Peripherally Acting μ-Opioid Receptor Agonists Attenuate Ongoing Pain-associated Behavior and Spontaneous Neuronal Activity after Nerve Injury in Rats

APPROXIMATELY 25 to 30% of chronic pain in patients is neuropathic in nature, afflicting 7 to 8% of adults in the general population.¹  These patients frequently experience ongoing pain and paroxysmal spontaneous pain, which are difficult conditions to treat. However, neuropathic pain-related behavior in animal models has commonly been inferred from observations of evoked reflex withdrawal responses to external stimuli. Clinically relevant behaviors suggestive of ongoing pain are more complex and have received less mechanistic investigation in preclinical studies. Because the underlying mechanisms and treatment strategies of ongoing pain may differ from those of evoked sensory hypersensitivity,^2,3  animal behavioral assays that rely on reflex responses do not always correlate well with analgesic effectiveness in humans. Accordingly, new drug candidates for neuropathic pain should also be studied comprehensively in experimental paradigms that assess the efficacy, safety, and mechanisms by which they inhibit ongoing pain.^4–6 

Improving opioid therapies for neuropathic pain remains a high research priority, largely because opioids are associated with severe central adverse effects that include sedation, addiction, and abuse.⁷  Recent studies suggested that targeting μ-opioid receptors in the peripheral nervous system (PNS) may alleviate pain hypersensitivity without causing central adverse effects.^8,9  Dermorphin [D-Arg2, Lys4] (1–4) amide,¹⁰  which is derived from the natural heptapeptide μ-opioid receptor agonist dermorphin, is highly hydrophilic and exhibits a restrictive penetration into the central nervous system (CNS) after systemic administration.^11,12  Our recent study showed that dermorphin [D-Arg2, Lys4] (1–4) amide attenuated mechanical and heat hypersensitivities in nerve-injured rats.¹³  Primary afferent neurons may also be an attractive target for the development of novel therapies for ongoing pain. Afferent drive after injury including ectopic discharges, and spontaneous activity in dorsal root ganglion and spinal wide-dynamic range neurons are considered to be important neurophysiologic correlates of ongoing pain.^14,15  One strategy would be to develop analgesics that act at spontaneously active sensory neurons and target “pain at its source” before the signals diverge over multiple pathways in the CNS.^8,16,17  Yet, whether peripherally acting μ-opioids alleviate ongoing neuropathic pain and inhibit its neurophysiologic correlates remains unclear.

Developing novel analgesics based on preclinical data has been fraught with difficulties that are likely multifactorial. One factor may be the insufficient predictive ability of outcome measures used in animal models that are based solely on reflex responses. In this multidisciplinary study, we used nonreflex outcome measures including conditioned place preference, wheel running activity, and spontaneous activity in dorsal horn and dorsal root ganglion neurons, to assess the efficacy, safety, and cellular mechanisms by which dermorphin [D-Arg2, Lys4] (1–4) amide inhibits ongoing pain-related manifestations in animal models of neuropathic injury. We further examined whether nerve-injured and naive rats develop antiallodynic tolerance and opioid-induced hyperalgesia to dermorphin [D-Arg2, Lys4] (1–4) amide after repeated systemic drug administrations. Our findings from complementary behavioral, electrophysiologic, and high-throughput in vivo calcium imaging studies suggest that acute systemic administration of dermorphin [D-Arg2, Lys4] (1–4) amide, a peripherally acting μ-opioid receptor agonist, suppresses spontaneously active small-diameter dorsal root ganglion and dorsal horn wide-dynamic range neurons, and induces peripheral inhibition of ongoing pain-related manifestations without adverse effects. However, long-term drug treatment may lead to the development of antiallodynic tolerance and opioid-induced hyperalgesia.

The animal behavioral and electrophysiology studies were conducted in adult male Sprague-Dawley rats (250 to 300 g, Harlan Bioproducts for Science, USA) and C57BL/6 mice (20 to 30 g, Jackson Laboratory, USA). For green fluorescent–calmodulin–M13 fusion protein calcium imaging, we utilized male and female pirt-green fluorescent–calmodulin–M13 fusion protein6 mice, which we generated by crossing Rosa26-loxP-STOP-loxP-green fluorescent–calmodulin–M13 fusion protein6 mice with pirt-cre mice. These mice exhibit green fluorescent–calmodulin–M13 fusion protein6 expression specifically in dorsal root ganglion and trigeminal ganglion neurons.¹⁸  Animals were housed under optimal laboratory conditions with a 12-h light/dark cycle and free access to food and water. Animals were acclimatized to laboratory conditions before the tests. All behavioral experiments were carried out between 9:00 am and 5:00 pm by an investigator blinded to the drug assignment. The experimental protocols were approved by the Animal Care and Use Committee of Johns Hopkins University (Baltimore, Maryland) and complied with the National Institutes of Health Guide for the Use of Experimental Animals to ensure minimal animal use and discomfort.

Dermorphin [D-Arg2, Lys4] (1–4) amide was purchased from United States Biological (USA). Loperamide hydrochloride, methylnaltrexone bromide, and 2-hydroxypropyl-β-cyclodextrin were purchased from Sigma-Aldrich (USA). Loperamide was dissolved in 20% 2-hydroxypropyl-β-cyclodextrin, made by diluting the 40% 2-hydroxypropyl-β-cyclodextrin/water solution (isotonic) with saline. Other drugs were purchased from Tocris Bioscience (United Kingdom). Stock solutions were freshly prepared as instructed by the manufacturer.

L5 spinal nerve ligation surgery was used for the induction of neuropathic pain in rats. The procedure was a modification of that described in our previous studies.^19,20  Briefly, rats were anesthetized with isoflurane (2%; Abbott Laboratories, USA) delivered through a nose cone. Under aseptic conditions, the skin was incised at the midline over the lumbar spine, and the L5, L6, and upper sacral vertebrae were exposed. The left transverse process of the L6 vertebra was removed, and the left L5 spinal nerve was exposed and dissected from the underlying tissue with fine forceps. The left L5 spinal nerve was then tightly ligated with a 6-0 silk suture and cut distally. The muscle layer was approximated with 4-0 silk suture and the skin closed with metal clips. After the surgery, the rats were returned to their cages, kept warm under a heat lamp, and monitored during recovery.

