Evaluating insulin secretagogues in a humanized mouse model with functional human islets

ME TAB O L IS M C LI N I CA L A N D E XP E RI ME N TAL 6 2 ( 2 01 3 ) 9 0–9 9

Available online at www.sciencedirect.com

Metabolism www.metabolismjournal.com

Evaluating insulin secretagogues in a humanized mouse model with functional human islets Jian Luo a,⁎, Kathy Nguyen b , Michael Chen b , Than Tran b , Jianqiang Hao c , Bole Tian d , Ingrid C. Rulifson b , Ying Zhang b , Lei Tian e , Yu Zhang f , Edwin Lopez b , Daniel C.-H. Lin b , Yingcai Wang b , Zhihua Ma b , Jonathan Houze b , Zhiguang Guo g,⁎⁎ a

NGM Biopharmaceuticals, Inc., South San Francisco, CA, USA Metabolic Disorders, Amgen Inc., South San Francisco, CA, USA c Department of Surgery, University of British Columbia, Vancouver, BC, Canada d Department of Surgery, West China Hospital, Sichuan University, Chengdu, China e Department of Surgery, First Affiliated Hospital of Guangxi Medical University, Nanning, China f Department of Urology, Dongzhimen Hospital, Beijing University of Chinese Medicine, Beijing, China g Sanford Research/Sanford Health, Department of Pediatrics, University of South Dakota, USA b

A R T I C LE I N FO Article history:

AB S T R A C T Objective. To develop a rapid, easy and clinically relevant in vivo model to evaluate novel

Received 18 May 2012

insulin secretagogues on human islets, we investigated the effect of insulin secretagogues

Accepted 17 July 2012

on functional human islets in a humanized mouse model.

Keywords: Islet Diabetes Human islet grafts Mice Insulin

Materials/Methods. Human islets were transplanted under the kidney capsule of streptozotocin (STZ)-induced diabetic mice with immunodeficiency. Human islet graft function was monitored by measuring non-fasting blood glucose levels. After diabetes was reversed, human islet transplanted mice were characterized physiologically by oral glucose tolerance and pharmacologically with clinically proven insulin secretagogues, glucagon-like peptide-1 (GLP-1), exenatide, glyburide, nateglinide and sitagliptin. Additionally, G protein-coupled receptor 40 (GPR40) agonists were evaluated in this model. Results. Long-term human islet graft survival could be achieved in immunodeficient mice. Oral glucose challenge in human islet transplanted mice resulted in an immediate incremental increase of plasma human C-peptide, while the plasma mouse C-peptide was undetectable. Treatments with GLP-1, exenatide, glyburide, nateglinide and sitagliptin effectively increased plasma human C-peptide levels and improved postprandial glucose concentrations. GPR40 agonists also stimulated human C-peptide secretion and significantly improved postprandial glucose in the human islet transplanted mice.

Abbreviations: GLP-1, glucagon-like peptide-1; GPR40, G protein-coupled receptor 40; STZ, streptozotocin; OGTT, oral glucose tolerance test. Disclosure statement: Jian Luo, Kathy Nguyen, Michael Chen, Than Tran, Ingrid C Rulifson, Ying Zhang, Edwin Lopez, Daniel C. Lin, Yingcai Wang, Zhihua Ma, and Jonathan Houze were employed by and shareholders in Amgen during the study period. Jianqiang Hao, Bole Tian, Lei Tian, Yu Zhang, and Zhiguang Guo were employed by the University of Minnesota during the study period. ⁎ Correspondence to: J. Luo, NGM Biopharmaceuticals, Inc., 630 Gateway Blvd., South San Francisco, CA 94080, USA. Tel.: +1 650 243 5587; fax: +1 650 583 1646. ⁎⁎ Correspondence to: Z. Guo, Sanford Research/Sanford Health, Department of Pediatrics and Surgery, University of South Dakota, 2301 East 60th Street North, Sioux Falls, SD 57104, USA. Tel.: +1 605 312 6030; fax: +1 605 312 6071. E-mail addresses: [email protected] (J. Luo), [email protected] (Z. Guo). 0026-0495/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.metabol.2012.07.010

ME TA BOL I SM C LI N I CA L A N D E XP E RI ME N TAL 6 2 ( 2 01 3 ) 9 0–9 9

91

Conclusions. Our studies indicate that a humanized mouse model with human islet grafts could mimic the in vivo characteristics of human islets and could be a powerful tool for the evaluation of novel insulin secretagogues or other therapeutic agents that directly and/or indirectly target human β cells. © 2013 Elsevier Inc. All rights reserved.

1.

