- Email: [email protected]
Theophylline-induced nociceptive behavioral response in mice: Possible indirect interaction with spinal N-methyl-d-aspartate receptors
Neurochem. Int. Vol. 22, No. 1, pp. 69-74, 1993
0197-0186/93 $5.00+0.00 Pergamon Press Ltd
Printed in Great Britain
THEOPHYLLINE-INDUCED NOCICEPTIVE BEHAVIORAL RESPONSE IN MICE: POSSIBLE INDIRECT INTERACTION WITH SPINAL N-METHYL-D-ASPARTATE RECEPTORS HIDEYO NAGAOKA,1 SHINOBU SAKURADA,2. TSUKASA SAKURADA,2 SATOSHI TAKEDA, 2 YOSHITO NAKAGAWA,l KENSUKE KISARA2 a n d YUICHIRO ARAI3 ~Department of Pharmacy, Yamagata University Hospital, lida-Nishi Yamagata 990, Japan 2Department of Pharmacology, Tohoku College of Pharmacy, Komatsushima, Aoba-ku, Sendai 981, Japan 3Department of Pharmacology, School of Medicine, Showa University, Tokyo 142, Japan (Received 22 February 1992 ; accepted 29 May 1992)
Abstraet--Intrathecal administration of an adenosine receptor antagonist, theophylline, elicited nociceptive behavior such as licking, biting and scratching in mice. This behavioral response was dose-dependently reduced by simultaneous injection of an adenosine receptor agonist, 5'-N6-ethylcarboxamidoadenosine, or a selective N-methyl-D-aspartate (NMDA) receptor antagonist, n-2-amino-5-phosphonovalerate. This theophylline-induced behavior was not significantly reduced by the substance P (SP) analogue, the neurokinin receptor antagonist, [D-Arg~, o-Trp 7'9, Leull]SP (spantide). These results suggest the possibility that theophylline-induced nociceptive behavior may be mediated through interactions with both spinal adenosine- and NMDA receptors separately, or only through interaction(s) with adenosine receptors localized on the axon terminals of excitatory amino acid neurons. Present data have failed to reveal involvement of SP.
Recent studies suggest that adenosine may be a neurotransmitter involved in the regulation of nociception at the level of the spinal cord (see e.g. Delander and Hopkins, 1986; Sawynok et al., 1986). Adenosinecontaining neurons or nerve terminals appear to be localized in the superficial layers of the dorsal horn of the spinal cord, as indicated by the distribution of adenosine deaminase activity (Geiger and Nagy, 1986), adenosine-like immunoreactivity (Braas et al., 1986), and adenosine receptor binding (Choca et al., 1987). Intrathecal (i.t.) administration of adenosine agonists produces antinociception, but the antinocieeptive action, in contrast, is attenuated by treatment with adenosine antagonists such as theophylline or 8-phenyltheophylline (Sawynok et al., 1986; Delander and Hopkins, 1986). These results indicate that the antinociceptive effects are mediated by interactions with adenosine receptors. Excitatory amino acids (EAA), primary L-glutamate and aspartate, are considered to function as excitatory neurotransmitters in various regions of
mammalian central nervous system (CNS). Postsynaptic receptors of mammalian brain EAA were first classified into two subtypes, N-methyl-D-aspartate (NMDA) and non-NMDA. Recently these have been classified into ionotropic and metabotropic subtypes ; the former being subdivided into N M D A , kainate and quisqualate receptors, based on different sensitivities to these agonists (Davies et al., 1980; Watkins and Evans, 1981). The quisqualate subtype is now referred to as the ~t-amino-3-hydroxy-5-methyl4-isoxazolepropionate (AMPA) receptor (Monaghan et al., 1989). These receptors are also thought to be involved in many physiological phenomena, such as the processing of sensory information, through coordinate movement patterns, cognitive processes, learning, and memory (Watkins et al., 1990). Behavioral effects that indicate activation of the nociceptive sensory pathway are brought about by i.t. injection of N M D A (Aanosen and Wilcox, 1986), tachykinins (Hylden and Wilcox, 1981), somatostatin (Seybold et al., 1982), bombesin (O'Donohue et al., 1984), 5-hydroxytryptamine (Fasmer and Post, 1983) or muscarinic agonists (Raffa et al., 1987). During preliminary studies, we found evidence that i.t.
*Author to whom all correspondence should be addressed. 69
injected theophyllinc, an adenosine receptor antagonist, produced licking, biting and scratching behavior in mice. The main purpose of the present study was to claril), the m e c h a n i s m by which such nocicepti\e behavioral response was produced by thcophylline.
EXPERIMENTAL
()saka. Japan). All other chemicals used ~erc of Ihc highest grade commercially available. Stalislical analyses of Ihe dala ~crc pcflormcd b~ Ilk' method of LilchlicM and Wflcoxon and Dunnctt's !cst for maltiple comparisons after analysis el" wlriancc (/\NOVA) All values are expressed as mean 4_SEM.
