Metabotropic Glutamate Receptor 5 Antagonism with Fenobam: Examination of Analgesic Tolerance and Side Effect Profile in Mice

• The metabotropic glutamate receptor 5 antagonist fenobam produces acute analgesia in rodents, and may be developed for clinical use

• Whether analgesic tolerance develops with chronic fenobam treatment has not been studied

• Two-week daily dosing, a duration that produces extreme tolerance to opioids, did not result in tolerance to fenobam's analgesic effect in mice

GLUTAMATE is the primary excitatory neurotransmitter in the mammalian nervous system. Abnormal glutamatergic signaling may play a prominent role in several disease processes, including chronic pain.1,2The G-protein coupled glutamate receptor metabotropic glutamate receptor 5 (mGlu5) is expressed at synapses throughout the nervous system, in particular the pain neuraxis,3and agents that modulate mGlu5 may have therapeutic potential in the treatment of pain. Significant inroads into the use of pharmacologic agents acting at mGluRs in human patients have been made recently,4,5and findings that activation of mGlu5 is pronociceptive6,–,8whereas its inhibition is antinociceptive9,–,11have led to the suggestion that antagonism of mGlu5 may have analgesic efficacy in humans. However, new pharmacologic agents that target mGlu5 will also have to possess acceptable safety profiles along with clinical efficacy. This is illustrated by the recent discontinuation of clinical trials involving the Addex Pharmaceuticals (Plan-les-Ouates, Switzerland) mGlu5 antagonist ADX-10059, which, despite fourfold improvement versus  placebo in inducing a pain-free state in migraineurs, resulted in unacceptable liver enzyme increases.12 

The recent finding that the clinically validated compound fenobam is a potent and selective mGlu5 negative allosteric modulator13has led to a resurgence in interest in testing the efficacy of fenobam as a treatment for various neurologic conditions. Originally developed in the 1970s as an anxiolytic with a then-unknown mechanism of action, fenobam was found to have a favorable safety profile;14,15however, initial results regarding its anxiolytic efficacy were mixed.16,17Further human clinical testing was discontinued in the early 1980s. Fenobam has subsequently been demonstrated to be analgesic in rodents10,11and moderately effective in managing some of the symptoms of fragile X syndrome in humans.4However, in these studies fenobam was only administered as a one-time dose, and future treatment strategies involving fenobam or other mGlu5 antagonists in pain conditions will likely require longer-term treatment. It is currently unknown whether tolerance develops to fenobam's analgesic effects. In addition, whereas the elevations in liver enzymes seen with ADX-10059 may be compound-specific, the effects of chronic mGlu5 antagonist administration on liver function and other vital systems are not well characterized in the literature.

Here, we performed a series of experiments to determine whether tolerance develops to the analgesic effects of fenobam and to assess for deleterious side effects. We test whether tolerance to the analgesic effects of fenobam develops after repeated dosing of up to 14 days. In addition, we assessed for alterations in liver function, gross liver histopathology, and hematologic abnormalities after chronic fenobam treatment. Sedation and impairment of motor coordination are common dose-limiting factors for some classes of analgesic drugs. Fenobam may act as a stimulant,10,16but it is unknown whether this stimulant effect is due to specific activity at mGlu5. Therefore, we also assessed the locomotor activity of mGlu5 knockout mice treated with fenobam. Finally, mGlu5 has been shown to play a role in weight gain and energy balance.18,19We sought to expand these results by determining if fenobam mediates any effects on appetitive behavior.

Experiments were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals  and were approved by the Animal Care and Use Committee of Washington University School of Medicine (St. Louis, Missouri). Male Swiss Webster mice (5 weeks old) were purchased from Taconic (Germantown, New York) and group-housed five animals to a cage for 1 week before any behavioral experiments were started. For experiments involving mice lacking mGlu5 (mGlu5 knockout; 6–9 weeks old), animals were bred in-house on a C57BL/6 background and compared with wild type (WT) littermates.20All C57BL/6 WT mice used in behavioral experiments were also derived from this colony. For experiments involving knockout animals the experimenter was blinded to genotype. Genotyping of mice bred in-house was performed using standard polymerase chain reaction techniques. All mice were group-housed on a 12/12-light/dark schedule with ad libitum  access to food and water, except as described in the next paragraphs during food deprivation experiments.

