Chronic Prenatal Exposure to Carbon Monoxide ... - Oxford Journals

Journal of Neuropathology and Experimental Neurology Copyright q 2000 by the American Association of Neuropathologists

Vol. 59, No. 3 March, 2000 pp. 218–228

Chronic Prenatal Exposure to Carbon Monoxide Results in a Reduction in Tyrosine Hydroxylase-Immunoreactivity and an Increase in Choline Acetyltransferase-Immunoreactivity in the Fetal Medulla: Implications for Sudden Infant Death Syndrome MARY TOLCOS, BSC (HONS), HUGH MCGREGOR, BSC (HONS), DAVID WALKER, PHD,

AND

SANDRA REES, PHD

Abstract. Maternal cigarette smoking during pregnancy is associated with a significantly increased risk of Sudden Infant Death Syndrome (SIDS). This study investigated the effects of prenatal exposure to carbon monoxide (CO), a major component of cigarette smoke, on the neuroglial and neurochemical development of the medulla in the fetal guinea pig. Pregnant guinea pigs were exposed to 200 p.p.m CO for 10 h per day from day 23–25 of gestation (term 5 68 days) until day 61–63, at which time fetuses were removed and brains collected for analysis. Using immunohistochemistry and quantitative image analysis, examination of the medulla of CO-exposed fetuses revealed a significant decrease in tyrosine hydroxylase-immunoreactivity (TH-IR) in the nucleus tractus solitarius, dorsal motor nucleus of the vagus (DMV), area postrema, intermediate reticular nucleus, and the ventrolateral medulla (VLM), and a significant increase in choline acetyltransferase-immunoreactivity (ChAT-IR) in the DMV and hypoglossal nucleus compared with controls. There was no difference between groups in immunoreactivity for the m2 muscarinic acetylcholine receptor, substance P- or met-enkephalin in any of the medullary nuclei examined, nor was there evidence of reactive astrogliosis. The results show that prenatal exposure to CO affects cholinergic and catecholaminergic pathways in the medulla of the guinea pig fetus, particularly in cardiorespiratory centers, regions thought to be compromised in SIDS. Key Words:

Brainstem; Catecholamines; Cholinergic; Cardiorespiratory nuclei; Met-enkephalin; SIDS; Substance P.

INTRODUCTION It is now widely accepted that suboptimal intrauterine conditions account for many of the neurological deficits that develop in the postnatal period (1–3). Although the type, severity, and timing of the insult in relation to gestational age are all likely to be important variables, relatively little is known of the outcome of specific prenatal insults. Maternal cigarette smoking during pregnancy is associated with low birth weight (4) and a significantly increased risk of Sudden Infant Death Syndrome (SIDS) (5); it is thought that neural development might be compromised in infants dying of SIDS (6). Although several subtle structural and neurochemical abnormalities have been reported in SIDS brains, particularly in the brainstem (7–12), it is still not certain which specific adverse prenatal conditions might result in these deficits. Carbon monoxide (CO), one of the agents primarily responsible for the adverse effects of cigarette smoke, readily crosses the placenta where it binds to hemoglobin causing fetal hypoxia. Despite the importance of understanding how CO exposure might affect fetal brain development, few studies have been undertaken in experimental animals. To date it has been demonstrated that there are alterations in noradrenaline and serotonin levels From the Department of Anatomy and Cell Biology (MT, SR), The University of Melbourne, Parkville, Victoria, and the Department of Physiology (HM, DW), Monash University, Clayton, Victoria, Australia. Correspondence to: Assoc. Prof. Sandra Rees, Department of Anatomy and Cell Biology University of Melbourne, Parkville, 3052, Victoria, Australia. Source of support: National SIDS Foundation of Australia (Grant #: G93-09/156).

in the brain following chronic prenatal CO exposure (13, 14) and in GABA (15) and dopamine (16) following combined pre- and postnatal CO exposure. As these were biochemical studies, the alterations in neurotransmitter levels could not be localized to specific nuclei. In qualitative structural studies, gross alterations in the cerebellum have been described after combined pre- and postnatal CO exposure (15) and in the cerebral white matter, brainstem, and basal ganglia after acute prenatal exposure (17). However brain development following chronic prenatal exposure has not been investigated. Therefore, in this study in the fetal guinea pig, we have investigated the structure and neurochemistry of the brainstem following chronic prenatal CO exposure during the last 60% of gestation. A considerable proportion of central nervous system development in the guinea pig occurs in utero as it does in the human, making this animal a particularly useful model for the human fetus. We have concentrated on this region since one of the leading hypotheses for the cause of SIDS is a dysmaturity or abnormal development of the brainstem nuclei controlling respiratory, cardiovascular, and arousal activities (6). Reactive astrogliosis has been consistently reported in the brainstem of SIDS infants (8, 10, 18, 19) and has been interpreted as a response to a hypoxic episode. Alterations in several neurotransmitter systems in the brainstem have been reported, including a decrease in the concentration of the catecholamine synthesizing enzymes phenylethanolamine-N-methyltransferase (PNMT) and dopamine-b-hydroxylase (DbH) (20), an increase in the concentration of substance P (SP) together with an increase in the ratio of SP to met-enkephalin (ME) (11), and a decrease in muscarinic cholinergic receptor binding

218

219

PRENATAL CARBON MONOXIDE EXPOSURE

in the arcuate nucleus (12), indicating a disturbance in cholinergic transmission. In the medulla of CO-exposed fetuses near term, we have quantitatively assessed the immunoreactivity of the neuropeptides substance P (SP), and met-enkephalin (ME); the immunoreactivity of the enzymes involved in the catecholamine (tyrosine hydroxylase, TH) and acetylcholine (choline acetyltransferase, ChAT) synthesising pathways; the immunoreactivity of the m2 muscarinic acetylcholine receptor (m2), the most abundant muscarinic receptor type in the brainstem (21); and the proliferation of reactive astrocytes using the intermediate filament marker, glial fibrillary acidic protein (GFAP). A number of medullary nuclei have been examined, particularly those directly or indirectly involved in the control of respiratory, cardiovascular, and arousal activities. MATERIALS AND METHODS Experimental Procedure Pregnant guinea pigs (n 5 6) were assigned to either the control (n 5 3) or experimental group (n 5 3) and were placed into their allocated chamber on days 23–25 of gestation. Guinea pigs in the experimental group were exposed to 200 p.p.m CO for 10 h/day until 61–63 days of gestation (term 5 68 days). It has recently been reported that prenatal exposure to 200 p.p.m CO throughout pregnancy results in fetal and maternal carboxyhemoglobin (COHb) levels of 13% and 8.5% respectively (22), which is consistent with levels found in humans who smoke heavily (23, 24). Exposure to CO occurred in a plastic chamber (100 cm 3 50 cm 3 50 cm) sealed with a perspex lid, with food and water provided ad libitum. To generate 200 p.p.m CO, 100% CO (BOC, Australia) was mixed with air (10 L/min) in a rotameter and delivered to the chamber (22). The concentration of CO in the chamber was monitored and recorded continuously by a CO sensor built into the side of the chamber and all gas exited the chamber via an externally vented passive extraction system. All control animals were placed in a separate chamber with the same dimensions as the CO chamber, but were continuously exposed to room air. These experiments were performed under the guidelines set down by the National Health and Medical Research Council of Australia, and the project was approved by the separate Animal Experimentation and Ethics Committees of the University of Melbourne and Monash University.

