Evaluation of different drug classes on transient sciatic nerve injury-depressed marble burying in mice

Abstract

A great need exists for the identification of new effective analgesics to treat sustained pain. However, most preclinical nociceptive assays measure behavioral responses evoked by noxious stimuli (i.e., pain-stimulated behavior), which presents a challenge to distinguish between motor impairing and antinociceptive effects of drugs. Here, we demonstrate that chronic constriction injury of the sciatic nerve (CCI) elicits common pain-stimulated responses (i.e., mechanical allodynia and thermal hyperalgesia) as well as reduces marble burying/digging behaviors that occur during the early stages of the neuropathy and resolve within one week. Whereas drugs representing distinct classes of analgesics (i.e., morphine, valdecoxib, and gabapentin) reversed both CCI-induced and CCI-depressed nociceptive measures, diazepam lacked antinociceptive effects in all assays and the kappa opioid receptor agonist U69593 reversed pain-stimulated, but not pain-depressed behaviors. In addition, we tested drugs targeting distinct components of the endocannabinoid system, including agonists at cannabinoid receptors type 1 (CB₁) and type 2 (CB₂), as well as inhibitors of the endocannabinoid regulating enzymes fatty acid amide hydrolase (FAAH), monoacylglycerol lipase (MAGL). Each of these drugs reversed all CCI-induced nociceptive measures, with the exception of the FAAH inhibitor that reversed pain-stimulated behaviors, only. These findings support the use of the mouse marble-burying assay as a model of pain-depressed behavior within the first week of sciatic nerve injury to examine candidate analgesics. These data also support existing preclinical research that cannabinoid receptor agonists and inhibitors of endocannabinoid regulating enzymes merit consideration for the treatment of pain.

1. Introduction

Pain is associated with numerous disease states and incurs a tremendous cost to society in terms of medical treatment and lost wages. Preclinical animal models of pain typically measure ‘pain-stimulated’ behavior, such as withdrawal responses from mechanical or thermal stimuli. Although numerous drugs elicit antinociceptive effects in pain-stimulated animal models, drug-induced motor impairment often confounds interpretation of these studies. Clinically relevant pain is commonly associated with functional impairment, which is not measured in typical preclinical pain-stimulated assays [43,51,53,64]. A cadre of assays examine pain-depressed behavior using nest building [17,53] burrowing [16], decreases in lever pressing for electrical intracranial self-stimulation (ICSS) [43,47], locomotion [7] and wheel running [48], but the need persists for better preclinical assays to identify new analgesic drugs that have increased translational implications. In the present study, we tested whether the mouse chronic constriction injury of the sciatic nerve (CCI) model of pain would alter marble-burying behavior. The marble-burying assay is sensitive to many environmental manipulations and different drug classes, including benzodiazepines, serotonin reuptake inhibitors, cannabinoids, and opioids. Although decreases in marble burying are often interpreted as evidence for anxiolytic-like activity [37], an alternative view is that it is a consequence of natural digging behavior [27].

First, we tested whether CCI alters marble-burying behavior and whether these changes correlate with time spent digging. Additionally, we measured traditional pain-stimulated behavior, including hypersensitive withdrawal responses from mechanical stimuli (von Frey filaments) and thermal stimuli (hot plate) [5,22]. As mice undergoing the spared nerve injury (SNI) procedure show increases in marble burying [24,55], we also compared marble burying in mice subjected to the CCI procedure versus the SNI procedure. Second, we tested the effects of a variety of pharmacological agents representing different drug classes in this assay. We tested traditional analgesics, including morphine, gabapentin, and a cyclooxygenase-2 (COX2)-selective inhibitor. Two negative control drugs were also included. Kappa opioid receptor agonists commonly produce apparent antinociception in traditional assays of pain-stimulated behavior but fail to produce effective analgesia in clinical trials [52,53]. We also evaluated diazepam, well known to reduce marble burying behavior, but lacks antinociceptive activity in animals and analgesic effects in humans [37,60].

Our final objective was to test whether drugs targeting different components of the endogenous cannabinoid system would alter CCI-depressed and CCI-stimulated behavior. While agonists of the cannabinoid 1 (CB₁) receptor and cannabinoid 2 (CB₂) receptor produce antinociceptive effects in numerous models of sustained pain [21,73], the motor-depressive effects produced by CB₁ receptor stimulation not only reflect a side effect of concern but render tests of pain-stimulated behavior difficult to interpret [14]. CB₂ receptor agonists are devoid of locomotor effects [59]. Inhibitors of the respective chief hydrolytic enzymes of N-arachidonoylethanolamine (anandamide; AEA) [19] and 2-arachidonylglyercol (2-AG) [46,65], fatty acid amide hydrolase (FAAH) [15] and monoacylglycerol lipase (MAGL) [20], produce antinociceptive effects in a variety of sustained pain models [1,25,32,34,35]. Thus, we tested a CB₁/CB₂ receptor agonist, a selective CB₂ receptor agonist, a FAAH and a MAGL inhibitor for alterations in CCI-induced marble burying behavior.

