It is known that ChemR23 agonists can alleviate hypersensitivity in animal models of inflammatory pain through both peripheral and central mechanisms
. These include a reduction in peripheral inflammation and a normalisation of potentiated spinal cord responses, respectively. In terms of central mechanisms, electrophysiological recordings from unidentified lamina II dorsal horn neurons have demonstrated that the ChemR23 agonists, RvE1 and chemerin, attenuate capsaicin potentiation of sEPSC frequency. However, as these findings were obtained using sEPSC recordings in unidentified neurons it is not known where in the spinal cord network or upon which neuronal subtypes that these effects are mediated. In this study we have investigated the ability of chemerin to modulate excitatory input to lamina I NK1R+ neurons following capsaicin potentiation and CFA inflammation. Our results have novelly revealed that chemerin can attenuate capsaicin potentiation of mEPSC frequency in lamina I NK1R+ neurons and presynaptically reduce monosynaptic C-fibre input to a subset of these neurons in inflammatory pain. Notably, chemerin was without effect in non-potentiated/control conditions. Given the essential role of lamina I NK1R+ neurons in the manifestation of inflammatory pain
, which is driven by C-fibres
, the chemerin attenuation of monosynaptic C-fibre input to these neurons suggests that the reported ability of ChemR23 agonists to attenuate inflammatory pain hypersensitivity may in part be due to presynaptic inhibition of monosynaptic C-fibre input to these key spinal cord output neurons.
We have demonstrated that chemerin can significantly reduce, but not eliminate, capsaicin potentiation of mEPSC frequency in rat lamina I NK1R+ neurons. However, Xu et al.
 report that chemerin, at the same dose used here, can completely prevent capsaicin potentiation of sEPSC frequency in unidentified mouse lamina II neurons. This difference likely reflects the different concentrations of capsaicin employed, 1 μ M in the present study, which resulted in a ∼44-fold increase in frequency, vs. 100 nM used by Xu et al.
, which increased frequency by only 2-fold. We did investigate use of 100 nM capsaicin, but found that this concentration did not reliably potentiate mEPSC frequency in lamina I NK1R+ neurons (data not shown). This dissimilarity may also reflect the different species employed, rat vs. mouse, different cell types targeted, lamina I NK1R+ neurons vs. unidentified lamina II neurons or the different recording approach, mEPSC vs. sEPSC recording, employed. Other groups have demonstrated that 100 nM capsaicin can potentiate mEPSC frequency in unidentified lamina I/II neurons
[31, 32], however potentiation of mEPSCs in lamina I NK1R+ neurons has only been reported using 1 μ M capsaicin
. Interestingly, Labrakakis and MacDermott
 report that 73% of lamina I NK1R+ neurons display an increase in mEPSC frequency in response to 1 μ M capsaicin, which is comparable to our findings, that 83% of neurons had capsaicin-sensitive input.
On the basis of anatomical expression data
 we hypothesised that ChemR23 would be expressed on a subset of monosynaptic C-fibre inputs to lamina I NK1R+ neurons. Indeed we provide supporting functional evidence for this expression pattern through our novel demonstration that chemerin presynaptically attenuates monosynaptic C-fibre input to a subset of these neurons in inflammatory pain. To unequivocally demonstrate that the chemerin effects reported here are mediated via ChemR23, we would ideally have shown blockade by a ChemR23 antagonist, but this was not possible as no such ligand is commercially available. Pertussis toxin (PTX), an inhibitor of G
i coupled GPCRs, the receptor family to which ChemR23 belongs, has been used by others to inhibit the RvE1 attenuation of capsaicin potentiated input in unidentified lamina II neurons
. While this approach could have been used to provide additional confirmation that the actions of chemerin were mediated by ChemR23, PTX inhibition of the chemerin response would only indicate that chemerin acted via a G
i coupled GPCR and not ChemR23 specifically.
