Retrograde study of projections from the tuberomammillary nucleus to the dorsal raphe and the locus coeruleus in the rat

Brain Research 1043 (2005) 65 – 75 www.elsevier.com/locate/brainres

Research report

Retrograde study of projections from the tuberomammillary nucleus to the dorsal raphe and the locus coeruleus in the rat Hyun S. Leea,*, Beob Y. Leeb, Barry D. Waterhousec a

Department of Premedical Science, College of Medicine, Konkuk University, Chungju, Chungbuk 380-701, South Korea b Department of Anatomy, College of Medicine, Konkuk University, Chungju, Chungbuk 380-701, South Korea c Department of Neurobiology and Anatomy, College of Medicine, Drexel University, 2900 Queen Lane, Philadelphia, PA 19129, USA Accepted 12 February 2005 Available online 1 April 2005

Abstract In the first series of experiments, a retrograde tracer, WGA-apo-HRP-gold (WG), was injected into the dorsal raphe (DR) or the locus coeruleus (LC) and adenosine deaminase immunostaining was subsequently performed for the tuberomammillary nucleus (TMN) in order to investigate projections from the TMN to the two brainstem monoaminergic nuclei. Following rostral DR injections, the majority of retrogradely labeled cells were located in the dorsomedial and ventrolateral subdivisions of the TMN. At middle DR levels, midline injections resulted in labeling mainly in the ventrolateral subdivision, whereas lateral wing injections produced labeling mostly in ventral and caudal TMN subdivisions. When injections were made in the caudal DR, only a few cells were observed along the rostro-caudal extent of the TMN. On the other hand, following rostral LC injections, labeled neurons were observed mainly in ventrolateral and ventral subdivisions of TMN. For principal LC injections, labeled cells were observed mostly in ventrolateral, ventral, and caudal TMN subdivisions, whereas for injections at caudal pole of LC, only a few cells were located along the rostro-caudal extent of the TMN. In the second series of experiments, an iontophoretic injection of fluorogold (FG) into the DR was paired with a pressure injection of WG into the LC to investigate the collateral distribution of TMN axonal fibers to DR and LC. Double-labeled cells were observed in ventrolateral, ventral, and caudal TMN subdivisions. The present study indicated that there exists a robust projection from the TMN to the DR or the LC and that some TMN neurons have axon collaterals projecting to both DR and LC. The TMN neurons with such axon collaterals might provide simultaneous, possibly more efficient, way of controlling the brainstem monoaminergic nuclei, thus influencing various sleep and arousal states of the animal. D 2005 Elsevier B.V. All rights reserved. Theme: Neural basis of behavior Topic: Monoamines and behavior Keywords: Tuberomammillary nucleus; Dorsal raphe; Locus coeruleus; Histamine cells; Serotonin cells; Norepinephrine cells

1. Introduction The tuberomammillary nucleus (TMN) in the posterior hypothalamus consists of several clusters of histaminergic cells: (1) a dorsomedial group, adjacent to the mammillary recess of the 3rd ventricle, (2) ventrolateral and ventral groups, laminae of tightly-packed cells located at the ventral, sub-pial surface, and (3) a caudal group, located * Corresponding author. Fax: +82 43 851 9329. E-mail address: [email protected] (H.S. Lee). 0006-8993/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2005.02.050

lateral to the lateral mammillary nucleus [40,42,53]. Neurons in the TMN as well as other two brainstem monoaminergic nuclei, the dorsal raphe (DR) and the locus coeruleus (LC), have common characteristics in that they exhibit maximal rates of discharge during waking, decrease their firing during non-rapid eye movement (NREM) sleep, and virtually become quiescent during rapid eye movement (REM) sleep [1,2,11,28,41,46,48]. The three monoaminergic nuclei are also similar in that they serve as major arousal centers and project extensively to the cerebral cortex [18,33,34].

66

H.S. Lee et al. / Brain Research 1043 (2005) 65 – 75

Using an anterograde tracing method, Steininger et al. [43] provided evidence for sub-regional organization of preoptic area projections to arousal-related monoaminergic cell groups; for example, ventrolateral preoptic nucleus (VLPO) provides a robust innervation to the histaminergic tuberomammillary nucleus (TMN), whereas VLPO/medial preoptic area and VLPO/lateral preoptic region send dense projections to the LC and the DR, respectively. The ventral portions of the medial preoptic area and the lateral preoptic nucleus correspond to dorsomedial and dorsolateral regions of the VLPO, respectively, which are collectively called ‘‘extended VLPO’’ by Lu et al. [26]. A functional differentiation between the VLPO and the extended VLPO has been suggested such that: (1) during NREM sleep, the sleepactive neurons in the VLPO might inhibit the activity of cells in the TMN, DR, and LC by releasing GABA and galanin; and (2) during the transitions from NREM to REM sleep, the firing of DR and LC is further decreased because of recruitment of inhibitory neurons in the extended VLPO [26,37,38,45]. Thus, one implication of the above findings is that TMN communicates with DR and LC during various sleep – waking states of the animal. Further support for this hypothesis derives from the fact that histamine-like immunoreactive fibers have been observed in the DR and the LC [13,22] and the TMN, in return, receives noradrenergic and serotonergic inputs from brainstem regions [7]. The present report is the first extensive, retrograde study of the projections from subdivisions of the TMN to those of the DR and/or the LC. A retrograde tracer, WGA-apo-HRPgold, was localized within specific subdivisions of the DR or the LC and subsequent adenosine deaminase (ADA) immunostaining was performed for TMN sections to identify neurons projecting to the DR or the LC. Collateral projections of TMN axonal fibers to the DR and LC were also analyzed by making paired injections of fluorogold (FG) into the DR and WG into the LC.

