Peripheral nerve injury induces upregulation of the calcium channel α2δ-1 structural subunit in dorsal root ganglia (DRG) and dorsal spinal cord of spinal nerve-ligated rats with neuropathic pain, suggesting a role of the calcium channel α2δ-1 subunit in central sensitization. To investigate whether spinal dorsal horn α2δ-1 subunit upregulation derives from increased DRG α2δ-1 subunit and plays a causal role in neuropathic pain development, we examined spinal dorsal hornα2δ-1 subunit expression with or without dorsal rhizotomy in spinal nerve-ligated rats and its correlation with tactile allodynia, a neuropathic pain state defined as reduced thresholds to non-noxious tactile stimulation. We also examined the effects of intrathecal α2δ-1 antisense oligonucleotides on α2δ-1 subunit expression and neuropathic allodynia in the nerve-ligated rats. Our data indicated that spinal nerve injury resulted in time-dependentα2δ-1 subunit upregulation in the spinal dorsal horn that correlated temporally with neuropathic allodynia development and maintenance. Dorsal rhizotomy diminished basal level expression and blocked injury-induced expression of the spinal dorsal hornα2δ-1 subunit and reversed injury-induced tactile allodynia. In addition, intrathecal α2δ-1 antisense oligonucleotides blocked injury-induced dorsal horn α2δ-1 subunit upregulation and diminished tactile allodynia. These findings indicate that α2δ-1 subunit basal expression occurs presynaptically and postsynaptically in spinal dorsal horn. Nerve injury induces mainly presynaptic α2δ-1 subunit expression that derives from increased α2δ-1 subunit in injured DRG neurons. Thus, changes in presynaptic α2δ-1 subunit expression contribute to injury-induced spinal neuroplasticity and central sensitization that underlies neuropathic pain development and maintenance.
Spinal sensitization derived from altered gene expression in the spinal dorsal horn is an important component in nerve injury-induced pain. Altered gene expression can result in spinal circuit reorganizations and changes in the excitatory or inhibitory tones that lead to neuropathic pain development. Among genes that undergo upregulation or downregulation after injury, changes in the voltage-gated calcium channel (VGCC) α2δ-1 subunit in dorsal root ganglion (DRG) neurons correlate with the onset and duration of tactile allodynia, a neuropathic pain state in which innocuous tactile stimulation elicits pain behavior (Luo et al., 2001; Newton et al., 2001; Wang et al., 2002; Valder et al., 2003). Because central axons of DRG neurons form synapses with dorsal horn neurons that are critical for sensory processing, injury-induced α2δ-1 subunit expression in DRG neurons may contribute to spinal neuroplasticity in neuropathic pain.
The VGCC consists of the channel-forming α1 and β, γ, and α2δ subunits. The α2δ subunit is a heavily glycosylated structural subunit containing covalently linked α2 and δ peptides that are encoded by the same gene (De Jongh et al., 1990). The α2δ subunit is important for functional assembly and expression of the VGCC (Mori et al., 1991; Williams et al., 1992; Brust et al., 1993; Gurnett et al., 1996; Kang et al., 2002). So far, four α2δ genes have been identified that encode the α2δ-1, α2δ-2, α2δ-3, and α2δ-4 subunits, respectively (Ellis et al., 1988; Klugbauer, 1999; Qin et al., 2002). These subunits have distinct tissue-specific expression patterns suggesting diversified functions associated with each individual α2δ subunit (Marais et al., 2001). This is also true for sensory processing, because α2δ-1 and α2δ-2 subunit mRNA is highly expressed in small DRG neurons but at a lower level in large DRG neurons, whereas α2δ-3 mRNA is relatively abundant in large DRG neurons but scarce in small sensory neurons (Yusaf et al., 2001). Expression and distribution of the calcium channel α2δ subunits in spinal cord is essentially unknown. Binding studies have shown that the α2δ-1 and α2δ-2, but not the α2δ-3, subunits bind with high affinities to gabapentin (Marais et al., 2001), a drug that has antineuropathic pain efficacy with unknown mechanisms (Taylor et al., 1998; Laird and Gidal, 2000).
