Differential expression of catecholamine synthetic enzymes in the caudal ventral pons

Differential Expression ofCatecholamine Synthetic Enzymes

in the Caudal Ventral Pons

ANN K. GOODCHILD,1 JACQUELINE K. PHILLIPS,2,3 JANUSZ LIPSKI,2

AND PAUL M. PILOWSKY1*1Hypertension and Stroke Research Laboratories, Departments of Physiology and

Neurosurgery, University of Sydney, Royal North Shore Hospital, Sydney, NSW, Australia2Department of Physiology, University of Auckland, Auckland, New Zealand

3Division of Veterinary and Biomedical Science, Murdoch University, Perth, WA, Australia

ABSTRACTThe analysis of colocalization of multiple catecholamine biosynthetic enzymes within the

ventrolateral part of the medulla oblongata of the rat revealed distinct subpopulations ofneurons within the C1 region (Phillips et al., J Comp Neurol 2001, 432:20–34). In extendingthis study to include the caudal pons, it was shown for the first time that the A5 cell groupcould be distinguished by the presence of immunoreactivity to tyrosine hydroxylase (TH),aromatic l-amino acid decarboxylase (AADC), and dopamine beta hydroxylase (DBH). A novelcell group was also identified. The cells within this new group were immunoreactive to DBHbut not TH, AADC, or phenylethanolamine N-methyltransferase (PNMT) and will be referredto as the TH-, DBH1 cell group. The TH-, DBH1 neurons were not immunoreactive for eitherthe dopamine or noradrenaline transporters, suggesting that these neurons do not take upthese transmitters. A5 neurons were immunoreactive for the noradrenaline transporter butnot the dopamine transporter (as previously shown). Retrograde tracing with cholera toxin Brevealed that the TH-, DBH1 neurons do not project to the thoracic spinal cord or to therostral ventrolateral medulla, but A5 neurons do. A calbindin immunoreactive cell group islocated in a region overlapping TH-, DBH1 cell group. However, only a few neurons wereimmunoreactive for both markers. The physiological role of the TH-, DBH1 cell groupremains to be determined. J. Comp. Neurol. 438:457–467, 2001. © 2001 Wiley-Liss, Inc.

Indexing terms: sympathetic; A5 cell group; noradrenaline; noradrenaline transporters

The presence of primary catecholamines in specific sub-populations of neurons in the central nervous system wasfirst described by Dahlstrom and Fuxe in 1964. They sug-gested that dopamine and noradrenaline accumulated inthese neurons and were released as neurotransmitters.Subsequently, adrenaline was also suggested to be syn-thesized, on the basis of the presence of immunoreactivityfor phenylethanolamine N-methyltransferase, and possi-bly released from the axons of neurons in the C1–C3regions of the medulla oblongata (Hokfelt et al., 1984).Four major enzymes are responsible for catecholaminesynthesis (Cooper et al., 1996): (1) tyrosine hydroxylase(TH), which catalyses the hydroxylation of l-tyrosine andis rate-limiting; (2) l-aromatic amino acid decarboxylase(AADC), which converts dopa to dopamine; (3) dopaminebeta hydroxylase (DBH), which converts dopamine to nor-adrenaline; and (4) phenylethanolamine N-methyl-transferase (PNMT), which converts noradrenaline to

adrenaline. When investigating catecholamine cell groups,it is usual to identify only one of these four major biosyn-thetic enzymes, with an underlying assumption that theneurons contain the full enzyme complement necessaryfor synthesis of the “end point” catecholamine (dopamine,noradrenaline, or adrenaline). For example, this has beenthe case with the A5 cell group, with some investigators byusing only the presence of DBH (Clark and Proudfit, 1993;Wrenn et al., 1996; Tavares et al., 1997; Madden et al.,1999). Others have relied on TH (Kwiat et al., 1993) im-munoreactivity alone to identify this group. However, re-

*Correspondence to: Professor Paul M. Pilowsky, Hypertension andStroke Research Laboratories, Block 3 Ground Floor, Royal North ShoreHospital, St. Leonards, 2065 Australia.E-mail: pilowsky@med.usyd.edu.au

Received 5 February 2001; Revised 18 May 2001; Accepted 26 June 2001

THE JOURNAL OF COMPARATIVE NEUROLOGY 438:457–467 (2001)

© 2001 WILEY-LISS, INC.

cently (Phillips et al., 2001) we reported that a subpopu-lation of neurons in the rostral ventrolateral medulla (theC1 region) was immunoreactive for TH, DBH, and PNMTbut lacked AADC immunoreactivity and failed to expressmRNA for AADC.

Within the spinal cord, medulla oblongata, pons, andhypothalamus, cell groups have been described that areimmunoreactive for AADC but not to other catecholaminebiosynthetic enzymes (Jaeger et al., 1984; Kitahama et al.,1990; Balan et al., 2000). However, AADC occurs in thecytoplasm of many tissues, including the adrenal medulla,kidney, liver, stomach, and brain (Costa et al., 1976; Rioset al., 1999; Flatmark, 2000). This broad distribution in-dicates that the function of AADC is not solely restrictedto monoamine biosynthesis (Cooper et al., 1996; Flatmark,2000). In contrast, in both the peripheral and centralnervous systems, the regional distributions of TH, DBH,and PNMT are closely associated only with catecholaminesynthesis (Pickel et al., 1975; Flatmark, 2000). Curiously,at select sites in both the peripheral and central nervoussystems, some neurons appear to have only one of thesecatecholamine biosynthetic enzymes present (Grzannaand Coyle, 1978; Ross et al., 1984; Foster et al., 1985; Parket al., 1986; Morris and Gibbins, 1987; Smeets and Gonza-lez, 2000). Within the CNS, neurons in the retina, and tworegions of the hypothalamus, are immunoreactive forPNMT but not for other monoamine synthesizing enzymes(Ross et al., 1984; Foster et al., 1985; Park et al., 1986;Simon et al., 1989). No other anomalous cell groups havebeen described within the CNS. In the guinea pig para-cervical ganglion, cells have been identified that are im-munoreactive for DBH but not for any other catechol-amine enzymes. Additionally, these cells do not showcatecholamine fluorescence nor do they take up cat-echolamines (Morris and Gibbins, 1987).

