- Email: [email protected]
Histochemical changes of substance P, FRAP, serotonin and succinic dehydrogenase in the spinal cord of rats with adjuvant arthritis
Life Sciences, Vol. 36, pp. 1247-1254 Printed in the U.S.A.
Pergamon Pres
HISTOCHEMICAL CHANGES OF SUBSTANCF P, FRAP, SEROTONIN AND SUCCINIC DEHYDROGENASE IN THE SPINAL CORD OF RATS WITH ADJUVANT ARTHRITIS
J. Schoenen ! '
J. Van Hees ~, J. Gybels ~, M. de Castro Costa ~ j.j. Vanderhaeghen ~
Department of Neurology and Clinical Neurophysiology, Institute of Medicine, Baviere Hospital, University of Liege, 66 Bd de la Constitution, 4020 Liege, Belgium. ~Department of Neurology and Neurosurgery, University of Leuven, 3000 Leuven, Belgium. ~ L a b o r a t o r y of Neuropathology and Neuropeptide Research, Brugmann and Erasme University Hospitals, Queen Elisabeth Medical Foundation, 1020 Brussels, Belgium. (Received in final form January 24, 1985) Summary Various histochemical changes were found in spinal segmen~ L4-L5 of rats with adjuvant arthritis, predominantly 30 days after inoculation. A slight to marked increase of substance P immunoreactivity occured in laminae I, II and X. FRAP activity was enhanced in lamina II. Serotonin immunoreactivity was heavier in laminae I, VIII and IX in a few animals. The intensity of the histoenzymological reaction for succinic dehydrogenase increased in certain laminae VIII and X neurons. At day 15 of the disease the increase of substance P and FRAP activities was chiefly restricted to the medial portion of the superficial dorsal horn. There was a significant positive correlation between the scratching behaviour of arthritic rats and the substance P immunoreactivity in laminae X and I. If one accepts that scratching is pain-related, the data are consistent with a possible role of substance P in the chronic pain associated with adjuvant arthritis. They leave undetermined the significance of the other histochemical changes. Adjuvant-induced arthritis in the rat is considered to be a suitable animal model for chronic pain (1,2) Arthritic rats develop a striking scratching behaviour, that varies during the course of the disease and is probably pain-related (2). In normal mice, scratching is evoked by intaathecal injections of substance P (3), a putative neurotransmitter in primary nociceptive neurons. An increased content of substance P was found by radioimmunoassay (RIA) in sciatic nerve, dorsal roots and, to a lesser degree, in the dorsal half of the spinal L4-L5 segment of arthritic rats (5). There is convincing evidence for a possible role of substance P in the peripheral manifestations of adjuvant arthritis since capsa~cin, which is known to deplete substance P as well as the fluoride resistant isoenzyme of acid phosphatase (FRAP) in the rat primary sensory neurons (6,7) attenuates the inflammatory response and
I
correspondence
0024-3205/85 $3.00 + .00 Copyright (c) 1985 Pergamon Press Ltd.
1248
Arthritis and Spinal Histochemical
Changes
Vol. 36, No. 13, 1985
the increase of substance P (8). If a correlation could be established between scratching and a substance P increase in the spinal cord, this might suggest that substance P is also involved in the central mechanisms of the chronic pain associated with adjuvant arthriti6. Substance P is however not the only chemical implied in pain regulation. In an animal model of chronic pain one might expect modifications of other chemical substances known to act the spinal level, such as FRAP, which is contained in a subpopulation of afferent C fibers (9), or serotonin. Descending serotonergic pathways modulate pain signals within the spinal cord (10). Serotonin may coexist with substance P in brainstem neurons projecting to the spinal cord (11). Increase in the serotonin content of midbrain and hypothalamus (12) as well as increased spinal synthesis of serotonin (13) have been reported in adjuvant arthritis. Moreover in arthritic rats various spinal neurons might undergo metabolic changes, e.g. spinothalamic and spino-reticular cells or interneurons and motoneurons since they are activated respectively during nociception or during motor activities such as scratching. Succinic dehydrogenase, a mitochondrial enzyme of Krebs'cycle can be a marker of the neuronal metabolic state. Its histoenzymologic reaction increases in motoneurons when their metabolism is activated e.g. after axotomy (14). We have explored with histochemical methods the spinal content of substance P, FRAP, serotonin and succinic dehydrogenase in arthritic rats at various stages of the disease. Possible correlations between quantified scratching behaviour and spinal histochemical changes within their spinal cord were examined. The results of a specific RIA used to measure substance P immunoreactivity on the same material will be reported elsewhere. Methods Male Wistar rats, weighing about 250 g at the start of experimentation, were used in this study. Arthritis was produced by intradermal inoculation in the base of the tail with 0.05 ml of a paraffin oil suspension of heat-killed Mycobacterium butyricum (5mg/ml) (15). Rats used as controls were inoculated with only the sterile vehiculum (paraffin oil). Two consecutive series of animals were studied. In a first pilot experiment, 2 control and 2 arthritic rats were prepared by three of us (J.V.H., J.G., M. de C), sacrified 40 days after inoculation and their spinal cords were blindly processed with histochemical methods by the two other authors (J.S., J.V.). In a subsequent experiment 18 arthritic rats were studied, 6 at each of the following stages of the disease : 15 days (El to E6), 30 days (CI to C6) and 60 days (BI to B6) after inoculation with Mycobacterium butyricum. Two control rats (D2 & D3) were sacrified 30 days after inoculation with the sterile vehiculum. The scratching behaviour was quantified at each stage of the disease as described elsewhere (2). The animals were killed by decapitation and the spinal cord was dissected at the level of the lumbar enlargement. Two tissue blocks were prepared from the L4-L5 segments, where the sciatic nerve enters the spinal cord. One block was immediately frozen and cut at 20 ~m on a cryostat for the histochemical demonstration of FRAP (method of Barka and Anderson) (16) and succinic dehydrogenase (tetrazolium salt method) (17). The other tissue block was fixed in paraformaldehyde and sectioned at 20 ~m on a freezing microtome in order to visualize substance P and serotonin with the unlabelled peroxidase-antiperoxidase complex method of immunohistochemistry (18).
