Afferent and efferent connections of the parapineal organ in lampreys A tract tracing and immunocytochemical studyкод для вставкиСкачать
THE JOURNAL OF COMPARATIVE NEUROLOGY 403:171–189 (1999) Afferent and Efferent Connections of the Parapineal Organ in Lampreys: A Tract Tracing and Immunocytochemical Study JULIÁN YÁÑEZ,1 MANUEL ANGEL POMBAL,2 AND RAMÓN ANADÓN3* 1Department of Cell and Molecular Biology, Faculty of Sciences, University of La Coruña, 15007-La Coruña, Spain 2Department of Functional Biology and Health Sciences, Faculty of Sciences, University of Vigo, 36200-Vigo, Spain 3Department of Fundamental Biology, Faculty of Biology, University of Santiago de Compostela, 15706-Santiago de Compostela, Spain ABSTRACT The neural connections of the parapineal organ of two species of lampreys were studied with the fluorescent dye 1,18-dioctadecyl-3,3,38,38-tetramethylindocarbocyanine perchlorate (DiI) and with immunocytochemistry. The lamprey parapineal organ consists of a vesicle and a ganglion that are connected to the left habenula. Labeling experiments included the application of DiI to the parapineal organ, left and right fasciculus retroflexus, left habenula, and the left pretectal region. Afferent parapineal fibers run in the left fasciculus retroflexus to the interpeduncular nucleus. The parapineal fibers of this fascicle arose from parapineal ganglion cells, whereas DiI application to the left habenula labeled both neurons of this ganglion and bipolar cells in the parapineal vesicle. Efferent neurons were observed in the left habenula, and bilaterally in the subhippocampal nucleus and the dorsal pretectum. Labeling with DiI also revealed a hippocampal projection. Immunocytochemical study of the parapineal vesicle revealed serotonin-immunoreactive cells in both species of lamprey, as well as substance P–immunoreactive (SP-ir) cells in sea lamprey and choline acetyltransferase–immunoreactive (ChAT-ir) cells in the river lamprey. The SP-ir cells and ChAT-ir cells formed a rich neuropil in the parapineal ganglion. Calretinin-ir cells were numerous in the ganglion. Neuropeptide Y–immunoreactive and ␥-aminobutyric acid–immunoreactive efferent fibers were observed in the parapineal organ. Neuropeptide Y–immunoreactive fibers originate in the subhippocampal nucleus, whereas ␥-aminobutyric acid–immunoreactive fibers might also arise in the pretectal nucleus. A few galanin-ir fibers were observed. These results indicate that the parapineal connections are completely different from those of the pineal organ. The possible homology between parapineal organs of vertebrates is discussed. J. Comp. Neurol. 403:171–189, 1999. r 1999 Wiley-Liss, Inc. Indexing terms: pineal complex; DiI; habenula; entopeduncular nucleus; GABA; Agnathans The pineal complex of lampreys consists of two photoreceptive organs, the pineal and parapineal organs (Meiniel, 1969; Meiniel and Collin, 1971; Cole and Youson, 1982). The pineal organ consists of a vesicle joined to the posterior commissure by a long pineal stalk, whereas the parapineal organ consists of a parapineal vesicle and a ganglionic mass (parapineal ganglion) associated with the tela choroidea, which extends between the parapineal and the habenula and through which runs a diffuse parapineal tract. It is notable that no similar ganglion has been found in other vertebrates. The nomenclature given to the ganglion and the tract entering the left habenula reflects two different interpretations. According to Studnicka (1905), r 1999 WILEY-LISS, INC. the parapineal ganglion should be considered as a part of the left habenula and, therefore, should be called the accessory habenular nucleus, and the tract, the habenular Grant sponsor: Spanish Education Ministry; Grant number: PB96–0945C03; Grant sponsor: Xunta de Galicia; Grant number: XUGA20012B96; Grant number: XUGA20002B97; Grant sponsor: University of Vigo; Grant number: 64102C734. *Correspondence to: Dr. R. Anadón, Department of Fundamental Biology, Faculty of Biology, University of Santiago de Compostela, Santiago de Compostela 15706, Spain. E-mail: email@example.com Received 27 January 1998; Revised 5 August 1998; Accepted 17 August 1998 172 J. YÁÑEZ ET AL. tract. Meiniel and Collin (1971) support this interpretation on the basis of the similar ultrastructural appearance of the cells of the ganglion and of the left habenula. According to Tretjakoff (1915), however, the parapineal ganglion and parapineal tract are different structures from the habenula. This question could be clarified by studying the connections of the parapineal organ. An experimental tracing study of the lamprey pineal complex has indicated that it has projections to the pretectum and mesencephalic tegmentum, without any distinction of pineal and parapineal components (Puzdrowski and Northcutt, 1989). Another study in larval lamprey (Yáñez et al., 1993) reported similar projections for the pineal organ. With regard to its function, the pineal complex, as a whole, has been considered as a sensory organ that conveys photic information to neural structures (Morita and Dodt, 1973, Pu and Dowling, 1981, Tamotsu and Morita, 1986; Samejima et al., 1989) and releases hormonal messages (Joss, 1977; Bolliet et al., 1993) that may be involved in sexual maturation (Joss, 1973a), changes in body coloration (Joss, 1973b, 1977), circadian rhythms (Morita et al., 1992), and metamorphosis (Cole and Youson, 1981). Immunocytochemical studies of the lamprey parapineal organ have reported the expression of molecules involved in light signal transduction (Vigh-Teichmann et al., 1983; Kuo et al., 1988; Tamotsu et al., 1990, 1994; Garcı́a-Fernández et al., 1997) and the presence of serotonin (a possible precursor of melatonin: Meiniel, 1978, 1980; Tamotsu et al., 1990), somatostatin (Yáñez et al., 1992), and glutamate (Debreceni et al., 1997), but no particular function has been assigned to this structure that is generally considered as a rudimentary or regressed sense organ (Cole and Youson, 1982). The aim of the present investigation with carbocyanine dye tracing was to determine the relation between the parapineal ganglion and vesicle with the habenula and other parts of the brain. In addition, the immunocytochemical study of the neuropeptides, neurotransmitter, and neurotransmitter synthesizing enzymes aimed to investigate the chemical nature and possible origins of some fiber systems of the parapineal organ. Because lampreys are living representatives of the earliest vertebrates, the Agnathans, better knowledge of the parapineal organ is Abbreviations FR H HY IPN LH MO OB OC ON OT P PH PP PC PT RH SH ST TC TH TL TO fasciculus retroflexus habenula hypothalamus interpeduncular nucleus left habenula medulla oblongata olfactory bulb optic chiasma optic nerve optic tract pineal organ primordium hippocampi parapineal organ posterior commissure parapinealopetal pretectal nucleus right habenula subhippocampal-thalamic nucleus striatum tela choroidea thalamus lateral telencephalic lobe optic tectum likely to contribute to understanding early evolution of the pineal complex. MATERIAL AND METHODS Animals Adult river lamprey (Lampetra fluviatilis) were caught in the spring in Söderham (Sweden) during their upstream migratory phase. They were maintained at 4–10°C in well-aerated freshwater aquaria before experiments. Upstream migrating adults, young adults, and ammocoetes of sea lamprey (Petromyzon marinus) were caught in the Rivers Miño and Ulla (northwest Spain) and maintained at room temperature. Before fixation, all lampreys were deeply anesthetized with tricaine methane sulfonate (Sigma, St. Louis, MO) and killed by decapitation. All the experiments conformed to the European Community’s guidelines on animal care and experimentation. The brains were then quickly exposed to fixative by carefully removing the skull roof to preserve the pineal complex. DiI labeling Twelve adult lampreys (L. fluviatilis, L.) and 6 young postmetamorphic sea lampreys (Petromyzon marinus, L.) were used for DiI labeling. The brain was fixed in cold 4% paraformaldehyde in 0.1 M phosphate buffer (PB), pH 7.2–7.4. Crystals of 1,18-dioctadecyl-3,3,38,38-tetramethylindocarbocyanine perchlorate (DiI; Molecular Probes, Eugene, OR) were applied either to the parapineal organ (three river lampreys, one sea lamprey), the left habenula (three river lampreys, one sea lamprey), the pretectal region (two river lampreys), the right fasciculus retroflexus (one river lamprey, three sea lampreys), or the left fasciculus retroflexus (three river lampreys, one sea lamprey) following the method of Yáñez et al. (1996). In brief, after fixation, the brain was embedded in 3% agarose, and the block was cut on a Vibratome to the desired level. A small crystal of DiI was then applied with the tip of an electrolytically sharpened insect pin or a micropipette. Then, the entire surface was sealed with melted agarose to avoid spillage