Fenobam was purchased from Tocris (Ellisville, MO) or was custom-synthesized for our use by Scynexis (Research Triangle Park, NC) and dissolved in dimethyl sulfoxide (DMSO; Sigma–Aldrich Chemical Company, St. Louis, MO) on the day of the experiment. All intraperitoneal injection volumes were 20 μl. Throughout all experiments the investigator was blinded to pharmacologic treatment. All other reagents were high-performance liquid chromatography grade and purchased from Sigma–Aldrich Chemical Company.

Locomotor activity was measured in an open-field chamber (42 l × 42W × 30H cm) attached to the VersaMax Animal Activity Monitoring System (AccuScan Instruments, Columbus, OH). For experiments involving drug-naive C57BL/6 WT and mGlu5 knockout mice littermates, animals were brought to the testing room in their home cage and acclimated for at least 2 h before testing. Locomotor activity was assessed individually by recording photobeam breaks for 60 min. Total distance traveled, time spent moving, and the number of beam breaks (horizontal activity) were calculated for the entire chamber. For experiments designed to assess the effects of fenobam on locomotor behavior in mGlu5 knockout mice, animals were habituated in their home cages for at least 2 h, and then injected intraperitoneally with either 30 mg/kg fenobam or vehicle (DMSO) immediately before placement in the chamber. Animals were allowed to explore the chamber for 90 min, locomotor activity was measured individually as described above, and the total distance traveled was calculated for the entire chamber.

Locomotor activity was measured in low light conditions using a zero maze (Stoelting Co., Wood Dale, IL) placed 70 cm off of the ground and consisting of two closed sections (wall height, 30 cm) and two open sections (wall height, 1.3 cm) on a circular track (diameter of track, 60 cm). C57BL/6 WT mice were habituated to testing room for 1 h before injections, and then injected intraperitoneally with fenobam (30 mg/kg) or vehicle (DMSO). Fifteen minutes after injection, mice were placed individually in the closed area of the zero maze for a 600-s trial. Movement during the trial was recorded using two digital video cameras (Logitech 9000 Pro webcam, Romanel-sur-Morges, Switzerland). Total distance traveled, number of entries into open sections, and time spent in the open sections was scored off-line using AnyMaze video tracking software (Stoelting Co.).

A series of three tests was performed on drug-naive C57BL/6 WT mice to assess the effects of fenobam on gross motor behavior and coordination. 21,22All tests were performed two times.

Each mouse was tested to see how long it could maintain its balance on a 0.75 cm-wide Plexiglass ledge without falling (60 s maximum). A score of 60 s was also assigned if the mouse traversed the entire 51-cm length of the ledge and returned to the starting point in less than 60 s without falling.

Mice were placed head upward at the top of a vertical metal rod (8 mm diameter, 55 cm height). The rod was finely textured with a file to provide a gripping surface. Mice were given a maximum of 120 s to turn 180 degrees and climb down to reach the bottom of the pole and place all four paws on the tabletop. Mice were required to reverse direction and actively climb down. If a mouse slid down the pole without reversing direction or fell down the pole it was given a score of 120 s.

Mice were placed on a wire mesh screen (16 squares per 10 cm², 47 cm high × 18 cm wide) oriented at 60 degrees. The screen was then inverted to 180 degrees so that mice were upside down. The time mice spent hanging on the screen was measured for 2 min. A maximum score of 120 s was given to an animal that did not fall.

