MetAP2 inhibitor treatment of high-fat and -fructose-fed dogs: impact on the response to oral glucose ingestion and a hyperinsulinemic hyperglycemic clamp

Abstract

We examined the methionine aminopeptidase 2 inhibitor fumagillin in dogs consuming a high-fat and -fructose diet (HFFD). In pilot studies (3 dogs that had consumed HFFD for 3 yr), 8 wk of daily treatment with fumagillin reduced food intake 29%, weight 6%, and the glycemic excursion during an oral glucose tolerance test (OGTT) 44%. A second group of dogs consumed the HFFD for 17 wk: pretreatment (weeks 0–4), treatment with fumagillin (FUM; n = 6), or no drug (Control, n = 8) (weeks 4–12), washout period (weeks 12–16), and fumagillin or no drug for 1 wk (week 17). OGTTs were performed at 0, 4, 11, and 16 wk. A hyperinsulinemic hyperglycemic clamp was performed in week 12; 4 chow-fed dogs underwent identical clamps. Kilocalories per day intake during the treatment period was 2,067 ± 50 (Control) versus 1,824 ± 202 (FUM). Body weights (kg) increased 1.9 ± 0.3 vs. 2.7 ± 0.8 (0–4 wk) and 1.2 ± 0.2 vs. −0.02 ± 0.9 (4–12 wk) in Control versus fumagillin. The OGTT glycemic response was 30% greater in Control versus fumagillin at 11 wk. Net hepatic glucose uptake (NHGU; mg·kg⁻¹·min⁻¹) in the Chow, Control, and fumagillin dogs was ~1.5 ± 0.6, −0.1 ± 0.1, and 0.3 ± 0.4 (with no portal glucose infusion) and 3.1 ± 0.6, 0.5 ± 0.3, and 1.5 ± 0.5 (portal glucose infusion at 4 mg·kg⁻¹·min⁻¹), respectively. Fumagillin improved glucose tolerance and NHGU in HFFD dogs, suggesting methionine aminopeptidase 2 (MetAP2) inhibitors have the potential for improving glycemic control in prediabetes and diabetes.

INTRODUCTION

Methionine aminopeptidase 2 (MetAP2) is a broadly expressed metalloenzyme that cleaves amino-terminal methionine residues from newly synthesized proteins (37). Positive metabolic effects of MetAP2 inhibitors have been observed in studies conducted in obese rodent models and humans. Substantial weight loss, at least partially due to a reduction in food consumption, has been observed in obese mice treated with MetAP2 inhibitors, in association with increases in adipose lipolysis and hepatic fatty acid oxidation (2, 18, 33). Additionally, glucose tolerance improved with 2 wk of MetAP2 inhibitor treatment in a rat model with obesity induced by hypothalamic damage (12). Moreover, in obese human adults with type 2 diabetes, a 6-mo trial of the MetAP2 inhibitor beloranib significantly reduced both body weight and HbA1c (30).

While data indicate that MetAP2 inhibitors reduce adipocyte size (18), little is known about their impact on metabolism in other tissues. In particular, the improvements in metabolic markers with MetAP2 inhibitor treatment raise the question of what impact MetAP2 inhibition might have on hepatic glucose uptake and metabolism under meal feeding conditions. The dog is virtually unique among animal models, in that it is possible to assess hepatic glucose uptake directly (23). The techniques required are too invasive for use in humans and too difficult to perform in rodents because of their small size. Our laboratory has demonstrated that a high-fat, high-fructose diet (HFFD) induces weight gain, impairs glucose tolerance, and reduces hepatic glucose uptake in response to hyperinsulinemic hyperglycemic conditions in mongrel dogs (6). We carried out the current studies to investigate the impact of MetAP2 inhibitor treatment on body weight, food intake, glucose tolerance, and hepatic glucose uptake in obese and glucose intolerance HFFD-fed dogs. Fumagillin, isolated from Aspergillus fumigatus, was utilized in these studies since it is a well-characterized, highly selective MetAP2 inhibitor appropriate for use as a prototype for chronic oral administration (18, 35).

RESEARCH DESIGN AND METHODS

Animal Care, Diet, and Surgical Procedures

Studies were carried out on adult male mongrel dogs obtained from a U.S. Department of Agriculture (USDA)-licensed vendor. The protocol was approved by the Vanderbilt University Institutional Animal Care and Use Committee, and the animals were housed and cared for according to Association for Assessment and Accreditation of Laboratory Animal Care and USDA guidelines. The dogs consumed a chow and meat diet [31% protein, 26% fat, and 43% carbohydrate, virtually all in the form of digestible starch (8)] or the HFFD (20% protein, 53% fat and 27% carbohydrate, with 17% of total energy supplied by fructose). We have previously reported the metabolic effects of chronic consumption of the HFFD (6). The HFFD was fed in amounts exceeding maintenance energy needs.

Pilot studies were carried out primarily to determine the appropriate fumagillin dosage in the mongrel dog. The pilot studies utilized three dogs that had been consuming the HFFD for ~3 yr and continued the diet throughout the current studies. Food intake and body weight in the pilot dogs were stable over the 3-mo period preceding fumagillin treatment. In the pilot dogs, fumagillin was administered orally at 0.1 mg·kg⁻¹·day⁻¹ for 5 wk and then at 0.2 mg·kg⁻¹·day⁻¹ for an additional 3 wk.

In the second set of studies, dogs that were fed the chow and meat diet underwent initial assessment and then were placed on the HFFD for 17 wk; we have previously shown that consumption of the HFFD induces glucose intolerance within 4 wk (8). At the end of the first 4 wk of HFFD consumption, these dogs were divided into two groups, fumagillin-treated (n = 6) and Control (n = 8 total, consisting of 2 placebo-treated and 6 untreated dogs). In the pilot study, there was no apparent enhancement of the effect of fumagillin with the 0.2 mg·kg⁻¹·day⁻¹ dosage, and therefore, the second set of dogs received 0.1 mg·kg⁻¹·day⁻¹ throughout their treatment. The placebo was the buffer solution used for delivery of the fumagillin, sodium carbonate neutralized with 8.5% phosphoric acid. No differences in the responses of the placebo-treated and the nonplacebo-treated Control animals were observed, and therefore, their results were combined into one Control group. The fumagillin or placebo treatment continued for 8 wk (weeks 4 to 12).

After 6 wk on the fumagillin or placebo/no treatment, the dogs in the second set of studies underwent surgery under general anesthesia to place sampling catheters in the femoral artery, hepatic portal vein, and left common hepatic vein; blood flow probes around the portal vein and hepatic artery; and catheters in a splenic and a jejunal vein to allow infusion into the portal circulation (7).

After 8 wk of daily oral dosing, fumagillin or placebo treatment was stopped, while HFFD consumption continued. After a 4-wk “washout” period (weeks 12 to 16), the fumagillin group resumed drug treatment for 1 wk before euthanasia and tissue harvest.

