Journal of Chemical Neuroanatomy, Vol. 5:427 440 (1992)

Immunohistochemical Study of the Catecholaminergic Innervation of the Spinal Cord of the Rat using Specific Antibodies against Dopamine and Noradrenaline P. Mouchet*, M. Manier and C. Feuerstein Laboratoire de Physiologie section Neurophysiologie, INSERM U.318, Pavilion de Neurologie, CH U de Grenoble, BP217, 38043 Grenoble cedex 09, France

ABSTRACT We have assessed the relative contributions of dopaminergic and noradrenergic descending systems to the catecholaminergic innervation of the rat spinal cord. Fibres and terminals were labelled with their own neurotransmitter by using specific antibodies raised against dopamine (DA) and noradrenaline (N A) respectively. For this purpose, immunohistochemistry according to the peroxidase anti-peroxidase technique was performed in different experimental conditions. Two group of rats received intracisternal 6-hydroxy-dopamine (6-OHDA) injections either with or without benzatropine pretreatment. Animals of a third group were not pretreated at all. While 6-OHDA induced a complete disappearance of spinal NA-like immunoreactivity (NA-LI), except for scarce residual fibres in the thoracic intermedio-lateral cell column, DA-like immunoreactivity (DA-LI) was unaffected by the lesion. This strongly suggests that the antisera used specifically labelled NA-containing and DA-containing fibres respectively. Spinal DA-L1 and NA-LI innervations differed markedly in their topographical distributions and in thc morphology of the corresponding fibres. DA-LI innervation was restricted to laminae I, 111 and IV and to the intermediate zone, especially the autonomic areas. In the ventral horn, it was sparse and more visible after acidification of the fixation solution. NA-LI innervation was much more widely spread. In addition, the organization of NA-LI fibres suggests that the innervation of the whole dorsal horn comes from a group of fibres travelling, at least partially, in the superficial dorsal horn. KEYWORDS" Descending pathways

Catecholaminergic systems

INTRODUCTION Since the beginning of the histochemical mapping of monoaminergic neurotransmitter systems, the spinal cord has been known to have a rich catecholaminergic innervation (Carlsson et al., 1964; Dahlstram and Ffixe, 1965). Catecholaminergic terminals were present in all parts of the spinal grey matter but with a higher density in the superficial dorsal horn and within autonomic and m o t o r nuclei. This innervation was also ascribed exclusively to fibres descending from the brainstem (Dahlstr6m and Ffixe, 1965). These observations were made using histofluorescence techniques which could not distinguish noradrenaline (NA) from dopamine (DA). So far, only biochemical data have been used to identify more exactly the neurotransmitter(s) used by these systems. Because spinal NA levels are nearly t e n times higher than DA levels (Magnusson and *Addressee for correspondence: P. Mouchet, Laboratoire de Physiologie seclion Neurophysiologie, INSERM U.318, Pavilion de Neurologie, CHU de Grenoble, BP 217, 38043 Grenoble cedex 09, France.

0891 0618/92/050427 14512.00 © 1992 by John Wiley and Sons Ltd

Rosengren, 1963; Laverty and Sharman, 1965), NA was claimed to be the only catecholamine transmitter and DA was thought to be its metabolic precursor in this part of the central nervous system, This idea was strengthened by the fact that biochemical assays carried out on spinal cord tissue could only detect adrenaline levels near the background range. Later, it was shown that adrenaline (H6k felt et al., 1974) and dopamine (Commissiong et al., 1978a) were also spinal neurotransmitters or modulators. Revisions were also made concerning the supraspinal origin of the fibres involved: noradrenergic fibres come from the pons (Westlund et al.. 1981), dopaminergic fibres from the hypothalamus (Bj6rklund and Skagerberg, 1979: Blessing and Chalmers, 1979; H6kfelt et al., 1979) and adrenergic fibres from the so-called C~ medullary cell group (Ross et al., 1981 ). However, all these results were obtained with indirect methods which did not differentiate DA from NA (the adrenaline is at the limit of detection) or which rely upon the detection of other compounds such as the enzymes of the catecholamine

DA

NA B

m

i!~ ~i ~iiI

~i

~ i~¸¸~II ii

i

:

;i,,

•

~!':~

" ~ ~i

Fig. 1. Consequences of 6-OHDA lesions as illustrated on transversal sections made at cervical level, (A) Non-treated animal, DA immunohistochemistry. (B) Non-treated animal, NA immunohistochemistry. (C) 6-OHDA-lesioned animal, DA immunohistochemistry. (D) 6-OHDA-lesioned animal, NA immunohistochemistry. The dorsal direction is upward. Scale-bar: 140 I~m.