Male Swiss Webster mice received chronic intraperitoneal injection with 30 mg/kg fenobam or vehicle (DMSO) once per day for 5 days. On the sixth day, mice were placed in transparent Plexiglas boxes (10 l × 10W × 15H cm, assembled by the Washington University School of Medicine Instrument Machine Shop) on a glass surface and acclimated for 2 h before any drug injection. Animals were then pretreated by intraperitoneal injection with vehicle or fenobam (30 mg/kg). Ten microliters of dilute formalin solution (5% dissolve in normal saline, Sigma-Aldrich Chemical Company) was injected subcutaneously into the plantar surface of the right hind paw. Some mice from the chronic-vehicle group were injected with fenobam, whereas others received vehicle, followed 5 min later by intraplantar formalin injection. Thus, three separate groups were analyzed: chronic vehicle/acute vehicle; chronic vehicle, acute fenobam; and chronic fenobam, acute fenobam, with chronic administration defined as 5 days. Spontaneous behavior was recorded using a digital video camera (Logitech 9000 Pro webcam, Romanel-sur-Morges, Switzerland) placed underneath the glass platform. Video recordings were scored after the collection of all data by an observer blinded to treatment (both chronic and acute) and the time spent in nocifensive behavior defined as licking, lifting, or flicking of the injected paw, and was scored in 5-min intervals for 45 min after paw injection. A separate set of experiments was performed exactly as just described, except animals were injected with fenobam (30 mg/kg) or vehicle for 14 days instead of 5 days, with formalin testing performed on the 15th day. Animals from this longer time period of injections were killed immediately after the cessation of the formalin test, and tissue obtained from them was used in the serum chemistry analysis and tissue histology described in the following paragraphs. All animals were weighed daily immediately before intraperitoneal injection.

Food deprivation studies were performed in a manner similar to that previously described 18. For experiments designed to assess the acute effects of fenobam on food intake, mice were weighed and housed individually in their home cages 3 days before food deprivation. Cage bedding was replaced with a raised wire mesh bottom. On the third day mice were weighed again and food deprived for 24 h starting 1 h after lights on until 1 h after lights on the next day. Mice were then weighed again and injected intraperitoneally with either 30 mg/kg fenobam or vehicle, and preweighed food was placed onto the cage bottom. Food intake was measured for each mouse at 15, 30, 60, 90, and 180 min, as well as 3 h before lights off (i.e. , 8 h later). To compensate for variation of body weight, mouse food intake was normalized to body weight. Mice were weighed to the nearest 0.1 g and food was weighed to the nearest 0.01 g.

All mice from the 14-day chronic fenobam injection study (see previous paragraphs) were administered an overdose of sodium pentobarbital (75 mg/kg) after cessation of the formalin test. Whole blood was obtained by transcutaneous cardiac puncture and placed into plasma separator tubes with lithium heparin for blood chemistries or EDTA tubes for complete blood counts (BD Microtainer, Franklin Lakes, NJ). Serum chemistry analysis and necropsy were performed immediately after tissue collection.

Complete blood counts were analyzed using the Hemavet 1700 (Drew Scientific, Waterbury, CT). Leukocyte differentials were determined by direct examination of blood smears. Serum chemistries were analyzed using the Vitros DT60 system dry reagent chemistry analyzer (Ortho Clinical Diagnostics, Rochester, NY).

Tissues were formalin-fixed, paraffin-embedded, and sectioned at 5 μm. Sections were prepared according to standard histologic techniques and stained with hematoxylin and eosin.

Statistical analysis of behavioral data was performed using Prism 5.0 (GraphPad Software, Inc., La Jolla, CA). All data collected over multiple time points from the open-field assay, the spontaneous formalin test, and the food deprivation studies were statistically analyzed using a Bonferroni correction multiple comparison test after a nonrepeated measures two-way ANOVA. The factors used for these analyses were genotype/treatment and time. Summed data from these tests were analyzed using a two-tailed Student t  test when comparisons were made between two groups or a Bonferroni multiple comparisons tests after a one-way ANOVA when comparisons were made between more than two groups. In all studies, the accepted level of significance was P < 0.05. In all figures, data are reported as mean ± SEM. Also listed in the text are 95% CIs of the difference between means.

Statistical analysis of data were performed using Prism 5.0 (GraphPad Software, Inc.). All data collected from chronically injected fenobam versus  vehicle-injected animals were statistically analyzed using an unpaired Student t  test. In all studies, the accepted level of significance was P < 0.05. In all figures data are reported as mean ± SEM.