R EStI,'fS
PROCEDURES
Adult male ddY mice weighing 22 26 g were used. Animals were housed in colony cages with food and water continuously available. Testing took place during the light perled era 12 12 h light dark cycle. The i.t. injection procedure was adapted from a method described by Hylden and Wilcox (1981). The lumbar puncture was made in unanesthetized mice at the L5-L6 intervertebral space with a 28 gauge needle connected to a 50 #1 Hamilton syringe. All drugs used were dissolved in sterile artificial cerebrospinal fluid (CSF), containing (g/[000 ml) : NaCI, 7.4: KCI, 0.19: MgCI2, 0.19 and CaCl2, 0.14. They were all administered in a total vol of 5/~1. Before the i.t. injection, the mice were adapted for 1 h to an individual plastic cage (22.0 x 15.0 x 12.5 cm) which served as the observation chamber. Immediately after the i.t. injeclion, the mice were placed in the transparent cage and the accumulated response time of scratching, biting, forepaw and hindpaw licking in each mouse was individually measured for 20 rain. The total duration (s) of these three responses measured in the first 20 min (I0 mice for each group) are expressed as "'behavioral response" in the present study. The effect of [D-Arg 1, D-Trp 79. Lcu~]SP (spantidc), D(--)-2amino-5-phosphonovalerate (D-APV) or 5'-N%ethylcarboxamidoadenosine tNECA) on these behavioral responses was determined after co-administration with various doses of theophylline (see the text). Drugs used in the present study were theophylline (Nakarai Tesquc, Inc., Kyoto, Japan), NECA (Sigma, St. Louis, Me, U.S.A.), D-APV (Cambridge Research Biochemicals, Cambridge, U.K.) and spantide (Peptide Institute, Inc.,
Behatriora[ response im/uced t~ i.t. adminislered th~,ophylline A d m i n i s t r a t i o n of theophylline into mice induced nociceptive behavioral response consisting of biting, licking a n d scratching. As shown in Fig. I(A), dosedependent increases in the total d u r a t i o n time of these responses were observed after i.t. a d m i n i s t r a t i o n of 0 . 1 2 5 1.0 n m o l / m o u s c theophylline. However, higher doses than those did not further increase the d u r a t i o n of such response. Thus, in further experiments, we usually used 0 . 2 5 1.0 nmol doses of theophyltinc. The behavioral effects peaked 10 20 rain after an injection and almost disappeared at 30 rain post-injection [Fig. I(B)]. As is also shown in this figure, the onset of the behavioral response was almost immediate for substance P (SP), that c o m p o u n d has been suggest to be a nociceptive neurotransmitter, whereas for thcophylline there was a time-lag of 80-150 s. Thus, the time-courses of this b e h a v i o r due to these two compounds were completely different [Fig. I(B)]. The difference was also found for behavioral responses induced by theophylline and SP. F o r example, theophylline-induced b e h a v i o r consisted p r e d o m i n a n t l y of licking and biting of hindlegs, tail a n d lower parts of
400 "2
g
B
350 -
100-
300-
"a
250
!
g 200 Ct 150
>
j /
--'O--Substance P C. t nmot -.O.--TheoPhylline 0.5 nrnol \~. -"a---TheophyUil~ 0.25 nmol xr.~ --O--TheophyUtne 0.125 nmol
75.
S 50u
loo
25.
50 0
nmo|)
Theophylline
0
i 0-5
5-110
' i 10-15 15 i_20 20-25
25-30
Time after injection (rnin)
Fig. 1. (A) Effects of various doses (0.125 40.0 nmol/mouse) of theophylline (TPH) administered i.t. into the mouse. The duration of licking, biting and scratching induced by TPH alone was determined during the first 20 rain after each injection. Total durations of these three responses during the first 20 rain are expressed as "Behavioral response" in this and following figs. Vertical bars show the SEM **P < 0.01 difference from CSF-controls. N = 10 for each group. (B) Time courses of the behavioral response induced by i.t. administration of TPH (0.25 1.0 nmol), SP (0.1 nmol) or CSF alone.
Theophylline-induced nociceptive response
When any of various doses of D-APV (0.125-2.0 nmol) was co-administered i.t. with theophylline, the behavioral response, as has been found for NECA, was significantly reduced dose-dependently (Fig. 3). The IDs0 values of o-APV for 0.25, 0.5 or 1.0 nmol doses of theophylline were 0.27 (0.17-0.43), 0.32 (0.19-0.54) and 0.74 (0.36-1.53) nmol, respectively. Spantide, a putative tachykinin antagonist, in doses ranging from 0.5 to 2.0 nmol failed to reduce the 0.5 nmol theophylline-induced response (Fig. 4).
the abdomen with a few episodes of scratching. In contrast, behavior induced by SP consisted predominantly of reciprocal hindlimb scratching directed towards the caudal part of the body, and a few episodes of biting or licking of hindlegs and lower parts of abdomen. Effects o/ N E C A , D-A P V and span tide on theophyllineinduced behavioral response
Administration of a single dose of the adenosine receptor agonist, NECA (0.25 16.0 pmol), the selective N M D A receptor antagonist, D-APV (0.125-2.0 nmol), or the tachykinin antagonist, spantide (0.5-2.0 nmol), per se failed to induce the behavioral responses that were induced by theophylline. However, when any of various doses of the adenosine recepor agonist, NECA (0.25-16.00 pmol), was co-administered i.t. with different doses of theophylline (0.25-1.00 nmol), NECA dose-dependently reduced the theophyllineinduced behavioral responses (Fig. 2). A potent inhibitor of the enzyme, 3',5'-cyclic nucleotide phosphodiesterase, papaverine (in doses ranging from 0.05 0.1 nmol/mouse, i.t.), also induced nociceptive behavior more extensively than theophylline. The doses of NECA, which selectively reduced nociceptive behavior induced by theophylline, however, could not block this papaverine-induced behavior to any appreciable extent (data not shown). The IDs0 values (doses that reduced the theophylline-induced response 50%) of NECA for 0.25, 0.5 or 1.0 nmol doses of theophylline were estimated to be 1.9 (0.96-3.76, confidence limit), 2.1 (0.37-11.64) and 2.5 (0.85-7.35) pmol, respectively. 350-
~ 0
200.
ca
150"
@
.~_ r@ 03
To our knowledge, this is the first demonstration that i.t. injected theophylline induced nociceptive behavior consisting of licking, biting and scratching in mice. This clearly indicates that theophylline may be an effective nociceptive agent. This behavior was selectively blocked by the adenosine receptor agonist, NECA, or the N M D A receptor antagonist, D-APV. The NECA was a more potent blocker than D-APV, as judged from the IDs0 values against the same doses of theophylline. This selective blockade may indicate the participations of both spinal adenosine- and N M D A receptors in this behavior. At least two classes of membrane adenosine receptors have been defined in the CNS, even so, subdivisions are possible. Adenosine receptor agonists such as NECA blocked the SP- or NMDA-induced nociceptive effect (Sawynok et al., 1986) and this blockade was reduced by the additional treatment with theophylline (Delander and Wahl, 1988). These findings thus clearly indicate possible antagonistic
+
250.
