Introduction

Depression and pain mutually aggravate the sub-thresholds for painful and depressive symptoms, conferring a clinically challenging healthcare problem [1–3]. Pharmacological treatment of comorbid depression and pain is one of the unmet medical needs worldwide. Previous attempts to treat comorbid depression and pain include the combined therapy with selective serotonin reuptake inhibitor (SSRI) anti-depressants (e.g., citalopram, Zoloft and Prozac) and non-steroidal anti-inflammatory drugs (NSAIDs) such as aspirin and ibuprofen. However, current anti-depressant therapies are dampened by several serious side effects such as emotional detachment, suicide tendency, sexual dysfunctions, hypertension risks, and weight gain [4, 5]. Moreover, various NSAIDs could unexpectedly diminish the effectiveness of SSRI anti-depressants [6, 7]. Such a disappointment is possibly due to the adverse interactions between these drugs. Based on previous pharmacological studies, SSRIs sequentially release inflammatory cytokines (e.g., tumor necrosis factor alpha (TNF-α), interferon gamma (IFN-γ)) and induce brain-derived neurotrophic factor (BDNF) expression against depression [7–9]. In contrast, the common analgesics resolve inflammation by inhibiting the release of TNF-α and IFN-γ [10, 11]. It appears to be difficult to relieve pain from brain and body simultaneously. As an alternative strategy, a clinical protocol of “sequenced treatment alternatives to relieve depression” was recently proposed to achieve the best therapeutic outcomes [7, 12]. Nevertheless, the side effects and the adverse drug-drug interactions have limited the combination therapy with two or more medicines for the treatment of multifactorial diseases.

Plant natural products are well documented for the efficacy in the treatment of complex chronic comorbidities (e.g., depression and pain) in China [13–15]. We have recently evaluated the neuroprotective and neurotrophic activities of a panel of plant natural products [16–18]. Of the tested herbal compounds, puerarin is an interesting C-glucosylated isoflavone derived from the herb Radix Puerariae Lobatae. Previous studies suggest that puerarin may be a potent antioxidant and anti-inflammatory agent [19–21]. In addition, puerarin could alleviate neuropathic pain via modulating the cell signals of P2X(3) receptors in dorsal root ganglion neurons [22]. Interestingly, we found that puerarin could not only coordinate with nerve growth factor (NGF) to promote neuronal differentiation but also protect neurons against oxidative injuries via inducing arginase-2 expression [18, 23]. These results highlight the potential of puerarin in the treatment of pain originated in the central or peripheral nervous system.

In the present study, we investigated the anti-depressant and analgesic activities of puerarin in a well-characterized mouse model of spared nerve injury (SNI). Depressive-like behaviors were assessed by forced swim test and tail suspension test. Pain responses were tested by measuring the sensitivity to the mechanical stimulation by von Frey filaments. We also explored the molecular mechanisms by determining the levels of reduced glutathione (GSH) and superoxide dismutase (SOD) and analyzing the expression of signaling molecules (i.e., ERK and CREB) and BDNF.

Materials and Methods

Antibodies and Biochemical Reagents

Antibodies against ERK, phospho-ERK, CREB, phospho-CREB, and GAPDH were purchased from Cell Signaling Technology (Boston, MA, USA). The anti-rabbit HRP-conjugated IgG secondary antibody was purchased from Sigma–Aldrich (St. Louis, MO, USA). Polyclonal rabbit anti-BDNF antibody was obtained from Santa Cruz Biotechnology Inc. (Santa Cruz, CA, USA). Alexa Fluor 594-conjugated goat anti-rabbit IgG secondary antibody was purchased from Invitrogen (Carlsbad, CA, USA). Protein Assay Dye Reagent Concentrate was purchased from Bio-Rad (Hercules, CA, USA). Select chemiluminescence (ECL) detection kit was purchased from GE Healthcare (Uppsala, Sweden). Puerarin, citalopram, ibuprofen, and other biochemical reagents unless otherwise indicated were purchased from Sigma–Aldrich (St. Louis, MO, USA).

Animal Husbandry

Adult male C57BL/6 N mice (8 weeks, 22–25 g) were supplied by the Laboratory Animal Unit of the University of Hong Kong. All experimental procedures were in compliance with the guidelines of the Committee on the Use of Live Animals in Teaching and Research of the University of Hong Kong (CULATR 3470-14). Animals were housed in a temperature- and humidity-controlled environment on a 12-h light–dark cycle and allowed free access to standard laboratory mice chow and drinking water. For all behavioral tests, each animal was tested only once to avoid potential carry-over effects. Mice from six different litters were used to minimize litter effects. Behavioral tests were performed between the hours of 09:00 to 15:00 on a 06:00 to 18:00 on-off light cycle to control circadian variations.