Previous work from our laboratory suggests that mGlu5 inhibition via  fenobam may affect locomotor activity.10We expanded upon previous findings to determine whether altered locomotor behavior after fenobam administration is mediated via  inhibition of mGlu5 or through an off-target effect of the drug. The open-field task was used to assess the effects of mGlu5 deletion and inhibition on locomotor activity. When the effects of genetic deletion of mGlu5 were examined in the open field, we found that drug-naïve mGlu5 knockout mice traveled a statistically significant farther distance in the open field over a 60-min time period in comparison with their drug-naïve WT littermates (7,365 ± 876.7 cm vs.  12,171 ± 1111 cm; 95% CI −7843 to −1770 cm; P < 0.001; fig. 1, A and B). To assess whether the increased locomotor activity after fenobam administration was due to inhibition of mGlu5, knockout mice and WT littermates were injected with fenobam (30 mg/kg) or vehicle (DMSO) and then immediately placed in the open field for 90 min. There was a statistically significant increase in the total distance traveled in the fenobam-treated WT mice and both the vehicle- and fenobam-treated mGlu5 knockout mice compared with vehicle-treated WT mice (95% CI between vehicle-treated WT mice and fenobam-treated WT, vehicle-treated knockout mice and fenobam-treated knockout mice −1,1735 to −2,044 cm, −9,924 to −541.8 cm, and −13,329 to −3,946 cm, respectively; P < 0.001; fig. 1, C and D). No differences were noted between vehicle- or fenobam-treated mGlu5 knockout animals.

Mice injected with fenobam (30 mg/kg) traveled a statistically significant farther distance during a 10-min period in an elevated zero maze when compared with vehicle-injected mice (41.44 ± 3.099 m vs.  15.59 ± 2.651; 95% CI −35.30 to −16.40 m; P = 0.0002). In addition, fenobam-injected mice more frequently entered the open sections of the zero maze (P = 0.001; fig. 2A), spent a statistically significant longer amount of time in the open section of the zero maze (P < 0.01, fig. 2B), and traveled a statistically significant farther distance as a percentage of total distance traveled in the open sections of the zero maze (i.e. , distance in open arms divided by total distance) (P < 0.01; fig. 2C) compared with vehicle-injected control mice.

When knockout mice were compared with WT control mice, no differences were noted in the total distance traveled (WT n = 8, 19.04 ± 1.633 meters vs.  knockout n = 7, 21.49 ± 2.770 meters; 95% CI −9.344–4.452 m; P > 0.05), the number of entries into the open sections of the zero maze (WT 13.75 ± 1.916 vs.  knockout 21.14 ± 4.512; 95% CI −15.70–4.450 entries; P > 0.05), the distance traveled as a percentage of total distance traveled in the open arms (WT 21.41 ± 3.300%vs.  knockout 18.99 ± 4.310%; 95% CI −9.223–14.06%; P > 0.05), knockout mice exhibited reduced time in the open sections of the maze (WT 86.19 ± 13.83 s vs . knockout 156.2 ± 24.43; 95% CI −126.6 to −11.36 s; P = 0.0229).

In comparison with vehicle, fenobam injection resulted in no physiologically or statistically significant differences in performance on three separate tasks designed to assess motor coordination. Mice injected with fenobam (30 mg/kg) performed equivalent to vehicle-injected mice when required to hang from an inverted screen for 120 s (95% CI for trial 1, 52.87–6.106; for trial 2, 33.56–25.41; fig. 3A), descend from a thin (8 mm diameter, 55 cm height) vertically-oriented metal pole (95% CI for trial 1, 48.85–49.06; for trial 2, 46.52–51.39; fig. 3B), and cross a thin (0.75 cm wide, 51 cm long) horizontally oriented ledge (fig. 3C). No physiologically or statistically significant difference between vehicle-injected and fenobam-injected C57 WT mice was observed over two consecutive trials.