2. Materials and methods A total of 44 Sprague – Dawley rats including both sexes and ranging in weight from 300 to 350 g were used in this study. Prior to surgery, each rat was anesthetized with an intraperitoneal injection of chloral hydrate (3.6% in distilled water, 1 ml/100 g body weight). All animals used in this study were treated according to guidelines approved by the institutional animal care and use committee and conformed to the NIH guidelines on care and use of animals in research. 2.1. DR injection sites The skull around bony lambda was removed and the superior sagittal sinus ligated with surgical sutures rostrally and caudally. Angiovasectomy was performed between suture points in order to expose the cerebral fissure at midline. Based on the atlas of Paxinos and Watson [32], the

rostral DR was approached for sites between 0.5– 1.0 rostral to bony lambda at the depth of 5.2– 5.4 mm. The midline at middle DR levels was targeted for sites between 0 –0.5 mm caudal to the lambda at the depth of 5.6– 5.8 mm, whereas the lateral wing subdivision was approached from 0.3– 0.4 mm lateral to the midline at the depth of 5.4– 5.6 mm. The caudal DR was targeted 1.0 –1.5 mm caudal to the lambda at the depth of 6.0 –6.2 mm. 2.2. LC injection sites Rats were placed in a stereotaxic apparatus. The incisor bar was set at 10.0 mm, thus tilting the skull surface 10– 20- downward from the horizontal plane. Based on the atlas of Paxinos and Watson [32], the rostral LC was targeted 1.1 mm lateral from the midline, 0.2 –0.4 mm caudal to the posterior border of the transverse sinus at the depth of 5.5– 5.7 mm, whereas the principal LC was 1.2 mm lateral from the midline, 0.6 –0.8 mm caudal to the sinus at the depth of 5.6– 5.8 mm. Finally, the caudal pole of LC was at 1.0– 1.2 mm caudal to the sinus, 1.2 mm lateral from the midline, and 5.7– 5.9 mm deep from the dural surface. 2.3. WG injection The WG was synthesized using inactivated WGA-HRP (Sigma, L-0390) and 10 or 20 nm (Sigma, G1527 or G1652) colloidal gold [4]. The injection apparatus consisted of a glass micropipette (tip diameter, 10 –15 Am) hydraulically linked to a 2.0 Al Hamilton syringe. We injected 0.02 Al of WG nine times (a total of 0.18 Al) at 3-min intervals into a single site within the DR or the LC and took out the pipette 30 min after the first injection trial. 2.4. FG injection To examine the collateral distribution of TMN axonal fibers to the DR and the LC, WG was pressure-injected into subdivisions of the LC, whereas FG was iontophoretically injected into midline at middle DR levels. The opposite combination was also performed but proved to be inappropriate because FG produced an extensive injection site with a mediolateral dimension of 1300 – 1500 Am; the LC, however, is a nucleus with a tight mediolateral dimension (150-350 Am). A solution of 1% FG (Fluorochrome Inc.) was prepared in saline, drawn into the tip of a glass micropipette (tip diameter, 10 – 15 Am) via capillary action, and deposited within the DR nuclei using a 2 AA alternating current applied on a 5-s duty cycle for 10– 20 min through a silver lead wire (Stoelting, 50880) inserted in the pipette. 2.5. Perfusion – fixation and silver enhancement reaction After a survival period of 48– 72 h following tracer injections, the animals were perfused using 150 ml of saline followed by 600 ml of fixative containing 4% paraformal-

H.S. Lee et al. / Brain Research 1043 (2005) 65 – 75

dehyde in 0.01 M phosphate-buffered saline (PBS, pH 7.4). The perfusion –fixation was completed with 100 ml of PBS containing 10% sucrose. The brain was then removed and stored in 30% sucrose solution in PBS overnight. A series of 40 Am sections were prepared using a cryostat. Following rinses with distilled water, the WG was detected using a commercial silver intensification kit (Sigma, SE-100) as described in L-Smith et al. [25]. 2.6. ADA immunocytochemistry Tissue sections through the TMN were processed for conventional ADA immunostaining to identify TMN neurons and the boundary of TMN [27]. Sections were washed in 0.1 M Tris-buffered saline (TBS, pH 7.4) and incubated in 0.5% H2O2 in TBS for 20 min to inhibit

67

endogenous peroxidase activity. After rinses, sections were incubated in 10% normal goat serum for 1 h to block nonspecific binding. After a brief rinse, free-floating sections were then incubated in 1:1000 dilution of rabbit anti-ADA (Chemicon, AB 176) for 48 h (4 -C). Sections were incubated in 1:100 dilution of goat anti-rabbit IgG. After rinses, sections were then incubated in rabbit peroxidase – anti-peroxidase (PAP) with a dilution of 1:400. After rinses with 0.1 M Tris buffer (TB, pH 7.4), sections were reacted with 3,3V-diaminobenzidine (DAB)-H2O2 kit (Vector, SK4100) for 1 –2 min (4 -C). 2.7. FG immunocytochemistry For animals with dual injection of WG and FG, silver enhancement reaction for WG was followed by FG

Fig. 1. In order to investigate the projection from the tuberomammillary nucleus (TMN) to the dorsal raphe (DR), a retrograde tracer, WGA-apo-HRP-gold (WG), was injected into rostral (A, R123), middle (B, R93), and caudal (C, R128) levels of the midline DR as well as lateral wing (D, R106) subdivision. The transition from rostral (A) to middle (B) DR sections was identified by the widening of cerebral aqueduct (Aq), whereas the transition from the middle (B) to caudal (C) DR regions was marked by the appearance of the 4th ventricle (4 V). Midline injections (A – C) were counterstained with neutral red, whereas lateral wing injection (D) was immunostained for 5-HT to examine whether the tracer was confined within the boundary of DR. An inset in panel (D) depicts higher magnification view of 5-HT-immunostained DR cells indicated by arrowheads at lower magnification. On the other hand, panel (E) represents iontophoretic injection of fluorogold (FG) into the midline DR (R78), which was combined with WG injection into the principal LC in order to investigate TMN cells with collateral distribution to the DR and the LC (Fig. 4; Table 2). Unlike the focal injection sites with WG (A – D), an iontophoretic injection of FG extended mediolaterally and dorsoventrally, thus a midline injection involved dorsomedial, ventromedial, and lateral wing subdivisions (E). At the iontophoretic injection sites, there are terminal boutons (E, inset; open arrows) as well as labeled cells (E, inset; FG cells). DRc, caudal dorsal raphe; DRdm, dorsomedial subdivision; DRlw, lateral wing subdivision; DRm, middle dorsal raphe; DRr, rostral dorsal raphe; DRvm, ventromedial subdivision; DTg, dorsal tegmental n.; mlf, medial longitudinal fasciculus. Scale bars = 50 Am.