Experimental Procedures

OGTTs.

Dogs were fasted 18 h before each OGTT; just before beginning the fast, each dog was fed a can of meat. On the morning of each OGTT, a catheter was introduced via a peripheral leg vein and threaded into a deep venous location to allow collection of samples for glucose, insulin, glucagon, and C-peptide (7). Following catheter insertion, there was a 60-min period of acclimation (no intervention) and a 10-min basal sampling period. The glucose load (0.9 g glucose/kg baseline weight; SolCarb, Medica Nutrition, Englewood, NJ) was then administered orally, and sampling continued for 180 min. The catheter was then removed.

The pilot dogs underwent three OGTTs: 1) before fumagillin treatment, 2) after 8 wk of treatment, and 3) 4 wk after treatment ceased. In the second set of studies, eight Control and four of the six fumagillin-treated dogs underwent four OGTTs: 1) before beginning the HFFD (week 0), 2) 4 wk after beginning the HFFD (before any treatment; week 4), 3) 7 wk after beginning fumagillin or placebo/no treatment (week 11), and 4) 4 wk after discontinuing the fumagillin or placebo treatment but while still consuming the HFFD (week 16).

Clamp study.

During week 12, the week after the third OGTT, the second set of dogs underwent a hyperinsulinemic hyperglycemic clamp experiment. The prestudy feeding and fasting conditions were the same as those for the OGTTs. On the morning of study, the sampling and infusion catheters and flow probes were removed from their subcutaneous pockets under local anesthesia, and the dog was then placed in the Pavlov harness. Vascular access was established in three peripheral veins. At −120 min, a priming dose of [3-³H]glucose (38 μCi) was administered, and a constant infusion of [3-³H]glucose (0.38 μCi/min) was begun. Each experiment consisted of a 100-min equilibration period (−120 to −20 min), a 20-min period of basal sampling (−20 to 0 min), and a 180-min experimental period divided into two subperiods (period 1, or P1, 0 to 90 min; period 2, or P2, 90 to 180 min). At time 0, a constant peripheral venous infusion of somatostatin (0.8 µg·kg⁻¹·min⁻¹) began, and continuous intraportal infusions of insulin (3 to 4 × basal; 1.2 mU·kg^-1.min⁻¹) and glucagon (basal; 0.55 ng·kg^-1.min⁻¹) were started. A peripheral infusion of 50% glucose was administered, with the infusion rate adjusted as needed to double the hepatic glucose load. In P2, 20% glucose was infused into the hepatic portal circulation at 4 mg·kg^-1.min⁻¹ to create a feeding signal (25), and the peripheral glucose infusion rate was adjusted as necessary to maintain the hepatic glucose load the same as in P1. At the end of the study, the infusions were discontinued, the catheters and flow probes were aseptically replaced in their subcutaneous pockets under general anesthesia, and the dogs were allowed to recover. The clamp data from both groups were compared with previously published data (8) from dogs undergoing identical clamps but consuming a standard diet of chow and meat (Chow group, n = 4).

Chronic blood monitoring.

Four fumagillin-treated dogs and the two placebo-treated Control dogs underwent chronic blood sampling to assess the safety of the drug treatment. After receiving a can of meat and 200 g chow at noon on the day preceding sampling, the dogs were then fasted overnight, with venous blood samples collected in the morning. These procedures were carried out to ensure that food intake before sampling was identical among all animals. Samples were collected during weeks 0, 2, 4, 6, 8, and 12 and analyzed for glucagon-like peptide 1 (GLP-1) and nonesterified fatty acid (NEFA) concentrations. In addition, samples were collected for platelet counts, prothrombin time, fibrinogen, D-dimer, triglycerides, and β-hydroxybutyrate (β-OHB) at 0, 4, 8, and 12 wk.

Analyses

Biochemical analyses.

Deep venous samples taken during the OGTTs were analyzed for glucose, insulin, glucagon, and C-peptide as previously described (13, 26). Arterial, hepatic portal vein, and hepatic vein samples from the clamp studies were analyzed for plasma glucose, [³H]glucose, glucagon, insulin, and NEFA levels and blood lactate, alanine, glycerol, and β-OHB concentrations as described previously (13, 26).

Calculations.

The glucose and insulin responses to the OGTTs were calculated as the change from basal during the 180-min period after intake of the glucose load (abbreviated as ΔAUC). Net hepatic substrate balances, net hepatic carbon retention, and nonhepatic glucose uptake (nonHGU) were calculated with the arteriovenous difference method, i.e., the difference between the inflowing load from the hepatic artery and portal vein and the outflowing load in a hepatic vein (6). Net hepatic glucose balance (NHGB) was calculated with an indirect method as previously described (6, 34) to reduce any error introduced by streaming of infusate in the portal vein. Unidirectional hepatic glucose uptake (HGU; utilizing [³H]glucose measurements) and hepatic sinusoidal plasma insulin and glucagon concentrations were calculated as previously described (6). Glycogen synthesis via the direct pathway was calculated by dividing hepatic [³H]-labeled glycogen at the end of the study by the average inflowing plasma [³H]glucose specific radioactivity (34). Net hepatic carbon retention was calculated from the sum of the balance data for glucose and its metabolites; we have demonstrated that this approach provides an index of glycogen synthesis (34).

Statistical analyses.

Data are expressed as means ± SE. Two-way ANOVA with or without repeated measures design was used (Systat Software, San Jose, CA), and post hoc analysis was performed using the Student-Newman-Keuls multiple comparisons test. A P value < 0.05 was considered significant.

RESULTS

Pilot Studies in Three Dogs Fed the HFFD Chronically

Food consumption, body weights, and blood testing.

Food consumption and body weight data for the fumagillin-treated pilot dogs that had consumed the HFFD long term are shown in Fig. 1, A and B, respectively. Fumagillin treatment was associated with reduced food intake in all three dogs (means ± SE of intakes before and during treatment: 1,677 ± 262 and 1,187 ± 258 kcal/day, respectively; P < 0.05). Food intake returned to pretreatment amounts after fumagillin treatment ended (1,705 ± 45 kcal/day during the 4 wk washout). Body weight tended to decline (P < 0.08) in all three dogs during treatment (33.6 ± 2.2 to 30.8 ± 3.4 kg before and at the end of treatment, respectively) and was no different from the pretreatment value at the end of the washout period. Pretreatment venous blood β-OHB was 51 ± 7 µmol/L, and it had risen to 85 ± 15 µmol/L by the end of the treatment period (P = 0.06). Other metabolite, blood chemistry, and blood count data showed no trend toward changes over time, and there were no differences evident between the two dosages of fumagillin (data not shown).

Fig. 1.