Vol. 36, No, 13, 1985
Arthritis and Spinal Histochemical
Changes
1249
The antiserum to substance P was raised in rabbits and tested for specificity by liquid and solid phase absorption tests (19). The characteristics of the antiserum to serotonin have been described by others (20). Both antisera were used in a 1/2000 dilution. The histochemical techniques were adapted in order to obtain submaximal staining, since previous studies (5, 12, 13, 14) suggested that enhanced rather than reduced activities were to be expected in arthritic rats. For each histochemical method tissue sections from control and arthritic animals were always processed simultaneously. This explains why only two control animals were used (in addition to 2 controls in the pilot experiment) and why the staining of substance P in control animals is several orders of magnitude fainter (see figure 2) than other published data (6,7). The animals were not perfused because fresh unfixed tissue was needed for the tetrazolium salt method and for a substance P RIA performed on the ramaining lumbar spinal cord. Consequently, red blood cells were stained with the immunocytochemical method, which hampered the photographic reproduction of the material, but not the study of immunoreactive fibers. Spinal cord sections adjacent to those used for histochemistry were stained with cresyl violet in order.to allow precise identification of cytoarchitectonic laminae (21). Statistical analyses of spinal histochemical changes in relation to the scratching behaviour were made using the Spearman rank correlation test. Results In the pilot-experiment, it was found by two of us that all the histochemical reactions were stronger in at least some spinal laminae of 2 rats. Substance P staining was intenser in laminae I, II and X. Lamina II FRAP activity was moderately stronger. Serotonin-immunoreactivity was heavier in laminae I, VIII and in the ventro-medial group of motoneurons. The reaction for succinic dehydrogenase was stronger in lamina VIII and X neurons than in motoneurons, while the opposite is found in the normal rat (14). Those 2 rats with stronger histochemical reactions had developped an arthritis since 40 days, while the 2 others were controls. In the subsequent experiment the histochemical reactions in the 3 groups of arthritic rats (15, 30 and 60 days after inoculation) were qualitatively evaluated using the 2 control rats as a reference. The results were expressed as slight, moderate or marked increase of activity in comparison to control rats and summarized in the diagram of figure 1 for substance P, FRAP and serotonin. The most pronounced changes were observed in the 6 rats sacrified 30 days after inoculation. Lamina I substance P immunoreactivity was increased markedly in 2 (CI, C3) (Fig. 2), moderately in | (C4) and slightly in 2 (C5,C6). Three rats (C|, C3, C4) had a moderate (fig. 3), two others (C5,C6) a slight increase of substance P in laminae II and X. Substance P was not modified in the other spinal laminae. In I animal (C2) it was similar to controls in the entire gray matter. In the 60 day group only | animal (BI) had moderately in increased substance P immunoreactivity in laminae I, II and X. The striking finding in the arthritic rats sacrified 15 days after inoculation was a moderate increase of substance P staining restricted to the medial portion of lamina II (animals El, E2, E3, E6). Lamina I substance P was moderately increased in 1 rat (El), slightly in another (E4). At this stage of the disease, substance P remained unchanged in lamina X of all animals and in all the spinal laminae of rat E5.
1250
Arthritis and Spinal Histochemical Changes
Relative intensity
Vol. 36, No. 13, 1985
i
iii
M
I X
SP (I,II,X) 3 2 1
:RAP (lO
IL i //ira
21 0
~-HT D2 D3 E1 E2 E3 E4 E5 E6 CONTROLS ARTHRITIS 15d
C1 C2 C3 C4 C 5 C 6 ARTHRITIS 30d
B1 B2 B3 B4 B5 B6 ARTHRITIS 60d
FIG. l Diagrammatic representation of substance P (SP), FRAP and serotonin (5-HT) activities in arthritic rats (hatched columns) studied 15 days (El to E6), 30 days (CI to C6), 60 days (BI to B6) after inoculation and in control animals (D2, D3) (black columns). The relative staining intensity is indicated on the abcissa using the control rats as a reference . 0 : intensity similar to controls; 1 : slight; 2 : moderate; 3 : marked increase of intensity. Subtance P : at 15 days im~unoreactivity in lamina I (left hemi-column) and II (right hemi-column) is represented; the medial (M) and lateral (L) portions of lamina II are shown. At 30 and 60 days lamina I is considered in the left hemi-colum, laminae II and X combined in the right hemi-column. FRAP : at 15 days the left hemicolumn represents activity in the medial (M) portion of lamina II, the right hemi-column in its lateral (L) portion. Serotonin : immunoreactivity is considered only in lamina I. The changes of lamina II FRAP activity were comparable to those of substance P (Fig. I). Thirty days after inoculation, FRAP was markedly increased in I rat (C3), moderately in 3 (CI, C4, C6) and unchanged in 2 animals (C2, C5). At 60 days, a marked increase of activity was ~und in | rat (BI), a moderate increase in 2 others (B3, B4). At 15 days, FRAP, like substance P, was chiefly enhanced in the medial portion of lamina II (animals El, E2, E3, E5) (fig. 3). Serotonin immunoreactivity was enhanced only in a few animals in lamina I and the medial ventral horn i.e. lamina VIII and the ventro-medial motoneuronal group. Changes were moderate in 2 rats at 30 days (CI, C3) in 2 rats at 60 days (B3, B4) and in I animal at 15 days (E2).
