Report

Inhibition of Hyperalgesia by Ablation of Lamina I Spinal Neurons Expressing the Substance P Receptor

See allHide authors and affiliations

Science  10 Oct 1997:
Vol. 278, Issue 5336, pp. 275-279
DOI: 10.1126/science.278.5336.275

Abstract

Substance P is released in the spinal cord in response to painful stimuli, but its role in nociceptive signaling remains unclear. When a conjugate of substance P and the ribosome-inactivating protein saporin was infused into the spinal cord, it was internalized and cytotoxic to lamina I spinal cord neurons that express the substance P receptor. This treatment left responses to mild noxious stimuli unchanged, but markedly attenuated responses to highly noxious stimuli and mechanical and thermal hyperalgesia. Thus, lamina I spinal cord neurons that express the substance P receptor play a pivotal role in the transmission of highly noxious stimuli and the maintenance of hyperalgesia.

A subpopulation of dorsal root ganglion neurons synthesize (1) and transport (2) substance P (SP) to the spinal cord, where it is released upon noxious stimulation of the innervated peripheral tissue (3). Although SP excites spinal cord nociresponsive neurons (4), the role that SP and the substance P receptor (SPR) play in signaling nociceptive information remains unclear. In the normal animal, SP, upon release from primary afferents, diffuses to and interacts primarily with SPR-expressing neurons located in lamina I of the spinal cord (5-7). A high proportion of spinothalamic and spinobrachial neurons located in lamina I express SPR (8), suggesting that these SPR-expressing neurons play a role in the ascending conduction of nociceptive information.

To investigate the functions of lamina I SPR-expressing neurons in nociceptive signaling, we selectively ablated these neurons by infusing a cytotoxin conjugated to SP into the intrathecal space of the spinal cord in rats. When SP binds to spinal cord neurons expressing the SPR, both SP and SPR are rapidly internalized (5-7). Using SP-induced internalization of SPR as a specific portal of entry into SPR-expressing spinal cord neurons, we conjugated SP to the ribosome-inactivating protein saporin (SAP). This substance P–saporin conjugate (SP-SAP), like other saporin conjugates, must be internalized to exert its toxicity as it inactivates and ultimately kills cells by blocking protein synthesis (9). We performed a series of correlative in vitro and in vivo studies to determine the specificity and toxicity of SP-SAP, as well as functional changes in somatosensory processing.

Competition binding studies with 125I-SP binding to membranes of the adult rat spinal cord demonstrated that SP-SAP [median inhibitory concentration (IC50) = 2.2 nM] was nearly equipotent with SP (IC50 = 2.0 nM) in blocking the binding of 125I-SP to the SPR (10), whereas neurokinin A (IC50 = 5 nM) was less potent, and SAP alone (1 μM) or the unrelated peptide endothelin-1 (1 μM) were totally inactive.

SP-SAP internalization and cytotoxicity were examined in primary cultures of neonatal rat spinal cord neurons (11), in which ∼15% of the neurons express the SPR (12). Both SP (10−7 M) and SP-SAP (10−7 M) induced a rapid and similar extent of SPR internalization that was blocked by 5 × 10−6 M of the nonpeptide SPR antagonist RP67580. Two hours after treatment with 10−7 M SP-SAP, but not 10−7 M SAP, SPR internalization was visualized with an antibody that recognized SPR (Fig. 1A), and intracellular accumulation of SAP was visualized with an antibody that recognized SAP (Fig. 1B).

Figure 1

Internalization and cytotoxicity of SP-SAP in primary cultures of neonatal spinal cord neurons (11). Confocal image of neurons where the SPR immunofluorescence (A, C, D) appears red, areas of concentrated SPR immunofluorescence appear yellow. (A, C, and D) SPR immunofluorescence in neurons 2 hours, 1 day, and 4 days, respectively, after treatment with SP-SAP. (B) Confocal image showing SAP immunofluorescence (yellow) 2 hours after SP-SAP treatment. These images were projected from 14 optical sections acquired at 0.8-μm intervals with a 60× lens. Bar, 25 μm.

One day after SP-SAP treatment, there was no significant loss of cultured SPR-immunoreactive neurons, although in most of these neurons the SPR-immunoreactivity was localized within intracellular endosomes (Fig. 1C). In contrast, 1 day after treatment with SP alone, most of the SPR neurons had recycled to the plasma membrane. Thus, within 24 hours after SP-SAP internalization, these neurons could no longer efficiently recycle SPR back to the plasma membrane. Four days after SP-SAP treatment, there was an 82% decrease in the number of SPR-immunoreactive neurons, at 7 days a 95% reduction, and at 10 days there were no SPR-immunoreactive neurons remaining in culture. At 4 and 7 days after treatment, the surviving SPR-immunoreactive neurons showed shrunken cell bodies, diffuse SPR immunoreactivity throughout the cytoplasm (Fig. 1D), and shortened dendritic processes. In contrast, nearby neurons that did not express SPR immunoreactivity, but did express the neuronal marker microtubule-associated protein–2 (MAP-2), appeared morphologically normal. Treatment of cultured spinal cord neurons with saline, SP, or SAP alone resulted in no significant morphological or cytotoxic changes in either the SPR-expressing neurons or the non-SPR, MAP-2–immunofluorescent neurons.

