Fine primary afferent fibers, which innervate nociceptors, terminate preferentially on neurons of the superficial spinal dorsal horn, particularly on substantia gelatinosa (SG; lamina II of Rexed) neurons [1,2]. SG neurons in the rat spinal dorsal horn play important roles in modulation of nociceptive information . Miniature excitatory postsynaptic currents (mEPSCs) and miniature inhibitory postsynaptic currents (mIPSCs) in SG are thought to be important for this modulation . It has been demonstrated that both spontaneous and noxious stimuli-evoked mEPSCs are mainly mediated by presynaptic glutamate release [3–6]. To our knowledge, however, little has been known about the origins of mEPSCs under physiological (normal) and noxiously stimulated situations. This led us to investigate the origins of spontaneous and noxious stimuli-evoked mEPSCs of the SG neurons in the spinal cord slice.
Capsaicin (vanilloid), a natural ingredient in chili peppers of the Capsicum family, can excite a subtype of primary sensory neurons or their terminals that are related to nociception, eliciting the sensation of burning pain [7,8]. The capsaicin receptor is a non-selective cation channel expressed on fine primary sensory neurons and their terminals . The confined distribution and selectivity of capsaicin for nociceptive primary afferent make it to be an important, if not indispensable, tool in the study of pain [9,10]. In the present work, we used capsaicin as the noxious stimuli for the fine primary afferent fibers.
MATERIALS AND METHODS
All protocols were approved by the Animal Care and Use Committee at The Fourth Military Medical University (Xi'an, China). The methods for slice preparation of adult rat spinal cord have been described elsewhere . Briefly, adult Sprague–Dawley rats (7–8 weeks old, weighing 250–280 g) were anaesthetized with urethane (1.5 g/kg, i.p.). The lumbosacral cord with dorsal and ventral roots was isolated. The pia-arachnoid membrane was removed, the spinal cord was then mounted on a microslicer (Dosaka EM, Kyoto, Japan) and covered with 95% O2/5% CO2 equilibrated ice-cold Krebs solution (in mM: NaCl 117, KCl 3.6, CaCl2 1.2, MgCl2 1.2, NaH2PO4 1.2, NaHCO3 25, and glucose 11). A 500 μm thick transverse slice with an attached dorsal root was cut and then preincubated in the chamber perfused with Krebs for 1–8 h before recording. Whole-cell voltage-clamp recordings were obtained at 35.0 ± 0.3°C in mixed-gas saturated Krebs with glass micropipettes filled with internal solution (in mM: K-gluconate 135, KCl 5, CaCl2 0.5, MgCl2 2, EGTA 5, HEPES 5 and Mg-ATP 5). The electrode had a resistance of about 8–10 MΩ.
For the neonatal capsaicin-treatment rat model, five rat pups were anesthetized by sodium pentobarbital and injected with 50 mg/kg, s.c. of capsaicin in vehicle (10% ethanol (v/v), 10% Tween (v/v) in 0.9% (w/v) saline) 24 h after birth . For the nerve transection rat model, four 4-week-old rats were anesthetized with sodium pentobarbital (40 mg/kg, i. p.). The right sciatic nerve was exposed at the middle thigh, tightly ligated, and cut just distal to the ligation site to prevent re-establishment of contact, then the wound was sutured . After a survival time of 5 weeks, the rats were subjected to the experimental procedure as described above. In the sciatic nerve ligation model, the recordings were made on the side ipsilateral to the operation site.
SG could be easily distinguished as a colorless band in the superficial dorsal horn under a binocular microscope. SG neurons in this study usually had a resting membrane potential negative than −55 mV. All neurons showed spontaneous mEPSCs at a holding potential of −70 mV where GABAA receptor- and glycine receptor-mediated mIPSCs were negligible . Strength of stimulation to elicit synaptic responses was 1.5 times higher than the threshold at a frequency of 0.1 Hz (duration 0.3 ms) . Signals were amplified with an Axon 200B amplifier (Axon Instruments, Foster City, CA, USA), analog data were digitized with an Digidata 1200A analog/digital interface (Axon Instruments). Data acquirement was controlled through the computer by the pCLAMP 6 (Axon Instruments). The mEPSCs events were analyzed with AxonGraph 3.5 (Axon Instruments). The program for mEPSCs analysis, detected miniature events if the difference between the baseline and a following current value exceeded a given threshold of 3 pA, and separating valleys were < 50% of adjacent peaks. The validity of this method was confirmed by measuring visually individual spontaneous currents on a fast time scale in several cases [5,6].