Vol. 36, No. 13, 1985
art. O~
Arthritis and Spinal Histochemical
Changes
"
~::,
#
1251
"
.
II p,
j-
, ~
- -
-
,
contr.
~,~
contr.
,"
III
,
i --
,,
III
FIG. 2 Substance (art. C|) low-power indicates of lamina
P immunoreactivity in the dorsal horn of arthritic rat C! and in a control animal (contr.). Upper & lower left : micrographs of transversely cut dorsal horn (the arrow lamina II). Upper & lower right : high-power micrographs II. Bars : 5 0 ~ m .
The histochemical activity of succinic dehydrogenase in laminae VIII and X neurons was enhanced markedly in 2 animals (C3, C4) (Fig. 3) and slightly in 2 others (C2, C6) 30 days after inoculation. At 60 days a moderate increase was found in 3 rats (B!, B3, B4) and at 15 days a slight increase in one (ES). Within the group of 18 arthritic rats correlations were sought between these histochemical changes and the quantitative data on scratching behaviour obtained independantly by 3 of the authors. Using the Spearman rank test, a strong correlation was found between scratching and increased lamina X substance P immunoreactivity (rs = 0.75) at a highly significant statistical level (one-tailed p < 0 . O O 0 5 ) . Correlation between scratching and increased substance P in lamina I was weak (rs = 0.485), but reached the level of significance (p
Arthritis and Spinal Histochemical Changes
1252
Vol. 36, No. 13, 1985
art. E1
~,~
~"~ X
-
•
dontff~~
o
arf,
~J
t
IV
M
CC
:
III
L
(
/
i
%
~ 4
1 "-j
,@.
J •
FIG. 3 Upper & lower left : Substance P immunoreactivity in lamina X of arthritic rat C3 and of a control animal (contr.). CC : central canal. Upper right : FRAP activity in lamina II of arthritic rat El. The activity is diffusely increased in lamina II, but more so in its medial (M) portion. Lower right : Succinic dehydrogenase activity in the ventral cord of arthritic rat C4. Neurons located in laminae VIII and X appear intensely stained. Bars : 50 ~m. Discussion The present study indicates that histochem~cal changes can be demons trated in the spinal cord of rats suffering from adjuvant induced arthritis. These changes, already suggested by a blindly conducted pilot-experiment on a few animals, were confirmed in a large number of animals sacrified at days 15, 30 and 60 of the disease. They were maximal 30 days after inoculation, i.e. in the last part of the acute phase according to clinical and behavioural parameters (2). There was great variability of the histochemical staining between animals studied at an identical phase of the arthritis. This is most probably related to the variable severity of the disease, which is demonstrated by the scattered values of joint circumferences and behavioural data (2). The most prominent modifications were observed for substance P and FRAP. Substance P immunoreactivity was increased in lamina I, and to a lesser degree in lamina II. This finding is in accordance with a radioimmunoassay study showing an increased content of substance P in the dorsal spinal cord of arthritic rats (5). Substance P activity was also enhanced in lamina X, the area around the central canal. The exact role of this area is not known, but its cells have anatomical and functional similarities with the specific nociceptive neurons in the outer layers of the dorsal horn (22). It contains a number of different peptides and could represent the caudal part of a
Vol. 36, No. 13, 1985
Arthritis and Spinal Histochemical Changes
1253
continuum running the entire length of the neuraxis and containing such structures as the hypothalamus, the periaqueductal gray or the solitary tract, all implied in pain and its visceral components (23). When compared to radioimmunoassay studies, immunocytochemical methods may lack sensitivity, but they have the advantage of detecting very localized changes in the tissue content of a neuropeptide (19). The enzyme FRAP, which is only demonstrable in rodent dorsal root ganglia and spinal cord, is found in a population of afferent C fibers distinct from those containing substance P (7). As substance P, it is depleted after neonatal capsa~cin treatment (7) and after peripheral nerve lesion (24). Unlike substance P, FRAP is not released from dorsal horn terminals after stimulation of peripheral nerve. Increased FRAP staining in arthritic rats can be due to a change in the concentration or possibly, in the activity of the enzyme. The histochemical method used in this study is not able to separate these two possibilities and detailed biochemical data on FRAP in rat lamina II are lacking. The majority of serotonergic afferents to the spinal cord originate in the hindbrain (25, 26). The nucleus raphe magnus innervates the dorsal horn where it modulates pain transmission. The nuclei raphe obscurus and pallidus send fibers to the ventral horns and can influence spinal reflexes. In our material serotonin irmnunoreactivity was heavier in the dorsal (lamina I) and in the ventral horn (laminae VIII and IX) of several arthritic rats, suggesting that both descending serotonergic systems may be hyperactive in these animals. It is improbable that the histochemical changes reported in the present study are related only to the systemic manifestations of adjuvant arthritis, since they occur in certain specific laminae and not in the entire spinal gray matter. There was no meningitis or vasculitis detectable in our material. Another finding arguing against non specific histochemical changes was the very localized increase of substance P and FRAP activity in the medial portion of lamina II 15 days after inoculation. The medial portion of the dorsal horn receives afferents from the distal parts of the limbs (27). In adjuvant arthritis the distal joints of the hindlimbs, i.e. the tibiotarsal joints, are the most affected and begin to swell 10 days after inoculation (2). There is thus a close somatotopical link between the peripheral manifestations of adjuvant arthritis and the histochemical changes in the spinal dorsal horn. Capsaicin attenuates the inflammatory response in arthritic rats, but also the RIA increase of substance P in peripheral nerve and dorsal spinal cord (8)° These data underline the possible role of substance P in the peripheral manifestations of adjuvant arthritis but leave open the question whether substance P is involved in the chronic pain associated with the disease. Our results may help answer this question, since we found a significant correlation between the scratching behaviour of arthritic rats and the increase of substance P staining in laminae I and X. Several criteria suggest that scratching is possibly a chronic pain-related behaviour in arthritic rats : it is significantly increased in these animals, even more so at a stage (30 days after inoculation) when peripheral inflammation begins to decline; it is depressed by morphine, this effect being blocked by naloxone (2), while pruritus, frequently associated with scratching in humans, cannot be suppressed by morphine (28). Increased substance P in laminae I and X may thus be relat~ to the chronic pain induced by adjuvant arthritis. Enhanced succinic dehydrogenase activity in laminae VIII and X neurons is also correlated with scratching. It remains to be determined if these metabolically hyperactive cells correspond to spino-thalamic, spino-reticular or internuncial neurons, all known to be located within these regions of the spinal gray matter (21).
