Adult male C57BL/6J mice (6–7 weeks old upon arrival), obtained from Animal Center of Chinese Academy of Sciences, were used in this study. The animals were housed in group (5–6 per cage) with water and chow available ad libitum, 12 h light cycles (on at 07:00 AM), and stable temperature (22 ± 0.5 °C) and relative humidity (60 ± 2%). In animal behavioral tests, operators were blind regarding the treatment cohort and animals were assigned to a cohort randomly. All animal handing and experiments were in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC). The animal study was reviewed and approved by the Ningbo University Committee on Animal Care and Use. The authors further ensure here that all efforts were made to minimize animal suffering and the number used.

A chronic constriction injury (CCI) of ipsilateral sciatic nerve was carried out in mice (7–8 weeks old at the time of surgery shown as day 0) as originally proposed by Bennett and Xie (1988) with minor modifications by us for adaptation to mice (Zhao et al., 2012). Briefly, surgery was performed following the induction of anesthesia by intraperitoneal (i.p.) injection of pentobarbital sodium (60 mg/kg). The right common sciatic nerve was exposed at the nerve of the mid thigh, and close to the sciatic nerve trifurcation, three ligations (4/0 silk threads, with 1 mm spacing appropriately) were tied loosely around it until a slight twitch was seen in the respective hindlimb. The surgical area was dusted with streptomycin and the incision was sutured. For sham animals, the sciatic nerves were isolated without ligations. The animals were placed under a heat lamp until they were awake. CCI or sham-operated mice were housed in separate cages and tested simultaneously. All surgical procedures were performed by the same operator.

The administration of isorhynchophylline (p.o., via gavage with a volume of 10 ml/kg) was initiated 15 days after the surgery of CCI or sham operation. At this time point, the CCI animals exhibited remarkable nocifensive hypersensitivity, evidenced by thermal hyperalgesia and tactile allodynia. For repetitive dosing, the sham and CCI animals were administered with isorhynchophylline or vehicle twice per day (10:00 AM and 6:00 PM, shown in Figure 1B) for two weeks (i.e., from day-15 to day-29 after surgery). After two weeks of isorhynchophylline regimen, the animals were co-administered with 5-hydroxytryptophan (5-HTP, a precursor of serotonin) or one of the 5-HT receptor antagonists (Figure 1B). These antagonists are WAY-100635 (a 5-HT_1A preferring antagonist), isamoltane (a 5-HT_1B preferring antagonist), ritanserin (a 5-HT_2A/2C preferring antagonist), and ondansetron (a 5-HT₃ preferring antagonist). The doses of these agents were used on the basis of previous studies (Wang et al., 2008; Zhao et al., 2012; Hu et al., 2017), with minor modifications to verify paucity of intrinsic activity per se in the present nocifensive tests. For i.p. injection, these agents were administered to animals 30 min before isorhynchophylline dosing with a volume of 10 mg/kg. Isorhynchophylline was purchased from Chengdu Mansite Pharmaceutical Co., Ltd. (Chengdu, China) with the purity more than 98%, and the other drugs were from Sigma-Aldrich. Isorhynchophylline was suspended in 0.5% sodium carboxymethyl cellulose and all the other drugs were dissolved in freshly sterile saline.

To localize the serotonergic receptors possibly involved, intracerebroventricular (i.c.v.), intrathecal (i.t.) or intraplantar (i.pl.) injection of WAY-100635 was carried out after two weeks of isorhynchophylline regimen. We did two i.c.v., i.t. or i.pl. injections on day-29, 30 min before the morning and the evening isorhynchophylline dosing. The mice were then challenged for nociceptive tests on the next morning, i.e., 38 h after the last isorhynchophylline dosing alone. These procedures are based on previous reports (Choucair-Jaafar et al., 2009; Yalcin et al., 2009; Zhao et al., 2012) regarding the analgesic mechanism of antidepressants.

Intracerebroventricular injections were performed as delineated by Haley and Mccormick (1957). Briefly, under ether anesthesia, mice were grasped firmly with the soft skin behind the head. A 29-gauge needle connecting a Hamilton syringe (50 μl) was inserted perpendicularly into mouse lateral ventricle, through the skull and around 2.0 mm into brain. The injection site was appropriate 1 mm lateral to the midpoint on a line between two anterior bases of the ears. After behavioral tests, the animals were decapitated rapidly and brains were removed and frozen in −70°C. The brains were sectioned in the coronal plane with the thickness of 10μm. The sections were then stained by cresyl violet to verify the injection sites. The successful injection rate was normally more than 95% in our group.