Drugs (tetrodotoxin (TTX), capsaicin, and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; Sigma, St. Louis, MO, USA)) were applied at known concentrations by switching the perfusion via a three-way tap. Data are presented as mean ± s.e. and statistical significance was determined using either Student's t-test or the Kolmogorov–Smirnov test with p < 0.05 indicating significance. In all cases n refers to the number of neurons studied.
All intact SG neurons in this database had spontaneous mEPSCs with an average frequency of 6.7 ± 0.6 Hz (n = 20) and amplitude of 37.7 ± 3.9 pA (n = 12), respectively. Fine primary afferent fibers in the dorsal horn degenerated after neonatal capsaicin-treatment or sciatic nerve ligation [12,14,15]. But the properties of mEPSCs (frequency, amplitude and half decay time) of intact SG neurons showed no difference to those either after neonatal capsaicin-treatment or sciatic nerve ligation (Table 1), indicating little dependence of mEPSCs on primary afferent fibers.
Figure 1a showed that TTX (1 μM, 3 min) reversibly blocked the dorsal root-evoked synaptic transmission (n = 5), indicating that at this concentration, TTX blocked the synaptic propagation. However, bath application of TTX (1 μM) longer than 4 min resulted in no noticeable effect on either membrane current or distribution of spontaneous mEPSCs in 15 of 16 SG neurons tested, as illustrated in Fig. 1b; the remaining one showed 15% frequency decrease. The results indicated that the spontaneous mEPSCs were mediated by neurotransmitter released spontaneously from presynaptic nerve; second-order neurons were not involved.
The efficacy of synaptic transmission is determined by transmitter release probability and postsynaptic responsiveness. Analysis of frequency and amplitude distributions of mEPSCs permits determination of the loci of experimental manipulation (i.e. presynaptic or postsynaptic) [16,17]. Fig. 2a showed that pretreatment of TTX (1 μM, 3 min) did not prevent capsaicin (2 μM, 60 s)-induced mEPSCs frequency increase, the results were similar to our previous report . In the presence of TTX, capsaicin increased the frequency of mEPSCs while had little effect on average amplitude profiles or mean amplitudes of mEPSCs (Fig. 2b), indicating the frequency increase was mediated by a presynaptic capsaicin action. The capsaicin-induced increase of mEPSCs frequency in the presence of TTX averaged to be 241 ± 27% (n = 9); this value was not distinct from that without TTX (260 ± 21%;n = 9;p > 0.05, unpaired t-test; see  for a similar result). The results indicated that in SG neurons in vitro, noxious stimuli-evoked frequency increase were mediated by primary afferent fiber excitation. All spontaneous and capsaicin-induced mEPSCs could be abolished by non-NMDA receptor antagonist CNQX (10 μM, 3 min), indicating that mEPSCs were mediated by non-NMDA receptor (data not shown; see [5,12] for review).
In the present study, we have analyzed the origins of spontaneous mEPSCs and capsaicin-induced frequency increase of mEPSCs. The properties of spontaneous mEPSCs recorded from intact (normal) rat showed no difference from those recorded from either neonatally capsaicin-treated or sciatic nerve ligated rat, indicating little involvement of fine primary afferent fibers in the spontaneous mEPSCs.