Introduction

Pancreatic β-cell dysfunction is responsible for the pathogenesis and progression of type 2 diabetes [1–3]. Inadequate secretion of insulin is a very early element in the development of type 2 diabetes. This β-cell defect is a consequence of β-cell loss and/or endocrine dysregulation of islet function [2]. Quantitation of insulin sensitivity and the major parameters of β-cell function (glucose sensitivity, rate sensitivity and potentiating factor) in a large number of individuals spanning the range from normal glucose tolerance to overt diabetes demonstrated that impaired β-cell glucose sensitivity is a characteristic feature of even minimally impaired glucose tolerance [4,5]. The use of insulin secretagogues such as sulfonylureas and related ATP-sensitive K+ channel blockers has been a critical part of treatment for patients with type 2 diabetes for a long time [6,7]. With the introduction of incretin based therapies [3,8–10], such as exenatide, a glucagon-like peptide-1 (GLP-1) analogue, and sitagliptin, a dipeptidyl peptidase-4 (DPP-4) inhibitor, there has been continued interest in novel drug targets that stimulate glucose dependent insulin secretion. Recently, several G protein-coupled receptors including GPR40, GPR119 and GPR120 have emerged as possible targets for treating diabetes [11]. Agonists of those GPRs stimulate not only pancreatic β cells to secrete insulin but also gut cells to secrete incretin hormones [12–19]. In drug discovery and academic research, rodent models are commonly used for in vivo evaluation because they are readily available [20,21]. However, studies have shown that there are striking species differences in islets with regard to both cytoarchitecture and function, especially between rodents and primates [22]. For a novel insulin secretagogue, the results generated from rodent models may be misleading. Thus, it is critical to demonstrate a secretagogue's specificity and function in relation to human islets, particularly in an in vivo setting. Transplanting isolated human islets into immunodeficient mice to evaluate their ability to revert the hyperglycemia induced by streptozotocin (STZ) has often been used to assess the quality and function of isolated human islets [23,24]. In light of this method, we explored and established the use of a humanized mouse model with functional human islets as a means for evaluating insulin secretagogues. GLP-1 is an incretin peptide secreted from intestinal L-cells in response to an oral glucose challenge or meal [25–27]. It directly stimulates glucose-dependent insulin secretion from β cells and improves glucose regulation. Exenatide is a currently marketed GLP-1 analog and has the same pharmacological actions as GLP-1 [28,29]. Nateglinide and glyburide are different pharmaceutical agents that directly stimulate insulin secretion from pancreatic islets [30]. Sitagliptin, a recently marketed anti-diabetic drug, indirectly increases

insulin levels by slowing down degradation of GLP-1 through inhibition of DPP-4 [31,32]. Metformin is a glucose lowering agent that reduces plasma glucose by suppression of hepatic glucose production and enhancement of insulin sensitivity in peripheral tissues [33]. GPR40, which is predominantly expressed in pancreatic β cells, regulates free fatty acidinduced insulin secretion [34,35]. GPR40 agonists have been investigated for discovery and development of novel glucose dependent insulin secretagogues [15,36–38]. Using humanized mice with functional human islets, we evaluated both these clinically proven drugs and novel insulin secretagogues on human insulin secretion in response to oral glucose challenge.

2.

Materials and methods

2.1.

Animals

Eight to nine-week-old male T-cell deficient nude mice were purchased from the National Institute of Health. The animals were housed at pathogen-free facilities under a 12-h light, 12h dark cycle and were allowed ad libitum access to regular chow and water. The animal use protocols were approved by the Animal Use and Care Committee at the University of Minnesota (protocol number: 0412A66332) and by Amgen San Francisco Animal Use and Care Committee (protocol number: 11-04r2amd06).

2.2.

Inducing diabetes in nude mice

Nude mice were injected with STZ (Sigma-Aldrich, St. Louis, MO, USA) at 240 mg/kg intravenously via the tail vein under anesthesia. Non-fasting blood glucose levels were checked one week later. Blood glucose levels were monitored with a Glucometer (Elite XL). Only diabetic mice with a non-fasted blood glucose level >400 mg/dL for two consecutive days were qualified and subjected to further procedures. These mice were treated with 0.5 U regular insulin and 0.5 U NPH insulin (Eli Lilly, Indianapolis, IN, USA) intraperitoneally daily to control the blood glucose concentrations until receiving the human islet transplantation.

2.3.

Islet transplantation

Human islets were obtained from the Islet Cell Resource Centers funded by the National Institute of Health and from Prodo Laboratories, Inc. (Irvine, CA, USA). The donor ages were between 30 and 59 years old. The purity of isolated islets was determined by dithizone staining. The viability was determined by a fluorometric cell viability assay using acridine orange and propidium iodide staining. The purity was >70% and the viability was >90%. Recipient mice were anesthetized

92

ME TAB O L IS M C LI N I CA L A N D E XP E RI ME N TAL 6 2 ( 2 01 3 ) 9 0–9 9

with Avertin (250 mg/kg, i.p.). The skin of the left lateral side was shaved and cleaned with Betadine. A 1 to 1.5 cm lumbar incision was opened perpendicular to the axis of the kidney across the left side. The left kidney was carefully pushed out through the lumbar incision using Q-tip. Using two small forceps, a small hole was opened in the lower half of the kidney capsule. A polyethylene tube (PE-50) containing 2000 or 3000 islet equivalent (IE) human islets was inserted beneath the kidney capsule and gently pushed from the lower pole to the upper pole. The human islets were then delivered to the upper pole of the kidney by a Hamilton syringe. The incision was closed by using continuous 5-0 Dexon absorbable suture with tapered needle.

2.4.

Assessment of islet graft function

ID, USA), sitagliptin (Merck, Whitehouse Station, NJ, USA), nateglinide (Novartis, Basel, Switzerland), glyburide (Sigma -Aldrich, St. Louis, MO, USA), metformin (Sigma -Aldrich, St. Louis, MO, USA), and GPR40 agonists, Amgen compounds A and B. Amgen compounds A and B were synthesized at Amgen with >95% purity. The dosage for each of the agents was based on our experience with these agents in rodent models and usually results in>75% of the maximum efficacy in similar settings. All agents were formulated in 1% methylcellulose (Sigma-Aldrich, St. Louis, MO, USA) with 1% Tween® 80 (Sigma-Aldrich, St. Louis, MO, USA). The experiments had a cross-over design. Mice were randomized to receive respective treatments or vehicle control on the testing day and subjected to OGTT. There was a minimum of 4-day rest between test days.

