5-HT_2A Receptors are Concentrated in Regions of the Human Infant Medulla Involved in Respiratory and Autonomic Control

Abstract

The serotonergic (5-HT) system in the human medulla oblongata is well-recognized to play an important role in the regulation of respiratory and autonomic function. In this study, using both immunocytochemistry (n=5) and tissue section autoradiography with the radioligand ¹²⁵I-1-(2,5-dimethoxy-4-iodo-phenyl)2-aminopropane (n=7), we examine the normative development and distribution of the 5-HT_2A receptor in the human medulla during the last part of gestation and first postnatal year when dramatic changes are known to occur in respiratory and autonomic control, in part mediated by the 5-HT_2A receptor. High 5-HT_2A receptor binding was observed in the dorsal motor nucleus of the vagus (preganglionic parasympathetic output) and hypoglossal nucleus (airway patency); intermediate binding was present in the nucleus of the solitary tract (visceral sensory input), gigantocellularis, intermediate reticular zone, and paragigantocellularis lateralis. Negligible binding was present in the raphé obscurus and arcuate nucleus. The pattern of 5-HT_2A immunoreactivity paralleled that of binding density. By 15 gestational weeks, the relative distribution of the 5-HT_2A receptor was similar to that in infancy. In all nuclei sampled, 5-HT_2A receptor binding increased with age, with significant increases in the hypoglossal nucleus (p=0.027), principal inferior olive (p=0.044), and medial accessory olive (0.038). Thus, 5-HT_2A receptors are concentrated in regions involved in autonomic and respiratory control in the human infant medulla, and their developmental profile changes over the first year of life in the hypoglossal nucleus critical to airway patency and the inferior olivary complex essential to cerebellar function.

INTRODUCTION

Serotonergic (5-HT) neurons in the medulla oblongata comprise a critical system involved in the modulation of autonomic and respiratory effector neurons in a state-dependent manner (Azmitia, 1999; Kinney et al., 2007; Kinney et al., 2001; Lovick, 1997; Mason, 2001; Morrison, 2001). These 5-HT neurons are organized in midline (raphé), lateral (extra-raphé), and ventral surface regions such that they virtually encircle and project to the adjacent tegmental nuclei in the medulla that mediate respiration, upper airway patency, heart rate, blood pressure, thermoregulation, and arousal (Kinney et al., 2007). The specific effects of 5-HT upon the autonomic and respiratory effector nuclei are mediated via different 5-HT receptor subtypes, of which the most information relevant to homeostasis is available for the 5-HT_1A, 5-HT_2A, and 5-HT₃ receptors (Cayetanot et al., 2002; Darnall et al., 2005; Helke et al., 1997; Helke et al., 1993; King and McCall, 1991; Knapp et al., 1998; Kubo et al., 1995; Lalley, 1994; Lalley et al., 1994; Li et al., 1999; Ling et al., 2001; Lovick, 1989; Nosjean and Guyenet, 1991; Onimaru et al., 1998; Paterson et al., 2006a; Pena and Ramirez, 2002; Penatti et al., 2006; Rose et al., 1995; Schwarzacher et al., 2002; Talley et al., 1997; Teng et al., 2003; Tryba et al., 2006; Wilken et al., 1997). There is increasing interest in the medullary 5-HT system in human pediatric neuropathology because of the evidence for 5-HT dysfunction in major developmental brainstem disorders, e.g., sudden infant death syndrome (SIDS) (Kinney et al., 2001; Kinney et al., 2005; Kinney et al., 2003; Machaalani et al., 2008; Ozawa and Okado, 2002; Panigrahy et al., 2000; Paterson et al., 2006b) and Rett syndrome (Paterson et al., 2005). Yet, little is known about the normative development of the different 5-HT receptor subtypes in the human brainstem—information that is essential towards defining 5-HT receptor pathology in pediatric disorders.

