`
`Review
`
`www.elsevier.com/locate/regpep
`
`Pharmacology of exenatide (synthetic exendin-4): a potential therapeutic
`for improved glycemic control of type 2 diabetes
`
`Loretta L. Nielsen 1, Andrew A. Young, David G. Parkes*
`
`Amylin Pharmaceuticals Inc., 9360 Town Centre Dr. Ste 110, San Diego, CA 92121, USA
`
`Received 12 June 2003; received in revised form 7 October 2003; accepted 14 October 2003
`
`Abstract
`
`to
`Exenatide (synthetic exendin-4), glucagon-like peptide-1 (GLP-1), and GLP-1 analogues have actions with the potential
`significantly improve glycemic control in patients with diabetes. Evidence suggests that these agents use a combination of mechanisms
`which may include glucose-dependent stimulation of insulin secretion, suppression of glucagon secretion, enhancement of h-cell mass,
`slowing of gastric emptying, inhibition of food intake, and modulation of glucose trafficking in peripheral tissues. The short in vivo
`half-life of GLP-1 has proven a significant barrier to continued clinical development, and the focus of current clinical studies has shifted
`to agents with longer and more potent in vivo activity. This review examines recent exendin-4 pharmacology in the context of several
`known mechanisms of action, and contrasts exendin-4 actions with those of GLP-1 and a GLP-1 analogue. One of the most provocative
`areas of recent research is the finding that exendin-4 enhances h-cell mass, thereby impeding or even reversing disease progression.
`Therefore, a major focus of this is article an examination of the data supporting the concept that exendin-4 and GLP-1 may increase h-
`cell mass via stimulation of h-cell neogenesis, stimulation of h-cell proliferation, and suppression of h-cell apoptosis.
`D 2003 Elsevier B.V. All rights reserved.
`
`Keywords: Pharmacology; Exenatide; Type 2 diabetes
`
`1. Introduction
`
`Exendin-4, the naturally occurring form of exenatide
`(synthetic exendin-4; AC2993), was originally isolated from
`the salivary secretions of the lizard Heloderma suspectum
`(Gila monster; Fig. 1) [1]. In the Gila monster, exendin-4
`circulates after the lizard bites down on its prey (ingestion of a
`meal) and thus represents the first example of an endocrine
`hormone secreted from salivary glands. [2]. It is unknown
`whether exendin-4 has a role in fuel homeostasis in the Gila
`monster [2]. Exendin-4 has a 53% amino acid sequence
`overlap with mammalian glucagon-like peptide-1 (GLP-1).
`In mammals, GLP-1 is processed from the proglucagon gene
`in L-cells in the small intestine [3]. Exendin-4 is transcribed
`from a distinct gene, not the Gila monster homologue of the
`mammalian proglucagon gene from which GLP-1 is
`expressed [4]. In mammals, exendin-4 is resistant to degra-
`dation by dipeptidyl peptidase-IV (DPP-IV) and has a much
`
`* Corresponding author. Tel.: +1-858-642-7290; fax: +1-858-334-1290.
`E-mail addresses: lnielsen@amylin.com (L.L. Nielsen),
`dparkes@amylin.com (D.G. Parkes).
`1 Alternate contact: Tel.: +1-858-642-7203; fax: +1-858-334-1203.
`
`0167-0115/$ - see front matter D 2003 Elsevier B.V. All rights reserved.
`doi:10.1016/j.regpep.2003.10.028
`
`longer plasma half-life than GLP-1, which is degraded by
`DPP-IV with a half-life of less than 2 min [5,6].
`Exendin-4 is not an analogue of GLP-1. In other words,
`the structure of the synthetic exendin-4 peptide (exenatide)
`was not created by sequential modification of the structure
`of GLP-1. However, exendin-4 and GLP-1 do share many
`glucoregulatory actions which may be mediated by the
`known pancreatic GLP-1 receptor [7]. Glucoregulatory
`actions of exendin-4 include glucose-dependent enhance-
`ment of insulin secretion [8 – 11], glucose-dependent sup-
`pression of inappropriately high glucagon secretion [10,12],
`slowing of gastric emptying [10,13] which may be para-
`doxically accelerated in people with diabetes [14], and
`reduction of food intake ([15,16]; Fig. 2). In addition,
`exendin-4 has been shown to promote h-cell proliferation
`and islet neogenesis from precursor cells in both in vitro and
`in vivo models [17 – 19]. These glucoregulatory actions of
`exendin-4, combined with enhanced pharmacokinetics, re-
`sult in very high in vivo potency relative to native GLP-1
`[11,20,21]. The putative mechanisms of action of exendin-4
`are compared and contrasted with the actions of GLP-1 and
`a long-acting GLP-1 analogue in the following sections.
