c-fos Expression in rat lumbar spinal cord during the development of adjuvant-induced arthritis

0306-4522/92$5.00+ 0.00 Pergamon Press Ltd 0 1992IBRO

NeuroscienceVol. 48, No. 4, pp. 985-993, 1992 Printed in Great Britain

c-fos EXPRESSION IN RAT LUMBAR SPINAL CORD DURING THE DEVELOPMENT OF ADJUVANT-INDUCED ARTHRITIS C. ABBADIE*

and J.-M. BESSON

Uniti de Recherches de Physiopharmacologie du Systime Nerveux, iNSERM U.161 and EPHE, 2 rue d’Al&a, 75014 Paris, France Abstract-A parallel clinical and behavioral study of adjuvant-induced arthritis in the rat showed four stages in the time-course of the disease: preclinical (first week), acute (weeks 2-4), post-acute (weeks 5-8) and recovery weeks 9-11) [Calvino et nl. (1987) Behau. Brain Res. 24, 1l-291. As several studies have reported the expression of the proto-oncogene c-j& in spinal cord neurons following acute noxious peripheral stimuli, the aim of this study was to quantitatively assess Fos-like immunoreactivity in lumbar spinal cord neurons at various times of adjuvant-induced arthritis development, i.e. one, two, three, 11 and 22 weeks post-inoculation. The total number of Fos-like immunoreactive neurons in the lumbar enlargement correlated with the observed development of adjuvant-induced arthritis, i.e. Fos-like immunoreactivity was absent at one week, moderate at two weeks, greatly increased at three weeks, decreased at 11 weeks and returned to control values at 22 weeks. At three weeks, at the peak of Fos-like immunoreactivity distribution and acute stage of hyperalgesia, maximal labeling was observed in L3 and L4 spinal segments. In these segments, the most densely labeled region was the neck (laminae V and VI) of the dorsal horn (55%) and the ventral horn (35%) as compared to the superficial laminae (laminae I and II; 5%) and the nucleus proprius (laminae III and IV; 5%). These data indicate that c-fos expression induced by chronic inflammation is better expressed in deeper laminae than in the superficial ones, and that the number of Fos-positive cells correlates with behavioral studies. Thus, the use of Fos-like immunoreactivity in the chronic inflammatory pain model seems to be an interesting tool to study possible effects of various pharmacological compounds such as analgesic or anti-inflammatory drugs.

An experimental model to study “chronic” pain is adjuvant-induced arthritis (AIA) in the rat (see references in Ref. 5), which resembles human rheumatoid polyarthritis. It is induced by subcutaneous injection of Freund’s adjuvant (killed Mycobacterium butyricum suspended in mineral oil) into the base

of the rat tai1.35,36This disease. predominantly affects distal hindlimb joints (ankles). Among the observed changes, the most marked are decreases in locomotion and increases in scratching behavior.*s14 An initial inflammatory response develops within hours, but dramatic clinical signs appear from the 10th day post-inoculation and the changes last over several weeks.10,i1,14The model of AIA which induces numerous biochemical changes at both the periphery and the central nervous system has been used for this study of potential analgesic and antiinflammatory drugs (see references in Ref. 5). In addition, from the electrophysiological point of view, the disease has been shown to induce dramatic modifications in the activity of both superficial (I and II) and deeper (V and VI) laminae of the dorsal horn

*To whom correspondence should be addressed. AIA, adjuvant-induced arthritis; -LI, like immunoreactive/immunoreactivity; NRST normal rabbit serum in phosphate-buffered saline with T&on-X; PB, phosphate buffer; PBS, phosphate-buffered saline.

Abbrrviufions:

neurons receiving noxious inputs. These modifications involve (i) the appearance of high spontaneous firing rate, frequently associated with bursting patterns and sometimes sudden dramatic increases in the absence of any intentional stimulation, and (ii) an increase of neuronal responsivity to mechanical stimulation.g26 These data are reminiscent of clinical observations in human rheumatoid disease, where flashes of pain can appear spontaneously, or can be caused by simple touch at the inflamed area. The aim of this study was to quantitatively evaluate the Fos-like immunoreactivity (-LI) in lumbar spinal cord neurons during AIA development. The expression of the proto-oncogene c-fos in spinal cord neurons following various acute noxious peripheral stimuli has been demonstrated in numerous studies.6~‘s~22~27*38~43~44 More interestingly, during acute inflammation, induced, either by plantar injection of complete Freund’s adjuvant*’ or by carageenan,34 an increase in Fos-LI was reported in dorsal horn neurons of the spinal cord; some of these neurons co-localized with dynorphin,34 or are at the origin of ascending pathways. *’ These two latter experimental conditions strongly differ from rheumatoid arthritis induced by injection of Freund’s adjuvant into the tail, which develops over several weeks and which mainly involves distal hindlimb joints.

985

986

C.

