete injections into the Vv, labeled somata in the rostral aspect of Dc are considered to be due to interruption of their descending fibers. The Dp and the nT are also reciprocally connected with the olfactory bulb. Although the specific functions of these regions are unknown, the Dp, which is considered a major olfacto-recipient area, is hypertrophied in gymnotiform fish, which is surprising given the small size of the olfactory bulb (Maler et al., 1991; Striedter, 1992; Sas et al., 1993). Noting the hypertrophy of other structures related to electrosensory/motor processing in gymnotiform 61 fish, it has been proposed that the Dp may play a role in electroreception (Maler et al., 1991; Sas et al., 1993). However, the existence of connections between the Dp and electrosensory/motor areas remains to be determined. The present study is in good agreement with previous studies with regard to diencephalic afferent cell groups to the Vv, although there may be some variation due to differences in cytoarchitectonic descriptions and injection placement. A major afferent projection from the preoptic area to the Vv has been described in the goldfish (Sloan, 1989), as have inputs from the hypothalamic nuclei, the dorsal thalamus, and the nucleus of the posterior tuberculum. Similarly, many of these projections (preoptic area, dorsal thalamus, posterior tubercular nucleus) appear similar to projections to the Vs in the land-locked salmon, although it is likely that the Vv was also included in the injection site (Shiga et al., 1985a). In the channel catfish, Striedter (1991) observed labeled cells and fibers in the Vv after application of tracer to the TA. However, neither labeled cells nor fibers were observed in the TA after injections into the ventral telencephalon in any fish (Shiga et al., 1985a; Sloan, 1989; this study). It is likely that this discrepancy is due to differences in interpretation of tracer spread. In both E. virescens (this study) and the goldfish (Sloan, 1989), a number of hypothalamic nuclei situated in close proximity to the TA are reciprocally connected with the Vv. In contrast to what has been reported in the goldfish (Sloan, 1989), projections with the habenula were not observed in E. virescens. However, a recent retrograde study in the goldfish showed that the Vd projects to the habenula (Villani et al., 1996). Functional considerations Besides heavy interconnectivity with each other and hypothalamic areas, the Vv and PPa/PPp of E. virescens are interconnected with the electrocommunicatory system. The Vv and preoptic area share reciprocal connections with the rostral, medial aspect of the CP/PPn (Wong, 1997a). Moreover, many hypothalamic nuclei, that had both labeled cells and fibers after injections of Neurobiotin into the Vv (PPa, PPp, Hl, Hc), are also reciprocally connected with the CP/PPn (pathway 1 in Fig. 11). These results complement and extend previous immunohistochemical studies on a related gymnotiform, A. leptorhynchus, that identified extensive peptidergic and monoaminergic fibers in both the Vv and the CP/PPn whose cell bodies of origin were hypothesized to be hypothalamic (see Johnston and Maler, 1992, for a review). Moreover, many of the cell groups identified to be connected with both the Vv and the CP/PPn also have cells that project to the pituitary (Johnston and Maler, 1992). Although further studies will be necessary to determine specifically the extent of overlap in individual cells of the hypothalamic nuclei that are connected with the CP/PPn, the basal forebrain, and the pituitary, the current study suggests that the Vv, rostral preoptic area, lateral and caudal hypothalamic nuclei, and CP/PPn constitute a network of intimately associated cell groups. A minor projection has also been identified from the Vv to the pretectal nE. Previous studies reported observing terminals in the nE, after massive injections of tracer into the telencephalon (Keller et al., 1990). The present study confirms this observation and identifies the cell bodies of origin to be specifically the Vv-v. 62 C.J.H. WONG Fig. 11. Relationship of cell groups, identified in the present study to be connected