Sub-anaesthetic dose of propofol attenuates mechanical allodynia in chronic post-ischaemic pain via regulation of PTEN/PI3K/IL-6 signalling

Abstract

Background: Propofol is an intravenous anaesthetic drug that has been shown to reduce inflammatory pain. Complex regional pain syndrome (CRPS) type I is a pain condition characterized by autonomic, motor and sensory disturbance. The chronic post-ischaemic pain (CPIP) model is a well-established model to recapture CRPS-I syndromes pre-clinically by non-invasive ischaemic-reperfusion (IR) injury. In this study, we investigated the analgesic effects of propofol and underlying mechanisms in mitigating CRPS pain using the CPIP model. Methods: Sub-anaesthetic dose of propofol (25 mg/kg) was intravenously delivered to the CPIP model and sham control. Nociceptive behavioural changes were assayed by the von Frey test. Molecular assays were used to investigate expression changes of PTEN, PI3K, AKT and IL-6 underlying propofol-mediated analgesic effects. Pharmacological inhibition was applied for PTEN/PI3K/AKT pathway manipulation. Results: Both pre- and post-operative administration of propofol attenuated mechanical allodynia induced by CPIP. Propofol could modulate PTEN/PI3K/AKT signalling pathway by increasing active PTEN and reducing phosphorylated PI3K, phosphorylated AKT and IL-6 expression in the spinal dorsal horn, which promoted pain relief in the CPIP model. Inhibition of PTEN with bpV abolished the analgesic effects produced by propofol in CPIP mice. Conclusion: Sub-anaesthetic dose of propofol administration resulted in the activation of PTEN, inhibition of both PI3K/AKT signalling and IL-6 production in the spinal cord, which dramatically reduced CPIP-induced pain. Our findings lay the foundation in using propofol for the treatment of CRPS with great therapeutic implications.

Introduction

Complex regional pain syndrome (CRPS) is a chronic pain condition affecting the distal limbs. Patients present with autonomic (change in skin colour and temperature, sweating abnormalities), motor (stiffness and weakness), sensory (pain and hyperalgesia) disturbances and oedema. These result in dysfunction of the affected limb and impairment in activities of daily living.^1,2 CRPS is difficult to treat and patients with this debilitating condition suffer long-term pain.^1,2

CRPS is a neuro-inflammatory condition whereas high levels of pro-inflammatory cytokines including IL-1β, IL-2, IL-6 and TNF-α have been found in CRPS patients.^2,3 The chronic post-ischaemic pain (CPIP) model is a well-established animal model to recapture chronic regional pain syndrome type I pre-clinically, especially in the acute inflammatory phase. ⁴ Chronic mechanical allodynia, cold allodynia, oedema and hyperaemia of CPIP last for at least 1 month, which are produced by prolonged ischaemic reperfusion injury using a tight-fitting Durometer O-ring over the hindlimb.^4,5 Levels of pro-inflammatory cytokines TNF-α, IL-1 and IL-6 are increased in CPIP rats, indicating pivotal roles of inflammation in CPIP-induced pain. ⁴

Propofol (2,6-diisopropylphenol) is the most commonly used intravenous anaesthetic drug. It has analgesic properties and has been shown to reduce acute postoperative pain.^6–10 Propofol reduced pain responses in the plantar incision and formalin-induced inflammatory pain models.^11,12 It has also been shown to reduce levels of pro-inflammatory cytokines TNF-α, IL-1 and IL-6, which as mentioned above are important in the pathophysiology of CPIP and CRPS.^11–13 More recently, propofol administered after CPIP induction was shown to attenuate pain by inhibiting free radical, hypoxia inducible factor and inflammasome. ¹⁴ However, the pain signalling mechanism was mediated via peripheral tissues and not at the spinal cord level. Propofol may be a potential therapeutic option for reducing CRPS-induced pain, but centrally mediated pain mechanisms have not been elucidated. It is also unclear whether propofol given prior to CPIP induction would have a preventive analgesic effect. In this study, we planned to investigate the preventive and therapeutic effects of propofol and also to explore its analgesic mechanisms mediated via the spinal cord.

The phosphatase and tensin homolog (PTEN) deleted from chromosome 10 is known as a tumour suppressor, which is highly expressed in neurons and astrocytes, playing important roles in resolving neuropathic pain.^15–19 In the chronic constriction injury (CCI) rats model, PTEN was downregulated in the spinal cord whereas restoration of PTEN expressions could reduce pain responses. ¹⁷ Loss of PTEN resulted in the activation of PI3K/AKT, a key pathological pathway that contributes to the development and maintenance of pain. ²⁰ Propofol has been shown to increase levels of PTEN in the dorsal hippocampus. ²¹ In cancer studies, propofol could inhibit cell migration and induce cell apoptosis in different types of tumours by repressing the PI3K/AKT pathway.^22–24 Therefore, propofol could reduce CPIP-induced pain by regulating PTEN and PI3K/AKT signalling in the spinal dorsal horn.

In this study, we investigated the preventive and therapeutic effects of propofol in CPIP-induced pain by pre- and post-operative administration. We hypothesize that propofol would reduce CPIP-induced pain by upregulating PTEN and inhibiting PI3K/AKT signalling in the spinal dorsal horn. Our results will reveal the analgesic efficacy and underlying molecular mechanisms of propofol for CPIP, which could provide translational value for future CRPS management.

