Catechol-O-methyltransferase in rat sensory ganglia and spinal cord

~

Pergamon

NeuroscienceVol. 73, No. 1, pp. 267-276, 1996 Copyright © 1996IBRO. Publishedby ElsevierScienceLtd Printed in Great Britain S0306-4522(96)00016-4 0306-4522/96$15.00+ 0.00

CATECHOL-O-METHYLTRANSFERASE IN RAT SENSORY GANGLIA AND SPINAL CORD T. K A R H U N E N , * t I. ULMANEN~/ and P. P A N U L A tDepartment of Biology, Abo Akademi University, FIN-20520 Turku, Finland :~Orion Corporation, Orion-Farmos, Orion Research, Helsinki, Finland Abstract--The localization of catechol-O-methyltransferase immunoreactivity in rat dorsal root ganglia and in the spinal cord and its co-existence with substance P, calcitonin gene-related peptide and fluoride-resistant acid phosphatase in dorsal root ganglion cells was examined with immunohistochemical and histochemical double-staining methods. Analysis of dorsal of dorsal root ganglia at both cervical and lumbar levels revealed catechol-O-methyltransferase immunoreactivity in numerous dorsal root ganglion cells. Double-staining studies showed that catechol-O-methyltransferase and substance P immunoreactivities were located in different cells with a few exceptions, whereas both catechol-O-methyltransferase and calcitonin gene-related peptide immunoreactivities were detected in about 10% of all labeled cells positive for one of the two markers at both levels studied. The great majority of fluoride-resistant alkaline phosphatase-positive cells were also immunoreactive for catechol-O-methyltransferase. Again, no difference was found between cervical and lumbar levels. Catechol-O-methyltransferase immunoreactivity was also found in the neuropil of the dorsal horn of the spinal cord. The staining was most intense in the superficial laminae (I-III) and overlapped partly with substance P and calcitonin gene-related peptide immunoreactivity. Western blotting analysis revealed that soluble catechol-O-methyltransferase was the clearly dominating form of the enzyme in dorsal root ganglia. The distribution pattern of catechol-O-methyltransferase in dorsal horn and sensory neurons suggests that the enzyme may modulate sensory neurotransmission. Copyright © 1996 IBRO. Published by Elsevier Science Ltd.

Key words: pain, catecholamines, substance P, calcitonin gene-related peptide, acid phosphatase.

The enzyme catechol-O-methyltransferase (COMT) is present in a variety of mammalian tissues, including the CNS. 21,22 Monoamine oxidase (MAO), together with COMT, is responsible for metabolic inactivation of catecholamine neurotransmitters. It is generally considered that M A O acts both intra- and extran e u r o n a l l y f while C O M T is located mainly extraneuronally. 3 C O M T has a wide range of catechol substrates. In addition to the neurotransmitters dopamine, norepinephrine and epinephrine, ~7,47:s C O M T also O-methylates their deaminated metabolites, 54 catechol hormones L2 and many drugs having a catechol group. 9 Earlier immunohistochemical studies have located C O M T only to the cytoplasm of glial cells in the CNS. 22 Recent immunohistochemical studies have also shown C O M T in neuronal elements in the rat brain 24 and the h u m a n b r a i n y With an antiserum against rat recombinant, E. coli-expressed *To whom correspondence should be addressed. Abbreviations: CGRP, calcitonin gene-related peptide; COMT, catechol-O-methyltransferase; DRG, dorsal root ganglion; FRAP, fluoride-resistant acid phosphatase; LI, like immunoreactivity; MAO, monoamine oxidase; MB-COMT, membrane-bound catecholO-methyltransferase; PB, phosphate buffer; PBS, phosphate-buffered saline; S-COMT, soluble catecholO-methyltransferase; SP, substance P; TH, tyrosine hydroxylase. 267

C O M T protein and immunoelectron microscopy, C O M T was detected in postsynaptic spines and astrocytes in different parts of the rat brain. 23 We also found that a subpopulation of neurons in rat spinal sensory ganglia were immunoreactive for COMT. 24 This was the only site where C O M T immunoreactivity could be detected in neuronal perikarya. The purpose of the present study was to identify the C O M T form present in dorsal root ganglia (DRGs) and the COMT-immunoreactive neurons in the spinal ganglia, and to correlate the distribution of C O M T to that of other known markers, substance P (SP), calcitonin gene-related peptide (CGRP) and fluoride-resistant alkaline phosphatase (FRAP).

EXPERIMENTAL PROCEDURES Sample preparation for immunoblotting Rat DRGs were homogenized in 10 mM sodium phosphate buffer, pH 7, containing 0.2 mM phenylmethylsulfonyl fluoride and the homogenate was centrifuged at 2500 r.p.m, for 20 min. The proteins in the supernatant were separated on 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, 34 followed by electrophoretic transfer onto polyvinylidene difluoride membranes Immobilon P, Millipore))3 The filters were treated with guinea-pig antiCOMT antiserum (diluted 1:200) in phosphate-buffered saline (PBS)/0.1% Tween 20 and 5% (w/v) non-fat dry milk for 1.5 h at room temperature, 18 followed by peroxidase-

268

T. Karhunen et al.

conjugated sheep anti-guinea-pig immunoglobulin G (Boehringer Mannheim, Mannheim, Germany) for 1 h at room temperature ~° and detection by the enhanced chemoluminescence Western blotting system (Amersham, U.K.). The films were quantiated by densitometry (Vitroscan XL, LKB).