`One of the most provocative areas of recent research is
`
`SUN EXHIBIT 1037, IPR2024-00107, PAGE 1
`
`
`
`78
`
`L.L. Nielsen et al. / Regulatory Peptides 117 (2004) 77–88
`
`Fig. 1. Comparison of amino acid sequences for exendin-4, mammalian GLP-1, and Gila monster GLP-1.
`
`based on observations that exendin-4 may improve h-cell
`mass, thereby impeding or even reversing disease progres-
`sion. Therefore, a major focus of this article will be to
`examine in detail
`the published reports supporting the
`concept that exendin-4 and GLP-1 may increase h-cell mass
`via stimulation of h-cell neogenesis, stimulation of h-cell
`proliferation, and suppression of h-cell apoptosis.
`The onset of type 2 diabetes is characterized by the
`emergence of postprandial (post-meal) hyperglycemia and
`subsequently, fasting hyperglycemia [22]. In most individ-
`uals, hyperglycemia results from a failure of pancreatic h-
`cells to secrete adequate insulin to compensate for insulin-
`resistance in peripheral
`tissues [23,24]. The fraction of
`glycosylated hemoglobin (A1C) in circulating red blood
`cells provides an accurate indicator of average glucose
`concentrations in the blood for the previous 3 months.
`A1C levels in healthy humans typically comprises 5 – 6%
`of total hemoglobin, while A1C values in people with
`poorly controlled diabetes generally exceed 9% [25].
`
`Results from the United Kingdom Prospective Diabetes
`Study (UKPDS) showed that a reduction in A1C was
`associated with a reduced risk of vascular complications,
`and also reaffirmed that type 2 diabetes is a progressive
`disease characterized by a continuous loss of h-cell function
`that current therapies cannot rectify. [26,27]. Exenatide is
`the USAN generic drug name for synthetic exendin-4, an
`investigational therapeutic being studied by Amylin Phar-
`maceuticals in partnership with Eli Lilly and Company, that
`may have a beneficial impact on the course of this disease.
`
`2. GLP-1 receptor
`
`The GLP-1 receptor (GLP-1R) is a seven-transmembrane
`domain, G-protein coupled receptor, initially described as
`the exendin receptor [28 – 30]. Distribution of the mamma-
`lian GLP-1R includes pancreatic periductal- and h-cells,
`kidney, heart, stomach, and brain [31]. The pancreatic
`
`Fig. 2. Theoretical overview of the primary, anti-diabetic mechanisms of action for exendin-4 (or exenatide). Data suggest that exendin-4 enhances glycemic
`control by enhancing glucose-dependent insulin secretion, suppressing glucose-dependent glucagon secretion, slowing gastric emptying, reducing food intake,
`stimulating h-cell health, and increasing the insulin sensitivity of peripheral tissues.
`
`SUN EXHIBIT 1037, IPR2024-00107, PAGE 2
`
`
`
`L.L. Nielsen et al. / Regulatory Peptides 117 (2004) 77–88
`
`79
`
`GLP-1 receptor binds exendin-4 and GLP-1 with equal
`affinity in in vitro assays, and both peptides stimulate the
`receptor equipotently as demonstrated by the production of
`cyclic adenosine monophosphate (cAMP) in human and rat-
`based receptor systems [32 – 34]. Therefore, differences in
`receptor affinity and activation cannot explain the difference
`in in vivo potency between the two peptides. Under phys-
`iological conditions, the GLP-1R recognizes GLP-1 specif-
`ically, with no significant binding to secretin, vasoactive
`intestinal peptide (VIP), or other closely related, mammalian
`peptide hormones [28,31].