Materials and methods

Animals

Age-matched 6–10 weeks old male C57BL/6N mice weighing 20–23 g were kept under 12 h/12 h light-dark cycle with ad libitum access to food and water. Animal experiments were approved and conducted under the guidelines from the Committee on the Use of Live Animals in Teaching and Research (CULATR) of The University of Hong Kong (CULATR permit number: 4714-18).

Experimental design

Chronic post-ischaemic pain model

The Chronic post-ischaemic pain (CPIP) model was induced as described previously.^4,25 Briefly, this was done by applying an occlusion rubber ring of 1.78 mm inner diameter, 5 mm outer diameter and 1 mm thickness (Nitrile O-rings, O-rings West, USA) just above and proximal to the ankle of the mice right lower limb for 3 hours to produce ischaemia, followed by the release of the occlusion rubber ring to allow reperfusion.^4,25

Propofol given before or after CPIP induction

Age-matched male mice were randomly assigned to one of the four experimental groups: (1) Sham group, (2) CPIP group (CPIP induction), (3) C + P group (CPIP induction with the treatment of propofol in intralipid) and (4) C + IL group (CPIP induction with the injection of intralipid vehicle). For inhibitory studies, each of the four mice groups was injected with either (a) normal saline only, (b) 800 nM dimethyl sulfoxide (DMSO) (Sigma) vehicle or (c) 800 nM PTEN inhibitor bpV (Abcam) in 200 μl phosphate buffered saline (PBS) intraperitoneally (i.p.) before propofol administration (Figure 1(a)). Propofol at a dose of 25 mg/kg was injected to C57BL/6N mice (10 mg/mL in 20 mL 1% Lipuro stock, permit no. HK-49677, B.Braun, Melsungen AG, Germany) in 10% intralipid (20% stock 500 mL, permit no. HK-35421, Fresenius Kabi AB, Uppsala, Sweden) intravenously via tail vein. In accordance with the practical guide for dose conversion between animals and humans, the dose of propofol used in animals in this study was equivalent to the human dose of 2 mg/kg.^26,27 Propofol was given either before or after CPIP induction to study the preventive and therapeutic effects, respectively. For experiment involving bpV treatment, bpV or DMSO was given immediately after CPIP once.

Figure 1.

Timeline of experimental procedure. (a) Pre‐CPIP Propofol; (b) Post‐CPIP Propofol.

Behavioural test

The experimental procedures were performed as outlined in Figure 1. To study the preventive effect, propofol was administered 30 min prior to CPIP induction only. Von Frey test was carried out at baseline (Day 0) and was repeated on postoperative day (POD) 3. To study the therapeutic effect, propofol was administered immediately after CPIP induction, and again on POD 1 and 2 right after von Frey test. Von Frey test was carried out at baseline (Day 0) and was repeated on POD1, 2 and 3. For both experiments, von Frey test was conducted 24 h after the last administration of propofol, and propofol was given immediately after von Frey test on POD 1, 2 and 3.

Mechanical allodynia

Mechanical allodynia of bilateral hind paw was tested using an electronic von Frey Anaesthesiometer (IITC Life Science Inc., California, USA).^25,28–30 Briefly, mice were numbered and placed individually in plexiglass boxes on a metal mesh floor platform. After 30-min habituation, a stimulation hair filament with a uniform 0.8 mm diameter tip was used as a probe applied perpendicularly onto the plantar surface of the hind paw. During the baseline assessment, the minimum filament force tested that could produce a brisk hindlimb withdrawal response within 6–8 s was 4G force. Hence, this stimulation filament was used throughout the von Frey test for quantitative comparison of Paw Withdrawal Threshold (PWT) for each of the experimental mice groups. For each hind paw, von Frey tests were performed in triplicate at 5-min intervals. The baseline test before the surgery was used as a reference and the results obtained for the sham group served as calibrator.

Western blotting

Spinal dorsal horns of bilateral lumbar L2- L5 were harvested on POD 3 upon completion of behavioural test. Tissue samples of the spinal cord were homogenized (Polytron, Kinematica, Switzerland) in 200 μl ice-cold lysis buffer of total protein extraction kit (Thermo Scientific, USA) with protease and phosphatase inhibitor cocktails (Roche, Germany). Protein concentrations were determined at the absorbance of 595 nm by the Bradford assay kit (Bio-Rad, USA). An equal amount of 20 μL proteins were separated by 12% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis and the separated proteins were transferred to polyvinylidene difluoride (PVDF) methanol activated membranes at 100V for 1 h (Bio-Rad, USA). After 1 h of 5% non-fat milk blocking, the proteins of interest were detected by incubating with either rabbit or mouse primary antibodies (Abcam, UK) at a concentration of 1:1000 at 4°C overnight. After washing with 0.01 M Tris-buffered saline Tween-20 once for 15 minutes and three times for 5 minutes each, the membranes were incubated for 1 h at room temperature with horseradish peroxidase (HRP)-conjugated secondary antibodies immunoglobulin G (IgG) of corresponding antibodies (1:1000, Cell Signalling, USA). The proteins of interest were detected by the enhanced chemiluminescence (ECL) method (Bio-Rad, USA) and were documented by ChemiDoc Imaging System with Image Lab Touch Software (Bio-Rad, USA). Primary antibodies against proteins of interest used included: anti-PTEN [Y184], anti-phosphoPTEN(phospho Serine 370), anti-PI3Kp85α, anti-phospho-PI3Kp85α(phospho Tyrosine 607), anti-PI3Kp110β, anti-AKT, anti-phosphoAKT(phospho Thr308), anti-NaKATPase and anti-GAPDH (Cell Signalling Technologies, USA). Secondary antibodies used included: HRP-conjugated anti-mouse secondary antibodies IgG, HRP-conjugated anti-rabbit secondary antibodies IgG, HRP-conjugated anti-mouse secondary antibodies IgG (Cell Signalling Technologies, USA). Antibodies were removed completely by stripping buffer (62.5 mM Tris-HCl, pH6.8, 2% SDS and 100 mM β-mercaptoethanol) at room temperature for 30 minutes in between proteins detection. All signals were confirmed to be completely erased by ECL before performing Western blotting.

