Movatterモバイル変換

Structures of rimonabant and TM38837.

Rimonabant [25] and TM38837 [17,26] were synthesized following previously published literature procedures, and pioglitazone was supplied by RenaSci Consultancy Ltd. The sodium salt of TM38837 was prepared by dissolution of the free acid in methanol, addition of one equivalent sodium hydroxide (from 1 M volumetric standard solution), evaporation to dryness, followed by trituration with isopropyl acetate/heptane (1:1), then filtration and drying of the resulting solid. All test compound batches were$>$98% purity by HPLC.

Dosing was p.o. using the same vehicle (0.1% Tween 80 and 1% hydroxypropyl methyl cellulose (HPMC) (Sigma No: H9262) in distilled water, pH = 7.4) for allin vivo studies according to the following protocol:

TM38837 (sodium salt) or rimonabant were ground using a pestle and mortar, then, the vehicle was added gradually, and the mixtures sonicated for 10 min. Slightly viscous solutions (TM38837) or white granular suspensions (rimonabant) were obtained, which were adequate for p.o. dosing. Suspensions with a pH above 8.0 were buffered with a few drops of 1 M HCl to produce a pH in the range 7.0–8.0. All drugs were administered using a minimal dose volume in the range of 1–3 mL/kg and formulations were prepared freshly within 2 h of each administration. Preparations were stored at room temperature and manually agitated immediately before each administration. All drug doses are expressed as the free base.

The use of this sodium salt suspension formulation enabled the administration of doses exceeding 10 mg/kg; however, exposure may be lower than for solution formulations with free acid and a solubilizing agent. Therefore, a preliminary PK study was run in lean Han–Wistar rats at p.o. doses of 5 mg/kg and 50 mg/kg (both n = 3) to compare exposure with previous studies in mice, for which solution formulations were used, and to select doses for the current study.

Female Han–Wistar rats for the PK and DIO studies were obtained from Charles River, Margate, Kent. Sprague Dawley rats for the cFOS induction study were obtained from Taconic, Denmark. Lean and obese animals for both studies in Zucker rats were obtained from Charles River, France.

For the rat DIO study, the exact statistical tests employed were dependent on the data obtained. The effects of chronic administration of drugs on body weight and daily food intake were compared using one-way analysis of covariance (ANCOVA), with baseline established as the covariate, followed by appropriate multiple comparisons tests (two-tailed) to compare the effects of each treatment group with control. ANCOVA, as a variation of analysis of variance (ANOVA), is used to determine differences between results from three or more unrelated samples or groups. but ANCOVA also considers the effects of other variables, in this case body weights and food intake of the test groups at baseline (i.e., before drug treatment). The advantage of using ANCOVA is that it improves precision and the likelihood of obtaining a significant finding by adjusting the results to account for differences in the groups at baseline. However, whilst ANCOVA (or ANOVA) will determine whether statistical differences exist between test groups that are not attributable to random chance, it will not indicate specifically which groups these differences exist between. Therefore, positive results were followed by post-hoc multiple two-tailed comparison tests, which identified differences between group means, without specifying the direction of the difference. A multiplet-test was sufficient when comparing between rimonabant- and vehicle-treated groups, whereas William’s test, which compares several treatment levels with a zero control in a one-factorial design, was used for comparing TM38837- and vehicle-treated groups.

The effects of drugs on plasma levels of various metabolic parameters were analyzed using ANCOVA, with body weight at baseline (Day 1) as covariate and, if appropriate, bleeding times and/or bleeding group as factors. These analyses examined and accounted for any potential differences between the three animals per group dosed again on Day 34 (for PK) and the remaining seven animals in each group, which were not dosed on Day 34. Plasma parameters were initially analyzed to assess the distribution of the data, as the statistical methods assumed that the data would be normally distributed with equal variance across the groups. If appropriate, log transformations and/or robust regression techniques were used to reduce the influence of outliers. Fat pad weights (retroperitoneal, inguinal, and combined (g)) were expressed as a % of the final body weight using the statistical tests described above.

Adjustments for variability in drug exposure, due to the use of suspension formulations, were not feasible for the analysis of any parameter, as plasma samples were only collected from test group animals for PK analysis at the end of the study. Thus, this may have been an important confounding factor that reduced the statistical significance of differences in various parameters between drug- and vehicle-treated groups in this and subsequent studies. However, based upon the performed PK studies, variability at the lower doses used in the DIO rat study was expected to be low and likely minimized by using freshly prepared formulations each day. It was also assumed that, for chronic studies, random order dosing of animals in each group would lead to minimal variation in the total dose received by each animal during the full course of each study.