Although calcium channel α2δ-2 subunit levels in spinal cord and DRG under a neuropathic pain condition remain to be determined, we observed enhanced dorsal horn and DRG α2δ-1 subunit expression in a rat neuropathic pain model (Luo et al., 2001, 2002). However, how these changes relate to neuropathic pain development and maintenance are not clear. Injury-induced spinal α2δ-1 subunit could derive from increased presynaptic expression of the DRG α2δ-1 subunit and/or trans-synaptic induction in dorsal horn cells, which affect different aspects of pain transduction, such as neurotransmitter release or postsynaptic neuronal excitability, respectively. Accordingly, we examined spinal cord α2δ-1 subunit expression in a rat neuropathic pain model derived from tight ligation of the L5/L6 spinal nerves, its relationship to injury-induced DRG α2δ-1 subunit expression, and its contribution to neuropathic allodynia.
Materials and Methods
Materials. The monoclonal antibodies against rabbit calcium channel α2 subunit were from Sigma (St. Louis, MO), and those against endothelial nitric oxide synthase (eNOS) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were from BD Biosciences (San Diego, CA) and Ambion (Austin, TX), respectively. Membrane extracts of rat brain and human endothelial cells were used as positive controls for the α2δ-1 subunit and eNOS, respectively. Tris-acetate gels (NuPAGE) and buffers were obtained from Invitrogen (Carlsbad, CA). Horseradish peroxidase-labeled secondary antibodies (mouse IgG) and its substrate and enhancer solutions were from Pierce (Rockford, IL). Other chemicals were from Sigma.
Animals. Male rats (Harlan Sprague Dawley; Harlan, Indianapolis, IN) were housed in separate cages and exposed to a 12 hr light/dark cycle with food and water ad libitum. All animal care and experiments were performed according to protocols approved by the Institutional Animal Care Committees of the University of California.
Neuropathic surgeries. Spinal nerve ligation was performed as described by Kim and Chung (1992). Briefly, the left L5/L6 lumbar spinal nerves were exposed in anesthetized male Harlan rats (100-150 gm) and ligated tightly with 6.0 silk sutures between the DRGs and the conjunction to form the sciatic nerve. In some animals, ipsilateral dorsal rhizotomy was performed during the same operation. Briefly, an incision was made at the L1 and L2 dorsal aspect, and a left hemilaminectomy was performed. After the dura was opened, the dorsal roots of L5/L6 DRG were exposed and cut to remove ∼2 mm in length (Luo et al., 2001). The anatomical connections of the rhizotomized dorsal roots were confirmed after tissue collections. Sham ligation and rhizotomy were performed in the same ways except that spinal nerves and dorsal roots were not ligated or transected, respectively.
Intrathecal antisense oligonucleotide treatment. Antisense and mismatch oligonucleotides (21 nucleotides each) with phosphothioate modification on three nucleotides each at the 5′ and 3′ ends were synthesized and purified with High Purity Salt Free method by MWG Biotech (High Point, NC). The antisense oligonucleotide sequence against a region containing the start codon ATG on the rat brain α2δ-1 gene (GenBank accession number M86621) was AGCCATCTTCGCGATCGAAG, and the mismatch control sequence was CGATACCTCGCTGGCTAAAG. These oligonucleotides were precipitated and washed in 75% ethanol solutions and dissolved in sterile saline before injecting into allodynic rats through intrathecal catheters (Yaksh and Rudy, 1976) in a total volume of 10 μl twice daily for 4 d, starting 3 weeks after the nerve ligation. The same amount of sterile saline was injected into some allodynic rats as control. Sterile saline (10 μl) was used to flash the catheter after each injection. Behavioral testing was performed before the first injection and daily before the first daily injections each day for 4 d as described below.
Behavioral testing. Tactile allodynia was tested as described by Chaplan et al. (1994). Briefly, rats were allowed to acclimate for at least 15 min in a clear plastic cage with a wire mesh bottom and were then tested for the 50% paw withdrawal threshold (PWT) to von Frey filaments (Stoelting, Wood Dale, IL) using a modified up-down method of Dixon (1980). A calibrated filament with a 2.0 gm buckling weight was applied to the plantar surface of the left hindpaw with a force causing the filament to bend. Paw lifting within 5 sec indicated a positive response and led to the use of the next weaker filament, whereas absence of a paw lifting after 5 sec led to the use of the next filament with increasing weight. This paradigm continued until six measurements starting from the one before the first positive response were made, or until four consecutive positive (assigned a score of 0.25 gm) or five consecutive negative (assigned a score of 15 gm) responses had occurred. The 50% paw withdrawal threshold was calculated from the resulting scores as described previously (Luo et al., 2001).