Immunofluorescence staining

On POD 3 after CPIP induction, transcardiac perfusion with 4% paraformaldehyde was performed and the mice’s lumbar L2-L5 spinal cord was isolated. After overnight 4% paraformaldehyde post-fixation, the tissues were immersed and cryoprotected in 30% sucrose and were embedded and cryopreserved in optimal cutting temperature (OCT) compound. The embedded and cryoprotected spinal cord tissues from different experimental groups were cryo-sectioned at 20 μm thickness (Leica CM3050S Cryostat, USA). After having permeabilized and blocked with 0.1% Triton-X 100 and corresponding serum for 30 min, the sections were washed three times for 5 minutes with PBS. The protein of interest of the tissue sections was incubated with double labelling with primary antibodies at 4°C overnight. The primary antibodies used are anti-PTEN [Y184] (Abcam #ab32199, 1:200), anti-phosphoPTEN (Abcam #ab195056, 1:150), anti-phospho-PI3Kp85α (Abcam #ab182651, 1:400), anti-phosphoAKT(Abcam #ab38449, 1:200), anti-GFAP (Sigma-Aldrich #MAB360, 1:500), anti-iba1(Santa cruz #sc-32725,1:500) and anti-NeuN (Sigma-Aldrich #MAB377, 1:500). After having washed the slices with PBS 5 minutes for three times, then the slices were incubated with anti-rabbit Alexa Fluor 488- and anti-mouse Alexa Fluor 568-conjugated secondary antibodies (1:500) (Thermo Fisher Scientific, USA) 1 h at room temperature. The sections were counter-stained with nuclear 4′,6-diamidino-2-phenylindole (DAPI) staining (Cell Signalling Technology, USA). Images were detected by Confocal Laser Scanning Microscopy LSM710 (Carl Zeiss, Germany) and were processed by ImageJ v1.52 (ImageJ, NIH, USA).

Statistical analysis

All data of both von Frey test and Western Blotting were expressed as mean ± standard error of the mean (SEM). The number of mice used was between 11 to 18 for behavioural test and 4 for Western Blotting from at least 3 independent experiments. The pixel density of protein bands of Western Blotting was measured by ImageJ v1.52 (ImageJ, NIH, USA). All the data were tested with the Shapiro-Wilk normality test for normal distribution. One-way or two-way analysis of variance (ANOVA) were performed with GraphPad Prism v7.04 software to determine the treatment effects between animal groups or inhibitors among animal groups under different treatments. This was followed by Tukey’s multiple comparisons test or Sidak multiple comparisons test. When only two groups were analysed, statistical significance was determined by the unpaired Student’s t test. Statistical significance was set at a p-value of less than or equal to 0.05.

Results

Propofol attenuated mechanical allodynia in the CPIP model

Previous findings showed that pre-emptive administration of propofol at a maintenance dosage of 0.6 mg·kg⁻¹·min⁻¹ for 1-h or 1.5 mg·kg⁻¹·min⁻¹ for 30-min produced preventive analgesic effects in the formalin-induced inflammatory pain and hind-paw incision post-operative pain, respectively.^11,12 However, whether propofol has a preventive analgesic effect and can produce long-lasting analgesia in CRPS-I has not been explored. Thus, we used the CPIP model to mimic CRPS type I.^4,31,32 The CPIP mouse model was successfully generated with decreased PWT upon mechanical stimulation of ipsilateral paw from 30.18 ± 10.65% (p = 0.0432) to 57 ± 11.48 (p = 0.0004) (One-way ANOVA with Tukey’s multiple comparisons test, n = 4–7) compared to the sham group on POD 3 after CPIP induction (Figure 2).

Figure 2.