To test further the assumption that neurons containingone of the enzymes involved in catecholamine synthesisalso contain the other “necessary” enzymes for the synthe-sis of “end” products, we have examined neurons in thecaudal pons. The previously defined A5 cells were found tobe immunoreactive for TH, AADC, and DBH, as well asNAT. A new cell group was identified that was immuno-reactive for DBH but not other catecholamine syntheticenzymes. We have also examined the expression of nor-adrenaline and dopamine transporters (NAT and DAT,respectively) and calbindin-D28k (CaBP) in these neu-rons, and tested whether they project to the spinal cord orto the rostral ventrolateral medulla.

METHODS

Immunohistochemistry

All experiments were conducted in accordance with theAustralian and New Zealand code of practice for the careand use of animals for scientific purposes as endorsed bythe National Health and Medical Research Council ofAustralia and the Health Research Council of New Zea-land. All protocols were approved by the Animal Care andEthics Committee of the Royal North Shore Hospital orthe University of Auckland.

Data were obtained from 18 adult male Sprague-Dawleyrats (290–500 g) and 3 juvenile Wistar Kyoto rats (3–5weeks, 55–120 g) of either sex. Animals were anesthetizedwith sodium pentobarbital (Nembutal, 60 mg/kg i.p.) or a

mixture of ketamine and xylazine (42 mg/kg, 5 mg/kgi.m.). An intracardiac injection of 5,000 units of sodiumheparin was followed by perfusion with 300–500 ml ofoxygenated tissue culture medium (DMEM/F12, SigmaD-8900) or saline (0.9% NaCl). Fixation was accomplishedby perfusion with a solution of 4% formaldehyde in sodiumphosphate (100 mM) buffered saline (0.9%) adjusted to pH7.4. Brains, spinal cords, or both, were removed and post-fixed for 1.5–18 hours before coronal sectioning of thebrainstem, thoracic spinal cord, or both, at 50-mm inter-vals, by using a vibrating microtome. Sections were incu-bated for 30 minutes in 50% ethanol in deionized water toenhance antibody penetration and then washed threetimes in 10 mM sodium phosphate buffer (pH 7.4) contain-ing 10 mM Tris and 0.9% NaCl (TPBS) before incubationin antibodies (Goodchild et al., 2000).

In two groups of adult animals, 3 days before perfusion,rats were anesthetized with sodium pentobarbitone (60mg/kg) and the retrograde tracing agent cholera toxin B(1%, CTB; List) was microinjected, by using glass micropi-pettes, into either the T1–2 segments of the spinal cord orinto the rostral ventrolateral medulla. Rats received bilat-eral spinal cord injections each of 200–250 nl. All spinalcord injections were confirmed histologically to be cen-tered around the intermediolateral cell column. For injec-tions into the rostral ventrolateral medulla, the animalswere place in a headholder, the head was flexed, and themuscles overlying the occipital plate and the atlanto-occipital membrane were removed. The membrane, dura,and arachnoid were opened, exposing the dorsal surface ofthe brainstem. To determine the site of tracer injection, aglutamate (100 mM) filled pipette was inserted and a sitegiving a pressor response greater than 40 mm Hg to 50-nlinjections was chosen. This pipette was replaced with onefilled with CTB and 30–40 nl was injected. The animalswere allowed to recover for 24–36 hours and then perfusedas described above.

Tyrosine hydroxylase, AADC, DBH, PNMT, CaBP,NAT, DAT, and CTB were detected immunocytochemi-cally (Table 1). Sections were incubated free-floating for24–72 hours at room temperature in different combina-tions of primary antibodies (Table 1) diluted in TPBS (pH7.4) containing 0.05% merthiolate (Sigma) with 5–10%normal horse or goat serum. After washing in TPBS (threechanges for 30 minutes each), sections were incubated insecondary antibodies (Table 1) in TPBS merthiolate and1–2% normal horse or goat serum for 4–18 hours. Sectionswere again washed, mounted on slides and cover-slippedwith ProLong Antifade (Molecular Probes) or glycerol car-bonate buffer (pH 8.5) containing 0.6% (w/v) phenylenedi-amine.

Microscopy

Sections were examined by using a Leica DML fluores-cence microscope using filter sets (Leica: A, L4, and TXwith excitation filters BP340-380, BP450-490, BP530-595,dichroic mirrors RKP400, RKP510, RKP600, and suppres-sion filters LP425, BP515-560, LP 615, respectively) thatdiscriminated between the fluorophores (FITC, CY3, orTexas Red, and AMCA). No “bleed-through” of fluores-cence from inappropriate fluorophores was observed. Sec-tions were also examined by confocal scanning micros-copy. Confocal microscopy was performed with a LeicaTCS NT confocal system with an argon krypton laser, byusing either a Plan Apochromatic 403/1.25 noradrenaline

458 A.K. GOODCHILD ET AL.

(NA) oil immersion lens, a 633/1.4 NA oil immersion lensor a Plan Fluotar 163/0.5 NA oil immersion lens. Opticalsections were acquired with excitation at 488 nm and 568nm by using a double dichroic beam splitter, a 580-nmdichroic filter, and then an FITC band pass filter. Imageswere averaged and superimposed by using the TCS NTsoftware package.

