PKMζ is essential for spinal plasticity underlying the maintenance of persistent pain

Abstract

Background: Chronic pain occurs when normally protective acute pain becomes pathologically persistent. We examined here whether an isoform of protein kinase C (PKC), PKMζ, that underlies long-term memory storage in various brain regions, also sustains nociceptive plasticity in spinal cord dorsal horn (SCDH) mediating persistent pain.

Results: Cutaneous injury or spinal stimulation produced persistent increases of PKMζ, but not other atypical PKCs in SCDH. Inhibiting spinal PKMζ, but not full-length PKCs, reversed plasticity-dependent persistent painful responses to hind paw formalin and secondary mechanical hypersensitivity and SCDH neuron sensitization after hind paw capsaicin, without affecting peripheral sensitization-dependent primary heat hypersensitivity after hind paw capsaicin. Inhibiting spinal PKMζ, but not full-length PKCs, also reversed mechanical hypersensitivity in the rat hind paw induced by spinal stimulation with intrathecal dihydroxyphenylglycine. Spinal PKMζ inhibition also alleviated allodynia 3 weeks after ischemic injury in rats with chronic post-ischemia pain (CPIP), at a point when allodynia depends on spinal changes. In contrast, spinal PKMζ inhibition did not affect allodynia in rats with chronic contriction injury (CCI) of the sciatic nerve, or CPIP rats early after ischemic injury, when allodynia depends on ongoing peripheral inputs.

Conclusions: These results suggest spinal PKMζ is essential for the maintenance of persistent pain by sustaining spinal nociceptive plasticity.

Figure 1

SCDH atypical PKC expression, and effects of PKMζ/PKC inhibition on formalin nociception. (a) Western blots and (b) histograms (% change) of atypical PKC showing i.pl. formalin increased PKMζ, PKCζ and PKCι/λ (n = 5/6 for vehicle/formalin) in SCDH (*P < 0.05). (c) Effect of ZIP on rotarod latencies, which significantly increased over training (*p < 0.05 from day 1). Latencies on day 6 (after ZIP) did not differ from day 5 (n.s.: non-significant difference). (d,e) Dose-dependent effects of pretreatment with ZIP (n = 5-7/dose) vs. scr-ZIP (n = 7) (d) or NPC-15437 (n = 5-7/dose) vs. vehicle (n = 8) (e) on formalin pain. ZIP dose-dependently reduced late phase (9-45 min) scores (†P < 0.05, *P < 0.01 red dash-boxed or individual symbols from scr-ZIP), but not early phase scores (0-6 min). NPC-15437 also significantly reduced late phase scores (†P < 0.05, *P < 0.01 for red dash-boxed or individual symbols), but not early phase scores. (f,g) Effect of ZIP (n = 6) or scr-ZIP (n = 4) pretreatment on von Frey paw-withdrawal thresholds (PWT, f) or plantar test paw-withdrawal latencies (PWL, g) in naïve rats. Neither ZIP nor scr-ZIP significantly affected either PWTs or PWLs. (h,i) Painful responses induced by formalin before and after ZIP (n = 5) or scr-ZIP (n = 5) (h), and NPC-15437 (n = 5) or vehicle (n = 7) (i), administered 25 min post-formalin. ZIP (*P <.05), but not NPC-15437, significantly attenuated persistent formalin pain. For these and subsequent graphs ZIP, scr-ZIP, NPC-15437 and vehicle were given i.t., and data is expressed as means ± s.e.m. unless otherwise stated.

Figure 2

SCDH atypical PKC expression, and effects of PKMζ/PKC inhibition on capsaicin-induced allodynia. (a) Western blots and (b) histograms (% change) of atypical PKC showing i.pl. capsaicin increased PKMζ, but not PKCζ or PKCι/λ, in SCDH from 2-24 h (*P < 0.05, n = 7/12 vehicle/capsaicin). c) Primary/secondary allodynia test sites in capsaicin-injected rats. d) Paw-withdrawal thresholds (PWTs) 24 h post-capsaicin after anesthesia (n = 5) or no-anesthesia (n = 5) of the primary site. Primary, but not secondary, allodynia was attenuated by anesthesia (*P < 0.05, from no-anesthesia). (e,f) Pre- and post-capsaicin PWTs (g) with ZIP, scr-ZIP, NPC-15437 or vehicle (n = 15,4,6,6, respectively) given 24 h post-capsaicin. ZIP (*P < 0.05), but neither scr-ZIP nor NPC-15437, significantly reversed capsaicin-induced allodynia 40-120 min post-drug. (g,h) Effect of ZIP given 20 min before testing on paw-withdrawal latencies, PWLs (s) and PWTs (g) 24 h after capsaicin or vehicle (n = 6,4, respectively). g) Heat hyperalgesia present at 2 h post-capsaicin (*P < 0.05, from vehicle), abated by 24 h, and PWLs were unaffected by ZIP. h) Mechanical allodynia 24 h post-capsaicin (*P < 0.05 from vehicle) was reversed by ZIP. (i,j) Effect of ZIP or scr-ZIP (n = 11, 8, respectively) 100 min post-capsaicin on PWLs and PWTs 20 min, 2 and 24 h post-capsaicin. i) Heat hyperalgesia (20 min-2 h, †P < 0.05 from baseline), abated by 24 h, and was unaffected by ZIP at 2-24 h post-capsaicin. j) Mechanical allodynia was significantly attenuated at 2 h, and reversed at 24 h, post-capsaicin by ZIP (*P < 0.05 from scr-ZIP).

