N-butyldeoxynojirimycin causes weight loss as a result of appetite suppression in lean and obese mice

Transcription

1 ORIGINAL ARTICLE doi: /j x N-butyldeoxynojirimycin causes weight loss as a result of appetite suppression in lean and obese mice D. A. Priestman, A. C. van der Spoel, T. D. Butters, R. A. Dwek and F. M. Platt Glycobiology Institute, Department of Biochemistry, University of Oxford, Oxford, UK Aim: To determine the mechanism of weight loss caused by high doses of N-butyldeoxynojirimycin (NB-DNJ) in healthy lean and leptin deficient obese (ob/ob) mice. Methods: Healthy lean and obese mice were treated with NB-DNJ by the following methods: admixed with their diet, delivered by subcutaneously implanted mini-pumps or by intraperitoneal or intracerebroventricular (ICV) injection. Daily changes in body weight and food intake were recorded during the experimental period. The effect of NB-DNJ treatment on subcutaneous adipose tissue and on epididymal fat pads was measured. Results: Lean mice treated with NB-DNJ, admixed with their diet, lost weight in the form of adipose tissue. This resulted in a 40% reduction in skin thickness (control, mm; NB-DNJ treated mm) and a reduction in epididymal fat pad weights after 5 weeks of treatment at 2400 mg/kg/day (control, ; NB-DNJ treated, as ratios of fat pad weight to total body weight). Following the depletion of adipose tissue mass, the mice grew normally and did not have any reduction in lean mass. Obese mice treated with NB-DNJ also lost weight or gained weight at a greatly reduced rate compared with non-treated controls. Body weights at 6 months of age were: lean control, g; lean NB-DNJ treated, g; obese control, g; obese NB-DNJ treated from 5 weeks of age, g; obese NB-DNJ treated from 12 weeks of age, g. Both the lean and obese groups of mice treated with NB-DNJ ate up to 30% less than untreated controls. Daily food intake (powder diet) were: lean control, g; obese control, g; lean NB-DNJ treated g; obese NB-DNJ treated, g. Mice treated with the N-substituted galactose imino sugar analogue, N-butyldeoxygalactonojirimycin (NB-DGJ) did not lose weight. Mice experienced similar weight loss or lack of weight gain when fed a restricted diet that mimics the drug-induced level of food consumption. Delivery of 2 nmol NB-DNJ by ICV injection into lean mice also caused similar reductions in food intake. Food intake: saline vehicle, g; NB-DNJ, g; NB-DGJ, g; 2-deoxyglucose, g. Conclusion: NB-DNJ causes weight loss as a result of reduced food consumption due to central appetite suppression. Keywords: appetite suppression, imino sugars, intracerebroventricular injection, N-butyldeoxynojirimycin, obesity Received 13 October 2006; returned for revision 28 November 2006; revised version accepted 28 November 2006 Introduction The imino sugar N-butyldeoxynojirimycin [(NB-DNJ), miglustat, Zavesca, Actelion Pharmaceuticals, Basel, Switzerland] has a variety of inhibitory activities. These include inhibition of the N-linked glycan-processing enzymes a-glucosidases I and II [1,2], the ceramidespecific glucosyltransferase involved in glycosphingolipid biosynthesis [3], the disaccharidases, sucrase and maltase [4] and the glycogen debranching enzyme [5]. Correspondence: Dr David A. Priestman, Department of Pharmacology, University of Oxford, Mansfield Road, Oxford OX1 3QT, UK. david.priestman@pharm.ox.ac.uk # 2007 Blackwell Publishing Ltd Diabetes, Obesity and Metabolism, 10, 2008, j 159

2 OA j N-butyldeoxynojirimycin causes weight loss as a result of appetite suppression D. A. Priestman et al. As an inhibitor of glycosphingolipid biosynthesis, NB- DNJ is presently used in clinical trials as a potential treatment for glycosphingolipid storage diseases with a neurological component, and is approved for use in non-neuronopathic type 1 Gaucher disease in Europe, Israel and the USA [6 9]. There are some side effects associated with the administration of NB-DNJ to both mice and humans. Among these is a significant reduction in weight gain seen in mice on relatively high doses of the drug (1200 mg/kg/day and above [3]) and weight loss in some individuals in clinical trials in humans [9,10]. In contrast to the effect of NB-DNJ, mice treated with the N-substituted galactose