Background

Postoperative pain is the most common health problem caused by tissue and nerve lesions after surgical injury. Approximately 80% of patients who experience surgical injury have acute pain postoperatively, and this progresses to severe chronic pain in over 20% of patients [1, 2]. However, effective treatments for postoperative pain are still scarce [3,4,5].

Neuroinflammation mediated by glia, especially microglia, induces central sensitization and plays a crucial role in the pathogenesis of pain [4]. Studies have shown that matrix metalloproteinases (MMPs) represent a novel mechanism and potential therapeutic target for neuroinflammation and pain [6,7,8]. The main MMPs involved in the generation and maintenance of pain include MMP-2 and MMP-9. After nerve injury, a rapid and transient increase in MMP-9 in early activated microglia by the cleavage of IL-1β and the activation of p38 induces neuropathic pain, while MMP-2 produces a delayed increase in maintaining neuropathic pain [8, 9]. Further studies have shown that MMP-9 inhibition suppresses astrocyte and microglia activation, and thus alleviates pain, while MMP-9 knockout significantly reduces neuralgia and tolerance to morphine [8,9,10]. Our previous study showed that paeoniflorin inhibits MMP9/2 and suppresses postoperative pain induced by plantar incisions [11]. Therefore, inhibition of MMP9-induced neuroinflammation may be a novel approach for the treatment of postoperative pain.

We explored signaling molecules upstream of MMP-9 to identify key targets for the inhibition of neuroinflammation. Apoptosis signal-regulating kinase 1 (ASK1), a member of the mitogen-activated protein kinase (MAP3K) family, activates JNK, MAPK, and p38 MAPK to induce the expression of MMP9 and amplify the inflammatory cascade; hence, it participates in the induction and maintenance of neuroinflammation [Behavioral analysis

Before baseline testing, all animals were placed in the testing environment for at least 2 days for acclimatization. The first researcher was responsible for numbering the animal groups and drugs, as well as for data analysis. The second researcher administered the drugs to the groups according to the assigned numbers, while the third researcher performed the behavioral tests. The staff involved were blinded. Mechanical sensitivity was detected using von Frey hairs (Woodland Hills, Los Angeles, CA, USA). The animals were placed in boxes with an elevated metal mesh floor for habituation 30 min before testing. A series of von Frey hairs (the pressure intensity of mice between 0.16 g and 0.008 g) with logarithmically incrementing stiffness were used to perpendicularly stimulate the plantar surface of each hind paw. Each mouse was tested three times and the average of the threshold was measured.

Thermal hyperalgesia was detected using an analgesia meter (UGO Basile, Gemonio, Varese, Italy). Briefly, each mouse was placed on a 55 °C hot-plate apparatus. The heat source was focused on a portion of the hind paw, which was flush against the glass, and a radiant thermal stimulus was delivered to that site. The stimulus was then shut off when the hind paw moved (or after 20 s to prevent tissue damage). The intensity of the heat stimulus was kept constant throughout all the experiments. The elicited paw movement occurred at a latency of between 9 and 14 s in control animals. Thermal stimuli were delivered three times to each hind paw at 5- to 6-min intervals. Behavioral tests were performed blindly.

Gelatin zymography

The lumbar spines (L4–L5) of mice were collected and analyzed at 24 h or 5 days after surgery. Mice were anesthetized, and spinal cord segments were rapidly dissected and homogenized in 1% Nonidet P-40 lysis buffer. The solubilized proteins were then resolved on gels (8% polyacrylamide gel containing 0.1% gelatin). After electrophoresis, each gel was incubated with 50 mL of develo** buffer for 48 h (37.5 °C) in a shaking bath. Finally, the gels were stained with Coomassie Brilliant Blue (1%, with 10% acetic acid and 10% isopropyl alcohol, diluted with double-distilled H2O).

Western blot analysis

Lumbar spine samples (L4–L5) were collected and analyzed at 24 h or 5 days after surgery. Cell samples were collected and analyzed 24 h after H2 treatment. Protein concentrations were determined by the BCA protein assay (Thermo Fisher Scientific, Waltham, MA, USA), and equal amounts of protein per lane were separated using 8–15% sodium dodecyl sulfate–polyacrylamide gel, and transferred to polyvinylidene fluoride membranes (Millipore Corp., Bedford, MA, USA). After being blocked with 5% bovine serum albumin and 5% skim milk for 2 h at room temperature, the membranes were incubated overnight at 4 °C with primary antibodies and then incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies for 2 h at room temperature. The primary antibodies used included thioredoxin (1:500), p-p38 (1:1000), p-JNK (1:1000), p-ERK1/2 (1:1000), p-ASK1(1:1000), IL-1b (1:500), and IBA-1 (1:1000). For loading control, the blots were probed with an antibody for β-actin (1:5000). The filters were then developed by enhanced chemiluminescence reagents (PerkinElmer, Waltham, MA, USA) with secondary antibodies, Sigma (St. Louis, MO, USA). Finally, data were acquired using the Molecular Imager (Gel DocTM XR, 170–8170) and analyzed using the associated software Quantity One 4.6.5 (Bio-Rad Laboratories, Hercules, CA).