Analysis

For quantitative analysis, only cell profiles that in-cluded a nucleus and one or more of the catecholaminebiosynthetic enzymes were counted. Analysis of single,dual, or triple labeling was done using 4003 magnifica-tion. In the adults, the rostrocaudal distribution of cate-cholamine biosynthetic enzymes was assessed in the cau-dal ventral pons. The region examined was the ventralquadrant of the rostral medulla and caudal pons. Cau-dally, the region was bounded by the spinal trigeminaltract and the facial nucleus and rostrally by the genu ofthe facial nerve and the midline. Rostrocaudally, the re-gion examined extended from approximately 11 to 9.3 mmcaudal to bregma. All neurons containing one or morecatecholamine synthetic enzymes were counted. For theretrograde tracing studies, only neurons containing cate-cholamine enzymes were assessed for CTB labeling. Datawere obtained from 50-mm-thick sections separated by150 mm.

Digital photographs were taken by using a Spot 2 cam-era (resolution 1,315 3 1,035 pixels, Diagnostic Instru-ments Inc.) attached to a Leica DML microscope. For dual-and triple-labeled images, each image was taken indepen-dently by using the appropriate fluorescence filter sets,artificially colored, and then merged by using commercialsoftware (SPOT).

Figures were assembled and labeled by using Corel-Draw (V8). Overall color balance and contrast of imageswere modified to enhance the quality of the images byusing standard bitmap handling software (SPOT); how-ever, no other adjustments were made.

RESULTS

Rostrocaudal distribution of catecholaminebiosynthetic enzymes in the

caudal ventral pons

In three adult animals, serial brainstem sections includ-ing the rostral half of the facial nucleus and the superiorolive, extending approximately 1.8 mm in the rostrocaudaldirection were subjected to triple labeling by using anti-bodies directed against TH, DBH, and PNMT. The resultswere very similar in all three animals, and the quantita-tive data from three adult animals are summarized inFigure 1A. Very few PNMT-immunoreactive (-IR) neuronswere seen in the defined region. Those that were observedwere probably associated with the more caudally locatedC1 cell group, because they also showed both TH and DBHimmunoreactivities (Howe et al., 1980). The PNMT-IRcells will not be discussed further, because they are thesubject of a previous report (Phillips et al., 2001).

Figure 1A shows that two main cell phenotypes wereseen in the region. One cell type was immunoreactive forboth TH and DBH, whereas the other cell type was im-munoreactive for DBH but not TH. The TH-negative,DBH-positive (TH-,DBH1) neurons extend only approxi-mately 600-mm rostrocaudally whereas the TH1,DBH1neurons, presumably the A5 cells, extend further ros-trally. Throughout the extent of the region examined, theintensity of the TH fluorescence was uniform. In contrast,the intensity of the DBH fluorescence was not (Fig. 1B–E).At the rostral boundary of the facial nucleus two charac-teristic types of DBH immunofluorescence were seen (Fig.1C,D). These two types were composed of a group of mod-erately immunoreactive cells with small, mostly roundedcell bodies and a smaller group of very brightly immuno-fluorescent neurons that had larger cell bodies, were mul-tipolar, and had thick proximal dendrites. The two groupswere intermingled. At more rostral levels, the latter pop-ulation of large, intensely fluorescent cells predominated(Fig. 1E). Dual confocal images of TH and DBH immuno-reactivity in adult animals are shown in Figure 2A,B.Figure 2C shows a merged image from a juvenile rat

TABLE 1. Primary and Secondary Antibodies Used for Immunofluorescent Detection of Catecholaminergic and Calbindin Neuronsin the Caudal Ventral Pons1

Antibody Species Dilution Source

Primary antibodiesAnti-TH

(1) Sheep 1:500 PelFreez(2) Mouse (monoclonal) 1:200 Boehringer Mannheim

Anti-AADC Rabbit 1:200–1:500 ChemiconAnti-DBH

(1) Rabbit 1:50 Chemicon(2) Mouse (monoclonal) 1:500 Chemicon

Anti-PNMT(1) Rabbit 1:5000–1:10000 Prof. P. Howe(2) Sheep 1:5000–1:10000 Prof. P. Howe

Anti-DAT Goat 1:500 Alpha DiagnosticsAnti-NAT Rabbit 1:200 Dr. David ChristieAnti-CaBP Mouse 1:500 Sigma

Secondary/detection antibodies2

FITC Donkey anti-sheep, -rabbit 1:200–1:1000 Sigma, JacksonTexas Red Donkey anti-goat, -sheep, -mouse, -rabbit 1:200–1:1000 Sigma, JacksonAMCA Donkey anti-sheep, -mouse, -rabbit 1:500–1:1000 JacksonBiotinylated Donkey anti-sheep, -mouse, -rabbit 1:500–1:1000 JacksonAvidin (FITC or CY3) 1:500–1:1000 Sigma

1TH, tyrosine hydroxylase; AADC, aromatic amino acid decarboxylase; DBH, dopamine beta hydroxylase; PNMT, phenylethanolamine-N-methyltransferase; FITC, fluoresceinisothiocyanate; DAT, dopamine transporter; NAT, noradrenaline transporter; CaBP, calbindin.2Secondary/detection antibodies were species-specific, affinity-purified, Fab2 fragments.