ABBADIEand

A detailed parallel clinical and behavioral study of AIA in the rat showed four stages in the time-course of the disease: preclinical (first week), acute (weeks 2-4), post-acute (weeks 5-8) and recovery (weeks 9-l l).’ Thus, our study took into consideration these four stages. For this purpose, at various times of AIA, we made an analysis of the segmental distribution of Fos-LI neurons in five spinal segments (L2-L6), and in the laminar distribution over four regions. Preliminary reports have partly described these data.‘,2 EXPERIMENTAL PROCEDURES

Experimental animals

Experiments were performed on 42 male Sprague-Dawley rats (Charles River, France), weighing 200-225g at the beginning of the experiment. Polyarthritis was induced by intradennal injection of Fnund’s adjuvant into the base of the tail (killed Mycobacteriwn butyricum suspended in mineral oil);% the inoculation was made at the breeding center (Charles River, France). Control animals received vehicle alone. Guidelines on ethical standards for investigations of experimental pain in animals were followed.13 The number of experimental arthritic animals was kept to a minimum. They were housed three to a large cage, the floor of which was covered with sawdust to minimize the possibility of interactions between rats placed in close contact. They were kept in an animal room at a constant temperature of 22”C, with a 12-h alternating light-dark cycle. Food and water were available ad libitum; the food was directly available on the sawdust in the cages to minimize the need for the animals to make potentially painful movements to obtain food. Rats were k&d at various post-inoculation periods: one, two, three, 11 and 22 weeks after Freund’s adjuvant injection. Every experimental group included seven animals. Immunohistochemistry

Animals were deeply anesthetized with pentobarbital (55 mg/kg, i.p.) and perfused intracardially with 200 ml of 0.1 M phosphate-buffered saline (PBS) followed by 500 ml of 4% paraformaldehyde in 0.1 M phosphate buEer (PB). The lumbar spinal cord was then removed and postfued for 4 h in the same fixative and cryoprotected overnight in 30% sucrose in PB. Frontal frozen sections of 40 pm were cut and collected in PB. The serial sections were immunostained for c-&-like protein according to the avidin-biotin-peroxidase method of Hsu et aLz’ The tissue sections were incubated for 30 min at room temperature in a blocking solution of 3% normal rabbit serum in PBS with 0.3% T&on-X NNRST). The sections were then incubated overnight at 4°C in the phmary ant&rum directed against the c-10s protein (OA 11-823, Cambridge Research Biochemicals). The Fos antibody was used at 1: 2000. The incubated sections were washed three times in 1% NRST and incubated in biotinylated rabbit anti-sheep IgG for 1 h at room temperature, then washed twice in 1% NRST and incubated for 1 h in avidin-biotin-peroxidase complex (Vectastain, Vector Laboratories). FinalIy, according to Mauro et al.,25modified by Menktrey et al.,m the sections were washed three times in PBS and developed in I-naphtol ammonium carbonate solution (89.5 ml 0.1 M PB, 10 ml 10% I-naphtol in absolute alcohol and 0.1 ml hydrogen peroxide) for 15min, and were washed three times in PB to stop the staining reaction. The sections were mounted on gelatin-subbed slides and air-dried for the stain to be intensified and made alcoholresistant through basic dye enhancement in 0.025% crystal

J.-M. BESON violet solution in PB for 3 min. After two short PB I inses to take off the excess of stain, sections were differentiated in 70% alcohol and the differentiation time was evaluated under the microscope. After being air-dried, the slides were coverslipped. To test the specificity of the primary antibody. controls were performed; preabsorption with the corresponding synthetic peptide (OP-1 I-3210, CRB) or omission of any stage in the protocol abolished the staining Counting of Fos -labeled cells Fos-LX was studied through five lumbar spinal levels, L2-L6. Tissue sections were first examined using dark-field microscopy to determine the segmental level according to Molander et al.,3z and the gray matter landmarks. The sections were then examined under light-field microscopy to localize Fos-positive cells. Labeled nuclei were counted using a camera lucida attachment. For each rat, three measures were made: (i) the total number of Fos-LI neurons for the five lumbar spinal segments, (ii) the number of Fos-LI neurons per lumbar level and (iii) the number of Fos-LI neurons per region in the spinal gray matter. According to Presley et al., 38 four regions were defined: superficial dorsal horn (laminae I and II; superficial), nucleus proprius (laminae III and IV; nucleus proprius), neck of ihe horsal.hom (laminae V and VI; neck) and the ventral erav flaminae VII. VIII. IX and X: ventral). All results a;e kxpressed as the sum’of the number of F&-L1 neurons in three sections per one segmental level. Statistical analysis was made to compare the numbers of labeled cells, using one-way analysis of variance for the total number in the whole lumbar enlargement in the different groups of animals, two-way analysis of variance for the different groups of animals and the spinal level, and three-way analysis of variance for the different groups of animals, the spinal level, and the region. For multiple comparisons, the PLSD Fisher’s test was used. The investigator responsible for plotting and counting the Fos-LI neurons was unaware of the treatment of each animal. RESULTS