For the brain cFOS study, statistical evaluation of the data was with ANOVA followed by appropriate post-hoc analysis (Tukey’s test) between control and treatment groups in cases where statistical significance was established (p 0.05).

For the studies in obese Zucker rats, body weight and weekly and overall body weight gain (g) were analyzed by two-way ANCOVA with treatment and cohort as factors and Day 1 body weight as a covariate. Meanwhile, food and water intake (g) were analyzed by ANCOVA with average daily food or water intake during the baseline phase (Days –6 to 0) as a covariate. Average food and water intake were analyzed using the same methods. Since the lean animals would be expected to have different baseline body weights and food and water intakes, this group was excluded from the analysis. For the steatosis study, the pair-fed groups were compared to the respective drug treated groups using multiplet-tests. For the glucose control study, plasma, insulin, and glycerol data were evaluated by a statistician, and a log transformation was used to calculate the area under the curve (AUC) when the data were not normally distributed. Analysis was by robust regression, which included the treatment and cohort as factors, along with Day 1 body weight, bleeding order, and Day 5 insulin as covariates. Statistical analysis was performed for each time point, and the AUC was calculated using the trapezoidal rule as total AUC and AUC above baseline (AUCB2) (N.B., robust regression is a broader approach to regression analysis than ANCOVA, which aims to minimize the impact of outliers and other violations of regression assumptions). Subsequent multiple comparisons against the vehicle group were by Williams’ test for the TM38837 groups and multiplet-test for the other groups. All tests were carried out as two-sided tests. The chance of a false positive was 5% for each compound for each time point. Data from all analyses are expressed as treatment group means$\pm$ standard error of the mean (SEM), and significant differences from the control groups are denoted by *p 0.05, **p 0.01, and ***p 0.001 (withp 0.05 considered the accepted level for statistical significance).

This study investigated the effects of chronic TM38837 treatment on body weight, food intake, fat pad weights, blood biochemical markers, and liver weights and histology in DIO rats, using rimonabant as comparator. Doses of the test drugs and rimonabant were based on those used in previous animal studies and were not anticipated to produce any adverse effects in the obese rats [16,17]. All dosing occurred at the onset of the 8 h dark period to maximize the impact of any inhibitory effects of the test drugs on food intake, as rats are nocturnal and the levels of food intake of the vehicle-treated control animals were expected to be highest during the dark period.

Female Han–Wistar rats (starting weight = 250–300 g, one set of 60 rats and 8 spares) were housed in pairs in polypropylene cages with solid floors and sawdust bedding. The cages were maintained at a temperature of 21$\pm$ 4 °C and 55$\pm$ 20% humidity, with the room on a reverse-phase light–dark cycle (lights off from 09:30 to 17:30 h), during which time, illumination was by red light. The animals had free access to a powdered high-fat diet (VRF1 plus 20% lard), ground chocolate, and ground peanuts (with continuous access to tap water) for 12 weeks to induce obesity, and were subsequently maintained on this diet throughout the study. The three components of the diet were contained in separate glass feeding jars with aluminum lids (Solmedia Laboratory Suppliers, Romford, Essex), each of which had a 3–4 cm hole to allow access to the food. After 12 weeks, the animals were individually housed in polypropylene cages with wire grid floors, allowing for the recording of the food intake for each rat. In each cage, there was a small amount of paper bedding for warmth, environmental enrichment, and to enable the animals to move off the wire grid floor. Beneath each cage, polypropylene trays with cage pads were placed to detect any food spillage. Animals were accustomed to these new conditions (individual housing and the wire grid floor cages with paper bedding) for approximately two further weeks, then, 65 animals underwent a 7-day baseline run-in period during which the vehicle was administered once daily by oral gavage. Toward the end of the baseline period, the animals were allocated into 6 weight-matched groups of 10 rats and thereafter dosed orally once daily with vehicle, rimonabant (10 mg/kg), or TM38837 sodium salt (1, 3, and 10 mg/kg) as suspensions. Spare animals remaining at the start of drug treatment on Day 1 (up to eight rats in total) were maintained for use in PK studies (these animals were housed under the same conditions as the feeding study). All dosing occurred at the onset of the 8 h dark period.

Rats, feeding jars, and water bottles were weighed (to the nearest 0.1 g) every day at the time of administration of vehicle or test drug. At each reading, the tray below each cage was examined for spilt food, which was returned to the appropriate jar before it was weighed. However, spillage of food from the feeding jars would normally be negligible. During the baseline and drug treatment period, all animals were examined daily at 0 h and at the end of the dosing period. Any overt behavioral physiological effects, or other relevant observations regarding the condition of the animals, were recorded manually. On the first day of drug administration (Day 1), food and water intake were also measured 2 h and 6 h after dosing at 0 h (any unusual behavioral or physiological effects observed during these readings were recorded manually). Variations in body weight and energy levels of the different types of food were accounted for by expressing the food intake results in terms of kJ/kg rat weight.