Propofol attenuates mechanical allodynia in both pre- and post‐CPIP mice by Von Frey Test. Chronic post-ischaemia pain (CPIP) C57BL/6N mice were induced by the application of occlusion rubber ring (Durometer O-rings, 1.78 mm diameter) just above and proximal to the right ankle of lower limb for 3 hours. Von Frey tests with 4G stimulation filament were performed as described in Figure 1. Bar chart shows the decrease in Paw Withdrawal Threshold (PWT) in both Pre and Post CPIP mice 3 days after CPIP. 0.5 mL bolus of 25 mg/kg propofol in 10% intralipid was injected intravenously via tail vein of the CPIP mice immediately before (a) or immediately after and repeated dose on day 1 and 2; (b) PWT was significantly increased with improvement in mechanical allodynia of the ipsilateral paw. Each bar or line represents mean ± SEM of number of mice between 4 to 14. Ordinate shows the percentage of PWT with sham as reference. One‐way ANOVA with Tukey post multiple comparisons was performed with GraphPad Prism V7.04 software among Baseline (mice before treatment), Sham (mice without CPIP and treatment), CPIP (mice with CPIP induction), C+P (CPIP mice treated with propofol in intralipid) and C+IL (mice injected with intralipid vehicle). Statistical significances were indicated as * with p value ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001 and **** ≤ 0.0001.

To evaluate the role of propofol in preventing and treating CPIP, a sub-anaesthetic propofol dose of 25 mg/kg (0.5 mL) in 10% intralipid was administered intravenously over 15 min via tail vein either before or immediately after CPIP induction followed by repeated injection on POD 1 and POD 2. Pre-emptive propofol administration significantly increased the PWT in CPIP mice by 32.41 ± 8.558% (p = 0.0024, n = 14) on POD 3 (Figure 2(a)). In addition, propofol given after CPIP induction also increased PWT by 42.01 ± 7.92% (p < 0.0001, n = 5) (One-way ANOVA with Tukey’s multiple comparisons test) on POD 3 as compared to CPIP mice with the injection of intralipid vehicle (Figure 2(b)). There was no significant difference with the contralateral side (data not shown). Altogether, propofol could prevent and treat mechanical allodynia induced by CPIP.

Effect of propofol on PTEN/PI3K/AKT expression in mice with CPIP

Phosphorylation promotes the inactive formation of PTEN while dephosphorylation induces translocation of PTEN to the plasma membrane where the active PTEN antagonizes PI3K/AKT signalling.^33–35 PTEN inactivation through increased phosphorylation of the C-terminal serine-threonine cluster (p-PTEN) was observed in the CCI model, which was highly associated with neuroinflammation and nociceptive sensitization. ³³ Propofol has also been reported to inhibit inflammation via modulating PI3K/AKT signalling in cancer and brain injury.^36–38

To investigate whether sub-anaesthetic dose of propofol attenuated CPIP-induced mechanical allodynia is through regulation of PTEN/PI3K/AKT, we first evaluated the changes of PTEN/PI3K/AKT signalling pathway in mice exposed to CPIP. The ratio of p-PTEN/total PTEN (1.52 ± 0.02 versus 1.00 ± 0.02, p = 0.0008, n = 4), p-PI3Kp85α/PI3Kp85α (1.36 ± 0.04 versus 1.00 ± 0.03, p < 0.0001, n = 4) and p-AKT/total AKT (3.14 ± 0.12 versus 1.00 ± 0.05, p < 0.0001, n = 4) were significantly elevated in the ipsilateral spinal dorsal horn of CPIP mice compared to the sham group (One-way ANOVA with Tukey’s multiple comparisons test, n = 4, CPIP vs Sham) (Figure 3(a)–(d)). This indicates that CPIP was associated with the inactivation of PTEN and increased PI3K/AKT activity. Injection of propofol in CPIP mice resulted in a significant decrease in the ratio of p-PTEN/total PTEN (1.39 ± 0.03 versus 1.5 ± 0.02, p = 0.0043, n = 4), p-PI3Kp85α/PI3Kp85α (0.84 ± 0.02 versus 1.36 ± 0.04, p < 0.0001, n = 4), p-AKT/total AKT (1.26 ± 0.07 versus 3.14 ± 0.12, p < 0.0001, n = 4) as compared with CPIP mice (Two-tailed unpaired t test, n = 4, C + P vs CPIP) (Figure 3). This suggests that propofol activated PTEN and inhibited PI3K/AKT signalling in the CPIP model.

Figure 3.

Propofol increases active PTEN but decreases phosphorylated PI3Kp85α and phosphorylated AKT in CPIP mice. (a) Western blotting shows protein bands (from top) of phosphorylated PTEN (p‐PTEN), total PTEN, phosphorylated PI3Kp85α (p‐PI3Kp85α), total PI3Kp85α, phosphorylated AKT (p‐AKT), total AKT and glyceraldehyde-3‐phosphate dehydrogenase (GAPDH) expression of 47, 54, 84, 84, 60, 56 and 37kDa respectively. Bar chart showing:(b) ratio of p‐PTEN over total PTEN protein; (c) ratio of p‐PI3Kp85α over total p‐PI3Kp85α protein; (d) ratio of p‐AKT over total AKT protein. Each bar represents mean ± SEM of number of 4 mice after glyceraldehyde-3‐phosphate dehydrogenase (GAPDH) normalization. Two‐way ANOVA with Tukey post multiple comparisons was performed with GraphPad Prism V7.04 software among Sham (S, mice without CPIP and treatment), CPIP (mice with CPIP induction), C+P (CPIP mice treated with propofol in intralipid) and C+IL (mice injected with intralipid vehicle). Statistical significances were indicated as * with p value ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001 and **** ≤ 0.0001 with Sham as reference.

