Distinct expression of spinal CRF-R1 and CRF-R2 receptors
Using a highly specific primer pair, both CRF-R1 and CRF-R2 mRNA were detectable in the dorsal part of the spinal cord (Fig. 1a). Concurrently, protein precipitation and separation of the dorsal part of the spinal cord (L4-L5) via gel electrophoresis and subsequent western blot revealed prominent CRF-R2 protein bands at the expected molecular weight (38 kDa), whereas CRF-R1 protein bands (56 kDa) were only faintly detectable (Fig. 1b).
Consistent with these findings, displacement experiments of radiolabeled [125J]-CRF-binding by CRF-R1- (NBI35965) and CRF-R2- (K41498) selective antagonists in membranes of the dorsal part of the spinal cord demonstrated clearly identifiable displacement with increasing concentrations of the CRF-R2-selective antagonist K41498 (IC50 = 500 pmol) (Fig. 1c) yet a lack of displacement with the CRF-R1 selective antagonist NBI35965 (Fig. 1c).
In cross-sectional spinal L4-L5 segments of naïve rats CRF-R2 immunoreactivity was strongly visible, whereas CRF-R1 immunoreactivity was only scarce (Fig. 2). CRF-R2 immunoreactivity was predominantly detectable in Rexed laminae I and II on both sides of the spinal dorsal horn (Fig. 2a) and in the immediate vicinity of the central canal (Fig. 2b), an area receiving descending projections from supraspinal nuclei [32], whereas CRF-R1 immuno-reactivity was not detectable in Rexed laminae I and II (Fig. 2d) and was only scarce around the central canal (Fig. 2e). CRF-R1 (Fig. 2f) as well as CRF-R2 (Fig. 2c) immunoreactivity, however, were clearly detectable in brain areas of known CRF-R1 expression such as hypothalamus confirming the specificity of the antibody [33, 34].
Distinct contribution of CRF-R2 and CRF-R1 receptors to the antinociceptive effects of intrathecal CRF in inflamed hindpaws
In Wistar rats with local hindpaw inflammation, i.t administration of increasing doses of CRF significantly and dose-dependently increased paw pressure thresholds (PPT), P < 0.001, one-way ANOVA and post-hoc Dunnett’s test) (Fig. 3a). This anti-nociceptive effect of i.t CRF was antagonized with increasing i.t doses of the CRF-R2 receptor selective antagonist K41498 (P < 0.001, one-way ANOVA and post-hoc Dunnett’s test) (Fig. 3b), but with the CRF-R1 receptor selective antagonist NBI35965 (P > 0.05, one-way ANOVA) (Fig. 3c).
To corroborate this finding, animals with hindpaw inflammation received the CRF-R2 selective agonist Ucn-2 intrathecally which also resulted in dose-dependent increased paw pressure thresholds (P < 0.001, one-way ANOVA and post-hoc Dunnett’s test) (Fig. 4a). As expected, this anti-nociceptive effect of i.t Ucn2 was antagonized by increasing i.t doses of the CRF-R2 receptor selective antagonist K41498 (P < 0.001, one-way ANOVA and post-hoc Dunnett’s test) confirming CRF-R2 receptor selectivity (Fig. 4b).
Intriguingly, i.t administration of either CRF (Fig. 5a) or Ucn-2 (Fig. 5b) together with the opioid receptor antagonist naloxone dose-dependently diminished the anti-hyperalgesic effect of both substances (P < 0.001, one-way ANOVA and post-hoc Dunnett’s test) indicating an opioid receptor-mediated effect.
Spinal cord areas of CRF receptor and opioid peptide co-expression
CRF-R2 immunoreactivity was found in the dorsal horn, the parasympathetic nucleus, and around the central canal of the spinal cord (Fig. 6). Double immunofluorescence confocal microscopy showed that CRF-R2 immunoreactivity within superficial laminae of the spinal dorsal horn predominantly overlapped with ENK-immunoreactivity (Fig. 7). Since most ENK-IR neurons within the dorsal horn characterize inhibitory interneurons [35–37], our results suggest that CRF-R2 was mainly expressed in inhibitory enkepalinergic interneurons of the dorsal horn of the spinal cord (Fig. 7). These ENK-ir spinal interneurons were found in close proximity of MOR derived from incoming CGRP-ir sensory neurons (Fig. 8).