~
C 0
DISCUSSION
+
300u
71
100. 5o. o
TPH 0.2,5 +
CSF
~
0,25
0.25 1.0
NECA
4.0
~
+
0.6
CSF
TI~
0.5
TPH 1.0
TPH
+
0.25
1.0
NECA
4.0
CSF
1.0
4.0 NECA
i 1.0
(nmol)
16.0 (pmol)
Fig. 2. Effects of various doses of 5'-N6-ethylcarboxamidoadenosine (NECA) on theophylline (TPH)induced licking, biting and scratching in mice. NECA was co-administered i.t. with TPH. The licking, biting and scratching induced by TPH started within 20 min after TPH alone or the combined injection. *P < 0.05, **P < 0.01 difference from TPH alone. N = 10 for each group.
72
] ]II)EY(> NA(iA()KA
+
350-
II II
300¢n
o~
25o200-
:. • >>:,:
150-
i i il i!iil
10050111
O"
TPH 0.25
TPH
0.25
TPH 0.5
+
TPH
0.5
TPH 1.0
+
CSF
0.125
0.25 D-APV
0.5
CSF
TPH
1.0
(nmol)
+
0.125
0.25 D-APV
0.5
CSF
0.5
1.0 D-APV
2.0 (n,mol)
Fig. 3. Effects of various doses of I)-( - )-2-amino-5-phosphonovalerate (D-APV) on theophylline (TPH)induced licking, biting and scratching in mice. D-APV at each dose was co-administered i.t. with TPH. The duration of induced licking, biting and scratching was determined daring the first 20 rain after injection. **P < 0.01 difference from TPH (0.5 nmol) alone. N = 10 for each group. interaction(s) of the latter two compounds at some spinal site(s). In addition, morphine enhanced adenosine release, and methylxanthine adenosine antagonists inhibited morphine-induced analgesia (Delander and Hopkins, 1986). These findings conclusively indicate the possible involvement of adenosine receptors in the modulation of spinal pain transmission. Taken together, the antagonism by NECA
350
~ 31111 aso
c o
I|
II
150
o
"~ .C
@ m
50 O
TPt.I 0.5
I"PH 0.5
0~5 1.0
(nmo~ 2 . 0 (nmoi)
Fig. 4. Effects of various doses of spantide on theophylline(TPH)-induced licking, biting and scratching in mice. Spantide was co-administered i.t. with TPH. The duration of licking, biting and scratching was determined during the first 20 rnin after TPH alone or the combined injection. N 10 for each group.
of the theophylline-induced nociceptive behavior is thus consistent with the hypothesis that theophylline may induce such nociception, at least in part, by direct interaction(s) with adenosine receptors. The indirect action of theophylline, due to the inhibition of Y,5'cyclic nucleotide phosphodiesterase with a concomitant cyclic A M P increase, could not be explained as a main nociceptive action, since papaverine, the potent inhibitor of this enzyme, also induced nociceptive behavior, but this behavior could not be blocked by the doses of NECA used, which significantly blocked theophylline-induced behavior. Despite the possible involvement of adenosine-A: receptors in agonist-induced spinal antinociception (Sawynok el a/., 1986 ; Delander and Hopkins, 1987), it is not clear at the moment at which subtype of the receptors the()phylline may selectively interact for its nociccptive effect. This is obscure because NECA is a non-selective agonist, and theophylline is a non-selective antagonist of membrane adenosine receptors (Daly el al.. 1985). As reported earlier (Sakurada el al., 1989), SP induced nociceptive behavior, that consisted mainly of licking and biting with a time course different from that due to theophylline. In addition, this SP-induced behavior was blocked by spantide and by several antagonistic SP analogues. Spantide also blocked the nociceptive behavior induced by the neurokinin-2 agonists, eledoisin and neurokinin A (Sakurada et al., 1989). Thus, spantide may not be a selective antagonist for neurokinin-I receptors, which are mainly responsible for mediating SP-induced nociccptive behavior. Despite the non-selectivity of spantidc
73
Theophylline-induced nociceptive response towards spinal neurokinin receptors, at the doses used in the present study it could not significantly block theophylline-induced behavior. In addition to simultaneous injection of spantide, its injection 20 rain before or after theophylline caused no antagonistic effect, excluding the possibility of a difference in time of onset of its antagonistic effect towards theophylline. Thus, the absence of blockade by spantide, and the different quality and time courses of behavior induced by SP and theophylline may suggest that theophylline-induced behavior is not mediated via neurokinin receptors. It was previously reported that EAA may be involved in spinal pain transmission (Aanosen and Wilcox, 1986 ; Sakurada et al., 1990). To support this, i.t. injection of N M D A induced nociceptive behavior (Aanosen and Wilcox, 1986; Hylden and Wilcox, 1981 ; Sakurada et al., 1990), which resembled that induced by SP, implying two different spinal nociceprive activating pathways that utilize these respectively as neurotransmitters (Aanosen and Wilcox, 1986). This NMDA-induced behavior was selectively blocked by D-APV, but not by selective neurokinin-1 receptor antagonists (Sakurada et al., 1990), which supports the hypothesis that N M D A may mainly induce nociception at N M D A receptors rather than at neurokinin receptors (Aanosen and Wilcox, 1986). As found in the present study, this selective antagonism by D-APV was also found for theophylline. This blockade may thus indicate a possible interaction(s) of theopylline with the spinal N M D A receptors, in addition to adenosine receptors. The time course and onset of NMDA-induced behavior, however, differed from those induced by theophylline, but were similar to those by SP which had rapid onset and subsidence (Sakurada et al., 1990). These may indicate different mechanisms of action by theophylline for causing nociceptive behavior, compared to that by SP or N M D A . The minor role of the N M D A receptors in theophylline-induced behavior may be supported by evidence of lower ICs0 values of N E C A than of DAPV, if intrinsic potencies of both antagonists are ignored. Both adenosine and EAA must be involved in interneuronal communication, because of the predominant localization of adenosine receptors on axon terminals of EAA neurons (for review, see Daval et al., 1991). Considering this evidence and the finding of N M D A receptors in the spinal cords (Aanonsen et al., 1991), it seems reasonable to hypothesize that, for inducing nociception, theophylline may mainly interact with presynaptic adenosine receptors on the axon terminals of EAA neurons and indirectly facili-
tate activity of postsynaptic spinal N M D A receptors. If this hypothesis is correct, the indirect action of theophylline may be blocked by the selective N M D A antagonist, D-APV, as found in the present study. The present data, however, have failed to reveal involvement of SP. There are possible interactions between SP and EAA in the spinal cord (Battaglia and Rustioni, 1988), so further studies on involvement of SP are needed to clarify this point.
