THE JOURNAL OF COMPARATIVE NEUROLOGY 355211-220 (1995)

5-HTIA Receptor Localization on the Axon Hillock of Cervical Spinal Motoneurons in Primates N.M. KHECK, P.J. GANNON, AND E.C. AZMITIA Department of Biology, (N.M.K., E.C.A.), New York University, New York, NY 10003 (N.M.K., E.C.A.); and Department of Otolaryngology, Mount Sinai School of Medicine, New York, NY 10029 (P.J.G.)

ABSTRACT Serotonin (5-HT) has direct and specific effects on the activity of spinal cord motoneurons. The 5-HTlA receptor has been shown to mediate motoneuron responses in spinal reflex pathways using the highly selective 5-HTIAreceptor agonist 8-OH-DPAT. We have developed an antipeptide antibody that recognizes a specific region (the second external loop) of the ~ - H T receptor. ~A This 5 - H T 1 receptor ~ antibody labels populations of neurons and glia in the primate cervical spinal cord. The highest receptor density is present in the superficial lamina of the dorsal horn, around the central canal, and on the axon hillock of large ventral horn motoneurons. The cellular labeling pattern on motoneurons shows a single, densely stained, tapering process emanating from the perikaryon. A more diffuse label is also present throughout the soma. Dendritic labeling was not apparent. These results suggest that post-synaptic 5 - H T 1receptors ~ may be involved in modulating spinal motoneuron activity at the key site of action potential initiation, the axon hillock, o 1995 WiIey-Liss, Inc. Indexing terms: serotonin, receptor antibody, autoreceptor,Macaca, immunocytochemistry

Descending serotonergic nerve fibers influence somatosensory, motor, and autonomic functions within the spinal cord (Andersen, 1984).Electrophysiological studies of brainstem and spinal motoneurons have demonstrated diverse and complex responses to serotonin (5-HT) that are dependent on the activation of specific postsynaptic receptors (Jackson and White, 1990; Rasmussen and Aghajanian, 1990; Larkman and Kelly, 1991). Cells in various sensory and motoneuron pools can be inhibited or excited by direct iontophoretic application of serotonin (Takahashi and Berger, 1990;Wang and Dun, 1990) or by electrical stimulation of the descending brainstem raphe nuclei (Fung and Barnes, 1989). Numerous serotonin receptor subtypes have been described in the spinal cord (Fone et al., 1991; Marlier et al., 1991a; Pubols et al., 1992; Thor et al., 1993). Radioligand binding of specific receptor agonists has allowed for regional localization and quantification of 5-HT receptor subtypes. A highly selective 5-HTIAreceptor agonist, 8-OHDPAT, an aminotetralin (Arvidsson et al., 1981),has been used extensively to localize 5 - H T 1receptor ~ binding sites in the rat brain and spinal cord. The dorsal horn and region around the central canal show the highest density of receptor binding sites (Marlier et al., 1991b; Thor et al., 1993). Undetectable levels of [3H]8-OH-DPATbinding are reported for the ventral horn motor nuclei, except for the O 1995 WILEY-LISS. INC.

dorsolateral nucleus of the pudendal nerve in the lumbar enlargement (Thor et al., 1993). In a previous study of 5-HT1 binding sites in the rat medulla, Thor et al. (1992) also reported low levels of [3H]8-OH-DPATin the facial and hypoglossal somatic motor nuclei. More recently, Wright et al. (1995) reported low levels of 5-HT1*mRNA in all cranial nerve motor nuclei in the rat brain. The 5-HTlA receptor may be involved in facilitation of motoneuron excitability, since 8-OH-DPAT can produce the 5-HT behavioral syndrome, which is thought to result from a direct serotonergic faciIitation of excitatory inputs to brainstem and spinal motoneurons (Jacobs, 1976). Several studies have recently demonstrated the involvement of 5-HTlA receptors in regulation of motor activity at the presynaptic level (by systemic and intracellular injections of 8-OH-DPAT into the dorsal and median raphe nuclei; Hillegaart et al., 1989; Hillegaart, 1990). At the postsynaptic receptor level, however, reports on the effects of ~ - H T ~ A agonists on motoneuron output are inconsistent. In wholecell recordings in neonatal rat spinal cord slices, administration of 8-OH-DPAT (in nanomolar concentrations) has Accepted September 12,1994. Address reprint requests to Nancy M. Kheck, Department of Biology, New York University, 100 Washington Square East, 1009 Main Bldg, New York, NY 10003.

212

N.M. KHECK ET AL.

been shown to mimic the effects of serotonin by directly exciting motoneurons (Takahashi and Berger, 1990). Conversely, other studies have reported that local application of 8-OH-DPAT in isolated neonatal rat spinal cord causes a hyperpolarizing response in motoneurons (Wang and Dun, 1990) or has no effect on motoneuron excitability (Jackson and White, 1990). Although the physiology and pharmacology of 5-HT and the 5-HT1.4receptor-mediated responses of sensory elements (Eide et al., 1990) and motoneurons (Jackson and White, 1990; Rasmussen and Aghajanian, 1990) have been well documented, the (cellular) localization of the 5-HT1* agonist binding site in brainstem and spinal cord motoneurons remains unknown. We have developed a 5-HT1* receptor antipeptide antibody that specifically recognizes amino acid sequence 170186 (Azmitia et al., 1992). This region on the second extracellular loop has no significant homology with any other published receptor or mammalian protein and is structurally linked to the agonist binding site. We are currently using this antibody to characterize the cellular localization of the ~ - H T receptor ~A throughout the rat, cat, and monkey brain and spinal cord. This paper presents our findings in the adult primate cervical spinal cord.