Immunofluorescence staining

The lumbar spines (L4–L5) of the mice were collected and analyzed 5 days postoperatively. After deep anesthesia, the animal was perfused transcardially with normal saline followed by 4% paraformaldehyde. The lumbar segment L4 and/or L5 were dissected and postfixed in the same fixative. The embedded blocks were sectioned to a thickness of 20 μm. The sections of each group (six animals in each group) were then incubated with rabbit antibody for IBA-1 (1:100, #ab178846, abcam, Cambridge, MA, USA), mouse antibody for MMP-9 (1:100, #10375-2-AP, Proteintech, Rosemont, IL USA), secondary antibody (1:300, Alexa Fluor 488 AffiniPure Donkey Anti-Rabbit/Mouse IgG, #711–545-152/#715–545-150, Jackson ImmunoResearch Laboratories, USA) at room temperature. After being washed three times with PBS, all slides were processed blindly and then studied under a confocal microscope (Olympus FV1000 confocal system, Olympus, Japan) to observe morphological details after staining. Images were randomly coded and fluorescence intensities were analyzed using Image Pro Plus 6.0 software (Media Cybernetics Inc., Rockville, MD, USA). The average green and red fluorescence intensity of each pixel was normalized to the background intensity in the same image.

Immunohistochemistry staining

The lumbar spines (L4–L5) of the mice were collected and analyzed at 24 h or 5 days after surgery. The samples were sectioned to 5 μm thickness, then incubated with the first antibodies for thioredoxin (1:100), mouse antibodies MMP-9 (1:100) and rabbit antibodies for IBA-1(1:300) in 10% donkey serum and 0.3% Triton-X100. After quenching endogenous peroxidase activity, the slides were washed in PBS and incubated with HRP conjugated secondary antibody for 2 h. Diaminobenzidine was used as a chromogen and counterstaining was performed with hematoxylin. Two independent pathologists evaluated all immunohistochemistry staining sections. The score for each slide was measured as the cross-product of the value of immunostaining intensity and the value of the proportion of positive-staining cells. The intensity of immunostaining was divided into four grades: 0, negative; 1, weak; 2, medium; 3, strong. The proportion of positive-staining cells was also divided into four grades: 1, 0 − 25%; 2, 26 − 50%; 3, 51 − 75%; and 4, > 75%. The score was calculated using the following formula: total score = intensity score × proportion score.

Statistical analysis

GraphPad Prism 6 software (GraphPad Software, San Diego, CA, USA) was used to conduct all statistical analysis. Continuous variables were presented as means ± SEM. Normally distributed alteration of protein detected expression and changes in behavioral responses were tested with Student’s t tests and one-way ANOVA. The differences in latency over time among groups were tested with two-way ANOVA. Bonferroni post hoc comparisons was performed between multiple groups. A criterion value of P < 0.05 was considered statistically significant.

Results

MMP-9 participates in the development of plantar incision-induced postoperative pain

To explore the role of MMP-9 in the process of plantar incision-induced postoperative pain, the activity of MMP-9 and the mechanical threshold of mice (WT or MMP-9−/−) were measured. As shown in Fig. 1A, compared with the sham group, plantar incision surgery significantly decreased the mechanical threshold of the mice for up to 5 days, and caused severe postoperative pain. Plantar incision surgery increased the activity and the expression of MMP-9 in the spinal cord of mice collected 5 days after the operation (Fig. 1B, C). Interestingly, MMP-9 knockout significantly alleviated postoperative pain compare to the WT + PI group (Fig. 1D). These results suggest that plantar incision surgery could be successfully used to establish a postoperative pain model, and that MMP-9 plays a vital role in the development of postoperative pain.

Fig. 1
figure 1

MMP-9 plays a crucial role in the development of plantar incision-induced postoperative pain. A Mechanical threshold of mice that received plantar incision surgery (n = 12). B MMP-9 activity was measured using zymography after surgery in the spinal cord of mice (n = 4). C MMP-9 expression in the spinal cord of mice was measured using western blot after surgery (n = 4). Lumbar spines (L4–L5) were collected and analyzed 5 days after surgery. D Mechanical threshold of wild-type or MMP-9 KO mice that underwent plantar incision surgery (n = 6). PI = plantar incision. Significant difference is revealed following two-way ANOVA (A and D) or unpaired Student’s t-test (B, C) (*P < 0.05, **P < 0.01, ***P < 0.001 vs. sham; &P < 0.05, &&P < 0.01 vs. PI group; Bonferroni post hoc tests)

Full size image

ASK1 mediates MMP-9 activation and facilitates the development of postoperative pain

Considering the important role of ASK1 in inflammation and apoptosis [29]. Our previous study showed that plantar incisions could increase MMP-9 activity, activate microglia, and promote the development of postoperative pain [11]. In this study, we further showed the importance of MMP-9 in the development of postoperative pain using MMP-9 knockout mice (Fig. 1D). Furthermore, MMP-9 co-localized with microglia IBA-1 and aggravated the cleavage of IL-1β (Fig. 4A). We also found that inhibition of ASK1 could ameliorate pain and reduce MMP-9 activity (Fig. 2), suggesting that MMP-9 and its upstream ASK1 are associated with the facilitation of postoperative pain. Regulation of ASK1/MMP9 in microglia may receive considerable attention as a potential therapeutic target for ameliorating postoperative pain.