Tissue Preparation Control and CO-exposed pregnant guinea pigs were removed from the chambers at 61–63 days of gestation and fetuses were delivered by Caesarean section after deeply anesthetizing the mothers and fetuses with Nembutal (Pentobarbitone Sodium, 60 mg/ml). Fetuses and placenta were weighed and the fetal head perfused via the left ventricle with isotonic saline followed by 4% paraformaldehyde in 0.1 M phosphate buffer (PB) at pH 7.4. Liver weights and crown-to-rump lengths were then measured. Brains were removed from the skull and postfixed for 4 h. Whole brain, brainstem (spino-medullary junction to just below the cerebellar peduncles) and cerebellar weights were taken

and blocks of brainstem were transferred into 20% sucrose in 0.1 M PB (pH 7.4) overnight prior to sectioning. The medulla of control and CO-exposed fetuses were serially sectioned at 40 mm on a Leitz-Wetzlar freezing microtome. Every 5th section was collected for structural analysis. Sections were mounted from 0.5% gelatin onto slides coated in 1% gelatin and stained with 0.01% thionin in acetate buffer (pH 4.4). Following dehydration in graded alcohols, clearing and coverslipping, sections of medulla from control and CO-exposed fetuses were qualitatively assessed for gross structural abnormalities such as areas of pallor, necrosis, infarction or neuronal loss, and gliotic aggregations. The intervening sections were collected and frozen in cryoprotectant (25) until used for immunohistochemical staining with the antibodies described below.

Immunohistochemistry Immunoreactivity for TH, ChAT, m2, SP, ME, and GFAP was localized on free-floating sections using the avidin-biotin peroxidase complex (Vector Laboratories, Burlingame CA) as described previously (26). The primary antibodies were obtained from the following sources and used at the following dilutions: mouse-anti-tyrosine hydroxylase (monoclonal, Boehringer Mannheim Biochemica, 1:1,000); goat anti-choline acetyltransferase (polyclonal, Chemicon, 1:500); rat anti-m2 muscarinic acetylcholine receptor (monoclonal, Chemicon, 1:1,000); rabbit-anti-substance P (polyclonal, kindly donated by Dr. R. Murphy: the production and characterization of this antibody has been described in Morris et al., 1986 (27), 1:1,000); rabbit-antimet-enkephalin (polyclonal, Penninsula Laboratories Inc, 1: 1,000) and rabbit anti-glial fibrillary acidic protein (polyclonal, Dako, 1:1,000). Sections were incubated either overnight (ChAT, m2, SP, GFAP) or for 48 h (TH, ME). Sections were then immersed in the appropriate secondary antibody (1:400, biotinylated anti-rabbit IgG; 1:200, biotinylated anti-mouse IgG; 1:200, biotinylated anti-goat IgG; Vector Laboratories) and after thorough washing placed in the appropriate dilution of the avidin-biotin complex (1:400, SP, ME, GFAP; 1:200, TH; 1:100 and Elite kit, m2; 1:50 and Elite kit, ChAT). Sections were reacted with 0.05% 3,39-diaminobenzidine (DAB) solution in 0.01% hydrogen peroxidase to produce a brown reaction product. Phosphate buffer was replaced with Tris phosphate buffered saline for ChAT immunostaining to reduce background staining. Sections immunostained with TH, ChAT and m2 were counterstained in 0.01% thionin in acetate buffer (pH 4.4) for 6 min and ME was enhanced with 0.01% osmium tetroxide for 30 s prior to dehydration and clearing. Negative controls were prepared for each antibody by replacing the primary antibody with PB or primary diluent in the incubation. In this procedure staining always failed to occur. For each antibody, sections from control and CO-exposed fetuses were stained simultaneously to ensure uniform conditions for subsequent quantitative and qualitative analysis.

Western Blotting At 62 days of gestation, a control and CO-exposed pregnant guinea pig were removed from the chambers and fetuses (3 control and 2 CO-exposed) were delivered by Caesarean section

J Neuropathol Exp Neurol, Vol 59, March, 2000

220

TOLCOS ET AL

after deeply anesthetizing the mothers and fetuses with Nembutal (Pentobarbitone Sodium, 60 mg/ml). The medulla (spinomedullary junction to just below the cerebellar peduncles) from each fetus was collected and homogenized in 50 mM Tris-HCl (pH 7.4) containing 50 mM NaCl and protease inhibitors (10 mM EDTA, 10 mM e-amino-n-caproic acid, 0.25 mM PMSF, 5 mM N-ethylmaleimide, 10 mM benzamidine, 500 KIU/ml aprotonin and 1 mg/ml pepstatin A). Samples were centrifuged (8 min; 10,000 g) and the supernatant was retained. The concentration of protein was determined using the bicinchoninic acid (BCA) protein assay reagent kit (Pierce, Rockford, IL) using BSA fraction V (Pierce, Rockford) as standards. Tissue preparations were then heated (958C for 5 min) in Laemmli’s sample buffer (Sigma Chemical Company, MO, USA) and separated by electrophoresis in 4%–15% gradient Tris-HCl polyacrylamide gel (BioRad; 40 mg protein/lane) in Laemmli’s running buffer (28). Proteins were electrophoretically transferred onto PVDF nitrocellulose membrane (Immobilon-P transfer membrane, Millipore Corp. Bedford, MA) in 25 mM Tris-glycine buffer containing 0.1% SDS and 20% methanol. Membranes were then blocked in ‘‘blotto’’ (8% (w/v) skim milk powder in PS with triton X-100) for 30 min and incubated in mouse-anti-TH (monoclonal, Boehringer Mannheim Biochemica, 1:1,000) for 48 h. Membranes were then washed in PB (3 3 10 min), incubated in biotinylated anti-mouse IgG (1:200; Vector Laboratories) for 45 min and after thorough washing (3 3 10 min) placed in the avidin-biotin complex (1:200; Vectastain Laboratories Burlingame CA) for a further 45 min. The bands were visualized using DAB as the chromagen and 0.01% hydrogen peroxide. To ensure that the primary anti-TH antibody was recognizing TH, adult maternal adrenal gland homogenates (prepared as for the medulla) and purified full-length TH protein of known molecular weight (0.05 mg and 56 kDa; Signal Transduction Products, San Clemente, CA) were also separated electrophoretically and immunoblotted as described. To control for nonspecific binding of the secondary antibody or the avidin-biotin complex to proteins, control lanes from which primary antibody was omitted were included. In this case, staining failed to occur. The density of anti-TH-IR bands were measured from control and CO-exposed medullary tissue using the Image Pro Plus image analysis system as described below. As there were insufficient numbers of samples to perform statistics, the results are expressed as an average percentage change from control values.

Quantitative Structural Analysis Every tenth thionin-stained section of medulla extending from the spino-medullary to the ponto-medullary junction was randomly projected onto an orthogonal 1 cm 3 1 cm grid. The cross-sectional area of the medulla was determined using a point counting technique (see Mallard et al, 1999 (29). In addition the volume of the medulla was determined using the principle of Cavalieri (30).

