Spider peptide Phα1β induces analgesic effect in a model of cancer pain

Abstract

The marine snail peptide ziconotide (ω‐conotoxin MVIIA) is used as an analgesic in cancer patients refractory to opioids, but may induce severe adverse effects. Animal venoms represent a rich source of novel drugs, so we investigated the analgesic effects and the side‐effects of spider peptide Phα1β in a model of cancer pain in mice with or without tolerance to morphine analgesia. Cancer pain was induced by the inoculation of melanoma B16‐F10 cells into the hind paw of C57BL/6 mice. After 14 days, painful hypersensitivity was detected and Phα1β or ω‐conotoxin MVIIA (10–100 pmol/site) was intrathecally injected to evaluate the development of antinociception and side‐effects in control and morphine‐tolerant mice. The treatment with Phα1β or ω‐conotoxin MVIIA fully reversed cancer‐related painful hypersensitivity, with long‐lasting results, at effective doses 50% of 48 (32–72) or 33 (21–53) pmol/site, respectively. Phα1β produced only mild adverse effects, whereas ω‐conotoxin MVIIA induced dose‐related side‐effects in mice at analgesic doses (estimated toxic dose 50% of 30 pmol/site). In addition, we observed that Phα1β was capable of controlling cancer‐related pain even in mice tolerant to morphine antinociception (100% of inhibition) and was able to partially restore morphine analgesia in such animals (56 ± 5% of inhibition). In this study, Phα1β was as efficacious as ω‐conotoxin MVIIA in inducing analgesia in a model of cancer pain without producing severe adverse effects or losing efficacy in opioid‐tolerant mice, indicating that Phα1β has a good profile for the treatment of cancer pain in patients.

It has been reported that, during the course of cancer, up to 90% of patients,suffer pain that impairs their quality of life. Moreover, cancer can be responsible for a state of painful hypersensitivity.1 The management of cancer‐associated pain is a special issue in cancer palliative care.2 Opioids are the mainstay treatment used for moderate to severe cancer‐related pain.1 However, the opioid therapy leads to distinct side‐effects that limit its use, including the development of analgesic tolerance and hyperalgesia.1 In patients non‐responsive to opioid therapy or who have developed related adverse effects, the use of adjuvant drugs, such as spinal delivery of ziconotide, is a valuable approach.3, 4

Ziconotide, the synthetic form of the ω‐conotoxin MVIIA (a peptide isolated from the marine snail Conus magus), is a selective and reversible inhibitor of the N‐type voltage gated calcium channels (NVGCCs).5 The analgesic efficacy of ziconotide was confirmed in patients with cancer pain refractory to other treatments, but its clinical use has been limited by the presence of several adverse effects.4 Diverse findings have shown the relevance of NVGCCs in the propagation of painful signals and substances capable of blocking these channels have been explored as novel analgesic drugs.6 Phα1β is a peptide purified from the venom of the armed spider Phoneutria nigriventer,7 which preferentially blocks NVGCCs.8 The analgesic effect of Phα1β has been shown in neuropathic and inflammatory pain models, with a therapeutic index wider than ω‐conotoxin MVIIA.9 However, the analgesic potential of Phα1β in cancer‐related pain refractory to opioid treatment is currently unknown.

Thus, the goal of this study was to evaluate the analgesic effect of Phα1β in a model of cancer pain. We also evaluated the adverse effects of this peptide in mice tolerant to morphine analgesia.

Materials and Methods

Animals

Male C57BL/6 mice (20–30 g) bred in‐house were used for these experiments. The study was approved by the Ethics Committee of the Federal University of Santa Maria (Santa Maria, Brazil) (number 23081.005024/2010‐88) and was carried out in accordance with ethical guidelines for the investigation of experimental pain in conscious animals set by the International Association for the Study of Pain.10 Behavioral evaluation was carried out blindly with respect to drug administration.