MATERIALS AND METHODS Macaque spinal cord tissue used in this study was obtained from Dr. William J. Doyle (Department of Otolaryngology, Children's Hospital of Pittsburgh). Monkeys were pretreated with pargyline (100 mg/kg, i.p.1 and then placed under deep barbiturate anesthesia (sodium pentobarbitol, 30-50 mgikg, i.p.) prior to transcardiac perfusion with saline (100 mlO.9% NaC1,37"C), followed by 3 liters of ice cold 4% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS; pH 7.4) with 0.1% MgS04, over the course of 20 minutes. Spinal cords of rhesus monkeys (n = 2; Macaca mulatta; adult females; approximately 3-4 kg) and cynomologus monkeys (n = 3; Macaca fascicularis; young adult females; 2-3 kg) were removed, blocked, and postfixed in 4% paraformaldehyde at 4°C for 8-12 hours. For cryoprotection, tissue blocks were immersed in 20% sucrose in 0.1 M PBS until they were infiltrated (24-36 hours). Spinal cord sections (from regions of the upper and midcervical enlargement) were cut (40 pm) on a frozen stage microtome for light microscopy (LM) and collected in 0.1 M PBS. Following two washes in PBS, and a 1 hour incubation in 1% normal goat serum (Amel), free-floating sections were incubated 48-72 hours in primary antibody directed against the 5 - H T 1receptor ~ at a dilution of 1:8,000 (empirically determined to give optimal signal with minimal background) in 0.1 M PBS containing 1%blocking serum and 0.3% Triton-X (Sigma). This polyclonal antibody was generated using a synthetic peptide with the amino acid sequence of a region on the second extracellular loop of the cloned rat 5-HTl* receptor protein (Albert et al., 1990). The peptide corresponding to amino acid sequence 170-186 was coupled to keyhole limpet hemocyanin (Sigma) via maleimidobenzoyl-N-hydrosuccinimide (Pierce Chem. Co.) as described by Azmitia et al. (1992) and injected into adult female New Zealand white rabbits. The antiserum was affinity purified using columns of immobilized recombinant protein A (Pierce Chem. Co.), and IgG concentrations were determined to be in the range of 0.2-0.8 mg/ml. The antiserum (at dilution of 1:8,000) labeled a major band of approximately 69 kd in immunoblot /gel electrophoresis of

solubilized protein from the hippocampus and brainstem of neonate rat pups. The specificity of the antibody for the peptide was established in serum dilution titer curves using 1251-labeledIgG to quantify binding affinities. At (5-HTlA antibody) dilution up to 1:10,000, radioimmunoassay binding of the native peptide (aa sequence 170-186) was obtained, whereas preimmune serum showed no affinity (Azmitia et al., 1992). Control alternate sections were incubated in primary antibodies against 5-HT (Incstar) and glial fibrillary acidic protein (GFAP; Sigma), at dilutions of 1:11,000 and 1:18,000, respectively. Sections were washed in 0.1 M PBS before and after incubation in 1) biotinylated goat antirabbit IgG (Vector) for 1 hour, followed by 2) avidin-biotin complex (Vector) for 30 minutes, and then 3) reacted with 0.05%diaminobenzidine (DAB) with 0.003% hydrogen peroxide at 4°C for 10-15 minutes. Following washes in dilute PBS (0.05 M) and mounting onto gelatinized slides, spinal cord sections were counterstained with methyl green, dehydrated, cleared in xylene, and coverslipped with DPX mountant (DBH). Slides were viewed on a Leitz microscope and photographed with Kodak T-Max 100 print film. Control sections were incubated in 1) 1% BSA alone, 2) preimmune serum (at dilutions of 1:1,000 to 1:6,000), and 3) a cocktail of primary antiserum and antigen-peptide. The antigen-peptide was serially diluted from concentrations of 0.75 mgiml to 0.075 pg/ml (comparable to the IgG concentration in our working dilutions) and preadsorbed with a nonvarying concentration of primary antiserum (diluted 1:6,000) for 4-6 hours under constant agitation, prior to incubation of tissue slices. Sections were then processed as described above. In each case there was no specific label detectable above background levels.

