Serotonin transporter messenger RNA expression in neural crest-derived structures and sensory pathways of the developing rat embryo

Pergamon

Neuroscience Vol. 89, No. 1, pp. 243 265, 1999 Copyright ~5 1998 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved PII: S0306-4522(98)00281-4 0306M522/99 $19.00+0.00

SEROTONIN TRANSPORTER MESSENGER RNA EXPRESSION IN NEURAL CREST-DERIVED STRUCTURES A N D SENSORY PATHWAYS OF THE DEVELOPING RAT EMBRYO S. R. H A N S S O N , * § I~. M E Z E Y t a n d B. J. H O F F M A N * ~ *Unit on Molecular Pharmacology, Laboratory of Cellular and Molecular Regulation, National Institute of Mental Health, Bethesda, MD 20892-4090, U.S.A. +Basic Neurosciences Program, National Institute of Neurological Diseases and Stroke, Bethesda, MD 20892, U.S.A. Abstraet---A growing body of evidence suggests that serotonin plays an important role in the early development of both neural and non-neural tissues from vertebrate and invertebrate species. Serotonin is removed from the extracellular space by the cocaine- and antidepressant-sensitive serotonin transporter, thereby limiting its action on receptors. In situ hybridization histochemistry was used to delineate serotonin transporter messenger RNA expression during rat embryonic development. Serotonin transporter messenger RNA was widely expressed beginning prior to organogenesis and throughout the second half of gestation. Strikingly, serotonin transporter messenger RNA was detected in neural crest cells, some of which respond to serotonin in vitro, and neural crest-derived tissues, such as autonomic ganglia, tooth primordia0 adrenal medulla, chondrocytes and neuroepithelial cells, in the skin, heart, intestine and lung. Within the peripheral sensory pathways, two major cells types were serotonin transporter messenger RNA-positive: (i) sensory ganglionic neurons and (ii) neuroepithelial cells, which serve as targets for the outgrowing sensory neurons. Several sensory organs (cochlear and retinal ganglionic cells, taste buds, whisker and hair follicles) contained serotonin transporter messenger RNA by late gestation. The expression of serotonin transporter messenger RNA throughout the sensory pathways from central nervous system relay stations [Hansson S. R. et al. (1997) Neuroscience 83, 1185 1201; Lebrand C. et al. (1996) Neuron 17, 823 835] to sensory nerves and target organs as shown in this study suggests that serotonin may regulate peripheral synaptogenesis, and thereby influence later processing of sensory stimuli, if the early detection of serotonin transporter messenger RNA in skin and gastrointestinal and airway epithelia correlates with protein activity, it may permit establishment of a serotonin concentration gradient across epithelia, either from serotonin in the amniotic fluid or from neuronal enteric serotonin, as a developmental cue. Our results demonstrating serotonin transporter messenger RNA in the craniofacial and cardiac areas identify this gene product as the transporter most likely responsible for the previously identified accumulation of serotonin in skin and tooth germ [Lauder J. M. and Zimmerman E. F. (1988) J. craniq/ac. Genet. devl Biol. 8, 265-276], and the fluoxetine-sensitive effects on craniofacial [Lauder J. M. et al. (1988) Development 102, 709-720; Shuey D. L. et al. (1992) Teratology 46, 367 378: Shuey D. L. et al. (1993) Anat. Embryol., Berlin 187, 75-85] and cardiac [Kirby M. L. and Waldo K. L. (1995) Circulation Res. 77, 211-215; Yavarone M. S. et al. (1993) Teratology 47, 573 584] malformations. Serotonin transporter messenger RNA was detected in several neural crest cell lineages and may be useful as an early marker for the sensory lineage in particular. The distribution of serotonin transporter messenger RNA in early development supports the hypothesis that serotonin may play a role in neural crest cell migration and differentiation [Lauder J. M. (1993) Trends" Neurosci. 16, 233-240], and that the morphogenetic actions of serotonin may be regulated by transport. The striking pattern of serotonin transporter messenger RNA throughout developing sensory pathways suggests that serotonin may play a role in establishing patterns of connectivity critical to processing sensory stimuli. As a target for drugs, such as cocaine, amphetamine derivatives and antidepressants, expression of serotonin transporter during development may reflect critical periods of vulnerability for fetal drug exposure. The widespread distribution of serotonin transporter messenger RNA during ontogeny suggests a previously unappreciated role of serotonin in diverse physiological systems during embryonic development. '.~ 1998 IBRO. Published by Elsevier Science Ltd. Ke 3 words: antidepressants, cocaine, in situ hybridization histochemistry, monoamines, neural crest cells, peripheral nervous system.

++To whom correspondence should be addressed at: Lilly Research Laboratories, Drop Code 0510, Lilly Corporate Center, Indianapolis, IN 46285, U.S.A. §On leave from the Division of Molecular Neurobiology, Wallenberg Neurocenter, University of Lund, Sweden. Abbreviations': E, embryonic day; EDTA, ethylenediaminetetra-acetate; 5-HT, serotonin; 5-HTT, serotonin transporter; SSRI. serotonin-specific re-uptake inhibitor. 243

244

S.R. Hansson et al.

In the adult mammalian periphery, serotonin (5-HT) is widely distributed, with significant levels of this chemical messenger associated with enterochromaffin cells of the gastrointestinal tract and blood platelets. Peripheral nerves containing 5-HT have been detected in the gut, heart, lung, kidney, spleen thyroid, in blood vessels and in mast cells of certain species. 5-HT activates at least 14 different subtypes of 5-HT receptors, as identified by molecular cloning. These receptor subtypes are found in peripheral tissues and exhibit partially overlapping distributions. 39 In the CNS, after 5-HT is released, its temporal and spatial distribution is limited by re-uptake into presynaptic terminals by a cocaineand antidepressant-sensitive plasma membrane 5-HT transporter (5-HTT). It is the activity of the 5-HTT which limits the action of 5-HT on its receptors and permits the recycling of the neurotransmitter. In blood platelets, which do not synthesize 5-HT, 5-HT is accumulated through the 5-HTT. Regulation of extracellular 5-HT by 5-HT re-uptake prevents desensitization of 5-HT receptors affecting motility in the gut. 9s In the lung, 5-HT re-uptake by endothelial cells is thought to remove blood-borne 5-HT, thereby preventing vasoconstriction. 9s In addition to its functions in the adult, a growing body of evidence suggests that 5-HT may act as a morphogen, affecting cell migration, epithelialmesenchyme interactions and differentiation. In vitro models have shown that 5-HT modulates cell migration of palatal mesenchyme, 114 cardiac mesenchyme,46,112 craniofacial mesenchyme, 86'87 and vascular smooth muscle and aortic endothelial cells.~ ~ Studies of in vitro migration of mesenchymal cells from the developing palate and heart have provided evidence that 5-HT plays an important role in palate development, where it appears to mediate mesenchyme contractility, migration and palate-shelf elevation.~ 14 In the developing heart, transient 5-HT uptake in the myocardium adjacent to the endocardial cushions has been demonstrated, 46'ss'~2 as well as a 5-HT-mediated inhibition of endocardial mesenchyme cell migration in vitro. ~~2 In the mouse, 5-HT has been shown to affect craniofacial morphogenesis, 5°'86'87 possibly through 5-HTlA receptormediated inhibition of cranial neural crest cell migration. 6s Furthermore, functional 5-HT2B receptors have been identified in neural crest cells, 19 although a physiological role has yet to be identified. In the CNS, 5-HT induces neurogenesis and neuronal differentiation, and inhibits growth cone mobility and synaptogenesis in both mammals and invertebrates. 5-HT causes release of the glial-derived growth factor S-100b, which stimulates outgrowth of 5-HT neurons 5'~°4"1°5 and cortical neurons which receive 5-HT innervation. 5 While unsubstantiated to date, 5-HT may serve a similar role in the outgrowth of peripheral neurons to target organs. Following disruption of the 5-HT system by p-chlorophenylalanine-induced 5-HT depletion,

maturation of target neuronal structures was delayed. 49'52 Exposure of neonatal rat pups to 5,7dihydroxytryptamine or p-chlorophenylalanine altered dentate granule cell morphology by reducing the dendritic spine density.l°8 11o The results of these in vitro studies are suggestive of the roles that 5-HT may play in early development. Monoamines, including 5-HT, are among the earliest transmitters present in the mammalian embryo; 47'4s'52 however, the ability to synthesize 5-HT appears to develop relatively late in rodents. The rate-limiting enzyme in 5-HT biosynthesis, tryptophan hydroxylase, has been detected at embryonic day 15 (El5), 5~ two days after 5-HT neurons were first detected, 79 but others have established conversion of 5-HT from tryptophan later, at two 1°2 or three s2 days before birth. 5-HT may be synthesized at a much earlier time; however, demonstration of tryptophan hydroxylase is lacking. The presence of 5-HT immunoreactivity may represent either synthesis or accumulation of maternal 5-HT through 5-HTTmediated uptake and storage by the vesicular monoamine transporter, but does not distinguish between these two possibilities. 5-HT may gain access to the embryo very early in development through the high-affinity, antidepressant-sensitive 5-HTT localized to the maternal-facing brush border membranes of the synciotrophoblast. 6'25'26 Mouse embryos cultured in 10mM 5-HT showed a gradient of fluoxetinesensitive 5-HT immunoreactivity in the placenta from the giant cells to the inner cell layers] 1~ Based on these studies, it may be possible that 5-HT is transferred to the embryo from maternal blood supply early in development and that access of 5-HT to the embryo is regulated by 5-HTT. Several lines of evidence suggest that the morphogenetic actions of 5-HT might be regulated by transport. In the embryo, antidepressant-sensitive accumulation of 5-HT has been demonstrated in the chick neural tube 54"~°° and notochord, 54'~°° and in mouse craniofacial epithelia, heart and thyroid. 5°'53'88 Both craniofacial and cardiac malformation have been identified in mouse embryos following treatment with 5-HT uptake inhibitors, as well as 5-HT receptor agonists and antagonists in whole embryo cultures. 46'87'j1~ In the chick, exposure to 5-HTT inhibitors lead to the abnormal development of the floor plate and notochord. 54'1°° Although the pharmacological evidence suggests a role for 5-HT transport in morphogenesis, demonstration of specific transport proteins has been lacking. Molecular cloning of transporters has provided an opportunity to address this latter issue in rat, 9"36'38 human, 56'57'78 murine is and D r o s o p h i l a 23 5-HTT cDNAs. The gene encodes an approximately 630amino acid protein with 12 putative transmembrane domains. This protein is a member of a large family of Na+/C1--dependent transporters with transmembrane domains that are relatively conserved, and

Ontogeny of 5-HT transporter mRNA expression amino- and carboxy-termini that are less well conserved. 3'1° The cloning of 5-HTT has permitted the localization of 5 - H T T m R N A in both adult and embryonic tissues. In the adult rodent, 5-HTT m R N A has been localized by northern blot analysis to the lung, gut, spleen and uterus, in addition to the brain. 36"83 5-HTT m R N A is abundant in the enteric 5-HT neurons of the myenteric plexus. Recently, 5-HTT m R N A and immunoreactivity have been localized to embryonic adrenal gland 84 and thyroid. 92 Recent pharmacological evidence demonstrating that fluoxetine can block activation of 5-HT2c 68 and nicotinic acetylcholine receptors 27 expressed in Xenopus oocytes emphasizes the importance of identifying proteins with specific molecular probes. As an initial step towards correlating the effects of 5-HT inhibitors with the specific 5-HTT, we have determined the spatial and temporal distribution of 5-HTT m R N A in the developing rat embryo using in situ hybridization histochemistry.