A and B: energy intake from high-fat and -fructose diet (HFFD) (A; data are the daily mean intake for the previous week) and body weights (B) for each individual animal (n = 3) that had consumed the HFFD for ∼3 yr. These dogs received fumagillin (FUM) for 8 wk, followed by a 4-wk washout period. C and D: energy intake (C) and body weight (D) in dogs that received a chow and meat diet initially (week 0) and then consumed the HFFD for 17 wk (data are means ± SE). After 4 wk of HFFD intake, the 2nd set of dogs received placebo/no drug treatment (Control; open symbols; n = 8) or FUM treatment (n = 4; filled symbols) during weeks 4–12. There was no FUM treatment between 12 and 16 wk. The FUM dogs were then returned to drug treatment for 1 wk (week 17). *P < 0.05 for fumagillin vs. Control at this time; †P < 0.05 for weights within the Control group, compared with the 4-wk measurement.

OGTT data.

The glucose responses to the OGTTs before fumagillin treatment, at the end of 8 wk of treatment, and 4 wk after treatment ended are shown in Fig. 2, A–C. The corresponding insulin responses are shown in Fig. 2, D–F. The ΔAUC for the glucose response (Fig. 2G) tended to be reduced by fumagillin treatment (P = 0.08 vs. pretreatment), and it increased significantly during the washout period (P < 0.05 vs. treatment). The ΔAUC of the insulin response tended to be reduced by drug treatment (P = 0.08 vs. pretreatment), but it did not differ between the washout and treatment periods (Fig. 2H). When the insulin response was calculated in relation to the glycemic response (Fig. 2I), there were no significant changes over time; nevertheless, all three pilot dogs showed a numeric increase while receiving fumagillin treatment and a reversal of that effect during the washout period.

Fig. 2.

Oral glucose tolerance test (OGTT) data from the 3 long-term high-fat and -fructose diet (HFFD) dogs. Glucose (A–C) and insulin (D–F) responses in the individual dogs before fumagillin (FUM) treatment (open circles), after 8 wk of treatment (closed circles), and 4 wk after fumagillin treatment ended (washout period; open triangles). Change from basal during the 180-min period after intake of the glucose load (ΔAUC_{0–180 min}) for glucose (G) and insulin responses (H) and change in insulin relative to change in glucose (I) for each OGTT. Each dog has a unique symbol in G–I. Units for I are µU/mL relative to mg/dL. P = 0.08 for the FUM treatment period vs. pretreatment in G and I. P = 0.05 for the FUM treatment period vs. pretreatment and for FUM treatment vs. washout in H. In G, P < 0.05 for differences between time points marked with different letters.

Second Set of Studies, Comparing Fumagillin-Treated and Control Dogs

Food consumption, body weights, and blood monitoring.

In the second set of studies, caloric intake increased in the first few days after the change to the HFFD, in comparison to the baseline chow and meat intake, averaging 2,774 ± 165 and 2,962 ± 168 kcal/day during the first week in the Control and fumagillin dogs, respectively (P = 0.23; Fig. 1C). HFFD intake declined in both groups between the first and fourth weeks of the diet’s consumption. During week 4, mean intake was 2,123 ± 121 and 2,041 ± 236 kcal/day in the Control and fumagillin groups, respectively (P = 0.35). HFFD intake remained relatively stable in the Control dogs during the 4- to 12-wk period, while it declined in the fumagillin group, so that the mean energy intake in the fumagillin group during the period between 4 and 12 wk was only 88% of that in the Control group (P < 0.05). There was no change in food intake between weeks 12 and 17 within the Control group. There was an initial increase in food intake in the fumagillin group after cessation of drug treatment, but by the end of the washout period this declined again to amounts no different from those during the final half of the treatment period; when fumagillin was reintroduced for the final week, food intake was again reduced compared with that in the Control dogs (P < 0.05).

Over the first 4 wk of HFFD consumption there was an increase in body weight of 1.9 ± 0.3 and 2.4 ± 0.5 kg in the Control and fumagillin groups, respectively (P = 0.10 between groups; P < 0.05 for weights at 4 vs. 0 wk within each group; Fig. 1D). While the Control dogs gained 1.2 ± 0.2 kg between weeks 4 and 12 (P < 0.01 for change within the group), the fumagillin dogs exhibited no weight change over that time period (P = 0.5 for the within-group comparison; P < 0.05 for between-group comparison of change during that time period). There were no significant differences between groups in weight or weight change during the washout period between 12 and 16 wk or during the final week of fumagillin/placebo reintroduction (week 17). Nevertheless, the fumagillin-treated dogs gained almost 1 kg between weeks 12 and 16, when there was no drug treatment.

Supplemental Table S1 (see https://doi.org/10.6084/m9.figshare.9985232.v1) shows the venous GLP-1, insulin, glucagon β-OHB, NEFA, and triglyceride concentrations, as well as the platelet count and prothrombin time, over the course of the interventions in the fumagillin- and placebo-treated dogs. While the number of placebo-treated dogs (n = 2) precludes any statistical comparisons, it appears that plasma insulin concentrations might have been slightly higher in the fumagillin vs. placebo dogs throughout the duration of the study. Where there were any abnormal values (e.g., elevated D-dimer concentrations in 1 placebo and 1 fumagillin-treated dog), the abnormality existed before the start of the drug/placebo treatment. There were no consistent differences between the placebo and fumagillin-treated dogs in any other blood parameters monitored chronically (complete blood count and serum chemistry; data not shown).

OGTT Data

Baseline (week 0).

The glucose response to the baseline OGTT, during chow and meat consumption, did not differ between groups, with an ΔAUC of 3,396 ± 359 and 3,936 ± 840 mg in the Control and fumagillin groups, respectively (P = 0.2; Table 1 and Fig. 3A). The ΔAUC of the insulin response (Fig. 3B) was higher (P < 0.01) in the fumagillin (2,844 ± 301 µU) dogs, compared with the Control dogs (1,811 ± 169 µU), and the insulin response expressed relative to the glycemic response was ~35% greater in the fumagillin vs. Control group (P = 0.07; Fig. 3C). The basal glucagon concentrations (before glucose intake) were significantly greater in the Control vs. fumagillin dogs (P < 0.05). However, the C-peptide and glucagon responses to glucose administration did not differ significantly between groups when evaluated with ANOVA (Table 2).

Table 1.