Non-fasting blood glucose levels were measured daily for 7 days following the transplantation and twice a week thereafter. Normoglycemia was defined as a persistent blood glucose level <200 mg/dL, and the first day of cure was the first of 3 consecutive days of <200 mg/dL. Mice that had non-fasting blood glucose <200 mg/dL for 3 consecutive measurements for over 2 weeks were qualified for enrollment for further experiments. A randomly selected group of mice was monitored for 14 weeks to evaluate the stability of the model and was subjected to nephrectomy to remove the left kidney bearing human islet grafts to determine its contribution to the control of the blood glucose level.

2.8.

2.5.

The significance of differences between groups was determined by the Kaplan–Meier analysis or a Student's t test. Data were shown as mean±SD. A value of P<0.05 was considered statistically significant.

High-fat diet in human islet transplanted nude mice

To determine whether obesity could be induced in nude mice with human islet grafts, we fed randomly selected human islet transplanted mice with a high-fat diet that had 60% fat (Research Diets Inc., New Brunswick, NJ). These human islet transplanted mice had been normoglycemic for at least 6 weeks before being given a high-fat diet.

2.6.

Oral glucose tolerance test (OGTT)

Naive nude or human islet transplanted nude mice were fasted for 4 h in the morning. Following baseline blood sampling, the animals were orally administered with glucose at 4 g/kg. Blood samples were collected via tail vein at 5, 15, 30, 60 and 120 min after the glucose load. The plasma samples were used for the measurements of plasma glucose, human Cpeptide and mouse C-peptide.

2.7.

Pharmaceutical agent evaluation

Naive nude or human islet transplanted nude mice were fasted for 4 h in the morning. Following a baseline blood glucose measurement, animals received their respective treatments subcutaneously or orally. The time of dosing is indicated in the figure legend. An oral glucose challenge (4 g/ kg) was administered and blood glucose measurements were taken at 0, 15, 30, 60, and 120 min via tail vein. Additional plasma samples were taken at 15 min for human C-peptide measurement. The treatments included GLP-1 (SigmaAldrich, St. Louis, MO, USA), exenatide (Eli Lilly, Indianapolis,

Plasma glucose and C-peptide assays

A colorimetric glucose oxidation assay using the Autokit Glucose Reagent from Wako Chemicals USA, Inc (Richmond, VA, USA) was used to measure plasma glucose concentrations. Human C-peptide EIA Kit and Mouse C-peptide II EIA Kit were purchased from ALPCO Diagnostics (Salem, NH, USA), and the measurement of plasma human C-peptide and mouse C-peptide levels was conducted according to the manufacturer's protocol.

2.9.

3.

Statistics

Results

3.1. Long-term human islet graft survival in immunodeficient mice Human islets were isolated from 6 pancreas donors. The average donor age was 46±10 years old and BMI was 31±4. The cold ischemia time was 12±3 h. The islet purity was 80%±7% and viability was 93%±2%. At least 8 islet transplantations were performed using 2000 IE or 3000 IE islets from each donor. Reversal of diabetes depended on the dose of transplanted islets. When 2000 IE islets were transplanted into each recipient nude mouse, diabetes was reversed in 21 of 42 recipient mice (50%) at 3 weeks and 22 of 42 recipient mice (52%) at 4 weeks posttransplantation (Fig. 1A). When 3000 IE islets were transplanted into each recipient nude mouse, diabetes was reversed in 25 of 28 recipient mice (82%, P<0.01 vs. 2000 IE islet transplanted mice) at 3 weeks posttransplantation and 26 of 28 recipient mice (93%, P<0.01 vs. 2000 IE islet transplanted mice) at 4 weeks posttransplantation. To determine long-term human islet graft function in recipient nude mice, we monitored blood glucose levels in 15 randomly selected recipient mice with normoglycemia over 14 weeks posttransplantation. Normoglycemia was achieved and maintained in these 15 mice after transplantation (Fig. 1B).

ME TA BOL I SM C LI N I CA L A N D E XP E RI ME N TAL 6 2 ( 2 01 3 ) 9 0–9 9

Mice with Normoglycemia (%)

100

3000 IE (N=28)

90

2000 IE (N=42)

80 70

transplanted mice were significantly higher than in transplanted mice given 2 g/kg (P<0.01). Thus, oral administration of glucose at the dose of 4 g/kg is more suitable for assessing insulin secretagogues in human islet transplanted nude mice.

60 50

3.3. High-fat diets in humanized mice with functional human islets

40 30 20 10 0 0

5 10 15 20 25 Days Post transplantation

A 700 Blood Glucose Level (mg/dL)

93

30

Islet Transplantation

600 500 Islet Grafts Removal

400 300 200

High-fat diets can induce obesity in C57BL/6 mice [39,40]. We determined whether obesity could be induced in human islet transplanted nude mice by feeding them a high-fat diet. The mean body weight in these mice (N=10) was 30.7±5.2 g before the diet and was 33.4±3.2 g after 8 weeks on a high-fat diet, which was not significantly different from the control mice that were on a regular diet (Fig. 2B). The body weight was 31.1± 4.0 g before and 32.6 ±3.2 g after 8 weeks on regular diet (N=8, P<0.05). A significant difference in blood glucose level was also not found before and after the high-fat diet. The mean blood glucose level in the high-fat diet fed mice was 108± 31 mg/dL before and 117±20 mg/dL after 8 weeks on the highfat diet. Thus, high-fat diet could not induce obesity in nude mice with human islet grafts.