To test whether tolerance to the analgesic effects of fenobam develops over time, Swiss Webster WT mice were injected with either vehicle or fenobam (30 mg/kg) for 5 days. On the sixth day, mice from the chronic-vehicle group were injected with either fenobam or vehicle, and mice from the chronic-fenobam group were injected with fenobam 5 min before performing a formalin test. Regardless of whether they had undergone chronic injection of fenobam or vehicle, mice receiving acute injections of fenobam demonstrated statistically significantly less spontaneous nocifensive behavior postformalin injection when compared with mice that had undergone chronic injection of vehicle mice and that received acute injections of vehicle (P < 0.001; fig. 4A). Acute pretreatment with fenobam (30 mg/kg intraperitoneal) statistically significantly reduced the time mice spent licking or lifting the formalin-injected paw during the first phase (95% CI of the mice treated with chronic vehicle/acute vehicle compared with chronic vehicle/acute fenobam and chronic fenobam/acute fenobam 14.57–179.0 and 30.16–200.8 s, respectively; P = 0.0097; fig. 4B) and the second phase compared with 5-day vehicle/vehicle mice, regardless of fenobam pretreatment (95% CI of the mice treated with chronic vehicle/acute vehicle compared with chronic vehicle/acute fenobam and chronic fenobam/acute fenobam 124.2–740.5 and 81.56–721.1 s, respectively; P = 0.0069; fig. 4C).

This experiment was then repeated with a second group of Swiss Webster WT mice, with the one alteration that fenobam or vehicle was injected for 14 days before performing a formalin test on the 15th day. Similar to the 5-day experiment previously discussed, mice undergoing acute treatment with fenobam demonstrated statistically significantly less spontaneous nocifensive behaviors postformalin injection when compared with 14-day vehicle/vehicle mice (P < 0.001; fig. 4D). Mice that were chronically treated with vehicle and then pretreated with fenobam exhibited statistically significantly reduced time spent licking or lifting in the first phase compared with 14-day vehicle/vehicle mice (95% CI 16.02–156.7 s; P = 0.0258; fig. 4E). Acute pretreatment with fenobam (30 mg/kg intraperitoneal) also statistically significantly reduced the time mice spent licking or lifting the formalin injected paw during the second phase compared with 5-day vehicle/vehicle mice, regardless of fenobam pretreatment (95% CI of the chronic and acutely vehicle-treated mice compared with chronic vehicle/acute fenobam and chronic fenobam/acute fenobam 34.97–575.3 and 65.50–583.9 s, respectively; P = 0.0154; fig. 4F).

Serum was collected from mice who received chronic injections of either vehicle or fenobam for 14 days. Serum chemistry values for cholesterol, total bilirubin, triglycerides, creatinine, total protein, alanine-aminotransferase, aspartate-aminotransferase, lactate dehydrogenase, and amylase were measured. No differences in any values measured were seen between fenobam- (30 mg/kg intraperitoneal for 14 days) and vehicle-injected (vehicle intraperitoneal for 14 days) mice (fig. 5A–I; 95% CIs: cholesterol: −23.87–10.17 mg/dl, total bilirubin: −0.07–0.10 mg/dl, triglycerides: −37.00–9.46 mg/dl, creatine: −0.043–0.021 mg/dl, total protein: −0.42–0.13 g/dl, alanine-aminotransferase: −83.38–289.5 units, aspartate-aminotransferase: −32.61–116.4 units, lactate dehydrogenase: −1,310–3,167 units, and amylase: −1,361–2,788 units; P > 0.05 for all comparisons).

Whole blood was also obtained from mice who received chronic injections of either vehicle or fenobam for 14 days and assayed for counts of leukocytes, erythrocytes, hemoglobin, hematocrit, mean corpuscular volume, mean corpuscular hemoglobin, mean corpuscular hemoglobin concentration, or platelets. No differences were found between mice injected daily with fenobam (30 mg/kg intraperitoneal for 14 days) or vehicle (vehicle intraperitoneal for 14 days). (fig. 6A–H; 95% CIs: leukocytes, −0.8969–1.052°K/cu mm, erythrocytes −3.187–0.8982 M/cu mm, hemoglobin −4.845–1.005 g/dl, hematocrit −18.84–4.038%, mean corpuscular volume −7.191 to 2.258 cu microns, mean corpuscular hemoglobin −2.358–0.4581 pg, mean corpuscular hemoglobin concentration −9.907–4.767%, and platelets −342.6–330.8°K/cu mm; P > 0.05 for all comparisons.)