Dosing was continued for 34 days, whereafter all animals, jars, and bottles were weighed again at the onset of the dark period to provide the final readings for the feeding study. Subsequently, three animals from each group were dosed again then killed 3 h later by rising CO₂, followed by cervical dislocation (Schedule 1 method). and terminal blood samples (ca. 8 mL per rat) were collected from these animals by cardiac puncture into EDTA-coated tubes (no protease inhibitors). Plasma was separated by centrifugation, divided into aliquots for the PK and plasma analyses, and then frozen. Additionally, various tissues were taken from these animals, including brain, liver, lungs, heart, two samples of white adipose tissue (WAT), and a muscle sample (quadriceps). These tissues were dissected out, weighed, rinsed with saline, blotted dry, and then frozen in liquid nitrogen. Likewise, 3 of the spare obese animals were dosed with TM38837 (10 mg/kg) and killed 3 h later, with collection of blood and tissue samples to provide Day 1 PK and tissue samples. Retroperitoneal (which normally covers an area from the groin, encases the kidneys and extends right up to the diaphragm on the dorsal surface of the peritoneal cavity) and inguinal (lower groin area sometimes extending to cover the whole of the surface of the gastrointestinal surface on the posterior side) fat pads were used because they are distinct, therefore, any drug-induced changes in the size of the fat pad can be easily detected. All plasma and tissue samples were stored at –80 °C prior to analysis.

The remaining 7 animals per group were not dosed again and killed after the final readings from the feeding study (i.e., approximately 24 h after the last dose) were recorded. Additionally, two lean Han–Wistar rats (which had been kept on normal chow for 8 weeks prior to the end of the study) were killed. Terminal plasma and tissue samples were collected from all animals as described above. Each liver from these rats was weighed and divided in two, with one half stained as formalin-fixed liver sections with H&E and the other with Sudan Black. An independent pathologist examined the sections for liver fat (unbiased, i.e., without knowing the experimental setup). The presence of fat in the liver cells was subjectively evaluated based on morphological evidence of rounded vacuoles in the cytoplasm.

Plasma levels of glucose, insulin, leptin, glycerol, triacylglycerol, non-esterified fatty acids (NEFAs), total cholesterol, adiponectin, and liver enzymes alanine aminotransferase (ALAT) and aspartate aminotransferase (ASAT) were determined in all samples taken at the end of the study using commercially available kits and reagents following the procedures described in the literature [16]. Additionally, total triglyceride, total cholesterol, and high-density lipoprotein cholesterol (HDL-C) were measured using a Vitros DT250 automatic analyzer with dedicated slides, and low-density lipoprotein cholesterol (LDL-C) was calculated using the Friedewald approximation:

LDL-C = Total cholesterol – HDL-C – VLDL-C

where very low-density lipoprotein cholesterol (VLDL-C) is estimated as 0.456$\times$ total triglyceride concentration expressed in mmol/L.

A total of 40 male Sprague Dawley rats (eight weeks of age, 250–280 g) were used in this experiment. Animals were acclimatized for one week. Prior to and throughout the study, the rats hadad libitumaccess to rodent chow (Altromin 1234) and tap water. The animal room environment was controlled (targeted ranges: temperature, 22$\pm$ 2 °C; relative humidity, 50$\pm$ 10%; light/dark cycle: 12 h light–12 h dark, with lights on from 07:00 to 21:00). For 3 days prior to the experimental dosing, all animals were handled daily and mock gavaged to accustom them to the experimental procedures. At Day 1, rats were randomized according to body weight to participate in one of five groups (n = 8), i.e., vehicle, rimonabant (10 mg/kg), TM38837 (10 mg/kg), TM38837 (30 mg/kg), or TM38837 (100 mg/kg).

Dosing was conducted at the beginning of the light phase and was randomized (3 rats injected at 15 min intervals). The compounds or vehicle were administered orally at timepoint t = 0 min by gavage (administration volume: 2 mL/kg) via a gastric tube connected to a syringe, thereby ensuring accurate dosing. At the designated timepoint (t = 150 min) following dosing, the animals were anaesthetized with a mixture of fentanyl and fluanisone (Hypnorm®) with midazolam (1.0 mL/kg) and transcardially perfused with heparinized saline (15,000 IU/L) followed by 10 min (30 mL/min) with 4% paraformaldehyde (4% paraformaldehyde in 0.1 M PBS pH 7.4). This method of euthanasia is widely used when it is necessary to clear blood and preserve brain tissue for immunostaining orin situ hybridization [27]. The brains were removed and postfixed overnight in 4% paraformaldehyde until histological processing. At the time of termination, a blood sample was collected into lithium-heparin tubes from 3 animals per group. Plasma was separated by centrifugation (4800$\times$g for 15 min) at 4 °C and divided into two aliquots, which were frozen on dry ice as rapidly as possible and stored at –80 °C prior to analysis.