The sciatic chronic constriction injury model was produced in mice as previously described.^21,22  Briefly, mice were anesthetized with isoflurane (2%; Abbott Laboratories) delivered through a nose cone. Under aseptic conditions, one side of the sciatic nerve was exposed at mid-thigh level by blunt dissection through the biceps femoris muscle. The sciatic nerve was further separated from the surrounding tissue and loosely tied with three nylon sutures (9-0 nonabsorbable monofilament; S&T AG, Switzerland). The distance between two adjacent ligatures was approximately 0.5 mm. After hemostasis was confirmed, the muscle layer was closed with a 6-0 silk suture, and the skin was stapled.

Animals were habituated for 30 min per day in an automated three-chamber box. During habituation, they had access to all chambers. The two larger chambers of this apparatus contained distinct visual (vertical stripes vs. triangular shapes) and tactile (smooth floor vs. grooved floor) cues. The third, smaller chamber was interposed between the other two and was devoid of overt spatial cues. On the preconditioning day, behavior was video recorded for 15 min while the animal was again free to explore all three chambers. The results were used to quantify any basal chamber preference or aversion in individual mice. In keeping with previous studies,^2,21  animals that spent more than 80% (greater than 720 s) or less than 20% (less than 120 s) of the total time in any given chamber were eliminated from further testing. The next day, animals received vehicle treatment in the morning session without anesthesia 10 min before being placed in one of the conditioning chambers for 45 min. Four hours later, the same animals received drug treatment (e.g., subcutaneous dermorphin [D-Arg2, Lys4] (1–4) amide: 5, 10 mg/kg; subcutaneous loperamide: 5 mg/kg) and after 10 min were restricted to the opposite conditioning chamber for 45 min. On the postconditioning test day, animals were placed in the same three-chamber box with access to all chambers but received no injection. Their behavior was recorded for 15 min and used to examine chamber preference or aversion. Pairing of drug or vehicle with a given chamber was counterbalanced between groups. An increase in postconditioning time spent in the drug-paired chamber, as compared with preconditioning time in the same chamber, indicated conditioned place preference. Different scores were calculated by subtracting preconditioning time from postconditioning time.

To assess mechanical hypersensitivity to punctuate mechanical stimuli in rats, we measured paw withdrawal threshold to von Frey filaments (Stoelting Co., USA). Each filament (0.38, 0.57, 1.23, 1.83, 3.66, 5.93, 9.13, 13.1 g) was applied for 4 to 6 s to the test area between the footpads on the plantar surface of the hind paw according to the up-down method as described previously.^13,23  The 1.83-g stimulus was applied first. If a positive response occurred, the next smaller von Frey hair was used; if a negative response was observed, the next higher force was used. The test was continued until: (1) the responses to five stimuli were assessed after the first crossing of the withdrawal threshold, or (2) the upper/lower end of the von Frey hair set was reached before a positive/negative response had been obtained. Abrupt paw withdrawal, licking, and shaking were regarded as positive responses. The pattern of positive and negative responses was converted to a 50% threshold value using the formula provided by Dixon.²⁴  To assess mechanical allodynia in mice, we applied calibrated von Frey monofilaments (0.1 and 0.45 g; Stoelting Co.) to the hind paw for approximately 1 s. Each stimulation was repeated 10 times to both hind paws, and paw withdrawal frequency was determined as described previously.²⁵ 

To test for signs of heat hypersensitivity, we used the Hargreaves test,²⁶ which measures paw withdrawal latency to radiant heat stimuli. Animals were placed under individual plastic boxes on a heated glass floor (30°C) and allowed to habituate for at least 30 min before testing. Radiant heat was applied to the plantar surface of each hind paw three times at 3- to 5-min intervals with a plantar stimulator analgesia meter (IITC model 390, USA). Paw withdrawal latencies were measured three times by an electronic timer, with at least 2 min between trials. A cut-off time of 20 s (rat) or 30 s (mouse) was used to avoid sensitization and damage to the skin. The average paw withdrawal latency of the three trials was used for data analysis.

The open field test was used to assess the effect of systemic drug administration on spontaneous exploration and locomotor activity of the rats.¹³  Rats were placed in an open field chamber (73 × 45 cm rectangular plastic box with a wall height of 33 cm) for 10 min. We analyzed parameters such as total distance traveled; mean travel speed; and number of border periphery, internal periphery, and center crossings in video recordings using SMART 3 software (Panlab Harvard Apparatus, Spain).

Voluntary activity of the mice was measured inside their cages by using Spontaneous Activity Wheels from Bioseb (USA). These wheels have a liquid crystal display and attach to a computer for data generation and analysis. Measurements included the distance run both ways; the number of wheel turns; average, minimum, and maximum speed; acceleration; and total time in the wheel as shown in a previous study.²⁷ 

We used a modified paradigm in which repeated systemic drug injections were used to induce opioid tolerance in spinal nerve ligation rats.²⁸  On days 5 to 9 post–spinal nerve ligation, we tested the rats to obtain baseline paw withdrawal thresholds in the morning. Then we injected dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg, subcutaneous) or saline (subcutaneous), and repeated the paw withdrawal threshold tests after 30 min. In the afternoon of days 5 to 8 post–spinal nerve ligation, we administered the same dose of dermorphin [D-Arg2, Lys4] (1–4) amide without conducting behavior tests.

We used a modified paradigm to examine opioid-induced hyperalgesia in naive rats.²⁹  We treated rats with systemic, fixed-dose dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg, subcutaneous) or saline once daily for 10 consecutive days. Each day, we measured thermal nociceptive threshold paw withdrawal latency and mechanical paw withdrawal threshold before drug injection (baseline), and at 30 min and 60 min after injection, to evaluate opioid-induced hyperalgesia and tolerance.

Rats were housed and habituated for three days in individual metabolic cages (Harvard Apparatus, USA) with known amount of food and water. Each rat then received twice-daily subcutaneous injections of loperamide (5 mg/kg), dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg), or vehicle (n = 5/group) for three consecutive days. Metabolic cages allow a separation of feces and urine and measurement of food and water intake. We measured the daily number of stools produced, urine volume, and amount of food intake at baseline and on each day of drug treatment.