Sample preparation for immunohistochemistry A total of eight male, adult Wistar rats (200-300 g, from the Animal Center of the University of Helsinki) were deeply anesthetized with sodium pentobarbital (Mebunat, Orion, Finland) and perfused transcardially with 0.9% saline followed by 200ml cold 4% paraformaldehyde (Merck, Darmstadt, Germany) in 0.1% phosphate buffer (PB), pH 7.4. For immunoelectron microscopy, fixation was done with 4% paraformaldehyde containing 0.5% glutaraldehyde (Fluka, Buchs, Switzerland). The spinal cord, the DRGs from the cervical and lumbar levels were removed immediately and postfixed in the same fixative for 2-4 h at 4°C. For immunohistochemistry, the tissues were then washed with 20% sucrose in PB, pH 7.4, at 4°C for at least overnight before use. Cryostat sections (20-pm-thick) were cut on gelatinized slides and tissues were processed for immunohistochemistry or sequentially first for immunofluorescence and then for FRAP histochemistry. For immunoelectron microscopy, the fixation was followed by washing in PB overnight at 4°C. The DRGs were embedded in gelatin and Vibratome sections (50/~m) were cut and collected in cold PB.

Immunohistochemical techniques An antiserum against recombinant, E. coli-expressed soluble COMT (S-COMT) protein was used. The COMT antiserum was produced in guinea-pig in this laboratory and its production, characterization and specificity have been described elsewhere. 24 The antiserum against COMT recognizes both the S-COMT and the membrane-bound COMT (MB-COMT). To study the relationship of the DRG neurons containing COMT immunoreactivity to those containing SP or CGRP immunoreactivity, double-labeling immunofluorescence was performed. The antiserum used for SP immunohistochemistry has been determined previously to be specific for this peptide. It does not cross-react with bombesin, ranatensin and somatostatin in immunohistochemical procedures, and it displays no cross-reactivity with somatostatin, bombesin or ranatensin in radioimmunoassay. 43 A commercial, specific antiserum against CGRP (RPN 1842, Amersham, U.K.) was used. The COMT antiserum and the normal guinea-pig serum used as a control were diluted 1 : 200, antiserum against SP 1 : 500 and antiserum against CGRP 1 : 1000 in PBS, pH 7.4, containing 0.25% Triton X-100 and 1% normal goat and/or swine serum, depending on the species in which the second antiserum was produced. Control sections for COMT were also incubated with COMT antiserum preabsorbed with an excess of the recombinant rat COMT protein in optimally diluted antiserum. In double-staining procedures, a mixture of two primary antisera was used in the first incubation step followed by sequential incubations with one secondary antiserum at a time. For immunofluorescence the sections were incubated with appropriate fluorescein- or rhodaminelabeled anti-immunoglobulin Gs (swine anti-rabbit immunoglobulin G, Dakopatts, Copenhagen, Denmark, diluted 1:40 for anti-SP and anti-CGRP, and goat antiguinea-pig immunoglobulin G, Cappel, West Chester, PA, U.S.A., diluted 1:I00 for anti-COMT) in PBS-Triton X-I00 for 1 h at room temperature followed by a wash in PBS. The sections were then coverslipped with glycerol-PBS (1 : 1) or with 1-5% sucrose in 20 mM Tris-maleate buffer (pH 5.0) if the sections were prepared for enzyme histochemistry. Samples were photographed with a Leitz Orthomat camera on Kodak T-Max film (400 ASA). Immunoreactive cells

were counted from complete sections of DRGs from three different animals. The number of sections analysed was seven on the cervical level and 14 on the lumbar level. Only cells with a distinct nucleus were counted. Cell counts between the two levels of the spinal cord were compared using Student's t-test or the alternate (Welch) t-test if the standard deviations differed significantly (InStat, GraphPad Software, San Diego, CA, U.S.A.).

lmmunoelectron microscopy Vibratome sections were incubated with COMT antiserum diluted 1:500 in PBS containing 1% normal goat serum and 0.05% saponin (Sigma, St Louis, MO, U.S.A.), for three to four days at 4°C. All dilutions and washing steps were done in PBS-saponin. The sections were washed and incubated in biotinylated secondary antiserum, diluted 1 : 1000, for 4 h at room temperature or overnight at 4°C. The sections were washed again and incubated in avidin coupled to horseradish peroxidase (Vector Lab., CA, U.S.A.), diluted 1 : 1000, overnight at 4°C. They were then washed with Tris-HCl buffer, pH 7.6, and reacted with 50 mg 3,Y-diaminobenzidine (Sigma) in 100 ml and 0.003% hydrogen peroxide in Tris-HCl buffer, pH 7.6, for 5 min. The sections were postfixed with 1% buffered osmium tetroxide for 1 h at room temperature and dehydrated through a graded ethanol series. They were then immersed twice in propylene oxide (Fluka) for 30 min each time and pre-embedded in a mixture of propylene oxide and Epon (Ladd, Burlington, VT, U.S.A.) overnight. The sections were first fiat-embedded and then embedded in Epon. Ultrathin sections were cut and examined under a Jeol SX100 transmission electron microscope without poststaining and using low acceleration voltage (40 kV).

