Intrathecal pretreatment with DHA alleviated inflammatory pain induced by CFA
To investigate a specific role of DHA in acute inflammatory pain, the model of intraplantar CFA (1mg ml−1, 20 μl) injection was employed and i.t. DHA (1, 10 and 100 μg) was administered 60 minutes prior to CFA intervention. First, we found no change in the baseline of peripheral mechanical and thermal sensitivity between groups (Fig. 1A-C). Intriguingly, DHA at 10 μg and 100 μg but not 1 μg reduced CFA-induced inflammatory pain, as characterized by the abrupt increase in paw withdrawal mechanical threshold (F (5, 42) = 134.6, P < 0.0001, n = 8, two-way ANOVA, Fig. 1A), the significant decrease in paw withdrawal mechanical frequency (F (5, 42) = 87.28, P < 0.0001, n = 8, two-way ANOVA, Fig. 1B) and the considerable increase in paw withdrawal thermal latency (F (5, 42) = 147.6, P < 0.0001, n = 8, two-way ANOVA, Fig. 1C). Such analgesia of DHA strongly started at 3 hours and lasted for 1-3 days following CFA administration in a dose-dependent manner. Noteworthy, DHA at 100 μg failed to affect normal nociceptive sensitivity and locomotor function in vehicle-treated mice (Fig. 1), suggesting that DHA at 100 μg was effective and safe for our model.
To explore whether SIRT3-associated nitroxidative insult is involved in anti-nociception of DHA, the alterations of spinal SIRT3, 3-nitrotyrosine and PRDX3 were examined on day 1 after CFA injection, when CFA mice showed most obvious pain to noxious stimuli and DHA exhibited a robust pain-inhibition. Our biochemical results revealed that CFA induced the reduction in SIRT3 expression and activity as compared to vehicle-treated animals (Fig. 2A-C). Surprisingly, DHA increased SIRT3 expression and activity in CFA-treated mice (Fig. 2A-C). Double staining also revealed that SIRT3 highly co-localized with a neuronal marker NeuN but not astrocytic marker GFAP and microglial marker IBA-1, thus suggesting the primary expression of SIRT3 by the spinal dorsal horn neurons (Fig. 2D). Further, immunostaining verified the decrease of SIRT3 expression in CFA mice and the increase of SIRT3 expression after DHA intervention (Fig. 2E). 3-nitrotyrosine is well recognized as a biomarker for endogenous peroxynitrite accumulation [24]. Western blot assay revealed that DHA markedly inhibited the over-expression of spinal 3-nitrotyrosine in CFA animals (Fig. 2A and B). As parallel, DHA protected against CFA-induced spinal PRDX3 hyperacetylation (Fig. 2A and B).
All these data suggested that CFA induced the rapid spinal SIRT3 inactivation, peroxynitrite formation, and PRDX3 hyperacetylation, whereas DHA reversed these alterations and thus attenuating the severity of acute inflammatory pain in mice.
Intrathecal injection of DHA impaired neuropathic pain after SNI
After the alleviation of acute pain by DHA was conformed, we investigated the efficiency of DHA at improving SNI-induced chronic neuropathic pain. First, mice received three injections of DHA (i.t., 100 μg) daily from days 4 to 6 (in the early phase) after SNI procedures. Von Frey tests showed that DHA significantly diminished SNI-induced mechanical allodynia, as reflected by the long-lasting increase in paw withdrawal threshold (F (2, 105) = 206.5, P = 0.003, n = 8, two-way ANOVA, Fig. 3A), and the significant decrease in paw withdrawal frequency (F (2, 105) = 222.8, P =0.001, n = 8, two-way ANOVA, Fig. 3B) in SNI animals. The robust anti-allodynia sustained for 1 week after termination of the third treatment. Separately, the Hargreaves test revealed that DHA produced transient inhibitory effects on the established thermal hyperalgesia, as indicated by the extension in paw withdrawal latency (F (2, 105) = 489.8, P < 0.0001, n = 8, two-way ANOVA, Fig. 3C) on day 7 after SNI operation. Also, a single treatment of DHA on 14 days after SNI produced a rapid and transient suppression of the established mechanical allodynia for 5 hours (F (1, 50) = 21.7, P < 0.0001, n = 6, two-way ANOVA, Fig. 3D) and heat hyperalgesia for 1 hour (F (1, 50) = 4.364, P = 0.0418, n = 6, two-way ANOVA, Fig. 3E).