245

were amplified at 95°C for 1 min, 55°C for 2 min and 72°C for 2.5 min for 35 cycles with a final extension at 72°C for 7 min. DNA templates were purified from agarose gels using GeneClean (Biol01). Complementary RNA probes were transcribed from 5 ng of gel-purified DNA template using [35S]UTP (Dupont NEN; 1300 Ci/mmol) and either T3 or T7 RNA polymerase according to manufacturer's instructions (Ambion MAXIscript) to generate sense and antisense probes, respectively.

Hybridization Prior to hybridization, tissue sections were fixed, dehydrated and delipidated as described previously. 12 Sections were hybridized (20-24 h, 55°C, humidified chamber) with 2 x l 0 6 c.p.m, of denatured [35S]cRNA probe per 80 ~tl hybridization buffer consisting of 20 mM Tris-HC1 (pH 7,4), 1 mM EDTA (pH 8.0), 300 mM NaC1, 50% formamide, 10% dextran sulfate, 1 x Denhardt's, 25 mg/ml yeast tRNA, 100 gg/ml salmon sperm DNA, 250 gg/ml total yeast RNA (fraction XI, Sigma), 100 mM dithiothreitol, 0.1% sodium thiosulfate and 0.1% sodium dodecyl sulfate. Slides were apposed to Hyperfilm Biomax MR (Kodak) for three days, then coated with nuclear track emulsion (NTB-3, Kodak). Slides were developed in Dektol (Kodak), fixed and counterstained with a Giemsa stain after an eight- to 12-week exposure at 4°C.

EXPERIMENTAL PROCEDURES

Section preparation

Microphotograph and figure preparation

Whole embryos were collected from timed-pregnant Sprague Dawley (Taconic Farms) rats on E8-E21 (plug day=E1). The embryos were frozen on powdered dry ice and stored at -80°C. Tissue sections (12 gm) were thawmounted on to silanized slides and stored at - 80°C prior to hybridization. Fresh frozen tissue rather than fixativetreated tissue was used in order to maximize the sensitivity for mRNA detection.

Microphotographs were prepared from an Axiophot microscope (Zeiss, Germany) equipped for dark-field and bright-field microscopy with a 35-mm camera and photographed using Kodak Tmax 400 technical film. Slides of photomicrographs were scanned from a Polaroid Sprintscan 35 slide scanner and captured electronically using either the NIH Image 1.61 software or Adobe Photoshop 3.0 at a 1:1 scale with 400-2500 dpi resolution, depending on the magnification of the original. Figures were assembled from captured images with Adobe Photoshop 3.0 and printed on matte-finished paper by a Fujix Pictrography 3000 (Fuji) printer.

RNA probes Two different non-overlapping probes were used corresponding either to the Y-untranslated and C-terminal portion of the open reading frame (Genbank accession no. M79450, nucleotides 1717 2201; 36 probe 1) or the 5'untranslated and N-terminal portion of the open reading frame (nucleotides 14404; probe 2). Specific DNA templates were generated by polymerase chain reaction from rat 5-HTT cDNA 36 using bipartite primers consisting of either a T7 RNA promoter and a downstream gene-specific sequence (anti-sense) or a T3 RNA promoter and an upstream gene-specific primer (sense). Polymerase chain reactions using 1 ng rat 5-HTT cDNA, I~M primers, 200gM of each deoxynucleotide triphosphate, 3 mM MgC12, 10 mM Tris (pH 8.3), 50 mM KCI and 2.5 units Taq polymerase (Boerhinger Mannheim)

RESULTS

Specificity of complementary RNA probes Probes directed against both the 3' (probe 1) and 5' (probe 2) ends of the 5-HTT m R N A showed an identical hybridization pattern in E l 8 embryos (Fig. 1) and in adult tissues (data not shown). Hybridization with probe 1 resulted in a better signal with less background (Fig. 1). Neither sense probe showed any specific hybridization (Fig. 1C, F). An

Abbreviations used in the figures I ll III

IV a at c ctx d drg e G h 1 li

neural plate neural tube yolk sac amnion aorta atrium cornea adrenal cortex dermis dorsal root ganglia epidermis gastrointestinal tract heart lens liver

IU

lz md md mx

nt P r r s sc

t

tc V

lung labyrinth zone adrenal medulla mandible maxilla neural tube penis raphe nuclei retina stomach subcutaneous layer tail telencephalon ventricle

246

S.R. Hansson et al.

Fig. 1. 5-HTT mRNA identified by specific cRNA probes. Dark-field (A) and bright-field (B) views of a cross-section through the torso of an El8 embryo, showing the lung and spinal column hybridized to probe 1. In dark field, developed silver grains appear white, while in bright field developed silver grains appear black. In an adjacent section through the same embryo hybridized with probe 2, dark-field (D) and bright-field (E) images demonstrated an identical pattern of hybridization. Dark field of hybridization to adjacent sections with the sense strand of either probe 1 (C) or probe 2 (F) showed no hybridization. Arrows indicate dorsal root ganglia, arrowhead indicates pleural lining and the asterisk marks the lumen of a pulmonary artery. The specificity of these probes was further emphasized by the identical hybridization patterns for at least four embryos at each age examined. An additional probe (data not shown) showed the same results as probes 1 and 2. Scale bars=335 gm (B, E). additional probe corresponding to the large glycosylated loop between transmembrane domains III and IV gave similar results, with much less intense hybridization signal (data not shown). These three regions of the m R N A were chosen based on the relative divergence between members of the Na+/ C1--dependent transporter gene family. Within this family, the monoamine transporters, 5-HT, dopamine and norepinephrine, are most highly related to one another. No hybridization with any of the three probes described above was detected in dopaminergic or noradrenergic neurons in the brain (data not shown), further confirming the specificity of these probes. For each time-point, four animals were analysed with consistently reproducible results. The use of multiple probes on sections from several animals with identical results demonstrates the specificity of these probes for the 5-HTT mRNA.

Early expression of serotonin transporter messenger RNA in extra-embryonic structures The placenta and associated membranes were examined from E8 to El4 (Table 1). 5-HTT m R N A was localized to the trophoblast giant cells of the placenta, especially on the side of the basal decidua, from E8 to El2 (data not shown). From El0 to El2, intense hybridization for 5-HTT m R N A was present in the cells of the expanding ectoplacental cone

(Fig. 2A, B). Initially, these cells formed a cluster within the placenta (Fig. 2A), from which they migrated to line the amniotic cavity (Fig. 2B-D). No sections were taken from this region after El2. By El4, the cells of the spongiotrophoblast zone in the placenta expressed the 5-HTT m R N A (Fig. 2E, F). In parallel with the ectoplacental cone, a strong hybridization signal was observed in the yolk sac (Table 1, Fig. 2C, D) surrounding the embryo.

Transporter messenger RNA expression appearing at embryonic days 10 and 11 Cardiovascular system. One of the earliest organs to develop, the heart, expressed 5-HTT m R N A by E11 (Table 2) in a large number of cells within the aortic wall and the heart tube (Fig. 3A, B). Between El0 and El5, positive cells remain in the walls of the heart and aorta (Fig. 3). Both the cell number and intensity of hybridization per cell in the heart increased until E18 and remained unchanged at E21. At E21, the signal is localized to the endocardium of the heart (data not shown). No hybridization was observed in atria or ventricles (Fig. 3E, F). Strong hybridization was detected at the root of larger vessels associated with the heart (Fig. 3E, F). Large vessels entering the heart showed roughly the same time-course of expression, with signal easily detected at El 1 (Table 2) and increasingly stronger

Ontogeny of 5-HT transporter mRNA expression

247

Table 1. Age-dependent expression of serotonin transporter messenger RNA in placenta, central and peripheral nervous systems, and sensory-related organs Structures Extra-embryonic structures Ectoplacental cone Yolk sac CNS-related structures Neural tube/CNS Retina Inner ea~. .

El0

Ell

El2

El3

El4

El5

E16

El7

El8

E21

+ +

+ +

+ +

ND ND

ND ND

ND ND

ND ND

ND ND

ND ND

ND ND

+

+

+

+ . .

+

+

+ + +

+ + +

+ + +

+ +

+ +

+ + +

+ + +

+ + + -

+ + + +

+

+ + + + +

.

.

Peripheral nervous system Ganglion Gasseri Dorsal root ganglia Adrenal medulla Sensory receptor system Vibrissa follicles Hair follicles Epidermis Tongue Tooth primordia

. + -

-

-

+ +

+ .

.

.

.

. + -

+ +

. + +

+ +

+ + +

+ indicates the presence of 5-HTT mRNA, - indicates no 5-HTT mRNA detected, blank indicates region not readily identifiable or developed at the given age, ND indicates that analysis was not determined.

up to El8. However, the most intense hybridization was seen in the ascending aorta, while the descending aorta and pulmonary veins (Fig. 3E, F) had less 5-HTT m R N A . Branches of the carotid artery, as well as vertebral and pulmonary arteries, were also strongly positive for 5-HTT m R N A . The majority of the m R N A appeared to be localized to the endothelium, but may be present in the smooth muscle as well.