Fluoride-resistant acid phosphatase After processing sections for COMT immunofluorescence and photography, the coverslips were removed and the slides were washed for 1 h in Tris-maleate buffer, pH 5.0. The sections were then reacted for FRAP with slight modification according to a method described previously. 39 Following the wash the sections were incubated in 20 mM Tris-maleate buffer, pH 5.0, containing 10mM sodium glycerophosphate, 1.5 mM lead nitrate and 0.5 mM sodium fluoride for 14~18h at room temperature. Incubations without the substrate as a control were also performed. The sections were then washed for 15 min with three changes of Tris-maleate buffer, rinsed with water and immersed in 2% ammonium sulfide for 2 min, rinsed with several changes of water, coverslipped in glycerol-water (3:1) and photographed using bright-field illumination without the condenser forelens to allow visualization of the cell borders. RESULTS

Western blotting Western blotting analyses o n D R G s indicated b o t h the 24,000 molecular weight S - C O M T a n d the 28,000 molecular weight M B - C O M T forms to be present in rat D R G s , a l t h o u g h S - C O M T was the p r e d o m i n a t i n g form in this tissue (Fig. 1), representing, o n the basis of densitometric scanning, a b o u t 7 5 % o f the total COMT.

Cateehol-O-methyltransferase immunoreactivity in dorsal root ganglia and its relation to substance P M a n y D R G cells exhibited C O M T i m m u n o reactivity (Fig. 2A) in the cytoplasm. The diameter of C O M T - i m m u n o r e a c t i v e cells m e a s u r e d from cells with a distinct nucleus was 22.741.7/~m (mean = 3 6 . 4 p m ; n = 4 8 ) . They were r o u n d e d or

COMT in the rat dorsal root ganglia slightly polygonal in shape (Fig. 2A). COMT (Fig. 2A) and SP (Fig. 2B) were stored in different cells, although the two markers were occasionally found co-localized (about 1% of labeled cells, not shown). Incubation of the sections with normal guinea-pig serum yielded no reaction (Fig. 2C). Antiserum preabsorbed with excess of recombinant COMT protein did not stain any structures (data not shown).

Catechol-O-methyltransferase immunoreactivity in dorsal root ganglia and its relation to calcitonin generelated peptide The distribution of COMT and CGRP immunoreactivity in DRGs is shown in Fig. 2D and E, respectively. CGRP immunoreactivity was prominent in both small and large neurons, whereas COMT immunoreactivity was not seen in large cells. Cells were counted from complete sections of ganglia at cervical and lumbar levels to determine the proportion of double-labeled, COMT-only and CGRPonly cells (Fig. 3). A total of 734 COMT- or CGRP-immunoreactive cells at the cervical level and 1544 at the lumbar level was counted. Ten per cent and 9% of all immunoreactive cells were positive for both markers at the cervical and lumbar levels, respectively (Fig. 3). No significant difference between the two levels of the spinal cord was detected.

Catechol-O-methyltransferase immunoreactivity in dorsal root ganglia and its relation to fluoride-resistant acid phosphatase The localization of COMT in relation to F R A P in DRGs is shown in Fig. 2F and G. Control sections

4030-

• MB "S

Fig. 1. Immunoblot analysis of COMT in rat DRGs. Proteins were separated on 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred to polyvinylidene difluoride membranes and stained with guinea-pig COMT antiserum followed by peroxidaselabeled secondary antibodies and substrate as described in Experimental Procedures. Amersham's colored "Rainbow" protein molecular weight markers were used. The position of the molecular weight standard is indicated, The 24,000 molecular weight S-COMT form is present in much higher concentrations than the 28,000 molecular weight MB-COMT. NSC 73/I--J

269

were used to test the specificity of the enzyme reaction. One group of control sections was incubated directly for FRAP. The reaction was stronger than in double-stained sections but the pattern of F R A P staining in these sections was similar to that seen in sections treated first for COMT immunofluorescence (data not shown). Another set of control sections was first processed for COMT immunofluorescence followed by F R A P incubation without substrate. There was no F R A P reaction in these slides (data not shown). Four hundred and fifty-six cells positive for COMT or F R A P at the cervical level and 830 at the lumbar level were counted. An analysis of micrographs processed for COMT and F R A P revealed that about 60% of cells positive for either one of these markers were reactive for both markers. Of the total of 810 FRAP-reactive cells counted, 90% were also immunoreactive for COMT. No significant difference was found between cervical and lumbar levels (Fig. 3).