Interestingly, our biochemical data discovered that DHA up-modulated spinal SIRT3 expression and activity, and inhibited 3-nitrotyrosine formation and PRDX3 hyperacetylation on day 7 after SNI intervention (Fig. 3F-H). These findings suggested that DHA suppressed the generation and maintenance of peripheral nerve injury-induced chronic neuropathic pain by reducing nitroxidation.
Intrathecal treatment with DHA mildly impaired chemotherapy-induced peripherally mechanical allodynia
Next, DHA (i.t., 100 μg) was administered daily for 3 consecutive days on days 4, 5, and 6 (in the early phase) after intraperitoneal injection of paclitaxel (6 mg kg-1) to validate its analgesic role in chemotherapy-induced chronic pain. We found that DHA elicited a short-term increase in paw withdrawal threshold on day 7 after paclitaxel exposure (F (2, 105) = 190.4, P =0.046, n = 8, two-way ANOVA, Fig. 4A), whereas there were no significant differences in paw withdrawal frequency and paw withdrawal latency (Fig. 4B and C). Also, a single delivery of DHA in the late phase (on day 14 after paclitaxel injection) only exhibited a transient and significant inhibition of the established mechanical allodynia for 1 hour (F (1, 50) = 6.813, P = 0.012, n = 6, two-way ANOVA, Fig. 4D and E). These data demonstrated that DHA slightly and transiently impaired the initiation and development of chemotherapy-induced mechanical allodynia, but not thermal hyperalgesia.
Intriguingly, spinal decreases of SIRT3 expression and activity, as well as 3-nitrotyrosine overload and PRDX3 hyperacetylation in paclitaxel-treated animals were reversed after i.t. DHA delivery (Fig. 4F-H), suggesting the involvement of SIRT3 and PRDX3 pathways in the analgesic effect of DHA in chemotherapy-induced peripheral neuropathy.
Intrathecal injection of DHA protected against bone cancer pain
To further examine the protective effect of DHA on bone cancer pain, repetitive DHA (i.t., 100 μg) was delivered on a daily basis from days 4 to 6 (in the early phase) after the injection of NCTC sarcoma cells into the distal femur condyle. Strikingly, DHA produced a long-term and significant inhibition of mechanical allodynia and thermal hyperalgesia, which persisted for more than 7 days, as manifested by the increase of paw withdrawal threshold (F (2, 105) = 484.4, P =0.005, n = 8, two-way ANOVA, Fig. 5A), the decrease of paw withdrawal frequency (F (2, 105) = 263.4, P =0.002, n = 8, two-way ANOVA, Fig. 5B) and the extension of paw withdrawal latency (F (2, 105) = 1228, P =0.033, n = 8, two-way ANOVA, Fig. 5C) in animals with cancer pain. Furthermore, a single treatment of DHA on day 14 after sarcoma exposure significantly inhibited the established mechanical allodynia for 3 hours (F (1, 50) = 27.97, P < 0.0001, n = 6, two-way ANOVA, Figure 5D) and heat hyperalgesia for 1 hour (F (1, 50) = 4.975, P = 0.0302, n = 6, two-way ANOVA, Fig. 5E). More importantly, mice exhibited the spinal elevations of 3-nitrotyrosine production, PRDX3 hyperacetylation and SIRT3 inactivation on day 7 after tumor cells injection, which was restrained by repeated administration of DHA (Fig. 5F-H).
Intrathecal pretreatment with DHA relieved spontaneous pain induced by NMDA
Subsequently, we assessed whether DHA was equally valid in NMDA-induced spontaneous pain. DHA (i.t., 100 μg) was administered 60 minutes before NMDA exposure (i.t., 1 nmol). Herein, we reported that spontaneous nociception-like behaviors including dramatic scratching and rolling emerged from 1 minute and were sustained for 9 minutes (Fig. 6A) in NMDA mice. Strikingly, NMDA only induced obvious pain behaviors from 2 to 5 minutes in female mice with DHA pretreatment (F (1, 154) = 783.8, P < 0.0001, n = 8, two-way ANOVA, Fig. 6A), suggesting that the unbearable pain was weakened and shortened by DHA. Besides, DHA pretreatment reduced NMDA-caused spinal SIRT3 inactivation, 3-nitrotyrosine formation, PRDX3 hyperacetylation (Fig. 6B-D).