Gastrointestinal tract. During early organogenesis ( E l l ) , 5-HTT m R N A was present in the intestinal tube (Table 2, Fig. 4A, B). 5-HTT m R N A was observed at El4, when the intestinal epithelium was first distinguishable. The intensity of the hybridization signal in the intestinal epithelium increases along its entire length throughout gestation (Fig. 4E-J), reaching the highest levels at E21 (Table 2). Beginning at El6, 5-HTT m R N A was also present in the smooth muscle layer, possibly in enteric ganglia. Notably, the epithelium in the stomach did not express 5-HTT m R N A during gestation (Fig. 4I, J), in contrast to the adult. 37 Transporter messenger RNA expression appearing at embryonic dav 12 Skin. 5 - H T T m R N A was first detectable in the Malphigian cell layer of the epidermis (Table 1) overlying the limb buds and in the lingual epithelium at E l 2 (Fig. 5A, B). The Malphigian layer consists of two cell layers, with the inner layer, stratum germinativum~ giving rise to cells of the epidermis. By E13, the epidermis over the entire body surface contained 5-HTT m R N A (data not shown). At this time-point,

positive cells were localized to the superficial germinal layer (Fig. 5C, D) of the epidermis, having two to three cell layers with no visible basal lamina. O f particular note, the epidermis of the developing snout exhibited a very intense hybridization signal at El3 (Fig. 7). By El8, the epidermis had an adult layer pattern, including a distinct basal lamina underlying the germinal layer. While 5-HTT m R N A - p o s i t i v e cells were still found in the germinal layer (basal layer of the epidermis; Fig. 5E, F), the number of cells and the intensity of hybridization in individual cells differed depending on which body surface was examined. F o r example, the hybridization signal intensity and the number of cells at the tip of the tail and the penis (Fig. 5G, H) appeared stronger from E l 6 to E21 than in skin on the trunk. In the dermis and the subcutaneous layer, 5-HTT m R N A was present in scattered cells. Based on their location and size, these cells were most likely blood-derived, either macrophages or mast cells. The number of positive subcutaneous cells also varied from region to region. For example, the subcutaneous layer under the mandible had more cells expressing the 5-HTT m R N A compared to other regions (data not shown).

Tongue. 5-HTT m R N A was first detectable in the lingual epithelium at E l 3 (Table 1, Fig. 6). Intense hybridization signal was detected over cell clusters within the epithelium (Fig. 6E, F), suggesting that the m R N A may be localized to developing taste buds. Within a given taste bud, only the lateral cells were positive. This level of expression persisted throughout gestation. At El6, 5-HTT m R N A - p o s i t i v e cells were also localized to specialized sensory structures within the muscle of the tongue (Fig. 6C).

248

S . R . Hansson et al.

~i!~!!¸~!ii~ii ~!iii¸ i~

Fig.

2.

Ontogeny of 5-HT transporter mRNA expression

249

Table 2. Age-dependent expression of serotonin transporter messenger RNA in peripheral organs Structures Cardiovascular system Heart Aorta Gastrointestinal tract Stomach Intestine Submandibular gland Thyroid

El0

Ell

El2

El3

El4

El5

El6

El7

El8

E21

+ +

+ +

+ +

+ +

+ +

+ +

+ +

+ +

+ +

+

+

+

+

+

. +

+ +

+ + +

.

. + +

.

+ indicates the presence of 5-HTT mRNA, - indicates no 5-HTT mRNA detected, blank indicates region not readily identifiable or developed at the given age.

Liver. ,At El2, the liver had a population of scattered, highly positive cells (data not shown). These cells, localized near the sinus lumen formed by the liver parenchyma, are most likely Kupfer cells, a specialized liver macrophage. This pattern remains unchanged up to E21. Tran~sporter messenger RNA expression appearing at embryonic day 13 Brown adipose tissue. The multilocular adipose tissue, also termed brown fat, located in the neck showed strong hybridization beginning at E l 3 (data not shown) and remained strong through El4. After this time, it was difficult to identify brown fat in the sections examined. Connective tissue. The developing bony structures of the head expressed 5-HTT m R N A beginning at El3 (data not shown). Of note was the signal in the developing palatal shelf at El6. Chondrocytes forming facial bones contained 5-HTT m R N A from E l 5 to E21. Snout and whiskers. The developing snout shows a uniformly strong signal at El3 (Table 1, Fig. 7B). By El6, these 5-HTT mRNA-positive cells migrate to form a condensed streak lining the future whisker pads (Fig. 7D). At El7, cells in the outer root sheet of the whisker follicles, specialized structures within the epidermis, showed a very strong hybridization signal (Fig. 7C, F-J). Similarly, cells surrounding the hair follicles on the body surface expressed 5-HTT

m R N A by E21 (Table 1). Based on their location, these are most likely Merkel cells, which form synapse-like junctions with axons of the innervating sensory neurons. The whisker pads are considered to be a "sticky" region for many probes. However, neither of the sense probes non-specifically hybridized to the whisker follicles. In addition, isolated Merkel cells expressed 5-HTT m R N A and transported 5-HT in a fluoxetine-sensitive m a n n e r (Hansson and Hoffman, unpublished observation).

Peripheral nervous system. In the head and neck region, most of the autonomic ganglia (Table 1), including ciliary, sphenopalatine and otic ganglia, are m R N A positive from El3 throughout gestation. The majority of neurons in the sympathetic chain and the dorsal root ganglia are also positive at this time (Fig. 8). In addition, 5-HTT m R N A was expressed in cells of the rami communicans running between the vertebrae (Fig. 8E, F). Transporter messenger RNA expression appearing at embo~onic day 15 Adrenal medulla. The adrenal medulla, composed of both ganglionic and chromaffin cells, expressed 5-HTT m R N A beginning at El5 in a scattered population of cells (Table 1). This signal is unchanged through E21 (Fig. 9C, D). Almost all cells showed strong hybridization in the well-demarcated adult adrenal medulla (Fig. 9E, F).

Fig. 2. 5-HTT mRNA in the developing placenta and early embryo. Bright field (A, B) shows 5-HTT mRNA in cells of the developing ectoplacental cone (arrowheads) at El0 (A) and E11 (B). Note that the positive cells seen at El0 (A) appear to have migrated to line the amniotic cavity (a) at Ell (B). No hybridization signal is detected in the placental cells (p). Bright-field (C) and dark-field (D) views of a section through an E11 embryo show that most of the cells lining the amniotic cavity (arrowheads) are 5-HTT mRNA positive. Positive cells are also found in the cells forming the yolk sac. Cells expressing 5-HTT mRNA are already present in the neural plate and the neural tube. No signal is seen in the amnion. Asterisks indicate cross-section through two ends of the embryo. Bright-field (E) and dark-field (F) views of a sagittal section through an El4 placenta indicate numerous 5-HTT mRNA-positive cells of the spongiotrophoblast zone (arrows) surrounding the labyrinth zone. Scale bars= 50 ~tm (A), 90 gm {B), 170 ~tm (C), 470gm (E, F).

O

ct~

Ontogeny of 5-HT transporter mRNA expression

Transporter messenger RNA expression appearing at embryonic day 16 Respiratory system. In continuity with the oral epithelium, the airway epithelium lining the sinuses of the dorsal nasal cavity, the trachea and the bronchi expressed 5-HTT m R N A beginning at El6 (data not shown). Cartilaginous structures such as the epiglottis and larynx expressed 5-HTT m R N A from El6 to E21. All cells of the pleural lining in the developing lung were intensely 5-HTT m R N A positive beginning at El6. Additional hybridization was apparent in the epithelium of the bronchi, while no signal was detectable in the lung mesenchyme. Transporter messenger RNA expression appearing at embryonic day 17 Ret&a. In the eye, the majority of cells in the ganglionic layer of the retina was strongly positive for 5-HTT m R N A beginning at El7 (Fig. 10C, D). No signal was detected in this region prior to El7 (Fig. 10A, B). The signal in the ganglionic cell layer subsequently remained unchanged through gestation (Table 1). Tooth primordia. Very strong hybridization was found in the dental pulp, both in the maxilla and mandible, beginning at El7 (Fig. 11). The signal remained unchanged throughout the period studied (Table 1). Inner ear. Beginning late in gestation (El7), 5-HTT m R N A was present in the cochlea of the inner ear (Table 1). A small group of ganglionic cells, the organ of Cortis, located in the epithelial lining of the cochlear duct near the edge of the basal membrane, expressed 5-HTT m R N A (Fig. 10E-H). In addition, the epithelial lining of the cochlear duct contained numerous mRNA-positive cells (Fig. 10E, F). Transporter messenger RNA expression appearing at embryonic day 2 l Thyroid gland, By E21, single positive cells were found in the interstitium between the thyroid-

251

producing follicles (Fig. 9A, B), suggesting that these might be calcitonin-producing cells (C-cells). This is the age at which follicular structure can first be identified. Prior to this time, the thyroid was not readily detectable in the sections examined. DISCUSSION 5-HT has been shown to affect cell migration in vitro and may be a morphogen in the developing mammalian embryo. Morphogenesis and migration of cells depend on appropriate spatial and temporal signals. The 5-HTT regulates the spatial and temporal distribution of 5-HT, and thereby, activation of 5-HT receptors. Thus, the localization of 5-HTT m R N A is suggestive of areas where 5-HT may play a role in cell migration or differentiation. In contrast to the adult rat, 9'36'38 we have found that 5-HTT m R N A is widely expressed throughout the developing embryo. From early gestation, 5-HTT m R N A was detected in the developing CNS in cells that are likely to be serotonergic neurons. 3° In the current study, 5-HTT m R N A was expressed in the sensory ganglia of the peripheral nervous system, in sensory receptor systems, in the enteric nervous system, in cartilage and bone, as well as in the gastrointestinal tract, the cardiovascular system and the respiratory system.