imino sugar analogue, N- butyldeoxygalactonojirimycin (NB-DGJ) and also an effective inhibitor of glycosphingolipid biosynthesis, did not lose weight [4]. This precludes the effects on glycosphingolipid biosynthesis being involved in the mechanism of weight loss. The other inhibitory activities mentioned above all have the potential to cause weight loss. For example, in vivo inhibition of disaccharidases in the gastrointestinal tract may result in reduced postprandial absorption of monosaccharides generated enzymically from sucrose and maltose. However, when mice treated with NB-DNJ were switched to a diet rich in glucose (65%) but lacking complex carbohydrate, body weights remained 15% less than controls [11]. Inhibition of glycogen debranching enzyme and thereby glycogenolysis could also result in a reduction in body weight: the inability to maintain blood glucose concentrations via the breakdown of glycogen would result in increased b-oxidation of fatty acids derived from white adipose tissue as an alternative fuel source. Nonetheless, results from a recent study on the inhibition of glycogenolysis by imino sugars in vitro and in vivo showed that N-nonyldeoxynojirimycin is a more potent inhibitor of glycogenolysis than NB-DNJ, yet did not cause weight loss [5]. This implies that inhibition of glycogen breakdown is unlikely to play a significant role in causing the reduced weight of NB-DNJ-treated mice. An additional possibility is that NB-DNJ, which is known to cross the blood brain barrier [12], may be acting centrally as an appetite suppressant, resulting in a decreased caloric intake and thereby causing weight loss. This mechanism of the effects of NB-DNJ to cause extensive glycosphingolipid depletion in mice in vivo was originally suggested in an earlier study [11]. However, until now, possible effects of NB-DNJ on satiety have remained unexplored. In the current study, the primary aim was to determine the mechanism of the reduced growth in mice treated with NB-DNJ. A further aim of the study was to investigate whether the weight loss caused by NB-DNJ extended to obese mice can thus be subsequently evaluated as a possible therapy for obesity. Materials and Methods All experiments were performed in accordance with UK Home Office Animals (Scientific Procedures) Act NB-DNJ was a gift from the Monsanto/Searle Company and Oxford GlycoSciences (Abingdon, Oxfordshire, UK) and NB-DGJ was purchased from Toronto Research Biochemicals (Downsview, Ontario, Canada). 2-Deoxyglucose was purchased from Sigma (Poole, UK). Treatment of Mice with Imino Sugars Control C57BL/6 (lean) and C57BL/6OlaHsd-Lep ob (obese) mice were obtained from Harlan (Bicester, UK) at 5 weeks of age. The mice were singly housed under standard non-sterile conditions and fed a diet of mouse chow pelleted or expanded ground RM1 diet (8.5% protein, 88.5% cereal, 0.5% soya oil, 2.5% supplements; SDS, Essex, UK) with water available ad libitum. For dietary treatment of the mice with NB-DNJ or NB-DGJ, the diet and the compound were admixed thoroughly as dry solids, stored at room temperature and used within 7 days. Unless indicated otherwise, mice were treated with a dose of 2400 mg/kg body weight per day in the diet, assuming an intake of around 5 g diet/day. Mice were maintained on diet with or without compound for up to 6 months of age and were weighed at regular intervals during the course of the experiment (see figure 2 legend for group sizes and further experimental details). To study the effect of compound on obese mice, three groups (n ¼ 5) of obese mice were monitored from 5 weeks of age. One group acted as control and were fed ground RM1 diet. At 6 weeks of age, the second group was started on NB-DNJ in the diet, while the third group started drug treatment at 12 weeks of age. All mice were then monitored for up to 6 months of age. Growth rates of control and NB-DNJ-treated lean mice were analysed statistically by univariate analysis of variance (ANOVA) using SPSS software to compare slopes of growth from 54 to 174 days of age. The effect of NB-DNJ on bone growth was also assessed by measuring tibia lengths in a group of 20 control mice at 1-year old, five mice at 2 years old and two mice 19.5 months old after 18 months on 2400 mg/ kg/day. Tibias were dissected from both hind-limbs and then boiled in distilled water for 20 min to clean the 160 j Diabetes, Obesity and Metabolism, 10, 2008, # 2007 Blackwell Publishing Ltd