General observations

With our procedures, Fos-LI nuclei appeared as dark-violet round structures, often containing nonstained nucleoli. No labeling or only a few labeled cells (
of adbvant -induced

Considering the mean total number of Fos-LI cells in the whole spinal segment from L2 to L6, there was a signifkant difference [F(5,36) = 16.2, P < O.OOOl]

c-fos expression in rat lumbar spinal cord

between the six groups of animals (Fig. 3). Fos-LI was nearly absent in control and at one week, moderate at two weeks, maximal at three weeks, and slightly decreased at 11 weeks, but significantly

981

decreased at 22 weeks. Thus, the total number of Fos-LI cells for one-week and 22-week animals did not differ from the control, but significantly differed for two weeks (P < O.Ol), three weeks and 11 weeks (P < 0.001). Considering the animals at three weeks as reference, which corresponds to the peak of this distribution, all groups except 11 weeks presented a significant lower number of Fos-positive cells (P < 0.01 for two weeks, and P < 0.001 for 22 weeks, one week and for the control). Segmental distribution The number of Fos-LI neurons differed according to the lumbar spinal level considered. Two-way analysis of variance showed a significant main effect between the six groups [F(5,180) = 51.8, P < O.OOOl], between the five spinal segments [F(4,180) = 10.5, P < O.OOOl],and a significant group x segment interaction [F(20,180) = 3.1, P < O.OOOl].This indicates that the development of AIA varied among the different segments. A significant increase of Fos-LI cells was observed at two, three and 11 weeks vs control but no change was observed at one and 22 weeks vs control (for significance, see Fig. 4) when the segmental distribution (L2-L6) was compared. On analysing the rostrocaudal distribution, we observed no significant difference between the five lumbar segments for control, one, two and 22 weeks, but a significant difference was noticed for three weeks [F(4,34) = 15.2, P < O.OOOl] and 11 weeks [F(4,34) = 4.6, P < O.OOS].At three weeks maximal labeling was present in L3-I.4 [60.5% (number of Fos-LI cells in L3-IA/total number of Fos-LI cells in the L2-L6 enlargement); 31 .l% in L3 and 29.4% in L4]. Thus, the number of Fos-positive cells was significantly higher in L3 and in L4 vs L2 (P < O.Ol), and vs L5 and L6 (P < 0.001). Laminar distribution

As already described in Experimental Procedures, we determined four regions in the spinal gray matter: superficial, nucleus proprius, neck and ventral. Threeway analysis of variance showed a highly significant effect between the six groups [F(5,720) = 99.6, P < O.OOOl], between the five spinal segments [F(4,720) = 21.3, P
Fig. 1. Photomicrographs illustrating the location of Fos-LI neurons in L3 segment at different times following Freund’s adjuvant injection. Two post-injection (p.i.) times are represented: two weeks p.i. (2 W; A) and three weeks (B). Note that the maximal number of Fos-LI nuclei is observed at three weeks which is the acute stage of hyperalgesia, and that the. region with the most intense staining corresponds to the neck of the dorsal horn &minae V and VI); (b) higher magnification photomicrograph of the section at three weeks illustrating Fos-LI in the neck of the dorsal horn.

988

C. ABBADIE and

L3

.:

i’ .. *

/

..a .. *

.. ::

~ .* . “$ 3 3 3 5 .*.* . ,/..-. 3

’

L6

c_~

1 ‘:...;.

lY__--_I

L5

BESSON

..* -(_‘j ..*-. ..: . .:.* <.

.. * . j”--”

L4 i

J.-M.

1’..

:

’

’ :.

.

1. *- -.

:

I .. .

” . ?...

1.. . :. ..

C’

k-L/-

1*.’. I

.

Fig. 2. Camera lucida drawings showing the rostrocaudal distribution of Fos-LI neurons in the lumbar spinal enlargement at different times post-Freund’s adjuvant injection. Five segments, from L2 to L6, and five post-injection times: one week (1 W) two, three, 11 and 22 weeks, are represented. Each scheme includes all labeled cells in 3 x 4Oqm sections; each dot represents one labeled cell. Note the increase in the number of Fos-LI cells at three weeks vs one and two weeks, and the decrease at 11 and 22 weeks vs 3 weeks. The most intense labehng is observed in L3 and L4 segments, and particularly in the neck of the dorsal horn (laminae V and VI). Control sections showed no Fos-LI and are not represented The boundaries of the superficial laminae and of the reticular part of the neck of the dorsal horn are outlined for orientation.

[F(60,720) = 2.5, P < O.OOOl]. This indicates that Fos-LI varied among the regions in the different spinal levels; but as maximal labeling was.observed in L3 and L4 (see above results), we considered, in the below analysis, only the regional distribution in these two segments. For all the groups of AIA animals, Fos-LI was maximal in the neck, rather high in the ventral horn,

and lower in the superficial laminae and the nucleus proprius. In this laminar analysis, similar results were also obtained considering the development of AIA, in the four determined regions, i.e. signitIcant increase at two, three and 11 weeks vs control and no change at one and 22 weeks vs control (for significance, see Fig. 5).