Serotonin transporter messenger RNA in neural crest cells and their derivatives Strikingly, most of the cells expressing 5-HTT m R N A in these systems are neural crest-derived cells. Neural crest cells are pluripotent precursors which migrate from along the outside of the neural tube to numerous locations in the body and differentiate depending on local cues. 13 In addition to forming serotonergic neurons, neural crest derivatives also give rise to cranial ganglia, adrenal chromaffin cells, connective tissue of the head and pigment cells of the epidermis. These are locations in which 5-HTT m R N A was detected during development.

Cardiovascular system. The circulatory system is the first functional system in the developing embryo. In rat heart, contractions are first evident on E l 5 ] 7 The neural crest cells give rise to sympathetic ganglia,

Fig. 3. 5-HTT mRNA in the developing heart and aorta. Bright-field (A) and dark-field (B) views of a cross-section through an E11 embryo. Asterisks indicate the double paired aorta. The heart tube (boxed) has intense hybridization for 5-HTT mRNA. Labeling is also seen in the aortic endothelium. Magnifications of the boxed area (A, B) are shown in bright-field (C) and dark-field (D) views, demonstrating that the hybridization signal is confined to the epithelium forming the developing heart. The presence of 5-HTT mRNA continues throughout gestation, as shown in bright-field (E) and dark-field (F) views of a cross-section through an E18 heart. The hybridization signal is found in the wall of the aorta wall, as well as in the endothelium of the pulmonary veins (stars). At the root of the aorta (arrowhead), high levels of 5-HTT mRNA are found in the sympathetic ganglia cells resident at this location. No signal was detected in either the atrium or the ventricle. Signal is present in the lung, as well as the liver. Scale bars= 100 lam (A), 40 ~tm (C), 300 p,m (E).

252

S. R. Hansson et al.

Fig. 4. 5-HTT m R N A in the embryonic gastrointestinal tract. Bright-field (A) and dark-field (B) views of a cross-section through an E11 embryo, indicating a very strong hybridization signal in the primitive gut (arrow). In addition, 5-HTT m R N A is present in the heart and aorta (also see Fig. 2). The telencephalon and the neural tube are shown for orientation. Bright-field (C) and dark-field (D) views of a cross-section through the intestine at El4 demonstrate hybridization in the developing epithelium (arrowheads) and smooth muscle layers. By El6, bright-field (E) and dark-field (F) views show that the 5-HTT mRNApositive cells are primarily seen in the epithelium lining the lumen (arrowhead). By E21, the 5-HTT m R N A expression in the epithelium is more obviously localized to villi (G: bright field; H: dark field). Note the absence of signal in the lamina propria in the middle of the villi. Bright-field (I) and dark-field (J) views of a low magnification at this same time-point (E21) demonstrate that no signal is present in the developing stomach at this age. Scale bars=200 ~am (A), 130 btm (C), 70 btm (E), 40 btm (G), 160 p.m (1).

Ontogeny of 5-HT transporter mRNA expression

253

Fig. 5. 5-HTT mRNA-positive cells are localized to specific layers of the skin and to the epithelial layer of the urethra. A series of bright-field (A, C, E) and dark-field (B, D, F) views of cross-sections through the skin from El2 (A, B), El6 (C, D) and E21 (E, F) embryos show the development of the different layers. The cells expressing 5-HTT mRNA are localized to the Malphigian cell layer (arrowheads) at El2 (A, B) and El6 (C, D). Note that the signal is located in the basal layer (arrowhead), or germinal layer, of the epidermis at E21, when the organization of layers most closely resembles mature skin. Scattered positive cells (arrows), most likely macrophages, are also seen in the subcutaneous layer. No signal was detected in the other epidermal layers or in the dermis. Bright-field (G) and dark-field (H) views of a longitudinal section through the penis at El6.5-HTT mRNA is present in the skin as well as in the epithelium lining the urethra (arrowheads). Similar to the entire epidermal surface of the embryo, the skin of the tail also expresses 5-HTT mRNA. Scale bars=20 gm (A)~ 15 gm (C), 40 gm (E), 180 gm (G). as well as contributing to the smooth muscle cells of the tunica media in the aortic arch 99 and the outflow septum. 46 The cardiac ganglia are located around the origin of the aorta, pulmonary trunk, veins and in the wall of the atria. The initial detection of 5-HTT m R N A at E l 0 in the heart tube and the subsequent presence of 5-HTT m R N A in the root of the aorta, atrial wall, pulmonary trunk and in other large vessels beginning at E l 3 suggest that this hybridiza-

tion is primarily localized in the cardiac ganglia rather than the myocardium itself. The 5-HTT m R N A expression in the developing aorta can be localized to two types of cells, smooth muscle cells in the aortic arch and, more rostrally, Type I cells of the carotid body, both of which are neural crest derived. O f these, Type I cells of the carotid body contain several monoamines and their biosynthetic enzymes. 43'm~ Expression of 5-HTT m R N A may be

254

S.R. Hansson et al.

A

Fig. 6. The epithelium of the developing tongue expresses high levels of 5-HTT mRNA. A bright-field view (A) of a sagittal section through the head of an El8 embryo indicating the area (box) shown at higher magnification in dark field (D). Similar views are shown in dark field for El3 (B) and El6 (C) embryos, demonstrating the presence of 5-HTT mRNA in the tongue epithelium beginning at mid-gestation. In addition, muscle spools, a specialization of muscle important in movement control, distinctly express 5-HTT mRNA. The mRNA-positive maxillary tooth primordium (arrowhead) is indicated in D. The tongue epithelium from D is shown at high magnification in bright field (E) and dark field (F) perpendicular to the surface of the tongue. Groups of cells (arrowheads) which appear to form part of taste buds express 5-HTT mRNA. Scale bars=300/am (B), 220/am (C), 10/am (E).

indicative of a serotonergic phenotype for these cells.43,101 Experimentally, two lines of evidence have suggested a role for 5-HT in heart development. In vitro exposure of mouse embryos to 5-HT uptake blockers resulted in cardiac malformations due to inhibition of cell proliferation in the cardiac mesenchyme, endocardium and m y o c a r d i u m ] 12 Exposure to cocaine

prenatally has been shown, in both humans and rodents, to result in an increased incidence of congenital cardiovascular abnormalities. 1,2,96 Since cocaine inhibits 5-HT, dopamine and norepinephrine transporters, the cardiovascular effects of cocaine may be partially mediated through the 5-HTT. The timing and localization of 5-HTT m R N A is consistent with these previously described effects of 5-HT

Ontogeny of 5-HT transporter mRNA expression

Fig. 7. 5-HTT mRNA is expressed in the supporting cells of the developing whisker. A bright-field view (A) of a sagittal section through the head of an El8 embryo indicating the area (box) shown at higher magnification in dark field (E). Similar views are shown in dark field for El3 (B) and El6 (D) embryos, highlighting 5-HTT mRNA-positive cells of the developing whisker follicles. At El3 (B), 5-HTT mRNA-containing cells are diffusely localized to the snout. As these cells begin to migrate and form the whisker pad (D), a clearly visible streak (arrows) of intense hybridization is apparent at El6. By El8 (E), 5-HTT mRNA-positive cells are confined to individual whisker follicles (arrows). Cells positive for 5-HTT mRNA are present in the tooth primordium (arrowhead in E) of the maxilla and the tongue (t) epithelium (D, E). Bright-field (C) and dark-field (F) views of El7 whisker pads sectioned in a plane perpendicular to the growth of the hair demonstrate the pattern of 5-HTT mRNA expression in the follicle cells surrounding each whisker (arrows). At a higher magnification of a single whisker, bright-field (G) and dark-field (H) views indicate that the positive cells are localized to the outer root sheet surrounding the whisker (star). In a cross-section perpendicular to the skin surface, bright-field (I) and dark-field (J) views show the hair follicle (arrowhead), extending from the epidermis to the dermis above the subcutaneous layer. Note that not all the cells forming the root sheet express 5-HTT mRNA. Scale bars=300 ~tm (B). 370 ~tm (D), 320 ~tm (E), 110 ~tm (F), 5 ~tm (G), 8 ~tm (I).

255

256

S . R . Hansson et al.

Fig. 8.5-HTT m R N A in dorsal root and sympathetic ganglia from mid to late gestation. Bright-field (A) and dark-field (B) views of an El3 embryo through the spine (s) demonstrate the localization of 5-HTT m R N A to the dorsal root ganglia (arrows), as well as the sympathetic chain located on the ventral side of the spine. Positive cells are also seen in the liver. Bright-field (C) and dark-field (D) views indicate expression of 5-HTT m R N A in a large proportion of the neurons in the dorsal root ganglia at El8. In addition, a strong hybridization signal appears in the sympathetic chain (arrowhead). Bright-field (E) and dark-field (F) views of an El8 embryo through the spine show intense hybridization in the dorsal root ganglia (arrows), as well as rami communicans passing between the vertebrae (vt), which are negative. 5-HTT m R N A is present in the majority of cells in pleural lining (arrowheads). Scale bars=220 ~tm (A), 50 pm (C), 100 ~tm (E).

a

Fig. 9.5-HTT m R N A expression in developing thyroid and adrenal glands. Bright-field (A) and dark-field (B) views of a section through the thyroid of an E21 embryo. Based on location and size, the 5-HTT mRNA-positive cells are most likely C-cells (arrows) situated between the thyroid follicles (arrowheads). No signal is found in the follicular epithelium. Bright-field (C, E) and dark-field (D, F) views of sections through E21 (C, D) and adult (E, F) adrenal glands. At E21, there are fewer positive cells in the medulla, which is more diffuse and less well delimited compared to the adult. Some of the positive cells located on the border of the cortex are indicated with arrows. In the adult rat, the positive cells are strictly localized to the medulla (md), where the majority of chromaffin and ganglionic cells contain 5-HTT mRNA. Scale bars=35 gm (A), 70 ~m (C), 220 gm (E).

258

S. R. Hansson

Fig. 10

et al.

Ontogeny of 5-HT transporter mRNA expression

259

Fig. 11. 5-HTT mRNA is present in the developing tooth primordia. Bright-field (A) and dark-field (B) views of a sagittal section through the oral cavity of an E17 embryo show very strong hybridization in the tooth primordia (arrowheads) of the maxilla and the mandible. The intense labeling is localized to the sensory nerves of the dental pulp. Additional hybridization signal is found in the skin overlying the snout, as well as the epithelium lining the tongue (t). Scale bar=270 gm (A).

and 5-HTT inhibitors on cardiac morphogenesis. 46'112 Thus, defects in the expression of 5-HTT may underlie certain malformation syndromes associated with the heart. In contrast to these results, 5-HT-specific re-uptake inhibitor (SSRI) administration during the first trimester was not associated with an increase in cardiac malformation 76 (see discussion below).