Before sectioning, the brains were transferred for 3 days to a 30% sucrose solution for cryoprotection. The brains were then divided into forebrain and hindbrain by a dorsoventral cut just rostral to the cerebellum. One-in-six series of 40 µm-thick frontal sections from the forebrain and hindbrain were cut on a freezing microtome and collected in potassium PBS (KPBS) (50 mM). One series of sections from each animal was immunoreacted for cFOS according to the following procedure: The sections were washed in KPBS for 3$\times$ 10 min, followed by 10 min of incubation in 1% H₂O₂ in KPBS. After a 20-minute incubation in 5% swine serum in KPBS containing 0.3% Triton X-100 (TX) and 1.0% BSA, the sections were incubated in rabbit anti-cFOS antibody (#94012; diluted 1:4000 in KPBS with 0.3% TX and 1% BSA) at 4 °C overnight [28]. The next day, the sections were washed for 3$\times$ 10 min in KPBS with 0.1% TX (KPBS-T) before incubation in a biotinylated donkey anti-goat antibody (Jackson Immuno Research Lab., Inc., RRID: AB_10633977) diluted 1:2000 in KPBS-T at room temperature for 60 min. After another rinse for 3$\times$ 10 min in KPBS-T, followed by 60 min of incubation in ABC-streptavidin horseradish peroxidase (DAKO cytomation), the sections were washed successively for 10 min each in KPBS-BT, KPBS, and Tris–HCl buffer (0.05 M, pH 7.6). The sections were assessed for peroxidase activity using diaminobenzidine tetrahydrochloride (DAB) as a chromogen (0.04% DAB and 0.003% H₂O₂ in Tris–HCl buffer). Sections were washed for 2$\times$ 10 min in ion-exchanged water, mounted on gelatinized slides, and embedded in Depex® cryoprotected in 30% sucrose and then cut into 40 µm-thick free-floating sections on a freezing microtome. One series of sections from each animal was immunoreacted against cFOS, and labeled cells were qualitatively assessed throughout the brain. Nine areas were selected for quantitative analyses based on the objectives of the study and a previous qualitative analysis. Areas showing a slight effect of TM38837 in the qualitative analysis were included in the quantitative analysis. The quantitative assessments were performed using a computerized image analysis system (Image Pro Plus 4.5 by Media Cybernetics, Rockville, MD, USA).

This study determined the effects of chronic TM38837 administration (56 days) on steatohepatitis in obese Zucker (fa/fa) rats, with rimonabant as comparator. The effects on body weight as well as food and water intake were also determined. The study design (e.g., animals, duration of study, dose of rimonabant, etc.) was based upon published procedures demonstrating reversal of hepatic steatosis in obese Zucker rats after treatment with a high dose (30 mg/kg) of rimonabant [23].

A total of 80 male obese Zucker rats (fa/fa) and 13 age-matched lean Zucker rats (6 weeks of age) were individually housed in polypropylene cages with free access to normal rat chow (Harlan Teklad Global Diet) and tap water at all times. All animals were maintained at 21$\pm$ 4 °C and 55$\pm$ 20% humidity on a 12-hour light/12-hour dark cycle (lights off from 16:00 to 04:00 during which time the room was illuminated by red light). After exposure for 4 weeks on the normal chow diet, at 10 weeks of age, all animals underwent a 7-day baseline run-in period during which time they were dosed with vehicle once daily by oral gavage. Towards the end of the baseline treatment, animals were weighed and allocated into 6 weight-matched treatment groups (n = 12) plus the lean control group (n = 12). At this stage, 8 obese and one lean animal were set aside. Thereafter, obese rats were orally dosed once daily with vehicle, rimonabant (30 mg/kg), or TM38837 sodium salt (3 and 10 mg/kg), according to their group, and the lean rats were dosed with vehicle. Additionally, two treatment groups were pair-fed to either the rimonabant or TM38837 (10 mg/kg) group. Where appropriate, animals were paired with counterparts of similar body weight and food intake. The pair-fed animals were one day out of phase with their drug-treated counterparts. These animals were dosed daily with vehicle and given a daily food allotment equal to that consumed by a drug-treated counterpart over the previous 24 h period. When a pair-fed animal did not consume all the food allocated to it, the food left in the hopper was weighed and replaced with that day’s quantity. All dosing was by oral gavage and occurred at the onset of the 12-h dark period, at approximately 14:00. Food intake, water intake and body weight were recorded daily for all animals. All animals were examined at the beginning and end of each dosing period and any overt behavioural/physiological or any other relevant observation regarding the condition of the animals recorded manually.