2. Methods

2.1. Animals

Adult male C57BL/6J mice (18–35 grams, Jackson Laboratory, Bar Harbor, ME) served as subjects in these experiments. Mice were housed four per cage in a temperature (20–22 °C), humidity (55 ± 10 %), and light-controlled (12 hour light/dark; lights on at 0600) AAALAC-approved facility, with standard rodent chow and water available ad libitum. All procedures adhered to the ARRIVE guidelines [33] as well as those of the Committee for Research and Ethical Issues of the International Association for the Study of Pain and were approved by the Institutional Animal Care and Use Committee (IACUC) of Virginia Commonwealth University. The sample sizes selected for each treatment group in each experiment were based on previous studies from our laboratory. Specifically, we used a total of 8–12 mice per experimental group where the only, or main, behavioral measure examined was marble burying behavior [37] and we used a total of 6 mice per experimental group where pain-evoked behaviors were examined in conjunction with marble burying behaviors, as this number of animals is sufficient for the detection of significant effects for pain-evoked behaviors [6,29,34,36,72]. In order to ensure reproducibility of our data, we voluntarily triplicated the morphine dose response study, including vehicle treatment, and finding no significant differences in the sets of data, collapsed them together, resulting in a study size of up to 18 mice per group for those experiments. A post-hoc power analysis was performed (paired t-test, PASS 15, NCSS LLC, Kaysville, UT) to verify that group sizes were adequate to detect significant effects.

2.2. Drugs

Gabapentin, diazepam, and the kappa-opioid agonist (+)-(5α,7α,8β)-N-Methyl-N-(7-(1-pyrrolidinyl)-1-oxaspiro(4.5)dec-8-yl)-benzeneacetamide (U69593) were purchased from Sigma (Sigma-Aldrich, St. Louis, MO). The COX2-selective inhibitor, valdecoxib was gifted by Ironwood Pharmaceuticals. Morphine sulfate and the pan cannabinoid agonist (−)-cis-3-(2-hydroxy-4-(1,1-dimethylheptyl)phenyl)-trans-4-(3-hydroxypropyl)cyclohexanol (CP55,940) [44] were obtained from the National Institute on Drug Abuse Drug Supply Program (Bethesda, MD). The CB₂ receptor selective agonist 3-cyclopropyl-1-(4-(6-((1,1-dioxidothiomorpholino)methyl)-5-fluoropyridin-2-yl)benzyl)imidazolidine-2,4-dione hydrochloride (LEI101) was synthesized by the van der Stelt laboratory as described previously [68]. The MAGL inhibitor MJN110 [11,54,72] and the FAAH inhibitor N-3-Pyridinyl-4-((3-((5-(trifluoromethyl)-2-pyridinyl)oxy)phenyl)methyl)-1-piperidinecarboxamide (PF3845) [2] were synthesized by the Cravatt laboratory as described previously. All drugs except morphine were dissolved in a vehicle solution consisting of a mixture of ethanol, alkamuls-620 (Sanofi-Aventis, Bridgewater, NJ), and saline (0.9 % NaCl) in a 1:1:18 ratio. Morphine was dissolved in sterile saline, (Hospira Inc, Lake Forest, IL). Each drug was given via the intraperitoneal (i.p.) route of administration. All drugs were administered in a volume of 10 μl/g body mass. The pretreatment times of drugs tested were based on previously reported time points in which maximal anti-allodynic effects occurred, and dose ranges were based on prior work (Table 1).

Table 1.

Description of the drugs tested in mechanical allodynia, thermal hyperalgesia, and marble burying assays.