3 D. A. Priestman et al. N-butyldeoxynojirimycin causes weight loss as a result of appetite suppression j OA bone of remaining connective tissue and muscle. Tibia length was then measured using calipers. Dietary intake of mice on powdered diets was measured on a daily basis, by preweighing diet put into the feeding hoppers and then weighing the remaining diet daily. The diet that the mice had spilt on the floor of the cage was carefully recovered by sieving and incorporated into the measurements. Skin Histology/thickness Six week old lean mice on a normal diet or treated with 2400 mg/kg/day NB-DNJ for 4 weeks (n ¼ 6 in each group) were killed and the hair removed by shaving followed by application of a proprietary depilatory cream. A small section of the skin was excised from the same area of the back of each mouse and processed for histology. Another portion of the skin (5 mm 2 ) was taken from each mouse, placed on a microscope slide and the skin thickness was measured using a micrometer gauge. Epididymal Fat Pads A dose response relationship for the effect of increasing concentrations of NB-DNJ on adipose tissue in male lean mice was determined by treating with doses ranging from 15 to 2400 mg/kg/day for 5 weeks. Mice were killed and their epididymal fat pads excised and weighed. The results are expressed as a ratio of epididymal pad fat mass to total body weight. Intraperitoneal Injections In order to facilitate measurement of dietary intake, to exclude possible effects of the bitter taste of imino sugar in the diet and also inhibitory effects on disaccharidases in the gastrointestinal tract, groups of lean control and obese mice 5 weeks old were given daily intraperitoneal injections of NB-DNJ at a dose equivalent to 2400 mg/kg/ day. A further group of obese mice was given intraperitoneal injections daily for 2 weeks from 10 weeks old. Lean control mice for each group were injected with an equal volume of isotonic saline vehicle. Dietary Restriction After assessment of dietary intake over 5 weeks in lean controls and 3 weeks in the obese mice, a restricted diet of 3.2 g/day for controls and 4.0 g/day for the obese mice was provided to compare with the effect of intraperitoneal injection. This equated to a 30% reduction in dietary provision for both groups of mice. Intracerebroventricular Injections To investigate a potential central anorexigenic mechanism for weight loss in mice treated with NB-DNJ, single intracerebroventricular (ICV) injections were performed. The mice were assessed for food intake and weight change. Mice were anaesthetized with isofluorane (4%) delivered by a vaporizer. Upon achievement of a surgical plane of anaesthesia, the animal was transferred to and secured in a Kopf small animal ultra precise stereotaxic instrument (David Kopf Instruments, Tujunga, CA, USA) fitted with rat ear bars, mouse adaptor and mouse anaesthesia mask. A small amount of lignocaine local anaesthetic gel was applied to the ear bars to reduce any postoperative discomfort. For the remainder of the procedure, anaesthesia was maintained with 2.5% isofluorane. Lacrilube (Allergan, High Wycombe, UK) was applied to the mouse eyes and a subcutaneous injection of the non-steroidal anti-inflammatory drug, Rimadyl (Pfizer, Kent, UK) was given for postoperative analgesia. For ICV injections, a small scalpel incision was made along the midline of the scalp from above the eyes to the occiput. A small hole was drilled in the skull, 1 mm lateral and 0.34 mm posterior to bregma. A 1 ml solution was infused, over a period of 1 min, into the lateral ventricle at a depth of 2.25 mm, via a glass needle pipette pulled to a thickness of <100 mm at the tip. Mice were injected with 2 nmol compound (NB- DNJ, NB-DGJ or 2-deoxyglucose) in isotonic saline vehicle which was injected into controls. The incision was then closed using VetBond acrylic suture fluid (3M Animal Care Products, St Paul, MN, USA). The whole procedure from induction of anaesthesia to full recovery lasted less than 7 min. Mice were allowed free access to a