To further examine whether propofol exerts its analgesic effect through direct or indirect association with neurons and/or astrocytes in this CPIP model, double immunofluorescent staining of PTEN or p-PTEN and GFAP in the ipsilateral dorsal horn of spinal cord of propofol-injected mice were carried out. GFAP is a prominent marker for reactive astrocytes in response to pain and inflammation. Active astrogliosis evidenced by remarkable intense fluorescent immunoreactivity of GFAP expression (green) in spinal cord dorsal horn of CPIP mice were observed, whereas administration of propofol resulted in a reduction in GFAP expression in CPIP mice (Figure 4(a) and (b)). Less astrocyte expression shown by immunofluorescent staining suggests that propofol reduces the extent of inflammation in mice with CPIP. Expression of p-PTEN in the spinal dorsal horn was increased in both CPIP and C + IL groups compared with the sham group, and propofol significantly reduced p-PTEN levels in CPIP mice (Figure 4(b)). Although both PTEN and p-PTEN were not exclusively expressed and co-localized in GFAP positive astrocytes, our results have demonstrated their global expressions in neurons. Moreover, the extent of inflammation indicated by astrogliosis has also been shown in different treatment groups.

Figure 4.

Immunofluorescence of PTEN and p‐PTEN expression in ipsilateral spinal cord dorsal horn of CPIP mice under the effect of Propofol. Propofol increases PTEN expression (a) and (c, magnification of inset) in CPIP mouse. PTEN (c) is red fluorescence in high intensity but not in p‐PTEN (b). GFAP is green fluorescence with highest intensity in CPIP and CPIP+Intralipid treatment groups. Blue is nuclei staining. Scale bars are 20 μm (a and b) and 5 μm (c).

Administration of propofol promoted translocation of cytoplasmic p-PTEN to the membrane with high fluorescence intensity (Figure 4(c)), suggesting activation of PTEN binding to the plasma membrane. This is in agreement with our Western blotting findings (Figure 3(a)). It has been reported that PTEN activation contributes to pain resolution in a neuropathic pain model and inhibition of PI3K/AKT attenuated pain responses.^17,19,20,39 These results suggest that sub-anaesthetic dose of propofol could produce analgesic effects for CRPS via modulating PI3K/AKT signalling.

Analgesic effects of propofol in CPIP was abolished by PTEN inhibitor bpV (phen)

To determine whether the analgesic effects produced by propofol in CPIP were via the activation of PTEN and/or the inhibition of PI3K/AKT signalling, we examined the pain responses of administering PTEN inhibitor bpV (800 nM) in CPIP mice. PTEN inhibitor bpV significantly decreased PWT by 24.8 ± 8.095% (p = 0.0195) compared to the DMSO-treated group in mice given propofol pre-emptively (Figure 5(a)). In mice given propofol after CPIP induction, PTEN inhibitor bpV reduced PWT by 26.71 ± 7.558% (p = 0.0021) (Unpaired t test, n = 10) (Figure 5(b)).

Figure 5.

PTEN inhibitor bpV exacerbates mechanical allodynia in CPIP mice and be prevented by pre‐CPIP Propofol treatment. Therapeutic analgesic effect of propofol can be abolished by bpV by Von Frey Test. Bar chart showing the PWT in both (a) pre‐ and (b) post‐CPIP mice 3 days after CPIP under the administration of propofol with or without inhibitors. Prior to propofol injection, each of four mice groups was immediately injected with either (1) normal saline without inhibitors, or (2) 800 nM DMSO vehicle, or (3) 800 nM PTEN inhibitor bpV in 200 μl PBS intraperitoneally. Each bar represents mean ± SEM of number of mice between 4 to 14. One‐way ANOVA with Tukey post multiple comparisons was performed with GraphPad Prism V7.04 software among Sham (mice without CPIP and treatment), CPIP (mice with CPIP induction) and C+P (CPIP mice treated with propofol in intralipid). Statistical significances were indicated as * with p value ≤ 0.05; **≤ 0.01; ***≤ 0.001 and ****≤ 0.0001 with Sham as reference.

Considering the functional significance of p-PTEN, p-PI3K and p-AKT, we further examined whether pharmacological inhibition of PTEN by bpV could alter the protein expressions involved in signalling activities in propofol administrated CPIP model. Western blotting studies showed that inhibition of PTEN by bpV in propofol-injected CPIP mice resulted in a statistically significant reduction of p-PTEN/total PTEN (0.60 ± 0.01 versus 1.25 ± 0.01, p < 0.0001, n = 4), a significant increase in p-PI3Kp85α/total PI3Kp85α (1.50 ± 0.09 versus 0.95 ± 0.03, p = 0.001, n = 4) and a significant decrease in p-AKT/total AKT (0.17 ± 0.01 versus 0.24 ± 0.01, p = 0.0013, n = 4) compared to DMSO vehicle control (Two-tailed unpaired t test, n = 4, C + P_bpV vs C + P_DMSO) (Figure 6(a)–(d)). Immunofluorescent studies showed that bpV significantly decreased PTEN expressions (red) and increased p-PI3Kp85α (red) and p-AKT (red) expressions in the ipsilateral spinal dorsal horn of CPIP mice given propofol compared to corresponding DMSO control mice (C + P_bpV vs C + P_DMSO) (Figure 7). These results suggest that administration of propofol before or after CPIP induction reduced mechanical allodynia by activating PTEN and inhibiting the PI3K/AKT signalling pathway.