of crystals, and the brain was left in fresh fixative for 7–20 days at 37°C in darkness. Vibratome sections (50- to 100-µm thick) of the brain were examined and photographed with a fluorescence microscope equipped with a rhodamine filter set. To reveal the parapineal connections, the parapineal organ was labeled following the previously described procedure. In an additional experiment, the tracer was applied to the parapineal vesicle without embedding the brain in agarose or cutting any part of the brain. To confirm the results of these experiments, DiI was applied to those projecting areas visualized: the caudal left habenula, the left and right fasciculus retroflexus at the level of the infundibulum, and the left pretectum at a level caudal to the posterior commissure. The nomenclature used in this study to name the brain areas follows that of Heier (1948) and Schöber (1964). Immunocytochemistry For immunocytochemistry, the brains were rapidly dissected out and fixed by immersion in the fixative appropriate for each antibody used (see also Pombal et al., 1997). A solution of 3% glutaraldehyde and 0.5% paraformaldehyde in PB or 5% glutaraldehyde in PB were used for ␥-aminobu- PARAPINEAL ORGAN CONNECTIONS IN LAMPREY tyric acid (GABA; eight river lampreys, two sea lampreys); 4% paraformaldehyde in PB for choline acetyltransferase (ChAT; four river lampreys); 4% paraformaldehyde and 0.2 picric acid in PB for serotonin (5-HT; five river lampreys, two sea lampreys), and substance P (SP; two sea lampreys, two river lampreys) and Bouin’s fluid for calretinin (CR; three river lampreys, two sea lampreys), neuropeptide Y (NPY; three river lampreys), and galanin (six river lampreys). Those brains fixed in Bouin’s fluid were subsequently dehydrated, embedded in paraffin, and cut (15-µmthick) either in the transverse, horizontal, or sagittal plane. The other brains were cryoprotected with 30% sucrose, frozen with liquid nitrogen, sectioned (15-µmthick) in the same planes on a cryostat, and processed according to the immunoperoxidase technique. Briefly, sections were incubated with one of the primary antisera raised in rabbits against GABA (Chemicon, Temecula, CA; Affinity, Mamhead, U.K.; or Sigma; Pombal et al., 1997; dilution, 1:1,000 to 1:2,000), 5-HT (Zhang et al., 1996; dilution 1:1,000), SP (Pombal et al., 1997; dilution 1:2,000), CR (7696, Swant, Bellinzona, Switzerland; Schwaller et al., 1993; dilution, 1:1,000), NPY (Sigma; dilution, 1:600), or galanin (Jiménez et al., 1996; dilution 1:1,500). They were subsequently incubated with goat anti-rabbit (Dakopatts, Glostrup, Denmark; dilution, 1:60) and rabbit peroxidase-antiperoxidase complex (Dakopatts; dilution, 1:300). The antisera were diluted in 0.1 M phosphatebuffered saline, pH 7.2–7.4, containing 0.3% Triton X-100. The peroxidase-antiperoxidase complex was developed by incubation in 3,38-diaminobenzidine (Sigma; 50 mg/100 ml) and 0.005% hydrogen peroxide in 0.05 M Tris buffer, pH 7.3. The same buffer was used to stop the reaction. For ChAT immunocytochemistry, the procedures used are the same as in Marı́n et al. (1997), by using a goat anti-ChAT serum (Chemicon; dilution, 1:100). RESULTS The distribution of afferents and efferents of the parapineal organ of lampreys revealed after DiI application to this structure or to some of its targets will be described first. A detailed description of the structures of the parapineal system immunoreactive for some neurotransmitters, neurotransmitter synthesizing enzymes, neuropeptides, and a calcium binding protein, are described next. Tract tracing methods The application of the carboindocyanine dye DiI to fixed tissue allows the tracing of connections of minute structures without affecting neighboring regions. As shown in Figure 1, the procedure of DiI application to the parapineal organ of lamprey specifically labeled this organ, and did not affect the pineal organ or any other brain structure. In two of the experiments the intention was to label only the parapineal vesicle, but this also led to labeling of the ganglion. Thus, the connections revealed by these experiments refer to the entire parapineal organ. Afferent connections Application of DiI to the parapineal organ. DiI application to the parapineal organ labeled the parapineal tract in the tela choroidea, the left (small) habenula (both cells and fibers), the left fasciculus retroflexus, and the efferent fibers running through the right habenula and the thalamus. The cells and neuropil of the right (large) habenula and the right fasciculus retroflexus were unlabeled. Both 173 cells and fibers were labeled in the left habenula, suggesting that the parapineal organ and left habenula are bidirectionally connected (see below). In addition, two efferent neuronal groups (located in the caudal telencephalon and the pretectum, respectively) were labeled bilaterally, as was a conspicuous tract that runs to the primordium hippocampi. The application of DiI to the parapineal organ did not produce any labeling in the pineal organ. The targets of the parapineal and the efferent nuclei, with the exception of the pretectal nucleus, thus, coincide with those reported for the habenula in larval sea lamprey (Yáñez and Anadón, 1994). Note that the connection between the parapineal and the habenula was not observed in that study. The parapineal tract consisted of numerous small nerve fascicles running parallel to each other in the tela choroidea (Fig. 1B). The left habenula was strongly fluorescent because of the presence of numerous labeled fibers, but a few labeled neurons were also observed (Fig. 1C,D). The labeling of this habenula contrasted with the lack of labeling of cells and fibers in the right habenula (excluding the efferent fibers that cross it caudally) (Fig. 1C). The left fasciculus retroflexus was intensely labeled. The fibers of the tract run compactly, and very few labeled fibers diverge from the fascicle, with the exception of a few fibers running in the dorsal thalamus-pretectum, caudal to the habenula. In the interpeduncular nucleus the fibers of this fascicle formed a conspicuous unpaired terminal field, probably branching and crossing the midline several times (Fig. 2A-E). In the midline, some fibers exhibit large irregular expansions (Fig. 2D). The parapinealofugal fibers extended only to the rostral part of the interpeduncular nucleus, and no fibers were seen to extend as far as the rhombencephalon proper. Application of DiI to the left habenula or the left fasciculus retroflexus. To determine whether the afferent fibers labeled after DiI application to the parapineal originate from the parapineal vesicle, the parapineal ganglion, or both, three types of labeling were carried out: one in the left habenula and the other two in the left or right fasciculus retroflexus at the level of the posterior tubercle. Application of DiI to the right fasciculus retroflexus produced no labeling of parapineal structures. Application of DiI to the left habenula or the left fasciculus retroflexus led to intense labeling of the parapineal ganglion, indicating that cells of this ganglion give rise to the afferent fibers running in the left fasciculus retroflexus (Fig. 3A-G). The application of DiI to the left habenula also led to intense labeling of numerous cells in the parapineal vesicle (Fig. 3A-D), whereas labeled cells were not observed in the vesicle after application to the left fasciculus retroflexus. The labeled cells in the parapineal vesicle were spindleshaped, with a dendrite extending to the parapineal lumen and ending in a small head, sometimes bearing an irregular apical process, and a basal process. Careful observation of the basal process suggests that it may branch at the transition to the parapineal ganglion, indicating that the labeled bipolar (long axon) cells might project both to the parapineal ganglion and left habenula. No fiber or cell of the pineal organ became labeled in these three sets of experiments, in agreement with the previously reported findings for the lamprey pineal organ (Yáñez et al., 1993). Efferent connections. Application of DiI to the parapineal organ led to labeling of fibers that run through the thalamus lateral to both habenulae and through the right 174 Fig. 1. DiI labeling of the parapineal organ of river lamprey. A: Transverse section showing the labeled parapineal vesicle (black asterisk) and parapineal ganglion (black star) just caudal to the application point. Note that both the pineal organ and dorsal telencephalon (white star) are unlabeled. B: Flat view of the tela choroidea between the parapineal ganglion and the habenula showing the fascicles of the parapineal tract. The position of the unlabeled pineal stalk, which runs in a groove of the tela, is also distinguishable J. YÁÑEZ ET AL. (arrow). C: Strong labeling of the left habenula (wide arrow) and of efferent fibers running through the thalamus and crossing