Previously we described the developmental distribution of the 5-HT_1A receptor in the human infant medulla (Paterson et al., 2004). In the following study, we examine the normative development and distribution of the 5-HT_2A receptor in the human medulla during the same developmental period. We chose to focus upon the 5-HT_2A receptor because extensive experimental analysis indicates a critical role for the 5-HT_2A receptor in respiration governed by the preBötzinger complex (Pena and Ramirez, 2002; Ramirez et al., 2004), gasping (Tryba et al., 2006), central cardiovascular regulation including the baroreceptor reflex (Comet et al., 2007; Dergacheva et al., 2007; Raul, 2003; Shen et al., 2007; Villalon and Centurion, 2007), the development of the medullary and spinal cord respiratory network (Belzile et al., 2002; Bou-Flores and Hilaire, 2000; Cayetanot et al., 2002; Kinkead et al., 2002), upper airway control (Cornea-Hebert et al., 1999; Fonseca et al., 2001; Ogasa et al., 2004; Zhan et al., 2002), and 5-HT-mediated adaption to intermittent hypoxia (Bach and Mitchell, 1996; Baker-Herman et al., 2004; Mitchell et al., 2001). In this study, we used tissue section autoradiography with ¹²⁵I-1-(2,5-dimethoxy-4-iodo-phenyl)2-aminopropane (¹²⁵I DOI) to determine quantitatively changes in the density of 5-HT_2A receptors in the human medulla during infancy and a combination of single-and double-label immunocytochemistry to determine the cellular localization of 5-HT_2A receptors in the medulla.

MATERIALS AND METHODS

Clinical Cases

Infant brainstem specimens were accrued from the autopsy services of the Department of Pathology, Children’s Hospital Boston, MA, and the office of the Chief Medical Examiner, San Diego, CA. This study was approved by the Committee on Clinical Investigation at Children’s Hospital Boston. For tissue receptor autoradiography, we examined the medullae of 7 infants. The cases ranged in age from 39 postconceptional weeks (term birth) to 82 postconceptional weeks (approximately 10 postnatal months), with a median age of 5 months. Fetal cases were not available. These infant cases had a median postmortem interval of 15 hours with a range of 8–20 hours. The causes of death were: congenital heart disease (n=3), acute pneumonia (n=1), intestinal obstruction (n=1), acute asphyxia (n=1), and Potter’s syndrome (n=1). The brainstems did not demonstrate pathologic changes by standard criteria. For immunocytochemical analysis, we examined formalin-fixed, paraffin-embedded sections of the medulla of 5 fetuses and infants ranging in age from 15 gestational weeks through 10 postnatal months, with a median age of 38 weeks (term birth). The median postmortem interval was 14 hours. The causes of death in this autopsy population were: extreme prematurity (n=2), congenital neuroblastoma (n=1), congenital heart disease (n=1), and pulmonary veno-occlusive disease (n=1). Again, in no case was there brainstem pathology by standard histopathologic criteria.

5-HT_2A Receptor Autoradiography

Tissue preparation

Our procedures for tissue preparation have been described in detail previously (Panigrahy et al., 2000; Paterson et al., 2004; Paterson et al., 2006b). Briefly, unfixed brainstems were stored frozen at −80°C, and subsequently sectioned at 20 μm on a Leitz motorized cryostat and thaw-mounted onto Superfrost Plus glass microscope slides (Thermo Fisher Sceintfic, NH).

¹²⁵I-DOI Binding to 5-HT_2A Receptors

The autoradiography procedure for determination of ¹²⁵I-DOI binding to 5-HT_2A receptors was performed on 20 μm sections of frozen medulla according to a previously described protocol (Lopez-Gimenez et al., 2002). Sections were preincubated in 50mM Tris-HCl (pH 7.4), 0.1% ascorbic acid, and 4mM CaCl₂ for 30 minutes at room temperature followed by incubation in the same buffer containing 86.3pM ¹²⁵I DOI (PerkinElmer Inc, Wellesley, Mass) for 60 minutes at room temperature to determine total 5-HT_2A receptor binding density. Nonspecific binding was determined by addition of 10μM ritanserin to the incubation solution. Sections were then washed 2 × 10 minutes in ice cold buffer and dried in warm air before being placed in cassettes and exposed to ¹²⁵I-sensitive film (Kodak BMR) for 42 hours along with a set of ¹²⁵I standards (Amersham) for conversion of optical density of silver grains to fmol/mg tissue.