Drug	Drug Class	Doses (mg/kg)	Pretreatment Time (min)	Reference(s)
Morphine	Opioid Receptor Agonist	0.3, 1, 3, 10	30	[72]
Gabapentin	Anticonvulsant	0.5, 20, 50	60	[36]
Valdecoxib	COX2 Selective Inhibitor	0.1, 2.5, 5	90	[26]
Diazepam	Benzodiazepine	0.3, 1, 3	60	[37,60]
U69593	K- Opioid Receptor Agonist	0.01, 0.032, 0.1	15	[8]
CP55,940	CB1/CB2 Receptor Agonist	0.01, 0.1, 0.2	30	[34]
LEI101	CB2 Receptor Agonist	5, 20, 40	120	[68]
MJN110	MAGL Inhibitor	0.125, 0.5, 1.25	90	[29,72]
PF3845	FAAH Inhibitor	3, 10, 30	60	[23]

2.3. Surgical Procedures

2.3.1. Chronic constriction injury (CCI) surgery

Following baseline (BL) behavioral assessment, the surgical procedure for chronic constriction of the sciatic nerve was completed as previously described [5] but modified for mouse [29]. In brief, the mice were anesthetized with isoflurane (induction 5% vol. followed by 2.0% in oxygen), and the mid to lower back and the dorsal left thigh were shaved and cleaned with 75% ethanol. Using aseptic procedures, the sciatic nerve was carefully isolated, and loosely ligated with three segments of 5-0 chromic gut sutures (Ethicon, Somerville, NJ). Sham surgery was identical to CCI surgery, but without the loose nerve ligation. The overlying muscle was closed with (1) 4-0 sterile silk suture (Ethicon, Somerville, NJ), and animals recovered from anesthesia within approximately 5 min. Use of chromic gut sutures in this model has been well characterized to produce bilateral allodynia, as well as robust upregulation of bilateral markers of inflammation in the dorsal horn of the spinal cord and the corresponding dorsal root ganglia [49,57,58,71]. Mice were randomly assigned to either CCI or sham surgical group, as baseline measures were not incorporated into the assignment in any way. Mice in both groups were re-assessed for allodynia and thermal hyperalgesia, as described above. Except for the characterization studies in Figures 1 and 2, a single acute administration of drugs or vehicle was employed, with a cross-over design for administration on days 3 and 6 after surgery.

Figure 1.

Mice subjected to CCI surgery bury fewer marbles and spend less time digging than sham mice. A) On post-surgical days 3 and 6, mice in the CCI condition bury fewer marbles than mice in sham surgery group, but marble burying is normalized by day 10. B) CCI surgery leads to reductions in time spent digging on post-surgical days 3 and 6, and digging behavior is again normalized by day 10. C) The time spent digging and number of marbles buried from panels A and B were highly correlated (r = 0.84). *** P < 0.0001, ** P < 0.005, * P < 0.05 vs. sham. Data reflect mean ± SEM, n=8 mice per group.

Figure 2.

Standard analgesics reverse pain-depressed marble burying and pain-stimulated allodynia and thermal hyperalgesia. Morphine, valdecoxib and gabapentin dose-dependently reverse CCI-induced A) decreases in marble burying, B) thermal hyperalgesia, C) allodynia. Filled symbols denote significance from CCI controls, (P < 0.05). D) Morphine (10 mg/kg) does not affect marble burying behavior in vehicle treated mice and does not alter U69593 (0.1 mg/kg)-induced decreases in marble burying. ** P < 0.001, vs. vehicle – vehicle. Data reflect mean ± SEM, n = 6–18 mice/group.

2.3.2. Spared nerve injury (SNI) surgery

Spared nerve injury surgery was performed as described by Decostered and Woolf [18]. Briefly, mice were anesthetized and prepared as described above for CCI surgery. Following exposure of the sciatic nerve, the tibial and common peroneal nerve branches were ligated using silk sutures and transected while leaving the sural nerve intact. As with CCI surgery, the overlying musculature was closed using (1) 4-0 sterile silk suture and the animals recovered from anesthesia within 5 minutes.

2.4. Marble burying assay

Mice were placed in Plexiglas cages (internal dimensions: 33 cm long × 21 cm wide × 19 cm high) with ~ 3 cm of woodchips (7090 Teklad Sani-Chips, Envigo, Somerset, NJ) on bottom and 20 marbles were spaced in grid-like manner (4×5). Marbles used in these experiments were clear, with exception to Digital Supplement Figure 2B where black marbles were used, but this did not noticeably influence digging behavior, or the data obtained. Following the 20-minute test period, subjects were carefully removed to minimize disturbance to the bedding. Marbles at least 75% covered with bedding were considered buried. In the model characterization studies, mice were recorded using Anymaze software (Stoelting Co., Wood Dale, IL) to determine time spent digging. Digging time was scored by an experimenter blinded with respect to group. Digging was operationally defined as use of the hindpaws to kick, or forcefully move, the woodchip bedding. This scoring was highly reliable, as a correlation of r = 0.98 was observed between two independent scorers. The experimenters were blinded with respect to surgical and treatment condition throughout all experiments. For the pharmacological assessment studies that examined both pain-evoked behaviors and marble burying, the order of testing was von Frey, hot plate, and then marble burying was analyzed last. For these studies, time spent digging was not captured. In order to minimize the number of mice needed to complete these studies, mice were tested with drug or vehicle on both day 3 and day 6 after CCI, in a counter-balanced fashion.