Dosing was continued for 8 weeks, whereafter animals were killed by a Schedule 1 method: 24 h after the last dose (drug and vehicle-treated animals on Day 57, and pair-fed animals on Day 58). Blood samples (approximately 140 µL) were collected into lithium heparin tubes from the tail vein of each animal immediately before it was killed. The samples were then centrifuged, and the plasma stored at –80 °C prior to analysis. Livers were dissected out, weighed, and inspected, with any gross pathology being recorded. Livers were dissected and samples, taken from the left lateral lobe and the right medial lobe, were bisected and each half either formalin fixed prior to H&E staining for histopathological analysis or cryopreserved for analysis of fat content by Oil Red staining. The slides were analyzed in a Leica DMRXe microscope equipped with a PC-based image analysis system. With this equipment, the positively stained Oil red was measured and expressed as percentage of positive stained area in relation to total analyzed liver area ($\times$100). The calculation was made from ca. 10 image fields (5 from each section level) for the 10$\times$ objective. For the H&E-stained slides, the presence of rounded intracellular vacuoles was considered as a sign of steatosis, and this was graded from mild (1+) to severe (3+) for each sample by a blinded pathologist.

This study determined the effects of chronic TM38837 administration (30 days) on glucose control in obese Zucker (fa/fa) rats, with rimonabant and pioglitazone as comparators. In this study, the high dose of rimonabant (30 mg/kg) was replicated from the previous study (section 2.7); however, higher doses of TM38837 (10 and 30 mg/kg) were used for a more direct comparison.

A total of 62 male obese Zucker rats (fa/fa) were individually housed in polypropylene cages with free access to normal rat chow (Harlan Teklad Global Diet) and tap water at all times. All animals were maintained at 21$\pm$ 4 °C and 55$\pm$ 20% humidity on a 12-hour light/12-hour dark cycle (lights off from 19:00 to 07:00 during which time the room was illuminated by red light). After exposure for 3 weeks on the normal chow diet, at 9 weeks of age, all animals underwent a 7-day handling period during which they were handled briefly as if to be dosed but not actually dosed. Subsequently, a 7-day baseline run-in period commenced, during which the animals were administered vehicle by oral gavage once daily, and food and water intake were recorded. On the afternoon of the first baseline day (Day -6), the animals were weighed and fasted at approximately 16:00. The following morning, a blood sample (120 µL) was collected into lithium-heparinized tubes (Sarstedt Microvette CB300 LH) from a tail vein bleed immediately prior to vehicle administration, and plasma was separated by centrifugation. Thereafter, food was presented again. Plasma samples were stored at –80 °C and subsequently analyzed for glucose, insulin, and glycerol using commercially available kits and reagents.

Toward the end of the baseline treatment, animals were weighed and allocated by a statistician into 7 weight-matched treatment groups (n = 8 or 9 animals per group), primarily based on baseline body weight and plasma levels of insulin, glycerol, and glucose. If deemed appropriate, food and water intake were also used. At this stage, two spare animals were set aside. Thereafter, obese rats were orally administered vehicle, TM38837 sodium salt (10 and 30 mg/kg), rimonabant (30 mg/kg), or pioglitazone (12 mg/kg) once daily according to their respective group. Additionally, two treatment groups were pair-fed to either the rimonabant group or the TM38837 (10 mg/kg) group as described for the previous study (section 2.7). In contrast to the previous study, for practical reasons, all dosing occurred at the onset of the 12 h light period. Due to the large number of animals in the study, the animals were divided into two cohorts and tested one day apart. Food intake, water intake, and body weight were recorded daily for all animals during the baseline and treatment periods.

Oral glucose tolerance tests (OGTTs) were performed on Days 1 and 29 after an overnight fast, starting at approximately 16:00. On the morning of the test, all animals were cannulated (lateral tail vein) before the start of the OGTT. Each animal was then dosed with either the vehicle or the test compound, and 120 min later, D-glucose (2 g/kg, p.o.). Baseline blood samples were taken immediately prior to compound administration (B1, 100 µL) and immediately before the glucose load (B2, 100 µL), then further blood samples were then taken at 15, 30, 45, 60, 120, and 240 min post-glucose administration. Plasma samples were analyzed for glucose, insulin, and glycerol levels to enable calculation of AUC and AUCB2 over the 0–240 period for each parameter. All blood samples were collected from the tail vein via the cannula and treated as for the baseline blood samples. Although pair-fed groups were one day out of phase with their drug-treated counterparts, all animals were bled on the same day. On Day 30, all animals were dosed, and the weights of food and water bottles recorded. Dosing was timed to enable the animals to be killed on Day 31 by a Schedule 1 method, 24 h after the final dose. Terminal blood samples were collected and treated as described in the previous study.