Peripheral nervous system. Multiple neural crest derivatives in the peripheral nervous system expressed 5-HTT m R N A . Dorsal root ganglion neurons, derivatives of the sensory lineage] 4's9 were 5-HTT m R N A positive throughout the embryonic period. Hybridization signal was also detected in cells of the sympathetic nervous system which originate from the sympathoadrenal precursors derived from a separate neural crest lineage. 4 5-HTT m R N A was also detected in the rami communicans, the sensory nerves connecting the dorsal root ganglia and the spinal cord. To date, monoamines have not been identified in rami communicans. It is possible that 5-HTT m R N A may also be present in the supporting

Schwann cells, which are derived from the neural crest through a cell-type specific lineage. 41'45"64 Neural crest cells begin to colonize the bowel at E 1 0 - E l l in rat. Derived from the vagal and sacral levels of neural crest cells, 5-HT neurons are among the earliest born of enteric neurons (E8.5 El4). 2s'29 5-HTT m R N A expression in the primitive gut is initially found in the majority of cells forming the intestinal tube (El 1), suggesting that these ceils may be the "transiently catecholaminergic" cells, some of which either are or will become 5-HT neurons. F r o m El6, the expression is clearly localized to the epithelial lining of the intestine and, most likely, to the intramural serotonergic neurons derived from the sympathoenteric lineage. 2s'29 During embryogenesis, the gastrointestinal lumen is constantly exposed to amniotic fluid which contains 5-HT, most likely originating from the maternal circulation. 1~ The enteric nervous system may also be a source of 5-HT in amniotic fluid or locally. In either case, if the 5-HTT in the epithelial lining were functional, a gradient of 5-HT could be generated, either locally or along the length of the intestine. Such a 5-HT

Fig. 10. Ganglionic cells of the retina and the inner ear express 5-HTT mRNA. Bright-field (A, C) and dark-field (B, D) views of sagittal sections through the eyes of El4 (A, B) and E17 (C, D) embryos indicate that 5-HTT mRNA is not present in the retina until El7, when layers first become discrete. The cells expressing 5-HTT mRNA are localized to the ganglionic cell layer (arrowhead). Additional hybridization signal is present in the skin of the eyelid (arrows). No signal is found in the cornea or the lens. Bright-field (E) and dark-field (F) views of a section through the cochlea of an E21 inner ear show 5-HTT mRNA-positive cells lining the cochlear duct (arrowhead). At the most lateral areas of the basal membrane, two symmetric groups of ganglionic ceils (arrows) are intensely mRNA positive. One group of these cells (boxed) is shown at higher magnification in bright field (G) and dark field (H) to demonstrate the cellular nature of the hybridization. Scale bars=60 lam (A), 150 ~tm (C), 240 ~tm (E), 40 ~tm (G).

260

S.R. Hansson et al.

gradient could act as a differentiative or migratory cue in the gut. Craniofacial structures. The developing bony structures of the head, including the palatine shelf and the facial skeleton, are derived from cranial neural crest cells. The presence of 5-HTT mRNA (El3 El6) precedes the critical time when the shelves of the palate begin to fuse (E15-E21).114 At the same time, 5-HT 3 receptor mRNA has also been detected in the developing bone, 93 5-HT 3 receptor agonists stimulated increases in expression of cartilage core protein from mandibular mesenchyme cells in vitro 67 and 5-HT stimulated the motility of palate cells in culture.ll4 Furthermore, cultured embryos treated with 5-HT uptake inhibitors (fluoxetine or sertraline) exhibited serious craniofacial malformations resulting from inhibition of cell proliferation. 86 Interestingly, it appears that there is a primary defect of neural crest cells in spontaneously occurring craniofacial malformations. 42 Many craniofacial malformation syndromes have associated malformations in the brain and cardiovascular system, 42 indicating that a defect in a common neural crest precursor may be responsible. Taken together, this evidence raises the possibility that the 5-HTT plays a crucial role in maintaining the appropriate levels of 5-HT during this critical period of craniofacial development. Thyroid. 5-HTT mRNA was evident in the interstitium of the thyroid gland at E21, when the hormone-producing follicles have formed. The neural crest-derived parafollicular cells are located in the interstitium between these follicles, suggesting that 5-HTT mRNA is present in these cells. Parafollicular cells arise from the same sympathoenteric lineage as the rostral enteric nervous system, including enteric serotonergic neurons,2~ and can be induced in vitro by nerve growth factor to acquire a serotonergic phenotype similar to neurons including expression of the 5-HTT. 2° Our results are in contrast to a recent study by Tamir et al., 92 demonstrating a functional 5-HT uptake in follicular cells as early as El3 in mouse (equivalent to El4 in rat), with no detectable uptake in parafollicular cells. It is possible that there are low levels of 5-HTT mRNA present in the follicular epithelium, below the sensitivity of our hybridization. Perhaps the mRNA detected in parafollicular cells is not actively transcribed or rat differs from mouse. Further studies should resolve these issues. In summary, our results demonstrate that, throughout gestation, there is 5-HTT mRNA expression in neural crest derivatives originating from several lineages. 5-HTT may be an early marker of neural crest cell progenitors for certain lineages (or sublineages). It should be emphasized that the detection of 5-HTT mRNA gene expression does not necessarily reflect the presence of protein or of functional protein activity. However, the results of this study do suggest that the 5-HTT may have a func-

tional impact on migration and differentiation of neural crest cells and their derivatives. Possible mechanisms include establishment of a 5-HT gradient as a reciprocal signal between neural crest cells and their target organs or allow for the accumulation of 5-HT and subsequent use as a borrowed transmitter to signal along the migration path. Sensor), p a t h w a y s

5-HTT mRNA was expressed in sensory organs such as skin, whiskers, taste buds, retina and cochlea, as well as in dorsal root ganglia and cranial ganglia. The sensory lineage of the neural crest gives rise to precursors that populate all of these areas. Within the sensory pathways, two major cell types were 5-HTT mRNA positive: (i) ganglionic sensory neurons and (ii) neuroepithelial cells. In the developing skin, 5-HTT mRNA was expressed in both the germinal cell layer and its precursors beginning at El2 and in the cells surrounding the hair follicles later in embryogenesis. The areas which showed most intense hybridization signal corresponded to those areas of the skin which receive the greatest sensory innervation, such as the snout, tail and penis. Furthermore. the dorsal root ganglion neurons, which relay somatic sensory information to the CNS, innervate the skin and form specialized sensory receptors at the level of the germinal cells and the hair follicles in the epidermis. These same sensory neurons not only express 5-HTT mRNA, but also express the vesicular monoamine transporter, 1°3 suggesting that these cells are capable of accumulating and storing 5-HT for regulated release. Therefore, re-uptake of 5-HT at the end organ may influence sensory innervation by locally decreasing 5-HT levels released from outgrowing neurons or by establishing a 5-HT concentration gradient of 5-HT from the amniotic fluid or from blood, as has been suggested. ~6 5-HTT mRNA expression was also found in scattered cells in deeper skin layers, such as the dermis and subcutaneous layer. The number of positive subcutaneous cells varied from region to region, with the subcutaneous layer under the mandible having more 5-HTT mRNA-positive cells than other regions. These regional differences suggest that the positive cells in the subcutaneous region may alternatively be adipocytes. 33'~4In addition, some of these cells are likely to be either blood-derived macrophages or mast cells, both of which are known to transport 5-HT. 9s In sensory epithelium, such as skin, whiskers, tongue, the nasal cavity and the mechano- and electroreceptors of the acoustic system, 73 75,8J.lo6 a family of specialized neuroepithelial cells form synaptic contacts with sensory neurons. During development, these cells may serve as targets for the outgrowing sensory neurons. Cells in the snout are intensely 5-HTT mRNA positive early (El3) and

Ontogeny of 5-HT transporter mRNA expression may be the precursors for the cells which ultimately form the outer root sheet of individual whisker follicles (and are 5-HTT m R N A positive). Based on their location, these cells are most likely Merkel cells. It has been suggested that Merkel cells release nerve growth factor to support the outgrowing nerves. 97 Interestingly, neurons of the ganglion Gasseri, which innervate the whisker pads, expressed 5-HTT m R N A as well. Therefore, 5-HT might be important in the formation of the Merkel cell-neurite synapse, in the same way that 5-HT has a role in central synaptogenesis 35"4952 (see Introduction). In the tongue epithelium, 5-HTT m R N A was localized to clusters of cells (El3) which will develop into the taste buds late in gestation and postnatally. 91 Similarly to the Merkel cells of the whisker follicles, the 5-HTT mRNA-positive cells of the taste bud are located at the taste bud periphery and make synaptic contact with adjacent taste cells and with innervating fibers. 22'24'44 Whether 5-HT acts as a neuromodulator in gustatory signaling or as a neurotrophic factor for the developing taste bud 85 is still not clear. A very intense hybridization signal was found in the tooth primordia of the maxilla and mandible. The tooth primordium, which is neural crest derived, gives rise to enamel-producing ameloblasts as well as innervating sensory neurons. 7172 In the mouse embryo, 5-HT 3 receptors have been demonstrated in the developing teeth. 93 Interestingly, 5-HT stimulates development of tooth germs from mandibular explant cultures. 66 Fluoxetine, a 5-HT uptake inhibitor, reverses this effect and blocks uptake of 5-HT into tooth germs and mandibular epithelium. Taken together, these data suggest that 5-HTT may be positioned to regulate tooth development by modulating 5-HT levels. In both the retina and cochlea, ganglionic cells are 5-HTT mRNA positive. 5-HT has been shown to inhibit neurite sprouting in goldfish retinal explants. 6° By stimulating these receptors with a specific 5-HTjA receptor agonist or by increasing extracellular levels of 5-HT with specific uptake blockers, the same inhibitory effect could be mimicked. 6° Inhibition of neurite sprouting may inhibit synapse formation or may, in fact. be necessary for differentiation and synaptogenesis. 5-HT immunoreactivity has been detected in the adult and developing amphibian retina, 6~'~j3 but the role of 5-HT in the mammalian retina remains unclear. Clarification of the types of cells and the localization of the transporter protein will be needed to further address the role of 5-HT in the retina. In the cochlea, two types of cells were 5-HTT m R N A positive, ganglionic cells in the organ of Corti and epithelial cells lining the cochlear duct. In the epithelium lining the cochlea, either chondrocytes or melatonin-producing cells may express 5-HTT mRNA. Melatonin has been suggested as a possible modulator of auditory signaling. 8 By regulating 5-HT levels, auditory signals might be modulated by providing the precursor to melatonin and conse-