Cell culture and procedure for tumor inoculation

B16F10 murine melanoma cells (CRL‐6475; ATCC, Manassas, VA, USA) were cultured in DMEM containing 10% FBS and 1% penicillin–streptomycin (10000 U/100 μg/mL) at 37°C with 5% CO₂ in a humidified atmosphere. For tumor inoculation, 20 μL of melanoma cells (2 × 10⁵ cells) suspended in PBS, or vehicle (PBS alone), was injected s.c. into the plantar region of the right hind paw11.

Assessment of painful hypersensitivity

The 50% paw withdrawal threshold (PWT in g) was measured using von Frey filaments, using the up‐down method as previously described.12, 13 The PWT was measured before (baseline) and 14 days after melanoma cell inoculation before drug treatments (time 0). The PWT was measured in the periphery of the melanoma mass, as described previously, because this region usually shows more sensitivity.14 A significant reduction in PWT at time 0 when compared with baseline characterized the development of painful hypersensitivity (hyperalgesia).

Treatment and experimental protocol

The intrathecal (i.t) injection of drugs was carried out as described previously.15 First, the time‐course effect and dose–response curve of Phα1β (10, 30, or 100 pmol/site, i.t.) and ω‐conotoxin MVIIA (10, 30, or 100 pmol/site, i.t.) in cancer‐related hyperalgesia was evaluated (Fig. 1a). A different group of animals was injected with vehicle (PBS, 5 μL/site, i.t.). The doses of Phα1β and ω‐conotoxin MVIIA were based on a previous study.9

Figure 1.

Long‐lasting and potent antihyperalgesic effect induced by intrathecal (i.t) injection of spider peptide Phα1β and ω‐conotoxin MVIIA in a model of cancer pain in mice. (a) Representative scheme of the experimental protocol. (b) Time‐course antinociceptive effect of Phα1β and ω‐conotoxin MVIIA (30 pmol/site, i.t.). (c) Injection with vehicle (PBS, 20 μL), Phα1β or MVIIA (30 pmol/site, i.t.) had no effect on mechanical hyperalgesia or mechanical thresholds. (d, e) Evaluation of antinociceptive effect of Phα1β or ω‐conotoxin MVIIA in different doses (10–300 pmol/site, i.t.) 1 h after i.t. injection. Development of mechanical hyperalgesia was seen 14 days after injection (time point 0). Data are expressed as the mean ± SEM of 6–7 mice. ***P < 0.001, when compared with vehicle treated group; ^# P < 0.001, when compared with baseline (B) values (one‐ or two‐way anova followed by Bonferroni's post‐hoc test or Student's t‐test). PWT, paw withdrawal threshold.

Second, mice were made tolerant to morphine by a treatment consisting of s.c. morphine injections three times a day (4‐h intervals) for three consecutive days with an increasing dose schedule. Day 1: 10, 10, and 15 mg/kg; day 2: 15, 15, and 20 mg/kg; day 3: 20, 20, 25 mg/kg, as described,11 with modifications. The control group received vehicle (saline 0.9%) at the same time points as the tolerant group. After the first injection of morphine or vehicle, the time‐course of the morphine (10 mg/kg, s.c.) analgesic effect was evaluated (15–120 min). On day 4, to confirm the development of morphine‐induced tolerance, mice were challenged with an injection of 10 mg/kg morphine and tested again. The analgesic effect of Phα1β (30 pmol/site, i.t.) in tolerant mice was evaluated from 15 to 480 min after injection (Fig. 2a).

Figure 2.