PK curves for TM38837 in Han–Wistar rat (i.v. and p.o.) are shown in Fig.2 (individual animals), and parameters are presented in Table1, including C57BL/6J mouse (p.o.) for comparison (used in previous DIO studies [16]). Good bioavailability was observed with the sodium salt suspension formulation, and scaling between the 5 mg/kg and 50 mg/kg was acceptable. Based upon comparison with the mouse data, it was concluded that a 10 mg/kg dose in rats would adequately match the 10 mg/kg dose previously used in mice, and this was selected as the top dose for the current study. It should be noted that these studies were performed in lean animals, whereas, as previously observed, it might be expected that the half-life and AUC would be greater in obese animals.

Fig. 2.

PK curves for TM38837 in Han–Wistar rat. (a) TM38837 sodium salt 1 mg/kg i.v. (b) TM38837 sodium salt 5 mg/kg (R4-6) and 50 mg/kg (R7-9) p.o. 4. PK, Pharmacokinetics.

Daily dosing with 1, 3, and 10 mg/kg TM38837 p.o. for 5 weeks produced significant dose-dependent decreases in body weight, when compared to the vehicle-treated controls (Fig.3a), whilst rimonabant (10 mg/kg p.o.) reduced body weight by a similar amount as the 10 mg/kg dose of TM38837. Notably, TM38837 treatment produced a smaller initial decrease in food intake than for rimonabant (Fig.3b) with differences in daily food intake between vehicle, rimonabant, and TM38837-treated rats becoming not significant by the end of the study. The observed reduction in body weight after chronic treatment with TM38837 and rimonabant was mirrored by a reduction in fat pad weights (Fig.3c), whilst there was no significant difference between treatment groups in liver weights relative to total body weights (Fig.3d).

Fig. 3.

Effects of chronic TM38837 and rimonabant administration in DIO rats. (a) Effect on body weight. Values are means (adjusted for differences between body weights of the different treatment groups at baseline (Day 1))$\pm$ SEM (calculated from the residuals of the statistical model), n = 10. Numbers represent % body weight change compared to the vehicle-treated control group on Day 34. Data were analyzed by ANCOVA with Day 1 body weight as a covariate and multiple comparisons against the control group by Williams’ test for TM38837 and multiplet-test for rimonabant (*p 0.05). (b) Effects on daily food intake. Values are means (adjusted for differences between daily food intakes of the different treatment groups at baseline (average of Days –6 to 0))$\pm$ SEM (calculated from the residuals of the statistical model), n = 10. Data were analyzed by one-way ANCOVA with Day 1 body weight as a covariate and multiple comparisons against the vehicle-treated control group by Williams’ test for TM38837 and multiplet-test for rimonabant (*p 0.05, **p 0.01, and ***p 0.001). (c) Effects on inguinal, retroperitoneal fat pad weights. Results are means (adjusted for differences between the body weights of the different treatment groups at baseline (Day 1)) + SEM (calculated from the residuals of the statistical model), n = 10. Data were analyzed by ANCOVA with treatment and PK sub-groups as factors and Day 1 body weight as a covariate. Multiple comparisons against the vehicle-treated control group were by Williams’ test for TM38837 and multiplet-test for rimonabant (*p 0.05, **p 0.01, and ***p 0.001). (d) Relative liver weights. Expressed as a % of final body weight. Results are means$\pm$ SEM (n = 7, except lean for which n = 2).

From the examination of the H&E-stained liver sections (Fig.4a), as expected, a higher content of fat was seen in the livers from obese vehicle-treated rats than in the livers from 2 lean control rats, although no fat was observed in sections from two obese vehicle-treated rats. A tendency towards reduced liver fat was evident when comparing TM38837-treated rats with those treated with vehicle or rimonabant, although the difference was not statistically significant. However, liver sections stained with Sudan Black (Fig.4b), for the identification of lipids in parenchymal cells, had very little consistency with the results obtained by H&E grading, with high levels of fat even observed in the two lean rats.

Fig. 4.

