α²Adrenoceptor–mediated Presynaptic Inhibition of Primary Afferent Glutamatergic Transmission in Rat Substantia Gelatinosa Neurons

This study was approved by the institutional Animal Use and Care Committee at Saga Medical School (Saga, Japan). The technique used for obtaining slice preparations from the rat spinal cord was the same as that described elsewhere. 27,30Briefly, adult male Sprague-Dawley rats (7–8 weeks old) were deeply anesthetized with urethane (1.2 g/kg). The lumbosacral spinal cord (L1–S3) was then removed and placed in preoxygenated (95% O²and 5% CO²) Krebs solution (117 mm NaCl, 3.6 mm KCl, 2.5 mm CaCl², 1.2 mm MgCl², 1.2 mm NaH²PO⁴, 25 mm NaHCO³, and 11 mm glucose) at 1–3°C; the rats were then immediately sacrificed by exsanguination. After cutting all of ventral and dorsal roots near the root entry zone, except for the L4 or L5 dorsal root on one side, the pia-arachnoid membrane was removed. The spinal cord was mounted on a Vibratome Series 1000® (Technical Products International, O'Fallon, MO), and then a 600- to 650-μm-thick transverse slice was cut with an attached dorsal root having a length of 6–14 mm. The slice was placed on a nylon mesh in the recording chamber and then perfused at a rate of 15–20 ml/min with Krebs solution maintained at 36 ± 1°C. Before the start of the experiment, the slice was preincubated for at least 1 h with Krebs solution.

Substantia gelatinosa neurons were identified by their location under a binocular microscope with light transmitted from below as reported previously 25,27,31; the SG was discernible as a relatively translucent band. Blind whole cell voltage clamp recordings were made from neurons that are located at the center of SG, with a patch pipette that was made up of a thin-walled fiberglass (1.5-mm OD) using a single-stage horizontal puller (P-97; Sutter Instrument, Novato, CA). The patch pipettes had a tip resistance of 5–10 MΩ when filled with a solution having the following composition: 110 mm Cs²SO⁴, 0.5 mm CaCl², 2 mm MgCl², 5 mm EGTA, 5 mm HEPES, 5 mm Mg-ATP, 5 mm tetraethylammonium, and 1 mm guanosine 5′-O-(2-thiodiphosphate), where guanosine 5′-O-(2-thiodiphosphate) and K⁺-channel blockers (Cs⁺and tetraethylammonium) were added to inhibit a hyperpolarizing effect of norepinephrine through the action of G proteins and to block an activation of K⁺channels, respectively. After making a rigid seal (resistances: 5–20 GΩ) in the cell-attached mode by a gentle suction given into the patch pipette, the membrane patch was ruptured by a brief period of more powerful suction, resulting in the whole cell configuration. Signals were amplified with a patch clamp amplifier (Axopatch 200B; Axon Instruments, Foster City, CA) in the voltage clamp mode. Data were low-pass filtered at 5 kHz, digitized at 333 kHz with an A/D converter, and stored and analyzed with a personal computer using the pCLAMP data acquisition program (version 6.0; Axon Instruments). The program used for analyzing miniature EPSCs (mEPSCs) detects spontaneous events if the difference between the baseline and a following current value exceeds a given threshold of 6 pA and separating valleys are less than 50% of adjacent peaks; the validity of the method was confirmed by measuring visually individual mEPSCs on a fast time scale in several cases. 30 

Orthodromic stimulation of the dorsal root to elicit EPSCs was performed with a suction electrode at 0.2 Hz unless otherwise mentioned. The strength of stimuli used was 1.2 times the threshold to elicit EPSCs, fearing a conduction block of action potentials in the dorsal root. The holding potential (V^H) used was −70 mV, at which glycine and γ-aminobutyric acid type A receptor–mediated synaptic currents were invisible. The duration of stimuli used was 0.1 ms throughout the experiments, and conduction velocities were calculated from the latency of monosynaptic EPSC and the length of dorsal root. Aδ-fiber and C-fiber evoked EPSCs were distinguished from each other on the basis of the conduction velocity of afferent fibers and stimulus threshold; they were considered as monosynaptic in origin when the latency remained constant and there was no failure during stimulation at 20 Hz for 1 s or when failures did not occur during repetitive stimulation at 1 Hz for 20 s, respectively, as reported previously. 25,26,32,33During these stimulations, a conduction block of action potentials did not occur when examined by using rat dorsal root ganglion neurons. 25,33Although in some neurons monosynaptic Aδ-fiber or C-fiber EPSCs were accompanied by polysynaptic Aδ-fiber or C-fiber EPSCs, such neurons were not used for analysis if the peak of the monosynaptic EPSCs was contaminated by the polysynaptic EPSCs.

