In this study we describe a novel mouse fracture model that resembles CRPS in; 1) traumatic etiology and natural history, 2) vascular changes, 3) mechanical sensitivity and guarding, 4) up-regulation of cutaneous inflammatory mediators, and 5) regional periarticular bone loss. After a tibia fracture and 3 weeks cast immobilization the WT mice developed chronic allodynia, unweighting, warmth, and edema (Figure 2), and these vascular and nociceptive changes gradually resolved over a period of 16 weeks (Figure 3).
In an earlier investigation we tested the hypothesis that exaggerated neurogenic inflammatory responses contribute to the vascular and nociceptive sequelae of fracture in rats. When fracture rats were treated with a single-dose of an SP NK1 receptor antagonist there was a 51% reduction in hindpaw spontaneous protein extravasation, a 56% decrease in edema, and a 40% reduction in von Frey fiber allodynia . The results of the current study confirm and extend these findings. SP deficient (Tac1−/−) fracture mice had 54% less von Frey allodynia, 70% less hindpaw unweighting, and no hindpaw warmth or edema, compared to WT fracture mice at 3 weeks post-injury (Figure 2). Time-course studies demonstrated further differences in the magnitude and duration of post-fracture nociceptive behaviors in the WT and Tac1−/− mice. The WT mice developed nociceptive behaviors persisting for 8–15 weeks after fracture, while in the SP deficient mice nociceptive behaviors persisted for only 4–6 weeks post-fracture (Figure 3). These data confirm that SP is a critical regulator for the development of nociceptive and vascular sequelae after fracture. Similarly, CGRP receptor deficient (RAMP1−/−) fracture mice had attenuated allodynia and hindpaw unweighting, and no hindpaw warmth, but post-fracture edema was not affected by the loss of CGRP signaling (Figure 2). The contribution of SP and CGRP signaling to post-fracture hindpaw warmth and the essential role of SP signaling in the development of post-fracture hindpaw edema correlates with the results of intradermal microdialysis studies in human subjects demonstrating that SP or CGRP perfusion induces a vasodilation response in the skin, but only SP perfusion can evoke a protein extravasation response . In addition, the sciatic nerve electrically evoked protein extravasation response was dramatically reduced in Tac1−/− mice and unchanged in the RAMP1−/− mice (Figure 1C), indicating that CGRP signaling does not contribute to neurogenic extravasation in mice. Spontaneous protein extravasation is up-regulated in CRPS affected limbs , leading to increased interstitial fluid flow and distal limb edema, an important component of CRPS .
When SP or CGRP is microdialyzed through the skin of normal volunteers  or in CRPS skin  there is no immediate painful response, leading us to postulate that SP and CGRP act as intermediate mediators in the development of post-traumatic pain. Increased levels of the inflammatory cytokines TNFα and IL-6 are observed in experimental blister fluid or skin biopsies from the affected, but not the contralateral limbs of CRPS patients [23–25]. TNFα, IL-1b, IL-6, and NGF are up-regulated in the hindpaw skin at 4 weeks after tibia fracture in rats and fracture rats treated systemically with the TNF inhibitor etanercept, the IL-1 receptor antagonist anakinra, or the NGF antibody tanezumab had reduced allodynia and hindlimb unweighting, indicating an important role for cytokine and growth factor signaling in the development of trauma induced chronic pain [26, 27, 29].
The TNF, IL-1, and NGF inhibitors we have tested in the fracture rat model are large molecules that can’t cross the blood-brain barrier, suggesting that the pronociceptive effects of these inflammatory mediators probably occur at the nociceptor level in the sensitized hindpaw skin. Pro-inflammatory cytokines and NGF can immediately evoke spontaneous firing and sensitization in primary sensory afferents [32–34]. Furthermore, we have observed that intraplantar injection of these inflammatory mediators into normal hindpaw skin rapidly induces nociceptive sensitization and have identified keratinocytes as the primary cellular source of these inflammatory mediators in the fracture hindpaw (Figure 5) . Robust keratinocyte proliferation and epidermal thickening was also observed in the fracture hindpaw . Collectively, these data indicate that up-regulated keratinocyte proliferation and/or inflammatory mediator expression results in increased cutaneous inflammatory mediator levels in the fracture hindpaw, with subsequent nociceptive sensitization.