`Based on studies using GLP-1R knockout (GLP-1R / )
`mice and in vitro receptor blockade in islets from these mice,
`a GLP-1 receptor appears to be involved in, and necessary
`for, the incretin/insulinotropic actions of exenatide and GLP-
`1, and the hepatic portal glucose sensor actions of GLP-1
`[3,35 – 38]. GLP-1R / mice exhibit fasting hyperglyce-
`mia and abnormal blood glucose excursions in response to
`glucose challenge [39]. However, care must be taken when
`ascribing specific physiological functions to the GLP-1
`receptor based on GLP-1R / mice, as these mice also
`exhibit abnormalities in the hypothalamic – pituitary – adrenal
`axis and alterations in the GIP (glucose-dependent insulino-
`tropic polypeptide) hormone path- ways which may com-
`pensate for the absence of the GLP-1 signal transduction
`pathway [40,41]. Both GLP-1 and GIP stimulate insulin
`secretion under conditions of hyperglycemia (incretin
`actions) in GLP-1R / mice, resulting in lowered post-
`prandial glucose excursions [3]. Truncated exendin-4(9 –
`39)-NH2 binds to and antagonizes mammalian pancreatic
`GLP-1 receptors [7] and is associated with an increased
`postprandial glucose excursion when administered to wild-
`type, male CD1 mice (5 Ag/mouse IP 20 min before oral
`glucose challenge; [35]). In contrast, administration of trun-
`cated exendin-4(9 – 39)-NH2 to GLP-1R / mice had no
`statistically significant effect on postprandial glucose excur-
`sions, although acute blood glucose levels (10 – 30-min post-
`challenge) trended lower after both oral and intraperitoneal
`glucose administration [35].
`While exendin-4 and GLP-1 appear to share certain
`glucose-lowering actions, it is apparent that not all actions
`of exendin-4 are predictable based on the known pharma-
`cology of GLP-1 [2]. For example,
`intraportal GLP-1
`infusion triggers firing of the hepatic vagal afferent nerves,
`while exendin-4 does not [42]. Exendin-4 sensitized differ-
`entiated 3T3-L1 adipocytes to insulin-dependent glucose
`uptake, while GLP-1 had no effect in the same assay [43].
`Exendin-4 may therefore exert at least some of its actions
`through a functionally different receptor, although this
`putative receptor has not yet been identified [38,42,43].
`
`3. Glycemic control
`
`Exendin-4, GLP-1, and GLP-1 analogues such as
`NN2211 have demonstrated abilities to control fasting and
`
`postprandial glucose excursions. The effects of GLP-1 on
`glycemic control in nonhuman models of diabetes have been
`reviewed elsewhere [44] and will not be extensively covered
`here, except for comparisons with exendin-4.
`Exendin-4 had potent activity in reducing plasma glucose
`when administered as a single intraperitoneal dose of 0.001
`to 10 Ag to hyperglycemic db/db mice, a model of type 2
`diabetes [11]. The maximal glucose-lowering effect of
`exendin-4 was sustained through the last time point mea-
`sured (4 h). Comparable administration of GLP-1 in the
`range of 1 – 1000 Ag also lowered glucose acutely; however,
`plasma glucose concentrations rapidly returned to their
`usual hyperglycemic levels. Exendin-4 lowered plasma
`glucose concentrations in a dose-dependent manner within
`1 h after injection, with a maximum reduction of 37%. The
`half-maximal effective dose (ED50) for exendin-4 averaged
`0.06 Ag/kg compared to 329 Ag/kg for GLP-1. In the ob/ob
`mouse model of diabetes, exendin-4 had an average ED50 of
`0.136 Ag/kg compared to 744 Ag/kg for GLP-1. Overall,
`exendin-4 was greater than 5000-fold more potent than
`GLP-1 in controlling hyperglycemia. In both strains of
`diabetic mice, the magnitude of the glucose lowering effect
`following a single dose of exendin-4 was related to the pre-
`existing plasma glucose concentration, i.e., exendin-4 activ-
`ity was glucose-dependent.
`The effects of prolonged 12- to 13-week treatment with
`once-daily intraperitoneal exendin-4 (24 nmol/kg; 100 Ag/
`kg) has been studied in both diabetic db/db and nondiabetic
`mice [20]. After 1 week, fasting blood glucose concentra-
`tions in diabetic mice had declined from 232 to 90 mg/dl in
`the exendin-4-treated group, but remained at 238 mg/dl in
`the vehicle-control group. Fasting blood glucose concen-
`trations were also lower in nondiabetic mice treated with
`exendin-4 (70 mg/dl) compared with vehicle-treated mice
`(135 mg/dl). At the completion of treatment, A1C values
`were 47% lower in exendin-4-treated diabetic mice than in
`vehicle-treated, averaging 4.7% and 8.8%, respectively.