RESULTS We examined 5-HT1* receptor immunoreactivity (IR) in the upper to midcervical spinal cord (approximately C2 to C5) in two species of macaques. The pattern of labeling observed in both species and within cervical spinal levels was consistent. The gross anatomical distribution of ~ - H T ~ A receptor immunoreactivity conforms with the pattern previously described for the distribution of 5-HT fiber terminals. The highest density of 5-HT1*receptor-IR was found in the superficial layers of the dorsal horns (laminae 1-11), with a particularly strong signal in the nerve fibers of lamina I1 (Figs. l a , 3a). In laminae I11 and IV, there was fiber labeling along the medial contours of the dorsal horn gray matter. A few cells in nucleus proprius (laminae IV-V) and in the intermediate gray (lamina VII) presented a distinct pattern of cellular labeling that is also characteristic of motoneurons in the ventral horn (Fig. lb). The dorsal commissural gray matter and the region around the central canal (lamina X) also show pronounced labeling of nerve fibers, ependymal cells, and occasionally some very densely stained cells in the ventrolateral commiss u r d gray area. We observed intense, punctate label on the soma and processes of these spindle-shaped cells. Their processes were often varicose and wrapped around large blood vessels of the ventral gray matter (Fig. 2b). The distribution of 5-HTlA receptor immunoreactivity in the dorsal horn and intermediate gray area assumes the characteristic V-shaped pattern described in the localization studies of 5-HTlAreceptor-ligand binding sites in rat spinal cord (Thor et al., 1993).

Fig. 1. Light microscopic photomicrographs of 40 km coronal section of macaque cervical spinal cord. Sections were processed for immunocytochemical demonstration of serotonin (5-HT),, receptor localization and counterstained with methyl green. a: Low-power micrograph showing localization of 5-HTIA receptor immunolabel within lamina I and I1 of the dorsal horn and in laminax. The two areas

marked by boxes correspond to plates a and b in Figure 2. b Medium-power micrograph of ventral horn showing the specific and dense localization of 5-HTIAreceptor immunoreactivity on the axon hillock of the majority of motoneurons. c (overleaf): High-power micrograph of labeled motoneurons. The arrows indicate 5-HT1* receptorimmunoreactive (IR) glial cells.

214

N.M. KHECK ET AL.

Figure 1. Continued

We observed in the ventral horn, a pattern of 5-HT1* receptor immunoreactivity not previously described: The perikarya of motoneurons show a diffuse gradient of label, which dramatically intensifies at the point where a single tapering process emerges. The nucleus was devoid of immunoreactivity and there was no dendritic labeling. We believe these processes are the axon hillocks of spinal motoneurons. The dense accumulation of chromagen in the proximal region of the hillock gradually diminished distal to the perikaryon (Fig. 2a). Linear measurements of the longitudinal axis of the axon hillock (proximally from its wide somatic origin and distally to where a gradual transition into a straight process of constant diameter indicated the axon initial segment proper) measured 29.4 +- 5.8 pm in a random sampling (n = 69) of motoneurons. In many of the motoneurons sampled, the continuation of the axon initial segment was also labeled, but these were not included in our measurements. Occasionally, within specific ventrolateral motoneuron pools, these processes appeared oriented in the same direction and their labeled axons could be traced to their exit of the gray matter within the ventral horn rootlets. Since the pattern of 5-HTlAreceptor-IR in the ventral horn is so localized (a single process emanating from each motoneuron), a counterstain (methyl green, cresyl violet) is essential to identify the placement of the cell bodies. Glial cells were also labeled by this 5 - H T 1 ~receptor antibody. We compared this label to adjacent spinal cord sections reacted with antibodies to GFAP, an astroglialreceptor-IR specific protein, in order to distinguish 5-HTLA glial cells from labeled neurons. Otherwise, characteristics

such as size, morphology, and location were used to assess the nature of labeled glia. Labeled astrocytes were interspersed between the fiber tracts of the dorsal and dorsolateral funiculi (Fig. 3b) and in the ventral white matter as well. Labeled glia (presumptive oligodendrocytes) were present throughout the gray matter, often in close association with motoneurons.

DISCUSSION The mammalian spinal cord receives a dense projection of 5-HT-containing fibers from cell bodies in the brainstem raphe nuclei (Dahlstrom and Fuxe, 1965). Descending serotonergic projections originate in the three most caudal raphe nuclei, designated as the raphe magnus, pallidus, and obscurus, and the medial reticular formation of the medulla (Bowker et al., 1982). The bulbospinal pathways have been shown to regulate a number of spinal activities, including

Fig. 2. High-magnification photomicrographs (the regions marked by boxes in Fig. l a correspond to Fig 2a,b. a: High-power micrograph of immunolabeled rnotoneurons showing the dense 5 - H T 1 receptor-IR ~ localized to the axon hillock, with a faint, diffuse label present on the cell body. Arrows indicate labeled glial cells. b High-power micrograph of 5-HTIAreceptor-IR cells located ventrolateral to the central canal. Arrowheads mark cellular processes that are associated with local blood vessels (bv). These neurons, which show dense immunolabel throughout their soma and cellular processes, may be caudally located representatives of nucleus raphe obscurus (La Motte et al., 1982; Azmitia and Gannon, 1986).

5-HTIA RECEPTORS AND PRIMATE SPINAL MOTONEURONS

Figure 2

215

216

N.M. KHECK ET AL.