Qualitative Immunohistochemical Analysis Qualitative assessment of the density and distribution of the immunoreactivity for TH, ChAT, m2, SP, ME and GFAP in the nuclei of interest was performed double blind by 2 independent


observers. Three levels of the medulla (caudal, mid- and rostral) were chosen (see Figure 3A–C in Tolcos and Rees, 1997 [26]) and a section at each level in the control and experimental fetuses was analyzed simultaneously using projectorscopes. These levels were chosen as they included the relevant areas of cardiorespiratory control, namely the NTS, DMV, hypoglossal nucleus, area postrema, ventral respiratory group in the ventrolateral medulla (VLM) and nucleus ambiguus (NAmb). Other nuclei and tracts were also qualitatively assessed and they included the spinal trigeminal tract and nucleus (SP, ME, TH), pyramidal tract (GFAP), dorsal and ventral spinocerebellar tract (GFAP), medial longitudinal fasciculus (GFAP), medial lemniscus (GFAP), inferior olivary (GFAP, ChAT), gracile (TH) and cuneate (TH) nuclei, ventrolateral medulla (SP), ventral reticular formation including the intermediate reticular nucleus (TH), hypoglossal, vagus, facial, and abducens nerves (ChAT), as well as the central canal and the lining of the fourth ventricle (GFAP). By analyzing both cardiorespiratory and noncardiorespiratory nuclei we were in a position to evaluate the specificity of any change which might have occurred.

Quantitative Immunohistochemical Analysis Analysis of immunoreactivity for TH, ChAT, m2, SP, ME and GFAP was performed quantitatively in nuclei in which changes were observed qualitatively. Sections from the caudal, mid- and rostral levels (see Figure 3A–C in Tolcos and Rees, 1997 [26]) were assessed quantitatively. The image was projected onto a computer monitor and a line drawn around the nucleus of interest for each section stained with antisera to either TH (NTS, DMV, hypoglossal nucleus, area postrema), ChAT (NTS, DMV, hypoglossal nucleus, facial nucleus), m2 (NTS, DMV, hypoglossal nucleus, area postrema, facial nucleus), SP (NTS, hypoglossal nucleus, area postrema), ME (NTS, hypoglossal nucleus, area postrema) and GFAP (NTS, DMV, hypoglossal nucleus, area postrema). In the case of TH-, m2- and GFAPIR, the NTS and DMV were measured together as it was difficult to accurately delineate the nuclei. In addition, the optical density of TH-IR and GFAP-IR was measured in the VLM and the medullary reticular formation respectively using sample areas of equal size positioned at random. The background optical density was measured from an area where staining was not evident on each section. This was then subtracted from optical density measurements from each nucleus or region. The background-adjusted optical density measurements were averaged for each animal and then averaged for control and CO-exposed groups. Immunoreactivity for all antibodies was assessed quantitatively using the Image-Pro Plus image analysis system (Media Cybernetics, Maryland). The image was acquired by using a Sony (CCD-IRIS/RGB) video camera head mounted to an Olympus microscope. For each objective, the system was calibrated for optical density measurements as follows. An area of the slide consisting of the glass slide and coverslip was aligned in the field of view. A black field was obtained by blocking the light source (infinite optical density), and a blank area of the slide was used as incident light. Using these values, the image analysis system converted light intensity into absolute optical density values (range 0 to 2.4). An electronic shading correction

221


was applied to each acquired image to compensate for any unevenness that might be present in the illumination. Sampling regions of varying stain intensity and a blank region of the slide tested the calibration. A reference slide was used at the start and end of the session to account for any discrepancies.

Counts of Immunoreactive Neurons In a quantitative study, the numerical density of TH-positive neurons in the region of the VLM, SP-positive neurons in the NTS and ME-positive neurons in the ventral reticular formation, were determined in sections taken at 3 matched levels in control (n 5 8) and CO-exposed (n 5 8) animals. These levels were caudal, mid- and rostral medulla for TH and mid-, lower, and upper medulla for SP and ME (see Figure 3A–D in Tolcos and Rees, 1997 [26]). It was not possible to count only cells with a nucleus since the intensity of the immunoreactive product made it impossible to visualize nuclei therefore all neuronal profiles were counted. Tyrosine hydroxylase-positive neurons were counted in the lower quadrant of each section. This area was demarcated by extending a line from the base of the central canal or the fourth ventricle horizontally to the lateral edge of the section and another vertically through the raphe nucleus and medial lemniscus to the ventral edge of the section. This is referred to as the region VLM in the Results section. The area of the left and right NTS, and the ventral reticular formation was determined using adjacent thionin-stained sections. The areas were then measured using a digitizing tablet (SigmaScan Pro 4, Jandel Scientific). The results are expressed as cell density (cells/mm2) for each section. The average density of TH-, SP- and ME-positive neurons for each animal was then determined by pooling the results from each level and calculating the mean value. For the CO-exposed group, values for each animal were corrected for the reduction in the volume of the medulla. Means were then taken for the control and CO-exposed groups, and expressed as the mean density (neurons/mm2) 6 the standard error of the mean (M 6 SEM). The number of ChAT-IR neurons in the left and right NAmb was determined by counting the positive neuronal profiles in equivalent sections from the caudal, mid- and 2 rostral levels in control and COexposed animals. An average was determined for the right and left NAmb, and then for each animal by pooling the results. Means were then taken for control and CO-exposed animals and the results are expressed as the mean number of ChATpositive neurons within these levels of the medulla 6 the standard error of the mean (M 6 SEM).

Statistical Analysis Fetal weights were expressed as the mean 6 standard error of the mean (M 6 SEM) and analyzed using an unpaired 2tailed t-test for independent samples assuming unequal variance. All immunohistochemical and structural data is expressed as the mean 6 the standard error of the mean (M 6 SEM) and statistical analysis performed using the Mann-Whitney U-test.

Preparation of Figures Black and white negatives (Copex Pan) were scanned onto an Apple Macintosh computer and figures were compiled using Adobe Photoshop 4.0. Background artifacts were removed and

TABLE 1 Data obtained at Postmortem from Control (n 5 8) and Carbon Monoxide (CO)-exposed (n 5 8) Fetuses Parameters

Control

Litter size 3.00 6 0.60 Body weight (g) 86.64 6 2.26 Brain weight (g) 2.49 6 0.05 Cerebellum weight (g) 0.29 6 0.01 Liver weight (g) 3.84 6 0.20 5.40 6 0.60 Placental weight (g) Crown-to-rump length (cm) 14.00 6 0.40 Brain weight: Body weight (%) 3 6 0.1 Brain weight: Liver weight (%) 66 6 3

CO-exposed 2.70 78.35 2.32 0.25 3.24 5.99 11.94

6 6 6 6 6 6 6

0.70 2.59* 0.06* 0.01 0.20* 0.20 0.40*

3 6 0.1 74 6 6

Values are mean 6 standard error of the mean. * p , 0.05.

image intensity levels were adjusted equally for control and experimental images.