1254
Arthritis and Spinal Histochemical Changes
Vol. 36, No. 13, 1985
Acknowledgments The authors are most grateful to Dr. H.W.M. Steinbusch for the generous gift of an antiserum to serotonin. This study was partly supported by research grant n ° 1.5.454.83F from the National Fund for Scientific Research (Belgium), of which J. Schoenen is a Research Associate. References I. F.C. COLPAERT, Life Sci. 24 120]-1209 (1979) 2. M. DE CASTRO COSTA, P. DE SUTTER, J. GYBELS and J. VAN HEES, Pain I0 173-185 (]981) 3. J.L.K. HYLDEN and G.L. WILCOX, Brain Res 217 212-215 (1981) 4. M. OTSUKA and S. KONISHI, Tins 6 317-320 (1983) 5. F. LEMBECK, J. DONNERER and J.C. COLPAERT, Neuropeptides ! 175-180 (]981) 6. G. JANCSO, G. KIRALY and A. JANCSO-GABOR, Nature 270 74]-743 (1977) 7. J.I. NAGY, S.P. HUNT, L.L. IVERSEN and P.C. EMSON,LNeuroscience 6 |923-1934 (1981) 8. F.C. COLPAERT, J. DONNERER and F. LEMBECK, Life Sci. 32 1827-1834 (1983) 9. J.I. NAGY and S.P. HUNT, Neuroscience 7 88-97 (1981) iO. A.I. BASBAUM and H.L. FIELDS, Ann. Rev~ Neurosci. 7 309-338 (1984) 11. T. HOKFELT, A. LJUNGDAHL, H. STEINBUSCH, A. VERHOFSTAD, G. NILSSON, E. BRODIN, B. PE[~NOW and M. GOLDSTEIN, Neuroscience 3 517-538 (1978) 12. R.D. SOFIA and H.B. VASSAR, Arch. Int. Pharmacodyn.--211 74-79 (1974) 13. J. WEIL-FUGAZZA, F. GODEFROY, M. BINEAU-THUROTTE and J.M. BESSON, Progress in Tryptophan and Serotonin Research, ed. by H.G. Schlossbergen, W. Kochen, B. Linzen and H. Steihart, 405-408, Walter de Gruyter & Co, Berlin (1984) 14. J. SCHOENEN, Bull. Ass. Anat. 58 415-425 (1973) 15. F. AWOUTERS, P.M.H. LENAERTS and C.J.E. NIEMEGEERS, Drug Res. 26 40-43 (1976) 16. T. BARKA and P.J. ANDERSON, J. Histochem. Cytochem. 10 741-753 (1962) 17. M.A. GEREBTZOFF, Bull. Acad. Roy. M~d. Belge 1 0 337-361 (1970) 18. L.A. STERNBERGER, P.H. HARDI, J.J. CUCULIS and H.G. MEYER, J. Histochem. Cytochem. 18 315-333 (1970) 19. F. VANDENSANDE, J. Neurosc. Meth. ] 3-23 (1979) 20. H.W.M. STEINBUSCH, A.A.J. VERHOFSTAD and H.W.J. JOOSTEN, Neuroscience 6 557-620 (1981) 2]. 7. SCHOENEN, L'organisation neuronale de la moelle ~pinigre de l'hormne. Editions Sciences et Lettres, Liege (1981) 22. R.L. NAHIM, A.M. MADSEN and G.J. GIESLER Jr, J. Comp. Neurol. 220 321-335 (1983) 23. S.J. GIBSON, J.M. POLAK, S.R. BLOOM, P.D. WALL, J. Comp. Neurol. 201 65-79 (1981) 24. J. SCHOENEN, C. BUDO and G. PONCELET, C.R.Soc. Biol. 162 2035-2037 (1968) 25. A. DAHLSTR~M and K. FUXE, Acta Physiol. Scand. 62 (suppl. 232) 1-55 (1964) 26. H.L. FIELDS and A.I. BASBAUM, Ann. Rev. Physiol. 40 217-248 (1978) 27. J.M. SPRAGUE and H. HA, Orsanization of the Spinal Cord, ed. by J.C. Eccles and J.P. Schad~. Progr. Brain Res 11-- 120-154 (1964) 28. J.H. HERNDON, Int. J. Dermatol. 14 477 (]975)
Pergamon Pres
HISTOCHEMICAL CHANGES OF SUBSTANCF P, FRAP, SEROTONIN AND SUCCINIC DEHYDROGENASE IN THE SPINAL CORD OF RATS WITH ADJUVANT ARTHRITIS
J. Schoenen ! '
J. Van Hees ~, J. Gybels ~, M. de Castro Costa ~ j.j. Vanderhaeghen ~
Department of Neurology and Clinical Neurophysiology, Institute of Medicine, Baviere Hospital, University of Liege, 66 Bd de la Constitution, 4020 Liege, Belgium. ~Department of Neurology and Neurosurgery, University of Leuven, 3000 Leuven, Belgium. ~ L a b o r a t o r y of Neuropathology and Neuropeptide Research, Brugmann and Erasme University Hospitals, Queen Elisabeth Medical Foundation, 1020 Brussels, Belgium. (Received in final form January 24, 1985) Summary Various histochemical changes were found in spinal segmen~ L4-L5 of rats with adjuvant arthritis, predominantly 30 days after inoculation. A slight to marked increase of substance P immunoreactivity occured in laminae I, II and X. FRAP activity was enhanced in lamina II. Serotonin immunoreactivity was heavier in laminae I, VIII and IX in a few animals. The intensity of the histoenzymological reaction for succinic dehydrogenase increased in certain laminae VIII and X neurons. At day 15 of the disease the increase of substance P and FRAP activities was chiefly restricted to the medial portion of the superficial dorsal horn. There was a significant positive correlation between the scratching behaviour of arthritic rats and the substance P immunoreactivity in laminae X and I. If one accepts that scratching is pain-related, the data are consistent with a possible role of substance P in the chronic pain associated with adjuvant arthritis. They leave undetermined the significance of the other histochemical changes. Adjuvant-induced arthritis in the rat is considered to be a suitable animal model for