the right habenula (long arrows). The small arrows indicate labeled neurons; dashed lines indicate the borders of the brain. D: Detail of the left habenula showing strong labeling of neuropil and a few habenular neurons (arrows). For abbreviations, see list. Scale bars ⫽ 50 µm in A,B,D, 100 µm in C. PARAPINEAL ORGAN CONNECTIONS IN LAMPREY 175 Fig. 2. Transverse sections of the interpeduncular nucleus of river lamprey showing labeled parapineal afferents. A: Transition from the left fasciculus retroflexus (arrow) to the rostral part of the interpeduncular nucleus. B: Rich plexus of parapinealofugal fibers in the rostral interpeduncular nucleus. C: Detail of the branching of parapineal fibers (arrow). D: Lateral expansions of parapineal fibers in the midline (arrow). E: Transition to the caudal part of the interpeduncular nucleus showing the end of parapineal fibers. Note the ventral unlabeled portion of the nucleus (asterisk). Star, raphe. Scale bars ⫽ 100 µm A,B,E; 50 µm in C,D. habenula in the habenular commissure (Fig. 1C). In the rostral thalamus, they formed a tract that apparently bifurcated between the primordium hippocampi and the subhippocampal nucleus, the rostrodorsal branch extending, as a band-shaped tract, through the inner part of the primordium hippocampi and ending freely, whereas the ventral branch was followed to a sparse population of small labeled neurons of the lobus subhippocampalisrostral thalamus which appears to give rise to these fibers (Fig. 4A–D). These cells have long thin dendrites that run through the neuropil. A few labeled cells of this population may even reach the ventral limit of the primordium hippocampi. This population corresponds to the subhippocampal-thalamic habenulopetal nucleus of Yáñez and Anadón (1994). Application of DiI to the parapineal organ also led to labeling of a nucleus of small neurons in the middle of the pretectal region, which extends rostrocaudally near the dorsal limit of the optic tract (Fig. 5A,B). The labeled cells of this nucleus exhibit long thin dendrites extending dorsolaterally to the subpial surface and forming a conspicuous dendritic plexus. Application of DiI to the region of the pretectal nucleus produced labeling of fibers in the habenulae and the proximal parapineal tract (Fig. 5C,D). This suggests that the pretectal nucleus is also habenulopetal. Unfortunately, in the experiments of this type, the tela choroidea was broken and fibers from the pretectal region could not be followed to the parapineal organ. A schematic representation of the afferent and efferent connections of the lamprey parapineal is shown in Figure 6. Immunocytochemistry of the parapineal organ To characterize the chemical nature of parapineal circuits, we carried out an immunocytochemical study with antisera to several neurotransmitters, neurotransmittersynthesizing enzymes, and neuropeptides, as well as to the 176 J. YÁÑEZ ET AL. Figure 3 PARAPINEAL ORGAN CONNECTIONS IN LAMPREY calcium-binding protein calretinin. Most of the results presented here were for the river lamprey, but some results were obtained by using sea lamprey. In the parapineal organ, we found immunoreactivity with antisera to the following substances: 5-HT, SP, ChAT, NPY, GABA, galanin, and CR. Cells. In the sea lamprey, but not in the river lamprey, the walls of the parapineal vesicle contained SP-immunoreactive (SP-ir) bipolar cells. The SP-ir cells, in general, stained faintly, and were more numerous in the dorsal, rostral, and lateral regions of the vesicle (Fig. 7A,B). These pale cells gave rise to more intensely stained fibers running from the outer region of the wall to the medial neuropil of the ganglion where they form a very rich SP-ir plexus. Only a few SP-ir fibers were observed in other neuropil areas of the ganglion and in the tela choroidea between the parapineal and the habenula. In the river lamprey, a few SP-ir fibers (probably efferent fibers) were observed in the ganglion and the tela choroidea, whereas the vesicle lacks any SP-immunoreactive structure. ChAT immunoreactivity has been observed in two groups of bipolar cells of the parapineal vesicle of river lamprey (Fig. 7C,D). One of the groups, the most numerous, occupies the dorsorostral wall of the vesicle, the other the ventrocaudal region. Fibers originated from these groups form numerous ChAT-immunoreactive (ChAT-ir) boutons among cells of the vesicle and the ganglion. No ChAT-ir fibers appear to run in the parapineal tract. 5-HT-immunoreactive (5-HT-ir) cells were mostly located in the dorsal, rostral, and lateral regions of the vesicle and were practically absent from the ventral wall (Fig. 8A,B). They were bipolar cells with a short process directed to the lumen and a basal process running near the pial surface. A few 5-HT-ir fibers enter the parapineal ganglion. No 5-HT-ir fiber was observed leading from the parapineal organ to the habenula. 5-HT immunoreactivity was observed in cells of the parapineal vesicle of both species of lamprey, but they were very scarce in the sea lamprey. CR-immunoreactive (CR-ir) neurons were observed in the parapineal ganglion of river lamprey but not in the vesicle (Fig. 8C). The CR-ir cells occupied rostral and lateral regions of the ganglion, where they may represent most of the cells. However, in caudal regions, there are many immunonegative cells. The positivity to CR was Fig. 3. Transverse sections showing the labeling of the parapineal organ after application of DiI to the left habenula (A–D) and the left fasciculus retroflexus (E–G). A: Labeling of bipolar cells ventral in the parapineal vesicle and strong labeling of the ganglion (black star) of a young sea lamprey. B: Differential interference contrast micrograph of the same section showing the pineal and the parapineal (asterisk) vesicle and ganglion (black star). C: Labeled bipolar neurons (arrow) in the parapineal vesicle of a river lamprey. Note also the strong labeling of the ganglion (star). D: Detail of a bipolar cell in the parapineal vesicle of sea lamprey showing the apical process (arrow) and possible branching (arrowhead) of the basal process. Star, parapineal ganglion. E: Transverse section through the posterior tubercle showing the point of DiI application to the left fasciculus retroflexus (black arrow). Star, optic tract; thin white arrow, posthabenular region; wide white arrow, posterior hypothalamic recess. F–G: Two different focal planes of a section of the parapineal ganglion (stars) in the same experiment as in E, showing the dendritic plexus, with some dendrites even extending toward the vesicle (asterisk), and labeled perikarya (small arrows in G). P, pineal organ. Scale bars ⫽ 50 µm in A-D,F,G; 250 µm in E. 177 weak but clear in perikarya, and rather diffuse in the neuropil and the parapineal tract. The left habenula exhibits large areas of CR-ir cells with a similar appearance (Fig. 8D), whereas immunoreactivity to CR in the right habenula is observed as patches of darker-stained cells. Efferent fibers. GABA-ir efferent fibers were observed in the parapineal organ of both species of lamprey, and NPY-immunoreactive (NPY-ir) efferent fibers were also observed in the river lamprey. These fibers enter the parapineal ganglion, as part of the parapineal tract and form a rather rich plexus of beaded fibers among its cells and in the neuropil (Figs. 9A-C, 10A,B). GABA-ir fibers, which appear to be the most numerous, also enter the walls of the vesicle in small number (Fig. 9A,B). Although double labeling was not carried out, comparison of the distribution of immunoreactive neurons in series of lamprey brains immunocytochemically stained for GABA or NPY with the retrogradely labeled nuclei after DiI application to the parapineal organ, allowed determination of the probable origin of either GABA- or NPY-ir efferent parapineal innervation. Many cells of the parapinealopetal subhippocampal-thalamic nucleus exhibited NPY immunoreactivity (Fig. 10C), indicating that this nucleus is the most probable origin of NPY-ir efferent fibers of the parapineal ganglion. Small GABA-ir neurons were observed both in the subhippocampal-thalamic nucleus and the region of the pretectal parapinealopetal nucleus (not shown), suggesting that efferent GABA-ir fibers of the parapineal organ may have a dual origin. Very few (one to four) galanin-ir fibers were observed to run through the parapineal tract and the ganglion of river lamprey, but not the vesicle (Fig. 10D,E). The possible origin of these galanin-ir fibers was not determined because none of the galanin-ir brain nuclei reported in a previous study of the river lamprey (Jiménez et al., 1996) corresponds with the efferent nuclei demonstrated by DiI application to the parapineal organ. DISCUSSION Technical comments DiI labeling is a very sensitive method for tracing neuronal projections (Holmqvist et al., 1992). When using this technique in unmyelinated systems, clear