The procedures for intrathecal injections were based on the method described by Hylden and Wilcox (1980). Shortly, a 29-gauge needle a Hamilton syringe (50 μl) was inserted into the sub-arachnoidal space between L5 and L6 vertebrate. The successful placement of the needle was verified by the sign of a tail flick movement. The volume for i.t. injection was 10μl.

The intraplantar injections were performed with ether anesthesia. A 29-gauge injection needle punctured the ipsilateral plantar skin and reached the subcutaneous space proximal to the footpad. The volume for i.pl. injection was 10μl.

Following two weeks of isorhynchophylline treatment, pharmacological manipulations were carried out to ablate spinal 5-HT or NA in mice. For chemically ablating spinal NA transmission, the mice were intrathecally injected with 6-OHDA (a catecholaminergic neurotoxin) at the dose of 20 μg per mouse (Suzuki et al., 2008; Zhao et al., 2012). To deplete descending 5-HT, p-chlorophenylalanine (PCPA, an inhibitor of 5-HT synthesis) was administered i.p. to mice for five successive days at the dose of 300 mg/kg/day (Tanabe et al., 2007; Zhao et al., 2012). 6-OHDA was dissolved in 5 μl of 0.9% saline containing ascorbic acid (100 μg/ml) and PCPA was suspended in 0.5% gum acacia/physiological saline.

After 3 weeks isorhynchophylline or vehicle treatment, the sham or CCI mice were decapitated, followed by the lumbar enlargements (approximately 1 cm) of their spinal cords being removed and tissues being dissected rapidly on an ice-chilled glass plate. The tissue samples were weighed, and then stored at −80°C. The contents of monoamines (NA, serotonin and dopamine) and their metabolites (5-HIAA and DOPAC) were measured as described previously by high-performance liquid chromatography (HPLC) with electrochemical detection (Zhao et al., 2014). Briefly, each spinal cord was homogenized by ultrasonication in 200 μl of 0.4 M perchloric acid (solution A). The homogenate was kept on ice for 1 h and then centrifuged at 12,000 × g at 4°C for 20 min. After discarding the pellet, an aliquot of 160 μl of supernatant was added to 80 μl of solution B (containing 0.2 M potassium citrate, 0.3 M dipotassium hydrogen phosphate and 0.2 M EDTA). The mixture was kept on ice for 1 h and then centrifuged at 12, 000 × g at 4°C for 20 min again. 20 μl of the supernatant was directly injected into an ESA liquid chromatography system equipped with a reversed-phase C₁₈ column (150 × 4.6 mm I.D., 5 μm) and an electrochemical detector (ESA CoulArray, Chelmstord, MA, USA.). The mobile phase consisted of 125 mM citric acid –sodium citrate (PH 4.3), 0.1 mM EDTA, 1.2 mM sodium octanesulfonate, and 16% methanol. The flow rate was 1.0 ml/min. the tissue levels of monoamines and their metabolites were expressed as nanograms per gram wet spinal tissue. The MAO activity in the tissue was assayed according to our established protocol (Zhao et al., 2014).

All values are presented as the mean ± S.E.M. Data were analyzed by multifactor analysis of variance (ANOVA) or one-way ANOVA. For multifactor-ANOVA, for example, the surgery (sham or CCI) and the treatment (vehicle or drug administration) were done as between-group factors. When needed, the time of measurement (time-course data) was done as within-subject factor. The Tukey test was used for post hoc comparisons. For one-way ANOVA, Student-Newman-Keuls test was used for multiple comparisons, followed by Student’s test to evaluate the difference between two groups at the same time. Differences with p < 0.05 were considered statistically significant.

The mononeuropathy produced by chronic constriction injury of the sciatic nerve evoked in mice ipsolateral thermal hyperalgesia and tactile allodynia, which were persistent and maintained throughout the experimental period (Figures 2A, B, left panels), without affecting the measurements of contra-lateral paws (data not shown). As expected, sham-surgery operation did not change the nocifensive reponsivity to tactile or thermal stimuli.