Drug Administration

Puerarin, citalopram, and ibuprofen were dissolved in 50 % 1,2-propylene glycol in saline. All the solutions were sterilized by passing through a 0.22-μm membrane filter (Pall Corporate, Port Washington, NY, USA).

Seventy mice were used for two sets of experiments: (1) determination of dose dependence and (2) comparison and combination with existing drugs. For experiment 1, 30 mice were randomly divided into five groups (n = 6): Sham group, SNI group, SNI + puerarin (30 mg/kg) group, SNI + puerarin (60 mg/kg) group, SNI + puerarin (120 mg/kg) group. For experiment 2, 40 mice were randomly divided into eight experimental groups (n = 5): Sham group, SNI group, SNI + puerarin group, SNI + citalopram group, SNI + ibuprofen group, SNI + puerarin + citalopram group, SNI + citalopram + ibuprofen group, SNI + puerarin + ibuprofen group. Based on previous experiments on drug doses [7, 18], mice received intragastric administration of drugs (e.g., puerarin, 120 mg/kg/day; citalopram, 10 mg/kg/day; ibuprofen, 70 mg/kg/day) for 8 or 14 consecutive days, while Sham and SNI animals received the same volume of 50 % 1,2-propylene glycol in saline. The related experiments were performed according to the protocol outlined in Fig. 1a.

Fig. 1
figure 1

Anti-depressant activities of puerarin, citalopram, and ibuprofen. a Experimental design. Mice were randomly divided into eight groups: Sham and seven spared nerve injury (SNI) groups. After 7-day post-SNI recovery, animals were treated with indicated drugs or vehicle for 8 or 14 days. On day 18 after surgery, mice were assessed for depressive behaviors by forced swim test (FST) and tail suspension test (TST). Pain response was examined by von Frey filaments on days – 1, 1, 7, 12, and 17. Brain tissues were collected for analyses of ERK, CREB, and BDNF. b FST evaluation of dose dependence. SNI animals were treated with puerarin (0, 30, 60, 120 mg/kg) whereas sham animals were treated with vehicle only. In FST evaluation, the immobility time was recorded during 4 min for each animal. The data (mean ± SEM, n = 6) were analyzed by one-way ANOVA followed by post hoc Dunnett’s test. ### p < 0.001 (SNI vs Sham); ***p < 0.001 (Drug vs SNI only). P30, puerarin (30 mg/kg); P60, puerarin (60 mg/kg); P120, puerarin (120 mg/kg). c TST evaluation of dose dependence. The animals were treated as described in Fig. 1b. In TST evaluation, the immobility time was recorded during 6 min for each animal. The data (mean ± SEM, n = 6) were analyzed by one-way ANOVA followed by post hoc Dunnett’s test. ### p < 0.001 (SNI vs Sham); ***p < 0.001 (Drug vs SNI only). P30, P60, and P120 were defined as in Fig. 1b. d FST comparison of puerarin with citalopram and ibuprofen. SNI animals were treated with puerarin (120 mg/kg), citalopram (10 mg/kg), ibuprofen (70 mg/kg), or vehicle, whereas sham animals were treated with vehicle only. FST was performed as described in Fig. 1b. The data (mean ± SEM, n = 5) were analyzed by one-way ANOVA followed by post hoc Dunnett’s test. ### p < 0.001 (SNI vs Sham); **p < 0.01; ***p < 0.001 (Drug vs SNI only). e TST comparison of puerarin with citalopram and ibuprofen. The animals were treated as described in Fig. 1d. TST was performed as described in Fig. 1c. The data (mean ± SEM, n = 5) were analyzed by one-way ANOVA followed by post hoc Dunnett’s test. ### p < 0.001 (SNI vs Sham); *p < 0.05; ***p < 0.001 (Drug vs SNI only). f FST evaluation of the combinations between puerarin, citalopram, and ibuprofen. SNI animals were treated with puerarin (120 mg/kg), citalopram (10 mg/kg), ibuprofen (70 mg/kg), in combination, or vehicle, whereas sham animals were treated with vehicle only. FST was performed as described in Fig. 1b. The data (mean ± SEM, n = 5) were analyzed by one-way ANOVA followed by post hoc Dunnett’s test. ### p < 0.001 (SNI vs Sham); *p < 0.05; ***p < 0.001 (Drug vs SNI only). g TST evaluation of the combinations between puerarin, citalopram, and ibuprofen. The animals were treated as described in Fig. 1f. TST was performed as described in Fig. 1c. The data (mean ± SEM, n = 5) were analyzed by one-way ANOVA followed by post hoc Dunnett’s test. ### p < 0.001 (SNI vs Sham); **p < 0.01; ***p < 0.001 (Drug vs SNI only). P puerarin, C citalopram, I ibuprofen

Full size image

SNI Model of Depression and Pain Comorbidity

SNI surgery was performed essentially as previously described [24]. Briefly, mice were anesthetized by intraperitoneal (i.p.) injection of ketamine (100 mg/kg) and xylazine (15 mg/kg). After the mouse skin on the lateral surface of the left thigh was cleaned and incised, sterile scissors were used to expose three branches of the sciatic nerves: namely sural, common peroneal, and tibial nerves, in the left thigh. The common peroneal and tibial nerves were tied with nonabsorbent 6-0 silk sutures at the point of trifurcation, whereas the sural nerve was untouched. The tied nerves were grabbed below the sutures with a pair of tweezers and cut above and below the tweezers. In Sham group, these nerves were dissected but not cut. At the end of surgery, muscle and skin layers were closed with sutures in distinct layers.