989

c-fos expression in rat lumbar spinal cord 300

oonaol 1w 2w

200

3w 11 w 22W

100

$

2

40

;

20

z E z’

0

Fig. 3. Histogram presenting the total number of Fos-LI neurons in the lumbar spinal enlargement, from L2 to L6, at different delays post-Freund’s adjuvant injection. Six groups of animals are considered: control animals and animals at five post-injection times: one week (1 w), two, three, 11 and 22 weeks (n = 7 in each group). Results are expressed as mean (+S.E.M.) number of Fos-LI neurons for five lumbar segments per 3 x 4Oqm sections. Note that the number of Fos-LI neurons increases at two weeks, is maximal at three weeks, slightly decreases at 11 weeks but importantly at 22 weeks. Significance is expressed using PLSD Fisher’s test, taking the control group as reference (**P co.01, ***p
At three weeks, the regional distribution in L3 and L4 was 4.5% in the superficial laminae, 6.2% in the nucleus proprius, 56.8% in the neck and 32.5% in the ventral horn. Thus, in this case, the number of Fos-LI neurons is significantly (P < 0.001) higher in the neck than in the three other regions (see Fig. 6).

DISCUSSION

This study demonstrates c-for expression in lumbar spinal cord neurons during the development of AIA. Previous data briefly related this expression following administration of Freund’s adjuvant, or

0

Superficial

Nucl. Proprius

Neck

Ventral

Fig. 5. Histogram presenting the number of Fos-LI neurons, in L3 and L4 segments, at different times post-Freund’s adjuvant injection, in different regions of the spinal gray matter. Six groups of animals are considered: control animals and animals at five post-injection times: one week (1 W), two, three, 11 and 22 weeks (n = 7 in each group). Four regions are defined: superficial dorsal horn (laminae I and II; superficial), nucleus proprius (laminae III and IV; nucleus proprius), neck of the dorsal horn (laminae V and VI; neck) and the ventral horn (laminae VII, VIII, IX and X; ventral). Results arc expressed as mean (+S.E.M.) number of Fos-LI neurons per 3 x 40-pm sections. Note that for all the groups of AIA animals, Fos-LI was maximal in the neck, rather high in the ventral horn and low in the superficial and in the nucleus proprius region. Significance is expressed using PLSD Fisher’s test, taking the control group as reference (*P< 0.05, **P < 0.01, ***P< 0.001).

carageenan,3*Mv42but they were not concerned with the time-course of AIA development nor with rostrocaudal and laminar distributions. In addition, our systematic investigation allows us to compare our data with a detailed clinical and behavioral study previously published by another group in our laboratory.’ In the lumbar spinal enlargement (L2-L6), we found that c-fos is expressed “spontaneously”, i.e. without any intentional stimulation in polyarthritic animals. Thus, the number of Fos-LI cells increases with the time-course of the development of AIA until three weeks, and then decreases and returns to control values at 22 weeks. These data are reminiscent

2w 3w 11 w 22W

~2

~3

~4

L5

L6

Fig. 4. Histogram presenting the rostrocaudal distribution in the lumbar spinal enlargement (from L2 to L6), of the number of Fos-LI neurons, at different delays post-Freund’s adjuvant injection. Six groups of animals are considered: control animals and animals at five post-injection times: one week (1 W), two, three, 11 and 22 weeks (n = 7 in each group). Results arc expressed as mean (* S.E.M.) number of Fos-LI neurons per 3 x 40-pm sections. Note that the number of Fos-LI neurons is significantly higher at three weeks all along the lumbar enlargement, and that maximal labeling is observed in L3 and L4. Significance is expressed using PLSD Fisher’s test, taking the control group as reference (+PcO.05, **P< 0.01, *+*P< 0.001).

Fig. 6. Diagram illustrating the laminar distribution in L3 and L4 segments, of the number of Fos-LI neurons, at three weeks post-Freund’s adjuvant injection. Four regions are considered: superticial dorsal horn (laminae I and II; superficial), nucleus proprius (laminae III and IV; nucleus pr~prius),neck of the dorsal horn (laminae V and VI; neck) and the ventral arav (laminae VII. VIII. IX and X: ventral). Values are expr&&d as the percentage of the number of Fos-LI cells in each region/total number of Fos-LI cells in the whole gray matter. Note that Fos-LI is maximal in the neck, and rather high in the ventral horn.