Drugs were dissolved in Krebs solution and then applied to the SG by an exchange of superfusing solution via  a three-way stopcock with one containing them at known concentrations without a change in superfusion rate and thus in temperature. Solutions in the recording chamber completely altered within 20 s. The drugs used in this work were (±)-norepinephrine (Aldrich Chemical Co., Milwaukee, WI), yohimbine hydrochloride, (−)-phenylephrine (Wako, Osaka, Japan), α-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA), (−)-isoproterenol hydrochloride, clonidine hydrochloride, oxymetazoline hydrochloride, guanosine 5′-O-(2-thiodiphosphate) (Sigma, St Louis, MO), and 6-cyano-7-nitroquinoxaline-2,3-dione (Tocris Cookson, St. Louis, MO).

Data are presented as mean ± SEM. Statistical significance was determined as P < 0.05 using either the paired or unpaired Student t  test unless otherwise mentioned. The Kolmogorov-Smirnov test was also used to compare mEPSCs in the absence and presence of norepinephrine in the distributions of their amplitudes and interevent intervals. In all cases, n refers to the number of neurons studied.

Whole cell patch clamp recordings were made from a total of 158 SG neurons. Stable recordings could be obtained from neurons in spinal cord slices maintained in vitro  for more than 12 h, and recordings were made from single neurons for up to 4 h. All experiments were performed at least 10 min later, enough time for guanosine 5′-O-(2-thiodiphosphate) and K⁺-channel blockers in patch pipette solutions to diffuse into SG neurons, after the establishment of whole cell configuration; norepinephrine (50 μm) did not change holding currents at −70 mV. In all SG neurons tested, mEPSCs were downward at −70 mV and were blocked by 6-cyano-7-nitroquinoxaline-2,3-dione (10 μm; n = 2), as reported previously, 27,30indicating the activation of AMPA receptors.

When stimulating the dorsal root with a strength of less than 200 μA, 82% (n = 75; this percentage was almost comparable to a value [79%] obtained previously 27) of 92 neurons examined could elicit monosynaptic EPSCs that had no failure and no change in latency when elicited at 20 Hz (fig. 1A, left). Conduction velocity values (3.8–15 m/s), estimated from the latency of EPSC and the length of dorsal root, were almost within a range of those of Aδ-fibers obtained from the experiment in dorsal root ganglion neurons, as reported previously. 25,32Monosynaptic Aδ-afferent EPSCs evoked at 0.2 Hz had a mean amplitude of 279 ± 23 pA (range, 59–884 pA; V^H=−70 mV;fig. 1A, right). On the other hand, when stimulated with a strength of more than 200 μA, 83% (n = 53; this percentage was also similar to a value [80%] obtained previously 27) of 64 neurons tested exhibited monosynaptic EPSCs that had no failures in response to stimuli at 1 Hz (fig. 1B, left), albeit a variability in EPSC latency was observed in some neurons. 25The EPSCs originated from afferent fibers whose conduction velocities were 0.3–0.8 m/s, a range of those of C fibers. 25,32The amplitude of monosynaptic C-afferent EPSCs evoked at 0.2 Hz averaged to be 253 ± 26 pA (range, 96–714 pA; V^H=−70 mV;fig. 1B, right). Thirty-one percent (n = 20) of the neurons tested exhibited both of the monosynaptic Aδ-fiber and C-fiber EPSCs, as seen in figure 1C. These EPSCs were completely inhibited by 6-cyano-7-nitroquinoxaline-2,3-dione (10 μm; n = 2), indicative of an involvement of AMPA receptors, as reported previously. 27,32 

Effects of norepinephrine (50 μm) on the monosynaptic Aδ-fiber and/or C-fiber EPSCs were examined in a total of 24 SG neurons. As seen from figures 2A–C, each of the EPSCs was inhibited in amplitude by norepinephrine in a reversible manner. These effects on Aδ-fiber and C-fiber EPSCs were, respectively, maximal at approximately 1 and 2 min following the application of norepinephrine (fig. 3A). When examined in many neurons, the inhibitions (measured at approximately 2 min following norepinephrine superfusion) of the peak amplitudes of Aδ-fiber and C-fiber EPSCs were, respectively, 50 ± 4% (n = 20) and 28 ± 4% (n = 8), with the values being significantly different from each other (P < 0.01), as seen in figure 3B. This greater sensitivity of Aδ-fiber EPSC than C-fiber EPSC to norepinephrine was also observed in a neuron in which both of the EPSCs were elicited (n = 5; for example, see fig. 2C).