100 0 0

B

1

2

3

4

5

6

7

8

9 10 11 12 13 14 15

3.4. Postprandial C-peptide profile in humanized mice with functional human islets

Weeks After Transplantation

Fig. 1 – (A) Reversal of diabetes in nude mice after human islet transplantation. Total 2000 IE or 3000 IE human islets were transplanted into each recipient mouse. (B) Blood glucose levels in human islet transplanted nude mice before and after transplantation. Recipient nude mice (n=15) were transplanted with isolated human islets under the left kidney capsule and the nephrectomy was performed to remove the islet grafts at week 14 posttransplantation.

To determine whether maintaining normoglycemia was due to human islet graft function, we performed the nephrectomy to remove the left kidney bearing human islet grafts. All recipient mice returned to hyperglycemia 1 week after the nephrectomy which confirmed human islet graft function. Histological analysis of the islet grafts was performed in selected mice and presence of the islet grafts under the kidney capsule was confirmed (data not shown).

3.2. OGTT in humanized mice with functional human islets Recipient nude mice that received 2000 IE human islets and had persistent normoglycemia for 2 weeks were used for evaluating OGTT. When mice were given glucose at 2 g/kg orally, significant difference in blood glucose levels was not found between human islet transplanted mice and naïve mice, although both human islet transplanted mice and naïve mice had significantly lowered blood glucose levels after oral glucose challenges at 30, 60, 90, and 120 min, as compared with the control diabetic mice without human islet grafts (Fig. 2A, P<0.01). However, when we increased the dose of glucose to 4 g/kg, blood glucose levels at 15 and 30 min in

To better characterize nude mice with human islet grafts, plasma human and mouse C-peptide levels were analyzed before and after oral glucose challenge and the results were compared with naive nude mice. The C-peptide assay was highly specific for the intended species. Few mouse C-peptide 2 could be detected in nude mice with human islet grafts, further confirming that the STZ effectively eliminated the majority of mouse β cells. As expected, there was no detection of human C-peptide in naive nude mice. In both naive nude mice (N=12) and nude mice with human islet grafts (N=18), immediate incremental increases of C-peptide following oral glucose challenge were observed with the C-peptide levels peaking at 15 min (Fig. 3A). The basal and postprandial human C-peptide levels were significantly higher in nude mice with human islet grafts than in naive nude mice (P<0.01). In association with the observation in plasma human C-peptide, the baseline glucose and the postprandial mouse C-peptide levels in nude mice with human islet grafts were significantly lower than in naive nude mice (P<0.01). Thus, nude mice with human islet grafts secreted human insulin in response to the oral glucose challenge.

3.5. Pharmaceutical agent evaluation in humanized mice with functional human islets To determine whether GLP-1 and exenatide improved OGTT in nude mice with human islet grafts, we subcutaneously (s.c.) gave GLP-1 and exenatide at 15 min before oral administration of 4 g/kg glucose. A significant increase in human C-peptide was observed in mice treated with 100 μg/kg GLP-1, compared to untreated control mice at 15 min after glucose challenge ( Fig. 4A, P<0.05). This change was associated with significant improvement of the postprandial glucose (Fig. 4B, P<0.05). Almost identical findings were seen with 10 μg/kg exenatide

94

ME TAB O L IS M C LI N I CA L A N D E XP E RI ME N TAL 6 2 ( 2 01 3 ) 9 0–9 9

700

Glucose Level (mg/dL)

600 500

Naive Mice with 2 g Glucose (N=8)

400

Diabetic Mice with 2 g Glucose (N=8) Transplanted Mice with 2 g Glucose (N=10)

300 Transplanted Mice with 4 g Glucose (N=12)

200 100 0 0

A

15

30 60 Time (Min)

120

40 35

Body Weight (g)

30 25 Before Diet 8 Weeks After Diet

20 15 10 5 0

B

Regular Diet (N=8) High-Fat Diet (N=9)

Fig. 2 – (A) Oral glucose tolerance in human islet transplanted nude mice and naive nude mice with a dose of 2 g/kg or 4 g/kg glucose. (B) Body weight change in human islet transplanted nude mice before and after feeding a high-fat diet for 8 weeks. Data were shown as mean±SD.

treatment (Fig. 4C and 4D). The significant elevation of human C-peptide resulting from exenatide treatment was also associated with clear postprandial glucose improvement (P<0.05). HIT mice - human C-peptide

1000

Plasma C-peptide (pM)

* *

750

*

*

HIT mice - mouse C-peptide 2 Naive mice - human C-peptide Naive mice - mouse C-peptide 2

*

500

*

250

0

Time (min) Fig. 3 – Plasma human and mouse C-peptide concentrations in human islet transplanted nude mice (N=12) and in naive mice (N=18) before and after orally administering 4 g/kg glucose (*P<0.01). An equal volume of plasma from 2 animals of the same group was pooled to increase the volume for the measurements. Data were shown as mean±SD.

To determine whether clinically approved oral insulin secretagogues improved OGT in nude mice with human islet grafts, we orally treated mice with 50 mg/kg nateglinide, or 10 mg/kg sitagliptin, or 10 mg/kg glyburide, at 60 min before oral administration of 4 g/kg glucose. All three drugs increased plasma human C-peptide to various degrees (Fig. 5A) following a glucose challenge and effectively improved postprandial glucose concentrations. Nateglinide seemed to be the most effective at the dosage used (Fig. 5B). However, the experiment was not designed to compare efficacy between agents. We orally treated humanized nude mice with 125 mg/kg metformin at 60 min before oral administration of 4 g/kg glucose. We found that metformin significantly improved postprandial glucose levels but did not affect plasma human C-peptide concentration (Fig. 6 A and 6B, P<0.05). Thus, metformin could improve OGT but not through stimulation of insulin secretion. Amgen compound A is a specific GPR40 agonist with 28 nM EC50 in an aequorin assay using GPR40 transfected CHO cells. We orally treated humanized nude mice with 10 mg/kg Amgen compound A at 60 min before oral administration of 4 g/kg glucose. In response to oral treatment of Amgen compound A, plasma human C-peptide in human islet transplanted nude mice was significantly higher at 15 min post glucose challenge when compared to vehicle control (Fig. 7A, P < 0.05).