990

C. ABBADIE and J.-M. Brsso~

of clinical, behavioral, electrophysiological and biochemical studies. From the behavioral point of view, the absence of significant labeling one week after the induction of the disease is in good agreement with the fact that during the preclinical stage the clinical signs and the behavioral modifications do not appear. In contrast, the increase in the number of Fos-LI neurons at two weeks and particularly at three weeks corresponds to the acute stage previously defined, i.e. lack of mobility and exploring behavior, dramatic increase in hindpaw and forepaw joint diameters, radiological abnormalities and hyperalgesia. Interestingly, the maximal labeling observed at three weeks corresponds to the peak of the above-mentioned behavioral modifications and to a marked increase in the volume of ventilation.” It must be remembered that the hyperventilatory response to adjuvant arthritis follows the same time-course as weight, paw diameters and the increase in fentanyl intake.‘? As will be discussed later, these observations indirectly suggest that Fos-LI could be used as a marker for chronic pain, at least for AIA. However, we still observed a sustained labeling 11 weeks after the induction of the disease which corresponds to the recovery stage described by Calvin0 et al.,* when behavioral scores return to control values. Nevertheless, it must be pointed out that at this time, postinoculation, mean joint diameters and radiological scores are still increased. In our experimental procedure, the suppression of Fos-LI was observed by our final chosen time-point, i.e. 22 weeks. Two hypotheses could explain the dissociation observed in the post-acute phase. AIA induces a sustained activation (several weeks) of this immediate-early gene as compared to the results obtained following acute stimulation, which is generally considered to be maximal a few hours after stimulation and which progressively diminishes. 22,38Fos-LI detected in AIA animals is reminiscent of electrophysiological data obtained both in the superficial layers and in the neck of the dorsal horn of polyarthritic rats.2h These neurons normally display a weak activity in healthy unanesthetized spinalized animals while four weeks after Freund’s adjuvant injection they display a high level of spontaneous activity, with bursts and dramatic increases in the absence of any intentional stimulation. Moreover, the neurons which receive noxious inputs exhibit an enlargement of their receptive field size, and very often respond maximally to light mechanical stimuli. The origin of these modifications could be due either to peripheral mechanisms, i.e. activation and/or sensitization of thin peripheral fibers induced by various chemical agents released at the periphery (for general references see Refs 16 and 20) during arthritis and/or to central components related to central nervous system plasticity described during inflammatory processes.” Whatever the origin of these modifications, the data obtained in the present study, and those reported by electrophysiological investigations, relate to the well-known clini-

cal observations in human rheumatoid discasc. I c. the flashes of pain which can spontaneously appear or pain reactions evoked by weak stimulation of the inflamed joint. Fos-LI was absent in mid-thoracic segment, but sparse Fos-LI was present in sacral, lower thoracic and cervical segments, and predominated in the lumbar segment. The greater number of stained cells were located in L3 and in L4 segments and so the segmental pattern of labeling follows the somatotopic organization of the afferent fibers that innervate the hindpaws.24,3’,32.40Thus, the maximal labeling observed in the lumbar enlargement corresponds to the clinical signs of arthritis, i.e. hindpaws are predominantly affected. In AIA animals. Fos-LI is mainly present in the neck (laminae V and VI) of the lumbar dorsal horn (z 55% of the total number of Fos-LI in L3 and L4 at three weeks), and to a lesser extent in the ventral horn (5!3.5%). Surprisingly, the number of labeled cells was low in the superficial laminae (z 5”/0). Thus. the labeling obtained three weeks after Freund’:, adjuvant injection into the base of the tail strongly contrasts with previous investigations performed during acute inflammation. Indeed, in these latter experiments, dense Fos-LI was observed in laminae I and II after complete Freund’s adjuvant injection?’ or carageenan injection.15.” This labeling was observed a few hours (l-24 h) after the injection, while in our experiment, maximal labeling appears at the “aculc” stage of AIA disease (three to four wseeks). The labeling observed after a short delay following carageenan or acute Freund’s adjuvant resembles the one described after various acute noxious stimulations 6.22.27.38.43.44 Our data indicates that c+.s induced by chronic inflammation is better expressed in deeper laminae than in the superficial laminae. and the question arises as to whether the Fos-LI we described could be related to the time-course ot the disease. From a functional point of view. the interpretation of Fos labeling should be made with caution. In fact, the expression of c‘-fos does not provide information on the function of the cell since both inocuous and noxious stimulation are able to induce Fos expression at the level of the dorsal horn2’ Interestingly, in the present study, it must be emphasized that only a few cells expressed Fos in laminae III and IV which contain numerous neuron> only activated by non-noxious stimuli. Thus, taking into account the various investigations performed with this technique, it is reasonable to assume from many electrophysiological studies that Fos-LI neurons located in the superficial laminae and in the neck of the dorsal horn seem to preferentially express c-@s. In support of this hypothesis, it must be pointed out that Fos labeling induced in dorsal horn neurons following formalin injection seems to correlate with behavioral manifestations.‘9,3R this labeling being dose-dependently depressed by systemic morphine administration38 and by intracerebroventriculai