Quantitative Analysis of brainstem autoradiograms

Film autoradiograms were generated according to standard laboratory procedure for development of light-sensitive film. For each specimen receptor binding density (expressed as the specific activity of tissue-bound ligand in fmol/mg protein) was analyzed in a total of eleven medullary nuclei, including the raphé obscurus (Rob), nucleus of the solitary tract (NTS), dorsal motor nucleus of the vagus (DMX), hypoglossal nucleus (HG), intermediate reticular zone (IRZ), gigantocellularis nucleus (GC), paragigantocellularis lateralis nucleus (PGCL), dorsal accessory olive (DAO), principle inferior olive (PIO), medial accessory olive (MAO) and arcuate nucleus (Arc) at a defined level of the brainstem according to the atlas of Olszewski and Baxter (1954) with the exception of the raphé nuclei, which were classified according to Tork and Hornung (1990). Total receptor binding was determined in 2 sections and non-specific receptor binding in 1 section for each nucleus analyzed. Specific receptor binding density was determined by subtracting nonspecific binding from total binding. Quantitative densitometry of autoradiograms was performed using an MCID 5+ imaging system (Imaging Research, Ontario, Canada). Linear regression analysis was used to determine the effect of age on ¹²⁵I-DOI binding across infancy. To examine a possible influence on postmortem interval (PMI) on binding density, linear regression analysis of binding (fmol/mg) versus PMI was also performed.

Photomicrograph production

Images of 5-HT_2A receptor binding in the human infant medulla were generated as TIFF files from the autoradiography film by the MCID 5+ imaging system (Imaging Research). All images were then imported into PhotoShop 6.0 (Adobe Systems, San Jose, CA) where they were scaled relative to each other and appropriate labels were added to form composite images.

Receptor Immunocytochemistry

Single-label immunocytochemistry for 5-HT_2Areceptors

Immunocytochemistry for 5-HT_2A receptors was performed using a goat polyclonal antibody SR2A A-15 (sc-15073, Santa Cruz Biotechnology Inc., Santa Cruz, CA) raised against amino acids 10–60 [protein accession number P28223] near the n-terminus of the human 5-HT_2A receptor that has previously been reported to label 5-HT_2A receptors in human tissue (Vikman and Edvinsson, 2006). Four micron thick formalin-fixed paraffin embedded sections of human infant medulla were de-paraffinzed before antigen retrieval was performed by incubating the tissue in a 1× citrate buffer solution (pH 6.0) in a microwave oven at 195°C for 10 min. Sections were allowed to cool for 10 min at room temperature before being washed in running water for 10 min. Sections were then washed 4×15 min in 1× phosphate buffered saline (PBS) with 0.1% Triton-X and treated for 30 min with 3% hydrogen peroxide. The sections were then blocked in 4% bovine serum albumin (BSA) for 60 min before being incubated in primary antibody (1:100) overnight at 4°C. Sections were then washed 3×20 min in 1xPBS with 0.1% Triton-X before being incubated in a biotinylated secondary antibody (Vectastain Elite Goat ABC Kit) for 60 min, rinsed briefly in PBS, and incubated in avidin-horseradish peroxidase solution (Vectastain Elite Goat ABC Kit, Vector Laboratories, Burlingame, CA) for 10 min. Staining was then developed by the addition of di-amino benzadine substrate (Vector Laboratories, Burlingame, CA) to the section for 1–5 min. DAB Enhancer solution (Zymed Laboratories) was then applied for 1–2 min to enhance immunostaining. Sections were then washed in running water before dehydration in a series of alcohols (80%, 90%, and 100%) cleared in xylene and coverslipped with permount. To determine the specificity of staining with the antibody, control sections were processed as above with omission of the primary and/or secondary antibody, which resulted in no staining. Pre-immune absorption was also performed by incubating the primary antibody in solutions containing 2, 4 and 8μg/ml of blocking peptide (sc-15073 P, Santa Cruz Biotechnology Inc., Santa Cruz, CA) before processing tissue sections as normal. No immunostaining for 5-HT_2A receptors was observed in these sections. Moreover, immunostaining was absent in brain tissue sections from 5-HT_2A receptor knockout mice that do not express the 5-HT_2A receptor protein. Each case had at least three positively stained sections at the selected mid and rostral medullary levels analyzed. Immunostained sections were visualized with an Olympus BX51 microscope (Olympus America Inc., Melville, NY) with image capture using an Optronics Microfire S99808 camera and Microfire 1.0 and Neurolucida 5.0 software (Microbrightfield, Colchester, VT). Images were captured as TIFF files and imported into Photoshop 6.0 (Adobe Systems, San Jose, CA) where they were scaled relative to each other and appropriate labels and scale bars were added and images were optimized to enhance clarity and contrast.