Spinal cord COMT immunoreactivity in the dorsal horn was apparent as fine, dot-like staining in the neuropil and it overlapped with that of SP and CGRP (Fig. 4A-D). Incubation of the spinal cord with normal guinea-pig serum did not show immunoreactivity (Fig. 4F), as was seen in a consecutive section incubated with COMT antiserum (Fig. 4E).

lmmunoelectron microscopy Immunoelectron microscopy revealed strong immunoreactivity throughout the cytoplasm and also in the nucleus of neurons in DRGs (Fig, 5A). Immunoreactivity was also prominent in the rough endoplasmic reticulum of these cells (Fig. 5A, B). Satellite cells surrounding the neurons were generally negative (Fig. 5A, D), but occasional patchy COMT immunoreactivity was seen in the cytoplasm of individual satellite cells surrounding the neurons (not shown). Immunoelectron microscopy also revealed nerve fibers immunoreactive for COMT in the ganglia (Fig. 5C). Immunoreactive neurons showed granular reaction product in the cytoplasm and a patchy staining along the plasma membrane (Fig. 5D). The neighboring satellite cells were frequently negative. COMT immunoreactivity was also detected in the cytoplasm and rough endoplasmic reticulum of non-neuronal cells (Figs 5E, 6B). Endothelial cells and processes of non-neuronal cells surrounding the capillaries in DRGs were immunoreactive for COMT (Fig. 6A, B). Figure 6C shows a negative endothelial cell in a section incubated with a control serum. DISCUSSION

Western blotting analysis The analysis of the western blots performed with the same antibody as for immunocytochemistry

270

T. Karhunen et al.

C

B

Q

Fig. 2. COMT immunoreactivity in DRG cells. Double staining of the ganglia with COMT antiserum (A) and SP antiserum (B). COMT immunoreactivity and SP immunoreactivity are located mainly in different cells. Arrows indicate the same cells containing either COMT or SP immunoreactivity. Incubation of the samples with normal guinea-pig serum (c) did not yield any reaction. Double staining of the samples with COMT antiserum (D) and CGRP antiserum (E) revealed partial co-localization. Cells immunoreactive for both COMT and CGRP are indicated with solid arrows and cells immunoreactive for either COMT or CGRP with open arrows. COMT immunoreactivity was seen in sections (F) which after immunostaining were processed for FRAP histochemistry (G). Most FRAP-positive cells were also immunoreactive for COMT (arrows). Scale bar = 50 pm.

COMT in the rat dorsal root ganglia

271

According to that study in the rat, up to 96% of CGRP-immunoreactive cells are also SP-immunoreactive at all different levels. In our study, COMT immunoreactivity in rat D R G cells was distributed almost entirely in cells distinct from those containing SP immunoreactivity, but up to about 26% of CGRP-immunoreactive neurons were also immunoreactive for COMT at both levels studied. Combining F R A P histochemistry with COMT immunohistochemistry revealed that the majority of Histochemistry of the spinal ganglia FRAP-positive cells were also COMT-immunoPrimary sensory neurons contain a variety of pep- reactive. The role of F R A P in D R G cells is not clear, tides, including SP 6'11'x2'14'35'57 a n d C G R P . 4'8'14'35'38'57 but its possible involvement in noeiceptive transCGRP-Iike immunoreactive (LI) peptide-containing mission is suggested29 based on its existence in small D R G cells and appropriate laminae of the dorsal neurons are possibly the most numerous among peptide-containing cells found in DRGs, constituting horn. In rat, biochemically different subpopulations about 60-70% of all L3 ganglion cells in rat} ° The of small type B cells have been defined, containing number of COMT-immunoreactive D R G neurons either SP, somatostatin or FRAP. x1'15'16'4°After colwas only slightly lower than that of CGRP-immuno- chicine pretreatment, however, co-existence of SP and reactive ones, which indicates that COMT is ex- F R A P in a small subpopulation of cells has been pressed in a significant population of D R G neurons. f o u n d : The largest diameter of the COMT-immunoD R G cells have been divided into subgroups by reactive cells was 30~0/~m and is thus within the various methods, including size and content of differ- size limits of type B cells (20-50/tin): 6 COMT iment chemical markers. Peptides are localized preferen- munoreactivity was not seen in large neurons, some tially to small neurons of the sensory ganglia.5'11'13'33'4° of which were immunoreactive for CGRP. Division of the cells into peptide- and non-peptideThe relatively recent finding that catecholaminergic containing groups, the latter also containing FRAP, traits are expressed by primary sensory neurons is has been suggested.~6 However, partial co-existence of based mainly on expression of tyrosine hydroxylase F R A P and CGRP has been reported and further (TH), the rate-limiting enzyme in catecholamine biodivision into three separate cell populations, those synthesis, in these neurons. 27,44,45A small population containing either CGRP and FRAP, CGRP and SP of cells displaying TH immunoreactivity has been or CGRP and somatostatin, has been proposed. 4 The found, constituting about 1% of neurons in rat L5 small cells are mainly labeled for CGRP, but the ganglia 26'44'45'55and 4--13 % of guinea-pig D R G cells at peptide is also found in large sensory neurons. 2° the lumbosacral level,n In rat, these cells appear to be

showed that the antiserum recognized the two known forms of COMT, the 28,000 molecular weight MBCOMT and the 24,000 molecular weight S-COMT, from rat DRGs. Densitometric quantification showed that the majority (about 75%) of the total COMT of DRGs represented the soluble form. This is in accordance with the western blot analyses of different brain areas, where 69-86% of the total COMT was found to represent the soluble fOrlTl. 3'47'52