Figure 3

5-HT1.4 RECEPTORS AND PRIMATE SPINAL MOTONEURONS

217

composition. The unusual, intensely ~ - H T ~ A -cells I R we observed in the ventral gray commissure may represent pre-synaptic 5-HTIAautoreceptors on caudal spinal raphe neurons, particularly given their limited appearance at upper cervical levels, morphology, and proximity to large blood vessels. We have also reported this identical somatodendritic pattern of 5-HTlAreceptor-IR in the dorsal raphe nuclei of the same macaque species (Azmitia et al., 1994). Glial cells This pattern of 5-HTlA receptor-IR is comparable to the receptor The presence of 5-HTlA receptor immunoreactivity on label of raphe neurons (with a different 5 - H T 1 ~ glial cells in the spinal cord was observed in both white and antibody) demonstrated by Sotelo et al. (1990) in the rat gray matter. This 5-HTIAreceptor antibody has previously brain. [3H]8-OH-DPATbinding is heavily localized t o this been shown to colabel a subpopulation of GFAP-immunore- region in the rat spinal cord, indicating a moderate to high active glial cells in the hippocampus, septum, and cortex in concentration of 5-HTlAreceptor binding sites in lamina x the rat brain (Whitaker-Azmitia et al., 1993). Stimulation (Marlier et al., 1991b; Thor et al., 1993). Our results in the of 5-HTlAreceptors (with specific agonists) has been shown macaque cervical spinal cord replicate this pattern of ~A to change the morphology of cultured rat brain-derived glial localization and show moderate levels of ~ - H T receptor cells and to promote the release of S-loop, a calcium immunoreactivity around the central canal. binding protein with trophic effects on select neurons and Ventral horn glial cells (Whitaker-Azmitia et al., 1990). S-loop is found in developing spinal motoneurons and glia and is known to In the ventral horn of the spinal cord, 5-HT-containing promote neuronal survival in vivo (Bhattacharyya et al., nerve terminals form a dense fiber plexus around large 1992). This raises the possibility that ~ - H T receptors ~A on motoneurons (Dahlstrom and Fuxe, 1965; Ulfhake et al., glial cells may subserve atrophic function in the developing 1987). Motoneurons in the lateral aspects of the ventral and adult spinal cord. horn appear to receive a heavier input of 5-HT terminals than those situated in the medial ventral horn (Dahlstrom Dorsal horn and Fuxe, 1965). This may correlate functionally to a more The superficiallayers of the dorsal horn (Rexed’slaminae lateral placement of extensor motoneuron pools and the I and 11)receive the heaviest 5-HT innervation (Marlier et more medial placement of flexor motoneuron pools. Immual., 1991a; Wallis et al., 1993). Radioligand binding studies nocytochemical electron microscopic (EM) investigations using r3H]8-OH-DPAThave also consistently identified the confirm that 5-HT axon terminals synaptically contact the dorsal horn (lamina 11) as the region of highest concentra- somata or dendrites of ventral horn motoneurons in rat and tion of 5 - H T I ~ receptors in the rat spinal cord (Marlier et cat spinal cord (Takeuchi et al., 1983; Ulfhake et al., 19871, ~ is although many boutons do not appear to make classical al., 1991b; Thor et al.,19931, where the 5 - H T 1receptor thought to modulate afferent nociceptive information. The synaptic contacts, and there are noted species differences. pattern of antibody labeling we observed in the macaque In rabbit brainstem trigeminal motoneurons, varicose 5-HT dorsal horn concurs with the autoradiographic data re- fibers are reportedly found in the neuropil in close proximported for the rat: Lamina I1 is the region with the most ity t o the soma, the dendrites, and the proximal portions of intense concentration of 5-HT1A receptor immunore- axons (Kolta et al., 1993). Takeuchi et al. (1983) described a activity. similar pattern of 5-HT fiber innervation of trigeminal motoneurons in rodents and dogs, whereas in monkeys the Central canal somata were surrounded with basket-like structures. In The region around the central canal (lamina X) also has a this study of the macaque spinal cord, we observed such substantial 5-HT terminal innervation (Azmitia and Gan- dense plexi of 5-HT fibers surrounding large ventral horn non, 1986; La Motte, 1988). In macaques, there are spinal motoneurons. Furthermore, we have often noted a dense 5-HT cell bodies located ventral and ventrolateral to the array of 5-HT axon terminals impinging on the axon central canal (CC), which send axonal processes to the hillocks of these large motoneurons that in adjacent secintermediate gray and medial portion of the ventral horn tions label positively for 5-HTlAreceptor-IR (Kheck et al., (La Motte et al., 1982). These cells are thought to be an 1994). extension of 5-HT raphe cell groups from the caudal There are several 5-HT receptor subtypes found in the medulla (La Motte et al., 1982), specifically raphe obscurus ventral horn. Functional studies (Takahashi and Berger, (Azmitia and Gannon, 1986). Some of the processes are 1990; Fone et al., 1991; Wallis et al., 1993) involving adjacent to large blood vessels or ependymal lining of the manipulations of motoneuron excitability with receptor central canal, suggesting that 5-HT is involved in the agonists and antagonists suggest the localization of 5-HT1 control of spinal blood flow or cerebrospinal fluid (CSF) high-affinity (nanomolar) receptors in the spinal cord on or around motoneurons. In autoradiographic mapping studies using L3H18-OH-DPAT, the distribution of the 5 - H T 1 ~ receptor has generally been shown to be very high in the dorsal horn, around the central canal, and absent in the Fig. 3. 5 - H T 1 ~receptor-IR in the dorsal horn at the level of the cervical enlargement. a: Fibers in lamina I1 show a particularly dense ventral horn (where 5 - H T 1 and ~ 5-HTz receptor subtypes band of ~ - H T receptor ~A labeling. The box in the dorsolateral white appear to predominate; Marlier et al., 1991b).A limitation matter corresponds to Figure 3b. b: High-magnification (oil immersion) of these studies is that receptor agonists such as 8-OHphotomicrograph of 5-HTIAreceptor-IR glial cells (arrowheads) inter~ sites spersed between fiber tracts of the dorsolateral funiculus. These glial DPAT exhibit high affinity only for 5 - H T 1binding cells are positive for glial fibrillary acidic protein (GFAP)-IR in adjacent that are actively coupled to G proteins (Hamon et al., 1990; sections. Zifa and Fillion, 1991). A 5-HTlA receptor antibody may