459CATECHOLAMINE ENZYMES IN THE VENTRAL PONS

Fig. 1. The rostrocaudal distribution of neurons in the ventrolat-eral medulla/pons showing immunoreactivity for the catecholaminebiosynthetic enzymes. A: Distribution of neurons showing immunore-activity for tyrosine hydroxylase (TH), dopamine beta hydroxylase(DBH), and phenylethanolamine-N-methyltransferase (PNMT). Onlytwo main cell groups are present in this region; these groups areTH1,DBH1 and TH-,DBH1 neurons. The mean values and SE areshown from three animals. B: A cartoon parasagittal section of therostral brainstem illustrating the location of the cell groups shown inA, C–F. The location of the facial nucleus (7th), the caudal periolivarynucleus (CPO), and the lateral superior olivary nucleus (LSO) are

shown. C–E: The distribution of DBH immunoreactivity in coronalsection at three rostrocaudal levels as indicated in B. Two populationsof neurons seem to be present. The large angular, intensely immuno-fluorescent neurons with intensely stained processes and the smaller,more faintly immunofluorescent neurons that form a more compactclump located a little more dorsally. F: Distribution of neurons show-ing immunoreactivity for TH and aromatic amino acid decarboxylase(AADC). Only one main cell population was present in this region.These neurons are TH1AADC1. The thin and bold lines representdata from two animals. Scale bar 5250 mm in E (applies to C–E).

Fig. 2. Confocal images showing immunoreactivity for catechol-amine biosynthetic enzymes in cells of the caudal ventral pons.A,A**: A, dopamine beta hydroxylase (DBH) immunoreactivity (red);A9, tyrosine hydroxylase (TH) immunoreactivity (green); A99, themerged image of A and A9. Two cell populations are evident: onecontaining both TH and DBH immunoreactivity (yellow) and oneshowing only DBH immunoreactivity. Many double-labeled processesare also evident in the region. B: Merged image as defined in A99. Twocell populations are evident: one TH1DBH1 and the other that isonly DBH1. C: Merged image as defined in A99. Image taken fromjuvenile rat by using a different antibody against DBH compared withthat used in images A99, B, and D. Processes and terminals within thefield contain DBH and/or TH immunoreactivity. D: Merged image asdefine in A99. Neurons are immunoreactive for DBH but not for TH.

There is also a population of terminals in the region that show im-munoreactivity only for DBH. E,E**: E, DBH immunoreactivity (red);E9, AADC immunoreactivity (green); E99, the merged image of E andE9. In E99 two populations of cells can be seen, those that are DBH1and AADC1 (yellow) and those that are only DBH1. F,F**: F, DBHimmunoreactivity (red); F9, noradrenaline transporter (NAT) immu-noreactivity (green), F99 the merged image of F and F9. The insetshows two cells located in the classically defined A5 region that areDBH1 and NAT1 (yellow in F99). The DBH1 cells in F are sur-rounded by NAT-immunoreactivity processes but themselves do notshow immunoreactivity to NAT. In contrast, many of the DBH-immunoreactive terminals in F are also NAT immunoreactive. Scalebars 5100 mm in A, 50 mm in B,E, 20 mm in C,D,F.

where a different DBH antibody was used (see Table 1). Inall of these cases, a cell group immunoreactive for DBHbut not TH was evident (n 5 10; 7 adult rats and 3 juvenileanimals). Figure 2D shows such cells at a higher magni-fication, and it is clear that, although some other dendriticprocesses and terminal fields were immunoreactive forboth TH and DBH, the cell bodies examined were onlyimmunoreactive for DBH.

The A6 cell group was seen located in the locus coer-uleus in many of the sections examined; however, no sys-tematic analysis of this region was conducted. These cellswere not included in any cell counts as they did not lie inthe region of caudal ventral pons described earlier.

Serial brainstem sections of the caudal ventral ponswere subjected to dual labeling by using antibodiesagainst AADC and TH in two adult rats, and AADC andDBH in one rat. In both rats examined for TH and AADCimmunoreactivity, only neurons immunoreactive for bothTH and AADC were seen (Fig. 1F). In contrast, whereDBH was used in conjunction with AADC, two popula-tions were seen (Fig. 2E,E99). One displayed the phenotypeof AADC1,DBH1 and the other AADC-,DBH1. Thesmall, rounded cell population that showed DBH immu-noreactivity did not show AADC immunoreactivity,whereas the A5 cell group was immunoreactive for TH,DBH, and AADC.

Do TH-, DBH1 neurons contain thenoradrenaline or dopamine transporters?

Sections from two adult animal brainstems were incu-bated with antibodies against DBH and NAT. Figure 2F–F99 shows that the DBH-IR neurons were not NAT-IR butwere surrounded by a very dense plexus of NAT-IR pro-cesses. The inset in Figure 2F shows two A5 neurons thatare very densely covered in NAT puncta. Previously, itwas reported that over 95% of neurons in the A5 cell groupare immunoreactive for NAT (Comer et al., 1998). Inagreement with this finding, most A5 neurons examinedin the present study were strongly NAT-IR. Sections fromtwo adult brainstems were incubated with antibodiesagainst DAT. Dopamine transporter immunoreactivitywas not observed in the TH-, DBH1 cell in the caudalventral pons, but was found in the nucleus tractus soli-tarius (NTS) and in the A6 cell group in the locus coer-uleus (data not shown).

Do TH-, DBH1 neurons projectto the spinal cord?

It is well established that the A5 cell group projects tothe thoracic spinal cord and synapses with preganglionicneurons (Clark and Proudfit, 1993). We examined poten-tial thoracic spinal cord projections for the TH-, DBH1cell group in three animals. The retrograde tracer CTBwas injected into the T1–2 thoracic segments. Regions ofthe brainstem containing the neurons of interest werethen subjected to dual labeling with antibodies directedagainst DBH and CTB. This labeling was quantified intwo animals (Fig. 3A). Two populations of DBH-IR neu-rons existed, one that was both CTB-IR and DBH-IR andone that was immunoreactive only for DBH. OccasionallyCTB-IR neurons were found that were not DBH-IR. Im-ages showing dual CTB and DBH labeling are shown inFigure 3B. It is clear that the A5 cells contained both CTBand DBH immunoreactivities; however, the small, round

DBH-IR cells that were intermingled with the A5 cellswere not CTB-IR.