Rats received subcutaneous injections of loperamide (5 mg/kg), dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg), or vehicle (n = 5/group) twice daily for three days. The distal colon motility assay was performed before drug treatment (baseline) and at 30 to 45 min after the first drug or vehicle injection on days 1 and 3. In each test, a 2-mm glass bead was gently inserted approximately 3 cm into the rectum of rats under brief and light anesthesia with isoflurane (1%) to avoid stress and discomfort. The animal was then placed in a white box and allowed to wake up. We recorded the time required for the glass bead to be expelled after insertion (i.e., expulsion time).

We conducted extracellular recordings of single dorsal horn neuronal activity as described in our previous studies.^13,23  Briefly, a laminectomy was performed at vertebral levels T12–L1 corresponding to lumbar enlargements at spinal segments L3–S1. During neurophysiologic recording, animals were anesthetized with 1.5% isoflurane and paralyzed by an intraperitoneal injection of pancuronium bromide (0.15 mg/kg; Elkins-Sinn Inc., USA) to facilitate controlled ventilation. Wide-dynamic range neurons with defined receptive fields in the plantar region of the hind paw were selected. Wide-dynamic range cells were identified by their characteristic responses to mechanical test stimulation at the skin receptive field,^30,31  and spontaneous activity of the neurons was examined before and after drug treatment. Analog data were collected with a real-time, computer-based data acquisition and processing system (CED Spike 2, United Kingdom).

Adult male and female pirt-green fluorescent–calmodulin–M13 fusion protein6 mice were anesthetized with 1.5% isoflurane, and the lumber L4 DRG ipsilateral to the nerve injury was exposed as described in our previous study.¹⁸  Care was taken not to remove the epineurium. Animals were placed under a 5.0× long-working distance objective lens of a confocal microscope (Leica LSI, Germany) for imaging. Time-lapse z-stacks of the intact dorsal root ganglion were acquired at 8 s/frame at 512 × 512 pixel resolution. Body temperature of the mice was maintained at 37.0°C for the duration of the imaging experiment. Electrical test stimulation (2 mA, 0.2 ms, 1 Hz) high enough to activate both A- and C-fibers was delivered through a pair of 30-gauge transdermal needles inserted into the glabrous skin of the hind paw. The train burst width was 8 s (time it takes to record one z-stack). Drug (dermorphin [D-Arg2, Lys4] (1–4) amide, 5μM) or vehicle at a volume of 1 μl was applied directly to the dorsal root ganglion while recording was in progress. After 24 s of baseline imaging, the test stimulus was applied to the hind paw for evoked response to electrical stimulation. For spontaneous activity recording, anesthesia level was reduced to 1% isoflurane and the dorsal root ganglion was imaged continuously for 5 min with no stimulation. Raw images were exported and analyzed with ImageJ (National Institutes of Health, USA) as previously described.¹⁸  An experimenter manually traced activated cells and determined cell size and relative fluorescent intensity. Calcium signal amplitudes were expressed as a ratio of fluorescence difference to basal fluorescence. Small-, medium-, and large-diameter neurons were defined as having somal areas of less than 450 μm², 450 to 700 μm², and greater than 700 μm², respectively.

The number of animals used in each study was based on our experience with similar studies.^13,21  We randomized animals to the different treatment groups and blinded the experimenter to drug treatment to reduce selection and observation bias. None of the variables had missing data. STATISTICA 6.0 software (StatSoft, Inc., USA) was used to conduct all statistical analyses. Two-way repeated measures ANOVA was used for statistical comparison in the conditioned place preference test. Two-way mixed model ANOVA was used for statistical comparison in the wheel running activity, open field test, colon motility assay, electrophysiology recording of wide-dynamic range neurons, and dermorphin [D-Arg2, Lys4] (1–4) amide tolerance assay. One-way repeated measures ANOVA was used for statistical comparison in the green fluorescent–calmodulin–M13 fusion protein imaging and opioid-induced hyperalgesia studies. The specific method for statistical comparison in each study is also given in the figure legends. A Bonferroni post hoc test was used to compare specific data points. Data are expressed as mean ± SD. All tests were two-tailed and P < 0.05 was considered significant in all tests.

Relief of the aversive state associated with ongoing pain may be rewarding and motivate animals to seek a context associated with that relief (negative reinforcement), a behavior that can be assessed by conditioned place preference.^6,32,33  Accordingly, we first examined whether subcutaneous injection of dermorphin [D-Arg2, Lys4] (1–4) amide at 5 mg/kg and 10 mg/kg, doses known to alleviate neuropathic mechanical and heat hypersensitivity,¹³  induced conditioned place preference in rats at 2 weeks after an L5 spinal nerve ligation. This time point corresponds to the peak maintenance phase of neuropathic pain in this model.³⁰  Spinal nerve ligation rats spent more time in the dermorphin [D-Arg2, Lys4] (1–4) amide–paired chamber in the postconditioning period (402.4 ± 61.3 s, 10 mg/kg, n = 9) than in the preconditioning period (322.1 ± 45.0 s, P = 0.009; fig. 1A). In addition, spinal nerve ligation rats spent less time in the saline-paired chamber in the postconditioning period (253.8 ± 42.3 s) than during preconditioning (325.5 ± 60.1 s, P = 0.027; fig. 1A). At both 10 mg/kg and 5 mg/kg (n = 8) doses, the difference scores for the dermorphin [D-Arg2, Lys4] (1–4) amide–paired chambers were significantly greater than those for the saline-paired chambers (10 mg/kg: 80.41 ± 71.9 s vs. –71.7 ± 79.8 s, P < 0.001; 5 mg/kg: 56.1 ± 94.3 s vs. –93.5 ± 59.8 s, P = 0.002; fig. 1, B and C). Importantly, systemic dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg) did not induce conditioned place preference in naive rats, which spent similar amount of time in the dermorphin [D-Arg2, Lys4] (1–4) amide–paired chamber in the post- and preconditioning period (333.4 ± 55.6 s vs. 301.9 ± 48.7 s, P = 0.176, n = 8; fig. 1A).