motor (ventral horn), nociceptive (dorsal horn), and autonomic (intermediate/central gray) functions (Andersen, 1984). We have examined these three 5-HT terminal regions in the cervical spinal cord for ~ - H T receptor-IR, ~A because they have major 5-HT input and are proposed to have postsynaptic 5-HT1 receptors (El-Yassir et al., 1988; Marlier et al., 1991b; Pubols et al., 1992; Thor et al., 1993).

218

N.M. KHECK ET AL.

have the distinct advantage of recognizing all 5-HTIA receptor binding sites, independent of their functional binding to second messengers (El Mestikawy et al., 1990; Riad et al., 1991). The rat ~ - H T ~receptor A shows a diverse profile of pharmacological, functional, and regulatory properties, which can vary between different brain regions and cell types. Studies of the adaptive regulatory mechanisms of 5 - H T 1 receptors ~ in response to various drug treatments (e.g., antidepressants, 5 - H T 1receptor ~ agonists) have demonstrated pronounced differences between pre- and postsynaptic 5 - H T 1 receptors ~ (Hamon et al., 1990; Yocca, 1990). Radja et al. (1992) reported regional heterogeneity of postsynaptic ~ - H T ~receptors A within the hippocampus (CA1 area vs. dentate gyrus), which could not be due to intrinsic properties of the binding sites, but suggests differences in intracellular signal transduction pathways. A northern blot analysis of 5 - H T 1 ~mRNAs in rat brain detected mRNA sequences of multiple sizes (3.9, 3.6, 3.3, and 2.2 kb) in some brain regions, suggesting differential transcriptional regulation of the receptor (Albert et al., 1990). More recently, Albert et al. (1994) report that 5-HTIAreceptors may be coupled to different G proteins and second messenger molecules in neurons and glial cells. Miquel et al. (1991) used a ~ - H T receptor ~A antipeptide antibody (aa 243-2681, a riboprobe, and a specific 5-HTIA radioligand, [12511BH-8-Me0-N-PAT, to map the distribution of 5 - H T 1mRNA ~ and receptor binding sites in the rat brain. The regional distribution of the 5 - H T 1 ~mRNA generally matched that of the 5 - H T 1 ~receptor protein; however, discrepancies were noted in several brain regions where the density of the 5-HTIAreceptor protein was high, and the in situ hybridization signal was notably low or absent, perhaps, as the authors suggest, resulting from a limited sensitivity of the hybridization procedure (Miquel et al., 1991). Collectively,these data demonstrate heterogeneity of the rat 5-HTlAreceptor at various levels. The highly localized 5-HTlAreceptor-IR we observed on ventral horn motoneuron axon hillocks may represent a distinct population of 5-HTlA receptor binding subunits that do not bind to 8-OH-DPAT, or perhaps are simply present at very low levels. Electrophysiological studies involving the direct excitation of motoneurons by serotonin do, however, report that 8-OH-DPAT can facilitate the depolarization of motoneuron membrane potential beyond firing threshold via local (Takahashi and Berger, 1990) and systemic (Jackson and White, 1990; Rasmussen and Aghajanian, 1990) administration. Thus, several lines of physiological evidence establish the functional significance of 5-HTlA receptors in motoneuron excitability. We believe that the localization of ~ - H T ~ receptor-IR A to the axon hillock and initial axon segment is functionally relevant and may provide the anatomical substrate for 5 - H T 1receptor~ mediated modulation of motor output.