1 1 2 1 1 2 4 4 0 1 3 2 2 2 5 3 0 2 5 1 4 10 9 8 2 3 1 0 3 7 5 6 2 4 3 1 6 15 14 2 1 5 2 0 4 12 10 1 2 4 5 3 3 5 4 4 1 4 6 2 4 7 15 12 2 3 6 5 6 5 12 11 7 28 5 1 5 33 6 2 32 18 7 3 Dorsomedial TMN 34 Ventrolateral TMN 19 Ventral TMN 9 Caudal TMN 5

Rostral Rostral Middle Middle Middle lat. wing Middle lat. wing Caudal Caudal Rostral LC Rostral LC Principal LC Principal LC Caudal pole LC Caudal pole LC DR DR midline midline Ipsi- Contra- Ipsi- Contra- DR DR Ipsi- Contra- Ipsi- Contra- Ipsi- Contra- Ipsi- Contra- Ipsi- Contra- Ipsi- ContraInjection sites

R131 R89 R98 R94 R96 R85 R128 R109 R106 R102

Midline injections at rostral, middle, and caudal DR levels are counterstained with neutral red [16]; lateral wing injection sites, however, are immunostained for 5-HT to examine whether the tracer is confined within the boundary of DR. Briefly, sections were incubated in 0.5% H2O2 in TBS for 20 min and subsequently in 10% normal rabbit serum for 1 h. They were then incubated in 1:100 dilution of rat anti-5-HT (Chemicon, MAB352) for 48 h (4 -C). After rinses, sections were incubated in 1:100 dilution of rabbit anti-rat IgG (Chemicon, AP132). Sections were then incubated in 1:400 dilution of rat PAP (Chemicon, PAP20). After rinses with TB, sections were reacted with DAB-H2O2 kit (Vector, SK-4100) for 1 – 2 min (4 -C). For all immunocytochemical procedures, positive tissue controls were performed using sections from other DBHor 5-HT-immunoreactive brain regions. Negative control experiments included omission of the primary antibody and substitution of an equivalent dilution of normal serum, or reaction of a series of sections with increasing dilutions of the primary antibody until all staining was lost. Similarly, omission of the secondary antibody or the PAP solution was also performed to indicate whether either the reagents or the procedures would give rise to non-specific staining.

R93

2.9. Serotonin (5-hydroxy-tryptamine, 5-HT) immunocytochemistry

R91

Tissue sections through LC injection site were processed for DBH immunostaining, using the PAP method of Sternberger [44]. Sections were washed in 0.1 M TBS (pH 7.4) and incubated in 0.5% H2O2 in TBS for 20 min. After rinses, sections were incubated in 10% normal rabbit serum (Vector, S-1000) for 1 h. After a brief rinse, freefloating sections were then incubated in 1:400 dilution of mouse anti-DBH (Chemicon, MAB 308) for 48 h (4 -C). Sections were incubated in 1:160 dilution of rabbit antimouse IgG (Chemicon, AP160). After rinses, sections were then incubated in mouse PAP (Chemicon, PAP14) with a dilution of 1:80. After rinses with 0.1 M TB (pH 7.4), sections were reacted with DAB-H2O2 kit (Vector, SK4100) for 1 –2 min (4 -C).

R123

2.8. Dopamine-b-hydroxylase (DBH) immunocytochemistry

R104

immunostaining. Sections were washed with 0.02 M potassium phosphate-buffered saline (KPBS, pH 7.4). They were then incubated in 1:500 dilution of rabbit antifluorogold (Fluorochrome Inc.) dissolved in KPBS, which contains 1% normal goat serum, 0.3% Triton, and 1% bovine serum albumin for 48 h (4 -C). Following rinses with KPBS, sections were then incubated in 1:250 dilution of biotinylated goat anti-rabbit antiserum (Vector, BA-1000) made with KPBS for 1 h. After rinses, sections were incubated in ABC complex for 1 h and then reacted with DAB-H2O2 kit (Vector, SK-4100) for 1 – 2 min (at 4 -C).

Cases

H.S. Lee et al. / Brain Research 1043 (2005) 65 – 75

Table 1 Distribution of retrogradely labeled cells in each subdivision of the tuberomammillary nucleus (TMN) following the injection of a retrograde tracer, WG, within DR or LC

68

H.S. Lee et al. / Brain Research 1043 (2005) 65 – 75

3. Results In an initial series of experiments (n = 33 rats), a retrograde tracer, WG, was injected into the subdivisions of DR or LC nuclei and subsequent ADA immunostaining was performed for TMN sections in order to investigate the projections from the TMN to the brainstem monoaminergic nuclei. In a second series of experiments (n = 11 rats), the possibility of collateral projections from TMN to the DR and the LC was assessed using a combined injection of FG into the DR and WG into the LC. 3.1. Projection from the TMN to the DR or the LC Among 21 cases of DR injection, eight representative cases are presented in detail; i.e., injection sites within rostral (R123, Fig. 1A), middle (R93, Fig. 1B), caudal (R128, Fig. 1C) levels of the midline DR as well as lateral wing (R106, Fig. 1D) subdivisions. In case of the lateral wing, injection was made unilaterally (Fig. 1D). Four injection cases at each