Spider peptide Phα1β was largely able to reduce hyperalgesia in mice tolerant to morphine antinociception and it was also able to partially restore the morphine antinociceptive effect in morphine‐tolerant mice. (a) Representative scheme of the experimental protocol. (b) Time‐course antihyperalgesic (15–120 min) effect of morphine (10 mg/kg, s.c.). (c) Tolerance to morphine antinociceptive effect induced by repeated morphine injection (3 days). On day 4, the mice that became tolerant developed mechanical hyperalgesia, which was not observed in non‐tolerant mice (vehicle treated). (d) Antinociceptive time‐course (15–480 min) effect of Phα1β (30 pmol/site, intrathecal [i.t.] injection) mice tolerant to morphine. Morphine (10 mg/kg, s.c) was injected at 480 min in tolerant mice pretreated with Phα1β (30 pmol/site) (arrow) and Phα1β partially restored morphine analgesia in tolerant animals at time 495 min. Data are expressed as the mean ± SEM of 6–9 animals. ***P < 0.001, when compared with vehicle group; ^# P < 0.001, when compared with baseline (B) values; ⁺ P < 0.001, when compared with values before the tolerance induction of the vehicle‐treated group; ^§ P < 0.001, when compared with values after the tolerance induction of the morphine‐treated group (one‐ or two‐way anova followed by Bonferroni's post hoc test, or Student's t‐test). PWT, paw withdrawal threshold.

Finally, we tested if the analgesic effect of morphine could be restored by the prior administration of Phα1β (30 pmol/site, i.t.) in mice tolerant to morphine. For that, we used the same protocol of morphine‐induced tolerance described earlier, and on day 4 injected Phα1β (30 pmol/site, i.t.) or vehicle (PBS) 2 h after the morphine challenge dose (10 mg/kg, s.c.). The mechanical threshold was then evaluated 15 min after morphine injection.

Evaluation of adverse effects

The development of behavioral adverse effects (tail serpentine‐like movements, whole‐body shaking, dynamic allodynia) and sedation were observed 1‐h post‐injection (i.t. in the peak of analgesic effect) of Phα1β (30 and 100 pmol/site), ω‐conotoxin MVIIA (30 and 100 pmol/site) or vehicle.16, 17 Sedation was assessed using the platform sedation test18 and animals were considered sedated when they presented an increase of at least 25% of fall time post‐drug when compared to pre‐drug latency.

Drugs

Native Phα1β was purified, as previously described.19 ω‐Conotoxin MVIIA was purchased from Latoxan (Valence, France). Morphine sulphate was purchased from Laboratório Cristália (São Paulo, Brazil).

Statistical analysis

The results are shown as the mean ± standard error (to PWT) or as geometric means accompanied by their respective 95% confidence limits to the ED₅₀ values. The ED₅₀ values were calculated by non‐linear regression using a dose–response equation. Other data were analyzed using a one‐way or two‐way anova, followed by Dunnet or Bonferroni tests, when appropriate. For analysis of the adverse effects, the χ ²‐test was used. Statistical significance was accepted at P < 0.05, using GraphPad Software 5.0 (GraphPad, La Jolla, CA, USA).

Results

Fourteen days after inoculation of B16‐F10 melanoma cells, a marked hyperalgesia was induced. This cancer‐related pain was fully reversed by i.t. treatment with Phα1β (Fig. 1b,d). The antihyperalgesic produced by Phα1β (30 pmol/site) developed quickly, beginning as earlier as 0.25 h after injection, was long‐lasting (up to 6 h), and peaked 1 h after injection (Fig. 1b). The Phα1β antihyperalgesic effect was dose‐dependent, with calculated values of ED₅₀ (and its 95% confidence limits) of 48 (32–72) pmol/site and maximal inhibition (I_max) of 100% at the dose of 100 pmol/site (Fig. 1d).

Comparatively, i.t. injection of ω‐conotoxin MVIIA (30 pmol/site) also induced an antihyperalgesic effect with a time‐course very similar to Phα1β (Fig. 1b). The ED₅₀ value for ω‐conotoxin MVIIA was 33 (21–53) pmol/site and the I_max was 100% at a dose of 100 pmol/site (Fig. 1e). Both ω‐conotoxin MVIIA and Phα1β reversed hyperalgesia in animals inoculated with tumor cells without altering the physiological detection of mechanical‐painful stimuli (Fig. 1c).