2.5. Assessment of pain-stimulated behavior

Mechanical allodynia and thermal hyperalgesia were used to assess nociceptive behavior following sham or CCI surgery (see above). Prior to surgery, mice were habituated to the testing environment. Next, von Frey monofilaments (North Coast Medical, Morgan Hills, CA) were used to establish baseline (BL) responses to light mechanical touch and to assess the development and presence of allodynia after surgery [29]. Specifically, the mice were placed on top of a wire mesh screen, with spaces 0.5 mm apart and habituated for approximately 30 min on four consecutive days before surgical procedures commenced. Mice were unrestrained, and singly placed under an inverted Plexiglas basket (8 cm diameter, 15 cm height), with a wire mesh top to allow for unrestricted air flow. The von Frey test utilizes a series of calibrated monofilaments, (2.83 – 4.31 log stimulus intensity) applied randomly to the left and right plantar surface of the hind paw for 3 s, using a modified “up-down” method [10]. Each paw was stimulated five times with each filament, starting with the 3.61 log stimulus intensity filament and increasing until the mouse responded five out of five times [29]. Lifting, licking or shaking the paw was considered a response. Three or more responses out of five stimulations at the 3.61 log stimulus intensity filament was coded as a positive response. If a positive response was detected, at the 3.61 log stimulus intensity filament, lower weight filaments, starting at 2.84 and then sequentially increasing were used to assess the sensory threshold for each paw until the paw responded five out of five times. After completion of allodynia testing, the mice were placed on a heated (52°C) enclosed Hot Plate Analgesia Meter (Columbus Instruments, Columbus, OH). The latency to jump or lick/shake a hind paw was assessed. A 30 s cut off time was used to avoid potential tissue damage [22]. For each assay, testing was performed in a blinded fashion. For dose-response analysis of each drug, data from the same sham-vehicle and CCI-vehicle groups are included in each appropriate graph depicting marble burying, mechanical allodynia and thermal hyperalgesia.

2.6. Data analysis

All data are presented as mean ± standard error (SEM). For allodynia testing, psychometric behavioral analysis was performed to compute the log stiffness that would have resulted in the 50% paw withdrawal rate, as previously described [67]. Briefly, thresholds were estimated by fitting a Gaussian integral psychometric function to the observed withdrawal rates for each of the tested von Frey hairs, using a maximum-likelihood fitting method [49,71]. Pearson correlations were conducted to examine the relationship between marble burying and digging time, and this data was transformed in a best-fitting curve analysis. Data were analyzed using appropriate inferential statistical analysis of t-tests, one-way analysis of variance (ANOVA), or two-way ANOVA [13]. Except where indicated, Tukey test was used for post hoc analyses of significant one-way ANOVAs. Multiple comparisons following two-way ANOVA were conducted with Bonferroni post hoc comparison.

3. Results

3.1. CCI surgery reduces marble burying and digging behaviors

The first experiment tested whether CCI surgery would alter marble burying and time spent digging compared with mice that had undergone sham surgery. As shown in Figure 1, mice in the CCI group buried fewer marbles than mice in the sham control (P < 0.01, Figure 1A), and displayed a reduction in time spent digging (main effect of surgery, P < 0.01, main effect of time, P < 0.05, Figure 1B) compared with controls. The reduction in these behaviors occurred on post-surgical days 3 and 6 and increased to levels of the control mice by days 10 and 13. Further analysis revealed a positive correlation of time spent digging and the number of marbles buried, (r = 0.84, P < 0.0001, Figure 1C). Locomotor speed (P = 0.4, Supplemental Digital Content Figure 1A) and distance traveled (P = 0.5, Supplemental Digital Content Figure 1B) did not differ between sham and CCI groups. To determine whether repeated exposure to the marble-burying assay accounted for the complete resolution of the observed effects, new groups of mice undergoing sham or CCI surgery were prepared and tested on post-surgical days 3 and 13, only. On day 3, mice in the CCI group buried significantly fewer marbles than the sham mice (P < 0.01, Supplemental Digital Content Figure 1C) and spent less time digging than mice in the sham group (P < 0.05). However, on post-surgical day 13 both groups buried a similar number of marbles (P = 0.40) and spent a comparable duration of time digging, (P = 0.39, Supplemental Digital Content Figure 1D). Thus, CCI surgery depressed marble burying and digging time for the first post-surgical week.