Axon hillock The axon hillock is the critical region on neurons where the action potential is initiated (Coombs et al., 1957; Palay et al., 1968). Furthermore, it is known to have considerably lower threshold characteristics than either the neuronal soma or the dendritic tree. Spinal cord neurons and a few other neurons, such as cortical pyramidal cells and cerebellar Purkinje cells, are unique in that the axon hillock receives a dense array of terminal boutons (Somogyi, 1977;

Peters and Harriman, 1990; Rethelyi and Lozsadi, 1990; Kolta et al., 1993). In neurons of the dorsal horn (lamina 11) of the rat spinal cord, for example, the majority of synapses on the axon hillock are symmetrical and are thought to be inhibitory in nature (Rethelyi and Lozsadi, 1990). In the cat, there are twice as many putative inhibitory as excitatory synapses on the axon hillock of ventral horn spinal motoneurons (Conradi, 1969). Axoaxonal contacts onto the axon hillock of cortical pyramidal neurons have also been shown to be primarily inhibitory in nature (Somogyi, 1977). Conversely, the parallel fiber inputs to the Golgi and basket neurons of the cerebellum are excitatory (Hamori, 1981). Collectively, this evidence suggests that synaptic contacts on the axon hillock serve to directly influence the initiation of action potentials in these neuronal cell types. Although there are few if any data available on the characteristics of the motoneuron axon hillock in Macaca fascicularis, vve have compared our data to data from studies on the cat and human spinal cord. The mean linear measure of the axon hillock in macaques, (29.4 ? 5.8 pm; n = 69) is longer than that reported for a similarly sized mammal, the domestic cat (12.0 4.5 pm; Conradi, 1969). Similarly, axon hillocks of spinal cord motoneurons in humans (Sasaki et al., 1990) are considerably longer than we report for macaques. These size differences may reflect such things as overall body size, evolutionary change, and different functional roles in species. However, differentiation of the transition between the axon hillock and initial segment is ambiguous at the light microscopic level, which may also account for some of the reported size variation. We have demonstrated for the first time that a dense population of 5-HT1A receptors is associated specifically with the axon hillock of cervical spinal cord motoneurons. We are confident that the cellular site labeled by our antibody is the axon hillock and not dendritic, based on the following evidence: 1) Without exception, only a single immunolabeled cellular process was present on all motoneuions observed and 2) the anatomical characteristics described for the axon hillock [a wide somatic origin, which rapidly tapers to a single thin process (Conradi, 196911were consistently present. The high density of receptors present at this functionally important cellular site suggests that they may be involved in the regulation of action potentials. We propose that the 5-HTlA receptors present on the motoneuron axon hillock might play a key role in serotonergic modulation of motor activity.

*

ACKNOWLEDGMENTS This work was supported in part by NIA grant 1 PO1 AG10208; Department of Biology, New York University, and Department of Otolaryngology, Mount Sinai School of Medicine, New York. The authors thank Dr. Monroe Yoder (NYU) for photographic assistance and Mr. Todd Anthony (NYLJ)for the Western analysis. We especially thank William J. Doyle, PhD, Director of Otolaryngology Research Labs, Children’s Hospital of Pittsburgh, for generously providing us with the monkey spinal cords that made this study possible. The tissues we obtained from Dr. Doyle were from animals used in a project funded by NIH grant DC 01260.

LITERATURE CITED Albert, P.R., $.-Y. Zhou, H.H.M.VanTo1, J.R. Bunzow, a n d 0 . Civelli (1990) Cloning, functional expression, and mRNA tissue distribution of the rat 5-hydroxytryptamineIAreceptor gene. J. Biol. Chem. 2655825-5832.