preweighed quantity of RM1 pellet diet with waterprovided ad libitum. Twenty-four hours following ICV injection, the mice and the remaining diet were weighed. ICV Mini-pump Implants The effect of long-term ICV delivery of imino sugars was determined by the implantation of Alzet mini-osmotic pumps (model 2004; DURECT, Cupertino, CA, USA) together with the brain infusion kit II, which features a high stability low profile cannula. The mini-pumps were loaded with 0.24 ml imino sugar or saline vehicle. The indwelling 28-gauge stainless steel cannula was cut to a length of 2.25 mm. Mice were anaesthetized as for the ICV injections and underwent similar surgical procedures throughout. After drilling a small hole in the cranium at the midline, 2 mm posterior to bregma, the mini-pump was positioned subcutaneously in the animal s flank and the cannula was implanted. This position # 2007 Blackwell Publishing Ltd Diabetes, Obesity and Metabolism, 10, 2008, j 161

4 OA j N-butyldeoxynojirimycin causes weight loss as a result of appetite suppression D. A. Priestman et al. and depth ensured delivery to the third cerebral ventricle. The implant was secured to the cranium using VetBond acrylic suture fluid, which was also used to close the surgical wound. The whole procedure lasted less than 9 min and mice were fully conscious and recovered within 3 min. For concentrations of imino sugar in mini-pumps see legend to figure 5. Results Effect of NB-DNJ on Lean Control Mouse Adipose Tissue In early experiments, gross dissection of NB-DNJ-treated mice showed a significant depletion of white adipose tissue. In addition, after corporal depilation, this group of mice had transparent skin making the visceral organs, ventrally, and the kidneys and spinal vertebrae, dorsally, clearly visible in contrast to the opaque, nontransparent skin of controls. A reduction in subcutaneous white adipose tissue was also clearly evident in histological sections of mouse skin. Sections taken from NB-DNJ-treated mice seemed completely devoid of the fat visible in the subcutaneous layer in the control (results not shown). This resulted in a 40% reduction in mouse skin thickness ( mm, control; mm, NB-DNJ-treated mice). Further results on body fat depletion were obtained in the study on mouse epididymal fat pads (figure 1). This discrete fat mass showed a dose-related response after 5-week treatment with NB-DNJ. Fig. 1 Dose response effect of NB-DNJ on epididymal fat pads. Mice (n ¼ 5 per group) were treated with NB-DNJ admixed with their diet for 5 weeks from 6 weeks old. Epididymal fat pads were then excised and weighed in pairs. Weights are expressed as a ratio to total body weight s.e.m. Data were analysed by one-way ANOVA with a Dunnett multiple comparison test using GRAPHPAD INSTAT software. Effect of Imino Sugars on Mouse Growth Figure 2 shows the effect of both NB-DNJ on the growth of healthy lean control and obese mice from 5 to 6 weeks of age. During the first week of treatment, control mice with NB-DNJ in their diet lost about 10% of their total body mass. Over the course of the following weeks, their weight stabilized. By 5 weeks of treatment, they weighed around 25% less than the lean mice on normal diet. They then started to gain weight. However, after 10 weeks, they remained 20% less in mass when compared with the age- and sex-matched control group. In contrast to the effect of NB-DNJ on growth, a previous study showed that NB-DGJ had no discernible effect and mice followed the same pattern of growth as untreated controls [4]. Effect of NB-DNJ on Obese Mice As NB-DNJ was so effective at reducing white adipose tissue in normal mice, a study was performed to discover Fig. 2 Growth curves of control and obese mice (mean s.e.m.). Mice were maintained on diet with or without NB- DNJ (2400 mg/kg/day), for up to 6 months of age. C57BL/6 mouse group numbers: n ¼ 12 to day 75, n ¼ 6 to day 112, n ¼ 3 to day 190. To study the effect of compound on obese mice, three groups (n ¼ 5) of ob/ob mice were monitored from 5 weeks of age. One group received diet containing NB-DNJ. Two groups were started on control diet and were maintained on this diet up to 12 weeks of age. One of the groups was supplied with NB-DNJ thereafter. 162 j Diabetes, Obesity and Metabolism, 10, 2008, # 2007 Blackwell Publishing Ltd