Figure 6.

Protein expression under the effect of co‐administration of propofol and inhibitors of PTEN by bpV, or DMSO vehicle control. (a) Western blotting shows protein bands (from top) of p‐PTEN, total PTEN, p‐PI3Kp85α, total PI3Kp85α, p‐AKT, total AKT and GAPDH expression of 47, 54, 84, 84, 60, 56 and 37kDa respectively. Bar chart showing: (b) ratio of p‐PTEN over total PTEN protein; (c) ratio of p‐PI3Kp85α over total PI3Kp85α protein; (d) ratio of p‐AKT over total AKT protein. Each bar represents mean ± SEM. Two‐way ANOVA with Tukey post multiple comparisons was performed with GraphPad Prism V7.04 software among CPIP (mice with CPIP induction) and C+P (CPIP mice treated with Propofol in intralipid). Statistical significances were indicated as * with p value ≤0.05; ** ≤0.01; *** ≤0.001 and **** ≤0.0001 with CPIP without inhibitor as reference by comparing between CPIP and CPIP with Propofol injection under various treatments.

Figure 7.

Immunofluorescence of PTEN, p‐PI3Kp85α and p‐AKT expression (upper to lower row) in ipsilateral spinal cord dorsal horn of CPIP mice under the effect of Propofol and PTEN inhibitor bpV. The protein expressions are in red fluorescence in different treatment groups including CPIP mice injected with DMSO vehicle control (left to right column), C+P mice injected with DMSO vehicle control, and bpV injected C+P mice. GFAP is green fluorescence and blue is nuclei staining. Scale bars are 20 μm.

PTEN expression and PI3K/AKT signalling activation were mainly in neuron cells

In order to identify the location of PTEN expression and PI3K/AKT signalling activation in spinal cord cells, co-immunofluorescent staining of PTEN, p-PTEN, p-PI3K or p-AKT with GFAP, Iba1 and Neun in the ipsilateral dorsal horn of spinal cord of mice with CPIP model were performed. GFAP, Iba1 and Neun are prominent marker for astrocytes, microglia and neurons, respectively. The results showed that PTEN, p-PTEN, p-PI3K and p-AKT were dominatingly co-stained with Neun (Figure 8), which indicated that CPIP induced PTEN/PI3K/AKT signalling activation was mainly in neuron cells. Moreover, a small amount of PTEN/p-PTEN/p-PI3K/p-AKT and GFAP co-staining was also identified. However, PTEN/p-PTEN/p-PI3K/p-AKT and Iba1 were not co-stained (Figure 8).

Figure 8.

Immunofluorescence of PTEN, p‐PTEN, p‐PI3Kp85α and p‐AKT (left to right column) co‐stain with GFAP, Iba1, NeuN (upper to lower row) in ipsilateral spinal cord dorsal horn of CPIP mice. The protein PTEN, p‐PTEN, p‐PI3Kp85α and p‐AKT are in red fluorescence. GFAP, Iba1 and NeuN are in green fluorescence and blue is nuclei staining. Scale bars are 15 μm.

Altered PTEN/PI3K/AKT signalling by propofol contributes to the regulation of IL-6 expression in CPIP mice

We focused on IL-6 as a downstream mediator of PTEN/PI3K/AKT in propofol-mediated analgesia. IL-6 is a pro-inflammatory cytokine involved in the pathogenesis of various pain upon stimulus. ⁴⁰ The expression of IL-6 was significantly increased in CPIP mice compared to sham group mice (8.21 ± 0.06 versus 1.00 ± 0.04. p < 0.0001, n = 4) (One-way ANOVA with Tukey’s multiple comparisons test, n = 4, CPIP vs Sham) (Figure 9(a) and (b)). Propofol significantly reduced IL-6 expression in mice with CPIP (4.36 ± 0.04 versus 8.21 ± 0.06, p < 0.0001, n = 4) as compared to the CPIP group (One-way ANOVA with Tukey’s multiple comparisons test, n = 4, C + P vs CPIP) (Figure 9(a) and (b)). Inhibition of PTEN by bpV in propofol-injected CPIP mice significantly increased IL-6 expressions (1.01 ± 0.07 versus 0.72 ± 0.01, p = 0.0286, n = 4) (Figure 9(c) and (d)) compared to corresponding DMSO-treated control (Two-tailed unpaired t test, n = 4, C + P_bpV vs C + P_DMSO). This indicates that activation of PTEN by propofol is required for the inhibition of IL-6, which functions as a major pathological factor in producing inflammatory chronic pain responses for CRPS-I.

Figure 9.