Another peripherally acting μ-opioid receptor–preferring agonist, loperamide (5 mg/kg, subcutaneous, n = 8),^9,19  also induced conditioned place preference in spinal nerve ligation rats, as indicated by the significant increase in time spent in the drug-paired chamber after conditioning (407.5 ± 60.9 s), as compared to the preconditioning period (304.3 ± 55.7 s, P = 0.003; fig. 1D). The difference score for the loperamide-paired chambers (103.2 ± 66.8 s) was also greater than that for the vehicle-paired chambers (–75.4 ± 100.4 s, P < 0.001; fig. 1E). Gabapentin (60 mg/kg, intraperitoneal, n = 8), which has been used clinically to attenuate neuropathic pain, was included as a positive drug control. Spinal nerve ligation rats spent more time in the gabapentin-paired chamber during the postconditioning period (370.8 ± 61.9 s) than during preconditioning period (272.1 ± 64.1 s, P = 0.026; fig. 1F). In addition, they spent less time in the saline-paired chamber in the postconditioning period (237.6 ± 68.2 s) than in the preconditioning period (344.1 ± 68.5 s, P = 0.016). The difference score for the gabapentin-paired chambers (98.9 ± 99.4 s) was significantly greater than that for the saline-paired chambers among spinal nerve ligation rats (–106.5 ± 95.3 s, P = 0.008; fig. 1G). Gabapentin did not induce conditioned place preference in naive rats (n = 8).

To extend our findings to other species, we conducted the conditioned place preference test in mice at 1 week after chronic constriction injury of the sciatic nerve. Similar to spinal nerve ligation rats, chronic constriction injury mice spent more time in the dermorphin [D-Arg2, Lys4] (1–4) amide–paired chamber after conditioning with a bolus injection of dermorphin [D-Arg2, Lys4] (1–4) amide (2 mg/kg, subcutaneous), as compared to that in the preconditioning period (437.8 ± 59.4 s vs. 351.3 ± 95.9 s, n = 8, P = 0.047; fig. 2A). In addition, chronic constriction injury mice spent less time in the saline-paired chamber in the postconditioning period (332.3 ± 38.2 s) than in the preconditioning period (415.5 ± 102.7 s, P = 0.034; fig. 2A).

Changes in voluntary wheel running may constitute an indicator of daily wellbeing and an objective way to measure the overall impact of aversive state after injury in rodents.²⁷  Continuous dermorphin [D-Arg2, Lys4] (1–4) amide (3 mg · kg^–1 · day^–1, subcutaneous, n = 6) or vehicle (n = 6) was delivered to mice with an implanted osmotic pump for seven consecutive days beginning on day 7 post–chronic constriction injury. Wheel running activity (distance and speed of travel) was sharply decreased in mice after injury, as compared to that after sham surgery (n = 5). Chronic constriction injury mice that received dermorphin [D-Arg2, Lys4] (1–4) amide ran longer distances and traveled at higher speeds on days 5 to 7 posttreatment than those that received vehicle (fig. 2, B and C). In a subgroup of mice that received sensory tests (n = 4/group), the ipsilateral paw withdrawal latency to heat stimulation was significantly decreased from preinjury baseline (vehicle: 22.5 ± 3.6 s; dermorphin [D-Arg2, Lys4] (1–4) amide: 17.8 ± 2.4 s) on day 7 post–chronic constriction injury (vehicle: 8.1 ± 0.9 s, P < 0.001; dermorphin [D-Arg2, Lys4] (1–4) amide: 6.7 ± 1.6 s, P < 0.001; fig. 2D). However, paw withdrawal latency in chronic constriction injury mice after seven days of dermorphin [D-Arg2, Lys4] (1–4) amide treatment (14.4 ± 2.1 s) was significantly increased from predrug level (P = 0.008). In contrast, paw withdrawal latency in chronic constriction injury mice after vehicle treatment (8.6 ± 2.2 s) remained significantly decreased from preinjury baseline (P < 0.001). The ipsilateral paw withdrawal frequency to mechanical stimuli in chronic constriction injury mice at day 7 after receiving vehicle treatment (von Frey, 0.1 g: 47.5 ± 12.6%; 0.47 g: 60.0 ± 14.1%) was significantly increased from preinjury baseline (0.1 g: 7.5 ± 5%, P < 0.001; 0.47 g: 12.5 ± 5.0%, P < 0.001; fig. 2E). In contrast, on day 7 after dermorphin [D-Arg2, Lys4] (1–4) amide treatment, the paw withdrawal frequency of chronic constriction injury mice (0.1 g: 17.5 ± 5.0%; 0.47 g: 30.0 ± 14.1%) was comparable to that at preinjury baseline (0.1 g: 10.0 ± 0%, P = 1.0; 0.47 g: 17.5 ± 12.6%, P = 1.0; fig. 2E) and in the sham-operated group (0.1 g: 12.5 ± 5.0%; 0.47 g: 22.5 ± 9.6%). Paw withdrawal frequency in chronic constriction injury mice on day 7 after dermorphin [D-Arg2, Lys4] (1–4) amide treatment was also significantly lower than that after vehicle treatment (0.1 g: P = 0.003; 0.47 g: P = 0.041).

We further tested whether the dose of dermorphin [D-Arg2, Lys4] (1–4) amide that induced conditioned place preference caused motor incoordination or constipation, two well-known μ-opioid-related side effects. Similar to our previous findings in the open field test, total distance traveled (in 10 min), center travel distance, time spent in the center, number of center crossings, and travel speed in the center and periphery were similar in spinal nerve ligation rats, before and 45 min after one bolus injection of dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg, subcutaneous, n = 6; fig. 3, A and B), and were comparable to those of the saline-treated group (n = 6). We then examined whether repeated systemic injection of loperamide (5 mg/kg, twice daily, subcutaneous), dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg, twice daily, subcutaneous) or vehicle (n = 6/group) for three consecutive days induced constipation or impaired distal colon motility. Compared to predrug baseline, loperamide significantly decreased the daily number of fecal pellets (baseline: 33.8 ± 8.1; day 1: 16.8 ± 4.8, P < 0.001; day 2: 20 ± 8.5, P = 0.006; day 3: 20.2 ± 3.3, P = 0.007) and food intake (baseline: 21.5 ± 4.5 g; day 1: 10.2 ± 5.8 g, P = 0.032; day 2: 13.5 ± 4.5 g, P = 0.045; day 3: 13.4 ± 3.1 g, P = 0.029; fig. 3C), suggesting the development of constipation. Dermorphin [D-Arg2, Lys4] (1–4) amide did not induce significant changes in these outcome measures from predrug baseline. We further examined distal colon motility after three twice-daily treatments with loperamide (5 mg/kg), dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg), or vehicle (n = 5/group). Loperamide significantly increased the expulsion time on day 3 (565.0 ± 78.3 s, P = 0.005; fig. 3D), as compared to baseline (219.7 ± 97.1 s). In contrast, dermorphin [D-Arg2, Lys4] (1–4) amide did not significantly change the time for bead expulsion on day 3 (225.0 ± 99.6 s), as compared to that on day 1 (185.7 ± 106.3 s, P = 1.0).