chronic pain (1,2) Arthritic rats develop a striking scratching behaviour, that varies during the course of the disease and is probably pain-related (2). In normal mice, scratching is evoked by intaathecal injections of substance P (3), a putative neurotransmitter in primary nociceptive neurons. An increased content of substance P was found by radioimmunoassay (RIA) in sciatic nerve, dorsal roots and, to a lesser degree, in the dorsal half of the spinal L4-L5 segment of arthritic rats (5). There is convincing evidence for a possible role of substance P in the peripheral manifestations of adjuvant arthritis since capsa~cin, which is known to deplete substance P as well as the fluoride resistant isoenzyme of acid phosphatase (FRAP) in the rat primary sensory neurons (6,7) attenuates the inflammatory response and
I
correspondence
0024-3205/85 $3.00 + .00 Copyright (c) 1985 Pergamon Press Ltd.
1248
Arthritis and Spinal Histochemical
Changes
Vol. 36, No. 13, 1985
the increase of substance P (8). If a correlation could be established between scratching and a substance P increase in the spinal cord, this might suggest that substance P is also involved in the central mechanisms of the chronic pain associated with adjuvant arthriti6. Substance P is however not the only chemical implied in pain regulation. In an animal model of chronic pain one might expect modifications of other chemical substances known to act the spinal level, such as FRAP, which is contained in a subpopulation of afferent C fibers (9), or serotonin. Descending serotonergic pathways modulate pain signals within the spinal cord (10). Serotonin may coexist with substance P in brainstem neurons projecting to the spinal cord (11). Increase in the serotonin content of midbrain and hypothalamus (12) as well as increased spinal synthesis of serotonin (13) have been reported in adjuvant arthritis. Moreover in arthritic rats various spinal neurons might undergo metabolic changes, e.g. spinothalamic and spino-reticular cells or interneurons and motoneurons since they are activated respectively during nociception or during motor activities such as scratching. Succinic dehydrogenase, a mitochondrial enzyme of Krebs'cycle can be a marker of the neuronal metabolic state. Its histoenzymologic reaction increases in motoneurons when their metabolism is activated e.g. after axotomy (14). We have explored with histochemical methods the spinal content of substance P, FRAP, serotonin and succinic dehydrogenase in arthritic rats at various stages of the disease. Possible correlations between quantified scratching behaviour and spinal histochemical changes within their spinal cord were examined. The results of a specific RIA used to measure substance P immunoreactivity on the same material will be reported elsewhere. Methods Male Wistar rats, weighing about 250 g at the start of experimentation, were used in this study. Arthritis was produced by intradermal inoculation in the base of the tail with 0.05 ml of a paraffin oil suspension of heat-killed Mycobacterium butyricum (5mg/ml) (15). Rats used as controls were inoculated with only the sterile vehiculum (paraffin oil). Two consecutive series of animals were studied. In a first pilot experiment, 2 control and 2 arthritic rats were prepared by three of us (J.V.H., J.G., M. de C), sacrified 40 days after inoculation and their spinal cords were blindly processed with histochemical methods by the two other authors (J.S., J.V.). In a subsequent experiment 18 arthritic rats were studied, 6 at each of the following stages of the disease : 15 days (El to E6), 30 days (CI to C6) and 60 days (BI to B6) after inoculation with Mycobacterium butyricum. Two control rats (D2 & D3) were sacrified 30 days after inoculation with the sterile vehiculum. The scratching behaviour was quantified at each stage of the disease as described elsewhere (2). The animals were killed by decapitation and the spinal cord was dissected at the level of the lumbar enlargement. Two tissue blocks were prepared from the L4-L5 segments, where the sciatic nerve enters the spinal cord. One block was immediately frozen and cut at 20 ~m on a cryostat for the histochemical demonstration of FRAP (method of Barka and Anderson) (16) and succinic dehydrogenase (tetrazolium salt method) (17). The other tissue block was fixed in paraformaldehyde and sectioned at 20 ~m on a freezing microtome in order to visualize substance P and serotonin with the unlabelled peroxidase-antiperoxidase complex method of immunohistochemistry (18).