RESULTS At 61–63 days of gestation, body, brain, and liver weights were significantly reduced in CO-exposed fetuses (n 5 8) when compared with controls (n 5 8) (Table 1). Measurements of the crown-to-rump length in CO-exposed fetuses also showed a significant decrease when compared with controls. No significant differences were found in either litter size, cerebellar, or placental weights. There were no significant differences in either brain to body or brain to liver ratios when CO-exposed fetuses were compared with controls (Table 1). Structural Analysis Analysis of thionin-stained sections of the medulla did not reveal any gross structural damage or areas of pallor, necrosis, infarction, or gliosis. The latter observation was supported by optical density measurements for GFAP-IR in the NTS/DMV (0.12 6 0.01 [control] vs 0.11 6 0.01 [CO-exposed]) and area postrema (0.34 6 0.03 [control] vs 0.32 6 0.07 [CO-exposed]). Quantitative analysis of thionin-stained sections of the medulla revealed a significant decrease in the volume (mm3) of the medulla in COexposed (82.58 6 3.92) compared with control (92.44 6 2.81, p , 0.05) fetuses. Immunohistochemistry Tyrosine Hydroxylase: In the medulla of control fetuses, TH-IR neurons processes and terminals were distributed predominantly in the NTS (A2/C2 groups), DMV, hypoglossal nucleus, intermediate and lateral reticular nuclei, ventrolateral medulla (A1/C1 groups), medullary reticular formation, gracile and cuneate nuclei, and spinal trigeminal nucleus, raphe nucleus, lateral paragigantocellular reticular and parvocellular reticular nuclei, and the J Neuropathol Exp Neurol, Vol 59, March, 2000

222

TOLCOS ET AL

TABLE 2 Optical Density Measurements (M 6 SEM) for Tyrosine Hydroxylase (TH) and Choline Acetyltransferase (ChAT)immunoreactivity in the Medulla of Control (n 5 8) and Carbon Monoxide (CO)-exposed (n 5 8) Fetal Guinea Pigs Nuclei NTS NTS/DMV DMV XII AP VLM Facial VLM cells (cells/mm2) NAmb cells

TH Control

CO-Exposed

ChAT Control

CO-exposed

m2 Control

CO-exposed

— 0.15 6 0.03 — — 0.37 6 0.02 0.10 6 0.01 — 10.6 6 0.8 —

— 0.07 6 0.01* — — 0.21 6 0.06* 0.05 6 0.01** — 7.9 6 0.4* —

0.24 6 0.02 — 0.18 6 0.02 0.27 6 0.01 — — 0.22 6 0.01 — 55.6 6 4.5

0.28 6 0.02 — 0.24 6 0.02* 0.33 6 0.02* — — 0.22 6 0.01 — 53.51 6 4.8

— 0.18 6 0.02 — 0.35 6 0.03 0.75 6 0.06 — 0.32 6 0.01 — —

— 0.16 6 0.02 — 0.30 6 0.02 0.69 6 0.1 — 0.29 6 0.02 — —

* p , 0.05; ** p , 0.01.

area postrema. The distribution is similar to that described in other species (31–34). The most striking differences in TH-IR in fibers and terminals in the medulla of CO-exposed fetuses compared with controls was a significant decrease in the optical density of staining in the NTS/DMV (particularly in the caudal extent of the nucleus), AP, and VLM (Table 2). This is illustrated by comparing Figure 1B (CO-exposed) with Figure 1A (control) and for the area postrema, Figure 1D (CO-exposed) with Figure 1C (control). There also appeared to be a decrease in TH-IR in the intermediate reticular nucleus in CO-exposed (Fig. 1B) fetuses when compared with controls (Fig. 1A). The levels of TH-IR were too low to give reproducible quantitative results so optical density measurements were not made. Other regions such as the gracile, cuneate, and spinal trigeminal nuclei did not show any differences in qualitative analysis. There was a significant difference in the numerical density of TH-positive neurons in the region of the VLM (refer to Materials and Methods for description) in CO-exposed fetuses compared with controls (Table 2). Choline Acetyltransferase: In the medulla of control fetuses ChAT-IR neurons, fibers and terminals were distributed in the hypoglossal nucleus, DMV, NAmb, medullary reticular formation, inferior olivary nucleus, medial subnucleus of the NTS, area postrema, lateral reticular nucleus, ventrolateral medulla, cuneate nucleus, lateral paragigantocellular nucleus, spinal trigeminal nucleus, facial nucleus, and superior salivatory nucleus. ChAT-IR fibers were also present in the abducens (VI), facial (VII), vagus (X), and hypoglossal (XII) nerves. The distribution is similar to that described in other species (35–38). Quantitative densitometry of transverse sections of the caudal, mid-, and rostral medulla of control and CO-exposed fetuses revealed a statistically significant increase in ChAT-IR within the DMV and the hypoglossal nucleus of CO-exposed animals but not in the NTS (Table 2). This is illustrated by comparing Figure 1F (CO-exposed) with Figure 1G (control). There was no significant J Neuropathol Exp Neurol, Vol 59, March, 2000

difference in optical density of the facial nucleus in control compared with CO-exposed fetuses. In a qualitative study of the cranial motor nerves (VI, VII, X, XII) and the inferior olivary nucleus, no consistent differences were observed in the intensity ChAT-IR fibers or neurons. No significant differences were found in the number of ChAT-IR neurons of the NAmb in CO-exposed fetuses compared with controls (Table 2). m2 Muscarinic Acetylcholine Receptor: Within the medulla of control fetuses, m2-IR was observed in fibers, terminals, and occasionally over cell bodies in the hypoglossal nucleus NTS, DMV, ventral and dorsal medullary reticular formation, lateral reticular nucleus, VLM, NAmb, spinal trigeminal nucleus, gracile and cuneate nuclei, medial longitudinal fasciculus, area postrema, arcuate nucleus, inferior olivary nucleus, raphe nuclei, external cuneate nucleus, gigantocellular, ventral gigantocellular, parvocellular, lateral paragigantocellular and intermediate reticular nuclei, Probst’s nucleus and bundle, Roller nucleus, intercalated nucleus, vestibular nuclei, facial nucleus, prepositus hypoglossal nucleus, abducens nucleus, predorsal bundle, and medial lemniscus. The distribution is similar to that described in other species (39–42). There was no significant difference in the optical density of m2-IR in the NTS/DMV, hypoglossal nucleus, area postrema, or facial nucleus of CO-exposed compared with controls. The results are summarized in Table 2. Substance P and Met-enkephalin: The distribution of SP- and ME-IR has been described previously in the medulla of the fetal guinea pig (26). Transverse sections of the caudal, mid- and rostral medulla were analyzed for alterations in the neurotransmitters/neuromodulators SP and ME. There were no statistically significant differences in optical density in any of the cardio-respiratory nuclei when control and CO-exposed animals were compared. The results are summarized in Table 3. There was no significant difference between control and CO-exposed animals in the density of SP-IR neurons in the