Glucose and insulin responses during the OGTTs in the second set of studies

	Basal	Time, min
		5	10	20	30	40	50	60	70	80	90	120	150	180
Glucose (week 0 – before HFFD)
Control	111 ± 2	113 ± 4	124 ± 5	143 ± 6	149 ± 7	146 ± 8	148 ± 8	151 ± 10	148 ± 10	150 ± 8	138 ± 5	124 ± 3	112 ± 6	106 ± 4
FUM	114 ± 3	123 ± 7	141 ± 8	160 ± 8	166 ± 8	168 ± 5	156 ± 11	151 ± 14	151 ± 8	150 ± 9	146 ± 5	124 ± 8	116 ± 4	114 ± 1
Glucose (week 4 of HFFD – before fumagillin treatment)
Control	110 ± 3	112 ± 3	128 ± 6	161 ± 13	194 ± 14	203 ± 13	206 ± 12	196 ± 13	183 ± 13	170 ± 15	151 ± 20	118 ± 17	102 ± 7	100 ± 4
FUM	110 ± 4	118 ± 5	140 ± 12	169 ± 23	195 ± 31	208 ± 24	209 ± 13	213 ± 13	200 ± 21	168 ± 24	134 ± 22	112 ± 14	113 ± 7	109 ± 4
Glucose (week 11 of HFFD – after 7 wk of treatment with fumagillin or no drug)
Control	108 ± 3	119 ± 5	136 ± 10	168 ± 13	194 ± 11	206 ± 9	201 ± 9	175 ± 15	156 ± 16	138 ± 12	140 ± 9	117 ± 12	101 ± 6	102 ± 6
FUM	117 ± 3	126 ± 6	152 ± 5	186 ± 4	201 ± 8	204 ± 13	194 ± 19	180 ± 20	163 ± 18	139 ± 11	121 ± 13	108 ± 8	107 ± 4	109 ± 4
Glucose (week 16 of HFFD – after 4 wk of fumagillin washout)
Control	108 ± 2	114 ± 5	138 ± 8	186 ± 9	212 ± 9	208 ± 11	201 ± 12	186 ± 9	158 ± 12	136 ± 15	118 ± 11	94 ± 10	95 ± 5	102 ± 3
FUM	113 ± 2	116 ± 6	118 ± 5	145 ± 8	190 ± 5	224 ± 6	241 ± 14	236 ± 29	216 ± 32	193 ± 25	153 ± 20	108 ± 7	102 ± 4	104 ± 4
Insulin (week 0 – before HFFD)
Control	5 ± 1	16 ± 7	18 ± 6	30 ± 7	27 ± 4	19 ± 2	26 ± 6	23 ± 5	21 ± 5	22 ± 3	16 ± 3	13 ± 3	7 ± 1	5 ± 1
FUM	5 ± 1	27 ± 3	39 ± 6	43 ± 9	32 ± 6	42 ± 5	32 ± 6	23 ± 4	32 ± 8	25 ± 8	36 ± 12	10 ± 4	4 ± 0	7 ± 1
Insulin (week 4 of HFFD – before fumagillin treatment)
Control	7 ± 1	13 ± 5	25 ± 6	31 ± 7	42 ± 5	42 ± 5	42 ± 5	32 ± 7	30 ± 3	34 ± 2	19 ± 4	11 ± 5	6 ± 1	6 ± 1
FUM	6 ± 1	26 ± 11	40 ± 14	38 ± 13	38 ± 19	57 ± 15	53 ± 12	51 ± 13	53 ± 17	49 ± 19	26 ± 8	9 ± 2	7 ± 3	6 ± 1
Insulin (week 11 of HFFD – after 7 wk of treatment with fumagillin or no drug)
Control	6 ± 1	18 ± 3	24 ± 4	36 ± 3	34 ± 3	39 ± 5	38 ± 4	28 ± 5	23 ± 6	16 ± 5	18 ± 4	13 ± 5	6 ± 1	5 ± 1
FUM	12 ± 1	41 ± 6	52 ± 6	46 ± 7	56 ± 5	67 ± 11	63 ± 6	49 ± 8	37 ± 17	28 ± 10	18 ± 6	11 ± 5	9 ± 1	12 ± 2
Insulin (week 16 of HFFD – after 4 wk of fumagillin washout)
Control	7 ± 1	25 ± 8	52 ± 16	49 ± 7	52 ± 6	53 ± 4	44 ± 7	37 ± 5	31 ± 6	23 ± 6	17 ± 4	7 ± 1	6 ± 1	7 ± 1
FUM	9 ± 1	10 ± 2	14 ± 3	35 ± 13	48 ± 16	51 ± 7	72 ± 18	59 ± 11	44 ± 6	45 ± 5	27 ± 4	8 ± 1	10 ± 2	8 ± 2

Data are means ± SE; n = 8 Control and n = 4 fumagillin (FUM). Units for glucose were mg/dL and for insulin were µg/mL. All dogs consumed the high-fat and -fructose diet (HFFD) after the oral glucose tolerance test (OGTT) in week 0. Control dogs received placebo (n = 2) or no pharmaceutical treatment (n = 6). Statistical analyses were not carried out on these data; see Fig. 3 for change from basal during the 180-min period after intake of the glucose load (ΔAUC_{0–180 min}) data, which did undergo statistical analysis.

Fig. 3.

Oral glucose tolerance test (OGTT) results in the 2nd set of dogs, newly introduced to the high-fat and -fructose diet (HFFD). Glucose (A), insulin (B), and insulin/glucose (C). Data are change from basal during the 180-min period after intake of the glucose load (ΔAUC_{0–180 min}) for the individual dogs (Control n = 8, open symbols; FUM n = 4, filled symbols; median shown as solid line); see Table 1 for values for the individual time points during each OGTT. Units for C are µU/mL relative to mg/dL. Where 2 time points within a group do not share a common letter, P < 0.05 for within-group differences at those time points. *P < 0.05 vs. Control at the same time.

Table 2.