Western blot. Frozen spinal cord and DRG samples were pulverized and extracted in 50 mm Tris buffer, pH 8.0, containing 0.5% Triton, 150 mm NaCl, 1 mm EDTA, and protease inhibitors. The cell extracts were applied to NuPAGE Tris-acetate gels under reducing conditions (0.05 m DTT) for electrophoresis and then transferred to nitrocellulose membranes (Schleicher and Schuell, Keene, NH) electrophoretically. The membranes were then incubated with α2 monoclonal antibodies in PBS with 0.1% Tween 20 for 1 hr at room temperature or overnight at 4°C after nonspecific binding sites were blocked with 5% low-fat milk in PBS. Secondary antibodies labeled with horseradish peroxidase were used to detect the antibody-protein complexes by incubating with the membrane at room temperature for 1 hr followed by washing, addition of chemiluminescent reagents, and x-ray radiography. In our experiments, the positive bands detected by the primary antibody reflect the α2 subunit only because the δ peptide separates from the α2 subunit under reducing conditions (Jay et al., 1991). The band densities were quantified by densitometry within the linear range of the film sensitivity curve.
Statistical analyses. Data were reported as means ± SE (SEM). Unpaired or paired Student's t tests were performed where significance was indicated by two-tailed p values <0.05.
Peripheral nerve injury induced upregulation of the α2δ-1 subunit in dorsal spinal cord that correlated with neuropathic tactile allodynia
Our previous experiments have shown that spinal nerve ligation-induced injury in rats causes upregulation of the α2δ-1 subunit in DRG neurons that correlates with neuropathic allodynia development and maintenance, suggesting an important role of DRG α2δ-1 subunit expression in neuropathic pain (Luo et al., 2001; Valder et al., 2003). To investigate whether nerve injury induces α2δ-1 subunit expression in the spinal cord that also correlates with neuropathic pain development, we examined the α2δ-1 subunit levels in rat spinal cord and hindpaw withdrawal thresholds to mechanical stimulation at different time points after L5/L6 spinal nerve ligation. Spinal nerve ligation-induced injury caused upregulation of the α2δ-1 subunit in dorsal (Fig. 1A,B), but not ventral (data not shown), spinal cord. The increase was most dramatic at time points when the nerve-injured rats exhibited the most severe tactile allodynia. When the nerve-ligated animals were recovering from the allodynia state, elevated α2δ-1 subunit started to diminish, and by 17 weeks after nerve injury, both dorsal horn α2δ-1 subunit levels and the tactile allodynia were returned to the control levels (Fig. 1B). In addition, the peak increase of α2δ-1 subunit in spinal dorsal horn (5 weeks after injury) lacked behind that in DRG, which reached the maximal level 1-2 weeks after injury (Luo et al., 2001). These data suggest that injury-induced changes in dorsal horn α2δ-1 subunit expression may derive from changes in the DRG α2δ-1 subunit and contribute to injury-induced spinal neuroplasticity mediating neuropathic pain development and maintenance.
Endogenous α2δ-1 subunit is likely expressed in spinal dorsal horn presynaptically and postsynaptically, and injury-induced presynaptic α2δ-1 expression may contribute to genesis and maintenance of neuropathic allodynia
The basal level of α2δ-1 subunit in spinal dorsal horn could be expressed at the presynaptic terminals and/or postsynaptic dorsal horn neurons and other cell types, thus contributing differentially to presynaptic and postsynaptic neurotransmission. Similarly, nerve injury can increase presynaptic α2δ-1 subunit expression through central transport of increased α2δ-1 subunit from DRG cell bodies and/or induce trans-synaptic expression of α2δ-1 subunit in postsynaptic dorsal horn neurons (Coull et al., 2003). To distinguish these possibilities, we examined nerve injury-induced dorsal horn α2δ-1 subunit expression and its relationships to DRG α2δ-1 subunit regulation and tactile allodynia in nerve-ligated rats with or without dorsal rhizotomy. Western blot analyses indicated that nerve injury induced upregulation of the α2δ-1 subunit in DRG and spinal dorsal horn 1 week after spinal nerve injury (Fig. 2). The increased dorsal horn α2δ-1 subunit mimicked the slower migration rate of the DRG α2δ-1 subunit (Fig. 2A) (Luo et al., 2001), suggesting that it may come from injured DRG neurons. Dorsal rhizotomy, which posted minimal effects on DRG α2δ-1 subunit expression, blocked the injury-induced dorsal horn α2δ-1 subunit upregulation to a level even lower than that of the endogenous α2δ-1 subunit seen in the contralateral side (Fig. 2). Importantly, dorsal rhizotomy also blocked spinal nerve ligation-induced tactile allodynia (Fig. 3). These data suggest that under normal conditions, the α2δ-1 subunit is likely expressed in both presynaptic and postsynaptic terminals in the dorsal horn, and the presynaptic α2δ-1 subunit is mainly transported from DRG neurons through their central axons. Nerve injury induces upregulation of the presynaptic dorsal horn α2δ-1 subunit that is originated from DRG.