95

ME TA BOL I SM C LI N I CA L A N D E XP E RI ME N TAL 6 2 ( 2 01 3 ) 9 0–9 9

*

3000

Control

450

GLP-1 100 µg/kg s.c.

400

2500 2000 1500 1000 500

Blood Glucose (mg/dL)

Human C-peptide (pM)

3500

350 300 250 200 150

1500

Control Exenatide 10 µg/kg s.c.

Human C-peptide (pM)

* 1250 1000 750 500 250

*

0 -15

0

15

30

45

400

60

75

90

105

120

Control Exenatide 10 µg/kg s.c.

300

200

*

100

15 min.

* *

*

0 -15

0

C

*

50

B

Blood Glucose (mg/dL)

A

*

100

0

15 min.

Control GLP-1 100 µg/kg s.c.

0

15

D

30

45

60

* 75

90

105

120

Time (min)

Fig. 4 – Oral glucose tolerance in human islet transplanted nude mice treated with GLP-1 or exenatide. (A) Human C-peptide concentrations in human islet transplanted nude mice with either vehicle (N=8) or GLP-1 (N=8) at 15 min after glucose challenge. (B) Blood glucose concentration with either vehicle (N=8) or GLP-1 (N=8). (C) Human C-peptide concentrations with either vehicle (N=13) or exenatide (n=13) at 15 min post glucose challenge. (D) Glucose concentration with either vehicle (N=13) or exenatide (N=13). Data were shown as mean±SD. *P<0.05, treatment vs. vehicle.

Accordingly, postprandial glucose levels in human islet transplanted nude mice were significantly lower with Amgen compound A treatment (Fig. 7B, P<0.05). Amgen compound B is a GPR40 agonist selective for the human GPR40 receptor. In an aequorin assay using human or

mouse GPR40, Amgen compound B was ~50× less potent on mouse GPR40 (EC50 of 6 nM and 280 nM on human and mouse GPR40, respectively). In addition, the maximal activity of Amgen compound B was identical to that of Amgen compound A on the human receptor, but only 50% of that of

Vehicle Netaglinide 50 mg/kg p.o. Sitagliptin 10 mg/kg p.o. Glyburide 10 mg/kg p.o.

Vehicle

2000

Sitagliptin 10 mg/kg p.o. Glyburide 10 mg/kg p.o.

* 1000

500

0.08

1500

Blood Glucose (mg/dL)

* Plasma C-peptide (pM)

400

Netaglinide 50 mg/kg p.o.

0

*

B

*

200

*

100

0 -60

15 min.

A

300

*

* *

*

* -30

0

30

* 60

* 90

*

120

Time (min)

Fig. 5 – Oral glucose tolerance in human islet transplanted nude mice treated with nateglinide, sitagliptin or glyburide. The testing agents were administered by gavage at 60 min before glucose challenge. Glucose (4 g/kg) was administered by gavage at 0 min. (A) Human C-peptide concentrations in human islet transplanted nude mice treated with vehicle, or nateglinide, or sitagliptin, or glyburide at 15 min post glucose challenge. (B) Blood glucose concentration in human islet transplanted nude mice treated with vehicle, or nateglinide, or sitagliptin or glyburide. There were 13 mice in each treated group. Data were shown as mean±SD. *P<0.05, treatment vs. vehicle.

96

ME TAB O L IS M C LI N I CA L A N D E XP E RI ME N TAL 6 2 ( 2 01 3 ) 9 0–9 9

Vehicle

Vehicle

400

Metformin125 mg/kg p.o. 1000

500

Blood Glucose (mg/dL)

Plasma C-peptide (pM)

1500

Metformin 125 mg/kg p.o. 300

200

*

100

0 -60

0

15 min.

-30

0

B

A

30

60

90

120

Time (min)

Fig. 6 – Oral glucose tolerance in human islet transplanted nude mice treated with metformin. The testing agent was administered by gavage at 60 min before glucose challenge. Glucose (4 g/kg) was administered by gavage at 0 min. (A) Human C-peptide concentrations in human islet transplanted nude mice with either vehicle or metformin at 15 min post glucose challenge. (B) Blood glucose concentration in human islet transplanted nude mice treated with either vehicle or metformin. There were 13 mice in each treated group. Data were shown as mean±SD. *P<0.05, treatment vs. vehicle.