5 D. A. Priestman et al. N-butyldeoxynojirimycin causes weight loss as a result of appetite suppression j OA if a similar effect would occur in obese mice also. The effect of long-term dosing on obese mice is also shown in figure 2. The control obese mice grew from just over 20 g at 5 weeks of age to about 60 g at 6 months of age, whereas the lean controls had reached about 27 g at the same age. The group of obese mice started on NB-DNJ treatment at 7 weeks old weighed 35 g after 6 months (40% less than their control littermates). A further group of obese mice started on compound at 12 weeks old, lost weight over 3 months, stabilizing their weight at a little less than 40 g. Effect of NB-DNJ on Lean Growth of Control Mice Although lean mice lost weight over the first 2 weeks of treatment, they then started to grow again, and continued to gain weight for the remainder of the study (figure 2). Between the ages of 2 and 6 months, mouse growth approximates a linear course, which can be analysed by least squares regression analysis. This was thus applied to the growth of the lean mice with and without NB-DNJ treatment, excluding the portion of the curve where the treated mice lost weight. Growth in both groups of mice as measured by the gradient of the line was similar and the data were analysed by univariate ANOVA to compare the slopes of growth statistically. SPSS software was used to process the data and resulted in a p-value of for the difference between the slopes of the two growth curves. This result shows that, after the original weight loss, the mice then grow normally, albeit weighing 4 g less than their untreated counterparts. This indicates that the mice are losing most of their adipose tissue mass initially, and that the treatment has little or no effect on the rate of lean growth. To establish further, whether NB-DNJ has an effect on bone growth of mice tibia lengths were measured in a large (n ¼ 20) group of lean control mice at 1 year of age, a smaller group (n ¼ 5) at 2 years of age and a comparison was made with two mice treated with NB-DNJ for 18 months. Tibia lengths in the treated mice were within the normal range (table 1), which indicates that imino sugar treatment had no effect on the lean growth of treated mice. Dietary Intake in Lean and Obese Mice In an earlier study on the effects of NB-DNJ to deplete glycosphingolipids in mouse tissues, it was suggested that the reduction in body weight might be a result of appetite suppression [11]. This hypothesis was tested by measuring dietary intake in both lean and obese mice fed on both powdered chow with or without imino sugar Table 1 Tibia lengths in lean control and NB-DNJ-treated mice Age (months) Weight (g) Tibia length (mm) (range) Control 12 (n ¼ 20) ( ) 24 (n ¼ 5) ( ) NB-DNJ The effect of NB-DNJ on bone growth was assessed by measuring tibia lengths in a group of 20 control mice at 12 months old, five mice at 2 years old and two mice 19.5 months old after 18 months on NB-DNJ (2400 mg/kg/day). See Materials and Methods for further experimental details. Tibia lengths in Control and NB-DNJ-treated mice. and with pelleted diet. To exclude any effects of the bitter taste of imino sugars and also as a means to bypass any disaccharidase inhibition in the gut, lean control and obese mice were given daily intraperitoneal injections of NB-DNJ for 5 and 3 weeks, respectively. In addition, a group of obese mice had NB-DNJ mini-pumps implanted and, after 4 weeks, when the pumps were empty, were then given intraperitoneal injections daily. This delivery route has the advantage that it allows easy assessment of dietary intake, as the mice can eat pelleted diet. The results for dietary intake are presented in figure 3. In all the groups of mice, whether on a pelleted or powdered diet, NB-DNJ treatment caused a significant reduction in food intake of around 30%. Control lean mice ate almost 10% more pellet diet by weight than powdered diet, whereas the obese mice ate 30% more pelleted diet on a daily basis. As a result of the larger intake of pelleted diet, the obese mice grew rather more rapidly than their powder-fed counterparts (results not shown). It is also notable that, though