Do TH-, DBH1 neurons project to theventrolateral medulla?

Four adult animals were used to determine whether ornot the TH-, DBH1 cell population projected to the rostralventrolateral medulla. Injections of CTB were made intosites within the ventrolateral medulla where microinjec-tion of l-glutamate 50 nl elicited a rise in arterial pressureof 54 6 5 mm Hg. In Figure 4A,B many large angularDBH-IR neurons are seen that contain CTB; however, thesmaller weakly fluorescent cell group intermingled withthe A5 cells did not contain CTB retrogradely transportedfrom the rostral ventrolateral medulla.

Do TH-, DBH1 neurons colocalize CaBP?

To determine whether or not the TH-, DBH1 cell pop-ulation was immunoreactive for CaBP, sections of interestwere incubated with antibodies directed against CaBPand DBH. This experiment was performed in three ani-mals (Fig. 5). As we have previously reported, the A5 cellswere not CaBP-IR (Goodchild et al., 2000). A CaBP-IR cellpopulation does overlap the region containing the TH-,DBH1 neurons; however, only one or two neurons persection were immunoreactive for both CaBP and DBH (seearrows in Fig. 5).

DISCUSSION

There were two major findings in this study. First, weshow for the first time that neurons of the A5 cell groupcontain all the enzymes required to synthesize noradren-aline (i.e., TH, AADC, and DBH). We also confirm thatthey contain NAT but not DAT and project to the rostralventrolateral medulla and the spinal cord. Second, a cellpopulation was identified in the caudal ventral pons thatis immunoreactive for DBH but not for any other catechol-amine biosynthetic enzyme. This cell population is neitherNAT-IR nor DAT-IR, nor do the majority of cells show anyCaBP-IR. Although the axonal projections of these cellsremain undefined, we have established that they do notproject to the thoracic spinal cord or to the rostral ventro-lateral medulla.

Distribution of catecholamine biosyntheticenzymes in the caudal ventral pons

In the present study, two populations of neurons wereidentified in the caudal ventrolateral pons of the adult andjuvenile rat: TH1, DBH1, and TH-, DBH1 neurons. Theuse of two different antibodies against DBH obtained fromseparate sources, indicates that the DBH immunoreactiv-ity is specific. The possibility that this cell group containsa protein immunologically similar to, but functionally dis-tinct from, DBH cannot be excluded. Thus, single cellreverse transcriptase-polymerase chain reaction or in situhybridization is needed to confirm that the cells containauthentic DBH message. Furthermore, only the TH-IRneurons also showed additional immunoreactivity forAADC. As described previously, no neurons in the regionwere PNMT-IR (Howe et al., 1980). The cell group identi-fied as immunoreactive for TH, AADC, and DBH is clearlythe previously defined A5 cell group. However, the neu-rons immunoreactive for DBH alone comprise a previously

462 A.K. GOODCHILD ET AL.

Fig. 3. Neurons in the caudal ventral pons that are dopaminebeta hydroxylase immunoreactive (DBH-IR) and project to thethoracic spinal cord. A: Rostrocaudal distribution patterns. Thethick and thin lines represent data from two animals. Two cellpopulations are seen in the region, one that is cholera toxin Bpositive (CTB1), DBH1 and the other that is just DBH1. The cellswere defined as DBH containing initially, and then their spinalprojection was determined. Some CTB1, DBH- neurons were also

evident in the region. B: Example of DBH- and CTB-immuno-reactive (-IR) neurons in a section of the caudal ventral pons. Twogroups of DBH-IR neurons are present, one intensely immunoflu-orescent and one weakly immunofluorescent. Double-labeled neu-rons are marked with asterisks. None of the weakly fluorescentneurons are CTB immunoreactive. There are a few CTB-IR neuronsthat are not DBH-IR (arrows). Scale bar 5 50 mm in B.

463CATECHOLAMINE ENZYMES IN THE VENTRAL PONS

undefined cell group that we shall refer to as the TH-,DBH1 cell group. The A5 and the TH-, DBH1 cell groupsare also clearly distinguishable on morphologic grounds.The A5 cells were of medium size and often multipolar asdescribed previously (Dahlstrom and Fuxe, 1964),whereas the neurons of the more dorsally located TH-,DBH1 cell group had smaller, circular soma that wereoften bi- or tripolar. The TH-, DBH1 neurons were found

only at the level of the most caudal A5 neurons extendingfor only approximately 600 mm rostrocaudally, whereasthe A5 cell group extends 1.6–2.0 mm (Dahlstrom andFuxe, 1964; Byrum et al., 1984). The catecholamine syn-thesizing properties of these cells remains to be clarified,as does the role of DBH in these cells. Although no otheramine synthesizing enzymes were detectable under theconditions of the present study, it is possible that, in other

Fig. 4. Projections from the caudal ventral pons to the rostralventrolateral medulla. Dopamine beta hydroxylase-immunoreactive(DBH-IR) neurons and neurons with rostral ventrolateral medulla(RVLM) projections are illustrated. A,B: Sections from two differentexperiments. The injection sites within the RVLM are shown in theinsets. Panels on the left hand side show DBH immunoreactivity, and

those on the right hand side show neurons that project to the RVLM(cholera toxin B [CTB]). Double-labeled neurons are indicated withasterisks. None of the faintly fluorescent DBH-IR neurons are CTB-IR. Neurons are present that are not DBH immunoreactive but areCTB-IR (arrows). NA, nucleus ambiguus. Scale bar 5 50 mm in B(applies to A,B).