To estimate the placement of the intrathecal catheter and the potential spread of the SP-SAP, we injected 10 μl of the dye Fast Green with the end of the intrathecal catheter placed at L4; 1 hour later, the dye had intensely labeled the spinal cord from spinal segments L2 to L5 (13). One hour after injection of 10 μl of 5.0 × 10−6 M SP-SAP, SPR internalization was observed in SPR-immunoreactive dendrites and cell bodies in lamina I of the spinal cord at spinal segments L2 to L5 (Fig.2). Internalization of SPR presumably reflected the sites where SP-SAP had bound to SPR and induced the internalization of both SPR and SP-SAP (5-7). After injection of SP-SAP, a significant loss of SPR immunoreactivity was first detected at 7 days after treatment. This loss of SPR immunoreactivity was confined to lamina I in spinal segments L2 to L5, and the loss of SPR immunofluorescence in lamina I was observed through 28 days after treatment (Fig. 3B), which was the last time point examined. In contrast, injection of saline, SP, or SAP alone produced no change in SPR immunoreactivity in lamina I in spinal segments L2 to L5 at any of the time points examined (Fig. 3A).

Figure 2

SPR immunofluorescence in lamina I neurons of the spinal cord after SP-SAP treatment (13). SPR immunofluorescence in lamina I neurons in sagittal sections of the L4 spinal segment of the spinal cord at 1 hour after infusion of saline (A), 1 hour after infusion of SP-SAP (B), and an absorption control for SPR immunofluorescence in a saline-infused animal (C). In these confocal images the SPR immunofluorescence appears red and areas of intense SPR immunofluorescence appear yellow. Infusion of SP-SAP induced a marked translocation of the SPR from the plasma membrane (A) into intracellular endosomes (B). These images were projected from 18 optical sections acquired at 0.7-μm intervals with a 60× lens. Bar, 25 μm.

Figure 3

Cytotoxicity after intrathecal infusion of SP-SAP in the spinal cord (13). Confocal images of SPR immunofluorescence in the spinal cord 28 days after infusion of saline (A) or SP-SAP (B), where the SPR immunofluorescence appears yellow. The only difference between saline- and SP-SAP–treated animals is the marked reduction in SPR immunofluorescence in lamina I (arrows) of the SP-SAP–treated animals. These images are 60-μm-thick tissue sections acquired with a 10× lens. Bar, 400 μm.

Twenty-eight days after injection of saline, SP, SAP, or SP-SAP, spinal cords and dorsal root ganglia (L4) were histologically examined (12) to determine which cell populations had been affected by these treatments (Table 1). Measurements were made of neuronal cell populations expressing SPR (labels lamina I, III to V, and the preganglionic sympathetic neurons at spinal segment T10), calbindin (labels a subset of lamina I and II neurons), ChAT (labels motor neurons), SP (labels cell bodies in the L4 dorsal root ganglia), as well as immunofluorescence for SP in lamina I (labels SP primary afferent inputs), MAP-2 (labels all neurons), glial fibrillary acidic protein (GFAP; labels astrocytes), and OX-42 (labels microglia).

Table 1

Cytotoxicity of intrathecally infused saline, SP, SAP, and SP-SAP in the L4 segment of the spinal cord at 28 days after treatment (12). Cell numbers and immunofluorescence levels were determined by confocal microscopy. In all instances the saline, SP, and SAP animals were not significantly different from normal untreated control animals, and thus only the values for the saline-, SAP-, and SP-SAP–infused animals are shown. The only significant difference in the SP-, SAP-, or SP-SAP–treated animals as compared with saline-treated controls was the loss of lamina I SPR-immunoreactive neurons and the loss of SPR immunoreactivity in lamina I of the spinal cord in the SP-SAP–treated animals. Data points are expressed as the mean ± SEM (n = 6), and significant differences were calculated by a one-way ANOVA and Bonferroni comparisons (*P < 0.01).

View this table:

Examination of neuronal markers and immunofluorescence intensity values showed that the only significant changes observed at 28 days after treatment with saline, SP, SAP, or SP-SAP was that SP-SAP treatment reduced the number of lamina I SPR–immunoreactive neurons in lamina I and in the levels of SPR immunofluorescence in lamina I (Fig.3 and Table 1). Infusion of SP-SAP produced an 85% reduction in the number of SPR immunofluorescent neurons in lamina I at spinal cord segment L4. The surviving 15% of the lamina I immunoreactive neurons showed shrunken cell bodies, shortened cell processes, and SPR immunoreactivity that was diffusely distributed throughout the cytoplasm with little SPR present on the plasma membrane. In contrast, there was not a significant reduction in the total number, or evidence of cytotoxicity, in SPR-immunoreactive neurons located in lamina III to V or X at the L4 spinal segment or in preganglionic sympathetic neurons at spinal segment T10 (Table 1).