Liver histology: grading of fat content from H&E and Sudan Black staining. Formalin-fixed sections from livers, collected 24 h after the last dose, were stained with (a) H&E or (b) Sudan Black and graded for liver fat content by a blinded pathologist. Grading was performed on two sections from each liver, and the result with the highest fat content recorded in each case.

From the plasma parameter analysis (Fig.5), a dose-dependent reduction in plasma leptin was observed for TM38837-treated animals, but no other statistically significant effects were observed, although TM38837 tended to decrease insulin and glycerol plasma levels at the high dose. No sign of liver toxicity was observed for rats treated with TM38837, as indicated by a lack of increase in the liver injury markers ALT and AST.

Fig. 5.

Effect of chronic TM38837 and rimonabant administration on various plasma parameters in DIO rats. Results are means (adjusted for differences between the treatment groups in body weight at baseline (Day 1) and bleeding order$\pm$ SEM (calculated from the residuals of the statistical model). Data were analyzed by one-way ANCOVA with body weights at baseline (Day 1) as covariate and treatment, bleeding times, and bleeding group as factors; comparisons to vehicle were by Williams’ test for TM38837 and multiplet-test for rimonabant (*p 0.05, ***p 0.001).

Plasma analysis indicated higher exposure of TM38837 than rimonabant on Day 34 whilst tissue homogenate analysis for TM38837 showed very low brain and high liver exposure (approximately 1% and 300% of plasma concentration, respectively) on Day 1 (Fig.6). There was no evidence for accumulation with similar plasma and brain levels observed between the first and last days of the study.

Fig. 6.

Plasma and tissue exposure of TM38837 and rimonabant in the DIO rat study. N.B. Plasma and tissue samples were collected 3 h post-dose on the indicated days.

cFOS immunopositive cell counts following drug or vehicle treatment are presented graphically in Fig.7a. Consistent with previously published findings, rimonabant was shown to induce cFOS in all areas investigated [29]. Conversely, a similar response was not observed for TM38837. The therapeutically effective dose of TM38837 in rats (10 mg/kg) and a 3-fold higher dose (30 mg/kg) did not lead to significantly increased numbers of cFOS-positive nuclei in any of the areas investigated. However, the highest dose of TM38837 (100 mg/kg) significantly increased cFOS in a manner comparable to rimonabant in some of the areas (i.e., the prefrontal cortex, nucleus accumbens shell, the anterior part of the bed nucleus of stria terminalis as well as in the caudal part of the solitary tract nucleus). Drug plasma concentrations (Fig.7b) for TM38837 (10 and 30 mg/kg) were roughly consistent with expected C_max values based upon preliminary PK studies; but were more variable for TM38837 (100 mg/kg), potentially due to delayed or variable absorption. Overall, this study indicated a lack of neuronal activation in the CNS induced by therapeutic doses of TM38837 and a different pattern at high doses compared with rimonabant.

Fig. 7.

Effects of different TM38837 doses and rimonabanton cFOS counts within selected brain regions. (a) cFOS nuclei. Statistical analysis: one-way ANOVA, Tukey’s multiple comparison test (p 0.05 * vs. vehicle; # vs. TM38837 10 mg/kg; ¤ vs TM38837 30 mg/kg; ^ vs TM38837 100 mg/kg). (b) Drug plasma exposure at termination: individual values.

As expected, obese vehicle-treated Zucker rats gained markedly more weight and consumed more food over the 8-week treatment period than the lean Zucker rats (Fig.8a). Meanwhile, obese rats treated with TM38837 (3 and 10 mg/kg) demonstrated a significantly lower total weight gain by the end of the study than the obese vehicle-treated rats. In contrast, no significant differences in total weight gain were seen between obese control rats and the pair-fed group to TM38837 (10 mg/kg), indicating a food intake independent effect of TM38837 on weight gain. Treatment with the high dose of rimonabant (30 mg/kg) induced a more marked reduction in total weight gain than for the TM38837-treated animals, and a much more prominent decrease in food intake during the initial week of the study. TM38837 treatment reduced food intake throughout the course of the study, but the effect was only significant in the first week (Fig.8b). In contrast to previous DIO rodent studies, the food intake of rimonabant-treated Zucker rats was significantly lower than that of obese vehicle-treated rats during almost the entire study. However, the dose of rimonabant was much higher than that used in the DIO studies (30 mg/kg vs. 10 mg/kg, respectively). As was observed for TM38837, the group pair-fed to rimonabant gained more weight than the rimonabant-treated group.

Fig. 8.