After two weeks of treatment as shown in Figure 2, isorhynchophylline strikingly attenuated in neuropathic mice the evoked thermal hyperalgesia (surgery × treatment × days interaction, F_7,282 = 9.8, p = 0.0023; post hoc: CCI + vehicle = CCI + IRN < sham on days 15–25, CCI + vehicle < all other groups on days 27–29; Figure 2A, left panel) and tactile allodynia (surgery × treatment × days interaction, F_7,282 = 13.1, p = 0.0008; post hoc: CCI + vehicle = CCI + IRN < sham on days 15–25, CCI + vehicle < all other groups on days 27-29; Figure 2B, left panel) in a dose-related manner (Figures 2A, B, right panels), showing statistical difference at doses of 15 mg/kg and 45 mg/kg. At the high dose of 45 mg/kg, isorhynchophylline showed recuperating effects on the evoked nociceptive hypersensitivities (heat hyperalgesia and tactile allodynia), without influencing the measures in sham-operated mice (Figures 2A, B). In the experimental session of acute treatment, isorhynchophylline did not show any favorable effect on these nocifensive measurements in sham-operated and neuropathic animals (Supplementary Material files Figure S1).

To determine whether isorhynchophylline-induced behavioral alterations in the Heagreaves test and von Frey test have relevance to the effect of isorhynchophylline on motor function, we performed another set of experiments to probe the effects of repetitive isorhynchophylline treatment on motor ability and locomotor activity in neuropathic mice. Our results show repetitive isorhynchophylline treatment, at the dose range of 5–45 mg/kg, did not affect the behavioral outcomes in the locomotor test (Supplementary Material files Figure S2A) and rota-rod test (Supplementary Material files Figure S2B). These results indicate isorhynchophylline-induced behavioral alterations in pain-related assays should not be interfered by potential effect on motor function.

As isorhynchophylline at the high dose of 45 mg/kg exerted disease-modifying effects on neuropathic hyperalgesia and allodynia, we evaluated how long can these actions last, and whether these actions can be re-initiated following isorhynchophylline cessation. As shown in Figure 3, we terminated isorhynchophylline dosing on day-30 when the hypernociceptive behaviors (heat hyperalgesia and tactile allodynia) were normalized by repetitive isorhynchophylline treatment. Although the interruption of isorhynchophylline dosing caused the relapse of neuropathic hyperalgesia (Figure 3A) and allodynia (Figure 3B), its therapeutic actions lasted for at least 3 days. On day-40, a new repetitive dosing of isorhynchophylline (45 mg/kg, p.o., twice per day) was initiated. In this session, isorhynchophylline again produced disease-modifying actions on neuropathic hyperalgesia (Figure 3A) and allodynia (Figure 3B) with remarkably less time of onset (around 10 days) compared with the first session of regimen.

As monoaminergic system, a vital ingredient of descending pathway in pain regulation (Millan, 2002), is involved in the antidepressant-like effects of isorhynchophylline (Xian et al., 2017), we reasoned that modification of descending monoamine transmission by isorhynchophylline may contribute its antinociceptive effects. We therefore investigated the effect of repetitive isorhynchophylline treatment on spinal levels of monoamines. Compared with vehicle treatment, isorhynchophylline dose-dependently increased, in neuropathic mice, the spinal level of 5-HT (F_3,38 = 17.3, p = 0.0026; Figure 4A), and decreased the ratio of 5-HIAA/5-HT (F_3,38 = 8.4, p = 0.0076; Figure 4C). However, isorhynchophylline at the dose range of 5–45 mg/kg, did not alter the levels of spinal 5-HT and the ratio of 5-HIAA/5-HT in sham-operated mice. Regarding spinal NA level (Figure 4D), there is a marginal but significant increase (F_3,38 = 12.7, p = 0.0376) in neuropathic mice following treatment with isorhynchophylline at high dose (45 mg/kg). The present isorhynchophylline regimen did not change the spinal levels of 5-HIAA (Figure 4B), dopamine (Figure 4E) and DOPAC (Figure 4F) in sham-operated and neuropathic mice. Consistent with previous studies (Vogel et al., 2003), there is a sharp serotonin-fall in vehicle-treated neuropathic mice, indicating the dearth of serotonin at spinal level may have relevance on peripheral nerve injury-induced mononeuropathy.

Escalated spinal 5-HT level and depressed ratio of 5-HIAA/5-HT imply the alteration of monoamine metabolism by isorhynchophylline. The monoamine in CNS is metabolized by two isoforms of MAO, i.e., MAO-A and MAO-B. We then evaluated the spinal MAO-A and MAO-B activities following repetitively oral isorhynchophylline treatment. As shown in Table 1, Neuropathic mice showed escalated activity of MAO-A but not MAO-B compared with sham mice, which is consistent with a previous study by Villarinho et al. (2013). Following repetitive oral treatment, isorhynchophylline with doses ranging from 5 to 45 mg/kg dose-dependently inhibited the escalated spinal MAO-A activity in neuropathic mice (F_3,38 = 14.2, p = 0.0072), and at the high dose of 45 mg/kg the MAO-A activity is reduced to the level essentially equivalent to that of sham mice. The present isorhynchophylline regimen did not influence the spinal MAO-B activity in both sham and neuropathic mice.