After nerve injury, a subpopulation of dorsal root ganglion neurons may develop spontaneous activity and ectopic discharge, the neurophysiologic correlates of spontaneous pain. We directly tested whether dermorphin [D-Arg2, Lys4] (1–4) amide attenuates both spontaneous and evoked activities in primary sensory neurons in pirt-green fluorescent–calmodulin–M13 fusion protein6 mice after nerve injury.¹⁸  Electrical test stimulation (2 mA, 0.2 ms, 1 Hz, 8-s burst) was delivered to pirt-green fluorescent–calmodulin–M13 fusion protein6 mice through transdermal needles inserted into the glabrous skin of the hind paw, and the cell sizes of activated neurons were recorded. Small-, medium-, and large-diameter neurons were defined as having somal areas of less than 450 μm², 450 to 700 μm², and greater than 700 μm², respectively. At 5 to 7 days after sciatic chronic constriction injury in mice, topical application of dermorphin [D-Arg2, Lys4] (1–4) amide (5 μM, 1 μl) to the lumbar L4 dorsal root ganglion reduced the evoked activities in small-diameter neurons at 10 min (15.5 ± 5.5), as compared to those at predrug baseline (28.2 ± 8.2, P = 0.009, n = 5 experiments; fig. 4A). Dermorphin [D-Arg2, Lys4] (1–4) amide did not significantly alter the evoked activities in medium- and large-diameter neurons (baseline: 24.3 ± 12.9, 10 min: 20.7 ± 11.2, P = 1.0). In chronic constriction injury mice before drug treatment, the spontaneous activity (cells/frame) was greater in small-diameter (1.5 ± 0.3) than in large-diameter neurons (0.1 ± 0.2, P = 0.037; fig. 4B). The small-diameter neurons showed a significant reduction in spontaneous activity at 5 min after dermorphin [D-Arg2, Lys4] (1–4) amide treatment (1.1 ± 0.4, P = 0.044; fig. 4B).

To further examine the site of action and cellular mechanisms by which systemic dermorphin [D-Arg2, Lys4] (1–4) amide inhibits ongoing pain, we performed in vivo electrophysiologic studies to examine changes in spinal wide-dynamic range neuronal response. Wide-dynamic range neurons receive converging afferent inputs in A-fibers and C-fibers and develop spontaneous activity after nerve injury. Systemic administration of saline (n = 8) in spinal nerve ligation rats did not change spontaneous activity of wide-dynamic range neurons. However, systemic administration of dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg, subcutaneous, n = 8) significantly reduced the spontaneous firing rate of wide-dynamic range neurons to 51.2 ± 46.4% (P = 0.039) and 53.2 ± 46.6% (P = 0.047) of predrug baseline at 20 and 30 min posttreatment (fig. 5A). Pretreatment with methylnaltrexone (5 mg/kg, n = 9, intraperitoneal), a peripherally acting μ-opioid receptor–preferring antagonist, 10 min before dermorphin [D-Arg2, Lys4] (1–4) amide treatment, blocked the inhibitory effect of dermorphin [D-Arg2, Lys4] (1–4) amide on spontaneous activity of wide-dynamic range neurons in spinal nerve ligation rats (fig. 5A).

In behavioral tests, pretreatment of spinal nerve ligation rats with methylnaltrexone (5 mg/kg, intraperitoneal, 10 min), but not saline, also blocked conditioned place preference to dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg, subcutaneous; fig. 5B). In the saline-pretreated group, spinal nerve ligation rats spent more time in the dermorphin [D-Arg2, Lys4] (1–4) amide–paired chamber after conditioning with dermorphin [D-Arg2, Lys4] (1–4) amide (434.7 ± 61.6 s), as compared to that during preconditioning (379.0 ± 77.1 s, P = 0.005, n = 7; fig. 5B). These rats also spent less time in the saline-paired chamber in the postconditioning period (259.1 ± 24.9 s) than during preconditioning (334.9 ± 60.9 s, P = 0.011). In addition, the difference score of time spent in the dermorphin [D-Arg2, Lys4] (1–4) amide–paired chamber (55.7 ± 28.7 s) was significantly higher than that in saline-paired chambers (–75.9 ± 46.3 s, P < 0.001; fig. 5C). In contrast, the difference score in the methylnaltrexone-pretreated group (–110.3 ± 69.9 s) was significantly lower than that in the saline-paired chambers (21.3 ± 102.2 s, n = 9, P = 0.034). Spinal nerve ligation rats pretreated with methylnaltrexone spent significantly less time in the dermorphin [D-Arg2, Lys4] (1–4) amide–paired chamber after conditioning (227.5 ± 68.2 s) than during preconditioning (324.2 ± 56.9 s, P = 0.011).

Unlike in nerve-injured rats, wide-dynamic range neurons in naive rats rarely showed spontaneous activity, similar to what we have reported previously.³⁰  The evoked responses of wide-dynamic range neurons to a supra-threshold electrical stimulus were separated into a short latency A-fiber component (0 to 100 ms), and a longer latency C-fiber component (100 to 500 ms; fig. 6A). In naive rats, dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg, subcutaneous, n = 8) attenuated the stimulus-response function (fig. 6B), and significantly decreased the total C-component (12.1 ± 8.4 action potentials) of wide-dynamic range neurons to graded intracutaneous electrical stimuli (0.1 to 5 mA, 2 ms) from baseline (29.2 ± 13.1 action potentials, P = 0.024; fig. 6C). The A-component was not changed by dermorphin [D-Arg2, Lys4] (1–4) amide. Pretreatment with methylnaltrexone (5 mg/kg, intraperitoneal, n = 6) 10 min before dermorphin [D-Arg2, Lys4] (1–4) amide blocked dermorphin [D-Arg2, Lys4] (1–4) amide–induced inhibition of the C-component (fig. 6, D and E).