A

B

Y

6O

6O

40

4O

20

2O

.Q "~ 40

c (J

0

I

2

EOMT/CGRP

3

1

2

3

COMT/FRAP

Fig. 3. Co-existence of COMT and CGRP (a) and COMT and FRAP (b) in DRG cells at the cervical (white bars) and lumbar (black bars) levels. Histograms show the percentages of labeled cells of all positive cells (error bars = S.E.M.; n = 7 complete sections at the cervical level and 15 complete sections at the lumbar level from three different animals). A: 1, double-labeled cells; 2, only COMT-immunoreactive cells; 3, only CGRP-immunoreactive cells. B: 1, double-labeled cells; 2, only COMT-immunoreactive cells; 3, only FRAP-positive cells. There was no significant difference between the cervical and lumbar levels. (If the standard deviations differed significantly, as was the case in columns A ! and B3, the Welch t-test was used, otherwise the Student's t-test was the choice.)

272

T. Karhunen et al.

distinct from D R G neurons expressing the peptidergic markers SP and somatostatin,204 as well as from FRAP-positive cells.44 In guinea-pig, TH-immunoreactive cells are distinct from DRG neurons expressing SP and CGRP, but display partly somatostatin immunoreactivity.32 Other enzymes involved in catecholamine synthesis, aromatic amino acid decarboxylase, dopamine-/~-hydroxylase (DBH) and phenylethanolamine-N-methyltransferase, have not been

dctcc~cd in

D R G ~ . ~'~

The lack of dopaminc-/~-

hydroxylase in TH-immunoreactive cells has led to the conclusion that these cells are probably dopaminergic. 27'4s Paraformaldehyde-induced fluorescence

studies to show catecholamines in rat DRG neurons have, however, given negative results in intact animals.45 In another study, after pretreatment of animals with pargyline, the MAO inhibitor, formaldehyde-induced catecholamine fluorescence was seen in numerous cells. 27 In more recent studies, however, no catecholamine-synthesizing enzymes other than TH nor DOPA or dopamine in D R G cells of guinea-pig sensory neurons have been found, and the dopaminergic nature of those TH-cot~taining ccll~ iu rat DRGs has been questioned. 32 Early studies did not reveal distinct MAO-positive or -negative cell populations in DRGs, but instead

Fig. 4. Distribution of COMT immunoreactivity in the spinal cord. Sections of the cervical spinal cord double-stained with COMT antiserum (a) and SP antiserum (b), and with COMT antiserum (c) and CGRP antiserum (d). Posterior horn of the spinal cord exhibits immunoreactivity for COMT and the two peptides. The most intense staining is in the superficial laminae of the dorsal horn. Staining of the consecutive sections with COMT antiserum (e) and normal guinea-pig serum (f) showed that there is no staining with the control serum. Scale bars = 100/~m.

Fig. 5. Electron microscopic immunohistochemistry of COMT in DRGs. (a) Electron mierograph of a DRG neuron showing strong COMT immunoreactivity in the cytoplasm. The nucleus (N) also expresses immunoreactivity for COMT. Reaction end-product is also seen in rough endoplasmic reticulum (solid arrow) and on outer mitochondrial membranes. Open arrows indicate the nuclear membrane. (b) Higher magnification of COMT immunoreactivity in rough endoplasmic reticulum. (c) Immunoelectron microscopy revealed reaction product in nerve fibers. (d) COMT immunoreactivity in neuronal cytoplasm and associated with the plasma membrane. (e) COMT immunoreactivity was also detected in nonneuronal cells. Reaction product is seen in the cytoplasm and in rough endoplasmic reticulum (solid arrow). Here the plasma membrane also shows immunoreactivity (open arrows). Scale bars = l ~ m (a, c, e); 500 nm (b, d).

274

T. Karhunen et al.

Fig. 6. (a) An electron micrograph showing a capillary exhibiting COMT immunoreactivity in endothelial cells. (b) A greater magnification of the wall of a capillary. COMT immunoreactivity is evident in an endothelial cell and in a cell process of a non-neuronal cell. (c) COMT immunoreactivity was no seen in endothelial cells in a section incubated with normal serum. Scale bars = 1 pm (a); 500 nm (b, c).