We next examined which subtypes of adrenoceptors are involved in the norepinephrine-induced inhibition of Aδ-fiber and C-fiber EPSCs by use of their agonists and antagonists. When superfused for 2 min, an α²adrenoceptor agonist, clonidine (10 μm), as well as norepinephrine, inhibited the peak amplitude of monosynaptically evoked Aδ-fiber and C-fiber EPSCs (by 24 ± 2%[n = 13] and 19 ± 3%[n = 11], respectively), as seen in figures 4Aa and Ba. An agonist of α^2Aadrenoceptor, oxymetazoline (10 μm), 34exhibited a similar action on Aδ-fiber and C-fiber EPSCs (inhibition: 34 ± 9%[n = 4] and 16 ± 2%[n = 4], respectively; these were not different from each other;P > 0.05;figs. 4Ab and Bb). On the other hand, an α¹adrenoceptor agonist, phenylephrine (10 μm), and a β adrenoceptor agonist, isoproterenol (40 μm), were without the inhibition on Aδ-fiber and C-fiber EPSCs, as seen in figures 4Ac, Bc, Ad, and Bd. Figure 5summarizes the effects of the adrenoceptor agonists on Aδ-fiber and C-fiber EPSCs. With respect to antagonists, an α²adrenoceptor antagonist, yohimbine (1 μm), superfused for 5 min prior to the application of norepinephrine (50 μm), greatly depressed its inhibitory effect on Aδ-fiber or C-fiber EPSCs (peak amplitude: 95 ± 3%[n = 12] and 97 ± 6%[n = 4] of control, respectively, in the presence of the antagonist), as seen in figures 6A and B. Altogether, these results indicate that the suppressive actions of norepinephrine on Aδ-fiber and C-fiber EPSCs are mediated by α²adrenoceptors.

Figures 7A and Bdemonstrate the effect of norepinephrine (50 μm) superfused for 2 min on mEPSCs recorded from SG neurons. Both mEPSC amplitude and frequency were unaffected by norepinephrine; they were 98 ± 7% (P > 0.05; n = 10) and 102 ± 1% (P > 0.05; n = 10) of control (17 ± 3 pA and 9 ± 2 Hz), respectively, as reported by Baba et al.  23When examined for cumulative distributions of the amplitude and interevent interval of mEPSC, they were also unaffected by norepinephrine, as seen from figure 7C. To determine any effect of norepinephrine on the sensitivity of SG neurons to l-glutamate, we examined whether an AMPA (10 μm) response is affected by norepinephrine (50 μm). As seen in figure 7D, norepinephrine did not affect the peak amplitude of the AMPA response (88 ± 5% of control, n = 12;P > 0.05).

The current study demonstrated that norepinephrine inhibits glutamatergic excitatory transmission to SG neurons in the spinal dorsal horn from the periphery. It was revealed here for the first time that both of Aδ and C primary afferent transmission are depressed by norepinephrine, although Travagli and Williams 24had not examined the action of norepinephrine on each of their transmissions to SG neurons of the spinal trigeminal nucleus. Both actions of norepinephrine were due to the activation of α²adrenoceptors, as reported for evoked EPSCs in spinal trigeminal SG neurons. 24This result may be consistent with the observations that an antinociceptive effect of either intrathecally administrated norepinephrine or electrical stimulation of sites near the A7 cell group and also of the periaqueductal gray is due to the activation of α²adrenoceptors. 4,14,16,17Although the α²adrenoceptors are subdivided into α^2A, α^2B, and α^2Creceptors, 34the α²action in the current study appears to be due to the activation of the α^2Atype because an α^2Aadrenoceptor agonist, oxymetazoline, 34reduced the Aδ and C afferent transmission. This idea is supported by a report of Stone et al.  35which demonstrated that a mouse expressing a point mutation in the α^2Areceptor was without α²agonist–mediated spinal analgesia in the tail-flick test. Although the activation of β adrenoceptors is known to enhance neurotransmitter release in many types of neurons, including sympathetic ganglion neurons, 36this was not the case in SG neurons because isoproterenol did not affect the transmission.