261

quently increasing melatonin levels. In the developing ear, 5-HT may play a role in synaptogenesis, as it does centrally. The expression of 5-HTT mRNA throughout the sensory pathways suggests that 5-HT may play a role in establishing patterns of connectivity critical to processing sensory stimuli. Spontaneous activity has been shown to be important for proper development of the visual 63 as well as the auditory system. 94 Furthermore, 5-HT depletion delayed the maturation of target neurons in the CNS. 49'52'1°8 ~o In the adult, the serotonergic system is known to modulate sensory information processing. 4° Notably, 5-HTT mRNA was expressed not only in sensory organs such as the retina, cochlea, whiskers, skin and the neurons of the sensory ganglia, but also transiently in the corresponding relay stations in the CNS, the trigeminal nucleus, 3° the medial and lateral geniculate and the ventrobasal thalamus. 3°'55 Disruption of the transiently serotonergic corticothalamic neurons disrupted barrel field formation in the somatosensory cortex. 55 Recently, functional 5-HT uptake has been demonstrated in embryonic Merkel cells in vitro (Hansson and Hoffman, unpublished observations). Both the ganglion Gasseri neurons (this study), which innervate the whisker pads, and the trigeminal nucleus, ~° which receives innervation from the ganglion Gasseri (this study), express 5-HTT m R N A and fluoxetinesensitive transporter binding sites (Hansson, Cabrera-Vera and Hoffman, unpublished observations). Furthermore, these brain areas 3°31"55 and the sensory ganglion neurons 32"m3 also express the vesicular monoamine transporters that would enable them to store accumulated 5-HT for subsequent regulated release. Taken together, these data suggest that regulation of 5-HT may be important to balance the ongoing neuronal activity and for synapse formation along the entire afferent pathway. Serotonin homeostasis in amniotic fluid

The localization of 5-HTT mRNA in placental trophoblasts, and in the yolk sac, suggests that 5-HTT may regulate the 5-HT concentration in the amniotic cavity surrounding the embryo. 5-HT regulates hormonal secretion from placental giant cells and may control invasion of these giant cells into the decidua. 26'1t~ Clearance of monoamines, in general, and 5-HT in particular has been shown in the placenta, s°'9° Our results are in agreement with other studies describing 5-HT uptake, 5-HTT mRNA 77'7S and 5-HT immunoreactivity in both human 26'77'78 and mouse placenta. ~J~ The source of 5-HT during early stages of development remains unclear; however, maternal blood or mast cells are possible sources. Although the placenta is thought to limit access of maternal 5-HT to the embryo by re-uptake, 2~' other studies have suggested that 5-HT is transported across the placenta to the inner cell layers, where it may gain access to the embryo. Early expression of 5-HTT

262

S.R. Hansson et al.

m R N A in the placenta may provide a source of 5-HT prior to embryonic synthesis. The appearance of 5-HTT m R N A in the placenta coincides with early organogenesis. In the embryo, early expression of 5-HTT m R N A in the skin and, later, to the airway epithelium and the gastrointestinal tract emphasizes a localization to surface areas continuously exposed to amniotic fluid. Expression of 5-HTT on these surfaces would allow an efficient clearance by the embryo itself. In newborn mice, injection of 5-HT induced multiple neural crest tumors, hyperplasia and cell death, 69'7°'8° indicating that the embryo might be equally sensitive to inappropriate levels of 5-HT. Alternatively, the 5-HT present in amniotic fluid may allow the establishment of a concentration gradient across epithelia as a developmental cue. It should be emphasized that the presence of 5-HTT m R N A does not necessarily reflect the presence of functional protein. Furthermore, a disjunction between m R N A synthesis, protein translation and functional protein may occur, as has been noted in a differentiated tertocarcinoma cell line. ~5 These possibilities will require further experimentation in specific pathways. If, indeed, 5-HTT is widely expressed in embryos, teratogenic effects would be expected as a result of exposure to SSRIs, cocaine or amphetamines. In contrast to the malformations observed in cultured murine embryos, there is limited evidence for teratogenic effects in intact animals or in humans. While various biochemical and functional effects on the 5-HT system have been described in progeny following in u t e r o administration of fluoxetine to pregnant rat dams, no evidence for morphological malformations have been noted. 7'5s'59 Furthermore, in humans, administration of SSRIs during the first trimester of pregnancy resulted in an increased risk of miscarriage, but without complications of major malformation. 76 However, the transporter may not be constantly blocked, effects may be seen later or may be more

subtle, or perturbation of the 5-HT system may result in compensatory responses either in the 5-HT system or in an interacting, related system. In support of this latter possibility, mice lacking the 5-HT~B receptor appeared normal by all major criteria until individual circuitry was tested or until the animal was stressed. 2j'62 As predicted from the intact animal studies, mice lacking the 5-HTT do not show any gross malformations; 1°7 however, there are compensatory changes in the 5-HT system, with a desensitized response to the 5-HTIA agonist despite normal levels of receptor. 1°7 Taken together, our data and those discussed here from other studies should stimulate further studies to resolve the role of the 5-HTT and antidepressants in development. CONCLUSION

The 5-HTT m R N A was widely expressed throughout the developing embryo beginning prior to organogenesis. The expression of this m R N A in derivatives of the neural crest suggests that the 5-HTT m R N A may be a useful marker for particular lineages of neural crest cells. The pattern of 5-HTT m R N A expression is consistent with this gene being responsible for previously described craniofacial and cardiac malformations following SSRI treatment. Furthermore, 5-HTT may regulate 5-HT levels in the neural crest cell microenvironment, affecting migration and/or differentiation. Especially noteworthy was the expression throughout the sensory pathways. Our results suggest that 5-HT may play a role in setting up patterns of connectivity critical to processing sensory stimuli. Drugs which block the 5-HTT, such as fluoxetine (Prozac), may not consistently result in malformation or other obvious congenital defects; however, the current study suggests that there may be critical periods of vulnerability when fetal exposure to these drugs might affect one system more than another. These observations provide a further guide for future studies of the role of 5-HT in particular pathways and cell types.

REFERENCES 1. Akbari H. M., Kramer H. K., Whitaker-Azmitia P. M., Spear L. P. and Azmitia E. C. (1992) Prenatal cocaine exposure disrupts the development of the serotonergic system. Brain Res. 572, 57-63. 2. Akbari H. M., Whitaker-Azmitia P. M. and Azmitia E. C. (1994) Prenatal cocaine decreases the trophic factor S-100 beta and induced microcephaly: reversal by postnatal 5-HT1A receptor agonist. Neurosci. Lett. 170, 141-144. 3. Amara S. G. and Kuhar M. J. (1993) Neurotransmitter transporters: recent progress. A. Rev. Neurosci. 16, 73-93. 4. Anderson D. J. (1993) Molecular control of cell fate in the neural crest: the sympathoadrenal lineage. A. Rev. Neurosci. 16, 129-158. 5. Azmitia E. C., Dolan K. and Whitaker-Azmitia P. M. (1990) S-100B but not NGF, EGF, insulin or calmodulin is a CNS serotonergic growth factor. Brain Res. 516, 354-356. 6. Balkovetz D. F., Tiruppathi C., Leibach F. H., Mahesh V. B. and Ganapathy V. (1989) Evidence for an imipramine-sensitive serotonin transporter in human placental brush-border membranes. J. biol. Chem. 264, 2195 2198. 7. Battaglia G., Cabrera T. M. and Van de Kar L. D. (1995) Prenatal cocaine produces biochemical and functional changes in brain serotonin systems in rat progeny. NIDA Res. Monogr. 158, 115-148. 8. Biesalski H. K., Welker H. A., Thalmann R. and Vollrath L. (1988) Melatonin and other serotonin derivatives in the guinea pig membranous cochlea. Neurosci. Lett. 91, 4146. 9. Blakely R. D., Berson H. E., Fremeau R. T. Jr, Caron M. G., Peek M. M., Prince H. K. and Bradley C. C. (1991) Cloning and expression of a functional serotonin transporter from rat brain. Nature 354, 66-70.

Ontogeny of 5-HT transporter mRNA expression 10.

263

Borowsky B. and Hoffman B. J. (1995) Neurotransmitter transporters: molecular biology, function, and regulation. Int. Rev. Neurobiol. 38, 139 199.

11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

24. 25. 26. 27. 28. 29. 30.

31. 32.