many studies reported faint staining in essentially all ergic terminals and neuronal somata or their satellite D R G cells (for review see Ref. 58). More recently, a cells were detected in guinea-pig D R G s ) ~ small, distinct cell population in rat DRGs containing Immunoelectron microscopy also revealed COMT MAO activity has been reported. 55 These cells were immunoreactivity in nerve fibers in DRGs. D R G cells very small in diameter, and predominated in lumbar containing F R A P can be demonstrated to have proDRGs as TH-LI-containing cells. Possible co-exist- jections in cutaneous and visceral nerves, 36'37 which ence of MAO activity and TH-LI was not studied in renders it possible that COMT synthesized in sensory detail. In comparison to these results, firstly, a dis- neurons may be active in, for example, the skin and tinct cell population containing clear COMT im- visceral organs. munoreactivity was found and, secondly, distribution of COMT immunoreactivity in both cervical and Spinal cord lumbar levels of DRGs was more extensive than either TH-LI or MAO. Noradrenaline and dopamine, arising from Evidence for the existence of catecholamines in supraspinal neurons, are both found in the dorsal D R G neurons is thus contradictory. Nerve terminals roots. 5° Noradrenergic terminations detected by containing catecholamine-synthesizing enzyme and immunocytochemical localization of dopaminecatecholamines have been observed by means of fl-hydroxylase56 or by aluminium-formaldehyde aldehyde-induced catecholamine fluorescence and im- fluorescence technique 49 are found throughout the munohistochemistry in DRGS of cat 42'51 and guinea- gray matter of the spinal cord, at all levels. The pig31'32 at the light microscopic level and also heaviest staining, however, is found in laminae I, II confirmed at the ultrastructural level in guinea-pig. 31 and V of the dorsal horn and in the ventral horn in They are considered to be confined to sympathetic regions containing large cells, possibly motoneunerves which enter and terminate in the ganglia. 32'51 rons. 56 Noradrenaline is found in higher concenOne possible role for COMT in D R G sensory cells trations than dopamine, 5° which is found in the might be to protect cell bodies from catecholamines superficial layers of the dorsal horn. 3°,49,5~A similar released from nearby fibers. Kayahara et al. 28 have distribution of COMT immunoreactivity in the dorshown electron microscopically synaptic terminals sal horn was detected in this study. The release of ending on cat cervical D R G cell bodies. In a recent catecholamines from the spinal terminals depresses study, however, no such contacts between noradren- spinal nociceptive transmission. 19 The pattern of

COMTin the rat dorsal root ganglia termination of FRAP-positive fibers is restricted to lamina II in the dorsal horn, 4° overlapping partly with C O M T immunoreactivity, suggesting that both may originate from primary afferents. The most intense C O M T staining was found in the upper layers of the dorsal horn, where C O M T may participate in inactivation of catecholamines in the descending pathways. Whether C O M T in the dorsal horn originates

275

from D R G cells remains to be studied with lesion techniques. The predominant localization of C O M T in the superficial laminae of the dorsal horn suggests a possible involvement in sensory neurotransmission. Acknowledgements--This study was supported by grants

from the Medical Research Council of The Academy of Finland and The Sigrid Juselius Foundation (to T. K. and P. P.).

REFERENCES

I. Ball P., Knuppen R.., Haupt M. and Breuer H. (1972) Kinetic properties of a soluble catechol-O-methyltransferase of human liver. Eur. J. Biochem. 26, 560-569. 2. Ball P., Knuppen R., Haupt M. and Breuer H. (1972) Interactions between oestrogens and catecholamines: influence of oestrogens on the effect of catecholamines and other catechols by the catechol-O-methyltransferase of human liver. J. clin. Endocr. Metab. 34, 736-746. 3. Broch O. Jr. and Fonnum F. (1972) The regional and subcellular distribution of catechol-O-methyltransferase in the rat brain. J. Neurochem. 19, 2049-2055. 4. Carr P. A., Yamamoto T. Y. and Nagy J. I. (1990) Calcitonin gene-related peptide in primary afferent neurons of rat: co-existence with fluoride-resistant acid phosphatase and depletion by neonatal capsaicin. Neuroscience 36, 751-760. 