Behavioral Test

Mice were tested for depressive-like behaviors and pain response using a battery of widely recognized methods as follows:

Tail Suspension Test

Animals were kept in a quiet experimental room. Tail suspension test was performed as previously described [25]. In brief, each mouse was suspended by its tail and secured by adhesive tape to the suspension bar (1 cm height, 1 cm width, 60 cm length), leaving a gap of 20–25 cm between mouse nose and apparatus floor. Adhesive tape was applied to the tail at the position of 2–3 mm to the very end. A blinded observer monitored the test for 6 min and recorded the immobility time with a stopwatch. The mouse was considered immobile only when the mice hung passively and completely motionless.

Forced Swimming Test

The mice were forced to swim in an acrylic plastic cylinder filled with water as described [26]. Briefly, mice were placed individually into plastic cylinders (height, 30 cm; diameter, 20 cm) filled with 15 cm of water, conditioned at 23 ± 2 °C. After 2-min habituation for mice, a blinded observer recorded the time when the mice remained floating passively or immobile in the water for a period of 4 min.

Von Frey Monofilament Assay

Von Frey monofilament test was performed as previously described [27]. In brief, mice were individually placed in red plastic chambers and habituated for 15 min prior to the test. Von Frey filaments were used to stimulate the lateral third of left paws of animals initially with the 1.0 g filament but subsequently with logarithmically incremental stiffness from 0.008 to 8.0 g 1 day before the surgery or on day 1, 7, 12, and 17 after the surgery. Positive response was defined by sudden paw withdrawal, sudden flinching, and sudden paw licking. Response in three out of five stimuli was regarded as a positive reaction and the threshold level was recorded.

Western Blot Analysis

At the end of drug treatment, the olfactory bulbs were removed, whereas the first 5 mm of the entire cortex was recovered as mouse forebrain cortex for further analyses. The cortex tissues were sliced into small pieces and then lysed in ice-cold RIPA buffer containing 150 mM NaCl, 1.0 % IGEPAL® CA-630, 0.5 % sodium deoxycholate, 0.1 % SDS, and 50 mM Tris, (pH 8.0) with sonication. The proteins were recovered by centrifugation at 13,000 rpm for 15 min at 4 °C. The protein concentrations were determined with protein assay dye reagent from Bio-Rad (Hercules, CA, USA). The brain proteins (70 μg) were resolved by gel electrophoresis on 12 % SDS-polyacrylamide gels and subsequently transferred onto PVDF membranes. Following overnight blocking with 5 % BSA in Tris-buffered saline containing 0.1 % Tween-20 (TBS-T) buffer at 4 °C, the blots were probed with specific primary antibodies, detected with HRP-conjugated secondary antibody, and finally visualized with Amersham™ ECL™ Select Western blotting detection reagent from GE Healthcare Biosciences (Uppsala, Sweden).

Determination of GSH Contents

GSH contents in brain tissues were determined by using a commercial glutathione kit (Jiancheng Bioengineering Ltd, Nanjing, China) according to the manufacturer’s instruction. Briefly, the brain tissue lysates were first treated with 5 % 5-sulfosalicylic acid solution to remove the proteins. Based on the reactivity of GSH to reduce 5,5′-dithiobis-(2-nitrobenzoic) acid (DTNB), the production of TNB was correlated with GSH contents. TNB levels were quantified by a colorimetric measurement of the absorbance at 412 nm. The GSH concentration was calculated as μmol/g proteins relative to that of GSH standard.

Assay of SOD Activity

The SOD activity in the brain tissues was measured by using a commercial SOD assay kit (Jiancheng Bioengineering Ltd, Nanjing, China) according to the manufacturer’s instruction. Briefly, the brain tissue homogenates were generated as described for Western blot analysis. The supernatants were incubated with a new tetrazolium compound, WST-1, a sodium salt of 4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate, xanthine oxidase, and hypoxanthine. A unit of SOD activity is defined as achieving 50 % inhibition of formazan formation. The SOD activity in the brain tissues was expressed as units/mg total proteins.