the obese mice fed on powdered chow ate the same quantity as control lean mice, the obese phenotype resulted in the extra weight gain seen in figure 2. A further experiment was done to observe weight gain in lean control and obese mice fed a restricted diet comparable in quantity to that eaten by NB-DNJ-treated animals. The mice were therefore provided with pellet diet weighing 30% less than normal ad libitum pellet intake. The reduced diet intake resulted in growth curves very similar to those for intraperitoneally injected control and obese mice (figure 4). This implies that the caloric restriction as a result of appetite suppression is sufficient to cause the reduced growth/ weight loss seen in drug-treated mice. ICV Injections To confirm a central anorectic effect of NB-DNJ, 2 nmol NB-DNJ, NB-DGJ and 2-deoxyglucose were injected directly # 2007 Blackwell Publishing Ltd Diabetes, Obesity and Metabolism, 10, 2008, j 163

6 OA j N-butyldeoxynojirimycin causes weight loss as a result of appetite suppression D. A. Priestman et al. Fig. 4 Growth curves for mice injected intraperitoneally with saline/nb-dnj or with a restricted diet (mean s.e.m., n ¼ 5 per group). At 11 weeks of age, one group of obese mice were given daily intraperitoneal injections of NB-DNJ (2400 mg/kg/day) for a 2-week period. the short term [13]. ICV mini-pump implants that were loaded with 500 mm NB-DGJ had no effect on either dietary intake or weight change (figure 5). This figure also shows the significant weight loss, caused by the reduction in food intake in the 24 h following ICV injection of NB-DNJ. Discussion Fig. 3 Dietary intake in lean control and obese mice (mean s.d.). Dietary intake was assessed daily over 7 days in each group of mice (n ¼ 5 per group, singly housed). i.p., intraperitoneal injection. See Materials and Methods for further experimental details. into the lateral cerebral ventricle of lean control mouse brain (approximately 100 mm final cerebroventricular concentration). Mice injected with both saline vehicle and NB-DGJ ate a normal quantity of pellet diet in the following 24 h, approximately 5% less in this experiment than that seen in the lean control mice in figure 3 (not statistically significant). In contrast, NB-DNJ-injected mice ate 20% less (p < 0.02 for both saline and NB-DGJ) and 2-deoxyglucose injections caused a 10% increase in feeding. 2-Deoxyglucose is a mild appetite stimulant in The molecular regulation of body weight and feeding behaviour is a highly complex process and recent discoveries regarding neuronal circuits and their hormonal regulation are leading to a much greater understanding of appetite control and energy homeostasis (for reviews see Wynne et al. [14] and Stanley et al. [15]). The discovery of leptin and a number of other gene products in rodent models has shown that similar mechanisms are involved in body weight regulation across different mammalian species. Multiple hormones, neurotransmitters and peptides are implicated in the inhibition of feeding behaviour in rodents, and ICV injections have confirmed a central appetite-suppression effect for at least 16 of these factors [16]. Although several studies have reported the effect of NB-DNJ to cause weight loss in growing mice and more recently in humans [9,10], the mechanism has never been established. A number of possible mechanisms have been proposed, including inhibition of a variety of enzyme activities as well as appetite suppression. In the present study, we were able to show a central anorectic effect of NB-DNJ by injecting directly into the lateral ventricle followed by careful assessment of dietary 164 j Diabetes, Obesity and Metabolism, 10, 2008, # 2007 Blackwell Publishing Ltd