Double-label immunocytochemistry

Double-label immunofluorescence was performed to determine the distribution and cellular localization of 5-HT_2A receptors relative to 5-HT neurons on 4 μm 4% paraformaldehyde fixed sections of medulla using the same 5-HT_2A goat polyclonal antibody used for single labeling and mouse-antihuman PH8 antibody (MAB5278, Millipore International, Temecula, CA) against tryptophan hydroxylase (TPOH) to identify 5-HT neurons. Tryptophan hydroxylase is the rate limiting enzyme in 5-HT synthesis and is a specific marker for 5-HT neurons labeling 5-HT cell bodies, fibers, and terminals. We have used this antibody to identify medullary 5-HT neurons in previous studies (Kinney et al., 2007; Paterson et al., 2006b). Double-label immunofluorescence was performed following the protocol for single-label immunocytochemistry except that the hydrogen peroxide blocking step was omitted. Sections were incubated simultaneously in 5-HT_2A goat polyclonal antibody (1:100) and PH8 mouse monoclonal antibody (1:8,000) overnight at 4 C. Sections were then washed 3×20 min in 1×PBS with 0.1% Triton-X before being incubated in Alex-Fluor 594 donkey anti-goat (A-11058, Invitrogen, Carlsbad CA) and Cy2 donkey anti-mouse (715-225-150, Jackson Immunoresearch Laboratories, West Grove, PA) fluorescent secondary antibodies for 1 hour at room temperature. Sections were then washed 3 × 20 min in 1XPBS with 0.1% Triton-X, and allowed to dry for 60 min at room temperature before being cover slipped in Fluoromount-G (Southern Bio-ctechnology). Omission of the primary antibody was used as a negative control. Immunofluorescence was visualized with an Olympus BX51 microscope (Olympus America Inc., Melville, NY) using FITC and TRITC filters with image capture using a Coolsnap fx camera (Photometrics, Tuscon, AZ) and MCID Elite 6.0 software (Imaging Analysis Inc., Ontario, Canada). Images were captured as TIFF files and imported into Photoshop 6.0 (Adobe Systems, San Jose, CA) where they were corrected for background immunofluorescence, scaled relative to each other, and appropriate labels and scale bars were added.

Analysis of the human infant medulla

The medullary 5-HT system, as defined by us, consists of the 5-HT neurons in the midline raphé (raphé obscurus, raphé pallidus, and raphé magnus), lateral extra-raphé (gigantocellularis, paragigantocellularis lateralis, intermediate reticular zone, subtrigeminalis, and lateral reticular nucleus), ventral surface (arcuate nucleus) and the regions in the medulla receiving major projections from these 5-HT neurons that are involved in autonomic and respiratory control including the hypoglossal nucleus (upper respiratory control), preBötC (respiratory rhythm genesis); human homologue in the PGCL (Gray et al., 1999; Rekling and Feldman, 1998; Smith et al., 1991), DMX (preganglionic parasympathetic output) and NTS (visceral sensory input). These nuclei are distributed from the caudal medulla (at the level of the area postrema) to rostral medulla (at the level of the paragigantocellularis lateralis), according to the standardized brainstem levels previously defined in our laboratory (Kinney et al., 2007; Kinney et al., 2001) based upon the human brainstem atlases of Olszewski and Baxter, (1954), Paxinos and Huang, (1995), and Tork and Hornung, (1990). These same standardized levels were analyzed in this study to determine the distribution of 5-HT_2A receptors in the human infant medulla using autoradiography and immunocytochemistry.