NTS, or of ME-IR neurons in the ventral medullary reticular formation (Table 3). Qualitative analysis revealed no apparent difference in SP- and ME-IR in the spinal trigeminal tract and nucleus or in SP-IR in the VLM. Western Analysis Using quantitative Western analysis, a 13% reduction in the level of TH protein was measured in the medulla of CO exposed fetuses in comparison to controls (C: 0.38 6 0.03 vs CO: 0.33 6 0.05; Fig. 2). DISCUSSION This study has investigated the structural and neurochemical development of the fetal brainstem in the guinea pig after chronic moderate exposure to CO throughout the last 60% of gestation. Significant reductions were observed in fetal body, brain and liver weights and crownto-rump length. There was also a significant reduction in the volume of the medulla in CO-exposed fetuses compared with controls. In contrast to intrauterine growthrestriction produced by chronic placental insufficiency (26), brain sparing was not evident, as there were no significant differences in brain to body and brain to liver ratios when CO-exposed fetuses were compared with controls. In brains from CO-exposed fetuses there were no gross structural alterations such as necrosis, infarction, or gliosis; the latter was confirmed by optical density measurements of GFAP-IR. To our knowledge this is the first study in which the structure of the brainstem has been examined following chronic exposure to moderate levels of CO. There have been reports of brain damage after chronic pre- and postnatal exposure to higher levels of CO (15) and to acute prenatal CO intoxication (17). In this study, the use of quantitative immunohistochemical analysis was essential for the identification of regionspecific changes in the levels of TH- and ChAT-IR in control compared with CO-exposed animals. Although this technique does not measure changes in the amount of protein, the reduction in TH-IR was supported using Western blot analysis. The reduction in TH as revealed using Western blot analysis (13%) was not as great as that from optical density measurements of single nuclei within the medulla (43%–53%). This is not unexpected as significant changes in specific nuclei are diminished when the entire medulla is homogenized. A major finding of this study was that prenatal CO exposure affected the catecholaminergic system in the brainstem as determined by a significant reduction in immunoreactivity for TH in the NTS, DMV, area postrema, intermediate reticular nucleus, and VLM. These results are consistent with those of Storm and Fechter, 1985 (13) who reported a decrease in the concentration of noradrenaline in the pons/medulla in the newborn rat following prenatal CO exposure. In the present study, the decrease in TH-IR in the VLM was associated with a significant

223

decrease in the number of TH-positive neurons in this region. Thus, the reduction in TH-IR is most likely largely due to this reduction in cell numbers although a decrease in the content of the enzyme within neurons and their processes, a decrease in the size of neurons and the extent of their dendritic arbors, or a decrease in catecholaminergic input to the nuclei could also contribute. Although other noncardiorespiratory regions of the medulla show TH-IR (i.e. gracile, cuneate, and spinal trigeminal nucleus) in both control and CO-exposed fetuses, the very sparse staining would not allow for accurate optical density measurements. At the level of qualitative assessment there did not appear to be any changes in these regions. It therefore appears that changes in TH-IR are seen in those regions of the medulla with a high catecholaminergic innervation. These are important regions involved in cardiorespiratory control. Since the nuclei mentioned above are involved in the control of cardiac and respiratory activities, any alteration in the catecholamine content of their afferent and efferent connections, or of the cell bodies themselves, could result in abnormal functioning of these pathways during development and after birth. It is relevant that the major decrease in TH-IR in the NTS was in the caudal extent of the nucleus, which receives chemosensory input (43–45). Consistent with this finding is our previous observation that neonatal guinea pigs from CO-treated pregnancies show altered respiratory responses to asphyxia and CO2 (22). In relation to the cholinergic system, CO exposure affected ChAT but not the m2 receptor expression in the medulla. Immunoreactivity for ChAT was significantly increased in the hypoglossal nucleus, and DMV, but not in the NTS or facial nucleus in CO-exposed fetuses compared with controls. The hypoglossal nucleus innervates the muscles of the tongue and is therefore important in the process of swallowing. Although the hypoglossal nucleus is not generally considered to be a primary respiratory center, populations of hypoglossal neurons have been shown to discharge during inspiration, expiration, and transitional phases of respiration (46). An increase in ChAT might directly or indirectly affect the control of swallowing, in relation to breathing, in the neonate. It has been suggested that the combination of impaired swallowing, deficient arousal mechanisms, and gastro-oesophageal reflux can result in apnoea in the human infant (47– 49). An interesting finding in the present study was the presence of a population of m2-IR neurons located on the ventral surface of the medulla. Studies have identified the presence of a chemosensitive region here in neonatal guinea pigs and rabbits (50). In the fetal guinea pig, this group of neurons appears to be located in a similar position to the arcuate nucleus in the human (12) and the respiratory chemosensitive fields in the cat (51–53). Furthermore, muscarinic cholinergic receptor binding has J Neuropathol Exp Neurol, Vol 59, March, 2000

224


TOLCOS ET AL

225


TABLE 3 Optical Density Measurements (M 6 SEM) for Substance P (SP) and Metenkephalin (ME)-immunoreactivity in the Medulla of Control (n 5 8) Carbon Monoxide (CO)-exposed (n 5 8) Fetal Guinea Pigs Nuclei NTS XII AP NTS cells (cells/mm2) MRF cells (cells/mm2)

SP Control 0.20 0.09 0.21 8.6

6 0.01 6 0.01 6 0.03 6 2.5 —

CO-exposed 0.19 0.09 0.22 8.8

Fig. 2. Western blot of tyrosine hydroxylase (TH) protein purified from the medulla of 62-day-old control and carbon monoxide (CO)-exposed fetuses. Quantitative western analysis revealed a 13% reduction in TH protein in the medulla of COexposed animals in comparison to controls. This result supports quantitative immunohistochemical analysis.

been localized to the arcuate nucleus in the human infant (12). In contrast to the effect of prenatal CO exposure on enzymes in the catecholaminergic and cholinergic systems in the medulla, the levels of the neuropeptides SP and ME were not affected. This perturbation might affect

6 0.02 6 0.01 6 0.03 6 1.2 —

ME Control

CO-exposed

0.08 6 0.01 0.04 6 0.01 0.07 6 0.01 — 21.1 6 2.8

0.10 6 0.01 0.05 6 0.01 0.07 6 0.01 — 19.7 6 2.7

synthetic enzymes differentially from neuropeptides or the effect might follow a different time course. As the animals were exposed to CO over several weeks it would be expected that if any affect were to have occurred it would have been observed within this time frame. Carbon monoxide may exert effects on the growing brain by inducing hypoxia, although it is possible that other mechanisms are involved as discussed below. Recent in vivo (54) and in vitro (55, 56) studies have shown that short-term (,24 h) hypoxia mediates an increase in TH mRNA by increasing the rate of gene transcription. Following exposure to chronic hypoxia during pregnancy in rats, White and Lawson, 1997 (57) reported a decrease in TH mRNA and protein in the dorsal medulla at birth. After birth however, there was an increase in TH mRNA and protein, a decrease in PNMT protein in the dorsal medulla, and an increase in TH mRNA in the ventral medulla. Other studies investigating the effects of chronic postnatal exposure to hypoxia have shown an increase in TH mRNA and/or protein in the caudal (58–61) and rostral NTS (59, 61), the DMV (61), the caudal dorsomedial medulla (62), the VLM (61, 62) and the locus ceruleus (62). Therefore it appears that hypoxia has different effects on the expression of tyrosine hydroxylase in the fetus, neonate, and adult. There have been some studies on the effects of hypoxia on the cholinergic system in the fetus, neonate, and adult and in general they reported a decrease in cholinergic markers across all ages. In a study of mouse fetal cortical cells, a decrease in ChAT activity was reported following exposure to severe chronic hypoxia (63). A number of in