The two types of sodium channels, TTX-sensitive and TTX-resistant, are expressed on dorsal root ganglion (DRG) neurons. Based on molecular and electrophysiological studies, at least two subtypes of TTX-resistant sodium channels (SNS/PN3 and SNS2/NaN) have been identified on DRG neurons which mediate primary afferent; these channels play important roles in nociceptive information transmission, some of the primary afferent might be inhibitory (see [18–22] for review). Capsaicin mainly acts on TTX-resistant but capsaicin-sensitive primary afferent fibers , facilitates glutamate release from primary afferent fiber central terminals , and at the same time blocks some dorsal root-evoked synaptic transmission . In the present study, the capsaicin-induced mEPSCs frequency increase persisted to the presence of TTX. Considering that capsaicin receptor was predominantly located on TTX-resistant fine primary afferent fibers  and the increase ratio showed no difference from that without TTX, the present data suggested that the increase of mEPSCs frequency was mediated by excitation of primary afferent fibers rather than the capsaicin action on dorsal horn interneurons.
Based on the present findings, we conclude that, in SG neurons of rat spinal dorsal horn, spontaneous mEPSCs originate predominantly from interneurons, whereas the noxious stimuli-evoked frequency increase are mainly mediated by increase of neurotransmitter release from fine primary afferent terminals.
It is well accepted that mEPSCs are mediated by presynaptic transmitter(s) release from neural terminals. For the SG neurons, there are three possible origins of the glutamatergic terminal input: primary afferent, interneuron and descending pain control system (mainly from nucleus raphe magnus and periaqueductal gray) . In the present study, the possible effect of descending system was not considered because of the limitation of the in vitro preparation. But we cannot rule out a possibility that under in vivo condition, mEPSCs in SG originate from descending system because it has been suggested that descending system may alter some silent synapses in the dorsal horn to the functional glutamatergic synapses . This idea, however, requires to be confirmed directly.
Supported by the National Natural Science Foundation of China (No.39625011, 39970239, 30000052), Foundation for University Key Teacher by the Ministry of Education of China and the National Program of Basic Research of China (G1999054000) to Y.Q.L.
1. Sugiura Y, Lee CL, Perl ER. Science 1986 234, 358–361.
2. Willis WD, Coggeshall RE. Sensory Mechanisms of the Spinal Cord.
2nd edn. New York: Plenum Press, 1991.
3. Yoshimura M, Nishi S. Neuroscience 1993 53, 519–526.
4. Ueda M, Kuraishi Y, Satoh M. Neurosci Lett 1993 155, 179–182.
5. Yang K, Kumamoto E, Furue H. et al
. Neurosci Lett 1998 255, 135–138.
6. Yang K, Li Y-Q. Neurosci Lett 2000 292, 21–24.
7. Caterina MJ, Schumacher MA, Tominaga M. et al
. Nature 1997 389, 816–824.
8. Tominaga M, Caterina MJ, Malmberg AB. et al
. Neuron 1998 21, 531–543.
9. Holzer P. Pharmacol Rev 1991 43, 143–201.
10. Szallasi A, Blumberg PM. Pharmacol Rev 1999 51, 159–211.
11. Yang K, Kumamoto E, Furue H. et al
. Brain Res 1999 830, 268–273.
12. Yang K, Kumamoto E, Furue H. et al
. Neuropharmacology 2000 39, 2185–2194.
13. Yoshimura M, Jessell T. J Physiol (Lond) 1990 430, 315–335.
14. Hökfelt T, Zhang X, Wiesenfeld-Hallin Z. Trends Neurosci 1994 17, 22–30.
15. Torsney C, Meredith-Middleton J, Fitzgerald M. Dev Brain Res 2000 121, 55–65.
16. Malgaroli A, Tsien RW. Nature 1992 357, 134–139.
17. Khakh BS, Henderson G. J Auton Nerv Syst 2000 81, 110–121.
18. Akopian AN, Sivilotti L, Wood JN. Nature 1996 379, 257–262.
19. Tate S, Benn S, Hick C. et al
. Nat Neurosci 1998 1, 653–655.
20. Akopian AN, Souslova V, England S. et al
. Nat Neurosci 1999 2, 541–548.
21. Fjell J, Hjelmström P, Hormuzdiar W. et al
. Neuroreport 2000 11, 199–202.
22. Szabat E, Soinila S, Happola O. et al
. Neuroscience 1992 47, 409–420.
23. Arbuckle JB, Docherty RJ. Neurosci Lett 1995 185, 70–73.
24. Li P, Zhuo M. Nature 1998 393, 695–698.