RESULTS

¹²⁵I DOI binding to 5-HT_2A receptors in the human infant medulla

The highest density of ¹²⁵I DOI binding was observed in the DMX (1.41 ± 0.25 fmol/mg) (Figs. 1 and 2). High binding was also present in the HG and the DAO. Intermediate levels of binding were observed in the NTS (0.70 ± 0.12 fmol/mg) and in the extra-raphé regions containing 5-HT neuron cell bodies including the GC, PGCL, and IRZ. In the olivo-cerebellar system, there was a differential distribution of binding intensity, with the highest level in the DAO (1.39 ± 0.20 fmol/mg), similar to that seen in the DMX (Figs. 1 and 2). There was intermediate binding in the MAO and PIO (Figs. 1 and 2). Negligible binding was observed in the Rob and in the Arc. Postmortem interval had no significant effect on binding.

Figure 1.

Autoradiographic images of ¹²⁵I DOI binding to 5-HT_2A receptors in transverse sections of the caudal (A) and rostral (B) human infant medulla. High 5-HT_2A receptor binding density is present in the DMX and HG, moderate binding density in the NTS and PGCL, and relatively low binding density in the ROB and Arc.

Figure 2.

Bar graph displaying 5-HT_2A receptor binding density in medullary nuclei presented from the highest to lowest. 5-HT_2A receptor binding density is highest in medullary nuclei without 5-HT neurons (e.g., DMX, HG).

Distribution of 5-HT_2A receptor Immunostaining in the human infant medulla

The pattern of 5-HT_2A immunoreactivity in medullary nuclei in the infant cases generally paralleled the distribution of 5-HT_2A receptor binding density observed with ¹²⁵I DOI. Intensely stained immunoreactive neurons were observed in the DMX and in motor neurons in the HG (Fig 3C and D). Intensely stained large, round neurons, were also observed in all subdivisions of the inferior olivary system (i.e., PIO, MAO, DAO) (Fig 3E). Moderately stained neurons of heterogeneous size and morphology were observed in the NTS and scattered throughout the extra-raphé regions (GC, PGCL, IRZ) (Fig 3F). Moreover, a group of small spherical 5-HT_2A immunoreactive neurons were frequently observed in the lateral extent of the PGCL (Fig 3G). 5-HT_2A immunoreactive neurons were also observed in the raphé obscurus and in the arcuate nucleus, both sites in which receptor binding to the radioligand ¹²⁵I-DOI was negligible. Immunostaining in the raphé obscurus was light and localized to spherical and fusiform neurons in the midline (Fig 3H). In the arcuate nucleus immunostaining was intense and localized to small, irregularly-shaped cells, consistent in morphology to astrocytes and to larger spherical neurons the major cell type in this region (Fig 3I). In virtually all immunopositive regions, the immunostaining was punctate and co-localized with cell bodies and/or processes, consistent with post-synaptic localization of the receptors.

Figure 3.

5-HT_2A receptor immunostaining in the human infant medulla at 10 postnatal months. Panels (A) and (B) show diagrams of horizontal sections of rostral and mid-medulla, respectively, showing the location of the component nuclei of the medullary 5-HT system at each level. 5-HT_2A receptor immunostaining is punctate and localized to neuronal cell bodies and processes, indicative of post-synaptic localization of the receptors. Figure shows intense immunostaining of neurons in the DMX (C), motor neurons in the HG (D), and spherical neurons in the PIO (E). Moderately stained neurons of heterogenous morphology were observed in the extra-raphé regions including the GC (F). Spherical 5-HT_2A immunoreactive neurons were observed in the lateral extent of the PGCL (G) consistent in location and morphology with preBötC neurons in rodents. Lightly stained spherical and fusiform 5-HT_2A receptor immunoreactive neurons were observed in the midline raphé (H). Midline of the medulla is labeled by double headed arrow; D, dorsal; V, ventral. Immunostaining to “classical” arcuate neurons (arrows) and astrocytes (arrow-heads) in the arcuate nucleus (I). VMS, ventral medullary surface. All images at ×40. Scale bar= 100 μm.