Rat Spinal Cord Catecholaminergic Innervation

DA

429

NA

Fig. 2. MorphologicaldifferencesbetweenDA-LI (A and C) and NA-LI (B and D) fibreslocatedin the dorsal horn as evidencedon sagittal thoracic sections. Arrowheadsin (A) and (C) point to some appendagesdistinctiveof DA-LI axons. Scale-bar:30 ~m. synthesis pathway. More recently, highly specific antibodies raised against DA (Chagnaud et at., 1987) and against NA (Geffard et al., 1986) became available. This prompted us to undertake the immunohistochemical study of rat spinal dopaminergic and noradrenergic systems labelled with their own neurotransmitters in order to get additional information about their topography and their differences.

the second group (ten animals) received 6-OHDA only. In a third group (eight rats), animals received no treatment at all. In addition, six rats were used for the examination of the longitudinal distribution of the catecholaminergic innervation. Three were treated with 6O H D A as stated above and the others received no treatment.

lmmunohistochemistry MATERIALS AND METHODS

Animals Experiments were performed on male Wistar rats weighing from 250 to 300 g. Two groups of animals received intracisternal 6-hydroxy-dopamine (6OHDA) injections (133 lag of 6-OHDA base in 20 lal of NaCI 0.9% containing 2% ascorbic acid). The rats of the first group (eight animals) were pretreated with benzatropine, a specific DA uptake inhibitor, at 25mg/kg given intraperitoneally 45 rain before 6-OHDA injection. The animals of

Animals were anaesthetized and perfused (10 rain) through the ascending aorta with 20 ml isotonic saline followed by 600 ml of 5% glutaraldehyde in 0.1 M-phosphate buffer (pH 7.4) containing 1% sodium metabisulphite. The spinal cord was then removed and blocks of tissue were taken at cervical (upper spinal cord and cervical enlargement), thoracic, lumbar and sacral levels. These blocks were immersed for 30 min at 4°C in the same fixative as that used for the perfusion. The blocks were then rinsed overnight at 4'C in phosphate buffer containing 20% sucrose and 1% sodium metabisulphite before being frozen.

DA

NA

Fig. 3. Spinal cord DA-LI (left) and NA-LI (right) innervations as visualized on transverse sections. (A) Thoracic dorsal horn. distal part (laminae I IV). Arrowhead points to a cluster of fibres descending within the dorso-lateral white matter. The arrows indicate the position of the superficial laminae. (B) Thoracic dorsal horn, laminae 1 Ill (indicated by the arrows). The medial direction is to the left. (C,D) Laminae Ill IV. The medial direction is to the left for (C), to the right for (D). (E) Thoracic spinal gray at T12 level. Arrows point to the intermedio-lateral cell columns. The dorsal direction is upward. (F) Thoracic spinal gray at T5 level. Arrow points to the intermedio-

DA

NA

A i

J

'0

) .

i

~

+;~,~

D

I

J J\

t

"-\

\

!

+.

Y.

F,

"

.pJ

•~

;o~'}

+J .•..

°

~*"'#"

t

++~

i

.4

t~+,,

+'t

t -.

\_

j-

-) ,

J

¢-

~,,+

-.. .... ,';',

+~',

,:"

-

] +, +.

-

i,+

t, +

•' %

"g +'J"

e

d

~ t

. ~

4-

,¢~ t

1~

Fig. 4. D A - L I (left) and N A - L I (right) i n n e r v a t i o n of some spinal a u t o n o m i c areas. (A and B) L a m i n a X at l u n l b a r (L3) Ic', el. ((' and DI Z o n e dorsal to the central canal at sacral (S2) level. (E and F) P a r a s y m p a t h e t i c nucleus al the S t level. The medial d i r e c l h m is u'ighba ard A s t e r i s k s indicate the central canal. Scale bar: 40 I.tm.