Oral administration of DHA relieves the established acute and chronic pain
Given that DHA is often administered via oral delivery for the treatment of malaria in clinical patients [25, 26], we next investigated whether oral DHA therapy was also effective against acute and chronic pain. According to previous reports, dose of DHA varying from 1 mg kg−1 to 100 mg kg−1 has been widely used in many animal models of different diseases with universally acknowledged efficacy and safety records [27]. Thus, we selected and administered DHA (50 mg kg−1) by gavage on 1 day following CFA injection, and on 14 days after SNI surgery, paclitaxel delivery and sarcoma cells implantation, respectively. Intriguingly, oral DHA therapy relieved the decrease of paw withdrawal mechanical threshold and paw withdrawal thermal latency due to peripheral inflammation, nerve injury and bone cancer (P < 0.05, n = 8, Fig. 7). Oral delivery of DHA also inhibited chemotherapy-induced mechanical allodynia (P < 0.05, n = 8, Fig. 7E). Such analgesia persisted for 3-5 hours after oral exposure to DHA in our acute and chronic pain models.
DHA inhibited spinal synaptic plasticity in inflammatory pain, neuropathic pain and cancer pain
Given the critical role of spinal synaptic plasticity in acute and chronic pain [19, 22], we examined whether DHA could modulate excitatory synaptic transmission in pain development. First, the frequency but not the amplitude of sEPSCs in lamina Ⅱo neurons of spinal cord slices on 1 day following peripheral CFA intervention was up-regulated (P = 0.016, Fig. 8A). Further, peripheral nerve injury increased the frequency of sEPSCs on 7 days after SNI exposure (P = 0.015, Fig. 8B). Strikingly, both CFA and SNI induced increases of sEPSCs were reduced by DHA pretreatment (P = 0.045 and P = 0.04, Fig. 8A and B), indicating the inhibition of neurogenic inflammation-induced synaptic plasticity by DHA (Fig. 8A and B). As parallel, the CFA and SNI potentiated the evoked EPSCs (eEPSCs) after the dorsal root stimulation (Fig. 8D and E), suggesting the unique contribution of inflammation and nerve trauma to glutamate release from central terminals of primary dorsal root ganglion (DRG) afferents in neuronal responsiveness following noxious stimulation. Furthermore, both CFA and SNI failed to increase the amplitude of the eEPSCs in the present of DHA (Fig. 8A and B), suggesting that DHA impaired noxious stimuli-induced synaptic activity via peripheral and presynaptic mechanisms. Additionally, recent studies have recapitulated the significance of dendritic spine morphogenesis in excitatory synaptic functional plasticity, after nerve injury, bone fracture and opioid treatment, which is indispensable for central sensitization and pain development [22-24]. Our results verified the increase of spine density following SNI surgery, paclitaxel injection and sarcoma cells implantation (P < 0.05, Fig. 8E-G). Intriguingly, DHA administration inhibited the up-regulated number of dendritic spines in mice with neuropathic pain and bone cancer pain (P < 0.05, Fig. 8E-G), demonstrating the suppression of structural spine plasticity by DHA. Collectively, these findings suggest that DHA may cause pain alleviation through reducing spinal synaptic plasticity.
Spinal SIRT3 inhibition eliminated the analgesic effect of DHA on acute and chronic pain
Then, we further tested whether SIRT3 is involved in the antinociceptive effect of DHA. The selective SIRT3 inhibitor 3-TYP was injected intrathecally (single injection, 10 μg) at 30 minutes after DHA treatment. Interestingly, the attenuation of CFA-induced mechanical allodynia and thermal hyperalgesia by DHA (i.t., 100 μg) was completely reversed after 3-TYP delivery (Fig. 9A and B). Furthermore, 3-TYP intervention impaired the protective effect of DHA on SNI-induced up-regulation of peripheral mechanical and heat sensitivity (Fig. 9C and D). The transient inhibition of paclitaxel-induced mechanical allodynia by DHA was significantly compromised after 3-TYP application (Fig. 9E and F). Also, spinal suppression of SIRT3 was sufficient to eliminate analgesic role of DHA in bone cancer pain mice with sarcoma cells implantation (Fig. 9G and H). Collectively, these behavioral data further illustrated that SIRT3 might be a therapeutic target of DHA analgesia in acute and chronic pain conditions.