Distribution of 5-HT_2A receptors relative to 5-HT neurons in the human infant medulla

We observed 5-HT neurons and fibers with TPOH immunofluorescence in the raphé, extra-raphé (GC, PGCL, and IRZ) and Arc at the ventral medullary surface, consistent with the distribution of 5-HT neurons previously described by us in the human infant medulla (Kinney et al., 2007; Paterson et al., 2006b). Double-label immunofluorescence revealed that 5-HT_2A receptors co-localized to the soma and dendrites of a subset of 5-HT (TPOH) immunoreactive neurons in the raphé and extra-raphé regions of the human infant medulla (Fig 4). However, not all 5-HT neurons expressed 5-HT_2A receptors, and not all neurons expressing 5-HT_2A receptor immunoreactivity were 5-HT (Fig 4.)

Figure 4.

Double-label immunofluorescent images showing localization of 5-HT_2A receptors and 5-HT neurons in the PGCL (×40). A. 5-HT_2A immunofluorescent staining; B. TPOH immunofluorescent staining; C. Merged images. 5-HT_2A receptors co-localized to the soma and dendrites of a subset of 5-HT immunoreactive neurons (arrows) in the medulla, but not all 5-HT neurons expressed 5-HT_2A receptors (arrowheads), and not all neurons expressing 5-HT_2A receptor immunoreactivity were 5-HT (*). ×40. Scale bar=100μm.

Developmental profile of 5-HT_2A expression in the human infant medullary 5-HT system

The developmental profile of 5-HT_2A receptor binding was analyzed across infancy to determine if there were any age-related changes in binding density in the component nuclei of the medullary 5-HT system. A trend for 5-HT_2A receptor binding density to increase with age was observed in all medullary nuclei examined, with statistically significant increases observed in the HG (p=0.027), PIO (p=0.044) and MAO (p=0.038) (Figs 5A-C). 5-HT_2A receptor immunostaining was present in the developing medulla as early as 15 gestational weeks with the distribution of receptors already “set” in the mature configuration at this age. Immunostaining was most intense in the HG (Fig 6A) and in the lateral extra-raphé where 5-HT_2A immunoreactive neurons were tightly clustered (Fig 6B). In all regions of the medulla at 15 weeks neurons were spherical and undifferentiated with few or no few immunoreactive processes observed (Fig 6A). At 22 weeks gestation 5-HT_2A receptor immunoreactivity was observed in distinct nuclei including the DMX (Fig 6C), PIO (Fig 6D) and arcuate nucleus (Fig 6E). By term (41 gestational weeks), the neurons had increased in size, developed immunoreactive processes and differentiated into distinct morphological subtypes, e.g., motorneurons in the HG (Fig 6F). By 10 postnatal months 5-HT_2A immunoreactive neurons of heterogeneous morphology were observed in different medullary regions, e.g., motorneurons in the HG (Fig 3B) and fusiform neurons in the in raphè and extra-raphé (Fig 3F). No obvious changes in the distribution or intensity of 5-HT_2A receptor immunostaining were observed from late gestation through infancy.

Figure 5.

Linear regression plots of the relationship between 5-HT_2A receptor binding density and postconceptional age (weeks) in HG (A), PIO (B) and MAO (C). In each nuclei, receptor binding density increases significantly with age. *p<0.05 linear regression analysis.

Figure 6.

Developmental expression of 5-HT_2A receptor immunocytochemistry in the human medulla from 15 gestational weeks to term (41 gestational weeks). Immunostaining in the HG (A) and extra-raphé (B) (i.e., GC/PGCL) at 15 weeks gestation. Neurons in the extra-raphé were tightly clustered together with no obvious distinction between nuclei. Neurons are spherical and undifferentiated. At 22 weeks gestation 5-HT_2A receptor immunoreactivity was observed in distinct nuclei including the DMX (C), PIO (D) and arcuate nucleus (E) with evidence of neuronal differentiation. Motor neurons in the HG expressing 5-HT_2A receptor immunoreactivity at term (41 weeks gestation) (F). All images at ×40 (scale bar=100μm) except D at ×20 (scale bar=200μm). VMS, ventral medullary surface.