`Nondiabetic mice also had a significant lowering of A1C
`values after exendin-4 treatment (3.1 F 0.07% vs. 3.5 F
`0.08%. in vehicle-control mice). These A1C data affirmed
`the beneficial effects of exendin-4 on long-term glycemic
`control.
`The effects of chronic intraperitoneal administration of
`exendin-4 on A1C and insulin sensitivity were further
`studied in male obese, diabetic Fatty Zucker (ZDF) rats
`[11]. These animals exhibit overt hyperglycemia starting at
`8 to 10 weeks of age, are insulin-resistant, obese, dyslipi-
`demic, and succumb to h-cell failure at 12 to 14 weeks of
`age. In a 5-week study using twice daily intraperitoneal
`injections of 100 Ag exendin-4, initial pretreatment A1C
`values were approximately 73% greater in obese ZDF rats
`(6.0 – 6.9%) than in lean non-diabetic control rats (3.6 –
`3.9%; Fig. 3). After 35 days, a significant decline in A1C of
`41.3% was seen in the ZDF rats receiving exendin-4. This
`was 2.4-fold greater than the reduction observed in exendin-
`4-treated control lean rats. Measurements of glucose uptake
`
`SUN EXHIBIT 1037, IPR2024-00107, PAGE 3
`
`
`
`80
`
`L.L. Nielsen et al. / Regulatory Peptides 117 (2004) 77–88
`
`doses of NN2211 (2 Ag/kg; 0.6 nmol/kg) or vehicle under
`conditions of hyperglycemic glucose clamp. h-Cell-reduced
`pigs treated with the GLP-1 analogue required an approx-
`imately 200% increase in the glucose infusion rate com-
`pared to the vehicle-control group, and exhibited a tem-
`porally coincident increase in plasma insulin concentrations
`(AUC 72 F 28% greater than vehicle) which was glucose-
`dependent. These data indicate an improvement in insulin-
`sensitivity after treatment with the GLP-1 analogue.
`NN2211 treatment also suppressed plasma glucagon con-
`centrations during hyperglycemic clamp, but not during
`euglycemia following termination of the clamp, demonstrat-
`ing the glucose-dependency of this anti-diabetic activity
`mediated through the GLP-1 receptor. Longer-term admin-
`istration of subcutaneous NN2211 also had effects on
`glycemia in h-cell-reduced pigs. Once-daily NN2211 (3.3
`Ag/kg) was administered for 4 weeks. Treatment with the
`GLP-1 analogue reduced postprandial glucose excursions at
`both 2 and 4 weeks, and slowed gastric emptying. Overall,
`treatment with a GLP-1 analogue in a pig model of human
`diabetes confirmed several glucoregulatory mechanisms of
`action previously observed after administration of exendin-4
`or GLP-1.
`In fasted Rhesus monkeys with type 2 diabetes, a single
`subcutaneous injection of exendin-4 caused a reduction in
`plasma glucose that was dose-dependently accelerated, with
`a mean ED50 value of 0.25 Ag/kg and a maximal glucose
`nadir of approximately 37% at a dose of 100 Ag/kg [11].
`Subsequently, plasma glucose concentrations tended to rise
`again towards control
`levels 2 to 3 h after exenatide
`injection.
`The ability of GLP-1 to control glucose excursions in
`preclinical diabetes models led to a series of GLP-1 clinical
`trials in humans, which have been summarized [44]. How-
`ever, it has become apparent that the short plasma half-life
`of GLP-1 was a significant barrier to clinical development.
`In particular, Zander et al. [46] demonstrated that subcuta-
`neous infusion of GLP-1 efficiently lowers plasma glucose
`in patients with type 2 diabetes, but must be given contin-
`uously to be effective. The resistance of exendin-4 to
`degradation by dipeptidyl peptidase-IV and other putative
`mechanisms which together result in a better pharmacoki-
`netic profile for exendin-4 offered one possible solution to
`this dilemma. A series of phase II trials of exenatide
`(synthetic exendin-4) have now been completed in greater
`than 300 subjects with type 2 diabetes (reviewed by Nielsen
`and Baron [47]). In general, a consistent pattern of safety
`and pharmacodynamics was observed. Dose-ranging studies
`identified an optimal exenatide glucose-lowering dose range
`of 0.05 – 0.2 Ag/kg when injected subcutaneously. Fineman
`et al. [48] reported notable findings from a phase II study of
`exenatide in patients with type 2 diabetes not attaining A1C
`goals V 8% with oral sulfonylureas and/or metformin.