In addition, a 100-count leukocyte differential was performed on all mice from which whole blood was obtained. No differences were noted in the number of segmented neutrophils (all values presented as mean number of cells per cubic mm ± SD) (vehicle = 506.9 ± 140.6; fenobam = 651.8 ± 428.1, P > 0.05), lymphocytes (vehicle = 1,194 ± 725.2; fenobam = 945.3 ± 632.5, P > 0.05), or monocytes (vehicle = 2.5 ± 6.1; fenobam = 5.1 ± 10.8, P > 0.05). No band forms, eosinophils, or basophils were observed from any mice in either injection group.

Several animals from each group were examined by necropsy. In addition, tissue histopathology was performed on brain, liver, and testes on all animals examined by necropsy. Additional organs that were noted to be abnormal by gross necropsy also underwent tissue histopathology on a case-by-case basis. In all animals, the hair, coat, and skin were normal, skeletal palpation revealed no evidence of malformation or trauma, there was no nasal or ocular discharge or diarrhea, and hydration and body fat were normal. Whole body and individual organ weights are detailed in table 1. No physiologically or statistically significant differences were found in total body weight or any organ weight between fenobam- and vehicle-treated mice. The major thoracic and abdominal organs and brain were examined in all animals. No abnormalities were noted in the respiratory system, digestive system, musculoskeletal system, urinary system, genital system, brain, thymus, spleen, lymph nodes, adrenal, thyroid, pituitary, or eye. Gross abnormalities and histopathologic examinations of the brain, liver, testes, and any other organs found to exhibit abnormalities on gross examination are noted in table 2. No physiologically or statistically significant differences were found between the fenobam- and vehicle-treated animals with regard to the type or prevalence of any abnormalities.

Animals injected with fenobam or vehicle for the 14-day tolerance study described previously were weighed daily before injection. When the percentage change in body weight was calculated as a function of starting weight for each animal, daily fenobam injection was found to have a statistically significant effect on percent change in body weight over a 14-day period, compared with vehicle (fig. 7A, P < 0.0001). No differences were found in the average starting weights of the animals (vehicle = 29.34 ± 0.77g, fenobam = 29.42 ± 0.65 g, 95% CI −2.112–1.946 g; fig. 7A, inset, P > 0.05).

The effects of fenobam on food intake after a 24-h food deprivation were assessed in WT Swiss Webster mice (fig. 7B). Fenobam (30 mg/kg) administered at the time of refeeding statistically significantly decreased the amount of food consumed, compared with vehicle-treated animals (fig. 7B, P < 0.001). Fenobam-treated mice consumed more food 60 min (95% CI −36.42 to −11.72 mg/g body weight), 90 min (95% CI −36.97 to −12.28 mg/g body weight) and 180 min (95% CI −26.17 to −1.470 mg/g body weight) after refeeding compared with vehicle-treated mice. However, by 8 h after the start of refeeding food intake of the fenobam-treated mice was not different from vehicle-treated animals (vehicle = 97.3 ± 8.8 mg/g; fenobam = 87.9 ± 7.8 mg/g; 95% CI −16.77–35.55 mg/g body weight; unpaired T -test P = 0.44). No differences in body weight were noted in animals before refeeding (Veh = 32 ± 0.82 g, Fenobam = 31.73 ± 1.25 g; 95% CI −3.052–3.585; unpaired Student t  test P = 0.86).

Activation of mGlu5 with the group I agonist (R,S)-3,5-dihydroxyphenylglycine is proalgesic,6whereas antagonism with fenobam results in decreased nociceptive behaviors.10,11These data provide evidence that mGlu5 may represent a viable therapeutic target for the treatment of pain. However, any future development of mGlu5 antagonists as analgesics in human patients will require both drugs that are selective for mGlu5 and for mGlu5 antagonism at sites outside the pain neuraxis to be devoid of clinically significant deleterious effects. Here we report that drug-naïve mGlu5 knockout mice exhibited increased locomotor activity at multiple time points when compared with WT littermates, and the level of locomotor activity of mGlu5 knockout mice was unchanged after either vehicle or fenobam injection. In contrast, when fenobam was injected into WT mice before placement in the open field, increased locomotor activity was exhibited at multiple time points compared with vehicle-treated mice. This suggests that the increased locomotor activity is due to antagonism of mGlu5. The fact that only a single difference (increased open arm time in knockout mice) was seen in drug-naive mGlu5 knockout mice compared with WT control mice may be due to a compensatory effect caused by genetic deletion of mGlu5 since birth in knockout animals.