C-peptide and glucagon responses during the OGTTs in the second set of dogs

	Basal	Time (min)
		10	20	30	50	60	70	90	120	150	180
C-peptide (week 0 – before HFFD)
Control	0.2 ± 0.1	0.6 ± 0.2	1.0 ± 0.2	0.9 ± 0.2	0.9 ± 0.2	0.9 ± 0.2	0.9 ± 0.2	0.7 ± 0.1	0.4 ± 0.1	0.2 ± 0.1	0.2 ± 0.1
FUM	0.2 ± 0.0	1.1 ± 0.1	1.2 ± 0.1	1.0 ± 0.1	1.1 ± 0.1	0.7 ± 0.2	1.0 ± 0.2	1.0 ± 0.4	0.3 ± 0.1	0.2 ± 0.0	0.2 ± 0.1
C-peptide (week 4 of HFFD – before FUM treatment)
Control	0.3 ± 0.1	0.9 ± 0.1	1.1 ± 0.2	1.4 ± 0.1	1.3 ± 0.2	1.1 ± 0.2	1.0 ± 0.2	0.8 ± 0.2	0.6 ± 0.3	0.3 ± 0.1	0.3 ± 0.1
FUM	0.2 ± 0.0	1.0 ± 0.3	1.1 ± 0.3	1.3 ± 0.4	1.4 ± 0.3	1.3 ± 0.2	1.3 ± 0.3	0.6 ± 0.2	0.2 ± 0.1	0.2 ± 0.2	0.2 ± 0.0
C-peptide (week 11 of HFFD – after 7 wk of treatment with FUM or no drug)
Control	0.2 ± 0.0	1.1 ± 0.3	1.2 ± 0.1	1.2 ± 0.3	1.3 ± 0.1	1.1 ± 0.1	0.9 ± 0.3	0.5 ± 0.1	0.3 ± 0.1	0.2 ± 0.1	0.2 ± 0.0
FUM	0.5 ± 0.1*	1.4 ± 0.2	1.3 ± 0.3	1.6 ± 0.2	1.6 ± 0.3	1.3 ± 0.3	1.1 ± 0.4	0.5 ± 0.1	0.3 ± 0.2	0.3 ± 0.1	0.4 ± 0.0
C-peptide (week 16 of HFFD – after 4 wk of FUM washout)
Control	0.2 ± 0.0	1.2 ± 0.2	1.3 ± 0.1	1.3 ± 0.1	1.4 ± 0.1	1.0 ± 0.1	0.8 ± 0.2	0.5 ± 0.2	0.2 ± 0.1	0.2 ± 0.1	0.2 ± 0.0
FUM	0.2 ± 0.1	0.4 ± 0.1	1.1 ± 0.4	1.3 ± 0.2	1.6 ± 0.4	1.1 ± 0.1	1.1 ± 0.1	0.7 ± 0.1	0.2 ± 0.0	0.2 ± 0.1	0.2 ± 0.1
Glucagon (week 0 – before HFFD)
Control	40 ± 4			34 ± 2		35 ± 3		36 ± 4	34 ± 3	35 ± 4	35 ± 2
FUM	26 ± 2*			24 ± 2		32 ± 4		27 ± 3	24 ± 2	26 ± 2	26 ± 2
Glucagon (week 4 of HFFD – before FUM treatment)
Control	32 ± 5			31 ± 3		29 ± 3		21 ± 2	28 ± 4	24 ± 2	30 ± 4
FUM	27 ± 4			27 ± 2		29 ± 3		32 ± 3	30 ± 3	30 ± 2	28 ± 2
Glucagon (week 11 of HFFD – after 7 wk of treatment with FUM or no drug)
Control	30 ± 3			27 ± 3		27 ± 3		27 ± 5	26 ± 3	23 ± 4	26 ± 4
FUM	25 ± 3			30 ± 3		30 ± 4		27 ± 3	25 ± 3	22 ± 4	23 ± 4
Glucagon (week 16 of HFFD – after 4 wk of FUM washout)
Control	31 ± 3			31 ± 4		27 ± 4		23 ± 5	27 ± 3	26 ± 3	29 ± 4
FUM	30 ± 3			27 ± 4		29 ± 4		27 ± 4	26 ± 3	30 ± 2	30 ± 4

Data are means ± SE; n = 8 Control and n = 4 fumagillin (FUM). Units for C-peptide are ng/mL and for glucagon were ng/L. All dogs consumed the high-fat and -fructose diet (HFFD) after the Oral glucose tolerance test (OGTT) in week 0. ANOVA revealed no significant differences between Control and fumagillin dogs during any of the OGTTs.

^*P < 0.05 between groups when basal values were compared with t test.

After 4 wk on the HFFD.

The glycemic excursion increased in both groups after 4 wk of HFFD consumption, with a very similar response in the two groups (ΔAUC: 6,500 ± 932 and 6,469 ± 1071 mg, Control and fumagillin, respectively; P = 0.64). Again, the ΔAUC of the insulin response was greater in the fumagillin group (2,543 ± 317 and 3,797 ± 8216 µU in Control and fumagillin, respectively; P < 0.05), regardless of the fact there was no difference between groups in treatment at this time. When the ΔAUC for the insulin response was expressed relative to that of the glucose response, the values did not differ significantly between groups, despite being ∼40% higher in the fumagillin group (0.45 ± 0.09 vs. 0.63 ± 0.17; P = 0.3) The C-peptide ΔAUC was ∼20% higher in the fumagillin vs. the Control group (P = 0.28), while the glucagon responses were similar between groups.

After 11 wk of HFFD (and 7 wk of fumagillin or Control treatment).

The glucose ΔAUC declined 12% (P = 0.18) between 4 and 11 wk in the Control group, while the ΔAUC for glucose in the fumagillin group fell ∼26% (P < 0.05 for week 11 vs. week 4 within the fumagillin group). Although the glucose ΔAUC at 11 wk was ∼30% greater in the Control vs. the fumagillin group, this did not reach statistical significance (P = 0.15). On the other hand, three of the four fumagillin-treated dogs exhibited a decrease in the insulin ΔAUC, compared with that during the OGTT conducted at week 4, although the change within the group was not significant. Moreover, the ΔAUC for the insulin response relative to the glucose response was ∼84% greater in the fumagillin group relative to the Control dogs (P < 0.05). The fasting C-peptide concentrations were 2.5 times higher in the fumagillin vs. the Control group (P < 0.05 when evaluated with t test). However, the C-peptide response to the glucose load was similar between groups, with ΔAUCs of 67 ± 5 and 71 ± 10 ng in the Control and fumagillin groups, respectively (P = 0.33) (Table 1). Again, the glucagon responses were not different between groups.

Posttreatment.

After 16 wk of HFFD consumption, as well as 4 wk off drug treatment in the fumagillin group, the glycemic response showed a tendency to be greater in the fumagillin dogs, compared with the Controls (ΔAUC: 5,453 ± 345 vs. 7,023 ± 1528 in Control vs. fumagillin, respectively; P = 0.08). The Control group demonstrated little change in the glycemic response in week 16 vs. week 11 (P = 0.33 within the group), but the glycemic response in the fumagillin dogs showed a strong tendency to increase over that time period (ΔAUC: 47% greater; P = 0.09). Neither the insulin responses (P = 0.19) nor the insulin response relative to the glucose response (P = 0.35) differed between groups. C-peptide and glucagon responses were also similar in the two groups.

Clamp Study Data

Hormone data.

The clamp study hormone data are shown in Table 3. The arterial insulin concentrations increased approximately three- to fourfold basal during the clamp periods in all groups, as designed, and the insulin concentrations did not differ significantly among the three groups at any time. Glucagon concentrations remained at basal levels during the clamp; again, the concentrations did not differ among the treatments. Cortisol concentrations also remained at basal levels in all groups during the clamp period, with no differences observed among the groups.

Table 3.