Intrathecal antisense oligonucleotide treatment diminished and blocked injury-induced α2δ-1 subunit in DRG and dorsal spinal cord, respectively, and reversed tactile allodynia
To investigate whether elevated α2δ-1 subunit in dorsal spinal cord and DRG plays a causal role in nerve injury-induced tactile allodynia, we treated the nerve-injured animals with antisense oligonucleotides complementary to a region in the α2δ-1 gene that contains the start codon ATG. As indicated in Figure 4, intrathecal treatment with antisense oligonucleotides at a dose of 30 μg per rat, twice daily for 4 d, but not with saline or mismatch oligonucleotide controls, diminished endogenous (contralateral side) as well as nerve injury-induced (ipsilateral side) α2δ-1 subunit expression in DRG, and blocked injury-induced α2δ-1 subunit upregulation in the spinal dorsal horn. In addition, the same antisense treatment also diminished injury-induced tactile allodynia in a dose-dependent manner (Fig. 5). However, the reversal of neuropathic allodynia after antisense oligonucleotide treatment was not complete, and we observed significant variations in the treatment outcome as indicated by the size of the error bars in Figure 5.
Our data indicated that α2δ-1 subunit is upregulated in dorsal, but not ventral, spinal cord by peripheral nerve injury signals. Because dorsal spinal cord contains afferent central terminals and dorsal horn neurons as well as interneurons participating in sensory processing, our data suggest that injury-induced upregulation of the α2δ-1 subunit may contribute to spinal neuroplasticity, leading to abnormal sensations in nerve-injured animals. Three lines of independent evidences from our study support this hypothesis. First, upregulation of the spinal α2δ-1 subunit only occurs in the dorsal, but not ventral, horn and correlates temporally with injury-induced tactile allodynia. Second, both dorsal horn α2δ-1 subunit upregulation and tactile allodynia in nerve-ligated animals can be blocked by dorsal rhizotomy. Third, intrathecal antisense, but not mismatch, oligonucleotides can reverse at least partially the injury-induced α2δ-1 subunit upregulation and tactile allodynia.
It is known that calcium channels are expressed in both presynaptic terminals and postsynaptic dorsal horn neurons (Ahlijanian et al., 1990). Thus, we would expect both presynaptic and postsynaptic α2δ-1 expression. The postrhizotomy reduction, but not elimination, of α2δ-1 subunit expression in dorsal horn ipsilateral to the injury to a level lower than that in the contralateral dorsal spinal cord confirms the presynaptic and postsynaptic expression of the α2δ-1 subunit and suggests that, under a normal condition, the presynaptic α2δ-1 subunit derives from DRG α2δ-1 subunit that may undergo axonal transport to the central terminals.
The blockade of dorsal horn α2δ-1 subunit induction with rhizotomy indicates that the injury-induced spinal α2δ-1 subunit increase is mostly presynaptic and derived from increased α2δ-1 subunit expression in injured DRG. In addition to modulation of DRG neuron excitability, the injury-induced DRG α2δ-1 subunit can be transported to central terminals in the spinal dorsal horn where it modulates presynaptic neurotransmission. This presynaptic neuroplasticity may underlie neuropathic pain development and maintenance. Alternatively, rhizotomy could block injury-induced postsynaptic induction of the α2δ-1 subunit that may contribute to enhanced excitability of dorsal horn neurons leading to neuropathic pain. However, our data support a presynaptic mechanism based on the following observations. First, injury-induced maximal increase in spinal cord α2δ-1 subunit expression lacks behind that in DRG (Fig. 1) (Luo et al., 2001). Second, injury-induced spinal dorsal horn α2δ-1 subunit mimics the slow migration pattern of the DRG α2δ-1 subunit in Western blots (Fig. 2A), and rhizotomy blocks the induction and reverses tactile allodynia (Figs. 2, 3). Third, spinal nerve ligation and rhizotomy have opposite effects on spinal α2δ-1 subunit expression, suggesting that these injuries do not independently regulate spinal cord α2δ-1 subunit expression through a common injury-related mechanism, including that mediated by injury factors such as neurotrophic factors and cytokines generated at the injury sites. Rather, the outcome of rhizotomy on spinal dorsal horn α2δ-1 subunit expression is more consistent with blocking the effects of spinal nerve ligation on dorsal horn α2δ-1 subunit expression that is most likely derived from injured DRG central axons.