Examination of the spinal cords treated with saline, SP, or SAP alone showed that these treatments did not produce a significant change in cell number, morphology, or fluorescence level of any of the cell markers examined (Table 1). Thus, the cytotoxicity of intrathecally infused SP-SAP was limited to the SPR-expressing lamina I neurons in spinal segments L2 to L5.

Intrathecal infusion of saline, SP, SAP, or SP-SAP produced no detectable changes in body weight, food intake, alertness, locomotion, or grooming behavior for 28 days after injection. Behavioral testing indicated that all animals had normal withdrawal latencies to heat applied to the plantar surface of the hindpaw before treatment with capsaicin (14). In untreated rats, intraplantar injection of 10 μg of capsaicin produced nocifensive behavior for a duration of ∼3 min and produced about a 50% decrease in withdrawal latency to heat and a 40 to 60% increase in the frequency of withdrawal from the mechanical stimuli (15). Animals treated with SP-SAP exhibited a significant attenuation of mechanical (85% decrease at day 28) and heat (60% decrease at day 28) hyperalgesia produced by intraplantar injection of capsaicin (Fig. 4, B and C). Additionally, there was a marked reduction (75% decrease at day 28) in the nocifensive behavior induced by unilateral injection of capsaicin into the hindpaw at days 7, 14, and 28 after intrathecal treatment (Fig. 4A). In contrast, infusion of saline, SP, or SAP produced no significant change in mechanical or thermal hyperalgesia, or in nocifensive behavior produced by capsaicin as compared with normal untreated animals in any corresponding time point examined (Fig. 4).

Figure 4

Three behavioral parameters after intrathecal infusion of saline, SAP, or SP-SAP at day 0 (14). There was no significant difference between normal untreated animals, SP-infused animals, or saline-infused animals at any of the time points examined, and thus only the values for the saline-, SAP-, and SP-SAP–infused animals are shown. All animals (normal, saline-, SP-, SAP-, and SP-SAP–treated) had normal withdrawal latencies to heat applied to the plantar surface of the hindpaw before treatment with capsaicin (14). (A) Nocifensive behavior during the first 5 min after unilateral injection of capsaicin. (B andC) Thermal and mechanical hyperalgesia at 5 min after intraplantar injection of capsaicin. The triangles at time 0 are the baseline measurements for the saline-treated animals after injection of capsaicin. Data for each experimental group and at each time point were obtained from separate animals (n = 6 for each group) and represent withdrawal responses of the paw injected with capsaicin. All data points are expressed as the mean ± SEM, and significant differences were calculated by a one-way analysis of variance (ANOVA) and Bonferroni comparisons (*P < 0.01).

Although we ablated only the SPR-expressing neurons, which constitute less than 10% of all lamina I neurons (6,7), capsaicin-induced nocifensive behavior and mechanical and thermal hyperalgesia were depressed by 60 to 90% (Fig. 4). Our assumption is that intrathecal infusion of SP-SAP is more cytotoxic to SPR-expressing lamina I cells than to SPR-expressing cells in laminae III to V because further spread of bioactive SP-SAP into deeper laminae was prevented by degradation of the SP moiety by characterized proteases. One reason that ablation of such a small percentage of lamina I neurons could produce such a large change in behavioral nociceptive responses may be that, because most lamina I spinothalamic and spinoparabrachial neurons express SPR (8) and internalize SP-SAP, SP-SAP treatment is ablating a major part of the system for the ascending conduction of nociceptive information.

Hyperalgesia produced by capsaicin is mediated in part by sensitization of spinothalamic neurons, and SP is involved in the excitation and sensitization of spinothalamic neurons (16) and the development of hyperalgesia (17). However, it has been surprisingly difficult to block noxious stimulus-evoked pain behavior with either SP antagonists (18) or “knockout” of SPR in mice (19). In this study we did not block or inactivate only the SPR, but rather we killed a specific population of SPR-expressing cells that also express a variety of other neurotransmitter receptors (20). These data suggest that while this small population of SPR-expressing neurons is pivotal in the maintenance of hyperalgesia, a variety of other non-SPR receptors expressed by these neurons are also involved in nociceptive signaling. Understanding the repertoire of receptors that are expressed by this small population of SPR-immunoreactive lamina I neurons and how these receptors interact to generate persistent pain states should provide valuable information on how chronic pain states are generated, maintained, and potentially managed.

Whether receptor internalization can serve as a specific portal for introducing other therapeutic compounds into other receptor-bearing cells remains to be determined. This approach is, however, promising as a substantial number of other receptors have been shown to undergo ligand-induced receptor internalization (21). Specific targeting of spinal neurons involved in transmitting chronic as opposed to acute pain may have substantial therapeutic potential because most analgesics block both acute and chronic pain (22), and tolerance and dependence are major problems in long-term treatment with narcotics (23). Because the present findings suggest that the conduction of mild pain can be dissociated from highly noxious and hyperalgesic pain, SP-SAP treatment may be therapeutically useful in the treatment of persistent pain. However, before such therapies can be considered, the long-term consequences of removal of the superficial SPR-immunoreactive neurons must be defined.

  • * To whom correspondence should be addressed. E-mail: manty001{at}maroon.tc.umn.edu

REFERENCES AND NOTES

View Abstract

Navigate This Article