A cyclic peptide targeted against PSD-95 blocks central sensitization and attenuates thermal hyperalgesia

Abstract

Post-synaptic density protein PSD-95 is emerging as a valid target for modulating nociception in animal studies. Based on the key role of PSD-95 in neuronal plasticity and the maintenance of pain behavior, we predicted that CN2097, a peptide-based macrocycle of nine residues that binds to the PSD-95 Discs large, Zona occludens 1 (PDZ) domains of PSD-95, would interfere with physiologic phenomena in the spinal cord related to central sensitization. Furthermore, we tested whether spinal intrathecal injection of CN2097 attenuates thermal hyperalgesia in a rat model of sciatic neuropathy. Results demonstrate that spinal CN2097 reverses hyperexcitability of wide dynamic range (WDR) neurons in the dorsal horn of neuropathic rats and decreases their evoked responses to peripheral stimuli (brush, low caliber von Frey and pressure), whereas CN5125 ("negative control") has no effect. CN2097 also blocks C-fiber long-term potentiation (LTP) in the dorsal horn, which is linked to neuronal plasticity and central sensitization. At a molecular level, CN2097 attenuates the increase in phosphorylated p38 MAPK, a key intracellular signaling pathway in neuropathic pain. Moreover, spinal injection of CN2097 blocks thermal hyperalgesia in neuropathic rats. We conclude that CN2097 is a small molecule peptide with putative anti-nociceptive effects that modulates physiologic phenomena related to central sensitization under conditions of chronic pain.

Fig 1. Structure of cellular probe CN2097

Fig 2. PDZ domain array

PDZ domain array probed with the CN5115, a TMR-labeled analog containing the PDZ domain-targeting portion of CN2097 (Arrays provided by Dr. Randy Hall, Emory University).

Fig 3. CN2097 decreases the firing of WDR neurons

(a) Representative example showing extracellular firing recorded from a single unit in the dorsal horn (lamina III) one week after CCI, at a time of persistent hyperalgesia. Individual action potentials were recorded and used to generate peristimulus time histograms, which show reproducibility of firing and recording under stable conditions in consecutive recording sessions 10 and 20 min before CN2097 treatment (left panels). Although application of CN2097 does not change the rate of spontaneous firing, evoked responses to several stimuli (e.g. brush, pressure and pinch) decreased 10 and 20 min after treatment with CN2097 (right panels), in addition to attenuation in the sustained discharge within 40 sec after pinch (afterdischarge). Numbers show mean frequency of responses per stimulus (Br, vF, Pr, Pi denote brush, von Frey filaments 0.6, 8 and 15 g, pressure, and pinch). (b) Mean WDR response rates (n=6–8 units, 1 unit/rat d7 after CCI) 10 min before and 20 min after CN2097 application. Responses to brush, low caliber von Frey and pressure stimuli decreased significantly, whereas responses to other stimuli were not significantly changed (* p<0.05). Inset shows estimation of the location of these units in the dorsal horn (laminae I–V) ipsilateral to CCI.

Fig 4. CN5125 (‘negative control’) does not affect the firing of WDR neurons

(a) Top two traces show representative example of extracellular firing recorded from a single unit in the dorsal horn one week after CCI. Individual action potentials and corresponding peristimulus time histograms are shown at 10 before, and 20 min after CN5125 (10 nmol) treatment in the same unit recorded (Br, vF, Pr, Pi denote brush, von Frey filaments 0.6 g, 8 and 15 g, pressure, and pinch). (b) Mean WDR response rates (n=5 units, 1 unit/rat d7 after CCI) 10 min before and 20 min after CN5125 application (10 nmol). Responses to stimuli were not significantly different.

Fig 5

(a) C-fiber field potentials (stimulus artifact and A-fiber field potential truncated, scale 150 ms). (b) CN2097 reduces C-fiber field potentials after conditioning stimuli. Traces show baseline C-fiber field potentials (before conditioning stimuli or TS) and 10 after TS. Note stable baseline and increased responses after TS in control rat (vehicle-treated, left panels) indicating induction of LTP compared to no change before and after TS in CN2097-treated rat (right panels).

Fig 6. CN2097 decreases the area, but not the latency of C-fiber field potentials

Mean C-fiber field potential areas and latencies were normalized to baseline (3 min before TS). Upper panels show timeline of evoked C-fiber field potentials before and after TS at t=0 (vertical dashed line). Note attenuated elevation in area in CN2097-teated rats compared to control at every time point after onset of TS (n=3 rats/group). Lower panels show values at time points corresponding to baseline and 10 min after TS. Histograms show that, in control group, areas significantly increase after TS (* p<0.05), whereas areas in CN2097-treated rats remain non-significantly different from baseline.

Fig 7. CN2097 reverses increased levels of phospho-p38 MAPK in CCI

Freshly isolated spinal cords from vehicle (aCSF) or CN2097-treated 7-day CCI rats were probed for phosphorylated and total p38 MAPK. (a) Representative immunoblots: Total and phosphorylated p38 MAPK after 15 minutes spinal treatment with CN2097. Western blots shown as representative of three independent experiments (n=6 rats). (b) Densitometric analysis: Ratio of phosphorylated p38 MAPK: total p38 MAPK shown as arbitrary units (* p<0.05).

Fig 8. CN2097 attenuates thermal hyperalgesia

Compared to baseline, paw withdrawal latencies (PWL) decrease significantly at d7 after CCI. (a) A single injection of CN2097 (10 nmol, i.t.) significantly increases PWLs in rats with CCI for at least 2 hrs post-injection (* p<0.05), demonstrating reversal of thermal hyperalgesia and suggesting that CN2097 is biologically active centrally. However, 1 nmol i.t. (b) and 200 nmol i.p. (c) had no effect on withdrawal latency in rats with CCI, and (d) 1 nmol i.t. did not have an affect in naïve rats. (e) Evans blue dye diffused throughout the spinal cord with maximal dye intensity (100%) within the vicinity of the distal tip of the catheter (L4-5) and decreasing diffusion gradients towards upper cervical levels (C6-7). Photo showing one cell in the spinal cord (arrows) with uptake of the TMR-labeled peptide. Labeling was typically vesicular in appearance and localized within the cytoplasm surrounding nuclear profiles in the grey matter. In general, the variance of soma diameters of the labeled cells indicates that more than one subtype of cells may accumulate the cyclic peptide, and that this modified peptide is effective in crossing the pia/brain barrier. White matter was mostly devoid of labeling (scale 35 µm).