In the present study, it is interesting to note that antagonism of CXCR4 by bath perfusion of DRG with AMD3100 results in significant inhibition of the BV-enhanced neuronal firing rate by reversing the lowered rheobase value to the normal level. This result suggests a maintaining role of SDF1–CXCR4 signaling in the primary nociceptor hyperexcitability at the neuronal cell body level caused by peripheral inflammatory pain state. The primary nociceptor neuronal cell body hyperexcitability caused by the BV-induced peripheral inflammatory pain state has already been demonstrated in our previous reports [
28,
29,
40,
41]. Regarding this, one question may arise to be asked: which cell types are possible sources of SDF1 and CXCR4 in the DRG? To answer this question, we further demonstrated that, by double immunofluorescent staining and/or Western blotting, SDF1 was exclusively up-regulated in GFAP-positive non-neuronal SGCs of the DRG following i.pl. BV injection, while CXCR4 was mainly co-localized with IB4, SP, and TRPV1, specific biomarkers of primary nociceptor neurons in the DRG. This at least suggests that the non-neuronal SGCs can be activated by peripheral inflammatory pain state that serve as a source of SDF1 release, allowing its binding to CXCR4 be over expressed in the primary nociceptor neuronal cell body caused by persistent firing initiated in the peripheral terminals [
33,
45‐
47]. The mechanism of the activation of non-neuronal SGCs may be due to the fractalkines released from DRG neurons which has been demonstrated by Souza and colleagues in i.pl. carrageenan injection-induced inflammatory pain model [
48]. It has been demonstrated that, under physiological state, both primary afferent neurons and non-neuronal SGCs in the DRG constitutively contain SDF1 that is maintained at a very low level [
9]. However, the source of SDF1 under pathological level may vary depending upon different conditions. In antiretroviral toxic neuropathy model, Bhangoo and colleagues have observed up-regulation of SDF1 and CXCR4 mRNA at 7 and 14 days after administration of antiretroviral drug 2, 3-dideoxycytidine, and the enhanced expression of SDF1 was mostly observed in the DRG non-neuronal cells [
49]. Similarly, Dubovy and colleagues have also demonstrated that the non-neuronal SGCs were the source of SDF1 by showing co-localization of SDF1 and glutamine synthetase in the DRG of animals with neuropathic pain induced by CCI [
15]. Based upon the above lines of evidence, it is highly suggested that the inducible release of SDF1 from non-neuronal SGCs by peripheral inflammatory and neuropathic pain conditions should result in development of intraganglionar inflammatory microenvironment, by which long-term hyperexcitability of primary nociceptor neurons can be maintained. As lines of supporting evidence, it has been demonstrated that the SGCs in the DRG when they had been activated by i.pl. carrageenan injection could produce some pro-inflammatory mediators (TNFα, IL-1β, and prostanoids) that can directly excite the primary nociceptive neurons [
48]. Moreover, SDF1 has been shown to significantly increase intracellular calcium concentrations in the isolated DRG neurons under several persistent pain conditions [
49‐
51]. It has also been demonstrated that exogenous SDF1 could lower the threshold for action potential generation and depolarize nociceptive DRG neurons in cultured DRG neurons [
21]. However, sensory neurons of the DRG may also be the source of SDF1 since increase in SDF1 mRNA or protein have been observed in DRG neurons of transgenic mice with high-fat diet-induced type II diabetic neuropathy [
50] and animals receiving repeated morphine treatment [
51]. Besides SDF1, other chemokines are also likely to be involved in primary sensory neuronal hyperexcitability, because up-regulation of CCL2, CCL3, and CXCL1 by pain and direct activation of DRG neurons by them have been seen in several previous reports [
6,
30,
31,
52,
53].
As for the source of CXCR4, a selective cognate receptor of SDF1, only primary sensory neurons have been targeted. Consistent with some previous reports [
9,
16,
21], in our present study, CXCR4 was constitutively present but overexpressed under peripheral inflammatory pain state in both non-peptidergic (IB4-positive) and peptidergic (SP-positive) primary nociceptor neurons. It is also co-localized with TRPV1, a thermonociceptor of primary sensory afferent. However, the overexpression of CXCR4 induced by intraplantar BV injection could be significantly prevented by pre-treatment with i.pl. AMD3100, a selective CXCR4 inhibitor, suggesting that the BV-induced overexpression of CXCR4 in the DRG cells may be partially mediated by peripheral SDF1–CXCR4 signaling in the skin which was trafficking from neuronal soma to the peripheral terminals. In the case of SDF1–CXCR4 signaling pathway, Reaux-Le and colleagues have found that CXCR4 receptor was constitutively present in both peptidergic (CGRP positive) and non-peptidergic (IB4 positive) DRG neuronal soma in rats [
9]. They have also demonstrated that CXCR4 can be localized in pre-synaptic components of both type I and type II glomeruli in the spinal dorsa horn under electron microscope, indicating that CXCR4 could be axonally transported to both peripheral and central terminals of the primary afferent neurons in DRG and exert its functions [
9]. Although we do not detect the protein level of CXCR4 in the skin directly in the current study, Reaux-Le and colleagues’ results can lend support to the rationale of peripheral administration of AMD3100 in our current study because CXCR4 immunoreactivities co-expressed in the CGRP-positive fibers have been shown to be present in the glabrous skin as well as in the DRG cells and project to the dermis [
9]. The functional nature of this increased CXCR4 receptor expression was identified by our behavioral pharmacology assays in which pre-treatment with i.pl. AMD3100 significantly prevented the development of the BV-induced persistent spontaneous pain-related behaviors and pain hypersensitivity. Moreover, bath perfusion of medium- and small-sized DRG neurons with AMD3100 also significantly reduced BV-induced tonic discharges by restoration of rheobase value to normal level, suggesting a maintaining role of SDF1–CXCR4 signaling in the primary nociceptor hyperexcitability. Since it is known that the decrease in rheobase value may reflect changes in persistent Na + conductance [
54,
55], the roles of TTX-resistant VGSC α subunits Nav1.8 and/or Nav1.9, which are selectively expressed in the primary nociceptor neurons, should be further investigated. As aforementioned, we have already demonstrated that Nav1.8 and Nav1.9 could be up-regulated in the small and medium-sized DRG cells by i.pl. melittin, the major toxin of whole bee venom, or CFA injection, resulting in increased persistent current mediated by Nav1.8 and Nav1.9 and enhanced firing rate of tonic, but not phasic, type of primary nociceptor neurons with lowered rheobase value [
28,
29,
41]. Down-regulation of Nav1.8 and Nav1.9 by anti-sense oligodeoxynucleotide, respectively, resulted in restoration of Nav1.8 and Nav1.9 current density and rheobase value, leading to reduction of hyperexcitability of primary nociceptor neurons and relief of persistent nociception and pain hypersensitivity induced by i.pl. injection of melittin or CFA [
28,
29]. In the present, our behavioral data also demonstrated that peripheral pre-treatment with A-803467, a selective Nav1.8 blocker, could inhibit the persistent nociception and reverse the primary pain hypersensitivity induced by i.pl. BV injection, indicating that increased expression of Nav1.8 in the DRG induced by i.pl. BV injection contributed to the development and maintenance of the BV-induced pain-related behaviors. Taken together, we proposed that there must be a functional link between SDF1–CXCR4 signaling and expression of Nav1.8 or Nav1.9 that maintains primary nociceptor neuronal hyperexcitability under peripheral inflammatory pain state.