ChemR23 agonists are proposed to reduce inflammatory pain in part by normalising potentiated spinal cord responses
[14, 33]. In the present data, chemerin reduced monosynaptic C-fibre input to a subset (∼44%) of lamina I NK1R+ neurons in CFA tissue but was without effect in control tissue. Interestingly, electrical stimulation of monosynaptic C-fibre input to lamina I projection neurons, that are likely to be NK1R+ neurons, in a manner which mimics the spontaneous firing pattern seen during inflammatory pain, results in the potentiation of C-fibre input to only a subset of these neurons
. It could therefore be hypothesised that neurons classified as chemerin responders were the subset that received potentiated C-fibre input. We therefore investigated whether there was a correlation between the initial peak amplitude of monosynaptic C-fibre eEPSCs (with greater amplitudes possibly signifying potentiated inputs) and the magnitude of the chemerin mediated change in peak amplitude. Our results found there to be no correlation between the initial eEPSC peak amplitude and the amplitude change (data not shown), however it should be recognised that this finding cannot confirm or refute the possibility that neurons classified as chemerin responders were those that received potentiated input, as it is not possible to directly distinguish potentiated inputs in this kind of population comparison study. Moreover, when the peak amplitude of monosynaptic C-fibre eEPSCs are compared between control and CFA tissue there is no significant potentiation observed overall (data not shown) as previously reported
It has previously been established that ChemR23 agonists do not alter acute pain sensitivity and have no effect upon sEPSC frequency, in unidentified lamina II neurons, in non-potentiated conditions
. In accordance, we have shown that chemerin does not alter basal mEPSC frequency or amplitude in lamina I NK1R+ neurons or the peak amplitude of monosynaptic C-fibre eEPCSs in these neurons in control tissue. It is proposed that ChemR23 activation reduces inflammatory pain hypersensitivity by normalising potentiated spinal cord responses, as opposed to a general reduction in sensory transmission
[14, 33], for which our findings provide additional support.
ChemR23 activation is proposed to normalise spinal cord inflammatory pain potentiation by inhibition of the extracellular signal-regulated kinase (ERK) pathway, which is key for central sensitisation
, both presynaptically in the central terminals of primary afferents and postsynaptically in dorsal horn neurons
. Inhibition of ERK, with the mitogen-activated protein kinase kinase inhibitors PD98059 and U0126, prevents capsaicin potentiation of sEPSCs, while application of RvE1 reduces capsaicin/TNF- α driven ERK phosphorylation in DRG cultures
. Therefore, the chemerin attenuation of capsaicin potentiated mEPSC frequency and chemerin mediated presynaptic inhibition of monosynaptic C-fibre inputs in inflammatory pain, reported here, may result from blockade of ERK mediated glutamate release from central terminals. Chemerin may also mediate its effects via inhibition of TRPV1, which is crucial for inflammatory pain
[35, 36], as RvE1 is a highly potent inhibitor of TRPV1, with different resolvins interestingly differentially modulating different TRP channels via GPCR activation
Chemerin is a natural ChemR23 ligand
[15, 16] and while it is currently unclear which cell types are responsible for its endogenous production and release, endothelial cells, keratinocytes, chondrocytes, platelets and osteoclasts have all been proposed as possible sources
[39–44]. Chemerin plays a key role in a number of physiological processes including adipocyte generation and metabolism
 and the chemotaxis of macrophages and dendritic cells
. However, it is not currently known whether endogenous chemerin plays any role in the modulation of inflammatory pain.
It is worth noting that chemerin is known to display high affinity binding with receptors other than ChemR23, namely the chemokine (C-C motif) receptor-like 2 (CCRL2) and G protein-coupled receptor 1 (GPR1), however these receptors are thought to play a limited role in cell signalling. Current evidence suggests that CCRL2 is not involved in cell signalling but may play a functional role in presenting chemerin to ChemR23
[40, 46, 47]. Interestingly, CCRL2 is known to be expressed in the spinal cord and wider CNS, predominantly in microglia
. Binding of chemerin to GPR1 results in receptor internalisation and signalling, but this signalling is weak
 and it has been speculated that GPR1 may act as a decoy receptor
. There is evidence that GPR1 is expressed in the CNS
[50–53], however spinal cord expression has not been investigated. It is possible, therefore, that inflammation-induced changes in the expression of either of these receptors could have influenced the results presented here.
While the evidence presented here supports the work by Xu et al.
 which revealed that drugs which target ChemR23 may be effective in the treatment of inflammatory pain, it is worth noting that chemerin or RvE1 may have limited therapeutic potential given that they are not metabolically stable and are rapidly inactivated in vivo
[33, 54, 55]. Interestingly, a stable analogue of RvE1, 19-(p-fluorophenoxy)-RvE1, has been shown to effectively reduce thermal hypersensitivity in the CFA inflammatory pain model for an extended time period compared to RvE1
. Stable chemerin analogues have been developed
, however their use has not yet been investigated in models of inflammatory pain. Further research into the use of these or new stable analogues of ChemR23 agonists should further establish ChemR23 as a promising target for the treatment of inflammatory pain.