IL‐6 expression under the effect of propofol with or without PTEN inhibitor bpV. Propofol downregulates inflammatory cytokine IL‐6 protein expression in CPIP mice. (a) and (b) Western blotting shows protein bands (from top) of interleukin 6 (IL‐6) and glyceraldehyde-3‐phosphate dehydrogenase (GAPDH) expression of 24 and 37kDa respectively. Bar chart showing (c) and (d) IL‐6 total protein. Each bar represents mean ± SEM of number of 4 mice after GADPH normalization or Na/K‐ATPase normalization for IL‐6. Two‐way ANOVA with Tukey post multiple comparisons was performed with GraphPad Prism V7.04 software among Sham (mice without CPIP and treatment), CPIP (mice with CPIP induction), C+P (CPIP mice treated with propofol in intralipid) and C+IL (mice injected with intralipid vehicle). Statistical significances were indicated as * with p value ≤ 0.05; **≤ 0.01; ***≤ 0.001 and ****≤ 0.0001 with Sham as reference in (c) and CPIP no inhibitor as reference in (d).

These results suggest that activation of PTEN and inhibition of PI3K/AKT reduced mechanical allodynia in the CPIP model. Reduction of downstream inflammatory cytokine IL-6 expression contributed to analgesia. Taken together, propofol attenuated mechanical allodynia when given before or after CPIP induction. The analgesic effect was mediated via the PTEN/PI3K/AKT signalling pathway by upregulating active PTEN and downregulating p-PI3K and p-AKT in CPIP mice, resulting in decreased inflammatory cytokine IL-6. PTEN inhibition with bpV exacerbated mechanical allodynia in CPIP mice. Propofol, however, increased active PTEN while decreasing phosphorylated PI3K and phosphorylated AKT, resulting in a decrease of IL-6 in CPIP mice.

Discussion

Our results revealed that administration of propofol before CPIP induction could prevent pain development whereas late propofol given after CPIP induction mitigated established mechanical allodynia. Mechanistically, propofol increased active PTEN and decreased phosphorylated PI3K, phosphorylated AKT and IL-6 expression, leading to reduced mechanical allodynia in CPIP mice. Inhibition of PTEN by bpV reversed its analgesic effect via PTEN/PI3K/IL-6 signalling, suggesting a potential molecular mechanism of propofol-mediated analgesia from central dorsal spinal level.

In this study, we managed to identify tumour suppressor gene PTEN, which was regulated by propofol in CPIP mice. It is known that phosphorylation promotes the inactive formation of PTEN while the dephosphorylation induces translocation of PTEN to the plasma membrane where the active PTEN antagonizes PI3K/AKT signalling.^33–35 PTEN inactivation through increased phosphorylation of the C-terminal serine-threonine cluster (p-PTEN) was observed in the CCI model, which is a model highly associated with neuroinflammation and nociceptive sensitization. ³³ In addition, propofol has been shown to inhibit inflammation via modulating PI3K/AKT signalling in cancer and brain injury.^36–38 Consistent with previous findings, our work demonstrates that propofol reduced ischaemia reperfusion-induced inflammatory pain in CPIP by regulating PTEN, PI3K/AKT and IL-6. Inflammatory pain is the prominent feature of acute CRPS. IR injury in the CPIP model results in increased levels of inflammatory cytokines, which activate nociceptive C-fibers and induce peripheral and central sensitization. ⁴¹ Our experiments showed that inhibition of PTEN abolished the analgesic effect produced by propofol and activated PI3K/IL-6 in propofol-treated CPIP mice. This suggests that the effect of propofol on pain reduction and potential downstream pain mediators PI3K, AKT and IL-6 was dependent on the upregulation of active PTEN. Our immunostaining data also showed that propofol reduced the level of p-PTEN and inhibited the PI3K-AKT signalling pathway mainly in neurons, secondarily in astrocytes, but not in microglial cells, suggesting that the effect of propofol is cell-type specific. Therefore, the results highlight the importance of PTEN in propofol-mediated pain signalling. Unlike most previous research on the analgesic mechanism of propofol from peripheral level, our study investigated its effect on CPIP through central manipulation in the spinal dorsal horn.

When administering propofol, we adopted a sub-anaesthetic dose before or after CPIP induction to study its preventive and therapeutic effect. The results suggest that a sub-anaesthetic dose of propofol was effective in attenuating mechanical allodynia in both conditions during the first 3 days of the early acute injury phase. This supports the use of sub-anaesthetic propofol for therapeutic treatment and for preventing chronic pain development. The significance of the study suggests an alternative potential treatment option using propofol under sub-anaesthetic dose is clinically important to minimize adverse effects such as oversedation, respiratory depression, apnoea and haemodynamic instability.^42–45 In the meantime, sub-anaesthetic dosage not only reduced sedation during pain relief, but also avoids prolonged duration of continuous intravenous infusion which can be replaced by single bolus injection.