`Twenty-eight days of exenatide treatment reduced A1C by
`approximately 0.9% compared to baseline ( p V 0.006). In
`addition, the proportion of patients achieving A1C V 7%
`
`Fig. 3. Plasma A1C in diabetic and normoglycemic rats. Long-term
`administration of exendin-4 reduced plasma A1C in diabetic, obese ZDF
`rats and normoglycemic lean littermate control rats ( p < 0.01 for obese rats
`and p < 0.002 for lean rats, respectively, compared with vehicle-treated lean
`rats). By day 35, A1C values had decreased 41% in obese ZDF rats treated
`with exendin-4. The relative reduction in A1C was 2.4-fold greater in obese
`than in lean exendin-4-treated rats, consistent with exendin-4 having a
`greater effect under hyperglycemic than normoglycemic conditions.
`Exendin-4 (100 Ag) or vehicle was injected intraperitoneally, twice daily,
`for 5 weeks. Initial A1C values were 6.44 F 0.25% in obese ZDF rats and
`3.73 F 0.37% in lean rats. Lean/Vehicle group, open triangles. Lean/
`Exendin-4 group, closed triangles. Obese/Vehicle group, open circles.
`Obese/Exendin-4 group, closed circles. N = 6 rats per group. (Adapted from
`Ref. [11]. Copyright n1999 American Diabetes Association. Reprinted
`with permission from The American Diabetes Association).
`
`into tissues under euglycemic and hyperinsulinemic con-
`ditions at the end of this study demonstrated an increase in
`insulin sensitivity (defined as plasma insulin concentration
`divided by the rate of glucose infusion) in exendin-4-treated
`obese ZDF rats (76%, p < 0.02) and lean, non-diabetic rats
`(51%, p < 0.05), compared with the saline-treated controls.
`In a second study, ZDF rats injected intraperitoneally once
`or twice daily with exendin-4 for 6 weeks had dose-
`dependent arrest or reversal of the trend for A1C to increase
`with increasing age/weight. Exendin-4 doses as low as 0.1
`Ag/rat
`twice daily were effective. Insulin sensitivity in-
`creased dose-dependently by up to 49% in the exendin-4-
`treated animals at the end of this study, as measured by
`glucose uptake under euglycemic hyperinsulinemic condi-
`tions. Whether this improvement in insulin sensitivity is a
`consequence of reduced glucose toxicity, or via direct
`actions of exendin-4 to improve glucose uptake into insu-
`lin-sensitive tissues, remains to be elucidated. Increases in
`total cholesterol levels were also significantly reduced in the
`exendin-4 treatment group over the 6-week period.
`Analagous effects on glycemic control have also been
`reported in a larger mammalian model of human diabetes,
`using the GLP-1 analogue NN2211 to activate the GLP-1
`receptor signal transduction pathway [45]. Chemical de-
`struction of h-cells in male Go¨ttingen minipigs caused the
`pigs to have impaired glucose tolerance or to be outright
`diabetic. Fasted h-cell-reduced pigs received intravenous
`
`SUN EXHIBIT 1037, IPR2024-00107, PAGE 4
`
`
`
`L.L. Nielsen et al. / Regulatory Peptides 117 (2004) 77–88
`
`81
`
`was fourfold greater after exenatide treatment compared
`with placebo. Given that A1C only fully reflects a change
`in glycemia 3 months after a sustained change has occurred,
`this reduction in A1C and enhanced ability to achieve
`clinically relevant A1C target values after only one month
`of therapy is of great clinical interest. Glucose profiles
`during ingestion of a mixed meal demonstrated that the
`marked, acute ability of exenatide to reduce postprandial
`glycemia was sustained over the 28-day observation period.
`Another recent report [49] examined the impact of 1 month
`of SC exenatide dosing on blood glucose levels in poorly
`controlled, community dwelling, insulin-naive patients with
`type 2 diabetes. Although interpretation of these study
`results is limited by the lack of a placebo control, exenatide
`treatment resulted in reduced A1C compared with pretreat-
`ment values.