The involvement of dorsal horn α2δ-1 subunit in neuropathic pain is further supported by our finding that intrathecal antisense oligonucleotides diminish and block, respectively, injury-induced DRG and dorsal horn α2δ-1 subunit expression and reverse partial injury-induced tactile allodynia. The incomplete reversal of tactile allodynia after antisense treatments may be caused by many factors, including pharmacodynamics, pharmacokinetics, and distributions of injected oligonucleotides. The following may also be possible: (1) the antisense oligonucleotides may not be site specific to completely block injury-induced α2δ-1 subunit in presynaptic terminals, because the proteins are made in DRG, and intrathecal antisense treatment can only diminish, but not abolish, injury-induced DRG α2δ-1 subunit upregulation; (2) changes in α2δ-1 subunit expression may mediate one of many pathways involved in neuropathic pain; and (3) the α2δ-1 subunit may be a cofactor or comediator, but not a determinant, in neuropathic allodynia. These possibilities may not be mutually exclusive.
The causal role of α2δ-1 subunit in tactile allodynia is not clear at the present time. Findings from biochemical and biophysical studies have indicated that this subunit is critical in functional assembly and expression of the voltage-gated calcium channels (Brust et al., 1993; Gurnett et al., 1996, 1997; Kang et al., 2002). Recent data from electron cryomicroscopy study on the L-type calcium channels have indicated that the extracellular α2 subunit protrudes from the membrane in close proximity to the channel forming α1 subunit (Wolf et al., 2003), supporting an important role of this subunit to calcium channel functions. Thus, it is possible that this subunit is a rate-limiting subunit in calcium channel assembly, and overexpression of this subunit may lead to enhanced expression of functional calcium channels. Increased dorsal horn presynaptic calcium channels can lead to enhanced and/or prolonged presynaptic terminal excitability, resulting in elevated excitatory neurotransmitter release after non-noxious stimulation, which evokes a pain response (tactile allodynia). This hypothesis is supported by the findings that calcium channels are expressed at the presynaptic terminals and involved in excitatory neurotransmitter release (Matthews and Dickenson, 2001; Kochegarov, 2003), nerve injury modulates dorsal horn release of excitatory neurotransmitters (Skilling et al., 1992; Noguchi et al., 1995; Malcangio et al., 2000), and gabapentin, an antihyperalgesic drug that binds to the α2δ-1 subunit, inhibits excitatory neurotransmitter release from presynaptic terminals only in hyperalgesic, but not normal, rat spinal cord and caudal trigeminal nucleus slices (Patel et al., 2000; Maneuf et al., 2001). Colocalization of this subunit with other calcium channel-forming subunits in the presynaptic terminals and in vivo electrophysiological studies will provide direct evidences to prove or disprove this hypothesis.
Alternatively, increased α2δ-1 subunit may play a yet identified role, in addition to being the calcium channel structural subunit, in neuropathic pain processing. This is supported by our findings that nerve injury only induces α2δ-1, but not other, calcium channel subunit expression (Luo et al., 2001). In addition, the structure of DRG α2δ-1 subunit differs from that of α2δ-1 subunits from spinal cord, brain, and skeletal muscle (Luo, 2000), suggesting distinct functions for the DRG and presumably the presynaptic α2δ-1 subunit in sensory processing. Nevertheless, our data support a contributory role of spinal plasticity mediated by the calcium channel α2δ-1 subunit in neuropathic pain development and maintenance.
This study was supported in part by National Institutes of Health Grants DE13270, DE14545, and NS40135 (Z.D.L.). We thank A. Boroujerdi and Dr. K.-W. Li for assistance in some experiments.
Correspondence should be addressed to Dr. Z. David Luo, Department of Anesthesiology, University of California, Irvine Medical Center, 101 The City Drive South, Orange, CA 92868. E-mail:.
Copyright © 2004 Society for Neuroscience 0270-6474/04/248494-06$15.00/0