464 A.K. GOODCHILD ET AL.

circumstances, such enzymes could be induced or arepresent at levels undetectable by these methods but insufficient quantities to synthesize catecholamines.

Our finding also emphasizes the caution that should beapplied when interpreting data when only one catechol-amine biosynthetic enzyme is used to define a population.Although many studies have defined the A5 cell group byusing antibodies directed against TH (Kwiat et al., 1993;Arvidsson et al., 1995; Bajic and Proudfit, 1999; Huangand Weiss, 1999) or DBH (Grzanna et al., 1987; Thor andHelke, 1988; Clark and Proudfit, 1993; Tavares et al.,1997; Madden et al., 1999) or AADC (Jaeger et al., 1984)no previous studies have simultaneously used antibodiesdirected against more than one catecholamine syntheticenzyme in identifying the A5 cell group, at least in mam-mals (Smeets and Gonzalez, 2000).

Role of DBH in the absence of othercatecholamine synthesising enzymes

The presence of DBH immunoreactivity indicates thatneurons could produce noradrenaline from dopamine.Cells of the A5 group also contain AADC, and these neu-rons are almost certainly the ones that typically showcatecholamine fluorescence in this region (Dahlstrom andFuxe, 1964).

However, it is possible that neurons such as the TH-,DBH1 cells can make noradrenaline by using alternatesynthetic pathways. These pathways include (a) the pro-duction of L-dopa after conversion of tyrosine by tyrosi-nase, although AADC would still be required to convertdopa to dopamine (Rios et al., 1999) or (b) the productionoctopamine from hydroxylation of tyramine by DBH (apathway that predominates in insects; Klemm et al., 1985)although AADC is still needed to convert tyrosine to tyra-mine. Because TH-, DBH1 neurons are not AADC-IR, it isunlikely that either of these alternate catecholamine bio-synthesis pathways are used.

The presence of other catecholamine synthetic enzymeswould not be necessary to synthesize noradrenaline if theTH-, DBH1 neurons were able to take up dopamine effi-ciently. The noradrenaline transporter (NAT) transportsnot only noradrenaline but also dopamine and adrenaline(Buck and Amara, 1994). The dopamine transporter(DAT) shows 67% amino acid identity with human NATbut does not transport noradrenaline efficiently (Cooper etal., 1996). Results of the present study show that immu-noreactivity for NAT is present on A5 cells (as has beendemonstrated previously; Comer et al., 1998) but not onthe TH-, DBH1 cells, whereas DAT is not present in theregion. These data indicate that dopamine is not trans-ported into TH-, DBH1 cells for later conversion intonoradrenaline.

If DBH in the TH-, DBH1 neurons is not utilized for thesynthesis of catecholamines what is its role in these cells?One proposal is that it may be an “artefact of develop-ment”, as in sympathetic ganglia, where DBH is found inboth TH1 and TH- cells during development (Ernsbergeret al., 2000). It may also simply be expression of a “super-fluous” protein (Bowers, 1994). Alternatively, DBH hasvery low substrate specificity converting any phenylethyl-amine to its corresponding phenylethanolamine (as in “b”above) and, thus, may be involved in some other pathway.

Do TH-, DBH1 cells project to the spinalcord or rostral ventrolateral medulla?

There is substantial evidence that the A5 cell groupprojects to both the thoracic spinal cord (Loewy et al.,1979, 1986; Blessing et al., 1981; Westlund et al., 1984;Kwiat and Basbaum, 1990; Clark and Proudfit, 1993; Ba-jic and Proudfit, 1999) and the rostral ventrolateral me-dulla (RVLM; Byrum and Guyenet, 1987). In such studies,catecholamine fluorescence, TH immunoreactivity, or ret-rograde transport of an antibody against DBH was used toidentify the A5 cell group. Our results agree with these

Fig. 5. Dopamine beta hydroxylase (DBH) and calbindin (CaBP) immunoreactivity in the caudalventral pons. Few DBH-immunoreactive (-IR) neurons are CaBP-IR (CaBP-IR). None of the brightlyfluorescent DBH-IR neurons are CaBP-IR but a few of the weakly fluorescent DBH-IR neurons are (blackand white arrows). Scale bar 5 50 mm.

465CATECHOLAMINE ENZYMES IN THE VENTRAL PONS

studies. However, Byrum et al. (1984) also identified cellsin the caudal A5 region that were spinally projecting butwere devoid of catecholamine fluorescence. This findingmay be the same group of spinally projecting neuronsidentified in the current study that were devoid of anycatecholamine synthetic enzymes. In contrast, the TH-,DBH1 cell group did not project to the upper thoracicspinal cord.

There is some confusion in the literature concerning thepossibility of A5 projections to the ventral medulla (By-rum and Guyenet, 1987; Tavares et al., 1997; Madden etal., 1999). The most recent evidence suggests that A5 toRVLM projections do exist, because anti–DBH-saporin re-sults in the death of A5 cells after injection into the C1region (Madden et al., 1999). The precise region or cells inthe ventral medulla that receive A5 input is unknown.The results of the present study confirm that A5 cellsproject to the rostral part of the ventral medulla but thatthe TH-, DBH1 neurons do not. Interestingly, Tavares etal. (1997) demonstrated a projection to the caudal ventralmedulla from DBH-IR neurons in the A5 region; therefore,projection of the TH-, DBH1 neurons to this area remainsa possibility.

Do TH-, DBH1 cells contain calbindin?

A striking finding in a previous study was that a smallbut distinct population of neurons in the caudal ventralpons contained the calcium binding protein calbindin(Goodchild et al., 2000). We showed that, although it wasadjacent to the A5 cell group, the calbindin neurons werenot TH-IR. Results of the present study indicate that onlyvery few of the TH-, DBH1 neurons are calbindin-IR. Thefunctional significance of the presence of calbindin is un-known at present but perhaps indicates that within theTH-, DBH1 cell group subpopulations of neurons exist.