REFERENCES
Aanosen L. M. and Wilcox G. L. (1986) Phencyclidine selectively blocks a spinal action of N-methyl-D-aspartate in mice. Neurosci. Lett. 67, 191--197. Aanosen L. M., Lei S. and Wilcox G. L. (1991) Excitatory amino acid receptors and nociceptive neurotransmission in rat spinal cord. Pain 41,309 321. Battaglia G. and Rustioni A. (1988) Coexistence of glutamate and substance P in dorsal root ganglion neurons of the rat and monkey. J comp. Neurol. 277, 302-312. Braas K. M., Newby A. C., Wilson V. S. and Snyder S. H. (1986) Adenosine-containing neurons in the brain localized by immunocytochemistry. J. Neurosci. 6, 1952 1961. Choca J. I., Proudfit H. K. and Green R. D. (1987) Identification of A~ and A2 adenosine receptors in the rat spinal cord. J. Pharmac. exp. Ther. 242, 905 910. Daly J. W., Padgett W., Shamim M. T., Butts-Lamb P. and Waters J. (1985) 1,3-Dialkyl-8-(p-sulfophenyl)xanthines; potent water-soluble antagonists for A~ and A2-adenosine receptors. J. Med. Chem. 28, 487-492. Daval J-L., Nehlig A. and Nicolas F. (1991) Physiological and pharmacological properties of adenosine : therapeutic implications. L(fe Sci. 49, 1435-1453. Davis J., Fransis A. A., Jones A. W. and Watkins J. C. (1980) 2-Amino-5-phosphovalerate (2-APV), a highly potent and specific antagonist at spinal NMDA receptors. Br. J. Pharmac. 70, 52 53. Delander G. E. and Hopkins C. J. (1986) Spinal adenosine modulates descending antinociceptive pathways stimulated by morphine. J. Pharmac. exp. Ther. 239, 88-93. Delander G. E. and Hopkins C. J. (1987) Involvement of A2 adenosine receptors in spinal mechanisms of antinociception. Eur. J. Pharmac. 139, 215 273. Delander G. E. and Wahl J. J. (1988) Behavior induced by putative nociceptive neurotransmitters is inhibited by adenosine or adenosine analogs coadministered intrathecally. J. Pharmac. exp. Ther. 246, 565 570. Fasmer O. B. and Post C. (1983) Behavioral responses induced by intrathecal injection of 5-hydroxytryptamine in mice are inhibited by a substance P antagonist, D-Pro 2, DTrpT9-substance P. Neuropharmacology 22, 1397 1400. Geiger J. D. and Nagy J. I. (1986) Distribution of adnosine deaminase activity in rat brain and spinal cord. J. Neurosci. 6, 2707 2714. Hylden J. L. K. and Wilcox G. L. (1981) Intrathecal substance P elicits a caudally-directed biting and scratching behavior in mice. Brain Res. 217, 212 215. Monaghan D. T., Bridges R. J. and Cotman C. W. (1989) The excitatory amino acid receptors : their classes, pharmacology, and distinct properties in the function of the
74
HII)~¢~o NAGAOKA
central nervous system. A. Rec. t'harnla~. Toxico/. 29, 365 402. O ' D o n o h u c T. L., Massari V. J.. Pazoles C J.. Chronwall B. M., Shults C. W., Quirion R., Chase T. N. and Moody T. W. (1984) A role for bombesin in sensory processing in the spinal cord. J. Neurosci. 4, 2956 2962. Raffa R. B., Mathiasen J. R., Goode T. L. and Vaught J. L. (1987) Evidence that reciprocal hindlimb scratching elicited in mice by intrathecal administration ofmuscarinic agonists is mediated through M 1 type reccptors. L[/i, Sci. 41,1831 I836. Sakurada T., Yamada T., Sakurada S., Kisara K. and Ohba M. (1989) Substance P analogues containing histidine antagonize thc behavioural effects of intrathecally coadminis|ered substance P in mice. Eur. J. Pharnlac. 174, 153 160. Sakurada T., M a n o m e Y., Tan-No K., Sakurada S. and
K isara K. 11990) Thc efli;cts of substance P anaiogncs on the scratching, biting and licking response reduced by intrathecal injection of N-methyl-l)-aspartate in mice. Br. ,I. Pharmac. !01,307 310. Sawynok J, Sweeney M. 1. and White T. D. (198t~) Classification of adenosine receptors mediating anlinociception in the rat spinal cord. Br..I. Pharmac. 88, 923 930. Seybold V. S., Hylden J. L. K. and Wilcox G. L. 11982) Intrathecal substance P and somatostatin m rats: behaviors indicative of sensation. Peptides 3, 49 54. Watkins J. C and Evans R. I t. ( 1981 ) Excitatory amino acid transmitters. ,4. Rct'. Pharmac. 7)~xicoL 21, 165 204. Watkins J. C , Krogsgaard-Larsen P. and Honore T. (19901 Structure activity relationships in the developlnent o1" excitatory amino acid receptor agonists and competitive antagonists, frend~, P/l~:;'ntac. 5;ci. 11, 25 33.