991

c-fos expression in rat lumbar spinal cord DAMG0.19 The labeling observed in this study probably reflects the signs of the spontaneous chronic pain, as described by behavioral reactions.*.“~‘4~37 According to numerous electrophysiological studies (see references in Ref. 4) we can speculate that the neurons we labeled were probably wide dynamic range neurons. Our data tend to suggest that neurons located in deeper laminae seem to be more sensitive in detecting changes occurring during chronic pain conditions. In this respect, a differential Fos expression in superficial laminae vs laminae III-VI has already been described under other circumstances. Indeed, morphine suppresses Fos-LI more so in laminae I and II than in deeper laminae.38~41Another finding is that laminar Fos-LI differs according to the delay following the stimulation: at short delay (1 h), c-f& is better expressed in laminae I and II, and then decreases; whereas later (2-4 h), staining remains constant in deeper laminae.22~38 The small number of labeled cells in the superficial laminae contrasts with our initial electrophysiological observations where we reported marked modifications of the electrophysiological properties of the neurons recorded at this level.26 In fact, such a dissociation could be due to experimental conditions since the present study was made in intact awake animals while our electrophysiological investigation was performed in a spinal preparation. Thus, it will be essential to try to evaluate the effect of descending control system on Fos-LI using various approaches. It must also be pointed out that the superficial laminae of the dorsal horn contain a certain proportion of nociceptive-specific neurons in which encoding properties in the medullary dorsal horn of awake monkey have been shown to be less pronounced than for wide dynamic range neurons.18 Thus, the level of Fos in superficial laminae cells in the absence of any intentional stimulation is possibly not intense enough to be detected by immunocytochemistry. By contrast, we observed a high proportion of Fos-LI cells in laminae I and II after moderate pressure to the ankle of arthritic animals.* Although speculative, another explanation could be that the Fos level in superficial lamina neurons is autoregulated to be able to respond to acute stimulation, whereas neurons in deeper laminae of the dorsal horn could respond to chronic inflammation. Among the various neurochemical changes occurring during the development of AIA, dynorphin levels are dramatically altered.23s29*30 On the other

hand, thermal noxious stimulation of the hindpaw or peripheral inflammation induced by carageenan produced an increase in prodynorphin gene expression which is preceded by an early induction of c-fos in lumbar spinal cord neurons,15.33 and carageenaninduced inflammation provokes dynorphin mRNA or peptide and Fos co-localization.34 Moreover, Sonnenberg et al. 39 showed that the Fos and Jun (product of the proto-oncogene c-&n) heterodimeric complex binds to a regulatory sequence of the proenkephalin gene and stimulates its transcription. Thus, the co-localization of proenkephalin and prodynorphin with Fos in dorsal horn neurons in polyarthritis (three weeks) and monoarthritis (one week) described by Weihe et aL4? could be partly due to the up-regulation of an opioid gene by c-fos, since levels of preprodynorphin mRNA and dynorphin are largely increased in the spinal dorsal horn during inflammation and hyperalgesia. This hypothesis could also be advanced for other regulation mechanisms since neurotransmitter (see references in Ref. 5) and neuroendocrine’ systems are widely disturbed in adjuvant arthritis. The increase in the number of Fos-positive cells observed during the acute phase of the disease could induce long-lasting physiopathological neural mechanisms (e.g. peptide synthesis or degradation). In addition to the above-mentioned data, it must also be emphasized that some of the Fos-LI cells encountered in the present study could project to supraspinal levels. In fact, a certain number of Fos-LI cells following subcutaneous inflammation of the plantar hindpaw could be retrogradely labeled after administration of protein-gold complex in various supraspinal regions.*’

CONCLUSION

These data indicate that c-fos expression induced by chronic inflammation is better-expressed in deeper than in the superficial laminae, and that Fos-LI correlates with behavioral studies. Thus, the use of Fos-LI in the chronic inflammatory pain model seems to be an interesting tool to study possible effects of various pharmacological compounds such as

analgesic or anti-inflammatory

drugs.

Acknowledgements-The authors gratefully acknowledge D. A. Dickenson for English revision of the manuscript and E. Dehausse for drawings and photography.

REFERENCE’3

1. Abbadie C., Morain F. and Besson J.-M. (1991) Expression of c-fos in rat lumbar spinal cord during the development of adjuvant-induced arthritis. Eur. J. Neurosci., Suppl. 4 (abstr.). 2. Abbadie C., Morain F. and Besson J.-M. (1991) Spontaneous and evoked expression of c-fos in rat lumbar spinal cord during the development of adjuvant-induced arthritis. Sot. Neurosci. Abstr. 17, 178.4. 3. Basbaum A. I., Menetrey D., Presley R. W. and Levine J. D. (1988) The contribution of the nervous system to experimental arthritis in the rat. In The Arthritic Rat as a Model of Clinical Pain? (eds Besson J.-M. and Guilbaud G.), pp. 41-54. Elsevier, Amsterdam. 4. Besson J.-M. and Chaouch A. (1987) Peripheral and spinal mechanisms of nociception. Physiol. Reu. 67, 67-186.

992

C.