Recent studies suggested that activation of transient receptor potential cation channel subfamily V member 1-expressing dorsal root ganglion neurons may play an important role in spontaneous pain.^2,34  To delineate the role of transient receptor potential cation channel subfamily V member 1-expressing neurons in dermorphin [D-Arg2, Lys4] (1–4) amide–induced amelioration of ongoing pain in nerve-injured rats, we injected adult rats at day 7 post–spinal nerve ligation with vehicle (n = 6) or resiniferatoxin (0.1 mg/kg, intraperitoneal, n = 7). Resiniferatoxin is an ultra-potent and selective transient receptor potential cation channel subfamily V member 1 agonist that desensitizes the receptors and decreases the excitability of transient receptor potential cation channel subfamily V member 1-expressing neurons.³⁴  Our recent study showed that heat hypersensitivity was abolished in spinal nerve ligation rats 7 to 9 days after resiniferatoxin treatment.¹³  Like in the previous experiment (fig. 5A), the spontaneous activity rates of wide-dynamic range neurons gradually decreased from 10 to 30 min after dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg, subcutaneous) in vehicle-pretreated spinal nerve ligation rats (fig. 7A). Strikingly, the baseline spontaneous activity rates of wide-dynamic range neurons were significantly lower in resiniferatoxin-pretreated spinal nerve ligation rats (0.5 ± 0.4 action potentials) than in vehicle-pretreated spinal nerve ligation rats (7.5 ± 8.7 action potentials, P = 0.039) and were not decreased after dermorphin [D-Arg2, Lys4] (1–4) amide treatment (fig. 7A).

In a behavior study, vehicle-pretreated spinal nerve ligation rats spent more time in the dermorphin [D-Arg2, Lys4] (1–4) amide–paired chamber after conditioning with systemic dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg, subcutaneous, 429.0 ± 58.7 s) than before conditioning (336.6 ± 74.4 s, n = 7, P = 0.007; fig. 7B). Additionally, the difference score for the dermorphin [D-Arg2, Lys4] (1–4) amide–paired chamber (92.4 ± 56.3 s) was significantly greater than that for the saline-paired chamber (–36.1 ± 71.4 s, P = 0.005; fig. 7C). In contrast to vehicle-pretreated spinal nerve ligation rats, those pretreated with resiniferatoxin (n = 11) did not show preference for the dermorphin [D-Arg2, Lys4] (1–4) amide–paired chamber after dermorphin [D-Arg2, Lys4] (1–4) amide conditioning (fig. 7, B and C).

On days 5 to 7 post–spinal nerve ligation, ipsilateral paw withdrawal threshold was significantly increased from the preinjection baseline (3.5 ± 1.9 g) at 30 min after systemic administration of dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg, subcutaneous, n = 8; day 5: 17.5 ± 6.2 g, P = 0.002; day 6: 13.2 ± 6.1 g, P = 0.008; day 7: 10.7 ± 7.3 g, P = 0.016).¹³  The drug effect gradually decreased after five days of repeated dermorphin [D-Arg2, Lys4] (1–4) amide injections (n = 8/group; fig. 8A), and was mostly gone by four to five days of drug treatment (i.e., days 8 to 9 post–spinal nerve ligation), suggesting the development of antiallodynic tolerance. Injection of saline did not significantly change paw withdrawal threshold in spinal nerve ligation rats (n = 8) or sham-operated rats (n = 8), as compared to preinjection baseline on day 5 post–spinal nerve ligation.

Repeated systemic administration of fixed-dose dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg, subcutaneous, n = 8) once daily for 10 consecutive days led to the development of mechanical (fig. 8, B and C) and heat (fig. 8, D and E) hypersensitivity in naive rats. Predrug paw withdrawal thresholds (baseline) on days 7 to 10 of dermorphin [D-Arg2, Lys4] (1–4) amide treatment were significantly decreased from that measured on day 1 (day 7: P = 0.026; day 8: P = 0.003; day 9: P = 0.010; day 10: P = 0.007; fig. 8C). Similarly, thermal nociceptive threshold (paw withdrawal latency) measured before drug injection on days 5 and 7 to 9 of dermorphin [D-Arg2, Lys4] (1–4) amide treatment was significantly lower than that on day 1 (day 5: P = 0.011; day 7: P = 0.047; day 8: P = 0.042; day 9: P = 0.032; fig. 8E). Mechanical paw withdrawal threshold after dermorphin [D-Arg2, Lys4] (1–4) amide treatment was significantly increased from predrug baseline on day 4 (P = 0.046) and day 6 (P = 0.013; fig. 8B). On days 4 to 9 of dermorphin [D-Arg2, Lys4] (1–4) amide treatment, paw withdrawal latencies at 30 min after dermorphin [D-Arg2, Lys4] (1–4) amide injection (day 4: P = 0.034; day 5: P < 0.001; day 6: P = 0.006; day 7: P = 0.044; day 8: P = 0.017; day 9: P = 0.039) were significantly increased from the respective predrug baseline (fig. 8D). Repeated saline injection in naive rats for 10 days did not change paw withdrawal latency or paw withdrawal threshold (fig. 8B–E; n = 6).

Critical to the use of peripherally acting μ-opioids for neuropathic pain treatment is an understanding of their effects on ongoing pain, and knowledge of the underlying neurobiological mechanisms. By examining conditioned place preference and wheel running activity, which are operant behavioral outcome measures, we showed that dermorphin [D-Arg2, Lys4] (1–4) amide treatment alleviates ongoing pain-associated behavior in nerve-injured animals through activation of μ-opioid receptors in the PNS, without adversely affecting gastrointestinal motility or exploratory activity. Furthermore, dermorphin [D-Arg2, Lys4] (1–4) amide attenuated spontaneous neuronal activity after nerve injury. Yet, repeated dermorphin [D-Arg2, Lys4] (1–4) amide treatments may induce antiallodynic tolerance and opioid-induced hyperalgesia.