0197-0186/93 $5.00+0.00 Pergamon Press Ltd
Printed in Great Britain
THEOPHYLLINE-INDUCED NOCICEPTIVE BEHAVIORAL RESPONSE IN MICE: POSSIBLE INDIRECT INTERACTION WITH SPINAL N-METHYL-D-ASPARTATE RECEPTORS HIDEYO NAGAOKA,1 SHINOBU SAKURADA,2. TSUKASA SAKURADA,2 SATOSHI TAKEDA, 2 YOSHITO NAKAGAWA,l KENSUKE KISARA2 a n d YUICHIRO ARAI3 ~Department of Pharmacy, Yamagata University Hospital, lida-Nishi Yamagata 990, Japan 2Department of Pharmacology, Tohoku College of Pharmacy, Komatsushima, Aoba-ku, Sendai 981, Japan 3Department of Pharmacology, School of Medicine, Showa University, Tokyo 142, Japan (Received 22 February 1992 ; accepted 29 May 1992)
Abstraet--Intrathecal administration of an adenosine receptor antagonist, theophylline, elicited nociceptive behavior such as licking, biting and scratching in mice. This behavioral response was dose-dependently reduced by simultaneous injection of an adenosine receptor agonist, 5'-N6-ethylcarboxamidoadenosine, or a selective N-methyl-D-aspartate (NMDA) receptor antagonist, n-2-amino-5-phosphonovalerate. This theophylline-induced behavior was not significantly reduced by the substance P (SP) analogue, the neurokinin receptor antagonist, [D-Arg~, o-Trp 7'9, Leull]SP (spantide). These results suggest the possibility that theophylline-induced nociceptive behavior may be mediated through interactions with both spinal adenosine- and NMDA receptors separately, or only through interaction(s) with adenosine receptors localized on the axon terminals of excitatory amino acid neurons. Present data have failed to reveal involvement of SP.
Recent studies suggest that adenosine may be a neurotransmitter involved in the regulation of nociception at the level of the spinal cord (see e.g. Delander and Hopkins, 1986; Sawynok et al., 1986). Adenosinecontaining neurons or nerve terminals appear to be localized in the superficial layers of the dorsal horn of the spinal cord, as indicated by the distribution of adenosine deaminase activity (Geiger and Nagy, 1986), adenosine-like immunoreactivity (Braas et al., 1986), and adenosine receptor binding (Choca et al., 1987). Intrathecal (i.t.) administration of adenosine agonists produces antinociception, but the antinocieeptive action, in contrast, is attenuated by treatment with adenosine antagonists such as theophylline or 8-phenyltheophylline (Sawynok et al., 1986; Delander and Hopkins, 1986). These results indicate that the antinociceptive effects are mediated by interactions with adenosine receptors. Excitatory amino acids (EAA), primary L-glutamate and aspartate, are considered to function as excitatory neurotransmitters in various regions of
mammalian central nervous system (CNS). Postsynaptic receptors of mammalian brain EAA were first classified into two subtypes, N-methyl-D-aspartate (NMDA) and non-NMDA. Recently these have been classified into ionotropic and metabotropic subtypes ; the former being subdivided into N M D A , kainate and quisqualate receptors, based on different sensitivities to these agonists (Davies et al., 1980; Watkins and Evans, 1981). The quisqualate subtype is now referred to as the ~t-amino-3-hydroxy-5-methyl4-isoxazolepropionate (AMPA) receptor (Monaghan et al., 1989). These receptors are also thought to be involved in many physiological phenomena, such as the processing of sensory information, through coordinate movement patterns, cognitive processes, learning, and memory (Watkins et al., 1990). Behavioral effects that indicate activation of the nociceptive sensory pathway are brought about by i.t. injection of N M D A (Aanosen and Wilcox, 1986), tachykinins (Hylden and Wilcox, 1981), somatostatin (Seybold et al., 1982), bombesin (O'Donohue et al., 1984), 5-hydroxytryptamine (Fasmer and Post, 1983) or muscarinic agonists (Raffa et al., 1987). During preliminary studies, we found evidence that i.t.
*Author to whom all correspondence should be addressed. 69
injected theophyllinc, an adenosine receptor antagonist, produced licking, biting and scratching behavior in mice. The main purpose of the present study was to claril), the m e c h a n i s m by which such nocicepti\e behavioral response was produced by thcophylline.
EXPERIMENTAL
()saka. Japan). All other chemicals used ~erc of Ihc highest grade commercially available. Stalislical analyses of Ihe dala ~crc pcflormcd b~ Ilk' method of LilchlicM and Wflcoxon and Dunnctt's !cst for maltiple comparisons after analysis el" wlriancc (/\NOVA) All values are expressed as mean 4_SEM.