Fenobam was originally developed as an anxiolytic,15and when administered to rodents13and humans15it is anxiolytic. Increased locomotion in an open-field enclosure could represent an anxiolytic effect of fenobam. To test this possibility, we assessed the effects of fenobam on drug-naive mice in an elevated zero maze. Fenobam-injected mice entered the open sections of the zero maze significantly more and spent significantly more time in the open sections when compared with vehicle-injected mice. They also traveled significantly farther in the zero maze as a whole and traveled significantly more of this distance in the open sections of the maze when compared with vehicle-injected mice. These findings suggest that the effects of fenobam are both stimulatory and anxiolytic. The fact that these findings were not seen in drug-naive mGlu5 knockout mice compared with WT control mice may be due to a compensatory effect due to genetic deletion of mGlu5 since birth in knockout animals.

Increased locomotor activity after mGlu5 antagonism with fenobam is in opposition to results seen with 2-methyl-6-(phenylethynyl)pyridine9,23and the related mGlu5 antagonist 3-[(2-methyl-1,3-thiazol-4-yl)ethynyl]pyridine,9which have been shown to reduce locomotor activity at analgesic doses. 2-Methyl-6-(phenylethynyl)pyridine has also been reported to impair motor coordination at analgesic doses.9Although our previous findings10did not show impaired motor coordination after fenobam injection, the discrepant findings between 2-methyl-6-(phenylethynyl)pyridine and fenobam compelled us to perform further assessment of the effects of fenobam on motor coordination. No differences were seen in fenobam- versus  vehicle-injected mice in the ledge crossing test or the vertical pole descent task. Both of these tasks require a mouse to navigate with limited foot purchase and execute complex turning behaviors without falling. A test of strength where mice hung upside down for 2 min also yielded no differences, further suggesting that although mGlu5 antagonism may result in increased locomotive behaviors, motor coordination remains intact. Thus, although fenobam may increase locomotor activity, the potentially more deleterious side effect of altered motor coordination appears to be absent at the tested analgesic dose. Furthermore, the stimulant side effects of fenobam could be beneficially exploited in patients who are also suffering from concurrent depression, assuming that any psychostimulant effects are not aversive. No major adverse reactions were reported by three healthy volunteers administered fenobam at doses of 150 mg.4Although these findings are encouraging they represent results from a limited number of subjects. Further human clinical testing will be necessary to confirm these data.

A major limitation of certain analgesics, such as opiates, is that tolerance to their analgesic effects occurs over time. For mGlu5 antagonists to be useful in humans they would have to avoid this limitation. Seven days of repeated dosing of the mGlu5 antagonist 3-[(2-methyl-1,3-thiazol-4-yl)ethynyl]pyridine has been demonstrated to result in analgesic tolerance in the formalin test.24Whether tolerance develops to the analgesic effects of fenobam is not currently known. Therefore, we injected groups of mice daily for either 5 or 14 days with an analgesic dose of fenobam (30 mg/kg) or vehicle. On the 6th or 15th day, respectively, mice were injected with either fenobam or vehicle, followed by a formalin test. Mice receiving chronic injections of fenobam did not display tolerance to the analgesic effects of acute fenobam on the test day. Fenobam was analgesic in both the first and second phase of the formalin test. Although this is not conclusive evidence that tolerance to fenobam would not develop if the drug was administered over a longer time course, it is nonetheless encouraging evidence that tolerance may not be a limiting factor in the use of fenobam as an analgesic.