ABBADIE

and J.-M. Bnssoh

5. Besson J.-M. and Guilbaud G. (1988) The Arthritic Rat as u Model of’ Clinical Pain’ Elsevier, Amsterdam. 6. Bullitt E. (1990) Expression of c-@-like protein as a marker for neuronal activity following noxious stirnu!atlc,il ifi the rat. J. camp. Neurol. 296, 517-530.

7. Calvin0 B., Besson J.-M., Mounier F., Kordon C. and Bluet-Pajot M.-T. (1992) Chronic pain induces a parLidl)xical increase in growth hormone secretion without affecting other hormones related to acute stress in the r<+t Pi,ir: (in press). 8. Calvin0 B., Crepon-Bernard M.-O. and Le Bars D. (1987) Parallel clinical and behavioral studies of adjuvant-induced arthritis in the rat: possible relationship with “chronic pain”. Behav. Brain Res. 24. 11 29. 9. Calvin0 B., Villanueia L. and Le Bars d. (1987) Dorsal horn (convergent) neurones in the intact anaesthetized arthritic rat. I. Segmental excitatory influences. Pain 28, 81-98. 10. Colpaert F. C., De Witte P., Maroli A. N., Awouters F., Niemegeers C. J. E. and Janssen P. A. (1980) Self-administration of the analgesic suprofen in arthritic rats: evidence of Mycohacterinm hutrricum-induced arthritis as an experimental model of chronic pain. Life Sci. 27, 921-928. Il. Colpaert F. C. and Van den Hoogen R. H. W. M. (1982) Ventilatory response to adjuvant arthritis in the rat Life. Sci. 31, 9577963. 12. Colpaert F. C. and Van den Hoogen R. H. W. M. (1983) Time course of the ventilatory response to adjuvant arthritis in the rat. Life. Sci. 33, 106551073. 13. Committee for Research and Ethical Issues of the I.A.S.P. (1980) Ethical standards for investigations of experimental pain in animals. Pain 9, 141-143. 14. De Castro Costa M., De Sutter P., Gybels J. and Van Hees J. (1981) Adjuvant-induced arthritis in rats: a possible animal model of chronic pain. Pain 10, 1733185. 15. Draisci G. and Iadarola M. J. (1989) Temporal analysis of increases in c-fos. preprodynorphin and preproenkephalin mRNAs in rat spinal cord. Molec. Brain Res. 6, 31-37. 16. Dray A. and Wood J. N. (1991) Nonopioid molecular signaling mechanisms involved in the nociception and antinociception. In Towards a New Pharmacatherapy af Pain (eds Basbaum A. I. and Besson J.-M.), pp. 21-34. John Wiley, Chichester. 17. Dubner R. (1991) Neuronal plasticity and pain following peripheral tissue inflammation or nerve injury. In Proceedings of the VIth World Congress on Pain (eds Bond M. R., Charlton J. E. and Woolf C. J.). pp. 263-276. Elsevier. Amsterdam. 18. Dubner R., Kenshalo D. R. Jr, Maixner W., Bushnell M. C. and Oliveras J. L. (1989) The correlation of monkey medullary dorsal horn neuronal activity and the perceived intensity of noxious heat stimuli. J. Neuroph.vsioI. 62, 450-457.

19. Gogas K. R., Presley R. W., Levine J. D. and Basbaum A. 1. (1991) The antinociceptive action of supraspinal opiords results from an increase in descending inhibitory control: correlation of nociceptive behavior and c-foes expression. Neuroscience 42, 6 17-628. 20. Handwerker H. 0. (1991) What peripheral mechanisms contribute to nociceptive transmission and hyperalgesia? In Towards a New Phurmucotherupy of Pain (eds Basbaum A. I. and Besson J.-M.), pp. 5519. John Wiley, Chichester. 21. Hsu S., Raine L. and Fanger H. (1981) Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabelled antibody (PAP) procedures. J. Histochem. Cytochem. 29, 577-580.

22. Hunt S. P., Pini A. and Evan G. (1987) Induction of c-fos-like protein in spinal cord neurons following sensory stimulation. Nature 328, 632-634. 23. Iadarola M. J., Brady L. S., Draisci G. and Dubner R. (1988) Enhancement of dynorphin gene expression in spinal cord following experimental inflammation: stimulus specificity, behavioral parameters and opioid receptor binding. Pain 35, 3 133326.