with the Vv (heavy arrows), cell groups connected with the CP/PPn (1) and cell groups that project to the pituitary (2). See text for details. For abbreviations, see list. Although one or two fibers were observed in other subnuclei of the nE, the most prominent subnucleus that receives input from the Vv was the nE<. Stimulation of the nE< in E. virescens elicits gradual decreases in EOD frequency (Keller and Heiligenberg, 1989), mediated by an inhibitory projection to the mesencephalic electromotor nucleus, the SPPn (Metzner, 1993). Besides a role in the jamming avoidance response (JAR) (Heiligenberg, 1991), gradual decreases in frequency of the EOD appear to be involved in social interactions in E. virescens (Hagedorn and Heiligenberg, 1985). In other gymnotiform fishes, the SPPn also mediates a variety of other modulations in the EOD (Keller et al., 1991; Kawasaki and Heiligenberg, 1989; Spiro, 1994; Heiligenberg et al., 1996). Further work will be necessary to determine the functional role of this descending projection from Vv to the nE<. The behavioral functions of the Vv and PPa have been examined in detail in a variety of teleost fishes (reviewed in Demski, 1983). In the goldfish, lesions of the posterior Vv and Vs severely impair male spawning behavior (Kyle and Peter, 1982; Kyle et al., 1982). Moreover, stimulation studies in many fishes have shown that the rostral preoptic area may be involved in the production of courtship displays (Demski and Knigge, 1971; Satou et al., 1984; Fine and Perini, 1994). Behavioral studies in E. virescens revealed that the EOD, which is normally at a fixed frequency, is modulated during courtship and agonistic encounters (Hopkins, 1974; Hagedorn and Heiligenberg, 1985). One prominent modulation produced, primarily by male fish, are brief interruptions (ca. 10–100 ms) in the EOD, which may serve as courtship signals to female fish. In E. virescens, electrical stimulation of the preoptic area elicits interruptions in the EOD (Wong, 1997b), suggesting that these basal forebrain pathways may also be involved in the production of electrocommunicatory signals. Connections with the medial olfactory bulb lend further credence to the notion that the basal forebrain may be involved in reproductive/ communicatory behavior. The medial olfactory tract, in the goldfish, has been implicated in the relaying of sex pheromonal information to the ventral telencephalon (see Dulka, 1993, for a review). Therefore, it will be of interest to determine whether gymnotiforms use pheromonal signals and if so, whether the Vv-d and Vs might integrate such signals with ascending thalamic electrosensory information (Wong, 1997a). Comparative aspects The present connectional results do not clarify specific homologies between the Vv and individual septal or basal forebrain nuclei of other vertebrates, and it is possible that such a one-to-one correspondence may not exist (Reiner and Northcutt, 1992; Northcutt, 1995). However, the Vv of E. virescens does share many connectional and functional similarities to the septal region of other vertebrates. One major cell group connected with the septal nuclei in other vertebrates is the amygdala (amphibians: Neary and Northcutt, 1990; rats: Krettek and Price, 1978; Swanson and Cowan, 1979). Labeled cells and fibers were observed within Vs, which may correspond to the amygdala, based upon topological and chemoarchitectonic criteria (Reiner and Northcutt, 1992; Northcutt, 1995; although, see Sas et al., 1993; Weld et al., 1994). Another major afferent region to the septum is the medial pallium (Neary, 1990; Northcutt and Ronan, 1992) or hippocampal formation in mammals (Swanson and Cowan, 1979). However, as the telencephalic hemispheres of ray-finned fishes develops by eversion and not evagination as in other vertebrate classes, recognition of homologous telencephalic cell populations between teleosts and other vertebrates is impaired (see Nieuwenhuys and Meek, 1990; Braford, 1995; Northcutt, 1995, for recent reviews). Although the telencephalic cell groups afferent to the Vv (Vid, nT, Dpm) do not readily correspond to the proposed topological homologue in teleost fishes