CaMKIIα may modulate fentanyl-induced hyperalgesia via a CeLC-PAG-RVM-spinal cord descending facilitative pain pathway in rats

Abstract

Each of the lateral capsular division of central nucleus of amygdala(CeLC), periaqueductal gray (PAG), rostral ventromedial medulla(RVM) and spinal cord has been proved to contribute to the development of opioid-induced hyperalgesia(OIH). Especially, Ca2+/calmodulin-dependent protein kinase IIα (CaMKIIα) in CeLC and spinal cord seems to play a key role in OIH modulation. However, the pain pathway through which CaMKIIα modulates OIH is not clear. The pathway from CeLC to spinal cord for this modulation was explored in the present study. Mechanical and thermal hyperalgesia were tested by von Frey test or Hargreaves test, respectively. CaMKIIα activity (phospho-CaMKIIα, p-CaMKIIα) was evaluated by western blot analysis. CaMKIIα antagonist (KN93) was micro-infused into CeLC, spinal cord or PAG, respectively, to evaluate its effect on behavioral hyperalgesia and p-CaMKIIα expression in CeLC, PAG, RVM and spinal cord. Then the underlying synaptic mechanism was explored by recording miniature excitatory postsynaptic currents (mEPSCs) on PAG slices using whole-cell voltage-clamp methods. Results showed that inhibition of CeLC, PAG or spinal CaMKIIα activity respectively by KN93, reversed both mechanical and thermal hyperalgesia. Microinjection of KN93 into CeLC decreased p-CaMKIIα expression in CeLC, PAG, RVM and spinal cord; while intrathecal KN93 can only block spinal but not CeLC CaMKIIα activity. KN93 injected into PAG just decreased p-CaMKIIα expression in PAG, RVM and spinal cord, but not in the CeLC. Similarly, whole-cell voltage-clamp recording found the frequency and amplitude of mEPSCs in PAG cells were decreased by KN93 added in PAG slice or micro-infused into CeLC in vivo. These results together with previous findings suggest that CaMKIIα may modulate OIH via a CeLC-PAG-RVM-spinal cord descending facilitative pain pathway.

Fig 1. Illustration of the experimental design.

Experiment 1: To investigate the effect of CeLC CaMKIIα antagonism on the level of CaMKIIα activity in CeLC, PAG, RVM and spinal cord regions and the behavioral hyperalgesia induced by fentanyl, rats were first implanted with CeLC cannulas before the induction of OIH by fentanyl. Then the Control group (n = 6) and OIH group (n = 6) received 50%DMSO (vehicle) 0.3 μl. OIH+KN92 10 nmol group (n = 6) and OIH+KN93 10 nmol group (n = 6) received equal volume (0.3 μl) of KN92 10 nmol and KN93 10 nmol respectively 6.5 h after the last injection of fentanyl. Experiment 2: To test whether spinal cord has a reverse effect on amygdala CaMKIIα signal, animals were randomly divided into four groups (n = 6). The Control group and OIH group received 50% DMSO (vehicle) 0.45 μl. OIH+KN92 45 nmol group and OIH+KN93 45 nmol group received equal volume (0.45 μl) of KN92 45 nmol and KN93 45 nmol respectively 6.5 h after the last injection of fentanyl. Experiment 3: To confirm whether p-CaMKIIα in vlPAG is involved in fentanyl-induced hyperalgesia and whether there is connection between PAG and other regions including CeLC, RVM and spinal cord in OIH, animals were randomly divided into four groups (n = 6). Control group and OIH group received 50% DMSO (vehicle) 0.3 μl. OIH + KN92 10 nmol group and OIH + KN93 10 nmol group received KN92 10 nmol and KN93 10 nmol respectively 6.5 h after the last injection of fentanyl. Experiment 4: To determine if there were differences between the Control rats and OIH rats in the synaptic transmission, and whether CaMKIIα modulates synaptic transmission in PAG neurons in OIH rats, the PAG slices of control rats (n = 6) and OIH rats (n = 6) were obtained 12 h after the last injection of saline or fentanyl. KN93 (10 μM) were added in the recording well 10 min after the baseline recording. Experiment 5: To further confirm whether inhibition of CeLC CaMKIIα activity has a direct effect on the enhanced synaptic transmission of vlPAG neurons in OIH rats, the Control+KN92 (n = 11) group and OIH+KN92 (n = 8) received CeLC injection of KN92 (10 nmol) and Control+KN93 group (n = 8) and OIH+KN93 (n = 8) group received CeLC injection of KN93 (10 nmol) respectively 6.5 h after the last injection of saline or fentanyl.