69

DR subdivision (R104, R91, R109, and R102) are also included in Table 1. The labeling patterns obtained in these cases are representative of those seen with remaining cases that have not been mentioned. Following rostral DR injections (R104 and R123), labeled cells were observed mainly in the dorsomedial as well as ventrolateral subdivisions at rostral TMN levels (Table 1). For midline injections at middle DR levels (R91 and R93), labeled cells were located mostly in the ventrolateral TMN subdivision, whereas they were located in ventral and caudal subdivisions following lateral wing injections (R102 and R106). For the latter injection cases, the ipsilateral projection was more pronounced than the contralateral one. When injections were made in the caudal DR (R109 and R128), only a few cells were observed along the rostro-caudal extent of the TMN. Among 12 cases of LC injection, six representative cases are presented in detail, i.e., injection sites within rostral (R85, Fig. 2A), principal (R98, Fig. 2B), and caudal pole (R89, Fig. 2C) subdivisions. All the LC injections were made unilaterally (Fig. 2). Three injection cases at each LC

Fig. 2. In order to investigate the projection from the TMN to the locus coeruleus (LC), a retrograde tracer, WG, was injected into rostral (A, R85), principal (B, R98), and caudal pole (C, R89) subdivisions of the LC. Injection sites were immunostained for dopamine-h-hydroxylase (DBH) to examine whether the tracer was confined within the LC core. DBH-immunostained cells at the injected side (left) were not easy to identify due to WGA-apo-HRP-gold granules. Thus, higher magnification views of DBH-immunoreactive LC cells (arrowheads) contralateral to the injection sites are shown (A – C, insets). Some of these cells also contain WG granules (B, DBH + WG), which represent LC cells projecting to the contralateral side. Cb, cerebellum; LCcp, caudal pole of LC; LCp, principal subdivision; LCr, rostral subdivision; me5, mesencephalic trigeminal tract; scp, superior cerebellar peduncle; 4 V, 4th ventricle. Scale bars = 50 Am.

70

H.S. Lee et al. / Brain Research 1043 (2005) 65 – 75

subdivision (R96, R94, and R131) are also included in Table 1. The labeling patterns obtained in these cases are representative of those seen with remaining cases that have not been mentioned. Following rostral LC injections (R85 and R96), labeled neurons were observed mainly in ventrolateral and ventral TMN subdivisions (Table 1). For principal LC injections (R94 and R98), labeled cells were observed mostly in ventrolateral, ventral, and caudal TMN

subdivisions (Table 1). Following injections at caudal pole of LC (R89 and R131), only a few cells were located along the rostro-caudal extent of the TMN (Table 1). For all LC injection cases, the ipsilateral projection was more pronounced than the contralateral one (Table 1). Representative examples of double-labeled (ADA + WG) TMN neurons following injections of WG into each DR subdivision were shown in Figs. 3A – D. DR-projecting

Fig. 3. For all TMN sections (A – H), adenosine deaminase (ADA) immunostaining was performed to identify TMN neurons (brown cells) as well as the boundary of TMN. Micrographs (A) – (D) represent retrogradely labeled cells at dorsomedial (A, R123), ventrolateral (B, R93; C, R128), and ventral (D, R106) subdivisions of the TMN, following the injection of WG into rostral (R104, R123), middle (R91, R93), caudal (R109, R128) levels of the midline DR as well as lateral wing subdivision (R102, R106). Micrographs (E) – (H) represent retrogradely labeled cells at ventral (E, R96; G, R89), ventrolateral (F, R98), and caudal (H, R98) subdivisions of the TMN, following the injection of WG into rostral (R85, R96), principal (R94, R98), and caudal pole (R89, R131) subdivisions of the LC. Insets in panels (A) – (H) represent higher magnification views of single- (ADA and WG) or double-labeled (ADA + WG) TMN cells indicated by arrowheads or asterisks at lower magnification. cp, cerebral peduncle; LM, lateral mammillary n.; ML, lateral portion of the medial mammillary n.; MRe, mammillary recess of 3rd ventricle; TMc, caudal TMN; TMdm, dorsomedial TMN; TMv, ventral TMN; TMvl, ventrolateral TMN. Scale bars = 25 Am.

9.5 0 18.2 19 0 4 D, double; FG, fluorogold; WG, WGA-apo-HRP-gold. a The number represents the percentage of double-labeled (D) cells over the combined number (D + WG) of double- (D) and single-labeled (WG), LC-projecting neurons. b The number indicates the percentages of double-labeled (D) neurons over the combined number (D + FG) of double- (D) and single-labeled (FG), DR-projecting neurons.

9 8 15.4 2 25.0 0 11.8 11 7.7 3 14.3 2 15 0 1 12 13.0 18 0 10 21.1 3 20 28.6 0 5 15.4 15 11.8 5 7.1 4 22 0 2 15 6.7 13 0 1 10.5 15 0 5 2 17 0 4

11.8 1 14 0 0 2

8.7 16.7 21 11.1 2 10 26.7 32 13.8 4 11 21.1 25 11.1 4 15 21.7 40 12.0 5 18 15.8 22 7.0 3 16 13.6 40 3 19

0 2 0 3 0 0 9 0 4 0 0 3 0 0 0 0 8 0 5 0 0 3 0 5 0 0 8 0 7 0

Dorsomedial TMN Ventrolateral TMN Ventral TMN Caudal TMN

FG %b Middle DR R 134

D WG %a Caudal pole LC FG %b Middle DR

R 107

D WG %a Caudal pole LC FG %b Middle DR

R83

D WG %a Principal LC FG %b Middle DR

R78

D WG %a Principal LC FG %b Middle DR

R 99

D WG %a Rostral LC FG %b Middle DR D WG %a Rostral LC

A series of experiments (11 injection cases) combining FG injection into the middle DR and WG into subdivisions of the LC were performed to investigate the collateral distribution of TMN axonal fibers to DR and LC. An iontophoretic injection of FG always produced a relatively large injection site with a mediolateral dimension of 1300 –1500 Am (Fig. 1E), thus such injections were only made along the midline at middle DR levels between the lateral wings. On the other hand, it was possible to inject WG into subdivisions of the LC (refer to Table 2 for LC injection sites). The opposite combination was also performed but proved to be inappropriate, since the core of the LC is confined within a narrow (approximately 150 – 350 Am) mediolateral dimension (Figs. 2A – C). Among 11 cases of injection combinations, a representative case (R78) is presented in Fig. 4 to show the injection sites (Figs. 4A1 – A4) and the location of double-labeled neurons in a rostro-caudal series of TMN sections (Fig. 4B). Double-labeled cells (asterisks) were observed in ventrolateral (Fig. 4B, sections 2 and 3), ventral (Fig. 4B, section 4), and caudal (Fig. 4B, section 5) TMN. They were mainly located ipsilateral to the LC injection site. Representative examples of TMN neurons (WG + FG) with dual projection to DR and LC were shown at ventral (R95, Fig. 5A; R107, Fig. 5C), ventrolateral (R95, Fig. 5D), and caudal (R78, Fig. 5E) subdivisions. The size of labeled cells was in the range of 20 –25 Am and the morphology was bipolar- (Figs. 5A and C) or spherical-shaped (Figs. 5D and E). A large number of single-labeled neurons (FG or WG) were also observed in the vicinity of double-labeled neurons (Figs. 5A, B, and D).