Bottaro D., Shepro D., Peterson S. and Hechtman H. B. (1985) Serotonin, histamine, and norepinephrine mediation of endothelial and vascular smooth muscle cell movement. Am. J. Physiol. 248, C252-C257. Bradley D. J., Towle H. C. and Young W. S. III (1992) Spatial and temporal expression of ct- and L3-thyroid hormone receptor mRNAs, including the [32-subtype, in the developing mammalian nervous system. J. Neurosci. 12, 2288 2302. Bronner-Fraser M. (1993) Environmental influences on neural crest cell migration. J. Neurobiol. 24, 233-247. Bronner-Fraser M. (1995) Origins and developmental potential of the neural crest. Expl Cell Res. 218, 405-417. Buc-Caron M. H., Launay J. M., Lamblin D. and Kellermann O. (1990) Serotonin uptake, storage, and synthesis in an immortalized committed cell line derived from mouse teratocarcinoma. Proc. natn. Acad. Sci. ~(S.A. 87, 1922 1926. Buznikov G. A., Shmukler Y. B. and Lauder J. M. (1996) From oocyte to neuron: do neurotransmitters function in the same way throughout development? Cell. molec. Neurobiol. 16, 537 559. Chacko K. J. (1976) Observations on the ultrastructure of developing myocardium of rat embryos. J. Morph. 150, 681 709. Chang A. S., Chang S. M., Starnes D. M., Schroeter S., Bauman A. L. and Blakely R. D. (1996) Cloning and expression of the mouse serotonin transporter. Brain Res. molec. Brain Res. 43, 185-192. Choi D. S., Ward S. J., Messaddeq N., Launay J. M. and Maroteaux L. (1997) 5-HT2B receptor-mediated serotonin morphogenetic functions in mouse cranial neural crest and myocardiac cells. Development 124, 1745 1755. Clark M. S., Lanigan T. M., Page N. M. and Russo A. F. (1995) Induction of a serotonergic and neuronal phenotype in thyroid C-cells. J. Neurosci. 15, 6167 6178. Crabbe J. C., Phillips T. J., Feller D. J., Hen R., Wenger C. D., Lessov C. N. and Schafer G. L. (1996) Elevated alcohol consumption in null mutant mice lacking 5-HT1B serotonin receptors. Nat. Genet. 14, 98 101. Delay R. J., Taylor R. and Roper S. D. (1993) Merkel-like basal cells in Necturus taste buds contain serotonin. J. eomp. Neurol. 335, 60(~613. Demchyshyn L. L., Pristupa Z. B., Sugamori K. S., Barker E. L., Blakely R. D., Wolfgang W. J., Forte M. A. and Niznik H. B. (1994) Cloning, expression, and localization of a chloride-facilitated, cocaine-sensitive serotonin transporter from Drosophila melanogaster. Proc. natn. Acad. Sci. U.S.A. 91, 5158 5162. Fujimoto S., Ueda H. and Kagawa H. (1987) Immunocytochemistry on the localization of 5-hydroxytryptamine in monkey and rabbit taste buds. Acta anat., Basel 128, 80-83. Ganapathy V. and Leibach F. H. (1995) Placental biogenic amines and their transporters. In Placental Toxicology (ed. Ramasastry B. V.), pp. 161 174. CRC, Boca Raton, FL. Ganapathy V., Ramamoorthy S. and Leibach F. H. (1993) Transport and metabolism of monoamines in the human placenta. Trophoblast Res. 7, 35 51. Garcia-Colunga J., Awad J. N. and Miledi R. (1997) Blockage of muscle and neuronal nicotinic acetylcholine receptors by fluoxetine (Prozac). Proc. natn. Acad. Sci. US.A. 94, 2041 2044. Gershon M. D. (1997) Genes and lineages in the formation of the enteric nervous system. Curr. Opin. ~urobiol. 7, 101 109. Gershon M. D., Chalazonitis A. and Rothman T. P. (1993) From neural crest to bowel: development of the enteric nervous system. J. Neurobiol. 24, 199-214. Hansson S. R., Mezey 1~. and Hoffman B. J. (1998) Serotonin transporter messenger RNA in the developing rat brain: early expression in serotonergic neurons and transient expression in non-serotonergic neurons. Neuroseienee 83, 1185 1201. Hansson S. R., Mezey I~. and Hofl'man B. J. (1998) Ontogeny of vesicular monoamine transporter mRNAs. VMATI and VMAT2. I. The developing rat central nervous system. Brain Res. devl Brain Res. 110, 135-158. Hansson S. R., Hoffmann B. J. and Mezey t~. (1998) Ontogeny of vesicular monoamine transporter mRNAs, VMAT1 and VMAT2. II. Expression in neural crest derivatives and their target sites in rat. Brain Res. devl Brain Res. 110, 159-174.

33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.

Hausman G. J. (1982) Adipocyte development in subcutaneous tissues of the young rat. Acta anat., Basel 112, 185 196. Hausman G. J., Campion D. R., Richardson R. L. and Martin R. J. (1981) Adipocyte development in the rat hypodermis. A m . . L Anat. 161, 85 100. Haydon P. G., McCobb D. P. and Kater S. B. (1984) Serotonin selectively inhibits growth cone motility and syuaptogenesis of specific identified neurons. Science 226, 561 564. Hoffman B. J. (1994) Expression cloning of a serotonin transporter: a new way to study antidepressant drugs. Pharmaeopsychiatry 27, 16-22. Hoffman B. J., Hansson S. R., Hunyady B. and Mezey l~. (1997) Localization of a serotonin transporter in epithelial and endocrine cells of the stomach and intestine. Gastroenterology 112, A l l 55. Hoffman B. J., Mezey E. and Brownstein M. J. (1991) Cloning o f a serotonin transporter affected by antidepressants. Science 254, 579--580. Hoyer D., Clarke D. E., Fozard J. R., Hartig P. R., Martin G. R., Mylecharane E. J., Saxena P. R. and Humphrey P. P. (1994) International Union of Pharmacology classification of receptors for 5-hydroxytryptamine (serotonin). Pharmac. Rev. 46, 157 203. Jacobs B. U and Azmitia E. C. (1992) Structure and function of the brain serotonin system. Physiol. Rev. 72, 165 229. Jessen K. R., Brennan A., Morgan L., Mirsky R., Kent A., Hashimoto Y. and Gavrilovic J. (1994) The Schwann cell precursor and its fate: a study of cell death and differentiation during gliogenesis in rat embryonic nerves. Neuron 12, 509--527. Johnston M. C. and Bronsky P. T. (1995) Prenatal craniofacial development: new insights on normal and abnormal mechanisms. Crit. Rev. oral BioL Med. 6, 25 79. Katz D. M., Adler J. E. and Black I. B. (1987) Catecholaminergic primary sensory neurons: autonomic targets and mechanisms of transmitter regulation. Fedn Proc. 46, 24~29.

264 44.

S.R. Hansson et al. Kim D. J. and Roper S. D. (1995) Localization of serotonin in taste buds: a comparative study in four vertebrates. ,L comp. Neurol. 353, 364-370.

45. 46. 47.

Kioussi C. and Gruss P. (1996) Making of a Schwann. Trends Genet. 12, 84-86. Kirby M. L. and Waldo K. L. (1995) Neural crest and cardiovascular patterning. Circulation Res. 77, 211-215. Lauder J. M. (1990) Ontogeny of the serotonergic system in the rat: serotonin as a developmental signal. Ann. N. E Acad. Sci. 600, 297-313. 48. Lauder J. M. (1993) Neurotransmitters as growth regulatory signals: role of receptors and second messengers. Trends Neurosci. 16, 233 240. 49. Lauder J. M. and Krebs H. (1978) Serotonin as a differentiation signal in early neurogenesis. Devl Neurosci. 1, 15 30. 50. Lauder J. M., Tamir H. and Sadler T. W. (1988) Serotonin and morphogenesis. I. Sites of serotonin uptake and binding protein immunoreactivity in the midgestation mouse embryo. Development 102, 709 720. 51. Lauder J. M., Wallace J. A. and Krebs H. (1981) Roles for serotonin in neuroembryogenesis. Adv. exp. Med. Biol. 133, 477 506. 52. Lauder J. M., Wallace J. A., Krebs H., Petrusz P. and McCarthy K. (1982) In vivo and in vitro development of serotonergic neurons. Brain Res. Bull. 9, 605-625. 53. Lauder J. M. and Zimmerman E. F. (1988) Sites of serotonin uptake in epithelia of the developing mouse palate, oral cavity, and face: possible role in morphogenesis. Z craniofac. Genet. devl Biol. 8, 265 276. 54. Lawrence I. E. Jr and Burden H. W. (1973) Catecholamines and morphogenesis of the chick neural tube and notochord. Am. Z Anat. 137, 199 207. 55. Lebrand C., Cases O., Adelbrecht C., Doye A., Alvarez C., El Mestikawy S., Seif I. and Gaspar P. (1996) Transient uptake and storage of serotonin in developing thalamic neurons. Neuron 17, 823-835. 56. Lesch K. P., Wolozin B. L., Estler H. C., Murphy D. L. and Riederer P. (1993) Isolation of a cDNA encoding the human brain serotonin transporter. J. neural Transm., Gen. Sect. 91, 67 72. 57. Lesch K. P., Wolozin B. L., Murphy D. L. and Reiderer P. (1993) Primary structure of the human platelet serotonin uptake site: identity with the brain serotonin transporter. J~ Neurochem. 60, 2319 2322. 58. Li Q., Brownfield M. S., Battaglia G., Cabrera T. M., Levy A. D., Rittenhouse P. A. and Van de Kar L. D. (1993) Long-term treatment with the antidepressants fluoxetine and desipramine potentiates endocrine responses to the serotonin agonists 6-chloro-2-[1-piperazinyl]-pyrazine (MK-212) and (+-)-l-(2,5-dimethoxy-4-iodophenyl)-2aminopropane HC1 (DOI). J. Pharmac. exp. Ther. 266, 83(~844. 59. Li Q., Levy A. D., Cabrera T. M., Brownfield M. S., Battaglia G. and Van de Kar L. D. (1993) Long-term fluoxetine, but not desipramine, inhibits the ACTH and oxytocin responses to the 5-HT1A agonist, 8-OH-DPAT, in male rats. Brain Res. 630, 148 156. 60. Lima L., Matus P. and Urbina M. (1994) Serotonin inhibits outgrowth of goldfish retina and impairs the trophic effect of taurine. J. Neurosci. Res. 38, 44~450. 61. Liu Q. and Debski E. A. (1993) Serotonin-like immunoreactivity in the adult and developing retina of the leopard frog Rana pipiens. J. comp. Neurol. 338, 391404. 62. Lucas J. J. and Hen R. (1995) New players in the 5-HT receptor field: genes and knockouts. Trendspharmac. Sci. 16, 246 252. 63. Mirmiran M. and Swaab D. F. (1986) Central neurotransmitter disturbances underlying developmental neurotoxicological effects. Neurotoxicology 7, 95-102. 64. Mirsky R., Stewart H. J., Tabernero A., Bradke F., Brennan A., Dong Z. and Jessen K. R. (1996) Development and differentiation of Schwann cells. Rev. Neurol., Paris 152, 308 313. 65. Moiseiwitsch J. R. and Lauder J. M. (1995) Serotonin regulates mouse cranial neural crest migration. Proc. natn. Acad. Sci. U.S.A. 92, 7182 7186. 66. Moiseiwitsch J. R. and Lauder J. M. (1996) Stimulation of murine tooth development in organotypic culture by the neurotransmitter serotonin. Archs oral Biol. 41, 161-165. 67. Moiseiwitsch J. R. and Lauder J. M. (1997) Regulation of gene expression in cultured embryonic mouse mandibular mesenchyme by serotonin antagonists. Anat. Embryol., Berlin 195, 71 78. 68. Ni Y. G. and Miledi R. (1997) Blockage of 5-HT2C serotonin receptors by fluoxetine (Prozac). Proc. natn. Acad. Sci. U.S.A. 94, 2036-2040. 69. Nozue A. T. (1988) Relationships between neural crest cells and serotonin in newborn mice. Anat. Anz. 166, 227 237. 70. Nozue A. T. and Ono S. (1991) Effects of catecholamine and serotonin in central nervous system in newborn mice with special reference to neural crest cells; presumptive evidence of neural crest origin. Anat. Anz. 173, 147-153. 71. Nozue T. (1973) Neural crest cells in early tooth germ of mice embryos. Okajimas Folia Anat. Jap. 50, 217-229. 72. Nozue T. and Kayano T. (1978) Effects of mitomycin C in postnatal tooth development in mice with special reference to neural crest cells. Acta anat., Basel 100, 85-94. 73. Nurse C. A. and Diamond J. (1984) A fluorescent microscopic study of the development of rat touch domes and their Merkel cells. Neuroscience 11, 509-520. 74. Nurse C. A. and Farraway L. (1989) Characterization of Merkel cells and mechanosensory axons of the rat by styryl pyridinium dyes. Cell Tiss. Res. 255, 125 128. 75. Nurse C. A., Mearow K. M., Holmes M., Visheau B. and Diamond J. (1983) Merkel cell distribution in the epidermis as determined by quinacrine fluorescence. Cell Tiss. Res. 228, 511-524. 76. Pastuszak A., Schick-Boschetto B., Zuber C., Feldkamp M., Pinelli M., Sihn S., Donnenfeld A., McCormack M., Leen-Mitchell M. and Woodland C. (1993) Pregnancy outcome following first-trimester exposure to fluoxetine (Prozac). J. Am. med. Ass. 269, 2246-2248. 77. Prasad P. D., Hoffmans B. J., Moe A. J., Smith C. H., Leibach F. H. and Ganapathy V. (1996) Functional expression of the plasma membrane serotonin transporter but not the vesicular monoamine transporter in human placental trophoblasts and choriocarcinoma cells. Placenta 17, 201 207. 78. Ramamoorthy S., Bauman A. L., Moore K. R., Han H., Yang-Feng T., Chang A. S., Ganapathy V. and Blakely R. D. (1993) Antidepressant- and cocaine-sensitive human serotonin transporter: molecular cloning, expression, and chromosomal localization. Proc. natn. Acad. Sci. U.S.A. 90, 2542-2546. 79. Rhoades R. W., Chiaia N. L., Bennett-Clarke C. A., Janas G. J. and Fisher C. M. (1994) Alterations in brainstem and cortical organization of rats sustaining prenatal vibrissa follicle lesions. Somatosensory Motor Res. 11, 1-17.