Vol. 36, No, 13, 1985
Arthritis and Spinal Histochemical
Changes
1249
The antiserum to substance P was raised in rabbits and tested for specificity by liquid and solid phase absorption tests (19). The characteristics of the antiserum to serotonin have been described by others (20). Both antisera were used in a 1/2000 dilution. The histochemical techniques were adapted in order to obtain submaximal staining, since previous studies (5, 12, 13, 14) suggested that enhanced rather than reduced activities were to be expected in arthritic rats. For each histochemical method tissue sections from control and arthritic animals were always processed simultaneously. This explains why only two control animals were used (in addition to 2 controls in the pilot experiment) and why the staining of substance P in control animals is several orders of magnitude fainter (see figure 2) than other published data (6,7). The animals were not perfused because fresh unfixed tissue was needed for the tetrazolium salt method and for a substance P RIA performed on the ramaining lumbar spinal cord. Consequently, red blood cells were stained with the immunocytochemical method, which hampered the photographic reproduction of the material, but not the study of immunoreactive fibers. Spinal cord sections adjacent to those used for histochemistry were stained with cresyl violet in order.to allow precise identification of cytoarchitectonic laminae (21). Statistical analyses of spinal histochemical changes in relation to the scratching behaviour were made using the Spearman rank correlation test. Results In the pilot-experiment, it was found by two of us that all the histochemical reactions were stronger in at least some spinal laminae of 2 rats. Substance P staining was intenser in laminae I, II and X. Lamina II FRAP activity was moderately stronger. Serotonin-immunoreactivity was heavier in laminae I, VIII and in the ventro-medial group of motoneurons. The reaction for succinic dehydrogenase was stronger in lamina VIII and X neurons than in motoneurons, while the opposite is found in the normal rat (14). Those 2 rats with stronger histochemical reactions had developped an arthritis since 40 days, while the 2 others were controls. In the subsequent experiment the histochemical reactions in the 3 groups of arthritic rats (15, 30 and 60 days after inoculation) were qualitatively evaluated using the 2 control rats as a reference. The results were expressed as slight, moderate or marked increase of activity in comparison to control rats and summarized in the diagram of figure 1 for substance P, FRAP and serotonin. The most pronounced changes were observed in the 6 rats sacrified 30 days after inoculation. Lamina I substance P immunoreactivity was increased markedly in 2 (CI, C3) (Fig. 2), moderately in | (C4) and slightly in 2 (C5,C6). Three rats (C|, C3, C4) had a moderate (fig. 3), two others (C5,C6) a slight increase of substance P in laminae II and X. Substance P was not modified in the other spinal laminae. In I animal (C2) it was similar to controls in the entire gray matter. In the 60 day group only | animal (BI) had moderately in increased substance P immunoreactivity in laminae I, II and X. The striking finding in the arthritic rats sacrified 15 days after inoculation was a moderate increase of substance P staining restricted to the medial portion of lamina II (animals El, E2, E3, E6). Lamina I substance P was moderately increased in 1 rat (El), slightly in another (E4). At this stage of the disease, substance P remained unchanged in lamina X of all animals and in all the spinal laminae of rat E5.
1250
Arthritis and Spinal Histochemical Changes
Relative intensity
Vol. 36, No. 13, 1985
i
iii
M
I X
SP (I,II,X) 3 2 1
:RAP (lO
IL i //ira
21 0
~-HT D2 D3 E1 E2 E3 E4 E5 E6 CONTROLS ARTHRITIS 15d
C1 C2 C3 C4 C 5 C 6 ARTHRITIS 30d
B1 B2 B3 B4 B5 B6 ARTHRITIS 60d
FIG. l Diagrammatic representation of substance P (SP), FRAP and serotonin (5-HT) activities in arthritic rats (hatched columns) studied 15 days (El to E6), 30 days (CI to C6), 60 days (BI to B6) after inoculation and in control animals (D2, D3) (black columns). The relative staining intensity is indicated on the abcissa using the control rats as a reference . 0 : intensity similar to controls; 1 : slight; 2 : moderate; 3 : marked increase of intensity. Subtance P : at 15 days im~unoreactivity in lamina I (left hemi-column) and II (right hemi-column) is represented; the medial (M) and lateral (L) portions of lamina II are shown. At 30 and 60 days lamina I is considered in the left hemi-colum, laminae II and X combined in the right hemi-column. FRAP : at 15 days the left hemicolumn represents activity in the medial (M) portion of lamina II, the right hemi-column in its lateral (L) portion. Serotonin : immunoreactivity is considered only in lamina I. The changes of lamina II FRAP activity were comparable to those of substance P (Fig. I). Thirty days after inoculation, FRAP was markedly increased in I rat (C3), moderately in 3 (CI, C4, C6) and unchanged in 2 animals (C2, C5). At 60 days, a marked increase of activity was ~und in | rat (BI), a moderate increase in 2 others (B3, B4). At 15 days, FRAP, like substance P, was chiefly enhanced in the medial portion of lamina II (animals El, E2, E3, E5) (fig. 3). Serotonin immunoreactivity was enhanced only in a few animals in lamina I and the medial ventral horn i.e. lamina VIII and the ventro-medial motoneuronal group. Changes were moderate in 2 rats at 30 days (CI, C3) in 2 rats at 60 days (B3, B4) and in I animal at 15 days (E2).