To explore whether the monoamine system is causally responsible for the antinociceptive effects of isorhynchophylline, we observed the influence of ablating spinal 5-HT or NA on the antinociceptive effects of isorhynchophylline. Consistent with our previous study (Zhao et al., 2014), a single intrathecal injection of 6-OHDA (20 μg per mouse) or five consecutive injection of PCPA (i.p., 300 mg/kg, once per day) reduced spinal NA or 5-HT content, respectively (data not shown). We then evaluated the effect of 5-HT or NA ablation by PCPA or 6-OHDA, respectively, on the antinociceptive effects of isorhynchophylline in neuropathic mice. As shown in Figure 5, PCPA (300 mg/kg, i.p.), after 5 days of successive injections, reverted the actions of isorhynchophylline on thermal hyperalgesia (surgery × treatment × days interaction, F_6,232 = 9.7, p = 0.0092; post hoc: CCI + vehicle < all other groups on days 29–33, CCI + vehicle = CCI + IRN < sham on days 34–35; Figure 5A) and tactile allodynia (surgery × treatment × days interaction, F_6,232 = 15.9, p = 0.0057; post hoc: CCI + vehicle < all other groups on days 29-33, CCI + vehicle = CCI + IRN < sham on days 34-35; Figure 5B), pointing to a critical role for serotonergic system in the present isorhynchophylline analgesia. In contrast, 6-OHDA-induced spinal NA ablation did not impact on the antinociception exerted by isorhynchophylline (Supplementary Material files Figure S3), indicating NA system may not be involved.

To further validate the crucial involvement of the serotonergic system, we tested whether the present isorhynchophylline analgesia can be affected by co-treatment with 5-HTP, a precursor of serotonin. As shown in Figure 5, a single administration of 5-HTP (10 mg/kg, i.p.) or ineffective dose of isorhynchophylline (5 mg/kg, p.o.) did not impact on the profiles measured in the Hargreaves test and von Frey test for both sham and neuropathic mice. However, co-administration of them yielded marked antinociceptive effects on thermal hyperalgesia (surgery × treatment × days interaction, F_6,246 = 6.4, p = 0.0067; post hoc: CCI + vehicle = CCI + IRN < sham on days 29–32, CCI + vehicle < all other groups on days 33–35; Figure 5C) and tactile allodynia (surgery × treatment × days interaction, F_6,246 = 21.6, p = 0.0029; post hoc: CCI + vehicle = CCI + IRN < sham on days 29–32, CCI + vehicle < all other groups on days 33–35; Figure 5D) in neuropathic mice, pointing toward a potentiating effect when isorhynchophylline was co-treated with 5-HTP. These results support the serotonergic mechanism underlying the present isorhynchophylline analgesia.

To disclose the subtype of 5-HT receptors implicated in the antinociceptive effects of isorhynchophylline, we asked whether the isorhynchophylline actions are influenced by subtype-selective 5-HT receptor antagonists, i.e., WAY-100635 (5-HT_1A receptor antagonist), isamoltane (5-HT_1B receptor antagonist), ritanserin (5-HT_2A/C receptor antagonist), and ondansetron (5-HT₃ receptor antagonist). Our results show that repetitive co-administration of 5-HT_1A receptor antagonist WAY-100635 offset the antinociceptive effects of isorhynchophylline on thermal hyperalgesia (surgery × treatment × days interaction, F_5,196 = 17.1, p = 0.0036; post hoc: CCI + vehicle < all other groups on days 29–32, CCI + vehicle = CCI + IRN < sham on days 33–34; Figure 6A) and tactile allodynia (surgery × treatment × days interaction, F_5,196 = 5.7, p = 0.0047; post hoc: CCI + vehicle < all other groups on days 29–32, CCI + vehicle = CCI + IRN < sham on days 33–34; Figure 6D). The effects of WAY-100635 on isorhynchophylline-induced antihyperalgesia (F_3,38 = 6.9, p = 0.0081; Figure 6B) and antiallodynia (F_3,38 = 10.2, p = 0.0028; Figure 6E) are dose-dependent (0.1, 0.3, and 1 mg/kg). In contrast, the other 5-HT antagonists (ondansetron, isamoltane, and ritanserin) did not influence the antinociceptive effects of isorhynchophylline on thermal hyperalgesia (Figure 6C) and tactile allodynia (Figure 6F), further consolidating the allegation that 5-HT₁ receptors are exclusively engaged in the present isorhynchophylline analgesia.

