The first neuron in the pain pathway, the primary afferent nociceptor, is an extremely important target for the development of novel pain therapeutics since: 1) nociceptors contain functionally important molecules not found in other cells (e.g., voltage-gated sodium channel, NaV1.8 (
Akopian et al., 1996,
Sangameswaran et al., 1996,
Renganathan et al., 2003)), 2) only a subpopulation of nociceptors may be involved in a given pain syndrome (
Joseph et al., 2008,
Joseph and Levine, 2010), which might allow for preservation of protective pain sensation, 3) analgesics working at this point in the pain pathway (i.e., in the primary afferent nociceptor) act before pain signals enter the central nervous system to diverge over multiple pathways (
Braz et al., 2005,
Bráz and Basbaum, 2009), 4) peripherally restricted analgesics avoid their many CNS-related side effects, and 5) blocking the pain signal in the periphery would prevent development of neuroplastic changes in the central nervous system (
Dubner, 2004,
Salter, 2005,
Sharif Naeini et al., 2005,
Eisenach, 2006,
May, 2008,
Descalzi et al., 2009,
Latremoliere and Woolf, 2009,
Seifert and Maihöfner, 2009,
Toyoda et al., 2009,
Asante et al., 2010).
The cardinal symptom in most pain syndromes is sensitivity to mechanical stimuli, pain made worse by movement or previous innocuous stimuli (e.g., mechanical hyperalgesia, mechanical allodynia, tenderness). Therefore, the present study focused on populations of nociceptors involved in mechanical hyperalgesia. We first determined which population(s) of nociceptors contributed to mechanical hyperalgesia. Our subdivision of nociceptors is based on the dorsal horn lamina in which they terminate. IB4(+) nociceptors terminate in lamina IIi (
Kitchener et al., 1993,
Molliver et al., 1995,
Gerke and Plenderleith, 2004), while IB4(−) (peptidergic) nociceptors terminate in lamina IIo and I (
Molliver et al., 1995). It has been suggested that these two populations of nociceptors contribute to different aspects of pain sensation (
Braz et al., 2005). In this study we injected NGF and GDNF, ligands for the TrkA and Ret receptors, respectively, which produce hyperalgesia by direct action on nociceptor terminals (
Malik-Hall et al., 2005,
Bogen et al., 2008) and are found on different populations of nociceptors (
Kashiba et al., 1998). In the rat Ret is only found on IB4(+) nociceptors while TrkA is on both IB4(+) and IB4(−) nociceptors (
Kashiba et al., 2001). We observed that both GDNF and NGF produce mechanical hyperalgesia. However, in the rat approximately one-third of TrkA(+) nociceptors are IB4(+). Therefore, to confirm that IB4(−) nociceptors contribute to mechanical hyperalgesia, which has recently been brought into question, in the mouse (
Basbaum et al., 2009), we demonstrated that a potent IB4(+) nociceptor toxin (
Vulchanova et al., 2001,
Tarpley et al., 2004) eliminated GDNF- but not NGF-induced
mechanical hyperalgesia. Thus, in contrast to what has been suggested for the mouse (
Basbaum et al., 2009), in the rat both the IB4(+) and IB4(−) nociceptors play a role in mechanical nociception and mechanical hyperalgesia. These observations may have clinical implications. For example, we have previously shown that IB4-saporin eliminates mechanical hyperalgesia in the early phase of the painful peripheral neuropathy produced by Oxaliplatin (
Joseph et al., 2008), a clinically distinct phase of the painful peripheral neuropathy associated with this important cancer chemotherapeutic drug (
Joseph et al., 2008). IB4-saporin also prevents development of hyperalgesic priming (
Joseph and Levine, 2010) a neuroplastic change in nociceptors leading to enhanced and markedly prolonged inflammatory hyperalgesia. Unfortunately, few models of clinical pain syndromes have been studied in terms of the population(s) of nociceptors involved.
A second important nociceptor mechanism, with well-established clinical significance, is as a site of action of opioid analgesics (
Aley and Levine, 1997c,
Walwyn et al., 2007,
Scherrer et al., 2009,
Stein and Zöllner, 2009,
Walwyn et al., 2009). While the major analgesic effect of opioids is likely mediated by its action at the central terminal, we have used the peripheral terminal to distinguish the effect of DOR and MOR agonists on mechanical hyperalgesia in populations of nociceptors. This approach also allowed us to exclude action on opioid receptors in neurons intrinsic to the spinal dorsal horn (
Zajac et al., 1989,
Besse et al., 1990a,
b,
Gillberg and Askmark, 1991,
Morinville et al., 2004,
Kline and Wiley, 2008). Since centrally but not peripherally administered opioids elevate nociceptive threshold in the absence of mechanical hyperalgesia, it also allowed us to exclude actions other than reversal of NGF- or GDNF-induced hyperalgesia. We found that a DOR but not a MOR agonist produced analgesia by action on IB4(+)/Ret(+) nociceptors (i.e., for GDNF hyperalgesia), similar to what has recently been reported in the mouse (
Scherrer et al., 2009). In contrast, both DAMGO and SNC inhibited mechanical hyperalgesia induced by NGF in IB4(−)/TrkA(+) nociceptors (i.e., in IB4-saporin treated rats).