Additionally, the ultrashort acting intravenous anaesthesia property of propofol is relevant in short-term pain relief. Though propofol has a quite safe record of usage especially the low rate of nausea and vomiting in patients after surgeries, over-dosage might still lead to severe complications varying from descriptive studies and randomized controlled trials including dose-dependent hypotension (most common), hypoxia, apnoea, arrhythmia, allergy with bronchospasm, hypertriglyceridemia and pancreatitis (uncommon), green discoloration of the urine and skin as well as the “propofol infusion syndrome” – severe metabolic acidosis, rhabdomyolysis and circulatory collapse induced by high dose infusions (rare but potentially fatal)^42–45 In this case, we introduced in this study by intravenous bolus administration of 10% intralipid diluted sub-anaesthetic dose of propofol in pain relief in CPIP mice model. The clinical relevance of this study is the introduction of propofol in CPIP pain management at its diluted sub-anaesthetic dosage and the investigation of the underlying pain related signalling pathway. We have demonstrated a new pain signalling pathway for propofol that involved the regulation of the PTEN/PI3K/IL-6 pathway in the spinal cord.

CRPS is a challenging clinical pain condition where the acute phase can develop into chronic debilitating pain. The mechanism of pain development and the pain-related signalling pathway involved in CRPS is complex, multifactorial and not well understood. Conventional treatment options have limited supporting evidence and efficacy. ⁴⁶ It is of paramount importance to have a better understanding of pain signalling involved in CRPS and explore an alternative effective strategy in CRPS pain management. In this study, we generated the CPIP mouse which can model CRPS Type 1. The typical signs of oedema and hyperaemia followed by hyperalgesia, mechanical allodynia and cold allodynia are similar to the clinical features in individuals with CRPS, especially during the acute phase. ⁴ Since CPIP is an animal model of CRPS, propofol may, therefore, contribute to providing an alternative strategy in CRPS pain management. Propofol has anti-inflammatory effects and reduced pro-inflammatory cytokines levels (TNF-α, IL-6 and IL-1β) that were induced by lipopolysaccharide. ³⁶ It also inhibits the N-methyl-D-aspartate (NMDA) receptor, potassium/sodium hyperpolarisation-activated cyclic nucleotide-gated channel 1 (HCN1) channels, and transient receptor potential cation channel, subfamily A, member 1 (TRPA1), which are possible analgesic mechanisms proposing propofol being a candidate agent in CPIP pain relief especially at the acute phase.^13,47,48

Our data showed that propofol treatment prevented CPIP-induced reduction of PTEN activity specifically in spinal neurons and astrocytes. It has been reported that most PTEN was co-localized with neuronal cells other than astrocytes or microglial cells, and upregulation of spinal PTEN has analgesic effect. ⁴⁹ Moreover, astrocyte-specific knockout of PTEN was shown to increase astrocyte proliferation ⁵⁰ and mechanical allodynia. ⁵¹ Importantly, specific astrocytic overexpression of PTEN could alleviate chronic constriction injury (CCI)-induced pain-like behaviour, ⁵¹ and that non-specific PTEN overexpression attenuated CCI-induced astrocyte activation. ¹⁷ These results, together with our present findings suggest that PTEN is a key regulator in preventing the development of chronic pain. Although our results showed that inhibition of PTEN was insufficient to increase p-AKT expression, it is plausible that the inhibition effect by bpV might be partial or incomplete. Moreover, loss of PTEN activity induced by bpV triggered other miRNA activity such as miR21 to compensate for the PTEN activity loss, resulting in a decrease in p-AKT. However, more studies still need to be carried out on the potential actions between PI3K and the downstream cytokine IL-6 release to investigate if AKT is actually involved in this signalling or there might be other mechanisms behind. It is also perhaps worthwhile to study the action of propofol in CPIP mice with sustained and elevated pain threshold under prolonged pain state induced by prostaglandin E2 (PGE2).^48,52 Other potential pathways that may be involved in the analgesic mechanisms of propofol in CPIP, which could also be of implication for future investigation include opioid receptors such as delta opioid receptor (DOR), kappa opioid receptor (KOR), mu opioid receptor (MOR), G protein-coupled receptors (GPCR) and CXCL12/CXCR4 axis or other pain associated ion channels, namely transient receptor potential cation channel, subfamily A, member 1(TRPA1) and vanilloid, member 1 (TRPV1) channels or inhibitory neuronal activities associated glycine and gamma-aminobutyric acid a receptor (GABAaR).^53–55

In conclusion, sub-anaesthetic dose of propofol reduced CPIP-induced mechanical allodynia when given before or after induction of the pain model, suggesting both a preventive and therapeutic effect. This was achieved by upregulation of PTEN and inhibition of PI3K/IL-6 signalling in the spinal dorsal horn. Whether AKT was actually involved in the pathway would require future investigation (Figure 10). Pharmacological blockade of PTEN abolished the analgesic effect of propofol, indicating that propofol-mediated analgesia was PTEN dependent. Our results provide molecular mechanisms underlying analgesic effects produced by propofol in a clinically relevant pain model. Based on propofol’s positive effect in the CPIP model, our findings could have translational value for the management of CRPS.

Figure 10.

Proposed Mechanism of Analgesic Effect of Propofol under the effect of bpV by regulation of IL‐6 via PTEN/PI3K/AKT signalling pathway. Phosphatase and Tensin homolog deleted from Chromosome 10 (PTEN). Phosphoatidylinositol‐3 kinase (PI3K). Protein Kinase B (AKT). Interleukin 6 (IL‐6).

ORCID iDs

Hei Lui Lhotse Ng https://orcid.org/0000-0001-5857-7548

Stanley Sau Ching Wong https://orcid.org/0000-0002-8763-5687