Figure 3

Effects of PKMζ inhibition on capsaicin-induced sensitization of spinal WDR neurons. (a,b) Spike rate (spikes/sec) of evoked responses of 2 WDR neurons to von Frey filaments pre- and post-capsaicin. ZIP (90 min post-capsaicin) reduced the capsaicin-induced sensitization of mechanically-evoked responses of the neuron in (a). The neuron in (b) was not sensitized by capsaicin, and ZIP did not reduce evoked responses to the mechanical stimuli. (c-e) Mechanically-evoked WDR neuron responses to 1 (c), 5 (d) or 20 (e) g stimuli pre- and post-capsaicin (% of baseline action potentials evoked/20 s). ZIP, but not scr-ZIP (n = 6, 5, respectively), significantly reversed capsaicin-enhanced evoked responses to 5-20 g stimuli at 30-60 min post-drug (*P < 0.05). (f-h) Mechanically-evoked WDR neuron responses to 1 (f), 5 (g) or 20 (h) g stimuli pre- (baseline, BL) and post-vehicle (before (Veh) and after ZIP (30-60 min) applied 90 min post-vehicle) (n = 5). Neither i.pl. vehicle, nor i.t. ZIP, significantly affected evoked responses of WDR neurons. i) Background activity and evoked responses to heat pre- and post-capsaicin. Although capsaicin significantly increased the background activity of WDR neurons (** P < 0.01 from Pre-capsaicin), it failed to increase their heat-evoked responses (P >0.05, n = 8). j) Background activity of WDR neurons pre- (baseline, BL) and post-capsaicin (before (Cap) and after ZIP (30-60 min) applied 90 min post-capsaicin, n = 6). Capsaicin significantly increased background activity (*P < 0.05), and ZIP failed to significantly reduce the capsaicin-induced increase in the background activity of WDR neurons (P > 0.05).

Figure 4

SCDH atypical PKC expression, and effects of PKMζ/PKC inhibition on spinal DHPG-induced allodynia. (a) Spike rate (spikes/sec) of evoked responses of a WDR neuron to 20 and 50 g von Frey filaments before and 5 or 25 min after DHPG. Note the increased evoked responses to both mechanical stimuli, and the prolonged after-discharges to the 50 g stimulus, both 5 and 25 min post-DHPG. Scale bar (20 sec). (b) Western blots and (c) histograms of % change in atypical PKC levels compared to vehicle illustrating that i.t. DHPG significantly increased PKMζ (n = 12), but not PKCζ (n = 12) or PKCι/λ (n = 8) in SCDH (vehicle, n = 11), 2 and 24 h post-DHPG combined (*P < 0.05). (d,e) Pre- and post-DHPG paw withdrawal thresholds (PWTs, g) for rats given i.t. ZIP or scr-ZIP (d), i.t. NPC-15437 or vehicle (e), 30 min prior to testing at 24 h post-DHPG, with additional testing 48 h post-drug. ZIP (*P < 0.05), but not scr-ZIP or NPC-15437, reversed DHPG-induced mechanical allodynia at both time-points (n = 5 for each drug). f) Pre-drug (base) and post-drug PWTs of CCI rats (n = 8), early CPIP rats (n = 9) and late CPIP rats treated with i.t. ZIP (n = 7), or late CPIP rats treated with i.t. scr-ZIP (n = 8). ZIP, but not scr-ZIP, reversed allodynia in late CPIP rats, while ZIP had no effects in CCI or early CPIP rats (* P < 0.01; ** P < 0.001 from scr-ZIP; † P < 0.05; †† P < 0.01 from pre-drug baseline).