and highly reproducible results can be obtained with a small number of animals (Yáñez and Anadón, 1996; Yáñez et al., 1993, 1996, 1997). Because this tracer diffuses along membranes of cell processes both anterogradely and retrogradely from the application site and bifurcation points, additional experiments may be necessary to confirm the afferent or efferent character of some labeled fiber systems. Moreover, because of the small size and closeness of the two parts of the parapineal organ (vesicle and ganglion), DiI application to these structures could not clearly differentiate their respective connections. Therefore, additional labeling of some parapineal targets, i.e., left habenula and left fasciculus retroflexus, was required to distinguish their respective connections. The results of GABA and NPY immunocytochemistry also served as a control of the results with DiI tracing on efferent systems, and excluded the possibility that these systems were labeled by transneuronal diffusion of DiI. From our previous experience with the DiI method (Yáñez et al., 1993, 1996, 1997; Yáñez and Anadón, 1994, 1996), it does not appear that transneuronal diffu- 178 Fig. 4. Transverse sections through the telencephalon showing neurons and fibers labeled after DiI application to the parapineal organ. A: Labeled neurons in the subhippocampal nucleus (SH). The arrow indicates the subhippocampal sulcus. B: Detail of labeled cells of this nucleus and their processes. C: Section at the region where parapinealopetal fibers bifurcate, giving rise to a collateral hippocam- J. YÁÑEZ ET AL. pal bundle (wide arrow). The thin arrow indicates the subhippocampal sulcus. D: Bundle of collaterals of parapineal efferent fibers running rostralward in the primordium hippocampi (PH). Asterisk, telencephalic ventricle; star, subventricular layer of neurons. In all figures, medial is to the left. Scale bars ⫽ 100 µm in A,C,D, 50 µm in B. PARAPINEAL ORGAN CONNECTIONS IN LAMPREY 179 Fig. 5. Transverse sections showing parapinealopetal neurons of the pretectum (PT) labeled after DiI application to the parapineal organ (A,B), and habenulopetal and parapinealopetal fibers after application of DiI to the pretectum (C,D). A,B: Rostral and caudal sections through the pretectum showing the parapinealopetal cells (thin arrows) and their dendritic plexuses. In A, note the labeled left fasciculus retroflexus (wide arrow). Asterisk, mesencephalic ventricle. C: Fibers labeled in the left (asterisk) and right habenulae (RH). Thick arrow, fiber passing to the parapineal tract; thin arrow, bundle of labeled fibers entering the right habenula. D: Labeled fibers (arrow) in the parapineal tract. The tela choroidea (asterisk) was obliquely sectioned; black star, dorsal telencephalon. Scale bars ⫽ 100 µm in A,B, 50 µm in C,D. sion was a relevant problem, providing that the fixation was good. those reported for the entire pineal complex in silver lamprey (Puzdrowski and Northcutt, 1989). Although these authors assumed that they labeled both the pineal and parapineal organs, in the light of the present results and those of Yáñez et al. (1993) on the projections of the lamprey pineal organ, it is evident that the projections reported by them correspond only to those of the pineal organ. The parapineal organ is connected to the habenulo- Connections of the lamprey parapineal organ and the question of homology The pineal and parapineal organs appear to have very different connections in lamprey. The neural projections of the parapineal organ reported here differ completely from Fig. 6. Schematic drawings of a lateral and a dorsal view of the brain of the lamprey (A,B) and four transverse sections through the brain (C–F) to show the location of the pineal complex and the distribution of parapineal-related structures (the pineal complex is not represented in B). The shaded areas in A and B represent the terminal fields of parapinealofugal fibers, other fibers were only represented in C–F. Single arrows indicate the pineal tract, double arrows represent the parapineal tract, and dots represent the location of parapinealopetal neurons. The vertical bars (C–F) in A indicate the levels of the sections. In A,B, rostral is at the left; in C–F, the left side of the brain is at the right. For abbreviations, see list. PARAPINEAL ORGAN CONNECTIONS IN LAMPREY 181 Fig. 7. Parapineal organs of sea lamprey (A,B) and river lamprey (C,D) immunostained for substance P (SP) and choline acetyltransferase (ChAT), respectively. A,B: Transverse sections through the rostral region of the vesicle and ganglion (A), and through the junction between the vesicle and ganglion (B). Note the pale SP-immunoreactive (SP-ir) cells (arrowheads in A and B), the abundance of SP-ir fibers in the medial neuropil of the ganglion (arrows in B) and their scarcity in the rostral neuropil (A). C: Transverse section through the vesicle showing the rostral group of ChAT-ir cells in the dorsal wall. D: Sagittal section through the parapineal organ showing the rostrodorsal (outlined arrow) and caudoventral (thin arrow) groups of ChAT-ir cells (rostral is at the left). Note the numerous boutons in both the vesicle and ganglion. Asterisks, parapineal vesicle; stars, parapineal ganglion. Scale bars ⫽ 50 µm in A–D. interpeduncular system, whereas in lampreys the pineal organ has no habenular connection and instead has projections to other diencephalic and mesencephalic regions. The origin of the efferents to the pineal and parapineal organs (a few tegmental neurons for the pineal, and the subhippo- campal-thalamic nucleus and the pretectal nucleus for the parapineal) is also completely separate. Therefore, these two organs do not seem to be directly related in lampreys and are probably involved in modulation of different systems by light. 182 J. YÁÑEZ ET AL. Fig. 8. Immunoreactivity to serotonin (5-HT) (A,B) and calretinin (C,D) in the parapineal organ (A–C) and left habenula (D) of river lamprey. A,B: Transverse sections through the rostral (A) and caudal region (B) of the vesicle showing 5-HT-ir cells and fibers (arrowheads). Asterisks, parapineal vesicle; star, parapineal ganglion. C: Sagittal section through the pineal complex showing CR-ir cells in the parapi- neal ganglion (star). Asterisk, parapineal vesicle. D: Sagittal section through the left habenula showing groups of CR-ir cells in the rostral part (open star) similar in appearance to those of the parapineal ganglion. In C,D, rostral is at the right. Scale bars ⫽ 50 µm in A,B, 100 µm in C,D. The present results involving connectivity and CR immunocytochemistry of the parapineal organ also support the idea, based on ontogenetic (Studnicka, 1905) and electron microscopic (Meiniel and Collin, 1971) studies, that the ganglion is a part of the left habenular structure, which undergoes a rostralward migration during development. In this sense, the presence of parapineal afferent fibers running in the left fasciculus retroflexus to the interpeduncular neuropil and of a bilateral efferent projection from the habenulopetal subhippocampal-thalamic nucleus to the parapineal organ are clearly habenular characteristics (Yáñez and Anadón, 1994). Calretinin staining of the neurons of the left habenula and the parapineal ganglion is also similar. The nomenclature used by Studnicka (accessory left habenular nucleus for the ganglion, and habenular tract or habenular lamina for the laminar tract that connects it with the left habenula), is very descriptive. Despite this, the traditional terms adopted here (parapineal ganglion and tract) are less confusing for new readers. Results seen after application of DiI to the left fasciculus retroflexus or the left habenula indicate that the cells of the parapineal vesicle do not project beyond the left PARAPINEAL ORGAN CONNECTIONS IN LAMPREY 183 Fig. 9. ␥-Aminobutyric acid (GABA) immunoreactivity in the parapineal organ of river lamprey. A: Sagittal section showing the parapineal tract (solid arrow), ganglion (star), and vesicle (asterisk). Numerous GABA-immunoreactive (GABA-ir) fibers are observed in the ganglion, but some also pass to the parapineal vesicle. Note also GABA-ir fibers running through the parapineal tract. Open arrow, pineal stalk. B: Transverse section through the parapineal ganglion (star) and vesicle (asterisk) showing GABA-ir fibers that also enter the walls of the vesicle (arrowheads). C: Detail of GABA-ir terminals in the parapineal ganglion. P, pineal organ. Scale bars ⫽ 50 µm in A,B, 10 µm in C. habenula. The afferents of the parapineal vesicle of the lamprey are similar to those of the parapineal organ of trout (Yáñez et al., 1996) and the parietal (parapineal) organ of lizards (Engbretson et al., 1981), which project exclusively to the left habenula. The asymmetrical relationship of the parapineal nerve with the left habenula has been reported in other reptiles by using nonexperimental methods (Dendy, 