Fig 2. Intra-CeLC KN93 micro-injection reversed fentanyl-induced hyperalgesia and inhibited the CaMKIIα activation in CeLC, vlPAG, RVM and spinal cord in OIH rats.

(A-B) Bar graphs show the measurements of the mechanical threshold of paw withdrawal (A) and thermal paw withdrawal latency (± SD) collected at baseline (before Cannulation), pre-drug (6h after the last injection of fentanyl or saline) and then post-drug (0.5 h after microinjection). (C-J) Bar graphs show the mean relative density of p-CaMKIIα to GAPDH in CeLC (C- D), vlPAG (E-F), RVM (G-H) and spinal cord (I-J). Control, OIH, OIH+KN92 45 nmol, and OIH+KN93 45 nmol group received 50% DMSO (vehicle), 50%DMSO (vehicle), KN92 10 nmol and KN93 10 nmol respectively through the CeLC cannula. *, compared with Control group; &, compared with OIH group; +, compared with OIH+KN92 45 nmol group, n = 6 for each group; One symbol: p < 0.05, Two symbols: p < 0.01, Three symbols: p < 0.001.

Fig 3. Intrathecal KN93 injection attenuated fentanyl-induced hyperalgesia and decreased the level of p-CaMKIIα in spinal cord but not CeLC in OIH rats.

(A, B) Graphical display of the mechanical threshold of paw withdrawal (A) and thermal paw withdrawal latency (B) collected at baseline, pre-drug (6 h after the last injection of fentanyl or saline), and post-drug (0.5 h after Intrathecal injection). (C-F) Representative Immunoblots of activated CaMKIIα (p-CaMKIIα) and histogram of relative density of p-CaMKIIα to GAPDH in the spinal cord (C-D) and CeLC (E-F). Control, OIH, OIH+KN92 45 nmol, and OIH+KN93 45 nmol group received i.t. injection with 50% DMSO (vehicle), 50% DMSO (vehicle), KN92 10 nmol and KN93 10 nmol respectively. Results are expressed as mean ± SD; *, compared with Control group; &, compared with OIH group; +, compared with OIH+KN92 45 nmol group, n = 6 for each group; One symbol: p < 0.05, Two symbols: p < 0.01, Three symbols: p < 0.001.

Fig 4. Intra-vlPAG KN93 micro-injection reversed fentanyl-induced hyperalgesia and inhibited CaMKIIα activation in PAG, RVM and spinal cord but not in CeLC.

(A-B) Graphical display of the mechanical threshold of paw withdrawal (A) and thermal paw withdrawal latency collected at baseline (before Cannulation), pre-drug (6 h after the last injection of fentanyl or saline) and then post-drug (0.5 h after microinjection). (C-J) Representative immunoblots of activated CaMKIIα (p-CaMKIIα) and histogram of relative density of p-CaMKIIα to GAPDH in the CeLC (C-D), vlPAG (E-F), RVM (G-H) and spinal cord (I-J). Control group (n = 6) and OIH group (n = 6) received 50% DMSO (vehicle) 0.3 μl. OIH+KN92 10 nmol group (n = 6) and OIH+KN93 10 nmol group (n = 6) received KN92 10 nmol and KN93 10 nmol dissolved in 0.3 μl of 50% DMSO respectively through an intra-vlPAG cannula 6.5 h after the last injection of fentanyl. Results are expressed as mean ± SD; *, compared with Control group; &, compared with OIH group; +, compared with OIH+KN92 45 nmol group, n = 6 for each group; One symbol: p < 0.05, Two symbols: p < 0.01, Three symbols: p < 0.001.

Fig 5. KN93 (added in the ACSF) reversed the enhanced mEPSCs in vlPAG neurons from the OIH rats.

(A) mEPSCs recordings in control and OIH group at baseline and after the application of KN93. (B) Individual traces (average) of mEPSCs obtained from respective recordings. Calibration: 1 s, 20 pA. (C-D). Bar graphs showing the frequency (C) and amplitude (D) of mEPSCs in vlPAG neurons in control and OIH group (12 h post induction). Results are expressed as mean ± SD; n = 6 for each group; *, p < 0.05.

Fig 6. Inhibition of mEPSCs in vlPAG neurons from the OIH rats by intro-CeLC injection of KN93.

(A) mEPSCs recordings in control and OIH group after the application of KN92 or KN93. (B) Individual traces (average) of mEPSCs obtained from respective recordings. Calibration: 1 s, 20 pA. (C-D) Bar graphs showing the frequency (C) and amplitude (D) of mEPSCs from vlPAG neurons in slices from control and OIH group (12 h post induction). Results are expressed as mean ± SD; n = 8–11 for each group; *, p < 0.05, **, p < 0.01.