Immunofluorescence Staining

At the end of drug treatment, the mice were anesthetized with a ketamine/xylazine mixture solution and perfused with 4 % paraformaldehyde in 0.01 M phosphate-buffered saline (PBS) (pH 7.4). Mouse brains were recovered, further fixed in 4 % paraformaldehyde for 3 days at 4 °C, and then incubated in 30 % sucrose in PBS overnight at 4 °C. Following frozen for 1 h at −80 °C, the brain tissues were cut into serial coronal sections with the thickness of ∼18 μm. The anterior cingulate cortex sections were blocked with 5 % normal goat serum and 0.3 % Triton X-100 in PBS for 2 h at room temperature and then incubated with polyclonal antibodies against BDNF overnight at 4 °C. After the removal of excessive antibodies, the brain sections were detected with Alexa Fluor 594-conjugated goat anti-rabbit IgG secondary antibody. The cell nuclei were subsequently stained with 4′-6-diamidino-2-phenylindole (DAPI) for 10 min. The immunofluorescence images were acquired on a fluorescence microscope (Carl-Zeiss Jena, Germany). The fluorescence intensity was measured using NIH ImageJ software (http://imagej.net/ImageJ2).

Data Analysis

The results were expressed in mean ± SEM (n = 5 or 6). The differences between the experimental groups were analyzed by two-tailed paired student t test or one-way analysis of variance (ANOVA), followed by post hoc Dunnett’s test with GraphPad Prism 6 software (La Jolla, CA, USA). The p values less than 0.05 were considered to be significantly different.

Results

Puerarin Ameliorated the Depressive-Like Behaviors in Mice with SNI Surgery

To examine the anti-depressant activity of puerarin, we evaluated the effect of puerarin on the behaviors of SNI mice in tail suspension test and forced swim test as previously described [7]. Our experiments were designed as outlined in Fig. 1a. SNI surgery was performed to induce peripheral neuropathic pain in mice. After the surgery, mice were allowed to recover spontaneously for 7 days. The animals were subsequently treated with puerarin, citalopram, and ibuprofen individually or in combination. On day 18 after surgery, mice were evaluated in forced swim test and tail suspension test (Fig. 1b–g). SNI surgery significantly prolonged immobile time than sham surgery in forced swim test (F value = 16.9, p < 0.001) and tail suspension test (F value = 13.83, p < 0.001). Firstly, based on the evaluation of dose dependence, puerarin (60, 120 mg/kg) significantly reduced immobile time in SNI mice compared with vehicle only in forced swim test (F value = 16.9, p < 0.001 for each dose) and tail suspension test (F value = 13.83, p < 0.001 for each dose). In contrast, puerarin (30 mg/kg) did not show strong activity. Secondly, the anti-depressant effects of puerarin, citalopram, and ibuprofen were compared in behavioral assessments. The time during which SNI mice remained immobile was reduced to a similar extent by puerarin (forced swim test: F value = 12.49, p < 0.01; tail suspension test: F value = 31.54, p < 0.001) vs citalopram (forced swim test: F value = 12.49, p < 0.001; tail suspension test: F value = 31.54, p < 0.001) but to a lesser extent by ibuprofen (forced swim test: F value = 12.49, p > 0.05; tail suspension test: F value = 31.54, p < 0.05). Thirdly, two drug combinations of puerarin, citalopram, and ibuprofen were also evaluated. Interestingly, puerarin and citalopram in combination significantly reduced the immobility time against SNI surgery (forced swim test: F value = 13.98, p < 0.001; tail suspension test: F value = 11.16, p < 0.001). Puerarin and ibuprofen in combination showed similar effects on SNI-induced immobility (forced swim test: F value = 13.98, p < 0.01; tail suspension test: F value = 11.16, p < 0.01). In contrast, ibuprofen profoundly attenuated the anti-depressant activity of citalopram (forced swim test: F value = 13.98, p < 0.05; tail suspension test: F value = 11.16, p > 0.05).