1911; von Haffner, 1953; Ortman, 1960; Kappers, 1965). All these parapineal organs of lamprey, teleosts, and reptiles have also in common their develop- ment from a rostral parietal vesicle (Hill, 1891; Nowikoff, 1910; Dendy, 1911; Meiniel and Collin, 1971; Corujo and Anadón, 1986), whereas the pineal organs develop from the posterior parietal vesicle of the embryo. The shared parapineal-left habenula projection, together with the same embryonic origin, are hard to explain if the parapineal appeared independently in these vertebrate groups, which suggests that these parapineal organs have evolved from a common precursor. The vesicular character of the parapineal organ is also shared by lampreys, coelacanths, 184 J. YÁÑEZ ET AL. Fig. 10. NPY-ir fibers in the parapineal organ (A,B), NPY-ir cells in the subhippocampal nucleus (C), and galanin-ir fibers in the parapineal ganglion (D,E) of river lamprey. A: Sagittal section showing NPY-ir fibers in the parapineal ganglion (star). Asterisk, parapineal vesicle. Rostral is at the left. B: Transverse section showing NPY-ir fibers in the parapineal ganglion (star). Asterisk, parapineal vesicle. C: Horizontal section passing through the striatum and thalamus showing numerous NPY-ir perikarya in the subhippocampal nucleus (open arrows). Rostral is at the left. Arrow, external sulcus between the telencephalic lobe and the thalamus. D,E: Transverse sections through the parapineal ganglion (black stars) (D, rostral; E, caudal) showing a few galanin-ir fibers (arrowheads). Asterisk, parapineal vesicle; open star in E, pineal atrium. For abbreviations, see list. Scale bars ⫽ 50 µm. Sphenodon, and lizards. It is, however, puzzling why other extant vertebrate groups (elasmobranchs, amphibians, birds, and mammals) lack a parapineal organ and why it has not been found in many bony fishes, although it is well developed in the coelacanth Latimeria chalumnae (Hafeez and Merhige, 1977). The hypothesis that all parapineal organs have a common origin suggests that their absence in some extant vertebrate groups must be attributable to regressive changes that occurred multiple times independently. Although a regressive tendency of the parapineal in benefit of the pineal might explain the scarce development of this organ in the trout and other teleosts, or the appearance during development of a vestigial rostral parietal organ (pre-epiphysis) in some birds and mammals, which exceptionally can persist in adults (Beccari, 1943, pp. 363), whether early members of these groups possessed this organ is not known. Because of its embryonic origin from a caudal parietal vesicle (Beccari, 1943) and its connections of pineal type (Eldred et al., 1980), the frontal organ of the frog does not appear to be a parapineal PARAPINEAL ORGAN CONNECTIONS IN LAMPREY organ. More studies are need to understand the phylogeny of parapineal organs. Parapineal photoreceptors Immunocytochemistry carried out with antisera to opsins and other visual proteins indicates that the lamprey parapineal vesicle is heterogeneous with respect to its photoreceptor cells (Vigh-Teichmann et al., 1983; Kuo et al., 1988; Tamotsu et al., 1990, 1994; Garcı́a-Fernández et al., 1997), in agreement with results of earlier ultrastructural studies (Meiniel and Collin, 1971; Cole and Youson, 1982). Our results demonstrated that different photoreceptor populations are labeled with antisera to SP, ChAT, and 5-HT presenting important interspecific differences. In sea lamprey, most SP-ir cells occupy different regions of the vesicle to those of the cells labeled by DiI application to the left habenula. Results from experiments using immunocytochemistry and DiI tracing indicate that the axons of most of these cells only extend as far as the parapineal ganglion neuropil, only a few SP-ir fibers are observed in the parapineal tract. Because of their bipolar appearance, mainly dorsal and lateral location in the vesicle, and the abundance of their processes in the ganglion, SP-ir cells may correspond to the type II photoreceptors revealed by electron microscopy in this species (Cole and Youson, 1982), although this should be investigated further. In the sea lamprey, SP-ir perikarya were reported in a region of the corpus striatum, which may include, in part, our subhippocampal nucleus, and SP-ir fibers were observed in the ventral habenular nucleus (Nozaki and Gorbman, 1986). The possibility that their corpus striatum also contributes with a few SP-ir fibers to the parapineal ganglion in the sea lamprey, thus, cannot be ruled out. In the river lamprey, the antiserum to SP did not reveal any cells in the parapineal organ, although a few SP-ir fibers were seen to run through the parapineal tract into the ganglion. These fibers were probably efferent. Differences between two species of lamprey (Petromyzon marinus and Entosphenus tridentatus) in both SP immunoreaction staining and distribution of SP-ir perikarya in the brain was also previously reported (Nozaki and Gorbman, 1986). The presence of SP-ir structures has been described in other parapineal organs. In trout, Ekström and Korf (1986) claim that the parapineal organ receives a distinct SP-ir innervation by means of the habenular nucleus. Tracing studies of trout parapineal organs, however, found no parapinealopetal neurons either in the habenula or other brain regions whereas the parapinealofugal fibers apparently extend to the same area of the left habenula that is positive to SP (Yáñez et al., 1996). These contradictory results in trout can only be explained if the parapineal perikarya were SP-ir and were, thus, the origin of the SP-ir parapineal tract and neuropil in the dorsal left habenula. This explanation would be in agreement with the lack of a neuropil in the parapineal of trout as revealed by electron microscopy (Rüdeberg, 1969). SP-ir fibers were also found in both the parietal organ and left habenula of lizard (Engbretson et al., 1982). Despite the parietopetal cells not being able to be characterized, the SP-ir fibers in the parietal organ were considered to be efferent because of the lack of SP-ir perikarya in the parietal organ and left habenula. Thus, all the parapineal organs studied appear to be related with SP-ir systems. On the other hand, the habenula of mammals (Ronnekleiv and Kelly, 1984; Con- 185 testabile et al., 1987), newt (Taban and Cathieni, 1983), and Apteronotus, an electric fish (Weld and Maler, 1992) contains SP-ir perikarya. Our ChAT immunocytochemical results suggest the presence of an intrinsic cholinergic system in the parapineal organ of river lamprey. ChAT-ir cells form two apparently separated groups, one in the rostrodorsal region and the other in the ventrocaudal region. The bipolar shape of the cells is compatible with them being photoreceptors. These groups of ChAT-ir cells give rise to nerve terminals in the vesicle and parapineal ganglion, and thus appear to form part of local circuits. Although the distribution of the rostral group of ChAT-ir cells is similar to that of 5-HT-ir photoreceptors, they probably represent separated populations on the basis of the different thickness and shape of cells and the different distribution of their fibers. ChAT-ir receptors are also present in the pineal organ of lampreys (Pombal et al., in press). ChAT-ir cells were observed in the pineal organ, but not in the parapineal organ, of a teleost (Ekström and Korf, 1986; Ekström, 1987b). However, the ChAT-ir cells of the teleost pineal organ have been considered as interneurons. The presence of 5-HT-ir photoreceptors in the pineal complex of lamprey has been previously reported by Meiniel (1978, 1980) and by Tamotsu et al. (1990), who consider these cells as modified (neuroendocrine) photoreceptors. 