to the medial pallium, the dorsolateral telencephalon (Northcutt, 1995), further examination of the connections of these regions will be necessary to fully assess their evolutionary relationship to the medial pallium. The most striking similarity between the teleost Vv and the septal nuclei of other vertebrates is the presence of connections with cell groups associated with the MFB. Like the septal nuclei of other vertebrates, the Vv is heavily connected with the preoptic area and hypotha- BASAL FOREBRAIN CONNECTIONS IN GYMNOTIFORM FISH lamic nuclei (teleost fish: Shiga et al. [1985a, 1985b]; Sloan ; this study; amphibians: Neary and Northcutt ; Neary ; reptiles: Bruce and Neary [1995a, 1995b]; birds: Berk and Butler ; Székely et al. ; mammals: e.g., Swanson and Cowan ). Moreover, there appear to be functional relationships in MFB cell groups. In a related gymnotiform fish, Sternopygus macrurus, the basal forebrain, preoptic area, and hypothalamic nuclei are immunoreactive for sex steroid receptors (Gustavson et al., 1994). Therefore, similar to what has been characterized in the rat (Cottingham and Pfaff, 1986), there appears to be a rich interconnectivity of MFB cell groups that are target regions for steroid action in gymnotiforms. Indeed among all classes of vertebrates, hormone-targeted cell groups appear to be the basal forebrain and preoptic nuclei in close association with gonadotrophin-releasing hormone (GnRH) systems (Demski, 1984). Cell groups that are connected with the Vv also have neurosecretory cells (PPa, PPp, Hl, Hc, Hv, nLTa), and have both cell bodies and fibers that are richly endowed with neuropeptides and monoamines (Johnston and Maler, 1992). Given the extensive anatomical overlap in MFB cell groups associated with neuroendocrine function, reproductive behavior, and electrocommunication, it will be of interest to determine their functional relationships and whether these might be mediated by transmitter-specific pathways. ACKNOWLEDGMENTS I thank Dr. R.G. Northcutt for engaging discussions and for providing the calretinin antibody used in these studies. Dr. T.H. Bullock, P. Holmes, Dr. C.H. Keller, Dr. R.G. Northcutt, and two anonymous reviewers provided helpful comments on this manuscript. Grace Kennedy and Andrea Thör provided excellent histological assistance. Additional thanks to Dr. T.H. Bullock, Dr. J.B. Graham, Dr. R.R. Hessler, the SIO Director’s Office, the NIH, the NIMH, and the NSF for allowing the Heiligenberg lab to remain open for the completion of this work. This work was supported by NIMH grant R37 MH26149-18, NINCDS grant RO1 NS22244-08, and NSF BNS-9106705 to Dr. W. Heiligenberg. LITERATURE CITED Baimbridge, K.G., M.R. Celio, and J.H. Rogers (1992) Calcium-binding proteins in the nervous system. Trends Neurosci. 15:303–308. Bass, A.H. (1981a) Organization of the telencephalon in the channel catfish, Ictalurus punctatus. J. Morphol. 169:71–90. Bass, A.H. (1981b) Olfactory bulb efferents in the channel catfish, Ictalurus punctatus. J. Morphol. 169:91–111. Bass, A.H. (1981c) Telencephalic efferents in the channel catfish, Ictalurus punctatus: Projections to the olfactory bulb and optic tectum. Brain Behav. Evol. 19:1–16. Bass, A.H. (1986) Electric organs revisited: Evolution of a vertebrate communication and orientation organ. In T.H. Bullock and W. Heiligenberg (eds): Electroreception. New York: John Wiley & Sons, pp. 13–70. Berk, M.L., and A.B. Butler (1981) Efferent projections of the medial preoptic nucleus and medial hypothalamus in the pigeon. J. Comp. Neurol. 203:379–399. Braford, M.R., Jr. (1995) Comparative aspects of forebrain organization in the ray-finned fishes: Touchstones or not? Brain Behav. Evol. 46:259– 274. Bruce, L.L., and T.J. Neary (1995a) Afferent projections to the ventromedial hypothalamic nucleus in a lizard, Gekko gecko. Brain Behav. Evol. 46:14–29. 63 Bruce, L.L., and T.J. Neary (1995b) Afferent projections to the lateral and dorsomedial hypothalamus in a lizard, Gekko gecko. Brain Behav. Evol. 46:30–42. Carr, C.E., L. Maler, W. Heiligenberg, and E. Sas (1981) Laminar organization of the afferent and efferent systems of the torus semicircularis of gymnotiform fish: Morphological substrates for parallel processing in the electrosensory system. J. Comp. Neurol. 