HRS and H2 have been confirmed to alleviate hyperpathia and activate autophagy in neuropathic pain models [24, 30]. In this study, HRS was used for the first time to treat postoperative pain. In accordance with the effects of H2 on p-ASK1, p-p38, and p-JNK (Fig. 3A-C), H2 decreased MMP-9 activity and expression, inhibited the expression of IL-1β, IL-6, and TNF-α (Fig. 3D-I), and effectively attenuated incision-induced postoperative pain (Fig. 3J).

Furthermore, we explore how H2 regulates the phosphorylation of ASK1, p38, and JNK. Trx1 is an endogenous 12 kDa multifunctional protein with two redox-active half-cysteine residues -Cys-Gly-Pro-Cys- [31]. It has been identified in all living cells and is related to cell proliferation and apoptosis, where it is responsible for protecting cells from oxidative stress by scavenging ROS [31, 32]. Experimental results show that intravenous administration of recombinant human thioredoxin and overexpression of Trx1 in transgenic mice confer resistance to ROS-induced cell death, ultimately decreasing brain damage in cerebral ischemia models [33, 34]. Trx1 overexpression extends antioxidant protection, attenuates mitochondrial damage, and prolongs survival during sepsis [35]. Furthermore, along with the decrease in Trx1 level, NLRP3 expression increases inflammation in injured tissue [36]. Since Trx1 exerts its role by interacting with its binding protein, ASK1, it inhibits the activation of ASK1 [14]; moreover, our data showed that H2 could increase the expression of Trx1 (Fig. 5A, D, E) and attenuate postoperative pain, which was abolished by the Trx1 inhibitor PX12 (Fig. 5G), these data indicate that Trx1 is an endogenous neuroprotective protein that is involved in proves through which H2 reduces postoperative pain. To further investigate the protection effects of H2, we collected and analyzed BV-2 cells after H2 treatment. We found that H2 could mimic the ASK1 inhibitor NQDI1 as it decreases the phosphorylation of ASK1, p38, and JNK, and reduces MMP-9 activity and expression (Fig. 6A–F). We also found that the protective effects of H2 were abolished by the Trx1 inhibitor, PX12, in BV-2 cells (Fig. 6G–N). These data demonstrate that H2 could attenuate postoperative pain by regulating the Trx1/ASK1/MMP-9 signaling pathway, and provides further details regarding the mechanism of H2 therapy.

H2 has been shown to be safe with few adverse effects. Compared to vitamin E and superoxide dismutase, H2 is a selective antioxidant that reduces cytotoxic oxygen radicals [18]. Inhalation of H2, drinking hydrogen water, injection of hydrogen saline, and direct incorporation of molecular hydrogen by diffusion, including eye drops, baths, and cosmetics, are the main methods of ingesting or consuming H2 [19]. Inhalation of H2 suppresses not only the initial brain injury, but also its progressive damage [18]. H2 has beneficial effects including the promotion of microglia M2 polarization and the reduction of inflammation [20, 23]. Interestingly, oral administration of hydrogen water was reported to alleviate neuropathic pain in mice by reducing oxidative stress. Since oxidative stress injury is an important pathological mechanism of postoperative pain [37], oral administration of hydrogen water may also be useful in attenuating postoperative pain. The oral pathway is more conducive to the promotion of the clinical application of H2.

One of the limitations of the research is that we did not consider gender differences in H2 therapy. We only considered the response of male mice to postoperative pain, while ignoring the difference in magnetic responses to pain. We will explore this in future research.

Conclusions

In summary, we demonstrated that H2 attenuates postoperative pain induced by plantar incisions by inhibiting the ASK1/JNK/p38/MMP9 signaling pathway and microglial activation, and this may be related to Trx1 (Fig. 7). Our findings suggest that H2 may be a potential drug for the treatment of postoperative pain.

Fig. 7
figure 7

H2 protects against plantar incision-induced postoperative pain by upregulating the Trx1/ASK1/MMP-9 signaling pathway. Plantar incision surgery increases ASK1/JNK/p38 phosphorylation mediated microglia activation, enhances MMP-9 activity and IL-1β cleavage, and subsequently induces neuroinflammation, which contributes to the progression of postoperative pain in mice. Inhalation of H2 and administration of HRS, which increases Trx1 expression, could ameliorate postoperative pain

Full size image