5-HT-ir cells are far more numerous in the parapineal organ of river lampreys (Tamotsu et al., 1990; present results) than in that of sea lamprey. Axons of 5-HT-ir cells are scarce, branch only occasionally, and do not appear to leave the parapineal organ, terminating near the parapineal vesicle surface or occasionally entering the ganglion, which is in agreement with the photoneuroendocrine hypothesis (Collin, 1971; Meiniel, 1980; Tamotsu et al., 1990). The dorsorostral location of most 5-HT-ir cells of river lamprey does not correspond to those cells labeled by DiI application to the left habenula, nor has any 5-HT-ir fiber been observed in the parapineal tract. Distribution of 5-HT-ir cells and fibers in the parapineal organ was also very different from that observed for SP-ir elements, suggesting that they form different systems. The salmon parapineal may contain some 5-HT-ir elements (Ekström and Ebbesson, 1989). It is, however, intriguing that the left habenular nucleus of Oncorhynchus kisutch contains a subnucleus of 5-HT-ir cells (Ekström and Ebbesson, 1988), apparently located in a position similar to the area innervated by parapineofugal fibers in Oncorhynchus mykiss (Yáñez et al., 1996). Ekström and Ebbesson (1988) suggested that they may represent displaced pineal photoreceptors but, if they were modified photoreceptors, comparison with our results in lamprey would suggest the parapineal organ as a more probable source for these cells. Unlike the parapineal organ of lamprey and salmon, no immunoreactivity to 5-HT has been observed in the parapineal organ of a teleost, Gasterosteus aculeatus, at any stage of development (van Veen et al., 1984). The bipolar parapineal cells labeled by DiI application to the left habenula form a population of long axon cells, as was also indicated by Samejima et al. (1989) and Tamotsu et al. (1994), who considered them as photoreceptors. These cells had been labeled in the pineal and parapineal organ of lamprey by application of horseradish peroxidase to the ‘‘transected end of the pineal tract’’ (Samejima et al., 1989; Tamotsu et al., 1994). Our DiI results in lamprey, however, indicate that the targets of pineal and parapineal 186 J. YÁÑEZ ET AL. long axon bipolar cells are different, the pineal cells not being labeled from the left habenula, and the converse being true for those of the parapineal organ. The long axon cells of lampreys are roughly similar to bipolar cells revealed in the trout pineal organ after horseradish peroxidase labeling of the proximal pineal stalk (Ekström, 1987a), although the targets of these cells are not known. The ventral region of the parapineal organ of L. japonica is characterized by the presence of cells immunoreactive to rod-opsin, s-antigen, or visinin, thus differing from the dorsal region (Tamotsu et al., 1990). For L. fluviatilis, rod-opsin was also reported in only a circumscribed area of the ventral region (Vigh-Teichmann et al., 1983). Because in the two species studied here long axon cells occupy the ventral region of the parapineal vesicle, they might correspond with cells containing one or more of these substances. Parapineal efferents An interesting result of the present investigation is that the parapineal ganglion is innervated by pretectal and telencephalic neuronal populations. Moreover, results of immunocytochemistry indicate that the parapineal organ receives at least two main types of efferent fiber through the parapineal tract, which were GABA- and/or NPYimmunoreactive. GABA-ir fibers also enter the proximal wall of the parapineal vesicle. This finding suggests that the parapineal organ activity is modulated by these two systems in lamprey. Comparison of distribution of GABAand NPY-ir brain nuclei with the results of application of DiI to the parapineal organ, indicates that the NPY-ir efferent fibers probably originate in the subhippocampalthalamic nucleus. Because GABA-ir cells were observed both in the pretectal and subhippocampal nuclei, GABA-ir efferent fibers might originate in either of these nuclei. It is also conceivable that some GABA-ir efferent fibers in the parapineal ganglion also may be NPY-ir, as are those found in the saccus vasculosus system of trout (Yáñez et al., 1997), but this does not appear to be the case for GABA-ir fibers projecting to the vesicle. However, additional types of efferent fibers immunoreactive to galanin (river lamprey), SP (river lamprey, sea lamprey?), and somatostatin (sea lamprey, Yáñez et al., 1992) have been observed in small numbers in the parapineal ganglion, and the list of substances may increase with future studies. The efferent innervation of the parapineal organ is noticeable in the context of the pineal complex of other vertebrates, but it is not if we consider that the parapineal ganglion of lamprey has habenular characteristics. The efferent system of the parapineal organ is notably similar to that the habenulopetal system. In a previous study of the larval sea lamprey habenula (Yáñez and Anadón, 1994), only a single habenulopetal nucleus, located in the subhippocampal-thalamic nucleus, was reported. Neurons of this nucleus, and of other telencephalic neurons, were also labeled in adult river lamprey after DiI application to the left habenula (present results). The main problem with interpretation of the telencephalo-habenular innervation, however, was to know which telencephalic fibers running through the habenular commissure are en passant fibers and that, in fact, connect with the habenula (Northcutt and Puzdrowski, 1988; Yáñez and Anadón, 1994). This was not a problem with the parapineal organ because of its terminal position. The application of DiI to the parapineal organ labeled a single nucleus in the telencephalon, the subhippocampal-thalamic nucleus. This type of application showed fewer bilateral labeled neurons in the subhippocampal-thalamic nucleus than the application to the habenula (Yáñez and Anadón, 1994; present results), suggesting that many neurons of the subhippocampalthalamic nucleus only have projections to the habenula. After DiI application to the parapineal organ, labeled fibers were observed to run rostralward to the primordium hippocampi (Johnston, 1902; Heier, 1948). Our results suggest that they are collateral branches of axons of neurons of the subhippocampal-thalamic nucleus. Further evidence that the cells of the subhippocampal-thalamic nucleus may project simultaneously to the olfactory bulb and the parapineal organ comes from two in vitro experiments of application of biotinylated dextran amine to the olfactory bulb (M.A. Pombal, unpublished data). These experiments labeled both neurons in this nucleus and efferent fibers in the parapineal organ that were similar to those revealed by GABA and NPY immunocytochemistry. In a similar way, cells of the main habenulopetal telencephalic nucleus of trout (entopeduncular nucleus) appear to project both to the habenula and the lateral telencephalon (Yáñez and Anadón, 1996). The different position of the entopeduncular-telencephalic projection in trout may be due to the telencephalic eversion occurring in teleosts (Northcutt, 1995). The characteristics shared by the subhippocampal-thalamic nucleus of lamprey and the entopeduncular nucleus of trout are more easily explained assuming that they are homologous structures than if they have appeared independently. These nuclei would relay information from basal telencephalic systems to both the habenula and dorsal telencephalon (for a comparison of the entopeduncular nucleus of trout with habenulopetal nuclei of other vertebrates, see Yáñez and Anadón, 1996). Moreover, the application of DiI to the parapineal organ revealed a pretecto-parapinealopetal (and habenulopetal) system. Because DiI application to the pretectum led to fibers in both habenulae being labeled, this nucleus appears to have been overlooked in a previous study of larval lamprey habenula (Yáñez and Anadón, 1994), probably because of the very short distance from this nucleus to the application point, the poor development of the pretectal region in larvae, or both. The cells of the parapinealopetal (and habenulopetal) pretectal nucleus extend dendrites through a pretectal region which in lamprey receives both primary visual and pineal fibers (Vesselkin et al., 1980; De Miguel et al., 1990; Puzdrowski and Northcutt, 1989; Yáñez et al., 1993). These results, thus, suggest that activity of the lamprey parapineal ganglion and vesicle is under control of two other photoreceptive organs, the pineal organ and the retina, that may modulate the input of the parapineal organ on the habenulo-interpeduncular system by means of this nucleus. Parapineal-interpeduncular projection One of the most interesting findings of our study is the afferent projection from the parapineal ganglion to the interpeduncular nucleus by means of the left fasciculus retroflexus, as well as the close relationship between the parapineal vesicle on one side and both this ganglion and the left habenula on the other. DiI application to the parapineal organ produces labeling of this fascicle which is rather similar to that previously reported after habenular labeling (Yáñez and Anadón, 1994). Unlike the habenula, the ganglion only has projections to the rostral (isthmic) PARAPINEAL ORGAN CONNECTIONS IN LAMPREY portion of the interpeduncular complex, traditionally included in the mesencephalon (Heier, 1948), whereas habenular fibers also run in the rhombencephalon proper up to the level of the trigeminal nerve. The caudal habenular fibers of lamprey might be comparable to habenular fibers ending in the raphe and dorsal tegmental nucleus in trout (Yáñez and Anadón, 1996) and lizard (Dı́az and Puelles, 1992). Because the parapineal ganglion and all portions of the habenular nucleus of lamprey have projections to the interpeduncular nucleus (present results; Yañez and Anadón, 1994), they can only be compared with the medial habenular nucleus of mammals (Herkenham and Nauta, 1977; Contestabile and Flumerfelt, 1981). Functional considerations The ‘‘parapineal organs’’ of