Because previous research using the SNI neuropathic pain model reported increases in marble burying [24,55], we next compared marble burying behavior among SNI, CCI, and sham-operated mice on days 3 and 14 post-surgery. Both the CCI and SNI groups exhibited significant decreases in marble burying compared to sham operated controls on day 3 (P < 0.05, Supplemental Digital Content Figure 2A), no group differences were observed on day 14 (P = 0.7, Supplemental Digital Content Figure 2B). However, mice undergoing SNI or CCI surgery exhibited significant ipsilateral mechanical allodynia (P < 0.0001, Supplemental Digital Content Figure 2C) and thermal hyperalgesia on day 13 compared to sham mice (P < 0.0001, Supplemental Digital Content Figure 2D).

3.2. Evaluation of standard analgesics and diazepam in CCI depressed marble burying and CCI evoked behaviors

We next examined if a panel of drugs (i.e., morphine, gabapentin, and valdecoxib) known to reverse CCI-induced mechanical allodynia and thermal hyperalgesia would also ameliorate CCI-depressed marble burying. Morphine significantly reversed CCI-induced decreases in marble burying (P < 0.05, Figure 2A), thermal hyperalgesia (P < 0.0001, Figure 2B), and mechanical allodynia (P < 0.0001, Figure 2C). Similarly, gabapentin significantly reversed CCI-induced decreases in the marble burying assay (P < 0.001, Figure 2A), thermal hyperalgesia (P < 0.0001, Figure 2B), and mechanical allodynia (P < 0.0001, Figure 2C). Similarly, valdecoxib reversed CCI-depressed marble burying (P < 0.001, Figure 2A), thermal hyperalgesia (P < 0.001, Figure 2B), and mechanical allodynia (P < 0.0001, Figure 2C).

None of the drugs altered marble burying in sham mice (Supplemental Digital Content Figure 3A; morphine: P = 0.7, valdecoxib: P = 0.1, gabapentin: P = 0.7). Morphine elicited antinociception in the hotplate test (P < 0.0001), but neither valdecoxib (P = 0.94) nor gabapentin (P = 0.06) significantly altered basal hotplate responses (Supplemental Digital Content Figure 3B). In sham mice, morphine elevated mechanical stimulus thresholds (P < 0.01), but gabapentin (P = 0.9) and valdecoxib (P = 0.06) did not significantly affect this measure (Supplemental Digital Content Figure 3C).

Next, we examined whether the standard analgesic morphine would reverse the depressive effects of the kappa opioid receptor agonist U69593 on marble burying. Naïve mice given U69593 (0.1 mg/kg) buried fewer marbles compared to the vehicle control mice (Figure 2D). Unlike its effects in reversing CCI-induced depression of marble burying, morphine (10 mg/kg) did not alter marble burying behavior in either vehicle or U69593-injected mice (main effect of U69593, P < 0.0001).

We next we sought to investigate if either the anxiolytic diazepam or U69593 would alter CCI-depression of marble burying or CCI-induced allodynia or thermal hyperalgesia. Both diazepam and U69593 produced further decreases in marble burying within the CCI group (P < 0.05, Figure 3A), but these drugs had differential effects on CCI-stimulated behavior. Whereas diazepam failed to alter CCI-induced thermal hyperalgesia (P = 0.4, Figure 3B), or mechanical allodynia (P = 0.6, Figure 3C), U69593 produced a dose-responsive reversal of CCI-induced thermal hyperalgesia (P < 0.0001, Figure 3B), and mechanical allodynia (P < 0.001, Figure 3C).

Figure 3.

Diazepam and U69593 decrease marble burying behavior but show differential effects in pain-stimulated behavioral assays. A) Neither diazepam nor U69593 reverse decreases in marble burying, but further decreased marble burying at the highest doses tested. B) U69593, but not diazepam reversed thermal hyperalgesia C) and allodynia. Filled symbols denote significance from CCI controls, (P < 0.05). Data reflect mean ± SEM, n = 6–8 mice/group.

Both diazepam and U69593 decreased marble burying in sham mice (diazepam: P < 0.0001, U69593: P < 0.01; Supplemental Digital Content Figure 4A), but neither drug altered hot plate latencies (diazepam: P = 0.1; U69593: P = 0.6; Supplemental Digital Content Figure 4B) or mechanical stimulus thresholds (U69593: P = 0.4; diazepam: P = 0.9; Supplemental Digital Content Figure 4C) in sham control mice.