5. Chan-Palay V. and Palay S. L. (1977) Immunocytochemical identification of substance P cells and their processes in rat sensory ganglia and their terminals in the spinal cord: light microscopic studies. Proc. natn. Acad. Sci. U.S.A. 74, 359%3601. 6. Dalsgaard C.-J., Risling M. and Cuello C. (1982) Immunohistochemical localization of substance P in the lumbosacral spinal pia mater and ventral roots of the cat. Brain Res. 246, 168-171. 7. Dalsgaard C.-J., Ygge J., Vincent S. R., Ohrling M., Dockray G. J. and Elde R. (1984) Peripheral projections and neuropeptide coexistence in a subpopulation of fluoride-resistant acid phosphatase reactive spinal primary sensory neurons. Neurosci. Lett. 51, 139-144. 8. Green T. and Dockray G. J. (1987) Calcitonin gene-related peptide and substance P in afferents to the upper gastrointestinal tract in the rat. Neurosci. l~ett. 76, 151-156. 9. Guldberg H. C. and Marsden C. A. (1975) Catechol-O-methyltransferase: pharmacological aspects and physiological role. Pharmac. Rev. 27, 135-206. 10. Harlow E. and Lane D. (1988) Immunoblotting. In Antibodies. A Laboratory Manual pp. 471--510. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 11. H6kfelt T., Elde R., Johansson O., Luft R. and Arimura A. (1976) Immunohistochemical evidence for separate populations of somatostatin- and substance P-containing primary neurons in the rat. Neuroscience 1, 131-136. 12. H6kfelt T., Johansson O., Kellerth J.-O., Ljungdahl A., Nilsson G., Nyg~.rds A. and Pernow P. (1976) Immunohistochemical distribution of substance P. In Substance P (eds von Euler U. S. and Pernow B.), pp. 117-145. Raven Press, New York. 13. H6kfelt T., Johansson O., Ljungdahl A., Lundberg J. M. and Schultzberg J. M. (1980) Peptidergic neurones. Nature 284, 515-521. 14. H6kfelt T., Wiesenfeld-Hallin Z., Villar M. and Melander T. (1987) Increase of galanin-like immunoreactivity in rat dorsal root ganglion cells after peripheral axotomy. Neurosci. Lett. 83, 217-220. 15. H6kfelt T., Vincent S., Dalsgaard C. J., Skirboll L., Johansson O., Schulzberg M., Rossell S., Pernow B. and Jansco G. (1982) Distribution of substance P in brain and periphery and its possible role as a co-transmitter. In Substance P in the Nervous System (eds Porter R. and O'Connor M.), Vol. 91, pp. 84-106. Ciba Foundation Symposium, Pitman, London. 16. Hunt S. P. and Rossi J. (1985) Peptide- and non-peptide-containing unmyelinated primary afferents: the parallel processing of nociceptive information. Phil. Trans. R. Soc. Lond. B308, 283-289. 17. Jeffery D. R. and Roth J. A. (1984) Characterization of membrane-bound and soluble catechol-O-methyltransferase from human frontal cortex. J. Neurochem. 42, 826-832. 18. Johnson D. A., Gautsch J. W., Sportsman J. R. and Elder J. H. (1984) Improved technique utilizing nonfat dry milk for analysis of proteins and nucleic acids transferred to nitrocellulose. Gene Anal. Tech. 1, 3 8. 19. Jones S. L. and Gebhart G. F. (1986) Quantitative characterisation of coerulospinal inhibition of nococeptive transmission in the rat. J. Neurophysiol. 56, 1397-1410. 20. Ju G., H6kfelt T., Brodin E., Fahrenkrug J., Fischer J. A., Frey P., Elde R. P. and Brown J. C. (1987) Primary sensory neurons of the rat showing calcitonin gene-related peptide immunoreactivity and their relation to substance P-, somatostatin-, galanin-, vasocactive intestinal polypeptide- and cholecystokinin-immunoreactive ganglion cells. Cell Tiss. Res. 247, 417-431. 21. Kaplan G. P., Hartman B. K. and Creveling C. R. (1981) Immunohistochemical localization ofcatechol-O-methyltransferase in circumventricular organs of the rat: potential variations in the blood-brain barrier to native catechols. Brain Res. 229, 323-335. 22. Kapaln G. P., Hartman B. K. and Creveling C. R. (1979) Immunohistochemical demonstration of catechol-O-methyltransferase in mammalian brain. Brain Rcs. 167, 241-250. 23. Karhunen T., Tilgmann C., Ulmanen I. and Panula P. (1995) Catechol-O-methyltransferase (COMT) in rat brain: immunoelectron microscopic study with an antiserum against rat recombinant COMT protein. Neurosci. Lett. 187, 57~0. 24. Karhunen T., Tilgmann C., Ulmanen I., Julkunen I. and Panula P. (1994) Distribution of catechol-O-methyltransferase enzyme in rat tissues. J. Histochem. Cytochem. 42, 1079-1090.