5-HTla RECEPTORS AND PRIMATE SPINAL MOTONEURONS Albert, P.R., P . Lembo, J. Storring, Y.F. Liu, A. Charest, C. Saucier, and H. Abdel-Baset (1994) The 5-HTIAreceptor: Signaling, desensitization, and gene transcription. Neuropsychopharmacology (in press). Andersen, E.G. (1984) The serotonin system of the spinal cord. In R.A. Davidoff (ed.): Handbook of the Spinal Cord. New York: Marcel1Dekker, pp. 241-274. Arvidsson, L.E., U. Hacksell, J.L.G. Nilsson, S. Hjorth, A. Carlsson, P. Lindberg, D. Sanchez, and H. Wikstrom (1981) 8-hydroxyv-2-(di-npropylamino)tetralin, a new centrally acting 5-hydroxytryptamine receptor agonist. J. Med. Chem. 241921-923. Azmitia, E.C., and P.J. Gannon (1986) The primate serotonergic system: A review of human and animal studies and a report on Mucucu fuscicularis. In S. Fahn, C.D. Marsden, and M. Van Woert (eds.): Neurology: Myoclonus, Vol. 43. New York Raven Press, pp. 407-468. Azmitia, E.C., I. Yu, H.M. Akbari, N. Kheck, P.M. Whitaker-hmitia, and D.R. Marshak (1992) Antipeptide antibodies against the 5-HT1A receptor. J. Chem. Neuroanat. 5:289-298. Azmitia, E.C., P.J. Gannon, N.M. Kheck, and P.M. Whitaker-Azmitia (1994) The cellular localization of the 5-HTlAreceptor in cortical, hippocampal, and brainstem neurons and glial cells in Mucaca fusciculuris. Neuropsychopharmacol. (in press). Bhattacharyya, A,, R.W. Oppenheim, D. Prevette, B.W. Moore, R. Brackenbury, and N. Ratner (1992) Sl00 is present in developing chicken neurons and schwann cells and promotes neuron survival in vivo. J. Neurobiol. 23t451-466. Bowker, R.M., K.N. Westlund, M.C. Sullivan, and J.D. Coulter (1982) Organization of the descending serotonergic projections to the spinal cord. Progr. Brain. Res. 57:239-265. Conradi, S. (1969) Observations of the ultrastructure of the axon hillock and initial segment of lumbosacral motoneurons in the cat. Acta Physiol. Scand. 332:65-84. Coombs, J.S., D.R. Curtis, and J.C. Eccles (1957) The generation of impulses in motoneurons. J. Physiol. 139:232-249. Dahlstrom, A,, and K. Fuxe (1965) Evidence for the existence of monoamine neurons in the central nervous system. 11. Experimentally induced changes in the intraneuronal amine levels of bulbospinal neuron systems. Acta Physiol. Scand. 64(Suppl. 247):l-36. Eide, P.K., N.M. Joly, and K. Hole (1990) The role of spinal cord 5-HT1A and 5-HT1B receptors in the modulation of a spinal nociceptive reflex. Brain Res. 536,195-200. El Mestikawy, S., M. Riad, A.M. Laporte, D. Verge, G. Daval, H. Gozlan, and M. Hamon (1990) Production of specific anti-rat 5-HT1A receptor antibodies in rabbits injected with a synthetic peptide. Neurosci. Lett. 118:189-192. El-Yassir, N., S.M. Fleetwood-Walker, and R. Mitchell (1988) Heterogeneous effects of serotonin in dorsal horn of rat: Involvement of 5HT1 receptor subtype. Brain Res. 456:147-158. Fone, K.C.F., A.J. Robinson, and C.A. Marsden (1991) Characterization of the 5-HT receptor subtype involved in the motor behaviors produced by intrathecal administration of 5-HT agonists in rats. Br. J. Pharmacol. 103,1547-1555. Fung, S.J., and C.D. Barnes (1989) Raphe produced excitation of spinal motoneurons in the cat. Neurosci. Lett. 193:185-190. Hamon, M., H. Gozlan, S. El Mestikawy, M.B. Emerit, F. Bolanos, and L. Schechter (1990) The central 5-HTlAreceptor: Pharmacology, biochemistry, functional, and regulatoryproperties. Ann. NYAcad. Sci. 600:114131. Hamori, J. (1981) Synaptic input to the axon hillock and initial segment of inhibitory interneurons in the cerebellar cortex of the rat. Cell Tissue Res. 217;553-562. Hillegaart, V. (1990) Effects of local application of 5-HT and 8-OH-DPAT into the dorsal and median raphe nuclei on motor activity in the rat. Physiol. Behav. 48:143-148. Hillegaart, V., M.L. Wadenberg, and S. Ahlenius (1989) Effects of 8-OH DPAT on motor activity in the rat. Pharmacol. Biochem. Behav. 32797-800. Jackson, D.A., and S.R. White (1990) Receptor subtypes mediating facilitation by serotonin of excitability of spinal motoneurons. Neuropharmacology 29:787-797. Jacobs, B.L. (1976) An animal behaviorial model for studying central serotonergic synapses. Life Sci. 19:777-786. Jacobs, B.L., and C.A. Fornal(1993) 5-HT and motor control: A hypothesis. TINS 16:34&-351.