Neurotransmitter release evoked by capsaicin injection or electrical nerve stimulation stimulates the production of TNFα, IL-1β, IL-6, and NGF in the skin of rats, but it is unknown which specific transmitters mediate these inflammatory effects [35, 36]. Previous studies have observed that SP and CGRP signaling can stimulate the secretion of IL-1β and NGF from keratinocyte cell cultures, in vitro evidence that these neuropeptides can directly stimulate keratinocyte inflammatory mediator release [37–39].
Figure 4 illustrates the robust increase in TNFα, IL-1β, IL-6, and NGF protein levels observed in the hindpaw skin at 3 weeks post-fracture in the WT mice. Cytokines are normally expressed transiently in response to injury and infection and the persistent regional elevation of cytokines observed at 3 weeks post-fracture indicates ongoing cytokine synthesis by the keratinocytes (Figure 5), which we postulate is the result of exaggerated neuropeptide signaling. In support of this hypothesis, Tac1 or RAMP1 gene deletion prevented the up-regulation of TNFα, IL-1β, and NGF in the injured hindlimb (Figure 4). There were no differences between the WT, Tac1−/− and RAMP1−/− mice in baseline cutaneous cytokine and NGF levels, evidence that the low basal levels of cytokines and NGF expressed in normal skin are not dependent on neuropeptide signal.
Both the Tac1−/− and RAMP1−/− fracture mice had increased IL-6 levels in the hindpaw skin, and there was residual, albeit attenuated, allodynia and unweighting in these transgenic fracture mice compared to WT fracture mice (Figures 2,4). We postulated that IL-6 signaling contributes to post-fracture allodynia and unweighting and that the increase in skin IL-6 levels after fracture in the Tac−/− and RAMP1−/− mice is the mechanism for the residual post-fracture allodynia and hindpaw unweighting observed in these strains. To test this hypothesis we treated WT fracture mice with an IL-6 receptor antagonist (TB-2-081), and this treatment partially reversed fracture-induced allodynia and unweighting (Figure 6), supporting the hypothesis that increased cutaneous IL-6 signal contributes to allodynia and unweighting after fracture.
This study provides the first in vivo evidence that SP and CGRP act as intermediate mediators in the post-traumatic inflammatory cascade by chronically up-regulating inflammatory proteins that can directly sensitize cutaneous nociceptors. In contrast to other inflammatory mediators examined in this study, the post-fracture up-regulation of IL-6 did not require CGRP signaling and was only partially dependent on SP signaling. The significance of IL-6 signaling was confirmed by antagonist studies demonstrating its role in the maintenance of post-fracture nociceptive behavior. Collectively, these data suggest that neuronal regulation of innate immunity can mediate inflammatory changes observed in CRPS skin.
After tibia fracture WT mice developed trabecular, and to a lesser extent cortical bone loss in the ipsilateral femur when compared to unfractured controls (Figure 7, Table 1). These data are consistent with the results observed in the rat tibia fracture model , as well as in CRPS patients who develop periarticular bone loss after fracture . The net effects of SP signaling on bone acquisition and resorption in vivo are unknown, but previously we observed that the SP NK1 receptor is expressed in both bone marrow stromal cells and bone marrow macrophages and SP signaling stimulates bone marrow stromal cell osteogenic activity as well as macrophage osteoclast differentiation and resorption activity in vitro
. The WT and SP deficient (Tac1−/−) mice had similar baseline bone parameters and no post-fracture differences were observed in skeletal integrity between groups, indicating that SP signaling does not regulate post-traumatic bone loss in mice (Figure 7 and Table 1).
CRPS patients usually have deep tissue pain, primarily tenderness over the distal limb joints and pain with range of motion [15, 42]. It is unknown whether the neuroinflammatory changes observed in the fracture limb skin may also be occurring in the joints or muscles of the injured limb, but there are some intriguing clues. The most prominent histological finding in CRPS synovial biopsies is the proliferation of synovial lining cells . Synoviocytes express NK1 receptors and very low concentrations of SP can stimulate synoviocyte proliferation in vitro
[43, 44]. Stimulated synoviocytes express IL-6 and CCL2, a chemokine which induces monocytes to leave the bloodstream and enter the surrounding tissue to become synovial macrophages, the primary source of TNFα and IL-1β in rheumatoid joints . Examining neuroinflammatory articular changes after fracture could potentially identify signaling mechanisms contributing to the articular tenderness and joint motion pain associated with CRPS.