Puerarin Reduced the Sensitivity to Mechanical Stimulation

To examine the analgesic activity of puerarin, we evaluated the effect of puerarin on mechanical allodynia-like behavior in SNI mice with von Frey monofilament as previously described [24, 28]. The drug treatments were carried out essentially as described for the assay of depressive-like behaviors. Mice were evaluated for the sensitivity to mechanical stimulation by von Frey monofilament test 1 day before surgery or on day 1, 7, 12, and 17 after SNI surgery. We compared the thresholds of pain response at the ipsilateral and contralateral hindpaws. For the evaluation of dose dependence as shown in Fig. 2a, SNI significantly reduced the thresholds of ipsilateral pain response to the mechanical stimulation in mice compared with sham surgery (p < 0.01 for test on day 1; p < 0.001 for each test on day 7, 12, and 17). In another set of the pain test for the comparison experiments as shown in Fig. 2b, c, SNI also reduced the thresholds of ipsilateral pain response to the mechanical stimulation in mice compared with sham surgery (p < 0.05, p < 0.01, and p < 0.05 for test on day 1, 7, and 12, respectively). In contrast, we did not detect any difference in such response at the contralateral hindpaws in these two sets of experiments. Furthermore, we analyzed the effects of different drugs on SNI-induced sensitization to pain stimulation by one-way ANOVA, followed by post hoc Dunnett’s test. As for the evaluation of dose dependence as shown in Fig. 2a, we discovered that puerarin at the dose of 60 mg/kg significantly improved SNI-induced pain behavior on day 12 (F value = 4.318, p < 0.05). For the comparison of three drugs as shown in Fig. 2b, we observed that puerarin at the dose of 120 mg/kg significantly improved SNI-induced pain behavior on day 12 (F value = 6.02, p < 0.05) and day 17 (F value = 4.477, p < 0.01). We also found that citalopram at the dose of 10 mg/kg significantly improved SNI-induced pain behavior on day 7 (F value = 2.986, p < 0.05) and day 17 (F value = 4.477, p < 0.05). In contrast, ibuprofen did not make any difference in SNI mice relative to sham animals. In the evaluation of two drug combinations as shown in Fig. 2c, interestingly, puerarin in combination with either citalopram or ibuprofen significantly increased the thresholds of pain response to mechanical stimulation on day 12 (F value = 5.338, p < 0.01; F value = 5.338, p < 0.05). On the other hand, citalopram in combination with ibuprofen significantly increased the thresholds of pain response to mechanical stimulation on day 1 (F value = 3.384, p < 0.05), day 7 (F value = 5.311, p < 0.01), and day 17 (F value = 3.252, p < 0.05).

Fig. 2
figure 2

Effects of puerarin, citalopram, and ibuprofen on SNI-induced sensitization to mechanical stimulation. a Evaluation of dose dependence. As outlined in Fig. 1a, SNI animals were treated with puerarin (0, 30, 60, 120 mg/kg) whereas sham animals were treated with vehicle only. Pain response was examined by mechanical stimulation to the ipsilateral and contralateral hindpaws with von Frey filaments on days 1, 1, 7, 12, and 17. The data (mean ± SEM, n = 6) were analyzed by one-way ANOVA followed by post hoc Dunnett’s test. ## p < 0.01; ### p < 0.001 (SNI vs Sham); *p < 0.05; **p < 0.01; ***p < 0.001 (Drug vs SNI only). P30 puerarin (30 mg/kg), P60 puerarin (60 mg/kg), P120 puerarin (120 mg/kg). b Comparison of puerarin with citalopram and ibuprofen. As outlined in Fig. 1a, SNI animals were treated with puerarin (120 mg/kg), citalopram (10 mg/kg), ibuprofen (70 mg/kg), or vehicle, whereas sham animals were treated with vehicle only. Pain response was assessed as described in Fig. 2a. The data (mean ± SEM, n = 5) were recorded and analyzed by one-way ANOVA followed by post hoc Dunnett’s test. # p < 0.05; ## p < 0.01 (SNI vs Sham); *p < 0.05; **p < 0.01 (Drug vs SNI only). c Evaluation of the combinations between puerarin, citalopram, and ibuprofen. As outlined in Fig. 1a, SNI animals were treated with puerarin (120 mg/kg), citalopram (10 mg/kg), ibuprofen (70 mg/kg), in combination, or vehicle, whereas sham animals were treated with vehicle only. Pain response was assessed as described in Fig. 2a. The data (mean ± SEM, n = 5) were recorded and analyzed by one-way ANOVA followed by post hoc Dunnett’s test. # p < 0.05; ## p < 0.01 (SNI vs Sham); *p < 0.05; **p < 0.01 (Drug vs SNI only). P puerarin, C citalopram, I ibuprofen

Full size image

Puerarin Marginally Reduced Oxidative Stress in SNI Mice

To elucidate the anti-depressant and analgesic mechanisms, we first examined the effects of puerarin, citalopram, and ibuprofen on redox homeostasis in SNI mice. We treated SNI mice with puerarin, citalopram, and ibuprofen, individually or in combination. On day 18 after surgery, the brain tissues were harvested and analyzed for the levels of GSH and SOD. As shown in Fig. 3a, SNI caused a significant decrease in GSH levels (p < 0.05). Citalopram significantly recovered the loss of GSH in SNI mice (p < 0.05), whereas puerarin and ibuprofen did not cause much alteration of GSH contents. Interestingly, puerarin, citalopram, and ibuprofen in combination effectively prevented the loss of GSH in SNI mice (p < 0.05 or p < 0.01). As shown in Fig. 3b, however, puerarin, citalopram, and ibuprofen, individually or in combination, increased the SOD activity to a certain extent but not to a significant extent.