A fascinating effect of mGlu5 deletion and pharmacologic inhibition is a significant effect on weight gain and postfast feeding behavior. mGlu5 knockout mice have been previously reported to weigh less than their WT littermates.18,19Differences in ad libitum  food intake were not found,18so this is not believed to account for the differences. However, both mGlu5 knockout mice and mice injected with 2-methyl-6-(phenylethynyl)pyridine exhibit significantly less refeeding behavior after a 14-h fast, suggesting that mGlu5 does play a role in appetite and feeding behavior.18Here we report that fenobam administration significantly decreases postfast refeeding. Fenobam's psychostimulant properties and appetite-suppressive effects should certainly be viewed as separate from its analgesic effects. However, rather than being dose limiting, it might be possible to exploit these off-target effects to beneficial clinical use in certain diseases that have pain as a major symptom. Osteoarthritis is a prime candidate as weight loss has been associated with both a decrease in risk of developing osteoarthritis25and weight loss and increased activity result in an improvement of existing osteoarthritis symptoms.26Again, further testing in human patients will be necessary to know for certain.

Future endeavors to bring mGlu5 antagonists to human patients will require that pharmacologic agents have acceptable safety profiles. This is underscored by the cessation of clinical trials involving the mGlu5 antagonist ADX-10059 due to unacceptable increases in liver enzymes.12Although these changes were thought to be compound-specific,27mGlu5 is expressed in liver tissue,28and thus mGlu5 antagonists as a class could theoretically cause damage to that tissue. Therefore, we sought to determine whether chronic administration of fenobam would also result in unacceptable changes to the liver, as well as other organs. Fenobam administration for up to 14 days resulted in no significant gross lesions in multiple organs and systems, including hepatobiliary, respiratory, digestive, urinary, nervous, or endocrine, when compared with vehicle-injected control subjects, or changes seen by histopathology in the brain, the liver, or the testes. In addition, no major changes were seen in standard blood chemistry measurements or in complete blood counts. Although some animals did demonstrate cardiomyopathy and testicular degeneration, these lesions can occur spontaneously as animals age.29Given that the animal with the significant testicular lesions was treated with vehicle, this is likely to be an incidental finding. An absence of brain lesions and a lack of altered motor coordination suggests an absence of adverse effects in the spinal cord; however, we did not directly verify this finding histologically. Many other lesions seen, including focal adhesions, liver capsule inflammatory infiltrates, and the darkened color of the seminal vesicles, were seen in both vehicle- and fenobam-injected mice and may simply be due to repeated intraperitoneal injections.30 

Finally, although these experiments were designed to test the analgesic tolerance of fenobam and better understand its toxicity profile, it is critical to note that only one dose was tested. This dose was chosen based on previous dose-response studies indicating that 30 mg/kg was an effective analgesic dose in mice.10Furthermore, this dose corresponds to the highest single dose of fenobam so far tested in human trials with different clinical indications.4,14,15Possible toxicity of fenobam at higher doses or administered for longer time periods have not been ruled out. In addition, as mGlu5 has been implicated in learning and memory,11future studies should test the effects of fenobam on these processes.

In conclusion, we report that fenobam, a clinically validated mGlu5 antagonist with analgesic and anxiolytic properties, has an acceptable safety profile in rodents after chronic daily dosing of 30 mg/kg for up to 14 days. In addition, we report that tolerance to the analgesic effects of fenobam does not develop during this time period. Furthermore, although this study and previous studies suggest that fenobam and other mGlu5 antagonists may have a psychostimulant effect, we continue to find no evidence that fenobam impairs motor coordination. Finally, we suggest that although fenobam may have effects outside of the pain neuraxis, at least some of its nonanalgesic effects could be used for beneficial clinical effect. Fenobam, and mGlu5 antagonists in general, represent promising candidates for the treatment of human pain conditions.

The authors acknowledge the work of Sherri Vogt, B.S., Research Technician, and Chang Shen Qiu, M.D., Laboratory Manager, Department of Anesthesiology, Washington University, St. Louis, Missouri, in maintaining our mouse colony. In addition, the authors acknowledge the work of Niecey Hinkle, B.S., Research Technician, Division of Comparative Medicine, Washington University, who worked on the hematology and clinical chemistry analyses. Finally, the authors thank Mena Morales, B.S., PhD Candidate, Program in Neuroscience and Department of Anesthesiology, Washington University, for extensive manuscript editing.