71

R 95

3.2. Collateral distribution of TMN axonal fibers to DR and LC

Cases

TMN neurons were observed at dorsomedial (R123, Fig. 3A), ventrolateral (R93, Fig. 3B; R128, Fig. 3C), and ventral (R106, Fig. 3D) subdivisions. The size of labeled neurons was in the range of 20 – 25 Am and the morphology was bipolar- (Figs. 3A and D) or multipolarshaped (Fig. 3B). The density of WG granules within labeled cells was either sparse (Fig. 3B, WG) or dense (Fig. 3C, upper inset; WG); in the latter case, it was often difficult to determine whether the cells were single- (WG) or double-labeled (ADA + WG). Representative examples of double-labeled (ADA + WG) TMN neurons following injections of WG into each LC subdivision were shown in Figs. 3E –H. LC-projecting TMN cells were observed at ventral (R96, Fig. 3E; R89, Fig. 3G), ventrolateral (R98, Fig. 3F), and caudal (R98, Fig. 3H) subdivisions. The size of labeled neurons was in the range of 20– 25 Am and the morphology was either oval (Fig. 3E) or multipolar-shaped (Fig. 3F). Single-labeled cells (WG or ADA) were also observed in the vicinity of double-labeled cells (Figs. 3E –H).

Table 2 The proportion of double-labeled neurons over the LC- or DR-projecting cells at each subdivision of the tuberomammillary nucleus (TMN) following paired injections of WG into the LC and FG into the DR

H.S. Lee et al. / Brain Research 1043 (2005) 65 – 75

72

H.S. Lee et al. / Brain Research 1043 (2005) 65 – 75

Fig. 4. The location of double-labeled neurons was illustrated in a rostro-caudal series of TMN sections (B1 – B5) following the combined injection (R78) of FG into middle DR (A1 – A2) and WG into the principal LC (A3 – A4). A single injection into the DR or the LC produced dense labeling at the center of the injection site (A1 or A3) and dispersed into adjacent sections (A2 or A4). Each asterisk (B2 – 5) represents one double-labeled TMN neuron, which has axon collaterals projecting to both DR and LC. Ipsi-/contra- indicate TMN ipsilateral/contralateral to the LC injection site, respectively. Arc, arcuate n.; LM, lateral mammillary n.; ML, lateral portion of medial mammillary n.; MRe, mammillary recess of 3rd ventricle; TMc, caudal TMN; TMdm, dorsomedial TMN; TMv, ventral TMN; TMvl, ventrolateral TMN.

Six injection cases (R95, R99, R78, R83, R107, and R134) are presented in detail with numerical analysis to assess the proportion of double-labeled cells over the total number of LC- or DR-projecting cells (Table 2). The labeling patterns obtained in these cases are representative of those seen in remaining cases that have not been mentioned. For rostral LC/middle DR injection combinations (R95 and R99), the proportion of double-labeled cells was in the range of 6.7 –15.8% of total LC- or DR-projecting neurons (Table 2). For principal LC/middle DR injection combinations (R78 and R83), the proportion of double-labeled cells was 11.1 – 28.6% of total LC- or DR-projecting neurons, whereas the proportion was 7.7 –26.7% for caudal pole LC/middle DR cases (R107 and R134).

4. Discussion Several reports have described connections between vigilance-related monoaminergic nuclei in the brain, i.e.,

the tuberomammillary nucleus, the dorsal raphe, and the locus coeruleus [7,13,18,22]. Specific details about organizational features of these pathways, however, have been lacking. The present study demonstrated that there exists a robust projection from the TMN to the DR or the LC. Following rostral DR injections, retrogradely labeled cells were located mainly in dorsomedial and ventrolateral TMN subdivisions (Table 1). With midline injections at middle DR levels, labeled cells were prominent in the ventrolateral TMN, whereas cells were located in ventral and caudal TMN subdivisions following lateral wing injections. When injections were made in caudal DR, only a few cells were observed along the rostro-caudal extent of the TMN. On the other hand, following rostral LC injections, labeled neurons were observed mainly in ventrolateral and ventral TMN subdivisions (Table 1). For principal LC injections, labeled cells were observed mostly in ventrolateral, ventral, and caudal TMN, whereas only a few cells were located along the rostro-caudal extent of the TMN following injections at caudal pole of LC.

H.S. Lee et al. / Brain Research 1043 (2005) 65 – 75

73

Fig. 5. Representative examples of single- (FG or WG) or double-labeled (WG + FG) neurons at ventral (A, R95; B – C, R107), ventrolateral (D, R95), and caudal (E, R78) tuberomammillary nucleus (TMN), following the combined injection of fluorogold (FG) into the middle DR and WGA-apo-HRP-gold (WG) into each subdivision of the LC (For LC subdivisions, refer to Table 2). Insets represent higher magnification views of single- or double-labeled TMN cells indicated by arrowheads or asterisks at lower magnification. cp, cerebral peduncle; LM, lateral mammillary n.; MRe, mammillary recess of 3rd ventricle; TMc, caudal TMN; TMv, ventral TMN; TMvl, ventrolateral TMN. Scale bars = 25 Am.