203:649–670. Cottingham, S.L., and D. Pfaff (1986) Interconnectedness of steroid hormonebinding neurons: Existence and implications. In D. Ganten and D. Pfaff (eds): Morphology of Hypothalamus and Its Connections. Berlin: Springer-Verlag, pp. 223–249. Demski, L.S. (1983) Behavioral effects of electrical stimulation of the brain. In R.E. Davis and R.G. Northcutt (eds): Fish Neurobiology. Ann Arbor: University of Michigan Press, pp. 317–359. Demski, L.S. (1984) The evolution of neuroanatomical substrates of reproductive behavior: Sex steroid and LHRH-specific pathways including the terminal nerve. Am. Zool. 24:809–830. Demski, L.S., and K.M. Knigge (1971) The telencephalon and hypothalamus of the bluegill (Lepomis macrochirus): Evoked feeding, aggressive, and reproductive behavior with representative frontal sections. J. Comp. Neurol. 143:1–16. Dulka, J.G. (1993) Sex pheromone systems in goldfish: Comparisons to vomeronasal systems in tetrapods. Brain Behav. Evol. 42:265–280. Dye, J.C., and J.H. Meyer (1986) Central control of the electric organ discharge in weakly electric fish. In T.H. Bullock and W. Heiligenberg (eds): Electroreception. New York: John Wiley & Sons, pp. 71–102. Echteler, S.M. (1984) Connections of the auditory midbrain in a teleost fish, Cyprinus carpio. J. Comp. Neurol. 230:536–551. Fine, M.L., and M.A. Perini (1994) Sound production evoked by electrical stimulation of the forebrain in the oyster toadfish. J. Comp. Physiol. A 174:173–185. Friedman, M.A., and M. Kawasaki (1997) Calretinin-like immunoreactivity in mormyrid and gymnarchid electrosensory and electromotor systems. J. Comp. Neurol. 387:341–357. Gustavson, S., H. Zakon, and G. Prins (1994) Androgen receptors in the brain, electroreceptors and electric organ of a wave-type electric fish. Soc. Neurosci. Abstr. 20:371. Hagedorn, M., and W. Heiligenberg (1985) Court and spark: Electric signals in the courtship and mating of gymnotoid fish. Anim. Behav. 33:254– 265. Heiligenberg, W. (1991) Neural Nets in Electric Fish. Cambridge, MA: MIT Press. Heiligenberg, W., T. Finger, J. Matsubara, and C.E. Carr (1981) Input to the medullary pacemaker nucleus in the weakly electric fish Eigenmannia (Sternopygidae, Gymnotiformes). Brain Res. 211:418–423. Heiligenberg, W., C.H. Keller, W. Metzner, and M. Kawasaki (1991) Structure and function of neurons in the complex of the nucleus electrosensorius of the gymnotiform fish Eigenmannia: Detection and processing of electric signals in social communication. J. Comp. Physiol. A 169:151–164. Heiligenberg, W., W. Metzner, C.J.H. Wong, and C.H. Keller (1996) Motor control of the jamming avoidance response of Apteronotus leptorhynchus: Evolutionary changes of a behavior and its neuronal substrates. J. Comp. Physiol. A 179:653–674. Hopkins, C. (1974) Electric communication: Functions in the social behaviour of Eigenmannia virescens. Behaviour 50:270–305. Hopkins, C.D. (1988) Neuroethology of electric communication. Annu. Rev. Neurosci. 11:497–535. Huang, Q., D. Zhou, and M. DiFiglia (1992) Neurobiotin, a useful neuroanatomical tracer for in vivo anterograde, retrograde and transneuronal tract-tracing and for in vitro labeling of neurons. J. Neurosci. Methods 41:31–43. Ito, H., and R. Kishida (1977) Tectal afferent neurons identified by the retrograde HRP method in the carp telencephalon. Brain Res. 130:142– 145. Jacobowitz, D.M., and L. Winsky (1991) Immunocytochemical localization of calretinin in the forebrain of the rat. J. Comp. Neurol. 304:198–218. Johnston, S.A., and L. Maler (1992) Anatomical organization of the hypophysiotrophic systems in the electric fish, Apteronotus leptorhynchus. J. Comp. Neurol. 317:421–437. Kawasaki, M., and W. Heiligenberg (1988) Individual prepacemaker neurons can modulate the pacemaker cycle of the gymnotiform electric fish Eigenmannia. J. Comp. Physiol. A 162:13–21. 