R EStI,'fS
PROCEDURES
Adult male ddY mice weighing 22 26 g were used. Animals were housed in colony cages with food and water continuously available. Testing took place during the light perled era 12 12 h light dark cycle. The i.t. injection procedure was adapted from a method described by Hylden and Wilcox (1981). The lumbar puncture was made in unanesthetized mice at the L5-L6 intervertebral space with a 28 gauge needle connected to a 50 #1 Hamilton syringe. All drugs used were dissolved in sterile artificial cerebrospinal fluid (CSF), containing (g/[000 ml) : NaCI, 7.4: KCI, 0.19: MgCI2, 0.19 and CaCl2, 0.14. They were all administered in a total vol of 5/~1. Before the i.t. injection, the mice were adapted for 1 h to an individual plastic cage (22.0 x 15.0 x 12.5 cm) which served as the observation chamber. Immediately after the i.t. injeclion, the mice were placed in the transparent cage and the accumulated response time of scratching, biting, forepaw and hindpaw licking in each mouse was individually measured for 20 rain. The total duration (s) of these three responses measured in the first 20 min (I0 mice for each group) are expressed as "'behavioral response" in the present study. The effect of [D-Arg 1, D-Trp 79. Lcu~]SP (spantidc), D(--)-2amino-5-phosphonovalerate (D-APV) or 5'-N%ethylcarboxamidoadenosine tNECA) on these behavioral responses was determined after co-administration with various doses of theophylline (see the text). Drugs used in the present study were theophylline (Nakarai Tesquc, Inc., Kyoto, Japan), NECA (Sigma, St. Louis, Me, U.S.A.), D-APV (Cambridge Research Biochemicals, Cambridge, U.K.) and spantide (Peptide Institute, Inc.,
Behatriora[ response im/uced t~ i.t. adminislered th~,ophylline A d m i n i s t r a t i o n of theophylline into mice induced nociceptive behavioral response consisting of biting, licking a n d scratching. As shown in Fig. I(A), dosedependent increases in the total d u r a t i o n time of these responses were observed after i.t. a d m i n i s t r a t i o n of 0 . 1 2 5 1.0 n m o l / m o u s c theophylline. However, higher doses than those did not further increase the d u r a t i o n of such response. Thus, in further experiments, we usually used 0 . 2 5 1.0 nmol doses of theophyltinc. The behavioral effects peaked 10 20 rain after an injection and almost disappeared at 30 rain post-injection [Fig. I(B)]. As is also shown in this figure, the onset of the behavioral response was almost immediate for substance P (SP), that c o m p o u n d has been suggest to be a nociceptive neurotransmitter, whereas for thcophylline there was a time-lag of 80-150 s. Thus, the time-courses of this b e h a v i o r due to these two compounds were completely different [Fig. I(B)]. The difference was also found for behavioral responses induced by theophylline and SP. F o r example, theophylline-induced b e h a v i o r consisted p r e d o m i n a n t l y of licking and biting of hindlegs, tail a n d lower parts of
400 "2
g
B
350 -
100-
300-
"a
250
!
g 200 Ct 150
>
j /
--'O--Substance P C. t nmot -.O.--TheoPhylline 0.5 nrnol \~. -"a---TheophyUil~ 0.25 nmol xr.~ --O--TheophyUtne 0.125 nmol
75.
S 50u
loo
25.
50 0
nmo|)
Theophylline
0
i 0-5
5-110
' i 10-15 15 i_20 20-25
25-30
Time after injection (rnin)
Fig. 1. (A) Effects of various doses (0.125 40.0 nmol/mouse) of theophylline (TPH) administered i.t. into the mouse. The duration of licking, biting and scratching induced by TPH alone was determined during the first 20 rain after each injection. Total durations of these three responses during the first 20 rain are expressed as "Behavioral response" in this and following figs. Vertical bars show the SEM **P < 0.01 difference from CSF-controls. N = 10 for each group. (B) Time courses of the behavioral response induced by i.t. administration of TPH (0.25 1.0 nmol), SP (0.1 nmol) or CSF alone.
Theophylline-induced nociceptive response
When any of various doses of D-APV (0.125-2.0 nmol) was co-administered i.t. with theophylline, the behavioral response, as has been found for NECA, was significantly reduced dose-dependently (Fig. 3). The IDs0 values of o-APV for 0.25, 0.5 or 1.0 nmol doses of theophylline were 0.27 (0.17-0.43), 0.32 (0.19-0.54) and 0.74 (0.36-1.53) nmol, respectively. Spantide, a putative tachykinin antagonist, in doses ranging from 0.5 to 2.0 nmol failed to reduce the 0.5 nmol theophylline-induced response (Fig. 4).
the abdomen with a few episodes of scratching. In contrast, behavior induced by SP consisted predominantly of reciprocal hindlimb scratching directed towards the caudal part of the body, and a few episodes of biting or licking of hindlegs and lower parts of abdomen. Effects o/ N E C A , D-A P V and span tide on theophyllineinduced behavioral response
Administration of a single dose of the adenosine receptor agonist, NECA (0.25 16.0 pmol), the selective N M D A receptor antagonist, D-APV (0.125-2.0 nmol), or the tachykinin antagonist, spantide (0.5-2.0 nmol), per se failed to induce the behavioral responses that were induced by theophylline. However, when any of various doses of the adenosine recepor agonist, NECA (0.25-16.00 pmol), was co-administered i.t. with different doses of theophylline (0.25-1.00 nmol), NECA dose-dependently reduced the theophyllineinduced behavioral responses (Fig. 2). A potent inhibitor of the enzyme, 3',5'-cyclic nucleotide phosphodiesterase, papaverine (in doses ranging from 0.05 0.1 nmol/mouse, i.t.), also induced nociceptive behavior more extensively than theophylline. The doses of NECA, which selectively reduced nociceptive behavior induced by theophylline, however, could not block this papaverine-induced behavior to any appreciable extent (data not shown). The IDs0 values (doses that reduced the theophylline-induced response 50%) of NECA for 0.25, 0.5 or 1.0 nmol doses of theophylline were estimated to be 1.9 (0.96-3.76, confidence limit), 2.1 (0.37-11.64) and 2.5 (0.85-7.35) pmol, respectively. 350-
~ 0
200.
ca
150"
@
.~_ r@ 03
To our knowledge, this is the first demonstration that i.t. injected theophylline induced nociceptive behavior consisting of licking, biting and scratching in mice. This clearly indicates that theophylline may be an effective nociceptive agent. This behavior was selectively blocked by the adenosine receptor agonist, NECA, or the N M D A receptor antagonist, D-APV. The NECA was a more potent blocker than D-APV, as judged from the IDs0 values against the same doses of theophylline. This selective blockade may indicate the participations of both spinal adenosine- and N M D A receptors in this behavior. At least two classes of membrane adenosine receptors have been defined in the CNS, even so, subdivisions are possible. Adenosine receptor agonists such as NECA blocked the SP- or NMDA-induced nociceptive effect (Sawynok et al., 1986) and this blockade was reduced by the additional treatment with theophylline (Delander and Wahl, 1988). These findings thus clearly indicate possible antagonistic
+
250.