Functional implications

The differences in the projection sites, as well as enzy-matic content, of the A5 and TH-, DBH1 cell groupsindicates a difference in function between these cellgroups. Previous studies support a role for A5 neurons inmediating chemoreceptor reflex function (Koshiya andGuyenet, 1994). However, our projection data do not sup-port a role for the TH-, DBH1 cells in cardiovascular orother sympathetic functions. Functions attributed to thisregion include modulation of respiration (Jodkowski et al.,1997), lacrimation, and salivation (Toth et al., 1999). Fu-ture studies are required to elucidate a role for the TH-,DBH1 cell group. Finally, this study in conjunction withour previous work (Phillips et al., 2001) suggests a cau-tious approach when defining a cell population by using asingle marker.

ACKNOWLEDGMENTS

The authors thank Tina Stasinopoulos and ElizabethMoon for valuable expert assistance. The authors appre-ciate the kind gift of the antibody against the noradrena-line transporter from Dr. David Christie of the Universityof Auckland, New Zealand. Grant support: NHMRC,NHF, and NSHRF.

LITERATURE CITED

Arvidsson U, Dado RJ, Reidl M, Lee J-H, Law PY, Loh HH, Elde R,Wessendorf MW. 1995. Delta-opioid receptor immunoreactivity: distri-

bution in brainstem and spinal cord, and relationship to biogenicamines and enkephalin. J Neurosci 15:1215–1235.

Bajic D, Proudfit HK. 1999. Projections of neurons in the periaqueductalgray to pontine and medullary catecholamine cell groups involved inthe modulation of nociception. J Comp Neurol 405:359–379.

Balan IS, Ugrumov MV, Calas A, Mailly P, Krieger M, Thibault J. 2000.Tyrosine hydroxylase-expressing and/or aromatic L-amino aciddecarboxylase-expressing neurons in the mediobasal hypothalamus ofperinatal rats: differentiation and sexual dimorphism. J Comp Neurol425:167–176.

Blessing WW, Goodchild AK, Dampney RAL, Chalmers JP. 1981. Cellgroups in the lower brain stem of the rabbit projecting to the spinalcord, with special reference to catecholamine-containing neurons.Brain Res 221:35–55.

Bowers CW. 1994. Superfluous neurotransmitters. TINS 17:315–320.Buck KJ, Amara SG. 1994. Chimeric dopamine-norepinephrine transport-

ers delineate structural domains influencing selectivity for cat-echolamines and l-methyl-4-phenylpyridinium. Proc Natl Acad SciUSA 91:12584–12588.

Byrum CE, Guyenet PG. 1987. Afferent and efferent connection of the A5noradrenergic cell group in the rat. J Comp Neurol 261:529–542.

Byrum CE, Stornetta R, Guyenet PG. 1984. Electrophysiological propertiesof spinally-projecting A5 noradrenergic neurons. Brain Res 303:15–29.

Clark FM, Proudfit HK. 1993. The projections of noradrenergic neurons inthe A5 catecholamine cell group to the spinal cord in the rat: anatom-ical evidence that A5 neurons modulation nociception. Brain Res 616:200–210.

Comer AM, Qi J, Christie DL, Gibbons HM, Lipski J. 1998. Noradrenalinetransporter expression in the pons and medulla oblongata of the rat:localisation to noradrenergic and some C1 adrenergic neurones. MolBrain Res 62:65–76.

Cooper JR, Bloom FE, Roth RH. 1996. The biochemical basis of neurophar-macology. New York: Oxford University Press.

Costa M, Furness JB, McLean JR. 1976. The presence of aromatic L-aminoacid decarboxylase in certain intestinal cells. Histochemistry 48:129–143.

Dahlstrom A, Fuxe K. 1964. Evidence for the existence of monoamine-containing neurons in the central nervous system. Acta Physiol Scand62:2–55.

Ernsberger U, Reissmann E, Mason I, Rohrer H. 2000. The expression ofdopamine beta-hydroxylase, and Phox2 transcription factors in sympa-thetic neurons: evidence for common regulation during noradrenergicinduction and diverging regulation later in development. Mech Dev92:169–177.

Flatmark T. 2000. Catecholamine biosynthesis and physiological regula-tion in neuroendocrine cells. Acta Physiol Scand 168:1–17.

Foster GA, Hokfelt T, Coyle JT, Goldstein M. 1985. Immunohistochemicalevidence for phenylethanolamine-N-methyltransferase-positive/tyrosinehydroxylase-negative neurones in the retina and the posterior hypo-thalamus of the rat. Brain Res 330:183–188.

Goodchild AK, Llewellyn-Smith IJ, Sun Q-J, Chalmers J, Cunningham A,Pilowsky PM. 2000. Calbindin-immunoreactive neurons in the reticu-lar formation of the rat brainstem: catecholamine content and spinalprojections. J Comp Neurol 424:547–562.

Grzanna R, Coyle JT. 1978. Dopamine-b-hydroxylase in rat submandibu-lar ganglion cells which lack norepinephrine. Brain Res 151:206–214.

Grzanna R, Chee WK, Akeyson EW. 1987. Noradrenergic projections tobrainstem nuclei: evidence for differential projections from noradren-ergic subgroups. J Comp Neurol 263:76–91.

Howe PRC, Costa M, Furness JB, Chalmers JP. 1980. Simultaneous dem-onstration of phenylethanolamine N-methyltransferase immunofluo-rescent nerve cell bodies in the rat medulla oblongata. Neuroscience5:2229–2238.