,a.

,

B

....

F

Fig. 5. Organization of NA-LI axons in the dorsal horn as illustrated on longitudinal sections. (A and C) Lumbar spinal cord (L3 L6), sagittal sections. (E) Thoracic segments (T4-T8), sagittal section. Arrows point to fibres leaving the superficia! part of the dorsal horn and going deeper in the horn. Some of them (C and E) are axonal collaterals. (B,D,F) Horizontal sections taken at the lumbar level (L3 L6). (D) Shows the same elements as seen in (B), but at higher magnification. Scale bar: 25 I,tm.

Rat Spinal Cord Catecholaminergic Innervation For some rats in each group these first steps were modified according to Holstege et al. (1990) by substituting 0.1 M-acetate buffer (pH 4) for phosphate buffer in the fixative and in the sucrose solution. For each spinal level, each block was cut according to one of three planes of section, transverse, horizontal or sagittal, with a microtome cryostat at - 20"C. All sections (20 ~tm thick) were transferred into Tris saline buffer 0.05 M containing 1% sodium metabisulphite (TBM). All steps of the immunohistochemical procedure were performed on free-floating sections. Sections were first incubated in a solution of 1% sodium borohydride in TBM for 30 min, followed by six washings in TBM (10 min each) and further incubation in TBM containing 3% normal rabbit serum (for DA immunohistochemistry) or 3% normal swine serum (for NA immunohistochemistry). They were then incubated for 12h or 35h at 4°C with antibodies against DA ( 1: l 0 000) or NA (1:15 000) in TBM containing 0.3% Triton X-100 and I% normal serum. Then sections were washed and incubated (1.5 h) in a rabbit anti-mouse IgG serum (Dakopatts, 1: 100) or a swine anti-rabbit IgG serum (Dakopatts, 1 : 100) diluted in saline 0.05 M-Tris-HC1 containing 0.3% Triton X-100 and 1% normal serum. The subsequent incubation was then performed in a mouse peroxidase anti-peroxidase (PAP) complex (Dakopatts, 1:100) or in a rabbit PAP complex (Dakopatts, 1:400)for 1 h. Lastly, sections were reacted with 0.05% diaminobenzidine (Sigma), 0.01% hydrogen peroxide (Merck) and imidazole (Sigma; 0.01 M final concentration) in 0.05 M-Tris-HCI buffer, pH 7.5. Alter a last rinse in phosphate-buffer saline, sections were mounted onto gelatine-coated slides, dehydrated in various ethanol baths, cleared in xylene and coverslipped in Eukitt. They were examined with a Zeiss microscope equipped with a camera lucida for accurate mapping of the patterns ofinnervation. At each level, sections were taken for Nissl staining with cresyl violet. The boundaries of the spinal laminae were determined from the Nissl-stained sections according to the reports of Molander et al. ( 1984, 1989). Quantitative analysis

A densitometric analysis of the densities of the DAlike immunoreactivity (LI) and NA-LI innervations was performed using a computerized image analyser type SAMBA 2005® (TITN, Alcatel, France). Following the examination of all the transverse sections (see below), four different sites were selected as typical: lamina II (lateral part), laminae III-IV, lamina VII in its ventral half and lamina IX. They were analysed at four rostro-caudal levels (C5, T12, L4 and $2). Due to the results of the 6-OHDA experiments (see below), the estimation was done in

433

non-treated rats, allowing measurements relative to both innervations to be performed in the same animals. Because of their very strong and often patchy labelling, autonomic areas and especially the intermedio-lateral cell column, were not included in this estimation. Three rats were used and the analysis was done on three sections for each rostro-caudal level of each rat. The immunostained image was extracted from background, which was defined by measuring optical density in spinal grey matter fields devoid of any immunoreactivity (local background). The validity of the threshold was further assessed by visual examination of the sections. Analysis was performed on fields of 3000 pm ~ for lamina II and 300001am 2 for the other regions, lamina II being thinner than the other studied zones. Results were expressed as a percentage of the surface area of the extracted image with respect to the whole analysed surface. Statistical analysis was performed using analysis of variance followed by a Tukey test. Antisera