~
C 0
DISCUSSION
+
300u
71
100. 5o. o
TPH 0.2,5 +
CSF
~
0,25
0.25 1.0
NECA
4.0
~
+
0.6
CSF
TI~
0.5
TPH 1.0
TPH
+
0.25
1.0
NECA
4.0
CSF
1.0
4.0 NECA
i 1.0
(nmol)
16.0 (pmol)
Fig. 2. Effects of various doses of 5'-N6-ethylcarboxamidoadenosine (NECA) on theophylline (TPH)induced licking, biting and scratching in mice. NECA was co-administered i.t. with TPH. The licking, biting and scratching induced by TPH started within 20 min after TPH alone or the combined injection. *P < 0.05, **P < 0.01 difference from TPH alone. N = 10 for each group.
72
] ]II)EY(> NA(iA()KA
+
350-
II II
300¢n
o~
25o200-
:. • >>:,:
150-
i i il i!iil
10050111
O"
TPH 0.25
TPH
0.25
TPH 0.5
+
TPH
0.5
TPH 1.0
+
CSF
0.125
0.25 D-APV
0.5
CSF
TPH
1.0
(nmol)
+
0.125
0.25 D-APV
0.5
CSF
0.5
1.0 D-APV
2.0 (n,mol)
Fig. 3. Effects of various doses of I)-( - )-2-amino-5-phosphonovalerate (D-APV) on theophylline (TPH)induced licking, biting and scratching in mice. D-APV at each dose was co-administered i.t. with TPH. The duration of induced licking, biting and scratching was determined daring the first 20 rain after injection. **P < 0.01 difference from TPH (0.5 nmol) alone. N = 10 for each group. interaction(s) of the latter two compounds at some spinal site(s). In addition, morphine enhanced adenosine release, and methylxanthine adenosine antagonists inhibited morphine-induced analgesia (Delander and Hopkins, 1986). These findings conclusively indicate the possible involvement of adenosine receptors in the modulation of spinal pain transmission. Taken together, the antagonism by NECA
350
~ 31111 aso
c o
I|
II
150
o
"~ .C
@ m
50 O
TPt.I 0.5
I"PH 0.5
0~5 1.0
(nmo~ 2 . 0 (nmoi)
Fig. 4. Effects of various doses of spantide on theophylline(TPH)-induced licking, biting and scratching in mice. Spantide was co-administered i.t. with TPH. The duration of licking, biting and scratching was determined during the first 20 rnin after TPH alone or the combined injection. N 10 for each group.
of the theophylline-induced nociceptive behavior is thus consistent with the hypothesis that theophylline may induce such nociception, at least in part, by direct interaction(s) with adenosine receptors. The indirect action of theophylline, due to the inhibition of Y,5'cyclic nucleotide phosphodiesterase with a concomitant cyclic A M P increase, could not be explained as a main nociceptive action, since papaverine, the potent inhibitor of this enzyme, also induced nociceptive behavior, but this behavior could not be blocked by the doses of NECA used, which significantly blocked theophylline-induced behavior. Despite the possible involvement of adenosine-A: receptors in agonist-induced spinal antinociception (Sawynok el a/., 1986 ; Delander and Hopkins, 1987), it is not clear at the moment at which subtype of the receptors the()phylline may selectively interact for its nociccptive effect. This is obscure because NECA is a non-selective agonist, and theophylline is a non-selective antagonist of membrane adenosine receptors (Daly el al.. 1985). As reported earlier (Sakurada el al., 1989), SP induced nociceptive behavior, that consisted mainly of licking and biting with a time course different from that due to theophylline. In addition, this SP-induced behavior was blocked by spantide and by several antagonistic SP analogues. Spantide also blocked the nociceptive behavior induced by the neurokinin-2 agonists, eledoisin and neurokinin A (Sakurada et al., 1989). Thus, spantide may not be a selective antagonist for neurokinin-I receptors, which are mainly responsible for mediating SP-induced nociccptive behavior. Despite the non-selectivity of spantidc
73
Theophylline-induced nociceptive response towards spinal neurokinin receptors, at the doses used in the present study it could not significantly block theophylline-induced behavior. In addition to simultaneous injection of spantide, its injection 20 rain before or after theophylline caused no antagonistic effect, excluding the possibility of a difference in time of onset of its antagonistic effect towards theophylline. Thus, the absence of blockade by spantide, and the different quality and time courses of behavior induced by SP and theophylline may suggest that theophylline-induced behavior is not mediated via neurokinin receptors. It was previously reported that EAA may be involved in spinal pain transmission (Aanosen and Wilcox, 1986 ; Sakurada et al., 1990). To support this, i.t. injection of N M D A induced nociceptive behavior (Aanosen and Wilcox, 1986; Hylden and Wilcox, 1981 ; Sakurada et al., 1990), which resembled that induced by SP, implying two different spinal nociceprive activating pathways that utilize these respectively as neurotransmitters (Aanosen and Wilcox, 1986). This NMDA-induced behavior was selectively blocked by D-APV, but not by selective neurokinin-1 receptor antagonists (Sakurada et al., 1990), which supports the hypothesis that N M D A may mainly induce nociception at N M D A receptors rather than at neurokinin receptors (Aanosen and Wilcox, 1986). As found in the present study, this selective antagonism by D-APV was also found for theophylline. This blockade may thus indicate a possible interaction(s) of theopylline with the spinal N M D A receptors, in addition to adenosine receptors. The time course and onset of NMDA-induced behavior, however, differed from those induced by theophylline, but were similar to those by SP which had rapid onset and subsidence (Sakurada et al., 1990). These may indicate different mechanisms of action by theophylline for causing nociceptive behavior, compared to that by SP or N M D A . The minor role of the N M D A receptors in theophylline-induced behavior may be supported by evidence of lower ICs0 values of N E C A than of DAPV, if intrinsic potencies of both antagonists are ignored. Both adenosine and EAA must be involved in interneuronal communication, because of the predominant localization of adenosine receptors on axon terminals of EAA neurons (for review, see Daval et al., 1991). Considering this evidence and the finding of N M D A receptors in the spinal cords (Aanonsen et al., 1991), it seems reasonable to hypothesize that, for inducing nociception, theophylline may mainly interact with presynaptic adenosine receptors on the axon terminals of EAA neurons and indirectly facili-
tate activity of postsynaptic spinal N M D A receptors. If this hypothesis is correct, the indirect action of theophylline may be blocked by the selective N M D A antagonist, D-APV, as found in the present study. The present data, however, have failed to reveal involvement of SP. There are possible interactions between SP and EAA in the spinal cord (Battaglia and Rustioni, 1988), so further studies on involvement of SP are needed to clarify this point.