Previous reports suggest that presumed serotonergic DR neurons play different roles in behavioral state control and that there is a functional differentiation within each DR subdivision [17,30,36,39,47,49,51]. The rostral DR is unique in that its neurons exhibit the smallest diurnal variation of c-Fos expression in the Mongolian gerbil [14]. Although the present study indicated that dorsomedial/ ventrolateral TMN project to the rostral DR, this DR subdivision, in turn, has few connections with the LC in the rat [16]. The rostral DR, however, provides extensive projections to the caudate– putamen and has a reciprocal connection with the dorsomedial part of the substantia nigra. Taken together, these observations suggest that sleep/ arousal-related information from the dorsomedial/ventrolateral TMN might influence activity in the rostral DR, which in turn plays an integrative role in state-dependent coordination of motor control [12,31]. Functional differences between midline and lateral wing subdivisions at middle DR levels have not been fully characterized. However, the lateral wing subdivision of DR is known to receive direct retinal afferents and sends projections to various subcortical, visual structures including lateral geniculate nucleus and superior colliculus [6,8,9,15,17,29,51,52]. The retina-lateral wing projection has been suggested to be involved in the temporal encoding

of light stimulation and the photic regulation of arousal and circadian functions [8]. The lateral wing also has major projections to subcortical somatosensory structures [17] as well as mid- to caudal LC [16]. The present study indicated that the lateral wing receives projections from ventral and caudal TMN. Thus, these TMN subdivisions might provide arousal-related information to subcortical, visual, or somatosensory structures through the lateral wing of the DR or via lateral wing-LC connections [5,16]. The TMN, however, has only a few connections with the caudal DR (Table 1). Previous work indicated that the caudal DR has a robust projection to the rostral LC in the rat [16]. In the Mongolian gerbil, the caudal ventral DR exhibits highly significant, diurnal variation of c-Fos expression; a possibility was suggested that the caudal DR might be regulated by light-activated afferents from the lateral wing via intranuclear interconnections, and thus be involved indirectly in the regulation of arousal and circadian function [14]. Furthermore, it was reported that corticotropin-releasing factor rapidly increases firing rates of a sub-population of serotonergic neurons in the ventral portion of the caudal DR in rats and that stress associated with isolation or daily restraint potentiates these responses [10,24]. Although the projection from the TMN to the caudal DR is meager, the TMN might still have some effect on the caudal DR via

74

H.S. Lee et al. / Brain Research 1043 (2005) 65 – 75

intranuclear interconnections within the DR [8,9,14] and subsequently affects output of the rostral LC [16], thus influencing sleep and arousal states of the animal. The present retrograde study is in general agreement with the results of previous anatomical studies [7,13,22]. Histamine-like immunoreactive fibers have been observed within the DR and the LC of the rabbit [13]. Likewise, a large number of very fine, short, and varicose histaminepositive fibers and terminal-like dots have been detected in the LC, LC alpha, and peri-LC alpha of cat [22]. The TMN, in return, receives noradrenergic and serotonergic inputs from the brainstem monoaminergic nuclei [7]. A biochemical basis for the functional significance of these projections has been suggested. It has been posited that histamine modulates the firing of serotonergic DR neurons through H1 receptors which are coupled to phosphoinositide hydrolysis [3,23]. It has also been suggested that histamine may exert an inhibitory effect on serotonergic DR neurons via H2-receptor subtype [19,20,21,50] or from an action directly or indirectly at the GABA receptor complex [20]. An electrophysiological study indicated that complete suppression of serotonergic DR unit activity during REM sleep might be caused by disfacilitation due to cessation of norepinephrine- or histamine-containing neuronal discharge [35]. The present study also indicated that a subset of ventrolateral, ventral, and caudal TMN neurons provide dual projections to DR and LC (Table 2). Although numerical analysis showed that the proportion of cells projecting to both LC and DR from TMN was small (6.7 – 15.8% of total LC- or DR-projecting cells for rostral LC/ middle DR injection combinations and 7.7– 28.6% of total LC- or DR-projecting cells for principal or caudal pole LC/ middle DR injection combinations), it should be indicated that only a fraction of TMN cells with such axon collaterals might be represented in our data. The DR is an extensive structure with a rostro-caudal dimension of approximately 2400 – 2600 Am [32]. An iontophoretic injection of FG at a middle DR level was not adequate to cover the entire extent of the DR. Furthermore, WG injection into subdivisions of the LC was confined within a narrow mediolateral and rostro-caudal dimension (150 – 350 Am). Nevertheless, the dual projections from the TMN to the DR and the LC might provide an anatomical substrate for transitions between various sleep and arousal states of the animal, considering the fact that the TMN plays a crucial role in the inducement of NREM sleep, whereas the DR and the LC have been involved in the maintenance of REM sleep [26,37,38,45]. The TMN neurons with such axon collaterals projecting to both DR and LC can further influence, simultaneously, the output of the brainstem monoaminergic nuclei. Based on our study, it can be concluded that there is a robust projection from the TMN to the DR or to the LC and that a portion of individual TMN cells communicate with both nuclei via axon collaterals. Although the specific mechanism has not been elucidated, it appears likely that