3.3. Modulation of the endocannabinoid system reverses both pain evoked and pain depressed behavior

Finally, because the endogenous cannabinoid system contains multiple targets that show promise for the treatment of sustained pain, we tested cannabinoid receptor agonists as well as inhibitors of FAAH and MAGL in CCI-induced depression of marble burying. In addition, the same mice were tested for mechanical allodynia and thermal hyperalgesia.

The mixed CB₁/CB₂ receptor agonist CP55,940 dose-dependently reversed marble burying (P < 0.001, Figure 4A), thermal hyperalgesia (P < 0.0001, Figure 4B), and mechanical allodynia (P < 0.0001, Figure 4C) in the CCI group. Likewise, the CB₂ receptor agonist LEI101 dose-dependently reversed CCI-induced decreases in marble burying (P < 0.05, Figure 4A), thermal hyperalgesia (P < 0.01, Figure 4B) and mechanical allodynia (P < 0.05, Figure 4C).

Figure 4.

Drugs targeting different components of the endogenous cannabinoid system differentially reverse CCI-induced nociceptive behavior. The mixed CB₁/CB₂ receptor agonist CP55,940, the selective CB₂ receptor agonist LEI101, and the MAGL inhibitor MJN110 dose-dependently reverse CCI-induced decreases in A) marble burying, B) thermal hyperalgesia, and C) mechanical allodynia. The FAAH inhibitor PF3845 does not reverse CCI-induced depression of A) marble burying but reverses CCI-induced B) thermal hyperalgesia and C) mechanical allodynia. Filled symbols denote significance from CCI controls, (P < 0.05). Data reflect mean ± SEM, n = 6 mice/group.

The MAGL inhibitor MJN110 reversed CCI-induced reduction in marble burying (P < 0.01, Figure 4A), thermal hyperalgesia (P < 0.001, Figure 4B), and mechanical allodynia (P < 0.0001, Figure 4C). In contrast, the FAAH inhibitor PF3845 did not reverse marble burying behavior (P = 0.1591, Figure 4A), but reversed thermal hyperalgesia (P = 0.001, Figure 4B) and mechanical allodynia (P = 0.0004, Figure 4C) in the CCI group. A summary of the drug effects on CCI-induced decreases in marble burying and pain-evoked behaviors can be found in Table 2.

Table 2.

Description of whether test drugs reverse, have no change, or further diminish CCI-induced alterations in marble burying, thermal hyperalgesia, and mechanical allodynia. Additionally, it is noted if these compounds currently have a clinical indication for the treatment of pain.

Drug	Marble Burying	Thermal Hyperalgesia	Mechanical Allodynia	Clinically Used?
Morphine	Reverse	Reverse	Reverse	Yes
Gabapentin	Reverse	Reverse	Reverse	Yes
Valdecoxib	Reverse	Reverse	Reverse	Yes
Diazepam	Further Diminish	No Change	No Change	No
U69593	Further Diminish	Reverse	Reverse	No
CP55,940	Reverse	Reverse	Reverse	No
LEI101	Reverse	Reverse	Reverse	No
MJN110	Reverse	Reverse	Reverse	No
PF3845	No Change	Reverse	Reverse	No

In sham mice, 2 mg/kg CP55,940 significantly decreased marble burying (P < 0.05, Supplemental Figure 5A), as well as elevated hot plate latencies (P < 0.05, Supplemental Digital Content Figure 5B) and withdrawal thresholds in the von Frey assay (P < 0.01, Supplemental Digital Content Figure 5C). Meanwhile, 40 mg/kg LEI101 and 1.25 mg/kg MJN110 did not alter the number of marbles buried (LEI101: P = 0.4, MJN110: P = 0.4, Supplemental Digital Content Figure 5A), hot plate latencies (LEI101: P = 0.3, MJN110: P = 0.3, Supplemental Digital Content Figure 5B), or paw withdrawal thresholds (LEI101: P = 0.6, MJN110: P = 0.1, Supplemental Digital Content Figure 5C) in sham mice. PF3845 produced a small, but significant decrease in the number of marble buried (P < 0.001, Supplemental Digital Content Figure 5A) and reduced hot plate latencies (P < 0.05, Supplemental Digital Content Figure 5B), but did not affect von Frey thresholds (P = 0.6, Supplemental Digital Content Figure 5C).