Fig. 3
figure 3

Effects of puerarin, citalopram, and ibuprofen on redox status. a Effects of puerarin, citalopram, and ibuprofen on the GSH contents in brains. As outlined in Fig. 1a, SNI mice were treated with puerarin (120 mg/kg), citalopram (10 mg/kg), and ibuprofen (70 mg/kg), alone or in combination, or vehicle, whereas sham animals were treated with vehicle only. On day 18 after surgery, the brain tissue lysates were prepared and precipitated to remove the proteins. The GSH contents in the supernatants (n = 3) were assayed by commercial kit as described in “Methods.” The results (mean ± SEM) were analyzed by one-way ANOVA followed by post hoc Dunnett’s test. # p < 0.05 (SNI vs Sham); *p < 0.05; **p < 0.01 (Drug vs SNI only). b Effects of puerarin, citalopram, and ibuprofen on the SOD activity in brains. Mice were treated was described in Fig. 3a. On day 18 after surgery, the SOD activity in the brain tissue lysates (n = 3) was assayed by commercial kit as described in “Methods.” The results (mean ± SEM) were analyzed by one-way ANOVA followed by post hoc Dunnett’s test

Full size image

Puerarin Promoted the Phosphorylation of ERK1/2 and CREB in SNI Mice

We subsequently examined the effect of puerarin on the ERK1/2 and CREB signaling pathways in SNI mice. Following the drug treatment for 8 days, the proteins were extracted from the brain tissues and analyzed by Western blot with specific antibodies. As shown in Fig. 4, SNI dramatically diminished the phosphorylation of ERK1/2 (p < 0.01). Interestingly, puerarin and citalopram significantly induced the expression of p-ERK in SNI mice (p < 0.05 and p < 0.01, respectively), whereas ibuprofen did not show any activity. As far as CREB signaling pathway is concerned, SNI also diminished the phosphorylation of CREB (p < 0.05). Remarkably, puerarin was the only drug to induce the phosphorylation of CREB (p < 0.001). Neither of citalopram and ibuprofen recovered the expression of p-CREB in SNI mice, even in combinations, for example, with puerarin.

Fig. 4
figure 4

Effects of puerarin, citalopram, and ibuprofen on the activation of ERK1/2 and CREB pathways. a Activation levels of ERK1/2 and CREB pathways after 14-day drug treatment. As outlined in Fig. 1a, SNI mice were treated with puerarin (120 mg/kg), citalopram (10 mg/kg), and ibuprofen (70 mg/kg), alone or in combination, or vehicle, whereas Sham animals were treated with vehicle only. On day 18 after surgery, the brain tissue lysates were prepared and analyzed by Western blotting for ERK1/2, phospho-ERK1/2, CREB, phospho-CREB expression, whereas GAPDH was analyzed as control. Representative blot was shown. b Quantitative analysis (n = 3). The signal intensities of protein bands were determined by a densitometric method. The signals from three independent experiments were analyzed by one-way ANOVA followed by post hoc Dunnett’s test. # p < 0.05; ## p < 0.01 (SNI vs Sham); *p < 0.05; **p < 0.01; ***p < 0.001 (Drug vs SNI only)

Full size image

Puerarin Induced the Expression of BDNF in SNI Mice

We further investigated the effect of puerarin on the expression of BDNF in SNI mice. When SNI mice were treated with puerarin, citalopram, and ibuprofen, alone or in combination, for 2 weeks, the proteins were extracted from the brain tissues, and analyzed by Western blot with specific antibodies. As shown in Fig. 5a, b, SNI dramatically reduced the levels of BDNF protein compared with sham surgery (p < 0.001). Puerarin and citalopram significantly recovered the loss of BDNF expression in SNI mice (p < 0.001), whereas ibuprofen showed somewhat activity. Importantly, puerarin in combination with citalopram or ibuprofen also effectively recovered the expression of BDNF (p < 0.001 for each combination). Ibuprofen reduced the activity of citalopram to a lesser extent although BDNF level was still significantly higher than that in SNI animals (p < 0.01). Thus, we further examined whether puerarin and citalopram could rapidly induce BDNF expression. We treated mice with puerarin and citalopram, alone or in combination, for 3 days. The expression of BDNF was detected by Western blot analysis. As shown in Fig. 5c, puerarin effectively induced the expression of BDNF in mice, whereas citalopram failed to induce the expression of BDNF. Nevertheless, citalopram appeared to be able to potentiate puerarin-induced expression of BDNF. Collectively, we proposed that puerarin could induce BDNF expression via sequential activation of ERK and CREB pathways (Fig. 5d).