Arterial hormone concentrations during the clamp studies

Hormone and Group	Basal (−20 to 0 min)	Period 1 (0 to 90 min)	Period 2 (90 to 180 min)
Insulin, µU/mL
Chow	8 ± 1	25 ± 5	27 ± 6
HFFD	10 ± 2	27 ± 2	29 ± 2
HFFD + FUM	9 ± 1	29 ± 5	29 ± 4
Glucagon, ng/L
Chow	33 ± 3	36 ± 5	36 ± 4
HFFD	40 ± 4	39 ± 5	32 ± 4
HFFD + FUM	32 ± 6	39 ± 5	40 ± 5
Cortisol, µg/dL
Chow	2 ± 0	2 ± 1	2 ± 0
HFFD	3 ± 0	3 ± 0	2 ± 0
HFFD + FUM	2 ± 0	2 ± 0	2 ± 0

Data are means ± SE; n = 4, 8, and 6 for Chow, high-fat and -fructose diet (HFFD), and HFFD + fumagillin (FUM), respectively. Basal period values are the mean of 2 sampling points, while data for Periods 1 and 2 are the means of 2, 3, or 5 sampling times (for cortisol, glucagon, and insulin, respectively). There were no significant differences among groups.

Glucose data.

Arterial plasma glucose concentrations during the basal period were ∼108 mg/dL in all groups, and glucose was infused into a peripheral vein during P1 to elevate the arterial concentrations to ∼220 mg/dL (Fig. 4A). During P2, when glucose was delivered into both the portal circulation and the peripheral vein, the arterial concentrations were decreased to ∼202 mg/dL. In this manner, the hepatic glucose load was equivalent, at approximately twofold basal, during the two clamp periods (Fig. 4B). The glucose infusion rates (GIR) required to maintain the clamp were not significantly different between the Chow group and either the HFFD or the HFFD + fumagillin groups (Fig. 4C). However, the GIR was numerically lower in the HFFD group than the other two groups at all but one sampling point during the clamp periods.

Fig. 4.

Clamp study glucose data. Arterial plasma glucose (A), hepatic glucose load (B), glucose infusion rate (C) (peripheral vein infusion during Period 1; peripheral and portal delivery during Period 2), net hepatic glucose balance (D), hepatic carbon retention/glycogen synthesis (E), and nonhepatic glucose uptake (F) in dogs receiving Chow (closed circles, n = 4; previously published data (8)), HFFD Control (open squares; n = 8), and HFFD + FUM (closed triangles, n = 6). Where a P value is shown, there was a main effect but no individual time points were identified as significantly different on post hoc testing. Significant results (P < 0.05) in post hoc tests are shown by symbols at individual time points: *HFFD vs. Chow; ‡HFFD + FUM vs. Chow

All groups exhibited net hepatic glucose output (NHGO) in the basal state, with similar rates evident among the groups (Fig. 4D). The Chow group shifted to net hepatic glucose uptake (NHGU) within 15 min of the start of P1 and exhibited a mean rate of 1.8 ± 0.7 mg·kg⁻¹·min⁻¹ over the last 30 min of P1. The HFFD control group displayed a reduction in NHGO but no shift to NHGU until the last sampling point in P1. The fumagillin-treated group also experienced a delay in the response, with no NHGU evident until the last 15 min of P1; the rate during that 15 min period was approximately half that in the Chow group. NHGU increased during P2 in the Chow dogs, averaging 3.4 ± 0.6 mg·kg⁻¹·min⁻¹ during the last 30 min of study. The rate in the HFFD control dogs was significantly less (0.5 ± 0.4 mg·kg⁻¹·min⁻¹ during the final 30 min; P < 0.05 between groups). The response of the HFFD + fumagillin-treated dogs was intermediate between the other two groups (1.7 ± 0.5 mg·kg⁻¹·min⁻¹). When NHGU data in the HFFD and HFFD + fumagillin groups over the course of P1 and P2 were compared by ANOVA, NHGU tended to be enhanced by fumagillin treatments (P = 0.1). Moreover, if the mean data during the last 30 min of P2 are compared with t tests, NHGU was greater (P < 0.05) in HFFD + fumagillin vs. the HFFD group.

Tracer-determined hepatic glucose uptake (HGU; data not shown) yielded similar results, with no significant difference among groups during P1, a blunting of HGU in the HFFD group, and an intermediate response with HFFD + fumagillin treatment. The rate of HGU during P2 was approximately twice as great in the HFFD + fumagillin dogs versus the HFFD group (1.0 ± 0.5 versus 0.5 ± 0.2 mg·kg⁻¹·min⁻¹, respectively), but this did not reach statistical significance (P = 0.19).

Hepatic carbon retention (an indicator of the carbon available for glycogen synthesis) was reduced ~50% by HFFD consumption, compared with Chow (P < 0.05; Fig. 4E). Fumagillin treatment resulted in an intermediate response in regard to hepatic carbon retention, with the rates not significantly different from either of the other groups. On the other hand, nonHGU did not differ among the groups over time (Fig. 4F).

Metabolite and NEFA data.

The lactate, alanine, glycerol, and NEFA concentrations and net hepatic balance data are shown in Table 4. Arterial blood lactate concentrations were similar in all groups during the basal period, and they increased in all groups during P1 and P2. However, the increase was more modest in the HFFD dogs than in the Chow group (P < 0.05 between groups; mean increase during the clamp periods was 1.7-fold basal in the HFFD dogs, compared with 2.3-fold basal in Chow dogs). The magnitude of the increase was very similar in the HFFD + fumagillin (2.2-fold basal) and Chow groups, but there was no significant difference between the HFFD and the HFFD + fumagillin groups. The Chow group exhibited net hepatic lactate uptake in the basal period, shifting to net hepatic lactate output during both clamp periods. On the other hand, both of the groups consuming the HFFD were in a state of net hepatic lactate output (NHLO) in the basal state. The HFFD group exhibited a low rate of NHLO throughout the clamp periods, with the rates significantly different from those in the Chow group. On the other hand, the HFFD + fumagillin group switched to a low rate of net hepatic lactate uptake during the clamp periods. While the rates in HFFD + fumagillin were significantly different from those in the HFFD dogs (P < 0.05), they did not reach significance in comparison to the Chow dogs (P = 0.14).

Table 4.