7 D. A. Priestman et al. N-butyldeoxynojirimycin causes weight loss as a result of appetite suppression j OA Fig. 5 Weight change vs. diet eaten 24-h post-icv compound/vehicle infusion (mean s.e.m.). Three groups of mice (saline n ¼ 19, NB-DGJ n ¼ 12, NB-DNJ n ¼ 12, 2-deoxyglucose n ¼ 12) were injected intracerebroventricularly with 2 nmol compound in 1 ml saline vehicle. Twenty-four hours following the injection, the mice were weighed and also a preweighed amount of diet was reweighed to assess dietary intake. Two further groups (n ¼ 5) had subcutaneous osmotic pump implants connected to a high stability, low profile, cannula positioned to ensure delivery of NB-DGJ or NB-DNJ to the third cerebral ventricle. The mini-pump implants contained 500 mm compound and a mean pumping rate of 0.22 ml/h, equivalent to 1.83 nmol/min. See Materials and Methods for further experimental details. intake and weight change. In a previous study on rats, feeding suppression was induced by intraventricular infusion of 1,5-anhydroglucitol (1-deoxyglucose) and 2-deoxyglucose [13]. Subsequent studies confirmed that the hydroxyl group on the glucopyranose ring, particularly at carbon 1 or 2, is involved in feeding modulation [17]. This is of particular interest here as the imino sugar NB-DNJ, is a glucose analogue lacking a hydroxyl group on carbon 1, and may suggest a comparable mechanism of action due to the structural similarity to 1,5- anhydroglucitol. However, in contrast to the action of 2-deoxyglucose, NB-DNJ did not cause an initial hyperphagic response in the present study. More recently, the same group has shown that neuroglucoprivation, induced by glucose analogues, increased the concentration and turnover rate of hypothalamic neuronal histamine [18]. Histamine plays an important role in feeding behaviour mediated through postsynaptic H1 and H2 receptors and the presynaptic H3 receptor. When histamine is injected intracerebroventricularly, it suppresses food intake [19]. Conversely, histamine antagonists increase food intake, providing an experimental opportunity to test whether the effects of N-substituted imino sugars on feeding behaviour are mediated through neuronal histamine. NB-DNJ inhibits glycosphingolipid biosynthesis and has been shown to be effective in mediating substrate reduction therapy in animal models of Tay-Sachs and Sandhoff disease [20,21]. It has also been shown to be effective in clinical trials for the treatment of patients with type I Gaucher disease [6,7,10]. In humans, the clinical trial in patients with Gaucher disease receiving mg, three times daily, showed a mean reduction in body weight of %. Three of 28 patients, one obese and two overweight (classified by body mass index), returned to normal weight. Only one patient classified as normal initially, became underweight by the end of the trial [10]. More recently, in a 24-month clinical trial with seven patients receiving NB-DNJ, mean body weight decreased from baseline by 5.6% [9]. The fact that the intraperitoneally injected mice eat 30% less than their age- and sex-matched controls suggests that both the bitter taste of imino sugars and possible effects on gut enzymes are unlikely to be factors in NB- DNJ-induced weight loss. It also supported the suggestion that there may be a central mechanism causing appetite suppression. The effect of NB-DNJ on lean mouse growth concurred with previous studies [4,11]. The leptin deficient ob/ob mouse is widely used as an animal model of obesity and weighs more than twice as much as a normal mouse, even when fed the same diet. The mouse has as much as a fivefold increase in fat content, is hypothermic, has decreased energy expenditure and is unusually efficient at converting metabolic fuels into fat [22]. This mouse proved to be a good model for investigating the effect of NB-DNJ to cause appetite suppression, which caused a greatly reduced rate of growth in 6 week old obese mice and weight loss in 12 week old animals. The lack of any effect of the galactose analogue, NB- DGJ, also implies that the significant inhibition of glycosphingolipid biosynthesis caused by both imino sugars plays no role in mouse growth. NB-DGJ could therefore be effectively used as a control compound when studying the effects of NB-DNJ on body weight of mice. While a well-defined effect on total body weight has only been seen in doses of 1200 mg/kg/day and above [11], a reduction in fat pad weight is apparent at doses as low as 15 mg/kg/day. This easily excisable discrete tissue may therefore prove useful in future studies to investigate the effects on adipose tissue reduction of other compounds. # 2007 Blackwell Publishing Ltd Diabetes, Obesity and Metabolism, 10, 2008, j 165