TMN neurons with such axon collaterals might provide a simultaneous, possibly more efficient, way of controlling the brainstem monoaminergic nuclei, thus influencing various sleep and arousal states of the animal. Acknowledgments The authors would like to thank Dr. W.C. Song for technical assistance. This work was supported by a KOSEF grant R04-2001-000-00010-0 to HSL and NIH grants NS32461 and NIDA017960 to BDW. References [1] G. Aston-Jones, F.E. Bloom, Activity of norepinephrine-containing locus coeruleus neurons in behaving rats anticipates fluctuation in the sleep – waking cycle, J. Neurosci. 1 (1981) 876 – 886. [2] G. Aston-Jones, F.E. Bloom, Norepinephrine-containing locus coeruleus neurons in behaving rats exhibit pronounced responses to nonnoxious environmental stimuli, J. Neurosci. 1 (1981) 887 – 900. [3] A. Barbara, J. Aceves, J.A. Arias-Montano, Histamine H1 receptors in rat dorsal raphe nucleus: pharmacological characterisation and linking to increased neuronal activity, Brain Res. 954 (2002) 247 – 255. [4] A.I. Basbaum, D. Menetrey, WGA-apo-HRP-gold: a new retrograde tracer for light- and electron-microscopic single- and double-label studies, J. Comp. Neurol. 261 (1987) 306 – 318. [5] C.W. Berridge, B.D. Waterhouse, The locus coeruleus – noradrenergic system: modulation of behavioral state and state-dependent cognitive processes, Brain Res. Rev. 42 (2003) 33 – 84. [6] J.M. Cedarbaum, G.K. Aghajanian, Afferent projections to the rat locus coeruleus as determined by a retrograde tracing technique, J. Comp. Neurol. 178 (1978) 1 – 16. [7] H. Ericson, A. Blomqvist, C. Kohler, Brainstem afferents to the tuberomammillary nucleus in the rat brain with special reference to monoaminergic innervation, J. Comp. Neurol. 281 (1989) 169 – 192. [8] K.V. Fite, S. Janusonis, Retinal projection to the dorsal raphe nucleus in the Chilean degus (Octodon degus), Brain Res. 895 (2001) 139 – 145. [9] K.V. Fite, S. Janusonis, W. Foote, L. Bengston, Retinal afferents to the dorsal raphe nucleus in rats and Mongolian gerbils, J. Comp. Neurol. 414 (1999) 469 – 484. [10] R.E. Grahn, M.J. Will, S.E. Hammack, S. Maswood, M.B. McQueen, L.R. Watkins, S.F. Maier, Activation of serotonin-immunoreactive cells in the dorsal raphe nucleus in rats exposed to an uncontrollable stressor, Brain Res. 826 (1999) 35 – 43. [11] J.A. Hobson, R.W. McCarley, P.W. Wyzinski, Sleep cycle oscillation: reciprocal discharge by two brainstem neuronal groups, Science 189 (1975) 55 – 58. [12] H. Imai, D.A. Steindler, S.T. Kitai, The organization of divergent axonal projections from the midbrain raphe nuclei in the rat, J. Comp. Neurol. 243 (1986) 363 – 380. [13] M. Iwase, I. Homma, S. Shioda, Y. Nakai, Histamine immunoreactive neurons in the brain stem of the rabbit, Brain Res. Bull. 32 (1993) 267 – 272. [14] S. Janusonis, K.V. Fite, Diurnal variation of c-Fos expression in subdivisions of the dorsal raphe nucleus of the Mongolian gerbil (Meriones unguiculatus), J. Comp. Neurol. 440 (2001) 31 – 42. [15] S. Janusonis, K.V. Fite, W. Foote, Topographic organization of serotonergic dorsal raphe neurons projecting to the superior colliculus in the Mongolian gerbil (Meriones unguiculatus), J. Comp. Neurol. 413 (1999) 342 – 355. [16] M.A. Kim, H.S. Lee, B.Y. Lee, B.D. Waterhouse, Reciprocal connections between sub-divisions of the dorsal raphe and the nuclear core of the locus coeruleus in the rat, Brain Res. 1026 (2004) 56 – 67.

H.S. Lee et al. / Brain Research 1043 (2005) 65 – 75 [17] M.L. Kirifides, K.L. Simpson, R.C.S. Lin, B.D. Waterhouse, Topographic organization and neurochemical identity of dorsal raphe neurons that project to the trigeminal somatosensory pathway in rat, J. Comp. Neurol. 435 (2001) 325 – 340. [18] E.M. Ko, I.V. Estabrooke, M. McCarthy, T.E. Scammell, Wake-related activity of tuberomammillary neurons in rats, Brain Res. 992 (2003) 220 – 226. [19] J.M. Lakoski, G.K. Aghajanian, Effects of histamine, H1- and H2receptor antagonists on the activity of serotonergic neurons in the dorsal raphe nucleus, J. Pharmacol. Exp. Ther. 227 (1983) 517 – 523. [20] J.M. Lakoski, G.K. Aghajanian, D.W. Gallager, Interaction of histamine H2-receptor antagonists with GABA and benzodiazepine binding sites in the CNS, Eur. J. Pharmacol. 88 (1983) 241 – 245. [21] J.M. Lakoski, D.W. Gallager, G.K. Aghajanian, Histamine-induced depression of serotoninergic dorsal raphe neurons: antagonism by cimetidine, a reevaluation, Eur. J. Pharmacol. 103 (1984) 153 – 156. [22] J.S. Lin, Y. Hou, K. Sakai, M. Jouvet, Histaminergic descending inputs to the mesopontine tegmentum and their role in the control of cortical activation and wakefulness in the cat, J. Neurosci. 16 (1996) 1523 – 1537. [23] J.S. Lin, K. Sakai, M. Jouvet, Evidence for histaminergic arousal mechanisms in the hypothalamus of cat, Neuropharmacology 27 (1988) 111 – 122. [24] C.A. Lowry, J.E. Rodda, S.L. Lightman, C.D. Ingram, Corticotropinreleasing factor increases in vitro firing rates of serotonergic neurons in the rat dorsal raphe nucleus: evidence for activation of a topographically organized mesolimbocortical serotonergic system, J. Neurosci. 20 (2000) 7728 – 7736. [25] I.J. L-Smith, P. Pilowsky, J.B. Mison, Retrograde tracers for light and electron microscopy, in: J.P. Bolam (Ed.), Experimental Neuroanatomy, Oxford Univ. Press, Oxford, 1992, pp. 31 – 59. [26] J. Lu, A.A. Bjorkum, M. Xu, S.E. Gaus, P.J. Shiromani, C.B. Saper, Selective activation of the extended ventrolateral preoptic nucleus during rapid eye movement sleep, J. Neurosci. 22 (2002) 4568 – 4576. [27] M. Mackiewicz, E.V. Nikonova, C.C. Bell, R.J. Galante, L. Zhang, J.D. Geiger, A.I. Pack, Activity of adenosine deaminase in the sleep regulatory areas of the rat CNS, Brain Res. Mol. Brain Res. 80 (2000) 252 – 255. [28] D.J. McGinty, R.M. Harper, Dorsal raphe neurons: depression of firing during sleep in cats, Brain Res. 101 (1976) 569 – 575. [29] R.R. Mize, L.H. Horner, Origin, distribution, and morphology of serotonergic afferents to the cat superior colliculus: a light and electron microscope immunocytochemistry study, Exp. Brain Res. 75 (1989) 83 – 98. [30] E. O’Hearn, M.E. Molliver, Organization of raphe-cortical projections in rat: a quantitative retrograde study, Brain Res. Bull. 13 (1984) 709 – 726. [31] D.A. Pasquier, T.L. Kemper, W.B. Forbes, P.J. Morgane, Dorsal raphe, substantia nigra and locus coeruleus: interconnections with each other and the neostriatum, Brain Res. Bull. 2 (1977) 323 – 339. [32] G. Paxinos, C. Watson (Eds.), The Rat Brain in Stereotaxic Coordinates, Academic Press, New York, 1998. [33] J. Pearson, G. Halliday, N. Sakamoto, J.P. Michel, Catecholaminergic neurons, in: G. Paxinos (Ed.), The Human Nervous System, Academic Press, New York, 1990, pp. 1023 – 1049. [34] V. Ramesh, M.M. Thakkar, R.E. Strecker, R. Basheer, R.W. McCarley, Wakefulness-inducing effects of histamine in the basal forebrain of freely moving rats, Behav. Brain Res. 152 (2004) 271 – 278. [35] K. Sakai, S. Crochet, Serotonergic dorsal raphe neurons cease firing by disfacilitation during paradoxical sleep, NeuroReport 11 (2000) 3237 – 3241. [36] K. Sakai, S. Crochet, Differentiation of presumed serotonergic dorsal