Arterial blood lactate, alanine, and glycerol and plasma NEFA concentrations, as well as net hepatic balance data, under basal conditions and during the clamp study

Parameter and Group	Basal (−20 to 0 min)	Period 1			Period 2
		30 min	60 min	90 min	120 min	150 min	180 min
Lactate, µmol/L
Chow	295 ± 9	525 ± 92	704 ± 117	691 ± 57	743 ± 51	784 ± 58	877 ± 61
HFFD	241 ± 14	281 ± 19*	331 ± 36*	415 ± 50*	469 ± 50*	483 ± 66*	485 ± 78*
HFFD + FUM	262 ± 26	466 ± 151	628 ± 186	595 ± 161	634 ± 168	614 ± 119	611 ± 130
Net hepatic lactate balance,a µmol·kg−1·min−1
Chow	−6.0 ± 1.0	1.9 ± 4.9	6.3 ± 3.7	5.3 ± 2.6	4.2 ± 2.1	4.2 ± 2.3	1.6 ± 2.0
HFFD	2.7 ± 0.3*	2.4 ± 0.5	1.2 ± 0.6*	1.3 ± 0.5*	1.1 ± 0.6*	1.3 ± 0.6*	2.0 ± 0.4*
HFFD + FUM	4.3 ± 0.3*	−0.6 ± 2.9	−1.3 ± 2.7	−0.5 ± 1.9	−0.7 ± 1.8	−0.7 ± 1.7	0.2 ± 1.5†
Alanine, µmol/L
Chow	276 ± 15	315 ± 22	336 ± 29	337 ± 22	327 ± 19	330 ± 24	328 ± 22
HFFD	226 ± 17	234 ± 15*	214 ± 15*	209 ± 19*	199 ± 17*	193 ± 21*	195 ± 19*
HFFD + FUM	196 ± 18*	221 ± 17*	227 ± 23*	213 ± 22*	205 ± 18*	200 ± 17*	186 ± 17*
Net hepatic alanine uptake, µmol·kg−1·min−1
Chow	2.3 ± 0.2	1.7 ± 0.4	1.8 ± 0.1	1.7 ± 0.1	2.6 ± 0.2	2.2 ± 0.2	3.0 ± 0.3
HFFD	1.1 ± 0.1*	1.3 ± 0.1	1.1 ± 0.1*	1.3 ± 0.1	1.3 ± 0.1*	1.5 ± 0.1*	1.4 ± 0.1*
HFFD + FUM	1.2 ± 0.1*	1.1 ± 0.1*	1.4 ± 0.2	1.3 ± 0.2	1.5 ± 0.2*	1.7 ± 0.2	1.7 ± 0.1*
Glycerol, µmol/L
Chow	81 ± 15	27 ± 2	35 ± 4	31 ± 4	35 ± 5	34 ± 9	32 ± 6
HFFD	87 ± 8	47 ± 4	43 ± 6	51 ± 6	40 ± 5	52 ± 9	44 ± 8
HFFD + FUM	137 ± 13*†	67 ± 5*†	63 ± 7*†	75 ± 8*†	73 ± 4*†	64 ± 6*	70 ± 6*†
Net hepatic glycerol uptake, µmol·kg−1·min−1
Chow	1.9 ± 0.4	0.6 ± 0.2	0.7 ± 0.2	0.7 ± 0.2	0.7 ± 0.2	0.6 ± 0.2	0.7 ± 0.1
HFFD	1.3 ± 0.1	0.6 ± 0.1	0.5 ± 0.1	0.6 ± 0.1	0.5 ± 0.1	0.8 ± 0.2	0.6 ± 0.1
HFFD + FUM	2.6 ± 0.3	1.1 ± 0.1	1.1 ± 0.2†	1.5 ± 0.3*†	1.3 ± 0.1†	1.2 ± 0.1†	1.6 ± 0.1*†
Arterial NEFA, µmol/L
Chow	828 ± 101	219 ± 85	144 ± 24	97 ± 13	91 ± 18	66 ± 9	52 ± 5
HFFD	810 ± 48	281 ± 43	138 ± 44	122 ± 34	87 ± 31	109 ± 35	94 ± 39
HFFD + FUM	866 ± 107	222 ± 54	118 ± 39	129 ± 48	81 ± 20	78 ± 27	96 ± 20
Net hepatic NEFA uptake, µmol·kg−1·min−1
Chow	2.1 ± 1.0	0.7 ± 0.5	0.4 ± 0.1	0.3 ± 0.1	0.2 ± 0.1	0.2 ± 0.2	0.2 ± 0.0
HFFD	2.0 ± 0.3	0.8 ± 0.2	0.1 ± 0.2	0.3 ± 0.1	0.1 ± 0.2	0.2 ± 0.3	0.2 ± 0.0
HFFD + FUM	1.9 ± 0.4	0.4 ± 0.2	0.3 ± 0.2	0.5 ± 0.4	0.2 ± 0.1	0.2 ± 0.2	0.2 ± 0.1

Data are means ± SE. Arterial lactate, alanine, and glycerol are from whole blood samples; nonesterified fatty acids (NEFAs) are from plasma samples; n = 4, 8, and 6 for the Chow [data previously published (8)], high-fat and -fructose diet (HFFD), and HFFD + FUM groups, respectively. Where no symbols are shown, there are no differences between groups.

^aNegative values for balance data indicate net hepatic uptake.

^*P < 0.05 vs. Chow by post hoc analysis.

^†P < 0.05 vs. HFFD by post hoc analysis.

Both the arterial alanine concentrations and net hepatic alanine uptakes were significantly lower in the HFFD and HFFD + fumagillin groups during the basal and clamp periods, compared with the Chow group, and the values did not differ between the HFFD and HFFD + fumagillin groups.

Arterial glycerol concentrations fell from ∼85 µmol/L in the basal period to 32–44 µmol/L by the end of P2 in the Chow and HFFD groups, with no differences between the groups in their responses. The basal period values were greater (P < 0.05) in the HFFD + fumagillin group than in the other groups. Although the HFFD + fumagillin group’s concentrations fell ∼50% below basal during the clamp periods, they remained significantly greater than those in the Chow and HFFD groups throughout P1 and P2. Net hepatic glycerol uptake data followed a pattern similar to those of the of the arterial concentrations, with a decline below basal in all groups, but greater rates of net uptake throughout the basal and clamp periods in the HFFD + fumagillin group (P < 0.05 for HGHG + fumagillin vs. both of the other groups).

Arterial NEFA concentrations were similar among all groups during the basal period, and they declined throughout the clamp periods, reaching 6–11% of basal in all groups by the end of P2, with no differences among the groups. Similarly, the rates of net hepatic NEFA uptake were not different in the basal period and the fell to ∼10% of basal rates by the end of P2; there were no differences in the responses among the three groups.

DISCUSSION

Although a number of medications have been approved for treatment of overweight and obesity, some of them are associated with side effects that limit their use in individuals with relatively common health problems including glaucoma, heart disease, depression, and poorly controlled hypertension (19, 28, 29). Thus the search continues for safe and effective pharmaceutical treatments for the control of body weight. It is worth noting that in a human trial utilizing ZGN-1061, closely related to the agent used in the current studies, no serious side effects were observed (20). The Food and Drug Administration has placed a clinical hold on the Investigative New Drug Application for the first US clinical trial of ZGN-1061 due to the possibility of cardiovascular safety risk based on the findings with the previous MetAP2 inhibitor beloranib.