Vol. 36, No. 13, 1985
art. O~
Arthritis and Spinal Histochemical
Changes
"
~::,
#
1251
"
.
II p,
j-
, ~
- -
-
,
contr.
~,~
contr.
,"
III
,
i --
,,
III
FIG. 2 Substance (art. C|) low-power indicates of lamina
P immunoreactivity in the dorsal horn of arthritic rat C! and in a control animal (contr.). Upper & lower left : micrographs of transversely cut dorsal horn (the arrow lamina II). Upper & lower right : high-power micrographs II. Bars : 5 0 ~ m .
The histochemical activity of succinic dehydrogenase in laminae VIII and X neurons was enhanced markedly in 2 animals (C3, C4) (Fig. 3) and slightly in 2 others (C2, C6) 30 days after inoculation. At 60 days a moderate increase was found in 3 rats (B!, B3, B4) and at 15 days a slight increase in one (ES). Within the group of 18 arthritic rats correlations were sought between these histochemical changes and the quantitative data on scratching behaviour obtained independantly by 3 of the authors. Using the Spearman rank test, a strong correlation was found between scratching and increased lamina X substance P immunoreactivity (rs = 0.75) at a highly significant statistical level (one-tailed p < 0 . O O 0 5 ) . Correlation between scratching and increased substance P in lamina I was weak (rs = 0.485), but reached the level of significance (p
Arthritis and Spinal Histochemical Changes
1252
Vol. 36, No. 13, 1985
art. E1
~,~
~"~ X
-
•
dontff~~
o
arf,
~J
t
IV
M
CC
:
III
L
(
/
i
%
~ 4
1 "-j
,@.
J •
FIG. 3 Upper & lower left : Substance P immunoreactivity in lamina X of arthritic rat C3 and of a control animal (contr.). CC : central canal. Upper right : FRAP activity in lamina II of arthritic rat El. The activity is diffusely increased in lamina II, but more so in its medial (M) portion. Lower right : Succinic dehydrogenase activity in the ventral cord of arthritic rat C4. Neurons located in laminae VIII and X appear intensely stained. Bars : 50 ~m. Discussion The present study indicates that histochem~cal changes can be demons trated in the spinal cord of rats suffering from adjuvant induced arthritis. These changes, already suggested by a blindly conducted pilot-experiment on a few animals, were confirmed in a large number of animals sacrified at days 15, 30 and 60 of the disease. They were maximal 30 days after inoculation, i.e. in the last part of the acute phase according to clinical and behavioural parameters (2). There was great variability of the histochemical staining between animals studied at an identical phase of the arthritis. This is most probably related to the variable severity of the disease, which is demonstrated by the scattered values of joint circumferences and behavioural data (2). The most prominent modifications were observed for substance P and FRAP. Substance P immunoreactivity was increased in lamina I, and to a lesser degree in lamina II. This finding is in accordance with a radioimmunoassay study showing an increased content of substance P in the dorsal spinal cord of arthritic rats (5). Substance P activity was also enhanced in lamina X, the area around the central canal. The exact role of this area is not known, but its cells have anatomical and functional similarities with the specific nociceptive neurons in the outer layers of the dorsal horn (22). It contains a number of different peptides and could represent the caudal part of a
Vol. 36, No. 13, 1985
Arthritis and Spinal Histochemical Changes
1253
continuum running the entire length of the neuraxis and containing such structures as the hypothalamus, the periaqueductal gray or the solitary tract, all implied in pain and its visceral components (23). When compared to radioimmunoassay studies, immunocytochemical methods may lack sensitivity, but they have the advantage of detecting very localized changes in the tissue content of a neuropeptide (19). The enzyme FRAP, which is only demonstrable in rodent dorsal root ganglia and spinal cord, is found in a population of afferent C fibers distinct from those containing substance P (7). As substance P, it is depleted after neonatal capsa~cin treatment (7) and after peripheral nerve lesion (24). Unlike substance P, FRAP is not released from dorsal horn terminals after stimulation of peripheral nerve. Increased FRAP staining in arthritic rats can be due to a change in the concentration or possibly, in the activity of the enzyme. The histochemical method used in this study is not able to separate these two possibilities and detailed biochemical data on FRAP in rat lamina II are lacking. The majority of serotonergic afferents to the spinal cord originate in the hindbrain (25, 26). The nucleus raphe magnus innervates the dorsal horn where it modulates pain transmission. The nuclei raphe obscurus and pallidus send fibers to the ventral horns and can influence spinal reflexes. In our material serotonin irmnunoreactivity was heavier in the dorsal (lamina I) and in the ventral horn (laminae VIII and IX) of several arthritic rats, suggesting that both descending serotonergic systems may be hyperactive in these animals. It is improbable that the histochemical changes reported in the present study are related only to the systemic manifestations of adjuvant arthritis, since they occur in certain specific laminae and not in the entire spinal gray matter. There was no meningitis or vasculitis detectable in our material. Another finding arguing against non specific histochemical changes was the very localized increase of substance P and FRAP activity in the medial portion of lamina II 15 days after inoculation. The medial portion of the dorsal horn receives afferents from the distal parts of the limbs (27). In adjuvant arthritis the distal joints of the hindlimbs, i.e. the tibiotarsal joints, are the most affected and begin to swell 10 days after inoculation (2). There is thus a close somatotopical link between the peripheral manifestations of adjuvant arthritis and the histochemical changes in the spinal dorsal horn. Capsaicin attenuates the inflammatory response in arthritic rats, but also the RIA increase of substance P in peripheral nerve and dorsal spinal cord (8)° These data underline the possible role of substance P in the peripheral manifestations of adjuvant arthritis but leave open the question whether substance P is involved in the chronic pain associated with the disease. Our results may help answer this question, since we found a significant correlation between the scratching behaviour of arthritic rats and the increase of substance P staining in laminae I and X. Several criteria suggest that scratching is possibly a chronic pain-related behaviour in arthritic rats : it is significantly increased in these animals, even more so at a stage (30 days after inoculation) when peripheral inflammation begins to decline; it is depressed by morphine, this effect being blocked by naloxone (2), while pruritus, frequently associated with scratching in humans, cannot be suppressed by morphine (28). Increased substance P in laminae I and X may thus be relat~ to the chronic pain induced by adjuvant arthritis. Enhanced succinic dehydrogenase activity in laminae VIII and X neurons is also correlated with scratching. It remains to be determined if these metabolically hyperactive cells correspond to spino-thalamic, spino-reticular or internuncial neurons, all known to be located within these regions of the spinal gray matter (21).