64 Kawasaki, M., and W. Heiligenberg (1989) Distinct mechanisms of modulation in a neuronal oscillator generate different social signals in the electric fish Hypopomus. J. Comp. Physiol. A 165:731–741. Kawasaki, M., L. Maler, G.J. Rose, and W. Heiligenberg (1988) Anatomical and functional organization of the prepacemaker nucleus in gymnotiform electric fish: The accommodation of two behaviors in one nucleus. J. Comp. Neurol. 276:113–131. Keller, C.H., and W. Heiligenberg (1989) From distributed sensory processing to discrete motor representations in the diencephalon of the weakly electric fish Eigenmannia. J. Comp. Physiol. A 164:565–576. Keller, C.H., M. Kawasaki, and W. Heiligenberg (1991) The control of pacemaker modulations for social communication in the weakly electric fish Sternopygus. J. Comp. Physiol. A 169:441–450. Keller, C.H., L. Maler, and W. Heiligenberg (1990) Structural and functional organization of a diencephalic sensory-motor interface in the gymnotiform fish Eigenmannia. J. Comp. Neurol. 293:347–376. Kita, H., and W.E. Armstrong (1991) A biotin-containing compound N-(2aminoethyl) biotinamide for intracellular labeling and neuronal tracing studies: Comparison with biocytin. J Neurosci. Methods 37:141–150. Krettek, J.E., and J.L. Price (1978) Amygdaloid projections to subcortical structures within the basal forebrain and brainstem in the rat and cat. J. Comp. Neurol. 178:225–254. Kyle, A.L., and R.E. Peter (1982) Effects of brain lesions on spawning behaviour in the male goldfish. Physiol. Behav. 28:1103–1109. Kyle, A.L., N.E. Stacey, and R.E. Peter (1982) Ventral telencephalic lesions: Effects on bisexual behavior, activity, and olfaction in the male goldfish. Behav. Neural Biol. 36:229–241. Lapper, S.R., and J.P. Bolam (1991) The anterograde and retrograde transport of Neurobiotin in the central nervous system of the rat: Comparison with biocytin. J. Neurosci. Methods 39:163–174. Levine, R.L., and S. Dethier (1985) The connections between the olfactory bulb and the brain in the goldfish. J. Comp. Neurol. 237:427–444. Maler, L., E. Sas, S. Johnston, and W. Ellis (1991) An atlas of the brain of the electric fish Apteronotus leptorhynchus. J. Chem. Neuroanat. 4:1–38. Matz, S.P. (1995) Connections of the olfactory bulb in the Chinook salmon (Oncorhynchus tshawytscha). Brain Behav. Evol. 46:108–120. Metzner, W. (1993) The jamming avoidance response in Eigenmannia is controlled by two separate motor pathways. J. Neurosci. 13:1862–1878. Murakami, T., Y. Morita, and H. Ito (1983) Extrinsic and intrinsic fiber connections of the telencephalon in a teleost, Sebastiscus marmoratus. J. Comp. Neurol. 2166:115–131. Neary, T.J. (1990) The pallium of anuran amphibians. In E.G. Jones and A. Peters (eds): Cerebral Cortex. Vol 8A. Comparative Structure and Evolution of Cerebral Cortex, part 1. New York and London: Plenum Press, pp. 107–138. Neary, T.J. (1995) Afferent projections to the hypothalamus in ranid frogs. Brain Behav. Evol. 46:1–13. Neary, T.J., and R.G. Northcutt (1990) Septal area connections in ranid frogs. Soc. Neurosci. Abstr. 16:129. Nieuwenhuys, R. (1963) The comparative anatomy of the actinopterygian forebrain. J. Hirnforsch. 6:171–192. Nieuwenhuys, R., and J. Meek (1990) The telencephalon of actinopterygian fishes. In E.G. Jones and A. Peters (eds): Cerebral Cortex, Vol 8A. New York: Plenum, pp. 31–73. Northcutt, R.G. (1995) The forebrain of gnathostomes: In search of a morphotype. Brain Behav. Evol. 46:275–318. Northcutt, R.G., and M.R. Braford, Jr. (1980) New observations on the organization and evolution of the telencephalon of actinopterygian fishes. In S.O.E. Ebbeson (ed): Comparative Neurology of the Telencephalon. New York: Plenum, pp. 41–98. Northcutt, R.G., and R.E. Davis (1983) Telencephalic organization in ray-finned fishes. In R.E. Davis and R.G. Northcutt (eds): Fish Neurobiology. Ann Arbor: University of Michigan Press, pp. 203–236. Northcutt, R.G., and M. Ronan (1992) Afferent and efferent connections of the bullfrog medial pallium. Brain Behav. Evol. 40:1–16. Peter, R.E., and J.N. Fryer (1983) Endocrine functions of the hypothalamus of actinopterygians. In R.E. Davis and R.G. Northcutt (eds): Fish Neurobiology. Ann Arbor: University of Michigan Press, pp. 165–201. Peter, R.E., and V.E. Gill (1975) A stereotaxic atlas and technique for C.J.H. WONG forebrain nuclei of the goldfish, Carassius auratus. J. Comp. Neurol. 