Since norepinephrine did not affect mEPSC amplitude and the sensitivity of SG neurons to AMPA, its actions on evoked EPSCs were presynaptic in origin. Since these actions were examined for monosynaptic transmission, the norepinephrine actions are suggested to be due to the activation of α², possibly α^2A, adrenoceptors existing in primary afferent central terminals. Consistent with this idea, an inhibition by norepinephrine of the release of l-glutamate from rat spinal cord synaptosomes is mediated by α²adrenoceptors, 37and the α^2Aadrenoceptor is expressed in primary afferent terminals in the rat. 20The action on C-fiber transmission in the current study may be consistent with the observation that capsaicin-induced release of l-glutamate from spinal cord synaptosomes is reduced by the activation of the α^2Aadrenoceptor 38because capsaicin is known to excite C fibers. Although a cellular mechanism for the norepinephrine action was not examined here, this would be due to an inhibition of voltage-gated Ca²⁺channels existing in nerve terminals because norepinephrine reduces Ca²⁺-channel currents in many types of neurons, including dorsal root ganglion neurons. 39There was a dissociation between the actions of norepinephrine on evoked EPSC and mEPSC in that the former was presynaptically depressed, while mEPSC frequency was unaffected, an observation different from that in spinal trigeminal SG neurons. 24One explanation for this discrepancy may be that norepinephrine acts differentially on evoked and spontaneous transmitter release mechanisms in the spinal dorsal horn, resulting in a distinct action on their EPSCs, because each of the releases in the SG is suggested to be mediated by different types of Ca²⁺channels. 40Alternatively, mEPSCs in spinal cord SG neurons may be produced by inputs not only from primary afferent fiber axon but also from interneuron axon, the former terminals having more α²adrenoceptors than the latter ones, where mEPSCs mainly originate from interneuron terminals. This idea may be consistent with the observation that many α^2Aadrenoceptors are colocalized in the spinal dorsal horn with neuropeptides, which are contained in primary afferent fibers. 20A similar discrepancy between spontaneous and evoked transmission has been seen in the actions of a μ-opioid receptor agonist 31and anandamide 28in spinal cord SG neurons. Although Pan et al.  41have very recently reported a clonidine-induced decrease in mEPSC frequency in SG neurons, a discrepancy between this and the current study may be due to the fact that different SG neurons were tested because they examined neurons in the outer layer of SG, while we investigated neurons located at the center of SG. The possibility cannot be ruled out that SG neurons exhibiting no effect of norepinephrine on mEPSCs in the current study (where the blind patch clamp technique was used) had located in the inner layer of SG because visually identified neurons in the inner layer of SG appeared to be without actions of clonidine on mEPSCs. 41 

The current study revealed for the first time that norepinephrine inhibits Aδ-fiber transmission more effectively than C-fiber transmission. There are two possible explanations for this result. One is that α²adrenoceptors are more densely expressed in Aδ-fiber than C-fiber terminals in the spinal dorsal horn. The other is that there is a different type of α²adrenoceptors in each of the Aδ-fiber and C-fiber terminals, although oxymetazoline could not discriminate between them in the extent of inhibition. This remains to be examined by using other agents regarding subtypes of α²adrenoceptors. The action of norepinephrine was the same as that of anandamide 28but different from those of baclofen 25and nociceptin 27in that Aδ-fiber transmission was more sensitive than C-fiber transmission. It is suggested that norepinephrine as well as anandamide may inhibit fast-conducting transmission more potently than slow-conducting pain transmission.

In conclusion, the current study provides a cellular basis for the antinociceptive action of norepinephrine through a mechanism in primary afferent terminals at the spinal cord level. Although norepinephrine may contribute to a prolongation of analgesia through its vasoconstrictive action in spinal anesthesia, the current finding of the inhibition of excitatory transmission supports its role as an important negative modulator of pain transmission to SG neurons together with a norepinephrine-induced hyperpolarization 21,22and enhancement of inhibitory transmission. 23 

The authors thank Tadahide Totoki, M.D., Ph.D. (Professor of Anesthesiology, Saga Medical School, Saga, Japan), for his encouragement during this study.