`
`4. Insulin secretion
`
`Glucose-dependent insulinotropism refers to the ability
`of agents such as exendin-4 and GLP-1 to stimulate insulin
`secretion during euglycemia or hyperglycemia, but not
`during hypoglycemia [2]. Glucose-dependent insulinotrop-
`ism has also been defined as the amplification of h-cell
`insulin secretion when circulating glucose concentrations
`are above, but not below, the normal range [8,9]. In animal
`models of diabetes, a predominant acute action of exendin-
`4 is glucose-dependent
`insulinotropism, resulting in an
`amplification of glucose/insulin secretion coupling. This
`action of exendin-4 contrasts with the action of non-
`glucose-dependent insulin secretagogues or hypoglycemic
`agents, such as sulfonylureas, which predominantly in-
`crease insulin secretion independent of prevailing glucose
`concentrations [50] and thus have a greater potential to
`induce hypoglycemia [27].
`Two in vitro models using pancreatic islets isolated from
`male Lewis rats have been used to examine the action of
`exendin-4 on insulin secretion [9]. In a static incubation
`model, islets were incubated in a medium containing either
`3 mmol/l (basal) or 10 mmol/l (elevated) glucose. Raising
`the glucose concentration from 3 to 10 mmol/l increased
`insulin secretion 9.8-fold ( p < 0.05). In the presence of
`elevated glucose (10 mmol/l) and a range of exendin-4
`doses (1 nmol/l – 1 Amol/l), insulin secretion was increased
`over basal levels by up to 19.6-fold ( p < 0.01). In a micro-
`physiometer model, increasing the glucose concentration
`from 3 mmol/l (basal) to 7.5 mmol/l for 15 min increased
`insulin secretion up to 6.4-fold. The addition of exendin-4
`(20 nmol/l) stimulated insulin secretion over and above the
`effect of 7.5 mmol/l glucose alone, to 13.5-fold over basal
`rate ( p < 0.01). Insulin secretion returned to previous levels
`within 5 min of cessation of exendin-4 perfusion.
`Thus, exendin-4 exerts direct effects on rat pancreatic
`islets in vitro to enhance glucose-stimulated insulin secre-
`tion. These effects do not appear to persist once exendin-4 is
`
`removed from the system, suggesting that the peptide may
`be binding and activating target receptors only while it is
`present
`in the perfusate. Furthermore,
`the insulinotropic
`action of exendin-4 in this system rapidly decreases when
`the ambient glucose is decreased back to 3 mmol/l, consis-
`tent with the glucose-dependence of insulinotropism being
`at least in part at the level of the pancreatic islet. Overall, the
`data support the interpretation that exendin-4 acts directly,
`but perhaps not exclusively, at the isolated pancreatic islet
`level to stimulate insulin secretion, and are consistent with
`published reports showing enhanced insulin secretion from
`islets exposed to GLP-1 [33,51 – 53].
`Intravenous administration of exendin-4 (0.4 nmol/kg)
`acutely increased plasma insulin concentrations in fasted
`Wistar rats, eliciting approximately double the insulin peak
`elicited by the same dose of GLP-1. The ED50 value for this
`insulinotropic activity was 0.014 nmol/kg for exendin-4
`compared with 0.19 nmol/kg for GLP-1 [20]. Similar results
`have been reported by Parkes et al. [9] examining the
`insulinotropic actions of exendin-4 during an intravenous
`glucose challenge (Fig. 4).
`Insulinotropic actions have also been reported for exen-
`din-4 in studies examining daily administration of the
`peptide (24 nmol/kg) to diabetic db/db mice for 12 – 13
`weeks [20]. Fasting plasma insulin concentrations in dia-
`betic mice treated with exendin-4 averaged 550% higher
`than vehicle-treated diabetic rats under conditions where the
`corresponding plasma glucose concentrations averaged 16
`and 29 mmol/l, respectively, compared to 7 – 8 mmol/l blood
`glucose in nondiabetic (young) db/db mice.
`In the human studies reported by Kolterman et al. [10],
`subcutaneous exenatide (synthetic exendin-4) rapidly low-
`ered both fasting and postprandial plasma glucose in
`patients with type 2 diabetes. The glucose-dependent insu-
`linotropism exhibited by exenatide was best illustrated by
`the data obtained in the fasting state, where there was a
`dose-dependent rise in serum insulin concentrations within
`the first 3 h after exenatide administration compared to
`placebo ( p < 0.001). In sharp contrast, placebo treatment
`resulted in relatively stable insulin concentrations through-
`out the 8-h period of observation. The rise and peak of
`serum insulin concentrations following exenatide adminis-
`tration coincided with the rapid decline of fasting glucose
`concentrations. After 3- to 4-h postdose, and coincident
`with reaching glucose nadir, mean serum insulin returned
`to baseline with little difference among groups. Insulin
`AUC(0 – 8 h) and Cmax values for all exenatide treatments
`increased in an apparently dose-dependent manner com-
`pared to placebo ( p < 0.05). Since exenatide concentrations
`remained elevated throughout the course of the assessment,
`the reduction in insulin beyond 3 h did not reflect a simple
`waning of exenatide effect due to lower circulating concen-
`trations of exenatide.