REFERENCES
Aanosen L. M. and Wilcox G. L. (1986) Phencyclidine selectively blocks a spinal action of N-methyl-D-aspartate in mice. Neurosci. Lett. 67, 191--197. Aanosen L. M., Lei S. and Wilcox G. L. (1991) Excitatory amino acid receptors and nociceptive neurotransmission in rat spinal cord. Pain 41,309 321. Battaglia G. and Rustioni A. (1988) Coexistence of glutamate and substance P in dorsal root ganglion neurons of the rat and monkey. J comp. Neurol. 277, 302-312. Braas K. M., Newby A. C., Wilson V. S. and Snyder S. H. (1986) Adenosine-containing neurons in the brain localized by immunocytochemistry. J. Neurosci. 6, 1952 1961. Choca J. I., Proudfit H. K. and Green R. D. (1987) Identification of A~ and A2 adenosine receptors in the rat spinal cord. J. Pharmac. exp. Ther. 242, 905 910. Daly J. W., Padgett W., Shamim M. T., Butts-Lamb P. and Waters J. (1985) 1,3-Dialkyl-8-(p-sulfophenyl)xanthines; potent water-soluble antagonists for A~ and A2-adenosine receptors. J. Med. Chem. 28, 487-492. Daval J-L., Nehlig A. and Nicolas F. (1991) Physiological and pharmacological properties of adenosine : therapeutic implications. L(fe Sci. 49, 1435-1453. Davis J., Fransis A. A., Jones A. W. and Watkins J. C. (1980) 2-Amino-5-phosphovalerate (2-APV), a highly potent and specific antagonist at spinal NMDA receptors. Br. J. Pharmac. 70, 52 53. Delander G. E. and Hopkins C. J. (1986) Spinal adenosine modulates descending antinociceptive pathways stimulated by morphine. J. Pharmac. exp. Ther. 239, 88-93. Delander G. E. and Hopkins C. J. (1987) Involvement of A2 adenosine receptors in spinal mechanisms of antinociception. Eur. J. Pharmac. 139, 215 273. Delander G. E. and Wahl J. J. (1988) Behavior induced by putative nociceptive neurotransmitters is inhibited by adenosine or adenosine analogs coadministered intrathecally. J. Pharmac. exp. Ther. 246, 565 570. Fasmer O. B. and Post C. (1983) Behavioral responses induced by intrathecal injection of 5-hydroxytryptamine in mice are inhibited by a substance P antagonist, D-Pro 2, DTrpT9-substance P. Neuropharmacology 22, 1397 1400. Geiger J. D. and Nagy J. I. (1986) Distribution of adnosine deaminase activity in rat brain and spinal cord. J. Neurosci. 6, 2707 2714. Hylden J. L. K. and Wilcox G. L. (1981) Intrathecal substance P elicits a caudally-directed biting and scratching behavior in mice. Brain Res. 217, 212 215. Monaghan D. T., Bridges R. J. and Cotman C. W. (1989) The excitatory amino acid receptors : their classes, pharmacology, and distinct properties in the function of the
74
HII)~¢~o NAGAOKA
central nervous system. A. Rec. t'harnla~. Toxico/. 29, 365 402. O ' D o n o h u c T. L., Massari V. J.. Pazoles C J.. Chronwall B. M., Shults C. W., Quirion R., Chase T. N. and Moody T. W. (1984) A role for bombesin in sensory processing in the spinal cord. J. Neurosci. 4, 2956 2962. Raffa R. B., Mathiasen J. R., Goode T. L. and Vaught J. L. (1987) Evidence that reciprocal hindlimb scratching elicited in mice by intrathecal administration ofmuscarinic agonists is mediated through M 1 type reccptors. L[/i, Sci. 41,1831 I836. Sakurada T., Yamada T., Sakurada S., Kisara K. and Ohba M. (1989) Substance P analogues containing histidine antagonize thc behavioural effects of intrathecally coadminis|ered substance P in mice. Eur. J. Pharnlac. 174, 153 160. Sakurada T., M a n o m e Y., Tan-No K., Sakurada S. and
K isara K. 11990) Thc efli;cts of substance P anaiogncs on the scratching, biting and licking response reduced by intrathecal injection of N-methyl-l)-aspartate in mice. Br. ,I. Pharmac. !01,307 310. Sawynok J, Sweeney M. 1. and White T. D. (198t~) Classification of adenosine receptors mediating anlinociception in the rat spinal cord. Br..I. Pharmac. 88, 923 930. Seybold V. S., Hylden J. L. K. and Wilcox G. L. 11982) Intrathecal substance P and somatostatin m rats: behaviors indicative of sensation. Peptides 3, 49 54. Watkins J. C and Evans R. I t. ( 1981 ) Excitatory amino acid transmitters. ,4. Rct'. Pharmac. 7)~xicoL 21, 165 204. Watkins J. C , Krogsgaard-Larsen P. and Honore T. (19901 Structure activity relationships in the developlnent o1" excitatory amino acid receptor agonists and competitive antagonists, frend~, P/l~:;'ntac. 5;ci. 11, 25 33.