← Fig. 1. Bright-field photomicrographs demonstrating the distribution and intensity of tyrosine-hydroxylase-immunoreactivity (TH-IR) (panels A–D) and choline-acetyltransferase-immunoreactivity (ChAT-IR) (panels E and F) in the medulla of control (panels A, C, E) and carbon monoxide (CO)-exposed (panels B, D, F) fetal guinea pigs. Exposure to CO prenatally resulted in a significant decrease in the optical density of TH-IR in the nucleus tractus solitarius (NTS), dorsal motor nucleus of the vagus (DMV), ventrolateral medulla (VLM) (panel B), and area postrema (AP) (panel D) when compared with controls (panels A and C). There was also a decrease in the numerical density of TH-IR neurons in the region of the VLM and an apparent reduction in TH-IR in the intermediate reticular nucleus (IRt) of CO-exposed fetuses (panel B) when compared with controls (panel A). An increase in the optical density of ChAT-IR was measured in the hypoglossal nucleus (XII) and DMV of CO-exposed fetuses (panel F) when compared with controls (panel E). Calibration bar: (panels A and B) 5 34 mm; (panels C and D) 5 21 mm; (panels E and F) 5 22 mm. Abbreviations: 4V, fourth ventricle; cc, central canal; MRF, medullary reticular formation; ts, tractus solitarius; Xn, vagus nerve; XIIn, hypoglossal nerve.


226

TOLCOS ET AL

vivo (64–67) and in vitro (68, 69) studies in the neonate and adult have shown that hypoxia suppresses the synthesis and presynaptic release of acetylcholine throughout the cerebral hemispheres. In these studies the only effect observed in the brainstem was a reduction in ChAT-IR cells in the pedunculopontine nucleus following acute severe hypoxia (70). On the other hand, in studies using a model that included hypoxia and ischemia, an increase in the density of ChAT-IR in the neuropil (71), and neurons (72) in the striatum has been reported. In addition to inducing hypoxia, it has been proposed that CO has a direct cytotoxic effect (73, 74). Carbon monoxide binds to many enzymes with a heme moiety, including mitochondrial cytochromes, thereby blocking energy production and interfering with cellular respiration. Further, it is now known that CO is formed endogenously in the nervous system by heme oxygenases which oxidatively cleave the heme ring releasing CO and biliverdin; CO has been implicated as a neuronal messenger activating soluble guanylate cyclase (75) and increasing the concentration of cGMP (76, 77). It is therefore conceivable that the alterations in TH- and ChAT-IR seen in our study occur as a result of the exogenous CO perturbing the neural pathways normally modulated by endogenously produced CO. Thus, we have shown that chronic moderate prenatal CO exposure results in an increase in ChAT-IR in the DMV and hypoglossal nucleus and a decrease in TH-IR in the NTS, DMV, area postrema, intermediate reticular nucleus, and VLM in the medulla of the fetal guinea pig. The changes in TH-IR correlate well with studies in SIDS brains where it has been reported that there is a decrease in the concentration of PNMT and DbH in the NTS, DMV, and the NAmb (20), and a reduction in TH-IR in the vagal nuclei and the ventrolateral reticular formation (78). Aspects of the cholinergic system are affected in the brainstem of SIDS infants, since it has been reported that there is a decrease in muscarinic receptor binding in the arcuate nucleus (12). In this study, we found no difference in the immunoreactivity of a specific muscarinic receptor subtype (m2). The alteration observed by Kinney et al, 1995 (12) might therefore be attributed to differences in subtypes other than m2. Alterations in the expression of neurotransmitters during fetal brain development will not only result in abnormal neurotransmission, but might also influence neuronal differentiation and synaptic plasticity, since it is now apparent that catecholamines (79–81) and acetylcholine (80, 82–85) play a trophic role in brain development. Such effects could be of relevance to the development of circuitry in cardiovascular and respiratory nuclei in the neonate, and thus could have implications for the relationship between cigarette smoking during pregnancy and abnormal brain development. J Neuropathol Exp Neurol, Vol 59, March, 2000

ACKNOWLEDGMENTS The authors wish to thank Mrs. Michelle Loeliger for her excellent technical assistance and Mrs. Kerryn Westcott for her involvement in the generation and care of the animals used in this study. We would also like to acknowledge Drs. Brian Key, James St. John, Richard Anderson, and Heidi Clarris for their help with the western blot analysis.

REFERENCES 1. Nelson KB, Ellenberg JH. Antecedents of cerebral palsy. Multivariate analysis of risk. N Engl J Med 1986;315:81–86 2. Kinney HC, Filiano JJ. Brainstem research in sudden infant death syndrome. Pediatrician 1988;15:240–50 3. Roberts GW. Schizophrenia: A neuropathological perspective. Br J Psychiatry 1991;158:8–17 4. Fredricsson B, Gilljam H. Smoking and reproduction. Short and long term effects and benefits of smoking cessation. Acta Obstet Gynecol Scand 1992;71:580–92 5. Mitchell EA, Ford RP, Stewart AW, Taylor BJ, Becroft DM, Thompson JM, et al. Smoking and the sudden infant death syndrome. Pediatrics 1993;91:893–96 6. Kinney HC, Filiano JJ, Harper RM. The neuropathology of the sudden infant death syndrome. A review. J Neuropathol Exp Neurol 1992;51:115–26 7. Naeye RL. Brain-stem and adrenal abnormalities in the suddeninfant-death syndrome. Am J Clin Pathol 1976;66:526–30 8. Takashima S, Armstrong D, Becker L, Bryan C. Cerebral hypoperfusion in the sudden infant death syndrome? Brainstem gliosis and vasculature. Ann Neurol 1978;4:257–62 9. Quattrochi JJ, Baba N, Liss L, Adrion W. Sudden infant death syndrome (SIDS): A preliminary study of reticular dendritic spines in infants with SIDS. Brain Res 1980;181:245–49 10. Kinney HC, Burger PC, Harrell FE, Jr., Hudson RP, Jr. ‘Reactive gliosis’ in the medulla oblongata of victims of the sudden infant death syndrome. Pediatrics 1983;72:181–87 11. Bergstrom L, Lagercrantz H, Terenius L. Post-mortem analyses of neuropeptides in brains from sudden infant death victims. Brain Res 1984;323:279–85 12. Kinney HC, Filiano JJ, Sleeper LA, Mandell F, Valdes-Dapena M, White WF. Decreased muscarinic receptor binding in the arcuate nucleus in sudden infant death syndrome. Science 1995;269:1446–50 13. Storm JE, Fechter LD. Prenatal carbon monoxide exposure differentially affects postnatal weight and monoamine concentration of rat brain regions. Toxicol Appl Pharmacol 1985;81:139–46 14. Storm JE, Fechter LD. Alteration in the postnatal ontogeny of cerebellar norepinephrine content following chronic prenatal carbon monoxide. J Neurochem 1985;45:965–69 15. Storm JE, Valdes JJ, Fechter LD. Postnatal alterations in cerebellar GABA content, GABA uptake and morphology following exposure to carbon monoxide early in development. Dev Neurosci 1986;8: 251–61 16. Fechter LD, Karpa MD, Proctor B, Lee AG, Storm JE. Disruption of neostriatal development in rats following perinatal exposure to mild, but chronic carbon monoxide. Neurotoxicol Teratol 1987;9: 277–81 17. Okeda R, Matsuo T, Kuroiwa T, Tajima T, Takahashi H. Experimental study on pathogenesis of the fetal brain damage by acute carbon monoxide intoxication of the pregnant mother. Acta Neuropathol (Berl) 1986;69:244–52 18. Takashima S, Becker LE. Developmental abnormalities of medullary ‘‘respiratory centers’’ in sudden infant death syndrome. Exp Neurol 1985;90:580–87 19. Yamanouchi H, Takashima S, Becker LE. Correlation of astrogliosis and substance P immunoreactivity in the brainstem of victims of sudden infant death syndrome. Neuropediatrics 1993;24:200–203