4. Discussion

Pain disrupts performance of otherwise routine behaviors such as housekeeping, social function, grooming, and can severely impact job performance as well as overall quality of life [41,45,62]. Despite pain-depressed normative life-oriented endpoints, most preclinical studies screening new analgesics employ a cadre of assays using pain-evoked behaviors, such as lifting or licking hind paws in response to light mechanical touch, cold, or heat [51]. Conversely, evaluation of pain-depressed behaviors offers parallel preclinical lines of evidence for screening potential therapeutics and has been hypothesized to predict clinical efficacy better than pain-stimulated pain assays [43,51,53,64]. Examples of pain-depressed assays include reductions in voluntary wheel running [70] and burrowing behavior [3] in rats after CCI or SNI surgery, respectively. We report that CCI or SNI surgery transiently decreases marble burying behavior and overall time spent digging. Moreover, established antinociceptive agents from distinct drug classes (i.e., morphine, gabapentin, and valdecoxib) and a variety of drugs targeting different components of the endocannabinoid system (i.e., cannabinoid receptors and MAGL) reversed CCI-induced depression of marble burying, as well as CCI-evoked behaviors of thermal hyperalgesia and mechanical allodynia.

Whereas CCI and SNI surgery led to mechanical allodynia and thermal hyperalgesia that persisted beyond the second post-surgical week, the decreased marble-burying effect occurred during the first post-surgical week, only. However, these decreases are likely not due to the surgical procedure itself, as sham mice display similar rates of marble burying to surgically naïve vehicle-vehicle treated mice (Figure 2D). Our findings agree with other evidence for robust and sustained expression of many pain-stimulated behaviors in comparison to expression of pain-depressed behaviors [12,43,61]. The high correlation between the number of marbles buried and the time spent digging is consistent with work from Gyertyan [27], who concluded that this assay reflects overall digging behavior. Other data suggest marble burying reflects a repetitive and perseverative behavior [66], rather than a model of anxiety or depression. Given that CCI surgery led to a transient decrease in marble-burying behavior compared with the long duration of hypersensitive withdrawal responses from mechanical and thermal stimuli [36], this may be related to inflammatory and pro-nociceptive mediators associated with the early-phase following the nerve ligation. Alternatively, the relatively quick resolution of this pain-depressed behavior compared with the pain-stimulated behaviors may reflect an adaptive response in which mice rely on such behaviors such as digging to survive [16]. Mechanical allodynia and thermal hyperalgesia are regulated largely via the spinothalamic tract [63], while numerous regions within the CNS (e.g., inputs from the hippocampus, prefrontal cortex, hypothalamus, thalamus, and spinal cord [16,40] regulate digging behavior. Thus, clear anatomical distinctions, specifically regarding higher order brain region inputs, may be responsible for the facilitated resolution of CCI-induced depression of marble burying, compared to CCI-induced thermal hyperalgesia or mechanical allodynia.

Morphine, gabapentin, and valdecoxib are reported to elicit antinociceptive effects in preclinical models of pain [26,34,63]. In the present study, each drug fully reversed both CCI-stimulated nociceptive behaviors and the depressive effects of CCI surgery on marble burying behavior, despite producing pharmacological effects through distinct mechanisms. Specifically, morphine dampens transmission of the sensory and affective components of nociception through activation of the mu opioid receptor, gabapentin dampens neuronal excitability, and valdecoxib elicits anti-inflammatory effects through the inhibition of COX-2. Similarly, the selective COX-2 inhibitor celecoxib decreases upregulation of P2X₃ receptors in dorsal root ganglia and decreases CCI-stimulated nociceptive behaviors when administered early after CCI [69]. Whereas U69593 reversed CCI-stimulated behaviors, it depressed marble burying irrespective of CCI surgery. Similarly, U69593 reverses lactic acid-induced stretching behavior, but not lactic acid-depressed nest building behaviors [53]. While morphine reversed CCI-depressed marble burying, which at higher doses might be due to enhanced locomotion rather than digging, it did not reverse drug-induced depression of marble burying by U69593. This pattern of findings suggests a degree of selectivity for morphine in reversing depression of marble burying by a pain stimulus, but not by a non-pain stimulus. It also highlights the sensitivity of the marble burying assay to drugs that increase locomotion. Diazepam was tested as another negative control, and it did not alter mechanical allodynia or thermal hyperalgesia, and only exacerbated CCI-induced suppression of marble burying behavior. The inhibitory effects of diazepam on marble burying behavior are well described [37,60].