Amgen compound A on the mouse receptor. We orally treated human islet transplanted nude mice with 1 or 10 mg/kg Amgen compound B at 60 min before oral administration of Vehicle

350

Compound A 10 mg/kg

300

Blood Glucose (mg/dL)

Human C-peptide (pM)

2000

4 g/kg glucose. Oral administration of Amgen compound B dose-dependently reduced postprandial glucose concentrations in human islet transplanted nude mice (Fig. 7C) but

*

1500

1000

500

0

Vehicle Compound A 10 mg/kg

250 200 150 100

*

50

*

0

-60

15 min

A

-30

B

*

*

0

30 60 Time (min)

90

120

Vehicle Compound B 1 mg/kg Compound B 10 mg/kg

300

* *

200

100

* * *

0

-60

C

-30

0

*

Compound B 10 mg/kg 300

200

100

*

*

60 30 Time (min)

Vehicle

400

Blood Glucose (mg/dL)

Blood Glucose (mg/dL)

400

* 90

0

120

-60

D

-30

0

30 60 Time (min)

90

120

Fig. 7 – Oral glucose tolerance in human islet transplanted nude mice treated with GPR40 agonists, Amgen compound A and Amgen compound B. The testing agent was administered by gavage at 60 min before glucose challenge. Glucose (4 g/kg) was administered by gavage at 0 min. (A) Human C-peptide concentrations in human islet transplanted nude mice treated with either vehicle (n=8) or Amgen compound A (n=8) at 15 min post glucose challenge. (B) Blood glucose concentration in human islet transplanted nude mice treated with either vehicle (n=8) or Amgen compound A at 10 mg/kg (n=8). (C) Blood glucose concentration in human islet transplanted nude mice treated with either vehicle (n=8) or Amgen compound B at 1 mg/kg (n=8) or Amgen compound B at 10 mg/kg (n=8). (D) Blood glucose concentration in naive nude mice treated with either vehicle (n=6) or Amgen compound B 10 mg/kg (n=6). Data were shown as mean±SD. *P<0.05, treatment vs. vehicle.

ME TA BOL I SM C LI N I CA L A N D E XP E RI ME N TAL 6 2 ( 2 01 3 ) 9 0–9 9

not in naive nude mice (Fig. 7D). These results confirmed that human islets were specifically responding to GPR40 agonist treatment.

4.

Discussion

Although human islet transplantation in immunodeficient mice with diabetes has been used as a bioassay to evaluate human islets for clinical transplantation, using this model for evaluating the effect of anti-diabetic drugs on human insulin secretion has not been reported. We found that human islet grafts could survive long-term in immunodeficient nude mice. We characterized the physiological response and evaluated the efficacy of both clinically proven and novel insulin secretagogues in response to oral glucose challenge in this model. Our results show that transplanted human islets are fully functional in that immediate incremental increases of human C-peptide were exhibited following glucose challenge and postprandial glucose levels were well controlled. The transplanted human islets were exclusively responsible for the control of the glucose as there was a sharp increase in blood glucose following removal of the human islet grafts. Measuring the plasma levels of C-peptide is a direct reflection of insulin levels in the circulation. There was few detectable mouse C-peptide in the circulation of the humanized nude mice. Obesity is a major cause of insulin resistance that contributes to an increased demand on insulin secretion. Establishing an obese mouse model with human islet grafts will be very valuable for investigating human islet function and for evaluating novel insulin secretagogues in the setting of obesity. Although high-fat diets can induce obesity in C57BL/6 mice [39,40], we found that the body weight did not significantly increase in humanized mice on a high-fat diet for 8 weeks. Our data confirmed an earlier study showing that germ-free mice are resistant to high fat-diet-induced obesity [41]. We are currently exploring other approaches to establish an obese mouse model with human islet grafts. Although we did not determine the levels of triglycerides, free fatty acids, cholesterol, leptin and adiponectin in this model, it will be very interesting to determine these metabolic changes and compare to an obese mouse model with human islet grafts. Because our humanized mice with human islet grafts responded well to OGTT, we evaluated their responses to pharmacological treatments with clinically proven or commercially available insulin secretagogues. We found that all five insulin secretagogues (GLP-1, exenatide, glyburide, nateglinide and sitagliptin) could stimulate human C-peptide secretion in the humanized mice. In other words, the agents directly or indirectly acted on the β-cells of the human islet grafts and stimulated secretion of the insulin. With the increase in insulin secretion, all five agents significantly reduced postprandial glucose concentrations. Conversely, metformin, a glucose lowering agent which does not act on β-cells to stimulate insulin secretion, failed to induce Cpeptide secretion despite a reduction of postprandial glucose. These results demonstrate that humanized mice with human islets are responsive to clinically proven pharmacological agents despite their action through different receptors or

97

mechanisms to stimulate insulin secretion and consequently improve the control of plasma glucose. The GPR40 receptor is a recently emerging target being investigated with respect to glucose dependent insulin secretagogues. Efforts in the pharmaceutical industry have led to the discovery of GPR40 agonists [15,36–38]. These agonists have been shown to effectively stimulate insulin secretion in both in vitro and in vivo rodent systems. The GPR40 receptor appears to be a viable and amenable target for small molecule development for use in treating type 2 diabetes. However, the validity of this target in humans will not be established until data are available to demonstrate that these agonists stimulate insulin secretion in human subjects. The results generated in the current studies establish that GPR40 agonists (Amgen compound A and B) act on human islets and stimulate insulin secretion as measured by human C-peptide in humanized mice. Furthermore, the enhancement of insulin secretion by this agonist was robust enough to lead to significantly better control of postprandial glucose. These findings provide strong evidence that continuing efforts may lead to the development of a GPR40 agonist with therapeutic utility. It is common practice in discovery to screen and optimize compounds against human selective receptors in an in vitro setting [42]. Selected compounds will eventually be evaluated in an in vivo system. In some cases, selected compounds such as DPP-4 inhibitors can only be evaluated in vivo. In this study, we show that the humanized mouse model with human islets is a specific and effective system that may be utilized to evaluate human-specific insulin secretagogues targeting islets. Compounds with no or minimum activity on rodent receptors can be assessed by using this model. Indeed, a human selective GPR40 agonist, Amgen compound B, showed great activity in humanized nude mice but no activity in naive nude mice. In addition to physiological and pharmacological evaluation, model stability is also a critical element for an animal model. Consistent with an earlier study [43], our data established that humanized mice with human islet grafts were very stable over the 14 week observation period until removal of the islet grafts. The stability of the model allows much greater flexibility in experiment design and generation of consistent results. Although some anti-diabetic drugs such as GLP-1 analogs and DPP-4 inhibitors can stimulate β-cell regeneration in rodent models [44–46], in vivo evaluation of human β-cell regeneration at the cellular and molecular levels in response to pharmacological agents is not feasible because of ethical and technical constraints. We and others have demonstrated that pharmacological agents can stimulate human β-cell regeneration in humanized mice [47–49]. Therefore, a humanized mouse model will also provide a powerful research tool for the in vivo study of human β-cell regeneration. In summary, we characterized and established humanized mice with human islets as a model for evaluating novel insulin secretagogues. We demonstrated that transplanted human islets are physiologically functional and pharmacologically responsive to clinically proven insulin secretagogues. Although establishing this humanized mouse model requires islet transplantation and availability