To test the hypothesis that mu and delta opioid receptors are co-expressed on nociceptors, we performed two experiments. First, we tested the hypothesis that the combination of low doses of mu and delta opioid agonists would interact to produce enhanced analgesia. In these experiments we found that the combination of low doses of DAMGO and SNC produce markedly greater analgesia for NGF, but not GDNF, hyperalgesia. These findings suggest that the action of DAMGO is restricted to MOR, with no action at DOR. Thus, the interaction between DAMGO and SNC, to produce enhanced analgesia, is likely due to an interaction between MOR and DOR signaling. To more directly test the hypothesis that MOR and DOR signaling occur in the same neuron, we show that for NGF hyperalgesia, there is cross-tolerance between DAMGO- and SNC-induced analgesia. That is, repeated administration of DAMGO produced cross-tolerance to SNC. Thus, we conclude that MORs and DORs are co-expressed in a functionally important population of TrkA(+) nociceptors. These findings are compatible with previous studies providing evidence that, mu and delta opioid receptors dimerize (
Lee et al., 1980,
Schiller et al., 1999,
Jordan et al., 2001,
Gomes et al., 2004,
Daniels et al., 2005), DOR modulates MOR analgesia (
Standifer et al., 1994,
Schiller et al., 1999,
Zhu et al., 1999,
Nitsche et al., 2002,
Gomes et al., 2004,
Fan et al., 2005,
Gallantine and Meert, 2005,
Chefer and Shippenberg, 2009,
Xie et al., 2009), and MOR and DOR are co-expressed in DRG neurons (
Ji et al., 1995,
Rau et al., 2005,
Wang et al., 2010). Using IB4-saporin to destroy IB4(+) nociceptors, we were able to show that this effect included action on IB4(−)/TrkA(+) nociceptors. These findings differ from those of Scherrer and colleagues (
Scherrer et al., 2009), who found in the mouse that IB4(+)/nonpeptidergic but not IB4(−)/peptidergic neurons mediate mechanical pain, and DOR (SNC) but not MOR (DAMGO) agonists, administered intrathecally produce analgesia against mechanical pain (
Scherrer et al., 2009).
Differences between the results in the present study and that of Scherrer and colleagues might be due, in part, to use of different species. Functional differences between the central and peripheral terminals of the nociceptor might also contribute, as Scherrer and colleagues used spinal and we intradermal administration of opioid agonists, since dorsal horn neurons also contain opioid receptors (
Zajac et al., 1989,
Besse et al., 1990a,
b,
Gillberg and Askmark, 1991,
Morinville et al., 2004,
Kline and Wiley, 2008). Also, differences in the pain models used, such as the use of the sensitized nociceptor in all the present experiments, may be important as CNS opioids have effects on nociceptive threshold in the absence of a sensitization state. Finally, use of VR1 as a selective marker for thermal nociceptors may also impact interpretation since VR1 may also function in mechanical transduction (
Gevaert et al., 2007,
Liedtke, 2007b,
Liedtke, 2007a,
Pedersen and Nilius, 2007,
Bielefeldt and Davis, 2008,
Yin and Kuebler, 2010). Of note, our findings in the rat are in agreement with a recent study in the mouse that reported coexpression of MOR and DOR on small-diameter dorsal root ganglion neurons, predominantly in IB4(+) population (
Ji et al., 1995,
Rau et al., 2005,
Wang et al., 2010). Differences between rats and mice, with respect to which nociceptors mediate mechanotransduction, and MOR and DOR agonist-induced analgesia, will require considerable additional studies in both species.
In conclusion, in this study we have found that intradermal injection of both endogenous Ret ligand GDNF, and TrkA ligand NGF, present on distinct populations of nociceptors, both produce
mechanical hyperalgesia. DOR agonist SNC but not MOR agonist DAMGO inhibits GDNF-induced hyperalgesia while both DAMGO and SNC both inhibit NGF hyperalgesia, even in rats pretreated with IB4-saporin. Co-administration of low doses of DAMGO and SNC produce marked analgesia, and repeatedly administered DAMGO produced cross-tolerance to the acute peripheral analgesic effect of SNC. These findings demonstrate that most nociceptors have a role in mechanical hyperalgesia, only the DOR agonist inhibits GDNF hyperalgesia mediated by IB4(+)/Ret(+) nociceptors, and MOR and DOR are co-expressed on a functionally important population of OB4(−)/TrkA(+) nociceptors. Since human DRG neurons do not bind IB4, the current use of NGF and GDNF to distinguish between physiologically important subpopulations of nociceptors, which have similar distribution to that in the rat (
Wetmore and Olson, 1995,
Josephson et al., 2001), may also allow more effective comparison of pain in preclinical models to pain syndromes in patients.