The present study also demonstrates that drugs targeting multiple components of the endocannabinoid system reverse both CCI-stimulated and depressed behaviors at approximately comparable doses for a given drug. Specifically, the CB₁/CB₂ receptor agonist CP55,940 or the selective CB₂ receptor agonist LEI101, reversed CCI-stimulated and CCI-depressed behaviors. This effect is consistent with a recent report that Δ⁹-tetrahydrocannabinol, another mixed CB₁/CB₂ receptor agonist, alleviated migraine pain-related depression of wheel running in rats [30]. However, Δ⁹-tetrahydrocannabinol and CP55,940 lacked efficacy in other pain-depressed assays, including i.p. acid-induced depression of feeding or wheel running in mice [48], i.p. acid-induced depression of feeding or positively reinforced operant behavior in rats, intraplantar formalin-induced depression of operant responding in rats, or noxious-heat-induced depression of operant responding squirrel monkeys [31,39,42,48]. The effects of LEI101 are consistent with previous reports showing that CB₂ receptor agonists reverse CCI-induced behaviors [36,71]. Thus, the effectiveness of cannabinoid receptor agonists in reversing pain-depressed behavior may depend on multiple procedural factors, such as the type of noxious stimulus/injury, the behavioral endpoint, and species.

Indirect modulation of cannabinoid receptors via inhibitors of endocannabinoid-regulating enzymes also holds promise as a potential strategy to treat pain. Specifically, MAGL inhibitors block 2-AG degradation leading to increased levels of this endocannabinoid, and consequently increased signaling at cannabinoid CB₁ and CB₂ receptors [29,34,36,66]. In the present study, the MAGL inhibitor MJN110 reversed CCI-induced behaviors in the von Frey, hot plate, and marble burying assays. Notably, while MJN110 did not affect marble burying in sham mice, the MAGL inhibitor JZL184 reduced marble burying in naïve mice [37]. As shown previously, MJN110 and JZL184 differentially alter rates of operant responding for food administration and locomotor behavior, which may be time- and dose-dependent [29].

The FAAH inhibitor PF3845 led to a different pattern of results than the other drugs acting on the endocannabinoid system. Although PF3845 reversed thermal hyperalgesia and allodynia, as reported elsewhere [6,23], it failed to reverse CCI-induced decreases of marble burying. Likewise, the FAAH inhibitor PF-0445784, did not reverse pain related decreases of burrowing behavior in a rat model of osteoarthritis [9], which is consistent with its failure in a clinical trial for osteoarthritis pain [28]. In contrast, another FAAH inhibitor, URB597, reversed acetic-acid depressed feeding and wheel running behaviors [48] and partially reversed lactic-acid depressed rates of intracranial self-stimulation [38]. Translation of preclinical studies to the clinic may be affected by multiple factors, including the type of noxious stimulus employed, the dependent measures of pain-depressed and pain-stimulated behavior, and differential pharmacokinetics and pharmacodynamics of drugs categorized in the same class of drugs.

In contrast to the observation that CCI decreased the number of marbles buried and time spent digging, others reported that SNI increased marble burying in mice beginning at two weeks post-surgery [55,74]. To ascertain whether the type of nerve injury model accounted for these disparate findings on marble burying, we compared the consequences of CCI and SNI surgery in the marble burying on days 3 and 14. Both surgeries elicited a similar pattern of effects in which marble burying was reduced on day 3 compared with the sham controls, and this pain-depressed behavior resolved by day 14. Although previous studies concluded that SNI-increased marble burying equated to “pain-induced anxiety”, the present study revealed a high correlation between marble burying and digging behavior. Similarly, other research suggests that marble burying reflects a non-goal directed digging behavior [66], which can be affected by numerous environmental and pharmacological manipulations. However, the present study did not assess digging behavior in experiments testing the various pharmacological agents and did not distinguish between goal-directed and incidental behavioral responses that resulted in the wood chips covering the marbles.

The marble-burying assay offers a straightforward procedure with sensitivity to pain-depressed behavior during the early stages of SNI and CCI-induced neuropathy, and is readily reversed by known analgesics (i.e., morphine, gabapentin, and valdecoxib). In addition, a variety of pharmacological agents targeting distinct components of the endocannabinoid system (i.e., cannabinoid receptors and MAGL) reverse both CCI-induced depression of marble burying behavior and CCI-stimulated nociceptive behavior, which adds credence for potential clinical efficacy. One important caveat of our findings is that this assay is only useful for one week after surgery, which may limit its preclinical drug discovery utility. Thus, this assay is not useful for measuring changes in chronic pain. Nonetheless, incorporation of pain-depressed behaviors, such as marble burying, in conjunction with pain-stimulated behaviors are relatively straightforward behavioral assays and may serve to identify new analgesic drugs with increased translational implications for treating patients suffering from pain.