To locate the functional involvement of 5-HT₁ receptors in the antinociceptive effects of isorhynchophylline, we injected WAY-100635 via three special delivery routes (i.t., i.c.v., and i.pl.) and tested its effect on the analgesia by isorhynchophylline. As shown in Figures 7A, D, i.t. injection of WAY-100635 counteracted, in a dose-dependent manner, the antinociceptive effects of isorhynchophylline on thermal hyperalgesia (Figure 7A, F_3,38 = 23.4, p = 0.0093) and tactile allodynia (Figure 7D, F_3,38 = 7.7, p = 0.0056) in neuropathic mice, without influencing the measures in sham mice. However, i.c.v. (5 μg in 2.5 μl, Figures 7B, E) or i.pl. (1.5 μg in 10 μl, Figures 7C, F) injection of WAY-100635 did not impact on the nociceptive responses to thermal (Figures 7B, C) and tactile (Figures 7E, F) stimuli in isorhynchophylline-treated sham and neuropathic mice. These results indicate the 5-HT₁ receptors at spinal site are responsible for the present isorhynchophylline analgesia in neuropathic mice.

As shown in Figure 8A, isorhynchophylline by itself did not produce any significant stimulation upon [³⁵S] GTPγS binding to 5-HT_1A CHO cell, with concentrations ranging from 0.1 nM to 10 μM. As expected, the 5-HT_1A agonist 8-OH-DPAT produced remarkable stimulant effect on 5-HT_1A receptors shown as a significant increase of [³⁵S] GTPγS binding to 5-HT_1A CHO cell, and this action is concentration-dependent with the range of 1–1,000 nM (Figure 8A). Intriguingly, we found that isorhynchophylline, at the concentration range of 1–100 nM, induced a statistically increase in the efficacy (E_max), but not in the potency (EC₅₀), with which 8-OH-DPAT stimulates [³⁵S]-GTPγS binding to 5-HT_1A expressing CHO cell membranes (Figures 8B–F and Table 2). Furthermore, we found that in the same membranes, isorhynchophylline at the concentration range of 1 nM-10 μM did not show any significant activity to displace the binding of [³H]-8-OH-DPAT (Supplementary Material files Figure S4A). Also within the effective concentration range of 0.1–10 nM for the enhancement of [³⁵S] GTPγS binding, isorhynchophylline did not influence the binding of 8-OH-DPAT to 5-HT_1A receptors (Supplementary Material files Figures S4B–D). Collectively, these in vitro data demonstrate that isorhynchophylline can modulate the activation of 5-HT_1A indirectly, but not by binding to their orthosteric sites directly.

A separate set of experiment was performed to evaluate whether isorhynchophylline is able to engender favorable effects on co-morbid depressive- and anxiety-like behaviors in neuropathic mice. Congruent with previous studies (Hu et al., 2009; Zhao et al., 2014), neuropathic mice developed co-morbidly depressive-/anxiogenic-like behaviors 2 weeks after the CCI surgery, as evidenced by increased immobility time in the tail suspension test (TST, a measurement of depressive-like behavior), increased latency to feeding in the novelty suppressed feeding test (NSFT, a measurement indicative of depressive- and anxiety-like behavior), and decreased staying time in lit part in the light-dark test (LDT, a measurement of anxiety-like behavior).

After 3 weeks of treatment, isorhynchophylline attenuated in CCI mice these co-morbidly behavioral phenotypes of depression and anxiety in a dose-dependent manner (5, 15, and 45 mg/kg). (a) In the TST, isorhynchophylline dose-dependently decreased the immobility time (F_3,38 = 13.0, p = 0.0127; Figure 9A); (b) In the NSFT, isorhynchophylline dose-dependently reduced the latency to feeding (F_3,38 = 6.2, p = 0.0071; Figure 9B); (c) In the LDT, isorhynchophylline dose-dependently increased the staying time in lit part (F_3,38 = 11.8, p = 0.0089; Figure 9C). These results reveal that isorhynchophylline displayed disease-modifying action up the co-morbidly depressive/anxiogenic-like behaviors specifically in neuropathic mice. Nonetheless, chronic isorhynchophylline treatment did not influence the measures of depressive/anxiogenic-like behaviors in sham-operated mice.