As shown in Figure 2(b), a single dose of morphine (10 mg/kg, s.c.) but not saline was able to induce an analgesic effect from 0.25 h to 1 h (100% inhibition at 0.5 h) after treatment in mice inoculated with tumor cells. Treatment with increasing doses of morphine or with saline was repeated three times a day for 3 days. Twenty‐four hours after the last repeated injection, we observed that the hyperalgesia induced by tumor cells was increased in mice treated with morphine, but not saline (50% PWT values of 0.015 ± 0.001 and 0.049 ± 0.007 g, respectively; P < 0.001, Student's t‐test) (Fig. 2c). The injection of a challenge dose of morphine (10 mg/kg, s.c.) was only able to produce a small antihyperalgesic effect (7 ± 3% inhibition at 0.5 h) in those animals repeatedly treated with morphine (Fig. 2c), suggesting the development of analgesic tolerance.

The injection of Phα1β (30 pmol/site, i.t.) was capable of reversing cancer‐related pain in morphine‐tolerant mice, in terms of efficacy and time‐course, at a similar level to that observed in non‐tolerant animals (Fig. 2d). Moreover, prior injection of Phα1β (30 pmol/site) was able to partially re‐establish the analgesic effect of morphine in tolerant mice (56 ± 5% at 0.25 h) (Fig. 2d).

Finally, we assessed the adverse effects produced by drugs in tumor cell‐inoculated animals. The i.t. injection of Phα1β (30 or 100 pmol/site) did not produce serpentine tail movements, body shakes, or sedation 1 h after it was given (Table 1). However, Phα1β injection produced only dynamic allodynia. In contrast, i.t. injection of ω‐conotoxin MVIIA induced dose‐related side‐effects (including serpentine tail movements, sedation, and dynamic allodynia) at doses were it produced an antihyperalgesic effect (Table 1). Sedation and allodynia were the more prevalent side‐effects observed, and the estimated dose that produced such toxicity in 50% of animals (DT₅₀) was 30 pmol/site.

Table 1.

Adverse effects in mice (n = 8) after spider peptide Phα1β or ω‐conotoxin MVIIA intrathecal injection (30 or 100 pmol/site)

Drug	Dose (pmol/site)	Adverse effect
		Body shake	Serpentine tail	Dynamic allodynia	Sedation
PBS	NA	0/8	0/8	0/8	0/8
Phα1β	30	0/8	0/8	2/8	2/8
	100	0/8	0/8	3/8*	0/8
MVIIA	30	0/8	1/8	4/8**	4/8**
	100	0/8	3/8*	6/8***	6/8***

Data expressed as the number of animals showing behaviors/total injected animals and were analyzed by χ ²‐test (n = 6–8). *P < 0.05; **P < 0.01; ***P < 0.001 versus control (PBS‐treated) groups. NA, not applicable.

Discussion

Although pain is not a major symptom of melanoma in the clinic, 7% of patients experienced pain; metastatic melanoma is associated with excruciating pain, and more than 50% of these patients require palliative care and morphine treatment.20, 21 Here, we observed that all mice inoculated with melanoma cells presented hyperalgesia 14 days after inoculation. This type of pain was seen in both inoculated (ipsilateral) and non‐innoculated (contralateral) hind‐paw of mice, indicating a further site than the primary tumor (50% PWT values of 1.974 ± 0.351 g and 2.197 ± 0.381 g at baseline or 0.053 ± 0.011 g and 0.417 ± 0.134 g, for the ipsilateral and contralateral paw, respectively; P < 0.001, Student's t‐test, n = 8–10). In fact, studies indicate that there is extensive metastasis 14 days after inoculation of B16‐F10 cells into paw of mice. We observed that all animals inoculated with melanoma cells showed lung metastasis 14 days after cancer induction (100% of animals, n = 30). Therefore, the model used in the present study was appropriate to study metastatic melanoma‐related pain.