219

Kheck, N.M., P.J. Gannon, B.L. Jacobs, and E.C. Azmitia (1994) ~ - H T I A receptor localization in the rat, cat, and primate spinal cord. SOC. Neurosci. Abstr. 24:1547. Kolta, A., R. Dubuc, and J.P. Lund (1993) An immunocytochemical and autoradiographic investigation of the serotonergic innervation of trigeminal mesencephalic and motor nuclei in the rabbit. Neuroscience 53,11131126. La Motte, C.C. (1988) Lamina X of primate spinal cord: Distribution of five neuropeptides and serotonin. Neuroscience 25:639-658. La Motte, C.C., D.R. Johns, and N.C. Lanerolle (1982) Immunohistochemical evidence of indoleamine neurons in monkey spinal cord. J. Comp. Neurol. 206r359-370. Larkman, P.M., and J.S. Kelly (1991) Pharmacological characterization of the receptor mediating 5-HT evoked motoneuronal depolarization in vitro. In J.R. Fozard and P.R. Saxena (eds.): Serotonin: Molecular Biology, Receptors, and Functional Effects. Basel; Birkhauser Verlag, pp. 310-321. Marlier, L., F. Sandillon, P. Poulat, N. Rajoafetra, M. Geffard, and A. Privat (1991a) Serotonergic innervation of the rat spinal cord: Light and electron microscopic immunocytochemical study. J. Neurocytol. 20:310322. Marlier, L., J.R. Teilhac, C. Cerruti, and A. Privat (1991b) Autoradiographic mapping of 5-HT1, 5-HTlA, 5-HT1B and 5-HT2 receptors in the rat spinal cord. Brain Res. 550t15-23. Miquel, M.-C., E. Doucet, C. Boni, S. El Mestikawy, L. Mathiessen, G. Daval, D. Verge, and M. Hamon (1991) Central serotoninlA receptors: Respective distributions of encoding mRNA, receptor protein, and binding sites by in situ hybridization histochemistry, radioimmunohistochemistry, and autoradiographic mapping in the rat brain. Neurochem. Int. 19:453465. Palay, S.L., C. Sotelo, A. Peters, and P.M. Orkand (1968) The axon hillock and initial segment. J. Cell Biol. 38:193-201. Peters, A., and K.M. Harriman (1990) Different kinds of axon terminals forming symmetric synapses with the cell bodies and initial axon segments of layer II/III pyramidal celis. I. Morphometric analysis. J. Neurocytol. 19:154-174. Pubols, L.M., N.A. Bernau, L.A. Kane, S.D. Dawson, A.L. Burleigh, and AS. Polans (1992) Distribution of 5-HT1 binding sites in cat spinal cord. Neurosci. Lett. 142:lll-114. Radja, F., G. Daval, M. Hamon, and D. Verge (1992) Pharmacological and physicochemical properties of pre- versus postsynaptic 5-hydroxytryptamine1, receptor binding sites in the rat brain: A quantitative aut.oradiographicstudy. J. Neurochem. 58: 1338-1346. Rasmussen, K., and G.K. Aghajanian (1990) Serotonin excitation of facial motoneurons-receptor subtype characterization. Synapse 5~324-332. Rethelyi, M., and D. Lozsadi (1990) Synapses upon the axon origin of dorsal horn neurons in the rat spinal cord. Neurosci. Lett. 117:20-25. Riad, M., S. El Mestikawy, D. Verge, H. Gozlan, and M. Hamon (1991) Visualization and quantification of central 5-HT1, receptors with specific antibodies. Neurochem. Int. 19:413-423. Sasaki, S., S. Maruyama, and M. Takeishi (1990) Observation of the proximal portions of axons of anterior horn cells in the human spinal cord. Acta. Anat. 139t26-30. Somogyi, P. (1977) A specific ‘axo-axonal’ interneuron in the visual cortex of rat. Brain Res. 136:345-350. Sotelo, C., B. Cholley, S. El Mestikawy, H. Gozlan, and M. Hamon (1990) Direct immunocytochemical evidence of the existence of 5-HTlAautoreceptors on serotonergic neurons in the midbrain raphe nuclei. Eur. J. Neurosci. 2: 1 1 4 6 1154. Takahashi, T., and A. Berger (1990) Direct excitation of rat spinal motoneurons by serotonin. J. Physiol. 423:63-76. Takeuchi, Y., M. Kojima, T. Matsura, and Y. Sano (1983) Serotonergic innervation of the motoneurons in the mammalian brainstem. Anat. Embryol. 167:321-333. Thor, K.B., A. Blitz-Siebert, and C.J. Helke (1992)Autoradiographic localization of 5-HTI binding sites in the medulla oblongata of the rat. Synapse 10:185-205. Thor, K.B., S. Nickolaus, and C.J. Helke (1993) Autoradiographic localization of 5-hydroxytryptaminelA, 5-hydroxytryptaminelB, and 5-hydroxytryptaminelCi2 binding sites in the rat spinal cord. Neuroscience 55235-252. Ulfhake, B., U. Arvidsson, S. Cullheim, T. Hokfelt, E. Brodin, A. Verhofstad,

220 and T. Visser (1987) An ultrastructural study of 5-hydroxytryptamine-, thyrotropin-releasing hormone- and substance P-immunoreactive axon boutons in the motor nuclei of the motor nucleus of spinal cord segments L7-Sl in the adult cat. Neuroscience23:917-929. Wallis, D.J., J. Wu, andX. Wang(1993) Descending inhibition in the neonate rat spinal cord is mediated by 5-hydroxytryptamine.Neuropharmacology 3273-83. Wang, M.Y., and N.J. Dun (1990) 5-hydroxytryptamineresponses in neonate rat motoneurons In vitro. J. Physiol. 430:87-103. Whitaker-Azmitia, P.M., R. Murphy, and E.C. Azmitia (1990) Stimulation of astroglial 5-HTIAreceptors releases the serotonergic growth factor, protein S-100, and alters astroglial morphology. Brain Res. 528:155-158.

N.M. KHECK ET AL. Whitaker-Azmitia, P.M., C. Clarke, and E.C. Azmitia (1993) Localization of 5-HTlA receptors to astroglial cells in adult rats: Implications for neuronal-glial interactions and psychoactive drug mechanism of action. Synapse 14:201-205. Wright, D.E., K.B. Seroogy, K.H. Lundgren, B.M. Davis, and L. Jennes (1995) Comparative localization of serotoninlA, 1C and 2 receptor subtype mRNAs in rat brain. J. Comp. Neurol. 351:357-373. Yocca, F.D. (1990) Neurochemistry and neurophysiology of buspirone and gepirone: Interaction at pre-synaptic and postsynaptic 5-HTlAreceptors. J. Clin. Psychopharmacol. 10:6S-l2S. Zifa, E., and G. Fillion (1991) 5-Hydroxytrypamine receptors. Pharmacol. R ~ V 44:401-4.58. .