Hokfelt T, Johannsson O, Goldstein M. 1984. Central catecholamine neu-rons as revealed by immunohistochemistry with special reference toadrenaline neurons. In: Bjorkland A, Hokfelt T, editors. Handbook ofchemical neuroanatomy. Amsterdam: Elsevier. p 157–276.

Huang J, Weiss ML. 1999 Characterization of the central cell groupsregulating the kidney in the rat. Brain Res 845:77–91.

Jaeger CB, Ruggiero DA, Albert VR, Park DH, Joh TH, Reis DJ. 1984.Aromatic L-amino acid decarboxylase in the rat brain: immunocyto-chemical localization in neurons of the brain stem. Neuroscience 11:691–713.

Jodkowski JS, Coles SK, Dick TE. 1997. Prolongation in expiration evokedfrom the ventrolateral pons of adult rats. J Appl Physiol 82:375–376

466 A.K. GOODCHILD ET AL.

Kitahama K, Denoyer M, Raynaud B, Borri-Voltattorni C, Weber M, Jou-vet M. 1990. Aromatic L-amino acid decarboxylase-immunohisto-chemistry in the cat lower brainstem and midbrain. J Comp Neurol302:935–953.

Klemm N, Nassel DR, Osborne NN. 1985. Dopamine- b-hydroxylase-likeimmunoreactive neurons in two insect species, Calliphora erythro-cephala and Periplaneta americana. Histochemistry 83:159–164.

Koshiya N, Guyenet PG. 1994. A5 noradrenergic neurons and the carotidsympathetic chemoreflex. Am J Physiol 267:R519–R526.

Kwiat GC, Basbaum AI. 1990. Organization of tyrosine hydroxylase- andserotonin-immunoreactive brainstem neurons with axon collaterals tothe periaqueductal gray and the spinal cord in the rat. Brain Res528:83–94.

Kwiat GC, Liu H, Williamson AM, Basbaum AI. 1993. GABAergic regula-tion of noradrenergic spinal projection neurons of the A5 cell group inthe rat: an electron microscopic analysis. J Comp Neurol 330:557–570.

Loewy AD, McKellar SD, Saper CB. 1979. Direct projections from the A5catecholamine cell group to the intermediolateral cell column. BrainRes 174:309–314.

Loewy AD, Marson AD, Parkinson D, Perry MA, Sawyer WA. 1986. De-scending noradrenergic pathways involved in the A5 depressor re-sponse. Brain Res 386:313–324.

Madden CJ, Ito S, Rinaman L, Wiley RG, Sved AF. 1999 Lesions of the C1catecholaminergic neurons of the ventrolateral medulla in rats usinganti-DbH-saporin. Am J Physiol 277:R1063–R1075.

Morris JL, Gibbins IL. 1987. Neuronal colocalization of peptides, cat-echolamines, and catecholamine-synthesizing enzymes in guinea pigparacervical ganglia. J Neurosci 7:3117–3130.

Park DH, Enrlich M, Evinger MJ, Joh TH. 1986. Strain differences indistribution of phenylethanolamine n-methyltransferase activity fromrat brain and adrenal gland. Brain Res 372:185–188.

Phillips JK, Goodchild AK, Dubey R, Sesiashvili E, Takeda M, Chalmers J,Pilowsky PM, Lipski J. 2001. Differential expression of catecholaminebiosynthetic enzymes in the rat ventrolateral medulla. J Comp Neurol432:20–34.

Pickel VM, Joh TH, Reis DJ. 1975. Ultrastructural localization of tyrosinehydroxylase in noradrenergic neurons of brain. Proc Natl Acad Sci USA72:659–663.

Rios M, Habecker B, Sasaoka T, Eisenhofer G, Tian H, Landis S, Chikara-ishi D, Roffler-Tarlov S. 1999. Catecholamine synthesis is mediated bytyrosinase in the absence of tyrosine hydroxylase. Neuroscience 19:3519–3526.

Ross CA, Ruggiero DA, Meeley MP, Park DH, Joh TH, Reis DJ. 1984. Anew group of neurons in hypothalamus containing phenylethanol-amine N-methyltransferase (PNMT) but not tyrosine hydroxylase.Brain Res 306:349–353.

Simon A, Versaux-Botteri C, Denoroy L, Vigny A, Nguyen-Legros J. 1989.Double antigen localization of two catecholamine enzymes and GABAin amacrine cells of the rat retina in semi-thin sections. J NeurosciMethods 27:181–189.

Smeets WJ, Gonzalez A. 2000. Catecholamine systems in the brain ofvertebrates: new perspectives through a comparative approach. BrainRes Rev 33:308–379.

Tavares I, Lima D, Coimbra A. 1997. The pontine A5 noradrenergic cellswhich project to the spinal cord dorsal horn are reciprocally connectedwith the caudal ventrolateral medulla in the rat. Eur J Neurosci9:2452–2461.

Thor KB, Helke CJ. 1988. Catecholamine-synthesizing neuronal projec-tions to the nucleus tractus solitarii, of the rat. J Comp Neurol 268:264–280.

Toth IE, Boldogkoi Z, Medveczky I, Palkovits M. 1999. Lacrimal pregan-glionic neurons form a subdivision of the superior salivatory nucleus ofrat: transneuronal labelling by pseudorabies virus. J Auton Nerv Syst7:45–54.

Westlund KN, Bowker RM, Ziegler MG, Coulter JD. 1984 Origins andterminations of descending noradrenergic projections to the spinal cordof monkey. Brain Res 292:1–16.

Wrenn CC, Picklo MJ, Lappi DA, Robertson D, Wiley RG. 1996. Centralnoradrenergic lesioning using anti-DBH-saporin: anatomical findings.Brain Res 740:175–184.

467CATECHOLAMINE ENZYMES IN THE VENTRAL PONS