Effects of 8 weeks of TM38837 and rimonabant treatments on body weight gain, insulin levels, and hepatic steatosis in obese Zucker rats. (a) Body weight gain. (b) Average food intake. For (a,b), values for obese animals are adjusted mean$\pm$ SEM (calculated from the residuals of the statistical model), and for lean animals, are simple mean$\pm$ SEM (n = 11–12). Data were analyzed by ANCOVA with baseline food intake (Days –6 to 0) as covariate. Comparisons with obese vehicle were by William’s test for TM38837 groups and multiplet-test for the other groups. Comparisons of rimonabant and TM38837 to their appropriate pair-fed groups were by multiplet-test (†††p 0.001). (c) Hepatic steatosis measured by Oil Red staining: Positively stained Oil Red was measured and expressed as a percentage of the positively stained area in relation to the total analyzed liver area ($\times$100). Student’s t-test: *p 0.05; **p 0.01; ***p 0.001; when compared to obese vehicle controls; §§,p 0.01; §§§,p 0.001: when compared to drug-treated counterparts. Individual values and the mean are also presented.

The results of the Oil Red staining are presented as a bar chart and as individual values in Fig.8c. As expected, the lean rats had normal livers with no observed steatosis, whereas the obese vehicle-treated rats had developed hepatic steatosis to various degrees, including three that had no steatosis. A reduction in the stained area was evident for TM38837-treated rats compared to vehicle-treated rats, although this difference was not statistically significant. However, the groups pair-fed to TM38837 and rimonabant had both developed a significantly higher degree of steatosis than their drug-treated counterparts.

Plasma marker analysis at the end of the study demonstrated a significant decrease in hyperinsulinemia in obese Zucker rats treated with TM38837 (10 mg/kg), which was not observed in animals pair-fed to TM38837 (10 mg/kg) (Fig.9). No other statistically significant effects were found for TM38837, although the plasma glycerol and ALT levels were significantly lower than those of the pair-fed rats. The reported reduction in plasma triglycerides by rimonabant treatment at 30 mg/kg was reproduced in this study, but not the published decrease in plasma ALT and TNF-alpha [23]. However, it is worth noting that the obese control rats did not exhibit higher plasma TNF-alpha levels compared to the lean rats in this study.

Fig. 9.

Plasma parameters in obese Zucker rats following 8 weeks of TM38837 and rimonabant treatments. Values are simple mean$\pm$ SEM. Data were analyzed by Student’st-test: *p 0.05; **p 0.01; ***p 0.001 vs. Vehicle §:p 0.05; §§:p 0.01; §§§:p 0.001 vs. the respective pair-fed counterpart.

No significant differences were found between groups for the area under the curve above baseline (AUCB2) of plasma glucose, insulin or glycerol after acute treatment with vehicle or test compounds, except for the high TM38837 dose (30 mg/kg) which demonstrated a significant reduction in plasma AUCB2_Insulin (Fig.10a). However, whilst no significant differences in glucose or glycerol AUCB2 were found between rats chronically treated with TM38837 and vehicle, TM38837 (10 and 30 mg/kg) and rimonabant both induced highly significant reductions in AUCB2_Insulin compared with vehicle treatment (Fig.10b). As in previous experiments, both TM38837 and rimonabant-treated animals showed highly significant reductions in weight gain that, unlike in the previous study, were replicated in the pair-fed animals (Fig.10c). In contrast, animals treated with pioglitazone showed a marked increase in weight gain although effects on glucose tolerance and hyperinsulinemia were markedly greater. No significant differences were found between the pair-fed and vehicle control groups with respect to either plasma glucose or insulin levels, indicating that the reduction in food intake (or weight gain) was not the main reason for changes to insulin levels following chronic TM38837 or rimonabant treatment.

Fig. 10.

Effects of chronic (4 weeks) TM38837 and rimonabant administration on glycemic control in obese Zucker rats. (a) Plasma glucose, glycerol, and insulin AUCB2 (0–240 min) on Day 1. (b) Plasma glucose, glycerol, and insulin AUCB2 (0–240 min) on Day 29. For (a,b), the glucose challenge was 120 min after compound dosing. Values are adjusted mean$\pm$ SEM (calculated from the residuals of the statistical model). A log transformation was used for the AUC calculations. Statistical analysis was by robust regression, including treatment and cohort as factors and Day 1 body weight, bleeding order, and Day 5 measurements as covariates for calculation of AUCB2. Comparisons against the vehicle group were by Williams’ test for TM38837 groups (B and D), and multiplet-test for the other groups (**p 0.01; ***p 0.001). (c) Overall body weight gain. Data were analyzed by ANCOVA, with body weight on Day 1 as covariate and treatment and cohort as factors. Comparisons against the vehicle group were by Williams’ test for TM38837 groups (B and D) and multiplet-test for the other groups (***p 0.001).

For clarity, details, and key findings for allin vivo studies have been summarized in Table2.