DISCUSSION

In the present study, we demonstrated that 5-HT_2A receptors are expressed in high density in medullary nuclei that are critical for autonomic and respiratory control. We identified high 5-HT_2A receptor binding density in the DMX (preganglionic parasympathetic output), and HG (airway patency), and intermediate binding in the NTS (visceral sensory input), GC, PGCL and IRZ. In addition, 5-HT_2A receptors are expressed in high density by neurons in the inferior olivary complex. The olivary complex is involved in somatomotor coordination (Azizi and Woodward, 1987; Ikeda et al., 1989; Nisimaru et al., 1991; Sugita et al., 1989; Urbano et al., 2006), but also has an underappreciated role in cardiorespiration via its connections with the cerebellum (Harper et al., 2000; Nisimaru et al., 1991; Nisimaru et al., 1998; Okahara and Nisimaru, 1991; Sugita et al., 1989). These observations suggest that, as in experimental animals, medullary 5-HT_2A receptors play important roles in 5-HT mediated regulation of autonomic and respiratory function. We also found that 5-HT_2A receptors are expressed in medullary cardiorespiratory-related nuclei as early as 15 gestational weeks, the earliest time-period studied by us, and that there are significant changes in 5-HT_2A receptor binding in the HG and the inferior olivary complex after birth. These observations suggest that the medullary 5-HT system is not fully developed at birth, but continues to mature throughout the first postnatal year. Below, we discuss these observations in relation to the roles of medullary 5-HT_2A receptors in putatively mediating respiratory and autonomic function in the human infant. We begin by considering the distribution of 5-HT_2A receptors in the human infant medulla.

Distribution of 5-HT_2A Receptors in Human Infant Medulla

In this study we observed that 5-HT_2A receptors are concentrated in high density in the HG, DMX and NTS, moderate to low density in the lateral extra-raphé regions (i.e., GC and PGCL), the inferior olivary complex, and in the Arc and in negligible density in the Rob. This distribution is consistent with previous studies describing 5-HT_2A receptors in the human brainstem (Burnet et al., 1995; Hall et al., 2000; Ozawa and Okado, 2002; Ozawa and Takashima, 2002) and with the distribution of 5-HT_2A receptors in the medulla in rats (Brandes et al., 2007; Cornea-Hebert et al., 1999; Fonseca et al., 2001; Haan et al., 1987; Huang and Pickel, 2002; Huang and Pickel, 2003; Jansson et al., 1998; Okabe et al., 1997; Wright et al., 1995). These observations suggest, that as in rats, 5-HT_2A receptors in the human infant medulla are concentrated in nuclei that receive significant innervation from 5-HT neurons in the caudal raphé and extra-raphé (Barnes and Sharp, 1999; Hamada et al., 1998; Mengod et al., 1996; Xu and Pandey, 2000), and thus, are predominantly postsynaptic to 5-HT axon terminals. This idea is supported by our previous demonstration of high density serotonin transporter (5-HTT) binding, a marker of 5-HT presynaptic terminals, in the DMX, NTS and HG (Paterson et al., 2004; Paterson et al., 2006b). The distribution of 5-HT_2A receptors in the human infant medulla also contrasts distinctly to the distribution of 5-HT_1A receptors: 5-HT_1A binding density is highest in the Rob (Paterson et al., 2004; Paterson et al., 2006a), 5-HT_2A binding is negligible in this region; similarly, 5-HT_1A receptors co-localize extensively with 5-HT neuronal cell bodies in the human infant medulla (Paterson et al., 2006a); Paterson et al., unpublished observations), while 5-HT_2A receptors localize predominantly to non-5-HT neurons. We propose that the differential distribution of 5-HT_1A and 5-HT_2A receptors underscores their different functional roles in the human infant medulla, i.e., 5-HT_1A somato-dendritic autoreceptor modulating 5-HT neuron function versus 5-HT_2A post-synaptic excitatory receptor mediating autonomic and respiratory function.