An alternative to treat severe refractory cancer pain, including that resulting from metastatic cancer, is i.t. injection of ziconotide (ω‐conotoxin MVIIA).22 Here, we observed that Phα1β possess an excellent analgesic effect in a model of metastatic melanoma‐related pain, being as potent, efficacious, and long‐lasting as ω‐conotoxin MVIIA to reduce hyperalgesia. The role of NVGCCs is well established in the transmission of painful impulses at the spinal cord, where they are highly expressed in presynaptic terminals of nociceptive neurons, allowing calcium influx and releasing algogenic neurotransmitters.5 Regarding cancer pain, depolarization‐induced calcium influx in the cell bodies of nociceptive neurons of mice with tumors was increased and even related with the animal's hyperalgesia.23 Therefore, blocking NVGCCs in nociceptive neurons by Phα1β and ω‐conotoxin MVIIA appears to be important in controlling pain associated with cancer.

Despite the excellent efficacy of ziconotide in patients with cancer‐related pain, its clinical use is limited by manifestation of adverse effects at analgesic doses, such as somnolence and paradoxal pain.24 Here, the i.t. injection of Phα1β produced minimal adverse effects (just paradoxal hyperalgesia) only at the highest dose, whereas ω‐conotoxin MVIIA showed adverse effects (such as sedation, motor dysfunction, and paradoxal hyperalgesia) at all doses. We assessed adverse effects based on behavioral symptoms and sedative effects because N‐type calcium channel blockers usually induce motor dysfunction when injected i.t.4 In addition, if a new substance induces motor dysfunction or sedation it could impair the detection of mechanical thresholds.25 As indicated, it was shown that Phα1β (i.t.) injection induced mild hypotension with a complete recovery after 3 h of injection.9 Furthermore, we observed that Phα1β (i.t.) continuous infusion in rats for 7 days did not induce motor dysfunction or sedation, and no histopathological alteration was observed in the spinal cord or brain (data not shown). Although both blockers can inhibit NVGCCs,26, 27 ω‐conotoxin is several times (~100‐fold) more selective than Phα1β.26, 27, 28 Thus, the ability of Phα1β to block other VGCCs in addition of the N‐type27 could explain their better therapeutic profile, as shown in metastatic melanoma‐related pain.

The mainstay of cancer pain control is associated to opioid therapy.2 Although described as distinct clinical phenomena, several animal studies have indicated a link between hyperalgesia and opioid‐induced tolerance.29 Here, we observed that after 3 days of repeated morphine injection, the mice developed hyperalgesia and complete analgesic tolerance to a morphine dose. These results were similar to that observed by Sasamura et al.11 who reported in the same cancer pain model that continuous injection of morphine leads to analgesic tolerance. The i.t. injection of Phα1β was able to reduce the mechanical hyperalgesia observed in mice subjected to cancer pain and tolerant to morphine analgesia, in a similar fashion to that observed in non‐tolerant mice. Accordingly, ω‐conotoxin MVIIA also possesses analgesic efficacy in animals tolerant to morphine analgesia, but was unable to restore morphine tolerance.30 Morphine produces analgesia by the activation of μ‐opioid receptor and, at least in part, the consequent closure of NVGCCs in the spinal cord.31 Of note, the ability of opioids to block NVGCCs is lost in tolerant animals.32 Thus, as Phα1β directly blocks NVGCC, it may explain its analgesic effect in morphine‐tolerant animals.

Another interesting result from this study was that prior injection of Phα1β in mice tolerant to morphine analgesia was able to partially restore analgesia, unlike ω‐conotoxin MVIIA.30 Interestingly, several calcium‐dependent intracellular kinases at the spinal cord, such as calcium–calmodulin kinase and protein kinase C, have been implicated in the development of opioid tolerance.33 Thus, the blockage of calcium influx by Phα1β could reduce the activation of such pathways and reverse opioid tolerance, as do calcium–calmodulin kinase and protein kinase C inhibitors.33, 34, 35

Collectively, our findings showed that Phα1β injection in a model of cancer pain produced long‐lasting, efficacious, and potent analgesic effects with a safer profile to that seen with ω‐conotoxin MVIIA, as well as great efficacy in morphine‐tolerant animals. These findings suggest that Phα1β is a suitable alternative to cancer pain relief in patients who are non‐tolerant or tolerant to morphine analgesia.

Conflict of Interest

The authors have no conflict of interest.

(Cancer Sci, doi: 10.1111/cas.12209, 2013)