Nerve-injured rats and mice both developed conditioned place preference to dermorphin [D-Arg2, Lys4] (1–4) amide. Loperamide, another peripherally acting μ-opioid receptors–preferring agonist,¹⁹  also induced conditioned place preference in nerve-injured rats. Importantly, dermorphin [D-Arg2, Lys4] (1–4) amide did not induce conditioned place preference in naive rats, suggesting that conditioned place preference in spinal nerve ligation rats results from the reward of pain relief, rather than a direct activation of brain reward circuitry, which would indicate CNS opioid penetration and an abuse potential.^35,36  The conditioned place preference to dermorphin [D-Arg2, Lys4] (1–4) amide was blocked by methylnaltrexone, a peripherally acting μ-opioid receptor–preferring antagonist, further suggesting a peripheral site of drug action.

The decrease in voluntary wheel running activity after nerve injury may serve as another objective measure of ongoing pain-related behavior²⁷  and might have multiple causes, including the aversive state associated with neuropathic pain, impaired locomotor function, and anhedonia. Travel distance decreased slightly from baseline in sham-operated mice, possibly because of incisional pain and decreased exploration after adaptation. Importantly, continuous subcutaneous infusion of dermorphin [D-Arg2, Lys4] (1–4) amide gradually increased the speed and distance traveled by nerve-injured mice, suggesting improvements in daily wellbeing and functional recovery. At the end of drug treatment, evoked mechanical and heat hypersensitivities were also significantly decreased, further suggesting dermorphin [D-Arg2, Lys4] (1–4) amide–induced pain relief.

In studies of pregnant sheep, dermorphin [D-Arg2, Lys4] (1–4) amide did not cross the placental barrier to a significant extent.^37,38  Additionally, dermorphin [D-Arg2, Lys4] (1–4) amide (10 mg/kg, subcutaneous) did not disturb locomotor function or reduce exploration,¹³  suggesting minimal central side effects after systemic administration. Loperamide is a substrate extruded by P-glycoprotein transporter from the brain endothelial cells, and can be removed quickly from CNS.^9,39–42  Nevertheless, CNS concentrations of loperamide and dermorphin [D-Arg2, Lys4] (1–4) amide were not examined in the current experimental settings. Therefore, we cannot exclude the possibility that drug effects may partially involve central drug action. The pharmacokinetic and pharmacodynamic properties of loperamide and dermorphin [D-Arg2, Lys4] (1–4) amide after subcutaneous administration warrant further investigation, especially after repetitive treatments.

Morphine and loperamide are known to decrease gastrointestinal secretion and motility.⁴³  Yet, three consecutive days of dermorphin [D-Arg2, Lys4] (1–4) amide administration did not reduce gastrointestinal motility. The reasons that dermorphin [D-Arg2, Lys4] (1–4) amide and loperamide have differing effects on the gastrointestinal tract are not currently clear. We speculate that dermorphin [D-Arg2, Lys4] (1–4) amide may be readily absorbed after subcutaneous administration and might not reach μ-opioid receptors in the gastrointestinal tract as effectively as loperamide.^44,45  Dermorphin [D-Arg2, Lys4] (1–4) amide is a tetra-peptide and has several polar groups that facilitate its absorption by the enterocytes.^11,37  In contrast, loperamide has three planar and nonpolar phenyl rings and a tertiary alcohol group, which facilitate access to the gut wall. The distribution of dermorphin [D-Arg2, Lys4] (1–4) amide (subcutaneous) in the gastrointestinal tract has not yet been determined and compared to that of other μ-opioid receptor agonists. Understanding the chemical properties that limit dermorphin [D-Arg2, Lys4] (1–4) amide’s inhibition of gastrointestinal function will be beneficial to the development of opioid analgesics with minimal gastrointestinal adverse effects (e.g., constipation).

Several additional experiments were conducted in response to reviewer concerns. Similar to loperamide,²⁸  repeated dermorphin [D-Arg2, Lys4] (1–4) amide treatments induced antiallodynic tolerance in spinal nerve ligation rats, and gradually increased pain sensitivity in naive rats, indicating opioid-induced hyperalgesia. Recent findings suggested that μ-opioid receptors expressed on nociceptive afferent neurons initiate morphine analgesic tolerance and opioid-induced hyperalgesia.²⁹  Yet, systemic morphine (3.0 mg/kg, subcutaneous) effectively inhibited neuropathic pain-related behavior in loperamide-tolerant spinal nerve ligation rats.²⁸  This finding suggests that central μ-opioid receptors are functional under conditions of peripheral opioid tolerance. The apparent discrepancies may be due to differences in species, pain models, drug doses, and experimental approaches used in these studies. For example, in addition to losing μ-opioid receptors at peripheral axons and soma of nociceptive afferent neurons, conditional μ-opioid receptor knockout mice exhibit loss of μ-opioid receptors on their central terminals, which may play an important role in initiating morphine tolerance and opioid-induced hyperalgesia.²⁹  In contrast, dermorphin [D-Arg2, Lys4] (1–4) amide and loperamide (subcutaneous) may not directly affect μ-opioid receptors in the spinal cord. Rather, they may change the function, but not ablate μ-opioid receptors in the PNS.²⁸  Although beyond the scope of the current study, it is important to examine the mechanisms of tolerance and opioid-induced hyperalgesia to peripherally acting μ-opioids, and further delineate roles of μ-opioid receptors in different compartments of nociceptive afferent neurons.

We also sought to identify the cellular mechanisms by which dermorphin [D-Arg2, Lys4] (1–4) amide inhibits ongoing pain. Both peripheral (e.g., ectopic discharge) and central mechanisms may contribute to ongoing neuropathic pain. After nerve injury, spinal wide-dynamic range neurons showed increased spontaneous activity, which might partially underlie ongoing pain.^30,46  Systemic dermorphin [D-Arg2, Lys4] (1–4) amide inhibited the spontaneously active wide-dynamic range neurons, an effect that might correlate with inhibition of ongoing pain. Importantly, this effect of dermorphin [D-Arg2, Lys4] (1–4) amide was blocked by methylnaltrexone, suggesting that dermorphin [D-Arg2, Lys4] (1–4) amide might inhibit ectopic peripheral inputs. Systemic dermorphin [D-Arg2, Lys4] (1–4) amide at the doses tested did not induce heat antinociception.¹³  However, it reduced the C-component of wide-dynamic range neurons in naive rats, an effect that was blocked by methylnaltrexone. It is possible that multiple neuronal circuitries or compensatory mechanisms contribute to transmission of heat nociception,^47,48  such as those mediated by nociceptive-specific dorsal horn neurons that are important to nociceptive pain.