Developmental Changes in 5-HT_2A Receptor Expression in the Human Infant Medulla

We observed a trend for 5-HT_2A receptor binding density to increase with age across infancy in all medullary nuclei examined, consistent with widespread postnatal changes in 5-HT_2A receptor expression observed in rat medulla (Liu and Wong-Riley, 2008). Significant age-related increases in binding were observed in the HG and component nuclei of the inferior olivary complex. While we did not observe any obvious change in the distribution or intensity of 5-HT_2A immunostaining in the medulla from mid-gestation through infancy, we did observe an increase in size and complexity of 5-HT_2A immunoreactive neurons, which was particularly evident in the HG. Literature reports indicate that the HG undergoes significant development in the postnatal period in both humans and rodents, increasing the size and complexity of its dendritic arborization with a concomitant increase in 5-HT immunoreactive fiber density (Nara et al., 1989; O’Kusky, 1998); (Berger et al., 1992)(Talley et al., 1997). In rats, hypoglossal neuron excitability and firing also change significantly during the postnatal period, with a progressive decrease in excitability observed with age (Bayliss et al., 1997; Berger et al., 1992; Berger et al., 1996). Developmental changes in neuronal receptor expression might be expected to accompany these functional changes and, assuming a similar developmental process in humans, may account for the age-related changes in 5-HT_2A receptors observed in the HG in this study. We also observed significant age-related increases in 5-HT_2A receptor expression in the inferior olivary complex. It receives input from 5-HT neurons in the caudal domain but the precise influence of 5-HT upon olivary function, especially via 5-HT_2A receptors, is unknown. Similarly, little evidence is available on the developmental expression of 5-HT_2A receptors in this nucleus, thus the significance of the developmental changes in 5-HT_2A receptor binding observed in this study are also unknown.

In a previous study investigating the developmental expression of 5-HT receptors in the human infant medulla, using ³H lysergic acid diethylamide (³H-LSD) autoradiography, we observed no significant age-related changes in receptor binding density (Paterson et al., 2004). However, ³H-LSD binds to multiple 5-HT receptor subtypes (5-HT_1A-1D and 5-HT₂), and thus, age-related changes in individual 5-HT receptor subtypes may have been masked. Indeed, in the same study, we observed an age-related increase in 5-HT_1A receptor binding in the HG in the same cases using ³H 8-OH-DPAT autoradiography (Paterson et al., 2004). This observation parallels the postnatal increase in 5-HT_2A receptor binding identified in the HG in this study. Moreover, we have also identified subtle changes in the distribution and morphology of 5-HT neurons in the medulla during development (Kinney et al., 2007). The developmental changes in 5-HT_2A receptors observed in this study, taken together with the studies described above, therefore, support the idea that the structure and neurochemistry of the medullary 5-HT system, and thus, 5-HT mediated homeostatic function, continue to develop during the first year of postnatal life.

Potential Limitations of the Study

A potential limitation of this study is the small sample size of fetal and infant medullae. “Normal” tissue specimens from human infants, as used in this study, are particularly hard to accrue due to the rarity of autopsies in this age group. The small sample size is particularly problematic when determining the effects of age on 5-HT_2A receptor binding levels in the medulla, as individual cases can have a disproportionate effect on statistical observations. The significant age-related changes in binding observed in the HG and PIO in this study therefore need to be confirmed in a larger dataset. Ongoing accrual of the cases needed for confirmation is in progress. However, the age-related increases in 5-HT_2A receptor binding observed in this study are supported by previous observations in our laboratory describing subtle developmental changes in 5-HT neurons (Kinney et al., 2007) and 5-HT receptors (Paterson et al., 2004) across infancy. Moreover, in the same cases analyzed in this study, we have previously observed statistically significant age-related reductions in nicotinic receptor binding in the medulla (Duncan et al., 2008), that we believe underscore the robustness of the observations made in this study, despite the small sample size. We therefore propose that the changes in 5-HT_2A receptor binding observed in this study are consistent with a pattern of developmental changes that occur in the medulla during the postnatal period.

6. Conclusions

In conclusion, the spatial distribution of 5-HT_2A receptors in the human fetal and infant brainstem is consistent with a role in cardiorespiratory regulation. The ongoing changes in their binding density in the hypoglossal nucleus and inferior olivary complex across infancy indicate that the medullary 5-HT system is not fully mature at birth, but it continues to develop postnatally. This extended maturation period suggests that the system is vulnerable to developmental insults both before and after birth, especially through the end of infancy. Consequently, different insults at single or multiple pre- and/or postnatal time-points will likely have differential effects upon the structural and/or neurochemical development of the system, and ultimately upon different functional outcomes, not only in early life but beyond. This study provides baseline information for the analysis of the 5-HT_2A receptors in pediatric brainstems disorders in early life.