three different groups of vertebrates (lampreys, teleosts, and lizards) share a similar relationship with the habenulo-interpeduncular system (see above), a system which appears to be one of the most constant in all vertebrates (Herrick, 1948). In mammals, the habenulo-interpeduncular system is involved in limbic functions, including autonomic and endocrine control mechanisms, and aspects of sexual and defensive behavior (Herkenham and Nauta, 1979; Sutherland, 1982; Glaser and Barfield, 1984; Meszaros et al., 1985). However, practically nothing is known about the functions of this system in non-mammals. Herrick (1948) suggested that this system is the principal component that controls feeding behavior in ‘‘lower’’ species, mediating its integration with olfactory activity. Our anatomical results in lampreys do not contradict this view, but unfortunately, there are no functional studies of the habenulo-interpeduncular system in lampreys or other nonmammalian vertebrates. Experiments in lizard indicate that the parietal eye is responsive to light (Miller and Wolbarsht, 1962; Dodt and Scherer, 1968; Engbretson and Lent, 1976), being modulated by the pineal gland (Engbretson and Lent, 1976). However, the parietal eye plays little or no role in maintenance of circadian melatonin rhythm (Underwood and Calaban, 1987), circadian electroretinogram rhythm (Shaw et al., 1993), daily thermoregulatory behavior (Tosini and Menaker, 1996), or the locomotor rhythm (Innocenti et al., 1993), all these rhythms being affected by pineal manipulation. Experimental studies on possible functions of the pineal complex in lamprey (Joss, 1973a,b; Cole and Youson, 1981) have not distinguished between the pineal and the parapineal organs. Thus, the role played by the strikingly asymmetrical projection of parapineal organs in the left habenulo-interpeduncular system is presently a mystery that must wait to be resolved by future studies. CONCLUSIONS The major conclusion of this study is that the parapineal organ of lamprey is a part of the afferents to the habenulointerpeduncular system, both being specifically connected to the left habenula and innervated by efferents from the two main habenulopetal nuclei (i.e., the subhippocampalthalamic and pretectal nuclei). The parapineal vesicle is mainly connected to the parapineal ganglion by short axon cells, but also has long projections to the left habenula. Both the vesicle and the ganglion possess cell populations and fibers that are heterogeneous with respect to neurotransmitters, neurotransmitter synthesizing enzymes, neuropeptides, and a calcium binding protein, some of which 187 (SP, ChAT, galanin, CR) have been found in the lamprey parapineal for the first time. Finally, this study proposes a general distinction between pineal and parapineal organs of vertebrates on the basis of the ontogeny and the very specific relation of ‘‘true’’ parapineal organs with the left part of the habenulo-interpeduncular system. ACKNOWLEDGMENTS The authors thank Prof. S. Grillner (Karolinska Institute) for providing river lamprey brains and laboratory facilities to one of us (M.A.P.), and A. González and O. Marı́n (University Complutense of Madrid) for helping with ChAT immunocytochemistry. We also thank Dr. Christine Francis for linguistic revision. This research was supported by Spanish Education Ministry Grant PB96– 0945-C03 to R.A. and J.Y.; by Xunta de Galicia Grants XUGA20012B96 to M.A.P. and XUGA20002B97 to R.A. and J.Y.; and by University of Vigo Grant 64102C734 to M.A.P. LITERATURE CITED Beccari N. 1943. Neurologia Comparata. Sansoni, Firenze. Bolliet V, Ali M A, Anctil M, Zachmann A. 1993. Melatonin secretion in vitro from the pineal complex of the lamprey Petromyzon marinus. Gen Comp Endocrinol 89:101–106. Cole W C, Youson J H. 1981. The effect of pinealectomy, continuous light, and continuous darkness on metamorphosis of anadromous sea lampreys, Petromyzon marinus L. J Exp Zool 218:397–404. Cole W C, Youson J H. 1982. Morphology of the pineal complex of the anadromous sea lamprey, Petromyzon marinus L. Am J Anat 165:131– 163. Collin J P. 1971. Differentiation and regression of the cells of the sensory line in the epiphysis cerebri. In: Wolstenholme GEW, Knight J, editors. The Pineal Gland. London: Longman. p 79–125. Contestabile A, Flumerfelt B A. 1981. Afferent connections of the interpeduncular nucleus and the topographic organization of the habenulointerpeduncular pathway: An HRP study in the rat. J Comp Neurol 196:253–270. Contestabile A, Villani L, Fasolo A, Franzoni M F, Gribaudo L, Oktedalen O, Fonnum F. 1987. Topography of cholinergic and substance P pathways in the habenulo-interpeduncular system of rat. An immunocytochemical and microchemical approach. Neuroscience 21:253–270. Corujo A, Anadón R. 1986. Morfogénesis del diencéfalo de la trucha arcoiris: I. Epitálamo. Acta Microscópica 9:55–66. De Miguel E, Rodicio C, Anadón R. 1990. Organization of the visual system in larval lampreys: an HRP study. J Comp Neurol 302:529–542. Debreceni K, Fejer Z, Manzano e Silva M J, Vigh B. 1997. Immunoreactive glutamate in the pineal and parapineal organs of the lamprey. Neurobiology 5:53–56. Dendy A. 1911. On the structure, development and morphological interpretation of the pineal organs and adjacent parts of the brain in the tuatara (Sphenodon punctatus). Philos Trans R Soc Lond 201:227–331. Dı́az C, Puelles L. 1992. In vitro HRP-labeling of the fasciculus retroflexus in the lizard Gallotia galloti. Brain Behav Evol 39:305–311. Dodt E, Scherer E. 1968. Photic responses from the parietal eye of the lizard Lacerta sicula campestris (De Betta). Vision Res 8:61–72. Ekström P. 1987a. Photoreceptors and CSF-contacting neurons in the pineal organ of a teleost fish have direct axonal connections with the brain: An HRP-electron-microscopic study. J Neurosci 7:987–995. Ekström P. 1987b. Distribution of choline acetyltransferase-immunoreactive neurons in the brain of a cyprinid teleost (Phoxinus phoxinus L.). J Comp Neurol 256:494–515. Ekström P, Ebbesson S O. 1988. The left habenular nucleus contains a discrete serotonin-immunoreactive subnucleus in the coho salmon (Oncorhynchus kisutch). Neurosci Lett 91:121–125. Ekström P, Ebbesson S O. 1989. Distribution of serotonin-immunoreactive neurons in the brain of sockeye salmon fry. J Chem Neuroanat 2:201–213. Ekström P, Korf H W. 1986. Putative cholinergic elements in the photosensory pineal organ and retina of a teleost, Phoxinus phoxinus L. 188 (Cyprinidae): Distribution of choline acetyltransferase immunoreactivity, acetylcholinesterase-positive elements and pinealofugally projecting neurons. Cell Tissue Res 246:321–329. Eldred W D, Finger T E, Nolte J. 1980. Central projections of the frontal organ of Rana pipiens, as demonstrated by the anterograde transport of horseradish peroxidase. Cell Tissue Res 211:215–222. Engbretson G A, Lent C M. 1976. Parietal eye of the lizard: Neuronal photoresponses and feedback from the pineal gland. Proc Natl Acad Sci USA 7:654–657. Engbretson G A, Reiner A, Brecha N. 1981 Habenular asymmetry and the central connections of the parietal eye of the lizard. J Comp Neurol 198:155–165. Engbretson G A, Brecha N, Reiner A. 1982 Substance P-like immunoreactivity in the parietal eye visual system of the lizard Uta stansburiana. Cell Tissue Res 227:543–554. Garcı́a-Fernández J M, Jiménez A J, González B, Pombal M A, Foster R G. 1997. An immunocytochemical study of encephalic photoreceptors in three species of lamprey. Cell Tissue Res 288:267–278. Glaser J H, Barfield R J. 1984. Blockade of progesterone-activated estrous behavior in rats by intracerebral anisomycin is site specific. Neuroendocrinology 38:337–343. Hafeez M A, Merhige M E. 1977. Light and electron microscopic study of the pineal complex of the coelacanth, Latimeria chalumnae Smith. Cell Tissue Res 178:249–265. Heier P. 1948. Fundamental principles in the structures of the brain: A study of the brain of Petromyzon fluviatilis. Acta Anat Suppl 8:1–213. Herkenham M, Nauta W J H. 1977. Afferent connections of the habenular nuclei in the rat: A horseradish peroxidase study, with a note on the fiber-of-passage problem. J Comp Neurol 173:123–146. Herkenham M, Nauta W J H. 1979. Efferent connections of the habenular nuclei in the rat. J Comp Neurol 187:19–48. Herrick C J. 1948. The Brain of the Tiger Salamander. Chicago: University of Chicago Press. Hill C. 1891. Development of the epiphysis in Coregonus albus. J Morphol 5:503–510 Holmqvist B I, Östholm T, Ekström P. 1992. DiI tracing in combination with immunocytochemistry for analysis of connectivities and chemoarchitectonics of specific neural systems in a teleost, the Atlantic salmon. J Neurosci Methods 42:45–63. Innocenti A, Minutini L, Foa A. 1993. The pineal and circadian rhythms of temperature selection and locomotion in lizards. Physiol Behav 53:911– 915. Jiménez A J, Mancera J M, Pombal M A, Pérez-Fı́garez J