PRENATAL CARBON MONOXIDE EXPOSURE 20. Denoroy L, Gay N, Gilly R, Tayot J, Pasquier B, Kopp N. Catecholamine synthesizing enzyme activity in brainstem areas from victims of sudden infant death syndrome. Neuropediatrics 1987;18: 187–90 21. Levey AI, Kitt CA, Simonds WF, Price DL, Brann MR. Identification and localization of muscarinic acetylcholine receptor proteins in brain with subtype-specific antibodies. J Neurosci 1991;11: 3218–26 22. McGregor HP, Westcott K, Walker DW. The effect of prenatal exposure to carbon monoxide on breathing and growth of the newborn guinea pig. Pediatr Res 1998;43:126–31 23. Mactutus CF. Developmental neurotoxicity of nicotine, carbon monoxide, and other tobacco smoke constituents. Ann N Y Acad Sci 1989;562:105–22 24. Heron H. The effects of smoking during pregnancy: A review with a preview. NZ Med J 1962;61:545 25. Watson RE, Jr., Wiegand SJ, Clough RW, Hoffman GE. Use of cryoprotectant to maintain long-term peptide immunoreactivity and tissue morphology [published erratum appears in Peptides 1986 May- Jun;7(3):545]. Peptides 1986;7:155–59 26. Tolcos M, Rees S. Chronic placental insufficiency in the fetal guinea pig affects neurochemical and neuroglial development but not neuronal numbers in the brainstem: A new method for combined stereology and immunohistochemistry. J Comp Neurol 1997;379: 99–112 27. Morris JL, Gibbins IL, Campbell G, Murphy R, Furness JB, Costa M. Innervation of the large arteries and heart of the toad (Bufo marinus) by adrenergic and peptide-containing neurons. Cell Tissue Res 1986;243:171–84 28. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680–85 29. Mallard C, Tolcos M, Leditschke J, Campbell P, Rees S. Reduction in choline acetyltransferase immunoreactivity but not muscarinicm2 receptor immunoreactivity in the brainstem of SIDS infants. J Neuropathol Exp Neurol 1999;58:255–64 30. Gundersen HJ, Jensen EB. The efficiency of systematic sampling in stereology and its prediction. J Microsc 1987;147:229–63 31. Saper CB, Reis DJ, Joh T. Medullary catecholamine inputs to the anteroventral third ventricular cardiovascular regulatory region in the rat. Neurosci Lett 1983;42:285–91 32. Vincent SR. Distributions of tyrosine hydroxylase-, dopamine-betahydroxylase-, and phenylethanolamine-N-methyltransferase-immunoreactive neurons in the brain of the hamster (Mesocricetus auratus). J Comp Neurol 1988;268:584–99 33. Arango V, Ruggiero DA, Callaway JL, Anwar M, Mann JJ, Reis DJ. Catecholaminergic neurons in the ventrolateral medulla and nucleus of the solitary tract in the human. J Comp Neurol 1988;273: 224–40 34. Blessing WW, Howe PR, Joh TH, Oliver JR, Willoughby JO. Distribution of tyrosine hydroxylase and neuropeptide Y-like immunoreactive neurons in rabbit medulla oblongata, with attention to colocalization studies, presumptive adrenaline-synthesizing perikarya, and vagal preganglionic cells. J Comp Neurol 1986;248:285–300 35. Armstrong DM, Saper CB, Levey AI, Wainer BH, Terry RD. Distribution of cholinergic neurons in rat brain: demonstrated by the immunocytochemical localization of choline acetyltransferase. J Comp Neurol 1983;216:53–68 36. Jones BE, Beaudet A. Distribution of acetylcholine and catecholamine neurons in the cat brainstem: A choline acetyltransferase and tyrosine hydroxylase immunohistochemical study. J Comp Neurol 1987;261:15–32 37. Lan CT, Wen CY, Tan CK, Ling EA, Shieh JY. Multiple origins of cerebellar cholinergic afferents from the lower brainstem in the gerbil. J Anat 1995;186:549–61

227

38. Sugaya K, McKinney M. Nitric oxide synthase gene expression in cholinergic neurons in the rat brain examined by combined immunocytochemistry and in situ hybridization histochemistry. Brain Res Mol Brain Res 1994;23:111–25 39. Wamsley JK, Lewis MS, Young WSd, Kuhar MJ. Autoradiographic localization of muscarinic cholinergic receptors in rat brainstem. J Neurosci 1981;1:176–91 40. Cortes R, Probst A, Palacios JM. Quantitative light microscopic autoradiographic localization of cholinergic muscarinic receptors in the human brain: Brainstem. Neuroscience 1984;12:1003–26 41. Kinney HC, Panigrahy A, Rava LA, White WF. Three-dimensional distribution of [3H]quinuclidinyl benzilate binding to muscarinic cholinergic receptors in the developing human brainstem. J Comp Neurol 1995;362:350–67 42. Mallios VJ, Lydic R, Baghdoyan HA. Muscarinic receptor subtypes are differentially distributed across brain stem respiratory nuclei. Am J Physiol 1995;268:L941–49 43. Donoghue S, Felder RB, Gilbey MP, Jordan D, Spyer KM. Postsynaptic activity evoked in the nucleus tractus solitarius by carotid sinus and aortic nerve afferents in the cat. J Physiol (Lond) 1985;360:261–73 44. Finley JC, Katz DM. The central organization of carotid body afferent projections to the brainstem of the rat. Brain Res 1992;572: 108–16 45. Vardhan A, Kachroo A, Sapru HN. Excitatory amino acid receptors in commissural nucleus of the NTS mediate carotid chemoreceptor responses. Am J Physiol 1993;264:R41–50 46. Withington-Wray DJ, Mifflin SW, Spyer KM. Intracellular analysis of respiratory-modulated hypoglossal motoneurons in the cat. Neuroscience 1988;25:1041–51 47. Newell SJ, Booth IW, Morgan ME, Durbin GM, McNeish AS. Gastro-oesophageal reflux in preterm infants. Arch Dis Child 1989;64: 780–86 48. Olafsdottir E. Gastro-oesophageal reflux and chronic respiratory disease in infants and children: Treatment with cisapride. Scand J Gastroenterol Suppl 1995;211:32–34 49. Jeffery HE, Page M, Post EJ, Wood AK. Physiological studies of gastro-oesophageal reflux and airway protective responses in the young animal and human infant. Clin Exp Pharmacol Physiol 1995;22:544–49 50. Wennergren G, Wennergren M. Neonatal breathing control mediated via the central chemoreceptors. Acta Physiol Scand 1983;119: 139–46 51. Trouth CO, Loeschcke HH, Berndt J. Histological structures in the chemosensitive regions on the ventral surface of the cat’s medulla oblongata. Pflugers Arch 1973;339:171–83 52. Schlafke M, Kille J, Loeschcke H. Elimination of central chemosensitivity by coagulation of a bilateral area on the ventral medullary surface in awake cats. Pflugers Arch 1979;378:231–41 53. Filiano JJ, Choi JC, Kinney HC. Candidate cell populations for respiratory chemosensitive fields in the human infant medulla. J Comp Neurol 1990;293:448–65 54. Czyzyk-Krzeska MF, Bayliss DA, Lawson EE, Millhorn DE. Regulation of tyrosine hydroxylase gene expression in the rat carotid body by hypoxia. J Neurochem 1992;58:1538–46 55. Czyzyk-Krzeska MF, Dominski Z, Kole R, Millhorn DE. Hypoxia stimulates binding of a cytoplasmic protein to a pyrimidine-rich sequence in the 39-untranslated region of rat tyrosine hydroxylase mRNA. J Biol Chem 1994;269:9940–45 56. Raymond R, Millhorn D. Regulation of tyrosine hydroxylase gene expression during hypoxia: Role of Ca21 and PKC. Kidney Int 1997;51:536–41 57. White LD, Lawson EE. Effects of chronic prenatal hypoxia on tyrosine hydroxylase and phenylethanolamine N-methyltransferase messenger RNA and protein levels in medulla oblongata of postnatal rat. Pediatr Res 1997;42:455–62