98

ME TAB O L IS M C LI N I CA L A N D E XP E RI ME N TAL 6 2 ( 2 01 3 ) 9 0–9 9

of human islets, it has great value for evaluating clinical candidate compounds that directly or indirectly target human islets. As a great tool for better understanding human islet biology, it could also have potential implication for providing insight into pathways, targets, and biochemical intervention points for improving human islet function.

Funding This study was funded by Amgen Inc.

[15]

[16]

[17]

[18] [19]

Author contributions Zhiguang Guo and Jian Luo designed the experiments, analyzed the data, and wrote the manuscript. Kathy Nguyen, Michael Chen, Than Tran, Jianqiang Hao, Bole Tian, Ingrid C Rulifson, Ying Zhang, Lei Tian, Yu Zhang, Edwin Lopez, Daniel C. Lin, Yingcai Wang, Zhihua Ma, and Jonathan Houze performed experiments and collected the data.

[20]

[21] [22]

[23] REFERENCES [24] [1] Prentki M, Nolan CJ. Islet beta cell failure in type 2 diabetes. J Clin Invest 2006;116:1802–12. [2] Wajchenberg BL. beta-Cell failure in diabetes and preservation by clinical treatment. Endocr Rev 2007;28:187–218. [3] Ashcroft FM, Rorsman P. Diabetes mellitus and the beta cell: the last ten years. Cell 2012;148:1160–71. [4] Ferrannini E, Gastaldelli A, Miyazaki Y, et al. beta-Cell function in subjects spanning the range from normal glucose tolerance to overt diabetes: a new analysis. J Clin Endocrinol Metab 2005;90:493–500. [5] Festa A, Williams K, D'Agostino Jr R, et al. The natural course of beta-cell function in nondiabetic and diabetic individuals: the Insulin Resistance Atherosclerosis Study. Diabetes 2006;55:1114–20. [6] Doyle ME, Egan JM. Pharmacological agents that directly modulate insulin secretion. Pharmacol Rev 2003;55:105–31. [7] Rendell M. The role of sulphonylureas in the management of type 2 diabetes mellitus. Drugs 2004;64:1339–58. [8] Ahren B, Schmitz O. GLP-1 receptor agonists and DPP-4 inhibitors in the treatment of type 2 diabetes. Horm Metab Res 2004;36:867–76. [9] Drucker DJ, Nauck MA. The incretin system: glucagon-like peptide-1 receptor agonists and dipeptidyl peptidase-4 inhibitors in type 2 diabetes. Lancet 2006;368:1696–705. [10] Cernea S, Raz I. Therapy in the early stage: incretins. Diabetes Care 2011;34(Suppl 2):S264–71. [11] Ahren B. Islet G protein-coupled receptors as potential targets for treatment of type 2 diabetes. Nat Rev Drug Discov 2009;8: 369–85. [12] Jones RM, Leonard JN, Buzard DJ, et al. GPR119 agonists for the treatment of type 2 diabetes. Expert Opin Ther Pat 2009;19: 1339–59. [13] Chu ZL, Jones RM, He H, et al. A role for beta-cell-expressed G protein-coupled receptor 119 in glycemic control by enhancing glucose-dependent insulin release. Endocrinology 2007;148:2601–9. [14] Chu ZL, Carroll C, Alfonso J, et al. A role for intestinal endocrine cell-expressed g protein-coupled receptor 119 in glycemic control by enhancing glucagon-like peptide-1 and

[25] [26] [27]

[28] [29]

[30]

[31] [32] [33] [34]

[35]

[36]

[37]

[38]

[39]