Fig. 5
figure 5

Effects of puerarin, citalopram, and ibuprofen on BDNF induction. a Western blot analysis of BDNF induction after 14-day drug treatment. As outlined in Fig. 1a, SNI mice were treated with puerarin (120 mg/kg), citalopram (10 mg/kg), and ibuprofen (70 mg/kg), alone or in combination, or vehicle, whereas Sham animals were treated with vehicle only. On day 18 after surgery, the brain tissue lysates were prepared and analyzed by Western blotting for BDNF expression, whereas GAPDH was analyzed as control. Representative blot was shown. b Quantitative analysis of BDNF induction after 14-day drug treatment (n = 3). The signal intensities of protein bands in Fig. 5a were determined by a densitometric method. The signals from three independent experiments were analyzed by one-way ANOVA followed by post hoc Dunnett’s test. ### p < 0.001 (SNI vs Sham); **p < 0.01; ***p < 0.001 (Drug vs SNI only). c Western blot analysis of BDNF induction after 3-day drug treatment. Mice were treated with puerarin (120 mg/kg) and citalopram (20 mg/kg), alone or in combination, for 3 days, whereas the controls were treated with vehicle only. The brain tissue lysates were prepared and analyzed by Western blotting for BDNF expression, whereas GAPDH was analyzed as control. Representative blot from two independent experiments was shown. d Potential mechanism. Puerarin sequentially activates ERK and CREB pathways, leading to the induction of BDNF expression

Full size image

In addition, we also examined BDNF levels in the brain tissue specimens by immunostaining with specific antibody. As shown in Fig. 6, BDNF-associated fluorescence was almost lost in SNI group compared with that in sham group (p < 0.05). Puerarin and citalopram, alone or in combination, effectively reverse the decrease in BDNF levels after SNI surgery (p < 0.001, p < 0.01, and p < 0.05, respectively). Puerarin and ibuprofen in combination showed marginal activity whereas ibuprofen alone or in combination with citalopram did not show any activity.

Fig. 6
figure 6

Immunofluorescence analysis of BDNF expression. a Representative image of immunofluorescence staining for BDNF expression. As outlined in Fig. 1a, SNI mice were treated with puerarin (120 mg/kg), citalopram (10 mg/kg), and ibuprofen (70 mg/kg), alone or in combination, or vehicle, whereas Sham animals were treated with vehicle only. On day 23 after surgery, brain tissues were collected and analyzed by immunostaining of BDNF sequentially with specific primary antibody and Alexa Fluor 594-conjugated secondary antibody, whereas DAPI was used to identify the cell nuclei as control. The images were captured under a fluorescence microscope (Carl-Zeiss, Jena, Germany). Arrows point to BDNF signals. b Quantitative analysis of BDNF levels. Fluorescence intensity of BDNF signals in the anterior cingulate cortex (ACC) were quantified by a densitometric method and expressed in mean ± SEM (n = 3). The difference between the experimental groups was analyzed by one-way ANOVA, followed by post hoc Dunnett’s test. # p < 0.05 (SNI vs Sham); *p < 0.05; **p < 0.01; ***p < 0.001 (Drug vs SNI only)

Full size image

Discussion

Depression and related psychiatric disorders frequently occur among patients with pain, especially those with chronic pain [27, 29]. Chronic pain ruins the quality of life and induces robust depressive-like behaviors [27, 30]. The interventions for chronic pain may hold promise to the control of the contextual psychosocial problems. To date, information is still restricted to the depression associated with specific infectious and inflammatory diseases such as Bacilli Calmette-Guerin inoculation and osteoarthritis [29, 31, 32]. Nevertheless, neuropathic pain is not well responsive to major conventional therapies including the combination of anti-depressant drugs and NSAIDs [7]. In the present study, we examined the effects of puerarin on depressive-like behaviors and pain responses in a SNI mouse model. Three key findings from the present study include (1) puerarin ameliorated depressive-like behaviors and pain responses in SNI mice; (2) puerarin exhibited dual anti-depressant and analgesic mechanisms compared with the existing drugs with either anti-depressant potential (e.g., citalopram) or anti-inflammatory activity (e.g., ibuprofen); (3) puerarin did not diminish the pharmacological activities of citalopram and ibuprofen.

Animal models of neuropathic pain are currently induced in mice and rats by SNI surgery and the chronic constriction injury [27, 33, 34]. The current study selected SNI mouse model for the evaluation of plant natural product puerarin for the management of chronic pain and depressive-like behaviors. Based on the results from forced swim test, tail suspension test, and von Frey filament test, our experiments confirmed that SNI rapidly induced comorbid pain and depressive-like behaviors. Based on the comparison of puerarin with existing drugs citalopram and ibuprofen for the anti-depressant activities, the immobility of SNI animals in forced swim test and tail suspension test was effectively reduced to a similar extent by puerarin vs citalopram, but not ibuprofen. When these drugs were used in combinations, puerarin and citalopram in combination maintained effective anti-depressant activity. Ibuprofen reduced the anti-depressant activity of citalopram but not that of puerarin. Presumably, puerarin and citalopram exhibited the anti-depressant activity via different mechanisms. Puerarin and ibuprofen in combination may not cause serious adverse drug-drug interactions. Moreover, we found similar phenomenon in testing the effects of puerarin, citalopram, and ibuprofen, alone or in combination, on pain response. Remarkably, puerarin showed better effect on pain response in SNI mice than citalopram and ibuprofen. Puerarin in combination with either citalopram or ibuprofen appeared to be more effective than puerarin alone. Such enhancement was evident after 3 days treatment but gradually disappeared in the treatment for prolonged period of time. Nevertheless, we did not find significant enhancement in the treatment with citalopram and ibuprofen in combination compared to single drug treatment. These results suggest that puerarin may be a dual anti-depressant and analgesic drug candidate with the potential to integrate with the existing drugs.