228

TOLCOS ET AL

58. Schmitt P, Pequignot J, Garcia C, Pujol JF, Pequignot JM. Regional specificity of the long-term regulation of tyrosine hydroxylase in some catecholaminergic rat brainstem areas. I. Influence of longterm hypoxia. Brain Res 1993;611:53–60 59. Soulier V, Dalmaz Y, Cottet-Emard JM, Kitahama K, Pequignot JM. Delayed increase of tyrosine hydroxylation in the rat A2 medullary neurons upon long-term hypoxia. Brain Res 1995;674:188–95 60. Dumas S, Pequignot JM, Ghilini G, Mallet J, Denavit-Saubie M. Plasticity of tyrosine hydroxylase gene expression in the rat nucleus tractus solitarius after ventilatory acclimatization to hypoxia. Brain Res Mol Brain Res 1996;40:188–94 61. Pepin JL, Levy P, Garcin A, Feuerstein C, Savasta M. Effects of long-term hypoxia on tyrosine hydroxylase protein content in catecholaminergic rat brainstem areas: A quantitative autoradiographic study. Brain Res 1996;733:1–8 62. Schmitt P, Soulier V, Pequignot JM, Pujol JF, Denavit-Saubie M. Ventilatory acclimatization to chronic hypoxia: Relationship to noradrenaline metabolism in the rat solitary complex. J Physiol (Lond) 1994;477:331–37 63. Sher PK. Chronic hypoxia in neuronal cell culture: Metabolic consequences. Brain Dev 1990;12:293–300 64. Shimada M. Alteration of acetylcholine synthesis in mouse brain cortex in mild hypoxic hypoxia. J Neural Transm 1981;50:233–45 65. Gibson GE, Peterson C, Sansone J. Decreases in amino acids and acetylcholine metabolism during hypoxia. J Neurochem 1981;37: 192–201 66. Saito K, Honda S, Tobe A, Yanagiya I. Effects of bifemelane hydrochloride (MCI-2016) on acetylcholine level reduced by scopolamine, hypoxia and ischemia in the rats and mongolian gerbils. Jpn J Pharmacol 1985;38:375–80 67. Chleide E, Ishikawa K. Hypoxia-induced decrease of brain acetylcholine release detected by microdialysis. Neuroreport 1990;1: 197–99 68. Ksiezak HJ, Gibson GE. Oxygen dependence of glucose and acetylcholine metabolism in slices and synaptosomes from rat brain. J Neurochem 1981;37:305–14 69. Flavin MP, Yang Y, Riopelle RJ. The effect of hypoxia on neurotransmitter phenotype of forebrain cholinergic neurons. Brain Res 1992;583:201–6 70. Tanaka H, Takahashi S, Miyamoto A, Oki J, Cho K, Okuno A. Effects of neonatal hypoxia on brainstem cholinergic neurons- pedunculopontine nucleus and laterodorsal tegmental nucleus. Brain Dev 1995;17:264–70


71. Burke RE, Kenyon N. The effect of neonatal hypoxia-ischemia on striatal cholinergic neuropil: A quantitative morphologic analysis. Exp Neurol 1991;113:63–73 72. Johnston MV, Hudson C. Effects of postnatal hypoxia-ischemia on cholinergic neurons in the developing rat forebrain: Choline acetyltransferase immunocytochemistry. Brain Res 1987;431:41–50 73. Goldbaum LR, Ramirez RG, Absalon KB. What is the mechanism of carbon monoxide toxicity? Aviat Space Environ Med 1975;46: 1289–91 74. Brown SD, Piantadosi CA. Recovery of energy metabolism in rat brain after carbon monoxide hypoxia. J Clin Invest 1992;89:666–72 75. Verma A, Hirsch DJ, Glatt CE, Ronnett GV, Snyder SH. Carbon monoxide: A putative neural messenger [see comments] [published erratum appears in Science 1994 Jan 7;263(5143):15]. Science 1993;259:381–84 76. Brune B, Ullrich V. Inhibition of platelet aggregation by carbon monoxide is mediated by activation of guanylate cyclase. Mol Pharmacol 1987;32:497–504 77. Brune B, Schmidt KU, Ullrich V. Activation of soluble guanylate cyclase by carbon monoxide and inhibition by superoxide anion. Eur J Biochem 1990;192:683–88 78. Obonai T, Yasuhara M, Nakamura T, Takashima S. Catecholamine neurons alteration in the brainstem of sudden infant death syndrome victims. Pediatrics 1998;101:285–88 79. Felten DL, Hallman H, Jonsson G. Evidence for a neurotropic role of noradrenaline neurons in the postnatal development of rat cerebral cortex. J Neurocytol 1982;11:119–35 80. Lipton SA, Kater SB. Neurotransmitter regulation of neuronal outgrowth, plasticity and survival. Trends Neurosci 1989;12:265–70 81. Levitt P, Harvey JA, Friedman E, Simansky K, Murphy EH. New evidence for neurotransmitter influences on brain development. Trends Neurosci 1997;20:269–74 82. Bear MF, Singer W. Modulation of visual cortical plasticity by acetylcholine and noradrenaline. Nature 1986;320:172–76 83. Mattson MP. Neurotransmitters in the regulation of neuronal cytoarchitecture. Brain Res 1988;472:179–212 84. Hohmann CF, Brooks AR, Coyle JT. Neonatal lesions of the basal forebrain cholinergic neurons result in abnormal cortical development. Brain Res 1988;470:253–64 85. Singer W. Role of acetylcholine in use-dependent plasticity of the visual cortex. In: Steriade M, Biesold D, eds. Brain cholinergic systems. Oxford: Oxford University Press, 1990:314–36 Received October 4, 1999 Accepted November 30, 1999