The DA antiserum used was monoclonal and raised in the mouse (Chagnaud et al., 1987). Polyclonal antibodies used for NA immunohistochemistry were obtained from rabbits (Geffard et al., 1986). RESULTS Comparison of the fixation conditions

For NA immunohistochemistry, there was no difference when substituting acetate buffer for phosphate buffer. With this substitution, the only visible difference concerned the DA immunolabelling of the ventral part of the spinal cord. Such a labelling was better observable after fixation at acidic pH. Accordingly, the observations described below were from animals perfused indifferently with one fixative or the other, except for the DA-like immunolabelling in the ventral horn and the ventral funiculus, which was preferentially described in rats perfused with glutaraldehyde in acetate buffer. 6-OHDA experiments Results concerning 6-OHDA experiments are mentioned first because information about the spinal NA-LI and DA-LI innervations is based on them. A detailed description of these innervations will be presented afterwards. While the spinal cord of untreated rats contained many NA-LI fibres and terminals (Fig. 1B), the cord of 6-OHDA-lesioned animals was almost completely devoid of such elements (Fig. 1D). The only exception was the presence of a few fibres near and in the thoracic intermedio-lateral cell column, albeit at a very low density when compared with the same innervation of untreated animals. In contrast,

434

P. Mouchet e t al.

CERVICAL • DA rR NA

0Z ...... 20.

muJ ,(0 olz

lO'

Ill-IV

VII

IX

Differences between NA-LI and DA-LI spinal elements

THORACIC •DA 20, ttl " mm 40

O~

lO

II

lll-IV

VII

IX

LUMBAR t~

•

20

DA

~NA

i! __Lj II

Ill-IV

VII

IX

SACRAL •

DA

mm

| II

Ill-IV

the distribution and density of DA-LI elements observed in lesioned animals were similar to those found in untreated normal rats (Fig. 1A,C). There was also no difference in this respect between untreated rats and 6-OHDA-lesioned animals pretreated with benzatropine. In this latter group, an almost complete disappearance of N A - L I immunoreactivity was observed in the spinal cord, as seen in animals treated with 6 - O H D A only. Consequently, observations concerning the N A - L I innervation of the spinal cord have been made with untreated rats only, while for the DA-LI innervation, sections obtained from animals of the three groups have been used.

VII

IX

Apart from their contrasting behaviour towards 6O H D A effects, N A - L I and D A - L I spinal elements differed markedly according to their topographic distribution and their morphology. The distribution of both innervations will be described at length below. Morphological differences were clearly apparent at sites where the two innervations could be compared (Fig. 2). While each antiserum labelled only fibres and terminals, D A - L I fibres tended to bear larger varicosities than did the N A - L I ones. More importantly, D A - L I axons often bore small, probably terminal, appendages which often possessed a terminal varicosity (Fig. 2A-C). N A - L I varicosities, however, were generally very thin and nearly always enshrined in the axonal path (Fig. 2B-D). N A - L I axons did not exhibit the bud-like structures characteristic of the D A - L I fibres. DA-LI innervation White matter

D A - L I fibres were present in the dorsolateral funiculus where they formed a bundle ofrostro-caudaUy oriented fibres from which axons could sometimes be seen running towards the dorsal grey matter. A few fibres were also seen in the ventral funiculus and were transversally oriented towards the tip of the ventral horn. Fig. 6. Densitometric estimation of the spinal cord DA-LI and NA-LI innervations at four different sites. The corresponding laminae are indicated by the numbers placed under the horizontal axis. Results (vertical axis) are the ratio (in per cent) of the surface of the extracted imageto the total surfaceanalysed. Vertical bars indicate the standard deviation of the mean. One-way analysis of variance followed by Tukey's multiple range test shows a significantdifference(P < 0.01) betweenthe NA and DA index for all sites except for laminae Ill-IV, for which P is only less than 0.05 at thoracic (T5), lumbar (1..4)and sacral ($2) levels and for which there is no significant difference at cervical level (C5). For each site and for each catecholamine there are no significant differences along the four rostrocaudal levels except between C5 and $2 for DA in laminae Ill-IV {P