`Fineman et al. [48] used the homeostasis model assess-
`ment (HOMA) [54] methodology to assess h-cell function
`at baseline and at Days 14 and 28 in human subjects with
`
`SUN EXHIBIT 1037, IPR2024-00107, PAGE 5
`
`
`
`82
`
`L.L. Nielsen et al. / Regulatory Peptides 117 (2004) 77–88
`
`in fasting and postprandial glucagon levels in patients with
`type 2 diabetes [55] and the known activity of glucagon
`with respect to maintaining hepatic glucose output [56], it is
`reasonable to extrapolate that glucagon suppression by
`exenatide contributed to the overall effect of lowered plasma
`glucose in both the fasting and postprandial periods. Like
`insulin, glucagon concentrations also returned towards
`baseline beyond 3-h post-injection during the fasting state
`when the prevailing plasma glucose concentration reached
`its nadir. As the changes in concentrations of insulin and
`glucagon are coincident and reciprocal,
`the effects of
`exenatide on the enhancement of insulin secretion and the
`suppression of glucagon secretion may be considered glu-
`cose-dependent. It is noteworthy that currently available
`insulin secretagogues, as well as exogenous insulin, do not
`suppress the paradoxical postprandial glucagon rise ob-
`served in patients with diabetes [57]. This inappropriate
`elevation in postprandial plasma glucagon concentrations
`results in an inappropriately low insulin-to-glucagon ratio in
`the hepatic portal vein, contributing to sustained rates of
`excess hepatic glucose production [58]. Thus, by virtue of
`its ability to enhance endogenous insulin and lower gluca-
`gon secretion, exenatide would tend to re-establish a more
`physiologic and favorable portal vein ratio of insulin-to-
`glucagon compared to currently available agents.
`
`6. Enhanced B-cell mass
`
`One of the most provocative areas of recent research is
`based on observations that exendin-4 may improve h-cell
`health, thereby impeding or even reversing disease progres-
`sion. A number of published reports support the concept that
`exendin-4 and GLP-1 increase h-cell mass via stimulation
`of h-cell neogenesis, stimulation of h-cell proliferation, and
`suppression of h-cell apoptosis (Table 1).
`The development of the endocrine pancreas is under
`multifactorial control, and many of the key proteins or
`transcription factors involved in each developmental step
`have now been identified [59]. Expression of the transcrip-
`tion factor PDX-1/IDX-1 is essential for development and
`regeneration of the endocrine pancreas, and stimulation of
`PDX-1 may be one of the intracellular mechanisms by
`which exendin-4 and GLP-1 can enhance h-cell mass
`leading to long-term improvement
`in glucose control.
`Exendin-4 and GLP-1 have been proposed to produce
`differentiation of pancreatic non-h-cells including pancreat-
`ic ductal epithelial cells, acinar cells, and nestin-positive
`peri-ductal cells (potentially ‘‘functional’’ islet stem cells),
`into immunohistologically identifiable insulin-producing h-
`cell buds and subsequently, fully functional islets. In vitro
`studies in human and animal cell preparations support this
`hypothesis, and give insight into the molecular pathways
`mediating these changes.
`Multipotential progenitor cells in the pancreatic islets and
`ducts, so-called nestin-positive islet-derived progenitor cells
`
`Fig. 4. Insulinotropic actions of exendin-4 and GLP-1: Concentration
`response. Male normoglycemic Harlan – Sprague – Dawley rats were
`anesthetized, then given a 2-h intravenous infusion of exendin-4 (3, 30,
`300, or 3000 pmol/kg/min; n = 4 – 8 per dose), or GLP-1 (3, 30, 300, or
`3000 pmol/kg/min; n = 4 – 8 per dose), or vehicle (1 ml/h, n = 8) at t = 30
`min. At t = 0 min, D-glucose (5.7 mmol/kg) was infused at a rate of 0.5 ml/
`min over 1 – 2 min. This infusion regimen resulted in identical plasma
`glucose pharmacokinetic profiles in all treatment groups (see Ref. [9]).