159:69–102. Reiner, A., and R.G. Northcutt (1992) An immunohistochemical study of the telencephalon of the Senegal bichir (Polypterus senegalus). J. Comp. Neurol. 319:359–386. Richards, S., and L. Maler (1996) The distribution of met-enkephalin in the brain of Apteronotus leptorhynchus, with emphasis on the electrosensory system. J. Chem. Neuroanat. 11:173–190. Sas, E., L. Maler, and M. Weld (1993) Connections of the olfactory bulb in the gymnotiform fish, Apteronotus leptorhynchus. J. Comp. Neurol. 335:486–507. Satou, M., Y. Oka, M. Kusunoki, T. Matsushima, M. Kato, I. Fujita, and K. Ueda (1984) Telencephalic and preoptic areas integrate sexual behavior in Himé salmon (landlocked red salmon, Oncorhynchus nerka): Results of electrical brain stimulation experiments. Physiol. Behav. 33:441– 447. Schnitzlein, H.N. (1962) The habenula and dorsal thalamus of some teleosts. J. Comp. Neurol. 118:225–268. Shiga, T., Y. Oka, M. Satou, N. Okumoto, and K. Ueda (1985a) An HRP study of afferent connections of the supracommissural ventral telencephalon and the medial preoptic area in Himé salmon (landlocked red salmon, Oncorhynchus nerka). Brain Res. 361:162–177. Shiga, T., Y. Oka, M. Satou, N. Okumoto, and K. Ueda (1985b) Efferents from the supracommissural ventral telencephalon in the Himé salmon (landlocked red salmon, Oncorhynchus nerka): An anterograde degeneration study. Brain Res. Bull. 14:55–61. Sloan, H.E. (1989) Neuroanatomical substrates for reproductive behavior in goldfish. Ph.D. Thesis. University of Kentucky, Lexington. Spiro, J.E. (1994) Modulation of the electric fish pacemaker rhythm at the level of the relay cell. Soc. Neurosci. Abstr. 20:379. Striedter, G.F. (1991) Auditory, electrosensory, and mechanosensory pathways through the forebrain of channel catfishes. J. Comp. Neurol. 312:311–331. Striedter, G.F. (1992) Phylogenetic changes in the connections of the lateral preglomerular nucleus in ostariophysan teleosts: A pluralistic view of brain evolution. Brain Behav. Evol. 39:329–357. Stroh, T., and G.K.H. Zupanc (1995) Somatostatin in the prepacemaker nucleus of weakly electric fish, Apteronotus leptorhynchus: Evidence for a nonsynaptic function. Brain Res. 674:1–14. Swanson, L.W., and W.M. Cowan (1979) The connections of the septal region in the rat. J. Comp. Neurol. 186:621–656. Székely, A.D., M.I. Boxer, M.G. Stewart, and A. Csillag (1994) Connectivity of the lobus parolfactorius of the domestic chicken. J. Comp. Neurol. 348:374–393. Villani, L., I. Zironi, and T. Guarnieri (1996) Telencephalo-habenulointerpeduncular connections in the goldfish: A DiI study. Brain Behav. Evol. 48:205–212. Von Bartheld, C.S., and D.L. Meyer (1986) Tracing of single fibers of the nervus terminalis in the goldfish brain. Cell Tissue Res. 245:143–158. Weld, M.M., and L. Maler (1992) Substance P-like immunoreactivity in the brain of the gymnotiform fish Apteronotus leptorhynchus: Presence of sex differences. J. Chem. Neuroanat. 5:107–129. Weld, M.M., S. Kar, L. Maler, and R. Quirion (1994) The distribution of tachykinin binding sites in the brain of an electric fish (Apteronotus leptorhynchus). J. Chem. Neuroanat. 7:123–139. Wong, C.J.H. (1997a) Afferent and efferent connections of the diencephalic prepacemaker nucleus in the weakly electric fish, Eigenmannia virescens: Interactions between the electromotor system and the neuroendocrine axis. J. Comp. Neurol. 383:18–41. Wong, C.J.H. (1997b) Neural Circuits Controlling Electrical Communication in Gymnotiform Fish. PhD. Thesis, University of California, San Diego. Zupanc, G.K.H., and W. Heiligenberg (1992) The structure of the diencephalic prepacemaker nucleus revisited: Light microscopic and ultrastructural studies. J. Comp. Neurol. 323:558–569. Zupanc, G.K.H., and L. Maler (1997) Neuronal control of behavioral plasticity: The prepacemaker nucleus of weakly electric gymnotiform fish. J. Comp. Physiol. A 180:99–111. Zupanc, G.K.H., and M.M. Zupanc (1992) Birth and migration of neurons in the central posterior/prepacemaker nucleus during adulthood in weakly electric knifefish (Eigenmannia sp.). Proc. Natl. Acad. Sci. USA 89:9539– 9543.