This study describes a unique approach to the neurobiology of the resolution of clinical inflammatory pain. The model of inflammatory pain used in this study, has been described previously and is an excellent model of naturally-occurring pain [1, 2]. Strict inclusion criteria were adhered to and from clinical examination animals recruited to the study were estimated to have had the infection for more than two weeks. Therefore, this model is representative of persistent pain. The condition 'footrot' is initiated by invasion of bacteria into the interdigital tissues inducing dermatitis, and as infection spreads the bacteria penetrate further into deep tissues causing extensive tissue damage and separation of the hoof horn, accompanied by inflammation and pain . To contain the clinical heterogeneity of diseased animals, only sheep with lameness scores ≥ 5 and pathology scores ≥ 2, were recruited. As expected, animals affected with unilateral inflammation displayed significant lameness (taken as a proxy indicator of spontaneous pain) and hyperalgesia, restricted to the inflamed limb. Previous studies in this laboratory revealed a significant relationship between lameness score and mechanical hyperalgesia, supporting a positive association between the magnitude of hyperalgesia and the intensity of inflammatory pain . The presence of mechanical hyperalgesia localised to the inflamed limb indicates an alteration in sensory nerve function and nociceptive processing occurring centrally, as a consequence of release of pro-inflammatory mediators at the site of inflammation and at spinal cord level. Previous studies have shown that parenteral and topical antibacterial treatment produces rapid recovery (within 5 days of treatment) from foot lesions and lameness in sheep with footrot . In the present study, antiseptic footbathing, trimming the hoof horn and parental antibiotics rapidly reversed both lameness and hyperalgesia, with more than 85% of animals showing significant improvements 3 days after treatment, despite the persistence of foot pathology and lymph node hyperplasia. These findings indicate that treating the primary source of infection and therefore reducing inflammation is effective and sufficient to reverse persistent pain hypersensitivity in this clinical model.
The bilateral induction of COX-2 protein in spinal cord from sheep with inflammation and subsequent decrease in expression 3 days after treatment as lameness and hyperalgesia were resolving, supports the theory derived from experimental models that that spinal COX-2 derived PGs play a key role in central mechanisms of hyperalgesia in clinical pain [33–36]. A bilateral and inter-segmental induction of spinal COX-2 in response to unilateral inflammation induced by injection of complete Freund's adjuvant (CFA) into the footpad in rat has been described previously [12, 15], suggested to be due to a humoral response induced by circulating pro-inflammatory mediators such as IL-1β. A humoral immune response could account for the bilateral increase in COX-2 protein in spinal cord observed in the present study, however, no corresponding increase in levels of COX-2 were detected in spinal cord segments distal to the inflamed limb afferent termination site (data not presented). It is likely that increased noxious input from activated primary afferents from the inflamed limb, driving activity of spinal dorsal horn neurons, contribute to induction of COX-2, perhaps through activation of substance P, which is known to lead to increased expression of COX-2 and release of PGE2 in spinal cord . Induction of COX-2 in spinal cord appears to be transient in most rodent models of inflammation, for instance, peaking 6 hours after hindpaw incision or intraplantar injection of carrageenan , and between 6-12 hours in response to injection of CFA into the footpad; returning to baseline levels a few days later [11, 12, 15], even though hyperalgesia and paw inflammation persists for several days beyond this period in this model [12, 39, 40]. These studies indicate that COX-2 in spinal cord plays key role in initiating central sensitization with acute inflammation. The elevated levels of COX-2 protein observed in spinal cord from sheep with persistent inflammatory disease, however, suggests that COX-2 derived PGs also contribute to maintaining the central component of hyperalgesia in this model. This is supported by a recent study by Prochazkova et al.  showing long-lasting up-regulation of COX-2 mRNA and protein in spinal cord in a model of osteoarthritis and persistent hyperalgesia.
Biosynthesis of Egr-1 mRNA and protein was increased in ipsilateral spinal cord from sheep with unilateral inflammation, also likely due to increased input from activated nociceptors in the inflamed limb. Egr-1 is primarily expressed in superficial dorsal horn neurons that receive input from small-diameter myelinated and unmyelinated afferent fibres [22, 25], and expression is rapidly induced in these laminae in response to a variety of acute experimental inflammatory stimuli including intraplantar injection of formalin [24, 26, 27] and carrageenan , and nerve-injury . Furthermore, the temporal pattern of Egr-1 expression is reported to parallel development of hyperalgesia and inflammation [23, 25], suggesting a link between Egr-1 expression and behavioural responses to inflammatory pain. This hypothesis was further strengthened by a study in Egr-1 knockout mice by Ko et al. , that reported that although acute nociception is unaltered, hypersensitivity induced by formalin or CFA was diminished in these mice. A role for spinal Egr-1 in the maintenance of inflammatory pain is supported by evidence that Egr-1 antisense treatment in rat resulted in deficits in the maintenance of mechanical allodynia . In the present study, the return of Egr-1 mRNA levels to baseline levels and decrease in protein expression 3 days after treatment, as lameness and hyperalgesia were resolving, further strengthens a role for Egr-1 in central neuronal plasticity underlying inflammatory or persistent pain.
Egr-1 activates a variety of downstream target genes by binding to the DNA sequence GCG(G/T)GGCG in the genes promoter region, including microsomal prostaglandin E synthase-1 (mPGES-1), the enzyme that couples with COX-2 to produce PGE2 , and COX-2 , suggesting an interaction between Egr-1 regulated transcription and inflammatory prostanoid production. This hypothesis is supported by evidence that inhibition of Egr-1 results in decreased COX-2 and mPGES-1 expression in LPS-stimulated murine macrophages , and induction of mPGES-1 mRNA following injurious ventilation in lung tissues is diminished in Egr-1 knockout mice . Inflammatory prostaglandins also seem to play an important regulatory upstream role, acting as initiators of Egr-1 activity. For instance, treatment of stimulated endothelial cells with a COX-2 selective inhibitor was reported to inhibit production of Egr-1 mRNA and protein , while PGE2 induces Egr-1 mRNA expression cementoblastic OCCM periodontal cells . The co-induction of COX-2 and Egr-1 in spinal cord from sheep with unilateral inflammation and hyperalgesia supports a link between the COX-2 signalling pathway and enhanced Egr-1 gene transcription, and may underlie persistent cell modifications at the spinal level and persistence of pain.