M, FernándezLlébrez P. 1996 Distribution of galanin-like immunoreactive elements in the brain of the adult lamprey Lampetra fluviatilis. J Comp Neurol 368:185–197. Johnston J B. 1902. The brain of Petromyzon. J Comp Neurol 12:2–86. Joss J M P. 1973a. Pineal-gonad relationships in the lamprey Lampetra fluviatilis. Gen Comp Endocrinol 21:118–122. Joss J M P. 1973b. The pineal complex, melatonin, and color change in the lamprey Lampetra. Gen Comp Endocrinol 21:188–195. Joss J M P. 1977. Hydroxyindole-O-methyltransferase (HIOMT) activity and the uptake of 3H-melatonin in the lamprey, Geotria australis Gray Gen Comp Endocrinol 31:270–275. Kappers J A. 1965. Survey of the innervation of the epiphysis cerebri and the accessory pineal organs of vertebrates. In: Kappers JA, Schadé JP, editors. Structure and Function of the Epiphysis Cerebri. Progress in Brain Research, vol. 10. Amsterdam: Elsevier. p 87–153. Kuo C H, Tamotsu S, Morita Y, Shinozawa T, Akiyama M, Miki N. 1988 Presence of retina-specific proteins in the lamprey pineal complex. Brain Res 442:147–151. Marı́n O, Smeets W J A J, González A. 1997 Distribution of choline acetyltransferase immunoreactivity in the brain of anuran (Rana perezi, Xenopus laevis) and urodele (Pleurodeles waltl) amphibians. J Comp Neurol 382:499–534. Meiniel A. 1969. Etude préliminaire de l’organe parapinéal de Lampetra planeri. Arch Anat Microsc Morphol Exp 58:219–237. Meiniel A. 1978. Présence d’indolamines dans les organes pinéal and parapinéal de Lampetra planeri (Pétromyzontoı̈des). C.R. Acad Sci Paris (D) 287:313–316. Meiniel A. 1980. Ultrastructure of serotonin-containing cells in the pineal organ of Lampetra planeri (Petromyzontidae). Cell Tissue Res 207:407– 427. J. YÁÑEZ ET AL. Meiniel A, Collin J P. 1971. Le complexe pinéal de l’ammocète (Lampetra planeri, Bl.). Z Zellforsch 117:345–380. Meszaros J, Gajewska S, Tarchalska-Krynska B. 1985. Habenulointerpeduncular lesions: The effects on pain sensitivity, morphine analgesia and open-field behavior in rats. Pol J Pharmacol Pharm 37:469–477. Miller W H, Wolbarsht M L 1962. Neural activity in the parietal eye of lizard. Science 135:316–317. Morita Y, Dodt E. 1973. Slow photic responses of the isolated pineal organ of lamprey. Nova Acta Leopold 38:331–339. Morita Y, Tabata M, Uchida K, Samejima M. 1992. Pineal-dependent locomotor activity of lamprey, Lampetra japonica, measured in relation to LD cycle and circadian rhythmicity. J Comp Physiol A 171:555–562. Northcutt R G. 1995. The forebrain of gnathostomes: In search of a morphotype. Brain Behav Evol 46:275–318. Northcutt R G, Puzdrowski R L. 1988. Projections of the olfactory bulb and nervus terminalis in the silver lamprey. Brain Behav Evol 32:96–107. Nowikoff M. 1910. Untersuchungen über den Bau, die Entwicklung und die Bedeutung des Parietalauges von Sauriern. Z Wiss Zool 96:118–207. Nozaki M, Gorbman A. 1986. Occurrence and distribution of substance p-related immunoreactivity in the brain of adult lampreys, Petromyzon marinus and Entosphenus tridentatus. Gen Comp Endocrinol 62:217– 229. Ortman R. 1960. Parietal eye and nerve in Anolis carolinensis. Anat Rec 137:386. Pombal M A, El Manira A, Grillner S. 1997 Organization of the lamprey striatum: Transmitters and projections. Brain Res 766:249–254. Pombal M A, Yáñez J, Marı́n O, González A, Anadón R. 1998. Cholinergic and GABAergic neuronal elements in the pineal organ of lamprey, and tract-tracing observations of differential connections of pinealofugal neurons. Cell Tissue Res. in press Pu G A, Dowling J E. 1981. Anatomical and physiological characteristics of pineal photoreceptor cell in the larval lamprey, Petromyzon marinus. J Neurophysiol 46:1018–1038. Puzdrowski R L, Northcutt R G. 1989. Central projections of the pineal complex in the silver lamprey Ichthyomyzon unicuspis. Cell Tissue Res 225:269–274. Ronnekleiv O K, Kelly M J. 1984 Distribution of substance p neurons in the epithalamus of the rat: An immunohistochemical investigation. J Pineal Res 1:355–370. Rüdeberg C. 1969. Structure of the parapineal organ of adult rainbow trout, Salmo gairdneri Richardson Z Zellforsch 93:282–304. Samejima M, Tamotsu S, Watanabe K, Morita Y. 1989. Photoreceptor cells and neural elements with long axonal processes in the pineal organ of the lamprey, Lampetra japonica, identified by use of the horseradish peroxidase method. Cell Tissue Res 258:219–224. Schwaller B, Buchwald P, Blümcke I, Celio M R, Hunziker W. 1993 Characterization of a polyclonal antiserum against the purified human recombinant calcium-binding protein calretinin. Cell Calcium 14:639– 648. Shaw A P, Collazo C R, Easterling K, Young C D, Karwoski C J. 1993. Circadian rhythm in the visual system of the lizard Anolis carolinensis. J Biol Rhythms 8:107–124. Schöber W. 1964. Vergleichend-anatomische Untersuchungen am Gehirn der Larven und adulten Tiere von Lampetra fluviatilis (Linné, 1758) und Lampetra planeri (Bloch, 1784). J Hirnforsch 7:107–209. Studnicka F K. 1905. Die Parietalorgane. In: Lehrbuch der Microskopischen Anatomie der Wirbeltiere. Jena: Gustav Fisher, pp. 1–256. Sutherland R J. 1982. The dorsal diencephalic conduction system: A review of the anatomy and functions of the habenular complex. Neurosci Biobehav Rev 6:1–13. Taban C H, Cathieni M. 1983. Distribution of substance P-like immunoreactivity in the brain of the newt (Triturus cristatus). J Comp Neurol 216:453–470. Tamotsu S, Morita Y. 1986 Photoreception in pineal organs of larval and adult lampreys, Lampetra japonica. J Comp Physiol A 159:1–5. Tamotsu S, Korf H-W, Morita Y, Oksche A. 1990. Immunocytochemical localization of serotonin and photoreceptor-specific proteins (rod-opsin, S-antigen) in the pineal complex of the river lamprey, Lampetra japonica, with special reference to photoneuroendocrine cells. Cell Tissue Res 262:205–216. PARAPINEAL ORGAN CONNECTIONS IN LAMPREY Tamotsu S, Samejima M, Morita Y. 1994. Subtypes of pineal photoreceptors in lamprey identified by means of tracing and immunocytochemical techniques. Adv Pineal Res 8:25–29. Tosini G, Menaker M. 1996. The pineal complex and melatonin affect the expression of the daily rhythm of behavioral thermoregulation in the green iguana. J Comp Physiol A 179:135–142. Tretjakoff D. 1915. Die Parietalorgane von Petromyzon fluviatilis. Z Wiss Zool 113:1–112. Underwood H, Calaban M. 1987. Pineal melatonin rhythms in the lizard Anolis carolinensis: II. Photoreceptive inputs. J Biol Rhythms 2:195– 206. van Veen T, Ekström P, Nyberg L, Borg B, Vigh-Teichmann I, Vigh B. 1984. Serotonin and opsin immunoreactivities in the developing pineal organ of the three-spined stickleback, Gasterosteus aculeatus L. Cell Tissue Res 237:559–564. Vesselkin N P, Ermakova T V, Repérant J, Kosareva A A, Kenigfest N B. 1980. The retinofugal and retinopetal systems in Lampetra fluviatilis: An experimental study using radioautographic and HRP methods. Brain Res 195:453–460. Vigh-Teichmann I, Korf H-W, Nürnberger F, Oksche A, Vigh B, Olson R. 1983. Opsin immunoreactive outer segments in the pineal and parapineal organs of the lamprey (Lampetra fluviatilis), the eel (Anguilla anguilla), and the rainbow trout (Salmo gairdneri). Cell Tissue Res 230:289–307. von Haffner K. 1953. Untersuchungen über die Entwicklung des Parietalorgans und des Parietalnerven von Lacerta vivipara und das Problem der Organe der Parietalregion. Z Wiss Zool 157:1–34. 189 Weld M M, Maler L. 1992. Substance P-like immunoreactivity in the brain of the gymnotiform fish Apteronotus leptorhynchus: Presence of sex differences. J Chem Neuroanat 5:107–129. Yáñez J, Anadón R. 1994 Afferent and efferent connections of the habenula in the larval sea lamprey (Petromyzon marinus L.): An experimental study. J Comp Neurol 345:148–160. Yáñez J, Anadón R. 1996. Afferent and efferent connections of the habenula in the rainbow trout (Oncorhynchus mykiss): An indocarbocyanine dye (DiI) study. J Comp Neurol 372:529–543. Yáñez J, Rodrı́guez-Moldes I, Anadón R. 1992. Distribution of somatostatinimmunoreactivity in the brain of the larval lamprey (Petromyzon marinus). J Chem Neuroanat 5:511–520. Yáñez J, Anadón R, Holmqvist B I, Ekström P. 1993 Neural projections of the pineal organ in the larval sea lamprey (Petromyzon marinus L.) revealed by indocarbocyanine dye tracing. Neurosci Lett 164:213–216. Yáñez J, Meissl H, Anadón R. 1996. Central projections of the parapineal organ of the adult rainbow trout (Oncorhynchus mykiss). Cell Tissue Res 285:69–74. Yáñez J, Rodrı́guez M, Pérez S, Adrio F, Rodrı́guez-Moldes I, Manso M J. Anadón R. 1997. The neuronal system of the saccus vasculosus of trout (Salmo trutta fario and Oncorhynchus mykiss): An immunocytochemical and nerve tracing study. Cell Tissue Res 288:497–507. Zhang W, Pombal M A, El Manira A, Grillner S. 1996. Rostrocaudal distribution of 5-HT innervation in the lamprey spinal cord and differential effects of 5-HT on fictive locomotion. J Comp Neurol 374:278–290.