276

T. Karhunen et al.

25. Katner A., Anglade P., Hirch E. C., Bounaix C., Damier P., Javoy-Agid F., Bromet N. and Agid Y. (1993). Catechol-O-methyltransferase in the human nigral complex and striatum: an immunohistochemical study. Soc. Neurosci. Abstr. 19, 927. 26. Katz D. M., Adler J. E. and Black 1. B. (1987) Catecholaminergic primary sensory neurons: autonomic targets and mechanisms of transmitter regulation. Fedn Proc. 46, 24-29. 27. Katz D. M., Markey K. A., Goldstein M. and Black I. B. (1983) Expression of catecholaminergic characteristics by primary sensory neurons in the normal adult rat in vivo. Proc. natn. Acad. Sci. U.S.A. 80, 3526-3530. 28. Kayahara T., Takimoto T. and Sakashita S. (1981) Synaptic junctions in the cat spinal ganglion. Brain Res. 216, 277-290. 29. Knyih~.r E. (1971) Fluoride resistant acid phosphatase system of nociceptive dorsal root afferents. Experientia 27, 1205 1207. 30. Kondo M., Fujiwara H. and Chikako T. (1985) Autoradiographic evidence for dopaminergic innervation in guinea pig spinal cord. Jap. J. Pharmac. 38, 442-444. 31. Kummer W. (1994) Sensory ganglia as a target of autonomic and sensory nerve fibers in the guinea pig. Neuroscience 59, 739-754. 32. Kummer W., Gibbins I. L., Stefan P. and Kapoor V. (1990) Catecholamines and catecholamine-synthesizing enzymes in guinea-pig sensory ganglia. Cell Tiss. Res. 261, 595-606. 33. Kummer W. and Heym Ch. (1986) Correlation of neuronal size and peptide immunoreactivity in the guinea-pig trigeminal ganglion. Cell Tiss. Res. 245, 657-665. 34. Laemmli U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage Ta. Nature Lond., 227, 680-685. 35. Lundberg J. M., Franco-Cereceda A., Hua X., H6kfelt T. and Fischer J. A. (1985) Co-existence of substance P and calcitonin gene-related peptide-like immunoreactivities in sensory nerves in relation to cardiovascular and bronchoconstrictor effects of capsaicin. Eur. J. Pharmac. 108, 315-319. 36. McMahan S. B. (1986) The localization of fluoride-resistant acid phosphatase (FRAP) in the pelvic nerves and sacral spinal cord of rat. Neurosci. Lett. 64, 305-310. 37. McMahon S. B., Sykova E., Wall P. D., Woof C.-J. and Gibson S. J. (1984) Neurogenic extravasation and substance P levels are low in muscle as compared to skin in rat hindlimb. Neurosci. Lett. 52, 235-240. 38. Merighi A., Polak J. M., Gibson S. J., Gulbekian S., Valentino K. L. and Peirone S. M. (1988) Ultrastructural studies on calcitonin gene-related peptide-, tachykinins- and somatostatin-immunoreactive neurones in rat dorsal root ganglia: evidence for the colocalization of different peptides in single secretory granules. Cell Tiss. Res. 254, 101-109. 39. Nagy J. I. and Daddona P. E. (1985) Anatomical and cytochemical relationships of adenosine deaminase-containing primary afferent neurons in the rat. Neuroscience 15, 799-813, 40. Nagy J. I. and Hunt S. P. (1982) Fluoride-resistant acid phosphatase-containing neurons in dorsal root ganglia are separate from those containing substance P or somatostatin. Neuroscience 7, 89-97. 41. O'Carrol A.-M., Tipton K. F., Sullivan J. P., Fowler C. J. and Ross S. B. (1987) Intra- and extrasynaptosomal deamination of dopamine and noradrenaline by the two forms of human brain monoamine oxidase. Implications for the neurotoxicity of N-methyl-4-phenyl-l,2,3,6-tetrahydropyridine in man. Biogenic Amines 4, 165-178. 42. Owman C. and Santini M. (1966) Adrenergic nerves in spinal ganglion of the cat. Acta physiol, scand. 68, 127-128. 43. Panula P., Hadjiconstantinou M., Yang H.-Y. T. and Costa E. (1983) Immunohistochemical localization of bombesin/gastrin-releasing peptide and substance P in primary sensory neurons. J. Neurosci. 3, 2021-2029. 44. Price J. (1985) An immunohistochemical and quantitive examination of dorsal root ganglion neuronal subpopulations. J. Neurosci. 5, 2051-2059. 45. Price J. and Mudge A. W. (1983) A subpopulation of rat dorsal root ganglion neurones is catecholaminergic. Nature 301, 241-243. 46. Rambourg A., Clermont Y. and Beaudet A. (1983) Ultrastructural features of six types of neurons in rat dorsal root ganglia. J. Neurocytol. 12, 47-66. 47. Rivett A. J. and Roth J. A. (1982) Kinetic studies on the O-methylation ofdopamine by human brain membrane-bound catechol-O-methyltransferase. Biochemistry 21, 1642-1740. 48. Rivett A. J., Francis A, and Roth J. A. (1983) Distinct cellular localization of membrane-bound and soluble forms of catechol-O-methyltransferase in brain. J. Neurochem. 40, 215-219. 49. Schroder H. D. and Skagerberg G. (1985) Catecholamine innervation of the caudal spinal cord in the rat. J. comp. NeuroL 242, 358-368. 50. Skagerberg G., Bjrrklund A., Lindvall O. and Schmidt R. H. (1982) Origin and termination of the diencephalo-spinal dopamine system in the rat. Brain Res. Bull. 9, 237-244. 51. Stevens R. T., Hodge C. J. and Apkarian V. (1983) Catecholamine varicosities in cat dorsal root ganglion and spinal ventral roots. Brain Res. 26, 151-154. 52. Tenhunen J. and Ulmanen I. (1993) Production of rat soluble and membrane-bound catechol-O-methyltransferase forms from bifunctional mRNAs. Biochem. J. 296, 595-600. 53. Towbin H., Staehelin T. and Gordon J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheet: procedure and some applications. Proc. natn. Acad. Sci. U.S.A. 76, 4350-4354. 54. Trendelenburg U., Cassis L., Grohmann M. and Langeloh A. (1987) The functional coupling of neuronal and extraneuronal transport and intracellular monoamine oxidase. J. neural Transm. 23, 91-101. 55. Vega J. A., Amenta F., Hernandez L. C. and Del Valle M. E. (1991) Presence of catecholamine-related enzymes in a subpopulation of primary sensory neurons in dorsal root ganglia of the rat, Cell molec. Biol. 37, 519-530. 56. Westlund K. N., Bowker R. M., Ziegler M. G. and Coulter J. D. (1983) Noradrenergic projections to the spinal cord of the rat. Brain Res. 263, 15-31. 57. Wiesenfeld-Hallin Z., Hfkfelt T., Lundberg J. M., Forssman W. G., Reinecke M., Tschopp F. A. and Fischer J. A. (1984) Immunoreactive calcitonin gene-related peptide and substance P coexist in sensory neurons to the spinal cord and interact in spinal behavioural responses of the rat. Neurosci. Lett. 52, 199-204. 58. Willis W. D. Jr. and CoggeshaU R. E. (1991) Sensory Mechanisms o f the Spinal Cord, p. 63. Plenum Press, New York. (Accepted 3 January 1996)