BDNF plays important roles on neuronal survival, neurogenesis, differentiation of neurons and synapses, and learning and memory [35, 36]. BDNF may attenuate chronic pain-induced depression via activating TrkB receptor at the cell surface of synaptic neurons [37]. The binding of BDNF to TrK receptor leads to the increased phosphorylation of protein kinases (e.g., MEK/ERK and PI3K/Akt) and transcriptional factors such as p-CREB for the regulation of synaptic function and neuron proliferation [38, 39]. Interestingly, BDNF, p-ERK1/2, and p-CREB were decreased in the hippocampus and prefrontal cortex of patients with depression and other psychiatric disorders [40, 41]. Increase in BDNF expression could largely improve depressive disorders [42]. Thus, BDNF was recently suggested to be an important biomarker for successful treatment of depressive-like behaviors [43, 44].

BDNF expression is regulated by several different mechanisms such as reactive oxygen species (ROS), ERK and p38 MAP kinase, and p-CREB [45–47]. The existing anti-depressant SSRIs ameliorate depressive-like behaviors via sequentially releasing inflammatory cytokines (e.g., TNF-α, IFN-γ) and inducing BDNF expression [7–9]. We explored the molecular mechanisms underlying the anti-depressant and analgesic activities of puerarin from three aspects: redox status, ERK1/2, and p-CREB, respectively. For the assessment of redox status, we determined the GSH levels and SOD activity for their anti-depressant roles as previously reported [48, 49]. In our experiments, SNI caused GSH depletion and, to a lesser extent, decreased SOD activity. Citalopram alone or in combination with puerarin or ibuprofen significantly elevated GSH levels. Ibuprofen alone showed somewhat activity but significantly increased GSH levels in the presence of puerarin. Interestingly, puerarin did not show much activity on either GSH levels or SOD activity. These results highlight the difference between puerarin and the existing drugs (i.e., citalopram and ibuprofen) in the regulation of intracellular redox status.

To probe the signaling molecules, we analyzed the phosphorylation of ERK1/2 and CREB by Western blot technique. We found that puerarin and citalopram, alone or in combination, significantly activated ERK1/2 pathway. On the other hand, puerarin alone profoundly induced the phosphorylation of CREB, whereas neither of citalopram and ibuprofen showed much activity. These results further revealed the difference between puerarin and the existing drugs (i.e., citalopram and ibuprofen) in the regulation of intracellular signaling pathways. Phosphorylation of CREB is an important signaling event in the neuronal plasticity and the anti-depressant response [50]. Finally, we examined the effects of puerarin, citalopram, and ibuprofen on BDNF induction by Western blot and immunostaining techniques. Chronic pain is known to induce various alterations of the cellular and molecular signals in the anterior cingulate cortex, leading to the development and aggravation of depression [51]. For this reason, we specifically examined the expression levels of BDNF in the anterior cingulate cortex of mouse brains. Not surprisingly, puerarin and citalopram, alone or in combination, significantly rescued SNI-induced loss of BDNF whereas ibuprofen alone showed marginal activity but did not undermine the activities of puerarin and citalopram. These results confirmed the common anti-depressant role of BDNF in SNI mice. It is noteworthy that puerarin rapidly induced the expression of BDNF. BDNF and anti-depressant drugs may differently but coordinately regulate the neuronal turnover, proliferation, and survival [52]. Interestingly, BDNF expression is regulated by ERK1/2 and CREB pathways, which are well-known to be the downstream targets of BDNF-TrkB signaling pathways [45, 53, 54]. Therefore, it is possible that puerarin promotes the cross-talks between BDNF, ERK, and CREB signaling pathways.

In conclusion, we have discovered the pharmacological potential of puerarin as a dual anti-depressant and analgesic drug candidate. Our results suggest that puerarin, citalopram, and ibuprofen may activate different anti-depressant and analgesic mechanisms. Importantly, puerarin may synergize with anti-depressant citalopram and non-steroid painkiller ibuprofen to resolve depression and pain via bolstering BDNF synthesis. Future work should verify the role of BDNF by using TrkB receptor inhibitor as well as the pharmacodynamics of puerarin towards the clinical translation of the experimental results from the present study.