`Plasma insulin was measured at 30, 15, 0, 5, 10, 15, 20, 30, 40, 50, 60,
`and 90 min. Peak insulinotropic effects of exendin-4 and GLP-1 occurred at
`a similar steady-state plasma concentration of approximately 1 to 2 nmol/l;
`thus, insulinotropic potencies were similar. However, the magnitude of the
`maximal effect was 63% greater with exendin-4 than with GLP-1
`( p < 0.03). Exendin-4, black circles. GLP-1, white rectangles. (Reprinted
`from Ref. [9]).
`
`type 2 diabetes receiving two or three daily subcutaneous
`injections of exenatide (0.08 Ag/kg). HOMA analysis
`revealed improved h-cell secretory function following exe-
`natide therapy. It is noteworthy that fasting values of plasma
`glucose and insulin, which were used to calculate HOMA,
`were obtained prior to the morning dose of exenatide when
`plasma concentrations of exenatide were negligible, sug-
`gesting a fundamental alteration in h-cell function following
`sustained exenatide exposure. Whether or not exenatide
`treatment increases pancreatic islet mass in humans awaits
`future studies, since h-cell function can improve indepen-
`dently of increased islet mass.
`
`5. Glucagon secretion
`
`Circulating glucagon was reduced after exenatide treat-
`ment in both the fasting and postprandial states in humans
`with type 2 diabetes [10]. The observation under fasting
`conditions supports the hypothesis that suppression of
`glucagon secretion is not merely a consequence of the
`slowing of nutrient presentation to the small
`intestine
`(gastric emptying). Given the well-documented elevations
`
`SUN EXHIBIT 1037, IPR2024-00107, PAGE 6
`
`
`
`L.L. Nielsen et al. / Regulatory Peptides 117 (2004) 77–88
`
`83
`
`Table 1
`Effects on pancreatic h-cells
`
`Model
`
`Agent
`
`Key findings
`
`Reference
`
`Nestin-positive islet-derived
`progenitor cells
`AR42J pancreatic acinar
`cell line
`Partially pancreatectomized
`rats
`Streptozotocin-treated
`Wistar rat
`Goto – Kakizaki rat
`
`Intrauterine growth-retarded
`neonatal Sprague –
`Dawley rats
`C57Bl/6 mouse, diabetic
`db/db mouse
`
`Streptozotocin-treated
`C57Bl/6 mouse,
`BHK-GLP-1-R cells
`Diabetic db/db mouse
`Diabetic db/db mouse
`
`Exendin-4, GLP-1
`
`Exendin-4, GLP-1
`
`Exendin-4
`
`Exendin-4, GLP-1
`
`Exendin-4, GLP-1
`
`Exendin-4
`
`Exendin-4, GLP-1
`
`Exendin-4
`
`Exendin-4, tNN2211
`Exendin-4
`
`Stimulated differentiation of progenitor cells into
`insulin-producing cells.
`Stimulated differentiation into cells expressing insulin,
`pancreatic polypeptide, and glucagon.
`Attenuated the development of diabetes; increased h-cell
`mass via differentiation (neogenesis) and replication.
`Increased pancreatic insulin content, h-cell mass, and
`h-cell proliferation; decreased plasma glucose.
`Increased pancreatic insulin content and h-cell mass;
`decreased plasma glucose.
`Increased h-cell mass and proliferation; increased
`Pdx-1 expression; decreased plasma glucose.
`
`Increased h-cell mass, h-cell neogenesis, and Pdx-1
`expression; increased insulin secretion; decreased plasma
`glucose; lower A1C; improved oral glucose tolerance.
`Increased h-cell mass; decreased h-cell apoptosis;
`increased plasma insulin; decreased plasma glucose;
`improved oral glucose tolerance; decreased cell apoptosis in vitro.
`Increased h-cell mass and proliferation; decreased plasma glucose.
`Increased pancreatic insulin content and h-cell mass; decreased
`h-cell apoptosis; improved hyperinsulinemia; enhanced glucose-
`stimulated insulin secretion; improved oral glucose tolerance;
`decreased plasma glucose.
`
`[60]
`
`[61]
`
`[17]
`
`[18]
`
`[19]
`
`[66]
`
`[62]
`
`[63]
`
`[64]
`[65]
`
`(NIPs), have been previously identified by Habener et al. In
`a recent study, they now report the expression of functional
`GLP-1 receptors on these progenitor cells, and show that
`GLP-1 stimulates the differentiation of NIPs into insulin-
`producing cells [60]. Co-incubation of these cells