Ontogeny of 5-HT transporter mRNA expression 80. 81. 82. 83. 84. 85. 86. 87.

88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114.

265

Robson J. M. and Senior J. B. (1964) The 5-hydroxytryptamine content of the placenta and foetus during pregnancy in mice. Br. J. Pharmac. 22, 38~391. Rosafi D., Nurse C. A. and Diamond J. (1984) Lectin-binding properties of the Merkel cell and other root sheath cells in perinatal rat vibrissae. Cell Tiss. Res. 236, 373-381. Schmidt M. J. and Sanders-Bush E. (1971) Tryptophan hydroxylase activity in developing rat brain. J. Neurochem. 18, 2549-2551. Schroeter S. and Blakely R. D. (1996) Drug targets in the embryo. Studies on the cocaine- and antidepressantsensitive serotonin transporter. Ann. N. Y. Acad. Sci. 801,239 255. Schroeter S., Levey A. I. and Blakely R. D. (1997) Polarized expression of the antidepressant-sensitive serotonin transporter in epinephrine-synthesizing chromaffin cells of the rat adrenal gland. Molec. cell. Neurosci. 9, 170 184. Seuwen K. and Pouyssegur J. (1990) Serotonin as a growth factor. Biochem. Pharmac. 39, 985-990. Shuey D. L., Sadler T. W. and Lauder J. M. (1992) Serotonin as a regulator of craniofacial morphogenesis: site specific malformations following exposure to serotonin uptake inhibitors. Teratology 46, 367-378. Shuey D. L., Sadler T. W., Tamir H. and Lauder J. M. (1993) Serotonin and morphogenesis. Transient expression of serotonin uptake and binding protein during craniofacial morphogenesis in the mouse. Anat. Embryol.. Berlin 187, 75 85. Shuey D. L., Yavarone M., Sadler T. W. and Lauder J. M. (1990) Serotonin and morphogenesis in the cultured mouse embryo. In Molecular Aspects of Development and Aging of the Nervous System (ed. Lauder J. M.), pp. 205-214. Plenum, New York. Siebe)-Blum M. (1989) Commitment of neural crest cells to the sensory neuron lineage. Science 243, 1608 1611. Sodha R. J., Proegler M. and Schneider H. (1984) Transfer and metabolism of norepinephrine studied from maternal-to-fetal and fetal-to-maternal sides in the in vitro perfused human placental lobe. Am. J. Obstet. Gynecol. 148, ,,74-481. State F. A., el-Eishi H. I. and Naga I. A. (1974) The development of taste buds in the foliate papillae of the albino rat. Aeta anat.. Basel 89, 452-460. Tamir H., Hsiung S. C., Liu K. P., Blakely R. D., Russo A. F., Clark M. S., Nunez E. A. and Gershon M. D. (1996) Expression and development of a functional plasmalemmal 5-hydroxytryptamine transporter by thyroid follicular cells. Endocrinology 137, 4475-4486. Tecott L., Shtrom S. and Julius D. (1995) Expression of a serotonin-gated ion channel in embryonic neural and nonneural tissues. Molec. cell. Neurosei. 6, 43 55. Thompson A. M. and Thompson G. C. (1995) Light microscopic evidence of serotoninergic projections to olivocochlear neurons in the bush baby (Otolemur garnettii). Brain Res. 695, 263-266. Verbeuren T. J. (1989) Synthesis, storage, release, and metabolism of 5-hydroxytryptamine in peripheral tissues. In The Peripheral Actions of5-Hydroxytryptamine (ed. Fozard J. R.), pp. 1 25. Oxford University Press, New York. Volpe J. J. (1992) Effect of cocaine use on the fetus. New Engl. J. Med. 327, 399-407. Vos P., Stark F. and Pittman R. N. (1991) Merkel cells in vitro: production of nerve growth factor and selective interactions with sensory neurons. Devl Biol. 144, 281 300. Wade P. R., Chen J., Jaffe B., Kassem I. S., Blakely R. D. and Gershon M. D. (1996) Localization and function of a 5-HT transporter in crypt epithelia of the gastrointestinal tract. J. Neurosci. 16, 2352-2364. Waldo K. L., Kumiski D. H. and Kirby M. L. (1994) Association of the cardiac neural crest with development of the coronary arteries in the chick embryo. Anat. Rec. 239, 315-331. Wallace J. A. (1988) Neurotransmitters and early embryogenesis. In Amino Acid Availability and Brain Function in Healt,~ and Disease (ed. Huether G.), pp. 431-439. Springer, Berlin. Wan G Z. Z., Stensaas L. J., Dinger B. and Fidone S. J. (1992) The co-existence of biogenic amines and neuropeptides in the type I cells of the cat carotid body. Neuroscience 47, 473-480. Wapnir R. A., Hawkins R. L. and Stevenson J. H. (1971) Ontogenesis of phenylalanine and tryptophan hydroxylation in rat brain and liver. Biol. Neonate 18, 85 93. Weihe E., Schafer M. K., Erickson J. D. and Eiden L. E. (1994) Localization of vesicular monoamine transporter isoforms (VMAT1 and VMAT2) to endocrine cells and neurons in rat. J. molec. Neurosci. 5, 149 164. Whitaker-Azmitia P. M. and Azmitia E. C. (1989) Stimulation of astroglial serotonin receptors produces culture media which regulates growth of serotonergic neurons. Brain Res. 497, 80-85. Whitaker-Azmitia P. M., Murphy R. and Azmitia E. C. (1990) Stimulation of astroglial 5-HT1A receptors releases the serotonergic growth factor, protein S-100, and alters astroglial morphology. Brain Res. 528, 155-158. Whitear M. (1989) Merkel cells in lower vertebrates. Archs Histol. Cytol. 52, 415-422. Wichems C. H., Andrews A. M., Bengel D., Lesch K. P. and Murphy D. L. (1997) Functional consequences of a disrupted serotonin transporter: pharmacological challenges and the 5-HTIA receptor. Soc. Neurosci. Abstr. 23, 980. Yan W., Wilson C. C. and Haring J. H. (1997) 5-HTla receptors mediate the neurotrophic effect of serotonin on developing dentate granule cells. Brain Res. devl Brain Res. 98, 185 190. Yah W., Wilson C. C. and Hating J. H. (1997) Effects of neonatal serotonin depletion on the development of rat dentate granule cells. Brain Res. devl Brain Res. 98, 177 184. Yan W., Young J. L. and Haring J. H. (1995) Evidence for the elaboration of multiple axons by developing dentate granule cells. Brain Res. devl Brain Res. 90, 194-197. Yavarone M. S., Shuey D. L., Sadler T. W. and Lauder J. M. (1993) Serotonin uptake in the ectoplacental cone and placenta of the mouse. Placenta 14, 149 161. Yavarone M. S., Shuey D. L., Tamir H., Sadler T. W. and Lauder J. M. (1993) Serotonin and cardiac morphogenesis in the mouse embryo. Teratology 47, 573 584. Zhu B. and Straznicky C. (1992) Large serotonin-like immunoreactive amacrine cells in the retina of developing Xenopus laevis. Brain Res. devl Brain Res. 69, 109 116. Zimmerman E. F. (1985) Role of neurotransmitters in palate development and teratologic implications. Prog. clin. biol. Res. 171,283 294. (Accepted 5 May 1998)