Recent studies suggested that “heat receptor” transient receptor potential cation channel subfamily V member 1-expressing primary sensory neurons might contribute to ongoing pain.^2,34  In line with this notion, dermorphin [D-Arg2, Lys4] (1–4) amide–induced conditioned place preference was prevented by systemic pretreatment of spinal nerve ligation rats with resiniferatoxin. Resiniferatoxin desensitizes the transient receptor potential cation channel subfamily V member 1 receptor after strong and prolonged activation, and then selectively decreases the excitability of transient receptor potential cation channel subfamily V member 1-expressing neurons.^2,34  Resiniferatoxin reversed heat hypersensitivity in spinal nerve ligation rats.¹³  Thus, it may ameliorate both heat hyperalgesia and ongoing pain by suppressing transient receptor potential cation channel subfamily V member 1-expressing dorsal root ganglion neurons, thereby preventing dermorphin [D-Arg2, Lys4] (1–4) amide from inducing further pain inhibition. Our electrophysiologic finding that resiniferatoxin pretreatment decreased the spontaneous activity of wide-dynamic range neurons in spinal nerve ligation rats supports the behavioral results. Because transient receptor potential cation channel subfamily V member 1 colocalizes with μ-opioid receptors in dorsal root ganglion neurons, dermorphin [D-Arg2, Lys4] (1–4) amide might inhibit ongoing pain and spontaneous neuronal activity by inhibiting transient receptor potential cation channel subfamily V member 1-expressing neurons.

Recent microneurography studies indicated that spontaneous activity in peripheral C-nociceptors might be a biomarker for, and underlie mechanisms of, ongoing pain in patients with painful polyneuropathy and fibromyalgia.^49,50  However, identifying spontaneously active dorsal root ganglion neurons in vivo has been challenging with conventional electrophysiologic recordings, as it requires sampling a large cell population. By generating pirt-green fluorescent–calmodulin–M13 fusion protein6 mice, we validated a high-throughput calcium imaging technique to examine dorsal root ganglion neuronal activity in vivo.^18,51,52  Green fluorescent–calmodulin–M13 fusion protein6 is a genetically encoded calcium indicator. The intensity of its green fluorescence increases robustly upon binding to intracellular calcium when the cell is active. Therefore, we used it to visualize neuronal activity.^18,53  The pirt promoter is expressed in almost all primary sensory neurons, but not in the CNS, glia, or other peripheral tissue.^18,54,55  The dorsal root ganglion is located outside of the blood-nerve barrier because the endothelium of vessels that supply dorsal root ganglion lacks tight junctions.^56,57  Therefore, dorsal root ganglion neuron soma can be reached by peripherally acting opioids that are administered systemically. Topical application of dermorphin [D-Arg2, Lys4] (1–4) amide to the ganglion attenuated both evoked and spontaneous excitation of dorsal root ganglion neurons. In line with findings from microneurography recording,^49,50  the spontaneously active cells were mainly small-diameter neurons known to be important for pain signaling. Our study supports the premise that spontaneous activity in dorsal root ganglion neurons might represent a useful biomarker for ongoing pain in rodents that can be examined by high-throughput green fluorescent–calmodulin–M13 fusion protein6 imaging.

The lack of effective therapies for neuropathic pain and the increasing morbidity associated with the use of opioids highlights the need for new treatment strategies. The demonstration that peripheral opioid mechanisms may have a role in alleviating neuropathic pain is relatively recent. Nevertheless, the pharmaceutical industry (Nektar, USA; NKTR-181) has begun to examine the clinical utility of a new class of opioids with low blood–brain barrier permeability. Current behavioral and electrophysiologic findings offer evidence that systemic administration of dermorphin [D-Arg2, Lys4] (1–4) amide attenuates ongoing neuropathic pain-related manifestations and spontaneous neuronal activity, by acting at a peripheral location. In light of sex-based differences in neuropathic pain,^58–60  future studies in female rats are needed to examine potential sexual dimorphism in ongoing pain inhibition by dermorphin [D-Arg2, Lys4] (1–4) amide.

Our integrated studies provide proof of concept for using peripherally acting μ-opioids to treat ongoing neuropathic pain, and begin to reveal the underlying neurophysiologic mechanisms. It remains to be determined whether dermorphin [D-Arg2, Lys4] (1–4) amide is more effective than currently used nonopioid medications, such as gabapentin, for neuropathic pain treatment. Potential peripheral side effects and the mechanisms underlying tolerance and opioid-induced hyperalgesia after long-term use of peripherally acting μ-opioids also need carefully investigation. Such studies will further establish the clinical translatability of peripherally acting μ-opioids. Intriguingly, a novel μ-opioid receptor agonist was recently developed by exploiting pathologic conformation dynamics of μ-opioid receptor-ligand interactions. The study demonstrated the feasibility of preferentially targeting the pathologic conformation of μ-opioid receptors in the PNS and produced “injury-restricted inhibition” of inflammatory pain without eliciting respiratory depression, sedation, or constipation.⁶¹  The knowledge gained from these studies is critical to developing peripheral opioid analgesics for neuropathic pain treatment that have minimal adverse effects and limited capacity to induce tolerance and opioid-induced hyperalgesia after prolonged use.

The authors thank Claire F. Levine, M.S. (Scientific Editor, Department of Anesthesiology and Critical Care Medicine, Johns Hopkins University, Baltimore, Maryland) for editing the manuscript, and Qian Huang, Ph.D. (postdoctoral fellow, Department of Anesthesiology and Critical Care Medicine, Johns Hopkins University, Baltimore, Maryland) for technical support.

This study was supported by grants from the National Institutes of Health (Bethesda, Maryland): R01NS70814 (Y.G.), R01NS26363 (S.N.R.), and R21NS99879 (Y.G.). This work was facilitated by the Pain Research Core funded by the Blaustein Fund and the Neurosurgery Pain Research Institute at Johns Hopkins University (Baltimore, Maryland).

The authors declare no competing interests.