Fumagillin treatment was associated with a significant reduction in food intake and a strong tendency toward a loss of body weight in the dogs consuming the HFFD on a chronic basis, with a mean decline of 1.9 ± 1.0 kg (P = 0.08), ~5.6% of body weight. In the second set of dogs consuming the HFFD over the shorter term, fumagillin treatment was associated with significantly lower food intake than that evident in the Control animals, with a stabilization of body weight, in contrast to the 5% increase in the Control group. While the changes in body weight were more modest than those reported in rodent models treated with similar agents (1, 18), they are consistent with results in humans treated with MetAP2 inhibitors over a comparable time period (20, 30). Moreover, the canine weight loss was consistent with the percentage achieved by humans in long-term diet and exercise programs designed to reduce body mass (4, 36). Two of the Control dogs in the second set of studies received placebo treatment, and their body weight and food intake changes, as well as their glycemic responses to the OGTTs, did not differ from that of the nonplacebo-treated Control dogs. The lack of metabolic impact is consistent with findings in humans receiving sodium bicarbonate treatment for renal disorders (16, 22, 32).

As anticipated, based on our previous findings (6), the glycemic and insulin responses to the OGTT worsened in all dogs during the first 4 wk of HFFD consumption, before any drug or placebo treatment. While the glycemic response declined slightly but not significantly in the Control group during the period between 4 and 11 wk (placebo or no drug treatment), it fell twice as much in the fumagillin-treated dogs, thus markedly reducing the diet-induced abnormality. Over the same time period, the drug treatment was associated with a 28% increase in the insulin response relative to the glycemic response, compared with <4% increase in the Control dogs. This occurred despite a numeric decline in the insulin response to the OGTT in three of the four fumagillin dogs. Both the reduction in the glycemic response and the enhancement of the insulin response relative to the glycemic response were reversed within 4 wk after treatment ended. The marked improvement in the insulin response relative to the glycemic response strongly suggests improved β-cell function brought about in response to fumagillin treatment, consistent with previous findings with MetAP2 inhibitor treatment (3, 12).

The dog model is particularly valuable in that it makes possible the quantification of hepatic glucose uptake, a process too invasive to be assessed in humans under usual conditions and too technically challenging to carry out in small rodents (17, 23). Thus these studies examined for the first time the impact of a MetAP2 inhibitor on NHGU and nonHGU under controlled conditions. The clamp studies, which allowed simultaneous matching of hormonal and glycemic concentrations, highlighted the differences among the groups. While the fumagillin treatment did not restore NHGU to the rate exhibited by the Chow dogs, it was able to bring about a stimulation of NHGU in comparison to that in the HFFD group that received no pharmaceutical treatment. Clearly there were no differences in insulin concentrations or other hormonal responses that would help to explain the improvement in NHGU. The enhancement of NHGU in the fumagillin-treated animals was associated with a 37% increase in hepatic glycogen storage (hepatic carbon retention), compared with the HFFD group, although this did not reach statistical significance. This enhancement of liver glycogen storage is consistent with that reported in a high-fat-fed rat model treated with a similar compound, compared with pair-fed control animals not given drug treatment (1). With the exception of inhaled or orally administered insulin analogs (11, 24), which remain uncommon in diabetes treatment or still under development (9), it is rare for diabetes medications to enhance NHGU. Subcutaneous delivery of the available insulin products is associated with increased peripheral (muscle and fat) glucose uptake but not with substantial stimulation of liver glucose uptake (10, 27). MetAP2 inhibitors hold promise as a therapy that can enhance liver glucose uptake and concomitantly stimulate hepatic glycogen storage in response to a meal, making them near-unique among the current oral diabetes therapeutic agents.

It is worth noting that glycerol concentrations were greater in the treated dogs during the basal period and throughout the clamp periods, consistent with an increased rate of lipolysis in the treated group. In contrast, NEFA concentrations in the fumagillin group were similar to those in the other two groups, but this is not unexpected, since NEFA but not glycerol can undergo reesterification in the fat tissue (5, 31). Positive effects on lipid metabolism, including reduction in fat cell size and the mass of white adipose tissue, have been reported in rodents and humans receiving MetAP2 inhibitor treatment (1, 14, 15). Moreover, a recent study in obese adults with type 2 diabetes demonstrated that 26 wk of treatment with the MetAP2 inhibitor beloranib significantly reduced triglyceride concentrations and increased HDL cholesterol, without altering total or LDL cholesterol levels (30). In a small group of overweight/obese but otherwise healthy subjects, 4 wk of treatment with fumagillin-1061 tended to reduce LDL cholesterol (20). Thus the impact of MetAP2 inhibitors on lipid metabolism appears promising but remains to be examined more thoroughly.

While glucose tolerance testing is common in clinical care and research, and glucose clamp studies are utilized in human research studies, the administration of glucose alone limits the application of our results. In future studies, it would be of benefit to examine the impact of fumagillin on mixed meal responses, as this would provide physiologic data with the potential to improve care of patients with prediabetes or diabetes. Direct examination of hepatic and pancreatic responses under physiologic conditions in the human is difficult because of the invasiveness of the procedures required. The structure of the liver, as well as the pancreas and islets, is similar in the dog and the human, making the dog a powerful model for these future investigations (17).

Beloranib has been associated with venous thromboembolism in clinical trials (21, 30). However, this problem has not been observed in human and animal trials with fumagillin-1061, closely related to the compound utilized in our canine studies (3, 20). Consistent with this, assessment of blood chemistry, complete blood counts, and clotting factors did not reveal any systematic or adverse changes with fumagillin treatment in the current studies. While continued assessment of the safety of MetAP2 inhibitors is important, the current data with fumagillin are promising in nature.

In conclusion, fumagillin was well tolerated during 8 wk of treatment, and it was associated with significant reduction in food consumption and modest impacts on body weight in dogs consuming a HFFD. It shows promise as a strategy to improve glucose tolerance, as well as hepatic glucose uptake and glycogen synthesis under feeding conditions. Its impact on liver glucose disposition suggests it many serve as a near-unique tool for management of impaired glucose tolerance and diabetes.

GRANTS

This work was supported by Zafgen, Inc. These studies utilized the Metabolic Physiology Shared Resource and the Hormone Assay and Analytical Services Core, supported by National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Grant DK-020593; the Hormone Assay and Analytical Services Core also receives support from NIDDK Grant DK-059637. A.D.C. is the Jacquelyn A. Turner and Dr. Dorothy J. Turner Chair in Diabetes Research.

DISCLOSURES

J.E.V. and T.E.H. are executives and shareholders of Zafgen.

AUTHOR CONTRIBUTIONS

J.E.V., T.E.H., and A.D.C. conceived and designed research; M.C.M., K.C.C., M.S., G.K., and B.F. performed experiments; M.C.M., K.C.C., and A.D.C. analyzed data; M.C.M., K.C.C., J.E.V., and A.D.C. interpreted results of experiments; M.C.M. prepared figures; M.C.M. drafted manuscript; M.C.M., K.C.C., G.K., J.E.V., T.E.H., and A.D.C. edited and revised manuscript; M.C.M., K.C.C., M.S., G.K., J.E.V., T.E.H., B.F., and A.D.C. approved final version of manuscript.