8 OA j N-butyldeoxynojirimycin causes weight loss as a result of appetite suppression D. A. Priestman et al. The results presented here show that weight loss in mice and, possibly in humans treated with high doses of NB-DNJ, is due to central appetite suppression. There may be other, as yet undiscovered, inhibitory activities of NB-DNJ, but the level of reduction in food intake seen in this study accounts for the levels of reduced growth/ weight loss in treated animals. It remains to be seen whether reduced dietary intake as a result of central appetite suppression is the likely sole cause of weight loss in human patients. Further studies will be required to investigate the precise molecular mechanism of action and a potential role for N-substituted imino sugars as a strategy for appetite suppression in the treatment of obesity in humans. Acknowledgements This work was supported by the Oxford Glycobiology Institute. References 1 Fischer PB, Collin M, Karlsson GB et al. The alphaglucosidase inhibitor N-butyldeoxynojirimycin inhibits human immunodeficiency virus entry at the level of post-cd4 binding. J Virol 1995; 69: Fischer PB, Karlsson GB, Dwek RA, Platt FM. N-butyldeoxynojirimycin-mediated inhibition of human immunodeficiency virus entry correlates with impaired gp120 shedding and gp41 exposure. J Virol 1996; 70: Platt FM, Neises GR, Dwek RA, Butters TD. N-butyldeoxynojirimycin is a novel inhibitor of glycolipid biosynthesis. J Biol Chem 1994; 269: Andersson U, Butters TD, Dwek RA, Platt FM. N-butyldeoxygalactonojirimycin: a more selective inhibitor of glycosphingolipid biosynthesis than N-butyldeoxynojirimycin, in vitro and in vivo. Biochem Pharmacol 2000; 59: Andersson U, Reinkensmeier G, Butters T, Dwek R, Platt F. Inhibition of glycogen breakdown by imino sugars in vitro and in vivo. Biochem Pharmacol 2004; 67: Cox T, Lachmann R, Hollak C et al. Novel oral treatment of Gaucher s disease with N-butyldeoxynojirimycin (OGT 918) to decrease substrate biosynthesis. Lancet 2000; 355: Lachmann RH. Miglustat. Oxford GlycoSciences/ Actelion. Curr Opin Investig Drugs 2003; 4: Elstein D, Hollak C, Aerts JM et al. Sustained therapeutic effects of oral miglustat (Zavesca, N-butyldeoxynojirimycin, OGT 918) in type I Gaucher disease. J Inherit Metab Dis 2004; 27: Pastores GM, Barnett NL, Kolodny EH. An open-label, noncomparative study of miglustat in type I Gaucher disease: efficacy and tolerability over 24 months of treatment. Clin Ther 2005; 27: Lachmann RH, Platt FM. Substrate reduction therapy for glycosphingolipid storage disorders. Expert Opin Investig Drugs 2001; 10: Platt FM, Reinkensmeier G, Dwek RA, Butters TD. Extensive glycosphingolipid depletion in the liver and lymphoid organs of mice treated with N-butyldeoxynojirimycin. J Biol Chem 1997; 272: Jeyakumar M, Dwek RA, Butters TD, Platt FM. Storage solutions: treating lysosomal disorders of the brain. Nat Rev Neurosci 2005; 6: Tsutsui K, Sakata T, Oomura Y, Arase K, Fukushima M, Hinohara Y. Feeding suppression induced by intraventricle III infusion of 1,5-anhydroglucitol. Physiol Behav 1983; 31: Wynne K, Stanley S, McGowan B, Bloom S. Appetite control. J Endocrinol 2005; 184: Stanley S, Wynne K, McGowan B, Bloom S. Hormonal regulation of food intake. Physiol Rev 2005; 85: Ahima RS, Osei SY. Molecular regulation of eating behavior: new insights and prospects for therapeutic strategies. Trends Mol Med 2001; 7: Sakata T, Kurokawa M. Feeding modulation by pentose and hexose analogues. Am J Clin Nutr 1992; 55: 272S 277S. 18 Yoshimatsu H, Chiba S, Tajima D, Akehi Y, Sakata T. Histidine suppresses food intake through its conversion into neuronal histamine. Exp Biol Med (Maywood) 2002; 227: Sakata T, Fukagawa K, Ookuma K et al. Modulation of neuronal histamine in control of food intake. Physiol Behav 1988; 44: Platt FM, Neises GR, Reinkensmeier G et al. Prevention of lysosomal storage in Tay-Sachs mice treated with N-butyldeoxynojirimycin. Science 1997; 276: Jeyakumar M, Butters TD, Cortina-Borja M et al. Delayed symptom onset and increased life expectancy in Sandhoff disease mice treated with N-butyldeoxynojirimycin. Proc Natl Acad Sci U S A 1999; 96: Friedman JM, Halaas JL. Leptin and the regulation of body weight in mammals. Nature 1998; 395: j Diabetes, Obesity and Metabolism, 10, 2008, # 2007 Blackwell Publishing Ltd