24. Lamotte C. C., Kapadia S. E. and Shapiro C. M. (1991) Central projections of the sciatic, saphenous, median, and ulnar nerves of the-rat demonstrated by transganglionic transport of choleragenoid-HRP (B-HRP) and wheat germ agalutinin-HRP (WGA-HRP). J. camp. Neurol. 311. 546-562. 25. h&ro A., German0 I., Giaccone G, Giordana M. T. and Schiffer D. (1985) I-Naphtol basic dye (I-NBD): an alternative to diaminobenzidine (DAB) in immunoperoxidase techniques. Hisiochemistry 83, 97-102. 26. Menetrey D. and Besson J.-M. (1982) Electrophysiological characteristics of dorsal horn cells in rats with cutaneous inflammation resulting from chronic arthritis. Pain 13, 343-364. 27. Menetrey D., Gannon A., Levin J. D. and Basbaum A. I. (1989) Expression of c-fos protein in interneurons and projection neurons of the rat spinal cord in response to noxious somatic, articular and visceral stimulation. J. camp Neural. 285, 177-195. 28. Menitrey D., de Pommery J., Baimbridge K. G. and Thomasset M. (1992) Calbindin-D28K (CaBP,,,)-like immunoreactivity in ascending projections: (I) trigemial nucleus caudials and dorsal vagal complex projections. Eur. J. Neurosci. 4, 61-69. 29. Millan M. J., Czlonkowski A., Pilcher C. W. T., Almeida 0. F. X., Millan M. H., Colpaert F. C. and Herz A. (1987) A model of chronic pain in the rat: functional correlates of alterations in the activity of opioid systems. J. ,Neeurosci. 7, 77-87. 30. Millan M. J., Millan M. H., Czlonkowski A., Hollt V., Pilcher C. W. T., Herz A. and Colpaert F. C. (1986) A model of chronic pain in the rat: response of multiple opioid systems to adjuvant-induced arthritis. J. Neurosci. 6, 899-906. 31. Molander C. and Grant G. (1986) Laminar distribution and somatotopic organization of primary afferent fibers from hindlimb nerves in the dorsal horn. A study by transganglionic transport of horseradish peroxidase in the rat. Neuroscience 19, 297-3 12.

32. Moland C., Xu Q. and Grant G. (1984) The cytoarchitectonic organization of the spinal cord in the rat: I. The lower thoracic and lumbosacral cord. J. camp. Neurol. Uo, 133-141. 33. Naranio J. R., Mellstriim B.. Achaval M. and Sassone-Corsi P. (1991) Molecular pathways of pain: Fos/Jun-mediated activaiion of a noncanonical AP-1 site in the prodynorphin gene. Neuron 6, 607-617. _ _ 34. Noeuchi K.. Kowalski K.. Traub R.. Solodkin A.. Iadarola M. J. and Ruda M. A. (1991) Dvnorohin exoression and Fos-like immunoreactivity’ following inflammation’induced hyperalgesia are colocalized inspinal cord neurons. Molec. Bruin Res. 10, 2277233.

c-fos expression in rat lumbar spinal cord

993

35. Pearson C. M., Waksman B. H. and Sharp J. T. (1961) Studies of polyarthritis and other lesions induced in rats by injection of mycobacterial adjuvant. J. exp. Med. 113, 485-509. 36. Pearson C. M. and Wood F. D. (1959) Studies of polyarthritis and other lesions induced in rats by injection of mycobacterial adjuvant. I. General clinical and pathologic characteristics and some modifying factors. Arthritis Rheum. 2, 440-459. 37. Pircio A. W., Fedele C. T. and Bierwagen M. E. (1975) A new method for the evaluation of analgesic activity using adjuvant-induced arthritis in the rat. Eur. J. Pharmac. 31, 207-215. 38. Presley R. W., Menetrey D., Levine J. D. and Basbaum A. I. (1990) Systemic morphine suppresses noxious stimulus-evoked Fos protein-like immunoreactivity in the rat spinal cord. J. Neurosci. 10, 323-335. 39. Sonnenberg J. L., Rauscher F. J. III, Morgan J. I. and Curran T. (1989) Regulation of proenkephalin by Fos and Jun. Science. 246, 1622-1625. 40. Sweet J. E. and Woolf C. J. (1985) The somatotopic organization of primary afferent terminals in the superficial laminae of the dorsal horn of the rat spinal cord. J. cotnp. Neural. 231, 66-77. 41. Tiille T. R., Castro-Lopes J. M., Coimbra A. and Zieglgiinsberger W. (1990) Opiates modify induction of c-fos proto-oncogene in the spinal cord of the rat following noxious stimulation. Neurosci. Letf. 111, 46-51. 42. Weihe E., Iadarola M. J., Nohr D., Miiller S., Millan M. J., Yanaihara N., Stein C. and Hem A. (1990) Sustained expression and colocalization of proenkephalin and prodynorphin opioids and c-fos protein in dorsal horn neurons revealed in arthritic rats. In New Lea& in Opioid Research (eds van Ree J. M., Mulder A. H., Wiegant V. M. and van Wimersma Greidanus T. B), pp. 92-94. Elsevier, Amsterdam. 43. Williams S., Evan G. I. and Hunt S. P. (1990) Changing patterns of c-fos induction in spinal neurons following thermal cutaneous stimulation in the rat. Neuroscience 36, 73-81. 44. Wisden W., Errington M. L., Williams S., Dunnett S. B., Waters C., Hitchcock S., Evan G., Bliss T. V. P. and Hunt S. P. (1990) Differential expression of immediate early genes in the hippocampus and spinal cord. Neuron 4,603-614. (Accepted 29 January 1992)