1254
Arthritis and Spinal Histochemical Changes
Vol. 36, No. 13, 1985
Acknowledgments The authors are most grateful to Dr. H.W.M. Steinbusch for the generous gift of an antiserum to serotonin. This study was partly supported by research grant n ° 1.5.454.83F from the National Fund for Scientific Research (Belgium), of which J. Schoenen is a Research Associate. References I. F.C. COLPAERT, Life Sci. 24 120]-1209 (1979) 2. M. DE CASTRO COSTA, P. DE SUTTER, J. GYBELS and J. VAN HEES, Pain I0 173-185 (]981) 3. J.L.K. HYLDEN and G.L. WILCOX, Brain Res 217 212-215 (1981) 4. M. OTSUKA and S. KONISHI, Tins 6 317-320 (1983) 5. F. LEMBECK, J. DONNERER and J.C. COLPAERT, Neuropeptides ! 175-180 (]981) 6. G. JANCSO, G. KIRALY and A. JANCSO-GABOR, Nature 270 74]-743 (1977) 7. J.I. NAGY, S.P. HUNT, L.L. IVERSEN and P.C. EMSON,LNeuroscience 6 |923-1934 (1981) 8. F.C. COLPAERT, J. DONNERER and F. LEMBECK, Life Sci. 32 1827-1834 (1983) 9. J.I. NAGY and S.P. HUNT, Neuroscience 7 88-97 (1981) iO. A.I. BASBAUM and H.L. FIELDS, Ann. Rev~ Neurosci. 7 309-338 (1984) 11. T. HOKFELT, A. LJUNGDAHL, H. STEINBUSCH, A. VERHOFSTAD, G. NILSSON, E. BRODIN, B. PE[~NOW and M. GOLDSTEIN, Neuroscience 3 517-538 (1978) 12. R.D. SOFIA and H.B. VASSAR, Arch. Int. Pharmacodyn.--211 74-79 (1974) 13. J. WEIL-FUGAZZA, F. GODEFROY, M. BINEAU-THUROTTE and J.M. BESSON, Progress in Tryptophan and Serotonin Research, ed. by H.G. Schlossbergen, W. Kochen, B. Linzen and H. Steihart, 405-408, Walter de Gruyter & Co, Berlin (1984) 14. J. SCHOENEN, Bull. Ass. Anat. 58 415-425 (1973) 15. F. AWOUTERS, P.M.H. LENAERTS and C.J.E. NIEMEGEERS, Drug Res. 26 40-43 (1976) 16. T. BARKA and P.J. ANDERSON, J. Histochem. Cytochem. 10 741-753 (1962) 17. M.A. GEREBTZOFF, Bull. Acad. Roy. M~d. Belge 1 0 337-361 (1970) 18. L.A. STERNBERGER, P.H. HARDI, J.J. CUCULIS and H.G. MEYER, J. Histochem. Cytochem. 18 315-333 (1970) 19. F. VANDENSANDE, J. Neurosc. Meth. ] 3-23 (1979) 20. H.W.M. STEINBUSCH, A.A.J. VERHOFSTAD and H.W.J. JOOSTEN, Neuroscience 6 557-620 (1981) 2]. 7. SCHOENEN, L'organisation neuronale de la moelle ~pinigre de l'hormne. Editions Sciences et Lettres, Liege (1981) 22. R.L. NAHIM, A.M. MADSEN and G.J. GIESLER Jr, J. Comp. Neurol. 220 321-335 (1983) 23. S.J. GIBSON, J.M. POLAK, S.R. BLOOM, P.D. WALL, J. Comp. Neurol. 201 65-79 (1981) 24. J. SCHOENEN, C. BUDO and G. PONCELET, C.R.Soc. Biol. 162 2035-2037 (1968) 25. A. DAHLSTR~M and K. FUXE, Acta Physiol. Scand. 62 (suppl. 232) 1-55 (1964) 26. H.L. FIELDS and A.I. BASBAUM, Ann. Rev. Physiol. 40 217-248 (1978) 27. J.M. SPRAGUE and H. HA, Orsanization of the Spinal Cord, ed. by J.C. Eccles and J.P. Schad~. Progr. Brain Res 11-- 120-154 (1964) 28. J.H. HERNDON, Int. J. Dermatol. 14 477 (]975)