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44] [45]

[46]

[47]

[48]

[49]

[50]

[51]

[52]

[53]

75

raphe neurons in relation to behavior and wake – sleep states, Neuroscience 104 (2001) 1141 – 1155. J.E. Sherin, P.J. Shiromani, R.W. McCarley, C.B. Saper, Activation of ventrolateral preoptic neurons during sleep, Science 271 (1996) 216 – 219. J.E. Sherin, J.K. Elmquist, F. Torrealba, C.B. Saper, Innervation of histaminergic tuberomammillary neurons by GABAergic and galaninergic neurons in the ventrolateral preoptic nucleus of the rat, J. Neurosci. 18 (1998) 4705 – 4721. K.L. Simpson, T.M. Fisher, B.D. Waterhouse, R.C.S. Lin, Projection patterns from the raphe nuclear complex to the ventricular system in the rat, J. Comp. Neurol. 399 (1998) 61 – 72. W.A. Staines, T. Yamamoto, P.E. Daddona, J.I. Nagl, Neuronal colocalization of adenosine deaminase, monoamine oxidase, galanin and 5-hydroxytryptophan uptake in the tuberomammillary nucleus of the rat, Brain Res. Bull. 17 (1986) 351 – 365. T.L. Steininger, M.N. Alam, R. Szymusiak, D. McGinty, Statedependent discharge of neurons in the rat posterior hypothalamus, Abstr.-Soc. Neurosci. 22 (1996) 689. T.L. Steininger, M.N. Alam, H. Gong, R. Szymusiak, D. McGinty, Sleep – waking discharge of neurons in the posterior lateral hypothalamus of the albino rat, Brain Res. 840 (1999) 138 – 147. T.L. Steininger, H. Gong, D. McGinty, R. Szymusiak, Subregional organization of preoptic area/anterior hypothalamic projections to arousal-related monoaminergic cell groups, J. Comp. Neurol. 429 (2001) 638 – 653. L.A. Sternberger (Ed.), Immunocytochemistry, Wiley, New York, 1986. R. Szymusiak, N. Alam, T.L. Steininger, D. McGinty, Sleep – waking discharge patterns of ventrolateral preoptic/anterior hypothalamic neurons in rats, Brain Res. 803 (1998) 178 – 188. M.E. Trulson, B.L. Jacobs, Raphe unit activity in freely moving cats: correlation with level of behavioral arousal, Brain Res. 163 (1979) 135 – 150. E.J. Van Bockstaele, A. Biswas, V.M. Pickel, Topography of serotonin neurons in the dorsal raphe nucleus that send axon collaterals to the rat prefrontal cortex and nucleus accumbens, Brain Res. 624 (1993) 188 – 198. G. Vanni-Mercier, K. Sakai, M. Jouvet, Waking-state specific neurons in the caudal hypothalamus of the cat, C. R. Acad. Sci., Ser. III 298 (1984) 195 – 200. M.J. Villar, M.L. Vitale, T. Ho¨kfelt, A.A. Verhofstad, Dorsal raphe serotonergic branching neurons projecting both to the lateral geniculate body and superior colliculus: a combined retrograde tracing immunohistochemical study in the rat, J. Comp. Neurol. 277 (1988) 126 – 140. M.L. Vizuete, E. Traiffort, M.L. Bouthenet, M. Ruat, E. Souil, J. Tardivel-Lacombe, J.C. Schwartz, Detailed mapping of the histamine H2 receptor and its gene transcripts in guinea-pig brain, Neuroscience 80 (1997) 321 – 343. B.D. Waterhouse, G.A. Mihailoff, J.C. Baack, D.J. Woodward, Topographic distribution of dorsal and median raphe neurons projecting to motor, sensorimotor, and visual cortical areas in the rat, J. Comp. Neurol. 249 (1986) 460 – 476. B.D. Waterhouse, B. Border, L. Wahl, G.A. Mihailoff, Topographic organization of rat locus coeruleus and dorsal raphe nuclei: distribution of cells projecting to visual system structures, J. Comp. Neurol. 336 (1993) 345 – 361. F.G. Wouterlood, R.P.A. Gaykema, Innervation of histaminergic neurons in the posterior hypothalamic region by medial preoptic neurons: an anterograde tracing with PHA-L combined with immunocytochemistry of histidine decarboxylase in the rat, Brain Res. 455 (1988) 170 – 176.