glucose-dependent insulinotropic peptide release. Endocrinology 2008;149:2038–47. Tan CP, Feng Y, Zhou YP, et al. Selective small-molecule agonists of G protein-coupled receptor 40 promote glucosedependent insulin secretion and reduce blood glucose in mice. Diabetes 2008;57:2211–9. Gao J, Tian L, Weng G, et al. Stimulating beta cell replication and improving islet graft function by GPR119 agonists. Transpl Int 2011;24:1124–34. Hirasawa A, Tsumaya K, Awaji T, et al. Free fatty acids regulate gut incretin glucagon-like peptide-1 secretion through GPR120. Nat Med 2005;11:90–4. Hansen HS, Rosenkilde MM, Holst JJ, et al. GPR119 as a fat sensor. Trends Pharmacol Sci 2012 [Epub ahead of print]. Gao J, Tian L, Weng G, et al. Stimulating beta-cell replication and improving islet graft function by AR231453, A gpr119 agonist. Transplant Proc 2011;43:3217–20. Luo J, Quan J, Tsai J, et al. Nongenetic mouse models of noninsulin-dependent diabetes mellitus. Metabolism 1998;47: 663–8. Shafrir E. Animal models of non-insulin-dependent diabetes. Diabetes Metab Rev 1992;8:179–208. Cabrera O, Berman DM, Kenyon NS, et al. The unique cytoarchitecture of human pancreatic islets has implications for islet cell function. Proc Natl Acad Sci USA 2006;103: 2334–9. Ricordi C, Gray DW, Hering BJ, et al. Islet isolation assessment in man and large animals. Acta Diabetol Lat 1990;27:185–95. Hering BJ, Kandaswamy R, Harmon JV, et al. Transplantation of cultured islets from two-layer preserved pancreases in type 1 diabetes with anti-CD3 antibody. Am J Transplant 2004;4:390–401. Drucker DJ. Glucagon-like peptides. Diabetes 1998;47:159–69. Drucker DJ. Minireview: the glucagon-like peptides. Endocrinology 2001;142:521–7. Aaboe K, Krarup T, Madsbad S, et al. GLP-1: physiological effects and potential therapeutic applications. Diabetes Obes Metab 2008;10:994–1003. Barnett A. Exenatide. Expert Opin Pharmacother 2007;8: 2593–608. Crasto W, Khunti K, Davies MJ. An update on exenatide, a novel therapeutic option for patients with type 2 diabetes. Drugs Today 2011;47:839–56. Fonseca VA, Kelley DE, Cefalu W, et al. Hypoglycemic potential of nateglinide versus glyburide in patients with type 2 diabetes mellitus. Metabolism 2004;53:1331–5. Drucker D, Easley C, Kirkpatrick P. Sitagliptin. Nat Rev Drug Discov 2007;6:109–10. Lyseng-Williamson KA. Sitagliptin. Drugs 2007;67:587–97. Hundal RS, Inzucchi SE. Metformin: new understandings, new uses. Drugs 2003;63:1879–94. Briscoe CP, Tadayyon M, Andrews JL, et al. The orphan G protein-coupled receptor GPR40 is activated by medium and long chain fatty acids. J Biol Chem 2003;278:11303–11. Itoh Y, Kawamata Y, Harada M, et al. Free fatty acids regulate insulin secretion from pancreatic beta cells through GPR40. Nature 2003;422:173–6. Rayasam GV, Tulasi VK, Davis JA, et al. Fatty acid receptors as new therapeutic targets for diabetes. Expert Opin Ther Targets 2007;11:661–71. Doshi LS, Brahma MK, Sayyed SG, et al. Acute administration of GPR40 receptor agonist potentiates glucose-stimulated insulin secretion in vivo in the rat. Metabolism 2009;58: 333–43. Lin DC, Zhang J, Zhuang R, et al. AMG 837: a novel GPR40/FFA1 agonist that enhances insulin secretion and lowers glucose levels in rodents. PLoS One 2011;6:e27270. Surwit RS, Kuhn CM, Cochrane C, et al. Diet-induced type II diabetes in C57BL/6J mice. Diabetes 1988;37:1163–7.

ME TA BOL I SM C LI N I CA L A N D E XP E RI ME N TAL 6 2 ( 2 01 3 ) 9 0–9 9

[40] Winzell MS, Ahren B. The high-fat diet-fed mouse: a model for studying mechanisms and treatment of impaired glucose tolerance and type 2 diabetes. Diabetes 2004;53(Suppl. 3): S215–9. [41] Backhed F, Manchester JK, Semenkovich CF, et al. Mechanisms underlying the resistance to diet-induced obesity in germ-free mice. Proc Natl Acad Sci USA 2007;104:979–84. [42] Landro JA, Taylor IC, Stirtan WG, et al. HTS in the new millennium: the role of pharmacology and flexibility. J Pharmacol Toxicol Methods 2000;44:273–89. [43] Deng S, Vatamaniuk M, Huang X, et al. Structural and functional abnormalities in the islets isolated from type 2 diabetic subjects. Diabetes 2004;53:624–32. [44] Xu G, Stoffers DA, Habener JF, Bonner-Weir S. Exendin-4 stimulates both beta-cell replication and neogenesis, resulting in increased beta-cell mass and improved glucose tolerance in diabetic rats. Diabetes 1999;48:2270–6. [45] Tian B, Hao J, Zhang Y, et al. Upregulating CD4+CD25+FOXP3+ regulatory T cells in pancreatic lymph nodes in diabetic NOD

[46]

[47]

[48]

[49]

99

mice by adjuvant immunotherapy. Transplantation 2009;87: 198–206. Tian L, Gao J, Hao J, et al. Reversal of new-onset diabetes through modulating inflammation and stimulating beta-cell replication in nonobese diabetic mice by a dipeptidyl peptidase IV inhibitor. Endocrinology 2010;151:3049–60. Suarez-Pinzon WL, Lakey JR, Rabinovitch A. Combination therapy with glucagon-like peptide-1 and gastrin induces beta-cell neogenesis from pancreatic duct cells in human islets transplanted in immunodeficient diabetic mice. Cell Transplant 2008;17:631–40. Tian L, Gao J, Weng G, et al. Comparison of exendin-4 on betacell replication in mouse and human islet grafts. Transpl Int 2011;24:856–64. Suarez-Pinzon WL, Rabinovitch A. Combination therapy with a dipeptidyl peptidase-4 inhibitor and a proton pump inhibitor induces beta-cell neogenesis from adult human pancreatic duct cells implanted in immunodeficient mice. Cell Transplant 2011;20:1343–9.